·No text on this page.

·No text on this page.

·Publuhed by the
JOSIAH MACY, JR FOUNDATION
565 Park Avenue, New York 21, N Y
1950
Price $1 60
Printed in the United Stales of America
B) Corlies Macy & Co, New York

·No text on this page.

·No text on this page.

·PARTICIPANTS Erghih Conference on Liver Injury DR CHARLES H BEST Charman Depatment of Phys ology Un vers ty of Toronto Toronto Canada DR WALLACE D ARMSTRONG Department of Plys olog cal Chem stry Un vers ty of Minnesota Med cal School Minneapol s Minnesota DR JESSE L BOLLMAN Department of Phys ology Mayo Foundation Rochester Minnesota DR FRANK FREMONT SMITH Jos sh Macy J Foundat on New York Caty New York DR HARRY GOLDBLATT Iost tute for Med cal Resea ch Cedars of Lebanon Hosp tal Los Aogeles Cal forn a DR PAUL GYÖRGY Depatment of Cincal Ped att cs Uo vesty of Pennsylvania School of Medcne Ph ladelph a Pennsylvan 2 DR FRANKLIN M HANGER Department of Medc de Columb a Un vesty College of Phys cans & Surgeons New Yok New Yok DR WS HARTROFT Banting & Bet Department of Med cal Research Un versty of Toronto To onto Canada DR FW HOOFBAUER Depa men of Med c ne va ser ty of M nnesota Med cal School M nneapols Minnesota DR DWIGHT J INGLE Research Department, The Upjohn Company Kalamazoo M ch gan DR MELVIN H KNISELY Department of Anatomy The Med cal College of the Sate of South Carol na Cha leston South Ca olioa

·No text on this page.

·PARTICIPANTS Eighth Conference on Ener Injury DR CHARLES H BEST, Charman Department of Physiology, University of Toronto Toronto, Canada DR WALLACE D ARMSTRONG Department of Physiological Chemistry Civersity of Minnesota Medical School Minneapolis, Afinnesota DR JESSE L BOLLMAN Department of Physiology, Mayo Fouadation Rochester, Minnesota DR. FRANK FREMONT SMITH Josiah Macy Jr Foundation New York City New York DR HARRY GOLDBLATT Institute for Medical Research, Cedars of Lebanon Hospital Los Angeles, California DR PAUL GYÖRGY Depa tment of Clinical Pediatrics University of Pennsylvania School of Medicine Philadelphia Pennsylvania DR FRANKLIN M HANGER Department of Medicine Columbia University, Caitege of Physicians & Surgeons New York New York DR WS HARTROFT Bunting & Best Department of Medical Research University of Toronto Toronto Canada DR FW HOFFBAUER Department of Medicine Upiersity of Minnesota Medical School Minneapolis Minnesota DR DWIGHT J INGLE Research Department, The Upjobs Company Kalamazoo Michigan DR MELVIN H KNISELY Department of Anatomy The Medical College of the State of South Carolina

·DR SIDNEY C MADDEN
Division of Pathology, Brookhaven National Laboratory, Associated Universities, Inc
Upton, LI, New York
DR J. MARKOWITZ
Department of Physiology University of Toronto
Toronto, Canada
DR JOHN R NEEFE
Department of Medicine, University of Peansylvania Hospital
Philadelphia, Pennsylvania
DR ARTHUR J PATEK, JR
Department of Medicine, Columbia University, College of Physicians & Surgeons
New York, New York
DR KLAUS SCHWARZ
Experimental Biology & Medicine, The National Institutes of Health
Bethesda 14, Maryland
DR EPHRAIM SHORR
Department of Medicine, Cornell University Medical College
New York, New York
DR DEWITT STETTEN, JR
Public Health Institute of New York
New York, New York
DR CECIL J WATSON
Department of Medicine, University of Minnesota Medical School
Minneapolis, Minnesota

·TABLE OF CONTENTS Erghih Conference On Liter Injury SECTION I The Liver Lobule Melem H Knisely Discussion 9 13 SECTION 11 "The Function of the Hepatic Artery | Alarkou stz 18 SECTION IN Massive Liver Necrosis (a) Paul Gyorgy (b) Klaus Schuarz Discussion 34 45 55 SECTION IV Recent Findings Concerning the Role of the Liver and Kidney in Circulatory Homeostatis Epiraim Shorr SECTION V Metabolic Behavior of the Eviscerate Rat Duight ] Ingle Discussion 86 107 SECTION VI Liver Glycogen and the Protection of Liver Cells | L Bollman 115 SECTION VII Locus of the Beginning of Dietary Cirrhosis Stanley Hartroft 126 Discussion 155

·No text on this page.

·SECTION 1 THE LIVER LOBULE Best I need not say that we conduct these conferences in a very informal manner The speakers may be interrupted for questions at any time We will start with Dr Knisely, who will discuss certain aspects of the liver lobule Knisely The purpose of this paper is to present a concise view of the selective removal of coated particles from the blood stream It is necessary to understand this subject in order to understand certain aspects of those conditions and diseases in which the blood of the patient is precipitated and agglutinated into a circulating sludge To present this subject it is necessary to describe first of all the structure and mechanical functioning of the liver lobule of frogs, and show how the phagocytic von Kupffer cells which line the sinusoids of the lobule selectively remove foreign particles from the blood stream Methods Living internal organs have been transılluminated in situ for microscopic study by means of fused quartz rods Motion pictures of the circulation in many frog organs have been taken through the microscope, thereby recording some of the observations made in mice, rats, guinea pigs, and kittens circulatory conditions in small vessels of pia arachnoid, smooth muscles of stomach and intestinal walls, and of uteri, spleens, and livers have been studied Problems and results. The structure and mechanical functioning of the living frog liver lobule has been studied with three purposes in mind (1) to find out how the hepatic artery and portal vein terminals are distributed to and within the lobule and how their activities may effect or control the chemical environment of the hepatıc parenchyma cells, (2) to find out how the lobule stores and releases blood, thereby controlling cuculating blood volume and cardiac output, (3) to find out how one phagocytic von Kupffer cell selectively removes one foreign particle from the circulating blood How does the phagocyte select what it takes? Why does it ignore what it does not tale? What factors determine the rates at which particles are brought to these stationary phagocytes? The hepatic artenole runs along the interlobular portal venule like a vine on a tree Both join the sinusoids Every sinusoid is connected to a terminal portal venule Every sinusoid, or almost every one, te ceives a branch the arterial sinus twig, from the hepatic artery Contractile arterio portal anastomoses interconnect the hepatic artery 9 ·and the portal venule like rungs of a ladder The sinusoids have a complete continuous cellular lining Each cell of this lioing is a phagocytic von Kupffer cell in the living frog liver the von Kupffer cell is not a star shaped cell suspended by processes like a spider in a tube Each sinusoid has an afferent inlet sphincter guarding its junction with the portal vein Each sinusoid has an efferent outlet sphincter guarding its junction with the central vein The living lobules also have the small sluice channels having one common outlet to the central vein or to the sublobular vein The degree of dilation or tonic contraction of the hepatic artery, arterio portal anastomoses, and afferent interlobular portal venule determines the relative composition of the blood which is supplied to the sinusoids, that is, of the blood which nourishes the hepatic parenchyma cells When the interlobular hepatic arteries shut off tightly throughout their lengths and the portal vein remains dilated, the sinusoids receive only pure portal vein blood When the terminal interlobular portal venules are tightly constricted and the hepatic arterioles dilated, the sinusoids receive only pure arterial blood When the hepatic artery is dilated and the arterial sinus twigs are shut off and the arterio-portal anastomoses are open, arterial blood passes into the terminals of the portal vein and is distributed by way of the portal vein tips to every sinusoid whose inlet sphincter is open With the arterio portal anastomoses closed and the portal vein terminal and hepatic artery terminals partly open the peripheral ends of the sinus oids receive a mixture of portal vein and arterial blood The kiod of blood, or the ratio of the mixture, and the volume of blood supplied to a given set of sinusoids per unit of time, is continually controlled and may be maintained constant for hours at a time These observations demonstrate beyond reasonable doubt, that th chemical environment of the hepatic parenchyma cells of each lobule is precisely controlled continuously and that it is forced to vary within controlled limits The tubular sinusoid lining membrane which is contractile through out its length and each cell of which is a phagocytic von Kupffer cell , has three distinct permeability phases In one the red cells enter the sinusoid separated by small volumes of plasma and the distance betteen the red cells does not become visibly less as the column of cells passes along the sinusoids During another extremely permeable phase the sinusoid lining membrane is probably permeable to all the colloids of the blood As the red cells pass along the sinusoids they come closer and closer together until the central two thưds, half, or third of the sinusoid contains only packed red cells moving sometimes slowly sometimes rapidly The sinusoids also exhibit an intermediate 10 ·permeability phase during which the sinusoid linings probably are permeable to proteins of small molecular seight such as albumea, but probably are not permeable to larger ones such as globulus or fibrinogen It is important to note that the experimental conditions do not cause the sinusoid linings to go into os stay in any one of their three perrocability phases The outlet sphincters of the individual sinusoids and those of Deysach s small sluice channels are the sluice mechanisms of the living frog liver Lach of these sphincters can dilate widely, contract tightly, shut, or maintain any intermediate degree of tonic contraction They control the volume of blood stored in the liver and backed up in the portal vein tributaries, and thereby control the circulating blood volume and cardiac output Thus far the outlet sphincters are the only contractile structures thich we have seen cause the liver to store or release blood (The anatomical distribution and mechanical behavior of the parts of the frog liver lobule s vascular system, outlined above, were demonstrated by motion pictures taken through the microscope ) Foreign particles suspended in the circulating blood are carried into the liver sinusoids at whatever rate blood is flowing into the sinusoids Hence the rate of supply of particles to hepatic phagocytes is faster when the inlet and outlet sphincters are open than when the outlet sphincters are closed during periods when the sinusoids are in blood storage phases, and the fastest flow rates occur when the hepatic astery is widely dilated sending blood in jetting pulses into, through and out of the sinusoids The phagocytic von Kupffer cells do not ingest every particle which comes into the hepatic sinusoids What factors determine the selec tivity of the selective phagocytic removal of particles from the blood stream? Why do the von Kupffer cells ingest some particles and miss or ignore others > When fine droplets of undiluted Higgins Indra ink are injected into the blood stream, each droplet immediately receives a visible coating of a clear glassy precipitate probably protein, derived from the frog's plasma Injected particles of Laolin graphite, and mercuric sulfide also receive protein coatings The coatings often, but not always, have long feathery fibrous streamers of newly formed glassy precipitate attached The coating which entirely surrounds the ink changes the droplet into a coated semisolid particle When a coated particle with a streamer ss cardied into the branching and anastomusing hepatic sinusoid system the tail of the streamer may whip around a sinusoid bifurcation or may drag along the sinusoid wall The moment this happens the streamer sticks to the sinusoid lining the tethered particle is then forced against the sinusoid wall by the flowing blood, and 11 ·the streamer and clear glassy coating, with its contained ink droplet, passes instantly into the cytoplasm of whatever sinusoid lining cell or cells they touch The coated particles appear to enter the cytoplasm as quickly as two mercury droplets which are rolled against each other merge into one The von Kupffer cell may then swell a little and, depending upon the pressure on each side of it, it may bulge a little either into the sinusoid lumen, or outward indenting the adjacent hepatic parenchyma cell cord A coated particle which has no streainers may pass some distance along the sinusoid system before striking the will hard enough to make firm contact with the lining But at what ever point it bumps or rubs against the linıng, it passes instantly into the sinusoid lining von Kupffer cell cytoplasm These phagocytes have not been seen to put out pseudopods Their ingestion processes are much too fast for that, and, thus far, appear rather to have the characteristics of simple strong surface tension phenomena Massive doses of heparin given frogs before injection of ink, kaolin, or graphite prevented the formation of coatings on particles of these materials, and the naked particles bumped, rolled, and slid along the inner surfaces of the sinusoid lining phagocytes without being ingested Normal frog red cells have no coatings and are not ingested From these experiments it seems necessary to conclude that in the frogs 1 Some kinds of particles, including red cells, will not be ingested by hepatic phagocytes unless they are coated 2 The hepatic phagocytes will ingest these coatings regardless of what is in them 3 The presence of the coating determines the selectivity of the selective phagocytic removal of particles from the blood stream by hepatic phagocytes 4 The von Kupffer cells can selectively ingest coated particles from blood which is flowing rapidly, that is, the old idea that blood must be stationary or flowing slowly to permit phagocytosis is not true, at least under the conditions of these experiments The number of particles which can be removed from blood per hour or day depends upon the rates at which particles suspended in blood are brought into hepatic sinusoids, and the coating which causes the phagocyte to select what it ingests, plus one other, the number of phagocytes which at any one time are ready to ingest particles We can say, from these experiments, that the rates at which par ticles suspended in the blood are carried into hepatic sinusoids de termine the maximum rates at which particles can be brought to 12 ·hepatic phagocytes, and this tate together with the selective surface factors which determine that a particle is phagocytizable and the factors of chance which cause a coated particle to touch a phagocyte determine the rates at which particles are removed from the blood by those phagocytes which are at that time capable of ingesting particles EDITOR The above abstract is published in detail in the following reference ? JKI1lel, w 4 L 1, 1948 A brief account of this work also appears in the following reference Kosely MH Bloch, E Hand Warner, L A Preliminary Account of the Structure and Mechanical Functioning of Living Frog Liver Lobules, and of Microscopic Observations of the Selective Phagocybe Removal of Coated Particles from Flowing Blood by the Sinusoid lunung von Kupffer cells Conference on Liver Injury, Transactions of the Fourth Meeting Josiah Macy, Je Foundation, New York, October 28 26, 1945, p 21-41 DISCUSSION Best Dr Krisely, I should like to ask a question celating to the arterio-Portal anastomoses you demonstrated Has anybody ever collected a sample of portal blood and demonstrated that it contamed more oxygen than the portal blood entering the liver? Knisely I don't know I doubt if anybody has trsed that Best it would be rather interesting I was just wondering how much oxygen those anastomoses really could contribute How much increase of portal blood oxygen could you get from that anastomosis? Knisely I am only talking about anatomy, the aspect of rate of Aow must be considered The portal vein gets shut off completely and there are torrents of arterial blood, or nothing at all if one wants to go back to the environment, as far as the supply of blood through the sinusoids goes, it is pure arterial or the artery can shut off to give pure portal Row There is no difficulty in controlling the anastomoses and twigs Here the ratio of mixture is controlled rather precisely for long periods of time Neefe For how large a portion of the lives may that appear? Knisely I cannot tell you that One is able to visualize only two or three at a time and these are all down at the edge of the liver Thus method permits you to see precisely what happens A method that gives a broader perspective is that described in the paper by S Soskin, H E Essex) F Herrick and F C Mann (Am ; Physiol 124, 558 (1938)) in which they used thermostromuhrs to measure the rate of flon Sometimes one finds as much as 90 percent of the blood 13 ·LE 1 will be by way of the artery and sometimes 90 percent by way of the postal vein Shorr Engel, Harrison and Long () Exper Med 79, 48 (1944)) have shown that the rat liver is very susceptible to anoxia in tivo, even short periods of oxygen lack (more than 15 minutes) leading to significant reductions in oxygen consumption But in the eviscerated rat with the hepatic artery left as the only blood supply to the liver, the oxidative needs are apparently entirely satisfied, for there is no reduction in the oxygen consumptions of such livers after as long as 4 hours, suggesting that no hypoxia of significance results under these conditions Best We must get Dr Bollman s opinion as to how well the studies with the thermostromuhr stand up in view of everything we now know Knisely You are questioning the thermostromuhr a little bit? Best Quite a lot, if you like Bollman There is no question but the thermostromuhr has its limitations as well as its tribulations However, when it is properly fitted and properly checked and rechecked, accurate readings may be obtained The blood vessel being studied is not cut, but it is encircled however, loosely, and a wide range of blood flow may be determined in a given blood vessel The main point that has been demonstrated concerning the total blood flow to the liver is that it may be greatly increased or decreased under many physiologic or pathologic condi tions, similar in this respect to the other organs of the body In addition both the arterial and portal blood flow may be altered to a degree independent of each other. Thus the total blood flow to the liver may be altered and the percentage supplied by the arterial and portal components differs greatly at different times, especially as influenced by the stage of digestion and exercise The overall average for the day is certainly an unknown quantity, but I would like to emphasize the extreme flexibility, both as to the total blood flow and to its component parts Knisely There is a piece of history here which may be worth reciting Some of the first papers that came out on the percentage of blood flow in the liver were by R Burton Opitz (Quart ) Exper Physiol 3, 297 (1910), 4, 93 103, 113 (1911), 3, 83 (1912) Many of the textbooks for medical students state that from a to a third of the blood entering the liver is by way the rest is by the portal vein, these data were taken fr papers He used Ludwig's stromuhr Let job for his day and age He measured the artery in one dog or cat, and then sh some 14 ·deasured the portal vein and added the two together Notice the allacy in adding together those two evidences and making a statement from the portal flow in one animal and the arterial How from another, because it is impossible to do both at the same time The assumption was made that that happened all the time it was an assumption that physiology was a constant thing over a long period Even though the stromuht figures may be doubted at the moment, I think we are ready to employ more delicate instruments if we can devise them We need to know the ratio of these two flows under many definable circumstances rather than with two animals Best Indeed we do CA Crandall, in collaboration with A Lipscomb, (Am J Physiol 148, 312 (1947)) has used the London can nula technic and his conclusions are quite at variance with those derived from the thermostromuhr work, so it is really, and I think Dr Bollman will agree, pretty much an open subject But regarding your present blood flow measurements, I suppose they are reasonably satisfactory Bollman The blood flows at any moment may be accurately determined and I would be satisfied with the results of the thermo stromuhrs in situ were again checked by mechanically pumping the same volumes of blood through the vessels I think there is a more important feature which Di Knisely will bring out I have heard him mention it before that the overall picture of blood Aow is the summation of blood flow to the individual lobules Some sinusoids may be filled and retain blood while others shunt blood rapidly into the venous system Some lobules receive practically 100 percent portal blood and many receive different mixtures of the two The overall average of the blood flow to and from the liver is the sum of all these variables the individual components of which seem to be in continuous variation There are probably but few points within the liver lobule where the blood flow is constant and continuous with blood of uniform composition Fremont South Isnt that highly pertinent to the many functions of the liver and the hundreds of chemical reactions it performs? Bollman Five hundred Knisely That was a guess, there are so many that all you can do is worry for the moment about some of them Fremont Smith This variability in the same part at different time as well as in different parts of the liver is a very important facto and more important than the overall in terms of being able to unde stand what is going on Knisely Highly controlled things' One of the most precisely cc trolled things I have ever run across in the living body!

·Best From another point of view, just to make an interesting controversy, what we want to know when we are studying sugar formation, is not really what happens in the individual lobules and the variation between them but the overall picture Fremont Smith Maybe the whole liver is not participating, so maybe you want to know how much of the liver is doing this and at what rate Best We all agree that we would like to know what each individual part is doing and also what the whole thing is doing Markou itz May I interject a note? The portal vein derives its blood from three huge arteries, the superior mesenteric, the celiac axis and the inferior mesenteric We have to postulate for a moment that the velocity of blood flow in the arteries is the same as in one small branch, the hepatic artery Even if you did not measure directly the blood flow in the hepatic vein, you would guess that the hepatic vein would have four or five times as much blood as the hepatic artery because otherwise where is the blood going? We must not forget that If we are inclined to think the hepatic artery carries more than the portal vein, what is happening to the blood pouring into the portal vein? Knisely It is sometimes stored for relatively long periods Markowitz If you tied it, there is evidence of considerable blood flow which is very acute in a minute Knisely That is true Fremont Smith Is there a possibility of variations in the arterial supply to the portal vein which might mean that if there is a con striction of those arteries from which the portal vein blood is derived you might have quite marked variations in the amount of blood entering the portal vein? Is that tenable? Alarkowitz Anyone who has done as I have done, in attempting to anastomose the splenic artery to the splenic vein recognizes how true that is in the dog you start with a splenic artery that has a diameter of a agarette and end up with something like the end of a match The operation is practically impossible "There must be a terrific variability in the caliber Best We have also a large capillary bed in the intestine Gyorgy And in the spleen Best We have here the possibility of storage apart from storage in the liver Fremont Smith May I go back to something earlier? It seems to me whether one is concerned with the overall picture or a particular 16 ·aspect, depends upon what you are interested in, and therefore there aeed be no controversy Obviously, as was pointed out, we need to know both in some situations we have got to know the overall, in others we have got to know the particular Best Dr Krisely, of course, cannot give us the picture of all the lobules simultaneously If so that would be the thing Kosely I would give up instantly on that Best So you want to know the representative picture of the variations We also want to know what is coming out of the hepatic veins from the overall metabolic point of view AVIOUS pous LOURNIS Luldntly, dild, I crte ler as in the kidney, the overall picture as well as the particular aspect may become Crucially important with respect to particular problemas Shorr What 15 suggested by De Knisely s description is that the liver has the appropriate vascular mechanism for purposefully altering the gaseous environment for homeostatic purposes There is nothing to say that only aerobic metabolic products and processes are of value to the organista Indeed the hepatic vasodepressor VDM (ferritin) with which my laboratory is concerned is formed by normal liver under anaerobic and activated under accobic conditions It is reassur ing to see that the vascular pattern in the liver is such as to permit the necessary shifts in oxygen tension which would be required by this metabolic pattern ·SECTION II THE FUNCTION OF THE HEPATIC ARTERY Best I will ask Dr Marlowitz to proceed with his presentation on the function of the hepatic artery in the dog Alarkou ?tz When I came back to Toronto from Georgetown Unnersity in 1932 and got a position with Dr Best, I remember tell IS him that what I was doing as studring the hepatic artery and its function I wanted to see if one could fierent hepatic necrosis by atroducing artenal blood into the portal blood and then tying the hepatic arteries I have been doing thus, with interruptions oning to the war for 15 years. In 12 or 13 rears I thought I had a few cases in which I tied the hepatic artery, arterialized the portal vein in one of a number of ways and the animal lied When I was on the point of repeating the crucal experiment, I alnas failed to confirm the finding One year I thought I could substitute for the hepatic artery by putting arterial blood into the portal rein and another year I could not I tried to arterialıze the portal vein in 12010us wars 1) One nay nhich was more successful than any was to do an anasto mosis bet een the aorta and the portal vein directly side by side using the method of Ick s tistula That has a high mortality First the portal vein has to be pushed a long way dwn and the two stay sutures tear out More usually the mortality is due to the fact that the portal sein becomes scutely linked and heause of the high pressure of arterial blood there is on such intraportul pressure to obstruct badly the blood flow from the bowel which becomes blue It is quste a thing to see the liver mos rod and the bonel blue 2) Another method is an arterialize the blood in the sena cava by doing a side to side in istimosis of the vena cava and aorta That is easy to do That blwod now being arterialized we would do a reverse Ich fistula Thit blot in the icna (113 having been shunted into the lover no would tie the hepatic artery o) One other way nonld be to pain the splenic artery to the splenic vein Thuat luaj's fue bause of thrombosis 4) Attempts to an istomose the renal artery to renal vein failed similarly In 1918 I pud i brick visit to N110 Clinic. There I SIN Grindlay Bollman and their as Transfie: tubing in their ·studies on lymph This svas a technic I thought I could use to join the splenic artery to the splenic vein, using the Transflex tubing, having first coated the tube with silicone Since you cannot be sure the technic is really aseptic, I thought we would gre the animals some penıcıllın So, as we planned, we used siliconed Transflex tubing and gave the animal penicillin to make sure that the animal would not succumb to infection in every instance we were faced with the result that although the hepatic artery was effectively tied the animal stayed alive However, when we came to examine the tubing, it was always thrombosed After debating this backward and forward, we could do nothing but repeat the experiments doing a control for the penallın We tied the hepatic artery in a number of animals, all the animals who failed to receive penicillin died All the animals who got penicillin effectively survived, except one which died of bile peritonitis Fremont Smith With the hepatic artery tied those that were given penıcıllın survived? Markou atz. Yes, that is something very unusual I think you might be interested to hear something of the literature I have a fuller account of it in my book (J Markowitz, Experimental Surgery, Williams and Wilkins Co, Baltimore, 1949) SB Wolbach and T Saiks (J Med Res 21, 267 (1909)) gave the whole point a new twist when they stated that 21 out of the 23 healthy dogs they examined, had livers containing spore bearing anaerobes They used extreme methods of insuring asepsis Ellis and Dragsted (ated from J Markowitz, A Rappaport and A T Scott, Am J Digestive Dis (1949) in press) stated that the liver normally housed spore bearing anaerobes which they thought were Welch s bacıllı Boyer and McFedridge (ated from Markowitz Rappaport and Scott, foc ait) later tested these findings and decided that the cause of death from liver autolysis was the liberation of toxic material from the enzymic post mortem degradation and was not due to bacterial decomposition There is nothing in the anatomy of the hepatic artery, as shown by Knisely s studies to indicate just why animals should die following its ligation A Nareth (Deuts Zeitschr Chir 135, 305 (1916)) tied the hepatic artery in dogs and found they all died However, he stated that when he anastomosed the hepatıc artery side to side, with the portal vein they survived ligation of the hepatic artery Hosever I cannot repeat his operative technic and I rould like to meet the man who can Nareth had three dogs out of seven that survived such arterialızation of the portal vein One dog lived 12 and one lived 62 days The one that lived 62 days had a fine cirrhosis I am not satisfied with his description of the post mortem data He says nothing 19 ·about the state of the stoma except this, that when you forced a little of Teichman's furd into the hepatic artery it appeared in the liver Whether you could dislodge a thrombus that way or not I don't knox Furthermore he says nothing about the condition of the gall bladder To find that the gall bladder is gangrenous is one of the criteria that the hepatic artery has been properly tied Usually the gangrenous process becomes walled off Occasionally it does not be come walled off and there is bile peritonitis If the gall bladder is not gangrenous we say a vessel was mussed and when you look for it you find it every time A few words may be said regarding the technic of hepatic artery ligation The hepatic artery gives off the gastroduodenal artery, the right gastric artery and then divides into the two hepatic branches after first giving off the cystic artery The main trunk is a pulsating loop along the portal vein When you ligate the pulsating loop proximal to the gastroduodenal and right gastric arteries, the animals invariably survive Haberer (cited from ſ Markowitz, Experimental Surgery, loc cit) said that in order to get necrosis of the liver after tying the hepatic artery you must tie the pulsating loop to the right of the portal vein, the gastroduodenal artery, and the right gastric. In practise, wherever you see an arterial vessel going to the liver, tie ut There is one here, and there is one there, and you tie them and hope You should actually be able to see the bifurcation in the portal fissure but it is not good enough to tie only the terminal trunks be cause sometimes a little hepatic attery courses behind them If you tie everything in sight you invariably get hepatic necrosis The microscopic appearance of the dogs liver following hepatic artery ligation is striking An animal operated upon 24 hours prior, and having received no penicillin, develops massive necrosis On the other hand, if the animal receives large doses of penicillin the liver appears normal The penicillin is given in the form of 200,000 units intraperitoneally as well as a large dose intramuscularly I don't think the dose is important so long as some is given Shorr Do you think the intra abdominal route is necessary? Alarkowitz I dont think so We did it for good measure Shorr When do you first obtain demonstrable culture, in six hours? Markou stz I cannot answer We have done them in 18 hours but We have not done them before that The organisms are anaerobes, large spore bearing bacillt On microscopic examination, the lives seems almost to be a pure culture of such bacalli Armstrong Do you have to continue the administration of penicillin in order to keep the animals going? 20 ·Markour. I think in some instances one injection of peniallin sill do it If given only one injection of penicillin they might five five or sıx days We hare given many injections just to make sure We have kept it up for ten days and all survived indefnitely thereafter with no further penicillin Dr Hattroft has studied many of the sections from these animals I should like to have him comment on some of his observations Hartroft Studies of the ligated vessels have shown fibrin in the center of the vessel * In many cases the force of the ligaturing has ruptured the internal elastic lamına Elastic tissue stains have been employed to demonstrate that and to produce evidence that the vessels are being completely obstructed One would expect to see interesting changes in the portal canal 1 thuk surely one must after some time One sees very little in an Animal sacrificed 48 hours after the hepatic astery ligation of pensialını has been employed The artery is small but, as Professor Knisely has already shows us these vessels can vary greatly in size the bile duct epithelium however, is beautifully preserved, there is no evidence of degenerative changes There is no evident abnormality present in the central region of a liver lobule from an animal killed 48 hours after the hepatic attery was located when that animal was given penıcıllın It will be interesting to follow these animals at longer periods of time In the necrotic liver that occurs when no penicillin is administered following ligation of the hepatic artery one sees branches of a portal vein just filled with masses of bacteria I believe these are branches of a portal venn but because the liver is so necrotic it is difficult to dentify the structures with certainty Markowitz The function of the hepatic artery so far as we can tell as to maintain the concentration of the orygen at such a height that these organisins normally resident there dont grow We have thus far been unable to find any ather function for the hepatic artery apart from the mechanical one of nourishing the gall bladder Ve have watched such animals for as long as 5 months This is a new hind of homeostasis I wrote Professor Knosely to ask if he knew when the hepatic artery first appears in the animal scale Does a lobster have a liepatic artery. The liver is meant to be supplied purely by venous blood I think as you go down the loner animals you find conspicuously the portal vein drains in the liver which has developed phylogenetically as an organ capable of living at a low Editor De Hartros? presented asenes of lantern slides illustrating the fustological findings 21 ·or reduced oxygen tension The oxygen tension in mixed venous blood is I think 42 mm of mercury The mixed arterial blood is 72 mm of mercury The latter oxygen tension corresponds to that which obtains at a height above the earth of about 65 kilometers and a tension of 42 mm of mercury corresponds to that at about 10 kilometers Translate that and think what would happen in the pulmonary alveolt if we were breathing that sort of an It would be the difference between Pikes Peak and Mt Everest We can live at Pikes Peak but I dont think we can live well at Mt Everest Dragstedt and his school have for years maintained that the liver harbors large numbers of anaerobic bacteria When necrotic liver tissue is left about these proliferate Thus in a sense the animal dies of gas gangrene in the liver We wont have it that when you tie the hepatic artery little areas of gangrene develop in the liver which permit the spores to germinate It is an obvious enough thing to say the spores in such areas germinate and once they get started nothing can stop them Hartroft has looked carefully for such areas of gangrene and there are no such areas The liver looks remarkably normal three months after this operation is not markedly reduced in size and is unscarred Aladden Penıcıllın is just given for the first ten days but the animal lives for three months Markowitz Lives indefinitely Whether it needs to be given for ten days or three days I dont know Madden Arent they reinfected with these organisms? Markowitz That is a thing which has puzzled me Maybe they become immune Madden You dont get reinfected to tetanus and typhoid Shorr What about the collateral circulation? Markowitz That is being studied by Hartroft He can enlarge on that Sborr We have seen that Alarkowitz Maybe they get collateral circulation but the fact is they die without the penicillin and live after they have it Shorr After that period they develop a collateral circulation We have noticed that after some months Alarkowitz Where the musculo phrenics? Shorr Through the diaphragm Markowitz As soon as an animal dies or as soon as we kill it Hartroft does a careful autopsy He makes sure the ligatures are 22 ·tight Whether the animals develop a collateral circulation in the ultimate future is another problem It is very likely they do Patek Have these organisms been shown to be penicillin sensitive in culture Markou itz The Welch bacillus is known to be sensitive The soldier and gas gangrene who is given penicillin has tapıd healing of his infection Whether this is the Welch bacillus, or is a variant, as the bacillus of kog cholera is a variant of the human, I don't know I think the organism is either a Welch baatlus or close to it Some people say they actually are Welch bacalle Best. Really the problem has been do the dogs die without per: cillin if you tie the hepatıc artery and do they live with it? There are 1 1 LIL Lok. in At a recent lecture in Toronto one of the surgeons stated that he had once, like many other surgeons, inadvertently tied the hepatic attery and the patient had died The physiologists looked smug They know how to save an animal when they tre the hepatic artery If this situation should anse again I think the surgeon would want to give the patient all the penicillin he could I dont know whether the patients would live but I know the dogs do Watson I want to know to what extent this could be transferred to human beings Markouitz Quste directly l'asson I had the impression that the human liver does not normally contain resident spore forming bacıllı, anaerobes Markourz I merely quote Dragstedt i directly have no interest or authority in that problem Watson Are you acquainted with the studies of I H Manson (Thesis A Consideration of Some Factors Responsible for Gas Bacillus Infections University of Minnesota, 1934)? He worked on this problem quite extensively and my recollection of the study is to the effect that he regularly got the anaerobes in dog lovers but hardly ever could get them in human livers Markouriz This is a non sequitur Why do human beings die when you tie the hepatic artery? Best That could be from quite a different cause ll'atson I think that deserves a lot of discussion The problem of hepatic artery obstruction in human beings is one of infarction of the liver 23 ·Miarkowitz Total infarction? Watson No For example we saw a case a few months ago of an individual who had a cancer of the pancreas with involvement of the hepatic artery It appeared not to be completely obstructed although the lumen was not more than a pin point That patient had multiple infarcts in the liver Also in periarteritis nodosa of the hepatic artery one sees complete obstruction with thrombus secondary to the disease together with multiple infarcts in the liver Markou ntz Where is the obstruction in the portal lumen? Watson It varies in location but it may be up in the liver in the branches of the hepatic artery with multiple thrombı but there may be multiple infarction in the liver here as well As I understand it the process in the dogs is entirely diffuse and they dont have any infarcts Miarkourtz Years ago I tried to tie one of the two main branches of hepatic artery You would think no problem would be simpler I did iwo dogs and both lued I gave up If you want an explanation Professor Knisely supplied it There is enough arterio venous anastomo sis You have retrograde flow Best We must bring in the rats If you tie their hepatic artery they dont die Ds Arthur Colwell Jr in my laboratory has confirmed this fact but one of his rats did die with a gangrenous liver Fremont Smith I would like to ask if you have infarcts as in the Case you have spoken of would that not be the ideal situation for at least the local growth of the anaerobic organisms and that therefore the absence of such growth would be some evidence against their presence in the human where infarcts had been seen? Is that a reason able line of approach? Best Just to set the discussion a bit straight perhaps the situation is if you tie the hepatic artery in the dog you get a liver that is full of these organisms It goes on to almost total necrosis does it not? The situation is absolutely different in the rat where if you tie the hepatıc you dont usually have to give penallı at all When the hepatic artery has been inadvertently tied in the human beings the picture is obviously quite different than that in the dog or in the rat Therefore none of us know the application of these results to the human being Alarkou itz There is a case by R R Graham and D Cannell (Brit J Surg 20 566 (1933)) in which they accidently tied the hepatic artery I think the patient died of total necrosis of the liver Armstrong With reference to your remark about the rat this animal is notably resistant to many sorts of infections The fact that 24 ·the rat s liver does not show any necrosis from the anaerobic organisms might be the result of this mysterious resistance which this animal certainly possesses with regard to pyogenic infections Markou itz I suggest that that is the explanation Shorr The presence of anaerobic organisms normally in the dog liver calls to mind the experiments of Aub and his colleagues on the possible role of bacterial toxins in traumatic shock They partially detached skeletal muscle of the hind limb of the dog, encased it in a sheath and constricted the limb proximally They collected the exudate of the muscle preparation after about 9 hours and injected it into normal dogs It produced a toxic type of shock The exudate was then shown by them, in collaboration with Dubos, to consist of a rich culture of Clostridium welchit It is uncertain whether the organisms were harbored in the skin or muscle From work in our own laboratory we suspect the skin, since anaerobic preparations of dog skeletal muscle in vitro removed with strict asepsis, failed to show these organisms and indeed very few bacteria after 5 hours in nitrogen Best Di Bollman have you something to contribute here? Bollman I have two things I would like to mention First I want to thank Dr Markowitz for doing this nice piece of aork I have been worried about the importance of the hepatic artery for years and have considered the use of Welch antitoxın to accomplish the same thing that Ds Markowitz has done with penicillin I am not sure that Welch antitoxin would be as effective but it is not necessary nou to attempt the study The other thing I would like to comment on is the subject which has already been mentioned the collateral arculation to the liver After reading Dr Markowitz s paper in the Proceedings of the Society for Experimental Biology and Nledicine ne ligated the hepatic artery and gave penıcıllın and the dogs lived just as he says they do not all of the arterial blood to the liver enters by way of the hepatic artery There are to additional places where small arteries enter the liver one from the circle around the attachment of the vena cava to the diaphragm and the other from the lower side of the liver near the entrance of the portal vein This supply in the dog is not suf ficient to maintain the liver free of lethal infection (Slide dernon strated ) This slide was made four days after ligation of the hepatic artery of a dog which had received penicillin and appeared well at the time The animal was anesthetized and I tied the aorta in the chest and just above the renal arteries before injecting the barium acacia mass This v ray plate shows the distribution of the barium in 25 ·Obviously it does not go the whole route Something happens when you tie the hepatic artery completely in the dog It strikes me as peculiar that when you increase the rate of venous blood flow, the blood still has a lot of oxygen, but you do not keep these animals alive You can get a tremendous flow We have been talking about the dynamics and the difference between arterial and venous tension If you increase the rate of flow you can make a lot of oxygen avail able There is something very fundamental in this survival after ligation Fremont Smith I wonder if the shock of the operation puts the animal for a short time immediately after the operation in a state where with the hepatic artery tied off, the organism can start to grow in the liver Then there is the fact that a short course of penicillin, probably too short to allow full collateral Circulation to develop still allows survival and shows that there we are dealing with a ter condition Are the blood pressures of these animals low immediately after you tie the hepatic artery? Markou 1tz You mean where we just te without the Eck's fistula They are normal animals Fremont Smith There is not very much shock? Alarkou riz There is no shock Fremont Smuh Anesthesia does things Anesthesia is nearly always followed by something Alarkourtz We did two kinds years ago I quoted my teacher, De Mann, who says that ether is the best anesthetic With the greatest of respect we switched over to nerbutal as it is easier There seems to be no greater shock than with ether Naturally they sleep longer with the nembutal Fremont Smith Both are quite bad for the liver, both the barbi turates and ether They do knock out a good many liver functions Best I might say we were anxious to check these experiments and at Dr Markowitz suggestion and in quite a different laboratory the same experiments were carried out, ligating the hepatic artery We picked a young surgeon from the department, a very skilled man and he failed several times to get his anunals to die after ligating the hepatic artery It turned out that he did not know how to tie the hepatic artery in dogs, and again and again left a little branch There is something more than just talk ahen Markowitz says he ties every thing he sees You more or less have to do this Even a man who operated a great deal on human lovers and biliary tracts is not an expert in the case of the dog until he has had a lot of experience 28 ·Bollman Dr Best has touched on one thing which I think would be well to bear in mind We have mentioned the oxygen supply by the hepatic artery, but we have little evidence that the loss of oxygen is the major factor Fremont Smith mentioned that other things may be in there the cons sinusords as we saw acterial supply was probably continue to fill the sinusoids and distend them to occupy the space formerly held by azterial blood From the appearance of the liver I would suspect that the emptying of many of the sinusoids sas delayed and that considerable stasis developed The oxygen content of the portal blood is los and is undoubtedly much lower when stasis has occurred in the sinusoids The same could be said of the nutrı tonal elements of the blood and the removal of metabolites may be impaired I do not like to blame all the changes on the lack of Oxygen until we have something more definue about oxygen and the other factors involved Many articles in the latesature use the term reduction of oxygen supply waterchangeably ** arterial igation and in some instances I am sure there has not even been a reduction of the amount of blood flowing to the organ involved Pulson Could you control that by running portal blood through the oxygenator for the first 24 hours? I dont know how difficult it would be technically Pollman it svould be difficult but it could be done Fremont Smith Is there enough glucose? What is the glucose content? Bollman Less than in the arterial blood Fremont Smith Oxygen has to have something to utilize and I wonder whether there is enough for it to utilize Best The glucose content is not lowered This oxygen deficit is I believe just a good guess at this stage Alarkou tz I am trying to reason in the most elementasy fashion I can If you double the blood flow by severse Eck fistula dogs die when you tre the hepatic artery even though the linee is getting more blood than ever Bollman The fact that the animals survive long after the peniallin has disappeared indicates that something has happened during the period after arter al ligation and the elimination of the penicillin I doubt that sufficient collateral artertal arculation has developed to supply oxygen equivalent to that formerly supplied by the hepatic artery 29 ·Armstrong Coming back to oxygen, I dont knon what the oxygen tension of portal blood is Will it differ greatly from mixed venous blood or not? It it does not, there is still a lot of oxygen in that blood and much higher amount than the bacteriologists can tolerate when growing Welch bacille Best I dont know what the oxygen tension of the portal blood of the dog is under these conditions and even less what the oxygen tension of the cells in which the organisms are growing They are growing in the cells I presume Armstrong The cells should have oxygen tensions somewhat related Watson Probably much lower Best We dont know how much the normal hepatic blood flow raises the oxygen tension of the cells Madden You can put a catheter in the hepatic vein and study all the variations Best If we did all the things Markowitz wants us to do we would not do anything else in the department Stetten If the degree of anaerobiosis is sufficient in the cell to permit the growth of Welch bacıllı it may be that some of the oxida tive functions are not operating like the one which converts beta carotene to vitamin A which is definitely an aerobic liver function Short The blood unic acid and ammonia levels furnish a good index in the dog as to whether the metabolic processes in the liver are oxidative or anaerobic The blood constituents are kept at normal levels as the result of oxidative processes within that organ When oxygen tensions are sufficiently reduced to initiate an anaerobic type of metabolism the blood uric acid and ammonia values rise The use of these criteria in your experiments would help determine the extent to which anaerobiosis occurs and its duration Bess This growth of bacteria might be really secondary to some other primary change caused by the procedure Markou 1tz Which is non fatal Best It is not fatal The study of the liver after complete ligation of the hepatic asfesy in dogs receiving penıcıllın is a fascinating thing For the first time you have got dogs without an obvious arterial blood supply to the liver Bollman I am not sure that you do have a livet deprived of arterial circulation I do knon that there is a Istrle arterial circulation remain ing I wish you would continue these studies and find out whether this collateral arterial circulation develops with time or always remains small 30 ·Best I dont know much about this but it seems to me you have a little knob of arterial arculation near the diaphragm from another source than the hepatic artery Hos much this develops after hepatic artery ligation I dont know Hartroft May I ask a question Coming back to De Knisely have you in any study seen direct connections between the hepatic system verns and the portal system veins? Amisely I have never seen any a sinusoid can dilate very widely It may be as much as 200 micra Whether that is pathological or just physiological I dont know Bollman With injection technics some direct communications be tween the branches of the portal vein and the hepatic veins have been demonstrated I have seen them very clearly I suspect that you have similar connections in your photographs where you have the many branches of the portal vein going in all directions an occasional branch seems to directly join the hepatic vein Hartroft I believe I have seen them in cirrhotic liver Bollman They are very marked in cirrhotic livers You may have quite a large anastomosis direct from the portal vein to the hepatic vein B Ansely M Prinzmetal EM Ornitz Jr Simkin and HC Bergman (Am J Physiol 152 48 (1948)) have been inject ing glass beads in animals and catching them on the other side 1 dont know whether they found they were large enough to pass them Watson They did Hofbauer Spheres from 60 to 180 micra were recovered from the lungs of rabbits after such spheres were injected into the portal vein Fremont Smith Another suggestion for your consideration is to put a Goldblatt clamp on the hepatic artery and turn it doan gradually and see how rapidly you could bring it down without getting the infection Shorr I can tell you what happens under those conditions as we have been studying the effects of partial constriction of the hepatic artery by Goldblatt clamps in our laboratory during the past 3 years After two to four months such an extensive collateral circulation develops to the liver from the diaphragmatic vessels that it is neces sary to go m and tighten the champ and even completely constrict the hepatic artery to maintain the animal in the condition initially induced by the constriction We cannot say hou soon an extensive collateral circulation develops after partial constriction of the hepatic artery but we suspect it occurs quite early 31 ·Armstrong coming back to oxygen, I don't know that the oxygen tension of portal blood is Will it differ greatly from mixed venous blood or not? If it does not, there is still a lot of oxygen in that blood and much higher amount than the bacteriologists can tolerate when growing Welch bacıllı Best I dont knos what the oxygen tension of the postal blood of the dog is under these conditions and even less what the oxygen tension of the cells in which the organisms are growing They are growing in the cells I presume Armstrong The cells should have orygen tensions somewhat related l'atson Probably much lover Best We dont know how much the normal hepatic blood flow raises the oxygen tension of the cells Aladden You can put a catheter in the hepatic vein and study all the variations Best If we did all the things Markowitz wants us to do we would not do anything else in the department Stetten If the degree of anaerobiosis is sufficient in the cell to permit the growth of Welch baallı it may be that some of the orda tive functions are not operating like the one which converts beta carotene to vitamin A, which is definitely an aerobic liver function Shorr The blood uric acid and ammonia levels furnish a good index in the dog as to whether the metabolic processes in the liver are oxidative or anaerobic. The blood constituents are kept at normal levels as the result of oxidative processes within that organ When oxygen tensions are sufficiently reduced to initiate an anaerobic type of metabolism the blood uric acid and ammonia values rise The use of these criteria in your experiments would help determine the extent to which anaerobiosis occurs and its duration Best This growth of bacteria might be really secondary to some other primary change caused by the procedure Alarkou itz Which is non fatal Best It is not fatal The study of the liver after complete ligation of the hepatic artery in dogs receiving penicillin is a fascinating thing For the first time you have got dogs without an obvious arterial blood supply to the liver Bollman I am not sure that you do have a liver deprived of arterial arculation, I do know that there is a little arterial circulation remain ing I wish you would continue these studies and find out whether this collateral arterial circulation develops with time or always remains small 30 ·Best I don't know much about this but it seems to me you have a Iittle knob of arterial arculation near the diaphragm from another source than the hepatic artery How much this develops after hepatic artery ligation I don't knon Hartrofs May I ask a question? Coming back to Di Knisely, have you in any study seen direct connections betsseen the hepatic system Teins and the portal system seis? Knusely 1 have never seen any a sinusord can dilate very widely It may be as much as 200 micra Whether that is pathological or just physiological I dont knon Bollman With Djection technucs some direct communications between the branches of the portal vein and the hepatic veins have been demonstrated I have seen them very clearly I suspect that you have similat connections in your photogsaphs where you have the many branches of the portal sein going in all directions, an occasional branch seems to directly join the hepatic vein Hastroft 1 behere I have seen them in arrhotic liver Bollman They are very marked in cirrhotic livers You may have quite a large anastumisis dures from the portal sem to the hepatic vein Knisely M Prinzme ai (M Oroitz, jc, B Simkia and HC Bergman (Am J Phus 112 48 (1948)) have been inject ing glass beads in animals and catching them on the other side 1 dont know whether they fund they a ere large enough to pass them Iratson They did Hofbauer Spheres from in i 180 micra nere recorered from the lungs of rabbits after suche spheres rere anzected into the portal vein Fremont Smith Anrther sulgestion for your consideration is to put a Goldblatt clamp on t e hepune astery and tum it doxn gradually and see hox rapidly we could bring it down sithout getting the infection Shory I can tell you what happens under those conditions, as we have been studyin the effects of partial constriction of the hepatic aztery by Goldblat lamps in our laboratory during the past 3 years After top to four months such an extensire collateral arculation develops to the liver form the diaphragmatic vessels that it is accessary to go in and nchten he clamp and es en completely constuct the hepatic artery tumainiz r ne abimal in the condition initially induced by the consisICIAS ? We wann say hou soon an extensive collateral arculation des clops after partial construction of the bepaac artery, but ne suspect it xcurs quite cariy 31 ·Gjorgy May I make one remark? You discussed the phylogenesis of the hepatic artery in the role of the arterial blood for the liver As a pediatrician I may call your attention to the fact that the fetal liver has quite a lot of arterial blood Therefore it must play a role in addition to anything else Alarkowitz I have considered that I thought, now the fetal liver has twice as much arterial blood, maybe we can put in tivice as much in the adult It cannot be done on that point of view I will stake my right arm The junction flaps around, and thromboses So while you can accomplish such vascular anastomoses in the groun or neck, or in an Eck s fistula, because the parts are stationary, you cannot do so in the portal system Gjorgy Certainly the very large liver is connected with this fetal circulation Best We have been tremendously interested recently in rats put on diets low in choline after the animals have been starved We have seen hemorrhagic kidneys in old rats after starving and refeeding them We have seen it and Salomon has seen this too I have been wondering this morning if you did a partial hepatectomy and when that animal recovered and when the liver was presumably regenerat ing very rapidly on a good diet, if you tied the hepatic artery and gave the animal penicillin would you save him? There might be a demand for something which would cause death under those condi tions in spite of penicillin Markowitz That would be a good experiment One of the first things which has to be done is to see whether such a liver, with a hepatic artery tied shows regeneration or does not show regeneration after removal of portions of the liver I suspect it will regenerate Let us call these dogs without the hepatic artery but with the portal ven as having a portal vein liver Let us call dogs ssthout a portal vein flow to the liver as having an 'hepatic artery livet Finally if the blood flow to the liver is intact, let us describe it by saying the animal has a portal vein and hepatic artery lever Fremont Smith You mean that those dogs with only the portal sein intact would show regeneration? Markowitz The dogs whose liver is nourished by the portal vein kull, I suspect show hepatic regeneration after removing chunks of the liver Fremont Smith I thought you said already that has not happened Markowitz In the Ecks fistula dog the liver does not regenerate with a In the terminology I propose, let us call that anuma! a d hepatic artery liver 32 ·Goldblatt In the light of what you told us today, would you not say that no one else before you has ever produced a portal vein liser? If you insist that a dog dies of the hepatic artery is occluded then no one before has produced a portal vein liver is that incorrect? Alarkourtz That is correct ·SECTION III MASSIVE LIVER NECROSIS Gjorg) From a pathological point of view necrosis and cirrhosis are the characteristic and specific manifestations of injury to the hepatic parenchyma Fat infiltration per se is not specific enough, and often too transient, without concomitant manifestations of tissue re action, to be considered in itself as a truly pathological condition of the liver Today we are discussing the etiology of acute necrosis, which can be massive or zonal The production of experimental hepatic necrosis (massive or zonal) dates back many years and decades Such acute injury of the liver was looked upon is the consequence of a toxic influence on liver tissue by numerous noxious agents Such factors, most diverse in character, included various chemicals, both organic and inorganic, such as carbon tetrachloride, chloroform, phosphorus, etc, certain drugs, foreign proteins, bacterial products, immune serums and infections From these experimental studies, which in general, tally well with clinical observations, it was concluded that in every instance in which the prolonged or repeated "action of an agent has resulted in some degree of curthosis, the acute effects have been degeneration and necrosis of hepatic cells (V H Moon, Arch Path, i8, 381, (1934)) From this point of view, necrosis and cirrhosis are the acute and chronic forms of hepatic injury, often in response to the same etiologic agent The response of the hepatic tissue to insults of various kinds manifests itself in necrosis when the insult is overa helming, or when the resistance of the hepatic parenchymi is reduced below the norm whereas continuous or sepeated exposure to the same agent given in doses which are in general below the necrogeruc level results in cirrhosis When cirrhosis follows interrupted, repeated exposures of hepatic tissue to a noxious agent, it may be inferred that the cajucious effect is achieved by proper timing of the insults which precluded the complete repair of the hepatic parenchyma between consecutive exposures without ever producing histologically demonstrable necrotic changes, and reduced the normally hugh regenerative power of hepatic parenchyma This Cirrhosis cannot be distinguished in many cases from ordinary Laennec s cirrhosis These well established facts should be kept in mind because it makes it very difficult to distinguish completely even in dietary necrosis or 31 ·dietary arrhosis tho phases of specific liver injury That is not on the program of today The question is Is there a dietary hepatic necrosis? Can it be produced? Can It be produced at will In 1939, Dc Goldblatt and I had published a paper on the occur rence of acute necrosis in rats lept on a semi synthetic diet not com plete in various members of the vitamin B complex in this group of observations acute necrosis nas not produced at will but was found in a relatively large percentage of the experimental rats (in 48 out of 300) It was never encountered in rats receiving the same basal diet supplemented with yeast extract With the use of diet rations in which the level of protein (casein) was kept at 10 percent or below hepatic injury became a regular occurrence produce experimental period up to 150 about 120 to 150 gm (or more tas reached and published almost simultaneously by four groups of investigators (cf P Gyorgy and H Goldblatt ) Exp Med 89 245 (1949)) in these experiments liver injury manifested itself either in the form of acute diffuse necrosis os of slonly developing diffuse fibrosis reminiscent even in the gross of Laennec s hepatic arrhosis in some rare instances zonal or massive acute necrosis and circhosis were en countered in the same animal It is not surprising that necrosis (zonal or massive) and cirrhosis produced by dietary means were first considered in the same relation to each other as the corresponding pathological changes in the liver occurring under the influence of noxious agents such as chloroform carbon tetrachlorde etc It has been assumed that necrosis is only the acute manifestation of a quantitatively more severe insult, than that causing cirrhosts which follows chronic and moderate hepatic injury A closer analysis of the etiologic dietary factors found in acute necrosis and in cirrhosis furnished results which seemed to be at least to some extent at variance with the unitarian theory of their etiology It was observed that supplements of choline os Iner fxuact to the experimental diet would prevcat cucslioses but 40t wnhout influence on or would even promote acute diffuse seanss Con versely stine given as supplemeat to the expecimental din certed a remarkable protective effect on the development of radiosis buvo seemed to enhance the production of hepatic cirrhosis On the other hand methionine or protein rich in methionine free ed both an diffuse necrosis and cirrhosis Protein lox in methion ne and at phim same time rich in stune such as peanut protein ni'l presenta necess 35 ·but will enhance, even when added to the diet in high relative proportion, the development of cirrhosis The rather complicated interrelations and effects of protein, methionine, cystine and choline as well as of other dietary factors (fats and vitamin E) to be discussed later, are surnmarized in Table I TABLE 1 Effect on Protein (Methionine Methionine Cystine Containing) Choline V1! E Far Beneficial Beneficial Injurious Beneficial 0 Injurious Currhosis Massive Necrosis Beneficial Beneficial Beneficial Oor Injurious Beneficial O or Injurious FS Daft, W H Sebrell and R D Lillie (Proc Soc Exp Bol and Med, 50, 1, (1942)) as well Himsworth and Glynn (cf Hımsworth, HP, The Liver and Its Diseases, Blackwell and Co, Ltd, Oxford, (1947)) have sharply distinguished massive hepatic necrosis from diffuse fibrosis (cirrhosis) They considered massive hepatıc necrosis as a sign of cystine (methionine) deficiency, and arrhosis more as a result of a deficiency of choline and its precursors, including methionine With the addition of choline to a diet low in casein, cirrhosis may be completely prevented but the incidence of necrosis, although accelerated and intensified was increased only slightly The irregular occurrence of massive necrosis in rats fed a diet low in casein was claimed by Himsworth (loc cit) to be due to the quality of the casein which even when given in small amount supplied methionine (and cystine) not too far below the prophylactically effective level In this connection it should be pointed out that the hemorrhagic foci ob served by T E Weichselbaum (Quart J Exp Physiol 25, 363, (1935)) in the livers of rats fed a cystine deficient diet may have represented necrotic foci, but were not described in detail or recog nized as necrosis By substituting yeast for casein the intake of sulfur containing amino acids may be further lovered Thus, it seemed to fulfill ex pectations, when reports appeared (H P Himsworth, loc ait, A Hock and H Fink, Ztschr physiol chem, 279, 187, (1943)), that in rats fed rations with yeast as the sole source of protein massive necrosis became a regular occurrence According to these observations the development of hepatic necrosis was again completely suppressed by prophylactic supplements of cystme or methionine 36 ·These experiments bring us up in chronological order to 1943/44 Our original basal diet contained casein (8 10%), and, as fat, lard (20 23%) usth cod liver oil (2%) mixed asth the ration In rats fed this low casein high lard diet, acute necrosis developed in about 40 50% of the experimental animals Substitution of percomorph oil, 3 drops once a week, for cod liver oil mixed with the diet * or of butter fat for lard reduced markedly the incidence of acute necrosis in rats kept on a low casein diet Thus, unsaturated fatty acids, as present in cod liver oil and the lard, seem to enhance the development of acute hepatic necrosis (See Table 1) In 1912, due to wartime restrictions we had to substitute hydrogenated cottonseed oil (Crisco) for lard From that moment on necrosis became conspicuous through is absence In studying the possible cause of the sudden disappearance of necrosis, we tned various other fats, and we always ran across one correlation, namely the vitamin E content of the fat It appeared, and evidence for this conclusion was presented at our Conference in 1947 (Liver Injury Transactions of the Sixth Conference, Josiah Macy, Jo Foundation New York, (1917)) that vitamin E sul present the development of dietary acute necrosis in rats receiving a diet low in sulfur containing amino acids Similar observations were presented during the war, unbeknown to me by Dr Schwarz from the laboratory of Professor Richard Kuhn, Heidelberg, Germany (K Schwarz, Ztschs physiol chem, 281, 101, & 109 (1911)) I am delighted and I am sure I can speak in the name of our whole group, that Dr Schwarz is with us today and will par ticipate in the discussion of the topic in question Dr Schwarz used in his experiments a special alkaline treated casein By feeding young weanling rats a semi synthetic diet similar to ours with the alkaline treated casein as the sole source of protein, massive necrosis of the liver developed in a large percentage of the experimental animals in contrast to our observations cystine and methionine were found by Schwarz ineffective in the prevention of this type of hepatic necrosis while on supplements of sitamine no necrosis of the ler developed We have tried to repeat the expert ments of D. Schwasz but nere unsuccessful (i Gyorgy and H] Goldblatt Exper Med 89 245 (1919)) This negative result was due probably to some differences in technic and perhaps eten more to the different casein samples used As stated before in our observations the occurrence of acute necrosis meser reached higher figures than 40 50% when rats were kept on a diet loss in casein and free from tocophero! This con trasted with the findings of Hock and Fink as well as Himsnorth 37 ·(loc ait), who reported a practically 100% incidence of massive necrosis in rats kept on a diet in which yeast furnished the only source of protein In order to solve this apparent discrepancy we carried out a large number of experiments using yeast rations in several com binations, such as high and low in yeast, in various fats (lard, Crisco), with or without cod liver oil, with or without tocopherol and/or cystine The yeast was a Type 50B dried Brewers yeast of Standard Brands Inc Again, the highest figure for incidence of massive necrosis, on a low yeast high lard cod liver oil ration (without cystine or tocopherol) was only about 40% A further group of 35 rats was put on a ration modeled after the experimental ration of Professor Himsworth (personal communication) American (Type 50B) yeast 18, corn starch 79 salt mixture, 3 Arachis oil o 5 ml and cod liver oil, 2 drops, were added to 8 gm of the dry mixture just before feed ing Further all animals received daily the usual vitamin B supplements on this diet All our experimental rats remained healthy, and at the end of the experimental period (200 days) showed at autopsy no signs of acute or healed hepatic necrosis Subgroups of these rats were exposed to various forms of stress (injections of adrenocortico trophic hormone, cold, exercise), and still no hepatic necrosis de veloped In the light of these completely negative results it appeared to us as the only logical conclusion, that the discrepancy between our ob servations and the positive findings of Himsworth might have been due to the fact -- other conditions having been more or less equal that different yeast brands were used in the respective rations Through the kindness of Professor H P Himsworth we obtained the same type of British Baker s yeast (United Yeast Co, Ltd, London) which was used in Professor Himsworths laboratory Fermentation tests showed no activity for the American yeast whereas the Batish yeast exhibited slight but definite activity The British yeast was given in form of the same ration as stated above for the American yeast to three groups of rats. Out of 18 rats fed the basal ration with the British yeast as the sole source of protein 15 animals died betw een the 100th and 145th and 2 more animals died between the 150th and 200th experimental day from acute hemorrhagic necrosis In the whole group only one rat survived the experimental period of 200 days and showed at autopsy normal liver The massive character of necrosis extending over many lobules and even lobes distinguished it from the usual more zonal form of necrosis (central and madzonal) seen in rats fed rations low in casein In all macroscopic and microscopic details the 38 ·findings conform to those given by Himstvorth for his animals he under identical conditions It was the first time that in our laborato massive necrosis was obsened as a truly regular occurrence In tw subgroups of animals consisting of 11 and 13 rats respectively, and fed the same basal diet (with the British yeast) supplements of tocopherol (3 mg daily) or cystine (50 mg daily) prevented the development of hepatic necrosis throughout The experimental period was extended to 200 days All these experiments with the two yeast rations aere carried out concurrently with rats of the same strain (Sprague Dawley) age, and identical dietary history Acute necrosis as it appeared irregularly in rats fed a diet low in casein, and with impressne regularity ja rats fed a ration containing a particulas brand of British yeast as sole source of protein may be prevented either by supplemeats of sulfus-containing amino acids or of tocopherol These observations are difficult to reconcile with the previously held view such identified massive hepatic necrosis with deficiency of cystine (methionine) Furthermore under special condi tions such as in the experiments of Dc Scha arz as well as in those recently reported by [ L. Hove D H Copeland and W D Salmon (red Proc 8 386 (1919)) cystine and methionine were ineffec tive in the presention of massive hepatic necrosis shich honeter still responded to the prophylactic administration of tocopherol in the experiments of Schwarz an alkalı treated casein and in those of Hove an oxidized (methionine free) casein was used This may explain the difference in response to cystine or methionine between these expert ments and those in which untreated regular casein served as source of protein The important role tocopherol plays in the etiology of massive hepatic necrosis and the fact that the body holds tenaciously to its tocopherol stores may explain to a large extent the difficulties several authors encountered in their effort to produce hepatic necrosis in rats Thus the pre experimental diet and its duration should exert a decisne influence on the development of massive hepatic necrosis (H P Himsworth and O Lindan Nature 163 30 (1919)) The preparatory period may be reduced to 14 21 days (mcan 23 24 S D 836 days) by feeding the usual yeast diet to young ueanling rats (23 days old) whose mothers acte put 10 days after the latters were bom on a sitamin I faced et When the female rats with young litters were kept through out the shole period of lactation on the usual stock diet containing vitamin I the de clopment of massic acute necrosis in the young rats recen ing the yeast diet is prolonged 39 ·to 60 days or more (personal communication of Professor Himsworth) The peculiar interrelation between cystine (methionine) and tocoph erol in the prevention of acute massive and zonal hepatic necrosis requires further elucidation The interchangeability of sulfur contain ing amino acids (cystine, methionine) and vitamin E as leading etio logical factors makes it difficult to accept pure deficiency as the basis of acute diffuse hepatic necrosis Furthermore, it has been amply proven that cystine when given in excessive doses (M X Sullivan, W C Hess and WH Sebrell Pub Health Rep USPHS 47, 75 (1932), D P Earle, Je and J Victor, Exp Med, 73, 161 (1941)) and often even not too greatly (A C Curtis and L H Newburgh, Arch Int Med, 39, 828 (1927)) in excess of the physiological norm will not prevent but in fact will promote the production of mas sive hepatic necrosis Methionine and cystine are known as detoxi fying agents in counteracting the noxious effect of hepatotoxic sub stances and related poisons Recently similar observations were re ported with regard to the effect of vitamin E in the particular case of carbon tetrachloride poisoning (E D Hove, Arch Biochem, 17, 467, (1948)) In rats kept on the usual low casein high fat diet supplemented with the carcinogenic dye N, N, dimethylaminoazoben zene (butter yellow) acute diffuse necrosis reached a much highe: incidence than in control groups fed the same diet without butte yellow Nevertheless in both groups supplements of cystine or toco pherol prevented the development of necrosis equally well (P Gyorgy and H Goldblatt locat) This may indicate that cystine and tocopherol neutralized first the effect of the exogenous poison butter yellow and simultaneously also that of some unrecognized endo genous metabolic poison Acute diffuse necrosis caused by large doses of cystine should also be considered as a toxic manifestation, perhaps initiated by the disturbed normal ratio of cystine and choline (P Gyorgy Am J Clin Path 14 67 (1944)) In this connection one should recall the enhancing effect of choline (also liver extract), on the production of hepatic necrosis in rats fed a ration low in cystine and methionine The phenomenon of hepatic necrosis provides an impressive illus tration of the difficulty encountered in distinguishing bettiseen mechan isms based primarily on deficiency or intoxication These mechanisms may be thoroughly intern oven and do not permit a distinct separation of the causative factors This brings us back to the question as to whether the interchangeability of cystine (methionine) and tocopherol may be primarily due more to an underlying detoxifying mechanism than to the restitution 1 40 ·of a defiaeacy condition If this were the case, then the regular occurrence of massive hemorrhagic hepatic necrosis seen in fats on the special British yeast ration might be traced either to presence of a toxic factor or to the low content of a special detoxifying con stituent in this yeast The chemical analysis of the American and Batish brands of yeast 15, of course, essential as a preliminary possible ansa er to the question What is the difference between these two brands of yeast? The analytical data for N and total s were ONG NGO SCO American yeast (Fleischmana s Type SOB) 783 British yeast (United Yease Co Ltd London) 676 065 The greater analytical value for total S in the American yeast mar indicate a higher content of cystine (or methionse) than that present in the British yeast The further possibility that the British yeast being an active yeast, might be less well digested and utilized (E L Price MM Marquette and H T Parsons ) Nutr, 34, 311 (1947)) than the inactive American yeast received no direct support in the expen ments of Himsworth and Lindan, who found no difference in the necrogenic effect of the Betish yeast when given alive or heat killed (personal communication of Professor Hemsworth) The average food intake in the group of rats fed the diet with the American or British yeast was nearly equal (in the average 65 gm dally in the first 14 weeks for the group with the American yeast and 63 gm dasly mn the first 14 seeks for the group sath the British yeast) On the other kand the gain in nesght during the total experimental period was distinctly higher in the group of animals receiving the American yeast than in the group of animals lept on the diet with the British yeast as source of protein Even when all these anous factors are con sidered the difference in sulfur intake about 25 percent does not appear to be distinct enough to give a satisfactory explanation for the complete absence of hepatic necrasts un one group and for its regular occurrence in the other further A Hock and H Fink (2 Natur forsch 26 187 (1947)) reported recently on a similar greater growth promoting activity of Breners feast compared with that of Balets jcast Nevertheless the incidence of massie hepatic necrosis nas apparently equally high in both groups of theit sats Obviously direct analytical data are necessary in order to express the Ssalues for either of the yeast samples as Cystone or methionine Such analytical work is at present in progress Horever it is north pointing out that in previous experiments (P Giorgy and H Goldblatt loc ot) cats 41 ·kept on a ration containing lard and American yeast as sole source of protein and in which the yeast intake was less than half as compared to the rats receiving the ration with the British yeast, the incidence of acute necrosis, in a rather milder form, barely reached 40 percent It should also be added, that in rats of both groups receiving Batish or American yeast, severe ulcero papillomas (to be described later in a separate communication) of the fore stomach were noticed which were absent in the groups receiving a supplement of cystine This indicates that even the American yeast could not have contained the necessary physiological minimum of cystine (methionine?) The other question which has been raised at the meeting of the Army Liver Commission by De Watson a few months ago, concerns the possible difference in the vitamin E content of the American and British yeast Generally, it is assumed that yeast is free from vitamin E All the yeast brands examined chemically in the laboratories of Dis tillation Products, Inc (Rochester, N Y) have been found according to personal communication of Dr P L Harris, to be free from toco pherol At least chemical isolation procedures did not give any indica tion for the presence of tocopherol ( The possibility, although perhaps remote, that yeast might contain some tocopherol like substance which cannot be extracted in the usual way has been investigated by a new technic We have recently shown (C S Rose and P Gyorgy, Fed Proc 8, 244, (1949)), that if a rat (we used rats of the Sprague Dawley strain) fed on a usual mixed grain diet is put on a vitamin E free diet, consisting of casein, salt , sugar, vitamins (with or without fat), beginning deficiency may be demonstrated after 3 7 days Red blood cells of such rats, washed and suspended in buffered saline solution will be hemolysed by minute amounts of dialuric acid less readily by ascorbic acid or cysteine Thıs hemolysis may be prevented by tocopherol added in vitro to the assay mixture Further, hemolysis will never occur - and I underline the word never — when red cells were taken from rats secesving tocopherol in their dret We have put rats on the usual necrogenic semi synthetic rations containing American and British yeast respectively and after 3 days we examined the red cells of the rats to see whether or not they would be hemolyzed by dealuric acid Hemolysis occurred with equal intensity regardless whether the blood was taken from rats receiving American or British yeast Thus, both brands of yeast must be essentially free from tocopherol The fact that yeast appears to be particularly conduave to the development of massive hepatic necrosis may be due not only to a deficiency of protective (detoxifying) factors, such as cystine (methio 42 ·niae) and vitamin E, but perhaps also to the presence of some un identified toxic factors in the yeast, or a more probably to the directly or indirectly supported production of some toxic metabolites Such toxic substances may originate in the intermediary metabolism, or under the sofluence of the intestinal flora, in the intestinal tract, in particular in the large intestine In any case, British yeast should pro. mote the production of such toxic factors As first pointed out by Himsworth (loc cit) the necrotic changes in the liver seem to prefer the left liver Himskorth saw in this difference a circumstantial evidence for the so-called trophopathic etiology of the condition Blood from the small intestine supplies mainly the sight liver Protective food constituents cystine, methio nine and as we know now also vitamin E are absorbed from the small intestine and directed to the right liver Blood from the Jarge intestine must be obviously low in or free from such protective food factors and in consequence, the deficiency condition --- necrosis should be more pronounced in the left liver On the other hand if toxic products originate in the large intestine, the left Iner should show more signs of injury than the right liver All these are, of course, purely hypothetical considerations How ever, as clinician I am impressed by the fact that a pure deficiency necrosis of the liver is unknown in clinical medicine The interplay of toxic factors tallies certainly better with clinical observations as far as hepatic necrosis is concerned Necrosis associated with viral hepa tatus is the most common form of this acute hepatic injury in the clinic Various chemical poisons, such as carbon tetrachloride phos phorus the unidentified cryptogenic poisons such as in eclampsia are other common etiological factors in the development of hepatic necrosis As detoxifying agents, vitamin E and cystine will probably act through the intermediary of some oxido-reductive reactions although the possibility of diferent points of attach resulting in the same effect ie in the prevention of hepatic necrosis cannot be excluded Necrosis of the laer is in most cases demonstrable in the gross In the few instances when it stays latent st may be brought out in a very ingenious say as devised by Professor R Kuhn Heidelberg (personal communication) Normal livers of freshly killed animals immersed in a 05% aqueous solution of phenyl tetrazolium chlonde will take up a bright red color (formazan formation) In contrast necrotic Iners or necrotic foci will retain their normal color after exposure to the dye Maybe this method which is based on Oxido ·reductive reactions will be of value in testing specimens of liver obtained through biopsy Watson Is it an external stain Gyorgy Yes Hanger The non injured cells take up the stain? Gyorgy That is correct Finally I cannot resist from calling attention to some practical implications of the research on hepatic necrosis, which appears to me rather intriguing especially because for me as a pediatrician, it has pediatric connotations In Rh incompatibility several research workers, among them Sir Leonard Parsons (Birmingham, England), N Philpott (Montreal) and myself have discussed the role of liver injury as a central patho genetic factor in the outcome of the disease, including especially the hemorrhagic manifestations and also kernicterus Hepatic injury fol lows some until yet unidentified phases of the antigen antibody reaç tron, which in itself occurs obviously as the first chain in the events linked with the syndrome of Rh incompatibility The hepatic injury 1s characterized chiefly by acute zonal or massive necrosis Protection of the liver may be attempted by methionine Philpott has already presented preliminary evidence in favor of this view We may now add the further possibility that during the sensitization process which is at the very bottom of Rh incompatibility, the normally low vitamın E stores of the fetus and the newborn are further depleted Hepatic necrosis is one of the characteristic sequelae of vitamin E deficiency under particular conditions Edema and pulmonary hemorrhage are often encountered in very severe human erythroblastosis and seen also in experimental vitamin E deficiency in animals Studies are in progress jointly with Dr Carl Bachman, Professor of Obstetrics, School of Medicine, University of Pennsylvania, on the combined use of methionine and vitamin E in the prevention of the unspecific sequelae of Rh incompatibility, such as hepatic necrosis, kernicterus, pulmonary hemorrhage and edema It is not expected that the anemia, based on immunological antigen antibody reaction, will be influenced by the administration of methionine and vitamin E In summary, I would like to conclude that it is possible to produce regularly by purely dietary means, massive necrosis of the liver in rats Further, it is possible to prevent such massive necrosis of the liver again with regularity, by supplements of sulfur containing amino acids or of vitamin E The final cause and pathogenesis of massive necrosis has still to be elucidated Best I have a question along the same line You say by dietary 44 ·means you produce necrosis, and yet you really favor now the presence of a toxic substance Those are not compatible in my opinion because you can produce almost anything by dietary means if you add some toxic substance Gyorgy The question of deficiency versus intoxication is difficult to answer If the experimental diet does not contain a poison as we understand poison and consists only of the usual food constituents such as casein yeast and fat as employed in nutritional research con ditions produced by such ration could be considered of dietary origin If we would add to the ration butter yellow, we added a poison Best This condition occurs if you use a certain peast but not with another yeast Gjorg) Our experimental yeast is used in England for baking purposes Fremont Smh May I suggest putting quotation marks around dietary? You would not object? Gyorgy No Fremont Smuth That would satisfy Dr Best? Best Yes for the time being Gyorgy If one adds raw egg white to the diet brotin will be bound and inactivated in the intestine Thus the egg white will appear as a toxic factor leading honeter to a deficiency condition Best I accept that as dietary Gyorgy On a diet low in casein one may get necrosis in 40% of the experimental animals There is the possibility that on such dict intermediary tovic metabolites are not detoxified through lack of sulfur containing amino acids or vitamin E It is a fair assumption that toxic metabolites are continuously produced in our body which require detoxication it alson Could you not classify them as absolutely and relatnely toxic Arsenic is absolutely toxic and ess white only relaticly poisonous Best But the mechanism is different Gyorgy One may go even further At the end of the last and at the beginning of the present century so-called autointoxication from the intestine was a much discussed subject in medicine Such intestinal poisons are metabolites or products of intestinal bacteria und special food constituents per se innocuous may promote their production in the intestine Best Dr Schwarz would you tell us about some of your studies? Seluarz It is both an honor and a pleasure for me to participate in this conference and I thank you for this opportunity 45 ·reductive reactions will be of value in testing specimens of liver obtained through biopsy Watson Is it an external stain? Gyorgy Yes Hanger The non injured cells take up the stain? Gyorgy That is correct Finally, I cannot resist from calling attention to some practical implications of the research on hepatic necrosis, which appears to me rather intriguing, especially because for me as a pediatrician, it has pediatric connotations In Rh incompatibility several research workers, among them Ser Leonard Parsons (Birmingham, England), N Philpott (Montreal) and myself have discussed the role of liver injury as a central patho genetic factor in the outcome of the disease, including especially the hemorrhagic manifestations and also kernicterus Hepatic injury fol lows some until yet unidentified phases of the antigen antibody reac tion, which in itself occurs obviously as the first chain in the events linked with the syndrome of Rh incompatibility The hepatic injury is characterized chiefly by acute zonal or massive necrosis Protection of the liver may be attempted by methionine Philpott has already presented preliminary evidence in favor of this view We may now add the further possibility that during the sensitization process which is at the very bottom of Rh incompatibility, the normally low vitamin E stores of the fetus and the newborn are further depleted Hepatic necrosis is one of the characteristic sequelae of vitamin E deficiency under particular conditions Edema and pulmonary hemorrhage are often encountered in very severe human erythroblastosis and seen also in experimental vitamin E deficiency in animals Studies are in progress jointly with Dr Carl Bachman, Professor of Obstetrics School of Medicine, University of Pennsylvania, on the combined use of methionine and vitamin E in the prevention of the unspecific sequelae of Rh incompatibility such as hepatic necrosis, kernicterus, pulmonary hemorrhage and edema It is not expected that the anemia, based on immunological antigen antibody reaction, will be influenced by the administration of methionine and vitamin E In summary, I would like to conclude that it is possible to produce regularly by purely dietary means, massive necrosis of the liver in rats Further, it is possible to prevent such massive necrosis of the liver again with regularity, by supplements of sulfuc containing amino acids or of vitamin E The final cause and pathogenesis of massive necrosis has still to be elucidated Best I have a question along the same line You say by dietary ·means you produce necrosis , and yet you really favor now the presence of a toxic substance Those are not compatible in my opinion because you can produce almost anything by dietary means if you add some toxic substance Gjorgy The question of deficiency versus intoxication is difficult to answer If the experimental diet does not contain a poison as we understand poison and consists only of the usual food constituents such as casein yeast and fat as employed in nutritional research con ditions produced by such cation could be considered of dietary origin If we would add to the ration butter yellon we added a poison Best This condition occurs if you use a certain yeast but not with inother yeast Gyorgy Our experimental yeast is used in England for baking urposes Fremont Smith May I suggest putting quotation marks around tary? You would not object> Fjorgy No 'remont Smith That would satisfy Dr Best? est Yes for the time being vorgy If one adds raw egg white to the diet biotin will be bound nactivated in the intestine Thus the egg white will appear as wuxic factor leading however to a deficiency condition Best I accept that as dietary Gyorgy On a diet low in casein one may get necrosis in 40% of the experimental animals There is the possibility that on such diet intermediary toxic metabolites are not detoxified through lack of sulfur containing amino acids or vitamin E It is a fair assumption that toxic metabolites are continuously produced in our body which require detoxication ivatson Could you not classify them as absolutely and relatively toxic? Arsenic is absolutely toxic and egg white only relatively poisonous Best But the mechanism is different Gyorgy One may go even further At the end of the last and at the beginning of the present century so called autointoxication from the intestine was a much discussed subject in medicine Such intestinal poisons are metabolites or products of intestinal bacteria and special food constituents per se innocuous may promote their production in the intestine Best Dr Schwarz would you tell us about some of your studies? Schu arz It is both an honor and a pleasure for me to participate in this conference and I thank you for this opportunity 45 ·The experiments on which I will base my discussion were conducted at the Kaiser Wilhelm Institute at Heidelberg between 1940 and 1945 Due to general events, complete publication of my research has been delayed However two papers appeared in 1944, one of which described the production of liver degeneration by dietary means, and the occurrence of liver protecting substances (K Schwarz, Ztsche physiol chem 281, 101 (1944)) while the second described the concentration and identification of vitamin E as a liver protecting factor (Ibid, 281, 109 (1944)) Unfortunately, these papers did not become available in this country until 1946 Other details have appeared subsequently (Ibid 283 106 (1948), Ibid 283, 186 (19-19)) I would like first to discuss Dc Gyorgy s experiments in relation to ours I was very much pleased to hear that Dc Gyorgy has been able to confirm our results with vitamin E With reference to the points where our results disagree I think there is much less discrepancy than Dr Gyorgy supposes Dr Gyorgy mentioned that sulfur containing amino acids had no protective effect in my experiments This is true only in part, namely in regard to the so called casein VI liver injury As I pointed out at the International Conference on Vitamin E (K Schwarz, Ann N Y Acad Sciences 52 (1949)) and is evident from publications, I had to destinguish between three different types of dietary liver injuries These are different in their histological and pathological aspects and in their relation to different liver protecting agents The damage which is produced by yeast as a main protein source in synthetic diets is the one applied by Dr Gyorgy In my experiments cystine and methionine had exactly the same protective action as in Dc Gyorgys tests when studying yeast liver damage (K Schwarz, Ztschr physiol chem 283 186 (1948)) In his experiments Ds Gyorgy has been unable to reproduce the casein VI liver damnage This might be due to technical differences 1 He used casein as a starting material which was not as highly purified as necessary It can be shown easily that crude casein has a protecting action against liver degeneration although the factor in or property of casein which is responsible for this action has not been identified If we assume — as seems probable — that this action is due to a protective substance, then it must be rather firmly bound to the protein since it is so difficult to remove The material used in our laboratory for the production of casein VI had been thrice repre cipitated from alkaline solution using the method of Hammersten and had been extracted with ether before the preparation of casein Vi was started 46 ·3 2 When he tried to produce the casein VI liver damage, Dr Gyorgy used a material named Vream instead of extracted butter fat (P Gjorgy and H J Goldblatt J Exp Med 89, 245 (1949)) The method of preparation of butter fat has been given in detail together with that of casein VI in the first of our publications Vream (Swift & Company) is a shortening of vegetable origin In Dr Gyorgys experiments, liver damage was inhibited by another vegetable shortening, namely Crisco (P Gyorgy, Sixth Cont on Liver Injury Josiah Macy, Jr Foundation, N Y (1947)), and it seems quite possible that Vream might have the same protective quality as Crisco In our experiments we watched not only the purity of the casein and fat, but we also used a specially refined sucrose which was tested microbiologically to be free of growth promoting factors In my experiments it was necessary to prevent the animals from getting too many protective substances during their nursing period if liver injury was to be obtained Therefore the breeding mothers were put on cooked rice and skimmed milk starting from the date of birth of their litter The young animals were put on experimental diets as soon as they weighed 28 to 30 grams The liver damage did not appear as early or as regularly when the mothers received stock diets or when the young animals were allowed to remain for a longer time with their mothers Several hundred days were needed to develop the liver damage when adult rats were put on experimental diets 4 Furthermore it is natural that differences occur among animals of different strains or litters In our experiments the rats were not inbred but the litters were always evenly distubuted over the experi mental and control groups In order to give a more complete picture I shall outline the develop ment of these experiments Between 1937 1939 efforts were made in Heidelberg to isolate the so called filtrate factor using lactobacillus plantarium (streptobacterium plantarum Orla Jensen 105) as a test organism The main component of this fraction finally turned out to be identical to pantothenic acid (E F Moller Ztsche physiol chem 260 246 (1939)) In highly purified concentrates of this principle, prepared from rice bran was able to separate a new growth promot ing factor which was necessary for our lactobacillus (E F Moller and K Schwarz Ber Dtsch Chem Ges 74, 1612 (1941)) The new factor was called H' we were finally able in 1941 to isolate the factor H and identify it as p amıno benzoic acid (R Kuhn and K Schwarz Ibid 74 1617 (1941)) Before we knew that factor H' was p amino benzoic acid, experi ments were started to determine whether it was a vitamin for the 47 ·rat During these experiments I happened to develop a method to produce fatal liver degeneration in young rats by dietary means ! tested the pamino benzoic acid content of a variety of caseins and found that a large amount of this factor was present As much or even more of this material could be found in crude casein as in the best natural sources, liver or yeast The content of p amino benzoic aad in casein preparations seems to be a good method to determine the purity of the material Even so called vitamin free and highly extracted samples of casein contain an amount of p amino benzoic acid which is easily detected by the growth test with lactobacillus plantarium Paminobenzoic aad seems to be firmly bound to the protein molecule and it is necessary to heat the casein under slightly alkaline conditions in order to eliminate factor H' In our procedure the casein was boiled in solution starting with pH 87 for three hours During the treatment the pH decreased slowly to 75 The substance which resulted after reprecipitation with acid, drying with alcohol and extraction with ether was named casein VI When young rats were fed on diets with this highly purified casein VI, more than 84 percent of them died within 12 to 80 days (average 42 days), in a quickly developing coma leading to death in à few hours Their livers were severely degenerated The injury appeared so suddenly that it was almost impossible to detect it before the onset The development of this disease was not related to p amino benzoic aad A daily dose of 1 milligram had no influence on the casein VI injury The content of this factor thus was used merely as titer for the purity of the casein The real reason for the fact that casein VI induces liver degenera tion, whereas the starting material does not, is not yet known Casein VI has the same optical rotation and nearly the same C, H, N, P, and S content as the Hammersten casein from which it was prepared casein VI itself did not seem to be toxic By varying the casein VI content of the diets, it could be shown that the incidence of iner degeneration fell with rising amounts of casein VI Apparently, the sulfur containing amino acids were not involved as causative factors in casein VI damage as they are when liver injury is developed on yeast diets, since the addition of cystine or methionine to the diet had no protective effect The methionine content of casein VI is prac tically the same as in the starting material The cystine content of casein is apparently too small to play an important role, nevertheless, we considered the possibility that the small amount of cystine present (04%) could be transformed into lanthronine during the alkaline treatment The addition of synthetic lanthionine to a diet sith puri fied, but not liver injuring casein did not induce Inet degeneration > 48 ·Taking all these points into consideration it seemed likely that no toxic agent was produced by the alkaline treatment but that a liver protecting factor was removed We were able to produce the casein VI liver injury as a matter of routine but a considerable variation occurred in the incidence of liver degeneration and in the time of survival in 35 experiments be tween 1941 and 1943 Regularly in September and October liver degeneration occurred in a smaller percentage of the animals and only after a longer time It has been mentioned as first stated by F S Daft WH Sebrell and R D Lillie (Proc Soc Exp Biol & Med 50 1 (1942)) (WH Sebrell Sixth Conf on Liver Injury Josiah Macy Jr Foundation p 71 (1917)) that liver degeneration has to be sharply separated from dietary liver cirrhosis in over 1300 animals dying from acute laer degeneration in our experiments not one was found to have a definite liver cirrhosis All our animals were given 1 milligram of chol ne chloride daily This was sufficient to protect against kidney hemor shage Larger daily amounts of choline (20 mg) did not influence the development of the casein VI injury We used casein VI liver damage between 1941 1943 as a method of routine to test the protecting activity of a large number of dif ferent materials and concentrates Several agents were found in nature which protected against the development of this type of liver degeneration Wheat germ was especially active in our prophylactic tests Using this as a starting material effects were made during 1941 1943 to isolate the active principle contained in this material The final con centrates contained 44 percent of a substance which was shonn to be vitamin [ I then found that synthetic dl alpha tocopherol acetate was active This was a surprising result because our animals had received a certain amount of vitamin E namely 50 mg per week as a matter of routine during all these experiments This was enough to protect against damage to the reproductive organs using normally purified casein but 17 or 18 times as much vitamin E was needed to protect against the development of liver damage in casein VI d ets After discovering the protective action of vitamin E upon the laer under the conditions of our dietary experiments I attempted to de termine whether other liver damages could also be influenced by vitamin E In the German veterinary literature an acute liver dystrophy has been described several times which arose in baby pigs when high levels of cod liver oil were fed (W Nikolaus Tierarztliche Rund schau 43 1 (1937) W Nikolaus Arch Tierheilkde 73 428 (1938) Tiedge Dtsch Tierarzt Wochenschr 45 132 (1937)) Incorpora 49 ·tion of 10 or 20 percent of cod liver oil into the scheme of our experimental diets — with highly purified, but not alkalıne treated casein — caused death of young rats after several weeks with marked liver damage There were gross and microscopic pathological differences between the casein VI damage and the cod liver oil damage Five milligram percent of dl alpha tocopherol acetate in the diet fully protected the animals against this disease The third type of liver damage, the so called yeast degeneration, was different in its origin as well as in its manifestations The injury occurred if large amounts of yeast were fed as the main protein source in synthetic diets This damage has been considered for a long time to be due only to the lack of cystine and methionine in yeast protein I found that vitamin E was quantitatively inhibiting In summacized form the three types of liver damage can be defined and differentiated as follows 1 Casein Vi Damage - This injury is produced by the incorpora tion of highly purified casein VI into synthetic diets The animals are in good physical condition and grow fairly well for several weeks They suddenly develop a peculiar coma, respiration becomes slower and deeper, the heart beat slackens, the body becomes cold and death ensues in a few hours The pathological picture is that of an acute liver degeneration The massive necrosis is localized in the central part of the lobules or at least starts there Hemorrhages in the liver are seen in almost every case Sometimes the major part of the organ 15 infiltrated with blood The degenerated parts show mostly a mod ecate fat infiltration Ceroid pigment but no fat is found in the endo thelium of the capillaries * Casein VI injury is not influenced by feeding sulfur containing amino acids The damage can be quantita tively inhibited by vitamin E A daily dose of about 130y of dI alpha tocopherol acetate is sufficient to protect most of the animals Five milligram percent of this substance in diets was protecting in every case Xanthine, which was found primarily to be active against chloro form damage (J C Forbes and J S McConnell, Proc Soc Exp Biol & Med 36, 359 (1937), R C Neale and H C Winter, J Pharmacol Exp Therapeut, 62, 127 (1938), H N Barett, D L McLean and E W McHenry Ibid 64 131 (1938), I C Forbes, Ibid 65, 287 (1939)), had a protective action The probability that there exists another protecting factor in crude casein has been outlined above I might mention that whey also has an inhibiting effect 1 * The histological examnations have been carried out by Prof Dr Dobberste o Tierarztliche Hochschule Berl n and Doz Dr Veltea Path Institut University of Heidelberg so ·could be seen during the coma The condition escally began at ca. so that many of the anumals were found dead 2 * mornig zu • compos on yeas 24 25€ 3 butte fat (extracted) scrode 638, sao 21 n sce K Shwa z 23h physiot be 281 103 2016 being observed The livers of anuals dying of the yeast injur 2 Cod Liver Oil Damage ---This injury is induced by diets con kaining normally purified casein and 10 to 20 percent of cod liver oil The animals die after several seeks The gross appearance as well as the histological picture of the livers is different from the casein VI damage No hemorrhages occur The livers are highly infiltrated with fat Microscopic examination reveals in many but not in all cases a degeneration which is evenly distributed over the entire lobule The severe fatty infltration exists in sharp contrast to the normal body fat depots which have disappeared almost completely The voluntary muscles of these animals are mostly degenerated Muscle degeneration is a well known finding in vitamin E deficiency The kidneys show a zone of calcification in the mtercorticomedullary region This is due to the hypervitaminosis D Cod liver oil damage to the liver and muscles is inhibited quantitatively by vitamin E The influence of vitamin E in this case might be explained by its well known action as an antioxidant It was first demonstrated by C. G Mackenzie J B Mackenzie and E V McCollum (Science 94 216 (1941)) by the work of H Dam (Proc Soc Exp Biol & Med 52 285 (1943) ) Mount Sinai Hospital 12 1021 (1946)) and by Mason and his group (I. F Filer Rumery and K E Mason First Conf on Biol Antioxidants Josiah Macy Jr Foundation (1916)) that certain features of cod liver oil intoxication have much in common with vitamin E deficiency symptoms and that this vitamın antagonizes the overburden ing of the body with cod liver ou It is not yet knowo other substances for example cystine methionine or xanthine are effective 3 Yeast damage (rat eclampsia) The injury which is caused by the administration of comparatively large amounts of seast as the source of amino acids in purked diets was first observed by A Hock and H Fink (Zschr z physiol chem 278 136 (1913) Ibid 279 187 (1944)) and by # P Himsworth and L E Gimn (Clin SCI S 93 (1944) Ibid's 133 (1944)) My aperurents on this subject have been carried out between 1912 and 1948 Young rats on diets contairung 24% yeast corresponding to 12 protein and 3% casein* died after several weeks with symu.cms which were similar to those in the other liver damage experiments . They lapsed into a coma followed in a few hours by death Hoteikt one remari able difference could be observed in a certain percerage convulsos very sunular macroscopically to those with caseur VI damage (onal 51 ·. areas of degeneration and other parts with hemorrhages were irregu larly distributed over the organ Normal tissue could hardly be found The color of the degenerated regions had a more pink tinge compared with the casein VI damage Histologically there existed a definite difference The degeneration in the casein VI injury is central, whereas the yeast injury begins in the periphery of the lobule A well developed fat infiltration is found not only in the liver cells themselves but also in star cells The kidneys of these animals show an alteration which has been defined by ) Dobberstein and A Hock (Ztsche physiol chem 280, 21 (1943)) as glomerulonephrosis These authors pointed out that the entire picture corresponds closely to the changes which can be found in human cases of eclampsia of pregnancy The term rat eclampsia' has been proposed for this type of dietary damage in the rat This proposition has been criticized by obstetricians who want to reserve the term for their own field regardless of the fact that eclampsia means any convulsive seizure of peripheral origin The term yeast injury is rather unsatisfactory because the same liver injury, or at least a very similar one, can be produced by several other dietary methods, 1 e by diets without yeast but with a similarly low content of cystine and methionine Diets with 10% or less casein cause liver damage Most of the classic experiments of Gyorgy and Goldblatt about liver injury were accomplished with this type of ration (P Gyorgy and H Goldblatt , J Exp Med 75, 355 (1942)) The regimen with 10% or less casein enabled these authors first to produce liver damage (necrosis and circhosis) regularly, and though necrosis and cirrhosis were not as clearly separated at this time it is quite obvious from their early work that yeast (of American origin) inhibits the development of necrosis The question, however, of whether this low casein damage is in all respects identical to the yeast injury — induced by European yeast — remains to be thoroughly investigated A Hock and H Fink (Ztsche Naturwiss 2b 203 (1947)) described several other diets containing 8 to 15 percent of protein in the form of potato protein, pea protein, cereal protein, and gelatin in various combinations which induced liver degeneration I should mention that these experiments were started as a result of discussion remarks of mine made in 1944 to the effect that we should consider the possibility that yeast might contain a substance with toxic effects The liver degenerations obtained by Hock and Fink could all be inhibited or at least beneficially influenced by cystine I think the fact that these damages can be produced by a variety of diets without yeast but poor in cystine and methionine, is an argument against the theory that yeast itself contains a toxic principle Yeast injury (rat eclampsia) was for a long time thought to be due 52 ·only to the defect of sulphur containing amino acids in the diet (H P Himsearth and O. Lindan, Nature 163, 30 (1949)) I found honever that 5 milligram percent of di alpha tocopherol acetate gave full protection against this damage in the same way as against the other types of liver damage described above * Rat eclampsia therefore cannot be considered to be due to a simple lack of sulfur-contaming amino acids It is a disease caused by the simultaneous absence of vitamin E on the one side and cystine methionine on the other Each of these substances alone is able to prevent the damage It is an injury which can arise only if two diverse, important, dietary com ponents are sufficiently supplied at the same time The term "am bogene" deficiency has been used to define this type of injury (K Schwarz, Ztscht physiol chem 283, 186 (1948)), which lately has become of growing interest in nutritional fesearch The phenomenon of an ambogene disease seems to indicate the existence of a close metabolic connection between the substances involved in our case the metabolic selation between vitamin E and sulfut containing amino acids is not yet known We find that one molecule of tocopherol has the same protective effect as 200 400 molecules of methionine or cystine This ratio might pechaps signify a catalyst substrate proportion The finding that at least some - American yeasts do not as casily induce liver injury in synthetic diets as European varieties is of considerable interest Dried yeast has been used more and more as a common foodstuff during the last few years in Europe In 1943 and 1944, we compared several German yeasts of different kinds and origins A quantitative difference but no fundamental distinction was found among brewers yeast (Loewenbraeu Muenchen), torula vesst grown on sulfit liquor (Zellstoft Cabrik Waldhof), and torula cults vated on a mixture of molasses and wood hydrolysate (Bergin A G Mannheim) All these induced liver degeneration, the animals from brewers yeast survived on the average longer and the inadence of Iner necrosis was smaller in this group We were unable to separate a toxic substance from yeast by thoroughly extracting with boiling water or fat solvents About 40% of the dry weight of yeast can be brought into solution by these means The residue of the extraction consists mostly of yeast protein When this residue was employed instead of yeast, liver degeneration sesuited to the same extent as with the untreated material The explanation for the quantitate differences among various yeasts, especially among European and American ones, might perhaps be found in the amount of sulfur containing amino acids present Most American yeasts are grown in a much richer medium than were the yeasts produced in Presented at the meeting on Protein July 1944 in the Plassenburg Kulmbach 33 ·Europe during the war The cystine and methionine content of Amer ican yeasts seems to be significantly higher In European materials cystine sometimes could not be detected at all This might be partly due to technical difficulties The exact determination of small amounts of cystme in material like yeast is problematic because cystine reacts during hydrolysis with carbohydrates and thus is destroyed It is obvious, however, that there is a definite difference between the amount of cystine and methionine present in American yeast and European yeast The content of organic sulfur in yeast is comparatively high It is sometimes higher than in casein, for example, and if casein is replaced by yeast protein in synthetic dicts the animals get more organic sulfur but nevertheless develop liver damage which can be prevented by sulfur containing amino acids This could support the toxicity theory if we assume that the organic sulfur in yeast is present mainly in the form of cystine and methionine However, the quantitative analysis of German yeast for methionine and cystine gave values which gen erally did not correspond to the total amount of organic sulfur present In some listances the deficit was as high as approximately 35 per cent of the total organic sulfur Therefore it seems possible that yeast contains a remarkable amount of organic bound sulfur which is not methionine or cystine The content of thiamine biotin, and adenyl thromethylpentose in yeast is too small to explain more than but a fess percent of this deficit We investigated the quantity of ergotho nine in yeast and found no cemarkable amount of this substance The question of toxic factors the toxicosis theory, has been dis cussed during the development of many brochemical problems especially those of medical application This was the case, for example, with the thyroid function, with the origin of pernicious anemia etc As far as different dietary liver injuries are concerned it can be stated that at least some of them, namely the low casein damage and the injuries caused by diets containing potato protein and gelatin etc are obviously not induced by special toxic agents We know on the other hand two forms of liver damage which are apparently caused by the presence of dietary constituents with deleterious effects the cod liver od damage and the laser degeneration produced by the administration of large amounts of cystine in rations This injury, induced by 5 or more percent of cystine in synthetic diets was the first dietary liver injury described (A C Curtis and L H Neuburgh Arch Int Med 29 828 (1927) N X Sullivan W C Hess and W H Sebrell Pub Health Rep USPHS 47 75 (1935) R D Lillie Jbid 47 83 (1935) D P Earle and ] Victor J Exp Med 73 161 (1941)) We must assume that cystine plays an important part in some metabolic reaction of the liver and that this function is so fundamental 54 ·that the liver cell degenerates if the reaction fails In view of this it is not difficult to reconcile the tro facts that on the one hand a small amount of cystine protects agausst certain liver degenerations while on the other hand a large amount of this substance causes the same The metabolic reactions in which cystine is involved are enzyme reac trons and cystine is the substrate It is quite a common and well known feature of many enzymes that they are inhibited by a surplus of substrate This would only be one of different possible ways to under stand the phenomenon In metabolısın ne always find a certain range within which every normal nutrient is harmless Iven sugar ss lajurious if this margin is exceeded The tolerance for normal dietary constituents often is te markably displaced in deficiency diseases if for example a rat is vitamin B, deficient a relatively small dose of sugar has a fatal effect It is debatable whether we should apply the terms toxic, toxine and intoxication in this case just as it seems questionable whether we should use them with the cod liver orl injury or the effect of large doses of cystine Perhaps it is better to define and classify these condi trons simply as dietary stress especially if we have no proof as is the case with the yeast injury and the casein VI damage w whether a toxic component seally is involved or not the more so since the term intoxication has many well established pharmacological and metabolic implications These might not be in accord with the character istic picture of metabolic pressure and final breakdown of cell func tions as we see in the different forms of dietary Iner degeneration DISCUSSION Best May tve have a general discussion at this point? Gyorgy I would like to make a few comments First of all I do not wish to give the impression that I assume a direct toxic factor in yeast I did not say there was one What I said was that the yeast might lead to the production of a toxic factor Where it originates I dont know but most probably through the interaction of the intestinal flora As previously stated I do not have the analytical figures for cystine and methionine in the British yeast However there are several des crepancies in the observations which make it difficult to explain the necrogenic effect of British yeast compared with the American yeast by the lower content of British yeast in sulfur-containing amino acids. + In accord v th the shes of the Cha man and the part peats of the Conference Dt Schwarz has prepared h s report a greater detal than he was able to preseat at the toe of the mee ng Consequently some of his data is not ava fable for commeat or dscussion by the goup --- Etor 35 ·ing, then Be that as it may, I do not want to postulate the presence of a direct toxic substance in yeast Schuarz I think it is very difficult to differentiate between intoxica tron and the lack of an important substance in the organ, especially of a substance which is working catalytically If such a substance is lack you will have, perhaps, an intoxication of something which is not changed because it is lacking The most famous example of that is vitamin B, and pure pyruvic acid Best You might have an enzyme system intoxicated by a poison and the same enzyme system completely inactivated by the lack of a vital factor It would be the same mechanism Schwarz That is what I wanted to say Patek From the remarks of Dr Gyorgy it seems that the old concept of autointoxication is not so fanciful I wonder whether it would be feasible to feed some animals a substance which would inhibit bacterial flora, to see if this would influence the production of necrosis Gyorgy In cooperation with Dr Joseph Stokes, Jr we are studying at present the influence of aureomycin, mainly as an intestinal anti septic, given by mouth, in a group of rats receiving the severely necro genic ration containing the British yeast, compared with a control group receiving the same diet without aureomycin The results will be forthcoming during the next three months * Best Dr Hoffbauer do you have any comments on this subject? Hofbauer What I have to say will not require much time At the conference last January, D. Best raised the question as to whether or not anyone had studied the porphyrin excretion in animals in which dietary liver injury was produced in our laboratory we have set about to do that As one aspect we were anxious to study animals in whom massive hepatic necrosis was produced Dr Gyorgy was kind enough to show me how to set about this problem and provided me with a supply of yeast from the United Yeast Company in Great Britain Since January we have observed the production of massive hepatic necrosis in 14 of 22 rats which were started on the diet This occurred within 35 to 80 days Other animals which we have sacrificed at intervals have not shown the lesion A few still remain alive Estima trons af wary and fecal coproporphyrın have been carried out by Dr Greenberg The estimations have been made at varying intervals We had hoped that we might be able to predict the onset of massive hepatic necrosis by an increase in the excretion of this pigment coproporphyrin which can be readily recognized by its fluorescent qualities + Added to the proof Aureomycia exerted a significant protective or at least a delaying effect on the development of this type of hepatic necto 3 ·In the paper reported by L E Glynn, H P Hemsworth and A Neuberger (Bsit ] Exp Path 26, 326 (1945)) mention is made that Professor Remington studied a few samples of unne from suarlar tats and failed to find an increase in coproporphyrn More recently C E Dent and C Remington (Biochem) 41, 253 (1947)) reported that animals mautained on allaline treated casein dset showed an increase of coproporphyrın We have not found, in any animal, a significant increase in either the urinary or fecal coproporphyrin These observations have not been carried out contmually but in various times during the course of the experiment In two animals we were able to obtain urine specimens for the two days immediately preceding death There was no siguficant increase in these animals We have definitely encountered lesions which are identical with what Dr Gyorgy his shown you We feel reasonably certain that, in the group of rats on the diet containing British yeast as the sole source of protein, massive necrosis of the liver has developed and yet there has been no significant increase in the urmary coproporphyrin excretion That may possibly be taken as some indirect evidence that we are not dealing with the usual type of chemical poisoning The common hepatotoxins usually result in an increase in the porphyrin excretion in the experimental animal Gjorgy This has been demonstrated in rats> #'atson With sulfanilamide, by C Remington and A W Hemmings (Lancet 1, 770 (1938)) Gyorgy That does not produce any necrosis as chlorofom and carbon tetrachloride Hofbauer Carboa tetrachloride, if not given in such massive doses that it damages the kidney appears to cause an increase u urinary coproporphyrin excretion We must also keep the factor of renal damage in mind 10 urinary pigment studies If the kidney is damaged the material may not appear even though it should be expected to on the basis of liver injury I believe the rat will respond to some of the porphyrinogenic materials by an increase in coproporphyrin in the Gyorgy Is there not a very acute onset of this type of ajury? Husn orth has noted this and I can confirm i The rats are running around and then suddenly they are dead Hofbauer It has been amazing to us As I perhaps mentioned thus afternoon one of the animals we had just removed from the metabolism cage used to collect the specimens seemed perfectly normal in all respects I was quite astounded to have the man in the laboratory call the up two hours later and say that the animal was dead I had no premonition that it as going to die urine 57 ·Be that as it may, I do not want to postulate the presence of a direct toxic substance in yeast Schwarz I think it is very difficult to differentiate between intoxication and the lack of an important substance in the organ, especially of a substance which is working catalytically If such a substance is lack ing then you will have, perhaps, an intoxication of something which is not changed because it is lacking The most famous example of that is vitamin B, and pure pyruvic acid Best You might have an enzyme system intoxicated by a poison and the same enzyme system completely inactivated by the lack of a vital factor It would be the same mechanism Schwarz That is what I wanted to say Patek From the remarks of Dr Gyorgy it seems that the old concept of autointoxication is not so fanciful I wonder whether it would be feasible to feed some animals a substance which would inhibit bacterial flora, to see if this would influence the production of necrosis Gyorgy In cooperation with Dr Joseph Stokes, Jr, we are studying at present the influence of aureomyan, mainly as an intestinal anti septic, given by mouth, in a group of rats receiving the severely necro genic ration containing the British yeast, compared with a control group, receiving the same diet without aureomycin The results will be forthcoming during the next three months * Best Dr Hoffbauer, do you have any comments on this subject? Hofbauer What I have to say will not require much time at the conference last January, Dr Best raised the question as to whether or not anyone had studied the porphyrin excretion in animals in which dietary liver injury was produced in our laboratory we have set about to do that As one aspect we were anxious to study animals in whom massive hepatic necrosis was produced Dr Gyorgy was kind enough to show me how to set about this problem and provided me with a supply of yeast from the United Yeast Company in Great Britain Since January we have observed the production of massive hepatic necrosis in 14 of 22 rats which were started on the diet This occurred within 35 to 80 days Other animals which we have sacrificed at intervals have not shown the lesion A few still remain alive Estima tions of urinary and fecal coproporphyrın have been carried out by Dr Greenberg The estimations have been made at varying intervals We had hoped that we might be able to predict the onset of massive hepatic necrosis by an increase in the excretion of this pigment coproporphyrin which can be readily recognized by its fluorescent qualities + Added to the proof Aureomycia exerted a significant protective or at least a delaying effect on the development of this type of hepatic necrosis-P Gyorgy 56 ·In the paper reported by LE Glynn, H P Himsworth and A Neuberger (Brit J Exp Path 26, 326 (1945)) mention is made that Professor Remington studied a few samples of urine from similar rats and failed to find an increase in coproporphyrın More recently C E Dent and C. Rimington (Biochem ) 41, 253 (1947)) reported that animals maintained on alkaline treated casein diet showed an increase of coproporphyrn We have not found, in any animal, a significant increase in either the urinary or fecal coproporphyrin These observations have not been carried out continually but in various times during the course of the experiment In two animals we were able to obtain urine specimens for the two days immediately preceding death There was no significant increase in these animals We have definitely encountered lesions which are identical with what Dr Gyorgy has shown you We feel reasonably certain that, in the group of rats on the diet containing British yeast as the sole source of protein, massive necrosis of the liver has developed and yet there has been no significant increase in the urinary coproporphyrın excretion That may possibly be taken as some indirect evidence that we are not dealing with the usual type of chemical poisoning The common hepatotoxins usually result in an increase in the porphyrin excretion in the experimental animal Gjorgy This has been demonstrated in rats? Watson With sulfanilamide, by C Remington and A W Hernmings (Lancet 1, 770 (1938)) Gjorgy That does not produce any necrosis as chloroform and carbon tetrachloride Hofbauer Carbon tetrachloride, if not given in such massive doses that it damages the kidney appears to cause an increase in urinary coproporphyrin excretion We must also keep the factor of renal damage in mind in urinary pigment studies If the kidney is damaged the material may not appear even though it should be expected to on the basis of liver injury I believe the rat will respond to some of the porphycnogenic materials by an increase in coproporphyrın in the Gyorgy Is there not very acute onset of this type of injury? Himsa orth has noted this and I can confirm it The rats are running around and then suddenly they are dead Hofbauer It has been amazıng to us As I perhaps mentioned this afternoon one of the animals we had just removed from the metabolism cage used to collect the specunens seemed perfectly normal in all respects I sas quite astounded to have the man in the laboratory call me up hvo hours later and say that the animal was dead I had no premonition that it was going to die urine 57 ·Watson I think that perhaps fits with the fact that you have not seen any increase in porphyrin at all even in the 48 hours before death If that laser had been appreciably damaged, I think you sould have found some increase It seems to me this must happen very abruptly I think the well being of the animal and lack of increase of porphyrın at least strongly suggests that Best 'The liver damage does not kill the rat Something else may be acting quite independently Watson I have asked Dr Hoffbauer to find out what the blood sugar is Gjorgy Nobody has done it Best You are not going to be able to do it You have extensive liver damage obviously Schu arz I might mention that it may perhaps be a very quickly developing intoxication I have been able to transfer the coma of the dying rat to mice by taking the blood of the animals that were just going to die I could make the mice die This was done with the animals of the casein VI group and the yeast group I did not try It with the cod liver oil animals because I did not have an opportunity It is very difficult to get these animals because most of them die at ought and the coma develops so quickly that you have only two or three hours of coma and then they are dead I have tried to find liver damage before they died I killed them after one neel, two weeks, and so on, and I did not find any liver damage I did find the liver damage only when the animals were going to die G701g) The last til o or three hours) Schuarz That is quite true Hofbauer Did I understand you correctly that blood from these animals, just at the time of death, 35 lethal for mice? Schuarz Not whole blood I prepared it Hoffbauer You ground up the liver? Schu arz I took only the blood and made a filtrate, I deproteinized the blood and injected this intraperitoneally into mice, very small mice, and they died Annstrong What did you use as a protein precipitant? Schuarz I prepared it by heating I did several other experiments, using I belies e phosphotungstic acid, by reprecipitating with the phos photungstic acid and by reprecipitating with sulfuric acid it was quite clear no other agent was in the filtrate Armstrong It seems to me as I heard this discussion that this question 58 ·may be reminiscent of the unsaturated essential fatty acids antioxidant story I should like to ask Dr Gyorgy if he has seen any evidence --- he has told us about the liver of scales on the tails Do the animals have hematuria ? Is there any evidence of unsaturated acid deficiency Gjorgy All have unsaturated fatty acids They all receive extra unsaturated fatty acid Schwarz My animals did have linoleate Gyorgy There were no tail changes, no kidney changes, no skin changes Goldblatt I should like to ask Dr Schwarz one question and that is just what is he referring to when he speaks of three types of liver injury? Can you, Dr Schr arz, translate that in terms of morphology, or are you speaking of some kind of a functional liver injury Schwarz I would say that you can differentiate beta een these three types of liver injury histologically One, the casein VI injury, is more central, the other, the yeast injury, is more on the periphery The yeast type is connected with typical changes in the kidney The kidneys show 'glomerulonephrosis The whole syndrome is very similar to the syndrome which you can see in human eclampsia The cod lever oil damage is distinguished by its heavy fatty in filtration of the liver cells, and is distributed homogeneously over the liver Thus, I believe one is able to differentiate all of these three types of damage Best If you let these go on, does not the peripheral go all through the lobule, and the central if you follow it does it not become general as well? Schuarz You can still differentiate it if you look over the liver, you may see several lobules uniformly invoked but then you see others where the degeneration is just beginning and there you see it is be ginning centrally or peripherally Best Both Dr Gyorgy and I thought immediately of vitamin B, It seems to be fairly obvious and we are anxious how to get a supply of alpha globulin from soybeans which is apparently completely free of vitamin B,z, because our diets have been full of this It is lipotropic It may spare choline Schu arz There is a possibility that vitamin B 2 may be involved That would give another factor which is involved in all these relations That might form an explanation, perhaps, for the differences between the several groups of authors too One is using crude casein and the other is using highly purified casein which is free of vitamin B,, 39 ·IV SECTION RECENT FINDINGS CONCERNING THE ROLE OF THE LIVER AND KIDNEY IN CIRCULATORY HOMEOSTASISI 2 Shorr Dr Knisely has made it abundantly clear from his studies on the liver, which he has just reported, that the regulation of the circula tion in that organ is a very precise and extremely delicate process The glimpse he gave us of the structure and function of the vascular bed of the liver would indicate that its complicated vascular architecture and the subtle changes which take place in the behavior of the blood vessels were not haphazard and that they were undoubtedly related to the homeostatic regulation of the many functions which this orgar serves This concept of homeostasis has come to exert a major influence in our thinking about bodily processes The stability of the internal en vironment is secured through the action of three major organ systems, the kidneys, lungs and the capillary bed Without detracting from the importance of the first two, the final homeostatic adjustment, which maintains a stable internal environment for the tissues, can be attributed o the function of the capillary bed It is for this final stage in the homeostatic process that the function of the lungs and kidneys is in the last analysis conditioned For that reason we have been particularly interested in studying the factors that regulate the function of the capillary bed Before discussing the role of the liver in this process it would be profitable to examine the present concepts as to the structure and function of this unit The original studies of Krogh have been recently extended by B W Zweifach R Chambers and their associates and the description that I am about to give is based on the observations of this latter group of investigators (Am | Anat 75 173 (1944)) Although the capillary bed may differ in some details from tissue to tissue, its basic structure appears to be fundamentally the same in all The archutectural unit consists of a thoroughfare channel the metarteriole which provides a direct connection between the arterial and venous sides of the central vascular tree The metarteriole is discontinuously surrounded by smooth muscle cells which are most numerous in the proximal and absent in the distal portion shich enters the venous channels All along its 1 Fphraim Short BW Zweifach R F Furchgott A Mazur S Baez and M A Payas Department of Medicine Cornell Laivers ty Medical College and The New York Hosp tal 2 These stu ie were supported by grants from The Josiah Macy Jr Foundation, Eli Lilly and Compant The Postley Hypertension Fund and The United States Public Health Service 60 ·length the metarterioles give rise to precapillary sphincters At the proximal end of the metarteriole these sphincters emerge at acute angles These angles become progressively less acute distally Each precapıllary is surrounded by a well defined ring of smooth muscle cells which acts as a very competent sphincter The precapillary sphinc ters in tuen give rise to the true endothelial capillaries through which Aurd exchange takes place The behavior of this architectural unit is characterized by periodic caliber changes of the metarterioles and precapıllary sphincters This intermittent relaxation and construction is termed vasomotion The furd exchange through the true capillaries is profoundly influenced by this vasomotion through its effects in altering the ratio of hydrostatic to osmotic pressure These muscular vessels of the capillary bed also undergo Auctuations in their reactivity which can be conveniently meas ured by their responsiveness to the topical application of epinephrine Arlor ola Wnuta Normol enculat on (Very olara) Ischemic rulat on Eh preden notes) Hyperemic crculotion (VDM predom nyes) FIGURE ! Diagrammatic representation of alterations in blood flow through capillary bed associated with predominance of VEN or VDM in the circulation Note ischemic circulation during VEM predominance resulting from constric tion of precapillary sphincters with restriction of blood flow to main thoroughfare channel (the metarteriole) Note also the overall capillary arculation during VDM predominance as a result of relaxation of precapiltary sphincters I have illustrated in Figure 1 the three typical states of the capillary bed which result from alterations in vasomotion and reactivity under the influence of the vasotropic substances VEM and VDM with which we have been concerned in the normal state when neither predominates there is an intermittent flow through the capillary bed resulting from the opening up of now one and now another precapillary sphincter There is a continuous flow through the metarterioles if, however conditions are set up which lead to increased vasomotion and increased 61 ·ascular hyperreactivity, this picture is profoundly altered Such a ondition is seen in the hyperreactive phase of hemorrhagic shoch in ssociation with the appearance in blood of the renal vasoexcitor, VEM (E Shoir, B W Zweifach, and R F Furchgott, Science 102, 489 (1945)) Vasomotion is increased and the constrictor phase of precapillary sphincteric activity prolonged at the expense of the dilator phase As a result of this augmented precapillary sphincteric constriction, blood flow through the true capillaries is profoundly reduced, the major portion of the blood flowing through the metarteriolar or thoroughfare channels directly from the arterial to the venous side If, however, the hepatic vasodepressor, VDM, predominates, as during the hyporeactive stage of shock, vasomotion is depressed, the vessels become hyporeactive and quite another picture is seen The dilator phase of the precapıllary vasomotion is now accentuated, the sphincters remain chiefly in the relaxed state and the capillary bed exhibits an overall flow which is in sharp contrast to the ischemic state previously described We are now in a position to consider the consequences of the prevalence of each of these types of vascular behavior on fluid exchange within the capillary bed I have represented these different situations in very diagrammatic manner in Figure 2 In this figure, which shows HYPO-REACTIVE (VOM) NORMAL HYPER-REACTIVE (VEM) PRECAPILLARY SPH NCTER ARTERIOLLA 0 CAP BCD O O VENULE CAPILLARY FLOW CAPILLARY FLOW CAPILLARY FLOW INTERMITTENT ISCHEMIC OVERALL FIGURE 2 Diagrammatic representation of alterations in blood flow and fluid exchange in the capillary bed during VEM and VDM predominance. The number of red blood cells represents the number of open endothelial capillaries The arrows represent the direction of the flow of fluid 62 ·a metarteriole, a precapıllary sphincter and one endothelial capillary vessel, the number of open capillaries is indicated by the number of red blood cells which are drann in the true capillaries, and the arrows indicate the movement of Auid in or out of the capillary bed Tuming first to the normal state, we observe that the movement of fluid through the metarterioles is outward where the pressure is highest 1e, at the proximal end, and the outward filtration decreases as the hydrostatic pressure is reduced by progressive resistance to the column of blood in its passage through this vessel in the true capillaries, hydrostatic pressure relationships are such as to make for outward filtration at the proximal end and for insard filtration at the distal end as the hydrostatic pressure is dissipated Attention should be called also to the action of the metarterioles, in favoring inward filtration at their distal ends This is a consequence of the wavelike progression of the smooth muscle contraction along the metarterioles The resultant rapid flow of blood through the metarterioles acts like a water pump to suck in fluid from surrounding tissues and tributary capillaries When, as in hyperreactive shock, vasomotion is increased, and the precapillaries are largely in the constricted phase with a correspond ingly reduced flow through the true capillaries, (as indicated by the smaller number of red blood cells in the diagram) all the factors making for inward filtration are favored 'The result is tissue ischemia, dehydration and hemodilution If, however the opposite condition of vascular hyporeactivity exists with sphincteric relaxation and an overall flow through the now open capillary bed then hydrostatic pressure predominates over a greater length of the capillary bed Outward filtration and edema formation are favored It is on this portion of the vascular bed, which is so essential for the maintenance of a stable internal environment for the cells, that the hepatorenal principles with which our studies have been concerned, act Their action is ordinarily of a chronic nature It appears to consist in subtle adjustments to the fluctuating needs of the tissues When the integrity of the circulation is endangered i e, after blood loss when the balance is precarious, it is believed that these humoral agents can produce acute changes of a beneficial or a deleterious nature When they participate in chronic circulatory disturbances, the predominance of one or the other type of capillary behavior calls forth compensatory vascular adjustments on the part of the central vascular bed It is in this way we believe, that these humoral agents can bring about their long term effects on blood pressure When these adjustments on the part of the central vascular bed to the chronic type of behavior exhibited by the capillary bed fail to occur then circulatory homeostasis breaks down and situations such as edema formation and peripheral circula tory failure arise 63 ·These vasotropic principles which participate in the regulation of capillary function comprise a vasoexcitor, VEM, which has its origen in the kidney and chiefly in the cortical portion, and a vasodepressor, VDM, originating in liver, spleen and skeletal muscle The vascular effects of VDM from these different tissues are indistinguishable These principles are quantitated by their effects in enhancng or depressing the response of the terminal vascular bed of normal anesthetized rats to topically applied epinephrine In this test, which was devised by Zweifach and Chambers (B W Zweifach, Afethods in Medical Research, edited by V R Potter, Year Book Publishers, Chicago, 1, 131 (1948)), the normal rat mesoappendix is exteriorized and the reactions of the vessels directly visualized before and after the injection of a test sample The conditions have been ascertained for their formation in zitro and under certain circumstances in ino In tutro experiments have shown that both VEM and VDM are normally formed by their tissues of origin only under anaerobic conditions They fail to appear under aerobic conditions Furthermore, under aerobic conditions normal liver inactivates VDM and normal kidney inactivates VEM In addition to establishing their tissues of origin and the conditions under which these factors are formed and inactivated we have succeeded in isolating and identifying the liver vasodepressor VDM (A Mazur and E Shorr, / Brol Chem 176 771 (1948)) This has been found to be ferritin or apoferritin, the iron free moiety of ferritin Ferritin is an interesting iron protein complex first described by Laufberger in 1937 and subsequently studied by Granıck and Michaelis in relation to its function in iron storage and transport It is a protein of a molecular weight of about 450 000 and exists in two forms ie, with and without iron The iron is not present as a porphyrın, but as fertic hydroxide The amount of iron is large constituting as much as 20% of the molecule by weight It is tightly bound within the molecule and can only be remov by strong reducing agents Both crystalline ferritin and apoferritin have similar and equal vasodepressor effects Hence, the vasodepressor action would seem to be related to the protein moiety rather than to the iron Profound vasodepressor effects are induced in the test rat weighing 100 125 gms with either substance in amounts of 0 0005 gamma of nitrogen Purification of VEM has not proceeded to the same extent Preparations have been obtained which are active in the rat test in amounts of approximately 0 0004 gamma of nitrogen It would appear to be a protein of relatively low molecular weight We have followed the appearance of these principles in a variety of pathological circulatory states with results which I shall notv describe 64 ·inefficiently returned to the active pusculation because of the reduced Eus is 1 teached It was further shown that bloud botne fict is are Tesponsible in Farge measure for these vascular episodes by the passe transference of corresponding effects to similar vessels in the mes arrendar of normal uts following the intravenous in ot biod Samples removed during the evolution of the sex sindr the Pror Soc Exper Biol & Med 36 127 (1944) R Chambers BV Zweifach B E Louenstem and R 1 Ice istas responsible for the occurrence of these humiril poinciples du The studes n our laboratory were designed to explore the 11, 12 the site of origin in specific tissues as well as the ent mental and experimental shock and to set up ppropriate (nditions for rent sling cellular factors concerned in their formation und in activ. The brst major arculatory disturbance which was studied in relation to these bepatorenal factors s.as experimental hemorrhagic and tru. mquc shock. Our studies were stimulated by the observation of Zscifach, Chambers and their associates, ustag 13scular reactions of the exteriorz omentum and mesentery as an index, that the shock syndrome, in anesthetzel anunals subjected to prolonged hemorrhagic hypotension, groceeds in two stages an mual compensatory and a subsequent decompensatory stage (B W Zseifach, R I Lec, C Hyman, and R. Chinsbers Inn Surg 120 232 (1944)) The initial compensatory phase was characterized by a hyperreactive condition of the terminal risculus bed. Thus hyperreactivity is manifested by an increase in the maternuttent constrictor activity (vasomotion) of the metastesioles and precapillaries The capillary ischemia shich results from the accentua Ast of the constrictor phase of vasomotion and the consequent restric tron of blood flow in the capillary bed to the thoroughfare channels the metarterioles serve to maintain an active venous return of blood from the tissues despite a reduced blood volume The ability to respond sat sfactonly to transfusion scems to depend upon the maintenance of this type of peripheral vascular activity Hosever, with the onset and prolongabon of more drastic hypotension, vascular changes gradually appear in the splanchnk areas visualized which are antagonistic to the compensatory vascular adjustments prevailing in the initial stage und which eventually disrupt the peripheral arculation During the decom pensaloty, stage there is a progressive reduction of cpinephrine leil urity and a slowing and finally complete suspension of spontanelus Faxrustzon in the metasteriales and precapillaries Hith a resultant te'acation of precapillary sphincters The predominance of de um pensatory 2ndvences on the periperal circulation results in an increasing Server ugen of blood into the capillary side channels from which it is blood pressure. Once dous decompensatory hyporeacrae phase bus Fully developed a state of urreversibility or failure to respond to trains (This ·required a correlation of in titro and in 1110 phenomena which was made possible by the rat mesoappendix test Experimental hemorrhagic or traumatic shock was induced in dogs and rabbits under pentobarbital anesthesia and the syndrome interrupted during the hyper- and hyporeactive phases Tissues were removed, appropriately sliced as for inicrorespiration studies, and extracted with saline The salıne extracts were then assayed for their vasotropic activity It was found that the vasoexcitor, VEM, which predominates in blood during the initial compensatory phase of shock had its origin in the kidney and chiefly in the cortex The vasodepressor, VDM, which predominates during the decompensatory hyporeactive phase suas found to arise in the liver, skeletal muscle and spleen During hemorrhagic shock the major role is played by VDM of hepatic origin During traumatic shock there is in addition a very considerable muscle VDM component This latter circumstance is believed to be responsible in large measure for the lessened tolerance to the equivalent degrees of hypotension of animals in traumatic as compared with those in hemorrhagic shock The mechanism responsible for the sequential appearance of these vasotropic factors appears to be the progressive tissue anoxia characteristic of shock The earlier appearance of VEM is ascribed to the rapidity with which the renal blood flow is curtailed following a reduction in blood volume Although during the initial phase of shock there is a parallel reduction in blood flow to the liver consequent on moderate hypotension, this does not appear to be sufficient to initiate anaerobic processes in this organ and no VDM formation takes place Apparently the blood flon to the liver is normally in excess of its oxidative require ments as shown by the fact that the blood flow can be restricted to the hepatıc artery with a maintenance of oxidative conditions (F L Engel, HC Harrison, and C N H Long, 1 Exper Med 79, 9 (1944)) Further support for this concept is provided by the work of Van Slyke and associates (D D Van Slyke, Ann New York Acad Sc 49, 593 (1948)) and Zweifach, Hershey, et al (B. W Zx eifach, S G Hershey, È A Rovenstine, R E Lee, and R Chambers, Surgery 18, 48 (1945)) who observed no rise in blood uric acid or ammonia with hemorrhagic hypotension of the moderate degree which leads to the development of peripheral vascular hyperreactivity Hosever, following the induction of a more drastic hypotension which results in vascular decompensatory hyporeactivity a more severe curtailment of blood flow occurs in both liver and kidney Kidney blood flow is virtually abolished and the release of VEM into the circulation thereby prevented The liver now becomes hypoxic and VDM formation is initiated to continue throughout the rest of the syndrome It is under these conditions of more drastic hypotension that Van Slyke et al, and Zweifach et al, 66 ·observed rise in blood unic acid, a constituent which is maintained at normal levels in the blood by oxidative processes in the liver Once VDM predominance and vascular hz porcactivity have persisted from 90 120 minutes the animal becomes sireversible to transfusion The question arose as to the reason for the failure of transfusions sufficient to restore oxidative conditions, at least temporarily within the organism The explanation for this failure is believed to reside in the progressive impairment of the mechanism by means of which the normal liver oxidatively inactivates VDM Studies of the liver, removed at intervals during the progression of the shock syndrome, showed that the VDM inactivation mechanism remained intact during the hyperreactive or reversible stage However, during the hyporeactive stage there occurred a progressive deterioration to total loss of thus hepatic inactivation mechanism (E. Short, B W Zweifach, and RF Furchgott Ann Neu York Acad Sc 47 571 (1948)) As a result, VDAI could not be destroyed on the restoration of oxidative conditions and indeed VDM formation actually took place under aerobic conditions in talo (R F Furchgott, B V Zweifach, and E Sharr Federnou Pror 8 201 (1949)) An additional factor of in the VDM inactivation system could not be restored This is shonn by the experiments of H A Frank A Seligman, and Fine ( Cim julestigation 25 22 (1946)) who found that viviperfusion of the liver via the splanchnic vein was ineffective of the perfusions were begun after the hyporeactive state had developed Other experiments have been carried out which have directly related the Inter VDM mechanisms to the capacity of the animal to recovec from traumatic procedures Shock as produced in rats by means of the Noble Collip drum (Shore B W Zuerfach and R F Furchgott, Science 102 489 (1945) and E Shorr and B W Zweifach unpub. lished data) 650 rotations of the drum are lethal for about 85ch of normal animals Honever it is possible to build up resistance by train ing Sublethal drummings are given and the num rotitions progressively increased in a short time the rats are able to tolerate 1000 rotations without fatality Experiments were carried out on the liver VDM mechanisms in pormal and estant rats There was a striking difference between the two groups the normal rats formed more Iner VDM during 500 to 700 sotations than did the resistant rats after 800 revolutions Livers removed from normal cats after drunming Kere less effective in VDM inactivation than livers removed after drumming resistant rats It was further found that the livers 67 ·Hepato-Renol Vasotropic Factors in Hemorrhagic Shock Houna 6 sod penrobarbital anesthesia XOF 20 BY mm Hg 404 Bleeding al35% body wk ul127 body wt. Terminal voacular bed VING! +++++ VOM VEM VEM formation in kidney 0 VDM formation in liver +++++ 0 Capacity of liver VDM inactivation system 2007 -Reversible to transfusion Irreversible phase FIGURE 3 of the resistant rats were able in most cases to tolerate a period of anoxia of 2 hours 11 litro and still preserve the capacity to inactivate VDM In contrast, livers from normal rats, like lıvers from normal dogs and rabbits, suffer profound deterioration of the VDM inactiva. tion system after a similar period of anoxia (Fig 3) From these and other correlative experiments a concept of shock has evoked which attributes a major role to these vasoactive principles of senal, hepatic, and skeletal muscle origins The initial reaction to hemorrhage, or extravasation of blood or Aud into traumatized limbs, leads to a reduction in blood volume and to adaptive vasoconstriction aluch involves splunchnic viscera as well as peripheral tissues The reduction in renal blood fion is profound and initiates the formation and release of VENÍ into the arculation This factor leads to the development of a specific kind of hyperreactivity in the termmal vascular bed which serves to restrict the flow in the true capillaries, send the blood through direct thoroughfare channels from the arterial to the venous side and thus maintain an effecise Cuculation in the face of a reduced blood solume As long as VEM predominance and vascular 68 ·hyperreactivity prevail the animal is responsive to transfusions Should more drastic hypotension ensue, there is a further restriction of blood flow to the splanchnic viscera which leads to hepatic hypoxia VDM formation and its release into the general arculation VDM now overwhelms VEM whose further supply is shut off by the virtual abolinon of renal blood flow As a result of the action of VDM hyperreactivity is replaced by peripheral vascular hyporeactivity of a decompensatory character The precapıllary sphincters are relaxed there 1$ an overall capillary blood flow and a progressive diversion of the reduced blood volume into an ever expanding capillary bed Through the local action of VDM upon the vessels of the splanchnic viscera total loss of the hepatic VDA inactivation mechanısın As a result the temporary restoranon of oxidative conditions by transfusions 15 ineffective in liberating the vascular bed from the decompensatory effects of VDN Transfusions exert only temporary beneficial effect on blood pressure and then are lost from the effective circulation by their sequestration in the splanchnic viscera The potential capacity of these abdominal viscesa to pool blood as a consequence of a local action of VDM on their vascular bed is believed to be quantitatively adequate to explain the ineffectiveness of blood replacement therapy at this stage From these studies on shock we then turned to an exploration of the possibility that the hepatorenal factors with which we are concerned experimental 120 (1948) Tr Am Physicians 60 28 (1947)) We produced experimental renal hyper clamp and assayed thin a few minutes was found in the renal vein blood and within a few hours in the peripheral circulation It was observed to be present in blood throughout the period of acute hypertension but when the ood pressure leveied out at chronic hypertensive values the blood severted to a neutral section At this point it seemed as if we were encountering a situation similar to that of renin a hich is present in measurable amounts in blood during the acute stage of renal hypertension but becomes undetectable when the chronic stage is reached However the development of procedures by which VEM and VDM could be fractionated in mixtures soon showed us that this is not the case (E Short and B W Zweifach federation Proc 8 146 (1949)) I shall describe these procedures briefly I have 1 69 ·already pointed out that normal kidney cortex can inactivate VEM on aerobic incubation in 1itro This property has been taken advantage of to wipe out VEM in mixtures of VEM and VDM Incubation of such a blood sample with slices of normal kidney for one hour in oxygen at 375°C serves to inactivate the VEM present and permit VDM to exert its effect unopposed in the test rat We have also prepared antisera to crystalline ferritin by its administration to rabbits (Å Mazur and E Short, Federation Proc 8 228 (1949)) The incubation of a blood sample containing VEM and VDM with antiferritin serum inactivates VDK and unmasks VEM Utilizing these procedures for the assay of blood from renal hypertensive animals throughout the evolution of the syndrome, we found that the neutral reaction to which the blood reverted during the chronic stage of hypertension was due to the progressive release of VDM by the liver until its concentration in the blood was such as neutralize the VEM which was being discharged by the kidney If one then follows the hypertensive animal for months and indeed years, during the chronic phase of the disease, one finds the invariable presence of both factors usually in a neutralizing ratio but occasionally with a temporary predominance of VEM or VDM In an occasional animal which has gone into the malignant phase we have obsen ed a resumption of VENT predominance It was now of interest to examine the mechanisms underlying the alterations in these humoral factors which followed partial constriction of the renal arteries The normal kidney never forms VEM under oxidative conditions but only during anaerobiosis The kidney in renal hypertension shows a very striking difference Within a few hours after the application of the Goldblatt clamp a metabolic alteration occurs as a result of which VEM is now formed under both aerobic and anaerobic conditions This inability to restrict VEM formation to anaerobiosis represents a critical metabolic defect which persists throughout the course of the disease Although its initiation probably is due to the renal hypoxia following partial constriction of the renal artery it is doubtful whether this explanation will hold for its per sistence inasmuch as there is no good evidence of persistent renal hypoxia following circulatory adjustments to the renal artery construc tion The oxygen consumption of these clamped kidneys in 11tro may remain within the normal range for the dog We have also studied renal hypertension in the rat and essential hypertension in man by the same methods in the rat the same defect appears in the kidney as was found in the dog and the humoral factors are present in similar fashion But we found as well a very specific alteration in the mesenteric vascular bed in renal hypertension in the rat (B X Zweifach S Rosenfeld and E Shorr Federation Proc 7, 70 ·139 (1948)). There is first of all an initial temporary increase in reactivity of the muscular vessels of the capillary bed as shown by an increased responsiveness to topical epinephrine This hyperreactivity is confined to the acute stage and disappears when the chronic phase 15 reached There is a parallel temporary increase in the sensitivity of the mesenteric vessels to the intravenous injection of VEM There is also a progressive hyperplasia of the capillary bed which reaches strik ing dimensions We believe this lesion is significant and a regular accompaniment of hypertension in the rat (Fig 4) Metabolic and Humoral Alterations in Hepato-Renal Vasotropic Factors in Renal Hypertension Weeks of hypertension TTTTTT Chronic stage Malignant! i stage Control period 1 3 5 7 9 J1 13_15 17 190 Acute stage ! Blood pressure (mm Hg) Clamps on both renal arter es VEM formation in vitro by slices from clamped kidney marked Marcuta M VEM in Oz VEM in N2 14 Hodnota Harhed Mesoappendix assay of untreated blood VEM in blood Moot لالالالا Torked liderate Mid Traces VOM Mesoappend xosudy of blood ofter aerobic incubauon with normal kidney si ces VEM VDM ratios in blood لللللللللLLLLبل Consult period Weeks of hypertension FIGURC 4 The relation of experimental renal to essential hypertension is still controversial it was therefore of interest to look for these same hepa torenal factors in essential hypertension in man utilizing the fractiona. tion procedures described above This was done with results shich nere identical to those obtained in our studies on renal hypertension in the ·dog (E Short and B W Zweifach, Tr Am Physicians 61, 350 (1948)) VEM and VDM were regularly present in all the cases studied The ratios were generally such as to yield a neutral reaction before fractionation but in a number of instances there was VEM predominance as in the dog This was most likely to occur in the more severe and malignant group Three cases of hypertension due to Cushings syndrome yielded essentially the same findings These studies were then extended to the hypertension of toxemia of pregnancy, the bloods being fractionated during the peak of hypertension (E Short and B W Zweifach, Program, 22nd Scientific Sessions, American Heart Association, page 7, June 1949) In all instances large amounts of VEM and VDM were present, sometimes one, sometimes the other factor predominating on the return to normotensive levels following delivery both factors either disappeared or were present in traces Opportunity was provided in a few instances of essential hypertension to study not only peripheral blood but also blood taken directly by catheters from the renal and hepatic veins The assays showed high con centrations of VEM in renal, and high concentrations of VDM in TABLE I Bio Assay of Blood of Patients with Essential Hypertension Dura tron Patient Sex Age Blood Pres sufe Renal Impair ment Rat Test of Plasma Un After treated Kidney Retin Cereb opathy Phenom Hyper Tacub yrs min min 2 0 Neutral t: VD mm Hg 0-3+0-4+ 172/99 0 0 180/110 0 182/110 0 2 0 H Neutral VDM 31 3 HD VOM? 11** 1+ 0 MA (f) 52 ET (1) 25 RP(f) 26 AP(f) 49 RC(m) 45 E(f) 37 EA(m) 39 CT() 41 RP(f) 32 tr VEM Neutral Neutral VDM 29 7 0 VDM 24 3 0 250/135 172/104 30/130 190/190 2²0/135 180/106 116/76 2+ H 2+ HD 2+ 0 2+ HD En tr VEM Neutral 6 11/2+ 7# 1+ VDM 36 VDM 21 VDN 33 VDM 24 Neutral 1+ Neutral 2 + H 0 VEM 27 Neutral 7## 1+ ** Cush ng s Syndrome with malignant adrenal cortical tumor # Aneurysm of L renal artery # # 3 months after L. nephrectomy TABLE I Vasotropic content in mild to moderate essential hypertension Bio assays carned out on peripheral blood before and after aerobic incubation with norma! Midney slices for one hour at 37 50 C to inactivate VEM and unmask VDV H="headache D = dizziness Es = encephalopathy 72 ·min min hepatic vein blood the mixed peripheral bloods being either neu ral or having some VEM predominance These findings would suggest that at least with respect to these specific defects in the VEM and VDM mechanisms there is no difference between experimental renal and essential hypertension in man (Tables I and II) TABLE II Bio Assay of Blood of Patients with Essential Hypertension Dura Blood Renal Rat Test of Plasma Pat ent tion Pres Impair Retin Cereb Ua After Ser Age Hyper sure ment opathy Phenom treated Kidney Incub yrs mm Hg 0-3+0 -4+ HS() 481 28*** 210/140 0 3+ HS Neutral VDM 30 MF(t) 761 22 194/90 1+ 3-4 0 Neutral VDM 30 MS(f) 45 g** 260/130 1+ 3+ H Neutral VDM 26 TE(m) 301 180/120 3+ VEM VDM 21 MK(1) 41 20. 172/132 1+ 3+ H En tr VEN VOM 21 43 22*•. 164/112 1+ 3+ 0 Neutral VD 32 SC(m) 50 4 228/130 2+ 3+ O tr VEM VDM 19 FH() 48 5 240/110 2+ 3+ HD En VEM 21 VDM 27 AG(m) 33 4 238/149 4+ 0 VEM 21 VDN 36 1 • Plasma incubated sn 95% 0,5% CO, but without k dney gave same reaction as un treated plasma ** Un lateral sympathectomy *** B fateral sympathectomy TABLE II Vasotropic content in moderate to severe essential hypertension Boassays carried out on peripheral blood before and after aerobic incubation with normal kidney si ces for one hour at 3750 C to inactivate VEM and unmask VDM H = headache S cerebral accident D = dizziness En = encephalopathy Fremont Smith What capillary beds have you studied? Short Only the mesentery and omentum Fremont Smith Do you know whether the hyperplasia is generalized and found in the skin and other parts of the body Short We do not It is a very striking picture and once we began to recognize it in our animals with experimental hypertension we began to look for it in the large numbers of rats which we use for the mesoappendix tests We found when the mesoappendix was exposed for testing purposes that occasional rats showed the same type of capillary hyperplasia of the vascular bed of the mesentery Several of these animals were saved and subsequently had blood pressure measure ments talen on them They were found to have spontaneous hyperten 73 ·sion The occurrence of this specific capillary hyperplasia in both spon taneous and experimental renal hypertension is of considerable interest The importance of the adrenal cortex for hypertension prompted investigation of the relation of the adrenal cortex to the renal VEM mechanisms in dogs, rabbits and rats (B V Zweifach, S Rosenfeld, S Baez and E Shorr, Factors Regulating Blood Pressure, Trans of the First Conference, Josiah Macy, Jo Foundation, New York, p 72, (1947)) It was observed that following ablation of the adrenal cortex the kidney lost the capacity to form VEM either in oxygen or nitrogen This property was not preserved by the maintenance of rats on salt but was maintained by supportive therapy with DCA In addition the kidneys of rats were capped with collodion gauze and the adrenals removed in two stages, the animals being supported by salt (B W Zweifach and E Short, Federation Proc 8, 175 (1949)) No hyper tension developed, there was no hyperplasia of the omental capillary bed and the kidneys were found to have lost their capacity to form VEM However, appreciable amounts of senin were present in the kidneys of these animals When such adrenalectomized rats with capped kidneys were supported by DCA, hypertension developed This interdependence of the renal VEM mechanisms and the adrenal cortex, as well as the very specific metabolic lesions in the renal VEM mechan isms in both essential and experimental renal hypertension, and the specific hyperplasia of the vascular bed of the mesentery in the rat--all these promise to be of great value in elucidating the relationship of the adrenals to hypertension The absolute dependence of hypertension upon the adrenals is clear What remains to be established is whether the adrenals serve as a necessary link in the chain of events leading to hypertension or whether hypertension is primarily the result of adrenal cortical overactivity Before discussing the relation of these humoral mechanisms to the problems of cirrhosis and edema I should like to comment briefly on the evidence for their participation in congestive heart farure in man In cooperation with Dr Leiter s group at the Montefiore Hospital, we undertook a study of patients with congestive failure due to theumatic heart disease (R Mokotoff, D J W Escher I S Edelman, J Grossman, L Leiter R E Weston BV Zweifach and E Short, Federation Proc 8 112 (1949)) In these patients mixed peripheral bloods as well as bloods obtained by catheterization of the renal and hepatic veins were studied for their content of hepatorenal factors Renal function studies including measurements of the oxygen tension of the arterial and venous blood from these organs shoned that the mean oxygen tension within these organs was loner than normal and that the AV difference was increased We consistently found VEM in the renal vein blood and 74 ·VDM in the hepatic vein blood, the mixed peripheral blood giving either a neutral reaction or a slight predominance of one or the other facto: It would appear then that in congestive failure the reduced oxygen tension prevailing in these organs is sufficient to initiate forma tion of these hepatorenal principles Direct evidence in support of this inference was provided by normal subjects who were breathing a mixture in which the oxygen tension was reduced to 10% After about 10 minutes in this atmosphere, VEM appeared in the renal venous blood The implications of these observations for the hemodynamics of conges tive heart failure still await clarification It is well accepted that there is an increased peripheral resistance in congestive failure It is also established that the filtration fraction is increased, most likely due to an increased constriction in the efferent arterioles of the glomerulus It is still a matter of speculation whether the local accumulation of VEM within the kidney may be responsible for this increased efferent arteriolar resistance it was of interest that one case in which VEM was not present in the renal venous blood had no increased filtration fraction However, for the present the role of VEM in vascular behavior within the kidney must remain conjectural Turning now to the aspects of our studies which are of particular interest to this group I shall describe our findings as to the involvement of these hepatorenal principles in cirrhosis and in edema formation Our initial studies nere stimulated by the observation of Dr Paul Gyorgy that animals in nutritional cirrhosis tolerate minor operative procedures, such as liver biopsies, very badly Apparently such animals, as well as human subjects with liver disease, are predisposed to shock Dr Gyorgy Lindly made available to us for these studies a group of rats made arrhotic by a low protein (10%) high fat diet In addition, we also studied a larger group ([ Shorr and B w Zweifach redera tron Proc 7 115 (1948)) Very significant changes were found to have occurred in these animals in contrast with the livers of normal rats their livers usually contained large amounts of VDM VDM was present in the blood just as it is in the blood of shocked animals although the arthotic animals were not subjected to any manipulative procedures The mesenteric vessels had a diminished responsiveness to epinephrine such as is seen during the hyporeactive stage of hemorrhagic or trau matic shock when VDA predominates in the blood The liver mechan ism for inactivating VDM under aerobic conditions in utro was regularly and significantly impaired It is this defect which is believed to be responsible for the continuous production of VDM by the livers of these cirrhotic rats The kidneys in addition were found to have sustained a very serious derangement of the VEM mechanism as a result of shich they were unable to form VEN in nitrogen or formed it in very trivial amounts Thus the hepatic and renal defects relating 75 ·to these vasotropic substances were such as to result in a state which predisposed these animals to shock Their inability to call on the renal VEM mechanisms did not permit them to take advantage of the com pensatory vascular behavior which this principle induces in response to circulatory stress resulting from shocking procedures The defects in their hepatic VDM mechanism had set up humoral and vascular situations comparable to those seen in animals in the hyporeactive or irreversible stage of experimental shock With respect to VEM and VDM therefore these animals exhibited a condition which could be appropriately termed a hepatorenal syndrome with profound circulatory implications Best What weight rats? Shorr In our initial studies we used rats weighing about 160 gms, in the later studies the weight of the cats ranged from 150 180 with an occasional one up to 220 gms Best Were they not getting any kidney lesions? Sborr We have taken tissue for sections but they have not as yet been run through They looked grossly normal Attempts were also made to correct these defects after they had been established by 4 months of this diet Although the number of expert ments is not large, two factors were found to correct the renal VEM defect These were methionine supplementation and the administration of DCA Following both procedures the kidneys of these animals regained the ability to form VEM The relation of diet particularly protein, to the renal VEM mechan isms was reinvestigated in more acute experiments with diets containing 5 and 10% protein as well as fasting (M A Payne and E Shorf, Federation Proc 8 125 (1949)) Adequate vitamin supplements were supplied throughout Fasting rats sustained a significant reduction in the capacity to form VEM by the 4th and a total loss by the 6th day of the fast Those receiving the 5% casein diet showed appreciable reductions in VEM formation by the 8th and complete abolition from the 12th through the 39th day With a 10% casein diet VEM forma tion did not take place after the 8th day DCA failed to prevent this defect whether given prophylactically from the start of the experiments or after the 14th day when this defect was well established It would appear then that the integrity of the renal VEM mechanism is dependent on at least two factors so far studied the dietary protein and the adrenal cortex In acute and extreme protein depletion DCA is unable to restore this defect The implication of these observations for nutri tional as well as human cirrhosrs are under active investigation As pointed out above the cirrhotic animal by virtue of these defects, is predisposed to circulatory failure This occurs through the loss of the 76 ·compensatory effects of VEM and through the presence of defects in the hepatic VDM mechanisms which are comparable to those believed to be responsible for the hyporeactive state of shock which becomes irreversible to transfusion Two other aspects of cirrhosis may also be involved with these de rangements, namely, antidiuresis and edema formation Both have been under active investigation in our laboratory These experiments have served to establish the fact that the administration of partially purified liver VDM concentrates as well as crystalline ferritin or apoferritin with which VDM has been identified, leads to profound antidiuresis in rabbits and dogs receiving oral or intravenous hydration (S Baez, A Mazur and E Shorr, Federation Proc. 8, 7 (1949)) The administra tion of 200 300 gamma of ferritin nitrogen/kg of body weight by intravenous infusion reduces urine flow to the extent of 60 80% during the period of infusion When the oral hydration method is used for producing diuresis there is a significant delay in reaching the peak of diuresis and in excreting the water load These antidiuretic phenomena following ferritin or apoferritin occur with no alteration in systemic blood pressure Animals can be immunized with heterologous feruitin When this occurs, and precipitins for ferritin appear in the blood stream, the antidiuretic effect of ferritin fails to occur (Fig 5) Rabbit No 6 Wt 285 kg Volume of umne (ml/min) orally 0 Infusion 150 cc (12cc) wote" Soline 1-Territin HSE, (140 IN/xg) 0 90 90 320 75060 210 240 270 Time in minutes Anu diuretic Iffect of Horse Spleen Ferritin FIGURES 77 ·Studies are in progress designed to clarify the manner in which this antidiuresis is achieved Three dogs with stalk resection have developed diabetes insipidus These animals which had previously given an antı diuretic response to homologous ferritin failed to do so after the de velopment of diabetes insipidus It is our tentative conclusion that VDM (ferrıtın) exerts its antidiuretic effect by stimulating the posterior pituitary to discharge its antidiuretic principle Fremont Smith There was no febrile response? Shorr There was no febrile response and no change in blood pressure during the experiment Hanger What is the source of the ferritin you use in the immuniza tron experiments ? Shorr Horse spleen ferritin which is injected into the rabbit We get good precipitin reactions in blood with these heterologous ferritins whereas we do not get antibody formation with homologous ferritinis Best What is the nature of the protein in ferritin Shorr There is nothing particularly characteristic of the amino acid composition of ferritin We have done amino acid analyses and have accounted for about 98 percent of the nitrogen without striking any thing peculiar except in one respect As compared with other proteins which are combined with iron the iron in ferritin is tightly bound to the protein The iron also appears to be contained within the structure of the molecule in such a way that it exerts no surface charge Apofer mitin has also been analyzed and there are no significant differences from ferritin in its composition as regards amino acid content Watson Is it known whether it is similar to the tronbound globulin in the circulation Shorr It is quite different The iron in the ironbound globulin of the blood is readily split off whereas the iron is tightly bound within the ferritin molecule and can only be split off by powerful reducing agents It was now of interest to see whether the findings in human cases with antidiuresis and edema were compatible with the concept that an accumulation of ferritin in the blood as a consequence of hepatic VDM derangements might be responsible for these antidiuretic phenomena in man To this end bio assays were carried out on blood from patients with arrhosis and antidiuresis and on subjects with nephrosis or in the nephrotic stage of chronic glomerulonephritis (E Shorr B W Zweifach A Mazur and M A Payne unpublished data) In many of these patients edema fluid was also analyzed The bloods and edema fluids were fractionated in the minner previously described for assaying VEM and VDM when they are present in mixtures (Ta' 1 & IV) 78 ·TABLE SIT The Octurrence of VDM 19 Decompensated Hepatic Cirrhosis Plasma Protein RAT MESO APPENDIX ASSAY- PLASMA Patent, Sex Age D agros Edema Asc<25 ת Album Globus To al After Control Incuba I on w Sal ne* After VDM Insct. wat on By Incubar on W th Ant Fers in Serun** MCM) 41 Laenne type + 20 23 43 neutral TC(m) 47 + 30 25 55 t + 30 79 f + 34 35 27 62 VDM 18 min VDM 24 VDM 24 VDM 27 VOM 22 VDM 18 VDV IS VDM 15 VDM 21 VDM 21 25 42 18(1) ND (1) M() 55 JF(m) 58 MMO) 48 MM(I) 51 67 +++ 32 33 65 + wy 27 68 + 26 39 63 o 38 44 8 Z VDM trace MG(1) 5+ Cholangiolotic type BS() (compensated) Lachnec tyre 0 0 25 72 VDM • Incuba od 91% 0.5% CO for 1 hour at 57 5°C ** Prered by mmpuh 2 ng tabb tawth ferr in ob a ned from human liels ·TABLE IV The Occurrence of VDM in the Nephrotic Syndrome RAT MESO APPENDIX ASSAY Plasma Proteins PLASMA EDEMA FLUID Patient, Sex Age Diagnosis Total After After VEM Control Inactivation Incuba By Incubation With Kidney Slices After VDMIN activation With Anti Ferritin Serum** After VEM After Inactivation Control By Incubation Incubation* With Kidney Slices tion + 0 17 08 25 08 VDM 27 min VDM VDM 28 min + + 11 25 36 18 min VDM 15 min VDM trace BM(m) 20 days lipord nephrosis CF(f) 5 yrs chronic glomerulo nephritis LF(f) 6 yrs lipoid nephrosis AM(m) 17 yrs chronic glomerulo nephritis AN() 5 yrs Lipoid nephrosis + 08 24 32 + + VDM 21 min VDM 19 min VDM 21 min VDM 21 + 17 16 33 VDM 25 min VDM 24 min VDM 32 min neutral VDM 29 min + + 04 29 VDM 35 min VDM 27 min VDM 21 min VDM 18 min * Incubated in 95% 0r5% CO, for 1 hour at 373°C ** Serum from rabbits immunized with ferritin from human liver ·rn mental Endan nara ne fallare Alle hjects with edema and VPM in the circula VIM was present in the blood or edema fiud in several patients with Cirrhosis but without ascites or antidiuresis, VDM was absent or present only in traces Opportunity was provided to study one case of nephrosis at the height of her edema, during treatment with intravenous serum albumin which content of the blood again rose to its original levels These observations are at least compatible with the concept that antıdwresis and edema, in a variety of clinical conditions including cirrhosis, may stem from hepatic defects in VDM mechanisms as a result of which VDM ac cumulates in the blood and stimulates the posterior pituitary to a dis charge of antidiuretic factors There would appear to be parallel derangements of the VEM mechanisms as a result of which they are unable to oppose the antidiuretic action of VDM This latter inference must be regarded as conjectural for the moment and a matter for future investigation Armstrong In cases of glomerulonephritis which you studied was there any hypertension? Sborr There was mild hypertension, 140/90 in one case, 160/98 in another However, don t you think that the hypertension in glomerulone pluratus is generally ascribed to anatomical changes in the renal blood vessels which increase peripheral resistance? We are now in a position to summarize the present status of these investigations as to the role of these liver and kidney principles in arcu fatory homeostasys We have established the conditions under which the metabolism of VEM and VDM is regulated under nal cond tions We have found that their metabolism is deranged in a variety of circulatory disturbances such as hemorrhagic and traumatic shock, essential and experimental renal hypertension nutritional and human arrhosis and in a variety of states associated with antidiuresis The locus of action of these vasotropic prinaples has been shown to be the terminal vascular bed They evert opposite effects on the behavior of the metarterioles and precapillary sphincters These effects consist in alterations in vasomotion and reactivity of these muscular units of the capillary bed These alterations have a profound influence on the flow through the true capillary bed and on the exchange of fuud shich takes 81 ·place between the blood within the capillary vessels, and the tissue fluids The preponderance of the vasoexcitor, VEM, makes for capillary ischemia and inward filtration VDM preponderance on the other hand makes for overall capillary flow and outward filtration Unless com pensatory vascular adjustments took place in the vascular bed central to the capillaries, VEM predominance would cause tissue dehydration and VDM predominance edema These principles therefore would appear to constitute a new homeostatic mechanism for the regulation of the peripheral circulation They would appear to constitute a major element in the homeostatic forces whose end is the preservation of the stability of the internal environment in a variety of circulatory disturb ances, these vasotropic systems undergo specific derangements of such character as to suggest that they are related to a number of phenomena characteristic of these circulatory states Several environmental factors have been shown to lead to derangements in these hepatorenal mecha nisms Those already studied are reduced oxygen tension, low protein intake and adrenal cortical hormones Low protein diets impair the integrity of both the renal and hepatic vasotropic systems The adrenal cortex is necessary for the integrity of the renal VEM mechanism Lth hepatic and renal systems be concerned not only with also with the renal excretory established in animals and appear to be mediated through the posterior pituitacy The possibility has been advanced that oliguric phenomena in human disease may be due to derangements in the VDM and VEM mechanisms as a result It is appreciated that much rernains to be done before a causal relation ship can be established These principles act in a chronic and subtle manner quite different from other vasoactive substances such as renu and angiotonin whose effects on blog pressure are more acute Markou itz Do you know Trueta s work and what do you think about it? Sborr I have rezer graph of my his associates and spent a whole even us and cons of the Trueta mechanism .th the ww1on The pros this postulated mechanism The cc relev I mechanism, evidence *t was the extraction extreme ais and seem to have de problem of ·man Thus we have an anatomical pattern for which no functional corollary has as yet been demonstrated at least to the satisfaction of most of the workers in the field of renal physiology Markouitz He gave evidence in his book which he published with Barclay, Franklin and Pritchard It seemed a very convincing story 1 spent a day at Oxford watching the work going on and it was very impressive Shorr i thunk the present difficulty is determining its role in tena! function Even in severe experimental shock, with reductions in blood How down to 3% of the normal, Van Slyke shoned no change whatsoever in the PAH extraction ratio Hanger Could I ask you jour opinion as to the mechanism of ascites in cirrhosis? Do you think it is due to an alteration of the capillanes of the peritoneum which disturbs the reabsorption of Auid, of do you think it is all an anu diuretic effect? Shorr I am inclined to favor the participation of positive antidiuretic factors as basic to the water retention with the localization of fluid within the peritoneal cavity as a result of local factors such as intra hepatic resistance to portal blood flou But this latter effect we would regard as subordinate to the factors influencing the renal excretion of fluid and salt There is also the possibility that the decompensatory effects of the chronic predominance of VDM on the capillary blood How would by favoring the hydrostatic pressure over the osmotic also contribute to the formation of tissue ederoa. Hanger That is what I wanted to know Shorr There is also some evidence which tve hope to supplement by more experiments that the hypervolemia of cirrhosis and infectious hepatitis may be related to VDÀI predominance in blood Through the Zweifach and C Shorr unpublished data) One is tempted to interpret this situation as follows for some unknown reason the granulosa tumor affects the hepatıc VDM mechanism and leads to VDN formation As the VDM accumulates in the blood stream it progressively opens up the capillary bed In order to fill the capillary bed and prevent circulatory collapse, there is a parallel increase in the blood volume This very interesting aspect of our studies needs amplification and we hope to report to you more fully at subsequent meetings of this group Fremont Smith Have you any evidence that VDM could still have an antidiuretic effect, if it were masked by equivalent amounts of VEM Shorr We suspect that in the presence of equivalent amounts of 83 ·VEM VDM would have no antidiuretic effect This is at present a matter of inference in that our patients with both VEM and VDM in blood show no delay in excreting the water load When we have obtained VEM in sufficiently pure form we shall carry out direct expert ments bearing on this point Fremont Smith Did you have one case of mild hypertension in the nephrotic group? This one case with a blood pressure 140/90 did not seem to fit in that it had antidiuresis with both VEM and VDM in the blood Shorr There were 2 patients with glomerulonephritis and mild hyper tension But there was no VEM present in either patients blood only VDM Fremont Smith Is VEM not present in the hypertension of glomeru lonephritis? Shorr We cannot say very much about glomerulonephritis We have studied a few severe cases in uremia but their bloods were so toxic for our animals that we could not assay them This same toxic effect is also exerted by blood from animals in the terminal stages of malignant renal hypertension Frenzont Smith What about eclampsia? Is that the same? Shorr In our eclamptıc cases whenever the bloods proved toxic we were able to assay them after dilution with saline Fremont Smith With what results? Shorr Both VEM and VDM were regularly present sometimes one sometimes the other predominating at the peak of the hypertension before delivery As I described above with the fall in blood pressure after delivery these factors disappeared or were present in only trace amounts Hoffbauer How potent is fermiten as an antidiuretic? Might this test be used in the place of the rat mesoappendıx assay for VDM? Shorr The rat test will detect ferritin in amounts of 0 0005 micro grams of nitrogen To obtain an antidiuretic effect we must inject 200 300 micrograms/kg Hence the antidiuretic test could not possibly be used for the assay of VDM in blood because the amounts present in the body fluids are more on the order of 0 001 micrograms/cc Watson Is ferritin plentiful enough so that you might propose to try it out on the human subject? Shorr We are accumulating a stock pile of crystalline human ferritin for just such purposes Gyorgy You probably could not use the antiserum for treatment of human disease because the supply might be quite limited Shorr We hope that with the help of our friends we will eventually 84 ·accumulate enough ferritin to produce enough antiferritin serum for human use We have already begun experiments with antiferriter serum in animals but it is too early to speak about that l'atson There is no evidence of ferritin or apoferritin in urine is there? Sborr Dr Mazur has accumulated some evidence that small amounts of fertitin, of the order of 10 to 15 micrograms of ferritin N per 24 hours may be excreted by the human subject This amount of fertitin is much too small to account for the antidiuretic effects obtained with urine concentrates in circhosis by Ralli et al (E P Ralli, 1 S Robson, D Clarke, and C L Hoagland'ſ Clin Imessigation 24, 316 (1945)) and other workers Fremont Smith You dont think that festitin is antidiuretic by virtue of its action on the kidney itself but through the release of antidiuretic factors from the posterior pituitary Shorr That is our present vies I can t say as yet that it has no direct action on the kidney but our sesults so far suggest that its oliguric effects are mediated through the anterior pituitary Markountz Do these dogs have polyuria? Short Yes Alarkowitz What was done? Shorr A resection of the stalk Markowitz You could still get diuresis Shorr Four animals were operated upon One by Dr Kendrick Hare and three by Dr Silvio Baez of our laboratory All four had the polyuna which is usually described to follow this procedure Alarkoustz How much diuresis did you get? Shory On the thund or fourth day after operation, diuresis was very marked-up to 23 liters per 24 hours The polyuria was gradually reduced and after a month of so was about 1 liter per 24 hours This is from 2 to 3 times the usual urinary output of the dog Markowitz I have not been able to do it I ll have to try agam Shorr We nere greatly helped by the advice of Dr Kendrick Hare who has been carrying out these procedures for many years Markou sız I have often wondered why a hypophysectomized dog did not have diabetes insipidus fitury Draketes arsipitas des not follow sest hypopitysturoy You have to leave in the anterior labe Markou siz That explains the whole thing I am sorry I was not clear about that in my mind 85 ·SECTION V THE METABOLIC BEHAVIOR OF THE EVISCERATE RAT Ingle Gentlemen, my interest in the eviscerate sat is concerned with two general problems First, what metabolic adjustments can this preparation carry outSecond, what factors influence the survival of the eviscerate animal? I think that this latter problem is one of the most challenging in physiology We have adequate replacement therapy for vital organs such as the pancreas, the adrenal cortex and the anterior pituitary, whereas the heart, lungs, and brain are essential for obvious reasons The liverless animal does not survive for very long and the causes of death have not been determined Could we prolong the survival of the eviscerate rat by improved intravenous feeding and would it be possible to extract anything from liver itself which would prolong survival) I started this work at the Mayo Institute twelve years ago You may well guess that my interest in liver was stimulated by Drs Mann and Bollman My beginning studies involved certain difficulties In order to keep the animals living I had to inject them frequently with glucose and in order to determine survival I had to either spend the night in the laboratory or carry the animals back and forth between the laboratory and my home My Friends in the laboratory exhibited unsympathetic amusement and the members of my family were neither sympathetic nor amused I dreamed of a time when I could have instrumentation which would permut us to feed the animals by continuous intravenous jajection and to record the time of death through the automatic measurement of some physical change If I may have the first slide, we will skip a number of years up to the time hen I acquired such a device (Figure 1) These are six es iscerate rats enclosed in a box with temperature constant at 265205 Centigrade The animals are fed intravenously by constant infusion We determine the time of death by amplifying the DC potential ictivates 1 relay switch which in turn le Micromax recorder The Micromax six rats We can record accurately the time that the heart stops beating within six minutes Nor for a little more detail about this procedure We anesthetize the animals with cyclopal, which is a rapidly acting barbiturate Intact cats recover rapidly from this anesthetic but the liverless rat does not recover from any kind of barbituric anesthesia, although it does recover yo 86 ·un FIGURE 1 from ether The animals are placed on their backs upon a little table above a funnel and a collection flask so that we can recover the urine The continuous injection machine was designed and built in our shops It handles six syringes and its blocks well accommodate any size syringe from a lcc to a 50cc capacity Syringes which will deliver their volume with the standard stroke of 65 mm must be ordered to a specified 87 ·diameter The machine has several different speeds so that the volume which can be delivered per unit time can be varied over a wide riage In all of the experiments to be described we have used a standard rate of delivery, 20cc per 24 hours The solutions all contain 09 percent sodium chloride with and without insulin and with various glucose loads Body weight was constant at 250 grams (Slide Table I) We have recently completed a study of the blood changes in these animals after 48 hours of infusion, a time at which they are approaching death The glucose load was 44 mgms of glucose per 100 grams of rat per hour (44/100/h) and the insulin dose was 4 units of regular insulin (Lilly) per 24 hours per rat I have listed here some of the most striking changes that we have seen There are other changes which may be important although less steiking The prothrombin time is short at the beginning of the experiment but after 48 hours the blood will not clot at all At the time of death there are many diffuse hemorrhages which almost always involve the bladder, frequently the kidneys and occasionally the lungs I have even seen hemorrhages in the brain The white cell count is markedly reduced This may be related to the fact that these animals are non resistant to infection Dr Armstrong mentioned yesterday that the rat is highly resistant to infection This is true of the normal rat but not of the liver less rat Within 24 hours the blood the body cavities and even the bone marrow are heavily charged with bacteria and we have not yet been fully successful in our attempts to preserve asepsis in our opera tions We can reduce the bacterial growth per unit time in all our animals and can carry some of them through the experiment without any bactectal infection The extent of infection is certainly a factor which influences the time of survival TABLE I Some changes in the blood of eviscerate rats Initial 49 Hours Prothrombin Time 15 Infinite (Seconds) White Cell Count 116 28 (Thousands/CMM) Total Plasma Protein 58 33 (GMS/100/CC) Bilirubin 00 44 (MGMS/100/CC) CO, Combining Power 625 350 (Volumes/100/CC) Short Have you ever used penicillin Ingle No After the experience of Dr Markowitz we will certainly try that The plasma protein is reduced This may be related to the development of ascites in the eviscerate rat Some of them virtually drown in their own fluid since the chest cavity fills so completely that 88 ·respiration may be impossible if the fluid is drained from the chest the animal may sutvive for a few hours longer but will eventually die for other reasons We have cafused the rats with collords (gelatin) in an effort to keep up the osmotic activity of the blood but it has not affected survival Gelatın seems to escape rapidly from the blood There is no measurable amount of bilirubin in normal fat blood but following evisceration the animal becomes severely jaundiced There is a definite tendency to deselop acidosis following evisceration Hanger Did you study blood platelets > To my know ledge no one ever studied blood platelets in eviscerated animals Do your smears give you any indication of a change? Ingle Miss Prestrud has carried out the hematological studies of our eviscerate rats Although platelet counts have not been done, it is her impression that the number of platelets were greatly reduced following evisceraton Until a more precise study has been carried out it will not be possible to reach a conclusion regarding platelets Armstrong Can you indicate on this slide where the plasma protein disappears > That seems to be faster than one would expect it from the isotope studies of its desintegration Iugle The Ausd which accumulates in the body cavibes contains significant amounts of protein It seems probable that it has come from the biood The other day at Detroit, Dr Roberts discussed changes in plasma proteins in the eviscerate rat and said that they did not find plasma proteins in the fluid which they recovered from the body cavities but their experiments were terminated within 24 hours as compared to 18 hours in our studies Dr Roberts noted that the albumin of the blood 1 . the presence of fibrin did show that there was a more saped loss of albumin than of globulin in our eviscerate rats and that our data are in agreement with the data of Roberts although the experimental condi tions were rather different Our animals were infused with glucose and insulin and the animals of Roberts were not Fremont Smith Hos much weight do the animals gaip? Ingle About 10 grams on the average Knasely Perhaps everyone else knows, but would you mind describing the surgical operation done on the animal in evisceration, hogy you do it, what do you take out? Ingle We remove the liver and the entire gut leaving only the Lidneys and the adrenals and in some experiments ne remove them also It is 89 ·a two stage operation The first stage involves total ligation of the vena cava between the liver and kidneys so that the animal builds up a collateral circulation The second stage or actual evisceration is carried out two or three weeks later To come back to the question of the missing plasma protein, Dr Roberts thinks that it may be used by the kidney for renal gluconeogene SIS This must be considered Shorr Do you know whether the lactic acid content of the blood or tissues rises? Ingle We have not determined that recently I think that Dr Bollman remembers the few experiments that we did when I was with him a number of years ago As I remember, there was no increase in the lactic acid of blood and muscle within the first few hours As a matter of fact, the increase in lactic acid which follows the stimulation of muscle is not sustained "The animal has the ability to utilize lactic acid in the absence of the liver Would you agree to that Dc Bollman? Bollman That is essentially what we have found in the eviscerated as well as in the hepatectomized rat and in the eviscerated dog Im mediately after operation there is a rise in the lactate content of the blood and also of the muscle This case is related to the anesthesia and surgery as it is similar to that found in sham operated animals On recovery from the operation the lactate levels of blood and muscle return to normal in the sham operated animals but in the eviscerate or liverless animals the level may decrease but does not return to normal levels There is usually a terminal rise of lactate which is of larger magnitude With exercise or muscle stimulation there is the expected increase in the lactate content of the muscle and blood and with rest or continuing muscular activity there is a disappearance of the lactate as occurs in the normal animal The level in the muscle and blood of the eviscerate animal returns to its pre exercise level which is higher than that of the intact animal Large amount of injected lactate will also disappear from the blood of the eviscerate animal There seems to be no real impairment of the animal s ability to metabolize lactate Short I was thinking in terms of the long experiment in which the circulatory phenomenon might be the limiting factor rather than the ability to oxıdıze in the presence or store in the presence of insulin and glucose If you have a limiting factor in terms of the oxygen supply to the tissues in the course of the deterioration of the animal you might get a lactic acid again I know Soskın, for example, in evisceration experiments, got back lactic acid The values rose at a very extraordinary rate, if you remember Bollman You are quite right As the animal fails in the shock like conditions with low blood pressure and blood volume there is a marked 90 ·rise in the lactic acid content of the blood and muscles This has little to do with the liver but is more related to the relative anaerobiosis of the tissues Of course of the blood could be arculated through a normal liver most of the lactate sould be removed but the impairment of Circulation and not the absence of the liver is the cause of the increased lactic acid Fremont Smith Do you know what happens to the blood volume> Does the blood volume stay about the same? Ingle We do not have any satisfactory measurements of blood volume The hematocrit does not change and the red cell count does not change but it is possible that there is a significant shift in the Volume of circulating blood Fremont Smut What would you say is the blood volume of these anmals normally or after you eviscerated them? Would it be 15 16 cubic centimeters Tungle We are able to recover 6 to 8 cc from these animals at the end of 24 hours which is not as good a recovery as you can get in 2 normal animal Goldblatt What weight Ingle 250 to 260 grams Sborr That would be 15 to 16 cubic centimeters Fremont Smith What I was trying to get at is that the animal gained 7 grams in weight of 7 cc of water and with the biood volume which you anticipate it seems to me that the gain in weight is enough to dilute your plasma by this much assuming that you leak whole plasmı out into the pleural cavity and other tissues I think one could probably make at least a guess that if there is a gross permeability change and leakage of whole plasma and if the hematocrit does not change (le blood volume remains about the same after the operation) then the retention of fluid in the animal as measured by the gain of weight of 7 grams would be enough to account for the dilution of the plasma proteins Ingle Yes That is a reasonable possibility Shory How much have you lost in the body cavities> Ingle We have not measured that Shorr Can that be 5 or 6 cubic centimeters? ingle It might be Sborr So that might reduce the blood volume Tremont S21111 But they have gained 7 ce of water which would be enough to restore the bicod volume to its original level But whole 91 ·degrees the survival time is decreased below the optimal values obtained at 265 degrees Shorr That happens to be the temperature that Dt Simms, who is doing long time aging studies on rats, has chosen as optimal for the maintenance of the rat, 78 degrees Ingle (Slide, Table III) Temperature has a marked effect upon the glucose requirement of the eviscerate fat At 26 degrees C the glucose requirement over a 3 hour period following evisceration is 64 mgms of glucose per 100 grams of sat per hour (64/100/h), but increasing the temperature to 38 degrees C brings the glucose requirement up to 150/100/h We have explored this temperature range thoroughly and a change of even 1 degree C will make a measurable difference in the glucose load which the animal can tolerate TABLE 111 The effect of temperature upon the tolerance of the eviscerate rat for glucose Glucose Requirement for a 3 hour period Degrees C mgms/100/h 26 64 38 150 Gyorgy How is this glucose measured? Ingle We determine the glucose load which sustains the blood glucose at its initial level of approximately 100 mgms percent We had not worked at this problem very long before I began to bring in one of my other major interests, that of the adrenal cortical hormones Our thinking about the nature of adrenal hormone action has become rather rigidly channelized and the textbook story would explain adrenal cortex physiology in terms of salt and water balance and in terms of hepatıc gluconeogenesis from protein The story of the relationship of the adrenal cortex to organic metabolism is something like this there is an increase in the need for the cortical hormones during any type of stress and the gland responds with an increase in its secretory activity The increased output of the hormones stimulates gluconeogenesis from protein in the liver thereby providing the extra carbohydrate energy needed to meet the increased requirement for carbohydrate energy during stress The liver is considered to be the site of action of the cortical hormones I doubt that any of the active investigators in the adrenal field would subscribe to such a simple story which does, never theless, appear in textbooks and in the classroom We set about to 94 ·360 340 OF 24 HOURS 3320 4300 280 ACE 260 2240 CONTROL 2220 200 180 160 BLOOD GLUCOSE MGMS AT END OF INFUSION 140 120 100 80 60 40 20 MGMS/100/H NO INSULIN 6-40+ INSULIN UP.JOHN.CODIL FIGURE 2 95 ·determine whether adrenal in the absence of the 1 w slowing that the 11 17 oxygena younus of the adrenal corter are dia betogenic even in normal rats and can cause the pouring out of much farger amounts of glucose than could possibly be accounted for in terms of gluconeogenesis from protein It seemed necessary from our data and from that of other investigators to conclude that the cortical hor mones must affect some phase of carbohydrate utilization the term utilization in a heande AN linis cart iw uc Ligure 2) We studied the effect of adrenal cortex extract upon the glucose tolerance of the eviscerate rat over a 24 hour period The glucose requirement of the rat falls during the first several hours following evisceration so that a glucose load of 4/100/h is sufficient to sustain a normal level of blood glucose by 24 hours Whern added the glucose requirement the u ullut attect glucose tolerance wwsic or insulin but in the presence of insulin it caused marked hyperglycemia and glycosuria in most of the animals (Slide Figure 3) We anticipated that the effect of adrenalectomy in the eviscerste rat should be more striking in animals treated with insulin than without insulin but the opposite was true In eviscerate fats given a glucose load of 4, 100/h the removal of the adrenal glands increased the rate at which thev ram 1 j 1. damals survived but the I w Lut una glucose were below normal whereas the non adrenalec tomized eviscerate rats developed hyperglycemia Adrenalectomy in eviscerate rats given a glucose load of 40/100/h with insulin failed to cause hypoglycemia in fact there was some decrease in glucose tolerance The changes were not great but we have seen it occur re peatedly and I have no doubt that it is a true difference Whenever you explore the dimension of load response in adrenalectomized animals when the load approaches a maximum the adrenally insufficient animal fails first In this experiment we pushed the glucose tolerance to its peak with usulin and the adrenally insufficient animal could not quite attain this peak At least that is the most rational general explanation that has occurred to us up to the present (Slide Figure 4) We at first assumed that the effect of adrenalectomy upon the glucose tolerance of the eviscerate rat was due entirely to the removal of the adrenal cortex We now have data which provide a 96

·AT END OF 24 HOURS
MGMS / PER CENT
ADRENALECTOMY
CONTROL
BLOOD GLUCOSE
INFUSION
OF
140
1201
1001
80
60
40
20
MGMS/100/H
NO
INSULIN
K-40-
INSULIN
UPJQHN.GO.Dull
[IGURE 3
97

·basis for suggesting that the absence of the medulla may have been a factor I was taught, and most of the textbooks teach, that epinephrine does not affect the level of blood glucose in the absence of the liver and yet there was quite a debate in the literature over this point by C F Con, (Physiol Rev 11, 143 (1931)), who claimed an extra hepatic action of epinephrme upon blood glucose and by S Soshin (Amer J Physiol 81, 382 (1927)) who densed it H P Himsworth and D B MCN Scott (J Physiol 93, 159 (1938)) reported that epine phrine accelerated the rate at which glucose disappeared from the blood of animals in which the liver was excluded from the circulation We have results that will agree with all three conclusions, depending upon the experimental conditions When the test period was limited to 2 hours, the administration of epinephrine with glucose and without insulin in the infusion fluid did not affect the level of blood glucose significantly When epinephrine was given in the presence of insulin it caused a striking rise in blood glucose Then as we repeated these 300 280 -EPINEPHRINE 20 PPM 2601 (1-CONTROL 240 220 BLOOD GLUCOSE AT END OF INFUSION MGMS / PER CENT 200 IBO 560 $40 120 100 80 60 J 40 20 *40- 4 MGH410014 UNITS INSULIN TIME OF INFUSION 16- *72-* K- -2 HOURS -24 HOURS 11210HM.CO FIGURE 4 98 ·experiments oser a period of 24 hours we found that epinephrie gren in the absence of insulin did indeed accelerate the rate of disappearance of glucose from the blood but in the presence of insulin the administra tion of epinephrine caused a rise in blood glucose Best Could that be related to blood pressure? Ingle We have not measured the blood pressure in these experiments but as I shall mention in a moment we have used a derivative of epmephrine which has a vasodepressor action and which gives the same story Shorr What happened to the carbohydrate content of muscle under these experimental conditions. Do you think that the epinephrine effect on the blood sugar might be due to the muscle having a high content of glycogen which could give rise to lactic acid and that this did not take place in the presence of insulin Ingle We have not measured the tissue glycogen at all, Dr Short That is a possibility in describing these changes in carbohydrate toler ance we cannot make any claim to know what happened to the carbohydrate that we administered or what shifts in carbohydrate stores occurred (Slide Figure 5) This gets more complicated as se go along There are many substances that affect the glucose tolerance of the eviscerate rat In fact we have not yet studied anything that fails to do so we have used posterior pituitary extract in these experiments and obtained exactly the same story that he obtained with epinephrine The results are not quite as striking but are all consistent The administration of posterior pituitary extract to rats given glucose and insulin for 2 hours causes a decrease in glucose tolerance Without insulin there was little or no effect Over a period of 24 hours the administration of posterior piturtary extract caused some decrease in glucose tolerance in the presence of insulm and some increase in tolerance in similar animals not given insulin We repeated the experiments using Prtressin and obtained exactly the same story Preparations of Pitocin gave small effects but it was found that this preparation of the orytocic principle contained about 20 percent as much pressor activity as did the Pstressin Up to this point it seemed probable that the effect of these hormones upon glucose tolerance ass closely related to pressor activity However we carried out an additional experiment using Isuprel which has a depressor action and the response to this pharmacologic agent was exactly the same as was obtained with epinephrine We still suspect that our results on glucose tolerance are based upon some action of these drugs upon the circulatory systern The point which I wish to make from these studies is that the eviscer e animal is capable of responding to these pharmacologic agents 99 ·Whether these results deserve any physiological interpretation I don't know, but the eviscerate animal is capable of adjusting to these hor mones, peculiar as the adjustments may be, and the kind of effect that we get is dependent upon the time over which we observe the animal and whether or not insulin is present - POSTERIOR PITUITARY 325 300 CONTROL 275 250 225 200 175 BLOOD GLUCOSE MGM PER CENT 150 125 100 75 50 25 PARTS PER 100 20 40 110 20 GLUCOSE MGM/007 64 65 64 16 16 INSULIN → NO INSULIN 2 HOURS INFUSION | 1020 1020 40 40 4 INSULIN NO INSULIN 24 HOURS INFUSION UPJOHN CO OLI in FIGURE 3 Bollman What are the dots on the slide? Ingle They show the range, the highest and the lowest values There is quite a lot of variability in most of them Here we happen to come out with a very small range, much smaller than is found at 24 hours We were just lucky in that series There are 20 pairs of animals repre sented each group (Slide, Figure 6) It has been known for many years that blood amino acids increase after removal of the liver It was demonstrated a few years ago by E G Frame and J A Russell (Endocrinology, 39, 420 (1946)) that the administration of insulto suppresses the rise to a significant extent In these experiments we were able to suppress the rise in blood amino acids completely by the administration of tasulin The initial value for whole blood is approximately 10 mgms per cent, and at the end of 24 hours without insulin the average value is nearly 30, but when insulin is given it is held at the initial level The extent 100

·of the rise and the degree to which insplin inhibits the rise is relate
to temperature as J A Russell and M Cappello (Endocrinology, 41
127 (19-19)) have recently demonstrated
MGMS / PER CENT
28
26
24
22
20
18
16
12
10
BLOOD AMINO ACIDS
8
6
4
2
UNITS INSULIN
0
4
CONTROL EVISCERATION
24 HOURS
FIGURE 6
UPJOHN
Gull

·(Slide Figure 7) We now come to a technic which I began to develop at the University of Minnesota in 1931 and which has now come through quite a period of evolution It is a method for measuring the ability of the rat to sustain work output in response to faradıc stimulation The rat is anesthetized with sodium phenobarbital the gastrocnemius muscle is weighted with 100 grams and is stimulated to contract 5 times per second The total distance that the weight is lifted is recorded on the work adder We are now interested in making the rat do just as much work as possible and we are passing the stimulus through both back legs thereby activating all of the muscles of the legs and back Even this did not satisfy us We have stimulated through the entire body of the animal and through various parts of the body but such animals fatigue rapidly and die within an hour The activation of both back legs causes the animal to develop hypoglycemia and death will occur from hypoglycemic shock in some animals unless glucose is administered Our present setup permits us to work 12 rats simul taneously and to feed them by constant intravenous infusion Shorr For how long a period of stimulation each time? Is this continuous ? ANESTHET ZED RAT SUPPORTS FOR RAT THREAD BOARD ELECTRODE ADJACENT TO GASTROCNEM US HEMOSTAT ON TBA ELECTRODE) UR NAL THREAD ON GASTROCNEM US TENDON GAUZE ROLL VEEDER COUNTER ANE WAY RECORD NG CLUTCH 100 9m FIGURE 7 102 ·Ingle No, intermittent, five times per second The duration of each stimulus is about 20 milliseconds Short This is going on continually throughout the study Ingle Yes (Slide, Figure 8) These data show that adrenal cortex extract can affect the work performance of the rat in the absence of the gut, Iner, kidneys and the adrenal glands Following evisceration all the animals were given enough glucose to prevent any degree of hypoglycemia, yet there was a striking effect of adrenal cortex extract upon the ability of these rats to work These data add to our evidence that the action of the adrenal cortical hormones, particularly upon vigor is not mediated by the liver or by any of the intra abdominal organs At the present time I do not have very much of an idea about the nature of the processes which the cortical hormones primarily affect Actually if you should go through the literature on adrenal cortex physiology seeking some meta bolic process which remains normal under all conditions of adrenal hormone insufficiency or excess the search might be in vain Armstrong Does this require whole adrenal extract or do any of the fractions such as Doca affect work capacity? Ingle We have used only whole adrenal extract in the eviscerate rat, but we know from studies on adrenalectomized rats that the 11 oxygenated and especially the 11 17-oxygenated adrenal compounds are far more potent than 11 desoxycorticosterone in sustaining work output I might say that the noch test was the procedure used to first demon strate to De Kendalls group that they were in possession of an active adrenal steroid Dr Mason isolated 17 hydroxy li dehydrocorticosterone in that laboratory in 1935 We demonstrated its activity immediately in the work test but nearly two years passed before any other laboratory found the compound to be active When 11 desoxycorticosterone became available, this was the first test used to demonstrate that its biological properties differed from those of the 11-oxygenated compounds It is of considerable personal interest to me to note that 17 hydroxy 11 dehydro corticosteronc (Mason s Compound L) is the most sought after of the corticosteroids and that its use is nos related to some of the most exciting developments in the clinic as well as in the laboratory Best When you said that these changes were not related to the supply of carbohydrate you did not climinate the effect of sugar Ingle That is true Certainly the exhaustion of carbohydrate stores in these adrenally insufficient animais can be a limsting factor When it is corrected the animal wall still fail for other reasons It can work a little longer when it is given glucose than when it develops hypo slycemia but when glucose is ghen other deficiencies which se do not understand become manifest 103

·6
4
RECORDER REVOLUTIONS
THOUSANDS
2
O
ACE
FIGURE 8
(Slide, Table IV) This final slide brings us to a story which seems
to me to be the most interesting that we have found We have studied
the effect of muscle work upon the tolerance of the eviscerate rat for
104

·glucose during a period of 2 hours In resting animals without insulin the glucose tolerance is best represented by a load of 16/100 If we add insulin, the glucose requirement is increased to 64/100/h If ne stimulate one leg without adding insulin the glucose requirement is increased from 16 to 48/100/and if lasulan is added the requirement goes to 110/100/h Now if we stimulate all of the musculature of both back legs without adding asalın the glucose requirement is increased ten fold over the resting state When insulin is added, there is another significant increase in glucose tolerance This may be entrely fortuitous but it is nevertheless interesting that the increment which insulin has made in the glucose tolerance n each of the three experimental conds tions is just about the same in each case TABLE W The effect of muscle work upon the tolerance of the eviscerste rat for glucose during a period of two hours Insulin Glucose Mascle Units Per Tolerance Samulation 24 Hours mgms/100/ O D 1 Hind Les 1 Hind Leg 2 Hind Legs 2 Hind Legs 4 0 16 64 0 43 4 110 160 220 Alarkourtz I think the point is important How long has that prepara tron been eithout any insulin when you begin that experiment ingle The pancreas was removed be evisceration and immediately following operation the animal was subjected to the stimulation of muscle for two hours Alarkou 112 It is not diabetic muscle? ingle No but I sul try to ansner that question in a few moments The working rat without insulin tolerates 160/100/h of glucose and the work output is normal When insulin is added hypoglycemia ensues and the rate of work is sharply decreased These observations suggest but do not prove that the pathuays of glucose utifızalıon fasored by insulia are competing Rith the contracting muscle for the carbohydrate We do not knon that is happening to the glucose but its disappear ance must be related to the activity of the muscle This suggests that the working muscle can utilize glucose perhaps in an optimal manner in the insulin dchaient rat 105 ·This brings us to the question just raised by Dr Alarkowitz What is the story on dsabetic muscle) We have used rats having alloxan diabetes and have selected animals which excrete as much glucose as is available from the carbohydrate glycerol of fat and protesn (estimating 'n L 1 would show up any kind of physiological deficiency any departure from normal and non we have sometimes thought that these totally diabetic rats work a little better than non diabetic rats The most work that we have obtained per unit time has been from diabetic rats In respect to glucose tolerance the effect of work upon glucose requirement is essentially the same as eve found in the above experiment All of the valves are shifted down a little The resting diabetic cat without insulin requires about 12 mgms of glucose per 100 grams of cat per hour instead of 16/100/h work causes a marked increase in the rate of disappearance of glucose from the blood not quite as great as in the previous experiment but the change is still very striking This brings me to the end of the story Perhaps I should mention a Isttle of what we hope to do Our studies as you have seen are exceed ingly superficial We do not know what is happening to the carbo hydrate except that it disappears from the blood I am very hopeful that Dr Stetten is going to come in with us on certain of our experi ments that have to do with the metabolic behavior of the eviscerate animal Certainly he is prepared to give these studses the depth that they deserve For our studies on the survival of the eviscerate rat ne now have apparatus that will accommodate 12 animals at a time and will permit us to push ahead more rapidly to determine the effects of other food factors in the infusion fluid and to test the possibility that the liver contains or secretes something that may affect the sun ival of the liverless animal Fremont Smith There are 500 chemical functions of the liver Ingle With that thought in mind it might be appropriate for me to close these temarks with a few lines from the Walrus and the Carpenter I have never quite become adapted to my mability to probe into each one of the myriads of problems that are presented in research The Walrus and the Carpenter Were walking close at hand They sept like anything to see Such quantities of sund If this sere only cleared an ay They said it nould be grand! 106 ·If seven maids with seven mops Swept at for half a year, Do you suppose, the Walrus said That they could get it clear? I doubt it, said the Carpenter, And shed a better tear DISCUSSION Stellen I would like to ask a question in regard to the efficiency of rauscle nork It would seem to me you have all the data here whereby you can compute the efficiency of the muscle work in the hind leg of the rat under various circumstances in other words did I understand it, with your Veeder counter technic plus the information about the weight you could calculate work directly if the drum was calibrated In one or two legs it came out pretty nearly 100 mg of glucose per hour per leg I wonder if you have some computations as to how much of that glucose oxidized to completion was conterted into mechanical svork Ingle We have not Dr Skadie once made some calculations which did not come out according to his expectations and which may have been due to an error inherent in our recording device We have recently had our plysıcs department determine the extent of error There us a substantial amount of coasting in the recorders We are able to keep the maximum relative error between recorders down to 15 percent and the average error is very much smaller than that well below 10 percent Shorr Could I ask you to dilate a bit on the diabetic animals? How 13.25 the diabetes achieved) Were these completely eviscerated? Were the Asdneys in or out? Was there loss of carbohydrate in the urine Ingle These animals were alloxan diabetic rats We have used some that Rere made diabetic by pancreatectomy During the tro hours of kor the animals have not excreted glucose in the urine Alarkourt: What evidence hase you that the alloxan diabetes was a nearly tatal diabetes o total diabetes Ingle We followed the glucose excretion of these rats that are force fed Our diabetic rats are force fed a fluid dier by stomach tube ture cach day so chase evact control over the food intake Me hon its carbohydrate content its protein content and our estimates of the amount of glucose that these animals should put out to be totally diabetic are based on the usual calculations which of course are open to some question if there is gluconeogenesis from fat then wc cannot say that these animals are totally diabetic Narcou iz It is a classical obseruation that the depancreatized Jos pots out neither more nor less sugar then worked on the treadmill 107 ·On many occasions I have seen pancreatectomized dogs under strychnine convulsion excrete more sugar than at rest This is due to the lactic aad liberated and subsequently resynthesized to sugar No one that I know has caimed that muscular exercise in total diabetics dogs or humaas, Jessens the glucose excretion or blood sugae That does not mean your findings are incorrect but it justifies our questioning these things in this way Ingle The only sort of observation that mught be in line with this is that clinicians a ho manage diabetes believe that the insulin require ment is less when these diabetics take substantial amounts of exercise Best There you have some msulin present That is what is bothering Dc Short Dc Darkowitz and bothering me The picture is completely different in the total diabetic animal You cannot train them They cannot do much work, and the worst thing you can do for them is to make them work, but give them a little insulin and the more work they do, within reason, the better they are in sugar tolerance, training and everything else There is a sharp dividing line in the previous literature, but your work is much more careful than the previous work Ingle I would admit the possibility that there may be some insulin left Best If you have a trace of insulin the picture is entirely different Gyorgy That is right Shorr I wonder whether ne should not entertain the possibility that the discrepancies between the depancreatized dog and alloxan diabetes as well as some of the phenomena seen under Ds Ingles experimental conditions might be attributable to a parallel suppression of anterior glycosuria to occur It is as if the alloxan animal in the basal state had small amounts of insulin present which could handle the basal carbo hydrate need by virtue of a decreased! uction of anti insulin factors by the pituitary or the pituitary drenal corte In this connection you may ndung by and Chandler of a progressive retum o al type of te metabolism in a few depancreatize happened for 15 20 days with the abolition of ketosis explanation for t the that the stress resulted i fant diabete or in additio tian changes the so called d to a pr this time! fat tona Is no estan 108 ·Were Attention may also be called to some of the metabolic characteristics of diabetic cardiac muscle Its R Q n sitro in the presence of glucose is in the neighborhood of 070, that of fat Yet at that RQ the turn over of high energy phosphates goes on at just about the same rate as in muscle from a normal dog burning at an RQ of 10 These observa trons, together with the extremely lou glycogen values sn the diabetic skeletal muscle, thren grave doubt on the necessity of diabetic skeletal muscle to depend on carbohydrate for energy Ingle We have gone much further in activating muscle here than putting a dog on a treadmill I would certainly agree to the need for caution in assuming that these animals were entirely without insulin At certain times I have held that it is virtually impossible to be certau that one has nd an animal of any hormone This is particularly true of the sex hormones, and to know when every bit of adrenal tissue has been removed or when every bit of thyroid tissue has been removed 15 difficult I am quite certain that many hypophysectomies have not rid the animal of every cell capable of secreting pitustary hormones I suppose that in the dog the removal of the pancreas can be seally commi It is not in the rat You can get pretty close to you section everything destroy sugu to hear of some studies on rats which owall of the usual criteria, had been completely thyroidectomized Yet when radioactive iodine was administered and the pictures were made of sts location the experimenters were able to find accessory bits of thyroid dissue lying down in the chest Shorr Do you think that the effects of hypoglycemia were exerted on the muscle per se or via the nerous system? Were you able to per fuse the brain separately with a glucose solution do you think you would find that the muscle contractibility would persist? A number of years ago Cattell and I found that we could get prolonged contracts bility with skeletal muscle strips from the dog for long periods electrical stimulation even in the absence of glucose from the medium Could impaired brain metabolism with hypoglycemia have been se sponsible for the muscle failure you obser eda ingle I should think so Most of that I know about that subject 15.25 learned in the Mayo Institute When I was there Di Corvin was working with Dr Bollman on the origin of hypoglycemic convulsions Those data indicated they were of nenous ongin Has not considerable noch been done in supporting that Dr Bollman? Bollman Some not an an ful lot thouch We would like to have I was very tu on more 109 ·On many occasions I have seen pancreatectomized dogs under strychnine convulsion excrete more sugar than at rest This is due to the lactic aad liberated and subsequently resynthesized to sugar No one that I know has claimed that muscular exercise in total diabetics, dogs or humans, lessens the glucose excretion or blood sugar That does not mean your findings are incorrect but it justifies our questioning these things in this way Ingle The only sort of observation that might be in line with thus is that clinicians who manage diabetes believe that the insulin requirement is less when these diabetics take substantial amounts of exercise Best There you have some insulin present That is what is bothering Dr Shore, Dr Markowitz and bothering me The picture is completely different in the total diabetic animal You cannot train them They cannot do much work, and the worst thing you can do for them is to make them work, but give them a little insulin and the more work they do, within reason, the better they are in sugar tolerance, training, and everything else There is a sharp dividing line in the previous literature, but your work is much more careful than the previous work Ingle I would admut the possibility that there may be some insulin left Best If you have a trace of insulin the picture is entirely different Gyorgy That is right Sborr I wonder whether we should not entertain the possibility that the discrepancies between the depancreatized dog and alloxan diabetes as well as some of the phenomena seen under Di Ingle's experimental conditions, might be attributable to a parallel suppression of anterior pituitary function in alloxan diabetes and in the eviscerated preparation The situation in the alloxan diabetic rat is suggestive of the Houssay animal, one has to load both with carbohydrate for any significant glycosuria to occur It is as if the alloxan anımal in the basal state had small amounts of insulin present which could handle the basal carbohydrate need by virtue of a decreased production of anti insulin factors by the pituitary or the pituitary controlled adrenal cortex In this connection you may recall the finding by Chambers and Chandler of a progressive return of a more normal type of carbohydrate metabolism in a few depancreatized dogs which happened to survive for 15 20 days with the abolition of glycosuria and ketosis While there is no established explanation for this phenomenon, the most likely is that this metabolic stress resulted in a depression of anterior pituitary diabetogenic activity or in addition adrenal exhaustion These same changes occur during the so-called premortal phase in dogs subjected to a prolonged fast At this time their RQ's rise from the level of fat tonards unity 1 108 ·Attention may also be called to some of the metabolic chacacteristics of diabetic cardiac muscle Its RQ in tiho in the presence of glucose is in the neighborhood of 0 70 that of fat Yet at that RQ the turn over of high energy phosphates goes on at just about the same rate as in muscle from a normal dog burning at an RQ of 10 These observa trons together with the extremely low glycogen values in the diabetic skeletal muscle thren grave doubt on the necessity of diabetic skeletal muscle to depend on carbohydrate for energy Ingle We have gone much further in activating muscle here than putting a dog on a treadmill I would certainly agree to the need for caution in assuming that these animals were entirely without insulin At certain times I have held that it is virtually impossible to be certain that one has nd an animal of any hormone This is particularly true of the sex hormones and to know when every bit of adrenal tissue has been removed or when every bit of thyroid tissue has been removed is difficult I am quite certain that many hypophysectomies have not rid the animal of every cell capable of secreting pituitary hormones I sup pose that in the dog the removal of the pancreas can be really complete it is not in the rat You can get pretty close to it with alloxan but unless you section everything you would not be able to say that all the islands were destroyed regardless of the manner in which the animal behaved I was very much interested a few months ago to hear of some studies on rats which according to all of the usual criteria had been completely thyroidectomized Yet when radioactive rodine was administered and the pictures were made of its location the experimenters were able to find accessory bits of thyroid tissue lying down in the chest Shorr Do you think that the effects of hypoglycemia were exerted on the muscle per se or via the nervous system? Were you able to per fuse the brain separately with a glucose solution do you think you would find that the muscle contractibility would persist? A number of years ago Cattell and I found that we could get prolonged contracts bility with skeletal muscle strips from the dog for long periods on electrical stimulation even in the absence of glucose from the medium Could impaired brain metabolism with hypoglycemia have been re sponsible for the muscle failure you observed? Ingle I should think so Most of what I know about that subject was learned in the Mayo Institute When I was there Dr Corwin was working with Dc Bollman on the origin of hypoglycemic convulsions Those data indicated they were of nervous origin Has not considerable work been done in supporting that Dr Boliman? Bollman Some not an awful lot though we would like to have more 109 ·Ingle We would like very much to know what our contracting muscle can use other than glucose We know that if we work an intact fasting rat, stimulate only one muscle, it can work for periods of from five to eight days with the muscle contracting continually at the rate of three times per second During that time the other musculature of the body atrophies The working muscle sustains its size and certainly gains some weight Whether the weight is water or whether it is actually hyper trophy we dont know All of the fat stores of the body disappear during that time The animal can mobilize its fats for utilization in some manner Whether it is conversion to glucose we dont know We have tried to improve work performance with certain preparations of fat emulsions given intravenously It never works It only makes the rat sick and it dies Best These that went five days, were they liverless? Ingle No those were not I think I got through without noting our longest survival which is 65 hours * Bollman That is just the same length of time as our liverless dog Markowitz Did he have a transfusion> Bollman That one had Markowitz The comparison is not quite correct Bollman It is not quite correct Markou itz You gave some prothrombın? Bollman He loses that very rapidly Best I was interested in De Ingle letting his mind run along the channels that there might still be a few molecules of a suprarenal hormone after removal of the gland I think all of us are getting wedded to the idea We cannot imagine the body without our pet substance I confess I have often thought that the depancreatized animal when it loses the last molecule of insulin dies immediately afterwards but I know that is not true You can make glycogen mn litro I remember thinking of those experiments that Di Mann did I dont know whether you were on that paper Di Bollman or not on glycogen deposition of the liver in the depancreatized dog You made a struggle to make it without insulin You got a little I considered the possibility that a few molecules of insulin might still have been present Seriously I dont really believe there was any insulin Bollman I wanted to make one comment I think we have gone a little bit afar I am sure De Ingle did not say the muscle was using the glucose in the test that he did for the blood sugar It is obvious that the glucose was disappearing from the entire animal not neces * Now 82 hours and 48 m nutes acco d og to the most recent report from D: Engle- Editor 110 ·Sarily that it was disappearing in the muscle which was being stimulated Along that line, I wanted to ask you if you checked the heart rate on your stimulated and non stimulated rats Don't you get an increased heart rate with stimulation of the muscle? Ingle Yes Best Do you have a record of the heart rate on this device? Ingle This is an all or none response We do have a little device though for measuring the heart rate in these rats Bollman I think it is perfectly justifiable to make the inference that the muscle is doing something that causes the greater disappearance of sugar from the blood We are going little afar to assume that the muscle is burning that sugar to do its work Ingle I would certainly agree with that one of the other things that I dream about having some day is a perfusion setup which will permit us to isolate a leg and perfuse it with oxygenated blood with added nutrients, and if we could keep such a preparation contracting for a period of say as long as eight or ten hours that would be sufficient time to tell whether the adrenal hormones act in such a preparation and what insulin may be doing to it I think it would still be a complex brochemical system but it would be much more simple than having all of the rat Bollman I have been impressed with the fact that most of the failures of muscle contraction by repeated stimulation have been related to circulatory failure rather than muscular failure As soon as you clamp the circulation to the muscle, the muscle is practically through At this rate of 3 or 5 stimulations per second it only has about a minute to go and then it is dead Best Have you any data on the rate of disappearance of body fat with the standard amount of exercise in your liverless as compared to the intact animal? Ingle No, I have not Shorr There is one muscle that can be shown to utilize fatty acid in ziro This is the smooth muscle of the rabbit intestine If such a rabbit gut preparation is provided with no substrate, and washed once or twice with Ringer solution, the amplitude of its contractions progres sively decreases When it has fallen to 5 10 percent of the original excursion, the addition of an effective substrate for the support of muscular contractility is promptly followed by an increase in the ampli tude of the contractions Maximal contractions are achieved with glucose, and this substrate can serve as a standard Using this prepara tion De R F Furchgott and I studied a variety of substrates, including odd and even numbered fatty acids (Proc Soc Exp Bol & Med 6i, 111 ·280 (1946)) Acetate had the same effect on amplitude as did glucose Even numbered fatty acid (4, 6 and 8 carbons) gave half the amplitude Odd numbered fatty acids (3, 5, 7 and 9 carbons) gave different but very interesting results Propionic acid did not restore contractions but actually inhibited the effect of butyric, apparently by competing with 2 carbon fragments for the surface of the enzyme systems involved As we went up the odd numbered series, increasing amplitudes were obtained, and with the 9 carbon pelargonic acid, the amplitude was almost as good as with butyric acid We felt that this was due to the increasing proportion of 2 to 3 carbon fragments The inhibitory effect of propionic acid on butyrıc could be overcome by adding appropriately greater amounts of butyric acid These experiments are taken to indicate that in smooth muscle, at any rate fatty acids can be utilized for con tractile energy Attempts by A V Hill, for example, working with the frog, to demonstrate a reduction in fat content of the contracting muscle were, I think, loaded against him because of low energy metabolism of the frog muscle hence the small changes in fat content which could take place over the period of the experiment But I think it should be quite possible, using the mammalian muscle strap to determine directly whether fat is utilized Best In smooth muscle? Shorr In skeletal muscle Best No one has ever demonstrated utilization of fat in skeletal muscle Shorr No one has ever tried to, except for the balance studies of Hill on the frog I believe, however, that the conditions holding for the mammalian muscle strip of the dog, higher energy production, hugher turnover, and the ability to contract for long periods of time—alí these are favorable for a direct proof or disproof of fat utilization This experiment should be carried out because of its great interest Markowitz No one has ever demonstrated it for the isolated heart perfused with Ringers solution It is benefited by glucose and glucose alone It has been tried for acetates and propionate, succinates and fatty acids (which are difficult to do because you precipitate the calcium) but the fact is that an isolated beating heart does have some re semblance to skeletal muscle, and the only thing which does it any good is glucose Shorr However, fat utilization has been shown to take place in the isolated heart lung preparation of the diabetic dog by Cruikshank and Startup (J Physiol 81, 153 (1934)) In the absence of insulin, the respiratory quotients were always in the neighborhood of fat 0 70 112 ·When insulin was added to the perfused blood, the RQ rose to unity The differences between these results and those just cited by Dr Markowitz may arise from difficulties presented by isolated heart per fusion experiments using Ringer solution The addition of fatty acids to a Ringer solution may not present this material to the heart in a physiological state We observed for example that when fat was extracted from blood by alcohol and ether and added to the rabbit intestinal strip, contractions were inhibited This inhibition could be overcome by the addition of serum albumin Whatever the cause, I am inclined to accept the work of Cruikshank and Startup, because of its more physiological character and the care with which it was cartied out Armstrong Does not the isolated heart utilize ketone bodies > Markourtz It does utilize ketone bodies I am talkiag about Ringers solution There is no blood in that I refer to the effect of Ringers solution alone, on the rabbit heart, the only thing that benefits that heart to my knowledge and I am very much open to correctionis glucose Whether ketone bodies can be used or not I would not debate They probably are used to benefit the heart It would not live any longer by the addition of ketone bodies To keep a surviving heart beating for long, one must add glucose and nothing else is of any use Shorr Dont you think, De Markowitz that this problem is compli. cated by the metabolic peculiarities of heart muscle Markowitz No Shorr Id like to ate from my own studies certain striking differences between diabetic skeletal and cardiac muscle as indicative of metabolic differences between these two tissues (Cold Spring Harbor Symposia on Quantitative Biology, 7 323 (1939)) Although the respiratory quotients with both tissues are usually in the neighborhood of 0 70, there is a significant differerce in their ability to glycolyze The skeletal muscle forms very little lactic acid under anaerobic conditions whether or not glucose is in the medium In contrast, cardiac muscle has a high rate of glycolysis which is significantly increased by the addition of glucose This difference appears to reflect in large measure the extremely low glycogen content of skeletal muscle in the diabetic dog and the high and indeed greater than normal content of the cardiac muscle from such an animal (N F Fisher and R W Lackey, Am J Physiol 72, 43 (1925)) The total fermentable carbohydrate in diabetic heart is likewise greater than normal in our experiments averaging 13 5 mg per gm as compared with 79 mg per gm in the normal controls These high glycogen and total fermentable carbohydrate values for the diabetic heart indicates that this tissue in contrast with skeletal muscle, retains the capacity to store glicogen in the absence of insulin even 113 ·though it cannot axıdıze glucose without the help of this hormone Apparently the carbohydrate metabolism of the various tissues is not equally affected in diabetes Indeed the brain can oxidize glucose without insulin Markourtz if blood is used from depancreatized dogs only, and a heart lung preparation is made from a depaacreatized dog that heart survives an hour or to and nothing you can do will make it survive longer unless you add insulin Best There is glucose Markowitz This is a heart lung prepacation with sugar in the blood but without insulin There must be some difference between diabetic muscle and non diabetic assuming that a heart muscle is a fair test object Those facts have never been disproven Knowlton and Starling made them I have been able to repeat them and my results were 50 entirely uniform I did not bother to report them The diabetic heart perfused with diabetic blood does not survive much longer than an hour Bess The amount of glycogen in the diabetic heart as compared to a normal heart is not very different Shorr The figures I cited above for total carbohydrate were about 30 percent bigher in the diabetic heart than in the normal I think the important thing is that the glycogen content is at least normal whereas the glycogen content of the diabetic skeletal muscle is negligible ·LIVER GLYCOGEN AND THE PROTECTION OF LIVER CELLS SECTION VI Best I think many of us who are interested in the protection of the liver found our allegiance divided between protein and perhaps methiofine and choline on the one hand and the storage of glycogen, with or without insulin on the other I was quite sure that I could not do justice to the carbohydrate side of it and I was anxious to have the field seviewed so I thought it would be a good idea to invite Ds Bollman to review it completely and tell us everything about it He has reacted a little bit to this He is not going to cover everything today He is going to take part of the hour to tell us about work on fat absorption Bollman Had I been able to find in the literature or perform expers ments myself which would furnish direct evidence and absolute proof that the resistance of the liver to hepatic toxic agents varies directly with the carbohydrate content of the diet or the glycogen content of the liver, I could give a short and satisfactory exposition Instead I will have to talk around the subject In terms of the daily caloric need of the body, the storage space for carbohydrate is limited and failing a source of supply of glucose for blood sugar as in the hepatectomized animal, life soon terminates The liver is able to fabricate glucose from non carbohydrate sources but such processes are accomplished only with the expenditure of energy The dset of all animals contains carbohydrate at least in concentrations equivalent to the proportion of carbohydrate contained in their body Most mammals including man consume a large excess of carbo hydrate in relation to the composition of the body which contains (man) about 45 times as much protein as it does carbohydrate (H K Mitchell, T S Hamilton, F R Steggerda and H W Bean, J Biol Chem 158 625 (1945)) Carbohydrate is known to spare protein and will reduce the nitrogen loss in the vilde when fed to othervise fasting animals Carbohydrate fed with protein all also allow a greater protein storage than when protein is fed alone la rats fed on a high carbohydrate diet sufficient to maintain their neight, Isotope studies indicated that about 69 percent of the liver glycogen was replaced daily (DeW Stetten Jt and J E Boxer J Biol Chem 155, 231 (1944)) Protection to the liver by carbohydrate may be accomplished in two ways, one in providing glucose for the maintenance of the entire body at no expense to the liver and second, by sparing protein which may 113 ·be used for rebuilding of the damaged liver and for other purposes Many clinical observations of liver diseases and also observations on experimental animals following surgical damage of the liver ie partial hepatectomy Ecks fistula biliary obstruction etc indicate better sur vival rates when carbohydrate is given Whether the glucose or carbo hydrate administration had any direct effect in sparing the liver or merely introduced factors which maintained life long enough for the liver to repair itself cannot be ascertained I am sure that many other factors are present and that life cannot be maintained indefinitely on glucose alone Certainly all the other food factors must also be supplied if the animal is to live and especially so if he has to repair a liver injury We know carbohydrate to any measurable amino acids of the diets more rich in protein are converted to urea by the liver and excreted in the urine and with diets relatively poor in protein the amino acids are conserved so that the actual protein content of the body is the same Since the liver is involved in both the destruc tion and the conservation of amino acids it is probable that the diseased or injured liver would not have the flexibility of a normal liver and that recovery and repair would be expedited more by supplying the liver with glucose and amino acids at some specific ratio which may not be as flexible as the normal dietary ratio I want to call attention to the constantly changing composition of the liver and its extreme flexibility in that it seems to function well over a wide range of chemical composition (Slide) This slide shows the changes in liver weight and glycogen protein and fat content of the liver following a meal Rats were trained to eat a full meal in one hour and on the day of the experiment were fed a diet containing 70% carbohydrate 20% protein and 10% fat at 9.00 am This is indicated as tie 0 on the slide and groups of rats were taken at subsequent intervals over the 24 hour period for analysis After feeding there 15 an increase of about 20% in the weight of the liver by about the 4th hour and a subsequent return to the original size in about 12 hours The glycogen content changes even more after the food is absorbed and later there is a decline below the fasting level with a return within 24 hours to the original fasting level The protein and fat content of the liver change in the same way but the changes are less extensive This second slide is an example of the thing that I am not supposed to say that carbohydrate definitely can injure the liver Best That is the reverse of what you are supposed to say Bollman That is the reverse of what I want to say I want to bring it out for this reason These sections were taken from a dog who had 116 ·recewed continuous intravenous injection of glucose for about 36 hours The amount of glucose was 50 given that he always had a high blood sugar, somes here between 200 and 300 mg per 100 cubic centimeters The dog was excreting a large amount of sugar The dog received about 3 grams of glucose per kilo per hour In the meantime the liver glycogen did accumulate At the time these sections were taken, after 36 hours of continuous infusions the liver was full of glycogen, the dog had a very definste bilirubinemia and a very marked dye retention This is ample proof, I believe, that carbohydrate can injure the liver Hanger You think the functional impairment is secondary to circula tory disturbances > Bollman I have no real idea as to the mechanism of the injury It would seen from the exammation of the liver that it is so full of glycogen it has neither time nor ability to do anything else This is obviously an excessive strain and is not the particular thing we were talking about Best You know of the active debate now going on a clinical circles as to the use of dextrose in the treatment of diabetic coma, one school assecting that the use of very large amounts of dextrose intravenously may kill the patient? Bollman Yes Gjorgy Do you agree? Best I dont agrée, no, without reservations I think that is another subject Watson From liver injury It was more related to the potassium situation and adrenal function Best Howard Root at the Academy one rught talked of liver injury Gyorgy From high glucose? Best in the treatment of diabetic coma Gyorgy The Boston group emphasizes the potassiun deficiency Best I think Root at that time had some specimens from animals to illustrate his talk Bollman I dont believe that any clinician would be idiotic enough to inject three grams of glucose per kg each hour for 36 hours That would mean the injection of about 40 liters of 10 percent glucose a day Watson I was thinking about a big excess I wont say 3 grams per kslo per hour In a number of places they are giving 20 percent glucose intravenously with an inlying plastic tube giving it day after day Best With vitamus Watson They give a lot of vitamins Best If you are losing a lot of sugar in the urme anyway it does not 117 ·grams, four make much difference whether you are giving three grams or two grams Some individuals eliminate more and others less Bollman That is right Neefe Do you think that the bilirubinemia and the dye retention could be the result of a competition of functions Bollman It is a failure of those functions Let me hasten to add that this is not a permanent injury The next day those animals would recover and the glycogen will fall fairly rapidly if you discontinue the injection You can, I should say, injure the liver by carbohydrate I also should add at the same time this glucose was given with about 2 per cent saline, but no other inorganic salts, and necessarily a large volume of fluid We washed a lot of things out Whether the liver injury was due to the things we washed out or the things we washed in it is still a question Of course that same thing comes up in clinical cases that you are doing two things You are putting some things in and also washing out some things, unless you use a perfectly balanced solution I know of no such solution (Slide) This represents a fatty liver which can be produced by reducing the protein of the diet This was from a dog that received 90 percent of his calories in the form of fat In about 8 weeks such dogs develop a fatty liver That fatty liver had perfectly good function as far as we could tell There was no dye retention Best What amount of fat is that? Bollman That one was about 22 percent My experience with stain ing suggests that one can make a fairly good guess as to the fat content in the dog liver when it increases from about four percent to eight or nine percent After it gets beyond eight or nine percent, an accurate estimation is impossible EDITOR Dr Bollman next reviewed certain data previously published (see Trans actions of the Fifth Conference on Isver Injury, Pages 11 to 14 1946 ) น hon na farmnd wh ch rats and a arc spl. adding oxygen to the spirometer and taking the Co. out with a CO absorber, we have essentially a Benedict table We can put our rats in the chamber and expose them to air plus any noxious agent that we wish to add to the air We may also take away the oxygen if we wish We have used this particular setup in two series of experiments so far One was to test the effect of ether with reference to the survival of the rats We put in the ether by recirculation and maintained the concentration at a level at which 50 percent of our well fed normal 118 ·fats would die within a period of 4 hours Then we tested a group of cats that had been fasted for four or five days and found that they succumbed at exactly the same concentration of ether as did the normal and well fed sat In other words, our feeding gave us absolutely no protection against this same concentration of ether That perhaps is not surprising but at any rate there was very definite evidence that the well fed rat was no better, as far as the ether exposure, than was the fasting rat Gyorgy What did they die from? Bollman The ether concentration Gyorgy It was not necessarily a liver death? Bollman No that is not necessarily a liver death by any means We have also done another series of experiments testing the effect of anoxia By merely replacing the oxygen with nitrogen, we dropped the oxygen content dona to about 5 percent The rats would succumb in about half an hour At a 50 percent level, they seem to be able to accommodate to this and will live five or six hours Hou ever, at a very low oxygen content we found no difference between fasted cats and the fed rats In a rather small series se injected glucose immediately before exposing them to this persod of anoxia and again there was no difference in the Survival The ability of the rat to withstand the anoxia certainly is not a function of the liver entirely but the lever is certainly involved The anoxic rats have livers that are nearly black Such lovers are large and swollen and certainiy appear anoxic Shorr Is this a 24 hour fast? Bollman No a longer period Shors Four days? Bollman The results after 24 hours were negative That is the reason we went to a longer period of time To get back to the original story, I think that we are all convinced that carbohydrate is very valuable food and does offer protection to the entire body and to the liver Best You have not found any evidence that that is so? Bollman We have no direct evidence that that is true Pavlov and Root were the first to demonstrate in 1890, that the Eck fistula animals survived much better without meat and on a high carbohydrate diet than on a high protein diet That is not proof but clinical observation We have clinical observations on postoperative dogs where the liver is involved It appears that with very severe cazury to the lives, we can better the animals chances of surviving if we give it glucose and maintain the animal on a high carbohydrate diet than if we omst glucose, and give them a high protein diet That is marked We have some 119 ·examples of that Let us take for example the dog that has had hus common bile duct ligated for about three months Such an animal is not a sturdy specimen If we put such dogs on a diet composed entirely of meat-not a bad diet for a dog-they usually die in less than a week If we give them a low protein, high carbohydrate diet consisting not entirely of carbohydrate by any means but containing protein and other food factors, we keep those animals alive for many months almost a year They usually develop a perforated duodenal ulcer and die Best What about the caloric value of the food that they are able to absorb in the two cases? If you tie the bile duct you get very little absorption of fat, I take it Bollman I really dont know just exactly what the caloric value of the food that they absorb would be I don t believe that that is a factor because we are talking about a week as compared to several months Markowitz Do they get ascites? Bollman They do Hanger You say these animals develop ascites in contrast to others? Bollman The meat fed animals develop ascites Gjorgy They have only meat, nothing but meat? Bollman Yes Best Meat extract will do that Bollman It is not protein It is meat, and meat extract will do the same thing, so I dont believe that the protein has anything to do with that. Hanger What is the mechanism of the ascites? Bollman I really dont know Gyorgy Cirrhosis of the liver Bollman Billary cirrhosis, yes Patek Would that be true of cooked meat as well as raw meat? Bollman Yes Watson Dont you think that must be a species difference? Have any of your clinicians ever cited any clear cut example of that in human arrhosis? I have watched for that for years and I have not seen it We feed such patients a lot of meat Bollman Let me add one other additional qualifying statement this particular development of ascites is under a very special set of circum stances The dogs have to have biliary obstruction for three months I have never seen it in less than three months They have a certain amount of liver damage, ascites will also occur if we continue these animals without meat for sooner or later they will develop ascites spontaneously 120 ·Best Isnt this the same story exactly as in Eck fistula dogs Bollman The same thing I think, and that is as much evidence! hate to admit it as I could find of the protective action of carbo hydrate on the liver Best We have assigned a very difficult title Liver Glycogen and the Protection of Liver Cells to Dr Bollman He has gone over the data and we have no evidence at all that glycogen does protect the liver Bollman I should have mentioned that Ravdın s group also studied the effect of carbon tetrachloride on the liver There was no relationship between the amount of glycogen in the liver and the protection that rats received Gyorgy What was the relationship in the case of fat? Bollman The idea that glycogen is reciprocal of fat in the liver is also erroneous because one can have very high glycogen present in a very fatty liver Best There are limits there? Bollman There are limits Best You cannot have a lot of glycogen in a very fatty liver for example 40 percent or even 30 percent You dont get glycogen in a very fatty lives You cannot have a lot of fat in a liver that is very rich in glycogen but in the intermediate stages you can in both cases Bollman In the intermediate stages yes At the extremes of course one is dealing with something else again Patek Is it possible that the protective effect of carbohydrate in certain instances might be due to a protein sparing effect> Best I dont know shether you should spare the protein Patek If one employed two series of rats on weighed diets with the same low protein content the remaining calories made up with carbo hydrate in one group as against fat in the second group it might sell be that the high carbohydrate diet would be more readily tolerated because of a protein sparing effect Best I think if I were going to go at this problem I would start with the resistance of the untreated diabetic animal to liver possons Gyorgy for the fatty liver? They have a fatty liver Best Yes they have a fatty liver but they go on to get arthosis The untreated diabetic animal will get cirrhosis if he just lives long enough in the old days the diabetics got arrhosis not the severe diabetics since they died but the muld diabetics frequently got cirrhosis I am talking of human beings Bollman One of the reasons I accepted this assigoment was not the fact that there was anything particular to say but I would like some suggestions to find out how we could check this 121 ·examples of that Let us take for example the dog that has had his common bile duct ligated for about three months Such an animal is not a sturdy specinen If we put such dogs on a diet composed entirely of meat--not a bad diet for a dog-they usually die in less than a week If we give them a low protein high carbohydrate dret, consisting not entirely of carbohydrate by any means but containing protein and other food factors, we keep those animals alive for many months, almost a year They usually develop a perforated duodenal ulcer and die Best What about the caloric value of the food that they are able to absorb in the two cases? If you tie the bile duct you get very little absorption of fat, I take it Bollman I really don't know just exactly what the caloric value of the food that they absorb would be I don't believe that that is a factor because we are talking about a week as compared to several months Markowitz Do they get ascites? Bollman They do Hanger You say these animals develop ascites in contrast to others? Bollman The meat fed anunals develop ascites Gyorgy They have only meat, nothing but meat? Bollman Yes Best Meat extract will do that Bollman It is not protein It is meat, and meat extract will do the same thing, so I don't believe that the protein has anything to do with that Hanger What is the mechanism of the ascites? Bollman I really don't know Gyorgy Cirrhosis of the liver Bollman Biliary cirrhosis, yes Patek Would that true of cooked meat as well as raw meat> Bollman Yes Watson Dont you think that must be a species difference? Have any of your clinicians ever cited any clear cut example of that in human arrhosis? I have watched for that for years and I have not seen at We feed such patients a lot of meat Bollman Let me add one other additional qualifying statement this particular development of ascites is under a very special set of circum stances The dogs have to have biliary obstruction for three months I have never seen it in less than three months They have a certain amount of liver damage, asates will also occur if we continue these animals without meat for sooner or later they will develop ascites spontaneously 120 ·Best Isn't this the same story exactly as in Eck fistula dogs Bollman The same thing I thinh, and that is as much evidence hate to admit it-as I could find of the protective action of carbo hydrate on the liver Best We have assigned a very difficult title, Liver Glycogen and the Protection of Liver Cells to De Bollman He has gone over the data and we have no evidence at all that glycogen does protect the liver Bollman I should have mentioned that Ravdın s group also studied the effect of carbon tetrachloride on the laer There was no relationship between the amount of glycogen in the liver and the protection that rats received Gyorgy What was the relationship in the case of fat? Bollman The idea that glycogen is reciprocal of fat in the liver is also erroneous because one can have very high glycogen present in a Tery fatty liver Best There are limits there? Bollman There are limits Best You cannot have a lot of glycogen in a very fatty liver, for example, 40 percent or even 30 percent You dont get glycogen in 2 very fatty liver You cannot have a lot of fat in a liver that is very rich In glycogen but in the intermediate stages you can in both cases Bollman in the intermediate stages yes At the extremes of course one is dealing with something else again Patek Is it possible that the protective effect of carbohydrate in certain instances might be due to a protein sparing effect? Best I dont know whether you should spare the protein Patek If one employed two series of rats on weighed diets tith the same low protein content the remaining calories made up with carbo hyd it might well be dily tolerated ata AAA ho bec Best I think if I were going to go at this problem I would start with the resistance of the untreated diabetic animal to liver poisons Gyorgy For the fatty liver? They have a fatty liver Best Yes they have a fatty Iner but they go on to get arrhosis The untreated diabetic animal will get arrhosis if he just lives long enough in the old days the diabetics got cirrhosis not the severe diabetics since they died but the mild diabetics frequently got arrhosis I am talking of human beings Bollman One of the reasons 1 accepted this assignment tras not the fact that there was anything particular to say but i would like some suggestions to find out how we could check this 121 ·say examples of that Let us take for example the dog that has had his common bile duct ligated for about three months Such an animal is not a sturdy specimen If we put such dogs on a diet composed entirely of meat--110t a bad diet for a dog-they usually die in less than a week If we give them a low protein high carbohydrate diet, consisting not entirely of carbohydrate by any means but containing protein and other food factors we keep those animals alive for many months, almost a year They usually develop a perforated duodenal ulcer and die Best What about the caloric value of the food that they are able to absorb in the two cases? If you tie the bile duct you get very little absorption of fat, I take it Bollman I really don't know just exactly what the caloric value of the food that they absorb would be I don t believe that that is a factor because we are talking about a week as compared to several months Markowitz Do they get ascites? Bollman They do Hanger You these animals develop ascites in contrast to others? Bollman The meat fed animals develop asates Gyorgy They have only meat, nothing but meat> Bollman Yes Best Meat extract will do that Bollman It is not protein It is meat, and meat extract will do the same thing, so I don't believe that the protein has anything to do with that. Hanger What is the mechanism of the ascites? Bollman I really dont know Gjorgy Cirrhosis of the liver Bollman Biliary Cirrhosis, yes Patek Would that be true of cooked meat as well as raw meat> Bollman Yes Watson Dont you think that must be a species difference? Have any of your clinicians ever cited any clear cut example of that in human arthosis? I have watched for that for years and I have not seen it We feed such patients a lot of meat Bollman Let me add one other additional qualifying statement this particular development of ascites is under a very special set of circum stances The dogs have to have biliary obstruction for three months I have never seen it in less than three months They have a certain amount of liver damage, asates will also occur if we continue these animals without meat for sooner or later they will develop ascites spontaneously 120 ·Best Isa t thus the same story exactly as in Eck fistula dogs > Bollman The same thing I think and that is as much evidence hate to admit it--as I could find of the protective action of carbohydrate on the liver Best We have assigned a very difficult title Laser Glycogen and the Protection of Liver Cells to Dr Bollman He has gone over the data and we have no evidence at all that glycogen does protect the liver Bollman I should have mentioned that Ravdın s group also studied the effect of carbon tetrachloride on the liver There was no relationship between the amount of glycogen the liver and the protection that rats received Gjorgy What was the relationship in the case of fat? Bollman The idea that glycogen is reciprocal of fat in the liver is ilso erroneous because one can have very high glycogen present in a ery fatty liver Best There are limits there? Bollman There are limits Best You cannot have a lot of glycogen in a very fatty liver for example 40 percent or even 30 percent You dont get glycogen in a very fatty lives You cannot have a lot of fat in a liver that is very rich in glycogen but in the intermediate stages you can in both cases Bollman in the intermediate stages yes At the extremes of course one is dealing with something else again Patek is it possible that the protective effect of carbohydrate in certain instances might be due to a protein sparing effect Best I dont know whether you should spare the protein Patek If one employed two series of rats on weighed diets with the same low protein content the remaining calories made up with carbo hydrate in one group as against fat in the second group it might well be that the high carbohydrate diet would be more readily tolerated because of a protein sparing effect Best I think if I were going to go at this problem I would start with the resistance of the untreated diabetic animal to loser poisons Gyorgy For the fatty liver? They have a fatty liver Best Yes they have a fatty liver but they go on to get cirrhosis The untreated diabetic animal will get cirrhosis if he just lives long enough in the old days the diabetics got cirrhosis not the severe diabetics since they died' but the mild diabetics frequently got cirrhosis I am talking of human beings Bollman One of the reasons I accepted this assignment was not the fact that there was anything particular to say but I would like some suggestions to find out hon we could check this 121 ·Best I think almost any clinician would say that the uatreated dsabetic was more susceptible to a lot of toxic influences than the well treated one One difference between the two is that one has Iner glycogen and the other has not Gyorgy Or one has fat and the other has no fat Best There are lots of other variables Bollman They do not make very good experimental animals I have tried that and I have tried the Eck fistula animals thinking that I could use alcohol or carbon tetrachloride The only trouble is that with that their susceptibility increased, all you have to do is blow a little chloro form at them and they turn up their toes and die Best You get a dramatic difference I should think between a treated diabetic and an untreated diabetic Bollman Of course Best You cannot say that this is due to depletion of lever glycogen or to an increase of fat Bollman That I think is very definite, there again is some indication of the value of carbohydrate Best I am not sure that it is I am not sure except that I believe you would have no difficulty in showing that with more or less specific hepatotoxins the insulin treated diabetic animal would survive more frequently than the untreated Bollman I thought everybody knew that Best Do they? I take it for granted that this would happen but I dont know where to find the data in the literature Bollman I tried it in dogs I never published it Best You should because it would be very valuable, perhaps more than anything we have unearthed today Bollman Thank you! Best I would not have undertaken this problem I am very grateful to Do Bollman for doing it Watson There are features from the clinical side which I think ought to be mentioned I would like to hear just a little bit of discussion about the effect of glucose in the patient with severe liver disease I had a most remarkable experience a year ago The patient was an alcoholic She had severe arthosis of the liver She came into the hos pital with ascites and edema We started her in on a high protein intake I think she was getting about 150 grams a day, part of which was casein and some meat We had some dificulty feeding her She was taking it reasonably well but she did not improve She persisted in having ascites and edeme and had quite a bit of jaundice and many 122 ·evidences of liver dysfunction We then gave her human albumin intra venously We gave about 900 grams in a period of about 12 days and go the serum albumin up to 52 grams percent but there was no diuresis In fact she had increasing oliguria of a mld degree and gained weight about ten pounds and her ascites increased We did not tap her dry at the outset Perhaps we could have gotten a better effect had we done so but in any event she got worse during this period We were getting in quite a lot of protein during this period as we had given 900 grams of alubumin intravenously in addition to the high protein diet Nevertheless she became semi comatose and everyone who was watching her thought she was going to die At that time we obtained a liver biopsy This was remarkable in that there was still a good deal of fat in her liver It was a severe cirrhosis and there were many spaces that were just empty which appeared to be the result of liver degenera tion There was no necrosis but considerable fibrosis and a great deal of cellular exudate We were at our wits ends and decided we would give her glucose She received 20 percent glucose intravenously in amounts of 300 grams a day She promptly improved and made a re markable recovery She had a good diurests and lost weight At present she is quite well and without jaundice ascites or edema I realize that she may have had a delayed benefit from the albumin If so it was delayed about 15 days During the period of albumin and high protein therapy not only did she not benefit but visibly she got worse Gyorgy Did you give injections of vitamins all the time? Watson No Gyorgy I am very much interested in that that is rather a common practice in many clinics Watson We don t often and I don t believe we did here Besi Did you stop the intravenous protein when you gave the sugar? Watson Yes the albumin was stopped for about 48 hours an inter val of just marking time trying to decide what to do Best Was the essential amino acid composition low in methionine? Watson Of the albumın? I would say it is good Gyorgy It is lower in isoleucine according to Fenster Watson The patient had a pretty good general protein intake by mouth in addition to the albumin intravenously I cite that case not in any attempt to say that glucose as the thing that saved her life but simply to show at least that glucose did her no harm She got 300 grams of glucose a day at a time when she was in coma I have seen that again and again I have seen a number of patients in hepatic coma at least come out of their coma when they got a lot of glucose intra enously so that I think we can say at the least that the glucose probably 123 ·isot doing them any harm Whether it is the thing that is bringing them out of coma I would hesitate to say because we do do other things We do give them vitamins in varying degree Gjorgj In coma especially, we are fond of niacin De Neefe did you not find in a few cases at least an indication for what we call the vitamin cocktail > Neefe I think that originated at the Mayo Clinic They gave large doses of nicotinic acid and thiamine and had dramatic responses I dont know whether they gave glucose or not I think that they did Parek Dr Chester Jones has maintained for years that infusions of glucose would produce remissions in subacute hepatitis and in the pre-chronic phases of cirrhosis Hanger We make no effort to keep up the protein intake during an episode of acute hepatic insufficiency We rely chiefly upon glucose by slow infusion Best It seems to me that there is very convincing clinical evidence that the glucose does a great deal of good but that the experimentalists have completely failed to find any reason for this Bollman That is correct Best Had Dr Stetten been here he would have given us a little talk on liver glycogen as a thing as obsolete as the vermiform appendix Gyorgy That is right a vestige Best Yet when you give the glucose you probably do get more liver glycogen A lot of other things also happeni so perhaps we have been wrong in just looking at liver glycogen to find the explanation of a beneficial effect of glucose Watson What observation would you suggest to the clinician who is going to be able to study just such a case as I mentioned? What other thing should one look at to try to find out hon glucose is beneficial if it is? Best Study it in the untreated diabetic Watson I was thinking of the patient with hepatic coma or pre coma What observations could one make to settle the question as to whether the glucose was having a rather precise effect? Best My thoughts are more in the experimental field where you have to eliminate certain organs to get a clear ansker Hanger Why do you think they come out of coma? Markou itz It might be circulatory effect Three hundred grams of glucose over the course of a day is bound to improve the blood flow, shall we say to the liver The blood 15 going around perhaps twice as fast while you are ging it 124 ·Best You spoke of the effect on the heart Markouitz That effect, sit, is only shown then no glucose whatever is in the infusion The addition of a trace of glucose is about as good as the addition of a gram Gyorgy It may be the nuacio and thiamine Watson The effect of the glucose on the brain might be very im portant because in hepatic coma there is brain injury You can see it with the microscope Knisely If you are going to have glucose bringing the patient out of coma, it has to go to the brain Markou itz For twenty five years I have been treating patients with liver disease with intravenous glucose I hate to think it accomplished nothing Best I dont think there is any intent that our lack of experimental information should discourage the clinicians in the use of glucose We should be stimulated to find some better reason, some better explana tion We really have never set up the proper experiments ·THE LOCUS OF THE BEGINNING OF DIETARY CIRRHOSIS SECTION VII Hastroft Evidence has been presented by L L Ashburn, R M Endicott, F S Daft and R O Lillie (Am J Path 23, 159 (1947)) to shon that the distribution of the fibrous trabeculae is centrolobular This has been supported by L E Glynn, H P Himsworth and o Ladan (Best J Exp Path 29, 2, (1948)) Recently this has also been accepted by P Gyorgy and H Goldblatt (Exp Med 89, 245 (1949)); who originally (Ibid 70, 185 (1939) and 75, 355 (1942)) considered the fibrosis to be periportal Yet many other investigators', who have described the microscopic appearance in this type of cirrhosis, indicated that the fibrous tissue surrounded the portal canals At one time, we also thought that possibly the distribution of fibrous tissue in the livers of our choline deficient rats might be postal, although it was stipulated that further evidence was needed The observations which follow were made for the purpose of obtaining additional information concernung this problem The results indicate that, rather than portal, the fibrosis appears to be definitely central in distribution The question is perhaps somewhat clarified, however, since some of the observations indicate a possible explanation for the divergent conclusions which have been reported in the interpretation of the microscopic features of this type of cirrhosis Some investigators (C L Connor and I L Chaikoff, Soc Exp Bio! and Med 39, 356 (1938)), L E Glynn, H P Himsworth and o Lindan (Brit J Exp Path 29, 1, (1948)) and others, support the view that the fatty infiltration of the liver which precedes dietary cirrhosis may be a primary etiological factor in stimulating fibrosis Others (P Handler and I N Dubin, Nutrition 31, 141 (1946)), (P Gyorgy and H Goldblatt, J Exp Med 89 245 (1949)) question this concept The observations to be described below, appear to give further support to the view of Connor and Chaikoff, Glynn et al Some of the findings also suggest a possible route by which the excess fat might escape from the liver during the later stages in the development of cirrhosis These 11 L Chaskoff CL Connor and S R Beskind (Am J Path 14 101 (1938) depan creatized dogs) CL Connor and I L Charkoff (Proc Soc Exp Biol & Med 39, 336 (1938) normal docs fed alcohol and high fat diets) IL Chutkoff and C L Cunnor (Proc Soc Exp Bol & Med 43 638 (1940) normal dogs fed high fat dicts alone) A R Rich and ID Hamilton (Bull Johas Hopkins Hosp 66 185 (1940) rabbits fed diets deficient ia veast) G Webster Clin lavestig 20 440 (1941) 21 385 (1942) rats fed low protein diets) H Blumberg and HS Grady (Arch Path 34 1035 (1942), rats fed low protein dicts) R I Holman ( Exp Med 81 399 (1945), dogs bled while fed on hgh fat drets) 126 ·aspects of the pathogenesis of dietary cirrhosis will now be presented in detail INVESTIGATIONS CONCERNING THE PATHOGENESIS OF DIETARY CIRRHOSIS IN CHOLINE DEFICIENT RATS Methods A diet deficient in choline and its precursorsa was fed ad lib for Serjods up to seven months to 120 young adult male rats (150 grams) of the Wistar strain This food mixture was relatively deficient in tocopherol At intervals separated by 24 hours during the first ten days, and twice weekly during the following three months of the experiment, two animals were selected at random and sacrificed During the remander of the experimental period, groups of five animals were killed at monthly intervals Under sodum amytal anesthesia', the hepatic vascular system in the heparinized animal after perfusion with normal saline, was injected under physiological pressure ita the beating left entricle with a filtered, diluted colloidal suspension of carbon Blocks of tissue selected from the right and the left lobes of the liver were 250 2 DIET Arachuza 12% Beef far 10 Gelatan 66 Corn oil 2 Casein 30% Cod liver Oil concentrate 0015% Fibrin Cystine 0 50% Salts so Cellufour Sucrose 57307 Vitamin powder 106 The composition of the salts used in the above per 100 gms was Calcium lactate 350 gms. Calcium carbonate 506 gms Ca(HPO.) H.O 34 60 gms КНРО NaH.PO. H.O Naci 9 34 gms MgSO, (anhydrous) 7 18 gms fe estrate 3 20 gms aad trace salts MnSO. 4H, 033 gms ZASO, THO 0035 gms CuSO, 3H,0 0 039 grs KI 000039 gms The Vitamin powder per 1000 gms of mixture contained Th 2 1 500 mgm Riboflavin Pyridoxine 200 mgms Calcium pantothenste 1 gm Nicolinc acid 2 gm and powdered sugar 997 05 gms The cod liver ol concentrate used (CLO conc) contained 50 000 U of Vat D and 200 000 IV of Vit A per gram (Supplied by Alessis Ayerst MicKenna and Harrison Montreal) 390 mg of solum amyral per kilogram of body weight administered intraperitoneally is a freshly prepared 100 sofution in dsledater "Hggins American India ink diluted with to parts of dsst lied walet A few drops of aromon a added to the mixture ensures suspension of the carbon particles 6 46 gms 18 76 gms 250 mgons 127

·No text on this page.

·1209
4
سر
T
مها
ده

·that 30 to 40 percent of the wet weight of the liver is lipid (L H Ridout, C C Lucas J M Patterson and C H Best, Science 103, 12 (1946)) Fatty livers produced in rats by one week of choline deficiency rapidly revert to normal when lipotropic factors are restored to the diet Within 24 to 48 hours, the cells at the periphery of the liver lobule are com pletely devoid of stainable fat (Figs 5 and 6) Centrolobulae lipid can still be seen, but this is mostly in the form of several small droplets instead of large spherules After three or four days on the normal diet, it is difficult to find any stainable fat throughout the sections (Fig 7) Liver cells which have been rapidly depleted of sudanophilic lipid in this manner, frequently give pale staining reactions and have a smooth, glassy type of cytoplasm These observations indicate that the cells in the central portion of the liver lobule are not only the first which become affected by fatty change in dietary cholme deficiency, but are also the last to retum to normal when lipotropic factors are restored to the food Circulatory Changes in the Fatty Liver In sections of livers which contain a high percentage of lipid, the degree of destension produced in parenchymal cells by large vacuoles of stanable fat is great (Fig 8) The cells have enlarged at the expense of the intervening sinusoids In routine sections the intralobular vascu lar channels are unconspicuous and difficult to demonstrate The lumina are often so small that their positions are only indicated by the flattened linings of endothelial cells Cleared liver slices (100 u), from rats injected intravenously with FIGURE 9 Normal lner from a control animal which has been injected with India Ink via the hepatıc arter, and portal vein The tissue was subsequently Fixed in formalin and a thick (100 w) sizce cut on the freezing microtome The slice was then dehydrated in alcohol and cleared in benzyl benzoate Note the appearance of the normal intralobular sinusoids x 50 FIGURE 10 The appearance of a very fatty irer from a rat fed a choline deficient diet for 7 days which recened the same treatment as the preparation shown in Figure 9 Note the compression of sinusords in the central and med portion of the lobules Ink in the central and med portion of the lobules Ink in the central veins demonstrates that the ischaemu was not complete x 50 FIGURE 11 The size and appearance of Iner cells in a parafin section of a normal control sat (150 200 grams) This is for comparison particulirly auth Figures 12 13 and 14 shich have recened the same photographic technic, and are at exktly the same magnification Paraffin section McGregor stain (azocarmiae anilin blue and orange G) x 860 Inele days FIGURE 12 Fatty vacuolation of individual loser cells in a mat less than of dietary choline debciencs The diameters of these celle take those of normal ones Paration section McGregor

·10
11
12

·ink, facilitate adequate visualization of the intralobular vessels (Fig. 9 & 10) An appreciable decrease in the diameters of centrolobular sinusoids is demonstrable in fatty livers Since the injection mass which came to the liver ila the portal vein and hepatic artery, reached the central veins, the centrolobular ischaemia can only have been partial Had it been complete necrosis of the surrounding parenchyma would have resulted The injected preparations indicate that there is a dimin ished blood flow in the central portions of the lobules This appears to be due to sinusoidal narrowing produced by pressure exerted on the channels from without by the swollen, fatty parenchyma It is not surprising that the effect of the extra sinusoidal pressure is most apparent toward the center of each lobule, for here the intra sinusoidal pressure is lower than at the periphery and the vessels are more readily com pressed A description of a similar type of circulatory impairment has been given by LE Glynn and HP Hemsworth (Clin Science, 6, 235 (1948)) following hydropic swelling of cells in the centrolobular regions of livers of rats treated with carbon tetrachloride Lipodrastaemata, Their Concept Structure and Formation Microscopic evidence of fibrous tissue proliferation in these livers was not infrequently observed in rats fed the basal diet for periods as short as 8 weeks In one case new fibroblasts were found after the animal had been fed the choline deficient diet only 6 weeks In the stage of early hepatic fibrosis, the spaces occupied by stainable fat achieve FIGURE 13 The large space occupying half of the field is a medium sized lipoduastaema (see text page 136) Note the very thin septum separating the two cells in the other portion of the picture If this septum should rupture, another lipodiastaema would be produced equal in size to that shown below The very blach, round or oval spaces represent injected in in sinusoids From an animal fed the deficient diet for 45 days Paraffin section of liver McGregor stain X 860 FIGURE 14 A giant lipodiastaema is shown Two nuclei of cells forming its wall are present at the right The diameter of this space is 7 8 times that of a normal liver cell (volume 350 400 times greater than normal liver cell) The Irser cell nucles have also increased in size but the small endothelial cells afford an index of comparatne size From a cat fed the deficient diet 51 days Paraffin section of lives McGregor stala x 860 FIGURE 15 The Lipodrastaema in the upper left portion possesses an outline resembling that of an hour glass This suggests that the lipoduastaema has just been formed from two smaller spaces by rupture of a septum between them which has incompletely retracted Paraffin section of liver of a rat fed a cholinedeficient diet 65 days McGregor stain FIGURE 16 A lipodrastaema with several compressed nuclei in the wall A fine fibril of connective tissue surrounds the structure Some of the flattened nuclei may belong to narrowed sinusoids around the lipodiastaemu Laser of a rat after 65 days of choline deficiency Paraffin section McGregor stain x 860 x 860

·13
红:14
15
6

·relatively enormous proportions Parenchymal cells in paraffia sections of livers of control rats averaged about 14 u in diameter (Fig 11) In livers of animals fed the choline deficient diet for twelve days, the diameters of the fatty cells seen in sections were increased, measuring 18 20 x (Fig 12) After longer periods (45 and 51 days in the cases of the animals from which the sections illustrated in Figures 13 and 14 respectively, were prepared) the spaces occupied by fat were found to have diameters measuring as much as 100 or more in the paraffin sections This is greater than 7 times the diameters of liver cells in Sections of normal rats, it is an astonishing fact that the volume of these large spaces must therefore be 350-400 times that of a normal hepatic cell The total liver weight, however, is increased to only two to three times the normal (J H Ridout, C C Lucas, J M Patterson and C H Best, Science, 103, 12 (1946)) This strongly suggests that these large spaces no longer represent the accumulation of stainable lipids in just one liver cell as in the earlier stages As vacuole is the term commonly employed when referring to the stainable fat of a single cell, it is not applicable to these giant spaces The name lipodiastaema (lipofat, diastaema--space between) is suggested for the latter Evidence will now be given which suggests that the fusion of several fat vacuoles, each orginally present in a single cell, produces a lipodiastaema soks, large affin sec is always a fraction of the total number of lens allameu me sice vi a npowmwema, the Laser would have a volume much greater than only double or triple the normal As has already been stated, this is not the case This evidence suggests that several cells probably take part in the formation of a lipodiastaema If this is so each lipodiastaema should contain several nuclei This appears to be true In thin (4 v) sections usually at least one nucleus (Fig 13) and frequently too (Fig 14) are seen This in 1 - 1.al nucleus one In paraffin sections the adjacent limiting membrances of two con tiguous Isser cells swollen by fat often form partitions bets een neigh bouring vacuoles These partitions are so stretched and thinned that even under the highest powers only a single, tenuous membrane can be resolved (Fig 13) Frequently these delicate partitions are broken in 136 ·the sections and their sesered ends separated by space These membranes are either ruptured during life or they are torn during the handling and prepatation of the tissue They are either fact or artefact Many such breaks doubtless are produced by the skeep of the microtome knife despite the support supplied these membranes by paraffin Horever, lipèdiastaemata in sections may sometimes have outlines which re semble those of hour glasses (Fig 15) This suggests that rupture of 1 septun originally separating fat vacuoles of to liver cells had occurred sufficiently long before the animals death to permit incomplete tetraction of the free ends of the severed membrane When the membranes are so stretched and thinned that they are easily torn by the microtome knife, this post mortem occurrence may be viewed in one sense as merely accomplishing a result which might have developed in the animals had death not supervened Having, introduced the concept of ispodsastaemata and presented evidence for their existence in fatty livers, these structure and formation will be considered in more detail as indicated, a lipodiastaema results from the rupture of a thin membrane which becomes stretched and compressed by expanding fat vacuoles which distend adjacent liver cells Dissolution of the membrane nhich had previously kept the fat of these cells in separate globules, allows them to coalesce into a single mass In this form, the fat is now contained in the lumen of an unlocu lar sac formed by the conjoined parent cells These are the cells, each of which previously contained a portion of the fat now in the form of a single unit Thus the lipodrastaema possesses a wall consisting of a continuous, single layer of thin, epithelial cells which surrounds a lumen filled usth lipid The size of cach lipodiastacma and the number of nude present in its wall will depend on the number of iner cells which at any time have contributed to its formation At the time of its formation, the smallest lipodiastaems would have a volume at least kice that of single liver cells distended with fat, the nuclei must number at least two In theory, the maximum size these structures could attain would be great it is apparent that tho selatively small Ispod: astaemata might form a single large one by a process similar to that by which liver cells combine to form new lipodiastaemata By extension of this principle, successive generations of Ispodiastaemata could be created of ever increasing magnitude Vith each nes generation, the total number of fipodsastaemata in any lobule would be reduced Diameters of lipodiastaemnata greater than 100 x are rare (Fig 16) Nucler, counted in serial sections have totalled as many as 80, in a few lipodrastaemata so studied There is the danger hos erer that large nucles present in successive sections may have been counted twice Plastic reconstructions in three-dimensions should be prepared for greater accuracy ·The structure of a lipodiastaemu might be compared to that of an acinus in the thyroid gland Both possess a central lumen filled with fat in the one case and colloid in the other The walls of both are formed by a continuous single layer of epithelial cells Unlike the thyroid acinus however the lipodiastaema may have an outlet draining its lumen Bile canalicul lie between liver cells not separated by sinusoids Dissolution of the septum between two such cells would imply that, in theory at least the lumen of the resultant lipodiastaema could be in continuity with the billary tree Passages which might appear to drain lumina of lipodiastaemata have not been observed in sections This does not prove their absence however for the walls of lipo diastaemata are so frequently thinned and stretched that small outlets passing through them might not be visualized even if present The concept of a lipodiastaema as presented here is that of a struc ture If this concept is correct what then is the function of this structure? It is a mathematical law that, of the total space occupied by a hollow sphere, the percentage of that space taken up by its walls must rapidly diminish as the size of the sphere increases Thus a single large sphere whose volume is equal to the sum of many separate small spheres occupres less space than does the group of small spheres even if the latter are crowded together The fusion of many small unicellular fat vacuoles into multicellular lipodrastaemata each of which contains a single large pool of fat faalitates maximum fat storage in a minimum of space according to the mathematical law just stated The high ratios FIGURE 17 A lipodiastaera with many nucle present in a single section of its wall This increase in nuclet has probably been produced by shrinkage of the structure rather than proliferation of the cells Lever of a rat choline deficient for 65 days Paraffin section McGregor stain x 860 FIGURE 18 Another atrophic lipodiastaema showing fibrous tissue spreading out to encircle the structure The central vein of the lobule lies just to the left of the field shown The fibrous tissue could be traced to the adventitial sheath of the vein Liver of a rat choline deficient for 65 days Paraffin section McGregor stain X 860 FIGURE 19 Two large lipodiastaemata cut 10 such a plane that they appear to have a longitudinal axis They are surrounded by a fine skein of connective tissue fibrils Their appearance is not unlike that of abnormal distended bile ducts They are not vessels for the animal was injected with india ink which titled all branches of rerns and arteries From & liver of a choline deficient sat (75 days) Paraffin section McGregor stain, x 420 process of shrinkage The nuclei of some 15 surrounded by a fine band of connective tissue which has an appealance not unlike that of a basement membrane From the liver of a rat choline deficient for 75 days Paraffin section McGregor stain x 860 138 ·318 2.20 ·1 5 - of fat to parenchyma that can be produced in the liver would probably not be possible without this device by which lipodrastaemata are formed If the fat had always to be stored in relatively small vacuoles, each in individual liver cells, the potential efficiency of the liver as a lipid storehouse might well be much less than it actually is Thus the function of lipodiastaemata is probably to enable the liver to store fat in large amounts The Fate of Lipodrastaemata in Cirrhosis In cases of advanced cirrhosis in both man and animals it has been well established that the fat content of the liver falls to almost normal values In microsections of these livers, few, if any, lipodiastaemata can be found Apparently th I the initial phases of cirrhosis only to by which they disappear will now be such Soon after the first appearance of large lipodiastaemata, the walls of some appear to contain increased numbers of nucler lying closely together (Fig 17) This is most frequently noted in the regions around central veins Mitotic figures in the cells of lipodiastaemata have not been observed at any stage The probable explanation of this apparent increase in the number of cells is that having attained their maximum size, the lipodiastaemata begin to shrink As their diameters decrease, the nude in their walls would approach one another so that more would be included in random sections of standard thickness (4 v) than would have been the case before the advent of shrinkage Further evi dence of atrophy of lipodiastaemata is afforded by extensions of connec tive tissue fibres from the adventitial sheaths of the nearby central veins FIGURE 21 Two atrophic lipodiastaemata surrounded by a fine netwock of connective tissue fibrils Note that the cells forming their walls are no longer thinned and stretched These might be compared to abnormal forms of bile ducts From the liver of a rat choline deficient for 75 days Paraffin section McGregor stain * 860 FIGURE 22 A further stage of atrophy is illustrated by the lipodiastaems in the center of the field The cells in the wall are finely vacuolated and appear to be degenerating Again this might be compared to an abnormal bile duct From the liver of a rat choline deficient for 75 days Paraffin section McGregor stain X 860 FIGURE 23 The collection of cells in the center of the field arranged in a somewhat acinar tanner represents the final result of atrophic changes in a lipodiastaema The lumen here is completely absent From the liver of a rat choline deficient for 90 days Paraffin section, McGregor stain x 860 FICURE 24 With complete atrophy of the lipodiastaemata the cells disappear leaving only hyalıne fragments (grey in photograph) between the connective tissue (black) Paraffia section of liver of a rat, 90 days choline deficient McGregor stain, x 860 140

·621
22
24
141

·(Fig 18) This suggests that the space occupied by the lipodiastaema before shrinkage began is being filled by encircling fibrous tissue When cut in the plane of their greatest diameters, these partially fibrosed and atrophic lipodiastaemata not infrequently resemble destended bile ducts (Fig 19) If the plane of section is tangential to the structure, the lumen of an atrophying lipodiastaema appears to be lined by several layers of cells (Fig 20) As the lumen becomes still smaller the walls lose their thanned and stretched appearance (Fig 21) This type of lipo diastaema may, in some instances, very closely resemble atypical bile ducts (Fig 22) In the final stages the lumen disappears altogether (Fig 23) with the result that only a collection of dedifferentiated liver cells is left These cells undergo further degeneration, eventually becom ing only hyalıne masses intermingled with fibrous tissue which has increased at the same time (Fig 24) It appears that in the manner the fibrous tissue spreads outward from the central veins preceded by a zone of atrophic lipodiastaemata Eventually all the lipodrastaemata disappear and are replaced by fibrous tissue which then surrounds only nodules of regenerating parenchyma The details of these last stages will be considered in a subsequent section (Figs 25 and 26) The factors which limit the growth of lipodrastaemata and are re sponsible for their atrophy and disappearance have not been demon strated It is likely, however, that the partial ischaemsa present in the centrolobular regions where this type of atrophy occurs may play an important role The loss of fat from those lipodiastaemata which atrophy could be either a cause or a result of their decrease in size The route by which the fat leaves the structures is unknown Absorption into the blood stream or lymph might be possible for this to occur, the fat would have to pass either through or between the epi Figure 25 This illustrates the growth of connective tissue around atrophuc lipodiastaemata The connective tissue fibres are young and have not stained intensely From the liver of a fat, 90 days choline deficient Parifin section, McGregor stain x 410 FIGURE 26 Degenerating liver cells (derived mostly from lipodiastaemata which have atrophed probably) are intermingled with connective tissue From the Inser of a rat 96 days choline deficient Parafin section, McGregor stain x 410 Figure 27 Lipodrastaemata fortration around a central vein, low power Those in the centre of the photograph are atrophying and becoming separated from the surrounding liver tissue From the liver of a rat choline deficient for 41 days Frozen section Sudan IV and haematoxylın x 80 FIGURE 28 High power view of a section from the inter shown in Figure 27 above A lipodsastaema lying adjacent to the central vein in the photograph is filled with fat (grey) A lipodiastaema undergoing atrophy is in the upper right portion From the liver of a rat choline defcient for 41 days Frozen section Sudan IV and hiematoxylın x 240 142 ·27 143 ·thelial cells forming the walls of lipodiastaemata In such an event it might be expected that a microscopic picture would be produced similar to that observed when fat traverses the epithelium of intestinal ville during fat absorption No suggestion of this has ever been found in frozen sections of livers containing atrophıc diastaemata It has been stated that theoretically the lumena of the lipodiastaemata could be in communication with the billary tree This raises the further possibility that the fat might leak out in the bile E R LeCount and E A Long (J Exp Med 19 233 (1914)) found the percentage of fat in the bile was increased in fatty degeneration of the liver in man It is planned to investigate this possibility further in choline deficient rats The Site of Earliest Proliferation of Fibrous Tissue It has been stated in the foregoing several times that the site of initial proliferation of fibrous tissue in these livers is around the central veins The evidence supporting this will now be presented Although liver lobules are most frequently cut in a tangential plane which does not include both portal and central areas recognizable in the same section favourable cuts may be found which pass through a section There is unds the veins shown at the edges of the field included in the photograph The portal region is quite free of fibrosis as shown in the enlargement (Fig 30) This indicates that the areas of early fibrosis are actually centrolobular and surround branches of the hepatic vein (Fig 31) The atrophic lipodiastaemata may bear a superficial resemblance to atypical bile ducts but in frozen sections such structures are filled with stainable fat (Figs 27 and 28) FIGURE 29 Small crescent shaped areas of fibrous tissue incompletely circle the portal vein in the center of the field The fibrous tissue appears as grey aceas in the upper right and lower left ortions of the field They surround ink filled veins (black) Liver of a rat choline deficient 56 days Paraffin section McGregor stain x 120 FIGURE 30 This is a high power of the central portion of Figure 29 above It shows that this area of the field (in which no fibrous tissue is present) is the portal region for the large portal veia small hepatic artery and bile duct are clearly shown x 860 FIGURE 31 This is a high power view of the portion of the section shown in the upper right area of Figure 29 It demonstrated early fibrous tissue prolifera tion (slight) around a central vein (black) in the midst of atrophic Lipodrastae FIGURE 32 The portal area in the center of the field is completely encircled by connective tissue extending around the central veins From the liver of a rat choline deficient for 79 days Paraffin section McGregor stain x 120 mata x 860 144

·29
.30
iܐ܀
31
32

·1 1 The fibrous tissue spreads out in the form of trabeculae which con nect nearby branches of the hepatıc vein one with another This creates an apparent reversal of the usually conceived lobular pattern for the portal canals now lie in the centers of islands of liver tissue which are encircled by the fibrous tissue extending from one central vein to another (Figs 32 and 33) If the plane of section passes through only the portal vein but m sses the smaller bile duct and hepatic artery and if atrophic lipodiastaemata in the centrolobular fibrotic regions are mistaken for bile ducts the encircling pattern of the fibrosis may erroneously be interpreted as portal We suggest that this is the explana tion for the contradictory statements in the literature concerning the distribution of these fibrous trabeculae The bile ducts in the fibrotic ares have been stated by some authors to be atypical or poorly formed (CL Connor and I L Chaikoff Proc Soc Brol & Med 39 356 34 1035 (1942)) s of fibrosis show e have illustrated (Fig 31 and others) Additional evidence of another type which supports the view that the fibrous tissue is centrolobular was obtained in the following man ner It has already been shown that when choline is restored to the det of choline deficient animals with fatty livers the only cells depleted of stainable lip d within the first 24 to 48 hours are those which sur round the portal veins The cells in the centrolobular regions remain fatty for a longer period A special group of animals rats in which hepatic fibrosis had previously been produced by choline deficiency were killed 24 hours and 48 hours after restoring lipotropic factors to their food Frozen sections of these livers showed that the fat had disappeared only from those cells farthest from the fibrous trabeculae T 2 f FIGURE 34 L ver of a rat fed the choline defic ent diet for five months The vessels (black) are nlled w some of the largest I podcastaemata $ pcture is for compar son wth FE IV and haematoxplio x 50 L FIGURE 36 Liver of a chol ne deficient rat a wh ch the sheath of ue nepais ve n was injected wth Ind 2 Ink The black ink is present throughout the strands of grey fibrous tissue McGregor stain Paraffin sect on x 100 146 ·'31 3.

·(Figs 34 and 35) The strands of fibrous tissue are surrounded by haloes of parenchymal cells which still contain stainable fat This strongly suggests that this fibrous tissue is en centrolobular, not peri portal, regions Tangential cuts of liver lobules frequently suggest that the fibrosis is 373dzonal for the resulting sections may sclude portions of the fibrous trabeculae but not the veins which they surround It can be demonstrated however, that the fibrous bands are always in continuity with the adventitial sheaths of the veins This purpose was served by the preparations made by injections of sok into the sheath of the hepatic vein in the sections the injection mass can be followed from the adventitial (Fig 36) that advan Lillte (Am ) Path 23, 159 (19:47)), leaves little doubt that the dis tribution of fibrous tissue is noaportal throughout the livers of rats in which dictacy acrhosis has been produced by choline deficiency ch CL EL 11 Parenchymal Proliferation in Dietary Circhosts Few of the liver cells adjacent to portal canals take part in the forma tion of lipodiastaemata for here these are rarely found (Figs 29 and 37) However, it is on these per portal portions of the lobules that mitotic figures are frequently observed They are not seen in appreciable numbers until the animals have been fed 2 choline deficient diet for approxsmnately three months or more As thus corresponds to the time when fibrosis of atrophic lipodiustaemata is actively progressing there is coexistent centrolobular degeneration and portal regeneration in these livers (Figs 37 38 and 39) The young lover cells rarely undergo more FIGURE 37 In the center of the field regenerating Isser tissue sear a portat canal is relatively devoid of fatty change Giant lipodiastaemata surrous the fibrous tissue around the central veins which are at the edges of the field Paraffin section McGregor stun x 50 From the liver of a rat choline-deficient for 5 months FIGURE 38 A mitotic figure in a portal area of a liver in the stage of early cirrhosis Paraffin section McGregor stain x 950 (This is from an animal of one of Professor E A Sellers experiments of 1947) FICURE 39 A mitone figure za a lsver celi fear a portal vein (not shoan) I This field 25 Paraffin section FIGURE 40 A cleared slice (100 u) of liver tissue melleu avascularly with India Ink Note the small islands of sinusoids sidely separated by relatively avascular bands of fibrous tissue The liver mas from a rat fed the choline deficient diet for six months x 50

·37
38
((
40

·than slight fatty change even though the animal is still consuming a choline deficient diet We cannot offer any explanation for this The absence of any appreciable degree of fatty vacuolation in the new tissue serves to differentiate it, in sections from the older parenchym. With completion of atrophy of the lipodiastaemata in the centrolobular ke gions the fibrous trabeculae surround irregulac islands formed almost entirely by regenerating livec tissue which contains little stainable fat On the surface of the organ the elevated nodules of parenchyma are separated by depressions in which run the fibrous bands This produces the appearance which is so characteristic of these lovers on gross examination Cleared slices (100 w) of these organs injected intra vascularly with ink reveal small groups of sinusoids which are widely separated by relatively avascular bands of cleared fibrous tissue (Tig 40) Cerord In 1942 R D Lille L L Ashburn WH Sebrell F S Daft and J V Lowry (U S Pub Health Rep 57, 502) gave the term cerord to the yellow pigment which is present in the form of large granules and globules in the fibrous trabeculae of arrhotic livers of choline deficient rats This pigment is insoluble in the usual fat solvents and is accord ingly retained in sections prepared with paraffin infiltration This prop erty plus its sudanophilic character renders it unique It also stains with most of the aqueous basic dyes and is acid fast The substance was reported almost simultaneously by R D Lillie F S Daft WH Sebrell (U S Pub Health Rep 57 562 (1941)), H Blumberg and E V AlcCollum (Science 93 598 (1941), J E Edn ards and y White (Nat Cancer Inst 2 157 (1941)) and P Gyorgy and H Goldblatt (J Exp Med 75 355 (1942)) Its fluorescent character was described by H Popper P Gyorgy and H Goldblatt (Arch Path 37 161 (1944)) K M Endocott and R D Lillie (Am Path 20 149 (1944)) reported its staining properties after investigation by a large number of methods and in addition described few of its cherrucat characteristics as exhibited in a crude extract K M Endscott (Arch Path 37 496 (1944)) later noted the similarity of ceroid to oxidized unsaturated fat He found that a substance similar to ceroid could be produced in cats by subcutaneous injection of cod Irver oil or linseed Oil Endscott also found ceroid lie pigment in a case of pneumonia in man following aspiration of cod liver oil Ssmslar cases in which acid fast material was present in the lung had been reported earlier by H Pinkerton (Acch Path 5 380 (1928)) and H Graef (Arch Path 28 613 (1939)) K M Endicott FS Daft and WH Sebrell (Proc Soc Exp Biol & Mied 57 330 (1944)) succeeded in producing dietary cirrhosis 150 ·Bithout cerard in rats fed diets low a cholme a - of both fat and cod liver oil They felt the latter substance was an important factor diets used in their previous experiments in which ceroid was found They noted that the ceroid producing diet of P. Gyorgy (Am J Clin Path 14, 67 (1944)) also contained cod liver out M Wachstein (Proc Soc Exp Biol & Med 59, 73 (1915)) found that the amount of cerord in airhosis produced in male weapling rats fed on a choline free, protein deficient diet for 90 days could be marked ly reduced if the intake of cod liver oil was restricted to one drop weekly In livers of other sats fed the same diet containing 2 percent cod liver oul, large amounts of ceroid were found However, in Iners of a few rats fed the same basal diet without any cod Liver oil, he found small amounts of ceroid Wachstein concluded, therefore, that although ceroid deposition was not caused solely by ingested cod liver oil it could markedly enhance formation of the pigment Ceroid was not found in cases of experimental selenium cirrhosis or of human carthosis examined by R D Lille, L L Ashburn, W H Sebrell, F S Daft and I y Lowry (U S Pub Health Rep 57, 502 (1942)) A similar examination of more than 200 normal and patho- Sogical human livers gave negative results (H Popper, P Gyorgy and H Goldblatt, Arch Path 37, 161 (1944)) Victor and A M Pappenheimer (J Exp Med 82 375 (1915)) called attention to the resemblance of ceroid to the acid fast pigment found in the uterine muscle and adipose tissue of rats fed diets deficient in vitamin E They found that the food mixtures reported in the litera ture to have produced large amounts of ceroid in dietary arrbosis were deficient in this vitamja They performed experiments which showed that the production of cerold in the livers of rats fed on a low protein diet, with or without the addition of excess l-cystine (5 percent), was transiently inhibited by the administration of alpha tocopherol Cod liver oil (5 percent) in the diet could not present this effect of toco pherol They also found that 1 percent choline chloride had an inhibit ing effect on the deposition of liver-ceroid resulting from a low protein containing excess cystine The authors do not discuss this finding in our opinion, these experiments are difficult to interpret as the lesels of the supplements (cystine choline) are rather high A M Pappenhesmer and ) Victor (Am J Path 22 395 (1916)). examined tissues from human autopsies and reported the presence of acid fast pigenent not only in lovers, but also in many other organs These findings which do not agree with earlier examinations of human tissuc (Lillie Ashburn Sebrell Daft and long, loc cit Popper, Gyorgy and Goldblatt loc cit may be open to question as the authors do not state of the pigment they found as sudanophilic and fluorescent 131 ·And on the front of numa nenn lae and than slight fatty change even though the animal is still consuming a choline deficient diet We cannot offer any explanation for this The absence of any appreciable degree of fatty vacuolation in the nex tissue serves to differentiate it, in sections, from the older parenchyma With completion of atrophy of the lipodustaemata in the centrolobular se gions, the fibrous trabeculae surround irregular islands formed almost entirely by regenerating Inter tissue which contains little stainable fat On the surface of the organ, the elevated nodules of parenchyma are separated by depressions in which run the fibrous bands This produces the appearance which is so characteristic of these lovers on gross examination Cleared slices (100 u) of these organis injected intra vascularly with ink reveal small groups of sinusoids which are sdely separated by relatively avascular bands of cleared fibrous tissue (Fig 40) Ceroid In 1942 R D Lillie, L L Ashburn V H Sebrell, F S Daft and J V Lowry (U S Pub Health Rep 57, 502) gave the term cesord to t € r ingly returned to sections prepared with pacathin ultration les property plus its sudanophilic character renders it unique It also stars with most of the aqueous basit dges and is acid fast The substance was reported almost sumultaneously by R D Lille, F S. Dift, W H Sebrell (U S Pub Health Rep 57, 562 (1941)), H Blumberg and E V McCollum (Science 93, 598 (1941)), ) E Edwards and ) White (] Nat Cancer Inst 2 157 (1941)), and P Gyorgy and H Goldblatt Exp Med 75 355 (19:12)). Its fluorescent character was described by H Popper P Gyorgy and H Goldblatt (Arch Path 37 161 (1944)) KM Endicott and R D Lille (Am J Path 20, 149 (1944)) reported its staining properties after latestigation by a large number of methods and in addition described a few of its chemical characteristics as exhibited in a crude extract K M Endicott (Arch Path 97 496 (1944)) later noted the similarity of ceroid to oxidized unsaturated fat He found that a substance similar to ceroid could be produced in rats by subcutaneous injection of cod Iner oil or linseed Ost Endicott also found ceroid he pigment in a case of pneumonia in idan following aspiration of cod liver oil Similar cases, a which acid fast material was present in the lung had been reported earlier by H Pinkerton (Arch Path $ 380 (1928)) and H Graef (Arch Path 28, 613 (1939)) K M Endicott F S Daft and W H Sebrell (Proc Soc Exp Biol & Med 57 330 (1944)) succeeded in producing dietary asshosis 150 ·without ceroid in rats fed diets low in choline and free of both fat and cod liver osl They felt the latter substance was an important factor in drets used in their previous experiments in which ceroid was found They noted that the cerosd producing diet of P Gyorgy (Am J Clin Path 14,67 (1944)) also contained cod liver oil M Wachstein (Proc Soc Exp Biol & Med 59, 73 (1945)) found that the amount of cerold in arrhosis produced in male weanling tats fed on a choline free, protein deficient diet for 90 days could be marked ly reduced if the intake of cod Iner oil as restricted to one drop weekly In livers of other rats fed the same diet containing 2 percent cod liver oil, large amounts of ceroid were found Honever, in livers of a few rats fed the same basal diet without any cod laver oil, he found small amounts of ceroid Wachstein concluded, therefore, that although ceroid deposition was not caused solely by ingested cod liver oil it could markedly enhance formation of the pigment Ceroid was not found in cases of experimental selenium archoses or of human cirrhosis examined by R D Lille, L. L Ashburn, W H Sebrell, F S Daft and ) v Lowry (U S Pub Health Rep 57, 502 (1942)) A similar examination of more than 200 normal and patho Jogical human livers gave negative results (H Popper P Gyorgy and H Goldblatt, Arch Path 37, 161 (1944)) I Victor and A M Pappenheimer () Exp Med 82, 375 (1915)) called attention to the resemblance of ceroid to the acid fast pigenent found in the uterine muscle and adipose tissue of rats fed diets deficient in vitamin C They found that the food mixtures reported in the litera ture to have produced large amounts of ceroid in dietary airthosis were debcient in this vitamin They performed experiments which showed that the production of ceroid in the livers of rats fed on a low protein dict with or without the addition of excess 1-cystine (5 percent), was transiently inhibited by the administration of alpha tocopherol Cod liver os (5 percent) in the diet could not prevent this effect of toco pherol They also found that 1 percent choline chloride had an inhibit ing effect on the deposition of liver-ceroid resulting from a low pratesa containing excess cystine The authors do not discuss this finding in our opinion these experiments are difficult to interpret as the levels of the supplements (cystine choline) are rather luch A M Pappenheimer and J Victor (Am J Path 22 395 (1916)). examined tissues from human autopsies and reported the presence of acid fast pisinent not only in liters but also in many other organs These findings which do not agree with earlier examinations of human tissue (Lille Ashburn Sebrelt Daft and Lorry, loc 6:4 Popper, Gyorgy and Goldblatt loc tu ) may be open to question as the authors do not state of the pigtnent they found stas sudanophile and fluorescent 231 ·as well as acid fast They pointed out that the administration of Jarge amounts of cod liver oil is known to favor destruction of vitamin E They concluded that the accumulating evidence suggesting a relation between vitamin E deficiency and cero.d, although strong, needed fur ther investigation P Gyorgy and H Goldblatt recently (J Exp Med 89, 245 (1949)) reported that tocopherol, even when given in very large doses (30 mgm daily), will not prevent the formation of ceroid when cod liver oil is ingested and will reduce only slightly the total quantity found in the cirrhotic livers of choline deficient rats They were able, however, to produce dietary cirrhosis without ceroid provided that cod liver oil, and to a lesser degree other sources of unsaturated fatty acids, were elimi nated from the experimental diet Masses of cerord were present in the fibrous trabeculae in the livers of those anunals in our experiments in which the basal diet alone was fed (Fig. 41) This diet is deficient in tocopherol, and although the mixture does not contain the crude form of cod lever oil, it is present in the form of a concentrate In a group of ten animals fed the same basal diet with a supplement of tocopherol little or no ceroid could be demonstrated when they were sacrificed at the end of five months (Fig 42) These results appear to support the conclusions of A M Pappenheimer and J. Victor (Am J Path 22, 395 (1946)) The apparent discrepancy between our observations and those of p Gyorgy and H Goldblatt (ſ Exp Med 89, 245 (1949)), concerning animals given supplements of tocopherol and cod liver oil may possibly be due FIGURE 41 Sudanophilic masses of ceroid appear as dark grey_masses in the fibrous tissue in which a few atrophic lipodiastaemata appear From the liver of a rat fed the basal, choline deficient diet for five months Paraffin section, Sudan IV and haematoxylın stain, x 500 FIGE rt + aholar of the to a Suulli to the cat whose liver is represented in this figure Paramn section and haematoxylin x 500 FIGURE 43 Fibrils of connective tissue are shown extending around and almost in foreneriting liser cell containing small amounts of cerold in the form with red blood cells i dla Le. re 42 hoa SATA X 860 152 ·41 3 ·to the difference in the form in which the latter substance was given, they used the crude product, whereas we employed the concentrate We have examined the livers of 40 cases of human cirrhosis for the presence of ceroid Paraffia sections were stained with Sudan IV and by the Ziehi Neelson method for demonstrating and fast material Little ceroid could be found in any of these laers This finding is in agreement with the results of similar surveys by Lillie, Ashburn et al (loc czt) and Popper et al (loc cit) but at variance with the report of Pappenheimer and Victor (loc cit) Finally, one or two observations were made concerning the formation of ceroid in the experimental animals The pigment first appeared in the form of small globules in the cytoplasm of regressing cells in walls of atrophic lipodiastaemata (Figs 43 and 44) When these cells have degenerated and been replaced by fibrous tissue, the pigment remains as clumped aggregates scattered through the fibrotic areas (Fig 41) Some of the pigment granules are taken up by macrophages Summary 1 A progressive series of changes, as followed by microsections of the fatty livers of choline deficient rats which were sacrificed at fre quent intervals during the development of typical cirrhotic lesions, has been described An attempt has been made to interpret these changes in terms of the pathogenesis of this type of cirrhosis 2 Small droplets of stainable fat appeared in the centrolobular cells of livers of rats killed after they had access to the basal choline deficient diet for only one day After seven to ten days, fat was demonstrable in every parenchymal cell During this period the small droplets initially present fused to form single large vacuoles that distended the cells in which they were contained 3 Adjacent portions of limiting membranes of neighboring cells became transformed into tenuous septa by a process of pression and stretching which accompanied the progressive enlargement of the intracellular fat vacuoles separated by the septa In the livers of anı mals sacrificed after several weeks of dietary choline deficiency, micro scopic evidence indicated that some of these thin septa had ruptured A number of septal breaks had the result of combining several liver cells into relatively enormous structures of a multicellular nature, these have been named lipodiastaemata These are sacs with unlocular lumina distended with fat The walls of the sacs consist of a continuous single layer of dedifferentiated liver cells The concept, formation, 5 This material was made available to the writer by kind permission of Professor William Boyd Department of Pathology and Bacteriology University of Toronto 154

·No text on this page.

·Shorr I wonder, D. Hartroft, whether your problem might not be helped by working with living rather than dead liver cells Some time ago, when I was interested in the differences in the behavior of liver slices and homogenates, I found that when liver was put through the Latapie mincer, a very high percentage of individual viable liver cells were obtained 'That they are little injured in the process was evident from their very steady oxygen consumption for several hours (Cold Spring Harbor Symposia on Quantitative Biology, 7, 323, (1939)) With this technic you should be able to study the individual living liver cell, see the distribution of fat droplets and determine how large a lipodiastaema can be and still be contained within a single cell, nr whether they are multinucleated giant cells Hartroft There is no doubt there are many nuclei in the walls of the lipodiastaemata because serial sections very definitely demonstrate that I have counted them The counts have gone up as high as 80 There is danger that I have counted some of them twice where portions appeared in neighboring sections The fact that lipodiastaemata measure very frequently 100 micra in stained sections and infrequently even higher, suggests I think, that in the fresh state they are larger because of shrinkage in the stained sections Watson Do you think it is the pressure of the lipodiastaemata that induces the beginning fibrosis or do you have any concept of the exact mechanism of origin of the fibrosis? Hartroft As to the origin of the fibrosis that may be demonstrated not only by using the injection of ink into the sheath of the veins but also by following the fibrous tissue through the sections In this manner we can demonstrate continuity between fibrous trabeculae, even in the earliest stages with the adventitial sheaths of the central veins In the oil immersion photomicrograph which I showed (Fig 18), the fibrous strands extend around the cell Why lipodsastaemata atrophy I do not know I might suggest that circulation possibly plays a part Watson Have you stained any sections with Weigert s stain? Hartroft Yes Watson Does it show anything unusual? Hartroft No Watson Perhaps you are acquainted with that paper of Urteaga Ballon (Arch Rev de Patologia y Clin 2, 171 (1948)) ? Have you seen his study on the use of Weigert s stain in human livers? He makes the claim that the so called fibrosis of the alcoholic fatty liver is not connective tissue at all but largely elastic tissue emanating from the walls of veins and that one can differentiate these cirrhoses strictly on the basis of Weigert s stain from those in which there is true fibrosis 156 ·1 That is not confirmed in the rat in my experience Watson I have not found anyone who has repeated Urteaga s study He believes it necessary to use Weigert s stain on fresh material and that old fixed tissue is of little or no value Goldblatt I have not done a Weigert stain on fresh liver tissue in fatty arrhosis so cannot comment on the matter of the site of origin of the connective tissue I must admit that I along with all others who first succeeded in producing experimental arthosis was gulty of an Tror nh ch we have all retracted viva voce et verbo scripto We all onsidered the arrhosis as primarily portal but we all agree now that originates around the central veins Since the work of Ashburn there has been no doubt that it is pri marily a central type of cirrhosis So as far as I am concerned I am completely in agreement with everything that Dr Hartroft has shown and everything that he has said But there are some things that I would like to ask him and then perhaps we can get things clarified One is from what he has said it is not quite clear to me whether he thinks there is a definite proliferation of connective tissue an absolute increase of collagenous material and where it comes from and even perhaps if he has some ideas shy it originates? What is the stimulus? At no time in his report has he mentioned the sord necrosis and that makes me wonder whether I am guilty of another error. There is very little doubt that as a result of a diet which as far as I know is xactly the same as that which induces the changes which Dr Hartroft tentioned we have been able to see necrotizing changes in the liver I see in practically all the livers which shon fibrosis remnants of hat I interpret as necrotic or extremely degenerated liver cells paren ...mal cells Inasmuch as Dr Hartroft did not even mention the word necrotic as applied to liver cells I wonder if he does not subscribe to the idea that there are necrotic liver cells no matter what their cause Whether they are the result of the fibrosis or help to induce the fibrosis that I realize is difficult to settle but I would like to knon from him whether he has observed such necrotic cells I have been referring to those as residual necrotic liver cells in my reports to Do Gyorgy without implying that they are remnants of originally necrotic cells or just ensnared liser cells that have become necrotic I am inclined to the former vies for the simple reason that I often see single ensnared normal liver cells surrounded by a mass of thick fibrous connective tissue Has Dr Hartroft observed such a thing as necrotic In er cells in these Incrs which he has examined? One other thing and that is what do you think is the site of forma tion and deposit of ceroid? Do you find it in Kupffer cells? Do you 157 ·think a lot of it that appears to be within cells is actually in Kup cells, or do you think that those cells are altered liver cells? Gjorgy Before Dr Hartroft answers, may I just add a few rema because they are pertinent It is true that at the meeting of the Endoc Study Section of the USPHS in Cleveland two years ago I mente that the cirrhotic changes in experimental dietary cirrhosis are not not portal but also not strictly central What I had in mind wa! differentiate between the purely central cardiac cirrhosis and the s stellate dietary cirrhosis originating around the central vein If die cirrhosis would be caused by anoxia as cardiac Circhosis it should strictly central Dr Hartroft stated en passant that in the presence of tocoph in the experimental diet one cannot expect production of cerord: conclusion is not in accord with our own observations Tocopherol not prevent the development of ceroid provided the experime cirrhosis producing diet contains cod liver oil or any unsaturated f acid as fat supplement Even in daily doses of 30 mg, tocopherol only slightly delay but not suppress the production of ceroid Cellular necrotic foci scattered over the lobules and enmeshed in bands of connective tissue have been called by us 'residual , recently cellular focal necrosis Hartroft First, Professor Goldblatt, you asked is there a defi: proliferation, an increase in collagen and what is the probable stimul The nicest evidence I think that one could get if there were a defii increase in fibroblasts and collagen, would be to see mitotic figure: the fibrous tissue cells I have never seen those On the other hand feel that there is a definite increase in the amount of fibrous tis because the appearance of the section gives everything that sugg that However, Professor Boyd s chief assistant, Dt Barry, from C land, who has been with us one or two years, made the rather ingerie suggestion that he thought it was the reticulum and supporting tis: around these cells shoved to the outside of the lipodiastaemata, t] gave rise to the deposition of fibrous trabeculae Nevertheless, if just from the appearance of the sections that there must be a defin increase in the amount of fibrous tissue I certainly think the wa which you mentioned on the estimation supports that Until there evidence to the contrary I probably will continue to believe there is definite increase in collagen and fibrous tissue