·No text on this page.

·MEMOIRS OF THE SOCIETY FOR ENDOCRINOLOGY No. 3

·No text on this page.

·MEMOIRS OF THE SOCIETY FOR ENDOCRINOLOGY
No. 3
THE TECHNIQUE AND
SIGNIFICANCE OF
OESTROGEN DETERMINATIONS
Proceedings of a conference field
jointly by the Society for Endocrinology
and the Endocrinological Section of the
Royal Society of fedicine, at the Royal
Society of Afedicine, London, on 17 February, 1954
Edited on behalf of the Society for Endocrinology by
P. ECKSTEIN & S. ZUCKERMAN
Will readers please note that beginning with this number, Memoirs of
the Society for Endocrinology will be published by the Cambridge
University Press.
Copies of Memoirs 1 to 3 and future issues of Memoirs can be obtained
from the Cambridge University Press, Bentley House, 200 Euston Road,
London, NW 1.
LAMDRIDGE
AT THE UNIVERSITY PRESS
1953

·PUBLISHED BY
THE SYNDICS OF THE CAMBRIDGE UNIVERSITY PRESS
London Office: Bentley House, NWI
American Branch. New York
Agents for Canada, India, and Pakistan: Macmillan
ALL RIGHTS RESERVED
COMPOSED BY MONOTYPING SERVICE LTD
PRINTED BY DJ JEFFERY ITD, HARTENDEN, HERTS

·CONTENTS
MORNING SESSION
Dr Raymond Greene in the Chair
ac
I
IO
11
22
J. B. BROWN: A new method for the determination of ocstrogens in urine and its
application to a study of the ocstrogen cxcrction in the menstrual cycle
Discussed by M. I. Stem, J. B Brown, P M. F. Bishop, P. C. Williams, E.
Diczfalusy and H Braunsberg
W.S. BAULD. Sources of crror in the chemical determination of urinary oestrogens
Discussed by F. T. G. Irunty and W. S Bauld
R. J Boscort: Somc factors influencing the composition of urinary oestrogen
fractions
Discussed by H Braunsberg, R. J Boscott, G. I. M. Swyer and M. I. Stern
HANNCLORE BRAUNSBERG, MARGARET I STERN and G. I M Swych. Experiences with
ocstrogen determination by fluorimetry after separation by partition
chromatography
Discussed by P. M Harlow, G I. M. Swyer and M. I Stern
GENERAL DISCUSSION G F. Marrian, A E. Kellie, M. Reiss, G Rumney, R. J
Boscott, J B Brown, W. S Bauld, H Braunsberg and M 1 Stern
23
39
41
47

·CONTENTS AFTERNOON SESSION Sir Charles Dodds, FRS, in the Chair page SO 6.2 E. DicZFALUSY: Characterization of the oestrogens in human semen by countercurrent distribution and fluorimetric analysis Discussed by M I. Stern, E. Dıczfalusy, H. Braunsberg and T. Mann FL MITCHELL. The estimation, recovery and normal levels of ocstradiol in human full term placentae Discussed by E Diczfalusy, F L. Mitchell, H. Braunsberg, G. Rumcy, W. S Bauld and R. J Boscott 64 70 73 Phyllis M HARLOW, MARJORIE SANDIFORD and Mary E Woon: The biological assay of ocstroncand ocstradiol in urine Discussed by J A Loramc, M E Wood, A. Klopper, R. J. Boscott, P. H. Jellinck and E Dıczfalusy 83 P M F. Bishop The value of oestrogen determinations to the clinician 85 GENERAL DISCUSSION A S. Parkes, E. Dıczfalusy, R J Boscott, H. Braunsberg, M I Stern, J. B Brown, G Rumncy and G I. M Swyer 90 SIR CHARLES DODDs Chairman's closing remarks OS List of Contributors 96 ·A NEW METHOD FOR THE DETERMINATION OF OESTROGENS IN URINE AND ITS APPLICATION TO A STUDY OF THE OESTROGEN EXCRETION IN THE MENSTRUAL CYCLE By J. B. BROWN J'rom The Chwical Endocrinology Research Unit (Afedical Research Council), University of Ehnburgh SUMMARY A convenient chemical nic thoal for estimating cestnol. oestronc and cestradiol-178 separately 111 human ure is descibed Evidence is presented for its specificity and accuracy when applicd to the urine of men and non-pregnant women The method was applied to a study of ocstrogen cxcretion during the normal menstrual cyck, and the results obtained are described Further research on the biological actions of the oestrogens depends to a considerable cxtent on the development of suitable assay nicthods. It was for this reason that, 41 years ago, Dr Bauld and I ser to work, under Prof. Marcian's supervision in Edinburgh, to devise an accurate chemical method for estimating oestriol, oestronc and ocstradiol-178 in non-pregnancy urinc. Because of obvious difficulties, it was decided after some time that it would be wiser for Dr Bauld and inyself to work more or less independently towards the same goal. As a result two very different but cqually satisfactory methods have been developed. The only major feature common to them both is the new Kober colour icthod by which thc oestrogens arc estimated. This colour method, in which the reagents are sulphuric acid and quiol, is a further development by Dr Bauld of a method alrcady described [Brown, 1952) and will be published clsewhere by Dr Bauld (1954). It is reliable in the presence of urinc cxtracts and is sensitive enough to estimate ocstrogens in non-pregnancy urine. This colour method has been the kcy to many otherwise baffling problems. Earlier mconsistencies in the behaviour of the oestrogens and incxplicable losses in extraction procedures were found to be due almost entirely to the inadequacy of the colour or fluorescence ficthods used at the time. METHIOD The present urinary ocstrogen method is summarized in the flow sheet (Fig. 1) and will be published in niore detail elsewhere. Urine is hydrolysed by boiling for 1 hr with 15 vol % of concentrated hydrochloric acid. Rosenmund (1948) and Van Bruggen (1948) have suggested that the partial loss of oestrogens which is said to occur during acid hydrolysis is duc to oxidation, and have added reducing agents to prevent this. The efficacy of this treatment could not be confirmed, not could better hydrolysis conditions be found than those just described. The hydrolysed urine is extracted with di-ethyl ether. The Editors accept 110 responsibility for the accounts of the papers read at the Symposuin. Manuscripts are published as received from the authors.

·J. B. BROWN The ether solublc acid fraction is then removed from the ether extract. Thus is usually performed at this stage by washing the ether extract with sodium bicarbonate solution or at a later stage by ether extracting aqueous solutions adjusted to pH 9 10 9 s (Cohen & Marrian, 1934; Engel, Slaunwhite, Carter & Nathanson, 1950). At the higher pH the acid fraction is removed more effectively, but some care is required to prevent loss of oestrol. However, it was found that oestriol is easily extracted with ether from aqucous solutions at a pH as high as 30.5, provided the aqueous solutions are saturated with sodium carbonate or bicarbonate at that pH. In fact, the partition coefficient of oestrol between ether and concentrated carbonate solution of pH 10.5 is the same as between ether and saturated sodium bicarbonate. Concentrated carbonate solution of pH 10.5, Urine hydrolysed with HCI extracted with di-cthyl ether Washed with Lilier extract Sard, sodium carbonate soln pH 10,5 NaOH, NaOH wash neutralized to pH 10 with satd NaHCO, and shaken again with the ethier extracts (3) NaHCO3 Erher er'aporated (4) Water Residue dissolved in ethanol Benzene and perrolcum ether (equal vols ) added extracted with water Benzenej petroleum ether water extract (oestrol) H200, and NaOH added. methylated with dimethyl sulphate at 37°, NaOH and H2O, added cyjracted with NaOH NaOH extract (oestroncoestradiol) HBO, added, methylated with dimethyl sulphate at 37° NaOH and H, O, added extracted with benzche extracted with petroleum ether Petroleuon eflies prors washed with water Benzcie crtrari washed with water chromatographed on alumina chromatographed on alumina Oesrrorit Oestriol methyl Ostrado! methyl ether cthier metliyl etter fig 1 flow sheet ·OESTROGENS IN URINE 3 which is much more effective than sodium bicarbonate, can therefore be used for washing thcaad fraction from cther extracts without any additional loss of ocstriol. Urine extracts contain substances, presumably polyhydric phenols, which oxidize to coloured products when dissolved in alkali and are not then re-cytractable with cthcr at pH > 7. The apparently involved alkalı washics shown in the flow sheet arc for the purposc of removing these substances. The acids arc removed by washing with concentrated carbonate solution of pH 10.5. The celcr cxtract is then shaken with sodium hydroxide The polyhydric phenols oxidize and, together with some of thic ocstriol, are extracted into the sodium hydroxide layer. The sodium hydroxide is then partly ncutralized by addmg saturated sodium bicarbonate solution The aqueous layer is now a concentrated carbonate solution of pH 10, and shaking agam with the ether layer cxtracts the oestriol back into the other. The aqucous layer is then discarded, and the other is washed with saturated sodum bicarbonate to neutralize any alkalı still present. This is necessary to prevent loss of oestriol in the subscquent water wash. The cther is then washed with a small amount of water to remove the bicarbonate. Thc cther is cvaporatcd and the residuc dissolved in cthanol. The phenolic fraction is usually separated from the neutral fraction, and ocstriol from oestrone and oestradiol, by partitions between benzend and water, and benzenc and sodium hydroxide. However, difficulty was experienced in getting quantitative recovcrics of ocstriol and ocstronc with these solvents, and it was concluded that their partition coefficients do not favour sufficiently the aqucous phases It is casy in this case to alter the partition coefficients in a favourablc manner by diluting the benzene with petrolcum cther, in which the ocstrogens arc rclatively insolublc. Tlie mixture of equal volumes of benzenc and petroleum cther was found to be very satisfactory. Using this as the organic phase instcad of benzenc alonc, better recoveries of ocseriol and ocstrone arc obtained with fewer extractions and without the formation of emulsions. Sodium hydroxide is added to the oestriol extract in preparation for the next step There was reason to bclicve that much of the material in the oestrogen fractions at this stage is not truly phenolic, and the next step was designed to remove this non-phenolic material. The principle used is phase change through chemical modification of the phenol group. After a few preliminary cxperiments it was found that methylation is a very convenient method for achieving thus, especially as the ocstrogen methyl ethers can be estimated directly with the Kober colour method The complete phase change obtained by methylation is shown by the fact that whereas oestrol is extracted from benzene with water, oestriol methyl ether is extracted from sodrum hydroxide with benzene; oestrone and oestradiol are extracted from benzcne with sodium hydroxide, while their methyl cthers are extracted from sodium hydroxide with petroleum ether. The oestrogen fractions, already in alkaline solution, are conveniently methylated with dimethyl sulphate, an effective methylating agent under these conditions. The usual methods of methylating phenols with dimethyl sulphate are, however, unsatisfactory when applied to these oestrogen extracts In strongly alkalıne solutions, niethylation is ·J. B. BROWN 4 slow compared to the rate at which the dimethyl sulphate is destroyed by hydrolysis. It was then found that this destruction is considerably retarded without seriously decreasing the rate of methylation, by methylating in borate buffer solutions of pH between 10 and 11.5. The reaction takes place slowly at room temperature but is conveniently accelerated at 37°C. At this temperature the reaction is complete in 30 min. Thc methylation nixture is then made strongly alkalıne to emphasize the alkalı insolubility of the true methoxy phenols, and hydrogen peroxide is added to destroy certain substances which form a portion of the urinary contamination. Oestriol methyl ether is then extracted with benzene and the ocstrone and oestradiol methyl ethers are extracted with petroleum ether. In each case the partition completely favours the organic phases The considerable purification of ocstrogen cxtracts obtamed by this methylation and phase change procedure is shown by the fact that the benzene and petroleum etler extracts are completely colourless and all the visible pigments are discarded, apparently unchanged, with the aqucous layers. The oestrogen methyl ether fractions are then further purified, and oestrone and ocstradiol are separated from one another by chromatography on alumına columns. Stimmel (1946] was the first to apply adsorption chromatography on alumina columns to the further purification and partition of the ocstrogen fractions However, considerable losses of oestriol occurred during his procedure, and these could not be corrected without considerably increasing the amount of interfering material in the oestriol fraction. The method therefore has not been favourably received. Tlus objection, however, does not apply to oestrogen methyl ethers, for, being less polar, they are more suited to adsorption chromatography procedures. Adsorption chromatography used as a quantitative procedure has been criticized and condemned by many workers (sce Engel, 1950] In spite of this, the present technique, developed in less than a month, has been in routine usc for more than 3 years without giving trouble. It seems to be completely reliable, provided the necessary precautions are taken and all factors are rigidly standardized. Mixtures of oestrone and oestradiol methyl ethers are resolved with difficulty on columns of active alumına, but are easily resolved on columns of less active alumına. A suitable activity was obtained by adding 9-10% of water to an alumina with an original activity equivalent to Brochiman No. II. The activity of this hydrated alumina is practically unaffected by small changes in water content and solvents used with it should be saturated with water. Both these factors contnbute to a robust chromatography procedure. Mutual displacement of adsorbed substances on the columns is avoided by using dilutc extracts, already considerably purified and containing only small concentrations of oestrogen methyl ethers. The chromatography procedure is a compromise between resolution of urinc fractions on the one hand and a reproducible, quantitative method with adequate safety margins on the other. Columns are prepared with 2 g of alumina The ocstronc-oestradiol fraction in petroleum cther is adsorbed on to the column. The column is then cluted with a mixture ·OTSTROGENS IN URINE 5 of benzenc ind petrolcum cther to remove the pre-oestrone fraction. Oestronc mcthyl cther is then clutcd sith a stronger cluant, and clution is continued with the same cluant to remove the inter ocstrono-oestradiol fraction, and the ocstradiol imethyl ether is then cluted with benzenc. By using cluants of increasing strengths, fractions are obtained in volumes small enough to be evaporated directly from the colour tubes. The oestriol fraction in benzenc is similarly adsorbed on another coluinn; a preoestriol fraction, which contams the only pigment band scen during clıromatography, is cluted with a dilute solution of cthanol in benzene and the ocstriol mcthyl ether with a more concentrated solution of cthanol in benzene. The methylated ocstronc, ocstradiol and oestriol fractions arc evaporated with quinol 1. TT 1 - colours in the Kober reaction and these are corrected for spectrophotometrically by the miciliod of Allen (1950). Colour densities are measured in a spectrophotometer at the wavelength 517 ne, the absorption maximum for tlic Kober red colour, and at two Onc person can convcniently do four completo estimations from hydrolysis of urine to colorimetry in 2 days Thus allows time for preparing reagents and washing glassware and can casily be doubled for limited periods. . 1 Specificity The criterion for validity of the spectroplotometric correction is that the wavelength absorption curves of the interfering yellow-brown colours should be lincar becween the sorption maxi- To determine WALIUA W wartrugu dusul puIon CUI VCS UI LIIe yellow-brown component are in fact lucar in this region, it was necessary to apply the method to urines in which low or zero oestrogen levels might be expected. Table I shows the results obtained in a series of urines from men, postmchopausal women, children, and patients suffering from Simmond's disease. Deviation from linearity is expressed in terms of micrograms of apparent ocstrogen excreted in 24 hr Completely zero levels, implying linearity, were rarely found. The results however are not inconsistent with those which might be expected if these urines contained small amounts of oestrogens The amount of interfering yellowbrown colour produced with these urine extracts is a considerablc proportion of the total colour, and it is necessary to rely over much on the spectrophotometric method to correct for this. For instance, in the case of the male urincs shown, the density of the yellow-brown component, as calculated by the correction formula, is about 9/roths of the total in the case of oestriol, 2/3rds in the case of oestrone, and 14/1 schs in the case of oestradiol. In the light of this contamination, the significance of the small apparent ·cascs 28 oestronc oestrone 1,0 6 J. B. BROWN oestrogen content of these urmes is somewhat doubtful, especially in the case of oestriol and oestradiol. Table i l'alues found in expected low titre urines (vg/24 hr) No of Range Mean Males oestrol 08-75 3.2 3 4-8.2 57 ocstradiol 00-24 I.S Post-menopausal women 9 Oestriol 06-68 3.0 08-7.1 27 Oestradiol 00-23 03 Children aged 8 years 4 oestriol 3-3.+ oestronc 0.2-10 OS ocstradiol 0.0-02 Simmond's disease 4 ocstriol 24 oestrone 00-36 1.7 oestradiol 0.0-1.5 03 There is a certain amount of evidence, however, that these values might represent truc ocstrogen levels. The substances being measured behave chromatographically as oestrogen methyl cthers. The Kober reaction, giving a colour with an absorption maxinium at 517 mji is known to be remarkably specific for the natural oestrogens. The wave length absorption curves of colours produced with these urine fractions are completely consistent with the presence of two components, one with an absorption maximum at 517 mu and the other with a linear wavelength absorption curve in this region of the spectrum (Fig. 2). 01 12-16 0 201 O 0 16 DENSITY 2012 O 08 8 004 400 560 500 520 540 WAVELENGTH(mu) Fig 2 Wavelength absorption curves of colours produced in the Kober reaction A, oestrone fraction froin of a 24 ho quale unne, calculated by correction formula to contain: 4 ug of strone octhyl cther, B. pure cestrone metly 1 cther 14 wg. C, background material calculated from Aminus ·OESTROGENS IN URINE Further support is afforded by results obtained biologically for the cxcrction of ocstrogens in malc urinc. Although these biological methods arc liable to considerable crrors, the results obtamcd are of the same order as those obtained with the chemical ricthod. For instance, Gallagher, Kenyon, Peterson, Dorfman & Koch (1937) found in men's urinc an average oestrogen concentration cquivalent to the excretion of 10 ug of ocstronc a day, results whuch compare favourably with those shown hcrc. Accuracy The accuracy of the method was tested by a scries of recovery experiments in which known amounts of oestriol, oestronc and ocstradiol-178 were added to portions of acidhydrolysed 24 hr malo urines. Blank determinations were made at the same timc on the same urine. Results are shown in Tablc 2 and arc expressed in terms of mean values and Table 2. Recoveries from and hydrolysed male urine. (Results shown as mean % S.D. and corrected for endogenous blunk values) ug added per 24 hr urine 25-35 36-60 Oestriol 88 108 (12) 87 – 11.7 (u) Oestronc 88 +38 (11) 84 £ 5.2 (12) 91 * 7.0 (1) 83 + 3.0 (12) 82 +61 (10) 86 + 6.6 (11) Oestradiol 80+ sos (12) standard deviations. The amounts of oestrogens added per 24 hr urine specimen are shown, and results are the percentage of this recovered after subtracting the urinc blank values. Recoveries are between 80 and 90%, even at levels corresponding to 4 18 per day, which was the lowest investigated Thc standard deviations, however, increasc as the amount of added oestrogen is decreased. This is not surprising, for at the concentrations being macasured the oestrogen colour densities are so low that considerable instrumental errors arisc during their measurement. Also, at these levels the apparent ocstrogen content of the blank forms a considerablc proportion of the whole, and errors in measuring this contribute to the total error. Thesc satisfactory recovery results show that the method, whatever it measures, is at least reliable for mcasuring differences in oestrogen concentrations in the same urine. RESULTS Oestrogen excretion during the itenstrual cycle The method was then applied to a study of oestrogen excretion during normal menstrual cycles in women Urines were collected daily throughout twelve menstrual cycles from ten apparently normal women whose ages ranged from 17 to 40 years. Pregnancy intervened in three of the cycles studied. Basal temperatures were taken daily throughout ·8 J. B. BROWN cach cycle in order to compare oestrogen levels with times of expected ovulation Fig 3 shows the results obtained from a typical cycle, and illustrates a number of features common to all the cycles studied. There are two peaks of oestrogen excretion. The sharper and usually the higher peak occurs about the 13th day (or more constantly, 15 days before the onset of menstruation) --just about the time of the temperature rise, and is presumably related to ovulation, u 99 TEMP.F 9B 97 ug OESTROGEN PER 24 HR URINE OESTRIOL S OESTRONE OESTAADIOL 0 -- 3 5 7 9 11 13 15 17 19 21 23 25 27 DAY OF CYCLE Fig 3 Excretion of oestriol, oestrone and oestradiol during a menstrual cycle Days of menstrual bleeding and basal temperature curve are shown Subject 1, dge 39, para 3 Tlus as followed by a sudden drop in oestrogen excretion, which then cises to a second peak at about the 21st day of the cycle. The levels then fall before the onset of menstruation and rcach their lowest about 2 days after the onset of menstruation These trends are shown by all thrce ocstrogen fractions The finding of two peaks of oestrogen excrction during the menstrual cycle is not new and has been recorded by many workers in the past, using bioassay methods. For instance, Smith, Smith & Picus (1938) found tivo peaks of oestrogen cxcrction, onc about the rath day and the other about the 18th day of a 24-day cycle. The oestrogen levels at the first peak (calculated from their data and assunung an oestronc : oestradiol ratio of 2 . I) were oestriol, 30 kg, oestrone plus ocstradiol, 20 jg, which are very sinular to thosc found here The relative amounts of oestriol, oestrone and oestradiol found in tliese menstri ·9 OLSTROGENS IN URINL cycle urines are interesting. The amount of oestradiol is usually about half that of the ocstrone in the same urine specimen, and daily changes in the concentration of the one are reflected in parallel daily changes in the concentration of the other. The relative amounts of oestriol and ocstronc are variable and the daily changes in thic ocstriol levels are not reflected directly a clic changes in thic oestrone levels but tend to lag a day behind them. This is most noticeable during the first oestrogen risc and at the first pcak, and is shown clearly in Fig. 3 Herc, thc oestronc and oestradiol peaks occur on the 13th day, but the corresponding oestriol peak docs not occur until the 14th day. Thesc rclationships between the aniounts of oestriol, oestronc and ocstradiol are also found following the single intramuscular injection of oestradiol into postmenopausal women. Fig. 4 shows the results obtained from one such experiment. The peak cxcrction of oestronc and ecstradiol occurs on the Ist day following the injection, and the oestradiol levels are about half the ocstronc levels. The oestriol cxcrction lags bchind that of the ocstronc and ocstradiol and reaches its peak on the 2nd day following the injection This lag phenonicnon was also observed under sinular conditions by Clayton (1949). 04- 02- OESTRIOL mg EXCRETED PER 24HR URINE 0 OESTRONE OESTRADIOL C 3 4 5 DAYS 20 mg OESTRADIOL IM Fig 4 Excretion of cestriol, oestrone and vestradiol following a single tramuskular injeLOkMI (1M1) of «xstradhe-173 (20 mg) Total oestrogen recovered was 16', of that meetud The results obtained from menstrual cycles at about the time of the first peak are thereforc completely consistent with those expected following the continuous injection of oestradiol, the amounts of which vary from day to day in a cyclic manner The cvidence therefore suggests that this chemical method affords a nieasure of the excretion of oestriol, oestronc and destradiol in thic urine of men and non-pregnant women.

·IO J. B. BROWN REFERENCES Aller, W.M (1950) J cu Endocrin 10, 71 Bauld, W S (1954] Biochem) 56, 426 Brown, J B (1982] J Endocrin 8, 196 Bruggen, J T van (1948] J Lab cluu Afed 33, 207 Clayton, B E (1949) PhD Thesis, University of Edinburgh Cohen, S L & Marrian, G F (1934) Biochem } 28, 1603 Engel, L L (1950) Recent Progr Hormone Res 5, 335 Engel. L L, Slauriwhite, W R, Carter, P. & Nathanson, I T (1930) J biol Chem 185, 255 Gallagher, T F, Kenyon, A T, Peterson, D H Dorfman, R I & Koch, F C (1937) J Amer med Ass 108, 586 Rosenmund, H (1948) Helv physiol acta 6, 349 Smith, G V S, Smuth, O W & Pincus, G [1938] Amer J Physiol 121, 98 Stimmel, B F (1946) J bol Chem 162,99 DISCUSSION Miss Stern. In your calibration graph you could measure down to 2 pg of oestradiol, and yet in the actual analysis, in the postmenopausal woman for instance, you measured down to o 2 ug. IOPS Dr Brown. Yes, but I agree that the Kober method is becoming unreliable at that level, so that we do not place too much importance on very low values, say below Miss Stern. But the actual analysis was on a 24-hr specimen ? Dr Brown: A fifth of a 24-hr sample. Aliss Storni, You added your oestrogens for recovery after acıd hydrolysis Do you thunk it possible that during hydrolysis there could be some destruction? We seem to have some evidence of that Dr Brown It is generally accepted that destruction is liable to occur during hydrolysis. The recovery experiments were not designed to check that, but just to see how our extraction procedures were working Dr Bishop. Has Dr Brown any suggestion why there should be this difference in oestriol and oestrone-ocstradiol excretion in nulliparous and parous women? Dr Brown We would like to know the answer to that ourselves. Mr Williams Have you done any determinations on urine from hysterectonized women? Dr Brown We haven't many figures on hysterectomuzcd women, and those we have are on post-menopausal cases Dr Dıczfalusy It may interest you that we have checked the partition data of Dr Brown on the methylated oestrogens, using the Craig countercurrent distribution, and our data so far agree completely. Dr Hamelore Braunsberg Have you any evidence that quantitative methylation occurs in the presence of impurities? You have added the oestrogens at stage 2 and recovered then agam in the presence of urinary residues? Dr Brown Yes, we have checked that.

·SOURCES OF ERROR IN THE CHEMICAL DETERMINATION OF URINARY OESTROGENS By W. S. BAULD From the Department of Biochemistry, University of Limburcle The investigations reported here were conducted simultancously with, but independendy of, those just reported by Dr Brown. As thic method finally devcloped will be described in detail clsewhere (Bauld & Marrian, 1954), the present discussion will be confined mainly to the major difficultics which arose and the steps taken to overcome them. METIIODS Thc original method (shown in outline in Fig. 1) was based on carlier investigations by Clayton (1949). The new developnients are marked m the figure and are as follows: *The oxidation of phenolic contaminants to more acidic products by shaking an alkalıne solution in the presence of air. **Extraction with cthcr of the oestriol fraction in N-NaOH to remove ncutral and nitrogenous contaminants. ***The use of partition chromatography for the separation and purification ofocstrone and oestradiol-178, and for the final purification of ocstriol. ****Chromatographic separation of neutrals and the lipophilic ocstrogens. The difficulties encountered in the application of the method will be discussed under three headings. Errors in colorimetric assay. Loss during partition. Destruction of oestrogens during the procedure. presence of Errors in colorimetric assay The introduction by Brown (1952 a) of the quinol-aqueous H.SO, reagents was a distinct advance in producing a convenient nicthod of oestrogen determination of much greater reliability than carlier modifications of the Kober (1931) rcaction (eg. Venning, Evelyn, Harkness & Browne, 1937). However, when the method in its original form was extended to the estimation of oestrogens in the urinary and solvent residues, certain forms of intereference with colour development were found. These findings are reported in detail elsewhere (Bauld, 19544] and will be summarized only very briefly here It was found that for oestriol to form quantitatively the yellow precursor of the final pink colour there had to be an effective reductant present in the first stage action For the determination of pure oestrogen thus requirement is met by the sulphonated quinol of Brown's reagents This weak reducing agent is, however, inadequate in the of the re- II ·12 W. S, BAULD (A)! Aliquot 24 hr urine boiled ihr with 15%. HCI Eiher (B) Aqucous (discard) Ether HCO; H,O (C) Ether Aqucous (discard) Residue in EtOH/benzerie (D) HO Lipophilic fraction Hydrophilic fraction (E) Made N-NaOH* Ether (H) Shaken with N-NaOH* Acidified Washed with CO,,H,O **Ether (F) Aqueous to pH93 (discard) Ether Residue Partition chromatogram Mobile phase--Benzene Stat phase-o8N-NaOH ****Neutrals ***Oestrone ***Oestradiol-17B Ether Aqucous (discard) Residuc (G) Partition chromatogram Mobile phase—ECI Stat phase-70% MOH ***Oustrol Tag : Outline of the basic method for extraction and purificate of ostrol, ex strone and ox stradiol-175 from urme presence of oxidants arising froni solvent or urinary residues. In a typical experiment 96% of oestriol was recovered on cxtraction from 13 N-H,SO, when the reagent for the colour reaction contained 1.4% (w/v) of free quinol, whereas only 65% was rccovered which the free quinol in the reagent was 0 3% (w/v) The difficulty was overcome by the addition of quinol to the rcagent immediately before usc. Morcover, for all the oestrogens the second stage of the reaction, the conversion of yellow to pink, appcars to be a graded oxidation: with insufficient oxidation, there is incomplete conversion and with excessive oxidation, fading of the pink occurs In the presence of various oxıdızıng and reducing contaminants it is therefore essential to have conditions of maximuni stability for this conversion. Since changes in the intensity of ·OLSTROGEN PETLRMINATIONS 13 thic pink occur in dilute H,SO, even at room temperature, the dilution with water immediately prior to colorimetry, as required in the original rcaction (Brown, 1952 a). was omitted. In the second stage, the important factors involved were shown to be: H.SO, concentration, time of locating, and composition of the reagent. Their cffects are shown in Fig. 2. In cach experiment, cleven samples of oestradiol-178 (25 mg) were licated at 100°C for 20 min with 4 ml. of 2% (w/v) quinol in 60% (v/v) 11,50,,cooled, diluted with 3.5 ml. of aqueous II,SOs, and rchcated. Optical densities obtained with the green (Ilford 604) and violet (Ilford 601) filters are plotted against time of reheating. With Type B H.SO, the conversion of yellow to pink (as cvidenced by alteration in thic optical densities obtained with both filters) at a second stage concentration of 60% was incomplete even with prolonged lcating. When the concentration was dropped to 46%, the conversion to pink was markedly facilitated, but the colour was now susceptable to fading. Morcover, a marked difference was found between the different types of analytical reagent H,SO,, both in case of conversion and in intensity of colour. That Type A H,So, is still not optimal is shown by the incrcased density obtained by the addition of Cu++ (280 pg) in the second stage of thc rcaction. Thesc results and simular findings with the other ocstrogens led to mmor but important changes in the original reagents and reaction described by Brown (1952 a) We are now convinced of the accuracy and reliability of the modified colour reaction in the presence of contaminants from solvents and urinc. Since the development of fluoresccncc m 11,50, merely represents the first stas of 040 030 ONE C4- FILTER 604 Os 03 020 020- oral CPTCAL DENS TY ***TYPE A 4,59 602 2nd, STAGE — TYPE A *602 -TRACE of cu . - TPE O . -607. Out! TYPE B . 014 012 chot 8 006 FILTER 604 04- CO2 TO TM OF HEAT 2M STAGE > Fız 2 Effect of variation in conditions of the secund stage on the contation of yellow to pink ·14 W. S. BAULD the Kober reaction, the considerations briefly discussed above will apply at least in part to fluorimetric methods of estimation of oestrogen. Brown (1952 b] stressed the necessity for a rcducing agent in the development of fluorescence by oestriol. Bates (1982) contested this point, but uscd oestronc as his example. Most methods mect the requirement by the use of ethanol as a solvent. Veldhuis (1953) found a fifteen-fold increase in the fluorescence of oestriol when ethanol was included in the H,SO, reagent. Conversion of the fluorescent compound to the pink derivativc-an cvent which in vicw of the low concentrations of ocstrogen used for fluonmetry would be demonstrablc only by decreased fluorescencc-is one of the possible forms of quenching'. This may cxplain why batches of H.SO, from different suppliers vary in their ability to produce fluorescence of ocstrogen (scc Engel, Slaunwhite, Carter & Nachanson, 1950). The instability of fluorescence methods in the presence of contaminants from solvents, urine or blood was recognized by Finkelstein (1982) and Veldhuis (1953), who correct for this form of quenching' by mcasuring the fluorescence of a known amount of ocstrogen in the presence and absence of an aliquot of their final extract. If fluorescence methods are to have a widc application their stability must be increased Loss during partition The variable factors affecting the elution of oestrogens from the partition chromatograms of the final method will be reported in detail clscwliere (Bauld, 1954 b). Two of the morc important causes of variation are shown in Fig 3. As the partition cocfficients of ocstronc and ocstradiol-178 between agucous alkali and benzenc depend upon temperature at which the solvents are cquilibrated, this factor becomes important in determinmg the clution pattern. The variation with percolation rate was shown to be the zc| OL STRONG OESTRADIOL 19 mt/n, 21°C. 20/16C -*-/,16C „Us OF OUSTROCEN PERMI OF ELUATE 6 100 20 24 20 32 is 32 560 84 66 12 16 DO 84 88 02 VOLUME OF ELUATE (MI) Ff 3 Effect of variations in percolaton rate of solvent, and in tcmxrature on clution pattern of estronc and cestradiol 173 Al columns WCIC 1 X 13 cm NaOH (OSN) and benzene were cquilibrated at the temperatures shown, the aqueous (stationary) phase was mured with celite sus in the proportiunts of 8 ml pet log of celure Sofures (1) Sorg) were applied in organic (mobile) phise m 31 ml washes, and collection of the cluste was begin after the third wash entered the chromatogram The mobile phax was changed at 40 mil to EC-tenzene 3/1 ·OESTROGEN DETERMINATIONS IS dependent upon the alteration in cross-sectional areas of the mobile and stationary pliases resulting from the variation in tlic firnmess with which the chromatograns were packed. The tendency to trailing clution boundaries seen in the figure is in accord with the non- Incar partition isotherm of ocstrogens in this system. In the final method the chromatogranis arc run in a rooni controlled at 18 +0.5°C; and the columns arc packed sufficiently firmly to give a percolation rate of 10 ml /hr with gravity flow. 'Destriction of oestrogens during the extraction procedure Because of possible instability of the colour reaction it is difficult to distinguish destruction', in the sense of modification of the ocstrogen molecule to make it nonchromogenic in thc Kober reaction, from apparent loss of oestrogen duc to interference with colour development. In the ensuing discussion, 'destruction' will be applıcd only to losses (in terms of colorimetric estimation) preventable by changes in the procedure prior to colormictry. Such losses were of four kinds. i 'Destruction' of oestrol on extraction from dilute HC! When the reliability of the projected method was tested by recovery of ocstrogens from dilute HCI, satisfactory results were obtained with oestronc and ocstradiol-178, but recoveries of only 64 + 11% were obtained in twenty-seven cxpcrinients with 25 mg of ocstriol added to 575 ml of 1.3 N-HCI. Investigation showed thac the principal loss occurred during the first stage of the purification procedure extraction with cthcr, washing with bicarbonate and water, and distillation of the extract. Yılds at this stage were not improved when various other methods of ether purification were substituted for the routinc onc of shaking with silver oxide (Werner, 1933), or when distillation in an atınosphere of nitrogen, or the addition of ascorbic or 1-amino-2-naphthol-4-sulphonic acids were employed. It was known (Rosenmund, 1948) that Fc+++ caused an apprcciable destruction of oestrogens on boiling in dilutc HC! It was found that completo destruction of ocstrol in dilute solution occurred on prolonged standing at room temperature with o 1 mg Fc+++ per 100 ml. of 1 3N-HCl. Even with immcdiatc cxtraction there were losses up to 50% Thus metal-catalysed destruction was prevented, however, by ascorbic acid (I part in 5000) Morcover, analytical grade reagents were used throughout this investig.2- tron, and no traces of Fe+++ could be demonstrated in the extracts. It was therefore considered unlikely that thus type of destruction was the causc of the low yields on cxtraction of ocstriol from dilute acid. Oestrol was recovered quantitatively in the first stage of the purification procedure (extraction with cther, washing with bicarbonate and water, distillation of the cther) when quinol (1 part in 5000) was added to the 1.3N-HCI at the start. This antioxidant, unlıke ascorbic acid and 1-amino-2-naphthol-4-sulphonic acid, is not completely removed from ether by the bicarbonate washes. Its cffectiveness therefore suggested that the destruction' took place during distillation of the solvent. In the absence of quinol, quantitative recoveries from this first stage were found when the dilute HCl was neutralized before extraction, or when ocstriol was extracted from water or 1.3N-H,SO. Thus suggested that HCl was not being completely removed from the ether by the bicarbonate washes and was destroying' oestriol during the dis- ·W. S. BAULD 1. the Kober reaction, the considerations bricfly discussed above will apply at least in part to fluorimetric methods of estimation of oestrogen, Brown (1932 b) stressed the necessity for a reducing agent in the developnient of fluorescence by oestriol. Bates (1952) contested this point, but used ocstronie as his cxamplc. Most incohods meet the requirement by the use of cthanol as a solvent. Veldhuis (1953) found a fifteen-fold increasc in the fluorescence of oestriol when cthanol was included in thcll,SO, reagent Conversion of the fluorescent compound to the pink derivative-an event which in vicw of the low concentrations of oestrogen used for fluorinctry would be demonstrable only by decrcased fluorescence-mis onc of the possiblc forms of quenching'. This may cxplain why batches of H,SO, from different suppliers vary in their ability to produce fluorescence of ocstrogen (sce Engel, Slaunwhite, Carter & Nathanson, 1950). The instability of fluorescence methods in the presence of contaminants from solvents, urinc or blood was recognized by Finkelstcın (1952) and Veldhuis (1953), who correct for this form of ‘qucnching' by measuring thic fluorescence of a known amount of oestrogen in the presence and absence of an aliquot of their final extract. If fluorescence nicthods arc to have a wide application their stability must be incrcasçd. Loss during partition Thc variable factors affecting thc elution of ocstrogens from the partition chromator grams of the final method will be reported in detail clsewhere [Bauld, 1954 b). Two of the more important causes of variation are shown in Fig. 3. As the particion coefficients of ocstrone and ocstradiol-178 bctwceu aquicous alkali and benzenc depend upon the temperature at which the solvents are cquilibrated, this factor becomes important in determining the clution pattern The variation with percolation rate was shown to be 20 OL STRONE OL STRADIOL 18 m/n, 2°C -70/60 --/* ,16 L9 OF DESTROGEN PER MI OF ELUATE 100 B 12 28 12 38 40 44 43 52 56 60 6 6B 17 VOLUME OF ELUATE (ml) Fig 3 Effect of variations in percolation rate of solvent, and in temperature, on elution pattera of ocstronc and oestradio- ·OESTROGEN DETERMINATIONS 17 shaken with a small amount of N-NaOH, and the mixture was then sluaken with blcarbonate to lower the pH of the aqucous phasc to 8.5 bcforc discarding the combined wash. This resulted in an appreciable purification of the final cxtracts. Edianolic solutions of the separate fractions of malc urine, obtained with this modification of the method, were evaporated to dryness by heating in a stream of air with and without added ocstrogens, and the colour reaction was carried out. The results for oestriol are shown in Fig. 4 and Table 1. OESTRIOL (150) CESTAIOL (2549) URNIE EK TOH REMOVED PRESENCE OF Quia (img) URINE BLANK, CCH REMONTO PRESEXCE OF Qual Obre -OESTE (259).URTE BAN, NO CU DNG MEMY 077 OF LOH 0404 004 obce Ost- 527 046- ALISNUI IN 032 028 024 0204 OG 042 008 Oo suo S60 600 WAVELENGTH Fig 4 Absorption spectra of colours produced in modified quinol-aqucous sulphuric acid caction (75 ml) by pure oestrol and ocstriol heated with che oestrol fraction of male urute The optical density at the peak of the oestrogen absorption was considerably lower for oestrol in the presence of urinary residues, whether a correction for the non-pink component was applied or not. The discrepancy was significantly less when the quinol required in the first stage of the colour reaction was added to the ethanolic solution before removal of the solvent. It would seem that heating oestrol with urinary residues causes a 'destruction' which is largely preventable by quinol In an effort to determine which group of compounds was causing this destruction', the alkalization wash of the original ether extract was acidified, extracted with ether, washed with bicarbonate and water, and the solvent removed by dıstıllation The residuc, ·16 W, S. BAULD tillation of the solvent. With the introduction of the strong alkalization of the ches extract (discussed later) no loss on cxtraction from HCl occurred. Since hydrolysis with this acid gives less contaminated urmary extracts than docs cquimolar H.SO, HCl has becn rctained in the final procedure. 2. 'Destruction' by heating with solrent residues. Other investigators have reported the destruction of oestrogen when an cthanolic solution is licated in a stream of air (sec Haani, 1980; Biteran & Sykes, 19531. Tlus finding was confirmed with certain batches of ethanol purified by azcotropic distillatio with benzcue and with ethanol containing appreciable quantities of acctaldchyde. The loss occurred cven when the ethanol was removed by licating 111 an atmosphere of otrogen under diminished pressure. However, it was found that ethanol, punfied during uinufacture by simple distillation only and rendered acetaldelıydc-free by a two-stage treatment with -phenylenediaminc, could be evaporated from ocserogcsis without causing any destruction. The cffect of traces of aldclıyde was ealanced by cther residucs. Routinely, therefore, aldchyde-frcc cthanolis used for the transfer of ocstrogens to tubes for colorsmetry, and as an additional precancion dhe qumol required in the first stage of the colour reaction is added. With these precautions no loss of ocstrogens occurs when the ethanol is removed by lcating in a stream of air. 3. 'Destruction' ly hcaling with srimary residues. When the method was applied to the recovery of ocstrogens added to acid-lıydrolysed male urme, additional purification stages were found desirable. The major group of brown contaminants arose from colourless precursors during the extraction procedure, the change occurring whenever che plucnole residues contacted alkali. Thus, in spite of shaking the benzene solution containing ocstrone and ocstradiol-178 with alkals, and subsequently washmg it with carbonate (Step H, Fig 1), a brown band was always left on tlic chromatogranı after clusion of the neutrals with benzeno. Waslug thc cther extracts with sulphite or bisulplure effected a purification at clus paraculat stage, but did not result 112 any apprectable decrease in the contamination of the final fractions The substances concerned appeared to be so suscepuble to oxidation to brown derivatives that purification was best achieved by removal rather than by reduction The casiest approach to this was an extension of the procedure for shakmg malkaline solution, followed by waskmg with weak alkalı. Accordingly, the initial other extract was washed once with bicarbonate and 0 Table 1 Corrected opnical densuy (Ecor) af colout produced in the modified quuml-apueses sulphunc acid reaction (7-5 mil vol) ly pure destrol and cestniał kcated in erlan} tik dhe ocstriol fraction of male 1135€ Eros 2 Loss Oestrol (-5 ug) 0.269 Oestriol (25 ug), urine blank, EtOH renionycd in pres na of qumol (i ng) Urte blank, EOH renoved in presence of quino 0034 0179 38 4 Oestrol (25 vg), urine blank, no qurol during removal of EtOH Urme blank, no qunol during removal of ECOH 1 2 0 281 17 3 022 S.

·OLSTROGIN DETERMINATIONS 19 buffer of pH 10.s prior to the alkalization a significant improvement in recovery resulted The use of a 10.5 buffer was introduced by Brown, as has already becn described. A 10.5 wash after alkalization did not iniprove the recoveries Ic thus appeared that the effectiveness of this wash lay in the fact that it removed substances from tlic urinary extracts which, if present during thic shaking with strong alkalı, facilitated the oxidation of ocstradiol-178. Expt. 3 demonstrated that the causes of destruction' were not completely removed because at lower levels of oestrogen the recoverics dropped to S1 8.2%. Prelınımary experiments established that some of the loss was due to alkalization of the benzene solution (Stage II. Fig. 1). Elimination of this stage of the punification gave recoveries of 85 + 5.8% when 10 pg of oestradhol-17% were added to i of a 24 hr specimen of hydrolysed male urinc. As shown by Expres. sand, however, the results became unsatisfactory as thcamount of added ocstrogen was decreased. There was a marked variation from urme to urinc. It was thought possiblc thiät this destruction of minute amounts of oestradiol-178 was occurring during the removal of cthanol after transfer to tubes for colorınıctry. This effect at the higher levels of ocstrogens luis already been discussed; it was shown to be largely overcome by the presence of qumol. Addition of 1 kg of ocstradiol-17B to this fraction of the urinary cxtract just before transfer for colorınıctry resulted in losses of 10-20% with certain urines even in tlic presence of quinol, thc losses being greater when the background colour in the fractions was high. Further purification was therefore rcquired. Saponification of the residuc from the chromatogram cluatc by refuxing for 30 min in N-NaOH gave a fraction (obtained by benzene cxtraction of the acidificd aqucous phase) which was considerably cleaner and did not destroy oestradiol-178 when heated with it in cthanol. Possibly the harmful agents are esters of phenolic acids. Expt. 7 showed that, with this final modification, oestradiol-17B (1 xg), when added to alıquot of hydrolyscd niale urine, was recovered to the cxtent of 84 6.9%. The outline of the final method is shown in Fig. 5 The factors affecting the recovery of ocstradiol-17ß also apply to ocstronc but to a lesser extent. Oestrone was found to be the most stable of the three ocstrogens, and with it no saponification was necessary. Table 3 Recovery of oestrol and oestrone alled to acid-hydrolysed male urine Oestrogen added to 1 of 24 lir No of Recou cry specimen of eypts Mean SD acid-lıydrolysed male urine Oestriol jo ug 12 2 12 I2 Oestrone 10 Hg 16 2 16 I 13 NB All recoveries are cnirected for endogenous oestrogen content 83 + 47 87 – 75 81 t 105 87 – 44 92 £ 41 90 + 70 ·W. S. BAULD 18 in amounts corresponding to 1/50 of the 24-lır urinc volume, caused a 'destruction of 25-26% of oestriol (10 ug) when a 1 ml. ethanolic solution was cvaporated to dryness by heating in a stream of air. Whicn the quinol required in the first stage of the colour reaction was added, the 'destruction' was only 8-10%. Since the urinary residues used in the experiment described in Fig. 4 were purified by alkalization of the initial ether extract, the indication was that the damaging agents were not completely removed by this step. Routinely, therefore, the cthanol used for thic transfer of the final fractions to tubes for colorimetry was removed in the presence of the quinol required in the colour reaction. 4. 'Destruction during extraction. Recovery of oestrogen added to purc solution and taken through the extraction procedure was found to be ncarly quantitative. Recovery from tirine, however, was found to be incomplete especially for oestradiol, and a summary of thc alterations in procedure nccessary for improvement of yıelds is shown in Table 2. Table 2. Recovery of ocstradiol added to acid-hydrolysed male urine Oestradiol added to aliquot No of % Recovery Expt. hydrolyscd male dctcr- Procedure Mean + SD. urine minations ug 25 25 33 21 Alkalization of cther at stage (C) 61 + 8.5 10 5 wash before alkalization of cthcr at 71 13.5 stage (C) IO 10 I 3 IS As in Expt 2 51 – 8.2 4 IO As in Expt 2, but no washing m stage (H) 85 58 S I 2 As in Expt. 4 70 +61 6 8 As in Expt 4 81 198 7 As in Expt 4, but with saponfication of 84 £ 69 ocstradiol eluate from chromatogram N 1 All recoveries are corrected for endogenous ocstrogen content *The basic incthod is outlined in Tig 1, and the changes made are noted below I 12 Expt i demonstrated that a definite loss occurred on extraction of oestradiol-178 from acid-hydrolysed male urine when the procedure shown in Fig. 1 was followed. When the oestrogen (25 rg) was added to the urinary fraction immediately before chromatography, recoveries were improved (100, 96%; 88, 92% in two experiments). Evidently the major loss was occurring during the preliminary purification Since it is known that the oxidation of strongly reducing phenols may facilitate the oxidation of other substances (sce Waters, 1946), it was thought possible that oxidative destruction of ocstradiol-178 was occurring during the alkalization of the initial clier extract In model experiments it was shown that addition of quinol (1-2 mg) to the ether before the alkalization increased the loss of oestradiol-17B, whercas quinol added after the alkalızation did not Moreover, the addition of quinol to a pure solution of oestradiol in ether, followed by alkalization, caused a 15-20% loss These experiments indicated the necessity for further purification of the ether extract before alkalization Expt. 2 showed that when the cther extract was washed with a concentrated ·21 0.11+ 1,1 1.0 0.002 O010 I 2 OESTROGEN DETERMINATIONS Saponification, however, did cffect a considerabic purification of the ostriol fraction and, for convenience, was incorporated into the procedure at an carlier stage Tlic rccoveries for oestroncand ocstriol are shown in Table 3. Investigations of the optimal conditions for hydrolysis are in progress. Most of the cxperiments were done with 1-2mmo-2-naphthol-f-sulphonic acid added to the urine before hydrolysis as suggested by Van Bruggen (1948), but no cvidence of any protective action by cus reagent Izs ct been obtimed. Typical analyses are shown in Table 4. In spite of the various purification stages, which remove more than 95% of the non-specific chromogens, thic specificity of the method Table 4. Typical mislyses of the sestrogen content of 24-lir specintens of urine Subject Type Ostrogen E, E E Ecorr ug/24 hr W.B Male Ocstrio 0249 0.222 0.134 0.030 0.296 0.26: 0.1001 0031 R2 Oestronc 0.379 0.348 0227 0045 0 371 0334 0207 00:45 7.0 Oestradiol 0.166 0.138 0.096 0007 O ISO 0.156 0.009 MF Menopausal Oestriol 0.112 0097 0.079 0.006 0.032 0.0.27 0018 Oestrone 0.280 0 245 0.189 0.332 0279 0.216 0005 Oestradiol 0.183 0.147 O.TOS 0.264 0.189 W.M 16th day of Oestriol 0.194 0223 0094 0 079 cycle 0.190 0214 0096 0071 20.5 Oestrone 0.416 0.009 02419 0.076 0.439 0.418 0 244 0 076 Ocstradiol 0 268 0 268 0.034 optica) density at 480 ml 0258 0257 0187 0034 60 E, - optical density at Sız s my for oestrol and oestro xe at Siso for oestradiol-173 E, = optical density at 945 mfor oestrol and ocstrone at Sis my for oestradiol-173 at the low excretion levels shown in the table, is largely dependent upon the specificity of the Kober reaction. The adequacy of the colour correction has already been discussed in full by Dr Brown. Some of the background colour arises from the effect of solvent residues on the quinol added during the transfer of the final fractions to tubes for colorimetry, as is evident from blank analyses Much of the residual chromogen, however, represents urinary residues which have not been removed in the purification procedures The niethod may readily be applied to the routine analyses of urine specimens. Four determinations are completed in 9-10 min-hr. A high resolving spectrophoto meter is the only special piece of apparatus required. The procedures involved are simple enough to be carried out by a competent technician given 3-4 weeks of special training, provided he is under the closc supervision of a research worker thorouglıly familiar with the sources of error of the method 0001 0001 0228 o 200 ·W. S. BAULD of 24 hr specimen Posler ht with 15 HICI Ezker Ether Aqueous (discard) pk 10 5 wasio pH 14-e*p!? 10 wash HICO,, 11,0 Ether Aqueous Residue in 6% EtOll-benzcuc feliscard} HO Hydroplulic fraction Made N-NAOH Lipophilic fraction Ether Residuc Partition cluomistogram 08 N-NaOH--Stat puasa Benzene Mob phase Ethict Reflux 30 min (Discard) Ether 0-8 ml Neutrals (discard) Aqueous to 9 3 10-30 ml 40-90 ml, 10% H2SO. SO Residue CO; Reflux 30 mm H20 in N-NaOH Acidify Residuc Oestrane Benzche Echer Aqueous Aqucows Ether Residuc Partition chromztogram 70% MeOH-Stat phase EtClMob phase 14-30 ml -Ostrio! Benzerie CO" Residuc Oestradul-178 Figs Outline of final method for the extractum and purification of oestu, OC Strone and oestradiol-378 from us ·THE SIGNIFICANCE OF COMPONENTS OF URINARY OESTROGEN FRACTIONS By R. J. BOSCOTT From the Department of Anatomy, University of Birmingham SUMMARY 1 The question of the identity of natural destrogens in human beings and the attempts to assess chemically the physiological estrogena state of man are discussed - Paper chroinatographic methods have been descloped for the qualitative analysis of uruary oestrogens, the majority of the clier-soluble acidic and phenolic compounds and certain artefacts in hydrolysed urines T. cs - + - ( the composition of uruary oestrogen fractions. The thesis that ocstriol, ocstradiol-17B and ocstrone are the only compounds of interest in assessing thc oestrogen state of human beings, both from the chnical and biochemical vicwpoints, is still a matter of conjecture. Thus the compound isolated by 17 some measure is required of the total ocstrogen production by the ovarics, adrenal and placenta Lieberman & Teich (1953) in their review on the biochemistry of steroid hornioncs, suggest that attack of the benzenoid ring of tlic oestrogen moleculc is an important step in the catabolism of oestrogens. This view is also supported by the preluminary report of Garst & Friedgood (1952) on the clieinistry of the compound they have isolated. Smith & Smith (1941; 1952) have published evidence that a urinary compound other than ocstronc can be converted by treatment with zinc and hydrochloric acid into a product with increased oestrogenic activity. Furthermore, Smith & Smith (1982) have reported the presence of an unidentificd labile oestrogen in the follicular fluid of women treated preoperatively with gonadotroplun. Previously, Huffman & Grollman (1947) suggested that 16-keto-oestradiol-178 may be the labile compound responsible for the SUCI as 7-keto-oestronc, may be responsible for the higher oestrogenic activity after hydrogenation with the nascent hydrogen produced by the zinc. Similar theories could be postulated in relation to the unidentificd oestrogen detected biologically in follicular Auid. With regard to these suggestions, Marrian (1949) and Pincus (1949) have pointed out 23 ·22 W. S. BAULD REFERENCES Bates, I Bruld, Baul3, Bauld, Ditman, J & Sykes, J F (1953] Senince, 117, 355 Brown, J 13 (1952.1) J Endocrm 8, 196 Brown, J B (19326] Chu Tondition Colloqurd on Tudocrinology, 2, 132 Bruggen, J Tan (1948) } Lab du Mid 33. 207 Clayton, B C (1949) l'hD Thesis, University of Edinburgh Engel, L , Sunwlute, W R, Carter, P & Nathansen, I T (1950) J biol Chem 185, 255 Funkelstein, M (1942] Ads endocr, Copenhagen, 10, 149 Hacını, E O (1950] Amer pharm. As 39. $44 Kob.r, S (1931) Blolie. Z 239. 204 Rosenmund, (1948) Huly playstol pharmacol act1, 6, 349 Veldhus, A (1953] J Mol Chem 202, 107 Venning, C H Exclyn, K A, Harkness, E V & Browne, J SL (19371 J bol Chema 120, 225 Waters, WA (1940) Roy Prager Chem. XLI, 152 Werner, A (1933] Analyst, 58,335 DISCUSSION Dr Prunty. May I ask Dr Bauld if he has ever obtained a negative figure for the corrected extinction--In other words, is there any cvidence that the absorption of his coloured background is not lincar with some urines? Dr Bauld Yes, we did obtain this carlicr on in the method, particularly in the case of ocstradiol That was one of the principal reasons why we had to introduce the final saponification stage That is, in thic ocstradiol determinations prior to the saponifications which we routinely do now, there was a negative correction of an extinction coefficient ofo 010-0.020 in certain urines. Dr Prunty So you tlunk there is reasonable cvidence that linear extinction occurs over the range you are correcting on? Dr Bauld With tlic residues as they are in the final method, yes.

·URINARY OESTROGEN IRACTIONS Afethol C1. The extracts were enzynically hydrolysed (sec method B), cxtracted and fractionated according to method A and Fig. 1. Method C2. The extracts were hydrolysed with lıydrochloric acid and worked up according to mcthod A and Fig. 1. Ether entrusts ( benzene luyer) Extracted with saturated NaHCO, contaifungos. N3,S,Os NaHCO, cxtracts- -Ether bach-wash-Ether solution Saturated with Naci, pH adjusted to 1-2 cxtracted with cther Ether washcl with saturatoul NaCl solution Buffered with 2 ml cach of os NaH,PO, and osNa,,os. Ether removed by distillation in the presence of the buffer. Residus + qucous layer diluted with benzene Benzene extracted with equal vols. of oos% Na $.03 Fraction 7 (Echer extract) Aqueous extracts -Benzone back-wash- Benzene solution Saturated with Naci pH adjusted to 1, cxtracted with ether Ether washed with saturated NaCl solution. Extracted with a solution containing 9% Na,Co, and 05% Na,s,o, Fraction II (Ether extract) Aqucous Na,CO, cxtracts --Benzene back-wash Benzene solution Fruction III (isolated as for Fraction I) Extracted with 2N KOH Aqueous extracts --- -Benzene back-wash- Neutral Fraction Fraction IV' (isolated as for Fraction I) Fig i Unnaty fractionation flow shett Arethod D The procedure of Engel (1950) was also used for the preparation of Fraction I and Fraction V (Engel's oestrogen fraction). Saturation of the hydrolysed urine with NaCl ·24 R. J. BOSCOTT that the influence of possible synergists or antagonists of ocstrogenic activity may invalidate conclusions derived from experiments based upon bioassay figures. In addition, Marrian & Bauld (1981) lave criticized the claims of Smith & Smith (1941). Valucs for the ratio of urinary oestrogens to pestrogen antagonists or thcir mctabolites (rather than ocstrogen values alone) may give a truer biochemical assessment of the general, physiological oestrogen state in fan. With this principle in mind, Salter, Hunm & Ocsterling (1948) measured ratios of urinary ocstrogens ('oestroids) to neutral 17-ketosteroids in man. Szcgo & Roberts (1953) have shown that cortisone and cortisol (Kendall's compound F) cxlubit antagonism towards the biochemical and physiological cffects of ocstrogen on the uteri of ovaricctomized rats. Since improved methods for the quantitative estimation of oestrogens and other steroidal metabolites are now available (Engel, 1950; scc also review by Lieberman & Teich, 1953), the work of Salter et al. (1948) should be extended to mclude ratios of the urinary adrenocortical steroids to ocstrogens. The objects of the present work were: (1) Development of improved methods for the extraction of urinary ocstrogens, acidic and phenolic compounds (ii) Complete analyscs by paper chromatography of urinary extracts containing these compounds. (ui) Study of factors influencing the composition of the urinary extracts. (iv) Investigation of the effects of acid hydrolysis conditions on acıd-labılc ocstrogenoestradiol-179. METHODS 1. Preparation of extracts Twenty four-hour collections of urine were preserved by the addition of s ml. of both chloroform and acetic acid. Sıx-lır altquots of urine were lıydrolysed according to Methods A, B, C, C., Dor E (below). Method A Six-hour aliquots of 24 hr collections of urme were subjected to hydrolysis by boiling under a reflux condenser for 30 min after the addition of 15 vols. % concentrated HCI, and a small volume of benzene. The cooled hydrolyscd urine was saturated with Naci and extracted four times with its quarter volume of ether. The ether extracts were fractionated, yıelding Fractions I-IV as shown diagrammatically in Fig. 1. Method B Urmes were borled for 10 mm at pH s, buffered to pH 6 2, and incubated for 36 hr with bacterial glycuronidase-(Sigma Chemical Company), eising 100 mg/400 ml. of uranc. Enzymic activity after the 36 hr hydrolysis period was checked by incubation of a sample with phenolphthalein glycuronide. The hydrolysed urine was then adjusted to pH 1 and fractionated according to Method A and Fig 1. Methods CI and C2 Aqueous solutions of urinary steroidal conjugates were prepared according to the mctliod of Edwards, Kellie & Wade (1953) ·URINARY OESTROGEN FRACTIONS 27 RESULTS 1 I of 3 24 200 2000 (A). Chromatography of Fraction I (Arethod A) An aliquot of urine cquivalent to from hr collection of human urinc was used. Fig. 2 and the accompanying key represents thic total number of compounds found in 200 hunian urines, which were detectable with tests 3-5. Far fewer compounds were detected on chromatograms of Fraction I obtained from human patients and rats fed highly purified dicts (cssentially free from vegetable components) than when a morc complex dict was given. It is interesting to note that a methylated phenolic acid vanillic acid was detected in Fraction I (method A-punfied dict), thus recalling die work of Maclagan & Wilkinson (1951) on the in vivo mcthylation of n-butyl 3 : 5- duodo-4-hydroxybenzoatc. 20% KCI frunt 30 ZOZ KCI 37°C (A) 54cm Mhr o CHCH COOH+H20(1) 24 cm 104 Mr Ftg 2 Map of phenolic acids found in Fraction Key Colour of the dyes (Test 4) Compounds s. 6, 17-yellow Compounds 15, 16, 25-orange-yellow Compounds, 2, 7, 18, 19, 22, 24, 27, 28, 33, 37-orange Compound 8-orange-pink Compounds 12, 14, 20-pale pink Compounds 3, 10, 21, 26, 29, 32, 34-red Compound 9-magenta Compounds 11, 23, 31-violet Compounds 4, 13-mauve Compound 36 (Test shyellow fading tv greysh blue Compound 38 (Test 2)-green fluorescence becoming pink-magenta in colour on long starling or after Test 3 ·R. J. BOSCOTT 26 was used to depress the formation of cmulsions during the extraction of the urine with ether. Mcthod E Method C1 (enzymic liydrolysis) but with fractionation according to method D 15 under investigation II. Paper chromatography A. Techniques employed Some of the paper chromatographic mcthods used have been described briefly clscwhere (Boscott 1952, 1954; Boscott & Bickel, 1953; Boscott & Cooke, 1954). The essential features of the methods used arc recorded in Figs. 2-8. Wlicnever possible, solutions of electrolytes were used instead of organic solvents for the development of the paper chromatograms. The use of electrolyte solutions saturated with an organic solvent --sce Figs. 2 and 7— as single phase' systemis prevents the formation of diffusc ocstrogen zones obtained when only thc clcctrolyte solution is used. A solution of s% sodium formate m water at rooni temperature may be used instead of 20% KCl at 3.4°C for the chromatography of Fraction I. Paper impregnated with a salt or other compound which is appreciably soluble in one component of a non-aqucous developing solvent mixture may be used for the chroínatography of many compounds with poor solubility in aqueous stationary organic solvents commonly used as the stationary phases in conventional paper chromatography. The hydrotropic agent, sodium p-toluenesulphonate, is appreciably soluble in absolute methanol, but only slightly in tolucne. Whatman papers Nos. 542 or 3MM were therefore sprayed with a half-saturated aqucous solution of sodium p-toluenesulphionate, leaving a i in margin unsprayed. The papers were dried and the urinary oestrogen extracts applied to the paper in acetone solution as narrow bands 1 in. from the unsprayed margin of the paper The unsprayed edge of the paper was placed in the developing solvent--1 vol absolute methanol to 9 vols, toluenc—without previous equilibration; chromatograpluc development then proceeded by the capillary ascent technique. Using paper No. 542, chroniatogram development took 12 hr for 30 cm development, whereas the tlucker paper No, 3MM took only s hr. An extension of this technique shows promise for the chromatography of neutral steroids and may be regarded as a development from the methods of Sakal & Merill (1953] and Boscott (1952) B. Tests for locating compounds on chromatograms Each chromatogram was investigated by the following tests in the order given. i Fluorescence of the compound at neutral pH in UV light (Mercury vapour lamp with Wood's glass filter) 2. Fluorescence and colour development with mucrodrops (1 mm diameter) of 90% (v/v) H.SO spaced at suitable intervals. 3. Fluorescence and colour changes after spraying with 9% NaCO3 solution 4. Formation of azo dyes by spraying with a sodium carbonate solution of diazotised p-amunophenyldiethylaminoethylsulphone (I.C I. 5091). 5. Colour changes after spraying with 2N-KOH shortly after test 4.

·URINARY OESTROGEN FRACTIONS 29 53 51 52 50 49 40 41 35 23 15 - Berald hydantoint ZOZ KCI 37" (50 cm) 12 CoMo (CH3COOH + H20) Ist developing solvent 22 cm (thr) Fig Map of compounds found in Fraction II, Method A, human urine Capillary ascent technique Kcy to Fig 3* Test Number 1 Spot Number 3 n 4 I OR Y(F) Reaction give BR(C) YG(F) PV(C) Y(C) SG(F) G(F) PICS V(C) B(F) OR PV 2 Oestriol 3 4 S. 6 R OR 7 8 9 OR(F) Y(F) B(F) B(F) B(F) G(F) BY(F) OR(F) || |||||||| to (BG(F) Y(C) BW(F) II 12 13 R(C) ·4 10 . 36 38 28 R, J. BOSCOTT Key of Fig 2 (continued) Filentily of the compounds, where known Compound - ferulic acid -vanillic acid 1.4 -5-hydroxy-ß undolylacctic acid (tentative) 18 -4-hydroxybenzoic acid 24 hty 26 28 29 30 34 -4-hydroxyphenylpropionic acid --phenylpyruvic acid ----ındolylacetic acid The significance of the tyrosinc and phenylalanine metabolites Nos. 24, 26, 28, 29 and 34 has been reported previously or is in press (Aterman, Boscott & Cooke, 1953; Boscott & Bickel, 1953; Boscott & Cooke, 1954). The excretion of these compounds is influenced by conditions with abnormal ascorbic acid metabolism, arthritis, gastro-intestinal disturbances, certain anaemias, pregnancy and liver disease. The excretion of compound 25 was found to be greatly incrcased in a case of essential hypertension. Compound 36 was obscrved only in phenylketonuric urine, as also was the vastly increased urinary excretion of compound 30. The tryptophane or tyrosine metabolite (Robinson, 1952) s-hydroxy-3-ındolylacetic acid (compound 14) has been observed in greatly increased amounts in urines from a small number of patients with both steatorrhoea and liver disease as well as one cancer patient with metastases in the liver. Compound 38-- B-indolylacetic acid-was detected (test 2) in abnormally high amounts in chromatograms of urines from patients with steatorrhoea and other gastrointestinal disorders Marrianolic acid has not been detected in Fraction I of the late pregnancy urines examined. If present in amounts of from 0.5-1 mg/24 hr specimen, it would have been detected close to compound 8, and given test rcactions similar to oestrol (Figs. 3 and 4) (B). Chromatography of Fraction II (1ſethods A, B, CI and C2) An aliquot of from one-tenth to one-fortieth of a 24-hr urine collection was used for the paper chromatography by the methods indicated in Figs. 3 and 4. The type of compounds detected is indicated in the keys accompanying Figs 3 and 4. Fraction II (method A) contained many more components than Fraction II (methods B, C1 and C2) Fraction II (method B and Cı), chromatographed on paper strips using 10% KÇI saturated with N-hexyl ether as developing solvent, contained several compounds giving positive reactions with alkalıne tetrazolium reagents. With the chromatographic developing solvent used cortisone and cortisol are known to have Rf values considerably higher than oestriol. Oestriol was detected readily in late pregnancy urines by this method, concomitant purple and red pigments being easily separated chromatographically. Five-substituted hydantoins arising as artefacts from the condensation of urea with certain a-ketoacids may occur in ction II (method A) but not in Fraction II (methods B and C) (Boscott & Bickel, 1953] The chromatographic detection of some of these compounds is shown in Fig. 4 5-(4-hydroxybenzal)-hydantoin arising from urea and ·31 URINARY OESTROGEN FRACTIONS 4-hydroxyphenylpyruvic acid was found in abnormal amounts in chromatogranıs of acid hydrolysed wrines from most patients with disturbed ascorbic acid metabolism, nutritional deficiencies, toxacmia of pregnancy and liver discasc. Indeed, Felix (1944) has employed test doses of 4-hydroxyphenylpyruvic acid as a liver function test in humans. A number of compounds give green fluorescence and pink colours with 90% sulphuric acid (test 2) reminiscent of the Kober rcaction for oestrogen (scc Fig. 3). The effect of dilure HCI on the compounds detected would serve to differentiatc ocstrogens and other compounds from those giving a green fluorescence micrcly in response to a change in pH. The interaction of the compounds detected, with the cellulose of the paper and the acid uscd, nay in part be responsible for the reactions given. (019 b VA ya S3 QR4 Greys OR 13 G3 24 MOS Grey YCS VF5 GIS IVF3 BYES QE 05 000000000 On :D 0:19 @BOX YCS b@.00, So02 01:00 -ba0f0000;60 - 00000000290 e 632 102 P44 Oestriot TCS X BYF1 ILI OFI vt Reference Consponde • 5 (ROM bow) - Namun Fruction I (Method A) Pregnancy w Wydtem x 5-64 OM BI Mydlantan Cu) Fogla uning (Hepat tu) Fig4 Chromatography of unnary oestrogens en Fraction !! A Whatman paper NS 842 Developing solvent-2K,CO, (J1 cm) Capulary ascent technique B Developing solsent---10% KCl in water saturated with N-heryl cther Key 2. 3. 4. 5. Test numbers, 1-Fluorescence SH–Shadow, C-Colour, 0-Odour, B-Blue, DG-Bluish greers, BV-Blush violet, BR-Browt, G-Greer, P-Pink, R-Red, OR-Orange, OR-P-Orange pink, V--Violet, Y-Yellow, M-Mauve, Si-Saimon red, SW-Sheet, WB-Whitish blue The polyphenolic compounds 29 and 43 (pyrocatechol, method A) are excreted generally in abnornal amounts by patients with nutritional deficiencies, liver disease, aplasse, microcytic and refractory megaloblastic anaemias. Compounds 41, 52,53 and S4 are frequently obtained from the urines of patients with steatorrhoea (method A). Interest attaches to the presence of resorcinol, compound fó (yellow dye test 3) in Fraction II, in view of its antithyroid properaes (Arnoct & Doniach, 1952) Relatively large amounts of resorcinol are found occasionally in urines. Rfvalues of polyphenolic compounds are given in Table 1.

·30 R. J. BOSCOTT Test Number Spot Number 2 بیه 4 4 IV(E) PV(C) B(F) Y(C) B(F) LYC). YG(F) s-p-hydroxybenzalhydantoin BV(F) 1(F) BV(F) DVF) 11111 III! BN(F) V(F) 11 BV(F) V(F) Polisi IIIIII! - V(F) BG(F) 23 2.4 25 26 27 28 29 30 31 32 33 34 35. 36 B(F) BŘ(C) BV(F) R(C) VF) 38 39 40 41 110 43 Pyrocatechol BG(C) BGE) D(F) YC) GF) 191111* *|||||||| 48 V(C) Y-BR(C) R(C) *For abbrcviations see Fig 4 ·33 URINARY OESTROGEN FRACTIONS Key of Fig s.* Test number Spot Nuniber 1 w + 5 Rcaction given BR(C) 1 2. G(F) GF) SOR 3. . $ 6 YG(F) BG(F) BG(F) G(F) 7. 8. G(F) B() (F) Y(C) VO) 9 10 11. - BV(F) Y(C) BIF) I BV(F) 13. 14 IS. OR(F) 16. DV(F) R(C) B(F) BG(C) P(C) B(F) B(F) OR R V(F) 17. OR(F) 18. B(F) 19 20 21 D(F) 22 23. 244 25 V(T) 26 27 G(F) 28 SH 29 30 31 G{F) 32 SW(0) 33 G(F) 3+ *For abbres nations see Fig 4 G(F) UT11161116 IIIIIIIIIIIIIIII!! B(F) V(F) OR compound 4 in relation to liver disease In general, chromatography of Fraction III (method A)—from liver disease patients revealed the presence of a larger number of fluorescent compounds than was found in Fraction III (method A) normal urines. Occasionally three compounds giving on paper a green fluorescence with 90% sulphuric acid, similar to oestrogens, were detected in this fraction. The possibility that the partition coefficient of oestradiol and oestrone distributed between benzene and 9% sodium ·32 R. ). BOSCOTT Table 1. Rf values of polyphenolic compounds Paper: Whiatnian No. 542 Developing solvent: 10% KCl solution in water Distance moved from origin, by solvent: 50 cm Compound Rf valuc Mcthod of Detection 0.73 Hydroquinone Resorcinol 0.70 069 0 66 0.64 "Test 3. Pyrocatechol 4-Methylcatechol Tolulıydroquinone 3-Methylcatechol Pyrogallol Orcinol 0.64 Test 3 0.62 Test 3. Brown colour STest 4. Yellow Test S. Salinon S Test 3. Bluc-green colour Test 4. Violet halo Brownish-violct colour ſ Brown colour Green fluorescence Test 3. Brown colour STest 4 Yellow Test s. Yellow Test 3 Bluc-violet fluorescence Test 4. Yellow Test 3. Violet fluorescence Test 4. Magenta Test 3. Yellow colour green Auorescence Test 4 Brown Test 3. Orange colour, bluc-violet Test 4. Yellow 0.60 Phloroglucinol 053 Protocatechuic acid 0.49 Gallic acid 0.38 Naphthoresorcinol 0.22 (C). Chromatograplıy of Fraction 111 (methods A, B, Ct and C2) The compounds found in Fraction III (method A) are illustrated in Fig. S. Fraction III (methods B, C and C2) contained fewer components than Fraction III (method A), Comments have been made [Aterman, Boscott & Cooke, 1953) on the significance of 34 32 30 28 20 20% KCI 37*C (34cm) rigin CSHS (CH3COOH + H2O) 124cm) (uhe) Fig Map of compounds in Fraction II, Method A, Whatman paper No 512 ·35 URINARY OESTROGEN FRACTIONS carbonate inay be modified by the hydrotropic action of abnormal amounts of other compounds extracted by sodium carbonate, should be considered, especially in view of che comments by Bates (1952) on co-solubility plicnomena influencing the extraction of ocstrogens froni organic solvents. (D). Chromatography of Fraction IV (methods A, B, C and C3) and Fraction V' (method D) Representative chromatograms of these fractions are illustrated in Figs. 6, 7 and 8. Seitrone Jestrol GF 2 PC2 GGF3 BGF3 BF3R RO R4 DVS AVF3 VFS OR4 OR 4 VF5 Vf3 BCF SORP AF 3 pv5 OR 4 BF3 883 14°Oezbron] 0 0.0 D1000000.00 2011 Ooecondo 068 YF3 R4 A.Oestone GFZ Mochis Destrad of 178 Oestradiol 17a 131 14 Equilenin Oestrad.l.17 VF2 BF3 YF 3 YF3 GF2 ]GF2 BF3 3453 BF3 OVES BAC3, ORC3, Starting ling BRC3 161? Reference Compounds Oestradiol-174 Fraction KZ After treatment Method I with 15 vols IHCE Synthetic diet Phenylhetonunc Male aged fys Fraction 2 Method C2 Toxaemua of Pregnancy Fraction Methods Tomaemua of pregnancy Fig 7 Chromatograms of urinary oestrogen fractions and reference compounds Whatman paper No 542 Descloping solvent--5% K,PO, in water, saturated with N-butanol Capillary ascenie technique (30 cm) (For key, see Fig Rf values of the oestrogens with two of the solvents used are given in Table 2. Equilenın is valuable as a reference marker on the chromatograms in view of the strong violet fluorescence it gives with alkalı (test 3) and the characteristic red colour (test 4) Some nineteen components have been observed in Fractions IV (method A) and V these tractions. The chromatographic methods illustrated in Fig. 6 were used primarily for the detection of compounds accompanying the oestrogens rather than the oestrogens themselves, which were not separated in sufficiently restricted zones Compound 17 (Fig. 6) ·34 R.). BOSCOTT 19 18 12 17 16 11 15 13 No Butanol cquilibrated 20% K3PO4 satution 22cm 5 Ist developing solvent 20% K3PO4 solution saturated with N-Amylalcohol 35cm Fug 6 Map of compounds in Fraction V, Method D, obtained from human pregnancy urine Key of Fig o* Test Number Spot Number I 2 3 4 S G(F) 2 R SOR 3 4 BR(C) PV(C) LOR(F) OR(F) G(E) B(F) BRC BV(F) BV(F) BV(F) B(P) B(F) S 6 7 8 9 PV IO Oestriol G(F) R(C) G(F) G(F) R(C) YG (weak) | |||||||||||||||IIBE! I Oestradiol-17B 12 Ocstrone OR OR PV PV V(F) BG(F) 13 14 IS 16 PV (funt) OR BR.-Y (faint) {Pic Y(C) W-BF 17 PV PV P 18 19 *For abbreviations see Fig 4 SW(O) ·URINARY OLSTROGEN TRACTIONS 37 Y4 GF2 Yo intensa GFT ODE ge 0000 JOS 2NKON ORFZ V81 Y - CRI (2NKOHT SVF1 ORY Oestrone PC-2 VFI Funt 2012 PC2 PY5 ART4-304 004 PV5 Equilenin M3 MS VF3R VFI OF ORYF?NK291 - GF AF3 Pestudio 170 VFI Oestradiol la CRP4 GFZ Phi Thi 652 NOPC2 075 YOR ORP4F3•v) AF R4-BF1 OFI 16-Ketooestradiol LUFI 6 F12 NEON) GF2 (stable) ORYS YGN L2NCO c's OR (Fant) Oestrol PCZE Intan Lab RYSI GFZPCZ OFI PVS Oestradiol - 178 Fraction Fraction V Fruction TV after HCI Method D Method C Further purified treatment Normal adult Hepatitis by Method D Adult female (equivalent to fraction Y, Method D Pregnancy urine 2NKOR CYFY Oan Starting line Fig & Chroznatograms of urinary ocstrogen fractions and reference compounds Whatman raper Ny 542, impregnated with Na-p-toluene sulphonate Developing solvent vols tolucne + I Vol nethanol Capıllary ascent technique (30 cm), shr *For abbreviations see Fig 4 Auoresced a brilliant sky blue are illustrated in Figs. 7 and 8. Further experiments of a similar nature may lead both to a better understanding of colour and fluorescence reactions for the storoidal oestrogens and to the recognition of factors leading to the destruction of oestrogens during urinary acid hydrolysis or extraction procedures. DISCUSSION Judging by the variations in composition of human urines, it would appear unprobable that consistently good recoveries would be achieved of 10-200 mg amounts of steroidal oestrogens added to the urine prior to acid hydrolysis The different amounts of protein, carbohydrate, polyphenolic compounds, iron and copper encountered in urines of patients with different pathological conditions might influence the degree of destruction of urinary oestrogens subjected to hydrolytic and extraction procedures. Indeed, the addıtion of zinc to the acıd hydrolysis medium (Smith & Smith, 1941] would destroy lughly reactive carbonyl compounds-for example, quinoncs, and aldehyde derivatives from sugars—which would otherwise react with the phenolic oestrogens The zinc would also remove traces of copper present which would tend to catalyse the oxidation of polyphenolic and other compounds to reactive quinones ·36 R, J. BOSCOTT Table 2. Rf values of oestrogens Rf values Method Compound Whatman Paper No 542 Developing solvent- 5% K PO, saturated with n-butyl alcohol Whatman paper No 3MM impregnated with sodium p-tolucne sulphonate. Developing solvent- Tolucne-metlianol (9) 002 O II 021 021 Oestriol 16-Keto-oestradiol-173 Oestradiol-17% Oestradiol-17 Oestrone A-Oestronc Equilin Equilenin Violet pigment (Fraction IV (Method A)) 0.40 0.40 O 30 028 O 35 026 0 30 022 OSO 047 0 47 0.46 000 0.08 may conceivably give a colour with the Kober reagent in view of its reaction with 90% sulphuric acid on paper (test 2) Fig 7 shows diagrammatically chromatograms of urinary oestrogen fractions, the components being detected by tests 2, 3 and 4. The developing solvent was prepared equilibrating 300 ml. of s% K,PO, solution with 100 ml. of n-butanol. The resulting more aqueous layer was used as the mobile phasc. Fig. 7 illustrates the use of paper impregnated with sodium p-toluenesulphonate and a non-aqueous organic solvent mixture for the chromatography of Fractions IV and V. Tests Ing were applied to the chromatograms A compound closely resembling 16-keto- Oestradiol-17B in Rfvalue, and fluorescence reactions (test 2) was detected in Fraction IV (method C2) late pregnancy urine, but not in urines from a small series of normal individuals and from one patient with hepatitis. Other compounds were also detected in the same zone as the 16-keto-oestradiol-178 like material by tests I and 3 30 L.. 7 . ' . oestrogen with acıd for .. udied chromatographically (Figs. 7 and 8). The acid solution of the oestrogen rapidly assumed a pink colour during boiling and fluoresced a brilliant green in UV light Towards the end of the heatnu maline The colour reaction so obtained was saturated will indul anu Lauluv ..... ts which ·URINARY OTSTROGEN FRACTIONS 39 Samples of the many rcfcrcice compounds used in these studies were donated by Dr K. Miescher (Ciba), Dr C. Djcrassi (fornictly Syntex), Dr W.J. Tindall (Organon), Dr F. Hartley (B.D.11. Ltd.), Dr E. B. Hershberg (Scherings, U.S.A.), Dr M. Huffnın, Prof. E. Lcdcrcr, Dr J. B. Jepson, and Dr H. G. Bray. Generous supplies of I.C.I. sogt were provided by Dr F. L. Rose (i C.I.). The author is greatly indebtci to Miss A. M. Coolc for valuable technical assistance. REFERENCES Amort, D G & Donuch, I (1952) Basham J 50. 47) Aterman, K. Bxcort RJ A Coole, W T. (1953) Sund! GLI Wed s. 385 latex, R W (1953] Ch Pound.fun Collaper on Indmeligy. 2. 115 Rascote, RJ (1952) Chem. E Ink 472 Ukyscott, R. (1954) Chem G Ind 244 Bwant, R. J & Backel, II (1931) Scand J din & 136 dr s. 180 Buscott, R. J & Cooke, W T.(1954Qum JJ (in press) Eduardk, R. W.H, Kellc. A E & Wade, A P (19531 Vom Sur Enfexrin No 2, 93 Engel, L L (190) Root Payo ilimuone Res 5, 345 Felux, (1944) 2 rhyrol Chem 281, 36 Gars, J 1 & Fredgrad, 11 D (1932) Sciomer, 116,65 Huffman, M N & Grollman, A (1947), Enjivamalgy. 41, 12 Liebermun, S & Tech, S (1953] Pumacol Rr S, 285 Maclagan, NF. & Wilkinson, J 11 (1951) Nature, Lon!, 168, 251 Martian, G F (120) keren Poker Humane Rrs 4. 58 Marian, G T & Bauld, W S 11951) Actu enjoyCopenhagen, 7. 240 Pearlman, W. (1919] In 71ic Hormones, Ed G Pincus & K v Thmann, Voll, P 400, New York Academia Press Pancus, G (194Recent Post Hormone Res 4. 57 Robinson, R. R (1952) Chem Ind 358 Sakal, E. B & Merall, CJ (1953) Sache, 117. 45! Salter, W T., Hunin, TD & Oesterling. M ) (19481 ) din Londom 8, 295 Slaunu hute, W RJ, Ekman, G. Engel L L. Nathansın, 1 T. Pincus, G & Carlo J (1951) Arts andorr, Copenhucen. 7. 321 Smith, O W & South, G V (1941) Endexronology. 28, 740 Snuth, O WA Smith, G V (1952) Recent Page Iformiune Res 7. 20) Szego, C. M & Roberts, S (1953) Recent Try Hmone Res 8. 419 Veldhus, A H (1953] J brol Chem 202, 107 DISCUSSION Dr Hannelore Bralınsberg: Have you found that samples of commercial ocstrol contain any fluorogenic impunity? Dr Boscott. In our samples, as far as I remember, we only get onc component, but we have seen trace impurities in some of the other ocstrogens. We find, for example, some traces of equilenin in some of the ocstrone samples, and we found a trace impurity in a sample of Ar-oestrone, and also a trace impunity in our 16-kcto-x-oestradiol. I also recall having separated from crude oestriol obtained from pregnancy urine, a few crystals of a compound with a melang point at least 2°C higher than that of ocstriol. This compound gave colour and fluorescence reactions very simular to oestrol More recently we paper-chromatographed thus material and found that this 'ocstriol always produced a tail, whereas pure oestriol formed a nice compact spot. We suspect that the tail may be ascnbed to part of the impurity present. Dr Styer. That is most interesting to us. We never succeeded in obtaining a sample of ·38 R. J, BOSCOTT Veldhuis (1953) has reported good recoveries of oestriol from a few micrograms of ocstriol glycuronide added to plasma and hydrolysed with bacterial glycuronidase The present writer has encountered some urines which inactivate B-glycuronidase more than others It is therefore advisable to test the urinc (with phenolphthalein glycuronide) for remaining enzymic activity after the standard period of enzynic hydrolysis has taken place. The addition to the urine of the enzyme in several portions at intervals, rather than all at once, appears to be advantageous. The use of an incrt atmosphere in the enzyme hydrolysis vessel is recommended to obviate the autoxıdation of the liberated urinary components. The enzymıc method of hydrolysis of ocstrogen conjugates used in conjunction with the method of Edwards, Kellie & Wade (1953) for the extraction of urinary steroidal conjugates, may prove to be the method of choice in the study of urinary oestrogens, since the destruction of labile compounds wou be miniinised, and few artefacts would be formed. A disadvantage of thc enzymic procedure appears to be an increased contamination of the ocstriol (Fraction II) with compounds-presumably adrenocortical steroidal metabolites-giving positive reactions with alkalıne tetrazolium reagents. The use of Engel's ocstrogen extraction procedure (1950) would minimize contamination of this nature, but would lead to destruction of alkalı-labilc oestrogens if such exist. Method II, C2 is preferable to method A ifa non-enzymic hydrolysis techmiquc is resorted to Fewer artefacts are then produced. The smaller number of components found in Fractions I-IV from human patients and experimental anımals fed synthetic or purificd diets leads to the conclusion that these diets should be given when metabolic studies on isotopically-labelled oestrogens are undertaken. The identification and recognition of the urinary non-steroidal metabolites of oestrogens would then prove less difficult, sifice minimal numbers of extrancous compounds unrelated to normal body constituents would be encountered. In the present studies, several urinary components have been identificd by classical means, others have been recognized by feeding specific precursors. Paper chromatography using tests 1-5 and supplementary microchemical tests, and the use of reference compounds have led to the tentative identification of the compounds detected in human urine. The studies on oestradiol-17may prove of interest in relation to the mechanism of the Kober and fluorescence rcactions for oestrogens. Further, similar studies are in progress on the compounds produced when oestrogens are treated with mineral acıds. An extension of the work of Slaunwlute, Ekman, Engel, Nathanson, Pincus & Carlo (1981), but based on paper chromatographic rather than countercurrent separation of oestrogen fractions, and their bioassay, are required for a more complete understanding of the nature of urinary oestrogens. The author wishes to thank Prof S Zuckerman for encouragement in this research The clinical material was kindly made available by physicians of the United Birminghain Hospitals Group, in particular Prof. H. C. McLaren, Prof A. P. Thomson, D. W.T. Cooke, Dr H. Bickel, and by Dr K. Aterman. Dr C. E Dent (University College Hospital, London), provided urines from a child receiving a synthetic diet.

·EXPERIENCES WITH OESTROGEN DETERMINATION BY FLUORIMETRY AFTER SEPARATION BY PARTITION CHROMATOGRAPHY By HANNELORE BRAUNSBERG, MARGARET I. STERN and G.I. M. SWYER Irom University College Hospital Afcdical School, London, Il'.C.1 METHOD The method used in our laboratory for the determination of oestrogens in body fluids consists of tlırec parts: I. Hydrolysis, cxtraction and partial purification by partition between scveral pairs of solvents Il Further purification of the phenolic niatcrial by partition chromatography on a celite column, III. Fluorimetric estimation of ocstronc, oestradiol and ocstriol in the three fractions obtained. The technique will first bc outlined, and then the significance of the results we have so far obtained will be discussed TECIINIQUE For urine, the hydrolysis and extraction used is essentially that described by Finkelstein (1953). The oestrogens are liberated from their conjugates by boiling with concentrated hydrochloric acid (15 vols / 100 vols urinc) and are extracted with other, after saturating the aqueous phase with sodium chloride. The ether extract is washed to remove strongly acidic compounds, reduced to a small volume, and cxtracted with aqucous N/1 caustic soda. The pooled aqucous extract is adjusted to pH 8-9 and extracted with ether. The residue obtained after evaporation of the cther contains the weakly acidic material including phenols. The phenolic residue is further purified by partition chromatography on a celite column containing aqueous caustic soda as stationary phase The solvents used for elution have been modified to yield purer ocstrogen fractions. Instead of benzene (Swyer & Braunsberg, 1951) a mixture of benzene and petroleum ether (4 1) is at present being used for the elution of oestrone and oestradiol, and the 3 i chloroform-butanol inixture (Stern & Swyer, 1982] has been replaced by a 17 3 chloroform-butanol muxture. The elution pattern was found to vary with temperature, and chromatogaphıc separation is now carried out at 22° +05°C The flow rate should not be faster than approximately 0.5 ml./mın to ensure equilibrium on the column. Fractions of approximately 4 ml. are collected by a mechanical automatic fraction collector. A preliminary fluorimetric analysis ('fractional marker assay') is carried out on a small aliquot, usually 1/8 of each fraction. By this mcans we are able to ascertain whuch of the 41 ·40 R. J. BOSCOTT purc oestriol. We suppose that all oestriol samples produced from pregnancy urine are impure, no matter how pure they are alleged to be. We wondered whether we could at some stage of our own work obtain some synthetic oestriol, which might be free from this troublesome contaminant, as we suppose it to be. Miss Stern: Might I add that we would be very glad to have some of this oestrol of Dr Boscott's, so that we could chromatograph it and sce if we get the same effect with our own samples, and find out if it is just one substance Dr Hannelore Braunsberg: I would go a step further and ask for some of your crystals of contaminant, too, for comparison. Dr Boscott: They are not absolutely free from other compounds; we have detected about five or sıx trace impurities in the unidentificd oestriol contaminant.

·OESTROGEN DETERMINATION BY FLUORIMETRY the highest and lowest quantities found during the proliferative and the progestational phases respectively are given (Tablc 3). In the same tablc arc shown the results of ocstrogen determinacions on the urines of various other female subjects. Table 3. Urinary excition of oestrogens by female stbjects (vg/24 hr) Subject Oostrong Oestradiol Oestriol Simmond's disease 33* Nonc detected by fractional marker assay Very small amount detected Nonc detected by fractional marker assay Nonc detected by fractional marker assay Oligonicnorchoca 39* 9* Free 2.5 Total 16+ 03 1.3 11.3* 14 34 Jaundicc, aged 14 years; spider nacvi, hydatid cyst (proliferatryc phase of cycle) Amenorrhoca, aged 25 years, tall, obese, acromegalic features Normal female proliferative phase (range) progestational phase (range) Adiposc sy nism Nil-17 39-11.5 3-191 22 o 5-43 1o* 5-41 5-97 Nonc detected by fractional marker assay 8* is Obesity, vidlısın, aged 35 ycars Post-menopausal 43* 19 5* None detected by fractional marker assay *Evidence of quenchung and'or non-speafic fluorescence In pregnancy, the urinary oestrogens are present in greatly increased amounts, so that methods far less sensitive than ours may be quite adequate. Indeed, before accurate fluorimetry can be undertaken, more than a thousand-fold dilution of the chromatographed extracts may be necessary, cven when only so ml of urine have been processed. Table 4 shows some of the results we have obtained An attempt has been made to estimatc the oestrogen content of a saniple of ovarian cyst fluid. After acid and alkaline hydrolysis, followed by extraction and cliomatography, no oestrogen could be detected fluorimetricallyTo another aliquot 20, 30 and so ug of oestrone, ocstradiol and oestriol respectively were added, followed by the same processing, recoveries of 20-50% only were obtained. The reasons for these losses are being investigated for they arc most gerinanc to the problem of blood oestrogen assay.

·HANNELORE BRAUNSBERG, MARGARET 1. STERN AND G. 1. M. SWYER strated; four showed some quenching but no 'non-specific' Auorescence; while twentyfour showed quenching and the presence of soine fluorescent impurity of twenty determinations on pregnancy uritics, thirteen were satisfactory, two were quenched without non-specific fluorescence, and in five there was both quenching and nonspecific fluorescence. Finally, of eleven determinations on male urnes (including those of two boys) only two were free from interference and nine showed quenching in the presence of Auorescent impurities. We are still inclined to be rather suspicious of the results of any determinations in wluch interference with fluorescence can be demonstrated, though the errors may sometimes be quite small. Further attempts to obtain purer fractions--such as the use of improved solvent inixtures for chromatography already mentioned-may lead to a high proportion of satisfactory results. With a highly sensitive fluorımcter it is sometimes possible to ‘dilute out' any quenching substanccs. RESULTS Table 2 gives the results of oestrogen dctcrmination on the urines of various niale subjects. Clearly, niany 'normal males' are niccded before an adequatc basclinc is established Morcover, determinations on successive daily outputs must be made beforc the day-to-day variation can be assesscd Tablc : Urinary excretion of oestrogens by male subjects (ug/24 hır) Subject and age cstrone Ocstradiol Ocstriol African (28 years)** S.5* None detected by fractional marker assay 11* 15 6+ African (29 years) African (39 years) 78+ 33 27 18* Boy (9 years), Adipose gynim Boy (12 ycars), Adipose gynism None detected by fractional marker assay approx 4 I3* *Evidence of quenching and/or non-specific fluorescence **Urine extracted with carbon tetrachloride, a poor solvent for oestradiol and destrol So far we have been able to carry out only a pilot survey of the urinary oestrogen overetion during the menstrual cycle. Since our methods have been undergoing con- -- -------+++ 1.

·OESTROGEN DETERMINATION BY FLUORIMETRY Finkelstem, A1.Ilesin, S & Kixh, W (1947) Prx Soc erp Bu, N1.64,4 Stern, ALI & Swyc, GIM (1952) Nur, Lont, 16). 7 Styer, GIM & Braunsbers, 11 (1931) Ltdocrin 7. Ir DISCUSSION Afiss IIarlow: May I ask Dr Swyer how hc detected the background fluorescence between oestronc and oestradiol. How do you know that it was not onc of those two compounds? Is it a different colour? Dr Stryer: With pure ocstronc and oestradiol there was practically no background fluorescence between them at all. Miss Storer. I think in this case there was only one cluate tube between the separations, and it probably contained a muxture of a small amount of oestronc and a small amount of oestradiol overlapping. Moss Harlowr: We also found thic different substance Dr Braunsberg mentioned with the oestriol (from Organon). Flow do you account for the higher than 100% recovery of Ocstriol? Dr Sieger. I am not sure that we can I would regard it as an experimental crror. Mfoss Harlour: We have found that also. It is rather surprising, sccing that part of the ocstriol added was removed with the greenish background fluorescence. Dr Srver: It all depends upon this question of what appcars to be non-specific fluorescent material which contaminates the oestrol, and until we are ablc to chromatograph some rcally purc ocstriol, I do not think we shall have the answer to that part of the problem.

·46 HANNELORE BRAUNSBERG, MARGARET I, STERN AND G. I, M. SWYER Table 4. Uninary excretion of oestrogens is pregnancy Subject and duration of pregnancy Ocstronc Oestradiol Oestriol (wecks) (ug/24 hr) (mg/24 hr) 22 33* 22 180 40* 108* II 03 325 19 Krohn's disease 27 30 33 204 365* 12 35 12 (approx) 240 124* Pre-eclampic r 28 toxacmia 134 193 33 5* 36 230* 1.25 14.7* *Evidence of quenching and/or non-specific fluorescence on the DISCUSSION Further studies must be directed, not only towards obtaining far more values in normal and abnormal patients, but also to a number of technical details We lack data optinum conditions for hydrolysis and extraction. Studies involving the addition of known amounts of conjugated ocstrogens, before hydrolysis and extraction, are still needed Although our present first mobile phase in the chromatography procedure cffects a better separation of ocstrone and oestradiol, the former still tends to be cluted too carly so that the oestrone fraction from some extracts may be contaminated. Further 51121plification of the extraction and chromatographic procedure would obviously be very desirable, though it is difficult to see how it can be achueved without loss of accuracy. The use of a monochromator and a more intense source of light in the actual measurcment of fluorescence will, we hope, reduce the incidence of interf and background fluorescence. Apparatus at present under construction should meet this need. Oestriol has presented a special problem of its own in the chroniatograply of a number of samples of oestriol from different sources, we have consistently found that the first eluate fraction contains a highly fluorogenic substance, presumably not oestrol The main fraction of the steroid is eluted in the cxpected place and the typical pink colour due to oestrol is observed when a sufficiently large quantity of the dried compound is heated with phosphoric acid On the other hand, the first fraction gives a fluorescence and colour more like that of oestrone and oestradiol when similarly treated. For this reason, we doubt the validity of all our oestrol measurements This problem still requires elucidation. The clinical value of oestrogen determinations will inevitably be limited until adequate methods for blood oestrogens have become avauable The provision of these might well be regarded as the ultimate requirement in this field REFERENCES Braunsberg, H, Osborn, S B & Stern, MI (1934) J Endocrm ti, 177 Finkelstein, M (1953Add emtort, Copenhagen, 10, 149 ·GENERAL DISCUSSION 49 Oestrogens in urine during the dienstrual cycle (148/24hr) Comparison of the methods of Brou'n and Bald Brown Bauld Subject Day of qcle Oestriol Oestrone Oestradiol Ocstriol Oestronc Oestradiol 178 178 W.M. 16 216 84 6.0 20 5 11.1 60 LR. TOO 96 36 65 8,0 3.4 E.M. IS 186 200 7S 190 21.0 60 D.P. 20 98 188 6.2 IS 16.0 4 S SA. 9 10+ 11.4 39 II.S 9.8 I.S ST 22 127 13.2 3.8 127 170 3.8 I was very interested in Dr Boscott's paper. It was good to learn that soniconc has at last commenced a thorough study of the nature of the phenols and plenolic acids in urine. Such a study will be interesting in itself; it may also be of great importance to those concerned in the determuation of urinary oestrogens, since some of these substances may be ones wluch interfere in the final deterniinations of oestrogens by the Kober or the fluorescence methods. However, I venture to suggest to Dr Boscott that he should not be content with paper chromatographic mcthods. My view is that while these methods are very valuable indeed for prelunary work, they are no substitutc for isolation and identification by the classical methods of organic chemistry. Dr Boscott raised the question of whether we should be satisfied with determinations of oestriol, oestronc and tradiol, and the same question arosc in Dr Swyer's paper as well as in the discussions. I think we should not be satisficd. There are various indications, the ones we heard this morning, and there are many more in the literaturc, that there are oestrogens—or rather Kober chromogens—in urine other than thesc thrce. These addıtional oestrogens must be isolated and identified and quantitative methods will also have to be worked out for them. Dr Boscott also referred in his paper to the Smiths' zinc-HCl method for hydrolysing urine. There is no doubt that in this procedure there occurs a reduction of oestrone to oestradiol and to 3-hydroxyoestratrienc. Possibly there may be also some inhibition of oxidative destruction of oestrogens durmg hydrolysis, and I believe that Dr Bauld might say that the zinc-HCl treatment may eliminate some of the substances which interfere in the Kober reaction In view of the complicated nature of what goes on in this procedure I believe that we would be well advised to forget about it for routine determination of oestrogens in urine.

·GENERAL DISCUSSION (MORNING SESSION) Prof Maman. I want to start my remarks by saying something about the first two papers in the programme-those of Drs Brown and Bauld. Dr Brown was kind cnough to mention my name in connection with the researches which he and Dr Bauld have been carrying out, but I want to make it clear that my own part in thus work was actually a very small one. Just over four years ago I asked them to try to develop a convenient and accurate method for the determination of urinary oestrogens which would be sensitive enough to be applied to samples of urine collected during the menstrual cycle. Apart from making this mutial request I have contributed little if anything to their work. My opinion of their work, which is perhaps somewhat biased, is that they have both achieved something very well worth while. Working independently, but nevertheless helping one another all the time, they have each devised a workable and convenient method for the reasonably accurate determination of as little as s ug cach of oestriol, oestronc and oestradiol in a 24 hr urine sample. Either of these methods can be used as a routine procedure for the daily determination of oestrogens throughout the menstrual cycle Dr Brown was very cautious about the significance of the values which he and Dr Bauld obtained for the ocstrogen excretion in men, However, there is considerable evidence that the substances which they were measuring in male urines were in all probability oestriol, oestronc and oestradiol, and their recovery experiments suggest that they were decerınınıng thcsc with tolerable accuracy I think, therefore, that Dr Brown's remarks on this matter may have been somewhat over-Cautious Both Dr Brown and Dr Bauld had misgivings about using the Allen correction formula for the elimination from their readings of the absorption due to the considerable 'non-specific' brown colour developed in the Kobcr reaction with extracts from urines of low oestrogen content I shared their misgivings, but my prosent view, based on their recovery experiments, is that what may appear to some as a musiise of this colour correction is in practice justifiable. In the final stages of their work Drs Brown and Bauld carried out a number of determinations by their different methods on the same urine specimens, and with their permission I am showing you some of the results which they obtained (see Table, p.49). In a few cases the agreement between the results obtained by the two methods, is not perfect, but in the majority it is amazingly good, considering the minute amounts of oestrogen which were being determined I think that Dr Brown's results on the menstrual cycle are very interesting indeed Previous workers have reported a few studies on oestrogen excretion throughout the cycle in wluch methods of rather dubious accuracy were used, but I believe that this is the first tinic that a considerable number of different cases have been studied by means of a method of proved accuracy. I am eagerly looking forward to the time when Dr Brown is able to study oestrogen excretion in menstrual abnormalitics 48 ·GENERAL DISCUSSION SI My second point concerns thic Kober rcaction. I am very surprised to hear this coming from Edinburgh, because I thought that this reaction had been well and truly interred many ycars ago. I think it is a pity that after such precise and meticulous work in chasing ang through a tremendous flow shect with excellent recovery, one should wreck oneself on a reaction like this. May I ask-and this may have already been triedwhether in the final methylation stage, after the ocstrogens have been separated and purified, it would not be possible to use a labelled carbon, and so increase the sensitivity of detection by a factor of ten or more? If radioactive methyl sulphate is unsuitable, acetic anhydride might be used to give just that little extra sensitivity to avoid thc Kober reaction. Thirdly, I want to mention my personal, liighly encouraging experience with gradient clution, particularly in ethanol-benzene on alunina. I find that this sharpens the separations considerably. It may also help to get rid of sonic of the objectionable chronogens that can intcrfcrc in the Kobcr reaction. Dr Reiss. Halkerston and Stich 113 our laboratory have had extensive experience with extraction of the sulphatasc and glycuronidase from these molluscs, and so far have only used them for the determination of ketosteroids. The results arc very encouraging. Onc gets about 30% niore ketosteroids by hydrolysis. And I was, thereforc, very pleased when Dr Kellie mentioned that he distrusts the very intensive acid hydrolysis of urines in oestrogen determinations. So far, we have myestigated only onc uninc with these mollusc extracts, and were surprised to get far more ocstrogens after this hydrolysis than after ordinary acid hydrolysis, but that might have been just an accident. Dr Rumey: In Prof. Zondek's laboratory in Jerisalcm 2 or 3 years ago'ocstnol was deternuncd in two normal menstrual cycles by the method of Finkelstein, and the results agreed very well with the results reported here. Onc observation rather interested us: the incrcase in basal temperature occurred at a time when the ocstrogen value was the lowest for the whole cycle, one day later oestrogen increased About the question of fluorimetry and thc Kober method, I think it is significant that the method reported by Dr Swyer employs 85% phosphoric acid and not sulphuric acid. This is very important because the American workers got into serious difficulties with sulphuric acid. The reaction with phosphoric acid is much more specific, and it has very great advantages over sulphuric acid. By taking s ml of urine and employing the Hilger fluorimeter in the final determination, one can get down to oor or even 0005 rg of oestrone. May I also add one word about the question of enzymic hydrolysis I know that Prof. Marrian has strong views about this But it appears to me that to use acid and then to spend many hours removing products forned by the acid hydrolysis wastes considerable time. If an enzymıc method can be utilized it would be very encouraging. Dr Boscott I would like to champion the cause of paper chromatography, by Bush at a recent Laurentian Conference in the paper chromatograms done by us we have used five tests, each test gives different colours or different coloured Muorescences. I might add that our I.C I reagent gives colours of azo-dyes ranging through the whole as was done ·So GENERAL DISCUSSION There are a number of points in the paper by Dr Braunsberg, Miss Stern and Dr Swyer on which I would like to comment. If I understand the procedure used correctly, one of the early stages in the purification process is the extraction of the oestrogens with N-NaOH from ether solution. This process would result in quantitative extraction of oestriol, but there would be very big losses of oestrone and quite appreciable losses of oestradiol These authors must be congratulated on developing an excellent single-column method for separating and purifying the three oestrogens in urine cxtracts which, I believe, will be very valuable indccd. I also think that their work on the technical side of fluorimetry is of great fundamental importance. I noticed that Dr Swyer referred to the peak of fluorogenic material which inmediately preceded the ocstriol peak on their chromatograms as being duc to ‘non-specific fluorogenic substances. It seems to me that tlus material may well be another oestrogen or another fluorogenic substance related to the oestrogens. In this connection I should niention that several workers in the United States have noted the presence of other fluorogenic substances in urine extracts in addition to ocstriol, ocstrone and oestradiol, while both Drs Brown and Bauld have detected the presence of a fourth Kober chromogen I think at this point I should say something about the relative metliods of fluorescence and Kober methods for the final determination of the oestrogens The former method has the great advantage over the latter in being considerably more sensitive. However, I know that it is the vicw of Drs Brown and Bauld that the Kober reaction is less prone to interference by trace impurities in solvents and by ‘muck' in the urinc extracts, and for this reason they decided to use it in preference to the fluorescence method It would be valuable if we could hear the views of others on this important point. Dr Kellie First of all, I would like to add my congratulations to Dr Brown and Dr Bauld for their cellent work. Only those of us who have tried to process 24-hr urine samples over a course of three menstrual cycles know the tremendous amount of work that goes into getting results which you can present in a matter of a few seconds. There are, however, a few points I would like to raise I am not experienced in the oestrogen ficld, but I would like to champion a very lowly anımal, the marine mollusc. There are two marine molluscs which exist in abundance on the shores of England—the limpet and the periwinkle Drs Dodgson and Spencer have shown that these two molluscs are a very potent source of aryl sulphatase. By a relatively simple process one can get aryl sulphatasc activity of the order of 4 x 10Ⓡu/g. The preparation is also a very good source of B-glycuronidase, with an activity of the order of 700,000 u /g, wluch is vastly greater than one can get froin any current American source It is, moreover, very chcap With regard to the use of aryl sulphatasc, I notice that in the papers which have been read, everyone rather guiltıly says, 'Wc hydrolysed the conjugates by boiling with 15% acid for 1 hr'. I suggest that this is very violent, and perliaps unwise, as there is such an excellent source of aryl sulphatase It may well be that, by using enzyme hydrolysis instead of acid hydrolysis, by the time one gets through to apply the Kober reaction, some of the troubles will disappear; perhaps there won't be such a high blank.

·GENERAL DISCUSSION S3 illusion in thinking that enzymic hydrolysis will give us pure extracts, becausc the main source of chromogens scems to be phenolic material, which would also be formed by czyınıc hydrolysis The final point: on gradient clution. This is unquestionably going to be the answer. If and when other oestrogens arc isolated, and identified in the urinary extracts, obviously there will be a limit to how many columns you can liave It seems to us that the method which l'rof. Morris lias developed, for gradient clution on a single partition chromato gram of the three known oestrogens, might well be advanced to the point to cover any other related compounds that we might find Dr Boscort Onc of the points that I think should be looked into is the recovery of ocstrogens from urines relatively rich in protein and carbohydrate, because here we have a lot of humin formation, and one docs not know what the intermediates are in this humin formation. Dr Bauld has shown that aldclydes will mterfere in the Kober row action. Is it possible that furfural arising from the reaction of acid with carbohydrate during acid hydrolysis may possibly attack the oestrogens in thcsc dirty urines? Onc recalls the zinc and hydrochloric acid values of toxacmic pregnancies being grcatly incrcascd, compared with ordinary hydrolysis. Is this due to the destruction of compounds arising during acıd hydrolysis which would tend to destroy oestrogens, for example the sugar and protein present in toxacmia of pregnancy urines? I think these points deserve further investigation. Dr Hannelore Braunsberg First I would like to thank Prof Marrian for hiis kind reception of our competitive efforts I was rather surprised, though, to hear that normal caustic soda is not supposed to extract all the oestrogens. I would be grateful for an indication of thic sort of losses we may have incurred. Sccondly, I quite agrec that what Dr Sivyer called the 'non-specific fluorescence coming out before the oestrone and then agani before thic ocstriol is interesting and my have considerablc significance. But I am sure it is due to a mixture of a lot of components, because it comes with the very front of the solvent, and it would need further purification and separation into various compounds Could we perhaps join hands in this cffort with the Edmburgh group? We are still rather in the dark about the valıdıty of our results, and it would be very useful to us if we could do somnc fluorimetric determinations on tlie same samples as done by Dr Bauld and Dr Brown, and see whether we can get answers in the same region I am sure we will find a lot of the purification steps of the Edinburgh group very use ful—although our problem is really quite different. We eliminate perhaps diffcrent compounds, because they interfere with the fluorimetric measurement, and not so much because they produce a colour-wc are not worried about any colours, because we use such small fractions and I am not sure how many of your steps were just to remove colour. I would also like to say a word in defence of fluorimetric methods. First of all, I was rather put off the Kobcr reaction when I read all the work done by Bandow in Germany His work followed on Kober's first publication, I think, and he investigated a multitude of organic compounds of all kinds, including steroids, and found that 11 the 1930s ·GENERAL DISCUSSION $2 of the spectrum, from yellow right through to mauve. By means of these relatively specific colour and fluorescence tests I think we are pretty certain of the compounds which we have identified only by paper chromatography, but we have also isolated quite a number of compounds I described. Dr Brown: One of the primary aims in developing our method was to produce something which could be used routinely. The present method is not difficult and four of us do forty to fifty determinations a weck, and have been doing so for the last 12 months. It was partly for this reason that we preferred acid hydrolysis to enzymic hydrolysis, it is casier to fit into a routine mcthod and is free from the uncertainties of present enzymic methods. As for the fuorescence versus the Kober reaction, I spent about 9 months on the fluorescence method at the same time as I was working on the Kobcr reaction, and the reason for choosing the latter was that it was the first to yield satisfactory results with urine cxtracts. Dr Bauld: I would like to comment on some of the points raised by Dr Kellic. I think it is about time we changed the name of the Kober reaction It has been m existence now since 1931, and until 2 ycars ago the initial thing that happened when anyone worked with the Kober reaction was that they modified it. In this period the Kober reaction got a well-deserved bad name, because in cach laboratory there was an art of conducting Kober assays-only one person could do it, and so on. I have spent over a year on Dr Brown's final modification and have found that a wide variety of factors--from the make of sulphuric acid to the size of tubes used—affects it But after the changes resulting from the year's work, we are confident of the stability and reproducibility of the quinolaqueous sulphuric acid reaction As far as sensitivity is concerned, 1c is as sensitive as any other colour rcaction in the literature, though much less sensitive than the fluorescence methods However, I think that we may be confusing sensitivity of Auorescence and sensitivity of fuorimetric measurement. There seems to be no end to it you just put another valve in the photoelectric multiplier and then measure not oos but o or jg How long this can go on and how real it is I don't know The second thing I would like to say is in connection with the acid hydrolysis. We may have appeared to be a bit gulty every time we say we boil this with acid for an hour. The reason we seem gulty is not because we have not done a great deal of investigation of acid hydrolysts, both Dr Brown and I have over the last 4 years looked into possible sources of error during the acid hydrolysis. My contention is that the apparent i destruction by acid hydrolysis is not duc to destruction of the oestrogens during lıydrolysis, but to modification of the urine residues which ac later stages in the procedure bring about a destruction of oestrogens In other words, I think the main effect of variations in the acid hydrolysis is on the background material, and hence on substances which go into ether extract and can destroy oestrogens subscquently during the purification procedure Our work with the final method so far gives no clear indication of a considerable destruction of oestrogens during acıd hydrolysis. I have had little cxperience myself with enzymıc hydrolysis; we tried to avoid it, as Dr Brown said, because we wanted to have a routine method But we may be under an ·SS GENLRAL DISCUSSION I rather think that Dr Boscott misunderstood niy remarks about paper chirontography. I have nothing against this on the contrary I believe that he is one of the most valuable now laboratory techniques which have been devised in recent y cars, and wc make very full use of it in Edinburgh. However, I reaffirm my belicf that it is neverthcless not an adequate substitute for isolation and characterization by the classical nicthods, where these can be employed. I think Miss Stern mentioned that she and her co-workers hydrolyse urinc by boiling for 20 min with 15 vols., of concentrated HCl. In our cxperience, based on many cxperiments carried out over the past 8 ycars, completc hydrolysis of the conjugated oestrogens in urinc rcquires not less than 40 min boiling with this concentration of HCI, and in order to be quite safc we cxtend this numimum timc to 60 min.

·S4 GENERAL DISCUSSION a very large number gave colours with sulphuric acid, and Auorescence in some cases, but not so many. He did not investigate phosphoric acid in detail. We know from Dr Boscott's work that a lot of substances will Auoresce, and I should be grateful for an indication from Dr Boscott what pathological states are likely to give us trouble and also, from the nutritional point of view, what kinds of things our patients should avoid before their urines are collected. Miss Stern. I should like to emphasize that the fluorescence method is something like 100 times more sensitive than thc Kober, not 10. Secondly, we are a little ashamed of boiling our urines with such strong acid—for 20 min, by the way, and not for 1 hr. We have tried some very preliminary experiments with Mr Hics on some enzyme hydrolyses, which were encouraging We have also done preluninary experiments on boiling some pure oestrogens with acid, the same strength as wc usc for urinc (15 vol. HCl/100 vol. H,O) and then extracting with other, without any extraction from ether with soda, but we still find quite big losses. Prof. Mfarrian: First of all I must comment with some vigour on Dr Kellie's remarks about the Kober reaction This reaction has never been 'interred' by those who know how to carry it out properly, and it would seem to me that Dr Kellie has been musinformed There is no satisfactory evidence to indicate that the Kober rcaction is nonspecific, on the contrary the available evidence shows that if it is properly carried out it is highly specific for the naturally occurring oestrogens The question of acid hydrolysis of urine versies enzymic lıydrolysis was raised in the discussion, and I must make two comments about this. First of all I would say that we in Edinburgh are by no means ignoring the possibility that there may be destruction of oestrogens during acid hydrolysis, and I would remind you that Cohen and I in 1935 were the first workers to suggest that such a destruction might occur. However, I think the present view of Drs Brown and Bauld, which is admittedly based on insufficient evidence, is that this destruction may not be as serious as Cohen and I and others originally thought. Sccondly, m reply to the suggestion made in the discussion that I am opposed to enzymic hydrolysis, I would remind you that as long ago as 1937 I stated (Cold Spring Harbor Symposia) that in my opinion the ultimate solution to the problem of the hydrolysis of the conjugated oestrogens would be the use of enzymes However, I feel rather strongly that at the present time the methods of enzymıc hydrolysis have not been sufficiently worked out to justify their use in the routine determination of urinary oestrogens, and for that reason we think it wiser m the meantime to continue to use acid hydrolysis In the same connection Dr Bauld raised an important point which I must underline and enlarge upon In his experience the urinary substances which interfere in the Kober reaction are not so much pigments formed during acid hydrolysis but phenolic substances which give rise to pigments by oxidation in the subsequent purification of the extracts Since such phenolic substances nught well be liberated from conjugates by enzymic hydrolysis, it is possible that the latter procedure might not be so advantageous from the port of of punity of the final fractions as many people suppose. view ·OI.STROGENS IN HUMAN SLMLN $7 'conjugated oestrogens as alcoliol-solublc, but insoluble in cther and 'protein-bound' oestrogens as alcohol-insoluble and ether-insoluble. 'Conjugated' oestrogens arc liberated by acid lydrolysis, 'protein-bound' ocstrogens by alkalı hydrolysis. The cxtracts are purificd by solvent partition, using the scheme developed by Engel, Slaunwhite, Carter & Nathanson (1950). The individual oestrogens in cach fraction are separated by countercurrent distribution and estimated by fluorimetry. The fluorimetric technique employed is a one-stage procedure. Alcoholic solutions of the oestrogens are heated with 88% (v/v) sulphuric acid, cluilled, and the resulting Auoresccncc is read in a Coleman fluorimctcr, Model 12 B, using a lamp filter transmitting at 4360 and a photocell filter with a maximal transmission at 5250 A (Baird interference filter). l'articulars of the techniquc are described elsewhere (Dıczfalusy, 1953). Since the fluorescence intensitics of the individual oestrogens are different, it is necessary to run on cach occasion standardization curves with all three oestrogens. Thc separation and simultancous characterization of the individual oestrogens is achieved by a discontinuous countercurrent procedure, as suggested by Engel (1950). Essentially, the procedure of comtercurrent distribution is based on a systematization of simultancous multiple extractions, solutes being particioned between series of immiscible solvent phases in separate contacting tubes in a simple countcrcurrent order. Since essential cqulibrium is attained at cach step, the results are particularly adaptable to exact mathematical interpretation. Thic countercurrent effect obtained by a discontinuous extraction process can bc illustrated by a simple example: contaming cqual voluincs of two and c (this will be the nobilc I be the stationary plase), and start a distribution of so parts of a hypothetical substance, which is assumed to be cxactly cqually soluble in both phases. The substance is partitioned between a and A, and when cquilibriuni is attained, there are forty parts in the upper as well as in the lower phase, as indicated in Step I in Fig 2. 1 2 2 BC I IN Fig 2 Schematic representation of discontinuous countercurrent distnbution of a solute whirse partition coefficient is unity, in a system consisting of two transfers ·CHARACTERIZATION OF THE OESTROGENS IN HUMAN SEMEN BY COUNTERCURRENT DISTRIBUTION AND FLUORIMETRIC ANALYSIS By E. DICZFALUSY From the Department of Women's Diseases, Karolinska Sjukuuset, Stockholm, Sweden SUMMARY An actempt was made to characterize the oestrogen pattern of hunan senen, using solvent partition, countercurrent distribution and fluorimetric analysis, The data obtained ndicate that at least three natural oestrogens, Ocstronc, ocstradiol-176 and oestriol are present in poolcd human semen The bulk of the scminul oestrogens was found to be present in the 'free' form During the past decade sonic data liave been reported in the literature indicating that hunuan senien may possess ocstrogenic activity (Green-Arnytage, Silberstein & Wechtel, 1947; Rusfeldt, 1948, McCullagh & Schaffenburg, 1951). On the other hand, negative findings have also been recorded (Bacsich, Sharnian & Wyburn, 1945) We decided to reuvestigate this problem and therefore collected over the last 4 years approximately half a gallon of human semen. The extraction procedure used in our laboratory, which was originally worked out for the extraction of placental oestrogens (Dıczfalusy, 1953) is shown schematically in Fig. 1. Thus 'free' ocstrogens arc characterized as alcohol-soluble and ether-soluble, SCONCIE Extracted by EtOH or BuOH Extract Residue Evaporated to dryness in atmosphcre of Ng Suspended in H2O Ether-soluble Ether-insoluble (Ether-soluble following acid hydrolysis) EtOH-, BuOHinsoluble Ether-insolubic (Ether-soluble following alkal hydrolysts) *Fric Conjugated" 'Profum-bound In all fractions Purification by solvent partition Separation by countercurrent distribution Estimation by fluorimetric analysis Fig Schematic representation of the procedure employed for the extraction of oestrogens from hunian semen 56 ·16 orSTROGENS IN HUMAN SEMEN 59 JYSTEH RAIS UM CHER 1 J+3% Econ K..21 K.390 o Tube No Fiz + Complete suparation of deorycorta osterone (1)OC) from progesterone (P) in a 24-ransfer countercurrent distbution in upper layer petroleum ather In kru cr layer 345.6 (v') ethanol Kus the partition coefficient for cach compound cakulated from the experimental curses according to Williamson & Cruz (19471 For all distribu ti un curses solid line cronmental, broken line theoretical Generally speaking, a satisfactory scparation of two substances in a 24-transfer distribution can be cxpected if the ratio of the greater partition cocfficient to the lower (the separation factor) is 3 or more. (In the example shown in Fig. 4. the scparation factor is 186). In a distribution consisting of 100 transfers, a ratio of 1 2 is already sufficient to obtain a certain degree of separation. The system most often used in our laboratory for the separation of natural ocstrogens was devised by Engel et al (1950) It consists of cqual volumes of soº, methanol and carbon tetrachloride. A 25-tube all glass machme, manufactured according to the specifications of Craig & l'ost (1949) is used. At the end of the distribution the contents of cach tube arc evaporated to dryness in an atmosphicre of nitrogen and at a reduced pressurc, the residuc is taken up in cthanol and duplicate alıquots are analyzed by fluorimetry. When crystalline oestronc, oestradiol-17B and oestriol were distributed in the system described by Engel et al. (1950), the pattern shown in Fig sВ was obtained; whereas the ethan extract of 1000 ml. human semen corresponding to the 'free' estrogens gave a very similar pattern, as it would appear from Fig. SA JYUTEM 2013 Scale 1 200mg HIEM Joy Eute M. Titolo Fig s Countercurrent distribution of the 'free' oestrogens obtained in an ethanol extract of pooled human semen (Fig SA) and of crystalline oestrogens (Fig $B) In upper layer yo", niethanol, on lower layer carbon tetrachdonde OE, s Oestrone, OE, 15 oestradiol-179, and OE, 13 oestrol ·58 E DICZFALUSY The lighter phase is then shifted across the heavier onc in such a way that in the next step a is extracted by B and A by b. The distribution of the substance will be now: twenty parts in a, twenty parts in B, and similarly twenty parts in b as well as in A (step 2 in Fig. 2). The mobile phase is shufted again, so that the twenty parts contained in a are partitioned between a and C; this operation will give ten parts in cach phase B as well as contained twenty units; in this tube the distribution remains unchanged, whereas the fraction left in A (20 units) will equally distribute itself between A and < (step 3 in Fig. 2) In this way a countercurrent distribution lias been carried out in a system consisting of three tubes The binomial character of the distribution can be illustrated graphically with the amount of solute contained in cach tube as ordinate and the tube number as abscissa, as shown in Fig. 3a. LA A ZN Fig 3 Effect of increasing number of transfers on relation of buomual distribution to normal curve If countercurrent distribution of the same substance is carried out in twenty-four or more tubes instead of three, a similar pattern will be obtained, the binomial distribution approaching more and more closely a normal curve (Fig. 3b) At this point an important term must be introduced, namely the partition coefficient or distribution constant (K), defined as the ratio of the solute in the upper phase to that in the lower phase (assuming equal volumes) In the above example it was assumed that the substance was cxactly equally soluble in both phases, 1 e it had a partition coefficient of unity. However, if the substance in question is more soluble in the stationary phase than in the mobile phase, more and more solute is retained in the lower phase so that the substance will accumulate in the first tubes of the distribution. On the other hand, if al.laco the inner whose than in the lower phase, it will accumuit of the two or more solutes may be more or less completely separated in this way, proviued their partition coefficients differ enough from each other A complete separation of deoxycorticosterone from progesterone in a system described previously [Dıczfalusy, 1952) is shown in Fig. 4.

·OI STROGINS IN HUMAN SEMEN 61 A 8 Py J73TCM: See $59 1 205% ETON 1 U *** Ko so RA Tudo Me Tode na 113 7 Couttercurrcit distributes of the contents of Tulis 12-16 of rigs sA and combined (tig 7A) and of crystalline vestradowl-179 (fag 71) Urrir layer 40% ethyl acetate, 650heranc; lower Layer 20 s. cthanol OE, is vestradol-77? A B JYSTEM 160 CBAC BOSCH 29% ION 11 LO 12 fo ke 300, K-75 . 2. Tuba Na Tudela Fig 8 Countercurrent distnbution of the contents of Tubes 17-24 of Figs 5 and 6 combud (nk 8A) and of crys- Kallinc oestrol (Fig 813) in the solvent system showti in Fig 7 OC, is vestradiol-175, and 01, is Ostriol previously and run in tlic Craig machine, the distribution pattern obtained did not provide any clear-cut cvidence for the presence of detectable amounts of oestrogen. Thus, although the presence of some oestrogenic material in the acid-hydrolysed fractions could not be excluded, the concentration of conjugated'ocstrogens in the extracts investigated was assumed to be rather low. This belicf was further strengthened by the negative results obtained in a screening test, in whuch thrcequarters of the 'conjugated oestrogen fraction obtained following extraction by butanol was injected into five oestrone-prined spayed mice Since in four animals out of five negative vaginal smcars were obtained, it would appear reasonable to assume that the concentration of 'conjugated' oestrogens in the extracts investigated was very low indeed. On the other hand, when the alcohol-insoluble material was submitted to alkalı hydrolysis, extracted and partitioned as described above, the distribution patter obtained following countercutrent distribution revealed the presence of both oestrone and oestriol, although in rather small amounts (8 9 ug oestrone and 4 6 ug oestrol/1000 ml. semen). The presence of oestrogens in this fraction is in accordance with previous experiments on placental tissue and blood (Diczfalusy, 1953), in which alkalı hydrolysis ·E. DICZOALUSY 60 According to the pattern shown in Fig. SA, 1000 ml. semch would contain approximately 60 pg of ocstrone, 10 ug of oestradiol-178 and 30 pg of ocstriol. When in another expcrinient 1700 ml. semen was extracted by butanol instead of cthanol, and the 'free' oestrogens distributed in the same system, the pattern shown in Fig. 6 was obtained. 401mg 30 20 10 EZ K. 213 Ez ? K.32 0 8 15 20 24 Tube No. Fag 6 Countercurrent distribution of the 'free' oestrogens obtamed in a butanol extract of human seinen Solvent system as in FIS S OE, is oestrone, OE, 15 oestradiol-173 Fig. O seems to indicate the presence of approximately 20 peg of oestrone and probably that of some oestradiol-17B (approx 9 ug), whereas the presence of the-triol cannot be ascertained from this distribution. It is of considerable interest to note the less favourable yield of oestrogen following extraction by butanol Further evidence for identity of the material with partition coefficients K = 203 (Fig. SA) and K = 2.13 (Fig. 6) was obtained by combning the contents of tubes 12-16 from both experiments and carrying out a second countercurrent distribution in a different solvent system. As pointed out by Engel (1950) this procedure may be compared with a mixed melting point determination, if the partition coefficients of the extracted material and that of an authentic sample of the crystalline compound agree in two different solvent systems, this is very suggestive of identity of substance, whereas a discrepancy of K-values in the second solvent system—like a depression in melting point, proves that the two substances are not identical. When the combined inaterial from tubes 12-16 (Figs sA and 6) was distributed in another solvent system consisting of cthyl acetatehexine as upper plase and aqueous ethanol as lower phase, and the calculated partition coefficient was compared with that of crystalline oestradiol-17B distributed in the same system, a close agreement of partition coefficients was found, indicating that the material is, in fact, identical with oestradiol-17B (Fig. 7) Again, when the contents of tubes 17-24 from the two experiments (sce Figs. 58 and 6) were combined and this material was distributed in the system shown in Fig. 7, the calculated partition coefficient (K 0.78) was found to be very ncarly identical with that of crystalline oestriol (K = 0.75), as it would appear from Fig. 8 When, following the removal of 'free' oestrogens, the residual seminal extracts were hydrolysed in 15% (v/v) hydrochloric acid for 30 min, extracted by ether, purified as

·OESTROGENS IN HUMAN SEMEN
63
Dr Mama: It would be very interesting to repeat
these
very
beautiful cxpcrinients on
spermatozoa and seminal plasma separately.
Dr Diczfalusy: It would take years.
Dr Mfann: In man, unlike the hog, the bulk of the semen is composed of scminal plasma.
So, your oestrogen may have nothing to do with the spermatozoa; it may be all in the
accessory
fluids.
Dr Dic=falusy: I don't think so McCullagh & Schaffenburg (1951) have separated
them, and state that the oestrogenic activity is due to the spermatozoa. I have no
experimental data.
Di Mfann. It would be very important to settle that point.

·E. DICZFALUSY 62 of ethanol- and butanol-extracted material invariably libcrated a small fraction of oestrogen. Whether the presence of this 'protein-bound' oestrogen fraction depends on adsorption, incomplete extraction, solubility of ocstrogen in protein solutions, or possıbly on a chemical bond, remains to be established. Further experimental details of the work described in outline in this paper are reported elsewhere (Diczfalusy, 1954). Part of the expenses of this investigation were defrayed by a research grant from A. B. Lco, Halsingborg, Sweden. REFERENCES Bacsich, P, Sharman, A & Wyburn, G M (1945] J Obste Gynaec , Bru Cmp, 52, 334 Craig, L C & Post, H O (1949) Anal Chem 21, 500 Dıczfalusy, E. (1952) Adid endocr, Copenhagen, 10, 373 D.czfalusy, E (1053) Acta endocr , Copenhagen, 12, Suppl XI Dıczfalusy, E (1994) Acta endorr, Copenhagen, 25, 317 Engel, L L (1950) Recent Progr Hormone Res 5,339 Engel, L L, Slaunwhite, W R Jr, Carter, P & Nathanson, 1 T (1950) J bol Clicm 185, 255 Grcen-Armytage, V B. Silberstein, F & Wechtel, G E (1947) J Obstet Gynact , Brit Emp, 54, 324 McCullagh, E P & Schaffenburg, CA (1951) J x Endocrw 11, 7.403 Rusfeldt, О (1948] Acus enduer, Copenhagert, 1, 217 Williamson, B & Craig. L C (1947) ſ biol Chem 168, 687 DISCUSSION Afiss Stem: I understand that the oestrogens you find in human semen arc largely in the free form. Was there evidence of any protein-bound oestrogens? Dr Dıczfalusy It depends on the definition of the term. If you accept the term 'proteinbound' oestrogens as indicating a chemical bond, we have never detected them. We simply cannot account for this fraction, so it is denoted very tentatively as “protein-bound ocstrogens'. Aliss Stern I was thinking of protein-bound' more as Samuels et al. did in their recent paper in the biol. Chem. (1954 206, 751] and found it loosely bound They used albumen in their experinents Dr Diczfalusy It might depend on failure of extraction techniques. I think we must be very cautious on this point. Dr Hannelore Braunsberg Were you quite satisfied with the caustic soda treatment? I prestime in the extraction of protein-bound material you heat the fluid with caustic soda Dr Diczfalusy. No, we leave it at room temperature for 24 hr with s% NaOH, it is added in cqual volunies, so you get a final concentration of 2 5% NaOH. Dr Hannelore Braunsberg. Does no destruction of oestrogens occur with this treatment? Dr Dıczfalusy We tried to do the same with crystalline compounds, and found no destruction, even when heating the crystalline oestrogens, I think Dr Bauld can confirm thus. On the other hand, I have not tried oestrogens added to extracts of body fluids Dr Hannelore Braunsberg In Prof Samuels' study, which Miss Stern just mentioned, they allowed the protein-binding to take place, and then just shook up thc protein-bound steroids (not just oestrogens, but all sorts of steroids) with ether and got them all out again.

·65 OESTROGLNS IN PLACENTAC fmnced placcatae Larenta Ethanol and butanol extraction Extract Residuc Evaporated, taken up in water, cther cttraction Ethict Water Alkalınc hydrolysis Acidification with cther extraction Acid hydrolysis. Etlier extraction 'Tree' fraction Conjugaidd fraction "Protein found fraction Eder extracts, washed with NalICO,, CCI added, extracted with N-NaOH NaOH to PH 9, cther extraction Ether evaporated, residue dissolved in benzene, extracted with water Water Benzene O strone and ocetradiol fraction Ostrol fraction Fig 1. Extraction of 'free', comunated' and 'protein-bound' oestrogens from tissue, and their purification and separation into oestronc-oestradiol and cestriol fract10% not The solvent systein for oestronc and ocstradiol was made from cqual volumes of light petroleum (boiling point range 100-120') and methanol, and for oestrol, 2 vol. of benzene to i vol, of methanol and 1 vol. of water. The solvent systems have been so arranged that the runs nray take place overnight, the Rf values in the respective systenis arc ocstrone, 0.2; ocstradiol-178, 007; and ocstriol, o 13 As a sınıple and sensitive method for the detection of the oestrogens on paper available, Folin and Ciocalteu's reagent (Folin & Ciocalteu, 1927) was investigated and found to be ideal for this purpose, this is a reagent which is used in routine biochemistry for the estimation of phenol The paper was lightly sprayed with the reagent diluted with water, and then placed in a tank, the atmosphere of which was saturated with ammonia vapour. The ocstrogens show blue spots on a pale blue background with a detection linuit of s ug/sq cm. This reaction may also be carried out quantitatively in solution for the colorimetric estimation of pure oestrogens, but as the colour is produced with all phenols, it is not suitable for use with extracts, in which phenolicimpurities are invariably present. The amount of the oestrogen in the spots was estimated by fluorimetry and, as a prior spraying with Folin & Ciocalteu's reagent would interfere, the positions of the spots were found by means of parallel runs with smaller announts of similar extract and of pure oestrogens The smallest arca of paper ensuring removal of all oestrogen in the up as ·THE ESTIMATION, RECOVERY AND NORMAL LEVELS OF OESTRIOL, OESTRONE AND OESTRADIOL IN HUMAN FULL-TERM PLACENTAE By F L. MITCHELL From the Jessop Hospital for Il'omien, Sheffield 3, and the Aledical Research Cou'ral Umt for Rescarci in Cell Metabolismi, The University, Sheffield 10 While much work has been carried out in recent years on the clinical estimation of Ocstrone, ocstradiol and oestriol in urinc, little work has been reported on the estimation and detection of the oestrogens in tissue. Dıczfalusy early last year (Dıczfalusy, 1953) reported the use of tite countercurrent method of Engel, Slaunwhite, Carter & Nathanson (1950) for the separation and purification of endogenous ocstrogens in placental tissuz, but as recently as 1952 Goldzieher & Roberts (1952) had to use over 8 kg of tissue and rely on only a chemical separation to identify the oestrogen in human testes. In an attempt to achieve the high degree of purification necessary for a fioul estimation by fluorimetry, a total of six chromatographic methods of separation and one chemical method of purification were tried before the technique of paper chromatography at elevated tenıperature described by Bush (1982) for adrenal steroids was investigat:d and found to give satisfactory results. By this means i Mg of each oestrogen may Le detected and estimated in up to 200 g of placental tissue. The tissue was minced and mixed with ethanol within 30 min of delivery; the oestrogens were extracted, separated into 'free', 'conjugated and protein-bound' extraces by a method simular to that used by Dıczfalusy (1953), and each extract was purified and separated into oestronc-oestradiol and oestriol fractions by the method shown in Fig. 1. This separation was found necessary because the high polarity of oestriol compared with oestrone and oestradiol entails the use of two separate solvent systems for chromatography. If greasy material was still present in an amount which interfered with chromatography, 1 e. more than approximately 1 mg per run, this was removed by washing a solution of the extract in 90% (v/v) methanol with 11-pentane [Ryan & Engel, 1953). All solvents were rigorously purified before use and operations above rooni temporature were carried out under nitrogen. Extracts were stored at -20°C when not aotually in usc The final fractions were taken up in small amounts of ethanol for transferring to the 18 in. strips of Whatman No. 541 filter paper Extracts from tissue up to 100 g may be run as one spot The chromatography was carried out by the descending method at a temperature of 32°, the temperature throughout the cabinet varying by not more than 0.5°. It was cssential that the inner walls of the tanks should be covered with wads of filter paper dipping into the mobile and stationary phases to chsure saturation of the atmosphere with solvent vapour. 64 ·OESTROGENS IN PLACENTAL 67 Table 1. Comparison of the amounts of oestrogen in paper chromatogram spots when assayed by anons methods Frec oestrogen (ug) per 400 g placenta Method of Assay Oestriol Oestronc Oestradiol Visual comparison of spots 10 UV absorption with approx. corr. for impunity 25 10 1 4 Biological 17 Tracc Fluorimetry 40.5 13.8 5.0 To arrive at the truc level of the oestrogens in the samples of placental tissue it was necessary to complete several large-scale recovery cxpcrinicnts. These experiments have only been carried outon thc'free' fraction. Amounts of soo ugand 30 pgofcachocstrogen were added to two similar batches of placental mince, cach batch weighing approximately 200 g. The recoveries were simular for the large and the small amounts of ocstrogen, and averages are shown in Table 2. Heating the tissue to 100° zn nitrogen before adding the oestrogen increased the recovery in all cascs, whilst hcating in air increased it further. Table 2. The recorcry of free' oestriol, cestrone and oestradiol adiled to full term luman placental tissue (The pure oestrogen in ethanolic solumen tras added to the mince immediately after the addition of 80% ethanol for extraction) Expt No of Treatment of Recovery of added 'free' oestrogen placentac placentae before pooled Oestriol Oestronc Oestradiol (6) (7 No cxtraction (",) Nonc 21.0 17.0 11.7 I 3 Heated 25 min at 100° in N, 52.0 24.9 19: 2 None 43-4 172 IS Nonc 18.5 45 ISO 5 2 Heated 25 min at 100° in air 990 130 700 26 7 13 5 13 7 Aver. recovery from untreated placentae Table 3 gives the experimental values found for the 'frce' oestrogens in different placentac compared with the estimated true values after correction for the low recoveries. It will be seen that though heating the mince in air causes increased recovery (Table 3), it results in the loss of a large amount of the original oestrogen Heating in nitrogen gives a smaller loss of endogenous oestrogen, which is counterbalanced by an uncreased recovery, to give an experimental value similar to that obtained for untreated placenta. The addition of thiourea to act as an antioxidant did not improve the yield.

·66 F, L. MITCHELL main runs was then cut out, together with similar areas from a blank part of the chromatogram, and standard amounts of oestrogen were placed on these blank papers. For quantitative work, the chromatography paper used was extracted with methanol for 48 hr, but, because at the end of this time methanol-soluble impurities were still extractable from the paper, each spot was extracted with the same amount of solvent for a standard time, and a special apparatus was constructed which dripped s ml. methanol over each paper for a period of 8 min. This ensured the extraction of all the oestrogen and the same aniount of paper impurity in both tests and standards. Fluorimetric assays were carried out by the method of Diczfalusy (1953). To prove the identity of the spots, extracts from 160 g of placental tissue were run on the chromatograms and the spots extracted. The ultraviolet absorption spectra were compared with those for purc oestrogens (Fig. 2), and a portion of the extracts were coupled with Fast Black Salt K followed by chromatography (Heftmann, 1950). In cach case well defined spots were produced, identical in all ways with spots produced by parallel runs of pure oestrogens. The spots have been assayed by four separate methods (including bioassay) with good agreement of results (Table 1). Recovery experiments with the three oestrogens added to the tissue mince showed that during the extraction and purification process, despite the precautions taken to purify all solvents, to reduce oxidation, and to store at a low temperature, a considerable loss of oestrogen took place. Very little loss was experienced during the chromatography and elution from the chromatogram, for when the oestrogens were added to the extract placed on the paper they were recovered as follows; oestriol 97%, oestrone 98%, and oestradiol 84%. These results also indicate that little quenching of fluorescence takes place by any tissue impurity present in the spots. OESTRONE 0.4 0.3 DESTRIOL E 02- 30vg/ml PURE OESTRONE ot OESTRADIOL 240 250 260 270 280 290 300 WAVELENGTH (my) Fig 2 Absorption spectra of oestrogen spots from 160 g placental tissue, eluted from the chromatograms and compared with pure oestrone Absorptions measured in ethanol against extraction from similar areas of blank chromatogranı paper ·Expt No of No. placentac pooled Table 4. Experincomal values for the conjugated and protein bound'ocstrogens found in placentae Treatincnt of 'Conjugated' oestrogens isolatců "Protein-bound' oestrogens isolated placentae before (wg/kg) (sg/kg) extractio!) Ocstriol Oestronc Ocstradiol Oestriol Ocstronc Oestradiol None 65 25 O 33 16 0 I 3 Ileated 25 min at 100° in N, 68 23 O 31 16 0 OESTROGENS IN PLACENTAL 3 I Nonc 70 34 27 27 None o 8 17 o 3 10 S Heated 25 min at 100° in air o 8 8 a 1 17 14 21 8 21 14 7 Aver content of untreated placentae Content as found by Dıczfalusy (1953) 31.+ 2.5 1.5 IOS O 69 ·68 Expt. Table 3 Experimental values for the 'free' oestrogens fomd in placentae, with values corrected by recovery experiments (cf. Table 2) No of Treatment of 'Free oestrogen isolated *Free' oestrogen corrected No, placentae placentae before (ug/kg) pooled (ug/kg) extraction Oestriol Oestrone Oestradiol Oestriol Oestrone Oestradiol None 157 108 748 635 O 3 Heated 25 min at 100° in N, 168 III O 317 317 2 None 220 109 13 $22 651 89 T. L, MITCHELL 4 (Heated 25 min at 100° in Ny+thioutca to o 1% Heated 25 mm at 100" an N, 2 288 275 194 194 IO 8 None 148 20 5 76 2 730 440 507 Heated 25 min at 100° in air 92 19 57 93 1.46 82 190 86 Aver. content of untreated placentae Content as found by Dıczfalusy (1953) 25 678 $84 170 1254 46.7 3.1 ·OESTROGENS IN PLACENTAE 71 experiment ocstronc gaso very poor recoveries. I just wondered where we should find dhe cause of the discrepancy. Dr Jetchell: I may be wrong, but wasn't Dr Engel working with enzynies? Dr Diczfaltesy: No, it was liver slices incubated. Delfitchell. But here we hope the enzymes are all destroyed by the addition of alcoliol before addition of the oestrogens. Also, the heating would take care of all the enzymes włuch were not destroyed by alcohol. We originally licated the tissues in the hope of removing proteins which interfered with partition coefficients by producing emulsions; the heating docs this to a large cxtent. It also has thus side-effect, which we thought was very interesting. Dr Diczfalusy: Would you be interested in a joint effort to reinvestigate this problem? I have stored some early placenta (a few kg) in deep freeze at -20°. I could send you some and we could try to carry out simultancous experiments Dr Mitchell: Certainly; that would be very interesting. It is onc of the drawbacks: you have worked on carly placenta, and wc have used full-term placenta. Dr Hamelore Braunsberg. I am interested in your comment on the matemal blood oestrogens How did you try to measure them, and were you able to detect anything at all? Dr Alitchell: We did them virtually by my method for placental mince. In cach casc wc got out spots which had the same Rf values as the three oestrogens, but we did not do any biological assays to characterize them. Dr Hannelore Braunsberg: flow did you show them up? With the Folin-Ciocaltcu? Dr Afrtchell: Yes. Dr Hannelore Braunsberg: That does not detect smaller amounts than about s ug, does it, after running? Dr Mitchell. No, that is truc. For these runs we uscd large amounts for detection For the other assays, although we cannot detect the spots by the Folin-Ciocaltcu rcagent, by using parallel runs, you cut out where you hope thic ocstrogens are, and you can estimate down to about 0.1 ug fluonmetrically, even though you cannot sce it. Dr Romncy: The mctabolism of natural oestrogens in vitro by cell-frce enzymatic extracts of rat and human liver has been recently studıcd by us The solvent system benzene/methanol/water was found convenient for the separation of all three natural oestrogens by paper chromatography, using the Folin-Ciocalteu reagent for detection, and fluorimetry with phosphoric acid for quantitative estimation Both oestrone and oestradiol were found to be partly interconverted, but neither gave rise to oestrol with either rat or human liver extract. A phenolic metabolite more polar than ocstriol was detected, and we are trying to identify this. Recently, we have also obtained the spleen from a case of Gaucher's disease. After incubation of oestronc with a homogenate of this tissue, no phenolic metabolite could be detected on the chromatogram Recovery of oestrone after incubation with boiled assue was 85% Dr Bauld. I would like to suggest that Fe+++ is the factor in Dr Mitchell's experiments which is able to destroy oestrogens in the placenta during the subsequent punfication process. In many of our experiments in Edinburgh traces of iron added to the water for our blank extractions, even at room temperature, caused massive destruction of ·70 F. L, MITCHELL The corrected amounts of 'free' oestrogen are between four and sixteen times more than the uncorrected values obtained by Diczfalusy (1953) which are similar to those obtained by Huffman, Thayer & Doisy (1940), except that the latter reported 38 ug of oestradiol/kg. The uncorrected values for the conjugated' and 'protein-bound' fractions (Table 4) are also somewhat higher than those found by Dıczfalusy (1953). To ascertain how much of the oestrogen in the tissue was due to the small amount of retained blood, the placental, maternal and foetal blood content was measured, and estimations were carried out on maternal blood. Though the sensitivity of the method was not sufficient to give accurate results on the 10 ml. volumes available, it was suffcient to show that the blood content of the placental tissue could not account for more than s ug/kg of each oestrogen (uncorrected), assuming that the foctal blood oestrogen is not liglier tliar the level in miternal blood Diczfalusy has carried out recovery experiments only on early placentae (3-5 lunar months) and has obtained recoveries of 40 to 90%. The results shown in Table 2 seem to indicate that full term placentae contain a factor which at some stage in the extraction and purification process is capable of destroying the oestrogens. From the recoveries obtained by Dıczfalusy (1953) thus factor does not appear to be present to the same extent in early placentae. It appears that the factor itself is partly destroyed by heating the tissue in air prior to extraction, though this procedure results in a large proportion of the ocstrogens themselves also being destroyed. To summarize briefly: a method has been developed for the estimation of the oestrogens in placental tissue by means of a preliminary extraction and chemical separation, followed by paper chromatography and fluorimetric assay. Recovery experiments using thus mode of assay show that the normal methods of extraction and purification of the 'free' oestrogens when applied to placental tissue give average losses of from 73 to 86%. Experimental details of the work described in outline in this paper have been reported clsewhere {Mitchell & Davies, 1954). REFERENCES Bush, 1 E (1952) Biochem } 50, 370 Dıczfalusy, E [1953] Actu endocr , Copenhagen, Suppl 12 Engel, L L, Slaunw lute, W R Jr, Carter, P & Nathanson, I E (1950) J bol Chem 285, 255 Folin, O & Ciocalteu, V (1927) J bol Chem 73, 627 Goldzıeher, J W & Roberts, I S (1952] ) thu Endocrin 12, 143 Hefmann, E (1950) Science, III, 571 Huffinan, M N Thayer, S A & Doisy, E A (1940) J bul Chem 133, 567 Mitchell, F L & Davies, R E (1954) Biochem }, 56, 6yo Ryan, K J & Engel, L L (1953) Endocrinology. 52, 277 DISCUSSION Dr Diczfalusy I am worried about the bad recoveries you get in full-term placenta Of course, Dr Engel noticed a very rapid inactivation or destruction or transformation of oestradiol when it was added to full-term placenta, and there was much less in early placenta. So that might be the explanation, but I am not quite convinced In our cxperience the best recoveries were invariably obtained with oestronc, whereas in your ·THE BIOLOGICAL ASSAY OF OESTRONE OESTRADIOL IN URINE By PHYLLIS M HARLOW, MARJORIE SANDINORD and MARY E. WOOD From the Department of Chemical Parhology, King's College Hospital, Denmark Hill, London, S.r. 5 SUMMARY 4 Preliminary investigations of both gonadotrophin and oestrone oestradiol cxcretion were ofreu dugnostic of either pituitary or ovanan failure, but single determinations of oestronc oestradiol were scldom of valus 5 Oestronc oestradiol excrction in normal prepubertal girls was well below normal menstrual cycle valucs Seven out of eight girls with precocious puberty did not show raised excretion of ocstronc oestradiol. In the cighth patient, who menstruated regularly, both gonadotrophin and Oestronc exstradiol were within the normal adult fange. 6 lixcreased amounts of oestrone ocstradiol scrc crcrcted by a patient with an adrenal tumour Six years ago, when this work began, thic use of ocstrogen cxcrction as an indication of ovarian and adrenocortical activity had been rcalısed, but had been little investigated. Chemical methods were not then capable of measuring the very small amounts of ocstrogens present in normal urinc, although pregnancy urines could be more satisfactorily analysed. It was thercforc necessary to use a biological method for measuring ocstrogens, thus, although laborious, was capable of detecting the fluctuations occurrmg in the normal menstrual cycle and of detecting the amounts found in the urine of children. The Allen-Doisy vaginal smcar technique is usually employed but has the disadvantage that animals respond differently to different oestrogens when these are injected subcutancously. Emmens (1941, 1950) has described an intra-vaginal method of admınıstration, but Biggers (1951) has shown that although oestrone, ocstriol and oestradiol may have a common slope and potency under certain conditions of intra-vaginal administration, the common slope of the dose response lene so obtained is too low for the method to be applied to practical problems A two-injection Allen-Doisy technique, as described by Emmens (1939) was used for the work described below, which took place over a period of 6 years. Oestrogens werc estimated in the urine of a series of normal subjects for comparison with urinary oestrogen values found in patients with various endocrine disorders. METHOD Specimens of urine were collected by each patient without preservative every 24 hr Acid hydrolysis was followed by extraction with ether. Oestrone and oestradiol were ·F. L. MITCHELL 72 ocstrogens, and there is a possibility that you are dealing with iron released by haenuolysis of maternal blood within the placenta This could be easily checked by seeing whether chelating agents do or do not improve your yields. Dr Mitchell. If it was from haemoglobin, Veldhuis would not have obtained his excellent recoveries of oestrogens. Dr Banld Veldhuis did his recoveries on plasma, not on wholc blood. Dr Boscott: I should like to congratulate Dr Mitchell on his very excellent paper, especially in view of the very troublesome compounds one finds in tissues such as placenta. Could he tell us what his experiences are in relation to interference from phospholipids, which I have found to be very troublesome indeed in other methods of steroid extraction. Do you get any interference from phospholipids, emulsions, etc? Dr Mitchell: We did. We have tried several methods of purifying the ocstrogens, and the manoeuvre which gave the best removal of phospholipids and the least trouble with emulsions was extracting the oestrogens from ether-CCl, by means of N-NaOH. Dr Boscott: Do you not get any fatty acids carried through? Dr Afitchell: Well, if they are carried through, there is no trouble with emulsions later on. Dr Boscott. You do not get emulsions at the ether-CCl, extraction stage? Dr Mitchell: No, with this technique there is no emulsion trouble anywhere. We had considerable trouble for 2 years with phospholipids, and it was the main reason for discarding the six chromatographic techniques which we tried. Eventually we found paper chromatography was the only answer. Of course, Do Diczfalusy gets good results, using countercurrent distribution, Dr Rumney: With the Folin-Ciocalteu reagent I tlunk we can detect 1-2 ug. Dr Mitchell We reckoned 0.5 mg/cm² or so; on a spot of perhaps 4 cmit would be about 2 g ·1 BIOLOGICAL ASSAY OF URINARY OSTROGENS 75 ual cycles of five young Two peaks of oestronc ic end of the cycle at a vaciable time before menstruation During a period values fell to less than 4 mg/24 hr. The range found in these cycles was 25-35 kg/24 hr. In some cycles the approximate time of ovulation was found by pregnancdiol determination (a) 25 OESTRONE OESTRADIOL »9/24hr 20 t5 10 5 LOWER NORMAL Limit DAYS OF CYCLE 10 15 20 25 2 PREGNANEDIOL mg/24h 1- 10 15 20 25 8 DAYS OF CYCLE (b) 20 OESTRONE OESTRADIOL P9/24hr 10 5 10 20 30 35 40 DAYS OF CYCLE 2- PREGNANEDIOL /24hr to is 20 25 75 30 35 40 4 DAYS OF CYCLE fog i Oestronc oestradiol and pregnanediol excretion in the moustrual cycle of a rormal woman, age 24 (8) 27-day cycle, (b) 41-day cycle The excretion of oestrone oestradiol and pregnanediol during two cycles from one woman are shown in Fig. 1, a and b Peaks in oestronc ocstradiol excretion occurred on the IIth and sth days before menstruation, while pregnanediol was maximal at the second peak on both occasions In the second cycle, which extended to 41 days, there appeared to be a delay in ovulation.

·74 PHYLLIS M. HARLOW, MARJORIE SANDIFORD AND MARY E. WOOD separated by a method previously described by Edwards, Gray & Wood (1990). The oestriol fraction was removed during the process. Ovariectomized albino micc were used for the assay. Each mouse was given two subcutaneous injections of nut oil containing a suitable aliquot of the oestrone oestradiol cxtract. Vaginal smears were taken three times to cover the period of oestrous response Oestrous and pro-oestrous changes were classed as positive. Standard doses of oestrone were included with each batch of unknowns in order to check the sensitivity of the mice. Slope of the dose response line Slopes of the dose response lines for pure oestrone, oestradiol and oestriol solutions were calculated from the weighted means of data obtained over a long period. The slopes for oestrone and oestradiol were satisfactory, but that for ocstriol was too low to permit accurate estimation A common slope for all three compounds was found to be invalid, but there was no significant difference between the slopes of oestrone and oestradiol (P = 0.8). Oestrone and oestradiol were therefore estimated together and tlic Oestriol fraction discarded. No significant difference was found between the slope of urinary oestrone oestradiol extracts and that of pure oestrone. Table 1. The mean slope of dose response lines Mean slope + S.E Oestrone 461 + 0.297 Oestradiol 4.38 + 0.796 Oestriol 1.87 + 0.236 x" Test Oestronc-oestradiol slopes x'=013, P= 8 Oestrone-ocstriol slopes XY S4, P < 001 For practical reasons the number of mice which could be used for cach estimation was limited, only fifteen or twenty mice were used for each dose level Two dose levels were given of each standard and unknown. The theoretical limits of error using these numbers were 69-144% and 73-137% respectively. Urinary pregnanediol was measured by the method of Sommerville, Gough & Marrian (1948). Urinary gonadotrophın was extracted by the rannic acid precipitation method described by Levin & Tyndale (1936). Later the kaolin adsorption method of Dekanski (1949) was substituted as the final extracts were less toxic. Gonadotrophın was assayed by a method depending on the increase in uterine weight produced in immature female rats. Later a more sensitive mouse method was used. RESULTS Normal menstrual cycles The excretion of oestronc oestradiol during the normal menstrual cycle was investi- ·BIOLOGICAL ASSAY OF URINARY OESTROGENS 77 Table 2 Primary amenorrhoea Oestrone ocstradiol g/24 h Gonadotrophin muu/24 hr* Condition Age 29 1.0 >06 Turner's syndrome Turer's syndrome 13 46 >73 Diabetic dwarf 20 19 3.1 <20 Alild hypothyroidism No rclated complaint 20 38 48 Detected once 181 .8-28 Mouse utetine units expected the oestronc oestradiol excretion was low and the gonadotrophun above normal limits. These patients have vestigial ovaries only and ocstrogens were produced presumably by the adrenal cortex. The two following cases showed a low cxcrction of both oestrone oestradiol and gonadotrophin. The duabetic girl, who showed slight pituitary dwarfism, subsequently took her own life. Autopsy showed that the ovaries were of normal size and the pituitary Was macroscopically normal, the adrenal cortex was atrophıcd. Insufficient production her only penods had been scanty and occurred at 6 months and a y car before the investigations. With continued thyroid treatment, however, her general condition improved and her periods became normal The delay might have been due to late production of gonadotrophin, apparently associated with thyroid deficiency In the last case, amenorrhoea was the only complaint. Oestrogen therapy had induced bleeding on the first occasion, but only a slight show was produced in subsequent months of treatment, thus did not persist after treatment ended Gonadotrophın and oestrone oestradiol excretion, measured some months after treatment ended, were within normal limits. Her uterus had remained small and was apparently unresponsive to oestrogens. Secondary amenorrhoea and oligomenorrhoca Erght patients with secondary amenorrhoca were investigated In five, oestrone oestradiol estimations were made over a month or more; one investigation only was made on each of the remaining threc. In the first case, a woman aged 32, periods had been regular until she was 2o, then at 3-6 monthly intervals until the last 2 yr, when they had ceased altogether. Values for both oestrone oestradiol and gonadotrophin excretion (Fig. 3a) were within normal limits for menstrual cycles, but neither showed any significant fluctuations, such as are ·76 PHYLLIS M. HARLOW, MARJORIE SANDIFORD AND MARY E, WOOD Oestrone oestradiol excretion in amenorrhoca In many cases of amenorrhoea the normal relationship between pituitary and ovarian hormones las esther been disturbed or has never been cstablished owing to primary insufficiency of one or more of the hormones concerned. Low excretion of both oestronc oestradiol and gonadotrophın may be due to pituitary lıypofunction, whereas low ocstrone oestradiol and excessive gonadotrophin indicates ovarian hypofunction. In other cases hormone excrction is normal, but the uterine mucosa may be insensitive to oestrogens and bleeding may never be produced. Primary amenorrhoea Urinary oestrone ocstradiol and gonadotrophın excretions were determined at intervals of 3 or 4 days for a month in two cases of primary amenorrhoea and values for both were consistently low. The results in one of these cases is shown in Fig. 2. There was no evidence of cyclical variation and such differences as do occur may be within the limits of experimental error. The investigations were made when the patient was 19 She had developed normally with the appearance of pubic hair and small breast development at 14, but her only period had been induced by a short course of stilboestrol which was not, however, taken regularly. In this case the amenorrhoea scems to be due to insufficient production of gonadotrophin, normal menstrual cycle values lying between S and 70 mouse uterine units (m.u.u.)/24 hr. OESTRONE CESTRADIOL 49/24h 10- SH LOWER NORMAL LIMIT 10 20 25 30 DAYS GONADOTROPHINS 10 muu/24 hr LOWER NORMAL LIMIT DAYS 10 15 20 25 30 Fig 2 Oestrone oestradiol and gonadutroplua excretion in a girl with primary amenorthoca, age 19 ..- .... 1 nvcrcted were measurable on one occasion. Large cyclical doses of otsitogens du vil giveu, vul uleeding was produced on only two Occasions The uterus remained infantile and appeared to be unresponsive to oestrogens. Pregnanediol was detected on two occasions, but was not investigated over the month Amenorrhoea was apparently due to insufficient oestrogen production and an insensitīve uterus It is sometimes possible to differentiate between such abnormalities by a single determination of ocstrone oestradiol and gonadotrophin excretion (Table 2). The first two subjects in Table 2 were examples of Turner's syndrome. As would be ·BIOLOGICAL ASSAY OR URINARY OSTROGENS 79 this treatinent her oestrone nestradiol excrction was less than 2.5 kg/24 hr on 4 consccutive estimations, although pregnancdiol was found. She later became pregnant without further treatment. Of the remaming two, one was dubotic and the other, after having amenorrhoca for 9 years, slon cd 110 response to progesterone or to ctlunyl oestradiol and ethisteronc. Abnormalities in the cxcrction of oestronc oestradiol were found in all these women with secondary amenorrhoea. In some the amounts caceted were very low, although not always accompanied by abnormal gonadotrophin cxcrction, in others normal quantities of oestronc ocstradiol wcrc cxcreted, but with no cyclical variations and no maintained peaks of cxcrction. In two cases oestrogen therapy made pregnancy possible, although regular periods could not be maintained on ceasing treatment. Ingular perods and sterility Tuo patients with irregular periods and sterility were investigated. The oestronc oestradiol and gonadotrophin cxcrction of the first of these is shown in Fig. 3b. This woman, aged 28, liad periods at irregular intervals They had occurred 12 days hcforc and 8 days after these investigations. Normal quantities of both gonadotrophın and ocstrogens were found. Gonadotrophın increased during the first part of the cycle and decreased in the second part, whereas values for oestronc ocstradiol were very low during the first 26 days and then showed a sudden risc with a peak about 15 days before the next period Therefore, apart from a delay in oestrogen production, the first part of the cycle appeared normal Unfortunately the last part, when a further risc in oestronc oestradiol would have been expected, was not investigated. The time of ovulation was not known, but it could well have occurred at the first oestronc oestradiol peak. The patient was later found to have a congenital abnormality of the uterus. In the second case, a womın aged 20, oestronc ocstradiol cxcrction was below normal during the first part of the cycle, but a pronounced increase occurred in the latter half. Gonadotrophin cxcrction was consistently subnormal cxcept at one point 4 days before the rise in ocstronc oestradiolcxcrction. Thus patient had a tubal insufflation immediately following the investigations and became pregnant shortly afterwards Therefore, although her excrction of both gonadotroplun and ocstronc ocstradiol differed from the normal pattern, the amounts produced were sufficient for pregnancy to occur Ocstrone oestradiol excretion 111 normal cinldren Table 3 Normal duldre Girls Oestronc oestradiol vg/24 hır Boys Ocstrone oestradiol Age Age kg/24 h 1.4 S 8 4 4 4 S 8 IO .9 14 9 1+ II 35 24 II 8 .6 7 II 11 13 ·78 PILYLIIS M. IIARIOW, MARJORIE SANDII ORD AND MARY L, WOOD OESTRONE OESTADIOL 10 ug/24hr 5 HA LOWER NORMAL LIMIT 20 25 30 35 DAYS 5 10 15 GONADOTROPKINS muu/24 hr 20 10 10 15 20 25 30 35 DAYS (6) OESTRONE O:STRADIOL 20 ug/24 hr 15 - 10- 5- LOWER NORMAL --- LIMIT 15 20 25 30 35 40 DAYS OF CYCLE GONADOTROPHINS 20-muw/24h 10- FIL 15 20 25 30 35 40 DAYS OF CYCLE Fig 3 Oestrone oestradiol and gonadotrophu excretion in two women with (a)s.condary amenorrhoea and sterility. age 32, (b) irregular periods and sterility, age 28 found in normal menstrual cycles. Her uterus and ovaries were both small, but hormones were obviously being produced by both ovarics and pituitary. Her failure to menstruate is associated with imbalance rather than msufficiency of hormones A similar cxcretion of ocstrone oestradiol was found in a girl of 19, who had had only 6 periods, all in the previous 2 yr Again no marked oestrone oestradiol excretion pcaks were found and all the values lay between 4 and 12.1 ug/24 hr The gonadotrophın excretion was measured by the rat niethod which is less sensitive and accurate than the mouse method. No pregnancdiol was found Like the first case she had very irregular periods, although they have since become more frequent There was both failure in ovulation and in regular production of ocstrogens. In the remaining six patients with secondary amenorrhoca, oestrone oestradiol excretion was below normal Three of these were probably prematurely nenopausal, as gonadotrophin values were abnormally lugh. One gul of 25 had had no periods for 3 yr, except those induced by oestrogen therapy. At a tinc when she was not receiving ·81 BIOLOGICAL ASSAY OF URINARY OLSTROGENS was not possible to obtain enough specuinens to see if thic amounts varicd as they do in the normal adult cycle, but they came within the adult rauge. Regular bleeding also occurred in another girl of 2, but the values were not abnormal for her age. Laparotomy at winch biopsics of both ovaries were taken showed that one contained a follicular cyst and the other a corpus lutcum. A third case was receiving dccp irradiation of the printary to inhibit thc gonadotropluc stimulus. Her Oestronc oestradiol excretion was normal for her age. These results suggest that in most cases of precocious puberty ocstronc oestradiol nicasurement is of little valuc as the amount crcrcted is not abnormal. There arc, lowcvcr, rarc cases of granulosa cell tumour of thic ovaries wherc a high ocstronc ocstradiol output would be dugnostic, but we did not have the opportunity to investigate such a casc. This nicthod of ocstronc oestradiol dctcrnimation becomes very inaccurate when the small amounts cxcreted by children are measured; nevertheless tliese amounts are considerablc if their small body weight is taken into account. Such quantities might be sufficient in combination with other factors to cause the changes found m precocious puberty. Oestrone ocstradiol excretion in adrenal carcinoma Carcinoma of the adrenal cortex is sometimes accompanied by a high output of oestrogens. We were able to obtam specimens from a woman aged 40 before and after removal of an adrenal tumour. The first estimation before the operation gave the abnormally higlı valuc of s2 ug in 24 hr. The second estimation, 2 months after the operation, showed a reduced valuc of 7.44g in 24 hr, which is well within normal menstrual cycle limits. This suggests that the tịnour had caused the high ocstrogen output. Tables Oistronie oestradiol action at patients with malignant tumours Ocstrone Condition Time of specimen oestradiol g/24 h Adrenal carcinoma Preoperative i month Postoperative 2 months $2.0 74 Granulosa cell tumour of the ovary Postoperative 4 months Postoperative 5 months 300 236 Oestrone oestradiol cxcretion in granulosa cell tumour of the ovary In the only case we investigated it had not been possible to obtain specimens before operation, when a large tumour, afterwards found to be a granulosa cell tumour of the ovary, was removed. Postoperative estimations were made to check the complete removal of the tumour. Values were high in specimens taken 4 and s months after operation, although the patient, aged 49, was menopausal, but there has been no recurrence of symptoms after 2 years.

·of 24 80 PHYLLIS M. HARLOW, MARJORIE SANDIFORD AND MARY E. WOOD The oestrone oestradiol excretion of normal children was measured using the whole hr urine specimens. These results are shown in Table 3 Apart from the two eldest girls, aged 11 and 13, who were probably approaching puberty, all the values for girls lay between 0.9 and 1.4 ug/24 hr which is well below the limit for the normal menstrual cycle. The oestrone Oestradiol excretion in prepubertal boys was not significantly different, the range being 06-2 4 ug/24 hr. Oestrone oestradiol excretion in children with precocious puberty The oestrone oestradiol excretion of several girls with precocious puberty was measured Two of these menstruated regularly and vaginal bleeding had occurred in some of the others; all showed breast enlargement and adult female growth of pubic hair The results are shown in Table 4. Table 4. Precocious puberty in girls Oestrons oestradiol Comments Time of Specimen Age Hg/24 hr 28th day sth day 17th day 3.5 2.4 so Menstruating regularly, gonadotroplın detected on 22nd, zıth, 27th days of consecutive cycles 24 $ 3.SI 17. 7.57 14 o 56 Recurrent uterine bleeding. Advanced physical development, follicular cyst in one ovary, corps luteum in the other 4 6 11.50 20. 7.51 1.6 06 Brcast enlargement and pubic hair No menstruatron, receiving deep irradiation of pituitary Vaginal bleeding. breast development Nipple pigmentation 22nd day 044 6 9 2 SI 0 58 Enlarged breasts and vaginal bleeding 2 ycars earlier. Monthly discliarge since and onset of 'Petit Mal' Precocious menstruation and sexual precocicy 7 7. 8.52 IL 8.52 04 06 26 9 52 30. 9.52 I 3 07 Advanced physical development, axıllary and pubic hair, brcast development, no gonadotrophin detected 6 07 07 31.10.52 Slight breast enlargement, advanced physical I.11.52 development In all cases except one the values found were either within or below the range found for normal children of the same age. The first case, however, showed high values for both gonadotrophin and oestrone oestradiol, each of which was measured three times during two consecutive cycles. The child menstruated regularly, although only 21. It ·BIOLOGICAL ASSAY OT URINARY OESTROGENS 83 for this work. We are most grateful to Mr Stanley Clayton, Dr S. Leonard Simpson, Dr Raymond Grocne and other clinicians whose patients uc luive investigated Tlianks are also due to Messrs Organon us ho love generously supplied us with purc ocstrogens. RETERENCES Diggers.) D (1951) J Candestin. 7, 163 Dekanski, J (1949) Brit j ex pul 30.272 Elands, AI S. Gry, C'HA Wood, M1 C. (1950) J Tam 6.440 Emmens, C W. (1939) Spec Ros Ser med Ros Cow, Lond. No 234 Eminens, C. W.119411 India, 2. 445 Emmen, CW (1950) J I Jorn 6 jo? Levn. L & Tyndale, 11 i 11936) Mor Sy oop Dul, NY, 34 Son l'etern-Bxrgar!, K A Tonnesen, MI (1948.4.11 mlr, Copenba 1,38 Smith, O W & Smith, G V (1952) Roent lapor Hurmir Rr 7 C) Sinetulle, r.GeuK, N & Afartun, G r 19 In Lupa 5.246 DISCUSSION Dr Lorame: Ded you use 4-point assays and how mizny ammals were employed per dose level of standard nd unknown preparation? Afoss Hood: Yes, we did 4-point assays. As a rule, fifteen animals were injected for cach dosc of standard and unknown, and twenty when we had enough material Dr Loraine. I wonder if you have any mformation on the index of precision (4) in these assays. Was that calculated? Miss Il'ood. It was not calculated, but we estimated that there was an error of + 30% at least, Dr Klopper: I was intrigued by your very low valucs for pregnanediol cxcretion in nornual cycles. You recorded several values of < 1 mg/24 hr. What was the method you used? Aliss ll’ood. We used the method of Sonimerville, Gough & Marrian, but we do not really consider that any value < 1 mg/24 hr is accurate Dr Klopper: If I remember rightly, Prof. Marrian put the limits of accuracy of hus method at some 2 mg/24 hr. My other point is: how did you estimate the substance you labelled 'TSH'? Did you use the Klinefelter method? Afiss Wood: We extracted our specimens by kaolin adsorption, and used the mcrease in uterinc weight in infantile mice for assay. Dr Klopper. Are you confident in labelling this material 'FSH'? Afiss Il’ood. No, we are not. We should llave altered the labels on some of our slides I actually referred to it as ‘gonadotrophen' in speaking We thuk it is FSH, but we would not like to commit ourselves. Dr Boscort I was interested in your results on oestrone and oestradiol in cases of adrenal tumours, because Migeon & Gardner have recently shown by fluorimetric and chemical methods that in certain patients with adrenal tumours there is a raised urinary oestrogen output Have you ever assayed oestrogens in human urines after treatment with ethyl ·82 PHYLLIS M, HARLOW, MARJORIE SANDIFORD AND MARY E. WOOD Oestronie oestradiol excretion in other conditions Single investigations were made in a number of other conditions, but as the amounts of ocstrogen found fell within normal range, they were of little valuc. These include oligo- and hypomenorrhoca (thrce cases), dysmenorrhoea (two cases), Cushing's syndrome (two cases) and gynaecomastia (three cases). {גוון DISCUSSION The error involved in the biological assay of oestrogens makes the use of large numbers of animals advisable if accurate results are to be obtained; this in turn necessitates the collection of large volumes of urine, particularly from patients with a low Ocstrogen output, in order to obtain sufficient matcrial for assay. Estimation of urinary oestrogens at frequent intervals during normal menstrual cycles showed that it was possible to obrain consistent results using comparatively small numbers of mice, and our findings compared well with thosc of Pedersen-Bjergaard & Tønnesen (1948), Smith & Smith (1982) and other workers It is clear from the study of oestrogen cxcretion in the conditions mentioned above tlat assay of oestrone oestradiol alone is seldom of much valuc. Simultaneous estination of 'urinary gonadotropluns is essential when investigating cases of amenorrhoea, a persistently high output of gonadotrophın being in itself diagnostic of ovarian farkure, and records of basal temperature or pregnanediol output are useful in detecting ovulation. In most cases of amenorrhoea only a scries of investigations will reveal the true condition In cases of precocious puberty or in adult women where amenorrhoea is associated with hirsutisnı it is usual first to estimate the urinary neutral 17-ketosteroids, as these bc greatly increased when the condition is secondary to adrenocortical hyperplasia or tumour, and may climinate the need for oestrogen assay. Oestrogen estimations should always be performed on more than one specinien from the same patient in order to take into account any cyclical or day-to-day variations, and it is preferable to have patients in hospital to ensure collection of complete urine specimens over periods of 2 or 3 days It follows that biological ocstrogen assays are of value in some conditions, but arc to be avoided if conclusive results can be obtained by other means The preparation and assay of urine extracts is laborious and time consuming and the maintenance of a large colony of ovaricctomized mice is expensive, in addition the experimental error, like that of most biological methods, is large We have seen that considerable advances have recently been made in the determination of urinary oestrogens by chemical methods. The results we have obtained by a less accurate method indicate some of the problems which require reinvestigation lising an improved technique. It is probable that, particularly where such small quantities as those found in chuldren and some cases of anienorrhoca are concerned, such methods will yield results of grcator significance. We wish to thank Prof C. H Gray for valuable advice and for providing the facilities ·THE VALUE OF OESTROGEN DETERMINATIONS TO THE CLINICIAN By P. M. F. BISHOP From Guy's Hospital, London, S.C.1, thic Chuisca Hospital for ll'omen, London, S.11.3 and the Pustgraduate Afedical School, Ducane Road, London, Il'12 The extent to which the clinician is likely to call for oestrogen determinations depends on many factors. An important one is his attitude towards the laboratory. There is, I think, a current tendency for the clinician to rack his brams to think up cvery conceivablc laboratory and other investigation in order to 'work up' the case as completely as posto think of the y-staffed steroid 10 -11, T... 1 Another factor is the degree of inconvenience to which they patient may be put. If one is investigating a case of gonadal or adrenocortical cunour—and such cases arc, of course, excessively rarc—the patient will be in hospital, 24 lur specimens of urine and frcquent samples of blood can be easily and conveniently collccccd, and the laboratory staff will be as enthusiastic and excited as the clinician concerning the outcome of the investigations. But af, on the other hand, one is minded to study the urinary steroid cxcrction pattern in a woman suffering from non-ovulatory cycles or metropathua lucmorrhagica, onc is Irable to encounter a certain amount of patient-resistance if one calls for frequent or serial 24 hr specimens. 'Ovcr-night' specimens, however, are relatively convenient to collect, and a co-operative patient will usually consent to supply these. Another factor which must be taken into account is how urinary (or blood) ocstrogen determinations compare in complexity of technique, expense and informative valuc with other available methods of demonstrating ostrogen activity. Take, as an example, menstrual disorders. Ocstrogen and oestrogen/progesterone activity may be studied by analysing basal temperature records, describing the hustological findings of a specimen obtained by endometrial biopsy and evaluating the cornification index and other features of daily vaginal smears, as well as by determining the total ocstrogen (and possibly the oestrone-oestradiol and oestrol ratios), and the pregnanediol output in the urine. In the first place it depends on what information is required if you want to decide whether ovulation is taking place, daily basal temperature records will provide a rough check and may even indicate not only that ovulation has occurred in any particular cycle, but also when it occured. Endometrial biopsy mflicts upon the patient anything from mild discomfort to quite severe pain or even an anaesthetic and a night in hospital. Patients in this country may submit to it one or twice, but not repeatedly. Performed on the first day of menstrual blceding it will provide information on whether progesterone has been present and whether the endometrium has responded satisfactorily to the ocstrogen/progesterone 85 ·84 PHYLLIS M. IARLOW, MARJORIE SANDIFORD AND MARY E. WOOD oestradiol? It is the most potent orally-active oestrogen in existence, but nothing is known of its metabolism. Could you throw any light on this subject? Miss Wood: We have estimated oestrone oestradiol 172 urine of patients who were being treated with comparatively small quantities of ethunyl oestradiol, and, as far as we could tell, it made very little impression on the ocstrogen output. But we did not do any proper quantitative recovery cxperiments. It disappeared; I was not being excrcted in the urine. Dr Boscott: Perhaps it is excrcted in the fieces. Dr Jellinck: I have done some experiments in which cthinyl ocstradiol was injected mto rabbits, and most of the activity scened to be in the facces. There was very little in thic urine at all, and there was no evidence that it was converted into either ocstronc or ocstradiol. Dr Diczfalusy. I wonder whether in your 4-point assays you have ever noticed a signficant deviation from parallelism for your urinary extracts? Afiss I'mod: No, we have not. We did quite a lot of preliminary work on thes, but we found no significant deviation Dr Diczfalusy: About this question of cthmyl oestradiol: Stinimel & May (in the J chis. Endocrin., 1951) claimed there is only cthinyl oestradiol in the urine after injection of 2 mg of tlus oestrogen, and that no oestronc, ocstradiolor pestssol was found. Miss Ilood We have not really done enough work on this to be able to say anything.

·YALUI OI OLSTROGIN DITLIMINATIONS 87 adrenocortical tumours in males who may or may not slow femimat1011, or in certain cases of gynaecomastia. Here the determination of the oestrogen is essential and complcmentary to the determination of the other hormo!ic. Oristian finnsaurs. Perliaps the most valuable use of urinary aestrogeri dctcrminations concerns the differcntirtion between ovariin tiimours producing sexual precocity and the far more comnion condition of constitutional precocious puberty. Such ovarian tumours arc tisuually granulosa cell tumours, though Scckel (1946) lzts described a case of cloniocpitheliona with a urinary oestrogen content cquivalent to Gooi ui. of oestronc per 24 hr and a high concentration of chorionic gonadotrophin Removal of thc tuniour resulted in disappearance of both C.G. and oestrogen from the urmc. Granulosa cell tumours occurring in childhood arc citreincly rarc, and Seckel was able to collect only thirty-onc cases from the litcrature between 1926 and 1944. High blood and urinary ocstrogen levels have been recorded in some of the recent cases, with absence of urinary gonado troplins, whicrcas in constitutional (or neurogenic) precocious puberty oestrogen levels do not exceed the normal adult range and gonadotrophms are usually demonstrablc. Testicular tunicuirs. Occasionally interstitial cell tumours wluch, when they occur in chuldhood, characteristically give rise to isoscxual precocity, are associated with Synaccomastia. Of tlic scventeen cases of interstinal cell tunours occurring before puberty which I have been able to collect from the literature (Bishop, 1954). two developed gynaecomastia (Huffman, 1941; Jolly, 1954). Other femınızıng interstitial cell tumours have becn descnbcd in adults (Monaschkin, 1926; Hunt & Budd, 1939; Melicow, Robinson, Ivers & Rainsford, 1949). Unfortunately, oestrogen determinations were not undertaken in these cases, but it sccnus possible that cxcess of this horinone was the cause of the feminization, for Maddock, Epstein & Nelson (1952) found increased quantities of urinary oestrogen in two cases in which the only functional testicular structures were clumps of Leydig cells. They also treated men with normal testes with chorionic gonadotrophin, which ization of tlic scnuniferous tubulos however, no general agicement ULUL Le Leyuig coils are the source of testicular oestrogen as well as androgen. Huggins & Moulder (1945) described Sertoli cell tumours in five dogs with enlarged nianimary glands and other signs of feminization. Two of the tumours yıclded appreciable quantities of destrogen. Teilum (1946) has published the only account of a Sertoli cell tumour in an otherwise normal man with descended testicles Unfortunately, oestrogen determinations were not undertaken in this case, but the patient complained of gynaecomastia and impotence. McCullagh & Schaffenburg (1952) suggest that the testicular oestrogen is elaborated by the germinal epithelium, possibly by the spermatıds and spermatozoa, and the observations of Diczfalusy to-day are of great interest in this connection. It is clearly important to carry out oestrogen determinations wherever possible in these rare cases of testicular tumour and in all cases of gynaecomastia, to discover the source of testicular ocstrogen and to define the different causes of gynaecomastia, which may result on the one hand from exogenous oestrogen administration and, on the other, from injection of chorionic gonadotrophin wluch induces physical signs of androgen production as well as gynaecomastia (or There is, ·86 P. M. F. BISHOP influence; or whether progesterone has been deficient in its concentration or effect on the endometrium; or whether it lias been completely absent, possibly with continuous, unopposed or even excessive influence of oestrogen The endometrial picture depends not only on the concentration of ocstrogen and progcstcrone, but the responsiveness of the endometrium to these hormones. A study of the cytology of vaginal smears is a somewhat time-consuming and exacting labour, if modern techniques are employed. Daily smears are essential in most cases They can be taken by the patient herself if she is moderately intelligent, lughly cooperative and somewhat introspective, and if she is provided with a series of microscopic slides to which paper clips have been attached to prevent the smears rubbing agamst each other; a diamond to cut the date of each smear onto the appropriate slıdc, and a pipette of the right length to reach the posterior vaginal fomix-attached to the Pipette is a rubber bulb to provide suction; and finally, a slıdc jar containing fixing fluid consisting of equal parts of ether and 95% alcohol. Analysis of the sinears reflects the effect of oestrogen and progesterone on the vaginal epithelium, this effect varying not only with the concentration of these hormones but also with the responsiveness of the vaginal mucosa Endometrial biopsy and vaginal smear studies often do not indicate the same degree of ocstrogenic or oestrogen/progesteronc effect This means that, in these cases, the endometrium and vaginal mucosa are not cqually responsive Finally, there arc the chemical and biological methods of estimating ocstrogens (and pregnanediol) in urine which have been so fully described and discussed to-day. Generally speaking, they require expensive and delicate equipment, high maintenance costs for solvents and other materials, or for the feeding and care of animals, rather skilled technicians in relation to the salary scale on which one is forced to operate and, finally, exceptionally skilled and well-trained steroid chemists and biologists such as the ladies and gentlemen who have been addressing you to-day. These inethods of estimating urinary ocstrogen give some idea not only of the quantity, but also of the quality of the ocstrogens which are being produced by tlc organism. but nevertheless they pin-point only one figure in a complex metabolic cycle wluch is dependent on such factors as rate of sccretion and circulation, extent of conjugation, efficiency of uptake by the target tissue wluch may involve enzyme systems and other complicated mechanisms--as well as renal clearance and the extent of recovery of hormone from the urine Like the various tests of thyroid function, the tests of oestrogene activity reflect different aspects of the problem. An estimation of the serum protein-bound 10dıne gives an indication of the concentration of circulating thyroid hormone A determination of the radioactive iodine uptake by the thyroid gland slows how rapidly and efficiently the gland receives its raw material. Radioactive iodine clearance studies show how quickly the hormone circulates and is excreted, and an estimation of the basal metabolic rate indicates how the organism as a whole is responding to the thyroid hormone with which it is being supplied. Yet another factor is the desirability of determuning hormone ratios, such as the gonadotrophindoestrogen ratio in order to locate the site of a lesion in the pituitary/ovarian axis; or the oestrogen/pregnanediol ratio in a case of menstrual disorder m order to evalu: the degrec of corpus luteum failure, or the oestrogen/androgen ratio in certain ·VALUE OF OESTROGEN DETERMINATIONS 89 gaard & Tomescn (1951) in which oestrogen and gonadotrophin assays were undertaken. Except in some cases of nietropathia Iacinorrlugica in which tlicy found excessive oestrogen output, all their other cases slowed low oestrogen production. These results rouglly coincide with those of de Allende & Orias (1950) who made a most detailed study of the vaginal smcars of women with micnstrual disorders and seldom cncountered high ocstrogen values in amenorrhoca, oligomenorrhoca or anovulatory cycles associated with menorrhagia, but did find high comification indices in cases of ructropathia haemorrhagica. Interesting as this information may be, I do not feel that we arc likely to make usc of routinc ocscrogen determinations in a gynaccological cndocrinc clinic. We shall, however, always have the odd patient whose steroid hormonc pattern we shall be studying, and I think the results of such studies may prove eventually to be of value. In conclusion, therefore, I would say that the methods which have been described to-day will not for some time become routine methods of clinical investigation, in the way that 17-ketosteroid, and possibly FSH dctcrminations, have become popular with the clinician. Furthermore, I think it is important that the steroid chemist and biologist should be shielded from the ununformed enthusiasm of his clinical collcague as far as ocstrogen determinations are conccmcd. On the other hand, I am dehghted to find that there are now availablc rchable methods of providing the clinician with accurate quantitative and qualitative information concerning oestrogen output in specially selected cascs. I believe this symposium has proved beyond doubt the need for endocrine laboratory units attached to clmics dealing exclusively with clinical endocrine material, and I am convinced that close co-operation between the clinic and the laboratory in such a sct-up will lead to the clarification of many diagnostic problems and may encourage rational as opposed to empirical endocrine therapy, especially in menstrual disorders. While, therefore, I applaud the enterprise of the promoters of this symposium, I warn them against the indiscrimunate enthusiasm of their clinical colleagues Oestrogen determinations are still, in my opinion, a rescarch problem and not a routine laboratory demand. REFERENCES Allende, 1 L C de & Ortas, O (1950) Cytology of the yfuman l'agira New York Hocber Bishop. P M F (1954) Recent Aduances in Endocrinology, 7th ed London Churchill Borth, R & de Watteville, H (1952) tutum G Horm Io, 141 Diczfalusy. E & Luft, R (1952) Ada endoct, Copenhagen, 9, 327 Hamburger, C (1938) Les Horntones Sexuelles Ed Brouha, L, P 345 Paris Heinemann Huffman, L F (1941] J Urol 45, 692 Huggins, C & Moulder, P V (1945] Cancer Res 3. 510 Hunt, V C & Budd, J W (1939) J Urol 42, 1242 Jolly, H K (193+) Serul Precoctty Springfield Thomas In the press Luft, R & Sjogren, B (1949) Acts endocr, Copenhagen, 3. 342 Maddock, W 0, Epstein, M & Nelson, WO (1952) Ann NY Add Sa, 55, 657 McCullagh E P & Schaffenburg. C A (1952) 4 NY 12 Sa SS. 694 Melicow, MM, Robinson, JN, Ivers, W & Rainsford, L (1949) J Urol 62, 672 Monaschkin, G B (1926] 2 Urol 20, 8 Pederseri-Bjergaard, K & Tønnesen, M (1951] Acta ondoa, Copenhagen, 7. 170 Seckel, H P G (1946) Aled Chu N Amer 30, 183 Smith, G V & Smith, O W (1948] Physiol Rey 28, 1 Teilum, G (1946] Acta pak microbuol scand 23, 252 Twombly, G H & Hocker, A F (1941) Surg Gynec Obstet 73.733 ·P. 3. F. BISHOP 'puberty mastitis') in adolescent youths. Chorioepuhehoma of the testis is usually associated with high ocstrogen and C.G. cxcrction, as well as gynaecomastia, and other testicular tumours such as semmomas and mixed cpithckomas have been shown to yield high oestrogen values [Hamburger, 1938; Twombly & Hocker, 1941). Adrenocortical tours. Whereas the characteristic feature of these tumours is the high output of 17-ketosteroids and percentage of dehydroepiandrosteronc, nmeteen cases have been described as occurring in men (one was a child of s) and giving rise to feminization, namely gynaecomastia, impotence, azoöspermia and testicular atrophy. In six of these cases opstrogen determinations have been undertaken and have been found to be uniformly lugh Diczfalusy & Luft (1982) submitted the urine in their case to countercurrent distribution and fluorimetric analysis250 ml. of urine yielded 34.5 mg of frec ocstriol. After hydrolysis 68.5 nig of oestrone, 18.4 mg of ocstradiol and 857 mg of oestriol were recovered from the same specimen of urine. Unfortunately, in only thrce of these six mses were 17-ketosteroid studies also undertaken. In one of these, reported by Luft & Sjögren (1949), the 17-ketosteroid output was disproportionately low compared with the yield of ocstrogens. It is clearly of great importance that conbined studies of 17-ketosteroids and ocstrogers should be carried out a fututc cases of this very rare condition. Pregnancy disorders. The intriguing thicory of the Smiths (1948) that many disorders of pregnancy, such as certain cases of habitual and threatened abortion, pregnancy vomiting, coxaemia, pre-eclampsia, cclampsia and diabetes are associated with an imbalance of placental hormone production, leading to excessive output of chorionic gonadotrophın and diminished quantities of urinary oestrogen and pregnanediol, and that thus imbala ance may be readjusted by administration of oestrogen (especially stulboestrol) and possibly progesterone, has stimulated the imagination of obstetricians and diabetic experts as well as steroid chemises, and is still the subject of serious, though essentially benign controversy. The modern techniques of oestrogen determination which have been laid before us to-day must obviously be applied to this problem, and tlus as osse of the very few conditions as which blood, as opposed to urinary, oestrogen estimation may prove valuable According to Borth & de Watteville (1952) 1.6 to 3 1.4. of oestrone/ ml of serum is the normal range in the last trimester Apart from the Smuhs' work there are other fascinating problems which confront the obstetrician and fus endocring laboratory, One e che sudden fal 11 Ostrogen oveput a cser of missed abortion and foeral deagh Mfenstrual distsırbances So far as I can see, the study of hormonal patiens in nenstrual disorders is an almost uncharted sca. Some of the pioneer pilots of those uncharted occans are here to-day It is obvious to any clinician who has prescribed hormone therapy en bluskers' for menstrual disorders as long as I have, that no further progress can be made without an endocrine laboratory unut There are few who have so far luad such facilities De Watteville, in my opinion, as the outstanding example, and for the results of hus experience I refer you to his recent article (Borth & de Watteville, 1952). In thus he adopts a rather conservative attitude to the clinical value of the work wluch lie revielys. He refers to the studios of 553 cases of menstrual disorder by Pedersen-Bjer- ·GENERAL DISCUSSION 91 cuestrogen m scmcn conies from the testis, but is rather a biological curiosity, and probably has no physiological function. That brings mc to Dr Mitchell's paper on the ocstrogen content of the placenta. Here again, the chief query raised by this paper to brologists is what the ocstrogen is doing in the placenta, and whether it Inns any physiological function. The first difficulty in dealng with tlus particulır question is the fact that different spccics vary so much m the oestrogen content of the placents, and it is difficult to visualize that it can have any function generally applicable to mammals. I think I am right in saying that, with thic definute exception of the cow and the possible exception of the horse, oestrogen is not known to be abundant in the placenta of mammals other than man. That raises the next and rather obvious comment that the cndocrmological characteristics of the various reproductive processes appear to vary morc from species to species than any of the other cndocrinological functions. As I was listenmg this afternoon, I recalled thic remarkable cudocrinological differences between the pregnant marc and the pregnant woman. Tlic former has a very large amount of gonadotrophm of the pregnant mare's scrum' type circuliting in the blood, but not much oestrogen which, however, appears in quantity in the urine. By contrast, in women, there is some gonadotrophun and some oestrogen circulatıng in the blood, and carly on a large amount of gonadotrophın and later a considerable amount of ocstrogen in the urine. The type of gonadotrophın is entirely different from that in the mare When the next paper was read by Miss Wood, I began to fecl entirely at home, the word 'biological assay' was mentioned, and the various difficulties were emphasized. One of the reasons it is so satisfactory to sec chemical supcrscding biological methods of oestrogen determination is that there are so many different factors which cause variation in the result of biological assay--a slightly different icdiod of comparing the unknown with the standard will produce very different results I well remember that my first contact with our Chairman, Sir Charles Dodds, arosc many years ago because we were obtaining different results in the biological assay of the same preparations of ocstrogens. We had begun to liave grave suspicions of cach other when it turned out that I'rof. Dodds' tcanı were using six mjections on rats, wlule we were using four injections on imce, a scrious discrepancy was thus explained There is another factor in biological assay which makes things very difficult, and that is the degree of purification of the preparation It is one of the anomalies of biological assay that the more purified the preparation, the less active it appears in many of the ordinary tests. That, of course, is because the purified substance is absorbed very rapidly and excreted without having time to do its Job Many of those undertaking biological assays for a chemist have got to troublc, because the better he worked the less they inade of his results Having said tlus about biological assay, I should add that I realize full well that those engaged in chemical assays also have their troubles, and I was interested labour for everybody concerned, and it is obviously not quite certain in his mind, and ·GENERAL DISCUSSION (AFTERNOON SESSION) Dr Parkes: I have been out of active participation in dluis particular subject for a number of years, and I am not as well up in the details as I ough to be to open thus general discussion. There are, however, a number of general remarks which I think can ficfully be made about the conference. The first relates to the interesting division of the two sessions and various papers between chemical and biological techniques. In the past, biological methods of assay had to be relied upon almost entucly. By contrast, in planning today's symposium the original idea was to devote the moming session to chemical techniques and the afternoon to biological tcchniques. In practice the chentical techniques have overfloweduto the afternoon session, and in only one paper have the authors dcpended entirely on biological assays. Thus, I think, is a very defimte sign of progress. In 1938 there was a long and arduous international conference on the establislinent of international standards for the pituitary hormones concerned with reproduction. At the end of thus conference we issued a report, sluch recorded a remark made by Sir Henry Dale that the ultimate aim of biological assay was self-abolition. I think, today has shown that tlus higluly desirable process has gone a long way. Thus conclusion seems to be backed up by the fact that if participacion m the discussions is any guide, then the mcmbers of tlus audience are more chemically than biologically munded. * Tummg to the papers of this afternoon's sessions, there is firse the interesting paper by Ds Diczfalusy on the oestrogen content of human semen. I have some particular interest in thus paper because years ago, following persistent reports of small amounts of oestrogen in semen, I was lured into doing some biological experiments, and after some İstelc work obtained results that were rather mconclusive. I was very glad to hear today that the definite answer was obtained, on what appears to have been abundant material, that oestrogens (all three of the common Destrogens, apparently)undoubtedly occur in human semen That is an interesting observati011. But the biologist imediately asks, 'Where does the stuff come fron? And what does it do there? The answer to thc first question secins preety clear, and has already been claborated by Dr Bishop. The oestrogen presumably comes from the testis, probably from the seruniferous tubules, and may be closely associatcd with the spermatozoon The question of what, if anythmg, 15 its plıysiological function in semen is, of course, much more difficult to answer. There lave beon various suggestions from time to time, but I think we must remember that the amounts present, in terms of biological reactions, arc very small indeed. Halfa gallon of human semen is not a physiological amount. The amount of ocstrogen available from physiological quantities of semes, although it migkc perhaps do something to a rar, is not hkely to produce an effect of any kind in its natural habitat. I suggest, therefore, that the *In view of Sir Charles D) kis's comments farct in cosmg the distussam Sougha o adu that in making these restacks I had in mind only the assay of knowls oestroke of osteogenix preparations Babykal testings, of course, usually unavoidable in the first qualtative examination of new compounds or preparatus for oestrogenic activny. 90 ·GENERAL DISCUSSION 93 methods of estimation when we know we are expecting rather larger differences than we think. A simplification at the expense of accuracy may be quite permissible when we arc looking for large differences. But I think in the end—and Dr Swyer made chis point in our paper this morning-only the blood values will have any meaning at all, or the blood together with urinary values. Dr Parles: I think it is quitc cvident that the information wanted is as to whar is going round in the circulation, not what is being excreted. Until what appears in the urine is shown to have some significant relation to what is going round inside, I think the whole position is not on sccuire ground. Miss Stern: In defenccofour micthod of dong oestrogen dererminations on urinc, ac have started it because urine was thic casiest fluid to get hold of and without much bother to the patient; with blood uc may nccd quite large amounts, Alo, one of the ultimate ains after examining oestrogen in blood is oestrogen metabolisın, and before we can do that, we must have accurate methods for determining oestrogatus in urinc. Dr Brown Whilc admncting that blood oestrogen methods are very desirable, I would like to put in a word for urinary oestrogen estimations. In all the menstrual cycles that se have been studying, the features have been so constant that we were able to predict, without any reference to basal temperature cures ctc., where ovulation occurred, when menstruation occurred, and so on. I think for the first time in history we are able to deterinine what the ovary is doing without having to rely on menstruation or any of its outward physical signs, and chat seems very valuable in itself. Dr Parkes: That is an instance of the type I mentioned, where you are able to show that what you measured in the unnc did give an indication of what was going on inside. That is what we want. Dr Romney. I do not want to provoke Dr Parkes any further in connection with the oestrogen production by the stallion Thus may be a curiosity, but I think it is established that the adrenal cortex produces, in addition to corticosteroids, oestrogens and androgens, while I think the late Dr Dingemanse showed that both androgens and oestrogens are produced by thcovary. Again, virilization in the Stein-Leventhal syndrome is believed to be due to ovarian androgens I do not see why, if the ovary can produce both oestrogens and androgens, the testis should not normally produce both these substances as well. Dr Parkes: There is no reason at all-in fact it does. May I make one more comment about this stallion business. Shortly after Zondek made that remarkable observation of the enormous concentration of oestrogen in the stallion urine, it became popular to suppose that the ordinary benign cnlargement of the prostate in man had some connection with the oestrogen story In association with Prof. Zuckerman, we went to some trouble to get the prostates of a number of stallions, with particular reference to the uterus masculinus which is quite a well-defined organ in the horse The point was that injected oestrogen, irrespective of what it may do to the prostate, undoubtedly causes changes in the uterus masculinus of experimental animals. Yet in the stallion we could find no changes whatever in the uterus masculinus suggesting that oestrogen had been acting on it. And yet the animal excretes 100,000 units a day. Dr Sitryer. An almost analogous situation arises in the case of the pregnant woman given ·92 GENERAL DISCUSSION it has never been quite certain in mme, whether the amount of uformation given abour the patient-I am not talking about academic or chemical or biological mattces-really justifies it. That is undoubtedly onc of the things that swill be cleared up in fixture. Part of this difficulty seems to be due to the fact that over the years there has been this intense concentration on what appears in the urine which, for all we know, may be merely what the organism does not want, and which may therefore not give much indication of what is going on inside the organisin.Years ago 1 mooted the idea that it would be a good thing to get away from uric and back to blood, and I thunk I am still justified in repearing that remark. But my chief function here is to provoke other people into speaking, and I hope I have at least donc that. Dr Dıczfalusy: I feel I must protest agunist Dr Parkcs' conclusions about the ocstrogen secretion of the testes He considered it in the nature of a biological curiosity that there is oestrogen in human semes, and that one simply cannot understand why it should be there. Now, I understand that it was a provocative remark because Dr Parkes wanted Ls to form an opinion on this problem. I do not see why one might not assunie that the testis, as an organ of internal sccrction, would secrcte oestrogens, and that these oestrom gens probably have an important function in acting on pituitary gonadotrophun sccretion. I suppose that was the reason why people looked so intensely for the existence of a second testicular hormone. Dr Parkes lf, as I take it, you refer to the assumed ocstrogenic secretion of the testes, that may or may not have some function to perform. What I meant to say was that it was not in the Icast evidenc to me what physiological function was performed by such part of that secrecson as got into the semien. Dr Diczfalusy I misunderstood. Dr Boscout. I wonder if Dr Parkes would hke to comment on another species difference He compared the pregnant marc with the human female How about the stallion conpaced with the human wale? The stallion excretes as much as 20 mg of oestrone/latrc, whereas the hunan male excretes very much less than that Dr Parkes I regard the stallion is very much to the point 13 showmg that these bsockerical curiosities of sccretion or excretion do occur and are very difficult to explaun in any rational manner, and indeed may have no rational explanation at all the fact that the stallion excretes in its urve 100,000 units a day or something similarly enormous m-25 just one of those things I do not think it means anything in particular Dr Holclore Bramsberg I should very much like to support Dr Parkes' 'back-to-blood movement We are all brasy studying the excretion of oestrogens, and there arc cccian aspects which nught be worth considering in that connection, such as the renal function, which seems to be ignored if you consider the presence of free ocstrogens as well as conjugaces, surely the degree and efficiency of renal function must bear some relation to the amount excreted, and that nught obscure any physiological effects Again, the dayto-day and hourly variations may do so, and even daily variations may obscure any physiological differences by our not doing enough daily specimens In connection with the clical value of any oestrogen excretion values, surely the difference betwcen normal and abnormal individuals is not yet established, and we may be able to simplfy our

·MEMOIRS OF TE SOCIETY FOR INDOCRINOLOGY
No 1, The Thyroid Gland, July 1953
No. 2. The Determination of Adrenocortical Steroids and their Metabolics, November 1953
No 3. The Technique and Significance of Ocstrogen Determinations, February 1955

·LIST OF CONTRIBUTORS BAULD, W. S. Department of Biochemistry, University of Edinburgh BISHOP, P. M. F. 122, Harley Street, London, W.1. BOSCOTT, R. J. Department of Anatomy, Medical School, University of Birmingham BRAUNSBERG, HANNELORE University College Hospital Medical School, London, W.C.1 Brown, J B. Clinical Endocrinology Research Unit, 2, Forrest Road, Edinburgh, 1. DrcZFALUSY, E. Department of Women's Diseases, Karolinska Sjukhuset, Stockholm, Sweden Dodds, Sir CHARLES, F.R.s. Courtauld Institute of Biochemistry, Middlesex Hospital, London, W.1 HARLOW, PHYLLIS M Department of Chemical Pathology, King's College Hospital, Denmark Hill, London, S.E.S. JELLINCK, P. H. Courtauld Institute of Biochemistry, Middlesex Hospital, London, W. 1. KELLIE, A. E Courtauld Institute of Biochemistry, Middlesex Hospital, London, W1. KLOPPER, A. Clinical Endocrinology Research Unit, 2, Forrest Road, Edinburgh, 1. LORAINE, J A. Clinical Endocrinology Research Unit, 2, Forrest Road, Edinburgh, 1 MANN, T, FRS. Molteno Institute, University of Cambridge, Downing Street, Cambridge. MARRIAN, G F., F.R.S. Department of Biochemistry, University of Edinburgh. MITCHELL, TL Jessop Hospital for Women, Sheffield, 3, and MR.C Unit for Research in Cell Metabolism, The Unuversity, Sheffield, 10 PARKLS, A. S National Institute for Medical Research, The Ridgeway, Mill Hill, London, NW 7. PRUNTY, F T G Department of Chemical Pathology, St Thomas's Hospital Medical School, London, S E.I REISS, M. Biochemical and Endocrinological Rescarch Department, Bristol Mental Hospitals, Fishponds, Bristol RUMNEY, G Hormone Rescarch Laboratory, Hadassah Medical School, Jerusalem, Israel SANDIFORD, MARJORIE Department of Chemical Pathology, King's College Hospital, Denmark Hill, London, S.ES STERN, MARGARET I University College Hospital Medical School, London, WC 1. SWYER, G. I M. University College Hospital Medical School, London, WCI WILLIAMS, P. C. Imperial Cancer Research Fund, Mill Hill, London, N.W.7. WOOD, MARY E. Biochemical Department, King's College Hospital, Denmark Hill, London, S E.S. 96 FRS

·No text on this page.