THE FORMATION OF ESTROGENS BY LIVER TISSUE IN VITRO by

David Richard Usher, B. Se.

A thesis submitted to the faeulty of Graduate Studies and Research in partial fulfilment of the requirements for the degree of Master of Science.

Department of Investigative Medicine, McGill University, Montreal. April 1961 ACKNOWLEDGMENTS

The author wishes to thank the Banting Research

Foundation for financial support1 and the Research

Director1 Dr. R. Hobkirk1 for much-appreciated advice and help througbout all aspects of this problem. Acknowledgment is also extended to Dr. R.H. Common for donation of the avian liver and to Mr. J. Knowles for assistance in the preparation of the figures. TABLE OF CONTENTS

SECTION PAGE 1- Estrogen Nomenclature 1

2- Introduction 4

3- H:l.storical Survey i) Earliest work 5 ii) Experimental hepatic posioning 6 iii) Splenic implantation techniques 7 iv) Vitamin and protein-deficiency effects 7 v) Enterohepatic circulation of estrogens 9 vi) Species differences 11 vii) Estrogen content in adult liver 11

viii) In vivo - in vitro deficiency studies 12 ix) Investigations of the enzyme systems 12 x) Estrogens and hepatic disease 13 xi) Sex difference 15 xii) Role or the retieulo- endothelial system 15 xiii) Early in vivo estrogen interconversion 16 xiv) Perfusion studies 17 xv) The effect of partial hepatectomy 17 xvi) Incubation with cu1tured liver ce11s 17 xvii) In vitro estradiol conversion to estrone 18 xviii) Review of bioassay procedures 18 xix) Chemica 1 assays 19

i SECTION PAGE 3- Historical Survey - cont'd. xx) Countercurrent distribution 20 xxi) Radioactive isotopes 21 xxii) Isotopie evidence for enterohepatic hypothesis 22 xxiii) Estrogens in the newborn 23

xxiv) Estriol formation 24

xxv) Androgen-estrogen conversion 24 xxvi) Progesterone relationships 28 xxvii) Estradiol metabolism in avian liver 29 xxviii) Hepatic sulfate conjugation 29 xxix) The role of the newer estrogens 29

xxx) Estriol metabolism 34

xxxi) Transmethylation reactions 34

4" Purpose of this Study 36

5- Methods i) Incubation 1 (adult liver): estradiol-cl4

and 16-ketoestradiol-c14 as substrates 37

(a) Incubation procedure 37

(b) Extraction procedures 37

(c) Counting procedure 39

(d) Countercurrent distribution 39

(e) Chromatography 40 (f) Methylation 41

ii SECTIO.N PAGE 5" Methods - cont'd. i) (g) Detection of contamination of 16-keto­ estradiol-c14 by labelled estriol 41 ii) Incubation 2 (newborn liver): estradiol-cl4 as substrate 42 iii) Countercurrent distribution of estriol methylMether standard 43 iv) Incubation 3 (newborn liver): 16-keto- estradiol-C14 as substrate 43 v) Incubation 4 (adult liver): testosterone-c14 and progesterone-c14 as substrates 44 vi) Incubation 5 (avian liver): estradiol-cl4 as substrate 46 (a) Separation of 6 fractions by column

chro~~atography 47

(b) Sodium borohydride reduction 49 (c) Countercurrent distribution of 16-epiestriol 50

6- Resulta

i) Incubation 1 51 (a) Estradiol-cl4 as substrate 51 (b) 16-Ketoestradiol-C14 as substrate 51 (c) Percentage contamination by labelled

es triol 52 ii) Incubation 2 52

iii SECTION PAGE 6- Resulta - cont 1 d.

iii) Incubation 3 53

iv) Incubation 4 54 (a) Testosterone-c14 as substrate 54 (b) Progesterone-c14 as substrate 54

v) Incubation 5 55

7... Discussion i) The problem of radiochemical purity as exemplified by the resulta or Incubation 5 57 ii) Significance of percentage conversion 59 iii) Definitive conversions in Incubation 1 59 iv) Pattern interpretation in Incubation 2 59 v) Incubation 3 60 vi) Lack of conversion of progesterone 60 vii) Difficulties in analysis or the testosterone incubation 60

8- Summary 62

9- Tables 63

10- Countercurrent distribution patterns and

radioautographs 70

11- Bibliography 78

iv ESTROGEN NOMENCLATUHB

(The trivial name is listed first, and then the proper name).

Estrio1

A1:3:5(10)-estratriene-3,16«1 17J-trio1

Es trone A1:3:5(10)"estratriene"3-o1-17-one

Estradio1-17p A1:3:5(10)_estratriene"3,17ft"dio1

Estradio1-17~

Â1:3:5(10)~estratr1ene-3,17~-dio1

16-Ketoestrone

61:3:5(10)-estratriene-3-o1"161 17-dione

16-Ketoestradio1-17p ~1:3:5(10)_estratriene-3 1 17p-dio1-16-one

16-Ketoestradio1-17~ ~1:3:5(10)_estratriene-3,17~-dio1-16-one

16p-Hydroxyestrone ~1:3:5(10).estratriene-3,16fl-diol-17-one

1~-Hydroxyestrone b1:3:5(10)-estratriene-3,16~-dio1-17-one 16.. ep1Estriol

61:3: 5(10)-estra triene .. 3,16! 1 17) -triol

17... ep1Estriol ~1:3:5(10) .. estratriene-3 1 16«,17~-triol

161 17-epiEstriol ~1:3:5(10) ... estratriene-3,16p,l7~-triol

6p-Hydroxyestradiol-17J 41:3:5(10) ... estratriene-3,6p,l7ft-triol

6p-Hydroxyestrone 61:3:5(10) ... estratriene-3,6p-diol-17-one

6c(-Hydroxyestradiol-l7f3

61:3: 5(10).estra triene-3,6«1 17fl ... triol

6~Ketoestradiol-17fi ~1:3:5(10).estratriene-3 1 17p-diol-6~one

2-Hydroxyestriol ~1:3:5(10) ... estratriene -2,3,16~1 17p-tetrol

2-Hydroxyestradio1-17fl ~1:3:5(10)_estratriene-2 1 3 1 17ft-triol

2-Methoxyestriol ~1:3:5(10).estratriene-2-methoxy-3,16~17J-triol

- 2 ... 2-Methoxyestrone Âl:3:5(10)_estratriene-2-methoxy-3-ol-17~one

2~Methoxyestradiol-17p ~1:3:5(10)_estratriene-2-methoxy-3 1 17p-diol

- 3 " INTRODUCTION

The capacity of both the human liver and hepatic tissue from various animal species to synthesize, interconvert and destroy many, if not a.ll, of the large number of estrogens now known to exist, has been amply demonstrated by numerous workers. Techniques have so changed and improved in the past decade that it is now possible to have definitive and quantitative concepts about the different reaction sequences. lt is fully realized that these same advances present grave difficulties (such as the determination of radiochemical purity) which will be discussed later in this work.

- 4 ... HISTORICAL SURVEY

Zondek (11 2) 1 in 19341 was the first to demonstrate any connection between estrogen metabolisa and the liver. He showed that after oral or subcutaneous administration of large amounts of estrogenic hormone (up to 5401 000 mouse units in women, and 401 000 mouse units in infantile rats) only 1 - 3% of the administered dose was recoverable in the urine. He also found that estrogenic hormone was inactivated after incubation with liver mince. H~ therefor~ postulated that the liver was probably responsible for the low in vivo recoveries.

No further work was published unti1 1937. In that year1 Israel et al (3) were the first, and are still one of the few groups, to have used perfusion techniques in this field. They reported that a heart-lung system did not effectively inactivate administered estrogen, whereas rapid inactivation occurred using a ~-lung-liver system. At the same time1 Engel and Navratil (4) proved that the liver was not the only organ able toœstroy estrogens in the case of cold-blooded animals, since they were able to demonstrate estrone inactiva­ tion in the hepatectomized frog. In the following year1 Parker and Tenney (5) analysed the estrogen content of fetal and maternai organa. They found that both fetal and maternai livers contained considerably more estrogens than did the placenta or other organs; and reasoned - rather vaguely -

- 5 - that the increased level of estrogens in pregnancy was not due to placenta! manufacture, but involved an increase in the "general cholesterol metabolism" of the liver and adrenal.

Haller et al (6-8) 1 working between 1939 and 19431 extended Zondek's in vitro observations. They showed that

~-estradiol, now known as estradiol-17J (9) and hereafter termed simply estradi>l, was eompletely inactivated by rabbit liver slices. Rat liver destroyed both estradiol and estriol, but renal tissue bad less effect, and after incubation with other tissues the estrogens still retained their original poteney. Working with non-pregnant and pregnant animals, Heller et al demonstrated that all three classieal estrogens (estradiol, estrone and estriol) were m8tabolized to the same extent by liver slices from either group. They postulated a hepatic threshold for exogenous estradiol destruction. Above this threshold level, greater biliary and urinary excretion would occur.

At this time, experimental poisoning o~ the liver (~ vivo or in vitro) was first attempted. Talbot (10) used 21-day old female rats, causing acute hepatic damage with carbon tetrachloride and ethanol. At the time of greatest prostrati on, the uterus was stimulated with estrogens. Increased uterine weight, along with oedema and congestion, proved that the liver was no longer capable of inactivating estrogens. Pincus and Martin (11) confirmed Talbot's work.

- 6 - Heller (7) extended this by showing that hepatic tissue treated with sodium cyanide (inhibiting oxidative enzymes) was less effective in inactivating estrone. He suggested that the liver contains an enzyme system reducing estrone to estradiol, plus an estradiol-destroying system, inactiva­ ting the estradiol formed. He also reasoned that estriol must be less affected by the liver, in order to explain its higher oral potency.

Biskind et al (12-18) did much in vivo work, using the technique of implantation of estrone pellets into adult, castrated, female rats. Normally, one would expect a con­ stant estrus to occur, but in this situation the estrone had to first pass through the liver. Estrus did occur, but only for a period of from three to fourteen days with constant anestrus thereafter. If the pellet was placed outside the portal circulation, constant estrus occurred (even if the pellet was placed in the liver). Testosterone propionate pellets were also inactivated when placed in the spleen, and reactivated when the spleen was transplanted outside the portal circulation. Partial inactivation occurred when these latter pellets were placed in the liver. Biskind et al suggested that part of the hormone may have been taken up by the hepatic venules before being acted upon by the liver cells. When the rats with intrasplenic estrone were kept on vitamin B deficient diets, estrus occurred. Vitamin B was also shown to be necessary for testosterone inactivation in

- 7 ~ the male rat. These workers later demonstrated that the liver or normal male rabbits could inactivate estrone, using the criteria or normal genitalia 91 days after splenic implantation.

Golden and Severinghaus (19) were actually the rirst to

employ this general method in 19381 when they transplanted rat ovaries into the mesentery with resulting anestrus. Retransplantation into the axilla caused estrus to recur.

The above techniques were adopted by others: Shipley and Gyorgy (20) induced liver damage by reeding low protein - high rat diets, causing impairment or estrone inactivation; Gyorgy (21) was able to rectiry the impairment with methionine or protein digests; Segalorr and Nelson (22) demonstrated that the point or attack was probably at the hydroxyl groups, since estradiol was partially protected by shielding these groups; Hooker et al (23) showed that monlcey liver did not inactivate estrogens, even arter administration or vitamin B; Drill and Preirrer (24) questioned the speciricity or vitamin B since their rats underwent estrus even with this supplement, when put into a state or inanition. Bernstorr (25) compared the weights or the uterus and the vagina, using castrate mice, and mice with an ovarian grart in their spleen. He round that the organ weights or the rtrst group were lighter than those or the second, thus reasoning that the estrogen rrom a splenic ovary is not totally inactivated by the liver. Vanderlinde et al (26) extended the work or Drill and

- 8 .. Pfeiffer (24), establishing the idea that the quantity and quality of protein intake were the primary factors in estrogen inactivation. In the same year (1950), Jailer and Seaman (27) completed this concept by showing that the liver of rats maintained on 50% casein retains the ability to inactivate estradiol, in spite of avitaminosis B or inani­ tion, while this is not true with animals on a 5 or 15% casein diet. On such low protein diets, vitamin C and gluta­ thione were also necessary for inactivation (28). Other workers (29-38) confirmed and extended most of this work.

Cantarow and his group, initially working with bile­ fistula dogs, surmised that estrogens underwent an entero­ hepatic circulation, similar to bile acids. They (39) gave large amounts of estrone intravenously to these dogs, and found that over 90% was excreted in the bile over the next three days. This threw doubt on the hypothesis of rapid hepatic inactivation or destruction of estrogens in the intact animal. Also, the liver appeared able to store large quantities of estrogens for 24 - 48 hours after the administration of a considerable dose of estrone (2501 000 I.u.). These findings were repeated, using estradiol (40). In the following year (1943), the same workers (41) again repeated the experiment with estradiol, but this time carried out tissue analysis one to two days after injection. Though large amounts of active estrogen were recoverable in the gall-bladder bile, none was round in the liver, spleen, gastro-intestinal wall or hepatic

- 9 - vein blood. Also 1 Cantarow et al noted that the liver of rats poisoned with carbon tetraehloride possessed the same eapaeity as normal liver for inaetivating estradiol in vitro. This did not support the hypothesis that the greater aetivity of exogenous and endogenous estrogens in the presence of hepatie damage was due to a deereased inactivation of estro­ gen by the damaged liver eells. Cantarow et al indieated that estrogens were removed from the blood by the hepatie eells1 eonverted to a substance possessing little or no estrogenie aetivity1 stored temporarily in the liver in this form, and "reaetivated" during the proeess of excretion by the hepatie eells. In 19441 they (42) reported that the blood estrogen eurve in dogs and humans was prolonged after hepatie damage, possibly due to a lessened eapaeity for removing estrogens from portal vein blood after enterohepatie circulation had oeeurred. Also in that year, they (43) laid the nucleus of the concept of hepatie androgen-estrogen eon­ version. The same experimental setup was used as in their early work, exeept that testosterone, androsterone or methyl­ androsterone was given, with resultant estrogen excretion in the bile in six out of nine dogs. Normally1 dog-bile has no estrogenic aetivity. Koch (44) gave testosterone propionate to normal men and women 1 and reported inereased urinary estrogens in both sexes.

In one experiment1 Cantarow et al (45) used Biskind's technique (12) 1 implanting a 15 mg. pellet in the spleen of a

- 10 N bile-fistula dog. They recovered estrogen from the bile for three weeks in large amotmts, but none was recoverable in the urine. This, they suggested, was another reason to doubt the concept of rapid in vivo hepatic inactivation, and was further support for enterohepatic circulation. In 1948, working with

DeMeio, Cantarow's group (46) turned to in vitr~ studies, and showed that diphosphopyridine nucleotide (DPN) plus nicotinamide caused a 50% increase in estradiol inactivation by liver homogenate, with cytochrome C having no further effect. This was confirmed and extended by Coppedge et al (47,48).

Earlier in vitro studies by Twombly and Taylor (49), in 1942, illustrated species differences, with human liver being less capable of estradiol inactivation than mouse or rat liver. They found this decreased capability to have no cor­ relation with the presence of cancer in these patients; and the livers of four strains of mice, differing in their sus­ ceptibility to spontaneous mammary cancer, did not differ significantly in their ability t.o destroy estradiol.

Also in 1942, Tenney and Parker (50) analysed normal and pathological human livers for estrogen, and found a wide range of values (0 - 172 mouse units/Kg. wet weight) with no correlation between the estrogen concentration and the degree of cirrhosis, age, sex or gonadal activity of the patients. Estrogen was occasionally found in the livers of post-menopausal females and normal males.

- 11 - Singher et al (51,52) performed in vivo - in vitro deficiency studies in 1944. They fed rats on thiamine and riboflavin - deficient diets and demonstrated that the liver slices from these animals could not inactivate estradiol, while the opposite was true with slices from control rats receiving these vitamins. The loss of inactivating ability paralleled the decrease in hepatic content of thiamine and riboflavin. Pyridoxin, pantothenic acid, biotin and vitamin A deficiencies had no effect. Singher et al also performed studies in which the liver of rats, maintained on a low protein diet for three months, failed to inactivate estradiol. Hormonal destruction did occur if the diets were supplemented with methionine. Thus, this group lent credence to both the low-protein and vitamin deficiency schools. Bencze (53) demonstrated increased estrone inactivation by liver homo­ genates taken from vitamin C pretreated female rats.

Engel and Rosenberg (54) reported a difference in the hepatic treatment of natural and artificial estrogens, inasmuch as acid liver extracts inactivated both estrone and stilboestrol, while alkaline extracts acted only on stilboestrol.

Levy (55), in 19471 investigated the enzyme system governing hepatic inactivation of estradiol. He showed that oxygen is necessary, and that cytochrome C is ineffective by itself. Fractionation experiments in strong centrifugai

" 12 - rields indicated an enzyme particle size or the order or that previously round ror cytochrome oxidase. The enzyme activity was inhibited by cyanide, azide and carbon monoxide (the latter inhibition being reversed by intense light or 436 mu) all or which points strongly to cvtochrome oxidase. He sug­ gested that the inactivation involved not onlv the cytochrome oxidase-cytochrome C system, but also an additional enzyme such as a dehydrogenase. Ledogar and Jones (56) supported the latter assumption, using the decolorization of methylene blue as evidence ror a dehydrogenase. Repke and Markwordt (57-59) achieved quantitative conversion of estrone to estra­ diol by incubating with glucose-6-phosphate, triphosphopyridine nucleotide (TPN) and glucose-6-phosphate dehydrogenase. Zondek (60) caused inactivation by heating to 70°C., and he stated that the enzymes disappeared in chronic hepatic disease.

Little has been said up to now about estrogens and hepatic disease. Glass et al (61,62) reported an increased excretion or administered estrogens in male patients with cirrhosis. Gilder and Hoagland (63), in 1946, determined 24-hour estrogen excretion in eleven young adult males with acute infectious hepatitis, at frequent intervals rrom onset to convalescence. In nine moderately severe cases, there was a signiricant increase with a return to normal in convalescence. Two years later, Lloyd and Williams (64) drew up a rather fanciful sch eme to explain all the endocrine changes of cirrhosis on the basis or hyperestrogenism, with little

- 13 - experimental evidence behind it. Bennet et al (65) 1 and Cameron (66), tried to correlate the presence of testicular atrophy, gynecomastia and increased urinary estrogens in cirrhotic males, without any great success. Cameron (66) also reported a much increased excretion ratio of estriol as compared to estradiol plus estrone, in five cirrhotics. Muller (67) noted that urinary estrogens in fifteen male cirrhotics were more than double the levels in normal men; but Stoa et al (68) reported excretion values in five patients with liver disease that were above, below and equal to normal levels.

Abriz.ouri and Gomez (69) reported that most male patients dying of chronic hepatitis had a~~ -spermia asso­ ciated with an increase in Leydig cells. This would not necessarily be due to an increased blood estrogen level. Zondek and Black (70) were not able to demonstrate any increased endogenous estrone blood levels, or urinary excretion levels, in patients with extensive liver damage. They carried out estrone-clearance tests, measuring the excess of estrone excretion for three days after estrone administration, and comparing this with normal values. The levels were increased only in the advanced stages or infectious hepatitis (c.f. 71) and in moderate to severe cases of cirrhosis. Pincus et al (72) published a wide range of estrogen-clearance values in liver disease; one" third of the patients showed an increased excretion, while

... 14 -

1 i the remainder had low or normal levels. West et al (73) reported that the rate of disappearance of administered testosterone from the blood was unaffected by the presence of cirrhosis, but there was less 17~ketosteroid formation, implying that the testosterone was metabolized by alternative pathways (c.f. 74,75). However, conversion to estrogens was not mentioned as a possible pathway. Charbonnier and Clement (76) pointed out that hepatitis or obstructive jatmdice would reduce or block the enterohepatic circulation of estrogens.

Kinsell et al (74) summarized the earlier findings by saying that there was no clear-cut evidence proving impair­ ment of estrogen-inactivation as an important factor in liver disease. In fact, Holmes (77), in 19561 indicated that partial hepatectomy, or liver damage (using ethionine or aminopterin) were associated with a decrease in the activity of administered estrogens. Kstriol was found to be the least dependent on the intact liver for activation (as compared to estrone and estradiol).

P. Engel (78), in 19411 stated that liver extracts from normal and castrate male guinea-pigs were twice as effective in inactivating estrogens as were extracts from normal, •astrate and pregnant female guinea-pigs, thus suggesting a sex difference.

In the same year, Zondek and Sklow (79) gave colloidal copper, intracardially, to infant rats, in order to block the

- 15 - reticulo-endothelial system. When estrone was now admin­ istered to these and to untreated control rats, there was the same amount of inactivation in each case. Therefore they indicated that the reticulo-endothelial system plays

no part1 and the liver cell itself must contain the inactivating factor.

Even·as early as 1937, work was being done on the in

vivo interconversion of estrogens. Pincus (80) 1 and Pincus and Zahl (81), were not able to recover estrone or estradiol after administration of estriol to rabbits, although estriol could be recovered after estrone was given to animals with a functional uterus. Also, administering estradiol to animais with functional ovaries caused increased estrone formation. Dorfman et al (82,83) later showed that the ovaries and uterus were not needed for conversion of estradiol to estrone.

Doisy et al (84) 1 in 19421 gave 500 mg. of estriol over a

five-day period to normal and castrate female monkeys 1 and found no increase in urinary estrone or estradiol. Thus estriol was considered an end-product of metabolism. However, Lipschutz et al (85) implanted 18 - 48 mg. of estriol in the spleen of castrate female guinea-pigs, and found that the liver was able to inactivate these large quantities, on the basis of a decrease in the "fibrous tumoral effect" (since subcutaneous estriol in these animals caused the formation of abdominal fibroids). Therefore they reasoned that estriol could not be an absolute end-product.

- 16 - Schiller and Pincus (86) 1 in 1943, carried out perfusion studies with rabbit liver, adding estradiol to the perfusate. About 7 - 12% of the recovered estrogenic activity was present as estradiol, while estrone was found in equal amounts and e9triol accounted for about half of these levels. Using rabbit heart, 90% of the activity was recovered as estradiol, with no estrone or estriol present. They extracted total estrogens into three groups: non-ketonic weak phenols ("estradiol"), ketonic weak phenols ("estrone") and strong phenols ("estriol"). Schiller and Pincus (87) also carried out in vivo work, assaying these three fractions in rat urine, before and after partial hepatectomy, and before and after administration of estrone. Normally there was an 18.7% average recovery of total estrogens after estrone was given. After partial hepatectomy, the "estradiol" fraction showed the greatest increase. When estrone was now administered to these animals, there was a 20% increase in the "estriol" fraction over non-hepatectomized controls, a 400% increase in "estradiol" and a 600% increase in "estrone". The total

estrogen recovery was now 65.5%. In a final perfusion study1 Schiller (88) added all three estrogens, in turn, to the perfusates, using rat liver. When estradiol or estrone was added, all three estrogens were recovered; but when estriol was used, it was the only one recovered.

An original technique was recently used by Ericksen and Velle (89), who incubated estrogens with cultured bovine

" 17 - hepatic cells. Estrone and estradiol were inter-convertible (with about 4% yield in either direction), but neither hormone lvas transformed to estriol.

Pearlman and DeMeio (90) incubated estradiol hemi­ succinate with a rat liver homogenate. Crystals obtained from the weakly-acid ketonic phenols showed no melting­ point depression on admixture with pure estrone. Crystals formed after acetylation caused no melting-point depression on admixture with estrone acetate. They calculated that there was a 6% conversion to estrone. The strongly-acid phenols were partitioned between benzene and sodium mono­ hydrogen phosphate, but the benzene-insoluble fraction (accounting for about half of the estrogenic activity) could not be crystallized. If estriol were present, it would have been in this fraction.

Until about 1948, measurement of the degree of estrogen inactivation or conversion was carried out solely by means of various bioassays. The classical Allen-Doisy method (91) was most popular, and consisted of determining the degree of vaginal carnification in ovariectomized rats. Increase in uterine weight was also used (10). Other methods were not quantitative, such as reddening of the sex-skin in monkeys (23), testicular atrophy in rabbits (17), or estrus occurence in castrate fema1e rats (12). By 1952, however, Su1man (92) bad announced a high1y sensitive method of intravagina1 assay,

- 18 - whieh eould deteet as little as 0.1 milliug. of estrone, 0.05 milliug. estriol and 0.005 milliug. estradiol. Even with sueh extreme sensitivities, there were still the dis­ advantages of expense, lengthiness and laek of speeifieity. If the extraction methods had been sueh as to separate out all the estrogens (ineluding the yet-unknown newer ones) laek of specifieity would not have been a problem. Also, eaeh determination involved a great number of injections in a large series of assay animals, thus causing difficulties in accurate quantitative evaluation (93). Finally, many of the newer estrogens bad little or no biological activity.

The idea of colorimetrie assays was conceived as far back as 19311 when Kober (94) outlined his elassical eolour reaction. Marrian (95), in 19301 had noticed that when sulfuric acid was added to an estrone solution, an orange colour with greenish fluorescence resulted. This latter observation was the basis for the fluorometric assays developed twenty to thirty years later. The biggest dis­ advantage of the early chemieal assays was their insuffieient sensitivity. In the case of urine, this meant working with larger volumes, with resultant interference from non-specifie ehromogens. Lieberman et al (93) measured the yellow eolour produeed when phenolie estrogens were eoupled with tetrazotized dianisidine (96), and were able to eonfirm earlier in vitro work. This method permitted measurement of between 10 and 150 ug. of estrogen, whieh was still not

- 19 " too sensitive.

In 19501 Engel et al (97) first used countercurrent distribution followed by fluorometric analysis in the determination of urinary estrogens. Three years later, Ryan and Engel (98) ineubated estrone and estradiol with rat liver slices, extracted with hot acetone and ethyl acetate, partitioned between pentane and 90% methanol, saponified and lowered to pH 9 1 and finally extracted the estrogens with ether. This relatively pure extract was then subjected to a 24-tube countercurrent distribution in 50% methanol/carbon tetraehloride. The estrone, with a partition coefficient (K) of 0.369 in this system, stayed in the first tvelve tubes, vhile the estradiol (K = 2.26) was recovered from tubes 14 to 24. Fluorometric analysis showed that interconversion of the estrogens readily occurred, to the extent of 0.1 - 14~, depending on the initial hormone concentration (always rather high, from 0.006 to 0.1 mg. hormone/mg. liver/ml. buffer). Ryan and Engel (99) also reported a 2 - 6% conversion of estradiol to estrone by cholangitic human liver, along with interconversion in a wide variety of other human tissues. This led them to question the concept that the liver was the only, or chief site of estrogen metabolism. No conversion of either hormone to estriol vas observed, and estriol was almost quantitatively recovered after incubation with testes and term placenta, bence if it were formed by these tissues in any detectable quantity

- 20 - (at least 5 ug.) from estrone or estradiol, it would bave been detected.

The next major breakthrough came in the use of radio­ active isotopes. Administration of an isotopically labelled estrogen or suspected estrogen percursor, and recovery of another estrogen with the label still attached, would be conclusive proof of conversion. Establishing the pathway of conversion was somewhat of a problem, but the certainty of radiochemical purity was much more difficult. This meant demonstrating that the label was definitely attached to the estrogen in question, and not to a contaminant with similar physico-chemical properties. It was necessary to show that the label and the estrogen "stayed together" after numerous procedures, such as: chromatography, radioautography, counter­ current distribution, derivative formation and repeated crystallization. The method of proving this estrogen-label kinship was by demonstration of a constant specifie activity (counts per min./mg. estrogen) throughout all these procedures. If most or all of the radioactivity was due to a contaminant, there would be a marked decrease in specifie activity once this impurity had been removed.

Beard et al (100) studied estrogens labelled with rl31, in 1949. Unfortunately, the label was not part of the normal molecular structure, it had a tendency to come off the estrogen, and it caused loss of biological activity. Szego (101)

- 21 - ineubated estrone~l6-c 14 , and Szego and Roberts (102) ineubated estradiol-16-cl4, with rat liver sliees in homo­ logous serum. The cl4 was reeovered in the serum proteins, and was only released (in an ether-soluble form) after partial protein hydrolysis. The substances bound to the protein were possibly still steroidal and were slightly estrogenic.

The liver was the only tissue active in promoting protein~ binding. The enzyme-system invovled here was studied by Riegel and Mueller (103). Jellinck (104) incubated rat liver slices with estrone-l6-cl4 and extracted ether-insoluble products, but only a small amount was protein-bound.

In 1954, Heard and O'Donnell (105) injected 409 mg. of cholesterol-C14 (57 million cpm) into a pregnant mare, but were unable to demonstrate any radioactivity in the urinary estrone, although such activity was present in the pregnane and allopregnane derivatives.

Wotiz et al (106), in 1955, were able to recover labelled

estrone, estradiol, progesterone, ~4-androstene-3,17-dione and testosterone, after ineubating acetate-l-cl4 with a testicular embryoma. The following year, they (107) recovered estriol-cl4 after incubation of testosterone-4-Cl4 with ovarian tissue, taken from a post-menopausal woman.

Doisy et al (108) provided isotopie evidence from Cantarow's enterohepatic hypothesis, by administering estrone-16-Cl4 to rats and finding that most of the cl4 was

- 22 " excreted in the bile. Sandberg and Slaunwhite (109) adminis­ tered c14-labelled estrone and estradiol to women, and ~ound

50% o~ the radioactivity in the bile and 7% in the ~eces. This explained the slow urinary excretion (so% o~ the cl4 was recovered in the urine, but over a 4 - 5 day period). Twombly and Levitz (110) obtained the same results using estrone-16-c14 sul~ate.

Beginning in 1955, Diczfalusy et al (111-115) published a series of articles on estrogens in the newborn baby. They reported that large amounts of estriol were exereted in new­ born boys, decreasing from an average of 7292 ug./liter urine/24 hours on the second day post-partum, to 70 ug. by the fifth day. Estriol was identified by formation of the 3-methyl ether, and 5 - 10% was non-conjugated. Estrone was measurable on the second day, but only traces were left by the third day. Human fetal liver was round to contain an average o~ 15 ug./kg. estrone, 30 ug./kg. estradiol and 346 ug./kg. estriol. Estriol was recoverable in vast quan­ tities ~rom the meconium or newborn boys: from 100 mg./kg. on the second day post-partum, decreasing to 0.5 mg./kg. by the sixth day. When 500 ug. of estradiol were administered to newborn boys, estriol, but almost no estradiol or estrone, was reeoveràble in the urine. From these experimenta, Dicz­ falusy et al concluded that estrogen metabolism in the newborn differs markedly from that in the adult. The urinary estriol excretion levels were later confirmed by Staemmler (116).

- 23 - In 195~, Engel et al (117) incubated human fetal liver with

estradiol-16-cl4 and achieved a 3.8% conversion to estriol1 maintaining their usual rigorous standards of radiochemical purity.

Levy and Fish (118) extended the observations of Ryan and Engel (98) by perfusing rat liver with estrone and recover­

ing estradiol-17~and estradiol-17p. Later, Hagopian and Levy (119) recovered labelled estriol after incubating rat

liver brei with estradiol-16~al4.

Dowben and Rabinowitz (120) incubated estradiol-16~cl4 with male adult human liver homogenate and claimed that there

was conversion to both estrone and estriol. Unfortunately1 paper chromatography was the only technique used to determine radiochemical purity, and the radioactivity levels of estriol were hard.ly ab ove background.

Bagget et al (122) incubated testosterone-4-cl4 with grossly normal ovarian slices from pre-menopausal women with carcinoma of the cervix. They reported a 1% conversion to estradiol, with no formation of labelled estrone or estriol.

Engel (123), in 1957, published a review on the biosyn~ thesis of estrogens, discussing androgenie origins, and deriving a possible reaction sequence for the formation of estradiol from testosterone (Fig. 1).

- 24 - r------·

OH

0 \

0

-- J ------·-- - - - · ------~ · - - · ------· ------· --

Figure 1. Hypotbetical reaction sequence for the formation of estradiol from testosterone (123).

- 25 - Engel quoted West et al (124), who administered testos­ terone to castrate, adrenalectomized women, and were able to demonstrate an increased urinary excretion of estrone and estradiol. This added weight to the idea that non-endocrine organs can perform this conversion in the human. Engel (123) administered estradiol to the same type of subjects, and recovered urinary estriol of the same order of magnitude as in normal women. He quoted Axelrod and ~1111er (125), and Schneider and Mason (126) as demonstrating the ability of animal livers to introduce 16a-hydroxy groups into suitable steroids (androgens in these cases).

Ryan (127), in 195~, reported another androgen-estrogen conversion. He recovered estriol after incubating

~5-androstene-3,16~,17p-triol with human placental microsomes in the presence of TPNH. It began to appear that estriol could be directly derived from androgens as an alternative pathway, though probably not the same androgens which led to the formation of estrone and estrarliol. Baggett et al (128) inoubated testosterone-4-Cl4 with stallion testes, human term placenta and human adrenal cortical carcinoma, and were able to recover labelled estrone and estradiol but not estriol. In the case of the placenta, initially it seemed that the estriol was labelled and it re~uired a large number of' counter­ current distributions in different systems before the hormone was separated from the label. Ryan (129), in 19591 proposed a scheme for estriol formation from C-16-hydroxylated neutral steroids (Fig. 2).

- 26 - ---OH ---OH

0 HO

OH

---OH ---OH ---OH

Figure 2. Estriol formation from C-16"hydroxylated neutral steroids (129).

- 27 - Longchampt e~ al (130) were able to trace a pathway from /). 4-androstene-3, 17-dione-cl4 to 19-hydroxy- ~4 -androstene- 3,17-dione-c14, and from there to estrone-c14, using human placenta! incubation.

Heard et al (121), on the basis of their work and others, derived a metabolic scheme relating androgens, estrogens, acetate and cholesterol:

~-----ACETATE------~ ! CHOLESTEROL------~

ESTRA.DIOL~ES'fRON~~---19-HYDROXY ... _64 -A.NDROSTEN.I!:-3 1 17-DIO.N~ ! . ~ f ESTRIOL TESTOSTERONE Â4 -ANDROSTEN~-3,17-DIONE

Savard et al (131) administered acetate-1 ... c 14 to a pregnant mare and analysed the activity of the estrone-c14 and allopregnanolone-C1 4 (a progeste-rone metab.olite) recovered in the urine. They hypothesized the pathway: acetate-----~ cholesterol~progesterone~l7«-hydroxyprogesterone------~ testosterone~estrogens. If they were derived from acetate, 10 carbon atoms of progesterone and 9 carbon atoms of estrone would be theoretically labelled, and this was found to occur.

As early as 19451 Masson and Hoffmann (132) had shown that the liver was capable of inactivating progesterone (comparing the progestational response after administering the hormone by gavage and subcutaneously). Later work (133) produeed numerous pregnane metabolites, but there was no attempt to

... 28 - look for estrogen conversion in either experiment.

Mitchell and Hobkirk (134) incubated estradiol~l6-C 14 with liver slices from a laying hen, and were able to demon­ strate a 6% conversion to estriol. They established radio" chemical purity, using paper chromatography, followed by radioautography, countercurrent distribution, and derivative formation followed by alumina chromatography, maintaining a constant specifie activity.

DeMeio et al (135) were able to form estrone-sulfate by incubating estrone, magnesium chloride, potassium sulfate and adenosine triphosphate (ATP) with ox or rabbit liver, in the presence of Krebs-Ringer Phosphate buffer. Using the same conditions, they obtained estradiol-sulfate from estradiol.

Jellinck (136) recovered a labelled water-soluble, ether-insoluble fraction, after incubation of estrone-16-cl4 with rat and human liver slices, and therefore he postulated the formation of non-steroid metabolites.

Nothing has been said yet about the role of the newer estrogens. Though much has been done in vivo and with non­ hepatic incubation systems, Breuer et al (137-150) were the only group to do extensive studies on the role of the liver in the biosynthesis and metabolism of these estrogens. Instead of working with radio-isotopes, they coupled the estrogens to Fast Black Salt K and made use of paper chromatography and

- 29 ~ microsublimation techniques. Breuer et al (1371 138) incubated

16~~hydroxyestrone and 16-ketoestradiol-17} with human liver slices, recovering estriol from both incubations, 17-epiestriol from the former, and 16~epiestriol from the latter incubation. Quantitatively, they found that the 17-keto group was pre­ ferentially reduced to a 17ft-hydroxy group (estriol) instead of a 17Q-hydroxy group with a ratio of 15:1, and the 16-keto group was preferentially reduced to a 16~-hydroxy group (2:1 ratio). It was shown that 16p-hydroxyestrone was equally metabolized in human or rat liver, but the resulting 16-epi­ estriol was formed in 400% greater quantity by the human liver (139). In the above incubation, 16-ketoestradiol-1~ was also formed, possibly from oxidation of the 16-epiestriol.

In 1958-591 Breuer et al (140,141) incubated 16-keto­ estrone with normal human liver and ovarian slices. Eight metabolites were recovered, 6 of which were characterized: (the percentage conversion is in parentheses, hepatic first, followed by ovarian) 16-ketoestradiol-17p (24%, 26%), 16p-hydroxyestrone (12%, 10%), 16-epiestriol (lü%, 6%),

16~hydroxyestrone (1%, 8%), estriol (6%, 2%) and

17-epiestriol (0.2%, 1%). Later, 16,17~epiestriol was shown to be a hepatic metabolite of 16-ketoestradiol-17~ and of 16J-hydroxyestrone (142).

Since all 4 triols (16oc,l7p; 16ft,l7p; 16~1 17~; 16p,l7d) were recovered from the urine of pregnant women by Breuer (143)

- 30 - in 1960, they can all be considered as normal in vivo metabolites. From the work already mentioned, a slightly revised scheme is shown, illustrating the metabolism of estrogens in adult human liver (Fig. 3).

Breuer et al also studied the formation of the 6-hydroxy and 6-keto estrogens (144, 145). Incubating estrone or estradiol with rat liver tissue produced 6p~hydroxyestradiol-17p. Incubation of this latter estrogen, under the same conditions, resulted in the formation of 6J-hydroxyestrone and 6-keto­ estradiol-17p. No 16-hydroxy or 16-keto compounds were isolated in either incubation.

Work was carried out on the relative activities of the enzymes concerned in the intermediate metabolism of the 16,17-substituted estrogens, after incubation with rabbit liver slices (146). It was found that 17p-dehydrogenase was more active than 16a, or 16p-dehydrogenase, while 17«-dehydrogenase was the least active of allo A scheme of the reactions is shown (Fig. 4). Similar work, incubating estradiol and estrone with rat liver slices, led to the conclusion t.hat 6-hydroxylase attacks both estradiol and estrone, while 16~-hydroxylase only acts on estradiol in this tissue (147).

In 1960, Breuer et al (148) were able to recover estriol and ~-hydroxyestradiol-17p after incubation of estradiol with human fetal liver, thus confirming and extending the work of

~ngel et al (117). Breuer had previously reported the same

- 31 - ~-- -- - .

r------~~------.tl.------· 1 1 1 1 1 1 /' Ào ~ -, 1 1 / ~ ' 1 _, / ~ 16-KETOESTRONE '',, 1 t --/ / r~ .... / ,;:0 ~~~~ J_T 16c<.- HVDROXYESTRONE 16- KETOESTRADIOL-17fi "';S~H v ESTRIOL 16-EPIESTRIOL

Figure 3. The metabolism or estrogens in adult human

liver. Solid arrows indieate known reactions1 broken arrows indicate postulated reactions. Bxcept for estradiol, only ring D of each estrogen is shown (141).

- 32 .. P - ---OH ----- P-OH tro 18-EPIE STRIOL ESTRIOL 16-KETOESTRADIOL-17f3 t 1r ~---OH iJo &OH 16 ~ -HYDROXYESTRONE 16 d..- HYDROXYESTRONE 16- KETOESTRONE l O,H O,H p1 - OH iro p-OH 16-KETOESTRADIOL-17lX 16,17-EPIESTRIOL 17-EPIESTRIOL

Figure 4. The intermediate metabo1ism of the 161 17-sub­ stituted estrogens in rabbit 1iver. Thick arrows indicate the 1argest percentage con- versions, whi1e broken arrows indicate postu1ated reactions. Only ring D or each estrogen is shown (146).

... 33 - transformations in rat liver (149). In the same period, Breuer and Pangels (150) carried out incubations with normal rabbit liver sliees, and demonstrated interconversion between estradiol-17C(, estrone and estradiol-17p.

Recently, evidence has accumulated to prove that estriol ean be further metabolized. King (151) incubated estriol with rat liver, and was able to recover 2-hydroxyestrio1 and 2-methoxyestriol in good yield. When there was no methyl donor in the incubation medium, only 2-hydroxyestriol was found. His work led him to suggest the following metabolic pattern:

DPN or TPN ~striol------~~ 2-hydroxyestriol plus oxygen Methionine, ATP Magnesium ions j and oxygen 2-methoxyestriol

He later found that rat-liver microsomes, plus the supernatant fraction, ean produce small amounts of 16-keto­ estradiol-17}) from estriol (152). MacRae et al (153) injected estriol-16~c 14 into a laying hen, and recovered labelled l6-ketoestradiol-17p and 16-~piestriol in the urine, and l6~epiestriol in large quantities in the feces.

Axelrod and Goldzieher (154) incubated various tissues with 2-hydroxyestradiol-17p plus DPN, TPN, DPNH, ATP and

- 34 ~ dl"methionine-methyl-c14. They were able to recover labelled

2~methoxyestradiol~l7J and 2-methoxyestrone after incubation with normal hurnan kidney (but after incubation with neoplastic kidney, only 2-methoxyestradiol-17p was recovered). They could not demonstrate any conversion after incubation with liver or adrenal, and therefore they considered that these tissues are not capable of carrying out this type of trans- methylation reaction.

- 35 - PU.tti?OSE OF THIS STUDY

This work was carried out to resolve four problems in areas where little or no studies had been done previously, or where earlier work had produced negative results.

These problems are: 1- The possibility of estriol biosynthesis from estradiol after incubation with adult human liver. No conversion had been shown to occur in any preceding work. 2- The capability of newborn human liver to form estriol. Prior studies were only with fetal tissue or live newborns. 3- The ability of human liver to synthesize estrogens from androgens or progesterone. All the preceding work had used other tissue or in vivo systems. 4- The capacity of avian liver to form various estrogens after incubation with estradiol. Earlier work demon­

strated conversion to estriol1 but notbing is known about the formation of estrone or the newer estrogens in such a system.

.. 36 .. ~THODS

Incubation 1: The liver was obtained from a 98 year-old female patient, 4! hours after death, and was placed in a deep freeze. Sixty g. of the thawed liver were homogenized in a Waring Blender with 60 ml. of Krebs-Ringer Phosphate buffer (155), pH 7, containing 100 mg.% of glucose. Twenty g. of this homogenate were added to each of two flasks. The first flask also con­ 14 tained 200 ug. of estradiol-16-C (approximately 8761 000 cpm) dissolved in 1 ml. of ethanol, along with 1.65 mg. of unlabelled estradiol. The labelled estradiol was obtained from Charles E. Frosst and Co., with a specifie activity of 2.7 microCuries/mg. The second flask contained 112 ug. of 14 16-ketoestradiol-17p-16-C (approximately 3 1 440,000 cpm) dissolved in ethanol, and 0.91 mg. of unlabelled 16-keto- estradiol-17p. The labelled 16-ketoestradiol-17J was obtained from Dr. Mortimer Levitz, and bad a specifie activity of 21.4 microCuries/mg. The flasks were incubated, with shaking in a water-bath for two hours, at 37 0 C. in air.

After incubation, 1.16 mg. of unlabelled estriol carrier were added to the flask initially containing estradiol as substrate. The contents of the flask were mixed with 50 ml. of cold ethanol, and the mixture was centrifuged. The ethanol layer was poured off, more cold ethanol was added,

- 37 - and the mixture was again centrifuged. The ethanolic extracts were pooled, and evaporated to dryness under reduced pressure. The d.ried extract was th en parti tioned in a fi ve­ tube eountercurrent distribution apparatus, between methanol­ water (70:30) and hexane, in order to remove fat.

The estrogen-containing methanol-water extract was evaporated to dryness, dissolved in 1.5 ml. of ethanol, and partitioned between benzene and water in the same manner as described for the hexane/methanol-water system. In the benzene/water system, estriol has a partition coefficient (K) of approximately 0.18, and is therefore almost quanti­ tatively retained in the aqueous phase. The majority of the other possible estrogen metabolites have much higher K values, and are largely retained in the benzene. The latter phase was therefore assayed for radioactivity and th~n discarded.

The aqueous phase was refluxed with 1 N sodium hydroxide for 30 minutes to achieve complete saponification, then cooled, and extracted once with one-third volume of et~er to remove neutral steroids. The ether was rejeeted and the pH of the aqueous layer was lmiered to 9 by the passage o:f carbon dioxide ~s. Estriol was then extracted with 4 x one-third volumes of ether, and the pooled ether fractions were washed with 1 x one-tenth volume of 8.5% sodium bicarbonate, and twice with water. The vH of the final wash was ascertained to be neutral. The ether was then evaporated to dryness 1

... 38 ... and the radioactivity level was determined.

The general counting procedure was as follows: the dried extract was dissolved in a known amount of ethanol, and 1 ml. of this solution was measured into a clean, stain- less-steel cupped planchette. The contents of the planchette were dried under an infra-red lamp, and the planchette was

placed in a windowless, gas~flow, proportional counter connected to a Nuclear-Chicago SG-186 scaler, with an operat- ing plateau around 1450 volts, and a fairly constant back­ ground radiation level of 30 cpm. Since the radioactivity of the ether extract was well above background, it was sufficient to have the scaler operate until 1000 counts were recorded. In genAral, enough counts were accumulated to give an error of no more than 5%.

The remainder of the ether extract was then subjected to a 25-tube countercurrent distribution, using chloroform: carbon tetrachloride:methanol:water (40:10:35:15). The radioactivity and the estriol levels were determined ~or tubes o, 6 - 18 and 25. From these results, the specifie activity (cpm/mg.) was calculated. ~striol levels were determined by the Kober reaction as modified by Bauld (156), and the Allen correction (157) was used.

Estriol carrier (1.32 mg.) was added after incubation to the flask initially containing 16-ketoestradiol-17p as substrate. The extracti on procedure was the same as

- 39 " described above until the ether extract was o-btained. At this point, descending paper chromatography was perf'ormed, using a Zaf'faroni-type system (158) with formamide as the stationary phase and chloroform as the mobile phase. Two strips were run, and onto each strip an amount of extract containing 36.8 ug. of estriol (and approximately 2760 cpm) was spotted. Beside each of these, 20 ug. of estriol standard were spotted. The system was run for 24 hours at 20°C. A narrow section was eut from each strip and stained by dipping in ~olin-Ciocalteu phenol reagent (159) and exposing to ammonium hydroxide vapour. An x-ray film was left in contact with one of the strips in the dark1 for one month, and a radioautograph was thus obtained. This was compared to the stained section. The steroid areas on the second strip (corresponding to the stained bands on the narrow section) were eluted with ethanol, and evaporated to dryness.

Column partition chromatography, using the Bauld technique

(156), was then carried out on the &~ate. The columns were packed with Celite, using methanol-water (70:30) as the stationary phase, and ethylene dichloride as the mobile phase. The ehBte was evaporated to dryness, and the specifie activity was determined. One-third of the ether extract was subjected to a 25-tube countercurrent distribution, and the contents of tubes 7 - 17 were analysed in order to determine the specifie activity.

- 40 ... The remainder of the ether extract was methy1ated according to the procedure of Brown (160), using dimethy1 sulfate and 20% sodium hydroxide in borate buffer (lowering the pH down to 10 - 11.5, which considerably retards the rate of hydrolysis, without seriously affecting the rate of methylation), at a temperature of 37°C. The estriol methyl­ ether was then dissolved in benzene (since the methylated estriol is no longer preferentially soluble in water) and transferred to a.n adsorption chromatographie column, prepared from standardized, deactivated alumina and water-saturated benzene. Finally, the estriol methyl-ether was eluted with 2.5% ethanol in benzene (12 ml.), and its specifie activity was determined.

An experiment was performed to discover the degree of contamination of the 16-ketoestradiol-l~-16-cl4 by labelled estriol. Fifty microliters of the ketoestradiol solution (containing approximately 43,000 cpm) plus 25 ug. of un­ labelled estriol were spotted onto one strip, 20 ug. of unlabelled estriol were spotted onto a second strip, and a paper chromatogram was run, using the formamide/chloro­ form system. The second strip was stained with Folin­

Ciocalteu phenol reagent plus anœonium hydroxide1 and the corresponding area on the first strip was eluted with ethanol and counted. The ratio of these counts to the original 43,000 cpm indicated the degree of contamination by estriol-cl4 •

..., 41 - Incubation 2: In this second incubation, human newborn liver was used.

A male baby 1 weighing 2000 g ., died at the age of 23 h·ours. Ten g. of liver plus 10 ml. of Krebs-Hinger Phosphate buffer (including 100 mg.% of glucose) and 200 ug. of estradiol-16-cl4

(containing approximately 9401 000 cpm) l'lere incubated together, beginning 6 hours after death. After incubation, 1 mg. of estriol carrier was added. The initial procedures were the same as in the first experiment, except that after incubation the homogenate was refluxed for 30 minutes with ethanol. When cool, this was centrifuged, pooled with the other ethanolic extracts, and evaporated to dryness. Counts were taken of the methanol-water fraction, both the benzene and water fractions, and the ether extract. A Kober reaction was also performed on the ether extract, so that the specifie activity could be determined.

Paper chromatography was carried out on 20% of the ether extract, with a 20 ug. estriol standard being run on a second strip. A radioautograph was made from the paper. The re­ mainder of the ether extract was subjected to a column chromatogram, and the specifie activity was determined. The eluted estriol was methylated, and the specifie activity of the estriol methyl-ether was determined. The detailed steps in each of the procedures mentioned in this paragraph were the same as in the first incubation.

... 42 ... The estriol methyl-ether ~vas then subjected to a 24-tube countercurrent distribution, using a methanol: water:carbon tetrachloride (25:25:50) solvent system (161). In this system, estriol metl1yl-ether has a partition coef­ ficient of 1.3.

Two-hundred and fifty ug. of a standard estriol-16-cl4 solution, containing 16,300 cpm, were methylated; and the resulting estriol methyl-ether was subjected to a counter­ current distribution as described in the preceding paragraph, in order to have a standard distribution pattern for compari­ son. In both distributions, specifie activities were determined for all tubes havin~ sufficient radioactivity and estrogenic content.

Incubation 3: In this experiment, a human, male, newborn liver was again used, the birth weight was 2900 g., and death occurred at 6 days of age. Incubation was begun 4i hours after dea.th, using 112 ug. of 16-ketoestrad.iol-17p-16-Cl4 (containing approximately 3,400,000 cpm) along with all the other materials and conditions as were used in previous incubations. The source and activity of the labelled estrogens were the same for all of the first three experiments. At the end of incubation, 1 mg. of unlabelled estriol was added as carrier.

The extraction procedures were the same as above and the radioactivity of all the fractions was determined, along

- 43 - with the specifie activity of the ether extract. Paper chromatography was carried out, using 20% of the ether extract, and from this a radioautograph was prepared.

Incubation 4: This experiment was carried out to determine whether human liver could convert testosterone or progesterone to estrogens in vitro. Two incubation systems were set up. The first contained 10 g. of adult human liver, 10 ml. of Krebs-Ringer Phosphate buffer (including lOO mg.% of glucose, 0.003 M fumaric acid, 0.003 M DPN and 0.003 M adenosine monophosphate) plus ~7 ug. of testosterone-16-C14 (with a level of 2 1 8001 000 cpm) having a specifie activity of 75 microCuries/mg. The second system contained 30.8 ug. 14 of progesterone-16-C (with a level of 1 1 1401 000 cpm) having a specifie activity of 65 microCuries/mg. In other respects, this system was the same as the first. The two steroids were obtained from the Radiochemical Centre, Amersham, U.K. After 2 hours in a shaking incubator at 370 c., 400 ug. of estrone and 750 ug. of estriol carrier we re added to each flask.

The preliminary extraction procedures, down to the benzene/water partition, were the same as in the second and third experiments. Since all three classical estrogens were of interest in this experiment, both the benzene and water phas e s were retained. The benzene phase wa s separated into

... 44 -

. - ··-·· - -- ··- --~------.1 an estradiol and an estrone fraction by evaporating to dry­ ness and transferring with 3 x 1 ml. of benzene to a Bauld partition col1w.m chromatogram (156) 1 using 0.8 N sodium hydroxide as the stationary phase and benzene as the mobile phase. The first 10 ml. of mobile phase were discarded and the next 30 ml. contained the estrone fraction. At this point, a 3:1 (v/v) mixture of ethylene dichlorine and benzene (60 ml., not equilibrated with a stationary phase) was then added to each reservoir. The elœ.te then collected contained the estradiol fraction.

All four fractions (two from each incubation system) were then evaporated to dryness and counted. Fjve hundred ug. of unlabelled estrnniol carrier were added to each of the estradiol fractions. All four fractions were then dissolved in 50 ml. of tolnene, and each was extracted with 2 x 50 ml. of 1 N sodium hydroxide, to separate the phenolic and neutral compounds. Acidification to below pH 9 was accomplished by the passage of carbon dioxide gas tllrough alkaline solutions. The estrogens werethen extracted with ether, washed with sodium bicarbonate and water, and evaporated to dryness. The specifie activity of each of the ether extracts was determined.

Methylation of tPe two estrone fractions was carried out 1 and the specifie activity of the two estrone methyl-ethers was determined.

The water phase from the benzene/water partition, con-

- 45 - taining the estriol, was next dealt with. Hot saponification, acidification and ether extraction were performed on both of these fractions (from the two incubation systems) in the same manner as described previously, and the specifie activities were determined.

Both estriol fractions, and the estradiol fraction from the progesterone incubation, were subjected to Bauld partition column chromatograms (156). The ehatee were collected and their specifie activities were determined. Finally, a 24-tube countercurrent distribution was performed on the estriol fraction from the testosterone incubation, with the specifie activity being determiried from tubes 6 - 14.

Incubation 5: In this final experiment, the liver of a laying hen was used. Incubation was be~m 45 minutes after death, using 5 ml. of liver homogenate, 5 ml. of Krebs-Ringer Phosphate buffer (containing lOO mg.% of glucose) and 3641 000 cpm of estradiol-16-cl4 • The procedure was the same as be~ore. On the completion of incubation, several carriers were added: estriol (125 ug.), estrone (100 ug.), 16-epiestriol (20.4 ug.), 2-methoxyestrone (96 ug.), 16-ketoestradiol-17f (100 ug.) and 16a-hydroxyestrone (79.2 ug.). Ethanolic extraction and hexane/methanol-water partition were carried out as in the previous incubations.

The methanol-water phase was evaporated to dryness and a

- 46 ~ Girard reaction (162) was performed at room temperature in order to separate the ketonic and non-ketonic components. After ether extraction, the non-ketonic estrogens were subjected to hot saponification, lowered to pH 9 with carbon dioxide, and re-extracted with ether. The ketonic fraction was acidified with 7 N sulfuric acid, and then extracted with ether. Both fractions were distilled to dryness, dissolved in specially purified methanol, and then spotted onto Celite at 750 c.

The extracts were next applied to Celite partition columns (162). The column used for the separation of the ketonie estrogens was prepared as follows: hexane-benzene (98:2) was equilibrated with methanol-water (70:30) at l6°C., and allowed to stand overnight. Ten g. of Celite were mixed with 10 ml. of stationary phase (methanol-water) and sufficient mobile phase to form a slurry. A small amount of asbestos was placed in the bottom of the column to prevent elution of the Celite, and the slurry was loosely packed to a level of 12 cm. The extract in the dried Celite was next added to the top of the column, the tube being rinsed three times with mobile phase which was also added to the column. A little fresh Celite was added and gently packed down, and 50 ml. of mobile phase were run through the column. The first 10 ml. of eluate were discarded, and the next 40 ml. contained the 2-methoxyestrone. At the time of solvent equilibration, hexane-benzene (50:50) was also equilibrated with methanol-

- 47 " water (70:30). One-hundred and thirty ml. of hexane­ benzene (50:50) were added to the column. The first 40 ml. of eluate contained the estrone fraction, and the next 90 ml. contained the ketolic fraction.

The column used for separation of the non-ketonic estrogens was prepared in the same rnanner, except for the amounts and concentrations of the solvants used. Hexane­ benzene (45:55) and pure benzene were both separately equili­ brated with methanol-water (70:30). Sixty ml. of hexane­ benzene (45:55) were added to the column, the first 10 ml. of eluate discarded, and the remaining 50 ml. constituted the estradiol fraction. At this point, 150 ml. of pure benzene were added, the 16-epiestriol corning off in the first 45 ml., and the estriol in the remaining 105 ml.

All six fractions were evaporated to dryness, and the specifie activity of each was determined.

The estrone fraction was extracted with 4 x 40 ml. of ether, and the ether extract shaken with 2 N sodium hydroxide (7.6 ml.). The aqueous phase, without removal from the separating funnel, was partially neutralized with molar sodium bicarbonate (30.4 ml.) and the phases were shaken again. The ether ·extract was next washed with 7.6 ml. of molar sodium bicarbonate, and then with 5 ml. of water, and distilled to dryness. The specifie activity was measured after the extract bad been run through a Hauld partition column.

- 48 - The estriol fraction was dissolved in 1.5 ml. of ethanol, transferred to a separating funnel with a total of 25 ml. of benzene (10, 10 and 5 ml.) and the estriol extracted with 2 x 25 ml. and 2 x 12.5 ml. of water. Saponification and ether extraction were then carried Ollt as before, the extract was run through a Hauld partition column, and the specifie activity was determined.

At this point, 108 ug. of 16«-hydroxyestrone and lOO ug. of 16-ketoestradiol-17f were added to a 40% aliquot of the ketolic fraction. To this, 140 mg. of sodium borohydride were added (to reduce the ketolic fraction to the correspond­ ing triols), the contents were dissolved in 24 ml. of ethanol

( to bring the final ethanol concentration to 20~o), and the flask was left overnight in the dark. The œxt day, con­ centrated acetic acid was added carefully to the solution, until effervescence ceased (indicating that neutrality was obtained). The estrogens were extracted with ether, care­ fully washed with 8.5% sodium bicarbonate until effervescence ceased, washed twice with 5 ml. of water, and taken to dryness. The ether extract was finally chromatographed on the type of column described on p.48, the 16-epiestriol and estriol fractions were collected, and their specifie activities were measured.

As far as the remainder of the originRl six fractions were concerned, 2-methoxyestrone was not dealt with any

- 49 - further because of insufficient activity. The 16-epiestrio1 was subjected to a 24-tube countercurrent distribution, using a benzene/water so1vent system, in which 16-epiestrio1 has a partition coefficient of 3.4. The contents of tubes 2, 4, 6, 8, 10 and 12 - 24 were ana1ysed for radioactivity, and for 16-epiestrio1 content.

- 50 - RESULTS

Incubation 1: A: Estradiol-16-cl4 as substrate.

The radioactivity o~ the ether extraet amounted to a total or 4460 epm. The results of the countercurrent distribution are shown in Table I and Fig. 5. The average specifie activity±standard deviation, of tubes 9- 151 vas 6160 ± 733 epm/mg., vhieh vould correspond to an 0.82% conversion from estradiol to estriol.

B: 16-ketoestradiol-17p as substrate.

At the stage o~ the ether extraet, there vere 400 ug. of estriol and 301 000 epm present, giving a specifie activity of 75,000 cpm/mg. Tvo bands were visible on the paper chromatogram. The ~irst was 1.5 - 2 cm. and the second vas 10.2 - 12.2 cm. from the starting line. The second band corresponded to the estriol standard. There vas a single darkened strip on the radioautograph, eorresponding to the second band on the paper (Fig. 6).

After column ehromatography, using the eluate of the estriol band on the second paper, 16.8 ug. of estriol vere recovered, with a radioaetivity level of 494 epm, eorres­ ponding to a specifie aetivity of 29,000 epm/mg.

" 51 " The countereurrent distribution (using one-third of the ether extract) is graphically displayed in Fig. 7, and the results are listed in Table II. On the basis of the contents of tubes 10 - 16, the average specifie activity was 30,000 ± 1770 epm/mg. After alumina adsorption ehromatography of the estriol methyl-ether, the total quantity of estriol methyl-ether eluted was 109.4 ug., and the total radioactivity was 2650 epm. Thus, the specifie activity vas 24,200 cpm/ag., which corresponds to a conversion of 0.93%.

In the experiment to determine the percentage contam­ ination of the original l6-ketoestradiol-17p-16-cl4 with labelled estriol, there were 75 cpm left after elution and counting of the estriol band. Since there vere 431 000 cpm initially, this would correspond to an 0.17% contamina­ tion, giving an actual percentage conversion of 0.76~.

Incubation 2: After the hexane/methanol-water partition, there were 495,500 cpm left in the methanol-water fraction. At the completion of the benzene/water partition, both fractions were counted. The benzene contained 3551 000 cpm, and the water contained 15,350 cpm. There were 121 325 cpm in the ether extract, along with 302 ug. of estriol, thus indicating a specifie activity at this stage of 41,000 epm/mg.

- 52 - The radioautograph band did not correspond to the estriol band from the paper ehromatograph.

There vere 152 ug. of estriol eluted from the eolumn, assoeiated vith a total of 1440 epm. Thus the specifie aetivity vas now redueed to 9500 epm/mg. After methylation, the specifie aetivity of the estriol methyl-ether vas

8100 epm/mg. (13~ ug. of estriol methyl-ether and 1113 epa). The results of the countereurrent distribution (performed after elution from the column and methylation) are shown in Table III and Fig. 8.

When labelled estriol standard (250 ug. and 161 300 epm) vas methylated, the resulting estriol methyl-ether contained 11,000 cpm. The findings from the countercurrent distribu­ tion are shown in Table IV.

Incubation 3: After the hexane/methanol-vater partition, there vere

29,000 cpm left in the hexane1 and 3,2201 000 cpm left in the methanol-water. At the eompletion of the benzene/vater partition, there vere 2 1 3501 000 cpm left in the benzene, and

3451 000 epm in the vater. The specifie aetivity of the ether extraet was 181 000 cpm/mg. (11,000 epm and 610 ug. of estriol). Radioautograpby demonstrated close agreement of the radio­ activity and the estriol band (Fig. 9).

- 53 - Incubation 4: A: Testosterone-16-c14 as substrate. The "estrone" fraction from the benzene/0.8 N sodiua hydroxide co1umn contained 621 000 cpm and 224 ug. of estrone.

The "estradio1" fraction contained 121 000 cpm. Since estradiol carrier had not yet been added at this stage1 the Kober reaction was not carried out. The "estrone" ether extract had a specifie activity of 5000 cpm/mg. (112 ug. of estrone and 560 cpm). The~stradio1" ether extract had a much lower specifie activity equa1 to 1300 cpm/mg. (365 ug. of estradio1 and 475 cpm) and therefore no further procedures were carried out on this fraction. The "estrone" fraction was methylated and the resu1ting estrone methy1-ether had almost no activity left (67 ug. of estrone methy1-ether and

60 cpm1 corresponding to a specifie activity of 900 cpm/mg.).

At the stage of ether extraction1 the "estriol" fraction had a specifie activity of 1700 cpm/mg. (612 ug. of estrio1 and 1040 cpm). After co1umn chromatography1 the specifie activity showed an unexpected increase to 3300 cpm/mg. (425 ug. of estriol and 1420 cpm). Countercurrent distribu­ tion at this stage gave the resulta shown in Table V and Fig. 10.

B: Progesterone-16-cl4 as substrate. The radioactivity levels of the "estrone" and "estra­ diol" fractions after the benzene/0.8 N sodium hydroxide

- 54 - oolumn vere 38,000 and 30,000 cpm respectively. At the stage of the ether extraction, the "estrone" fraction bad a specifie activity of 12,000 cpm/mg. (70 ug. of estrone and 840 cpm), the "estradiol" fraction contained 389 ug. of estradiol and 7400 cpm equalling a specifie activity of 19,000 cpm/mg., and the "estriol" fraction bad a specifie activity of 3000 cpm/mg. (483 ug. of estriol and 1450 cpm).

The estrone was methylated and the estrone methyl-ether had no significant activity left (50 cpm). After column chromatography, the estriol bad only 180 cpm and the estradiol only 80 cpm.

Incubation 5: The activities of the six estrogens after column ehromatography are shown in Table VI. After reduction of the ketolic fraction vith sodium borohydride and subject­ ing the triols thus formed to column chromatography, nearly all the radioactivity disappeared, the estriol component containing 40 cpm and the 16-epiestriol containing 110 cpm. The specifie activity vas decreased nearly 200 times, from 1,196,000 epm/mg. for the ketolic fraction to only 1200 and lOOOcpm/mg. for the estriol and 16-epiestriol reduction components.

There vas some decrease in the specifie activity of the estriol and estrone fractions after they were subjected to Bauld partition chromatography. The estriol was redueed

- 55 - to 12,400 cpm/mg. (48.3 ug. and 600 cpm) and the estrone to 115,100 cpm/mg. (44.0 ug. and 5065 cpm).

Countercurrent distribution of the 16-epiestrio1 fraction gave the resulta shown in Table VII and in Figs. 11 and 12.

- 56 - DISCUSSION

In practise, proor of radioehemical purity is the most difficult problem in isotopie tracer work of this type. There are no absolute standards, no point - short or degradation procedures - at which one ean say beyond all doubt that the label is definitely attaehed to the estrogen in question.

It is sometimes easier to establish the opposite situation, demonstrating that the label is distinctly separate from the estrogen, and no conversion has taken place. But caution must be used. In the countercurrent distribution of the 16-epiestriol rrom Incubation 5 (Fig. 11), it is evident that the main radioactivity peak is at tube 22, 4 tubes apart from the estrogen peak at tube 18. However, there is a second radioaetivity hump (one-quarter the height of the primary one) with its peak at tube 18, whieh might well represent labelled 16-epiestriol. Re-scaling the graph provides us with two almost, but not quite, superimposable eurves (Fig. 12). It eould even be argued that the contaminant, providing most or the total cpm, is the sole cause of the lack of perfeet fit. Thus we have come nearly full cirele, rrom a seemingly sure negative result to a likely, but not certain, positive one.

- 57 ~ In the case of the ketolic fraction from the same incubation, a slightly dirferent picture emerges. With •ost of this fraction lost after column chromatography, but a very appreciable number of cpm remaining, this resulted in an incredibly high specifie activity, greater than tbat of the estradiol at the same stage. It required further procedures to show that the counts were very nearly all due to contamination. Again, one cannot state that there was no conversion, since, if the counts remain­ ing with the now-reduced ketolic fraction vere not !!!! due to contamination, the yield was over 1%. With only a total of 150 cpm left, further procedures were out of the question. Insutficiency of counts was also encountered vith 2-methoxyestrone, but conversion was quite doubtful here, and at best would be less than 0.2%. There was an unexpectedly lov activity associated with estriol, since Mitchell and Hobkirk (134) bad achieved a 6% yield after fairly rigorous purification, using the same incubation system.

Althougb the liver of many species bas been shown to be capable of converting estradiol to estrone, this was not previously attempted with avian liver. On the basis of only a slight decrease in specifie activity after column chromatography in two different systems, it seems likely that this transformation does occur in the laying hen (maximum conversion of 3%).

- 5e - lt is difficult to say when1 if ever1 a percentage conversion could be regarded as insignificant. Breuer recovered only 0.2 ug. of 17"epiestriol after incubation vith lOO ug. of 16~ketoestrone (141) 1 while Ryan (163) demonstrated a 40 " 60% conversion of estrone to

~4-androstene-3,17-dione. From one point of view1 it would be possible to say that the latter experiment vas 300 times more significant than the former. However, both investigations are of equal importance in elucidating new metabolic pathways.

Although countercurrent distribution was the only procedure performed on the estradiol extract in the first incubation, the two curves showed close coineidence, and 0.82% conversion to estriol may be assumed. This is the first time that such a reaction has been shown to occur in adult human liver. The 16-ketoestradiol-1~ extract exhibited a fairly constant specifie activity throughout several dif­ ferent procedures. The countercurrent distribution patterns ve~not nearly as impressive this time, but in toto there is good evidence that there vas conversion to estriol.

The first experience with newborn liver (Incubation 2) posed a problem of pattern-interpretation. The formation of the methyl-ether could possibly take place on other than the aromatic 3-hydroxyl group, which may explain the biphasic estrogen curve. However, the radioaetivity pattern is a

- 59 - complete enigma, and with the exception of a slight cor­ respondance betveen tubes 7 and 11 (vhich may be coinciden­ tal) there is complete dissociation from the estrogen. The estrogen results vere not just due to technieal error, sinee exaetly the same pattern oeeurred using a methyl-ether derived from an estriol standard. The faet that the radio­ autograph was also negative strongly suggests that conver­ sion was nil.

One vould have expeeted the degree of conversion to lie between the 3.~% that Engel noted vith fetal liver, and the o.s2% found earlier in this work vith the adult specimen.

On the basis of radioautography alone, the estriol was labelled in Incubation 3. However, this eannot be eonsidered a suffieient test of radioehemieal purity by itself, and - sinee no further procedures were earried out - it is impos­ sible to drav any definite conclusions.

After analysis of the results of the fourth incubation, it ean be fairly safely concluded that progesterone is not eonverted to estrogens by adult human liver in vitro. Unless it were possible to obtain starting material of very high specifie aetivity, one eould not be absolutely certain of this.

The testosterone incubation was less straightforvard. The estrone was fairly certainly unlabelled, sinee there was

- 60 - a steep drop in specifie activity after each stage, until a point vas reached where the radioactivity was only slightly above background. Sinee estradiol carrier vas not added until after the first ehromatography, the specifie aetivity was only determined after ether extraction in this case, and there vere too few eounts to proeeed any further. Unfortunately, it is impossible to say that conversion did not oecur, since there was some radioactivity left. The estriol fraction poses the knottiest problem, beeause it is theoretically ineonceivable that the specifie activity eould inerease unless outside contamination oeeurred, or the steroid was broken into a smaller molecule, in whieh case it could no longer reaet with the Kober reagent. Even contamination (unless it resulted during the determination of the specifie actiYity) would not explain the inordinately high reading in tube 9 of the countereurrent distribution (Table V), sinee it vould spread itself out over several tubes. If the tube 9 radio­ activity reading is disregarded, there is nov some corres­ pondance with the estrogen curve, though not suffieient to draw any definite conclusions.

In summary, none of the fractions isolated from Incu­ bation 4 eould be definitely shown to be derived from the starting material.

- 61 - SUMMARY

1. Normal adult human liver is capable o~ eonverting estradiol and 16.. ketoestradiol-17p to estriol. 2. This capacity is either decreased or absent in the

case o~ newborn human liver. 3. The adult liver does not seem to be able to metabo­ lize progesterone or testosterone to estrogens. 4. The liver of a laying ben, on the basis of this and previous work, transforma estradiol into estriol, estrone and possibly 16-epiestriol. The ketolie estrogens and

2-methoxyestrone were, i~ present at all in this final incubation, in too small a concentration to be properly determined. "

.. 62 .. TABLE I

Countercurrent distribution or the ether extract f"rom Incubation 1...A:

Tube Number Estrio1 cp• s2ecific Activitl (ug.) ( cpm/mg.)

0 40 2 28

4 15

6 2.6 30 7 a.o 43 7100 8 12.3 88 7200 9 21.0 108 5200 10 30.5 178 5800

11 38.5 217 5600 12 35.5 228 6400 13 36.0 210 5800 14 25.3 190 7500 15 19.5 133 6800 16 10.3 95 9200 17 4.0 38 9500 18 1.8 25 20 9

22 9

25 0

- 63 ... TABLK ll

Counterourrent diBtribution (using one-third ot the ether extract) froa Incubation 1-B:

Tube Nuaber &~triol cpa S~eoific ActiYitl (u,;.) (cpa/ag.) 1 245 3 200

~ 210 7 0 236

8 4.0 244

g ~.1 398 10 14.4 478 33,000

11 17.8 515 2Q1 000 12 24.0 728 30,000 13 19.1 535 27,000 loi 17.0 490 29,000 15 10.0 310 31,000 16 5.7 174 31,000 17 o.e 157 19 lOO 21 80

23 70 25 15

- 64 - TABLE III Countercurrent distribution (after e1ution from the eo1umn and methy1ation) from Incubation 2:

Tube .Nu.mber Estrio1 cpm S~eeific Aetivitl aeililt-ether (cpm/mg.) (ug.) 1 21 2 10 3 11 4 8 5 0 18 6 1.0 18 7 2.5 10 4,000 8 7.5 18 2,400 9 11.4 38 3,300 10 15.9 55 3,500 11 10.8 40 3,700 12 14.6 73 5,000 13 10.5 93 8,900 14 8.6 143 16,600 15 2.2 70 31,800 16 2.1 115 55,300 17 0 70 18 0 53 19 0 4:8 20 35 21 46 22 18 23 18 .. 65 • 24 5 TABLE IV

Countereurrent distribution of the 1abe11ed estrio1 standard (Incubation 2):

Tube .Number Estrio1 epm SEeeifie Aetivitl methyl-ether (epm/mg.) (ug.) 7 5.2 585 1131000 8 9.0 1065 118,000 9 14.0 1360 97,000 10 20.5 1630 80,000 11 10.1 1675 166,000 12 16.2 830 51,000 13 11.6 825 71_.000 14 7.3 660 90,000 16 1.8 175 97,000

18 0.6 20 33~000

- 66 .. TABLE V

Countercurrent distribution of the estrio1 fraction (after eo1umn chromatography) from Incubation 4-A:

Tube ~umber Estrio1 epm SEeeifie Activitl (ug.) (epm/mg.) 6 9.4 63 6700 7 15.7 68 4300 8 32.0 123 3800 9 54.3 298 5500 10 52.6 160 3000 11 56.9 128 2300 12 41.5 98 2400 13 32.5 45 1400 14 18.5 48 2600 16 3.6 0

- 67 - TABLE VI

Aetivities of the six estrogens from Incubation 51 after eo1uan ehromatography:

Estrogen epm sEeeifie Activitz ~ (epm/mg.) Estrio1 87.2 2,220 25,500 Es trone 73.0 11,350 155,000 Estradio1 219.0 181,800 830,000 Keto1ic fraction 7.7 9.,210 1,196,000 2-Metho:xyestrone 33.3 270 8,100 16-eEiEstrio1 27.5 3,840 139,600

.. 68 .. TABLE VII

Countercurrent distribution of the 16~epiestrio1 fraction from Incubation 5:

Tube Number 16-e~iEstrio1 cpm S~ecific Activitl ug.) (cpm/mg.) 2 0.84 7

4 0.70 9 6 Oe70 4 8 0.70 7 10 0.28 9 12 0.56 23 13 1.12 20 17,900 14 2.38 33 13,900 15 4.35 49 11,300 16 7.01 65 9,200 17 10.23 106 10,400

18 13.49 110 8•140 19 5.61 81 14,400 20 5.61 210 37,400 21 4.77 334 70,100 22 2.67 443 166,000 23 1.68 437 262,000

24 1.12 168 150.000

- 69 - •• E8TRIOL .., C.P.M. 36 234

08

28 182

24 le6 !.

220 OOp a: :v 1- c.o !'= 1&1 16 104

78

52

4 26

4 8 12 16 20 TUBE NUMBER

Figure 5. Countereurrent distribution of the ether

extract from Incubation 1~.

- 70 ... Figure 6. Radioautograph, ·using the paper chromatogram from Incubation 1-B. The area between the two dark 1ines, on the paper strip at the 1ef\ corresponds to the estrio1 band.

- 71 - ••• ESTRIOL ooo C.P. M. 2

24 720

C!) ~2 ..J 2 ~ 16 Cl) "'

• 4

0 4 8 12 16 TUBE NUMBER

Figure 7. Countercurrent distribution (using one•third of the ether extract) from Incubation 1-B.

.. 72 .. - ESTftiOL METHYL-ETHER 18 --o-- C.P.M. 182

16 IÎ Il 1 1 14 1 1 126 1 \ • 1 1 1'\ 1 1 1 \ ..,a: 1 1 1 \ % \ 1 \ l­ Ill 1 1 \ 1 .J \ 1 \ >­ 0 % v \ 1- 1 . \ .., 1 ::Il 1 \ \ \ 1 •• .J 1 \ ' ...... 2 \ 1 \ /. a: -J \ / \ :;; 4 v \ .., \ \ 2 \...._ Il ' ' / ·--· \ \ '·--•--t~' \ 0 12 20 24 TUBE HUMBER

Figure s. Countercurrent distribution (after elution from the column and methylation) fro• Incubation 2.

.. 73 • Figure 9. Radioautograph, using the paper chromatogram from Incubation 3. The area between the two dark lines, on the paper strip at the left, corresponds to the estriol band.

~ 74 " -- ---. - .. - -- - - ··- ...• - ·-· -·-· - .. ···· ··- ·-- · ...... ·-·--

80 170

--- ESTRIOL 54 -~-C.P.M. 2

{ 126 1 1 108 1 1 1 1 1 1 1 ,..P o"'

18

12

6

0 4 12 20 24 TUBE NUMBER

Figure 10. Countercurrent distribution of the estriol fraction (after column chromatography) from Incubation 4-A.

- 75 .. ------·· · -···-- ·--- ... .

--MS-EPIESTRIOL --o-- C.P.M. 80

o--o 14 ,' 1 420 1 1 1 1 1 1 12 1 1 360 •1 '1 1 1 c; 10 / 1 300 :::l.. 1 1 1 1 1 1 Q 8 1 1 a: J 1 ...co ~ 6 'll80 ILl 1 !! 120

2 60 o------o----- 0 12 16 24 TUBE NUMBER

Figure 11. Countercurrent distribution of the 16-epi­ estrio1 fraction from Incubation 5.

- 76 - - 16-EPIESTRIOL --o-- C.P.M.

112 r" 1 1 96 1 1 1 1 a•o 1 1 .J 1 2 8 0: 1i ln 1 1 Ill 1 ~ 6 ,o 11.1 1 1 • ~ 1 1 4 l 32 1 /~-ol / 2 / / ----G-....., __ -o" 0 ... -..o---­0 0 8 20 TUBE NUM8ER

Figure 12. Countercurrent distribution of the 16-epiestrio1 fraction from Incubation 5. The radioactivity curve is expanded so that the peak in tube 18 corresponds to the 16-epiestriol peak in the same tube. Thus, the higher cpm in tubes 20 - 24 fal1 outside the graph.

- 77 - • BIBLIOGRAPHY

1~ Zondek, B., Skand. Arch. Physiol. 70, 133 (1934). 2- Zondek, B., Laneet 227, 356 (1934).

3~ Israel, S.L., Meranze, D.R., and Johnston, C.G., Am. J. Med. Sei. 194, 835 (1937). 4- Engel, P. and Navratil, E., Biochem. Ztsehr. 292, 434 (1937).

5" Parker, F. and Tenney, B., Endocrinology 231 492 (1938). 6- Helier, C.G., Helier, E.J. and Severinghaus, E.L.,

Proc. Am. Phys. Soc. April 261 1939, p.ll4. 7- Helier, C.G., Endocrinology 26, 619 (1940). 8- Helier, C.G. and Heller, E.J., Endocrinology 32, 64 (1943). 9- Fieser, L.F. and Fieser, M., Steroids, Waverley Press Ine., Baltimore, Md., p.467 (1959).

10- Talbot, N.B., Endocrinology 25 1 601 (1939). 11- Pincus, G. and Martin, D.W., Endocrinology 27, 838 (1940). 12- Biskind, G.R. and Mark, J.,

Bull. Johns Hopkins 65 1 212 (1939). 13- Biskind, G.R., Endocrinology 28, 894 (1941).

14- Biskind, M.S. and Biskind, G.R., Science 941 462 (1941).

15- Biskind, M.S. and Biskind, G.R. 1 Endocrinology 31,. 109 (1942).

16- Biskind, G.R. and Meyer, M.A.,

Proe. Soc. Ex~tl. Biol. Med. 53, 91 (1943).

17... Biskind, 1-1. s. and Biskind, G.R., Endoerinolo~l 32, 97 (1943).

18- Biskind, M.s. and Biskind, G.R., Endocrinolo~l 30, 819 (1942).

.. 78 .. 19" Golden, J.B. and Severinghaus, E.L.,

Proc. Soc. Ex~tl. Biol. Med. 391 361 (1938). 20- Shipley, R.A. and Gyorgy, P.,

Proc. Soc. Ex~tl. Biol. Med. 571 52 (1944).

Gyorgy, P., Proc. Soc. Ex~tl. Biol. Med. 60, 344 (1945). 22- Segaloff, A. and Nelson, w.o.,

Proc. Soc. Exptl. Biol. Mede 481 33 (1941).

23~ Hooker, C.W., Drill, V.A. and Pfeiffer, C.A., Proc. Soc. Exptl. Biol. Med. 65, 192 (1947). 24- Drill, V.A. and Pfeiffer, C.A.,

Endocrinology 381 300 (1946).

25~ Bernstorf, E.c., Endecrinology 49, 302 (1951). 26- Vanderlinde, R.E. and Vesterfeld, w.w.,

Endocrinologz 471 265 (1950). 27- Jailer, J.W. and Seaman, L.,

Proc. Soc. Exptl. Biol. Med. 731 70 (1950). 28- Vasington, F.D., Parker, A., Headley, w. and Vanderlinde,

R.E., Endocrinology 621 557 (1958). 29- Bernstorf, E.c., Proc. Soc. Exptl. Biol. Med. 69, 447 (1948). 30- Mayer, G. and Soumireu, J., C.R. Soc. Biol. Paris 142, 964 (1948). 31" Canivenc, R. and Mayer, G., C.R. Soc. Biol. Paris 145, 713 (1951). 32- Grayhack, J.T. and Scott, w.w.J., Endocrinology 48, 453 (1951).

- 79 - 33- Kirgis, H. and Rothchild, 1., Endocrinology 50, 269 (1952). 34- Ferret, P., Rev. lnt. Hepatol. 2, 503 (1952). 35- Wenner, M.R.,

Bull. Fed. Soc. Gyne~. Obstet. 4, 865 (1952). 36- Kullander, s., Acta Endocrinol. (Kbh) 17, Suppl. 22, 1 (1954).

37- Kullander, s., Acta Endocrinol.- (Kbh) _21, 204 (1956). 38- Kullander, s., Acta Endocrinol. (Kbh) 23, 131 (1956). 39- Cantarow, A., Rakoff, A.E., Paschkis, K.E. and Hansen, L.P., Proc. Soc. Exptl. Biol. Med. 49, 707 (1942). 40- Cantarow, A., Rakoff, A.E., Paschkis, K.E., Hansen, L.P. and Walkling, A.A., Endocrinology 31, 515 (1942).

41- Cantarow, A., Rakoff, .A.E., Paschl

42- Rakoff, A.~., Cantarow, A., Paschkis, K.E., Hansen, L.P.

and Walkling, A.A., Endocrinology 341 370 (1944).

43- Paschkis, K.~., Cantarow, A., Rakoff, A.E., Hansen, L.P. and Walkling, A.A., Proc. Soc. Exptl. Biol. Med. 55, 127 (1944).

44.. Koch, F.C., in Heard, R.D.H.,

Rec. Prog. Hormone Research 4 1 25 (1949).

45- Cantarow, A., Paschkis, K.~. and Rakoff, A.E., Science 101, 558 (1945).

46- DeMeio, R.H·., Rakoff, A.E., Cantarow, A. and Paschkis, K.E.,

~ndocrinology 43, 97 (1948). 47- Coppedge, R.L., Segaloff, A • ., Sarett, H.P. and Altschul.,

A.M • ., J. Biol. Chem. 1731 431 (1948).

- 80 - 48- Coppedge, R.L., Segalofr, A. and Sarett, H.P., J. Biol. Chem. 182, 181 (1950). 49- Twombly, G.H. and Taylor, H.c., Cancer Research 2, 811 (1942).

50~ Tenney, B. and Parker, F., J. Clin. Endocrinol. 2, 293 (1942). 51" Singher, H.o., Kensler, C.J., Taylor, H.C., Rhoads, C.P.

and Unna, K., J. Biol. Chem. 1541 79 (1944). 52- Unna, K., Singher, H.O., Kensler, C.J., Taylor, H.C. and Rhoads, C.P., Proc. Soc. Exptl. Biol. Med. 55, 254 (1944). 53- Benoze, E., Mag. Noorv. Lapja 18, 248 (1955). 54- Engel, P. and Rosenberg, E., EndocrinoloQ: 37, 44 (1945). 55.. Levy, H., Arch. Biochem. 14, 325 (1947). 56- Ledogar, J.A. and Jones, H.W ., Science 112, 536 (1950). 57- Repke, K. and Markwordt, F., Naturwissenschaften 41, 258 (1954). 58- Repke, K. and Markwordt, F., Arch. Exp. Path. Pharmak. 223, 271 (1954). 59- Repke, K. and Markwordt, F., Arch. Exp. Path. Pharmak. 224, 341 (1955).

60- Zondek, H., Minerva Ginec. 31 188 (1951). 61- Glass, s.J., Edmundson, H.A. and Soll, S.N., Endocrinology 27, 749 (1940). 62- Glass, S.J., Edmundson, H.A. and Soll, S.N., J. Clin. Endocrinol. 4, 54 (1944). 63- Gilder, H. and Hoagland, C.L. Proc. Soc. Exptl. Biol. Med. 61, 62 (1946).

- 81 .. 64- Lloyd, c.w. and Williams, R.H., Am. J. Med. 4, 315 (1948). 65- Bennett, H.s., Haggenstoss, A.H. and Butt, H.R., Am. J. Clin. Path. 20, 814 (1950). 66- Cameron, C.B., J. Endocrinol. 15, 199 (1957). 67- Muller, J., Acta Endocrinol. (Kbh) 28, 205 (1958). 68- Stoa, K.F., Bassoe, H.H. and Emberland, R., Acta Endocrinol. (Kbh) 28, 357 (1958). 69- Abrizouri, M. and Gomez, M.F., Revista de Investigacion Clinica, Mexico 2, 41 (1950). 70- Zondek, B. and Black, R., J. Clin. Endocrinol. 7, 519 (1947). 71- Brzezinski, A. and Sulman, F.C., Harefuah 42, 118 (1952). 72- Pincus, I.J., Rakoff, A.E., Cohn, E.M. and Tumen, H.J.,

Gastroenterology 191 735 (1951). 73- West, C.D., Tyler, F.H., Brown, H. and Samuels, L.T.,

J. Clin. Endo~rinol. 11, 897 (1951). 74- Kinsell, L.W., Margen, s., McCallie, U.P. and Michaels, G.D., Stanford Med. Bull. 11, 46 (1953). 75- Worter, J., Schwartz, J. and Weill, J.P., Schweiz. Med. Wschr. 83, 467 (1953). 76- Charbonnier, A. and Clement, D., Rev. Int. Hepatol. 2, 114 (1952). 77- Ho1mes, W.N., Acta Endocrinol. (Kbh) 23, 39 (1956).

78- ~ngel, P., ~ndocrinology 291 290 (1941). 79- Zondek, B. and Sklow, J., Proc. Soc. Exptl. Biol. Med. 46, 276 (1941).

- 82 " 80- Pincus, G., Cold Spring Harbour Syœposia in Quant. Biol. 5, 44 (1937). 81- Pincus, G. and Zahl, R.A., J. Gen. Physiol. 20, 879 (1937). 82- Beard, R.D.H., Bauld, w.s. and Dorfman, R.I., J. Biol. Chem. 141, 709 (1941). 83- Fish, W.R. and Dorfman, R.I., J. Biol. Chem. 143, 15 (1942). 84- Doisy, E.A., Thayer, S.A. and Van Bruggen, J.T., Federation Proc. 1, 202 (1942). 85- Lipschutz, A., Becker, c., Mello, R.F. and Riesco, A., Science 101, 410 (1945).

86- Schiller, J. and Pincus, G., Science 98 1 410 (1943). 87- Schiller, J. and Pineus, G., Endocrinology 34, 203 (1944).

88- Schiller, J., ~ndocrinolosz 36, 7 (1945). 89- Ericksen, s. and Velle, w.,

Acta Endocrinol. (Kbh) 341 27 (1960). 90- Pearlman, W.H. and DeMeio, R.H.,

J. Biol. Chem. 1791 1141 (1949). 91- Doisy, E.A., in Sex and Interna! Secretions by E. Allen, Williams and Wilkins, Baltimore, Md., p.846 (1939).

92- Sulman, F.G., Endocrinology 501 61 (1952). 93- Lieberman, s., Tagnon, H.J. and Schulman, P., J. Clin. Invest. 31, 341 (1952).

94- Kober, s., Biochem. Ztschr. 2391 209 (1931).

95- Marrian, G.F., Biochem. J. 241 1021 (1930).

- 83 - 96- Bender, A.E. and Wilson, A., Biochem. J. 142, 741 (1942). 97- Engel, L.L., Slaunwhite, R., Carter, P. and Nathanson, I.T., J. Biol. Chem. 185, 255 (1950). 98- Ryan, K.J. and Engel, L.L., Endocrinology 53, 277 (1953). 99- Ryan, K.J. and Engel, L.L., Endocrinology 53, 287 (1953).

100- Albert, s., Beard, R.D~H., Leblond, C.P. and Saffran, J., J. Biol. Chem. 177, 247 (1949). 101- Szego, C.M., Endocrinology 52, 669 (1953). 102- Szego, C.M. and Roberts, s., J. Biol. Chem. 221, 619 (1956). 103- Riegel, I.L. and Mueller, G.c., J. Biol. Chem. 210, 249 (1954). 104- Jellinck, P.H., Bioehim. Biophys. Acta 41, 37 (1960). 105- lleard, R.D.H. and O'Donnell, V.J., Endocrinology 54, 209 (1954). 106- Wotiz, H.H., Davis, J.W. and Lemon, H.M., J. Biol. Chem. 216, 677 (1955). 107- Wotiz, H.H., Davis, J.w., Lemon, H.M. and Gut, M., J. Biol. Chem. 222, 487 (1956).

108- Valcourt, A.J., Thayers, S.A., Doisy Jr., ~.A., Elliott, W.H. and Doisy Sr., E.A., Endocrinology 57, 692 (1955). 109- Sandberg, A.A. and Slaunwhite, W.R., J. Clin. lnvest. 36, 1266 (1957). 110- Twomb1y, G.H. and Levitz, M., Am. J. Ob. Gyn. 80, 889 (1960).

111~ Diczfa1usy, E.,

Symposium deut. ges. ~ndokrino1. 3, 147 (1955).

R 84 " 112... Diczfa1usy, E.~ Ti11inger, K. and Westman, Ae,

Acta Endocrino1. (Kbh) 261 303 (1957). 113- Diczfa1usy, E., Ti11inger, K. and liestman, A.,

Acta Endocrino1. (Kbh) 261 313 (1957). 114- Diczfalusy, E., Tillinger, K. and Westman, A., Acta Endocrinol. (Kbh) 28, 169 (1958). 115- Diczfa1usy, E., Menini, E., Ti1linger, K. and Westman, A.,

Acta Endocrinol. (Kbh) 301 539 (1959).

116" Staemm1er, H., Klin. Wochschr. 381 97 (1960).

117- Engel, L.L., ~aggett, B. and Balla, M., Bioehim. Biophys. Acta 30, 209 (1958). 118- Levy, L. and Fish, C.A., in The Hormones, Ed. G. Pincus

and K.V. Thimann, Academie Press, ~.Y., Vol. III, p.600 (1955).

1198 Hagopian, M. and Levy, L., Biochim. Biophys. Acta 301 641 (1958). 120- Dovben, R.M. and Rabinowitz, J.L.,

~ature 178, 696 (1956). 121- Heard, R.D.H., Bligh, E., Conn, M., Jellinck, P.,

1 0 Donne11, V., Rao, B. and Webb, J.1 Rec. Prog. Hormone Research 12, 45 (1956). 122- Baggett, B., Engel, L.L., Savard, K. and Dorfman, R.I.,

Federation Proc. 141 175 (1955).

123" Engel, L.L., Cancer 101 711 (1957). 124" West, C.D., Damast, B.L., Sarro, S.D. and Pearson, O.H., J. Biol. Chem. 218, 409 (1956).

- 85 .. 125" Axelrod, L.R. and Miller, L.L., J. Biol. Chem. 219, 455 (1956). 126- Schneider, J.J. and Mason, B.L., J. Biol. Chem. 172, 771 (1948). 127" Ryan, K., Endocrinology 63, 392 (1958). 128" Baggett, B., Engel, L.L., Balderas, L., Lanman, G., Savard, K. and Dorfman, R.I., Endocrinology 64, 600 (1959). 129" Ryan, K., J. Biol. Chem. 234, 2006 (1959). 130- Longchampt, J., Gual, c., Ehrenstein, M. and Dorfman, R.I., Endocrinology 66, 416 (1960). 131- Savard, K., Andrec, K., Brooksbank, B.W.L., Reyneri, c., Dorfman, R.I., Beard, R.D.H., Jacobs, R. and Solomon, s.s., J. Biol. Chem. 231, 765 (1958). 1328 Masson, G. and Hoffmann, M.M., Endocrinology 37, 111 (1945). 133- Taylor, w., Biochim. Biophys. Acta 15, 592 (1954). 134- Mitchell, J.E. and Hobkirk, R., Biochem. Biophys. Researeh Communications 1, 74 (1959). 135- DeMeio, R.H., Lewycka, c., Wizerkaniuk, M. and

Salciunas, o., Biochem. J. 681 1 (1958). 136- Jellinck, P.B., Bioehem. J. 71, 665 (1959). 137- Breuer, H., Nocke, L. and Knuppen, R., z. Physiol. Chem. 311, 275 (1958). 138- Breuer, H., Nocke, L. and Knuppen, R., Bioehim. Biophys. Acta 33, 254 (1959).

- 86 - 139- Breuer, B., Knuppen, R. and Nocke, L., Biochem. J. 71, 26P (1959). 140- Breuer, H. and Knuppen, R., Nature 182, 1512 (1958). 141- Breuer, H., Knuppen, R. and Pange1s, G., Acta Endoerino1. (Kbh) 30, 247 (1959). 142- Breuer, H. and Nocke, L., Biochim. Biophys. Acta 36, 271 (1959). 143- Breuer, H., Nature 185, 613 (1960). 144- Breuer, H., Nocke, L. and Knuppen, R.,

~aturwissensehaften 45, 397 (1958). 145- Breuer, H., Nocke, L. and Knuppen, R., z. Physio1. Chem. 315, 72 (1959). 146- Ureuer, H., Knuppen, R. and Pange1s, G., z. Physio1. Chem. 317, 248 (1959). 147- Breuer, H., Noeke, L. and Pangels, G., Acta Endoerinol. (Kbh) 34, 359 (1960). 148- Breuer, H., Knuppen, R., Ort1epp, R., Pange1s, G.

and Puck, A., Biochim. Biophys. Acta 401 560 (1960). 149- Breuer, H. and Knuppen, R., Biochtm. Biophys. Acta 39, 408 (1960). 150- Breuer, B. and Pangels, G., Acta Endocrino1. (Kbh) 33, 532 (1960). 151- King, R.J.B., Biochem. J. 74, 22P (1960).

152- King, R.J.B., Biochem. J. 76, 7P (1960). 153- MacRae, H.F., Dale, D.G. and Common, R.B., Can. J. Biochem. Physio1. 38, 523 (1960).

- 87 - 154- Axelrod, L.R. and Goldzieher, J.w., J. Clin. Endocrinol. Metab. 21, 211 (1961). 155- Krebs, H.A. and Eggleston, L.V., Biochem. J. 34, 442 (1940). 156- Bauld, w.s., Biochem. J. 63, 488 (1956). 157- Allen, W.M., J. Clin. Endocrinol. 10, 71 (1950). 158- Zaffaroni, A. and Burton, R.B.,

J. Biol. Chem. 1931 749 (1951). 159- Folin, o. and Ciocalteu, v•, J. Biol. Chem. 73, 627 (1927). 160- Brown, J.B., Biochem. J. 60, 185 (1955). 161- Diczfalusy, E., Acta Endocrinol. (Kbh) 20, 216 (1955). 162- Givner, M.L., Bauld, w.s. and Vagi, K., Biochem. J. 77, 400 (1960).

163- Ryan, K., Biochim. Biophys. Acta 271 658 (1958).

- 88 -