Evidence for a Two-Pool System Governing the Excretion of Radioactive Urinary Estrogen Conjugates During the First Eight Hours Following the Administration of Estrone-6,7-3H to Male Subjects. Probable Role of the Enterohepatic Circulation CAROLYN M. HOWARD, HUGH ROBINSON, FREDERICK H. SCHMIDT, JAMES R. McCORD, AND JOHN R. K. PREEDY Departments of Medicine and Biometry, Emory University School of Medicine, Atlanta, Georgia 30322 ABSTRACT. The presence of an enterohepatic circulation for estrogens in the human is well established, but its possible function as an anatomic pool in which estrogens are distributed and from which they may be transferred has not been explored. To obtain evidence for a pool effect attributable to the enterohepatic circulation, a single dose of estrone-6,7-3H was administered to 4 normal male subjects, and urine was collected in hourly periods for 8 hr. Samples were analyzed for radioactive unconjugated estrone, and for radioactive sulfates and glucosiduronates of estrone, estradiol-17/3 and estriol. Radioactive estrone glucosiduronate was found to be quantitatively the most important conjugate excreted during the 8-hr period, exceeding the excretion of the other radioactive conjugates

I

T IS fully established that estrogens undergo an enterohepatic circulation in man, and certain aspects of this system have been well studied (1-5). Thus, the nature and concentration of estrogens in bile have been investigated (2-4), and the metabolism of estriol and estriol conjugates by the intestine has been intensively explored (5, 6). However, there is little information regarding the functioning of the enterohepatic circulation as a whole. The circulation has finite dimensions, and must presumably behave to some extent as a pool1 in Received April 15,1969. Supported in part by Research Grant AM 02221 and Training Grant T 15040 from the National Institute of Arthritis and Metabolic Diseases, and by Grant FR 00231 from the Special Research Resources Branch of the NIH. Address requests for reprints to Dr. John R. K. Preedy, Department of Medicine, 69 Butler Street, S.E., Atlanta, Ga. 30303. 1 In the present context, the term "pool" is used to describe some subunit of total body fluid; the

by a factor of 10. The excretion of each estrogen was plotted against time, and the curves so obtained were fitted using a mathematical model consisting either of 1 or the sum of 2 gamma density functions. Unconjugated estrone which does not enter the enterohepatic circulation was excreted in the form of 1 gamma density function, suggesting one pool of distribution. In contrast, the various estrogen sulfates and glucosiduronates which do enter the enterohepatic circulation were usually excreted in the form of the sum of 2 gamma density functions, suggesting an additional pool of distribution for these conjugates. Several considerations suggest that this additional pool is likely to be the enterohepatic circulation. (J Clin Endocr 29: 1618, 1969)

which estrogens are distributed and from which they are transferred into the systemic circulation. Such transfer should contribute to the concentration in the systemic circulation of all estrogens which undergo enterohepatic circulation, and this should in turn be reflected in the rate of urinary excretion of these estrogens. If the pool effect of the enterohepatic circulation is substantial, it should be possible to detect and characterize it by appropriately designed tracer experiments using sequential urine analyses. To test for such an effect estrone-6,7-3H was given to a group of normal males, and all urine excreted thereafter was collected in hourly periods for eight hours. Each urine sample was analyzed for labeled unconjugated estrogens, and for the labeled forms of a number of estrogen conjugates known to occur in the bile (2-6). Amounts relevant dimension is therefore volume and not mass.

1618

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 January 2017. at 11:11 For personal use only. No other uses without permission. . All rights reserved.

December 1969

ESTROGEN EXCRETION IN MALES

1619

3 scribed (8). Authenticity of estrone-6,7H. The 3 original sample of estrone-6,7- H was submitted to column partition chromatography at intervals during the study and the purity established as described by Crowell, Howard and Preedy (8). Aliquots of the solution injected were similarly tested in each case. Specificity of methods. Although the methods Materials and Methods used this study for hydrolyzing estrogen sulFour normal males in the age range 20-28 fates in and glucosiduronates have been were selected. There was no evidence of any extensivelyestrogen employed by ourselves (8, 12) and disease process or any endocrine abnormality. by others (9-11), and are thought to be specific There was no gynecomastia. During the period for sulfates and for glucosiduronates, respecof study the subjects carried out their normal tively, they do not distinguish between the activities and ate a normal diet. 3 sulfates of any one estrogen, where more than About 10 juCi authentic estrone-6,7- H was sulfate is possible, or between glucosidurodissolved in 0.1 ml 95% aqueous ethanol and one nates of any one estrogen, where more than one diluted to 25 ml with 0.9% w/v aqueous NaCl glucosiduronate is possible. Although there is with sterile precautions. Exactly 20 ml of this at present no evidence for the natural occursolution was injected intravenously into each of rence in the human of estrogen sulfates other the male subjects at 8:00 AM (time zero—Table than the 3-monosulfates, the occurrence of more 1; Fig. 2-7). An aliquot of the remaining soluone glucosiduronate of an estrogen is well tion was taken for radioactivity measurement than for example, estriol-3 glucosiduroto determine the amount of radioactivity in- established; and estriol-16 glucosiduronate. Consejected, and for further analysis (see below). nate quently, in the present study the moieties reThereafter, all urine excreted was collected for ferred "estradiol-17/3 glucosiduronate" and 8 consecutive hourly periods (7 in the case of "estriolto as glucosiduronate" may include more subject AL), and stored at —15 C until analyzed. than one glucosiduronate. On the other hand, Each urine sample was analyzed for radioactive the designations "estrone sulfate" and "estrone unconjugated estrone, estradiol-17/3 and estriol, glucosiduronate" refer, respectively, to one and for radioactive estrone glucosiduronate, substance only, since estrone has only one estrone sulfate, estradiol-17/3 glucosiduronate, hydroxyl group. It should be noted that estradiol-17/3 sulfate, estriol glucosiduronate estrogen sulfoglucosiduronates and n-acetyl gluand estriol sulfate (for meaning of the terms cosaminides are not measured by the present "sulfate" and "glucosiduronate" in the present methods. context, see below). Identification of parent estrogens. Identification of radioactive urinary unconjugated esMethods of analysis. All urine samples from the trone, and of radioactive estrone and estradiolsame subject were analyzed together, to min- 17/3 and estriol released by the hydrolysis proimize random variation associated with the cedures was carried out as described by Crowell, analytical procedures. Turner, Schmidt, Howard and Preedy (12). Unconjugated estrone, estradiol-17fi and estriol. Recoveries. Using the present methods, it was The urine samples were adjusted to pH 9 and found that recoveries of nonradioactive estrone extracted with ether, followed by the standard sulfate added to urine were in the range 74-82 % method of Preedy and Aitken (7), as described (using solvolysis), and of estradiol-17/3 17-gluby Crowell, Howard and Preedy (8). Estrone, cosiduronate in the range 67-80 % (using gluestradiol-17(3 and estriol glucosiduronates. One curonidase hydrolysis). half of the aqueous phase from the above separation was used. Glucuronidase hydrolysis was Physical model. The hypothesis to be tested carried out, followed by the standard method states that the enterohepatic circulation funcof Preedy and Aitken (7) as previously de- tions as an additional pool for estrogen conjuscribed (8). Estrone, estradiol-17/3 and estriol gates, and that the effects of this pool can be sulfates. The other half of the aqueous phase detected by sequential urine analysis. An apfrom the separation referred to above was used. propriate physical model describing this hySolvolysis was carried out as described by pothesis is depicted in Fig. 1. In the present Crowell, Howard and Preedy (8), followed by context, a pool is considered to have finite the standard method of Preedy and Aitken (7). dimensions and to be identifiable anatomically. Radioactivity determinations. These were carried A chemical compartment represents the site or out using a Packard Model 500 liquid scintilla- sites at which some chemical transformation tion counter under the conditions already de- occurs. It is dimensionless and unidentified

excreted were plotted against time, and a mathematical model consisting either of one or the sum of two gamma density functions was fitted to the curves so obtained.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 January 2017. at 11:11 For personal use only. No other uses without permission. . All rights reserved.

HOWARD ET AL.

1620 NPUT E

E

S y s t e m ic

Compartment

C i r c u 1 a t i on EC

Pool I

EC A

j

EC R EC B


0, is approximately the fraction of the input to the channel that leaves the last stage at a time between t and t + At after instantaneous injection into the first stage. Let gi(t) denote this function for the quantity aQ of input E that has already been defined separately for the models in Fig. 7 and 8, and let g^(t) denote this function for the corresponding quantity bQ. Then gi(t)At and g2(t)At are the fractions of the quantities aQ and bQ, respectively, that leave the system of 2 parallel channels at a time between t and t + At after entrance of the quantity Q of input E into the system. From this, the total quantity of E leaving the system between t and t + At is seen to be [aQgi(t) + bQg2(t)JAt (equation 7) This can be abbreviated by f (t) At, so that f(t) = aQgi(t) + bQgs(t) (equation 8) which is clearly a non-normalized density function satisfying f(t) >0 on 0