Metabolism of the Adrenal Cortical Hormones

A . CATABOLISM; URINARY EXCRETORY PRODUCTS AS INDICES OF ADRENOCORTICAL FUNCTION

While relatively little information has been gained concerning the individual fate of the many steroids which have been isolated in crystal-line state from the gland (Table III, page 557), cortical function has clearly been associated with the excretion of a large number of metab-olites, many of which are of undetermined chemical structure. These may be classified in two groups: (1) those which may be quantitatively determined by the application of relatively simple colorimetric or bio-assay techniques (and in this group fall an uncertain number of cortical metabolites of unknown structure, detectable only by virtue of a specific biological or chemical property), and (2) metabolites of established chemical constitution, estimable gravimetrically only after a usually tedious process of isolation. The neutral 17-ketosteroids of urine occupy an intermediate position, in that the whole may be determined colori-metrically while the principal components of the mixture are known chemical substances which may be estimated individually by isolation.

In nearly all instances, the connection between urinary excretion and cortical function has been arrived at through the rise or fall in the output of a specific metabolite or group of metabolites (possessing a common property) under experimental or clinical conditions or hyper- or hypo-adrenocortical function.

The following laboratory reactions, applicable to the water-insoluble, lipide-soluble, neutral products of urine, give a measure of the excretion of metabolites classified under group 1 but reflect cortical functions of quite different kinds (Table X X I I , page 604) : (a) 17-ketosteroids (Table X X I V , page 608), (6) cortin, (c) reducing substances, (d) 17-ketosteroids generated on periodic acid oxidation of the non-ketonic alcohols of urine, and (e) formaldehyde generated on periodic acid oxidation of the neutral fraction of urine. And in group 2⁵, the following urinary excretory prod-ucts (Table X X V , page 613) have definitely been associated (Table X X I I I , page 606) with cortical function: (/) A⁵-androstene-3(/3),17(a;)-diol (CXLVIII), (g) A⁶-androstene-3(/3),16,17-triol (CXLIX), (h) A⁵ -pregnene-3(0),2Oa-diol (CL), (i) 3(<x),20û:-diol (CLI), (j) pregnane-3(«),17,20-triol (CLII), and (k) pregnane-3(a),17-diol-20-one (CLIII).

6 T o the list of adrenal metabolites should be added A5 -pregnene-3(/3),17"j3"-diol-20-one (formula 1 below) isolated (5.3 m g . / l . ) and characterized b y Hirschmann and Hirschmann (100) from the urine of a b o y with adrenal carcinoma. The spatial arrangement at C-17 is that of the natural adrenal compounds (probably 1 7 ( a ) O H , although formulated 17 (β) O H b y the authors). Accompanying the diolone was the

H O H O

Of these, pregnane-3(û;),20a-diol (CLI) is unique, in that this compound is also the chief metabolite in man of the corpus luteum hormone pro-gesterone ( X X I V ) , and that there exists a simple gravimetric method (253) for its estimation (as 3-monoglucuronide), which stands in contrast to the more elaborate chemical procedures requisite to the determination of the excretion of the other substances mentioned. In addition to com-pounds / to k, several other very probable cortical metabolites have been separated from urine, but, to date, their relationship to cortical function may only be surmised, or they are characterized only by physical con-stants and elementary composition. It is highly significant that of the ten or more recognized excretory products of the cortical hormones, only two, 11-hydroxy-androsterone (CXLVI) and ll-keto-etiocholan-3(a )-ol-17-one (CXLVII) (members of the 17-ketosteroid group a) have been shown conclusively to be oxygenated at C-ll, but there can be no reason-able doubt that one or more of the components of " urinary cortin,"

which possesses carbohydrate activity, is also an 11-oxygenated com-pound which further fulfills the structural requirements at present associated with cortical activity of any type (Section VI).

Reactions a to e, which may be carried out routinely without chemical isolations, provide criteria of quite different kinds of cortical secretory function, which need not and do not run strictly parallel. The neutral 17-ketosteroid colorimetric estimate gives a measure of the metabolites of the testicular and adrenocortical hormones excreted as 17-ketones;

the quantity of cortical origin is about 10 mg. per day. Undoubtedly these products arise both from the catabolic breakdown of members of the C21 series and from the normal physiological secretion from the

corresponding D-homo rearrangement product (formula 2), 17a-methyl-A6 -2)-homo-androstene-3(j3),17a-diol-17-one (6.4 m g . / l . ) , which almost certainly arises from 1 in the course of the processing of the urine. The close chemical relationship of 1 to the adrenal constituents 17"/3"-hydroxyprogesterone ( X X I ) and allopregnatie 3(<*),-17"/S"-diol-20-one ( X X , page 558), and the fact that 1 has not been encountered in normal urine leave little doubt that this excretory product arises from the adrenal cortex.

Ο 4^ TABLE XXII DAILY URINARY EXCRETION OF CORTICAL METABOLITES Subject Normal male average. . Normal female average Hypoadrenalism Addison's disease Panhypopituitarism. Hyperadrenalism Hirsutism Cushing's syndrome. Virilism Stress Burns Postoperative Late pregnancy

Criterion of cortical function 17-Keto- steroids,0 mg. 15 10 0-7 0-4 15-30 10-40 40-250 20-30* 20-30* 10-20*

Cortin,6 mg. 0.062 0.039 0 to 0.015 0 0.050 to 0.065 0.2 to 0.7 0.1 to 0.5* 0.1 to 0.2* 0.1 to 0.4

Reducing substances, mg. Heard et al.c 1.5 1.3 0.4 to 0.6 4.8 3.0 to 4.0 2.8 to 3.2

Talbot et al.d 0.24 0.24 0.02 to 0.26 0.10 to 0.17 0.23 to 0.32 0.90 to 12.0 0.15 to 0.57 0.34 to 1.70 0.34 to 1.70

Generated on HIO4 oxidation 17-Keto- steroids,' mg. 10-16

Formalde- hyde/ mg. 0.5 to 0.8 0.15 21.0

a ο

α Expressed as the color equivalent of androsterone. 6 Data of Venning et al. (92,254,251); expressed as the biological equivalent of Kendall's Compound Ε (VIII), compared by th power to cause deposition of glycogen in the liver of the adrenalectomized mouse. e Data of Heard, Sobel, and Venning (92); ascertained on extraction of the urine at pH 1.0 and expressed as the reducing equiva lent of desoxycorticosterone. d Data of Talbot et al. (250); ascertained by extraction of urine without acidification and expressed as mg. of "corticosteroid." « Data of Talbot and Eitingon (249) ; ascertained by extraction of urine with butanol and hydrolysis of the conjugates wit barium chloride followed by hydrochloric acid or with liver enzyme, followed by barium chloride, followed by hydrochloric acid expressed as the color equivalent of androsterone. ' Data of Lowenstein, Corcoran, and Page (136); estimated by determination of formaldehyde generated on periodic acid oxida tion of the neutral extract of acidified urine and calculated as the equivalent of dehydrocorticosterone (XVI). 0 A slight rise, inconstantly observed immediately after damage, and followed in one or two days by a decline to normal or sub normal values (60,64,246). A The upper limit is recorded by the Zimmermann reaction, and is probably due to increased output of 20-ketosteroids ; no ris is observed by the antimony trichloride method (Venning, 254). * Dependent on the severity of the stress, the rise, unlike that of the 17-ketosteroids, may persist for many weeks (137).

606 R . D. H . H E A R D

gland of members of the C I⁹ series. The active cortin-like substance or substances are excreted in very much smaller amounts (of the order of 0.1 mg. per day) and presumably represent an overflow of cortical hor-mones from the circulation. Their chemical nature is not known, but it is to be anticipated that the groupings essential to high cortical activity (Section VI) are contained in the molecules. The urinary excretion of neutral water-insoluble, chloroform-soluble lipides which reduce cupric ion or phosphomolybdic acid also varies directly with degree of cortical function. As reducing capacity is conferred by the primary a-ketol grouping (side chain types 6 and e, page 563), it is assumed that the determination provides a measure of the output of metabolites possessing

T A B L E X X I I I

U R I N A R Y E X C R E T I O N OF R E C O G N I Z E D A D R E N O C O R T I C A L M E T A B O L I T E S0

Nor-Metabolite Formula mal urine

a Expressed as mg./l., and excluding 17-ketosteroids and artifacts arising from the compounds listed (see footnote5 page 6 0 2 ) .

these side chain types. The quantity normally excreted seems to be of the order of 1 mg. per day. One compound has been isolated which is strongly reducing and which also exhibits gluconeogenic activity, but its structure has not been elucidated. Two reactions with periodic acid, applicable to the appropriate urine residues, give further quantitative information concerning two other classes of metabolites, namely the 17-20 glycols (side chain types a and d, page 563), and the 20-21 ketols and glycols (side chain types a, î>, e, and / , page 563). The first-mentioned class, which is non-reducing and is contained in the neutral non-ketonic alcoholic fraction of urine, yields, with periodic acid, the corresponding 17-ketosteroids, which can then be estimated colori-metrically in the usual way, while the second gives rise to formaldehyde, which may also be determined colorimetrically. In Table X X I I , the excretion of cortical metabolites, as determined by each of these five

607 procedures a to e, is compared under various conditions of hypo- and hyperadrenalism.

Fewer quantitative data pertain to the metabolites of group compounds / to k, mainly because of the rather elaborate chemical isola-tions requisite to the determination of all except pregnane-3(a),20a-diol.

Table X X I I I shows the excretion of these products in a limited number of normal individuals and in a few cases of adrenocortical hyperfunction.

1. 17-Ketosteroids

The relationship between the neutral 17-ketosteroids of urine and testicular and cortical function is more fully dealt with in Chapter X I I . In general, the view is accepted that these excretory products (formulated in Table X X I V ) arise almost entirely from the adrenal cortex in the normal female and mainly from the cortex in the male. In Addison's disease, the output of urinary 17-ketosteroids is markedly diminished.

According to Fraser et al. (63) the titer is practically zero in the female and about one-third normal in the male, the latter residual proportion presumably representing the contribution from the testis; in the data of Friedgood (66), however, this difference between the male and female Addisonian is less sharply defined. In panhypopituitarism values as low if not lower than in Addison's disease are encountered; this is difficult to reconcile with the fact that even in the totally hypophysectomized animal some adrenal function is still retained. Conversely in certain types of hyperadrenalism, particularly in cases of virilizing cortical tumors, the output may reach levels many times normal.

There is very little information bearing directly on the important problem of which of the many adrenal steroids (Table III, page 557) may function as precursors of one or more of the urinary 17-ketosteroids of cortical origin. The ease with which the 17-hydroxylated compounds (side chain types a to d, page 563) may be ruptured in vitro to the corre-sponding 17-ketone has long suggested that the same reaction may proceed in vivo. In guinea pigs treated with 5-mg. quantities of allopregnane-3(a),17^aa"-diol-20-one and of Kendall's Compound Ε (VIII) an increase in 17-ketosteroid output (antimony trichloride color assay) from about 0.15 to 0.45 mg. per day has been observed (186); the increment accounts for roughly 5-10% of the substance administered in each case. On the other hand, in rabbits each given orally over three days 225 mg. of the 3,21-diacetate of allopregnane-3(/3),17"/3'',21-triol-20-one (the diacetate of Reichstein's Compound Ρ, X ) , or given subcutaneously the same quantity of the unacetylated substance, the excretion of the correspond-ing 17-ketosteroid, isoandrosterone, (CXLIII) could not be demonstrated, nor could any ketonic or non-ketonic transformation product be isolated

608 R. D. H. HEARD

following acid hydrolysis of the urine, systematic fractionation in the usual way, and chromatographic separation (32). Neither progesterone ( X X I V ) nor desoxycorticosterone ( X X I I ) , both constituents of the adrenal cortex, cause any significant rise in 17-ketosteroid output.

T A B L E X X I V U R I N A R Y 1 7- K E T O S T E R O i D Sa

0 0 0

CXLVII

° Exclusive of artifacts arising from the above compounds or their water-soluble conjugates.

Throughout gestation when the progesterone secretion of the corpus luteum of pregnancy and of the placenta increases enormously, the urinary 17-ketosteroids remain relatively constant (44,77,79,137,187,254);

also the subcutaneous or oral administration to individual male or female rabbits of up to 500 mg. of progesterone does not alter significantly the

17-ketosteroid excretion (103). Desoxycorticosterone acetate adminis-tered at a dose level of 10 mg. per day to both male and female Addison-ians does not lead to any consequential change in titer (35) ; similarly, in adult male rabbits each treated with 500 mg. of desoxycorticosterone acetate no significant increase in urinary 17-ketosteroids is observed (103). Following the administration of adrenocortical extracts to human subjects, no convincing alteration in urinary 17-ketosteroids has been noted, but in this connection it should be pointed out that the con-centration of adrenal steroids in commercial extracts is not high on a weight/volume basis, and that the quantity given may be insufficient to reflect the metabolic pathway. Stimulation (in hypopituitarism in man) of the adrenals with corticotrophin does however lead (252) to a signifi-cant increase (five to seven times) in the output of 17-ketosteroids (as determined by the Zimmermann color reaction).

2. Cor tin

The cortin-like properties of extracts of human urine were first observed by Perla and Marmorston-Gottesman (182) in 1931 and by Grollman and Firor (78) in 1932. Ample confirmation followed, with the demonstration that the excreted principle or principles possess activity in the adrenalectomized animal with respect to (a) protection against histamine poisoning (182) and exposure to cold (39,262), (b) time survival and life maintenance (38,78), (c) prevention of water (220) and potassium (55) intoxication, (d) prolongation of the work performed by the gastrocnemius muscle (228), and (e) the deposition of hepatic glycogen (38,137,256,258). The chemical nature of the active metab-olite (s) is not established, although Venning, Hoffman, and Browne (256) have isolated, as acetate, a crystalline ketone, melting point 234-236°, which reduces alkaline silver diamine and exhibits cold pro-tection; also the compound promotes glycogen storage in the liver (105).

As urine concentrates are corrective in both the electrolyte and carbo-hydrate disturbances associated with loss of adrenal function, the excre-tion of active substances in both the 11-desoxy- and 11-oxy-series is implied. Presumably also the active metabolites are excreted in con-jugated form, as extraction of urine at pH 1.0 leads to preparations with approximately twice the potency of those attained on processing at neutrality (92,137,257,258); more vigorous hydrolytic treatment, such as boiling in acid medium, destroys the active excretory products.

The quantity of cortin normally eliminated per day fluctuates ( ± 50 % of the mean) from time to time in the same individual, possibly due to variation in state of activity or excitement. Compared by the power to cause glycogen deposition in the liver of the adrenalectomized animal

610 R. D. H. HEARD

(42,45,137,213,258) and expressed as the biological equivalent of Ken-dall's Compound Ε (VIII), normal adult females excrete from 29 to 55 μg. (average, 39), with no correlation apparent with the phase of the menstrual cycle, and normal adult males, from 45 to 90 Mg. (average 62) (92,137,257). Exercise and activity may raise the amount excreted (137). No glycogenic activity is detectable in the urine of newborn males (1-4 days), but at 2.5 years of age the excretion has attained the adult female level, and at about 6 years the adult male value is reached (92,137,257). Subnormal to zero values are associated with hypo-adrenalism (Addison's disease and panhypopituitarism), and, conversely, with cortical hyperfunction (Cushing's syndrome) and under stress (thermal trauma, the postoperative state, and gestation), values six to ten times normal are encountered (92,137,257).

Experimentally, the administration of whole adrenal cortical extract to the dog (80), monkey (41), and man (40,256,261) causes a rise in urinary cortin, the increment representing approximately 10% of the hormones contained in the original extract (effect on salt and water metabolism and cold exposure test). In the monkey, active material (cold exposure test) is still excreted following gonadectomy but not adrenalectomy (41). Stimulation of the adrenal cortex in man by the administration of pituitary adrenocorticotrophin leads to a ten- to twenty-fold increase in cortin output (cold exposure test, 137).

3. Neutral Lipide-Soluble Reducing Substances

Talbot et al. (250) and Heard and Sobel (91) have quantitatively standardized the reduction of cupric ion and of phosphomolybdic acid, respectively, by adrenal steroids possessing a reducing group in the molecule, and have developed colorimetric methods for the estimation of small quantities of these compounds. In the first-mentioned test the reaction is given by a primary a-ketol grouping (side chain types b and e, page 563), and in the second by the a-ketol and/or the α,β-unsaturated 3-ketone groupings. When applied to extracts of urine, general paral-lelism between urinary reducing power and cortical function is observed (Table X X I I ) . By the procedure of Talbot et al. (250), the urine is extracted without previous treatment with acid, and the colorimetric determination is made on that portion of the neutral lipide-soluble ketones which is extractable from benzene with water. Heard, Sobel, and Venning (92) acidify the urine to pH 1.0 prior to extraction, an operation which, as in the estimation of urinary cortin (page 609), leads to an increase (four to five times) in the quantity of metabolites recovered, and they apply the colorimetric estimation directly to the total neutral extract. By either technique the excretion of reducing substances is

611 approximately one-third to one-half normal in hypoadrenalism and two to ten times normal in hyperadrenalism, under stress, and in late pregnancy. Normal adult values fluctuate within ± 5 0 % of the mean;

by the procedure of Heard, Sobel, and Venning (92), males range from 1.1 to 2.1 (average 1.5) mg. per day (calculated as desoxycorticosterone), and females from 1.0 to 2.0 (average 1.3). With increasing age in male children, the normal adult level is not attained as early in life as the out-put of cortin (92); by age 7, urinary reducing capacity is only 50% of the adult value, while cortin excretion has already reached the adult level (see page 610). Also, in convalescence from damage, the normal biologi-cal titer is attained more rapidly than that of the reducing substances (92). In the dog (92), extirpation of the adrenals diminishes the output by about 66%, and the intravenous injection of cortical extract causes within a few hours a sharp rise and fall, the total additional excretion representing 3.4% of the reducing capacity of the original extract; also, stimulation of the adrenals by the subcutaneous administration of whole anterior lobe extract containing adrenocorticotrophin leads to a twofold increase over normal, which persists for several days. In man (hypo-pituitarism), corticotrophin causes a sixfold rise in the output of reducing substances (252).

It must be assumed that at least one third of the water-insoluble, lipide-soluble reducing substances normally excreted are of extra-adrenal origin, as in Addison's disease and panhypopituitarism (92,250), and, following bilateral adrenalectomy in the dog (92), a minimum residual titer of at least this proportion is retained.

4. 17-Ketosteroids Generated on Periodic-Acid Oxidation of the Neutra Nonketonic Lipides

As pointed out in Section III, C, 2, steroid glycols hydroxylated at C-17 and C-20 (i.e., side chain types a and d) are readily oxidized by periodic acid to the corresponding 17-ketones. Accordingly, 17-keto-steroids formed from the neutral lipides of urine by this treatment should be an index of the quantity of metabolites with these types of side chains which are excreted.

The principle has been applied by Talbot and Eitingon (249). Be-cause of the sensitivity of the tertiary 17-hydroxyl group and of con-jugated alcohols to dehydration on treatment with strong acid, the conjugates are first extracted from urine with butanol, washed until neutral, and then hydrolyzed (a) by refluxing with barium chloride at pH 6.0, or (Jo) with rat liver enzyme, or (c) by refluxing for ten minutes in 15 volumes per cent hydrochloric acid, or progressively by b plus a plus c, b plus c, or a plus c. After separation of the hydrolyzates into

612 R. D. H . H E A R D

ketonic and non-ketonic portions with Girard reagent, colorimetric assay of the ketonic fraction and of the neutral products of the oxidation of the non-ketonic fraction with periodic acid then give, respectively, a measure of the "preformed" 17-ketosteroids and of those formed by the oxidative treatment. Following hydrolytic procedures a or b, a significant quan-tity of 17-ketosteroids is generated from non-ketonic material on oxida-tion with periodic acid. As none is observed after hydrolysis c, it follows that the precursors of the formed 17-ketosteroids are destroyed on being subjected to strong acid treatment. That the precursors are excreted in conjugated state is evidenced by the significantly lower values encountered on processing the urine without any form of hydrolysis.

In one case of adrenocortical hyperplasia and two cases of adreno-cortical carcinoma, the daily excretion of non-ketonic substances oxidiza-ble with periodic acid to 17-ketosteroids amounted to the equivalent of 10-16 mg. of the latter as compared to the excretion of about 0.4 mg. by normal individuals; in the three pathological subjects, the output of pre-formed 17-ketosteroids was, respectively, 25, 60, and 200-375 mg. (249).

Fieser, Fields, and Lieberman (57) have also observed the formation of 17-ketosteroids on periodic acid oxidation of the non-ketonic alcohols derived from commercially processed human pregnancy urine subjected to acid hydrolysis. The periodic acid oxidation of 90 g. of this fraction (from 467 1. of urine) yielded 4.14 g. (4.6%) of ketonic material which could not be crystallized after repeated chromatographic fractionation but which contained the equivalent of 0.87 g. of androsterone by colori-metric assay. The output of precursors is thus about 2 mg./L, a level somewhat higher than that encountered by Talbot and Eitingon (249) in normal individuals (0.4 mg. per day) and suggestive of increased adrenal cortical activity during pregnancy, which is also evidenced by other criteria of cortical function (Table X X I I , page 604). According to Dobriner and Lieberman (137) however, the quantity of 17-keto-steroids generated with periodic acid from the non-ketonic fraction of urine does not in general parallel the excretion of active cortin.

An exploration has been carried out (57) of the possibility of separat-ing 1,2- or 1,3-glycols from urine as their acetals with ra-hydroxybenzalde-hyde so that the condensation products could be selectively extracted from a mixture with alkali. While acetal formation readily took place in model experiments with A⁴-pregnen-17"a:,"20,21-triol-3-one and A⁵-pregnene-3(0),2O,21-triol, only 0 and 20%, respectively, of the original steroid could be recovered from the acetals by acid hydrolysis, presuma-bly due to the high sensitivity of α-glycols to dehydration.

At least one non-ketonic metabolite has been isolated which gives rise

613 to a 17-ketosteroid on oxidation with periodic acid, namely, pregnane-3(a),17,20-triol (CLII): in all probability the 3(ß)-hydroxy epimer of CLII constitutes another.

5. Formaldehyde Generated on Periodic Acid Oxidation of Neutral Lipides Cleavage of 20-21-a-ketols or glycols (side chain types a, b, e, and / , page 563) releases formaldehyde, the estimation of which may serve to

5. Formaldehyde Generated on Periodic Acid Oxidation of Neutral Lipides Cleavage of 20-21-a-ketols or glycols (side chain types a, b, e, and / , page 563) releases formaldehyde, the estimation of which may serve to