A Discussion of the Mechanism of Action ofSteroid Hormones*

GERALDC. MUELLER

(McArdie Memorial Laboratory, The Medical School, University of Wisconsin, Madison 6, Wis.)

The steroids may be classified as hormoneswhen they originate in one group of cells and regulate the physiologic behavior of other cells in thesame multicellular organism. Through the classicalmethods of endocrinology, involving extirpationof various organs and hormone replacement, anumber of primary sites of synthesis of suchsteroids have been identified as well as the chemical agents which they produce. That other steroidhormones will be found which are selectively produced in cells not nearly so susceptible to removalfrom the host seems highly probable, although itdoes not pertain to the present discussion.

As hormones, the steroids elicit biological responses in susceptible cells and tissues; the latterare often referred to as target organs when the response is sufficiently obvious and characteristic ofa particular hormone. These responses, for themost part, represent rather complicated changesof physiology in a group of cells and usually require a considerable period of time for overt manifestation. This physiologic behavior, which thesteroid agents modify, is the product of interactionof many enzyme systems and other functionalunits of the cell. Therefore, when we inquire howhormones act we are really asking how do thesesubstances induce shifts in the pre-existing balances of integrated enzyme systems. Unfortunately, in many cases the basic systems are not yetwell established and their manner of interactiononly inferred. Obviously then, the question ofmechanism must be asked repeatedly in the lightof each fundamental advance in our knowledge ofnucleic acid and protein synthesis, carbohydratemetabolism, lipide systems, steroid biochemistry,and other areas of cell physiology.

In attempting to assess present knowledge on* Studies from this laboratory presented in this paper

were supported by the Alexander and Margaret StewartTrust, grant C-1897 from the United States Public HealthService, and an institutional research grant from the AmericanCancer Society.

Presented at the second meeting of the Scientific ReviewCommittee of the American Cancer Society, held at the West-chester Country Club, Rye, New York, December 13 and 14,1956.

the mechanism of action of steroid hormones, oneis confronted with a vast sea of observationswhich range from the molecular level to the behavioral realm. It is the purpose of this review toconsider criteria for establishing relevance amongthese varied findings. In a general section, generalaspects of hormone action will be discussed; steroidhormones will be presented as a molecular class ofcompounds, and some of the characteristics of thebiological responses will be related for representative hormones. Turning to units of cellular structure and function, a discussion of basic controlmechanisms will be presented with respect to theirpossible implication in hormone action. In a moreexperimental section an analysis of studies on themechanism of action of estrogenic hormones willbe undertaken; in fact, the estrogens will serve asa model system throughout the paper, since (in theopinion of the author) more is known concerningtheir mechanism of action than for any othersteroid. Relevant findings with other steroids willbe reported in less detail.

A number of unanswered questions will beraised in the attempt to point out major deficiencies in our knowledge; it is hoped that these willprovoke discussion and experimentation. In thepresent state of knowledge one can hardly consider hormone action without resorting to speculation; if in the course of this activity certain "strawmen" are set up, perhaps they will kindle the fires

of research.Before proceeding, the thoughtful review by

Hechter (37) on the general subject of hormone action is recommended to the reader's attention.

Similarly, the manuscripts of Roberts and Szego(75) and Lieberman and Teich (52) contributenobly to the documentation of the steroid biochemistry.

I. THE HORMONALAGENTSteroid hormones belong in a class of com

pounds having a common cyclopentanoperhydro-phenanthrene nucleus. Biosynthetically they areall derivable from cholesterol or a common intermediate by a series of reactions involving the re-

490

Research. on August 13, 2020. © 1957 American Association for Cancercancerres.aacrjournals.org Downloaded from

MUELLER—Mechanismof Actioii of Steroid Hormones 491

moval of all or part of the side chain, selective de-hydrogenation, and strategic oxidation of certainpositions on the nucleus (18). Some of these reactions are highly developed processes in the adrenals, ovaries, testes, and placenta and are themselves regulated by the action of nonsteroid hormones referred to as trophins. Other steps in theinterconversion of biologically active steroids maybe carried out in a variety of tissues (18, 61).

A certain degree of correlation exists betweenthe chemical structure of the steroid and the ability to elicit a biological response; this permits aloose classification, which is shown in Table 1.Biological responses are mentioned which permitassay or reveal characteristic effects. However, itis to be emphasized that there is considerableoverlap in this regard, which may result frommetabolic interconversion, tolerance by the receptor system of multiple sites of potential specificityin the steroid, or metabolic sparing of a naturalanalog. High biological activity in general is associated with fusion of rings A and B in the transposition relative to the angular methyl groups; thisaffords a more planar molecule from the rear aspect. A side chain at position 17 in progesteroneand in the adrenal cortical steroids is invariablycis to the angular methyl groups. In those steroidswithout a side chain, the ßhydroxyl group at position 17 is associated with higher biological activitythan the ketone; and an a hydroxyl group drastically reduces activity. A similar relationship existsfor adrenal steroids with reference to oxygénationat carbon 11. In general, the more polar functionsare found protruding from the ßaspect of themolecule.

The steroids, as hydrocarbons, are lipophilic,but this property is diminished by hydroxyl substitution. The hydroxyl groups in turn are revers-ibly convertible to the ketonic forms by specificsteroid dehydrogenases (59) ; the ketones are considerably less polar.

One of the most striking properties of steroids istheir ability to associate with proteins; this propensity appears to play a part in the transport ofsteroids in the body fluids (7, 59, 83). Aqueoussolutions of protein, particularly albumin, are ableto solubilize remarkable quantities of steroid (6,20); in this latter case, however, the associationappears to be nonspecific, and the steroid can bereadily recovered by extraction with organic solvents. A more rigid binding with protein in which acovalent linkage appears to be established enzy-matically has been described by Riegel andMueller (72).

Steroids possessing oxygen functions are alsosubstrates for the glucuronide- and sulfate ester-

synthesizing systems (51). Such compounds possess detergent-like properties; however, most ofthem have not been well characterized or investigated from a functional point of view. Informationon the reactivity of the carbonyl group of steroidsin biological systems is rather limited to reduction;whether or not Schiff base formation (such as involved in transamination), aldol condensations,enol-ester or mercaptal formation occurs will require further research. It seems that, at this stageof our knowledge, the participation of a steroid asa substrate in all these types of metabolic reactionsshould be surveyed, since the biologically activeform of these agents cannot be defined as yet. It isquite possible that the steroids may express biological action through entering into such reactionmechanisms.

II. UNITS OF STRUCTUREANDFUNCTIONFor this discussion it is appropriate to consider

some of the factors involved in the expression ofbiological function; this is of interest, since weseek ultimately to explain how steroid hormonesmodify function. It must be recognized first thatbiological function is composed at all the differentlevels of organization; i.e., the animal, organ, cell,subcellular particulates, and the multiple enzymesystems. The contributions to function at eachlevel may be measured by the spectrum of products and the rate at which they are formed. Thatis to say that, for any cell, the type of products andthe quantity contributed to the environment in afinite period of time is characteristic of that cell ina specific environment; and, similarly, the type andrate of production of products of a subcellular par-ticulate are characteristic of that unit in its environment; and so it continues down through thesuccessively lower levels of organization to thefundamental unit of catalytic activity, the enzyme.

At each level we have a condition of semi-isolation of varying degree wherein a definite barrier tocommunication exists. Some of these barriers arerecognized cytologically as the cell wall, nuclearsurface, mitochondrial membrane, endoplasmicreticulum, and Palade granules; many others remain to be studied. In this sense it seems appropriate to visualize the cell as being made up of atremendous number of compartments, each ofwhich contains a specific association of enzymes.Thus, to mention a few, we have the tricarboxylicacid cycle-electron transport system concentratedin the mitochondria; diphosphopyridine nucleo-tide synthesis and pyrophosphorolysis of uridinecoenzymes concentrated in nuclei; protein andcholesterol synthesis concentrated in the endo-

Research. on August 13, 2020. © 1957 American Association for Cancercancerres.aacrjournals.org Downloaded from

TABLE 1

STRUCTUREANDFUNCTIONOFREPRESENTATIVESTEROIDHORMONESCLASS AND BEPHESESTAIIVE

COMPOUNDS

Progestational:

CH,i 3C=0

(Progesterone)

Estrogenici

OH

(Estradici)

Androginie:

(Testosterone)

Electrolyte activity:CH2OH

C=0

(Deoxy corticosterone)

O CH,OHII I

HO HC C=O

(Aldosterone)Carbohydrate activity:

CH,OH

BITES OF

SPECIFICITY

O

CHsC (side chain)

PregneneNucleus

(i.e., aßunsaturatedKetone)

Phenolic A-ring

|3-OH at position 17

Androstane (5aH)

Androstene nucleus

(A4or A6)

21-hydroxyl

18-carboxyl11-hydroxyl21-hydroxyl

17 a-hydroxyl11-keto or |3-hy-

droxyl

BlOX¿>GICAL RESPONSES WHICH

PERMIT ASSAY

Supports traumatic deciduomata development (4)

Hypertrophy of stremai nuclei in en-dometrium (40)

Increases carbonic anhydrase in urogenital tissue (55)

Vaginal cornification (23)Inhibition of water and hypertrophy

of uterus (3)Vaginal opening (41, 76)

Promotes comb growth (26)Promotes growth of seminal vesicle

and prostate (11, 60)Myotrophic activity in levator ani

muscle (21)

Promotes sodium retention in adre-nalectomized animals (16)

Promotes glycogen deposition in liverof adrenalectomized rodents (85)

Depresses circulating eosinophils inadrenalectomized mice (79)

Lymphocyte destruction (24)

(Hydrocortisone)

Research. on August 13, 2020. © 1957 American Association for Cancercancerres.aacrjournals.org Downloaded from

MUELLER—Mechanismof Action of Steroid Hormones 493

plasmic reticulum-Palade granule fractions; andglycolysis, amino acid activation, and manyothers localized in the operationally defined solublefraction. Each of these is probably made up ofmany smaller compartments, the smallest of whichmay be single enzyme molecules loosely localizedon a surface (interface) or may constitute a catalytic surface amid the structure of a contractile,protein, or a semipermeable membrane.

With this state of semi-isolation, productsformed in one area have limited access to enzymesystems localized in another. As soon as utilizationin the second area exceeds the diffusion or transport of products from the first, a gradient is estab-

surfaces (involving a three-point contact), reactions can be catalyzed which are thermodynami-cally possible. The enzyme acts by lowering theamount of activation energy required for the con-

From AnotherEnzyme

To AnotherEnzyne

S = Activated Subitr.ii floleciilesPAP'3 Product floleculesC »Coeniyne

CHART 1.—Hypothetical surface of an enzyme

lished which is maintained dynamically in accordance with the function of both areas. This type ofinteraction is of course not confined to only twoareas, but exists simultaneously between all.

In each fundamental compartment the enzymesof a multiple enzyme system oscillate within asteady state. The particular steady state attainedis determined by the amounts and varieties of specific enzymes and the concentrations of substratesand products which are maintained over the periodof measurement.

1. The catalytic unit: the enzyme.—Before considering some of the properties of multiple enzymesystems let us take a look at a single enzyme, sincethis is the fundamental machine permitting theflow of energy involved in product formation(Chart 1). An enzyme is a protein carrier of a specific catalytic surface. Through specificity studiesit has been established that this surface is complementary with the structure and configuration of aparticular substrate molecule. Through the formation of enzyme-substrate complexes upon these

version.The over-all velocity of an enzyme-catalyzed re

action is determined chiefly by the inherent properties of the enzyme, the amount of the enzyme,and the substrate and product concentrations.Thus, at low substrate levels the velocity may belimited by the number of active molecules thatcollide with the enzyme surface. With increasingsubstrate concentration, the frequency of collisionis higher and the velocity increases in a mannercharacteristic of the relationship between rates ofassociation and dissociation of the product fromthe enzyme. From the standpoint of rapid alterations of velocity in a biological system, variationof substrate concentration is probably the mostimportant. However, in certain systems the enzyme activity may be released from an "inhibited"

form or position. Cori has indicated that this maybe the case with phosphorylase which is broughtinto high activity with muscle contraction (15).The experimental findings are strongly suggestivethat contractile protein, possibly adherent to amembrane or surface of multiple enzyme compartment, may mechanically bring these enzymes intothe environment of their substrates. If this is truein muscle, one might ask whether a vestige of sucha control mechanism might also be operative inother tissues.

Accessory factors such as pH, equilibrium constant, oxidation-reduction potential, and free energy are also implicitly involved. Similarly, if theenzyme is missing a cofactor or activating ion, aprofound decrease in catalytic activity will be observed and finally (endogenous) inhibitors mayalso play a role.

2. Some properties of multiple enzyme systems.—The fact that two enzymes attacking the samesubstrate may have different Michaelis constants(substrate concentration necessary for half maximum velocity) and different turnover numbers allows the substrate concentration to determine thedegree of competition between the two alternativepathways. Thus, one enzyme may be saturated ata substrate concentration which permits only alow reaction velocity of the second enzyme, eventhough the latter under optimal substrate concentration has a higher turnover rate. In such a system the steady state is determined by the level atwhich the substrate is maintained. With this inmind, let us turn to a diagram, Chart 2, and a discussion of a multiple enzyme system presented byGreenberg (33). In this scheme substrates A, B, O,

Research. on August 13, 2020. © 1957 American Association for Cancercancerres.aacrjournals.org Downloaded from

494 Cancer Research

K, and R are beginning substrates which have diffused into the environment of our multiple enzymesystem from other compartments. These are introduced into a major synthetic chain along which Ais converted sequentially to C —¿�>F —¿�>J —¿�»P. Thesequence O —¿�»P —¿�>E represents the activation of

substrate O such as known for fatty acids, aminoacids, and nucleotides. GHa—G is a reversible oxidation-reduction carrier system which is reoxi-dized by substrate R. Y is a side product. Eachnumber represents an enzyme-catalyzed reaction.Reaction (5) occupies a strategic position. If in theintact system only traces of intermediates C, F,and J are present during the formation of P, it follows that each of the enzymes may be acting at alow and different level of saturation : thus a steadystate exists, limited by reaction (1) involving substrates A and B. In this case either the enzyme orsubstrate may be limiting. If, however, the rate ofreaction (5) is decreased below a critical level, thesequence of reactions from (1) to (4) is disturbed.The system then seeks a new steady state in accordance with the fundamental properties of eachenzyme and the reaction it catalyzes. IntermediateF may increase while product P diminishes. If reactions (1) to (4) are reversible, the new steadystate will be characterized by a difference in thesaturation levels for each enzyme. Of greatest importance is that, in the establishment of the newsteady state, the ratios between the formation ofproducts F, P, and Y are changed. They accumulate, accordingly, to form different gradientsthrough which they diffuse to other multiple enzyme systems in other compartments. Here theycontribute to the substrate environments whichhelp determine, in turn, the steady state for a different group of enzymes and a different spectrumof products.

Thus, the compartmental nature of the cell increases the versatility to respond to certain stimuli(substrates) and provides for maintenance of orderamong the multitudes of enzyme reactions. Feedback mechanisms are implicit among such systems. Communication between these semi-isolatedorganized systems operating in different steadystates requires time; therein lies a considerablestability of the systems with respect to time.

3. Some additional comments on organization.—An integral aspect of the compartmental structureis that a local concentration of enzymes and substrate can be obtained. For enzymes operating in asequence it would appear that the efficiency mightbe increased if they were organized spatially inclose proximity to one another. Such appears to bethe case for mitochondria. This particle contains agroup of enzymes which will oxidize acetate to car

bon dioxide and water, and trap the energy inadenosinetriphosphate. Attempts to fractionatethis group of enzymes have demonstrated thatelectron transport and oxidative phosphorylationare intimately coupled (12, 31, 50). In fact, Greenand associates have isolated submitochondrial particles which appear to be organized aggregates ofenzymes with lipide backbones (56, 57, 88). Intheir "electron transport particle" (Chart 3), the

enzymes behave as if they are linked sequentiallywith their prosthetic groups in sufficiently closeapposition so that electron transfer proceedsthrough the chain with little opportunity for random interaction. High efficiency is observed. The

RH

CHART2.—Diagram of a multiple enzyme system.Letters refer to precursors and products ; numerals refer to

enzymes catalyzing specific reactions. (Reprinted with permission from G. R. Green berg, Problems in the Study of Multiple Enzyme Systems. In: D. T. Green [ed.], Currents in Biochemical Research. New York: Interscience Publishers, Inc.,1956.)

same situation seems to hold for the other substrates and products. Disruption of the organizedstructure by deoxycholate allows the escape ofcytochrome C (56) with simultaneous loss of ability to catalyze the oxidation of reduced diphospho-pyridine nucleotide by oxygen. This type of organization may be an example of maintaining maximal substrate concentration at or near the catalytic surfaces of enzymes acting in a sequence.

In this respect one begins to wonder about thematrix in which enzymes lie: the high phospho-lipide content of the mitochondria, the high cholesterol content of the endoplasmic reticulum, theinositol lipides in the nucleus, and the glycopro-teins characteristic of certain cell surfaces. Whatrole do these materials, which have the propertiesof forming membranes, play in regulating cell

Research. on August 13, 2020. © 1957 American Association for Cancercancerres.aacrjournals.org Downloaded from

MUELLER—Mechanism of Action of Steroid Hormones 495

function? Certainly they are suspect sites ofsteroid hormone action.

III. SOMEGENERALASPECTSOFHORMONEACTION

Any explanation of the mode of action of steroidhormones must answer two basic questions (37)arising out of the nature of many hormonal responses:

A. How can trace amounts of hormone produceprofound biological responses without contributing significant amounts of energy or mass to thesystem?

B. How can the apparent specificity of hormoneaction be explained?

1. Consideration of Question A.—One must accept that in the interaction of the hormone with

D Me—

0 —¿�Me—¿�

i /•'

^ I ini

-M.-fT]-M.-

-Me

Lip.de - Lipide

D-DPNH IMf «Met.l

c Deriyarogen3 > CyKxniomei

CHART8.—Diagram of an electron transfer chain.Indicates a high degree of spatial organization of enzymes

situated in a lipide matrix. (Reprinted with permission fromD. E. Green, Structural and Enzymatic Pattern of the ElectronTransfer System. In: O. H. Gaebler [ed.], Enzymes: Units ofBiological Structure and Function. New York: AcademicPress, Inc., 1956.)

the biological system the hormone acts as a particle of mass, that primarily it can only alter thestate of another particle of mass through molecularinteraction. Therefore, in order to produce theprofound biological responses observed, the hormone must affect a component of an amplificationsystem; such loci would appear to lie within theoperation of multiple enzyme systems. Withinthese parameters a hormone might act initially byaltering either the effective concentration of anenzyme or the effective concentration of a substrate.

The level of active enzyme could be altered in a

number of ways without invoking new enzymesynthesis. If a hormone reacts with an enzyme toeither increase or decrease the Michaelis constantit would be reflected in the velocity of the catalyzed reaction with a given substrate concentration. It also follows that, if the hormone combinedwith incomplete enzyme as a coenzyme or prosthetic group converting inactive to active enzyme,the initial over-all conversion of substrate to product would be increased accordingly; the oppositeeffect on the total process would be observed if thehormone were an inhibitor.1

Considering the compartmental nature of thecell, with enzymes fixed on or within membranous structures, a hormone might act by increasing the accessibility of pre-existing enzyme tosubstrate. Acting as a surface-active material ordetergent, it could induce permeability changes ina membrane or compartment around an enzymeand thus indirectly alter the concentrations ofthose substrates diffusing in or out of the compartment. Such an interaction must, however, accountfor the specific relationships of hormone structureand biological activity. The low concentration ofhormone required for activity (about 2 X 10~9 M

for estradici) also points to the existence of specific receptors rather than any general detergent-like action. More probably the hormone would introduce electrostatic changes in a membrane analogous to the changes which take place in neuralmembranes or contractile proteins. In addition, itmust be kept in mind that cell structures are beingcontinually renewed. Thus, the hormone might interfere with either the synthesis or destruction ofsuch a structure; however, this again would referthe initial site of action of a hormone to a specificenzyme. So far, practically no evidence can bebrought to bear either pro or con, although thearea appears highly promising.

On the other hand, the synthesis of new enzymes is certainly involved in the over-all hyper-trophic effect of certain hormones. Whether this isa primary or secondary effect of the hormone canonly be speculated upon at this time. While littleis known of the factors which regulate de novosynthesis of enzymes in mammalian cells, there is reason to believe, from the studies of Knox et al. (45,46, 47) on tryptophan peroxidase, that the principles of induced biosynthesis of enzymes, established in microbiology, also apply to mammaliansystems. These principles, largely derived from thework of Monod, Cohn, Lwoff, Spiegelman, andStanier, have been summarized by Monod (63).

1In 19«,D. E. Green (30) first formally presented the concept that hormones might act as coenzymes or inhibitors ofspecific enzymes.

Research. on August 13, 2020. © 1957 American Association for Cancercancerres.aacrjournals.org Downloaded from

496 Cancer Research

When supplied with adequate precursors forprotein synthesis and energy, genetically competent E. coli can be induced to synthesize /3-galacto-sidase by the simple addition of a /3-galactoside tothe culture medium. The inducer may be a substrate for the enzyme, an inhibitor, or have littleaffinity at all. The minimum requirement is asteric similarity to the actual substrate. In addition, the inducer is not used up in the process ofinducing the de novo synthesis of new enzymes.

As a new active enzyme is formed it begins toact on available substrate, thus giving rise tohigher concentrations of product; the latter in turnacts as inducer for the synthesis of a second enzyme, and so on. This process of sequential induction has been shown to apply in galactose andtryptophan metabolism in bacteria (80) and in thecytochrome system of yeast (78). The inductionprocess is controlled in two ways: through theavailability of inducer and a type of feed-back inhibition exerted by products of the induced en-zymic reaction. In addition, competition can takeplace between the induction of separate enzymesystems. The important principle is that changesin relative substrate concentration, howeverbrought about, can set in motion a sequential induction of enzyme synthesis. The balance of powerbetween interacting enzymes within the cell isthen changed, giving rise to an alteration in biological function.

It should be kept in mind that this process involves actual protein synthesis and, very possibly,concomitant ribonucleic acid synthesis (36). Also,as far as it is known the presence of an inducer isrequired; this is normally the substrate. Thus, it isprobable that hormonally induced enzyme synthesis would be associated with a primary increase inthe substrate concentration. Once again, the site ofaction of the hormone would appear to be on anenzyme or its compartment. However, the possibility cannot be excluded that the hormone actsdirectly on the nucleic acid. In fact, it has been reported that estradici (5-10 Mg/ml) facilitates theaction of deoxyribonuclease on the decompositionof deoxyribonucleoproteins (1).

2. Consideration of question B.—The target organ concept has arisen in endocrinology becausethe biological response in certain tissues completely overshadows happenings in others. For example,the response of the uterus to estrogens is dramatic;but when one looks beyond, he sees that estrogensstimulate mitoses in the skin (10), suppress theproduction of gonadotrophins by the pituitary(34), cause development of pituitary tumors (62),stimulate fat synthesis in the liver (70), promoteprotein synthesis in the connective tissue of the

pubic symphysis (28), and change the histologieappearance of the salivary glands (27). These areonly a few of the responses obtained in differenttissues. While many of them may be complicatedby other humoral factors, it does appear that thespecificity of hormones for certain tissues may bemore apparent than real.

It is not possible to go into the subject of differentiation and morphogenesis in this discussion;it is only necessary to recognize that the end result of such processes is a segregation of potentialities for function in certain cells. Thus, it is possiblethat a hormone acts similarly in all cells, but thatthe composition of the enzyme population, beingquantitatively different for various cells, gives riseto a different response in light of the foregoing discussion. Of course, if such differences in enzymepopulation involve the initial receptors or the"hormone signal," the target organ concept would

be upheld. Attempts thus far to show selective localization of exogenous steroids in target organshave not been very successful; however, selectivelocalization may not be needed.

The recent findings of Huggins and Jensen indicate the complexity of the hormone specificity (41,42). The authors carried out a very fine study onthe ability of various steroids to depress estrone-induced uterine growth. It was observed that 6-keto and 16-hydroxyl derivatives of 3-hydroxy-estratriene ("impeded" estrogens) inhibited the

response in the uterus but not in the vagina. Fromthis, it would appear that the hormones and thedepressor steroid are eliciting responses throughdifferent mechanisms in each of these tissues. However, until such problems can be attacked in isolated systems, the question of hormone specificitymust be pondered. From the present position, theapparent specificity seems to reflect the effect ofquantitative differences of the spectrum of enzymes available for the response.

IV. EXPERIMENTALAPPROACHESTOTHE MECHANISMOF STEROIDACTION

In the foregoing discussion an attempt has beenmade to set up a framework for understanding thegeneral nature of any biological function; obviously, this is incomplete. However, it was doneso that one might begin to establish some relevanceamong the multitudinous observations which havebeen made on the responses of each steroid hormone.

In this regard, it can be stated that any visibleresponse (i.e., changes in macroscopic or microscopic appearance, weight, mechanical activity,and behavior) is likely to be many scientific milesremoved from the initial action of the hormone.

Research. on August 13, 2020. © 1957 American Association for Cancercancerres.aacrjournals.org Downloaded from

MUELLER—Mechanismof Action of Steroid Hormones 497

This is particularly true when the observationshave been made in tissues where cell reproductionor destruction has taken place over a considerableperiod of time. Accordingly, interpretation of thehormonal response under such circumstancesshould take into account the varied composition ofthe tissue. In fact, it is frequently difficult to knowwhat constitutes an adequate control tissue insuch cases. Nonetheless, these observations establish the important cumulative effects of hormonalstimulation and thus provide one parameter of theresponse.

At this time there do not seem to be any shortcuts to deciphering the mechanism of action ofsteroid hormones. For those hormones concerning

TABLE 2

EFFECTOFA SINGLEDOSEOFESTRADIOLONTHEACCUMULATIONOFRNA

ANDDNAPERUTERUSA single injection of 5 jig. estradici was

given. Derived from data by Telfer (84).

Hours afterestradiol

06

214872

DNA (mg.)per uterus

1.351.331.711.51a.08

RNA (mg.)per uterus

0.3970.3590.7751.160.939

which no clues are evident as to the area of metabolism they affect (he., glucose metabolism inthe case of insulin), it is necessary first to establish it. This may be approached by surveyingsegments of metabolism at the earliest time periodfollowing the hormonal administration that onecan be reasonably certain a response exists. Preferably this should be within a few hours (within 6hours).

By utilizing isotopically labeled precursors, representative metabolic pathways can be studied inthe animal or in tissue slices, and the presence orabsence of a hormonal influence on a specific segment of metabolism can be assessed. Once an effect has been established, the problem becomes oneof interpreting the finding. Surveys of other metabolic processes add perspective to the original result. However, one is very dependent on the stateof fundamental knowledge of the processes underconsideration. In the light of available knowledge,the most sensitive reaction must next be studied incell-free systems with the properties of enzymesand multiple enzyme systems in mind. The problem then becomes one of determining whether theobserved response is due to substrate effects or alteration of effective enzyme concentration. Thisshould be considered in light of the discussion inSection III.

Another approach is to study the hormone itself: the metabolic conversions it undergoes, theenzymes that attack it, and the possible functionof such systems in metabolism. In essence, thisconstitutes an approach opposite to the first. Bothapproaches are to be recommended, since in thefinal analysis an explanation of steroid hormoneaction requires both a knowledge of the initial interaction of the hormone with certain tissue constituents and knowledge on the dynamic chain ofevents which take place in the expression ofbiological response.

V. STUDIESOF THE MECHANISMOF ACTIONOF ESTROGENS

In light of the foregoing discussion, let us nowconsider studies on the mechanism of action ofestrogenic hormones. This class of hormones,while perhaps not completely representative of thenonaromatic steroids, does constitute a group of

zoo

o(E

oo2o

150

100

50

I oUJo* -50

UH PROTEIN /UM THYMINE

6 12 18 . 24

HOURS AFTER ESTRADIOL

CHART 4.—Alteration in uterine composition followingsingle dose of estradiol.

Ten fig. estradiol injected intravenously at zero time;DNA measured as Amólesthymine. RXA measured as /uñólesuridine and expressed as ratio, Miioles uridine per /¿molesthymine during the first 24 hours following hormonal treatment.2

highly active modifiers of biological function. Oneof their chief attributes is the rapidity and magnitude of the biological response. In addition, theextremely low level of the hormone required foreffect points to a specific receptor mechanism.

A. The biological response.—-A synopsis of the

physiological effects induced by estrogens has beengiven by Pincus (69). The metabolic interactionwith other steroids in the reproductive tract hasbeen summarized in the inclusive reviews bySzego and Roberts (73, 83).

Research. on August 13, 2020. © 1957 American Association for Cancercancerres.aacrjournals.org Downloaded from

498 Cancer Research

Among all the physiological effects of estrogens(a few listed in Section IV), the most striking is itsaction upon the uterus. In response to a singlephysiological dose of a natural estrogen, theatrophie uterus of the ovariectomized female rat israpidly converted to an actively growing tissue;this mobilization of latent growth processes ischaracterized in the first 4 hours by a gross hy-peremia and imbibition of water in the uterine tissue (3); later, increased mitotic activity (2) andconcomitant hypertrophy of the uterus, with increases in ribonucleic acid and protein (3, 84) ; and,finally, new cell production as reflected by increased deoxyribonucleic acid content per uterus(84). To better orient these responses, a chart hasbeen prepared from the data of Telfer (84) (Table2), and additional data from our laboratory areshown in Chart 4 on the early time response interms of protein and ribonucleic acid.2

The whole process takes about 72 hours andpresents a sequential mobilization of cell machinery from growth. The first 6 hours may be regarded as the induction phase; while there has beena dramatic increase in the water content of thetissue (wet weight may double), little or no accumulation of protein or ribonucleic acid occursduring this period (Chart 6). The period from 6 to24 hours may be referred to as the RNA (ribonucleic acid) accumulation phase. Protein accumulation also follows in the wake of the rising RNAcontent per uterus. If the hormonal influence is extended, a DXA (deoxyribonucleic acid) synthesisphase occurs sometime between 40 and 72 hours.The latter transition requires further definition asto time and duration of estrogen dosage requiredto achieve this point.

The water imbibition process during the induction phase constitutes a major problem in estrogenresearch. Cole has concluded that the water imbibition is due to an increase of osmotically activegroups from within the tissue (14). Since the evidence supported a nearly equal increase in negative and positive charges, it was suspected thatprotein was the primary source. However, noclear identification of the osmotically active unitshas been arrived at. Kaiman (44) has demonstrated that uptake of Na22 was greater than could

be explained by the water imbibition and that anincreased adsorption of I131-labeled albumin was

suggestive of greater permeability. Our own studies showed that potassium uptake paralleled thewater imbibition.3

Szego and Roberts have demonstrated that the*K. Jervell, C. R. Diniz, and G. C. Mueller, manuscript

in preparation.3Unpublished data.

water imbibition with critical levels of estradicican be largely blocked by stimulating adrenal cortical activity with ACTH or administering micro-crystalline hydrocortisone acetate or cortisoneacetate intravenously (83). Even though the initial imbibition of water was largely prevented byhydrocortisone acetate, these authors noted thatsignificant growth of the uterus took place subsequently. The anterior pituitary was not necessaryfor the initial response of the subsequent growth.

While these findings on water imbibition havenot been resolved and await basic research onwater and electrolyte transport, they do assure usthat by this time some very fundamental changeshave taken place which may yield to biochemicaldissection.

B. Alterations in uterine metabolism induced byestradiol.—In this section a special effort will be

made to present evidence of early alterations incertain areas of uterine metabolism by estrogenichormones. For the most part the data have beenderived from experiments with isotopie precursorsdesigned to show the sequence of metabolicchanges with respect to time of estrogen treatment. In the procedure utilized in this laboratory,ovariectomized rats are injected with a single doseof estradiol via the tail vein. After varying periodsthe rats are sacrificed and the uterine horns removed and cut into small segments. These segments are incubated at 37°C. in a glucose-bal

anced salt solution containing a radioactive precursor in an atmosphere of oxygen. After incubation the enzymatic reactions are stopped with acidand the tissue fractionated into the following fractions; acid-soluble (small molecules), lipide, nucleic acid, and protein. Specific isolations are carried out on these main fractions. The respired carbon dioxide is collected when indicated.

1. Influence of pretreatment with estrogen onthe incorporation of amino acids into protein: Theeffect of a single injection of estradiol (10 /ng.) onthe rate of incorporation of glycine-2-C14 into the

protein of surviving uterine segments is shown inChart 5 (64). It should be noted that with estradiolthe rate of incorporation of glycine is acceleratedin a linear manner over the first 12 hours; thereafter, gradually, this stimulatory process begins tosubside so as to reach a peak at 20 hours. An estradiol deficiency appears to arise at this time, because if a second injection of estradiol is given at16 hours the acceleration continues in a linearmanner well beyond this point. If the second injection is made on the descending limb of the curve, acomplete reversion is obtained, with a primary response similar to that obtained with the initialdose of estradiol. The failure to respond more rap-

Research. on August 13, 2020. © 1957 American Association for Cancercancerres.aacrjournals.org Downloaded from

MUELLER—Mechanismof Action of Steroid Hormones 499

idly at this point would argue against the accumulation of any apo-enzyme which awaits activation by estradici.

The stimulation of amino acid incorporation occurred with all amino acids tested, i.e., glycine,alanine, serine, lysine, and tryptophan; thus, itseems to be a part of a general protein anabolic response. Since this is one of the earliest metabolicchanges with estrogen treatment (discernible asearly as 1 hour) it should be highly interesting todetermine whether the estrogen effect concerns theactivation of amino acids by the soluble fraction ofthe cell or incorporation of the activated forms intothe ribonucleoprotein granules of the microsome

200

20HOURS

CHART5.—The effect of a single dose of estradici on therate of incorporation of glycine-2-C" into protein of survivinguterine segments.

Ten ng. estradici injected intravenously at zero time. Atthe indicated times uterine segments were incubated withglycine-2-C14for 2 hours. Data are presented in counts/min/mgprotein (64).

fraction (53). For reasons to be discussed later, theactivation process is suspected.

2. Dose-response curve: With a sensitive indicator of estradiol action it was possible to determine dose-response relationships with glycine incorporation as the end-point. The results of suchan experiment are given in Chart 6 (64). Only 0.1ng of estradiol/200-gm rat is required to produce amaximal response when the measurements aremade at 6 hours; however, in order for the activation to continue beyond 6 hours, larger doses ofhormone are needed. It should be noted that with0.1 /ig. the increase obtained between 6 and 20hours was very small; thus, it would appear thatsome receptor was titrated progressively as long asactive estrogen was available. Beyond that point asteady state characteristic of the amount of receptor titrated was maintained for a number of hours.

Can this involve hormone utilization? Onlycareful studies on the amount and form of theestrogen in the uterus with respect to time anddosage will provide an answer. To do this, estradiol with very high specific activity will be required as well as good micro-isolation technics. Itis to be expected that this study may aid in theidentification of the receptor mechanism.

3. Early alterations in carbohydrate and lipidemetabolism induced by estradiol: Szego and Roberts (74, 83) have studied extensively the changesin aerobic and anaerobic metabolism of carbohydrates in rat uteri following estrogen treatment.Within 4 hours after hormone injection, bothaerobic and anaerobic glycolysis were elevated significantly. This increase was not accompanied bya concomitant enhancement of respiration; theextra glucose uptake was reflected in lactic acidaccumulation. By 20 hours, when the uterus waswell into the RNA accumulation phase, the respiration was substantially increased, as well as glycolysis. Data from this laboratory on glucose-l-Cu,

250

0 .001 .01 O.I 1.0 10./ «- ESTRAOIOL

CHART6.—The effect of pretreatment with various dosesof estradiol on the incorporation of glycine-2-C14 into proteinof surviving uterine segments.

Hormone injected at zero time; incorporation study madeafter 6 and 20 hours; 2-hour incubation period (64).

pyruvate-2-C14, and acetate-1-C14 are in essentialagreement with these findings (65); it was notedthat the combustion to carbon dioxide was not significantly affected. Thus, it appears that the earlyestrogen effects in the uteri do not involve alterations in the operation of the tricarboxylic cycle or,therefore, the energy yield from such reactions.

On the other hand, the conversion of the isotopeinto the nonvolatile lipide fraction is stimulated asmuch as 250 per cent by 6 hours. This finding is ofinterest in view of recent studies indicating that

Research. on August 13, 2020. © 1957 American Association for Cancercancerres.aacrjournals.org Downloaded from

500 Cancer Research

fatty acid synthesis takes place in the soluble fraction of the cell (48) and that cholesterol synthesisis concentrated in the large microsomes (9). Theunusual sensitivity of cholesterol synthesis to es-tradiol treatment has been shown by Emmelot andBosch, who observed more than a 30-fold stimulation of acetate incorporation into cholesterol inmouse uterus at 20 hours after hormone administration (22). These workers have used this systemingeniously to study the endogenous production ofestrogens by ovarian tumors (8).

In view of the discussion of the compartmentalnature of metabolism (Section II), these findingson lipide metabolism are highly interesting. Theremarkable estrogenic stimulation of cholesteroland fatty acid synthesis, which are indicated to besoluble and microsomal systems in other tissues,and the lack of stimulation of oxidative pathways,which are located in mitochondria in other tissues,

-14C O2 INTO ADENINE

700

100-

6 12 18HOURS AFTER ESTRADIOL

24

CHART7.—Theeffect of pretreatment with estradici on theincorporation of radioactive carbon dioxide into adenine ofthe mixed nucleic acids of surviving uterine segments.

Ten /ig. estradiol injected at zero time; incubation periodwas 1 hour. Data expressed as counts/min//miole adenine.8

suggest a high selectivity in the site of the estrogenic attack. Before one can be sure, however, agreat amount of work will have to be done towardisolation of the systems in a cell-free state.

4. Early estrogen effects on pathways of ribonu-cleic acid synthesis: In the process of atrophy following ovariectomy, the uterus of the rat decreases to a very low level of RNA per unit ofDNA (84).2 Following a single administration ofestradiol one observes little or no change in RNAcontent for 6 hours, after which interval it rapidlyaccumulates, as shown in Table 2 and Chart 4

(84) .2When incorporation studies were done withradioactive carbon dioxide during this period, therate of labeling of the nucleic acid adenine wasfound to be increased at least 150 per cent by 6hours (Chart 7).2 The rate of incorporation fell offrapidly after 12 hours, even though the RNA wasstill accumulating at this time. Apparently, theestrogen had set in motion a process in which theinitial precipitating effects were wearing off andwere becoming obscured by subsequent metabolicchanges. It would be very interesting to know thecharacter of this response with varying duration of

500Lu-J4000SE

§300Ü2•x

200CL;^

100GLYCINEFORMATEFORMATE-C14A•

fp~,•.dti4t/

1t

'///t

1//

1:

ri!11<///rtt

t/^y|/J^/^/^/^tAr*a////^//^/sÕG^LYCINE-2-C14+u,'''p;;;s

.G=GUANINE;A A»ADENINE/

r1/«O

CONTROLE2TREATEDGr"\

r^-^/>+.n['/''s+CHART8.—Theeffect of pretreatment with estradiol on the

incorporation of labeled formate and glycine into nucleicacid purines of surviving uterine segments.

Uterine segments were from rats pretreated for 6 hourswith a single 10-^g. dose of estradiol. Incubation period was2 hours with labeled formate or glycine. Data expressed ascounts /min /Vmole purine (67).

estrogenic treatment (i.e., small doses over extended periods).

Results with radioactive formate and glycine(Chart 8) parallel the response obtained withradioactive carbon dioxide at 6 hours and, in addition, have revealed a high sensitivity of "one-carbon" metabolism to estrogen influence (67). Thisis particularly well illustrated with serine-3-C14 asthe precursor (Chart 9) of the "one-carbon" fragment (38). With 6 hours' pretreatment the in

vitro labeling of the nucleic acids purines was 4.4times the control value. The inclusion of a non-radioactive pool of formate in the reaction mixtureacted as a trap for the one-carbon intermediateand decreased the incorporated radioactivity inthe nucleic acid purines strikingly (38). On examination of the pool it was clear that one estrogeneffect occurs at or near the activation step for

Research. on August 13, 2020. © 1957 American Association for Cancercancerres.aacrjournals.org Downloaded from

MUELLER—Mechanismof Action of Steroid Hormones 501

serine; this, however, does not account for thelarge effect observed on the total pathway; theremust be other steps which are highly susceptible tothe action of estradiol. Similar results have beenobtained with glycine-2-C14 (67).

Another finding in these experiments with labeled serine, glycine, and formate was that variouspathways for a single precursor are not altered tothe same extent by estrogen pretreatment. Thisappears to point to a selective sensitivity of competing enzymes to the action of the hormone; how-

24ICROMOLE

—¿�roCDO|«Os^o.4GLYCINEFORMATE.:-.1lA!¿^<\NNSSNNNSNaC3CONJRQL

EZaTREATEDAi1¿G= GUANINE

A =ADENINEG

GñAÖAñ\m+

+•++

CHART9.—Theeffect of pretreatment with estradiol on theincorporation of serine-3-C14 into nucleic acid purines ofsurviving uterine segments.

Uterine segments were from animals pretreated for 6 hourswith a single 10-jug.dose of estradiol. Incubated with serine-8-Cu for 2 hours. Data expressed as counts/min/Vmolepurine (88).

ever, if two enzymes with widely different Michaelis constants were competing under conditions oflimiting substrate concentration, an alteration ofthe substrate concentration might well account forthe observed results. One can only proceed toanalyze the situation stepwise in line with the discussion in Sections II and III, keeping strongly inmind the compartmental nature of the cell and theproperties of multiple enzyme systems.

The segmental nature of estrogen influence onmetabolism is also highly apparent when one compares the effects on carbon dioxide incorporationinto pyrimidines (Chart 10) with the results justreported for purines. Prior to 6 hours, little or noeffect was seen on the total radioactivity incorporated into pyrimidines. Thus, pyrimidine synthesis, which from studies on liver appears to origi

nate in the mitochondria, does not reflect the initial estrogen influence when carbon dioxide is theprecursor.

5. Effect of estradiol pretreatment on serinealdolase activity of the uterus: The high sensitivity of one-carbon metabolic pathways in the uterine segment to estrogen action prompted a studyof serine synthesis in uterine homogenates. Attempts to convert radioactive formate to serinefailed, owing to difficulty in the activation offormate. However, homogenates fortified withtetrahydrofolic acid, nonradioactive formaldehyde, and glycine-2-C14 synthesized serine readily.The process was linear with time and proportionalto tissue concentration; accordingly, uterine homogenates from rats pretreated with a single dose

INTO URIDINE

200

OCCh

ISO

8 too

IÜ- Zifì

^ OLUO

è?-50

/JM URIDINE /THYMINE

TOTAL C. P. M. INURIDINE /

THYMINE

C.P.M.

I6 12 18 24

HOURS AFTER ESTRADIOL.CHART10.—The effect of pretreatment with estradiol on

incorporation of radioactive carbon dioxide into uridine (R^A)of surviving uterine segment.

Ten fig. estradiol injected at zero time: incubation period,1 hour. Data expressed as total counts/min in uridine //uniólethymine. Ratio of uridine to thymine (RXA/DNA) is alsoshown.8

of estradiol were assayed for serine aldolase activity. As demonstrated in Chart 11, serine aldolaseactivity increases linearly with time of estrogenpretreatment.4

Thus, the estrogen-induced alterations in one-4A. Herranen and G. C. Mueller, manuscript in prepara

tion.

Research. on August 13, 2020. © 1957 American Association for Cancercancerres.aacrjournals.org Downloaded from

502 Cancer Research

carbon metabolism may in part be explained byincreased serine aldolase activity. This introducesfurther questions whether the observed results aredue to activation of pre-existing enzyme (unmasking) or due to de novo synthesis of a new enzyme.Answers to these questions should place us closerto an understanding of estrogen action.

C. In vitro effects with estrogens on intact tissues.—¿�Amajor obstacle to progress in steroid hormoneresearch has been the inability to obtain physiological effects when a hormone is added to in vitrosystems. As a result, the question of hormone activation and possible dependency on other tissueshas arisen. However, in the case of estrogen,Biggers, Claringbold, and Hardy have demonstrated that estrone produces a typical keratiniza-tion of mouse vagina in tissue culture; thus theyhave excluded or de-emphasized a number of theories of estrogen action which involve vascularchanges or hormonal activation by other tissues(5). A similar situation was reported by Kahn,who also observed an antagonistic effect of vitamin A on the estrogen action (43). Further studyof this important finding promises to be interesting.

In studying factors which modify mitotic activity, Bullough has reported that the addition ofrelatively high levels of estrogens to mouse skinincubated in saline containing glucose yields moremitoses in a 4-hour period (10). He concludes thatthis effect results from a stimulation of the glucoki-nase activity of this tissue which normally limitsthe glucose uptake in skin and the availability ofenegy for mitoses. The estrogen effect can be bypassed by using fructose instead of glucose; apparently, fructokinase is not limiting or estrogen-sensitive.

The addition of estradici, estrone, or estriol tosurviving uterine segments failed to influence theuptake of labeled formate into protein, whereas the2-hydroxy and the 4-hydroxy derivatives of estradici gave a significant "estrogen-like" stimulation

to this process (66). It remains to be establishedwhether or not the other metabolic pathways for"one-carbon" fragments are stimulated in a man

ner typical of estradici in vivo. Of even greater interest is the question of whether or not thesehydroxylated derivatives of estradici will effect anin vitro activation of serine aldolase.

D. In vitro effect of estrogens on enzyme systems.—¿�Workingwith human endometrium tissue,Hagerman and Villee demonstrated that naturalestrogens stimulated the metabolism of glucoseand pyruvate toward more complete oxidation(35). This very fine study was extended to theplacenta, where a soluble diphosphopyridine nu

cleo tide-linked isocitric dehydrogenase has beenfound which is stimulated by the addition of estradici, estrone, equilin, and equilenin (86). It wasalso stimulated to a much lesser extent by estriol,progesterone, diethylstilbestrol, and testosterone.The relatively high specificity for natural estrogenhas permitted the assay of estradiol-17)3 andestrone (29). While the actual manner in which thehormone participates in this system is unclear, it ispossible that it may provide a model system ofenzyme activation by a hormone.

The very fundamental work of Talalay and associates, using purified enzymes of bacterial origin,has established the molecular specificity for thecombination of steroids with (3-hydroxysteroid de-

25

20

go

Bx

U 5

CONVERSION OF GLYCINE-2-C14 INTO SERINE 8Y

UTERINE HOMOGENATES

02468IOI2MI6I8HOURS AFTER ESTROCEN INJECTION

'CHART11.—Effectof estradici pretreatment on the levels

of serine aldolase in the rat uterus.Ten per cent homogenates were prepared from uterine

horns of rats pretreated with l<Vg- estradici for the indicatedtime. Reaction mixture contained 10 mg. tissue, 3 /¿molesglycine-2-C14 (1.5 mc/fimoles), 3 /»molesformaldehyde, and1.5 /mioli's tetrahydrofolic acid. Incubation period was 10minutes. Data expressed as counts/min in serine/mg proteinresidue.4

hydrogenase (59). It was shown that planar molecules of the androstene and 5 a-androstane serieshad intermediate affinities, while the natural estrogens had a very high affinity for the enzyme surface. This work appears to be related to that ofVillee in that both cases involve dehydrogenationreactions.

The possibility that the estrogens might exerttheir biological effects in this manner is of courseintriguing. However, at this time no evidence isavailable on the operation of such mechanismswithin the tissue. In Villee's studies, the relatively

high specificity for natural estrogens argues againstthis reaction as being a primary step in the mechanism of estrogen action, since certain synthetic

Research. on August 13, 2020. © 1957 American Association for Cancercancerres.aacrjournals.org Downloaded from

MUELLER—Mechanism of Action of Steroid Hormones 503

compounds of high biological activity were notactive in his system.

Further work in the very important area of hormone-enzyme interaction should be pursued, sinceit is difficult to escape the conclusion that the effective action of the estrogen must be exerted on anenzyme acting in a multiple enzyme system.

VI. COMMENTSONTHE MECHANISMOF ACTIONOF OTHERSTEROIDHORMONES

While many descriptions of biological responseshave been made for other steroid hormones, veryfew reports concerning their fundamental mechanisms of action have appeared. Accordingly, it ispossible at this time only to point up a few problems where metabolic attacks are urgently needed.The tremendous importance of early biochemicalstudies in responsive tissues can scarcely be overemphasized. Of equal importance is the establishment of dose-response relationships in terms of themetabolic processes with respect to time.

A. Androgenic hormones.—An inclusive reviewof the metabolic effects of androgens has beenmade by Roberts and Szego (73), and the morerecent aspects of the relationship of androgens toreproduction have been emphasized by Mann(58). For a very comprehensive account of thewhole subject of androgens the reader is referredto the compendium prepared by Dorfman andShipley (17).

While many toxicity studies with androgens onenzyme systems have been carried out (summarized, 73) it is not possible at this time to establishthe relevance of these experiments to the problemof androgenicity. It appears that it will be necessary first to establish the metabolic parameters ofthe androgenic response.

Three sites of rapid and relatively specific response are the chick (or capon) comb, and theseminal vesicles and prostate in the rat. The chickcomb appears to offer some advantage for biochemical attack, since it has been demonstratedthat the glucosamine and mucopolysaccharidecontents are increased strikingly in response toandrogens (54, 77). This is an invitation to beginthe metabolic dissection at this point in terms ofthe reactions involved in polysaccharide synthesis.Considerable knowledge on the participation ofuridine coenzymes in these reactions already exists. In addition, the isotopie precursors and isolation technics are available.

Attention is again drawn to this area by the riseof glucuronidase in the kidney following androgentreatment (25). Possibly the rise of this enzyme inthe kidney is an induced enzyme synthesis in response to higher glucuronide levels in the kidney.Whatever may be the answer, this area appears

vulnerable to attack with our present tools andtechnics.

In this same light, the finding of fructose, citricacid, and ergothionine in the fluids of the accessoryorgans of reproduction (58) offers other leads forthe study of the mechanism of androgen action.The recent observation by Hers that fructose formation in the seminal vesicles involves a direct reduction of glucose to sorbitol with a reoxidation tofructose without involving the phosphorylatedsugars suggests another opportunity for biochemical attack on androgen action (39).

The demonstration by Lasnitzki that ventralprostate proliferates in tissue culture under theinfluence of added testosterone is a move in thedirection of a controlled environment for a hormonal response (49). To be of biochemical value,however, such preparations have to be expandedto large-scale cultures.

B. Progestational hormones.—Studies on themechanism of action of progesterone are complicated by a relative dependency on estrogen priming; however, under standardized conditions thetraumatic deciduomata (4) provides a rapid andspectacular biological response. This tissue shouldlend itself to a sequential analysis. The correlationbetween carbonic anhydrase level and progesterone treatment indicates again a segmental effecton metabolism (55).

The rather striking activation of mitochondria!adenosinetriphosphatase from liver (87) is an interesting lead; however, at this time the relationship to progestational effects is unclear. It is of interest that progesterone also depresses glycine incorporation in surviving uterine segments whenadded to the reaction flask (64). The inhibited tissue relaxes or becomes flaccid under its influence.

C. Adrenal cortical hormones.—Foran excellentsummary of the literature on adrenal corticalphysiology, the reader is referred to a review byNoble (68). Despite the multitude of publicationson adrenal cortical steroids, the fundamentalmechanism of action has resisted elucidation.While the 11-oxygenated corticosteroids have effects on carbohydrate and protein metabolism, nocommon denominator has emerged.

The striking involution of the thymus glandunder hydrocortisone treatment now appears toresult mainly from blockage of new cell formation.Thus, the incorporation of P32 into deoxyribonu-cleic acid is highly inhibited, while the labeling ofribonucleic acid proceeds at a nearly normal rate(13). Since there is also increased cell fragmentation, one cannot rule out completely cell destruction as an important factor in thymus maintenance. WTithlow doses of cortisone the incorporation of labeled amino acids into thymus is also in-

Research. on August 13, 2020. © 1957 American Association for Cancercancerres.aacrjournals.org Downloaded from

504 Cancer Research

hibited; in contrast, the incorporation into theliver is simultaneously stimulated.8

The studies of Dougherty and associates on inflammation have revealed that hydrocortisone isconcentrated in the cytoplasm of fibroblasts withina few hours after topical application. These cellsround up and exhibit increased cytoplasmic baso-philia and begin to look more epithelioid (19).Under the influence of the hydrocortisone they resist damage from cytotoxic substances in areas ofinflammation (19); the authors suggest that thecells become impermeable to such factors.

Thus, while the adrenal steroids exert strikingeffects on "connective tissue" elements, particu

larly in so-called connective tissue diseases, verylittle insight has been gained as to the fundamentalsite of action. It appears that a really good testsystem will have to be established before muchprogress can be made. We are especially limited byour lack of knowledge concerning the formation ofcollagen and the materials included in the "groundsubstance" lying between cells.

Our knowledge of the mechanism of action ofaldosterone and deoxycorticosterone in regulatingelectrolyte balance is similarly deficient. Here itappears that fundamental knowledge on ion transport will have to be established before much progress can be expected. These hormones should constitute excellent tools for this study.

The studies of Stolkowski and Reinberg (71, 81,82) demonstrate an effect of deoxycorticosteroneon potassium transport in isolated systems. Intheir studies the corticosteroids inhibited the entrance of potassium, whereas the hydrogen ionproduction from glucose metabolism acts to displace potassium from the cell. While it is too earlyto conclude anything concerning deoxycorticosterone action, this approach promises to lead to somefundamental discoveries.

CONCLUDINGCOMMENTSIn this discussion, the mechanism of action of

steroid hormones has been considered as a problem in fundamental cell physiology. The units ofstructure and function are the enzymes as they existin specific association with other enzymes in semi-isolated compartments. In this state the individualenzymes compete as members of multiple enzymesystems; the amount of each enzyme, their inherent catalytic activity, and substrate availability determine a particular steady state characteristic of a certain level or type of function.

To modify function, a steroid hormone must actwithin these confines so as to induce a shift in thebalance of power between various cell processes.

5Personal communication from Dr. R. K. Boutwell.

To accomplish this, it appears necessary that theactive form of the steroid hormone interact primarily with an enzyme to produce an activation orinhibition of this unit for the catalysis of a specificprocess. This alteration, in turn, is amplified andexpressed through the establishment of a differentsteady state among the various multiple enzymesystems of the cell.

The importance of the compartimentai nature ofthe cell for organized function is emphasized. It issuggested that communication between such lociof function may be subject to hormonal influencethrough altered rates of formation or destruction oflimiting membranes or interfaces. It is also suggested that electrostatic effects, analogous tochanges taking place in nerve cell membranes or incontractile proteins, may be responsible for activating certain enzymes or increasing the accessibility of enzymes to substrates. This area of research offers exciting opportunities for discovery.

It is proposed that attempts to explain steroidhormone action in molecular terms will profit byfirst searching out the segments of metabolismwhich are responsive to the early action of a hormone. Guided by the results of such surveys onintact tissues, one can next begin to analyze theobserved effects in cell-free systems. In Section V, asummary of studies on the mechanism of action ofestrogens was presented as an example of this approach. The data support the conclusion that synthetic pathways involving carboxyl activation and"one-carbon" transfer are very sensitive to estro

gen action. A preliminary attempt to analyze theestrogen effect on one-carbon metabolism hasdemonstrated that the level of serine aldolase increases rapidly in uterine tissue in response toestradici. The relevance of this finding to themechanism of action of estrogen as well as the further analysis of other sensitive pathways in cell-free systems awaits study.

Despite the fact that at this time we do notknow the mechanism of action for any steroid hormone, this author feels highly optimistic concerning the future. The much needed information onfundamental cell processes is being accumulated ata tremendous rate; it remains for the discerning investigator to put these findings to use along thelines discussed. In this manner, the steroid hormones can be expected to serve as useful tools inthe study of important problems in health and disease.

ACKNOWLEDGMENTS

The author wishes to thank Drs. A. Herranen, K. Jervell,and I. Riegel for their stimulating discussions and directassistance in the preparation of this manuscript.

Research. on August 13, 2020. © 1957 American Association for Cancercancerres.aacrjournals.org Downloaded from

MUELLER—Mechanismof Action of Steroid Hormones 505

REFERENCES1. AGBELL,I., and PERSSON,H. The Action of Oestradiol on

Desoxyribonucleoprotein as the Cause of Its InhibitoryEffect upon Mitosis in the Sea Urchin Embryo. Biochim. &Biophys. Acta, 20:543-46,1956.

2. ALLEN,E.; SMITH,G. M.; and GARDNER,W. U. Growthof Ovaries and Genital Tract in Response to Hormonesas Studied by the Colchicine Technique. Anat. Ree., 67(Suppl.3):S-4,1937.

3. ASTWOOD,E. B. A Six-hour Assay for the QuantitativeDetermination of Estrogen. Endocrinology, 23:25-31.1938.

4. . An Assay Method for Progesterone Based upon theDecidua! Reaction in the Rat. J. Endocrino!., 1:49-55,1939.

5. BIOOERS,J. D.; CLARINOBOLD,P. J.; and HARDY,M. H.The Action of Oestrogens on the Vagina of the Mouse inTissue Culture. J. Physiol., 131:497-515, 1956.

6. BISCHOFF,F., and STAUFFER,R. D. The Dispersion ofTestosterone in Aqueous Bovine Serum Albumin Solution. J. Am. Chem. Soc., 76:1962-65, 1954.

7. BISCHOFF,F.; STAUFFER,R. D.; and GRAY,C. L. Physio-C'hernical State of the Circulating Steroids. Am. J. Physiol.,177:65-68, 1954.

8. BOSCH,L., and EMMELOT,P. The Metabolic Activity ofthe Uterus in Ovariectomized Mice Bearing OvarianTumors of the Granulosa Cell Type. Archiv, fürGynek.fin press).

9. BUCHER,X. L. R., and MCGARHAHAN,K. Rat LiverMicrosomes in Cholesterol Biosynthesis. Fed. Proc.,14:187, 1955.

.10. BULLOCOH,W. S. Hormones and Mitotic Activity. In:Vitamins and Hormones. New York: Academic Press, Inc.,1955.

11. CALLOW,R. K., and DEANESLY,R. Effect of Androsteroneand of Male Hormone Concentrates on the AccessoryReproductive Organs of Castrated Rats, Mice and GuineaPigs. Biochem. J., 29:1424-45, 1935.

12. CHANCE,B. Interaction of Adenosine Diphosphate withthe Respiratory Chain. In: Enzymes: Units of BiologicalStructure and Function. New York: Academic Press, Inc.,1956.

18. CLARK,I., and STOERK,H. C. The Uptake of P* by NucleicAcids of Lymphoid Tissue Undergoing Atrophy. J. Biol.Chem., 222:285-92, 1956.

14. COLE,D. F. The Effects of Oestradiol on the Rat Uterus.J. Endocrino!., 7:12-23, 1950.

15. CORI,C. F. Problems of Cellular Biology. In: D. E. GREEN(ed.), Currents in Biochemical Research. New York:Interscience Publishers, 1956.

16. DORFMAN,R. I.; POTTS,A. M.; and FEIL, M. L. Studies onthe Bioassay of Hormones: The Use of Radiosodium forthe Detection of Small Quantities of Desoxycorticosterone.Endocrino!., 41:464-69, 1947.

17. DORFMAN,R. I., and SHIPLEY,R. A. Androgens. NewYork: John Wiley & Sons, Inc., 1956.

18. DORFMAN,R. I., and UNGAR,F. Metabolism of SteroidHormones. Minneapolis, Minn.: Burgess Publishing Co.,1953.

19. DOUGHERTY,T. F., and SCHNEEBELI,G. L. The Use ofSteroids As Anti-Inflammatory Agents. Ann. N.Y. Acad.Sc., 61:328-48, 1955.

aO. EIK-NES, K.; SCHELLMAN,J. A.; LUMRY,R.; and SAMUELS,L. T. The Binding of Steroids to Protein. J. Biol. Chem.,206:411-19, 1954.

21. EISENBERG,E., and GORDON,G. S. The Levator Ani Muscle of the Rat as an Index of Myotropic Acting of Steroid»!Hormones. J. Pharmacol. & Exper. Therap., 99:38-44,1950.

22. EMMBLOT,P., and BOSCH,L. The Influence of Oestrogenson the Protein and Lipid Metabolism of the MouseUterus Studied with Acetate-1-C". Ree. Trav. Chim.,73:874-77, 1954.

23. EMMENS,C. W. Hormone Assay. New York: AcademicPress, Inc., 1950.

24. FELDMAN,J. D. The In Vitro Reaction of Cells to AdrenalCortical Steroids with Special Reference to Lymphocytes.Endocrino!., 46:552-62, 1950.

25. FISHMAN,W. H. ßGlucuronidase and the Action ofSteroid Hormones. N.Y. Acad. Sc., 54:548-57, 1951.

26. FRANK,R. T.; KLEMPNER,E.; HOLLANDER,F., and KRISS,B. Detailed Description of Technique for Androgen Assayby the Chick Comb Method. Endrocrinology, 31:63-70,1942.

27. FRANTZ,M. J., and KIRSCHBAUM,A. Sex Hormone Secretion by Tumors of the Adrenal Cortex of Mice. CancerResearch, 9:257-66, 1949.

28. FRIEDEN,E. H., and MARTIN,A. S. Uptake of Glycine-1-C14 by Connective Tissue. J. Biol. Chem., 207:133-42,1954.

29. GORDON,E. E., and VILLEE,C. A. An In Vitro Assay forEstradiol-17/3 and Estrone. Endocrinol., 58:150-57, 1956.

30. GREEN, D. E. Enzymes and Trace Substances. In: Advances in Enzymology. New York: Interscience Publishers,1941.

31. . Fatty Acid Oxidation in Soluble Systems of Animal Tissues. Biol. Rev., 29:330-66, 1954.

32. . Structural and Enzymatic Pattern of the ElectronTransfer System. In: Enzymes: Units of Biological Structure and Function. New York: Academic Press, Inc., 1958.

33. GBEENBERG,G. R. Problems in the Study of MultipleEnzyme Systems. In: Currents in Biochemical Research.New York: Interscience Publishers, 1956.

34. CREEP, R. O., and JONES, I. C. Steroid and PituitaryHormones. In: E.S. GORDON(éd.),A Symposium on SteroidHormones. Madison, Wis.: Univ. of Wisconsin Press, 1950.

35. HAGERMAN,D. D., and VILLEE,C. A. Effects of Estradiolon the Metabolism of Human Endometrium In Vitro.J. Biol. Chem., 203:425-31, 1953.

36. HALVORSON,H., and JACKSON,L. The Rektion of RiboseNucleic Acid to the Early Stages of Induced EnzymeSynthesis in Yeast. J. Gen. Microbio!., 14:26-37, 1956.

37. HECHTER,0. Concerning Possible Mechanisms of Hormone Action. In: Vitamins and Hormones. New York:Academic Press, Inc., 1955.

38. HERRANEN,A., and MUELLER,G. C. Effect of Estradiolon the Metabolism of Serine-3-C1* in Surviving UterineSegments. J. Biol. Chem., 223:369-75, 1956.

39. HERS, H. G. Le Mécanismede la transformation deglucose en fructose par la vésiculesséminales.Biochim. &Biophys. Acte, 22:202-3, 1956.

40. HOOKER,C. W., and FORBES,T. R. A Bioassay for MinuteAmounts of Progesterone. Endocrinology, 41:158-69,1947.

41. HUGGINS,C., and JENSEN, E. V. The Depression of Es-trone-induced Uterine Growth by Phenolic Estrogens withOxygenated Functions at Positions 6 or 16; The ImpededEstrogens. J. Exper. Med., 102:335-46, 1955.

42. . The Depression of Growth of the Uterus, Adrenals,and Ovaries by Fluorinated Steroids in the PregnaneSeries. Ibid., pp. 347-60.

43. KAHN,R. H. Effect of Oestrogen and Vitamin A on VaginalCornification in Tissue Culture. Nature, 174:317, 1954.

44. KALMAN,S. M. Some Studies on Estrogens and UterinePermeability. J. Pharm. & Exper. Therap., 116:442-48,1955.

45. KNOX,W. E. Two Mechanisms Which Increase in Vim theLiver Tryptophan Peroxidase Activity. Specific Enzyme

Research. on August 13, 2020. © 1957 American Association for Cancercancerres.aacrjournals.org Downloaded from

506 Cancer Research

Adaptation and Stimulation of the Pituitary-Adrenal 67.System. Brit. J. Exper. Path., 32:46-2-69, 1951.

46. KNOX, W. E.; ACERBACH,V. H.; and LIN, E. C. C.Enzymatic and Metabolic Adaptations in Animals. 68.Physiol. Rev., 36:164-251, 1956.

47. KNOX,W. E., and MEHLER,A. H. The Adaptive Increaseof the Tryptophan Peroxidase-Oxidase System of Liver. 69.Science, 113:237-88, 1951.

48. LANGDON,R. G. The Requirement of TriphosphopyridineNucleotide in Fatty Acid Synthesis. J. Am. Chem. Soc., 70.77:5190-92,1955.

49. LASNITZKI,I. The Effect of Testosterone Propionate onOrgan Cultures of the Mouse Prostate. J. Endocrino!., 71.12:236-40, 1955.

50. LEHNINOER,A. L. Physiology of Mitochondria. In:Enzymes: Units of Biological Structure and Function. 72.New York: Academic Press, 1956.

51. LEWT, G. A. Glucuronide Metabolism, with Special Reference to the Steroid Hormones. In: Vitamins and Hor- 78.mones. Vol. 14. New York: Academic Press, Inc., 1956.

52. LIEBERMAN,S., and TEICH, S. Recent Trends in the Biochemistry of the Steroid Hormones. Pharm. Rev., 6:285- 74.380, 1953.

53. LITTLEFIELD,J. W.; KELLER, E. B.; GROSS, J.; and 75.ZAMECNIK,P. C. Studies on Cytoplasmic Ribonucleo-protein Particles from the Liver of the Rat. J. Biol. Chem., 76.217:111-23, 1955.

54. LUDWIG,A. W., and BOAS,N. F. The Effects of Testosterone on the Connective Tissue of the Comb of theCockerel. Endocrinol., 46:291-98,1950. 77.

55. LUTWAK-MANN,C. Carbonic Anhydrase in the FemaleReproductive Tract. Occurrence, Distribution and Hormonal Dependence. J. Endocrinol., 13:26-38,1955.

56. MACKLEH,B., and GREEN,D. E. Studies on the Electron 78.Transport System. II. On the "Opening" Phenomenon.Biochim. & Biophys. Acta, 21:1-6, 1956.

57. . Studies on the Electron Transport System. III. 79.The Properties of Two Modified Forms of ETP. Ibid., pp.6-13.

58. MANN,T. Male Sex Hormone and Its Role in Reproduction. In: Recent Progress in Hormone Research, Vol. 12. 80.New York: Academic Press, Inc., 1956.

59. MARCUS,P. I., and TALALAY,P. On the Molecular Speci- 81.ficity of Steroid-Enzyme Combinations. Proc. Royal Soc.,s.B, 144:116-32, 1956.

60. MATHIESON,D. R., and HATS,H. W. The Use of Castrate 83.Male Rats for the Assay of Testosterone Propionate. Endo-cinology, 37:275-79, 1945.

61. METER, A. S. Conversion of 19-Hydroxy-A4-androstene- 83.3,7-dione to Estrone by Endocrine Tissue. Biochim. &Biophys. Acta, 17:441-42, 1955.

62. METER, R. K., and CLIFTON,K. H. Effect of Diethyl 84.Stilbestrol-Induced Tumorigenesis on the Secretory Activity of the Rat Anterior Pituitary Gland. Endocrinol.,68:686-93, 1956. 85.

63. MONOD,J. Remarks on the Mechanism of Enzyme Induction. In: Enzymes: Units of Biological Structure andFunction. New York: Academic Press, Inc., 1956. 86.

64. MUELLEK,G. C. Incorporation of Glycine-2-C14into Protein by Surviving Uteri from EstradioI-Treated Rats.J. Biol. Chem., 204:77-90,1953.

65. . The Effect of Estrogen Pretreatment on Amino 87.Acid Incorporation in Surviving Rat Uteri. J. Clin.Endocrinol., 13:836-1954.

66. . In Vitro Stimulation of Incorporation of Formate- 88.C14 in Surviving Uterine Segments by HydroxylatedOestradiols. Nature, 176:127,1955.

MUELLER,G. C., and HERRANEN,A. Metabolism of 1-Carbon Fragments by Surviving Uteri from EstradioI-Treated Rats. J. Biol. Chem., 219:585-94, 1956.NOBLE, R. L. Physiology of the Adrenal Cortex. In:G. PINCUSand K. V. THIMANN(eds.), The Hormones,Vol. 3. New York: Academic Press, Inc., 1955.PINCUS,G. The Physiology of Ovarian and Testis Hormones. In: The Hormones, Vol. 3. New York: AcademicPress, Inc., 1955.RANNET,R. E., and CHAIKOFF,I. L. Effect of FunctionalHepatectomy upon Estrogen-Induced Lipemia in theFowl. Am. J. Physiol., 165:600-603, 1951.REINBERG,A., and STOLKOWSKI,J. Action de la désoxy-corticostéronesur le K cellulaire (mise en évidenceàl'aidede radio-potassium). Ann. Endocrinol. 13:599-609, 1952.RIEGEL,I. L., and MUELLER,G. C. Formation of a Protein-Bound Metabolite of Estradiol-16-C14 by Rat LiverHomogenates. J. Biol. Chem., 210:249-57, 1954.ROBERTS,S., and SZEGO,C. M. Steroid Interaction in theMetabolism of Reproductive Target Organs. Physiol.Rev., 33:593-629, 1953.

. The Influence of Steroids on L'terine Respirationand Glycolysis. J. Biol. Chem., 201:21-30, 1958.

—¿�.Biochemistry of Steroid Hormones. In: Ann. Rev.Biochem. Stanford, Calif.: Annual Reviews, Inc., 1955.SEVERIXGHACS,E. L. The Immature Rat Uterus in theAssay of Estrogenic Substances, and a Comparison ofEstradici, Estrone and Estriol. Endocrinol., 24:85-44,1939.SHILLEB,S.; BENDITT,E. P.; and DOHFMAN,A. Effect ofTestosterone and Cortisone on the Hexosamine Contentand Metachromasia of Chick Combs. Endocrinol., 60:504-10, 1952.SLONIMSKI,P. P. Formation des enzymes respiratoireschez la levure. ActualitésBiochimique, 17, Desoer,Liège,1953.SPEIRS,R. S., and METER, R. K. A Method of AssayingAdrenal Cortical Hormones Based on a Decrease in theCirculating Eosinophil Cells of Adreualectoiuized Mice.Endocrinol., 48:316-26, 1951.STANIER,R. Y. Problems of Bacterial Oxidative Metabolism. Bact. Rev., 14:179-91, 1950.STOLKOWSKI,J. and REINBERG,A. Les Modalitésde l'appauvrissement des cellules en potassium sous l'action dela désoxycorticostérone.Arch. Se. Physiol., 8:51-60, 1954.

. Recherches sur le méchanismede l'action de

quelques hormones corticostéroidessur le potassium cellulaire. Ann. Endocrinol., 17:137-59, 1956.SZEGO,C. M., and ROBERTS,S. Steroid Action and Interaction in Uterine Metabolism. Recent Prog. Hormone Research, 8:419-69,1953.TELFER, M. A. Influence of Estradici on Nucleic Acids,Respiratory Enzymes, and the Distribution of Nitrogenin the Rat Uterus. Arch. Biochem., 44:111-19, 1953.VENNING,E. H.; KAZMIN,V. E.; and BELL, J. C. Biological Assay of Adrenal Corticoids. Endocrinol., 38:79-89, 1946.VILLEE,C. A., and GORDON,E. E. The Stimulation byEstrogens of a DPN-Linked Isocitric Dehydrogenasefrom Human Placenta. Bull. Soc. Chim. Belg., 65:186-201, 1956.WADE,R., and JONES, H. W. Effect of Progesterone onMitochondrial Adenosmetriphosphatase. J. Biol. Chem.,220:547-51, 1956.ZIEGLER,D.; LESTER,R.; and GREEN, D. E. OxidativePhosphorylation by an Electron Transport Particle fromBeef Heart. Biochim. & Biophys. Acta, 21: 20-85, 1956.

Research. on August 13, 2020. © 1957 American Association for Cancercancerres.aacrjournals.org Downloaded from

1957;17:490-506. Cancer Res   Gerald C. Mueller  A Discussion of the Mechanism of Action of Steroid Hormones

  Updated version

  http://cancerres.aacrjournals.org/content/17/5/490.citation

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/17/5/490.citationTo request permission to re-use all or part of this article, use this link

Research. on August 13, 2020. © 1957 American Association for Cancercancerres.aacrjournals.org Downloaded from