Androgenic Control of Nucleic Acid and Protein Synthesis in Male Accessory Genital Organs

H. G. WILLIAMS - ASHMAN Brady Urological Institute, The Johns Hopkins Hospital, and Department of Pharmacology and Experimental Therapeutics, The Johns Hopkins School of Medicine, Baltimore, Maryland

ABSTRACT A survey is given of experimental studies on the influence of treatment with androgenic hormones in vivo on various intermediary reactions involved in ribonucleic acid (RNA) and protein synthesis in the prostate gland and seminal vesicle, with particular reference to the control of the growth and functional differentiation of these organs by testosterone and related steroids. Studies on the influence of androgens on RNA metabolism and protein biosynthesis in mouse kidney, certain muscles, and some other extragenital tissues are also considered.

An ever-increasing amount of experi- comprehensive molecular theories of sex mental effort has been expended over the hormone action. last 5 years toward examining various in- (1) It is well established that common termediate reactions involved in ribonu- pathways exist for the biosynthesis of cleic acid and protein synthesis in male androgens and estrogens in the ovary, accessory genital glands. Much of this testis, adrenal cortex, and placenta. These work has been carried out vis-a-vis the all- two categories of sex hormones can be embracing effects of androgens on the elaborated in the aforementioned steroid growth and functional differentiation of factories in both male and female mam- these organs. The strategy and tactics of mals. Progesterone appears to be a com- these researches have often resembled mon intermediate in the transformation of those employed in comparable investiga- cholesterol into both androgenic and estro- tions on the action of estrogenic hormones genic steroids, as well as adrenocortical on various structures in the female genital hormones. tract. (Comparatively little attention has (2) Whereas estrogenic activity is a been given to the nature and chronology property of many types of nonsteroidal of molecular events that underlie the bio- molecule, very few substances that are not logical actions of gestagens.) The remark- steroids have been found to be androgenic, able progress in our understanding of the and those that have are only feebly active. role of various forms of ribonucleic acid in (3) The natural ovarian estrogens, and gene expression and protein synthesis, to- their more potent non-steroidal synthetic gether with the failure of many attempts congeners (such as diethylstilbestrol), ex- to explain the action of sex hormones in ert their estrogenic effects in doses which terms of their direct effects on either iso- are two to three orders of magnitude lower lated enzyme systems or on the permeabil- than those at which testosterone displays ity of cell or organelle membranes, has its major androgenic actions. naturally focused attention on the sex hor- (4) Although there is evidence that monal control of the synthesis and turn- estradiol-17B can be concentrated in cer- over of specific enzymes and structural tain female secondary sexual tissues, and proteins. can induce uterine growth without under- Before considering some phenomenolog- going any chemical change (Jensen, '63), ical details of these processes in the male metabolic transformation products of tes- reproductive tract, it may be well to list tosterone and A'-androstene-3,17-dione are certain aspects of the chemical physiology readily detectable in male accessory sexual of androgens and estrogens which must organs soon after administration of these ultimately be taken into account by any steroids in physiological doses (Harding

J. CELL. AND COMP. PHYSIOL., 66: 111-124. 111 112 H. G. WILLIAMS - ASHMAN and Samuels, ’62; Jensen, ’63; Pearlman, and loss of their endoplasmic reticulum ’63). Whether selective concentration of (Brandes and Groth, ’63; Deane and androgens in tissues of the male genital Porter, ’60; Harkin, ’57; Szirmai, ’62; tract (cf. Butenandt, Gunther, and Turba, Price and Williams-Ashman, ’61). There ’60) is of widespread significance remains also occurs a marked fall in the oxygen controversial . consumption and certain respiration- (5) By and large, the sex genotype has coupled synthetic activities by slices of relatively little influence on the reactivity these tissues (Huggins, ’47; Nyden and of many mammalian cells to androgens Williams-Ashman, ’52; Wicks and Villee, and estrogens. ’64). Within a few days after excision (6) Some of the physiological actions of the testes, a decline occurs in the level of estrogens on the female genital tract of certain respiratory enzymes and in the (e.g., uterine hyperemia and water imbibi- mitochondria1 population density (Edel- tion) are manifest more quickly than most man, Brendler, Zorgniotti, and Edelman, of the known actions of androgens on ’63; Price and Williams-Ashman, ’61; Wil- male accessory sexual tissues. liams-Ashman, ’62). Castration also leads (7) It is extremely difficult to distin- to a loss of cytoplasmic basophilia (largely guish in any rigorous fashion between a due to RNA). In some male accessory “target” and “non-target” tissue for andro- genital glands, such as the mouse seminal gens and estrogens. Szirmai (’62) has pro- vesicle (Deane and Porter, ’60), androgen vided an excellent discussion of variations withdrawal does not result in a very from one species to another in the sensi- marked decline in the number of RNA- tivity of different connective tissue, epithe- rich granules per unit volume of inter- lial and muscle cells to sex hormones, with cisternal cytoplasm, although with the special reference to the embryology and cytoplasmic shrinkage and loss of endo- location of these tissues. He points out plasmic reticulum, the total number of that only in relatively few instances can ribosomes per epithelial cell is profoundly androgens and estrogens act in the same reduced. In other glands, such as the rat way on homologous or identical structures; ventral prostate, there occurs a more usually their effects are mutually antagon- marked post-castrate decrease in the epi- istic. In this connection, there comes to thelial cell ribosomal population density mind such striking phenomena as the an- (Brandes and Groth, ’63; Price and Wil- tagonism by estrogens of androgen action liams-Ashman, ’61). All of these morpho- on prostatic epithelia, and the uterotrophic logical and biochemical changes are action of testosterone (Huggins, ’47), as well as the ability of estrogens to induce rapidly reversed by treatment of re- squamous metaplasia and fibromuscular cently castrated animals with testoster- hypertrophy in some male accessory glands one (cf. Williams-Ashman, Liao, Han- (Price and Williams-Ashman, ’61 ). cock, Jurkowitz, and Silverman, ’64). It is worth noting that “electron dense bodies” ANDROGENS AND REACTIONS INVOLVED appear in the supranuclear region of pros- IN RIBONUCLEIC ACID AND PROTEIN tatic epithelial cells as a result of orchiect- METABOLISM IN THE PROSTATE omy or estrogen treatment (Brandes and GLAND AND SEMINAL Groth, ‘63). It is conceivable that these VESICLES structures are lysosomes, and that they may perform a scanvenging function when General morphological and biochemical the prostatic cells dwindle after removal considerations of androgens from the circulation. Since The most remarkable morphological the report of Hertz and Tullner (’53), change in male accessory glands which much too little attention has been given to occurs soon after orchiectomy of postpu- the biochemistry of the post-castrate regres- beral males is the shrinkage of the cyto- sion of the prostate and seminal vesicles, plasm of their tall columnar epithelial which is obviously a complex and far from cells, accompanied by a massive collapse passive process. ANDROGENS AND NUCLEIC ACID AND PROTEIN SYNTHESIS 113

Amino acid penetration and resulted in: (1) an approximately twofold activation increase in the quantity of ribosomal mate- As yet there are no compelling reasons rial extractable by deoxycholate treatment of crude cell particulate preparations (the to believe that the effects of androgens on RNA:protein ratio of the ribosomes and protein synthesis in male accessory glands the gross base composition of the riboso- are in any large measure dictated by ac- mal RNA were unaffected by testosterone tions of these hormones on either the ac- administration), and (2) an increase in tive transport of amino acids, or on the the capacity of the isolated ribosomes to enzymatic synthesis of transfer RNA- support the transfer to protein-like mate- amino acids. Recent reviews by Riggs rial of valine-14C or phenylalanine-14C de- ('64) and Tomkins and Maxwell ('63) rived from the corresponding preformed emphasize, however, the paucity of experi- transfer RNA-amino acids. mental work on the influence of andro- Evidence in support of the latter was gens on these processes in mammalian tis- obtained under experimental conditions in sues. Kochakian, Tanaka, and Hill ('61) which both the rate and extent of the reported that the activity of amino acid- aminoacyl transfers were proportional to activating enzymes (measured by 32PP-ATP the quantity of ribosomal material added exchange in the absence and presence of to the reaction mixtures, and where the amino acid mixtures) in guinea pig pro- levels of GTP, ATP, Mg++ ions, soluble state and temporal muscle changed in di- transfer enzymes and sulfhydryl com- rect proportion to the alterations in mass pounds were not rate-limiting. The marked of the tissues induced by castration or diminution in aminoacyl transfers by ribo- androgen treatment. On the contrary, the somes from control as compared with tes- specific activity of these enzymes in guinea tosterone-treated castrates was nullified pig seminal vesicle was markedly dimin- either by addition to the isolated ribosomal ished by orchiectomy and restored by in- systems of appropriate synthetic polyribo- jection of testosterone. In experiments nucleotides rich in codons for the amino with isolated rat seminal vesicle slices, Wil- acid tested (poly UG for valine and poly U son ('62) observed little influence of tes- for phenylalanine) or by addition of pros- tosterone in vivo on either the penetration tatic nuclear RNA isolated from testoster- of labeled amino acids into the slices or on one-treated animals ( Williams-Ashman et the formation of transfer RNA-amino al., '64; Liao, '65a). These findings, to- acids. gether with failure to observe any marked difference in the ability of either ribosomes Aminoacyl transfers by ribosomes or soluble protein fractions from the pros- From studies on isolated seminal vesicle tates of the untreated or androgen-treated slices, Wilson ('62) concluded that testos- castrates to degrade synthetic template terone affects protein biosynthesis in this RNA's, suggested that the increased amino- organ primarily by enhancing the incor- acyl transfer capacity of prostatic ribo- poration into microsomal ribonucleoprotein somes from the androgen-treated animals of aminoacyl residues derived from trans- was due to increased levels of template fer RNA-amino acids. Recently, a series of RNA associated with these ribonucleopro- investigations have been carried out on the tein particles. Two aspects of these investi- amino acid-incorporating capacities of iso- gations are worthy of note. First, the ribo- lated prostatic ribonucleoprotein particles somes studied in vitro had been detached (Liao and Williams-Ashman, '62; Silver- from the endoplasmic reticulum by treat- man, Liao, and Williams-Ashman, '63; ment with detergents. Further studies on Williams-Ashman, and Liao, '63; Wil- the amino acid-incorporating activity of liams-Ashman et al., '64; Liao, '65a). the ribosomes in association with the lipo- These experiments disclosed that injection protein membranes of the endoplasmic of testosterone into recently castrated rats reticulum may be enlightening. Second, it over periods of 48-72 hours (during which has not been possible under a variety of interval the androgen caused the fresh circumstances to demonstrate any mean- weight of the rat ventral prostate to double) ingful in uitro effects of testosterone or 4- 114 H. G. WILLIAMS - ASHMAN androstene-3-17-dione on aminoacyl trans- in the kidney of sexually mature males; fers by the isolated prostatic ribosomal the enzyme could not be detected in renal systems. extracts of females or prepuberal males. Experiments of rather similar design But if testosterone was injected for 7 days, (Breuer and Florini, '65; Florini and then the esterase isozyme was present in Breuer, '65) hint that testosterone can in- kidney extracts from both female and ju- crease the number as well as the amino venile male mice. Other kidney esterases acid-incorporating capacity of ribosomes separable by starch electrophoresis were in the skeletal muscles of orchiectom- not affected by sex hormones. Shaw and ized rats. The increased protein-synthesiz- Koen ('63) concluded that testosterone in- ing capacity of the ribosomes was related duced the synthesis of the mouse kidney to the levels of template RNA associated esterase isozyme, which was not found in with the particles. Sucrose density gradi- other organs. ent analyses revealed that the testosterone- Further investigations on the control of induced elevation in the aminoacyl-trans- protein biosynthesis by testosterone would fer activity of the muscle ribosomes was be greatly facilitated if methods could be paralleled by an increase in the popula- developed for measurement of incorpora- tion of polyribosomes. tion of amino acids into specific proteins Extensive investigations by Kochakian by cell-free enzyme preparations from male on mouse kidney (Kochakian, '62; Kocha- accessory genital organs. kian, Hill, and Aonuma, '63) and guinea pig male accessory glands (Kochakian, '64) Synthesis and template activity of are also consistent with the view that the ribonucleic acids effects of testosterone on protein biosyn- The ability of androgens to increase the thesis in these organs are contingent upon total RNA levels and RNA:DNA ratio in more primary changes in the synthesis or male accessory glands and some other intracellular translocation of ribosomal and susceptible organs has been widely docu- messenger RNA's. The possibility that male mented (Frieden, '64; Kochakian, and Har- sex hormones also influence protein synthe- rison, '62; Williams-Ashman et al., '64). sis at the level of translation of messenger Excessive doses of testosterone can in- RNA's cannot, however, be overruled by crease the RNA: DNA ratio in these tissues the limited experimental evidence avail- to well above the normal values. A care- able at present. ful study by Liao ('65a) demonstrated that Most studies on the effects of androgens in adult rats sacrificed 70 hours after on protein biosynthesis have dealt with the orchiectomy, the injection of testosterone incorporation of radjoactive amino acids over this interval increased the levels of to- into ill-defined mixtures of polypeptide tal and ribosomal RNA, but did not alter the material. It is well known that testosterone content of nuclear RNA per unit amount treatment in vivo can differentially influ- of DNA, in the rat ventral prostate. The ence the activity of many enzymes in male base composition of the nuclear RNA accessory sexual tissues (Mann, '64; Price (A:U:C:G = 16:22:26:36) and of the ri- and Williams-Ashman, '61 ; Williams- bosomal RNA (A:U:C:G = 18:20:24:38) Ashman, '62, '64, '65a, b). Particularly was not affected by the androgen treat- striking increases in the levels of D-amino ment, nor was the sedimentation profile of acid oxidase and B-glucuronidase in mouse the prostatic nuclear RNA in a sucrose kidney are known to follow injection of an- density gradient (the latter being very simi- drogens (Frieden, Harper, Chin, and Fish- lar to the sedimentation profile of isolated man, '64). But most of these reported andro- prostatic ribosomal RNA). It was shown, gen-induced changes in tissue protein levels however, that the template activity of the are quantitative rather than qualitative in refined prostatic nuclear RNA, measured nature. However, an interesting qualita- with both bacterial and prostatic riboso- tive action of testosterone on an isozyme mal amino acid-incorporating systems, was of esterase in mouse kidney was recently markedly increased within as little as 24 described by Shaw and Koen ('63). They hours after treatment of recently castrated showed that this isozyme was present only rats with testosterone. Similar increases ANDROGENS AND NUCLEIC ACID AND PROTEIN SYNTHESIS 115 in the relative template activity of RNA phosphate as precursors, but only the nu- isolated from prostatic ribosomes were also clear RNA polymerase system was inhibited demonstrated in the E. coli system, al- by low levels of actinomycin D in vitro. though the inherent template activity of Initial experiments (Hancock, Zelis, Shaw. the ribosomal RNA from both groups of and Williams-Ashman, ’62) showed that animals was much less than that of the the RNA polymerase activity of crude pros- corresponding preparations of nuclear tatic nuclear “aggregate” enzyme prepara- RNA. tions of recently orchiectomized rats was Much more rapid effects of androgens increased after administration of testos- on RNA synthesis in rat seminal vesicles terone over periods of 4-5 days. The ef- were recently discovered by Wicks and fects of the androgen were, however, much Kenney (’64). They show that within an more pronounced when the polymerase re- hour or so after injection of testosterone actions were studied in media of low ionic into rats castrated 12-15 hours previously, strengths, the control activities being the rate of incorporation of 32Pinto vesicu- markedly enhanced by raising the salt con- lar RNA was increased by 50%,and con- centration. More recently, Liao (’65b) tinued to rise until an approximately 2- to found that the RNA polymerase activity 3-fold increase was attained. The base of prostatic nuclear extracts was signifi- composition of the pulse-labeled RNA was cantly increased within 1 hour after in- intermediate between that of total seminal jection of testosterone into castrates; no vesicle RNA and “DNA-like RNA.” Wicks effect of testosterone on the RNA polymer- and Kenney (’65) have also reported that ase of crude nuclear extracts of liver or in rat seminal vesicle, little turnover of thymus was demonstrable. Again, the ef- the phosphorus of the -CCA termini of fects of testosterone on prostatic RNA transfer RNA’s occurred after pulse polymerase activity were much more pro- labeling with 32P. But 90 minutes after nounced in media of low ionic strengths. injection of testosterone, there occurred a Although the RNA polymerase activity of 2- to 3-fold increase in the synthesis of such crude prostatic nuclear preparations transfer RNA in this organ. is inhibited by actinomycin D and by ex- The latter experiments and the afore- posure to DNase( and requires the simul- mentioned studies on prostatic ribosomes taneous presence of all four major ribonu- suggest, then, that one of the earliest cleoside triphosphates), it has not been known effects of androgens on male ac- possible, despite repeated attempts (Han- cessory reproductive organs is to increase cock, Jurkowitz, and Jurkowitz, ’65), to the synthesis of messenger, ribosomal, and rid the enzyme preparations of DNA in transfer RNA’s. A detailed examination of such a fashion that their ability to catalyze enzymatic pathways for the incorporation RNA synthesis can be clearly stimulated of nucleotides into RNA in the rat ventral and directed by exogenous DNA. There is prostate (Hancock, Zelis, Shaw, and Wil- considerable evidence that the RNA poly- liams-Ashman, ’62; Hancock, Jurkowitz, merase activity of the prostatic nuclear and Jurkowitz, ’65; Williams-Ashman and preparations is limited not by the catalytic Liao, ’63; Williams-Ashman et nl., ’64) re- capabilities of the activating protein, but vealed the presence of only two distinct rather by the priming ability of the DNA enzyme systems : ( 1 ) a DNA-dependent that is firmly bound to the extract. The RNA polymerase associated solely with the fact that increasing the ionic strength of cell nuclei, and (2) enzyme( s) which were the reaction mixtures increases the base- associated with both nuclear and cyto- line activity of such preparations and di- plasmic fractions that catalyzed the addi- minishes the effects of testosterone treat- tion of cytidylate and adenylate residues ment hints that androgen in vivo affects to the termini of preexisting transfer RNA the levels or activity of substances associ- chains (the activity of the latter enzyme( s) ated with prostatic DNA (conceivably his- was orders of magnitude greater than that tones) that limit its priming activity in the of the nuclear RNA polymerase in the pros- RNA polymerase reaction. In marked con- tates of normal animals). Both of these en- trast to the prostatic RNA polymerase sys- zyme systems utilized ribonucleoside tri- tem, it has proved relatively easy to isolate 116 H. G. WILLIAMS - ASHMAN the RNA polymerase of rat testis free from rats largely prevents the striking increases DNA, and to purify the enzyme as a poly- in prostatic and seminal vesicle weight nucleotide-free protein (Ballard and Wil- which occur 72 hours after daily subcuta- liams-Ashman, '64). As yet, it has not neous injections of testosterone propionate been possible to influence the catalytic ac- (1 mg per 100 g of body weight). Ange- tivity of this resolved testicular RNA poly- letti, Salvi, and Tacchini ('64) reported merase by addition of testosterone or other that over a period of 5 days, concurrent sex hormones in vitro. administration of actinomycin D almost Research on the influence of androgens completely blocked the testosterone-in- and other hormones on the synthesis and duced increase in the weight, soluble pro- utilization of ribonucleic acids has been tein level, and activities of protease, severely impeded by lack of precise meth- a-amylase, and nerve growth factor of ods for estimation of discrete RNA's, par- mouse submaxillary glands. ticularly with respect to the template func- Frieden et al. ('64) have studied some tion of various messenger RNA's (or effects of actinomycin D administration polycistronic messengers) in protein bio- on certain biochemical concomitants of the synthesis. Kidson and Kirby ('64) have renotropic action of testosterone in mice. recently developed countercurrent distribu- They reported that in experiments of 2-3 tion methods for separation of rapidly days' duration, the injection of actinomy- labeled RNA's in rat liver. Complex and cin D (total of four doses of 200 ug per kg reproducible patterns were observed under body weight) injected 1 day prior to and well-controlled conditions. Testosterone concurrently with testosterone completely was among the hormones that induced inhibited the increase in the kidney B-glu- rapid, selective, and reversible changes in curonidase activity resulting from the an- the rapidly labeled RNA profiles. Kidson drogen treatment. The testosterone-in- and Kirby ('64) are of the opinion that duced incorporation of leucine-14C into the various polyribonucleotide fractions isolated kidney slices was, however, not af- separable by their techniques are largely messenger RNA's, and application of such fected by actinomycin D over this time methods to the problem of the nature of period, although the antibiotic depressed the androgenic control of RNA synthesis the base-line leucine incorporation values in male accessory organs might be of great in the controls that did not receive testos- value. terone. Similar findings were obtained when gl~cine-*~Cand a~ginine-'~Cwere INHIBITORS OF RNA AND PROTEIN used as the labeled amino acids. These SYNTHESIS AS ANTAGONISTS experiments on the influence of actinomy- OF TESTOSTERONE ACTION cin D on the effect of testosterone on In comparison with studies on the ac- amino acid-incorporation by kidney slices tions of estrogens, comparatively little are somewhat difficult to interpret, be- work has been published on the influence cause the rate-limiting steps for entry of of inhibitors of RNA and protein synthesis the amino acids into protein-like material on the action of testosterone on the acces- have not been precisely clarified in such sory glands. The ability of puromycin to isolated kidney preparations. Frieden et depress aminoacyl transfers by isolated ri- al. ('64) also found that actinomycin D bosomes, and of actinomycin D to depress had little effect upon the increase in renal the DNA-dependent synthesis of RNA by @-glucuronidaseactivity due to testosterone nuclear RNA polymerase preparations in administration if actinomycin treatment male accessory glands, is well documented was begun after the first injection of the (Williams-Ashman, Liao, Hancock, Jurko- androgen. witz, and Silverman, '64). As might be ex- Breuer and Florini ('65) state in a pre- pected, the author has observed in unpub- liminary communication that actinomycin lished experiments that intraperitoneal in- D blocks the testosterone-induced increase jection of actinomycin D (25 pg per 100 g in the ribosomal and messenger RNA con- of body weight) into recently castrated tent of skeletal muscle of castrated rats. ANDROGENS AND NUCLEIC ACID AND PROTEIN SYNTHESIS 117

ANDROGENS AND DNA SYNTHESIS that testosterone may increase the entry of Relatively little attention has been paid fluorouracil into RNA at the expense of to the biochemistry of androgen-induced its conversion into fluorodeoxyuridylic changes in DNA synthesis in accessory re- acid. Dorfman ('63) has described the productive organs. The classical experi- anti-androgenic action of 5-fluorouracil in ments of Burkhart ('42) using the colchi- the cock's comb. cine technique showed that following Much interest attaches to further inves- injection of a single dose into castrate rats, tigations on the interrelationships between cell hypertrophy and nuclear enlargement RNA and DNA synthesis in male accessory were observable at 23 hours, whereas mito- genital organs. In this regard, it would be tic activity began at 35 hours and reached particularly interesting to examine various a maximum at 43 hours. Wicks and Villee reactants in the DNA polymerase system, ('64) observed that 14C derived from gly- cine-2-14C was hardly incorporated into and also enzyme systems responsible for DNA by seminal vesicle slices prepared the production of deoxyribonucleoside tri- from castrated rats, but that detectable phosphates. Very recently, Weiss, Zager- incorporation of the radioisotope into DNA man, and Kokolis ('65) reported that tes- was manifest with slices obtained from tosterone induces large increases in the animals 36 and 48 hours after injection of activity of thymidine kinase and thymidy- testosterone propionate, but not before that Iate synthetase in mouse seminal vesicle. time (marked changes in oxygen consump- Any comprehensive biochemical explana- tion, uptake of gly~ine-'~Cinto acid-soluble tion for the stimulation by androgens of material, and incorporation of glycine-14C cell division in accessory reproductive or- into RNA were observed in seminal vesicle gans must account for the ultimate cur- slices within 18-24 hours after administra- tailment of the steroid-induced growth of tion of the androgen). Very recently, Shep- these organs. pard, Tsien, Mayer, and Howie ('65) re- ported that treatment of orchiectomized CONCLUSIONS rats with methandrostenolone caused an The experimental evidence available to- increased uptake of thymidine-'H into the day is by and large consistent with the view DNA of the levator ani, ventral prostate, that androgenic hormones initiate and and seminal vesicles; no effect was ob- maintain the functional differentiation of served in adrenal, thymus, or leg muscle, the prostate gland and seminal vesicles - and a decreased uptake of thy~nidine-~Hand their elaboration of all sorts of secre- in the kidney was noted. The effects of tory products (Mann, '64; Price and Wil- this anabolic agent were observable only liams-Ashman, '61) - by regulating the after 2 days of treatment. biosynthesis of rate-limiting enzymes and Of related interest is the finding of structural proteins. Furthermore, certain of Cantarow and Zagerman ('64) that the the effects of testosterone on the incor- testosterone-induced growth of seminal poration of amino acid into proteins by vesicles in castrated rats is inhibited by cell-free extracts of these tissues seem to simultaneous treatment of the animals be contingent upon more primary changes with 5-fluorouracil. However, the extent in the ribosomal population density and in of growth inhibition by the pyrimidine the levels of template RNA's. It is fast be- analogue could be overcome by high doses coming apparent that an increased synthe- of the androgen at all tolerated levels of sis of transfer, ribosomal, and messenger- fluorouracil administration. Cantarow and like RNA's is one of the earliest detectable Zagerman ('64) point out that the princi- metabolic events in intact male accessory pal effect of fluorouracil in animal tissues glands of castrates following the injection is to inhibit DNA synthesis, via the for- of testosterone. In accord with this are the mation of fluorodeoxyuridylic acid, a po- early increases in RNA polymerase activity tent inhibitor of thymidylate synthetase. of crude nuclear extracts of the prostate However, it is also known that fluorouracil after stimulation by testosterone in vivo. can be incorporated into RNA, and it is Such changes in nuclear RNA polymerase possible, as these investigators conjecture, activity, however, are in all probability a 118 H. G. WILLIAMS - ASHMAN reflection of ill-defined perturbations in the polymerase actiuities of both types of prep- priming ability of the DNA bound to the aration were, however, measured at both preparations, and it simply is not known low and high ionic strengths. how far these effects are biochemically re- BONNER:Pardon me. When you test moved from the primary receptors for an- at high ionic strength, then at all times drogenic hormones. Nor is it possible, after testosterone administration the activ- from the very limited data at hand, to come ity of the system is higher than that of the to any sure conclusions as to the degree system tested at low ionic strength, and is of selectivity of activation by testosterone constant throughout the experiment. It of the synthesis of complementary RNA would seem to me again that this implies copies of specific genes. Certainly there that we are here testing something that is are many reasons to suppose (Williams- a combination of RNA polymerase plus Ashman et al., '64; Williams-Ashman, '62, template activity. The fact that KNA- '64, '65a, b) that the receptors for andro- making activity is constant in the high genic and other steroid hormones are prob- salt concentration medium would seem to ably proteinaceous in nature. As a guide to me to imply that it is differences in tem- further experimentation, it still seems rea- plate activity of the DNA that are being sonable to surmise that such proteins may manifested. be in some way connected with regulation WILLIAMS-ASHMAN: That's precisely of synthesis of specific polyribonucleotides, what I tried to imply in my talk. or conceivably with the translocation of BONNER: I haven't come to my question! various RNAs from the cell nucleus to the O.K. So, fine; we are at one! Now we go cytoplasmic sites of protein biosynthesis. to this purification of the soluble RNA Progress along these lines will undoubtedly polymerase. You can get a nice RNA poly- be contingent, inter alia, upon develop- ment of much better understanding of the merase. Others have also shown that the biochemical intimacies of RNA synthesis capability of the organ to synthesize RNA's and intracellular transport in mammalian rises %fold over a period of 60 minutes cells, and of the interrelationships between after hormone treatment. RNA- and DNA-synthesizing systems. WILLIAMS-ASHMAN: The increased RNA polymerase activity following testosterone ACKNOWLEDGMENT treatment in vivo was measured in the Work in the authors' laboratories as re- prostate, not in the testis. ported in this paper was supported by BONNER: What I wanted to find out was Grants CA-06545 from the National Cancer whether in the case of the organ from Institute and HD-01453 from the National which you purified the RNA polymerase, it Institute of Child Health and Human De- is now possible to determine if the RNA velopment. polymerase itself varies, or if it is template activity. OPEN DISCUSSION WILLIAhlS-ASHMAN : We have not ex- amined the effects of testosterone in uiuo BONNER: Well, first of all, Professor on the RNA polymerase activity of testis. Williams-Ashman, I would like to discuss Testosterone does not have any effect on this elegant experiment concerning what the purified testicular enzyme. you refer to as RNA polymerase from pro- TOMKINS%: Have you looked into any state. You show that at different times, other circumstances except the one that after the administration of testosterone, the you showed? Does the total nuclear RNA ability of this crude system to make RNA stay the same? increases, provided only that the prepara- WILLIAMS- ASHMAN: The RNA : DNA tion was isolated at low ionic strength. ratio of isolated nuclei remains at the value WILLIAMS-ASHMAN:No, it was not of 0.25. tested in this way. The nuclear prepara- tions were isolated from both castrated ani- 1 James Banner, Division of Biology, California In- stitute of Technology, Pasadena California. mals and castrated animals treated with 2 Gordon M. Tomkins, Nationil Institute of Arthritis and Metabolic Diseases. National Institutes of Health, testosterone in the same fashion. The RNA Bethesda, Maryland. ANDROGENS AND NUCLEIC ACID AND PROTEIN SYNTHESIS 119

WOOL3: There is one puzzling aspect to ous expressions of the enzyme, you did not the results. Incorporation of radioactivity find any effect. I would like to know from sRNA charged with 14C-phenylalanine whether you have done this experiment into protein is decreased when ribosomes in conditions where the system was strictly from the prostate of castrate animals are DNA-dependent, and concentration of en- used to catalyze protein synthesis. If that zyme was not limiting. This is a very defect is to be accounted for by a de- difficult experiment to do. I think that you ficiency of template RNA, as you suggest, have demonstrated that steroids have no then I should have expected that addition effect on the enzymes. I just wonder of artificial template RNA (that is to say, whether you have examined the effect of polyuridylic acid) would not merely re- the steroid on the template activity of the store synthesis to normal, as you found, but DNA. This can be done only when DNA would actually result in an amount of syn- is limiting in the system. thesis that was apparently greater than WILLIAMS-ASHMAN:We have not ex- that observed with ribosomes from the amined the effect of testosterone under the prostate of hormone-treated animals to conditions you mention. But we have per- which polyuridylic acid was also added. formed a number of different types of ex- The basis of my prediction is that ribo- periment, including examination of the somes from the prostate of castrate ani- effect of preincubating the enzyme with tes- mals have less natural template RNA; tosterone under conditions where consider- therefore, they should bind greater able polymerase activity is lost, the assays amounts of artificial template RNA (poly- being initiated by later addition of DNA. uridylic acid). The artificial template RNA I should emphasize, however, that our in- contains only codons for phenylalanine; in ability as yet to demonstrate any in vitro natural template RNA the codon for phe- action of testosterone on the soluble testicu- nylalanine appears, on the average, only lar enzyme preparations does not neces- once in each 20 triplets. Since the only sarily mean that the enzyme cannot be in- radioactive amino acid is phenylalanine, fluenced by testosterone under other ex- there should be far greater incorporation perimental conditions. of radioactivity into protein when ribo- NOVELLI5: I would like to make a com- somes from the prostate of castrate ani- ment on this poly U experiment. If you do mals are used. Can you account for the a very brief incubation at reduced temper- discrepancy between observation and pre- ature, and then take your reaction mix- diction? ture and put it on a sucrose gradient, so WILLIAMS-ASHMAN: I think that one you can see where the incorporation is can still conclude from these experiments taking place, you will see that with rat liver that the capacity of unprogrammed pro- all of the endogenous incorporation is in static ribosomes to react with the syn- the polysomal region. The monomers do thetic template RNA is not diminished by not incorporate the amino acid. When the androgen withdrawal under these experi- same experiment is repeated with poly U, mental conditions. It seems to me that a the incorporation in the polysomal region possible explanation for the increased remains unchanged, but the bulk of the amino acid incorporation in the absence poly U-stimulated incorporation is associ- of poly U by ribosomes from testosterone- ated with the monomers. treated castrates is that relatively higher So I would interpret your experiment to levels of template RNA’s are associated indicate that both preparations have a com- with the total ribosomal population. parable amount of competent monomers. Ts’o 4: I would like to make a com- ment on the very elegant research you have done on this soluble, purified mammalian 3 Ira G. Wool, Department of Physiology, University of Chicago, Chicago, Illinois. polymerase. You mentioned the fact that 4 Paul 0. P. Ts’o, Department of Radiological Sci- when you added testosterone and tried to ence, Johns Hopkins University, Baltimore, Maryland. 5G. David Novelli, Biology Division, Oak Ridge examine the effect of the hormone on vari- National Laboratory, Oak Rldge, Tennessee. 120 H. G. WILLIAMS - ASHMAN

HANSON’:In line with Dr. Bonner’s tones, or to inactivation of inhibiting comments, there was one point which I factors associated with the aggregates, such did not understand. You were using the as RNase. purified polymerase, with DNA of an un- BOND’: I have been looking at a liver specified kind. Down in the table you protein component, and its response to showed an item which said “minus am- both testosterone and estradiol which pro- monium sulphate,” and the activity fell by vides an interesting experimental model. half. If this were “clean” DNA, one would The protein which I have designated as a think there was an ionic strength effect sex-associated protein is present in large in addition to the removal of the histones. amount (about 3% of the soluble or super- WILLIAMS-ASHMAN:Phillip Ballard has natant protein) in the liver of the male demonstrated a small (about 50% ) ac- rat and appears by similar evaluating cri- tivating effect of relatively low ionic teria to be absent from the liver of the strengths (about 0.2) on the RNA-poly- female. I will return to this point mo- merase activity of soluble testicular en- mentarily. zyme preparations, which exhibit an ab- Discussion Figure A shows the sex-as- solute requirement for exogenous DNA. sociated protein chromatographically. The I think that it is very likely that this is an chromatogram was developed on DEAE- effect of ionic strength on interaction of cellulose by equilibrium elution with a the RNA polymerase protein with the DNA, single buffer. Consequently, the vast ma- since “clean” DNA preparations were used jority of the cellular proteins remained on in these experiments. The effects of high the column at the time that the chromato- ionic strengths (approximately 1.O) on the gram was completed. The first large peak RNA polymerase activity of prostatic nu- is a complex consisting of basic and rela- clear extracts are, I believe, of quite a dif- tively uncharged proteins, the so-called ferent nature. In the latter situation, the fall-through components. The second RNA polymerase activity is dependent on smaller peak represents the sex-associated DNA firmly bound to the prostatic nuclear protein. This component cannot be de- extracts. Here the effects of high salt con- 6 John B. Hanson, Department of Agronomy, College centrations could well be due to displace- of A riculture, Universlty of Illinoig. Urbana, Illmols. 7 lfooward E. Bond,. Carcinogenesls Studies Branch, ment of inhibiting substances such as his- National Cancer Instltute, Bethesda, Maryland.

9 0.6- I a- 0a 0.4- N ANDROGENS AND NUCLEIC ACID AND PROTEIN SYNTHESIS 121 tected in the chromatogram of the female male shows a tendency toward a slow and rat liver. variable loss of the material with time fol- Discussion Figure B describes some of lowing castration. The normal male the responses of the sex-associated pro- treated with estradiol loses the protein tein to steroid treatment. The castrated rapidly; in the experiment shown here,

CHANGES IN THE SEX- ASSOCIATED PROTEIN COMPONENT WITH RESPECT TO EXPERIMENTAL TREATMENT AND AGE :ASTRATED MALES CASTRATED MALES TREATED WITH ESTRADIOL

12 Ii 14 25 3 5 7 10 12 35465 n,n TIME AFTER CASTRATION1 (days) TIME AFTER TREATMENT(days) VORMAL MALES TREATED WITHn JNTREATED Mi :S OF DIFFERENT ESTRADIOL AGES C D

35 60 9a z ll 239 TIME AFTER INJECTION (days) AGE (days) Discussion Figure B 122 H. G. WILLIAMS - ASHMAN most of the material had been lost by 9 It would be of fundamental concern to days after treatment. The last figure on know if the sex-associated protein exists as this slide shows the relationship of the a circulating entity in blood. Thus far it sex-associated protein to the age" of the ani- has not been detected there, either chro- mals. As you can see, there is only a small matographically or by the much more sen- amount of the material present at 35 days, sitive immunochemical techniques. whereas the normal complement of the There does not appear to be a necessary protein is present at 60 days of age. A adrenal involvement in the sex-associated point at 43 days of age was omitted from phenomenon. The adrenalectomized fe- this slide. It has a value about intermedi- male still responds to testosterone treat- ate between the other two. Clearly, there ment with an increase in the protein, is an increase in amount of the protein as- although the response is rather poor com- sociated with sexual maturity of the ani- pared to that in the intact animal. Adre- nalectomy appears to cause no significant mal. change in the amount of protein found in Corresponding experiments with testo- the male. The adrenalectomized male re- sterone-treated females have given opposite sponds similarly to the intact male when effects. Both the normal and castrated fe- treated with estradiol; the protein disap- male responded by producing the protein pears rapidly from the liver. in amounts approaching that seen in An interesting response occurred when males. The protein did not appear in the the hormones were administered to ani- untreated female castrate. mals simultaneously. Estradiol was given These studies recently have been con- in oil, whereas the testosterone was given firmed by [D.] Barzilai and [G.] Pincus as a combination of testosterone propionate (Proc. SOC.Exptl. Biol. Med., 118: 57-59, and testosterone phenylacetate. The hor- 1965). They also extended the investiga- mones were administered in a single treat- tions to show a response from progester- ment with the result that the testosterone one treatment which caused the appear- was a longer-lasting repository. Adrenal- ance of the protein in the female and an ectomized females showed no response un- increase in the level found in the castrated til about 20 days after hormone treatment, male. It is conjectural, though, as to when the sex-associated protein began ap- whether this is from a direct response to pearing in substantial amounts. In simi- progesterone or from a metabolite of pro- larly treated males, the protein disappeared gesterone with androgen activity. rapidly after hormone treatment, but at The sex-associated protein has been puri- about 20 days it returned in appreciable fied to the extent that a highly specific amounts. antiserum has been produced. The re- Generally, it appears that I have de- agent produces a single precipitin band scribed a protein that is sex hormone-de- when diffused in agar against the liver pendent. There appears to be a direct- soluble proteins. The sensitivity of the acting antagonism between the hormones agar diffusion techniques is such that it and a reciprocal-dose relationship. I be- has been possible to detect the protein in lieve that the phenomenon offers the pos- livers of normal female rats. The concen- sibility of studying a highly specific re- trations are about 32 times less in the fe- sponse to hormone action at the molecular male than in the male. Both single and level, in addition to studying the physio- double diffusion quantitative agar tech- logical consequences of that action. niques give good agreement on this figure WILLIAMS-ASHMAN: I appreciate an op- which, nevertheless, lacks precision be- portunity to hear about this excellent work. cause of its logarithmic nature. Stability This is another example of an almost problems associated with preparing the qualitative effect of a sex hormone on for- highly purified protein have prevented the mation of a specific protein in an extra- accumulation of sufficient material to use genital tissue. It brings to mind the work more sophisticated quantitative methods of IC. R.] Shaw and [A. L.] Koen on the and, thereby, to arrive at the absolute induction by testosterone of a specific ester- amounts present in tissue. ase isozyme in mouse kidney, and Olga ANDROGENS AND NUCLEIC ACID AND PROTEIN SYNTHESIS 123

Greengards studies on the influence of by nuclei of the prostate gland. Biochim. Bio- estrogens on plasma phosphoprotein for- phys. Acta, 55: 257-260. Harding, R. W., and L. T. Samuels 1962 The mation in male chickens. uptake and subcellular distribution of CI4-la- beled steroid in rat ventral prostate following LITERATURE CITED in vivo administration of testosterone-4-C14. Angeletti, P. U., M. L. Salvi, and G. Tacchini Endocrinology, 70: 109-113. 1964 Inhibition of the testosterone effect on Harkin, J. C. 1957 An electron microscopic the submaxillary gland by actinomycin-D. Ex- study of the castration changes in the rat pro- perientia, 20: 612-613. state. Endocrinology, 60: 185-199. Ballard, P., and H. G. Williams-Ashman 1964 Hertz, R., and W. W. Tullner 1953 Quantitative Isolation of a soluble RNA polymerase from rat aspects of regression of prostate and uterus of testis. Nature, 203: 150-151. the rat following hormonal deprivation. J. Clin. Brandes, D., and D. P. Groth 1963 Functional Endocrinol., 13: 832. ultrastructure of rat prostatic epithelium. In, Huggins, C. 1947 Androgens and anaplasia. Biology of the Prostate and Related Tissues, ed., Yale J. Biol. Med., 19: 319-330. E. P. Vollmer. Natl. Cancer Inst. Monograph Jensen, E. F. 1963 Comments on Dr. Pearl- 12, pp. 47-52. man’s paper: Comparison of androgens and Breuer, C. B., and J. R. Florini 1965 Effects estrogens as to their fate in target tissues. In, of hormones on rat muscle protein synthesis Biology of the Prostate and Related Tissues, in vitro. Federation Proc., 24: 600. ed., E. P. Vollmer. Natl. Cancer Inst. Mono- Burkhart, E. Z. 1942 A study of the early ef- graph 12, pp. 317-321. fects of androgenic substances in the rat by aid Kidson, C., and K. S. Kirby 1964 Selective al- of colchicine. J. Exptl. Zool., 89: 135-165. teration of mammalian messenger-RNA synthe- Butenandt, A., H. Gunther, and F. Turba 1960 sis, Evidence for differential action of hor- Zur primaren Stoffwechselwirkung des Testos- mones on gene transcription. Nature, 203: terons. Hoppe-Seylers. Z. Physiol. Chem., 322: 599-603. 28-37. Kochakian, C. D. 1962 Intracellular regula- Cantarow, A., and A. J. Zagerman 1964 Flu- tions in the kidney by androgens. Am. Zool., orouracil inhibition of testosterone-stimulated 2: 361-366. growth of seminal vesicles. Proc. SOC.Exptl. 1964 Effect of castration and testoster- Biol. Med., 115: 1052-1054. one on protein biosynthesis in guinea pig tis- Deane, €I. W., and K. R. Porter 1960 A com- sue preparation. Acta Endocrinol., (Suppl.), parative study of cytoplasmic basophilia and 92: 1-16. the population density of ribosomes in the Kochakian, C. D., and D. G. Harrison 1962 secretory cells of mouse seminal vesicle. Z. Regulation of nucleic acid synthesis by andro- Zellforsch, 52: 697-711. gens. Endocrinology, 70: 99-108. Dorfman, R. I. 1963 The anti-androgenic ac- Kochakian, C. D., J. Hill, and S. Aonuma 1963 tivity of 5-fluorouracil. Steroids, 2: 555-559. Regulation of protein biosynthesis in mouse Edelman, J. C., H. Brendler, A. W. Zorgniotti, and kidney by androgens. Endocrinology, 72: 354- P. M. Edelman 1963 Effects of castration on 363. mitochondria of rat ventral prostate. In, Biol- Kochakian, C. D., R. Tanaka, and J. Hill 1961 ogy of the Prostate and Related Tissues, ed., E. Regulation of amino acid activating-. enzvmes of P. Vollmer. Natl. Cancer Inst. Monograph 12, guinea pig tissues by androgens. Am. J. Phys- pp. 275-280. iol., 201: 1068-1072. Florini, 5. R., and C. B. Breuer 1965 Amino Liao, S. 1965a Influence of testosterone on acid incorporation into protein by cell-free template activity of prostatic ribonucleic acids. preparation from rat skeletal muscle. 111. Com- J. Biol. Chem., 240: 1236-1243. parisons of activity of muscle and liver ribo- 1965b Rapid effect of testosterone on somes. Biochemistry, 4: 253-257. ribonucleic acid polymerase activity of rat ven- Frieden, E. H. 1964 Sex hormones and the tral prostate. Endocrinology. In press. metabolism of amino acids and proteins. In, Liao, S., and H. G. Williams-Ashman 1962 An Actions of Hormones on Molecular Processes, effect of testosterone on ainino acid incorpora- ed., G. Litwack and D. Kritchevsky. John Wiley tion by prostatic ribonucleoprotein particles. and Sons, New York, pp. 509-559. Proc. Natl. Acad. Sci. U. S., 48: 1956-1964. Frieden, E. H., A. A. Harper, F. Chin, and W. H. Mann, T. 1964 The biochemistry of semen and Fishman 1964 Dissociation of androgen-in- of the male reproductive tract. John Wiley and duced enzyme synthesis and amino acid in- Sons, New York. corporation in mouse kidney by actinomycin-D. Nyden, S. J., and H. G. Williams-Ashman 1953 Steroids, 4: 777-786. Influence of androgens on synthetic reactions in Yoncock, R. L., M. Shaw-Jurkowitz, and L. Jurko- ventral prostate tissue. Am. J. Physiol., 172: witz 1965 Studies on the RNA polymerase 588-600. in rat ventral prostate. Arch. Biochem. Bio- Pearlman, W. H. 1963 Metabolism in vivo of phys., 110: 124-132. A4-androstene-3,17-dione-H3 and its localiza- Hancock, R. L., R. F. Zelis, M. Shaw, and H. G. tion in the rat. In, Biology of the Prostate and Williams-Ashman 1962 Incorporation of ribo- Related Tissues, ed., E. P. Vollmer. Natl. Can- nucleoside triphosphates into ribonucleic acid cer Inst. Monograph 12, pp. 309-315. 124 H. G. WILLIAMS - ASHMAN

Price, D., and H. G. Williams-Ashman 1961 Wicks, W. D., and C. A. Villee 1964 Studies The accessory reproductive glands of mammals. on the course of action of testosterone propion- In, Sex and Internal Secretions, ed., W. C. ate on the rat seminal vesicle. Arch. Biochem. Young, 3rd edition. Williams and Wilkins, Biophys., 106: 353-359. pp. 366-488. Williams-Ashman, H. G. 1962 Chemical ap- Riggs, T. R. 1964 Hormones and the transport proaches to the function of the prostate gland of nutrients across cell membranes. In, Ac- and seminal vesicles. In, On Cancer and Hor- tions of Hormones on Molecular Processes, ed., mones: Essays in Experimental Biology. Uni- G. Litwack and D. Kritchevsky. John Wiley versity of Chicago Press, Chicago, pp. 325- and Sons, New York, pp. 1-57. 346. Shaw, C. and A. L. Koen Hormone- R., 1963 1964 Some experimental approaches to induced esterase in mouse kidney. Science, the molecular basis of sex hormone action. In, 140: 70-71. Cellular Control Mechanisms and Cancer, ed., Sheppard, H., W. H. Tsien, P. Mayer, and N. P. Emmelot and 0. Muhlbock. Elsevier Pub- Howie 1965 Metabolism of the accessory lishing Company, Amsterdam, pp. sex organs of the immature male rat: Changes 103-123. in nucleic acid composition and uptake of 1965a Ribonucleic acid and protein th~midine-~Hinduced by castration and meth- synthesis in male accessory reproductive glands androstenolone. Biochem. Pharmacol., and its control by testosterone. In, Mecha- 14: 41- nisms of Hormone Action, ed., P. Karlson. 51. Silverman, D. A., S. Liao, and H. G. Williams- Georg Thieme Verlag, pp. 221-227. Ashman 1963 Influence of testosterone and 1965b New facets of the biochemistry polyuridylic acid on the incorporation of phe- of steroid hormone action. Cancer Res., 25: nylalanine into peptide linkage by prostatic 1096-1120. ribosomes. Nature, 199: 808-809. Williams-Ashman, H. G., and S. Liao 1963 In- Szirmai, J. A. 1962 Histological aspects of the corporation of amino acids into protein by cell- action of androgens and estrogens. In, Protein free extracts of the prostate gland: Effect of Metabolism, ed., F. Gross. Heidelberg, Springer testicular hormones and polyribonucleotides. Verlag, pp. 45-74. In, Biology of the Prostate and Related Tissues, Tomkins, G. M., and E. S. Maxwell 1963 Some ed., E. P. Vollmer. Natl. Cancer Inst. Mono- aspects of steroid hormone action. Ann. Rev. graph 12, pp. 281-296. Biochem., 32: 677-708. Williams-Ashman, H. G., S. Liao, R. L. Hancock, Weiss, A., A. Zagerman, and P. Kokolis 1965 L. Jurkowitz, and D. A. Silverman 1964 Tes- Relationship of testosterone to thymidine ana- ticular hormones and the synthesis of ribonu- bolic enzymes. Federation Proc., 24: 330. cleic acids and proteins in the prostate gland. Wicks, W. D., and F. T. Kenney 1964 RNA In, Recent Progress in Hormone Research, ed., synthesis in rat seminal vesicles: Stimulation G. P~CUS.XX, pp. 247-292. by testosterone. Science, 144: 1346-1347. Wilson, J. D. 1962 Localization of the bio- 1965 Stimulation of transfer RNA svn- chemical site of action of testosterone on pro- thesis by steroid hormones. Federation Pr&., tein synthesis in the seminal vesicle of the rat. 20: 600. J. Clin. Invest., 41: 153-161.