seminars in CELL&DEVELOPMENTAL BIOLOGY,Vol9,1998:pp417᎐421Article No.sr980253 Spermatogenesis:its regulation by testosterone andFSHBarry R.ZirkinIt has been clear for decades that testosterone is a necessary prerequisite for the maintenance of established spermatogenesis in the adult testes,and for the restoration of spermatogenesis in testes induced experimentally to become azoospermic.In contrast,the possible involvement of FSH in regulating spermatogenesis in the adult mammal,though also studied extensively,continues to be debated in the literature.The mechanisms by which testosterone and FSH regulate spermatogenesis remain uncertain as well.This review focuses on our current understanding of the involvement of testosterone and of FSH in the maintenance and restoration of spermatogenesis in the adult testes,and discusses the emerging concept that testosterone and FSH serve as cell survival factors.Key words:spermatogenesis r testosterone FSHᮊ1998Academic PressS PERMATOGENESIS IS A COMPLEX,cyclic process that involves germ cells undergoing mitotic divisions, meiosis and terminal differentiation,and interacting over time and space within the seminiferous tubules. At any location within a tubule,one or two genera-tions of spermatogonia,one or two generations of spermatocytes,and one or two generations of sper-matids are present.The various generations of germ cells form the stages of the cycle of the seminiferous epithelium,defined as specific sets of cellular associa-tions that succeed each other in place,over time. From the Division of Reproductive Biology,The Johns HopkinsUniversity,School of Hygiene and Public Health,Baltimore,MD 21205,USAᮊ1998Academic Press1084-9521r98r040417q05$30.00r0The result of this arrangement is the continuous and asynchronous production of male gametes.The production of appropriate numbers of sper-matozoa depends upon stimulation of the testes by the gonadotropic hormones,follicle-stimulating hor-Ž.Ž.mone FSH and luteinizing hormone LH,both produced by the pituitary gland in response to go-Ž.nadotropin-releasing hormone GnRH from the hypothalamus.In response to LH,testosterone is produced by the Leydig cells.It has been clear for decades that testosterone is a necessary prerequisite for the maintenance of established spermatogenesis in the adult testis,and for the restoration of sper-matogenesis in testes induced experimentally to be-come azoospermic.The possible involvement of FSH in regulating spermatogenesis in the adult mammal also has been studied extensively,but its role contin-ues to be debated in the literature.The mechanisms by which testosterone and FSH regulate spermato-Žgenesis remain uncertain as well for reviews,see refs .1᎐4.Androgens and spermatogenesisLeydig cells present in the interstitial compartment are the source of testosterone in the testes.It is not surprising,therefore,that there is a considerably higher concentration of testosterone within the testes than in blood serum.5Interestingly,in a study of rats in which the local production of testosterone by Leydig cells was suppressed,administered testos-terone became highly concentrated in the testes.6 This surprising result indicates that testosterone be-comes concentrated within testicular compartments even under circumstances in which there is no local production of the steroid.As yet,the explanation for this phenomenon is not known,417B.R.ZirkinThe average testosterone concentration in ratŽ.seminiferous tubulefluid STF is approx.50᎐60Žy7.5ng r ml 1.7=10M,which is considerably higherŽthan the K of testicular androgen receptors3= dy9.10M.This suggests that the testosterone concen-tration in the seminiferous tubules may be considerably in excess of that required to maintain established spermatogenesis in the adult rat.Indeed, studies of the quantitative relationship between intra-testicular testosterone concentration and the number of advanced spermatids per testis have proven this to be the case.Thus,quantitatively complete spermatogenesis has been shown experimentally to be maintained at a STF testosterone concentration of 20ng r ml,which represents a reduction of60%from controls.6Similarly,in adult ratsfirst rendered azoospermic with a contraceptive dose of testos-terone and then administered additional testos-terone,the intra-testicular testosterone concentration required for the restoration of quantitatively com-plete spermatogenesis also was20ng r ml.7These observations indicate that spermatogenesis can be maintained or restored,quantitatively,at testosterone concentrations far lower than those normally present within the testis,but still an order of magnitude higher than that present in serum.Why are such high testosterone concentrations re-quired to maintain or restore spermatogenesis?One possibility is that not all of the intra-testicular testos-terone is free to bind to androgen receptors.Indeed, Roberts and Zirkin8reported that in rats receiving contraceptive testosterone implants,the concentra-Ž.tion of androgen-binding protein ABP in the STF was about the same as the intra-testicular testosteroneŽconcentration that resulted from the implants30᎐40 .nM;and,moreover,that ABP binding of testos-terone was capable of suppressing testosterone-dependent gene transcription.Binding of testos-terone thus might explain why a contraceptive dose fails to sustain spermatogenesis.However,remaining unanswered is the question of why so much testos-Ž.terone70nM,or20ng r ml relative to androgen receptor concentration,is required minimally to maintain or restore spermatogenesis.In fact,does testosterone function via the andro-gen receptor?Testosterone concentration in the STF of the rat testes is approximately one order of magni-tude greater than the concentration of5␣-Ž.5 dihydrotestosterone DHT,the major regulatory steroid for most androgen-dependent organs of the male reproductive tract.If intra-testicular androgenssupport spermatogenesis via the androgen receptor, then DHT,with its greater affinity for,and slower rate of dissociation from,the androgen receptor should be more effective in maintaining spermato-genesis than testosterone.In fact,Chen et al9demon-strated that under experimental conditions in which endogenous testosterone production was suppressed, the minimal average STF concentration of DHT re-quired to quantitatively maintain spermatogenesis Ž.7᎐8ng r ml was only one-third the required testos-terone concentration.These results indicate that DHT maintains spermatogenesis more effectively than testosterone,which is consistent with the greater Ž.three to four-fold affinity of DHT than testosterone for the androgen receptor.Does DHT play a role in regulating spermatogenesis?In a study designed to determine whether DHT might be involved in the restoration of spermiogenesis in vivo,O’Donnell et al10reported that5␣-reductase inhibition signifi-cantly suppressed the hourly production rate of round spermatids,in a setting in which testosterone concen-tration was relatively low.Whether or not DHT playsa role in regulating spermatogenesis under normalcircumstances of high intra-testicular testosterone re-mains an open question.FSH and spermatogenesisThough the involvement of FSH in spermatogenesis has been studied extensively,considerable uncer-tainty remains about the circumstances under which it is required,and what it does.The uncertainty stems in part from apparent species differences in the effects of FSH on spermatogenesis,and in part on the time of the life cycle during which FSH is either administered or withdrawn.Numerous studies through the years have reported that both testos-terone and FSH are required for quantitatively nor-mal spermatogenesis in non-human primates,human adults,and seasonally breeding rodents.3,11᎐13There also have been numerous reports that FSH is re-quired for the initiation of spermatogenesis in imma-ture rats.3However,an important recent study re-ported that FSH-deficient male mice,though their testes were small,were fertile,14arguing against the necessity for FSH even for the initiation of spermato-genesis.Through the years,controversy about the involve-ment of FSH has come largely from studies of the adult rat.The administration of testosterone to rats at418Spermatogenesis:its regulation by testosterone and FSHthe time of hypophysectomy,or after hypophysec-tomy-induced testicular regression,fails to maintain or restore spermatogenesis,quantitatively,2suggest-ing that pituitary factors in addition to LH may participate in the regulation of spermatogenesis.A number of studies of this kind have concluded that it is the absence of FSH rather than the absence of other pituitary factors that explains the inability of testosterone to sustain or restore spermatogenesis after hypophysectomy,a conclusion that is consistent with reported synergy between FSH and testosterone in the adult.3For example,the administration of recombinant FSH,by itself or together with testos-terone,to hypophysectomized adult rats has been shown in some studies to prevent germ cell loss,andŽ.in others to restore spermatogenesis qualitatively to germ cell-depleted testes.3,15᎐17A conclusion from these studies,taken together,is that FSH can affect adult spermatogenesis.The question is:Does FSH affect adult spermatoge-nesis under non-experimental conditions?We reported that the administration of testosterone to adult rats actively immunized against GnRH main-tained spermatogenesis,quantitatively,despite the absence of detectable FSH.18A similar conclusion, that FSH has little or no effect on spermatogenesis, had been reached earlier by Dym et al19,who re-ported that the administration of FSH antiserum to adult rats did not affect testes weight or germ cell number.Also,as demonstrated by our laboratory, testosterone administration to rats made azoospermic by active immunization against GnRH,restored sper-matogenesis quantitatively,again in the absence of detectable FSH.20Taken together,these reports sug-gested that FSH is not required either for the mainte-nance or restoration of spermatogenesis,at least un-der circumstances in which normal,high levels of testosterone are present in the testes.From more recent studies,it has become clear that FSH may stimulate early events in spermatogenesis, including spermatogonial proliferation and meiosis, but only testosterone is able to sustain complete spermatid differentiation.21᎐23Perhaps most convinc-ingly,recent analyses of spermatogenesis in a setting in which there are defects in the genes encoding FSH and its receptor,have shown that spermatogenesis and male fertility are possible without FSH action.24,25 Taken together,these studies argue that under cer-tain circumstances,FSH may play a role in maintain-ing or restoring quantitatively complete spermato-genesis.However,it is becoming apparent that com-plete spermatogenesis can occur,and fertility can be realized,in the absence of FSH.How do testosterone and FSH regulatespermatogenesis?It is well established that acute testosterone with-drawal depletes the testes of specific germ cells at particular stages of the cycle of the seminiferous epithelium;rapid effects typically occur on pachytene spermatocytes and spermatids at stages VII᎐VIII.3 Significantly,the conversion of spermatids of step7 to step8is inhibited by testosterone withdrawal,and restored when testosterone levels are replenished.22 Why are cells lost,and why do particular cells fail to proceed in development?In part,testosterone with-drawal results in changes in the adhesion of sper-matids to the Sertoli cells with which they are associ-ated,26probably via effects on the Sertoli cell cytoskeletal components,actin and vinculinfila-ments,and thus on the junction between the sper-matids and Sertoli cells.Loss of spermatid adhesion, in turn,precludes further maturation of these cells.Additionally,Sharpe et al3have shown that androgen depletion markedly reduces the level of synthesis of particular proteins by tubules at defined stages of the cycle.As yet,the identity and function of these pro-teins,which apparently are regulated by androgen, are not known.Finally,testosterone is known to sti-mulate the secretion of some well known Sertoli cell proteins,including ABP,and to inhibit the produc-Ž.3 tion of others e.g.plasminogen activator.Unfortu-nately,it is not known if and how these androgen-de-pendent proteins function in the regulation of sper-matogenesis.It has been appreciated only recently that the germ cell losses that occur normally during spermatogene-sis,or after experimental manipulation,are attributable to apoptosis.27᎐33Apoptosis,a mode of physiological cell death,originally was characterized by morphological changes that include nuclear and cytoplasmic condensation,membrane blebbing,re-lease of apoptotic bodies,and phagocytosis of these apoptotic bodies by neighboring cells.One of the precipitating events associated with the onset of apoptosis is endonuclease-catalyzed cleavage of cellu-lar DNA at internucleosomal sites,which generates DNA fragments in size multiples of approx.185base pairs.A number of studies28᎐32have shown that the withdrawal of testosterone from rat testes leads to increased germ cell apoptosis.These data suggest419B.R.Zirkinthat testosterone in some way may function as a cell survival factor,protecting germ cells from apoptotic death.Unfortunately,the molecular mechanisms by which testosterone does this,have received very little attention.A recent study has implicated the Fas system in germ cell apoptotic death in the rat testes.33Fas is a transmembrane receptor protein that belongs to the TNF r NGF receptor family.It contains a domain that is capable of initiating apoptosis when stimulated byŽ.its ligand,Fas ligand FasL,which is a TNF-related transmembrane protein.The FasL r Fas interaction triggers the death of cells expressing Fas,and is associated downstream with activation of interleukin-1Ž.34converting enzyme ICE.FasL,has been localized to Sertoli cells,and Fas to germ cells in the rat testis.33Expression of the FasL and Fas genes and their protein products have been shown to be up-regulated following xenobiotic exposure of rats,lead-ing to germ cell apoptosis.33It is possible that intra-testicular testosterone concentration affects the Fas system,and that this system thus provides a mecha-nism by which testosterone ultimately can affect germ cell apoptosis via its site of action,the Sertoli cell. The protooncogene product Bcl-2,when elevated in cells in vivo or in vitro,exerts an anti-apoptotic effect.35It has been established that there is a family of Bcl-2proteins,some of which inhibit while others promote apoptosis.Interactions between these groups of proteins within cells have been shown to antago-nize their opposing functions and thereby to modu-late the sensitivity of cells to apoptosis.35It is now established that apoptosis occurs in response to testosterone withdrawal,and that Bcl-2family mem-bers are present in mammalian testes.36,37 However,little attention has been paid to the possi-bility that the Bcl-2proteins might be involved in the molecular regulation of apoptosis within the testes. There is some evidence that FSH also may be involved in suppressing cell death.Russell et al16 have shown that rhFSH partially prevents degenera-tion of germ cells in the testes of hypophysectomized rats.McLachlan et al38similarly have reported that rhFSH partially restores spermatogenesis in rats,by supporting the survival of spermatogonia and pro-moting subsequent maturation steps through round spermatids.A recent study reported that im-munoneutralization of FSH in both immature and adult rats causes increased apoptosis,with spermato-gonia and pachytene spermatocytes particularly be-ing susceptible.39These studies all support a role for FSH in cell survival.ConclusionsThough it has been clear for decades that testos-terone is a necessary prerequisite for spermato-genesis,the mechanisms by which it acts remain uncertain.This uncertainty has resulted in part from the recognition that the intra-testicular testosterone concentration required minimally to maintain or re-store spermatogenesis,is considerably higher than the K of androgen receptors.Why such high testos-dterone concentrations are required in the testes is not known.Even more uncertain is the role that FSH plays in spermatogenesis.Numerous studies have re-ported that FSH is required for the initiation of spermatogenesis,and,in the adult,that it acts syner-gistically with testosterone to maintain or restore spermatogenesis.However,a recent study reported that FSH-deficient male mice are fertile,arguing against the necessity for FSH even for the initiation of spermatogenesis.Consistent with this,studies have reported that qualitatively complete spermatogenesis and fertility are possible in settings in which there are defects in the genes encoding FSH and its receptor.These studies argue that FSH may play a role in maintaining or restoring quantitatively complete spermatogenesis.However,qualitatively complete spermatogenesis can occur,and fertility can be real-ized,in the absence of FSH.Both testosterone and FSH suppress germ cell apoptotic death,and thus are considered to play roles in cell survival.The molecu-lar mechanisms by which either testosterone or FSH accomplish this remain to be determined.ReferencesŽ.1.Roberts KR,Zirkin BR1991Androgen regulation of sper-matogenesis in the rat.Ann NY Acad Sci637:90᎐106Ž.2.Zirkin BR1993Regulation of spermatogenesis in the adultmammal:gonadotropins and androgens,in Cell and Molecu-Ž.lar Biology of the Testis DesJardins C,Ewing LL,eds pp166᎐188.Oxford University PressŽ.3.Sharpe RM1994Regulation of spermatogenesis,in TheŽ.Physiology of Reproduction,2nd ed Knobil E,Neill JD,edspp1363᎐1434.Raven Press,New York4.McLachlan RI,Wreford NG,O’Donnell L,deKretser DM,Ž.Robertson DM1996The endocrine regulation of spermato-genesis:independent roles for testosterone and FSH.J En-docrinol148:1᎐95.Turner TT,Jones CE,Howards SS,Ewing LL,Zegeye B,Ž.Gunsalus GL1984On the androgen microenvironment ofmaturing spermatozoa.Endocrinology115:1925᎐1932Ž.6.Zirkin BR,Santulli R,Awoniyi CA,Ewing LL1989Mainte-nance of advanced spermatogenic cells in the adult rat testis:quantitative relationship to testosterone concentration withinthe testis.Endocrinology124:3043᎐30497.Awoniyi CA,Santulli R,Sprando RL,Ewing LL,Zirkin BR 420Spermatogenesis:its regulation by testosterone and FSHŽ.1989Restoration of advanced spermatogenic cells in the experimentally regressed rat testis:quantitative relationship to testosterone concentration within the testis.Endocrinology 124:1217᎐1223Ž.8.Roberts KP,Zirkin BR1993Androgen binding proteininhibition of androgen-dependent transcription explains the high minimal testosterone concentration required to main-tain spermatogenesis in the rat.Endocrine J1:41᎐47Ž.9.Chen H,Chandrashekar V,Zirkin BR1994Can spermato-genesis be maintained quantitatively in intact adult rats with exogenously administered dihydrotestosterone?J Androl 15:132᎐13810.O’Donnell L,Stanton PG,Wreford NG,Robertson DM,Ž.McLachlan RI1996Inhibition of5␣-reductase activity im-pairs testosterone-dependent restoration of spermiogenesis in adult rats.Endocrinology137:2703᎐2710Ž.11.Matsumoto AM1989Hormonal control of spermatogenesis,Ž.in The Testis,2nd ed Burger H,deKretser DM,eds pp 181᎐196.Raven Press,New YorkŽ.12.Weinbauer GF,Nieschlag E1993Hormonal control ofspermatogenesis,in Molecular Biology of the Male Reproduc-Ž.tive System deKretser DM,ed pp99᎐142.Academic Press, New YorkŽ.13.Marshall GR,Zorub DS,Plant TM1995Follicle-stimulatinghormone amplifies the population of differentiated sper-matogonia in the hypophysectomized testosterone-replacedŽ.adult Rhesus monkey Macaca mulatta.Endocrinology 136:3504᎐3511Ž.14.Kumar RR,Wang Y,Lu NE,Matzuk,MM1997Folliclestimulating hormone is required for ovarian follicle matura-tion but not male fertility.Nat Genet15:201᎐204Ž.15.Kerr JB,Maddocks S,Sharpe RM1992Testosterone andFSH have independent,synergistic and stage-dependent effects upon spermatogenesis in the rat testis.Cell Tissue Res 268:179᎐18916.Russell LD,Corbin TJ,Borg KE,De Franca LR,Grasso P,Ž.Bartke A1993Recombinant human follicle-stimulating hor-mone is capable of exerting a biological effect in the adult hypophysectomized rat by reducing the numbers of degener-ating germ cells.Endocrinology133:2062᎐2070Ž.17.Sinha Hikim AP,Swerdloff RS1995Temporal and stage-specific effects of recombinant human follicle-stimulating hormone on the maintenance of spermatogenesis in gonadotropin-releasing hormone antagonist-treated rats.En-docrinology136:253᎐261Ž.18.Awoniyi CA,Zirkin BR,Chandrashekar V,Schlaff WD1992Exogenously administered testosterone maintains spermato-genesis in adult rats actively immunized against GnRH.En-docrinology130:3283᎐328819.Dym M,Raj HGM,Lin YC,Chemes HE,Kotite NJ,NayfehŽ.SN,French FS1979Is FSH required for maintenance of spermatogenesis in adult rats?J Reprod Fertil Suppl 26:172᎐18120.Awortiyi CA,Santulli R,Chandrashekar V,Schanbacher BD,Ž.Zirkin BR1989Quantitative restoration of advanced sper-matogenic cells in adult male rats made azoospermic by active immunization against luteinizing hormone or go-nadotropin-releasing hormone.Endocrinology125: 1303᎐1309Ž.21.Singh J,Handelsman W1996The effects of recombin-ant FSH on testosterone-induced spermatogenesis inŽ.gonadotrophin-deficient hpg mice.J Androl17:382᎐393 22.O’Donnell L,McLachlan RI,Wreford NG,Robertson DMŽ.1994Testosterone promotes the conversion of round sper-matids between stages VII and VIII of the rat spermatogenic cycle.Endocrinology135:2608᎐261423.McLachlan RI,Wreford NG,deKretser DM,Robertson DMŽ.1995The effects of recombinant follicle-stimulating hor-mone on the restoration of spermatogenesis in thegonadotropin-releasing hormone-immunized adult rat.En-docrinology136:4035᎐4043Ž.24.Tapanainen JS,Aittomaki K,Huhtaniemi IT1997Newinsights into the role of follicle-stimulating hormone in re-production.Ann Med29:265᎐26625.Tapanainen JS,Aittomaki K,Min J,Vaskivuo T,HuhtanierniŽ.IT1997Men homozygous for an inactivating mutation ofŽ.the follicle-stimulating hormone FSH receptor gene presentvariable suppression of spermatogenesis and fertility.NatGenet15:205᎐20626.O’Donnell L,McLachlan RI,Wreford NG,deKretser DM,Ž.Robertson DM1996Testosterone withdrawal promotesstage-specific detachment of round spermatids from the ratseminiferous epithelium.Biol Reprod55:895᎐901Ž.27.Allen DJ,Harmon BV,Roberts SA1992Spermatogonialapoptosis has three morphologically recognizable phases andshows no circadian rhythm during normal spermatogenesis inthe rat.Cell Prolif25:242᎐250Ž.28.Tapanainen JS,Tilly JL,Vihko KK,Hsueh AJW1993Hor-monal control of apoptotic cell death in the testis:go-nadotropins and androgens as testicular cell survival factors.Mol Endocrinol7:643᎐65029.Billig H,Furuta I,Rivier C,Tapanainen J,Parvinen M,HsuehŽ.AJW1995Apoptosis in testis germ cells:developmentalchanges in gonadotropin dependence and localization toselective tubule stages.Endocrinology136:5᎐12Ž.30.Sinha-Hikim AP,Wang C,Leung A,Swerdloff RS1995Involvement of apoptosis in the induction of germ cell de-generation in adult rats after gonadotropin-releasing hor-mone antagonist treatment.Endocrinology136:2770᎐2775Ž.31.Henriksen K,Hakovirta H,Parvinen M1995Testosteroneinhibits and induces apoptosis in rat seminiferous tubules in astage-specific manner:In situ quantification in squash prepa-rations after administration of ethane dimethane sulfonate.Endocrinology136:3285᎐3291Ž.32.Marathe GK,Shetty J,Dighe RR1995Selective im-munoneutralization of luteinizing hormone results in theapoptotic cell death of pachytene spermatocytes and sper-matids in the rat testis.Endocrine J3:705᎐709Ž.33.Lee J,Richburg JH,Younkin SC,Boekelheide K1997TheFas system is a key regulator of germ cell apoptosis in thetestis.Endocrinology138:2081᎐2088Ž.34.Enari M,Hug H,Nagata S1995Involvement of ICE-likeprotease in Fas-mediated apoptosis.Nature375:78᎐81Ž.35.Korsmeyer SJ1995Regulators of cell death.Trends Genet11:101᎐10536.Krajewski S,Krajewska M,Shabaik A,Wang H-G,Irie S,FongŽ.L,Reed JC1994Immunohistochemical analysis of in vivopatterns of Bcl-X expression.Cancer Res54:5501᎐550737.Krajewski S,Krajewska M,Shabaik A,Miyashita T,Wang HG,Ž.Reed JC1994Immunochemical determination of in vivodistribution of Bax,a dominant inhibitor of Bcl-2.Am JPathol145:1323᎐133638.McLachlan RI,Wreford NG,deKretser DM,Robertson DMŽ.1995The effects of recombinant follicle-stimulating hor-mone on the restoration of spermatogenesis in the go-nadotropin-releasing hormone-immunized adult rat.En-docrinology136:4035᎐4043Ž.39.Shetty J,Marathe GK,Dinghe RR1996Specific im-munoneutralization of FSH leads to apoptotic cell death ofthe pachytene spermatocytes and spermatogonial cells in therat.Endocrinology137:2179᎐2182421。
