HAIR. Department of Zoology, University of Sheffield, Sheffield, England

Transcription

1 THE JOURNAL OF INVESTIGATIVE DERMATOLOGY, 67 :98-15, 1976 Cpyright 1976 by The Williams & Wilkins C. Vl. 67, N.1 Printed in U.S.A. F. J. EBLING, D.Se., PH.D. Department f Zlgy, University f Sheffield, Sheffield, England The psychlgic imprtance f hair t man is in inverse rati t its physical functin. Except fr scalp hair and desultry areas f sexual hair, mst f man's hair fllicles are vestigial. Three prblems f hair grwth remain t be slved: (1) hw the intermittent activity f hair fllicles in bth animals and man is cntrlled; (2) hw the male hrmne alters the hair cycle in human skin; and (3) why larger hairs are prduced by teststerne in sme areas f the bdy when in sme individuals the hair fllicles in the scalp regress. Studies in which skin grafts frm rats f different ages were exchanged shwed that hair fllicles are innately prgrammed but can be slwly influenced by systemic factrs. Sterid hrmnes, especially estrgens, slw dwn the mult cycle whereas thryid hrmnes accelerate it. What establishes the innate rhythm remains prblematical. The fact that plucking ut the club hair initiates activity in resting fllicles has been explained by the hypthesis that the mittic inhibitr which accumulates during anagen is nrmally used up r dispersed during telgen r by wunding. Hwever, cntrary t this thery, fllicular activity is nt prlnged by epilatin during anagen. Mrever, if rats are epilated within ne r tw days f eruptin, nly club hairs are remved since frceps cannt grasp the tips f the new hairs. Such epilatin des nt affect the anagen in prgress, but remarkably enugh the subsequent resting phase is shrtened. Bth sexual hair and male-pattern baldness depend n andrgenic hrmnes. Target rgans f teststerne cnvert the hrmne t active metablites, chiefly 5a-dihydrteststerne. In skin, hwever, 5a-dihydrteststerne may nt be the nly active tissue andrgen. The majr metablite f teststerne incubated with hair rts is andrstenedine, and hirsute wmen withut ther bvius endcrine abnrmality smetimes excrete high levels f andrstanedil. Bth sterids stimulated the sebaceus glands f hypphysectmized-castrated rats, which, hwever, shwed nly a limited respnse t teststerne. The andrgenic sterids, the enzymes that cnvert them t their active metablites, and the prteins that bind them are undubtedly very imprtant t the prblems f the grwth f sexual hair and f male-pattern baldness. Hair perfrms n vital functin whatsever in man whse bdy culd be perpetually depilated r shaved withut any physilgic disadvantages. But, the psychlgic functins f hair seem almst immeasurable. Scalp hair is a majr scial and sexual display feature f the human bdy: fr wmen, the crwning glry f decently expsable femininity; fr men, a traditinal symbl f masculinity. If the lack f scalp hair is a disaster fr wmen, excessive facial r bdy hair beynd the cultural nrm set by the sciety in which they live can be almst as distressing. It is little cmfrt t such a wman t be tld she has idipathic hirsutism! Hair cannt, therefre, be scientifically ignred. But hw are the cmplex and seemingly insuperable prblems f hair grwth t be slved? Is it pssible t make any useful hyptheses abut human hair fllicles frm experiments carried ut n animal mdels? Reprint requests t Dr. F. J. Ebling, Department f Zlgy, The University, Sheffield S1 2TN, England. 98 The evlutinary histry f hair is s cnfused as t utterly mislead the cmparative zlgist, let alne the bichemist with perhaps a less subtle appreciatin f natural selectin. Hw ften, fr example, d we read f sme material being tested n hair grwth in the guinea pig as if the experiment were smehw related t the alleviatin f human-pattern alpecia? The difficulty arises frm the versatile functin f the hair fllicle. We d nt knw hw r where hairs riginated in evlutin r what was their riginal purpse. Perhaps, as Madersn [1] and thers have suggested, they first arse as part f mechanreceptr units, develping in the hinge regins between the scales f sme reptilian ancestr. Certainly, the evlutinary success f the warm-blded mammals wes much, if nt all, t the prperties f a hairy pelage as a heat insulatr. T judge by the ppularity f fur cats and wl garments, the fiber prduced by the hair fllicle still hlds its wn against the artifacts f mdern technlgy. These cnsideratins d nthing t explain any

2 July 1976 selective advantages f hair t man. Human hair fllicles are still sense rgans and t tug the hair causes pain, but men d nt explre their envirnments with vibrissae. And if the survival f the hairy mammals is due t their ability t keep warm, the emergence f glbal man depended n his capacity t keep cl. The highly vascular human skin, which can switch frm radiating heat t cnserving it by shunting the bld flw belw a layer f insulating fat and, cnversely, can increase heat lss by evaprating sweat frm millins f glands, is a far mre adaptable structure than the slwly multing animal pelage. Mst f ur hair fllicles are vestigial. At best, ur hairy heritage is nw represented by scalp hair, perhaps as a prtectin against the nnday trpical sun befre the inventin f hats, and by desultry areas f s-called sexual hair. These are clearly f tw types. The male beard must be cmpared t the lin's mane and ther similar structures; it is a secndary sexual character, whatever that may mean. Presumably it was riginally mre cncerned with aggressive displays against ther males than with sex appeal t females, which may explain why many f us shave it ff in ur suppsedly nnaggressive, peaceful, and cperative scieties. The remainder f the sexual hair, the pubic and axillary, cmmn t males and females alike, is, I believe, mre likely t be cncerned with disseminating glandular dr than (as sme authrs have suggested) with preventing frictin burns in weight lifting, sexual intercurse, and ther heavy exercise. Chase and Eatn [2,3] and Argyris [4-6] were amng thse wh pineered the experimental study f the cntrl f the hair fllicle, the ne stressing the rle f inhibitrs, the ther that f wund hrmnes. Van Sctt and c-wrkers were respnsible fr magnificent balsa wd recnstructins f pilsebaceus apparatus [7] as well as fr imprtant studies f alpecia areata [8,9]. Pinkus [1] wrte an authritative accunt f the embrylgy f the hair fllicle, drawing largely frm the riginal bservatins f his illustrius father. Kligman [11] prvided a classic accunt f the human hair cycle which included the first descriptin f catagen in scalp fllicles as well as ne f the first descriptins f human hair lss in terms f the dynamics f the fllicle [12]. Hair grwth was and has remained ne f the fremst interests f William Mntagna, an interest with which he has successfully infected a lng line f clleagues and pupils. I shall describe my wn data cncerning hair grwth and shall leave anatmy, embrylgy, histlgy, histchemistry, and electrn micrgraphy :ltrictly alne. I must mit the interesting and lmprtant experimental studies f Chen [13] and Oliver [14,15] n the interactins between the dermal papilla and the epidermis in the inductin f hair fllicles. Fr me, there are tw majr prblems t cnsider. What cntrls the cyclic activity f the hair fllicle? Hw des male hr- 99 mne alter the cycle in the human skin and, in particular, why des it prduce larger hairs in sme areas f the bdy but, in sme individuals, cause the fllicles in the scalp t regress? THE CONTROL OF THE CYCLE That virtually all hair fllicles underg cycles f activity in which an active phase (anagen) alternates with a resting phase (telgen) is knwn t all. Such cycles ccur nt nly in species which shw well-marked patterns f mult as the animals grw up r as the seasns change, but als in species, such as the guinea pig, which d nt. Prbably there is n sharp distinctin between the frms. Fllicular activity in guinea pigs is nt synchrnized within each regin t the same extent as it is in rats r mice, but neither des it, at least in yung animals, ccur at randm. Pulse labeling with [35S ]cysteine reveals that activity is t sme extent synchrnized within each f several different fllicle' types which are, hwever, ut f phase with each ther [16]. The fllicles in the nenatal human scalp appear t shw a measure f synchrny [17], as d the fllicles within each tri grup thrughut the bdy [18]. What cntrls the cyclic activity f hair fllicles? Experiments n labratry rats, in which the mults are nt nly very well marked in adlescence but cntinue thrughut life even under cnstant envirnmental cnditins, have thrwn sme light n the prblem. Befre each mult, fllicle activity starts in the ventral regin and mves drsally acrss the flanks ; prductin f the new hairs is fllwed by shedding f the clubs frmed in the previus cycle. There are varius ways, as we shall see, in which the passage f the mult can be accelerated r delayed, but it always ccurs in sequence (fr reviews, see [19,2]). Leaving aside the prblem f what triggers the start f the mult, the hypthesis that it spreads by prpagatin, that each active fllicle smehw stimulates its immediate neighbr, is as attractive as it is errneus. The experimental evidence against it is clear-cut. When a full-thickness skin autgraft abut 15 mm wide is made in the flank f a rat while the fllicles are in telgen, the advancing frnt f the next wave f activity n the graft remains behind that n the adjacent flank [21]. If the wave f activity were prpagated, ne wuld expect that fllicles drsal t the graft wuld nt becme active until the frnt advancing ver the graft had reached the drsal margin. In fact, they becme active synchrnusly with thse n the adjacent ungrafted flank. If hair fllicles are independent f each ther, is each fllicle then inherently prgrammed, r des each fllicle individually respnd t a cmmn systemic stimulus? When the psitin f fllicles is altered by rtating r transpsing grafts, the fllicles largely cntinue in the rhythms f their riginal sites, althugh a small site effect can be demnstrated. Hwever, this experimental evidence des nt reslve the issue since it can be

3 1 EBLING argued that fllicles d nt necessarily have an innate prgram, nly a site-characteristic respnse t sme systemic factr. The prblem can be slved nly by exchanging hmgrafts between rats f different ages in different stages f the mult cycle. It is necessary t cmpare each hmgraft with aut grafts made n the ppsite flanks f the dnr and recipient, respectively. Such experiments (Fig. 1) shw that the hmgrafts at first cntinue in the same rhythm as the dnr autgrafts, but that gradually ver many weeks they cme int phase with autgrafts n the recipients [22]. S d the hair cycles f parabitically jined rats f unequal age [23]. It appears, therefre, that fllicles are innately prgrammed, but that they can als be slwly influenced by systemic factrs. The finding is perhaps nt surprising since multing is influenced by external changes in envirnment, f which the phtperid seems t be the Vl. 67, N.1 mst imprtant, thugh temperature may exert a small mdifying influence (see [2]). While a definitive statement f the systemic factrs respnsible fr mediating between envirnment and fllicle cannt be given, several hrmnal systems prfundly affect the mult cycle. Sterid hrmnes, especially estrgens, act as a brake, whereas thyrid hrmnes accelerate. Fr example, in intact female rats, generatin 3 (i.e.,. the secnd pstnatal replacement f hair) f fllicular activity starts in the ventral regin at abut 63 days f age, reaches the drsum by 9 days, and is cmplete by a little after 15 days. In the spayed animal, the wave reaches the drsum by 66 days and is cmplete at abut 92 days, that is, abut 6 days earlier than in the intact. If estradil is implanted, the wave des nt reach the drsum by 11 days [24,25]. In cntrast, thyrxine greatly accelerates the 123 autgraftil ~~~I ~+~ ~I~: 4:t~+~1---- yunger rats 2 3 lder rats 3 4 autgraft~i ~ ~'~P ~r~; ~: :~: ~t-::~ : :~JL-- lder rats 3 4 yunger rats autgraftl + + I 1: : , I I I I I I I age In days FIG. 1. Successive eruptins f hair n hmgrafts exchanged between rats aged 28 and 49 days, respectively, and n aut grafts made at the same time. Shade limits represent plus r minus 1.5 times the standard errrs fr a minimum f 9 bservatins f the sequential eruptins.

4 July 1976 passage f the mult. Thus, intact female rats given 1 j.lg f thyrxine per day cmpleted the mult in 96 days, at abut the same age as untreated spayed rats. On the ther hand, treatment with prpylthiuracil greatly delayed the mult s that the mid-drsum was nt reached until after 17 days and the mult had nt been cmpleted at 26 days [19,26]. The daily rate f elngatin and the definitive length f the hairs are significantly affected by hrmnes, i.e., they are reduced by estrgens and increased by thyrxine. After pulse-labeling with 35 [ 8 ]cysteine, the duratin f anagen in each particular regin can be reduced by thyrxine and, perhaps paradxically, als by estrgen [27]. But these effects are slight cmpared with thse n the duratin f telgen. Hrmnes apparently exercise cnsiderable cntrl ver quiescent fllicles, but nce anagen starts, the hrmnal effect becmes relatively slight. We must cnclude, therefre, that the hair fllicle is cntrlled by bth lcal and systemic factrs. Each fllicle has an intrinsic cycle characteristic f its regin, but especially in respect t the initiatin f anagen and t a limited extent during activity, it is at the same time susceptible t hrmnal systemic influences. But what cntrls the intrinsic cycle? Chase [2] and Chase and Eatn [3] pineered the hypthesis that a mittic inhibitr accumulates during anagen and is gradually used up r dispersed during telgen. This has much in cmmn with Bullugh's thery that cell divisin is nrmally held in check by tissue-specific chalnes [28]. The best evidence f the inhibitr hypthesis is that anagen can be initiated by plucking the club hair frm a resting fllicle r by wunding, the inference being that an accumulated inhibitr is remved r dispersed. It might then be imagined that systemic hrmnes culd act by mdifying the derma!' envirnment t influence the rate f dispersal f the inhibitr. In fact, there are prfund changes in the dermis during the fllicular cycle, including regular fluctuatins in cllagen and fat [29] as well as mast cells, histamine, and sertnin [3]. Hwever, under hrmnal treatments, the changes in the hair fllicle and in the dermis remain strictly in phase, s it is, in my view, impssible t claim that either ne is the cause f the ther. Fllicular activity, revealed by the uptake f tritiated thymidine, starts abut 1 hr after epilatin [31 J. The external eruptin f hairs, i.e., the emergence f mntrichs and awls (abut 1 t 2 rj. ays earlier than auchenes and zigzags) ccurs in ' he drsal flank f intact rats abut 11 days after.'l lucking [27] and n the midflank a little sner,fig. 2). Keratinizatin cmmences at the tips f "he new hairs 6 t 8 days befre eruptin. The :espnse f the fllicle t plucking is slightly but.)ignificantly affected by hrmnes, the interval being shrtened by spaying r thyrxine and lengthened by estrgen r prpylthiuracil. A meticulus study by pulse labeling with [35 8 ]cysteine 11 "" "", ",, 3 " "", "'" """, "" "", CDO 2 "", ",,,,, z,, l- e.. =>, 6", X: 1 ~ x ', x, I- x x ~, x 2 3 DAYS TO ERUPTION - e.. ", " 4 4: := w w I-,, I/),, >4: 1 NON-EPILATED FLANK FIG. 2. The effect f epilatin f fllicles in the midflank f intact female rats n the next eruptin. The number f days t eruptin n the epilated flank is pltted against the time t eruptin n the cntrl nnepilated flank. Plucking f club hairs mre than 1 days befre expected eruptin induces eruptin after 8 t 1 days. Plucking f club hairs less than 1 days befre expected eruptin, when fllicles are already active, has little r n effect n the time f eruptin. Plucking grwing hairs advances next eruptin even mre than plucking club hairs. Plucking G 1 club hairs frm alngside newly erupted G2 hairs des nt affect their grwth but shrtens the ensuing telgen., G2 club hairs;., G2 grwing hairs; 1::., G 1 club hairs frm alngside recently erupted G2 hairs; x, G 1 club hairs. reveals that epilatin has n effect n the ultimate length, du:r:atin f grwth, r mean rate f grwth f the hairs [27]. This hlds fr rats under hrmnal treatments as well as fr nrmal animals; i.e., hrmnal treatments have exactly the same effect n active fllicles regardless f whether the activity was spntaneus r induced by epilatin. Epilatin, which acts slely t break telgen, can be cmpared t the setting f an alarm clck: the time at which the alarm ges ff (r the hairs erupt) is changed withut altering the duratin f the peel (r the grwth f the hairs). All this reveals nthing f hw the active hair fllicles enter catagen r, indeed, f why epilatin f telgen hairs induces anagen. If the cause f quiescence were the build-up f an inhibitr, ne might expect that the fllicle culd be kept in cntinuus activity by plucking the grwing hairs and thus dispersing the inhibitry substance. But this des nt happen; if anything, remving the grwing hairs shrtens rather than lengthens the duratin f anagen [32). Hwever, such evidence against the inhibitry thery is by n means cnclusive since it can, with sme factual supprt, be argued that t wrench ut grwing hairs causes cnsiderable damage t the fllicle [33,34]. Indeed, repeated plucking reduces the number f fllicles with nrmal grwing hairs and, accrding t Hamiltn and Ptten [35], destrys the stimulusrespnsive cells. 8me new experimental evidence relating t this prblem can nw be ffered. If rats are epilated

5 12 Vl. 67, N.1 EBLING within 1 r 2 days f eruptin, the nly remvable hairs are clubs, since the tips f the new hairs d nt yet prtrude far enugh t be grasped by frceps. Such epilatin has n effect whatsever n the duratin f the anagen in prgress [27]. The new hairs cntinue t elngate, and the fllicles t enter catagen after the nrmal interval, cmpletely ignring the remval f the ld club. A remarkable cnsequence, hwever, is that the ensuing telgen phase is shrter than nrmal, that is t say, the next eruptin is advanced. Figure 3 shws the mean advance f successive eruptins after a single epilatin carried ut at varius stages f the cycle. Plucking f club hairs abut the time f spntaneus eruptin des nt affect it (first eruptin), but the next eruptin (secnd eruptin) is abut 6 days earlier. This evidence cannt easily be recnciled with the view that inhibitr builds up t a maximum twards the end f anagen and disperses during telgen t the pint where the fllicle again becmes active. On the cntrary, in this experimental situatin, initiatin f activity des nt appear t be directly related t the terminatin f the previus anagen but t a chain f events set in mtin by the remval f club hairs when anagen was already under way. In summary, we knw little abut the mechanism f the intrinsic cycle except that the facts cannt all be explained by the inhibitr build-up hypthesis. It has, hwever, been clearly established that during the resting phase the initiatin f anagen can be prfundly influenced by systemic factrs such as hrmnes which can advance Te lgen Angcn I - V f G2 ST ATE OF Angt:!n VI algi fo LLI CLES.AT Telgen Angen EPILA TI ON FIG. 3. The advance f successive eruptins f hair after a single epilatin in the midflank, shwn in relatin t the age f the rats and the stage f the cycle at which epilatin was carried ut. Means ± SEM are fr grups f 4 t 6 intact female rats. The first eruptin after epilatin is advanced, except where club hairs were plucked frm fllicles already in anagen. The effect was greatest when grwing hairs were plucked. The secnd eruptin is even mre advanced, by an additinal 1 days r mre, than the first. The third eruptin remains advanced by abut the same amunt as the secnd. ' r retard it by several weeks r by plucking the club which virtually verrides the ther cntrls. D epilated fllicles remain permanently ut f phase r, like transplanted fllicles n grafts, is the timing f eruptin slwly influenced by systemic factrs? Evidence suggests that fllicles d mre than simply remain ut f phase (Fig. 3). In the cycle immediately succeeding the induced ne, the effect is enhanced and anagen becmes even further advanced. The cyclic activity f the hair fllicle clearly remains a fascinating prblem f experimental bilgy with the final slutin still a glittering prize. But what is the relevance, if any, f these animal experiments t human prblems? D they, fr example, shed any light n cnditins in which scalp hair becmes thin? Certain types f transient alpecia have been ascribed t alteratins in the hair cycle. Pstpartum shedding f hair, fr example, appears t invlve the simultaneus precipitatin int catagen f large numbers f fllicles, suppsedly as a cnsequence f the withdrawal f circulating hrmnes which prevented the nrmal prcess in late pregnancy when an exceptinally high prprtin f fllicles are in anagen [36 J. A similar kind f hair lss smetimes ccurs within 2 t 3 mnths f febrile episdes such as t yphid fever, scarlet fever, r pneumnia. Kligman [12 J prpsed that the phenmenn be called " telgen effluvium" and believed that sme hair lss f apparently psychgenic rigin is f this type. He recunts the case f a prisner, under his clse surveillance, wh had been implicated in a murder and had std trial n three ccasins, each time escaping the death penalty n a legal technicality. His uncertain status in this wrld cntinued fr a perid f abut 3 years, until in a furth trial he was cnvicted f murder in the first degree. Abut a mnth later he presented transverse depressins (Beau's lines) in the prximal prtins f all his fingernails and became shwered with scales frm an extrardinary amunt f simple dandruff. Abut 1 weeks after the sentence he started t cmplain f hair lss, and by fanatically saving every hair, he even managed t cnvince Kligman that he had shed between 6 and 15 hairs per days fr a perid f 3 weeks. Kligman recrds that telgen cunts were abut 5 %, that substantial restitutin started abut 8 weeks after shedding, and that when the prisner was subsequently pardned his hair regrew cmpletely. Hair is als rapidly shed in alpecia areata. Lesins usually appear t start a fcal pint and spread utwards t frm patches, but the prcess can ccur diffusely. A study [37] in which hairs were successively plucked in a series f cncentric rings frm develping lesins shwed that at an early stage 1% f the hairs remved frm the center f the patch were club hairs and that a wave f telgen appeared t mve centripetally. Tw ther features f alpecia areata are the ccurrence f shrt prtruding club hairs with a frayed pint, ften knwn as "exclamatin marks," and the

6 July appearance as the lesins prgress, f many fllicles in a state f arrested anagen [8,9,38]. The simplest hypthesis t explain all these facts is that a transitry failure near the matrix causes the anagen hair t shed and that shearing f hair frm fllicles mving int catagen prduces the truncated prtruding exclamatin mark hairs (Fig. 4). Since the fllicles already in telgen remain unaffected, the nly remaining hairs that can be plucked are club hairs. Hwever, we cannt be sure that the disrder des nt als precipitate fllicles,i,' ~ int catagen as Eckert et al riginally prpsed.,/ LATE [37]. At least, a knwledge f the hair cycle enables. A]AGEN. AIAGEN us t understand smewhat the prgress f the lesin, even if it prvides n clue as t its cause. The facts abut male type baldness, such as they are, are widely knwn and can be rapidly summa.. rized. The clinical manifestatins, namely, the gradual replacement ver symmetrically patterned ANAGEN TELOGEN TELOGEN areas f the scalp f gd quality terminal hairs by csmetically useless fluff, has been amply defig. 4. Simplest explanatin f phenmena bserved scribed, nt nly in man but in even mre detail in in develpment f lesins in alpecia areata. splendid studies f the stump-tailed macaque [39,4]. Metablically, the fllicles d nt appear altered. The nly etilgic clue is that, even in adult human males has a very much greater ability subjects with an hereditary dispsitin t early than abdminal skin t cnvert teststerne [4-14C] baldness, the male testicular hrmne teststerne t 5a-dihydrteststerne [44]. What is perhaps is necessary fr its manifestatin. This single fact even mre interesting is that fetal skin frm the verrides all thers. genital areas has the same ability. Why, then, des teststerne, which n the face Male type hrmnes are unquestinably the agents which at puberty cause grwth f pubic and and sme ther regins f the bdy causes fllicles axillary hair and transfrm the vellus fllicles f t switch their hair prductin frm vellus t the face int a beard. An indirect demnstratin f terminal, apparently have the ppsite effect in this fact cncerns a yung Englishman [41] wh the scalp? Adachi et al [45] attempted t frmulate a had t spend perids f several weeks n a remte island and cnsequently had n pprtunity fr hypthesis t explain the develpment f pattern sexual activity. He nticed, hwever, that his alpecia. Frm plucked grwing scalp hairs, they beard grew faster (as measured by weighed shav- extracted adenyl cyclase, the enzyme respnsible ings) n the day befre he was t leave the island. fr prducing cyclic AMP, which is accepted as a Grwth reached a maximum during the first days cmmn mediatr f hrmnal actin at the celluf resumed sexual intercurse but declined t lar level and is nrmc>.lly increased by hrmnal nrmal within 4 t 6 days. That the effects were stimulatin. They shwed that adenyl cyclase due t fluctuatins in andrgen levels was sup- culd be inhibited in vitr by 5a-dihydrteststerprted by a clear crrelatin between beard grwth ne, but nt by teststerne. They suggested that and sebum prductin. After the results were such a decrease wuld limit the energy metablism described in Nature [41], a ppular rgan f the and prtein synthesis in hair fllicles in viv. Leaving aside the questin f why 5a-dihyBritish press reprted "Sexual intercurse makes the beard grw." The Times f Lndn, ever drteststerne shuld inhibit rather than enhance respectable, reprted "Sexual abstinence makes adenyl cyclase in the scalp, it is clear that this the beard grw." Neither was crrect. The true hypthesis requires that bald skin shuld cnvert interpretatin f the annymus results is that teststerne mre readily than hairy skin. The evidence f Bingham and Shaw [46,47] suggests anticipatin makes the beard grw. The respnse f target rgans t teststerne is that this may be true, even if it des nt firmly dependent upn its cnversin t metablites, f establish it. They incubated male scalp skin frm which 5a-dihydrteststerne is the mst impr- bald and hairy sites with [4-14 C ]teststerne. In urn t (fr review, see [42]). 5a- Reductin f tests- each f fur dnrs, bth the uptake and metablism f teststerne were greater in the bald than in 1 erne was first established fr the prstate, but it :!ppears t be equally necessary fr the respnses f the hairy site. Hwever, althugh 5a-dihydrtes'Jther rgans, including the seminal vesicles, pre- tsterne frmed abut 2% f the ttal radiactiv.mtial glands, and skin [43]. The limitatin f ity, less than that due t unchanged teststerne, mre than 1% was due t andrstanedine. We.~ exual hair t certain areas is almst certainly may cnclude, therefre, that a scalp remains :'elated t the lcal capacity f the skin t carry ut :;his cnversin. Fr example, the scrtal skin f hairy because it lacks 5a-reductase but that there : 7 : 7..' "

7 14 Vl. 67, N.1 EBLING are n qualitative differences between the actual metablites prduced by hairy and bald areas. The prblem has, hwever, several ther facets. First, it shuld nt be assumed that 5a-dihydrteststerne is necessarily the nly active andrgen in skin. When hair rts are incubated with labeled teststerne, andrstenedine is the majr metablite [48,49]. In additin t 5a-dihydrteststerne and andrstenedine, andrstanedine, andrsterne, and andrstanedils are frmed in target rgans. 5a-Andrstane-31J, 171J-dil, as well as andrstenedine, significantly stimulates the sebaceus glands f hypphysectmized-castrated rats [5] even thugh such glands give nly a very limited respnse t teststerne, pssibly because f a failure f 5a-reductin. When tissues are incubated in vitr with [3H ]teststerne, the 31J, 17,B-dil is a majr metablite f human scalp and back skin [51] and sme wmen with idipathic hirsutism r acne vulgaris excreted several times the nrmal amunts f andrstanedil even thugh their levels f urinary teststerne were nrmal [52 ]. The pssibility that mre than ne metablite f teststerne is hrmnally active is reinfrced by the evidence that different andrgenic sterids can affect target cells in different ways. Thus, in the seminal vesicles r preputial glands f the rat, bth 5a-andrstane-31J, 171J-dil and 5a-andrstane-3a, 17,B-dil have similar effects n increases in DNA as 5a-dihydrteststerne, but the 3a-dil has greater effect n cell metablism [53]. In the sebaceus glands, t, cell divisin and intracellular lipid synthesis appear t respnd differentially t different sterids [54]. A further pint t cnsider is that the respnse f the target rgan t andrgens depends n the specific and selective attachment f the active sterid t receptr prteins in the cell. The prcess in the prstate, fr example, is believed t invlve the immediate binding f the sterid t a cytplasmic receptr and its transfer, within 3 min, t the nucleus where it initiates new prtein synthesis. A defect in the andrgen receptr results in a failure f respnse; such a defect appears t explain the end-rgan insensitivity t andrgens in the hereditary disrder knwn as testicular feminizatin [55 ]. Specific cytsl receptrs have been identified in the prstate [56] and the hamster sebaceus gland [57]. The pssibility that 5a-andrstanedils culd be bund t a specific receptr in the cytplasm and influence events at the translatinal as distinct frm the transcriptinal level has been cnsidered by Rbel et al [56]. These authrs have established that 5a-andrstane-3,B, 17,B-dil, as well as 5a-dihydrteststerne, is retained in a micrsmal binding prtein in rat ventral prstate, but they cnclude that the physilgic significance f this is nt yet established. The issue is, perhaps, cnfused by the demnstratin that cytplasmic extracts f canine prstate in vitr have a specific receptr prtein fr 5a-andrstane-3a, 17 a-dil [58] and a claim that rat prstate cytsl will bind nly 3a, 17{j-dil and nt 3{j, 17{j in viv, and neither in vitr [59]. These authrs cncluded that the 3a,171J-dil exerts its bilgic effects mainly by cnversin t 5a-dihydrteststerne. Thugh mre needs t be dne, abstruse discussin f exactly which active andrgens are invlved and where they are bund is premature. The slutin t the prblem f pattern baldness lies with the sterid mlecules, the enzymes which cnvert them t active metablites, and the prteins which bind them. REFERENCES 1. Madersn PFA: When? why? and hw?: Sme speculatins n the evlutin f the vertebrate integument. American Zlgist 12: , Chase HB: Grwth f the hair. Physil Rev 34: , Chase HB, Eatn GJ: The grwth f hair fllicles in waves. Ann NY Acad Sci 83: , Argyris TS: The grwth-prmting effects f wunds n hair fllicles already stimulated by plucking. Anat Rec 143: , Argyris TS: Wund healing and the cntrl f grwth f the skin, Advances in Bilgy f Skin, vl 5, Wund Healing. Edited by W Mntagna, RE Billingham. Oxfrd, Pergamn, 1964, pp Argyris TS: Chalnes and the cntrl f nrmal, regenerative and neplastic grwth f the skin. American Zlgist 12: , Mntagna W, Van Sctt EJ: The anatmy f the hair fllicles, The Bilgy f Hair Grwth. Edited by W Mntagna, RA Ellis, New Yrk, Academic, 1958, pp Van Sctt EJ: Mrphlgic changes in pilsebaceus units and anagen hairs in alpecia areata. J Invest Dermatl 31:35-43, Van Sctt EJ, Ekel TM: Gemetric relatinships between the matrix f the hair bulb and its dermal papilla in nrmal and alpecic scalp. J Invest Dermatl 31: , Pinkus H: The embrylgy f hair, The Bilgy f Hair Grwth. Edited by W Mntagna, RA Ellis. New Yrk, Academic, 1958, pp Kligman AM: The human hair cycle. J Invest Dermatl 33:37-316, Kligman AM: Pathlgic dynamics f human hair lss. I. Telgen effluvium. Arch Dermatl 83: , Chen J: Dermis, epidermis and dermal papillae interacting, Advances in Bilgy f Skin, vl IX, Hair Grwth. Edited by W Mntagna, RL Dbsn. Oxfrd, Pergamn, 1969, pp Oliver RF: The vibrissa dermal papilla and its influence n epidermal tissues. Br J Dermat181:55-65, Oliver RF: The inductin f hair fllicle frmatin in the adult hded rat by vibrissa dermal papillae. J Embryl Exp Mrphl 23: , Jacksn D, Ebling F J: The activity f hair fllicles and their respnse t estradil in the guinea-pig Cavia prcellus L. J Anat 111:33-316, Pecrar V, Astre I, Barman JM: Cycle f the scalp hair f the new-brn child. J Invest Dermatl 43: , Saith M, Uzuka M, Sakamt M: Human hair cycle. J Invest Dermatl 54:65-81, Ebling FJ, Jhnsn E: The cntrl f hair grwth. Sympsium f the Zlgical Sciety f Lndn 12:97-13, Ebling FJ, Hale PA: The cntrl f the mammalian mult. Memirs f the Sciety fr Endcrinlgy 18: Ebling FJ, Jhnsn E: Hair grwth and its relatin t

8 July 1976 vascular supply in rtated skin grafts and transpsed flaps in the albin rat. J Embryl Exp Mrphl 7:417-43, Ebling FJ, Jhnsn E: Systemic influence n activity f hair fllicles in skin hmgrafts. J Embryl Exp Mrphl 9: , Ebling FJ, Hervey GR: The activity f hair fllicles in parabitic rats. J Embryl Exp Mrphl 12: , Jhnsn E: Quantitative studies f hair grwth in the albin rat. I. Nrmal males and females. J Endcrinl 16:337-35, Jhnsn E: Quantitative studies f hair grwth in the albin rat. II. The effect f sex hrmnes. J Endcrinl 16: , Ebling FJ, Jhnsn E: The actin f hrmnes n spntaneus hair grwth cycles in the rat. J Endcrinl 29: , Hale PA, Ebling F J: The effects f epilatin and hrmnes n the activity f rat hair fllicles. J Exp Zl 191:49-62, Bullugh WS: Mittic and functinal hmestasis: a speculative review. Cancer Res 25 : , Ebling FJ, Hale PA: The cmpsitin f female rat skin in relatin t regin, age, hair grwth cycle and hrmnes. J Endcrinl 36:177-21, Mretti G, Baccaredda-By A, Rebra A: Bichemical aspects f hair grwth, Advances in Bilgy f Skin, vl IX, Hair grwth. Edited by W Mntagna, RL Dbsn. Oxfrd, Pergamn, 1969, pp Pt ten CS: Tritiated thymidine incrpratin int hair fllicle matrix and epidermal basal cells after stimulatin. J Invest Dermatl 56: , Jhnsn E, Ebling F J: The effect f plucking hairs during different phases f the fllicular cycle. J Embryl Exp Mrphl 12: , Argyris TS: Hair grwth induced by damage, Advances in Bilgy f Skin, vl IX, Hair Grwth. Edited by W Mntagna, RL Dbsn. Oxfrd, Pergamn, 1969, pp Silver AF, Chase HB, Arsenault CT: Early anagen initiated by plucking cmpared with early spntaneus anagen, Advances in Bilgy f Skin, vl IX, Hair Grwth. Edited by W Mntagna, RL Dbsn. Oxfrd, Pergamn, 1969, pp Hamiltn E, Ptten CS: The effect f repeated plucking n muse skin cell kinetics. J Invest Dermatl 62:56-562, Lynfield YL: Effect f pregnancy n the human hair cycle. J Invest Dermatl 35: , Eckert J, Church RE, Ebling F J: The pathgenesis f alpecia areata. Br J Dermatl 8:23-21, Van Sctt EJ: Evaluatin f disturbed hair grwth in alpecia areata and ther alpecias. Ann NY Acad Sci 83:48-49, Un H, Adachi K, Mntagna W: Mrphlgical and bichemical studies f hair fllicle in cmmn baldness f stump-tailed macaque (Macaca specisa), Advances in Bilgy f Skin, vl IX, Hair Grwth. Edited by W Mntagna, RL Dbsn. Oxfrd, Pergamn, 1969, pp Takashima I, Mntagna W: Studies f cmmn baldness f the stump-tailed macaque (Macaca specisa). VI. The effect f teststerne n cmmn baldness. Arch Dermatl 13: , Annymus: Effects f sexual activity n beard grwth in man. Nature (Lnd) 226:869-87, King RJB, Mainwaring WIP: Sterid-Cell Interactins. Lndn, Butterwrths, Wilsn JD, Walker JD: The cnversin f teststerne t 5a-andrstan-17,B-l-3-ne (dihydrteststerne) by skin slices f man. J Clin Invest 48: , Flamigni C, Cllins WP, Kullapis EN, Craft I, Dewhurst CJ, Smmerville IF: Andrgen metablism in human skin. J Clin Endcrinl 32: , Adachi K, Takayasu S, Takashima I, Kan M, Knd S: Human hair fllicles: metablism and cntrl mechanisms. Sc Csmet Chern 21:91924, Bingham KD, Shaw DA: Male pattern baldness and the metablism f andrgens by human scalp skin. J Sc Csmet Chern 24: , 1973a 47. Bingham KD, Shaw DA: The metablism f teststerne by human male scalp skin. J Endcrinl 57: , 1973b 48. Sansne-Bazzn G, Reisner, RM, Bazzan, G: Cnversin f teststerne-1,2-3h t andrstenedine- 3H in the islated hair fllicles f man. J Clin Metab 34: , Schweikert HU, Wilsn JD : Regulatin f human hair grwth by sterid hrmnes. 1. Teststerne metablism in islated hairs. J Clin Endcrinl Metab 38: , Ebling FJ, Ebling E, McCaffery V, Skinner J: The respnses f the sebaceus glands f the hypphysectmized-castrated male rat t 5a-andrstanedine and 5a-andrstane-3,B, 17,B-dil. J Invest Dermatl 6: , Karunakaran MD, Pchi PE, Strauss JS, Clark SJ, Wtiz HH: In vitr metablism f 3H-teststerne by scalp and back skin: cnversin f teststerne t 3,B-andrstanedil. J Endcrinl (in press) 52. Mauvais-Jarvis P, Charransl G, Bbas-Masn F: Simultaneus determinatin f urinary andrstanedil and teststerne as an evaluatin f human andrgenicity. J Clin Endcrinl 36: , Schmidt H, Giba-Tziampiri, V Rtteck G, Vigt KD: Metablism and mde f actin f andrgens in target tissue f male rats. Acta Endcrinl (Kbh) 73: , Ebling FJ: Hrmnal cntrl and methds f measuring sebaceus gland activity. J Invest Dermatl 62: , Bardin CW, Bullck LP: Testicular feminizatin: studies f the mlecular basis f a genetic defect. J Invest Dermatl 63:75-84, Rbel P, Blndeau J-P, Baulieu E-M: Andrgen receptrs in rat ventral prstate micrsmes. Bichim Biphys Acta 373:1-9, Adachi K: Receptr prteins fr andrgen in hamster sebaceus glands. J Invest Dermatl 62: , Evans CR, Pierrepint CG: Demnstratin f a specific cytsl receptr in the nrmal and hyperplastic canine prstate fr 5a-andrstane-3a, 17 adil. J Endcrinl 64: , Krieg M, Hrst H-J, Sterba M-L: Binding and metablism f 5a-andrstane-3a, 17/3-dil and f 5a-andrstane-3,B, 17,B-dil in the prstate, seminal vesicles and plasma f male rats: studies in viv and in vitr. J Endrcrinl 64:529-53B, 1975