Internal and external timing puberty

Internal and external determinants of the timing of pubertyin the female

D. L. Foster, S. M. Yellon and Deborah H. OlsterReproductive Endocrinology Program, Department ofObstetrics and Gynecology and Division of

Biological Sciences, The University ofMichigan, Ann Arbor, Michigan 48109, U.S.A.

Summary. A working hypothesis is proposed to account for the timing of puberty infemale sheep. In the immature female, the frequency of LH pulses is low, and ovarianfollicles do not develop to an advanced stage. During the pubertal transition, thefrequency of LH pulses increases to drive follicular development and the productionof oestradiol which evokes the gonadotrophin surge and ovulation. Central to thehypothesis is the hypothalamic pulse generator for GnRH that directs the pattern andlevel of LH secretion. Growth-related cues are monitored to regulate the activity of theGnRH pulse generator. When a sufficient body size is attained, the frequency of LHpulses increases both because the sensitivity to oestradiol inhibitory feedback decreasesand because the GnRH pulse generator can be accelerated by the steroid. This increasein LH pulse frequency occurs provided the female has experienced the requisiteexposure to photoperiod, i.e. the long days of summer followed by the short days ofautumn. These photoperiodic cues are transduced by the pineal gland into a humoralsignal which is an increased nocturnal production of melatonin. Failure to grow to theappropriate body size or to experience the appropriate exposure to photoperiod leadsto a maintenance of the prepubertal anovulatory condition because the GnRH pulsegenerator operates at low frequency.

Introduction

Considerable knowledge has accumulated about many of the growth-related physical changes thatoccur in our own species during the transition into adulthood (Tanner, 1978). Furthermore, a

broad understanding of the hormonal control of puberty in children has been attained throughclinical investigations (Clayton, 1985, for review) based on concepts derived from animal studies inlaboratory conditions (for reviews see Ojeda, 1980 (rat); Norman, 1983 (macaque)). Yet, it shouldbe recognized that we have little information on how growth is linked mechanistically to theneuroendocrine systems governing the pattern and level of secretion of the hormones involved inthe pubertal process. Moreover, the mechanisms timing sexual maturity in the human being may beless complex than those in other species. This is because, unlike man, who has various degrees ofcontrol over his own environment, fertility in the vast majority of mammals is influenced greatly byseasonal changes in their surroundings. Mammals of most species exhibit a seasonal pattern ofreproduction such that conception is restricted to only part of the year. This represents an

additional constraint to their sexual maturation and the young female mammal of a seasonallybreeding species not only has to be sufficiently mature to begin reproduction, but also needs to matefor the first time at an optimal time of year so that the young are born in the seasons most favour¬able for their survival and nourishment. Sexual maturation of the domestic sheep (Ovis aries)

*Present address: Department of Physiology, Division of Perinatal Biology, Loma Linda University School of Medi¬cine, Loma Linda, CA 92350, U.S.A.

fPresent address: Department of Obstetrics and Gynecology, Columbia University College of Physicians andSurgeons, 630 West 168th Street, New York, New York 10032, U.S.A.

Downloaded from Bioscientifica.com at 02/02/2022 12:02:10AMvia free access

provides an excellent example of how multiple inputs into the brain are used to influence the onsetof fertility.

Foster & Ryan (1981) reviewed our understanding of the endocrine mechanisms governing thetransition into adulthood in the female lamb and the factors responsible for the timing of this tran¬sition were discussed briefly. The present paper intends to emphasize the latter and to incorporateinto the working hypothesis for the transition into adulthood how internal and external factorstime puberty. Again, we use as the model the Suffolk sheep, a breed in which the female exhibits adistinct breeding season (autumn and winter) and non-breeding season (spring and summer) (seeKarsch et al, 1984, for review). Much of the recent information is derived from studies conducted our laboratory. We have, as far as possible, differentiated concepts that we feel are firmlysupported by experimental evidence from those that are relatively speculative.

Hypothesis for puberty

An hypothesis for the neuroendocrine control of puberty in the female sheep is presentedschematically in Text-fig. 1. Central to the hypothesis is the gonadotrophin-releasing hormone(GnRH) pulse generator, the neural oscillator(s) that ultimately dictates patterns of secretion ofluteinizing hormone (LH) from the anterior pituitary gland. In the postpubertal lamb or adult ewe

during the follicular phase of the oestrous cycle, the frequency of the GnRH pulse generator is

Text-fig. 1. Hypothetical control of pulsatile LH secretion in the developing female sheepthrough modulation of the GnRH pulse generator by internal cues (melatonin, metabolicsignals). The internal cues mediate the effects of external factors (photoperiod, nutrition).Having experienced the appropriate photoperiod sequence (long-day melatonin then short-daymelatonin) and attained a sufficient physiological size (assessed by metabolic cues) forpregnancy, the LH pulse frequency increases to produce a preovulatory follicle. See text fordetails of hypothesis. (Redrawn from Foster, Olster & Yellon, 1985.)


relatively high, and LH pulses are produced at intervals of < 1 h. This high-frequency of LH pulsesdrives the follicle to increase its secretion of oestradiol to levels that induce the massive discharge ofgonadotrophin required for ovulation. In the immature lamb, the GnRH pulse generator is clearlyoperative because it has the ability to stimulate LH secretion at hourly intervals. However, it doesnot normally do so because the system is hypersensitive to the inhibitory feedback action of thesmall amounts of oestradiol secreted by the ovary. The resultant frequency of the GnRH pulsegenerator is therefore low, and LH pulses are produced at intervals of 2-3 h or greater. This low LHpulse frequency, although capable of effecting transient rises in oestradiol, is insufficient fordevelopment of the preovulatory follicle and its attendant oestradiol signal for the preovulatorygonadotrophin surge. During pubertal transition, the response to oestradiol inhibition is reduced,and the GnRH pulse generator is allowed to run at its inherently high frequency. As a consequence,the pituitary gland produces the requisite pattern of circulating LH necessary for successfulcompletion of the first follicular phase culminating in ovulation.

The activity of the GnRH pulse generator, and hence the timing of first ovulation, is influencedby internal and external factors. Metabolic cues, either substrates or products of metabolism ormetabolic hormones themselves, are continuously monitored by the brain. The resultant infor¬mation relating to physiological size and well being is transferred to the GnRH pulse generator.Metabolic signals that are translated as inadequate body size or a poor level of nutrition maintainthe neural oscillator in a relatively slow state so that preovulatory follicles do not develop. How¬ever, in a developing seasonally breeding female, attainment of the appropriate physiological sizeand metabolic well being are not the only requirements for enhancing the frequency of the GnRHpulse generator. There is a seasonal determinant as well. The young female must first conceive inthe autumn or winter so that the 5-month gestation period will yield birth in spring or summer. Theimmature lamb monitors daylength, an annually consistent environmental cue, to determine whenthe predicted season for first conception has arrived. It uses the long days of spring and summer totime puberty in the short days of autumn. The pineal gland is involved in this timekeeping function.Rather than having antigonadal or progonadal determinants, the pineal gland simply transducesphotoperiod cues into hormonal signals that can be used to determine length of day. The nocturnalincrease in secretion of melatonin by the pineal serves to code daylength. The durations of thenightly rises in circulating melatonin are used by some as yet unidentified target tissue(s) as infor¬mation, and cumulation of this information over time provides a 'photoperiod history'. In thisregard, the developing female lamb is exposed first to long-day melatonin patterns during springand summer and then to short-day melatonin patterns during autumn. This sequence of melatoninpatterns provides the necessary 'photoperiod experience' to the normally growing spring-bornlamb for it to decrease its response to oestradiol negative feedback. The attendant increase inactivity of the GnRH pulse generator then initiates the first reproductive cycle in the autumn.

The next three sections of the paper provide evidence to support many aspects of the foregoinghypothesis.

Evidence for sequence of events during puberty

Various lines of evidence have led to the view that the hypothalamic-hypophysial-ovarian systemof the prepubertal lamb is relatively mature and that ovulation does not occur because there isinsufficient tonic LH secretion. The rationale for this hypothesis stems from our earlier studieswhich have been reviewed in detail previously (Foster & Ryan, 1981). The salient features of thisrationale are outlined below. Firstly, high-frequency LH pulses (< 1/h) have been detected in thepostpubertal lamb (Foster, Lemons, Jaffe & Niswender, 1975b) and mature ewe (Karsch, Foster,Bittman & Goodman, 1983) during the follicular phase of the cycle when one or more large folliclesdevelop and oestradiol secretion becomes maximal. By contrast, the frequency of LH pulses in thepubertal lamb is low, and the interval between LH pulses is 2-3 h or greater (Foster et al, 1975b).


Secondly, the ovary is well developed long before puberty (first ovulation, ~ 25-35 weeks of age).Administration of a physiological dose of LH at hourly intervals for 48 h to immature lambs (18-20weeks) replicates the sequence of endocrine events that occur during the follicular phase of the cycleof the adult, including the sustained increase in circulating oestradiol (Foster, Ryan & Papkoff,1984) (Text-fig. 2). This suggests that the ovary is able to develop follicles to an advanced stageand to produce increased levels of oestradiol if provided with an appropriate stimulus (hourlyLH pulses). Thirdly, the preovulatory gonadotrophin surge system is able to function in theprepubertal lamb, but it remains inactive in the absence of sustained high physiological levels ofcirculating oestradiol. This is readily apparent by the induction of gonadotrophin surges inresponse to the endogenous oestradiol produced during treatment with exogenous LH (Text-fig. 2).Furthermore, it has been determined that the system that regulates the surge of gonadotrophin inthe prepubertal lamb (19 weeks) is as sensitive as that of the mature female to the stimulatoryfeedback action of exogenous oestradiol (Foster, 1984).

200

100

50

E£ 10

ÍTT1mr^

1 VÌ°'' \ìos -°-V Ni

0-2' J_L J_L

muOnset ofLH surge

\

J_L

E

Sg1CO

Sb

1

0-5

6 0 10 14 20 24Hours relative to first LH injection

Text-fig. 2. Patterns of circulating LH and oestradiol at various times before and during hourlyinjections (arrows) of purified ovine LH in a prepubertal female sheep, 18 weeks of age. Thedose per injection was 15-5 NIH-LH-S1 equivalents of the preparation Papkoff G3-256DA(2-5 NIH-LH-S1; < 001 NIH-FSH-S1); blood samples were collected every 20 min. HighLH values between Hours 22 and 24 reflect the initial portion of a preovulatory LH surge thatwas followed by an 11-day luteal phase (3 corpora lutea). (Redrawn from Foster et al, 1984.)

The foregoing evidence that initiation of the sequence of events leading to first ovulation in thelamb is dependent on a high frequency of LH pulses raises the question of what role follicle-stimulating hormone (FSH) may play in the timing of puberty. By 10-12 weeks of age, circulatingFSH is within the adult range (Foster et al, 1975b; Fitzgerald & Butler, 1982); this finding, inassociation with the evidence that administration of LH alone can initiate the ovulatory sequence(Text-fig. 2), leads to the conclusion that, against a background of sufficient FSH secretion, LHmay serve as a key timing hormone for puberty. The view that FSH may play a permissive rolein follicular maturation in the adult sheep has been proposed elsewhere (Karsch et al, 1984).


Nevertheless, a subtle increase in circulating FSH has been detected during the pubertal period inblood samples collected every 4 h ( . D. Ryan & D. L. Foster, unpublished). The physiologicalsignificance, if any, of the modest FSH rise during puberty remains to be clarified.

If the increase in frequency of LH pulses is responsible for the initiation of the pubertal fol¬licular phase, then the key to our understanding of sexual maturity is the consideration of thesystem that governs GnRH secretion. Partly on the basis of work on fully mature females, it isassumed that the frequency of LH pulses reflects the activity of the hypothalamic GnRH pulsegenerator. In the ewe, each LH pulse in the peripheral circulation has been shown to be associatedwith a pulse of GnRH in perfusate from the median eminence (Levine, Pau, Ramirez & Jackson,1982) or in hypothalamic-hypophysial portal blood (Clarke & Cummins, 1982). In the rhesusmonkey, LH pulses occur simultaneously with volleys of electrical activity recorded from multipleelectrode arrays chronically implanted in the medial basal hypothalamus (Kesner, Kaufman,Wilson & Knobil, 1984). Although similar relationships between changes in hypothalamic electri¬cal activity, GnRH discharges and LH pulses have not been demonstrated during development,there is no reason to suspect that the general principles of neurosecretory control of LH secretionwould not hold for the prepubertal individual. On the other hand, developmental changes insynaptogenesis and cytoarchitecture may occur to increase activity of the GnRH pulse generatorat puberty. In the hypothalamus of the laboratory rat approaching puberty, synaptogenesisaccelerates (Matsumoto, 1976) and a preliminary report indicates that the proportion of GnRHcells with spine-like processes on the soma and dendrites increases (Wray & Hoffman-Small, 1984).

Bearing in mind the foregoing considerations of hypothalamic control of LH secretion, it couldbe postulated that the immature lamb cannot produce high-frequency GnRH pulses. This is not thecase, because in the absence of the ovaries (ovariectomy at 2 weeks), hourly LH pulses are manifestby 7-9 weeks of age (Foster, Jaffe & Niswender, 1975a). Thus, the potential to produce rapid LHpulses is attained by the lamb early in postnatal life, but this is not expressed in the intact lamb untilpuberty, some 30 weeks after birth. A decrease in sensitivity to oestradiol negative feedback hasbeen used to explain the appearance of rapid LH pulses in the intact animal (Foster & Ryan, 1981,for review). This explanation is based upon the finding that the same physiological blood con¬centrations of exogenous oestradiol that suppress circulating LH to low values in ovariectomizedlambs become ineffective in this regard during the pubertal period (Foster & Ryan, 1979; Foster,1981).

Is there enough evidence for the proposed sequence of events that occurs during the transitioninto adulthood in the female sheep? For the most part, it is felt that there is. However, additionalstudies of pulsatile LH secretion are necessary to address certain issues relating to neuroendocrinechanges that occur during puberty. For example, detailed studies of the patterns of circulating LHduring the first follicular phase should be undertaken to evaluate better the time course of thepostulated increase in LH pulse frequency. Is there a progressive increase in LH pulse frequencyduring the week or so before the first preovulatory gonadotrophin surge? Or are there periods whenLH pulse frequency increases transiently, periods that become progressively longer until a

preovulatory follicle is developed? Such patterns also bear on the question whether the theory ofa decrease in feedback sensitivity to oestradiol and the consequent expression of hourly pulses ofLH are sufficient to explain the initiation of the first follicular phase. Certainly, it has beendemonstrated that exogenous LH administered at 60-min intervals to the prepubertal lamb caninduce a follicular phase culminating in ovulation (Text-fig. 2). However, the frequency of endo¬genous LH pulses during puberty could be even greater. In the fully mature ewe, LH pulse fre¬quency in the follicular phase is nearly 2-fold higher than that in the absence of steroid negativefeedback (i.e. ovariectomized ewe), and the data suggest that oestradiol contributes to the very highfrequency of LH pulses in the normal follicular phase (Karsch et al, 1983). Evidence thatoestradiol may also accelerate LH pulse frequency in the young lamb will be presented in a latersection of this paper. This raises the possibility that oestradiol may be inhibitory to pulsatile LHsecretion before puberty and stimulatory afterwards.


The above considerations suggest that the action of oestradiol on the GnRH pulse generatorchanges during the transition into adulthood. If this proves to be the case, then we must expand theworking hypothesis to account for how a high frequency of LH pulses is manifest during pubertybeyond the simple notion of a decrease in sensitivity to oestradiol negative feedback. We must alsoaccount for how oestradiol increases LH pulse frequency. One explanation may be that theinherent neuroendocrine activity of the GnRH pulse generator increases during puberty and thatthis modifies the feedback action of oestradiol. For example, when bouts of activity occur at60-90-min intervals, oestradiol exerts a potent inhibitory feedback action. However, when therhythm of the GnRH pulse generator increases even further to give a bout of activity at ~45-minintervals, oestradiol then has a stimulatory action. Studies that use level of nutrition to alter thetime of first ovulation have provided some evidence that the ability of oestradiol to inhibit or stimu¬late LH pulse frequency may be related to the inherent activity of the GnRH pulse generator (seebelow). Clearly, we must examine changes in LH pulse frequencies in other model systems ofdevelopment in the lamb to resolve these important issues of the dynamics and mechanismsunderlying the initiation of the pubertal follicular phase.

Evidence that multiple factors time puberty

As indicated at the onset, the time of puberty in most mammals is determined not only by theattainment of an appropriate body size, but by seasonal factors as well. This is illustrated for thelamb by the data presented in Text-fig. 3. In the study shown, Suffolk lambs were born in the spring(March) and raised outdoors under natural conditions. Some lambs were fed ad libitum, and theyovulated at the normal age ( ~ 30 weeks, Group A). Other lambs were undernourished from thetime of weaning (10 weeks of age, Groups B, C, and D), and puberty was delayed relative to well-fed lambs. When undernourished lambs were given unlimited amounts of food during autumn and

Text-fig. 3. Influence of season and growth on age at first ovulation (a, mean ± s.e.m.; b,individuals, histograms). The lambs were born in spring (March) and were raised outdoors(-, photoperiod). They were either fed ad libitum after weaning (10 weeks of age) or wereplaced on a restricted diet of similar composition; at various times (arrows), ad-libitum feedingwas begun in food-restricted lambs. The general limits of the anoestrous and breeding seasonsof the adult are indicated at the top. (Redrawn from Foster & Ryan, 1985.)


early winter, the first ovulation occurred within a few weeks (Groups and C). However, whenad-libitum feeding was delayed until late winter and early spring, the lambs grew well beyond thenormal size for puberty, but ovulation did not occur (Group D). The lambs remained anovulatoryduring the summer, and then in autumn reproductive cycles began despite the absence of furthergrowth. These results may be interpreted against the background of seasonal reproduction in theadult female Suffolk sheep (Text-fig. 3a). It appears that the developing spring-born lamb mustgrow rapidly to achieve the appropriate size for reproduction during the first breeding season

(September-February) (Text-fig. 3, Group A), although the size requirement decreases as thebreeding season progresses (Groups and C). If substantial growth is forestalled until theanoestrous season (March-August), first ovulation will not occur until the beginning of the nextbreeding season when the females are well over the normal age for puberty (Group D).

The foregoing study raises several provocative questions concerning the timing of puberty inthe female lamb. For example, what specific growth factors regulate the initiation of the follicularphase at puberty? What aspect of the natural environment modifies the time of puberty, and is it thesame factor that times the breeding and anoestrous seasons of the adult? How are internal andexternal cues integrated to govern the transition into adulthood in the appropriate season? Answersto many of these questions are beginning to emerge. The next section provides evidence to supportthe hypothesis that the GnRH pulse generator of the developing lamb is sensitive to level of nutri¬tion, and, therefore, to metabolic status. A later section will detail our understanding of howphotoperiod, a synchronizer of breeding activity in the mature sheep, can play an important role inthe timing of puberty in the developing female.

Evidence for timing of puberty by nutrition and growth

The study referred to in the previous section (Text-fig. 3) showed that prolonged undernutritionafter weaning can prevent the initiation of reproductive cycles at the normal age. Perhaps thefrequency of LH pulses is low in such growth-retarded lambs. In a recent investigation (Foster &Olster, 1985), two models were used to determine the effects of nutrition on the ability to generateLH pulses in the absence of steroid feedback (ovariectomized lamb, Text-fig. 4b) and in its presence(oestradiol-treated, ovariectomized lamb, Text-fig. 4c). Ovariectomy of undernourished lambs pro¬duced an initial increase in concentrations of circulating LH, but with continued food restrictionLH secretion was reduced to low levels. This was due to a slow rate of LH discharge. Ad-libitumfeeding produced a rapid weight increase and, in intact lambs, puberty was evident within a fewweeks (Text-fig. 4a); in untreated ovariectomized lambs, LH pulse frequency increased markedly.Therefore, level of nutrition, even in the absence of steroid feedback, can have a strong modulatoryeffect on the activity of the GnRH pulse generator.

The influence of nutrition on LH secretion in the presence of steroids appears to be more com¬

plicated (Text-fig. 4c, Weeks 30-40). First, oestradiol reduces the amplitude of LH pulses regardlessof the level of nutrition. Second, when the ability to produce LH pulses is reduced in the absence ofsteroids during undernutrition, oestradiol eliminates pulsatile LH secretion altogether. Third,during increased nutrition, when the ability to produce LH pulses is enhanced in the absence ofsteroids, oestradiol appears to accelerate LH pulse frequency. In fact, the sampling frequency used(every 12 min) was often insufficient to permit clear definition of the small and rapid LH pulses (seediscussion in paper by Foster & Olster, 1985). The high frequency of LH pulses in this modelappeared when reproductive cycles were initiated in intact lambs during ad-libitum feeding and thismay reflect the development of a fully mature GnRH pulse generator that is capable of producingthe very high frequency of LH pulses that drives the follicular phase (Karsch et al, 1983).

The results obtained from the various experiments considered above suggest that one cause ofthe anovulatory condition of the undernourished lamb is a deficiency in tonic LH secretion. Indeed,a lower LH pulse frequency and amplitude have been reported in lambs on a low energy diet


July. Aug. Sep. Oct. Nov. Dec.50

;r 40

5 30>,- om 20

15

10

I 5

fc10

5

0

(a) -1

-

Ad libitum

«-Restricted-

Ovariectomy(N = 13)

(c) Ovariectomy+ E2(N = 12)

«JL> oo-ooo-o

J_1_

= 6

A = 28weeks

cœamuuaujLMinx>

= 33weeks

C = 37weeks

^ ^

20 30Age (weeks)

œœaxorxorjrxccœcD caxorxœoaxairœcç

20

10

0E

10 "ôi

010

40 4 0Time (hours)

Text-fig. 4. Left, mean ( ± s.e.m.) concentrations of circulating LH in ovariectomized lambs (c)with or (b) without chronic treatment with oestradiol (1-2 pg/ml by Silastic implant). Between10 and 28 weeks of age, level of food was restricted to maintain (a) average body weight of20 kg; after 28 weeks of age some females were fed ad libitum, while others were maintained onfood-restricted diet. Histogram in (a) shows onset of reproductive cycles during ad-libitumfeeding in intact lambs that were initially undernourished (10-28 weeks). Right, detailedpatterns of circulating LH in representative females at 28, 33 and 37 weeks of age. Samples wereobtained every 12 min; open circles indicate undetectable values ( < 0-25 ng/ml). (Redrawnfrom Foster et al, 1985.)

(Fitzgerald, Michel & Butler, 1982). According to the hypothesis being developed, the GnRHpulse generator of undernourished lambs would not have the ability to produce rapid LH pulses.Further, the small quantities of oestradiol from small antral follicles would suppress the LHsecretory system even more. Under this condition, the young malnourished female cannot developthe high-frequency LH pulses required to produce the preovulatory follicle and the sustainedoestradiol rise for induction of the preovulatory gonadotrophin surge.

Some evidence also suggests that the preovulatory gonadotrophin surge system may beimpaired in the growth-retarded lamb. This is based upon the finding that a physiological dose ofexogenous oestradiol produced a very small LH surge in undernourished ovariectomized lambsthat were pretreated with oestradiol for several months (Text-fig. 5b). In the absence of chronicoestradiol pretreatment, an acute challenge with oestradiol can produce a normal LH surge in suchgrowth-retarded lambs (Text-fig. 5a). Perhaps the relatively inactive GnRH pulse generator inmalnourished lambs exposed to small quantities of oestradiol does not permit the pituitary gland tobuild up sufficient LH reserves for release by the positive feedback action of oestradiol.

It is clear from the studies described above that undernutrition and subsequent refeeding can

exert pronounced effects on the pattern and level of LH secretion and can alter the timing ofpuberty. The postulated changes in activity of the GnRH pulse generator are speculative owing to


-4 0 4 -4 0 4Hours relative to LH peak

Text-fig. 5. LH surges induced by oestradiol (5-6 pg/ml, Silastic implant) in 40-week-old,undernourished ovariectomized lambs (body weight, 20 kg). In (a) lambs received no previousoestradiol treatment. In (b) lambs were pretreated with low levels of oestradiol (1-2 pg/ml,Silastic implant) from the time of ovariectomy (20 weeks of age). Text-fig. 4(a) shows weightsand concentrations of LH before 40 weeks of age. (Redrawn from Foster et al, 1985.)

çü

co Ece

ç"ô>>üco Eco

100

50

0100

50

0

100

50

0

Breeding Anoestrus Breedingseason season

(b) Spring-born( = 8)

birth

(c) Autumn-born( = 6)

birth 48-50 weeks

(e) Autumn-born(N = 9)

birth33-37weeks

Oic >.COQ

SI

^ OcL Feb.. Jun. „Oct.^- Dec. Apr. Aug. Dec.

Text-fig. 6. Month and age at first ovulation in (a) natural photoperiod in (b) spring-born(March) and (c) autumn-born (October) lambs and in (d) an artificial, seasonally reversedphotoperiod in (e) autumn-born lambs. The general times of the adult breeding and anoestrousseasons are indicated at the top. First ovulation was based upon the presence of luteal-phaselevels of circulating progesterone. In (c) and (e) the hatched bar provides a reference for the agerange (26-35 weeks) over which ovulation was initiated in control spring-born lambs in (b).(From Foster & Ryan, 1981.)


the absence of measurements of GnRH production rates or electrical activity of the hypothalamus.More problematical is the question of whether or not an approach involving nutritional manipu¬lations is valid for the study of normal development. Certainly, many lambs, as do other mammals,mature under conditions of suboptimal nutrition, and therefore availability of food may indeed bea critical determinant for initiation of puberty (Bronson, 1985). Relatively few hypotheses havebeen developed to explain the link between growth and activity of the reproductive system. Furtherconsiderations of this difficult matter may be found elsewhere (Frisch, 1984; Cameron et al, 1985).

Evidence for the timing of puberty by photoperiod

Attainment of a 'critical' body size is not the sole determinant of puberty in the female sheep. Thiswas evident in the previous section by the anovulatory condition of lambs in which growth to theappropriate size for puberty was delayed by inadequate nutrition until the anoestrous season

(Text-fig. 3, Group D). A similar conclusion can be made when season of birth, instead of level ofnutrition, is altered (Foster, 1981). This is exemplified for lambs born in the autumn (Text-fig. 6c)instead of spring (Text-fig. 6b), the usual season of birth. In natural conditions, autumn-bornlambs attain the appropriate size for puberty during the anoestrous period the following spring andsummer. However, they do not exhibit reproductive cycles at that time but wait for an additional20 weeks until the autumn breeding season before ovulating. This delay in puberty can be almosttotally prevented by rearing autumn-born females in an annually reversed artificial photoperiod(Text-fig. 6d, e).

Does photoperiod alter the timing of puberty in developing lambs much the same as nutritiondoes through its influence on tonic LH secretion? It appears that it generally does, althoughdetailed comparisons between the models used to study nutrition and photoperiod have not yetbeen made. The response to oestradiol inhibitory feedback has been examined in spring-born andautumn-born lambs treated chronically with oestradiol (implanted capsule, ~ 2 pg/ml blood levels)at ovariectomy (Text-fig. 7b). In the autumn-born lamb, the period of high responsiveness tooestradiol negative feedback is prolonged for an additional 20 weeks relative to that in the spring-born lamb. This delayed decrease in response to oestradiol inhibition most probably accounts forthe 20-week delay in initiation of ovulation in the autumn-born female (Text-fig. 6c). The pro¬longed period of hyper-responsiveness to oestradiol does not result from an obvious alteration inability to secrete LH at high levels. Removal of all steroid inhibition produces LH concentrations inthe autumn-born female which are greater than those in the developing spring-born female (Text-fig. 7a). This model appears to be unlike the nutritional model in which the ability to secrete LH,evaluated in the absence of steroids, determined the efficacy of oestradiol inhibitory feedback.However, before firm conclusions can be made, detailed studies of pulsatile LH secretion are

necessary. The lack of change in mean levels of LH in blood samples obtained at infrequent inter¬vals does not preclude the occurrence of reciprocal changes in LH pulse frequency and amplitude(Goodman & Karsch, 1981).

Studies of delayed puberty and its restoration to normal by artificial light in the autumn-bornlamb provide compelling evidence that daylength cues serve to modify the onset of reproduction infemales born in an inappropriate season. What aspect of the natural photoperiod cycle does thewell-nourished female use to synchronize puberty in the autumn? Perhaps the lamb simply moni¬tors daylength and, when it becomes sufficiently short, reproductive cycles begin. This hypothesispredicts that spring-born lambs would not exhibit puberty at the normal age (25-35 weeks; Text-fig. 8a) if they were raised under long days from birth. Indeed, when lambs were maintained con¬

tinuously in an artificial photoperiod of 15L:9D, the onset of reproductive cycles was perturbed(Text-fig. 8b); many lambs did not exhibit luteal function during the first year after birth, and thosethat did had predominantly isolated short luteal phases.


Text-fig. 7. Mean ( ± s.e.m.) concentrations of circulating LH in lambs after ovariectomy(arrows) (a) without or (b) with chronic treatment with oestradiol ( ~ 2 pg/ml, Silasticimplants). The females were born in spring (March) or autumn (October) and were raised in anatural environment. Cross-hatch rectangles identify ages at onset of reproductive cycles inintact lambs. (Redrawn from Foster & Ryan, 1985.)

As another test of the hypothesis that lambs simply require short daylengths for puberty,spring-born females were reared entirely under an artificial photoperiod of 9L:15D. The predictedoutcome of the experiment was that they would exhibit precocious puberty. This was not the case,and, in fact, the initiation of ovulations was delayed in the lambs reared in short days (Text-fig. 9a);most lambs remained anovulatory during the first year, and regular oestrous cycles did not occur

until the second year (data for second year shown by Yellon & Foster, 1985). An alternativehypothesis was then tested, namely that a sequence of photoperiods is involved in the timing ofpuberty. This hypothesis was formulated from the fact that under natural conditions, spring-bornlambs normally experience the increasing daylengths of spring and summer before the decreasingdaylengths of autumn when they begin to ovulate. Lambs were therefore raised in various combi¬nations of long and short days (Yellon & Foster, 1985) and both long days and short days were

necessary. Remarkably, as little as 1 week of long days, at Week 21 of age, was effective in initiatingreproductive cycles at the normal age (Text-fig. 9b). Greater periods of long days presented at

younger ages (e.g. birth-4 weeks) were largely ineffective in producing puberty during the first year(Text-fig. 9c). The failure of long days during very early life to initiate repeated reproductive cyclesunder subsequent short days is intriguing. As will be discussed, perhaps this reflects an immaturityof the system used for photoperiodic time measurement. Nevertheless, the results of the studieswith artificial photoperiod lead to the conclusions that the lamb maintains a 'photoperiod history'and that it uses the long days of spring and summer as a reference to time puberty to the short daysof autumn.


15

co"D

S 9oco

Mar. May Jul. Sep. Nov. Jan. I I I I I-1-1-1-1-1-r

12-/

(a) Natural jPubertal period

photoperiod (N = 3) ,,-*.Mblt/ \ MrVlfWli\l.

2

-

3o

15

12

(b) Artificial long days (N = 6)AnovulatoryAnovulatoryAnovulatory

_

20Age (weeks)

40

Text-fig. 8. (a) Concentrations of circulating progesterone in 3 spring-born lambs raised in anatural photoperiod; open circles denote undetectable concentrations ( < 0-2 ng/ml). Theblocks below each progesterone cycle are used for coding purposes with normal luteal phasecycles represented by large blocks and short luteal-phase cycles represented by small blocks, (b)Schematic progesterone cycles in spring-born lambs raised entirely in artificial long days(15L:9D); each horizontal line shows data for an individual female with the beginning of theline designating the start of the period of blood sample collection (twice weekly). (Redrawnfrom Yellon & Foster, 1985.)

The pineal gland of the adult female sheep, through its production of melatonin, is a necessarylink in the system that uses photoperiod to synchronize breeding to the autumn and winter (Karschet al, 1984, for review). This is also the case for photoperiodic timing of puberty in the lamb.Denervation of the pineal gland by bilateral removal of the superior cervical ganglia (see Text-fig. 1for pathway) early in life prevents puberty occurring at the normal age (Text-fig. 10). The nightlymelatonin rise that normally occurs (Text-fig. 10a, inset) was abolished in ganglionectomized lambs(Text-fig. 10b, inset). More direct evidence of a role for melatonin in the photoperiodic timing ofpuberty has recently been obtained. Ganglionectomized lambs infused nightly with melatonin toreplicate a sequence of short days, long days and short days exhibit puberty at the normal age(data not shown; see Yellon & Foster, 1984a). Pharmacological treatments, such as the continuousadministration of melatonin by Silastic devices inserted subcutaneously or intravaginally tootherwise untreated spring-born female lambs, have produced intriguing results. Puberty was

delayed in lambs kept in a natural photoperiod when melatonin was administered chronically start¬ing at 3-4 weeks of age (Kennaway & Gilmore, 1984) or 8 weeks of age (Nowak & Rodway, 1984);by contrast, puberty was advanced when continuous melatonin treatment was started at 19 weeksof age (Nowak & Rodway, 1984). Perhaps the continuous administration of melatonin is inter¬preted by the lamb as 'long nights' (i.e. short days). If this were the case, then the effects of chronic


CO- k_

O) 12ocok.

OI _

12

(a)

Mar. May Jul. Sep. Nov. Jan1 ! ' ' ! I I I I 1 I—r

¡PubertaTpiriodAnovulatoryAnovulatoryAnovulatoryAnovulatoryAnovulatoryAnovulatoryAnovulatory

_Q

_n_ _

(b) II :

_Eh_ rTlr-rT·

_II_nur

-(c)

9-

Anovulatory _

^J_

J_20 40

Age (weeks)Text-fig. 9. Schematic progesterone cycles in (a) spring-born lambs raised entirely in short days(9L: 15D), (b) raised in short days except for Week 22 when they were exposed to long days(15L:9D) or (c) exposed to 4 weeks of long days and then maintained in short days. Largeblocks designate cycles with normal luteal phases and small blocks cycles with short lutealphases; each horizontal line shows data for an individual female (see Text-fig. 8 for examples).(Redrawn from Yellon & Foster, 1985.)

melatonin treatment early and late in natural photoperiod would be compatible with the previouslydiscussed findings obtained by raising lambs in artificial photoperiods (Text-fig. 9). In this respect,spring-born lambs kept outside would experience several nights of brief melatonin exposure (longdays) for puberty to occur in response to prolonged melatonin (short days). This interpretationassumes that the duration of melatonin exposure codes daylength and that some duration existsbeyond which the female considers the night to be long, and, therefore, the day to be short. Clearly,this 'duration hypothesis' remains to be tested thoroughly in the sheep (see discussions by Rollag,O'Callaghan & Niswender, 1978; Bittman, Dempsey & Karsch, 1983; Bittman & Karsch, 1984).

The foregoing discussion of the role of melatonin in the photoperiodic regulation of the timingof puberty begs the question of when the female lamb becomes sensitive to photoperiod. One indexof this would be the age at which a stable melatonin rhythm develops (low concentrations ofcirculating melatonin during the light phase of the photoperiod, high concentrations during thedark phase). As shown in Text-fig. 11, for lambs raised in a sequence of artificial light treatments,the 16- to 20-week-old female can readily produce melatonin in a pattern that accords with the


Jun. Aug. Oct. Dec.

15

o

&£= 15oCOw

I 12

IntQCt Control (N = 6) sunset sunrise

(a) \ 600 '

"Sr/12.00 24.00Hours

Ganglionectomized (N = 5)(b) \Anovulatory

AnovulatoryAnovulatory

Dec.21

\ •*wu i|H

200 ¡

~^/12.00 24.00Hours

10 20 30 40

Age (weeks)Text-fig. 10. Mean ( ± s.e.m.) concentrations of circulating melatonin during a 24-h period at40 weeks of age in a natural photoperiod in (a) (inset) postpubertal female lambs and in (b)(inset) lambs in which the superior cervical ganglia had been removed at 6 weeks of age. Datafor progesterone cycles in each group of lambs are coded (large blocks, normal luteal phasecycles; small blocks, short luteal-phase cycles); each horizontal line shows data for anindividual female (see Text-fig. 8 for examples). (Redrawn from Foster et al, 1985.)

Text-fig. 11. Mean ( ± s.e.m.) concentrations of circulating melatonin during a 24-h period(insets) in 6 lambs at 16 weeks of age in artificial short days (9L:15D) and again at 20 weeks ofage in artificial long days (15L:9D). (S. M. Yellon & D. L. Foster, unpublished.)

prevailing photoperiod. Other studies in artificial photoperiods indicate that the photoperiod-entrained melatonin rhythm becomes established by about 10 weeks of age (Yellon & Foster,1984a). At younger ages (i.e. 3 or 6 weeks) melatonin is produced, but high levels are often presentduring the light phase, as well as the dark phase of the photoperiod. The inability of the very younglamb to produce a pattern of melatonin appropriate to the prevailing photoperiod may explain why


Text-fig. 12. Mean ( ± s.e.m.) concentrations of circulating melatonin during a 24-h period at16 and 40 weeks of age (insets) in 5 lambs with delayed puberty raised in continuous short days(9L:15D). None of the lambs had exhibited reproductive cycles by 40 weeks of age when the24-h melatonin pattern was determined (see Text-fig. 10a, for time of puberty in naturalphotoperiod and pattern of melatonin at 40 weeks). (Redrawn from Foster et al, 1985.)

exposure to long days before 10 weeks of age is relatively ineffective in inducing puberty undersubsequent short days (see Text-fig. 9c). The photoperiodic signal is presumably not transducedinto the appropriate melatonin signal; whether the target tissue for melatonin is responsive duringthe neonatal period has not been determined.

The development of the melatonin rhythm has also been studied under natural conditions.Rodway, Swift, Nowak, Smith & Padwick (1985) reported that, in spring-born lambs, peakmelatonin concentrations during the night varied inversely with daylength. However, that pre¬liminary communication did not indicate the age when a stable melatonin rhythm was established.Examination of several melatonin patterns obtained from developing lambs in our laboratory hasled to the conclusion that a stable melatonin rhythm develops later in the natural environment(after 20 weeks) than in artificial long days or artificial short days (by 10 weeks) (Yellon & Foster,1982, 1984b). Perhaps the invariant periods of light and dark provided by the artificialphotoperiods act as a stronger entraining agent than do the gradually changing natural daylengths.

It is tempting to speculate that the photoperiodic timing of puberty rests with the pineal glandand its ability to transduce signals concerning the length of day. This reasoning is based on thefindings that abolition of the melatonin rhythm (ganglionectomy, Text-fig. 10b) or production ofan 'inhibitory' rhythm (artificial long days from birth, Text-fig. 8b) delays puberty. This leads tothe prediction that the delayed puberty which occurs in lambs raised entirely in short days is due tothe failure of the pineal gland to translate this photoperiod into an appropriate melatonin rhythm.As shown in Text-fig. 12, this may not be the case. An apparently normal melatonin rhythm thatreflected the prevailing photoperiod was evident at 40 weeks of age in short-day lambs with delayedpuberty; in these anovulatory females, the melatonin pattern was obtained several weeks afterreproductive cycles had begun in lambs raised in natural photoperiod (Text-fig. 10a). Such a find¬ing raises the possibility that a short-day melatonin pattern in the female lamb does not necessarilyensure that normal puberty will occur. Perhaps in lambs reared in short days the target tissue for


Weeks of age (spring-born)

20 30 40 50 0 10Weeks of age (autumn-bom)

Text-fig. 13. Concept of short-day photorefractoriness to explain long-day requirement fortiming of puberty under short days. Models used are spring-born lambs raised (a) entirely inartificial short days (9L:15D), (b) in a sequence of artificial short days, long days (15L:9D, one

week), short days, and (c) in natural photoperiod and (d) autumn-born lambs raised in naturalphotoperiod. ·, indicates birth.

melatonin is unresponsive to the short-day pattern, and, hence, the female is 'photorefractory' toshort days. According to the hypothesis being developed, the function of the long-day melatoninpattern would be to induce responsiveness of the target tissue to the subsequent short-day pattern.

Text-figure 13 presents a hypothetical scheme to illustrate some of the foregoing concepts andis based upon experimental models and observations in natural conditions. We propose that thefemale lamb is born photorefractory to short days. In the absence of any long days, puberty willeventually occur because, beyond some physiological age, the normally growing female spon¬taneously 'breaks' short-day photorefractoriness. An alternative explanation is that she rejectsphotoperiod as a cue for maturation. Thus, lambs reared entirely in short days or entirely in longdays or in the absence of photoperiod cues (ganglionectomized) exhibit puberty, albeit well beyond1 year of age (Text-fig. 13a) (Yellon & Foster, 1984b, 1985). Experimentally, a brief period oflong-day melatonin can break photorefractoriness to short days and allow the recognition of theshort-day melatonin pattern that is necessary to initiate puberty during the first year (Text-fig. 13b).Under natural conditions, it is the long-day melatonin patterns that are generated during springand summer that enable the developing lamb to recognize the short-day patterns of autumn thatsignal for the onset of reproductive cycles (Text-fig. 13c). In the autumn-born lamb, a dual mechan¬ism may exist to postpone puberty to a much greater age than would be the case had the lamb beenborn in spring (Text-fig. 13d): it first experiences short days and then long days, the oppositesequence of that experienced by the spring-born lamb. The autumn-born female lamb therefore


maintains short-day photorefractoriness longer because long days are not experienced until a muchlater age. The first exposure to long days occurs during the pubertal period (25-35 weeks) and thisactively inhibits the initiation of cycles. Eventually, as daylength decreases, the autumn-born lambbegins to ovulate because the long-day inhibition diminishes and short-day photorefractoriness isbroken.

Conclusions

The hypothesis presented at the outset reflects our attempt to identify major systems governing thetempo of sexual maturation in the female sheep. Various lines of evidence support the conceptualunderpinnings of the proposed scheme, but it must be continually modified as new informationaccumulates. Better explanations must be sought to account for the working of the GnRH pulsegenerator and how its frequency is slowed by oestradiol at one stage of development and yetaccelerated at another. Moreover, what neuroanatomical correlates can explain the relatively'mature' preovulatory gonadotrophin surge system during much of the prepubertal period whenthe GnRH pulse generator is undergoing developmental modifications? Our knowledge must begreatly expanded concerning how the GnRH pulse generator is influenced by internal and externalcues. A tremendous gap exists in our understanding of what type of growth-related signals aredetected by the hypothalamus that allow it to modify its activity. Finally, it has become apparentthat the photoperiodic timing of first ovulation is much more complex than we had originallyenvisioned. In this regard, many unanswered questions remain, including those relating to theidentification of the target tissue for melatonin and the neuroanatomical basis for photorefractori¬ness. Clearly, we have only begun to appreciate the intricate interplay between the multiple inputsinto the developing brain that enable the lamb to perceive its physiological body size and to predictthe optimal time of year for initiation of reproduction.

We thank Dr Fred J. Karsch and his fellows and students for their conceptual and technicalassistance in many of the developmental studies; Dr Jane E. Robinson for her constructive scien¬tific criticism of the review; Ms Diane E. Belleba for assistance in preparation of the manuscript;Mr Douglas D. Doop for his expert and conscientious care of the experimental animals; our

colleagues Dr Gordon D. Niswender, Colorado State University, Dr Leo E. Reichert Jr, AlbanyMedical College of Union University, and Dr Harold Papkoff, University of California-SanFrancisco, for the generous supply of antisera and purified ovine LH. The Reproductive Endo¬crinology Program at The University of Michigan provided the setting for invigorating discussionsof experimental design and interpretation; the National Institutes of Child Health and HumanDevelopment, the Ford Foundation and The University of Michigan (Rackham School ofGraduate Studies and Biomedicai Research Council) provided funds for the research.

References

Bittman, E.L. & Karsch, F.J. (1984) Nightly durationof pineal melatonin secretion determines the repro¬ductive response to inhibitory day length in the ewe.Biol. Reprod. 30, 585-593.

Bittman, EX., Dempsey, R.J. & Karsch, F.J. (1983)Pineal melatonin secretion drives the reproductiveresponse to daylength in the ewe. Endocrinology 113,2276-2283.

Bronson, F.H. (1985) Mammalian reproduction: an

ecological perspective. Biol. Reprod. 32, 1-26.

Cameron, J.L., Hansen, P.D., McNeil, T.H., Koerker,D.J., Clifton, D.K., Rogers, . V., Bremner, W. J. &Steiner, R. . (1985) Metabolie cues for the onset ofpuberty in primate species. In Adolescence inFemales, pp. 59-78. Eds C. Flamigni, S. Venturoli &J. Givens. Year Book Medical Publishers, Chicago.

Clarke, I.J. & Cummins, J.T. (1982) The temporal rela¬tionship between gonadotropin releasing hormone(GnRH) and luteinizing hormone (LH) secretion inovariectomized ewes. Endocrinology 111, 1737-1739.


Clayton, R.M. (1985) Role of GnRH in the maturationof pituitary gonadotropin function. J. Reprod. Fert.75,307-315.

Fitzgerald, J. & Butler, W.R. (1982) Seasonal effects andhormonal patterns related to puberty in ewe lambs.Biol. Reprod. 27, 853-863.

Fitzgerald, J., Michel, F. & Butler, W.R. (1982) Growthand sexual maturation in ewes: dietary and seasonaleffects of modulating luteinizing hormone secretionand first ovulation. Biol. Reprod. 27, 864-870.

Foster, D.L. (1981) Mechanisms for delay of first ovula¬tion in lambs born in the wrong season (fall). Biol.Reprod. 25, 85-92.

Foster, D.L. (1984) Preovulatory gonadotropin surgesystem of prepubertal female sheep is exquisitely sen¬sitive to stimulatory feedback action of estradiol.Endocrinology 115, 1186-1189.

Foster, D.L. & Olster, D.H. (1985) Effect of restrictednutrition on puberty in the lamb: patterns of tonicluteinizing hormone (LH) secretion and competencyof the LH surge system. Endocrinology 116, 375-381.

Foster, D.L. & Ryan, K.D. (1979) Endocrine mechanismsgoverning transition into adulthood: a markeddecrease in inhibitory feedback action of estradiol ontonic secretion of luteinizing hormone in the lambduring puberty. Endocrinology 105, 896-904.

Foster, D.L. & Ryan, K.D. (1981) Endocrine mechanismsgoverning transition into adulthood in female sheep.J. Reprod. Fert., Suppl. 30, 75-90.

Foster, D.L. & Ryan, K.D. (1985) Puberty in the lamb:sexual maturation of a seasonal breeder in a changingenvironment. In Control of the Onset of Puberty II,(in press). Eds P. C. Sizonenko, M. L. Aubert &M. M. Grumbach. Williams & Wilkins Co.,Baltimore (in press).

Foster, D.L., Jaffe, R.B. & Niswender, G.D. (1975a)Sequential patterns of circulating LH and FSH infemale sheep during the early postnatal period: effectofgonadectomy. Endocrinology 96, 15-22.

Foster, D.L., Lemons, J.A., Jaffe, R.B. & Niswender,G.D. (1975b) Sequential patterns of circulatingluteinizing hormone and follicle-stimulating hor¬mone in female sheep from early postnatal lifethrough the first estrous cycles. Endocrinology 97,985-994.

Foster, D.L., Ryan, K.D. & Papkoff, H. (1984) Hourlyadministration of luteinizing hormone induces ovula¬tion in prepubertal female sheep. Endocrinology 115,1179-1185.

Foster, D.L., Olster, D.H. & Yellon, S.M. (1985)Neuroendocrine regulation of puberty by nutritionand photoperiod. In Adolescence in Females:Endocrinological Development and Implications on

Reproductive Function, pp. 1-21. Eds C. Flamigni,S. Venturoli & J. Givens. Year Book MedicalPublishers, Chicago.

Frisch, R.E. (1984) Body fat, puberty and fertility. Biol.Rev.59, 161-188.

Goodman, R.L. & Karsch, F.J. (1981) A critique ofthe evidence on the importance of steroid feedbackto seasonal changes in gonadotrophin secretion. J.Reprod. Fert., Suppl. 30, 1-13.

Karsch, F.J., Foster, D.L., Bittman, E.L. & Goodman,R.L. (1983) A role for estradiol in enhancing luteiniz¬ing hormone pulse frequency during the follicular

phase of the estrous cycle of the sheep. Endocrinology113, 1333-1339.

Karsch, F.J., Bittman, E.L., Foster, D.L., Goodman,R.L., Legan, S.L. & Robinson, J.E. (1984) Neuro¬endocrine basis of seasonal reproduction. RecentProgr. Horm. Res. 40, 185-232.

Kennaway, D.J. & Gilmore, T.A. (1984) Effects ofmelatonin implants in ewe lambs. J. Reprod. Fert. 70,39^45.

Kesner, J.S., Kaufman, J.-M., Wilson, R.C. & Knobil, E.(1984) Circhoral activation of the hypothalamicluteinizing hormone (LHRH) releasing system is notthe consequence of an LH feedback loop. Biol.Reprod. 30, Suppl. 1, p. 32, Abstr.

Levine, J.E., Pau, K.-Y., Ramirez, V.D. & Jackson, G.L.(1982) Simultaneous measurement of luteinizinghormone-releasing hormone and luteinizing hor¬mone release in unanesthetized, ovariectomizedsheep. Endocrinology 111, 1449-1455.

Matsumoto, A. (1976) Developmental changes in synap-tic formation in the hypothalamic arcuate nucleus offemale rats. Cell Tiss. Res. 169, 143-156.

Norman, R.L. (1983) Neuroendocrine Aspects of Repro¬duction. Academic Press, New York.

Nowak, R. & Rodway, R.G. (1984) Advancement of thebreeding season of ewes and lambs by intravaginalapplication of melatonin. J. Steroid Biochem. 20,1461, Abstr.

Ojeda, S.R. (1980) Recent advances in the endocrinologyof puberty. Endocrine Reviews 1, 228-257.

Rodway, R.G., Swift, A.D., Nowak, R., Smith, J.A. &Padwick, J.A. (1985) Plasma concentrations ofmelatonin and the onset of puberty in the femalelamb. Anim. Reprod. Sci. (in press).

RoUag, M.D., O'CaUaghan, P.L. & Niswender, G.D.(1978) Serum melatonin concentrations during differ¬ent stages of the annual reproductive cycle in ewes.Biol. Reprod. 18, 279-285.

Tanner, J.M. (1978) Foetus into Man, Physical Growthfrom Conception to Maturity. Harvard UniversityPress, Cambridge, Massachusetts.

Wray, S. & Hoffman-Small, G.E. (1984) Postnatalmaturation of the LHRH system. J. Steroid Biochem.20, 1420, Abstr.

Yellon, S.M. & Clayton, J.A. (1983) Evidence that thepineal times puberty in the lamb. Biol. Reprod. 28,Suppl. 1, p. 46, Abstr.

Yellon, S.M. & Foster, D.L. (1982) Photoperiod-mediated melatonin patterns in the lamb before andafter puberty and in delayed puberty. Endocrinology110, Suppl. 1, p. 117, Abstr.

Yellon, S.M. & Foster, D.L. (1984a) Melatonin rhythmsmediate photoperiodic timing of puberty in thefemale lamb. Biol. Reprod. 30, Suppl. 1, p. 107, Abstr.

Yellon, S.M. & Foster, D.L. (1984b) Maturation ofmelatonin rhythms in female lambs under artificialand natural photoperiods. Soc. Neurosci. Abstr. 10,820, Abstr.

Yellon, S.M. & Foster, D.L. (1985) Alternatephotoperiods time puberty in the female lamb.Endocrinology 116, 2090-2097.

Received 4 April 1985


Documents

Internal and external timing puberty