Evolutionary and Functional Aspects of Pituitary Gonadotropins in the Green Turtle, Chelonia Mydas 1

Transcription

1 AMER. ZOOL., 20: (1980) Evolutionary and Functional Aspects of Pituitary Gonadotropins in the Green Turtle, Chelonia Mydas 1 PAUL LICHT Department of Zoology, University of California, Berkeley, California SYNOPSIS. Information on the pituitary gonadotropins in Chelonia mydas represents some of the most complete data available for any reptile and thus provides an important basis for evaluating evolutionary processes in tetrapod endocrine physiology. The two gonadotropins isolated from Chelonia pituitary glands show clear chemical and immunological homologies to mammalian follicle-stimulating hormone (FSH) and luteinizing hormone (LH). However, receptor studies and biological tests indicate that the functions of these hormones may be different in turtles and mammals. In particular, Chelonia LH shows an unusual ability to interact with FSH receptor sites and to stimulate physiological functions normally attributed to FSH. Results with Chelonia LH demonstrate that errors may arise from using mammalian hormones to investigate reproduction in turtles. Measurements of endogenous gonadotropin levels in the plasma of breeding and nesting Chelonia provide a different perspective of the potential roles of the FSH and LH from that obtained in physiological tests. In particular, FSH and LH secretion are markedly dissociated during the nesting cycle; FSH has only a transient peak during oviposition, whereas LH, along with progesterone, displays a pronounced "ovulatory" surge in the day following nesting. Preliminary studies with synthetic gonadotropic releasing factors in Chelonia suggest that these may be useful in inducing reproductive changes. INTRODUCTION The hormones of the green sea turtle, Chelonia mydas, figure prominently in the development of evolutionary concepts in reproductive endocrinology, since data on the biochemistry and physiology of pituitary gonadotropins in this turtle are among the most extensive available for any reptile. This situation arose fortuitously when relatively large quantities of pituitary glands from the commercial slaughter of sea turtles became accessible for some of the first studies on gonadotropins from non-mammalian tetrapods (Licht et al., 1976). The approach taken here will be to examine the characteristics of the C. mydas gonadotropins in the context of hormonal evolution and then in terms of the reproductive physiology of the turtle itself. BIOCHEMICAL CHARACTERISTICS OF PITUITARY GONADOTROPINS Basic methodology and results of fractionation studies on the pituitary gonadotropins of tetrapods have been reviewed in 1 From the Symposium on Behavioral and Reproductive Biology of Sea Turtles presented at the Annual Meeting of the American Society of Zoologists, December 1979, at Tampa, Florida. 565 detail by Licht et al. (1977c); species for which comparative data are now available include representatives of several orders of mammals, reptiles, amphibians and birds. It was originally proposed that reptiles might differ from mammals in possessing only a single type of gonadotropic hormone. However, fractionation studies established that the existence of a dual gonadotropin system is a common feature of all classes of tetrapods and, hence, presumably represents a primitive tetrapod condition. Only the reptilian order Squamata (snakes and lizards) appears to deviate from this pattern; snakes probably have only a single pituitary gonadotropin that lacks clear homologies to either of the two molecules found in other orders (Licht etal., 1979a). Having established that most tetrapods, including C. mydas, possessed at least two chemically distinct gonadotropic hormones, it was necessary to ascertain their chemical nature and the degree of homology among hormones from different species. Using the mammalian model and associated terminology as a reference, the extent to which the hormones of C. mydas resemble mammalian folliclestimulating hormone (FSH) and luteiniz-

2 566 PAUL LICHT ing hormone (LH) in structure and function had to be determined. Identification of gonadotropins depended on a combination of biochemical, immunochemical and biological properties; however, for heuristic purposes, each of these facets will be considered separately. Biochemical composition Certain biochemical attributes of a hormone may be inferred from its behavior during fractionation (see Licht et al., 1977c for review). The use of standardized fractionation procedures for diverse species has revealed a strong resemblance in the two major types of gonadotropic molecules isolated from each class. For example, one gonadotropic material (designated FSHlike on the basis of mammalian hormones) tends to be unadsorbed on Amberlite CG- 50 and Sulphoethyl Sephadex C-50 and is adsorbed on DEAE-cellulose. Under the same conditions, a second gonadotropic fraction (LH-like) shows the opposite behavior in each of the three chromatographic systems. This difference allows the rather complete separation of the total gonadotropic activity into two discrete fractions, each being relatively free of cross-contamination with the other. The two types of hormones have similar characteristics in gel filtration on Sephadex G-100 suggesting that they are all of comparable molecular weight. Pronounced differences in relative mobility (R f ) on polyacrylamide disc gel electrophoresis between the two gonadotropins are also seen in each species. In C. mydas, as in mammals, the LH-like fraction has a very retarded R f, while the FSH-like material shows the same relatively high R f as do typical mammalian FSH preparations. Thus, even by this rather crude biochemical characterization, the turtle pituitary seems to possess a pair of gonadotropins that are comparable in structure to the FSH and LH of mammals (and other tetrapods). Amino acid and carbohydrate composition of the two isolated gonadotropic fractions also support the existence of structural homologies among tetrapod classes. The two fractions purified from the green turtle have distinctive amino acid composition, and each is typical of that expected for FSH or LH based on comparative studies (see Licht et al., 1977c for details). Differences in gross composition between the Chelonia and mammalian hormones are no greater than are found among different species of mammals; differences between each of the green turtle and snapping turtle hormones are as great as between those from a turtle and mammal. Analyses of carbohydrate composition yield important insights into the physiological activities of the gonadotropins. Carbohydrate contents of Chelonia hormones fall within the range of values observed for other species of gonadotropins. However, while FSH typically has a higher amount of carbohydrate than LH, this situation is reversed in the case of Chelonia hormones. Particularly salient is the high sialic acid content of Chelonia LH, since this moiety can affect biological potency by influence on clearance rates (see Licht and Papkoff, 1974 for data on turtle hormones). Subunit nature Chemical homologies between the Chelonia and mammalian LH molecules are further evidenced from the demonstration that both share a common subunit structure (Licht et al., 1978). The subunit composition of mammalian glycoprotein hormones (gonadotropins and thyrotropin) is a fundamental feature of their structurefunction relationships. Previous studies demonstrated that LH from the snapping turtle (Chelydra serpentina) also possesses a subunit nature and that these subunits are sufficiently like those of ovine LH to allow recombination between species (Papkoff et al., 1976). Further investigations with subunits prepared from the LH of the green turtle helped to clarify the structural basis for some of the unique biological properties of the turtle and mammalian hormones (Licht et al., 1978). The important point in the present biochemical context is that the amino acid composition of the two subunits of Chelonia LH (designated a and fi) correspond closely to those of their mammalian counterparts. As in mammalian hormones, the individual turtle subunits are by themselves essentially inert

3 biologically, and their recombination generates the full biological potency of the molecule. Moreover, the formation of biologically potent "hybrid" molecules by combination of turtle LHa with mammalian LH/3 and vice versa argues for considerable structural homology between the two species of LH. The behavior of these hybrids is discussed further under biological properties. Immunochemistry Immunochemical properties of the two green turtle gonadotropins again substantiate their distinctiveness from one another, provide an assessment of their degree of purification, and support their tentative identification as structural homologues of FSH and LH. Two basic types of immunochemical approaches have been of value. One involves the behavior of Chelonia hormones in tests with heterologous antisera; in particular, an antiserum raised against ovine FSH and used with ( 125 I)hFSH as radioligand for RIA (Licht and Bona- Gallo, 1978). Immunoneutralization tests confirm that this antiserum crossreacts with the biologically active form of Chelonia FSH, and RIA is used to quantify the crossreactivity. The Chelonia FSH is only slightly less active than mammalian FSH in this mammalian RIA system, whereas Chelonia LH is essentially inactive; the crossreactivity of Chelonia LH indicates a maximum contamination of only ca. 0.2% FSH. The Chelonia FSH is about 70-fold more potent than the crude pituitary extract. A second line of immunochemical evidence comes from the use of antisera raised in rabbits against the purified Chelonia hormones and their subunits (Licht et ai, 19776). RIA studies with these homologous antisera underscore the distinctiveness of the FSH and LH: Crosscontamination is less than a few percent; and each is considerably more potent than crude pituitary extracts. The antiserum to Chelonia FSH has only limited crossreactivity with heterologous species in RIA, but immunoneutralization or direct binding studies substantiate its relatedness to the FSH from mammals (Licht, 1978). The GONADOTROPINS IN GREEN TURTLES 567 Chelonia-LH RIA crossreacts with a greater diversity of species: All chelonian, crocodilian and avian LH preparations crossreacted to a high degree; mammalian, especially marsupial, LH had limited crossreactivity; and gonadotropins from squamate reptiles and amphibians were inactive. The broad phylogenetic immunorelatedness of the LH preparations is a function primarily of the a-subunit; RIA based on the Chelonia LH /3-subunit crossreacts completely within LHs from all turtles but little with nonchelonian hormones (Licht, 1978). This situation contrasts with findings for mammalian glycoproteins in which the immunological determinants underlying crossreactivity among hormones appear to reside largely in the j8- subunit (see review in Licht et al., 1917b). BIOLOGICAL PROPERTIES Since there appears to have been sufficient conservatism in the biochemistry of gonadotropins to allow identification of structural counterparts in turtles and mammals, one might also expect some equivalence in the physiological roles of the two hormones in both classes. However, available comparative data indicate significant divergence in the physiology of gonadotropins among tetrapods. Several issues must be considered in approaching the study of gonadotropic function. It is important to determine the actions of gonadotropins in homologous species as well as to ascertain the extent to which their actions in heterologous species are indicative of normal functions. Since there is little direct information on the actions of Chelonia hormones in Chelonia itself, inferences must be made from the responses that they elicit in other species. From the standpoint of the practical application of gonadotropin therapy for regulating reproduction in the green turtle, one must resolve whether mammalian (or any other heterologous) gonadotropins can be used to control chelonian reproduction. The potential problems in the use of heterologous hormones are generally described by the concept of "species specificity" and may relate to either the quantitative or

4 568 PAUL LICHT v s» : Binding of IZi l-che/on/a FSH to Chelonia ovary p-fsh N^hFSH *,,, 1.,.,1,, X LH - 1. Ng unlabelled hormone FIG. 1. Competitive activities of human (h) and Chelonia (Ch) FSH and LH in inhibiting binding of radioiodinated Chelonia FSH to Chelonia ovarian tissues. Reproduced from Licht et al. (1977i). qualitative actions of the hormones (see Licht and Papkoff, 1978). Hormone-receptor interactions The only direct information on the physiological responses of Chelonia to its own hormones comes from data for gonadotropic binding sites on testes and ovaries. The interaction of the gonadotropins with these binding sites (often referred to as receptors) presumably represents the first step in the mediation of biological action by the hormones. Thus, the characteristics of these receptors should provide some insight into the specificity of actions of the two gonadotropins. The evidence for interactions between hormones and gonadal receptors in Chelonia is in striking contrast to the model that has been developed for the gonadotropic receptors in mammals. In mammals, the gonads appear to have two distinctive sets of receptor sites that differ in their tissue localization within the gonad and in their specificity for FSH or LH in accordance with established differences in the sites and nature of physiological responses induced by the two gonadotropins. Such a dual receptor system does not seem to exist in Chelonia testes or ovaries. Initial receptor studies in Chelonia (and other turtles) were based on the use of a radiolabeled human FSH, ( 125 I)hFSH, as ligand to identify and characterize receptors (Licht and Midgley, 1976). When H p N N mammalian hormones are tested in competition with this ligand, the Chelonia testis and ovary (like those of several other turtles tested) have the expected receptor specificity: Only FSH, and not LH, compete for the binding of the ( 125 I)hFSH. However, profoundly different results are obtained when other species of gonadotropins, especially those from Chelonia, are tested. In brief, Chelonia LH is essentially equipotent to or even more potent than FSH in competing for the presumptive FSH-receptor sites; in other words, the receptor sites on the Chelonia (and other turtle) gonads fail to discriminate between Chelonia FSH and LH. This situation has since been observed with other species of receptor and LH, but without clear phylogenetic trends. For example, LH from the snapping turtle, Chelydra serpentina, does not possess this ability to compete for binding of ( 125 I)hFSH. It is also noteworthy that mammalian "FSH" receptors cannot discriminate between Chelonia FSH and LH (Licht and Midgley, 1976). We considered the likelihood that these results with ( 125 I)hFSH in turtles might represent an "artifact" caused by the anomalous behavior of a heterologous hormone; i.e, perhaps the hfsh bound to both the FSH- and LH-receptors in the turtle. Thus, it was with special interest that we undertook direct binding studies with radiolabeled Chelonia FSH; these represent the first such direct binding studies of a nonmammalian hormone (Licht et al., 1977a). Surprisingly, the Chelonia gonadal receptor sites continued to exhibit relatively little FSH/LH specificity, even with the homologous FSH as radioligand (Fig. 1). The potency of Chelonia LH in competing for the binding of ( 1K l)chelonia FSH to receptors in Chelonia is about 30% that of FSH; and in other turtle receptors, the Chelonia LH is 300% as active as FSH; results are virtually identical to those obtained in assays with ( 125 I)hFSH. Unfortunately, parallel radioreceptor tests to identify LH binding sites directly could not be undertaken due to the inability to demonstrate binding of radiolabeled mammalian or Chelonia LH to any turtle (or nonmammalian) gonadal tissues.

5 However, we did recently succeed in obtaining binding of ( 125 I)turkey LH to the Chelonia ovary (Bona-Gallo and Licht, 1979), and competitive inhibition studies indicate the same lack of gonadotropic specificity as is observed with FSH as radioligand: Chelonia FSH and LH are essentially equipotent in competing for the binding of ( t25 I)turkey LH. Thus, all data point to the existence of a single class of gonadotropin receptors in the Chelonia gonad. Bioassays Although Chelonia hormones have not yet been tested for biological activities in Chelonia, some insight into their potential actions may be gained from studies of their effects in other species. It is clear from such comparative assays that the structure of Chelonia hormones must differ in important respects from mammalian hormones. When tested in mammalian bioassays, the potencies of Chelonia hormones (especially LH) may be several orders of magnitude lower than those of purified mammalian hormones (e.g., Licht et al., 1977c). This lack of activity is probably related to the low affinity for gonadal binding sites (Licht and Midgley, 1976). The demonstration that Chelonia LH may be as active or even more potent than mammalian hormones in several, nonmammalian bioassays also argues that the species specificity in mammalian assays has a structural basis rather than being due to differences in hormone purities. There is also evidence of significant species specificity in the response of turtles to mammalian hormones; and these data are directly relevant to the issue of whether mammalian hormones should be used in these reptiles. The most conspicuous example of such specificity concerns the regulation of steroid secretion by chelonian testes and ovaries. Tests of mammalian gonadotropins in assorted turtles led to the conclusion that the secretion of testosterone by testes and of progesterone by ovaries (e.g., Fig. 2) are primarily FSH-dependent (reviewed by Licht et al., 1977c; see also Lance and Callard, 1978). Owens et al. (1978) used FSH- GONADOTROPINS IN GREEN TURTLES JS Dose (>jg/cc) FIG. 2. Stimulation by ovine (NIH) or Chelonia mydas (Cm.) gonadotropins of in vitro progesterone secretion by preovulatory follicular tissues taken from the snapping turtle, Chelydra serpentina. Modified from a figure in Licht and Crews (1976). induced testicular androgen secretion as a technique for sex determination in young C. mydas. Thus, the turtles would appear in opposition to mammals (and some other tetrapods) in which these aspects of steroid secretion are highly LH specific. While there is no doubt that turtles differ from mammals insofar as FSH is highly potent in stimulating steroid secretion, the apparent lack of response to LH is probably an "artifact" of species specificity toward mammalian LH. Steroid secretion from the turtle testes and ovaries clearly responds to LH preparations derived from various nonmammalian, and especially reptilian, species (Crews and Licht, 1975; Licht and Crews, 1976; Tsui and Licht, 1977); in fact, even some mammalian (e.g., equine) LHs may be potent (Licht, unpublished). Moreover, the LH from Chelonia is among the most potent of the gonadotropins tested in turtle steroid assay systems; Chelonia LH is considerably more potent than Chelonia FSH (Fig. 2). Purified gonadotropins from another turtle, Chelydra serpentina, are more nearly equipotent (Licht and Crews, 1976; Tsui and Licht, 1977). Studies with the subunits of the mammalian and Chelonia LH molecules eluci-

6 570 PAUL LICHT date the structural basis for the species specificity noted above. When hybrid molecules are formed between the two species of subunits, the activity of the molecule is determined largely by the nature of the /3-subunit (Licht et al., 1978). Hence, the combination containing ovine LH-/3 (with Chelonia LHa) is active in mammals but not turtles, while the reverse is true for the hybrid containing Chelonia LH-/3 (and ovine LH-a). However, the regeneration of activity is lower for the former combination, indicating that the Chelonia LH-a may not be able to substitute fully for the ovine LH a-subunit. Biological tests of Chelonia hormones in several nonmammalian species also indicate that the Chelonia LH may have some unusual properties that are not shared by other nonmammalian or even turtle species. In general, the Chelonia hormones are found to be relatively potent compared to other species when tested in several standard assays (Licht et al., 1977a, c). Chelonia FSH has the expected biological profile: It is active only in assays known to be sensitive to FSH or "total" gonadotropin and inactive in tests such as amphibian ovulation that are considered LH specific. However, the biological profile of Chelonia LH is less consistent. It has the expected activity in standard LH specific assays (e.g., amphibian ovulation), but it also shows unexpected activity in other types of bioassays. For example, previous tests suggested that the Anolis lizard bioassay, involving testis growth and androgen secretion, is primarily an FSH dependent system (based on studies with mammalian, amphibian and even turtle gonadotropins). However, Chelonia LH is almost as potent as the FSH in this system (a phenomenon later observed with several other species of hormones; e.g., from alligator and turkey). The basis for the unusually high potency of Chelonia LH in this assay is probably, in part, its high sialic acid content which leads to a significantly increased half-life in the lizard (Daniels et al., 1979). These results imply that the apparent FSH specificity of the lizard bioassay may not entirely be due to receptor specificity. Physiologial studies involving tests of Chelonia hormones in heterologous species, as well as limited data for the action of mammalian hormones in Chelonia (Owens, 1976; Owens et al., 1978), lend support to the general hypothesis developed for other reptilian systems that suggests that the reproductive endocrinology of these vertebrates differs significantly from the classical mammalian scheme. Despite the demonstration of two distinctive gonadotropins in the turtle pituitary, there is little evidence to support the existence of a dual gonadotropic system from a functional point of view; i.e., both LH and FSH seem to have similar functions. Both hormones appear to share a common receptor site and to elicit comparable physiological responses when injected into reptiles. The only evidence for distinctive actions comes from a few assays with nonreptilian species {e.g., amphibian ovulation). These findings have important implications for hormone therapy since they indicate that the turtle system may be less complex than that of mammals; i.e., one hormone could be used to accomplish full reproductive stimulation. However, new information related to endogenous patterns of circulating FSH and LH in Chelonia now raise serious doubts about this conclusion. PROFILES OF CIRCULATING GONADOTROPINS Seasonal patterns The availability of purified Chelonia hormones made possible the development of the first homologous radioimmunoassay systems for measuring plasma levels of FSH and LH in a reptile (Licht et al., 19776). These RIAs have been used to study plasma gonadotropins in relation to gonadal sex steroids during the breedingnesting cycles of both captive (Licht et al., 19796) and wild (Licht et al., 1980) populations. Since these data have recently been published in detail, only certain features that are salient to gonadotropin physiology will be emphasized here. It is noteworthy that comparable results were obtained for the captive breeding population at the Cayman Turtle Farm and for three wild nesting populations (on Aves Island, Surinam and Mexico).

7 GONADOTROPINS IN GREEN TURTLES 571 Studies in males were unsuccessful in regard to plasma gonadotropins; plasma FSH and LH remain at undetectable levels in all samples. A somewhat surprising finding is that plasma testosterone in samples taken during the prebreeding season, several months before the onset of copulation, is twice as high as in copulating individuals. Females have more dramatic changes in gonadotropins than males. Lance et al. (1979) reported that immuno-lh (measured by a heterologous avian LH RIA) is slightly higher in the plasma of nesting Chelonia than in internesting animals (about 7 days later). Plasma progesterone (PRO) is also elevated (2-3 fold) in nesting samples, whereas estradiol (E) is depressed and testosterone (T) is unchanged. Our measurements on these same blood samples confirm the results for LH (steroids were not measured) (Licht et al., 19796). Additional measurements on a larger series of nesting and internesting turtles confirmed further the difference in plasma LH but did not show the expected change in T or E. Moreover, data obtained from this broader series of samples for the entire nesting cycle provide a somewhat different perspective for evaluating changes in both gonadotropins and steroids (Licht et al., 19796, 1980). For example, plasma LH and PRO begin to rise in females at about the time of mating, well in advance of the nesting season. Surprisingly, plasma E peaks before the estrous period and T is highest during copulation. Seasonal profiles of the three ovarian steroids in Chelonia contrast in several respects with those observed in two other turtles examined, Chrysemys picta (Callard et al., 1978) and Chelydra serpentina (Lewis et al., 1979). While plasma LH and PRO may be slightly elevated in nesting Chelonia, the values at the time of oviposition prove to be relatively inconsistent, and differences between samples taken at the time of nesting and at the midpoint of the internesting period are relatively small compared to the dramatic changes observed immediately after nesting. In animals destined to renest in the next days, plasma LH and PRO rise sharply beginning about 6-9 hr Do,s, pos'-ovipos ' cp Fic. 3. Plasma gonadotropins (FSH and LH) and progesterone (PRO) during the breeding season and internesting cycle of two female Chelonia mydas. The continuous curves represent sequential samples taken once or twice daily starting at the time of oviposition (0 days) up until a second prebreeding season (P), during mating (M), or during an earlier nesting (N) are shown on the left. Reproduced from Licht et al. (19796). after nesting and persisting for about hr (Fig. 3). During this brief period, titers of both hormones increase to fold baseline levels, and they attain higher values than recorded in any reptile. Although the exact timing of ovulation is unknown for C. mydas, a variety of evidence suggests that these changes in LH and PRO represent the "ovulatory surge" that is responsible for ovulation and luteinization (see also discussion by Owens in this issue). We have observed comparable hormonal surges in mating animals, several weeks before the first nest of the season; the rise in LH and PRO is absent after the last nest of the season. From a practical standpoint, it should be mentioned that beached females that are turned on their backs before they had oviposited fail to show these post-nesting hormone surges, but females turned after laying is completed have the expected changes in plasma hormones. Thus, this ovulatory surge is apparently blocked by the presence of oviducal eggs, and handling per se probably has little effect (Licht et al., 1980). Profiles of plasma FSH during the

8 572 PAUL LICHT e'- S 8 Q (131 kg) cf 722 (109 kg) ICO Time After Inaction of LH/FSH-RH FIG. 4. Plasma immunoreactive LH following a single intravascular injection of 1 mg gonadotropin releasing hormone (FSH/LH-RH) into young adult Che- Ionia mydas; sex and body weight are as indicated. Samples were taken immediately before injection and at 10, 30, 60 and 120 min thereafter. Based on unpublished data from Woods and Licht. breeding and nesting season contrast markedly with those described for LH. The most conspicuous difference occurs during the nesting cycle. FSH levels are generally low and difficult to measure except at the time of oviposition. Females sampled in the morning before nesting, emerging from the water to nest or even during the early stages of nest digging still have low plasma FSH titers (Licht et al, 1980). FSH rises abruptly with the onset of oviposition and disappears rapidly, with a half-life of about 15 min, following the completion of oviposition. FSH then remains low throughout the rest of the internesting period (Fig. 3). This transient FSH surge occurs whether or not a subsequent LH/PRO surge is destined to occur; i.e., even at the time of the last nest of the season (Licht et al., 1979ft). The pronounced synchrony between the peaks of FSH and LH during the nesting cycle has important implications. First, the conspicuous dissociation between the timing of the FSH and LH peaks suggests the existence of independent mechanisms for regulating the pituitary release of the two hormones. This hypothesis is interesting in light of the large body of information pointing to the existence in mammals of a single hypothalamic gonadotropin releasing factor (FSH/LH-RH or Gn-RH). Second, the pronounced differences in the temporal relationship among the two gonadotropins and other features of ovarian activity, notably PRO secretion and ovulation, tend to contradict the physiological studies in reptiles. The evidence from endogenous secretory patterns suggests that even though the two hormones may have the same potential activities when administered exogenously, they may play distinctive roles in vivo; it seems likely that LH is primarily responsible for ovulation and related increases in PRO secretion. Thus, the nature of ovarian regulation in this turtle may be more like that of the mammal than previously thought. The extent to which these findings can be extrapolated to other turtles is unclear in view of the differences in biological profiles between LH from Chelonia and other species (see discussion of bioassays and receptor assays). The role of the transient FSH surges during oviposition can only be speculated upon at this time. One possibility suggested by mammalian studies is that the FSH may act to "prime" the ovary for the subsequent actions of LH. In any case, these data demand that any attempt to control reproduction in Chelonia through hormone therapy must consider that a complex series of sequential hormone injections may be required. Clearly, more information is now needed on the mechanisms controlling the onset and termination of these surges in the two gonadotropins and how they affect details of ovarian function. Gonadotropin releasing hormone (Gn-RH) The RIA systems developed for the Chelonia gonadotropins also allow assessment of the potential for employing synthetic releasing hormone to control reproductive activity an idea of special interest in view

9 of the evidence for independent regulation of FSH and LH in Chelonia. Moreover, such hormone therapy has recently proven highly valuable for the artificial control of reproduction in species for which supplies of pituitary gonadotropin are unavailable; e.g., Gn-RH has become an important element in fish culture. Only limited data exist for the action of Gn-RH in a reptile; some increase in immunoreactive plasma LH and progesterone was reported for the painted turtle (Callard and Lance, 1977). Initial attempts to demonstrate an action of Gn-RH in Chelonia have met with limited but promising success. In the first test, six young adult males (weighing about 115 kg) were injected with 100 fig Gn-RH (three received a second injection of 50 /xg, 30 min later). No change was observed in either gonadotropin although in four animals plasma testosterone increased slightly (1.5-2 fold) after an hour (Owens and Licht, unpublished). In a second experiment, three animals injected intravascularly with a much higher dose (1 mg) of Gn-RH had a significant elevation in plasma LH within 10 min (Fig. 4); the rise was greatest in the two females and was related to the initial plasma levels (from Woods and Licht, unpublished). The male showed about a 3-fold rise in plasma testosterone between 60 and 120 min, and one female exhibited a rise (3-fold) in plasma PRO. This dosage of Gn-RH is relatively large compared to that required to elicit responses in some large mammals. It has been our experience that reptiles (including other turtles and snakes and lizards) are relatively insensitive to the synthetic releasing factor, and it may well be that species specificity is a problem here. It will, thus, be of interest to test some of the other more potent analogues of Gn-RH that are now available, while giving more attention to the reproductive status of the animal; i.e., sex, age and season. ACKNOWLEDGMENTS The preparation of this manuscript and collection of unpublished data was aided by a grant (PCM ) from the National Science Foundation. GONADOTROPINS IN GREEN TURTLES 573 REFERENCES Bona-Gallo, A. and P. Licht Differences in the properties of FSH and LH binding sites in the avian gonad revealed by homologous radioligands. Gen. Comp. Endocrinol. 37: Callard, I. P. and V. Lance The control of reptilian follicular cycles. In J. H. Calaby and C. H. Tyndale-Biscoe (eds.), Reproduction and evolution, pp Australian Acad. Sci, Canberra City. Callard, 1. P., V. Lance, A. R. Salhanick, and D. Barad The annual ovarian cycle of Chrysemys picta: Correlated changes in plasma steroids and parameters of vitellogenesis. Gen. Comp. Endocrinol. 35: Crews, D. and P. Licht Stimulation of in vitro steroid production in turtle ovarian tissue by reptilian, amphibian, and mammalian gonadotropins. Gen. Comp. Endocrinol. 27: Daniels, E. L., P. Licht, and A. Bona-Gallo Relation between biological potency and clearance rates of gonadotropins in the lizard Anolis carolinensis. Gen. Comp. Endocrinol. 38: Lance, V. and I. P. Callard Steroidogenesis by enzyme-dispersed turtle (Chrysemyspicta) ovarian cells in response to ovine gonadotropins (FSH and LH). Gen. Comp. Endocrinol. 34: Lance, V., D. W. Owens, and I. P. Callard Radioimmunoassay of plasma progesterone, testosterone, total estrogens and immunoreactive gonadotropin in the nesting and non-nesting green sea turtle, Chelonia mydas (L.). Experientia 35: Lewis, J., I. Y. Mahmoud, and J. Klicka Seasonal fluctuating of plasma progesterone and oestradiol-17/3 in the female snapping turtle, Chelydra serpentina. J. Endocrinol. 80: Licht, P Studies on the immunochemical relatedness among tetrapod gonadotropins and their subunits with antisera to sea turtle hormones. Gen. Comp. Endocrinol. 36: Licht, P. and A. Bona-Gallo Immunochemical relatedness among pituitary follicle-stimulating hormones of tetrapod vertebrates. Gen. Comp. Endocrinol. 36: Licht, P., A. Bona-Gallo, and E. L. Daniels. 1977«. In vitro binding of radioiodinated sea turtle (Chelonia mydas) follicle-stimulating hormone to reptilian gonadal tissues. Gen. Comp. Endocrinol. 33: Licht, P. and D. Crews Gonadotropic stimulation of in vivo progesterone production in reptilian and amphibian ovaries. Gen. Comp. Endocrinol. 29: Licht, P., S. VV. Farmer, A. Bona-Gallo, and H. Papkoff. 1979a. Pituitary gonadotropins in snakes. Gen. Comp. Endocrinol. 38: Licht, P., S. VV. Farmer, and H. Papkoff Further studies on the chemical nature of reptilian pituitary gonadotropins: FSH and LH in the American alligator and green sea turtle. Biol. Reprod. 14:

10 574 PAUL LICHT Licht, P., S. W. Farmer, and H. Papkoff Biological activity of hybrid combinations of ovine gonadotropin and steroids associated with breed- Licht, P., W. Rainey, and K. Cliffton Serum and sea turtle LH subunits. Gen. Comp. Endocrinol. 35: das. II. Mating and nesting in natural populaing activities in the green sea turtle Chelonia my- Licht, P., D. S. MacKenzie, H. Papkoff, and S. W. tions. Gen. Comp. Endocrinol. 40: Farmer. 1977A. Immunological studies with the Licht, P., J. Wood, D. W. Owens, and F. Wood. gonadotropins and their subunits from the green 1979i. Serum gonadotropin and steroids associated with breeding activities in the green sea sea turtle Chelonia mydas. Gen. Comp. Endocrinol. 33: turtle Chelonia mydas. I. Captive animals. Gen. Licht, P. and A. R. Midgley, Jr Competition Comp. Endocrinol. 39: for the in vitro binding of radioiodinated human Owens, D. W The endocrine control of reproduction and growth in the green sea turtles follicle-stimulating hormone in reptilian, avian and mammalian gonads by nonmammlian gonadotropins. Gen. Comp. Endocrinol. 30:346- na, Tucson. Chelonia mydas. Ph.D. Diss., University of Arizo Owens, D. W., J. R. Hendrickson, V. Lance, and I. Licht, P. and H. Papkoff Phylogenetic survey P. Callard A technique for determining of the neuraminidase sensitivity of reptilian gonadotropin. Gen. Comp. Endocrinol. 23:415- munoassay. Herpetologica 34: sex of immature Chelonia mydas using a radioim Papkoff, H., S. W. Farmer, and P. Licht Isolation and characterization of follicle stimulating Licht, P. and H. Papkoff Species specificity as a key to structural evolution in pituitary glycoprotein hormones. In P. J. Gaillard and H. H. units from snapping turtle (Chelonia serpentina) hormone and luteinizing hormone and its sub- Boer (eds.), Comparative endocrinology, pp pituitaries. Endocrinology 98: Elsevier/North-Holland, Amsterdam. Tsui, H. W. and P. Licht Gonadotropin regulation of in vitro androgen production by rep- Licht, P., H. Papkoff, S. W. Farmer, C. H. Muller, H. W. Tsui, and D. Crews. 1977c. Evolution in gonadotropin structure and function. Rec. Progr tilian testes. Gen. Comp. Endocrinol. 31:422- Hormone Res. 33: