Department of Veterinary Physiology,3 Veterinary Medical Science, and Laboratory ... Laboratory of Animal Physiology,6 Faculty of Agriculture, Meiji University, ...
BIOLOGY OF REPRODUCTION 56, 847-851 (1997)
Different Female Reproductive Phenotypes Determined by Human Growth Hormone (hGH) Levels in hGH-Transgenic Rats' Akihiro Ikeda,3 Yoshiki Matsumoto,s Kyu-Tae Chang,3 Takahiro Nakano,3 Shigemi Matsuyama, 3 Keitaro Yamanouchi, 3 Akihiko Ohta,6 Masugi Nishihara, 3 Hideaki Tojo,4 Fumihiko Sasaki, 5 and Michio Takahashi 2 ,3 Department of Veterinary Physiology,3 Veterinary Medical Science, and Laboratory of Applied Genetics, 4 Institute of Animal Resource Science, The University of Tokyo, Bunkyo-ku, Tokyo 113, Japan Department of Anatomy,5 Osaka Prefecture University, College of Agriculture, Mozuume-machi, Sakai, Osaka 591, Japan Laboratory of Animal Physiology, 6 Faculty of Agriculture, Meiji University, Tama-ku, Kawasaki-shi 214, Japan ABSTRACT The effect of continuous human GH (hGH) secretion on female reproduction was studied in adult female transgenic rats expressing the hGH gene with a mouse whey acidic protein (mWAP) promoter. Two lines of transgenic female rats carrying the mWAP/hGH gene were established and used in the study. One was characterized by relatively high levels of serum hGH (high line), and the other had relatively low levels (low line). High-line female rats had recurring, pseudopregnancy-like estrous cycles accompanied by increased serum progesterone levels for 2 wk after ovulation, and they were fertile. In these rats, luteinization occurred spontaneously without cervical stimulation, probably due to high levels of serum hGH, which has prolactin (PRL)-like activity inthe rat. Although low-line female rats had recurring, regular 4-day estrous cycles, they were sterile. In these rats, pseudopregnancy could not be induced by mating or by mechanical cervical stimulation. PRL surges following cervical stimulation were not detected, and PRL secretion was not induced by a dopamine antagonist, metoclopramide. The ovarian tissue in this line had a much higher number of corpora lutea and grew much heavier than in normal littermates, suggesting impairment of PRL-induced structural luteolysis. Suppression of PRL secretion in low-line rats was, at least in part, due to a marked decrease in the number of lactotrophs in the pituitary. The present study shows that the serum hGH level plays a crucial role in regulating luteal function in female transgenic rats expressing the hGH gene. INTRODUCTION Investigations of transgenic mice that secrete foreign growth hormone (GH) in many extrapituitary sites are contributing to understanding of the physiological roles of GH [1-5]. Mice made transgenic for human [6], rat [7] or bovine [8] GH have accelerated growth rates. The female offspring are usually sterile [6, 9, 10], even though most male offspring are fertile [11]. Bartke et al. [10] reported that female mice expressing a mouse metallothionein-1/human (h)GH hybrid gene were sterile because luteal function failed as a consequence of reduced prolactin (PRL) levels, and suggested that relatively low levels of hGH were sufficient to inhibit endogenous PRL release but not sufficient to support luteal function. Accepted October 29, 1996. Received July 2, 1996. 'Supported in part by a Grant-in Aid from the Scientific Research found of the Ministry of Education, Science and Culture, Japan, and a grant for a pioneering research project in biotechnology from Ministry of Agriculture, Forestry and Fisheries, Japan. 2 Correspondence: Michio Takahashi, Department of Veterinary Physiology, Veterinary Medical Science, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113, Japan. FAX: 81-3-3815-4266.
847
We established two lines of transgenic rats expressing hGH at high or low levels [12]. Because we found that these two lines of rats expressed completely different reproductive phenotypes, the present study was undertaken to identify the mechanism responsible for these reproductive characteristics based on the PRL-like activity of hGH in transgenic rats. MATERIALS AND METHODS Animals and Generation of Transgenic Rats Wistar strain rats, purchased from Imamichi Institute for Animal Reproduction (Tsuchiura, Japan), were housed in a room at 23 1°C with a lighting schedule of 14L:10D (lights-on at 0500 h). Two lines of adult (9-17 wk old) transgenic female rats carrying the mouse whey acidic promoter (mWAP)/hGH gene were used in this study [12]. One line was characterized by relatively high levels of serum hGH (high line), and the other had relatively low levels (low line). To obtain these female rats, male transgenic rats from each strain (a heterozygote) were mated with normal adult female rats, resulting in transgenic (low line, n = 29; high line, n = 16) and nontransgenic female littermates. The nontransgenic littermates of low-line rats were used as normal rats (n = 21). Vaginal smears were taken every day for cytological estrous evaluation. On the evening of proestrus, some rats were placed with a fertile male, while others were cervically stimulated with a glass rod to induce pseudopregnancy. To raise PRL levels, anterior pituitary tissue taken from a normal adult male rat was transplanted underneath the kidney capsule of each of 4 low-line female rats at 50-65 days of age. After confirmation that a rat was having recurring pseudopregnancy-like cycles, she was placed into the cage with a normal fertile male for 1-3 days starting from the day on which nucleated cells dominated the vaginal cytology. Establishment of pregnancy and occurrence of delivery were monitored. To test whether a dopamine antagonist could induce PRL secretion, the same 4 female low-line transgenic rats and 4 normal rats at 60 days of age were given an i.p. injection of metoclopramide (Wako Chemicals, Tokyo, Japan; 3 mg/5 ml per kg body weight) at 1000-1100 h, and blood samples were collected for PRL assay from the tail vein without anesthesia at 0 min (before the treatment) and 30 and 60 min later. One ovary was taken from each of 4 low-line and 3 high-line transgenic rats, and from 4 normal rats, at 80 days of age and weighed. The ovaries from one low-line and one
848
IKEDA ET AL.
TABLE 1. Characteristics of female transgenic rats used in this study.
Line High hGH female Low hGH female
Serum hGHa (ng/ml)
Relative growth ratio,
468.3 _ 153.5
1.788
21.37 + 1.15
1.068
Estrous cycleb Prolonged (12-14)c Normal 4 days
120 100
No. of rats becoming pregnant
80 60 40 20
3/3 0/8
0 C
aAs shown in a previous report [12].
bConfirmed by observation of daily vaginal smear. c Range of cycle length in days.
E a a)
normal rat were prepared as described below (Morphometric Study), and the section at the maximal diameter of the tissue was used for morphological observation.
cl0
g)
Blood Sampling and Hormone Measurements For progesterone assays, high-line female rats were placed under diethyl ether anesthesia, and blood samples were collected at 1100 h daily, for 2 wk, from the tail vein. Serum progesterone was measured by RIA with an antibody raised in a rabbit against 4-pregnene-1 l-ol-3,20-dione hemisuccinate:BSA as described previously [13]. For PRL assay during estrous cycle or pseudopregnancy, blood samples were collected by decapitation at 1700 h on the day of proestrus or at 0500 h on the day of estrus from the rats that had received cervical stimulation. Serum PRL concentrations were determined by double-antibody RIA using materials supplied by the NIDDK Hormone Distribution Program (Baltimore, MD). The reference standard for the rat (r)PRL assay was NIDDK-rPRL-RP3. The standard curve was linear between 0.5 and 64 ng/ml. Samples with values above 64 ng/ml were reassessed at dilutions ranging from 1:2.5 to 1:10. Morphometric Study Pituitary glands from low-line, high-line, and normal rats (n = 5 for each group) were fixed in 10% neutral buffered formalin, dehydrated in a graded series of alcohol solutions, and embedded in paraffin wax. The tissues were cut sagittally into 6-pum-thick sections. Morphological methods for measuring the volume of the anterior lobe, percentage, and absolute number of PRL-immunoreactive cells have been described previously [14]; sagittal sections from the lateral, parasagittal, and midsagittal parts of the pituitary gland from control and transgenic rats were stained immunohistochemically with antiserum to rPRL (1: 1000). Sections were mounted on microscopic glass slides, Epon was removed with sodium ethoxide (C2H5ONa) in ethanol, and the sections were immersed in 4% H 2 02 solution for removal of osmium. The sections were washed in PBS and treated with antiserum for 0.5 h, rinsed in PBS, and incubated with peroxidase-goat anti-rabbit immunoglobulin G serum (Zymed Laboratories, San Francisco, CA). All of these semithin sections were incubated with 0.02% 3,3'-diaminobenzidine tetrahydrochloride solution containing 0.005% H 20 2 for about 30 min and observed under a light microscope. The sections were measured morphometrically with an image analyzer (Cosmozone-S; Nikon, Tokyo Japan). About 1500 pituitary cell nuclei were counted in the three parts of the anterior lobe of each animal. In addition, the nuclei of cells showing positive immunoreactivity with PRL antiserum were counted and expressed as percentages of the total number of pituitary cells.
4
6
8
10
12
14
22
4 4
61 6
8
10
12 12
14
2
4
6
8
10
12
1,4
120 100 80
6O60
I
20 40 20
O.
E
2
120 100
o0
I
80
60 40 20 0
i (Day)
FIG. 1. Representative serum progesterone changes in high-line female transgenic rats. Day 0 is defined as the day of ovulation as predicted by vaginal smear. Each panel shows data for an individual rat.
Statistical Analysis The data for Figure 2a were analyzed by ANOVA followed by Student's t-test. The data on ovarian weight, the data on PRL levels presented in Figure 2b, and the morphometric data presented in Figure 5 were analyzed by ANOVA followed by Duncan's multiple range test. In all statistical tests, a difference was considered significant if p < 0.05. RESULTS
As was the case in a previous investigation [12], highline rats used in this study had relatively high levels of serum hGH (468.3 ± 153.5 ng/ml), while low-line rats had relatively low hGH levels (21.37 + 1.15 ng/ml). As shown in Table 1, vaginal smear cytology revealed a prolonged estrous cycle in high-line rats (n = 5), but 3 of these rats mated and were fertile. On the other hand, low-line rats (n = 8) had recurring regular 4-day cycles and mated, but were sterile. Pseudopregnancy could not be induced in these low-line rats by cervical stimulation or normal mating. In high-line rats, serum progesterone levels gradually increased after ovulation (Fig. 1). In cervically stimulated low-line rats, PRL surges did not occur as is normally expected on the evening of proestrus or in the morning on the day of estrus (Fig. 2a). Treatment with the dopamine antagonist metoclopramide could not stimulate PRL secretion in low-line rats; this contrasted with findings in the normal littermates, whose PRL secretion was markedly stimulated (Fig. 2b). The ovarian weights in low-line rats,
TRANSGENIC RATS EXPRESSING hGH GENE
849
A ---% ZU
15
E a
10C
-I 0n 5
Low line
Normal
B 8CD0
Low line
--, #
+c
Normal
60
S a
C
40
-j CL 0.
20D0-
0
30
60
minute FIG. 2. A) PRL levels in low-line transgenic rats and normal littermates at 1700 h on the day of proestrus (n = 4 for each group) and at 0500 h on the day of estrus (n = 4 for each group). Cervical stimulation was given at 1800 h on proestrous day. Each column represents the mean SE. *, Different from values for transgenic rats (p < 0.05). B) PRL responses to metoclopramide in low-line transgenic rats and normal littermates (n = 4 for each group). #, Different from values in the respective 0-min controls (p < 0.01). *, Different from values in transgenic rats (p < 0.01).
high-line rats, and normal littermates were 160.3 18.3 mg, 45.2 + 11.2 mg, and 38.7 + 2.25 mg, respectively. The ovarian weight of low-line rats was significantly (4-fold) higher than that of normal littermates, and the histology of the ovaries of the low-line animals indicated the presence of numerous corpora lutea (Fig. 3). Three of four pituitary-grafted low-line female rats that were tested became pregnant and delivered 9.0 + 1.52 pups. All of the pups were born alive and appeared grossly normal. Immunohistochemical analysis for PRL in the pituitary indicated that the number and percentage of PRL-producing cells decreased in both lines of transgenic rats (Fig. 4). The percentages and total numbers of PRL-immunoreactive cells are shown in Figure 5, a and b. The percentages of PRL-immunoreactive cells in high-line rats and low-line rats decreased to 62% and 74%, respectively, compared to the value in age-matched normal rats (Fig. 5a). The total cell number also decreased in both lines of transgenic rats (to 25% of normal values in high-line and to 30% of normal values in low-line rats) (Fig. 5b).
FIG. 3. Histology of the ovary in low-line transgenic rat (A) and normal littermate (B). Note the heavily accumulated corpora lutea in the ovary of the transgenic rat. X13.
DISCUSSION Extended estrous cycles and elevated serum progesterone levels were observed in high-line rats. This finding contrasts with observations in transgenic mice expressing bovine (b)GH, whose luteal function is not stimulated [15]. Because bGH is believed to have only GH, not PRL, activity in rodent species, the results for the bGH transgenic mice could be interpreted as indicating that GH cannot stimulate luteal function through the GH receptor. On the other hand, hGH has been shown to bind the PRL receptor as well as the GH receptor in the rat [16]. Although GH actions of hGH cannot be ruled out completely, the stimulation of luteal function in high-line rats can be attributed to hGH binding with PRL receptors in the rat [16]. This cause-and-effect relationship in high-line rats seems to be parallel to that observed in pituitary-isografted rats, where spontaneous luteinization occurs after each ovulation due to PRL secretion from the graft [17]. These pseudopregnancies recur spontaneously. Pituitary-isografted rats are fertile; high-line rats can mate between two pseudopregnancy-like cycles and are similarly fertile.
850
IKEDA ET AL. Percentage Io0/\
%-/oI
iA n
.u
b 30-
a a
2010-
0Low line High line Normal
Numberr b
(x10 5 ) 30-
t
20-
10-
a
0Low line High line
Normal
FIG. 5. Percentages and numbers of PRL-positive cells in pituitary tissue. Each column represents the mean + SE. Values with different superscripts are significantly different from each other (p < 0.05).
FIG. 4. Pituitary cells immunoreactive with antisera to PRL in control (A), high-line (B), and low-line (C) female rats. Nuclei are stained with hematoxylin (purplish blue). Immunoreactive cells were stained brown. x750.
In contrast to high-line rats, low-line rats had recurring regular 4-day estrous cycles but were infertile. Neither cervical stimulation nor natural mating could induce pseudopregnancy. These results suggest two points: first, that hGH levels in our low-line rats were inadequate to functionally luteinize corpora lutea after each ovulation, and, secondly, that the PRL surge was somehow impaired. When the pituitary was isografted to low-line rats to boost and maintain high peripheral PRL levels, the animals had recurring pseudopregnant-like cycles and became fertile. The recovery of fertility via supplementation with PRL indicated the inadequacy of hGH levels in low-line rats for rescuing luteal function. In cervically stimulated low-line rats, the expected spontaneous PRL surge on the evening of proestrus and the morning of estrus was completely missing. Furthermore, the dopamine receptor antagonist [18] metoclopramide could not induce PRL secretion in low-line transgenic female rats, indicating that these transgenic rats were incapable of responding to the surge-inducing mechanism of the hypothalamus [19]. Thus, infertility in the low-line rats could be attributed to the absence of either an induced or a spontaneous PRL surge. Abnormal increases in ovarian weight in low-line rats were also probably due to the absence of a cyclic PRL surge on each proestrous day, since this PRL surge is known to be a prerequisite for the induction of structural luteolysis [20]. The normal ovarian weight of the high-line rats indicates that hGH at that level is effective in inducing structural luteolysis. Immunohistochemical analysis of the anterior pituitary in both high- and low-line rats indicated that the number
TRANSGENIC RATS EXPRESSING hGH GENE
of lactotrophs and somatotrophs (data not shown for somatotrophs) decreased to a large extent. Although the possibility cannot be excluded that some neuronal disorders were induced in these transgenic rats, the main reason for the inability of the transgenic rats to express a PRL surge would be either the marked decrease in the number of lactotrophs or the insensitivity of lactotrophs to the withdrawal of dopamine inhibition. Though there was a 20-fold difference in peripheral hGH levels as well as various phenotypic differences between low- and high-line rats, the decrease in the number of lactotrophs occurred to exactly the same extent in both lines. The normal differentiation process of lactotrophs must be disturbed by nonprogrammed expression of GH and/or PRL activity from the transgene, and this process may be highly sensitive to GH and/or PRL activity. Because PRL secretion from the pituitary was severely suppressed in our two lines of transgenic rats, differences in the reproductive characteristics described above may be solely attributed to the difference in peripheral hGH levels. The different reproductive phenotypes observed in this study occurred similarly in transgenic mice expressing the hGH gene [10, 21]. Transgenic mice carrying a construct of the metallothionein/hGH chimeric gene showed relatively lower blood hGH levels than transgenic mice carrying a construct of phosphoenolpyruvate carboxykinase (PEPCK)/ hGH chimeric gene, and were infertile. The transgenic mice carrying the (PEPCK)/hGH chimeric gene construct showed much higher hGH levels and were fertile [21]. These observations and our present results indicate that the differential responsiveness of luteal function to low and high peripheral hGH is similar between female mice and rats. ACKNOWLEDGMENT The authors thank NIDDK for providing the materials with which to perform the RIA and gratefully acknowledge Mrs. Ford for proofreading the manuscript.
5.
6. 7.
8.
9.
10.
11.
12.
13. 14.
15. 16.
REFERENCES 1. Palmiter RD, Brinster RL. Germ-line transformation of mice. Annu Rev Genet 1986; 20:465-499. 2. Bchini O, Andres AC, Shubaur B, Mehtali M, Lemeur M, Lathe R, Gerlinger P. Precocious mammary gland development and milk protein synthesis in transgenic mice ubiquitously expressing human growth hormone. Endocrinology 1991; 128:539-546. 3. Chandrashekar V, Bartke A, Wagner TE. Neuroendocrine function in adult female transgenic mice expressing the human growth hormone gene. Endocrinology 1992; 130:1802-1808. 4. Orian JM, Lee CS, Weiss LM, Brandon MR. The expression of a metallothionein-ovine growth hormone fusion gene in transgenic mice
17. 18. 19. 20. 21.
851
does not impair fertility but results in pathological lesion in the liver. Endocrinology 1989; 124:455-463. Orian JM, Snibson K, Stevenson JL, Brandon MR, Herington AC. Elevation of growth hormone (GH) and prolactin receptors in transgenic mice expressing ovine GH. Endocrinology 1991; 128:12381246. Palmiter RD, Norstedt G, Gelinas RE, Hammer RE, Brinster RL. Metallothionein-human growth hormone fusion genes stimulate growth of mice. Science 1983; 222:809-814. Palmiter RD, Brinster RL, Hammer RE, Trumbauer ME, Rosenfeld MG, Birnberg NC, Evans RM. Dramatic growth of mice that develop from eggs microinjected with metallothionein-growth hormone fusion genes. Nature 1982; 300:611-615. McGrane MM, Devente J, Yun J, Bloom J, Park E, Wynshaw-Boris A, Wagner T, Rottman FM, Hanson RW. Tissue-specific expression and dietary regulation of a chimeric phosphoenolpyruvate carboxykinase/bovine growth hormone gene in transgenic mice. J Biol Chem 1988; 263:11443-11451. Naar EM, Bartke A, Majumdar SS, Buonomo FC, Yun JS, Wagner TE. Fertility of transgenic female mice expressing bovine growth hormone or human growth hormone variant genes. Biol Reprod 1991; 45:178-187. Bartke A, Steger RW, Hodges SL, Parkening TA, Collins TJ, Yun JS, Wagner TE. Infertility in transgenic female mice with hGH expression: evidence for luteal failure. J Exp Zool 1988; 248:121-124. Bartke A, Naar EM, Johnson L, May MR, Cecim M, Yun IS, Wanger TE. Effects of expression of human or bovine growth hormone genes on sperm production and male reproductive performance in four lines of transgenic mice. J Reprod Fertil 1992; 95:109-118. Ikeda A, Matsuyama S, Nishihara M, Tojo H, Takahashi M. Changes in endogenous growth hormone secretion and onset of puberty in transgenic rats expressing human growth hormone gene. Endocr J 1993; 41:523-529. Matsuyama S, Shiota K, Takahashi M. Possible role of transforming growth factor-E3 as a mediator of luteotropicaction of prolactin in rat luteal cell cultures. Endocrinology 1994; 127:1561-1567. Sasaki E Kawai T, Ohta M. Immunohistochemical evidence of neurons with GHRH or LHRH in the arcuate nucleus of male mice and their possible role in the postnatal development of adenohypophysial cells. Anat Rec 1994; 240:255-260. Cecim M, Kerr J, Bartke A. Infertility in transgenic mice overexpressing the bovine growth hormone gene: luteal failure secondary to prolactin deficiency. Biol Reprod 1995; 52:1162-1166. Posner BI, Kelly PA, Shiu RPC, Paud R, Friesen HG. Studies of insulin, growth hormone and prolactin binding: tissue distribution, species variation and characterization. Endocrinology 1974; 95:521531. de Greef WJ, Zeilmaker GH. Regulation prolactin secretion during the luteal phase in the rat. Endocrinology 1978; 102:1190-1198. Yamauchi J, Takahara J, Ofuji T Effect of metoclopramide on rat prolactin secretion in vivo. Life Sci 1977; 20:1581-1583. Leong DA, Frawley LS, Neill JD. Neuroendocrine control of prolactin secretion. Annu Rev Physiol 1983; 45:109-127. Wuttke W, Meites J. Luteolytic role of prolactin during estrous cycle of the rat. Proc Soc Exp Biol Med 1971; 137:988-991. Milton S, Cecim M, Li YS Yun JS, Wanger TE, Bartke A. Transgenic female mice with high human growth hormone levels are fertile and capable of normal lactation without having been pregnant. Endocrinology 1992; 131:536-538.
