Prolactin and Paternal Behavior in the Biparental California Mouse, Peromyscus californicus

Transcription

1 HORMONES AND BEHAVIOR 23, (1989) Prolactin and Paternal Behavior in the Biparental California Mouse, Peromyscus californicus DAVID J. GUBERNICK* AND RANDY J. NELsoNt *Department of Psychology, Indiana University, Bloomington, Indiana 47405, and tdepartments of Psychology and Population Dynamics, The Johns Hopkins University, Baltimore, Maryland Relatively little is known about hormonal mechanisms underlying paternal behavior in mammals. Male California mice, Peromyscus cafifornicus, display extensive parental care toward their young. Parental behavior of fathers, expectant fathers (males living with their pregnant partner), and virgin males was assessed in a IO-min test with a 1- to 3-day-old alien pup. Few virgin males acted parental (19%) compared to fathers one day postpartum (80%) and expectant fathers (56%). Plasma prolactin levels were significantly elevated in fathers 2 days postpartum compared to expectant fathers and virgin males. Paternal prolactin levels were similar to those of mothers. There were no differences between groups in levels of plasma testosterone. These data suggest, contrary to other reports, that protactin is a likely correlate of paternal behavior in rodents. o 1989 Academic Press. Inc. Male parental care is relatively uncommon in mammals and is found predominantly among rodents, carnivores, and primates (Kleiman and Malcolm, 1981). Information about male parental care is accumulating rapidly, but relatively little is known about the proximate causation of paternal behavior in any mammal (Daly and Wilson, 1978; Elwood, 1983b). The proximate causation of maternal behavior in mammals has been investigated extensively (Rosenblatt and Siegel, 1981; Elwood, 1983a). Whereas, the hormonal basis of maternal behavior has been well documented, primarily in rodents (Rosenblatt and Lehrman, 1%3; Rosenblatt, Siegel, and Mayer, 1979; Rosenblatt and Siegel, 1981), hormonal changes associated with male parental care in mammals have been relatively unexplored. The few studies of hormonal correlates of paternal behavior in rodents used species in which males do not normally display parental care, but are induced to act parental either through repeated exposure to infants Current address: Department of Psychology, University of Wisconsin, Madison, WI X/89 $1.50 Copyright by Academic Press, Inc. All rights of reproduction in any form reserved.

2 204 GUBERNICK AND NELSON (e.g., Bridges, 1983; Tate-Ostroff and Bridges, 1985) or hormonal manipulation (Lubin, Leon, Moltz, and Numan, 1972). Under such conditions, males are hormonally less responsive than females (Samuels and Bridges, 1983) or in one case, hormonal changes are correlated with the male s behavior (Brown and Moger, 1983; but see Jakubowski and Terkel, 1986 for contradictory findings). Paternal behavior can be observed in approximately 6% of rodent genera (Kleiman and Malcolm, 1981), but most instances are poorly documented. Male California mice, Peromyscus californicus, display extensive paternal care toward their young from birth to weaning (Gubemick and Alberts, 1987). Fathers and mothers spend substantial and equivalent amounts of time in the nest and in physical contact with pups throughout lactation. Males devote more time than females to licking pups. Males retrieve their young and build nests. Fathers display all parental behaviors exhibited by mothers, except lactation. We determined whether fathers differed behaviorally from nonfathers in their responsiveness to infants and whether there were concomitant hormonal differences between fathers and nonfathers. METHODS To determine whether nonfathers would exhibit spontaneous parental behavior, adult virgin males (n = 16) and males living with a pregnant female (expectant fathers, n = 18) were presented with a single l- to 3- day-old pup for 10 min. For comparison, we also tested fathers (n = 15) one day postpartum with a l- to 3-day-old alien pup. The pup was placed in the center of the male s home cage. Each male was tested individually and only once. A male was considered parental if he spent one minute or more licking the pup or crouched over it in a nursing posture. Parental mice typically spent from 2 to 6 min licking a pup or huddled over it. Nonparental animals either ignored or attacked the pup. When attacks occurred, we immediately hit the top of the home cage to disrupt the aggression and we terminated the test. The pup was removed and returned to its parents, and was reared normally. This was a fairly successful procedure (Gubernick and Alberts, in press), but about 10% of attacked pups were still lost. Each pup was used only once. The lomin test was sufficiently long to elicit parental behavior reliably. Attacks occurred typically within 1 min of pup presentation. Animals that ignored the test pup in the first 10 min continued to ignore the pup 30 min later. Additional groups of males were used to collect blood. Blood samples (0.5 to 1.5 ml) were obtained from fathers (n = 11) 2 days postpartum, expectant fathers 10 days prior to parturition (n = lo), and individually isolated adult virgin males (n = 14). For comparison, blood samples were taken from mothers 2 days postpartum (n = 11). Animals were anesthetized with methoxyflurane (metofane) and blood was obtained

3 PROLACTIN AND PATERNAL BEHAVIOR 205 within 3 min by heart puncture. Blood was collected from all animals between 1400 and 1600 hr. Samples were centrifuged at 3000 rpm for 30 min at 4 C. Plasma aliquots were stored at - 70 C until assayed. Prolactin and testosterone levels in blood plasma were determined by radioimmunoassay (RIA). Because prolactin levels were substantially elevated in the California mouse compared to other rodents (see Results), blood samples were also obtained from 11 adult, virgin male Peromyscus maniculatus bairdii for interspecies comparison and as an additional control for possible effects of our blood collection procedures. Radioimmunoassay Prolactin was measured in duplicate lo-p.1 aliquots of plasma with double antibody RIA developed with a sensitive rabbit first antibody raised against deer mouse prolactin (Peromyscus maniculatus bairdii) (Marr, Colosi, Desjardins, and Talamantes, 1983). The second antibody was sheep anti-rabbit y-globulin obtained from Antibodies, Inc. (Davis, CA). Final hormone concentrations are expressed as nanogram equivalents of a purified deer mouse prolactin per milliliter of California mouse blood plasma. The standard curve was based upon nine values ranging from 0.1 to 10 rig/tube. The lower limit of this assay was 10 rig/ml and the upper limit was 90 rig/ml blood sample. A few values exceeded the upper limit and were extrapolated by computer. All plasma samples were measured in a single assay. The within-assay coefficient of variation (CV) for pools of plasma obtained from intact or castrated males or from intact, nulliparous, or lactating female mice (eight aliquots each) was 6.2,7.1,6.0, and 8.0%, respectively. Five samples of 10 different dilutions of pooled intact lactating female blood plasma were made and the resulting curve was parallel to the inhibition curve of purified deer mouse prolactin. Testosterone was measured in 25~1 blood plasma aliquots with a double antibody RIA procedure validated for use with small volumes of Peromyscus blood plasma (Desjardins and Lopez, 1983) that relied upon a specific rabbit antiserum to a 19-O-carboxymethyl-ether derivative of testosterone (Radioassay Systems Laboratories, Carson, CA). All samples were analyzed during a single assay. The within-assay CV was 7.0% for 10 samples of blood plasma obtained from pooled blood of several intact male California mice. The testosterone assay has been validated with Peromyscus several times. In the current study, we ran pooled castrated male, intact male, and intact female plasma. Castrated male and female plasma had nondetectable values of testosterone. Five samples of 10 different dilutions of pooled normal male blood plasma were made and the resulting curve was parallel to the inhibition curve of highly purified testosterone. We used x2 tests to detect differences between groups in the number

4 206 GUBERNICK AND NELSON of males that acted parental versus nonparental (attack + ignore). Oneway analysis of variance was used to determine differences between groups in plasma levels of prolactin and testosterone. Planned comparisons were at 0.05 level of significance. RESULTS There was a significant difference between groups in the number of parental and nonparental males [x2(2) = 11.85, P < (Table 1). More fathers and expectant fathers were parental compared to isolated males (~ $1) = and 4.86, P s < 0.01 and 0.05, respectively]. There was no difference between the number of fathers and expectant fathers that acted parental [x*(l) = There was a significant difference in prolactin levels between groups [F(3, 42) = 4.15, P <.05]. Plasma prolactin titers of mothers, fathers, expectant fathers, and isolated males are shown in Fig. 1. Prolactin was elevated significantly in mothers ( rig/ml) and fathers ( rig/ml) as compared to expectant fathers ( rig/ml) and nonfathers (593.3? 82.9) (P s <.05 in every case). There were no differences between mothers and fathers or between expectant fathers and nonfathers. In contrast to the California mouse, plasma prolactin levels of male P. maniculutus bairdii were markedly lower ( nshl). Plasma testosterone levels were characteristically low in the California mouse and did not differ between fathers (30.0 t 6.8 pg/ml), expectant fathers ( pg/ml), and isolated males (53.9? 27.3 pg/ml). Testosterone appears not to be associated with spontaneous male parental care in adult virgin, Rockland-Swiss albino mice (Svare, Bartke, and Gandelman, 1977). DISCUSSION Several studies have found either no prolactin responses by sensitized male rats to pup stimulation (Samuels and Bridges, 1983; Sodersten and Eneroth, 1984) or decreased levels of prolactin (Brown and Moger, 1983; however, Jakubowski and Terkel (1986) were unable to find any relation between prolactin and induced male parental behavior in rats). In contrast, we found a striking difference in prolactin levels of Peromyscus californicus TABLE 1 Number of Males Exhibiting Parental or Nonparental Behavior toward Test Pup Behavior Isolated male Expectant father Father Parental Nonparental ) (16) (18) (15)

5 PROLACTIN AND PATERNAL BEHAVIOR 207 Mothers Fathers Expectant Isolated Fathers Males FIG. 1. Plasma prolactin titers (rig/ml) (means + 1 SEM) of California mice (Peromyscus culifornicus). Blood samples were obtained from mothers (N = 11) and fathers (N = 11) 2 days postpartum, from expectant fathers within 10 days of parturition (N = lo), and from isolated, adult virgin males (N = 14). Fathers were in contact with pups prior to blood collection. fathers compared to nonfathers. Prolactin was elevated in fathers relative to expectant fathers and virgin males. Elevated prolactin levels in Peromyscus californicus fathers are probably not due to recency of mating because postpartum mating occurs on the day of birth (Gubernick, 1988) and blood samples were taken 2 days later. Prolactin levels in the California mouse are substantially higher than in other rodents. Since blood samples were obtained within 3 min of anesthesia, it is unlikely that some nonspecific stress associated with blood collection significantly affected prolactin levels. Furthermore, analysis of blood obtained from P. manicdutus bairdii under identical collection procedures revealed markedly lower prolactin levels. Thus the high levels of prolactin reported here for the California mouse are apparently more typical for this species. Expectant fathers displayed a higher incidence of paternal behavior than isolated virgin males, indicating that the onset of paternal behavior occurs during the pregnancy of the male s mate. A similar pregnancy effect has been found for male gerbils (Elwood, 1977). It is possible that prolactin levels are significantly elevated in expectant fathers that act parental. Because we did not behaviorally test males prior to blood collection both parental and nonparental expectant fathers (and virgin males) were used for hormonal assays. Future studies will examine more closely the prepartum onset of paternal behavior and the dynamic endocrine

6 208 GUBERNICK AND NELSON changes of expectant fathers throughout the pregnancy of the male s partner. Suckling stimulation and olfactory stimulation from pups enhance prolactin secretion in female mammals (Grosvenor, 1965; Koranyi, Phelps, and Sawyer, 1977; Stem and Siegel, 1978). Tactile stimulation and olfactory cues from the pups might affect prolactin levels in fathers. Parental male California mice huddle over pups and pups actively probe the ventrum of the male with their snout and attempt to suckle, although males lack nipples. Males typically respond to pup probing by exhibiting an arched nursing posture and remaining quiescent. It is unlikely that mere exposure to a pup would elevate prolactin in virgin males. Virgin male rats induced to act parental through repeated daily exposure to pups fail to show a prolactin response (Jakubowski and Terkel, 1986; Tate-Ostroff and Bridges, 1985) or exhibit decreased levels of prolactin (Brown and Moger, 1983). Pituitary prolactin has been implicated in the induction of maternal behavior in rats (Bridges, DiBiase, Loundes, and Doherty, 1985; Bridges and Dunckel, 1987). Whether increased prolactin is a cause of paternal behavior in Peromyscus californicus or a consequence of caring for young remains to be determined. Only one other study to date has examined hormonal correlates of paternal behavior in a mammal that normally displays male care (Dixson and George, 1982). As in the California mouse, parental male marmosets (a New World primate) exhibited elevated prolactin levels compared to males living with pregnant or nonpregnant females. Testosterone levels did not differ between groups. Elevated prolactin levels in fathers may be a pattern characteristic of mammalian species in which males normally care for their young. The California mouse provides a model system to explore hormonal mechanisms underlying the onset and maintenance of paternal behavior in mammals. Males begin to exhibit parental care prior to birth of their young, fathers display extensive care of their offspring from birth to weaning (Gubemick and Alberts, 1987), and fathers differ hormonally from nonfathers. Comparative analyses with other, closely related, rodents may provide insight into the evolution of hormonal influences on male parental behavior. ACKNOWLEDGMENTS We thank Rusty Schmidt and Laurie Mason for technical assistance and Dr. F. Talamantes for providing purified deer mouse prolactin and the first antibody. This research was supported by a NICHD grant (HD-21233) to David J. Gubemick and NIH Biomedical Support Grant to Randy J. Nelson. REFERENCES Bridges, R. S. (1983). Sex differences in prolactin secretion in parental male and female rats. Psychoneuroendocrinology, 8, 109-l 16.

7 PROLACTIN AND PATERNAL BEHAVIOR 209 Bridges, R. S., DiBiase, R., Loundes, D. D., and Doherty, P. C. (1985). Prolactin stimulation of maternal behavior in female rats. Science 227, Bridges, R. S., and Dunckel, P. T. (1987). Hormonal regulation of maternal behavior in rats: Stimulation following treatment with ectopic pituitary grafts plus progesterone. Biol. Reprod. 37, Brown, R. E., and Moger, W. H. (1983). Hormonal correlates of parental behavior in male rats. Horm. Behav. 17, Daly, M., and Wilson, M. I. (1978). Sex, Evolution, and Behavior. Duxbury Press, North Scituate, MA. Desjardins, C., and Lopez, M. J. (1983). Environmental cues evoke differential responses in pituitary-testicular function in deer mice. Endocrinology 112, Dixson, A. F., and George, L. (1982). Prolactin and parental behaviour in a male New World primate. Nature (London) 299, Elwood, R. W. (1977). Changes in the responses of male and female gerbils (Meriones unguiculutus) towards test pups during the pregnancy of the female. Anim. Behav. 25, Elwood, R. W. (Ed.) (1983a). Parental Behaviour in Rodents. Wiley, New York. Elwood, R. W. (1983b). Paternal care in rodents. In R. W. Elwood (Ed.), Parental Behaviour in Rodents, pp Wiley, New York. Grosvenor, C. E. (1965). Evidence that exteroceptive stimuli can release prolactin from the pituitary gland of the lactating rat. Endocrino/ogy 76, Gubernick, D. J. (1988). Reproduction in the California mouse, Peromyscus cakjornicus. J. Mammal. 69, Gubernick, D. J., and Alberts, J. R. (1987). The biparental care system of the California mouse, Peromyscus californicus. J. Comp. Psycho/. 101, Gubemick, D. J., and Alberts, J. R. (In press). Postpartum maintenance of paternal behaviour in the biparental California mouse, Peromyscus cahfornicus. Anim. Behav. Jakubowski, M., and Terkel, J. (1986). Nocturnal surges and reflexive release of prolactin in parentally behaving virgin female and male rats. Horm. Behav. 20, Kleiman, D. G., and Malcolm, J. R. (1981). The evolution of male parental investment. In D. J. Gubemick and P. H. Klopfer (Eds.), Parental Care in Mamma/s, pp Plenum, New York. Koranyi, L., Phelps, C. P., and Sawyer, C. H. (1977). Changes in serum prolactin and corticosterone in induced maternal behavior in rats. Physiol. Behav. 18, Lubin, M., Leon, M., Moltz, H., and Numan, M. (1972). Hormones and maternal behavior in the male rat. Horm. Behav. 3, Marr, G. A., Colosi, P., Desjardins, C., and Talamantes, F. (1983). Development and characterization of a homologous radioimmunoassay for deer mouse (Peromyscus maniculafus bairdii) prolactin. Life Sci. 33, Rosenblatt, J. S., and Lehrman, D. S. (1963). Maternal behavior of the laboratory rat. In H. L. Rheingold (Ed.), Maternal Behavior in Mammals, pp Wiley, New York. Rosenblatt, J. S., and Siegel, H. I. (1981). Factors governing the onset and maintenance of maternal behavior among nonprimate mammals. In D. J. Gubemick and P. H. Klopfer (Eds.), Parental Care in Mammals, pp Plenum, New York. Rosenblatt, J. S., Siegel, H. I., and Mayer, A. D. (1979). Progress in the study of maternal behavior in the rat: Hormonal, nonhormonal, sensory, and developmental aspects. In J. S. Rosenblatt, R. A. Hinde, C. Beer, and M. C. Busnel (Eds.), Advances in the Study of Behavior, Vol. 10, pp Academic Press, New York. Samuel% M. H., and Bridges, R. S. (1983). Plasma prolactin concentrations in parental male and female rats: Effects of exposure to rat young. Endocrinology 113, Sodersten, P., and Eneroth, P. (1984). Effects of exposure to pups on maternal behavior,

8 210 GUBERNICK AND NELSON sexual behavior and serum prolactin concentrations in male rats. J. Endocrinol. 102, Stem, J. M., and Siegel, H. I. (1978). Prolactin release in lactating, primiparous and multiparous thelectomized and maternal virgin rats exposed to pup stimuli. Biol. Reprod. 19, Svare, B., Bartke, A., and Gandelman, R. (1977). Individual differences in the maternal behavior of male mice: No evidence for a relationship to circulating testosterone levels. Horm. Behav. 8, Tate-Ostroff, B., and Bridges, R. S. (1985). Plasma prolactin levels in parental male rats: Effects of increased pup stimuli. Horm. Behav. 19,