The Role of Vasopressin in the Genetic and Neural Regulation of Monogamy

Document Sample
The Role of Vasopressin in the Genetic and Neural Regulation of Monogamy Powered By Docstoc
					Journal of Neuroendocrinology, 2004, Vol. 16, 325±332

The Role of Vasopressin in the Genetic and Neural Regulation of Monogamy
M. M. Lim, E. A. D. Hammock and L. J. Young
Center for Behavioral Neuroscience and Department of Psychiatry and Behavioral Sciences, Emory University, Atlanta, GA, USA. Key words: vole, pairbond, microsatellite, reward, ventral pallidum.

Abstract Arginine vasopressin modulates pairbond formation in the monogamous prairie vole (Microtus ochrogaster). Our laboratory has investigated the genetic and neural mechanisms by which vasopressin and its V1a receptor (V1aR) regulate social attachment between mates. Non-monogamous vole species show strikingly different distribution patterns of brain V1aR expression compared to monogamous species, and these patterns are thought to arise from species differences in the respective promoter sequences of the V1aR gene. Individual differences in prairie vole V1aR patterns may also re¯ect individual differences in promoter sequences. Pharmacological and genetic manipulation of the speci®c brain regions that express V1aR in the `monogamous pattern' allows multilevel examination of the neural circuits that underlie pairbond formation in monogamous species. For example, V1aR are expressed in brain regions involved in reward circuitry in monogamous vole species and have been implicated in pairbonding. V1aR are also highly expressed in regions implicated in the olfactory processing of sociosexual behaviour. We hypothesize that both circuits of reward and olfactory memory underlie the cognitive mechanisms that control pairbonding. When used in conjuction, genetic and cellular analyses of a complex social behaviour can provide a coherent framework with which to examine the role of the vasopressin system in species evolution and neural control of behaviour.

As a neuromodulator in the brain, the neurohypophysial hormone arginine vasopressin plays a number of different roles in regulating complex social behaviours in many species (1). One example is the expression of monogamous behaviour, such as long-term, selective pairbond formation between adult mates. The monogamous prairie vole (Microtus ochrogaster) forms pairbonds and thus is an excellent animal model to study the role of vasopressin and its brain receptor subtype, V1a, in pairbonding (2). In male prairie voles, central infusion of vasopressin accelerates pairbonding, while the selective antagonist for the V1a subtype of the vasopressin receptor (V1aR) blocks pairbond formation (3). However, it is unclear how vasopressin acts to stimulate this complex social behaviour. One clue has arisen from the discovery of the neuroanatomical sites of action of vasopressin. Dramatic species differences in V1aR distribution exist between prairie voles and non-monogamous vole species, such as the montane vole (Microtus montanus). Prairie and montane voles are closely related congeners, but interestingly, show a strikingly different pattern of V1aR expression in the brain (Fig. 1A±D) (4). Furthermore, central infusion of vasopressin into montane voles has no effect on their social behaviour (5). This strongly suggests that the speci®c location of V1aR within certain brain regions are what

modulate pairbonding, and not merely the presence of vasopressin itself. Thus, the endogenous release of vasopressin in the brain during mating and cohabitation would activate different neural circuits in a monogamous versus a non-monogamous species. In the past 5 years, our laboratory has focused mainly on two lines of research on vasopressin systems in the prairie vole model, using comparative approaches to examine different aspects of pairbond formation. First, we have investigated the genetic contribution to species differences in pairbond formation. Speci®cally, we examined the species differences in the 5H regulatory region of the V1aR gene and the regulation of V1aR expression in the brain and behaviour (Fig. 1E). Second, we have focused on several candidate brain regions expressing V1aR in prairie, but not montane voles, with the hypothesis that one or more of these regions might reveal the neural circuits by which vasopressin stimulates the pairbond.
The genetics of the pairbond

Genetic differences exist between any two species. We are interested in which of those de facto genetic differences leads directly to changes in social structure. Because there is already strong

Correspondence to: Dr Miranda Lim, 954 Gatewood Road Yerkes Research Center, Emory University Atlanta, GA 30329, USA (e-mail: mmlim@ # 2004 Blackwell Publishing Ltd

326 V1aR and social attachment in voles

Fig. 1. V1aR distribution in prairie vole brain (A,C) and montane vole brain (B,D). Note the disparity of V1aR binding in the ventral pallidum (VP), mediodorsal thalamus (MDthal) and amygdala (Amyg) between the two species. (E) A schematic showing the prairie and montane vole V1aR gene and upstream regulatory region. The prairie vole sequence contains a repetitive expansion, or microsatellite polymorphism, in the promoter region of the gene. The striped boxes represent exons in the coding region of the gene. Sections are taken from age-matched adult male prairie and montane voles, although female voles appear identical. Scale bar ˆ 1 mm. LS, lateral septum.

evidence for V1aR involvement in social behaviour, in a speciesspeci®c manner, we have hypothesized that V1aR is a good target gene to observe possible species differences. Therefore, we compared the V1aR gene sequences of monogamous and nonmonogamous vole species. We hypothesized that we would observe species differences in the regulatory portion of the gene, but not in the coding region, because the phenotypic species differences are in the neuroanatomical distribution of V1aR and not in the af®nity for selective ligands (4). The alignment of the prairie and montane vole V1aR shows that the coding

sequences are 99% identical between the two species, meaning that the protein itself is virtually the same. However, as hypothesized, there is a striking divergence between the two species in the sequence upstream of the V1aR coding region. Speci®cally, in the prairie vole, there is approximately 500 bp of a highly repetitive sequence located at 662 bp upstream of the transcription start site, which is absent in the montane vole (Fig. 1E) (5). Interestingly, a similar sequence expansion is also found in the monogamous pine vole (Microtus pinetorum), and is absent in the non-monogamous meadow vole (Microtus pennsylvanicus) (5). We hypothesize that
# 2004 Blackwell Publishing Ltd, Journal of Neuroendocrinology, 16, 325±332

V1aR and social attachment in voles 327 this species-speci®c repetitive region contributes to the observed species differences in neuroanatomical distribution of V1aR. This kind of repetitive region, also referred to as microsatellite DNA, has been shown experimentally to affect the expression of other genes such as the dopamine transporter and the serotonin transporter (6, 7). It is likely that this microsatellite sequence can modulate the expression of the V1aR gene as well. Not only can this microsatellite modify gene expression, but also it is potentially unstable and prone to mutation by expanding and contracting. This instability has exciting implications for evolution of behaviour. It can easily be imagined how changes at this locus could rapidly lead to differences in neuroanatomical V1aR gene expression patterns, which may directly affect the ability to form a pairbond. Is the V1aR gene responsible for pairbonding behaviour in this species? Can this single gene polymorphism determine the entire pattern of receptors in the brain of a monogamous species, and subsequently affect behaviour? To demonstrate experimentally that there is a direct relationship between the prairie vole V1aR gene structure and V1aR distribution in the brain, mice transgenic for the prairie vole V1aR gene were created (5). Of note, mice have very high coding region homology with the prairie vole V1aR gene and very low non-coding homology. The transgene contained 2.2 kb of the 5H ¯anking region (including the microsatellite), the entire coding sequence including the intron, and 2.4 kb of the 3H ¯anking region of the prairie vole V1aR gene. These transgenic mice showed a pattern of V1aR binding remarkably similar, although not identical, to prairie voles, and were qualitatively different from their wild-type litter mates (5). For example, high levels of V1aR binding were observed in the olfactory bulb, thalamus and cingulate cortex in both transgenic mice and prairie voles, but not in wild-type mice. In addition, when injected with vasopressin, the transgenic mice responded with increased af®liative behaviour, much like prairie voles, whereas the wild-type mice had no changes in social behaviour, much like the nonmonogamous montane voles (5). It is logical to attribute the af®liative response to vasopressin to the similarity in V1aR pattern in the brain between prairie voles and the transgenic mice. However, two caveats must be noted. First, the transgenic mice did not display elevated V1aR binding in some of the brain regions that were hypothesized to be critical for pairbond formation in prairie voles, such as the medial amygdala and ventral pallidum. Second, these mice do not display partner preference, as prairie voles do, suggesting that there are other factors, other genes and/or environmental cues that may mediate the distribution of V1aR and behaviour. Nevertheless, it is intriguing to see that species differences in the regulatory region of the V1aR gene can have such a large impact on V1aR distribution in the brain and af®liative behaviours. The results from this experiment were consistent with our hypothesis that regulatory regions of the V1aR gene, which includes the microsatellite, could direct the V1aR expression pattern throughout the brain, which has a direct behavioural consequence. Further evidence of the role of the species-speci®c microsatellite in generating speci®c molecular and behavioural phenotypes comes from within the prairie vole species itself. The V1aR microsatellite region of the prairie vole is variable in length between individuals of the species. The range of microsatellite length is approximately Æ50 bp from the median length (unpublished observations). In addition, there is striking variability of the
# 2004 Blackwell Publishing Ltd, Journal of Neuroendocrinology, 16, 325±332

distribution of V1aR in the brain of prairie voles across individuals (Fig. 2) (8, 9). There is also signi®cant covariation of V1aR distribution across some brain regions, suggesting a shared mechanism of regulation in those regions (8). We have hypothesized that this intra-speci®c variability in microsatellite length correlates with brain V1aR density in certain brain regions, and these indices might correlate with individual differences in social behaviour, such as length or duration of pairbonding (9). Studies are currently underway to assess the relationships between microsatellite length, gene expression and behaviour. Preliminary studies suggest that there are multiple relationships between microsatellite length and V1aR expression, V1aR expression and behaviour, and microsatellite length and behaviour (10). It is now clear that speciesspeci®c changes in the V1aR microsatellite are capable of modifying gene expression. Using transcription reporter assays, we have observed a functional role for the species-divergent microsatellite in a cell-type dependent manner (11). Additionally, further studies performed in vitro are being used to examine the effects of intra-speci®c variability of allele lengths. In sum, we hypothesize that length variation at this microsatellite within a population produces variation in receptor distribution patterns, which then produces variable, and therefore selectable, behavioural traits. Given the vast variability in V1aR distribution across species, it is reasonable to predict that there might be relationships between the V1aR gene structure in humans and social behaviour. Thibonnier et al. (12) have shown that the human V1aR gene (AVPR1a) also contains a microsatellite in the 5H regulatory region, and that there are multiple alleles of varying length within the human population. There is also an association between a speci®c microsatellite allele in AVPR1a and autism (13). While autism is at the extreme end of the social behaviour spectrum, other minor allele variations might also contribute to more subtle individual differences in social behaviour in humans. In general, the lack of conservation of regulatory regions of the V1aR gene across mammalian species suggests an interesting model for the evolution of species-typical social behaviours, such as monogamy and promiscuity. Possibly, mutations in regulatory regions of genes involved in behaviour are prone to positive selection and the gene products are consequently modi®ed in their temporo-spatial distribution pattern, which undoubtedly alters their functional capacity within the organism. Such a `modular behavioural gene', where the coding region is conserved but the regulatory regions are not, could result in rapid phenotypic changes, leading directly to divergent behavioural phenotypes and allowing for rapid adaptation to any given ecological niche.
Neural substrates of the pairbond

Previous work suggests that V1aR distribution can be dramatically affected by a single gene polymorphism. However, it is unlikely that all of the brain regions that express V1aR are critical for pairbonding; some may function in other behaviours, or localize in redundant or nonfunctional areas whose expression and behavioural effects were merely not selected against. Neuroanatomical mapping of vasopressin shows a remarkably conserved pattern across species, with many instances of regions that do not `match' with V1aR expression in the brain, which is highly diverse across taxa (1, 4, 14, 15). Therefore, it is a challenge to determine which brain regions are speci®cally involved in pairbond formation in

328 V1aR and social attachment in voles

Fig. 2. Examples of individual differences in V1aR distribution among prairie voles. At the level of the olfactory bulbs (A,B), variability exists in the density of V1aR in the internal plexiform and/or mitral layers (plex) as well as the accessory olfactory bulb (AOB). More caudally (C,D), variability exists in the cingulate cortex (cing), various nuclei of the thalamus (thal), as well as the cortex (ctx). In the hindbrain (E,F), V1aR variability is observed in the lateral geniculate nucleus of the thalamus (LGN thal). Studies are currently in progress to examine the relationship of brain V1aR distribution to behaviour, as well as microsatellite genotype. Sections were taken from age-matched adult male prairie voles, although similar diversity exists in female voles. Scale bar ˆ 1 mm.

prairie voles. It might be expected that the critical brain regions contain both vasopressin ®bres and V1aR. It also might be predicted that the critical brain regions in pairbonding are those that differentially express V1aR between prairie and montane voles. Based on these premises, what candidate brain regions could be involved in pairbond formation? We have performed several experiments using Fos induction as a marker of neural activation to identify candidate brain regions during mating and pairbond formation in prairie voles. Signi®cant Fos induction was observed

in several limbic brain areas including the ventral pallidum, nucleus accumbens, bed nucleus of the stria terminalis (BNST), medial preoptic area, medial amygdala, and mediodorsal thalamus (16). All these regions, except the nucleus accumbens, express V1aR in prairie voles and could therefore be a potential site for vasopressin to facilitate pairbond formation. Of note, the medial amygdala and the BNST contain vasopressin-producing cell bodies that project vasopressin ®bres to several brain regions (17). Because the medial amygdala, BNST and medial preoptic
# 2004 Blackwell Publishing Ltd, Journal of Neuroendocrinology, 16, 325±332

V1aR and social attachment in voles 329 area comprise a well-characterized neural circuit for sexual behaviour in many other species (18±21), the Fos induction observed in prairie voles is probably a re¯ection of sociosexual, and not necessarily social attachment, behaviour. By contrast, the ventral forebrain (ventral pallidum and nucleus accumbens) activation may be more related to pairbonding behaviour. In particular, the ventral pallidum is a region that highly expresses V1aR in prairie voles, but not montane voles (Fig. 1A,C) (4). The ventral pallidum also contains vasopressin ®bre immunoreactivity, consistent with vasopressin release in the region to activate V1aR (22). Other more distantly related vole species, such as the monogamous pine vole (Microtus pinetorum) and the non-monogamous meadow vole (Microtus pennsylvanicus), show a similar species difference in V1aR distribution in the ventral pallidum (4). V1aR are also highly expressed in the ventral forebrain in other monogamous species of mammals. For example, the monogamous California mouse (Peromyscus californicus) and the monogamous marmoset monkey both have a high density of V1aR in the ventral pallidal region, whereas the closely related non-monogamous Peromyscus leucopus and the non-monogamous rhesus monkey both have low levels of V1aR binding there (23). This is consistent evidence that V1aR in the ventral pallidum are associated with monogamous social structure across taxa, and thus may represent a potential mechanism for the convergent evolution of the pairbond. To test the hypothesis that Fos induction observed in the ventral pallidum was associated with vasopressinergic neurotransmission, we examined Fos induction in animals in which we speci®cally manipulated V1aR density. We used an adeno-associated viral (AAV) vector gene transfer to increase V1aR expression in the ventral pallidum of prairie voles. AAV vectors are an ef®cient means by which gene expression can be manipulated in animals. A parvovirus, AAV typically infects cells (including neurones) and inserts its own DNA into the genome of the host cell. By deleting the native AAV genes and replacing them with our gene of interest, it is possible to create site-speci®c, transgenic expression in the infected area (24). We constructed an AAV vector containing the prairie vole V1aR gene sequence downstream of a neuronespeci®c enolase promoter, which directs expression in all neurones (AAV-V1a). We injected the AAV-V1a unilaterally in the ventral pallidum of male prairie voles and allowed them to cohabitate and mate with another female, simulating the initial steps in pairbond formation. A similar injection of a LacZ control AAV vector was performed on the contralateral side. We hypothesized that asymmetric Fos induction in the ventral pallidum would re¯ect vasopressinergic neurotransmission and signify that V1aR activation in this region occurred during the pairbonding process. Our results demonstrate that Fos induction was in fact signi®cantly increased on the side of the ventral pallidum that received the AAV-V1a injection (16). This is consistent with our hypothesis that vasopressinergic neurotransmission in the ventral pallidum might be involved in pairbond formation. There is also more direct experimental evidence demonstrating that these reward circuits in the ventral forebrain are involved in pairbonding. Another experiment using AAV-V1a vector gene transfer examined the behavioural effects of over-expression of V1aR bilaterally in the ventral pallidum of male prairie voles (25). This resulted in approximately a two-fold increase in the amount of V1aR expressed in neurones in the region, which persists for at least 4 months postsurgery. These animals with arti®cially
# 2004 Blackwell Publishing Ltd, Journal of Neuroendocrinology, 16, 325±332

elevated V1aR in the ventral pallidum showed increased af®liative behaviour, as measured by olfactory investigation and huddling, compared to control animals injected into the caudate-putamen. Furthermore, male prairie voles with increased V1aR in the ventral pallidum, but not the caudate-putamen, developed partner preferences with an abbreviated period of cohabitation without mating (Fig. 3) (25). Thus it appears that increased V1aR activation in the ventral pallidum leads to increased social contact, as well as accelerated pairbond formation. Recent work in our laboratory has used site-speci®c pharmacological blockade of V1aR in the ventral pallidum to demonstrate the necessity of these receptors in pairbond formation. Male prairie voles were injected bilaterally into the ventral pallidum with either arti®cial cerebrospinal ¯uid or the selective V1aR antagonist, d(CH2)5[Tyr(Me)]AVP (0.05 ng/ml, 1 ml per side). We injected the same dose intracerebroventricularly (i.c.v) as a control, to show that this dose was not effective i.c.v., and was in fact site-speci®c. Animals that received antagonist injections into the ventral pallidum did not show a partner preference for their mate, as opposed to the animals that received vehicle or the antagonist i.c.v (16). By contrast, animals that received antagonist injections into the medial amygdala and the mediodorsal thalamus showed normal partner preference formation (16). This experiment con®rms the critical role of V1aR in the ventral pallidum for the development of the pairbond.
Neural mechanisms of the pairbond

What is the functional signi®cance of the involvement of the ventral pallidum in pairbond formation? The ventral pallidum is a key brain region in the neural circuitry of reward and addiction in the mesolimbic dopamine reward pathway. The ventral pallidum has topographically organized connections to limbic and motor brain regions, via the medial and lateral subdivisions (26). For example, the medial ventral pallidum mainly projects to the ventral tegmental area and mediodorsal thalamus, and receives mostly GABAergic projections from the shell of the nucleus accumbens (27, 28). In addition, the neurochemistry of the medial ventral pallidum is more similar to the BNST, medial preoptic area, and lateral hypothalamus than it is to the globus pallidus, and thus is considered by some as part of the `extended amygdala' (26, 28). Thus, the medial ventral pallidum is thought to be more involved in limbic aspects of motivation and reward. The clear topographic organization of the ventral pallidum coupled with dual functions in reward and motor output have led many to conclude that the ventral pallidum is critically involved in reinforcement-driven motor behaviours (29). In support of this, we have observed a thin, discrete band of V1aR mRNA in the medial ventral pallidum using in situ hybridization, suggesting that cell bodies in the medial subdivision synthesize V1aR and therefore are sensitive to vasopressinergic neurotransmission (22). This is consistent with our hypothesis that V1aR in the ventral pallidum might mediate the limbic or motivational aspects of pairbond formation. The extensive projections of the ventral pallidum with the mesolimbic dopamine pathway play a critical role in mediating the rewarding effects from drugs of abuse. For example, cocaine self-administration into the ventral pallidum leads to an increase in extracellular dopamine concentration in the ventral pallidum (30). In addition, infusion of psychostimulants directly into the ventral pallidum leads to the development of a conditioned place

330 V1aR and social attachment in voles

Fig. 3. Viral vector-mediated over-expression of V1aR in the ventral forebrain facilitates af®liative behaviour. (A) Endogenous V1aR expression is dense in the ventral pallidum (VP), but not the caudate-putamen (CP). (C) Viral vector-mediated over-expression of V1aR in the ventral forebrain. (B) Juvenile af®liation is enhanced in animals over-expressing V1aR in the ventral pallidum (V1aR-VP) compared to control animals (LacZ-VP/CP and V1aR-CP) (P < 0.05, Student's ttest). (D) Partner preference formation is facilitated in animals over-expressing V1aR in the ventral pallidum (V1aR-VP), whereas control animals did not form pairbonds under the same conditions (LacZ-VP/CP and V1aR-CP) (P < 0.05, Student's t-test).

preference for the environment in which the injections were experienced (30). Depletion of dopamine speci®cally in the ventral pallidum prevents this cocaine-induced place preference (31). Experimental evidence also indicates the role of the ventral pallidum in natural reinforcement because lesions of the ventral pallidum signi®cantly attenuate conditioned place preference to sucrose (32), while electrical stimulation in this region induces Fos expression in other reward regions (33). It is possible that V1aR activation in the ventral pallidum of monogamous voles facilitates pairbonding through this reward circuitry. The ventral pallidum is heavily interconnected with the nucleus accumbens, which is another major region in the reward pathway. Interestingly, a parallel story exists with oxytocin and oxytocin receptors (OTR) in the nucleus accumbens in prairie versus montane voles. Prairie voles have many more OTR in the nucleus accumbens than montane voles, analogous to the density of V1aR observed in the ventral pallidum (34). In addition, parallel to the V1aR antagonist experiments in male prairie voles, OTR antagonist microinjected into the nucleus accumbens blocks partner preference formation in female prairie voles, while oxytocin infused i.c.v. facilitates female partner preference (35, 36). It is noteworthy that OTR in the nucleus accumbens and V1aR in the ventral pallidum are anatomically adjacent to one another in the prairie vole, and that these brain regions are heavily interconnected structures (22). Although oxytocin and vasopressin systems may initially appear sexually dichotomous, with oxytocin affecting females and vasopressin affecting males, it is possible

that these two neuropeptide systems converge upon a common neural circuit at the level of the ventral forebrain, via cross-talk between the nucleus accumbens and ventral pallidum, to produce the same functional outcome, a pairbond. Drug addiction studies have advanced the current view that nucleus accumbens dopamine is critical for the rewarding effects of drugs of abuse, such as psychostimulants and opiates, as well as for natural reinforcers such as food and sex (37). Dopamine release into the nucleus accumbens plays a role in the anticipatory and arousing aspects of male rat sexual behaviour (38). In prairie voles, similar studies have shown dopamine release in the nucleus accumbens during mating using microdialysis in males, as well as increased dopamine metabolite turnover using tissue punches of the nucleus accumbens in females (39, 40). Recent work has found that both OTR and dopamine are necessary for pairbond formation (41). It is likely that the functional signi®cance of OTR in the nucleus accumbens and V1aR in the ventral pallidum in monogamous voles results from their association with this neural circuit for natural reinforcers. What cognitive mechanisms might occur during pairbond formation? Both oxytocin and vasopressin are involved in learning and memory processes, and both neuropeptides play a critical role in the formation of social memories in rats and mice (42, 43). Transgenic mice lacking oxytocin (44) and Brattleboro rats lacking vasopressin both lack the ability to form social memories (45). Mutant mice lacking V1aR receptors also show the same phenotype (46). Central OTR antagonism blocks social memory in
# 2004 Blackwell Publishing Ltd, Journal of Neuroendocrinology, 16, 325±332

V1aR and social attachment in voles 331

Fig. 4. A proposed neural circuit for pairbonding in male prairie voles. (1) Social olfactory information from the accessory olfactory bulb (AOB) is relayed to the medial amygdala (MeA). (2) The MeA sends vasopressinergic projections to the ventral pallidum (VP). (3) The VP is heavily interconnected with another reward nucleus, the nucleus accumbens (NAcc). (4 and 5) The VP also projects to other regions in the reward pathway, such as the mediodorsal thalamus (MDThal) and the ventral tegmental area (VTA). The convergence of inputs onto the MeA, where social memories are formed, and the VP, which activates reward circuitry, reinforces the association between the partner and reward.

wild-type mice, while V1aR blockade prevents social recognition in normal rats (44, 45). One hypothesis to explain the pharmacological effects of oxytocin and vasopressin on partner preference in the prairie vole is that oxytocin and vasopressin may facilitate pairbonding by improving social recognition, while OTR antagonist might induce social amnesia. However, this hypothesis still does not explain why prairie voles form monogamous relationships whereas mice and rats do not. One of the cognitive mechanisms underlying pairbond formation in monogamous species may involve the formation of a learned association between the memory of the partner and reward. The conditioned stimulus, the partner, is paired with the unconditioned stimulus, mating, to result in a conditioned response of the prairie vole seeking to preferentially be with its mate. Overnight cohabitation and mating in prairie voles is inferred to release oxytocin and vasopressin in the brain, which could potentially modulate social memory circuits to imprint the mate with a speci®c `pheromonal signature' (47). Additionally, dopamine neurotransmission occurs in the ventral forebrain during mating, and this is necessary for partner preference formation in prairie voles (39, 40). It is possible that the neuronal inputs from reward circuitry and the inputs from social memory both converge at the level of the ventral forebrain, to create a conditioned `partner' preference, analogous to conditioned place preference. Speci®c examples of the neural projections in this proposed circuit have been identi®ed. For example, during courtship, olfactory information is processed by the medial amygdala, which is critical for social memory. The medial amygdala has projections to the nucleus accumbens (48), and also sends vasopressinergic projections to the ventral pallidum (49). The nucleus accumbens and ventral pallidum are reciprocally connected and project to other limbic regions involved in reward, such as the ventral tegmental area and the mediodorsal thalamus (Fig. 4) (28, 50, 51). The scarcity of OTR and V1aR in the ventral forebrain of nonmonogamous species may explain their inability to form partner preferences after mating. Conceivably, the integration of two basic neural circuits, one for reward and the other for social memory, could provide the proximate mechanism for a complex social behaviour such as pairbonding, and permit the evolution of a monogamous species.
# 2004 Blackwell Publishing Ltd, Journal of Neuroendocrinology, 16, 325±332

The prairie vole model for studying the role of vasopressin and V1aR in social behaviour has proven useful at both the level of the gene and the brain. In conjuction, both genetic and neural analyses of a complex social behaviour can provide a coherent framework by which to examine the neural control of behaviour and the evolution of species-typical social behaviours. Understanding the molecular and neural substrates of complex behaviours in several species may eventually contribute to the understanding of social behaviour in our own species.
This research was supported by NIH MH65050 to MML, MH67397 to EADH, MH56897 and MH 64692 to LJY, and NSF STC IBN-9876754 and the Yerkes Center Grant RR00165.

1 Goodson JL, Bass AH. Social behavior functions and related anatomical characteristics of vasotocin/vasopressin systems in vertebrates. Brain Res Brain Res Rev 2001; 35: 246±265. 2 Insel TR, Young LJ. The neurobiology of attachment. Nature Rev Neurosci 2001; 2: 129±136. 3 Winslow JT, Hastings N, Carter CS, Harbaugh CR, Insel TR. A role for central vasopressin in pair bonding in monogamous prairie voles. Nature 1993; 365: 545±548. 4 Insel TR, Wang ZX, Ferris CF. Patterns of brain vasopressin receptor distribution associated with social organization in microtine rodents. J Neurosci 1994; 14: 5381±5392. 5 Young LJ, Nilsen R, Waymire KG, MacGregor GR, Insel TR. Increased af®liative response to vasopressin in mice expressing the vasopressin receptor from a monogamous vole. Nature 1999; 400: 766±768. 6 Michelhaugh SK, Fiskerstrand C, Lovejoy E, Bannon MJ, Quinn JP. The dopamine transporter gene (SLC6A3) variable number of tandem repeats domain enhances transcription in dopamine neurons. J Neurochem 2001; 79: 1033±1038. 7 Lesch KP, Bengel D, Heils A, Sabol SZ, Greenberg BD, Petri S, Benjamin J, Muller CR, Hamer DH, Murphy DL. Association of anxiety-related traits with a polymorphism in the serotonin transporter gene regulatory region. Science 1996; 274: 1527±1531. 8 Phelps SM, Young LJ. Extraordinary diversity in vasopressin (V1a) receptor distributions among wild prairie voles (Microtus ochragaster): patterns of variation and covariation. J Comp Neurol 2003; 466: 564±576. 9 Hammock EA, Young LJ. Variation in the vasopressin V1a receptor promoter and expression: implications for inter- and intraspeci®c variation in social behaviour. Eur J Neurosci 2002; 16: 399±402.

332 V1aR and social attachment in voles
10 Hammock EAD, Lim MM, Nair HP, Young LJ. Genetic control of variability in brain vasopressin receptor V1a expression and behavior in prairie voles. Society for Neuroscience, New Orleans, LA, 2003. Abstract Viewer/Itinerary Planner CD-ROM. Washington, DC: Society for Neuroscience, 2003. 11 Hammock EAD, Young LJ. Functional microsatellite polymorphism associated with divergent social structure in vole species. Mol Biol Evol 2004, in press. 12 Thibonnier M, Graves MK, Wagner MS, Chatelain N, Soubrier F, Corvol P, Willard HF, Jeunemaitre X. Study of V(1)-vascular vasopressin receptor gene microsatellite polymorphisms in human essential hypertension. J Mol Cell Cardiol 2000; 32: 557±564. 13 Kim SJ, Young LJ, Gonen D, Veenstra-VanderWeele J, Courchesne R, Courchesne E, Lord C, Leventhal BL, Cook EH Jr, Insel TR. Transmission disequilibrium testing of arginine vasopressin receptor 1A (AVPR1A) polymorphisms in autism. Mol Psychiatry 2002; 7: 503±507. 14 Young LJ, Toloczko D, Insel TR. Localization of vasopressin (V1a) receptor binding and mRNA in the rhesus monkey brain. J Neuroendocrinol 1999; 11: 291±297. 15 Wang Z, Toloczko D, Young LJ, Moody K, Newman JD, Insel TR. Vasopressin in the forebrain of common marmosets (Callithrix jacchus): studies with in situ hybridization, immunocytochemistry and receptor autoradiography. Brain Res 1997; 768: 147±156. 16 Lim MM, Young LJ. Vasopressin-dependent neural circuits underlying pair bonding in the monogamous prairie vole. Neuroscience, in press. 17 Wang Z. Species differences in the vasopressin-immunoreactive pathways in the bed nucleus of the stria terminalis and medial amygdaloid nucleus in prairie voles (Microtus ochrogaster) and meadow voles (Microtus pennsylvanicus). Behav Neurosci 1995; 109: 305±311. 18 Coolen LM, Peters HJ, Veening JG. Distribution of Fos immunoreactivity following mating versus anogenital investigation in the male rat brain. Neuroscience 1997; 77: 1151±1161. 19 Fernandez-Fewell GD, Meredith M. c-fos expression in vomeronasal pathways of mated or pheromone-stimulated male golden hamsters: contributions from vomeronasal sensory input and expression related to mating performance. J Neurosci 1994; 14: 3643±3654. 20 Michael RP, Clancy AN, Zumpe D. Effects of mating on c-fos expression in the brains of male macaques. Physiol Behav 1999; 66: 591±597. 21 Reyna-Neyra A, Camacho-Arroyo I, Cerbon MA, Gonzalez-Mariscal G. Mating modi®es c-fos expression in the brain of male and female rabbits. Neurosci Lett 2000; 284: 1±4. 22 Lim MM, Murphy AZ, Young LJ. Ventral striato-pallidal oxytocin and vasopressin V1a receptors in the monogamous prairie vole (Microtus ochragaster). J Comp Neurol 2004; 468: 555±570. 23 Young LJ, Frank A. Beach Award. Oxytocin and vasopressin receptors and species-typical social behaviors. Horm Behav 1999; 36: 212±221. 24 Nair HP, Young LJ. Application of adeno-associated viral vectors in behavioral research. Methods 2002; 28: 195±202. 25 Pitkow LJ, Sharer CA, Ren X, Insel TR, Terwilliger EF, Young LJ. Facilitation of af®liation and pairbond formation by vasopressin receptor gene transfer into the ventral forebrain of a monogamous vole. J Neurosci 2001; 21: 7392±7396. 26 Zahm DS. The ventral striatopallidal parts of the basal ganglia in the rat À II. Compartmentation of ventral pallidal efferents. Neuroscience 1989; 30: 33±50. 27 Zahm DS, Williams E, Wohltmann C. Ventral striatopallidothalamic projection. IV. Relative involvements of neurochemically distinct subterritories in the ventral pallidum and adjacent parts of the rostroventral forebrain. J Comp Neurol 1996; 364: 340±362. 28 Zahm DS, Heimer L. Two transpallidal pathways originating in the rat nucleus accumbens. J Comp Neurol 1990; 302: 437±446. 29 Napier TC, Mitrovic I, Churchill L, Klitenick MA, Lu XY, Kalivas PW. Substance P in the ventral pallidum: projection from the ventral striatum, and electrophysiological and behavioral consequences of pallidal substance P. Neuroscience 1995; 69: 59±70. 30 Gong WDN, Justice JB Jr. Conditioned place preference and locomotor activation produced by injection of psychostimulants into ventral pallidum. Brain Res 1996; 707: 64±74. 31 Gong WDN, Justice JB Jr. 6-Hydroxydopamine lesion of ventral pallidum blocks acquisition of place preference conditioning to cocaine. Brain Res 1997; 754: 103±112. 32 McAlonan GM, Robbins TW, Everitt BJ. Effects of medial dorsal thalamic and ventral pallidal lesions on the acquisition of a conditioned place preference: further evidence for the involvement of the ventral striatopallidal system in reward-related processes. Neuroscience 1993; 52: 605±620. 33 Panagis G, Nomikos GG, Miliaressis E, Chergui K, Kastellakis A, Svensson TH, Spyraki C. Ventral pallidum self-stimulation induces stimulus dependent increase in c-fos expression in reward-related brain regions. Neuroscience 1997; 77: 175±186. 34 Insel TR, Shapiro LE. Oxytocin receptor distribution re¯ects social organization in monogamous and polygamous voles. Proc Natl Acad Sci USA 1992; 89: 5981±5985. 35 Williams JR, Insel TR, Harbaugh CR, Carter CS. Oxytocin administered centrally facilitates formation of a partner preference in female prairie voles (Microtus ochrogaster). J Neuroendocrinol 1994; 6: 247±250. 36 Young LJ, Lim MM, Gingrich B, Insel TR. Cellular mechanisms of social attachment. Horm Behav 2001; 40: 133±138. 37 Robbins TW, Everitt BJ. Neurobehavioural mechanisms of reward and motivation. Curr Opin Neurobiol 1996; 6: 228±236. 38 Fumero MMB, Gonzalez-Mora JL. Voltammetric and microdialysis monitoring of brain monoamine neurotransmitter release during sociosexual interactions. Behav Brain Res 1995; 71: 69±79. 39 Gingrich B, Liu Y, Cascio C, Wang Z, Insel TR. Dopamine D2 receptors in the nucleus accumbens are important for social attachment in female prairie voles (Microtus ochrogaster). Behav Neurosci 2000; 114: 173±183. 40 Aragona BJ, Liu Y, Curtis JT, Stephan FK, Wang Z. A critical role for nucleus accumbens dopamine in partner-preference formation in male prairie voles. J Neurosci 2003; 23: 3483±3490. 41 Liu Y, Wang. Both dopamine and oxytocin receptors are required for partner preference formation in female prairie vole. Society for Neuroscience, Orlando, FL, 2002. Abstract Viewer/Itinerary Planner CD-ROM. Washington, DC: Society for Neuroscience, 2002. 42 Popik P, van Ree JM. Neurohypophyseal peptides and social recognition in rats. Prog Brain Res 1998; 119: 415±436. 43 Ferguson JN, Young LJ, Insel TR. The neuroendocrine basis of social recognition. Front Neuroendocrinol 2002; 23: 200±224. 44 Ferguson JN, Young LJ, Hearn EF, Matzuk MM, Insel TR, Winslow JT. Social amnesia in mice lacking the oxytocin gene. Nature Genet 2000; 25: 284±288. 45 Engelmann M, Wotjak CT, Neumann I, Ludwig M, Landgraf R. Behavioral consequences of intracerebral vasopressin and oxytocin: focus on learning and memory. Neurosci Biobehav Rev 1996; 20: 341±358. 46 Bielsky IF, Bao-Hu S, Szegda KL, Westphal H, Young LJ. Profound impairment in social recognition and reduction in anxiety in vasopressin V1a receptor knockout mice. Neuropsychopharmacology, in press. 47 Wang Z, Young LJ, De Vries GJ, Insel TR. Voles and vasopressin: a review of molecular, cellular, and behavioral studies of pair bonding and paternal behaviors. Prog Brain Res 1998; 119: 483±499. 48 Coolen LM, Wood RI. Bidirectional connections of the medial amygdaloid nucleus in the Syrian hamster brain: simultaneous anterograde and retrograde tract tracing. J Comp Neurol 1998; 399: 189±209. 49 de Vries GJ, Miller MA. Anatomy and function of extrahypothalamic vasopressin systems in the brain. Prog Brain Res 1998; 119: 3±20. 50 Mogenson GJ, Ciriello J, Garland J, Wu M. Ventral pallidum projections to mediodorsal nucleus of the thalamus: an anatomical and electrophysiological investigation in the rat. Brain Res 1987; 404: 221±230. 51 Ungerstedt U. Stereotaxic mapping of the monoamine pathways in the rat brain. Acta Physiol Scand (Suppl) 1971; 367: 1±48.

# 2004 Blackwell Publishing Ltd, Journal of Neuroendocrinology, 16, 325±332

Shared By: