Robust age, but limited sex, differences in mu-opioid receptors in the rat brain: relevance for reward and drug-seeking behaviors in juveniles

Male and female Wistar rats were obtained from Charles River Laboratories (Raleigh, NC) at 22 or 56 days of age and housed under standard laboratory conditions (12-h light/dark cycle, lights on at 7:00 am, food and water available ad libitum, 22 °C, 60% humidity). Upon arrival at our facility, rats were housed in standard rat cages (26.7 × 48.3 × 20.3 cm). Twenty-two-day-old rats were housed in same-sex groups of 3–4 until brain collection for receptor autoradiography at 35 days of age (juvenile group). Fifty-six-day-old rats were housed in same-sex pairs until brain collection for receptor autoradiography at 84 days of age (adult group). All experiments were conducted in accordance with the NIH Guide to the Care and Use of Laboratory Animals and approved by the Boston College Institutional Animal Care and Use Committee (IACUC).

Rats (juvenile males: n = 13; juvenile females: n = 13; adult males: n = 12; adult females: n = 12) were euthanized using CO₂ inhalation and brains were removed, rapidly frozen in methylbutane on dry ice, and stored at −45 °C. Brains were cut on a cryostat into 16-µm coronal sections and mounted onto slides in eight adjacent series. Collection began at approximately 3.72 mm anterior to bregma and ended at approximately 8.52 mm posterior to bregma (Paxinos and Watson 2007). Sections were then frozen −45 °C until receptor autoradiography was performed. MOR autoradiography was conducted using the MOR-specific agonist [³H]D-Ala²-MePhe⁴-Gly-ol⁵ enkephalin (DAMGO; Perkin Elmer, Boston, MA) as tracer. In brief, slides were thawed and air-dried at room temperature, followed by pre-incubation for 30 min in 50 nM Tris–HCl (pH 7.4) containing 0.9% NaCl. The slides were then exposed to tracer buffer (4 nM [³H]D-Ala²-MePhe⁴-Gly-ol⁵ enkephalin and 50 mM Tris) for 60 min. Non-specific binding was assessed in adjacent brain sections, by incubation in tracer buffer with the addition of 1 µM of the MOR antagonist naloxone (Sigma-Aldrich, St. Louis, MO). All slides were then washed three times, for 5 min each, in ice-cold Tris–HCl, air-dried, and exposed to Biomax MR films (VWR International, Pittsburgh, PA) for 16 weeks. Brain sections of juvenile and adult male and female rats were processed together and balanced across incubation chambers and exposure to films.

Autoradiography films were digitized using a Northern Light Illuminator (InterFocus Imaging, Cambridge, UK) and optical densities of MOR binding were measured in coronal sections using ImageJ (NIH, http://​imagej.​nih.​gov/​ij/​). The data were converted to dpm/mg tissue (disintegrations per minute/milligram tissue) using a [³H] standard microscale (American Radiolabeled Chemicals Inc., St. Louis, MO). Because non-specific binding was undetectable (Fig. 1), film background values were subtracted from total binding values to yield specific binding values. Binding densities were calculated by taking the mean of bilateral measurements in a fixed number of sections per region of interest per rat. The total number of measurements depended on the size of the region of interest and ranged from 4 to 13. MOR binding density was measured in a total of 33 brain regions (see Fig. 2 for receptor autoradiograms and schematic diagrams indicating the brain regions in which MOR binding was quantified). All abbreviations of brain regions are in accordance with Paxinos and Watson (2007), except for the nucleus accumbens core and nucleus accumbens shell, where we added the subdivisions anterior core (aAcbC), anterior shell (aAcbSh), dorsomedial shell (dmAcbSh) and ventral shell (vAcbSh) to delineate the separate areas analyzed, as well as for the laterodorsal thalamic nucleus where we used the abbreviation LDTN to refer to the dorsomedial and ventrolateral parts of the nucleus combined, the lateral posterior thalamic nucleus where we used the abbreviation LPTN to refer to the mediorostral and laterorostral parts combined, and the molecular layer of the dentate gyrus where we used the abbreviation moDGp to refer to the more posterior part of the region.

For all statistical analysis, PASW/SPSS Statistics (Version 22.0) was used. A one-way ANOVA followed by Bonferroni post-hoc testing was used to compare MOR binding density across all brain regions analyzed and collapsed across age and sex. Two-way ANOVAs were used to test for age and sex differences in MOR binding density in each brain region. The false discovery rate (FDR) procedure (Hochberg and Benjamini 1990) was used to correct for multiple comparisons (age, sex, and interaction). This resulted in an FDR α < 0.0130 (based on 99 comparisons). Significant interaction effects were followed by Bonferroni post hoc tests (reflecting t tests pre-adjusted for multiple comparisons) to examine differences among groups. Significant age or sex effects were followed by Cohen’s D to calculate the effect size of age differences (overall and separately for males and females) and of sex differences (overall and separately for juveniles and adults). A subsequent independent samples t test was run to determine whether the effect size of age differences was different between males and females for all brain regions. Significance for independent samples t tests was set at p < 0.05.

MOR binding density varied greatly by brain region [F _(32,1527) = 260.8; p < 0.001; Fig. 3], with an approximate tenfold difference between the highest and lowest MOR binding density within the 33 regions that were analyzed. Binding density was highest in the apical subnucleus of the interpeduncular nucleus (IPA) and stria medullaris of the thalamus (sm) and lowest in the lateral periaqueductal gray (LPG), CA1 region of the hippocampus (CA1), and lateral septum (LS) (Fig. 3). Notably, MOR binding densities did not predict where age, sex, and age × sex interaction effects were found.

Age differences in MOR binding density were found in 12 of the 33 brain regions analyzed (see Table 1 for complete statistics). MOR binding density was higher in juveniles compared to adults in 11 brain regions: 4 telencephalic brain regions [LS, the CA1 and CA2/3 sub-regions of the hippocampus, and the posterior bed nucleus of the stria terminalis (BNSTp; Fig. 4a)], 6 diencephalic brain regions [the dorsal lateral geniculate nucleus (DLG), the LPTN, the posterior thalamic nuclear group (Po), the paraventricular thalamic nucleus (PV), the reticular thalamic nucleus (Rt), and the nucleus reuniens (Re); Fig. 5a] and the caudal subnucleus of the interpeduncular nucleus in the mesencephalon (IPC; Fig. 5b). MOR binding density was higher in adults than juveniles only in the moDGp (Fig. 4b). Finally, the effect sizes of age differences were similar between the sexes [t _(1,22) = −1.21; p = 0.24; Fig. 6a, b].

Sex differences in MOR binding density were observed in 2 of the 33 brain regions analyzed (see Table 1 for complete statistics). In detail, MOR binding density was higher in females than in males in the posteromedial cortical amygdaloid nucleus (PMCo; Fig. 7a) with the effect size being larger in adults than in juveniles (Fig. 7c). In contrast, MOR binding density was higher in males than in females in the LS (Fig. 7b) with the effect size being larger in juveniles than in adults (Fig. 7c).

Significant age × sex interaction effects on MOR binding density were found in two brain regions (see Table 1 for complete statistics). In detail, in the lateral subnucleus of the interpeduncular nucleus (IPL), MOR binding density was higher in juveniles compared to adults in males (p < 0.001), but not in females (p = 0.534; Fig. 8a). In the Po, adult males had significantly lower MOR binding density than adult females (p < 0.05), while there was no sex difference in juveniles (p = 0.112; Fig. 8b).

Despite robust MOR binding, no age or sex differences were found in 19 of the 33 brain regions analyzed, including 10 telencephalic regions, 3 diencephalic regions, and 6 mesencephalic regions (see Table 1 for statistics and list of brain regions).

MOR binding density varied substantially across regions of the rat brain, independent of age and sex. In general, MOR binding was observed in the same brain regions as previously reported in the adult male rat using MOR autoradiography (Mansour et al. 1986, 1987; Temple and Zukin 1987; McLean et al. 1986; Mansour et al. 1994) or MOR immunohistochemistry (Ding et al. 1996), with the latter suggesting that MOR binding density resembles MOR protein expression. Moreover, the patterns of MOR binding density in the rat brain are largely consistent with those reported in other mammalian species (Daunais et al. 2001; Hurd and Herkenham 1993; Voorn et al. 1996; Ragen et al. 2015a, b). In particular, dense MOR binding in striatal and amygdala sub-regions has been observed in species ranging from rats, voles, and guinea pigs to non-human primates and humans (Daunais et al. 2001; Hurd and Herkenham 1993; Voorn et al. 1996; Ragen et al. 2015a, b; Inoue et al. 2013; Resendez et al. 2013; Sharif and Hughes 1989). Furthermore, the notable absence of MOR binding in the central nucleus of the amygdala observed in rats is consistent with reports in titi monkeys and macaques (Daunais et al. 2001; Ragen et al. 2015a, b). Even patterns of MOR binding density within individual brain regions appear to be consistent across species in some cases. For example, here we show that in rats MOR binding density is higher in the anterior than in medial/posterior portions of the caudate putamen (CPu) and MOR binding density is lower in the ventral part compared to the dorsomedial and anterior parts of the nucleus accumbens shell. This pattern is in line with MOR binding density in the CPu in prairie voles, meadow voles and macaques (Resendez et al. 2013; Daunais et al. 2001) and in the nucleus accumbens shell in prairie and meadow voles (Resendez et al. 2013). Given the assumption that differences in MOR binding density may reflect differences in MOR activation, the consistency of these MOR binding density patterns across species could indicate that the functions of MORs in these sub-regions are conserved. Indeed, the neural circuitry underlying pleasure and reward (including the nucleus accumbens and CPu) is highly evolutionarily conserved, and evidence for the involvement of MORs in pleasure and reward can be found in all of the above-mentioned species (Berridge and Kringelbach; 2015; Resendez et al. 2013; Trezza et al. 2011; Hsu et al. 2013; Ragen et al. 2015b; Barr et al. 2010).

Age differences in MOR binding density were observed in 12 out of 33 brain regions analyzed, including the lateral septum, as well as sub-regions of the hippocampus, BNST, thalamus, and interpeduncular nucleus. Importantly, the direction of these age differences was largely uniform with denser MOR binding in juveniles than adults. Moreover, these age differences were found in both sexes. Previous studies in rats have shown that, in many brain regions, MOR binding density is highest around postnatal day 12 compared to earlier ages and to adulthood (Recht et al. 1985; Kornblum et al. 1987; Moon Edley and Herkenham 1984; Spain et al. 1985). The decline in MOR binding density after postnatal day 12 has been suggested to be the result of increased synaptic pruning in the third and fourth weeks of postnatal life in the rat (Kornblum et al. 1987). However, synaptic elimination is a developmental process that continues well into the pubertal period (Andersen et al. 2000; Huttenlocher and Dabholkar 1997; Geröcs et al. 1986). Therefore, it is plausible that the brain regions in which we observed higher MOR binding densities in juvenile than in adult rats are those in which synaptic pruning is not complete until later in development. If so, one might expect to see a similar decline in the binding density of other types of receptors in the same brain regions. Yet, oxytocin receptor binding density in the BNSTp has been found to be higher in adult compared to juvenile rats (Smith et al. 2017). Furthermore, dopamine D1, D2, and D4 receptor binding has been shown to be lower in adult compared to juvenile rats in the nucleus accumbens (Tarazi and Baldessarini 2000), while we did not observe an age difference in MOR binding density in this region. Therefore, if pruning is causing a decline in receptors, these findings suggest that it may be specific to synapses expressing only certain types of receptors and not others.

Irrespective of the underlying cause of the age-dependent decline in MOR binding density, it is likely that higher MOR binding density in juveniles allows for higher MOR activation, which may have relevance to the facilitation of juvenile-specific behaviors. One region of particular interest is the lateral septum, because it shows the most robust age difference in MOR binding density, is reciprocally connected to the mesolimbic reward system (Swanson 1982), and it is involved in social play (Veenema et al. 2013; Bredewold et al. 2014, 2015), a highly rewarding and juvenile-specific behavior (Vanderschuren et al. 2016). Surprisingly, the role of MORs in the lateral septum in the regulation of social or non-social rewarding behavior has not been studied. However, MORs in the lateral septum have been implicated in the regulation of anxiety. Specifically, MOR activation increases anxiety-related behavior in adult mice (Le Merrer et al. 2006). This corresponds to the overall role of the lateral septum, as activation of this region results in anxiogenic effects (Anthony et al. 2014; Veening et al. 2009). It would, therefore, be interesting to determine whether higher MOR binding density in the lateral septum of juvenile versus adult rats has implications for the age-specific regulation of anxiety and rewarding social behaviors.

Age differences in MOR binding density were particularly evident in the thalamus, with six out of nine sub-regions showing higher MOR binding density in juvenile compared to adult rats. Although the functional role of MORs in these thalamic sub-regions is unknown, interesting and testable hypotheses can be generated based on the function of each of these sub-regions. For example, the nucleus reuniens receives input from the prefrontal cortex and relays it to the hippocampus (Ito et al. 2015; Hallock et al. 2016). Disruption of this pathway by blockade of the reuniens impairs spatial navigation, learning, and memory (Ito et al. 2015; Davoodi et al. 2009). Because MOR activation has an overall inhibitory effect on the thalamus (Brunton and Charpak 1998; Nakahama et al. 1981; Benoist et al. 1986), it is possible that denser MOR binding in the nucleus reuniens allows for higher MOR activation, which, in turn may mediate a greater inhibition of this circuit in juveniles than adults. Furthermore, the paraventricular nucleus of the thalamus mediates the aversive effects of opiate withdrawal through connections with the nucleus accumbens (Zhu et al. 2016). Because juveniles are less sensitive to the aversive effects of withdrawal than adults (Doremus-Fitzwater and Spear 2007; Hodgson et al. 2010) it is plausible that denser MOR binding allows for higher MOR activation in the paraventricular thalamic nucleus, which in turn may inhibit signaling in this pathway, resulting in reduced withdrawal symptoms in juveniles. These hypotheses will need to be tested in future studies.

Finally, age differences in MOR binding density were found in the caudal sub-region (higher in juveniles of both sexes) and in the lateral sub-region (higher in juveniles, but only in males) of the interpeduncular nucleus. This nucleus is densely interconnected with the lateral habenula (Sutherland 1982). This habenulo-interpeduncular pathway exerts a chronic inhibitory influence over the mesolimbic reward pathway (Nishikawa et al. 1986). In fact, it has been suggested that these two pathways jointly regulate the rewarding properties of drugs (Ellison 1994). Given the increased susceptibility to drug seeking behavior during the juvenile period (Spear 2000), it would be of interest to determine whether higher MOR binding density in the interpeduncular nucleus leads to higher susceptibility to MOR activation within the habenulo-interpeduncular pathway in juveniles versus adults. This, in turn, might result in higher susceptibility to drug-seeking behaviors in juveniles.

Sex differences in MOR binding density were found in the lateral septum (higher in males) and posterior cortical nucleus of the amygdala (higher in females). These sex differences were already present in juveniles, suggesting a pre-pubertal age of onset. Interestingly, both brain regions have sexually dimorphic features. The intermediate lateral septum, in which MOR binding was measured, contains more cells in females than in males (Segovia et al. 2009), while the posterior cortical nucleus of the amygdala contains more cells in males than in females (Vinader-Caerols et al. 1998). Thus, the direction of the sex difference in cell number is opposite to the direction of the sex difference in MOR binding density in both regions. Interestingly, the sex difference in posterior cortical nucleus volume and cell number is present prior to puberty (Akhmadeev and Kalimullina 2014) and in gonadectomized rats (Vathy et al. 2003), suggesting that this sex difference does not depend on circulating gonadal hormones.

Further research is required to determine the functional implications of the sex differences in MOR binding density in these two brain regions. The lateral septum is implicated in the regulation of emotional, motivational, and social behaviors (Sheehan et al. 2004). As such, higher MOR binding density in the lateral septum of males versus females may allow for higher MOR activation in males, which, in turn, may be involved in the regulation of any of these behaviors in a sex-specific way. The posterior cortical nucleus of the amygdala receives olfactory information from the main and accessory olfactory bulbs (Scalia and Winans 1975; Kevetter and Winans 1981), as well as the medial amygdala (DiBenedicitis et al. 2014) and mediates olfactory-guided social behaviors such as sexual behavior (Maras and Petrulis 2008). It would, therefore, be interesting to determine whether MORs in this brain region are involved in the sex-specific regulation of olfactory-guided social behaviors.

In contrast to our hypothesis, MOR binding density was similar between males and females in the vast majority of brain regions analyzed. A limitation of the current study is that we did not measure the effect of estrous phase on MOR binding density. However, MOR binding density variability (as interpreted by the average standard deviation of binding density in each brain region) was no greater in females than in males, suggesting that it is unlikely that estrous phase had a large impact on the absence of sex differences in MOR binding density. This suggests that the sex differences in opioid sensitivity and response in rats (Lynch and Carroll 1999; Cicero et al. 2003) are less likely to be due to sex differences in MOR binding. However, it is possible that sex differences occur in the downstream signaling pathways of the MOR. In support, estrogens have been shown to inhibit MOR-mediated signaling via a protein kinase A pathway (Wagner et al. 1998). Further work is needed to determine whether other aspects of the MOR system show sex differences and if so, whether these underlie the observed sex differences in MOR-mediated addictive behaviors.