Oxytocin receptor activation does not mediate associative fear deficits in a Williams Syndrome model

Abstract Williams Syndrome results in distinct behavioral phenotypes, which include learning deficits, anxiety, increased phobias and hypersociability. While the underlying mechanisms driving this subset of phenotypes is unknown, oxytocin (OT) dysregulation is hypothesized to be involved as some studies have shown elevated blood OT and altered OT receptor expression in patients. A “Complete Deletion” (CD) mouse, modeling the hemizygous deletion in Williams Syndrome, recapitulates many of the phenotypes present in humans. These CD mice also exhibit impaired fear responses in the conditioned fear task. Here, we address whether OT dysregulation is responsible for this impaired associative fear memory response. We show direct delivery of an OT receptor antagonist to the central nervous system did not rescue the attenuated contextual or cued fear memory responses in CD mice. Thus, increased OT signaling is not acutely responsible for this phenotype. We also evaluated OT receptor and serotonin transporter availability in regions related to fear learning, memory and sociability using autoradiography in wild type and CD mice. While no differences withstood correction, we identified regions that may warrant further investigation. There was a nonsignificant decrease in OT receptor expression in the lateral septal nucleus and nonsignificant lowered serotonin transporter availability in the striatum and orbitofrontal cortex. Together, these data suggest the fear conditioning anomalies in the Williams Syndrome mouse model are independent of any alterations in the oxytocinergic system caused by deletion of the Williams locus.

processing. [3][4][5][6][7][8][9][10][11][12] Interestingly, however, individuals with WS possess relatively intact expressive language and verbal skills, 13,14 as well as heightened sensitivity and emotional response to music. 2,15 One of the most striking phenotypes of individuals with WS is hypersociability and strong social motivation, [16][17][18] despite high non-social anxiety 19 and deficits in social cognition and awareness. 20 A substantial body of research indicates the neuropeptide oxytocin (OT) plays a key role in mediating the regulation of social behavior and cognition, fear conditioning and extinction, observational fear, 21 fear modulation via social memory 22 and anxiety in humans and rodents. [23][24][25] Given the aberrant social behavior and anxiety in individuals with WS, recent studies have tested the hypothesis that OT is dysregulated in WS. Indeed, one study found elevated blood levels of OT in individuals with WS compared with controls. 26 However, the findings on the OT receptor (OXTR) have been contradictory. One study suggested increased gene expression, 27 while another demonstrated downregulation and hypermethylation of OXTR in WS. 28 While we did not see altered social behavior in a recent application of the standard social approach task, we did see differences in freezing during a conditioned fear task. 29 Another mouse model, which deletes the entire WS-homologous region, has also shown alterations in fear conditioning, 30 and individuals with WS have heightened phobias and non-social anxieties. Alterations in the brain OT system play an important role in social fear conditioning, contextual fear-induced freezing and social fear extinction. 31,32 Additionally, peripheral administration of an OT receptor agonist has been shown to inhibit fear-induced freezing, 33 and evoked OT release via channelrhodopsins also results in attenuation of fear. 34 In this study, we investigated whether OT dysregulation is a mechanism underlying the fear conditioning phenotype following deletion of the WSCR using the mouse experimental system. We employed the model reflecting the most common deletion found in WS patients: the hemizygous loss of the entire genomic region between the Gtf2i and Fkbp6 genes. 30 These heterozygous Complete Deletion (CD) mice show reduced freezing in fear conditioning recall, which is consistent with the expected consequences of OT elevation. Therefore, we probed whether OT activity could be responsible for the decreased expression of associative fear memory in CD mice. Further, to complement the prior human studies of OXTR expression in peripheral cells, 27,28 we tested whether OXTR expression differs in CD versus wild type (WT) mice across the brain, but we found no differences after statistical correction in this system, nor in a second neurotransmitter system (serotonin, 5HT), which had previously been shown to cooperate with OT in social learning, 25 and can be influenced by OT. 35 Together, these data suggest there is not a direct role for the OT system in associative fear learning in WS.

| Animals
CD mice contain a hemizygous deletion of the WSCR and were maintained on the C57BL/6J background (Jackson #000664). 30 Animals were bred by crossing CD heterozygotes to C57BL/6J WT animals to produce heterozygous CD experimental mice along with WT littermates for the control group. Tissue collection and genotyping PCR occurred in the second postnatal week. Mice were housed by sex and treatment, when relevant, and were kept on a 12:12 h light/dark schedule with food and water provided ad libitum.
All studies were approved by and conducted in accordance with the Institutional Animal Care and Use Committee at Washington University in St. Louis. All behavioral testing occurred during the light phase and was conducted by a female experimenter. Four independent cohorts were used in this study. Cohort 1 included 13 CD and 11 WT male mice from 8 independent litters and was used to assess blood OT levels via ELISA. Cohort 2 comprised 10 CD (4 females (F), 6 males (M)) and 10 WT (8 F, 2 M) mice from four independent litters and were behaviorally examined as adults (postnatal day [P] 97-106) with the conditioned fear task. Cohort 3 comprised 14 CD (8 F, 6 M) and 29 WT (12 F, 17 M) mice from 16 independent litters and served to evaluate the role of the OT system in associative fear and avoidance learning as adults (P68-118). Cohort 4, comprising 22 WT (12 F, 10 M) and 14 CD (5 F, 9 M) from 11 independent litters, was used to evaluate OXTR and serotonin transporter (SERT) expression in specific brain regions of interest. Tissue was collected post-mortem to confirm initial genotyping results.

| Oxytocin ELISA
Blood was drawn from the retro orbital sinus of isofluraneanesthetized mice at P30 using heparinized glass capillary tubes. Samples were collected in 1.8 ml EDTA-coated tubes, spun at 1600 g for 5 min at 4 C, then split into two aliquots, placed on dry ice and stored at À80 C until use. An ELISA kit was used for colorimetric quantification of OT per the manufacturer's protocol (ADI-900-153A, Enzo Life Sciences, Farmingdale, NY, USA). Prior to use, samples were diluted 1:3 with 200 μl of assay buffer. Absorbance measurements were read at 405 nm and OT concentration was calculated using a standard curve produced using the provided OT standards.

| Conditioned fear task
Associative fear and avoidance learning were evaluated in the CD mice using the conditioned fear paradigm (Figure 1(A)), as described in our previous studies. 36 Briefly, each mouse was habituated to and tested in an acrylic apparatus, which measured 26 cm Â 30 cm Â 30.5 cm tall and     50 Thus, because of the suggested increase in OT production in WS, we sought to determine whether the CD mouse model also had altered associative fear learning. Previously a decrease in overall freezing time during fear conditioning was shown in the CD model using only male mice; 30 here we replicate and expand on these findings by confirming the phenotype in both sexes.
We found that CD mice responded to shock during conditioning (Day 1) by increasing freezing to the same extent as WT mice (main effect of minute, F(2, 36) = 77.81, p = 8.5 Â 10 À14 ). There was no main or interaction effect of genotype observed (Figure 1(B), complete statistical analysis available in Table S1). WT and CD mice both

| CD conditioning deficits are not reversed by central infusion of an oxytocin receptor antagonist
We next sought to determine if CD mice had elevated OT levels in the blood, as had been reported in patients. 26 We used the same ELISA approach, but did not see a significant difference between genotypes

| Autoradiography reveals no changes in oxytocin receptor density or distribution in CD mice
In parallel to the experimental approach, we investigated OXTR availability in the mouse brain through a discovery-based approach, as elevated levels in the amygdala might influence fear conditioning. 50 Particularly, given opposite directions of effect of the OTA in WT and CD mice on Day 1 freezing (Figure 2(B)), we suspected genotype differences in OXTR expression in regions related to fear learning. Furthermore, it is of interest to study this binding given the findings of OXTR dysregulation in brains of humans with the WSCR deletion. 27,28 Therefore, we conducted an autoradiography study using radiolabeled OVT ligand on coronal sections of WT and CD brains (Figure 3(A) Figure 3(B)), as hypersociability and cognitive impairments are characteristic of WS. We found no significant differences in OXTR binding between genotypes within regions of interest in CD and WT brains when corrected for multiple testing (Figure 3(C), Supplemental Table 2).
There was a nominally significant change in the LSN (p = 0.034), but it did not meet the corrected experimentwise critical alpha level (α = 0.006). Given the role of the LSN in fear and anxiety and future focused studies of this region may be warranted.

| Autoradiography reveals no changes in serotonin transporter density or distribution in CD mice
Finally, with no changes in OXTR availability, we examined an additional alternative neurotransmitter: serotonin (5HT). Disruption to the We measured SERT binding in several brain regions (Figure 4(A)).
We focused on amygdalar regions relevant to the human postmortem studies 63 and fear conditioning phenotypes, 64 assessing BLA, central amygdala (CeA) and LA regions independently based on findings that they differ in the amount of serotonergic innervation in some species. 65 We then included other areas where SERT has been implicated in behaviors relevant to the WS phenotype, patient findings and knowledge of 5HT biology. These included the nucleus accumbens (NAc) based on 5HT's role on social reward in this region, 25 the BNST for its role in adaptive anxiety, 66

| DISCUSSION
We and others have previously observed altered fear conditioning in WS models. 29,30 Expanding on previous studies that used only male mice, we found, using both sexes, that CD animals on a C57BL/6J background had a suppressed fear response to the context and cue presented in the fear conditioning paradigm. Thus, CD mice enable investigation of the underlying circuit disruptions mediating this phenotype. We hypothesized that the altered associative fear memory response in CD mice was because of the increased availability of OT based on human findings of elevated OT in WS. 26 We focused our efforts on functional studies, which would be definitive with regards In addition, given some of the prior work suggesting OXTR receptor gene expression and methylation in WS patient blood cells, we were curious if receptor availability was also altered in the brain following deletion of these genes. We used autoradiography because it can measure the availability of the receptor at the surface, which should reflect protein level and localization changes in addition to changes in gene expression. Further, it provides an opportunity for spatially informed analyses. As such, it is the best single measure for assessing if the OXTR is modulated in a conserved way by these mutations. We assessed regions previously associated with fear conditioning and those where OT had been shown to modulate social reward (CPu). Overall, we found no significant changes in OXTR binding in the CD mouse brain compared with controls, although there was a nominally significant difference in the LSN prior to correction, in the same direction of effect as seen in WS patient blood cells. The LSN is an interesting center integrating a variety of fear and anxiety signals, for example, playing a role in how stressful social cues are received. 71 In addition, the LSN has been associated with fear-enhancing effects of the OXTR, but as modulation of the OXTR did not alter contextual conditioned fear responses, it is thought to be through indirect means. 72 These data motivate future studies focusing on the role of the LSN in behavioral abnormalities observed in the CD mouse model. Despite ruling out a direct role for OT in fear conditioning deficits in this model, altered OT might play a role in the increased social motivation in this population. While it is of interest scientifically to assess this, hypersociality is not as much of a concern therapeutically as other phenotypes, such as learning deficits, ADHD, phobias and anxiety. Beyond OT and serotonin, dopamine has also been implicated in WS, 7 and has been previously connected to fear conditioning, 73 other anxiety-avoidance tasks, 74 and ADHDrelated hyperactivity. 75 Thus, it is possible deletion of the WSCR disrupts dopamine signaling to result in these behavioral alterations. Understanding the roles of any of these systems in patientrelated phenotypes of the CD mice might help highlight potential treatments in WS, as a wide range of therapeutics working on these systems are currently available.