Increased anxiety and decreased sociability in adulthood following paternal deprivation involve oxytocin in the mPFC

Early adverse experiences often have devastating consequences on adult emotional and social behavior. However, whether paternal deprivation (PD) during the pre-weaning period affects brain and behavioral development remains unexplored in socially mandarin vole (Microtus mandarinus). We found that PD increased anxiety-like behavior and attenuated social preference in adult males and females; decreased prelimbic cortex OT-immunoreactive fibers and paraventricular nucleus OT positive neurons; reduced levels of medial prefrontal cortex (mPFC) OT receptor protein in females and OT receptor and V1a receptor protein in males. Intra-prelimbic cortical OT injections reversed anxiety-like behavior and social preferences affected by PD, whereas injections of OT and OT receptor antagonist blocked this reversal. These findings demonstrate that PD leads to increased anxiety-like behavior and attenuated social preferences with involvement of the mPFC OT system. The prelimbic cortex OT system may be an important target for the treatment of disorders related to early adverse experiences.


Introduction
Social attachments are necessary in many species as they facilitate reproduction, increase survival, 30 provide a sense of security and reduce feelings of stress and anxiety (Coria-Avila et al., 2014). In 31 humans, attachment is especially important during early development because disruption of filial 32 attachment in children (e.g., abuse, neglect, death of a parent) increases their vulnerability to mood 33 and anxiety disorders at a later age (Brown et al., 1977;Agid et al., 1999;Bernet and Stein, 1999;34 Reinherz et al., 1999). 35 Neonatal social or paternal deprivation (PD) has been proven to exert a profound and persistent 36 influence on the physiological and behavioral development of offspring. Neonatal rodents depend 37 on their parents physiologically and emotionally. For example, PD impairs sociability (Bambico et  38  2014) to affect social behavior. If pre-weaning PD changes social preference and emotion, we 28 predict that it should also affect levels of OT and OTR. 29 OTRs and V1aRs are found in the medial prefrontal cortex (mPFC) (Smeltzer et al., 2006;30 Lieberwirth and Wang, 2016). Several studies suggest that the mPFC is critical for the expression 31 of anxiety-like behavior ( A recent report found that a subset of mPFC neurons elevate discharge rates when approaching a 38 strange mouse but not when approaching non-social objects (Lee et al., 2016) indicating 39 involvement of the mPFC in social behavior. However, whether OT in the PLC is involved in the 40 manifestations of pre-weaning deprivation on emotion and social preference remains unclear. 41 Using socially monogamous mandarin voles we investigated the effects of PD from PND 14-21 42 on emotion and social preference, and levels of OT and OTR in specific brain regions. We then 43 tested whether microinjection of OT into the mPFC can recover the effects of pre-weaning PD. 44 We hypothesized that disruption of early emotional attachment between pups and fathers affects 1 anxiety-like behavior and social preference in mandarin voles at adulthood, and that the OT 2 system is likely involved in this process. 3 4

5
Experiment 1: Effect of pre-weaning PD on anxiety-like behavior and social 6 preference 7 It has previously been shown that offspring who experience neonatal maternal separation or early 8 deprivation display high levels of anxiety-like behavior ( that PD group males and females had fewer PVN OT-IR neurons than the PC group (both p < 0.01) 5 (Fig. 2), and reduced mPFC OT-IR fibers than PC group males and females (both p < 0.01) (Fig.  6 3). Females possessed more PVN OT-IR neurons (p < 0.01) and mPFC OT-IR fibers (p < 0.01) 7 than males (Fig. 2, 3). NAcc OT-IR fibers did not differ between groups (Fig. 4). 8 9 Experiment 3: Effect of pre-weaning PD on mPFC OTR-IR and V1aR-IR, serum 10 OT and corticosterone (CORT) concentration 11 PD reduced OT-IR fibers in the PLC (Experiment 2 protein in the mPFC than PC group females (t(6) = 2.648, p < 0.05), but not in the NAcc or PVN. 23 No group differences were noted for levels of female V1aR protein in the mPFC, NAcc or PVN. 24 PD had no effect on AVP neuropeptide in the mPFC, NAcc or PVN regardless of sex (Fig. 5). 25 PVN OT signaling is necessary and sufficient for social buffering effects (Smith and Wang,26 2014) and is anxiolytic (Neumann et al., 2000) by suppressing hypothalamic-pituitary-adrenal 27 (HPA) axis function. To assess whether PD reduces OT and increases CORT in serum, three hours 28 following the social preference test, experimental mandarin voles from the PC (n = 6 males, 6 29 females) and PD (n = 6 males, 6 females) groups were anesthetized with pentobarbital sodium (30 30 mg/kg i.p.). Serum OT and CORT concentrations were monitored. 31 Two-way ANOVA revealed an interaction between treatment x sex for serum OT (F(1,20) = 32 5.055 p < 0.05) and CORT (F(1,20) = 11.755, p < 0.01). The post-hoc test indicated that 33 pre-weaning PD significantly reduced serum OT (p < 0.01) and increased serum CORT (p < 0.01) 34 concentrations only in females. Female serum OT levels were much higher than males in the PC 35 group (p < 0.05). Female serum CORT concentrations were also much higher than males in the PD 36 group (p < 0.01) (Fig. 6). . Therefore, 43 we tested the hypothesis that microinjection of OT in the PLC restores anxiety-like behavior and social preference altered by pre-weaning PD. To this end, we implanted bilateral injection 1 cannulas into the PLC. After three days of recovery, subjects received an intra-PLC injection ( Fig.  2 7A-C) of CSF (male = 6, female = 6), OT (male: 1 ng OT = 6, 10 ng OT = 6; female: 1 ng OT = 6, 3 10 ng OT = 6) or OT plus OTA (male: 10 ng OT/10 ng OTA = 6, 10 ng OT/100 ng OTA = 5; 4 female: 1 ng OT/10 ng OTA = 5, 1 ng OT/100 ng OTA = 5). Levels of anxiety-like behavior and 5 social preference were then measured. 6 One-way ANOVA showed that microinjection of OT in the PLC increased the percentage of 7 time spent in the central area for males (10 ng: p < 0.01) and females (1 ng: p < 0.05 and 10 ng: p 8 < 0.01) exposed to pre-weaning PD, while OT plus either dose of OTA had no effect (except male 9 10 ng OT/10 ng OTA, p < 0.05). No differences were found between treatment groups for total 10 distance travelled (Fig. 7

20
The present study found that pre-weaning PD possibly disrupts emotional attachment in mandarin 21 vole pups, as evidenced by increased levels of anxiety-like behavior and attenuated social 22 preference in male and female adults (Experiment 1). PD reduced mPFC OT-IR fibers and PVN 23 OT-IR neurons (Experiment 2), and decreased mPFC OTR protein in females, OTR and V1aR 24 protein in males, and reduced serum OT and increased CORT in females (Experiment 3). We then 25 demonstrated that intra-PLC OT injection restored anxiety-like behavior and social preference 26 altered by pre-weaning PD (Experiment 4). 27 In this study, PND 14-21 PD increased anxiety-like behavior and reduced social preference in 28 both sexes. This is consistent with studies in animals and other humans that early severe 29 deprivation is associated with behavioral abnormalities ( that the disruption of attachment between pups and fathers during PND 14-21 increases levels of 36 anxiety and reduce levels of sociability. 37 The present study found that PND 14-21 PD reduced mPFC OT-IR fibers and OTR protein in previous study found that OT in the PL region of the mPFC decreased anxiety regardless of sex, 43 and neither AVP nor OTR-A affected anxiety-like behavior (Sabihi et al., 2014b). Blocking OTR 1 in the mPFC enhances postpartum anxiety, but has no effect on anxiety in virgin females (Sabihi et  2 al., 2014a). OTR knockout mice display deficits in social approach behavior (Nishimori et al. 2016). It is probable that OT neurons releasing Glutamate activate local GABA-interneurons in 10 the mPFC, leading to reduced anxiety and regulated social behavior. Thus, we conclude that 11 disruption to attachment between pups and fathers increases levels of anxiety and impairs social 12 preference via a decrease in OT-IR fibers and OTR protein in the mPFC. 13 PND 14-21 PD decreased V1aR protein in males but had no effect on levels of V1aR protein in 14 females. Sex differences in the OT system may therefore be implicated in sex-specific regulation 15 of impaired social behavior (Smeltzer, 2006; Dumais and Veenema, 2016). V1aR and OTR show 16 distinct and largely nonoverlapping expression in the rodent brain (Dumais and Veenema, 2016). 17 Female prairie voles had higher densities of OTR binding but lower densities of V1aR binding 18 than males in the mPFC (Smeltzer, 2006 We also found that disruption of early emotional attachment reduced OT-IR neurons in the PVN. 32 These results suggest that OT magnocellular neurons in the hypothalamic nuclei may have 33 reduced OT into the peripheral circulation via the posterior pituitary (Ludwig and Leng, 2006). 34 During some prosocial behavior, OT is released into plasma and centrally in females ( A major finding in the current study is that OT microinjection directly into the PLC of males 2 (10 ng) and females (1 ng and 10 ng) restored changes to anxiety-like behavior and social 3 preference resulting from PD, while voles treated with OT plus either dose of OTA did not exhibit 4 reversal of any kind (except for the male 10 ng OT/10 ng OTA group). This is consistent with 5 previous findings that OT in the PLC reduces anxiety-like behavior in both sexes (Sabihi et al.,6 2014b). Injection of highly specific OTA into the PLC region of the mPFC increases anxiety-like 7 behavior in postpartum females (Sabihi et al., 2014a)  interaction between OT and DA to regulate social behavior in the mPFC, and mPFC OT may be a 20 promising target for treating emotional and social disorders induced by adverse early experiences. 21 Mandarin voles can be used as an animal model to investigate the effects of early emotional 22 attachment disruption on the adult brain and behavior and underlying mechanism (He et al., 2017). 23 Disruption to early emotional attachment between pups and fathers impairs emotional and social 24 behavior and leads to OT system dysfunction in the brain. We provide intriguing evidence that 25 site-specific OT action in the PLC has potential beneficial effects on the recovery of emotional and 26 social dysfunction induced by the disruption to early emotional attachment. The modulation of OT 27 on emotion and social behavior was sex-specific. Therefore, OT may be potentially targeted to 28 ameliorate social and emotional deficits resulting from early adverse experiences. 29 30

Open field test (OFT)
2 To assess the impact of pre-weaning PD on adult anxiety-like behavior, F2 mandarin voles were 3 observed in the OFT on PND 70. Subjects were placed in a center of an open-field arena (50 cm x 4 50 cm x 25 cm), and the duration and distance moved within the center or periphery was recorded 5 using an automated system (SocialScan 2.0, Clever Sys, Reston, VA, USA). Measures include the 6 proportion of time spent in the central area and total distance moved in the OFT. 7 8 Social preference test (SPT) 9 Immediately after the OFT, the SPT was carried out. The SPT was based on the social 10 approach-avoidance paradigm previously described (Qiao et al., 2014). Briefly, prior to testing 11 voles were placed in a box. The test box (50 cm length x 50 cm width x 24 cm height) was 12 constructed of white glacial polyvinylchloride. After 5 min of habituation, an empty wire-mesh 13 cage (object stimulus; 10 cm length x 10 cm width) was placed near one side wall of the arena for 14 10 min, which was then exchanged for a cage containing an stranger same-sex con-specific (social 15 stimulus) for an additional 10 min. Behavioral responses to an empty cage or to a cage with a 16 stimulus individual were videotaped, and scored and quantified afterwards using OBSERVER 17 v5.0 (vNoldus, NL). Measures included the time spent investigating the object and social stimulus. 18 Data are presented as investigation time/total time (10 min) x 100%. 19 20

21
Blood was collected directly in microcentrifuge vials from the heart (Cao et al., 2014). After 22 clotting, blood was centrifuged at 6000 rpm for 30 min at 4°C. Supernatant was collected. Serum 23 OT and CORT levels were monitored by a vole-specific enzyme-linked immunosorbent assay 24 (Shanghai Xitang Biotechnology, Shanghai, China), according to the manufacturers' instructions. 25 The resultant absorbance was measured at 450 nm using a Metertech microplate reader (BioTek 26 Instruments, Winooski, USA) after the reader was zeroed using the blank well. Variation between 27 duplicate measurements was less than 5%. (NAcc) and 40X (mPFC) objective using the cell counter (PVN) and the value of integrated 6 optical density (IOD) plugin in Image-pro-plus; these were averaged across three nonoverlapping 7 sections in an evenly spaced series per animal. 8 9 Western blotting 10 Brains were immediately extracted and frozen in dry ice. Coronal sections (200 µm) were cut on a 11 cryostat and frost mounted onto microscope slides. Bilateral tissue punches with a 1 mm diameter 12 were taken from the entire mPFC (Cg and PLC), NAcc and PVN and stored at -80°C until 13 processing. Total proteins were extracted with RIPA lysis buffer containing protease inhibitors 14 (R0010, Solarbio Biotechnology). Protein samples were separated by sodium dodecyl 15 sulphate-polyacrylamide gel electrophoresis and transferred to PVDF membranes (Millipore, 16 Billerica, MA, USA). The membrane was incubated with the following diluted primary antibodies: 17 OTR (ab181077, 1:2000, Abcam), V1aR (GTX89114, 1:7000, GeneTex), AVP (AB1565, 1:4000, 18 Millipore), β-Tubulin (CW0098M, 1:5000, ComWin Biotechnology), at 4°C overnight. Following 19 washing, the membrane was incubated with horseradish peroxidase-conjugated secondary 20 antibodies (1:10000, ZhongShan Goldenbridge Biotechnology); membranes were revealed with 21 ECL (WBKLS0500, Millipore) and exposed on Luminescent Imaging (Tanon 6200 Luminescent  22 Imaging Workstation, Tanon). Quantification was performed using ImageJ software, and all 23 signals were normalized within the same membrane to β-Tubulin. 24 25 All microinjections were made with a 33-gauge needle that extended 1 mm below the guide 33 cannula and delivered at a rate of 0.1 μl/min as previously described (Young et al., 2014). Data 34 were excluded if tips were located in other brain regions. The final size of each group was 6 35 except for the 10 ng OT/100 ng OTA male group which numbered 5, and the 1 ng OT/10 ng OTA 36 and 1 ng OT/100 ng OTA female groups (both n = 5). 37 38 Data analysis 39 Parametric tests were used because all data were normally distributed according to one-sample 40

13
Figure 7 -supplement 5 14 Table 5 Summary of microinjection of OT into the PLC on anxiety-like behavior and social 15 preference analysis.