《神经科学 NEUROSCIENCE》：Ventral Hippocampus Modulates Anxiety-Like Behavior in Male But Not Female C57BL/6 J Mice

No of Pages 9 06 September 2019 ARTICLE IN PRESS NEUROSCIENCE BR● RESEARCH ARTICLE Cheng Wang et al. /Neuroscience 418 (2019)XXx-XXX Ventral Hippocampus Modulates Anxiety-Like Behavior in Male But Not Female c57BL/6J Mice Cheng Wang, Yu Zhang, Shan Shao, Shuang Cui, You Wan. b, c, and Ming Yi b, * Neuroscience Research Institute, Peking University, 38 Xueyuan Road, Beiing 100083, China Department of Neurobiology, School of Basic Medical Sciences, Peking University, 38 Xueyuan Road, Bejing 100083, China Key Laboratory for Neuroscience, Ministry of Education/National Health Commission, Peking University, 38 Xueyuan Road, Beijing 100083, China Co-innovation Center of Neuroregeneration, Nantong University, Nantong 226001, China Abstract-Remarkable sex difference has been observed in emotional processing including anxiety. The hippocam pus, its ventral pole in particular, modulates anxiety like behavior in rodents. However, most researches have been per- formed in male animals only, leaving hippocampal modulation of anxiety in females poorly defined In the present study, we showed that excitotoxic lesioning of the ventral hippocampus with ibotenic acid produced anxiolytic effects in three behavioral tests(novelty-suppressed feeding, marble burying, and elevated-plus maze )in male but not female C57BL/6 J mice. Locomotion in the open field remained similar after lesioning in either sex More c-Fos-positive neu rons were observed in the ventral hippocampus in male than in female mice after exploration in an elevated plus maze, indicating stronger enrollment of this region in anxiety-like behavior in males. These results reveal significant biological sex difference in ventral hippocampal modulation on anxiety in mice and provide a new sight for anxiety modulation and hippocampal function. 2019 IBRo. Published by Elsevier Ltd. All rights reserved Key words: ventral hippocampus, anxiety, sex difference. INTRODUCTION Furthermore, local field potentials in the ventral hippocam- Anxiety disorders are prevalent( Kessler et al., 2010)and pus show increased correlation in theta-frequency cover 7. 3% of the population(Baxter et al., 2013). Signifi- oscillations with medial prefrontal cortex in anxiogenic cant sex difference has been observed in mood disorders. environments in mice(Adhikari et al., 2010). Optogenetic women are more likely than men to experience anxiety activation and inhibition of basolateral amygdala to ( Kessler et al., 1994: Craske and Stein, 2016)or depression ventral hippocampus projections increase and decrease (Nolen-Hoeksema and Girgus, 1994)from early adoles- anxiety-related behaviors in mice, respectively(Felix-Ortiz cence through adulthood. Brain regions in the limbic system et al., 2013). More recently, anxiety cells"have been dis underlie emotional processing(Rosen and Schulkin, 1998) covered in ventral hippocampal CA1 in mice (Jimenez In rodents, ventral hippocampus shows dense connection etal.2018) with affect-related regions including amygdala, prefrontal However, most of these studies were performed only in cortex and hypothalamus(Bannerman et al., 2004; Fanse male animals. To our knowledge, effects of ventral hippo- low and Dong, 2010). Excitotoxic lesioning of the ventral campal lesioning on anxiety in female rodents have not hippocampus in rats reduces anxiety-related behaviors in been reported Previous work has shown that males have elevated plus-maze and hyponeophagia tests( Kjelstrup larger hippocampal(Ruigrok et al., 2014; Meyer et al et al., 2002; Bannerman et al., 2003), whereas optogenetic 2017)and amygdala(Giedd et al., 1996: Goldstein et al activation of granule cells in the ventral dentate gyrus sup. 2001: Ruigrok et al., 2014)volumes, as well as greater presses innate anxiety in mice( Kheirbek et al., 2013 within-hemispheric connectivity(Ingalhalikar et al., 2014 than females in both mice and humans. The present study aims to explore potential sex difference in hippocampal modulation of anxiety, by examining anxiety-like behavior HEy L co in male and female C57BL/6 J China te+8613718823306. ing of the ventral hippocampus and assessing hippocampal sses:yuan@hsc.pku.edu.cn(YouWan)mingyi@hsc.pku neuronal activation with c-Fos staining after an anxious https://doi.org/10.1016/j-neuroscience.2019.08.032 0306-4522@ 2019 IBRO. Published by Elsevier Ltd. All rights reserved

NEUROSCIENCE RESEARCH ARTICLE Cheng Wang et al. / Neuroscience 418 (2019) xxx–xxx Ventral Hippocampus Modulates Anxiety-Like Behavior in Male But Not Female C57BL/6 J Mice Cheng Wang, a Yu Zhang, a Shan Shao, a Shuang Cui, a You Wana,b,c,d,⁎ and Ming Yia,b,⁎⁎ a Neuroscience Research Institute, Peking University, 38 Xueyuan Road, Beijing 100083, China b Department of Neurobiology, School of Basic Medical Sciences, Peking University, 38 Xueyuan Road, Beijing 100083, China c Key Laboratory for Neuroscience, Ministry of Education/National Health Commission, Peking University, 38 Xueyuan Road, Beijing 100083, China d Co-innovation Center of Neuroregeneration, Nantong University, Nantong 226001, China Abstract—Remarkable sex difference has been observed in emotional processing including anxiety. The hippocam￾pus, its ventral pole in particular, modulates anxiety-like behavior in rodents. However, most researches have been per￾formed in male animals only, leaving hippocampal modulation of anxiety in females poorly defined. In the present study, we showed that excitotoxic lesioning of the ventral hippocampus with ibotenic acid produced anxiolytic effects in three behavioral tests (novelty-suppressed feeding, marble burying, and elevated-plus maze) in male but not female C57BL/6 J mice. Locomotion in the open field remained similar after lesioning in either sex. More c-Fos-positive neu￾rons were observed in the ventral hippocampus in male than in female mice after exploration in an elevated plus￾maze, indicating stronger enrollment of this region in anxiety-like behavior in males. These results reveal significant biological sex difference in ventral hippocampal modulation on anxiety in mice and provide a new sight for anxiety modulation and hippocampal function. © 2019 IBRO. Published by Elsevier Ltd. All rights reserved. Key words: ventral hippocampus, anxiety, sex difference. INTRODUCTION Anxiety disorders are prevalent (Kessler et al., 2010) and cover 7.3% of the population (Baxter et al., 2013). Signifi- cant sex difference has been observed in mood disorders: women are more likely than men to experience anxiety (Kessler et al., 1994; Craske and Stein, 2016) or depression (Nolen-Hoeksema and Girgus, 1994) from early adoles￾cence through adulthood. Brain regions in the limbic system underlie emotional processing (Rosen and Schulkin, 1998). In rodents, ventral hippocampus shows dense connection with affect-related regions including amygdala, prefrontal cortex and hypothalamus (Bannerman et al., 2004; Fanse￾low and Dong, 2010). Excitotoxic lesioning of the ventral hippocampus in rats reduces anxiety-related behaviors in elevated plus-maze and hyponeophagia tests (Kjelstrup et al., 2002; Bannerman et al., 2003), whereas optogenetic activation of granule cells in the ventral dentate gyrus sup￾presses innate anxiety in mice (Kheirbek et al., 2013). Furthermore, local field potentials in the ventral hippocam￾pus show increased correlation in theta-frequency oscillations with medial prefrontal cortex in anxiogenic environments in mice (Adhikari et al., 2010). Optogenetic activation and inhibition of basolateral amygdala to ventral hippocampus projections increase and decrease anxiety-related behaviors in mice, respectively (Felix-Ortiz et al., 2013). More recently, “anxiety cells” have been dis￾covered in ventral hippocampal CA1 in mice (Jimenez et al., 2018). However, most of these studies were performed only in male animals. To our knowledge, effects of ventral hippo￾campal lesioning on anxiety in female rodents have not been reported. Previous work has shown that males have larger hippocampal (Ruigrok et al., 2014; Meyer et al., 2017) and amygdala (Giedd et al., 1996; Goldstein et al., 2001; Ruigrok et al., 2014) volumes, as well as greater within-hemispheric connectivity (Ingalhalikar et al., 2014), than females in both mice and humans. The present study aims to explore potential sex difference in hippocampal modulation of anxiety, by examining anxiety-like behavior in male and female C57BL/6 J mice with excitotoxic lesion￾ing of the ventral hippocampus and assessing hippocampal neuronal activation with c-Fos staining after an anxious experience. *Correspondence to: Key Laboratory for Neuroscience, Ministry of Education/National Health Commission, Peking University, 38 Xueyuan Road, Beijing 100083, China Tel.: +86 137 1882 3306. E-mail addresses: ywan@hsc.pku.edu.cn (You Wan) mingyi@hsc.pku. edu.cn (Ming Yi) https://doi.org/10.1016/j.neuroscience.2019.08.032 0306-4522/© 2019 IBRO. Published by Elsevier Ltd. All rights reserved. 1 NSC 19240 No of Pages 9 06 September 2019

No of Pages 9 06 September 2019 ARTICLE IN PRESS 2 Cheng Wang et al. / Neuroscience 418(2019)XXx-XXX EXPERIMENTAL PROCEDURES cage, and the food pellet was weighed again to obtain the Animals amount of food consumed Adult weight-matched (26-30 g)male and female C57BL/6 mice at the age of 6-8 weeks were housed 4-6 per cage Marble burying test under a 12 h light/dark cycle with food and water provided ad libitum(except before the novelty-suppressed feeding The marble burying test(Zhang et al., 2017)was performed test). Corn cobs were laid in the cage as bedding. All experi- 7 days after the novelty-suppressed feeding test. Twenty ments were conducted in a blind manner and in accordance marbles, with 4 x 5 arrangement, were put on cob bedding ith the institutional animal care and use committee of materials at a depth of 5 cm in home cage. Each mouse Peking University Health Science Center( LA2016061). All was left single in its home cage for 30 min before placed mice were handled for three days before behavioral experi in a random corner of the test apparatus facing the chamber ments 32 male and 36 female mice were used in behavioral wall. A marble with two-thirds or more in the cob bedding experiments Mice were randomly allocated to sham and was treated as a buried marble. the number of buried mar- lesion groups(16 male and 18 female mice in each group) bles was counted after 30 min For c-Fos staining, 6 male and 6 female mice were used Open field test Hippocampal lesioning 7 days after the marble burying test, each mouse was Mice were anesthetized with 1% pentobarbital sodium, and placed in a 60 x 60 x 60 cm open field chamber with 30 Ix positioned in a stereotaxic instrument( RWD Life Science illumination and allowed to explore freely for 5 min (Jiang Shenzhen, China): 0.4 ul of ibotenic acid(pH 7. 4, 10 mg et al., 2018). Locomotive activity was videotaped and the ml, Sigma-Aldrich, USA) was injected into bilateral ventral total distance traveled in the field was measured using the hippocampus(AP-3.40 mm, ML +3.0 mm, DV-32 mm SMART software(v2. 5.21, Panlab). The chamber was relative to Bregma)(Paxinos and Franklin, 2001) through a cleaned by 75% ethanol between tests 1-HL Hamilton microsyringe within 5 min, at a flow rate of 0.08 Hl/min. Sham lesion was performed by injecting the same amount of normal saline at the same flow rate After Elevated plus-maze test stereotaxic injection, the needle was left in place for another 5 min to allow for drug diffusion before slow withdrawal. Seven days after the open field test, mice performed the Mice were allowed to recover for 7 days before further elevated plus-maze test(Zheng et al., 2017). The maze consisted of two open(5 x 30 cm) and two closed arms (same size with 15 cm walls)and was placed 50 cm above the floor. In a room with 30 lx illumination each mouse was Novelty-suppressed feeding test placed onto the central area heading towards the same The timeline of behavioral experiments was shown in Fig. 1 open arm. Animal activity was videotaped and the time After recovery from surgery, mice were first examined in the spent in open arms and the percent of entries into open novelty-suppressed feeding test as previously reported arms in the following 5 min were assessed by hand-score (Zhang et al., 2017). The apparatus consisted of a chamber The maze was cleaned by 75% ethanol between tests (40 x 40 cm) filled with cob bedding materials at a depth of 5 cm. Under 300 Ix illumination, a standard chow pellet (5 g) was placed on a piece of white filter paper(10 cm in Estrus cycle identification diameter) positioned in the center of the arena. Each mouse was left single in a new cage for 30 min after 24 h food The estrus cycle of female mice consisted of proestrus estrus, metestrus and diestrus phases, and was defined deprivation before being placed in a random corner of the by vaginal cytology after each behavioral test. Proestrus chamber facing the chamber wall. The amount of time and estrus were allocated to high-estradiol phase(HEP), was recorded. After the first bite or a 10 min cut-off time while metestrus and diestrus were allocated to low- estradiol phase(LEP). Vaginal smears were obtained by the mouse was transferred to the prior new cage. A new sin gle food pellet, which was weighed in advance, was placed toothpicks scraping gently in vagina after each behavioral test. 95%alcohol was used to fix vaginal smears for in the cage. The mouse was allowed to eat the food pellet 15 min after drying in air for 15 min. Harris-Shorr staining for 5 min. After that the mouse was returned to its home was carried out to stain vaginal epithelial cells. Vaginal smears were put in Shorr stain solution for 10 min, then put in 95%, 95%, 100% and 100% alcohol for 1-2 min in turn, finally put in xylol for 10 min. After that, vaginal smears were sealed by neutral gum, dried in a 37C oven. Vaginal Fig. 1. Timeline of behavioral experiments. NSF: novelty-suppressed cytology was observed with a light microscope(Leica feeding test, MBT: marble burying test, OFT: open field test, EPM: ele DMI 4000B, Germany) as previously described(Nelsor ated plus-maze test. etal,1982)

EXPERIMENTAL PROCEDURES Animals Adult weight-matched (26–30 g) male and female C57BL/6 mice at the age of 6–8 weeks were housed 4–6 per cage under a 12 h light/dark cycle with food and water provided ad libitum (except before the novelty-suppressed feeding test). Corn cobs were laid in the cage as bedding. All experi￾ments were conducted in a blind manner and in accordance with the Institutional Animal Care and Use Committee of Peking University Health Science Center (LA2016061). All mice were handled for three days before behavioral experi￾ments. 32 male and 36 female mice were used in behavioral experiments. Mice were randomly allocated to sham and lesion groups (16 male and 18 female mice in each group). For c-Fos staining, 6 male and 6 female mice were used. Hippocampal lesioning Mice were anesthetized with 1% pentobarbital sodium, and positioned in a stereotaxic instrument (RWD Life Science, Shenzhen, China); 0.4 μl of ibotenic acid (pH 7.4, 10 mg/ ml, Sigma-Aldrich, USA) was injected into bilateral ventral hippocampus (AP -3.40 mm, ML ± 3.0 mm, DV -3.2 mm relative to Bregma) (Paxinos and Franklin, 2001) through a 1-μL Hamilton microsyringe within 5 min, at a flow rate of 0.08 μl/min. Sham lesion was performed by injecting the same amount of normal saline at the same flow rate. After stereotaxic injection, the needle was left in place for another 5 min to allow for drug diffusion before slow withdrawal. Mice were allowed to recover for 7 days before further experiments. Novelty-suppressed feeding test The timeline of behavioral experiments was shown in Fig. 1. After recovery from surgery, mice were first examined in the novelty-suppressed feeding test as previously reported (Zhang et al., 2017). The apparatus consisted of a chamber (40 × 40 cm) filled with cob bedding materials at a depth of 5 cm. Under 300 lx illumination, a standard chow pellet (5 g) was placed on a piece of white filter paper (10 cm in diameter) positioned in the center of the arena. Each mouse was left single in a new cage for 30 min after 24 h food deprivation before being placed in a random corner of the chamber facing the chamber wall. The amount of time passed before the mouse approached and ate the pellet was recorded. After the first bite or a 10 min cut-off time, the mouse was transferred to the prior new cage. A new sin￾gle food pellet, which was weighed in advance, was placed in the cage. The mouse was allowed to eat the food pellet for 5 min. After that, the mouse was returned to its home cage, and the food pellet was weighed again to obtain the amount of food consumed. Marble burying test The marble burying test (Zhang et al., 2017) was performed 7 days after the novelty-suppressed feeding test. Twenty marbles, with 4 × 5 arrangement, were put on cob bedding materials at a depth of 5 cm in home cage. Each mouse was left single in its home cage for 30 min before placed in a random corner of the test apparatus facing the chamber wall. A marble with two-thirds or more in the cob bedding was treated as a buried marble. The number of buried mar￾bles was counted after 30 min. Open field test 7 days after the marble burying test, each mouse was placed in a 60 × 60 × 60 cm open field chamber with 30 lx illumination and allowed to explore freely for 5 min (Jiang et al., 2018). Locomotive activity was videotaped and the total distance traveled in the field was measured using the SMART software (v2.5.21, Panlab). The chamber was cleaned by 75% ethanol between tests. Elevated plus-maze test Seven days after the open field test, mice performed the elevated plus-maze test (Zheng et al., 2017). The maze consisted of two open (5 × 30 cm) and two closed arms (same size with 15 cm walls) and was placed 50 cm above the floor. In a room with 30 lx illumination, each mouse was placed onto the central area heading towards the same open arm. Animal activity was videotaped and the time spent in open arms and the percent of entries into open arms in the following 5 min were assessed by hand-score. The maze was cleaned by 75% ethanol between tests. Estrus cycle identification The estrus cycle of female mice consisted of proestrus, estrus, metestrus and diestrus phases, and was defined by vaginal cytology after each behavioral test. Proestrus and estrus were allocated to high-estradiol phase (HEP), while metestrus and diestrus were allocated to low￾estradiol phase (LEP). Vaginal smears were obtained by toothpicks scraping gently in vagina after each behavioral test. 95% alcohol was used to fix vaginal smears for 15 min after drying in air for 15 min. Harris–Shorr staining was carried out to stain vaginal epithelial cells. Vaginal smears were put in Shorr stain solution for 10 min, then put in 95%, 95%, 100% and 100% alcohol for 1–2 min in turn, finally put in xylol for 10 min. After that, vaginal smears were sealed by neutral gum, dried in a 37 °C oven. Vaginal cytology was observed with a light microscope (Leica DMI 4000B, Germany) as previously described (Nelson et al., 1982). Fig. 1. Timeline of behavioral experiments. NSF: novelty-suppressed feeding test, MBT: marble burying test, OFT: open field test, EPM: ele￾vated plus-maze test. 2 Cheng Wang et al. / Neuroscience 418 (2019) xxx–xxx NSC 19240 No of Pages 9 06 September 2019

ARTICLE IN PRESS No of Pages 9 06 September 2019 Cheng Wang et aL. Neuroscience 418(2019)Xx Histology of hippocampal lesioning Slides were rehydrated through phosphate buffer saline After the elevated plus-maze test, mice were anesthetized and stained in Cresyl violet solution for 8 min, then put in with 1% pentobarbital sodium(50 mg/kg, ip), intracardially 75%,75%,80%,80%,90%,100%and100% alcohol in turn perfused with 50 ml normal saline and 50 ml 4% parafor (1 min for each step). Finally, slides were put in xylol for maldehyde(PFA, in 0.1 M phosphate buffer, pH 7. 4)in turn 2 min twice and sealed by neutral gum Lesion sites in ven Brains were post-fixed with PFA for 24 h, and cryoprotected tral hippocampus were confirmed by microscopic inspection in 20% and 30% sucrose solutions in turn. Coronal brain ( Leica DMI 4000B, Germany) sections were cut at 30-um using a cryostat microtome (Leica 1950, Germany)and mounted on positive charged Immunofluorescence and quantification of c-Fos plus slides, which were dried in 37C oven for a week expressIo Male Female activation. male and female mice were sacrificed either 90 min after a 5-min exploration in the elevated plus-maze(anxiety group) or directly from the home cage(control group) Estrus cycle of female mice were 2.46mm identified before sacrifice. To exclude influences from the estrus 70 mm 少少 ycle(Marcondes et al., 2001),we selected female mice showing sim lar levels of open arm exposure to males, representing similar level of 92m innate anxiety. Of the 6 female mice used for c-Fos staining, 3 were in the estrus phase and 3 were in the metestrus phase. Procedures of per 3.16mm fusion and post-fixing were similar to described above. 30-um sections were sliced coronally through ventral 3.28mm hippocampus, amygdala and prelim- bic cortex. Free-floating sections were washed in PBs (5 min x 3 346m times), blocked for 60 min with blocking-buffer containing 3% bull erum albumin and.3% triton x- in pbs. and in 3.52mm bated with the rabbit anti-C-Fos(Cell Signaling Technology 2250S, USA) in 4C for 24 h. The primary anti body was dissolved 1: 200 in 3.64mm blocking-buffer. Sections were then washed in PBS(10 min x 3 times and incubated with Alexa fluor 594-conjugated goat anti-rabbit anti- body(Invitrogen A11032, USA) at room temperature for 60 min, fol- lowed by PBS-washing(10 min x 3 times), and finally mounting and cover-slipping on microscope slides after incubated with DAPI(1: 1000) The secondary antibody was dis- Fig. 2. Histology of ventral hippocampal excitotoxic lesion. Schematic of ventral hippocampal lesion size solved 1: 500 in blocking-buffer in male(A)and female mice( B). Coronal sections were made from AP-2.5 mm to AP-3. 6 mm relative Images were taken on a Leica respectively. No lesions were detected in the slice of -2.46 mm in either male or female mice. (C, D) STED laser scanning microscope Representative slices of dorsal and ventral hippocampus from male(C) and female(D)mice after ibotenic using a 10 x objective for quantifica acid lesioning. Dorsal hippocampus was intact (left), whereas ventral hippocampus was lesioned (right). tion of c-Fos immunoreactivity in Filled arrows indicated borders between intact and damaged tissues. brain regions. Identical image

Histology of hippocampal lesioning After the elevated plus-maze test, mice were anesthetized with 1% pentobarbital sodium (50 mg/kg, i.p.), intracardially perfused with 50 ml normal saline and 50 ml 4% parafor￾maldehyde (PFA, in 0.1 M phosphate buffer, pH 7.4) in turn. Brains were post-fixed with PFA for 24 h, and cryoprotected in 20% and 30% sucrose solutions in turn. Coronal brain sections were cut at 30-μm using a cryostat microtome (Leica 1950, Germany) and mounted on positive charged plus slides, which were dried in 37 °C oven for a week. Slides were rehydrated through phosphate buffer saline and stained in Cresyl violet solution for 8 min, then put in 75%, 75%, 80%, 80%, 90%, 100% and 100% alcohol in turn (1 min for each step). Finally, slides were put in xylol for 2 min twice and sealed by neutral gum. Lesion sites in ven￾tral hippocampus were confirmed by microscopic inspection (Leica DMI 4000B, Germany). Immunofluorescence and quantification of c-Fos expression To assess hippocampal neuronal activation, male and female mice were sacrificed either 90 min after a 5-min exploration in the elevated plus-maze (anxiety group) or directly from the home cage (control group). Estrus cycle of female mice were identified before sacrifice. To exclude influences from the estrus cycle (Marcondes et al., 2001), we selected female mice showing simi￾lar levels of open arm exposure to males, representing similar level of innate anxiety. Of the 6 female mice used for c-Fos staining, 3 were in the estrus phase and 3 were in the metestrus phase. Procedures of per￾fusion and post-fixing were similar to described above. 30-μm sections were sliced coronally through ventral hippocampus, amygdala and prelim￾bic cortex. Free-floating sections were washed in PBS (5 min × 3 times), blocked for 60 min with blocking-buffer containing 3% bull serum albumin and 0.3% triton X- 100 dissolved in PBS, and incu￾bated with the rabbit anti-c-Fos (Cell Signaling Technology 2250S, USA) in 4 °C for 24 h. The primary anti￾body was dissolved 1:200 in blocking-buffer. Sections were then washed in PBS (10 min × 3 times) and incubated with Alexa Fluor 594-conjugated goat anti-rabbit anti￾body (Invitrogen A11032, USA) at room temperature for 60 min, fol￾lowed by PBS-washing (10 min × 3 times), and finally mounting and cover-slipping on microscope slides after incubated with DAPI (1: 1000). The secondary antibody was dis￾solved 1: 500 in blocking-buffer. Images were taken on a Leica STED laser scanning microscope using a 10× objective for quantifica￾tion of c-Fos immunoreactivity in brain regions. Identical image Fig. 2. Histology of ventral hippocampal excitotoxic lesion. Schematic of ventral hippocampal lesion size in male (A) and female mice (B). Coronal sections were made from AP −2.5 mm to AP −3.6 mm relative to Bregma. Gray and black zones in schematic figure represented maximum and minimum lesion areas, respectively. No lesions were detected in the slice of −2.46 mm in either male or female mice. (C, D) Representative slices of dorsal and ventral hippocampus from male (C) and female (D) mice after ibotenic acid lesioning. Dorsal hippocampus was intact (left), whereas ventral hippocampus was lesioned (right). Filled arrows indicated borders between intact and damaged tissues. Cheng Wang et al. / Neuroscience 418 (2019) xxx–xxx 3 NSC 19240 No of Pages 9 06 September 2019

No of Pages 9 06 September 2019 ARTICLE IN PRESS Cheng Wang et al. / Neuroscience 418(2019)XXx-XXX acquisition settings were maintained for all subsequent ima MaeB□ Female ging of C-Fos o Sham For quantification of c-Fos expression, three slices were ■ Lesion ounted for ventral hippocampal CAl (AP -3.3 mm 3. 4 mm, and-3.5 mm relative to Bregma), basolateral amygdala(AP-1.5 mm, -1.6 mm, and-1.7 mm relative to Bregma) and prelimbic cortex(AP 1.9 mm, 1.8 mm, and 1.5 mm relative to Bregma), respectively. c-Fos expression was quantified for each of these brain regions by ImageJ (Version 1.52, National Institutes of Health, USA). The num- bers of c-Fos-positive neurons in left and right hemispheres LEP was pooled since no significant differences were observed (data not shown) 150 Statistical analysis All data were analyzed and plotted using the GraphPad Prism 6 software Data were expressed as mean t SEM(standard error of the mean). Males and females were analyzed sepa rately. The assumption of normality and equality of variance vas tested by Shapiro-Wilk test and F-test, respectively. All E6月088 data met the normal distribution and had equal variances in each group. Unpaired Student's t test was perfomed in males Two-way ANOVA with Bonferroni post hoc test was performed in females, with lesion and estrus cycle as the two factors Fig 3. Ventral hippocampal lesioning relieves anxiety-like behavior in Probability values of P <.05 were considered to represent sig the novelty-suppressed feeding test in male but not female mice. Bilat- nificant differences eral ventral hippocampal lesioning decreased feeding latency in the novelty-suppressed feeding test in male(A)but not female( B)mice despite significant influences from estrogen levels in females. Similar amount of food was consumed in two groups of male(C)and femal RESULTS (D)mice. "P<.05, "P<.01. "P<.001, unpaired t test or two-way ANOVA with Bonferronis test n 15/group for males, n 8/group for Reduced anxiety in male, but not female, mice females in HEP and LEP, respectively. HEP: high-estradiol phase. with ventral hippocampal lesions LEP: low-estradiol phase. To examine biological sex differences in hippocampal mod- ulation of anxiety-like behavior, we first performed excito- consumption were observed in either biological sex [male toxic lesions of bilateral ventral hippocampus by local t(28)=0.30, P=765, unpaired t test, Fig 3C; female injection of ibotenic acid. Fig. 2A and Fig. 2B showed sche- estrus cycle effect: F(1, 28)=0.3 33. P=. 568: lesion eff matic illustration of damages from male and female mice F(, 28)=0.05, P= 818; interaction- (r1, 28). g respectively. The lesion extended throughout bilateral ven- 899, two-way ANOVA with Bonferroni post hoc test tral hippocampus and included all cell fields, with minimal Fig. 3D], indicating similar levels of hunger and motivation extrahippocampal damages. 2 male and 4 female mice The same conclusion could be reached after food consump were excluded from analysis due to incorrect lesion site tion was normalized to body weight [male: t(28)=0.52, P resulting in 30 male and 32 female mice in final statistical analysis. Male and female mice had an average of 56% 0 25, P= 623: lesion effect: F(1. 28)=0.46, P=505 (40%-77%)and 60%(37%0-70%)bilateral tissue loss 607, unpaired t test; female: estrus cycle((. 22.5 action:F(, 28)=1.01, P=.323, two-way ANOVA with Bon- espectively. No significant differences of lesion size were ferroni post hoc test observed between males and females [t(6o)=1.07, P In the marble burying test, bilateral ventral hippocampal 291, unpaired t test]. excitotoxic lesion resulted in fewer buried marbles in male In the novelty-suppressed feeding test, bilateral ventral mice[ 28)=2.59, P=016, unpaired t test, Fig 4A], indicat hippocampal excitotoxic lesioning shortened the feeding ing decreased anxiety level. Again, we failed to observe latency in male mice [t(28)=2.21, P=.036, unpaired t test, such effects in female mice, despite a significant effect of Fig 3A], indicating reduced anxiety. By sharp contrast, in the estrus cycle [estrus cycle effect female mice, the same lesion did not produce anxiolytic P<.001: lesion effect: F( 28)=0.22, P=642; interaction: effects in this test [lesion effect: F( 28)=0.23, P=64 F(, 28)=2.55, P= 121, two-way ANOVA with Bonferroni interaction: F(1, 28)=0.38,P=541, Fig 3B], despite a sig post hoc test, Fig. 4B]. nificant effect of the estrus cycle [F(1. 28)=597.00 In the elevated plus-maze test, male mice spent more P<. 001, two-way ANOVA with Bonferroni post hoc test] time [t( 28)=0. 29, P=023, Fig. 5A] and exhibited more as previously reported(Mora et al., 1996: Seeman, 1997; entries into the open arms[t( 28)=2.53, P=018, unpaired Gangitano et al., 2009). No significant differences in food t test, Fig. 5C] after ventral hippocampal excitotoxic lesions

acquisition settings were maintained for all subsequent ima￾ging of c-Fos. For quantification of c-Fos expression, three slices were counted for ventral hippocampal CA1 (AP -3.3 mm, −3.4 mm, and − 3.5 mm relative to Bregma), basolateral amygdala (AP -1.5 mm, −1.6 mm, and − 1.7 mm relative to Bregma) and prelimbic cortex (AP 1.9 mm, 1.8 mm, and 1.5 mm relative to Bregma), respectively. c-Fos expression was quantified for each of these brain regions by ImageJ (Version 1.52, National Institutes of Health, USA). The num￾bers of c-Fos-positive neurons in left and right hemispheres was pooled since no significant differences were observed (data not shown). Statistical analysis All data were analyzed and plotted using the GraphPad Prism 6 software. Data were expressed as mean ± SEM (standard error of the mean). Males and females were analyzed sepa￾rately. The assumption of normality and equality of variance was tested by Shapiro–Wilk test and F-test, respectively. All data met the normal distribution and had equal variances in each group. Unpaired Student's t test was performed in males. Two-way ANOVA with Bonferroni post hoc test was performed in females, with lesion and estrus cycle as the two factors. Probability values of P < .05 were considered to represent sig￾nificant differences. RESULTS Reduced anxiety in male, but not female, mice with ventral hippocampal lesions To examine biological sex differences in hippocampal mod￾ulation of anxiety-like behavior, we first performed excito￾toxic lesions of bilateral ventral hippocampus by local injection of ibotenic acid. Fig. 2A and Fig. 2B showed sche￾matic illustration of damages from male and female mice, respectively. The lesion extended throughout bilateral ven￾tral hippocampus and included all cell fields, with minimal extrahippocampal damages. 2 male and 4 female mice were excluded from analysis due to incorrect lesion site, resulting in 30 male and 32 female mice in final statistical analysis. Male and female mice had an average of 56% (40%–77%) and 60% (37%–70%) bilateral tissue loss, respectively. No significant differences of lesion size were observed between males and females [t(60) = 1.07, P = .291, unpaired t test]. In the novelty-suppressed feeding test, bilateral ventral hippocampal excitotoxic lesioning shortened the feeding latency in male mice [t(28) = 2.21, P = .036, unpaired t test, Fig. 3A], indicating reduced anxiety. By sharp contrast, in female mice, the same lesion did not produce anxiolytic effects in this test [lesion effect: F(1, 28) = 0.23, P = .64; interaction: F(1, 28) = 0.38, P = .541, Fig. 3B], despite a sig￾nificant effect of the estrus cycle [F(1, 28) = 597.00, P < .001, two-way ANOVA with Bonferroni post hoc test] as previously reported (Mora et al., 1996; Seeman, 1997; Gangitano et al., 2009). No significant differences in food consumption were observed in either biological sex [male: t(28) = 0.30, P = .765, unpaired t test, Fig. 3C; female: estrus cycle effect: F(1, 28) = 0.33, P = .568; lesion effect: F(1, 28) = 0.05, P = .818; interaction: F(1, 28) = 0.02, P = .899, two-way ANOVA with Bonferroni post hoc test, Fig. 3D], indicating similar levels of hunger and motivation. The same conclusion could be reached after food consump￾tion was normalized to body weight [male: t(28) = 0.52, P = .607, unpaired t test; female: estrus cycle effect: F(1, 28) = 0.25, P = .623; lesion effect: F(1, 28) = 0.46, P = .505; inter￾action: F(1, 28) = 1.01, P = .323, two-way ANOVA with Bon￾ferroni post hoc test]. In the marble burying test, bilateral ventral hippocampal excitotoxic lesion resulted in fewer buried marbles in male mice [t(28) = 2.59, P = .016, unpaired t test, Fig. 4A], indicat￾ing decreased anxiety level. Again, we failed to observe such effects in female mice, despite a significant effect of the estrus cycle [estrus cycle effect: F(1, 28) = 108.80, P < .001; lesion effect: F(1, 28) = 0.22, P = .642; interaction: F(1, 28) = 2.55, P = .121, two-way ANOVA with Bonferroni post hoc test, Fig. 4B]. In the elevated plus-maze test, male mice spent more time [t(28) = 0.29, P = .023, Fig. 5A] and exhibited more entries into the open arms [t(28) = 2.53, P = .018, unpaired t test, Fig. 5C] after ventral hippocampal excitotoxic lesions. Fig. 3. Ventral hippocampal lesioning relieves anxiety-like behavior in the novelty-suppressed feeding test in male but not female mice. Bilat￾eral ventral hippocampal lesioning decreased feeding latency in the novelty-suppressed feeding test in male (A) but not female (B) mice, despite significant influences from estrogen levels in females. Similar amount of food was consumed in two groups of male (C) and female (D) mice. *P < .05, **P < .01, ***P < .001, unpaired t test or two-way ANOVA with Bonferroni's test. n = 15/group for males, n = 8/group for females in HEP and LEP, respectively. HEP: high-estradiol phase, LEP: low-estradiol phase. 4 Cheng Wang et al. / Neuroscience 418 (2019) xxx–xxx NSC 19240 No of Pages 9 06 September 2019

06 September 2019 ARTICLE IN PRESS No of Pages 9 Cheng Wang et al. Neuroscience 418(2019)XXx-XXX This effect was not observed in female mice [time in open AMale Female arms: estrus cycle effect: F(, 28)=23.66, P <.001; lesion 200 effect:F(n,28)=0.01,P=.973; interaction:F(1,28)=0.11 P=. 741, Fig 5B: entry into open arms: estrus cycle effect 568,P=024; lesion effect: F(1,28)=0.54,P= 469; interaction: F(, 28)=0.07, P=797, two-way ANOVA with Bonferroni post hoc test, Fig 5DI Bilateral ventral hippocampal excitotoxic lesions did not affect exploring behavior in male [t(28)=0.78, P= 441 unpaired t test, Fig. 5E] or female [estrus cycle effect: Fc 2.52. P=.124: lesion effect. 0.08,P=.778; nteraction: F(1. 28)=0.02, P=.904, Fig 5F]mice, indicated D by similar total arm entries in the plus maze. This was further confirmed by the open field test, where lesion effects were observed in neither biological sex [male: t(28)=0.86 P= 398, unpaired t test, Fig. 6A; female: estrus cycle effect: F(. 28)=0.05, P=828; lesion effect: F(1. 28) 370,P=065; interaction:F(1,28)=0.51,P=482,to way ANoVA with Bonferroni post hoc test, Fig 6B Overall, these findings indicate anxiolytic effects of ventral ppocampal excitotoxic lesions in male but not female mice HE Stronger ventral hippocampal enrollment under F innate anxiety in male than female mice Several brain regions are involved in anxiety modulation such as ventral hippocampus(Kjelstrup et al., 2002; Ban- nerman et al., 2003), amygdala(Davis, 1992; Davidson 2002)and prefrontal cortex (Jinks and Mc Gregor, 199 E Shah et al., 2004; Stern et al., 2010). One possible mechan- ism of the biological sex difference revealed above is that an anxiogenic context activated more hippocampal neurons in males than females. We observed increased c-Fos expres- LEP sion in ventral hippocampal CAl [VCA1; exposure effect F(, 31)=367. 30, P <.001: biological sex effect: F(1. 31) Fig. 5. Ventral hippocampal lesioning relieves an interaction: elevated plus-maze test in male but not female mice. Bilateral ventral hippo- Fig. 7A-C], basolateral amygdala (exposure effect: F, campal lesioning increased time spent in open arms in male (A)but not 10.82, P=.003; biological sex effect: F( 31)=4.22 female( B)mice, and increased entries into open arms in male(C)but not =048; interaction:F(1,31=273,P=.109,Fg.7DF female(D)mice. Bilateral ventral hippocampal lesioning had limited effect arm entries in male(E)or female(F) 05.*P<.01 P<.001, unpaired t test or two-way ANOVA with Bonferroni's test n 5/group for males, n=8group for females in HEP and LEP, respectively Male HEP: high-estradiol phase, LEP: low-estradiol phase. B and prelimbic cortex [exposure effect: F( 31)=277.20 P<.001; biological sex effect: F(, 31)=0.01, P=.960 interaction: F(, 31)=0.03, P= 868, Fig. 7G-I] in both male and female mice. More interestingly, we detected signifi cantly more c-Fos-positive neurons in VCAl in males than females(Fig. 7C), indicating stronger enrollment of ventral hippocampal neurons in male mice under innate anxiety LEP Fig. 4. Ventral hippocampal lesioning relieves anxiety like behavior in the marble burying test in male but not female mice. Bilateral ventral hippocam- DISCUSSION pal lesioning resulted in fewer marbles buried in male(A)but not female(B) mice, despite a significant effect of estrogen levels in females."P <.05 Substantial evidence supports sex difference in anxiety dis P<.01, "P<.001, unpaired t test or two-way ANOVA with Bonferroni's orders( Kessler et al., 1994: Nolen-Hoeksema and Girgus, test n=15/group for males, n= 8/group for females in HEP and LEP, 1994: Craske and Stein, 2016). But the neural correlates respectively. HEP: high-estradiol phase, LEP: low-estradiol phase remain insufficiently understood. In the present study, we

This effect was not observed in female mice [time in open arms: estrus cycle effect: F(1, 28) = 23.66, P < .001; lesion effect: F(1, 28) = 0.01, P = .973; interaction: F(1, 28) = 0.11, P = .741, Fig. 5B; entry into open arms: estrus cycle effect: F(1, 28) = 5.68, P = .024; lesion effect: F(1, 28) = 0.54, P = .469; interaction: F(1, 28) = 0.07, P = .797, two-way ANOVA with Bonferroni post hoc test, Fig. 5D]. Bilateral ventral hippocampal excitotoxic lesions did not affect exploring behavior in male [t(28) = 0.78, P = .441, unpaired t test, Fig. 5E] or female [estrus cycle effect: F(1, 28) = 2.52, P = .124; lesion effect: F(1, 28) = 0.08, P = .778; interaction: F(1, 28) = 0.02, P = .904, Fig. 5F] mice, indicated by similar total arm entries in the plus maze. This was further confirmed by the open field test, where lesion effects were observed in neither biological sex [male: t(28) = 0.86, P = .398, unpaired t test, Fig. 6A; female: estrus cycle effect: F(1, 28) = 0.05, P = .828; lesion effect: F(1, 28) = 3.70, P = .065; interaction: F(1, 28) = 0.51, P = .482, two￾way ANOVA with Bonferroni post hoc test, Fig. 6B]. Overall, these findings indicate anxiolytic effects of ventral hippocampal excitotoxic lesions in male but not female mice. Stronger ventral hippocampal enrollment under innate anxiety in male than female mice Several brain regions are involved in anxiety modulation, such as ventral hippocampus (Kjelstrup et al., 2002; Ban￾nerman et al., 2003), amygdala (Davis, 1992; Davidson, 2002) and prefrontal cortex (Jinks and McGregor, 1997; Shah et al., 2004; Stern et al., 2010). One possible mechan￾ism of the biological sex difference revealed above is that an anxiogenic context activated more hippocampal neurons in males than females. We observed increased c-Fos expres￾sion in ventral hippocampal CA1 [vCA1; exposure effect: F(1, 31) = 367.30, P < .001; biological sex effect: F(1, 31) = 20.84, P < .001; interaction: F(1, 31) = 25.31, P < .001, Fig. 7A-C], basolateral amygdala [exposure effect: F(1, 31) = 10.82, P = .003; biological sex effect: F(1, 31) = 4.22, P = .048; interaction: F(1, 31) = 2.73, P = .109, Fig. 7D-F] and prelimbic cortex [exposure effect: F(1, 31) = 277.20, P < .001; biological sex effect: F(1, 31) = 0.01, P = .960; interaction: F(1, 31) = 0.03, P = .868, Fig. 7G-I] in both male and female mice. More interestingly, we detected signifi- cantly more c-Fos-positive neurons in vCA1 in males than females (Fig. 7C), indicating stronger enrollment of ventral hippocampal neurons in male mice under innate anxiety. DISCUSSION Substantial evidence supports sex difference in anxiety dis￾orders (Kessler et al., 1994; Nolen-Hoeksema and Girgus, 1994; Craske and Stein, 2016). But the neural correlates remain insufficiently understood. In the present study, we Fig. 4. Ventral hippocampal lesioning relieves anxiety-like behavior in the marble burying test in male but not female mice. Bilateral ventral hippocam￾pal lesioning resulted in fewer marbles buried in male (A) but not female (B) mice, despite a significant effect of estrogen levels in females. *P < .05, **P < .01, ***P < .001, unpaired t test or two-way ANOVA with Bonferroni's test. n = 15/group for males, n = 8/group for females in HEP and LEP, respectively. HEP: high-estradiol phase, LEP: low-estradiol phase. Fig. 5. Ventral hippocampal lesioning relieves anxiety-like behavior in the elevated plus-maze test in male but not female mice. Bilateral ventral hippo￾campal lesioning increased time spent in open arms in male (A) but not female (B) mice, and increased entries into open arms in male (C) but not female (D) mice. Bilateral ventral hippocampal lesioning had limited effect on total arm entries in male (E) or female (F) mice. *P < .05, **P < .01, ***P < .001, unpaired t test or two-way ANOVA with Bonferroni's test. n = 15/group for males, n = 8/group for females in HEP and LEP, respectively. HEP: high-estradiol phase, LEP: low-estradiol phase. Cheng Wang et al. / Neuroscience 418 (2019) xxx–xxx 5 NSC 19240 No of Pages 9 06 September 2019