pmc.ncbi.nlm.nih.gov

Sexually dimorphic neuronal responses to social isolation

Abstract

Many species use social networks to buffer the effects of stress. The mere absence of a social network, however, may also be stressful. We examined neuroendocrine, PVN CRH neurons and report that social isolation alters the intrinsic properties of these cells in sexually dimorphic fashion. Specifically, isolating preadolescent female mice from littermates for <24 hr increased first spike latency (FSL) and decreased excitability of CRH neurons. These changes were not evident in age-matched males. By contrast, subjecting either males (isolated or grouped) or group housed females to acute physical stress (swim), increased FSL. The increase in FSL following either social isolation or acute physical stress was blocked by the glucocorticoid synthesis inhibitor, metyrapone and mimicked by exogenous corticosterone. The increase in FSL results in a decrease in the excitability of CRH neurons. Our observations demonstrate that social isolation, but not acute physical stress has sex-specific effects on PVN CRH neurons.

DOI: http://dx.doi.org/10.7554/eLife.18726.001

Research Organism: Mouse

eLife digest

Many species, including humans, use social interaction to reduce the effects of stress. In fact, the lack of a social network may itself be a source of stress. Recent research suggests that young girls are more sensitive to social stress than boys. This could mean that social networks are more important for females in general, and that young females from different species, such as mice, may be more sensitive to social isolation than males. However, few studies have examined how social isolation affects the brain cells that control the release of stress hormones.As such, it remains unknown whether isolating individuals from their social group impacts on the brain in sex-specific ways.

Senst, Baimoukhametova et al. now show that the brains of young male and female mice react differently to social isolation. Less than a day after separation from their littermates, the activity in the brain cells of female mice became markedly different from that of isolated males. In contrast to social isolation, the physical stress of being made to swim produced similar changes in the brains of both male and female mice. Further experiments then showed that the changes in the brain cells that control the release of stress hormones required a signalling chemical called corticosterone, which is produced in response to stressful situations. This suggests that, in repsonse to soical isolation, the females are experiencing more stress than the males.

Following on from this work, one future challenge will be to investigate if reuniting a social group erases the effects of social isolation on the brain. Further experiments could also examine the behavioural and physiological effects of social isolation, including how females respond to later stressful events.

DOI: http://dx.doi.org/10.7554/eLife.18726.002

Introduction

Survival threats trigger a repertoire of behavioral and endocrine adjustments (Cannon, 1932). The endocrine changes rely on the immediate engagement of the hypothalamic pituitary adrenal (HPA) axis which is controlled by corticotropin releasing hormone (CRH) neurons in the paraventricular nucleus of the hypothalamus (PVN). Although it is essential for survival, the dysregulation of the stress response is implicated in the emergence of numerous neuropsychiatric diseases (Tost et al., 2015). Interestingly, many of these psychopathologies show sex-specific differences which may be a consequence of sexual dimorphism of the neuroendocrine response to stress (Bale and Epperson, 2015). This sexual dimorphism may result in different response sensitivity to the same stressor and/or differences in which stimuli males and females perceive as stressful. Human studies demonstrating that young girls exhibit greater corticosteroid (CORT) stress reactivity to social stress tests than boys (Gunnar et al., 2009; de Veld et al., 2012) suggest, for example, that young females are more responsive to changes in social situations. Consistent with this idea, are demonstrations that, in comparison to males, females take greater advantage of social support and group dynamics to manage stress (Taylor et al., 2000). Here we hypothesized that disrupting the social network would elicit neurobiological changes preferentially in female mice.

Results

To test this idea, mice were housed in same-sex groups (3–5 animals per group) from weaning (p21) until the day of the experiment (p22–35). This developmental window allows us to interrogate the contributions of sex differences independent of the effects of circulating gonadal hormones (Nelson et al., 1990). Mice were either in same-sex groups, pairs, or isolated from littermates for sixteen to eighteen hours. We then prepared hypothalamic brain slices and examined synaptic and intrinsic properties of identified CRH neurons in the PVN in acute brain slices (Wamsteeker Cusulin et al., 2013b) (Figure 1a). Activation of these cells is obligatory for launching the neuroendocrine response to stress (Denver, 2009).

Figure 1. Social isolation alters intrinsic properties of CRH neurons in female, not male mice.

(a) Experimental timeline (top), schematic (middle) showing coronal brain section of PVN (red) with a fluorescent image of tdTomato immunofluorescence (bottom). Panels on right show DIC image with electrode (top), tdTomato cells (middle) and recorded neuron (bottom) (b) Summary bar graphs indicate no difference in FSL for single and group-housed males (left, one-way ANOVA, p=0.9). FSL in single females is significantly longer than group-housed females (one-way ANOVA, p<0.0001). FSL in group-housed females and males are not different (p>0.05). (c) Traces show responses to a single +80 pA depolarizing step (holding potential = −102 mV) in group (dark blue) and single-housed (light blue) males. FSL calculated from the start of the depolarizing pulse (dotted line) to the first spike is plotted for all cells in (c). The relative frequency distributions are overlaid by the relative cumulative distributions. (d) Traces show responses to the same depolarizing step protocol as in (c) but in group (dark orange) and single-housed (light orange) females. The relative frequency distributions are overlaid by the relative cumulative distributions.

DOI: http://dx.doi.org/10.7554/eLife.18726.003

Figure 1.

Figure 1—figure supplement 1. No sex differences in basal properties of CRH neurons.

Figure 1—figure supplement 1.

(a) Left: Input resistance of individual cells from isolated males and females (male: 797 ± 32 MΩ, n = 157, vs female: 868 ± 23 MΩ, n = 273; unpaired t-test, p=0.06). Right: Spike thresholds of individual cells from isolated males and females (male: −51.5 ± 1.0 mV, n = 63 vs female: −51.3 ± 0.7 mV, n = 99; unpaired t-test, p=0.86).

Figure 1—figure supplement 2. No sex differences in excitatory post-synaptic currents.

Figure 1—figure supplement 2.

(a) Representative traces of spontaneous synaptic activity from individual cells from single-housed males and females with bath application of picrotoxin (100 µM). (b) Summarized data of cumulative fraction of amplitude (pA) and inter-event intervals. There were no significant differences in either the spontaneous amplitude (males: 34.14 ± 1.30pA, n = 12; females: 33.10 ± 1.19pA, n = 45, p>0.05) or frequency (males: 2.99 ± 0.61 Hz, n = 12; females: 2.50 ± 0.25 Hz, n = 45, p>0.05). (c) Peak scaled sample trace of a paired-pulse recording from a male (blue) and a female (orange). Right: Summary data of PPRs recorded from males and females. There was no significant difference in the paired pulse ratio (PPR) of EPSCs (males: 0.72 ± 0.05, n = 9; females: 0.80 ± 0.04, n = 32, p>0.05). Scale bars are 50 pA/10 ms.

Figure 1—figure supplement 3. No sex differences in inhibitory post-synaptic currents.

Figure 1—figure supplement 3.

(a) Representative traces of spontaneous synaptic activity from individual cells from males and females with bath application of DNQX (10 µM). (b) Summarized data of cumulative fraction of amplitude (pA) and inter-event intervals. There were no differences in sIPSC amplitude (males: 27.61 ± 0.89 pA, n = 43; females: 26.77 ± 0.76 pA, n = 66, p>0.05) or frequency (males: 2.25 ± 0.27 Hz, n = 43; females: 1.85 ± 0.14 Hz, n = 66, p>0.05). (c) Peak scaled sample trace of a paired-pulse recording from a male (blue) and a female (orange). Right: Summary data of PPRs recorded from males and females. There was no significant difference in the PPR of IPSCs (males: 0.65 ± 0.02, n = 29; females: 0.70 ± 0.02, n = 57, p>0.05. Scale bars are 50 pA/10 ms.

The cellular heterogeneity in the PVN has necessitated various approaches to distinguish cell types. One of the most widely used has been to identify cells based on their voltage responses to a protocol of current steps. This reveals an ‘electrical fingerprint’ that has been used to distinguish among magnocellular (oxytocin, vasopressin) (Tasker and Dudek, 1991) and parvocellular (CRH, TRH, somatostatin) neurosecretory cells (Hoffman et al., 1991) and pre-autonomic neurons. Reports of changes in the intrinsic properties of neurons following experimental manipulations ex vivo (Shah et al., 2010; O'Leary et al., 2014; Kourrich et al., 2015), however, raise the possibility that electrical fingerprints are dynamic and potentially unreliable as a classification tool. We have recently described a transgenic reporter mouse in which CRH neurons can be identified and targeted directly for electrophysiological recording (Wamsteeker Cusulin et al., 2013b). In male mice, the electrical fingerprint of CRH cells was consistent with previous reports on parvocellular neurosecretory cells (Wamsteeker Cusulin et al., 2013b). These cells had a linear current voltage relationship and, unlike the magnocellular neurosecretory cells, did not have a long delay to first spike (first spike latency (FSL)) when depolarized from hyperpolarizing membrane potentials.

We first examined FSL of CRH neurons from group-housed vs isolated mice. FSL can play a key role in integrating synaptic events and influence probability of firing (Molineux et al., 2005) and changes in the underlying currents can tune neuronal frequency (Ellis et al., 2007). Social isolation had no effect on FSL in males (malegroup: 48.7 ± 2.1 ms, n = 53 vs malesingle: 46.1 ± 1.8 ms, n = 76, p=0.9; Figure 1b,c). By contrast, FSL in isolated females was significantly longer than FSL in group-housed females (femalegroup: 45.1 ± 1.9 ms, n = 42; femalesingle: 65.6 ± 1.9, n = 177, p<0.0001; Figure 1b,d). We then conducted experiments in which female mice were housed in pairs prior to electrophysiology experiments. This manipulation revealed an intermediate phenotype, with pair-housed females exhibiting FSLs that were longer than those observed in group housed female mice, but shorter than those observed in single-housed female mice (femalepaired: 56.5 ± 2.0 ms, n = 105, p=0.0029 vs femalesingle and p=0.0017 vs femalegroup, F4,471 = 10.6). There was no difference in the input resistance of isolated males vs isolated females (malesingle: 796 ± 31 MΩ, n = 157, vs femalesingle: 869 ± 23 MΩ, n = 273; unpaired t-test, p=0.063; Figure 1—figure supplement 1) or spike threshold (malesingle: −51.5 ± 1.0 mV, n = 63 vs femalesingle: −51.3 ± 0.7 mV, n = 99; unpaired t-test, p=0.86; Figure 1—figure supplement 1). We also examined properties of glutamate and GABA synapses on CRH neurons in hypothalamic slices from single-housed females and males. We observed no differences in basal glutamate (Figure 1—figure supplement 2) or GABA (Figure 1—figure supplement 3) synaptic transmission. These observations indicate that social isolation alters the intrinsic properties of CRH neurons in female, but not age-matched male mice.

Next, we hypothesized that female mice, but not male mice, may interpret social isolation as a stress. If true, we reasoned that subjecting single-housed male mice to an acute swim stress should also increase FSL. In females, however, the prior isolation should occlude the effects of an acute stress. To test this idea, single -housed male and female mice were subjected to swim stress for 20 min and intrinsic properties of CRH neurons were assessed (Figure 2a). Following swim stress, FSL in male mice (maleswim) was 64.0 ± 4.6 ms, n = 37, Figure 2b,c; this is significantly longer than FSL in the single males reported above (maleswim vs malesingle, 1 way ANOVA, p=0.0004). FSL in single female mice subjected to swim stress was 65.8 ± 4.9 ms, n = 32, Figure 2b,c; this is not different from FSL in female mice subjected to social isolation (femaleswim vs femalesingle, p=0.67). Finally, FSL in single males subjected to swim was not different from FSL in single females subjected to swim (p=0.9, Figure 2b,c). This suggests that either increases in FSL following social isolation occlude further effects of stress on CRH neurons or that females do not respond to swim stress. To test the effects of swim stress independent of any potential effects of social isolation, we conducted experiments in which group-housed females and males were subjected to swim stress. Following swim, FSL in group-housed males (malegroup swim: 58.8 ± 2.8 ms, n = 23, Figure 2d) was significantly longer than FSL in naïve, group -housed males (p<0.001). FSL of group-housed males subjected to swim was not different than FSL in females subjected to swim (femalegroup swim:58.7 ± 1.9 ms, n = 27, p>0.05 vs malegroup, Figure 2d). Similarly, CRH neurons in group-housed females subjected to swim had a significantly longer FSL than CRH neurons in naive, group-housed females (p=0.0001). These observations demonstrate that both females and males show equivalent sensitivity to an acute physical stress (i.e. forced swim). Additionally, a prior social isolation occludes the effects of subsequent physical stress on FSL in females, but has no effect on FSL in males.

Figure 2. Females and males show equivalent sensitivity to an acute physical stress.

Figure 2.

(a) Experimental timeline. (b) Neuronal responses to +80 pA depolarizing pulse in single-housed males and females. (c) Summary data show that following swim stress, FSL in single males and single females is not different (p>0.05). FSL in single males subjected to swim is significantly longer than FSL in naive single males (p<0.0001). FSL in single females subjected to swim is not different than single females (p>0.05). (d) Summary data show that FSL in group-housed males and females is not different (p>0.05). Group-housed males subjected to swim have longer FSL than group-housed, naive males (p<0.001) and group-housed females subjected to swim have longer FSL than group-housed, naive females (p=0.0001).

DOI: http://dx.doi.org/10.7554/eLife.18726.007

Since stress increases CORT, we examined the relationship between circulating CORT and FSL. In order to assess CORT during a defined temporal window, we subjected single male mice to swim stress and obtained samples thirty minutes after the protocol. We noted a positive correlation between plasma CORT and FSL (r2 = 0.3, p=0.024, n = 17, Figure 3a). CORT, however, is one of many signaling molecules that is altered in response to stress. To determine whether CORT is necessary for increases in FSL following social isolation, we conducted experiments in which female mice were given access to drinking water containing the CORT-synthesis inhibitor, metyrapone prior to, and during, isolation (Figure 3b). Metyrapone treatment blocked the effects of social isolation on FSL in female mice (femalesingle metyrapone = 47.2 ± 2.0 ms, n = 24, p=0.001 vs femalesingle, Figure 3c). Next, to determine whether CORT is necessary for increases in FSL following swim stress, we administered metyrapone prior to swim stress in single males (Figure 3d). This eliminated the stress-induced increase in FSL in males (malemetyrapone = 50.2 ± 2.4 ms, n = 35, vs malevehicle = 66.7 ± 2.4, n = 23, p=0.007, Figure 3e). These observations demonstrate that CORT is necessary for stress-induced changes in FSL in both males and females. In order to determine whether CORT is sufficient for increasing FSL, we incubated slices from either single-housed males or group-housed females with 100 nM CORT for 1 hr prior to electrophysiological assessment. This incubation time is sufficient to mimic CORT-dependent changes in synaptic metaplasticity observed after acute stress (Wamsteeker Cusulin et al., 2013a). CORT-incubation increased FSL in slices from single-housed males (59.2 ± 4.1 ms, n = 36, p<0.021 vs malesingle) and in slices from group-housed females (64.1 ± 4.8 ms, n = 32, p<0.0001 vs femalegroup, Figure 3f). Collectively, these observations demonstrate that CORT is both necessary (during stress) and sufficient (in the absence of stress) to increase FSL in male and female mice.

Figure 3. CORT is necessary and sufficient for increasing FSL.

Figure 3.

(a) CORT measurements in male mice subjected to stress are plotted against FSL. (b) Experimental timeline. (c) FSL following pre-treatment with the CORT synthesis inhibitor metyrapone in drinking water prior to and during isolation in females (orange). The gray bar shows data from Figure 1 and is shown here for comparative purposes. (d) Experimental timeline. (e) Left bar shows FSL in male mice pretreated with Met prior to swim stress. Right bar shows FSL in male mice pretreated with vehicle prior to swim stress. (f) FSL is increased following incubation of slices from single males (p<0.021 vs malesingle) or group-housed females (p=0.002 vs femalegroup) with CORT. There is no significant difference in FSL between the two CORT-treated groups.

DOI: http://dx.doi.org/10.7554/eLife.18726.008

Next we investigated the underlying conductance that controls FSL in single-housed males and females. A rapidly activating, rapidly inactivating potassium (K) conductance contributes to FSL in a number of different cell types throughout the nervous system (Meng et al., 2011). The channels responsible for this conductance are largely inactive at resting membrane potential, but this inactivation can be removed by a membrane hyperpolarization (Yellen, 2002; Maffie and Rudy, 2008). To test whether similar K channels contribute to the FSL in CRH neurons, we conducted experiments in which we did not deliver a membrane hyperpolarization prior to a depolarizing current step. The absence of this hyperpolarizing pre-pulse decreased the FSL in CRH neurons from socially-isolated females (FSLnoHP: 37.8 ± 3.1 ms vs FSLHP: 55.3 ± 4.5 ms, paired t-test, p=0.0022, Figure 4—figure supplement 1). Rapidly activating and inactivating K currents are sensitive to millimolar concentrations of 4-aminopyridine (4-AP) (Alexander et al., 2015), and consistent with this, we observed shorter FSL in the presence of 2 mM 4-AP (FSL4-AP: 36.8 ± 5.6 ms vs FSLcontrol: 62.8 ± 7.6 ms unpaired t-test, p=0.015 n = 8, Figure 4a). We obtained voltage clamp recordings and characterized the currents responsible for regulating FSL (malesingle n = 11, femalesingle n = 12). These currents were largely inactive at resting membrane potential, inactivated quickly and required membrane hyperpolarization to relieve inactivation (Figure 4b). They were partially blocked by 2 mM 4-AP, fully blocked by 6 mM 4-AP (n = 4, Figure 4c), but were unaffected by a lower concentration of 4-AP(500 uM, data not shown) (Anderson et al., 2010). In addition, these currents were insensitive to 20 mM TEA (data not shown). Neurons from socially-isolated females showed no difference in activation or recovery from inactivation in comparison to isolated males (Figure 4d). There was no difference in current density in socially-isolated females when compared to socially-isolated males (data not shown), arguing against an increase in channel number or conductance. We next asked whether channel kinetics might explain the differences in FSL. There was no difference in the activation kinetics between males and females, but currents from female mice had a longer decay time constant in comparison to males (τfemale single: 55.8 ± 5.6 ms, n = 11, vs τmale single: 41.9 ± 2.9 ms, n = 11, unpaired t-test, p=0.019, Figure 4e,f). These findings indicate that a slowing in the decay of a rapidly activating, rapidly inactivating voltage-gated K current is causative for the increase in FSL.

Figure 4. Slower decay of 4-AP sensitive, voltage-gated K current in isolated females.

(a) FSL is decreased by bath application of 2 mM 4-AP. (b) Voltage clamp traces from a male CRH neuron in response to different current step protocols as shown. (c) The rapidly activating and inactivating current was completely blocked by 6 mM 4AP. (d) The activation and inactivation curves are shown for single-housed females and males. (e) Currents at half-maximal activation, fit with a double-exponential are shown from males and females. (f) The slow inactivation time constant in malesingle and femalesingle is summarized (unpaired t-test, p=0.02).

DOI: http://dx.doi.org/10.7554/eLife.18726.009

Figure 4.

Figure 4—figure supplement 1. FSL is sensitive to membrane hyperpolarization prior to depolarization.

Figure 4—figure supplement 1.

(a) Traces on left show voltage responses to a depolarizing current step with (orange) and without (gray) a hyperpolarizing pre-pulse. The absence of a hyperpolarizing pre-pulse decreases FSL as summarized in b (n = 29, p<0.01).

Next, we asked whether these differences in the biophysical properties of the rapidly inactivating K current observed in socially-isolated females compared to males would be sufficient to alter the spike output of CRH neurons. We conducted experiments in which depolarizing current steps (20 pA intervals) were preceded by a hyperpolarizing pre-pulse. CRH neurons from female isolated mice required a greater depolarization to generate a spike compared to CRH neurons from socially isolated males. This is revealed as a rightward shift in the relationship between spike probability and current step (female: I50 = 49.5 ± 1.2 mV, n = 104 vs male: I50 = 55.0 ± 1.6 mV, n = 88, p<0.0001, K-S statistic). Finally, we compared the excitability of CRH neurons from isolated females and males. We used a protocol (hyperpolarizing pre-pulse followed by a family of depolarizing steps) that allowed for maximal activation of the rapidly activating K current described above. The rightward shift in the F-I plot indicates that CRH neurons from isolated females had lower excitability in comparison to isolated males (p<0.0001, K-S statistic, Figure 5b).

Figure 5. Longer FSL decreases spike probability and neuronal firing.

Figure 5.

(a) Traces show firing properties of CRH neurons from individually housed male (blue) and female (orange) mice. Right panel shows the probability of eliciting a spike in response to varying current steps (K-S statistic, p<0.0001). (b) Responses from male and female CRH neurons to varying current steps. Right panel shows spike frequency in the first 50 ms of current step (K-S statistic, p<0.0001).

DOI: http://dx.doi.org/10.7554/eLife.18726.011

Discussion

Our findings demonstrate that brief social isolation (<24 hr) affects the biophysical properties of PVN CRH neurons in a sexually dimorphic fashion. We report an increase in FSL and a decrease in neuronal excitability in socially-isolated female, but not in age-matched, male mice. By contrast, an acute physical stress (swim) increased FSL in group-housed females and single-housed males. In addition, social isolation occluded the effects of swim stress on female mice. CORT is both necessary for the isolation (females) and swim stress (males and females) induced increases in FSL and sufficient to increase FSL in the absence of stress. Finally, we noted the effect of social isolation was graded in female mice with paired mice exhibiting FSLs that were intermediate to those observed from group-housed or single mice.

We examined the underlying conductance and report that it is sensitive to millimolar concentrations of 4-AP and insensitive to millimolar concentrations of TEA. In addition, FSL was profoundly sensitive to pre-hyperpolarizing pulses, which remove channel inactivation and make rapidly inactivating K channels available for opening in response to depolarizations. The features of these channels are consistent with Kv4 channels previously described in the PVN (Lee et al., 2012). They do, however, exhibit a more hyperpolarized activation threshold in comparison to kinetically similar 4-AP sensitive channels reported in PVN parvocellular neurosecretory cells in rat (Luther and Tasker, 2000). We failed to note any differences in overall current density or voltage-dependent activation or recovery from inactivation. We did, however, observe an increase in the decay time constant of rapidly inactivating K currents in socially-isolated female mice compared to male mice, providing a potential underlying mechanism for the prolonged FSL in response to depolarization. The longer FSL we observed is not a classically defined characteristic of CRH neurons and, in fact, the absence of this delay has been used previously to categorize putative CRH neurons (Hoffman et al., 1991; Wamsteeker Cusulin et al., 2013b).

Our findings demonstrate that the biophysical characteristics of neurons are flexible and exquisitely sensitive to specific experiences. Whilst synaptic changes in the PVN (Inoue et al., 2013; Kuzmiski et al., 2010; Wamsteeker Cusulin et al., 2013a) and other brain regions have been investigated, this is, to the best of our knowledge, one of the first reports of rapid intrinsic plasticity in neurons following acute stress. Our observations add to a number of observations that intrinsic (synapse autonomous) plasticity can have a profound effect on the output of neural networks (Shah et al., 2010; Kourrich et al., 2015;O'Leary et al., 2014). In addition, as electrical fingerprints have been used to identify specific cell types, investigators should use caution when interpreting the data as the physiological or psychological state of the animal may have a dramatic impact on these fingerprints. Similarly, since neurons in the PVN can express multiple neuromodulators, caution must also be exercised when interpreting data from a genetically identified cell population. We have previously shown, using immunohistochemistry, that overlap between the genetically identified CRH cell population in PVN and non-CRH neurons is minimal (<5% for vasopressin/CRH or oxytocin/CRH; <<1% for thyrotropin releasing hormone/CRH or somatostatin/CRH)(Wamsteeker Cusulin et al., 2013b).

These findings demonstrate that males and females react differently to some stressors but not others. Specifically, social isolation has definitive, CORT-dependent effects on CRH neurons in female, but not male mice. By contrast, both males and females show indistinguishable responses to an acute physical swim stress. These findings highlight the importance of carefully considering both the sex of the animal and the stressor modality when designing studies investigating neurobiological effects of stress. Indeed recent work demonstrating sex-specific expression of fear responses in rodents (Gruene et al., 2015) serves as a corollary to our observations and when taken together suggests that strategies for coping with stress are sex-specific. In particular, our findings highlight the importance of a social network in females and provides a neurobiological framework for further testing the underlying thesis of ‘tend and befriend’ strategy for females in response to stress (Taylor et al., 2000). Finally, as our observations report sex-specific differences in response to stress in the first two weeks post-weaning, they provide clear evidence for biological effects on the stress axis that are independent of circulating gonadal hormones (McCarthy and Arnold, 2011). This exquisite sensitivity of the brain during the pre-adolescent period (Foilb et al., 2011) raises intriguing questions about the long-term consequences of subtle environmental and social manipulations on stress responses and behavior in males and females later in life.

Materials and methods

Animal care and stress paradigm

All protocols received approval from the University of Calgary Animal Care and Use Committee in accordance with the guidelines of the Canadian Council on Animal Care guidelines (Protocol # AC13-0027). B6(Cg)-Crhtm1(cre)Zjh/J (Crh-IRES-Cre)mice and B6.Cg-Gt(ROSA)26Sortm14(CAG-TdTomato)Hze/J (Ai14) mice, whose generation has been detailed previously (Madisen et al., 2010; Taniguchi et al., 2011), were obtained from Jackson Laboratories (stock number 012704 and 007914 respectively). These were maintained as colonies of homozygous mice, with one backcrossing to C57BL/6J background strain following their arrival. Genotyping was used to identify mutants using PCR procedures provided by the supplier. The following primers were used to identify Crh-IRES-Cre mutants: 5′-CTT ACA CAT TTC GTC CTA GCC and 5′- CAA TGT ATC TTA TCA TGT CTG GAT CC-3′ and (468 base pair resultant PCR band). To identify Ai14 mutants: 5′-GGC ATT AAA GCA GCG TAT CC-3′ and 5′-CTG TTC CTG TAC GGC ATG G -3′ were used (196 base pair band). The age of pre-adolescent mice (post-natal day 21–35) was determined according to previous literature demonstrating that in C57/BL mice the onset of puberty occurs at approximately 5 weeks of age (Nelson et al., 1990; Mayer et al., 2010). Mice were individually housed on a 12 hr:12 hr light:dark cycle (lights on at 7:00) with ad libitum access to food and water. Pairs of either homozygous Crh-IRES-Cre or Ai14 genotypes were mated, and the resulting heterozygous Crh-IRES-Cre;Ai14 offspring used in subsequent experiments. Sixteen hours prior to the acute stress protocol or slice preparation, mice were housed either individually, in same-sex pairs or same-sex groups (3–5 mice per group). Single or group -housed mice were randomly assigned to naive or stress conditions. For stress experiments, mice were exposed to a forced swim stress (between 8:00 and 9:30 during the light phase) consisting of 20 min in a glass cylinder (14 cm internal diameter) filled with 30–32°C water. Following one hour of recovery in their home cage, mice were anesthetized and brain slices were prepared as described below.

Slice preparation

Experimental animals were anaesthetized with isoflurane and decapitated. The brain was quickly removed, and coronal brain slices (250 μm) containing the PVN of the hypothalamus were obtained using a vibrating slicer (Leica, Nussloch, Germany) while submerged in ice cold slicing solution (0°C, 95% O2/5% CO2 saturated), containing (in mM): 87 NaCl, 2.5 KCl, 0.5 CaCl2, 7 MgCl2, 25 NaHCO3, 25 D‐glucose, 1.25 NaH2PO4, 75 sucrose. Slices were then allowed a recovery period, of a minimum 60 min, in artificial cerebrospinal fluid (aCSF) (32.5°C, 95% O2/5% CO2 saturated) containing (in mM): 126 NaCl, 2.5 KCl, 26 NaHCO3, 2.5 CaCl2, 1.5 MgCl2, 1.25 NaH2PO4, 10 glucose. The CORT synthesis inhibitor, metyrapone was dissolved in polyethylene glycol and injected i.p. 60 min prior to swim stress at a dose of 75 mg/kg in a volume of 50 μL. For female isolation experiments, CORT was dissolved in the drinking water (500 μg/ml) and given for 24 hr prior to isolation and during the entire isolation period.

Electrophysiology

Hypothalamic slices were transferred to a recording chamber and superfused with 30–32°C aCSF at a flow rate of 1–2 ml/min. Slices were visualized using an upright microscope (BX51WI, Olympus) fitted with infrared differential interference contrast optics. CRH neurons were identified by their expression of tdTomato. Whole‐cell patch clamp recordings were obtained from CRH neurons using borosilicate glass microelectrodes with tip resistance between 2–5 MΩ. The normal intracellular solution contained (in mM): 108 K‐gluconate, 2 MgCl2, 8 Na‐gluconate, 8 KCl, 1 K2‐EGTA, 4 K2‐ATP, and 0.3 Na3‐GTP buffered with 10 mM HEPES. For images of filled cells, 0.2 mM Alexa-488 hydrazide and 10 mg·mL−1 biocytin were added to internal solution. Microscope images were captured using a Retiga EXi camera (Qimaging) and processed using ImageJ.

Recordings were amplified using a Multiclamp 700B amplifier (Molecular Devices, Union City, CA), low‐pass filtered at 1 kHz, and digitized at 10 kHz using the Digidata 1322 (Molecular Devices). Data were recorded (pClamp 9.2; Molecular Devices) and stored on a computer for offline analysis. During all experiments, initial access resistance (Ra) was below 20 MΩ. Cell membrane properties were monitored for the duration of the experiments and only recordings in which changes to Ra Cm did not exceed 15% were accepted for analysis. For synaptic experiments, the cell membrane was voltage clamped at −80 mV. Spontaneous inhibitory ionotropic GABAA receptor postsynaptic currents (sIPSCs) were isolated by blocking AMPA‐ and kainate‐receptor‐mediated glutamatergic synaptic transmission with 6,7‐dinitroquinoxaline‐2,3‐dione (DNQX, 10 μM). When measuring spontaneous excitatory postsynaptic currents (EPSCs), the GABAA channel blocker picrotoxin (100 μM) was included in the bath to isolate excitatory currents mediated by AMPA and kainate receptors. These spontaneous currents representing stochastic transmitter release were analyzed using MiniAnalysis 6.0.3 (Synaptosoft, Decatur, GA). Event detection was set at three times the baseline noise and confirmed as synaptic events by eye.

To determine RMP, cells were recorded in zero current (I = 0) mode. Latency to spike initiation and firing thresholds were both measured using a current clamp depolarization step protocols. First, a baseline current injection that maintained the membrane voltage near −80 mV was chosen individually for each cell, which served to exclude the possible confounding influence of variable resting membrane potentials between cells. Next, a 200 ms/30 pA hyperpolarizing current injection was given followed by 250 ms/20 pA depolarizing current steps up to 140 pA. The latency time was measured as the duration from the point of initiating the depolarizing pulse to the initiation of the first spike. Firing threshold was determined as the membrane potential at the initiation of the first spike. Both firing threshold and membrane potential are corrected for a liquid junction potential of 12 mV, as calculated with solution ion concentrations.

In experiments examining cell properties following 4-AP administration, baseline recordings were obtained, and then a ten-minute treatment period was allowed before secondary data were obtained. Synaptic currents were evoked by paired afferent stimulation (every 5 s with an interstimulus interval of 50 ms) and analyzed using Clampfit 9.2 (Molecular Devices). Evoked postsynaptic current (ePSC) amplitudes were calculated from the baseline (current before the first evoked response) to peak of each evoked response. The paired pulse ratio (PPR) was calculated using the ratio of the amplitudes of the evoked pair (peak 2/peak 1) from a minimum of a one-minute epoch within each cell. To isolate the voltage gated fast inactivating K current the following cocktail was applied to the slices for at least of 15 minutes prior recording: Bicuculline 10 µM, DNQX 10 µM, dAPV 50 µM, TTX 1 µM, TEA 20 mM.

Data analysis

Each group represents a minimum of three animals for pharmacology experiments, or a minimum of four for stress experiments. Data points are presented as mean ± SEM. Statistical analyses for sIPSCs, and current clamp step data were performed in GraphPad Prism 4 using a one way ANOVA for multiple groups followed by a post-hoc Tukey’s multiple comparisons test; unpaired student's t-test were used for two group comparisons, and a K-S statistic for comparing two distributions.

Drugs

Drugs were dissolved into aCSF daily prior to experiments from frozen aliquots stored at −20°C and added to the bath by perfusion pump. The drugs were dissolved in accordance with guidelines either in DMSO, PEG, ethanol or distilled water. DNQX, 4-AP, d APV and metyrapone were obtained from Tocris (Tocris Cookson, Ellisville, MO). Picrotoxin, bicuculline, TEA and corticosterone were obtained from Sigma (Sigma‐ Aldrich, St. Louis, MO). TTX was obtained from Alomone labs (Jerusalem BioPark (JBP), Hadassah Ein Kerem, P.O.Box 4287 Jerusalem 9104201, Israel).

Acknowledgements

We thank all members of the Bains lab for thoughtful discussions and input. We are grateful to Ms. Cheryl Sank technical assistance and Mr. Rodney Barasi for animal husbandry. This work was supported by a CIHR Operating Grant (86501) to JSB. LS is supported by the Leaders in Medicine program in the Cumming School of Medicine. TLS is supported by a University of Calgary Eyes High Postdoctoral Fellowship and an Alberta Innovates Health Solutions Fellowship.

Funding Statement

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Funding Information

This paper was supported by the following grants:

  • Leaders in Medicine, Cumming School of Medicine to Laura Senst.

  • Alberta Innovates - Health Solutions Postdoctoral Fellowship to Toni-Lee Sterley.

  • University of Calgary Eyes High Postdoctoral Fellowship to Toni-Lee Sterley.

  • Canadian Institutes of Health Research 86501 to Jaideep Singh Bains.

  • Fondation Brain Canada to Jaideep Singh Bains.

Additional information

Competing interests

The authors declare that no competing interests exist.

Author contributions

LS, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article.

DB, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article.

T-LS, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article.

JSB, Conception and design, Analysis and interpretation of data, Drafting or revising the article.

Ethics

Animal experimentation: All experiments were approved by the University of Calgary Animal Care and Use Committee in accordance with Canadian Council on Animal Care guidelines: Protocol # AC13-0027.

References

  1. Alexander SP, Catterall WA, Kelly E, Marrion N, Peters JA, Benson HE, Faccenda E, Pawson AJ, Sharman JL, Southan C, Davies JA, CGTP Collaborators The concise guide to PHARMACOLOGY2015/16: voltage-gated ion channels. British Journal of Pharmacology. 2015;172:5904–5941. doi: 10.1111/bph.13349. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Anderson D, Mehaffey WH, Iftinca M, Rehak R, Engbers JD, Hameed S, Zamponi GW, Turner RW. Regulation of neuronal activity by Cav3-Kv4 channel signaling complexes. Nature Neuroscience. 2010;13:333–337. doi: 10.1038/nn.2493. [DOI] [PubMed] [Google Scholar]
  3. Bale TL, Epperson CN. Sex differences and stress across the lifespan. Nature Neuroscience. 2015;18:1413–1420. doi: 10.1038/nn.4112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Cannon WB. The Wisdom of the Body. New York: WW Norton & Co; 1932. [Google Scholar]
  5. de Veld DM, Riksen-Walraven JM, de Weerth C. The relation between emotion regulation strategies and physiological stress responses in middle childhood. Psychoneuroendocrinology. 2012;37:1309–1319. doi: 10.1016/j.psyneuen.2012.01.004. [DOI] [PubMed] [Google Scholar]
  6. Denver RJ. Structural and functional evolution of vertebrate neuroendocrine stress systems. Annals of the New York Academy of Sciences. 2009;1163:1–16. doi: 10.1111/j.1749-6632.2009.04433.x. [DOI] [PubMed] [Google Scholar]
  7. Ellis LD, Krahe R, Bourque CW, Dunn RJ, Chacron MJ. Muscarinic receptors control frequency tuning through the downregulation of an A-type potassium current. Journal of Neurophysiology. 2007;98:1526–1537. doi: 10.1152/jn.00564.2007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Foilb AR, Lui P, Romeo RD. The transformation of hormonal stress responses throughout puberty and adolescence. Journal of Endocrinology. 2011;210:391–398. doi: 10.1530/JOE-11-0206. [DOI] [PubMed] [Google Scholar]
  9. Gruene TM, Flick K, Stefano A, Shea SD, Shansky RM. Sexually divergent expression of active and passive conditioned fear responses in rats. eLife. 2015;4:e18726. doi: 10.7554/eLife.11352. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Gunnar MR, Wewerka S, Frenn K, Long JD, Griggs C. Developmental changes in hypothalamus-pituitary-adrenal activity over the transition to adolescence: normative changes and associations with puberty. Development and Psychopathology. 2009;21:69–85. doi: 10.1017/S0954579409000054. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Hoffman NW, Tasker JG, Dudek FE. Immunohistochemical differentiation of electrophysiologically defined neuronal populations in the region of the rat hypothalamic paraventricular nucleus. The Journal of Comparative Neurology. 1991;307:405–416. doi: 10.1002/cne.903070306. [DOI] [PubMed] [Google Scholar]
  12. Inoue, Baimoukhametova DV, Füzesi T, Wamsteeker Cusulin JI, Koblinger K, Whelan PJ, Pittman QJ, Bains JS. Noradrenaline is a stress-associated metaplastic signal at GABA synapses. Nature Neuroscience. 2013;16:605–612. doi: 10.1038/nn.3373. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Kourrich S, Calu DJ, Bonci A. Intrinsic plasticity: an emerging player in addiction. Nature Reviews Neuroscience. 2015;16:173–184. doi: 10.1038/nrn3877. [DOI] [PubMed] [Google Scholar]
  14. Kuzmiski JB, Marty V, Baimoukhametova DV, Bains JS. Stress-induced priming of glutamate synapses unmasks associative short-term plasticity. Nature Neuroscience. 2010;13:1257–1264. doi: 10.1038/nn.2629. [DOI] [PubMed] [Google Scholar]
  15. Lee SK, Lee S, Shin SY, Ryu PD, Lee SY. Single cell analysis of voltage-gated potassium channels that determines neuronal types of rat hypothalamic paraventricular nucleus neurons. Neuroscience. 2012;205:49–62. doi: 10.1016/j.neuroscience.2011.12.031. [DOI] [PubMed] [Google Scholar]
  16. Luther JA, Tasker JG. Voltage-gated currents distinguish parvocellular from magnocellular neurones in the rat hypothalamic paraventricular nucleus. The Journal of Physiology. 2000;523 (Pt 1):193–209. doi: 10.1111/j.1469-7793.2000.t01-1-00193.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Madisen L, Zwingman TA, Sunkin SM, Oh SW, Zariwala HA, Gu H, Ng LL, Palmiter RD, Hawrylycz MJ, Jones AR, Lein ES, Zeng H. A robust and high-throughput Cre reporting and characterization system for the whole mouse brain. Nature Neuroscience. 2010;13:133–140. doi: 10.1038/nn.2467. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Maffie J, Rudy B. Weighing the evidence for a ternary protein complex mediating A-type K+ currents in neurons. The Journal of Physiology. 2008;586:5609–5623. doi: 10.1113/jphysiol.2008.161620. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Mayer C, Acosta-Martinez M, Dubois SL, Wolfe A, Radovick S, Boehm U, Levine JE. Timing and completion of puberty in female mice depend on estrogen receptor alpha-signaling in kisspeptin neurons. PNAS. 2010;107:22693–22698. doi: 10.1073/pnas.1012406108. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. McCarthy MM, Arnold AP. Reframing sexual differentiation of the brain. Nature Neuroscience. 2011;14:677–683. doi: 10.1038/nn.2834. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Meng X, Lu Q, Rinzel J. Control of firing patterns by two transient potassium currents: leading spike, latency, bistability. Journal of Computational Neuroscience. 2011;31:117–136. doi: 10.1007/s10827-010-0297-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Molineux ML, Fernandez FR, Mehaffey WH, Turner RW. A-type and T-type currents interact to produce a novel spike latency-voltage relationship in cerebellar stellate cells. Journal of Neuroscience. 2005;25:10863–10873. doi: 10.1523/JNEUROSCI.3436-05.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Nelson JF, Karelus K, Felicio LS, Johnson TE. Genetic influences on the timing of puberty in mice. Biology of Reproduction. 1990;42:649–655. doi: 10.1095/biolreprod42.4.649. [DOI] [PubMed] [Google Scholar]
  24. O'Leary T, Williams AH, Franci A, Marder E. Cell types, network homeostasis, and pathological compensation from a biologically plausible ion channel expression model. Neuron. 2014;82:809–821. doi: 10.1016/j.neuron.2014.04.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Shah MM, Hammond RS, Hoffman DA. Dendritic ion channel trafficking and plasticity. Trends in Neurosciences. 2010;33:307–316. doi: 10.1016/j.tins.2010.03.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Taniguchi H, He M, Wu P, Kim S, Paik R, Sugino K, Kvitsiani D, Kvitsani D, Fu Y, Lu J, Lin Y, Miyoshi G, Shima Y, Fishell G, Nelson SB, Huang ZJ. A resource of Cre driver lines for genetic targeting of GABAergic neurons in cerebral cortex. Neuron. 2011;71:995–1013. doi: 10.1016/j.neuron.2011.07.026. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Tasker JG, Dudek FE. Electrophysiological properties of neurones in the region of the paraventricular nucleus in slices of rat hypothalamus. The Journal of Physiology. 1991;434:271–293. doi: 10.1113/jphysiol.1991.sp018469. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Taylor SE, Klein LC, Lewis BP, Gruenewald TL, Gurung RA, Updegraff JA. Biobehavioral responses to stress in females: tend-and-befriend, not fight-or-flight. Psychological Review. 2000;107:411–429. doi: 10.1037/0033-295X.107.3.411. [DOI] [PubMed] [Google Scholar]
  29. Tost H, Champagne FA, Meyer-Lindenberg A. Environmental influence in the brain, human welfare and mental health. Nature Neuroscience. 2015;18:1421–1431. doi: 10.1038/nn.4108. [DOI] [PubMed] [Google Scholar]
  30. Wamsteeker Cusulin JI, Füzesi T, Inoue W, Bains JS. Glucocorticoid feedback uncovers retrograde opioid signaling at hypothalamic synapses. Nature Neuroscience. 2013a;16:596–604. doi: 10.1038/nn.3374. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Wamsteeker Cusulin JI, Füzesi T, Watts AG, Bains JS. Characterization of corticotropin-releasing hormone neurons in the paraventricular nucleus of the hypothalamus of Crh-IRES-Cre mutant mice. PLoS One. 2013b;8:e18726. doi: 10.1371/journal.pone.0064943. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Yellen G. The voltage-gated potassium channels and their relatives. Nature. 2002;419:35–42. doi: 10.1038/nature00978. [DOI] [PubMed] [Google Scholar]

In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses. A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.

[Editors’ note: a previous version of this study was rejected after peer review, but the authors submitted for reconsideration. The previous decision letter after peer review is shown below.]

Thank you for submitting your work entitled "Sexually dimorphic neuronal responses to social isolation stress" for consideration by eLife. Your article has been favorably evaluated by a Senior Editor and four reviewers, one of whom, Peggy Mason, is a member of our Board of Reviewing Editors, and another is Danny Winder. Our decision has been reached after consultation between the reviewers. Based on these discussions and the individual reviews below, we regret to inform you that your work will not be considered further for publication in eLife.

The reviewers agreed that this manuscript addresses an important and understudied question regarding sex differences in the response of rodents to social isolation stress. Unfortunately, there were fundamental issues that decreased the reviewers' enthusiasm for this study.

1) Non-isolated animals were not studied. So the effect of social isolation on membrane properties of PVN CRH neurons is in fact not discernible from the experiments reported. That isolated females are different from isolated males may be due to sex, isolation, or an interaction of the two.

2) The metyrapone and RU486 experiments are done in males with FS instead of with isolated females. Also no vehicle controls.

3) Studied neurons may include neurons with up-regulated CRF transcript in response to isolation. Thus the population of neurons studied may not be as specific as suggested.

4) The observations reported are largely independent of each other and no causal relationship has been demonstrated.

The first two concerns, which were most critical to the editorial decision, are addressable with new experiments. Should the authors perform those experiments and substantially revise the manuscript to address the remaining concerns, then a new submission would be of interest.

Reviewer #1:

The fundamental problem with this study is that non-isolated animals are not studied. So the opening sentence of the discussion (that social isolation alters membrane properties of PVN CRH neurons in female rats) is simply not tested by the experiments. What can be said is that isolated females are different from isolated males. But whether that difference is due to sex, isolation or an interaction is unknown. Similarly, it is not known (although it is stated) that male PVN CRH neurons are unchanged by isolation.

The metyrapone and RU486 experiments were done on males and only males and only on FS. Why? There is no guarantee that the mechanisms of stress responses – when engaged – are the same in males and females.

Reviewer #2:

This manuscript starts off with an interesting and potentially important observation – that social isolation increases CRH PVN neuron first spike latency (FSL) in females, but not males. In addition, the authors nicely show that this effect can be partially reversed by pair housing (Figure 1). However, it's unclear to me why the authors then followed up by exploring corticosterone modulation of swim stress FSL increases in males and not females (Figure 2). Why didn't they look at CORT modulation of their initial finding (i.e. FSL in socially isolated females)? It therefore becomes difficult to directly link the findings in Figure 1 and Figure 2. Moreover, the experiments in Figure 3 also seem to be only loosely connected to the primary story. Based on the title of the paper, it is unclear why most of the mechanistic work (all of the CORT-related experiments) is done in males and not females, and the paper as a whole feels very disjointed, even if the individual experiments are more or less experimentally sound.

Additionally, there are a number of instances in which the manuscript or figure caption text is inconsistent or incomplete with respect to the figures. For example, "FSL in female mice subjected to swim stress was not different from FSL in female mice subjected to social isolation (Figure 2A)," but Figure 2A compares female mice to male mice in the swim stress condition. Second, the manuscript text for Figure 3 refers to current characterization from voltage clamp recordings from male, female, and male + CORT groups, but traces are only shown from one male neuron, and the pink and blue traces are never identified (I'm assuming it's control and 4AP conditions, but this is not explicitly stated and the fact that the 4AP condition is represented in gray to the right makes things somewhat confusing). The leftward shift in the activation curve of socially isolated females is referenced in the manuscript as Figure 3A, but this should be Figure 3B.

Reviewer #3:

In this manuscript, Senst et al. explore sex-specific alterations in the excitability of CRF neurons in the hypothalamus. They present an interesting series of studies demonstrated a relationship between sex, social interaction and first spike latency (FSL) that is intriguing, and implicate SGK1 regulation of Kv4 potassium channels in the process. The work is timely, interesting, and well done, though the premise is slightly oversold. The second sentence of the manuscript states, "an unproven corollary is that the absence of a social network may itself be stressful". I would argue that this has already been widely studied and that "social isolation" is widely viewed as stressful. Less well characterized are the sex-specific differences in responses to this type of stress, which the authors nicely delve into here.

The presentation of comparisons is at first confusing, as the data seem to be presented in a counter-intuitive fashion. Rather than presenting data from group housed males and females, then comparing that with singly housed animals, the authors have presented the isolated data first. It would seem more logical and easier for the reader (at least this one) to follow to start with the "control" condition first in presentation, but this raises another issue, which is that it appears that group housed animals (beyond the paired) were not reported. This seems an important oversight to correct. That is, to demonstrate intrinsic plasticity has occurred by direct comparison of female group-housed to isolated, then to compare with males.

The description of the "paired-housed" experiments in the third paragraph of the Results is confusing, as it could be interpreted as implying the animals were isolated first, and then pair-housed. As such, does this represent a reversal of the plasticity? What if animals are only group housed? If such a "reversal" experiment was not performed, do the authors think subsequent pair-housing would reverse the isolation effects?

The authors show a correlation between cort levels and FSL in male mice. This initiates the mechanistic studies on GR and SGK1 relative involvement in FSL plasticity. However, little attention appears to be paid to the converse, that is, does the FSL plasticity observed lead to alterations in cort levels. This is a difficult "chicken-and-egg" issue, that nonetheless deserves more attention in the Discussion since the implicit idea is that this plasticity will lead to altered HPA output. The authors describe FSL plasticity as a means for negative regulation of HPA output, but since the latency is positively correlated with increased cort, this does not seem to be directly supported by the data.

Reviewer #3 (Additional data files and statistical comments):

Is the window current difference pointed to statistically significant? This needs to be explicitly described in the manuscript.

Reviewer #4:

This manuscript addresses an important and understudied question regarding sex differences in the response of rodents to social isolation stress. The authors show clear differences in the response of PVN CRF neurons to social isolation, with CRF neurons from females showing a marked decrease in excitability as measured by the latency to first spike onset (FSL) in response to a transient depolarizing current injection compared to male CRF neurons. The authors show that a similar response can be induced in males in response to an acute non-social stressor, suggesting that the response is sex- and stressor specific. In subsequent experiments the authors show that the increased FSL was mimicked by systemic cortisol administration, but was not blocked by a glucocorticoid receptor antagonist. In contrast, application of an inhibitor of serum glucocorticoid kinase-1 (SGK1) prevented the cortisol-induced increase in FSL. The authors then argue that mechanistically, the increased FSL, is due to a cort-SGK1-dependent modulation of an underlying Kv4 channel-mediated IA current.

The main concerns with the manuscript are that 1) the authors fail to address the potential confound of putative CRF neuronal heterogeneity in the PVN. Many cells in the brain transiently express mRNA for peptides but fail to express the mature protein, for example a subclass of oxytocin (OT) neurons in the PVN co-express CRF mRNA. Hence the transgenic mouse may show significant ectopic expression of the reporter protein used to "identify" CRF neurons. Significantly a subpopulation of OT neurons have properties identical to those of putative CRF neurons in the female PVN following isolation stress (Luther & Tasker, 2000), hence social isolation may upregulate CRF mRNA expression in these neurons and bias the sampling in subsequent recordings. 2) The authors fail to establish a cause-effect relationship for any of their independent observations. 3) In the absence of a dose-response relationship for 4-AP the assumption that the transient outward current is mediated by Kv4 channels is groundless. Many other transient outward K+ channels are blocked by lower concentrations of 4-AP. Also what evidence do the authors have that these cells even express Kv4 channels? 4) SGK1 activation modulates the activity of multiple ion channels that could affect the FSL including Kv1.1 – 1.5, M-channels, and ASIC1 channels, and the author provide no evidence for a direct/specific effect of SGK1 on Kv4 channel activity. Consequently, the manuscript reads like a collection of independent observations that have been rather tenuously linked together into a story that is mostly speculation.

[Editors’ note: what now follows is the decision letter after the authors submitted for further consideration.]

Thank you for submitting your article "Sexually dimorphic neuronal responses to social isolation" for consideration by eLife. Your article has been favorably evaluated by a Senior Editor and four reviewers, one of whom, Peggy Mason (Reviewer #1), is a member of our Board of Reviewing Editors, and another is Danny Winder (Reviewer #2 in the resubmission).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

Your finding that isolation causes stress in females but not males is a striking and provocative example of sexual dimorphism. Yet the follow up experiments do not convincingly investigate this sex difference since, for example, there are no non-isolated control groups. Also without female CORT data or swim stress in group-housed animals, the sexual dimorphism is restricted to whether isolation is sufficient or not.

On the other hand, the mechanistic insight into the stress response, afforded by the FSL result, is exciting. The authors are urged to change the emphasis (and title) accordingly.

Reviewer #1:

This rewrite is very well done. The experiments are well motivated and clear in the revision. I have only a few comments.

Two portions of the text are a bit challenging to follow. First the contribution of the rapidly inactivating K channels. This is a broad audience. Please remind the reader what the meaning of each finding is and how the results lead step by step to the conclusion.

Second, the negative feedback part is confusing to me. I was happily going along thinking that I was learning about the stress response to social isolation (F) or swim in cold water (M) and then in the Discussion I read that the whole paper is interpreted as the negative feedback on the stress response. Obviously, I, as is true of most readers, do not think about stress responses as much as the authors do. The authors are requested to take this into account and connect the dots. Is the stress response so fleeting that any attempt to study it is thwarted and automatically delegated to the longer-lasting negative feedback stage?

In the first experiment on group- vs. single-housed mice, what age are these animals? Are these adults? If not, when in the P22-35 window do these animals come from? Were there any differences across this broad range of isolation times?

In the Abstract, it should be stated that exposure of male mice to acute physical stress had the same effect.

The authors may consider rephrasing the conclusion that stress is perceived differently by males and females. This reviewer suggests: males and females react differently to stress.

Why was Metyrapone administered p.o. to females and by injection to males?

Interesting questions for the future include: Are females similarly sensitive to physical stressors as are males? Would grp housed females show shortened FSL after a swim? The authors could consider adding in such speculation as their Discussion is laudably direct and concise.

Reviewer #2:

This revised version is very significantly improved over the initial submission. I have no significant concerns to add.

Reviewer #3:

Overall I think this is an improved manuscript. Figure 1 and Figure 3 are more complete, and I think the significance is high. However, I still find concerning gaps in the connectivity of the individual experiments and the conclusions drawn.

I am still confused by the logic behind the 2nd set of experiments. If the hypothesis is that FSL changes in isolated females are due to stress, then how does stressing isolated males test this hypothesis? (Results, fourth paragraph). It seems to me that it could just be a matter of degree – i.e. isolation isn't sufficiently stressful to increase FSL in males, but isolation + swim stress is. From my reading, the authors did not test swim stress alone in either group. Conversely, the lack of FSL increase in response to swim stress in isolated females seems more like a ceiling effect than an occlusion. To be able to conclude that the isolation-induced FSL increase in females is due to stress, the authors should be able to show the same effect in group housed females exposed to swim stress. Additionally, Figure 2C shows the CORT/FSL correlation for males only, when the primary hypothesis seems to be for females. If this correlation wasn't significant in females, that should be stated. To me, these were the missing experiments and data sets in the first version of the paper and they appear to still be missing.

I can't find the data described at the end of the fourth paragraph of Results (CORT effects in vitro). First, why wouldn't the authors show these data, and second, why does it appear that this experiment was only done in isolated males? Response point 6 says that they didn't do it in females because FSL is already increased due to isolation, but they could have done it in non-isolated animals to show that CORT can have an effect on FSL in females. As with the first version of this manuscript, these kinds of decisions make it seem like the authors are throwing together a bunch of data without considering how each experiment contributes to the overarching conclusions. If the goal of this project is to define sexual dimorphism in CRH neuronal responses to stress and stress hormones, they need to do these experiments in both sexes and experimentally dissect out the roles of isolation vs. swim stress, since in many experiments these conditions are combined.

Reviewer #4:

The authors have gone a long way in addressing the concerns raised in the initial review of this revised manuscript. The addition of the non-stressed control groups, and the metyrapone experiments in females have strengthened the manuscript considerably. Similarly, the removal of the SGK1 data and the inclusion of the CORT experiments in females have bolstered the mechanistic aspect of the study. Consequently, I have no major concerns with the manuscript. The study is timely, and the identification of sex-specific responses of the PVN circuitry to stressors is of considerable interest.

[Editors’ note: the author responses to the first round of peer review follow.]

The reviewers agreed that this manuscript addresses an important and understudied question regarding sex differences in the response of rodents to social isolation stress. Unfortunately, there were fundamental issues that decreased the reviewers' enthusiasm for this study.

1) Non-isolated animals were not studied. So the effect of social isolation on membrane properties of PVN CRH neurons is in fact not discernible from the experiments reported. That isolated females are different from isolated males may be due to sex, isolation, or an interaction of the two.

We have now conducted additional experiments with male and female animals housed in same-sex groups (3-5 mice per group). There is no difference in first spike latency (FSL) if males are isolated or in a group. By contrast, females show an inverse correlation between group size and FSL. FSL increases as we decrease from group housing (3-5) to pair housing to single housing. These data are now part of a new Figure 1.

2) The metyrapone and RU486 experiments are done in males with FS instead of with isolated females. Also no vehicle controls.

The metyrapone experiments have now been done in females as follows: Females in a group (3) were given access to drinking water containing metyrapone for one day and then isolated for sixteen hours with continued access to metyrapone in the water. The FSL in CRH neurons recorded from these animals was not different from group-housed females or from males pre-treated with metyrapone and then subjected to swim stress. These data are now reported in Figure 3.

We have also conducted vehicle control experiments. Vehicle (PEG) administration had no effect on the stress-induced increase in FSL. For the female isolation experiments, metyrapone was dissolved in drinking water, so the control group is the isolated females.

3) Studied neurons may include neurons with up-regulated CRF transcript in response to isolation. Thus the population of neurons studied may not be as specific as suggested.

This is a point that can be raised with every study that uses a genetic approach to target a specific cell population. The alternatives to this approach are to use post-hoc immunohistochemistry to identify the cell. This is notoriously unreliable and in particular, immunohistochemistry for CRH rarely labels more than 75% of the cell population. One could use single-cell PCR approaches, but here the reviewer’s concerns would be the same. There is no perfect method, but we have carefully quantified potential overlap with other cell populations in PVN in a previous paper (Wamsteeker Cusulin et al., 2013). We see less than 5% overlap with oxytocin neurons and td-tomato in PVN. More than 95% of the CRH immunopositive neurons are td tomato positive. We do agree with the reviewer that some caution is warranted, so we have added a comment in the Discussion to reflect this point.

4) The observations reported are largely independent of each other and no causal relationship has been demonstrated.

Our additional experiments with group housed females and males as well as those showing the effects of isolation on FSL were blocked by inhibition of CORT synthesis strengthen the links between the findings. In addition, we have taken the reviewer comments about the flow and logic of the manuscript very seriously and modified the organization of the text and figures accordingly. In summary, we now show that:

1) Social isolation increases FSL in females, but not males.

2) The impact of social isolation was blocked by administering the CORT-synthesis inhibitor, metyrapone to females prior to and immediately after social isolation. This suggested that CORT was necessary for the increase in FSL.

3) We then asked whether acute stress, which increases CORT, would similarly increase FSL in males. It did.

4) If the increase in FSL in socially-isolated females is due to CORT, then effects of additional swim stress should be occluded. They were.

5) We show a positive correlation between CORT and FSL.

6) To determine whether CORT is sufficient, we incubated naïve males slices with CORT for 60 minutes and showed that this increased FSL. Female slices were not used because FSL is already increased in socially-isolated females.

7) Mechanistically, this increase in FSL relies on a Kv channel that is sensitive to millimolar concentrations of 4-aminopyridine. There was no change in current density after stress, but we do report significantly longer decay kinetics in female CRH neurons.

8) The functional consequences of this intrinsic plasticity is a decrease in excitability of CRH neurons from socially isolated females.

The first two concerns, which were most critical to the editorial decision, are addressable with new experiments. Should the authors perform those experiments and substantially revise the manuscript to address the remaining concerns, then a new submission would be of interest.

Reviewer #1:

The fundamental problem with this study is that non-isolated animals are not studied. So the opening sentence of the discussion (that social isolation alters membrane properties of PVN CRH neurons in female rats) is simply not tested by the experiments. What can be said is that isolated females are different from isolated males. But whether that difference is due to sex, isolation or an interaction is unknown. Similarly, it is not known (although it is stated) that male PVN CRH neurons are unchanged by isolation.

We have conducted additional experiments to address this question directly. Specifically, males and females were housed in same sex groups (3-5 animals per group). This manipulation had no effect on FSL in males (compared to single housed), but there was a significant decrease in FSL in females (compared to single-housed). The data from these experiments are now shown in a new Figure 1. Additional experiments in which females were pair-housed showed an intermediate FSL phenotype. Collectively, these observations indicate a graded effect of the social network in females that is absent in males. In order to make the study easier to understand, we have re-written the entire manuscript, re-organized the figures and conducted additional experiments to address reviewer concerns.

The metyrapone and RU486 experiments were done on males and only males and only on FS. Why? There is no guarantee that the mechanisms of stress responses – when engaged – are the same in males and females.

We conducted these experiments in males because single-housed females already showed an increase in FSL, and therefore additional effects of swim stress were occluded. We have now conducted additional metyrapone experiments in the females (Figure 3). Given the additional experiments that have significantly strengthened the link between social isolation and changes in FSL, we feel the RU486 experiments divert attention from the main hypothesis and consequently, we have removed them from the manuscript.

Reviewer #2:

This manuscript starts off with an interesting and potentially important observation – that social isolation increases CRH PVN neuron first spike latency (FSL) in females, but not males. In addition, the authors nicely show that this effect can be partially reversed by pair housing (Figure 1). However, it's unclear to me why the authors then followed up by exploring corticosterone modulation of swim stress FSL increases in males and not females (Figure 2). Why didn't they look at CORT modulation of their initial finding (i.e. FSL in socially isolated females)? It therefore becomes difficult to directly link the findings in Figure 1 and Figure 2. Moreover, the experiments in Figure 3 also seem to be only loosely connected to the primary story. Based on the title of the paper, it is unclear why most of the mechanistic work (all of the CORT-related experiments) is done in males and not females, and the paper as a whole feels very disjointed, even if the individual experiments are more or less experimentally sound.

As described in the response to the editor, we have now conducted metyrapone experiments in females prior to and immediately after social isolation. This eliminated the increase in FSL. These data are now included in Figure 3. The CORT modulation was done in males because an acute stress in males mimicked the effects of social isolation in females. This hinted at a common mechanism. Furthermore, the fact that acute stress had no effect on FSL in females suggests that isolation has occluded the effects of acute elevations of CORT.

We have thought carefully about the reviewer’s comment about the ‘disjointed’ aspects of the paper and the fact that experiments in Figure 3 appear only loosely connected. Accordingly, and because of new data, we have re-organized the paper and removed the data in Figure 3 because we agree that they were largely diversionary.

Additionally, there are a number of instances in which the manuscript or figure caption text is inconsistent or incomplete with respect to the figures. For example, "FSL in female mice subjected to swim stress was not different from FSL in female mice subjected to social isolation (Figure 2A)," but Figure 2A compares female mice to male mice in the swim stress condition. Second, the manuscript text for Figure 3 refers to current characterization from voltage clamp recordings from male, female, and male + CORT groups, but traces are only shown from one male neuron, and the pink and blue traces are never identified (I'm assuming it's control and 4AP conditions, but this is not explicitly stated and the fact that the 4AP condition is represented in gray to the right makes things somewhat confusing). The leftward shift in the activation curve of socially isolated females is referenced in the manuscript as Figure 3A, but this should be Figure 3B.

We apologize for these errors. We have corrected the manuscript accordingly.

Reviewer #3:

In this manuscript, Senst et al. explore sex-specific alterations in the excitability of CRF neurons in the hypothalamus. They present an interesting series of studies demonstrated a relationship between sex, social interaction and first spike latency (FSL) that is intriguing, and implicate SGK1 regulation of Kv4 potassium channels in the process. The work is timely, interesting, and well done, though the premise is slightly oversold. The second sentence of the manuscript states, "an unproven corollary is that the absence of a social network may itself be stressful". I would argue that this has already been widely studied and that "social isolation" is widely viewed as stressful. Less well characterized are the sex-specific differences in responses to this type of stress, which the authors nicely delve into here.

We have re-written parts of the Introduction to make this point clearer.

The presentation of comparisons is at first confusing, as the data seem to be presented in a counter-intuitive fashion. Rather than presenting data from group housed males and females, then comparing that with singly housed animals, the authors have presented the isolated data first. It would seem more logical and easier for the reader (at least this one) to follow to start with the "control" condition first in presentation, but this raises another issue, which is that it appears that group housed animals (beyond the paired) were not reported. This seems an important oversight to correct. That is, to demonstrate intrinsic plasticity has occurred by direct comparison of female group-housed to isolated, then to compare with males.

We have re-organized the Results to start with the ‘control’ condition (group-housed males or females) and now include additional experiments with metyrapone in females (see comments to editor and reviewer 1 above).

The description of the "paired-housed" experiments in the third paragraph of the Results is confusing, as it could be interpreted as implying the animals were isolated first, and then pair-housed. As such, does this represent a reversal of the plasticity? What if animals are only group housed? If such a "reversal" experiment was not performed, do the authors think subsequent pair-housing would reverse the isolation effects?

The animals were not isolated first. They were either in pairs or single. We did not attempt a reversal experiment. Given the addition of new data with males or females in groups, the paired female data, showing an effect that is intermediate, now serve to further support our overall argument.

The authors show a correlation between cort levels and FSL in male mice. This initiates the mechanistic studies on GR and SGK1 relative involvement in FSL plasticity. However, little attention appears to be paid to the converse, that is, does the FSL plasticity observed lead to alterations in cort levels. This is a difficult "chicken-and-egg" issue, that nonetheless deserves more attention in the Discussion since the implicit idea is that this plasticity will lead to altered HPA output. The authors describe FSL plasticity as a means for negative regulation of HPA output, but since the latency is positively correlated with increased cort, this does not seem to be directly supported by the data.

We have now conducted experiments in which we block CORT production in females prior to isolation. Following this manipulation, FSLs are the same as those observed in group housed females or males. We have removed the SGK data because we agree with a comment raised by reviewer 2 that these were diversionary. Instead, we have bolstered the link between isolation, stress, CORT and FSL.

Reviewer #3 (Additional data files and statistical comments):

Is the window current difference pointed to statistically significant? This needs to be explicitly described in the manuscript.

We have re-analyzed the window current data, and although the curve-fits appear to show differences, we are not convinced of its biological relevance given that the curves do not always fit the data points accurately. Consequently, we are no longer making any statements about window current differences. We have conducted additional analyses on the kinetics of the fast-inactivating current in males and females. We report that currents from isolated females have significantly longer time constants (Figure 4). Given the robust increase in FSL and the absence of any changes in current density and only modest changes in inactivation/activation thresholds, we propose that kinetic changes in this current are likely responsible for the longer FSL in isolated females.

Reviewer #4:

The main concerns with the manuscript are that 1) the authors fail to address the potential confound of putative CRF neuronal heterogeneity in the PVN. Many cells in the brain transiently express mRNA for peptides but fail to express the mature protein, for example a subclass of oxytocin (OT) neurons in the PVN co-express CRF mRNA. Hence the transgenic mouse may show significant ectopic expression of the reporter protein used to "identify" CRF neurons. Significantly a subpopulation of OT neurons have properties identical to those of putative CRF neurons in the female PVN following isolation stress (Luther & Tasker, 2000), hence social isolation may upregulate CRF mRNA expression in these neurons and bias the sampling in subsequent recordings.

We have previously shown that 96% of the CRH immunopositive neurons in PVN are also tdTomato positive (Wamsteeker Cusulin et al., 2013). These cells are neuroendocrine (terminals project outside the blood-brain barrier) as 86% of td-Tomato cells are labelled by intravenous (tail vein) injection of fluorogold. There is virtually no overlap (3%) between OT neurons and td-Tomato neurons; 5% overlap between VP and td-Tomato and <<1% overlap between either TRH (3 cells in 5 mice) or somatostatin (0 cells in 5 mice).

The reviewer’s astute comment, however, raises a larger issue that can, and perhaps should, be raised with every manuscript that uses a genetic approach to identify neurons. That concern is about how we actually describe or categorize neurons. If a neuron does not make a peptide, for example, CRH when sampled, does that mean it’s not a CRH neuron? We did not set out to resolve this question and we do not think anyone has resolved this in a satisfactory fashion. If the reviewer can offer a better alternative of sampling CRH neurons, we’re open to suggestions. Second, we are well aware of the Luther and Tasker paper (Luther & Tasker, 2000) and in fact it is that paper (and others) that have led the field to the conclusion that there are distinct fingerprints that can be used to distinguish OT/VP neurons from non-magnocellular neuroendocrine cells (CRH, TRH, somatostatin) and preautonomic neurons. We are arguing here that it is not so simple and that the plasticity of intrinsic properties is something that should be considered.

2) The authors fail to establish a cause-effect relationship for any of their independent observations. 3) In the absence of a dose-response relationship for 4-AP the assumption that the transient outward current is mediated by Kv4 channels is groundless. Many other transient outward K+ channels are blocked by lower concentrations of 4-AP. Also what evidence do the authors have that these cells even express Kv4 channels?

The reviewer makes a good point. Based on pharmacological overlap, it is difficult to discriminate between specific Kv channel subtypes. At this stage, based on our ability to block >30% of the current at 1-2 mM 4-AP, we can only conclude that these are either Kv1.4 or Kv4. We choose these as candidates based on work from Lee et al. indicating that Kv1.2, Kv1.3, Kv1.4, Kv4.1, Kv4.2 are expressed in neuroendocrine cells in PVN(Lee et al., 2012). We had shown this in the manuscript. Kv1.2 and 1.3 exhibit characteristics of classic Kv channels with delayed activation and no inactivation and are not sensitive to 4-AP, so we can rule these out as candidates. Kv1.4 and Kv4 are sensitive to 4-AP (fampridine), therefore we cannot discriminate between these options. Consequently, we have not suggested a specific K channel subtype and instead, refer to them transient outward K+ channels.

4) SGK1 activation modulates the activity of multiple ion channels that could affect the FSL including Kv1.1 – 1.5, M-channels, and ASIC1 channels, and the author provide no evidence for a direct/specific effect of SGK1 on Kv4 channel activity.

Based on comments from reviewer 2, we have removed the SGK1 data from the manuscript.

[Editors’ note: the author responses to the re-review follow.]

Your finding that isolation causes stress in females but not males is a striking and provocative example of sexual dimorphism. Yet the follow up experiments do not convincingly investigate this sex difference since, for example, there are no non-isolated control groups.

Non-isolated control groups are shown in Figure 1, Figure 2 and Figure 3.

Also without female CORT data or swim stress in group-housed animals, the sexual dimorphism is restricted to whether isolation is sufficient or not.

We have also conducted additional experiments in which group housed female mice were subjected to swim stress. As expected, this increased FSL. We also conducted experiments in which slices from group housed female mice were incubated with CORT. This increased FSL. The results are consistent with our model and more detailed information is available in specific responses to reviewer 2.

Reviewer #1:

This rewrite is very well done. The experiments are well motivated and clear in the revision. I have only a few comments.

Two portions of the text are a bit challenging to follow. First the contribution of the rapidly inactivating K channels. This is a broad audience. Please remind the reader what the meaning of each finding is and how the results lead step by step to the conclusion.

This is a very good point and we have made changes throughout the manuscript to provide a succinct conclusion at the end of each experiment, followed by a sentence that provides the rationale for the next series of experiments.

Second, the negative feedback part is confusing to me. I was happily going along thinking that I was learning about the stress response to social isolation (F) or swim in cold water (M) and then in the Discussion I read that the whole paper is interpreted as the negative feedback on the stress response. Obviously, I, as is true of most readers, do not think about stress responses as much as the authors do. The authors are requested to take this into account and connect the dots. Is the stress response so fleeting that any attempt to study it is thwarted and automatically delegated to the longer-lasting negative feedback stage?

We apologize for this lack of clarity. We have made a number of changes to the discussion which are in blue. We are in agreement with the reviewer that the negative feedback discussion is confusing (specifically for a broad audience) and have removed this from the manuscript.

In the first experiment on group- vs. single-housed mice, what age are these animals? Are these adults? If not, when in the P22-35 window do these animals come from? Were there any differences across this broad range of isolation times?

We did examine the effects of age on FSL in socially isolated females. There were no significant differences across this range (p=0.23) (see Author response image 1).

Author response image 1.

Author response image 1.

DOI: http://dx.doi.org/10.7554/eLife.18726.012

In the Abstract, it should be stated that exposure of male mice to acute physical stress had the same effect.

Done.

The authors may consider rephrasing the conclusion that stress is perceived differently by males and females. This reviewer suggests: males and females react differently to stress.

Done.

Why was Metyrapone administered p.o. to females and by injection to males?

Metyrapone has relatively short half-life and the effects of the injection are rapid, but transient. So, injection was appropriate prior to swim stress in the males, but because we did not know when, during isolation, CORT peaked, we decided to administer metyrapone through the drinking water in the females to maintain a constant dose throughout the experiment.

Interesting questions for the future include: Are females similarly sensitive to physical stressors as are males? Would grp housed females show shortened FSL after a swim? The authors could consider adding in such speculation as their Discussion is laudably direct and concise.

Since reviewer 3 also suggested a similar experiment, we have now done this and include the data in the manuscript. Group-housed females were subjected to swim stress. This increased FSL (see response to reviewer 2 for additional details). FSL in males subjected to swim stress was not different from FSL in females subjected to swim or social isolation. So, one additional conclusion we can draw is that physical stressors have the same effect on males and females, but the response to social isolation is sexually dimorphic. We have now expanded the Discussion according to the suggestions provided here.

Reviewer #3:

Overall I think this is an improved manuscript. Figure 1 and Figure 3 are more complete, and I think the significance is high. However, I still find concerning gaps in the connectivity of the individual experiments and the conclusions drawn.

I am still confused by the logic behind the 2nd set of experiments. If the hypothesis is that FSL changes in isolated females are due to stress, then how does stressing isolated males test this hypothesis? (Results, fourth paragraph). It seems to me that it could just be a matter of degree – i.e. isolation isn't sufficiently stressful to increase FSL in males, but isolation + swim stress is. From my reading, the authors did not test swim stress alone in either group. Conversely, the lack of FSL increase in response to swim stress in isolated females seems more like a ceiling effect than an occlusion. To be able to conclude that the isolation-induced FSL increase in females is due to stress, the authors should be able to show the same effect in group housed females exposed to swim stress.

We apologize for the confusion. In order to further clarify things, we have now conducted the experiment suggested by the reviewer. Group-housed females were subjected to swim stress. FSL in CRH neurons was significantly longer (58.7 ± 1.9 ms, n=27) than FSL from unstressed, group-housed females (45.1 ± 1.9 ms, n=42, p=0.0001). In addition, we have subjected group housed males to swim. FSL was significantly longer (58.8 ± 2.8 ms, n=23) than FSL from unstressed, group-housed males (48.7 ± 2.1 ms, n=53, p<0.0001). These data are now included in a revised Figure 2.

Additionally, Figure 2C shows the CORT/FSL correlation for males only, when the primary hypothesis seems to be for females. If this correlation wasn't significant in females, that should be stated. To me, these were the missing experiments and data sets in the first version of the paper and they appear to still be missing.

The reviewer makes a good point. We focused on CORT immediately after swim in males because this give us a relatively confined temporal window in which to make the measurements. In the isolated females, although we can say that CORT is necessary (metyrapone experiment), we do not have the ability to monitor CORT levels in vivo in real time. In the absence of this information, it would be difficult to draw any conclusions from either a correlation or lack of correlation in this group.

I can't find the data described at the end of the fourth paragraph of Results (CORT effects in vitro). First, why wouldn't the authors show these data, and second, why does it appear that this experiment was only done in isolated males? Response point 6 says that they didn't do it in females because FSL is already increased due to isolation, but they could have done it in non-isolated animals to show that CORT can have an effect on FSL in females. As with the first version of this manuscript, these kinds of decisions make it seem like the authors are throwing together a bunch of data without considering how each experiment contributes to the overarching conclusions.

We have now conducted additional experiments in which slices from group-housed females were incubated with CORT. FSL in CRH neurons in this group was 64.1 ± 4.8 ms, n=32. This is significantly longer than FSL from non-CORT treated CRH neurons of group housed female mice (45.1 ± 1.9 ms, n=42, p<0.0001). These data are now included in a revised Figure 3.

If the goal of this project is to define sexual dimorphism in CRH neuronal responses to stress and stress hormones, they need to do these experiments in both sexes and experimentally dissect out the roles of isolation vs. swim stress, since in many experiments these conditions are combined.

We apologize for our lack of clarity. We have now conducted additional experiments in which group-housed males were also subjected to swim stress. This caused an increase in FSL. Importantly, the FSL in single males subjected to swim is not different than grouped males subjected to swim. These data are now included in a revised Figure 2.