Comparison of inflammatory and behavioral responses to chronic stress in female and male mice

Abstract

Major depressive disorder (MDD) is a debilitating disease with a high worldwide prevalence. Despite its greater prevalence in women, male animals are used in most preclinical studies of depression even though there are many sex differences in key components of depression, such as stress responses and immune system functions. In the present study, we found that chronic restraint stress-induced depressive-like behaviors are quite similar in male and female mice, with both sexes displaying increased immobility time in the tail suspension test and reduced social interactions, and both sexes exhibited deficits in working and spatial memories. However, in contrast to the similar depressive-like behaviors developed by male and female mice in response to stress, they displayed different patterns of pro-inflammatory cytokine increases in the periphery and the brain, different changes in microglia, and different changes in the expression of Toll-like receptor 4 in response to stress. Treatment with (+)-naloxone, a Toll-like receptor 4 antagonist that previously demonstrated anti-depressant-like effects in male mice, was more efficacious in male than female mice in reducing the deleterious effects of stress, and its effects were not microbiome-mediated. Altogether, these results suggest differential mechanisms to consider in potential sex-specific treatments of depression.

1. INTRODUCTION

Major depressive disorder (MDD) is a common and debilitating disease that afflicts approximately 21% of women and 11% of men in the United States (Belmaker, 2008; Kessler et al., 2005). Despite the higher prevalence of MDD in women, most preclinical studies have used only male rodents as research tools (Beery and Zucker, 2011). This is partly because it is commonly thought that behavioral measurements in females can be variable because of the influence of different phases of the estrous cycle. Another problem has been that several commonly used animal models measuring depression-like behaviors were developed for male rodents. For example, the widely used social defeat model used to induce stress and depression-like behaviors has traditionally been limited to male rodents because female rodents do not attack intruders, although specific strains or modifications of the traditional protocols have been used with female rodents (Hammels et al., 2015; Harris et al., 2018; Hollis and Kabbaj, 2014; Steinman and Trainor, 2017; Takahashi et al., 2017; Trainor et al., 2011). Another widely used model, learned helplessness, has been reported to work well in male but not female rats (Dalla et al., 2008), and chronic unpredictable intermittent restraint stress disrupted spatial memory in male but not female rats (Peay et al., 2020). However, some methods widely used to induce behavioral impairments associated with depression have been shown to be reliable for studying female rodent responses to stress, such as chronic restraint stress, subchronic variable stress, and chronic unpredictable stress (Autry et al., 2009; Belovicova et al., 2017; Bowman et al., 2001; Hibicke et al., 2017; Hodes et al., 2015b).

Inflammation contributes to stress-induced depression-related behaviors (Hodes et al., 2015a; Medina-Rodriguez et al., 2018), but these findings have largely used only male rodents (Bekhbat and Neigh, 2018). In male mice, Toll-like receptor 4 (TLR4) is a key mediator of stress-induced inflammation and depression-like behaviors in male mice (Cheng et al., 2016; Dantzer et al., 2008; Medina-Rodriguez et al., 2020c). Mice deficient in TLR4 or pharmacological blockade of TLR4 with the TLR4 antagonist (+)-naloxone confer resilience to stress in male mice (Cheng et al., 2016; Medina-Rodriguez et al., 2020c). However, differences in inflammation between female and male rodents have been reported. For example, the injection of the TLR4 ligand lipopolysaccharide in adult mice induces greater increases of the cytokines IL6, TNFα and IL10 in male than in female mice (Kuo, 2016). In addition, the response to anti-inflammatory and antidepressant treatments is different between male and female rodents. For example, glucocorticoids have more potent effects as anti-inflammatory molecules in male than female rats (Duma et al., 2010) and ketamine shows long-lasting antidepressant-like effects in males but not in females subjected to chronic mild stress for 5 weeks while females responded better to acute treatment with ketamine (Franceschelli et al., 2015) and females also show an increased sensitivity to ketamine since they respond to doses that do not exert any effects in males (Carrier and Kabbaj, 2013; Dossat et al., 2018; Sarkar and Kabbaj, 2016).

In the last few years, evidence has been reported that the gut microbiome participates in the communication between the immune system and the brain to affect mood, in a bidirectional circuit known as the gut-brain axis (Foster et al., 2017; Ma et al., 2019). The absence of gut microbes, as in germ-free mice or mice treated with antibiotics to deplete bacteria, induces altered responses to stress (De Palma et al., 2015; Sudo et al., 2004) and colonization of germ-free mice with bacteria from stressed mice promotes depressive-like behaviors (Chevalier et al., 2020). In addition, the lack of the gut microbiome in germ-free mice is associated with defects in the immune system (Bauer et al., 1963). These findings point out an important role of the gut microbiome in the control of the immune system and the consequential outcome on stress-induced depression-like behaviors. However, although it is know that males and females show differences in their gut microbiota composition, which induces differences in the immune system (Fransen et al., 2017), very little is known about the importance of these sex-related differences of microbes and immunity in the development of depression-like behaviors.

In the present study we used chronic restraint stress to induce depression-like behaviors in male and female mice to test two hypotheses, the hypothesis that female mice differ from male mice in detrimental effects of stress, and the hypothesis that the absence of microbes affects behaviors in male and female mice. In addition, we used the TLR4 antagonist (+)-naloxone and TLR4 knockout mice to demonstrate that TLR4 mediates detrimental effects of stress in both female and male mice and we discovered that the beneficial effects of (+)-naloxone in dealing with stress are independent of the microbiome.

2. MATERIALS AND METHODS

2.1. Mice and drug administration

Male and female C57BL/6 wild-type mice were obtained from Taconic. C57BL/6 TLR4 global knockout (TLR4^−/−) mice were initially provided by Dr. Suzanne M. Michalek (University of Alabama at Birmingham). TLR4^−/− mice were bred using the following paradigm: TLR4^+/− × TLR4^+/− to obtain 25% TLR4^−/−, 25% wild-type, and 50% TLR4^+/− mice. Littermates were used at 8–12 weeks of age for experiments. A total of 242 mice were used, and mice were housed in groups of 3–5 in the University of Miami vivarium in light and temperature-controlled rooms and were treated in accordance with NIH and the University of Miami Institutional Animal Care and Use Committee regulations (protocols 17–172 and 20–143). The experiments were carried out with littermates that were randomized to their treatment groups. Where indicated, mice were treated intravenously (i.v.) via a tail vein with the TLR4 antagonist (+)-naloxone (5 or 10 mg/kg; provided by Dr. Kenner Rice) or saline.

For antibiotic treatments, specific-pathogen-free (SPF) mice were gavaged with a solution of neomycin (100 mg/kg), metronidazole (100 mg/kg), and vancomycin (50 mg/kg) twice daily for 7 days prior to chronic restraint stress. Ampicillin (1 mg/mL) was also provided ad libitum in the drinking water for 7 days before chronic restraint stress and was maintained until the behavioral assessments were finished to avoid bacterial recolonization. These conditions produced a germ free-like phenotype (Medina-Rodriguez et al., 2020b; Reikvam et al., 2011).

Experiments were carried out between 0600 and 1200 in a normally lighted laboratory, and behavioral apparatus was always deeply cleaned with 70% ethanol between tests.

2.2. Bacterial evaluation

To evaluate bacterial depletion genomic DNA was purified from stools before and after antibiotic treatment using the Quick-DNA Fecal/Soil Microbe Miniprep (Zymo Research, Irvine, Calif.) according to the manufacturer’s instructions and Eubacteria 16S was measured by SYBR green RT- qPCR in a Jena Analytika instrument and the results were quantified by the 2-ΔΔCt method respect to a negative control. Primers: 5’-ACTCCTACGGGAGGCAGCAGT-3’; and 5’- ATTACCGCGGCTGCTGGC -3’.

2.3. Chronic restraint stress

For chronic restraint stress, mice were restrained individually in a 50 mL ventilated conical tube for 2 h per day (~0900–1100) for 2 weeks, as we previously described (Beurel et al., 2013), and behavioral tests were carried out beginning one day after the last day of chronic restraint stress, with the timing shown in Suppl Figure 1. Mice were treated with (+)-naloxone (5 or 10 mg/kg; i.v.) or saline as vehicle 1 hr before testing for activity in a novel open field, and later habituated to the social interaction apparatus. Mice were treated each day with (+)-naloxone (5 or 10 mg/kg; i.v.) or saline 1 hr prior to performing behavioral tests that were carried out in the order of sociability and social novelty, coordinate spatial memory, novel object recognition, categorical spatial memory, temporal order recognition, and the tail suspension test, as shown in Suppl. Figure 1.

2.4. Tail suspension test

Mice were suspended by the tail on an automated tail suspension test cubicle (33 × 31.75 × 33 cm; Med Associates, St Albans, VT, USA) for a period of 6 min and the time they were moving was measured automatically by the apparatus (Liu et al., 2003). The immobile time during the last 4 min was used to determine the immobility time for each mouse. Mice were treated with the TLR4 antagonist (+)-naloxone (5 or 10 mg/kg; i.v.) or vehicle 1 hr prior to the tail suspension test.

2.5. Open field activity

For open field activity measurements (Beurel et al., 2013), mice were placed in a Plexiglas open field instrument (San Diego Instruments) outfitted with photo beam detectors under soft overhead lighting, and activity was monitored during 30 min using activity monitoring software (San Diego Instruments). Mice were exposed, or not, to chronic restraint stress and were treated with (+)-naloxone (5 or 10 mg/kg; i.v.) or saline 1 hr before testing for activity in a novel open field. The number of beam breaks was measured automatically by the apparatus and were calculated for each 5 min period. The total number of rearings was also calculated by the apparatus. The time in seconds spent in the periphery and the center of the open field was calculated with the PAS Data reporter (San Diego Instruments) considering the center to be 8×8 squares out of a total of 16×16 squares.

2.6. Social interaction

Diminished social interaction is associated with depression and was measured as we previously described (Medina-Rodriguez et al., 2020a). Mice were treated with the TLR4 antagonist (+)-naloxone (5 or 10 mg/kg; i.v.) or vehicle 1 hr prior to the test. Social interactions were measured using a sociability apparatus, which is a rectangular, transparent, Plexiglas box (24 cm × 19 cm, 19 cm high) divided into three equal sized chambers with doors. Chambers 1 and 3 had a wire cage; Chamber 2 in the middle was empty. Mice were habituated individually by being placed in Chamber 2 and were allowed to freely explore the entire apparatus for 25 min the day prior to testing, and, separately, stimulus mice were habituated for 20 min to the wire cage in Chamber 1. A new stimulus mouse was used with each new set of mice being tested for sociability. Testing consisted of 5 min rehabituation followed by 10 min access to all chambers with an unfamiliar stimulus mouse (age- and sex-matched) in the wire enclosure in Chamber 1. Each session was videotaped and videos were analyzed for the number of nose contacts with the stimulus mouse.

2.7. Social novelty

After the social interaction test, a new stimulus mouse (age- and sex-matched) was placed in the wire enclosure in Chamber 3 while the previously presented mouse remained in the wire enclosure in Chamber 1. Mice were allowed to explore for 10 min, which were videotaped and videos were analyzed for the number of nose contacts with the new stimulus mouse and the previously presented mouse.

2.8. Novel Object Recognition

Novel object recognition was measured as described previously (Pardo et al., 2016) by allowing each mouse individually to explore two identical objects for 5 min, and after a 5 min period in an opaque chamber, mice were allowed to explore an unused familiar object and a novel object for 5 min. Time spent exploring each object (sniffing or touching the object with its nose, vibrissa, mouth or forepaws) was quantified from video recordings. Mice were treated with the TLR4 antagonist (+)-naloxone (5 or 10 mg/kg; i.v.) or vehicle 1 hr prior to the test.

2.9. Temporal Order Recognition

Temporal order recognition was carried out as previously described (Pardo et al., 2016). Each mouse underwent three sessions exploring three sets of objects (Object sets 1, 2, 3) for 5 min each, with a resting interval of 5 min between each set of objects. During the test session, the mouse was allowed to explore a copy of Object 1 and a copy of Object 3 for 5 min and videotaped. Time spent exploring each object (sniffing or touching the object with its nose, vibrissa, mouth, or forepaws) was quantified from video recordings. Mice were treated with the TLR4 antagonist (+)-naloxone (5 or 10 mg/kg; i.v.) or vehicle 1 hr prior to the test.

2.10. Coordinate spatial processing

For the coordinate spatial processing task (Pardo et al., 2016), each mouse was allowed to explore two novel objects that were 45 cm apart for 15 min. After 5 min in an opaque chamber, each mouse was allowed to explore the same two objects that had been moved closer together (30 cm) for 5 min. The exploration ratio was calculated as time (exploring during the 5 min test session)/(exploring during the 5 min test session plus the last 5 min of the habituation session). Mice were treated with the TLR4 antagonist (+)-naloxone (5 or 10 mg/kg; i.v.) or vehicle 1 hr prior to the test.

2.11. Categorical spatial processing

For the categorical spatial processing task, each mouse was allowed to explore two novel objects that were 45 cm apart for 15 min. After 5 min in an opaque chamber, each mouse was allowed to explore the same two objects whose position had been switched. The exploration ratio was calculated as time (exploring during the 5 min test session)/(exploring during the 5 min test session plus the last 5 min of the habituation session). Mice were treated with the TLR4 antagonist (+)-naloxone (5 or 10 mg/kg; i.v.) or vehicle 1 hr prior to the test.

2.12. Cytokine measurements

The hippocampus and prefrontal cortex were dissected from mice subjected to chronic restraint stress or non-stressed mice (control) 1, 4 and 24 hr after stress or three hr after the last behavioral test, and homogenized in lysis buffer containing 20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton-X-100, 10 μg/ml leupeptin, 10 μg/ml aprotinin, 5 μg/ml pepstatin, 1 mM phenylmethanesulfonyl fluoride, 1 mM sodium vanadate, 50 mM sodium fluoride, and 100 nM okadaic acid as previously described (Cheng et al., 2016). Blood samples were quickly collected at the time of sacrifice, centrifuged to obtain the plasma, and stored at −80°C until use. The cytokines TNFα, IL-6, IL-17 and IFN-β were measured by enzyme-linked immunosorbent assay (ELISA; eBioscience or Biolegend) using 150 μg protein according to the manufacturer’s instructions. Briefly, the cytokine-specific capture antibody was immobilized onto an ELISA plate to capture the targeted cytokine, which is then detected by a cytokine-specific biotinylated antibody. The cytokine is quantified using a colorimetric reaction based on activity of avidin-horseradish peroxidase bound to the biotinylated detection antibody. Cytokine concentrations were determined with a microplate reader (SpectraMax M3).

2.13. Quantitative real-time polymerase chain reaction (qRT-PCR)

The hippocampus was dissected from mice subjected to chronic restraint stress or non-stressed mice (control) 1, 4 and 24 hr after stress or, in a different cohort of mice, 12 days after chronic stress once all the behavior tests were finished (2 hr after tail suspension test). RNA from the hippocampus was extracted with TRIzol Reagent (Life Technologies) and cDNA was synthesized with ImProm-II^™ Reverse Transcriptase and random primers (Promega). TLR4, Iba-1, CD11b and CD68 expression were measured by SYBR green RT- qPCR or Taqman gene expression assay in a Jena Analytika instrument and the results were quantified by the 2-ΔΔCt method. Primers used: TLR4 (Mm00445273_m1; Thermo Fisher Scientific); Iba-1 5’- GTCCTTGAAGCGAATGCTGG -3’ and 5’- CATTCTCAAGATGGCAGATC -3’; CD11b 5’- CCTTGTTCTCTTTGATGCAG -3’ and 5’- GTGATGACAACTAGGATCTT -3’; CD68 5’- CTCATCATTGGCCTGGTCCT -3’ and 5’- GTTGATTGTCGTCTGCGGG -3’. Values were normalized to GAPDH 5’- AGGTCGGTGTGAACGGATTTG -3’ and 5’- TGTAGACCATGTAGTTGAGGTCA -3’

2.14. Data and statistical analyses

Behavioral quantifications were blinded. Statistical significance was analyzed with a one-way, two-way or three-way analysis of variance (ANOVA) with Sidak’s, Dunnett or Tukey’s post-hoc test for multiple comparisons or with Student’s t-test using GraphPad Prism software. p<0.05 was considered significant. Histograms represent mean ± SD.

3. RESULTS

3.1. Effects of stress and (+)-naloxone treatment on the tail suspension test

The tail suspension test has been widely used to screen for potential antidepressant drugs (reviewed in (Cryan et al., 2005)). Corroborating our previous results in male mice (Medina-Rodriguez et al., 2020c), administration of 5 mg/kg (+)-naloxone exerted an antidepressant-like effect resulting in decreased immobility time in the tail suspension test (Fig 1A). Chronic restraint stress exposure did not induce more immobility time in the tail suspension test and the antidepressant-like effects of 5 mg/kg (+)-naloxone were the same regardless of exposure to chronic restraint stress. In female mice, administration of 5 mg/kg (+)-naloxone had no significant effect on immobility time, whereas treatment with 10 mg/kg (+)-naloxone had a significant effect on reducing immobility time (Fig 1B), indicating that female mice require a higher dose for (+)-naloxone to be effective. Comparison between male and female mice behavior before treatment showed no significant differences.

Figure 1. Female mice require a higher dose than male mice of the TLR4 antagonist (+)-naloxone to obtain the same antidepressant-like effect in the tail suspension test (TST).

Wild-type male (A) and female (B) mice subjected to chronic restraint stress (CRS) for two weeks or control (CT) mice were tested in the TST 1 hr after administration of (+)-naloxone (5 mg/kg for males and 5 mg/kg or 10 mg/kg for females; i.v.) or vehicle and the immobility time was measured. Each point represents an individual mouse. Bars represent means ± SD. Detailed statistics are in suppl. table 1. *p<0.05, **p<0.01. n=7–10 mice/group.

3.2. Effects of chronic restraint stress on cytokines

Basal levels of four cytokines, (TNFα, IL-17A, IL-6, and IFNβ), were compared in male and female mouse plasma, prefrontal cortex (PFC), and hippocampus, two brain regions involved in depression-like behaviors (Hastings et al., 2004). As summarized in Table 1, measurements of these cytokines in unstressed mice revealed that the basal levels of TNFα were lower in female mice than in male mice in the plasma and PFC (Fig 2A, B), but the basal levels of TNFα in the hippocampus were equivalent in male and female mice (Fig 2C). Basal levels of IL-17A and IL-6 were the same in male and female mice in plasma, PFC and hippocampus (Fig 2D–I). Basal levels of IFNβ were higher in the PFC of female mice than male mice, but were equivalent in the plasma and hippocampus (Fig 2J–L).

Table 1.

Comparison of basal levels of TNFα, IL-17A, IL-6, and IFNß in unstressed, control female mice and male mice in plasma, prefrontal cortex (PFC) and hippocampus.

	SAMPLE	TNFα	IL-17A	IL-6	IFNβ
Levels in female mice compared to male mice	PLASMA	↓
	PFC	↓			↑
	HIPPOCAMPUS

Figure 2. Comparisons of basal levels of inflammatory cytokines in male and female mice in plasma, prefrontal cortex (PFC) and hippocampus.

Basal levels of inflammatory cytokines (TNFα (A-C), IL-17A (D-F), IL-6 (G-I) and IFNβ (J-L)) were measured by ELISA in plasma (A, D, G, J), PFC (B, E, H, K) and hippocampus (C, F, I, L) of wild-type male and female mice. Each point represents an individual mouse. Bars represent means ± SD. Detailed statistics are in suppl. table 1. *p<0.05, **p<0.01, n=3–5 mice/group.

In order to study whether the inflammatory response to chronic stress differs by sex, male and female mice were subjected to 2 weeks of daily restraint stress and TNFα, IL-17A, IL-6, and IFNβ were quantified 1, 4 and 24 hr after the last restraint session in plasma, PFC, and hippocampus. These different time intervals were used to identify changes that occurred in response to the acute restraint period, which would likely be short-lived and transient, and changes resulting from the chronic stress that were long-lasting that would be evident throughout the 24 hr post-treatment interval. As summarized in Table 2, male mice but not female mice exhibited significant increases in TNFα in the plasma. TNFα in the PFC increased in male mice at 1 hr but at 24 hr in female mice (Fig 3B). Hippocampal TNFα increased at 4 hr in female but not in male mice. IL-17A increased in male mice 1 hr (plasma, PFC) or 4 hr (hippocampus) after chronic restraint stress (Fig 3D–F), but IL-17A levels did not change in female mice at any of the times examined in these three tissues (Fig 3D–F). After chronic restraint stress, IL-6 levels increased in the plasma at 1 hr in both male and female mice (Fig 3G), did not change at any time in the PFC in male or female mice (Fig 3H), and increased in the hippocampus at 1 hr in female mice but not in male mice (Fig 3I). IFNβ increased in plasma of male mice 4 hr after chronic restraint stress (Fig 3J) and increased in the PFC 1 hr after chronic restraint stress in male mice (Fig 3K), while there were no changes in the hippocampus (Fig 3L). Male mice vs. female mice comparisons revealed significant differences in TNFα at 1 and 24 hr in PFC, in IL-17A at 1hr in plasma and PFC, and at 4 hr in the hippocampus, in IL-6 at 4 hr in plasma and at 1 hr in the hippocampus, and IFNβ at 1 hr in PFC and 24 hr in the hippocampus (Figure 3). These results demonstrated that the inflammatory response to chronic restraint stress differs between male and female mice, showing different time-dependent responses in the circulating and brain levels of inflammatory cytokines.

Table 2.

Time-dependent changes in the levels of TNFα, IL-17A, IL-6, and IFNß in male and female mice in plasma, prefrontal cortex (PFC) and hippocampus after chronic restraint stress.

SAMPLE		TIME (hr)	TNFα	IL-17A	IL-6	IFNβ
PLASMA	MALES	1		↑	↑
		4	↑			↑
		24
	FEMALES	1			↑
		4
		24
PFC	MALES	1	↑	↑		↑
		4
		24
	FEMALES	1
		4
		24	↑
HIPPOCAMPUS	MALES	1
		4		↑
		24
	FEMALES	1			↑
		4	↑
		24

Figure 3. Chronic restraint stress differentially induces orchestrated time and region-dependent changes in male and female mice.

Wild-type control (CT) male and female mice were subjected to chronic restraint stress for two weeks and plasma, prefrontal cortex (PFC) and hippocampal samples were taken 1h, 4h and 24h after chronic restraint stress. Inflammatory cytokines (TNFα (A-C), IL-17A (D-F), IL-6 (G-I) and IFNβ (J-L)) were measured by ELISA. Each point represents an individual mouse. Bars represent means ± SD. Detailed statistics are in suppl. table 1. *p<0.05, **p<0.01. n=3–5 mice/group.

3.3. Effects of chronic restraint stress on microglia

Microglia are a major source of inflammatory cytokines in the CNS and microglial responses have been linked to both the stress response and to depression and to animal models of depression (Bottcher et al., 2020; Yirmiya et al., 2015; Zhang et al., 2018). To study whether microglial changes in the brain differ between male and female mice in response to chronic restraint stress, we quantified the expression of two microglial markers, Iba-1 and CD68 in the hippocampus at 1, 4 and 24 hr after chronic restraint stress (Fig 4). The expressions of Iba-1 and CD68 were the same in untreated male and female mice (Fig 4A,B). After chronic restraint stress, in male mice hippocampal Iba-1 expression decreased at 4 hr and was subsequently restored at 24 hr (Fig 4C). However, in female mice, Iba-1 expression was decreased earlier, only 1 hr after chronic restraint stress, and this decrease was maintained for 24 hr after stress (Fig 4C). These findings corroborate previous studies showing that Iba-1 expression can decrease after chronic stress (Kreisel et al., 2014; Tong et al., 2017) and show that microglial responses to stress are different between male and female mice (significant differences at 24h). Oppositely, the marker of activated microglia CD68 was increased in male and female mice 24 hr after stress (Fig 4D).

Figure 4. Chronic restraint stress differentially affects microglial markers expression in male and female mice.

(A-B) Basal expression of microglial markers Iba-1 (A) and CD68 (B) were measured by PCR in the hippocampus of wild-type male and female mice. (C-D) Wild-type control (CT) male and female mice were subjected to chronic restraint stress for two weeks and hippocampal expression of microglial markers Iba-1 (C) and CD68 (D) were measured 1h, 4h and 24h after chronic restraint stress. Each point represents an individual mouse. Bars represent means ± SD. Detailed statistics are in suppl. table 1. *p<0.05, **p<0.01. n=3–7 mice/group.

3.4. Influence of microbiome on responses to chronic restraint stress

To investigate whether the effects of chronic restraint stress on microglia are influenced by the composition of the microbiome, male and female mice were treated with antibiotics to induce a germ free-like phenotype(Medina-Rodriguez et al., 2020b; Reikvam et al., 2011) and then subjected to chronic restraint stress for two weeks. Microbiota depletion in male and female mice was confirmed in stools collected before (CT) or after antibiotic treatment by Eubacteria 16S qRT-PCR (Fig 5A). A depletion of 99.95% of bacteria was observed in male and female mice (Fig 5A). Iba-1 and CD68 expression were assessed in hippocampal samples taken from mice sacrificed 1, 4 and 24 hr after chronic restraint stress to compare with the results from control mice (Fig. 4) that had an intact microbiome. Microbiota depletion did not alter Iba-1 expression compared to unstressed control mice, and did not alter changes induced by chronic restraint stress, with a reduction of Iba-1 expression 4 hr after chronic restraint stress in male mice and in female mice a decrease 1 hr after stress that was maintained for 24 hr (Fig 5B). These results indicated that microbiome depletion does not affect microglial Iba-1 expression in control or stressed male or female mice. Differing from Iba-1 expression, activation of microglia measured by CD68 expression was reduced by antibiotics treatment in both male mice and female mice, indicating a decrease in microglial activation in the germ free-like mice (Fig 5C). In addition, CD68 expression was not increased after chronic restraint stress in microbiome-depleted mice in contrast to the increased CD68 expression previously observed in control male and female mice after chronic restraint stress (Fig 5C). These findings indicate that microglial expression of CD68, which is a marker of microglial activation, needs the presence of gut microbes to be triggered by stress in both male and female mice. Comparisons between male and female mice did not show differences.

Figure 5. Chronic restraint stress-altered microglial CD68 expression is influenced by the microbiome.

(A) Eubacteria levels in fecal samples from male and female mice treated with antibiotics for 1 week to deplete bacteria were measured by PCR and compared with non-treated mice (CT). (B-C) Wild-type SPF male and female mice were treated with antibiotics (Antib) for 1 week and then subjected to chronic restraint stress for two weeks and hippocampal expression of microglial markers Iba-1 (B) and CD68 (C) was measured 1h, 4h and 24h after chronic restraint stress and compared with control (CT) mice with intact microbiomes. Each point represents an individual mouse. Bars represent means ± SD. Detailed statistics are in suppl. table 1. *p<0.05, **p<0.01, ***p<0.001. n=3–4 mice/group.

3.5. Effects of chronic restraint stress and (+)-naloxone treatment on depression/anxiety-like behaviors

To study whether molecular and cellular differences between male and female mice after stress affect their behavioral responses to stress, several behavioral tests were performed after chronic restraint stress. In addition, some mice were treated with the TLR4 antagonist (+)-naloxone that we recently reported reverses some deleterious behavioral effects of stress in male mice (Medina-Rodriguez et al., 2020c). Diminished social interactions frequently occur in patients with depression and in rodents exhibiting depressive-like behaviors (reviewed in (Kupferberg et al., 2016)). Male mice displayed reduced sociability after chronic restraint stress in the three-chambered test (Fig 6A), as indicated by reduced nose contacts with a stimulus mouse (Fig 6B), and this was repaired by (+)-naloxone treatment, as previously reported (Medina-Rodriguez et al., 2020c). Female mice also displayed reduced sociability after chronic restraint stress, but differed from male mice in that (+)-naloxone treatment (5 mg/kg) had no effect (Fig 6C). However, administration of a higher dose of (+)-naloxone (10 mg/kg) to female mice provided the same protection observed in male mice with the lower dose of (+)-naloxone of restoring sociability behavior (Fig 6C). These measurements were extended to evaluate social interactions with a second novel mouse (social novelty; Fig 6A), and both male mice and female mice displayed more nose contacts with the novel mouse than with the familiar mouse (Fig 6D,E). Male mice displayed less sociability with the novel mouse after chronic restraint stress, and this impairment was reversed by (+)-naloxone treatment (Fig 6D). Female mice also displayed less sociability with the novel mouse after chronic restraint stress, but treatment with 5 mg/kg of (+)-naloxone only partially reversed this impaired sociability (Fig 6E), whereas a greater effect was observed after administration of a higher dose (10 mg/kg) of (+)-naloxone (Fig 6E).

Figure 6. TLR4 blockade with (+)-naloxone reverses chronic stress-induced impairments in depression-like behaviors.

(A) Scheme of sociability and social novelty tests. (B-J) Wild-type male and female mice subjected to chronic restraint stress (CRS) for two weeks or control (CT) mice were tested to measure depression-like behaviors 1 hr after administration of (+)-naloxone (5 mg/kg for males and 5 mg/kg or 10 mg/kg for females; i.v.) or vehicle. (B-C) The number of nose contacts with an unfamiliar mouse were measured in the three chambered sociability test in male (B) and female (C) mice. (D-E) The number of nose contacts with a familiar vs. a novel mouse were measured in the social novelty test in male (D) and female (E) mice. (F-G) The percentage of time spent in the center and in the periphery of an open field were compared between groups in male (F) and female (G) mice. (H) The locomotor activity of male (top) and female (bottom) mice was measured as the number of beam breaks in an open field during 30 min. (I-J) The number of rearing events in the open field was quantified in male (I) and female (J) mice. Each point represents an individual mouse. Bars represent means ± SD. Detailed statistics are in suppl. table 1. *p<0.05, **p<0.01, ***p<0.001, n=7–15 mice/group.

Chronically stressed mice have been reported to spend less time in the center of a novel open field compared with control mice (Fitzgerald et al., 2019; Kaufmann and Brennan, 2018). In accordance with this, chronic restraint stress caused a reduction in the time spent in the center of a novel open field for male mice (Fig 6F), and this stress-induced reduced center-square time was reversed by (+)-naloxone treatment. Female mice exposed to chronic restraint stress also spent less time in the center of a novel open field than did control female mice, but differently from male mice this behavior was unaffected by 5 mg/kg (+)-naloxone treatment (Fig 6G). However, administration of a higher dose of naloxone (10 mg/kg) effectively repaired the stress-induced reduced center-square time (Fig 6G). These behavioral effects of chronic restraint stress were not due to altered locomotor activity because the total number of beam breaks in the novel open field and the rearing behavior of male and female mice were not altered by chronic restraint stress (Fig 6H–J). Other variables such as dominance/subordinance and olfaction were not identified, which might be considered a potential limitation to the study. (+)-Naloxone treatment of control mice did not alter any of these behaviors (Fig 6B–J). Comparison between males and females did not show differences in any of these behavioral tests after chronic restraint stress (Fig 6).

3.6. Effects of chronic restraint stress and (+)-naloxone treatment on cognition

Novel object recognition (Fig 7A) was previously reported to be impaired in male mice exposed to chronic restraint stress, and this was prevented by (+)-naloxone treatment (Medina-Rodriguez et al., 2020c). Both male mice (Fig 7B) and female mice (Fig 7C) displayed impaired novel object recognition after chronic restraint stress. This impairment in working memory was repaired in male mice by (+)-naloxone treatment (5 mg/kg), and in female mice by treatment with 10 mg/kg, but not with 5 mg/kg, (+)-naloxone (Fig 7C). Comparison between males and females did not show differences in their behavior after chronic restraint stress (Fig 7B–C).

Figure 7. Treatment with (+)-naloxone reverses chronic stress-induced cognitive deficits.

(A-L) Wild-type male and female mice subjected to chronic restraint stress (CRS) for two weeks or control (CT) mice were tested to measure working and spatial memories 1 hr after administration of (+)-naloxone (5 mg/kg for males and 5 mg/kg or 10 mg/kg for females; i.v.) or vehicle. (A) Scheme of novel object recognition test. (B-C) Novel object recognition test: the time exploring a novel object vs. a familiar object was compared for each experimental group in male (B) and female (C) mice. (D) Scheme of temporal order recognition test. (E-F) Temporal order recognition test: the time exploring an object presented earlier at the beginning of the test vs. an object recently presented was compared for each experimental group in male (E) and female (F) mice. (G) Scheme of coordinate spatial processing test. (H-I) Coordinate spatial processing test: each mouse was allowed to explore two novel objects that were 45 cm apart for 15 min. After 5 min in an opaque chamber, each mouse was allowed to explore for 5 min the same two objects that had been moved closer together (30 cm). The exploration ratio was calculated as time (exploring during the 5 min test session)/(exploring during the 5 min test session plus the last 5 min of the habituation session). The exploration ratio was measured for each experimental group in male (H) and female (I) mice. (J) Scheme of categorical spatial processing test. (K-L) Categorical spatial processing test: each mouse was allowed to explore two novel objects for 15 min. After 5 min in an opaque chamber, each mouse was allowed to explore the same two objects that had been switched from their original positions for 5 min. The exploration ratio was calculated as time (exploring during the 5 min test session)/(exploring during the 5 min test session plus the last 5 min of the habituation session). The exploration ratio was measured for each experimental group in male (K) and female (L) mice. Each point represents an individual mouse. Bars represent means ± SD. Detailed statistics are in suppl. table 1. **p<0.01, ***p<0.001, n=7–15 mice/group.

Temporal order recognition, in which mice spend less time exploring an object recently explored compared with an object previously presented (Fig 7D), was impaired in both male mice (Fig 7E) and female mice (Fig 7F). (+)-Naloxone (5 mg/kg) treatment completely reversed the impairment in temporal order memory in male mice. This dose of (+)-naloxone partially reversed the impairment in female mice, but a higher dose of 10 mg/kg completely reversed the impairment (Fig 7F). Comparison between males and females did not show differences in their behavior after chronic restraint stress (Fig 7E–F).

Two tasks were used that measure spatial memory, the coordinate and categorical spatial processing tasks. Coordinate spatial processing (Fig 7G) involves moving two objects from 45 to 30 cm apart and comparing exploration time to the exploration time during the last 5 min of the habituation phase. Control male and female mice explored the objects more time after they were moved, which results in an exploration ratio >0.5 (Fig 7H, I). After chronic restraint stress, male and female mice displayed decreased exploration ratios (Fig 7H, I). (+)-Naloxone treatment totally prevented the stress-induced impaired coordinate spatial processing memory in male mice (Fig 7H). No significant effect in female mice was observed after treatment with 5 mg/kg (+)-naloxone, but a total restoration of the spatial memory was observed when treated with 10 mg/kg (+)-naloxone (Fig 7I). Comparison between males and females did not show differences in their behavior after chronic restraint stress (Fig 7H–I).

The categorical spatial processing task involves switching the position of two objects and comparing exploration times to the last 5 min of the habituation phase (Fig 7J). Control male and female mice explored the objects more time after their positions were switched, and this was impaired after chronic restraint stress (Fig 7K, L). Treatment with (+)-naloxone restored the impaired categorical spatial processing memory in male mice (Fig 7K), but the effect in female mice treated with 5 mg/kg (+)-naloxone (Fig 7L) was not significant. Administration of a higher dose of (+)-naloxone was necessary to result in a significant effect in female mice (Fig 7L). Treatment of control male and female mice with (+)-naloxone did not exert any effect on these measurements of memory (Fig 7). Comparison between males and females did not show differences in their behavior after chronic restraint stress (Fig 7K–L).

Altogether these data show that male and female mice display similar cognitive impairments in response to chronic restraint stress, but a higher dose of (+)-naloxone is required in female mice than male mice to achieve ameliorative effects on stress-induced impaired cognition.

3.7. Effects of chronic restraint stress and (+)-naloxone treatment on TLR4 and inflammation

Because TLR4 is a major mediator of the inflammatory response to stress (Cheng et al., 2016), we measured TLR4 expression in the hippocampus 1, 4 and 24 hr after chronic restraint stress (Fig 8). Male and female mice had the same basal levels of TLR4 expression (Fig 8A), but there was a larger increase of TLR4 expression in female mice than male mice in response to stress (Fig 8B). Microbiota depletion by antibiotic treatment decreased TLR4 expression in the hippocampus in male and female mice, and blocked TLR4 expression increases following chronic restraint stress (Fig 8C), suggesting a regulatory role for the microbiome in the stress-induced increase in TLR4 expression. No differences between male and female mice were found after the antibiotic treatment (Figure 8C)

Figure 8. TLR4 expression is increased by chronic restraint stress in mice with intact microbiomes but not in mice with depleted bacteria, and changes in TLR4 expression and microglial markers induced by chronic restraint stress are diminished by TLR4 blockade by (+)-naloxone.

(A) Basal expression of TLR4 was measured by PCR in the hippocampus of wild-type male and female mice. (B) Wild-type control (CT) male and female mice were subjected to chronic restraint stress for two weeks and hippocampal expression of TLR4 was measured 1h, 4h and 24h after chronic restraint stress. (C) Wild-type SPF male and female mice were treated with antibiotics (Antib) for 1 week to deplete their microbiome and then subjected to chronic restraint stress for two weeks and hippocampal expression of TLR4 was measured 1h, 4h and 24h after chronic restraint stress and compared with SPF control (CT) mice with intact microbiomes. (D-I) Wild-type male and female mice subjected to chronic restraint stress (CRS) or control (CT) mice were sacrificed after a battery of behavioral tests, which was12 days after the last session of CRS. Mice that received (+)-naloxone treatment (5 mg/kg for males and 5 mg/kg or 10 mg/kg for females; i.v.) while they were subjected to the behavioral tests were compared to those injected with vehicle. TLR4 (D males, E females), Iba-1 (F males, G females) and CD68 (H males, I females) expression was measured. Each point represents an individual mouse. Bars represent means ± SD. Detailed statistics are in suppl. table 1. *p<0.05, **p<0.01, ***p<0.001. n=3–10 mice/group.

We tested if stress-induced changes in TLR4 expression were long-lasting by measuring it 12 days after the end of the chronic restraint stress after completing behavioral measurements. No changes in hippocampal TLR4 expression were observed at this time in male mice (Fig 8D). However, increased hippocampal expression of TLR4 was found in stressed female mice compared to control mice (Fig 8E), thus, there was a higher TLR4 expression in female compared to male mice after chronic restraint stress (p=0.0268), and this increase was reduced by administration of 10 mg/kg of (+)-naloxone (Fig 8E).

There were also long-term changes in levels of the microglial marker Iba-1. Chronic restraint stress caused reduced Iba-1 in both male and female mice, and this was normalized by administration of (+)-naloxone (Fig 8F, G). Chronic restraint stress also caused long-lasting increases in the activated microglial marker CD68, which was higher in female than male mice (p<0.001), which were blocked by (+)-naloxone treatment (Fig 8H, I). Altogether these results may suggest that stress induces a stronger effect on microglia activation in female than male mice, with increased TLR4 expression that requires a higher dose of (+)-naloxone to be blocked.

We measured the levels of the cytokines TNFα, IFN-β, IL-17A and IL-6 twelve days after the end of the chronic restraint stress, after all the behavioral tests were completed, with or without (+)-naloxone treatment. There was increased TNFα in the hippocampus of male and female mice (Fig 9A, B) and in the PFC of male, but not female, mice (Fig 9C,D), and the increases were eliminated by (+)-naloxone treatment. IFN-β remained unchanged in the hippocampus of male and female mice after chronic restraint stress (Fig 9E, F), but was increased in the PFC of male and female mice (Fig 9G,H), and the increases were eliminated by (+)-naloxone treatment. IL17-A was increased in the hippocampus and PFC of male, but not female, mice (Fig 9I–L), and (+)-naloxone treatment did not attenuate the increases. IL-6 was increased only in the hippocampus of female mice (Fig 9M–P), and the increase was not eliminated by (+)-naloxone treatment. These results after (+)-naloxone treatment are in agreement with our previous report that TLR4 blockade by (+)-naloxone involves a decrease in TNFα and IFN-β in the hippocampus to exert its antidepressant-like and procognitive effects in stressed male mice, while other inflammatory cytokines such as IL-17A or IL-6 are not affected by (+)-naloxone (Medina-Rodriguez et al., 2020c) and also showed that a higher dose of (+)-naloxone is required to be effective in female mice than in male mice. Male mice vs. female mice comparisons resulted in significant differences in TNFα in PFC (p<0.001) and IL-17A in the PFC (p=0.024), after chronic restraint stress (Figure 9).

Figure 9. Effects of (+)-naloxone treatment on cytokines induced by chronic restraint stress.

(A-P) Wild-type male and female mice subjected to chronic restraint stress (CRS) or control (CT) mice were sacrificed 12 days after the last session of CRS after being subjected to a battery of behavioral tests. Levels of TNFα (A-D), IFN-β (E-H), IL-17A (I-L) and IL-6 (M-P) in the hippocampus and PFC of mice that received (+)-naloxone treatment (5 mg/kg for males and 5 mg/kg or 10 mg/kg for females; i.v.) while they were subjected to the behavioral tests were compared to those injected with vehicle. Each point represents an individual mouse. Bars represent means ± SD. Detailed statistics are in suppl. table 1. *p<0.05, **p<0.01, ***p<0.001. n=4–5 mice/group.

To further evaluate the mediation by TLR4 of chronic stress-derived behavioral effects, we tested the behavioral effects of chronic restraint stress in TLR4 knockout mice (TRL4^−/−). TLR4 deficiency in male and female mice had an antidepressant-like effect in the tail suspension test since TRL4^−/− mice had decreased immobility time with respect to WT mice after chronic restraint stress (Fig 10A). Following chronic restraint stress, both sociability and social novelty interactions were greater in both TRL4^−/− male and female mice than stressed wild-type mice (Fig 10B, C), indicating a regulatory role of TLR4 in the induction of stress-mediated depression-like behaviors. In addition, working and spatial memories measured by novel object recognition, temporal order recognition, coordinate and categorical spatial processing tests were evident in TRL4^−/− male and female mice after chronic restraint stress, opposite to their impairments in wild-type mice (Fig 10D, G). These results, together with the ameliorative effects of treatment with the TLR4 antagonist (+)-naloxone, indicate that stress-induced TLR4 activation mediates these impairments in cognition. The antidepressant-like and pro-cognitive effects observed in the absence of TRL4 might be due at least in part to the lack of microglial activation after stress since CD68 expression in TRL4^−/− male and female mice subjected to chronic restraint stress did not change compared to their non-stressed counterparts (Fig 10H). Male mice vs. female mice comparisons did not show significant differences.

Figure 10. TLR4 knockout male and female mice are resilient to chronic restraint stress.

(A-G) Wild-type (WT) and TLR4 knockout (TLR4^−/−) male and female mice subjected to chronic restraint stress (CRS) were tested in the tail suspension test (A), sociability (B), social novelty (C), novel object recognition (D), temporal order recognition (E), coordinate spatial processing (F), and categorical spatial processing (G). (H) Microglial activation was quantified by measuring CD68 expression levels in TLR4^−/− male and female mice subjected or not to chronic restraint stress. Each point represents an individual mouse. Bars represent means ± SD. Detailed statistics are in suppl. table 1. *p<0.05, **p<0.01, ***p<0.001, n=3–5 mice/group.

3.8. Effects of chronic restraint stress and (+)-naloxone treatment in microbiome-depleted mice

To investigate whether the effects of chronic restraint stress and their reversal by (+)-naloxone treatment are influenced by the composition of the microbiome, male and female mice were treated with antibiotics to induce a germ free-like phenotype (Medina-Rodriguez et al., 2020b; Reikvam et al., 2011) Mice were then subjected to chronic restraint stress for two weeks, and behavioral tests were carried out. First, we noted that the treatment with antibiotics alone was sufficient to alter some behaviors, and these changes were similar in male and female mice. Thus, while antibiotic treatment did not affect sociability (Fig 11A), the time spent in the center of an open field (Fig 11B), or spatial memory (Fig 11C, D), it impaired working and temporal memories (Fig 11E, F).

Figure 11. TLR4 blockade with (+)-naloxone reverses chronic stress-induced impairments independently of the microbiome.

(A-F) Control (CT) wild-type male and female mice, male and female mice subjected to antibiotic treatment for 1 week to deplete the microbiome (Antib), male and female mice subjected to chronic restraint stress for 2 weeks (CRS), male and female mice treated with antibiotics and then subjected to chronic restraint stress (Antib+CRS) and male and female mice treated with antibiotics, subjected to chronic restraint stress and treated with (+)-naloxone (5 mg/kg for males and 5 mg/kg or 10 mg/kg for females; i.v.; Antib+CRS+(+)-naloxone) were tested to measure depression and anxiety-like behaviors and cognition. The number of nose contacts with a unfamiliar mouse in the three chambered sociability test (A), the percentage of time spent in the center of an open field (B), the ability to realize that two objects have been moved closer together (Coordinate spatial processing) (C), or that two objects have been switched from their original position (Categorical spatial processing) (D), the time exploring a novel object vs. a familiar object (Novel object recognition) (E), and the time exploring an object presented earlier at the beginning of the test vs. an object recently presented (Temporal order recognition) (F), were measured, Each point represents an individual mouse. Bars represent means ± SD. Detailed statistics are in suppl. Table1. *p<0.05, **p<0.01, ***p<0.001, n=7–8 mice/group.

After CRS, some behaviors remained impaired despite the depleted microbiome, including sociability (Fig 11A) and working and temporal memories (Fig 11E, F), and treatment with (+)-naloxone still effectively reversed the impairments in these behaviors, indicating that the beneficial effects of (+)-naloxone on these behavioral responses to stress are independent of the microbiome. Microbiome depletion in CRS-treated mice increased the time spent in the center of an open field to become equivalent to that of non-stressed mice (Fig 11B), and also normalized behaviors in the coordinate and categorical spatial processing measurements (Fig 11C, D), indicating that the microbiome contributes to these behavioral impairments induced by chronic stress. Male mice vs. female mice comparisons did not show significant differences.

4. DISCUSSION

Studying the differences and similarities in females and males in response to stress, one of the main risk factors triggering depression, is needed due to the higher prevalence of depression in women than men (Belmaker, 2008; Kessler et al., 2005). However, most preclinical research concerning depression utilizes male rodents, in part due to the general belief that female rodents are more variable than males (Beery and Zucker, 2011), and because female rodents do not display some of the most widely used models of depression-like behavior, such as learned helplessness (Dalla et al., 2008) and the social defeat model (Hammels et al., 2015). Since stress is the most prevalent risk factor for depression (de Kloet et al., 2005), we studied the effects of chronic restraint stress on female and male mice behaviors and inflammatory responses, which revealed both similarities and differences between the groups. The hippocampus and PFC were studied because of the widely known involvement of these two brain regions in the pathophysiology of depression, e.g., depression is associated with a volumetric decrease of these brain regions (reviewed in (Belleau et al., 2019; Duman and Aghajanian, 2012)). Similarities in female and male mice observed in the present study include equivalent basal levels in most of the measured inflammatory cytokines and similar increases in response to chronic restraint stress, although differentially orchestrated depending on the time and brain region examined, similar changes in microglial markers, and similar behavioral responses, including depression-like behaviors and cognitive impairments. Observed differences include different increases in expression levels of TLR4 in response to chronic restraint stress, which is higher in female than male mice, and different efficacy of (+)-naloxone treatment, which was tested for reversing the deleterious effects of stress by blocking TLR4 and was found to be effective in both sexes but a higher dose was needed in female mice to restore microglial status and pro-inflammatory cytokines levels and to reach the same antidepressant-like and procognitive effects as those observed in male mice with a lower dose. Thus, although the measured behavioral outcomes are generally similar in female and male mice after chronic restraint stress, the immune processes underlying depression-like behaviors and cognitive impairments induced by stress differ somewhat, as well as the (+)-naloxone treatment needed to revert the deleterious effects of stress.

The activation of TLR4 by chronic restraint stress led to a generalized increase of proinflammatory cytokines in plasma, PFC and hippocampus in both female and male mice. However, not all the cytokines responded in the same way or at the same time point in female and male mice. The literature comparing cytokines between females and males is variable, likely due in part to these regional and time-dependent differences. For example, while some studies report that the TLR4 ligand lipopolysaccharide induces greater increases of the cytokines IL-6, TNFα and IL-10 in male mice than in female mice (Kuo, 2016), the lipopolysaccharide challenge that leads to depressive-like behavior increased expression of TNFα and IL-6 in the hippocampus and brainstem of female, but not male, rats (Tonelli et al., 2008). Interestingly, IFNβ one of the main cytokines previously reported to mediate the effect of (+)-naloxone in male mice and microglial cell lines (Medina-Rodriguez et al., 2020c; Wang et al., 2016) was found at the same levels in female and male mice in plasma and hippocampus, but the basal level in the PFC was higher in female mice compared to male mice.

The TLR4 antagonist (+)-naloxone needed to be administered at a higher dose to female mice than to male mice to exert the same antidepressant-like effects in the measured depression-like behaviors and to counteract cognitive impairments induced by chronic restraint stress. Also 10 mg/kg (+)-naloxone was necessary in female mice to totally revert IFNβ increases after chronic restraint stress, while in male mice this was achieved with 5 mg/kg (+)-naloxone. This lesser effect of (+)-naloxone in female mice than male mice might also be due to higher basal level of IFNβ observed in female mice compared to male mice. The same (+)-naloxone concentration-dependent pattern was observed for reducing stress-induced TNFα, which was previously reported to be reduced when TLR4 is blocked by (+)-naloxone in male mice (Medina-Rodriguez et al., 2020c; Wang et al., 2016). These finding agree with previous preclinical and clinical studies which showed that females require more morphine than males to produce similar levels of analgesia (Boyer et al., 1998; Cepeda and Carr, 2003; Kepler et al., 1989) since morphine not only activates opioid receptors but also TLR4 (Farzi et al., 2015; Wang et al., 2012). The lack of effect of (+)-naloxone in reducing IL-17A or IL-6 after chronic stress in male mice is also in accordance with our previous results using learned helplessness as a different paradigm of stress to induce depression-like behaviors (Medina-Rodriguez et al., 2020c). The higher expression of TLR4 observed here in female mice compared to male mice might contribute to the need of a higher dose of the antagonist (+)-naloxone to block TLR4-mediated effects, and it matches with previous findings that in general show that innate and adaptive immune responses in females are greater than in males (reviewed in (Klein and Flanagan, 2016)). TLR7 has also been reported to be expressed at a higher level in female than male leukocyte populations (Souyris et al., 2018). Additionally, TLR4 expression and pro-inflammatory cytokine production are decreased in murine macrophages after removal of endogenous estrogens, and 17β-estradiol reverses this effect (Rettew et al., 2009). Oppositely, monocytes/macrophages express more TLR4 in the absence of endogenous testosterone (Rettew et al., 2008). These previous observations are in agreement with our finding of a higher stimulated expression of TLR4 in female mice compared to male mice in response to chronic restraint stress. However, reports of sex-dependent inflammatory responses are variable, for example administration of the TLR4 ligand lipopolysaccharide in adult mice induces greater increases of the cytokines IL6, TNFα and IL10 in male mice than in female mice (Kuo, 2016).

Chronic stress has been associated with alterations of microglia (Bollinger et al., 2016; Kreisel et al., 2014), which are the major TLR4 expressing cells in the CNS (Lehnardt et al., 2003). The decrease in Iba-1 expression found in the present study is in agreement with previous reports that Iba-1 expression decreases after chronic stress (Kreisel et al., 2014; Tong et al., 2017) and show that microglial responses to stress are also slightly differentially regulated over time in female and male mice. However, results of studies on microglial activation vary depending on the stress model used and timing, and the microglial state seems to vary over the course of stress (reviewed in (Wang et al., 2022; Woodburn et al., 2021a)). It was recently proposed that while synapse loss continues during chronic stress, microglia, which might mediate the synapse loss, decrease actions on neuronal remodeling over time (Woodburn et al., 2021b). CD68 expression, which corresponds to activated microglia, seems equally increased in female and male mice and corroborates studies showing that activated microglia mediate some stress responses (Ramirez et al., 2017; Sugama et al., 2019; Zhang et al., 2018). The fact that Iba-1 expression and CD68 are different suggest that there are dynamic changes in microglia induced by stress, as has been previously reported (Gong et al., 2018; Kreisel et al., 2014; Woodburn et al., 2021b). In fact, changes in Iba-1 which seems initially transient and disappeared 24 hr after the last session of chronic restraint stress, appeared again when measured 12 days later, correlating with also apparently transient changes observed in some cytokines 24 hr after chronic restraint stress that increased again later when measured after 12 days, which suggests that microglia might be primed due to chronic stress and respond rapidly when subjected to a mild stress later, such as the behavioral tests (Kreisel et al., 2014; Lloyd et al., 2019). A potential explanation to this increased sensitivity to acute stress after previous chronic stress exposure might be related to the hypothalamic-pituitary-adrenal (HPA) axis. In humans, it is believed that alterations of cortisol levels at the time of a psychological trauma result in damage to the hippocampal neurons that may persist for many years after the original trauma and this could involve differences in stress responsivity (Lucassen et al., 2014). Interestingly, again the dose of (+)-naloxone that was effective in male mice failed to restore microglial marker expression in female mice, in which a higher dose was necessary. This finding may help to explain the different effects of (+)-naloxone observed in male mice compared with female mice, as the lack of antidepressant and procognitive effects observed in female mice after (+)-naloxone treatment with the lower dose might be due to the inability of this dose to totally restore Iba-1 and CD68 expression in microglial cells. The important roles of TLR4 and microglial dynamics in controlling responses to stress was further supported by the results of experiments that used TLR4 knockout female and male mice, which were both resistant to chronic restraint stress-mediated changes in microglia and behavioral impairments. These results are in agreement with previous studies reporting the importance of the immune system, and particularly TLR4 and microglial disturbances, in depression (Liu et al., 2014; Yirmiya et al., 2015) and extend our previous studies in TLR4 knockout male mice, which were reported to be resistant to the learned helplessness model of depression (Cheng et al., 2016; Medina-Rodriguez et al., 2020c).

Alterations in the microbiome are being implicated as contributors to many CNS illnesses, including depression (Du Toit, 2019; McGuinness et al., 2022; Medina-Rodriguez et al., 2020b; Yang et al., 2020). Although further investigation is needed, microbiome depletion via antibiotic treatments decreased the expression of both TLR4 and CD68, which corroborates previous results that antibiotic treatments decrease microglia inflammatory responses via a pathway involving TLR4 (Zusso et al., 2019). Regarding the role of the microbiome in modulating susceptibility to stress-induced depression-like behaviors, our observations that microbiome depletion by antibiotic treatments differentially affects the relevant behaviors is in agreement with previous studies showing that anxiety-like behaviors and spatial memories are unaltered by antibiotic treatment, whereas it impairs novel object recognition (Frohlich et al., 2016). The fact that not all the behaviors studied show the same responses to microbiome depletion supports the idea that the effects of microbiome depletion are not generalized in the brain and different brain regions and pathways likely differentially contribute to individual behavioral outcomes following stress. However, although these particular observations require further investigations, both female and male mice were affected similarly following microbiome depletion and we also found that the beneficial effects of (+)-naloxone in coping with stress appear to be independent of the microbiome.

In addition, this study extends our previous finding of an antidepressant-like effect of (+)-naloxone in male mice (Medina-Rodriguez et al., 2020c). Here, we demonstrated that (+)-naloxone has many similar effects in female mice. (+)-Naloxone is the stereoisomer of (−)-naloxone, which is a drug widely used in humans to treat opioid overdose, is CNS-penetrant and may be administered safely to human subjects, with the advantage that (+)-naloxone, unlike (−)-naloxone, does not block opioid receptors, avoiding the side effects of blocking opioid receptors (Wang et al., 2016). Although further studies might be needed since the high number of conditions, time points, and inclusion of both sexes for each experiment in the present work often limited the sample size, the present findings increase the potential translation to the clinic of (+)-naloxone as a potential treatment for depression. This study also was limited because it used mice, and other variables such as dominance/subordinance and olfaction were not identified, which might be considered a potential limitation to the study.

In summary, it is important to increase the use of female rodents in preclinical studies of psychiatric disorders, such as depression, to avoid biased results since many behavioral and neurological differences previously have been reported between females and males in response to stress (Autry et al., 2009; Dalla et al., 2008; Hodes et al., 2015b; Peay et al., 2020). This study contributes to evidence of sex-related differences and similarities between female mice and male mice in stress-induced depression-like behaviors, and corroborates that TLR4 mediates many of the detrimental effects of stress, and provides further evidence that males and females should not be considered equal in the search for treatments for depression.

Supplementary Material

1

Suppl. Figure 1. Experimental timeline.

2

Suppl. Table 1. Detailed statistics for Figures 1–11.

Highlights:

• Depression prevalence is higher in women than men

• Stress induces different patterns of inflammatory mediators in male and female mice

• Antidepressant-like effects of (+)-naloxone are greater in male than female mice

• Sex-specific treatments might be needed to treat depression