Chronic pain is one of the most prevalent and debilitating conditions affecting as many as 20% of the population worldwide (Goldberg and McGee, 2011). It is the foremost cause of long-term sick leave, and disability imposing a profound health care burden that not only affects the individual but also permeates throughout all avenues of social, work, and family life (Lynch, 2011; Gaskin and Richard, 2012). However, despite the prevalence of chronic pain, it remains challenging to treat, in part because it is associated with a high incidence of comorbid conditions, including anxiety and depression (Nicholson and Verma, 2004). Approximately 50% of patients diagnosed with major depressive disorder reportedly suffer from chronic pain, and the risk of developing comorbid depression in chronic pain patients is greater among women (Radat et al., 2013). Further, patients suffering from chronic pain and depression have a poorer prognosis than patients suffering from each illness alone (Arnow et al., 2006; Munce and Stewart, 2007). Clinically, it is well known that chronic pain and depression are related; however, the neurobiological changes responsible for the co-occurrence of these conditions are not well-understood, and the sex-specific mechanisms that might help explain the high prevalence of these conditions in women remain unknown.

Microglia, a principal immune cell, comprise approximately 5–12% of the glial cell population and emerge from either erythromyeloid precursors of the embryonic yolk sac or invade the central nervous system (CNS) via myeloid progenitors and proliferate during embryonic and postnatal development (Hristovska and Pascual, 2015). Microglia interact with and prune synapses during healthy brain development to modify their structure and function (Chen et al., 2018b). As such, in a healthy resting state, microglia actively sense the environment to maintain normal physiological conditions; however, following tissue damage or inflammation, microglia transition from an inactive ramified state to an active amoeboid state – characterized by enlarged cell bodies and shortened processes and increased expression of microglial markers such as ionized calcium-binding adapter molecule 1 (IBA-1) and CD11b (Ohsawa et al., 2004; Ji et al., 2013; Chen et al., 2018b). A few studies have demonstrated that chronic pain hypersensitivity is dependent on microglia in male but not female rodents (Sorge et al., 2015; Chen et al., 2018b; Mapplebeck et al., 2018). Further, spinal microglial activation promotes BDNF release and pain hypersensitivity in male but not female rats – an effect that is conserved in humans (Dedek et al., 2021).

An abundance of research has focused on spinal mechanisms; however, there is evidence to suggest that brain microglia may contribute to the development and maintenance of chronic pain and its associated comorbidities. A study by Miyamoto et al. (2017) showed that intracerebroventricular injections of minocycline – a broad-spectrum antibiotic known to inhibit glial cell activity – reduced microglial activation in the anterior cingulate cortex (ACC) and reversed tactile allodynia. In a rat model of diabetic neuropathy, ammoxetine, a novel, potent serotonin, and norepinephrine reuptake inhibitor, reversed mechanical allodynia and depressive-like behavior that coincided with decreased spinal microglial activation (Zhang et al., 2018). Furthermore, a recent study using a rodent model of neuropathic pain, called chronic constriction injury (CCI), revealed that male mice developed affective disorders, albeit at a delayed time point, which was associated with activation of microglia characterized by increased numbers and cell body size in the prefrontal cortex, amygdala, and hippocampus (Barcelon et al., 2019).

There is also evidence to suggest that brain astrocytes may be involved in regulating pain sensitivity and affect following injury (Chen et al., 2018b). Several studies have demonstrated that activated astrocytes in the ACC and hippocampus are involved in the maintenance of chronic pain and the onset of depression (see Li et al., 2019 for a review). Narita et al. (2006) demonstrated that peripheral nerve injury in mice increased glial fibrillary acidic protein (GFAP) expression in the ACC, which was associated with increased affective behaviors. Similarly, Ikeda et al. (2013) used a mouse model of inflammatory pain and showed increased astrocytic expression in the ACC resulted in increased affective behavior that was attenuated with the administration of L-α-aminoadipate (L-AA), an astroglia toxin. Further, the administration of antidepressant drugs inhibits hippocampal astrocyte activation and reduces nerve injury-induced pain hypersensitivity (Zhu et al., 2009). However, there seems to be a paucity of research on astrocytes in chronic pain and depression, necessitating further investigation.

The primary goal of the present study was to identify whether sex influenced the development of depressive-like behavior in mice following spared nerve injury (SNI). A secondary goal was to characterize whether brain glial activation was increased in a time-, sex-, and-, brain region-dependent manner following injury. Based off clinical studies showing a greater degree of comorbid depression in female chronic pain patients (Munce and Stewart, 2007), we hypothesized that female mice would display greater and more robust signs of depressive-like behavior, which would be associated with increased glial activity. We used a battery of behavioral tests and examined two early time points (2 and 3 weeks) following nerve injury. Microglia and astroglia in the ACC (ventral and dorsal compartments) and the nucleus accumbens were quantified given data supporting a role for these regions in clinical anhedonia and response to treatment (Pizzagalli et al., 2001; Wacker et al., 2009; Carl et al., 2016).

The current study used a mouse model of chronic neuropathic pain and investigated the development of depressive-like behavior in male and female mice. We also characterized changes in microglia number and astrocyte density as a measure of activation following injury. Overall, there were sex differences in the onset of depressive-like behaviors in male and female mice such that males displayed immobility on the forced swim test at an earlier time point than females, and only female mice developed immobility on the tail suspension task following nerve injury. In the contralateral vACC, microglia cell number was significantly greater for female SNI mice than sham. No other changes in glial cells were apparent.

The onset of the depressive-like phenotype was different in male and female mice, which mirrors the human literature on the prevalence of depression and shows different onset rates for each sex (Romans et al., 2007). In our study, males developed the depressive phenotype 2-week following injury, while females developed depressive- and anxiety-like behavior at 3-week, suggesting that initially, females may be more resilient than males, or changes in affect may take longer to manifest in females. However, depressive-like behaviors have been shown to manifest in male mice at later time points (i.e., 8-week following injury) when the chronic constriction injury model of neuropathic pain was used (Barcelon et al., 2019). We do not believe that the effects of nerve injury on FST immobility are related to movement because nerve injury did not affect walking or general movement in the OFT. However, the possibility exists that depending on sex, behavioral tests such as the FST and TST might be more sensitive for the detection of subtle behavioral differences. As another behavioral measure, we assessed anhedonia using the sucrose preference test; however, nerve injury did not alter sucrose intake. We do not believe that the lack of effect for the sucrose preference test was related to our testing procedures because during pilot studies sucrose intake was decreased when we used a positive control (i.e., restraint stress, data not shown). Anhedonia is challenging to induce in mice, and various pain models fail to cause a robust change in sucrose preference following injury (Schwartz et al., 2014). The sucrose preference test did reveal lower baseline preference for female than male mice. This is interesting because there have been reported sex differences in sucrose preference, but they typically are in the opposite direction with female rodents showing a greater preference (McCall et al., 2013). However, a few studies have shown that behavioral taste responses to dilute sucrose solutions are decreased by estrogen, whereas ovariectomy abolishes this effect (Curtis et al., 2004). This suggests that higher levels of estrogen in female mice may elevate the threshold for gustatory detection of sweet taste, a possibility that has been previously suggested (Curtis et al., 2005). Thus, sex differences in this assay may depend on estrogen-modulation of taste responses to specific concentrations of sucrose in these testing procedures.

Human studies are far more complicated than the current mouse experiments and the prevalence of comorbid depression with pain may depend on how and where the patients were assessed and the criteria for depression used, such as severity, assessment method and sample selection (Okifuji and Turk, 2016). While many studies have examined the prevalence of comorbid pain and depression (Tunks et al., 2008; Rayner et al., 2016), few studies fail to specifically examine coping strategies when patients suffer from both conditions as well as recovery time for pain and/or depression. A human study by Rovner et al. (2017) demonstrated that although the severity of chronic pain was the same for females and males, females were more accepting of the pain, remained more active, and reported fewer mood disturbances than males, supporting the idea that initially, females are better able to cope with chronic pain. In humans, females are more likely to endorse rumination (Meints et al., 2017) and rely on social support (Rovner et al., 2017) than males. These coping strategies may be the result of female patients reporting greater levels of pain dismissal (Igler et al., 2017), which may contribute to delayed-onset depression. Unfortunately, no study using a human population has examined whether there is a sex-difference in the relationship between chronic pain duration and the onset of depression. Thus, it is possible that female patients are likely to cope with pain better than males, but when coping strategies fail, such as lack of social support the consequence may be increased prevalence of depression.

In addition, pain severity, pain duration and number of pain locations are associated with the recurrence of depressive and anxiety disorders (Gerrits et al., 2014; Sharpe et al., 2017). To our knowledge, no single study has examined whether male patients recover from comorbid depression quicker than females; however, inflammatory markers are predictive of depression in men, but not in women suggesting that the etiology of depressive disorders within the patient population are different between genders (Ernst et al., 2021). In our study, increased inflammation may have contributed to the early behavioral signs of depression in male mice, however, we did not measure general inflammation at the 2-week time point and cannot be certain that this was the precise mechanism. Further, male mice recovered from the depressive phenotype at 3-week, suggesting that males may adapt to chronic pain and eventually recover from affect-related disorders. These results are in line with studies in mice (Sellmeijer et al., 2018) and may explain why there is a higher incidence of depression in women with chronic pain than men (Munce and Stewart, 2007).

While several research groups have used microglial cell body size to quantify microglia activation (Taylor et al., 2017; Guneykaya et al., 2018), we used microglial cell number as a less subjective metric (Barcelon et al., 2019). In the contralateral vACC female SNI mice displayed significantly more microglia cell bodies than sham controls whereas SNI mice did not display overall changes in astrocyte intensity. Previous research has indicated that sex differences exist for microglia in the brain, but following nerve injury, the degree of microglia and astrocyte activation is similar between the sexes (Chen et al., 2018a). A possible explanation for this is that following nerve injury, cortical microglia may not be activated in the same way that spinal microglia have been characterized (Coull et al., 2005). Recent evidence suggests that there is an interaction between the spinal dorsal horn and the ACC in pain modulation, but mechanisms in the spinal cord do not necessarily transfer to cortical mechanisms (Tsuda et al., 2017). Previous reports characterizing brain microglia have mainly used CCI-induced nerve injury in male mice, suggesting that injury type may impact microglial or astrocyte activation in the brain differently between the sexes (Taylor et al., 2017).

We used Iba1 and GFAP as molecular markers for activated microglia and astrocytes, respectively, because of their roles in demonstrating morphological features at different microscopic levels (Shapiro et al., 2008; Sofroniew and Vinters, 2010). However, limitations do exist with these two markers and should be noted. Iba1 is not only a marker for activated microglia but also other macrophages that are recruited during nerve injury as well as other subpopulations of microglia (i.e., ramified microglia) (Ohsawa et al., 2004; Shapiro et al., 2009). So, Iba1 immunohistochemical expression may not be limited to just activated microglia and thus may affect analysis and interpretation of results. It is challenging to acquire quantitative data with the IHC techniques used in the current paper. While some, but not all papers complement their IHC results with qPCR for verification, we did not do this and is a limitation of the present study. In addition, GFAP does not always label healthy CNS astrocytes (Sofroniew and Vinters, 2010), which could alter the difference observed in astrocyte density between nerve-injured and control mice. Further, we considered using Sholl analysis to capture astrocyte complexity; however, in our staining the larger size of the astrocytes and their lack of clear definition, made it difficult to determine whether branches were from one astrocyte or many. The center point of many astrocytes was also not always evident, making other quantification methods difficult. Thus, it would be important to consider the use of complementary techniques or other activated microglia and astrocyte markers, such as TMEM-119 (a transmembrane protein found only on microglia) (Bennett et al., 2016) and SOX9 (expressed exclusively by astrocytes) (Sun et al., 2017), respectively. These markers may be better suited for detecting active glia states and understanding reactive changes in glial cells.

Since, spinal nociceptive afferents are expected to innervate the contralateral side of the brain, our analysis considered whether lateralization occurred in any of the brain regions following injury as previously shown (Taylor et al., 2017). Our results stand in contrast to Taylor et al. (2017), where robust lateralization and regional differences were uncovered; however, we did find a lateralization effect for the number of microglia in the vACC of female SNI mice. Here, microglia cell number per square millimeter was significantly greater in SNI versus sham mice. Notably, there were minimal changes in microglia, and astrocytes even though several other studies have shown microglia changes in the thalamus, amygdala, ventral tegmental area (VTA), nucleus accumbens, bed nucleus of the stria terminalis, and periaqueductal gray following peripheral nerve injury (Taylor et al., 2015; Ni et al., 2016; Liu et al., 2017). Given that we found minimal differences, we did not pursue a mechanistic line of inquiry and we do not know whether glial changes in the ventral ACC are related to the depressive-like behavior in female SNI mice. As with some of our behavioral results, comparing between the 2- and 3-week time points is difficult because there may be inherent baseline expression differences between the mice. There may also be inherent sex differences in microglial expression as previously shown with male microglia being more numerous in the cortex, hippocampus, and amygdala (Guneykaya et al., 2018).

Further investigations may want to explore whether microinjections of glial inhibitors into either the ipsilateral or contralateral hemisphere reverse the pain or depressive phenotype. In line with this, a previous study showed that microglia and astrocytes were increased in the ACC of nerve injured mice and microinjections of minocycline into the contralateral ACC partially reversed mechanical allodynia (Cooper et al., 2018). However, the side of nerve injury (i.e., left vs. right) plays a big role in whether functional pain responses are altered by brain region and hemispheric manipulations. For instance, inactivation of the right or bilateral central amygdala (CeA) attenuates mechanical allodynia and hyperalgesia when SNI is performed on the left side of the body, while inactivation of the left CeA has no effect. The same paper also showed that following right-sided SNI, mechanical allodynia was attenuated only by inactivation of the left CeA, while mechanical hyperalgesia was reduced by left, right and bilateral inactivation of the CeA (Cooper et al., 2018). There is also evidence showing that overproduction of interleukin-1β, a cytokine that activates microglia is a common mechanism underlying the generation of neuropathic pain, memory deficits, and depressive-like behavior in mice (Gui et al., 2016). Thus, a potential strategy may be to target upstream activators of microglia, rather than focus on direct glial inhibition.

Overall, the most important aspect of the current study was the demonstration that SNI induced depressive- and anxiety-like behavior differently in male and female mice. However, this study encourages further research on comorbid pain and depression using both sexes as there are clear behavioral sex differences and understanding the sex-specific mechanisms should be further explored.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

All procedures were performed in accordance with the guidelines of the Canadian Council on Animal Care and approved by the University of Toronto Animal Care Committee.

LM and VM conceived and designed the experiments and wrote the article. VM and CC performed the experiments. LM supervised the acquisition of results. VM, NL, and LM analyzed the data. All authors edited and commented on the final version of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.