Sex-specific immune mechanisms in PTSD symptomatology and risk: a translational overview and perspectives

Abstract

Altered immune function in patients with posttraumatic stress disorder (PTSD) may play a role in the disorder pathophysiology and onset. Women are more likely to develop PTSD, suggesting potential sex-specific inflammatory mechanisms underlying the dichotomous prevalence and risk of PTSD in men and women. In this review we examine the available literature to better assess the state of knowledge in the field. In humans, increased systemic inflammation is found in both men and women with PTSD, but seems to be at a greater extend in women. Despite the existence of few clinical studies taking account of sex as a factor in the observed immune changes in PTSD, challenges in the study of sex-specific immune function in humans include: controlling for confounding variates such as the type of trauma and the ethnicity; and limited methodologies available to study central nervous system (CNS)-relevant changes. Thus, preclinical studies are a valuable tool to provide us with key insights on sex-specific peripheral and CNS immune mechanisms underlying PTSD. Available preclinical studies reported increased systemic and CNS inflammation, as well as elevated trafficking of monocytes from the periphery to the brain in both male and female rodents. To date, psychological trauma-induced inflammation is more robust in female vs male rodents. Limitations of preclinical studies include animal models hardly applicable to female rodents, and hormonal changes across estrus phases that may affect immune function. The present review: (1) highlights the key findings from both human and animal studies, (2) provides guidance to address limitations; and (3) discuss the gap of knowledge on a more complex intertwined interaction between the brain, neurovascular, and systemic units.

1. Introduction

Posttraumatic stress disorder (PTSD) is a complex psychiatric disorder that can develop after experiencing severe psychological trauma. PTSD symptomatology includes intrusive thoughts (i.e., flashbacks, reoccurring memories or dreams related to the traumatic event), avoidance and intense fear behaviors, mood and cognitive impairments (i.e., sleep disturbances, irritability, anhedonia, and social isolation), as well as hyperarousal (Brunello et al., 2001). PTSD is associated with comorbidities such as major depressive and anxiety disorders and is also linked to increased risk for chronic illness, accelerated aging and premature mortality. Globally, the prevalence of PTSD is 6–9%, with rates up to 20% in the military personnel and first responders (Koenen et al., 2017). The current available pharmacotherapies (e.g., antidepressants) have limited efficacy and high rate of relapse and dropout. Thus, research groups in the field aim to better understand the mechanisms of higher susceptibility to psychological trauma and increased risk for PTSD, with the ultimate goal of identifying biomarkers of diagnosis and risk, and developing novel therapeutic alternatives.

The hypothalamo-pituitary-adrenal (HPA) axis plays an important role in the systemic and central nervous system (CNS) homeostasis in response to stress. In patients with PTSD, altered sensitivity to glucocorticoids is observed via elevated corticotrophin releasing factor (CRF) and CRF1 receptor brain levels (Mendoza et al., 2016). The HPA axis also contributes to the response to stress by moderating the peripheral and CNS inflammatory reactions. In the last decade, the contribution of immune function to the development of PTSD has aroused growing interest. Clinical evidence shows that PTSD is associated with altered levels of inflammatory regulators. Specifically, higher plasma levels of the pro-inflammatory cytokines interleukin-6 (IL-6), interleukin-2 (IL-2), interleukin-1β (IL-1β), interferon-γ (IFN-γ), tumor necrosis factor-α (TNF-α), and C-reactive protein (CRP), as well as elevated white blood cell (WBC) counts have been repeatedly found in patients with PTSD compared to healthy individuals (Passos et al., 2015) Still, there are discrepancies across studies in regards to the specific markers that are increased. Another group of investigations suggest these increased levels of inflammatory markers may not be only a consequence of PTSD but may also play a causal role (Hori and Kim, 2019). Indeed, peripheral or CNS inflammation can contribute to the onset of PTSD symptomatology by inducing brain changes at morphological, functional and cognitive levels (Hori and Kim, 2019). Microglia cells (i.e., the primary immune cells of the CNS) are involved in the PTSD-associated neuroinflammation via persistent activation, as well as overproduction of proinflammatory cytokines in individuals with PTSD (Enomoto and Kato, 2021). Moreover, PTSD-related neuroinflammation also contributes to an increased risk of dementia, cognitive dysfunction, cardiovascular diseases, diabetes, atherosclerosis and autoimmune diseases (Levine et al., 2014; Hori and Kim, 2019). Yet, the observed alterations in immune biomarkers relevant to PTSD are not consistent between studies. These discrepancies may be partially explained by confounding factors that are modulated by biological sex in the samples of these studies (Yang and Jiang, 2020).

Such sex-specific mechanisms underlying PTSD symptomatology and risk would be consistent with relevant sex differences regarding response to stress. Indeed, women are twice more likely to develop PTSD than men, female sex being one of the main risk factors for developing chronic PTSD after experiencing psychological trauma. Higher rates of drug abuse and high-risk sexual behaviors are also found in women with a history of past trauma vs men (Stevens et al., 2003). Moreover, empirical evidence suggests that women have a higher perception of threat and loss of control when being exposed to stress (Olff et al., 2007). In regards to the experience of traumatic event itself, the incidence of traumatic events is higher in men, but women are more likely to experience sexual assault and sexual abuse during childhood, which could contribute to the elevated risk for the development of PTSD symptoms (Tolin and Foa, 2006). Still, studies focusing on a specific type of trauma and examining PTSD symptomatology reported that women have a greater tendency to meet criteria for PTSD and to live with more severe symptoms compared to men. These findings suggest that there might be sex-specific biological mechanisms underlying the differences in risk of PTSD in men vs women.

These higher rates of PTSD in women may be mediated by a sex-dependent HPA axis response to trauma (Rohleder et al., 2001). Indeed, activation of the HPA axis and release of pro-inflammatory cytokine following psychosocial stress-induced production of corticosteroids were modulated by hormonal changes between men vs women (Mendoza et al., 2016). Elevated levels of progesterone were also associated with enhanced recall of negative memories (Felmingham et al., 2012), and circulating levels of 17-β estradiol, another female sex hormone, was associated with deficient fear extinction (Rohleder et al., 2001).

Recently, the differences between males and females regarding the inflammatory response to trauma has aroused growing interest as a promising avenue to understand the sex-specific mechanisms in PTSD. Given the small number of studies on the topic, as well as the heterogeneity in the examined inflammatory markers examined, meta-analyses are premature at this stage. Hence, the aim of this review is to examine the current state of literature about the sex-dependent inflammatory markers in PTSD, and to better assess the biological mechanisms underlying the higher prevalence of PTSD in women vs men. To find the current state of clinical literature on sex-dependent immune changes in PTSD, a search using MEDLINE PUBMED database was conducted. For human studies, key words “PTSD” and “inflammation” were used. Exclusion criteria were the absence of a PTSD symptom scale to assess a diagnosis and no consideration of sex effect when both men and women were included. Overall, 17 clinical studies were included in this review. For preclinical studies, key words “PTSD”, “inflammation”, “animals”, “blood-brain barrier” were used in our search strategy. Only preclinical investigations using a validated animal model of PTSD (Deslauriers et al., 2018) were included in this review for a total of 16 articles. Preclinical studies using a validated rodent model but using pharmacological and/or genetic approaches with a validated animal model are referred to in the discussion to highlight potential underlying mechanisms. We will also discuss perspectives and provide guidance for further investigations.

2. Sex-specific immune changes in PTSD: a clinical overview

Numerous studies reported altered immune status in individuals with PTSD, but few of them has explicitly taken account of biological sex in their experimental design, as some have included only men or women while others included both, adding sex as a covariate in the analyses (Table 1).

Table 1.

Sex-specific immune changes associated with PTSD in humans

Reference	n total (n with PTSD)	Ethnicity	Country	Type of trauma	Men	Women
Lindqvist et al., 2017	61 men (31)	NS	U.S.	War	↑ circulating IL-6 and hsCRP in PTSD; No changes in circulating TNF-α	NA
Bruenig et al., 2017	299 men (159)	NS	U.S.	War	↑ circulating TNF-α in PTSD	NA
Spivak et al., 1997	38 men (12)	NS	Israel	War	↑ circulating IL-1β in PTSD; No changes in circulating IL-2R	NA
Miller et al., 2017a	32 men (16)	NS	U.S.	War	↑ circulating hsCRP in PTSD	NA
Solomon et al., 2017	116 men (101 are prisoner of war, with or without PTSD	NS	Israel	War (captivity)	↑ circulating CRP and metabolic syndrome in PTSD	NA
Lindqvist et al., 2014	102 men (51)	52.9% Hispanic	U.S.	War	↑ circulating TNF-α and IFN-γ in PTSD; No changes in circulating CRP, IL-1β, IL-6, and IL-10	NA
Eraly et al., 2015	2555 men (quantitative score for PTSD)	76.5% non-Hispanic	U.S.	War	PTSD diagnosis after combat associated with ↑ CRP circulating levels pre-deployment	NA
Imai et al., 2018	105 women (40)	Asian	Japan	Violence for most of the participants	NA	↑ circulating IL-6 in PTSD; No changes in circulating TNF-α, CRP, and IL-1β
Sumner et al., 2017	524 women (174)	96.8% Caucasian	U.S.	NS	NA	PTSD associated with overall ↑ circulating TNFRII, hsCRP, and ICAM-1 at the first visit; ↑ VCAM-1 across time
Renner et al., 2022	53 women (17)	NS	Germany	NS	NA	PTSD diagnosis associated with ↑ circulating IL-6 (vs controls and individuals with MDD) No changes in circulating IL-10
Heath et al., 2013	55 women (7 with probable PTSD)	83.5% AA	U.S.	Violence	NA	↑ circulating CRP in PTSD
Powers et al., 2019	55 women* (18) *with type II diabetes	AA	U.S.	NS	NA	↑ circulating CRP in PTSD
Pace et al., 2012	36 women (12)	NS	U.S.	Childhood abuse	NA	Severity of PTSD symptomatology positively correlated with NF-κB activity in PBMCs
Spitzer et al., 2010	1464 men (16) 1585 women (39)	NS	Germany	NS	PTSD associated with ↑ circulating CRP Female sex associated with overall ↑ CRP
Miller et al., 2017b	253 men (142) 23 women (21)	72.8% Caucasian	U.S.	War	Severity of PTSD symptomatology positively correlated with circulating CRP	Women with PTSD have ↑ circulating CRP vs men with PTSD
de Oliveira et al., 2017	16 men (11) 66 women (33)	63% Caucasian	Brazil	NS	↑ circulating IL-6 and IL-10 in PTSD No effects of sex
Sumner et al., 2018	350 women (175)	96.8% Caucasian	U.S.	NS	NA	PTSD associated with ↑ circulating TNFRII and ICAM-1 at the first visit; ↑ VCAM-1 across time; No changes in circulating CRP

AA, American-African; NA, non-applicable; NS, not specified

CRP, C-reactive protein; hsCRP, high-sensitivity C-reactive protein; ICAM-1, InterCellular Adhesion Molecule-1; IFN-γ, interferon-γ; IL-1β, interleukin-1β; IL-10, interleukin-10; IL-2R, interleukin-2 receptor; IL-6, interleukin-6; MDD, major depressive disorder; NF-κB, nuclear factor-κB; PBMCs, peripheral blood mononuclear cells; TNF-α, tumor necrosis factor-α; TNFRII, Tumor necrosis factor receptor II; VCAM-1, Vascular Cell Adhesion Molecule-1

Three large studies that have included only men with PTSD were performed in war veterans and reported inconsistent results. Male veterans diagnosed with PTSD displayed higher circulating levels of the pro-inflammatory cytokines IL-6 and CRP compared to veterans without PTSD, whereas no differences were found in levels of the pro-inflammatory modulators TNF-α or IFN-γ, and anti-inflammatory cytokine IL-10 (Lindqvist et al., 2017). In contrast to these studies, higher circulating levels of TNF-α were found in male veterans with PTSD vs those without PTSD (Bruenig et al., 2017). In line with these two human studies, other investigations including smaller sample sizes have consistently showed elevated expression of inflammatory markers in men with PTSD. Spivak and colleagues compared circulating levels of IL-1β and IL-2 receptors (IL-2R) in veterans with PTSD to veterans without PTSD. Male veterans diagnosed with PTSD exhibited higher levels of IL-1β, but no changes in IL-2R levels (Spivak et al., 1997). Moreover, two supplemental investigations reported higher circulating levels of high-sensitivity CRP (hsCRP) or CRP in male war veterans with PTSD vs non-PTSD veterans (Spivak et al., 1997; Miller et al., 2017; Solomon et al., 2017). Interestingly, in an extensive study including only male war veterans (n=2555), higher circulating concentration of CRP before deployment was associated with PTSD diagnosis after return from combat (Eraly et al., 2014), suggesting that immune function may be not only related to the symptomatology per se, but may be a pre-disposing causal factor for PTSD risk following severe psychological trauma.

Few studies examined the association between systemic inflammation and PTSD patients in samples that included only women. Higher levels of IL-6, but not IL-1β, TNF-α and CRP were found in women diagnosed with PTSD following violence-related trauma vs controls (Imai et al., 2018). In two studies focusing on women’s health, unchanged or higher CRP circulating levels were associated with PTSD diagnosis (Sumner et al., 2017; 2018). Another investigation in women found no association between circulating levels of IL-6 and PTSD diagnosis, but reported increased circulating levels of the anti-inflammatory cytokine IL-10 in women with PTSD vs those without PTSD (Renner et al., 2022). In trauma-exposed African-American women, elevated circulating levels of CRP have been reported in two distinct studies (Heath et al., 2013; Powers et al., 2019). Interestingly, PTSD symptomatology severity in childhood abuse-exposed women correlated with increased activity of the inflammatory modulator nuclear factor-κB (NF-κB) in peripheral blood mononuclear cells. This increased activity of NF-κB was linked to reduced sensitivity of monocyte to glucocorticoids (Pace et al., 2012). Altogether these findings may suggest that immune alterations in both men and women with PTSD correlate with symptomatology severity. Still, the literature remains scarce, and further investigations are needed to establish a strong relationship.

In regard to studies examining the sex-dependent immune alterations in PTSD, the available literature remains scarce. First, an extensive study including 1464 men and 1585 women reported higher CRP circulating levels in women with PTSD vs men with PTSD (Spitzer et al., 2010). In line with these findings, Miller and colleagues observed higher circulating levels of CRP levels in 21 female veterans with PTSD compared to 142 male veterans with PTSD (Miller et al., 2018). In contrast to these two studies, no differences in circulating levels of IL-6 and IL-10 were observed in 33 men vs 8 women with PTSD (De Oliveira et al., 2018). Of note, these three studies (Spitzer et al., 2010; De Oliveira et al., 2018; Miller et al., 2018) did not have a primary focus on sex-dependent effects, and their observations were part of secondary analyses, which could explain the gap in the number of men vs women included. Thereby, there is an important lack, not to say the absence of literature examining levels of immune markers between men and women living with PTSD as a primary objective.

Clinical neuroimaging studies and postmortem studies are also useful approaches to examine sex differences in immune mechanisms in PTSD, as these procedures enable us to gather additional information that is not available in blood or cerebrospinal fluid. In postmortem brain tissues from female individuals, PTSD was associated with lower expression of the microglial marker 8-kDa translocator protein (TSPO), in BA11, but not BA25 or frontal, cortical region. This difference was not observed in male samples (Bhatt et al., 2020). Another study also used positron emission tomography to evaluate the levels of monoamine oxidase (MAO)-B, an index of astrocyte levels in women with or without PTSD. They found that PTSD was associated with loss of astrocytes in corticolimbic regions (Gill et al., 2022). Still, postmortem and neuroimaging studies evaluating sex dimorphisms in inflammation markers of PTSD risk and symptomatology are further needed.

In regards to sex differences in response to available treatments for PTSD, Danzi and Greca showed that gender was not a clear moderator for PTSD treatment with cognitive-behavioral therapy (Danzi and La Greca, 2021). In line with these results, Connor and al. showed that sex was not a predictor of response to PTSD treatment with the potential antidepressant brofaromine (Connor et al., 2001). These findings suggest that sex may not predict or regulate treatment response, but further studies are essential to validate these effects. Moreover, further investigations should address questions on the pathways (e.g., neuroimmune mechanisms) underlying the effects of biological sex on response to treatment.

With limited approaches available to study neuroinflammation in association with PTSD in humans, recent studies focused on the blood-brain barrier (BBB) permeability and integrity to provide key insights on if and how the BBB interact with the contribution of systemic-CNS inflammation to PTSD risk and symptomatology. Indeed, the BBB is the primary interface between the periphery and the CNS, and its function and integrity are modulated by pericytes and endothelial tight junctions (Abbott et al., 2010). The BBB may therefore interact with the intertwined peripheral-CNS modulation of immune response to trauma. Sumner and colleagues assessed circulating levels of the endothelial markers of BBB function ICAM-1 (i.e., InterCellular Adhesion Molecule-1) and VCAM-1 (i.e., Vascular Cell Adhesion Molecule-1) in women with PTSD. They reported increased circulating levels of VCAM-1 across time in women with PTSD. In regard to circulating levels of ICAM-1, the first study showed higher levels at the first visit in women with PTSD, while the other investigation showed higher overall levels of ICAM-1 (i.e., independently of time) in PTSD group (Sumner et al., 2017; 2018). Still, no studies explored the sex-specific differences in markers of BBB function and integrity. Moreover, limitations in regard to the study of neuroinflammatory and neurovascular mechanisms underlying PTSD pathophysiology in humans make preclinical studies a useful tool and current valuable source of information that could provide insights into the neuroimmune pathways involved in sex-specific differences in PTSD.

3. Sex-specific inflammatory mechanisms in PTSD: key insights from preclinical studies

3.1. Neuroinflammatory pathways in animal models of PTSD

Animal models are essential to gain a better understanding of the peripheral-CNS interactions underlying the development of PTSD, and to provide key insights on potential prophylactic and/or therapeutic alternatives for PTSD (Deslauriers et al., 2018). Translational research requires methods that: mimic the underlying causal mechanisms of the disorder (construct validity); reproduce the behavioral and biological phenotypes associated with the condition (face validity); and show a response to pharmacological treatment that is similar to human patients (predictive validity). Current validated animal models of PTSD mostly rely on behavioral/biological changes (face validity) and response to medication (predictive validity) (see (Deslauriers et al., 2018; Bienvenu et al., 2021) for reviews comparing currently available animal models of PTSD)”. In the last decade, a large number of animal models of PTSD supported the hypothesis of increased neuroinflammation following exposure to psychological trauma (Table 2). Most of these studies focused on immune changes found in the hippocampus, PFC (prefrontal cortex), and amygdala, all relevant to PTSD pathophysiology (Deslauriers et al., 2018).

Table 2.

Sex-specific immune changes in animal models of PTSD

Model	Species	Males		Females		References
		Periphery	CNS	Periphery	CNS
Predator stress/scent	Mouse	↑ Ang II and IL-61	Acute: ↑ TNF and IL-1 in hypothalamus, HP, and midbrain (adult) 2; ↑ IL-1β and IL-6 in HP (1 wk)3 Chronic (28 days): TNF in midbrain (adolescent); impaired immune response (IL-1 and TNF) to LPS in midbrain and HP (2 wks)2	NA	↑ IL-1β and IL-6 in HP (1 wk)3 ↑ NALP3 in HP and amygdala of stressed female vs male mice3	1Xue et al., 2022; 2Barnum et al., 2012; 3Ghosh et al., 2021
Underwater trauma	Mouse	NA	No changes in NALP3 in HP and amygdala		↑ NALP3 in HP and amygdala	Ghosh et al., 2021
Repeated predator + psychological stress (daily cage change)	Mouse	NA	↑ IL-1β and NALP3 in PFC, HP and amygdala (3 wks)	NA	NA	Wilson et al., 2013
Fear conditioning	Mouse	↑ IFN-γ and IL-6 (0 and 10 min after re-exposure to the CS; normalized after 60 min)	NA	NA	NA	Young et al., 2018
Single prolonged stress	Rats	↑ TNF-α and IL-1β1,4	↑ IL-1β and IL-6 in HP (2 wks)2; ↑ TNF-α and IL-1β in HP3; ↑ IL-1β and IFN-γ in mPFC and HP4; ↑ IFN-γ in amygdala4; ↑ IL-6, TNF-α, PGE2, NO in the HP5	NA	NA	1Yamanashi et al., 2020; 2Peng et al, 2013; 3Lee et al., 2016; 4S.C. Wang et al., 2018; 5M. Wang et al., 2018
Repeated social defeat	Mouse	↑ production of Ly-6C monocytes1; ↑ IL-6 and IL-17A in T-lymphocytes2; ↑ IL-2, IL-6, TNF-α, IL-17A, IL-22, MCP-1, CXCL2, IL-103	Recruitment of monocytes from the spleen to the brain (8 days)4; ↑ IL-1β and TNF-α in microglia (8 days)4	NA	NA	1Wohleb et al., 2013; 2Moshfegh et al., 2019; 3Elkhatib et al., 2020; 4Wohleb et al., 2014
Resident-intruder paradigm	Rats	↑ hypertensive response to Ang II1	↑ ACE, AT1R, TNF-α, IL-6, and IL-1β, and CD11b in the lamina terminalis1 ↑ Iba1 and IL-1β in the vHP of rats that received microbiota from rats vulnerable to RDS2	NA	NA	1Xue et al., 2019; 2Pearson-Leary et al., 2020
Social isolation	Mouse	NA	↓ PPAR-α in HP of mice that developed aggressive behavior; ↑ TLR-4, NF-κB, TNF-α and MCP-1 in the HP of aggressive mice	NA	NA	Matrisciano et al., 2021

NA, non-applicable

ACE, angiotensin-converting enzyme; AT1R, angiotensin type 1 receptor; Ang II, angiotensin II; CNS, central nervous system; CS, conditioned stimulus; CXCL2, chemokine (C-X-C motif) ligand 2; HP, hippocampus; Iba1, ionized calcium-binding adapter molecule 1; IFN-ɣ, interferon-ɣ; IL-1, interleukin-1; IL-1β, interleukin-1β; IL-10, interleukin-10; IL-17A, interleukin-17A; IL-2, interleukin-2; IL-22, interleukin-22; IL-6, interleukin-6; MCP-1, monocyte chemoattractant protein-1; mPFC, medial prefrontal cortex; NALP3, NACHT, LRR and PYD domains-containing protein 3; NF-κB, nuclear factor-κB; NO, nitric oxide; PFC, prefrontal cortex; PGE2, prostaglandin E2; PPAR-α, peroxisome profilerator-activated receptor-⍺; TLR4, toll-like receptor 4; TNF-⍺, tumor necrosis factor-⍺; vHP, ventral hippocampus

The well-known single prolonged stress (SPS) model, that consists of the acute sequential exposure to restraint, forced swim and ether-induced loss of consciousness (Lisieski et al., 2018), has been widely used to study stress-induced neuroinflammation relevant to PTSD in male rodents. Lee et al. (2016) found increased hippocampal expression of IL-1β and TNF-α in stressed males. These inflammatory changes, along with SPS-induced anxiety-like behavior were reduced by a 14-day long pre-treatment with the non-selective cyclooxygenase (COX) inhibitor ibuprofen, supporting the association between neuroinflammation and the development of anxiety-like phenotype following trauma. In addition to increased hippocampal levels of IL-6 and TNF-α, Wang et al. (2018) reported increased hippocampal levels of the inflammatory mediators prostaglandin E2 (PGE2) and nitric oxide (NO) following SPS. The pharmacological inhibition of COX-2, an enzyme implicated in prostaglandin synthesis and cell injury, alleviated SPS-induced anxiety-like response and cognitive deficits. The protective effects of COX-2 inhibition on stress-induced behavioral changes correlated with reduced neuronal apoptosis and protein levels of TNFα, IL-6, PGE2 and NO in the hippocampus. In a slightly modified SPS protocol that includes an inescapable foot shock in addition to the other stressors, anxiety-like behavior was associated with increased hippocampal protein levels of IL-1β and IL-6 in male rats 2 weeks following exposure (Peng et al., 2013). S.C. Wang et al. (2018) investigated SPS-induced neuroinflammation in the hippocampus, but also in the medial prefrontal cortex (mPFC) and amygdala, two other crucial brain regions associated with the pathophysiology of PTSD. Male SPS rats displayed increased hippocampal, amygdala, and mPFC levels of IL-1β and IFN-γ compared to non-stressed rats. Interestingly, intranasal injection of oxytocin, a neuropeptide involved in social behavior and sexual reproduction reduced SPS-induced elevation of hippocampal IL-1β and IFN-γ, and enhanced fear extinction, whereas no beneficial effect of oxytocin was found in the amygdala (Wang, S. C. et al., 2018). These findings are of great interest for the field and are in line with a clinical study reporting an association between intranasal oxytocin and reduced PTSD symptoms (Frijling, 2017). Intriguingly, the same oxytocin regimen increased hippocampal and mPFC levels of IFN-γ in non-stressed rats (Wang, S. C. et al., 2018). Altogether these findings suggest that administration of oxytocin may be a promising alternative therapeutic for PTSD patients via the reduction of neuroinflammation and facilitation of fear extinction.

Another widely used animal model called repeated social defeat (RSD) consists of exposing a resident (experimental) C57BL/6 mouse to a bigger aggressive CD-1 mouse for multiple days. This robust and highly recognized model of depression also induces anxiety symptoms as well as biological and behavioral phenotypes relevant to PTSD, such as changes in amygdala activity, peripheral/CNS inflammation, hyperarousal, and anhedonia (Golden et al., 2011; Deslauriers et al., 2018). In male mice, anxiety-like behavior is associated with increased recruitment of bone marrow-derived macrophages in the brain up to 8 days following the last RSD procedure. Increased expression of IL-1β and TNF-α was also found in enriched CD11b+ (i.e., microglia) brain cells of RSD-exposed mice (Wohleb, Mckim, et al., 2014), suggesting a crucial role of microglia activation and related production of cytokines in response to RSD. As the contribution of the gut-brain axis to higher sensitivity to stress has emerged (Belkaid and Hand, 2014), the interaction of microbiota with response to stress has been studied. Interestingly, rats receiving microbiota from RSD-susceptible (i.e., developing anxiety-like behavior) rats exhibited enhanced microglial activation, and higher IL-1β expression in the ventral hippocampus (Pearson-Leary et al., 2020). Importantly, the same animals displayed higher depression-like phenotype in the forced-swim test, but no changes in stress-associative anxiety-like behavior (i.e., social interaction test). These findings suggest that the experience of psychological stress or trauma itself is a prerequisite to the development of anxiety-like behavior. As the role of the microbiota in immune function has been well established (Belkaid and Hand, 2014), this study provides novel insights on the intertwined interaction between microbiota, inflammation and animal behavior relevant to PTSD.

Social isolation (i.e., multiple weeks of single housing) is a well-known chronic model that induces depressive- and anxiety-like behaviors, as well as social deficits. In addition to these latter phenotypes, this model reproduces few core aspects of PTSD symptomatology including fear extinction deficits, increased fear response, aggressive behavior and hypofunction of the hypothalamic-pituary-adrenal (HPA) axis (Aspesi and Pinna, 2019; Pinna, 2019). Social isolation can be perceived as a chronic stress inducing state, which leads to maladaptive stress response, and is therefore a suitable model to study PTSD vulnerability (Pinna, 2019). Male socially isolated mice developing aggressive behavior exhibited reduced hippocampal expression of peroxisome profilerator-activated receptor-α (PPAR-α). PPAR-α, a ligand activated nuclear receptor mostly associated with homeostasis and metabolism, is also known to regulate inflammation via indirect inhibition of toll-like receptor 4 (TLR4)/κB signaling (Matrisciano and Pinna, (2021). Importantly, DNA methylation of the PPAR-α gene was increased in socially isolated mice, along with elevated hippocampal expression of immune markers TLR-4, NF-κB, TNF-α and monocyte chemoattractant protein-1 (MCP-1), all involved in the PPAR-α anti-inflammatory pathway (Matrisciano and Pinna, 2021). This study provides evidence of an epigenetic regulation of the PPAR-α gene in the brain following stress, leading to increased neuroinflammation in association with higher susceptibility to stress measured by aggressive behavior.

The predator stress (PS) model (i.e., protected or unprotected exposure to a predator, or exposure to a predator’s scent) demonstrates ethological validity through its unpredictability, and generates biological and behavioral changes relevant to PTSD pathophysiology (Deslauriers et al., 2018). In male mice, the combination of repeated protected PS with daily psychological stress (i.e., cage change) leads to a persistent (i.e., 3 wks after the last exposure) increase in IL-1β and NALP3 inflammasome in the hippocampus, PFC, and amygdala (Wilson et al., 2013).. Importantly, Barnum and colleagues studied the effect of acute vs chronic exposure to PS in both adolescent and adult male mice. They first reported that acute PS exposure during adulthood leads to increased neuroinflammation in the following hours via the overexpression of tumor necrosis factor (TNF) and IL-1 genes in the hypothalamus, hippocampus, and midbrain, whereas no or little effect was found in the PFC and the spleen (Barnum et al., 2012). In regard to age-dependent effects, chronic PS during adolescence increased mRNA expression of TNF in the midbrain. Of note, the same study has also investigated the effect of acute immune challenge in PS-exposed or unexposed mice. Following immune challenge, overexpressed IL-1 and TNF genes in the midbrain and hippocampus of only non-stressed, not chronically stressed mice were observed (Barnum et al., 2012). Altogether these findings suggest that: (1) chronic stress induces subchronic neuroinflammatory state in male mice; and (2) and impairs immune adaptive response to a new, acute subsequent immunological challenge, that could affect the immune system’s ability to clear pathogenic agents.

Very few preclinical studies compared the sex-dependent neuroimmune responses following stress. In the PS model, anxiety-like behavior has been associated with increased hippocampal levels of IL-1β and IL-6 in both male and female mice 1 week following exposure vs control mice (Ghosh et al., 2021). Interestingly, for all behavioral assessments and immune parameters, stressed females displayed higher anxiety-like behavior and inflammation vs the corresponding male group. The same group also used another PTSD-like model, namely the underwater trauma model (i.e. physical restraint with brief underwater immersion), NLRP3 inflammasome and cleaved Caspase 1, which both play a role in the maturation of IL-1β were upregulated in the hippocampus and amygdala of stressed females vs male groups (Ghosh et al., 2021). Higher activation of the upstream regulator of NLRP3 Bruton’s tyrosine kinase (BTK) in the same brain regions of stressed females vs non-stressed females and male groups (Ghosh et al., 2021). These results provide us with valuable insights on the inflammation as a potential mechanism underlying sexual dimorphism observed in PTSD (i.e., higher prevalence in women). Of note, pharmacological inhibition of NLRP3 or BTK significantly reduced the inflammatory and behavioural responses to physical stress similarly in both males and females, further establishing the link between inflammation and anxiogenesis (Ghosh et al., 2021). Yet. further investigations are needed to thoroughly elucidate the sex-dependent neuroimmune mechanisms of enduring response to psychological trauma. The influence of gonadal hormones on PTSD risk has aroused growing interest in the field, and remains to be further investigated. Nevertheless, interesting preclinical data have shown that acute treatment of ovariectomized (OVX) female rats with estradiol (E₂) facilitated fear extinction (Graham and Daher, 2016). Conversely, another group found that regardless of the estrous cycle phase, PS lead to the development of anxiety-like behavior (elevated plus maze, startle response, open field) in female rats. Independently of stress exposure, OVX rats displayed greater startle response and anxiety-like behavior in the open field test vs. sham control rats (Zoladz et al., 2019). These results indicate that while ovarian hormones seem to influence anxiety states, they do not prevent the development of stress-induced PTSD-like symptomatology.

An increasing number of studies further investigated the contribution of the BBB to psychological trauma response and PTSD symptoms. In the RSD procedure, exposed male mice displayed reduced BBB integrity, as measured by the expression of the tight-junction protein claudin-5 (cld-5). Interestingly, a loss of BBB integrity via cldn-5 downregulation also results in greater sensitivity to subsequent stress (Cathomas et al., 2019). In the SPS procedure, acute (i.e., 24 hrs) cognitive impairment was associated with increased hippocampal BBB leakage, as measured by NaFlu extravasation across the BBB, as well as reduced expression of the tight junction proteins cld-5 and occludin in male rats. This elevated BBB disruption was also related to increased microglial activation in the same brain region (Ni et al., 2022). Interestingly, all the BBB, immune, and behavioral parameters were normalized 7 days after SPS (Ni et al., 2022), suggesting an acute transient loss in BBB integrity and memory function. Pharmacological microglial inhibition prior to SPS preserved working memory and hippocampal BBB integrity (Ni et al., 2022), suggesting a direct and causal role of microglia activation in BBB disruption and memory impairment following chronic psychological stress. Furthermore, IL-1 receptor type-1 (IL-1R1) knockout mice and mice deficient in key chemokine receptors (i.e., chemokine receptor-2 (CCR2) or fractalkine receptor (CX3CR1)) did not show recruitment of circulating monocytes to the brain after RSD, in parallel with a lack of anxiety-like behavior, as measured in the open field and light-dark box tests (Wohleb et al., 2013; Wohleb, Mckim, et al., 2014; Wohleb, Patterson, et al., 2014). These findings suggest that the recruitment of peripheral inflammatory monocytes to the brain across the BBB is a necessary step of the stress-primed immune response leading to the development of anxiety-like behavior. More importantly, BBB disruption after stress strongly suggest a contribution of peripheral inflammation in stress-induced neuroinflammatory and behavioral changes.

Sex-specific neuroinflammatory response to stress is a hot topic in the field of PTSD research, especially in regards to the role of microglia in behavioral phenotypes. Few preclinical studies comparing this role in both female and male rodents have yielded interesting results. Acute and chronic restraint stress induced greater heterogeneity in microglial morphological state across the corticolimbic circuitry (i.e. orbitofrontal cortex, PFC, dorsal hippocampus, basolateral amygdala) in males, whereas chronic stress induced greater morphological coupling (i.e. less heterogeneity) in females (Bollinger et al., 2017). Furthermore, an inescapable tail shock protocol potentiated microglial IL-6 and IL-1β mRNA expression following LPS challenge in males, but not females, suggesting that neuroinflammatory response in females may not depend on microglial reactivity (Fonken et al., 2018). Still, locus coeruleus (LC)-specific depletion of microglia in female rats reduced burying behavior (i.e., hypervigilant behavior) and prevented the increase of IL-1β protein expression in the LC following witness stress (e.g. witnessing social defeat stress) (Pate et al., 2023). In mice, DREADD technology has been used to generate aggressiveness in male mice toward females and to mimic social defeat stress in female rodents (Takahashi et al., 2017). In addition to anxiety-like behavior and social avoidance, female mice submitted to this protocol exhibited elevated release of IL-6 and monocytes in the blood, an accumulation of monocytes into the brain, as well as microglial activation (Yin et al., 2019).

3.4. Systemic inflammation in animal models of PTSD

As peripheral markers are still being investigated in various psychiatric disorders such as depression and panic disorder (Kim et al., 2020), preclinical studies examining peripheral immune response to psychological trauma are critical to provide insights on the pathophysiology of PTSD, and to facilitate the identification of accessible biomarkers of diagnosis and risk (Table 2). In the auditory fear conditioning paradigm consisting of multiple tone-footshock pairings (conditioned stimulus; CS), followed by an extinction phase during which the same, unpaired tone is presented, fear memory is quantified via freezing behaviors. Young and colleagues (2018) have demonstrated, immediately and 10 minutes after CS re-exposure that fear conditioned male mice displayed increased circulating levels of IFN-γ and IL-6, which normalized after an hour. Interestingly, during memory recall 24h after extinction learning, IL-6 knockout and IL-6 antibody-treated mice exhibited minimized freezing behavior vs wildtype mice. These findings suggest that a short, transient increase in peripheral inflammation during fear recall may prevent long-term extinction of fear learning.

In the SPS rodent model, anxiety-like phenotype in males was accompanied with elevated peripheral levels of TNF-α and IL-1β 2 weeks after exposure (Yamanashi et al., 2020). Repeated peripheral injection of the NALP3 inflammasome inhibitor beta-hydroxybutyrate (BHB) reduced SPS-induced anxiety-like behavior and TNF-α sera levels. These findings support the therapeutic potential of anti-inflammatory intervention for PTSD. Another preclinical study reported that administration of oxytocin successfully reduced both CNS (IL-6) and peripheral (TNF-α and IL-6) immune responses to SPS (Wang, S. C. et al., 2018). These findings highlight a distinct regulation of inflammation in the periphery vs the brain in stress response and supports the relevance of studying both these components in the context of PTSD.

In a RSD model in which male rats are exposed to a bigger resident male for 3 consecutive days, stressed rats exhibited upregulated expression of several inflammatory cytokines (TNF-α, IL-6, and IL-1β), and of the CD11b microglial marker in the lamina terminalis, a region partly lacking the standard BBB. Interestingly, increased levels of renin-angiotensin system (RAS) components angiotensin type 1 receptor (AT1R) and angiotensin converting enzyme (ACE) were observed in the lamina terminalis of SDS-exposed rats (Xue et al., 2019). Interestingly, enhanced hypertensive and neuroinflammatory response was observed in stressed vs control mice following the administration of angiotensin II (Ang II), a crucial component of RAS. In line with these findings, PTSD patients are at higher risk of developing hypertension (Edmondson and Von Känel, 2017). Altogether, the upregulated expression of RAS components and cytokines following stress suggests a potential interaction between RAS and immune function in the increased susceptibility to trauma. Moreover, pharmacological inhibition of ACE or TNF-α prevented the hypertensive and inflammatory responses to Ang II in stressed mice, highlighting the accumulative and moderating effect of both RAS and immune function on hypertension following stress. In a recent study, Xue et al. (2022) reported extended hypertensive response to Ang II in association with anxiety-like behavior and increased peripheral inflammation (i.e., Ang II and IL-6) in male rats exposed to PS. Again, pretreatment with ACE or TNF-α inhibitors prevented all stress-induced physiological and behavioral effects, supporting the emerging hypothesis that ANG II-moderated RAS plays a causal role in the enhanced peripheral inflammation anxiety-like behavior following stress.

The exposition of male mice to RSD paradigm increased the presence of the pro-inflammatory Ly6C^Hi subset of peripheral monocytes (Wohleb et al., 2013). T-lymphocytes extracted 2 days following RSD also showed increased production of inflammatory cytokines (i.e., IL-6, IL-17A) (Moshfegh et al., 2019). Interestingly, systemic inflammation has been associated with depression-like, but not anxiety-like behavior, suggesting that RSD-induced immune changes in the CNS or periphery may be specific to depressive-like phenotype, but does not correlate with anxiety-like behavior. In another study, increased plasma levels of numerous inflammatory regulators (e.g., IL-6, TNFα, MCP-1, and IL-10) were found 2 days after RSD in male rodents (Elkhatib et al., 2020). In recent years, evidence has established a link between gut microbiota and immune function as well as multiple psychiatric diseases such as anxiety and depression (Pearson-Leary et al., 2020). Pearson-Leary and collaborators (2020) found that 9 days after receiving microbiota from rats susceptible to RSD, male rats exhibited increased neuroinflammation, but not systemic immune function vs controls or rats receiving microbiota from resilient rats. However, elevated circulating levels of the indirect marker of BBB permeability S100β was detected in rats receiving microbiota from susceptible rats (Pearson-Leary et al., 2017), suggesting that transferring microbiota from vulnerable rats to naïve rats is sufficient to induce BBB disruption. The absence of peripheral inflammation detected 9 days after the transfer of the microbiota coupled with increased S100β suggest that inflammatory molecules may have already passed through the BBB into the CNS, contributing to neuroinflammation and depression-like behavior. Future investigations remain to be conducted to unravel the cascade of events across time.

Though a large number of animal models of PTSD exist to date, several limitations complicate the application of preclinical findings into clinical practice. These include the use of non-ethological stressors which affects the construct validity of the model (e.g., electrical shock, single prolonged stress), paradigms that can hardly be applied to females (e.g., RSD) and a general lack of uniformity in the reported behavioral parameters, thus affecting the consistency of results across research groups (Deslauriers et al., 2018).

4. Discussion

Though there is an overall consensus of increased immune status in PTSD, clinical studies investigating the mechanisms of enduring susceptibility to psychological trauma and PTSD risk face multiple challenges. The sex, type of trauma and ethnicity are confounding factors that may be sources of high variability across studies. There is also an important lack of research with sex-specific immune changes as the primary outcomes. Moreover, approaches to study neuroinflammation and BBB functionality in response to trauma in humans are limited. Preclinical studies are therefore necessary for our understanding of peripheral-CNS interactions in behavioral phenotypes relevant to PTSD. Up to date, the available literature clearly shows that psychological trauma consistently induces systemic inflammation in both humans and animals, and neuroinflammation has been repeatedly shown in animal models. Across several animal models covered in this review, the pro-inflammatory cytokine IL-1β is consistently upregulated in the CNS following trauma (Barnum et al., 2012; Peng et al., 2013; Wilson et al., 2013; Wohleb, Patterson, et al., 2014; Lee et al., 2016; Wang, S. C. et al., 2018; Ghosh et al., 2021). Additional experiments using IL-1β-deficient mice support the premise that IL-1β signaling is necessary for the recruitment of monocytes to the brain and the development of anxiety-like behavior (Wohleb, Patterson, et al., 2014), providing us with key elements on peripheral mechanisms underlying enduring risk for PTSD. Further, pharmacological or genetic inhibition of IL-6 or TNF-α showed protective effects, such as sustained fear extinction, reduced anxiety-like behavior, and prevention of stress-induced systemic and CNS inflammation (Young et al., 2018; Xue et al., 2019; Xue et al., 2022). Other animal studies have highlighted the therapeutic potential of reducing the inflammatory response through knockout models or direct inhibition of key inflammatory regulators such as COX-2 and the NLRP3 inflammasome (Wang, M. et al., 2018; Yamanashi et al., 2020). Inhibiting these regulators preserved the BBB and neuronal integrity, and the stress-induced cognitive deficits and anxiety-like behavior. Alternatively, prophylactic chronic systemic administration of the non-selective COX inhibitor ibuprofen was also sufficient to reduce trauma-induced inflammation and anxiety-like behavior in rats (Lee et al., 2016). Altogether these findings support the contribution of upregulated cytokines to PTSD pathophysiology and symptomatology.

Recently, growing interest has aroused in the role of BBB in the enduring response to psychological trauma and PTSD risk. The BBB represents a crucial mediator between peripheral and CNS inflammation. Microglial activation contributes to BBB injury and cognitive decline following stress (Ni et al., 2022). Peripheral recruitment of circulating monocytes into the CNS is also necessary for the development of anxiety-like behavior following stress (Wohleb et al., 2013; Wohleb, Mckim, et al., 2014; Wohleb, Patterson, et al., 2014). These findings suggest that both peripheral and CNS inflammation may regulate BBB integrity, leading to increased PTSD symptoms and cognitive impairments. Yet, the causal role of both systemic and CNS inflammation, as well as of the BBB as the periphery-CNS interface in the pathophysiology of PTSD need to be further elucidated.

In animal models of PTSD, the inconsistent time points of presented inflammatory and behavioral parameters are an important source of discrepancies between studies. Furthermore, the lack of studies monitoring the stress-induced immune response across time limits our understanding of the pathophysiology of PTSD. Notably, Young et al. (2018) have shown a transient peripheral immune response driven by IL-6 following reminiscence of trauma, which has stabilized after 60 minutes. In another study, stress-induced loss of BBB integrity and cognitive function normalized levels 7 days after stress exposure (Ni et al., 2022). These findings illustrate the adaptable and dynamic aspect of the pathophysiological response to psychological trauma, which further complicates the characterization of the mechanisms implicated in the onset of PTSD. Nevertheless, few studies using similar time points have yielded interesting data. Moshfegh et al. (2019) and Elkhatib et al. (2020) used the RSD paradigm to assess peripheral immune and behavioral responses at consistent time points in male mice. No association between cytokine expression in T-lymphocytes and anxiety-like behavior was found in one study (Moshfegh et al., 2019), whereas the other study showed an association between peripheral inflammation and anxiety-like behavior (Elkhatib et al., 2020). These results suggest that overall systemic inflammation, via immune cells other than T-lymphocytes play a role in the onset of anxiety-like behavior.

Though the higher prevalence of PTSD in women vs men is well established, the literature documenting the sex-specific mechanisms underlying PTSD remains scarce. Human studies addressing the sex differences in PTSD must consider the distinct type of trauma experienced by men and women. Indeed, the type of trauma, and its severity may significantly alter the biological response to trauma and associated PTSD risk. Notably, females are more likely to experience sexual assaults, whereas men are more likely to experience physical assaults, and combat- or work-related accidents (Tolin and Foa, 2006). Sex differences also apply to PTSD symptomatology clusters, with women displaying more re-experiencing and anxiety symptoms (Olff, 2017), and men suffering mainly from irritability and alcohol abuse (Green, 2003). As men being more likely to endorse a combat- or work-related trauma, they are also at higher risk to experience a traumatic brain injury (TBI). The higher risk for PTSD onset following a TBI-associated traumatic event is known, but the impact of TBI on PTSD symptomatology (i.e., categories of symptoms and severity) has not been extensively studied (Vasterling et al., 2018). In a double-hit animal model consisting of chronic variable stress (CVS) combined to the closed head model (CHM) of TBI, Fesharaki-Zadeh and collaborators (2020) reported that CVS alone is sufficient to generate anxiety-like behavior, but only the combination of both CVS and CHM caused cognitive and memory deficits. These impairments were associated with microglia activation in the hippocampus, suggesting that PTSD combined with TBI can lead to distinct inflammatory effects and additional behavioral deficits. Thus, consideration of TBI is needed in further studies investigating sex-specific mechanisms of PTSD symptoms and risk following TBI-related or unrelated trauma.

Preclinical studies also present limitations to study sex-dependent immune effects of psychological trauma. First, RSD paradigm favors the use of male rodents as aggression between males and females, or between female rodents, is rare (Trainor et al., 2017). To solve this issue, the pharmacological activation of a subdivision of the ventromedial hypothalamus using DREADDs technology to generate aggressivity has been developed (Takahashi et al., 2017). With this approach, social defeat generates anxiety-like behavior, social avoidance as well as central and peripheral immune responses in female mice (Yin et al., 2019). Still, sex-dependent mechanisms remain to be compared within the same experimental design. Second, hormonal changes across estrus phases in female rodents may affect their immune status (Deslauriers et al., 2018). As vaginal sampling to control for estrus phase is stressful and may add a confounding bias in models of stress, one solution may be to consider collecting vaginal fluid immediately before the sacrifice. Yet, this solution is still challenging in weeks-long experimental designs, as estrus phases change over 4–5 days in rodents (Deslauriers et al., 2018). Alternatively, gonadectomy is useful to study the role of gonadal hormones on stress response (Pooley et al., 2018). Less invasive procedures such as the administration of selective inhibitors or estrogen receptor modulators may be used in future investigations aiming to study sex differences in PTSD risk.

Recent studies highlighted the modulating effects of stress on microbiota and RAS components. First, the transfer of microbiota from stress-susceptible rats to naïve rats lead to increased CNS, but not peripheral inflammation (Pearson-Leary et al., 2020). Moreover, stress exposure in animals resulted in the upregulation of renin-angiotensin system (RAS) components and in extended hypertensive and inflammatory response to Ang II (Xue et al., 2019; Xue et al., 2022). Pharmacological inhibition of ACE, an essential component of RAS prevented the stress-induced behavioral and biological changes, highlighting the essential role played by the RAS system in stress response. Altogether both clinical and preclinical findings suggest that the association between inflammation and PTSD risk involves a more complex intertwined interaction between the brain, BBB, and systemic units (i.e., immune, gut, and RAS components). The consideration of sex-specific changes is mandatory in future investigations to take account of the overall multifactorial context of trauma response and PTSD risk.

Figure 1. Summary of findings related to sex differences in immune response in both humans and preclinical models relevant to core symptoms of PTSD.

CRP, C-reactive protein; ICAM-1, intercellular adhesion molecule 1; IL-1β, interleukin-1β; IL-6, interleukin-6; IL-10, interleukin-10; NF-κB, nuclear factor kappa B; NLRP3, NOD-like receptor family, pyrin domain containing 3; TNF-α, tumor necrosis factor-α; VCAM-1, vascular cell adhesion molecule 1. The schematic figure was created on Mind the Graph platform (www.mindthegraph.com).