Immunomodulatory T cell death associated gene-8 (TDAG8) receptor in depression-associated behaviors

Abstract

Converging evidence supports neuroimmune factors in depression psychopathology. We previously reported reduced depression-like behavior in immunomodulatory G-protein-coupled receptor, T cell death-associated gene-8 (TDAG8) deficient mice. Here, we expand on those findings by investigating depression- and anxiety-associated behaviors, and cytokine profiles in TDAG8-deficient mice. TDAG8-deficiency reduced depression- and anxiety-associated behaviors in the forced swim test (FST), open-field test and elevated zero maze. Interestingly, cytokine expression, particularly IL-6, was attenuated within hippocampus and spleen in TDAG8-deficient mice following the FST. There were no differences in immune-cell frequencies. Collectively, these data suggest a contributory role of TDAG8 in neuroimmune regulation and depression-associated physiology.

1. Introduction

Major depressive disorder (MDD) is a severe mood disorder affecting an estimated 350 million people worldwide. MDD results in a decreased quality of life, significant morbidity and mortality and is a large socioeconomic burden[1-3]. Despite high prevalence, the pathophysiological mechanisms underlying MDD remain uncertain. Current pharmaceutical treatments, such as monoamine reuptake inhibitors, have a delayed onset of treatment effect and lack efficacy in a large proportion of patients[4-6]. While the role of neurotransmitter abnormalities in depression has been evidenced by clinical and preclinical studies[7,8], the heterogeneity of presentation, course, and treatment responses infer that additional mechanisms may be involved.

Numerous clinical and preclinical studies demonstrate that cytokines play an important role in the pathogenesis of depression[9-12]. Individuals with depression have significantly increased levels of circulating cytokines including interleukin 6 (IL-6), TNF alpha (TNF-α), and IL-1β[13-16] which correlate with the progression and severity of depression[17]. Stress and negative emotion upregulate cytokine concentrations in humans[18]. In rodents, administration of cytokines at levels associated with inflammation or lipopolysaccharide (LPS), which promotes cytokine release, leads to depression-like behaviors[19-21]. Interestingly, treatment with antidepressants reduces plasma cytokines in both humans and rodents[13,22-24]. Despite these findings, the mechanisms by which immune dysfunction affects depressive-like behavior are not fully understood. Identification of novel immune-modulatory targets that can regulate depression-associated behaviors may offer increased mechanistic understanding and lead to novel therapies for depression.

The T cell death-associated gene-8 (TDAG8) receptor is a proton sensing G-protein-coupled receptor predominantly expressed on immune cells in both the CNS and periphery[25-28]. Previous studies from our lab [27] and others [29,30] reported activation of TDAG8 by extracellular acidification suggesting its potential role in CNS and systemic homeostasis[27,29-32]. TDAG8 regulates the release of cytokines in vitro[33,34] and has been reported to facilitate inflammation-related hyperalgesia and allergic airway pathophysiology[33,35,36] suggesting its recruitment in modulating immune function and inflammatory responses. In a previous study, we reported attenuated depression-like behavior in TDAG8-deficient mice suggesting a potential role for TDAG8 in depression-associated pathophysiology[37]. In the current study, we extended our previous findings to investigate the association of TDAG8 receptor with cytokines relevant to depressive pathophysiology at baseline and following a depression-relevant stressor, the forced swim test (FST). Additionally, we replicated our effects on the FST and sucrose preference tests and expanded our examination of behavioral effects of TDAG8 deficiency on anxiety-associated behaviors.

2. Methods

2.1. Animals

TDAG8^−/− mice (Dr. Owen Witte, UCLA) were generated on a BALB/c background as described previously[38]. Briefly, TDAG8 expression was eliminated by replacing a 2.5 kb fragment containing exon 2-derived TDAG8 coding sequences with a 3.4-kb fragment encoding promoter-less internal ribosomal entry site EGFP sequence. All experiments were performed on 2-5 months old male mice bred in-house and carrying wild-type (TDAG8^+/+) or knockout (TDAG8^−/−) allele. Male mice were chosen for consistency with and replication of our previous observations[28,37]. Mice were housed under standard temperature (23–28 °C) and light conditions (14h-light, 10h-dark cycle; on at 0600h) with ad libitum food and water. Behavioral experiments were performed from 8a to 1p. Studies were approved by the Institutional Animal Care and Use Committees of University of Cincinnati in vivariums accredited by the Association for the Assessment and Accreditation of Laboratory Animal Care (AAALAC) which follow the National Institutes of Health guide for the care and use of Laboratory animals.

2.2. Behavioral Studies

Behavioral studies were performed using 4 separate cohorts. Cohort 1 was trained and tested for sucrose preference. One week later they underwent the open field test (OFT), followed by the forced swim test (FST) one week after the OFT. Experiments were performed in this order to minimize any effects of previous stress effects on later behaviors. Cohort 2 was used for cytokine quantification following exposure to the FST. To confirm effects on anxiety-like behaviors, a separate cohort (Cohort 3) underwent the elevated zero maze (EZM). Behavior-naive mice from both genotypes (Cohort 4) were used for tissue collection for flow cytometry. Details below.

2.2.1). Sucrose preference

Sucrose intake and preference were assessed on singly housed mice as previously described[37] with modifications. Mice were allowed to acclimate for 3 days to two water bottles. On the fourth day water was replaced with sucrose solution (1%) in one of the bottles for the next five days. The bottles were reversed daily to control for side preference. Sucrose and water intake were measured daily. Sucrose preference was calculated based on the ratio of sucrose solution consumed and total liquid consumed in a 24 h period and averaged across all days. Consumption was normalized to body weight.

2.2.2). Forced swim test (FST)

The FST was performed as previously described[37]. Briefly, mice were placed into a beaker (30 cm height × 20 cm diameter) filled approximately 20 cm with water (23-25 °C) for 6 min and behavior was videotaped. Immobility (absence of all movements except those required to maintain floating) was quantified by a trained observer blind to genotype[39].

2.2.3). Open Field Test (OFT)

Mice were exposed to a dimly lit open field apparatus that consisted of a white 50 × 50 × 22cm Plexiglas box as previously described[40]. Each mouse was placed in a corner facing the center and allowed to freely explore for 5 min. The test was recorded from a camera mounted above the OFT. Videos were scored for time spent in the center areas (defined as the middle 50% of the total arena or an approximately 35 × 35 cm square) and distance traveled using CleverSys TopScan (Reston, VA) software.

2.2.4). Elevated zero maze (EZM)

The EZM was performed as previously described with modifications[41]. Briefly, the elevated circular maze (Stoelting Co., Wood Dale, IL) was divided into four quadrants of equal lengths with two opposing open quadrants lit to 30 lux with halogen lamps and two opposing closed quadrants with black acrylic walls 20 cm in height. Mice were placed in a closed quadrant and allowed to explore for 5 min, during which they were recorded with an overhead video camera. Time spent in the open quadrants, latency to enter an open quadrant and distance travelled were scored using CleverSys TopScan software (Reston, VA).

2.3. Cytokine Assay

Tissue concentration of cytokines IL-6, IL-1β and TNF-α were measured using the Bio-Plex® Mouse Cytokine Assay (Bio-Rad, USA) as described previously[28]. Selection of these cytokines was based on previous evidence supporting their association with depression and associated comorbidities[9,11,42,43]. Briefly, tissue was collected immediately following exposure to FST or from home-caged TDAG8^+/+ and TDAG8^−/− mice via rapid decapitation. Snap frozen brain and spleen samples were stored at −80°C until dissection. Samples from brain regions (hippocampus, amygdala, subfornical organ (SFO) and hypothalamus) were micro-dissected from 2.3 mm slices on a cryostat (−20°C). These regions were selected based on their association with affective behavior and physiology. The SFO was selected based on its relevance as a key brain-body interface for immunomodulation and our previous data supporting TDAG8-associated neuroimmune responses within this area[28]. Spleen and brain samples were homogenized in a dounce homogenizer with a tight fitting pestle in 500 μl of PBS supplemented with a complete ULTRA protease inhibitor cocktail tablet (Roche) and centrifuged at 14,000 RPM for 15min at 4°C. The supernatant was collected for cytokine analysis. Samples were run in duplicate and normalized to total protein content. Respectively, intra-assay and inter-assay CVs, and assay ranges were as follows: IL-6: 3%, 16%, 0.74-12,053 pg/ml; IL-1β: 4%, 7%, 10.36-60,631 pg/ml; and TNF-α: 3%, 6%, 5.8-59,626 pg/ml as reported by the manufacturer.

2.4. Spleen and Whole Brain Immune-cell Frequency Quantification

Spleens or whole brains were dissected and homogenized, and immune cells populations were quantified by flow cytometry as previously described[28,44]. The overall frequency of splenic dendritic cells populations (conventional DCs: 7-AAD^negCD45⁺CD11c⁺CD11b⁺Gr1^neg; and plasmacytoid dendritic cells: 7-AAD^negCD45⁺CD11c⁺CD11b^negGr1⁺), the frequency of activated DCs (CD86⁺ or MHC Class II⁺) or the frequency of splenic granulocytes (7-AAD^negCD45⁺CD11c⁺CD11b⁺Gr1⁺) were determined by flow cytometry. In whole brains, the overall frequency of microglia (7-AAD^negCD45^dimCD11b⁺) or macrophages (7-AAD^negCD45^brightCD11b⁺) and the frequency of activated (CD86⁺ or MHC Class II⁺) microglia or macrophages were determined by flow cytometry. AF488-conjugated anti CD80 (clone 16-10A1), PE-conjugated anti CD86 (clone GL1), PE-Cy7-conjugated anti CD11b (clone M1/70), APC-conjugated MHC Class II (clone 14-4-s), AlexaFluor700-conjugated CD45 (clone 30-F11), APC-Cy7-conjugated anti Gr1 (clone RB6-8C5), Pacific Blue-conjugated anti CD11c (clone N418) and 7-AAD were all purchased from Thermo Fisher Scientific.

2.5. Statistics

Shapiro-Wilk tests were used to check for normality. Student’s t-tests, Mann-Whitney (ranked) or Two-way ANOVA followed by Holm-Sidak’s post hoc tests were used as appropriate to determine statistical significance between groups (p<0.05). Welch’s corrections were used to correct for unequal variance. The EZM was performed subsequent to the OFT to confirm the results from the OFT. Therefore, we hypothesized a priori that TDAG8-deficient mice would elicit reduced anxiety-like behaviors and the duration in the open arm was analyzed by one-tailed t test. The Grubb’s test was used to identify and remove outliers with no more than one subject removed from any group. Prism software was used to analyze all data (GraphPad Software, Inc., La Jolla, CA).

3. Results

3.1. TDAG8 receptor deficiency attenuates depression-like behaviors

Consistent with previous observations [37], immobility in the forced swim test was significantly decreased in TDAG8^−/− mice compared to TDAG8^+/+ mice (Fig 1A; t=2.475, df= 15, p=0.026) suggesting antidepressant-like effects of TDAG8 deficiency.

Figure 1:

TDAG8-deficient mice show reduced depression- and anxiety-like behaviors (A) Immobility in the forced swim test was significantly lower in TDAG8^−/− (^−/−) mice as compared to TDAG8^+/+ (^+/+) mice. (B) There were no differences in sucrose preference nor (C) total fluid consumption. (D) TDAG8-deficient mice exhibit increased center time in TDAG8^−/− mice in the open field test (OFT) compared to TDAG8^+/+ mice. (E) Locomotion (distance traveled) in the OFT was not significantly different between genotypes. Within the elevated zero maze (EZM) (F) TDAG8-deficient mice exhibit significantly increased time in the open arms and (G) a significantly decreased latency to enter the open arm. (H) There were no differences in distance traveled within the EZM. Data represents mean ± SEM *p<0.05 (n=6-10/group).

TDAG8^−/− mice were also tested in the sucrose preference test to assess anhedonia-like behavior, relevant to depressive physiology. Consistent with previous observations[37], there were no differences in sucrose preference (Fig 1B; t=0.048, df= 16, p=0.962) nor total fluid consumption (Fig 1C; t=0.048, df= 16, p=0.962) in TDAG8^−/− mice compared to TDAG8^+/+ mice. There were no differences in sucrose consumption (data not shown; mean±SEM: TDAG8^+/+=140.9 ± 7.07; TDAG8^−/− = 133.3 ± 16.64; t=0.417 using Welch’s correction, df=9, p=0.686).

3.2. TDAG8 receptor deficiency attenuates anxiety-like behavior in the open field test and elevated zero maze

Given comorbidity between depression and anxiety disorders, we next investigated anxiety-relevant behavior in TDAG8^+/+ and TDAG8^−/− mice using the OFT and EZM [45]. Within the OFT, TDAG8^−/− spent significantly more time in the center of the open field (Fig. 1D, p = 0.039; Mann-Whitney U = 17). There was no effect of genotype on total distance traveled within the OFT (Fig. 1E, t = 1.928, df = 16, p = 0.072). Within the EZM, TDAG8^−/− spent significantly more time in the open arms of the maze (Fig. 1F, t = 1.892, df = 13, p = 0.041). Additionally, the latency to enter the open arm was significantly reduced in TDAG8^−/− mice (Fig. 1G, t = 2.504, df = 13, p = 0.026). As in the OFT, there was no effect of genotype on total distance traveled (Fig. 1H, t = 0.841, df = 13, p = 0.4154).

3.3. Tissue-specific attenuation in pro-inflammatory cytokines at baseline and post forced swim stress in TDAG8^−/− mice

Tissue- and cytokine- selective differences were observed between genotypes both peripherally and in the CNS either at baseline or following a stressor, the FST. We first evaluated cytokine levels within brain tissues. Within the hippocampus there was a significant overall effect of swim (F(1,13)=4.952 p=0.044) and interaction between genotype and swim on IL-6 expression (Fig 2A; F(1,13)=6.317 p=0.026). There was no effect of genotype (F(1,13)=2.414 p=0.144). Post hoc tests revealed a significant increase in hippocampal IL-6 following the FST only within TDAG8^+/+ mice, an effect that was blunted in TDAG8^−/− mice. There were no significant differences in IL-1β (Fig 2B; Genotype F(1,13)=0.193 p=0.668; Swim F(1,13)=0.516 p=0.485; Interaction F(1,13)=0.484 p=0.499) or TNF-α (Fig 2C; Genotype F(1,13)=0.731 p=0.408; Swim F(1,13)=0.136 p=0.718; Interaction F(1,13)=0.338 p=0.571) expression within the hippocampus. An overall effect of genotype was observed on IL-1β expression within the subfornical organ (SFO; Fig 2E; F(1,12)=5.620 p=0.035) with TDAG8^−/− mice showing reduced IL-1β compared to TDAG8^+/+ mice. There were no significant effects of swim or interaction of swim and genotype (Swim F(1,12)=0.124 p=0.731; Interaction F(1,12)=0.287 p= 0.602). There was a significant effect of swim on TNF-α expression within the SFO (Fig 2F; F(1,12)= 12.340 p=0.004). Post hoc tests revealed a significant swim-evoked reduction in TNF-α expression within TDAG8^+/+ (p<0.05), but not TDAG8^−/− mice. There were no significant effects of genotype or interaction of swim and genotype (Genotype F(1,12)=2.476 p=0.142; Interaction F(1,12)=0.151 p=0.705). There were no differences in IL-6 expression within the SFO (Fig 2D; Genotype F(1,12)= 1.815 p=0.203; Swim F(1,12)= 1.659 p=0.222; Interaction F(1,12)= 1.357 p=0.267). IL-6 and TNF-α within the amygdala were reduced in TDAG8 ^−/− mice (IL-6: Fig 2G, Genotype F(1,12)=4.025, p=0.068 (trending effect); TNF-α: Fig 2I, Genotype F(1,12)=5.175, p=0.042). There were no effects of swim or interaction of swim and genotype (IL-6: Swim F(1,12)=0.040 p=0.844; Interaction F(1,12)=0.088 p= 0.772; TNF-α: Swim F(1,12)= 1.676 p=0.220; Interaction F(1,12)=0.050 p=0.828). No differences were noted in IL-1β within the amygdala (Fig 2H; Genotype F(1,12)=3.101 p=0.104; Swim F(1,12)=0.993 p=0.339; Interaction F(1,12)= 1.915 p=0.192). Finally, there were no significant effects of genotype or swim on hypothalamic expression of IL-6 (Fig 2J; Genotype F(1,14)=0.005 p=0.943; Swim F(1,14)=0.464 p=0.507; Interaction F(1,14)=0.021 p=0.887), IL-1β (Fig 2K; Genotype F(1,14)=2.499 p=0.136; Swim F(1,14)=2.922 p=0.110; Interaction F(1,14)= 1.454 p=0.248), or TNF-α (Fig 2L; Genotype F(1,14)=0.010 p=0.920; Swim F(1,14)= 1.866 p=0.194; Interaction F(1,14)=0.891 p=0.361).

Figure 2.

Differential expression of cytokines in TDAG8^+/+ and TDAG8^−/− mice under baseline (control) and post forced swim stress (swim) within the hippocampus (A-C), subfornical organ (SFO) (D-F), amygdala (G-I),hypothalamus (J-L) and spleen (M-O). (A) In the hippocampus, IL-6 was upregulated post swim in TDAG8^+/+, but not in TDAG8^−/− mice. There were no effects of genotype or swim for (B) IL-1β or (C) TNF-α. In the SFO, there were no effects of genotype or swim on IL-6 (D). (E) IL-1β expression was significantly reduced in TDAG8^−/− mice. (F) Significant swim stress-evoked reduction in TNF-α expression was observed in TDAG8^+/+ but not in TDAG8 mice. In the amygdala, TDAG8^−/− mice elicited lower expression of (G) IL-6 and (I) TNF-α. There were no effects on (H) IL-1β expression. Within the hypothalamus, there were no effects of genotype or swim on (J) IL-6, (K) IL-1β or (L) TNF- α. (M) In the spleen, IL-6 was significantly downregulated in TDAG8^−/− mice. There were no effects on (N) IL-1β. (O) There was a significant increase in TNF-α in TDAG8^−/− mice. Data are mean ± SEM *p<0.05 **p<0.01, (J) # p=0.067 (n=4-6/group)

Given contributions of the spleen in immune dysfunction associated with stress and depression-like behaviors, we next assessed cytokine levels within this tissue[46,47]. TDAG8^−/− mice had significantly reduced IL-6 levels in the spleen. There was a significant overall effect of genotype (Fig 2M; F(1,20)= 14.15 p=0.001), though no effects of the swim or genotype x swim interaction were noted (Swim: F(1,20)= 1.213 p=0.284; Interaction F(1,20)=2.341 p=0.142). There were no significant effects of TDAG8 deficiency on IL-1β expression within spleen (Fig 2N; Genotype F(1,20)=0.022 p= 0.883; Swim F(1,20)= 1.850 p=0.189; Interaction F(1,20)=0.110 p=0.744). However, TDAG8^−/− mice had higher TNF-α expression in the spleen. There was a significant effect of genotype (Fig 2O; F(1,20) = 6.129 p=0.022), though no effects of the swim or a genotype x swim interaction were noted (Swim F(1,20)=0.083 p= 0.777; Interaction F(1,20)=0.599 p=0.448).

In order to determine whether differences in cytokines were the result of differences in immune-cell populations, we quantified the frequency of immune cell population in spleen and whole brain by flow cytometry using behaviorally-naïve mice[48]. Comparison of the frequency of splenic dendritic cells populations (DCs) with robust immunostimulatory potential (conventional dendritic cells, cDCs CD45⁺CD11c⁺CD11b⁺Gr1^neg) versus those which drive the induction of regulatory, anti-inflammatory immune responses to exogenous antigens (plasmacytoid dendritic cells, pDCs - CD45⁺CD11c⁺CD11b^negGr1⁺) in TDAG8^+/+ and TDAG8^−/− mice demonstrated no significant alteration in the overall frequency of either dendritic cell subset (Fig 3A-B; cDCs: t=0.191, df= 16, p=0.851; pDCs: t=0.113, df= 16, p=0.912) or the frequency of activated, CD86⁺ or MHC Class II⁺ DCs (Fig 3F-G, J-K; cDC CD86⁺: t=0.026, df=16, p=0.979; pDC CD86⁺: t=0.260, df= 16, p=0.798; cDC MHC Class II⁺: t=0.124, df=16, p=0.903; pDC MHC Class II⁺: 1=0.117, df=16, p=0.908). No differences were observed in the frequency of splenic granulocytes (polymorphonuclear neutrophils, PMN; CD45⁺CD11c⁺CD11b⁺Gr1⁺) (Fig 3E; t=1.225, df=13, p=0.242 using Welch’s correction). Similarly, no differences were observed in the overall frequency of microglia (CD45^dimCD11b⁺) or macrophages (CD45^brightCD11b⁺) in the whole brain of TDAG8^+/+ and TDAG8^−/− mice (Fig 3C-D; microglia: t=0.666, df=20, p=0.513; macrophages: t= 1.403, df=9, p=0.194), or in the frequency of activated (CD86⁺ or MHC Class II⁺) microglia or macrophages (Fig 3H-I,L-M; microglia CD86⁺: t=0.021, df=20, p=0.984; microglia MHC Class II⁺: t=0.275, df=20, p=0.786; macrophages CD86⁺: t=0.191, df=20, p=0.850; macrophages MHC Class II⁺: t=0.256, df=20, p=0.800).

Figure 3.

There were no effects of TDAG8 deficiency (−/−) on frequency (Freq) of (A) conventional dendritic cell (cDC), (B) plasmacytoid dendritic cells (pDC), (C) microglia, (D) macrophage or (E) splenic granulocytes (polymorphonuclear neutrophils, PMN) in spleen (A, B, E) or whole brain samples (C, D). There were also no effects of genotype on the frequency of activated, CD86⁺ (F) cDCs, (G) pDCs, (H) microglia, or (I) macrophages nor activated, MHC Class II⁺ (J) cDCs, (K) pDCs, (L) microglia, or (M) macrophages. Data are mean ± SEM *p<0.05 **p<0.01, (J) # p=0.067 (n=9-11 per group).

4. Discussion

Here, we report a role of the TDAG8 receptor in the regulation of depression- and anxiety-relevant behaviors within the FST, OFT and EZM. Additionally, TDAG8 appears to regulate cytokine expression in a tissue-selective manner under tonic condition as well as post swim stress. These differences do not appear to be an outcome of population differences between immune cell types, but rather regulation of cytokine turnover and release[38]. These data replicate and expand on our previously reported association of TDAG8 deficiency with attenuated depression-like behaviors in mice[37] suggesting a contributory role of TDAG8 in neuroimmune regulation, and depression and anxiety-associated physiology.

Previous studies have reported a discrete expression profile for the TDAG8 receptor, with enriched expression in immune cells of the brain and peripheral tissues, including microglia, macrophages and lymphocytes[26,28,38]. However, expression in other tissues (e.g. lung, dorsal root ganglion and intestine) has also been reported[32,36,49]. TDAG8, an acid-sensing receptor activated by extracellular acidification, potentially contributes to ionic homeostasis and adaptive responses to homeostatic perturbations such as acid- base balance. Accordingly, in previous studies we reported recruitment of microglial TDAG8 in fear-associated responses to carbon dioxide inhalation, a homeostatic threat[28]. Behavioral phenotypes observed in the current study extend these findings to diverse behaviors that do not involve obvious pH fluctuations suggesting that TDAG8-mediated regulation of behavior may not be confined to acid-base threats and imbalance. It is possible that TDAG8 deficiency contributes to an altered physiological state under tonic conditions that may influence behavioral performance. Contribution of homeostatic pathways to behavior and physiology is recognized [50], but not fully understood. Alternatively, deprivation of TDAG8 may have contributed to adaptations or compensatory changes that may have led to altered behaviors. Improved performance in depression- and anxiety-associated paradigms in TDAG8-deficient mice suggests regulation of common effector mechanisms that may influence behavior in these tests. Contribution of pH-sensing mechanisms in diverse behavioral phenotypes is also supported by previous studies on other acid-sensing proteins. For example, consistent with our data, acid sensing ion channels 1A and 3 (ASIC1, ASIC3) may regulate depression and anxiety-related behaviors[51,52]. Although behavior is often investigated as an outcome of exteroceptive sensory experiences, it is highly likely that internal homeostatic milieu under basal conditions contributes to behavioral performance.

Consistent with our previous study[37], no effect of TDAG8 deficiency on sucrose preference was observed. Effects on sucrose preference are generally observed following stress (e.g. chronic mild stress) [53], and it would be important to investigate whether TDAG8 deficient mice exhibit differences in sucrose preference following stress. Contrary to previous observations, no effect on sucrose consumption was observed in this study. Although the exact reason for this discrepancy is not evident, age and metabolic factors may have contributed as the current cohort was younger (2 month) compared to 6 month old mice used in the previous study. Similarly, in a previous study, we reported no effects of TDAG8 deficiency on anxiety-like behavior in the elevated plus maze (EPM)[28], which is inconsistent with the effects on OFT and EZM reported here. However, these differences may be an outcome of relative sensitivities of these tests. For example, the EZM was designed to overcome the difficulty in interpreting time spent in the central cross-section of the EPM and may represent a more sensitive test[54].

The relevance of cytokines in maintenance of homeostasis and regulation of behavior and physiology is well recognized[55]. Cytokines are prime mediators of neuroimmune crosstalk and strong evidence supports their role in peripheral and central pro-inflammatory mechanisms of depression, anxiety, cognition and memory associated behaviors[55,56]. Given the immune cell-predominant expression of TDAG8, we hypothesized potential TDAG8-mediated regulation of cytokine expression at baseline and following intense stress (forced swim test). We investigated pro-inflammatory cytokines IL-6, IL-1β and TNF-α in brain regions associated with regulation of stress and emotional behaviors, as well as a peripheral tissue, the spleen, that plays a central role in peripheral immune to brain communication and behavioral responses to stress[57]. Our data revealed genotype–dependent differences in cytokine expression that were tissue and cytokine selective, and not the result of differences in immune-cell frequency. Lack of a TDAG8 genotype effect on immune cell frequency and phenotypes has also been reported in a previous study[30]. TDAG8-associated alterations in cytokine concentrations in the absence of differences in cell phenotypes and frequencies suggest that the receptor may have regulatory effects on cytokine turnover and release, however, this remains to be investigated[58,59].

Significant differences were evident for IL-6 in the hippocampus. Interestingly, the FST-evoked increase in hippocampal IL-6 protein expression was significantly blunted in TDAG8^−/− mice suggesting a role for the receptor in stress-evoked cytokine release and potentially function. Acute forced swim stress induced increases in hippocampal IL-6 and association with immobility behavior have been reported previously[60], potentially through modulation of neuronal activity. The ability of TDAG8 deficiency to block IL-6 upregulation following the FST may be of relevance to depression-like behavior as IL-6 has long been thought to contribute to the pathogenesis of depression[61-64]. Previous studies have shown that IL-6 deficiency reduces immobility within the FST [61] and has been widely investigated in clinical and preclinical studies of depression[42,62-64]. Importantly, IL-6 is downregulated following antidepressant treatment in both humans and rodents[13,65]. TDAG8 regulation of IL-6 expression may be of interest to depression pathophysiology.

Interestingly, we observed significant reduction of IL-6 in spleen tissue of TDAG8-deficient mice at baseline. On the contrary, higher spleen TNF-α concentration was associated with TDAG8 deficiency. Altered splenic cytokines in TDAG8-deficient mice suggest potential immunomodulatory influence of TDAG8 on basal inflammatory tone. Contribution of splenic inflammation in stress-related behavior and pathophysiology has been reported[66], and splenectomy was found to attenuate peripheral as well as central inflammatory responses and depression and anxiety-related behaviors[57,67]. It would be of interest to investigate contributions of TDAG8 in stress-evoked immune and behavioral phenotypes. Consistent with our previous finding[28], the SFO, a BBB-deprived sensory circumventricular area elicited significant genotype-dependent reduction in IL-1β concentration. TDAG8 within the SFO likely contributes to neuroimmune homeostasis given the central role of this area in homeostatic balance[68], immunogenic responses[69], and behavioral regulation[28,70]. Amygdala cytokines also showed reduced expression in TDAG8-deficient mice consistent with a lower basal inflammatory milieu. Overall, our data suggest regulatory influences of TDAG8 on basal cytokine expression and responsivity that appears to be highly region and cytokine selective. Altered cytokines in TDAG8-deficient mice may contribute to behavior; however, underlying mechanisms remain to be investigated.

TDAG8 regulation of behavioral phenotypes observed in the current study likely involves neuroimmune mechanisms. Previously, our group reported microglia-specific expression of TDAG8 with no neuronal or astrocytic co-localization[28]. Since microglia-neuron signaling and inter-regulatory communication has been identified as an important contributor to physiology and behavior[71,72], we speculate that TDAG8-mediated regulation of neuroimmune homeostasis may have influenced diverse behavioral phenotypes observed in the current study. Although our previous data observed predominant localization of TDAG8 within the circumventricular organs under basal conditions, microglial TDAG8 expression (and recruitment) in other areas following psychogenic stress and immunogenic threats needs to be investigated.

In conclusion, the studies presented here identify an association of TDAG8 in the regulation of behaviors associated with depression and anxiety, as well as an altered immune profile in the CNS and periphery. Given the relevance of immune mediators in stress and psychiatric conditions such as depression, investigation of central and peripheral TDAG8 mechanisms can provide novel insights and potential intervention strategies.

Highlights.

• TDAG8-deficient mice show attenuated depression-like and anxiety-like behavior

• TDAG8^−/− mice had attenuated post-stress interleukin-6 (IL-6) in hippocampus

• Reduced basal IL-6 expression in spleen was observed in TDAG8^−/− mice

• TDAG8 receptor regulates neuroimmune homeostasis and depression-associated behaviors