FDA-approved cannabidiol [Epidiolex®] alleviates Gulf War Illness-linked cognitive and mood dysfunction, hyperalgesia, neuroinflammatory signaling, and declined neurogenesis

Animals and study design

The rats used in this study were 8-week-old male Sprague Dawley rats (n = 53; Harlan, Indianapolis, IN, USA) and were provided with ad libitum access to standard chow containing 24.3% crude protein, 4.7% fat, 40.2% carbohydrates, and 4% crude fiber (Teklad rodent diet 8604; Inotiv, West Lafayette, Indiana, USA), as well as water. All studies conducted in this investigation received approval from the Institutional Animal Care and Use Committee of Texas A&M University (Animal Use Protocol number: 2022-0166D). The rats were randomly assigned to either the naïve control group (n = 18) or the GWI group (n = 35). A schematic diagram outlining the procedures and analyses conducted in this investigation is depicted in Fig. 1a. Rats in the GWI group received oral administration of PB, dermal application of DEET and PER, as well as daily restraint stress for 15 min for 28 d, following previously described protocols [28–31, 33]. Six months after exposure to chemicals and stress, a cohort of rats from the GWI group underwent vehicle (VEH, sterile water) treatment (GWI-VEH group; n = 17) for 16 weeks (5 d/week), while another cohort received Epidiolex^® treatment (GWI-CBD group; n = 18) during the same period (5 d/week). The decision to administer VEH or CBD to GWI rats starting from 6 months post-exposure was based on previous findings that at this time point these animals exhibit significant cognitive and mood impairments, along with increased oxidative stress and neuroinflammation [28]. A separate cohort of rats subjected solely to stress was not included in parallel, since an independent study conducted in our laboratory demonstrated that exposure to stress alone (i.e., 15 min of restraint stress daily for 28 d) does not lead to cognitive impairments, neuroinflammation or decline in hippocampal neurogenesis (see Additional file 1: Results and Figs. S1–S5).

During the 7th week of VEH/CBD treatment, the rats received 5’-bromodeoxyuridine (BrdU) injections (100 mg/kg) for 5 d to label newly generated cells and neurons in the hippocampus. Starting from the 11th week after VEH/CBD treatment, animals in all groups were tested with a series of neurobehavioral tasks during the daytime. After 16 weeks of VEH/CBD treatment, subgroups of rats were deeply anesthetized using isoflurane, and their brains were fixed with 4% paraformaldehyde via intracardiac perfusions. The fixed brain tissues were then processed for immunohistochemical and immunofluorescence analyses. Additional subgroups of rats were deeply anesthetized using isoflurane followed by decapitation. Fresh brains were harvested, snap-frozen, and stored at − 80 °C until used for biochemical assays. The investigators who performed the neurobehavioral tests, immunohistochemical or immunofluorescence quantifications, as well as biochemical assays remained blinded to group identification of animals or samples throughout the study period.

Analyses of motor, cognitive, and mood function and nociception in GWI rats

Changes in motor function were evaluated in animals from all groups using an open field test (OFT). Then, 5 neurobehavioral tests were employed to investigate the cognitive and mood function as well as nociception. An object-in-place test (OIPT) was used to test the ability of animals to associate objects with previously encountered locations [55]. Proficiency in this task relies on an interaction between the hippocampus, the perirhinal cortex, and the medial prefrontal cortex. An object location test (OLT) was utilized to assess the capacity of animals to discern subtle changes in their immediate environment, a cognitive function dependent on the integrity of the hippocampus [55]. A pattern separation test (PST) examined the ability of animals to discriminate similar but not identical representations in a non-overlapping manner. Maintenance of this function depends upon the integrity of the dentate gyrus (DG) in the hippocampus, including the extent of neurogenesis [56, 57]. The degree of anhedonia, a depressive-like behavior, was evaluated through a sucrose preference test (SPT) [58, 59]. Hyperalgesia was ascertained using an electric Von Frey apparatus (Maze Engineers), which measured the force [millinewtons (mN)] exerted by the monofilament applied perpendicularly to the paw with an incremental increase until eliciting the paw withdrawal reflex [60].

Tissue processing and immunohistochemistry

The fixed brain tissues were processed for cryostat sectioning, and 30-µm thick coronal brain tissue sections were stored at − 20 °C in a cryobuffer until further use, as described in our previous reports [28–31, 61–63]. Serial sections (every 15th or 20th) through the septotemporal axis of the entire hippocampus from each animal were processed for immunohistochemical studies to identify glial fibrillary acidic protein (GFAP)⁺ astrocytes, ionized calcium-binding adapter molecule-1 (IBA-1)⁺ microglia, newly born cells and neurons expressing BrdU or doublecortin (DCX). The primary and secondary antibodies employed are listed in Additional file 1: Table S2. The avidin–biotin complex reagent and the 3,3′-diaminobenzidine and Vector Gray substrate kits were supplied by Vector Labs (Burlingame, CA, USA). The brain tissue sections were mounted on gelatin-coated slides, air-dried, counterstained with nuclear fast red or hematoxylin (Vector Labs), dehydrated, cleared, coverslipped with permount solution, and examined under a Nikon E600 microscope [28–31].

Evaluation of activated microglia

A dual immunofluorescence technique was utilized to visualize microglia expressing both IBA-1 (a marker of all microglia) and cluster of differentiation 68 (CD68, a marker for activated microglia) in various regions of the hippocampus, following the methods described in our previous publications [17, 28, 30, 31]. The primary and secondary antibodies employed for IBA-1 and CD68 immunofluorescence studies are listed in Additional file 1: Table S2. Z-section analysis was performed using a Nikon confocal microscope to quantify the percentages of microglia expressing IBA-1 and CD68. We counted the total numbers of microglia as well as the percentages of microglia expressing CD68 per unit area (approximately 216 μm²), which were then converted into percentages representing microglia expression of CD68. For these measurements, three consecutive sections (every 20th section) were analyzed in each animal (n = 6) [17, 18].

Quantification of IBA-1⁺ microglial elements

The area fractions (AFs) of microglia (IBA-1⁺ structures) were quantified in the DG and hippocampal CA1 and CA3 subfields (n = 6) using ImageJ. The AF of thresholded pixels was expressed as a percentage of the total image area, which was measured in square millimeters. The microscopic images from different hippocampal subfields were digitized using a Nikon E600 microscope equipped with a 20 × objective lens and connected to a digital video camera linked to a computer. Each grayscale image was imported into ImageJ, where a binary image was generated by selecting an appropriate threshold value that highlighted all IBA1⁺ immunostained elements while eliminating background noise. The area occupied by the IBA1⁺ structures (AF) in the binary image was then measured using the Analyze command within the software. The AF represents the proportion of highlighted pixels in the image. For each animal, separate calculations were performed to determine the AF of IBA1⁺ structures in each hippocampal region, using data from all chosen serial sections before determining the means and standard errors for each group (n = 6). Three serial sections (every 20th section) were analyzed per animal (n = 6) for these measurements [17, 18].

Quantification of NLRP3 inflammasome complex in microglia

We next probed the formation of NLRP3 inflammasome complexes in microglia through a triple immunofluorescence method for NLRP3, ASC, and IBA-1 [30, 31, 63, 64]. The primary and secondary antibodies employed for NLRP3, ASC, and IBA-1 triple immunofluorescence studies are listed in Additional file 1: Table S2. The sections were next examined through 2-μm thick, Z-section analysis using Leica THUNDER 3D Imager. The total number of NLRP3 inflammasome complexes (i.e., structures displaying NLRP3 and ASC co-expression) per unit area (about 216 μm²) of the CA3 subfield was measured using three representative sections per animal (n = 5). Also, the percentages of IBA-1⁺ microglia containing NLRP3-ASC complexes were measured.

Preparation of hippocampal tissue lysates for biochemical assays

Each hippocampus was microdissected and lysed by sonication for 15–20 s at 4 °C in a tissue extraction reagent containing protease and phosphatase inhibitor. Following centrifugation of the resulting solution for 10 min at 15,000g, the supernatant was aliquoted and kept at − 80 °C until further use. The details of the reagents are listed in Additional file 1: Table S1.

Quantification of markers of the NLRP3 inflammasome pathway, the oxidative stress, the JAK/STAT pathway, anandamide, and astrocyte hypertrophy

The hippocampal lysates from rats belonging to different groups were processed to quantify the concentrations of various markers associated with NLRP3 inflammasome activation and proinflammatory cytokines (n = 6). Individual enzyme-linked immunoassay (ELISA) kits, as detailed in our previous studies [30, 64] and Additional file 1: Table S1, were used for this purpose.

The hippocampal lysates were also processed to measure the levels of MDA, PCs, catalase (CAT), and superoxide dismutase (SOD) using assay kits as detailed in our previous studies [30, 31] and Additional file 1: Table S1. The concentrations of these markers were normalized to the total protein content in the respective tissue lysates. Subsequently, we employed ELISA to quantify the mediators and markers associated with JAK/STAT signaling, including IL-6, interferon-γ (IFN-γ), phosphorylated JAK-1 (p-JAK-1), and phosphorylated STAT1 (p-STAT1). Additionally, we assayed the concentration of endogenous cannabinoid anandamide using a kit provided by MyBioSource (San Diego, CA, USA). The assay kits employed for these measurements are listed in Additional file 1: Table S1.

The AFs of astrocytes (GFAP⁺ structures) were determined in the DG and hippocampal CA1 and CA3 subfields (n = 6) using ImageJ. Three serial sections (every 20th) were employed in each animal (n = 6) for these measurements [17, 18].

Measurement of newly born cells positive for BrdU and newly born neurons expressing DCX

The entire hippocampus was sectioned into 30-µm slices (every 15th slice) for visualization of BrdU⁺ newly born cells or DCX⁺ newly born neurons in the subgranular zone-granule cell layer (SGZ-GCL) of the hippocampus. A total of 6 sections were analyzed per marker per animal, with a sample size of n = 5–6. Stereological quantification of BrdU⁺ cells and DCX⁺ neurons was performed using a StereoInvestigator system (Microbrightfield, Williston, Vermont, USA), following the methodology described in our previous reports [31, 59, 65, 66]. Within each animal, BrdU⁺ cells and DCX⁺ neurons were counted from randomly and systematically selected frames measuring 40 μm × 40 μm (0.0016 mm² area) in each chosen section under a 100 × oil immersion objective lens. The selection frequency of serial sections across groups was determined based on the density distribution of BrdU⁺ cells and DCX⁺ neurons, with every 15th section being included. The StereoInvestigator system’s optical fractionator component used a systematic random sampling scheme to determine the number and density of frames for analysis in each section. The dissector height for counting frames was set at 8 μm to ensure accurate cell counting. By selecting frames randomly and systematically throughout the SGZ-GCL in each serial section, we ensured that every BrdU⁺ cell or DCX⁺ neuron had an equal chance of being counted despite any uneven distribution within the SGZ-GCL among different experimental groups. The average Gunderson coefficient error values ranged from 0.126 to 0.198 for BrdU⁺ cells across all groups, and between 0.111 and 0.156 for DCX⁺ neurons.

Quantification of neuronal differentiation of BrdU⁺ newly born cells, and net hippocampal neurogenesis

We subsequently quantified the percentages of newly born cells (i.e., BrdU⁺ cells) that underwent neuronal differentiation in the SGZ-GCL using dual immunofluorescence for BrdU and neuron-specific nuclear antigen (NeuN), as well as Z-section analyses [28, 31]. For these investigations, we utilized mouse anti-NeuN, mouse anti-BrdU, donkey anti-mouse IgG labeled with Alexa Flour 488, and donkey anti-rat IgG labeled with Alexa Flour 594. The primary and secondary antibodies employed for BrdU immunostaining and BrdU-NeuN dual immunofluorescence are listed in Additional file 1: Table S2. Two-micrometer thick optical Z-sections were taken using a Nikon confocal microscope from 3 representative sections at anterior, mid, and posterior levels of the hippocampus in each animal (n = 5, 25–30 BrdU⁺ cells/animal). Subsequently, by combining the absolute numbers of BrdU⁺ cells in the SGZ-GCL obtained through stereology with the percentages of BrdU⁺ cells expressing NeuN, we estimated the total number of mature neurons added to the GCL over 5 d in each animal.

Statistical analyses

Data were expressed as means ± SEM. For comparing the two groups, either a two-tailed, unpaired Student’s t-test or the Mann–Whitney U test was employed, depending on the presence of significant differences in standard deviations between the groups. When dealing with comparisons involving three or more datasets, one-way analysis of variance (ANOVA), followed by Newman–Keuls multiple comparison post hoc tests, was used. In case any individual group failed the normality test (Shapiro–Wilk test), the Kruskal–Wallis test with Dunn’s post-hoc tests were utilized. To compare body weights across groups over 16 weeks, two-way repeated measures ANOVA was employed. Throughout all comparisons, statistical significance was defined as P < 0.05.

CBD treatment did not cause weight loss in GWI rats

We monitored the body weights of GWI rats receiving either VEH or CBD throughout a 16-week treatment regimen, weekly. Throughout the entire 16-week treatment period, we did not observe any statistically significant differences in body weights between the two groups (Additional file 1: Fig. S6). Furthermore, no adverse effects were observed in GWI rats receiving CBD.

CBD treatment did not impact motor function in GWI rats

Evaluation of the total distance traveled and the velocities of movement in an OFT revealed no differences in motor function among the naïve, GWI-VEH, and GWI-CBD groups (Fig. 1b). Thus, long-term CBD treatment did not induce motor impairments.

CBD treatment improved associative recognition memory function in GWI rats

Naïve control rats demonstrated the ability to associate objects with specific locations based on previous encounters. This was evident from their preference to explore the 2 objects that swapped positions in T3 (i.e., OINP) over the 2 objects that remained in their original locations from T2 (OIFP; P < 0.01, Fig. 2b). Rats in the GWI-VEH group spent more time exploring OIFP during TOETs compared to OINP (P < 0.05, Fig. 2b), implying impaired recognition memory associated with spatial information. In contrast, rats in the GWI-CBD group allocated significantly higher percentage of their TOET towards the objects that swapped positions (OINP; P < 0.01, Fig. 2b), confirming intact spatial recognition memory. Comparison of the discrimination index (DI) for OINP using one-way ANOVA across groups revealed significant differences (Fig. 2c). The OINP-DI was significantly lower in the GWI-VEH group compared to the naïve control group (P < 0.05, Fig. 2c). Conversely, GWI rats receiving CBD displayed similar OINP-DI as observed in the naïve control group (P > 0.05, Fig. 2c). The duration of TOETs during T2 and T3 did not differ significantly between groups (P > 0.05, Fig. 2d). Thus, CBD treatment improved associative recognition memory function in GWI rats.

CBD treatment alleviated object location memory dysfunction in GWI rats

In this test, animals proficient in object location memory show a greater inclination to inspect the object that has been relocated to a novel location (OINP) instead of the OIFP. In T3, naïve rats preferred the OINP over the OIFP (P < 0.05), whereas GWI rats treated with VEH showed impairment as they spent similar amounts of time exploring both the OIFP and OINP (P > 0.05, Fig. 2f). Conversely, GWI rats that received CBD exhibited increased percentages of their TOET with the OINP (P < 0.01, Fig. 2f), mirroring the behavior of naïve control rats. One-way ANOVA analysis of the OINP-DI values across groups confirmed the proficiency of the GWI-CBD group for object location memory, as their DI values matched those of animals in the naïve control group (P > 0.05, Fig. 2g). However, the DI values in the GWI-VEH group were lower than those of animals in both naïve and GWI-CBD groups (P < 0.05, Fig. 2g). The TOETs in T2 and T3 were comparable among all groups (P > 0.05, Fig. 2h). Thus, CBD treatment alleviated dysfunction in object location memory observed in GWI rats.

CBD treatment eased pattern separation memory dysfunction in GWI rats

In the PST, animals proficient in pattern separation exhibited a greater preference for exploring the NO on P2 compared to the FO on P2 in T4. Impaired pattern separation was observed in rats belonging to the GWI-VEH group, as they did not show a significant preference for NO on P2 over FO on P2 (P > 0.05), unlike the naïve control group which demonstrated a strong preference for NO on P2 (P < 0.001, Fig. 3b). In contrast, rats in the GWI-CBD group spent a higher percentage of their TOET with NO on P2 (P < 0.05, Fig. 3b), exhibiting behavior similar to that of rats in the naïve control group. Differences between groups were also evident from the one-way ANOVA analysis of the DI values for the NO on P2 across groups (P = 0.05, Fig. 3c). The TOETs in T2 and T3 showed no significant differences among groups (P > 0.05, Fig. 3d). Thus, CBD treatment effectively improved pattern separation ability in GWI rats.

CBD treatment alleviated anhedonia in GWI rats

Naïve control rats preferred sucrose water over standard water, indicating the absence of anhedonia (P < 0.0001, Fig. 3e). Rats in the GWI-VEH group displayed anhedonia, as evidenced by their equal preference for sucrose-containing water and regular water (P > 0.05, Fig. 3e). Remarkably, GWI rats treated with CBD exhibited a preference for sucrose-containing water compared to standard water, suggesting the absence of anhedonia (P < 0.01, Fig. 3e). The sucrose preference rate (SPR) was reduced in GWI rats that received VEH compared to naïve control rats (50.26% vs. 74.55%, P < 0.0001; Fig. 3f). Additionally, GWI rats treated with CBD showed significantly improved SPR than GWI-VEH rats (64.18% vs. 50.26%, P < 0.01; Fig. 3f). However, the overall SPR in the GWI-CBD group was still lower than that in the naïve control group (P < 0.05, Fig. 3f). Furthermore, there were no significant differences in the total fluid consumption between different groups (P > 0.05, Fig. 3g). Thus, while CBD treatment in GWI rats resulted in an increase of 13.92% sucrose preference, it did not fully restore their sucrose preference to levels observed in the naïve control group. Nonetheless, CBD treatment effectively alleviated anhedonia or depressive-like behavior in rats with chronic GWI.

CBD treatment reduced hyperalgesia in GWI rats

The presence of persistent pain has emerged as one of the most incapacitating symptoms associated with chronic GWI [67, 68]. In line with this, rats in the GWI-VEH group exhibited significantly reduced paw withdrawal thresholds compared to those in the naïve control group during the Von Frey test (P < 0.01, Fig. 3h). In contrast, rats in the GWI-CBD group demonstrated paw withdrawal thresholds matched the thresholds that were comparable to those observed in the naïve control group (P > 0.05, Fig. 3h). Consequently, CBD treatment effectively reduced hyperalgesia in chronic GWI rats.

CBD treatment diminished microgliosis and activation of microglia in chronic GWI rats

We utilized ImageJ to analyze the AF of IBA-1⁺ microglial elements in the hippocampus. Figure 4a depicts the IBA-1⁺ microglial structures in the hippocampal CA1 and CA3 subfields. One-way ANOVA revealed significant differences in AFs of IBA-1⁺ elements between groups in both CA1 and CA3 regions, as well as when considering the entire hippocampus (P < 0.01, Fig. 4a). However, no such differences were observed in the DG region (P > 0.05, Fig. 4a). In comparison to the naïve control group, higher levels of IBA-1⁺ microglial elements were observed in the GWI-VEH group within CA1, CA3 subfields, and throughout the entire hippocampus (P < 0.05, Fig. 4a). In contrast, CBD treatment in GWI rats normalized the extent of IBA-1⁺ microglial elements to those in the naïve control group (P > 0.05, Fig. 4a). The AF of IBA-1⁺ microglial elements was lower in both CA1 subfield and entire hippocampus for GWI-CBD group compared to the GWI-VEH group (P < 0.01, Fig. 4a). The increased density and hypertrophy of microglia soma likely contributed to higher AFs of microglial elements observed in the hippocampus of GWI-VEH group compared to naïve control group. Moreover, there was a comparable level of AFs between GWI-CBD and naïve control groups since CBD treatment reduced both density and hypertrophy of microglia soma. Qualitatively, microglial morphology resembled that seen in the naïve control group for GWI-CBD rats. Thus, CBD treatment effectively mitigated hyperplasia and hypertrophy exhibited by microglia soma in GWI rats’ brains.

We also examined the presence of activated microglia in the CA1, CA3, and DG subfields of the hippocampus (Fig. 4b). Activated microglia were visualized through dual immunofluorescence staining for IBA-1 and CD68. One-way ANOVA analysis revealed significant differences between groups in the CA1 and CA3 subfields as well as the entire hippocampus (P < 0.05, Fig. 4b). However, no such differences were observed in the DG subfield (P > 0.05, Fig. 4b). Post hoc tests conducted on the CA1 and CA3 subfields and the whole hippocampus uncovered that rats in the GWI-VEH group exhibited higher percentages of activated microglia compared to those in the naïve control group (P < 0.05, Fig. 4b). Treatment with CBD reduced these percentages of the IBA-1⁺ cells expressing CD68 to levels comparable to those seen in naïve controls (P > 0.05, Fig. 4b), implying mitigation of activated microglia phenotypes by CBD treatment. Notably, there was a significant reduction in percentages of activated microglia in both the CA1 subfield and the entire hippocampus following CBD treatment compared to the GWI-VEH group (P < 0.01, Fig. 4b). Thus, CBD treatment effectively decreased the percentages of activated microglia in the hippocampus.

CBD treatment inhibited the activation of NLRP3 inflammasomes in chronic GWI rats

We investigated the formation of NLRP3 inflammasome complexes and their activation in the hippocampus. Using triple immunofluorescence staining for NLRP3, ASC, and IBA-1, we initially observed the activation of NLRP3 inflammasome complexes specifically in microglia (Fig. 5a). One-way ANOVA analysis uncovered significant differences between groups regarding both the number of inflammasome complexes per unit area and the percentages of microglia with inflammasome complexes (P < 0.001, Fig. 5a). Compared to the naïve control group, the GWI-VEH group exhibited an elevated number of inflammasome complexes as well as a higher percentage of microglia expressing inflammasome complexes (P < 0.001, Fig. 5a). Treatment with CBD resulted in a reduction in both the number of inflammasome complexes per unit area and the percentages of microglia expressing inflammasome complexes in the hippocampus (P < 0.01, Fig. 5a). Consequently, CBD treatment effectively decreased overall incidence rates of NLRP3 inflammasome complex formation within microglia located in the hippocampus.

Next, we quantified the concentrations of the transcription factor NF-κB p65 and mediators involved in NLRP3 signaling (NLRP3, ASC, and cleaved caspase-1) in the hippocampus. One-way ANOVA analysis revealed significant differences among groups for these markers (P < 0.05, Fig. 5b). The concentrations of these proteins were found to be upregulated in the GWI-VEH group compared to the naïve control group (P < 0.05, Fig. 5b). Notably, CBD treatment normalized the concentrations of these proteins to levels observed in the naïve control group (P > 0.05, Fig. 5b). Furthermore, there were significant variations between groups regarding IL-1β and IL-18, which are end products of NLRP3 inflammasome activation (P < 0.05, Fig. 5b). The concentration of IL-18 was significantly increased in the GWI-VEH group (P < 0.01, Fig. 5b) but it returned to a level similar to that seen in the naïve control with CBD treatment. Additionally, compared to GWI-VEH, GWI-CBD showed a lower concentration of IL-1β (P < 0.05, Fig. 5b). Overall, CBD treatment in GWI rats reduced concentrations of the transcription factor NF-κB p65 responsible for triggering the formation of NLRP3 inflammasome complexes and proteins that facilitate NLRP3 inflammasome activation (NLRP3, ASC, and cleaved caspase-1), leading to decreased levels of proinflammatory cytokines IL-1β and IL-18 in the microenvironment. Thus, CBD treatment effectively suppresses chronic inflammation perpetuated by NLRP3 inflammasome activation.

CBD treatment diminished oxidative stress in chronic GWI rats

We investigated the efficacy of CBD in diminishing oxidative stress in rats with chronic GWI by quantifying concentrations of oxidative stress (MDA and PCs) and antioxidant markers (SOD and CAT). One-way ANOVA demonstrated significant differences between groups for MDA, PCs, and CAT (P < 0.05, Fig. 5c). However, no significant differences were observed between groups for SOD (P > 0.05, Fig. 5c). Post hoc analysis revealed that the concentration of MDA was significantly higher in the GWI-VEH group compared to the naïve control group (P < 0.001, Fig. 5c). Although not statistically significant, the concentration of PCs was also higher in the GWI-VEH group than in the naïve control group (P > 0.05, Fig. 5c). In contrast, both MDA and PCs concentrations were reduced in the GWI-CBD group compared to the GWI-VEH group (P < 0.05), resembling levels seen in the naïve control group (P > 0.05, Fig. 5c). The concentration of CAT was lower in the GWI-VEH group compared to the naïve control group (P < 0.05), while it remained similar to that of the naïve control group after CBD treatment in the GWI-CBD group (P > 0.05, Fig. 5c). Thus, CBD treatment reduced oxidative stress in rats with GWI.

CBD treatment mitigated JAK/STAT signaling in chronic GWI rats

We investigated whether CBD treatment can modulate the activation of JAK/STAT signaling in GWI rats by quantifying proteins that trigger (IL-6 and IFN-γ) and activate (p-JAK-1 and p-STAT-1) JAK/STAT pathway. One-way ANOVA analysis of IL-6 and IFN-γ concentrations across groups revealed significant differences between groups (P < 0.01, Fig. 6a). Notably, both IL-6 and IFN-γ were significantly upregulated in the hippocampus of the GWI-VEH group (P < 0.05, Fig. 6a). However, CBD treatment normalized the concentrations of IL-6 and IFN-γ to levels observed in the naïve control group (P > 0.05, Fig. 6a). The decreased levels of IL-6 and IFN-γ mediated by CBD led to reduced concentrations of p-JAK-1 and p-STAT-1 in the GWI-CBD group. While one-way ANOVA analysis revealed significant differences between groups for p-JAK-1 and p-STAT-1 (P < 0.01, Fig. 6b), post hoc analysis showed that concentrations of p-JAK-1 and p-STAT-1 were significantly upregulated in the hippocampus of the GWI-VEH group compared to those in the naïve control group (P < 0.05, Fig. 6b). Notably, CBD treatment normalized the concentrations of p-JAK-1 and p-STAT-1 in the GWI-CBD group to levels in the naïve control group (P > 0.05, Fig. 6b). Thus, CBD treatment can suppress the activation of the JAK/STAT signaling in GWI rats by reducing the concentrations of IL-6 and IFN-γ.

CBD treatment did not alter endogenous anandamide levels in chronic GWI rats

The effect of CBD treatment on endogenous cannabinoid N-arachidonoylethanolamine (or anandamide) in chronic GWI was also investigated. One-way ANOVA analysis did not reveal any significant differences between the groups (P > 0.05, Fig. 6c), suggesting that exogenous CBD did not affect the concentration of endogenous cannabinoids in chronic GWI.

CBD treatment diminished hypertrophy of astrocytes in chronic GWI rats

Using ImageJ, we examined the AF of GFAP⁺ astrocytes in different subfields of the hippocampus, including DG, CA1, and CA3. The representative images displayed GFAP⁺ astrocytes specifically in the CA1 and CA3 subregions of the hippocampus (Fig. 7). Statistical analysis using one-way ANOVA revealed significant differences among groups regarding GFAP⁺ structures within these regions (P < 0.05, Fig. 7). Subsequent post hoc analysis demonstrated that compared to the naïve control group, GWI-VEH group exhibited elevated levels of GFAP⁺ structures in both CA1 and CA3 subfields as well as throughout the entire hippocampus (P < 0.05, Fig. 7). Although ANOVA analysis indicated statistical differences between groups in DG, no significant variations were observed through post hoc tests (Fig. 7). Notably, CBD treatment normalized astrocyte features as evidenced by comparable levels of astrocytic elements between GWI-CBD group and naïve control group within these regions (P > 0.05, Fig. 7). Thereby, CBD treatment reduced astrocyte hypertrophy in the hippocampus of GWI rats.

CBD treatment improved hippocampal neurogenesis in chronic GWI rats

Using BrdU labeling, we quantified the numbers of newly born cells and neurons in the 8th week of VEH/CBD treatment and survived for 8 weeks. Figure 8a displays examples of BrdU-labeled newly born cells. Examples of BrdU⁺ newly born cells that differentiated into mature neurons from naïve control, GWI-VEH, and GWI-CBD groups are shown with dual immunofluorescence for BrdU and NeuN (Fig. 8b). One-way ANOVA analysis comparing the total number of BrdU⁺ newly born cells and net neurogenesis in the SGZ-GCL across groups revealed significant differences between the groups (P < 0.0001; Fig. 8a, c). However, there were no group differences in the differentiation of newly born cells into neurons between groups (P > 0.05, Fig. 8b). Post hoc tests indicated that the GWI-VEH group had decreased numbers of newly born cells and net neurogenesis compared to the naïve control group (P < 0.0001; Fig. 8a, c). However, CBD treatment increased these measures of neurogenesis in GWI rats compared to the GWI-VEH group (P < 0.01; Fig. 8a, c). Thus, a 9-week CBD intervention could enhance hippocampal neurogenesis in GWI rats.

We also quantified the status of hippocampal neurogenesis after 16 weeks of either VEH or CBD treatment by assessing the number of DCX⁺ newly born neurons in the SGZ-GCL. The morphology and distribution of DCX⁺ neurons in the SGZ-GCL from different groups are depicted in Fig. 8d. One-way ANOVA analysis revealed significant differences in DCX⁺ new neuron counts among the groups (P < 0.01, Fig. 8d). The GWI-VEH group exhibited a lower number of DCX⁺ newly born neurons compared to the naïve group (P < 0.01, Fig. 8d). However, CBD treatment restored the levels of DCX⁺ new neurons to those observed in the naïve control group (P > 0.05, Fig. 8d). Thus, 16 weeks of CBD treatment can enhance hippocampal neurogenesis to levels comparable to those seen in naïve rats.

Mechanisms of CBD-mediated modulation of oxidative stress

Our previous studies have analyzed the brain tissues of rats with chronic GWI at various time points. These studies have identified persistent oxidative stress as one of the primary features of the pathophysiology of chronic GWI [16, 28, 30, 31]. Evidence of oxidative stress in chronic GWI includes elevated levels of MDA and PCs, reduced antioxidants, and increased expression of genes encoding proteins linked to heightened oxidative stress, mitochondrial dysfunction, and dysregulated nuclear factor erythroid 2-related factor 2 (Nrf-2) signaling [16, 28]. This study has shown that CBD can mitigate oxidative stress by reducing concentrations of MDA and PCs while increasing levels of the antioxidant enzyme CAT. The effectiveness of CBD treatment in reducing oxidative stress depends on its duration and dose [69]. CBD indirectly reduces oxidative stress by regulating intracellular pathways such as inflammation [70]. It also directly affects cellular redox status by interrupting free radical chain reactions, chelating transition metal ions, weakening lipid peroxidation, and boosting antioxidant enzyme activity [69–71]. The hydroxyl groups, the pentyl chain in the phenolic ring, and the methyl group on the cyclohexene ring of CBD are responsible for its direct redox-modulating effect [42]. Furthermore, CBD regulates Nrf2 transcriptional activity by downregulating inhibitors such as Kelch ECH associating protein 1 (Keap1) and transcriptional repressor BTB and CNC homology 1 (Bach1), while upregulating activators including Krüppel-associated box-associated protein 1 (KAP1), cyclin-dependent kinase inhibitor p21, and sequestosome-1 (p62) [72]. Additionally, CBD can reduce oxidative stress by altering the endocannabinoid concentration [69]. The current study has demonstrated that CBD is capable of reducing neuroinflammation, lipid peroxidation, and protein oxidation, while simultaneously increasing the antioxidant enzyme CAT. Interestingly, CBD treatment did not have an impact on the concentration of anandamide, suggesting that it can effectively alleviate oxidative stress in GWI without interfering with the endogenous endocannabinoid system. This finding differs from the results in a previous study examining the effects of CBD in schizophrenia patients, where doses escalated to 800 mg/d within the first week and continued for 3 weeks [73]. The discrepancy between this prior study and our current research regarding the effects of exogenous CBD treatment on endogenous anandamide likely stems from variations in species and dosages employed.

Mechanisms of CBD-mediated modulation of chronic neuroinflammation

Previous studies have consistently reported that chronic neuroinflammation is a prominent feature of GWI in the brain [2, 17, 28–31]. In GWI rats, chronic neuroinflammation is characterized by persistent astrocyte hypertrophy and increased percentages of activated microglia presenting CD68 and NLRP3 inflammasome complexes, elevated levels of proteins involved in NLRP3 inflammasome activation, as well as heightened concentrations of proinflammatory cytokines. Indeed, GWI rats receiving VEH treatment in this study demonstrated most of the above pathophysiological changes in the hippocampus. However, the hippocampus of CBD-treated GWI rats displayed normalization of astrocyte hypertrophy, percentages of microglia with CD68 or NLRP3 inflammasome complexes, levels of proteins mediating NLRP3 inflammasome activation (NF-κB p65, NLRP3, ASC, and cleaved caspase-1), concentrations of end products of activated NLRP3 inflammasomes (IL-1β and IL-18), as well as levels of proteins activating JAK/STAT signaling (IL-6, IFN-γ, p-JAK-1, and p-STAT1), which were restored to naïve control levels. These findings indicate that CBD can effectively attenuate neuroinflammatory signaling in chronic GWI.

CBD treatment can reduce neuroinflammation through multiple mechanisms [74–79]. Firstly, it exerts an inhibitory effect on NF-κB signaling, which is responsible for activating genes encoding proinflammatory cytokines such as IL-1, IL-6, TNF-α, and IFN-γ, as well as NLRP3 inflammasome formation [74–76, 78, 79]. Several studies have demonstrated that CBD can inhibit the NF-κB p65 element involved in this process. Consistent with these findings, long-term CBD treatment to GWI rats in this study normalized the levels of NF-κB p65, IL-1β, IL-6, and IFN-γ to those observed in naïve control levels. Secondly, CBD directly inhibits NLRP3 inflammasome activation [74, 75, 79], which plays a significant role in chronic neuroinflammation associated with GWI [30, 31]. This effect is mediated through the inhibition of genes encoding NLRP3 and caspase-1 (mediators of NLRP3 inflammasome activation), as well as IL-18 and IL-1β (end products of NLRP3 inflammasome activation), Toll-like receptor adaptor myeloid differentiation primary response 88 (MYD88), IFN-γ receptors 1 and 2 (IFNGR1 and IFNGR2), and mitogen-activated protein kinases [74, 75, 79]. Notably, CBD exhibits comparable efficacy to MCC950, a small molecule inhibitor targeting NLRP3, in terms of inhibiting NLRP3 inflammasome activation and production of IL-1β [79]. The results obtained from this study are consistent with the aforementioned activities attributed to CBD since it reduced both the incidences of microglia containing NLRP3 inflammasome complexes and the concentrations of NF-κB p65, NLRP3, ASC, cleaved caspase-1, IL-1β, and IL-18 within the hippocampus.

Another mechanism by which CBD can reduce neuroinflammation is through modulating JAK/STAT signaling [78]. Previous in vitro studies have shown that CBD can downregulate JAK/STAT pathways by decreasing the concentrations of IL-6 and IFN-γ [78], which are the activators of JAK/STAT signaling [80–82]. Studies conducted on models of arthritis, multiple sclerosis, and type 1 diabetes have also shown the ability of CBD to inhibit the JAK/STAT activator IFN-γ [73, 83–86]. The present study has validated the efficacy of long-term CBD treatment in suppressing the activation of JAK/STAT signaling, as evidenced by reduced levels of IL-6, IFN-γ, p-JAK-1, and p-STAT1. Given that phosphorylation of STAT1 leads to robust proinflammatory effects, diminished phosphorylation of STAT1 observed in this study confirmed the potential of CBD for reducing neuroinflammation through modulation of JAK/STAT signaling. Thus, prolonged administration of CBD treatment holds significant promise for mitigating chronic neuroinflammation in GWI by attenuating activated microglia, NLRP3 inflammasome activity, and JAK/STAT signaling pathways. These findings align with previous research suggesting that CBD possesses anti-neuroinflammatory properties via inhibition of proinflammatory cytokines TNF-α, IL-1β, IL-6, and IFN-γ [87–89].

Potential mechanisms of CBD-mediated improvements in hippocampal neurogenesis

This study also demonstrated the neurogenic effects of CBD in chronic GWI. These effects were evident through an increase in the number of newly born cells (i.e., dentate granule cells) labeled with BrdU, which were generated during the 7th week of CBD treatment and survived for extended periods in the GCL (net hippocampal neurogenesis). Additionally, there was an increase in DCX⁺ newly born neurons observed in the SGZ-GCL, indicating an improved status of neurogenesis following 16 weeks of CBD treatment. While it is known that decreased oxidative stress and neuroinflammation mediated by CBD can increase neurogenesis by enhancing the proliferation of stem/progenitor cells in the hippocampus because of a more favorable microenvironment [90, 91], it is important to consider the potential direct effects of CBD on improving hippocampal neurogenesis. This possibility arises from the activity of CBD on cannabinoid receptors 1 and 2 (CB1 and CB2), which have been shown to influence adult neurogenesis [92, 93]. Previous studies have also demonstrated the dose-dependent beneficial effects of CBD on the generation, survival, and maturation of newly born neurons in the hippocampus [94–97]. Furthermore, CBD treatment has been found to elevate the brain-derived neurotrophic factor concentration, a critical factor capable of enhancing hippocampal neurogenesis [98–100]. Moreover, CBD can inhibit glycogen synthase kinase-3β, a negative regulator involved in maintaining hippocampal neurogenesis through the Wnt/β-catenin pathway [44, 45]. Therefore, multiple mechanisms may contribute to enhanced neurogenesis observed in GWI rats as a result of CBD administration. However, further studies are required to determine the specific mechanisms underlying CBD-mediated increased neurogenesis in GWI.

Functional impacts of CBD-mediated reduced oxidative stress and neuroinflammation and increased neurogenesis in the hippocampus

Our findings also provide evidence supporting the efficacy of CBD in improving cognitive and mood functions, as well as reducing hyperalgesia in a rat model of chronic GWI. Notably, improvements were observed in hippocampus-dependent cognitive tasks such as object location memory and pattern separation function [55, 56]. Additionally, CBD treatment resulted in enhanced associative recognition memory, a cognitive task that relies on the integrity of multiple brain regions including the medial prefrontal cortex, hippocampus, and perirhinal cortex [101]. Furthermore, CBD administration reversed anhedonia and mitigated hyperalgesia in GWI rats. The amelioration of cognitive and mood impairments following CBD treatment is likely attributed to its antioxidant properties, anti-inflammatory effects, and promotion of neurogenesis. This hypothesis is supported by previous studies demonstrating consistent cognitive and mood deficits associated with conditions characterized by increased oxidative stress, chronic neuroinflammation, and reduced hippocampal neurogenesis [28, 30, 31, 102, 103]. Moreover, since no significant differences were observed in body weights between the GWI-VEH and GWI-CBD groups during the 16-week treatment period, it is unlikely that the alleviation of anhedonia in the GWI-CBD group compared to the GWI-VEH group can be attributed to calorie-seeking behavior. Instead, this finding suggests that anhedonia relief may be due to reductions in neuroinflammation along with maintenance of higher levels of hippocampal neurogenesis. Persistent oxidative stress can lead to neuroinflammation, while chronic neuroinflammation can contribute to continuous oxidative stress through increased release of reactive oxygen species. While either oxidative stress or neuroinflammation alone can induce cognitive and mood impairments, their combined presence, as seen in chronic GWI, exacerbates their detrimental effects on cognitive and mood function.

The hyperalgesia observed in GWI rats in this study is consistent with several investigations demonstrating that exposure to GWI-related chemicals, such as PB, PER, and chlorpyrifos, can lead to chronic pain in rodent models of GWI [104–108]. Notably, studies have indicated that the presence or absence of PB during the exposure period determined whether clinical signs of chronic pain would endure. Changes in the expression of voltage-dependent sodium channel 1.9 and transient receptor potential vanilloid-1 (TRPV1) and nociceptor excitability are associated with chronic pain related to PB exposure [106]. Numerous clinical and preclinical studies have shown that CBD can alleviate pain through its antihyperalgesic effects, primarily by activating TRPV1 [109–112]. Additionally, a study has demonstrated the ability of CBD to increase the withdrawal threshold to mechanical stimulation in neuropathic rats via TRPV1 receptors [113]. This study provides evidence that long-term CBD treatment in GWI rats can elevate paw withdrawal thresholds, supporting the antihyperalgesic effect of CBD. However, further research is necessary to determine whether increased TRPV1 expression underlies the alleviation of hyperalgesia mediated by CBD in GWI rats.

Ethics approval and consent to participate

The animal care and experimental procedures were conducted per the animal protocol approved by the Institutional Animal Care and the Use Committee (IACUC) of Texas A&M University School of Medicine (Animal Use Protocol number: 2022-0166D).

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.