Permethrin exposure primes neuroinflammatory stress response to drive depression-like behavior through microglial activation in a mouse model of Gulf War Illness

Abstract

Gulf War Illness (GWI) is a chronic multisymptom disorder that affects approximately 25-32% of Gulf War veterans and is characterized by a number of symptoms such as cognitive impairment, psychiatric disturbances, chronic fatigue and gastrointestinal distress, among others. While the exact etiology of GWI is unknown, it is believed to have been caused by toxic exposures encountered during deployment in combination with other factors such as stress. In the present study we sought to evaluate the hypothesis that exposure to the toxin permethrin could prime neuroinflammatory stress response and elicit psychiatric symptoms associated with GWI. Specifically, we developed a mouse model of GWI, to evaluate the effects of chronic permethrin exposure followed by unpredictable stress. We found that subjecting mice to 14 days of chronic permethrin exposure followed by 7 days of unpredictable stress resulted in the development of depression-like behavior. This behavioral change coincided with distinct alterations in the microglia phenotype, indicating microglial activation in the hippocampus. We revealed that blocking microglial activation through Gi inhibitory DREADD receptors in microglia effectively prevented the behavioral change associated with permethrin and stress exposure. To elucidate the transcriptional networks impacted within distinct microglia populations linked to depression-like behavior in mice exposed to both permethrin and stress, we conducted a single-cell RNA sequencing analysis using 21,566 single nuclei collected from the hippocampus of mice. For bioinformatics, UniCell Deconvolve was a pre-trained, interpretable, deep learning model used to deconvolve cell type fractions and predict cell identity across spatial datasets. Our bioinformatics analysis identified significant alterations in permethrin exposure followed by stress-associated microglia population, notably pathways related to neuronal development, neuronal communication, and neuronal morphogenesis, all of which are associated with neural synaptic plasticity. Additionally, we observed permethrin exposure followed by stress-mediated changes in signal transduction, including modulation of chemical synaptic transmission, regulation of neurotransmitter receptors, and regulation of postsynaptic neurotransmitter receptor activity, a known contributor to the pathophysiology of depression in a subset of the hippocampal pyramidal neurons in CA3 subregions. Our findings tentatively suggest that permethrin may prime microglia towards a state of inflammatory activation that can be triggered by psychological stressors, resulting in depression-like behavior and alterations of neural plasticity. These findings underscore the significance of synergistic interactions between multi-causal factors associated with GWI.

Fig. 1

Brain and Plasma Concentrations over time. (A) Plasma levels of trans-Permethrin reached a maximum concentration at 24 h after acute injection (I.P.) of 200 mg/kg of permethrin. (B) Plasma levels of cis-Permethrin also reached a maximum concentration at 24 h. (C) Similarly, plasma levels of the permethrin metabolite 3-PBA reached a maximum concentration 24 h after injection. (D) Brain tissue levels of trans-Permethrin reached a maximum concentration at 2 h after acute injection. (E) Brain tissue levels of cis-Permethrin also reached a maximum concentration at 2 h after acute. (F) Brain tissue levels of the permethrin metabolite 3-PBA reached a maximum concentration at 24 h after acute injection. In each group, there were 5 mice, and for brain tissue, 2 replicates (separate brain hemispheres) were utilized per time point. The data are presented as means ± SEM

Fig. 2

Permethrin primes stress response to induce depressive-like behavior. (A) Experimental scheme. (B-C) Exposure to permethrin followed by stress induced depressive (C) but not anxiety-like (B, right panel) behaviors as measured via forced swim and open field tests respectively. (B, left panel) Bar graphs showed locomotion activity with no alterations among all experimental groups. (D) Cx3Cr1^CreeEr/hM4Di-DREADD mice expressing the Gi receptor on microglia were used to demonstrate that microglial inhibition via selective ligand JHU 37,160 is sufficient to prevent behavioral changes resulting from permethrin and stress exposure. Statistical analyses were performed using Two-Way ANOVA (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.001, compared to permethrin exposure followed by stress). In each group, there were13 mice and each dot represents an individual mouse. Data are expressed as the means ± SEM

Fig. 3

Alterations in microglial morphology induced by permethrin and stress in brain, specifically in the hippocampus prefrontal cortex. (A) Representative immunofluorescence images of IBA-1+ (red) cells in hippocampus. (B-D) Bar graphs represented soma volume (B), branch length (C) and number of terminal points (D) per microglia in hippocampus. (E-F) Sholl analysis showed that microglia in the group exposed to both permethrin and stress displayed reduced branching complexity compared to other experimental groups. (E) Link graph showed the quantification of the number of intersections at increasing radii, with measurements taken at 5 μm intervals. (F) Bar graph represented the Area Under the Curve (AUC) of the line graphs shown in (E). (G) Representative immunofluorescence images of IBA-1 + cells in prefrontal cortex. (H-J) Sholl analysis showed that microglia in the prefrontal cortex exhibited no apparent morphological changes. Bar graphs represented soma volume (H), branch length (I) and number of terminal points (J) per microglia in prefrontal cortex. (K) Link graph showed the quantification of the number of intersections in the prefrontal cortex. (L) Bar graph represented the AUC of the line graphs. The area outlined with a white box in the top panel is magnified in the middle panel. The bottom panel shows representative images of three-dimensional (3D) reconstructions of microglia from the middle panel Scale bars indicate 50 (top panel) and 5 (middle and bottom panel) um. Statistical analyses were performed using Two-Way ANOVA (*p < 0.05, **p < 0.01, compared to permethrin exposure followed by stress. In each group, there were 4-5mice and each dot represents an individual mouse. Data are expressed as the means ± SEM

Fig. 4

Characterization of GWI-associated brain cell population through single-cell sequencing. (A) UCDSelect was used to project annotations from a reference mouse cortex / hippocampus atlas onto novel dataset. Microglial cluster was confirmed by expression of canonical marker genes, including inpp5d, Tgfbr1, Apbb1ip. (B) Cell density plots for each experimental group depict a uniform distribution of cell density, with increasing intensity of red indicating specificity to the respective condition. (C) Differential expression analysis against predicted clusters to identify conserved markers specific to each putative cell annotation

Fig. 5

The microglia population exhibits significant enrichment in pathways linked to axon development, calcium ion transport, and neurotransmission. (A and B) Bubble plots of Gene Ontology (GO) category enrichment results in microglia cell populations for biological process (A) and molecular function (B). The color of the points reflects the − log10​ adjust p-value, with more significant p-values appearing as more intensely colored points. The size of each point corresponds to the percentage of gene sets within each GO category, with larger points indicating a higher percentage. (C) Venn diagram showing the number of overlapping significantly differentially expressed genes (DEGs) specifically enriched under different conditions: vehicle exposure followed by stress (green), permethrin exposure followed by no stress (red), and permethrin exposure followed by stress (pupple) (left panel). The middle panel shows the number of overlapping DEGs after sorting by an absolute z-score greater than 8.8 in the permethrin exposure followed by stress condition, and then how many genes are shared across the different experimental conditions. The right panel shows the log2 fold change values of selected DEGs expressed in the permethrin exposure followed by stress and permethrin exposure followed by no stress conditions

Fig. 6

CA3 neuronal cell population exhibits significant enrichment in pathways linked to synaptic plasticity. (A) Venn diagram showing the number of overlapping significantly differentially expressed genes (DEGs) in DG, CA1, CA2, and CA3 regions specifically enriched under different conditions: vehicle exposure followed by stress (green), permethrin exposure followed by no stress (red), and permethrin exposure followed by stress (purple) (left panel). (B) Bar graphs represent the number of total genes altered and the proportion of genes represented in the Venn diagram showing the overlapping significantly DEGs within DG, CA1, CA2, and CA3. (C and D) Bubble plots of GO category enrichment results in neuronal cell populations in CA3 for biological process (C) and molecular function (D). The color of the points reflects the − log10 adjust p-value, with more significant p-values appearing as more intensely colored points. The size of each point corresponds to the percentage of gene sets within each GO category, with larger points indicating a higher percentage

Fig. 7

Olink proteomic analysis in the GWI mouse model. (A-F) Box graphs represented significant changes in protein expression levels of TGFα (A), Ahr (B), IL-1β (C), RGMa (D), Snap29 (E), and Ddah1 (F) in the hippocampus across all experimental groups. Statistical analyses were performed using Two-Way ANOVA (*p < 0.05; **p < 0.01, ***p < 0.001; ****p < 0.0001; compared to the permethrin followed by stress group, n = 7–9 mice for each group). Data are expressed as the means ± SEM