Microglia Mediate Metabolic Dysfunction From Common Air Pollutants Through NF-κB Signaling Free

Numerous studies have shown a positive association between long-term ambient air pollution exposures and increased risk of type 2 diabetes (T2D) (1). Among the common air pollutants, volatile organic compounds (VOCs) stand out because of their widespread presence indoors and outdoors, mainly emitted during vehicle operation and fuel production and by consumer products and exposure to smoke (2–5). However, research on the metabolic consequences of exposure to these nonoccupational VOCs remains limited (6–10).

Mutual relationships between metabolic abnormalities and the immune system were reported in obesity, T2D, and cardiovascular disease (11). Accordingly, human studies show an association between exposure to various VOCs, including benzene, and increased systemic inflammation (12–14). Previously, we reported that chronic exposure to benzene at levels mimicking smoking led to insulin resistance, hyperglycemia, and increased expression of inflammatory genes in the liver and the hypothalamus, specifically in male mice (15). These findings highlight the potential implications of benzene exposure in the development of metabolic diseases.

The hypothalamus receives input from various hormonal, environmental, and nutritional signals, which regulate peripheral glucose metabolism and energy balance (16). In the context of metabolic disorders, neuroinflammatory responses characterized by dysregulated glial cells (microglia and astrocytes) in the hypothalamus are well described (17,18). It was suggested that hypothalamic inflammation precedes, and mechanistically contributes to, the whole-body metabolic imbalance (17). Microglia, the central nervous system (CNS)-resident immune cells, exhibit sensitivity to stressors and can be activated by inhaled components of urban air pollution through direct and indirect pathways (19,20). This stress drives microglia to release interleukin 1α (IL-1α), tumor necrosis factor α (TNF-α), and C1q, inducing astrocytes to acquire a proinflammatory phenotype (21). We have previously demonstrated that acute exposure of primary glial cells to benzene induced the upregulation of inflammatory IL-1 and TNF-α gene expression in a manner similar to lipopolysaccharide-driven inflammation (15,22). However, the impact of microglial activation in response to VOC exposure on peripheral metabolic control remains unknown. We hypothesized that microglia might serve as the mechanistic link between VOC exposure and metabolic dysfunction.

Our study adopts a comprehensive approach, incorporating continuous real-time glucose monitoring (CGM), extensive hypothalamic and microglial transcriptomics, and innovative global and adult-onset microglia-specific mice models with IKK/NF-κB pathway loss of function. For the first time, our study establishes microglial IKK/NF-κB signaling as a crucial mediator of the adverse metabolic effects induced by VOC exposure.

All mice were on C57BL/6J background and maintained under standard laboratory conditions with standard chow low-fat diet (Research diets). For some studies, males and females were used, or only the male mice were used, as indicated in RESULTS. Additional information on Cx3cr1^Cre, Tmem^119CreER, Cx3cr1^GFPΔIKK, and Tmem^119ERΔIKK mice breeding and tamoxifen injections is provided in Supplementary Methods.

The mice in inhalation chambers using FlexStream automated Perm Tube System (KIN-TEK Analytical, Inc.) were exposed to benzene concentration of 50 ppm for 6 h/day (acute) or up to 4 weeks (chronic) as described before (15). Additional information is provided in Supplementary Methods.

To assess VO₂ consumption, VCO₂ production, respiratory exchange ratio (RER), and heat production, mice were placed in PhenoMaster metabolic cages (PhenoMaster, TSE system). Animals were individually housed during 12-h dark cycle (6:00 p.m. to 6:00 a.m.). The mice were acclimatized for 48 h, and data were collected for 72 h, while food and water were provided ad libitum.

The commercial Libre 2 glucose monitor device (Abbott Diabetes Care) was engaged to obtain the assembled glucose sensor. CGM provides blood glucose readings automatically every 15 min. Mice were anesthetized by inhalation with 2.5% isoflurane; the sensor was inserted using a 23-gauge needle to pierce the shaved caudal surface of the skin. Animals were allowed 3 days for acclimatization. Subsequently, the sensor was scanned periodically using the reader to obtain the data. For the glucose tolerance test (GTT), mice were fasted for 6 h (5:00–11:00 a.m.) and intraperitoneally injected with d-glucose at 2 g/kg⋅BW. Blood glucose levels were measured as before (23).

Mice were anesthetized (i.p.) with avertin and transcardially perfused with PBS (pH 7.5) followed by 4% paraformaldehyde. Brains were postfixed, sunk in 30% sucrose, frozen in optimal cutting temperature medium, and sectioned coronally (30 µm) using a Leica 3050S cryostat. Four series were collected and processed for immunohistochemistry as previously described (23). Additional information is provided in Supplementary Methods.

Hypothalamus samples were lysed with 0.75 mL 2-mercaptoethanol added lysis buffer (PureLink RNA Mini Kit). Samples were centrifuged at room temperature. RNA was eluted using 15–20 μL RNase-free water. Additional information on microglia isolation, RNA extraction, and sequencing is provided in Supplementary Methods. All RNA sequencing (RNA-seq) data are available at the Sequence Read Archive at National Center for Biotechnology Information under accession number PRJNA1035111.

Statistical analyses for differentially expressed mRNAs were performed pairwise using EdgeR in the software R (3.2.2). For all other experiments, results are expressed as the mean ± SE and were analyzed using Statistica software (version 10). Graphs were generated using GraphPad Prism software. An analysis of t test was used when comparing only treatment versus control; variance (ANOVA) with repeated measurements was used to analyze metabolic data. Other parameters were analyzed by two-way ANOVA. All data were further analyzed with Newman-Keuls post hoc analysis. The level of significance (α) was set at 5%.

All data sets generated and/or analyzed during the current study are available upon request from the lead contact. All RNA sequencing (RNA-seq) data are available at the Sequence Read Archive at National Center for Biotechnology Information under accession number PRJNA1035111.

We conducted a rigorous meta-analysis of clinical studies to explore the connections between environmental benzene exposure and metabolic health outcomes. This analysis focused on studies that measured urinary benzene metabolites (particularly trans, transmuconic acid [t,t-MA]) and provided adjusted odds ratios (OR) concerning metabolic disease risk. Our comprehensive literature search employed the PubMed research database and included key terms such as (“volatile organic compound” OR “benzene exposure”) AND (“metabolic disease” OR “insulin resistance”) in the abstract or title. This search yielded a total of 128 articles. We excluded 68 articles because of the absence of relevant metabolic measurements, 38 articles for being methodological, and 14 articles for using animals. Eight studies met our stringent criteria including measurements of urinary t,t-MA levels with readjusted OR for metabolic disease risk, using parameters such as blood glucose, insulin, or cholesterol levels. The chosen studies encompassed 26,313 subjects, spanning various demographics including pregnant women, children, young adults, adults, and elderly individuals (6–8,10,24–27). As indicated in the forest plot (Fig. 1), we found significant effects of environmental benzene exposure, at levels pertinent to highly polluted urban settings, on the risk of developing metabolic disease. Notably, higher levels of benzene exposure were directly associated with an increased risk of insulin resistance, with a calculated meta OR of 1.47, 95% CI (1.33, 1.63), P < 0.001. Our meta-analysis provides the most robust evidence that benzene exposure at environmental levels in humans is directly associated with metabolic disease.

In a previous study, we demonstrated that a 4-week exposure to benzene at levels mimicking smoking triggered hyperglycemia and hyperinsulinemia, specifically in male mice without affecting body weight (15). To follow early metabolic changes, we employed a CGM system, enabling real-time tracking of blood glucose levels. After CGM implantation, male mice underwent a 3-day acclimatization period (Fig. 2A). We continuously measured blood glucose levels for 7 days. No significant changes in blood glucose levels were detected at the initial 3 days of exposure (Supplementary Fig. 1A–C). However, starting from day 4 of exposure, a significant elevation in blood glucose levels was observed in mice exposed to benzene (P < 0.001) (Fig. 2B and Supplementary Fig. 1D–F). After 7 days of exposure, a substantial increase of about 20% change (P < 0.01) in blood glucose levels compared with day 1 was observed (Fig. 2C). Moreover, we observed significant reductions in VO₂ consumption, VCO₂ production, RER, and heat production measured during the dark cycle, after the initial 6 h of benzene exposure (Fig. 2D–G). No differences in food intake or activity levels were detected (Fig. 2H and I). The alterations in energy homeostasis remained evident even 7 days after the 4-week exposure in male mice (Supplementary Fig. 1G and H), indicating the prolonged effects of benzene exposure. Intriguingly, benzene exposure led to increased serum corticosterone levels (measured after 24 h or on day 4 of exposure) in male mice (Supplementary Fig. 1I). However, the levels of corticosterone were normalized after 4-week exposure, indicating the involvement of other non–stress-related mechanisms to the disruptions in energy homeostasis (Supplementary Fig. 1I). Notably, no differences in energy homeostasis (Supplementary Fig. 2), weight, or glucose tolerance were observed between females exposed to benzene and the control group (15).

To assess the impact of benzene exposure on hypothalamic transcriptome before the onset of metabolic phenotype, we conducted an analysis of differential gene expression in the hypothalamus of male mice acutely exposed to benzene for 24 h, comparing them to filtered air–exposed mice using bulk RNA-seq. The RNA-seq analysis revealed 405 significantly upregulated and 332 downregulated genes in the whole hypothalamus of benzene-treated mice (Fig. 3A). Remarkably, gene ontology (GO) analysis (biological processes) identified the response to insulin as one of the most enriched pathways in response to benzene exposure (Fig. 3B). Among the top genes involved in response to insulin were Sh2b2, Ogt, Klf15, Lpin2, Rab31, Ifg1r, Agt, Pik3ca, Cpeb2, Srsf5, Nucks1, Cry1, Cry2, Slc2a1, Cited1, Pdk2, Sos2, and Ptpn11 (Fig. 3C). Some of these genes have been demonstrated to stimulate insulin signaling, while others are involved in insulin-induced activation of mitogen-activated protein kinases (MAPK) mediated by Ras (28). Moreover, the interaction network of GO terms revealed that, in addition to response to insulin, the top markedly enriched GO terms are cellular response to organic substance, cellular response to endogenous stimulus, and response to nitrogen compound (Fig. 3D), demonstrating a rapid hypothalamic response to chemicals. Building on previous findings that NF-κB signaling increases in response to organic substances in air pollution (29), and elevated levels of Ikbkb and Ikbke in the hypothalamus (15), we cross-referenced NF-κB–related genes from the Mouse Genome Informatics Gene Ontology Project (Supplementary Table 2) with differentially expressed genes (DEGs) from our study. Our functional enrichment analysis identified that these NF-κB interacting genes from our DEGs are associated with the activation of the immune response and the regulation of NF-κB signaling (Fig. 3E and Supplementary Table 3), indicating the involvement of NF-κB signaling in benzene-induced hypothalamic inflammation.

To confirm that insulin signaling is perturbed upon short benzene exposure, male mice were acutely exposed to benzene for 6 h while fasting, and brains were collected after receiving a single dose of insulin (i.p.; 3 international units/kg body weight) or saline injection. In control mice, insulin-stimulated FoxO1 phosphorylation led to an increase in its cytoplasmic translocation in both the arcuate nucleus (ARC) (Fig. 3F and G) and the ventromedial nucleus of the hypothalamus (Supplementary Fig. 3A and B). In contrast, following benzene exposure, there was a significant increase in cytoplasmic FoxO1 levels (P < 0.001), which did not exhibit further responsiveness to insulin (Fig. 3G and Supplementary Fig. 3A and B). Consistent with FoxO1 cytoplasmic localization, pAkt levels were significantly increased upon insulin stimulation in control mice (Supplementary Fig. 3C and D), while benzene-exposed mice, which had increased basal pAkt levels, did not respond to insulin stimulation (Supplementary Fig. 3C and D). A similar response pattern to benzene exposure was also observed with MAPK signaling, a pathway implicated in the development of insulin resistance (30). In control mice, insulin stimulation increased MAPK phosphorylation (Supplementary Fig. 3E–G). However, in mice exposed to benzene, phosphorylated MAPK was significantly elevated as compared with unstimulated control (P < 0.001) and did not further respond to subsequent insulin stimulation (Supplementary Fig. 3E–G). Consistent with our prior findings (15), we found significant alterations in the morphology of astrocytes and microglia within the hypothalamus. We observed an increase in the extension of branches and end points of these cells (P < 0.001) (Fig. 3H–L), suggesting a higher level of activity in response to benzene. There were no significant differences in microglia numbers in the cortex or hippocampal CA3 subregion after the acute exposure (Supplementary Fig. 3H–K).

To test the link between benzene-induced central inflammation and impaired glucose metabolism, we used the fractalkine receptor (Cx3cr1)-driven Cre, widely expressed in immune cells including microglia, on ROSA26-GFP background and crossed them with IKKβ^lox/lox mice (31,32). The expression levels of IKKβ in control (IKK^flox/flox) and Cx3cr1^GFPΔIKK mice were assessed histologically and by quantitative PCR, demonstrating a significant decrease in IKKβ expression in exposed microglia of Cx3cr1^GFPΔIKK mice (Supplementary Fig. 7). In control mice, acute, 6-h exposure to benzene significantly increased TNFα expression levels in microglia, with approximately 80% of GFP⁺ stained hypothalamic microglia showing TNFα production (GFP⁺/TNFα⁺), indicating their inflammatory status (Fig. 4A and B). In contrast, Cx3cr1^GFPΔIKK male mice exposed to benzene had similar numbers of GFP⁺/TNFα⁺ microglia and TNFα gene expression as filtered air control mice (Fig. 4A and Band Supplementary Fig. 4A), and no aberrations in microglia morphology as compared with benzene-exposed controls (Fig. 4D–F), suggesting the absence of microglial activation. As previously, the numbers of microglia were not increased by acute exposure in control animals (15) but were similarly reduced in unexposed or exposed Cx3cr1^GFPΔIKK mice compared with control (Fig. 4C). Accordingly, while benzene-exposed control mice exhibited an increase in GFAP⁺ astrocyte numbers and alterations in astrocytes’ morphology characterized by the numbers and length of GFAP⁺ processes within the hypothalamus, astrocyte in benzene-exposed Cx3cr1^GFPΔIKK male mice was not significantly affected (Supplementary Fig. 4B–G). Finally, mice lacking the IKKβ/NF-κB signaling pathway in immune cells resisted benzene-triggered metabolic impairment. Cx3cr1^GFPΔIKK mice exposed to benzene during the light cycle exhibited similar VO₂ and VCO₂ as filtered air-exposed controls compared with benzene-exposed control mice (Fig. 4G and H). Additionally, while benzene exposure induced glucose intolerance in control mice, Cx3cr1^GFPΔIKK male mice exhibited normal glucose tolerance, after 4 weeks of exposure (Fig. 4I). Further, insulin stimulation promoted similar responses in benzene-exposed Cx3cr1^GFPΔIKK males, indicated by FoxO1 translocation to the cytoplasm, comparable to control Cx3cr1^GFPΔIKK mice (Supplementary Fig. 4H–K), with appropriate insulin-stimulated increase in MAPK phosphorylation (Supplementary Fig. 4L and M).

To specifically determine the role of central inflammatory responses to benzene exposure, we used CD11b⁺ magnetic-activated cell sorting (CD11b-MACS) to isolate microglia from the brains of adult male mice exposed to benzene or filtered air for 24 h and performed bulk RNA-seq analysis (Fig. 5A). The purity of the CD11b-MACS positive microglia was verified by flow cytometry (Supplementary Fig. 5A). The CD11b-MACS positive fraction was composed of ∼80% CD11b⁺ CD45⁺ cells, as compared with 10–20% of the input (Supplementary Fig. 5B) (two-way ANOVA, **P < 0.01). In addition, because CD45 is expressed by nonmicroglial cells, including monocytes and macrophages, we further assessed the purity of samples with a partial deconvolution method, CIBERSORTx (33). CIBERSORTx analysis indicated an average of 2.1% nonmicroglial (astrocytes, neurons, oligodendrocytes) cells per sample (Supplementary Fig. 5). RNA-seq analysis showed 1,531 differentially expressed genes in microglia of benzene-treated mice, where 749 transcripts were significantly elevated and 782 were significantly reduced as compared with control (Fig. 5B). Gene set enrichment analysis revealed that exposure to benzene elicited a significant upregulation of pathways related to chemotaxis, response to nicotine, energy derivation by oxidation of organic compounds, downregulation of negative regulation of interleukin 6 production, and negative regulation of interleukin 1 production in microglia (Fig. 5C). Notably, we found enrichment of genes associated with the positive regulation of I-κ B kinase/NF-κ B signaling and cytokines production (Fig. 5D and Table S4).

To test whether NF-κB pathway activation specifically within microglia is critical for the observed hyperglycemia and insulin resistance, we crossed tamoxifen-inducible Tmem119-Cre^ERT2 mice (34) with IKKβ^lox/lox mice (TMEM119^ERΔIKK). Expression levels of IKKβ were significantly decreased and not detected histologically in TMEM119^ERΔIKK microglia compared with control (Supplementary Fig. 7). The 14-week-old TMEM119^ERΔIKK male mice displayed average body weight and blood glucose levels, comparable to their controls (Supplementary Fig. 6A and B). Remarkably, the TMEM119^ERΔIKK males were entirely resistant to benzene-induced microgliosis, as evidenced by the minimal TNFα production observed in the hypothalamic microglia of TMEM119^ERΔIKK mice (Fig. 6A and B). Furthermore, there were no detectable changes in energy homeostasis parameters in TMEM119^ERΔIKK mice, compared with control animals, whether exposed to air or benzene (Fig. 6C–E). Remarkably, the TMEM119^ERΔIKK males remained protected against benzene-induced hyperglycemia after 4 weeks of benzene exposure while maintaining euglycemia, as indicated by the GTT (Fig. 6F). Furthermore, the elevated expression of inflammatory genes related to NF-κB signaling and insulin-responsive genes detected in the hypothalamic transcriptome of benzene-exposed control mice was mostly restored in TMEM119^ERΔIKK mice, reflecting air-exposed controls (Fig. 6G and H). Consistent with this, benzene exposure increased the expression of TNFα, IL-1, IL-6, and S100b in isolated microglia from control mice but not in TMEM119^ERΔIKK microglia (Supplementary Fig. 6C–F). Thus, the microglia-specific ablation of the NF-κB pathway protects against benzene-induced metabolic disturbances.

We report a meta-analysis, including a wide range of demographics such as the elderly, children, young and mature adults, and pregnant women (6–8,10,24–27), which shows a strong association between levels of benzene urinary metabolites, and indicators of insulin resistance. These data align with recent findings from a cross-sectional study, showing a strong correlation between environmental exposure to mixtures of VOCs indoors and outdoors and the development of insulin resistance within the broader population (35). Importantly, we reveal the mechanisms underlying VOC-induced insulin resistance. Hypothalamic microglia play an early and critical role in these metabolic disturbances following benzene exposure. Accordingly, we demonstrate that activation of the NF-κB pathway in microglia is essential for initiating of neuroinflammatory response. Indeed, inhibition of the NF-κB pathway prevented the metabolic abnormalities triggered by exposure to benzene. Our study highlights the remarkable sensitivity of the hypothalamus to environmental stressors, particularly VOCs. It identifies microglial NF-κB signaling as a promising target for preventing metabolic impairments induced by airborne toxins.

Findings in this study highlight the hypothalamus as a critically sensitive organ, immediately vulnerable to VOC exposure. Specifically, rapid benzene exposure impaired basal and insulin-stimulated FoxO1 cellular translocation in the hypothalamus and led to elevations in hypothalamic pAkt and MAPK signaling, indicating impaired hypothalamic insulin signaling (30). In support, hypothalamic RNA-seq analysis conducted 24 h after exposure to benzene revealed the most enriched GO terms related to responses to insulin and alterations in key biological processes related to immune response. Indeed, the disruptions in hypothalamically controlled energy homeostasis are observable within the first 6 h of exposure. Consistent with that, we demonstrate elevated blood glucose levels on the fourth day of exposure, gradually progressing to severe hyperglycemia. This implies that hypothalamic dysregulation precedes and mechanistically contributes to the whole-body metabolic imbalance. Interestingly, the reductions in energy homeostasis persisted even a week after the cessation of the 4-week exposure, suggesting the involvement of multiple hypothalamic signaling processes (36). Our previous study demonstrated an increase in the expression of hepatic genes associated with gluconeogenesis and lipid metabolism in benzene-exposed male mice; thus we cannot rule out the importance of peripheral organs in contributing to the metabolic imbalance induced by benzene exposure (15).

In the current study and our previous research, we found that female mice resisted benzene-induced metabolic imbalance (15). This resistance could stem from several factors, related to sex hormone differences and enzymatic activities. Females possess higher estrogen levels, which may protect them from metabolic imbalances induced by endocrine disruptors like benzene (37). Furthermore, differential expression of cytochrome P450 can influence the level of benzene-induced metabolic alterations (38). Similarly, benzene exposure rapidly induced the expression of benzene metabolite CYP2E1 in male mice, indicating sex differences in benzene metabolism (15).

VOCs can reach the CNS in just a few minutes through direct transport or via systemic inflammation initiated in the lung tissue immune cells (39). The proximity of the hypothalamus to the third ventricle, a region with permeable barriers, may contribute to its increased vulnerability to exposure (19,20). This is consistent with the rapid transcriptomic changes observed in our study and the reported increase in the hypothalamic benzene metabolite gene Cyp2e1 expression (15). This rapid activation is also confirmed in cultured glial cells (40). While the half-life of Cyp2e1 in glial cells is not well documented, CYP2E1 generally has a half-life of 6 h or less in various peripheral organs, including the brain. It is quickly degraded after exposure cessation (41). This suggests that Cyp2e1 degradation and synthesis rates are comparably rapid in various biological contexts. In male rats, deleting the Cyp2e1 gene prevents diet-induced obesity, fatty liver, and insulin resistance by increasing energy expenditure (42). On the other hand, elevated Cyp2e1 expression can lead to oxidative stress and potentially impact glucocorticoid metabolism, including corticosterone (43). Previously, we reported elevated Cyp2e1 expression in the hypothalamus and liver of males, but not females, after 4 weeks of exposure, correlating with impaired metabolic function only in male mice (15). Importantly, corticosterone levels normalized after the 4-week exposure. Studies directly linking CYP2E1 and corticosterone levels in the context of systemic metabolic dysfunctions are limited. However, increased CYP2E1 might influence corticosterone levels indirectly through its role in oxidative stress and metabolic processes (44). More studies will be needed to assess CYP2E1 levels in various tissues and correlate them with serum corticosterone and systemic metabolic function in control and IKK-deleted microglia mice following benzene exposure.

GO analysis of the isolated microglia following 24 h of benzene exposure indicates a distinct shift toward an inflammatory phenotype, with strong enrichment in genes related to NF-κB signaling. This suggests that benzene exposure can prime microglia, particularly those located in the hypothalamus. In support, changes in hypothalamic glia branching patterns, such as increased ramification, can lead to excessive release of proinflammatory cytokines and reactive oxygen species, contributing to metabolic dysfunction (17). However, inflammatory profile and branching patterns of microglia in other brain regions may also contribute to the metabolic phenotype. Our study specifically focused on the hypothalamus because of its critical role as a central metabolic region in the brain. However, benzene exposure may significantly impact various brain regions and neuroinflammatory responses, potentially affecting cognitive functions, memory, and other neurophysiological processes. Future thorough investigations of microglial and astrocyte status in multiple brain regions, including the cortex and hippocampus, will be necessary.

The generation of Cx3cr1^GFPΔIKK mice, in which IKK was eliminated from systemic immune cells, including microglia, provided further evidence. These mice were protected against alterations in glucose metabolism and energy expenditure induced by exposure. Further, we demonstrated that inducible microglia-specific IKK-knockout mice (TMEM119^ERΔIKK) showed a blunted neuroinflammatory response and no metabolic impairments following benzene exposure. Furthermore, we observed a limited overlap in hypothalamic genes regulated by response to insulin or NF-κB signaling between the control group and TMEM119^ERΔIKK animals exposed to benzene. These findings were supported by previous reports showing that activation of IKKβ/NF-κB in hypothalamic microglia and astrocytes is linked to metabolic imbalance in obese mice (45,46). These data, using the TMEM119^ERΔIKK mice, provide direct evidence for the involvement of NF-κB–induced microglial activation in the benzene-induced peripheral metabolic imbalance.

A recent study suggested that the neuroinflammatory response in microglia provides a protective mechanism against insulin resistance. Using the high-fat diet–fed Cx3cr1^CreER/+::Ikk^fl/fl inducible mice, the authors showed that loss of microglial IKKβ protected the mice from weight gain; however, paradoxically, it worsened glucose tolerance. Conversely, stimulating microglial inflammatory signaling improved glucose tolerance (47). It is possible that microglia response to high-fat diet feeding differs from their response to environmental stressors. Further investigations are essential to elucidate these differences in microglial responses.

In summary, our study provides novel insights into the mechanisms underlying metabolic imbalance upon benzene exposure, highlighting the potential contribution of environmental contaminants to the rising prevalence of chronic metabolic health conditions.

Funding. This study was supported by American Diabetes Association 1-lB-IDF-063, Center for Urban Responses to Environmental Stressors P30ES036084, National Institutes of Environmental Health Sciences R01ES033171, National Institute of Aging RF1AG078170, and Center for Leadership in Environmental Awareness and Research P42ES030991 for M.S. L.K. was further supported by 5T32GM142519-02 and L.S. was supported by 5T32HL120822-09. Services were provided by Genomic Sciences Core of the Oklahoma Nathan Shock Center P30AG050911; the Wayne State University Microscopy, Imaging and Cytometry Resources Core P30CA22453 and R50CA251068-01; and Wayne State University Genomics Core.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. L.K.D., H.S.M.J., L.S., A.L.T.d.S., L.K., S.S., and R.S. carried out the research and reviewed the manuscript. U.K. designed and ran investigation for the CGM study. A.M. performed and analyzed the meta-analysis. M.S. conceptualized the study, acquired funding for the study, and supervised the study. M.S. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Prior Presentation. The abstracts presenting these data were showcased at the 82nd Scientific Sessions of the American Diabetes Association, 3–7 June 2022, New Orleans, LA.