Adult male C57BL/6J mice (14 weeks old) were purchased from Liaoning Changsheng Biotechnology co., Ltd. (Benxi, China). Benzene was provided by Sigma-Aldrich. Hydroquinone was provided by Fluka. The blood routine test kit was provided by Drew Scientific (Miami Lakes, USA), and the kits for detecting triglyceride (TG), total cholesterol (TC), non-esterified fatty acid (NEFA), high-density lipoprotein cholesterol (HDL-C) and low-density lipoprotein cholesterol (LDL-C) were obtained from Nanjing Jiancheng Bioengineering Institute (Nanjing, China). Leptin and adiponectin ELISA kits were provided by Mlbio (Shanghai, China).

The adult male C57BL/6J mice were housed in a specific pathogen-free (SPF) animal room in a 12 h dark/light cycle at 20 ± 3°C and 50% humidity. Mice were allowed free access to food and water. The mice were exposed to various doses of benzene and each group was housed separately in polypropylene cages with a maximum of five mice in each cage. Benzene exposure manner and doses were designed according to our previous study (21). Briefly, the mice were randomly divided into four groups (n = 11-12), and benzene was diluted in corn oil. Mice were administrated with benzene by oral gavage at doses of 0, 1, 10 and 100 mg/kg for a consecutive 4 weeks (6 times/week). The body weight and food intake of mice were monitored daily. This study was conducted in strict accordance with the guidelines for the care and use of laboratory animals issued by the State Science and Technology Commission of the People’s Republic of China. The experimental scheme was approved by the Laboratory Animal Ethics Committee, School of Public Health.

Urinary SPMA detection was performed according to our previous protocol (21). Briefly, urine samples were collected using metabolic cages after the last gavage. Then after centrifugation at 3000 rpm for 5 min, the supernatant was collected for detecting the SPMA level by liquid chromatography/electrospray tandem mass spectrometry (LC-MS/MS) (Agilent, Santa Clara, USA). Urinary levels of SPMA were normalized to creatinine, which was measured by the creatinine detection kit (Nanjing Jiancheng Bioengineering Institute, Nanjing, China).

Blood was collected by inferior vena cava puncture. A blood routine test was conducted by an automatic blood cell analyzer according to our previous protocol (21). According to our previous study (21) and research conducted by Lan etal. (22). (23), cell counts of white blood cells (WBC), neutrophils, lymphocytes and monocytes showed high sensitivity to benzene-induced hematotoxicity. Therefore, in this study, these four indexes were tested preferentially to reflect the hematotoxicity of benzene. For biochemical analysis, plasma levels of TG, TC, HDL-C, LDL-C and NEFA were measured using commercial kits, respectively.

The body composition of mice was measured by the Niumag small animal body composition analyzer (Suzhou Niumag Analytical Instrument Corporation, Suzhou, China), according to the manufacturer’s guidelines. Then mice were anesthetized with pentobarbital sodium, and diverse adipose depots were carefully dissected and weighed, including two white adipose tissues (WATs) and one brown adipose tissue (BAT). Epididymal and inguinal subcutaneous adipose depots were representative WATs, while interscapular brown fat was dissected as representative BAT. Then, we calculated the fat content of each adipose depot (fat content = fat pad weight/body weight × 100%) and WAT fat content (the sum of the above two WAT depots) for each mouse.

The distal adipose tissue of the epididymis was fixed with 4% paraformaldehyde (PFA). Then tissues were sectioned after paraffin embedding and then stained with hematoxylin and eosin (H&E) (24). All tissue sections were observed under a biological microscope and photographed. The adipocytes of adipose tissue were analyzed using Image J software (National Institutes of Health, Bethesda, USA).

In vitro experiments were conducted using adipocytes differentiated from preadipocyte 3T3-L1. The differentiation protocol was according to Hsu’s report with minor modifications (25). Briefly, two days post confluence the 3T3-L1 cells were incubated in adipogenesis-inducing medium (AIM) cocktails (DMEM medium containing 1 μmol/L dexamethasone, 0.5 mM IBMX, 10 μg/mL insulin, 1% penicillin, 1% streptomycin and 10% FBS) for 3 days. Next, cells were cultured in adipogenesis maintaining medium (AMM) (DMEM medium containing 10 μg/mL insulin, 1% penicillin, 1% streptomycin and 10% FBS) for 6 days with the medium changed every 3 days. Then cells were maintained in DMEM (1% penicillin, 1% streptomycin and 10% FBS) for another 2 days. Fully differentiation is usually achieved by day 11 (ID11). Then mature adipocytes were exposed to HQ at 0, 1, 5 and 25 μM for 24 hours, respectively. The doses of HQ treatment were designed according to our previous study and cytotoxicity was evaluated by trypan blue staining (26). The levels of intracellular lipid and released-TG were determined to reflect the lipid accumulation and lipolysis activity, respectively. The main genes related to lipid metabolism were also detected. The endocrine function of adipocytes was assessed by examining the levels of leptin and adiponectin at both mRNA and protein levels.

The plasma samples were centrifuged at 3000 rpm and 4°C for 20 min, then the supernatants were taken for measurement of blood lipids and adipokines according to the manuals of commercial kits. Cell culture medium was collected and centrifuged at 3000 rpm and 4°C for 20 min, then the supernatant was taken for measurement of adipokines according to the manuals of commercial kits.

The mRNA levels of adipokines and lipometabolism-related genes were detected. Total RNA was isolated from epididymal fat tissues or adipocytes with TRIzol Reagent (Invitrogen, Carlsbad, USA). Then the Takara primescript™ RT reagent (Takara Biotechnology Co., Ltd., Dalian, China) was used to reverse transcribe 1 ng RNA. With Quantum Studio™ real-time PCR software, SYBR Green (TOYOBO Biotechnology Co., Ltd., Shanghai, China) was used for RT-qPCR. The target gene primer pairs were synthesized by Tsingke Biotechnology Co., Ltd. (Beijing, China). The primer sequence was shown in Supplementary Table 1 (27–29). The 2^−ΔΔCt method was used to determine the relative gene expression (30).

Statistical analyses were performed in SPSS 25.0 software (IBM, Armonk, USA). All data were presented by mean ± standard deviation (SD). One-way analysis of variance (ANOVA) was used to compare the differences between groups. LSD multiple comparison test was used for pairwise comparison, and polynomial contrast procedure was conducted to analyze the changing trend. Pearson correlation coefficient was used to analyze the relationship between fat content or plasma adipokine levels, and peripheral blood leukocyte parameters. All tests were performed by a two-sided test, p < 0.05 was considered statistically significant.

Adult male C57BL/6J mice were exposed to benzene at doses of 0, 1, 10 and 100 mg/kg bw for four weeks. During the period of benzene treatment, no obvious changes in food intake and body weight were observed ( Figures 1A, B ). At the end of the benzene-treatment period, the urinary SPMA level of mice, corrected by creatinine, was increased in a dose-dependent manner (p_trend < 0.001, Figure 1C ). The urinary SPMA levels of 1 and 10 mg/kg groups were equivalent to that of adult male mice exposed to 1 and 10 ppm benzene for 6h/d, respectively (data not shown). Findings indicated that the cell count of WBC, neutrophils and lymphocytes decreased in a dose-dependent manner (all p_trend < 0.05, Figure 1D ). Exactly, compared with the control group, the WBC count of 1, 10 and 100 mg/kg group was decreased by 41.0% (p < 0.05), 38.4% (p < 0.05) and 66.7% (p < 0.001), respectively ( Figure 1D ). While the cell counts of neutrophils and lymphocytes showed significant decrease in 1 and 100 mg/kg (all p < 0.05, Figure 1D ). These results were consistent with our previous study (21), indicating that we successfully established a benzene-induced hematotoxicity model in mice. Subsequently, the effect of benzene exposure on adipose tissue was observed in these mice.

In our study, the plasma levels of TG, TC, HDL-C, LDL-C and NEFA were also detected. Findings showed no significant changes in blood TG, HDL-C and LDL-C levels ( Table 1 ). However, the plasma TC level decreased from 1 mg/kg dose group (p _trend < 0.05), and TC content in 100 mg/kg group was 14.3% lower than that of the control group (p < 0.05). More interestingly, the plasma NEFA level showed a significant upward trend from 1 mg/kg (p _trend < 0.05), with 20.6% higher in 100 mg/kg group than that of the control group (p < 0.05). As it is well-known that adipose tissue is the major source of NEFA, the increased blood NEFA level may suggest increased lipolysis upon benzene exposure. These results indicate that benzene exposure can induce dyslipidemia.

To explore the effect of benzene exposure on body fat content, a non-invasive body composition analysis was firstly conducted. The results showed obvious alterations in body composition after benzene exposure, with a significant decrease in body fat content. Compared with the control group, average body fat content in 1, 10 and 100 mg/kg groups decreased by 27.15%, 25.74% and 27.34% (all p < 0.05), respectively ( Figure 2A ). Meanwhile, higher content of lean meat in 1 and 100 mg/kg groups (both p < 0.05, Figure 2B ) and lower content of free water in all three benzene exposed groups compared with the control group (all p < 0.05, Figure 2C ) were observed. These results indicate that benzene exposure can alter the body composition of mice and dramatically decrease body fat content even in the 1 mg/kg group.

Based on the significant decrease in total body fat content upon benzene exposure, we further investigated which adipose tissue subtype is more sensitive to benzene toxicity. The following results indicated that WAT may be more vulnerable than BAT. Significant decreases in various subtypes of WAT fat content were observed even at the lowest dose group, 1 mg/kg ( Figures 3A–C ), while the content of interscapular brown adipose tissue (iBAT) only decreased in the 10 mg/kg group (20.9% lower than the control group, p < 0.05, Figure 3D ). Compared with the control group, the eWAT fat content in 1 mg/kg, 10 mg/kg and 100 mg/kg groups decreased by 34.4%, 31.1% and 31.7% ( Figure 3B ), and the inguinal white adipose tissue (iWAT) fat content in 1 mg/kg, 10 mg/kg and 100 mg/kg groups decreased by 22.6%, 22.9% and 23.4%, respectively (all p < 0.05, Figure 3C ). Total WAT fat content was calculated by adding the content of the above two white adipose depots, which was 29.8%, 27.9% and 28.4% lower than the control group in 1 mg/kg, 10 mg/kg and 100 mg/kg groups (all p < 0.05, Figure 3A ), respectively. These findings indicate that the white subtype of adipose tissue is more sensitive to benzene-induced toxicity, even at the lowest concentration of 1 mg/kg, which may account for the loss of body fat after benzene exposure.

Given the adverse effect of benzene exposure on WAT, we next analyzed the pathological changes of WAT. Histological analysis was conducted using the eWAT, and shrinking adipocytes were observed ( Figures 4A, B ). As shown in Figure 4B , the area of adipocytes in 1 mg/kg, 10 mg/kg and 100 mg/kg group was 26.5% (p < 0.05), 35.8% (p < 0.001) and 36.2% (p < 0.001) less than that of the control group, respectively. For further information, we analyzed the adipocyte composition of eWAT. Findings indicated that despite benzene exposure having no obvious effect on the total number of adipocytes, the percentage of small adipocytes increased in a dose-dependent manner (p _trend < 0.001) while that of large adipocytes decreased from the 1 mg/kg group ( Figure 4C ). Compared with the control group, the percentage of small adipocytes with an area of (3-10) ×10³ μm² in the 1 mg/kg group was 29.9% higher than that of the control group (p < 0.05, Figure 4C ). Meanwhile, the percentage of the large adipocytes with an area > 20×10³ μm² in the 1 mg/kg group decreased by 63.4% when compared to the control group (p < 0.05, Figure 4C ). Similarly, increasing in large adipocyte percentage and decreasing in small adipocyte percentage, were observed in 10 mg/kg and 100 mg/kg groups, respectively (both p < 0.05, Figure 4C ). In summary, these results exhibited that although benzene exposure has no impact on the total number of adipocytes in eWAT, a considerable number of adipocytes became smaller upon benzene exposure. All these findings suggest that an imbalance of lipid homeostasis may occur after benzene exposure, and the shrinking adipocytes may explain why benzene exposure decreased the WAT fat content in mice even at the lowest dose.

Based on the above findings, we further investigated the expression levels of critical genes involved in pivotal lipid metabolism signaling. Peroxisome proliferator-activated nuclear receptor gamma (PPARγ) is abundantly expressed in adipose tissue and plays a key role in the regulation of adipogenesis and lipometabolism, especially in lipid homeostasis regulation (31, 32). In this study, PPARγ mRNA expression was significantly down-regulated in 1 mg/kg (p < 0.05) and 10 mg/kg groups (p < 0.001, Figure 5A ). Another three critical genes involved in lipogenesis, Cd36, Tcf71 and Zfp43, were all down-regulated in 1 mg/kg,10 mg/kg and 100 mg/kg groups (all p < 0.05, Figures 5C–E ). On the other hand, three pivotal genes involved in lipolysis were also detected. Significant down-regulation of LPL and Lipe2 mRNA expression was observed in 1 mg/kg, 10 mg/kg and 100 mg/kg groups (all p < 0.05, Figures 5G, H ). These results showed that both lipid synthesis and lipid lysis signalings were attacked by benzene exposure, and the disturbance of lipid metabolism signalings may explain why the adipocytes become smaller in benzene exposed groups.

In our study, mature adipocytes were obtained from 3T3-L1 cells. The findings indicated that up to ID11 more than 95% of cells were fully differentiated into adipocytes ( Figure 6A ). Trypan blue staining showed that 24h-HQ-exposure had no obvious cytotoxicity (data not shown) on adipocytes even at 25μM. However, the lipid content decreased significantly in HQ-treated cells ( Figure 6B ), and the level of TG released from cells into the medium increased in a dose-dependent manner (p_trend < 0.05, Figure 6C ). These results indicated that HQ exposure might increase lipolysis and subsequently reduce the lipid content of adipocytes, which was consistent with our in vivo findings.

The RT-qPCR results displayed similar changes in PPARγ mRNA expression to that of in vivo experiment. Exactly, PPARγ mRNA level was significantly down-regulated in 1 μM (p < 0.05), 5 μM (p < 0.05) and 25 μM groups (p < 0.001, Figure 7A ). Four crucial genes related to lipogenesis were also detected, among which Fasn was down-regulated in all HQ-exposed groups (all p < 0.05, Figure 7B ). However, Cd36 mRNA expression was up-regulated in 1 μM group (p < 0.05) and down-regulated in 5 μM (p < 0.05) and 25 μM groups (p < 0.001, Figure 7C ). In addition, no significant change was observed in Tcf7l1 and Zfp423 ( Figures 7D, E ). Three pivotal genes involved in lipolysis were also detected. Significant down-regulation of Plin1 mRNA expression was observed in all benzene-exposed groups (all p < 0.001, Figure 7F ), while the mRNA expression of LPL (both p < 0.05, Figure 7G ) and Lipe2 (both p < 0.001, Figure 7H ) was significantly down-regulated in 5 μM and 25 μM groups. The results showed that HQ exposure could disturb the expression of critical lipometabolism genes. All these findings suggest that the intermediate metabolite of benzene disturbs the lipid homeostasis of adipocytes, which is in accordant with the findings of in vivo experiment.

The above findings indicated that benzene exposure can induce obvious adverse effects on WAT. Since WAT has important roles in regulating the function of distant organs or related diseases in an endocrine manner, we further evaluated the effect of benzene exposure on the endocrine function of WAT. The biomarkers for the endocrine activity of adipose tissues, leptin and adiponectin, were detected. Results showed that the mRNA expression level of adiponectin decreased significantly in 10 mg/kg group compared with the control group (p < 0.001, Figure 8A ), while no significant change was observed in plasma adiponectin level ( Figure 8B ). By contrast, leptin exhibited higher sensitivity. Compared with the control group, leptin mRNA expression in 1 mg/kg, 10 mg/kg and 100 mg/kg group decreased by 79.20% (p < 0.05), 90.83% (p < 0.001) and 42.00% ( Figure 8A ), respectively. In addition, when compared to the control group, the plasma leptin level in 10 mg/kg and 100 mg/kg (both p < 0.05) groups also showed significant decrease ( Figure 8B ). Similar findings were obtained from in vitro experiments. Compared with adiponectin, the biomarker leptin showed better sensitivity. The leptin mRNA expression was down-regulated in adipocytes exposed to HQ at 1 μM (p < 0.05), 5 μM (p < 0.05) and 25 μM groups (p < 0.001, Figure 8C ), while significant decrease in culture medium content of leptin was observed from 5 μM group (p < 0.05, Figure 8D ). Together with the in vivo findings, these results indicate that exposure to benzene or its toxic metabolite HQ disturbs the endocrine function of white adipocytes.

As WAT is reported to be involved in hematological diseases through endocrine, we next analyzed the relationship between fat content (body fat content, iBAT fat content, eWAT fat content, iWAT fat content and WAT fat content) and plasma adipokines (leptin and adiponectin) levels with peripheral blood leukocyte parameters (WBC, neutrophil, lymphocyte and monocyte), respectively. As shown in Table 2 , there were significant positive relationships between total body fat content and WBC, as well as neutrophil and monocyte count (all p < 0.05). And the WAT fat content, especially eWAT fat content, also showed positive correlations with WBC, neutrophil and monocyte count (p < 0.05). In addition, significant positive correlations between plasma leptin levels and WBC counts, as well as lymphocyte counts were also observed (both p < 0.05) ( Table 3 ). The findings suggest that WAT, especially the visceral WAT, may affect the blood routine during benzene exposure, and leptin may play an important role in this process.