Adult (6–8 weeks) C57BL/6J male mice (Beijing Vital River Laboratory Animal Technology Co., Ltd., Beijing, China) were housed in groups of five, had ad libitum access to food and water and were maintained on a 12:12-h light/dark cycle (lights on from 8:00 a.m. to 8:00 p.m.). Mice were given a single dose of 4 mg/kg lipopolysaccharides (LPS, L2630, Sigma, Saint Louis, Missouri, USA) or same volume of saline via intraperitoneal (i.p.) injections. All husbandry and experimental procedures in this study were approved by the Animal Care and Use Committees of the Shenzhen Institute of Advanced Technology (SIAT), Chinese Academy of Sciences (CAS), China.

Mice plasma samples were collected at the afternoon of the experimental day. Plasma testosterone levels were measured using an Iodine[¹²⁵I] Radio-Immunoassay Assay (KIP1709, DIasource, Louvain-la-Neuve, Belgium). Blood samples from the LPS-treated and control mice were collected with ethylenediaminetetraacetic acid (EDTA)-pretreated centrifuge tubes and gently mixed at room temperature before centrifugation to prevent blood coagulation. Supernatant plasma was collected and stored at −80°C prior to use. Standard testosterone samples were used to prepare a standard curve. Both standard testosterone samples and plasma samples were incubated with rabbit anti-testosterone antibody at 37°C for 1 h. Then, donkey anti-rabbit separating agent was added, and the mixture was kept still at room temperature for another 15 min prior to centrifugation at 3,500×g for 10 min. The absorbance was normalized to standard testosterone to calculate each sample (Beijing North Institute of Biotechnology Co., Ltd., Beijing, China).

Plasma samples were analyzed with Endocrine Multiplex Assay (MADKMAG-71K, Merck Millipore, Burlington, Massachusetts, USA) to determine levels of interleukin 6 (IL6), tumor necrosis factor alpha (TNFα), monocyte chemoattractant protein 1 (MCP1), according to the manufacturer’s protocol. Briefly, after incubation of the assay buffer for 10 min at room temperature, experimental plasma samples or gradient dilution standard samples were added to the wells of a 96-well plate and supplemented with assay buffer, serum matrix, and then neat samples. Dyed antibody-bound beads were then added and shaken gently before being left to incubate overnight at 4°C. Next, detection antibodies were added to each well following a washing step and incubated with agitation for 30 min at room temperature. Streptavidin–phycoerythrin was added to the plate and was incubated for another 30 min. The plate was washed for three times, and sheath fluid was added to resuspend the beads for 5 min. Fluorescent intensity quantification was performed using a Luminex analyzer (Luminex 200, Merck Millipore, Burlington, Massachusetts, USA). Data were analyzed using MILLIPLEX software (Analyst.V5.1, Merck Millipore, Burlington, Massachusetts, USA).

Plasma luteinizing hormone (LH) and follicle-stimulating hormone (FSH) were measured using ELISA kit (EM1188 and EM1035, FineTest, Wuhan, China) according to the manufacturer’s instructions. Briefly, experimental plasma samples and gradient dilution standard samples were incubated with biotin-labeled antibody in the pre-coated plate, respectively, for 45 min at 37°C. The plates were washed with wash buffer for three times; then, the horseradish peroxidase –streptavidin conjugate was added to the plates and incubated for 30 min at 37°C. The plates were washed five times with wash buffer; TMB substrate was added and incubated at 37°C in dark for 15 min. When the standard wells showed apparent gradient, stop solution was added to each well and subjected to a 450 nm in a microplate reader (Synergy H1, Bio-Tek, Santa Clara, California, USA) immediately. The absorbance was normalized to standard LH and FSH, respectively, to calculate the value for each sample.

Forty-eight hours after LPS injection, mice were put under deep anesthesia and transcardially perfused 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS), using saline-injected mice as controls. Histological analysis of H&E and immunofluorescence staining was performed as previously reported (26). For H&E staining, fixed testes and epididymis were harvested and post-fixed with 4% PFA for 24 h, then embedded in paraffin, sliced into 4-μm sections with a microtome (RM2016, Lecia, Wetzlar, Germany), and subjected to deparaffinization, hydration, and then stained with hematoxylin and eosin (AS1055A and AS1094, Aspen, Wuhan, China). For c-fos mapping, the brains were harvested, post-fixed with 4% PFA, and cryoprotected with 30% sucrose in PBS. For immunofluorescence staining, mouse brains were dehydrated in 30% sucrose and then sliced into 30-µm sections using cryostat (CM1860, Lecia, Wetzlar, Germany). Brain sections were stored at −20°C with a cryoprotectant prior to histological analyses. Tissue sections were first washed with PBS to remove embedding medium; antibody staining was performed in 24-well plates on floating tissue sections. Sections were blocked with normal goat serum for 1 h at room temperature and then incubated overnight in primary antibodies at 4°C, followed by PBS washing and incubation with secondary antibodies at room temperature for 2 h. The primary antibodies used in this study were rabbit anti-c-fos (2250, Cell Signaling Technology, Danvers, Massachusetts,USA, 1:500). Alexa Fluor@488-conjugated AffiniPure fab fragment goat anti-rabbit secondary antibody (111-547-003, Jackson ImmunoResearch, Philadelphia, Pennsylvania, USA, 1:200) was used. Sections were then incubated with 4′,6-diamidine-2′-phenylindole dihydrochloride (DAPI, D9542, Sigma, Saint Louis, Missouri, USA, 0.4 mg/ml) for 5 min. Confocal images were taken with a confocal microscope (LSM 880, Zeiss, Oberkochen, Germany) and scanning images were taken with a virtual microscopy slide scanning system (VS120, Olympus, Tokyo, Japan).

To explore expression levels of blood–testis–barrier-related, steroidgenesis and inflammatory genes, testicular tissues, epididymis tissues, adrenal gland and hypothalamus tissues from LPS-treated and control animals were collected and digested to perform real-time quantitative PCR analysis. Total RNA was extracted from the testis, epididymis tissue or hypothalamus tissue with Trizol reagent (15596018, Invitrogen, Carlsbad, California, USA) and quantified using a micro-spectrophotometer (NanoDrop One, Thermo Scientific, Massachusetts, USA). One microgram of total RNA was used for cDNA synthesis using ReverTra Ace qPCR RT Kit (FSQ-101, TOYOBO, Osaka, Japan). Real-time quantitative PCR was performed using SYBR^® Green Real-Time PCR Master Mix (QPK-201, TOYOBO, Osaka, Japan), with a LightCycler^® 480 Instrument (Roche, Basel, Switzerland). β-Actin was used to normalize expression levels of the genes tested. The primers used in this study are shown in Table 1 .

An open-field test (OFT) was used to assess anxiety-like behavior and locomotor activity in an open-field arena (50 cm × 50 cm × 50 cm). For analysis, the center of the arena floor covering 25% of the total floor area was defined as the center area (i.e., 25 cm × 25 cm). The OFT was used to assess behavior following LPS stress. Mice were first allowed to freely explore the open field for 1–2 min prior to a 10-min observation session. The number of entries to the center area and time spent in the center during the 10-min session were recorded and analyzed using Anymaze software (Stoelting, Chicago, Illinois, USA). For elevated plus maze (EPM) test, mice were put on a four-armed plus maze, at the center of two open and two closed arms (white PVC, 30 cm length × 5 cm width per arm) raised 50 cm above the floor, for 10-min sessions. The OFT and EPM were cleaned between mice with 20% ethanol solution. The number of entries to the open arms and time spent in the open arms were recorded and analyzed by Anymaze software (Stoelting, Chicago, Illinois, USA). To access the effect of low dose of LPS treatment on animal behaviors, an extra group of mice were treated with 0.4 mg/kg LPS for 48 h, besides the 4 mg/kg LPS treatment.

All data in this study were analyzed as mean ± SEM. Statistical significance was calculated using GraphPad Prism (GraphPad Software, San Diego, California, USA) and labeled as ns, p > 0.05, *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. Student’s t-test or Kruskal–Wallis with Dunn post-hoc test was used as appropriate. All n values in this study refer to the number of mice used in each experiment, as shown in each figure legend.

Adult male C57BL/6J mice were intraperitoneally injected with a single dose of either 4 mg/kg LPS in 0.9% saline or with 0.9% saline only. After 48 h, bodyweight decrease in the LPS group was significantly larger than that of the control group (p < 0.0001, Figure 1A ). There were higher plasma levels of inflammatory cytokines such as TNFα (p < 0.001), IL6 (p < 0.01), and MCP1 (p < 0.001) in the LPS group compared to the controls ( Figures 1B–D ), gene expression analysis of testis also revealed local testicular inflammation ( Supplementary Figure 1 ). In summary, acute stress caused by LPS administration resulted in a significant physiological disturbance in terms of inflammation.

H&E staining revealed immature male gametes in the lumen of seminiferous tubules of the testis ( Figure 2A ) and in the lumen of the epididymis ( Figure 2B ) 48 h after LPS treatment. Quantification of the gametes in the epididymis showed a statistical difference between the LPS treatment group and the saline control group (p < 0.05, Figure 2C ). The BTB lays the structural foundation of the spermatogenesis niche (27). Real-time quantitative PCR analysis of the testicular tissue revealed lower expression levels of genes coding for BTB-related proteins, including intercellular adhesion molecule 1 (Icam-1, p < 0.05), tight junction protein 1 (Tjp1, p < 0.05), and gap junction alpha-1 protein (Gja1, p < 0.05) in the LPS-treated groups compared to controls ( Figure 2D ). Using the epididymis tissue as a control, we found no such difference in gene expression levels of Icam-1, Tjp1, or Gja1 in the LPS-treated epididymis, indicating that LPS administration led to a dysfunction of cell–cell interaction in testes, rather than the epididymis ( Figure 2E ).

Radioimmune assay (RIA) showed that LPS-treated group had lower plasma testosterone levels than the controls (p < 0.05, Figure 3A ). We then performed real-time quantitative PCR to evaluate gene expression levels of key steroidogenesis enzymes using another batch of mice. Reduction in steroidogenic acute regulatory protein (StAR, p < 0.05), cholesterol side-chain cleavage enzyme (P450scc, p < 0.001), 3β-hydroxysteroid dehydrogenase (3β-HSD, p < 0.01), and cytochrome P450 family 11 subfamily B member 1 (Cyp11b1, p < 0.05) was found in the LPS treatment group compared to controls ( Figure 3B ). In the adrenal glands, another primary steroidogenesis organ, only one steroidogenic gene had reduced expression (3β-HSD, p < 0.01), whereas expression of the other three genes were not different between groups (StAR, p = 0.8129; P450scc, p = 0.1979; Cyp11b1, p = 0.1740; Figure 3C ). These results suggest that LPS treatment caused an impairment in male sex hormone production. The testes seem to be more vulnerable to the LPS-caused stress than the adrenal glands in terms of steroidogenic gene expression levels.

We observed abnormal behaviors after LPS treatment. When subjected to OFT, LPS-treated mice showed decreases in total distance travelled, entries to center, and time in center ( Supplementary Figures 2A, B ). In the EPM test, LPS treatment also induced decrease in total distance travelled and a trend in Entries to Open arms and Time in Open arms ( Supplementary Figures 2C, D ). To clarify the behavioral output of the LPS treatment, controls, 0.4 mg/kg LPS-treated group, and 4 mg/kg LPS-treated group were evaluated with OFT and EPM. The lower dose of LPS treatment caused no significant difference in locomotion and an increase in anxiety-like behavior ( Supplementary Figures 2E–G ). The abnormal behavior caused by LPS treatment suggested neurotoxicity of LPS as reported before. Gene expression analysis of hypothalamic tissues revealed elevated levels of inflammatory cytokines ( Supplementary Figure 3 ). Considering the importance of the HPG axis in regulating testicular function (28), we wanted to know whether LPS treatment induced any change in neuronal activity in the hypothalamus, especially the medial preoptic area (mPOA) where most the gonadotropin-releasing hormone (GnRH) neurons are located. Immunostaining of c-fos early response gene as an indicator of neuronal activation showed significant increase in the mPOA of the LPS group ( Figures 4A, B p < 0.05). ELISA results showed changes in circulating FSH (p < 0.05) and LH (p < 0.001) levels after LPS treatment ( Figures 4C, D ), suggesting LPS-treatment resulted in disturbed HPG-axis functions ( Figure 4E ).

Besides mPOA, c-fos immunofluorescent staining showed broad activation of brain nuclei in the LPS-treatment group than controls, including nuclei of the forebrain, thalamus, hypothalamus, brainstem, and cortex. Cell counting and analysis showed that there was significantly higher number of c-fos-positive cells in the LPS group compared to the saline control group in the following brain nuclei: the anterior cingulate cortex (ACC), lateral septum (LS), basolateral amygdala (BLA), paraventricular nucleus of the hypothalamus (PVN), lateral habenular nucleus (LHb), ventral tegmental area (VTA), locus coeruleus (LC), Barrington’s nucleus (BAR), and the nucleus of the solitary tract (NTS) ( Figures 5A, B ). Interestingly, most of the LPS-activated brain nuclei that we observed are participants in anxiety circuits and involved in stress responses (29), which was in line with the behavioral outputs that we observed ( Supplementary Figure 2E–G ).