Sex-dependent effects of adipocyte STAT3 inhibition on the inflammatory response during severe sepsis

Abstract

Introduction:

Sepsis is a dysregulated host response to infection that can lead to life-threatening organ dysfunction. Clinical and animal studies consistently demonstrate that female subjects are less susceptible to the adverse effects of sepsis, demonstrating the importance of understanding how sex influences sepsis outcomes. The signal transducer and activator of transcription 3 (STAT3) pathway is a major signaling pathway that facilitates inflammation during sepsis. STAT3 is abundantly expressed in white adipose tissue, however little is known about the contribution of white adipose tissue STAT3 activation during sepsis. We hypothesize that adipocyte STAT3 inhibition during severe sepsis will exaggerate the inflammatory response and impact organ injury, in a sex-dependent manner.

Methods:

We generated STAT3 flox/flox (WT) and adipocyte STAT3 knock out (A-STAT3 KO) mice using Cre-lox technology. Studies were done in 12–16-week-old male and female mice. Polymicrobial sepsis was induced by cecal ligation and puncture (CLP). Control non-septic mice did not undergo CLP (0h CLP). Tissues were harvested 18h after CLP. Body composition was determined by Echo magnetic resonance imaging (MRI). Energy metabolism was determine by indirect calorimetry. White adipose tissue morphology was determined by hematoxylin and eosin staining, while STAT3 activation in the white adipose tissue was determined by Western blot analysis and immunohistochemistry staining of STAT3 activation/phosphorylation at tyrosine 705. Plasma cytokines (TNF-α, IL-6, and leptin) were determined by luminex assay. Neutrophil infiltration of the lung and liver were assessed by myeloperoxidase activity assay. Histological signs of organ injury on lung and liver tissue were assessed by hematoxylin and eosin staining. Liver injury was further assessed by measuring plasma alanine and aspartate aminotransferase. In a separate cohort of mice, sepsis was induced by CLP and mice were monitored every 6–12 hours over a 7-day period to assess survival rate.

Results:

We demonstrate that neither body composition nor energy metabolism is altered with adipocyte STAT3 inhibition in male or female mice, under non-septic conditions. Sepsis was associated with reduced adipocyte size in female WT and A-STAT3 KO mice, suggesting that this event is STAT3 independent. Sepsis did not alter adipocyte size in male WT and A-STAT3 KO mice, suggesting that this event is also sex-dependent. Although STAT3 phosphorylation at tyrosine 705 expression is negligible in male and female A-STAT3 KO mice, septic female WT and A-STAT3 KO mice have higher white adipose tissue STAT3 activation than male WT and A-STAT3 KO mice. Adipocyte STAT3 inhibition did not alter the proinflammatory cytokine response during sepsis in male or female mice, as measured by plasma TNF-α, IL-6, and leptin levels. Adipocyte STAT3 inhibition reduced lung neutrophil infiltration and histological signs of lung injury during sepsis in male mice. On the contrary, adipocyte STAT3 inhibition had no effect on lung neutrophil infiltration or lung injury in female mice. We further demonstrate that neither liver neutrophil infiltration nor histological signs of liver injury is altered by adipocyte STAT3 inhibition during sepsis, in male or female mice. Lastly, adipocyte STAT3 inhibition did not affect survival rate of male or female mice during sepsis.

Conclusions:

Our study demonstrates that sex influences white adipose tissue STAT3 activation and morphology during sepsis, which is not dependent on the presence of functional STAT3 in mature adipocytes. Furthermore, genetic inhibition of adipocyte STAT3 activation in male, but not female mice, results in reduced lung neutrophil infiltration and lung injury during sepsis. The results from our study demonstrate the importance of considering biological sex and the white adipose tissue as potential sources and targets of inflammation during sepsis.

INTRODUCTION

Sepsis is a dysregulated host response to infection and is the leading cause of morbidity and mortality in the intensive care unit (1,2). Sepsis and progression to septic shock are a continuum of several pathophysiologic states that can result in organ damage and failure. The development of organ failure remains a threat to the survival of septic patients. Several clinical and animal studies have shown that female subjects are less likely to develop sepsis, are less susceptible to adverse outcomes and have lower mortality compared to men (3–5).

Overweight and obesity, a condition characterized by excessive fat accumulation in white adipose tissue, is a major global health problem (6). Sex differences are also observed during obesity, where the prevalence of obesity is higher in women than in men (7). The association between obesity and sepsis outcomes has been inconsistent in the literature. Some studies show that obesity increases the risk of developing sepsis (8,9) and increases sepsis-mortality (10,11). Other studies show reduced mortality among septic patients with obesity (12,13), which is termed the “obesity paradox”. Whether obesity is detrimental or protective remains unclear, but these findings led to our interest in understanding the biological significance of white adipose tissue during sepsis and highlight the importance of understanding how the influence of sex affects diseases processes.

White adipose tissue has traditionally been viewed as an energy reservoir, however this concept has been revised. The discovery of leptin, the first adipocyte-derived cytokine, in 1994 demonstrated that white adipose tissue is an active endocrine organ. White adipose tissue is now known to produce a variety of pro- and anti-inflammatory cytokines, several of which are common to those produced during sepsis. The Janus kinase (JAK)/signal transducer and activator of transcription-3 (STAT3) pathway is a major signaling pathway involved in regulating the cellular responses of cytokines during sepsis (14). Cytokines produced during infection (e.g., IL-6 and leptin) bind to their corresponding receptors and activate phosphorylation of STAT3 at tyrosine and serine residues (15). STAT3 then homo- or heterodimerizes, binds to specific promoter sequences in the nucleus, and regulates transcription of cytokine responsive genes. The embryonic lethality of STAT3 knockout mice prevents physiological studies (16). To overcome this limitation, tissue-specific conditional knockout models have been developed to study the function of STAT3. The contribution of STAT3 during sepsis has been studied in various organ systems including the lungs, liver, heart, and kidneys. STAT3 is highly expressed in mature adipocytes and stromal vascular cells of white adipose tissue (17,18), however little is known about the role of adipocyte STAT3 during sepsis.

The association between obesity and sepsis outcomes (detrimental or protective), provides evidence that white adipose tissue may contribute to the inflammatory response during sepsis. In the current study, we investigate the contribution of adipocyte STAT3 during sepsis, while also examining the influences of sex on sepsis outcomes. We hypothesize that adipocyte STAT3 inhibition during severe sepsis will exaggerate the inflammatory response and impact organ injury, in a sex dependent manner.

MATERIALS AND METHODS

Animals

To create mice with adipocyte STAT3 activation deletion, Cre-lox technology was used. Mice expressing an adiponectin Cre promotor (B6. FVB-Tg [Adipoq-Cre] 1Evdr/J; Jackson Laboratory, Bar Harbor, ME) were crossed with STAT3 floxed (B6.129S1-STAT3^tm1Xyfu/J) mice. Mice were genotyped by polymerase chain reaction. Genotypes used for experimentation were STAT3^{f/f_Adipoq-Cre−} (wild-type, WT) and STAT3 ^{f/f_Adipoq-Cre+} (adipocyte-specific STAT3 knockout, A-STAT3 KO) mice. The resulting genetically modified mice had deletions on exons 18, 19, and 20 of STAT3, which coincide with the SH2 domain important for phosphorylation/activation of STAT3. Both male and female mice were used at 12–14 weeks of age. Mice were housed under a 14:10 hour light-dark cycle in the animal facility at Cincinnati Children’s Hospital Medical Center (CCHMC). All mice received free access to water and a standard chow diet (Formulab no. 5008, 16% kcal provided by fat). The investigations conformed to the Guide for the Care and Use of Laboratory Animals (19) and were approved by the Institutional Animal Care and Use Committee at CCHMC.

Body composition

Body composition was examined with an EchoMRI Whole Body Composition Analyzer (Echo Medical Systems, Houston, TX). The EchoMRI is a quantitative nuclear magnetic resonance instrument that provides measurements of whole-body composition including body fat, lean mass, and total body water in live mice.

Metabolic assessment

Metabolic studies in live mice were conducted at CCHMC. Mice were housed individually in metabolic chambers under a 12:12 hour light-dark cycle with free access to food and water. Mice were acclimated to the environment of the metabolic chambers two days before data collection and data was collected for three days. Oxygen consumption (VO₂), carbon dioxide production (VCO₂), respiratory exchange ratio (RER), energy expenditure, locomotor activity, and food consumption were assessed in a metabolic monitoring system (Sable Systems International, Promethion Core System, Las Vegas, NV, USA).

Cecal ligation and puncture model

Polymicrobial sepsis was induced by cecal ligation and puncture (CLP) as previously described (20). Mice were anesthetized with 2% isoflurane inhalation. After opening the abdomen, the cecum was ligated with a 3.0 silk suture 3 – 4 mm distal from the ileocecal valve and double punctured with a 21-gauge needle. Fecal material was expressed into the peritoneum, and the abdominal incision was closed with silk running sutures and liquid topical adhesive. Immediately after the procedure, mice were fluid resuscitated with sterile saline (1ml) injected subcutaneously. Lidocaine hydrochloride (0.25%, 1.6 μl/g total dose) was administered at the surgical site, and buprenorphine (0.5 mg/kg) at the scruff of the neck to minimize pain, both were injected subcutaneously. Ceftriaxone (25 mg/kg) and metronidazole (12.5 mg/kg) was administered by intraperitoneal injection for antimicrobial coverage. Mice were sacrificed and tissue was harvested 18 hours after CLP (18h CLP). Non-septic mice did not undergo CLP (0h CLP) and served as controls.

Sample collection

Mice were anesthetized by 2% isoflurane inhalation. Blood was collected by cardiac puncture, centrifuged at 3000 rpm for 10 minutes at 4°C, and plasma was collected and stored at −80°C. Sections of epididymal white adipose tissue, lung, and liver samples were collected and stored in 10% buffered formalin for 48 hours then transferred to 70% ethanol for immunohistochemistry and histology, as described below. Remaining tissues (epididymal white adipose tissue, lung, and liver) were collected, flash-frozen in liquid nitrogen, and stored at −80°C until processing.

White adipose tissue cell separation

Approximately 150 mg of epididymal white adipose tissue was dissected and minced in 1.5 ml of digestion buffer (HBSS, 1% Pen-Strep, 4% BSA, 1.5 mg/ml Collagenase [Roche: 11088793001]) to facilitate quick digestion. The sample was incubated at 37°C and continuously inverted for 45 minutes and vigorously shaken by hand every 15 minutes. The digested tissue suspension was filtered through a 100 μm filter and centrifuged at 500 rpm for 2 minutes at room temperature. Adipocyte collection: The top layer of adipocytes was collected, washed with 1 ml of PBS, and centrifuged again at 500 rpm for 2 minutes at room temperature. The clean adipocytes were transferred to 150 μl of whole cell protein extraction buffer for protein isolation, as described below. Stromal vascular fraction collection: The bottom layer of the digested tissue suspension was centrifuged at 1200 rpm for 5 minutes at room temperature. The supernatant was discarded and the pellet containing stromal vascular cells were resuspended with 150 μl of whole cell protein extraction buffer for protein isolation, as described below.

Protein isolation and Western blot analysis

Protein was isolated from epididymal white adipose tissue, epididymal white adipose tissue adipocyte fraction, and epididymal white adipose tissue stromal vascular fraction, as indicated. Samples were homogenized in a whole cell protein extraction buffer (50 mM Tris-HCL, 250 mM NaCl, 3 mM EDTA, 3 mM EGTA, 1 % Triton X-100, 0.5% NP-40, 10% Glycerol, 1 mM PMSF, 2 mM PNPP, 3 mg/ml Aprotinin and 0.1 mM sodium orthovanadate). Homogenates were centrifuged at 8500 rpm for 30 minutes at 4°C. Supernatant was collected and total protein was quantified using a Bio-Rad protein assay kit. The Invitrogen NuPAGE gel electrophoresis system was used for all Western blotting. 30 μg of protein extract/sample were loaded onto 10% Bis-Tris gels (Invitrogen) and were separated by gel electrophoresis. Gels were transferred to a 0.45 μm pore size nitrocellulose membrane (Invitrogen) and incubated in SuperSignal^™ Western Blot Enhancer (Thermo Fisher Scientific) for 30 minutes at room temperature. Membranes were washed with 0.1% tris buffered saline with tween (TBST), blocked in 5% milk in 0.1% TBST for 1 hour, and incubated with primary antibody overnight at 4°C. Membranes were washed with 0.1% TBST and incubated with secondary antibody for 1 hour. Membranes were washed with 0.1% TBST and then incubated with SuperSignal^™ West Dura Extended Duration Substrate (Thermo Fisher Scientific). Membranes were imaged using the Bio-Rad ChemiDoc XRS+ gel documentation system and analyzed using ImageLab v. 5.1 software (Bio-Rad, Hercules, CA). The following antibodies were used: phospho (p) STAT3 tyrosine 705 (1:1000 dilution) (Cell Signaling Technology, Danvers, MA), STAT3 (1:1000 dilution) (Cell signaling Technology, Danvers, MA), and Beta Actin (1:500) (Santa Cruz Biotechnology, Dallas, TX).

Immunohistochemistry and histology

White adipose tissue morphology: epididymal white adipose tissue obtained from each mouse were fixed and embedded in paraffin blocks. A section of paraffin embedded tissue was cut for each mouse/sample using a microtome. Hematoxylin and eosin staining and conventional light microscopy (Upright Zeiss Axio Imager) was performed, and representative images for each sample were taken at 40x magnification using AxioVision LE software. Images were analyzed using NIS Elements imaging software (Nikon). In brief, all adipocytes in each image were selected and measured to determine lipid droplet surface area. The average lipid droplet surface area was calculated for each sample and the frequency of each adipocyte size was counted. White adipose tissue STAT3 phosphorylation at tyrosine 705: an additional section of epididymal white adipose tissue was cut for each sample, and automated immunohistochemistry was performed (Ventana Medical Systems, Tucson, AZ) using an anti-pSTAT3 tyrosine 705 antibody (1:100 dilution) (Cell Signaling, #9145). STAT3 phosphorylation at tyrosine 705 was visualized using conventional light microscopy (Olympus BX40 Binocular Microscope) and images were taken at 40x magnification using Q capture pro 7 software. Whole slide digital images were obtained from the resultant slides using an Aperio AT2 digital slide scanner (Leica Microsystems, Buffalo Grove, IL). Digital images were analyzed using Image Scope^™ software (Aperio version 12) using the positive pixel count algorithm as described by Gannon et al. (21). In brief, the algorithm generated three classes of output values (weak positive, positive, and strong positive) based on pixel intensity which represents staining intensity. Only positive and strong positive values were included in the analysis. Three similar-sized areas were randomly selected on each digital slide and the average value of those three areas was used as the final value for comparison. Lung and liver injury score: lung and liver samples were obtained from each mouse and were fixed and embedded in paraffin blocks. Hematoxylin and eosin staining and conventional light microscopy (Olympus BX40 Binocular Microscope) was performed, and representative images of each sample were taken at 40x magnification using Q capture pro 7 software. To objectively quantify organ injury, whole slides were scored by four independent observers blinded to the experimental protocol. Lung injury score was determined by alveolar congestion, hemorrhage, infiltration of leukocytes in airspace or alveolar wall, and thickness of alveolar wall or hyaline membrane formation. Liver injury score was determined by sinusoid congestion, infiltration of red blood cells and inflammatory cells, presence of lipid droplets, and necrosis. Severity of organ injury was calculated for each variable based on the percentage of the sample/slide that was injured. Score 0 = no injury, score 1 = minimal injury (0–25% of the section), score 2 = mild injury (25–50% of the section), score 3 = significant (50–75% of the section), and score 4 (more than 75% of the section). The injury score for each variable was summed to represent the total score (total score 0 – 16) and the average of each independent observer was calculated.

Myeloperoxidase Activity

Myeloperoxidase (MPO) activity, a peroxidase present in granules of neutrophils, was measured as a marker for neutrophil infiltration, in the lung and liver. Tissues were homogenized in 0.5% hexa-decyl-trimethyl-ammonium bromide dissolved in 10 mM potassium phosphate buffer (pH 7.0) and were centrifuged at 4000 rpm for 30 minutes at 4°C. An aliquot of supernatant was allowed to react with a solution of tetra-methyl-benzidine (1.6 mM) and hydrogen peroxide (0.1 mM). The rate of change in absorbance was measured by spectrophotometry at 650 nm. Myeloperoxidase activity was defined as the quantity of enzyme degrading 1 μmol hydrogen peroxide/min at 37°C and was expressed in units per 100 mg of tissue.

Alanine aminotransferase and aspartate aminotransferase assays

Plasma alanine aminotransferase and aspartate aminotransferase were measured using standard enzyme assay kits (Sekisui Diagnostics, Lexington, MA).

Plasma cytokine and adipokine analysis

Plasma cytokines were measured with a multiplex assay kit (EMD Millipore) using the manufacturer’s protocol.

Survival Study

In a separate study, sepsis was induced in 12–16-week-old WT and A-STAT3 KO mice at a high severity by CLP, as described above. To minimize pain buprenorphine (0.5 mg/kg) was administered every 12 hours following CLP procedure. Mice were monitored every 6–12 hours over a 7-day period or until all mice were deceased to assess survival rate. An injury severity scale was used to assess the severity of symptoms. Mice were scored on a scale of 0 to 3 for posture, feces consistency, eye appearance, hair coat, and labored breathing/gasping every 6–12 hours. A score of 0 referred to no signs of illness, with 1, 2, and 3 referring to minimum, mild, and severe symptoms. The scores were tallied for the four categories for each mouse, and if the total score was 8 or higher, the mouse was euthanized to minimize suffering.

Statistical Analysis

Data were analyzed using SigmaPlot for Windows Version 13 (SysStat Software, San Jose, CA). Results are presented as mean ± SD for parametric data and median and interquartile range for nonparametric data in text. In figures, data are presented as median and interquartile ranges to best convey the variability in the results. Student’s t-test was used for parametric data comparisons among two groups. Welch’s t-test and Mann-Whitney Rank Sum test were used for nonparametric data comparisons among two groups. Two-way ANOVA was used for parametric data comparisons involving two independent variables with the Holm-Sidak method for multiple comparisons. Two-way ANOVA on Ranks was used for nonparametric data. Data were log-transformed to meet assumptions, as indicated in figure legends. Data for metabolic studies were generated using CalR, a web-based analysis tool for indirect calorimetry experiments. Data for survival analysis were analyzed by Kaplan-Meier Survival Analysis log-rank test. Significance was considered as p≤0.05. N values represent individual animals. Specific details for statistical analysis and n value are noted in each figure legend.

RESULTS

Adipocyte STAT3 inhibition does not affect body composition

We used EchoMRI to measure the body composition of WT and A-STAT3 KO mice. In both sexes, there were no genotype differences in body weight, body fat, lean mass, free water, or total water (Fig. 1A–B and Fig. S1A–B). As expected, female mice had significantly lower body weight, lean mass, and total water than male mice in both genotypes (Fig. 1A–B and Fig. S1A–B). These data suggest that adipocyte STAT3 is not involved in body weight homeostasis under normal (non-septic) conditions.

Figure 1. Adipocyte STAT3 inhibition does not affect body composition.

Body composition of (A) male and (B) female WT and A-STAT3 KO mice. Vertical box represents 25^th percentile (bottom line), median (middle line), and 75^th percentile (top line) values, whiskers represent 10^th and 90^th percentiles. Data were analyzed by Student’s t-test, Mann-Whitney rank sum test when normality was not assumed and Welch’s t-test when equal variance was not assumed. White boxes = WT, gray boxes = A-STAT3 KO (n = 5–14 mice/group).

Adipocyte STAT3 inhibition does not alter energy metabolism

Body weight homeostasis is influenced by energy metabolism, which is a general process where living cells obtain and use energy from macronutrients. To better understand the functional role of STAT3 in mature adipocytes, we examined the energy metabolism of WT and A-STAT3 KO mice using indirect calorimetry. Cells require oxygen to release energy from macronutrients and through this reaction, carbon dioxide is produced. For both sexes, there was no difference in oxygen consumption or carbon dioxide production between WT and A-STAT3 KO mice (Fig. 2A, 3A, S2A and S3A). In male mice, the mean RER for WT and A-STAT3 KO mice was 0.8, indicating that mice are utilizing both carbohydrate and fat for energy (Fig. 2C and Fig. S2C). Similar findings in mean RER were observed in female mice (Fig. 3C and Fig. S3C). As expected, energy expenditure, food intake, and locomotor activity in male (Fig. 2D–F and Fig. S2D–F) and female (Fig. 3D–F and Fig. S3D–F) mice were not altered with A-STAT3 inhibition. This data suggests that adipocyte STAT3 is not involved in energy metabolism.

Figure 2. Effects of adipocyte STAT3 inhibition on energy metabolism in male mice.

Full day, light cycle, and dark cycle of (A) oxygen consumption, (B) carbon dioxide production, (C) respiratory exchange ratio, (D) energy expenditure, (E) food intake, and (F) locomotor activity. Vertical box represents 25^th percentile (bottom line), median (middle line), and 75^th percentile (top line) values, whiskers represent 10^th and 90^th percentiles. Data was generated using CalR: a web-based analysis tool for indirect calorimetry experiments. Data were analyzed using Student’s t-test. White boxes = WT, gray boxes = A-STAT3 KO (n = 7–9 mice/group).

Figure 3. Effects of adipocyte STAT3 inhibition on energy metabolism in female mice.

Full day, light cycle, and dark cycle of (A) oxygen consumption, (B) carbon dioxide production, (C) respiratory exchange ratio, (D) energy expenditure, (E) food intake, and (F) locomotor activity. Vertical box represents 25^th percentile (bottom line), median (middle line), and 75^th percentile (top line) values, whiskers represent 10^th and 90^th percentiles. Data was generated using CalR: a web-based analysis tool for indirect calorimetry experiments. Data were analyzed using Student’s t-test or Mann-Whitney rank sum test when normality was not assumed. White boxes = WT, gray boxes = A-STAT3 KO (n = 7–10 mice/group).

Sepsis alters adipocyte cell size in a sex-dependent manner

We used hematoxylin and eosin staining to examine the morphological differences of epididymal white adipose tissue in non-septic (0h CLP) and septic (18h after CLP) WT and A-STAT3 KO mice. To quantify adipocyte cell size, we measured the mean lipid droplet surface area and frequency of cell size using NIS Elements image analysis software. As expected, at 0h CLP, there were no genotype differences in lipid droplet surface area for both male (Fig. 4A, 4C, S4A and S4C) or female mice (Fig. 4B, 4D, S4B and S4D). In male mice, lipid droplet surface area was not altered with adipocyte STAT3 inhibition in non-septic or septic mice (Fig. 4A, 4C, S4A and S4C). In female mice, lipid droplet surface area in septic WT mice was significantly smaller compared to non-septic WT mice (144 ± 129 vs. 2,186 ± 776 μm², p ≤ 0.05) (Fig. 4B, 4D, S4B and S4D). On the contrary, lipid droplet surface area was not altered in female A-STAT3 KO mice 18h after CLP, compared to non-septic A-STAT3 KO mice (Fig. 4B, 4D, S4B and S4D). In female mice, there were no genotype differences in lipid droplet surface area during sepsis (Fig. 4B, 4D, S4B and S4D). Interestingly, we observed sex differences during sepsis and found that female septic mice had a lower mean lipid droplet surface area than male septic mice, in both WT (144 ± 129 vs. 2043 ± 647 μm², p ≤ 0.05) and A-STAT3 KO mice (908 ± 958 vs. 2307 ± 588 μm², p ≤ 0.05) (Fig. 4E and Fig. S4E).

Figure 4. Hematoxylin and eosin staining of epididymal white adipose tissue.

Representative images of epididymal white adipose tissue hematoxylin and eosin staining at 0h and 18h after CLP in (A and C) male WT and A-STAT3 KO mice, (B and D) female WT and A-STAT3 KO mice, and (E) male and female, septic WT and A-STAT3 KO mice. Images at 40x magnification. Scale bar represents 20 μm in length. Vertical box represents 25^th percentile (bottom line), median (middle line), and 75^th percentile (top line) values, whiskers represent 10^th and 90^th percentiles. Data were analyzed by two-way ANOVA with Holm-Sidak method for all pairwise multiple comparisons. *P ≤ 0.05 vs. 0h within genotype, &P ≤ 0.05 vs. males within genotype. White boxes = WT, gray boxes = A-STAT3 KO (n = 3–10 mice/group).

When comparing the frequency of lipid droplet surface area, male mice had lipid droplet sizes that spanned the spectrum of small to large sized cells (Fig. 5A). The percentage of small cells measuring less than 400 μm² was 3.4% in male WT mice without sepsis, 5% in male WT mice with sepsis, 4% in male KO mice without sepsis and 0% in male KO mice with sepsis. In comparison, female mice had a higher percentage of cells measuring less than 400 μm²: 0% in female WT mice without sepsis, 89% in female WT mice with sepsis, 24% in female KO mice without sepsis and 46% in female KO mice with sepsis (Fig. 5B). Taken together, this data suggests that sex influences white adipose tissue morphology during sepsis.

Figure 5. Frequency of epididymal white adipose tissue lipid droplet surface area.

Frequency of lipid droplet surface area of WT and A-STAT3 KO mice at 0h and 18h after CLP in (A) male and (B) female mice. Solid black line = WT 0h CLP, solid red line = A-STAT3 KO 0h CLP, dash black line = WT 18h CLP, and dash red line = A-STAT3 KO 18h CLP (n= 3–10 mice/group).

Sex influences white adipose tissue STAT3 activation during sepsis

We used immunohistochemistry to determine STAT3 activation/phosphorylation at tyrosine 705 in epididymal white adipose tissue of WT and A-STAT3 KO mice at 0h and 18h after CLP. Positive pixel count was quantified using Image Scope, an image analysis software. STAT3 activation was not observed in adipocytes of WT or A-STAT3 KO non-septic male (Fig. 6A, 6C, S5A and S5C) or female mice (Fig. 6B, 6D, S5B and S5D). In male mice, STAT3 activation was significantly higher in septic WT mice compared to non-septic WT controls (Fig. 6A, 6C, S5A and S6C). No changes were observed in septic A-STAT3 KO mice compared to non-septic A-STAT3 KO mice, but septic A-STAT3 KO mice had significantly less STAT3 activation than septic WT mice (Fig. 6A, 6C, S5A and S5C). Surprisingly, residual STAT3 activation can be observed in male A-STAT3 KO mice during sepsis (Fig. 6A and Fig. S5A). Like males, sepsis significantly increased STAT3 activation in female septic WT mice compared to non-septic WT mice (Fig. 6B, 6D, S5B and S5D). No significant differences were observed in female septic A-STAT3 KO mice compared to non-septic A-STAT3 KO mice (Fig. 6B, 6D, S5B and S5D). As expected, female septic A-STAT3 KO mice had significantly less STAT3 activation than female septic WT mice (Fig. 6B, 6D, S5B and S5D). Like males, residual STAT3 activation can be observed in female A-STAT3 KO mice during sepsis (Fig. 6B and Fig. S5B).

Figure 6. Sex influences white adipose tissue STAT3 activation during sepsis.

Representative images of epididymal white adipose tissue STAT3 phosphorylation at tyrosine 705 staining in (A and C) male WT and A-STAT3 KO mice at 0h and 18h after CLP, (B and D) female WT and A-STAT3 KO mice at 0h and 18h after CLP, and (E) male and female, septic WT and septic A-STAT3 KO mice. Images at 40x magnification. Scale bar represents 25 μm in length. Vertical box represents 25^th percentile (bottom line), median (middle line), and 75^th percentile (top line) values, whiskers represent 10^th and 90^th percentiles. Data were analyzed by two-way ANOVA on Ranks with Holm-Sidak method for all pairwise multiple comparisons. *P ≤ 0.05 vs. 0h within genotype, #P ≤ 0.05 vs. WT within 18h CLP, &P ≤ 0.05 vs. males within genotype. White boxes = WT, gray boxes = A-STAT3 KO (n = 3–9 mice/group).

We also observed sex differences during sepsis and found that female septic mice had higher STAT3 activation compared to male mice in both WT and KO mice. (Fig. 6E and Fig. S5E). These data indicate that white adipose tissue STAT3 activation is reduced in A-STAT3 KO mice during sepsis, and that STAT3 activation in the white adipose tissue is less pronounced in male mice.

STAT3 phosphorylation at tyrosine 705 is inhibited in adipocytes but not the stromal vascular fraction of A-STAT3 KO mice

Next, we investigated if residual STAT3 activation on A-STAT3 KO mice was located in the adiponectin-Cre targeted mature adipocytes or in the stromal vascular fraction of the adipose tissue that includes endothelial, immune cells, adipocyte progenitors and others. Therefore, we separated mature adipocytes from the stromal vascular fraction in epididymal white adipose tissue to investigate if activation of STAT3 is inhibited in adipocytes of A-STAT3 KO mice during sepsis, while also determining the cellular source of STAT3 activation in A-STAT3 KO mice during sepsis. We used Western blot analysis to examine STAT3 phosphorylation at tyrosine 705 in adipocytes and stromal vascular cells of WT and A-STAT3 KO mice at 0h and 18h after CLP. To our surprise, in both male and female mice, STAT3 was phosphorylated at tyrosine 705 in adipocytes of non-septic (0h) WT mice, but not adipocytes of septic (18h) WT mice (Fig. S6A and S6C). Consistent with the adipocyte fraction, STAT3 was phosphorylated at tyrosine 705 in the stromal vascular fraction of non-septic (0h) mice, but not septic (18h) mice (Fig. S6B and S6D). However, STAT3 was phosphorylated at tyrosine 705 in the stromal vascular fraction of both WT and A-STAT3 KO mice (Fig. S6B and S6D). These findings confirm our experimental methods that STAT3 tyrosine 705 phosphorylation occurs in adipocytes of WT mice but not adipocytes of A-STAT3 KO mice (Fig. S6A and S6C) and demonstrate that activation of STAT3 is inhibited in adipocytes of A-STAT3 KO mice but not in the stromal vascular fraction.

Effects of adipocyte STAT3 inhibition on plasma cytokine levels during sepsis

We examined the importance of adipocyte STAT3 inhibition on the systemic proinflammatory cytokine response during sepsis, as measured by plasma levels of TNF-α, IL-6, and leptin in WT and A-STAT3 KO mice at 0h and 18h after CLP. In both WT and A-STAT3 KO male mice, plasma TNF-α was higher at 18 hours after CLP compared to their respective non-septic mice (Fig. 7A and Fig. S7A). Similar findings were observed in female mice (Fig. 7B and Fig. S7B). In male mice, plasma IL-6 increased 18 hours after CLP in both WT and A-STAT3 KO mice compared to their respective non-septic controls (Fig. 7C and Fig. S7C). Similar findings were observed in female mice (Fig. 7D and Fig. S7D). In male mice, plasma leptin increased 18 hours after CLP compared to their respective non-septic controls in both WT and A-STAT3 KO (Fig. 7E and Fig. S7E). Plasma leptin was not altered in female mice (Fig. 7F and Fig. S7F). There were no genotype differences in plasma cytokine levels, indicating that adipocyte STAT3 inhibition does not alter plasma cytokine levels during sepsis.

Figure 7. Effects of adipocyte STAT3 inhibition on plasma cytokine levels during sepsis.

(A) Male plasma TNF-α log transformed, (B) female plasma TNF-α log transformed, (C) male plasma IL-6 log transformed, (D) female plasma IL-6 log transformed, (E) male plasma leptin log transformed, and (F) female plasma leptin log transformed, in WT and A-STAT3 KO mice at 0h and 18h after CLP. Vertical box represents 25^th percentile (bottom line), median (middle line), and 75^th percentile (top line) values, whiskers represent 10^th and 90^th percentiles. Data were analyzed by two-way ANOVA with Holm-Sidak method for all pairwise multiple comparisons. *P ≤ 0.05 vs. 0h within genotype. White boxes=WT, gray boxes=A-STAT3 KO (n=5–10 mice/group).

Adipocyte STAT3 inhibition reduces lung neutrophil infiltration and lung injury during sepsis, in a sex-dependent manner

We investigated the effects of adipocyte STAT3 inhibition on organ injury by measuring MPO activity, a marker for neutrophil infiltration, in the lungs of WT and A-STAT3 KO mice at 0h and 18h after CLP. In male mice, lung MPO activity 18h after CLP was significantly higher compared to non-septic mice in both WT (602 ± 212 vs. 82 ± 30 Units/100 mg tissue, p ≤ 0.05) and A-STAT3 KO mice (367 ± 130 vs. 94 ± 40 Units/100 mg tissue, p ≤ 0.05) (Fig. 8A and Fig. S8A). In female mice, lung MPO activity 18h after CLP was significantly higher compared to non-septic mice in both WT (340 ± 122 vs. 62 ± 25 Units/100 mg tissue, p ≤ 0.05) and A-STAT3 KO mice (325 ± 123 vs 91 ± 28 Units/100 mg tissue, p ≤ 0.05) (Fig. 8B and Fig. S8B). Interestingly, in male mice, lung MPO was higher in septic WT mice compared to septic A-STAT3 KO mice (602 ± 212 vs 367 ± 130 U/100 mg tissue, p ≤ 0.05) (Fig. 8A and Fig. S8A).

Figure 8. Effects of adipocyte STAT3 inhibition on lung neutrophil infiltration during sepsis.

Lung MPO in (A) male and (B) female WT and A-STAT3 KO mice at 0h and 18h after CLP. Vertical box represents 25^th percentile (bottom line), median (middle line), and 75^th percentile (top line) values, whiskers represent 10^th and 90^th percentiles. Data were analyzed by two-way ANOVA on Ranks with Holm-Sidak method for all pairwise multiple comparisons. *P ≤ 0.05 vs. 0h within genotype, #P ≤ 0.05 vs. WT within 18h after CLP. White boxes=WT, gray boxes=A-STAT3 KO (n=3–10 mice/group).

Hematoxylin and eosin staining was performed to assess histological signs of lung injury, which include alveolar congestion, hemorrhage, infiltration of leukocytes in the airspace or alveolar walls, and thickness of the alveolar walls. In male mice, lung injury was higher 18h after CLP in WT mice compared to non-septic WT mice (Fig. 9A, 9C, S9A and S9C). On the contrary, sepsis did not change lung injury score in A-STAT3 KO mice (Fig. 9A, 9C, S9A and S9C). As expected, septic A-STAT3 KO mice had significantly less lung injury than septic WT mice (Fig. 9A, 9C, S9A and S9C). In female mice, lung injury was higher 18h after CLP compared to non-septic controls in both WT and A-STAT3 KO mice (Fig. 9B, 9D, S9B and S9D). Unlike males, there were no genotype differences in lung histology in female mice during sepsis (Fig. 9B, 9D, S9B and S9D). Taken together, these results show that adipocyte STAT3 inhibition reduces lung injury during sepsis, in a sex-dependent manner.

Figure 9. Hematoxylin and eosin staining of the lung.

Representative images of lung hematoxylin and eosin staining at 0h and 18h after CLP in (A and C) male WT and A-STAT3 KO mice and (B and D) female WT and A-STAT3 KO mice. Images at 40X magnification. Scale bar represents 100 μm in length. Vertical box represents 25^th percentile (bottom line), median (middle line), and 75^th percentile (top line) values, whiskers represent 10^th and 90^th percentiles. Data were analyzed by two-way ANOVA with Holm-Sidak method for all pairwise multiple comparisons. *P ≤ 0.05 vs. 0h within genotype, #P ≤ 0.05 vs. WT within 18h after CLP. White boxes = WT, gray boxes = A-STAT3 KO (n = 4–7 mice/group).

Adipocyte STAT3 inhibition does not affect liver neutrophil infiltration, liver injury, or survival rate during sepsis

We also investigated the effects of adipocyte STAT3 inhibition on liver injury by measuring liver MPO activity in WT and A-STAT3 KO mice at 0h and 18h after CLP. In male mice, liver MPO activity was higher 18h after CLP compared to non-septic controls in both WT and A-STAT3 KO mice (Fig. 10A and Fig. S10A). In female mice, liver MPO was higher 18h after CLP compared to non-septic controls in both WT and A-STAT3 KO mice (Fig. 10D and Fig. S10D). There were no genotype differences during sepsis in male or female mice (Fig. 10A, 10D, S10A and S10D).

Figure 10. Effects of adipocyte STAT3 inhibition on liver neutrophil infiltration and plasma ALT and AST levels during sepsis.

Liver MPO in (A) male and (D) female WT and A-STAT3 KO mice, plasma ALT levels in (B) male and (E) female WT and A-STAT3 KO mice, and plasma AST levels in (C) male and (F) female WT and A-STAT3 KO mice at 0h and 18h after CLP. Vertical box represents 25^th percentile (bottom line), median (middle line), and 75^th percentile (top line) values, whiskers represent 10^th and 90^th percentiles. Data were analyzed by two-way ANOVA or two-way ANOVA on ranks with Holm-Sidak method for all pairwise multiple comparisons. *P ≤ 0.05 vs. 0h within genotype, #P ≤ 0.05 vs. WT within CLP time. White boxes = WT, gray boxes = A-STAT3 KO (n = 3–10 mice/group).

Hematoxylin and eosin staining was performed to assess histological signs of liver injury, which include sinusoid congestion, infiltration of red blood and inflammatory cells, presence of lipid droplets, and necrosis. In male mice, liver injury was higher 18h after CLP compared to non-septic controls in both WT and A-STAT3 KO mice (Fig. 11A, 11C, S11A and S11C). In female mice, liver injury was higher 18h after CLP compared to non-septic controls in both WT and A-STAT3 KO mice (Fig. 11B, 11D, S11B and S11D). Consistent with liver neutrophil infiltration, there were no genotype differences in liver histology in male or female mice during sepsis (Fig. 11A–D and Fig. S11A–D).

Figure 11. Hematoxylin and eosin staining of the liver.

Representative images of liver hematoxylin and eosin staining at 0h and 18h after CLP in (A and C) male WT and A-STAT3 KO mice and (B and D) female WT and A-STAT3 KO mice. Images at 40x magnification. Scale bar represents 100 μm in length. Vertical box represents 25^th percentile (bottom line), median (middle line), and 75^th percentile (top line) values, whiskers represent 10^th and 90^th percentiles. Data were analyzed by two-way ANOVA with Holm-Sidak method for all pairwise multiple comparisons. *P ≤ 0.05 vs. 0h within genotype. White boxes = WT, gray boxes = A-STAT3 KO (n = 4–7 mice/group).

We further examined the effects of adipocyte STAT3 inhibition on liver injury by measuring plasma ALT and AST levels, enzymes released into the circulation upon hepatocyte injury. In male mice, plasma ALT was higher 18h after CLP compared to non-septic controls in both WT (55 ± 10 vs. 12 ± 3 U/L, p ≤ 0.05) and A-STAT3 KO mice (65 ± 9 vs. 20 ± 11 U/L, p ≤ 0.05) (Fig. 10B and Fig. S10B). Interestingly, plasma ALT was significantly higher in septic A-STAT3 KO mice compared to septic WT mice (65 ± 9 vs. 55 ± 10 U/L, p ≤ 0.05) (Fig. 10B and Fig. S10B). Furthermore, in female mice, plasma ALT was significantly higher in non-septic A-STAT3 KO mice, compared to non-septic WT mice (16 ± 3 vs. 8 ± 2 U/L, p ≤ 0.05) (Fig. 10E and Fig. S10E). Additionally, plasma ALT was higher 18h after CLP compared to non-septic control mice in both WT (68 ± 2 vs. 8 ± 2 U/L, p ≤ 0.05) and A-STAT3 KO mice (60 ± 9 vs. 16 ± 3 U/L, p ≤ 0.05) however there were no genotype differences during sepsis (Fig. 10E and Fig. S10E). Plasma AST levels were not affected by sepsis or adipocyte STAT3 inhibition in male or female mice (Fig. 10C, 10F, S10C and S10F). The results from liver neutrophil infiltration and liver injury scores suggest that adipocyte STAT3 has no role in liver injury during sepsis, in male or female mice. However, results became conflicting with increased plasma ALT in male septic A-STAT3 KO mice, compared to male septic WT mice during sepsis. To determine the effects of adipocyte STAT3 inhibition on survival rate following sepsis, we induced sepsis in male and female WT and A-STAT3 KO mice and monitored survival of the mice. Our results indicate that adipocyte STAT3 inhibition does not alter survival rate after sepsis onset, in male or female mice (Fig. 12A–B).

Figure 12. Survival of WT and A-STAT3 KO mice following CLP.

Sepsis was induced in (A) male and (B) female WT and A-STAT3 KO mice and mice were monitored for survival. Data were analyzed by Kaplan-Meier Survival Analysis log-rank test. White boxes = WT, gray boxes = A-STAT3 KO (n = 6–7 mice/group).

DISCUSSION

In the current study, we explored the impact of sex on body weight homeostasis and sepsis outcomes in mice with genetic inhibition of adipocyte STAT3 activation. We demonstrate that neither body composition nor energy metabolism is affected by adipocyte STAT3 inhibition in male or female mice, under normal (non-septic) conditions. Sepsis was associated with reduced adipocyte cell size in female WT and A-STAT3 KO mice compared to male counterparts. Adipocyte STAT3 inhibition did not alter the pro-inflammatory cytokine response during sepsis, as measured by plasma TNF-α, IL-6, and leptin levels. We demonstrated that adipocyte STAT3 inhibition reduced lung neutrophil infiltration and lung injury in male mice. On the contrary, adipocyte STAT3 inhibition had no effect on lung neutrophil infiltration or lung injury in female mice. We further demonstrate that neither liver neutrophil infiltration nor liver injury is altered by adipocyte STAT3 inhibition during sepsis, in male or female mice. Survival rate of septic WT and A-STAT3 KO mice was evaluated to understand the significance of adipocyte STAT3 inhibition on mortality during sepsis. We showed that adipocyte STAT3 inhibition does not affect survival of male or female mice during sepsis. In conclusion, our data demonstrate that adipocyte STAT3 inhibition regulates the inflammatory response during sepsis with differences by sex.

Adipocyte STAT3 inhibition and energy metabolism

Previous studies have suggested that STAT3 plays an important role in regulating mammalian body weight and energy homeostasis. Leptin, an adipocyte-derived hormone, regulates body weight by suppressing food intake and increasing metabolic rate (22). Mice deficient in leptin (ob/ob) (23) or resistant to the actions of leptin (db/db) (24) are severely obese and hyperphagic. Leptin acts through its receptor, notably LepRb, which is capable of activating the JAK2/STAT3 pathway (25).

STAT3 is highly expressed in pre- and mature adipocytes of the white adipose tissue (17,18), which is the primary site of leptin production and the site of body fat storage. The contribution of STAT3 in adipocyte function is not fully understood. We showed that adipocyte STAT3 inhibition had no effect on body composition or energy metabolism under normal conditions, in male or female mice, similar to findings on body composition and energy metabolism in male mice found by Reilly et al. (26). These findings indicate that the effects of STAT3 in the regulation of body weight homeostasis is not dependent on the presence of functional STAT3 in adipocytes and is not influenced by sex. It should be noted that STAT3 is involved in body weight homeostasis when activated in the hypothalamus. Previous studies have shown that leptin signaling activates STAT3 in the hypothalamus and hypothalamic STAT3 activation is required for regulating body weight homeostasis (27–29). Mice with hypothalamic deletion of STAT3 have increased food intake, body weight, adiposity, and leptin resistance (30). Taken together, these studies suggest that STAT3 may have an important regulatory role in metabolism only when it is activated centrally.

Sex influences white adipose tissue STAT3 activation and morphology during sepsis

During sepsis, we observed sex differences in STAT3 phosphorylation at tyrosine 705 and white adipose tissue morphology. The white adipose tissue is a heterogeneous tissue. Besides mature adipocytes, the white adipose tissue contains cells of the stromal vascular fraction, which include preadipocytes, fibroblasts, endothelial cells, and immune cells, which all express STAT3. Studies to examine STAT3 activation in the white adipose tissue during sepsis, by immunohistochemistry analysis, revealed that septic A-STAT3 KO mice had negligible STAT3 phosphorylation at tyrosine 705. However, we observed white adipose tissue STAT3 activation in septic A-STAT3 KO mice during sepsis. This suggests that STAT3 activation in septic A-STAT3 KO mice is caused by activation of STAT3 in stromal vascular cells. This finding is consistent with Cernkovich et al. (31) who reported residual white adipose tissue STAT3 activation in mice with genetic inhibition of adipocyte STAT3 activation. Therefore, we separated mature adipocytes from the stromal vascular fraction and confirmed that STAT3 activation is inhibited in mature adipocytes but not stromal vascular cells of male and female A-STAT3 KO mice. To our surprise, we found that STAT3 is activated in adipocytes of non-septic WT mice but not septic WT mice, in both male and female cohorts. Normally, STAT3 remains inactive under non-septic conditions and is activated during sepsis, so it is unclear why STAT3 phosphorylation is enhanced in non-septic adipocytes and stromal vascular cells, compared to septic adipocytes and stromal vascular cells. It is possible that this is related to the reduction in total STAT3 levels observed in septic mice. Reduction in total STAT3 levels during sepsis may limit detection of STAT3 phosphorylation by Western blot analysis.

To our surprise, female WT and A-STAT3 KO mice had higher STAT3 phosphorylation at tyrosine 705 expression in the white adipose tissue than male WT and A-STAT3 KO mice, as revealed by immunohistochemistry analysis. Furthermore, female WT and A-STAT3 KO mice had smaller adipocytes during sepsis than their respective non-septic controls; this was not observed in male mice. Also, when comparing morphology of male and female mice, female mice had smaller adipocytes than male mice during sepsis, which was independent of adipocyte STAT3 activation. While the mechanism of the sex differences observed in this study are unknown, it is known that biological sex influences the immune response during infection by regulating exposure, recognition, and clearance of microorganisms (32). The sex of an individual is defined by their sex chromosomes, reproductive organs, and their hormonal sex steroid levels (32). It is possible that there is a chromosomal and/or hormonal influence on white adipose tissue STAT3 activation and morphology during sepsis. Studies examining the impact of sex hormone levels, such as ovariectomy and orchiectomy, and evaluation of X- and Y-linked genes activated during sepsis are necessary to examine the mechanisms of the sex differences observed in this study and are beyond the scope of the current work.

Systemic proinflammatory cytokine response and organ injury during sepsis

Sepsis increased plasma TNF-α and IL-6 in both WT and A-STAT3 KO mice, however the proinflammatory cytokine response was not alter with adipocyte STAT3 inhibition in male or female cohorts. Plasma leptin increased in male WT and A-STAT3 KO mice during sepsis, but not female WT or A-STAT3 KO mice. It is possible that there is a correlation between white adipose tissue morphology and leptin production during sepsis. Female mice experience a reduction in adipocyte size during sepsis that may lessen leptin production by adipocytes during sepsis.

Although adipocyte STAT3 inhibition had no effect on the systemic pro-inflammatory cytokine response during sepsis, we found that lung neutrophil infiltration and the histological signs of lung injury were reduced in male A-STAT3 KO mice compared to WT mice during sepsis. On the contrary, adipocyte STAT3 inhibition did not alter lung neutrophil infiltration or lung injury in female mice, suggesting that this event may be sex-dependent. The connection between adipocyte STAT3 inhibition and reduced lung injury in male mice was not determined by this investigation. The white adipose tissue is an endocrine organ capable of producing a variety of adipokines, non-coding RNAs, and extracellular vesicles that can communicate with local and distal tissues to influence inflammation during sepsis. These products are derived from both adipocytes and cells of the stromal vascular fraction. It is possible that cytokine and/or chemokine release from adipocytes are altered with adipocyte STAT3 inhibition during sepsis in male mice. For instance, adipocytes produce several cytokines (e.g., granulocyte-colony stimulating factor), chemokines (e.g., monocytes chemoattractant protein-1 and chemokine receptor ligand 12), and miRNAs (e.g., miR-125b) that regulate neutrophil production and mobilization (33–37). Future studies evaluating genes activated by adipocyte STAT3 during sepsis may provide some clarity.

Lung injury in female mice was unaffected by adipocyte STAT3 inhibition during sepsis. It is possible that the effects of adipocyte STAT3 inhibition on lung injury in female mice cannot be detected in our study. There is evidence that organ injury is less pronounced in female mice during sepsis. In our study, we found that during sepsis, female WT mice have less MPO activity than male WT mice (statistical analysis was not shown). Female mice have less renal mitochondrial injury than male mice in a comorbid model of sepsis (38). There is a significant reduction in renal mitochondrial complex I, II, and III activity in septic male mice that initially underwent uninephrectomy to establish renal disease, whereas female mice were less effected; only complex I was reduced (38). Female mice have less cardiac dysfunction than male mice with sepsis, as measured by ejection fraction, fractional shortening, and fractional area change (39). Furthermore, female mice experience milder liver injury than male mice during sepsis, as evidenced by lower levels of serum ALT and AST, reduced liver infiltration of inflammatory cells, and less liver necrosis (40). Further studies are necessary to understand the contribution of sex on lung injury during sepsis.

We further showed that adipocyte STAT3 inhibition did not affect liver neutrophil infiltration or liver injury during sepsis in male or female mice. Despite these results, in male mice liver injury, as measured by plasma ALT levels, were higher in A-STAT3 KO mice compared to WT mice during sepsis. On the contrary, in female mice, plasma ALT levels were higher in non-septic A-STAT3 KO mice compared to non-septic WT mice. ALT is considered the most liver-specific enzyme in detecting organ injury (41). However, ALT is also produced by cardiac muscle, skeletal muscle, and kidneys (41). It is possible that we are seeing an additive effect of ALT production by other tissues not evaluated in study (cardiac muscle, skeletal muscle, and kidneys).

To address the importance of adipocyte STAT3 in sepsis, survival rates of WT and A-STAT3 KO mice were examined following CLP, however survival rates of WT and A-STAT3 KO mice were comparable in both male and female cohorts. These findings suggests that although adipocytes STAT3 inhibition reduces lung injury during sepsis in male mice it does not alter sepsis survival.

Conclusion

In conclusion, our study demonstrates that sex influences white adipose tissue STAT3 activation and morphology during sepsis, which is not dependent on the presence of functional STAT3 in mature adipocytes. Furthermore, genetic inhibition of adipocyte STAT3 activation in male, but not female mice, results in reduced lung neutrophil infiltration and lung injury during sepsis. The results from our study demonstrate the importance of considering biological sex and the white adipose tissue as potential sources and targets of inflammation during sepsis.

Supplementary Material

Supplmentary Figure 1

Supplemental Figure 1. Adipocyte STAT3 inhibition does not affect body composition. Body composition of (A) male and (B) female WT and A-STAT3 KO mice. Each individual data point represents one mouse. Data were analyzed by Student’s t-test, Mann-Whitney rank sum test when normality was not assumed and Welch’s t-test when equal variance was not assumed. White circles = WT, black circles = A-STAT3 KO (n = 5–14 mice/group).

Supplmentary Figure 2

Supplemental Figure 2. Effects of adipocyte STAT3 inhibition on energy metabolism in male mice. Full day, light cycle, and dark cycle of (A) oxygen consumption, (B) carbon dioxide production, (C) respiratory exchange ratio, (D) energy expenditure, (E) food intake, and (F) locomotor activity. Each individual data point represents one mouse. Data was generated using CalR: a web-based analysis tool for indirect calorimetry experiments. Data were analyzed using Student’s t-test. White boxes = WT, gray boxes = A-STAT3 KO (n = 7–9 mice/group).

Supplmentary Figure 3

Supplemental Figure 3. Effects of adipocyte STAT3 inhibition on energy metabolism in female mice. Full day, light cycle, and dark cycle of (A) oxygen consumption, (B) carbon dioxide production, (C) respiratory exchange ratio, (D) energy expenditure, (E) food intake, and (F) locomotor activity. Each individual data point represents one mouse. Data was generated using CalR: a web-based analysis tool for indirect calorimetry experiments. Data were analyzed using Student’s t-test or Mann-Whitney rank sum test when normality was not assumed. White boxes = WT, gray boxes = A-STAT3 KO (n = 7–10 mice/group).

Supplmentary Figure 4

Supplemental Figure 4. Hematoxylin and eosin staining of epididymal white adipose tissue. Representative images of epididymal white adipose tissue hematoxylin and eosin staining at 0h and 18h after CLP in (A and C) male WT and A-STAT3 KO mice, (B and D) female WT and A-STAT3 KO mice, and (E) male and female, septic WT and A-STAT3 KO mice. Images at 40x magnification. Scale bar represents 20 μm in length. Each individual data point represents one mouse. Data were analyzed by two-way ANOVA with Holm-Sidak method for all pairwise multiple comparisons. *P ≤ 0.05 vs. 0h within genotype, &P ≤ 0.05 vs. males within genotype. White boxes = WT, gray boxes = A-STAT3 KO (n = 3–10 mice/group).

Supplmentary Figure 5

Supplemental Figure 5. Sex influences white adipose tissue STAT3 activation during sepsis. Representative images of epididymal white adipose tissue STAT3 phosphorylation at tyrosine 705 staining in (A and C) male WT and A-STAT3 KO mice at 0h and 18h after CLP, (B and D) female WT and A-STAT3 KO mice at 0h and 18h after CLP, and (E) male and female, septic WT and septic A-STAT3 KO mice. Images at 40x magnification. Scale bar represents 25 μm in length. Each individual data point represents one mouse. Data were analyzed by two-way ANOVA on Ranks with Holm-Sidak method for all pairwise multiple comparisons. *P ≤ 0.05 vs. 0h within genotype, #P ≤ 0.05 vs. WT within 18h CLP, &P ≤ 0.05 vs. males within genotype. White boxes = WT, gray boxes = A-STAT3 KO (n = 3–9 mice/group).

Supplmentary Figure 6

Supplemental Figure 6. STAT3 phosphorylation at tyrosine 705 is inhibited in adipocytes but not the stromal vascular fraction of A-STAT3 KO mice. Representative Western blot analysis for epididymal white adipose tissue (A) male adipocyte fraction, (B) male stromal vascular fraction, (C) female adipocyte fraction, and (D) female stromal vascular fraction pSTAT3 Tyr705, STAT3, and Beta Actin in WT and A-STAT3 KO mice at 0h and 18h after CLP. n = 3–4 mice/group.

Supplmentary Figure 7

Supplemental Figure 7. Effects of adipocyte STAT3 inhibition on plasma cytokine levels during sepsis. (A) Male plasma TNF-α log transformed, (B) female plasma TNF-α log transformed, (C) male plasma IL-6 log transformed, (D) female plasma IL-6 log transformed, (E) male plasma leptin log transformed, and (F) female plasma leptin log transformed, in WT and A-STAT3 KO mice at 0h and 18h after CLP. Each individual data point represents one mouse. Data were analyzed by two-way ANOVA with Holm-Sidak method for all pairwise multiple comparisons. *P ≤ 0.05 vs. 0h within genotype. White boxes=WT, gray boxes=A-STAT3 KO (n=5–10 mice/group).

Supplmentary Figure 8

Supplemental Figure 8. Effects of adipocyte STAT3 inhibition on lung neutrophil infiltration during sepsis. Lung MPO in (A) male and (B) female WT and A-STAT3 KO mice at 0h and 18h after CLP. Each individual data point represents one mouse. Data were analyzed by two-way ANOVA on Ranks with Holm-Sidak method for all pairwise multiple comparisons. *P ≤ 0.05 vs. 0h within genotype, #P ≤ 0.05 vs. WT within 18h after CLP. White boxes=WT, gray boxes=A-STAT3 KO (n=3–10 mice/group).

Supplmentary Figure 9

Supplemental Figure 9. Hematoxylin and eosin staining of the lung. Representative images of lung hematoxylin and eosin staining at 0h and 18h after CLP in (A and C) male WT and A-STAT3 KO mice and (B and D) female WT and A-STAT3 KO mice. Images at 40X magnification. Scale bar represents 100 μm in length. Each individual data point represents one mouse. Data were analyzed by two-way ANOVA with Holm-Sidak method for all pairwise multiple comparisons. *P ≤ 0.05 vs. 0h within genotype, #P ≤ 0.05 vs. WT within 18h after CLP. White boxes = WT, gray boxes = A-STAT3 KO (n = 4–7 mice/group).

Supplmentary Figure 10

Supplemental Figure 10. Effects of adipocyte STAT3 inhibition on liver neutrophil infiltration and plasma ALT and AST levels during sepsis. Liver MPO in (A) male and (D) female WT and A-STAT3 KO mice, plasma ALT levels in (B) male and (E) female WT and A-STAT3 KO mice, and plasma AST levels in (C) male and (F) female WT and A-STAT3 KO mice at 0h and 18h after CLP. Each individual data point represents one mouse. Data were analyzed by two-way ANOVA or two-way ANOVA on ranks with Holm-Sidak method for all pairwise multiple comparisons. *P ≤ 0.05 vs. 0h within genotype, #P ≤ 0.05 vs. WT within CLP time. White boxes = WT, gray boxes = A-STAT3 KO (n = 3–10 mice/group).

Supplmentary Figure 11

Supplemental Figure 11. Hematoxylin and eosin staining of the liver. Representative images of liver hematoxylin and eosin staining at 0h and 18h after CLP in (A and C) male WT and A-STAT3 KO mice and (B and D) female WT and A-STAT3 KO mice. Images at 40x magnification. Scale bar represents 100 μm in length. Each individual data point represents one mouse. Data were analyzed by two-way ANOVA with Holm-Sidak method for all pairwise multiple comparisons. *P ≤ 0.05 vs. 0h within genotype. White boxes = WT, gray boxes = A-STAT3 KO (n = 4–7 mice/group).