In accordance with the Declaration of Helsinki, informed consent for all human participants was obtained following the Institutional Review Board–approved protocols at the Forsyth Institute and the University of Kentucky. Peripheral blood was obtained from normoglycemic subjects who were lean (avg: 41.2 years; BMI 23.8 kg/m²) or obese/normal glucose-tolerant (NGT; avg: 46.6 years; BMI 39.7 kg/m²). A1c or HbA1c is a blood test that measures average blood sugar levels. The A1c test showed that both lean and obese subjects were glucose-tolerant. Additional characteristics are shown in Table 1. Exclusion criteria were smoking, recent use of antibiotics, or anti-inflammatory medications, that is, NSAIDs/steroids, colds/flu within the past 2 weeks, type 2 diabetes or diabetes medications, and any history of cancer, hyperglycemia, and auto-immune diseases.

Fifty milliliter of peripheral blood was collected into acid/citrate/dextrose containing tubes by venous puncture. PBMCs were purified by histopaque and frozen at −80°C under controlled cooling conditions in a Mr. Frosty apparatus (Nalgene). For multi-week storage, cells were moved to −190°C following 1–2 days at −80°C. CD4⁺ T cells were isolated from PBMCs by negative selection using MACS columns (Miltenyi Biotech). PBMCs or isolated CD4⁺ T cells were stimulated in vitro for 40 h with T cell–activating αCD3/αCD28 Dynabeads (Thermo Fisher Scientific, 11132D) at 2 μL Dynabeads per 100k cells. In some cultures, cells obtained from lean adults were treated with either 300 µM palmitate (pal) (C16:0) coupled to fatty acid–free bovine serum albumin (BSA) at a ratio of 2 mol palmitate to 1 mol BSA, or 300 µM oleate (Ole). The fatty acid concentrations mimic concentrations achievable in human serum. 1% BSA was used as vehicle (Veh). The following treatments were added to some cultures as indicated: 100 mM trehalose (Tre) or 3 µM Mitocur-1 (Mtcur) for 40 h. The ROS scavenger, tempol (Temp; 100 µM), was added 3 h after Dynabeads to avoid excessive cell death. A minimum of 3 hours without ROS quenchers (i.e., tempol) is required post activation for survival of the cells. Thus, cells remained in Temp for 37 h. The late-stage autophagy blocker bafilomycin (Baf; 1 µM) was added to some cultures treated with Tre for the last 30 min, 1 h, or 4 h of incubation.

All treatments were added to our standard culture media of RPMI with 5 mM glucose (normoglycemic), 10% heat-inactivated FBS, and antibiotics. Supernatants were collected and stored at −80°C and shipped on dry ice to the University of Kentucky for performing the Luminex bio-plex assay to assess the levels of cytokines. Cells were assayed as outlined below.

In vitro activated CD4⁺ T cells obtained from NGT subjects and cells from lean subjects after in vitro fatty acid treatments (±Trehalose,± Mtcur,± tempol, ± Baf) were plated on coverslips coated with poly-D-lysine in 6-well or 12-well plates. After 40 h, the cells were briefly centrifuged (1,200 rpm, 10 min), washed two times with ×1 PBS, and incubated in 4% paraformaldehyde for 30 min at RT. The coverslips were washed ×2 with PBS and 0.1% triton X-100 (PBST) and were blocked for at least 30 min in 5% BSA/PBST. Antibodies to TOM20 (Santacruz Biotechnology, Dallas, TX), p-STAT3 Ser 727 (Cell signaling technology, Danvers, MA), and LC3 (Millipore Sigma, Burlington, MA) were added at 1:50 dilution with incubation overnight at 4°C. The coverslips were washed ×2 with PBST and incubated with fluorophore-tagged secondary antibodies (anti-mouse Alexa 488 or anti-rabbit Alexa 680) (Rockland Immunochemicals, Limerick, PA) for 2 h at RT. The coverslips were washed ×2 with PBST and mounted on glass slides using Fluoromount-G (Southern Biotech, Birmingham, AL). Cell imaging under a 63× oil immersion lens was performed in a Zeiss confocal microscope. Seven to ten fields per slide were imaged were imaged on N = 3–4 slides, and data were analyzed using ImageJ (Kirber et al., 2007), (Vereb et al., 2000). Pearson’s colocalization coefficient (PCC) is a well-established method that quantifies the degree of overlap between fluorescence in the channels. The PCC is unaffected by changes to the offset and independent of gain. The PCC has a range of +1 (perfect correlation) to −1 (perfect but negative correlation). The PCC is not sensitive to differences in signal intensity between the components of an image caused by different labeling with fluorochromes, photobleaching, or different settings of amplifiers (Adler and Parmryd, 2010).

Immunoblotting quantified protein expression was published (Bharath et al., 2020). Briefly, 30 μl of ×1 cell lysis buffer (Cell signaling technology, Danvers, MA) was added to 1 × 10⁶ cells and incubated on ice for 20 min. Cells were centrifuged at 13,000 rpm for 20 min, and the supernatant was collected. A bicinchoninic assay (Thermo Fisher Scientific, 23225) was used to assess the protein concentration. Fifteen μg protein was loaded onto polyacrylamide gels, and electrophoresis was performed at 100 V for 1 h. The transfer of protein to the polyvinylidene difluoride (PVDF) membrane was performed at 35 V for 5 h. The membrane was blocked for 30 min at room temperature (RT) in the blocking buffer containing 2% bovine serum albumin in TBST followed by overnight incubation at 4^○C in the respective primary antibodies. The membrane was washed ×3 with 1X TBST and incubated with secondary antibodies for 2 h at RT and then imaged. Table 2 lists the antibodies used in this study. All antibodies were used at a dilution of 1:500 except β-actin, which was used at 1:10,000. We quantified protein expression on Western blots using Image studio lite (Licor, Lincoln, NE).

ROS was quantified using 10 µM chromomethyl 2′,7′ –dichlorofluorescin diacetate (CM-H₂DCFDA), fluorescence (Sigma-Aldrich, D6883), 5 µm MitoSox red (Thermo Fisher Scientific, M36008), and 10 µM dihydroethidium fluorescence (DHE) (Thermo Fisher Scientific, D11347) and normalized to cell number. 50 µM tert-butyl hydrogen peroxide (TBHP) (Sigma Aldrich, 416665) was used as a positive control for DCFDA assay. ROS was quantified in the activated PBMCs from NGT subjects and lean subjects after fatty acid treatments (± Tre and ± Temp).

Cell viability was assessed via the Alamar blue assay and trypan blue exclusion according to the manufacturer’s directions after different treatments.

Cytokine production was assessed in supernatants by the Luminex multiplex assay (Milliplex human Th17 25-plex kit, Millipore) and by an enzyme-linked immunosorbent assay (Quantikine ELISA, R&D systems, D1700 and D6050) as we published (Bharath et al., 2020). Outcomes from wells with <35 beads read for each analyte were excluded from the analysis. The plates were washed between incubations using a BioTek 406 Touch plate washer (BioTek) and read using the Luminex FlexMap 3D system (Luminex).

Data are presented as the mean ± standard error of the mean (SEM). The nonparametric Kruskal–Wallis and Dunn’s post hoc tests or one-way ANOVA (for cytokines) followed by Bonferroni post hoc analysis were performed using Graph-Pad Prism 7.03 (GraphPad Software). The Mann–Whitney test compared means of two values. Significance was accepted when p < 0.05.

P-Serine (p-Ser) 727 STAT3, a modification required for the translocation of STAT3 into mitochondria, was assessed in CD4⁺ T cells and PBMCs from NGT middle-aged adults (Table 1). We detected significantly more p-Ser 727 STAT3 (Figure 1A) and higher mitochondrial accumulation of STAT3 (Figure 1B) in CD4⁺ cells from NGT adults. To model the possibility that obesity-induced excess in circulating fatty acids promotes mitoSTAT3 accumulation, cells from lean subjects were treated with a physiologically relevant dose of palmitate or oleate prior to analysis. Palmitate induced a robust increase in p-Ser 727 STAT3 (Figure 1C, in PBMCs) in the cytoplasm and the nucleus of CD4⁺ T cells, as evidenced by fluorescence observed in the periphery of the cells and p-Ser 727 STAT3 fluorescence merging with the nuclear stain (data not shown) in a confocal analysis (Figure 1D). Nuclear stain in general is not shown in the images to delineate the STAT3 fluorescence. Confocal microscopy indicated more colocalization of p-Ser-727 STAT3 with mitochondrial translocase of the outer protein membrane, TOM20, following palmitate or oleate treatment of CD4⁺ T cells from lean subjects (Figure 1E). Cell viability was not affected by treatment with fatty acids (Supplementary Figure S3). Collectively, these data support the interpretation that mitochondrial translocation of STAT3 occurs in response to a challenge by saturated and monounsaturated fatty acids in vitro, with likely parallels in vivo.

To assess the functional consequence of the blocking mitochondrial translocation of STAT3, cells were incubated for 40 h with the mitoSTAT3 inhibitor Mtcur. Lower phosphorylation (Figure 2A), mitochondrial colocalization (Figure 2B), and lower levels of proinflammatory cytokine production (Figure 2C) were observed in mitocur-treated cells from NGT and in lean cells co-treated with Mtcur and palmitate (Figure 2D). Mtcur did not alter cytokine production in cells treated with oleate (Figure 2C). The bio-plex assay indicated Mtcur did not uniformly lower cytokine production (Supplementary Figure S1). These data support the conclusion that mitochondrially located STAT3 promotes inflammation in NGT adults and in palmitate-treated cells from lean subjects.

To identify the mechanism(s) that promote mitochondrial localization of STAT3, we assessed the processes that are known to promote similar STAT3 translocation in other cell types, including the cellular recycling process of autophagy. We initially tested the effect of trehalose, a nutraceutical known to enhance autophagy on some cell types, on STAT3 translocation and cytokine production. Trehalose treatment of cells from NGT adults prevented obesity-induced phosphorylation and mitochondrial translocation of STAT3, while lowering inflammatory cytokine production (Figures 3A–C). Trehalose also prevented palmitate-induced Ser 727 STAT3 phosphorylation (Figure 4A), STAT3 mitochondrial translocation (Figure 4B), and inflammation (Figure 4C) but was unable to prevent oleate-induced STAT3 phosphorylation and mitochondrial translocation (Figures 4D,E). Supplementary Figure 2A shows the production of other cytokines in response to palmitate and oleate ± trehalose. We conclude that trehalose lowers STAT3-mitochondrial translocation and Th17-associated cytokine production in T cells exposed to saturated but perhaps not monounsaturated lipid excess.

To identify how the pro-autophagy actions of trehalose account for functions shown in Figure 4, we assessed autophagy in cells activated ± trehalose by immunofluorescence and staining for microtubule associate protein-1 light chain 3B (LC3B) punctuate structures (puncta). Palmitate reduced LC3 puncta formation (Figure 5A), but oleate did not (Figure 5B), indicating saturated but not monounsaturated fatty acids compromise autophagy. However, the autophagic flux assessment using the late-stage autophagy inhibitor bafilomycin A1 (Baf) showed that trehalose did not enhance palmitate-compromised autophagy but instead stalled the clearance of the autophagosome (Figure 5C). Furthermore, autophagy was not impaired in cells from NGT adults (Figure 5D), showing that the effect of trehalose on STAT3 activation and mitochondrial translocation is not secondary to impaired autophagy. These data also indicate (unsurprisingly) that palmitate treatment does not perfectly mimic obesity-associated changes in CD4⁺ Tcell physiology.

Another well-characterized function of trehalose is its ability to function as an antioxidant, as reported in many studies. Trehalose reduced obesity (Figure 6A) and palmitate-induced (Figure 6B) increase an cellular peroxide levels. Peroxide production did not increase upon oleate treatment; however, a numerical increase in peroxide was observed in cells cotreated with oleate and trehalose (Figure 6B). Neither mitochondrial superoxide nor cellular superoxide was altered in NGT or after treatment with fatty acids, and trehalose did not further affect superoxides in such cultures (Figure 6C and Figure 6D). Together, these data point to the ability of trehalose to lower mitochondrial STAT3 translocation and thus perhaps the production of Th17-associated cytokines by lowering cellular peroxides.

To identify the mechanism that obesity and fatty acids use to promote the mitochondrial translocation of STAT3 and to independently test if trehalose lowers mitoSTAT3 due to the ability to scavenge peroxides, we tested the impact of tempol (Temp), a well-established ROS scavenger, on p-Ser727-STAT3 localization and cytokine production. Temp reduced NGT (Figure 7A) and palmitate-induced peroxide production (Figure 7B). Interestingly, Temp slightly increased mitochondrial superoxide production in NGT (Figure 7C) but had no effect on palmitate- or oleate-treated cells (Figure 7D). Temp prevented mitochondrial translocation of p-Ser 727 STAT3 in cells from NGT adults (Figure 7E) and in both palmitate- or oleate-treated cells (Figure 7F). Temp reduced IL-17 production in cells from NGT (Figure 7G) and palmitate-treated cells from lean subjects (Figure 7H). Collectively, these data show that either scavenging peroxides (but not superoxides) or blocking mitoSTAT3 (Figure 2) prevents IL-17A secretion. We conclude that the mitochondrial translocation of STAT3 favors cellular peroxide formation, which promotes inflammatory cytokine production by T cells primed by high saturated fatty acids.