Endotoxin (lipopolysaccharide, LPS, O55:B5) and brefeldin A were from Sigma-Aldrich (Taufkirchen, Germany). 12-O-tetradecanoylphorbol-13-acetate (TPA) was from Enzo Life Sciences (LÖrrach, Germany) and A23187 from AppliChem (Karlsruhe, Germany). Cyclosporin A (CsA) and FK506 were purchased from Calbiochem-Novabiochem (Bad Soden, Germany). Human IL-22 and interferon (IFN)γ were obtained from Peprotech Inc. (Frankfurt, Germany). Murine (αCD3-#17A2, αCD28-#37.51) and human (αCD3-#OKT3, αCD28-#28.2) agonistic anti-CD3 and anti-CD28 antibodies were from BioLegend (San Diego, CA, USA).

All animal procedures were approved by local authorities (“Regierungspräsidium Darmstadt”) and are in accordance with National Institutes of Health guidelines. For experiments, 10- to 12-week-old C57Bl/6 male mice (MFD-Diagnostics GmbH, Wendelsheim, Germany) were transferred individually into cages early morning. The body weight and core temperature was determined using laboratory scales and a TH-5 + RET-3 mouse thermometer with rectal probes (Physitemp Instruments Inc., Clifton, NJ, USA). Mice were injected i.p. with LPS (1 µg/g body weight) or PBS and kept at either standard room temperature (RT, 23°C) or at 4°C with access to water only (36). After 5 h, mice underwent short isoflurane (Abbott, Wiesbaden, Germany) anesthesia and were sacrificed. Liver, lungs, spleen, colon, cecum, and blood plasma were snap frozen in liquid nitrogen and stored at −80°C.

For isolation of PBMC, heparinized blood was taken from healthy donors. This procedure and the respective consent documents were approved by the “Ethik Kommission” of the University Hospital Goethe-University Frankfurt. PBMC were isolated from peripheral blood using Histopaque-1077 (Sigma-Aldrich) according to the manufacturer’s instructions. Untouched CD3⁺ T-cell were isolated from PBMC using the Pan-T-cell isolation kit according to the manufacturer’s instructions (Miltenyi, Bergisch Gladbach, Germany). Mean purity was 96.0 ± 0.7% (n = 37) determined by FACS analysis using anti-CD3-PerCP/Cy5.5-#UCHT1 (BioLegend). Monocyte-depleted PBMC were generated using anti-CD14 beads (Miltenyi) with a mean depletion efficiency of 98.1 ± 0.7% (n = 7) as assessed by FACS analysis using an anti-CD14eFluor450-#61D3 antibody (eBioscience, Frankfurt, Germany). Cells were resuspended in RPMI 1640 supplemented with 10 mM HEPES, 100 U/ml penicillin, 100 µg/ml streptomycin, and 1% human serum (Life Technologies, Darmstadt, Germany) and seeded at 3 × 10⁶ cells/ml in round-bottom polypropylene tubes (Greiner, Frickenhausen, Germany).

Spleens obtained from 8- to 12-week-old male C57Bl/6 mice (MFD-Diagnostics GmbH) were excised and transferred to 5 ml ice-cold RPMI 1640 medium without FCS. Tissue was destroyed over a nylon cell strainer (70 µm; BD Biosciences, Heidelberg, Germany). Cell suspensions were centrifuged at 500 g for 5 min at 4°C and resuspended in 2 ml 0.83% NH₄Cl for 2 min at RT. Red blood cell lysis was stopped by adding 10 ml cold RPMI 1640 medium without FCS. Splenocytes were collected by centrifugation, washed once with RPMI and resuspended in RPMI 1640, supplemented with 10% heat-inactivated FCS and 100 U/ml penicillin, 100 µg/ml streptomycin. Cells were seeded on 24 well polystyrene plates (Greiner) with 0.5 ml media in a concentration 6 × 10⁶ cells/ml.

Jurkat T cells (ATCC-TIB-152) were obtained from the American Type Culture Collection (Manassas, VA, USA) and cultured in RPMI 1640 (Life Technologies) supplemented with 100 U/ml penicillin, 100 µg/ml streptomycin, and 10% heat-inactivated FCS (Life Technologies). For experiments, cells were seeded on 6-well polystyrene plates (Greiner) at a density of 2.5 × 10⁶ cells/ml. DLD1 colon epithelial/carcinoma cells (Center of Applied Microbiology and Research, Salisbury, UK), Caco2 colon epithelial/carcinoma cells, and HepG2 hepatoma cells (German Collection of Microorganisms and Cell Cultures, Braunschweig, Germany) were maintained in DMEM supplemented with 100 U/ml penicillin, 100 µg/ml streptomycin, and 10% heat-inactivated FCS (Life Technologies). For experiments, cells were plated on 6-well polystyrene plates (Greiner) and used in subconfluent condition.

According to the protocols indicated in the figure legends, cells (PBMC, epithelial-like cell lines) were cultivated in parallel at 30, 33, or 37°C incubator temperature. Incubators used for different temperatures were switched occasionally in order to exclude incubator effects on cell behavior different from incubator temperature.

Murine and human IL-22 (R&D-Systems, Wiesbaden, Germany) and human IL-8 (BD Biosciences) secretion were determined by ELISA according to the manufacturer’s instructions.

Peripheral blood mononuclear cells were kept as unstimulated control or stimulated for 7 h with agonistic anti-CD3 (0.2 µg/ml)/-CD28 (0.02 µg/ml) antibodies at 37 or 33°C. Thereafter, brefeldin A (2 µg/ml) was added for another 4 h, followed by intracellular staining and flow cytometry. After harvesting, PBMC were reconstituted in FACS buffer (1 × PBS + 1% FCS) and stained with surface marker antibody (anti-CD4-PE-Cy7-#SK3, eBioscience) for 30 min on ice. Thereafter, PBMC were fixed and permeabilized [BD Cytofix/Cytoperm Kit (BD Biosciences)], followed by resuspension in FACS buffer, and intracellular staining (2 h on ice) using IL-22-PE-#22URTI or IFNγ-FITC-B27 (both eBioscience) and flow cytometry with gates set to exclude cell debris.

Total RNA was extracted from homogenized mouse tissue or cultured cells using Tri-Reagent according to the manufacturer’s instructions (Sigma-Aldrich). Tissues were homogenized using OMNI TIP Homogenizing KIT (Kennesaw, GA, USA). 0.5 µg RNA was transcribed using random hexameric primers and Moloney Murine Leukemia Virus Reverse Transcriptase (Thermo Scientific, Darmstadt, Germany) according to the manufacturer’s instructions. cDNA was amplified using assay-on-demand kits (Taqman probes/assay kit from Thermo Scientific) and an AbiPrism 7500 Fast Sequence Detector (Thermo Scientific). During real-time PCR, changes in fluorescence are caused by the Taq polymerase degrading a probe containing a fluorescent dye [glyceraldehyde 3-phosphate dehydrogenase (GAPDH): VIC; all others: FAM]. Two initial steps at 50°C for 2 min and 95°C for 20 s were followed by 40 cycles at 95°C for 3 s and 60°C for 30 s. Target mRNA was normalized to that of GAPDH and quantified by the 2^–ΔCT method (raw data, Figures 1, 2, 4 and 5) or the 2^–ΔΔCT method (fold-induction, Figure 6). The following probes were used: hs-GAPDH (4310884E), hs-IL-22 (Hs01574152_g1), hs-IL-10 (Hs99999035_m1), hs-IFNγ (Hs00174143_m1), α1ACT (Hs00153674_m1), hs-IL-8 (Hs00174103_m1), hs-IL-2 (Hs00174114_m1), mm-GAPDH (4352339E), mm-IL-22 (Mm00444241_m1), mm-MIP2 (Mm00436450_m1), mm-IL-10 (Mm00439614_m1), mm-TNF-α (Mm00443285_m1), mm-IFNγ (Mm01168134_m1), and mm-IL17A (Mm00439618_m1). Primers and probe for IL-18BPa were designed using Primer Express (Applied Biosystems) according to AF110798: forward, 5′-ACCTCCCAGGCCGACTG-3′; reverse, 5′-CCTTGCACAGCTGCGTACC-3′; probe 5′-CACCAGCCGGGAACGTGGGA-3′. GAPDH was not a target of regulation by hypothermia under all conditions investigated (data not shown).

Immunoblot analysis for cellular STAT1/3 in DLD1, Caco2, and HepG2 cells was performed as previously described (37) using total cell lysis buffer [150 mM NaCl, 1 mM CaCl₂, 25 mM Tris–Cl (pH 7.4), 1% Triton X-100] supplemented with protease inhibitor cocktail (Roche Diagnostics, Mannheim, Germany), DTT, Na₃VO₄, PMSF (each 1 mM), and NaF (20 mM). Antibodies: total STAT3-#124H6 (mouse monoclonal antibody); total STAT1, pSTAT1-Y701-#D4A7 (both rabbit polyclonal antibodies); pSTAT3-Y705-#D3A7 (rabbit monoclonal antibody); all from Cell Signaling, Frankfurt, Germany. For detection of total STAT1 or total STAT3 blots were stripped and reprobed.

Isolation of nuclei (38) for immunoblot analysis of nuclear NFAT-c2 in human T-cells was performed by lysis using nuclear extraction buffer A (10 mM HEPES at pH 7.9, 10 mM KCL, 0.1 mM EDTA, 0.1 mM EGTA) supplemented with protease inhibitor cocktail (Roche Diagnostics). After 10 min on ice and addition of Triton X-100, nuclei were collected by centrifugation at 12,000 g for 1 min at 4°C. Pellets containing nuclei were resuspended in nucleic-lysis buffer C (20 mM HEPES at pH 7.9, 0.4 M NaCl, 1 mM EDTA, 1 mM EGTA, 25% glycerin) supplemented with protease inhibitor cocktail (Roche Diagnostics). Antibodies: NFAT-c2-#4G6-G5, mouse monoclonal antibody (Santa Cruz Biotechnology, Heidelberg, Germany); β-actin-#AC15, mouse monoclonal antibody (Sigma-Aldrich). For detection of NFAT-c2 and β-actin on the same blot, the blot was cut. Data quantifications were performed by Quantity-One analysis software (Bio-Rad, Munich, Germany).

Data are shown as group median, means ± SD, or means ± SEM and presented as [raw data], [fold-induction], [percent], [pg/ml], [ng/ml], and [Adj.Vol. INT*mm2]. The D’Agostino–Pearson normality test was used to assess data distribution. Statistical analysis was performed on raw data as indicated in the legends by one-way ANOVA with post hoc Bonferroni correction, unpaired Student’s t-test, or Mann–Whitney U-test. Differences were considered significant in case of P values below 0.05 (Prism 5.0, GraphPad, La Jolla, CA, USA).

Experimental endotoxemia is regarded a standard model for the hyper-inflammatory phase of sepsis. Both, rodent endotoxemia (39) and sepsis (40) associate with enhanced IL-22 production. To investigate il22 gene expression under the influence of cold stress, PBS-treated control mice and mice undergoing endotoxemia were exposed to either an ambient temperature of 4°C or to RT (23°C) for 5 h. Notably, a low endotoxin dosage of 1 µg/g, by itself unable to induce hypothermia, was administered for induction of systemic inflammation. Analysis of rectal temperature after 5 h at 4°C revealed that only endotoxemic mice (n = 9) but not PBS-treated control mice (n = 8) developed mild-to-moderate hypothermia with a significant drop from 35.9± 0.3°C to a core temperature of 30.9 ± 0.9°C (P < 0.001, Student’s t-test). Under conditions of RT, core body temperature of PBS-treated control (n = 8) or endotoxemic mice (n = 9) was indistinguishable at 35.8 ± 0.4 or 35.5 ± 0.4°C, respectively. In accord with previous in vivo data on IL-10 and anti-inflammatory properties of hypothermia (26–35), enhanced splenic il10 gene expression was observed in endotoxemic mice exposed to 4°C (Figure 1A). To verify the hypothesis that IL-22 is affected by hypothermia, these same specimens were analyzed for expression of this cytokine. As shown in Figure 1B, splenic il22 gene expression was significantly upregulated in endotoxemic mice exposed to 4°C ambient temperature. On the contrary, in endotoxemic mice, upregulation of splenic IL-17A mRNA [0.40 × 10^–4 versus 0.42 × 10^–4 (median of target gene expression normalized to GAPDH) for RT versus 4°C ambient temperature (n = 9), not significantly different by Mann–Whitney U-test; IL-17A mRNA, undetectable in PBS-treated control mice—irrespective of ambient temperature] and tumor necrosis factor (TNF)-α mRNA [6.7 ± 0.9- versus 6.4 ± 1.7-fold-induction for RT (n = 8) versus 4°C ambient temperature (n = 9)] was unaffected by changes in ambient temperature. Since local IL-22 is crucial for maintenance of intestinal barrier function (41) and sepsis pathogenesis (16), colonic IL-22 expression was determined and found likewise upregulated upon cold stress (Figure 1C). Induction of colonic TNF-α mRNA by endotoxemia was, in accord with observations in spleen, not further increased by changes in ambient temperature (data not shown). Notably, as compared to RT, an ambient temperature of 4°C without endotoxemia failed to significantly affect splenic and colonic il22 expression. Specifically, using the current protocol, IL-22 mRNA was neither detectable in total colonic RNA obtained from PBS-treated control mice exposed to RT (n = 6) nor in splenic specimens irrespective of the tested ambient temperature (n = 8). In colonic tissues obtained from PBS-treated control mice exposed to 4°C, il22 expression was very low and barely detectable with a median of 0.01 × 10^–5 (n = 6) for IL-22 mRNA normalized to that of GAPDH. Notably, very low il22 gene expression in colonic tissue of healthy untreated mice concurs with previous observations (42).

Exposure of mice to cold stress engages a complex systemic response that involves, among others, activation of the β-adrenergic/cAMP-axis (20) with its documented potential for immunomodulation (43), possibly by upregulation of IL-10 (43, 44). In order to evaluate whether hypothermia is able to directly upregulate il22 expression on the level of cultured murine leukocytes, freshly isolated splenocytes were stimulated using agonistic anti-CD3/-CD28 antibodies in the context of an ambient temperature of either 37 or 33°C—the latter condition resembling mild hypothermia. Figure 1D demonstrates that mild hypothermia amplifies IL-22 mRNA expression by activated splenocytes. Moreover, we also verified the potential of hypothermia to upregulate IL-22 protein release (Figure 1E). Whereas the murine functional IL-8 homolog and stress-responsive parameter macrophage inflammatory protein (MIP)-2 displayed similar upregulation by hypothermia (Figure 1F), expression of IFNγ mRNA was retarded under these same conditions (Figure 1G). In light of the proinflammatory pathological functions of IFNγ (45), this latter observation agrees with immunosuppressive properties of hypothermia (46). Taken together, data suggest that cold stress and hypothermia can serve as a cofactor enhancing murine il22 gene expression. Since stimulatory effects of hypothermia on IL-22 are detectable on cell culture level (Figures 1D,E), upregulation of the cytokine in vivo may also be independent from activation of the β-adrenergic/cAMP-axis.

To broaden aforementioned murine in vitro and in vivo data, experiments were performed using human PBMC. We confirm previous observations (47) on IL-22 secretion by PBMC in response to stimulatory anti-CD3/-CD28 antibodies at physiological 37°C. Herein, we demonstrate that cultivation of PBMC at 33°C significantly amplifies IL-22 protein secretion (Figure 2A). Enhanced expression of IL-22 under the influence of mild hypothermia and anti-CD3/-CD28-stimulation was likewise detectable on mRNA level (Figure 2B). Upregulation of IL-22 by hypothermia was even more pronounced by further reducing the cultivation temperature to 30°C (Figure 2C) indicating a concurrent response upon more severe hypothermic culture conditions. IL2 gene expression was, like IL22, enhanced upon cultivation at 33°C (Figure 2D). Notably, regulation by mild hypothermia displayed specificity since neither expression of IFNG (Figure 2E) nor that of IL10 (Figure 2F) was affected by cultivation of anti-CD3/-CD28-stimulated PBMC at 33°C. Unresponsiveness of IL-10 concerning hypothermia in vitro contrasts with robust IL-10 upregulation in vivo [(28–35) and Figure 1A]. IL-10 production in vivo is, however, induced by the β-adrenergic/cAMP-axis (44), which is activated by acute hypothermia (20). This likely explains the observed differences compared to in vitro cultured PBMC. In contrast to IFNγ and IL-10, mRNA expression (Figure 2G) and release (Figure 2H) of IL-8 was increased by mild hypothermia. This latter observation concurs with the notion that IL8 is a stress-responsive gene (48).

Stimulation of human PBMC by treatment with anti-CD3/-CD28 antibodies and associated IL-22 secretion has been linked to the CD4⁺ memory T-cell compartment being the chief IL-22 PBMC-source under conditions of polyclonal activation (47). Herein, intracellular cytokine staining demonstrated that upregulation of IL-22⁺ cells by anti-CD3/-CD28 in total PBMC (Figure 3A) and the CD4⁺ PBMC fraction (Figure 3B) is further amplified by mild hypothermia. In contrast, exposure to 33°C did not significantly affect numbers of IFNγ⁺ cells in total PBMC (Figure 3C) thus confirming aforementioned mRNA data (Figure 2E).

Monocytes have previously been identified as crucial cellular component supporting T-cell-derived IL-22 production by PBMC in response to diverse stimuli (49, 50). To specify their role in mild hypothermia enhancing IL-22 production, PBMC were depleted from monocytes. As shown in Figure 4A, depletion of monocytes entirely prevented upregulation of IL22 expression under the influence of hypothermia. To further investigate this matter, whole T-cells were isolated from PBMC and cultivated thereafter at either 37 or 33°C. Analysis after stimulation of isolated T-cells by anti-CD3/-CD28 antibodies actually revealed no significant difference in IL22 expression between both ambient temperatures (Figure 4B, left panel). Notably, whole PBMC from these same donors simultaneously cultivated displayed amplified IL-22 mRNA expression at 33°C (Figure 4B, right panel). Human leukemic Jurkat T-cells activate IL22 gene expression in response to polyclonal stimulation (18). In accord with aforementioned data on isolated primary T-cells, mild hypothermia left IL22 expression by TPA/A23187-stimulated Jurkat T-cells unaffected (Figure 4C). Taken together, observations indicate that upregulation of IL-22 by mild hypothermia is not a T-cell autonomous process but requires interactions between T cells and additional PBMC subsets, especially monocytes.

The calcineurin/NFAT inhibitors cyclosporin A (CsA) and FK506 (tacrolimus) (51) potently suppress IL22 expression by activated human Jurkat T-cells, primary T-cells, and PBMC (18, 52). This observation concurs with inhibition of T-cell-derived IL-22 in psoriatic skin of tacrolimus-treated mice (53) and downregulation of IL-22 in patients undergoing CsA therapy (54, 55). Moreover, NFAT-c2 (51) binding to a specific site within the human IL22 promoter contributes to gene activation (18), conceivably by cooperation with an adjacent binding site for ATF2-jun heterodimers (56). Since aforementioned data suggested IL-22 as NFAT-inducible cytokine, CsA was tested in the current experimental protocol. In fact, coincubation with CsA not only impaired IL22 expression by activated PBMC upon normothermia but entirely prevented potentiation of gene induction at 33°C (Figure 5A). In contrast, CsA failed to suppress anti-CD3/-CD28-induced IL-8, a monocyte-derived inflammatory chemokine that was determined to control for unspecific inhibitory effects of the agent on PBMC cytokine expression. Interestingly, CsA actually amplified IL8 expression upon hypothermia (Figure 5B), which corresponds to observations on human smooth muscle cells where stimulation of activator protein-1 by CsA enforces IL-8 (57). Notably, the alternate NFAT antagonist FK506 (tacrolimus) displayed very similar inhibitory action on IL-22 (Figure 5C). As expected, we confirm previous observations (58) on potent suppression of IFNγ production by PBMC under the influence of CsA (Figure S1 in Supplementary Material).

In order to more directly investigate the role of NFAT-c2 in this context, immunoblot analysis was performed using T-cells isolated instantly after stimulation of PBMC at 33 or 37°C. Herein, we demonstrate that hypothermia increases nuclear NFAT-c2 accumulation, thus transcription factor activation, in primary human T-cells after activation by anti-CD3/-CD28 (as part of whole PBMC). Notably, results obtained from five different donors indicated the most prominent effect of low ambient temperature at 5–7 h after onset of polyclonal T-cell stimulation (Figure 5D). Densitometric quantification of these experiments is shown in Figure 5E. Data altogether suggest that exposure of PBMC to mild hypothermia in CsA/FK506-sensitive manner supports IL22 expression, which associates with and is likely mediated by enhanced nuclear NFAT-c2.

To investigate effects of hypothermia on IL-22 biological activity, IL-22-responsive epithelial-like human Caco2 and DLD1 colon carcinoma cells (59) as well as HepG2 hepatoma cells (13) were adjusted for 6 h to an ambient temperature of either 30 or 37°C followed by stimulation with IL-22. As detected by analysis of pSTAT3 after 1 h, early IL-22 signal transduction was significantly enhanced upon hypothermia in all three cell types investigated. Notably, this effect vanished at the later 3-h time point. Figure 6 displays densitometric quantification of results (three independently performed experiments per cell line) obtained from DLD1 (Figure 6A), Caco2 (Figure 6B), and HepG2 cells (Figure 6C). Representative immunoblots for each cell line are shown in Figure 6D. To determine consequences of temperature-sensitive STAT3 activation for downstream gene regulation, prototypic IL-22-inducible α1-antichymotrypsin (α1ACT) (1, 14) was investigated in DLD1 cells. Time course analysis revealed enhanced expression of α1ACT mRNA at 30°C, a phenomenon that started at 4 h and persisted thereafter (Figure 6E). The specificity of hypothermia supporting STAT3 was assessed by analysis of IFNγ-mediated activation of STAT1 using the same experimental protocol. As shown by immunoblot analysis (Figure 6F) and respective densitometric quantification of three independently performed experiments (Figure 6G), STAT1 activation was, in stark contrast to STAT3, not enhanced but moderately inhibited by exposure of DLD1 cells to hypothermia. Moreover, expression of prototypic STAT1-dependent IL-18 binding protein (IL-18BP) mRNA expression (60, 61) was curbed by cultivation at 30°C (Figure 6H).