A role for heat shock factor 1 in hypercapnia-induced inhibition of inflammatory cytokine expression

Abstract

Hypercapnia, elevated levels of CO₂ in the blood, is a known marker for poor clinical prognosis and is associated with increased mortality in patients hospitalized with both bacterial and viral pneumonias. Although studies have established a connection between elevated CO₂ levels and poor pneumonia outcomes, a mechanistic basis of this association has not yet been established. We previously reported that hypercapnia inhibits expression of key NF-κB–regulated, innate immune cytokines, TNF-α, and IL-6, in LPS-stimulated macrophages in vitro and in mice during Pseudomonas pneumonia. The transcription factor heat shock factor 1 (HSF1) is important in maintaining proteostasis during stress and has been shown to negatively regulate NF-κB activity. In this study, we tested the hypothesis that HSF1 activation in response to hypercapnia results in attenuated NF-κB–regulated gene expression. We found that hypercapnia induced the protein expression and nuclear accumulation of HSF1 in primary murine alveolar macrophages and in an alveolar macrophage cell line (MH-S). In MH-S cells treated with short interfering RNA targeting Hsf1, LPS-induced IL-6 and TNF-α release were elevated during exposure to hypercapnia. Pseudomonas-infected Hsf1^+/+ (wild-type) mice, maintained in a hypercapnic environment, showed lower levels of IL-6 and TNF-α in bronchoalveolar lavage fluid and IL-1β in lung tissue than did infected mice maintained in room air. In contrast, infected Hsf1^+/− mice exposed to either hypercapnia or room air had similarly elevated levels of those cytokines. These results suggest that hypercapnia-mediated inhibition of NF-κB cytokine production is dependent on HSF1 expression and/or activation.—Lu, Z., Casalino-Matsuda, S. M., Nair, A., Buchbinder, A., Budinger, G. R. S., Sporn, P. H. S., Gates, K. L. A role for heat shock factor 1 in hypercapnia-induced inhibition of inflammatory cytokine expression.

Hypercapnia, elevated levels of CO₂ in the blood, has been recognized as a marker of poor prognosis in patients with chronic and/or severe lung diseases, such as chronic obstructive pulmonary disease and acute respiratory distress syndrome (1–7). Elevated CO₂ levels are also associated with increased mortality in patients hospitalized with community-acquired bacterial and viral pneumonias (8, 9). Although clinical studies have established a connection between elevated CO₂ levels and poor patient outcomes, they have not yet revealed the mechanistic basis for that association. Work from our laboratory and others, using both cell culture systems and in vivo animal models, supports the hypothesis that exposure to hypercapnia impairs the host’s innate immune response and host defense (10–12).

In human and murine macrophages, we reported that hypercapnia attenuates the LPS-induced expression of TNF-α and IL-6– and NF-κB–regulated cytokines important for host defense in a pH-independent manner (11). We further showed that hypercapnia suppresses IL-6 and TNF-α expression in the lungs of mice infected with Pseudomonas aeruginosa, which is associated with increased mortality and impaired bacterial clearance (10). Interestingly, decreased expression of TNF-α, IL-6, and other NF-κB–regulated cytokines is also observed in macrophages exposed to “heat shock,” a modest increase in temperature of 1–2°C (13–16).

Heat shock induces expression of heat shock proteins (HSPs), such as HSP70, which function as molecular chaperones that maintain proteostasis within the cell during stress-invoking stimuli, such as bacterial infections (17, 18). Induction of HSPs is a highly conserved response regulated by the transcription factor heat shock factor 1 (HSF1), which binds to heat shock elements in promoter regions of target genes (15, 16, 19). HSF1 has been shown to inhibit gene transcription of TNF-α, IL-6, and IL-1β, at least in part, by binding to heat shock elements in the promoter regions of those non–heat shock genes, highlighting the function of HSF1 as an integrator of stress responses in cells (16, 19, 20).

In this study, we show that hypercapnia activates HSF1 in murine macrophages. Knockdown of Hsf1 in a murine alveolar macrophage–like cell line (MH-S) prevents the hypercapnia-mediated inhibition of LPS-induced IL-6 and TNF-α expression. Consistent with these findings, we show that mice with reduced HSF1 expression (Hsf1^+/−) are protected from hypercapnia-induced inhibition of IL-6, TNF-α, and IL-1β during Pseudomonas pneumonia. Taken together, these results suggest a previously unrecognized role for HSF1 as a mediator of CO₂ signaling in LPS-stimulated murine macrophages and in the Pseudomonas-infected mouse lung.

MATERIALS AND METHODS

Cell culture

MH-S cells were obtained from American Type Culture Collection (ATCC; Manassas, VA, USA) and propagated in RPMI-1640 medium (ATCC), supplemented with 10% fetal bovine serum, 1% penicillin-streptomycin, and 0.003% 2-mercaptoethanol at 37°C in humidified 5% CO₂/95% air.

Heat shock exposure

MH-S cells were maintained at 37°C in humidified 5% CO₂/95% air. At 80% confluence, cells were transferred to 42°C for 1 h, then allowed to recover at 37°C for an additional 1 h. After recovery, cells were stimulated with Escherichia coli K12 LPS (InvivoGen, San Diego, CA, USA) at 1 ng/ml for 6 h.

In vitro exposure to elevated CO₂

Cells were exposed to hypercapnia in a humidified C-174 cell-culture chamber equipped with a ProCO₂ regulator (BioSpherix, Parish, NY, USA), housed within a standard cell-culture incubator, maintained at 37°C. The regulator was set to maintain the hypercapnic atmosphere at 15% CO₂, 21% O₂. The pH and partial pressures of CO₂ and O₂ were measured with a pHOx Plus blood gas analyzer (Nova Biomedical, Waltham, MA, USA), which was calibrated daily. Cells were placed in the corresponding conditions overnight (∼16 h) before proceeding with experimental protocols.

Lactate dehydrogenase assay

MH-S cells were plated in a 96-well plate in triplicate and incubated overnight at 37°C, 5% CO₂. After exposure to hypercapnia or heat shock, cellular death was quantitated using a commercially available lactate dehydrogenase (LDH) cytotoxicity assay (Thermo Fisher Scientific, Waltham, MA, USA). The percentage of cytoxicity was calculated as 100 × (treated LDH activity − spontaneous LDH activity)/(maximum LDH activity − Spontaneous LDH activity).

Immunofluorescence

MH-S cells were fixed with 4% paraformaldehyde and blocked using bovine serum albumin with Triton X-100. Cells were incubated with rabbit anti-HSF1 antibody (Cell Signaling Technology, Danvers, MA, USA) at 1:400 dilution and mouse anti-HSP70 antibody (Abcam, Milton, Cambridge, United Kingdom) at 1:50 dilution overnight at 4°C. Cells were washed with PBS and then incubated with Alexa Fluor-555–conjugated donkey anti-rabbit and Alexa Fluor-488–conjugated goat anti-mouse secondary antibodies (Thermo Fisher Scientific) for 1 h. Nuclei were stained with DAPI at 1:1000 dilution. Images were captured with a fluorescence microscope (Nikon, Tokyo, Japan), and fluorescence intensity was quantified using ImageJ [National Institutes of Health (NIH), Bethesda, MD, USA], with results presented as corrected total cell fluorescence, the integrated density after subtraction of background fluorescence.

Cellular fractionation

MH-S cells were plated on 60-mm plates and incubated at 37°C, 5% CO₂. Once cells reached ∼80% confluence, the cells were subjected to the experimental conditions described above. A commercially available kit, NE-PER Nuclear and Cytoplasmic Extraction Kit (Thermo Fisher Scientific) was used to isolate cytoplasmic and nuclear cellular fractions for immunoblot assays.

Short interfering RNA transfection

MH-S cells were seeded in 6-well plates at a density of 1 × 10⁶ cells/well and transfected with On-TargetPlus SmartPool HSF1 short interfering RNA (siRNA; Thermo Fisher Scientific) or nontargeting negative control siRNA (Thermo Fisher Scientific) using Lipofectamine RNAiMAX transfection reagent (Thermo Fisher Scientific), as described in the manufacturer’s instructions. Briefly, before transfection, HSF1 or nontargeting siRNA was added to the transfection reagent, diluted in OptiMEM medium (Thermo Fisher Scientific), and incubated for 20 min to allow formation of transfection complexes. The siRNA complexes were added to the cells and incubated for 48 h, followed by the addition of fresh medium. The efficiency of the siRNA knockdown was determined by quantitative PCR and immunoblot to be ∼50%. Transfected cells were then exposed to normocapnia or hypercapnia overnight before LPS stimulation.

Protein preparation and Western blotting

MH-S cell extracts and lung tissue homogenates were lysed using NP-40 lysis buffer (Thermo Fisher Scientific) containing a protease inhibitor cocktail (Thermo Fisher Scientific). Protein extracts were centrifuged at 14,000 rpm for 10 min at 4°C. Total protein (20–30 µg/lane) was separated on 4–20% SDS-PAGE gels (Bio-Rad, Hercules, CA, USA) and transferred to polyvinylidine fluoride (Bio-Rad) or nitrocellulose membranes (Bio-Rad). Membranes were blocked with 5% dry, nonfat milk in Tris-buffered saline with Tween. Membranes were incubated with rabbit anti-HSF1 antibody (Cell Signaling Technology) at 1:1000 dilution, rabbit anti-HSP70 antibody (R&D Systems; Bio-Techne, Minneapolis, MN, USA) at 1:2000 dilution, mouse anti-IκBα antibody (Cell Signaling Technology) at 1:1000 dilution, or rabbit anti-p65 antibody (Cell Signaling Technology) at 1:1000, followed by horseradish peroxidase–conjugated secondary antibodies at 1:10,000 dilution (Bio-Rad). Blots were developed by chemiluminescence using SuperSignal West Dura substrate (Thermo Fisher Scientific), and quantification of protein amounts was performed using the Odyssey Fc imaging system (Li-Cor Biosciences, Lincoln, NE, USA). Densitometric analysis was performed with Image Studio Software (v.2.1.10; Li-Cor Biosciences).

Mice

Heterozygous, male Hsf1^+/− (Hsf1^tm1Ijb) mice were obtained from The Jackson Laboratory (Bar Harbor, ME, USA) and bred with wild-type C57BL/6 mice. Male, 6–10-wk-old Hsf1^+/− mice and wild-type littermate controls were used for all in vivo studies. The studies were performed according to a protocol approved by the Institutional Animal Care and Use Committee of Northwestern University (Evanston, IL, USA), according to NIH guidelines for the use of rodents.

Determination of cytokine concentrations by ELISA

Levels of IL-6 and TNF-α were determined in cell-free culture supernatants and in bronchoalveolar lavage (BAL) fluid from mice by R&D Systems ELISAs (Bio-Techne). IL-1β level in lung tissue homogenates was determined with a mouse eBioscience ELISA kit from Thermo Fisher Scientific.

Murine CO₂ exposure

Mice were exposed to normoxic hypercapnia (10% CO₂, 21% O₂) in a BioSpherix A environmental chamber. O₂ and CO₂ concentrations in the chamber were maintained at the indicated levels using ProOx C21 O₂ and CO₂ controllers (BioSpherix). Age-matched mice simultaneously maintained in air served as controls in all experiments.

Luciferase transfection in vivo

With the assistance of W. E. Balch (Scripps Research Institute, La Jolla, CA, USA), we created an adenoviral vector encoding destabilized luciferase (ad-Luc1) to allow transfection of lung tissue with luciferase adenoviral vector, which serves as a sensor of proteomic stress (21, 22). Mice were sedated with isoflurane and intubated using a 20-gauge angiocatheter. After confirmation of correct angiocatheter placement, ad-Luc1 virus (1 × 10¹⁰ plaque-forming units stock concentration), diluted in 50% dialysis buffer vehicle and 50% surfactant (Abbott Laboratories, Abbott Park, IL, USA) to achieve a dose of 1 × 10⁸ plaque-forming units in 50 µl/mouse, was intratracheally administered. To ensure survival of the animals, some were ventilated for ∼10 min after vector instillation (respiratory rate, 90 breaths/min; tidal volume, 150 μl; positive end–expiratory pressure, 5 cmH₂O; oxygen level, 100%).

Determination of luciferase activity

At the conclusion of the experiments, mice were euthanized with Euthasol (Anada, Isehara, Japan). Lungs were perfused with PBS and resected en bloc, then homogenized in 1 ml of PBS. Lung homogenate (50 μl) was used in a commercially available luciferase assay (Promega, Madison, WI, USA). Luciferase activity was quantified with a Veritas microplate luminometer (Promega).

Mouse pneumonia model

The nasal aspiration mouse model of acute pneumonia was used as described in Comolli et al. (23). P. aeruginosa strain PA103 was grown to an optical density at 600 nm of 0.002 (1.0–1.8 × 10⁶ colony-forming units/ml). For infection studies, mice preexposed to 10% CO₂ were removed from the hypercapnia chamber for bacterial inoculation, which was performed in air. Mice were lightly anesthetized with a ketamine-xylazine mixture (60 μl, 1:1 ketamine to xylazine, diluted in sterile PBS), then inoculated with 50 μl of bacterial suspension (6–10 × 10⁴ colony-forming units) via the nares.

After inoculation, mice were allowed to fully recover from anesthesia, then returned to the 10% CO₂-exposure chamber or maintained in air, according to the protocol for each experiment. Mice were euthanized 7 h after bacterial inoculation for determination of IL-6 and TNF-α secretion in BAL fluid and IL-1β levels in the lung tissue. For immunofluorescence (IF) evaluation of HSF1 and HSP70 expression in alveolar macrophages, mice exposed to air or hypercapnia were euthanized 24 h after exposure without infection. Cytospin was performed with BAL fluid, followed by IF staining of cells, as previously described.

Statistical methods

Data are presented as means ± sem. Statistical analysis was performed with Prism 6 (GraphPad Software, La Jolla, CA, USA) for Macintosh OSX (Apple, Cupertino, CA, USA). The Mann-Whitney U test, unpaired Student t test, or 2-way ANOVA with Bonferroni posttest were used, as appropriate. A value of P < 0.05 was considered significant.

RESULTS

Hypercapnia activates HSF1

To determine whether hypercapnia activates the transcription factor HSF1, MH-S cells were exposed to 5% for normocapnia (NC) or 15% CO₂ for hypercapnia (HC) for 16 h. We found that hypercapnia increased HSF1 protein expression and nuclear accumulation (Fig. 1A) as well as HSP70 protein expression (Fig. 1B) compared with macrophages exposed to NC. Similarly, heat shock exposure, a well-established activator of HSF1, increased protein expression and nuclear accumulation of HSF1 and HSP70 protein expression in MH-S cells (Fig. 1C, D).

Figure 1.

Hypercapnia and heat shock increase HSF1 protein expression and nuclear accumulation associated with increased HSP70 expression. MH-S cells were cultured in 5% CO₂ [normocapnia (NC)] or 15% CO₂ [hypercapnia (HC)] for 16 h. Cells were then fixed and stained for HSF1 and imaged with an IF microscope. A) HSF1 protein expression was determined via quantification of fluorescence. B) Alternatively, cells were lysed after 16 h exposure to 5 or 15% CO₂ for determination of HSP70 expression in whole cell lysate. For heat shock experiments, MH-S cells were exposed to 42°C for 1 h, then allowed to recover for 1 h, or were maintained continuously at 37°C as controls. C) Cells were fixed, stained, and imaged with an IF microscope to determine HSF1 protein expression. D) HSP70 protein expression was determined via immunoblot of whole cell lysate. IF images are representative images of n ≥ 3 independent experiments. For immunoblots, bar graphs represent means ± sem, n ≥ 3 independent experiments.

HSF1 activation via hypercapnia and heat shock inhibits IL-6 and TNF secretion

To evaluate the effect of HSF1 activation on cytokine secretion, MH-S cells were exposed to HC or heat shock, followed by stimulation with LPS. In MH-S cells exposed to HC, the IL-6 and TNF-α secretions were decreased in a similar manner to that of macrophages exposed to heat shock (Fig. 2A, B). Of note, hypercapnia-mediated inhibition of both IL-6 and TNF-α release was not associated with the degradation of Iκ-Bα (Fig. 2C) or the nuclear translocation of p65/RelA (Fig. 2D) in LPS-stimulated macrophages, indicating that elevated CO₂ does not impair NF-κB activation in these key regulatory steps. Moreover, the differences in cytokine secretion in these cells were not due to cellular death as shown in Table 1.

Figure 2.

Hypercapnia and heat shock exposure reduces IL-6 and TNF secretion in macrophages independent of NF-κB regulation. A) MH-S cells were cultured in 5 or 15% CO₂ for 16 h, then stimulated with LPS (1 ng/ml) in normocapnia (NC) and hypercapnia (HC), respectively; after which, supernatant levels of IL-6 and TNF were determined by ELISA. B) Alternatively, cells were exposed to 42°C for 1 h, followed by recovery at 37°C for 1 h, or were maintained at 37°C as controls, then stimulated with LPS (1 ng/ml) for 6 h; after which, IL-6 and TNF release was quantified by ELISA. To determine NF-κB activation, MH-S cells exposed to 5 or 15% CO₂ for 16 h, then stimulated with LPS (1 ng/ml) in NC or HC cells for 1 h. After LPS stimulation, cellular fractionation was performed to isolate cytoplasmic and nuclear compartments. C, D) IκBα (C) and RelA/p65 (D) were determined in the cytoplasmic and nuclear fractions, respectively, using immunoblot. Bar graphs represent means ± sem, n ≥ 3 independent experiments.

TABLE 1.

Cell death in MH-S macrophages exposed to NC or HC and heat shock

Exposure	% Cytotoxicity	P
NC, 37°C	2.75 ± 1.07
HC	3.45 ± 0.67	n.s.a
Heat shock (42°C)	4.163 ± 0.91	n.s.a

^aCompared to NC, 37°C control;

n.s., not significant.

Hypercapnia-induced inhibition of IL-6 requires HSF1 expression

To determine whether HSF1 is required for the inhibition of cytokine expression by elevated CO₂, we knocked down HSF1 mRNA expression in MH-S cells using siRNA. As shown in Fig. 3A, HSF1 siRNA reduced HSF1 protein expression in MH-S cells by ∼60% as compared with cells transfected with nontargeting siRNA. In macrophages treated with nontargeting siRNA, hypercapnia exposure inhibited LPS-induced IL-6 secretion by nearly 50% as compared with normocapnia exposure (Fig. 3B). In contrast, hypercapnia-induced inhibition of IL-6 expression was abrogated in macrophages treated with HSF1 siRNA (Fig. 3B).

Figure 3.

HSF1 is required for hypercapnia-mediated IL-6 inhibition. A) MH-S cells were treated with HSF1 siRNA or control siRNA for 48 h (at 37°, 5% CO₂). Cells were then exposed to NC or HC for an additional 16 h; after which, HSF1 protein expression was determined by immunoblot; bar graphs represent means ± sem, n ≥ 3 animals. B) Alternatively, cells were treated with HSF1 or control siRNA for 72 h, exposed to normocapnia (NC) or hypercapnia (HC) for 16 h, then stimulated with LPS (1 ng/ml) for 6 h in NC or HC, respectively. IL-6 levels were determined in cellular supernatant by ELISA. Bar graphs represent means ± sem in duplicate, n ≥ 3 animals.

Hypercapnia increases HSF1 expression and activates the proteostatic network in the murine lung

Given our in vitro findings, we moved to a murine mouse model of hypercapnia to further evaluate the effect of elevated carbon dioxide on HSF1. Mice exposed to normoxic hypercapnia (10% CO₂, 21% O₂) had increased levels of total HSF1 protein in whole lung homogenate compared with mice exposed to normocapnia (Fig. 4A). Next, because HSF1 is a key transcription factor for maintaining proteostasis during organismal stress, we used a luciferase tool to assess the changes in proteostatic activity in the lung of hypercapnia-exposed mice. The American firefly (Photinus pyralis) luciferase (Fluc) is a 60-kDa protein requiring chaperones for folding and maximal enzymatic activity and has been used in multiple cell lines and in Caenorhabditis elegans as a sensor for proteostatic activity during stress (21, 24). During cellular stress, such as in heat shock and hypercapnia exposure, key chaperone proteins are up-regulated in efforts to maintain proteostasis. The increased chaperone proteins bind to Fluc, resulting in improved folding and increased luciferase activity. To demonstrate this in the murine lung, we used an adenoviral vector carrying Fluc (Ad-Luc1) to transfect cells in the murine lung, as described elsewhere (22, 25). At 5 d after ad-Luc1 transfection, the mice were exposed to air or normoxic hypercapnia for up to 7 d; after which, lungs were harvested, and luciferase activity was quantified. In hypercapnia-exposed mice, there was a 2.5-fold increase in luciferase activity, indicating increased chaperone activity induced by hypercapnia (Fig. 4B). This correlated with increased HSF1 protein expression and nuclear accumulation as well as increased HSP70 protein expression in alveolar macrophages from hypercapnia-exposed mice, as compared with air-exposed controls (Fig. 4C).

Figure 4.

HSF1 and HSP70 expression is increased in hypercapnia-exposed mice. Wild-type C57BL/6 mice were exposed to normoxic hypercapnia (21% O₂, 10% CO₂) or air, as a control, for 3 d and then euthanized. A) HSF1 expression was determined in homogenized lung tissue by immunoblot. B) In a separate set of experiments, C57BL/6 mice were infected intratracheally with an ad-Luc1 adenovirus vector; 5 d after transfection, animals were exposed to air or normoxic hypercapnia (21% O₂, 10% CO₂). Mice were euthanized after exposure for 7 d to hypercapnia (or air as the control), and luciferase activity was measured in whole lung homogenates. C) C57BL/6 mice were exposed to normoxic hypercapnia (21% O₂, 10% CO₂), or air as control, for 3 d, Then, mice were euthanized, BAL was performed, and BAL cells were fixed, immunostained, and imaged with fluorescence microscopy to determine HSF1 and HSP70 expression. Scale bar, 10 µm. Bar graphs represent means ± sem, n ≥ 3 animals in all experiments.

HSF1 is necessary for the hypercapnic inhibition of proinflammatory cytokine release during P. aeruginosa infection

To investigate the clinical relevance of our findings, we turned to an in vivo model of P. aeruginosa pneumonia in mice. We exposed wild-type mice and Hsf1 heterozygous mice to air or normoxic hypercapnia (10% CO₂, 21% O₂) for 3 d, which results in a P_aCO₂ of ∼75 mmHg, with maximal renal compensation of respiratory acidosis (arterial pH, ∼7.28). Additionally, we previously showed that in the absence of infection, exposure to 10% CO₂ caused no overt distress to the animals or evidence of lung injury (10).

In mice heterozygous for Hsf1 (Hsf1^+/−), baseline expression of HSF1 protein and HSP70 in lung homogenates was ∼65 and 50% less, respectively, than that of their wild-type littermate controls (Fig. 5A). We exposed Hsf1^+/− and wild-type littermate control mice to room air or normoxic hypercapnia (21% O₂, 10% CO₂) for 3 d. BAL was performed after CO₂ exposure in the absence of infection. Using IF microscopy, we observed increased HSF1 expression, nuclear accumulation, and HSP70 expression in alveolar macrophages from hypercapnia-exposed, wild-type mice (Fig. 5B); that increase was attenuated in Hsf1^+/− mice exposed to hypercapnia. Next, we exposed Hsf1^+/− and wild-type littermate control mice to air or normoxic hypercapnia (21% O₂, 10% CO₂) for 3 d before inoculation with P. aeruginosa, followed by continued exposure to hypercapnia or air, respectively. Mice were euthanized 7 h after infection for BAL, and lung tissue was harvested. Similar to our previous findings, Pseudomonas infection elicited significant macrophage and neutrophil influx in both air- and 10% CO₂–exposed animals, which did not differ between Hsf1^+/− mice and wild-type controls (Table 2). In wild-type mice exposed to air, Pseudomonas infection markedly increased lung IL-6 and TNF-α levels in the BAL fluid and the IL-1β secretion in the lung tissue, whereas the increase in cytokines was attenuated in wild-type animals exposed to 10% CO₂ (Fig. 5C). Conversely, Hsf1^+/− mice were protected from hypercapnic inhibition of IL-6, TNF-α, and IL-1β in the lungs after Pseudomonas infection (Fig. 5C).

Figure 5.

HSF1 is necessary for hypercapnia-induced inhibition of IL-6 and TNF in vivo. A) HSF1 and HSP70 protein expression in lung homogenates of wild-type C57BL/6 and HSF1^+/− mice was determined by immunoblot; bar graphs represent means ± sem, n ≥ 3 mice. B) Wild-type and HSF1^+/− C57BL/6 mice were exposed to normoxic hypercapnia (10% CO₂, 21% O₂), or to air as a control, for 3 d, and were then euthanized. BAL was performed, and BAL cells fixed, immunostained, and imaged using fluorescence microscopy to quantify HSF1 and HSP70 expression. Scale bar, 10 µm. C) IL-6 and TNF protein concentration was determined in cell-free BAL fluid and IL-1β levels in lung homogenate by ELISA. Bar graphs represent means ± sem, performed in duplicates, n ≥ 3 mice.

TABLE 2.

Differential BAL fluid cell count from Pseudomonas-infected mice exposed to air or hypercapnia

Cells	Air		P	10% CO2		P
	WT	HSF1+/−		WT	HSF1+/−
Neutrophils (%)	93.0 ± 1.5	90.3 ± 3.7	n.s.	83.0 ± 3.5	92.7 ± 2.2	n.s.a,b
Macrophages (%)	7.3 ± 1.2	11.3 ± 3.5	n.s.	18.7 ± 3.5	6.7 ± 2.0	n.s.a,b

^aCompared to air controls;

^bcompared to wild-type controls;

n.s., not significant.

DISCUSSION

We previously observed that hypercapnia impairs macrophage function and worsens outcomes of Pseudomonas pneumonia associated with impaired IL-6 and TNF-α secretion and impaired macrophage and neutrophil function (10, 11, 26). However, the mechanism by which hypercapnia impairs innate immune function has yet to be established. HSF1 is a conserved transcription factor that has been shown to have a role in immunity. In C. elegans and animal models, HSF1 is independently involved in important effector functions for optimal immune response to bacterial pathogens (27–30). In this study, we examined the effect of hypercapnia on activation of the proteotoxic stress response, focusing on activation of HSF1.

The central finding of our study is that hypercapnia increases expression and activation of HSF1, resulting in HSF1-dependent inhibition of IL-6 and TNF-α. The role of HSF1 in transcriptional regulation of inducible heat shock genes, such as HSP70, has been well studied in the heat shock response, the prototypical model of HSF1 activation (13, 31–33). We used an adenoviral vector containing luciferase as a sensor of proteostatic stress to show that hypercapnia serves as a less-potent activator of HSF1 in murine lungs. Although it has been well established that HSF1 serves as a sensor for cellular and organismal stress, the exact mechanism or mechanisms by which HSF1 senses stressors and regulates transcription is not clearly understood. The inability of HSF1-dependent inhibition of NF-κB–regulated cytokines in hypercapnic environments to alter IκBα expression or nuclear accumulation of p65/RelA, 2 important regulatory steps in NF-κB signaling, is consistent with previous studies. These results suggest that HSF1 may have a unique role in mediating the effects of hypercapnia at the cellular and organismal level (11, 12).

There are many cellular stressors that activate HSF1, including a decrease in extracellular pH (19, 34, 35). In our in vitro model, the pH of the medium after hypercapnia exposure was ∼7.2. However, we previously demonstrated that buffered medium does not abrogate hypercapnia-induced inhibition of IL-6 and TNF-α expression, suggesting this is a pH-independent effect of hypercapnia (11). We also showed that exposure of animals to hypercapnia for 3 d resulted in an arterial blood pH of 7.28, with maximal renal compensation, which is also associated with IL-6 and TNF-α inhibition after Pseudomonas infection (10, 11). Consistent with these findings, Helenius et al. (36) showed that in Drosophila, hypercapnia exposure leads to increased susceptibility to bacterial infection independent of pH. In addition, Nichol et al. (37) demonstrated that pH buffering did not improve susceptibility to infection in rats. These findings taken together suggest that CO₂ may serve as an activator of HSF1 independent of acidosis.

Multiple mechanisms that regulate NF-κB activity by HSF1 have been described; these include induction of heat shock proteins, such as HSP70, which modulate the stability and function of regulatory proteins, such as IκBα (15, 32, 38). In our in vitro studies, hypercapnia caused increased expression and nuclear accumulation of HSF1 with associated IL-6 and TNF-α inhibition, similar to heat shock exposure. This inhibition was not associated with changes in the protein expression of IκBα or nuclear accumulation of p65/RelA, indicating that the inhibitory effect of HSF1 is downstream of p65/RelA activation in this model. Our results are similar to the findings by Cummins et al. (12), in which hypercapnia did not alter IκBα expression in LPS-stimulated fibroblasts. In our animal model, HSP70 protein expression was increased by hypercapnia with a correlating increase in luciferase activity, suggesting that HSF1-induced chaperones may be altered by hypercapnia.

Genetic studies in yeast and Drosophila have shown that HSF1 orchestrates cellular programming that results in alterations in many genes other than heat shock genes (39, 40). In addition, Mendillo et al. (41) showed that HSF1 drives transcriptional regulation of genes distinct from heat shock genes in human cancer cells. The mechanisms by which HSF1 impairs gene expression include 1) direct repression by binding at or near promoter regions of target genes, and 2) facilitating the binding of other transcription factors that serve as negative regulators of gene activity (14, 38, 42, 43). In a genome-wide RNA interference screen, Helenius et al. (44) observed that zinc finger homeodomain 2 (Zfh2), a zinc-finger transcription factor in Drosophila, is a mediator of hypercapnic immune regulation associated with mitigated reductions in antimicrobial peptides and resistance to infection in flies exposed to CO₂. These observations, along with our observation that hypercapnia-induced inhibition of NF-κB–dependent cytokine expression occurs independent of p65/RelA nuclear accumulation, suggest that HSF1-mediated inhibition of IL-6 and TNF-α may be mediated by transcriptional regulation of these genes by HSF1 (15, 16, 20, 42).

The mechanism by which HSF1 senses stress and regulates transcriptional activity remains unclear. There is evidence that, in the setting of heat shock, HSF1 independently serves as a thermosensor; however, in our model, sensing of temperature changes does not explain the inhibition of IL-6 and TNF-α that we have observed (45). Carbon monoxide (CO), a molecule that similarly diffuses across cellular membranes, regulates HSF1-dependent up-regulation of HSP70 expression in endothelial cells and fibroblasts, supporting a potential role for HSF1 in CO₂-mediated inhibition of innate immune responses (46). Our findings are consistent with genetic studies in yeast and C. elegans, which showed that the stress response, regulated by HSF1, is an integrator of various metabolic processes in organisms (39, 40).

In summary, we have shown that hypercapnic inhibition of IL-6 and TNF-α is dependent on HSF1. We have demonstrated that hypercapnia may exert its effects through many pathways that regulate NF-κB signaling. There is mounting evidence suggesting that hypercapnia exerts its effects independent of changes in pH, which suggests the presence of CO₂ sensors in mammalian cells that have not been fully elucidated (10–12, 36). The current findings implicate HSF1 as a potential CO₂ sensor and regulator of the proteostatic stress response, including regulation of heat shock and non–heat shock genes important in the immune response to infection. Further studies are required to determine the mechanism by which HSF1 senses CO₂ and mediates its effects.

Glossary

ad-Luc1

adenoviral vector encoding destabilized luciferase

BAL

bronchoalveolar lavage

HC

hypercapnia

HSF1

heat shock transcription factor 1

HSP

heat shock protein

IF

immunofluorescence

LDH

lactate dehydrogenase

NC

normocapnia

P_aCO₂

partial pressure of carbon dioxide in arterial blood

siRNA

short interfering RNA

AUTHOR CONTRIBUTIONS

S. M. Casalino-Matsuda, P. H. S. Sporn, and K. L. Gates designed research; Z. Lu, A. Nair, A. Buchbinder, and K. L. Gates performed the research; S. M. Casalino-Matsuda and G. R. S. Budinger contributed analytic tools; Z. Lu, G. R. S. Budinger, P. H. S. Sporn, and K. L. Gates analyzed data; and P. H. S. Sporn and K. L. Gates wrote the paper.