For in vivo studies, lung specific IL6 over-expressor (IL6+) transgenic mice breeders were obtained from the University of Pittsburgh (gift from Dr. Stephen Chan). In the IL6+ mice, the Clara cell 10-kD promoter constitutively drives lung-specific IL6 over-expression (3). Age-matched (3 – 4-month-old) IL6+ and littermate IL6- wild-type (WT) control male and female mice were used for the study in accordance with protocols approved by the Tufts University Institutional Animal Care and Use Committee. Mice were housed in the animal facility at Tufts University School of Medicine and provided food and water ad libitum. Mice were anesthetized with Ketamine (100 mg/kg) and xylazine (10 mg/kg) and right ventricular systolic pressures were recorded as previously described (13). Mice were then euthanized by exsanguination and lung and heart tissues were isolated. Wet and dry tissue weights were recorded. For dry weights, the tissue was placed in a 60°C incubator for 24 hours and then weighed.

The IL6+ mice were identified by genotyping mouse DNA by PCR as described previously (3). Briefly, genotyping of mice was performed by PCR using the following primer pair: forward primer 5’-GCAACAAAAAGTGGGTAAATG-3’ and reverse 5’-CTCCAAAAGACCAGTGATGAT-3’. PCR was carried out using the REDExtract-N-Amp PCR kit (MilliporeSigma, Burlington, MA) and the PCR products were resolved on 1.2% agarose gels using the FlashGel System (Cat# 57067; Lonza).

For in vitro studies, recombinant human IL6 was purchased from R&D Systems (Minneapolis, MN). TG2 small molecule inhibitor, ERW1041E (Cat# 5095220001) and pyruvate kinase isoform M2 inhibitor, Shikonin (Cat# S7576) were purchased from MilliporeSigma and both were dissolved in dimethyl sulfoxide (DMSO; MilliporeSigma) as described previously (10).

Hydroxyproline in right lung lobes and cardiac right ventricles was measured using high-performance liquid chromatography as previously described (14). For quantitative assessment of fibrogenic remodeling, hydroxyproline measurements are reported as amounts (μg) per wet weight of tissue.

Immunohistochemistry was carried out on formalin-fixed and paraffin-embedded lung, left ventricle (LV) and right ventricle (RV) sections following standard methodology. For qualitative assessment of fibrogenic remodeling, lung, and heart (LV and RV) sections were stained with Picrosirius stain (Cat# 26357-02; Electron Microscopy Sciences, Hatfield, PA). All imaging was performed using light microscopy (Nikon Eclipse E800 microscope; Nikon Instruments, Melville, NY) and Spot Imaging software (Sterling Heights, MI).

Frozen lung and heart tissues were used for total RNA extraction by TRIzol (Invitrogen, Thermo Fisher Scientific, Waltham, MA) according to the manufacturer’s instructions. Briefly, 1μg total RNA was reverse transcribed using a High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific). Real-time polymerase chain reaction (PCR) analysis was performed using 2x SYBR Green Master Mix (Thermo Fisher Scientific) on an ABI Prism 7900 Sequence Detection System (Thermo Fisher Scientific) as described previously (10). Human and mouse specific primer sets (IDT Technologies, Coralville, IA) were used (10). 18s ribosomal RNA levels were used to normalize the cycle threshold (Ct) values. A ΔΔCt method was used to express the relative quantification of specific genes as previously described (10).

Frozen tissues were homogenized, and cell culture lysates were lysed in NP-40 lysis buffer (Boston Bio Products, Ashland, MA) in presence of protease and phosphatase inhibitor cocktails (MilliporeSigma). The lysed samples were then separated into two fractions by centrifugation, the soluble supernatant and the insoluble pellet fraction comprising the extracellular matrix (ECM) proteins. Bradford Assay (Bio-Rad Laboratories) was used to quantify the total protein content. An equal amount of protein lysates or the ECM fractions were then reduced in SDS sample buffer (Boston Bio Products) and processed for SDS-PAGE analysis as described (30). For Western blot analysis, antibodies against target protein transglutaminase 2 (TG2; Cat# sc-48387; Santa Cruz Biotechnology, Dallas, TX); fibronectin (Fn; Cat# sc-8422, Santa Cruz Biotechnology); serotonin (Cat# S5545, MilliporeSigma); Stat3-phospho (Cat# 9145; Cell Signaling Technology, Danvers, MA); Stat3 (Cat# 4904; Cell Signaling Technology); pyruvate kinase M2 (PKM2; Cat# 3198; Cell Signaling Technology); type 1 collagen (Col1; Cat# ab34710, Abcam, Boston, MA); α-smooth muscle actin (α-SMA; Cat# sc-32251; Santa Cruz Biotechnology) and β-actin (Cat# 4970, Cell Signaling Technology) were used. Horseradish peroxidase (HRP) tagged secondary antibodies (Santa Cruz Biotechnology) were used for primary antibody target detection. ECL substrate (Thermo Fisher Scientific) was used to visualize the protein bands. Quantitative image analysis with densitometry was performed using Image J software (NIH).

For in vitro studies, primary mouse lung fibroblasts isolated from WT and IL6+ mice by explant culture; de-identified human PA adventitial fibroblasts isolated from healthy donors at the University of Colorado (gift from Dr. Kurt Stenmark) and control human PA adventitial fibroblasts purchased from a commercial supplier (Catalog #3120; Sciencell Research Laboratories) were used. Fibroblasts were grown in fibroblast growth media supplemented with 5% fetal bovine serum (FBS), fibroblast growth factors and antibiotics (Cat. #2331; ScienCell Research Laboratories, Carlsbad, CA) and used at passage 2-4. Cells were maintained at 37°C in a humidified 5% CO₂ incubator. Cells were serum starved 1 day prior to treatment and maintained in reduced-serum media (0.2% FBS) for the duration of the cell treatment studies.

Mouse lung fibroblasts were grown on cover slips in fibroblast growth media as described above. Cells were then serum-starved overnight and incubated with sterile phosphate buffered saline (PBS; vehicle) or recombinant human IL6 protein (Cat# 206-IL; R&D Systems). After 24 hours, followed by a brief PBS wash, cells were fixed with 4% formaldehyde (Tousimis, Rockville, MD) and permeabilized with 0.1% Triton X-100 (Thermo Fisher Scientific). Fixed cells were then blocked with 5% bovine serum albumin (BSA) in PBS and incubated overnight at 4⁰C with 1% BSA (diluent; secondary antibody control) or alpha smooth muscle actin antibody (α-SMA; Cat# sc-32251; Santa Cruz Biotechnology) followed by AlexaFluor 555 conjugate secondary antibody (Thermo Fisher Scientific) for 1 hour in blocking buffer (1% BSA). Cover slips were then mounted on to slides using Vectashield Antifade mounting medium with DAPI (Cat# H-1200; Vector Laboratories, Burlingame, CA) and sealed. Nikon Eclipse E800 fluorescence microscope (Nikon Instruments, Melville, NY) equipped with Spot Imaging software (Sterling Heights, MI) was used to visualize and image the stained cells.

Human pulmonary artery adventitial fibroblast cells were seeded in fibroblast growth media as described above on a 96-well cell culture plates (MilliporeSigma). Fibroblast growth media was replaced with reduced-serum fibroblast basal media (0.2% FBS). Next day, a TG2 substrate, 50µM biotin cadaverine (N-(5-Aminopentyl) biotinamide, trifluoroacetic acid Salt) (Cat# A1594; Thermo Fisher Scientific) was then added to the wells. After 24 hours, cell plates were washed with PBS and blocked with 1% BSA in PBS for 1 hour. Cells were then incubated with Streptavidin-HRP conjugate (Cat# GERPN1231; Millipore Sigma) in blocking buffer for 1 hour followed by washing with PBS. For extracellular TG2 activity, cultured cell plates were de-cellularized with 0.25M NH₄OH (MilliporeSigma) in 50mM Tris pH 7.4 for 10 minutes at room temperature as described previously (15). Cells were then incubated with a chromogenic substrate and visualizing reagent, TMB solution (Cat# ab171522; Abcam). After 5 minutes, Stop Solution for TMB (Cat# ab171529; Abcam) was added to each well and the 450nm absorbance was read in a Tecan SPECTRAFluor Plus microplate reader (Artisan Technology Group, Champaign, IL) equipped with Magellan Data Analysis software (Tecan).

Data are expressed as mean ± standard error of the mean (SEM) or as median with inter-quartile range. Statistical analysis was performed by unpaired t-test or analysis of variance (ANOVA) using a post hoc Tukey’s multiple comparison test on GraphPad Prism. P value of < 0.05 was considered statistically significant.

Over-expression of IL6 in mouse lungs significantly elevated right ventricular systolic pressure (RVSP; Figure 1A ; 31.2 ± 0.79 mmHg), over wild-type mice (20.0 ± 0.62 mmHg), as previously reported (3). Fulton index (RV/LV+S ratio) and ratio of RV/body weights were both significantly greater than in the respective wild-type controls ( Figures 1B, C ). Additionally, the ratio of LV+S/body weights did not differ between IL6+ and WT mice ( Figure 1D ). These results suggest that lung-specific IL6 over-expression enhances both the RVSP and RV tissue remodeling without altering LV+S mass.

Although we found no significant differences in body weights between the IL6+ (25.78 ± 0.97 gms) and WT mice (24.3 ± 0.94gms; Figure 2A ), we observed that both wet ( Figure 2B ) and dry ( Figure 2C ) lung weights (left and right lung lobes) were significantly increased in IL6+ mice compared to WT littermates by 3 months of age. Furthermore, the ratios of wet-to-dry weights of both left and right lungs were significantly increased ( Figure 2D ). On the other hand, the trends toward increases in cardiac RV and LV wet and dry weight ratios in IL6+ mice were not significantly different ( Figure 2E ) compared to the wild-type littermates.

Lung-specific overexpression of IL6 led to a significant elevation in hydroxyproline levels ( Figure 3A ) in right lung lobes of IL6+ mice (344.45 ± 38.58 µg) compared to those of WT littermate controls (139.45 ± 7.17 µg). Although not significant (p=0.08; Figure 3B ), we also observed that hydroxyproline levels were moderately elevated in RVs of IL6+ mice (555.53 ± 27.7 µg) compared to WT controls (484.77 ± 22.8 µg).

Interestingly, as seen in Figure 3C , the IL6+ lung images show remarkable enlargement of air spaces compared to WT lungs. In addition, we show tissue fibrosis as evidenced by increased collagen accumulation in the left lungs and RVs of lung-specific IL6+ transgenic mice. SiriusRed staining of collagen fibers showed more fibrosis in lungs ( Figure 3C ) and RVs ( Figure 3D ) of IL6+ mice than wild-type littermates. As expected, we did not find a marked increase in collagen staining in the LVs of IL6+ mice ( Figure 3E ).

We next investigated if glycolysis participates in IL6 mediated tissue remodeling. Quantitative PCR analysis demonstrated that the rate-limiting glycolytic enzymes, 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 3 (PFKFB3) and pyruvate kinase isoform M2 (PKM2) and the master transcriptional regulator for glycolytic markers, hypoxia-inducible factor 1-alpha (HIF1α) are markedly up-regulated in the lungs of IL6+ transgenic mice as compared to age-matched WT control mice ( Figure 4A ), suggesting that glycolysis is markedly stimulated by the presence of IL6. Similarly, fibrogenic markers including transglutaminase 2 (TG2), collagen 1A1 (Col1) and fibronectin mRNA levels are elevated in lungs of these IL6+ mice ( Figure 4B ). In addition, although it did not reach statistical significance, there was a trend toward increased mRNA levels of lysyl oxidase (LOX) in lungs of IL6+ mice compared to WT littermates ( Figure 4B ).

To establish the relevance of lung specific IL6 over-expression on RV remodeling, we next evaluated the expression of fibrogenic markers in the RVs of IL6 transgenic mice. As anticipated, fibrogenic markers including TG2 and Fn mRNA expression are significantly increased in RVs of IL6+ mice compared to WT mice ( Figure 4C ).

Next, we investigated the effect of IL6 on STAT3 activation, a canonical transcriptional mediator of IL6 downstream signaling and promoter of pulmonary fibrosis (16). Immunoblot analysis confirmed that the IL6 over-expression increased the ratio of phosphorylated-STAT3/total STAT3 in IL6+ transgenic mouse lungs compared to WT mice ( Figure 5A ). We next confirmed that IL6 over-expression induced the protein expression of glycolytic and fibrogenic markers in the lungs of IL6 transgenic mice as compared to wild-type controls. The lung specific IL6 over-expression induced PKM2 in the mouse lung tissue lysates ( Figure 5B ). In addition, we found that IL6 induced more TG2 expression and activity (sFN levels) in the extracellular matrix fraction of lung tissue lysates of the IL6 transgenic mice than wild-type control mice ( Figure 5C ).

To investigate the link between IL6 and fibrogenic remodeling, the induction of a fibroblast cell phenotype was assayed by immunocytochemistry. Lung fibroblasts isolated from both transgenic IL6+ mice and wild-type control mice were grown on coverslips for 72 hours. Wild-type cells were incubated with or without human IL6 recombinant protein. Fibroblasts were then stained with diluent solution ( Figure 6A ) or α-smooth muscle actin (α-SMA) antibody to detect a myofibroblast phenotype ( Figures 6B-D ). Increased expression of α-SMA was evident in the fibroblasts isolated from WT mice treated with recombinant IL6 protein ( Figure 6C ) and IL6+ transgenic mice ( Figure 6D ) compared to the vehicle-treated WT fibroblasts ( Figure 6B ).

To investigate a role for IL6 in human pulmonary arterial (PA) remodeling; in vitro studies with human pulmonary artery adventitial fibroblasts were carried out to assess effects of IL6 on glycolysis and TG2 activity in human PA adventitial fibroblasts. The results of these studies show that exogenous addition of human recombinant IL6 cytokine induces mRNA expression of glycolytic enzyme and fibrogenic markers in fibroblasts ( Figure 7A ). After stimulation with IL6 for 24 hours, we observed concentration-dependent marked upregulation of PKM2, TG2, TGFβ1 and Col1 mRNA expression ( Figure 7A ). Additionally, we found that IL6 stimulation led to activation of TG2 in four independent primary human PA adventitial fibroblast cell lines (n=4; Figure 7B ). We further showed that pretreatment with a small molecule TG2 inhibitor significantly blocked the IL6-induced TG2 activity in these primary human fibroblasts ( Figure 7B ).

Previous studies from our lab showed that TG2 is implicated in glycolysis-mediated fibrogenesis in lung and RV fibroblasts (10). To assess the functional role for glycolytic enzyme, PKM2 on IL6-mediated TG2 activation and fibrogenic signaling in vitro, we exposed human PA adventitial fibroblasts to IL6 in the presence of vehicle control or Shikonin, a pharmacological inhibitor of PKM2 (17), for 96 hours. IL6 significantly induced TG2 activity in both cell lysates ( Figure 8A ) and the ECM ( Figure 8B ). Furthermore, Shikonin significantly reduced IL6-induced TG2 activity in cell lysates and the ECM ( Figures 8A, B ). Immunoblotting confirmed that the IL6-induced fibrogenesis is mediated by PKM2 in these fibroblasts ( Figure 8C ). We found that pretreatment with Shikonin at varying (0.5 – 2.5 µM) concentrations attenuated the IL6-induced protein expression of fibronectin, Col1 and α-SMA in fibroblasts ( Figure 8C ).