Differential responses to folic acid in an established keloid fibroblast cell line are mediated by JAK1/2 and STAT3

Abstract

Keloids are a type of disordered scar formation which not only show heterogeneity between individuals and within the scar itself, but also share common features of hyperproliferation, abnormal extra-cellular matrix deposition and degradation, as well as altered expression of the molecular markers of wound healing. Numerous reports have established that cells from keloid scars display Warburg metabolism-a form of JAK2/STAT3-induced metabolic adaptation typical of rapidly dividing cells in which glycolysis becomes the predominant source of ATP over oxidative phosphorylation (OxPhos). Using the JAK1/2 inhibitor ruxolitinib, along with cells from patients with STAT3 loss of function (STA3 LOF; autosomal dominant hyper IgE syndrome) we examined the role of JAK/STAT signaling in the hyperproliferation and metabolic dysregulation seen in keloid fibroblasts. Although ruxolitinib inhibited hyperactivity in the scratch assay in keloid fibroblasts, it paradoxically exacerbated the hyper-glycolytic state, possibly by further limiting OxPhos via alterations in mitochondrial phosphorylated STAT3 (pSTAT3Ser727). In healthy volunteer fibroblasts, folic acid exposure recapitulated the exaggerated closure and hyper-glycolytic state of keloid fibroblasts through JAK1/2- and STAT3-dependent pathways. Although additional studies are needed before extrapolating from a representative cell line to keloids writ large, our results provide novel insights into the metabolic consequences of STAT3 dysfunction, suggest a possible role for folate metabolism in the pathogenesis of keloid scars, and offer in vitro pre-clinical data supporting considerations of clinical trials for ruxolitinib in keloid disorder.

Conflict of interest statement

The authors have declared that no competing interests exist.

Figures

Fig 1. EMT in keloid fibroblasts was dependent on JAK signaling and glycolysis.

(A) Wound closure over time for keloid (KEL-) and healthy volunteer (HV-) fibroblast cell lines (FB) in the scratch assay of wound repair. (B and C) Scratch assay results after treatment with the glycolysis inhibitor 2DG or the oxidative phosphorylation inhibitor rotenone for keloid (B) and HV (C) fibroblast cell lines. (D) Scratch wound repair in KEL-FB, HV-FB, contrasted against FB from patients with autosomal dominant hyper IgE syndrome due to STAT3 loss of function (STAT3 LOF) or KEL-FB treated with the JAK1/2 inhibitor ruxolitinib (Ruxo). Representative images (E) and quantitation at 20 hours (F) of wound closure in indicated cells with ruxolitinib treatment. Results are representative of three independent experiments and displayed as mean ± SEM for triplicate wells. * = p

Fig 2. Abnormalities in metabolic balance were influenced by STAT3.

Seahorse assay results for a healthy volunteer (HV) fibroblast line, a keloid (Kel) fibroblast line, or fibroblasts form patients with autosomal dominant hyper IgE syndrome due to STAT3 loss of function (STAT3LOF) with and without treatment with the JAK1/JAK2 inhibitor ruxolitinib (Ruxo): (A) basal extracellular acidification rate (ECAR), (B) mitochondrial ATP production, (C) basal oxygen consumption rate (OCR), (D) OCR tracing for indicated cell types during basal, oligomycin (Oligo) stimulation, carbonyl cyanide-4-(trifluoromethoxy)phenylhydrazone (FCCP) stimulated, and rotenone/antimycin A (Rot/AA) stimulation. (E) spare respiratory capacity (SRC), and (F) basal ECAR divided by basal OCR (ECAR/OCR ratio). Results are representative of three independent experiments and displayed as mean ± SEM for 5–6 wells per condition. * = p

Fig 3. Keloid fibroblasts have alterations in transcriptional regulators of metabolism.

(A and B) Immunofluorescent (IF) representative images (A) and quantitation of HIF1α staining signal co-occurring with DAPI nuclear marker in each identified cell line (B), DAPI not shown to allow nuclear localization assessment. (C) Representative IF images for cells stained for TOM70 (red), phosphorylated STAT3Ser727 (pSTAT3Ser727; green), and DAPI (blue). (D and E) Signaling intensity for TOM70 (D) as well as pSTAT3Ser727 associated with TOM70 (mitochondrial) or not associated with TOM70 (nuclear and cytoplasmic; Nuc + Cyto) (E) per cell. Results are representative of two independent experiments and displayed as mean ± SEM for triplicate wells per condition. * = p <0.05; **** = p < 0.0001 versus HV with diluent condition as determined by ANOVA with Sidak adjustment.

Fig 4. Folic acid exposure recapitulates some of the keloid phenotype in healthy fibroblasts.

Seahorse assay results for a healthy volunteer (HV) fibroblast (FB) line, a keloid (Kel) fibroblast line, or fibroblasts form patients with autosomal dominant hyper IgE syndrome due to STAT3 loss of function (STAT3LOF) with and without treatment with 50ug/mL folic acid: (A) basal extracellular acidification rate (ECAR), (B) basal oxygen consumption rate (OCR), (C) mitochondrial ATP production, (D) spare respiratory capacity (SRC), and (E) basal ECAR to OCR ratio are shown. (F-G) basal ECAR to OCR ratio (F), OCR (grey, dotted lines) and ECAR (black, solid lines) raw values (G) for HV fibroblasts pre-treated with increasing doses of folic acid. (H) Scratch assay wound closure over time for indicated fibroblasts (FB) with and without 100μg/mL of folic acid (FA), (I and J) FB treated with folic acid (FA) with and without 50ug/mL of ruxolitinib (Ruxo) for the healthy (I) or keloid derived (J) FB cell lines. (K) Signal intensity for phosphorylated STAT3Ser727 (pSTAT3Ser727) associated with TOM70 staining (mitochondrial) or not associated with TOM70 (nuclear and cytoplasmic; Nuc + Cyto). (L) Representative image and close-up (inset) for co-stain for TOM70 (red), pSTAT3Ser727 (green), and DAPI (blue) as enumerated in K. (M and N) representative images (M) and quantitation of HIF1α staining signal co-occurring with DAPI nuclear marker in each identified cell (N). Results are representative of three independent experiments and displayed as mean ± SEM for 5–6 wells per condition (A–G) or triplicate wells (H–K, N). * = p <0.05; ** = p <0.01; *** = p < 0.001, **** = p < 0.0001 versus HV with diluent condition unless otherwise indicated as determined by ANOVA with Sidak adjustment.