Hsp90 as a Myokine: Its Association with Systemic Inflammation after Exercise Interventions in Patients with Myositis and Healthy Subjects

Abstract

Compelling evidence supports the health benefits of physical exercise on the immune system, possibly through the molecules secreted by the skeletal muscles known as myokines. Herein, we assessed the impact of exercise interventions on plasma Heat shock protein 90 (Hsp90) levels in 27 patients with idiopathic inflammatory myopathies (IIM) compared with 23 IIM patients treated with standard-of-care immunosuppressive therapy only, and in 18 healthy subjects undergoing strenuous eccentric exercise, and their associations with the traditional serum markers of muscle damage and inflammation. In contrast to IIM patients treated with pharmacotherapy only, in whom we demonstrated a significant decrease in Hsp90 over 24 weeks, the 24-week exercise program resulted in a stabilization of Hsp90 levels. These changes in Hsp90 levels were associated with changes in several inflammatory cytokines/chemokines involved in the pathogenesis of IIM or muscle regeneration in general. Strenuous eccentric exercise in healthy volunteers induced a brief increase in Hsp90 levels with a subsequent return to baseline levels at 14 days after the exercise, with less pronounced correlations to systemic inflammation. In this study, we identified Hsp90 as a potential myokine and mediator for exercise-induced immune response and as a potential biomarker predicting improvement after physiotherapy in muscle endurance in IIM.

Keywords: cytokines; exercise; heat shock protein 90; idiopathic inflammatory myopathies; myokines; skeletal muscles.

Figure 1

Plasma Hsp90 levels in patients with idiopathic inflammatory myopathies treated with standard-of-care pharmacotherapy (control group, CG, n = 23) or with additional non-pharmacological intervention (intervention group, IG, n = 27) for 24 weeks. (A) Differences in plasma Hsp90 levels over 24 weeks in both IG and CG (lines represent the mean, whiskers represent the standard error of the mean). (B) An increase in plasma Hsp90 levels over 12 weeks predicted an improvement in muscle endurance assessed by Functional Index-2 (FI-2) over the entire 24-week intervention. Lower baseline plasma Hsp90 levels in IG predicted an increase in plasma Hsp90 over weeks 0–12 (C) and 0–24 (D). (E) An increase in plasma Hsp90 levels over weeks 0–12 in IG predicted an increase in plasma Hsp90 over weeks 0–24. Higher baseline plasma Hsp90 levels in CG predicted a decrease in plasma Hsp90 over weeks 0–12 (F) and 0–24 (G). (H) A decrease in plasma Hsp90 levels over weeks 0–12 in CG predicted a decrease in plasma Hsp90 over weeks 0–24. W, week.

Figure 2

Associations of changes in systemic levels of Hsp90 and selected inflammatory cytokines/chemokines in patients with idiopathic inflammatory myopathies treated with standard-of-care pharmacotherapy (control group, CG, n = 23) or with additional non-pharmacological intervention (intervention group, IG, n = 27) for 24 weeks. Lower baseline plasma Hsp90 levels in IG predicted an increase in serum IL-10 (A) and a decrease in MIP-1β (B) over weeks 0–24. (C) Changes in plasma Hsp90 levels in IG over weeks 0–24 correlated positively with changes in serum IL-6 over weeks 0–24. Changes in plasma Hsp90 levels in IG over weeks 0–12 correlated positively with changes in serum IL-2 (D) and negatively with MIP-1β (E) over weeks 0–24. (F) Lower baseline plasma Hsp90 levels in CG predicted a decrease in serum IL-10 over weeks 0–24. Changes in plasma Hsp90 levels in CG over weeks 0–24 correlated positively with changes in serum IL-8 (G) and TNF (H) over weeks 0–24. (I) Changes in plasma Hsp90 levels in CG over weeks 0–12 correlated positively with changes in serum IL-7 over weeks 0–24. W, week.

Figure 3

Systemic levels of Hsp90 and selected traditional markers of muscle damage and inflammation in healthy volunteers (n = 18) upon strenuous exercise. Systemic levels of Hsp90 (A) and of AST, ALT, CK, LD, and myoglobin (B) from baseline to 14 days after the intervention. Baseline plasma Hsp90 levels correlated positively with baseline serum levels of CK (C), ALT (D), and IL-4 (E). Plasma Hsp90 levels at 30 min correlated positively with serum levels of myoglobin (F) and MCP-1 (G) at 30 min. The red line (A) and the black/gray lines (B) represent the mean and the whiskers (A,B) represent the standard error of the mean.

Figure 4

Associations of changes in systemic levels of Hsp90, traditional markers of muscle damage and inflammation in healthy volunteers (n = 18) upon strenuous exercise. The change in plasma Hsp90 from baseline to 1 h was inversely correlated with change from baseline to 1 h in serum levels of CK (A), myoglobin (B), and ALT (C). (D) A decrease in Hsp90 from 30 min to 14 days was associated with a decrease in LD from 30 min to 14 days. (E) An increase in Hsp90 from baseline to 30 min was associated with an increase in IL-4 from baseline to 30 min. (F) A decrease in Hsp90 from 30 min to 14 days was associated with a decrease in IL-4 from 30 min to 14 days.

Figure 5

Signaling pathways of Hsp90 in the inflammatory process. Solid blue arrows indicate signaling pathways that upregulate Hsp90 expressions in the inflammatory process. IL-6 upregulates Hsp90 via both the JAK-STAT3 and the NF-IL6 pathways. INF- γ increases Hsp90 expressions through the JAK-STAT1 pathway. Other cytokines/chemokines such as TNF and IL-1 can upregulate Hsp90 by activating the NF-kB complex. Dashed green arrows denote relevant pathways influenced by Hsp90. Hsp90 binds and stabilizes JAK and STAT proteins, as well as the nonreceptor tyrosine kinase HCK, which is known to regulate innate immune response. Reciprocatively, Hsp90 is required to induce NF-kB. HCK, hematopoietic cell kinase; IL, interleukin; INF, interferon; JAK, Janus kinase; MAPK, mitogen-activated protein kinase; NF-kB, nuclear factor kappa-light-chain-enhancer of activated B cells; Ras, rat sarcoma protein, or Ras GTPase protein; STAT, signal transducer and activator of transcription.