Growth differentiation factor-15 is associated with muscle mass in chronic obstructive pulmonary disease and promotes muscle wasting in vivo

Abstract

Background: Loss of muscle mass is a co-morbidity common to a range of chronic diseases including chronic obstructive pulmonary disease (COPD). Several systemic features of COPD including increased inflammatory signalling, oxidative stress, and hypoxia are known to increase the expression of growth differentiation factor-15 (GDF-15), a protein associated with muscle wasting in other diseases. We therefore hypothesized that GDF-15 may contribute to muscle wasting in COPD.

Methods: We determined the expression of GDF-15 in the serum and muscle of patients with COPD and analysed the association of GDF-15 expression with muscle mass and exercise performance. To determine whether GDF-15 had a direct effect on muscle, we also determined the effect of increased GDF-15 expression on the tibialis anterior of mice by electroporation.

Results: Growth differentiation factor-15 was increased in the circulation and muscle of COPD patients compared with controls. Circulating GDF-15 was inversely correlated with rectus femoris cross-sectional area (P < 0.001) and exercise capacity (P < 0.001) in two separate cohorts of patients but was not associated with body mass index. GDF-15 levels were associated with 8-oxo-dG in the circulation of patients consistent with a role for oxidative stress in the production of this protein. Local over-expression of GDF-15 in mice caused wasting of the tibialis anterior muscle that expressed it but not in the contralateral muscle suggesting a direct effect of GDF-15 on muscle mass (P < 0.001).

Conclusions: Together, the data suggest that GDF-15 contributes to the loss of muscle mass in COPD.

Keywords: Atrophy; COPD; Electroporation; GDF‐15; Muscle mass.

Figure 1

Serum GDF‐15 is elevated in patients with COPD and is associated with muscle size and exercise capacity. Serum levels of GDF‐15 were determined in patients with COPD and age‐matched controls and compared with RF_CSA and exercise performance in two cohorts of COPD patients. (A) Serum GDF‐15 was elevated in COPD patients (n = 46) compared with healthy controls (n = 21, Mann–Whitney, P < 0.001). (B, D) RF_CSA was inversely correlated (Pearson's r) with serum GDF‐15 in two independent cohorts of COPD patients and controls [(B) RBH patients and controls r = −0.381, P < 0.001; (D) St George's patients only r = −0.357, P = 0.018]. (C, E) Exercise capacity was inversely correlated with serum GDF‐15 in two independent cohorts of COPD patients and controls [(C) RBH patients and controls measured as 6 MW, r = −0.466, P < 0.001; (E) St George's patients only, measured as ISW, r = −0.488, P < 0.001]. Patients are shown as black circles, and controls are shown as open circles. COPD, chronic obstructive pulmonary disease; GDF, growth differentiation factor; RBH, Royal Brompton Hospital.

Figure 2

GDF‐15 is elevated in COPD patients with muscle atrophy. Serum GDF‐15 levels were compared in patients defined as having muscle atrophy (as described in the Methods section, n = 20) and those with a normal muscle size (n = 69) for both the RBH and St George's cohorts. Patients with atrophy had higher serum GDF‐15 levels than those without atrophy in both cohorts (RBH, P = 0.011, and St George's, P = 0.022, Mann–Whitney test). COPD, chronic obstructive pulmonary disease; GDF, growth differentiation factor; RBH, Royal Brompton Hospital; SGH, St George's Hospital.

Figure 3

Expression of GDF‐15 and myostatin is elevated in the muscle of patients with COPD. The expression of (A) GDF‐15 and (B) myostatin was determined by real‐time quantitative PCR in skeletal muscle biopsies of patients (n = 49) and controls (n = 21) from the Royal Brompton Hospital cohort as described in the Methods section. The expression of both mRNAs was higher in the patients than in the controls (GDF‐15, P < 0.001; myostatin, P < 0.001 Mann–Whitney). (C) Muscle GDF‐15 expression and (D) myostatin expression were also weakly correlated with exercise capacity measured as 6 MW in this cohort (r = 0.250, P = 0.04, and r = 0.301, P = 0.012, respectively, Pearson's r). COPD, chronic obstructive pulmonary disease; GDF, growth differentiation factor.

Figure 4

Expression of GDF‐15 in the skeletal muscle of mice does not reduce growth but does cause muscle wasting. Mice were electroporated with either a control vector into both tibiales anteriores (pCAGGS1/pCAGGS2, n = 8) or a control vector into one tibialis anterior and a GDF‐15 expression vector into the contralateral tibialis anterior (pCAGGS3/GDF‐15, n = 8) and then left for 14 days. (A) There was no difference in the body weight of the mice 14 days after electroporation and (B) no difference in the growth rate of the mice. (C) There was no difference in the proportionate size of the tibialis anteriores in control mice (pCAGGS1 and pCAGGS2), but the GDF‐15‐expressing tibiales anteriores were smaller than the control tibiales anteriores in the experimental animals (pCAGGS3 and GDF‐15 P = 0.033 paired t‐test). Control tibiales anteriores from the experimental mice were not smaller than the tibiales anteriores from the control mice, suggesting that there was no systemic effect of GDF‐15. GDF, growth differentiation factor. TA, (Tibialis anteriores)

Figure 5

GDF‐15 over‐expression causes muscle fibre atrophy. Tibialis anteriores muscles from mice (n = 8) over‐expressing GDF‐15 in one tibialis anterior were mounted onto cork and then sectioned as described in the Methods section. Myofibre diameter was measured in sections taken from levels 6–8 (representing the middle of the muscle), and minimum feret's diameter was measured. (A) Average fibre diameter was smaller at all three levels of the muscle in the GDF‐15‐expressing tibialis anterior (P = 0.02 at level 6, P < 0.001 at levels 7 and 8, unpaired t‐test). (B) Fibres were binned into 12 bins of fibre size ranging from less than 15 µm to more than 65 µm, and the proportion of fibres in each bin was determined. Consistent with the reduction in average fibre diameter, there was a leftward shift in fibre‐type distribution profile of the GDF‐15‐expressing muscle compared with the contralateral muscle. (C) Representative images of control (pCAGGS) and GDF15‐over‐expressing tibialis anteriors stained with haemotoxylin and eosin. (D) Representative images of sections from control (pCAGGS, left‐hand image) and GDF‐15 over‐expressing (right‐hand image) stained for GDF‐15. GDF, growth differentiation factor.