Specific targeting of TGF-β family ligands demonstrates distinct roles in the regulation of muscle mass in health and disease

Abstract

The transforming growth factor-β (TGF-β) network of ligands and intracellular signaling proteins is a subject of intense interest within the field of skeletal muscle biology. To define the relative contribution of endogenous TGF-β proteins to the negative regulation of muscle mass via their activation of the Smad2/3 signaling axis, we used local injection of adeno-associated viral vectors (AAVs) encoding ligand-specific antagonists into the tibialis anterior (TA) muscles of C57BL/6 mice. Eight weeks after AAV injection, inhibition of activin A and activin B signaling produced moderate (∼20%), but significant, increases in TA mass, indicating that endogenous activins repress muscle growth. Inhibiting myostatin induced a more profound increase in muscle mass (∼45%), demonstrating a more prominent role for this ligand as a negative regulator of adult muscle mass. Remarkably, codelivery of activin and myostatin inhibitors induced a synergistic response, resulting in muscle mass increasing by as much as 150%. Transcription and protein analysis indicated that this substantial hypertrophy was associated with both the complete inhibition of the Smad2/3 pathway and activation of the parallel bone morphogenetic protein (BMP)/Smad1/5 axis (recently identified as a positive regulator of muscle mass). Analyses indicated that hypertrophy was primarily driven by an increase in protein synthesis, but a reduction in ubiquitin-dependent protein degradation pathways was also observed. In models of muscular dystrophy and cancer cachexia, combined inhibition of activins and myostatin increased mass or prevented muscle wasting, respectively, highlighting the potential therapeutic advantages of specifically targeting multiple Smad2/3-activating ligands in skeletal muscle.

Keywords: BMP; activin; hypertrophy; muscle; myostatin.

Fig. 1.

Myostatin and activins synergize to regulate muscle mass. The right tibialis anterior (TA) muscles of 6- to 8-wk-old male C57BL/6 mice were injected with AAVs encoding for modified activin A prodomain, activin B prodomain, and/or myostatin prodomain (left TA muscles were injected with equivalent doses of an AAV lacking a transgene). (A) Eight weeks post-AAV injection, the TA muscles were harvested and weighed (n = 4–6, paired Student’s t test, data groups with different letters achieved significance of P < 0.05). (B) Western blot analysis was used to assess prodomain expression and the phosphorylation of mTOR, S6RP, and Smad1/5. (C) Hematoxylin and eosin staining of TA muscles was performed on cryosections. (Scale bar: 100 µm.) (D) Muscle fiber size was quantified (n = 3, one-way ANOVA with Tukey's post hoc test, data groups with different letters achieved significance of P < 0.05; at least 150 myofibers were counted per TA muscle). (E) qRT-PCR was used to assess mRNA levels of Fbxo30 and Fbxo32 in response to activin/myostatin inhibition (n = 5, paired Student’s t test, data groups with different letters achieved significance of P < 0.05). (F) The right tibialis anterior (TA) muscles of 6- to 8-wk-old male C57BL/6 mice were injected with AAVs encoding for the activin B/myostatin prodomains, Smad6, or the three vectors combined (left TA muscles were injected with equivalent doses of an AAV lacking a transgene). Eight weeks post-AAV injection, the TA muscles were harvested and weighed (n = 4–6, one-way ANOVA with Tukey's post hoc test, data groups with different letters achieved significance of P < 0.05).

Fig. S1.

Specificity of the myostatin prodomain and effect of inhibiting activin and myostatin signaling on protein synthesis, protein degradation, and Smad1/5 pathways. (A) The myostatin prodomain blocks myostatin- or GDF11-induced activation of a Smad2/3-responsive luciferase reporter in HEK293T cells. (B) Demonstration that the standard dose of AAV:Act B prodomain (10¹⁰ vg) used in this study is saturating for endogenous activins (n = 5–6, paired Student’s t test, data groups with different letters achieved significance of P < 0.05). (C) Densitometric analysis of Western blots probed for total and phosphorylated forms of mTOR, S6RP, and Smad1/5 in response to activin and myostatin inhibition (n = 5, one-way ANOVA with Tukey’s post hoc test, data groups with different letters achieved significance of P < 0.05). (D) Expanded Western blot (n = 4–5 TAs) of S6RP phosphorylation in response to prodomain treatment. (E) qRT-PCR analysis was used to assess mRNA expression of Trim63 (Murf1) in muscles of WT mice treated with prodomains (n = 5–6, paired Student’s t test, data groups with different letters achieved significance of P < 0.05). (F) Western blot analysis of TA muscles was used to assess prodomain and Smad6 expression and the phosphorylation of Smad1/5. For the prodomain Western blot, two lanes, representing samples not related to the results for this paper, were spliced out (black line).

Fig. S2.

Verification of RNA-Seq analysis of prodomain-treated TA muscles. (A) qRT-PCR analysis was used to assess mRNA expression of Actc1, Klhl38, and Igfn1 in TA muscles of WT mice treated with activin B and/or myostatin prodomains. (B) The effect of Smad6 on prodomain-induced changes in the mRNA expression of Actc1, Klhl38, and Igfn1 in TA muscles of WT mice (n = 5, one-way ANOVA with Tukey’s post hoc test, data groups with different letters achieved significance of P < 0.05).

Fig. 2.

The effects of inhibiting Smad2/3-activating ligands in dystrophic mice. The right tibialis anterior (TA) muscles of 4- to 6-wk-old male and female C57BL/6 dystrophin^−/− mice (mdx) and dystrophin^−/−/utrophin^+/− mice (het) were injected with AAV vectors encoding for the modified activin B prodomain and/or the myostatin prodomain (left TA muscles were injected with equivalent doses of empty AAV). (A and B) Eight weeks post-AAV injection, the TA muscles were harvested and weighed (n = 5–10, paired Student’s t test, data groups with different letters achieved significance of P < 0.05). (C) Masson’s trichrome staining of TA muscles was performed on cryosections. (Scale bar: 100 µm.) (D) Muscle fiber size quantified (n = 3, paired Student’s t test, data groups with different letters achieved significance of P < 0.05; at least 150 myofibers were counted per TA muscle). (E) Despite increases in muscle mass with AAV injection, Western blot analysis indicated that there was very little prodomain expression by the experimental endpoint. (F) qRT-PCR analysis of fibrosis and atrophy genes CTGF, Fn1, Col3a1, Mfap4, and Igfn1 in the presence or absence of Smad2/3 signaling (n = 4–8, paired Student’s t test, data groups with different letters achieved significance of P < 0.05).

Fig. 3.

Persistent inhibition of Smad2/3 activating ligands increases mass and reduces fibrosis in muscles of dystrophic mice. The right tibialis anterior (TA) muscles of 4- to 9-wk-old male and female C57BL/6 dystrophin^−/− mice (mdx) and dystrophin^−/−/utrophin^+/− mice (het) were injected with AAVs encoding for the activin B and myostatin prodomains (left TA muscles were injected with equivalent doses of empty AAV). (A) Western blot analysis indicated the continued expression of prodomains 2 wk post-AAV injection. (B) TA muscles were harvested and weighed (n = 5–10, paired Student’s t test, data groups with different letters achieved significance of P < 0.05). (C) qRT-PCR analysis of fibrosis and atrophy genes CTGF, Fn1, Col3a1, Mfap4, and Igfn1 in the presence or absence of Smad2/3 signaling (n = 4–8, paired Student’s t test, data groups with different letters achieved significance of P < 0.05).

Fig. S3.

Skeletal muscle expression of fibrosis- and atrophy-associated genes and circulating levels of TGF-β proteins in WT and dystrophic mice. qRT-PCR analysis was used to assess mRNA expression of (A) fibrosis (CTGF, Fn1, Col3a1, and Mfap4) and atrophy (Igfn1 and Mss51) associated genes, and (B) TGF-β superfamily genes (Mstn, Gdf11, Inhba, Inhbb, Tgfb1, Tgfb2, and Tgfb3) in TA muscles of WT, mdx, and het mice. For the TGF-β genes, prodomains had no effects on expression. (C) Specific ELISAs were used to measure circulating TGF-β levels (all experiments, n = 5–6, one-way ANOVA with Tukey’s post hoc test, data groups with different letters achieved significance of P < 0.05; N.D., not detected).

Fig. 4.

Prevention of local muscle wasting in colon-26 tumor-bearing mice by inhibition of Smad2/3-activating ligands. C26 tumor fragments were implanted s.c. into 8- to 10-wk-old male BALB/c mice, and simultaneously the TA muscles were injected with AAVs encoding for activin B and/or myostatin prodomains (control mice received sham surgery and injection of equivalent doses of empty AAV). (A) After 18 d, when C26 tumor-bearing mice required euthanasia, TA muscles were harvested and weighed (n = 6–16, one-way ANOVA with Tukey’s post hoc test, data groups with different letters achieved significance of P < 0.05). (B) Hematoxylin and eosin staining on TA muscles was performed on cryosections of sham, cachectic, and activin B plus myostatin prodomain-treated mice. (Scale bar: 100 µm.) (C) Muscle fiber size quantified (n = 3, one-way ANOVA with Tukey’s post hoc test, data groups with different letters achieved significance of P < 0.05; at least 150 myofibers were counted per TA muscle). (D) qRT-PCR analysis was used to assess mRNA expression of Fbxo32, Trim63, Ky, Igfn1, Csrp3, and Tnfrsf12a (n = 5, one-way ANOVA with Tukey’s post hoc test, data groups with different letters achieved significance of P < 0.05).

Fig. S4.

Catabolic effects of C26 tumors in mice. C26 tumor fragments were implanted s.c. into 8- to 10-wk-old male BALB/c mice, and simultaneously the TA muscles were injected with AAVs encoding for activin B and/or myostatin prodomains (control mice received sham surgery and injection of equivalent doses of empty AAV). (A) Body weights were assessed and lean and fat masses measured by EchoMRI analysis and plotted as percentage change from the day of tumor implantation (n = 14–16, unpaired Student’s t test, data groups with different letters achieved significance of P < 0.05). (B and C) Western blot analysis was used to assess prodomain expression and the phosphorylation of Smad1/5, Smad2/3, and S6RP.