Apoptosis-inducing factor regulates skeletal muscle progenitor cell number and muscle phenotype

Abstract

Apoptosis Inducing Factor (AIF) is a highly conserved, ubiquitous flavoprotein localized in the mitochondrial intermembrane space. In vivo, AIF provides protection against neuronal and cardiomyocyte apoptosis induced by oxidative stress. Conversely in vitro, AIF has been demonstrated to have a pro-apoptotic role upon induction of the mitochondrial death pathway, once AIF translocates to the nucleus where it facilitates chromatin condensation and large scale DNA fragmentation. Given that the aif hypomorphic harlequin (Hq) mutant mouse model displays severe sarcopenia, we examined skeletal muscle from the aif hypomorphic mice in more detail. Adult AIF-deficient skeletal myofibers display oxidative stress and a severe form of atrophy, associated with a loss of myonuclei and a fast to slow fiber type switch, both in "slow" muscles such as soleus, as well as in "fast" muscles such as extensor digitorum longus, most likely resulting from an increase of MEF2 activity. This fiber type switch was conserved in regenerated soleus and EDL muscles of Hq mice subjected to cardiotoxin injection. In addition, muscle regeneration in soleus and EDL muscles of Hq mice was severely delayed. Freshly cultured myofibers, soleus and EDL muscle sections from Hq mice displayed a decreased satellite cell pool, which could be rescued by pretreating aif hypomorphic mice with the manganese-salen free radical scavenger EUK-8. Satellite cell activation seems to be abnormally long in Hq primary culture compared to controls. However, AIF deficiency did not affect myoblast cell proliferation and differentiation. Thus, AIF protects skeletal muscles against oxidative stress-induced damage probably by protecting satellite cells against oxidative stress and maintaining skeletal muscle stem cell number and activation.

Figure 1. Atrophy in Hq skeletal muscle.

Six-week old, AIF hypomorphic, harlequin (Hq) mutant mice (pdcd8^{Hq /Y}; B) show reduced body size, lordokyphosis and reduced hair density in comparison with wild-type littermate (WT; A). The hematoxylin-eosin staining (C, D) on transverse sections of 3 month-old WT (C) and Hq (D) soleus muscles show atrophy in Hq mice. Bar = 20 µm. Quantitation of fiber cross sectional area (CSA) (E) confirms the atrophy in Hq soleus and EDL muscles of 3-month old mice of indicated genotype. Sections of WT (F) and Hq (G) EDL (F, G) muscles were immunostained using anti-laminin and bis-benzimide to reveal sarcolemma and nuclei. Arrows in F and G show myonuclei inside the sarcolemma. The number of myonuclei is reduced in Hq soleus and EDL muscles, as quantitated in H. (I) Dry and wet weights of individual skeletal muscles and cardiac muscle of 2 month-old Hq and WT mice confirm a muscle weight loss in Hq mice. gastroc.: gastrocnemius.Bar = 25 µm. * indicates P<0.05 vs WT.

Figure 2. Oxydative stress in Hq skeletal muscle.

The 8OHdG immunostaining (A, B) on transverse sections of 3 month-old WT (A) and Hq (B) soleus muscles show increased free radical damage in Hq mice. Bar = 10 µm. (C) Real-time PCR analysis of gclm transcript abundance, gene overexpressed in oxidative stress models in soleus and EDL from 2-month old wild-type and Hq mice, Western blot (D) and real-time PCR (E) analyses of endogenous Nrf2 (D) and NQO1 (E) reveal dysfunctions in the Nrf2 redox signaling.

Figure 3. AIF deficiency induces a slow fiber type switch, associated to an increased MEF2 activity.

Immunofluorescent analysis of type I myofibers on transverse sections of WT EDL muscle (A) and Hq EDL muscle (B). Bar = 60 µm. Quantification of type I and type II fibers in WT and Hq soleus (C) and EDL (E) muscles confirms the slow fiber type switch in Hq mice. Quantitation of fiber cross sectional area (CSA) specifies the atrophy of some fiber types in soleus (D) and EDL (F) muscles. * indicates P<0.05 vs WT. (G) β-Galactosidase staining of soleus and EDL muscles isolated from 2-month-old MEF2 indicator mice (3×MEF2-lacZ) or Hq mice crossed with MEF2 indicator mice (3×MEF2-lacZ×Hq). Expression of the lacZ transgene depended on MEF2 activity. Blue staining indicates augmented MEF2 activity in tissues. Up-regulation of desMEF2-lacZ transgene expression in 2 month-old Hq soleus and EDL muscles occurs without any significant changes in the abundance of major MEF2 isoforms (H). We notice that MEF2C is differentially spliced in soleus and EDL muscles.

Figure 4. Delayed skeletal muscle regeneration in Hq mice.

Hematoxylin-eosin histological staining of transverse sections of regenerating EDL (A–F) and soleus (G, H) muscle from WT (A, C, E and G) and Hq (B, D, F and H) mice 3- (A, B), 5- (C, D) and 10 days (E–H) following cardiotoxin injury. Bar = 10 µm. Arrows in A and D point out newly formed myotubes with central nuclei. (I) Quantitation of cross sectional area (CSA) of myotubes in regenerating soleus and EDL muscles 3, 5, 10 and 20 days following cardiotoxin injection confirms the delay in muscle regeneration in Hq mice.

Figure 5. A slow fiber type switch in regenerating Hq muscles.

Quantification of fibers expressing type I, type II, embryonic and neonatal MyHC in WT and Hq soleus (A, C, E, G) and EDL (B, D, F, H) muscles 3 (A, B), 5 (C, D), 10 (E, F) and 20 (G, H) days following cardiotoxin injury indicates a slow fiber type switch and an increase in immature regenerating myofibers in Hq mice, the latter corroborating the delay in skeletal muscle regeneration of Hq mice. * indicates P<0.05 vs WT.

Figure 6. Loss of satellite cells in AIF deficient mice.

(A) Representative picture of a Pax7-positive satellite cell attached to a freshly isolated myofiber from WT EDL muscle. (B) Quantification of the number of Pax7-positive satellite cell nuclei per 100 myofiber nuclei, indicates a significant reduction of Pax7-positive satellite cells in isolated Hq EDL myofibers. Measurements were made from six different animals of each genotype by counting the number of Pax7-positive cells per fiber and the total number of myofiber nuclei per fiber as visualized by Bis-benzimide staining. Pax7 (C) and M-cadherin (E) antibody staining of WT EDL muscles show Pax7⁺ (arrow in C), M-cadherin⁺ nuclei (arrow in E) and myonuclei (arrowheads in C and E) located under the basement membrane, marked by anti-laminin staining from WT EDL cross sections. Quantification of the number of Pax7-positive (D) and M-cadherin-positive (F) satellite cell nuclei per 100 myofiber cross sections indicates a significant reduction of Pax7 positive satellite cells in Hq EDL muscles and a reduction of M-cadherin⁺ cells in soleus and EDL Hq muscles. (G, I) Representative pictures of an activated satellite cell (arrow in G), co-immunostained with MyoD and Pax7 antibodies and of satellite cells expressing either Pax7 and M-cadherin (arrow in I, top panel) or M-cadherin only (arrow in I, lower panel) on Hq EDL cross sections. (H) Quantification of the number of activated satellite cells (Pax7 and MyoD positive nuclei) and of non-activated satellite cells (Pax7-positive MyoD-negative nuclei) indicated a higher amount of activated satellite cells in both Hq soleus and EDL muscles compared to WT muscles. (J) Among the M-cadherin-positive satellite cells, about 30% are Pax7-negative in the WT soleus only. Measurements in D, F, H and J were made from five different animals of each genotype on soleus and EDL muscle sections. Nuclei were stained with Bis-benzimide. Bar = 10 µm (A), 25 µm (C, E, G, I).

Figure 7. AIF deficiency decreases satellite cell clonogenicity, without affecting their proliferation and their differentiation.

(A) Representative picture of WT and Hq microcolony grown for 4 days. Cells were stained with anti-MyoD and anti-BrdU antibodies and Bis-benzimide to reveal myogenic S-phase and total nuclei, respectively. Bar = 10 µm. (B) Quantification of the percentage of cycling cells in WT and Hq myogenic colonies of at least 2 cells, grown for 3 or 4 days, after 2 h BrdU pulse and stained as in A, shows no difference in the proliferation rate of Hq and WT mpc. (C) Assessment of WT and Hq microcolony formation after 3 or 4 days in culture and stained as in A, indicates a delay in the microcolony formation of Hq satellite cell primary cultures. (D) Differentiated WT and Hq myocytes plated at high density and grown for 3 days were detected by immunostaining for myogenin and sarcomeric MyHC. Total nuclei were stained with Bis-benzimide. Bar = 15 µm. (E, F) After 3 days in culture, the myotube diameter (E) and the number of nuclei per myofiber (F) were similar in WT and Hq primary cultures. At least 400 myotubes were analyzed. n.s. not significant.