Six1 regulates stem cell repair potential and self-renewal during skeletal muscle regeneration

Abstract

Satellite cells (SCs) are stem cells that mediate skeletal muscle growth and regeneration. Here, we observe that adult quiescent SCs and their activated descendants expressed the homeodomain transcription factor Six1. Genetic disruption of Six1 specifically in adult SCs impaired myogenic cell differentiation, impaired myofiber repair during regeneration, and perturbed homeostasis of the stem cell niche, as indicated by an increase in SC self-renewal. Six1 regulated the expression of the myogenic regulatory factors MyoD and Myogenin, but not Myf5, which suggests that Six1 acts on divergent genetic networks in the embryo and in the adult. Moreover, we demonstrate that Six1 regulates the extracellular signal-regulated kinase 1/2 (ERK1/2) pathway during regeneration via direct control of Dusp6 transcription. Muscles lacking Dusp6 were able to regenerate properly but showed a marked increase in SC number after regeneration. We conclude that Six1 homeoproteins act as a rheostat system to ensure proper regeneration of the tissue and replenishment of the stem cell pool during the events that follow skeletal muscle trauma.

Figure 1.

Satellite cells express Six1. (A) qRT-PCR analysis indicated expression of SIX family transcripts by freshly FACS-sorted SCs (Satellites), myogenic cells cultured in growth medium (Myoblasts), or induced to differentiate by serum removal for 3 d (Myotubes). Error bars indicate standard deviations. (B) Single myofibers isolated from EDL muscles of C57BL/6 mice. Myofibers were cultured in floating conditions and immunolocalized for Six1 and Pax7 or Myogenin proteins at different times after isolation. All quiescent, dividing, or differentiating SCs expressed Six1. Bars, 10 µm.

Figure 2.

Six1 gene disruption does not influence SC quiescence, activation, or proliferation. (A) Single myofibers isolated from EDL muscles of control (Tg:Pax7-CreERT2::Six1^flox/+) and Six1KO (Tg:Pax7^CreERT2/Six1^flox/flox) mice 1 wk after TM injection. Six1 protein expression is lost in Six1KO SCs (arrows). (B) The majority of SCs from Six1KO EDL and TA muscles are negative for Six1 expression. (C) Quantification of quiescent sublaminar Pax7⁺ SCs per EDL myofibers isolated from control and Six1KO mice 6 wk after TM injection. Six1 loss does not perturb SC quiescence in vivo. (D) EDL myofibers from control and Six1KO animals were cultured for 2 d to visualize SC activation (Pax7⁺/Ki67⁺). Six1 loss does not perturb SC activation ex vivo. (E) EDL myofibers from control and Six1KO animals were cultured for 3 d. SC descendants were immunolocalized for both Pax7 and Myogenin proteins. Six1 loss does not perturb SC proliferation ex vivo. (F) Primary myoblasts were isolated from control and Six1KO limb muscles. qRT-PCR analysis indicated expression of Six1, Pax7, and Six4 transcripts. Six1 gene disruption does not have an impact on Pax7 and Six4 expression levels. Error bars indicate standard deviations. *, P < 0.001. Bars, 10 µm.

Figure 3.

Six1 gene disruption perturbs myogenic differentiation of SC descendants ex vivo. 1 wk after TM treatment, EDL myofibers from control and Six1KO mice were plated on Matrigel, and cultures were analyzed after 6 and 9 d of culture ex vivo. (A) Myogenic cells grown for 6 d were immunolocalized for Desmin (myoblast marker) and MyHC (differentiation marker) proteins. (B) Six1KO cells exhibit limited differentiation potential ex vivo compared with control cells. (C) Myogenic cells grown for 9 d were immunolocalized for Six1, Pax7 (undifferentiated state marker), and MyHC (differentiated state marker) proteins. (D) Six1KO cells fuse less efficiently and form smaller myotubes compared with control cells. (E) qRT-PCR analysis indicated expression of Six1, CalcR, Pax7 (SC markers), and Myh1, Myh4 (differentiation markers) transcripts by differentiated myogenic cells. (F) Six1KO cell cultures generate more Pax7⁺ cells compared with control cells. (G) SC-derived myogenic cells grown for 9 d were immunolocalized for Pax7 and both MyoD and Ki67 proteins. (H) Six1KO cells generate more “reserve” cells (Pax7⁺/MyoD⁻/Ki67⁻; arrows) compared with control cells. Error bars indicate standard deviations. *, P < 0.02. Bars, 10 µm.

Figure 4.

Six1 expression by SCs is necessary for proper skeletal muscle regeneration. (A) 3 d after TM treatment, TA muscles of control and Six1KO mice were injured by a single CTX injection and analyzed at various times during the regeneration process. (B) Cryosections of 4-d regenerating TA muscles. Immunolocalization of MyHC emb proteins marks the newly formed myofibers. (C) Regenerating myofibers of Six1KO animals are smaller compared with controls. (D) Regenerating myofibers of Six1KO animals contain fewer nuclei compared with controls. (E) Cryosections of 7-d regenerating TA muscles. Laminin staining shows basal lamina of myofibers. Strong MyHC emb staining marks a population of small, delayed myofibers. (F) 7-d regenerating muscles of Six1KO animals exhibit a significant proportion of lagged myofibers compared with controls. (G) 7-d regenerating control and Six1KO muscles do not contain significantly different amounts of Pax7⁺ cells. (H) Cryosections of regenerated TA muscles 14 d after CTX injection. Laminin staining shows basal lamina of myofibers. Note the abnormal accumulation of matrix in Six1KO muscles. (I) Quantification of muscle fiber caliber in 14-d regenerated TA muscles. Regenerated Six1KO muscles contain smaller fibers compared with regenerated control muscles. (J) Cryosections of regenerated TA muscles 14 d after CTX injection. Dystrophin staining shows myofibers sarcolemma. Shown is the percentage of Six1⁺ myonuclei. (K) Regeneration of the muscle tissue results in a higher number of nuclei per myofiber on cross sections in controls but not in Six1KO animals. Error bars indicate standard deviations. *, P < 0. 01. Bars, 50 µm.

Figure 5.

Six1 activates MyoD and Myogenin expression by SCs in vivo. (A) qRT-PCR analysis of differentiating myogenic cells shows transient up-regulation of Six1, MyoD, and Myogenin expressions during the first day after serum removal. (B) qRT-PCR analysis of differentiating myogenic cells shows that Six1 silencing decreases MyoD and Myogenin expression but not Myf5 expression ex vivo. (C) EDL single myofibers were cultured for 3 d, and immunolocalized for Pax7 and MyoD protein expression. Representative myogenic cell clusters are shown. (D) Percentage of SC descendants at the surface of cultured myofibers. Loss of Six1 decreases the proportion of committed cells (Pax7⁻/MyoD⁺) and increases the proportion of undifferentiated (Pax7⁺/MyoD⁻) cells in clusters. (E) Cryosections of 4-d regenerating TA muscles. Laminin staining shows basal lamina of myofibers. Immunolocalization of Myogenin proteins marks differentiating myonuclei. (F) Six1 gene disruption in SCs results in decreased Myogenin⁺ nuclei numbers during muscle regeneration. (G) qRT-PCR analysis of Six1, MyoD, and Myogenin transcripts levels by 4-d regenerating TA muscles. Six1 gene disruption decreases MyoD and Myogenin expression in vivo. (H) Schematic representations of Myogenin, MyoD, and Myf5 regulatory regions region. Shown are the localization and sequences of E-box (bHLH binding) and MEF3 (SIX binding) sites. (I) qRT-PCR analysis of locus enrichment in ChIP assays from differentiating myogenic cells. MyoD and Six1 proteins are bound to the MyoD and Myogenin upstream regulatory elements, but not on Myf5 enhancer. Error bars indicate standard deviations. *, P < 0.04. Bars: (C) 20 µm; (E) 50 µm.

Figure 6.

Six1 limits SC self-renewal in vivo. (A) 3 d after TM treatment, TA muscles of control and Six1KO mice were injured by a single CTX injection and analyzed 30 d after the injury. (B) Single myofibers isolated from 30-d regenerated EDL muscles of control and Six1KO animals. Renewed Pax7⁺ SCs are located in sublaminar position around host myofibers in both control and Six1KO muscles. (C) Six1 is expressed by centrally located myonuclei and renewed SCs in control myofibers but not in Six1KO myofibers. Six1KO myofibers contain a higher number of renewed SCs (arrows). (D) Cryosections of 30-d regenerated TA muscles. Immunolocalization of Pax7 proteins mark quiescent SCs (arrows). (E) The SC pool is increased 2.4-fold in regenerated Six1KO TA muscles. (F) TA muscles of non-TM treated control and Six1KO mice were injured by a single CTX injection. Mice were then subjected to TM administration between 7 and 11 d after injury. Muscles were analyzed 30 d after the injury. (G) Cryosections of 30-d regenerated TA muscles. Immunolocalization of Pax7 proteins mark quiescent SCs (arrows). (H) Although the size of regenerated myofibers of control and Six1KO animals are similar, the SC pool is increased 2.1-fold in regenerated Six1KO TA muscles when TM was administrated after myogenesis has occurred. Error bars indicate standard deviations. *, P < 0.02. Bars: (B) 20 µm; (C) 10 µm; (D and G) 50 µm.

Figure 7.

Six1 gene disruption increases SCs self-renewal, but does not perturb the orientation of SC divisions. (A) FACS-sorted SCs were plated ex vivo and fixed after the first division. Typical doublets of sister SCs with Pax7^+/+ or Pax7^+/− gene signature are shown. (B) Six1KO SCs have higher self-renewal potential. (C) Immunolocalization of Pax7 and Ki67 proteins on myofibers separated from 4-d regenerating EDL muscles. (D) Quantification of SC division orientation. Six1 gene disruption does not have an impact on the rate of planar-to-perpendicular divisions. (E) EDL myofibers were separated from 4-d regenerating EDL muscles, and immunolocalized for Pax7 and Myf5 protein expression. Six1 gene disruption does not have an impact on the Myf5-negative satellite stem cell population. (F) FACS-sorted SCs separated on the basis of Myf5-Cre–driven reporter fluorescence and plated ex vivo. Immunolocalization of Pax7 and YFP proteins on YFP⁻ (stem) and YFP⁺ (committed) myoblasts. (G) qRT-PCR analysis indicated expression of Six1 transcripts by YFP⁺ and YFP⁻ SCs and myoblasts. Error bars indicate standard deviations. *, P < 0.01. Bars: (A) 10 µm; (C and F) 50 µm.

Figure 8.

Six1 negatively regulates ERK signaling in SCs. (A) qRT-PCR analysis of Six1, Ang1, Tie2, Etv4, and Dusp6 expression in proliferating myoblasts. Six1 gene disruption or silencing increases Ang1 and Etv4 transcription levels and decreases Dusp6 expression. (B) qRT-PCR analysis of Six1 and Dusp6 expression in proliferating myoblasts. Six1 overexpression increases Dusp6 transcription. (C) Real-time PCR analysis of locus enrichment in ChIP from proliferating myoblasts. Six1 proteins are bound to the MEF3 element upstream of the Dusp6 gene. (D) Immunolocalization of Pax7 and Dusp6 proteins on 3-d cultured EDL myofibers. Proliferating Six1KO SCs do not express Dusp6. (E) Detection of phosphorylated ERK from control (gray) and Six1KO (fushia) myoblasts. pI values are plotted against signal intensities. Different ERK isoforms and the relative increases in signal intensity of Six1KO over control are indicated (n = 3). Six1KO myoblasts present elevated ERK1 signaling ex vivo. (F) Immunolocalization of Pax7 and phosphorylated ERK1/2 proteins on 7-d regenerating EDL muscles. Six1KO SCs present elevated ERK signaling in vivo. (G) Immunolocalization of Pax7 and phosphorylated ERK1/2 proteins on 6-d cultured EDL myofibers. Six1KO SCs present elevated ERK signaling ex vivo. (H) EDL myofibers from control and Erk1^−/− animals. Immunostaining indicated that quiescent SCs express Pax7⁺ and have a correct sublaminar position in mutant muscles. Pax7⁺ sublaminar SCs were scored on EDL single fibers (left) and on TA cryosections (right). The SC pool is diminished in muscles from Erk1^−/− mice compared with controls. Error bars indicate standard deviations. *, P < 0.05. Bars, 10 µm.

Figure 9.

Dusp6 is required for restoring the SC pool during regeneration. TA and EDL muscles of control and Dusp6^−/− mice were injured by a single CTX injection and analyzed 30 d after the injury. (A) Single myofibers isolated from 30-d regenerated EDL muscles. Renewed Pax7⁺ SCs are located around host myofibers (arrows). (B) The SC pool is increased 2.4-fold on regenerated in Dusp6^−/− myofibers compared with control myofibers. (C) Cryosections of 30-d regenerated TA muscles. Immunolocalization of Pax7 and Laminin proteins allows visualization of sublaminar renewed SCs (arrows). (D) The SC pool is increased twofold within regenerated Dusp6^−/− muscles compared with control muscles. (E) Quantification of muscle fiber caliber in uninjured and regenerated TA muscles of Dusp6^−/− and control animals. No muscle defects were observed in mutant mice. Error bars indicate standard deviations. *, P < 0.004. Bars, 50 µm.

Figure 10.

Proposed model for the role of Six1 in regulating muscle tissue repair and SC niche occupancy.