Efficient generation of iPS cells from skeletal muscle stem cells

Abstract

Reprogramming of somatic cells into inducible pluripotent stem cells generally occurs at low efficiency, although what limits reprogramming of particular cell types is poorly understood. Recent data suggest that the differentiation status of the cell targeted for reprogramming may influence its susceptibility to reprogramming as well as the differentiation potential of the induced pluripotent stem (iPS) cells that are derived from it. To assess directly the influence of lineage commitment on iPS cell derivation and differentiation, we evaluated reprogramming in adult stem cell and mature cell populations residing in skeletal muscle. Our data using clonal assays and a second-generation inducible reprogramming system indicate that stem cells found in mouse muscle, including resident satellite cells and mesenchymal progenitors, reprogram with significantly greater efficiency than their more differentiated daughters (myoblasts and fibroblasts). However, in contrast to previous reports, we find no evidence of biased differentiation potential among iPS cells derived from myogenically committed cells. These data support the notion that adult stem cells reprogram more efficiently than terminally differentiated cells, and argue against the suggestion that "epigenetic memory" significantly influences the differentiation potential of iPS cells derived from distinct somatic cell lineages in skeletal muscle.

Figure 1. Isolation and differentiation potential of myofiber-associated skeletal muscle precursors (SMPs) and Sca-1+ mesenchymal progenitors.

(A) Myofiber-associated cells were isolated from mouse hind limb muscles and triceps. SMPs were isolated by FACS as CD45− Mac1− Sca1− β1-integrin+ CXCR4+, while bipotent, adipogenic/fibrogenic cells were identified and isolated as CD45− Mac1− Sca-1+ (abbreviated Sca1+; [3], [6]). All cell populations were double-sorted, yielding highly purified cell populations that subsequently were seeded onto irradiated MEFs in the presence of doxycyline (see Figure 2). (B) SMPs but not Sca1+ cells can differentiate into myosin heavy chain+ (MyHC+) cells, indicating their myogenic differentiation potential (bottom panel). Conversely, Sca1+ cells but not SMPs can form adipogenic cells, as indicated by Oil-Red-O staining (top panel). CXCR4− cells are a mixture of myoblasts (bottom panel) and fibroblasts, some of which can adopt an adipocytic fate under adipogenic culture conditions (top panel).

Figure 2. SMPs and Sca-1+ mesenchymal progenitors reprogram into pluripotent iPS cells with high efficiency.

(A–F) iPS cells generated from SMPs (“SMP iPS” cells) show embryonic stem cell-like morphology (A), and stain for the pluripotency markers Oct 4 (B) and SSEA-1 (C). Brightfield and corresponding fluorescence images (red) are shown for B and C. SMP iPS cells also show alkaline phosphatase activity (D). When SMP iPS cells were injected into C57Bl/6 blastocysts, the resulting chimeras had agouti coat color (E), indicating contribution of the SMP iPS cells to chimeric tissues. Offspring from matings between SMP iPS cell chimeric mice and C57Bl/6 mice have brown coat color (F), demonstrating the capacity of SMP iPS cells to provide germline transmission.

Figure 3. Clone-sorted myofiber-associated stem cells reprogram at high efficiency.

(A) Experimental strategy for clone-sorting and reprogramming sorted cells. Myofiber-associated cells from transgenic mice carrying dox-inducible transgenes (Oct4, Sox2, c-Myc, Klf4) and labeled with constitutively-expressed Tdtomato , were double-sorted for purity, and seeded into 96-well plates, at one cell per well, on irradiated MEFs. After three weeks in dox-containing media, emergent colonies were trypsinized and passaged in the absence of dox, and iPS lines were established which showed embryonic stem cell-like morphology, stained for the pluripotent marker SSEA-1, and showed alkaline phosphatase activity (data not shown).

Figure 4. Increased efficiency of reprogramming of SMPs and Sca1+ cells is independent of initial levels of endogenous expression of reprogramming factors.

Quantitative real-time PCR using primers that detect endogenously expressed Oct4, Sox2, Klf4, and cMyc mRNAs was performed on freshly sorted SMPs, Sca1+ cells and CXCR4− cells (A) and on reprogrammed SMP and Sca1+ iPS cells and mES cells (three different lines tested for each cell type, B). None of the freshly-isolated myofiber-associated cell populations expresses Oct4 or Sox2 at appreciable levels compared to pluripotent cells. SMPs and CXCR4− cells express Klf4 and c-Myc at significantly higher levels than in Sca1+ cells. Data are normalized for beta2-microglobulin expression and presented as mean +/− S.E.M. for the indicated number of replicates. *p<0.05; **p<0.001.

Figure 5. Teratomas from SMP iPS and Sca1+ iPS cells do not show biased differentiation into skeletal muscle by histology.

(A) Representative image of an H&E stained section of a teratoma formed from SMP iPS cells (A). The presence of skeletal muscle in teratomas from mES, SMP iPS or Sca1+ iPS was assessed by histology, and is shown as the percent area of the tumor occupied by muscle (B). No significant differences were detected in the fraction of striated muscle present in teratomas derived from mES, SMP iPS or Sca1+ iPS cells.

Figure 6. Teratomas from SMP iPS and Sca1+ iPS cells do not show biased differentiation into skeletal muscle by Western blot.

(A) Representative Western blots of teratomas from SMP iPS, Sca1+ iPS, or mES cells, stained with the indicated antibodies against skeletal muscle myosin heavy chain (MyHC), cytokeratin 8 (CK8), cytokeratin 14 (CK14) or GAPDH as a control. Three different lines were used for each pluripotent cell type to generate the teratomas. (B) Western blots were quantified by optical densitometry using ImageJ, and results reported for MyHC (top), CK8 (middle) and CK14 (bottom) as mean intensity (arbitrary units), normalized to the loading control (GAPDH). Error bars reflect standard error. No significant differences were detected in the presence of MyHC (a mesodermal and skeletal muscle marker) or CK8 (an endodermal marker) among any of the three sources of pluripotent cells. SMP iPS showed 2–3 fold greater representation of CK14 (an ectodermal marker) than Sca1+ iPS or mES. N.D. – not detectable.

Figure 7. Sca1+ cells but not SMPs contain a PDGFRα+ population.

Sca1+ cells or SMPs were stained for PDGFRα and analyzed by FACS. Representative FACS plots show that only Sca1+ cells (left) contain a substantial population of PDGFRα+ cells (∼30% of Sca1+ cells), whereas SMPs (right) are PDGFRα−.