Distinct populations of adipogenic and myogenic Myf5-lineage progenitors in white adipose tissues.

Brown adipose tissues (BAT) are derived from a myogenic factor 5 (Myf5)-expressing cell lineage and white adipose tissues (WAT) predominantly arise from non-Myf5 lineages, although a subpopulation of adipocytes in some WAT depots can be derived from the Myf5 lineage. However, the functional implication of the Myf5- and non-Myf5-lineage cells in WAT is unclear. We found that the Myf5-lineage constitution in subcutaneous WAT depots is negatively correlated to the expression of classical BAT and newly defined beige/brite adipocyte-specific genes. Consistently, fluorescent-activated cell sorting (FACS)-purified Myf5-lineage adipo-progenitors give rise to adipocytes expressing lower levels of BAT-specific Ucp1, Prdm16, Cidea, and Ppargc1a genes and beige adipocyte-specific CD137, Tmem26, and Tbx1 genes compared with the non-Myf5-lineage adipocytes from the same depots. Ablation of the Myf5-lineage progenitors in WAT stromal vascular cell (SVC) cultures leads to increased expression of BAT and beige cell signature genes. Strikingly, the Myf5-lineage cells in WAT are heterogeneous and contain distinct adipogenic [stem cell antigen 1(Sca1)-positive] and myogenic (Sca1-negative) progenitors. The latter differentiate robustly into myofibers in vitro and in vivo, and they restore dystrophin expression after transplantation into mdx mouse, a model for Duchenne muscular dystrophy. These results demonstrate the heterogeneity and functional differences of the Myf5- and non-Myf5-lineage cells in the white adipose tissue.


Adipose SVF cell isolation and culture
The white adipose SVF cells were isolated using collagenase digestion followed by density separation ( 17,27 ). Briefl y, the anterior subcutaneous WAT (asWAT), inguinal WAT (ingWAT), and epididymal WAT (eWAT) were collected and minced into 2 ‫ف‬ 5 mm 2 pieces. The WAT pieces were then digested in 1.5 mg/ml collagenase at 37°C for 1.5 ‫ف‬ 2 h. The digestions were terminated with DMEM containing 10% fetal bovine serum (FBS), and then fi ltered through 100 µm fi lters to remove connective tissues and undigested trunks of tissues. Cells were then centrifuged at 450 g for 5 min to separate the SVF cells in the sediment and the lipidcontaining adipocytes in the fl oating layer. The freshly isolated SVF cells from the WAT were seeded and cultured in growth medium containing DMEM, 20% FBS, 1% penicillin/streptomycin (P/S) at 37°C with 5% CO 2 for three days, followed by feeding with fresh medium every two days. For adipogenic differentiation, the cells were induced with induction medium containing DMEM, 10% FBS, 2.85 µM insulin, 0.3 µM dexamethasone (DEXA), and 0.63 mM 3-isobutyl-methylxanthine (IBMX) for three days upon confl uence, and then differentiated in differentiation medium containing DMEM, 10% FBS, 200 nM insulin, and 10 nM T3 for four days until adipocytes matured. For myogenic differentiation, the cells were induced with DMEM, 2% horse serum, 1% P/S for six days upon confl uence. To ablate Myf5-lineage cells in culture, the SVF cells from WAT and BAT of the Myf5-Cre/ Rosa26-iDTR mice were treated with diphtheria toxin (DT, 200 ng/ml) for 48 h. To avoid the effect of cell density on adipogenic or myogenic differentiation, the control and DT-treated cells were induced to differentiate when they reached 90% confl uence.

Muscle myoblast isolation and culture
Myoblast cells were isolated using type I collagenase and dispase B digestion ( 28 ). Briefl y, the skeletal muscles near the as-WAT from the Myf5-Cre/Rosa26-tdTomato mice were collected, minced, and digested. The digestions were stopped with F-10 Ham's medium containing 20% FBS, and then the cells were fi ltered through 70 m and centrifuged at 450 g for 5 min. The pelleted cells were seeded on collagen-coated dishes and cultured in growth medium containing F-10 Ham's medium with 20% FBS, 4 ng/ml basic fi broblast growth factor (bFGF), and 1% P/S at 37°C with 5% CO 2 . The medium was refreshed every two days. Cells were trypsinized with 0.25% trypsin under close monitoring only to lift off myoblasts (but not fi broblasts) during passages. Enriched myoblasts were used for transplantation (5 × 10 4 cells per muscle).

FACS
SVF cells were isolated from SAT tissues of Myf5-Cre/Rosa26-tdTomato mice as described above. The red fl uorescent protein (RFP)-positive (tdTomato + ) and RFP-negative (tdTomato Ϫ ) SVF cells represent Myf5-lineage and non-Myf5-lineage cells, respectively. Freshly isolated SVF cells were labeled with lineage markers CD45, CD31 TER-119, and CD11b (Lin) conjugated with PE-Cy7, and stem cell antigen 1 (Sca1) conjugated with FITC, as described ( 27 ). All antibodies were purchased from eBioscience. After staining, SVF cells were isolated and fi ltered through 30 µm fi lter before sorting. Nonlabeled SVF cells from wild-type (WT) mice were used as negative control for gating purposes. From the Lin- ( 15 ). However, a recent study by Sanchez-Gurmaches and colleagues provides evidence that a subset of white adipocytes is derived from the Myf5-lineage mesenchymal progenitors ( 16 ). The relative contribution of Myf5-lineage cells to WAT appears to vary among different WAT depots ( 16 ). Interestingly, Pax3 (an upstream regulator of Myf5 during myogenesis) lineage cells also contribute to a subset of white adipocytes in different depots ( 17 ). Due to the Myf5-lineage origin of brown adipocytes, it is plausible to hypothesize that the Myf5-lineage progenitors are more likely to give rise to the adaptive beige adipocytes. However, until now the functional implication of Myf5-lineage and non-Myf5-lineage adipocytes has not been explored. It has been widely accepted that WAT contains an abundant and accessible source of adult stem cells with multiple differentiation potentials (18)(19)(20). Cells from the stromal vascular fraction (SVF) of WAT can differentiate into adipocytes, myocytes, osteocytes, chondrocytes, cardiomyocytes, hepatocytes, epithelial cells, endothelial cells, and even neuron-like cells (20)(21)(22)(23)(24). It has been shown that adipose tissue-derived progenitors have myogenic potential in vitro and in vivo ( 25,26 ). Our previous study further demonstrates that the myogenic progenitors are enriched in the non-aP2-lineage population of SVF cells in WAT ( 27 ). A longstanding question is whether the multilineage potential of WAT SVF cells results from a multipotent mesenchymal stem cell population or from separate subpopulations of unipotent progenitors. If the latter is true, what are the unique markers for each subpopulation of progenitor cells?
In this study, we used cell-lineage labeling, cell ablation, fl uorescence-activated cell sorting (FACS), and cell transplantation to demonstrate the phenotype and function of the Myf5-lineage and non-Myf5-lineage progenitors in various depots of subcutaneous adipose tissue (SAT), a tissue that has tremendous plasticity in cold-induced browning. We showed that the Myf5-lineage SVF cells contain subpopulations of adipogenic and myogenic progenitors that can be prospectively isolated based on Sca1 expression. We found that the Myf5-lineage adipocytes expressed lower levels of brown and beige adipocyte markers than the non-Myf5-lineage adipocytes within the same SAT depots. We further analyzed the Myf5-lineage myogenic progenitors of the SAT SVF cells and found that the Myf5-lineage progenitors are the only population that has the myogenic differentiation potential. The FACS-purifi ed, Myf5-lineage myogenic progenitors are capable of restoring dystrophin expression in mdx mice and form exclusively slow fi bers after transplantation into wild-type (WT) recipient mice. These results provide new insights into the roles of Myf5lineage cells in white adipose tissue and suggest that Myf5lineage cells in the BAT and WAT are functionally distinct.

Animals
All procedures involving mice were performed in accordance with Purdue University's Animal Care and Use Committee. Mice were housed in the animal facility with free access to standard rodent chow and water. All mice were from Jackson Laboratory one-way ANOVA, as appropriate. Effects were considered significant at P < 0.05.

Myf5 lineage contributes to both mature adipocytes and immature SVF cells
To investigate the progeny of Myf5-lineage cells in various tissues, we conducted lineage-tracing experiments using Myf5-Cre driver and Rosa26-tdTomto reporter mice ( 29,30 ), in which Myf5-lineage cells were labeled by tdTomato, an improved RFP. Consistent with previous reports ( 15 ), the vast majority of cells in muscle and BAT cross-sections were RFP-positive (supplementary Fig. I, A, B), suggesting that the Myf5 lineage contributes to these tissues. This observation also demonstrates the effectiveness of our lineage tracing model. RFP-positive cells were also detected in the heart, brain, and spleen, but not in the liver, lung, or kidney (supplementary To further characterize whether the Myf5 lineage gives rise to adipose progenitors in WAT, we isolated the SVF cells and found RFP-positive cells in the SVF cells ( Fig. 1A -C ). Quantitative analysis indicated that there are about 31, 11, and 14% RFP-positive cells in the SVF of as-WAT, ingWAT, and eWAT, respectively ( Fig. 1D-F ). To examine whether at least some of the RFP-positive SVF cells are adipogenic progenitors, we fi rst examined the expression of Sca1, an established marker for adipose stem cells ( 31,32 ). FACS analysis (refer to Fig. 3 for details) indicated that 13, 4, and 4% of freshly isolated Lin-negative SVF cells in asWAT, ingWAT, and eWAT were double-positive for RFP + Sca1 + , respectively (supplementary Fig. II). We next induced the SVF cells to undergo adipogenic differentiation. After induction and differentiation, about 45% and 6% of mature adipocytes were RFP-positive in the asWAT ( Fig. 2A , B ) and ingWAT ( Fig. 2C, D ), indicating that Myf5-lineage SVF cells have the adipogenic differentiation potential in vitro. Together, these results suggest that a subpopulation of WAT adipocytes and SVF cells are derived from Myf5-lineage precursors.

Myf5-lineage adipocytes express lower levels of BAT and beige markers in subcutaneous WAT
Given that BAT adipocytes are derived from the Myf5 lineage ( 15 ), we hypothesized that Myf5-lineage adipocytes in WAT should express higher levels of BAT and beige signature genes. To test this hypothesis, we fi rst examined whether the relative abundance of Myf5-lineage adipocytes in different WAT depots affects the expression of BAT signature genes, including Ucp1 , Prdm16 , Cidea , and Ppargc1a . Contrary to what we expected, we found that (FITC) signal. Sorted cells were cultured and differentiated to examine their adipogenic and myogenic potential.

Total RNA extraction, cDNA synthesis, and real-time PCR
Total RNA extraction, cDNA synthesis and real-time PCR were performed as described ( 17,27 ). Briefl y, total RNA was extracted from cells using Trizol Reagent according to the manufacturer's instructions. RNA was treated with RNase-free DNase l to remove contaminating genomic DNA. The purity and concentration of total RNA were measured by a spectrophotometer (Nanodrop 3000, Thermo Fisher) at 260 nm and 280 nm. Ratios of absorption (260/280 nm) of all samples were between 1.8 and 2.0. Then 5 g of total RNA were reversed transcribed using random primers and MMLV-reverse transcriptase. Real-time PCR was carried out in a Roche Lightcycler 480 PCR System with SYBR Green Master Mix and gene-specifi c primers. Primer sequences are from published papers ( 12,17,27 ). Ct value of 18S rRNA was used as internal control and 2 Ϫ ⌬ ⌬ CT method was used to analyze the relative expression levels of varies genes.

Protein extraction and Western blot analysis
The protein extraction and Western blot were conducted as previously described ( 27 ). Briefl y, total protein was isolated from cells using RIPA buffer containing 50 mM Tris-HCl (pH 8.0), 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, and 0.1% SDS. Protein concentrations were determined using Pierce BCA Protein Assay Reagent (Pierce Biotechnology, Rockford, IL). Proteins were separated by SDS-PAGE, transferred to a polyvinylidene fl uoride (PVDF) membrane (Millipore Corporation, Billerica, MA), and incubated with the primary antibodies overnight. The Myog and GAPDH antibodies were from Santa Cruz Biotechnology (Santa Cruz, CA), and MF20 was from the Developmental Studies Hybridoma Bank (Developmental Studies Hybridoma Bank, Iowa City, IA). The secondary antibody (anti-rabbit IgG or anti-mouse IgG, Santa Cruz Biotechnology) was diluted 8,000-fold. Immunodetection was performed using enhanced chemiluminescence (ECL) Western blotting substrate (Pierce Biotechnology, Rockford, IL) and detected with a Gel Logic 2200 imaging system (Carestream).

Immunostaining and image acquisition
Immunostaining was performed as previously described ( 27 ). Cells or tissue sections were fi xed with 4% PFA, and the fi xed cells or sections were blocked with blocking buffer containing 5% goat serum, 2% BSA, 0.2% triton X-100, and 0.1% sodium azide in PBS for 1.5 h. Then the samples were incubated with primary antibodies diluted in blocking buffer overnight. After washing with PBS, the samples were incubated with secondary antibodies and DAPI for 45 min at room temperature. Fluorescent images were captured using a Leica DM 6000B fl uorescent microscope.

Transplantation
The SVF cells and the FACS-purifi ed Sca1 Ϫ RFP + cells from as-WAT of Myf5-Cre/Rosa26-tdTomato mice were mixed with 1 ng/ml bFGF and injected into tibialis anterior (TA) muscles of 6-to 8-week-old WT or mdx mice. After 5 or 21 days, the recipient mice were euthanized, and the TA muscles were collected and examined for RFP expression using a fl uorescent microscope and staining. The myoblasts from the skeletal muscles near the asWAT were also transplanted into WT mice as the control.

Data analysis
All experimental data are presented as means ± SEM. Comparisons were made by unpaired two-tailed Student t -test or was inversely correlated to the abundance of RFP-positive adipocytes in the asWAT and ingWAT ( Fig. 2G ).
The inverse correlation between the abundance of Myf5-lineage adipocytes and BAT/beige gene expression the abundance of RFP-positive adipocytes in the asWAT and ingWAT was inversely correlated to the mRNA levels of the BAT marker genes ( Fig. 2E, F ). Likewise, the expression level of beige adipocyte markers Tmem26 and Tbx1  analysis demonstrated that the adipogenic and myogenic activities were not shared by a common stem cell population but rather resided in distinct progenitor cell populations ( Fig. 3C-F ). We attempted to further examine the lineage origin of myogenic progenitors in WAT depots. We isolated SVF cells from the asWAT of Myf5-Cre/Ro-sa26-tdTomato mice and induced them to undergo myogenic differentiation. More than 95% of the differentiated myotubes were RFP-positive ( Fig. 4A , B ), indicating that they have been predominantly derived from the Myf5lineage cells.
To verify whether the Myf5-lineage cells were necessary for the myogenic potential of WAT SVF cells, we conducted lineage ablation. SVF cells cultured from asWAT of Myf5-Cre/Rosa26-iDTR mice were treated with DT to ablate the Myf5-lineage cells, then grown to confl uence and induced to undergo myogenic differentiation. Ablation of Myf5lineage SVF cells nearly abolished the emergence of myotubes ( Fig. 4C, D ). In the control groups, numerous myosin heavy-chain-expressing (MF20-positive) myotubes were readily detectable after differentiation ( Fig. 4E ). By contrast, DT treatment nearly eliminated the MF20-positive myotubes ( Fig. 4F ). The mRNA levels of myogenic markers, such as Myod , Myog , eMHC , Myf5 , Myf6 , Des , Cav3 , and Cadh15 were also signifi cantly decreased after DT treatment ( Fig. 4G ). Western blot confi rmed that the protein levels of sarcomeric myosin heavy chain and Myog were almost decreased to undetectable levels after DT treatment ( Fig. 4H ). Similar results were observed in ingWAT SVF cells after the Myf5-lineage ablation (supplementary Fig. IV, A-C). These results indicate that Myf5-lineage cells are essential for the myogenic differentiation of the WAT SVF cells.
We further confi rmed these results by FACS analysis. When RFP + Sca1 + , RFP + Sca1 Ϫ , RFP Ϫ Sca1 + , and RFP Ϫ Sca1 Ϫ cells were grown to confl uence, only the RFP + Sca1 Ϫ cells spontaneously formed small myotubes ( Fig. 5A -D ). Six days after serum withdrawal induced differentiation, long mytubes were readily detectable in the RFP + Sca1 Ϫ fraction, with a few short myotubes in the RFP + Sca1 + fraction and no myotubes in the two other fractions ( Fig. 5E-H ). Consistently, the expression levels of myogenic genes Myod , Myog , and eMHC were more than 100 times higher in the RFP + Sca1 Ϫ fraction than in the RFP Ϫ Sca1 + fraction ( Fig. 5I ). Together, these results demonstrate that the myogenic differentiation potential of the WAT SVF cells resides in the Myf5-lineage progenitors.

Myf5-lineage SVF cells effi ciently differentiate into muscles in vivo after transplantation
To examine the myogenic potential of Myf5-lineage SVF cells in vivo, we fi rst transplanted the RFP + Sca1 Ϫ SVF cells from the asWAT of Myf5-Cre/Rosa26-tdTomato mice to cardiotoxin (CTX)-injured TA muscle of WT mice. Five days after transplantation, we detected RFP-positive myofi bers in the recipient mice ( Fig. 6A -C ). Moreover, cultured Lin Ϫ RFP + Sca1 Ϫ SVF cells predominantly expressed MyoD and Pax7 ( Fig. 6D-I ), two well-established myogenic progenitor-cell markers ( 34,35 ). These results provided strong evidence that the Myf5-lineage SVF cells represent in different WAT depots may be due to depot-specifi c differences other than Myf5-lineage contribution. To address this possibility, we examined whether

Sca1
Ϫ ; all Lin Ϫ ) of SVF cells ( Fig. 3A , B ). We then induced them to undergo adipogenic differentiation and examined the expression of BAT/beige markers in the differentiated adipocytes. Only Sca1-positive cells had adipogenic potential, and the RFP Ϫ Sca1 + and RFP + Sca1 + cells had similar adipogenic potency ( Fig. 3C, D ), as is also evident from their identical expression levels of Adipoq and Leptin (Refer to Fig. 3G ). Interestingly, the RFP + Sca1 Ϫ cells gave rise to numerous myotubes after differentiation under adipogenic conditions ( Fig. 3F ), whereas the RFP Ϫ Sca1 Ϫ cells failed to form adipocytes or myotubes ( Fig. 3E ). Consistent with our earlier observations, the RFP + Sca1 + SVF gave rise to adipocytes expressing signifi cantly lower levels of BAT marker genes Ucp1 , Prdm16 , Cidea , and Ppargc1a compared with the RFP Ϫ Sca1 + descendant adipocytes ( Fig. 3G ). Additionally, the RFP + Sca1 + -descendant adipocytes expressed significantly lower levels of beige marker genes CD137 , Tmem26 , and Tbx1 compared with the RFP Ϫ Sca1 + -derived adipocytes ( Fig. 3H ). Similar BAT/beige marker expression patterns were observed in Myf5-lineage and non-Myf5-lineage adipocytes derived from ingWAT (supplementary Fig. III,  A-D). Together, these data provide compelling evidence that the Myf5-lineage adipocytes are less brown than the non-Myf5-lineage adipocytes within the same WAT depot.
To further confi rm these results, we established the Myf5-Cre/Rosa26-iDTR mouse model ( 17,33 ). In this model, Myf5-Cre induces the expression of DT receptor (DTR), which is normally not expressed by murine cells, and renders the Myf5-lineage cells sensitive to DT. Thus, DT treatment should selectively ablate all Myf5-lineage cells but not the non-Myf5-lineage cells. SVF cells cultured from Myf5-Cre/Rosa26-iDTR mice were treated with DT to ablate the Myf5-lineage cells, then grown to confl uence and induced to undergo adipogenic differentiation. Ablation of Myf5-lineage SVF cells did not affect the accumulation of lipids and the expression of the mature adipocyte markers Adipoq and Leptin , but it signifi cantly upregulated the mRNA levels of BAT markers Ucp1 , Prdm16 , Cidea , Ppargc1a , and Ppara ( Fig. 3I ). The beige cell markers Tmem26 and Tbx1 were also increased signifi cantly after ablation of Myf5-lineage SVF cells ( Fig. 3J ). Taken together, the interdepot comparison, intradepot analysis of Myf5-lineage and non-Myf5-lineage adipocytes and lineage ablation studies demonstrate that Myf5-lineage adipocytes express lower levels of brown and beige adipocyte markers than the non-Myf5-lineage adipocytes.

Myf5-lineage progenitors are necessary for the myogenic potential of SVF cells
Adipose-derived stem cells have the potential to differentiate into multiple lineages, including adipogenic, chondrogenic, and myogenic differentiations ( 23 ). Our FACS Ϫ myogenic progenitors, we transplanted the RFP + Sca1 Ϫ cells into mdx mice that lack the dystrophin gene, thus representing a model for Duchenne muscular dystrophy. We grafted freshly sorted RFP + Sca1 Ϫ cells into TA muscles of mdx mice (50,000 cells/mouse). After 21 days, dystrophin-positive fi bers were detected in the recipient mice, while there were nearly no dystropinpositive fi bers in the sham-operated contralateral TA muscles ( Fig. 6J-L ). These results demonstrate that the RFP + Sca1 Ϫ cells represent a myogenic progenitor population that can be used to treat muscular dystrophy diseases.

DISCUSSION
In this study, we dissected the phenotypic and potential functional differences of Myf5-lineage and non-Myf5-lineage adipocytes in WAT depots. We discovered that the Myf5-lineage adipocytes in WAT express lower levels of BAT and beige marker genes than do the non-Myf5-lineage adipocytes. In addition, we showed that the Myf5-lineage SVF cells in WAT contain two distinct subpopulations of adipogenic and myogenic progenitors, and we provided proof-of-principle evidence that the adipose-derived myogenic progenitors can be used to restore muscle function in an animal model of muscular dystrophy.
One important question in developmental biology is whether lineage origin underlies cell function. The analysis of Myf5-lineage origin provides mixed answers to this question. Our results clearly indicate that among the subcutaneous WAT depots and within single WAT depots, the answer is "Yes". Through lineage tracing and ablation, FACS, differentiation assay, and gene expression analysis, we provided compelling evidence that the Myf5-lineage and non-Myf5-lineage WAT adipocytes have distinct phenotype and function. On the other hand, our fi nding that the Myf5-lineage WAT adipocytes are less brown than the non-Myf5-lineage WAT adipocytes is inconsistent with the  kidney. The identifi cation of Myf5-lineage adipocytes in both subcutaneous and visceral WAT is consistent with the recent observation by Sanchez-Gurmaches and colleagues ( 16 ). Together, our results suggest that lineage origin does underscore cell function within the same tissue but that cells derived from the same lineage marker may have distinct function in different tissues/organs.
As brown adipocytes are derived from the Myf5-cell lineage ( 15 ), we hypothesized that Myf5-lineage adipocytes in WAT may represent the adaptive beige adipocytes that share physiological function and gene expression with BAT. However, our data show that the Myf5-lineage adipocytes in WAT adipocytes in BAT and WAT may have originated from two distinct populations of Myf5-expressing progenitors that are phenotypically and functionally unrelated. In addition, as Myf5 is expressed by several mesodermal and neural cell lineages during development ( 15,(36)(37)(38)(39)(40), it is possible that the Myf5-lineage cells in BAT and WAT are from spatially unrelated embryonic primordial structures. Using the highly sensitive Rosa26-TdTomato reporter that contains a strong CAG promoter knocked into the ubiquitously expressed Rosa26 gene locus ( 30 ), we found that Myf5-lineage cells contribute to the skeletal muscle, adipose, brain, heart, and spleen, but not to liver, lung, or Myf5-lineage SVF cells, as more than 95% of myotubes found in WAT SVF cultures arose from the Myf5 lineage. In addition, ablation of the Myf5-lineage SVF cells abolished the myogenic potential of WAT SVF cells. The in vitro and in vivo myogenic differentiation potential of adipose-derived cells has also been reported ( 25,26 ). However, what has been unclear is whether the multilineage differentiation potential of the so-called "adipose-derived stem cells" (ADSC) resides in a bona fi de multipotential stem-cell population or mixed subpopulations of unipotent progenitors. Our FACS analysis results showing that the Sca1-positive subpopulation was adipogenic and Sca1-negative subpopulation was myogenic support the notion that adipose SVF cells consist of mixed populations of unipotent progenitor cells with adipogenic, myogenic, and other lineage potentials.
Although it has been shown that myogenic cells can be derived from adipose tissues, whether the adipose-derived myogenic progenitors are equivalent to skeletal myoblasts is unclear. We have identifi ed several common features but also differences between adipose-and skeletal muscle-derived myogenic cells. The adipose-derived myogenic cells express common skeletal myoblast markers Pax7 and MyoD ( 34,35 ). Like the skeletal myoblasts, the adiposederived myogenic cells can effi ciently differentiate into myotubes in vitro and in vivo, and they can restore dystrophin expression in mdx mice. However, the adipose-derived (asWAT) myogenic cells differentiated exclusively into slow myofi bers, whereas the skeletal myoblasts isolated from the adjacent muscles differentiated exclusively into fast myofi bers. The slow myofi ber-specifi c differentiation of adipose-derived progenitors may represent an express lower levels of BAT and beige adipocyte marker genes. We draw this conclusion through multiple lines of evidence. First, the higher abundance of Myf5-lineage adipocytes in the asWAT than in ingWAT is inversely correlated to the relative expression levels of BAT/beige adipocyte marker genes. Second, Myf5-lineage and non-Myf5-lineage adipocytes isolated by FACS from the same WAT depot expressed different levels of BAT/beige maker genes, with higher expression levels in the non-Myf5-lineage adipocytes. Third, ablation of the Myf5-lineage adipocytes resulted in elevated expression of BAT/beige-specifi c genes. Our results are in line with the observation by Seale et al. that the adrenergic stimulation-induced adaptive beige adipocytes in the subcutaneous WAT are derived from non-Myf5 lineage ( 15 ). Although a previous study has shown that both Myf5-positive and Myf5-negative precursors in WAT respond to ␤ 3-adrenoreceptor stimulation ( 16 ), our current results suggest that the Myf5-lineage WAT adipocytes are perhaps less adaptive/recruitable, as they express lower levels of BAT and beige adipocyte-specifi c genes. Taken together, these data indicate that Myf5-lineage cells contribute less signifi cantly to beige adipocytes or browning of WAT.
We identifi ed two subpopulations of Myf5-lineage progenitors in WAT: the Sca1-positive adipogenic and Sca1negative myogenic progenitors. Our results are consistent with recent reports demonstrating Sca1 as a positive selection marker for adipose stem cells and a negative selection marker for muscle stem cells ( 31,32,41,42 ). The myogenic potential of Myf5-lineage SVF cells in WAT is not surprising given that Myf5-lineage precursors give rise to muscle tissues ( 40 ). Importantly, we show that the myogenic potential of adipose progenitors resides only in the advantage in therapeutic applications. It has been well established that slow myofi bers are more resistant to damage under stress conditions, such as muscular dystrophy and diabetes (43)(44)(45). In addition, slow myofi bers express high levels of the mitochondrial biogenesis-determinant genes Pgc1 ␣ and PPAR ␦ ; therefore, they are more sensitive to insulin-regulated glucose uptake and may provide beneficial metabolic effects (46)(47)(48)(49). Expression of CD34 or Sca1, two stem-cell markers, is associated with the myogenic potential of progenitor cells derived from bone marrow, blood, and embryonic vasculature (50)(51)(52)(53). In addition, expression of CD34 defi nes the majority of quiescent satellite cells in adult skeletal muscle ( 54 ). However, our results from the FACS-sorted cells show that the Sca1-positive SVF cells were predominantly adipogenic, with little or nearly no myogenic activity. Strikingly, the RFP + Sca1 Ϫ SVF cells form myotubes spontaneously, suggesting that the myogenic progenitors in WAT SVF cells are negative for Sca1. Consistent with our results, previous studies have shown that the ability to spontaneously form skeletal myotubes in vitro resides in the Sca1-negative SVF cells ( 25 ). Other groups have also shown that noncultured, adipose-derived, CD45-negative side-population cells were enriched with progenitors that gave rise to myofi bers in vivo ( 55 ). Our recent data further demonstrate that the myogenic progenitors in WAT SVF cells were from the non-aP2-lineage cells ( 27 ). Taken together, these results indicate that the Myf5-lineage Sca1-negative cells are the myogenic progenitors in WAT SVF and that these cells are distinct from the other myogenic cells derived from regenerating muscle, bone marrow, or other tissues that have been reported to express Sca1.
In conclusion, we demonstrate in this study that Myf5lineage progenitors in subcutaneous WAT give rise to both adipogenic and myogenic lineages. In addition, the Myf5lineage adipocytes express lower levels of BAT and beige cell markers than do the non-Myf5-lineage adipocytes. Furthermore, the Myf5-lineage myogenic progenitors in WAT can effi ciently differentiate into myofi bers in vitro and in vivo, but they are distinct from the myogenic progenitors from skeletal muscles in differentiation potential toward slow myofi bers. These results provide novel insights into the adipogenic and myogenic differentiation potential of Myf5-lineage cells in WAT. Such knowledge may be useful for treating metabolic and muscle diseases.