Type-1 pericytes participate in fibrous tissue deposition in aged skeletal muscle

CALL FOR PAPERS Cellular Mechanisms of Tissue Fibrosis

Type-1 pericytes participate in fibrous tissue deposition in aged skeletal muscle

Alexander Birbrair,1,2 Tan Zhang,1 Zhong-Min Wang,1 Maria Laura Messi,1 Akiva Mintz,3

and Osvaldo Delbono1,2

1Department of Internal Medicine-Gerontology, Wake Forest School of Medicine, Winston-Salem, North Carolina;2Neuroscience Program; Wake Forest School of Medicine, Winston-Salem, North Carolina; and 3Departments of Radiologyand Neurosurgery, Wake Forest School of Medicine, Winston-Salem, North Carolina

Submitted 7 June 2013; accepted in final form 24 September 2013

Birbrair A, Zhang T, Wang ZM, Messi ML, Mintz A, Delbono O.Type-1 pericytes participate in fibrous tissue deposition in aged skeletalmuscle. Am J Physiol Cell Physiol 305: C1098–C1113, 2013. Firstpublished September 15, 2013; doi:10.1152/ajpcell.00171.2013.—Inolder adults, changes in skeletal muscle composition are associatedwith increased fibrosis, loss of mass, and decreased force, which canlead to dependency, morbidity, and mortality. Understanding thebiological mechanisms responsible is essential to sustaining andimproving their quality of life. Compared with young mice, aged micetake longer to recover from muscle injury; their tissue fibrosis is moreextensive, and regenerated myofibers are smaller. Strong evidenceindicates that cells called pericytes, embedded in the basement mem-brane of capillaries, contribute to the satellite-cell pool and musclegrowth. In addition to their role in skeletal muscle repair, after tissuedamage, they detach from capillaries and migrate to the interstitialspace to participate in fibrosis formation. Here we distinguish twobona fide pericyte subtypes in the skeletal muscle interstitium, type-1(Nestin-GFP�/NG2-DsRed�) and type-2 (Nestin-GFP�/NG2-DsRed�), and characterize their heretofore unknown specific roles inthe aging environment. Our in vitro results show that type-1 andtype-2 pericytes are either fibrogenic or myogenic, respectively.Transplantation studies in young animals indicate that type-2 peri-cytes are myogenic, while type-1 pericytes remain in the interstitialspace. In older mice, however, the muscular regenerative capacity oftype-2 pericytes is limited, and type-1 pericytes produce collagen,contributing to fibrous tissue deposition. We conclude that in injuredmuscles from aging mice, the pericytes involved in skeletal musclerepair differ from those associated with scar formation.

pericytes; skeletal muscle; fibrous tissue; aging

MOST CHRONIC DISEASES ARE characterized by excessive fibroustissue deposition, altered tissue architecture, and organ dys-function (49, 76). Typically, fibrosis increases in all organswith aging (27, 43). Following skeletal muscle injury, stemcells proliferate, differentiate, and fuse into myoblasts withinjured myofibers during the repair phase (39). Connectivetissue accumulation in the skeletal muscle has been reportedwhen the injury involves the basal lamina (31). While theconnective tissue contributes to cytoarchitecture stability dur-ing the healing process, excessive fibrogenic cells proliferationand fibrous tissue deposition may interfere with muscle regen-eration, leading to incomplete functional recovery (69). Com-pared with young, aged mice take longer to recover from

muscle injury; their regenerated myofibers are smaller, andtissue fibrosis is significantly increased (13).

The biological process underlying fibrous tissue depositionwith aging is incompletely understood. Various cell types havebeen proposed to produce fibrous tissue in several organsincluding skeletal muscle (29), resident fibroblasts (3), bonemarrow-derived circulating fibrocytes (14), epithelial cells(52), endothelial cells (79), and more recently, pericytes (28).Besides their role in initiating a fibrogenic response to pathol-ogy in various organs, pericytes act as multipotent stem cells intissue repair (23). In the skeletal muscle, they contribute to thesatellite cell pool and muscle growth (26).

Based on markers and morphology, pericytes have beenidentified as a heterogeneous cell population (11). However,their diverse differentiation potential was not explored until wedemonstrated their heterogeneity in the skeletal muscle (7, 10).Whether a specific pericyte subtype contributes to skeletalmuscle fibrosis with aging is unknown.

Here, by using a Nestin-GFP/NG2-DsRed transgenicmouse, we show the presence of two bona fide pericyte sub-types, type-1 (Nestin-GFP�/NG2-DsRed�) and type-2 (Nestin-GFP�/NG2-DsRed�), in the skeletal muscle. These cells ex-press the pericyte markers platelet-derived growth factor re-ceptor-� (PDGFR-�) and CD146 and are associated withmicrovessels. Our in vitro studies show that type-2 but nottype-1 form muscle cells when exposed to myogenic differen-tiation medium. Additionally, we found that type-1 but nottype-2 pericytes are fibrogenic when exposed to transforminggrowth factor-� (TGF-�) in culture. Transplantation studiesindicate that type-2 pericytes contribute only to new muscleformation after injury but not to fibrogenesis tissue in the oldmouse. Of relevance is the fact that type-2 pericyte regenera-tive myogenic capacity varies depending whether they areinjected in either young or old animals, indicating that the hostenvironment is crucial for efficient muscle regeneration. More-over, type-1 pericytes do not form muscle in vivo but contrib-ute to increased muscle fibrous tissue in older mice.

We conclude that type-1 pericytes contribute to musclefibrous tissue formation with aging and envision the possibilityof using type-2 and type-1 pericyte as a cellular target toimprove skeletal muscle repair and reduce fibrosis in oldermammals, respectively.

MATERIALS AND METHODS

Animals. Nestin-GFP transgenic mice (our colony) were main-tained homozygous for the transgene on the C57BL/6 genetic back-ground (56). Aging Friend Virus B (FVB) mice (our colony) have

Address for reprint requests and other correspondence: O. Delbono, WakeForest School of Medicine, Dept. of Internal Medicine, Section on Gerontol-ogy and Geriatric Medicine, Medical Center Boulevard, Winston-Salem, NC27157 (e-mail: [email protected]).

Am J Physiol Cell Physiol 305: C1098–C1113, 2013.First published September 15, 2013; doi:10.1152/ajpcell.00171.2013.

0363-6143/13 Copyright © 2013 the American Physiological Society http://www.ajpcell.orgC1098

mailto:[email protected]

long been used as a model of aging skeletal muscle in our laboratory(9, 66); young (3–5 mo), middle-aged (11–14 mo), and old (22–26mo) mice were used. NG2-DsRed transgenic mice expressingDsRed-T1 under the control of the NG2 promoter (10) were pur-chased from the Jackson Laboratory (Bar Harbor, ME). �-Actin-DsRed transgenic mice expressing red fluorescent protein variantDsRed.MST under the control of the chicken �-actin promoter cou-pled with the cytomegalovirus (CMV) immediate-early enhancer (9)were purchased from the Jackson Laboratory. All tissues of �-actin-DsRed transgenic mice fluoresce red. Nestin-GFP mice were cross-bred with NG2-DsRed and �-actin-DsRed mice to generate, respec-tively, Nestin-GFP/NG2-DsRed and Nestin-GFP/�-actin-DsRed dou-ble-transgenic mice. All mice colonies were housed in a pathogen-freefacility of the Animal Research Program at Wake Forest School ofMedicine (WFSM) under 12:12-h light/dark cycle and fed ad libitum.Both male and female homozygous mice were used, and their agesranged from 3 to 5 mo. The WFSM Animal Care and Use Committeeapproved handling and procedures.

Primary antibodies. Table 1 shows the antibodies, their dilution,and source.

Immunohistochemistry. To detect DsRed and GFP fluorescence orDsRed fluorescence from 3-mo-old Nestin-GFP/NG2-DsRed mice orFVB mice of different ages injected with DsRed� pericytes, respec-tively, extensor digitorum longus (EDL) and tibialis anterior (TA)muscles were dissected; fixed in 4% paraformaldehyde (PFA) over-night; immersed in 10, 20, and 30% sucrose solutions for 60, 45, and30 min, respectively; embedded in OCT; and rapidly frozen in liquidnitrogen to prepare 10-�m thick cryosections. Muscle sections werefixed with 4% PFA for 30 min, then permeabilized in 0.5% TritonX-100 (Sigma, St. Louis, MO), and blocked to saturate nonspecificantigen sites using 5% (vol/vol) goat serum/PBS (Jackson Immunore-search Laboratories, West Grove, PA) overnight at 4°C. The next day,the sections were incubated with primary antibodies at room temper-ature for 4 h and visualized using appropriate species-specific sec-

ondary antibodies conjugated with Alexa Fluor 488, 568, or 680 at1:1,000 dilution (Invitrogen, Carlsbad, CA). Muscle sections werecounterstained with Hoechst 33342, mounted on slides using Fluo-rescent Mounting Medium (DakoCytomation, Carpinteria, CA), andexamined under fluorescence microscopy.

Fluorescence-activated cell sorting. Fluorescence-activated cellsorting (FACS) was carried out on a BD FACS (Aria Sorter, San José,CA) at 4°C and a pressure of 20 psi, using a laser at the 488-nm line,a 530/30 band-pass filter, a 100-�m sorting tip, and 34.2-kHz drivefrequency. The sorting apparatus was sterilized with 10% bleach. Thisinstrument allowed us to characterize cells by size as well as fluores-cence. Data acquisition and analyses were performed using BD FACSDiva 5.0.3 software, gated for a high level of GFP, DsRed, or APCexpression. The clear separation of GFP� from GFP� cells (7) andDsRed� from DsRed� cells as well as the low flow rate explains theease and accuracy of sorting (10). Sorted cells were reanalyzed toconfirm their fluorescence profile (7, 10).

Pericyte subtypes isolation from Nestin-GFP/NG2-DsRed miceskeletal muscle by FACS. Cells were sorted immediately from a poolof hindlimb muscles after skeletal muscle dissection and dissociation.Hindlimb muscles from young-adult (3–5 mo) Nestin-GFP/NG2-DsRed transgenic mice were prepared as described before (10).Briefly, muscles were carefully dissected away from the surroundingconnective tissue and minced, then digested by gentle agitation in0.2% (wt/vol) type-2 collagenase in Krebs solution at 37°C for 2 h,and dissociated by trituration and resuspension in 0.25% trypsin/0.05% EDTA in PBS for 15 min at 37°C. After centrifuging at 1,500rpm for 5 min, the supernatant was removed, and the pellet wasresuspended in growth medium. Aggregates were removed by passingthem through a 40-�m cell strainer before sorting. Cells were centri-fuged at 1,500 rpm for 5 min. The supernatant was removed, and thepellet was resuspended in 1% FBS in PBS and analyzed for GFP andDsRed fluorescence to sort the different cell populations based onthese two markers. The gate was set using cells isolated from C57BL6

Table 1. Antibodies, concentration, and source

Antibody Dilution Source Location

Rabbit anti-PDGFR-� 1:250 Dr. W. Stallcup Sanford-Burnham Medical Research Institute, CARat anti-mouse CD146 1:250 BioLegend (cat. no. 134702) San Diego, CARat anti-CD31 (PECAM-1) 1:100 BD Biosciences (cat. no. 558736) San Jose, CARabbit anti-type I collagen 1:1,000 AbD Serotec (cat. no. 2150–1410) Raleigh, NC,Mouse anti-MHC (MF 20) 1:2,000 Developmental Studies Hybridoma Bank,

University of IowaIowa City, IA

Rabbit anti-NG2 Chondroitin sulfate proteoglycan 1:100 Chemicon-Millipore (cat. no. AB5320) Temecula, CAMouse anti-Pax7 1:100 Developmental Studies Hybridoma Bank,

University of IowaIowa City, IA

Mouse anti-FSP1 1:100 Novus Biologicals (cat. no.H00006275-M01)

Littleton, CO

Mouse anti-fast myosin skeletal heavy chain 1:100 Abcam (cat. no. ab51263) Cambridge, MA

PDGFR-�, platelet-derived growth factor receptor-�; PECAM-1, platelet endothelial cell adhesion molecule-1; MHC, myosin heavy chain; FSP1,fibroblast-specific protein 1; NG2, neuron-glial antigen 2.

Table 2. Genes, GenBank accession numbers, coding regions, and primers

GeneGenBank Accession

Numbers Coding Regions Forward Primer and Positions Reverse Primer and Positions

CD146 NM_023061.2 CDS: 33–1797 AAGAGGAGAGCACCGATGAA (922–941) TTACTTTCTGCCTCGCAGGT (1147–1128)NG2 NM_139001.2 CDS: 87–7070 GCACGATGACTCTGAGACCA (3020–3039) AGCATCGCTGAAGGCTACAT (3242–3223)PDGFR-� NM_001146268.1 CDS: 430–3729 CCGGAACAAACACACCTTCT (2511–2530) TATCCATGTAGCCACCGTCA (2656–2637)Pax7 NM_011039.2 CDS: 58–1569 CATCCTTAGCAACCCGAGTG (1215–1234) AGTAGGCTTGTCCCGTTTCC (1567–1548)Myf5 NM_008656.5 CDS: 204–971 AGACGCCTGAAGAAGGTCAA (483–502) TGGAGAGAGGGAAGCTGTGT (899–880)Col1a1 NM_007742.3 CDS: 100–4461 CACCCTCAAGAGCCTGAGTC (3765–3784) GTTCGGGCTGATGTACCAGT (3998–4017)S100a4 (FSP1) NM_011311.2 CDS: 53–358 TTGTGTCCACCTTCCACAAA (87–106) GCTGTCCAAGTTGCTCATCA (225–244)Scx NM_198885.3 CDS: 165–788 TGGCCTCCAGCTACATTTCT (529–548) TGTCACGGTCTTTGCTCAAC (746–765)GAPDH NM_008084.2 CDS: 51–1052 GTGGCAAAGTGGAGATTGTTGCC (118–140) GATGATGACCCTTTTGGCTCC (407–387)

C1099ROLE OF PERICYTE SUBTYPES IN OLD SKELETAL MUSCLE

AJP-Cell Physiol • doi:10.1152/ajpcell.00171.2013 • www.ajpcell.org

C1100 ROLE OF PERICYTE SUBTYPES IN OLD SKELETAL MUSCLE


wild-type mice. Isolated type-1 (Nestin-GFP�/NG2-DsRed�) andtype-2 (Nestin-GFP�/NG2-DsRed�) pericytes were used to track thefate of those cells in vitro under myogenic or fibrogenic conditions.

Reverse transcription-polymerase chain reaction. To detect themRNA expression in cells, total RNA was isolated using TRIzolreagent (Life Technologies, Carlsbad, CA), RNA was dissolved insterile, RNase-free water (Invitrogen) and quantitated spectrophoto-metrically at 260 nm. Reverse transcription-polymerase chain reaction(RT-PCR) was performed in accordance with the manufacturer’sinstructions using the SuperScript III First-Strand synthesis system forRT-PCR system (Invitrogen). For each experiment, equivalentamounts of intact RNA (0.1–0.2 �g) were used. As negative controls,the RT reactions were performed in the absence of RNA (only water)or reverse transcriptase. The cDNA was amplified by PCR using theprimers included in Table 2. PCR Master Mix was purchased fromPromega (Madison, WI). Each PCR reaction contained 1� PromegaPCR Master Mix, 1 �M of each primer, and the cDNA of the cellsused in each case (Nestin-GFP�/NG2-DsRed�, Nestin-GFP�/NG2-DsRed�, and Nestin-GFP�/NG2-DsRed� cells). The volume of eachreaction was brought up to 50 �l with water. DNA amplification wascarried out as follows: denaturation at 94°C for 2 min, followed by 35cycles of 94°C for 1 min, 60°C for 1 min, and 72°C for 2 min. After35 cycles, the reactions were incubated at 72°C for 7 min to increasethe yield of amplification. PCR products were verified with DNA 2%agarose gel electrophoresis.

Myogenic induction in vitro. Freshly isolated pericyte subtypes(2.5 � 103 cells/cm2) were cultured on laminin-precoated plates(Invitrogen) for 2 days in growth medium [DMEM-high glucose(Invitrogen), supplemented with 2% L-glutamine, 50 U/ml penicillin,50 mg/ml streptomycin, and 10% (vol/vol) FBS (Invitrogen)], fol-lowed by 14 days in differentiation medium [DMEM (Invitrogen)containing 2-mM L-glutamine (Invitrogen) and 1% penicillin/strepto-mycin (Invitrogen), supplemented with 2% horse serum (Invitrogen)]in a 5% CO2 atmosphere (80, 81). Medium was changed every 3 daysuntil elongated, multinucleated myotubes appeared. After day 14 inculture, cells were fixed in 4% PFA at room temperature, and myosinheavy chain (MHC) expression was analyzed and quantified.

Fibrogenic induction in vitro. Fibrogenic differentiation was in-duced in fibrogenic medium for 5 days as described elsewhere (29).Briefly, freshly isolated pericyte subtypes were plated onto laminin-coated plates (Invitrogen) and cultured for 5 days in DMEM supple-mented with 2% horse serum 10% (vol/vol) (Invitrogen), with 2%L-glutamine, 50 U/ml penicillin, and 50 mg/ml streptomycin, supple-mented with 2.5 ng/ml of TGF-�1. Medium was changed twice. Afterday 5 in culture, cells were fixed in 4% PFA at room temperature, andtype I collagen expression was analyzed and quantified.

Immunocytochemistry. Cultured cells were fixed with 4% PFA for30 min, then permeabilized in 0.5% Triton X-100 (Sigma), andblocked to saturate nonspecific antigen sites using 5% (vol/vol) goatserum/PBS (Jackson Immunoresearch Laboratories) overnight at 4°C.

Fig. 1. Two pericyte subtypes are associated with skeletal muscle microvessels. A: histological analysis of pericyte subtypes in the skeletal muscle of doubletransgenic Nestin-GFP/NG2-DsRed mice. Whole lumbricalis muscle examined immediately after dissection. NG2-DsRed, Nestin-GFP, brightfield, and mergedimages are shown. NG2-DsRed� cells can be detected; some overlap with Nestin-GFP fluorescence. B: longitudinal section of EDL muscle fromNestin-GFP/NG2-DsRed mice illustrating the 2 pericyte subtypes, type-1 (Nestin-GFP�/NG2-DsREd�; white arrow) and type-2 (Nestin-GFP�/NG2-DsRed�;yellow arrow). A–C show the same muscle area for different channels [Nestin-GFP (N), NG2-DsRed (N), Hoechst (H), brightfield (BF), merged fluorescence,and merged fluorescence and brightfield images]. C: immunofluorescence on transverse sections from the same muscle shown in B, stained with anti-platelet-derived growth factor receptor-� (PDGFR-�), anti-CD146, and anti-CD31 antibodies, confirms the presence of 2 types of pericytes, both PDGFR-��/CD146� and associated with microvessels (CD31�). Panels show identical muscle areas from left to right: PDGFR-� (1st line), CD146 (2nd line), or CD31 (3rdline, orange), Nestin-GFP� (green), NG2-DsRed (red), Hoechst (blue), brightfield, and merged images. White arrows indicate type-1 pericytes (Nestin-GFP�/NG2-DsRed�), and the yellow arrows indicate type-2 (Nestin-GFP�/NG2-DsRed�).

Fig. 2. Satellite cells and fibroblasts do not express NG2-DsRed. Skeletal muscle from Nestin-GFP/NG2-DsRed transgenic mice shows Pax7 (satellite cellsmarker), fibroblast-specific protein 1 (FSP1; a fibroblast marker), Nestin-GFP, NG2-DsRed, and Hoechst staining in the same region. Brightfield and mergedimages are also shown. White arrow shows a satellite cell (Nestin-GFP�/Pax7�) that does not express NG2-DsRed; the yellow arrow indicates aNestin-GFP�/FSP1� but NG2-DsRed� fibroblast.



The next day, the cells were incubated with primary antibody at roomtemperature for 4 h and visualized using appropriate species-specificsecondary antibody conjugated with Alexa Fluor 680 at 1:1,000dilution (Invitrogen). They were counterstained with Hoechst 33342reagent at 1:2,000 dilution (Invitrogen) to label the DNA and mountedon slides for fluorescent microscopy with Fluorescent MountingMedium (DakoCytomation).

Isolation of type-1 and type-2 DsRed� pericytes. Hindlimb musclecells were isolated from young adult (3–5-mo-old) Nestin-GFP/�-actin-DsRed mice as described above (9). After being counted, cellswere centrifuged at 1,500 rpm for 5 min and resuspended in 100-�l1% FBS in PBS/106 cells. First, an aliquot was collected for use asunlabeled control (labeled with only the secondary APC anti-rabbit,without the primary anti-NG2 antibody) to set the gate. The remainingcells were incubated with the primary APC anti-mouse NG2 antibodyfor 45 min and washed in 1% FBS in PBS. They were then incubatedfor 30 min with APC anti-rabbit secondary antibody, washed in PBSwith 1% FBS, and run on a BD FACS flow cytometer (Aria Sorter).Sorting was done based on GFP and APC fluorescence. IsolatedNestin-GFP�/NG2-APC�/�-actin-DsRed� and Nestin-GFP�/NG2-APC�/�-actin-DsRed� cells were used in cell fate tracking experi-ments to evaluate muscle and fibrous tissue formation in vivo.

Muscle injury and cell transplantation. Skeletal muscle regenera-tion was studied in TA muscle injured by intramuscular injection ofbarium chloride (BaCl2) as described previously (32). Specifically,young, middle-aged, and old FVB mice were anesthetized withisoflurane/O2 inhalation. TA muscles were injected with 50 �l of1.2% BaCl2 dissolved in sterile PBS 1 day before cell transplantation.At 24 h postinjury, type-1 (Nestin-GFP�/NG2-APC�/�-actin-DsRed�) or type-2 (Nestin-GFP�/NG2-APC�/�-actin-DsRed�) peri-cytes were isolated from donor Nestin-GFP/�-actin-DsRed mice,resuspended in PBS (3 � 105 cells per TA), and slowly injected intothe damaged muscle of the acceptor mice. As controls, injured TAmuscles of all groups were injected with PBS. As our transplantationexperiments were performed in immune-competent (FVB) mice, im-munosuppression was induced in all mice to suppress immune rejec-tion of the transplanted cells as described before (6). Briefly, allanimals received daily subcutaneous injections of cyclosporine A (15mg/kg sc; Novartis, East Hanover, NJ) begining 2 days before oftransplantation and continuing for the duration of the experiment.Mice were killed 14 days postinjection, and TA muscles ere collected

and processed for immunohistochemistry as described above. Thearea and number of newly formed DsRed� myofibers were quantifiedin the muscle sections. Also, the area with collagen was quantified inold compared with young muscle sections. DsRed� cells positive totype I collagen, which were localized at the interstitial muscle spacewere counted in muscle sections.

Microscopy, cell imaging, and counting. An inverted motorizedfluorescent microscope (IX81; Olympus, Tokyo, Japan) with anOrca-R2 Hamamatsu CCD camera (Hamamatsu, Japan) was used forimage acquisition. Camera drive and acquisition were controlled by aMetaMorph Imaging System (Olympus). Ten arbitrary microscopicfields were counted in each immunostained plate or tissue section, andvalues were pooled from parallel duplicates per time point andindividual experiment.

Statistical analysis. Results are expressed as the means � SE.Statistical significance was assessed using Student’s t-test or one-wayANOVA followed by Holm-Sidak or Tukey post hoc tests usingSigmaPlot11 (Systat Software; San Jose, CA). P � 0.05 was consid-ered significant.

RESULTS

Two bona fide pericyte subpopulations associated to mi-crovessels are present in the skeletal muscle. Pericytes arecommonly recognized by their location and gene expressionpatterns rather than a precisely defined phenotype. Neuron-glial 2 chondroitin sulfate proteoglycan (NG2) has been used todetect them (62). We conducted a histological analysis ofskeletal muscle from Nestin-GFP/NG2-DsRed mice using Nes-tin and NG2 regulatory elements to control GFP and DsRedexpression, respectively. We discovered two types of pericytes;one expressed Nestin-GFP in the interstitium and the other didnot. We termed them type 1 (Nestin-GFP�/NG2-DsRed�) andtype 2 (Nestin-GFP�/NG2-DsRed�) (Fig. 1, A and B).

We confirmed that both types express two pericytic markersin addition to NG2 proteoglycan: a cell-surface tyrosine kinasereceptor (the platelet-derived growth factor receptor PDGFR-�/CD140; Ref. 28) and a melanoma-specific cell-adhesionmolecule, CD146, also known as MCAM, M-CAM, and

Fig. 3. Isolation of skeletal muscle pericyte subtypes from Nestin-GFP/NG2-DsRed mice by sorting. Representative flow cytometry dot plot showing GFP vs.DsRed fluorescence with the gate set using cells isolated from skeletal muscle of wild-type mice (A). B: Nestin-GFP/NG2-DsRed skeletal muscle-derived cellswere divided into 4 populations: Nestin-GFP�/NG2-DsRed-, Nestin-GFP�/NG2-DsRed� (type-1 pericytes), Nestin-GFP�/NG2-DsRed� (type-2 pericytes), andNestin-GFP�/NG2-DsRed- cells. C: single cells in culture dishes immediately after sorting. Localization of Nestin-GFP� (green) and NG2-DsRed� (red) cellsin freshly sorted fractions. All cells in the fraction of type-1 pericytes (Nestin�/NG2�) were Nestin-GFP negative and NG2-DsRed positive and Nestin-GFP andNG2-DsRed positive in the fraction of type-2 pericytes (Nestin�/NG2�).



MUC18 (23). While neither type expressed the endothelial cellmarker platelet endothelial cell adhesion molecule-1 (PECAM-1), a transmembrane glycoprotein also known as CD31 (23)(Fig. 1C), they surrounded CD31�-labeled blood vessels (or-ange); pericytes typically surround small vessels with endothe-lial cells (23).

Like pericytes, satellite cells express GFP in Nestin-GFPtransgenic mice (25) but not NG2-DsRed (Fig. 2) in Nestin-GFP/NG2-DsRed mice. As fibroblasts are also present in theskeletal muscle interstitium, we examined whether pericytesexpress fibroblast-specific protein 1 (FSP1; Ref. 42). Ourresults indicate that they do, but FSP1� fibroblasts do notexpress NG2-DsRed (Fig. 2).

Pericytes have been reported to have fibrogenic (28, 29) andmyogenic (23, 26) potential, although myogenic potential hasnot been reported for fibrogenic progenitors (29). Whetherdistinct pericyte subtypes contribute to either muscle repair orthe excessive fibrogenesis that occurs in older adults is un-known. Here, we propose that the muscle vasculature fromadult and aged mammals contains two types of pericytes, oneresponsible for muscle tissue repair, the other for fibrous tissueaccumulation with aging (28).

Type-1 pericytes are fibrogenic, while type-2 pericytes aremyogenic in vitro. To determine whether fibrogenic or myo-genic potential arises from the distinct, lineage-committedpericyte subtypes described above (10), we isolated type-1 andtype-2 pericytes from mononucleated cells dissociated fromNestin-GFP/NG2-DsRed mouse skeletal muscle by cell sorting(Fig. 3) and cultured them separately in myogenic (see Fig. 5,A–C) or fibrogenic (see Fig. 5, D–G) induction medium.

Our previous work show that, in contrast to type-2 pericytes,satellite cells do not express NG2 proteoglycan and are locatedbeneath the basal lamina (53, 77). To confirm the purity of ourpericyte subpopulations, we examined the expression of spe-cific markers in our isolated sorted cells by RT-PCR (Fig. 4, Aand B). All pericytes expressed CD146, NG2 and PDGFR-�mRNA but not the satellite cell markers Pax7 (48, 61, 65, 68)or Myf5 (5). Type I collagen mRNA was expressed in satellitecells (Nestin-GFP�/NG2-DsRed�) as reported (71), and type-1pericytes (Nestin-GFP�/NG2-DsRed�), but absent in type-2pericytes (Nestin-GFP�/NG2-DsRed�). Additionally, peri-cytes did not express the fibroblast markers FSP1(42) andScleraxis (22, 55).

To investigate myogenesis in vitro, we examined the expres-sion of MHC, a marker of myogenic lineage (63), in cellsexposed to myogenic differentiation medium for 14 days toallow myotube formation. We found that type-2 (Nestin-GFP�/NG2-DsRed�) but not type-1 (Nestin-GFP�/NG2-DsRed�) pericytes differentiated into myotubes (Fig. 5B). Thepercentage of nuclei per MHC� myotubes normalized to totalnuclei in type-1 and type-2 pericyte cultures was 0.34 � 0.10(n � 3 replicates, 500 nuclei counted) and 60 � 13% (n � 3replicates, 500 nuclei counted), respectively (Fig. 5C).

To examine the fibrogenic potential of the two skeletalmuscle pericyte subpopulations, we exposed them to fibrogenicmedium containing TGF-�1, a profibrotic cytokine of well-established potency that is overexpressed in most fibrotictissues (19). Type-1, but not type-2 pericytes, differentiatedinto FSP1� fibroblast-like cells (Fig. 5G) and expressed type Icollagen (Fig. 5E). The percentage of type-1 and type-2 peri-cyte-derived type I collagen� cells per total nuclei was 68 �

8.1 (n � 3 replicates, 200 cells counted) and 0.9 � 0.3% (n �3 replicates, 200 cells counted), respectively (Fig. 5F).

Type-2 pericytes are myogenic in vivo. As purified type-2pericytes demonstrated myogenic differentiation in vitro, wenext examined their myogenic potential in skeletal muscle.Immediately after cell sorting, we transplanted the pericytesubtypes freshly isolated from the muscles of Nestin-GFP/�-

Fig. 4. Nestin-GFP and NG2-DsRed cells gene expression. A: representativeRT-PCR agarose gel from 3 experiments showing CD146, NG2, PDGFR-�,Pax7, Myf5 expression, and control GAPDH. The pericyte markers CD146,NG2, and PDGFR-� were present in Nestin-GFP�/NG2-DsRed� and Nestin-GFP�/NG2-DsRed� cells but absent in Nestin-GFP�/NG2-DsRed- cells. Thesatellite cell markers Pax7 and Myf5 were detected only in Nestin-GFP�/NG2-DsRed- cells. B: RT-PCR agarose gel representative of 3 experiments, showingtype I collagen (Col1a1), FSP1, Scleraxis (Scx) expression, and controlGAPDH. Type I collagen gene was expressed in Nestin-GFP�/NG2-DsRed�

and Nestin-GFP�/NG2-DsRed- cells, but absent in Nestin-GFP�/NG2-DsRed� cells. The fibroblast markers were not present in Nestin-GFP�/NG2-DsRed�, Nestin-GFP�/NG2-DsRed�, or Nestin-GFP�/NG2-DsRed� cells.cDNA from whole skeletal muscle dissociated cells presorting (WSM) wasused as a positive control.



actin-DsRed transgenic (Tg) mice into the TA muscles ofwild-type mice of different ages (young, middle-aged, and old)(Fig. 6B) using BaCl2 injection (Fig. 6) to induce well-confinedmuscle degeneration without damaging the basal membrane(16). DsRed fluorescence (9) allowed us to identify donorfibers in the host muscle (Fig. 6A). The purity of these cellpopulations was confirmed by flow cytometric reanalysis asreported (10). After 2 wk, very few interstitial DsRed� cellswere observed in the muscles from young mice injected withtype-1 pericytes, and no DsRed� myofibers were detected (Fig.6C). In contrast, transplanted muscles from animals injectedwith type-2 pericytes showed many newly formed DsRed�

myofibers (Figs. 6D and 7A).Regenerative capacity of type-2 pericytes decreased when

transplanted into aged skeletal muscle. Aged microenviron-ments or niches inhibit the regenerative capacity of tissue stemcells (74). Whether they influence the differentiation potentialof pericyte subpopulations is unknown. To address this ques-tion, we injected young, middle-aged, and old mice with type-2pericytes and found that their myogenic regenerative capacitydecreased in the aged microenvironment; the same number ofcells formed fewer myofibers in old mice compared with theiryounger littermates. Quantitative analysis revealed that Nestin-GFP�/NG2�/�-actin-DsRed� cells participated in the forma-tion of 474 � 39 (n � 3 muscles), 316 � 23 (n � 3 muscles),and 79 � 30 (n � 3 muscles) DsRed� myofibers per mm2 inyoung, middle-aged, and old animals, respectively (Fig. 7, Aand B). Compared with young mice, middle-aged, and old miceshowed a 33 and 83% decrease in DsRed� cells, respectively.Additionally, the cross-sectional area of the newly formedfibers was smaller in old than young and middle-aged mice.The area of DsRed� myofibers was 2,123 � 211, 984 � 54,and 707 � 54 �m2 in young, middle-aged, and old animals,respectively (Fig. 7, A and C). Compared with young mice,middle-aged and old mice showed a 53 and 66% decrease inthe DsRed� area, respectively.

Skeletal muscle exhibits marked plasticity; myofiber typereadily converts in response to distinct stimuli (12, 34, 38, 40).During muscle formation, fast myosin is expressed before slowmyosin appears (41). To determine DsRed� myofiber type, weexamined their MHC isoform expression. Our results show thattype-2 pericytes fused to fast (type II) myofibers (Fig. 8A). Asthe adult TA muscle is composed of only fast fibers (36), whilesoleus muscles are a mixture of fast and slow fibers (30), weused the soleus muscle as a control for our antibody (Fig. 8B).

In addition to fusing to developing myofibers, skeletal mus-cle pericytes have been reported to enter the satellite-cellcompartment and express satellite-cell markers (26). Sincesatellite cells do not express NG2 proteoglycan, they did notcontaminate the pool of injected type-2 pericytes, and any

DsRed� satellite cells detected after transplantation had tooriginate from the pericyte subtype. Indeed, in examining theexpression of the satellite-cell marker Pax7 in DsRed� cellsderived from the transplanted type-2 pericytes, we did findDsRed�/Pax7� cells in TA muscles from young animals,although very rarely (0.92 � 0.92 DsRed�/Pax7� cells permm2), indicating that besides myofibers, type-2 pericytes maygenerate satellite cells (Fig. 7D).

Type-1 but not type-2 pericytes contribute to fibrous tissueformation in skeletal muscle from old mice. Because type-1pericytes showed fibrogenic differentiation potential in vitro,we next evaluated their in vivo capacity to form fibrous tissuein skeletal muscle. Fibrous tissue accumulation is observedonly when myogenic regeneration fails due to aging or disease(29) and not in response to injury in young healthy mice.Therefore, we transplanted type-1 and type-2 pericytes into theinjured skeletal muscle of old mice to analyze their fibroticpotential (54). Previous works have measured collagen accu-mulation in intact aging skeletal muscle using biochemical andhistochemical techniques (1, 33, 37, 47, 57).

We sorted the pericyte subtypes from �-actin-DsRed/Nestin-GFP mice using a rabbit anti-NG2 proteoglycan and APCanti-rabbit secondary antibodies. Since every cell expressesDsRed (10), even when transdifferentiated into another celltype, this technique allowed us to track isolated type-1 andtype-2 pericytes after injection (Fig. 9A).

Two weeks after cell transplantation, muscles were har-vested and stained for type I collagen (Fig. 9B). We observedfew DsRed� mononucleated cells in the interstitium of mus-cles injected with type-2 pericytes; almost no DsRed� cellsexpressing type I collagen (9.4 � 9.4 DsRed�/type I collagen�

cells/mm2; n � 3 muscles; Fig. 9, B and C); and newly formedDsRed� myofibers (Fig. 9B). In contrast, the interstitial con-nective tissue of muscles transplanted with type-1 pericyteswas rife with DsRed� cells expressing type I collagen (1,772 �228 DsRed�/type I collagen� cells per mm2: n � 3 muscles) andFSP1 (Fig. 9D) and had no regenerated DsRed� myofibers,indicating that type-1 pericytes are fibrogenic in vivo (Fig. 9, Band D).

DISCUSSION

Pericytes are found in vascularized tissues and recognizedby their histological location rather than a precisely definedphenotype. However, some antigens, such as neuron-glial 2chondroitin sulphate proteoglycan (NG2; Ref. 23), PDGFR-�(28), and a melanoma-specific cell-adhesion molecule CD146(23) have been identified as pericytic molecular markers. Wedistinguished two bona fide subpopulations of pericytes inskeletal muscle, types 1 and 2. They express these markers and

Fig. 5. Type-1 pericytes produce type I collagen, while type-2 pericytes differentiate into myotubes in vitro. Myogenic and fibrogenic induction of freshly isolatedpericyte subtypes from Nestin-GFP/NG2-DsRed mouse muscle. A: time frame of myogenic differentiation in vitro: freshly isolated pericytes (Fig. 3) werecultured for 2 wk in myogenic differentiation medium. B: after 14 days in differentiation medium, Nestin-GFP�/NG2-DsRed� and Nestin-GFP�/NG2-DsRed�

cells were stained with anti-myosin heavy chain (MHC) antibody. C: percentage of MHC� nuclei derived from each pericyte subtype was counted and normalizedto the total number of nuclei (n � 3). Data are means � SE. D: time frame for fibrogenic in vitro differentiation: freshly isolated pericytes (Fig. 3) were culturedfor 5 days in fibrogenic medium containing TGF-�. E: after 5 days in fibrogenic medium, Nestin-GFP�/NG2-DsRed� and Nestin-GFP�/NG2-DsRed� cells werestained with anti-type I collagen antibody. F: percentage of type I collagen� cells derived from each pericyte subpopulation was counted and normalized to thetotal cell number (n � 3 preparations). G: Nestin-GFP�/NG2-DsRed� and Nestin-GFP�/NG2-DsRed� cells were stained with FSP1 antibody at day 5 infibrogenic conditions.



Fig. 6. Skeletal muscle type-2 pericyte fate in vivo. A: pericytes subtypes from Nestin-GFP/�-actin-DsRed double-transgenic mice were isolated and sorted usingan anti-NG2 proteoglycan APC antibody. All cells are DsRed� and can be tracked in vivo. Representative dot plots showing GFP vs. APC fluorescence withthe gate set using unlabeled cells before and after labeling with NG2 APC antibody. B: transplantation of isolated type-2 DsRed� pericytes into injured skeletalmuscles. C: representative tibialis anterior (TA) muscle section from a young mouse showing transplanted type-1 pericytes (DsRed�) remaining in the interstitalspace. Panels show identical muscle areas. Nuclei were stained with Hoechst 33342. D: whole TA muscle visualized immediately after dissection from young,middle-aged, and old mice 2 wk after type-2 DsRed� pericytes transplantation. Brightfield, DsRed fluorescence, and merged images are shown. Notice thescarcity of DsRed� cells in the muscle from old animals.



Fig. 7. Type-2 pericytes participate in the formation of fewer and smaller myofibers in injured old mice. A: 2 wk after transplantation, DsRed� myofibers (red)are present in the muscle of mice injected with type-2 pericytes. Panels show identical muscle areas from top to bottom: DsRed fluorescence (first line), Hoechst(blue), brightfield, merged fluorescence, and merged fluorescence and brightfield images. B and C: quantification of the data illustrated in A. B: number ofDsRed�/type I collagen� cells normalized to the cross-sectional area at day 14 of cell transplantation into injured skeletal muscle from old mice (n � 3 muscles).For data analysis, we used one-way ANOVA followed by Holm-Sidak test. C: cross-sectional area of DsRed� myofibers in young, middle-aged, and old mice(n � 3 muscles). For data analysis, we used one-way ANOVA followed by Tukey test. D: representative TA muscle section from a transplanted mouse (as inA), showing a rare type-2 pericyte (DsRed�) expressing the satellite cell marker Pax7. Panels show identical muscle areas. Nuclei were stained with Hoechst33342. *P � 0.05.



surround the endothelial cells in the microvasculature. Al-though pericytic differentiation in several tissues supports theirmultipotency (23), whether pericyte subpopulations are differ-entially committed to specific lineages is not known.

We found that both in vitro and in vivo, type-2 (Nestin-GFP�/NG2-DsRed�) pericytes are myogenic, while type-1 arefibrogenic. Future studies targeting their specific capacitiescould establish a therapeutic role, either to enhance myogen-esis or to decrease fibrous tissue accumulation, in skeletalmuscle from aging mammals (Fig. 10A).

Type-1 pericytes as cellular targets for antifibrotic therapyin skeletal muscle. Pericytes have been found to contribute tofibrous tissue production in several organs, including muscle(28). Once activated, they escape from the microvessels intothe interstitium and participate in the formation of scar tissue,progression of fibrosis, and deterioration of organ function(28). However, they have also been associated with regenera-tive functions. In skeletal muscle, for example, they participatein vasculogenesis and contribute to the satellite-cell pool andmuscle growth (23, 24, 26). Thus removing all pericytes toprevent or reduce fibrosis would impede tissue repair afterinjury.

Here, we adduce data indicating that a distinct type ofpericyte directs each of these separate outcomes. Type 1contributes to fibrosis in aging skeletal muscle, while only type2 is myogenic. Instead of the whole pericyte population, futurestudies should target type 1 for antifibbrotic therapy in skeletalmuscle and, further, determine whether specific pericyte sub-populations vary in their function and regenerative capacity inother organs.

Age-dependent mechanisms may activate different pericytesubtypes. Fibrous tissue accumulates in the skeletal musclewith aging (75), but the mechanisms remain poorly understood(21). Some studies suggest that the decrease in, and reducedfunction of, stem cells play an essential role (58); others thatsignificant changes to the skeletal muscle pericyte microenvi-ronment impair regeneration (35). Some pericytes may notexpress a specific receptor that mediates the signaling pathwayrequired for their differentiation, resulting in the emergence ofa subpopulation with poor sensitivity to a specific agonist.Reduced expression of the notch-ligand delta affects notchsignaling to impair muscle regeneration (21). The TGF-�, Wnt,and IGF pathways have also been associated with age-depen-dent impairment of muscle regeneration (4, 18).

Fig. 9. Type-1 pericytes are fibrogenic in vivo in old mouse skeletal muscle. A: time frame for type-1 or type-2 DsRed� pericyte isolation and injection intoinjured skeletal muscles from old wild-type mice. B: TA muscle cross sections from old mice injected with type-1 or type-2 DsRed pericytes 2 wk aftertransplantation. Panels show identical muscle areas from left to right: DsRed, type I collagen, Hoechst, brightfield, merged fluorescence, and merged fluorescenceand brightfield images. Anti-type I collagen staining confirms that type-1 pericytes stay in the interstitium and do not differentiate into muscle cells; type-2pericytes do not express type I collagen and fuse in DsRed� myofibers. C: quantification of the data illustrated in B. Number of DsRed�/type I collagen� cellsnormalized to the cross-sectional area at day 14 of cell transplantation into injured skeletal muscle from old mice (n � 3 muscles). D: TA muscle cross-sectionsfrom old mice injected with type-1 DsRed� pericytes 2 wk after transplantation. All panels show the same muscle area for different channels (FSP1 staining,DsRed, Hoechst, brightfield, merged fluorescence, and fluorescence and brightfield merged images). Type-1 pericytes are located in the interstitium and expressthe fibroblast marker FSP1.

Fig. 8. DsRed� muscle fibers in muscles transplanted with DsRed� type-2 pericytes are fast (type II). A: all fibers express fast myosin heavy chain in TA muscleinjected with DsRed� type-2 pericytes. B: soleus muscle, used as a staining control, also shows some fibers expressing fast myosin heavy chain. A and B showthe same area for different channels: Fast MHC staining (orange), DsRed (red), Laminin (green), Hoechst (blue), brightfield, merged fluorescence images, andfluorescence and brightfield merge image; n � 3.



The constitutively active PDGFR--receptor knockin mouseexhibits fibrosis both systemically and in the skeletal muscle(60). Consequently, it could be used to determine whetheraging-impaired signaling through PDGFRs affects differentpericyte subtypes (2) and how extrinsic and intrinsic pericytechanges impair muscle regeneration. Cell-intrinsic changesmay be reversible or irreversible but in any case represent anadditional source of heterogeneity. One pericyte subtype maybe more prone to senescence or apoptosis, and the agedenvironment may select for a subtype with distinct regenerativepotential (20). Pericyte subtype characterization also suggestsfurther investigation of reported heterogenous stem cells (72)and their role in tissue regeneration in various organs. Futurestudies should determine whether the fate of the pericytechanges when they are exposed to such ligands as IGF-I,PDGF-AA, Wnt, and delta and whether their differentiationpotential remains unchanged after exposure to PDGFR--,TGF-�R-, IGFR-, frizzled-, or notch-Fc chimeric receptors,which compete with pericyte receptors for ligands in vitro.

Host microenvironment is critical for type-2 pericytemyogenicity. Total muscle size decreases over the lifespan(51), and old muscle regenerative capacity is severely impaired

(21), which can lead to disability, particularly in patients withother diseases or organ impairments. Note that when exposedto a younger microenvironment or experimental activation ofspecific signaling, old muscle can regain its regenerative ca-pacity (21).

Our data show that the myogenic regenerative capacity oftype-2 pericytes decreases when injected in older host animals:the same number generated fewer and smaller myofibers. Wesuggest that the deleterious age-muscle microenvironmentcould be modified to improve type-2 pericyte-dependent mus-cle regeneration.

Total muscle decrease with aging is associated with fewerfibers (51), but many studies also report myofiber atrophywith a substantial decrease in fiber size (73). Reducedregenerative potential of muscle stem cells (15) is one ofseveral mechanisms proposed to explain age-related muscleloss. Here, we show that the regenerative capacity of type-2pericytes is impaired in the old. To what extent this impair-ment leads to fiber loss or atrophy compared with the effectsof other myogenic stem cells in skeletal muscle has not beenestablished.

Fig. 10. Diagram of the role of pericyte subtypes in healing young and old skeletal muscle. Two subpopulations of pericytes are associated with blood vesselsin the skeletal muscle: type-1 (yellow) and type-2 (green). A: we propose that during muscle repair, type-2 pericytes together with satellite cells contribute tomyogenesis, producing larger myofibers in young than in old mice, while type-1 pericytes participate in the fibrous tissue accumulation observed in aged mice.B: classical model of the pericyte as a multipotent stem cell able to differentiate into fat, fibrous tissue, muscle, or neural cells. C: our model proposes thatpericytes are heterogeneous and multipotent, but their subtypes are oligopotent and their ability to differentiate more restricted.



In general, pericytes are multipotent, but their subpopula-tions exhibit lineage restriction. Strong evidence indicates thatpericytes are multipotent stem cells, able to form varioustissues. They have been shown to improve heart functionfollowing myocardial infarction in animal models (45) and toform myotubes with high efficiency (23, 26), and their osteo-genic differentiation and chondrogenic and adipogenic poten-tial are also confirmed (23). They accelerate wound healing(78) and contribute to fibrous tissue formation (28, 29). Theirrole in forming and stabilizing engineered blood vessels (24)suggests a use in vasculogenic therapy, and they can beinduced to develop into neural cells (44).

Pericytes are heterogeneous in location, origin, and mor-phology (82), and here we show that their molecular markerexpression along the skeletal muscle vasculature also varies.However, in contrast to the current concept of skeletal musclepericytes as multipotent stem cells with several differentiationpotentials (Fig. 10B), the two oligopotent capillary pericyteswe identified differ. Type-1 is responsible for ectopic adi-pocyte deposition (8) and fibrous tissue accumulation withaging, while type-2 accounts for regenerative capacity and,under special conditions, may be induced to differentiate intoneural lineage (10). This new knowledge will advance tissuerepair and regenerative medicine.

Limitations of this study. Our transplantation studies are thefirst to show that pericytes participate differentially in myo-genesis or fibrosis in vivo. However, their contribution relativeto other myogenic and fibrogenic cells in skeletal muscle,namely, satellite cells and tissue-resident fibroblasts, respec-tively, remains to be studied. Although lack of muscle regen-eration has been reported after satellite cell ablation (50, 54,67), their interaction with fibroblasts is required for efficientmuscle regeneration (59), and cross talk with other cell types,including pericytes, cannot be ruled out. Skeletal muscle fibro-blasts do not participate directly, contributing myonuclei, tomuscle regeneration. However, since their ablation alters theexpansion of satellite cells and impairs muscle regeneration(59), research is needed to determine whether ablating type-2pericytes, which do contribute myonuclei by fusion, can affectmuscle regeneration.

Endogenous pericytes contribute to growth of a significantpercentage of muscle fibers (26). Additionally, conditions thattrigger skeletal muscle regeneration, such as muscle injury,dramatically increase pericyte contribution to skeletal myogen-esis (26). Reports showing that selectively ablating satellitecells permanently hampers subsequent attempts to regenerateskeletal muscles (50, 59, 67) do not contradict our resultsbecause satellite cells may be required to induce other celltypes to adopt a myogenic fate, as suggested by Sambasivan etal. (67). The relative contribution of pericytes to muscle repairis unknown since they fuse with myofibers and may generatesatellite cells (26).

Multinucleated myofibers arise from the fusion of manycells with myogenic capacity in developing and regeneratingmuscle (17, 46, 64, 70). A recent study describes the ability ofpericytes resident in postnatal skeletal muscle to fuse withmyofibers and enter the satellite cell compartment (26). Here,we show that myofibers become DsRed� after injection withtype-2, but not type-1, pericytes because these pericytes fuse tothe myofibers during tissue repair. Whether one or many type-2pericytes fuse with individual damaged myofibers is unknown.

We also show that, in vitro, only type-2 pericytes form myo-tubes.

Quantifying the contribution of endogenous pericyte sub-types to myogenesis and fibrous tissue formation with aging isenabled by recombination-based lineage tracing, and ablating apericyte subpopulation would show whether satellite cell myo-genic capacity is affected by its absence. The only marker wefound differentially expressed in pericyte subpopulations isNestin-GFP, which is also expressed in satellite cells (25), sotracking pericyte fate or ablating a subtype in vivo will requirethe discovery of new markers expressed in a pericyte subpop-ulation but not other cells. Future work characterizing theexpression profiles of type-1 and type-2 pericytes will definethese novel markers.

ACKNOWLEDGMENTS

We thank Dr. G. N. Enikolopov from Cold Spring Harbor Laboratory (ColdSpring Harbor, NY) and Dr. W. Stallcup from the Sanford-Burnham MedicalResearch Institute (La Jolla, CA) for sharing with us the Nestin-GFP mouseand the rabbit anti-PDGFR-� antibody, respectively. Dr. James Wood, fromthe WFSM Comprehensive Cancer Center, contributed expertise on FlowCytometry to our project.

GRANTS

The present study was supported by a Wake Forest Pepper Center PilotProject and PUSH grant from the Wake Forest Comprehensive Cancer Center(to O. Delbono and A. Mintz), National Institute on Aging Grants AG-13934and AG-15820 (to O. Delbono), Wake Forest Claude D. Pepper OlderAmericans Independence Center (P30-AG21332), and a Glenn/AFAR Schol-arship for Research in the Biology of Aging (to A. Birbrair).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: A.B., A.M., and O.D. conception and design ofresearch; A.B., T.Z., Z.-M.W., and M.L.M. performed experiments; A.B. andM.L.M. analyzed data; A.B., A.M., and O.D. interpreted results of experi-ments; A.B. and T.Z. prepared figures; A.B. and O.D. drafted manuscript;A.B., A.M., and O.D. edited and revised manuscript; O.D. approved finalversion of manuscript.

REFERENCES

1. Alnaqeeb MA, Al Zaid NS, Goldspink G. Connective tissue changes andphysical properties of developing and ageing skeletal muscle. J Anat 139:677–689, 1984.

2. Andrae J, Gallini R, Betsholtz C. Role of platelet-derived growth factorsin physiology and medicine. Genes Dev 22: 1276–1312, 2008.

3. Barnes JL, Glass WF 2nd. Renal interstitial fibrosis: a critical evaluationof the origin of myofibroblasts. Contrib Nephrol 169: 73–93, 2011.

4. Barton-Davis ER, Shoturma DI, Musaro A, Rosenthal N, SweeneyHL. Viral mediated expression of insulin-like growth factor I blocks theaging-related loss of skeletal muscle function. Proc Natl Acad Sci USA 95:15603–15607, 1998.

5. Beauchamp JR, Heslop L, Yu DS, Tajbakhsh S, Kelly RG, Wernig A,Buckingham ME, Partridge TA, Zammit PS. Expression of CD34 andMyf5 defines the majority of quiescent adult skeletal muscle satellite cells.J Cell Biol 151: 1221–1234, 2000.

6. Biernaskie J, Sparling JS, Liu J, Shannon CP, Plemel JR, Xie Y,Miller FD, Tetzlaff W. Skin-derived precursors generate myelinatingSchwann cells that promote remyelination and functional recovery aftercontusion spinal cord injury. J Neurosci 27: 9545–9559, 2007.

7. Birbrair A, Wang ZM, Messi ML, Enikolopov GN, Delbono O.Nestin-GFP transgene reveals neural precursor cells in adult skeletalmuscle. PLoS One 6: e16816, 2011.

8. Birbrair A, Zhang T, Wang ZM, Messi ML, Enikolopov GN, Mintz A,Delbono O. Role of pericytes in skeletal muscle regeneration and fataccumulation. Stem Cells Dev 22: 2298–2314, 2013.



9. Birbrair A, Zhang T, Wang ZM, Messi ML, Enikolopov GN, Mintz A,Delbono O. Skeletal muscle neural progenitor cells exhibit properties ofNG2-glia. Exp Cell Res 319: 45–63, 2013.

10. Birbrair A, Zhang T, Wang ZM, Messi ML, Enikolopov GN, Mintz A,Delbono O. Skeletal muscle pericyte subtypes differ in their differentia-tion potential. Stem Cell Res 10: 67–84, 2013.

11. Bondjers C, He L, Takemoto M, Norlin J, Asker N, Hellstrom M,Lindahl P, Betsholtz C. Microarray analysis of blood microvessels fromPDGF-B and PDGF-Rbeta mutant mice identifies novel markers for brainpericytes. FASEB J 20: 1703–1705, 2006.

12. Booth FW, Thomason DB. Molecular and cellular adaptation of musclein response to exercise: perspectives of various models. Physiol Rev 71:541–585, 1991.

13. Brack AS, Conboy MJ, Roy S, Lee M, Kuo CJ, Keller C, Rando TA.Increased Wnt signaling during aging alters muscle stem cell fate andincreases fibrosis. Science 317: 807–810, 2007.

14. Bucala R, Spiegel LA, Chesney J, Hogan M, Cerami A. Circulatingfibrocytes define a new leukocyte subpopulation that mediates tissuerepair. Mol Med 1: 71–81, 1994.

15. Buford TW, Anton SD, Judge AR, Marzetti E, Wohlgemuth SE,Carter CS, Leeuwenburgh C, Pahor M, Manini TM. Models ofaccelerated sarcopenia: critical pieces for solving the puzzle of age-relatedmuscle atrophy. Ageing Res Rev 9: 369–383, 2010.

16. Caldwell CJ, Mattey DL, Weller RO. Role of the basement membranein the regeneration of skeletal muscle. Neuropathol Appl Neurobiol 16:225–238, 1990.

17. Capers CR. Multinucleation of skeletal muscle in vitro. J BiophysBiochem Cytol 7: 559–566, 1960.

18. Carlson ME, Conboy MJ, Hsu M, Barchas L, Jeong J, Agrawal A,Mikels AJ, Agrawal S, Schaffer DV, Conboy IM. Relative roles ofTGF-beta1 and Wnt in the systemic regulation and aging of satellite cellresponses. Aging Cell 8: 676–689, 2009.

19. Clouthier DE, Comerford SA, Hammer RE. Hepatic fibrosis, glomer-ulosclerosis, and a lipodystrophy-like syndrome in PEPCK-TGF-beta1transgenic mice. J Clin Invest 100: 2697–2713, 1997.

20. Collins CA, Zammit PS, Ruiz AP, Morgan JE, Partridge TA. Apopulation of myogenic stem cells that survives skeletal muscle aging.Stem Cells 25: 885–894, 2007.

21. Conboy IM, Conboy MJ, Wagers AJ, Girma ER, Weissman IL,Rando TA. Rejuvenation of aged progenitor cells by exposure to a youngsystemic environment. Nature 433: 760–764, 2005.

22. Covas DT, Panepucci RA, Fontes AM, Silva WA Jr, Orellana MD,Freitas MC, Neder L, Santos AR, Peres LC, Jamur MC, Zago MA.Multipotent mesenchymal stromal cells obtained from diverse humantissues share functional properties and gene-expression profile withCD146� perivascular cells and fibroblasts. Exp Hematol 36: 642–654,2008.

23. Crisan M, Yap S, Casteilla L, Chen CW, Corselli M, Park TS,Andriolo G, Sun B, Zheng B, Zhang L, Norotte C, Teng PN, Traas J,Schugar R, Deasy BM, Badylak S, Buhring HJ, Giacobino JP, LazzariL, Huard J, Peault B. A perivascular origin for mesenchymal stem cellsin multiple human organs. Cell Stem Cell 3: 301–313, 2008.

24. Dar A, Domev H, Ben-Yosef O, Tzukerman M, Zeevi-Levin N, NovakA, Germanguz I, Amit M, Itskovitz-Eldor J. Multipotent vasculogenicpericytes from human pluripotent stem cells promote recovery of murineischemic limb. Circulation 125: 87–99, 2012.

25. Day K, Shefer G, Richardson JB, Enikolopov G, Yablonka-Reuveni Z.Nestin-GFP reporter expression defines the quiescent state of skeletalmuscle satellite cells. Dev Biol 304: 246–259, 2007.

26. Dellavalle A, Maroli G, Covarello D, Azzoni E, Innocenzi A, Perani L,Antonini S, Sambasivan R, Brunelli S, Tajbakhsh S, Cossu G. Peri-cytes resident in postnatal skeletal muscle differentiate into muscle fibresand generate satellite cells. Nat Commun 2: 499, 2011.

27. Duffield JS. The elusive source of myofibroblasts: problem solved? NatMed 18: 1178–1180, 2012.

28. Duffield JS, Lupher M, Thannickal VJ, Wynn TA. Host responses intissue repair and fibrosis. Annu Rev Pathol 8: 241–276, 2013.

29. Dulauroy S, Di Carlo SE, Langa F, Eberl G, Peduto L. Lineage tracingand genetic ablation of ADAM12(�) perivascular cells identify a majorsource of profibrotic cells during acute tissue injury. Nat Med 18: 1262–1270, 2012.

30. Farnbach GC, Brown MJ, Barchi RL. A maturational defect in passivemembrane properties of dystrophic mouse muscle. Exp Neurol 62: 539–554, 1978.

31. Fechner G, Bajanowski T, Brinkmann B. Immunohistochemical alter-ations after muscle trauma. Int J Legal Med 105: 203–207, 1993.

32. Ge Y, Sun Y, Chen J. IGF-II is regulated by microRNA-125b in skeletalmyogenesis. J Cell Biol 192: 69–81, 2011.

33. Goldspink G, Fernandes K, Williams PE, Wells DJ. Age-relatedchanges in collagen gene expression in the muscles of mdx dystrophic andnormal mice. Neuromuscul Disord 4: 183–191, 1994.

34. Gollnick PD, Matoba H. Identification of fiber types in rat skeletalmuscle based on the sensitivity of myofibrillar actomyosin ATPase tocopper. Histochemistry 81: 379–383, 1984.

35. Gopinath SD, Rando TA. Stem cell review series: aging of the skeletalmuscle stem cell niche. Aging Cell 7: 590–598, 2008.

36. Hamalainen N, Pette D. The histochemical profiles of fast fiber types IIB,IID, and IIA in skeletal muscles of mouse, rat, and rabbit. Histochem Soc41: 733–743, 1993.

37. Haus JM, Carrithers JA, Trappe SW, Trappe TA. Collagen, cross-linking, and advanced glycation end products in aging human skeletalmuscle. J Appl Physiol 103: 2068–2076, 2007.

38. Hood DA. Invited Review: contractile activity-induced mitochondrialbiogenesis in skeletal muscle. J Appl Physiol 90: 1137–1157, 2001.

39. Hurme T, Kalimo H. Activation of myogenic precursor cells after muscleinjury. Med Sci Sports Exerc 24: 197–205, 1992.

40. Jarvis JC, Mokrusch T, Kwende MM, Sutherland H, Salmons S.Fast-to-slow transformation in stimulated rat muscle. Muscle Nerve 19:1469–1475, 1996.

41. Jerkovic R, Argentini C, Serrano-Sanchez A, Cordonnier C,Schiaffino S. Early myosin switching induced by nerve activity in regen-erating slow skeletal muscle. Cell Struct Funct 22: 147–153, 1997.

42. Joe AW, Yi L, Natarajan A, Le Grand F, So L, Wang J, Rudnicki MA,Rossi FM. Muscle injury activates resident fibro/adipogenic progenitorsthat facilitate myogenesis. Nat Cell Biol 12: 153–163, 2010.

43. Kapetanaki MG, Mora AL, Rojas M. Influence of age on wound healingand fibrosis. J Pathol 229: 310–322, 2013.

44. Karow M, Sanchez R, Schichor C, Masserdotti G, Ortega F, HeinrichC, Gascon S, Khan MA, Lie DC, Dellavalle A, Cossu G, GoldbrunnerR, Gotz M, Berninger B. Reprogramming of pericyte-derived cells of theadult human brain into induced neuronal cells. Cell Stem Cell 11: 471–476, 2012.

45. Katare R, Riu F, Mitchell K, Gubernator M, Campagnolo P, Cui Y,Fortunato O, Avolio E, Cesselli D, Beltrami AP, Angelini G, EmanueliC, Madeddu P. Transplantation of human pericyte progenitor cellsimproves the repair of infarcted heart through activation of an angiogenicprogram involving micro-RNA-132. Circ Res 109: 894–906, 2011.

46. Konigsberg IR, McElvain N, Tootle M, Herrmann H. The dissociabil-ity of deoxyribonucleic acid synthesis from the development of multi-nuclearity of muscle cells in culture. J Biophys Biochem Cytol 8: 333–343,1960.

47. Kragstrup TW, Kjaer M, Mackey AL. Structural, biochemical, cellular,and functional changes in skeletal muscle extracellular matrix with aging.Scand J Med Sci Sports 21: 749–757, 2011.

48. Kuang S, Charge SB, Seale P, Huh M, Rudnicki MA. Distinct roles forPax7 and Pax3 in adult regenerative myogenesis. J Cell Biol 172: 103–113, 2006.

49. Leask A. Signaling in fibrosis: targeting the TGF beta, endothelin-1 andCCN2 axis in scleroderma. Front Biosci 1: 115–122, 2009.

50. Lepper C, Partridge TA, Fan CM. An absolute requirement for Pax7-positive satellite cells in acute injury-induced skeletal muscle regenera-tion. Development 138: 3639–3646, 2011.

51. Lexell J, Taylor CC, Sjostrom M. What is the cause of the ageingatrophy? Total number, size and proportion of different fiber types studiedin whole vastus lateralis muscle from 15- to 83-year-old men. J Neurol Sci84: 275–294, 1988.

52. Markwald R, Eisenberg C, Eisenberg L, Trusk T, Sugi Y. Epithelial-mesenchymal transformations in early avian heart development. Acta Anat(Basel) 156: 173–186, 1996.

53. Mauro A. Satellite cell of skeletal muscle fibers. J Biophys Biochem Cytol9: 493–495, 1961.

54. McCarthy JJ, Mula J, Miyazaki M, Erfani R, Garrison K, FarooquiAB, Srikuea R, Lawson BA, Grimes B, Keller C, Van Zant G,Campbell KS, Esser KA, Dupont-Versteegden EE, Peterson CA.Effective fiber hypertrophy in satellite cell-depleted skeletal muscle.Development 138: 3657–3666, 2011.

55. Mendias CL, Gumucio JP, Davis ME, Bromley CW, Davis CS, BrooksSV. Transforming growth factor-beta induces skeletal muscle atrophy and



fibrosis through the induction of atrogin-1 and scleraxis. Muscle Nerve 45:55–59, 2012.

56. Mignone JL, Kukekov V, Chiang AS, Steindler D, Enikolopov G.Neural stem and progenitor cells in nestin-GFP transgenic mice. J CompNeurol 469: 311–324, 2004.

57. Mohan S, Radha E. Age-related changes in rat muscle collagen. Geron-tology 26: 61–67, 1980.

58. Morrison SJ, Spradling AC. Stem cells and niches: mechanisms thatpromote stem cell maintenance throughout life. Cell 132: 598–611, 2008.

59. Murphy MM, Lawson JA, Mathew SJ, Hutcheson DA, Kardon G.Satellite cells, connective tissue fibroblasts and their interactions arecrucial for muscle regeneration. Development 138: 3625–3637, 2011.

60. Olson LE, Soriano P. Increased PDGFRalpha activation disrupts con-nective tissue development and drives systemic fibrosis. Dev Cell 16:303–313, 2009.

61. Oustanina S, Hause G, Braun T. Pax7 directs postnatal renewal andpropagation of myogenic satellite cells but not their specification. EMBOJ 23: 3430–3439, 2004.

62. Ozerdem U, Grako KA, Dahlin-Huppe K, Monosov E, Stallcup WB.NG2 proteoglycan is expressed exclusively by mural cells during vascularmorphogenesis. Dev Dyn 222: 218–227, 2001.

63. Parlakian A, Gomaa I, Solly S, Arandel L, Mahale A, Born G,Marazzi G, Sassoon D. Skeletal muscle phenotypically converts andselectively inhibits metastatic cells in mice. PLoS One 5: e9299, 2010.

64. Pietsch P. Differentiation in regeneration. I. The development of muscleand cartilage following deplantation of regenerating limb blastemata ofAmblystoma larvae. Dev Biol 3: 255–264, 1961.

65. Relaix F, Montarras D, Zaffran S, Gayraud-Morel B, Rocancourt D,Tajbakhsh S, Mansouri A, Cumano A, Buckingham M. Pax3 and Pax7have distinct and overlapping functions in adult muscle progenitor cells. JCell Biol 172: 91–102, 2006.

66. Renganathan M, Messi ML, Delbono O. Overexpression of IGF-1exclusively in skeletal muscle prevents age-related decline in the numberof dihydropyridine receptors. J Biol Chem 273: 28845–28851, 1998.

67. Sambasivan R, Yao R, Kissenpfennig A, Van Wittenberghe L, PaldiA, Gayraud-Morel B, Guenou H, Malissen B, Tajbakhsh S, Galy A.Pax7-expressing satellite cells are indispensable for adult skeletal muscleregeneration. Development 138: 3647–3656, 2011.

68. Seale P, Sabourin LA, Girgis-Gabardo A, Mansouri A, Gruss P,Rudnicki MA. Pax7 is required for the specification of myogenic satellitecells. Cell 102: 777–786, 2000.

69. Stauber WT. Factors involved in strain-induced injury in skeletal musclesand outcomes of prolonged exposures. J Electromyogr Kinesiol 14: 61–70,2004.

70. Stockdale FE, Holtzer H. DNA synthesis and myogenesis. Exp Cell Res24: 508–520, 1961.

71. Urciuolo A, Quarta M, Morbidoni V, Gattazzo F, Molon S, GrumatiP, Montemurro F, Tedesco FS, Blaauw B, Cossu G, Vozzi G, RandoTA, Bonaldo P. Collagen VI regulates satellite cell self-renewal andmuscle regeneration. Nat Commun 4: 1964, 2013.

72. Van Keymeulen A, Rocha AS, Ousset M, Beck B, Bouvencourt G,Rock J, Sharma N, Dekoninck S, Blanpain C. Distinct stem cellscontribute to mammary gland development and maintenance. Nature 479:189–193, 2011.

73. Verdijk LB, Koopman R, Schaart G, Meijer K, Savelberg HH, vanLoon LJ. Satellite cell content is specifically reduced in type II skeletalmuscle fibers in the elderly. Am J Physiol Endocrinol Metab 292: E151–E157, 2007.

74. Villeda SA, Luo J, Mosher KI, Zou B, Britschgi M, Bieri G, Stan TM,Fainberg N, Ding Z, Eggel A, Lucin KM, Czirr E, Park JS, Couillard-Despres S, Aigner L, Li G, Peskind ER, Kaye JA, Quinn JF, GalaskoDR, Xie XS, Rando TA, Wyss-Coray T. The ageing systemic milieunegatively regulates neurogenesis and cognitive function. Nature 477:90–94, 2011.

75. Visser M, Goodpaster BH, Kritchevsky SB, Newman AB, Nevitt M,Rubin SM, Simonsick EM, Harris TB. Muscle mass, muscle strength,and muscle fat infiltration as predictors of incident mobility limitations inwell-functioning older persons. J Gerontol A Biol Sci Med Sci 60:324–333, 2005.

76. Wynn TA. Cellular and molecular mechanisms of fibrosis. J Pathol 214:199–210, 2008.

77. Yin H, Price F, Rudnicki MA. Satellite cells and the muscle stem cellniche. Physiol Rev 93: 23–67, 2013.

78. Zebardast N, Lickorish D, Davies JE. Human umbilical cord perivas-cular cells (HUCPVC): A mesenchymal cell source for dermal woundhealing. Organogenesis 6: 197–203, 2010.

79. Zeisberg EM, Tarnavski O, Zeisberg M, Dorfman AL, McMullenJR, Gustafsson E, Chandraker A, Yuan X, Pu WT, Roberts AB,Neilson EG, Sayegh MH, Izumo S, Kalluri R. Endothelial-to-mes-enchymal transition contributes to cardiac fibrosis. Nat Med 13: 952–961, 2007.

80. Zhang T, Birbrair A, Delbono O. Nonmyofilament-associated troponinT3 nuclear and nucleolar localization sequence and leucine zipper domainmediate muscle cell apoptosis. Cytoskeleton (Hoboken) 70: 134–147,2013.

81. Zhang T, Birbrair A, Wang ZM, Taylor J, Messi ML, Delbono O.Troponin T nuclear localization and its role in aging skeletal muscle. Age(Dordr) 35: 353–370, 2011.

82. Zimmermann KW. Der feinere bau der blutkapillaren. Z Anat Entwick-lungsgesch 68: 29–109, 1923.



Type-1 pericytes participate in fibrous tissue deposition in aged skeletal muscle

Documents

Transcript of Type-1 pericytes participate in fibrous tissue deposition in aged skeletal muscle