Msc

A Population of Cells Resident Within Embryonic and Newborn Rat Skeletal Muscle Is Capable of Differentiating Into Multiple Mesodermal Phenotypes

Paul A. Lucas, Ph.D. Department of Surgery, Mercer University School of Medicine at the Medical Center of Central Georgia, Macon, GA 31201 Andrew F. Calcutt, M.D. Department of Surgery, Mercer University School of Medicine at the Medical Center of Central Georgia, Macon, GA 31201 Sheila S. Southerland, B.A. Department of Surgery, Mercer University School of Medicine at the Medical Center of Central Georgia, Macon, GA 31201 J. Alan Wilson, M.D. Department of Surgery, Mercer University School of Medicine at the Medical Center of Central Georgia, Macon, GA 31201 Richard L. Harvey, M.D. Department of Surgery, Mercer University School of Medicine at the Medical Center of Central Georgia, Macon, GA 31201 Debra Warejcka, Ph.D. Department of Surgery, Mercer University School of Medicine at the Medical Center of Central Georgia, Macon, GA 31201 Henry E. Young, Ph.D. Division of Basic Medical Sciences, Mercer University School of Medicine, Macon, GA 31207 Work should be attributed to: Department of Surgery, Mercer University School of Medicine at the Medical Center of Central Georgia, Macon, GA 31201 Correspondence and reprint requests should be addressed to: Paul A. Lucas, Ph.D. Department of Surgery, Hospital Box 140, Medical Center of Central Georgia, 777 Hemlock Street, Macon, GA 31201 Phone: 912-752-2093 Fax: 912-752-2047 Running title: Mesenchymal stem cells in rat tissue

1

ABSTRACT We have demonstrated a population of putative mesenchymal stem cells in the connective tissue surrounding embryonic avian skeletal muscle. These cells differentiate into at least 5 recognizable phenotypes in culture: fibroblasts, chondrocytes, myotubes, osteoblasts, and adipocytes. We have isolated a similar population of cells from fetal and newborn rat skeletal muscle. Cells from rat leg muscle were dissected, minced, then enzymatically digested with a collagenase-dispase solution. The dissociated cells were plated and allowed to differentiate into two recognizable populations: myotubes and stellate mononucleated cells. The cells were then trypsinized, filtered through a 20 µm filter to remove the myotubes, frozen at -80°C, then thawed and replated. In culture the cells maintained their stellate morphology. However, under treatment with dexamethasone, a non-specific differentiating agent, 7 morphologies emerged: cells with refractile vesicles that stained with Sudan black B (adipocytes), multinucleated cells that spontaneously contracted in culture and stained with an antibody to myosin (myotubes), round cells whose extracellular matrix stained with Alcian blue, pH 1.0 (chondrocytes), polygonal cells whose extracellular matrix stained with Von Kossa's stain (osteoblasts), cells with filaments that stained with an antibody to smooth muscle a-actin (smooth muscle cells), cells that incorporated acetylated-low density lipoprotein (endothelial cells), and spindle-shaped cells that grew in a swirl pattern (fibroblasts). The initial population is tentatively classified as putative mesenchymal stem cells. The presence of these cells point to the existence of stem cells in the post-embryonic mammal that could provide a basis for tissue regeneration as opposed to scar tissue formation during wound healing. Key words: wound healing, mesenchyme, stem cells, dexamethasone

2

INTRODUCTION: Skeletal muscle has a limited capacity for repair in response to injury. Understanding this repair involves knowledge of both the cells involved and the local factors that influence the cells. Myosatellite cells are mononucleated, quiescent, myogenically committed stem cells located between the muscle sarcolemma and the overlying basement membrane (1,2). These cells are thought to be involved in muscle maintenance and provide the stem cells for muscle repair. Following focal injury to muscle fibers, these cells proliferate and migrate, within limits, to the site of injury and fuse to form new myotubes. Muscle regeneration studies involving muscle mincing, muscle grafting, and focal injury to the myofibers by temperature extremes, ischemia, anesthetics, and snake venom have all noted formation of regenerating myofibers within the connective tissue scar immediately adjacent to intact myofibers (3-8). However, a localized migration of satellite cells does not explain the results of other studies demonstrating the formation of myofibers embedded within the connective tissue scar at some distance from intact myofibers (9,10). Satellite cells also do not explain the numerous studies showing the de novo induction, using demineralized bone matrix or proteins derived from it, of cartilage and bone in muscle tissue (11-14). Indeed, minced muscle explants grown on demineralized bone matrix or with media containing soluble bone proteins demonstrate the differentiation of cartilage (15,16). These studies imply the existence of a more primitive stem cell located within skeletal muscle that is capable of differentiating into skeletal muscle, cartilage, and bone. A population of cells from day 11 embryonic chick leg muscle capable of differentiating into cartilage, skeletal muscle, bone, fat, and connective tissue has been reported (17). The current study was designed to determine the existence of a similar population of cells in rat skeletal muscle.

3

METHODS AND MATERIALS: Cell culture: The use of animals in this study was in compliance with the guidelines of Mercer University and the National Research Council's criteria for humane care as outlined in "Guide for the Care and Use of Laboratory Animals" prepared by the Institute of Laboratory Animal Resources and published by the National Institutes of Health (NIH Publication No. 85-23, Revised 1985). Leg muscle was harvested from day 19 rat fetuses. Rat dams at 19 days of gestation were euthanized and the fetuses removed under sterile conditions. The fetuses were placed in sterile 100 mm culture dishes (Falcon, Fisher, Norcross, GA) where the skin of the legs was carefully dissected. The muscle of the limbs was carefully removed, care being taken not to dissect any tendon, ligament, major nerve, major blood vessel, or bone. The muscles removed included the gluteal group, the quadratus, the gastrocnemius, and the soleus. Muscle pieces removed from the fetuses were placed in a separate 100 mm culture dish in 10 ml of Eagle's Minimal Essential Media with Earle's salts (EMEM) (GIBCO, Grand Island, NY) supplemented with 10% preselected horse serum (Sigma Chemical Co., St. Louis MO, lot #17F-0082). The muscle tissue was then carefully minced using sterile curved scissors. The solution was transferred to a sterile 50 ml conical centrifuge tube and centrifuged at 50 x g for 20 minutes. The supernatant was discarded and an estimate made of the pellet volume. The cell pellet was resuspended in 7 volumes of EMEM and 2 volumes collagenase/dispase solution to enzymatically release the cells (17). The collagenase solution consisted of 37,500 units of collagenase (CLS-I Worthington Biochemicals, Freehold, NJ) in 50 ml EMEM added to 100 ml dispase solution (Collaborative Research, Bedford, MA). The final concentrations were 250 units/ml collagenase and 33.3 units/ml dispase. The tissue suspension was transferred to a sterile 100 ml media bottle containing a magnetic stir bar stirred at 37° C for 1 hour; until the tissue was digested. The suspension was then transferred to centrifuge tubes and centrifuged at 300 x g for 20 minutes. The supernatant was discarded and the cells resuspended in 20 ml of EMEM with 10% horse serum. The cells were filtered through a 20 µm Nitex filter to obtain a single cell suspension, centrifuged at 150 x g for 10 minutes, the supernatant discarded, and the pellet re-suspended in 10 ml of EMEM plus 10% horse serum. The cells were counted on a hemocytometer and plated at 100,000 cells per 100 mm culture dish. The dishes had been precoated with 1% bovine gelatin (EM Sciences, Cherry Hills NJ) The procedure for obtaining newborn rat muscle was identical to that described above except that one day old rats were used. The rats were again euthanized using C02 inhalation. The rats were soaked in 70% ethanol for 2 minutes, then brought to the sterile hood and placed on sterile paper towels. At this time the hind quarters and back legs were skinned and the thigh, buttock, and leg muscle were removed, taking only the thickest part of the muscles. Again, care was taken so that no tendons, major nerves, or major blood vessels were included in the muscle prep. At this point the tissue was treated identical to the embryonic cells discussed above. Once the cells isolated from the tissue were plated, they were fed EMEM + 10% horse serum every other day. The cells had grown to confluence after approximately 8 days in culture and were found to consist of multinucleated myotubes and a large population of mononucleated cells. The cultures were released from the dish with 0.025% trypsin in Dulbecco's Phosphate Buffered Saline (DPBS) with 0.01% ethylenediaminetetraacetic acid (EDTA) and filtered 4

through a 20 µm filter. This filtration removed the myotubes leaving only the population of mononucleated cells. These cells were then frozen in aliquots of 1 ml containing 106 cells in EMEM + 10% horse serum and 7.5% DMSO (Sigma). Cryopreservation was performed in freezing chambers (Fisher Scientific, Norcross, GA) to slow freeze the cells to -80°C (18). After being frozen for at least 24 hours, aliquots of the frozen cells were thawed and plated at a density of 20,000 cells per 16 mm well in 24-well gelatin-coated culture plates (Corning Glass Works, Corning, NY) in EMEM + 10% horse serum. These cells were designated as secondary cultures. Some wells were maintained in the same media to allow for a control group, while the experimental wells, beginning on day 1 in culture, were treated with the media supplemented with dexamethasone (Sigma) at concentrations ranging from 10-10M to 106M for up to 5 weeks (19). At one week intervals during culture, cultures were fixed and assayed for phenotypes as described below. A separate series of experiments were performed using cells isolated from newborns to determine the number of of doublings the cells could undergo while retaining differentiation capabilities. Cells were thawed plated into 100 mm culture dishes as described above. The cultures were fed EMEM + 10% horse serum and allowed to grow to confluence. Once they had reached confluence, they were released from the dishes with trypsin as described above. Cell number averaged 3.0 to 3.5 x 106 cells per dish. Some cells were replated and tested with dexamethasone as described above while the remainder were refrozen (second passage). The frozen 2nd passage cells were then thawed. Some of the cells were plated at 20,000 cells/well in a 24 well gelatin-coated culture plate and treated with dexamethasone as described above. The remaining cells were plated in gelatin-coated 100 mm culture dishes and maintained until they reached confluence. These cells were trypsinized and frozen (3rd passage). This procedure was repeated through 17 passages. Assays for Phenotypes: 1.Mineralized Tissue. The presence of calcified tissue was assayed by Von Kossa's staining of calcium phosphate essentially as described by Humason (20). Briefly the culture medium was removed and the plates rinsed twice with DPBS. The cells were fixed with 0.5 ml of 10% formalin (Sigma) for 3 to 5 minutes, then rinsed four times with distilled water. Then 0.5 ml of freshly prepared 2% silver nitrate (Sigma) solution was added and the cells were incubated in the dark for ten minutes. Following incubation, the silver nitrate solution was removed and the cells rinsed five times with distilled water. Approximately 0.5 ml of distilled water was left on each well. The plate was exposed to bright light for 15 minutes with a white background underneath it to reflect light. The plates were again rinsed five times with distilled water and then dehydrated quickly with 100% ethanol. The plates were made permanent with glycerine jelly (17). Confirmation of the presence of calcium phosphate was performed by pretreating selected cultures with 1% w/v [ethylene bis (oxyethylenenitrilo)]-tetraacetic acid (EGTA) (Sigma), a specific calcium chelator, in Ca2+, Mg2+-free buffer for 1 hr prior to incubation in the silver nitrate solution (20). 2. Cartilage. Cultures were stained with Alcian blue (Roboz Surgical Instrument, Rockville, MD), pH 1.0. The fixed wells were stained with 0.5 ml Alcian blue, pH 1.0, for 30 minutes, then removed from the wells. Unbound stain was removed by rinsing the wells seven times with tap water or distilled water. The cultures were preserved under glycerine jelly. 3. Fat. Sudan black B (Asbey Surgical Co., Washington, DC) staining for saturated neutral lipid (20) was performed in the following manner: All media was aspirated from the

5

culture wells and each well was washed twice with one ml of DPBS. Then 0.5 ml of 70% ETOH was added to break cell membranes. After one minute, the alcohol was aspirated and the wells washed twice with DPBS. The cells were then incubated twice for 5 minutes in 100% propylene. Next, the cells were incubated twice for 10 minutes with 0.5 ml of Sudan black B per well. Stain differentiation was performed by rinsing the cells repeatedly with 0.5 ml of each of the following solutions until each solution was clear: Propylene: Water 90:10, 85:15, and 70:30. The cells were washed twice for one minute using distilled water, then made permanent with glycerine jelly. 4. Muscle. The cells were stained with the MF-20 antibody to skeletal muscle myosin (Hybridoma Bank, Ames, IA) using a modified procedure of Young et al (21). Each step is preceded by 2 rinses with DPBS unless noted. After another rinse, 0.5 ml of cold methanol (20°C) was applied for 5 minutes to fix the cells. This was followed by a 5 minute incubation with 0.5 ml of 1% v/v Triton-X100/0.05% w/v sodium azide in DPBS to solubilize cell membranes and inhibit endogenous peroxidases, respectively. A primary blocker of 20% goat serum was applied for 30 minutes in a 37°C incubator. The primary IgG of 1:200 dilution of MF-20 (0.4 ml/well) was then incubated for 1 hour. A secondary blocker of 0.5 ml of 20% goat serum was applied for 30 min and was followed by 0.4 ml of 1:7500 dilution of biotinylated goat anti-mouse IgG (Leinco, St. Louis, MO), also incubated for 30 minutes at 37°C. A tertiary blocker, consisting of 20% goat serum, was applied for 30 min and removed, then 0.4 ml of 1:3750 dilution of Streptavidin-horseradish peroxidase (Leinco) was added and incubated at 37°C for 30 minutes. At this point the cells were rinsed and 0.5 ml of ABTS-peroxidase substrate (Kirkegaard and Perry Labs, Gaithersburg, MD) was added for 30 minutes incubation at ambient temperature in the dark. After incubation, 200 µl of ABTS solution was removed from the cells and placed in a well of a 96-well ELISA plate (Falcon) containing 10 µl of 0.03% sodium azide. The ELISA plate was read on a Titer Tek spectrophotometric plate reader using a 405 nm filter. After the aliquot of ABTS solution had been removed, the cells were rinsed twice with 0.5 ml DPBS, then twice with 0.5 ml distilled water. Chromagen (Sigma) was added as per the instructions in the staining kit to selected wells for future photography. Once the color developed, 25 µl of 0.05% sodium azide was added per well to stop the reaction. The wells were then rinsed and made permanent with glycerine jelly. The ABTS was removed from the remaining wells and DNA content analyzed using the in situ diaminobenzoic acid (DABA) procedure of Johnson-Wint and Hollis (21) as previously described (22). Thus, the absorbance for the myosin content and the DNA content were obtained on the same wells. 6. Smooth Muscle. Smooth muscle was assayed by staining with an antibody to smooth muscle a-actin using a kit from Sigma. 7. Endothelial Cells. Endothelial cells were identified by their ability to take up low density lipoprotein as described by Voyta et al. (24). Cells were washed 5 times with Dulbecco's Minimal Essential Medium (high glucose) (DMEM) (GIBCO) supplemented with antibiotics. The cells were incubated for 4 hr. at 37°C with 10 µg per ml of 1,1'-dioctadecyl-3,3,3',3'tetramethyl-indocarbocyanine perchlorate (DiI-Acyl-LDL) (Biomedical Technology, Stoughton, MA). The wells were then washed 6 times with EMEM + 10% horse serum and viewed on a Nikon Diaphot with fluorescent attachment.

6

RESULTS On day one in culture, the primary cultures of the mononucleated rat cells isolated from newborn animals consisted of essentially two distinct morphologies: stellate shaped cells and bipolar cells (Figure 1A). By day three in culture, the bipolar cells had begun fusing to form small multinucleated myotubes, while the stellate shaped cells had proliferated but remained essentially unchanged in morphology (Figure 1B). The small myotubes had become large, multinucleated, branched, spontaneously contracting myotubes by the 8th day of culture. Meanwhile, the stellate cells continued to have no change in morphology and were easily located among the myotubes (Figure 1C). After trypsin release, filtration, freezing, and replating, the stellate cells alone were present in secondary culture and there were no myotubes. Control cultures (no dexamethasone) continued to have cells with a stellate morphology throughout the culture period (Figure 2A). From approximately two weeks, cultures treated with 10-8 through 10-6 M dexamethasone contained cells with large vesicles of varying sizes which were refractile in appearance under phase contrast microscopy (Fig. 2B). These cells stained with Sudan black B stain, indicating the presence of saturated neutral lipids, and have thus been identified as adipocytes (Fig. 2C). These cells did not stain with antibodies to myosin or smooth muscle a-actin. Multinucleated cells that spontaneously contracted in culture also appeared between one and two weeks in culture (Fig. 3A,B). The multinucleated cells stained with an antibody to myosin, confirming their identity as myotubes (Fig. 3C). When the myogenic response was quantified, all concentrations of dexamethasone exhibited myosin levels statistically significant (p<0.05) above controls, with the maximal response at 10-7M dexamethasone (Fig. 4). This assay also demonstrated that dexamethasone did not act as a general mitogen in this system; DNA content per well was constant across the treatment range (Fig. 4). After three weeks in culture small collections of very rounded cells, all of similar size, with a refractile extracellular matrix appeared in the wells treated with 10-9 to 10-6 M dexamethasone. These aggregates, which stained with Alcian blue at pH 1.0, were tentatively identified as chondrocytes (Fig.5A-C). Several different morphologies of chondrocytes were noted. Some nodules appeared to merge into the surrounding cell layer (Fig. 5A). Other nodules appeared elongated and had fibroblast-like cells at the edges (Fig. 5B). A third morphology consisted of chondrocytes in a nodule without any surrounding stellate or spindle-shaped cells (Fig. 5C). Cell aggregates of polygonal cells appeared after four weeks in culture. They were most common in the wells treated with 10-9 to 10-10 M dexamethasone but appeared in small numbers at all concentrations of dexamethasone. These cells had a dense extracellular matrix that appeared quite dark under phase contrast microscopy, and the matrix stained with Von Kossa's stain (Fig. 6A). It was found that the staining could be prevented by pre-treatment with EGTA (data not shown). All of this indicated a calcified extracellular matrix. Therefore these cells were tentatively identified as osteoblasts. Additionally, areas of Von Kossa's positive staining material could be seen in nodules with the morphological and staining characteristics of cartilage (Fig. 6B). It is uncertain whether this represented calcified cartilage or osteoblasts in proximity to chondrocytes. By 4 weeks of treatment with dexamethasone, cells of roughly parallelogram shape containing fibers were observed. These cells were most numerous at 10-7 and 10-6M 7

dexamethasone. The fibers stained with an antibody to smooth muscle a-actin and were identified as smooth muscle cells (Fig. 7). Also by 4 weeks of treatment with dexamethasone, cells of polygonal shape but without discernible extracellular matrix appeared in the 10-7 and 10-6M dexamethasone cultures. These cells took up DiI-Acyl-LDL into cytoplasmic vesicles (Fig. 8) and have thus been identified as endothelial cells. The incubation period with DiI-Acyl-LDL was limited to 4 hr., and the smooth muscle cells did not exhibit staining (data not shown). Areas of spindle-shaped cells with a granular matrix that stained lightly with Alcian blue, pH 1.0 appeared at 10-10 to 10-8M dexamethasone treatment (Fig. 6C). The cells grew in swirl patterns. On the basis of the morphology and staining pattern, the cells were tentatively identified as fibroblasts. These cells predominated in secondary cultures grown in both horse and fetal bovine serum that was not pre-selected. In addition, when secondary cultures were grown in fetal bovine serum, the cells would no longer form cartilage, skeletal muscle, adipocytes, or bone when treated with dexamethasone (data not shown). Cultures from cells isolated from the skeletal muscle of near term fetuses behaved virtually identically to cells isolated from newborn animals. In order to assess the longevity of the cells in culture, and determine if there were limits to their response to dexamethasone, cells from the newborn rats were passaged in culture and tested with dexamethasone after each passage. Each passage consisted of plating 100,000 cells/100 mm culture dish. At confluence, the dishes contained between 3 and 3.5 x 106 cells. Thus, each passage represents about 5 cell doublings. Currently, we have tested cells through the 17th passage, or about 85 cell doublings. There has been no difference in the response to dexamethasone between any of the passages and the secondary cells discussed above.

8

DISCUSSION A previous study reported the isolation of a population of cells from day 11 chick embryonic leg muscle. When treated with dexamethasone, these cells gave rise to the differentiated mesodermal phenotypes of fat, cartilage, bone, muscle, and fibroblats. The cells were designated putative mesenchymal stem cells (17). Clonal analysis demonstrated individual populations that exhibited similar pluripotential properties and were thus designated pluripotent mesenchymal stems cells (24). The present study was undertaken to determine if a corresponding population of cells can be isolated from rat skeletal muscles. The isolation of these putative stem cells involved a two step process similar to that used for the chick cells. The first step was a primary culture of cells obtained from enzymatic digestion of the extracellular matrices of skeletal muscle, releasing the connective tissue-resident cells. This primary culture was designed to allow lineage committed cells, particularly satellite cells for the myoblast lineage, to differentiate. The primary culture contained large, multinucleated myotubes in addition to a population of mononucleated cells (Fig. 1). The second step was the isolation of the mononuclear cells, which were thought to be the mesenchymal cells, by filtration (to remove the myotubes), freezing, and then reculturing the thawed cells into the secondary culture. The major difference between this method and that used for the chick was that the chick cultures required the use of stage specific embryo extract and horse serum while the rat cultures required only horse serum (17,18). It is important to use a preselected serum since extensive testing has demonstrated that most lots of serum, including fetal bovine serum, result in fibroblasts in the secondary culture (data not shown). We speculate that serum contains an agent(s) that induces the committment of the secondary cells to the fibroblast lineage. Once, committed, the cells can no longer respond to other inductive agents to produce other mesodermal lineages. The secondary cells maintained a stellate morphology when cultured in media alone (Fig. 2A). No distinct phenotypes were observed. However, with dexamethasone treatment several phenotypes appear, depending upon both the length of treatment and the dose of dexamethasone. The first phenotype to appear was multinucleated, spontaneously contracting cells in either a linear or branched morphology. These cells stained with an antibody to skeletal myosin and have thus been identified as skeletal myotubes (Fig. 3A-C). The myogenic response to dexamethasone was quantified (Fig. 4). Myogenesis appeared at all concentrations of dexamethasone tested, but was maximal at 10-7 M. The data also showed that dexamethasone did not act as a mitogen in this system. The second phenotype appeared at about 2 weeks' treatment with dexamethasone and consisted of round cells containing what appeared to be large droplets in the cytoplasm. These droplets stained with Sudan black B for neutral lipids and these cells have been identified as adipocytes (Fig. 2B and C). Adipocytes were present at 10-8 to 10-6 M dexamethasone, and were present in the same cultures as the myotubes. However, the adipocytes did not stain for skeletal muscle myosin. By 3 weeks in culture, cell aggregates with a refractile matrix were present in cultures treated with 10-9 - 10-6 M dexamethasone. The extracellular matrix of these cell aggregates stained with Alcian blue, pH 1.0, indicating the presence of sulfated glycosaminoglycans and which identified these cells as chondrocytes (Fig. 5A-C). The cartilage nodules had different morphologies: some nodules merged into the cell layer (Fig. 5A), others contained fibroblastlike cells (Fig. 5B), some nodules were solitary without any surrounding stellate or spindle9

shaped cells (Fig. 5C), and some nodules stained for mineral with Von Kossa's stain (Fig. 6B). These morphologies resemble the in vivo cartilages hyaline, fibro, articular, and growth plate, respectively. However, further tests are necessary before any definitive statement can be made. By 4 to 5 weeks, cell aggregates of polygonal cells appeared in the secondary cultures. These cells had a dense extracellular matrix that stained for calcium phosphate and were therefore tentatively identified as osteoblasts. The appearance of skeletal muscle, fat, cartilage, and bone in the secondary cultures correlates with the response seen in the chick mesenchymal stem cells (17,24). The response is identical even to the concentrations of dexamethasone where the phenotypes appeared and the timing of their appearance. The response is also identical to that seen by Grigoriadis et al. (19). These investigators treated clonal cells derived from day 19 fetal rat calvaria with dexamethasone in the presence of fetal bovine serum. Again, both the timing and the concentrations of dexamethasone correlate with the results in this study. However, two additional phenotyes not observed in the previous two studies were identified. Large irregular shaped cells appeared at 10-7 and 10-6 M dexamethasone by 4 weeks in culture. These cells stained with a monoclonal antibody to smooth muscle a-actin and were identified the cells as smooth muscle cells. Also present at the same time and dexamethasone concentrations were round cells that incorporated acyl-low density lipoprotein and were thus identified as endothelial cells (Figs 7 and 8). The cells in secondary culture required dexamethasone to form the observed phenotypes. It should be noted that several phenotypes were present within an individual culture well. The identity of those phenotypes depended upon the dexamethasone concentration as described above. The mechanism of action of dexamethasone on the cells is unclear. Dexamethasone is known to stimulate myogenesis (25-28), osteogenesis (29-31), adipogenesis (32-34), and chondrogenesis (35-38) in cultures of embryonic or lineage-committed cells. In addition, dexamethasone and other glucocorticoids have been shown to have effects on the production by cells of local mediators such as prostaglandins (39,40). Therefore, dexamethasone may 1) be affecting the cell sensitivity to other regulatory factors present in the serum, 2) increasing the production by the cells of regulatory agents, or 3) non-specifically stimulating initiation of gene expression that leads to differentiation. This latter view is supported by the observations: 1) several phenotypes were observed at each concentration of dexamethasone, 2) dexamethasone induced phenotypic expression across a wide range of concentrations (Fig. 4), and, 3) some of the cells remained undifferentiated. However, whatever the mechanism, it remains that dexamethasone regulates the differentiation potential of this cell population. It is generally recognized that during development there exists a mesenchymal stem cell which is pluripotent and which eventually differentiates into all of the cell phenotypes of the mesodermal lineage (41-44). These phenotypes include skeletal myotubes, adipocytes, chondrocytes, osteoblasts, fibroblasts, etc. However, there are indications that cells capable of differentiating into multiple mesenchymal phenotypes may survive into the post-natal animal. In the adult Ambystoma salamander, cells have been identified that, when the limb is amputated, formed the regeneration blastema that eventually restored all the mesodermally-derived limb structures (45,46). Mammalian marrow stroma contains cells that are capable of differentiating into cartilage and bone (30,47). Healing of peritoneal mesothelium apparently involves the differentiation of new mesothelial cells from "primitive mesenchymal cells" underlying the

10

mesothelium (48) and the cells that cover vascular prosthesis are speculated to arise from "circulating stem cells" (49). However, the most extensive evidence for stem cells in adult mammals arises from the studies involving the de novo bone induction in ectopic sites following the implantation of demineralized bone matrix or proteins isolated from the matrix. Implantation of demineralized bone matrix either intramuscularly or subcutaneously results in the formation first of cartilage (59 days post-implantation), then bone (10-17 days post-implantation), and finally hematopoiesis in the induced bone (14-21 days post-implantation) (12,13). Normally, of course, bone will never form at either site. The respondent cells have been referred to as "mesenchymal cells" (13,50). Further studies demonstrated that minced muscle grown in vitro on demineralized bone matrix resulted in the induction of cartilage in the muscle (15). Nathanson found that the connective tissue of a number of tissues could be induced to form cartilage when grown on demineralized bone matrix in vitro (51). All these studies suggest a resident mesenchymal stem cell which, under appropriate stimulation, is capable of differentiating into various mesodermal phenotypes. The secondary cells have the characteristics of a population of mesenchymal stem cells. The secondary cells are quiescent, exhibiting no discernible phenotype in control cultures. Yet the cells have the capability of differentiating into a number of phenotypes when stimulated. In addition, the secondary cultures were taken through 85 cell doublings, and they maintained their ability to differentiate into multiple phenotypes. Eighty-five cell doublings is far beyond Hayflick's limit of 50 cell doublings prior to cell death (52) and indicates the cells have an unlimited mitogenic potential, which is a criterion for a pluripotent stem cell. Together, these data support the hypothesis that these cells represent a population of mammalian mesenchymal stem cells. The secondary cells could represent a population of lineage committed cells, each cell committed to a specific lineage, such as a myosatellite cell, but not yet expressing the terminal phenotype, and/or pluripotential stem cells. The initial primary culture was designed to eliminate lineage-committed cells by allowing them to differentiate. There were no detectable myotubes, adipocytes, smooth muscle cells, or endothelial cells in the secondary cultures maintained in control medium; the mesodermal phenotypes expected from cells isolated from skeletal muscle. Even in the presence of serum, the secondary cells required stimulation with dexamethasone in order to differentiate, which argues that the cells were not lineage committed cells programmed to differentiate. The most likely possibility is that the secondary cultures contain at least some pluripotent mesenchymal stem cells, mixed with some lineage-committed cells. Clonal analysis is required to conclusively determine the identity of the stem cells (pluripotent and progenitor) in this population and these experiments are underway. However, pluripotent mesenchymal stem cells were found during such an analysis of cells isolated from fetal chicks (24) and fetal rat calvaria (19). Since the secondary cells behave identically to those, it seems plausible to hypothesize that pluripotent cells exist in the secondary cells as well. We postulate the existence of a pluripotent mesenchymal stem cell that can differentiate into a variety of adult morphologies. The existence of such a population of mesenchymal stem cells offers a variety of possibilities. Perhaps the foremost of these is the potential use of these cells for tissue repair. Wound healing results in scar tissue formation (53, 54). Scar tissue, while filling the space made by the wound, does not have the physiological or mechanical properties of the original differentiated mesodermal tissues. However, the resident mesenchymal stem cells, with their potential to form a number of mesodermal tissues, could provide a means of

11

regenerating these tissues following trauma. At present the cells will be used to test both intrinsic and extrinsic factors that may direct the commitment or differentiation of mesenchymal stem cells to specific phenotypes. Once these factors have been clearly identified, then they, in combination with mesenchymal stem cells that can be grown ex vivo, could be used to achieve tissue regeneration following trauma or other damage to body tissue.

12

ACKNOWLEDGMENTS The authors wish to thank John Reichert and John Knight for photographic assistance and Lorrie Smith for assistance in preparing the manuscript. This work was funded by grants from the Clinical Research Center of the Medcen Foundation.

13

LEGENDS TO FIGURES Figure 1. Primary Cultures of Cells Isolated from Newborn Rat Skeletal Muscle. Phase contrast microscopy, original magnification = 100x. A. Cells on day 1 of culture. b = bipolar cell, s = stellate cell. B. Cells on day 3 of culture. m = small myotube, s = stellate. C. Cells on day 8 in culture. s = mononucleated stellate cell, m = large multinucleated myotube, small arrows point to nuclei. Figure 2. Secondary Culture of Cells Isolated from Newborn Rat Skeletal Muscle. Phase contrast microscopy, original magnification = 100x. A. Control culture after 5 weeks in culture. Arrows point to stellate cells. B. Culture treated with 10-7 M dexamethasone after 3 weeks in culture. Arrows point to cells with large refractile vesicles. C. Culture treated with 10-7 M dexamethasone after 3 weeks in culture and stained with Sudan black B. Arrows point to stained cells (adipocytes). Figure 3. Secondary Culture of Cells Isolated from Newborn Rat Skeletal Muscle. A. Phase contrast of culture treated with 10-8 M dexamethsone after 2 weeks in culture, original magnification = 100x. s = stellate cell, m = myotube. Small arrows point to nuclei. B. View shown in A, original magnification = 200x, showing multinucleated myotube. Arrows point to nuclei. C. Bright field of culture treated with 10-7 M dexamethasone after 2 weeks in culture and stained with an antibody to myosin. Arrows point to nuclei. Note partially contracted myotube in lower right corner. m = myotube. Figure 4. Myosin Assay of Secondary Cultures. Assay was performed after 2 weeks in culture. Points represent mean ± SEM of 4 samples. Visual observation indicated that the increased myosin content is due to an increased number of myotubes, but does not represent changes in size or number of nuclei. Figure 5. Secondary Culture of Cells Isolated from Newborn Rat Skeletal Muscle. Phase contrast microscopy of cultures stained with Alcian blue, pH 1.0, after 4 weeks in culture. Original magnification = 200x. A. Culture treated with 10-6 M dexamethasone, showing large cartilage nodule. C = cartilage. B. Culture treated with 10-8 M dexamethasone, showing cartilage nodule with fibroblast-like cells. C = cartilage. C. Culture treated with 10-9 M dexamethasone, showing cartilage nodules without surrounding cells. C = cartilage. Figure 6. Secondary Culture of Cells Isolated from Newborn Rat Skeletal Muscle. Phase contrast microscopy after 5 weeks in culture. Original magnification = 200x. A. Culture treated with 10-10 M dexamethasone, stained with Von Kossa's stain. b = bone. B. Culture treated with 10-9 M dexamethasone, stained with Von Kossa's stain, showing nodule of cartilage with stained mineral. c = cartilage, b = mineralized matrix. C. Culture treated with 10-8 M dexamethasone stained with Alcian blue, pH 1.0, showing granular pattern of spindle-shaped cells. Figure 7. Secondary Culture of Cells Isolated from Newborn Rat Skeletal Muscle. Phase contrast microscopy after 5 weeks in culture. Original magnification = 100x. A. Control culture stained with antibody to smooth muscle a-actin. B. Culture treated with 10-6 M dexamethasone stained with antibody to smooth muscle a-actin. Arrows point to cells positive for antibody. Figure 8. Secondary Culture of Cells Isolated from Newborn Rat Skeletal Muscle. Phase contrast microscopy after 5 weeks in culture. Original magnification = 200x. A. Phase contrast photograph of culture treated with 10-6M dexamethasone and incubated with acetylated LDL.

14

Arrows point to corresponding fluorescent cells in (B). B. Same field as (A) using fluorescence microscopy.

15

REFERENCES 1. Mauro A. Satellite cells of skeletal muscle fibers. J Biophys Biochem Cytol 1961;9:493-95. 2. Allbrook D. Skeletal muscle regeneration. Muscle Nerve 1981;4:234-35. 3. Grounds MD, McGeachie, JK. Myogenic cell replication in minced skeletal muscle isografts of swiss and balb/c mice. Muscle Nerve 1990;13:305-13. 4. Carlson BM. Regeneration of entire skeletal muscles. Fed. Proc 1986;45:1456-1460. 5. Schultz E, Jaryszak DL, Valliere, CR. Response of satellite cells to focal skeletal injury. Muscle Nerve 1985;8:217-223. 6. Jarvinen M. Healing of a crush injury in rat striated muscle. Acta Path Microbiol Scand Sect A 1975;83:269-282. 7. Benoit PW, Belt WD. Destruction and regeneration of skeletal muscle after treatment with a local anesthetic, bupivacaine. (Marcain). J Anat 1970;1970:547-556. 8. McGeachie JK, Grounds MD. Initiation and duration of muscle precursor replication after mild and severe injury to skeletal muscle. Cell Tiss Res 1987;248:125-130. 9. Levander, G. Induction phenomena in tissue regeneration. Baltimore: Williams and Wilkins, 1964. 10. Carlson, BM. Relationship between tissue and epimorphic regeneration of skeletal muscle In: Mauro A, Bischoff R, Carlson, BM, editors. Muscle regeneration. New York: Raven Press, 1979:57-71. 11 Klein-Ogus C, Harris JB. Preliminary observations of satellite cells in undamaged fibers of the rat soleus muscle assaulted by a snake-venom toxin. Cell Tiss Res 1983;230:671-676. 12. Urist MR. Bone: formation by autoinduction. Science 1965;150:893-899. 13. Reddi AH, Anderson WA. Collagenous bone matrix-induced endochondral ossification and hemaopoiesis. J Cell Biol 1976;69:557-572. 14. Wang EA, Rosen V, D'Alessandro JS, Bauduy M, Cordes P, Harada T, Israel DI, Hewick RM, Kerns DM, LaPan P, Luxenberg DP, McQuaid D, Moutsatsos IK, Nove J, Wozney JM. Recombinant human bone morphogenetic protein induces bone formation. Proc Natl Acad Sci USA 1990;87:2220-2224. 15. Urist MR, Terashima Y, Nakagawa M, Stamos, C. Cartilage tissue differentiation from mesenchymal cells derived from mature muscle in tissue culture. In Vitro 1978;14:697-706. 16. Lucas PA, Syftestad GT, Caplan AI. A water-soluble fraction from adult bone stimulates the differentiation of cartilage in explants of embryonic muscle. Differentiation 1988;37:47-52. 17. Young HE, Ceballos EM, Smith JC, Lucas PA, Morrison DC. Isolation of embryonic chick myosatellite and pluripotent stem cells. J Tiss Cult Meth 1992;14:85-92. 18. Young HE, Morrison DC, Martin JM, and Lucas PA. Cryopreservation of embryonic chick myogenic lineage-committed stem cells. J Tiss Cult Meth 1991;13:275-284. 19. Grigoriadis AE, Heersche JNM, Aubin JE. Differentiation of muscle, fat, cartilage, and bone from progenitor cells present in a bone-derived clonal cell population: effect of dexamethasone. J Cell Biol 1988;106:2139-2151. 20. Humason G. Animal tissue techniques, 3rd ed. San Francisco: WH Freeman and Co, 1972. 21. Johnson-Wint B, Hollis S. A rapid in situ deoxyribonucleic acid assay to determine cell number in culture and tissue. Anal Biochem 1982;122:338-344. 22. Young HE, Sippel J, Putnam LS, Lucas PA, Morrison DC. Enzyme-linked immuno-culture assay. J Tiss Cult Meth 1992;14:31-36.

16

23. Voyta JC, Via DP, Butterfield E, Zetter BR. Identification and isolation of endothelial cells based on their increased uptake of acetylated-low density lipoprotein. J Cell Biol 1984;99:20342040. 24. Young HE, Ceballos EM, Smith JC, Mancini ML, Wright RP, Ragan BL, Bushell I, Lucas PA. Pluripotent mesenchymal stem cells reside within avian connective tissue matrices. In Vitro Cell Dev Biol 1993;29A:723-736. 25. Allen RE, Dodson MV, Luiten LS, Boxhorn LK. A serum-free medium that supports the growth of cultured skeletal muscle satellite cells. In Vitro Cell Dev Biol 1985;21:636-640. 26. Ball EH, Sanwal BD. A synergistic effect of glucocorticoids and insulin on the differentiation of myoblasts. J Cell Physiol 1980;102:27-36. 27. Guerriero VJr, Florini JR. Dexamethasone effects on myoblast proliferation and differentiation. Endocrinology 1980;106:1198-1204. 28. Smith TJR, Dana R, Krichevsky A, Bilezikian JP, Schonberg M. Inhibition of ß-adrenergic responsiveness in muscle cell cultures by dexamethasone. Endocrinology 1981;109:2110-2116. 29. Tenenbaum HC, Heersche JNM. Dexamethasone stimulates osteogenesis in chick periosteum in vitro. Endocrinology 1985;117:2211-2217. 30. Owen ME, Joyner CJ. Clonal analysis in vitro of osteogenic differentiation of marrow CFUF. J Cell Science 1987;87:731-738. 31. Bellows CG, Heersche JNM, Aubin JE. Determination of the capacity for proliferation and differentiation of osteoprogenitor cells in the presence and absence of dexamethasone. Dev Biol 1990;140:132-138. 32. Greenberger JS. Corticosteroid-dependent differentiation of human marrow preadipocytes in vitro. In Vitro 1979;15:823-828. 33. Houner H, Schmid P, Pfeiffer EF. Glucocorticoids and insulin promote the differentiation of human adipocyte precursor cells into fat cells. J Clin Endocrin Metab 1987;64:832-835. 34. Schiwek OR, Loffler G. Glucocorticoid hormones contribute to the adipocyte activity of human serum. Endocrinology 1987;120:469-474. 35. Bernier SM, Goltzman D. Regulation of the expression of the chondrocyte phenotype in a skeletal cell line. J Bone Miner Res 1993;8:475-484. 36. Zimmerman B, Cristea R. Dexamethasone induces chondrogenesis in organoid culture of cell mixtures from mouse embryos. Anat Embryol 1993;187:67-73. 37. Grigoriadis AE, Aubin JE, Heersche JNM. Effects of dexamethasone and vitamin D3 on cartilage differentiation in a clonal chondrogenic cell population. Endocrinology 1989;125:21032110. 38. Zalin RJ. The role of hormones and prostanoids in the in vitro proliferation and differentiation of human myoblasts. Exp Cell Res 1987;172:265-281. 39. Foster SJ, Perkins JP. Glucocorticoids increase the responsiveness of cells in culture to prostaglandin E1. Proc Natl Acad Sci USA 1977;74:4816-4820. 40. Williams IH, Polakis SE. Differentiation of 3T3-L1 fibroblasts to adipocytes. The effect of indomethacin, prostaglandin E1 and cyclic AMP on the process of differentiation. Biochem Biophys Res Commun 1977;77:175-186. 41. Wilmer EN. Cytology and evolution, 2nd edition. New York: Academic Press, 1970. 42. Sulston J, Schierenberg E, White J, Thomson J. The embryonic cell lineage of the nematode Caenorhabditis elegans. Dev Biol 1983;100:64-119. 43. Owen ME. Marrow stromal stem cells. J Cell Sci 1988;10[Suppl]:63-76.

17

44. Caplan AI. Mesenchymal stem cells. J Orthop Res 1991;9:641-650. 45. Young HE, Bailey CF, Markwald RR, Dalley BK. Histological analysis of limb regeneration in postmetamorphic adult Ambystoma. Anat Rec 1985;212:183-194. 46. Young HE, Dalley BK, Markwald RR. Glycoconjugates in normal wound tissue matrices during the initiation phase of limb regeneration in adult Ambystoma. Anat Rec 1989;223:231241. 47. Friedenstein AJ. Precursor cells of mechanocytes. Int Rev Cytol 1973;47:327-355. 48. diZarega GS. Contemporary adhesion prevention. Fertil Steril 1994;61:219-235. 49. Scott SM, Barth MG, Gaddy LR, Ahl ET. The role of circulating cells in the healing of vascular prostheses. J Vasc Surg 1994;19:585-593. 50. Urist MR. Bone morphogenetic protein, bone regeneration, heterotopic ossification, and the bone-bone marrow consortium. Bone Min Res 1989;6:57-112. 51. Nathanson MA, Hilfer SR, Searls RL. Formation of cartilage by non-chondrogenic cell types. Develop Biol 1978;64:99-117. 52. Hayflick L. The limited in vitro lifetime of human diploid cell strains. Exp Cell Res 1965;37:614-636. 53. Clark RAF. Basics of cutaneous wound repair. J Dermatol Surg Oncol 1993;19:693-706. 54. Orgill D, Demling RH. Current concepts and approaches to wound healing. Crit Care Med 1988;16:899-908.

18