- Email: [email protected]
Isolation and characterization of multipotential mesenchymal stem cells from feline bone marrow
Experimental Hematology 30 (2002) 879–886
Isolation and characterization of multipotential mesenchymal stem cells from feline bone marrow Douglas R. Martin, Nancy R. Cox, Terri L. Hathcock, Glenn P. Niemeyer, and Henry J. Baker The Scott-Ritchey Research Center, College of Veterinary Medicine, Auburn University, Auburn, Ala., USA (Received 10 January 2002; revised 22 March 2002; accepted 18 April 2002)
Objective. Although several types of stem cells have been isolated from rodent and human tissues, very few data exist on stem cell isolation from nonrodent animals, which seriously limits the advancement of stem cell biology and its ultimate translation to human clinical applications. Domestic cats are used frequently in biomedical research and are the preferred species for studies of normal physiology and disease, particularly in neuroscience. Therefore, the objective of this study was to characterize mesenchymal stem cells (MSC) from feline bone marrow for use in research on the application of stem cells to human health problems for which cats are the preferred model. Methods. Mesenchymal stem cells from feline bone marrow were isolated by standard methodology developed for other species and characterized according to morphology, growth traits, cell-surface antigen profile, and differentiation repertoire in vitro. Results. Feline mesenchymal stem cells exhibit a fibroblast-like morphology with bipolar or polygonal cell bodies and possess a cell-surface antigen profile similar to their rodent and human counterparts. Feline MSC exist at a frequency of 1 in 3.8 105 bone marrow mononuclear cells and are capable of differentiation to adipocytic, osteocytic, and neuronal phenotypes when exposed to appropriate induction media. Conclusions. Mesenchymal stem cells isolated from feline bone marrow possess several traits typical of MSC from other species. Characterization of feline mesenchymal stem cells will facilitate future studies of stem cell biology and therapeutics for which the domestic cat is an indispensable model. © 2002 International Society for Experimental Hematology. Published by Elsevier Science Inc.
The defining characteristics of stem cells are self-renewal and the ability to differentiate into one or more specialized cell types [1]. Traditionally, stem cells have been divided into two major groups: 1) embryonic stem cells, derived from the inner cell mass of fertilized ova which can differentiate into any cell type, and 2) tissue-specific stem cells, derived from specific organs. Tissue-specific stem cells once were believed to have differentiation capacity limited to the organ of origin but now are recognized to be capable of differentiation into cells of other tissues. For example, mesenchymal stem cells (MSC), also known as bone marrow stromal cells, are capable of differentiation in vitro to marrow and non-marrow cell types, including: adipocytes,
Offprint requests to: Henry J. Baker, D.V.M., Professor and Director, The Scott-Ritchey Research Center, College of Veterinary Medicine, Auburn University, AL 36849; E-mail: [email protected]
chondrocytes, osteocytes [2–5], myocytes [6,7], astrocytes, and neurons [8–11]. Differentiation of MSC to a variety of cell types, including astrocytes and neurons, has been demonstrated in vivo as well [3,12–15]. Although present in very small numbers in the bone marrow [16–18], MSC are capable of substantial proliferation and expansion in culture [2,17,19]. Undifferentiated MSC exhibit a fibroblast-like morphology and a characteristic pattern of cell-surface antigens [2,8] as well. The criteria listed above, including the ability to differentiate to many cell types, have been used to define a prototypic mesenchymal stem cell phenotype which is consistent among a variety of species, including mouse [16], rat [8], dog [20], and human [2]. The majority of research on stem cells, including MSC, has been done on cells derived from rodent and human tissues. However, investigation of stem cells from nonrodent mammals is essential for progress to be made toward the ultimate application of stem cells to human health problems.
0301-472X/02 $–see front matter. Copyright © 2002 International Society for Experimental Hematology. Published by Elsevier Science Inc. PII S0301-472X(02)0 0 8 6 4 - 0
880
D.R. Martin et al./Experimental Hematology 30 (2002) 879–886
The domestic cat provides a number of important advantages for biomedical research, particularly stem cell transplantation in fetal and postnatal animals. There is substantial knowledge about both prenatal and postnatal hematopoiesis; fetal, neonatal, and adult immunology; and bone marrow transplantation in domestic cats. Although inbred rodents provide an experimental advantage for stem cell transplantation because they are histocompatible, cats heterologous for histocompatibility loci more closely mimic the challenges and limitations that are likely to be encountered in human patients receiving allogeneic cell transplants [21]. Cats have been used extensively for studies of normal physiology and disease, particularly of the nervous system, which presents a critically important opportunity for research on stem cell transplantation therapy of neurological disorders [22]. Cats are models for approximately 60 inherited human diseases [22–24], including lysosomal storage diseases, diabetes, heart muscle disorders, polycystic kidney disease, and retinal atrophy. Also, cats are important models of acquired immunodeficiency diseases [25–27]. Because several feline viral diseases adversely affect bone marrow as well as other organ systems in the cat, a substantial base of information is available concerning cell markers, particularly of the hematopoietic and immune system of this species, which provides opportunities for experimental assessment of the differentiation fate of transplanted stem cells. In addition, cats are maintained easily in laboratory breeding colonies, produce 4 to 6 progeny per breeding, and have multiple breeding cycles per year with a short gestation of 63 1 days. Cats respond well to intrauterine surgery and are large enough to permit frequent sampling of tissues and body fluids [22]. The purpose of this study was to isolate and characterize MSC from normal domestic cats as the essential first step toward their use in feline models of development, physiology, and disease.
Materials and methods Materials All tissue culture media, supplements, and differentiation reagents were purchased from Sigma (St. Louis, MO, USA) unless otherwise noted. Percoll was purchased from Amersham Pharmacia (Piscataway, NJ, USA). All antibodies for flow cytometry were shown by the manufacturer to cross react with feline cells. Antibodies to CD9 and CD45 were purchased from Serotec (Raleigh, NC, USA), while all other antibodies were purchased from VMRD (Pullman, WA, USA). Antibodies to neuronal proteins used for immunohistochemistry were as follows: mouse anti-pig neurofilament-M (160 kDa, Accurate Chemical, Westbury, NY, USA), rabbit anti-human trkA (nerve growth factor receptor, Santa Cruz Biotechnology, Santa Cruz, CA, USA) and mouse anti-human III-tubulin (Sigma). Isolation of mesenchymal stem cells from feline marrow Feline bone marrow was harvested antemortem by bone marrow aspiration from the greater trochanter of the femur or greater tubercle of the humerus, or at necropsy by flushing the shaft of a femur. Bone marrow was collected into 1–5 volumes Iscove’s modified Dulbecco’s medium (IMDM) containing 200 units/mL heparin, filtered (100 m), and centrifuged at 900g to pellet cells. After two rinses in phosphate-buffered saline (PBS), 108 cells were loaded onto 12.5-mL Percoll (1.073 g/mL in 0.15 M NaCl) and spun at 1100g for 30 minutes. Mononuclear cells were collected at the Percoll interface, rinsed twice in PBS, and seeded at 2 105/cm2 in Dulbecco’s Modified Eagle’s Medium (DMEM) (1 g/L glucose) with 10% fetal bovine serum (FBS; Hyclone, Logan, UT, USA) (Passage 0). Three lots of FBS were screened for their ability to support feline mesenchymal stem cell proliferation for 2 to 3 passages, followed by successful differentiation to the neuronal morphology. The lot of FBS that was optimal for proliferation also allowed induction to a neuronal phenotype and was chosen for subsequent studies. Nonadherent cells were removed and media replaced 1 to 3 days after initial plating. After 7 to 12 days in culture, isolated colonies of mesenchymal stem cells were apparent.
Table 1. Cell-surface antigen profile of feline vs. rodent/human MSC Cell-surface antigen (lineage expression) CD4 (M, monocyte, T helper) CD8 (T cytotoxic/suppressor cell) CD9 (myeloid) CD13 (granulocyte, monocyte) CD14 (M, monocyte, some granulocytes) CD18 (leukocyte) CD41/61 (megakaryocyte, platelet) CD44 (brain, erythrocyte, leukocyte, platelet) CD45 (leukocyte) MHC I (almost all nucleated cells) MHC II (antigen-presenting cells) *No data available for CD8 or CD13 expression on rodent or human MSC. † MSC tested with anti-CD61, not anti-CD41/61. M macrophage.
Feline MSC
Rodent/human MSC [2,8]
ND* ND* †
D.R. Martin et al./Experimental Hematology 30 (2002) 879–886
These cells were trypsinized and replated at 8000/cm2, and fresh medium was added every 3 to 4 days (Passage 1). At passage 2, media was changed to DMEM (4.5 g/L glucose) with 20% fetal bovine serum (Dr. D. Woodbury, Robert Wood Johnson Medical School, Piscataway, NJ, personal communication). Flow cytometric determination of cell-surface antigen profile Feline MSC were trypsinized (0.05% trypsin-EDTA), resuspended in staining media (1% bovine serum albumin, 0.2% sodium azide in PBS), and stained on ice according to the manufacturer’s recommendations with the monoclonal antibodies listed in Table 1. Positive cells were detected with a 1:100 dilution of FITC-conjugated goat anti-mouse IgG (Jackson ImmunoResearch, West Grove, PA, USA) on a Coulter Epics Elite fluorescence-activated cell sorter. Differentiation protocols Adipocytic differentiation [3] was achieved by allowing cells to remain at confluency for 3 to 7 days, followed by incubation in Alpha Minimal Essential Medium ( -MEM) supplemented with 10% FBS, 10% normal rabbit serum, 10 nM dexamethasone, 5 g/mL insulin, and 50 M 5,8,11,14-eicosatetraynoic acid. After 3 days, cells were cytocentrifuged and stained for lipid droplets with Oil Red O (0.3% in isopropanol with 0.4% dextrin). The osteocytic phenotype [28] was induced by seeding cells at 6000–8000/cm2, followed 1 day later by incubation in DMEM containing 10% FBS, 100 mM dexamethasone, 10 nM -glycerophosphate, and 0.25 mM L-ascorbic acid. Differentiation proceeded for 3 weeks with media changes twice weekly. Induced cells were stained for calcium with Alizarin Red S (10%, pH 4.2) or tested for alkaline phosphatase activity (Sigma Diagnostic kit #104). The long-term neuronal differentiation protocol of Woodbury et al. [8] was modified by the substitution of -MEM for DMEM in the induction media. Specifically, cells were expanded to 70% to 80% confluency in DMEM with 10% FBS, and basic fibroblast growth factor (10 ng/mL) was added for 24 hours. Thereafter, MSC were incubated in neuronal induction media ( -MEM, 2% dimethylsulfoxide [DMSO], 200 M butylated hydroxyanisole, 25 mM potassium chloride, 2 mM valproic acid, 10 M forskolin, 1 M hydrocortisone, 5 g/mL insulin, and 2 mM L-glutamine with no serum) [8]. After 24 hours of neuronal induction, cells were stained for expression of the neuron-specific proteins neurofilament M (NF-M), nerve growth factor receptor (trkA), or III-tubulin, as described below.
881
of morphology. Phenotypically, mesenchymal stem cell colonies appeared as fibroblast-like cells, bipolar or polygonal in morphology as compared with contaminating monocyte colonies, which appeared as large, round cells with irregular boundaries and large nuclei.
Results Isolation of feline MSC In initial experiments, the total amount of marrow cells collected per bone (humerus or femur) ranged from 4.25 108 to 1.13 109 (mean 6.7 108 cells/bone, n 8 cats). After density gradient centrifugation, the bone marrow mononuclear cell (BMMC) population collected at the Percoll interface represented 10.6% to 23.4% of the total bone marrow isolate (mean 16.1%, n 8 cats). Feline MSC grew as isolated colonies 7 to 10 days after initial plating and were trypsinized for further expansion. After replating, cells were fibroblast-like, appearing polygonal or spindleshaped with long processes (Fig. 1). Morphologically, feline MSC appeared very similar to their rodent and human counterparts [2,8]. Mesenchymal stem cell isolation experiments were conducted on at least 15 normal cats, with each marrow sample yielding similar results in terms of colony and cellular morphology. Enumeration and expansion of MSC in feline bone marrow By comparing the number of mesenchymal stem cell colonies to the total number of bone marrow mononuclear cells seeded after Percoll centrifugation, the frequency of MSC was determined to range from 1 in 4.7 104 to 1 in 5.9 105, which compares favorably to data generated from mouse and human studies [16–18] (Table 2). Initial attempts
Immunohistochemistry of neuronally differentiated MSC Cells induced to the neuronal phenotype were fixed in 4% paraformaldehyde for 12 minutes, rinsed with PBS, and blocked in 10% normal goat serum for 1 hour. Fixed cells were stained in 1.5% goat serum with neurofilament M (1:40–1:200), trkA (1:200), or III-tubulin (1:200) at 4 C overnight. Positive cells were visualized with a FITC- or TRITC-conjugated secondary antibody (Jackson ImmunoResearch) and counterstained with propidium iodide or DAPI (Vector Laboratories, Burlingame, CA, USA). Enumeration of MSC in feline bone marrow To evaluate the percentage of MSC in feline bone marrow, mononuclear cells isolated by Percoll centrifugation were plated at 2 105 cells/cm2 in 6-well plates and incubated for 10 days. Isolated colonies were stained with methylene blue (0.167% in methanol) and enumerated, with mesenchymal stem cell colonies distinguished from monocyte contamination by microscopic evaluation
Figure 1. Morphological characterization of feline bone marrow stromal cells. In growth media (DMEM 20% FBS), feline MSC are long, flattened cells of a fibroblastic morphology. Scale bar 25 m.
882
D.R. Martin et al./Experimental Hematology 30 (2002) 879–886
Table 2. Frequency of MSC in bone marrow mononuclear cell fraction from different species Source of MSC
Frequency in BMMC
Feline (CR643)* Feline (N359)* Feline (N367)* Human [17] Mouse [16] Mouse [18]
1 in 5.9 105 1 in 4.9 105 1 in 4.7 104 1 in 1 105–1 106 1 in 3.3 105–3.3 106 1 in 2.2 104–3.3 105
*Numbers represent individual cats from which marrow was harvested.
to grow feline MSC at low cell densities (3, 10, 30, and 100 MSC/cm2 at passage 1) were unsuccessful. To assess the expansion capacity of feline MSC, initial cultures were generated by seeding a 75-cm2 flask with 1.5 107 bone marrow mononuclear cells. In an average culture, the frequency of MSC is 1 in 3.8 105 BMMC (mean of 3 cats reported in Table 2). Therefore, an initial culture of 1.5 107 BMMC contains approximately 40 MSC. Initial colonies of MSC were counted and found to have expanded to 5.3 1.2 105 (mean standard deviation, n 3) cells after 9 to12 days, a 13,250 3000-fold increase. After trypsinization of the initial cultures, MSC expanded approximately threefold for the first 2 passages and twofold thereafter (data not shown). Most cultures were proliferative to at least passage 6, indicating that an average mesenchymal stem cell can proliferate to a minimum of 7 107 MSC in vitro. The expansion potential of feline MSC may be considerably greater, however, since rigorous longevity studies to assess the maximum attainable passage number have not been performed. Cell-surface antigen profile Feline mesenchymal stem cell cell-surface antigen profile was ascertained by staining with feline-specific monoclonal antibodies followed by flow cytometry as shown in Table 1 and Figure 2. Feline MSC possess an array of cell-surface antigens similar to MSC from rodents or humans. Feline MSC are strongly positive for CD9, CD44 (Pgp-1 or HCAM), and MHC-I, while negative for CD45 (leukocyte common antigen). Differentiation of feline MSC To determine whether feline mesenchymal stem cells, like MSC from other species, are capable of differentiation to multiple cell types, feline cells were placed into induction media specific for the generation of adipocytes [3], osteocytes [28], or neurons [8]. To generate adipocytes, feline MSC at passages 3 through 5 were grown to confluency and incubated for 3 days prior to addition of adipocyte induction media. Sixty hours after induction, the confluent monolayer detached from the tissue culture flask and floated as a homogeneous mass in the culture media (n 2). Very few cells ( 1%) remained adher-
Figure 2. Flow cytometric determination of cell-surface antigen profile for feline MSC. Feline bone marrow stromal cells were stained with felinespecific antibodies to a variety of cell-surface antigens. The percentage of cells positive for each antibody is listed in the corresponding histogram. MSC from 4 separate cats were tested with similar results. Representative results from 1 cat are shown.
ent. The detached monolayer was dissociated into a single cell suspension with a serological pipette and cell number/ viability were assessed by trypan blue exclusion. The majority of the cells ( 95%) remained viable. Cytocentrifugation of the cells followed by Oil Red O staining revealed that 92.5% of feline MSC produced lipid droplets after incubation in adipogenic media compared to only 5.5% of control cells in normal growth media (Fig. 3). Approximately 700 cells from at least 5 microscopic fields were scored for acquisition of an adipocytic phenotype as detected by staining with Oil Red O [3]. Subsequent experiments (n 3), in which feline MSC were seeded at lower densities and not allowed to reach confluency prior to induction, resulted in an adherent monolayer of lipid-producing cells. It was estimated that at least 70% of the adipocyte-induced cells stained positively for Oil Red O, although rigorous determination of percentages was not possible due to the confluency of the monolayer at the time of staining (data not shown). When grown in osteocyte-specific induction media, feline MSC grew as tightly packed, spindle-shaped cells for approximately 2 weeks, at which time clusters of round, calcium-producing cells formed above the tightly packed monolayer. Calcium formation was demonstrated by staining with Alizarin Red S, a marker for the osteocytic phenotype (Fig. 4). Alkaline phosphatase activity was undetectable when tested on undifferentiated or osteocyte-induced MSC from three separate cats on cumulative days 1, 4, 6, 9, 11, 13, 16, and 18 postinduction. However, a positive control cell line (canine osteosarcoma #C0585-6, Dr. Lauren Wolfe, Auburn University College of Veterinary Medicine) exhibited robust alkaline phosphatase activity (data not shown).
D.R. Martin et al./Experimental Hematology 30 (2002) 879–886
Figure 3. Adipocyte differentiation of feline MSC. After incubation at confluency for 3 days followed by the addition of adipocyte differentiation media for 3 days, feline MSC were cytocentrifuged, fixed in 10% neutral buffered formalin for 25 minutes, and stained in Oil Red O to reveal lipid vacuoles. MSC at passage number 3 were used for this experiment. (A) Uninduced control MSC. (B) MSC induced to the adipocyte lineage. Scale bar 10 m.
The ability of feline MSC to differentiate to a cell type of disparate dermal origin was demonstrated by induction to the neuronal phenotype. When incubated in neuronal induction media, 75% to 85% of feline MSC rapidly (less than 1 hour) acquired a neurotypic morphology, with multipolar, rounded cell bodies and long, narrow processes that contacted other cells in the culture vessel (Fig. 5). The neuronal induction protocol was repeated at least 15 times on MSC from at least five different cats. MSC were induced successfully at passage numbers ranging from 1 to 6. Neuronal induction of MSC at higher passage numbers has not been attempted. In addition to acquiring a neuronal morphology, feline MSC were stained 24 hours postinduction and found to express neuron-specific proteins such as neurofilament M (160 kDa), trkA (nerve growth factor receptor), and III-tu-
Figure 4. Osteocyte differentiation of feline MSC. After a 3-week incubation in osteocyte induction media, feline MSC stained positively for calcium with Alizarin Red S, indicating osteocytic differentiation. MSC from 3 separate cats were used for osteocyte induction experiments. Four independent experiments were conducted with MSC at passages 2 through 4, and one representative result is shown. (A) Uninduced feline MSC. (B) Feline MSC induced to the osteocytic phenotype. Scale bar 20 m.
883
Figure 5. Morphological characterization and neuronal differentiation of feline MSC. (A) In growth media, feline MSC appear as long, flattened cells of a fibroblastic morphology. (B,C) After a preinduction treatment with basic fibroblast growth factor followed by addition of neuronal induction media, 75%–85% of feline MSC change to a neuronal morphology with long, thin processes and round cell bodies. Experiments were performed 4 times, and cells were counted from 10 microscopic fields. Scale bars (A,B) 25 m. Scale bar (C) 10 m.
bulin at levels that appeared to be much greater than control cells (Fig. 6). Discussion We have a long-standing interest in therapeutic strategies for the gangliosidoses, a subset of lysosomal diseases that are fatal neuronopathic disorders of humans and several other species. Mesenchymal stem cells are attractive candidates for stem cell transplantation–based therapy for the gangliosidoses, in which the primary affected organs are the nervous system, thymus, and liver. The feline gangliosidoses are authentic replicas of the human counterparts [29– 32], and the characterization of feline MSC as reported here provides the essential foundation for testing the fate, distribution, and therapeutic benefit of MSC transplantation for these diseases. Based on morphology, cell frequency rate and cell-surface antigen profile, mesenchymal stem cells from feline bone marrow are very similar to MSC from rodent or human sources. When isolated by standard methodology developed for other species [2,8], individual colonies of feline MSC become apparent by 1 week after plating. Feline MSC comprise a minor percentage of the total bone marrow mononuclear cell fraction, similar to that reported for other species (approximately 1 in 4 105 cells, Table 2). Morphologically, undifferentiated feline MSC are fibroblastlike with spindle-shaped or polygonal cell bodies (Fig. 1), similar to their rodent or human counterparts. It is estimated that one feline mesenchymal stem cell can proliferate to a minimum of 7 107 MSC in six passages. The expansion
884
D.R. Martin et al./Experimental Hematology 30 (2002) 879–886
Figure 6. Expression of neuronal markers after differentiation of feline MSC. Feline MSC were pre-induced in basic fibroblast growth factor, then induced to the neuronal phenotype with induction media. Twenty-four hours later, cells exhibiting the neuronal phenotype stained positively for (A) the 160-kDa neurofilament M, (B) trkA, the nerve growth factor receptor, and (C) III tubulin. Cells that did not change morphologically served as internal negative controls for staining (arrows). Scale bars 50 m.
potential of human MSC has been estimated to range from 5.5 108 to 1.2 109 in studies that propagated MSC for 10 to 25 passages [19,33]. Although initial studies suggest that feline MSC are not capable of attaining such high passage numbers, rigorous longevity studies have not been performed to date. Also, preliminary data indicate that growth factor supplementation can enhance the proliferation of feline MSC (data not shown). Therefore, a firm conclusion about the expansion potential of feline MSC awaits further investigation. Feline MSC are CD9, CD44, MHC-I, CD4, CD45, and MHC-II, a cell-surface antigen profile similar to MSC from rodents and humans [2,8] (Table 1, Fig. 2). Although rodent/human MSC are positive for CD61 (integrin III, vitronectin receptor ), the feline antibody used in this study was raised against the CD41/61 complex and may recognize a CD41- or complex-specific epitope. Therefore, until a CD61-specific antibody for feline cells is obtained, no conclusion can be drawn about the presence of this cellsurface antigen on feline cells. Also, no data were available regarding the expression of CD8 or CD13 on rodent/human MSC, although feline mesenchymal stem cells are negative for both CD8 and CD13, as shown in these experiments. In addition to the general characteristics described above, another defining feature of mesenchymal stem cells is multipotentiality, or the ability to acquire multiple cellular phenotypes when exposed to the appropriate stimuli. The ability of MSC to differentiate to cells of a mesodermal origin has been well documented and includes adipocyte, osteocyte, chondrocyte [2–5,16], and myocyte [6,7,16] phenotypes. Our studies support and extend these observations to include feline mesenchymal stem cell differentiation to adipocyte and osteocyte phenotypes (Figs. 3 and 4), as indicated by lipid accumulation and calcium deposition, respectively. Alkaline phosphatase activity often is used as an early marker of osteocytic differentiation. In human and mouse MSC, alkaline phosphatase activity is present at baseline levels in undifferentiated cells and at significantly increased
levels in osteocyte-induced cells prior to calcium deposition [2,5,16]. However, no alkaline phosphatase activity was detected in feline MSC either prior to or following induction to the osteocytic phenotype, a phenomenon that may be explained by the unusually short half-life of the feline enzyme. Since elevated alkaline phosphatase activity is a diagnostic indicator of liver or bone disease in both humans and domestic animals, it is known that the serum half-life of feline alkaline phosphatase is unusually short ([34] and Dr. J. Spano, Auburn University College of Veterinary Medicine, personal communication). Because we detected robust alkaline phosphatase activity in a canine osteosarcoma cell line used as a positive control, we believe that feline alkaline phosphatase simply is undetectable with our assay system. A substantial body of evidence is beginning to emerge that suggests that rodent and human MSC are capable not only of differentiation to mesodermal cell lineages but also of “transdifferentiation” to cell types derived from ectoderm. For example, with two different induction protocols, Sanchez-Ramos et al. [9] and Woodbury et al. [8] generated cells of a neuronal morphology that expressed neuron-specific/selective proteins such as neurofilament M, tau, neuron-specific enolase, trkA, nestin, and neuron-specific nuclear protein (NeuN). Similarly, when treated with neuronal induction media, 75% to 85% of feline MSC rapidly exhibited a neuronal morphology with round, multipolar or bipolar cell bodies and long, thin processes that contacted other cells in the culture dish (Fig. 5). Neuronally induced feline MSC expressed NF-M, trkA, and III-tubulin (a neuronspecific isoform of tubulin), at levels greater than control MSC (Fig. 6). Additional support for the transdifferentiative capacity of MSC comes from transplantation studies in which mesenchymal stem cells were injected into experimental animals. For example, Kopen et al. [3] found that murine MSC injected into the lateral ventricle of newborn mice integrated into the brain parenchyma, appeared morphologically similar to astrocytes, and expressed glial fibrillary acidic protein (GFAP). A very small percentage of the transplanted MSC expressed neurofilament, suggesting that neuronal differentiation may have occurred as well. In other studies, adherent cells from rat bone marrow, which are morphologically similar to MSC, were transplanted to treat functional deficits resulting from traumatic injury to brain or spinal cord [14,15]. When stained for lineage-specific markers, 4.9% of transplanted MSC expressed the neuronal protein NeuN and 6.7% expressed the astrocytic marker GFAP. In these studies, mesenchymal stem cell transplantation improved the recovery of experimental subjects in a statistically significant manner based on various tests of neurologic function. Similarly, when stroke was induced in rats by occlusion of the middle cerebral artery, mesenchymal stem cell infusion into the internal carotid artery or into the tail vein significantly improved functional recovery. Immunostaining of brain sections revealed that MSC migrated preferentially to the ischemic site, where 5% to 10% of
D.R. Martin et al./Experimental Hematology 30 (2002) 879–886
transplanted MSC expressed GFAP while 1% to 2% expressed the neuronal markers NeuN or MAP-2 (microtubule associated protein-2) [12,13]. The purpose of this initial study was to isolate MSC from a feline source and to verify the identity of these cells by comparison to MSC from other species. The following factors provide convincing evidence that the cells of this study are feline MSC: 1) published methods for MSC isolation from other species were used to isolate the cells in this study, 2) the cells are very similar to their rodent and human counterparts based on morphology, growth traits, frequency rates in bone marrow, and cell-surface antigen profile, and 3) the cells are multipotential, capable of differentiating to adipocytes, osteocytes, and neurons in vitro. Although the chondrogenic potential of MSC from other species is well established [2,3], no attempt was made to differentiate feline MSC to chondrocytes. Additionally, no attempt was made to evaluate the differentiation fates assumed by feline MSC after transplantation. Since feline MSC do not proliferate adequately when plated at low cell densities (3, 10, 30, and 100 MSC/cm2 at passage 1), cloning of individual feline MSC for transplantation experiments may be possible only after immortalization. A more detailed analysis of the differentiation repertoire of feline MSC will be the focus of future studies. In conclusion, mesenchymal stem cells possess several properties that make them attractive tools for the study of development, physiology, and disease. MSC are harvested easily from bone marrow with standardized isolation techniques and have a high proliferative potential in culture. Also, MSC possess a characteristic morphology and cellsurface antigen profile, which allow accurate confirmation of cellular identity. Importantly, MSC are capable of differentiation to a wide variety of cell types and have been shown not only to survive after transplantation but also to migrate to an area of brain injury. The phenotype of feline MSC, as reported herein, concurs with and strengthens the mesenchymal stem cell prototype developed from several other species. Isolation of feline MSC also makes possible the application of stem cell technology to a variety of valuable feline experimental models.
4. 5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
Acknowledgments This work was supported in part by grant R01 JL69139-01 issued by the National Heart, Lung and Blood Institute, National Institutes of Health, Bethesda, MD, USA.
20.
21.
22.
References 1. Donovan PJ, Gearhart J (2001) The end of the beginning for pluripotent stem cells. Nature 414:92 2. Pittenger MF, Mackay AM, Beck SC, et al. (1999) Multilineage potential of adult human mesenchymal stem cells. Science 284:143 3. Kopen GC, Prockop DJ, Phinney DG (1999) Marrow stromal cells migrate throughout forebrain and cerebellum, and they differentiate into
23.
24.
885
astrocytes after injection into neonatal mouse brains. Proc Natl Acad Sci U S A 96:10711 Gerson SL (1999) Mesenchymal stem cells: no longer second class marrow citizens. Nat Med 5:262 Jaiswal N, Haynesworth SE, Caplan AI, Bruder SP (1997) Osteogenic differentiation of purified, culture-expanded human mesenchymal stem cells in vitro. J Cell Biochem 64:295 Negishi Y, Kudo A, Obinata A, et al. (2000) Multipotency of a bone marrow stromal cell line, TBR31-2, established from ts-SV40 T antigen gene transgenic mice. Biochem Biophys Res Commun 268:450 Wakitani S, Saito T, Caplan AI (1995) Myogenic cells derived from rat bone marrow mesenchymal stem cells exposed to 5-azacytidine. Muscle Nerve 18:1417 Woodbury D, Schwarz EJ, Prockop DJ, Black IB (2000) Adult rat and human bone marrow stromal cells differentiate into neurons. J Neurosci Res 61:364 Sanchez-Ramos J, Song S, Cardozo-Pelaez F, et al. (2000) Adult bone marrow stromal cells differentiate into neural cells in vitro. Exp Neurol 164:247 Banu N, Reitter J, Vonschild E, Biber K, Rosenzweig M, Pykett M (2000) Neuronal, mesenchymal and hematopoietic cells derived from CD34, lin cells from adult bone marrow. Exp Hematol 28:46 Deng W, Obrocka M, Fischer I, Prockop DJ (2001) In vitro differentiation of human marrow stromal cells into early progenitors of neural cells by conditions that increase intracellular cyclic AMP. Biochem Biophys Res Commun 282:148 Li Y, Chen J, Wang L, Lu M, Chopp M (2001) Treatment of stroke in rat with intracarotid administration of marrow stromal cells. Neurology 56:1666 Chen J, Li Y, Wang L, Lu M, Zhang X, Chopp M (2001) Therapeutic benefit of intracerebral transplantation of bone marrow stromal cells after cerebral ischemia in rats. J Neurol Sci 189:49 Lu D, Mahmood A, Wang L, Li Y, Lu M, Chopp M (2001) Adult bone marrow stromal cells administered intravenously to rats after traumatic brain injury migrate into brain and improve neurological outcome. Neuroreport 12:559 Chopp M, Zhang XH, Li Y, et al. (2000) Spinal cord injury in rat: treatment with bone marrow stromal cell transplantation. Neuroreport 11:3001 Phinney DG, Kopen G, Isaacson RL, Prockop DJ (1999) Plastic adherent stromal cells from the bone marrow of commonly used strains of inbred mice: variations in yield, growth, and differentiation. J Cell Biochem 72:570 Haynesworth SE, Goshima J, Goldberg VM, Caplan AI (1992) Characterization of cells with osteogenic potential from human marrow. Bone 13:81 Friedenstein AJ, Gorskaja JF, Kulagina NN (1976) Fibroblast precursors in normal and irradiated mouse hematopoietic organs. Exp Hematol 4:267 Conget PA, Minguell JJ (1999) Phenotypical and functional properties of human bone marrow mesenchymal progenitor cells. J Cell Physiol 181:67 Kadiyala S, Young RG, Thiede MA, Bruder SP (1997) Culture expanded canine mesenchymal stem cells possess osteochondrogenic potential in vivo and in vitro. Cell Transplant 6:125 Haskins M, Baker HJ, Birkenmeier E, et al. (1991) Transplantation in animal model systems. In RJ Desnick (ed.): Treatment of Genetic Diseases. New York: Churchill Livingstone, p. 183 Griffin B, Baker HJ (2001) Domestic cats as laboratory animals. In JG Fox, LC Anderson, FM Lowe, FW Quimby (eds.): Laboratory Animal Medicine, 2nd edition. New York: Academic Press, p. 460 Menotti-Raymond M, David VA, Lyons LA, et al. (1999) A genetic linkage map of microsatellites in the domestic cat (Felis catus). Genomics 57:9 Migaki G (1982) Section VIII. Compendium of inherited metabolic diseases in animals. In: R Desnick, D Patterson, DG Scarpeooi (eds.):
886
25. 26.
27. 28.
29.
30.
D.R. Martin et al./Experimental Hematology 30 (2002) 879–886 Animal Models of Inherited Metabolic Disease, 1st edition. New York: A.R. Liss, p. 473 Linenberger ML, Abkowitz JL (1995) Haematological disorders associated with feline retrovirus infections. Baillieres Clin Haematol 8:73 Rohn JL, Gwynn SR, Lauring AS, Linenberger ML, Overbaugh J (1996) Viral genetic variation, AIDS, and the multistep nature of carcinogenesis: the feline leukemia virus model. Leukemia 10:1867 Willett BJ, Flynn JN, Hosie MJ (1997) FIV infection of the domestic cat: an animal model for AIDS. Immunol Today 18:182 Conget PA, Minguell JJ (2000) Adenoviral-mediated gene transfer into ex vivo expanded human bone marrow mesenchymal progenitor cells. Exp Hematol 28:382 Baker HJ, Reynolds GD, Walkley SU, Cox NR, Baker GH (1979) The gangliosidoses: comparative features and research applications. Veterinary Pathology 16:635 Baker HJ, Walkley SU, Rattazzi MC, Singer HS, Watson HL, Wood
31.
32.
33.
34.
PA (1982) Feline gangliosidoses as models of human lysosomal storage diseases. Progress in Clinical & Biological Research 94:203 Baker HJ, Walkley SU, Rattazzi MC, Singer HS, Watson HL, Wood PA (1982) Feline gangliosidoses as models of human lysosomal stroage diseases. In RJ Desnick, DF Patterson, DG Scarpelli (eds.): Animal Models of Inherited Metabolic Diseases. New York: Alan R. Liss, p. 203 Baker HJ Jr., Lindsey JR, McKhann GM, Farrell DF (1971) Neuronal GM1 gangliosidosis in a Siamese cat with -galactosidase deficiency. Science 174:838 Bruder SP, Jaiswal N, Haynesworth SE (1997) Growth kinetics, selfrenewal, and the osteogenic potential of purified human mesenchymal stem cells during extensive subcultivation and following cryopreservation. J Cell Biochem 64:278 Kramer JW, Hoffmann WE (1997) Clinical enzymology. In JJ Kaneko, JW Harvey, ML Bruss (eds.): Clinical Biochemistry of Domestic Animals, 5th ed. San Diego: Academic Press, p. 303
Isolation and characterization of multipotential mesenchymal stem cells from feline bone marrow Douglas R. Martin, Nancy R. Cox, Terri L. Hathcock, Glenn P. Niemeyer, and Henry J. Baker The Scott-Ritchey Research Center, College of Veterinary Medicine, Auburn University, Auburn, Ala., USA (Received 10 January 2002; revised 22 March 2002; accepted 18 April 2002)
Objective. Although several types of stem cells have been isolated from rodent and human tissues, very few data exist on stem cell isolation from nonrodent animals, which seriously limits the advancement of stem cell biology and its ultimate translation to human clinical applications. Domestic cats are used frequently in biomedical research and are the preferred species for studies of normal physiology and disease, particularly in neuroscience. Therefore, the objective of this study was to characterize mesenchymal stem cells (MSC) from feline bone marrow for use in research on the application of stem cells to human health problems for which cats are the preferred model. Methods. Mesenchymal stem cells from feline bone marrow were isolated by standard methodology developed for other species and characterized according to morphology, growth traits, cell-surface antigen profile, and differentiation repertoire in vitro. Results. Feline mesenchymal stem cells exhibit a fibroblast-like morphology with bipolar or polygonal cell bodies and possess a cell-surface antigen profile similar to their rodent and human counterparts. Feline MSC exist at a frequency of 1 in 3.8 105 bone marrow mononuclear cells and are capable of differentiation to adipocytic, osteocytic, and neuronal phenotypes when exposed to appropriate induction media. Conclusions. Mesenchymal stem cells isolated from feline bone marrow possess several traits typical of MSC from other species. Characterization of feline mesenchymal stem cells will facilitate future studies of stem cell biology and therapeutics for which the domestic cat is an indispensable model. © 2002 International Society for Experimental Hematology. Published by Elsevier Science Inc.
The defining characteristics of stem cells are self-renewal and the ability to differentiate into one or more specialized cell types [1]. Traditionally, stem cells have been divided into two major groups: 1) embryonic stem cells, derived from the inner cell mass of fertilized ova which can differentiate into any cell type, and 2) tissue-specific stem cells, derived from specific organs. Tissue-specific stem cells once were believed to have differentiation capacity limited to the organ of origin but now are recognized to be capable of differentiation into cells of other tissues. For example, mesenchymal stem cells (MSC), also known as bone marrow stromal cells, are capable of differentiation in vitro to marrow and non-marrow cell types, including: adipocytes,
Offprint requests to: Henry J. Baker, D.V.M., Professor and Director, The Scott-Ritchey Research Center, College of Veterinary Medicine, Auburn University, AL 36849; E-mail: [email protected]
chondrocytes, osteocytes [2–5], myocytes [6,7], astrocytes, and neurons [8–11]. Differentiation of MSC to a variety of cell types, including astrocytes and neurons, has been demonstrated in vivo as well [3,12–15]. Although present in very small numbers in the bone marrow [16–18], MSC are capable of substantial proliferation and expansion in culture [2,17,19]. Undifferentiated MSC exhibit a fibroblast-like morphology and a characteristic pattern of cell-surface antigens [2,8] as well. The criteria listed above, including the ability to differentiate to many cell types, have been used to define a prototypic mesenchymal stem cell phenotype which is consistent among a variety of species, including mouse [16], rat [8], dog [20], and human [2]. The majority of research on stem cells, including MSC, has been done on cells derived from rodent and human tissues. However, investigation of stem cells from nonrodent mammals is essential for progress to be made toward the ultimate application of stem cells to human health problems.
0301-472X/02 $–see front matter. Copyright © 2002 International Society for Experimental Hematology. Published by Elsevier Science Inc. PII S0301-472X(02)0 0 8 6 4 - 0
880
D.R. Martin et al./Experimental Hematology 30 (2002) 879–886
The domestic cat provides a number of important advantages for biomedical research, particularly stem cell transplantation in fetal and postnatal animals. There is substantial knowledge about both prenatal and postnatal hematopoiesis; fetal, neonatal, and adult immunology; and bone marrow transplantation in domestic cats. Although inbred rodents provide an experimental advantage for stem cell transplantation because they are histocompatible, cats heterologous for histocompatibility loci more closely mimic the challenges and limitations that are likely to be encountered in human patients receiving allogeneic cell transplants [21]. Cats have been used extensively for studies of normal physiology and disease, particularly of the nervous system, which presents a critically important opportunity for research on stem cell transplantation therapy of neurological disorders [22]. Cats are models for approximately 60 inherited human diseases [22–24], including lysosomal storage diseases, diabetes, heart muscle disorders, polycystic kidney disease, and retinal atrophy. Also, cats are important models of acquired immunodeficiency diseases [25–27]. Because several feline viral diseases adversely affect bone marrow as well as other organ systems in the cat, a substantial base of information is available concerning cell markers, particularly of the hematopoietic and immune system of this species, which provides opportunities for experimental assessment of the differentiation fate of transplanted stem cells. In addition, cats are maintained easily in laboratory breeding colonies, produce 4 to 6 progeny per breeding, and have multiple breeding cycles per year with a short gestation of 63 1 days. Cats respond well to intrauterine surgery and are large enough to permit frequent sampling of tissues and body fluids [22]. The purpose of this study was to isolate and characterize MSC from normal domestic cats as the essential first step toward their use in feline models of development, physiology, and disease.
Materials and methods Materials All tissue culture media, supplements, and differentiation reagents were purchased from Sigma (St. Louis, MO, USA) unless otherwise noted. Percoll was purchased from Amersham Pharmacia (Piscataway, NJ, USA). All antibodies for flow cytometry were shown by the manufacturer to cross react with feline cells. Antibodies to CD9 and CD45 were purchased from Serotec (Raleigh, NC, USA), while all other antibodies were purchased from VMRD (Pullman, WA, USA). Antibodies to neuronal proteins used for immunohistochemistry were as follows: mouse anti-pig neurofilament-M (160 kDa, Accurate Chemical, Westbury, NY, USA), rabbit anti-human trkA (nerve growth factor receptor, Santa Cruz Biotechnology, Santa Cruz, CA, USA) and mouse anti-human III-tubulin (Sigma). Isolation of mesenchymal stem cells from feline marrow Feline bone marrow was harvested antemortem by bone marrow aspiration from the greater trochanter of the femur or greater tubercle of the humerus, or at necropsy by flushing the shaft of a femur. Bone marrow was collected into 1–5 volumes Iscove’s modified Dulbecco’s medium (IMDM) containing 200 units/mL heparin, filtered (100 m), and centrifuged at 900g to pellet cells. After two rinses in phosphate-buffered saline (PBS), 108 cells were loaded onto 12.5-mL Percoll (1.073 g/mL in 0.15 M NaCl) and spun at 1100g for 30 minutes. Mononuclear cells were collected at the Percoll interface, rinsed twice in PBS, and seeded at 2 105/cm2 in Dulbecco’s Modified Eagle’s Medium (DMEM) (1 g/L glucose) with 10% fetal bovine serum (FBS; Hyclone, Logan, UT, USA) (Passage 0). Three lots of FBS were screened for their ability to support feline mesenchymal stem cell proliferation for 2 to 3 passages, followed by successful differentiation to the neuronal morphology. The lot of FBS that was optimal for proliferation also allowed induction to a neuronal phenotype and was chosen for subsequent studies. Nonadherent cells were removed and media replaced 1 to 3 days after initial plating. After 7 to 12 days in culture, isolated colonies of mesenchymal stem cells were apparent.
Table 1. Cell-surface antigen profile of feline vs. rodent/human MSC Cell-surface antigen (lineage expression) CD4 (M, monocyte, T helper) CD8 (T cytotoxic/suppressor cell) CD9 (myeloid) CD13 (granulocyte, monocyte) CD14 (M, monocyte, some granulocytes) CD18 (leukocyte) CD41/61 (megakaryocyte, platelet) CD44 (brain, erythrocyte, leukocyte, platelet) CD45 (leukocyte) MHC I (almost all nucleated cells) MHC II (antigen-presenting cells) *No data available for CD8 or CD13 expression on rodent or human MSC. † MSC tested with anti-CD61, not anti-CD41/61. M macrophage.
Feline MSC
Rodent/human MSC [2,8]
ND* ND* †
D.R. Martin et al./Experimental Hematology 30 (2002) 879–886
These cells were trypsinized and replated at 8000/cm2, and fresh medium was added every 3 to 4 days (Passage 1). At passage 2, media was changed to DMEM (4.5 g/L glucose) with 20% fetal bovine serum (Dr. D. Woodbury, Robert Wood Johnson Medical School, Piscataway, NJ, personal communication). Flow cytometric determination of cell-surface antigen profile Feline MSC were trypsinized (0.05% trypsin-EDTA), resuspended in staining media (1% bovine serum albumin, 0.2% sodium azide in PBS), and stained on ice according to the manufacturer’s recommendations with the monoclonal antibodies listed in Table 1. Positive cells were detected with a 1:100 dilution of FITC-conjugated goat anti-mouse IgG (Jackson ImmunoResearch, West Grove, PA, USA) on a Coulter Epics Elite fluorescence-activated cell sorter. Differentiation protocols Adipocytic differentiation [3] was achieved by allowing cells to remain at confluency for 3 to 7 days, followed by incubation in Alpha Minimal Essential Medium ( -MEM) supplemented with 10% FBS, 10% normal rabbit serum, 10 nM dexamethasone, 5 g/mL insulin, and 50 M 5,8,11,14-eicosatetraynoic acid. After 3 days, cells were cytocentrifuged and stained for lipid droplets with Oil Red O (0.3% in isopropanol with 0.4% dextrin). The osteocytic phenotype [28] was induced by seeding cells at 6000–8000/cm2, followed 1 day later by incubation in DMEM containing 10% FBS, 100 mM dexamethasone, 10 nM -glycerophosphate, and 0.25 mM L-ascorbic acid. Differentiation proceeded for 3 weeks with media changes twice weekly. Induced cells were stained for calcium with Alizarin Red S (10%, pH 4.2) or tested for alkaline phosphatase activity (Sigma Diagnostic kit #104). The long-term neuronal differentiation protocol of Woodbury et al. [8] was modified by the substitution of -MEM for DMEM in the induction media. Specifically, cells were expanded to 70% to 80% confluency in DMEM with 10% FBS, and basic fibroblast growth factor (10 ng/mL) was added for 24 hours. Thereafter, MSC were incubated in neuronal induction media ( -MEM, 2% dimethylsulfoxide [DMSO], 200 M butylated hydroxyanisole, 25 mM potassium chloride, 2 mM valproic acid, 10 M forskolin, 1 M hydrocortisone, 5 g/mL insulin, and 2 mM L-glutamine with no serum) [8]. After 24 hours of neuronal induction, cells were stained for expression of the neuron-specific proteins neurofilament M (NF-M), nerve growth factor receptor (trkA), or III-tubulin, as described below.
881
of morphology. Phenotypically, mesenchymal stem cell colonies appeared as fibroblast-like cells, bipolar or polygonal in morphology as compared with contaminating monocyte colonies, which appeared as large, round cells with irregular boundaries and large nuclei.
Results Isolation of feline MSC In initial experiments, the total amount of marrow cells collected per bone (humerus or femur) ranged from 4.25 108 to 1.13 109 (mean 6.7 108 cells/bone, n 8 cats). After density gradient centrifugation, the bone marrow mononuclear cell (BMMC) population collected at the Percoll interface represented 10.6% to 23.4% of the total bone marrow isolate (mean 16.1%, n 8 cats). Feline MSC grew as isolated colonies 7 to 10 days after initial plating and were trypsinized for further expansion. After replating, cells were fibroblast-like, appearing polygonal or spindleshaped with long processes (Fig. 1). Morphologically, feline MSC appeared very similar to their rodent and human counterparts [2,8]. Mesenchymal stem cell isolation experiments were conducted on at least 15 normal cats, with each marrow sample yielding similar results in terms of colony and cellular morphology. Enumeration and expansion of MSC in feline bone marrow By comparing the number of mesenchymal stem cell colonies to the total number of bone marrow mononuclear cells seeded after Percoll centrifugation, the frequency of MSC was determined to range from 1 in 4.7 104 to 1 in 5.9 105, which compares favorably to data generated from mouse and human studies [16–18] (Table 2). Initial attempts
Immunohistochemistry of neuronally differentiated MSC Cells induced to the neuronal phenotype were fixed in 4% paraformaldehyde for 12 minutes, rinsed with PBS, and blocked in 10% normal goat serum for 1 hour. Fixed cells were stained in 1.5% goat serum with neurofilament M (1:40–1:200), trkA (1:200), or III-tubulin (1:200) at 4 C overnight. Positive cells were visualized with a FITC- or TRITC-conjugated secondary antibody (Jackson ImmunoResearch) and counterstained with propidium iodide or DAPI (Vector Laboratories, Burlingame, CA, USA). Enumeration of MSC in feline bone marrow To evaluate the percentage of MSC in feline bone marrow, mononuclear cells isolated by Percoll centrifugation were plated at 2 105 cells/cm2 in 6-well plates and incubated for 10 days. Isolated colonies were stained with methylene blue (0.167% in methanol) and enumerated, with mesenchymal stem cell colonies distinguished from monocyte contamination by microscopic evaluation
Figure 1. Morphological characterization of feline bone marrow stromal cells. In growth media (DMEM 20% FBS), feline MSC are long, flattened cells of a fibroblastic morphology. Scale bar 25 m.
882
D.R. Martin et al./Experimental Hematology 30 (2002) 879–886
Table 2. Frequency of MSC in bone marrow mononuclear cell fraction from different species Source of MSC
Frequency in BMMC
Feline (CR643)* Feline (N359)* Feline (N367)* Human [17] Mouse [16] Mouse [18]
1 in 5.9 105 1 in 4.9 105 1 in 4.7 104 1 in 1 105–1 106 1 in 3.3 105–3.3 106 1 in 2.2 104–3.3 105
*Numbers represent individual cats from which marrow was harvested.
to grow feline MSC at low cell densities (3, 10, 30, and 100 MSC/cm2 at passage 1) were unsuccessful. To assess the expansion capacity of feline MSC, initial cultures were generated by seeding a 75-cm2 flask with 1.5 107 bone marrow mononuclear cells. In an average culture, the frequency of MSC is 1 in 3.8 105 BMMC (mean of 3 cats reported in Table 2). Therefore, an initial culture of 1.5 107 BMMC contains approximately 40 MSC. Initial colonies of MSC were counted and found to have expanded to 5.3 1.2 105 (mean standard deviation, n 3) cells after 9 to12 days, a 13,250 3000-fold increase. After trypsinization of the initial cultures, MSC expanded approximately threefold for the first 2 passages and twofold thereafter (data not shown). Most cultures were proliferative to at least passage 6, indicating that an average mesenchymal stem cell can proliferate to a minimum of 7 107 MSC in vitro. The expansion potential of feline MSC may be considerably greater, however, since rigorous longevity studies to assess the maximum attainable passage number have not been performed. Cell-surface antigen profile Feline mesenchymal stem cell cell-surface antigen profile was ascertained by staining with feline-specific monoclonal antibodies followed by flow cytometry as shown in Table 1 and Figure 2. Feline MSC possess an array of cell-surface antigens similar to MSC from rodents or humans. Feline MSC are strongly positive for CD9, CD44 (Pgp-1 or HCAM), and MHC-I, while negative for CD45 (leukocyte common antigen). Differentiation of feline MSC To determine whether feline mesenchymal stem cells, like MSC from other species, are capable of differentiation to multiple cell types, feline cells were placed into induction media specific for the generation of adipocytes [3], osteocytes [28], or neurons [8]. To generate adipocytes, feline MSC at passages 3 through 5 were grown to confluency and incubated for 3 days prior to addition of adipocyte induction media. Sixty hours after induction, the confluent monolayer detached from the tissue culture flask and floated as a homogeneous mass in the culture media (n 2). Very few cells ( 1%) remained adher-
Figure 2. Flow cytometric determination of cell-surface antigen profile for feline MSC. Feline bone marrow stromal cells were stained with felinespecific antibodies to a variety of cell-surface antigens. The percentage of cells positive for each antibody is listed in the corresponding histogram. MSC from 4 separate cats were tested with similar results. Representative results from 1 cat are shown.
ent. The detached monolayer was dissociated into a single cell suspension with a serological pipette and cell number/ viability were assessed by trypan blue exclusion. The majority of the cells ( 95%) remained viable. Cytocentrifugation of the cells followed by Oil Red O staining revealed that 92.5% of feline MSC produced lipid droplets after incubation in adipogenic media compared to only 5.5% of control cells in normal growth media (Fig. 3). Approximately 700 cells from at least 5 microscopic fields were scored for acquisition of an adipocytic phenotype as detected by staining with Oil Red O [3]. Subsequent experiments (n 3), in which feline MSC were seeded at lower densities and not allowed to reach confluency prior to induction, resulted in an adherent monolayer of lipid-producing cells. It was estimated that at least 70% of the adipocyte-induced cells stained positively for Oil Red O, although rigorous determination of percentages was not possible due to the confluency of the monolayer at the time of staining (data not shown). When grown in osteocyte-specific induction media, feline MSC grew as tightly packed, spindle-shaped cells for approximately 2 weeks, at which time clusters of round, calcium-producing cells formed above the tightly packed monolayer. Calcium formation was demonstrated by staining with Alizarin Red S, a marker for the osteocytic phenotype (Fig. 4). Alkaline phosphatase activity was undetectable when tested on undifferentiated or osteocyte-induced MSC from three separate cats on cumulative days 1, 4, 6, 9, 11, 13, 16, and 18 postinduction. However, a positive control cell line (canine osteosarcoma #C0585-6, Dr. Lauren Wolfe, Auburn University College of Veterinary Medicine) exhibited robust alkaline phosphatase activity (data not shown).
D.R. Martin et al./Experimental Hematology 30 (2002) 879–886
Figure 3. Adipocyte differentiation of feline MSC. After incubation at confluency for 3 days followed by the addition of adipocyte differentiation media for 3 days, feline MSC were cytocentrifuged, fixed in 10% neutral buffered formalin for 25 minutes, and stained in Oil Red O to reveal lipid vacuoles. MSC at passage number 3 were used for this experiment. (A) Uninduced control MSC. (B) MSC induced to the adipocyte lineage. Scale bar 10 m.
The ability of feline MSC to differentiate to a cell type of disparate dermal origin was demonstrated by induction to the neuronal phenotype. When incubated in neuronal induction media, 75% to 85% of feline MSC rapidly (less than 1 hour) acquired a neurotypic morphology, with multipolar, rounded cell bodies and long, narrow processes that contacted other cells in the culture vessel (Fig. 5). The neuronal induction protocol was repeated at least 15 times on MSC from at least five different cats. MSC were induced successfully at passage numbers ranging from 1 to 6. Neuronal induction of MSC at higher passage numbers has not been attempted. In addition to acquiring a neuronal morphology, feline MSC were stained 24 hours postinduction and found to express neuron-specific proteins such as neurofilament M (160 kDa), trkA (nerve growth factor receptor), and III-tu-
Figure 4. Osteocyte differentiation of feline MSC. After a 3-week incubation in osteocyte induction media, feline MSC stained positively for calcium with Alizarin Red S, indicating osteocytic differentiation. MSC from 3 separate cats were used for osteocyte induction experiments. Four independent experiments were conducted with MSC at passages 2 through 4, and one representative result is shown. (A) Uninduced feline MSC. (B) Feline MSC induced to the osteocytic phenotype. Scale bar 20 m.
883
Figure 5. Morphological characterization and neuronal differentiation of feline MSC. (A) In growth media, feline MSC appear as long, flattened cells of a fibroblastic morphology. (B,C) After a preinduction treatment with basic fibroblast growth factor followed by addition of neuronal induction media, 75%–85% of feline MSC change to a neuronal morphology with long, thin processes and round cell bodies. Experiments were performed 4 times, and cells were counted from 10 microscopic fields. Scale bars (A,B) 25 m. Scale bar (C) 10 m.
bulin at levels that appeared to be much greater than control cells (Fig. 6). Discussion We have a long-standing interest in therapeutic strategies for the gangliosidoses, a subset of lysosomal diseases that are fatal neuronopathic disorders of humans and several other species. Mesenchymal stem cells are attractive candidates for stem cell transplantation–based therapy for the gangliosidoses, in which the primary affected organs are the nervous system, thymus, and liver. The feline gangliosidoses are authentic replicas of the human counterparts [29– 32], and the characterization of feline MSC as reported here provides the essential foundation for testing the fate, distribution, and therapeutic benefit of MSC transplantation for these diseases. Based on morphology, cell frequency rate and cell-surface antigen profile, mesenchymal stem cells from feline bone marrow are very similar to MSC from rodent or human sources. When isolated by standard methodology developed for other species [2,8], individual colonies of feline MSC become apparent by 1 week after plating. Feline MSC comprise a minor percentage of the total bone marrow mononuclear cell fraction, similar to that reported for other species (approximately 1 in 4 105 cells, Table 2). Morphologically, undifferentiated feline MSC are fibroblastlike with spindle-shaped or polygonal cell bodies (Fig. 1), similar to their rodent or human counterparts. It is estimated that one feline mesenchymal stem cell can proliferate to a minimum of 7 107 MSC in six passages. The expansion
884
D.R. Martin et al./Experimental Hematology 30 (2002) 879–886
Figure 6. Expression of neuronal markers after differentiation of feline MSC. Feline MSC were pre-induced in basic fibroblast growth factor, then induced to the neuronal phenotype with induction media. Twenty-four hours later, cells exhibiting the neuronal phenotype stained positively for (A) the 160-kDa neurofilament M, (B) trkA, the nerve growth factor receptor, and (C) III tubulin. Cells that did not change morphologically served as internal negative controls for staining (arrows). Scale bars 50 m.
potential of human MSC has been estimated to range from 5.5 108 to 1.2 109 in studies that propagated MSC for 10 to 25 passages [19,33]. Although initial studies suggest that feline MSC are not capable of attaining such high passage numbers, rigorous longevity studies have not been performed to date. Also, preliminary data indicate that growth factor supplementation can enhance the proliferation of feline MSC (data not shown). Therefore, a firm conclusion about the expansion potential of feline MSC awaits further investigation. Feline MSC are CD9, CD44, MHC-I, CD4, CD45, and MHC-II, a cell-surface antigen profile similar to MSC from rodents and humans [2,8] (Table 1, Fig. 2). Although rodent/human MSC are positive for CD61 (integrin III, vitronectin receptor ), the feline antibody used in this study was raised against the CD41/61 complex and may recognize a CD41- or complex-specific epitope. Therefore, until a CD61-specific antibody for feline cells is obtained, no conclusion can be drawn about the presence of this cellsurface antigen on feline cells. Also, no data were available regarding the expression of CD8 or CD13 on rodent/human MSC, although feline mesenchymal stem cells are negative for both CD8 and CD13, as shown in these experiments. In addition to the general characteristics described above, another defining feature of mesenchymal stem cells is multipotentiality, or the ability to acquire multiple cellular phenotypes when exposed to the appropriate stimuli. The ability of MSC to differentiate to cells of a mesodermal origin has been well documented and includes adipocyte, osteocyte, chondrocyte [2–5,16], and myocyte [6,7,16] phenotypes. Our studies support and extend these observations to include feline mesenchymal stem cell differentiation to adipocyte and osteocyte phenotypes (Figs. 3 and 4), as indicated by lipid accumulation and calcium deposition, respectively. Alkaline phosphatase activity often is used as an early marker of osteocytic differentiation. In human and mouse MSC, alkaline phosphatase activity is present at baseline levels in undifferentiated cells and at significantly increased
levels in osteocyte-induced cells prior to calcium deposition [2,5,16]. However, no alkaline phosphatase activity was detected in feline MSC either prior to or following induction to the osteocytic phenotype, a phenomenon that may be explained by the unusually short half-life of the feline enzyme. Since elevated alkaline phosphatase activity is a diagnostic indicator of liver or bone disease in both humans and domestic animals, it is known that the serum half-life of feline alkaline phosphatase is unusually short ([34] and Dr. J. Spano, Auburn University College of Veterinary Medicine, personal communication). Because we detected robust alkaline phosphatase activity in a canine osteosarcoma cell line used as a positive control, we believe that feline alkaline phosphatase simply is undetectable with our assay system. A substantial body of evidence is beginning to emerge that suggests that rodent and human MSC are capable not only of differentiation to mesodermal cell lineages but also of “transdifferentiation” to cell types derived from ectoderm. For example, with two different induction protocols, Sanchez-Ramos et al. [9] and Woodbury et al. [8] generated cells of a neuronal morphology that expressed neuron-specific/selective proteins such as neurofilament M, tau, neuron-specific enolase, trkA, nestin, and neuron-specific nuclear protein (NeuN). Similarly, when treated with neuronal induction media, 75% to 85% of feline MSC rapidly exhibited a neuronal morphology with round, multipolar or bipolar cell bodies and long, thin processes that contacted other cells in the culture dish (Fig. 5). Neuronally induced feline MSC expressed NF-M, trkA, and III-tubulin (a neuronspecific isoform of tubulin), at levels greater than control MSC (Fig. 6). Additional support for the transdifferentiative capacity of MSC comes from transplantation studies in which mesenchymal stem cells were injected into experimental animals. For example, Kopen et al. [3] found that murine MSC injected into the lateral ventricle of newborn mice integrated into the brain parenchyma, appeared morphologically similar to astrocytes, and expressed glial fibrillary acidic protein (GFAP). A very small percentage of the transplanted MSC expressed neurofilament, suggesting that neuronal differentiation may have occurred as well. In other studies, adherent cells from rat bone marrow, which are morphologically similar to MSC, were transplanted to treat functional deficits resulting from traumatic injury to brain or spinal cord [14,15]. When stained for lineage-specific markers, 4.9% of transplanted MSC expressed the neuronal protein NeuN and 6.7% expressed the astrocytic marker GFAP. In these studies, mesenchymal stem cell transplantation improved the recovery of experimental subjects in a statistically significant manner based on various tests of neurologic function. Similarly, when stroke was induced in rats by occlusion of the middle cerebral artery, mesenchymal stem cell infusion into the internal carotid artery or into the tail vein significantly improved functional recovery. Immunostaining of brain sections revealed that MSC migrated preferentially to the ischemic site, where 5% to 10% of
D.R. Martin et al./Experimental Hematology 30 (2002) 879–886
transplanted MSC expressed GFAP while 1% to 2% expressed the neuronal markers NeuN or MAP-2 (microtubule associated protein-2) [12,13]. The purpose of this initial study was to isolate MSC from a feline source and to verify the identity of these cells by comparison to MSC from other species. The following factors provide convincing evidence that the cells of this study are feline MSC: 1) published methods for MSC isolation from other species were used to isolate the cells in this study, 2) the cells are very similar to their rodent and human counterparts based on morphology, growth traits, frequency rates in bone marrow, and cell-surface antigen profile, and 3) the cells are multipotential, capable of differentiating to adipocytes, osteocytes, and neurons in vitro. Although the chondrogenic potential of MSC from other species is well established [2,3], no attempt was made to differentiate feline MSC to chondrocytes. Additionally, no attempt was made to evaluate the differentiation fates assumed by feline MSC after transplantation. Since feline MSC do not proliferate adequately when plated at low cell densities (3, 10, 30, and 100 MSC/cm2 at passage 1), cloning of individual feline MSC for transplantation experiments may be possible only after immortalization. A more detailed analysis of the differentiation repertoire of feline MSC will be the focus of future studies. In conclusion, mesenchymal stem cells possess several properties that make them attractive tools for the study of development, physiology, and disease. MSC are harvested easily from bone marrow with standardized isolation techniques and have a high proliferative potential in culture. Also, MSC possess a characteristic morphology and cellsurface antigen profile, which allow accurate confirmation of cellular identity. Importantly, MSC are capable of differentiation to a wide variety of cell types and have been shown not only to survive after transplantation but also to migrate to an area of brain injury. The phenotype of feline MSC, as reported herein, concurs with and strengthens the mesenchymal stem cell prototype developed from several other species. Isolation of feline MSC also makes possible the application of stem cell technology to a variety of valuable feline experimental models.
4. 5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
Acknowledgments This work was supported in part by grant R01 JL69139-01 issued by the National Heart, Lung and Blood Institute, National Institutes of Health, Bethesda, MD, USA.
20.
21.
22.
References 1. Donovan PJ, Gearhart J (2001) The end of the beginning for pluripotent stem cells. Nature 414:92 2. Pittenger MF, Mackay AM, Beck SC, et al. (1999) Multilineage potential of adult human mesenchymal stem cells. Science 284:143 3. Kopen GC, Prockop DJ, Phinney DG (1999) Marrow stromal cells migrate throughout forebrain and cerebellum, and they differentiate into
23.
24.
885
astrocytes after injection into neonatal mouse brains. Proc Natl Acad Sci U S A 96:10711 Gerson SL (1999) Mesenchymal stem cells: no longer second class marrow citizens. Nat Med 5:262 Jaiswal N, Haynesworth SE, Caplan AI, Bruder SP (1997) Osteogenic differentiation of purified, culture-expanded human mesenchymal stem cells in vitro. J Cell Biochem 64:295 Negishi Y, Kudo A, Obinata A, et al. (2000) Multipotency of a bone marrow stromal cell line, TBR31-2, established from ts-SV40 T antigen gene transgenic mice. Biochem Biophys Res Commun 268:450 Wakitani S, Saito T, Caplan AI (1995) Myogenic cells derived from rat bone marrow mesenchymal stem cells exposed to 5-azacytidine. Muscle Nerve 18:1417 Woodbury D, Schwarz EJ, Prockop DJ, Black IB (2000) Adult rat and human bone marrow stromal cells differentiate into neurons. J Neurosci Res 61:364 Sanchez-Ramos J, Song S, Cardozo-Pelaez F, et al. (2000) Adult bone marrow stromal cells differentiate into neural cells in vitro. Exp Neurol 164:247 Banu N, Reitter J, Vonschild E, Biber K, Rosenzweig M, Pykett M (2000) Neuronal, mesenchymal and hematopoietic cells derived from CD34, lin cells from adult bone marrow. Exp Hematol 28:46 Deng W, Obrocka M, Fischer I, Prockop DJ (2001) In vitro differentiation of human marrow stromal cells into early progenitors of neural cells by conditions that increase intracellular cyclic AMP. Biochem Biophys Res Commun 282:148 Li Y, Chen J, Wang L, Lu M, Chopp M (2001) Treatment of stroke in rat with intracarotid administration of marrow stromal cells. Neurology 56:1666 Chen J, Li Y, Wang L, Lu M, Zhang X, Chopp M (2001) Therapeutic benefit of intracerebral transplantation of bone marrow stromal cells after cerebral ischemia in rats. J Neurol Sci 189:49 Lu D, Mahmood A, Wang L, Li Y, Lu M, Chopp M (2001) Adult bone marrow stromal cells administered intravenously to rats after traumatic brain injury migrate into brain and improve neurological outcome. Neuroreport 12:559 Chopp M, Zhang XH, Li Y, et al. (2000) Spinal cord injury in rat: treatment with bone marrow stromal cell transplantation. Neuroreport 11:3001 Phinney DG, Kopen G, Isaacson RL, Prockop DJ (1999) Plastic adherent stromal cells from the bone marrow of commonly used strains of inbred mice: variations in yield, growth, and differentiation. J Cell Biochem 72:570 Haynesworth SE, Goshima J, Goldberg VM, Caplan AI (1992) Characterization of cells with osteogenic potential from human marrow. Bone 13:81 Friedenstein AJ, Gorskaja JF, Kulagina NN (1976) Fibroblast precursors in normal and irradiated mouse hematopoietic organs. Exp Hematol 4:267 Conget PA, Minguell JJ (1999) Phenotypical and functional properties of human bone marrow mesenchymal progenitor cells. J Cell Physiol 181:67 Kadiyala S, Young RG, Thiede MA, Bruder SP (1997) Culture expanded canine mesenchymal stem cells possess osteochondrogenic potential in vivo and in vitro. Cell Transplant 6:125 Haskins M, Baker HJ, Birkenmeier E, et al. (1991) Transplantation in animal model systems. In RJ Desnick (ed.): Treatment of Genetic Diseases. New York: Churchill Livingstone, p. 183 Griffin B, Baker HJ (2001) Domestic cats as laboratory animals. In JG Fox, LC Anderson, FM Lowe, FW Quimby (eds.): Laboratory Animal Medicine, 2nd edition. New York: Academic Press, p. 460 Menotti-Raymond M, David VA, Lyons LA, et al. (1999) A genetic linkage map of microsatellites in the domestic cat (Felis catus). Genomics 57:9 Migaki G (1982) Section VIII. Compendium of inherited metabolic diseases in animals. In: R Desnick, D Patterson, DG Scarpeooi (eds.):
886
25. 26.
27. 28.
29.
30.
D.R. Martin et al./Experimental Hematology 30 (2002) 879–886 Animal Models of Inherited Metabolic Disease, 1st edition. New York: A.R. Liss, p. 473 Linenberger ML, Abkowitz JL (1995) Haematological disorders associated with feline retrovirus infections. Baillieres Clin Haematol 8:73 Rohn JL, Gwynn SR, Lauring AS, Linenberger ML, Overbaugh J (1996) Viral genetic variation, AIDS, and the multistep nature of carcinogenesis: the feline leukemia virus model. Leukemia 10:1867 Willett BJ, Flynn JN, Hosie MJ (1997) FIV infection of the domestic cat: an animal model for AIDS. Immunol Today 18:182 Conget PA, Minguell JJ (2000) Adenoviral-mediated gene transfer into ex vivo expanded human bone marrow mesenchymal progenitor cells. Exp Hematol 28:382 Baker HJ, Reynolds GD, Walkley SU, Cox NR, Baker GH (1979) The gangliosidoses: comparative features and research applications. Veterinary Pathology 16:635 Baker HJ, Walkley SU, Rattazzi MC, Singer HS, Watson HL, Wood
31.
32.
33.
34.
PA (1982) Feline gangliosidoses as models of human lysosomal storage diseases. Progress in Clinical & Biological Research 94:203 Baker HJ, Walkley SU, Rattazzi MC, Singer HS, Watson HL, Wood PA (1982) Feline gangliosidoses as models of human lysosomal stroage diseases. In RJ Desnick, DF Patterson, DG Scarpelli (eds.): Animal Models of Inherited Metabolic Diseases. New York: Alan R. Liss, p. 203 Baker HJ Jr., Lindsey JR, McKhann GM, Farrell DF (1971) Neuronal GM1 gangliosidosis in a Siamese cat with -galactosidase deficiency. Science 174:838 Bruder SP, Jaiswal N, Haynesworth SE (1997) Growth kinetics, selfrenewal, and the osteogenic potential of purified human mesenchymal stem cells during extensive subcultivation and following cryopreservation. J Cell Biochem 64:278 Kramer JW, Hoffmann WE (1997) Clinical enzymology. In JJ Kaneko, JW Harvey, ML Bruss (eds.): Clinical Biochemistry of Domestic Animals, 5th ed. San Diego: Academic Press, p. 303