WO1997039104A1 - Cryopreservation and extensive subculturing of human mesenchymal stem cells - Google Patents

Abstract

Disclosed is a cryopreserved preparation of an isolated, homogeneous population of viable human mesenchymal stem cells obtained from periosteum, bone marrow, cord blood, peripheral blood, dermis, muscle, or other known sources of mesenchymal stem cells. After restoration from cryopreservation, the human mesenchymal stem cells can differentiate into cells of connective tissue types, including bone, cartilage, adipose, tendon, ligament, muscle, dermis, and marrow stromal connective tissue which supports the differentiation of hematopoietic stem cells. The cryopreserved preparation of human mesenchymal stem cells binds to antibodies produced from hybridoma cell lines SH2, SH3, and SH4, which have the ATCC accession numbers HB 10743, GB 10744, and HB 10745, respectively.

Description

CRYOPRESERVATION AND EXTENSIVE SUBCULTURING OF HUMAN MESENCHYMAL STEM CELLS

This application is a continuation-in-part of U.S. provisional application serial no. 60/015,712, filed April 17, 1996.

Use of the term "stem" cell has generally been reserved for those cells possessing the ability to self-replicate and give rise to daughter cells which undergo a unidirectional, terminal differentiation process (Hall and Watt, 1989). Three adult tissues in which stem cells have been extensively studied include the epidermis, gastrointestinal epithelium, and the hematopoietic compartment of bone marrow. Of these, hematopoietic stem cells are perhaps the best characterized (Lemischka et al., 1986; Sachs, 1987; Spangrude et al. , 1988), and are noted for their ability to give rise to multiple cellular phenotypes through lineage progression of daughter progenitor cells. Also present in adult bone marrow is a population of mesenchymal stem cells (MSCs) which give rise to multiple mesodermal tissue types including bone, cartilage, tendon, muscle, fat, and a marrow stromal connective tissue which supports the differentiation of hematopoietic stem cells (Dexter and Testa, 1976; Owen, 1985, 1988; Beresford, 1989; Caplan, 1991; Caplan et al., 1993).

Recently, a technique for the isolation and extensive subcultivation of human marrow-derived MSCs has been developed, along with a series of monoclonal antibody probes which react with the surface of these cells both in vitro and in situ (Haynesworth et al., 1992a, b; Bruder et al., 1995a). These and other studies demonstrate that purified, culture-expanded human MSCs are capable of differentiating along the osteogenic (Haynesworth et al. , 1992b; Bruder et al. , 1995b), chondrogenic (Lennon et al, 1996; Johnstone et al. , 1996), adipogenic (Pittenger, unpublished observations), and hematopoietic supportive stromagenic lineages (Majumdar et al. , 1995). Further characterization of these cells documents their complement of cell surface and extracellular matrix molecules (Haynesworth et al. , 1995), as well as their secretory cytokine profile under standard growth conditions, or following stimulation by selected inductive molecules (Haynesworth et al., 1996). While the precise mechanisms regulating commitment of these and other stem cells remain elusive, we have demonstrated that under appropriate tissue culture conditions, entire MSC populations can be directed into the osteogenic lineage. The cellular and molecular events of this differentiation cascade have been presented in a companion paper (Jaiswal et al., 1997), however, the observations may be summarized as follows. MSCs exposed to optimal concentrations of dexamethasone, ascorbic acid, and /3-glycerophosphate in vitro assume a cuboidal moφhology, upregulate alkaline phosphatase enzyme activity, express osteoblastic cell surface antigens, modulate the synthesis of osteocalcin mRNA in response to 1, 25-(OH)₂ vitamin D₃, and deposit a mineralized extracellular matrix characteristic of osteoblasts and terminally differentiated osteocytes reproducibly within 16 days (Bruder et al., 1995a, b; Jaiswal et al. , 1997).

In view of the increasing interest in aging research and studies of cellular senescence, we have sought to examine the behavior of human MSCs (hMSCs) against the backdrop of stem cell biology and Leonard Hayflick's seminal work describing the limited proliferative capacity of normal human cells (Hayflick and Moorehead, 1961; Hayflick, 1965). In brief, Hayflick demonstrated that fibroblastic cells derived from fetal lung tissue were capable of approximately 50 population doublings, while similar cells derived from adult lung lost their replicative potential after only 20 population doublings. Furthermore, cryogenic storage for as long as 27 years had no effect on the behavior of these cells (Hayflick, 1989). Summary of the Invention

With the above in mind, the studies reported here addressed the growth kinetics, self-renewing capacity, and the osteogenic potential of purified hMSCs during extensive subcultivation and following cryopreservation. By maintaining adult hMSCs in the log phase of growth, this study demonstrates that hMSCs have a relatively high replication capacity compared to other adult cell types. Despite the eventual loss of replication capacity with extensive passaging, the osteogenic potential of hMSCs is conserved throughout all subcultivations, and no differences in replicative or osteogenic potential are observed following cryogenic preservation.

Thus, the invention provides a cryopreserved preparation comprising an isolated, homogeneous population of viable human mesenchymal stem cells which can differentiate into cells of more than one connective tissue type upon being restored from cryopreservation. Such stem cells bind to antibodies produced from hybridoma cell lines SH2 (ATCC accession number HB 10743), SH3 (ATCC accession number HB 10744), and SH4 (ATCC accession number HB 10745). The mesenchymal stem cells in the cell preparation have preferably been culture- expanded by serial passaging prior or subsequent to cryopreservation.

These cells can be culturally expanded, for example, in BGJ_b medium containing 10% fetal serum or in a chemically defined medium which does not require serum. Suitable media for culture expansion of these cells are described in U.S. patent number 5,486,359, issued January 23, 1996, and suitable chemically defined media which do not require the presence of serum are described in U.S. application serial number 08/464,599, filed June 5, 1995.

Particularly preferred are such cryopreserved mesenchymal stem cell preparations when mesenchymal stem cells were obtained from periosteum, bone marrow, cord blood, peripheral blood, dermis, muscle or other known sources of mesenchymal stem cells. Also preferred are such cryopreserved preparations wherein the human mesenchymal stem cells, upon being restored from cryopreservation, can differentiate into cells of connective tissue-types, including bone, cartilage, adipose, tendon, ligament, muscle and dermis.

Brief Description of the Drawings

Figure 1. Protocol showing cell cultivation for serial passaging of hMSC cultures. Primary hMSC cultures were trypsinized as described herein and half of the cells were replated in 60 cm² dishes at a density of 5x10³ cells per cm². The remaining cells were cryopreserved overnight, thawed, and replated at a density of δxlO³ cells per cm² in 60 cm² dishes. At 80-90% confluence, cells were trypsinized and replated in 60 cm² dishes for subculturing and in 6-well plates at a density of 3xl0³ cells per cm² for in vitro osteogenesis assays. Cells used for in vitro osteogenesis assays are referred as being derived from passage 1 through X.

Figure 2. Growth curves of hMSCs in primary culture. Percoll-fractionated hMSCs from whole bone marrow cell suspensions were seeded at IO⁴ nucleated cells per cm² in 35 mm plates. hMSCs were allowed to attach for 3 days, after which the culture medium was changed on a twice weekly or daily schedule. Each day for 13 days, triplicate cultures were harvested for calculation of cell number as described below. The results represent the mean cell number ± SD from one representative donor and experiment. * p< 0.05

Figures 3A-3D. Growth curves of hMSC cultures at passages 1, 4, 7 and 10. Cultures were seeded at 2 x IO³ cells per cm² in 35 mm plates and fed on a twice weekly (solid circles) or daily (open circles) schedule. Each day for 9 days, one culture from each group was harvested for calculation of cell number as described below. The results represent the mean cell number ± SD of three donor preparations. * p < 0.05

Figures 4A-4B. Growth curve comparisons of hMSCs at passages 1, 4, 7 and 10. At each passage, cells were seeded at 2 x 10³ cells per cm² in 35 mm plates and fed on a twice weekly (A) or daily (B) schedule. Each day for 9 days, one culture from each group was harvested for calculation of cell number as described below. The results represent the mean cell number ± SD of three donor preparations. * p<0.05 (compared to P-l cultures).

Figure 5. Population doubling potential of hMSCs. Cumulative population doublings were calculated from the initial number of developing hMSC colonies and the number of hMSCs harvested at each passage. Results are presented as the mean cumulative populating doublings ± SD of hMSCs derived from four different donors.

Figures 6A-6B. APase activity, calcium deposition and cell proliferation in hMSC cultures grown in Control or OS Medium on days 4, 8, 12 and 16. Samples were harvested at the indicated days, and APase activity, calcium deposition and cell number were determined as described below. The results represent the mean ± SD of triplicate cultures of one representative experiment. *p<0.05, **p< 0.001 (compared to Control).

Figures 7A-7B. APase activity and cell proliferation on days 8 and 12, respectively, as a function of passage number and subsequent assay in Control or OS Medium. Cells were seeded at 3x10^s per cm² and switched to OS Medium the following day. Cultures were assayed for APase and cell proliferation as described below. The results represent the mean ± SD of duplicate cultures of one representative donor. *p<0.05 (compared to Control).

Figures 8A-8H. Effect of OS on cell moφhology, APase expression, and mineral deposition in 16 day hMSC cultures derived from first, fourth, seventh, and ninth passages. All specimens were stained by the APase and von Kossa histochemical techniques as described below. Phase contrast microscopy demonstrates the spindle-shaped moφhology of uniformly dense Control MSCs (A,B,C,D), while cells grown with OS are polygonal, APase positive, and exhibit nodular aggregates with von Kossa staining mineral (E,F,G,H). The formation of nodular aggregates in OS-treated cells frequently results in the exposure of bare tissue culture plastic visible in e.g, and h. Importantly, the moφhology of the cells cultured in either Control or OS Media is indistinguishable from first passage through ninth passage ( ~30 population doublings). (Bar, 200 μm).

Figures 9A-9B. APase activity as a function of passage number and cryopreservation in Control and OS-treated hMSC cultures on day 8. Cells from cryopreserved and non-cryopreserved samples were seeded at 3X10³ per cm² and switched to OS Medium the following day. On day 8, cultures were assayed for APase as described below. The results represent the mean ± SD of duplicate cultures of one representative donor. Upper and lower panels show the APase activity of cultures grown in Control Medium and OS Medium, respectively, for cryopreserved and non-cryopreserved cells. *p <0.05 (compared to non- cryopreserved cells).

Figures IOA- 1 OB. Cell proliferation as a function of passage number and cryopreservation in Control and OS-treated hMSC cultures on day 12. Cells from cryopreserved and non-cryopreserved samples were seeded at 3x10^s per cm² and switched to OS Medium the following day. On day 12, cultures were assayed for cell proliferation as described below. The results represent the mean ± SD of duplicate cultures of one representative donor. Upper and lower panel shows the cell number of cultures grown in Control Medium and OS Medium, respectively, for cryopreserved and non-cryopreserved cells. *p <0.05 (compared to non- cryopreserved cells).

Detailed Description of the Invention

To provide a viable cryopreserved product for human use, controlled, reproducible processes for the freezing, thawing and administration of the MSCs are required. To permit simple handling at the clinical site (either for direct infusion to a patient or further formulation for infusion or combination with a matrix), the MSCs should ideally be cryopreserved at high concentration in a small volume of a medium which is acceptable for injection and which does not affect the functionality of the cells. The examples detailed below present the results obtained for cryopreservation of human MSCs frozen to -80°C in a Forma CryoMed mechanical freezer at a controlled rate of 1 °C per minute at the cell concentration and in the media compositions described. Frozen cells were transferred to a -150°C freezer and stored for the time periods described.

The majority of media formulations currently used for cryopreservation of cells contain dimethyl-sulfoxide (DMSO). Media formulations used for cryopreservation of cells for clinical use contain DMSO in combination with HETA- Starch and/ or human serum components and/or other protein bulking agents. Historical research and development studies on cryopreservation of MSCs used a freezing medium composed of 90% fetal bovine serum + 10% DMSO. This latter medium, therefore, serves as a positive control with which to assess process efficiency.

As detailed above, a range of components/ additives are acceptable in cryopreservation medias used for clinical cell processing. The cost and safety of some additives is, however, a consideration. Our aim has been to rapidly develop an economic and safe media for clinical use. In order to eliminate the disease transmission issues associated with allogeneic human serum lots, we opted to evaluate autologous human serum as the bulk stabilizing agent. Peripheral blood stem cell grafts of ≥ 500ml containing up to 10% DMSO are routinely infused directly into patients. Even high dose MSC infusions are likely to be ≤ 50ml, therefore, use of up to 10% DMSO in the cryopreservation medium should not pose any issue of excessive toxicity to the patient.

With the above considerations, culture expanded human MSCs have been frozen in the following media compositions:

(i): 1:1 mixture of 15% DMSO Cryoprotective Medium (Bio-

Whittaker Part#12-DMEM-low glucose (Life Technologies

Part#11885) (ii) 90% autologous serum + 10% DMSO (iii) 95 % autologous serum + 5 % DMSO

(iv) 90% DMEM-lg + 10% DMSO

(v) 85% DMEM-lg + 5% autologous serum + 10% DMSO

(vi) 90% fetal bovine serum + 10% DMSO

Cryopreserved human MSCs are rapidly thawed in a 37°C water bath until only a small ice crystal remains. The 1.0ml cell suspension is then aseptically transferred to 10ml of complete hMSC medium and centrifuged at 200xg for 5 minutes at 20-23°C. The supernatant is quickly aspirated and the pellet resuspended in 1.0ml of complete hMSC medium. Viable cells are counted by Trypan blue exclusion (mix an aliquot of the cell suspension 1: 1 with 0.4% Trypan blue and count un-stained cells on a Hemacytometer). Cells are then seeded at approximately 5,400 viable cells per cm² to assess plating and expansion capability.

Example 1 Cryopreservation of Human MSCs

MSCs were frozen at 2 x 10⁶ cells per ml in the media formulations detailed in section (1.2) using the process described above. After 24 hours storage at - 150°C, the cells were thawed using the procedure described above and analyzed for:

(i) viability by trypan blue dye exclusion as described above;

(ii) days to reach confluency upon plating at 5,400 viable cells cm² in T- flasks. Average time to confluency in standard culture conditions was 5 - 6 days. The results shown in Table 1 demonstrate that good viability is retained on thawing MSCs cryopreserved in the various media.

-θ - Table 1

Viability of Cryopreserved MSCs Frozen at

2 x IO⁶ CeUs per ml

After 24 Hours of Storage at -150°C

Example 2

Cryopreservation of human MSCs at 2 x IO⁶ cells per ml followed by 7 days of storage at -150°C

MSCs were frozen at 2 x IO⁶ cells per ml in the media formulations detailed in section (1.2) using the process described above. After 7 days storage at -150°C, the cells were thawed using the procedure described above and analyzed for:

(i) viability by trypan blue dye exclusion as described above;

(ii) days to reach confluency upon plating at 6,000 cells cm² in T-flasks The results shown in Table 2 demonstrate that good viability is retained on thawing MSCs cryopreserved in various media.

Table 2 Viability of Cryopreserved MSCs Frozen at 2 x IO⁶ Cells per ml After 7 Davs of Storage at -150°C

Example 3

Viability of Cryopreserved Human MSCs Frozen at Different Cell

Concentrations in Buffer (a) 1:1 (Biowhittaker) :DMEM-lg

MSCs were frozen at 2, 5, 10 and 20 x IO⁶ cells per ml in medium (a) the 1 : 1 Biowhittaker : DMEM-low glucose medium detailed above using the process described above. After 7 days storage at -150°C, the cells were thawed using the procedure described above and analyzed for:

(i) viability by trypan blue dye exclusion as described above;

(ii) days to reach confluency upon plating at 5,400 cells cm² in T-flasks. The average for standard expansion to confluency is 5 - 6 days. The results shown in Table 3 demonstrate that good viability is retained on thawing MSCs cryopreserved at different cell concentrations.

Table 3

Viability of Cryopreserved MSCs Frozen at Different Cell

Concentrations in Buffer

(a) 1:1 Biowhittaker :DMEM-low glucose

Example 4

Differentiation and Proliferation of

Cryopreserved Human MSCs

Materials

Dexamethasone (Dex), sodium j3-glycerophosphate G3-GP), Percoll, penicillin/streptomycin antibiotic, DMSO, alkaline phosphatase diagnostic kit #85, and calcium diagnostic kit #587 were purchased from Sigma Chemical Co. (St. Louis, MO). DMEM-LG (DMEM) was purchased from GIBCO (Grand Island, NY), L-ascorbic acid-2-phosphate (AsAP) from Wako Chemical (Osaka, Japan), and fetal bovine serum from Biocell Laboratories (Rancho Dominguez, CA) following an extensive testing and selection protocol (Lennon et al., 1996). All other routine reagents used were of analytical grade.

Cell preparation and culture methods

Fresh bone marrow (10 ml) was obtained by routine iliac crest aspiration from normal human donors after informed consent. Human MSCs were isolated from these marrow aspirates using methods modified from those described previously (Haynesworth et al., 1992a). Briefly, 10 ml of marrow was added to 20 ml of DMEM containing 10% fetal bovine serum from selected lots (Control Medium), and centrifuged to pellet the cells and remove the fat layer. Cell pellets were then resuspended and fractionated on a density gradient generated by centrifuging a 70% Percoll solution at 13,000 xg for 20 min. The hMSC- enriched low density fraction was collected, rinsed with Control Medium, plated at IO⁷ nucleated cells per 60 cm² dish in Control Medium, and cultured at 37 °C in a humidified atmosphere containing 5 % CO₂. Adherent hMSCs represent approximately 1 in 10^s nucleated cells in this low density fraction. Non-adherent cells were removed on day 3 at the time of the first medium change, and fresh Control Medium was changed twice weekly thereafter. When culture dishes became near-confluent, cells were detached with 0.25% trypsin containing ImM EDTA for 5 min at 37°C. For determining growth kinetics in primary cultures, the low density fraction described above was plated at 1.6 x IO⁶ cells per 10 cm² dish in multiple dishes. On day 3, the culture medium was removed along with the non-adherent cells, and fresh Control Medium was added. Starting on day 3 and continuing each day until day 14, three dishes were used to quantify the number of adherent hMSCs as described below. Measurements were obtained on primary cultures from three different marrow donor samples. For measuring growth kinetics of serially passaged cultures, primary hMSCs cultivated on 60 cm² dishes were trypsinized, counted on a hemacytometer, and a portion were replated at a density of 3x10^s cells per cm² on sixty 10 cm² dishes, randomly divided in two groups. Following overnight attachment, cells were maintained for ten days with medium changes occurring daily in one group of dishes and twice weekly in the other group. Starting the day after plating (day 0) and continuing each day until day 9, three dishes from each group were used to determine cell number. The remaining dishes of primary hMSCs which were not replated for the above growth kinetics assay were subcultivated at a 1:3 split ratio onto new 60 cm² dishes. These cells were allowed to replicate until they became near confluent, at which time a portion of cells were again replated on sixty 10 cm² dishes. The remaining 60 cm² dishes were split 1:3, and serially passaged as above. This scheme of serial subcultivation, with growth kinetics assays at each passage, was continued until the cells became senescent. The number of adherent hMSCs in primary cultures was determined by counting the number of MSC colonies, each of which represent the progeny of a single hMSC (Haynesworth et al., 1992a, b). Meticulous record-keeping of the number of cells present at the start and end of each passage facilitated calculation of the number of population doublings for any given passage number. Growth curves and population doubling calculations represent experiments performed using hMSCs from four marrow donors, and is presented as the mean ± SD of the cell number measurements obtained for all three marrow donors combined.

Experiments to determine the osteogenic potential of hMSCs during extensive subcultivation and following cryopreservation were performed as follows. Aliquots of trypsin-released primary hMSC cultures were cryopreserved in FBS with 10% DMSO in liquid nitrogen, thawed 24 hours later, tested for viability by Trypan Blue exclusion, and plated in three 60 cm² dishes at 5xl0³ cells per cm². A separate aliquot of the primary cells were identically detached and replated to perform direct comparisons of fresh versus cryopreserved hMSCs obtained from the same donor. When culture dishes from both groups (cryopreserved and unfrozen original) became near-confluent, cells were detached with trypsin (as described above), and replated in three 60 cm² dishes at a density of 5x10^s cells per cm² for continued passaging and in 6-well tissue culture plates at 3x10^s cells per cm² for in vitro osteogenic assays as described previously (Jaiswal et al., 1996). This procedure for replating the cells was repeated with similar plating densities at each subculture when cells grown in the 60 cm² dishes became near-confluent.

Detached cells were counted using a hemacytometer at the end of each passage in order to calculate the number of population doublings. The protocol for cell handling, subculture and osteogenic assays is diagramatically presented in Figure 1. Cells were subcultured in this manner for up to 10 passages, and their in vitro osteogenic potential was measured at each passage in assays performed on days 4, 8, 12 and 16 as described below. For these in vitro osteogenic assays, the cells grown in 6-well plates were provided fresh Control Medium one day after plating (Day 0) and subsequently grown in the absence or presence of an optimized mixture of Osteogenic Supplements (OS) (100 nM dexamethasone, 10 mM /3-GP and 0.05 mM AsAP) as previously described (Jaiswal et al. , 1997). Media changes were performed twice weekly, with a medium volume of 2 ml per well.

Cell Proliferation Assay

Cell proliferation was measured in triplicate cultures using a modification (Lennon et al., 1995) of the crystal violet dye-binding method (Westergren- Thorsson et al. , 1991). Cultures were rinsed twice with Tyrode's balanced salt solution, fixed with 1 % gluteraldehyde (v/v) in Tyrode's for 15 min, rinsed twice with deionized water and air-dried. Cultures were then stained with 0.1 % crystal violet (w/v) in water for 30 min. After washing, crystal violet dye was extracted from the cells by 4 hr rotary incubation at 25°C with 1 % Triton X-100 (v/v in water). Absorbance of the resulting Triton extract was read at 595 nm on a Bio- Rad Microplate reader. Absorbance values were converted into absolute cell numbers based on established standard curves.

Alkaline Phosphatase Assay

Alkaline Phosphatase (APase) enzyme activity of the cell layer was measured in triplicate cultures by rinsing twice with Tyrode's balanced salt solution, and then incubating the cells with 5 mM p-nitrophenyl phosphate in 50 mM glycine, 1 mM MgCl₂, pH 10.5, at 37°C for 5 to 20 min. APase enzyme activity was calculated after measuring the absorbance of the reaction product formed, p-nitrophenol (pNP), at 405 nm on a Bio-Rad Microplate reader. Enzyme activity was expressed as nmol of pNP produced per min per dish.

Histochemical Analyses

Alkaline phosphatase histochemistry was performed for 1 hour at 25°C as recommended by the manufacturer's instructions contained in Sigma Kit #85. During incubation, culture dishes were protected from drying and direct light. Dishes were rinsed with deionized water, and air-dried prior to viewing. Selected specimens were subsequently stained for mineral by the von Kossa method. Cell layers were fixed with 10% formalin for 1 hr, incubated with 2% silver nitrate solution (w/v) for 10 min in the dark, washed thoroughly with deionized water and then exposed to bright light for 15 min.

W

Calcium Assay

Cell layers were rinsed twice with PBS and scraped off the dish in 0.5 N HCl. The calcium was extracted from the cell layers by shaking for 4 hours at 4°C, then centrifuging at 1,000 xg for 5 min. The resulting supernatant was used for quantitative calcium determination according to the manufacturer's instructions contained in Sigma Kit #587. Absorbance of samples was read at 575 nm on a Beckman Spectrophotometer. Total calcium was calculated from standard solutions prepared in parallel, and expressed as μg per dish.

Statistics

Statistical analyses were performed using the Student's / test.

RESULTS

Growth Characteristics of Primary hMSC Cultures

Human MSCs were introduced into culture after enrichment from whole bone marrow cell suspension using a gradient composed of 70% Percoll. After plating, hMSCs were allowed to attach for three days at which time the culture medium was replaced with fresh medium. At the early stage of primary culture, adherent hMSCs can be observed as sparsely distributed individual spindle-shaped cells. In all of our previous studies, culture adherent hMSC have been allowed to mitotically expand in primary culture with Complete Medium changes occurring twice weekly. When hMSCs in primary cultures expand to form colonies of several hundred cells that collectively cover 80-90% of the culture plate, the cells are passaged at a 1:3 dilution after being detached from the culture substratum with trypsin. The length of time from introduction of hMSCs into culture until their harvest for subcultivation into first passage is generally 12 to 14 days. In this study we measured the rate of division of hMSCs in primary cultures processed by our standard twice weekly medium change, and in cultures where the medium was changed daily.

The growth curves of hMSCs in primary cultures exhibit a lag phase of 6 to 8 days (Figure 2). Upon visual observation on day 2, most of the adherent hMSCs were seen as individual, spindle-shaped cells that were sparsely distributed across the plate indicating that little mitotic expansion had taken place during the three day attachment period. In addition to the spindle-shaped hMSCs, round nonadherent cells were also observed. These nonadherent cells comprise the majority of cells originally seeded onto the plate. A small percentage of these nonadherent cells appear to non-specifically stick to the plates during me early days of culture without actually spreading out across the dish and becoming adherent to the substrate. Occasionally, these cells contribute to a transiently high background when cultures are analyzed for cell number by the crystal violet assay on days 2 through 4. By day 5 these cells are easily removed during the process of changing culture medium, and consequently, the background is reduced to undetectable levels. The lower limit of the crystal violet assay to accurately measure cell number is 1000 cells as determined by standard curve. As such, we stained plates with crystal violet dye and viewed adherent cells directly by phase contrast microscopy to complement the cell number measurement generated from the crystal violet colorimetric assay.

On day 3 a few small colonies of 4 to 8 hMSCs were observed, however, these early dividing cells represented only a small percentage of the total number of adherent cells (-500), which would eventually begin to divide and form larger colonies. Between days 4 and 6 the number of observable hMSC colonies increased, and a few large colonies could be observed resulting from the cells which had begun dividing by day 3. By day 8, the majority of hMSC colonies were established. Some of these colonies contained only 4 to 8 cells, while other colonies were very large and contained hundreds of cells. The pattern of formation of early, intermediate and late developing colonies was similar in both twice weekly and daily fed cultures. Colorimetric quantitation of hMSC cell number showed an exponential increase in cell number between days 8 and 10, followed by a plateau phase of slower cell growth from days 11 through 13. Although the media changing schedule did not appear to influence the initiation or growth of colonies during the lag phase, when hMSC cultures entered into log phase, daily fed samples generated a steeper growth curve and resulted in the formation of significantly more cells as compared to cultures fed twice weekly on days 10 through 13.

Growth Characteristics of Serially Passaged hMSC Cultures

Growth curves were generated at each passage beginning with passage 1 until culture senescence, which occurred between passages 10 and 15 depending on the donor. Cells were allowed to divide for 10 days with medium changes occurring twice weekly or daily. Figure 3 shows the growth curves obtained at passages 1, 4, 7 and 10 to illustrate the similarities and differences of hMSC growth characteristics in early, middle and late passaged cultures. At each passage, hMSC growth curves depict an initial lag phase of 24-36h. This was followed by a log phase in which the hMSCs mitotically divided at exponential rates for 4 to 6 days, depending on the passage from which the cells were derived. The log phase was followed by a plateau phase where mitotic division continued through day 9 of culture, but at a slower rate.

hMSC cultures that were fed daily grew at faster rates and generated significantly more cells by the end of the 9 day growth period as compared to cultures supported by twice weekly medium changes (Figure 3). For example, at passage 1 , the number of cells in daily fed cultures were significantly greater on days 5, 7 and 8 (p<0.05), and appreciably but not significantly higher on day 9 (p=0.12) compared to twice weekly fed cultures. Daily fed cultures from passage 4 contained significantly more cells on days 3 through 9 (p<0.05), while daily fed passage 7 cultures contained more cells on days 6 through 9 (p<0.05), as compared to cultures fed twice weekly. By passage 10, however, as the cells approached the limits of their replicative potential, wide variability was observed in the number of cells on each day during the 9 day culture period for both daily and twice weekly fed cultures, resulting in no statistical differences in cell number.

With increasing passage number, the hMSC growth rates were slower and the number of cells generated by the end of 9 days in culture was reduced. Figure 4 illustrates that by passage 4, a decline in the number of cells generated during the 9 day culture period was observable. For example, 9 days after replating cells derived from passages 1 , 4, 7, and 10, the mean yields for twice weekly fed cultures were 2.7x10^s, l. lxlO⁵, 0.88x10^s, and 0.68x10^s cells, respectively. This reduction did not result in significantly fewer cells in twice weekly fed cultures at any time point in passage 4 (Fig. 4a), however, in daily fed cultures at this passage, proliferation was compromised between days 5 and 8 (p<0.05) when compared to the growth of cells derived from passage 1. At passage 7 and beyond, the growth of cells had waned in comparison to passage 1 cultures at nearly every time point beyond day 5 (p<0.05). This slowing of cell proliferation as a function of increasing passage number was independent of the feeding schedule used to maintain the cells.

Self Renewal Capacity of hMSCs

The total population doubling potential was calculated from the number of population doublings determined for each passage from four different donors. Starting with adherent hMSC colonies in primary culture, population doublings were readily calculated until the time of replicative senescence, which began between tenth and fifteenth passage, depending on the donor. For primary cultures, the initial number of hMSCs that attached to the plate and divided to form colonies was estimated as the average number of colonies which form in primary culture based on our colony count database of over fifty donors (data not shown). All other determinations represent the actual number of hMSCs counted at the end of each passage prior to introducing the cells back into culture for the next passage. The mean cumulative population doublings for the four donor hMSC preparations was 38 ± 4 (Figure 5). On average, 11.2 population doublings took place in primary culture, accounting for 29.4% of the mean cumulative number of population doublings. The average starting number of hMSCs was 500 cells, and the average final number at the end of primary culture was 6.1x10^s. For passages 1 through 10, the average number of population doublings at each passage was about 2, which was expected since the cultures were replated when they became 90-95 % confluent at each passage. After tentii passage, the number of population doublings declined as senescence was reached in the different donor preparations. By passage 12, hMSCs in two of the four donor preparations had stopped dividing while cells in the other two cultures continued to divide very slowly and were passaged prior to achieving 90% confluence to determine if replating would stimulate further cell division. As observed in Figure 5, few additional population doublings were generated by cells subcultured from passages 12 through 15.

Continuous Subculturing of hMSC Cultures

When primary cultures of hMSCs were subcultivated, cells attached uniformly to the culture dishes at approximately 30% confluence and reached 80- 90% confluence in 5 days, at which time, cells were subcultured again. As passage number increased, the time between initial plating and subsequent subculturing increased from 5 days to approximately 10 days by tenth passage. (Figure 4a). Uniform attachment and spindle-shaped moφhology of hMSCs was observed on cells at every passage until the hallmarks of cellular senescence appeared during the final subculture. These features include an increased duration of the cell cycle's GI phase, cessation of mitotic activity, accumulation of cellular debris and stress fibers, a broad flattened moφhology, and ultimately, total degeneration of the culture. Retention of the MSC phenotype following serial passaging has previously been confirmed on all cells by positive cell surface immunostaining with monoclonal antibodies SH-2, SH-3, and SH-4 (Haynesworth et al., 1992a).

Induction of Osteogenesis by OS in hMSC Cultures

As detailed in a companion report to this (Jaiswal et al. , 1997), MSCs cultured with OS undergo a dramatic change in cellular moφhology from that of spindle-shaped to cuboidal, which is accompanied by an increase in APase activity and mineral deposition. This increase in APase activity and mineral deposition was consistent for all marrow donors, and is illustrated by one representative experiment in Figure 6a. A significant increase in APase activity was observed after 4 days of OS treatment with maximal activity occurring on day 12, followed by a decline by day 16. This late decrease in APase activity of OS cultures is reproducible, and correlates with increasing mineral deposition and terminal differentiation of cells into osteocytes. Although the timing of peak APase activity varies from day 8 through 12 for different donors and samples, virtually all cultures show a decrement in activity by day 16. These cultures were also studied for their ability to elaborate mineralized extracellular matrix when grown in the presence or absence of OS. No calcium deposition was detected either by Von Kossa staining or the sensitive colorimetric quantitative calcium assay in Control MSC cultures. As shown in Figure 6a, MSCs grown with OS showed a significant calcium deposition as early as day 8 (1.32 ±0.05 μg/dish), with a further increase on days 12 (16.26±2.94 μg/dish) and 16 (31.18 ± 1.52 μg/dish). As illustrated in Figure 6b, treatment with OS also significantly increased cell proliferation on days 8, 12 and 16. The results presented here and below reflect experiments performed with MSCs derived from one donor, although similar results were obtained with MSCs derived from other donors which were cryopreserved and/or extensively subcultured.

Osteogenic Differentiation and Cell Proliferation as a Function of Passage Number in Control and OS-Treated hMSC Cultures.

Figure 7 illustrates the APase activity and cell proliferation on days 8 and 12, respectively, of serially subcultured hMSCs grown in the absence and presence of OS. Relatively little variability occurred in the low basal APase activity (range, 3.42±0.29 to 8.91 ±0.49 nmol pNP/min/dish) of these cells even after subculturing nine times. Human MSC cultures treated witii OS showed a significant increase in APase activity on day 8 in cells derived from passages 1 through 10 (Figure 7a). Interestingly, the lowest APase activity was observed in cells derived from passage 10 in this donor (> 30 population doublings), which parallels the point where mitotic activity decreased and cells began to degenerate in a manner similar to Hayflick's Phase III senescence (Hayflick, 1965). Similar to Control cultures, OS cultures derived from passage 10 had the lowest absolute value of APase activity. Although the absolute level of APase activity on day 8 varied from one passage to another, a 3-5 fold increase in APase activity was invariably observed in these serially passaged, OS-treated hMSCs. Furthermore, despite that fact that cells derived from tenth passage expressed the lowest level of APase activity in both Control and OS-treated cultures, the fold stimulation observed in OS-treated cultures was comparable to that obtained in cultures derived from passages 1 through 9. Additional evidence supporting the osteogenic differentiation of these extensively subcultured MSCs is provided in Table 1 , which indicates the maximum fold stimulation of APase activity per cell for each passage.

Figure 7b depicts cell proliferation by day 12 in the hMSC cultures derived from sequential passages. The cell number in Control cultures from passage 2 through 9 was approximately 2-4X10³ cells on day 12, or about 10-fold greater than the number of cells originally seeded on the dish, and clearly represents cells in the plateau phase of growth as described above. Furthermore, the decline in cell proliferation as a function passage number is reproduced in this series of experiments, with cell numbers from assays at passages 7 and 10 significantly (p< 0.001) lower than those from passage 1. Interestingly, although the baseline rate of cell division decreased as a function of passage number, the addition of OS to these MSCs caused a characteristic and significant increase in cell number at every passage tested (p<0.05). Visual inspection as well as cell counting of the MSC cultures derived from tenth passage revealed that at this stage of subculture for this MSC donor, the rate of cell proliferation decreased markedly. Additional subcultivation and osteogenic assays were not performed beyond tenth passage since the cells had become senescent and degeneration ensued. The relatively low values for APase activity of Control and OS-treated cultures derived from tenth passage, when expressed as nmol pNP/min/dish, are a direct result of the low cell numbers present at day 8 in extensively subcultured cells. The cell proliferation data from this hMSC donor, with respect to the self- renewing capacity of approximately 30 population doublings, lies within the range detailed in experiments presented in Figure 5. Spindle-shaped hMSCs at the start of cell culture in each passage became cuboidal within 48 hours of OS treatment. On day 4, Control cultures contained only few (< 1 %) APase-positive cells, whereas approximately 30-40% of cells were APase-positive when cultured in OS Medium. By day 16, Control cultures grew as a uniform sheet of cells in a whirling pattern with about 5 % APase positive spindle-shaped cells. OS-treated cultures displayed strong APase staining in approximately 90% of all cells after 8 days of treatment for all passages. As the cells grew to confluence by day 12, multilayered bone-like nodular aggregates were observed throughout the dish when cultured in OS Medium. After 16 days in OS Medium, the cells reproducibly formed a mineralized matrix. Figure 8 illustrates the Control and OS cultures from early, mid, and late passages when stained for APase and mineral after 16 days of culture. The critical feature in these photomicrographs is the observation that the moφhologic and developmental changes which hMSCs undergo in response to OS do not vary as a function of the passage number from which the starting cells were derived. Additionally, the quantitative changes in APase activity occurring on days 4 through 16 of Control and OS-treated samples from all subcultures (data not shown) duplicate the phenomenon described in Figure 6a using cells derived from first passage. That is, one can not distinguish cells derived from first passage through ninth passage based on the biochemical and moφhologic changes in response to OS-induced osteogenic differentiation.

Osteogenic Differentiation and Cell Proliferation in Serially Passaged hMSC Cultures Following Cryopreservation.

In an effort to determine whether the self-renewing capacity and/or osteogenic potential of hMSCs preserved in liquid nitrogen is similar to that of the original unfrozen cultures, we performed the above series of analyses on cryopreserved cultures in parallel with unfrozen cultures as outlined in Figure 1. The results of that comparison showing Control and OS-treated cultures are presented in Figure 9. Figure 9a shows that basal APase activity on day 8 in both cryopreserved and non-cryopreserved cultures were not significantly different from each other (p>0.05), except for passages 1, 2, and 4. Basal APase activity in cryopreserved and non-cryopreserved cultures ranged from 2.20±0.03 to 9.29±0.04, and 2.88±0.25 to 11.22±0,83 nmol pNP/min/dish, respectively. Figure 9b illustrates that treatment of cryopreserved, serially passaged hMSC cultures with OS markedly enhanced the APase activity as previously described for unfrozen cultures. Of particular interest is the fact that the variability observed in absolute APase activity at day 8 in original unfrozen cells from different passages is mirrored by changes in the absolute value of APase activity of cells which were cryopreserved. For example, assays performed on both fresh and frozen OS-treated cells derived from third passage show APase activity of approximately 12 nmol pNP/min/dish, whereas APase activity on OS-treated cultures derived from sixth passage is closer to 34 nmol pNP/min/dish. Importantly, although the absolute value of APase activity varies from passage to passage, the phenomenon of a 3-5 fold increase in APase activity of OS-treated cells derived from cryopreserved cultures exactly duplicates the results obtained from experiments using cells from any given passage which were never frozen.

As illustrated in Figure 10a, the proliferative potential of MSCs following cryopreservation is not significantly different (p> .05) than that of cells which were never frozen. Continued growth was observed for 10 passages, until division began to slow and cells eventually degenerated as described above. Again, in this donor, the cell number at day 12 in samples derived from passage 7 and 10 was significantly (p< 0.005) lower than the cell number in samples derived from passage 1. Furthermore, the characteristic mitogenic response of MSCs to OS presented in Figure 7b is retained following cryopreservation (Figure 10b). Here again, the results obtained with cryopreserved cells mirror the observations made with cell preparations which were never frozen.

DISCUSSION

In the present study, we have provided further characterization of a population of cells derived from human bone marrow which are found in very low abundance, possess high self-renewal capacity, and retain their undifferentiated phenotype through senescence, or until such time that inductive cues direct the cells into a restricted terminal differentiation pathway. These cells are referred to as human Mesenchymal Stem Cells (MSCs) because they are known to give rise to multiple mesodermal tissues including bone (Haynesworth et al., 1992b; Bruder et al. , 1995b) cartilage (Lennon et al. , 1996; Johnstone et al. , 1996), fat (Pittenger, unpublished observations), and hematopoietic supportive stroma (Majumdar et al. , 1995). We report here on the growth kinetics of purified, culture-expanded human MSCs, and that their behavior is consistent with other normal human diploid cells. We present additional data which demonstrates full retention of their osteogenic potential following cryopreservation and extensive subculture. Together, these investigations support the characterization of this population of cells as true stem cells, and further point the way to novel therapeutic options for self cell therapy.

MSC cultures are initiated as primary cultures of fibroblastic cells which grow out of marrow cell suspensions by selectively attaching to tissue culture plastic and forming clonal colonies through cell division. The process of colony growth from the newly adherent fibroblastic cell population is not uniform. Instead, some cells appear to quickly give rise to colonies soon after adherence to the culture dish, whereas other cells do not yield colonies until several days of culture. The rate of cell division is also variable with some colonies enlarging rapidly after the initial cell division, while other colonies expand slowly. This bi- modality in the rate of colony formation is well documented for fibroblastic cells in cultures derived from clonal and mass cell origin (Martin and Sprague, 1970; Martinez et al. , 1978; Matsumura et al. , 1979; Mitsui and Schneider, 1976). Fibroblastic cells from fetal and adult skin that replicate rapidly to form large colonies have been shown to possess higher population doubling capacity than cells that divide slowly and form small colonies (Smith et al., 1978). Similar observations were made by Mets and Verdonk (1981) in their study of in vitro aging characteristics of human bone marrow stromal cells. These investigators distinguished two adherent cell types in the primary cell population. Type I cells were spindle-shaped and rapidly dividing, while Type II cells were broader and divided slowly if at all. The majority of cells in primary cultures were Type I cells, however, during subcultivation the appearance of Type II cells formed an increasing fraction of the total cell population. These observations led to the inteφretation that Type I cells are progenitor cells which gradually give rise to Type II cells through tangential or asymmetric cell division, and that Type II cells represent a terminally differentiated state of non-dividing cells.

Our observations also indicate that primary hMSC cultures consist of a mixed population of rapidly and slowly dividing cells. During early subcultivations, spindle-shaped, rapidly dividing cells generate a near homogeneous population. However, after 5-6 passages, the moφhology of the cells becomes heterogeneous with broad cells distributed among the spindle- shaped cells. This likely occurs as some daughter cells, generated from the asymmetric cell division of spindle-shaped cells, acquire a broad moφhology and reduced replication capacity while other daughter cells retain the spindle-shape and high replication capacity. Differences in population doubling potential among human fibroblastic daughter cells from derived from the division of a single cell have been documented to differ by as many as eight population doublings (Smith and Whitney, 1980). The source of this heterogeneity is not clearly understood. Since the majority of hMSCs are broad in moφhology by 10th passage, these observations are consistent with those of Mets and Verdonk (1981) who noted that spindle-shaped rapidly dividing fibroblasts give rise to broad-shaped fibroblasts with diminished replication capacity. Our observations do not support the inteφretation by Mets and Verdonk that the nondividing broad cells represent a terminally differentiated population. Instead these broad cells, like the spindle- shaped cells of early passage, do not express osteogenic traits under Control medium conditions, but indeed retain the capacity to differentiate along the osteogenic lineage when exposed to OS, and to also increase their rate of division in response to OS.

At each passage, hMSCs followed a normal growth curve consisting of a lag phase followed by a log phase of exponential cell growth, ending with a plateau phase in which the growth rate declined. In daily fed cultures, the length of the exponential growth phase is similar to that of cultures fed twice weekly. However, the steeper slope of the growth curves during this phase results in a higher density of cells before the rate of growth begins to decline. The rate of growth and level of density-dependent inhibition of further growth is likely influenced by the concentration of serum derived growth factors in the culture medium. In cultures fed twice weekly, the concentration of mitogenic factors becomes rate-limiting as the cells metabolize the culture medium, whereas cells maintained in daily fed cultures are exposed to a consistently high concentration of serum-derived factors. As the number of population doublings increases, the rate of population growth declines. Studies on other types of fibroblastic cells have shown this process to involve a percentage of cells leaving the cell cycle at each successive division, and resulting in a lower net percentage of cells in the population which continue to divide (Martin and Sprague, 1970; Martinez et al., 1978; Matsumura et al. , 1979; Mitsui and Schneider, 1976; Mets and Verdonk, 1981). Data from these studies do not support the opposing hypothesis that the decline in population growth rate is due to a uniform decline in the rate of cell division throughout the cultures (Karatza et. al, 1984).

Regardless of the tissue source, fibroblastic cell growth follows a predictable sequence with increasing age in vitro. These events have come to be known as the Fibroblast Model or Hayflick Model of cellular age in vitro (Hayflick and Moorehead, 1961; Hayflick, 1965). Likewise, hMSCs possess many of the growth characteristics of other types of fibroblasts in culture. hMSCs pass through the four stages which characterize the Fibroblast Model. These events begin with the outgrowth of fibroblastic cells from bone marrow cell suspension, and proceeds with a period of rapid cell division (Phase I). This is followed by a period during which proliferation slows and signs of cellular deterioration begin (Phase II). During the next stage, proliferation declines further and ultimately stops (Phase III). The final stage is characterized by a period that can last for several months in which the cells do not divide but slowly degenerate resulting in gradual death of the culture (Phase IV). The mean total of population doublings of marrow-derived hMSCs from the adult donors used in this study was 38 ± 4. This number is greater than the average of 20 population doublings which occurs in adult lung fibroblasts (Hayflick, 1965) and keratinocytes (Rheinwald and Green, 1975). However, it is less than the 50 population doublings obtained from fetal lung fibroblasts (Hayflick, 1965), and the 70 population doublings observed with human umbilical vein endothelial cells (Thornton et al., 1983). These data suggest a population doubling potential of hMSCs which is intermediate to what has been reported for other human cells derived from either embryonic or adult tissues. Since the population doubling potential has been used widely as an index of biological age, the higher population doubling potential of hMSCs suggests that these cells are of a younger biological age than other normal adult cells, and as such, may provide an in situ source for mesenchymal progenitor cells well into adulthood.

As with other culture adherent cells, the mechanism for the loss of population doubling potential after a finite range of population doublings is not well understood. Several mechanisms for the cause of replication senescence have been put forth, including the accumulation of genetic damage (Szilard, 1959; Orgel, 1973), the shorting of telomeres (Hariey et al., 1990) and the activation of tumor suppressor (Sager, 1989) or senescence genes which code for proteins that block the cell cycle (Wang et al., 1994). In addition, replication senescence has been described as a type of terminal cell differentiation (Bell et al. , 1978). Future studies will be needed to determine which of these mechanisms are involved in the replicative senescence of hMSC and whether the replication senescence of all types of fibroblasts are controlled by the same mechanisms.

The events which occur during osteogenic differentiation of hMSCs, that is, commitment, lineage progression, and elaboration of a mineralized matrix, have been documented in a companion report to the current study (Jaiswal et al. , 1997). One of the hallmarks of this process is the significant elevation in APase activity which peaks between days 8 and 12 of culture, depending on the donor, and the passage from which the cells were derived. Experiments testing the osteogenic potential of hMSCs as a function of passage number clearly indicate that such elevations in APase activity in response to OS consistently occur in the same high proportion of cells derived from every passage (Figure 7a). The formation of characteristic mineralized nodular aggregates throughout the culture dish was also observed in specimens from every passage. Furthermore, the fact that the basal APase activity remains low in Control samples derived from extensively passaged cells supports our assertion that prolonged tissue culture does not result in the selective outgrowth of a subpopulation of MSCs with a predilection to become osteogenic. Although comprehensive studies showing other mesenchymal developmental potentials of hMSCs at every passage from a single donor have not been performed, chondrogenesis (Johnstone et al., 1996; Lennon et al., 1996) and stromagenesis (Majumdar et al. , 1995) are routinely observed in preparations derived from first through fourth passage. Beyond the osteoinductive effect which OS has on hMSCs, OS also acts as a mitogen (Figure 6b). Figure 7b illustrates that this mitogenic response of hMSCs to OS is also conserved following serial subculture. These data, together with studies of in vivo osteogenesis of passaged hMSCs (Haynesworth et al., 1992b), demonstrate that the osteogenic potential and mitogenic response of hMSCs to OS following extensive subculture is not diminished. Therefore, it appears that the replicative senescence which occurs in late passage is not due to terminal differentiation of Type II cells, as suggested by Mets and Verdonk (1981). Certain techniques relating to the way the hMSCs are initially isolated, grown, and subcultivated undoubtedly contribute to the peφetuation of these cells without lineage progression, not the least of which is the selection of an appropriate lot of fetal bovine serum (Lennon et al. , 1996).

We also sought to determine whether cryopreservation affected either the proliferative capacity of these cells, or their developmental potential. Such cryogenic manipulation is particularly interesting in view of the fact that other investigators use cryopreservation in their technique for selecting and isolating mesenchymal stem cells from tissues other than bone marrow (Young et al., 1991, 1993, 1995). In our hands, cell recovery following thawing routinely exceeds 95 % , and is consistent with the viability of numerous cell lines and hybridomas routinely maintained in our laboratory. Therefore, results obtained from experiments with these cells cannot be construed to reflect the activity of a subpopulation of the originally frozen hMSCs. As presented in Figure 9a, cryopreserved cells which are subcultured in parallel with cells never frozen possess the same low basal APase activity. Importantly, in response to OS (Figure 9b), the increase in APase activity of cryopreserved hMSCs mirrors OS's effect on non-cryopreserved cells and, for the most part, is not statistically different. That is, in passages where the absolute activity is either relatively low or high, both fresh and frozen cells from the same donor behave identically. The same phenomenon holds true with respect to the proliferation and mitogenic response of cells cultured in the absence or presence of OS (Figure 10). These observations suggest, as others have previously shown (Hayflick and Moorehead, 1961; Hayflick, 1965, 1989), that cells have a "memory", which is primarily influenced by their duration and manipulation in tissue culture rather than their calendar age or the process of cryopreservation.

In view of the above observations, it may now be possible to explore clinical alternatives using self cell therapies which have never before been available (Caplan and Bruder, 1997). Since the reduction of bone mass in osteoporosis, normal aging, and a variety of other diseases has been linked with a diminution in the number and activity of marrow-derived osteoprogenitor cells (Tabuchi et al., 1986; Tsuji et al, 1990; Egrise et al., 1992; Quarto et al., 1995; Kahn et al., 1995), one idealized therapeutic goal would be to rejuvenate the supply of osteogenic progenitor cells, or MSCs. In the absence of clinical pharmacologic factors capable of mitotically stimulating the dwindling supply of endogenous MSCs in these patient populations, one alternative is to harvest the patient's own rare MSCs from their marrow, mitotically expand them ex vivo, and reinfuse them back into the host. The present study demonstrates that MSCs can be cryopreserved and expanded over one billion fold ex vivo without a loss in their osteogenic potential (Table 1). Therefore, for the treatment of diseases based on an inadequate supply of MSCs, we would propose the periodic administration of cryopreserved, autologous MSCs based on the safe intravenous infusion protocol recently reported (Lazarus et al. , 1995). While such massive proliferation of cells may be required for life-long dosaging in these metabolically compromised patients, a sufficient number of cells for the regeneration of local bone defects can be obtained in one or two passages (Bruder et al. , 1994; Kadiyala et al. 1997). Successful application of this technology for the treatment of metabolic, genetic, and structural skeletal abnormalities shall serve to guide aspects of our future efforts.

Cited Literature

Bell, E., Marek, L.F., Levinstone, D.S., Merrill, C, Sher, S., Young, I.T. , and Eden, M. (1978) Loss of division potential in vitro: Aging or differentiation. Science 202: 1158-1163.

Beresford, J.N. (1989) Osteogenic stem cells and the stromal system of bone and marrow. Clin. Orthop. Rel. Res. 240:270-280.

Bruder, S.P., Fink, D.J., and Caplan, A.I. (1994) Mesenchymal stem cells in bone formation, bone repair, and skeletal regeneration therapy. J. Cell Biochem. 56(3) :283-294.

Bruder, S.P., Lawrence, E.G., and Haynesworth, S.E. (1995a) The generation of monoclonal antibodies against human osteogenic cells reveals embryonic bone formation in vivo and differentiation of purified mesenchymal stem cells in vitro. Trans. Ortho. Res. Soc. 20. 8.

Bruder, S.P., Eames, B.F., and Haynesworth, S.E. (1995b) Osteogenic induction of purified human mesenchymal stem cells in vitro: Quantitative assessment of the osteoblastic phenotype. Trans. Ortho. Res. Soc. 20. 464.

Caplan, A.L (1991) Mesenchymal Stem Cells. J. Orthop. Res. 9:641-650.

Caplan, A.I. and Bruder, S.P. (1997) Cell and molecular engineering of bone regeneration. In Textbook of Tissue Engineering. R. Lanza, R. Langer, and W. Chick, eds. R.G. Landes and Co., Georgetown, pgs.603-618.

Caplan, A.L, Fink, D.J., Goto, T., Linton, A.E., Young, R.G., Wakitani, S., Goldberg, V.M., and Haynesworth, S.E. (1993) Mesenchymal stem cells and tissue repair. In The Anterior Cruciate Ligament: Current and Future Concepts. D.W. Jackson, ed. Raven Press, Ltd., New York. 405-417. Dexter, T.M., and Testa, G.M. (1976) Differentiation and proliferation of haemopoietic cells in culture. In Methods in Cell Biology, Vol 14. D.M. Prescott, ed. Academic Press, New York. 387-405.

Egrise, D., Martin, D., Vienne, A., Neve, P., and Schoutens, A. (1992) The number of fibroblastic colonies formed from bone marrow is decreased and the in vitro proliferation rate of trabecuiar bone cells increased in aged rats. Bone 73:355-361.

Hall, P.A. and Watt, F.M. (1989) Stem cells: The generation and maintenance of cellular diversity. Development. 706:619-633.

Hariey, C.B., Futcher, A.B., and Greider, C.W. (1990) Telomeres shorten during ageing of human fibroblasts. Nature 345: 458-460.

Hayflick, L. (1965) The limited in vitro lifetime of human diploid cell strains. Exp. Cell Res. 37:614-636.

Hayflick, L. (1989) Antecedents of aging research. Exp. Gerontol. 24:355-365.

Hayflick, L. and Moorehead, P. (1961) The serial cultivation of human diploid cell strains. Exp. Cell Res. 25:585-621.

Haynesworth, S.E., Baber, M.A., and Caplan, A.I. (1992a) Cell surface antigens on human marrow-derived mesenchymal cells are detected by monoclonal antibodies. Bone 73: 69-80.

Haynesworth, S.E., Goshima, J., Goldberg, V.M., and Caplan, A.I. (1992b) Characterization of cells with osteogenic potential from human marrow. Bone. 73:81-88. Haynesworth, S.E., Baber, M.A., and Caplan, A.I. (1995) Characterization of the unique mesenchymal stem cell phenotype in vitro. Trans. Ortho. Res. Soc. 20:7.

Haynesworth, S.E., Baber, M.A., and Caplan, A.I. (1996) Cytokine expression by human marrow-derived mesenchymal progenitor cells in vitro: Effects of dexamethasone and IL-lα. J. Cell Physiol. 766(3;:585-592.

Jaiswal, N., Haynesworth, S.E., Caplan, A.I., and Bruder, S.P. (1997) Osteogenic differentiation of purified, culture-expanded human mesenchymal stem cells in vitro. J. Cell Biochem., 64:295-312.

Johnstone, B., Yoo, J.U., Barry, F.P. (1996) In vitro chondrogenesis of bone marrow-derived mesenchymal cells. Trans. Ortho. Res. Soc. 27: 64.

Kadiyala, S., Jaiswal, N., and Bruder, S.P. (1997) Culture Expanded, Bone Marrow Derived MSCs Can Regenerate a Critical Sized Segmental Bone Defect, Tissue Engineering, Vol.3 (in press).

Kahn, A., Gibbons, R., Perkins, S., and Gazit, D. (1995) Age-related bone loss: A hypothesis and initial assessment in mice. Clin. Orthop. Rel. Res. 373:69-75.

Karatza, C, Stein, W.D., and Shall, S. (1984) Kinetics of in vitro ageing of mouse embryo fibroblasts. J. Cell Sci. 65: 163-175.

Lazarus, H.M. , Haynesworth, S.E., Gerson, S.L., Rosenthal, N.S., and Caplan, A.I. (1995) Ex vivo expansion and subsequent infusion of human bone marrow- derived stromal progenitor cells (mesenchymal progenitor cells): Implications for therapeutic use. Bone Marrow Transpl. 76:557-564.

Lemischka, I.R., Raulet, D.H., and Mulligan, R.C. (1986) Developmental potential and dynamic behavior of haematopoietic stem cells. Cell. 45:917-927. Lennon, D.P., Haynesworth, S.E., Young, R.G., Dennis, J.E., and Caplan, A.I.

(1995) A chemically defined medium supports in vitro proliferation and maintains the osteochondral potential of rat marrow-derived mesenchymal stem cells. Exp. Cell Res. 279:211-222.

Lennon, D.P., Haynesworth, S.E., Bruder, S.P., Jaiswal, N., and Caplan, A.I.

(1996) Development of a serum screen for mesenchymal progenitor cells from bone marrow. In Vitro Animal Cell. Dev. Bio., 32:602-611.

Majumdar, M.K., Haynesworth, S.E., Thiede, M.A., Marshak, D.R., Caplan, A.L, and Gerson, S.L. (1995) Culture-expanded human mesenchymal stem cells (MSCs) express cytokines and support hematopoiesis in vitro. Blood, S6(70):494a, #1966.

Martin, G.M., Sprague, C.A., and Epstein, CJ. (1970) Clonal selection, attenuation and differentiation in an in vitro model of hypeφlasia. Lab. Invest. 23:86-92.

Martinez, A.O., Norwood, T.H., Prothero, J.W., and Martin G.M. (1978) Evidence for clonal attenuation and differentiation in an in vitro model of hypeφlasia. Am. J. Pathol. 74: 137-154.

Matsumura, T., Pfendt, E.A., and Hayflick, L. (1979) DNA synthesis in the human diploid cell strain WI-38 during in vitro ageing: An autoradiography study. J. Gerontol. 34:323-327.

Mets, T., and Verdonk, G. (1981) In vitro aging of human bone marrow-derived stromal cells. Mech. Ageing Dev. 76: 81-89.

Mitsui, Y., and Schneider, E.L. (1976) Characterization of fractionated human diploid fibroblast cell populations. Exp. Cell Res. 703: 23-30. Orgel , L.E. (1973) Ageing of clones of mammalian cells. Nature 243:441-445.

Owen, M. (1985) Lineage of osteogenic cells and their relationship to the stromal system. In Bone and Mineral/3. W.A. Peck, ed. Elsevier, Amsterdam, 1-25.

Owen, M. (1988) Marrow stromal stem cells. J. Cell Sci. Suppl. 70:63-76.

Quarto, R., Thomas, D., and Liang, T. (1995) Bone progenitor cell deficits and the age-associated decline in bone repair capacity. Calcif. Tissue Int. 56: 123- 129.

Rheinwald, J.R., and Green, H. (1975) Serial cultivation of strains of human epidermal keratinocytes: The formation of keratinizing colonies from single cells. Cell 6:331-344.

Sachs, L. (1987) The molecular control of blood cell development. Science. 235:1374-1379.

Sager, R. (1989)Tumor suppressor genes: The puzzle and the promise. Science 246:1406-1412.

Smith, J.R., Pereira-Smith, O., and Schneider, E.L. (1978) Colony size distribution as a measure of in vivo and in vitro age. Proc. Natl. Acad. Sci. 75(3): 1353-1356.

Smith, J.R., and Whitney, R.G. (1980) Intraclonal variation in proliferative potential of human diploid fibroblasts: Stochastic mechanism for cellular aging. Science 207:82-84.

Spangrude, G.J., Heimfeld, S. Weissman, LL. (1988) Purification and characterization of mouse hematopoietic stem cells. Science. 724:58-61. Szilard, L. (1959) On the nature of the aging process. Proc. Natl. Acad. Sci. 45: 30-45.

Tabuchi, C, Simmon, D.J., Fausto, A., Russell, J., Binderman, I., and Avioli, L. (1986) Bone deficit in ovariectomized rats. J. Clin. Invest. 78:637-642.

Thornton, S.C, Mueller, S.N., and Levine, E.M. (1983) Human endothelial cells: Use of heparin in cloning and long-term serial cultivation. Science 270:623-625.

Tsuji, T., Hughes, F.J., McCulloch, C.A., and Melcher, A.H. (1990) Effect of donor age on osteogenic cells of rat bone marrow in vitro. Mech. Ageing Dev. 57: 121-132.

Wang, E., Lee, M., and Pandey, S. (1994) Control of fibroblast senescence and activation of programmed cell death. J Cell. Biochem. 54: 432-439.

Westergren-Thorsson, G., ...nnerevik, P-O., Fransson, L-A., and Maelstrδm, _. (1991) Proliferation of cultured fibroblasts is inhibited by L-iduronate-containing glycosaminoglycans. J. Cell Physiol. 747:523-530.

Young, H.E., Morrison, D.C, Martin, J.D., and Lucas, P.A. (1991) Cryopreservation of embryonic chick myogenic lineage-committed stem cells. J. Tissue Cult. Methods 73:275-284.

Young, H.E., Ceballos, E.M., Smith, J.C, Mancini, M.L., Wright, R.P., Ragan, B.L., Bushell, I., and Lucas, P.A. (1993) Pluripotent mesenchymal stem cells reside within avian connective tissue matrices. In Vitro Cell. Dev. Biol. 29,4:723-736. Young, H.E., Mancini, M.L., Wright, R.P., Smith, J.C, Black, A.C, Reagan, C.R., and Lucas, P.A. (1995) Mesenchymal stem cells reside within connective tissues of many organs. Dev. Dynam. 202: 137-144.

Claims

What Is Claimed Is:

1. A cryopreserved preparation comprising an isolated, homogeneous population of viable human mesenchymal stem cells which can differentiate into cells of more than one connective tissue type upon being restored from cryopreservation.

2. The mesenchymal stem cell preparation of claim 1 wherein an antibody produced from a hybridoma cell line selected from the group consisting of SH2 (ATCC accession number HB 10743), SH3 (ATCC accession number HB 10744) and SH4 (ATCC accession number HB 10745), which antibody binds to said mesenchymal stem cells upon their being restored from cryopreservation.

3. The mesenchymal stem cell preparation of claim 2 wherein said mesenchymal stem cells have been culture-expanded by serial passaging.

4. The mesenchymal stem cell preparation of claim 3 wherein said mesenchymal stem cells adhere to a plastic surface when cultured in a complete or serum-free medium upon being restored from cryopreservation.

5. The mesenchymal stem cell preparation of claim 3 wherein said mesenchymal stem cells retain the potential to differentiate into more than one connective tissue type after population expansion in culture upon being restored from cryopreservation.

6. The mesenchymal stem cell preparation of claim 1 wherein said mesenchymal stem cells are obtained from a source selected from the group consisting of bone marrow, periosteum, cord blood, peripheral blood, dermis and muscle.

7. The mesenchymal stem cells of claim 6 wherein said mesenchymal stem cells adhere to a plastic surface when cultured in a complete or serum-free medium upon being restored from cryopreservation.

8. The mesenchymal stem cell preparation of claim 1 wherein said mesenchymal stem cells are non-embryonic cells.

9. The mesenchymal stem cell preparation of claim 1 wherein one of said connective tissue types is bone.

10. The mesenchymal stem cell preparation of claim 1 wherein one of said connective tissue types is cartilage.

11. The mesenchymal stem cell preparation of claim 1 wherein one of said connective tissue types is adipose.

12. The mesenchymal stem cell preparation of claim 1 wherein one of said connective tissue types is tendon.

13. The mesenchymal stem cell preparation of claim 1 wherein one of said connective tissue types is ligament.

14. The mesenchymal stem cell preparation of claim 1 wherein one of said connective tissue types is dermis.

15. The mesenchymal stem cell preparation of claim 1 wherein one of said connective tissue types is muscle.

16. The mesenchymal stem cell preparation of claim 1 wherein one of said connective tissue types is a marrow stromal connective tissue which supports the differentiation of hematopoietic stem cells.

17. The mesenchymal stem cell preparation of claim 1 wherein the human mesenchymal stem cells were cryopreserved in a composition selected from the group consisting of those comprising: (a) 1:1 Biowhittaker: DMEM-low glucose 955; (b) 90% autologous serum + 10% DMSO845; (c) 95% autologous serum + 5% DMSO875; (d) 90% DMEM-lg + 10% DMSO875; (e) 85% DMEM-lg + 5% autologous serum + 10% DMSO955 and (f) 90% fetal bovine serum + 10% DMSO.