Stromal Cell-Associated Hematopoiesis: Immortalization and Characterization of a Primate Bone Marrow-Derived Stromal Cell Line

Home , Stromal cell

By S.R. Paul, Y-C. Yang, R.E. Donahue, S. Goldring, and D.A. Williams

An elucidation of the interaction between the bone marrow etic cells in long-term cultures for up to 4 weeks as measured microenvironment and hematopoietic stem cells is critical to by in vitro progenitor assays. Studies were undertaken to the understanding of the molecular basis of stem cell self characterize the products of extracellular matrix biosynthe- renewal and differentiation. This interaction is deaendent, at sis and growth factor synthesis of this cell line, designated least in part, on direct cell to cell contact or cellulafipdhesion PU-34. In contrast to most murine bone marrow-derived to extracellular matrix proteins. Long-term bone marrow stromal cell lines capable of supporting hematopoiesis in cultures (LTMC) provide an appropriate microenvironment vitro that have been examined, the extracellular matrix for maintenance of primitive hematopoietic stem cells and a produced by this primate cell line includes collagen types 1, means of analyzing this stem cell-stromal cell interaction in 111, and V, and fibronectin rather than collagen type IV and vitro. Although LTMC have been successfully generated from laminin. Growth factor production analyzed through RNA murine and human bone marrow, only limited success has blot analysis, bone marrow cell culture data, and factor- been reported in a primate system. In addition, few perma- dependent cell line proliferation assays includes interleukin-6 nent stromal cell lines are available from nonmurine bone (IL-6). IL-7, granulocyte-macrophage colony-stimulating fac- Downloaded from http://ashpublications.org/blood/article-pdf/77/8/1723/605635/1723.pdf by guest on 24 September 2021 marrow. Because the primate has become a useful model for tor (GM-CSF), G-CSF, M-CSF, leukemia inhibitory factor, and large animal bone marrow transplant studies and, more a novel cytokine designated IL-I1. This immortalized primate specifically, retroviral-mediated gene transfer analysis, we bone marrow stromal cell line may be useful in ‘maintaining have generated immortalized bone marrow stromal cell lines early progenitor cells for experimental manipulation without from primate bone marrow using gene transfer of the Simian the loss of reconstituting capacity and as a potential source virus large T (SV40 LT) antigen. At least one stromal cell line of novel hematopoietic growth factors. has demonstrated the capacity to maintain early hematopoi- 0 1991 by The American Society of Hematology.

HE HEMATOPOIETIC microenvironment (HM) con- marrow transplantation in which reconstitution of HSC can T sists of an organized array of endothelial cells lining be studied in a large outbred specie^.'^"^ Cloned human the marrow vasculature with a network of fibroblasts, hematopoietic growth factors are biologically active on adventitial reticular cells, macrophages, and adipocytes in primate cells, allowing the evaluation of cytokine effects on the hematopoietic cords.’’2Direct contact with some or all primate LTMC and reconstitution after bone marrow of these cells by hematopoietic stem cells (HSC) has been transplantation.’8~21In addition, hematopoietic cytokines demonstrated in situ and in vitro, and may play a critical may also be useful during in vitro manipulations of HSC role in stem cell proliferation and differentiati~n.~.~How- required for gene transfer experiment^^^,*^ while cells mak- ever, cellular interactions between the HSC and the mi- ing up the HM may provide a useful environment for croenvironment are not well understood, largely due to the maintaining transduced HSC cells for prolonged periods in heterogeneity of the cells making up the HM. Although a vitro, without the loss of reconstituting ca~acity.’~ number of growth factors affecting hematopoietic cells have LTMC are made up of a heterogeneous papulation of recently been the role of these factors in the stromal cells.” The role of the individual cell types in this complex culture system may be studied by. the development maintenance of the stem cell compartment is not clear. of cloned stromal cell lines. Murine stromal cell lines Indeed, stem cell sunrival in vitro is dependent, at least in generated by repeated in vitro passage have been demon- part, on direct stem cell contact to the HM’,’ or cellular strated to support the proliferation of pre-B lymphocytes, adhesion to extracellular matrix proteins elaborated by myeloid progenitors, and factor-dependknt cell these ~ells.8.~Furthermore, recent studies by several labora- .. tories suggest that several known growth-regulating factors may affect hematopoietic cells indirectly through effects on From the Howard Hughes Medical Institute, The Children’s Hospi- stromal cells.’o-’2Therefore, it appears that the optimum tal and Dana Farber Cancer Institute, Massachusetts General Hospi- environment for maintenance, proliferation, and differenti- tal, Harvard Medical School, Boston; and the Genetics Institute, ation of stem cells both in vivo and in vitro is dependent on Cambridge, M. a complex interaction between hematopoietic growth fac- Submitted August 6, 1990; accepted December IS, 1990. S.R.P. is supported by National Institutes of Health (N1H)-NHLBI tors and the cells and matrix proteins of the HM. training Grant 5T32 HL07574 and the David Abraham Fellowship of Long-term marrow cultures (LTMC) provide an in vitro the DFCI. DA.W is supported by CIA Award 5 KO8 HLO1554-02 and method of studying the HM.1.’3,14Murine LTMC have been Program Project Grants NIH CA39542-05 and NIH HL32262. shown to be capable of supporting colony-forming units- Primate work supported by NIH Center Grant No. RR00168. spleen (CFU-S), reconstituting HSC, and myeloid and Address reprint requests to David A. williams, MD, Assistant lymphoid hematopoiesis in vitro for up to 4 months. Professor of Pediatrics, Howard Hughes Medical Institute, Children’s Although LTMC have also been successfully generated Hospital, Enders 7, 300 Longwood Ave, Boston, M 02115. using human bone marrow,I5 cultures of primate bone The publication costs of this article were defr’ayed in part by page charge payment. This article must therefore be hereby marked marrow have yielded inferior long-term results because of “advertisement” in accordance with 18 U.S.C.section 1734solely to the lack of stable stroma in these cultures. Primate LTMC indicate this fact. would be useful for several reasons. The nonhuman pri- 0 1991 by The American Sociery of Hematology mate is a desirable experimental model for autologous bone 0006-497119117708-0016$3.00/0

Blood, Vol77, No 8 (April 151,1991: pp 1723-1733 1723 1724 PAUL ET AL

However, few stromal cell lines from human bone marrow selected by addition of 0.5 mgimL of G418 (dry powder) (GIBCO). have been characteri~ed~*~~and none are available from Well-isolated G418-resistant colonies were picked on day 21 and primates. Such stromal cell lines could provide a unique expanded into multiwel1 plates in complete long-term culture resource for the isolation, characterization, and cloning of medium. Once established, the cell lines were maintained at 37°C. cytokines, which may affect hematapoietic cells?" For One such cell line was extensively analyzed and has been desig- example, interleukin-7 (IL-7) was identified from a murine nated PU-34. Analysis of support of hematopoiesis by stromal cell lines. stromal cell line isolated from long-term B-lymphoid cul- Confluent PU-34 cells were incubated with 10 pgimL mitomycin C ture~.~~ (Boehringer-Mannheim Co, Mannheim, Germany) for 2 hours at Work in our laboratory and by other investigators has 37°C and 5% CO,. Subsequently, cells were trypsinized, washed shown that retrovirus vectors containing oncogenes can extensively, and replated on 0.1% gelatin coated 25-cm2 tissue immortalize primary cells from several tissues of adult culture flasks (Corning Glass Works, Corning, NY)at a concentra- animals and We hypothesized that an in vitro tion of 1 x lo6cells/flask. Primate bone marrow mononuclear cells culture system could be established using immortalized were depleted of adherent cells by two passages on fresh tissue cloned primate bone marrow stromal cells generated using culture flasks in 24 hours. These low-density bone m irrow cells such recombinant retrovirus vectors as previously described depleted of plastic-adherent cells were charged onto confluent Downloaded from http://ashpublications.org/blood/article-pdf/77/8/1723/605635/1723.pdf by guest on 24 September 2021 with murine bone marrow.32We report here the immortal- mitomycin C-treated PU-34 cells. Nonadherent cells were removed ization of multiple primate bone marrow stromal cell lines at weekly intervals and assayed in semisolid medium for the presence of progenitor cells. Cultures were plated with lo5nonad- using a recombinant retrovirus vector containing the SV40 herent cells/mL in IMDM, 0.9% methylcellulose, 1% bovine serum LT antigen and neo phosphotransferase (Neo) gene. One albumin (BSA), mol/L P-mercaptoethanol, 100 U/mL penicil- such line, PU-34, was capable of maintaining primitive lin, and 100 pg/mL streptomycin. Methylcellulose cultures were hematopoietic progenitor cells in long-term culture. This supplemented with 10 ng/mL granulocyte-macrophage colony- cell line was further characterized to establish the features stimulating factor (GM-CSF) and 2 U/mL erythropoietin (supplied responsible for this effect. by Genetics Institute, Cambridge, MA). Cultures were plated in 1 mLvolumes in 35-mm tissue culture plates in 5% CO, at 37°C and colonies were counted on day 12. Support of progenitors was MATERIALS AND METHODS calculated based on the number of CFU per lo5cells and the total Generation ofpnmate stromal cell lines. Details of the recombi- number of recovered nonadherent cells per flask at each weekly nant retroviral vector, U19, have been previously rep~rted.~' feeding. These data do not reflect adherent CFU present in or on Briefly, the SV40 early region viral sequences from the Bgl I site to the stromal cells. the Hpa I site were inserted into the BamHI site of the pZip Neo Protein analysrs. Extracellular matrix components (ECM) and SV(X) 1 vector.34 An amphotropic producer line, U19BL, was intermediate filament proteins were analyzed by two methods: generated by infection of $AM cell? with ecotropic viral harvest immunofluorescence microscopy and sodium dodecyl sulfate poly- from the U19-5 cell line36and produces recombinant amphotropic acrylamide gel electrophoresis (SDS-PAGE). PU-34 cells were SV40 large T virus at a titer of 5 X lo3 G418 resistant CFU/mL grown to confluence in Labtek chambers (Nunc, Inc, Naperville, when assayed on NIW3T3 cells. The producer line was maintained IL) and fixed in 3.7% formaldehyde for 4 minutes followed by in Dulbecco's modified Eagle medium (DME) supplemented with 100% methanol for 1 minute. Cells were rinsed once each with 10% fetal calf serum (FCS), 100 U/mL petlicillin, and 100 pg/mL water and phosphate-buffered saline (PBS), then 0.5 mL of the streptomycin (all from GIBCO, Grand Island, NY). primary antibody was added and incubated for 1hour at 3PC. The Adult macaques were anesthetized with 0.5 mL Ketamine HCI primary antibodies consisted of 1:50 dilution of Pab 41937(SV40 (Parke-Davis, Morris Plains, NJ) given intramuscularly and placed LT) (kindly provided by Dr Ed Harlow, Cold Spring Harbor in a ventral recumbent position. The posterior superior iliac crest Laboratories, NY);1:100 dilution of anti-fibronectin; 1:50 dilution was identified, the hair was shaved, and the site prepared surgically of anti-laminin; a 1:100 dilution of anti-type IV cellagen (Collabo- with betadine solution and 100% ethanol. An 18-gauge disposable rative Research, Bedford, MA); a 1:5 dilution of anti-vimentin; 1:3 Jamshedi bone marrow needle (Baxter Care Corp, Valencia, CA) dilution of anti-neurofilament; 1:50 dilution of anti-cytokeratin; 1:3 was inserted and 5 mL of bone marrow aspirated. These cells were dilution of anti-desmin (all Boehringer Mannheim); and a 1:5 subjected to Ficoll-Paque (Pharmacia, Uppsala, Sweden) density dilution of anti-Factor VIII-associated antigen (Dako, Santa centrifugation at 1,500 rpm for 30 minutes. LTMC were estab- Barbara, CA). After 1 hour of incubation, the primary antibody lished from the mononuclear cells in Iscove modified Dulbecco was aspirated, the cells were washed thrice with PBS, and 0.5 mL of medium (IMDM) with 10% horse serum, 10% FCS, 100 UimL the appropriate rhodamine-conjugated secondary ptibody was penicillin, and 100 pghnL streptomycin (all from GIBCO) (com- added and incubated with the cells for 1 hour at 37°C. Goat plete long-term culture media). The cultures were fed weekly by antirabbit IgG (Tago, Burlingame, CA, 1:lOO dilution) was used as removal of 75% of medium and cells followed by addition of fresh a secondary antibody for anti-fibronectin, anti-Factor VIII- medium. Cultures were maintained at 33°C and 5% CO,. associated antigen, anti-laminin, gnti-neurofilament, and anti-type Supernatants containing the U19 virus were harvested from IV collagen. Goat antimouse IgG (Tago; 1:lOO dilution) was used confluent plates of U19BL producer cells 18 hours after the as the secondary antibody for anti-desmin, anti-vimentin, and Pab addition of fresh medium, filtered through 0.45-km filters (Nal- 419 primary antibodies. Omission of the relevant first antibody with gene, Rochester, NY) and frozen at -70°C until use. Adherent inclusion of 5% FCS served as controls for nonspecific staining. cells of primate LTMC were infected 7 and 10 days after establish- After incubation with the second antibody, slides were washed ment with 2 mL of viral stock in the presence of 8 WgjmL of twice with PES, once with water, tha two drops of gelvatol added polybrene (Aldrich Chemical Co, Inc, Milwaukee, WI) at 33°C for and cover slips were placed on each slide. Immunofluorescent- 2.5 hours. Subsequently, the cultures were fed in the usual manner. stained cells were viewed with a Ziess fluorescence microscope Simultaneous titer of the same viral stock was determined on using a 546-nm excitor filter with a 15-nm band path, FT 580 NIW3T3 as previously described." Beginning 14 days after estab- mirror, and 590-nm barrier filter. lishment of long-term cultures, G418-resistant stromal ceHs were SDS-PAGE analysis of metabolical!y labeled proteins was per- STROMAL-ASSOCIATED HEMATOPOIESIS 1725 formed as previously described.)* Briefly, PU-34 cells were plated wells were then pulsed with 0.5 pCi of tritiated thymidine (Du- at a density of 2 X 105 cellddish in 3.5-cm tissue culture dishes. pont). After 4 hours of incubation the cells were harvested and After 2 days, cells were incubated for 24 hours at 37°C in 5% C02in counted. IMDM without serum with 50 pg/mL P-aminopropionitrile, 50 TF-1 cells, a human erythroleukemia line dependent on IL-3, pg/mL ascorbic acid, and 2 mmol/L glutamine (all Sigma Chemical IL-4, IL-5, IL-6, GM-CSF, and LIF, were maintained in RPMI Co, St Louis, MO) with 25 pCi of L-[5-3H] proline (30 mCi/mol; supplemented with 5% FCS, 100 UimL penicillin, 100 UimL Amersham-Searle, Corp, Arlington Heights, IL). Labeled medium streptomycin, and 200 mmoliL glutamine. Samples of PU-34 CM proteins were collected at 24 hours and analyzed by SDS-PAGE were added into 96-well plates containing 5 x 103 TF-1 cellsiwell in (5% acrylamide) with or without reduction with 0.5% P-mercapto- the presence or absence of neutralizing antibody to human ethan01.’~ 14C-labeled rat tail tendon collagen was used as a GM-CSF or human IL-6 (both supplied by Genetics Institute). molecular weight marker. The labeled collagen in the medium was Plates were wrapped in parafilm and incubated for 3 days. The further characterized after pepsinization at 4°C and the collagens wells were then pulsed with 0.5 pCi of tritiated thymidine. After 4 analyzed by SDS-PAGE (5% acrylamide) with delayed reduction hours of incubation, the cells were harvested and counted. to distinguish alpha ](I) from alpha 1(III) collagen chains.” T1165 cells, a murine plasmacytoma line dependent on murine Fluorograms of the gels were prepared as described el~ewhere.’~ IL-1, IL-6, murine GM-CSF, and tumor necrosis factor (TNF),

Growth factor anulysis. Poly A+ RNA was prepared by stan- were maintained in RPMI supplemented with 10% FCS, 2 mmoliL Downloaded from http://ashpublications.org/blood/article-pdf/77/8/1723/605635/1723.pdf by guest on 24 September 2021 dard methods using the guanidinium isothiocyanate procedure39 glutamine, 100 mg/mL penicillin, 100 pg/mL streptomycin, 5 x from uninduced PU-34 cells and from PU-34 cells that had been mol/L p-mercaptoethanol and rh IL-6. T1165 cells were plated at 5 induced for 24 hours with 2 UimL recombinant human (rh) IL1-a x 10‘ cells/well in 96-well tissue culture plates with samples of (Cistron, Pinebrook, NJ). RNA samples were fractionated on a 2.2 PU-34 (CM) in the presence or absence of antibody to human IL-6 mol/L formaldehyde-1% agarose gel, transferred to nylon filters (supplied by Genetics Institute). Cells were incubated for 2 days, (Gene Screen, New England Nuclear Research Products, Boston, then pulsed for 4 hours with tritiated thymidine. Labeled cells were MA), and hybridized with ”P-labeled cDNA probes for GM-CSF, harvested and counted. G-CSF, 1L-3, IL-4, IL-5, IL-6, IL-7, IL-9, and leukemia inhibitory Human bone marrow cells were obtained from normal donors in factor (LIF) (DNAs supplied by Genetics Institute, Cambridge, accordance with the guidelines of the Human Investigation Com- MA). Prehybridization, hybridization and posthybridization washes mittee of the Children’s Hospital, Boston, MA. Human bone of filters were performed as recommended by the manufacturer. marrow progenitor assays were performed with mononuclear cells Filters were exposed to X-ray film in the presence of a calcium isolated using Ficoll-Paque seperation and depleted of plastic- tungstate intensifying screen at -70°C. adherent cells by two passages on fresh tissue culture flasks over 24 Conditioned media (CM) from uninduced or induced PU-34 hours. Mononuclear cells, los, were plated in IMDM with 0.3% (IL1-a for 48 hours) were tested for cytokine activity on a series of agarose, 20% FCS, 100 UimL penicillin, 100 pgimL streptomycin, factor-dependent cell lines through analysis of tritiated thymidine and mol/L p-mercaptoethanol at 37°C and 5% CO,. These incorporation. Factor dependent cell lines included the human cultures were plated with various dilutions of PU-34-conditioned megakaryoblastic leukemia cell line, Mo7E4’; a human erythroleu- media and colonies were evaluated after 10 days of incubation. kemia cell line, TF14’;two murine IL-3-dependent cell lines, DA2” Analysis of constitutive expression of M-CSF was performed using and 32D43; and a murine plasmacytoma cell line, T1165.MThe murine CFU-M assay. Freshly isolated C3HMeJ bone marrow was assays and culture conditions of these lines have been described in plated at 1 x 105imLin IMDM methylcellulose (0.9%) containing the individual references. Briefly, the Mo7E cell line is a human 20% FCS, 100 U/mL penicillin, 100 &mL streptomycin, and megakaryoblastic leukemia line dependent on IL-3, IL-4, IL-9, and moliL P-mercaptoethanol. Colonies containing greater than 50 GM-CSF. Mo7E cells were maintained in DME with 10% FCS, 100 cells were counted after 10 days at 37°C and 5% CO,. UimL penicillin, 100 pgimL streptomycin, and 200 mmol/L glu- Erythroid activity was evaluated through burst promoting activ- tamine. The cells were resuspended in 96-well plates to a final ity (BPA) assays. Peripheral blood progenitor cells from leukopaks concentration of 5 x lo’ cellsiwell. Samples of PU-34 CM were were acquired from patients undergoing platelet pheresis in the added in the presence or absence of neutralizing antibody to blood component lab of the Children’s Hospital in Boston, MA, in human GM-CSF (supplied by Genetics Institute, Boston, MA). accordance with guidelines of the Human Investigation Commit- Plates were wrapped in parafilm and incubated for 3 days. The tee. Peripheral cells were separated over a Ficoll-Paque gradient. wells were pulsed with 0.5 pCi of tritiated thymidine (Dupont, The mononuclear cells were then washed and resuspended at 5 x Wilmington, DE) for 4 hours after which the cells were harvested 10’ cellsimL and placed on ice. Cold-induced aggregates (Agg’) and counted with an LKB flatbed scintillation counter (Pharma- were collected, layered on FCS, and allowed to settle by unit cia, Piscataway, NJ). gravity through the FCS at 4°C. The resulting cells were incubated CM from PU-34 was also tested for mitogenic activity on the with carbonyl iron and phagocytic cells magnetically separated. DA-2 cell line, which is responsive to murine IL-3, murine IL-4, These Fe- cells were further depleted of monocytes/macrophages G-CSF, and LIF. DA-2 cells were maintained in RPMI supple- by adherence depletion (Ad-) on tissue culture plates at 37°C for 1 mented with 5% FCS and 1,000 UimL rhLIF (supplied by Genetics to 2 hours. The resulting Agg+, Fe-, Ad- cells were plated at IO5 Institute). Samples of PU-3konditioned media were added into cellsiml in IMDM, 0.9% methylcellulose, 30% FCS, mol/L 96-well tissue culture plates containing lo4DA-2 cellsiwell. Plates p-mercaptoethanol, 100UimL penicillin, and 100 pgimL streptomy- were wrapped in parafilm and incubated for 3 days. Subsequently cin. Conditioned media from IL1-ct-induced PU-34 cells were each well was pulsed with 0.5 mCi of tritiated thymidine. After 4 added at dilutions of 1:lO and 150 and burst forming units- hours of incubation, the cells were harvested and counted. erythroid (BFU-E) were counted at 12 days. 32-D cells, which are dependent on IL-2, murine IL-3, G-CSF, and murine GM-CSF, were maintained in RPMI supplemented RESULTS with 5% FCS, 100 U/mL penicillin and 100 pdmL streptomycin. Samples of PU-34 CM were added in the presence or absence of Generation of cell lines. Infection of primate LTMC neutralizing antibody to human G-CSF (supplied by Genetics with the amphotropic U19 virus and selection with G418 led Institute) into 96-well plates containing 2 x lo4 32-D cells/well. to the appearance of multiple macroscopic G418-resistant Plates were wrapped in parafilm and incubated overnight. The colonies. Such colonies consistently led to the establish- 1726 PAUL ET AL

Table 1. Immunofluorescence Analysis of PU-34 Stromal Cell Line

Primary Antibody Fluorescence

Pab 419 P 800 Vimentin - Fi bronectin . ._Q Cytokeratin -5 600 Desmin 0 Laminin ,E 400 Collagen IV - Factor VIII-associated Ag $ 200

0 ment of permanent cell lines and 15 cell lines were easily 1 2 3 4 established from the initial infection of primate LTMC. All weeks in culture

of the cell lines established exhibited nuclear-specific SV40 Downloaded from http://ashpublications.org/blood/article-pdf/77/8/1723/605635/1723.pdf by guest on 24 September 2021 LT antigen immunofluorescence staining using Pab 419 antibody (Table 1).The cell lines demonstrated fibroblast- like morphology with large round nuclei and prominent B nucleoli on phase microscopy. Figure 1 demonstrates the l0O0 1 morphology of a representative line, PU-34. None of these 5 8001 cell lines accumulated obvious fat vacuoles in culture, even in the presence of hydrocortisone. Three cell lines were chosen for preliminary analysis based on growth velocity (PU-22, PU-27, and PU-34). PU-34 supported hematopoietic cells for prolonged periods in vitro during this initial screening and was extensively studied. Support of hematopoiesis in long-term cultures by stromal cell line PU-34. Confluent flasks of mitomycin C-treated PU-34 cells were charged with primate bone marrow cells 1 2 depleted of plastic-adherent cells. Treatment with mitomy- 3 cin C prevents further proliferation while maintaining weeks in culture cellular function of the PU-34 stromal cells. In a typical Fig 2. Support of primate progenitor CFU by PU-34 cell line. experiment, 860 progenitors (535 myeloid, 325 erythroid) Nonadherent cells produced during long-term cultures initiated on were plated onto PU-34 (Fig 2A). Subsequently, nonadher- PU-34 were harvested weekly and assayed in methylcellulose cultures ent cells continuously shed into the media in these cultures for myeloid (N)and erythroid (0)progenitors (A) or myeloid progeni- included multilineage, myeloid, and erythroid progenitors tors only with cultures sacrificed and replated at week 3 (B). Colonies and mature granulocytes and monocytes. Assayed in meth- (>50 cells) were counted 10 to 12 days after plating. These results represent two representative and independent experiments.

Fig 1. Photomicrograph of PU-34 immortalized stromal cell line. Cells display a fibroblastic morphology with a large round nucleus IN) and prominent nucleoli. Note abscence of fat vacuoles. (Original magnification ~250.) STROMAL-ASSOCIATED HEMATOPOIESIS 1727 ylcellulose, large numbers of myeloid and erythroid progen- were trypsinized, hematopoietic cells separated by Ficoll- itor-derived colonies were present in the nonadherent cells Paque density centrifugation, and the low-density cells for up to 4 weeks after initiation, with a cumulative increase replated in methylcellulose cultures. The adherent cells in colony-forming cells in each culture. As seen in Fig 2A, from cultures represented in Fig 2B gave rise to a mean of from the initially plated 860 progenitors, 555 were recov- 21 colonies/flask, representing approximately 30% of the ered at week 1 in the nonadherent fraction. Subsequently, total CFU/flask at week 3. Individual progenitor colonies an additional 400 were recovered by week 4, with erythroid plucked from methylcellulose and stained with Wright- colonies present throughout the culture period. In addition, Giemsa stain after cytocentrifugation consisted of granulo- characteristic refractile hematopoietic cells were present in cytic and monocytic cells. Some multilineage colonies the adherent fraction of the cultures. containing erythroid and megakaryocytic cells were also Morphologically, these cultures appeared indistinguish- evident up to 2 weeks in culture (Fig 3A and B). These data able from standard human LTMC, with refractile he- suggest that PU-34 supported the survival of multilineage, matopoietic cells scattered throughout the stroma rather myeloid, and erythroid progenitor cells as well as sustained than in prominent nests (cobblestone areas) seen in murine myeloid differentiation in vitro. Similar attempts to main- LTMC. In some experiments (Fig 2B), the adherent cells tain colony forming progenitor cells in primary primate Downloaded from http://ashpublications.org/blood/article-pdf/77/8/1723/605635/1723.pdf by guest on 24 September 2021

Fig 3. (A) Progenitor-derived mixed primate colony grown in methylcellulose from nonadherent cells of long-term culture established on PU-34. The darker cells of the main colony (at arrow head) include hemoglobinized cells and erythroid precursors by Wright-Giemsa and benzidine staining of cytospin preparations. (Original magnification x40.) (6) Cytospin preparation of primate progenitor colony derived from nonadherent cells of long-term culture established on PU-34. Stained with Wright-Giemsa. G. granulocytes; M, macrophages; Mk, megakarocyte, E, erythroid precursor. (Original magnification ~1,000.) 1728 PAUL ET AL

LTMC were unsuccessful due to premature peeling of the Table 2. Constitutive Expression of Hematopoietic Growth Factors adherent stromal cells. Experiments performed with nonad- by PU-34 herent mononuclear cells from both human and rhesus Growth Factor % Activity primate bone marrow aspirates showed virtually identical Growth Activity PU-34 PU-34 CM Attributable to Factor Bioassay* CM (1:lO) +(Antibody) Growth Factor support of cells capable of colony formation for 3 weeks ~ (data not shown). Plating of nonadherent cells from these LIF DA-2 5,427t 5,816 (a-GCSF) 100% cultures yielded predominantly myeloid colonies and some IL-6 T1165 67,176t 2,935 (a-IL-6) 96% 418 (a-GM-CSF) 69% erythroid and mixed colonies. GM-CSF MO7-E 1,365t M-CSF CFU-M 15$ ND 100% Phenotypic analysis of primate stromal cell line, PU-34. The phenotypic analysis of stromal cell lines included Abbreviation: ND, not determined. demonstration of intermediate filament proteins, extracellu- *See text for full details. lar matrix components, and constitutive and induced growth tTritiated thymidine incorporation (see Materials and Methods). factor production. Intermediate filaments and the compo- SMonocyte/macrophage colonies/105 plated murine marrow cells (see Materials and Methods). nents of the ECM were evaluated through immunofluores- Downloaded from http://ashpublications.org/blood/article-pdf/77/8/1723/605635/1723.pdf by guest on 24 September 2021 cence microscopy and SDS-PAGE analysis. Immunofluores- cence studies demonstrated the presence of vimentin and blot analysis of poly A+ RNA obtained from IL-la- fibronectin and the absence of cytokeratin, desmin, lami- induced PU-34 (Fig 5). The expression of G-CSF and LIF nin, collagen type IV, and Factor VIII-related antigen were confirmed by bioassay using DA-2 cells in the pres- (Table 1). SDS-PAGE was used to analyze the types of ence and absence of antibody to human G-CSF (Fig 6A). collagens present in the ECM. Collagen types I, 111, and V The expression of GM-CSF was confirmed by bioassay were clearly demonstrated and the absence of collagen type using TF-1 cells and Mo7-E cells in the presence or absence IV (as suggested by the immunofluorescence data) was of antibody to human GM-CSF (Fig 6B and C). The confirmed by SDS-PAGE analysis (Fig 4). As seen in Fig 4, expression of IL-6 was confirmed by bioassay using T116.5 digestion of the medium procollagens with pepsin followed cells in the presence or absence of antibodies to human by separation with SDS-PAGE with delayed reduction IL-6 (Fig 6D). No expression of IL-3, IL-4, IL-5, IL-9, or made it possible to distinguish alpha 1(I), alpha 2(I), and TNFa was noted by RNA analysis. We did not evaluate alpha l(II1) collagen chains. Faint bands representing IL-7 through bioassays. alpha 1(V) and alpha 2(V) are also demonstrated. No band IL1-a-induced PU-34 CM also stimulated the growth of with mobility suggestive of collagen type IV could be CFU-GM colonies in human progenitor assays (Table 3A) detected. These findings demonstrate that PU-34 expresses and BFU-E colonies in human erythroid cultures (Table a fibroblastic phenotype consistent with a mesenchymal 3B), even in the presence of excess antibodies to GM-CSF, origin. the only demonstrated source of BPA in PU-34. Stimula- Growth factor production by PU-34. Growth factor pro- tion of BFU-E by PU-34 CM was equivalent to that seen duction was evaluated by Northern blot analysis and by with addition of optimal doses of rhGM-CSF, but nearly biologic assays. PU-34 constitutively produced M-CSF, 50% of BFU-E colonies remained after treatment with LIF, and small amounts of IL-6 and GM-CSF by bioassay equivalent amounts of neutralizing antibody to GM-CSF. (Table 2). The expression of G-CSF, GM-CSF, IL-6, IL-7, No further neutralization of BFU-E formation was demon- and LIF were demonstrated by the appearance of appropri- strated with larger amounts of neutralizing antibody to ate size cDNA-hybridizing transcripts noted on Northern GM-CSF. In contrast, addition of anti-GM-CSF antibody completely abrogated the growth of BFU-E stimulated by rhGM-CSF. Antibody-resistant BFU-E demonstrates residual BPA-like activity in PU-34 CM and may represent , *- a1 (111) another novel cytokine activity. Dilutions of PU-34 CM also stimulated the growth of murine megakaryocyte colonies (data not shown), which could be accounted for bycombina- tions of the known hematopoietic growth factors present in PU-34 CM (LIF, IL-6, IL-11).

DISCUSSION The microenvironment that supports HSC renewal both in vitro and in vivo is composed of multiple cell types with a complex extracellular LTMC provide an in vitro environment that closely mimics the HSC-HM interactions Fig 4. SDS-PAGE analysis of seen in vivo.',2,14*45We have previously shown that recombi- 'H-Prolindabeled proteins made nant retrovirus vectors provide an efficient and rapid by PU-34. Electrophoresisof pep- method of immortalizing stromal cells from multiple tissues sinized samples under nondena- of an adult mouse, including adult bone Be- turing conditions shows types 1, 111, and V collagen production by cause primary LTMC have been problematic in the primate PU-34. system, we hypothesized that immortalized cell lines may STROMAL-ASSOCIATED HEMATOPOIESIS 3729

LL v, 0 t c3

2.3 - 2.3 - 2.3 - 2.31 2.3 - 2.1 - 2.1- - 2.1- - 2.1 - 2.1 - Downloaded from http://ashpublications.org/blood/article-pdf/77/8/1723/605635/1723.pdf by guest on 24 September 2021

Fig 5. Northern blot analysis of poly-A + RNA 0.5- 0.5 - 0.5 - from IL-I-induced PU-34 cells. UP-labeled probes 0.5- used are shown at top of each lane. Arrows show 0.5- location of hybridizing transcripts and molecular size markers are shown at left.

provide a means of studying primate HSC-stromal cell ful infection of reconstituting HSC of larger species is a interactions in vitro. In addition, such cell lines would be critical goal of current experiments before the application useful for support of stem cells into which foreign genetic of gene transfer technology toward human somatic gene sequences had been inserted via retrovirus vectors. Success- therapy. A major obstacle in the extension of gene transfer A '' B 60

E;a 30 oa - PO

--._--. ---.---..._.____.~ , 0 0 - 1:lO 1:50 1:250 1:1250 1:io 1:50 1:250 1:1250 Dilution Dilution c 70

60 I \

0 1:lO 1:50 1:250 1:1250 150 1:250 1:1250 Dilution Dilution

Fig 6. Tritiated thymidine incorporation by-factor dependent cell lines in response to CM from PU-34 cells. Response of (A) DA-2 cells in the presence or absence of antibody to G-CSF; (B) TF-1 cells in the presence or absence of antibody to GM-CSF; (C) Mo7-E cells in the presence or absence of antibody to GM-CSF; and (D) TI165 cells in the presence or absence of antibodies to IL-6. The dilution of PU-34 CM is noted on the abscissa and the counts incorporated on the ordinate. (-04 PU-34. (--A--)PU-34 + antibody. 1730 PAUL ET AL

Table 3. Support of (A) Myeloid and (B) Erythroid Progenitor Growth cultures on PU-34, we postulate that reconstituting primate by PU-34 CM and human HSC will survive in vitro on PU-34 for pro- A. CSF* CFU-GM/105t longed periods. PU-34(1:16) 15,14 Multiple cell types, including fibroblasts, have been PU-34 (1:16) + anti-G-CSF 12,14 shown to comprise the adherent population in LTMC. PU-34 + anti-G-CSF + anti-GM-CSF 0.0 Although bone marrow explants have given rise to adherent fibroblasts in short-term culture (CFU-F)yl Perkins and 8. BFU-E/IO~* Fleischman5’ have recently shown that the progeny of None o,o,o murine CFU-F are endothelial-like cells. In addition, Zuck- 0.0.0 erman and Wicha have demonstrated an extensive extracel- rHGM-CSF (5 ng/mL) 6,3,10 lular network of fibronectin, laminin, and collagen IV 4,8,12 throughout the life of adherent stroma of primary murine rHGM-CSF + anti-GM 0.0.1 LTMC.53The cell type giving rise to permanent cell lines 0,I.O that support murine hematopoiesis express collagen type Downloaded from http://ashpublications.org/blood/article-pdf/77/8/1723/605635/1723.pdf by guest on 24 September 2021 PU-34 (1: 10) 8.7.6 IV and laminin, markers associated with endothelial lin- (1:50) 6.5.8 eages.2627S2 In contrast, analysis of the extracellular matrix products of PU-34 shows the presence of collagen types I, PU-34(1:10) + anti-GM 4,2,3 (1:50) + anti-GM 3.3,5 111, and V with no collagen type IV or laminin, findings consistent with a fibroblastic phenotype. Nevertheless, this *Conditioned media (dilutions) or growth factor (concentration) primate cell line is capable the support of hematopoiesis added. of tAssayed on human bone marrow cells. and replacement of the complex LTMC microenvironment. *Assayed on peripheral blood monoculear cells. Lim et a1 have reported that cells expressing collagen type IV can be grown out of human bone marrow using special- experiments from the murine system to larger species (dog, ized culture conditions and termed the progenitor cell for primate, and presumably human) is the inefficiency of these colonies CFU-RF.” It is intriguing that the lines infection of long-term reconstituting stem cells. The use of generated from primate bone marrow long term cultures drug-resistant stromal cell lines, such as PU-34, may ad- are uniformly fibroblast-like. These results raise the impor- dress this limitation by providing an optimal environment tant possibility that the cells in the microenvironment that for reconstituting stem cells during prolonged in vitro are the critical determinant of the stroma’s capacity to manipulation required for infection and election.'^,^^ support hematopoiesis may be different in murine and Although many murine bone marrow stromal cell lines human or primate LTMC. This difference could explain the have been generated and characterized, few stromal cell lack of cobblestone areas in human and primate cultures as lines are available from human bone marrow and none from well as the differences in duration of hematopoiesis noted primate bone marrow. The few human stromal cell lines when human cultures are compared with murine cultures. available have been generated using transfection of SV40 Alternatively, the support of hematopoiesis may not de- LT antigen by DNA-mediated transfection methods?8329 pend on interactions with specific stromal cell types. Al- The technique using recombinant retrovirus vectors used in though results of Roberts et aIs4would support this interpre- these studies has proven to be a more efficient means of tation, the localization of hematopoiesis in the medullary gene transfer into mammalian cells including HSCZ4and cavity in adults suggests some specificity to this microenvi- cells making up the HM.323a-49 ronment. Studies are currently underway to address these The stromal line PU-34 was generated from a randomly questions using immortalized cell lines derived from picked G418-resistant colony derived from primate LTMC CFU-RF colonies. In addition, systematic analysis of immor- infected with a defective recombinant retrovirus expressing talized stromal cell lines derived from adult mice in our the SV40 LT antigen. PU-34 demonstrated the capacity to laboratory has shown that lines derived from the bone replace the HM of LTMC in support of both primate and marrow, but not skin, lung, or kidney, are capable of human multipotent progenitor and committed cells. We maintaining CFU-S populations in vitro.46 have previously demonstrated that an immortalized murine The interaction of the microenvironment and hematopoi- bone marrow stromal cell line was capable of supporting etic cells may be dependent, at least in part, on local day 12 CFU-S for up to 5 weeks in vitro3’ and reconstituting secretion of growth factors. In LTMC, reentry into the S HSC for up to 3 weeks.46Unfortunately, an assay equivalent phase of primitive hematopoietic progenitors occurs after to the CFU-S assay does not exist in the primate system. each weekly medium change.55Recent work suggests that However, studies aimed at demonstrating the capacity of this may be due to a yet undefined component of horse PU-34 to support primate reconstituting HSC are under- serum inducing cytokine expression by stromal cells. Prece- way. Because PU-34 and the murine stromal cell line noted dent for such a model is suggested by the demonstrated above differ in morphology and cell type (see below) the capacity of regulatory molecules, such as platelet-derived support of reconstituting HSC population by PU-34 re- growth factor (PDGF), ILl-a, and transforming growth mains an important unanswered question. However, be- factor p (TGF-P) to modulate cytokine production by cause multipotential colonies are supported in long-term mesenchymal cell^.'^^"^^^^^^ Therefore, there are likely to be STROMAL-ASSOCIATED HEMATOPOIESIS 1731 soluble regulatory factors that influence the activities of of supporting primate and human multipotent, myeloid, hematopoietic cells indirectly through effects on cellular and erythroid progenitor cells for up to 3 weeks in vitro. In components of the HM. Many of these known growth addition, this stromal cell line has also provided a source for regulating factors are produced by PU-34, such as G-CSF, the molecular cloning of a novel hematopoietic growth IL-6, LIF, GM-CSF, and IL-7. However, novel growth factor. The role of this factor in the capacity of this cell line factors may also play a positive or negative regulatory role to support hematopoiesis as well as the full spectrum of its in the support of hematopoiesis. Residual BPA activity not biologic activity has not been established. The use of accounted for by known cytokines is present in IL1-a- immortalized stromal cell lines could have important impli- induced PU-34-conditioned media. In addition, a residual cations for somatic gene therapy in large animal models proliferative activity was noted at saturating concentrations because the efficiency of gene transfer into primitive HSC of anti-11-6-antibody in the T1165 bioassay. This residual appears to be markedly lower in larger animal species when activity provided a useful biologic assay for cloning this compared with the mou~e.'~?~~In addition, the use of cytokine activity produced by PU-34. The cDNA encoding retroviral vectors containing immortalizing oncogenes pro- T1165 mitogenic activity was isolated by mammalian expres- vides a novel and efficient approach to developing cell lines sion cloning methods and has been designated IL-11 and that might help to dissect the mechanisms by which stem- Downloaded from http://ashpublications.org/blood/article-pdf/77/8/1723/605635/1723.pdf by guest on 24 September 2021 has been reported in detail in a separate comm~nication.~~stromal cell interactions regulate HSC self renewal. This cytokine can stimulate murine B cell and megakaryocyte development in synergy with IL-3. Although murine ACKNOWLEDGMENT megakaryocyte stimulatory activity has been demonstrated We thank Fran Bennett, James Calvetti, Agnes Ciarletta, Tom using PU-34 CM, the combination of LIF, IL-6, and the Gessner, JoAnn Giannotti, Tonya Harris, David Larson, Ann novel cytokine described above may account for the ob- Leary, Mark Metzger, Susan Ricciardi, and Maribel Rios for served support of murine megakaryocytic colonies. Once technical help; Dr P.J. Jat for providing the $2 U195 producer cells neutralizing antibodies to LIF and the novel cytokine and to Giorgio Trinchieri for providing the IL-6 antibody; and Dr S. Clark and members of our laboratories for helpful discussions. described here are available, the potential presence of Primates used in this study were housed at the New England novel megakaryocyte stimulatory activity in PU-34 CM may Regional Primate Research Center and we would like to thank the be addressed more fully. staff for assistance and the care of these animals and Dr Prabhat We have immortalized multiple primate bone marrow Sehgal for helpful discussions. We also thank Dorothy Giarla for stromal cell lines and demonstrated that one line is capable preparation of the manuscript.

REFERENCES 1. Dexter TM, Allen TD, Lajtha LG: Conditions controlling the tory role of a variety of mesenchymal cell activities and inhibitory proliferation of haemopoietic stem cells in vitro. J Cell Physiol role of TGF-B. Blood 75:96,1990 91:335,1977 12. Billips LG, Petitte D, Landreth S: Bone marrow stromal cell 2. Allen TD, Dexter TM: Cellular interrelationships during in regulation of B lymphopoiesis: Interleukin-1 (IL-1) and IL-4 vitro granulopoiesis. Differentiation 6:191,1976 regulate stromal cell support of pre-B cell production in vitro. 3. Allen TD: Hematopoietic microenvironments in vitro. Ultra- Blood 75:611,1990 structural aspects in microenvironments in hematopoietic and 13. Whitlock CA, Witte ON Long term culture of B lympho- lymphoid differentiation. Ciba Found Symp 84:38,1981 cytes and their precursors from murine marrow. Proc Natl Acad Sci 4. Lichtman MA: The ultrastructure of the hemopoietic environ- USA 79:3608,1982 ment of the marrow: A review. Exp Hematol9:391,1981 14. Greenberger JS: Sensitivity of corticosteroid-dependent, 5. Weiss L Hemopoiesis in mammalian bone marrow, in Mi- insulin-resistant lipogenesis in marrow preadipocytes of mutation croenvironments in Hematopoietic and Lymphoid Differentiation. diabetic-obese mice. Nature 275:752,1978 Ciba Found Symp 845,1981 15. Gartner S, Kaplan HS: Long-term culture of human bone 6. Clark SC, Kamen R: The human hematopoietic colony- marrow cells. Proc Natl Acad Sci USA 77:4756,1980 stimulating factors. Science 236:1229,1987 16. Berenson RJ, Andrews RG, Bensinger WI, Kalamasz D, 7. Metcalf D: The molecular biology and functions of the Knitter G: Antigen CD 34+ marrow cells engraft lethally irradi- granulocyte-macrophage colony-stimulating factors. Blood 67257, ated baboons. J Clin Invest 81:951,1988 1986 17. Kantoff PW, Gillio AP,McLachlin JR, Bordignon C, Eglitis 8. Roberts R, Gallagher J, Spooncer E, Allen TD, Bloomfield F, MA, Kernan NA, Moen RC, Kohn DB, Yu S-F, Karson E, Dexter TM: Heparan sulphate bound growth factors: A mechanism Karlsson S, Zwiebel JA, Gilboa E, Blaese RM, Neinhuis A, for stromal cell mediated haemopoiesis. Nature 332:376, 1988 O'Reilly RJ, Anderson WF: Expression of human adenosine 9. Saksela 0, Moscatelli D, Sommer A, Rifkin DB: Endothelial deaminase in non-human primates after retroviral mediated gene cell-derived heparan sulfate binds basic fibroblast growth factor transfer. J Exp Med 166:219,1987 and protects it from proteolytic degradation. J Cell Biol 107:743, 18. Nienhuis AW, Donahue RE, Karlsson S, Clark SC, Agricola 1988 B, Antinoff N, Pierce JE, Turner P, Anderson WF, Nathan DG: 10. Delwiche F, Raines E, Powell J, Ross R, Adamson J: Recombinant human granulocyte-macrophage colony stimulating Platelet-derived growth factor enhances in vitro erythropoiesis via factor (GM-CSF) shortens the period of neutropenia after autolo- stimulation of mesenchymal cells. J Clin Invest 76:137,1985 gous bone marrow transplantation in a primate model. J Clin 11. Cashman JD, Eaves AC, Raines EW, Ross R, Eaves CJ: Invest 80573,1987 Mechanisms that regulate the cell cycle status of very primitive 19. Umemura T, al-Khatti A, Donahue RE, Papayannopoula T, hematopoietic cells in long term human marrow culture I. Stimula- Stamatoyannopoulos G: Effects of interleukin-3 and erythropoi- 1732 PAUL ET AL etin on in vitro erythropoiesis and F cell formation in primates. clonal antibodies specific for simian virus 40 tumor antigens. J Blood 74:1561,1989 Virol39:861, 1981 20. Donahue RE, Wang EA, Stone DK, Kamen R, Wong GG, 38. Goldring SR, Roelke MS, Petrison KK, Bhan AK: Human Sehgal P-K, Nathan DG, Clark SCStimulation of hematopiesis in giant cell tumors of bone; Identification and characterization of primates by continuous infusion of recombinant human GM-CSF. cell types. J Clin Invest 79:483, 1987 Nature 321:872,1986 39. Maniatis T, Fritsch EF, Sambrook J: Molecular cloning: A 21. Gillio AP, Bonilla MA, Potter GK, Gahrilove JF, O’Reilly Laboratory Manual. Cold Spring Harbor, NY,Cold Spring Harbor RJ, Souza LM, Welte K: Effects of recombinant humangranulocyte- Laboratory, 1982, p 188 colony stimulating factor on hematopoietic reconstitution after 40. Avanzi GC, Lista P, Giovinazzo B, Miniero R, Saglio G, autologous bone marrow transplantation in primates. Transplant Bennetton G, Coda R, Cattoretti G, Pegoraro L Selective growth Proc 19:153,1987 response to IL-3 of a human leukemic cell line with megakaryoblas- 22. Bodine DM, Karlsson S, Nienhuis AW Combination of tic features. Br J Haematol69:359,1988 interleukins 3 and 6 preserves stem cell function in culture and 41. Kitamura T, Tange T, Terasawa T, Chiba S: Establishment enhances retrovirus-mediated gent transfer into hematopoietic and characterization of a unique human cell line that proliferates stem cells. Proc Natl Acad Sci USA 8633897, 1989 dependently on GM-CSF, IL-3 or erythropoietin. J Cell Physiol

23. Lim B, Apperley JF, Orkin SH, Williams DA Long-term 140:323,1989 Downloaded from http://ashpublications.org/blood/article-pdf/77/8/1723/605635/1723.pdf by guest on 24 September 2021 expression of human adenosine deaminase in mice transplanted 42. Godard CM, Augery YL, Ginsbourg M, Jasmin C: Estab- with retrovirus-infected hematopoietic stem cells. Proc Natl Acad lished cell lines of mouse marrow adherent cells producing differ- Sci USA 86:8892,1989 entiation factor(s) for the granulocyte-macrophage lineage. In 24. Apperley JF, Williams DA. Gene therapy: Current status Vitro 192397,1983 and future directions. Br J Haematol75:148, 1990 43. Greenberger JS, Eckner RJ, Sakakeeny M, Marks P, Reid 25. Perkins S, Fleischman RA: Hematopoietic microenviron- D, Nabel G, Hapel A, Ihle JN, Humphries KC: Interleukin ment origin, lineage, and transplantability of the stromal cells in 3-dependent hematopoietic progenitor cell lines. Fed Proc 42: long-term marrow cultures from chimeric mice. J Clin Invest 2762,1983 81:1072, 1988 44. Nordan RP, Potter M: A macrophage-derived factor re- 26. Zipori D: Hemopoietic microenvironments, in Testa NG, quired by plasmacytomas for survival and proliferation in vitro. Gale RP (eds): Bone Marrow Damage. New York, NY,Dekker, Science 233:566,1986 1988, p 57 45. Dexter TM, Moore MAS: In vitro duplication and cure of haemopoietic defects in genetically anaemic mice. Nature 269:412, 27. Gimble JM: The function of adipocytes in the bone marrow 1977 stroma. New Biol2:304,1990 46. Rios M, Williams DA Systematic analysis of the ability of 28. Harigaya K, Handa H: Generation of functional clonal cell stromal cell lines derived from different murine adult tissues to lines from human bone marrow stroma. Proc Natl Acad Sci USA support maintenance of hematopoietic stem cells in vitro. J Cell 823477,1985 Physiol145:434,1990 29. Singer Charbord P, Keating A, Nemunaitis .I,Ravgi G, JW, 47. Stead RB, Kwok WW, Storb R, Miller AD: Canine model Wight TN, Lopez JA, Roth GJ, Dow LW, Fialkow PJ: Simian virus for gene therapy: Inefficient gene expression in dogs reconstituted 40-transformed adherent cells from human long-term marrow with autologous marrow infected with retroviral vectors. Blood cultures-Cloned cell lines produce cells with stromal and he- 71:742, 1988 matopoietic characteristics. Blood 70464,1987 48. Corey CA, DeSilva A, Williams DA: Erythropoiesis in 30. Namen AE, Lupton S, Hjerrild K, Wignall J, Machizuki DY, murine long term marrow cultures following transfer of the Schmierer A, Mosley B, March CJ,Urdal D, Gillis S: Stimulation erythropoietic cDNA into marrow stromal cells. Exp Hemqtol of B-cell progenitors by cloned murine interleukin-7. Nature 18:201, 1990 333:571,1988 49. Anklesaria P, Sakakeeny MA, Klassen V, Rothstein L, 31. Jat PS, Cepko CL, Mulligan RC, Sharp PA: Recombinant Fitzgerald TJ, Appel M, Greenberger JS, Holland CA: Expression retroviruses encoding simian virus 40 large T antigen and polyoma of a selectable gene transferred by a retroviral vector to hematopoi- virus large and middle T antigens. Mol Cell Biol6:1204,1986 etic stem cells and stromal cells in murine continuous bone marrow 32. Williams DA, Rosenblatt MF, Beier DR, Cone RD: Gener- cultures. Exp Hematol 15:195,1987 ation of murine stromal cell lines supporting hematopoietic stem 50. Friedenstein AJ,Chailakhjan RK, Lalykina KS: The develop- cell proliferation by use of recombinant retrovirus vectors encoding ment of fibroblast colonies in monolayer cultures of guinea pig simian virus 40 large T antigen. Mol Cell Biol8:3864,1988 bone marrow and spleen cells. Cell Tissue Kinet 3:393,1970 33. Cone RD, Grodzicher T, Jaramillo M: A retrovirus express- 51. Lim B, Izaguirre CA, Aye MT, Huebsch L, Drouin J, ing the 12s adenoviral E1A gene product can immortalize epithe- Richardson C, Minden MD, Messner HA: Characterization of lial cells from a broad range of rat tissue. Mol Cell Biol8:1036,1988 reticulofibroblastoid colonies (CFU-RF) derived from bone mar- 34. Cepko CL, Roberts RE, Mulligan RC: Construction and row and long-term marrow culture monolayers. .I Cell Physiol application of a highly transmissible murine retrovirus shuttle 127:45,1986 vector. Cell 37:1053,1984 52. Perkins S, Fleischman RA: Stromal cell progeny of murine 35. Cone R, Mulligan RC: High efficiency gene transfer into bone marrow fibroblast colony-forming units are clonal endothelial- mammalian cells: Generation of helper-free recombinant retrovi- like cells that express collagen IV and laminin. Blood 75:620, 1990 rus with broad mammalian host range. Proc Natl Acad Sci USA 53. Zuckerman KS, Wicha MS: Extracellular matrix production 81:6349,1984 by the adherent cells of long-term murine bone marrow cultures. 36. Jat PS, Sharp PA Large T antigens of simian virus 40 and Blood 61540,1983 polyomavirus efficiently establish primary fibroblasts. J Virol 59: 54. Roberts RA, Spooncer E, Parkinson EK, Lord BI, Allen TD, 746,1986 Dexter TM: Metabolically inactive 3T3 cells can substitute for 37. Harlow E, Crawford LV, Pim DC, Williamson NW: Mono- marrow stromal cells to promote the proliferation and develop- STROMAL-ASSOCIATED HEMATOPOIESIS 1733 ment of multipotent haemopoietic stem cells. J Cell Physiol 57. Chasman J, Eaves AC, Eaves CJ:Regulated proliferation of 132:203,1987 primitive hematopoietic progenitor cells in long term human 55. Toksoz D, Dexter TM, Lord BI, Wright EG, Lajtha LG: The marrow cultures. Blood 66:1002,1985 regulation of hematopoiesis in long-term bone marrow cultures. 11. Stimulation and inhibition of stem cell proliferation. Blood 55:931, 58. Paul SR, Bennett F, Calvetti JA, Kelleher K, Wood CR, 1980 O’Hara RM Jr, LeaIy AC, Sibley B, Clark SC, Williams DA, Yang 56. Sing OK, Keller JR, Ellingsworth LR, Ruscetti FW:Trans- Y-C Molecular cloning of a cDNA encoding interleukin-11, a forming growth factor beta selectively inhibits normal and leuke- novel stromal cell-derived lymphopoietic and hematopoietic cyto- mic human bone marrow cell growth in vitro. Blood 72:1504,1988 kine. Proc Natl Acad Sci USA 87:7512,1990 Downloaded from http://ashpublications.org/blood/article-pdf/77/8/1723/605635/1723.pdf by guest on 24 September 2021