We have used dogs to study gene transfer into hematopoietic stem cells, because of the applicability of results in dogs to human transplantation and the availability of canine disease models that mimic human diseases. Previously we reported successful gene transfer into canine marrow repopulating cells, however, gene transfer efficiency was low, usually below 0.1% (Kiem et al, Hum Gene Ther 1996; 7: 89). In this study we have used CD34-enriched marrow cells to study different retroviral pseudotypes for their ability to transduce canine hematopoietic repopulating cells. Cells were divided into two equal fractions that were cocultivated for 72 h with irradiated packaging cells producing vector with different retroviral pseudotypes (GALV, amphotropic or 10A1). The vectors used contained small sequence differences to allow differentiation of cells genetically marked by the different vectors. Nonadherent and adherent cells from the cultures were infused into four dogs after a myeloablative dose of 920 cGy total body irradiation. Polymerase chain reaction (PCR) analysis of DNA from peripheral blood and marrow after transplant showed that the highest gene transfer rates (up to 10%) were obtained with the GALV-pseudotype vector. Gene transfer levels have remained stable now for more than 18 months. Southern blot analysis confirmed the high gene transfer rate. Interference studies on canine D17 cells revealed that 10A1 virus behaved like an amphotropic virus and was not able to use the GALV receptor. In summary, our results show improved gene transfer into canine hematopoietic repopulating cells when CD34-enriched cells are transduced by cocultivation on a GALV-pseudotype packaging cell line in combination with a GALV-pseudotype vector. Furthermore, these results demonstrate that the monoclonal antibody to canine CD34 used in this study is able to enrich for hematopoietic repopulating cells.
We have used the dog to study gene transfer into hematopoietic stem cells because a number of spontaneous canine diseases can serve as models for human diseases. These models can be used to study the clinical effectiveness of gene therapy for genetic diseases such as metabolic (alpha-L-iduronidase deficiency) or hematologic (pyruvate kinase deficiency) disorders.1,2 In contrast to recent reports in nonhuman primates,3,4 gene transfer efficiencies in the dog have remained low.5,6 In this study we have therefore tested whether techniques employed in the nonhuman primates would also result in improved gene transfer efficiencies in the dog.
Improved gene transfer into human hematopoietic progenitor cells and baboon marrow repopulating cells has been reported using GALV-pseudotype retroviral vectors compared with amphotropic vectors.4,7,8,9 The increased gene transfer rate has been correlated with increased expression of the corresponding retroviral receptor on target cells.4,10 Retroviruses that can utilize both the GALV receptor (Pit-1) and the amphotropic receptor (Pit-2), such as the 10A1 virus, may therefore be able to improve further gene transfer rates into hematopoietic stem cells.11
Increasing the vector to target ratio has also been described to improve gene transfer efficiency in hematopoietic cells.7 In humans and monkeys this has been achieved mainly by enriching for CD34+ cells. Monoclonal antibodies (MAb) to canine CD34 have become available recently, thus enabling us to explore CD34 enrichment in our canine model for stem cell gene transfer.12,13 MAb to canine CD34 recognize approximately 2% of canine bone marrow cells,13 similar to the percentage of CD34+ cells in human bone marrow.14,15
In the current study we have used CD34-enriched marrow cells to compare gene transfer rates between GALV-pseudotype vectors and amphotropic or 10A1- pseudotype vectors in their ability to transduce canine marrow repopulating cells. As compared with our previously reported gene transfer results in dogs, we found that gene transfer efficiency was considerably higher using CD34-enriched marrow cells in combination with a GALV-pseudotype vector.
Transduction and engraftment of CD34-enriched marrow cells
Four dogs were transplanted with transduced CD34-enriched marrow cells. In three dogs transduction by cocultivation with PG13/LN cells (GALV-pseudotype) was compared with that with PA317/LNX cells (amphotropic pseudotype) and in one dog PG13/LN cocultivation was compared with cocultivation with PT67/LNX cells (10A1-pseudotype). Gene transfer efficiencies and vector titers in the culture medium after cocultivation are shown in Table 1. Even though vector titers from all packaging cell lines were similar (5 × 105 CFU/ml), retroviral titers from medium harvested after cocultivation on irradiated packaging cells varied among the different cell lines. Vector titers retrieved from irradiated PA317/LNX cells were usually lower than titers from irradiated PG13/LN cells. Cocultivation with PT67/LNX, however, resulted in a slightly higher titer that PG13/LN.
All animals engrafted and the mean time to achieve an ANC >500/μl was 13 days (Table 2 and Figure 1). Two animals were killed after a follow-up of 6 months and the two animals with the higher gene transfer rates were observed long term. Two dogs, E409 and E475, were treated with rcG-CSF (10 μg/kg/day) and rcSCF (25 μg/kg/day) for 5 days before marrow cells were harvested. There were no obvious differences in engraftment. All animals were free of helper virus as determined by PCR for amphotropic and GALV envelope sequences (data not shown).
Comparison of gene transfer efficiencies between cocultivation with PG13/LN or PA317/LNX cells
Based on improved gene transfer in human hematopoietic progenitor cells and in baboons with GALV-pseudotype vectors, we hypothesized that this pseudotype might also improve gene transfer into canine marrow repopulating cells. In three animals we compared the gene transfer efficiency between a GALV-pseudotype vector (cocultivation with PG13/LN cells) and an amphotropic vector (cocultivation with PA317/LNX cells). In all three animals gene transfer rates into CFU-C (Table 1) and into hematopoietic repopulating cells (Figures 2 and 3) were higher with PG13/LN. However, vector titers measured in the culture medium after coculture of hematopoietic cells on PA317/LNX cells were lower than the corresponding PG13/LN titers, possibly explaining the overall lower gene transfer efficiency observed with PA317/LNX. Gene transfer levels in peripheral blood cells after transplantation were initially ⩾10% in all three animals (Figures 2 and 3). In two of the three animals, E342 and E409, gene transfer levels subsequently decreased to 0.1–1% and persisted at these levels until animals were killed at 6 months after transplant (data not shown). The third animal, E476, had persistent gene transfer levels in peripheral blood mononuclear cells and granulocytes between 2 and 5% for more than 18 months (Figures 4 and 5). Since vector titers in culture medium after coculture of canine CD34+ cells on PA317/LNX cells were considerably lower than vector titers after coculture on PG13/LN, we were not able to evaluate the influence of the pseudotype on the gene transfer efficiencies obtained with these two vector producing cell lines. However, the overall gene transfer levels obtained in these animals were considerably higher than previously reported results in dogs.
Comparison of gene transfer efficiencies between cocultivation with PG13/LN and PT67/LNX
In mouse cells 10A1 virus has been shown to use both the Pit-1 and Pit-2 receptors equally well.16 To determine whether 10A1-based vectors could increase gene transfer efficiency into canine hematopoietic stem cells by using both receptors, we have compared gene transfer efficiency between a GALV- (cocultivation with PG13/LN) and a 10A1-pseudotype vector (cocultivation with PT67/LNX) (animal E475). Gene transfer into both CFU-C (Table 1) before transplantation and into hematopoietic repopulating cells (Figures 4, 5 and 6) were superior with PG13/LN, even though the vector titer in culture medium from the PT67/LNX coculture was higher than from the PG13/LN coculture. These results suggested that 10A1-pseudotype vector was not able to use the GALV receptor on canine CD34-enriched cells. The gene transfer levels for LN (PG13) were in the 10% range and have persisted long-term for more than 18 months so far (Figures 4, 5 and 6). PCR analysis of DNA from Ficoll–Hypaque-separated PB mononuclear cells and granulocytes, as well as from DM5-sorted granulocytes revealed similar levels of gene transfer (Figure 5).
GALV receptor utilization by 10A1 pseudotyped vector in the canine cell line D17
The 10A1 virus can use both Pit-1, the GALV receptor and Pit-2, the amphotropic receptor, for entry into mouse cells. Thus, one might predict that 10A1-pseudotype vectors should infect canine hematopoietic cells as well as GALV vectors, which is not what we found. To study receptor utilization by 10A1, GALV and amphotropic vectors on dog cells, we performed interference analysis by using D17 canine cells infected with different replication-competent viruses and studied their susceptibility to vectors having different pseudotypes (Table 3). Both amphotropic and 10A1 viruses blocked entry by amphotropic- and 10A1-pseudotype vectors, showing that these viruses use the same receptor for cell entry in the dog. This is in contrast to mouse cells where 10A1 blocks amphotropic vector entry, but amphotropic virus does not block 10A1 vector entry, showing that 10A1 virus uses a receptor in addition to that used by amphotropic virus. Entry of GALV vectors into D17 cells was inhibited by GALV, as expected, but was not substantially affected by the presence of amphotropic virus, showing that GALV can use a different receptor than amphotropic virus. 10A1 virus was able substantially to block GALV vector entry, indicating that 10A1 can bind to Pit-1, but does not use Pit-1 for entry. Likewise, GALV had a small effect on amphotropic and 10A1 vector entry showing some binding of GALV to the common receptor used by amphotropic and 10A1 viruses. While the results of the interference analysis in these cells are somewhat complicated and show some overlap in receptor binding, it is clear that 10A1 virus does not effectively use the GALV receptor in dog cells. Thus, it is not surprising that 10A1 vectors work only as well as amphotropic vectors in dog hematopoietic cells.
Successful gene therapy has been limited by the low gene transfer efficiencies observed in human gene therapy trials. This was in contrast to the high gene transfer efficiency observed in human hematopoietic progenitor cells and in the mouse and emphasized the need for large animal studies to investigate stem cell gene transfer protocols.17,18,19 We have used the dog to study gene transfer strategies for hematopoietic stem cell gene transfer,5,6,20 given that, in contrast to nonhuman primates, there are many spontaneous canine disease models to study gene therapy, eg alpha-L-iduronidase deficiency, pyruvate kinase deficiency and various immunodeficiencies.1,2,21
In previous studies we showed that GALV-pseudotype vectors increased gene transfer into human progenitor cells and baboon marrow repopulating cells in comparison to amphotropic vectors.4,5,7 We, therefore, investigated whether GALV-pseudotype vectors would also increase gene transfer into canine marrow repopulating cells. To accomplish this, we have used a competitive repopulation assay that allowed us to study several variables within one animal since we had previously seen considerable differences in gene transfer rates among different animals. In three dogs we compared transduction of marrow CD34+ cells with a GALV-pseudotype vector to transduction with an amphotropic vector. Gene transfer was consistently higher with the GALV-pseudotype vector. However, this could have been due to the lower virus titer obtained after coculture on irradiated PA317/LNX cells compared with PG13/LN cells. The observation that PA317 cells were more sensitive to irradiation (data not shown) may also explain the lower gene transfer rates observed in our previously published animals in which cocultivation on irradiated PA317 cells was part of the transduction protocol.5 The packaging cells were irradiated to prevent growth of the cells in transplanted animals. We have not compared transductions with unirradiated PA317 and PG13 cells since, in contrast to findings made with human and baboon sera, PG13 packaging cells survived treatment with dog serum, and thus might survive in recipient animals if not irradiated.
In one animal we compared transduction with a 10A1-pseudotype vector with transduction with a GALV-pseudotype vector. Even though the vector titers in medium obtained after cocultivation on irradiated packaging cells was slightly higher with the 10A1-pseudotype vector, gene transfer levels in peripheral blood and marrow cells after transplant were higher with the GALV-pseudotype vector, suggesting that 10A1-pseudotype vector was not able to use the GALV receptor in canine hematopoietic repopulating cells. This was confirmed by interference analysis in canine D17 cells. Similar results have also been observed for human fibroblasts.16 Since the 10A1 pseudotype vector mainly uses the amphotropic receptor for entry, this result would also suggest that GALV pseudotype vectors are more efficient than amphotropic vetors in transducing canine hematopoietic stem cells.
We have used cocultivation in this study to maximize both gene transfer into, and survival of, hematopoietic repopulating cells since optimal culture conditions for the maintenance of CD34+ cells in suspension without stromal support have not yet been established in the dog. To minimize stem cell loss we infused both the adherent and the nonadherent cells as we have done in our studies in the baboon.4 In previously published studies in nonhuman primates and the NOD/SCID mouse model using cocultivation, the adherent layers were not infused and this may explain the superior results reported here.22,23
The overall gene transfer rates reported here are considerably higher than our previously reported results in dogs.5,20 Since we have made several changes in our transduction protocol in the current study (the use of CD34-enriched marrow cells and different transduction conditions including 72-h cocultivation with GALV-pseudotype vector versus cocultivation plus transduction in a long-term culture system using amphotropic vectors), it is unclear how much of the improved gene transfer efficiency in our current dogs is due to the CD34 enrichment and how much to the transduction conditions including the retroviral pseudotype. The long-term persistence of marked marrow and peripheral blood cells after transplant though suggests that the CD34 antibody used in this study is able to enrich for hematopoietic repopulating cells in the dog.
In summary, we have shown considerably increased gene transfer into canine hematopoietic repopulating cells using CD34-enriched marrow cells and a GALV-pseudotype vector. The improved gene transfer rates will be useful to study therapeutic effects of stem cell gene therapy in canine disease models.
Materials and methods
Dogs were raised and housed at the Fred Hutchinson Cancer Research Center (FHCRC) under conditions approved by the American Association for Accreditation of Laboratory Animal Care. All animals were provided with commercial chow and chlorinated tap water ad libitum. Marrow draws were performed after animals had been anesthetized with a combination of ketamine-HCl. Animals received broad-spectrum antibiotics and recombinant canine granulocyte colony-stimulating factor (G-CSF) after transplant until absolute neutrophil count (ANC) >1000/μl. As preparation for transplantation all animals received a single myeloablative dose, 920 cGy, of total body irradiation (TBI) administered from two opposing 60Cobalt sources at 7 cGy per minute.6
Retrovirus vectors and cell lines
The LN and LNX vectors carried the bacterial neomycin phosphotransferase (neo) gene conveying G418 resistance, and were identical with the exception of different length sequences between the neo gene and the 3′ vector long terminal repeat (LTR).4 This made it possible to use a single pair of primers to amplify different length sequences by PCR that could distinguish cells genetically marked by the different vectors. Packaging cell lines were generated using the PA317 (amphotopic),24 PT67 (10A1)16 and PG13 (GALV) cells25 and individual producer clones with equal titers were selected for the LN/LNX comparisons. Vector titers were 5 × 105 colony-forming units (CFU)/ml on canine D17 and human Hela cells.
Transduction of canine CD34 cells
Dog marrow was harvested by needle aspiration from humeri and femora. Buffy coat cells were labeled with biotinylated MoAB IH6 (IgG1 anti-canine CD34)13 40 μg/ml at 4°C for 20 min, washed and then incubated at 1 × 108 cells/ml with streptavidin microbeads (Milteny Biotec, Auburn, CA, USA) for 30 min at 4°C. Cells were then separated using an immunomagnetic column technique according to the manufacturer’s instructions. Two animals (E405 and E475) were treated with 25 μg/kg recombinant canine stem cell factor (SCF) and with 10 μg/kg recombinant canine granulocyte colony- stimulating factor (G-CSF) daily for 5 days before harvesting marrow (growth factors were kindly provided by Amgen, Thousand Oaks, CA, USA).
Equal numbers of CD34-enriched cells were placed in to 75-cm2 canted-neck flasks (Corning, Corning, NY, USA) that had been seeded 6 to 8 h earlier with irradiated (2500 cGy) packaging cells. Cells were grown in Dulbecco’s modified Eagle’s medium (DMEM) containing SCF (50 ng/ml), G-CSF (50 ng/ml) and protamine sulfate (8 μg/ml). After a 72-h cocultivation, adherent and nonadherent cells were collected from the cultures and infused intravenously into the irradiated animals.
Analysis of neo gene expression by CFU-C assay
CD34-enriched cells were cultured in a double layer agar culture system as previously described.4 Briefly, isolated cells were cultured in alpha minimal essential medium supplemented with 25% fetal bovine serum (FBS; Hyclone, Logan, UT, USA), 0.1% bovine serum albumin (BSA; fraction V; Sigma, St Louis, MO, USA), 0.3% (wt/vol) agar (Difco, Detroit, MI, USA), and overlaid on medium with 0.5% agar (wt/vol) containing 100 ng/ml of recombinant canine SCF, G-CSF and 4 U/ml Epo (provided by Dr Ian McNiece, Amgen). Cells were plated at 1000–5000 cells per dish. Cultures were incubated at 37°C in 5% CO2 in a humidified incubator. All cultures were performed in triplicate. Colonies grown with and without G418 (0.5 to 2 mg/ml of active drug) were enumerated at day 14 of culture using an inverted microscope. Colonies were picked after 10–14 days and DNA was prepared from colonies as described.4
PCR analysis after transplant
Genomic DNA from peripheral blood (PB) and marrow cells was prepared using DNAzol (MRC, Cincinnati, OH, USA). PB granulocytes were obtained from the buffy coat and PB lymphocytes were recovered from the interface fraction after Ficoll–Hypaque gradient separation as described.6 PB granulocytes were also obtained by flow cytometric selection of DM5-positive cells. Anti-DM5 antibody has been shown specifically to recognize canine myeloid cells.26 Amplification conditions for LN and LNX vectors have been described.4 Briefly, 100 to 500 ng (depending on the signal intensity) of genomic DNA were amplified with either LN/LNL6 2229 and LN/LNL6 3210 or with neo 350 and neo 1150 using 2.5 U of Taq polymerase (Perkin Elmer Cetus, Norwalk, CT, USA). Conditions were optimized to ensure linear amplification in the range of the intensity of the positive PCR samples: denaturation at 95°C, followed by 31 cycles of 62°C annealing (1 min), and 95°C denaturation (1 min) with a final extension at 72°C for 7 min. For the detection of LN and LNX, PCR was run in the presence of 10 mCi/ml 32P deoxycytidine triphosphate and PCR products were separated on a 6% polyacrylamide gel. To correct for differences in actual DNA amount between samples, all samples were analyzed for the β-actin gene using 100 ng of genomic DNA and the following primers: actin-1, 5′ TCC TGT GGC ATC CAC GAA ACT 3′ and actin-2, 5′ GAA GCA TTT GCG GTG GAC GAT 3′. Conditions for β-actin PCR were the same as for the neo gene, except that only 16 cycles were run.
Southern blot analysis after transplant
Restriction digests of genomic DNA were performed with either XbaI or SacI and HindIII. XbaI restriction digest revealed full-length integrated retroviral DNA only. SacI and HindIII restriction digests differentiated between LN and LNX. LNX had a HindIII site and LN did not.
Detection of helper virus
After transplantation, peripheral blood mononuclear cell DNA was screened by PCR for GALV and amphotropic envelope sequences. Positive control log dilutions of DNA from PG13/LN and PA317/LNX packaging cells into normal dog DNA were run concurrently. The sequences of the primers used to amplify the amphotropic and GALV envelope gene has been described.4 PCR conditions were as described above for the neo gene.
On day 1, infected and uninfected canine D17 cells were seeded at 5 × 104 per 3-cm diameter well of a six-well culture dish. On day 2, the medium was replaced with 2 ml of medium containing polybrene (4 μg/ml), and the pseudotyped LAPSN vectors. On day 4, cells were stained for alkaline phosphatase-positive foci as described previously.27 Results are expressed as the number of foci per milliliter.4
Taylor N et al. Correction of interleukin-2 receptor function in X-SCID lymphoblastoid cells by retrovirally mediated transfer of the gamma-c gene Blood 1996 87: 3103–3107
Shull RM et al. Bone marrow transplantation in canine mucopolysaccharidosis I: effects within the central nervous system J Clin Invest 1987 79: 435–443
Dunbar CE et al. Improved retroviral gene transfer into murine and rhesus peripheral blood or bone marrow repopulating cells primed in vivo with stem cell factor and granulocyte colony-stimulating factor Proc Natl Acad Sci USA 1996 93: 11871–11876
Kiem H-P et al. Gene transfer into marrow repopulating cells: comparison between amphotropic and gibbon ape leukemia virus pseudotyped retroviral vectors in a competitive repopulation assay in baboons Blood 1997 90: 4638–4645
Kiem H-P et al. Long-term persistence of canine hematopoietic cells genetically marked by retrovirus vectors Hum Gene Ther 1996 7: 89–96
Kiem H-P et al. Retrovirus-mediated gene transduction into canine peripheral blood repopulating cells Blood 1994 83: 1467–1473.
von Kalle C et al. Increased gene transfer into human hematopoietic progenitor cells by extended in vitro exposure to a pseudotyped retroviral vector Blood 1994 84: 2890–2897
Bauer TR Jr, Miller AD, Hickstein DD . Improved transfer of the leukocyte integrin CD18 subunit into hematopoietic cell lines using retroviral vectors having a gibbon ape leukemia virus envelope Blood 1995 86: 2379–2387
Bunnell BA et al. High efficiency retrovirus-mediated gene transfer into human and nonhuman primate peripheral blood lymphocytes Proc Natl Acad Sci USA 1995 92: 7739–7743
Orlic D et al. The level of mRNA encoding the amphotropic retrovirus receptor in mouse and human hematopoietic stem cells is low and correlates with the efficiency of retrovirus transduction Proc Natl Acad Sci USA 1996 93: 11097–11102
Miller DG, Miller AD . A family of retroviruses that utilize related phosphate transporters for cell entry J Virol 1994 68: 8270–8276
McSweeney PA et al. Canine CD34: cloning of the cDNA and evaluation of an antiserum to recombinant protein Blood 1996 88: 1992–2003
McSweeney PA et al. Characterization of monoclonal antibodies that recognize canine CD34 Blood 1998 91: 1977–1986
Civin CI et al. Antigenic analysis of hematopoiesis. III. A hematopoietic progenitor cell surface antigen defined by a monoclonal antibody raised against KG-1a cells JImmunol 1984 133: 157–165.
Andrews RG, Singer JW, Bernstein ID . Monoclonal antibody 12.8 recognizes a 115-kd molecule present on both unipotent and multipotent hematopoietic colony-forming cells and their precursors Blood 1986 67: 842–845
Miller AD, Chen F . Retrovirus packaging cells based on 10A1 murine leukemia virus for production of vectors that use multiple receptors for cell entry J Virol 1996 70: 5564–5571
Dick JE et al. Introduction of a selectable gene into primitive stem cells capable of long-term reconstitution of the hemopoietic system of W/WV mice Cell 1985 42: 71–79
Nolta JA, Smogorzewska EM, Kohn DB . Analysis of optimal conditions for retrovirus-mediated transduction of primitive human hematopoietic cells Blood 1995 86: 101–110
Moritz T et al. Fibronectin improves transduction of reconstituting hematopoietic stem cells by retroviral vectors: evidence of direct viral binding to chymotryptic carboxy-terminal fragments Blood 1996 88: 855–862
Schuening FG et al. Retrovirus-mediated gene transduction into long-term repopulating marrow cells of dogs Blood 1991 78: 2568–2576
Whitney KM, Goodman SA, Bailey EM, Lothrop CD Jr . The molecular basis of canine pyruvate kinase deficiency Exp Hematol 1994 22: 866–874
Van Beusechem VW, Valerio D . Gene transfer into hematopoietic stem cells of nonhuman primates (review) Hum Gene Ther 1996 7: 1649–1668
Larochelle A et al. Identification of primitive human hematopoietic cells capable of repopulating NOD/SCID mouse bone marrow: implications for gene therapy Nature Med 1996 2: 1329–1337.
Miller AD, Buttimore C . Redesign of retrovirus packaging cell lines to avoid recombination leading to helper virus production Mol Cell Biol 1986 6: 2895–2902
Miller AD et al. Construction and properties of retrovirus packaging cells based on gibbon ape leukemia virus J Virol 1991 65: 2220–2224
Sandmaier BM et al. Biochemical characterization of a unique canine myeloid antigen Leukemia 1991 5: 125–130
Miller DG, Edwards RH, Miller AD . Cloning of the cellular receptor for amphotropic murine retroviruses reveals homology to that for gibbon ape leukemia virus Proc Natl Acad Sci USA 1994 91: 78–82
This work was supported in part by grants HL36444, HL03701, DK42716, and DK47754 awarded by the National Institutes of Health, DHHS, Bethesda, MD. HPK is a Markey Molecular Medicine Investigator. We thank Eric Bell, Barbara Johnston, DVM, and the staff of the Fred Hutchinson Cancer Research Center Clinical Hematology laboratory for their excellent assistance. We also wish to acknowledge the assistance of Harriet Childs in preparing the manuscript.
About this article
Cite this article
Kiem, HP., McSweeney, P., Bruno, B. et al. Improved gene transfer into canine hematopoietic repopulating cells using CD34-enriched marrow cells in combination with a gibbon ape leukemia virus–pseudotype retroviral vector. Gene Ther 6, 966–972 (1999). https://doi.org/10.1038/sj.gt.3300925
- gene transfer
- amphotropic pseudotype
- canine CD34
Journal of Virology (2010)
Gene Therapy (2010)
Derivation and Characterization of Canine Embryonic Stem Cell Lines with In Vitro and In Vivo Differentiation Potential
Stem Cells (2009)
Journal of Veterinary Medical Science (2008)