Effects of HOXB4 Overexpression on Ex Vivo Expansion and Immortalization of Hematopoietic Cells from Different Species


  • Xiao-Bing Zhang,

    1. Clinical Research Division, Fred Hutchinson Cancer Research Center, Seattle, Washington, USA
    Search for more papers by this author
  • Jeffrey L. Schwartz,

    1. Department of Radiation Oncology, University of Washington School of Medicine, Seattle, Washington, USA
    Search for more papers by this author
  • R. Keith Humphries,

    1. Terry Fox Laboratory, British Columbia Cancer Agency and Department of Medicine, University of British Columbia, Vancouver, British Columbia, Canada
    Search for more papers by this author
  • Hans-Peter Kiem M.D.

    Corresponding author
    1. Clinical Research Division, Fred Hutchinson Cancer Research Center, Seattle, Washington, USA
    2. Department of Medicine, University of Washington, Seattle, Washington, USA
    • Correspondence: Hans-Peter Kiem, M.D., Fred Hutchinson Cancer Research Center, 1100 Fairview Avenue N., Mail Stop D1-100, P.O. Box 19024, Seattle, Washington 98109-1024, USA Telephone: 206-667-4425; Fax: 206-667-6124
    Search for more papers by this author


Overexpression of the human HOXB4 has been shown to induce the expansion and self-renewal of murine hematopoietic stem cells. In preparation for clinical studies, we wished to investigate the effects of HOXB4 on cells from other species, in particular preclinical large animals such as dogs and nonhuman primates. Thus, we transduced CD34+ cells from nonhuman primates, dogs, and humans with a HOXB4-expressing gammaretroviral vector and a yellow fluorescent protein-expressing control vector. Compared with the control vector, HOXB4 overexpression resulted in a much larger increase in colony-forming cells in dog cells (28-fold) compared with human peripheral blood, human cord blood, and baboon cells (two-, four-, and fivefold, respectively). Furthermore, we found that HOXB4 overexpression resulted in immortalization with sustained growth (>12 months) of primitive hematopoietic cells from mice and dogs but not from monkeys and humans. This difference correlated with increased levels of retrovirally overexpressed HOXB4 in dog and mouse cells compared with human and nonhuman primate cells. The immortalized cells did not show any evidence of insertional mutagenesis or chromosomal abnormalities. Competitive congenic transplantation experiments showed that HOXB4-expanded mouse cells engrafted well after 1 or 3 months of expansion, and no leukemia was observed in mice. Our findings suggest that the growth promoting effects of HOXB4 are critically dependent on HOXB4 expression levels and that this can result in important species-specific differences in potency.

Disclosure of potential conflicts of interest is found at the end of this article.


Hematopoietic stem cells (HSCs) have the unique property of being able to self-renew and differentiate into cells of all hematopoietic lineages. Expansion of HSCs has numerous potential clinical applications [1, 2]. Thus, many investigators have been interested in studying strategies for HSC expansion [3, [4], [5], [6], [7], [8], [9], [10], [11]–12]. The use of the homeodomain-containing protein HOXB4 to expand HSCs is one of the most promising strategies. HOXB4 is normally expressed in immature hematopoietic progenitor cells and is rapidly downregulated during differentiation [13]. Accumulating evidence from mouse models suggests that HOXB4 overexpression enhances HSC self-renewal and allows for in vivo expansion [14, 15] and ex vivo expansion of HSCs [16, 17]. Importantly, in contrast to the strong leukemogenic capacity of other members of the HOX gene clusters [18], HOXB4-expanded HSCs retained their normal differentiation and long-term repopulation potential, and no hematologic abnormalities have been detected in mice that were transplanted with HOXB4-transduced HSCs [14, 16]. Furthermore, two research groups recently reported successful ex vivo expansion of HSCs using a soluble recombinant HOXB4 protein [17, 19]. Taken together, HOXB4 is one of the most attractive candidates for HSC expansion. Based on the reported potency of HOXB4 to stimulate the ex vivo expansion of mouse HSCs, we wished to study its effects on dog and nonhuman primate hematopoietic stem/progenitor cells, since we have used these animals as clinically relevant large animal models for stem cell transplantation and gene therapy studies [20, [21]–22].

Materials and Methods

Viral Vectors

Derivation of the gammaretroviral vector plasmid MSCV-HOXB4-ires-GFP from the original murine stem cell virus (MSCV) backbone provided by Dr. Robert Hawley [23] has been previously described [14]. The control vector, MSCV-ires-YFP, was generated by inserting the coding region of the enhanced yellow fluorescent protein reporter gene in place of the enhanced green fluorescent protein reporter gene of the previously described MSCV-ires-GFP vector [14]. The vectors were used to transiently transfect 293T-based Phoenix-GALV packaging cells. The resulting virus-containing medium was used to transduce Phoenix-RD114 packaging cells as described before [24]. A high-titer helper virus-free clone was selected. Virus-containing medium was collected in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin from monolayers of Phoenix-RD114 producer cells after incubation for 12 hours at 37°C. Viral particles were concentrated by centrifugation at 7,277g at 4°C for 16–20 hours and resuspension of the pellet in 1% of the original volume. For titer determination, serial dilutions of viral supernatants were assayed on HT1080 cells. Titers were typically 1 × 107 to 2 × 107 infectious particles per milliliter for concentrated Phoenix-RD114 retroviral vector stocks. For transduction of mouse cells, GPE86 pseudotyped virus was used as previously described [14].

CD34+ Cell Enrichment and Transduction

Nonhuman primates were housed in the University of Washington National Primate Center and dogs and mice at the Animal Health Resources unit of the Fred Hutchinson Cancer Research Center. All of the experiments were previously approved by the Institutional Animal Care and Use Committee. For mouse bone marrow (BM) extraction, the mice were humanely sacrificed. For experiments with human, nonhuman primate, and dog cells, CD34+ cells were enriched by magnetic activated cell sorting. For the mouse studies, bone marrow Sca-1+ cells from C57BL/6 mice were used. Retroviral transductions of CD34+ cells were carried out on a RetroNectin-coated surface in nontissue culture treated well plates. Dilutions of concentrated Phoenix-RD114 vectors were used for transduction of human, nonhuman primate, and dog cells at a multiplicity of infection (MOI) of 1–2. Cells were stimulated for 48 hours before transduction. Cells were cultured in Iscove's modified Dulbecco's medium (IMDM) supplemented in 12.5% horse serum, 12.5% FBS, 10−6 M hydrocortisone, 10−4 M β-mercaptoethanol, 1% glutamine, 1% penicillin/streptomycin in the presence of Flt3-L, canine stem cell factor (SCF), and canine granulocyte–colony-stimulating factor (G-CSF), each at 50 ng/ml. For transduction, cells were exposed to retroviral vectors for 4 hours in the presence of growth factors. After overnight culture in medium containing growth factors, cells were re-exposed to the same MOI of concentrated vector for 4 hours. Immediately after this second exposure, cells were washed and seeded for long-term liquid culture. Mouse Sca-1+ cells were cultured in IMDM/10% FBS supplemented with murine SCF, human thrombopoietin (TPO), and Flt3-L, each at 100 ng/ml. Mouse cells were transduced similarly, except that GPE86 pseudotyped retroviral vectors were used.

Colony-Forming Cell Assay

Assays for progenitor cells were performed in a 2-layer agar culture system in minimal essential medium/20% FBS supplemented with SCF, IL-3, IL-6, G-CSF, and granulocyte-macrophage colony-stimulating factor at 100 ng/ml and erythropoietin at 4 U/ml. After 2 weeks of culture, colonies with more than 50 cells were scored.

Long-Term Culture

Cells were maintained in the same culture medium as used for transduction in tissue culture treated 24-well plates (Nunc, Rochester, NY, http://www.nuncbrand.com). Cultures were split every week at a ratio of 1:2–1:10, and cell density was adjusted to 2–5 × 105 per milliliter with fresh medium and cytokines. For the immortalization of dog and mouse cells, HOXB4-transduced dog cells were cultured in IMDM supplemented in 12.5% horse serum, 12.5% FBS, 10−6 M hydrocortisone, 10−4 M β-mercaptoethanol, 1% glutamine, 1% penicillin/streptomycin in the presence of TPO, Flt3-L, and canine SCF, each at 100 ng/ml; HOXB4-transduced mouse Sca-1+ cells were cultured in IMDM/10% FBS supplemented with murine SCF, human TPO, and Flt3-L, each at 100 ng/ml.

Flow Cytometry

Flow cytometric analysis was performed for green fluorescent protein (GFP)/yellow fluorescent protein (YFP) expression. More than 20,000 events were analyzed for each sample. Nontransduced cells were used as a control for the gating of positive cells. For the phenotypic analysis of mouse cells, fluorescent phycoerythrin (PE)-conjugated monoclonal antibodies rat anti-mouse c-kit, CD3, CD90.2, and Ly-6G were used. For the phenotypic analysis of dog cells, mouse antihuman monoclonal antibodies CD21 (clone CA2.1D6; Serotec Ltd., Oxford, U.K., http://www.serotec.com) and CD14 (clone TÜK4; Dako, Glostrup, Denmark, http://www.dako.com), which cross-react with dog cells, were used [25]. Canine-specific monoclonal antibodies DM5 and CD3 (clone 17.6B3) were kindly provided by Drs. Peter Moore and Brenda Sandmaier (University of California, Davis, and Fred Hutchinson Cancer Research Center, Seattle, respectively) [26]. For CD34 staining, cells were labeled with biotinylated monoclonal antibody 1H6 (IgG1 anti-canine CD34) and incubated with streptavidin/PE-conjugated antibody after washing. Isotype controls were also included in these experiments. All other antibodies were purchased from Becton, Dickinson and Company (Franklin Lakes, NJ, http://www.bd.com).

Competitive Repopulation Assay

C57BL/6 Pep3 (CD45.1) and C57BL/6 mice were purchased from the Jackson Laboratory (Bar Harbor, ME, http://www.jax.org) and housed and bred at mouse facilities at the Fred Hutchinson Cancer Research Center. HOXB4-transduced BM Sca-1+ cells (1 × 107) from C57BL/6 (CD45.2) that were cultured for 1, 3, or 9 months were mixed with 1 × 105 CD45.1 fresh competitor BM cells and transplanted via tail vein into the lethally irradiated CD45.1 recipients; five animals were used for each experimental condition. At various time points (3–33 weeks after transplantation), peripheral blood was collected from the retro-orbital venous plexus and analyzed by fluorescence-activated cell sorting (FACS) for the relative contribution of donor (CD45.2) engraftment. Peripheral blood cells were treated with ammonium chloride-potassium bicarbonate buffer and, after being washed with phosphate-buffered saline containing 2% FBS cells, were stained with CD45.2-FITC, CD11b (Mac-1)-PE, CD3-PE, Grl-APC-Cy7, Ter119-PE-Cy7, Thy1.2-PE, or B220-PerCP. All of these antibodies were purchased from Becton, Dickinson and Company. Flow cytometry was performed on Vantage II (Becton, Dickinson).

Western Blot

Human, nonhuman primate, and dog CD34+ cells and mouse Sca-1+ cells were transduced with HOXB4 and FACS sorted to 99% purity. During 3 weeks of culture, fractions of the GFP/YFP-expressing cells were maintained at more than 95%. Protein was extracted by RIPA buffer and 1% protease inhibitor cocktail (Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com; P8340) was freshly added. Protein concentration was determined by Bradford assay and 20 μg of total protein per lane was separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and subsequently blotted onto polyvinylidene difluoride membranes. HOXB4 was detected after incubation with supernatant from I12 (hybridoma secreting rat anti-mouse HOXB4 antibody), which cross-reacts with human HOXB4 (I12; Developmental Studies Hybridoma Bank, Iowa City, IA, http://www.uiowa.edu/∼dshbwww). Visualization of bound antibodies was performed with the goat-anti-rat horseradish peroxidase-conjugated secondary antibody with subsequent chemiluminescence reaction (enhanced chemiluminescence) (Amersham Biosciences, Piscataway, NJ, http://www.amersham.com).


Cytospin preparations from liquid cultures were made using a cytospin centrifuge (Shandon, Pittsburgh). The cytospins were fixed with methanol and stained with a Wright-Giemsa stain (Sigma). Digital pictures were taken on a Nikon microscope using an oil immersion lens at ×100 magnification.

Statistical Analysis

Data are presented as the mean ± SD. The Student's t test (2-sided) was applied for paired samples, with equal variance assumed. Differences of p < .05 were considered statistically significant.


HOXB4 Confers a Greater In Vitro Advantage to Canine CD34+ Cells than to Human or Nonhuman Primate CD34+ Cells

Studies by Antonchuk et al. [14] have shown that HOXB4 overexpression in mouse BM cells results in a dramatic increase in the proportion of HOXB4GFP+ cells and a dramatic expansion of clonogenic progenitors in vitro. Here, we investigated the effects of HOXB4 overexpression on CD34+ cells from humans, baboons, and dogs. CD34+ cells were transduced with a HOXB4-expressing gammaretroviral vector as described in Materials and Methods. After 4 weeks in culture, the percentage of HOXB4-overexpressing dog cells (detected by expression of the linked GFP reporter gene) increased from 30%–76% (n = 6), whereas the proportion of control YFP-transduced dog cells decreased from 53%–30% (n = 6) (p < .01, paired t test). HOXB4 also conferred a similar growth advantage to transduced baboon cells (GFP+ from 29%–63% and YFP+ from 33%–26%; n = 4, p < .01). The difference observed for human CD34+ cells from peripheral blood (n = 5) and cord blood (n = 3) was not significant. Transduction with the HOXB4-vector also augmented total cell expansion three- to fivefold in suspension cultures when compared with YFP-transduced controls (p < .05) (Fig. 1).

Figure Figure 1..

Ex vivo expansion of HOXB4- and YFP-transduced cells in long-term liquid culture. GFP (for HOXB4) and YFP marking and the fold expansion of total cells in long-term liquid culture. CD34+ cells from human PB (n = 5), human CB (n = 3), baboon bone marrow (n = 5), and dog PB (n = 6) were transduced with HOXB4 or YFP retroviral vectors. Cells were kept in liquid culture for 4 weeks. Cell numbers were counted every week. GFP/YFP marking was tracked by fluorescence-activated cell sorting, and positive cells were gated according to nontransduced control. Culture media were changed regularly to keep cells at optimum densities. Shown is the mean ± SD of 3–6 independent experiments. Abbreviations: CB, cord blood; GFP, green fluorescent protein; PB, peripheral blood; YFP, yellow fluorescent protein.

The effect of HOXB4 on colony-forming cell (CFC) progenitors is shown in Figure 2. In human cord blood, HOXB4 overexpression induced a fourfold increase in the yield of CFCs relative to the YFP control measured at 17 days of in vitro expansion. Greater yields of CFCs were also seen for baboon and dog HOXB4-overexpressing cells compared with the YFP control. Maximum differences compared with control cells were fivefold for baboon cells (day 24, p < .05) and 28-fold for dog cells (day 10, p < .05).

Figure Figure 2..

CFC expansion of HOXB4 and YFP-transduced CD34+ cells. CD34+ cells from human peripheral blood (n = 5), human cord blood (n = 3), baboon bone marrow (n = 4), and dog PB (n = 6) were transduced with HOXB4 or YFP vectors and cultured for 4 weeks. CFC assays were performed every week. Shown is the mean ± SD of 3–6 independent experiments. Abbreviations: CB, cord blood; CFCs, colony-forming cells; PB, peripheral blood; YFP, yellow fluorescent protein.

Differential HOXB4 Protein Expression Profiles in HOXB4-Transduced Cells

The highest level of HOXB4-induced CFC expansion was observed in dog cells, and this was associated with a 3- to 10-fold higher expression of the HOXB4GFP transgene as determined by flow cytometric analysis for GFP of transduced dog cells compared with human and monkey cells (Fig. 3). To determine whether the differences in HOXB4-mediated expansion were associated with differences in HOXB4 protein expression, we performed Western blot analyses. First, the specificity of the antibody was tested. No band was observed from protein extracted from YFP-transduced packaging cells, indicating the specificity of I12 anti-HOXB4. HOXB4 expression levels in transduced cells were ∼100-fold higher than endogenous expression levels (not shown). To compare the HOXB4 protein expression in cells from different species at different time points, FACS-sorted HOXB4 transduced human, baboon, dog, and mouse cells were subjected to Western blot analysis. As shown in Figure 4A, 3 days after transduction, the expression levels in cells from humans and baboons were lower than those from dogs and mice. Furthermore, 10 and 17 days after transduction, the expression levels in human and baboon cells dramatically decreased. However, no significant change in HOXB4 expression was detected in dog and mouse cells.

Figure Figure 3..

Representative flow cytometry diagrams of HOXB4-transduced cells from baboons, humans, and dogs. All cells were transduced with Phoenix GALV-pseudotyped vectors. Shown are cells cultured for 10 days after transduction. Similar medium fluorescence intensity (MFI) for baboon (122) and human (117) transduced cells was observed, whereas HOXB4-overexpressing dog cells showed a higher MFI (433). Cells cultured for 3 or 17 days after transduction had the same green fluorescent protein expression patterns. Abbreviations: Bab, baboon; Hum, human.

Figure Figure 4..

Differential HOXB4 protein expression levels in HOXB4-overexpressing cells from different species. (A): HOXB4 expression levels in HOXB4-transduced human, baboon, and dog CD34+ cells and mouse Sca-1+ cells. Human CB, human peripheral blood, baboon, and dog CD34+ cells and mouse bone marrow Sca-1+ cells were transduced with HOXB4 retroviral vectors and transduced cells (green fluorescent protein+) were sorted by fluorescence-activated cell sorting. Cells were kept in culture with regular medium change and cytokine addition regularly. Cells were harvested for Western blot analysis 3, 10, or 17 days after transduction. The same quantities of extracted protein were added in each lane. Membranes were stained with GelCode Blue Stain Reagent (Pierce, Rockford, IL, http://www.piercenet.com) as loading controls. Shown is one representative experiment of four experiments with similar results. (B): HOXB4 expression levels in HOXB4-transduced human, baboon, and dog CD34+ cells and mouse Sca-1+ cells that were cultured in medium supplemented with stem cell factor (SCF), thrombopoietin (TPO), and Flt3-L. Transduced cells were sorted and kept in culture. Cells were harvested for Western blot 3 or 10 days after transduction. Shown is one representative experiment of three experiments with similar results. (C): Similar half-life of HOXB4 in transduced dog and human cells. Transduced cells were sorted and cultured in medium supplemented with SCF, TPO, and Flt3-L for 2–3 days. The HOXB4 intensity decreases by half around 1 hour after addition of cycloheximide, an inhibitor of protein synthesis, indicating a half-life of ∼1 hour. Abbreviations: Bab, baboon; CB, cord blood; Hum, human; min(s), minute(s).

Because the differentiation-promoting factors G-CSF and IL-3 were used in the human and monkey cell cultures and canine G-CSF was used in the dog cell cultures, different culture conditions might in part contribute to the observed differential HOXB4 expression profiles. To address this possibility, we cultured all cells in medium supplemented with SCF, TPO, and Flt3-L; 3 days after transduction, protein was extracted for Western blot analysis. Again, we observed three- to fivefold lower expression levels of HOXB4 in human and monkey cells relative to dog and mouse cells (Fig. 4B). Furthermore, the expression levels in human cells decreased over time in culture, suggesting that the lower expression levels of HOXB4 in human and monkey cells are independent of culture conditions.

Differential protein expression levels may result from different degradation rates of the same protein in different cells. Thus, we examined the HOXB4 half-life by treating transduced cells with cycloheximide, a protein synthesis inhibitor [27, 28]. As shown in Figure 4C, HOXB4 protein in human and dog cells had a similar half-life of ∼1 hour. These data suggest that differential regulation of HOXB4 at the protein level does not explain the differential HOXB4 protein expression patterns.

HOXB4 Overexpression Immortalizes Dog and Mouse Hematopoietic Cells

To further study the proliferation potential of dog cells, we cultured HOXB4-overexpressing cells in prolonged ex vivo liquid cultures with TPO, SCF, and Flt3-L, a cytokine combination commonly used for hematopoietic stem/progenitor cell cultures. Under these culture conditions, dog cells grew robustly for more than 12 months in five independent experiments, suggesting immortalization of dog hematopoietic cells by HOXB4 overexpression. Most of the immortalized cells did not express differentiated cell surface markers such as the myeloid markers CD14 and DM5 or the lymphoid markers CD21 and CD3. The expression of CD34, a hematopoietic stem/progenitor cell marker, was also at very low levels (Fig. 5A). Morphologically, most immortalized cells were blast-like cells with a high nucleus/cytoplasm ratio (Fig. 5B). CFC assays revealed that these cells could form colonies with a clonogenicity of 1%–2% (Fig. 5C). Cytokine dependence of these cells was also tested. Canine SCF, but not TPO or