|Home | About | Journals | Submit | Contact Us | Français|
Successful genetic treatment of most primary immunodeficiencies or hematological disorders will require the transduction of pluripotent, self-renewing hematopoietic stem cells (HSC) rather than their progeny in order to achieve enduring production of genetically corrected cells and durable immune reconstitution. Current ex vivo transduction protocols require manipulation of HSC by culture in cytokines for various lengths of time depending upon the retroviral vector that may force HSC to enter pathways of proliferation, and possibly differentiation, that could limit their engraftment potential, pluripotentiality and long-term repopulating capacity.
We have compared the ability of normal CD34+ cells cultured in a standard cytokine cocktail for 18 hours or 4.5 days to reconstitute XSCID dogs following bone marrow transplantation in the absence of any pre-transplant conditioning with that of freshly isolated CD34+ cells. CD34+ cells cultured under standard γ-retroviral transduction conditions (4.5 days) showed decreased engraftment potential and ability to sustain long-term thymopoiesis. In contrast, XSCID dogs transplanted with CD34+ cells cultured for 18 hours showed a robust T cell immune reconstitution similar to dogs transplanted with freshly isolated CD34+ cells, however, the ability to sustain long-term thymopoiesis was impaired. These results emphasize the need to determine ex vivo culture conditions that maintain both the engraftment potential and “stem cell” potential of the cultured cells.
A prerequisite for any ex vivo approach to hematopoietic stem cell (HSC) gene therapy is the ability of the transduced HSC to successfully engraft in the recipient following their re-infusion. The major limitation of the ex vivo approach to gene therapy is the need for ex vivo manipulation of HSC with culturing in cytokines for various lengths of time depending upon the retroviral vector. CD34, a cell surface glycoprotein, is an antigen expressed on a subpopulation of hematopoietic cells that contain both stem cells, presumably pluripotent stem cells, and early committed progenitors that are capable of multilineage engraftment in humans, mice, nonhuman primates, and, more recently, dogs [1–8]. It is clear from a large body of clinical and experimental data that a population of cells within the CD34+ population is both pluripotent and capable of self-renewal, and selection based upon CD34 does not deplete the graft of significant numbers of HSCs. Therefore, enumeration of CD34+ cells is the current “surrogate” for determining the stem cell content of a human, nonhuman primate and canine bone marrow grafts and are the current targets for human and canine gene therapy involving diseases of the hematopoietic system.
Because γ-retroviruses require active replication of the target cells for transduction, the typical γ-retroviral ex vivo transduction protocol used in human gene therapy clinical trials consists of a one to two day pre-stimulation in cytokines followed by three days of transductions to “maximize” the number of transduced cells [9–15]. Lentiviral vectors have gained considerable attention as potential vectors for HSC since they have been shown to be capable of transducing quiescent CD34+ cells, however maximal transduction occurs as the cell enters G1 of the cell cycle [16, 17]. Therefore, most studies use either an 18 hour (overnight) or 48 hour transduction protocol in the presence of various cytokine cocktails. Since HSCs are typically quiescent, ex vivo culture in cytokines that allow for gene transfer to occur may force HSCs to enter pathways of proliferation, and possibly differentiation, that could limit their engraftment potential, pluripotentiality and long-term repopulating capacity. Currently, the most controversial and important issue regarding the clinical use of ex vivo manipulated cells is the question of whether exhaustion of stem cells might result from growth factor stimulation ex vivo which has considerable significance for both gene therapy and HSC expansion protocols .
Over the past few years, evidence has accumulated that cultured HSC have a reduced ability to engraft in murine, xenogeneic and large animal transplant models. These animal studies were performed in recipients following pre-transplant conditioning that is known to damage bone marrow stromal cells [19–21] and decrease HSC homing to or survival in the bone marrow following transplantation .
Competitive repopulation studies in mice have revealed significantly reduced engraftment with expanded cells compared with freshly isolated bone marrow (BM) cells [23, 24]. Transplantation of ex vivo cultured autologous BM CD34+ cells in myeloablated baboons resulted in delayed short-term engraftment (recovery of normal neutrophil and platelet counts) compared to freshly isolated BM CD34+ cells . Autologous transplantation of irradiated cats with cultured bone marrow cells resulted in an engraftment rate of 40% compared with 100% for non-cultured cells .
Xenotransplantation of cultured human cord blood (CB) cells in SCID or NOD/SCID mice have been reported to result in delayed and significantly reduced engraftment compared with fresh CD34+ cells [27–33]. When fresh and expanded human CB CD34+ cells were transplanted together in a competitive repopulation assay, the fresh CD34+ cells significantly out-competed the expanded cells . An important finding from these xenogeneic studies is that the reduced in vivo repopulating capacity was evident as early as 24 to 48 hours in culture [29, 30, 33].
None of the previous studies have evaluated the effect of ex vivo culture of HSC on engraftment in the absence of pre-transplant conditioning or in a large animal model of a relevant human disease, or on the reconstitution of immune function. We have previously shown that freshly isolated bone marrow CD34+ cells from normal DLA-identical donors are capable of sustained B and T cell immune reconstitution of XSCID dogs following allogeneic BMT without any pre-transplant conditioning . The purpose of this study was to determine the effect of short-term ex vivo culture, similar to that used for γ-retroviral and lentiviral ex vivo transduction, of purified bone marrow CD34+ cells on their ability to reconstitute normal T cell function and long-term thymopoiesis in nonconditioned XSCID dogs.
The XSCID dogs used in this study were derived from a breeding colony established from a single carrier female [35, 36]. All affected dogs have the same γc mutation, a four bp deletion in exon 1, and were diagnosed shortly after birth by a PCR based mutation detection assay using DNA isolated from whole blood [36, 37]. DLA-matched or DLA-haploidentical normal littermate donors for transplantation were determined by PCR assay for highly polymorphic MHC class I and class II microsatellite marker polymorphisms [38–40].
Bone marrow cells were collected from normal littermate donors following euthanasia by removing a segment of the femur, flushing the marrow into a sterile petri dish containing HBSS without calcium and magnesium (Mediatech, Fisher Scientific, Philadelphia, PA), and filtered through a fine mesh filter [8, 36]. The cells were centrifuged and washed twice in HBSS and resuspended in PBS for isolation of CD34+ cells. Isolation of bone marrow CD34+ cells was performed as previously described . Briefly, unfractionated marrow cells were resuspended at a final concentration of 1 × 108 cells/ml in a PBS solution containing 2% horse serum and incubated with anti-canine CD34 antibody 1H6 at 20 μg/108 cells . Cells were incubated with anti-mouse IgG MACS magnetic micro-beads followed by selection on varioMACS columns according to the manufacturer’s protocol (Miltenyi, Auburn, CA). Aliquots of positively selected cells were labeled with PE-labeled streptavidin (Jackson Immunoresearch, West Grove, PA) and analyzed on a FACSCalibur flow cytometer (Becton Dickonson, San Jose, CA) to determine the purity of the eluted cells. The purity of the cells was >96%.
On day 1 purified CD34+ cells were placed in Human Dexter Medium (HDM) containing 72% Isocove’s DMEM (Mediatech, Fisher Scientific, Philadelphia, PA), 12.5%FBS, 12.5% Horse serum, 1% Penicillin-Streptomycin (Invitrogen, Carlsbad, CA), 1% sodium pyruvate (Mediatech, Fisher Scientific, Philadelphia, PA), 10uM β-mercaptoethanol and 1uM hydrocortisone (Sigma Chemical, St. Louis, MO), stimulated with 100ng/ml canine granulocye-colony-stimulating factor (G-CSF), 100ng/ml canine stem cell factor (SCF), 100ng/ml human FLT-3 Ligand (FLT3-L) (all kindly supplied by Amgen, Thousand Oaks, CA), and 50ng/ml human thrombopoietin (TPO; PeproTech, Rocky Hill, NJ) and plated in 75mm × 117mm single well plates (Nalge Nunc International, Rochester, NY) at 4 × 105 cells/cm2. For cells following the γ-retroviral vector transduction protocol, plates were incubated for 2 days at 37°C, 5% CO2. On days 3 and 4 cells were washed once, fresh media and cytokines were added, and plates were incubated 4 hours at 37°C, 5% CO2. After the 4 hour incubation on day 4, cells were washed twice in saline and a cell count was taken. For cells following the lentiviral vector transduction protocol, freshly isolated CD34+ cells were incubated in HDM and the same cytokine cocktail describe above overnight at 37°C, 5% CO2. After the overnight incubation, cells were washed twice in saline and a cell count was taken. An aliquot of CD34+ cells from each set was labeled with PE-labeled streptavidin (Jackson Immunoresearch, West Grove, PA) and analyzed on a FACSCalibur flow cytometer (Becton Dickinson, San Jose, CA) to determine the purity of the CD34+ cells.
Ten XSCID dogs were transplanted with freshly isolated DLA-matched CD34+ bone marrow cells, including 5 from a previous study  and 5 unique to this current study, and 6 XSCID dogs were transplanted with freshly isolated DLA-haploidentical CD34+ bone marrow cells. Four XSCID dogs were transplanted with CD34+ bone marrow cells cultured for 4.5 days, and two XSCID dogs were transplanted with CD34+ bone marrow cells cultured for 18 hrs (Table 1). All XSCID dogs were <2 weeks of age at the time of transplantation. None of the dogs received any pre-transplant conditioning.
Quantitation of peripheral blood lymphocyte subpopulations was determined on EDTA blood samples using a standard whole blood staining and lysis protocol . The murine mAb used in this study were FITC- or PE-labeled CD3 (CA17.3G9), CD4 (CA13.1E4), CD8 (CA9.JD3), CD21 (CA2.1D6), and CD45RA (CA4.1D3) [42, 43]. Analysis gates were adjusted to 1% positive staining with directly labeled isotype controls. For each sample, 10,000 cells were analyzed using a Becton Dickinson FACSCalibur (Becton Dickinson, San Jose, CA).
The response of peripheral blood lymphocytes to in vitro mitogenic stimulation with PHA-P (5 μg/ml; Sigma, St. Louis, MO) was performed as previously described using a flow cytometric bromodeoxyuridine (BrdU) assay (BD Bioscience, San Jose, CA) for assessing cell proliferation as per manufacturer’s instructions . Secondary staining was performed using an allophycocyanin (APC)-labeled anti-mouse IgG1 antibody for flow cytometric detection.
DNA was purified from peripheral whole blood using Generation DNA purification kit (Gentra Systems, Minneapolis, MN) according to the manufacturers instructions. To detect signal joint TRECs, a real-time quantitative PCR method was was performed on the 7500 Real-time PCR system (Applied Biosystems, Branchburg, NJ). Each PCR reaction was performed in a 50ul solution containing DNA from cells of interest, 1.0X Taqman universal master mix (Applied Biosystems, Branchburg, NJ), 900nM forward and reverse primers, and 250nM Taqman probe. The sequences of the primers and probe used are the following: forward primer 5′-GACGCCACATGCCTTTCAA - 3′; reverse primer 5′-GCGCCAGCTGCAGGGTTTT - 3′ and the probe 5′-6FAM-CTCACCTCTGTGCACGGTGATA-TAMRA - 3′ producing a 131bp product. As an internal control measurement, to normalize for the input DNA, HPRT was amplified in every sample tested (forward primer 5′-GGATTTGAAATTCCAGACAAGTTTGTTG- 3′; reverse primer 5′-GTGAGAAAAAGAAGCAATTACTTACATTC- 3′ and the probe 5′-6FAM-GCCCTTGACTATAATGAATACTTCAGG-TAMRA- 3′. A standard curve was created, by cloning the Sj fragment into a pGEM-T Easy Vector (Promega) and the number of Sj copies present in a given population was calculated by including a dilution series of this standard in each PCR experiment. PCR was performed under the following conditions: 50°C for 2 minutes followed by 95°C for 10 minutes after which 50 cycles of amplification were carried out (95°C for 15s, 60°C for 1 minute). For each sample the Ct-value, defined as the minimal number of cycles necessary to exceed the threshold, was measured and applied to the standard curve.
Serum IgG was determined by radial immunodiffusion using canine IgG specific plates (Bethyl Lab. Inc., Montgomery, TX). When serum IgG concentrations reached normal levels, dogs were immunized with a modified-live canine parvovirus (CPV) and canine distemper virus vaccine (CDV) and serum IgG specific antibody titers were determined by either a passive hemagglutination (for CPV) or serum neutralization (for CDV) assay by the Cornell University Veterinary Diagnostic Laboratory.
Six XSCID dogs were transplanted with allogeneic CD34+ bone marrow cells from normal donors that were cultured for either 4.5 days or 18 hours in a cytokine cocktail used for ex vivo retroviral transduction. Table 1 illustrates the dose of cells and haplotype of the donor for each of the XSCID recipients. The purity of the isolated CD34+ cells ranged from 94 to 97%. The dose of cells chosen was based on studies using freshly isolated allogeneic CD34+ cells from both DLA-matched and DLA-haploidentical donors that result in sustained engraftment of both donor-derived B and T cells with reconstitution of normal B and T cell function. Fig. 1 illustrates that T cell reconstitution as measured by CD45RA+ (naïve) T cells is similar in XSCID dogs transplanted with freshly isolated CD34+ cells from both DLA-matched and DLA-haploidentical donors at doses of ≥5 × 106 cells/kg (range 5 to 40 × 106/kg). None of the dogs in either experiment received any pre-transplant conditioning.
Prior to transplantation, all six dogs had the typical XSCID phenotype of low to absent peripheral T cells and normal or elevated numbers of peripheral B cells. During the first 3 months following transplantation, all 6 dogs demonstrated an increase in the proportion of peripheral T cells (Fig. 2A). Some XSCID dogs, like some XSCID boys, can develop autologous phenotypically mature, but non-functional, T cells with age because of peripheral expansion through a γc-independent pathway. However, these T cells have an activated (CD45RA−) phenotype with the dogs remaining severely T cell lymphopenic and not surviving past 3 months of age. Since T cells derived from transplanted stem cells or T cell progenitors would be expected to develop through a thymic-dependent pathway and have a naïve (CD45RA+) phenotype, we examined the CD45 isoform usage in the transplanted dogs as a measurement of T cell reconstitution. Between 1 and 1.5 months post transplant the majority of peripheral T cells in all the dogs expressed a naïve (CD45RA+) phenotype (Fig. 2B). However, 2 of the 4 dogs transplanted with cells cultured for 4.5 days showed a rapid conversion to a CD45RA− phenotype typical of untreated XSCID dogs. These same two dogs failed to develop normal numbers (>1000/μl) of CD45RA+ T cells (Fig. 2C). The other two dogs in this group, R1789 and X299, maintained normal proportions and numbers of CD45RA+ peripheral T cells through 8 months post transplant although the absolute numbers of CD45RA+ T cells remained less than normally observed in XSCID dogs transplanted with freshly isolated CD34+ cells. At 8 months post transplant the proportion and number of CD45RA+ peripheral T cells started to decline. The kinetics of T cell reconstitution differed in these two dogs with one dog, X299, showing a rapid increase in the number of CD45RA+ peripheral T cells, whereas the other dog, R1789, showing a more gradual increase in the number of CD45RA+ peripheral T cells. Both of the dogs transplanted with cells cultured for 18 hours developed and sustained numbers of CD45RA+ peripheral T cells similar to dogs transplanted with freshly isolated cells. The CD4/CD8 ratio in the four dogs that successfully engrafted followed a similar kinetics as seen in XSCID dogs transplanted with freshly isolated CD34+ cells (data not shown).
Proliferation in response to mitogenic stimulation was used to assess T cell function in the dogs that successfully engrafted. Fig. 3 illustrates that the two dogs transplanted with cells cultured for 4.5 days, R1789 and X299, and both of the dogs transplanted with cells cultured for 18 hrs demonstrated normal proliferation.
Thymic function was assessed in all 4 of the dogs that successfully engrafted using a TREC assay to evaluate thymic output. All the dogs showed high levels of TRECs during the first four months following transplantation (Fig. 4A). The number of TRECs in the two dogs transplanted with cells cultured for 4.5 days, dogs 1789 and X299, started declining between 4 to 6 months post transplant. The dogs transplanted with cells for 18 hours, dogs 1844 and 1847, also showed a decline in TREC levels however at 8 months post transplant the TREC levels were higher than those observed in the dogs transplanted with the cells cultured for 4 ½ days at 6 months post transplant. These results contrast with those in two dogs recently transplanted with with freshly isolated cells, dogs R1751 and R1752, in whom TREC levels have remained stable for at least 12 months post transplant. Fig. 4B illustrates the histology of the thymus at 12 months post transplant showing that dogs transplanted freshly isolated cells possess a normal thymic histology characterized by numerous thymocytes and a well defined cortex and medulla. The dogs transplanted with cells cultured for 4.5 days had a thymic morphology very similar to that of an untreated XSCID while the thymic morphology of dogs transplanted with cells cultured for 18 hours, although not normal, still had an appearance of a normal thymus with a well defined cortex and medulla.
Serum IgG concentrations in dogs that successfully engrafted reached low normal values between 6 to 8 months post transplant (data not shown). These dogs were immunized with a modified-live vaccine to canine parvovirus and canine distemper virus and, in contrast to untreated XSCID dogs, demonstrated a significant increase in IgG specific antibody titers to each virus. All the dogs had pre-immunization specific IgG antibody titer to canine distemper virus (CDV) of <1:10 and to canine parvovirus (CPV) of <1:4. Following two doses of vaccine, the IgG specific antibody titers to CDV ranged from 1:160 to 1:1280 CDV and from 1:128 and 1:640 for CPV. No adverse reaction to the vaccine was observed.
The disappointing results in attempting to duplicate the γ-retroviral gene transfer experiments in mice to large animals and humans in the 1990s were often attributed to “low” transduction efficiencies. Therefore, considerable effort was devoted to increasing transduction efficiencies in CD34+ cells from large animals and humans that resulted in the first successful human clinical gene therapy trial in 1999 . Current protocols for ex vivo transduction of HSC with γ-retroviral vectors require ex vivo manipulation and culture of the cells in various cytokine cocktails for up to 5 days. Numerous murine transplant studies, including xenotransplant studies using human CD34+ cells in NOD/SCID mice, and nonhuman primate transplant studies have shown that ex vivo culture of HSC for as little as 24 to 48 hours may have a detrimental effect on their short-term and long-term engraftment potential, however these studies were performed in myeloablated recipients [reviewed in 45]. Although the NOD/SCID xenotransplant model has become the “gold standard” for evaluating human hematopoiesis in vivo, this model is limited because of it’s short life span precluding long-term studies  and does not support thymopoiesis so T cell development cannot be studied .
In this study we addressed the effect of ex vivo culturing of normal canine CD34+ bone marrow cells under conditions similar to those used for ex vivo gene transfer on their ability to reconstitute normal T cell function and sustain long-term thympoiesis in nonconditioned XSCID dogs following allogeneic BMT. Although the focus of this study was to address ex vivo culture conditions used for ex vivo gene therapy, we chose to perform these studies in an allogeneic BMT setting since it eliminates one major variable associated with gene therapy studies, namely the transduction efficiency of the γ-retroviral or lentiviral vector. This is the first study to address this question in a large animal model of a relevant human disease and in the absence of any pre-transplant conditioning, and is also the first to demonstrate the effect of ex vivo culture on thymopoiesis.
The dose of cells used in this study was the upper limit used in our previous studies using freshly isolated CD34+ bone marrow cells [8, 48]. Only two of the four XSCID dogs transplanted with CD34+ cells cultured for 4.5 days successfully engrafted with each dog showing a different kinetics of T cell reconstitution. Although both of these dogs developed numbers of CD45RA+ (naïve) T cells within the range for age-matched normal dogs, they were substantially lower than that observed in XSCID dogs transplanted with freshly isolated CD34+ bone marrow cells. Both dogs showed a gradual decline in the proportion and absolute number of peripheral CD45RA+ T cells to the lower limits of age-matched normal dogs (>750/μl) by 12 months following transplantation. These results differ substantially from those observed in XSCID dogs successfully immune reconstituted with freshly isolated CD34+ bone marrow cells from both DLA-matched and haploidentical donors that maintain normal numbers of peripheral T cells ranging from 1950/μl to 2190/μl through 5 to 7.5 years post transplant [8, 48]. In contrast, both XSCID dogs transplanted with CD34+ cells cultured for 18 hours successfully engrafted with a robust T cell reconstitution with sustained numbers of peripheral CD45RA+ T cells similar to that observed using freshly isolated CD34+ cells.
The dogs that successfully engrafted donor cells exhibited T and B cell function as demonstrated by a normal T cell proliferative response following mitogenic stimulation and the ability to produce antigen-specific IgG antibody following immunization.
Evaluation of thymopoiesis by TREC analysis and the histologic appearance of the thymus, however, demonstrated differences between both groups of XSCID dogs treated with cultured CD34+ cells compared with XSCID dogs transplanted with freshly isolated CD34+ cells. XSCID dogs transplanted with freshly isolated CD34+ cells exhibit a thymus that has a normal histiologic appearance and increasing numbers of TRECS at 12 months post transplant suggesting active thymopoieis is still occurring. XSCID dogs transplanted with CD34+ cells cultured for 4.5 days showed a rapid decrease in TRECs and a thymus that had a histologic appearance of a dysplastic XSCID thymus 12 months post transplant. In contrast, the XSCID dogs transplanted with CD34+ cells cultured for 18 hours showed a more gradual decrease in TRECs with a thymus that, although not normal, had a well-defined cortex and medulla. These results suggest that the ability to sustain long-term thymopoiesis is impaired under culture conditions used for ex vivo gene therapy, and that ex vivo culturing of the HSC under conditions used for γ-retroviral also affects their engraftment potential.
Our results are consistent with the results of 21 XSCID boys treated by ex vivo γ-retroviral gene therapy [10, 13, 14, 49–52]. Two patients failed to engraft gene-corrected cells and an additional 4 patients had suboptimal levels of T cell immune reconstitution. Long-term follow-up of the 15 patients who reconstituted normal or low normal gene-corrected T cells have shown a gradual decline in T cell numbers starting approximately 2 years after treatment with a concomitant decrease in CD45RA+ (naïve) T cells [50–52] suggesting that the ability to sustain long-term thymopoiesis is impaired under current conditions used for ex vivo γ-retroviral gene therapy. These results can be best explained by the engraftment of transduced committed T cell progenitors, and probably few, if any, transduced HSC capable of self-renewal. If this were the case, the continual production of new T cells would be expected to decrease over time as appears to be the case.
Central to optimizing ex vivo gene therapy, is determining culture conditions that maintain both the engraftment potential and “stem cell” potential of the transduced (cultured) cells. The true test, however, of whether these culture conditions preserve stem cell activity and engraftment potential is evaluating their performance in vivo, preferably in large animal models under conditions of no conditioning . For over 50 years, the dog has historically been the pre-clinical animal model for BMT with many of the advances made in this model being directly transferable to human clinical BMT protocols [53–57]. A major advantage of using the XSCID dog model to test ex vivo culture conditions associated with gene therapy is that BMT can be performed in the absence of any pre-transplant conditioning thereby eliminating any potential effects of the conditioning regimen itself on either the bone marrow or thymic microenvironments.
This study was supported by NIH grants RO1 AI43745 and RO1 RR02512. The authors would like to thank Patty O’Donnell for excellent supervision of the XSCID dog colony. We would like to also thank Amgen for supplying cytokines used in this study.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.