Search

We have recently reported that the RD114-pseudotyped MFGS-gp91phox vector achieves unprecedented levels of correction of the NADPH-oxidase gp91phox (approved gene symbol CYBB) defect in CD34(+) cells from patients with X-linked chronic granulomatous disease in the NOD/SCID mouse model. Considering clinical use of this vector, we transplanted autologous mobilized peripheral blood CD34(+) progenitor cells, transduced with the RD114-MFGS-gp91phox vector, into two healthy rhesus macaques following nonmyeloablative conditioning. The moderately high levels of in vivo marking seen in the first months following transduction decreased and stabilized at about 8 months posttransplant. Marking for both healthy animals after 15 months was 0.3 to 1.3 vector copies per 100 cells in lymphocytes, neutrophils, and monocytes. Vector insertion analyses performed by linear amplification-mediated PCR and sequencing identified 32 and 45 separate insertion sites in the animals. Identical insertion sites were found in myeloid cells and lymphocytes, demonstrating the successful transduction of lymphomyeloid progenitors. Some inserts landed in the vicinity of genes controlling cell cycle and proliferation. Statistical analyses of insertion sites 1 year posttransplant suggest a high diversity of insertion sites despite low marking.

Recent studies have suggested a remarkable potential of adult stem cells from a variety of organs to give rise to cells of disparate organs, but evidence of such potential at a clonal level is lacking in most if not all studies to date. To assess directly the hematopoietic potential of muscle-derived cells in a relevant large animal, we initiated retroviral-tagging studies in the rhesus macaque to allow tracking at the clonal level by integration site analysis. Four rhesus macaques underwent transplantation with transduced muscle-derived cells after lethal irradiation followed by delayed infusion of an autologous hematopoietic graft. The first animal showed no evidence of hematopoietic recovery and, despite infusion of the backup hematopoietic graft, succumbed due to complications of prolonged cytopenias. In the remaining three animals, the overall contribution of retrovirally tagged muscle-derived cells toward hematopoiesis was exceedingly low. Retroviral integration site analysis among clonally derived muscle cells and bone marrow cells in vivo in one animal suggests a common source. These results demonstrate that harvesting disparate organs for cellular therapy is currently highly inefficient at best.

Mutagenesis activation of the LMO-2 gene was observed to be associated with the development of lymphoid leukemia several years after two infants with X-SCID had their immune systems restored by gene therapy with a murine onco-retrovirus vector encoding the human IL2 receptor common gamma chain. Just recently, a third infant treated with the same vector was reported to also have developed T cell leukemia, however, insertion site analysis is pending in this case. These observations highlight the importance of the conductance of large animal marking studies which follow the insertional pattern of vectors planned for human treatments. We have recently reported that a high titer RD114 pseudotyped MFGS-gp91phox vector achieves unprecedented levels of correction of the oxidase defect in CD34+ cells from patients with X-linked chronic granulomatous disease in the xenotransplant NOD/SCID mouse model. In anticipation of considering clinical use of this vector, we transduced G-CSF-FLT3L fusion protein mobilized peripheral blood CD34+ cells (CD34+ PBPC) of two healthy rhesus macaques with the RD114-MFGS-gp91phox. Transduced CD34+PBPC were transplanted on culture day 4 into two non-myeloablative (2|[times]|300 Rads) irradiated animals. The in vivo marking decreased over the first months and stabilized at about 8 months post transplant. Marking for animal J976 and D406 at 15 and 19 months, respectively is 1.3% and 0.3% in B lymphocytes, 0.3% and 0.3% in neutrophils and 0.7% and 0.4% in T lymphocytes at which time the animals remained healthy. Vector insertion analysis was performed using LAM-PCR and gel electrophoresis. Analyses of cloned and sequenced integration sites revealed 44 integrations sites in monkey J976, and 33 integration sites in monkey D406 with no predominant clone seen to emerge over time. 27% (J976) and 12% (D406) of the insertion sites were identified as coding regions (intron or exon). Nested PCR with insertion site specific primers enabled recovery of identified sites from different cell lineages at different time points. In summary, we detected a high number of integration sites in two rhesus macaques transplanted with gene-modified CD34+ cells despite low marking due to non-myeloablative conditioning.

With the goal of reducing toxicity associated with conditioning for transplantation of genetically modified autologous hematopoietic stem cells (HSCs), others and we have explored non-myeloablative approaches. In this work, we sought to identify an alternative to irradiation that would allow dose dependent engraftment along a less steep dose response curve. Busulfan is a widely employed conditioning agent, yet the oral formulation is erratically absorbed, necessitating cumbersome pharmacokinetic monitoring. The recent availability of an intravenous formulation prompted our assessment in the murine congenic transplantation model. To assess hematopoietic toxicity, busulfan liquid at 1, 5, 10, 15, and 20mg/kg was injected into 10-12 week-old C57BL6 mice intraperitoneally (IP) in cohorts of 5 mice. Complete blood counts were obtained twice a week for 1 month. Busulfan produced dose dependent toxicity that was most pronounced in the third week after infusion. All mice recovered to baseline without hematopoietic support. To determine the level of engraftment attainable with IP busulfan, twenty million nucleated bone marrow cells from 6-8 week-old B6.SJL-Ptprca-Pep3b-/BoyJ donors (CD45.1) were injected into 10-12 week-old recipient C57BL6 mice (CD45.2), one day after 10 (n=8) or 20 (n=5) mg/kg busulfan. Donor chimerism was determined among lymphocytes and granulocytes at 4, 8, and 13 weeks post-transplant. At 13 weeks post transplant, mean donor lymphoid and myeloid engraftment were 28.6% and 25.8% respectively for the 10mg/kg dose, and 68.7% and 76.1% for the 20mg/kg dose. As busulfan is primarily toxic against metabolically quiescent cells in the bone marrow producing myelosuppressive effects that are delayed, we reasoned that delaying the infusion of HSCs until the nadir of peripheral blood counts might improve engraftment, and performed a delayed infusion time course study to test this hypothesis. Twenty million nucleated bone marrow cells from 6-8 week-old donor mice, C57BL6 (CD45.2), were injected at 1, 10, 13, 15, 17, 20, and 22 days after a single IP injection of 10mg/kg busulfan into 10-12 week-old recipient mice, B6.SJL-Ptprca-Pep3b-/BoyJ (CD45.1). There were 5 mice per cohort. Donor chimerism was determined among lymphocytes and granulocytes at 4, 8, and 13 weeks post-transplant. Groups of mice were analyzed using analysis of variance (ANOVA), and confirmed with post-hoc Games/Howell analysis. The delayed infusion time course spanning days 1 to 22 following busulfan demonstrated that both lymphoid and myeloid engraftment were significantly improved from a mean of 28.6% (day -1) to 55.2% (day -15) and 25.8% (day -1) to 58.8% (day -15), respectively (Figure 1), p

Wild-type murine onco-retroviruses cause cancer in part by insertional mutagenesis, but until recently, cancers had not been observed in humans or animals subjected to marking or treatment with the highly modified murine retrovirus vectors in current use. Recently, mutagenesis activation of the LMO-2 gene was observed to be associated with the development of lymphoid leukemia several years after two infants with X-SCID had their immune systems restored by gene therapy with a therapeutic onco-retrovirus vector. Insertional mutagenesis activation of other oncogenes associated with development of cancers has been reported in vector-marked mice. These observations highlight the importance of encouraging many investigators to conduct large animal marking studies which follow the insertional pattern of vectors planned for human treatments. We have recently reported that a very high titer RD114 pseudotyped MFGS-gp91phox vector achieves unprecedented levels of correction of the oxidase defect in CD34+ cells from patients with X-linked chronic granulomatous disease in the NOD/SCID mouse model. In anticipation of considering clinical use of this vector, we transduced mobilized peripheral blood CD34+ cells (PBSC CD34+) of two healthy rhesus macaques with the RD114-MFGS-gp91phox. Transduced PBSC CD34+ with an average two vector copies per cell were transplanted on culture day 4 into two non-myeloablative (2× 300 Rads) irradiated animals. The moderately high levels of in vivo marking (>6%) seen in the first months following gene marking decreased and stabilized at about 8 months post transplant. Marking for animal J976 and D406 at 15 and 19 months, respectively is 1.3% and 0.3% in B lymphocytes, 0.3% and 0.3% in neutrophils and 0.7% and 0.4% in T lymphocytes at which time the animals remained healthy. Vector insertion analysis was performed by LAM-PCR, using gel electrophoresis and GeneScan analysis to demonstrate that despite the relative low marking at 10 months and later animal D406 had more than 25 and animal J976 had more than 40 separate insertional events. 34% of integration sites are located in coding regions. No predominant clone was seen to emerge over time and no patterns of insertion site preferences were detected up to this point.