Bleeding Phenotype of Murine Bernard Soulier Syndrome Is Potentially Corrected by Non-Myeloablative Hematopoietic Stem Cell Transplantation

Abstract 3340

Bernard Soulier Syndrome (BSS) is an inherited bleeding disorder caused by a defect in the platelet glycoprotein (GP)Ib/IX complex. Platelet transfusion is the primary treatment for hemorrhage but is often limited because of the potential for provoking allo-immunization and refractoriness. We have previously shown that the macrothrombocytopenia and prolonged bleeding of murine BSS (GPIbα^null) can be corrected by myeloablative total body irradiation (TBI) and transplantation of hematopoietic stem cells (HSCs) transduced with a lentivirus that expresses hGPIbα under integrin αIIb promoter control (2bIbαLV). For application of this gene therapy approach to the treatment of human patients, it is important to minimize treatment-related adverse effects. In this study, we established a transgenic mouse that expresses hGPIbα at levels similar to the average levels observed in 2bIbαLV transduced HSC recipients that can be used as a consistent source of bone marrow HSCs (hGPIbα^tg+). We undertook two approaches for evaluation of HSC transplantation using hGPIbα^tg+ mouse bone marrow (BM) cells; 1) determine the percentage of hGPIbα^tg+ HSCs necessary for therapeutic benefit and 2) determine if non-myeloablative strategies can achieve hGPIbα expression levels sufficient to correct the bleeding phenotype of GPIbα^nullmice.

In order to determine the percentage of HSCs required to achieve therapeutic effects, mixed BM chimeras were generated by reconstituting lethally irradiated GPIbα^null mice with variable mixtures (5% - 100%) of BM cells isolated from hGPIbα^tg+ and GPIbα^null mice. Tail bleeding time assays performed after BM reconstitution demonstrated that 10% hGPIbα^tg+ BM mononuclear cells mixed with 90% GPIbα^null BM mononuclear cells were sufficient to correct the tail bleeding time (n=5). Platelet analysis showed that the bleeding phenotype was rescued in recipients having higher than 40 × 10³/μl of hGPIbα^tg+ platelets, which corresponds to 6.5% of wild type mouse platelet counts (630 ± 57 × 10³/μl, n = 8). These results suggest therapeutic potential for non-myeloablative conditioning regimens for the treatment of murine BSS.

Thus we next tested the transplantation of hGPIbα^tg+ mouse BM cells into GPIbα^null recipients conditioned with two non-myeloablative doses of busulfan (25 mg/kg on days -2 and -1). Analysis of platelets by flow cytometry showed that, as expected, the percentage of hGPIbα-positive platelets in busulfan-conditioned recipients was lower than those transplanted after lethal TBI conditioning (63.2% ±36.6%, n = 15 vs. 95.2% ±1.7%, n = 8). However, tail bleeding time assays showed that bleeding times of busulfan-conditioned recipients were corrected and not significantly different from lethally irradiated recipients (2.7 ± 3.1 minutes, n = 15 vs 1.9 ± 1.5 minutes, n = 8), while no untreated GPIbα^null mice (n=8) stopped bleeding within 10 minutes. Platelet counts were also significantly increased in busulfan-conditioned recipients compared to untreated GPIbα^null mice (381 ± 132 × 10³/μl, n = 11 vs. 181 ± 35 × 10³/μl, n = 11). Thus non-myeloablative doses of busulfan conditioning and transplantation of hGPIbα^tg+mouse BM cells were shown to achieve relevant levels of engraftment in murine BSS with significantly corrected bleeding times and platelet counts.

Antibody response to hGPIbα and immune-mediated thrombocytopenia was documented in 3 of 15 recipient mice, suggesting immunogenicity of hGPIbα transgene protein in GPIbα^nullmice partially immunosuppressed with busulfan. However, immunotolerance was identified without treatment and antibody disappeared in all three mice 3 to 7 months after BM transplantation.

In conclusion, non-myeloablative doses of busulfan conditioning may be a reasonable alternative to TBI for the BMT-based treatment of BSS and could be utilized in non-myeloablative autologous gene therapy in human BSS.