Non-myeloablative conditioning with busulfan before hematopoietic stem cell transplantation leads to phenotypic correction of murine Bernard-Soulier syndrome

Abstract

Background: Bernard-Soulier syndrome (BSS) is an inherited bleeding disorder characterized by macrothrombocytopenia. Platelet transfusion is used for the management of bleeding, but repeated transfusion often results in alloimmunization. We have recently shown phenotypic correction of murine BSS (GPIbα(null) ) using lethal radiation conditioning followed by hematopoietic lentivirus-mediated gene transfer.

Objectives: For application of gene therapy to treatment of human patients, it is important to minimize treatment-related side effects. The objective of this study is to model a clinically relevant non-myeloablative hematopoietic stem cell (HSC) transplantation strategy.

Methods: Using transplantation of bone marrow (BM) HSCs from transgenic mice that express hGPIbα (hGPIbα(tg+/+) ), we sought to (i) determine the percentage of hGPIbα(tg+/+) HSCs required for therapeutic benefit, (ii) evaluate the efficacy of non-myeloablative conditioning using busulfan, and (iii) test the ability of anti-thymocyte globulin (ATG) to prevent/reduce undesirable immune responses.

Results: Transplantation of 10-20% hGPIbα(tg+/+) BM HSCs mixed with GPIbα(null) BM HSCs into irradiated GPIbα(null) mice was sufficient to correct bleeding time (n = 5). Transplantation of hGPIbα(tg+/+) BM HSCs into busulfan-conditioned GPIbα(null) mice corrected bleeding time in 21 of 27 recipients. Antibody response to hGPIbα and immune-mediated thrombocytopenia was documented in eight of 27 recipients, suggesting immunogenicity of hGPIbα in busulfan-conditioned GPIbα(null) mice. However, these antibodies disappeared without treatment within 30 weeks after transplantation. A combination of busulfan plus ATG conditioning successfully prevented antibody development and significantly increased therapeutic engraftment.

Conclusion: A conditioning regimen of busulfan in combination with ATG could potentially be used in non-myeloablative autologous gene therapy in human BSS.

Keywords: Bernard-Soulier syndrome; bleeding time; blood platelets; bone marrow transplantation; genetic therapy.

Fig. 1

Establishment of a transgenic mouse that expresses hGPIbα. hGPIbα transgenic mice were produced by lentivirus-mediated transgenesis, then cross-bred onto a GPIbα^null background. Transgene heterozygous mice (hGPIbα^tg+/−) were bred to homozygosity (hGPIbα^tg+/+) and were analyzed for (A) hGPIbα expression, (B) platelet count, and (C) platelet size. hIbα^Tg is a previously established hGPIbα transgenic model (7). Wild type C57BL/6 (WT) and GPIbα^null mice were used as controls.

Fig. 2

Analysis of mixed BMT recipients. Lethally (1100 cGy) irradiated GPIbα^null mice were transplanted with GPIbα^null and hGPIbα^tg+/+ BMMNCs mixed at various ratios. (A) Tail bleeding time assay of the recipients. (B) The number of hGPIbα^tg+/+ platelets was calculated by multiplying total platelet count by the percentage of hGPIbα-positive platelets, and was plotted with error bars representing the standard deviation. (C) Bleeding phenotype was rescued in recipients having at least 40 × 10³ µL⁻¹ hGPIbα^tg+/+ platelets in peripheral blood (40 × 10³ µL⁻¹ corresponds to 6.5% of wild type mouse platelet counts). *P < 0.05

Fig. 2

Analysis of mixed BMT recipients. Lethally (1100 cGy) irradiated GPIbα^null mice were transplanted with GPIbα^null and hGPIbα^tg+/+ BMMNCs mixed at various ratios. (A) Tail bleeding time assay of the recipients. (B) The number of hGPIbα^tg+/+ platelets was calculated by multiplying total platelet count by the percentage of hGPIbα-positive platelets, and was plotted with error bars representing the standard deviation. (C) Bleeding phenotype was rescued in recipients having at least 40 × 10³ µL⁻¹ hGPIbα^tg+/+ platelets in peripheral blood (40 × 10³ µL⁻¹ corresponds to 6.5% of wild type mouse platelet counts). *P < 0.05

Fig. 3

Busulfan-conditioned BM transplantation. GPIbα^null mice were conditioned with either lethal total body irradiation (1100 cGy) or non-myeloablative busulfan (25 mg kg⁻¹ on days −2 and −1), then transplanted with hGPIbα^tg+/+ BMMNCs. (A) Blood samples of the recipients were analyzed after bone marrow reconstitution, and the percentage of hGPIbα-positive platelets at 19 weeks after transplantation is shown. (B) Tail bleeding time assay was performed 16–20 weeks after transplantation. *P < 0.05, ns: not significant

Fig. 4

FACS analysis of humoral immune response against hGPIbα. Plasma samples from busulfan-conditioned GPIbα^null recipient mice were analyzed for the presence of antibodies against hGPIbα, and 8 recipients, designated R1 – R8, tested positive for antibodies. (A) Normal human platelets were incubated with GPIbα^null recipient mouse plasmas collected at various time points or C57BL/6 wild type mouse plasma as controls, stained with phycoerythrin (PE)-labeled anti-mouse IgG, and analyzed by flow cytometry. (B) The antibodies found in the GPIbα^null recipient mice bind to normal human platelets and hGPIbα^tg+/+ mouse platelets but not C57BL/6 wild type mouse platelets, demonstrating that the antibody is recognizing hGPIbα. “plt” means platelets. Representative data of R4 plasma collected 19 weeks after transplantation are shown. (C, D) Time course showing the percentage of hGPIbα-positive platelets (C) and the number of hGPIbα^tg+/+ platelets (D) in the recipients that presented humoral immune responses to hGPIbα (n = 8).

Fig. 4

FACS analysis of humoral immune response against hGPIbα. Plasma samples from busulfan-conditioned GPIbα^null recipient mice were analyzed for the presence of antibodies against hGPIbα, and 8 recipients, designated R1 – R8, tested positive for antibodies. (A) Normal human platelets were incubated with GPIbα^null recipient mouse plasmas collected at various time points or C57BL/6 wild type mouse plasma as controls, stained with phycoerythrin (PE)-labeled anti-mouse IgG, and analyzed by flow cytometry. (B) The antibodies found in the GPIbα^null recipient mice bind to normal human platelets and hGPIbα^tg+/+ mouse platelets but not C57BL/6 wild type mouse platelets, demonstrating that the antibody is recognizing hGPIbα. “plt” means platelets. Representative data of R4 plasma collected 19 weeks after transplantation are shown. (C, D) Time course showing the percentage of hGPIbα-positive platelets (C) and the number of hGPIbα^tg+/+ platelets (D) in the recipients that presented humoral immune responses to hGPIbα (n = 8).

Fig. 5

ATG treatment prevents immune responses against newly introduced hGPIbα. GPIbα^null mice were treated with busulfan-only (25 mg kg⁻¹ on days −2 and −1), or with ATG (10 mg kg⁻¹, day −2) before busulfan conditioning, and then hGPIbα^tg+/+ BMMNCs were transplanted. (A) Comparison of engraftment following various conditioning regimens. Flow cytometric analysis performed 10–12 weeks after transplantation showed that all 11 recipients conditioned with busulfan plus ATG presented over 80% hGPIbα-positive platelets. (B) Platelet count in GPIbα^null recipients conditioned with ATG and busulfan was significantly higher than that of the recipients conditioned with busulfan only. (C, D) Time course showing the percentage of hGPIbα-positive platelets (C) and platelet counts (D) in the GPIbα^null recipients conditioned with ATG and busulfan (n = 11). **P < 0.01, ns: not significant

Fig. 5

ATG treatment prevents immune responses against newly introduced hGPIbα. GPIbα^null mice were treated with busulfan-only (25 mg kg⁻¹ on days −2 and −1), or with ATG (10 mg kg⁻¹, day −2) before busulfan conditioning, and then hGPIbα^tg+/+ BMMNCs were transplanted. (A) Comparison of engraftment following various conditioning regimens. Flow cytometric analysis performed 10–12 weeks after transplantation showed that all 11 recipients conditioned with busulfan plus ATG presented over 80% hGPIbα-positive platelets. (B) Platelet count in GPIbα^null recipients conditioned with ATG and busulfan was significantly higher than that of the recipients conditioned with busulfan only. (C, D) Time course showing the percentage of hGPIbα-positive platelets (C) and platelet counts (D) in the GPIbα^null recipients conditioned with ATG and busulfan (n = 11). **P < 0.01, ns: not significant

Fig. 5

ATG treatment prevents immune responses against newly introduced hGPIbα. GPIbα^null mice were treated with busulfan-only (25 mg kg⁻¹ on days −2 and −1), or with ATG (10 mg kg⁻¹, day −2) before busulfan conditioning, and then hGPIbα^tg+/+ BMMNCs were transplanted. (A) Comparison of engraftment following various conditioning regimens. Flow cytometric analysis performed 10–12 weeks after transplantation showed that all 11 recipients conditioned with busulfan plus ATG presented over 80% hGPIbα-positive platelets. (B) Platelet count in GPIbα^null recipients conditioned with ATG and busulfan was significantly higher than that of the recipients conditioned with busulfan only. (C, D) Time course showing the percentage of hGPIbα-positive platelets (C) and platelet counts (D) in the GPIbα^null recipients conditioned with ATG and busulfan (n = 11). **P < 0.01, ns: not significant