Myeloid Expansion after Insertional Activation of MDS1/EVI1, PRDM16 and SETBP1 in a Successful Chronic Granulomatous Gene Therapy Trial.

Successful gene therapy trials of ADA-SCID and SCID-X1 demonstrated the curative potential of oncoretroviral gene transfer. Integration of the retroviral vectors used in these studies has been thought to be a random process but severe side effects in gene therapy and in vitro studies revealed preferred insertion of these vectors mainly around transcription start sites. In SCID patients proliferation advantage of gene corrected cells was one reason for the success of the trials, whereas in the most recent chronic granulomatous disease (CGD) gene therapy trial corrected cells do not have any selective advantage therefore the two patients received mild busulfan treatment before transplantation. High efficiency transduction and conditioning have helped in the successful correction of the patients. Peripheral blood granulocytes show a stable expression (>10%) of the transgene (gp91phox) in patient 1 (15 months post treatment) as well as in patient 2 (11 months post treatment). We reasoned that, unlike T cells, which have the capability to proliferate independent of their bone marrow progenitors, granulocytes more directly reflect the influence of retrovirus insertion, and should therefore allow to closely monitor clonal fate in vivo and its potential relation to vector insertion. To study the clonality of the corrected myelopoiesis, the long term activity of individual cell clones, and the distribution of integration sites in active cells we carried out high sensitive LAM-PCR. The highly polyclonal composition of transduced cells forming myelopoiesis caused the sustained expression of gp91phox. Individual clones carrying the transgene could be detected at multiple time points. To define whether corrected cells have a proliferation advantage due to their vector integration we started large-scale sequencing and mapping of involved insertion sites. We here present >700 unique mappable integration sites of the two treated patients. The distribution of the SFFV based retroviral vector integration sites in this trial turned non random 5 months after transplantation. Corrected long-term myelopoiesis expanded 3- to 5- fold in the two patients due to activating common integration sites (CIS) in the zinc finger transcription factor homologs MDS1/EVI1, PRDM16, or in SETBP1, suggesting that these genes influence regulation of normal long-term hematopoiesis in humans. Our data indicate that the therapeutic benefit in this trial was activated through insertional side effects, therefore our findings have important implications in novel gene therapy approaches.