Hematopoietic Gene Therapy Using
Retroviral Vectors: Location, Location,
Location

By Derek L. Persons, M.D., Ph.D.

The use of retroviral gene transfer to correct lympho-hematopoietic disorders results in permanent alteration of target cell chromosomal DNA by the vector integration event. Although this facilitates the stable, ongoing production of genetically modified, differentiated hematopoietic cell progeny from primitive cells, the location of the vector insertion in the host DNA may, in some instances, be problematic. In fact, replicating retroviruses in mice are well-recognized for their potential to cause oncogenic transformation due to alteration of endogenous gene expression and function by proviral insertion. Until recently, it was unknown if such a scenario was possible with a non-replicating retroviral vector. Several simultaneous session presentations yesterday and today report on the location of vector insertions in both human clinical trials and pre-clinical studies and their potential to cause alteration of “neighborhood” genes and cellular function.

In the Gene Therapy: Clinical Results Simultaneous Session on Monday, Dr. Cavazzana-Calvo and co-workers (Abstract 533) reported functional immunological reconstitution in nine of 10 children with SCID-X1 (due to the absence of expression of the common gamma chain) that were treated with retroviral gene correction of bone marrow CD34+ cells. Although this study constituted the first instance of correction of a human lympho-hematopoietic disorder by gene therapy, two patients subsequently developed clonal T -cell proliferations. Both patients required treatment of the lymphoproliferative syndrome with chemotherapy, and in one patient additional treatment with allogeneic stem cell transplantation was necessary. In both patients, a single retroviral insertion near or in the LMO2 transcription factor gene was observed and caused increased expression of LMO2, possibly due to gene activation by the enhancer contained within the retroviral LTR. Dysregulated LMO2 expression was felt to play a significant role in causing the monoclonal T-cell proliferation syndrome. In another session, Dr. Hu and colleagues reported (Gene Transfer: Biology and Marking; Abstract 700) that all the other patients in this trial demonstrated a stable, polyclonal profile of vector insertions. Monitoring of the diversity of insertion sites in these patients is ongoing. In contrast to the results in the SCID-X1, Dr. Auiti and co-workers reported, in an update on the results of the ADA-SCID gene therapy trial (Abstract 531), that a polyclonal pattern of vector integrations in genetically corrected T lymphocytes was observed in all four treated patients. All four patients experienced immunological reconstitution and to date no serious adverse events have been observed. Given these reports, it remains unclear as to the precise factors that led to the adverse events in the SCID-X1 trial and whether these mechanisms will be operational in ADA-SCID and other candidate disorders for gene therapy.

In today’s Gene Transfer: Biology and Marking Studies Simultaneous Session (8:00 a.m. – 10:00 a.m.), Dr. Dave and colleagues will present data (Abstract 878) suggesting a cooperative effect of dysregulated expression of both LMO-2 and the common gamma chain in causing T-cell leukemia in the AKXD mouse retroviral insertional mutagenesis model. Evidence will be presented showing that the LMO-2 and the common gamma chain loci are frequent targets of retroviral insertion in the T-cell leukemias that develop in this model. These data raise the question of whether dysregulated common gamma chain expression was contributory to the T-cell lymphopro-liferation syndrome observed in the two patients in the SCID-X1 trial. Given the potential for the activation of endogenous genes by the viral LTR contained in most retroviral vectors, Hanawa and co-workers will report studies in this same session (Abstract 873) that evaluate whether lentiviral vectors utilizing a strong cellular promoter in lieu of a viral LTR might diminish the likelihood of transcriptional activation of loci near the vector insertion site. Data will be presented that suggest that as many as one in 3000 integration events may lead to nearby transcriptional activation and that this frequency can be reduced up to 50-fold depending on choice of promoter contained within the vector.

In what is likely the most extensive analysis of vector insertion sites in the context of stem cell-targeted gene transfer and transplantation, Dr. Calmels and colleagues reported in the Gene Transfer: Biology and Marking Studies Simultaneous Session yesterday (Abstract 697) on the mapping of 200 retroviral insertion sites in the rhesus autologous, retrovirally-transduced stem cell transplantation model. In the analysis of insertion sites from five animals, 40% of insertions were intragenic. Interestingly, several vector insertions sites were independently common in more than one animal. An insertion site in the myelodysplasia syndrome gene 1, which is involved in recurrent translocations in myeloid leukemia, occurred independently in three different animals. Two other insertion sites also involved leukemia translocation sites. Despite the presence of these insertion sites, no animals have demonstrated alteration of hematopoiesis and overall vector insertion site profiles are polyclonal with follow-up of 30-64 months.