388. Use of the cHS4 Chromatin Insulator to Reduce Oncoretrovirus Vector-Mediated Genotoxicity

Abstract

Pre-clinical studies and clinical trials have demonstrated that vectors based on retroviruses can induce malignant transformation by trans-activation of proto-oncogenes near sites of vector integration. We have been investigating a class of DNA elements known as chromatin insulators for their ability to reduce this risk. Chromatin insulators are DNA elements that act in cis to form functional boundaries between differentially expressed chromosomal loci, blocking both the positive effects of enhancers and the repressive effects of heterochromatin. Although chromatin insulators such as the prototypic cHS4 element can provide an indirect benefit by reducing the rate of vector silencing and thus the need for high vector copy numbers, their ability to directly prevent the transformation potential of virus vectors has not been demonstrated. In order to address this issue, we transduced the IL-3 dependent cell line 32D with uninsulated and insulated oncoretrovirus vectors, and compared the rate with which |[ldquo]|transformed|[rdquo]| colonies grew in the absence of IL-3. We used the MSCV-based GFP reporter vector MGPN2 either alone or flanked with the cHS4 chromatin insulator (Emery et al., PNAS 97:9150). From 5 independent experiments, a total of 33 IL-3 independent colonies from an estimated 64,000 transduction events were observed for the uninsulated vector, giving an overall conversion rate of 1 in 2122 |[plusmn]|1183 transduction events. In contrast, only 8 IL-3 independent colonies from an estimated 76,000 transduction events were observed for the same vector flanked with the cHS4 element, giving an overall conversion rate of 1 in 9624 |[plusmn]|3202 transduction events (p=0.002). Untransduced cells typically generated no such colonies with a sensitivity limit of |[sim]| 1 in 106. Replating studies revealed this IL-3 independent phenotype was epigenetically unstable, with only 2 |[plusmn]|3% of cells from expanded clones retaining the ability to reform colonies in the absence of IL-3. These clones also often failed to express the vector-derived LTR-GFP reporter cassette after a short time in culture, suggesting that the variegated phenotype may reflect a high rate of LTR silencing. Despite this highly variegated phenotype in culture, 7 out of 7 clones tested to date were fully capable of forming tumors when transplanted into congenic C3H/HeJ mice. Recipients (5 for each clone) survived an average of 31|[plusmn]|8 days, succumbing to a variety of lymphoproliferative diseases including the differentials of myeloid dysplasia, myeloid leukemia, and malignant lymphoma. Control mice (n=10) transplanted with similar numbers of untransduced 32D cells failed to develop tumors by 24 weeks post-transplant. In all cases fully IL-3 independent cell lines could be readily derived from sites of tumor formation (spleen or lymph nodes). Recent pilot studies suggest that the rate of transduction-mediated 32D transformation can also be directly assessed by tumor formation. We are currently working to clone vector integration sites with the goal of identifying the source of the transformation events at the genetic level. Taken together, these studies provide the first functional evidence that the cHS4 chromatin insulator can significantly reduce the rate of oncoretrovirus vector-mediated genotoxicity.