80. Insulation of Oncoretrovirus Vectors with the cHS4 Chromatin Insulator Is Associated with an Open Chromatin Histone Code

Abstract

We and others have reported that the cHS4 chromatin insulator can improve the expression of integrating gene transfer vectors by reducing the impact of silencing chromosomal position effects. Independent studies indicate that the cHS4 element prevents the encroachment of silencing heterochromatin by recruiting high levels of histone acetyl transferases, which in turn acetylate key histone moieties and enforce an open chromatin structure. This is true even in lineages where the neighboring beta-globin locus is transcriptionaly inactive, suggesting that the chromatin opening activity of the cHS4 element is independent of transcriptional activity in adjoining regions. However, previous studies in mouse bone marrow transduction and transplantation models demonstrated that a certain fraction of vector provirus, although flanked with the cHS4 element, nevertheless fail to express the vector transgene. It is unclear whether this lack of expression is associated with a failure of the cHS4 element to maintain an open chromatin state, or a failure of the vector transgene to take advantage of this open chromatin state. In order to determine whether the cHS4 element increases the frequency of vector expression by increasing the level of histone acetylation, and whether vector silencing is associated with a failure of this activity, we analyzed vector provirus in mice transplanted with uninsulated and insulated reporter vectors by chromatin immunoprecipitation (ChIP). Consistent with our previous studies, we found that 4 mice transplanted with the uninsulated vector expressed the LTR-GFP cassette only 9|[plusmn]|11% of the time, compared to 30|[plusmn]|17% for 5 mice transplanted with the same vector flanked with the cHS4 element. Cross-linked chromatin from steady-state spleen cells was precipitated with antibodies to acetylated histone 3 (H3K9/K14), and the ratio of vector provirus sequences in the precipitated and unprecipitated fractions was determined by quantitative PCR and compared to that of an internal control (GAPDH). We found that chromatin associated with uninsulated provirus was acetylated at only 3.8|[plusmn]|2.4 fold of control, while chromatin associated with insulated provirus was acetylated at 12.3|[plusmn]|6.6 fold of control (p=0.04). More detailed mapping indicated a notable peak of acetylated histone associated with the cHS4 elements in the vector LTRs (averaging 20.4|[plusmn]|10.8 fold of control versus 7.1|[plusmn]|6.3 fold for more internal sequences, p=0.001), while no such bias was observed for the uninsulated vector. In one recipient the insulated vector was only expressed in 3.7% of provirus-containing cells. We found that chromatin associated with provirus in this mouse was only acetylated at 2.9 fold of control, or 5-fold lower than in the other mice of this group (p<<0.001). Taken together, these studies suggest that the cHS4 chromatin insulator increases the frequency of provirus expression by maintaining an open chromatin structure via histone acetylation, that this activity appears to peak near the cHS4 elements, and that the failure of some insulated provirus to be expressed is associated with a failure of this activity. Future studies will be needed to determine whether alternative insulators will provide a more robust level of protection.