The 3′ End of the Chicken Hypersensitive Site-4 Insulator Has Properties Similar to the 5′ Insulator Core and Is Necessary in Conjunction with the Core for Full Insulator Activity

Abstract

Genetic correction of hematologic defects is currently impeded by inefficient vector technology. We find that vectors that insulate the correcting transgene from position effects and genotoxicity compromise viral titers. Here we present an improved vector system which utilizes a modified insulator element, without sacrificing viral titers. Specifically, our genetic and epigenetic analysis of the 1.2kb chicken β-globin hypersensitive site-4 (cHS4) insulator reveal heretofore unknown activities in regions of the chicken β-globin insulator element outside the canonical and well studied 250bp 5′ “core” element. Previously, the core insulator activity was mapped to CTCF and USF-1/2 binding sites, located only in the 5′ 250bp core. However, we find that the 5′ 250bp core alone is ineffective at shielding from position effects when it flanks transgenes, and gammaretrovirus/lentivirus vectors. In contrast, the entire 1.2kb cHS4 efficiently insulates, but significantly lowers titers of lentiviral vectors. To identify insulating activities which might be appended to the 5′ 250bp core to properly insulate transgene expression cassettes without sacrificing viral titers, we performed a structure-function analysis of the cHS4 insulator placed within the 3′LTR of a lentivirus containing the regulatory and coding sequences of human b-globin (Table 1). We compared single-copy clonal progeny of mouse erythroleukemia cells (MEL) and primary transduced and transplanted hematopoietic stem cells for position effects. Additionally, we studied repressive and activating histone marks over the transgene promoter and cHS4 in the different proviruses. Our data indicate that while all vectors containing the core reduced the coefficient of variation (CV) of human b-globin (HbA) expression, several constructs suggested that cHS4 sequences in the most 3′ 400bp (furthest from the core) may be critical to full length insulator activity. We next analyzed HbA expression in vector-corrected thalassemia mice, and generated single copy secondary CFU-S, the gold standard for studying chromatin position effects (Table 2). While all vectors containing the cHS4 core provided some ‘insulator’ activity when compared to the uninsulated vector control (conceivably by reducing the CV/clonal variegation) the full length 1.2kb insulator vector provided maximum shielding from position effects, with nearly 2.5-fold higher HbA expression compared to the uninsulated vector. These data were confirmed in secondary CFU-S. Epigenetic analyses of the vector b-globin promoter revealed that transcriptionally repressive histone modifications were decreased, and activating histone modifications increased when the last 400bp sequences of cHS4 were present. Notably, vectors carrying only the 3′ 400bp sequences of cHS4 reduced clonal variegation in MEL cells and secondary CFU-S, but did not increase HbA expressing cells both in vitro and in vivo (Table 2). However, full insulator activity was restored in MEL clones when both the 5′ 250bp core was combined with the 3′ 400bp element. The addition of the 3′ 400bp element to the core was accompanied with a significant enrichment of active histone marks and minimal repressive histone marks in the provirus as seen with the 1.2Kb insulator. These data consolidate the known insulating activity of the 5′ 250bp core element with a novel 3′ 400bp element which (together) constitutes a new insulated vector system with excellent insulating properties and viral titers. Our data have important implications in the design of gene therapy vectors, where optimal insulator activity can be achieved with a minimal reduction in viral titers.Table 1. Single copy MEL clones showing effect of insulator sequences on position effects in β-globin carrying lentivirusesMEL ClonesUninsulated5′ 250bp Core5′ 400bp2 Cores5′800bp1.2Kb3′ 400bpHbA+ cells (%)59±549±651±352±349±684±3**59±3CV of HbA expression53±437±3*40±2*42±1*42±337±1*38±1*Table 2. Effect of insulator sequences on the differentiated progeny of transduced and transplanted thalassemia hematopoietic stem cellsPrimary transplants (24 wks)MockUninsulated5′250 bp5′400bp2 cores1.2Kb3′ 400bp^RBC (M/μl)6.6±0.48.6±0.58.5±0.37.9±0.27.5±0.28.9±0.27.9±0.2^Hematecrit (%)24±232±336±136±135±238±134±1^MCHC (g/dL)24±120±129±229±227±233±1*28±1^Reticulocyte (%)29±211±410±212±711±137±112±2Vector Copy (VC)/Cell01.2±0.180.90±0.120.97±0.131.2±0.260.63±0.100.76±0.16HbA (%)/VC (HPLC)19±626±430±322±343±3**18±3HbA+ RBC/VC (FACS)40±856±562±751±9100±5**32±7^CV618±171375±36*384±29*294±29**281±16**270±44**Secondary CFU-S (single copy)^HbA+cells/CFUS (%)37±453±453±542±485±2**8±1***^CV98±1570±5*75±3*91±557±2**59±2***P<0.05; **P<0.01; ***P<0.001; HbA = human β and mouse α globin tetramers Statistical analysis was performed by ANOVA (Dunnell's multiple comparison test to the control vector sBG)*Data has not been nomialized far vector copies; CFU-S represent those screened for carrying single integrants by qPCRNote: Copy number of vector with 1.2kb cHS4 insert is significantly lower compared to uninsulated vector