Poor in vivo transduction of myocardium by lentiviral vectors is likely caused by a post-entry block of transduction

Abstract

Abstract Funding Acknowledgements Type of funding sources: Foundation. Main funding source(s): Stichting LSH-TKI Introduction In cardiac gene therapy, two prototypic viral vectors are typically considered for long-term transgene expression, i.e. lentivirus (LV) and adeno-associated virus (AAV) vectors. Compared to AAV vectors, LV vectors have the advantages of a larger insert capacity and the ability to confer permanent transgene expression due to integration of the vector genome in the host cell’s chromosomal DNA . However, standard VSV G protein-pseudotyped LV vectors (VSV G-LV vectors) has shown suboptimal transduction of cardiac myocytes in vivo. Purpose In this study, we aimed to improve LV-mediated transduction of cardiomyocytes in vivo by evaluating LV vectors carrying various different envelopes and compared the most potent variants with AAV1 capsid-pseudotyped AAV2 vectors (AAV2/1 vectors). Methods We made a panel of LV vectors pseudotyped with envelope proteins from 76 different viruses carrying an expression cassette for enhanced green fluorescent protein (eGFP). Transduction efficiencies in neonatal rat ventricular myocytes (NRVMs), human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) and cardiomyocyte progenitor cells (CMPCs) were determined by fluorescence-activated cell sorting (FACS) as percentage eGFP-positive cells. Male NOD -SCID mice aged 3-6 months were injected directly into the left ventricular myocardium with eGFP-encoding VSV G-LV and AAV2/1 vectors at doses of 1.4 x 10^7 infectious units and 2.0 x 10^10 genome copies, respectively, or PBS as a control. Mice were sacrificed one and four weeks after surgery. Per heart, left ventricular and eGFP-positive area was determined from 40 – 50 slides and the ratio was used as a measure of transduction efficiency. Results Of the panel of pseudotyped LVs none was able to consistently outperform VSV G-LV vectors, which reached transduction efficiencies in vitro of 58% in NRVMs to 96% in hiPSC-CMs and 100% in CMPCs . In AAV2/1 vector-injected hearts, the eGFP expressing area in the left ventricle increased from 1.07% in week 1 to 5.42% in week 4. In contrast, eGFP expression in LV-injected hearts was 0.31% and 0.16% after 1 week and 4 weeks, respectively (figure 1). Conclusion In vitro screening of a panel of pseudotyped LV vectors showed standard VSV G-LV vectors as top performer. Despite encouraging in vitro results and previously reported findings, VSV G-LV vectors was unable to provide for robust in vivo transduction. The observation that none of the 76 different envelopes yielded a meaningful improvement of lentiviral transduction rates together with the discrepancy between in vitro and in vivo transduction efficiencies, suggest that a post-entry block of transduction likely represents an important component of the poor in vivo performance.Figure 1in vivo transduction