300. AAV Vector-Mediated In Vivo Reprogramming of Various Cell Types in Adult Mice

The 2006/7 discovery that somatic cells of different tissue origins can be converted into induced pluripotent stem cells (iPSC) by mere introduction of a cocktail of “reprogramming” factors, such as Oct4, Klf4, Sox2 and c-Myc (OKSM), has revolutionized the field of regenerative medicine. A major part of this success story is due to the fact that delivery of the OKSM combination (or variants thereof) to cultured target cells can be accomplished using virtually any of the existing non-viral or viral gene transfer systems. Here, we specifically engineered Adeno-associated viral (AAV) vectors for OKSM delivery, based on their superior safety profile and their typical persistence as episomal DNA that is lost during cell division, which we considered beneficial for reprogramming. A third advantage is that AAV vectors can be pseudotyped with natural or synthetic viral capsids exhibiting high specificities and efficiencies in a given cell. Indeed, we reproducibly found that OKSM expression from four AAV vectors – each encoding a single factor under the strong and ubiquitous SFFV promoter, and pseudotyped with the potent chimeric AAV-DJ capsid – resulted in reprogramming of murine embryonic fibroblasts. Notably, the up to 0.2% reprogramming efficiency obtained with our AAV vectors is comparable to the best lentiviral system. Curiously, and defying our original prediction, we observed by PCR that the AAV vector DNA persisted in the reprogrammed cells even after extended cultivation for up to 21 passages. Using a combination of LAM-PCR and SureSelect technologies, we were then able to identify a variety of OKSM vector integration sites in the mouse genome, including a previously reported hot spot for AAV vector insertion. We speculate that the unusual high frequency of genomic AAV integration in iPSC is either related to mutational inactivation of genes that inhibit reprogramming, or merely a by-product of the relaxation of the chromatin structure triggered by OKSM expression. Based on the robust ex vivo efficiency, we next asked whether our OKSM vectors could also mediate reprogramming in vivo, and therefore applied different doses of AAV8-pseudotyped variants to adult mice via tail vein injection. Strikingly, four to eight weeks later, we found various small and large teratomas in their livers that we identified through histological analysis and isolation of residual iPSC. Moreover, by week eight, one mouse showed signs of extra-hepatic reprogramming in the pancreas and the thoracic cavity, congruent with the in vivo tropism of AAV serotype 8. These features make our novel AAV vector-mediated in vivo reprogramming strategy comparable to our previously published OKSM-transgenic mice (Nature 2013, 502:340-5), with the added benefits that AAV vectors can be applied in any mouse strain or other animals, and can be re-targeted to specific tissues, which opens up entirely novel avenues for stem cell research.