Previously, we reported the construction of a prototype hybrid adenoretroviral vector, AdLTR-luc, which included Moloney murine leukemia virus (MoMLV) elements flanking (|[sim]| 2.7 kb 5|[prime]|; |[sim]| 1 kb 3|[prime]|) transgene cassette (Zheng et al., Nat. Biotech., 2000). Like traditional adenoviral vectors, this hybrid adenoretroviral vector is capable of transducing a wide variety of dividing and non-dividing cell types, but like retroviral vectors AdLTR-luc is able to integrate in the host cell genome resulting in long-term transgene expression. To further develop this unique vector, and improve production efficiency, we modified MoMLV sequences in the original hybrid vector. We report here the construction of a second generation hybrid vector, containing either of two transgenes: human erythropoietin (hEPO) or enhanced green fluorescence protein (EGFP). In addition, the transgene cassette included the human elongation factor 1 alpha (EF1a) promoter. Two resulting vectors, AdLTR2EF1a-hEPO and AdLTR2EF1a-EGFP, used 0.8 kb MoMLV elements (0.2 kb envelope and 0.6 kb LTR sequences in a 5|[prime]| to 3|[prime]|direction) upstream, and 0.6 kb LTR sequence downstream, of the transgene. The selection of these fragments was based on a series of gel shift assays following incubation of nuclear protein extracts with different MoMLV sequences. Production of both vectors was efficient and comparable to that obtained with first generation adenoviral vectors (|[sim]|1012 - 1013 particles/ml by Q-PCR). To evaluate if these constructs were able to integrate into genome, the rat submandibular gland A5 cell line was transduced with AdLTR2EF1a-EGFP at 100 particles/cell. After 10 days, colonies were counted and 8.1% (of 869) exhibited green fluorescence while none of the colonies (|[sim]| 900 each) obtained after similar transduction with two conventional adenoviral vectors, AdCMV-EGFP and AdEF1a-EGFP, exhibited green fluorescence. PCR assays on DNA from the cloned cells transduced with AdLTR2EF1a-EGFP were consistent with possible transgene integration into genomic DNA. To test if these constructs could mediate long-term transgene expression in vivo, AdLTR2EF1a-hEPO was delivered into rat submandibular glands at 109 particles/gland (only one gland targeted). Three of 5 rats showed elevated hEPO and hematocrit levels for two months, and in one rat elevated hEPO and hematocrit levels persisted until the end of the experiment (5 months). Transduction with the conventional adenoviral vectors, AdCMV-hEPO and AdEF1a-hEPO, resulted in elevated hEPO levels for |[sim]| 1 week. PCR assays indicated that after AdLTR2EF1a-hEPO transduction, the EF1a-hEPO sequence was found in extracted DNA at 5 months endpoint from targeted submandibular glands of all three rats expressing hEPO for at least two months. In two of these samples, PCR assays also suggested that an integration event occurred. We detected anti-hEPO antibody in sera from the two rats whose hEPO levels decreased after two months. However, in one rat with persistent high levels of hEPO for 5 months, there was no evidence of anti-hEPO antibody. Together, these results suggest that this second generation hybrid adenoretroviral vector, which is more efficiently produced than the original hybrid vector, can mediate long-term transgene expression in vivo.
Read full abstract