Top of pageAbstract Background: The circumstances governing the extent to which unintegrated lentiviral DNA can persist and serve as a template for transcription and protein expression are unclear. Methods: We constructed and validated class I mutants for FIV IN (D66V and D66V+D118A) and compared them to analogous class I HIV-1 IN mutants. We documented class I properties, and performed further characterizations with lentiviral vectors and with full length viruses bearing these mutations. Results: The D66V mutation in FIV IN, alone or in combination with D118A, did not affect Gag/Pol precursor expression or proteolytic processing, particle formation, or RT production. The mutations did block integration and transduction of dividing fibroblasts (5 log reduction, RT activity-normalized comparisons). In rats injected subretinally with RT-normalized stocks, retinal pigment epithelium showed extensive transduction, as did some cells in the outer nuclear layer, persisting to at least 3 months for wild-type (WT), compared with only rare positive outer nuclear layer cells for D66V vectors. Results were different in primary post-mitotic rat neurons (retinal ganglion cells), where class I-mutant and WT vectors produced fully equivalent transduction. Transgene expression in growth-arrested fibroblasts was not inhibited at all by class I mutations (97% vs. 99% of cells transduced with WT vector at MOI = 5, comparing RT activity-normalized stocks), and expression level per cell was equal to WT. Northern blotting for reporter gene mRNA fully confirmed the protein expression results and ruled out pseudotransduction, as did use of mock vectors. mRNA levels assayed in Northern blots remained high at 8 days in growth-arrested cells infected with class I mutants. Similar cell cycle-dependent expression was seen with class I mutant HIV vectors and full length FIV. Southern blotting revealed persistence of vector DNA only in arrested cells transduced with the class I mutant vector. Quantitative real-time PCR corroborated this, showing accumulation of 2-LTR circles with the class I mutant vectors. Release from aphidicolin resulted in stable transgene expression in cells transduced with WT FIV and HIV vectors. Cells were monitored for 37 days, and expression was durable to repeated passage. In contrast, release of cells transduced with equal RT units of class I IN mutant FIV or HIV vectors resulted in steady decline of expression, from 97% to 0% of cells by day 10. Loss of transgene expression correlated to a loss of vector DNA evaluated by Southern blotting and a loss of mRNA evaluated by Northern blotting. Conclusions: In non-cycling cells, unintegrated FIV and HIV-1 vector DNAs reside in a compartment that is fully accessible to high-level transcription from internal promoters. Reporter gene expression is equivalent to that of integrated DNA if cells are growth-arrested. mRNA and protein expression levels correlate with abundant, persistent unintegrated DNA. Application of these findings to basic research will be discussed.