Background:  The Ref1 and Lv1 post-entry restrictions in human and monkey cells have been analyzed for lentiviruses in the primate and ungulate groups, but no data exist for the third group (feline).

Methods:  We compared feline immunodeficiency virus (FIV) to TRIM5a-restricted (HIV-1, equine infectious anemia virus [EIAV]) and unrestricted (NB-tropic murine leukemia virus [NB-MLV]) retroviruses across wide ranges of viral inputs in cells from multiple primate and non-primate species. We also characterized restrictions conferred to permissive feline and canine cells engineered to express rhesus and human TRIM5a proteins, and we achieved stable RNAi knock-down of endogenous rhesus TRIM5a in FRhK4 cells.

Results:  Expression of rhesus or human TRIM5a proteins in feline cells restricted FIV, impairing pseudotyped vector transduction and viral replication, but rhesus TRIM5a was more restricting than human TRIM5a. Stable RNAi knock-down of endogenous rhesus TRIM5a resulted in marked increases in FIV and HIV-1 infectivities, but in no effect on NB-MLV. A panel of non-primate cell lines varied widely in susceptibility to lentiviral infection, but RT-normalized FIV and HIV-1 vectors varied concordantly. In contrast, in human and monkey cells, relative restriction of FIV compared with HIV-1 varied from none to substantial, with a marked relative infectivity deficit for FIV vectors observed in human T-cell lines. Endogenous and introduced TRIM5a restrictions of FIV could be titrated by co-infections with FIV, HIV-1, or EIAV virus-like particles. Arsenic trioxide had complex, and also TRIM5a-independent, enhancing effects on lentiviral but not NB-MLV infection. A canine cell line did not support restriction by human TRIM5a, and only limited restriction by rhesus TRIM5a.

Conclusions:  Rhesus and human TRIM5a proteins restrict FIV in a saturable manner. Stable knock-down of rhesus TRIM5a released HIV-1 and FIV from LV1 restriction. FIV VLP saturated LV1 restriction less effectively than HIV virus-like particles (VLP). Degree of restriction by introduced TRIM5a proteins was dependent on the cell line. A canine cell line supported restriction much less effectively than other cells, raising the possibility that these proteins may not function autonomously or that a canine factor may interfere.