Background : RNAi provides new opportunities for therapeutic intervention. A limitation of the technique has been the dependence on transfection for oligonucleotide or plasmid siRNA delivery. Only certain cell lines can be transfected, and efficient transfection into primary cells is virtually impossible. In contrast, retroviral gene delivery is effective in most cell lines and many primary cells.

Methods: Oligonucleotides encoding CCR5 were ligated into lentiviral vector pRDI292 carrying the H1 promoter and a puromycin selection marker. The packaging cell line 293T was transfected with 4 plasmids (pGag/Pol, pRev, pVSV and pRDI292) to generate lentiviruses. The efficacy of the various siRNA was evaluated in GHOST cells expressing CCR5 by using FACS analysis and an anti-CCR5 antibody. The most effective siRNA lentivirus was used to transduce a collection of purified CD4 memory T-cells from healthy blood donors. Characterization of the collection included determination of CCR5/CCR2 genotype and CCR5 expression. The outcome of infection with R5 strain pNL4-3BE (MOI=0.3) was monitored by p24 in culture supernatant.
Results: The five siRNA lentiviruses resulted in 37 to 95% CCR5 silencing at 72h post-transduction of GHOST cells. The most effective, LV466, was used to transduce a collection of primary cells representative of the spectrum of CCR5 expression and permissiveness to infection (Table). LVpuro is the control vector.

Primary	Genotype*	CCR5 surface density (% pos. cells)			p24 pg/mL @7d‡
CD4 T-cells	CCR5 /proCCR5 /CCR2	Baseline	Post-transduction† LVpuro            LV466		LVpuro	LV466
Donor 11	wtD32 /p1p1 /wt64I	0	14.4	1.8	2184	146
Donor 101	wtwt /wtp1 /wt64I	7.1	39.5	3.5	16361	124
Donor 42	wtwt /p1p1 /wt64I	18.8	34.4	3.9	125645	159
Donor 22	wtwt /wtwt /wtwt	39.0	56.2	9.5	425810	12815

*wt=wild type, D32=deletion in CCR5, p1=59029C allele in CCR5 promoter, 64I= Isoleucin in CCR2 codon 64. † 7 days post-transduction, mean values from data in triplicate. ‡Average from duplicate experiments.

LV466 diminished infection of CD4 T-cells by up to 3 log. The amount of p24 in supernatant of donor cells #11, #101 and #42 decreased between days 5 and 7, corresponding to the input inoculum, indicating maximal silencing.

Conclusions: Stable RNAi can be effectively achieved in primary CD4 T-cells, a key target of HIV-1. However, cells with the highest expression of CCR5 may fail to achieve optimal silencing, thus underscoring the need to further improve delivery, integration or expression of lentiviral vectors.