RNAi as Antiviral Therapy: The HIV-1 Case - RNA Interference from Biology to Therapeutics - page 222

encode the early viral proteins Tat, Rev, and Nef. The idea behind this strategy is
that an early block in viral gene expression will severely impact the subsequent
expression of unspliced and singly spliced mRNAs that encode the structural HIV-1
proteins. Second, one could target HIV-1 genomic regions that are represented in all
spliced viral mRNAs, which is the case for small sequence stretches in the 5¢ UTR
and 3¢ UTR [ 63 ]. Third, as target RNA structure can effectively neutralize an RNAi
attack [ 64, 65 ], focusing on the “accessible” RNA domains is important. For HIV-1,
this strategy is clearly facilitated by the recent description of the secondary structure
of the complete HIV-1 RNA genome [ 66, 67 ]. Fourth, an important selection crite-
rion when targeting a variable virus like HIV-1 concerns the variability of the target
sequence among different virus isolates. One should ideally select targets that are
highly conserved because one wants to inhibit as many virus strains as possible. We
discovered that targeting of highly conserved genome regions may also restrict the
evolution of viral escape mutants because well-conserved sequences exhibit an
important function in HIV-1 biology, as RNA signal and/or protein-coding sequence.
RNAi-induced sequence variation may, thus, be expected to have an impact on the
viral replication capacity and fitness [ 68 ] .
We performed an extensive RNAi screen against highly conserved HIV-1 genome
sequences, which yielded about 20 potent shRNA inhibitors [ 69 ] . Stable shRNA-
expressing T cell lines were generated that were subsequently infected with HIV-1,
which yielded four durable shRNA inhibitors that restricted virus replication in a
transformed T cell line for at least 3 months [ 70 ]. Other groups have also screened
large sets of anti-HIV shRNAs [ 71 ], and effective inhibitors were described that target
regulatory sequences [ 68, 71 ] or protein-coding sequences in the gag [ 69, 73- 75 ] , pol
[ 73, 76, 77 ] , vif [ 72 ] , tat [ 69, 77- 79 ] , rev [ 69, 78, 79 ] , vpu [ 73 ] , env [ 75 ] , and nef
genes [ 72 ]. Follow-up analyses should include prolonged culturing of stably trans-
duced T cells to score the impact on cell viability. To address safety in more detail, the
off-target effects of the antiviral shRNAs on human mRNAs can be evaluated [ 80 ] .
An important lesson to be learned from various siRNA investigations is the
necessity for inclusion of appropriate controls. Several preclinical antiviral studies
used EGFP-specific siRNA as a control. Favorable therapeutic effects were observed;
however, these were later attributed to nonspecific induction of TLR7/8 interferon
pathway innate immunity triggered by the antiviral-specific sequence but not the
EGFP [ 81 ]. Another lesson comes from a study on an siRNA therapeutic for the
treatment of age-related macular degeneration in the eye. The siRNA exhibited a
therapeutic effect, but this was not likely elicited by the RNAi mechanism since the
charged siRNA molecule cannot easily penetrate cells. Instead, the clinical effect
occurred through TLR-3 signaling [ 82 ]. These two examples illustrate the impor-
tance of selecting the correct controls to ensure RNAi-specificity. For HIV-1 thera-
pies that target the viral genome, exclusive specificity can be demonstrated by the
selection of escape variants with a single point mutation in the target sequence. The
sequence-specificity of a particular RNAi effector must be demonstrated in vitro
and subsequently in appropriate in vivo models before translation into the clinical
test phase [ 4, 83, 84 ]. Guidelines for the proper testing and selection of potent and
safe shRNA inhibitors against HIV-1 have been formulated [ 70 ] .