Silencing of HIV-1 genes using siRNA-expressing genetic constructs

Abstract

Today antiviral chemotherapy is the main methodto control HIV infection. Common medications influ-ence primarily the key enzymes of HIV-1, such asreverse transcriptase, integrase, and protease. How-ever, in spite of a large quantity of existing anti-HIVdrugs, antiviral therapy is often ineffective. The mainreason for ineffectiveness of treatment is adaptivemutagenesis of the virus, causing the appearance ofnew HIV-1 variants resistant to antiviral drugs. Thetoxicity and high cost of such drugs also present aserious problem.It is known that RNA interference is a powerfulnatural mechanism involved in the regulation of geneexpression and protection of cells against foreignRNA; it is also called ancient immunity operating atthe RNA level [1]. Recently, it has been found thatRNA interference is regulated during the developmentof an organism [2]. Viruses have the capacity to regu-late RNA interference. In the course of evolution, theyacquired the ability to turn off RNA interference in thehost cell to ensure their own reproduction [3]. Severalyears ago, we proposed artificial RNA interference forboth the investigation of its mechanisms and the use ofsilencing of HIV-1 genes in gene therapy. Now severalresearch groups are working in this field [4–6]. In thispaper, we report the results of testing three geneticconstructs for suppression of HIV-1 production inhuman cells.We induced artificial RNA interference using shorthairpin genetic constructs. Such constructs containshort palindromic sequences separated by a shortloop. The palindromic regions of RNA moleculesform hairpins. Owing to the existing cell mechanismof RNA monitoring, such double-stranded RNAs(dsRNAs) are detected and processed by Dicer(RNase III) to yield double-stranded small interferingRNA (siRNA). At the second stage of RNA interfer-ence, only the antisense siRNA strand remains withinthe RNA-induced silencing complex (RISC). A nucle-otide sequence of the siRNA within RISC is used tosearch for complementary regions in cell RNAs. In thecase of an HIV-1-specific siRNA, RISC recognizes anavailable single-stranded region in viral transcriptsand cuts it.The efficiency of siRNA is determined by a num-ber of factors. In particular, it depends on the siRNAnucleotide composition, which should meet the fol-lowing empiric requirements [7, 8]: a GC content of30–52%, at least three A/U in positions 15–19 of thesense strand, lack of internal repeats, A in positions 19and 3 of the sense strand, U in position 10 of the sensestrand, lack of G/C in position 19 of the sense strand,lack of G in position 13 of the sense strand, etc. Theavailability of target RNA is the second factor deter-mining the efficiency of siRNA. If the mRNA regioncomplementary to siRNA is unavailable (for example,it is bound to a protein or included in a duplex), thenit cannot be degraded.We began our study when these requirements wereunknown, so we proceeded from the assumptions thatthe best targets in HIV-1 transcripts are the most con-served regions within the functionally importantdomains and that biologically active siRNAs can beisolated in direct tests for suppression of virus produc-tion.We searched for HIV-1 conserved regions with theuse of the BLAST program and multiple sequencealignment. Within the AF316544 sequence of HIV-1subtype A, we found the most conserved regions cor-responding to the protease, reverse transcriptase, andTat domains. These regions were chosen as targets forsiRNA-encoding constructs. The resulting siRNAs