Intracellular susceptibility to ribozymes in a tethered substrate-ribozyme provirus model is not predicted by secondary structures of human immunodeficiency virus type 1 RNAs in vitro.

We have assessed the sensitivity of different sites in HIV-1 genomic RNA to ribozymes. Ribozymes targeted to sequences in U5, Pol, Env, RRE, or R were positioned into nef of an infectious HIV-1 provirus. When these proviral DNAs were introduced into HeLa CD4+ cells, recombinant viruses that contain ribozymes tethered to genomic RNA or viral mRNAs were produced. The growth kinetics of ribozyme-containing viruses in CD4+ lymphocytes (MT4 cells) were distinctly delayed when compared to control viruses. On the basis of the ability of a particular ribozyme to inhibit virus replication, we inferred intracellular ribozyme-sensitive sites. We found that although ribozyme sensitivity in vitro could be correlated with predicted secondary structures of target RNAs, such in vivo correlations could not be made when using the HIV provirus model. We conclude that both Zuker algorithm computer modeling of substrate RNA secondary structures and in vitro cleavage efficiencies cannot be reliably used to determine HIV-1 ribozyme sensitive sites in vivo.