Recombinant AAV-2 harboring gfp-antisense/ribozyme fusion sequences monitor transduction, gene expression, and show anti-HIV-1 efficacy.

Vector-mediated delivery of potentially antivirally active genes is a key step in somatic gene therapy including therapeutic approaches against AIDS. A crucial technical prerequisite is to monitor DNA transfer into target cells. Here, we describe recombinant infectious particles derived from the adeno-associated virus type 2 (AAV-2) that are suitable to deliver effective HIV-1-directed antisense and ribozyme genes into target cells. To monitor transduction, we designed and tested a number of fusions between indicator-coding sequences of luciferase or gfp with effective HIV-1-directed antisense or ribozyme sequences. The combination of an indicator function and an antiviral func- tion in cis allows successful identification of transduced cells and measurement of effects on the replication of HIV-1 in antisense/ribozyme-expressing cells only. The fusion genes were shown to express the indicator genes. Inhibition of HIV-1 replication mediated by the antisense/ribozyme portion of the fusion transcripts was similar to parental constructs and neither acute nor long-term toxicity of fusion genes and their gene products was observed. These results suggest the use of rAAV constructs described here as tools to study the transducibility of target cells, gene expression and efficacy of HIV-1-directed antisense and ribozyme genes.

Length variation of helix III in a hammerhead ribozyme and its influence on cleavage activity.

The previously described HIV-1 directed hammerhead ribozyme 2as-Rz12 can form with its target RNA 2s helices I and III of 128 and 278 base pairs (bp). A series of derivatives was made in which helix III was truncated to 8, 5, 4, 3, and 2 nucleotides (nt). These asymmetric hammerhead ribozymes were tested for in vitro cleavage and for inhibition of HIV-1 replication in human cells. Truncation of helix III to 8 bp did not affect the in vitro cleavage potential of the parental catalytic antisense RNA 2as-Rz12. Further truncation of helix III led to decreased cleavage rates, with no measurable cleavage activity for the 2 bp construct. All catalytically active constructs showed complex cleavage kinetics. Three kinetic subpopulations of ribozyme-substrate complexes could be discriminated that were cleaved with fast or slow rates or not at all. Gel purification of preformed ribozyme-substrate complexes led to a significant increase in cleavage rates. However, the complex cleavage pattern remained. In mammalian cells, the helix III-truncated constructs showed the same but no increased inhibitory effect of the comparable antisense RNA on HIV-1 replication.

The subcellular localization and length of hammerhead ribozymes determine efficacy in human cells.

The length requirements of the antisense portion of hammerhead ribozymes for efficacy in living cells was investigated. The HIV-1tat-directed asymmetric hammerhead ribozyme alphaYRz195 was used with a 195 nt 3'-antisense arm and a 3 nt 5'-antisense portion as well as a set of successively 3'-shortened derivatives thereof. In the 3'-antisense arm a minimum length of 20 complementary nucleotides was required for efficient association with a 645 nt target RNA transcript in vitro(for all constructs kass ranged between 0.3 and 1.8x104/M/s). The cleavage rate constants (kcleav) were independent of the length of the antisense flank and ranged between 0.8 and 1.2x10-4/s. However, the length of the antisense arms, as well as the mode of delivery and the subcellular location of the ribozymes, had a dramatic effect on efficacy in HIV-1-producing human cells. When proviral HIV-1 DNA and ribozymes were co-microinjected into the nucleus of human cells, a minimum length of 51 nt in the antisense arm was necessary for antisense- and ribozyme-mediated inhibition of HIV-1 replication. Ribozymes with shorter antisense arms were almost ineffective. Conversely, short chain ribozymes, including those with chemical modifications, were superior to long chain ribozymes when co-microinjected into the cytoplasm. When transfected, all ribozymes showed an antisense effect as well as an additional ribozyme-mediated increase in inhibition. Consequences for the design and application of ribozymes are discussed.

A spermidine-induced conformational change of long-armed hammerhead ribozymes: ionic requirements for fast cleavage kinetics.

The catalytic activity of the trans cleaving hammerhead ribozyme 2as-Rz12, with long antisense flanks of 128 and 278 nt, was tested under a wide range of different reaction conditions for in vitro cleavage of a 422 nt RNA transcript derived from human immunodeficiency virus type 1 (HIV-1). Depending on the reaction conditions, in vitro cleavage rates varied by a factor of approximately 100. Increasing concentrations of magnesium up to 1 M were found to enhance the reaction. Sodium when added simultaneously with magnesium showed an inhibitory effect on the cleavage reaction. Addition of sodium during pre-annealing, however, produced a stimulating effect. It was found that the additional inclusion of spermidine during pre-annealing further increased the reaction rate markedly. In accordance with accelerated cleavage, it was possible to identify a distinct, spermidine-induced conformer of the ribozyme-substrate complex. Under the most favourable conditions cleavage rates of 1/min were obtained, which are in the range of rates obtained for conventional hammerhead ribozymes with short antisense flanks. A comparison of thermodynamic data for short- and long-armed hammerhead ribozymes suggested that the activation entropy became unfavourable when helices I and III formed a long chain ribozyme-substrate complex. We conclude that in the absence of spermidine folding into the active conformation is impaired by increased friction of long helices, resulting in relatively low cleavage rates in vitro.