In many eukaryotes, expression of nuclear-encoded mRNA can be strongly inhibited by the presence of a double-stranded RNA (dsRNA) corresponding to exon sequences in the mRNA (refs 1,2). The use of this "RNA interference" (RNAi) in mammalian studies had lagged well behind its utility in lower animals because uninterrupted RNA duplexes longer than 30 base pairs trigger generalized cellular responses through activation of dsRNA-dependent protein kinases. Recently it was demonstrated that RNAi can be made to work in cultured human cells by introducing shorter, synthetic duplex RNAs (approximately 20 base pairs) through liposome transfection. We have explored several strategies for expressing similar short interfering RNA (siRNA) duplexes within cells from recombinant DNA constructs, because this might allow long-term target-gene suppression in cells, and potentially in whole organisms. Effective suppression of target gene product levels is achieved by using a human U6 small nuclear RNA (snRNA) promoter to drive nuclear expression of a single RNA transcript. The siRNA-like parts of the transcript consists of a 19 base pair siRNA stem with the two strands joined by a tightly structured loop and a U1-4 3' overhang at the end of the antisense strand. The simplicity of the U6 expression cassette and its widespread transcription in human cell types suggest that this mode of siRNA delivery could be useful for suppressing expression of a wide range of genes.

Several types of small RNAs have been proposed as gene expression repressors with great potential for use in gene therapy. RNA polymerase III (pol III) provides an ideal means of expressing small RNAs in cells because its normal products are small, highly structured RNAs that are found in a variety of subcellular compartments. We have designed cassettes that use human pol III promoters for the high-level expression of small RNAs in the cytoplasm, nucleoplasm, and nucleolus. The levels and subcellular destinations of the transcripts are compared for transcripts expressed using the U6 small nuclear RNA (snRNA), 5S ribosomal RNA (rRNA), and the 7SL RNA component of the signal recognition particle. The most effective location for a particular inhibitory RNA is not necessarily predictable; thus these cassettes allow testing of the same RNA insert in multiple subcellular locations. Several small interfering RNA (siRNA) inserts were tested for efficacy. An siRNA insert that reduces lamin expression when transcribed from the U6 snRNA promoter in the nucleus has no effect on lamin expression when transcribed from 5S rRNA and 7SL RNA-based cassettes and found in the nucleolus and cytoplasm. To test further the generality of U6-driven siRNA inhibitors, siRNAs targeting HIV were tested by co-transfection with provirus in cell culture. Although the degree of HIV-1 inhibition varied among inserts, results show that the U6 cassette provides a means of expressing an siRNA-like inhibitor of HIV gene expression.

No abstract

Double-stranded RNAs (ds RNAs) are thought to be the cytoplasmic determinants responsible for the phenomenon of transmissible hypovirulence in the chestnut blight fungus Endothiaparasitica [Murr.] Anderson. The three major ds RNA components associated with the North American hypovirulent strain, Grand Haven 2, were characterized with respect to molecular-hybridization specificity and RNase T1-digestion patterns. The large (L-RNA; %9 kilobase pairs) and middle-sized (M-RNA; -3,5 kilobase pairs) ds RNA components cross-hybridized under stringent conditions and exhibited indistinguishable partial and complete RNase T1 digestion patterns relative to their 5' and 3' termini. These results suggest that M-RNA was derived from L-RNA by an internal deletion event. The small (S-RNA; =1 kilobase pair) RNA was unrelated to L-and M-RNA by these criteria. However, all three ds RNA components contained RNase Ti-resistant oli-

No abstract