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-I fat-directed asymmetric hammerhead ribozyme αYRz195 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 k(ass) ranged between 0.3 and 1.8 x 10 ⁴/M/s). The cleavage rate constants (k(cleav) were independent of the length of the antisense flank and ranged between 0.8 and 1.2 x 10 ^-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.