5–2.0-fold compared with those of the wild-type sequence (Fig. 2b). However, steady-state levels of the mutant wt-L that showed a wild-type-like phenotype were similar to those of the wild-type sequence, indicating that the mutant wt-L mRNA is processed by RNase III. We further investigated RNase III cleavage

activity on these mutant sequences via primer extension analyses (Fig. 2c). Mutant sequences that resulted in a higher degree of resistance to chloramphenicol were not cleaved by RNase III, while the mutant sequence (wt-L) that showed a wild-type-like phenotype was mainly cut only once at cleavage site 3, located PLX4032 in vivo to the 5′-terminus of the stem loop. Interestingly, we found that a base substitution at the RNase III cleavage site on the RNA strand to the 3′-terminus in wt-L mutant RNA in one of mutants tested here (SSL-1) abolished RNase III cleavage activity at both target sites. To further characterize the molecular basis of RNase III cleavage on bdm mRNA,

we synthesized a model hairpin RNA (bdm hp-wt) that has a nucleotide sequence between +84 and +170 nt from the start codon of bdm, encompassing RNase III cleavage sites 3 and 4-II in bdm mRNA (Fig. 1a) and used for biochemical analyses in vitro. Two additional mutant bdm hairpin RNA transcripts that contained mutations at the RNase III cleavage Z-VAD-FMK price sites derived from wt-L and SSL-1 bdm′-′cat mRNA (bdm-hp-wt-L and bdm-hp-SSL-1, respectively) were also synthesized for comparison. Incubation of the 5′-end-labeled bdm-hp-wt transcript with purified RNase III generated two major RNA fragments that corresponded to cleavage sites 3 and 4-II, while the bdm-hp-wt-L transcript was predominantly Mannose-binding protein-associated serine protease cleaved at the cleavage site 3 and bdm-hp-SSL-1 was not cleaved (Fig. 3a). These results confirmed the results of primer extension analyses on in vivo bdm′-′cat mRNA. Interestingly, RNase III cleavage of the bdm-hp-wt transcript with a radiolabeled 3′-end yielded the major cleavage product generated from the cleavage at 4-II, indicating that a majority of the initial cleavages of bdm-hp-wt

transcripts by RNase III occur at the site 4-II, and this decay intermediate is further cleaved at site 3 (Fig. 3b). A similar result, albeit less dramatic, was observed in the in vivo analysis of wild-type bdm′-′cat mRNA, which showed the synthesis of more cDNAs from the bdm mRNA cleavage products generated by RNase III cleavage at site 4-II. RNase III cleavage of the 3′-end-labeled bdm-hp-wt-L transcript produced the major cleavage product generated from the cleavage at site 3 (Fig. 3b). To test whether the altered RNase III cleavage activities on bdm-hp-wt and its derivatives are related to its RNA-binding activity, an EMSA was performed. One major band corresponding to the RNase III–RNA complex was observed when lower concentrations of RNase III (20 and 40 nM) were reacted with RNA (indicated as A in Fig.