The role of the OOP antisense RNA in coliphage lambda development.

We have made a derivative of bacteriophage lambda that makes no OOP antisense RNA. The mutant phage carries a point mutation that inactivates the OOP promoter, po. The phages lambda + and lambda po- have identical plaque morphologies, one-step growth curves, and frequencies of lysogenization of a sensitive host. OOP RNA synthesis is weakly repressed by the Escherichia coli LexA protein. Consonant with this inducibility of OOP RNA synthesis by ultraviolet light, we find a two-fold greater phage burst following ultraviolet induction of a lambda + than of a lambda po- prophage. In lambda + infections, OOP RNA causes two cleavage events in cll mRNA: one is in the 3'-end of the coding region, and the second is in the intercistronic region between the cll and O genes. The cll gene fragments are subject to additional hydrolytic events, and cll mRNA levels are several-fold lower in lambda + than in lambda po- infections late in the infection cycle. However, O mRNA levels are almost unaffected by the po- mutation.

RNase III-dependent hydrolysis of lambda cII-O gene mRNA mediated by lambda OOP antisense RNA.

The 77-nucleotide OOP antisense RNA of bacteriophage lambda complements lambda cII-O mRNA in a region that includes 55 nucleotides at the 3' end of the cII gene and 22 nucleotides in the intercistronic region between the cII and O genes. OOP RNA, produced from multicopy plasmids, inhibits lambda cII gene expression by approximately 100-fold through an RNase III-dependent mechanism. Using primer extension analysis of cellular RNA isolated from an induced lambda lysogen that contains an OOP DNA plasmid, we have identified a cleavage site in cII-O mRNA within the region of complementarity with OOP RNA, at 13 nucleotides from the 3' end of that region. Ribonuclease protection experiments demonstrate that almost all cII-O mRNA in this overlap region is cleaved when OOP RNA is overproduced in RNase III+ cells but not in RNase III- cells. RNA fragments are detected that extend into the O gene from the cleavage sites, while the sister fragments that extend into the cII gene cannot be detected and must be eliminated by additional hydrolytic events. Differences in levels of uncleaved mRNA between RNase III+ and RNase III- cells are much less at several hundred nucleotides to either side of the target region. An alternate OOP RNA-dependent hydrolytic process occurs in RNase III- cells that results in cleavages in one of two regions, one close to the cleavage site observed in RNase III+ cells, and the second several nucleotides beyond the end of the complementary region between OOP RNA and cII-O mRNA. In this latter case, the fragments that extend into the cII gene are stable, while the sister O gene fragments are destroyed, in direct contrast to the RNase III-dependent process.

The cleavage specificity of RNase III.

We determined sites in lambda cII mRNA that are cleaved by RNase III in the presence of lambda OOP antisense RNA, using a series of OOP RNAs with different internal deletions. In OOP RNA-cII mRNA structures containing a potential region of continuous double-stranded RNA bounded by a non-complementary unpaired region, RNase III cleaved the cII mRNA at one or more preferred sites located 10 to 14 bases from the 3'-end of the region of continuous complementarity. Cleavage patterns were almost identical when the presumptive structure was the same continuously double-stranded region followed by a single-stranded bulge and a second short region of base pairing. The sequences of the new cleavage sites show generally good agreement with a consensus sequence derived from thirty-five previously determined cleavage sequences. In contrast, four 'non-sites' at which cleavage is never observed show poor agreement with this consensus sequence. We conclude that RNase III specificity is determined both by the distance from the end of continuous pairing and by nucleotide sequence features within the region of pairing.

OOP RNA, produced from multicopy plasmids, inhibits lambda cII gene expression through an RNase III-dependent mechanism.

OOP RNA is a major short (77 bases) transcript that is made from bacteriophage lambda DNA both in vivo and in vitro. OOP RNA is synthesized in the opposite direction to mRNA for the lambda cII gene, and the final 55 bp of the OOP region overlaps the 3' end of the cII gene. We find that a multicopy plasmid containing an OOP DNA fragment inhibits cII expression from a derepressed prophage by approximately 100-fold, using an in vivo assay in which cII protein activates galactokinase synthesis from a cII-dependent promoter on a multicopy plasmid. A large inhibitory effect is also observed when the po promoter for OOP RNA is replaced by the strong lambda pL promoter, but not when po is deleted. Plasmids that provide a large excess of "anti-OOP" RNA (RNA that is complementary to OOP RNA) make OOP RNA a less effective inhibitor of cII expression. Inhibition by the OOP DNA plasmid is not observed in an Escherichia coli strain deficient in RNase III. We propose that the 3' end of cII mRNA and OOP RNA form a double-stranded complex that is a substrate for the host enzyme RNase III, resulting in degradation of cII mRNA. Deletion studies on the OOP DNA plasmid indicate that no specific sequence between the promoter and terminator stem structure is required for the inhibitory effect. Lambda cII expression from an induced prophage is increased twofold in the presence of a large excess of anti-OOP RNA. This experiment, in which the prophage is the sole source of OOP RNA, suggests a physiological role for OOP RNA in regulating cII-gene expression.