Antisense RNA regulation by stable complex formation in the Enterococcus faecalis plasmid pAD1 par addiction system

Abstract

The par stability determinant, encoded by the Enterococcus faecalis plasmid pAD1, is the only antisense RNA regulated postsegregational killing system identified in gram-positive bacteria. Because of the unique organization of the par locus, the par antisense RNA, RNA II, binds to its target, RNA I, at relatively small, interspersed regions of complementarity. The results of this study suggest that, rather than targeting the antisense bound message for rapid degradation, as occurs in most other antisense RNA regulated systems, RNA I and RNA II form a relatively stable, presumably translationally inactive complex. The stability of the RNA I-RNA II complex would allow RNA I to persist in an untranslated state unless or until the encoding plasmid was lost. After plasmid loss, RNA II would be removed from the complex, allowing translational activation of RNA I. The mechanism of RNA I activation in vivo is unknown, but in vitro dissociation experiments suggest that active removal of RNA II, for example by a cellular RNase, may be required.

FIG. 1.

Genetic organization of the par locus. The par locus is approximately 400 bp in size, delimited by the promoters for the RNA I and RNA II genes located on the right and left ends, respectively, and is shown by a heavy black line. The relative lengths of the RNA I and RNA II transcripts are depicted as arrowed lines below and above the genetic map, respectively. The fst toxin-encoding gene is depicted on the RNA I gene and transcript with a hatched box. The direct repeats DRa and DRb and the inverted repeat of the bidirectional rho-independent transcription terminator are shown by white arrows. These sequences provide the complementary regions required for interaction of RNA I and RNA II, since transcription occurs in opposite directions across the repeats. Note also that the DRa and DRb sequences overlap the translation initiation region of fst in RNA I. The sequence of the toxic Fst peptide is shown below the map.

FIG. 2.

Stabilization of RNA II by RNA I. Samples were taken from cultures at time intervals following the addition of rifampin to stop transcription initiation. RNA was purified, fractionated, transferred to nylon membranes, and probed with radiolabeled oligonucleotides specific for RNA I, RNA II, and E. faecalis 5S rRNA, as described in Materials and Methods. The 5S rRNA probe was added as a loading control. Time points are 0, 5, 10, 20, and 40 min after the addition of rifampin, shown from left to right on all gels. (A) Stabilization of RNA II in trans. Lanes 1 to 5, OG1X(pDAK611, pAM401); lanes 6 to 10, OG1X(pDAK611, pDAK704). The lower panel shows an overexposure of the RNA II bands. (B) Stabilization of RNA II in cis. Lanes 1 to 5, OG1X(pDAK102); lanes 6 to 10, OG1X(pDAK102Δ).

FIG. 3.

Gene dosage effects of par RNA levels. RNA was purified from logarithmic-phase OG1X cells containing the par RNAs in various contexts, fractionated, transferred to nylon membranes, and probed with radiolabeled oligonucleotides specific for RNA I, RNA II, and E. faecalis 5S rRNA, as described in Materials and Methods. Lanes: 1, OG1X(pDAK611); 2, OG1X(pDAK704, pDAK611); 3, OG1X(pDAK2300K); 4, OG1X(pDAK607).

FIG. 4.

Quantitative Northern blot analysis of par RNAs. Total RNA harvested from OG1X(pDAK2300K) (1.5 μg) was fractionated and transferred to a nylon membrane. Lane 1 was probed with oligonucleotides complementary to RNA I and RNA II. Lane 2 was probed with RNA I only. Lane 3 was probed with RNA II only. Densitometry revealed the following intensities for each lane. Lane 1, 2,231 net counts for RNA I and 2,253 net counts for RNA II; lane 2, 2,801 net counts for RNA I; lane 3, 3,041 net counts for RNA II.

FIG. 5.

Dissociation of RNA I-RNA II complex in vitro. In vitro-produced RNA I and RNA II were allowed to form complexes at various ratios. The preformed complexes were then challenged with a 10-fold (lanes 3 to 6) or 100-fold (lanes 7 to 10) molar excess of labeled RNA II relative to RNA I. Lane 1, labeled RNA II; lane 2, labeled RNA II with unlabeled RNA I at a 1:1 ratio; lanes 3 to 6, unlabeled preformed RNA I-RNA II complex at 1:1, 1:2.5, 1:5, and 1:10 ratios, respectively, challenged with a 10-fold molar excess of labeled RNA II relative to RNA I; lanes 7 to 10, unlabeled preformed RNA I-RNA II complex at 1:1, 1:2.5, 1:5, and 1:10 ratios, respectively, challenged with 100-fold molar excess of labeled RNA II relative to RNA I. The large black spot at the bottom of lanes 7 to 10 is due to the 10-fold-larger amounts of labeled challenge RNA II used in these assays than in those shown in lanes 3 to 6. Counts from the shifted bands as determined by ImageQuant software are as follows: lane 3, 518,812; lane 4, 110,869; lane 5, 47,118; lane 7, 422,190; lane 8, 132,348; lane 9, 51,742.

FIG. 6.

Model of the regulation of par PSK. In plasmid-containing cells (A) most RNA I (lined crescent) and RNA II (hatched rectangle) are in complex, as suggested by their 1:1 ratio observed in cells carrying wild-type par. Excess RNA II may be transcribed in order to ensure translational repression of RNA I, but any RNA II not in complex is rapidly degraded by a cellular RNase (hatched Pac-Man). RNA II is more slowly removed from the RNA I-RNA II complex, perhaps by a different RNase (lined Pac-Man). In cells that retain the plasmid (B), RNA II removed from the complex is rapidly replaced by newly synthesized RNA II. However, in cells that lose the plasmid (C), RNA II cannot be replaced, RNA I is translated by ribosomes (bilobed shaded shapes), and the toxin kills the cell (rectangles in the membrane). Empty circles represent pAD1 DNA.