Ribonucleases J1 and J2: two novel endoribonucleases in B.subtilis with functional homology to E.coli RNase E

Abstract

Many prokaryotic organisms lack an equivalent of RNase E, which plays a key role in mRNA degradation in Escherichia coli. In this paper, we report the purification and identification by mass spectrometry in Bacillus subtilis of two paralogous endoribonucleases, here named RNases J1 and J2, which share functional homologies with RNase E but no sequence similarity. Both enzymes are able to cleave the B.subtilis thrS leader at a site that can also be cleaved by E.coli RNase E. We have previously shown that cleavage at this site increases the stability of the downstream messenger. Moreover, RNases J1/J2 are sensitive to the 5' phosphorylation state of the substrate in a site-specific manner. Orthologues of RNases J1/J2, which belong to the metallo-beta-lactamase family, are evolutionarily conserved in many prokaryotic organisms, representing a new family of endoribonucleases. RNases J1/J2 appear to be implicated in regulatory processing/maturation of specific mRNAs, such as the T-box family members thrS and thrZ, but may also contribute to global mRNA degradation.

Figure 1

thrS mRNA leader: regulation by antitermination and cleavage sites. Under threonine starvation conditions, uncharged tRNA^Thr interacts with the thrS leader mRNA as illustrated, thereby increasing read-through of the terminator by stabilizing the antiterminator structure. Two RNA cleavage sites are indicated by arrows. Cleavage 1 in the AG-box (see text) was used to trace the novel endoribonuclease activity during purification. Cleavage 2, occurring next to the T-box has previously been shown to stabilize thrS read-through transcripts in vivo by leaving a stable secondary structure close to the 5′ end of the mRNA (24). E.coli RNase E is able to cleave site 2 in vivo and in vitro (19).

Figure 2

Purification of the nuclease responsible for the cleavage in thrS mRNA leader in vitro. (A) The thrS leader was transcribed in vitro using plasmid pHMS17 linearized by ClaI as template and E.coli RNA polymerase (see Materials and Methods). The thrS transcripts, one prematurely terminated at the leader terminator and a read-through transcript, were incubated in 50 μl post-transcriptional assays without the addition of proteins, with 15 μg of a B.subtilis ribosomal high salt wash (HSW) or with the most active enriched protein fraction eluting from a Superdex 200 (S 200) column. Scissors indicate the processed transcript. (B) Mapping of the processing site shown in (A) in the thrS specifier domain. Primer extension was performed with ³²P-labelled oligonucleotide HP238 on thrS mRNA synthesized in vitro in the presence of B.subtilis HSW. Arrows indicate the processing sites within the thrS specifier domain, referred to as cleavage 1 (Figure 1). A sequencing ladder was run in parallel using the same primer. (C) Purification of the endoribonucleolytic activity responsible for cleavage in the thrS specifier domain. The most active fractions from the heparin (Hep) and S 200 columns were separated on a 15% SDS–polyacrylamide gel and silver stained. Molecular weight marker proteins (M) were run in the first lane. (D) Cleavage of the thrS leader transcript by purified YkqC and YmfA proteins. thrS mRNA was prepared as described in (A) and incubated in 50 μl post-transcriptional assays without proteins, with 15 μg of a B.subtilis HSW or with 200 ng of the purified proteins, YmfA, YkqC and a mixture of YmfA and YkqC. The processed transcripts are indicated by scissors and a number (1: cleavage in the AG-box; 2: cleavage near the T-box, see Figures 1 and 3). (E) Cleavage of a thrS leader transcript generated with T7 RNA polymerase from a PCR fragment (see Materials and Methods) by purified YkqC and YmfA proteins. The fully synthesized thrS mRNA was incubated in 10 μl post-transcriptional assays without any proteins or with 200 ng of the purified proteins, YmfA and YkqC. The processed transcripts are indicated by scissors.

Figure 3

YmfA and YkqC cleave near the T-box. (A) In vitro transcription was performed as described in Materials and Methods with a PCR fragment corresponding to the thrS leader mRNA as template and T7 RNA polymerase. The terminated and full-length transcripts were purified on a 5% polyacrylamide gel and subsequently used together (lanes 1–3) or separately (lanes 4–6 for the terminated transcript and lanes 7–9 for the full-length transcript) in 10 μl post-transcriptional assays. Incubations were performed as described in Materials and Methods without any proteins or with 200 ng of the purified proteins, YmfA and YkqC. The processed transcript is indicated by scissors. A sequencing ladder was run in parallel using primer HP43, which starts at the transcriptional start site of thrS. (B) T-box processing site in the terminator and antiterminator conformations. T-box is indicated in bold.

Figure 4

Sensibility of the YkqC (A) and YmfA (B) processing activity to the 5′ end phosphorylation state of the substrate. Equal amounts of ³²P uniformly labelled tri- (lanes 1–7) and monophosphorylated (lanes 8–14) thrS leader transcripts were used as substrates in 50 μl post-transcriptional processing assays with 1 μg of YkqC or YmfA. Samples (5 μl) were taken at 0 min (lanes 1 and 8), 1 min (lanes 2 and 9), 2.5 min (lanes 3 and 10), 5 min (lanes 4 and 11), 10 min (lanes 5 and 12), 15 min (lanes 6 and 13) and 20 min (lanes 7 and 14) after addition of purified YkqC (A) or YmfA (B), directly diluted with 2.5 μl of 3× gel loading buffer to stop the reaction and run on a 5% polyacrylamide gel. The processed transcript is indicated by scissors and a number (1: cleavage in the AG-box; scissor 2: cleavage near the T-box).

Figure 5

Processing of the thrZ mRNA leader in vitro and in vivo. (A) Genetic organization map of the thrZ leader. Grey squares indicate the three T-box sequences. The PCR fragment used for in vitro transcription (see B) and the transcripts generated are also shown. (B) In vitro transcription was performed as described in Materials and Methods with T7 RNA polymerase and a PCR fragment corresponding to the thrZ leader downstream of the second transcription terminator and the beginning of the coding sequence as template (see A). The two transcripts generated, i.e. the full-length read-through transcript and a shorter one, prematurely terminated at the leader terminator, were incubated in 10 μl post-transcriptional assays without any proteins or with 200 ng of the purified proteins, YmfA and YkqC. The processed transcripts are indicated by a scissors symbol. A sequencing ladder was run in parallel using primer HP 821, which starts at the transcriptional start site of thrZ mRNA leader synthesized in vitro. Arrows indicate the processing site. T-box is indicated in bold. (C) Processing of the thrZ mRNA leader in vivo. Primer extension was performed with ³²P-labelled oligonucleotide HP814 (5′-GTCCGGAAGCTGAATGTG-3′) on total RNA from B.subtilis SSB30 (lane 1) and SSB345 (lane 2) as described in Materials and Methods. A sequencing ladder was run in parallel using the same primer. The transcriptional start site of thrZ (+1), the processing site near the T-box (scissor) and the non-specific stop at the terminator structure (T) are indicated.

Figure 6

Domain structure of B.subtilis RNases J1/J2. (A) RNases J1/J2 belong to the β-CASP family recently described by Callebaut et al. (37) as a separate group within the metallo-β-lactamase superfamily. These proteins appear to be specialized towards nucleic acids and are characterized by the presence of three highly conserved amino acids, namely motifs A, B and C. Motifs B and C are located in the RNA metabolizing metallo-β-lactamase domain (RMMBL). (B) Comparison of motifs A, B and C between RNases J1 and J2.

Figure 7

The phylogenetic distribution of RNase J orthologues in prokaryotes. Subfamilies of organisms where at least one member contains an RNase J are encircled. Grey-shaded circles indicate that every single member of this family contains an RNase J orthologue. Ratios refer to the number of members of a subfamily that possess one orthologue to RNase J1/J2 over the total number of members. Subfamilies of the Firmicutes often contain several RNase J orthologues (indicated by an asterisk). Within the Bacillales and Lactobacillales groups, there is always one orthologue clearly more related to RNase J1 than RNase J2, with the others being equally similar to both. Subfamilies of the Proteobacteria containing the classical RNase E and G complement are marked by the letter E.