Search tips
Search criteria 


Logo of jbacterPermissionsJournals.ASM.orgJournalJB ArticleJournal InfoAuthorsReviewers
J Bacteriol. 2005 July; 187(14): 4698–4706.
PMCID: PMC1169516

Addition of Poly(A) and Heteropolymeric 3′ Ends in Bacillus subtilis Wild-Type and Polynucleotide Phosphorylase-Deficient Strains


Polyadenylation plays a role in decay of some bacterial mRNAs, as well as in the quality control of stable RNA. In Escherichia coli, poly(A) polymerase I (PAP I) is the main polyadenylating enzyme, but the addition of 3′ tails also occurs in the absence of PAP I via the synthetic activity of polynucleotide phosphorylase (PNPase). The nature of 3′-tail addition in Bacillus subtilis, which lacks an identifiable PAP I homologue, was studied. Sizing of poly(A) sequences revealed a similar pattern in wild-type and PNPase-deficient strains. Sequencing of 152 cloned cDNAs, representing 3′-end sequences of nontranslated and translated RNAs, revealed modified ends mostly on incomplete transcripts, which are likely to be decay intermediates. The 3′-end additions consisted of either short poly(A) sequences or longer heteropolymeric ends with a mean size of about 40 nucleotides. Interestingly, multiple independent clones exhibited complex heteropolymeric ends of very similar but not identical nucleotide sequences. Similar polyadenylated and heteropolymeric ends were observed at 3′ ends of RNA isolated from wild-type and pnpA mutant strains. These data demonstrated that, unlike the case of some other bacterial species and chloroplasts, PNPase of Bacillus subtilis is not the major enzyme responsible for the addition of nucleotides to RNA 3′ ends.

Polyadenylation is an important posttranscriptional modification of prokaryotic, eukaryotic, and organellar RNA. In Escherichia coli, polyadenylation has been shown to play a role in the process of decay, and the enzymes responsible for both polyadenylation and degradation are known (reviewed by Grunberg-Manago [15]). Poly(A) polymerase I (PAP I), an enzyme belonging to the nucleotidyltransferase family, adds poly(A) extensions to the 3′ ends of mRNAs, as well as to tRNA and rRNA (22). It has been shown that such extensions aid in the 3′-to-5′ exoribonucleolytic degradation of RNAs, particularly for the degradation by polynucleotide phosphorylase (PNPase) of RNAs containing structured ends, such as transcription terminators (1, 9, 13, 16, 20, 35). Coordination of endo- and exonucleolytic activities is thought to occur through the physical interaction of RNase E and PNPase, which, together with RNA helicase RhlB and enolase, form the degradosome (reviewed in reference 8). Physical interaction between PAP I and RNase E has also been observed (28), suggesting that polyadenylation may also be a part of the coordinated decay process.

PNPase can act as a 3′-to-5′ phosphorolytic exoribonuclease or as an RNA polymerase, depending on the availability of phosphate and ribonucleoside diphosphates. The possibility of in vivo polymerase activity of PNPase has gained more attention recently. In an E. coli strain deficient for PAP I, Mohanty and Kushner observed PNPase-dependent addition of heteropolymeric 3′ tails in E. coli (24). In Streptomyces coelicolor, cDNAs cloned from two mRNAs and from 23S and 16S rRNA exhibited heteropolymeric extensions (2). S. coelicolor PNPase was shown by Sohlberg and colleagues to have poly(A) polymerase activity in vitro, and it is likely that PNPase is responsible for 3′-end addition in this organism (32). In Synechocystis and spinach chloroplast, Rott and colleagues have reported that 3′-end addition is carried out by PNPase, resulting in heterogeneous, poly(A)-rich tails (30).

The molecular mechanism of RNA processing in Bacillus subtilis may be different from that of E. coli. Sequence homologues of E. coli RNase E, RNase II, and oligoribonuclease, major enzymes involved in mRNA decay, are absent from the B. subtilis genome (reviewed by Condon [10]). In contrast to E. coli, PNPase is believed to be the main enzyme responsible for 3′-to-5′ exonucleolytic activity in B. subtilis (11). However, PNPase is not essential for viability; a pnpA insertional mutant grows well at 37°C (34). B. subtilis has a single gene coding for a recognizable RNA nucleotidyltransferase, which Raynal et al. have demonstrated specifies tRNA CCA-adding activity, rather than poly(A) polymerase activity, in vitro (29). Although no gene encoding a poly(A) polymerase in B. subtilis has been identified, considerable polyadenylation of B. subtilis RNA has been demonstrated by Sarkar and colleagues (Gopalakrishna and Sarkar [14] and Karnik et al. [19]). Thus far, there has been only one report—also by Cao and Sarkar—concerning the sequence of 3′ tails added posttranscriptionally to a specific transcript in B. subtilis (5).

A major question addressed in the current report was whether PNPase functions in B. subtilis as the polyadenylation enzyme. A large number of 3′ tails, added to nontranslated and translated RNAs, was sequenced in both wild-type (wt) and PNPase-deficient (pnpA) strains. Our results suggest that mature RNAs and degradation intermediates contain modified 3′ ends consisting of either poly(A) or heteropolymeric tails but that PNPase plays, at best, a minor role in these modifications.


Bacterial strains.

The prototrophic B. subtilis strain PY79 (36) was used as the wt strain for isolation of RNA used to make cDNA for cloning. B. subtilis CV2 (33) carries a copy of the B. thuringiensis cry1Aa gene. Strains PY79 and CV2 were transformed with total DNA from strain BG119, which has a kanamycin resistance gene cassette replacing a portion of the pnpA gene encoding PNPase (34). Selection for resistance to 5 μg/ml kanamycin yielded pnpA derivatives of PY79 and CV2. For the poly(A) sizing assays (Fig. (Fig.1),1), the wt strain was BG1, which is trpC2 thr-5. The pnpA-disrupted strain for the experiment in Fig. Fig.1A1A was BG119 (pnpA disrupted by a kanamycin resistance cassette), which has been described previously (34). To make the cca disruption (Fig. (Fig.1B),1B), the cca gene from BG1 was amplified and cloned as a HindIII-EcoRI fragment into pGEM-9Zf(+) (Promega). A 485-base-pair EcoRV-PstI fragment in the cca gene coding sequence was replaced with a spectinomycin resistance gene, generating plasmid pIO2. Plasmid pIO2 DNA was linearized with ScaI and used to transform BG1 to spectinomycin resistance, with selection for growth on 200 μg/ml spectinomycin, giving BG254. The pnpA cca double mutant strains (Fig. (Fig.1B,1B, lanes 3 and 4) were constructed by transformation of BG254 with genomic DNA from BG119 or BG116 (same as BG119 but with the pnpA gene disrupted by a chloramphenicol resistance cassette [34]). The pnpA inducible strain (Fig. (Fig.1A,1A, lane 4) was BG302, which was constructed by transformation of BG119 with plasmid pNP18 DNA. Plasmid pNP18 is a derivative of pDR67 (17) that has the pnpA coding sequence cloned as an EcoRI-XbaI fragment. The isopropyl-β-d-thiogalactopyranoside (IPTG)-inducible Pspac promoter of pDR67 drives pnpA expression in this construct.

FIG. 1.
Poly(A) sizing assay. (A) Lane 1, labeled 18-nt oligonucleotide size marker; lane 2, RNA from wt strain; lane 3, RNA from pnpA strain BG119; lane 4, RNA from pnpA strain BG302 containing an IPTG-regulated pnpA gene at the amyE locus and grown in the presence ...

Growth of bacteria and isolation of RNA.

B. subtilis strains were grown at 37°C in LB medium, and aliquots were withdrawn during log-phase (250 Klett units with red filter). For analysis of cry1Aa mRNA, which is expressed during sporulation in B. subtilis (33), aliquots from CV2 and CV2 pnpA were withdrawn 2 h after the onset of stationary phase. Total RNA was isolated using RNeasy mini columns, as described by the manufacturer (QIAGEN). For the experiments shown in Fig. Fig.1,1, RNA was isolated by the hot phenol method described by Köhrer and Domdey (21).

Poly(A) tail length assays.

Poly(A) tail lengths were determined as described previously (3, 4). RNA (10 μg) was end labeled with [32P]pCp and RNA ligase. Labeled RNAs were then digested with a combination of RNase A and RNase T1; this combination cleaves all phosphodiester bonds in the RNAs except those between adjacent A residues. The poly(A) tails that remained following RNase digestion were separated on 12% polyacrylamide gels and visualized by autoradiography.

RT-PCR analysis of polyadenylation and cloning of 3′ ends from specific transcripts.

Five hundred nanograms of total RNA was ligated to 200 ng of a hybrid RNA-DNA anchor oligonucleotide (12) (pUUUAACCGCATCCTTCTCT [RNA shown in italics]; Dharmacon) in 20 μl using 40 units of T4 RNA ligase (Amersham Pharmacia Biotech). Five microliters of the ligation reaction mixture was the template for reverse transcription-PCRs (RT-PCRs) using the Superscript one-step RT-PCR system (Invitrogen), as outlined by the manufacturer. It should be noted that the RT-PCRs were not quantitative. The primers for the reverse transcription reactions (see Fig. Fig.2A)2A) were as follows: oligonucleotide a, 5′-AATTCCAAGAATTCGAGAAGGATGCGGTTAAA-3′; oligonucleotide b, same as primer a but with three additional T residues at the 3′end; and oligonucleotide c, same as primer a but with eight additional T residues at the 3′end. For amplification of the various RNAs, gene-specific primers (70 ng) internal to the gene of interest were designed to be located 200 to 300 bases upstream from the 3′ end of the gene. The primers were as follows: 5′-TATAATGAATTCTGAGACAGTTCGGTCCCTATC-3′ for 23S rRNA; 5′-TCCAAAGGCGGCATAGCCAAG-3′ for tRNACys-Leu; 5′-CTACGTCTTCGGATATGGCTGAGT-3′ for rnpB; 5′-TGACAAAGCTGGCAAACTA-3′ for rpsD; and 5′-AAAACTGCAGGAGGTGCGTACACTTCTCGT-3′ for cry1Aa.

FIG. 2.
Strategies used in the cloning of RNA 3′ ends. (A) An anchor oligonucleotide (hatched rectangle) was ligated to total RNA, followed by reverse transcription using a complementary oligonucleotide (black rectangle) to generate cDNA. Reverse transcription ...

Cloning and sequencing of cDNA.

Five microliters of the RT-PCR mixtures (described above) was used to clone into the ZeroBlunt Topo PCR (Invitrogen) or pGEM-T Easy (Promega) cloning vectors, according to the manufacturer's procedure. To ensure independence of the clones, multiple RT-PCRs were carried out on the same RNA, as well as on different RNA preparations. Plasmid DNA from at least five clones obtained for any given reaction was digested with restriction enzymes to verify the presence of a cloned fragment, as well as to screen for sizes of inserts. Fragments of different sizes were chosen for sequencing. For more extensive analysis of rnpB transcripts, a number of fragments of the same size were sequenced.

For RT-PCRs using oligonucleotide c, clear amplification products were not observed in a conventional 30-cycle PCR. When these reactions were further reamplified, discrete products were obtained. Fifty-six clones from these products, arising from 23S rRNA, rnpB RNA, and cry1Aa mRNA and containing 3′ tails, were sequenced. Given the possibility of generating artifacts in a double amplification, we also cloned and sequenced 69 more clones from the initial RT-PCRs. Several clones identical to those obtained after double amplification were obtained from the initial PCRs, suggesting that no artifacts had been obtained in the double amplification. The data presented in Fig. Fig.33 to to55 contain tail sequences from both single and double amplifications.

FIG. 3.
Sequences from cloned 3′ ends of stable RNAs. (A) 23S rRNA. (B) tRNACys-Leu. Coding genes are depicted by shaded rectangles. Double dash to the left indicates that the depicted gene is truncated towards the 5′ end. For 23S rRNA, the double ...
FIG. 5.
Sequences from cloned 3′ ends of mRNAs. The figure is set up in the same format as Fig. Fig.33.


Poly(A) tails in wt and pnpA strains.

To analyze the level of tails and the length of tails globally, a poly(A) sizing assay was performed on RNA isolated from wt and pnpA mutant strains. In this assay, total RNA is 3′ end labeled with T4 RNA ligase and [32P]pCp, followed by digestion with RNase A and T1, which cleave after C, G, and U residues, but not after A residues. The results in Fig. Fig.1A,1A, lanes 2 and 3, demonstrated that polyadenylation was occurring in wt and pnpA mutant strains, with a tail length of up to 40 nucleotides (nt) in both strains. Lane 4 in Fig. Fig.1A1A contained RNA isolated in a pnpA strain that had an IPTG-inducible pnpA gene at the amyE locus and that was grown in the presence of IPTG and showed a tail pattern similar to that of the wt.

The cca gene is the only one in B. subtilis that clearly belongs to the RNA nucleotidyltransferase gene family that includes poly(A) polymerases. It was shown to encode a tRNA CCA-adding activity in vitro (29). To test whether the cca activity might be involved in the addition of RNA tails in vivo, double mutants of B. subtilis that were deficient in both pnpA and cca were constructed. Lanes 3 and 4 in Fig. Fig.1B1B show the results of the poly(A) sizing assay for RNA isolated from two versions of the pnpA cca double mutant. Clearly, neither of these activities was necessary for the wt pattern of polyadenylation.

Generation of a collection of cDNA ends from nontranslated and translated RNAs.

To analyze the specific nature of RNA 3′ ends in B. subtilis, an oligonucleotide was ligated to the 3′ ends of total RNA, followed by annealing of a complementary oligonucleotide that was used to prime cDNA synthesis and to amplify the cDNA in conjunction with a gene-specific forward primer (Fig. (Fig.2A).2A). Gene-specific oligonucleotides were designed to amplify the 3′ ends of 23S rRNA, tRNACys-Leu (the last two genes of the trnD operon), rnpB RNA (the RNA subunit of RNase P), and rpsD and cry1Aa mRNAs. Reverse transcription was carried out using one of three different oligonucleotides, all complementary at their 5′ ends to the anchor ligated to the RNA but carrying either eight, three, or no additional T residues at the 3′ end (Fig. (Fig.2A).2A). Oligonucleotide a, which is simply complementary to the ligated anchor oligonucleotide, is not selective, but oligonucleotides b and c are selective for the presence of adenosine residues at the site of ligation of the anchor oligonucleotide. Figure Figure2B2B shows the results of RT-PCRs carried out using this strategy on RNA obtained from wt and pnpA strains. Since the RNA ligation reactions from the RNAs derived from the wt and pnpA strains may have had different efficiencies, the most meaningful comparison was between amplifications using the same ligation reaction but different oligonucleotides (a, b, and c). There was a two- to fivefold difference in the amount of RT-PCR product amplified with nonselective oligonucleotide a compared to RT-PCR product using oligonucleotide b. Little or no amplification was observed for reactions carried out with oligonucleotide c. Control PCR amplification of clones possessing no poly(A) or with an 8-A residue extension were carried out to verify that oligonucleotides a, b, and c were all able to amplify to a similar extent (data not shown). The difference in the amount of amplification using oligonucleotides a, b, and c was likely due to the relatively small percentage of RNAs containing a stretch of adenosines adjacent to the anchor oligonucleotide. In fact, only one clone out of 48 clones obtained with oligonucleotide a contained a nonencoded sequence: clone 2918 from 23S rRNA (Fig. (Fig.3A),3A), possessing an 8-nt heteropolymeric tail. Furthermore, a screen of 500 individual rnpB clones, recovered using oligonucleotide a, yielded no clones that could be amplified using oligonucleotide b or c, suggesting that none had polyadenylated ends. These results indicated the requirement for a selective step in order to clone modified ends. Thus, in subsequent experiments, cDNA clones for sequencing were obtained using either oligonucleotide b or c.

In the case of rnpB RNA, for which many more clones were isolated, a counterselection strategy was used to isolate clones that did not contain the native 3′ end (Fig. (Fig.2C).2C). For this, the RT-PCR products were digested with EcoRI, which cleaves at a site around nt 260. Amplicons that contained sequences including the EcoRI site were thus selected against in the subsequent cloning.

Polyadenylated and poly(A)-rich ends of truncated transcripts.

A total of 152 clones containing added 3′ tails were sequenced (Table (Table1).1). Interestingly, many clones analyzed corresponded to incomplete transcripts. For 23S rRNA, most of the 3′ ends were located more than 10 nt upstream of the expected transcription termination site (Fig. (Fig.3A).3A). The tRNACys-Leu clones were closer to the 3′ end, but only one of the clones included the transcription terminator (clone 259 [Fig. [Fig.3B]).3B]). None of the cloned rnpB, rpsD, and cry1Aa cDNAs with 3′ tails included the transcription terminator (Fig. (Fig.44 and and5).5). We think it likely that our collection of 3′-end clones mainly represents RNA molecules undergoing decay, although these could also be the result of stalled or prematurely terminated transcription.

FIG. 4.
Sequences from cloned 3′ ends of rnpB RNA. The figure is set up in the same format as Fig. Fig.3.3. The asterisk indicates that clones were also recovered using the EcoRI strategy (Fig. (Fig.1C).1C). One version of the sequence ...
Summary of sequenced clones with 3′ tails

Two types of tails were observed: those containing only A residues and those containing predominantly A residues but also C, G, and U residues. Poly(A) ends ranged in size from 1 nt to a maximum of 29 nt, with a mean of 5 nt. Heteropolymeric ends ranged in size from 8 to 113 nt, and the adenosine residues were most often clustered, with other residues interspersed. The length of clustered adenosines in the heteropolymeric tails varied, with 8- and 9-nt-long stretches occurring most frequently, but longer strings of poly(A) were also observed. With one notable exception (rpsD clone 235 [Fig. [Fig.5A]),5A]), we did not observe stretches of contiguous non-A nucleotides longer than three residues, suggesting a high preference for A during polymerization. There did not seem to be any specificity to the sites at which tails were added, although in a number of cases, triplets of A, U, or C were observed within 8 nt of the addition site.

For mRNAs, five of the seven different clones analyzed for rpsD were heteropolymeric. On the contrary, for cry1Aa mRNA, only one 3′ addition (at position 3407) was found to be heteropolymeric, while 3′ additions at seven other sites contained only poly(A) (Fig. (Fig.55).

The composition of heteropolymeric tails was close to 90% adenosine, with the order of preference A > U > C > G (Table (Table2).2). The percent composition determined for B. subtilis heteropolymeric tails was quite similar to what was observed in E. coli (24) (Table (Table22).

Polyribonucleotidic tail composition

Nature of 3′-end addition in wt versus PNPase mutant strains.

Both polyadenylated and heteropolymeric ends were obtained in wt and pnpA backgrounds. However, heteropolymeric tails were found in the pnpA strain at only one site on rnpB RNA and one site on rpsD mRNA. No heteropolymeric tails were recovered in the pnpA strain for 23S rRNA, tRNACys-Leu, or cry1Aa mRNA (Fig. (Fig.33 to to5;5; see Discussion).

With respect to the locations of cloned ends, surprisingly, we did not observe a clear pattern distinguishing tails derived from wt and pnpA strains. One might have expected that the absence of PNPase, the major 3′-to-5′ exoribonucleolytic activity, would result in more abundant 3′-proximal sequences. Nevertheless, the 3′ ends of many clones occurred at identical positions or a few bases away in transcripts from wt and pnpA strains. For example, 23S rRNA clones were obtained twice at position 2924 in the wt strain and six times at position 2926 in the pnpA strain (Fig. (Fig.3A).3A). Similarly, clones with tails added at position 217 for tRNACys-Leu (Fig. (Fig.3B),3B), at positions 256, 374, 393, and 398 for rnpB RNA (Fig. (Fig.4),4), and at position 3514 for cry1Aa mRNA (Fig. (Fig.5B)5B) were isolated in both strains. Overall, there was a broad distribution of sites of 3′-tail addition in the wt and pnpA strains.

Sequences of multiple rnpB clones.

Several RNAs with complex heteropolymeric tails were obtained from the rnpB clones, including the clones with 3′ additions at position 256 isolated in both wt and pnpA strains (Fig. (Fig.4).4). To explore this further, additional rnpB clones were sequenced. Many of the rnpB clones obtained had 3′ additions at similar locations, but these differed in sequence. For example, five clones from the wt strain with additions at position 237 had different numbers of adenosine residues or had a heteropolymeric end (Fig. (Fig.4).4). One particular endpoint, at position 256, was found repeatedly, with 19 clones recovered from each strain. These clones were recovered using both selective oligonucleotide c and an enrichment strategy that included EcoRI digestion after amplification with oligonucleotide a (Fig. (Fig.2C;2C; see Materials and Methods). Clones with tails consisting of nine adenosine residues at position 256 were recovered (only in the pnpA strain), but most of these clones had heteropolymeric ends (Table (Table3).3). The pattern consisted of a heteropolymeric octamer, heptamer, and pentamer, each followed by poly(A) stretches between 7 and 12 nt, with slight variations in some clones. Heteropolymeric tails at other positions in rnpB isolated from the wt strain had different sequences. The similarity of the heteropolymeric tails at position 256 and the fact that heteropolymeric tails had different sequences at other positions, suggested that the tail composition depended on the site of addition. Nevertheless, even at the same site of addition, poly(A) and heteropolymeric ends were observed (23S rRNA position 2924 and rnpB position 374).

Composition of tails at rnpB position 256


This study represents the first large-scale analysis of the nature of posttranscriptionally added 3′ ends in B. subtilis, including data from 152 sequenced 3′ tails (Table (Table1).1). A major observation was that poly(A) and heteropolymeric tails were found in both wt and pnpA strains. The poly(A) sizing assay (Fig. (Fig.1)1) made it clear that there was little or no difference in the global pattern of 3′-end tail additions between the wt strain and strains deleted for the genes encoding PNPase or nucleotidyltransferase. Of the 17 different clones with heteropolymeric tails found in this study (not including the unique case of rnpB nt 256 clones), only one example was found in the pnpA strain (rpsD clone 235; Fig. Fig.4).4). This might be interpreted as indicating a lower level of heteropolymeric tail addition in the pnpA strain and perhaps implicating PNPase in this mode of addition. However, the multiple isolations, from independent experiments, of heteropolymeric tails at rnpB position 256 (Fig. (Fig.4;4; Table Table2)2) showed that addition of such tails was not PNPase dependent. Furthermore, using a different strategy to clone 16S rRNA 3′ ends, we found heteropolymeric tails in four out of six clones from the wt strain and seven out of eight clones from the pnpA strain (P. Bralley and G. H. Jones, unpublished results). This differs greatly from the observation in E. coli, where the analysis of tails in a wt strain revealed homogeneous poly(A) sequences with only an occasional nucleotide other than A (7, 18, 23). Only in a pcnB strain (lacking PAP I) were appreciable levels of heteropolymeric tails observed (24).

For B. subtilis clones containing poly(A) tails (in wt and pnpA strains), the average length was about 5 nt when including clones isolated from all procedures or 7 nt when including only clones isolated using oligonucleotide c. This is considerably shorter than what has been reported by others for E. coli strains that have PAP I, where the average length of the poly(A) tails was closer to 20 nt (6, 16, 23-25). This difference may reflect distinct properties of the E. coli and B. subtilis PAPs. On the other hand, variations in tail lengths between these studies may reflect the choice of target RNAs and the nature of oligonucleotides used in the amplification protocols. In this regard, we note that the longest poly(A) tail we obtained was 29 residues (tRNACys-Leu clone 259), whereas the sizing assays (Fig. (Fig.1)1) suggest that longer poly(A) tails should be present.

It should be stressed that the sequencing results obtained in this study cannot be used to draw conclusions about the frequency or location of particular types of tails in vivo. Clones were chosen for sequencing on the basis of the presence of inserted fragments, and for some genes, clones with different-sized inserts were chosen rather than multiple clones with similar-sized inserts. Furthermore, the use of RNA ligase to attach an oligonucleotide to the 3′ ends of total RNA may have introduced an inherent bias for particular RNA sequences or structures as sites of ligation (23). Thus, although it is tempting to speculate that frequently observed sites of 3′-tail attachment represent endonuclease cleavage sites or pause sites in 3′-to-5′ exonucleolytic decay, it is difficult to draw conclusions from this study about the relationship between mRNA decay and 3′-tail addition, other than the fact that 3′ tails are associated with decay intermediates. Alternatively, as mentioned above, it is possible that the observed 3′ ends are not the result of mRNA decay at all but rather derive from transcriptional stalling or premature termination.

In studies on processing of E. coli 23S rRNA, 3′ tails were found to be attached to the mature transcript in most cases (22, 24). In our experiments, we found that the locations of the 3′ tails varied considerably from at or near the mature 3′ end to sites several hundred nucleotides upstream, and this was true in both wt and pnpA strains (Fig. (Fig.3).3). A larger data set of 23S rRNA clones will be needed to determine whether tail addition occurs more frequently on partially degraded molecules. If this were the case, it would suggest that the addition of 3′ tails occurs primarily as a part of the decay process and is secondary to initiation of decay.

One might have expected that the absence of PNPase, the major 3′-to-5′ exoribonuclease in B. subtilis (11), would lead to a higher concentration of RNA ends located closer to the transcription termination site, but this was not the case (Fig. (Fig.33 to to5).5). In recent studies on exoribonuclease mutant strains of B. subtilis, we found an accumulation of short mRNA decay fragments that did not contain 3′-proximal sequences (27). This and other observations led to the conclusion that decay is initiated by an endoribonuclease cleavage and not by attack from the 3′ end. As such, the abundance of RNA decay fragments with 3′ ends close to the transcription termination site would be expected to be similar in wt and pnpA strains.

A very interesting finding was the recovery of multiple examples of heteropolymeric tails attached to rnpB RNA, with only slight differences in sequence (Table (Table3).3). If these sequences are the result of a template-independent polymerase activity, the similar length and sequence of these additions are truly striking. There was a possibility that these sequences resulted from the ligation of other RNAs to the ends of rnpB transcripts. However, sequences similar to those shown in Table Table33 were not identified in a BLAST search of the B. subtilis genome. One could propose that these 3′ tails are the result of a template-dependent RNA polymerase activity, as has been proposed in the case of plastid ndhD mRNA (37). However, the structure and sequence peculiarities that lead to the addition of specific 3′-tail sequences in the plastid ndhD case are not present in B. subtilis rnpB RNA.

Whether a single enzyme with a preference for adenosine carries out both the addition of poly(A) and heteropolymeric ends or whether there are specific enzymes generating each type of tail remains to be discovered. Evidence for two PAP activities in B. subtilis comes from the work of Sarkar et al., who reported the existence in B. subtilis of poly(A) polymerase activity in a pnpA::mini-Tn10 background that resolved as two peaks on a Sephacryl column (31). In our study, the disparate nature of poly(A) and heteropolymeric tails, in terms of length and nucleotide composition, implies that more than one activity capable of adding 3′ tails may be found in B. subtilis.

Taken together, our data suggest strongly that an as-yet unidentified RNA polymerase, rather than PNPase, is responsible for 3′-end addition in B. subtilis. Although an RNase PH-encoding gene is present in B. subtilis and theoretically could be capable of adding nucleotides under low-phosphate conditions (26) on the basis of the precedent in E. coli (24), it is unlikely that this is the poly(A) polymerase of B. subtilis. It is noteworthy that the poly(A) tails associated with RNAs from mutants of S. coelicolor lacking RNase PH are indistinguishable from those associated with RNAs from wild-type strains (Bralley and Jones, unpublished). Fractionation of B. subtilis cell extracts and assay of poly(A) addition in vitro are currently under way in an effort to identify the B. subtilis PAP.


This work was supported by CONACYT grant 36950 to G.O.-A. and a fellowship to J.C.-G. Work in the laboratory of G.H.J. was supported by MCB-0133520 from the National Science Foundation and in the laboratory of D.H.B. by Public Health Service grant GM-48804 from the National Institutes of Health.

We thank Irina Oussenko for construction of the cca-disrupted strains and Jackeline L. Arvizu-Gómez for help in the screening for clones with tails.


1. Blum, E., A. J. Carpousis, and C. F. Higgins. 1999. Polyadenylation promotes degradation of 3′-structured RNA by the Escherichia coli mRNA degradosome in vitro. J. Biol. Chem. 274:4009-4016. [PubMed]
2. Bralley, P., and G. H. Jones. 2002. cDNA cloning confirms the polyadenylation of RNA decay intermediates in Streptomyces coelicolor. Microbiology 148:1421-1425. [PubMed]
3. Bralley, P., and G. H. Jones. 2003. Overexpression of the polynucleotide phosphorylase gene (pnp) of Streptomyces antibioticus affects mRNA stability and poly(A) tail length but not ppGpp levels. Microbiology 149:2173-2182. [PubMed]
4. Bralley, P., and G. H. Jones. 2001. Poly(A) polymerase activity and RNA polyadenylation in Streptomyces coelicolor A3(2). Mol. Microbiol. 40:1155-1164. [PubMed]
5. Cao, G. J., and N. Sarkar. 1993. Poly(A) RNA in Bacillus subtilis: identification of the polyadenylylation site of flagellin mRNA. FEMS Microbiol. Lett. 108:281-285. [PubMed]
6. Cao, G. J., and N. Sarkar. 1992. Poly(A) RNA in Escherichia coli: nucleotide sequence at the junction of the lpp transcript and the polyadenylate moiety. Proc. Natl. Acad. Sci. USA 89:7546-7550. [PubMed]
7. Cao, G. J., and N. Sarkar. 1997. Stationary phase-specific mRNAs in Escherichia coli are polyadenylated. Biochem. Biophys. Res. Commun. 239:46-50. [PubMed]
8. Carpousis, A. J., A. Leroy, N. Vanzo, and V. Khemici. 2001. Escherichia coli RNA degradosome. Methods Enzymol. 342:333-345. [PubMed]
9. Coburn, G. A., and G. A. Mackie. 1999. Degradation of mRNA in Escherichia coli: an old problem with some new twists. Prog. Nucleic Acid Res. Mol. Biol. 62:55-108. [PubMed]
10. Condon, C. 2003. RNA processing and degradation in Bacillus subtilis. Microbiol. Mol. Biol. Rev. 67:157-174. [PMC free article] [PubMed]
11. Deutscher, M. P., and N. B. Reuven. 1991. Enzymatic basis for hydrolytic versus phosphorolytic mRNA degradation in Escherichia coli and Bacillus subtilis. Proc. Natl. Acad. Sci. USA 88:3277-3280. [PubMed]
12. Elbashir, S. M., W. Lendeckel, and T. Tuschl. 2001. RNA interference is mediated by 21- and 22-nucleotide RNAs. Genes Dev. 15:188-200. [PubMed]
13. Goodrich, A. F., and D. A. Steege. 1999. Roles of polyadenylation and nucleolytic cleavage in the filamentous phage mRNA processing and decay pathways in Escherichia coli. RNA 5:972-985. [PubMed]
14. Gopalakrishna, Y., and N. Sarkar. 1982. The synthesis of DNA complementary to polyadenylate-containing RNA from Bacillus subtilis. J. Biol. Chem. 257:2747-2750. [PubMed]
15. Grunberg-Manago, M. 1999. Messenger RNA stability and its role in control of gene expression in bacteria and phages. Annu. Rev. Genet. 33:193-227. [PubMed]
16. Hajnsdorf, E., F. Braun, J. Haugel-Nielsen, and P. Régnier. 1995. Polyadenylylation destabilizes the rpsO mRNA of Escherichia coli. Proc. Natl. Acad. Sci. USA 92:3973-3977. [PubMed]
17. Ireton, K., D. Z. Rudner, K. J. Siranosian, and A. D. Grossman. 1993. Integration of multiple developmental signals in Bacillus subtilis through the Spo0A transcription factor. Genes Dev. 7:283-294. [PubMed]
18. Johnson, M. D., J. Popowski, G. J. Cao, P. Shen, and N. Sarkar. 1998. Bacteriophage T7 mRNA is polyadenylated. Mol. Microbiol. 27:23-30. [PubMed]
19. Karnik, P., Y. Gopalakrishna, and N. Sarkar. 1986. Construction of a cDNA library from polyadenylated RNA of Bacillus subtilis and the determination of some 3′-terminal sequences. Gene 49:161-165. [PubMed]
20. Khemici, V., and A. J. Carpousis. 2004. The RNA degradosome and poly(A) polymerase of Escherichia coli are required in vivo for the degradation of small mRNA decay intermediates containing REP-stabilizers. Mol. Microbiol. 51:777-790. [PubMed]
21. Köhrer, K., and H. Domdey. 1991. Preparation of high molecular weight RNA. Methods Enzymol. 194:398-405. [PubMed]
22. Li, Z., S. Pandit, and M. P. Deutscher. 1998. Polyadenylation of stable RNA precursors in vivo. Proc. Natl. Acad. Sci. USA 95:12158-12162. [PubMed]
23. Mohanty, B. K., and S. R. Kushner. 1999. Analysis of the function of Escherichia coli poly(A) polymerase I in RNA metabolism. Mol. Microbiol. 34:1094-1108. [PubMed]
24. Mohanty, B. K., and S. R. Kushner. 2000. Polynucleotide phosphorylase functions both as a 3′ right-arrow 5′ exonuclease and a poly(A) polymerase in Escherichia coli. Proc. Natl. Acad. Sci. USA 97:11966-11971. [PubMed]
25. Mohanty, B. K., V. F. Maples, and S. R. Kushner. 2004. The Sm-like protein Hfq regulates polyadenylation dependent mRNA decay in Escherichia coli. Mol. Microbiol. 54:905-920. [PubMed]
26. Ost, K. A., and M. P. Deutscher. 1990. RNase PH catalyzes a synthetic reaction, the addition of nucleotides to the 3′ end of RNA. Biochimie 72:813-818. [PubMed]
27. Oussenko, I. A., T. Abe, H. Ujiie, A. Muto, and D. H. Bechhofer. 2005. Participation of 3′-to-5′ exoribonucleases in the turnover of Bacillus subtilis mRNA. J. Bacteriol. 187:2758-2767. [PMC free article] [PubMed]
28. Raynal, L. C., and A. J. Carpousis. 1999. Poly(A) polymerase I of Escherichia coli: characterization of the catalytic domain, an RNA binding site and regions for the interaction with proteins involved in mRNA degradation. Mol. Microbiol. 32:765-775. [PubMed]
29. Raynal, L. C., H. M. Krisch, and A. J. Carpousis. 1998. The Bacillus subtilis nucleotidyltransferase is a tRNA CCA-adding enzyme. J. Bacteriol. 180:6276-6282. [PMC free article] [PubMed]
30. Rott, R., G. Zipor, V. Portnoy, V. Liveanu, and G. Schuster. 2003. RNA polyadenylation and degradation in cyanobacteria are similar to the chloroplast but different from Escherichia coli. J. Biol. Chem. 278:15771-15777. [PubMed]
31. Sarkar, B., G. J. Cao, and N. Sarkar. 1997. Identification of two poly(A) polymerases in Bacillus subtilis. Biochem. Mol. Biol. Int. 41:1045-1050. [PubMed]
32. Sohlberg, B., J. Huang, and S. N. Cohen. 2003. The Streptomyces coelicolor polynucleotide phosphorylase homologue, and not the putative poly(A) polymerase, can polyadenylate RNA. J. Bacteriol. 185:7273-7278. [PMC free article] [PubMed]
33. Vazquez-Cruz, C., and G. Olmedo-Alvarez. 1997. Mechanism of decay of the cry1Aa mRNA in Bacillus subtilis. J. Bacteriol. 179:6341-6348. [PMC free article] [PubMed]
34. Wang, W., and D. H. Bechhofer. 1996. Properties of a Bacillus subtilis polynucleotide phosphorylase deletion strain. J. Bacteriol. 178:2375-2382. [PMC free article] [PubMed]
35. Xu, F., S. Lin-Chao, and S. N. Cohen. 1993. The Escherichia coli pcnB gene promotes adenylylation of antisense RNAI of ColE1-type plasmids in vivo and degradation of RNAI decay intermediates. Proc. Natl. Acad. Sci. USA 90:6756-6760. [PubMed]
36. Youngman, P., J. B. Perkins, and R. Losick. 1984. Construction of a cloning site near one end of Tn917 into which foreign DNA may be inserted without affecting transposition in Bacillus subtilis or expression of the transposon-borne erm gene. Plasmid 12:1-9. [PubMed]
37. Zandueta-Criado, A., and R. Bock. 2004. Surprising features of plastid ndhD transcripts: addition of nonencoded nucleotides and polysome association of mRNAs with an unedited start codon. Nucleic Acids Res. 32:542-550. [PMC free article] [PubMed]

Articles from Journal of Bacteriology are provided here courtesy of American Society for Microbiology (ASM)