Implication of membrane localization of target mRNA in the action of a small RNA: mechanism of post-transcriptional regulation of glucose tr
http://www.100md.com
基因进展 2005年第3期
Division of Biological Science, Graduate School of Science, Nagoya University, Chikusa, Nagoya 464-8602, Japan
Abstract
Accumulation of phosphosugars such as glucose-6-phosphate causes a rapid degradation of ptsG mRNA encoding the major glucose transporter IICBGlc in an RNase E/degradosome-dependent manner. The destabilization of ptsG mRNA is caused by a small antisense RNA (SgrS) that is induced by phosphosugar stress. In this study, we analyzed a series of ptsG-crp translational fusions to identify the mRNA region required for the rapid degradation of ptsG mRNA. We found that the ptsG-crp mRNA is destabilized in response to phosphosugar stress when it contains the 5' portion of ptsG mRNA corresponding up to the first two transmembrane domains (TM1 and TM2) of IICBGlc. The destabilization of ptsG-crp mRNA was largely eliminated by frameshift mutations in the transmembrane region. The IICBGlc-CRP fusion proteins containing more than two transmembrane domains were localized at the membrane. The efficient destabilization of ptsG-crp mRNA was restored when TM1 and TM2 of IICBGlc were replaced by part of the LacY transmembrane region. We conclude that the membrane-targeting property of IICBGlc protein rather than the particular nucleotide or amino acid sequence is required for the efficient degradation of ptsG mRNA in response to metabolic stress. The stimulation of ptsG-crp mRNA degradation was completely eliminated when either the hfq or sgrS gene is inactivated. The efficient mRNA destabilization was observed in the absence of membrane localization when translation was reduced by introducing a mutation in the ribosome-binding site in the cytoplasmic ptsG-crp mRNA. Taken together, we conclude that mRNA localization to the inner membrane coupled with the membrane insertion of nascent peptide mediates the Hfq/SgrS-dependent ptsG mRNA destabilization presumably by reducing second rounds of translation.
[Keywords: mRNA localization; mRNA degradation; protein targeting; glucose transporter; small RNA; Hfq]
Received October 8, 2004; revised version accepted December 8, 2004.
External glucose is transported and phosphorylated by the phosphoenolpyruvate (PEP): sugar phosphotransferase system (PTS) in bacteria (Meadow et al. 1990; Postma et al. 1996; Tchieu et al. 2001). The PTS consists of two common cytoplasmic proteins, enzyme I and HPr, and a series of sugar-specific enzyme II complexes (EIIs). The glucose-specific EII (glucose transporter) of Escherichia coli consists of a cytoplasmic component IIAGlc and a membrane component IICBGlc. The phosphoryl group of PEP is sequentially transferred to enzyme I, to HPr, to IIAGlc, to IICBGlc, and finally to glucose as it is translocated across the membrane. The resulting glucose 6-phosphate (G6P) is metabolized by the Embden-Meyerhof glycolytic pathway and by the pentose-phosphate pathway to produce numerous intermediary metabolites and energy (Fraenkel 1996).
Recent studies have demonstrated that the ptsG gene encoding IICBGlc is regulated in quite an intriguing manner at both transcriptional and post-transcriptional levels depending on physiological conditions (Plumbridge 2002). The ptsG transcription is regulated simultaneously by two global transcription factors, CRP (cAMP receptor protein)-cAMP and Mlc, in response to glucose availability (Kimata et al. 1997, 1998). Glucose uptake induces ptsG transcription by modulating the Mlc-mediated negative regulatory pathway (Kimata et al. 1998; Plumbridge 1998). The glucose transporter itself modulates the localization and activity of Mlc. Dephosphorylated IICBGlc generated by glucose uptake binds Mlc, resulting in sequestration of this global transcription factor to the membrane (Lee et al. 2000; Tanaka et al. 2000; Nam et al. 2001). The membrane localization itself but not binding to IICBGlc is directly responsible for the inactivation of Mlc (Seitz et al. 2003; Tanaka et al. 2004).
In addition to the transcriptional control, the expression of ptsG is regulated at the post-transcriptional level in response to glycolytic flux. We found previously that the ptsG mRNA is dramatically destabilized when the glycolytic pathway is blocked either by mutations at its early stages or by treatment with a nonmetabolizable glucose analog (Kimata et al. 2001). The physiological relevance of this novel post-transcriptional regulation of ptsG expression is rather obvious because it prevents glucose uptake to avoid too much accumulation of hexose phosphates that are potentially toxic to cells (El-Kazzaz et al. 2004). On the other hand, the mechanism by which the block of the glycolytic pathway leads to the destabilization of ptsG mRNA is apparently complicated and intriguing.
Several important aspects have been uncovered concerning the mechanism and/or pathway of the destabilization of ptsG mRNA in response to the metabolic stress. First, the major endoribonuclease RNase E is primarily responsible for the destabilization of ptsG mRNA under the stress condition (Kimata et al. 2001). Second, the accumulation of G6P, fructose 6-phosphate, or -methylglucoside 6-phosphate somehow triggers the RNase E-mediated-destabilization of ptsG mRNA (Morita et al. 2003). Third, the destabilization of ptsG mRNA in response to stress induced by the intracellular accumulation of phosphosugars such as G6P is dependent on enolase, a major component of the RNase E-based degradosome (Morita et al. 2004). Fourth, the destabilization of ptsG mRNA is dependent on an RNA chaperone Hfq (Morita et al. 2004). Finally, an Hfq-binding small RNA, SgrS, mediates the destabilization of ptsG mRNA (Vanderpool and Gottesman 2004). It has been shown that SgrS is induced in response to phosphosugar accumulation resulting in the degradation of ptsG mRNA presumably through SgrS-ptsG pairing (Vanderpool and Gottesman 2004).
In spite of this progress, several important questions remain unanswered regarding the mechanism by which the ptsG mRNA is destabilized in response to phosphosugar stress. In particular, little is known about the features of ptsG mRNA that are responsible for the destabilization of mRNA. In this paper, we addressed the question of what features of ptsG mRNA are required for the rapid degradation of ptsG mRNA. We found that the region from positions +70 to +319 relative to the ptsG transcriptional start site is sufficient for the message destabilization. This region includes the 3' one-third of the untranslated leader region corresponding to the predicted target for SgrS and the first 70 codons of ptsG. Mutational studies demonstrated that the membrane-targeting property of the translated protein is required for the rapid degradation of ptsG mRNA in response to phosphosugar stress. The data presented in this paper suggest that mRNA localization to the inner membrane coupled with the membrane insertion of nascent peptide allowed Hfq/SgrS to act efficiently on the ptsG mRNA by occluding subsequent rounds of translation leading to the RNase E-dependent message degradation.
Results
The plasmid-derived ptsG mRNA is destabilized in an RNase E-dependent manner by phosphosugar stress
We used the cloned ptsG gene to analyze the features of ptsG mRNA responsible for rapid degradation in response to phosphosugar stress. We first examined the expression of ptsG in wild-type cells harboring pIT499 carrying the intact ptsG gene (Fig. 1). Total RNAs from cells with and without pIT499 grown in LB medium containing glucose were analyzed by Northern blotting. Semiquantitative analyses revealed that the level of ptsG mRNA in wild-type cells harboring pIT499 was more than 10-fold higher than that in cells without the plasmid (Fig. 2A, lanes 1-4). Then, we examined the expression of ptsG from pIT499 in pgi cells lacking phosphoglucose isomerase. The ptsG mRNA was destabilized, resulting in a dramatic decrease in the level of the full-length mRNA (Fig. 2A, lane 5). The reduction of ptsG mRNA in pgi cells harboring pIT499 was abolished at 42°C but not at 30°C when the rne allele (ams-1) encoding a temperature-sensitive RNase E was introduced (Fig. 2B). These results imply that the ptsG mRNA generated from pIT499 was destabilized in an RNase E-dependent manner in response to G6P accumulation. The rapid degradation of ptsG mRNA was also observed when the wild-type cells harboring pIT499 were grown in LB medium containing -methylglucoside (MG) (Fig. 2C) as observed previously in cells without the plasmid (Kimata et al. 2001).
Figure 1. (A) The ptsG gene and its variants. The map of the ptsG region is shown at the top along with relevant restriction sites. The open box represents the coding region of IICBGlc. The thick lines below the map indicate the cloned regions on plasmids. The arrows represent the start and direction of transcription from two promoters (PptsG and Pbla). Plasmid pIT499 carries the 3.2-kb fragment containing the intact ptsG gene with the native promoter (PptsG). Plasmid pTH121 carries the Pbla-ptsG2 gene in which the ptsG gene with its entire leader region is under the control of Pbla, while pTH111 carries the Pbla-ptsG1 gene in which the ptsG gene lacks the 5' portion up to +56 of the leader region. (B) Nucleotide sequence around the ptsG leader region. The nucleotide corresponding to the transcription start site (+1) of pptsG is encircled. Numbering is relative to the transcription start site. Arrows show the 5'-ends of Pbla-ptsG2 and Pbla-ptsG1. The Shine-Dalgarno sequence (SD) and the translation initiation codon are boxed. The amino acid sequence of the N terminus of IICBGlc is shown below the nucleotide sequence. The nucleotides complementary to SgrS RNA proposed by Vanderpool and Gottesman (2004) are shown by asterisks. The vertical arrow denotes the nucleotide substitution in the ptsG-crp fusion gene (see Fig. 6).
Figure 2. Expression and response to phosphosugar stress of plasmid-borne ptsG. (A) IT1568 (wild-type) (lane 1), AS16 (ptsG) harboring the ptsG plasmid pIT499 (lanes 2-4), and AS18 (ptsG pgi) harboring pIT499 (lane 5) were grown to A600 = 0.5. Glucose was added at a final concentration of 1% and incubation was continued for 20 min. Total RNAs were prepared, and indicated amounts of RNA samples were subjected to Northern blot analysis using the ptsG probe. (B) AS20 (ptsG pgi ams-1) harboring pIT499 was grown at 30°C to A600 = 0.5. Glucose was added at a final concentration of 1%, and incubation was continued at the indicated temperature for 20 min. Total RNAs were prepared and 2 μg of RNA samples was subjected to Northern blot analysis using the ptsG probe. (C) AS16 harboring pIT499 was grown to A600 = 0.5. Glucose (Glc) or -methylglucoside (MG) was added at a final concentration of 1%, and incubation was continued for 20 min. Total RNAs were prepared and 2 μg of RNA samples was subjected to Northern blot analysis using the ptsG probe.
The first two-thirds of the ptsG leader is dispensable for the mRNA destabilization
The ptsG mRNA possesses a 5'-untranslated leader region (5'-UTR) of 103 nt preceding the structural gene encoding IICBGlc (Fig. 1B). It is likely that certain structural features of the ptsG promoter and/or leader region would be involved in the destabilization of ptsG mRNA. Recently, we demonstrated that the ptsG mRNA exhibited the normal response to G6P accumulation when the ptsG promoter was replaced with a foreign bla promoter on the chromosome (Morita et al. 2004). This ptsG allele (Pbla-ptsG2) produces an extended form of ptsG mRNA in which several nucleotides derived from the bla region are attached to the 5'-end of the wild-type ptsG mRNA (see Fig. 1). This implies that the ptsG promoter is not required for the destabilization of ptsG mRNA and that the modification of the original 5'-end of ptsG mRNA does not affect the stability control by phosphosugar stress. We confirmed this by showing that the mRNA derived from the cloned Pbla-ptsG2 gene on plasmid pTH121 is normally destabilized in the presence of MG (Fig. 3, lanes 1,2). Previously, we constructed a plasmid pTH111 in which the ptsG gene lacking the first 56 nt of 5'-UTR is under the control of the bla promoter (Takahashi et al. 1998). The bla promoter to drive the ptsG in pTH111 is the same as that used in ptsG2. The ptsG allele in pTH111 was designated as Pbla-ptsG1. By using this construct, we examined whether the 5' portion of the ptsG 5'-UTR is required for the destabilization of ptsG mRNA. The 5'-truncated ptsG mRNA was stably expressed when cells harboring pTH111 were grown in the presence of glucose while it was rapidly degraded when cells were grown in the presence of MG (Fig. 3, lanes 3,4). Further truncation of the 5'-UTR at least up to +70 affected neither the expression of ptsG mRNA nor its response to phosphosugar stress (H. Kawamoto, T. Morita, and H. Aiba, unpubl.). We conclude that the first 70 nt of the ptsG UTR is not required for the stability control of ptsG mRNA by phosphosugar stress.
Figure 3. Expression and response to phosphosugar stress of the variant ptsG mRNAs. (A) AS16 harboring the Pbla-ptsG2 plasmid pTH121 (lanes 1,2) or the Pbla-ptsG1 plasmid pTH111 (lanes 3,4) were grown with 1% Glc (lanes 1,3) or MG (lanes 2,4) to A600 = 0.6. Total proteins were prepared and samples equivalent to 0.0075 A600 units were subjected to Western blot analysis using anti-IIB antibody. (B) Cells were grown to A600 = 0.5. Glc (lanes 1,3) or MG (lanes 2,4) was added at a final concentration of 1%, and incubation was continued for 20 min. Total RNAs were prepared and 3 μg of RNA samples was subjected to Northern blot analysis using the ptsG probe.
The region up to the transmembrane segment II is sufficient for the response to phosphosugar stress
To define further the features of ptsG mRNA responsible for the rapid degradation in response to phosphosugar stress, a series of ptsG-crp translational fusions encoding hybrid proteins with N-terminal regions of IICBGlc and the entire CRP was constructed (Fig. 4). IICBGlc consists of a transmembrane IIC domain of 400 residues and a cytoplasmic IIB domain of 90 residues (Buhr and Erni 1993). The IIC domain contains eight transmembrane segments (TM) along with an N-terminal amphipathic sequence (Buhr and Erni 1993). The fusions were designed to produce IIC-CRP proteins in which the N-terminal 14, 40, 72, 107, and 139 residues of IICBGlc were joined to the CRP ORF. These positions correspond to those just after the N-terminal amphipathic sequence, TM-I, TM-II, TM-III, and TM-IV, respectively. Cells harboring a control plasmid pHA7MK expressing the crp mRNA or each of five different plasmids carrying ptsG-crp fusions were grown in the presence of either glucose or MG. Total RNAs and proteins were prepared and analyzed by Northern and Western blotting, respectively (Fig. 5). When cells were grown in the presence of glucose, all fusion genes produced significant amounts of hybrid proteins and mRNAs (Fig. 5A,B, lanes 3,5,7,9,11). On the other hand, the levels of hybrid proteins and ptsG-crp mRNAs were dramatically reduced when cells harboring pMM12, pMM13, or pMM14 were grown in the presence of MG (Fig. 5A,B, lanes 4,6,8,10,12). The reduction of hybrid proteins and ptsG-crp mRNAs by MG were no longer observed when RNase E was inactivated (data not shown). These results clearly indicate that the ptsG-crp chimeric mRNAs were rapidly degraded in response to phosphosugar stress as in the case of the native ptsG mRNA. The presence of MG weakly reduced the expression of the ptsG-crp mRNAs when cells harboring pMM10 or pMM11, while it did not affect the expression of crp mRNA. Taken together, we conclude that the information responsible for the rapid degradation in response to phosphosugar stress is contained within the region from positions +70 to +349 of the ptsG mRNA. This region consists of the 3' one-third of the leader region (+70 to +103) including the Shine-Dalgarno sequence and the region corresponding to the N-terminal portion of IICBGlc containing the first two transmembrane segments along with the amphipathic sequence (+103 to +349). The nucleotide position +349 corresponds to amino acid position 72.
Figure 4. (A) Membrane topology of IICBGlc. The deduced membrane topology of IICBGlc is taken from Buhr and Erni (1993). The portions corresponding to IIC and IIB domain are labeled. The N-terminal amphipathic helix and eight transmembrane (TM) domains are represented as rectangles. Arrowheads represent the sites (amino acid residue number of IICBGlc) to which the crp gene is fused. (B) Schematic representation of ptsG-crp translational fusion genes. Open and filled rectangles represent ptsG and crp coding sequences, respectively. The numbers in open rectangles represent amino acid positions of IICBGlc. The thick lines represent the part of the ptsG leader region. Arrows represent the direction of transcription driven by Pbla.
Figure 5. Expression and response to phosphosugar stress of ptsG-crp translational fusions. (A) IT1568 harboring pHA7MK (lanes 1,2), pMM10 (lanes 3,4), pMM11 (lanes 5,6), pMM12 (lanes 7,8), pMM13 (lanes 9,10), or pMM14 (lanes 11,12) was grown with 1% Glc or MG to A600 = 0.6. Total proteins were prepared, and indicated amounts of samples were subjected to Western blot analysis using anti-CRP antibody. (B) Cells were grown to A600 = 0.5. Glc or MG was added at a final concentration of 1%, and incubation was continued for 20 min. Total RNAs were prepared and indicated amounts of RNA samples were subjected to Northern blot analysis using the crp probe.
The membrane-targeting property of the transmembrane segments is important for the rapid degradation of ptsG-crp mRNA
It has been proposed that a 32-nt-long region from positions +77 to +107 of ptsG mRNA is the direct target of SgrS RNA because it is partially complementary to that of SgrS at nucleotide positions 157-187 (Vanderpool and Gottesman 2004). Our observation that the ptsG14-crp mRNA containing +70 to +142 region is moderately reduced in response to the phosphosugar stress is consistent with the proposed base-pairing model by Vanderpool and Gottesman (2004). However, it is also clear that this region alone is not sufficient for the efficient destabilization of ptsG mRNA. To examine the role of the transmembrane region in the mRNA destabilization, we introduced one nucleotide insertion at position +145 and one nucleotide deletion at position +349 in the ptsG72-crp fusion gene to generate the ptsG72FS-crp gene (Fig. 6A). These frameshift mutations, which create no intervening nonsense codons, cause only a slight change in the nucleotide sequence but a dramatic change in the amino acid sequence of TM-I and TM-II. Northern blot analysis revealed that the level of mRNA produced from the ptsG72FS-crp gene was only moderately reduced by MG (Fig. 6B, lanes 4,5). This suggests that the amino acid sequence but not nucleotide sequence of the TM-I/TM-II region is required for the efficient destabilization of ptsG mRNA.
Figure 6. Expression and response to phosphosugar stress of ptsG-crp translational fusion variants. (A) Schematic representation of altered ptsG-crp translational fusions. The rectangles in white and black represent the ptsG and crp coding regions, respectively. The rectangles in light and dark gray represent TM I-II of IICBGlc with frameshift mutations and LacY TM I-II, respectively. The frameshift mutations just before and after TM I-II are shown as +A (insertion) and -T (deletion) in the ptsG72FS-crp gene. (B) Northern blot analysis of altered ptsG-crp mRNAs. IT1568 harboring pMM12 (lanes 1,2), pMM12FS (lanes 3,4), or pMM10Y (lanes 5,6) was grown to A600 = 0.5. Glc or MG was added at a final concentration of 1%, and incubation was continued for 20 min. Total RNAs were prepared and 3 μg (lanes 1,2,5,6) or 1 μg (lanes 3,4) of RNA samples was subjected to Northern blot analysis using the crp probe.
The next question is how the amino acid sequence of the TM-I/TM-II region participates in the marked destabilization of ptsG mRNA. One possibility is that a particular amino acid sequence within this region somehow plays a role in the regulation of mRNA stability. Alternatively, the membrane-targeting property of the transmembrane segments may be needed for the destabilization of ptsG mRNA. To examine these possibilities, the region corresponding to TM-I and TM-II in the ptsG72-crp was replaced with that corresponding to two transmembrane segments derived from LacY (LacY TM-I-TM-II) to construct the ptsG14Y-crp allele (Fig. 6A). Northern blot analysis indicated that the mRNA produced from the ptsG14Y-crp allele was stably expressed in the presence of glucose, while it was markedly destabilized in the presence of MG (Fig. 6B, lanes 5,6). This strongly suggests that the membrane-targeting property rather than a particular amino acid sequence itself within TM-I and TM-II of IICBGlc is crucial for the rapid degradation of ptsG mRNA.
Proteins generated from mRNAs sensitive to phosphosugar stress are membrane-associated
To confirm the role of the membrane-targeting property of the transmembrane segments in the rapid degradation of ptsG mRNA, we performed a cell fractionation assay by Western blotting using anti-CRP antibody. As expected, the wild-type CRP was found exclusively in the soluble cytoplasmic fraction (Fig. 7, lanes 1-3). Under the same condition, IIC72-CRP possessing TM-I and TM-II generated from pMM12 appeared predominantly in the membrane fraction (Fig. 7, lanes 10-12), while IIC14-CRP lacking the transmembrane segments generated from pMM10 appeared predominantly in the soluble fraction (Fig. 7, lanes 4-6). IIC40-CRP possessing only TM-I was found in both soluble and membrane fractions (Fig. 7, lanes 7-9). As expected, IIC72FS-CRP generated from pMM12FS appeared almost exclusively in the cytoplasmic fraction (Fig. 7, lanes 13-15). These results again support our claim that the rapid degradation of ptsG mRNA in response to phosphosugar stress is tightly coupled with the membrane-targeting property of translated proteins. In addition, it should be noted that the expression levels of fusion proteins attached to membrane are greatly reduced compared to those of cytoplasmic proteins (e.g., cf. lanes 10 and 13 in Fig. 7). The reduced expression of membrane proteins could be due to either inefficient translation or reduced stability of proteins. However, we observed no significant difference in the protein stability among these proteins (data not shown). This suggests that the translation of membrane proteins is less efficient.
Figure 7. Subcellular localization of proteins. IT1568 harboring pHA7MK (lanes 1-3), pMM10 (lanes 4-6), pMM11 (lanes 7-9), pMM12 (lanes 10-12), or pMM12FS (lanes 13-15) was grown with 1% Glc to A600 = 0.6. Total proteins (T) were prepared and fractionated by centrifugation to obtain supernatant (S) and pellet (P). The fractionated protein samples were subjected to Western blot analysis using anti-CRP antibody. The proteins samples used are 0.005 A600 unit (lanes 1-6), 0.075 A600 unit (lanes 7-12), and 0.01 A600 unit (lanes 13-15).
The destabilization of ptsG mRNA in response to phosphosugar stress is dependent on Hfq and a small RNA (SgrS)
It is highly possible that Hfq is involved in the destabilization of ptsG mRNA in response to phosphosugar stress because this RNA chaperon has been shown to participate in post-transcriptional regulation by affecting stability and/or translation of several mRNAs often with small RNAs (Gottesman 2002, 2004; Storz et al. 2004; Valentin-Hansen et al. 2004). To test this possibility, we examined the effect of the hfq mutation on the expression of ptsG and ptsG-crp mRNAs in response to phosphosugar stress. The rapid degradation of ptsG mRNA in response to phosphosugar stress was abolished in cells lacking Hfq, while the hfq mutation did not affect the expression of ptsG mRNA in the presence of glucose (Fig. 8A,C, lanes 1-4). This implies that the destabilization of ptsG mRNA is dependent on Hfq. Recently, Vanderpool and Gottesman (2004) have discovered that the synthesis of SgrS, an Hfq-binding small RNA, is induced by phosphosugar stress and acts as an antisense RNA to mediate the destabilization of ptsG mRNA. We extended their results by using strains expressing Pbla-ptsG and ptsG-crp fusion genes. SgrS was highly induced in cells harboring ptsG plasmids in the presence of MG where the Pbla-ptsG and ptsG72-crp but not ptsG72-crp mRNAs are destabilized (Fig. 8B,D). The destabilization of Pbla-ptsG and ptsG72-crp mRNAs no longer occurs in sgrS cells (Fig. 8A,C, lanes 5,6). We also observed that the destabilization of ptsG14Y-crp mRNA in the presence of MG was abolished in hfq or sgrS cells (Fig. 8E). The level of SgrS RNA under the condition of phosphosugar stress is dramatically reduced in the hfq cells, suggesting that Hfq is also involved in the stabilization of SgrS RNA (Fig. 8B,D, lane 4). The reduction in the level of the SgrS (RyaA) RNA in an hfq mutant was previously observed (Zhang et al. 2003). The stabilization of RNA by Hfq has been shown for several other Hfq-binding small RNAs (Sledjeski et al. 2001; Moller et al. 2002; Masse et al. 2003; Moll et al. 2003).
Figure 8. Effects of hfq or sgrS mutation on the degradation of ptsG and ptsG-crp mRNAs. (A,B) AS16, AS21 (ptsG hfq::cat), and AS23 (ptsG sgrS) harboring pTH111 were grown to A600 = 0.5. Glc, or MG was added at a final concentration of 1%, and incubation was continued for 20 min. Total RNAs were prepared and 3 μg of RNA samples was subjected to Northern blot analysis using ptsG or sgrS probes. (C,D) IT1568, TM122 (hfq::cat), and TM542 (sgrS) harboring pMM12, and IT1568 harboring pMM12FS were grown to A600 = 0.5. Glc, or MG was added at a final concentration of 1%, and incubation was continued for 20 min. Total RNAs were prepared and 3 μg of RNA samples was subjected to Northern blot analysis using crp or sgrS probes. (E) IT1568, TM122, and TM542 harboring pMM10Y were grown to A600 = 0.5. Glc or MG was added at a final concentration of 1%, and incubation was continued for 20 min. Total RNAs were prepared and 3 μg of RNA samples was subjected to Northern blot analysis using the crp probe.
The SgrS-mediated degradation of ptsG-crp fusion mRNA lacking TM domains is enhanced when its translation is reduced
As shown above, the action of SgrS/Hfq on the target ptsG-crp mRNAs is weak in the absence of membrane localization presumably due to an efficient translation. To investigate a possible link between translation efficiency and SgrS/Hfq action, we introduced a single nucleotide substitution at position +90 in the Shine-Dalgarno sequence of ptsG72FS-crp (Fig. 1B). We examined the expression of the mutant fusion gene (ptsG72FSM-crp) in the presence of either glucose or MG by Western and Northern blotting. As expected, the expression of the fusion protein was dramatically decreased even in the normal condition (Fig. 9A, lane 3). In addition, the mRNA level was also markedly reduced in the presence of glucose presumably due to a reduced stability caused by inefficient translation (Fig. 9B, lane 3). Interestingly, however, the mutation markedly enhanced the degradation of mRNA in the presence of MG (Fig. 9B, lane 4), therefore in the presence of SgrS, resulting in no detectable fusion protein (Fig. 9A, lane 4). Thus, the SgrS-mediated degradation of the cytoplasmic fusion mRNA is apparently stimulated when its translation is slowed down. In other words, SgrS/Hfq retains the ability to act efficiently in the absence of membrane localization depending on the translation status of the target mRNA.
Figure 9. Effect of down-mutations in the ribosome-binding site on the response to phosphosugar stress of the cytoplasmic ptsG-crp mRNA. (A) IT1568 harboring pMM12FS (lanes 1,2) or pMM12FSM (lanes 3,4) was grown with 1% Glc or MG to A600 = 0.6. Total proteins were prepared, and the indicated amounts of samples were subjected to Western blot analysis using anti-CRP antibody. (B) Cells were grown to A600 = 0.5. Glc or MG was added at a final concentration of 1%, and incubation was continued for 20 min. Total RNAs were prepared, and the indicated amounts of RNA samples were subjected to Northern blot analysis using the crp probe.
Discussion
A class of small noncoding RNAs is involved in the regulation of gene expression at post-transcriptional levels by altering translation or stability of target mRNAs in bacteria (Gottesman 2002, 2004; Storz et al. 2004). These RNAs are believed to act through base-pairing with target mRNAs in an Hfq-dependent manner. For example, OxyS RNA inhibits the translation of target mRNAs by occluding the ribosome-binding sites (Altuvia et al. 1998; Zhang et al. 1998). In contrast, the DsrA and RprA RNAs stimulate rpoS translation by preventing formation of a translation-inhibitory hairpin structure in the rpoS 5'-untranslated region (Lease et al. 1998; Majdalani et al. 1998, 2002). Recent studies have shown further that small RNAs could alter the stability of target mRNAs. Typically, RyhB RNA induces a rapid degradation of its target mRNAs in an Hfq- and RNase E-dependent manner (Masse and Gottesman 2002; Masse et al. 2003). SgrS (RyaA) RNA is one of a group of newly identified small noncoding RNAs that bind to Hfq (Zhang et al. 2003). The work by Vanderpool and Gottesman (2004) demonstrated that SgrS mediates the destabilization of ptsG mRNA presumably through base-pairing with ptsG mRNA. They have shown that SgrS is expressed in response to phosphosugar accumulation and that SgrS expression is correlated with greatly reduced levels of ptsG mRNA. They have also found that the expression of sgrS is regulated by a novel transcriptional activator SgrR that may activate sgrS transcription in response to high intracellular levels of phosphosugars.
The major question addressed in the present study is what features of ptsG mRNA are responsible for the rapid decay by RNase E under conditions of phosphosugar accumulation. We first showed that the 5' portion up to +70 in the ptsG mRNA is dispensable for the destabilization of message. Then, we used a series of ptsG-crp translational fusions to specify the downstream boundary in the ptsG mRNA required for the response to phosphosugar stress. We found that the region up to the second transmembrane segment of IICBGlc is sufficient for the efficient message destabilization. The region required for the efficient destabilization of message covers the region complementary to SgrS RNA (see Fig. 1B). Thus, our data are consistent with the SgrS-ptsG mRNA base-pairing model by Vanderpool and Gottesman (2004).
However, our results indicate that the ptsG region complementary to SgrS RNA alone is not sufficient for the strong destabilization of the message because the ptsG-crp fusion mRNAs missing the transmembrane region are only moderately decreased in the stress conditions where SgrS is highly expressed. Our data clearly demonstrate that the ptsG region up to the second transmembrane segment is necessary for the efficient destabilization of ptsG mRNA. One simple explanation for this result is that the mRNA region corresponding to two transmembrane segments somehow are involved in the interaction with SgrS and/or components of RNase E degradosome. Another possibility is that the translated nascent peptide of the transmembrane segments of IICBGlc plays some roles in the SgrS-ptsG interaction and/or the action of RNase E degradosome. However, our experiments clearly exclude these possibilities because the transmembrane segments of IICBGlc can be replaced with those derived from an unrelated protein LacY. Thus, the simplest interpretation of our data is that membrane targeting ability of translated protein is crucial for the destabilization of ptsG mRNA in response to phosphosugar stress.
The sugar-specific permeases (EIIs) of PTS generally contain N-terminal amphipathic sequences of 13-20 residues that are believed to function in membrane insertion possibly cooperatively with the first hydrophobic transmembrane segment (Saier 1989; Saier et al. 1989; Yamada et al. 1991). Recently, it has been shown that mannitol permease (MtlA) is recognized by the bacterial signal recognition particle resulting in cotranslational targeting of the growing MtlA polypeptide to the inner membrane (Beck et al. 2000). IICBGlc has an N-terminal amphipathic sequence of 13 residues preceding eight transmembrane segments (Buhr and Erni 1993). Therefore, it is very likely that IICBGlc is also inserted into membranes in a cotranslational manner. If this is the case, the ptsG mRNA under translation is expected to localize near the inner membrane because translation is also coupled with transcription in bacteria. Taken together, we propose that mRNA localization to the inner membrane coupled with the membrane insertion of nascent peptide is a critical event for the Hfq/SgrS- and RNase E-mediated destabilization of the ptsG mRNA.
There are several possible models to explain the requirement of mRNA localization to the inner membrane for the rapid degradation of ptsG mRNA in response to phosphosugar stress. Firstly, the preferential localization of RNase E-based degradosome near the membrane could explain this. In this regard, there is a report that RNase E seems to be predominantly located near the cytoplasmic membrane (Liou et al. 2001). We showed recently that enolase within the degradosome is required for the destabilization of ptsG mRNA (Morita et al. 2004). Thus, one possible mechanism by which enolase stimulates ptsG mRNA degradation would be to modulate the membrane localization of RNase E. Further experiments are certainly needed to examine whether RNase E is indeed located near the membrane and whether enolase affects subcellular localization of RNase E. Secondly, the requirement of the mRNA localization to the inner membrane for the rapid degradation of message can be explained if SgrS or SgrS/Hfq complex were preferentially localized near the inner membrane by some mechanism. It is unlikely that SgrS or SgrS/Hfq itself is localized near membrane. However, additional factors including enolase could mediate the localization of SgrS/Hfq to the inner membrane. The third possibility would be that the action of either SgrS/Hfq or RNase E is somehow inhibited in the particular ptsG-crp fusion mRNA when it is in the cytoplasm by an unknown factor, while this inhibition is relieved when the mRNA is localized near the membrane. Thus, it is plausible that additional factor(s) other than RNase E, enolase, Hfq, SgrS, and SgrR are also involved in the regulation of stability of ptsG mRNA in response to phosphosugar stress. A simple attractive idea is that the ribosome itself is acting as an additional player for the regulation of stability of ptsG mRNA. Consistent with this idea, we have shown that the reduced translation of the cytoplasmic target mRNA by introducing single base substitutions in the ribosome-binding site allowed SgrS/Hfq to act efficiently in the absence of membrane localization.
Taken together, we would like to propose the following scenario concerning the regulation of translation and stability of ptsG mRNA (Fig. 10). When Mlc is inactivated in the presence of glucose, ptsG transcription is induced. As shown by Vanderpool and Gottesman (2004), and confirmed here, no SgrS RNA is made under nonstress conditions. Translation of ptsG mRNA occurs normally and the newly synthesized nascent polypeptide of IICBGlc would be cotranslationally targeted into the inner membrane. Under stress conditions, SgrS RNA is efficiently transcribed. We hypothesize that new rounds of translation of ptsG mRNA already being translated would be less effective compared to the first round of translation because the membrane localization of message may somehow affect the entry of new ribosomes. The fact that the expression level of IIC72-CRP retaining the membrane-targeting ability is markedly lower compared to that of IIC72FS-CRP, the nonmembrane counterpart of IIC72-CRP, supports this assumption. The inhibition of the ribosome entry may allow SgrS along with Hfq to act on the ptsG mRNA, resulting in SgrS-ptsG pairing, and therefore the RNase E-dependent degradation of the ptsG mRNA. When the target mRNA lacks the membrane-localization ability, the entry of new ribosomes occurs effectively, which, in turn, prevents the action of SgrS/Hfq on the ptsG mRNA. It remains to be studied whether the destabilization of ptsG mRNA is caused directly by SgrS-mRNA base-pairing or is an indirect effect of the inhibition of translation of ptsG by SgrS-mRNA base-pairing.
Figure 10. A model for ptsG mRNA degradation in response to the phosphosugar stress. In nonstress conditions, ptsG mRNA is translated and IICBGlc protein is targeted to cytoplasmic membrane in a cotranslational manner, resulting in localization of mRNA near the membrane. New rounds of ribosome entry into the ptsG mRNA being translated would be inhibited due to the membrane localization of message. Under stress conditions, SgrS RNA is highly transcribed by the action of SgrR. The inhibition of the ribosome entry may allow SgrS/Hfq to act on the ptsG mRNA, resulting in SgrS-ptsG pairing and therefore the RNase E-dependent degradation of the ptsG mRNA. When the target mRNA lacks the membrane-localization ability, the entry of new ribosomes occurs effectively, which, in turn, prevents the action of SgrS/Hfq on the ptsG mRNA.
In conclusion, the present study has shed a light on a unique feature, implication of mRNA localization to membrane, of small RNA-mediated mRNA degradation. It is certainly interesting to know whether this is true for other membrane-associated messages that may be regulated by small RNAs in bacteria. Finally, this novel mechanism could be relevant to the action of small regulatory RNAs in eukaryotes too.
Materials and methods
Media and growth conditions
Cells were grown aerobically at 37°C unless specified otherwise in LB medium. Antibiotics were used at the following concentrations when needed: ampicillin (50 μg/mL); chloramphenicol (15 μg/mL), and tetracycline (15 μg/mL). Bacterial growth was monitored by determining the optical density at 600 nm.
Bacterial strains and plasmids
The E. coli K12 strains and plasmids used in this study are listed in Table 1. IT1568 (W3110 mlc) was used as a parent wild-type strain. Other strains were constructed by one-step gene inactivation protocol (Datsenko and Wanner 2000). When necessary, the cat marker was eliminated by using an FLP expression plasmid pCP20. Plasmid pHA7MK was constructed from pHA7M by introducing a KpnI site just after the initiation codon of CRP. To construct pMM10, pMM12, pMM13, pMM14, and pMM15, DNA regions between a BamHI site upstream of Pbla and various sites corresponding to amino acid positions 14, 40, 72, 103, and 140 of IICBGlc in pTH111 were amplified by PCR and cloned between BamHI and KpnI sites of pHA7MK.
Table 1. Bacterial strains and plasmids used in this study
Northern blotting
Total RNAs were isolated from cells grown to mid-log phase as described (Aiba et al. 1981). For Northern blot analysis, indicated amounts of total RNAs were resolved by 1.2% agarose gel electrophoresis in the presence of formaldehyde and blotted on to Hybond-N+ membrane (Amersham Biosciences). The mRNAs were visualized using digoxigenin (DIG) reagents and kits for nonradioactive nucleic acid labeling and detection system (Roche Molecular Biochemicals) according to the procedure specified by the manufacturer. The following DIG-labeled DNA probes were prepared by PCR using DIG-dUTP: a 305-bp fragment corresponding to the 5' region of ptsG; a 621-bp fragment corresponding to the crp; and a 150-bp fragment corresponding to sgrS.
Western blotting
Bacterial cells grown in LB medium containing appropriate antibiotics(s) or carbon(s) were harvested and suspended in 100 μL of SDS-PAGE loading buffer (62.5 mM Tris-HCl at pH 6.8, 2% SDS, 10% glycerol, 5% -mercaptoethanol, 0.1% bromophenol blue). The sample was heated at 100°C for 5 min. Indicated amounts of protein samples were subjected to a 12% or 15% polyacrylamide-0.1% SDS gel electrophoresis and transferred to Immobilon membrane (Millipore). The membranes were treated with anti-IIB (Tanaka et al. 2000) or anti-CRP (Ishizuka et al. 1993) antibodies. Signals were visualized by the ECL system (Amersham).
Protein localization assay
Cells harboring indicated plasmids were grown in 20 mL of LB medium to A600 of 0.6. A 0.5-mL aliquot of cell cultures was directly sonicated for 5 sec twice at 0°C, with a 10-sec interval in between. The sonicated samples were centrifuged at 15, 000g (4°C) for 3 min. The pellets were dissolved in 30 μL of SDS sample buffer and used as the membrane fraction. The supernatant fraction was precipitated with cold trichloroacetic acid (5% final concentration). The precipitates were collected by centrifugation, washed with 70% acetone, dissolved in 30 μL of SDS sample buffer, and used as the cytoplasmic fraction.
Acknowledgments
We thank Jacqueline Plumbridge, Susan Gottesman, and Carin Vanderpool for critical reading of the manuscript and for helpful comments. This work was supported by Grants-in-Aid from the Ministry of Education, Culture, Sports, Science and Technology of Japan and by Ajinomoto Co., Inc.
References
Aiba, H., Adhya, S., and de Crombrugghe, B. 1981. Evidence for two functional gal promoters in intact Escherichia coli cells. J. Biol. Chem. 256: 11905-11910.
Altuvia, S., Zhang, A., Argaman, L., Tiwari, A., and Storz, G. 1998. The Escherichia coli OxyS regulatory RNA represses fhlA translation by blocking ribosome binding. EMBO J. 17: 6069-6075.
Beck, K., Wu, L.F., Brunner, J., and Muller, M. 2000. Discrimination between SRP- and SecA/SecB-dependent substrates involves selective recognition of nascent chains by SRP and trigger factor. EMBO J. 19: 134-143.
Buhr, A. and Erni, B. 1993. Membrane topology of the glucose transporter of Escherichia coli. J. Biol. Chem. 268: 11599-11603.
Datsenko, K.A. and Wanner, B.L. 2000. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. 97: 6640-6645.
El-Kazzaz, W., Morita, T., Tagami, H., Inada, T., and Aiba, H. 2004. Metabolic block at early stages of the glycolytic path-way activates the Rcs phosphorelay system via increased synthesis of dTDP-glucose in Escherichia coli. Mol. Microbiol. 51: 1117-1128.
Fraenkel, D.G. 1996. Glycolysis. In Escherichia coli and Salmonella typhimurium: Cellular and molecular biology (ed. F.C. Neidhardt), pp. 189-198. ASM Press, Washington, DC.
Gottesman, S. 2002. Stealth regulation: biological circuits with small RNA switches. Genes & Dev. 16: 2829-2842.[Free Full Text]
____. 2004. The small RNA regulators of Escherichia coli: Roles and mechanisms. Annu. Rev. Microbiol. 58: 303-328.
Ishizuka, H., Hanamura, A., Kunimura, T., and Aiba, H. 1993. A lowered concentration of cAMP receptor protein caused by glucose is an important determinant for catabolite repression in Escherichia coli. Mol. Microbiol. 10: 341-350.
Kimata, K., Takahashi, H., Inada, T., Postma, P., and Aiba, H. 1997. cAMP receptor protein-cAMP plays a crucial role in glucose-lactose diauxie by activating the major glucose transporter gene in Escherichia coli. Proc. Natl. Acad. Sci. 94: 12914-12919.
Kimata, K., Inada, T., Tagami, H., and Aiba, H. 1998. A global repressor (Mlc) is involved in glucose induction of the ptsG gene encoding major glucose transporter in Escherichia coli. Mol. Microbiol. 29: 1509-1519.
Kimata, K., Tanaka, Y., Inada, T., and Aiba, H. 2001. Expression of the glucose transporter gene, ptsG, is regulated at the mRNA degradation step in response to glycolytic flux in Escherichia coli. EMBO J. 20: 3587-3595.
Lease, R.A., Cusick, M.E., and Belfort, M. 1998. Riboregulation in Escherichia coli: DsrA RNA acts by RNA:RNA interactions at multiple loci. Proc. Natl. Acad. Sci. 95: 12456-12461.
Lee, S.J., Boos, W., Bouche, J.P., and Plumbridge, J. 2000. Signal transduction between a membrane-bound transporter, PtsG, and a soluble transcription factor, Mlc, of Escherichia coli. EMBO J. 19: 5353-5361.
Liou, G.G., Jane, W.N., Cohen, S.N., Lin, N.S., and Lin-Chao, S. 2001. RNA degradosomes exist in vivo in Escherichia coli as multicomponent complexes associated with the cytoplasmic membrane via the N-terminal region of ribonuclease E. Proc. Natl. Acad. Sci. 98: 63-68.
Majdalani, N., Cunning, C., Sledjeski, D., Elliott, T., and Gottesman, S. 1998. DsrA RNA regulates translation of RpoS message by an anti-antisense mechanism, independent of its action as an antisilencer of transcription. Proc. Natl. Acad. Sci. 95: 12462-12467.
Majdalani, N., Hernandez, D., and Gottesman, S. 2002. Regulation and mode of action of the second small RNA activator of RpoS translation, RprA. Mol. Microbiol. 46: 813-826.
Masse, E. and Gottesman, S. 2002. A small RNA regulates the expression of genes involved in iron metabolism in Escherichia coli. Proc. Natl. Acad. Sci. 99: 4620-4625.
Masse, E., Escorcia, F.E., and Gottesman, S. 2003. Coupled degradation of a small regulatory RNA and its mRNA targets in Escherichia coli. Genes & Dev. 17: 2374-2383.
Meadow, N.D., Fox, D.K., and Roseman, S. 1990. The bacterial phosphoenolpyruvate: Glycose phosphotransferase system. Annu. Rev. Biochem. 59: 497-542.
Moll, I., Afonyushkin, T., Vytvytska, O., Kaberdin, V.R., and Blasi, U. 2003. Coincident Hfq binding and RNase E cleavage sites on mRNA and small regulatory RNAs. RNA 9: 1308-1314.
Moller, T., Franch, T., Hojrup, P., Keene, D.R., Bachinger, H.P., Brennan, R.G., and Valentin-Hansen, P. 2002. Hfq: A bacterial Sm-like protein that mediates RNA-RNA interaction. Mol. Cell 9: 23-30.
Morita, T., El-Kazzaz, W., Tanaka, Y., Inada, T., and Aiba, H. 2003. Accumulation of glucose 6-phosphate or fructose 6-phosphate is responsible for destabilization of glucose transporter mRNA in Escherichia coli. J. Biol. Chem. 278: 15608-15614.
Morita, T., Kawamoto, H., Mizota, T., Inada, T., and Aiba, H. 2004. Enolase in the RNA degradosome plays a crucial role in the rapid decay of glucose transporter mRNA in the response to phosphosugar stress in Escherichia coli. Mol. Microbiol. 54: 1063-1075.
Nam, T.W., Cho, S.H., Shin, D., Kim, J.H., Jeong, J.Y., Lee, J.H., Roe, J.H., Peterkofsky, A., Kang, S.O., Ryu, S., et al. 2001. The Escherichia coli glucose transporter enzyme IICB(Glc) recruits the global repressor Mlc. EMBO J. 20: 491-498.
Plumbridge, J. 1998. Expression of ptsG, the gene for the major glucose PTS transporter in Escherichia coli, is repressed by Mlc and induced by growth on glucose. Mol. Microbiol. 29: 1053-1063.
____. 2002. Regulation of gene expression in the PTS in Escherichia coli: The role and interactions of Mlc. Curr. Opin. Microbiol. 5: 187-193.
Postma, P.W., Lengeler, J.W., and Jacobson, G.R. 1996. Phosphoenolpyruvate: Carbohydrate phosphotransferase systems. In Escherichia coli and Salmonella typhimurium: Cellular and molecular biology. (ed. F.C. Neidhardt), pp. 1149-1174. ASM Press, Washington, DC.
Saier Jr., M.H. 1989. Current studies on the bacterial phosphotransferase system in the Saier laboratory (La Jolla, California, summer, 1988). FEMS Microbiol. Rev. 5: 183-192.
Saier Jr., M.H., Werner, P.K., and Muller, M. 1989. Insertion of proteins into bacterial membranes: Mechanism, characteristics, and comparisons with the eucaryotic process. Microbiol. Rev. 53: 333-366.
Seitz, S., Lee, S.J., Pennetier, C., Boos, W., and Plumbridge, J. 2003. Analysis of the interaction between the global regulator Mlc and EIIBGlc of the glucose-specific phosphotransferase system in Escherichia coli. J. Biol. Chem. 278: 10744-10751.
Sledjeski, D.D., Whitman, C., and Zhang, A. 2001. Hfq is necessary for regulation by the untranslated RNA DsrA. J. Bacteriol. 183: 1997-2005.
Storz, G., Opdyke, J.A., and Zhang, A. 2004. Controlling mRNA stability and translation with small, noncoding RNAs. Curr. Opin. Microbiol. 7: 140-144.
Takahashi, H., Inada, T., Postma, P., and Aiba, H. 1998. CRP down-regulates adenylate cyclase activity by reducing the level of phosphorylated IIA(Glc), the glucose-specific phosphotransferase protein, in Escherichia coli. Mol. Gen. Genet. 259: 317-326.
Tanaka, Y., Kimata, K., and Aiba, H. 2000. A novel regulatory role of glucose transporter of Escherichia coli: Membrane sequestration of a global repressor Mlc. EMBO J. 19: 5344-5352.
Tanaka, Y., Itoh, F., Kimata, K., and Aiba, H. 2004. Membrane localization itself but not binding to IICB is directly responsible for the inactivation of the global repressor Mlc in Escherichia coli. Mol. Microbiol. 53: 941-951.
Tchieu, J.H., Norris, V., Edwards, J.S., and Saier Jr., M.H. 2001. The complete phosphotranferase system in Escherichia coli. J. Mol. Microbiol. Biotechnol. 3: 329-346.
Valentin-Hansen, P., Eriksen, M., and Udesen, C. 2004. The bacterial Sm-like protein Hfq: A key player in RNA transactions. Mol. Microbiol. 51: 1525-1533.
Vanderpool, C.K and Gottesman, S. 2004. Involvement of a novel transcriptional activator and small RNA in the response to phosphosugar stress. Mol. Microbiol. 54: 1076-1089.
Yamada, Y., Chang, Y.Y., Daniels, G.A., Wu, L.F., Tomich, J.M., Yamada, M., and Saier Jr., M.H. 1991. Insertion of the mannitol permease into the membrane of Escherichia coli. Possible involvement of an N-terminal amphiphilic sequence. J. Biol. Chem. 266: 17863-17871.
Zhang, A., Altuvia, S., Tiwari, A., Argaman, L., Hengge-Aronis, R., and Storz, G. 1998. The OxyS regulatory RNA represses rpoS translation and binds the Hfq (HF-I) protein. EMBO J. 17: 6061-6068.
Zhang, A., Wassarman, K.M., Rosenow, C., Tjaden, B.C., Storz, G., and Gottesman, S. 2003. Global analysis of small RNA and mRNA targets of Hfq. Mol. Microbiol. 50: 1111-1124.(Hiroshi Kawamoto, Teppei )
Abstract
Accumulation of phosphosugars such as glucose-6-phosphate causes a rapid degradation of ptsG mRNA encoding the major glucose transporter IICBGlc in an RNase E/degradosome-dependent manner. The destabilization of ptsG mRNA is caused by a small antisense RNA (SgrS) that is induced by phosphosugar stress. In this study, we analyzed a series of ptsG-crp translational fusions to identify the mRNA region required for the rapid degradation of ptsG mRNA. We found that the ptsG-crp mRNA is destabilized in response to phosphosugar stress when it contains the 5' portion of ptsG mRNA corresponding up to the first two transmembrane domains (TM1 and TM2) of IICBGlc. The destabilization of ptsG-crp mRNA was largely eliminated by frameshift mutations in the transmembrane region. The IICBGlc-CRP fusion proteins containing more than two transmembrane domains were localized at the membrane. The efficient destabilization of ptsG-crp mRNA was restored when TM1 and TM2 of IICBGlc were replaced by part of the LacY transmembrane region. We conclude that the membrane-targeting property of IICBGlc protein rather than the particular nucleotide or amino acid sequence is required for the efficient degradation of ptsG mRNA in response to metabolic stress. The stimulation of ptsG-crp mRNA degradation was completely eliminated when either the hfq or sgrS gene is inactivated. The efficient mRNA destabilization was observed in the absence of membrane localization when translation was reduced by introducing a mutation in the ribosome-binding site in the cytoplasmic ptsG-crp mRNA. Taken together, we conclude that mRNA localization to the inner membrane coupled with the membrane insertion of nascent peptide mediates the Hfq/SgrS-dependent ptsG mRNA destabilization presumably by reducing second rounds of translation.
[Keywords: mRNA localization; mRNA degradation; protein targeting; glucose transporter; small RNA; Hfq]
Received October 8, 2004; revised version accepted December 8, 2004.
External glucose is transported and phosphorylated by the phosphoenolpyruvate (PEP): sugar phosphotransferase system (PTS) in bacteria (Meadow et al. 1990; Postma et al. 1996; Tchieu et al. 2001). The PTS consists of two common cytoplasmic proteins, enzyme I and HPr, and a series of sugar-specific enzyme II complexes (EIIs). The glucose-specific EII (glucose transporter) of Escherichia coli consists of a cytoplasmic component IIAGlc and a membrane component IICBGlc. The phosphoryl group of PEP is sequentially transferred to enzyme I, to HPr, to IIAGlc, to IICBGlc, and finally to glucose as it is translocated across the membrane. The resulting glucose 6-phosphate (G6P) is metabolized by the Embden-Meyerhof glycolytic pathway and by the pentose-phosphate pathway to produce numerous intermediary metabolites and energy (Fraenkel 1996).
Recent studies have demonstrated that the ptsG gene encoding IICBGlc is regulated in quite an intriguing manner at both transcriptional and post-transcriptional levels depending on physiological conditions (Plumbridge 2002). The ptsG transcription is regulated simultaneously by two global transcription factors, CRP (cAMP receptor protein)-cAMP and Mlc, in response to glucose availability (Kimata et al. 1997, 1998). Glucose uptake induces ptsG transcription by modulating the Mlc-mediated negative regulatory pathway (Kimata et al. 1998; Plumbridge 1998). The glucose transporter itself modulates the localization and activity of Mlc. Dephosphorylated IICBGlc generated by glucose uptake binds Mlc, resulting in sequestration of this global transcription factor to the membrane (Lee et al. 2000; Tanaka et al. 2000; Nam et al. 2001). The membrane localization itself but not binding to IICBGlc is directly responsible for the inactivation of Mlc (Seitz et al. 2003; Tanaka et al. 2004).
In addition to the transcriptional control, the expression of ptsG is regulated at the post-transcriptional level in response to glycolytic flux. We found previously that the ptsG mRNA is dramatically destabilized when the glycolytic pathway is blocked either by mutations at its early stages or by treatment with a nonmetabolizable glucose analog (Kimata et al. 2001). The physiological relevance of this novel post-transcriptional regulation of ptsG expression is rather obvious because it prevents glucose uptake to avoid too much accumulation of hexose phosphates that are potentially toxic to cells (El-Kazzaz et al. 2004). On the other hand, the mechanism by which the block of the glycolytic pathway leads to the destabilization of ptsG mRNA is apparently complicated and intriguing.
Several important aspects have been uncovered concerning the mechanism and/or pathway of the destabilization of ptsG mRNA in response to the metabolic stress. First, the major endoribonuclease RNase E is primarily responsible for the destabilization of ptsG mRNA under the stress condition (Kimata et al. 2001). Second, the accumulation of G6P, fructose 6-phosphate, or -methylglucoside 6-phosphate somehow triggers the RNase E-mediated-destabilization of ptsG mRNA (Morita et al. 2003). Third, the destabilization of ptsG mRNA in response to stress induced by the intracellular accumulation of phosphosugars such as G6P is dependent on enolase, a major component of the RNase E-based degradosome (Morita et al. 2004). Fourth, the destabilization of ptsG mRNA is dependent on an RNA chaperone Hfq (Morita et al. 2004). Finally, an Hfq-binding small RNA, SgrS, mediates the destabilization of ptsG mRNA (Vanderpool and Gottesman 2004). It has been shown that SgrS is induced in response to phosphosugar accumulation resulting in the degradation of ptsG mRNA presumably through SgrS-ptsG pairing (Vanderpool and Gottesman 2004).
In spite of this progress, several important questions remain unanswered regarding the mechanism by which the ptsG mRNA is destabilized in response to phosphosugar stress. In particular, little is known about the features of ptsG mRNA that are responsible for the destabilization of mRNA. In this paper, we addressed the question of what features of ptsG mRNA are required for the rapid degradation of ptsG mRNA. We found that the region from positions +70 to +319 relative to the ptsG transcriptional start site is sufficient for the message destabilization. This region includes the 3' one-third of the untranslated leader region corresponding to the predicted target for SgrS and the first 70 codons of ptsG. Mutational studies demonstrated that the membrane-targeting property of the translated protein is required for the rapid degradation of ptsG mRNA in response to phosphosugar stress. The data presented in this paper suggest that mRNA localization to the inner membrane coupled with the membrane insertion of nascent peptide allowed Hfq/SgrS to act efficiently on the ptsG mRNA by occluding subsequent rounds of translation leading to the RNase E-dependent message degradation.
Results
The plasmid-derived ptsG mRNA is destabilized in an RNase E-dependent manner by phosphosugar stress
We used the cloned ptsG gene to analyze the features of ptsG mRNA responsible for rapid degradation in response to phosphosugar stress. We first examined the expression of ptsG in wild-type cells harboring pIT499 carrying the intact ptsG gene (Fig. 1). Total RNAs from cells with and without pIT499 grown in LB medium containing glucose were analyzed by Northern blotting. Semiquantitative analyses revealed that the level of ptsG mRNA in wild-type cells harboring pIT499 was more than 10-fold higher than that in cells without the plasmid (Fig. 2A, lanes 1-4). Then, we examined the expression of ptsG from pIT499 in pgi cells lacking phosphoglucose isomerase. The ptsG mRNA was destabilized, resulting in a dramatic decrease in the level of the full-length mRNA (Fig. 2A, lane 5). The reduction of ptsG mRNA in pgi cells harboring pIT499 was abolished at 42°C but not at 30°C when the rne allele (ams-1) encoding a temperature-sensitive RNase E was introduced (Fig. 2B). These results imply that the ptsG mRNA generated from pIT499 was destabilized in an RNase E-dependent manner in response to G6P accumulation. The rapid degradation of ptsG mRNA was also observed when the wild-type cells harboring pIT499 were grown in LB medium containing -methylglucoside (MG) (Fig. 2C) as observed previously in cells without the plasmid (Kimata et al. 2001).
Figure 1. (A) The ptsG gene and its variants. The map of the ptsG region is shown at the top along with relevant restriction sites. The open box represents the coding region of IICBGlc. The thick lines below the map indicate the cloned regions on plasmids. The arrows represent the start and direction of transcription from two promoters (PptsG and Pbla). Plasmid pIT499 carries the 3.2-kb fragment containing the intact ptsG gene with the native promoter (PptsG). Plasmid pTH121 carries the Pbla-ptsG2 gene in which the ptsG gene with its entire leader region is under the control of Pbla, while pTH111 carries the Pbla-ptsG1 gene in which the ptsG gene lacks the 5' portion up to +56 of the leader region. (B) Nucleotide sequence around the ptsG leader region. The nucleotide corresponding to the transcription start site (+1) of pptsG is encircled. Numbering is relative to the transcription start site. Arrows show the 5'-ends of Pbla-ptsG2 and Pbla-ptsG1. The Shine-Dalgarno sequence (SD) and the translation initiation codon are boxed. The amino acid sequence of the N terminus of IICBGlc is shown below the nucleotide sequence. The nucleotides complementary to SgrS RNA proposed by Vanderpool and Gottesman (2004) are shown by asterisks. The vertical arrow denotes the nucleotide substitution in the ptsG-crp fusion gene (see Fig. 6).
Figure 2. Expression and response to phosphosugar stress of plasmid-borne ptsG. (A) IT1568 (wild-type) (lane 1), AS16 (ptsG) harboring the ptsG plasmid pIT499 (lanes 2-4), and AS18 (ptsG pgi) harboring pIT499 (lane 5) were grown to A600 = 0.5. Glucose was added at a final concentration of 1% and incubation was continued for 20 min. Total RNAs were prepared, and indicated amounts of RNA samples were subjected to Northern blot analysis using the ptsG probe. (B) AS20 (ptsG pgi ams-1) harboring pIT499 was grown at 30°C to A600 = 0.5. Glucose was added at a final concentration of 1%, and incubation was continued at the indicated temperature for 20 min. Total RNAs were prepared and 2 μg of RNA samples was subjected to Northern blot analysis using the ptsG probe. (C) AS16 harboring pIT499 was grown to A600 = 0.5. Glucose (Glc) or -methylglucoside (MG) was added at a final concentration of 1%, and incubation was continued for 20 min. Total RNAs were prepared and 2 μg of RNA samples was subjected to Northern blot analysis using the ptsG probe.
The first two-thirds of the ptsG leader is dispensable for the mRNA destabilization
The ptsG mRNA possesses a 5'-untranslated leader region (5'-UTR) of 103 nt preceding the structural gene encoding IICBGlc (Fig. 1B). It is likely that certain structural features of the ptsG promoter and/or leader region would be involved in the destabilization of ptsG mRNA. Recently, we demonstrated that the ptsG mRNA exhibited the normal response to G6P accumulation when the ptsG promoter was replaced with a foreign bla promoter on the chromosome (Morita et al. 2004). This ptsG allele (Pbla-ptsG2) produces an extended form of ptsG mRNA in which several nucleotides derived from the bla region are attached to the 5'-end of the wild-type ptsG mRNA (see Fig. 1). This implies that the ptsG promoter is not required for the destabilization of ptsG mRNA and that the modification of the original 5'-end of ptsG mRNA does not affect the stability control by phosphosugar stress. We confirmed this by showing that the mRNA derived from the cloned Pbla-ptsG2 gene on plasmid pTH121 is normally destabilized in the presence of MG (Fig. 3, lanes 1,2). Previously, we constructed a plasmid pTH111 in which the ptsG gene lacking the first 56 nt of 5'-UTR is under the control of the bla promoter (Takahashi et al. 1998). The bla promoter to drive the ptsG in pTH111 is the same as that used in ptsG2. The ptsG allele in pTH111 was designated as Pbla-ptsG1. By using this construct, we examined whether the 5' portion of the ptsG 5'-UTR is required for the destabilization of ptsG mRNA. The 5'-truncated ptsG mRNA was stably expressed when cells harboring pTH111 were grown in the presence of glucose while it was rapidly degraded when cells were grown in the presence of MG (Fig. 3, lanes 3,4). Further truncation of the 5'-UTR at least up to +70 affected neither the expression of ptsG mRNA nor its response to phosphosugar stress (H. Kawamoto, T. Morita, and H. Aiba, unpubl.). We conclude that the first 70 nt of the ptsG UTR is not required for the stability control of ptsG mRNA by phosphosugar stress.
Figure 3. Expression and response to phosphosugar stress of the variant ptsG mRNAs. (A) AS16 harboring the Pbla-ptsG2 plasmid pTH121 (lanes 1,2) or the Pbla-ptsG1 plasmid pTH111 (lanes 3,4) were grown with 1% Glc (lanes 1,3) or MG (lanes 2,4) to A600 = 0.6. Total proteins were prepared and samples equivalent to 0.0075 A600 units were subjected to Western blot analysis using anti-IIB antibody. (B) Cells were grown to A600 = 0.5. Glc (lanes 1,3) or MG (lanes 2,4) was added at a final concentration of 1%, and incubation was continued for 20 min. Total RNAs were prepared and 3 μg of RNA samples was subjected to Northern blot analysis using the ptsG probe.
The region up to the transmembrane segment II is sufficient for the response to phosphosugar stress
To define further the features of ptsG mRNA responsible for the rapid degradation in response to phosphosugar stress, a series of ptsG-crp translational fusions encoding hybrid proteins with N-terminal regions of IICBGlc and the entire CRP was constructed (Fig. 4). IICBGlc consists of a transmembrane IIC domain of 400 residues and a cytoplasmic IIB domain of 90 residues (Buhr and Erni 1993). The IIC domain contains eight transmembrane segments (TM) along with an N-terminal amphipathic sequence (Buhr and Erni 1993). The fusions were designed to produce IIC-CRP proteins in which the N-terminal 14, 40, 72, 107, and 139 residues of IICBGlc were joined to the CRP ORF. These positions correspond to those just after the N-terminal amphipathic sequence, TM-I, TM-II, TM-III, and TM-IV, respectively. Cells harboring a control plasmid pHA7MK expressing the crp mRNA or each of five different plasmids carrying ptsG-crp fusions were grown in the presence of either glucose or MG. Total RNAs and proteins were prepared and analyzed by Northern and Western blotting, respectively (Fig. 5). When cells were grown in the presence of glucose, all fusion genes produced significant amounts of hybrid proteins and mRNAs (Fig. 5A,B, lanes 3,5,7,9,11). On the other hand, the levels of hybrid proteins and ptsG-crp mRNAs were dramatically reduced when cells harboring pMM12, pMM13, or pMM14 were grown in the presence of MG (Fig. 5A,B, lanes 4,6,8,10,12). The reduction of hybrid proteins and ptsG-crp mRNAs by MG were no longer observed when RNase E was inactivated (data not shown). These results clearly indicate that the ptsG-crp chimeric mRNAs were rapidly degraded in response to phosphosugar stress as in the case of the native ptsG mRNA. The presence of MG weakly reduced the expression of the ptsG-crp mRNAs when cells harboring pMM10 or pMM11, while it did not affect the expression of crp mRNA. Taken together, we conclude that the information responsible for the rapid degradation in response to phosphosugar stress is contained within the region from positions +70 to +349 of the ptsG mRNA. This region consists of the 3' one-third of the leader region (+70 to +103) including the Shine-Dalgarno sequence and the region corresponding to the N-terminal portion of IICBGlc containing the first two transmembrane segments along with the amphipathic sequence (+103 to +349). The nucleotide position +349 corresponds to amino acid position 72.
Figure 4. (A) Membrane topology of IICBGlc. The deduced membrane topology of IICBGlc is taken from Buhr and Erni (1993). The portions corresponding to IIC and IIB domain are labeled. The N-terminal amphipathic helix and eight transmembrane (TM) domains are represented as rectangles. Arrowheads represent the sites (amino acid residue number of IICBGlc) to which the crp gene is fused. (B) Schematic representation of ptsG-crp translational fusion genes. Open and filled rectangles represent ptsG and crp coding sequences, respectively. The numbers in open rectangles represent amino acid positions of IICBGlc. The thick lines represent the part of the ptsG leader region. Arrows represent the direction of transcription driven by Pbla.
Figure 5. Expression and response to phosphosugar stress of ptsG-crp translational fusions. (A) IT1568 harboring pHA7MK (lanes 1,2), pMM10 (lanes 3,4), pMM11 (lanes 5,6), pMM12 (lanes 7,8), pMM13 (lanes 9,10), or pMM14 (lanes 11,12) was grown with 1% Glc or MG to A600 = 0.6. Total proteins were prepared, and indicated amounts of samples were subjected to Western blot analysis using anti-CRP antibody. (B) Cells were grown to A600 = 0.5. Glc or MG was added at a final concentration of 1%, and incubation was continued for 20 min. Total RNAs were prepared and indicated amounts of RNA samples were subjected to Northern blot analysis using the crp probe.
The membrane-targeting property of the transmembrane segments is important for the rapid degradation of ptsG-crp mRNA
It has been proposed that a 32-nt-long region from positions +77 to +107 of ptsG mRNA is the direct target of SgrS RNA because it is partially complementary to that of SgrS at nucleotide positions 157-187 (Vanderpool and Gottesman 2004). Our observation that the ptsG14-crp mRNA containing +70 to +142 region is moderately reduced in response to the phosphosugar stress is consistent with the proposed base-pairing model by Vanderpool and Gottesman (2004). However, it is also clear that this region alone is not sufficient for the efficient destabilization of ptsG mRNA. To examine the role of the transmembrane region in the mRNA destabilization, we introduced one nucleotide insertion at position +145 and one nucleotide deletion at position +349 in the ptsG72-crp fusion gene to generate the ptsG72FS-crp gene (Fig. 6A). These frameshift mutations, which create no intervening nonsense codons, cause only a slight change in the nucleotide sequence but a dramatic change in the amino acid sequence of TM-I and TM-II. Northern blot analysis revealed that the level of mRNA produced from the ptsG72FS-crp gene was only moderately reduced by MG (Fig. 6B, lanes 4,5). This suggests that the amino acid sequence but not nucleotide sequence of the TM-I/TM-II region is required for the efficient destabilization of ptsG mRNA.
Figure 6. Expression and response to phosphosugar stress of ptsG-crp translational fusion variants. (A) Schematic representation of altered ptsG-crp translational fusions. The rectangles in white and black represent the ptsG and crp coding regions, respectively. The rectangles in light and dark gray represent TM I-II of IICBGlc with frameshift mutations and LacY TM I-II, respectively. The frameshift mutations just before and after TM I-II are shown as +A (insertion) and -T (deletion) in the ptsG72FS-crp gene. (B) Northern blot analysis of altered ptsG-crp mRNAs. IT1568 harboring pMM12 (lanes 1,2), pMM12FS (lanes 3,4), or pMM10Y (lanes 5,6) was grown to A600 = 0.5. Glc or MG was added at a final concentration of 1%, and incubation was continued for 20 min. Total RNAs were prepared and 3 μg (lanes 1,2,5,6) or 1 μg (lanes 3,4) of RNA samples was subjected to Northern blot analysis using the crp probe.
The next question is how the amino acid sequence of the TM-I/TM-II region participates in the marked destabilization of ptsG mRNA. One possibility is that a particular amino acid sequence within this region somehow plays a role in the regulation of mRNA stability. Alternatively, the membrane-targeting property of the transmembrane segments may be needed for the destabilization of ptsG mRNA. To examine these possibilities, the region corresponding to TM-I and TM-II in the ptsG72-crp was replaced with that corresponding to two transmembrane segments derived from LacY (LacY TM-I-TM-II) to construct the ptsG14Y-crp allele (Fig. 6A). Northern blot analysis indicated that the mRNA produced from the ptsG14Y-crp allele was stably expressed in the presence of glucose, while it was markedly destabilized in the presence of MG (Fig. 6B, lanes 5,6). This strongly suggests that the membrane-targeting property rather than a particular amino acid sequence itself within TM-I and TM-II of IICBGlc is crucial for the rapid degradation of ptsG mRNA.
Proteins generated from mRNAs sensitive to phosphosugar stress are membrane-associated
To confirm the role of the membrane-targeting property of the transmembrane segments in the rapid degradation of ptsG mRNA, we performed a cell fractionation assay by Western blotting using anti-CRP antibody. As expected, the wild-type CRP was found exclusively in the soluble cytoplasmic fraction (Fig. 7, lanes 1-3). Under the same condition, IIC72-CRP possessing TM-I and TM-II generated from pMM12 appeared predominantly in the membrane fraction (Fig. 7, lanes 10-12), while IIC14-CRP lacking the transmembrane segments generated from pMM10 appeared predominantly in the soluble fraction (Fig. 7, lanes 4-6). IIC40-CRP possessing only TM-I was found in both soluble and membrane fractions (Fig. 7, lanes 7-9). As expected, IIC72FS-CRP generated from pMM12FS appeared almost exclusively in the cytoplasmic fraction (Fig. 7, lanes 13-15). These results again support our claim that the rapid degradation of ptsG mRNA in response to phosphosugar stress is tightly coupled with the membrane-targeting property of translated proteins. In addition, it should be noted that the expression levels of fusion proteins attached to membrane are greatly reduced compared to those of cytoplasmic proteins (e.g., cf. lanes 10 and 13 in Fig. 7). The reduced expression of membrane proteins could be due to either inefficient translation or reduced stability of proteins. However, we observed no significant difference in the protein stability among these proteins (data not shown). This suggests that the translation of membrane proteins is less efficient.
Figure 7. Subcellular localization of proteins. IT1568 harboring pHA7MK (lanes 1-3), pMM10 (lanes 4-6), pMM11 (lanes 7-9), pMM12 (lanes 10-12), or pMM12FS (lanes 13-15) was grown with 1% Glc to A600 = 0.6. Total proteins (T) were prepared and fractionated by centrifugation to obtain supernatant (S) and pellet (P). The fractionated protein samples were subjected to Western blot analysis using anti-CRP antibody. The proteins samples used are 0.005 A600 unit (lanes 1-6), 0.075 A600 unit (lanes 7-12), and 0.01 A600 unit (lanes 13-15).
The destabilization of ptsG mRNA in response to phosphosugar stress is dependent on Hfq and a small RNA (SgrS)
It is highly possible that Hfq is involved in the destabilization of ptsG mRNA in response to phosphosugar stress because this RNA chaperon has been shown to participate in post-transcriptional regulation by affecting stability and/or translation of several mRNAs often with small RNAs (Gottesman 2002, 2004; Storz et al. 2004; Valentin-Hansen et al. 2004). To test this possibility, we examined the effect of the hfq mutation on the expression of ptsG and ptsG-crp mRNAs in response to phosphosugar stress. The rapid degradation of ptsG mRNA in response to phosphosugar stress was abolished in cells lacking Hfq, while the hfq mutation did not affect the expression of ptsG mRNA in the presence of glucose (Fig. 8A,C, lanes 1-4). This implies that the destabilization of ptsG mRNA is dependent on Hfq. Recently, Vanderpool and Gottesman (2004) have discovered that the synthesis of SgrS, an Hfq-binding small RNA, is induced by phosphosugar stress and acts as an antisense RNA to mediate the destabilization of ptsG mRNA. We extended their results by using strains expressing Pbla-ptsG and ptsG-crp fusion genes. SgrS was highly induced in cells harboring ptsG plasmids in the presence of MG where the Pbla-ptsG and ptsG72-crp but not ptsG72-crp mRNAs are destabilized (Fig. 8B,D). The destabilization of Pbla-ptsG and ptsG72-crp mRNAs no longer occurs in sgrS cells (Fig. 8A,C, lanes 5,6). We also observed that the destabilization of ptsG14Y-crp mRNA in the presence of MG was abolished in hfq or sgrS cells (Fig. 8E). The level of SgrS RNA under the condition of phosphosugar stress is dramatically reduced in the hfq cells, suggesting that Hfq is also involved in the stabilization of SgrS RNA (Fig. 8B,D, lane 4). The reduction in the level of the SgrS (RyaA) RNA in an hfq mutant was previously observed (Zhang et al. 2003). The stabilization of RNA by Hfq has been shown for several other Hfq-binding small RNAs (Sledjeski et al. 2001; Moller et al. 2002; Masse et al. 2003; Moll et al. 2003).
Figure 8. Effects of hfq or sgrS mutation on the degradation of ptsG and ptsG-crp mRNAs. (A,B) AS16, AS21 (ptsG hfq::cat), and AS23 (ptsG sgrS) harboring pTH111 were grown to A600 = 0.5. Glc, or MG was added at a final concentration of 1%, and incubation was continued for 20 min. Total RNAs were prepared and 3 μg of RNA samples was subjected to Northern blot analysis using ptsG or sgrS probes. (C,D) IT1568, TM122 (hfq::cat), and TM542 (sgrS) harboring pMM12, and IT1568 harboring pMM12FS were grown to A600 = 0.5. Glc, or MG was added at a final concentration of 1%, and incubation was continued for 20 min. Total RNAs were prepared and 3 μg of RNA samples was subjected to Northern blot analysis using crp or sgrS probes. (E) IT1568, TM122, and TM542 harboring pMM10Y were grown to A600 = 0.5. Glc or MG was added at a final concentration of 1%, and incubation was continued for 20 min. Total RNAs were prepared and 3 μg of RNA samples was subjected to Northern blot analysis using the crp probe.
The SgrS-mediated degradation of ptsG-crp fusion mRNA lacking TM domains is enhanced when its translation is reduced
As shown above, the action of SgrS/Hfq on the target ptsG-crp mRNAs is weak in the absence of membrane localization presumably due to an efficient translation. To investigate a possible link between translation efficiency and SgrS/Hfq action, we introduced a single nucleotide substitution at position +90 in the Shine-Dalgarno sequence of ptsG72FS-crp (Fig. 1B). We examined the expression of the mutant fusion gene (ptsG72FSM-crp) in the presence of either glucose or MG by Western and Northern blotting. As expected, the expression of the fusion protein was dramatically decreased even in the normal condition (Fig. 9A, lane 3). In addition, the mRNA level was also markedly reduced in the presence of glucose presumably due to a reduced stability caused by inefficient translation (Fig. 9B, lane 3). Interestingly, however, the mutation markedly enhanced the degradation of mRNA in the presence of MG (Fig. 9B, lane 4), therefore in the presence of SgrS, resulting in no detectable fusion protein (Fig. 9A, lane 4). Thus, the SgrS-mediated degradation of the cytoplasmic fusion mRNA is apparently stimulated when its translation is slowed down. In other words, SgrS/Hfq retains the ability to act efficiently in the absence of membrane localization depending on the translation status of the target mRNA.
Figure 9. Effect of down-mutations in the ribosome-binding site on the response to phosphosugar stress of the cytoplasmic ptsG-crp mRNA. (A) IT1568 harboring pMM12FS (lanes 1,2) or pMM12FSM (lanes 3,4) was grown with 1% Glc or MG to A600 = 0.6. Total proteins were prepared, and the indicated amounts of samples were subjected to Western blot analysis using anti-CRP antibody. (B) Cells were grown to A600 = 0.5. Glc or MG was added at a final concentration of 1%, and incubation was continued for 20 min. Total RNAs were prepared, and the indicated amounts of RNA samples were subjected to Northern blot analysis using the crp probe.
Discussion
A class of small noncoding RNAs is involved in the regulation of gene expression at post-transcriptional levels by altering translation or stability of target mRNAs in bacteria (Gottesman 2002, 2004; Storz et al. 2004). These RNAs are believed to act through base-pairing with target mRNAs in an Hfq-dependent manner. For example, OxyS RNA inhibits the translation of target mRNAs by occluding the ribosome-binding sites (Altuvia et al. 1998; Zhang et al. 1998). In contrast, the DsrA and RprA RNAs stimulate rpoS translation by preventing formation of a translation-inhibitory hairpin structure in the rpoS 5'-untranslated region (Lease et al. 1998; Majdalani et al. 1998, 2002). Recent studies have shown further that small RNAs could alter the stability of target mRNAs. Typically, RyhB RNA induces a rapid degradation of its target mRNAs in an Hfq- and RNase E-dependent manner (Masse and Gottesman 2002; Masse et al. 2003). SgrS (RyaA) RNA is one of a group of newly identified small noncoding RNAs that bind to Hfq (Zhang et al. 2003). The work by Vanderpool and Gottesman (2004) demonstrated that SgrS mediates the destabilization of ptsG mRNA presumably through base-pairing with ptsG mRNA. They have shown that SgrS is expressed in response to phosphosugar accumulation and that SgrS expression is correlated with greatly reduced levels of ptsG mRNA. They have also found that the expression of sgrS is regulated by a novel transcriptional activator SgrR that may activate sgrS transcription in response to high intracellular levels of phosphosugars.
The major question addressed in the present study is what features of ptsG mRNA are responsible for the rapid decay by RNase E under conditions of phosphosugar accumulation. We first showed that the 5' portion up to +70 in the ptsG mRNA is dispensable for the destabilization of message. Then, we used a series of ptsG-crp translational fusions to specify the downstream boundary in the ptsG mRNA required for the response to phosphosugar stress. We found that the region up to the second transmembrane segment of IICBGlc is sufficient for the efficient message destabilization. The region required for the efficient destabilization of message covers the region complementary to SgrS RNA (see Fig. 1B). Thus, our data are consistent with the SgrS-ptsG mRNA base-pairing model by Vanderpool and Gottesman (2004).
However, our results indicate that the ptsG region complementary to SgrS RNA alone is not sufficient for the strong destabilization of the message because the ptsG-crp fusion mRNAs missing the transmembrane region are only moderately decreased in the stress conditions where SgrS is highly expressed. Our data clearly demonstrate that the ptsG region up to the second transmembrane segment is necessary for the efficient destabilization of ptsG mRNA. One simple explanation for this result is that the mRNA region corresponding to two transmembrane segments somehow are involved in the interaction with SgrS and/or components of RNase E degradosome. Another possibility is that the translated nascent peptide of the transmembrane segments of IICBGlc plays some roles in the SgrS-ptsG interaction and/or the action of RNase E degradosome. However, our experiments clearly exclude these possibilities because the transmembrane segments of IICBGlc can be replaced with those derived from an unrelated protein LacY. Thus, the simplest interpretation of our data is that membrane targeting ability of translated protein is crucial for the destabilization of ptsG mRNA in response to phosphosugar stress.
The sugar-specific permeases (EIIs) of PTS generally contain N-terminal amphipathic sequences of 13-20 residues that are believed to function in membrane insertion possibly cooperatively with the first hydrophobic transmembrane segment (Saier 1989; Saier et al. 1989; Yamada et al. 1991). Recently, it has been shown that mannitol permease (MtlA) is recognized by the bacterial signal recognition particle resulting in cotranslational targeting of the growing MtlA polypeptide to the inner membrane (Beck et al. 2000). IICBGlc has an N-terminal amphipathic sequence of 13 residues preceding eight transmembrane segments (Buhr and Erni 1993). Therefore, it is very likely that IICBGlc is also inserted into membranes in a cotranslational manner. If this is the case, the ptsG mRNA under translation is expected to localize near the inner membrane because translation is also coupled with transcription in bacteria. Taken together, we propose that mRNA localization to the inner membrane coupled with the membrane insertion of nascent peptide is a critical event for the Hfq/SgrS- and RNase E-mediated destabilization of the ptsG mRNA.
There are several possible models to explain the requirement of mRNA localization to the inner membrane for the rapid degradation of ptsG mRNA in response to phosphosugar stress. Firstly, the preferential localization of RNase E-based degradosome near the membrane could explain this. In this regard, there is a report that RNase E seems to be predominantly located near the cytoplasmic membrane (Liou et al. 2001). We showed recently that enolase within the degradosome is required for the destabilization of ptsG mRNA (Morita et al. 2004). Thus, one possible mechanism by which enolase stimulates ptsG mRNA degradation would be to modulate the membrane localization of RNase E. Further experiments are certainly needed to examine whether RNase E is indeed located near the membrane and whether enolase affects subcellular localization of RNase E. Secondly, the requirement of the mRNA localization to the inner membrane for the rapid degradation of message can be explained if SgrS or SgrS/Hfq complex were preferentially localized near the inner membrane by some mechanism. It is unlikely that SgrS or SgrS/Hfq itself is localized near membrane. However, additional factors including enolase could mediate the localization of SgrS/Hfq to the inner membrane. The third possibility would be that the action of either SgrS/Hfq or RNase E is somehow inhibited in the particular ptsG-crp fusion mRNA when it is in the cytoplasm by an unknown factor, while this inhibition is relieved when the mRNA is localized near the membrane. Thus, it is plausible that additional factor(s) other than RNase E, enolase, Hfq, SgrS, and SgrR are also involved in the regulation of stability of ptsG mRNA in response to phosphosugar stress. A simple attractive idea is that the ribosome itself is acting as an additional player for the regulation of stability of ptsG mRNA. Consistent with this idea, we have shown that the reduced translation of the cytoplasmic target mRNA by introducing single base substitutions in the ribosome-binding site allowed SgrS/Hfq to act efficiently in the absence of membrane localization.
Taken together, we would like to propose the following scenario concerning the regulation of translation and stability of ptsG mRNA (Fig. 10). When Mlc is inactivated in the presence of glucose, ptsG transcription is induced. As shown by Vanderpool and Gottesman (2004), and confirmed here, no SgrS RNA is made under nonstress conditions. Translation of ptsG mRNA occurs normally and the newly synthesized nascent polypeptide of IICBGlc would be cotranslationally targeted into the inner membrane. Under stress conditions, SgrS RNA is efficiently transcribed. We hypothesize that new rounds of translation of ptsG mRNA already being translated would be less effective compared to the first round of translation because the membrane localization of message may somehow affect the entry of new ribosomes. The fact that the expression level of IIC72-CRP retaining the membrane-targeting ability is markedly lower compared to that of IIC72FS-CRP, the nonmembrane counterpart of IIC72-CRP, supports this assumption. The inhibition of the ribosome entry may allow SgrS along with Hfq to act on the ptsG mRNA, resulting in SgrS-ptsG pairing, and therefore the RNase E-dependent degradation of the ptsG mRNA. When the target mRNA lacks the membrane-localization ability, the entry of new ribosomes occurs effectively, which, in turn, prevents the action of SgrS/Hfq on the ptsG mRNA. It remains to be studied whether the destabilization of ptsG mRNA is caused directly by SgrS-mRNA base-pairing or is an indirect effect of the inhibition of translation of ptsG by SgrS-mRNA base-pairing.
Figure 10. A model for ptsG mRNA degradation in response to the phosphosugar stress. In nonstress conditions, ptsG mRNA is translated and IICBGlc protein is targeted to cytoplasmic membrane in a cotranslational manner, resulting in localization of mRNA near the membrane. New rounds of ribosome entry into the ptsG mRNA being translated would be inhibited due to the membrane localization of message. Under stress conditions, SgrS RNA is highly transcribed by the action of SgrR. The inhibition of the ribosome entry may allow SgrS/Hfq to act on the ptsG mRNA, resulting in SgrS-ptsG pairing and therefore the RNase E-dependent degradation of the ptsG mRNA. When the target mRNA lacks the membrane-localization ability, the entry of new ribosomes occurs effectively, which, in turn, prevents the action of SgrS/Hfq on the ptsG mRNA.
In conclusion, the present study has shed a light on a unique feature, implication of mRNA localization to membrane, of small RNA-mediated mRNA degradation. It is certainly interesting to know whether this is true for other membrane-associated messages that may be regulated by small RNAs in bacteria. Finally, this novel mechanism could be relevant to the action of small regulatory RNAs in eukaryotes too.
Materials and methods
Media and growth conditions
Cells were grown aerobically at 37°C unless specified otherwise in LB medium. Antibiotics were used at the following concentrations when needed: ampicillin (50 μg/mL); chloramphenicol (15 μg/mL), and tetracycline (15 μg/mL). Bacterial growth was monitored by determining the optical density at 600 nm.
Bacterial strains and plasmids
The E. coli K12 strains and plasmids used in this study are listed in Table 1. IT1568 (W3110 mlc) was used as a parent wild-type strain. Other strains were constructed by one-step gene inactivation protocol (Datsenko and Wanner 2000). When necessary, the cat marker was eliminated by using an FLP expression plasmid pCP20. Plasmid pHA7MK was constructed from pHA7M by introducing a KpnI site just after the initiation codon of CRP. To construct pMM10, pMM12, pMM13, pMM14, and pMM15, DNA regions between a BamHI site upstream of Pbla and various sites corresponding to amino acid positions 14, 40, 72, 103, and 140 of IICBGlc in pTH111 were amplified by PCR and cloned between BamHI and KpnI sites of pHA7MK.
Table 1. Bacterial strains and plasmids used in this study
Northern blotting
Total RNAs were isolated from cells grown to mid-log phase as described (Aiba et al. 1981). For Northern blot analysis, indicated amounts of total RNAs were resolved by 1.2% agarose gel electrophoresis in the presence of formaldehyde and blotted on to Hybond-N+ membrane (Amersham Biosciences). The mRNAs were visualized using digoxigenin (DIG) reagents and kits for nonradioactive nucleic acid labeling and detection system (Roche Molecular Biochemicals) according to the procedure specified by the manufacturer. The following DIG-labeled DNA probes were prepared by PCR using DIG-dUTP: a 305-bp fragment corresponding to the 5' region of ptsG; a 621-bp fragment corresponding to the crp; and a 150-bp fragment corresponding to sgrS.
Western blotting
Bacterial cells grown in LB medium containing appropriate antibiotics(s) or carbon(s) were harvested and suspended in 100 μL of SDS-PAGE loading buffer (62.5 mM Tris-HCl at pH 6.8, 2% SDS, 10% glycerol, 5% -mercaptoethanol, 0.1% bromophenol blue). The sample was heated at 100°C for 5 min. Indicated amounts of protein samples were subjected to a 12% or 15% polyacrylamide-0.1% SDS gel electrophoresis and transferred to Immobilon membrane (Millipore). The membranes were treated with anti-IIB (Tanaka et al. 2000) or anti-CRP (Ishizuka et al. 1993) antibodies. Signals were visualized by the ECL system (Amersham).
Protein localization assay
Cells harboring indicated plasmids were grown in 20 mL of LB medium to A600 of 0.6. A 0.5-mL aliquot of cell cultures was directly sonicated for 5 sec twice at 0°C, with a 10-sec interval in between. The sonicated samples were centrifuged at 15, 000g (4°C) for 3 min. The pellets were dissolved in 30 μL of SDS sample buffer and used as the membrane fraction. The supernatant fraction was precipitated with cold trichloroacetic acid (5% final concentration). The precipitates were collected by centrifugation, washed with 70% acetone, dissolved in 30 μL of SDS sample buffer, and used as the cytoplasmic fraction.
Acknowledgments
We thank Jacqueline Plumbridge, Susan Gottesman, and Carin Vanderpool for critical reading of the manuscript and for helpful comments. This work was supported by Grants-in-Aid from the Ministry of Education, Culture, Sports, Science and Technology of Japan and by Ajinomoto Co., Inc.
References
Aiba, H., Adhya, S., and de Crombrugghe, B. 1981. Evidence for two functional gal promoters in intact Escherichia coli cells. J. Biol. Chem. 256: 11905-11910.
Altuvia, S., Zhang, A., Argaman, L., Tiwari, A., and Storz, G. 1998. The Escherichia coli OxyS regulatory RNA represses fhlA translation by blocking ribosome binding. EMBO J. 17: 6069-6075.
Beck, K., Wu, L.F., Brunner, J., and Muller, M. 2000. Discrimination between SRP- and SecA/SecB-dependent substrates involves selective recognition of nascent chains by SRP and trigger factor. EMBO J. 19: 134-143.
Buhr, A. and Erni, B. 1993. Membrane topology of the glucose transporter of Escherichia coli. J. Biol. Chem. 268: 11599-11603.
Datsenko, K.A. and Wanner, B.L. 2000. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. 97: 6640-6645.
El-Kazzaz, W., Morita, T., Tagami, H., Inada, T., and Aiba, H. 2004. Metabolic block at early stages of the glycolytic path-way activates the Rcs phosphorelay system via increased synthesis of dTDP-glucose in Escherichia coli. Mol. Microbiol. 51: 1117-1128.
Fraenkel, D.G. 1996. Glycolysis. In Escherichia coli and Salmonella typhimurium: Cellular and molecular biology (ed. F.C. Neidhardt), pp. 189-198. ASM Press, Washington, DC.
Gottesman, S. 2002. Stealth regulation: biological circuits with small RNA switches. Genes & Dev. 16: 2829-2842.[Free Full Text]
____. 2004. The small RNA regulators of Escherichia coli: Roles and mechanisms. Annu. Rev. Microbiol. 58: 303-328.
Ishizuka, H., Hanamura, A., Kunimura, T., and Aiba, H. 1993. A lowered concentration of cAMP receptor protein caused by glucose is an important determinant for catabolite repression in Escherichia coli. Mol. Microbiol. 10: 341-350.
Kimata, K., Takahashi, H., Inada, T., Postma, P., and Aiba, H. 1997. cAMP receptor protein-cAMP plays a crucial role in glucose-lactose diauxie by activating the major glucose transporter gene in Escherichia coli. Proc. Natl. Acad. Sci. 94: 12914-12919.
Kimata, K., Inada, T., Tagami, H., and Aiba, H. 1998. A global repressor (Mlc) is involved in glucose induction of the ptsG gene encoding major glucose transporter in Escherichia coli. Mol. Microbiol. 29: 1509-1519.
Kimata, K., Tanaka, Y., Inada, T., and Aiba, H. 2001. Expression of the glucose transporter gene, ptsG, is regulated at the mRNA degradation step in response to glycolytic flux in Escherichia coli. EMBO J. 20: 3587-3595.
Lease, R.A., Cusick, M.E., and Belfort, M. 1998. Riboregulation in Escherichia coli: DsrA RNA acts by RNA:RNA interactions at multiple loci. Proc. Natl. Acad. Sci. 95: 12456-12461.
Lee, S.J., Boos, W., Bouche, J.P., and Plumbridge, J. 2000. Signal transduction between a membrane-bound transporter, PtsG, and a soluble transcription factor, Mlc, of Escherichia coli. EMBO J. 19: 5353-5361.
Liou, G.G., Jane, W.N., Cohen, S.N., Lin, N.S., and Lin-Chao, S. 2001. RNA degradosomes exist in vivo in Escherichia coli as multicomponent complexes associated with the cytoplasmic membrane via the N-terminal region of ribonuclease E. Proc. Natl. Acad. Sci. 98: 63-68.
Majdalani, N., Cunning, C., Sledjeski, D., Elliott, T., and Gottesman, S. 1998. DsrA RNA regulates translation of RpoS message by an anti-antisense mechanism, independent of its action as an antisilencer of transcription. Proc. Natl. Acad. Sci. 95: 12462-12467.
Majdalani, N., Hernandez, D., and Gottesman, S. 2002. Regulation and mode of action of the second small RNA activator of RpoS translation, RprA. Mol. Microbiol. 46: 813-826.
Masse, E. and Gottesman, S. 2002. A small RNA regulates the expression of genes involved in iron metabolism in Escherichia coli. Proc. Natl. Acad. Sci. 99: 4620-4625.
Masse, E., Escorcia, F.E., and Gottesman, S. 2003. Coupled degradation of a small regulatory RNA and its mRNA targets in Escherichia coli. Genes & Dev. 17: 2374-2383.
Meadow, N.D., Fox, D.K., and Roseman, S. 1990. The bacterial phosphoenolpyruvate: Glycose phosphotransferase system. Annu. Rev. Biochem. 59: 497-542.
Moll, I., Afonyushkin, T., Vytvytska, O., Kaberdin, V.R., and Blasi, U. 2003. Coincident Hfq binding and RNase E cleavage sites on mRNA and small regulatory RNAs. RNA 9: 1308-1314.
Moller, T., Franch, T., Hojrup, P., Keene, D.R., Bachinger, H.P., Brennan, R.G., and Valentin-Hansen, P. 2002. Hfq: A bacterial Sm-like protein that mediates RNA-RNA interaction. Mol. Cell 9: 23-30.
Morita, T., El-Kazzaz, W., Tanaka, Y., Inada, T., and Aiba, H. 2003. Accumulation of glucose 6-phosphate or fructose 6-phosphate is responsible for destabilization of glucose transporter mRNA in Escherichia coli. J. Biol. Chem. 278: 15608-15614.
Morita, T., Kawamoto, H., Mizota, T., Inada, T., and Aiba, H. 2004. Enolase in the RNA degradosome plays a crucial role in the rapid decay of glucose transporter mRNA in the response to phosphosugar stress in Escherichia coli. Mol. Microbiol. 54: 1063-1075.
Nam, T.W., Cho, S.H., Shin, D., Kim, J.H., Jeong, J.Y., Lee, J.H., Roe, J.H., Peterkofsky, A., Kang, S.O., Ryu, S., et al. 2001. The Escherichia coli glucose transporter enzyme IICB(Glc) recruits the global repressor Mlc. EMBO J. 20: 491-498.
Plumbridge, J. 1998. Expression of ptsG, the gene for the major glucose PTS transporter in Escherichia coli, is repressed by Mlc and induced by growth on glucose. Mol. Microbiol. 29: 1053-1063.
____. 2002. Regulation of gene expression in the PTS in Escherichia coli: The role and interactions of Mlc. Curr. Opin. Microbiol. 5: 187-193.
Postma, P.W., Lengeler, J.W., and Jacobson, G.R. 1996. Phosphoenolpyruvate: Carbohydrate phosphotransferase systems. In Escherichia coli and Salmonella typhimurium: Cellular and molecular biology. (ed. F.C. Neidhardt), pp. 1149-1174. ASM Press, Washington, DC.
Saier Jr., M.H. 1989. Current studies on the bacterial phosphotransferase system in the Saier laboratory (La Jolla, California, summer, 1988). FEMS Microbiol. Rev. 5: 183-192.
Saier Jr., M.H., Werner, P.K., and Muller, M. 1989. Insertion of proteins into bacterial membranes: Mechanism, characteristics, and comparisons with the eucaryotic process. Microbiol. Rev. 53: 333-366.
Seitz, S., Lee, S.J., Pennetier, C., Boos, W., and Plumbridge, J. 2003. Analysis of the interaction between the global regulator Mlc and EIIBGlc of the glucose-specific phosphotransferase system in Escherichia coli. J. Biol. Chem. 278: 10744-10751.
Sledjeski, D.D., Whitman, C., and Zhang, A. 2001. Hfq is necessary for regulation by the untranslated RNA DsrA. J. Bacteriol. 183: 1997-2005.
Storz, G., Opdyke, J.A., and Zhang, A. 2004. Controlling mRNA stability and translation with small, noncoding RNAs. Curr. Opin. Microbiol. 7: 140-144.
Takahashi, H., Inada, T., Postma, P., and Aiba, H. 1998. CRP down-regulates adenylate cyclase activity by reducing the level of phosphorylated IIA(Glc), the glucose-specific phosphotransferase protein, in Escherichia coli. Mol. Gen. Genet. 259: 317-326.
Tanaka, Y., Kimata, K., and Aiba, H. 2000. A novel regulatory role of glucose transporter of Escherichia coli: Membrane sequestration of a global repressor Mlc. EMBO J. 19: 5344-5352.
Tanaka, Y., Itoh, F., Kimata, K., and Aiba, H. 2004. Membrane localization itself but not binding to IICB is directly responsible for the inactivation of the global repressor Mlc in Escherichia coli. Mol. Microbiol. 53: 941-951.
Tchieu, J.H., Norris, V., Edwards, J.S., and Saier Jr., M.H. 2001. The complete phosphotranferase system in Escherichia coli. J. Mol. Microbiol. Biotechnol. 3: 329-346.
Valentin-Hansen, P., Eriksen, M., and Udesen, C. 2004. The bacterial Sm-like protein Hfq: A key player in RNA transactions. Mol. Microbiol. 51: 1525-1533.
Vanderpool, C.K and Gottesman, S. 2004. Involvement of a novel transcriptional activator and small RNA in the response to phosphosugar stress. Mol. Microbiol. 54: 1076-1089.
Yamada, Y., Chang, Y.Y., Daniels, G.A., Wu, L.F., Tomich, J.M., Yamada, M., and Saier Jr., M.H. 1991. Insertion of the mannitol permease into the membrane of Escherichia coli. Possible involvement of an N-terminal amphiphilic sequence. J. Biol. Chem. 266: 17863-17871.
Zhang, A., Altuvia, S., Tiwari, A., Argaman, L., Hengge-Aronis, R., and Storz, G. 1998. The OxyS regulatory RNA represses rpoS translation and binds the Hfq (HF-I) protein. EMBO J. 17: 6061-6068.
Zhang, A., Wassarman, K.M., Rosenow, C., Tjaden, B.C., Storz, G., and Gottesman, S. 2003. Global analysis of small RNA and mRNA targets of Hfq. Mol. Microbiol. 50: 1111-1124.(Hiroshi Kawamoto, Teppei )