Title: Accumulation of Glucose 6-Phosphate or Fructose 6-Phosphate Is Responsible for Destabilization of Glucose Transporter mRNA inEscherichia coli
Abstract: Previously we found that a mutation in either pgi or pfkA, encoding phosphoglucose isomerase or phosphofructokinase A, respectively, facilitates degradation of the ptsG mRNA in an RNase E-dependent manner in Escherichia coli (1Kimata K. Tanaka Y. Inada T. Aiba H. EMBO J. 2001; 20: 3587-3595Crossref PubMed Scopus (116) Google Scholar). In this study, we examined the effects of a series of glycolytic genes on the degradation of ptsG mRNA and how the mutations destabilize the ptsG mRNA. The conditional lethal mutation ts8 in fda, encoding fructose-1,6-P2 aldolase just downstream ofpfkA in the glycolytic pathway, caused the destabilization of ptsG mRNA at the nonpermissive temperature. Mutations in any other gene did not destabilize the ptsGmRNA; rather, they reduced the ptsG transcription mainly by affecting the cAMP level. The rapid degradation ofptsG mRNA in mutant strains was completely dependent upon the presence of glucose or any one of its compounds, which enter the Embden-Meyerhof glycolytic pathway before the block points. A significant increase in the intracellular glucose-6-P level was observed in the presence of glucose in the pgi strain. An overexpression of glucose-6-phosphate dehydrogenase eliminated both the accumulation and the degradation of ptsG mRNA in the pgi strain. In addition, accumulation of fructose-6-P led to the rapid degradation of ptsG mRNA in a pgi pfkA mutant strain lacking glucose-6-P. We conclude that the RNase E-dependent destabilization ofptsG mRNA occurs in response to accumulation of glucose-6-P or fructose-6-P. Previously we found that a mutation in either pgi or pfkA, encoding phosphoglucose isomerase or phosphofructokinase A, respectively, facilitates degradation of the ptsG mRNA in an RNase E-dependent manner in Escherichia coli (1Kimata K. Tanaka Y. Inada T. Aiba H. EMBO J. 2001; 20: 3587-3595Crossref PubMed Scopus (116) Google Scholar). In this study, we examined the effects of a series of glycolytic genes on the degradation of ptsG mRNA and how the mutations destabilize the ptsG mRNA. The conditional lethal mutation ts8 in fda, encoding fructose-1,6-P2 aldolase just downstream ofpfkA in the glycolytic pathway, caused the destabilization of ptsG mRNA at the nonpermissive temperature. Mutations in any other gene did not destabilize the ptsGmRNA; rather, they reduced the ptsG transcription mainly by affecting the cAMP level. The rapid degradation ofptsG mRNA in mutant strains was completely dependent upon the presence of glucose or any one of its compounds, which enter the Embden-Meyerhof glycolytic pathway before the block points. A significant increase in the intracellular glucose-6-P level was observed in the presence of glucose in the pgi strain. An overexpression of glucose-6-phosphate dehydrogenase eliminated both the accumulation and the degradation of ptsG mRNA in the pgi strain. In addition, accumulation of fructose-6-P led to the rapid degradation of ptsG mRNA in a pgi pfkA mutant strain lacking glucose-6-P. We conclude that the RNase E-dependent destabilization ofptsG mRNA occurs in response to accumulation of glucose-6-P or fructose-6-P. phosphoenolpyruvate-dependent sugar phosphotransferase system enzyme II complex of the PTS cAMP receptor protein fructose-1,6-diphosphate aldolase tryptone broth Luria-Bertani broth In bacteria, a number of sugars represented by glucose are transported into the cells coupled with their phosphorylation by the phosphoenolpyruvate-dependent sugar phosphotransferase system (PTS)1 (2Meadow N.D. Fox D.K. Roseman S. Annu. Rev. Biochem. 1990; 59: 497-542Crossref PubMed Scopus (302) Google Scholar, 3Postma P.W. Lengeler J.W. Jacobson G.R. Neidhardt F.C. 2nd Ed. Escherichia coli and Salmonella: Cellular and Molecular Biology. ASM Press, Washington, D. C.1996: 1149-1174Google Scholar, 4Saier Jr., M.H. Ramseier T.M. Reizer J. Neidhardt F.C. 2nd Ed. Escherichia coli and Salmonella: Cellular and Molecular Biology. ASM Press, Washington, D. C.1996: 1325-1343Google Scholar), whereas the translocation of some other sugars such as lactose is catalyzed by non-PTS transport systems. In either case, the incorporated sugars are metabolized primarily by the Embden-Meyerhof glycolytic pathway and by the pentose phosphate pathway to produce numerous intermediary metabolites as well as energy in cells (5Fraenkel D.G. Neidhardt F.C. 2nd Ed. Escherichia coli and Salmonella: Cellular and Molecular Biology. ASM Press, Washington, D. C.1996: 189-198Google Scholar). The PTS in Escherichia coli consists of two common cytoplasmic proteins, enzyme I and HPr (histidine-containing protein of the PTS), as well as an array of sugar-specific enzyme II complexes (EIIs). The glucose-specific EII (glucose transporter) consists of cytoplasmic protein IIAGlc and membrane receptor IICBGlcencoded by crr and ptsG, respectively. The phosphoryl group from phosphoenolpyruvate is transferred sequentially to enzyme I, HPr, the EIIs, and finally glucose as it is translocated across the membrane. In addition to sugar transport and phosphorylation, the PTS plays important regulatory roles in a variety of cellular activities. This is particularly evident for the glucose-specific PTS. For example, IIAGlc regulates both the transport of non-PTS sugars and the activity of adenylate cyclase depending on its phosphorylation state (2Meadow N.D. Fox D.K. Roseman S. Annu. Rev. Biochem. 1990; 59: 497-542Crossref PubMed Scopus (302) Google Scholar, 3Postma P.W. Lengeler J.W. Jacobson G.R. Neidhardt F.C. 2nd Ed. Escherichia coli and Salmonella: Cellular and Molecular Biology. ASM Press, Washington, D. C.1996: 1149-1174Google Scholar, 4Saier Jr., M.H. Ramseier T.M. Reizer J. Neidhardt F.C. 2nd Ed. Escherichia coli and Salmonella: Cellular and Molecular Biology. ASM Press, Washington, D. C.1996: 1325-1343Google Scholar). The former process, called inducer exclusion, is fully responsible for the glucose-lactose diauxie that is a prototype of catabolite repression (6Inada T. Kimata K. Aiba H. Genes Cells. 1996; 1: 293-301Crossref PubMed Scopus (165) Google Scholar, 7Kimata K. Takahashi H. Inada T. Postma P. Aiba H. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 12914-12919Crossref PubMed Scopus (118) Google Scholar). A striking recent discovery regarding the regulatory function of PTS is that IICBGlc, depending on its phosphorylation state, interacts with Mlc to modulate the cellular localization and activity of this global repressor protein (8Tanaka Y. Kimata K. Aiba H. EMBO J. 2000; 19: 5344-5352Crossref PubMed Scopus (124) Google Scholar, 9Lee S.J. Boos W. Bouche J.P. Plumbridge J. EMBO J. 2000; 19: 5353-5361Crossref PubMed Scopus (122) Google Scholar, 10Nam 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. Seok Y.J. EMBO J. 2001; 20: 491-498Crossref PubMed Scopus (105) Google Scholar). The expression of the ptsG gene encoding IICBGlcis regulated by two global control systems at the level of transcription initiation (7Kimata K. Takahashi H. Inada T. Postma P. Aiba H. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 12914-12919Crossref PubMed Scopus (118) Google Scholar, 11Kimata K. Inada T. Tagami H. Aiba H. Mol. Microbiol. 1998; 29: 1509-1519Crossref PubMed Scopus (73) Google Scholar, 12Plumbridge J. Mol. Microbiol. 1998; 29: 1053-1063Crossref PubMed Scopus (94) Google Scholar). First, it is under positive control by CRP-cAMP; hence, the ptsG expression is absolutely dependent on this complex. Second, the transcription of theptsG gene is negatively regulated by a global repressor, Mlc. Recent studies have established that external glucose inducesptsG transcription by modulating the Mlc-mediated regulatory pathway (7Kimata K. Takahashi H. Inada T. Postma P. Aiba H. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 12914-12919Crossref PubMed Scopus (118) Google Scholar, 11Kimata K. Inada T. Tagami H. Aiba H. Mol. Microbiol. 1998; 29: 1509-1519Crossref PubMed Scopus (73) Google Scholar, 12Plumbridge J. Mol. Microbiol. 1998; 29: 1053-1063Crossref PubMed Scopus (94) Google Scholar). When glucose is transported into the cell, IICBGlc is dephosphorylated and binds Mlc, resulting in the sequestration of Mlc at the membrane (8Tanaka Y. Kimata K. Aiba H. EMBO J. 2000; 19: 5344-5352Crossref PubMed Scopus (124) Google Scholar, 9Lee S.J. Boos W. Bouche J.P. Plumbridge J. EMBO J. 2000; 19: 5353-5361Crossref PubMed Scopus (122) Google Scholar, 10Nam 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. Seok Y.J. EMBO J. 2001; 20: 491-498Crossref PubMed Scopus (105) Google Scholar). Another unexpected discovery regarding the regulation of ptsG expression is that a mutation in either pgi or pfkA, encoding glycolytic enzyme phosphoglucose isomerase or phosphofructokinase A, respectively, leads to a rapid degradation of ptsG mRNA and that RNase E is responsible for the degradation of ptsGmRNA (1Kimata K. Tanaka Y. Inada T. Aiba H. EMBO J. 2001; 20: 3587-3595Crossref PubMed Scopus (116) Google Scholar). Thus, ptsG expression also is regulated at the level of mRNA degradation, presumably in response to the glycolytic flux. However, it remains unclear how the glycolytic flux is involved in the destabilization of ptsG mRNA, and little is known about the mechanism of the stimulation of ptsG mRNA degradation except that RNase E is a major player in this process. In this study, we first investigated whether mutations in glycolytic genes other than pgi and pfkA affect the stability of ptsG mRNA. We found that the conditional lethal mutation ts8 in fda, encoding fructose-1,6-P2 aldolase (Fda) (13Singer M. Rossmiessl P. Cali B.M. Liebke H. Gross C.A. J. Bacteriol. 1991; 173: 6242-6248Crossref PubMed Google Scholar), stimulates degradation of ptsG mRNA but that mutations in any other glycolytic gene did not affect ptsG mRNA degradation. Then we addressed the question of how the mutations destabilize ptsGmRNA. The degradation of ptsG mRNA in mutant cells occurred only in the presence of glucose or any one of the glycolytic intermediates upstream of the block. More specifically, the degradation of ptsG mRNA was associated with elevated levels of glucose-6-P and/or fructose-6-P. We propose that accumulation of glucose-6-P or fructose-6-P is responsible for ptsGmRNA degradation. The physiological relevance of this novel feedback regulatory system at the post-transcriptional step is discussed below. The bacterial strains used in this study are listed in TableI. A series of disruption mutants were constructed by the one-step gene inactivation protocol that is based on the high efficiency of the phage λ Red recombinase (14Datsenko K.A. Wanner B.L. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 6640-6645Crossref PubMed Scopus (11304) Google Scholar). The FLP recognition target-flanked resistance gene was eliminated by using an FLP expression plasmid, pCP20 (14Datsenko K.A. Wanner B.L. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 6640-6645Crossref PubMed Scopus (11304) Google Scholar). To transfer the ts8 fda allele to other backgrounds by P1 phage, the catgene was inserted 2.7 kb downstream of the chromosomal ts8 fda gene in CAG417. IT1568 carrying a frameshift mutation in themlc gene was isolated spontaneously. The zwf gene including the promoter region was amplified by PCR using primers 5′-CCCAAGCTTGTGCCGCACTTTGCGCGCT-3′ and 5′-CCCAAGCTTGGCCTGTAACCGGAGCTCA-3′. The amplified DNA fragment was digested with HindIII and cloned intoHindIII-digested pBR322 to construct plasmid pTM6. Cells were grown at 37 °C in LB medium or TB medium (15Miller J. Experiments in Molecular Genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1972: 433Google Scholar) supplemented with kanamycin (15 μg/ml), tetracycline (15 μg/ml), chloramphenicol (15 μg/ml), and/or ampicillin (50 μg/ml) when needed. Bacterial growth was monitored by determining the optical density at 600 nm.Table IBacterial strains used in this studyStrainRelevant genotypeSourceW3110Wild typeLaboratory stockTM54W3110Δpgi::catThis studyCAG417CSH57A fda ts8C. GrossTM191W3110fda ts8 yggC756::catThis studyGW20W3110ams1 zce726::Tn10Ref. 23Wachi M. Umitsuki G. Nagai K. Mol. Gen. Genet. 1997; 253: 515-519Crossref PubMed Scopus (63) Google ScholarTM234W3110ams1 zce726::Tn10 fda ts8 yggC756::catThis studyTM238W3110pfkA::Tn5This studyTM84W3110ΔtpiA::catThis studyTM61W3110Δpgk::catThis studyTM81W3110ΔgpmBΔgpmA::catThis studyTM154W3110ΔpykAΔpykF::catThis studyIT1568W3110mlcThis studyTM162W3110mlcΔpgi::catThis studyIT1186W3110mlc ptsI::Tn5Ref. 18Inada T. Takahashi H. Mizuno T. Aiba H. Mol. Gen. Genet. 1996; 253: 198-204Crossref PubMed Scopus (32) Google ScholarTM145W3110mlc Δpgi::cat ptsI::Tn5This studyIT1165W3110mlc pfkA::Tn5Ref. 1Kimata K. Tanaka Y. Inada T. Aiba H. EMBO J. 2001; 20: 3587-3595Crossref PubMed Scopus (116) Google ScholarTM109W3110mlc pfkA::Tn5Δpgi::catThis studyTM163W3110mlcΔzwf::catThis studyTM164W3110mlcΔpgm::catThis studyTM239W3110mlc ΔpgiΔzwf::catThis studyTM240W3110mlc ΔpgiΔzwf::catThis study Open table in a new tab Total cellular RNAs were isolated from exponentially growing cells as described (16Aiba H. Adhya S. de Crombrugghe B. J. Biol. Chem. 1981; 256: 11905-11910Abstract Full Text PDF PubMed Google Scholar). The RNAs were resolved by 1.2% agarose gel electrophoresis in the presence of formaldehyde and blotted onto a Hybond-N+ membrane (Amersham Biosciences) as described (17Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989: 202-203Google Scholar). The mRNAs were visualized using digoxigenin reagents and kits for nonradioactive nucleic acid labeling and detection systems (Roche Molecular Biochemicals) according to the procedure specified by the manufacturer. The 305-bp digoxigenin-labeled DNA probe corresponding to the 5′-ptsG region was used. Bacterial cells were grown in LB medium to A 600 = 0.6. Two milliliters of the culture were taken and centrifuged at 12,000 ×g for 20 s at 4 °C. The pellet was suspended in 40 μl of 5 mm NaCl and immediately heated at 100 °C for 5 min. After adding 160 μl of ethanol, the mixture was chilled at −70 °C for 30 min and centrifuged at 14,000 × gfor 30 min at 4 °C. The resulting supernatant was dried up, dissolved in 20 μl of H2O, and used for cAMP assay. The determination of cAMP by gel mobility shift assay has been described previously (18Inada T. Takahashi H. Mizuno T. Aiba H. Mol. Gen. Genet. 1996; 253: 198-204Crossref PubMed Scopus (32) Google Scholar). The cellular concentration of cAMP was calculated on the assumption that an A 600 of 1.4 corresponds to 109 cells/ml (15Miller J. Experiments in Molecular Genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1972: 433Google Scholar) and that the volume of a cell is 2 × 10−12 ml (19Joseph E. Bernsley C. Guiso N. Ullmann A. Mol. Gen. Genet. 1982; 185: 262-268Crossref PubMed Scopus (46) Google Scholar). Bacterial cells were grown in LB medium toA 600 = 0.6 unless otherwise specified. One milliliter of the culture was centrifuged at 10,000 ×g for 2 min at room temperature. The pellet was suspended in 100 μl of H2O, immediately added to 50 μl of 5 m HClO4, and chilled on ice. After adding 100 μl of 2.5 m K2CO3, the mixture was centrifuged at 14,000 × g for 10 min at 4 °C. The supernatant was used for the glucose-6-P assay. The assay was performed according to the method of Hogema et al. (20Hogema B.M. Arents J.C. Bader R. Postma P.W. Mol. Microbiol. 1999; 31: 1825-1833Crossref PubMed Scopus (31) Google Scholar). The intracellular concentration of glucose-6-P was calculated as in the case of the cAMP assay. Among glycolytic genes other thanpgi and pfkA, we first focused on fda, which encodes Fda just downstream of pfkA in the Embden-Meyerhof glycolytic pathway (Fig.1). It is known that the Fda activity in a strain carrying a temperature-sensitive fda allele (ts8 or h8) was essentially undetectable at 42 °C (13Singer M. Rossmiessl P. Cali B.M. Liebke H. Gross C.A. J. Bacteriol. 1991; 173: 6242-6248Crossref PubMed Google Scholar). The ts8 mutation was shown to inhibit stable RNA synthesis and cell growth depending on glucose metabolism at the nonpermissive temperature (21Singer M. Walter W.A. Cali B.M. Rouviere P. Liebke H.H. Gourse R.L. Gross C.A. J. Bacteriol. 1991; 173: 6249-6257Crossref PubMed Google Scholar). To examine whether the ts8mutation affects the expression and/or degradation of ptsGmRNA, the ts8 allele was transferred to a W3110 background, and the expression of ptsG mRNA in the presence of glucose was analyzed by Northern blotting using aptsG DNA probe. The full-length ptsG mRNA was expressed stably in the ts8 fda strain at 30 °C as in the wild-type strain, whereas the pgi mutation dramatically reduced the full-length ptsG mRNA, resulting in an extensive smear because of the stimulation of RNase E-mediated mRNA degradation (Fig. 2, lanes 1–3). When the temperature was shifted to 42 °C, the ts8 fda strain gave a smear of degradation intermediates ofptsG mRNA (Fig. 2, lane 4), whereas the temperature upshift to 42 °C did not affect the Northern pattern in the wild-type and pgi strains (data not shown). This implies that the loss of Fda activity leads to the rapid degradation ofptsG mRNA in the presence of glucose. The introduction of an rne allele, ams1 (22Kuwano M. Ono M. Endo H. Hori K. Nakamura K. Hirota Y. Ohnishi Y. Mol. Gen. Genet. 1977; 154: 279-285Crossref PubMed Scopus (63) Google Scholar, 23Wachi M. Umitsuki G. Nagai K. Mol. Gen. Genet. 1997; 253: 515-519Crossref PubMed Scopus (63) Google Scholar), encoding a temperature-sensitive RNase E into the ts8 fda mutant prevented the rapid degradation of ptsG mRNA at the nonpermissive temperature (Fig. 2, lane 5). Thus, theptsG mRNA is destabilized in an RNase E-dependent manner in the ts8 fda strain at 42 °C as in the case of the pgi and pfkAstrains (1Kimata K. Tanaka Y. Inada T. Aiba H. EMBO J. 2001; 20: 3587-3595Crossref PubMed Scopus (116) Google Scholar).Figure 2Effects of mutations in glycolytic genes on the expression of ptsG mRNA. W3110 (lane 1), TM54 (lane 2), TM84 (lane 6), TM61 (lane 7), TM81 (lane 8), and TM154 (lane 9) were grown in LB medium with 1% glucose at 37 °C. Cellular RNAs were prepared at A 600= 0.6. In the case of TM191 (lanes 3 and 4) and TM234 (lane 5), cells were grown in LB medium with 1% glucose at 30 °C to A 600 = 0.5, and the temperature was shifted to 42 °C (lanes 4 and5) or kept at 30 °C (lane 3). Cellular RNAs were prepared after further incubation for 10 min. The RNA samples (15 μg) were subjected to Northern blot analysis. WT, wild type.View Large Image Figure ViewerDownload Hi-res image Download (PPT) To examine the effects of mutations in other glycolytic genes on the degradation of ptsG mRNA, a series of deletion mutants of glycolytic genes downstream of fda were constructed. We successfully disrupted tpiA, pgk, gpm, andpyk encoding triose-phosphate isomerase, phosphoglycerate kinase, phosphoglycerate mutase, and pyruvate kinase, respectively, although we failed to disrupt gap andeno encoding glyceraldehyde-3-phosphate dehydrogenase and enolase, respectively. The ptsG mRNA expression in each mutant was analyzed by Northern blotting. As shown in Fig. 2, each of these mutations affected the expression of ptsG mRNA to various extents. However, the Northern patterns of the ptsGmRNA in these mutant strains were essentially the same as those of the wild-type, indicating that these mutations do not affect significantly the degradation of ptsG mRNA (Fig. 2,lanes 6–9). We conclude that mutational blocks only in early stages of the glycolytic pathway produce a signal that leads to the rapid degradation of ptsG mRNA. The effects of mutations in the glycolytic genes on ptsG expression also were analyzed by Western blotting. The level of IICBGlc was correlated with that of the full-length ptsG mRNA in each strain (data not shown). It is likely that the mutations in the glycolytic genes downstream of fda reduce theptsG expression at the level of transcription initiation presumably by affecting the CRP-cAMP and/or Mlc pathways becauseptsG transcription is regulated by these two global systems (7Kimata K. Takahashi H. Inada T. Postma P. Aiba H. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 12914-12919Crossref PubMed Scopus (118) Google Scholar, 11Kimata K. Inada T. Tagami H. Aiba H. Mol. Microbiol. 1998; 29: 1509-1519Crossref PubMed Scopus (73) Google Scholar, 12Plumbridge J. Mol. Microbiol. 1998; 29: 1053-1063Crossref PubMed Scopus (94) Google Scholar). However, the introduction of the mlc mutation did not affect the ptsG mRNA level in each mutant strain in the presence of glucose, suggesting that the Mlc pathway is not involved in the reduction of ptsG expression (data not shown). We then determined the intracellular cAMP levels in each mutant strain both in the presence and absence of glucose. We observed that the reduction in the intracellular cAMP level by glucose was more significant in the tpiA, gpm, and pykstrains than in the wild-type strain (TableII). Thus, the reduced expression ofptsG mRNA in the tpiA, gpm, andpyk strains particularly in the presence of glucose appears to be at least in part due to the reduction of the intracellular cAMP level. It remains to be studied why ptsG expression is reduced in the pgk strain, in which no significant reduction of cAMP was observed.Table IIIntracellular cAMP levels in exponentially growing cellsStrainGenotypeIntracellular cAMP−Glucose+GlucoseμmμmW3110Wild type5.0 ± 0.62.0 ± 0.2TM54Δpgi4.7 ± 1.03.6 ± 0.1TM84ΔtpiA3.6 ± 0.40.7 ± 0.1TM61Δpgk5.5 ± 0.82.6 ± 0.6TM81ΔgpmAΔgpmB4.5 ± 0.31.4 ± 0.2TM154ΔpykAΔpykF3.5 ± 0.50.7 ± 0.2Intracellular cAMP levels were determined as described under "Experimental Procedures." Each value is the average of three independent experiments. Open table in a new tab Intracellular cAMP levels were determined as described under "Experimental Procedures." Each value is the average of three independent experiments. The ptsG mRNA is highly expressed in the mlc strain without the addition of glucose to the medium, and the ptsG mRNA expression is moderately reduced in the presence of external glucose because of the reduction in cAMP and CRP levels (11Kimata K. Inada T. Tagami H. Aiba H. Mol. Microbiol. 1998; 29: 1509-1519Crossref PubMed Scopus (73) Google Scholar). When the ptsGmRNA was analyzed in the pgi mlc double mutant growing in LB medium, rapid degradation of ptsG mRNA occurred in the presence of external glucose as expected (Fig.3, lane 4). However, a significant degradation of ptsG mRNA also was observed even without the addition of external glucose (Fig. 3, lane 3). This raises a question as to whether the rapid degradation ofptsG mRNA is dependent upon glucose metabolism. To test this, we introduced the ptsI mutation into thepgi mlc strain and examined the expression ofptsG mRNA by Northern blotting. The ptsGmRNA was stabilized in the pgi mlc ptsI triple mutant where the uptake of glucose is prevented (Fig. 3, lanes 5and 6). It is apparent that the degradation ofptsG mRNA observed in the pgi mlc double mutant growing in LB medium without external glucose was due to trace amounts of endogenous glucose in the LB medium. In fact, the rapid degradation of ptsG mRNA no longer occurred in themlc pgi and mlc pfkA double mutant strains without the addition of glucose when TB medium was used (Fig.4, A and B,lane 1). The addition of glucose caused a dramatic degradation of ptsG mRNA in both strains growing on TB medium (Fig. 4, A and B, lane 2). These results clearly indicate that the destabilization ofptsG mRNA in mutant strains occurs depending on the glucose uptake and metabolism.Figure 4Effects of various carbon sources on the degradation of ptsG mRNA. TM162 (A) and IT1165 (B) were grown in TB medium with and without 0.1% compounds indicated to A 600 = 0.6. Cellular RNAs were prepared, and 15 μg of each of the RNA samples was subjected to Northern blot analysis. The compounds used are glucose (Glc), lactose (Lac), mannitol (Mtl), xylose (Xyl), fructose (Fru), glycerol (Gly), glucose-6-P (G6P), and fructose-6-P (F6P).View Large Image Figure ViewerDownload Hi-res image Download (PPT) To test whether glucose generated internally also could lead to the destabilization ofptsG mRNA in the mutant strains, the effect of the addition of lactose on the degradation of ptsG mRNA was examined. The addition of lactose in the culture medium stimulated the degradation of ptsG mRNA in both the pgi mlcand pfkA mlc strains (Fig. 4, A and B,lane 3). Thus, internally produced glucose, which is believed to be phosphorylated by glucokinase encoded by glk(5Fraenkel D.G. Neidhardt F.C. 2nd Ed. Escherichia coli and Salmonella: Cellular and Molecular Biology. ASM Press, Washington, D. C.1996: 189-198Google Scholar), also leads to the destabilization of ptsG mRNA in the mutant strains. We also examined the effect of the addition of several glycolytic intermediates and/or carbon sources on theptsG mRNA degradation in pgi mlc andpfkA mlc double mutant strains. The addition of glucose-6-P, which is taken up by UhpT (24Ambudkar S.V. Anantharam V. Maloney P.C. J. Biol. Chem. 1990; 265: 12287-12292Abstract Full Text PDF PubMed Google Scholar), caused the rapid degradation ofptsG mRNA in both strains (Fig. 4, A andB, lane 8). On the other hand, the addition of fructose-6-P, which also is taken up by UhpT, failed to cause the rapid degradation of ptsG mRNA in the pgi mlcstrain (Fig. 4 A, lane 9), although it caused the degradation of ptsG mRNA in the pfkA mlcstrain (Fig. 4 B, lane 9). The addition of mannitol and xylose, which are taken up by mannitol-specific PTS and non-PTS xylose permease, respectively (25Lin E.C.C. Neidhardt F.C. 2nd Ed. Escherichia coli and Salmonella: Cellular and Molecular Biology. ASM Press, Washington, D. C.1996: 307-342Google Scholar), and then converted to fructose-6-P, also caused the rapid degradation of ptsGmRNA in the pfkA mlc (Fig. 4 B, lanes 4 and 5) but not in the pgi mlc (Fig.4 A, lanes 4 and 5) strains. However, the addition of fructose, which enters the Embden-Meyerhof pathway past fructose-6-P, did not affect the stability of ptsG mRNA in either strain (Fig. 4, A and B, lane 6). Likewise, the stability of ptsG mRNA in mutant strains was not affected by the addition of glycerol (Fig. 4,A and B, lane 7), which also enters the central carbon metabolism past fructose-1,6-P2. We conclude that the mutational block in the glycolytic pathway destabilizes ptsG mRNA depending on the accumulation of glycolytic intermediates up to the block points. Among glycolytic intermediates, glucose-6-P would be the most likely participant in the destabilization ofptsG mRNA because this intermediate is expected to accumulate at least in the pgi strain growing in the glucose medium. In fact, it is known that the glucose-6-P level in thepgi strain markedly increases when grown in the presence of glucose (26Levi B. Werman M.J. J. Nutr. Biochem. 2001; 12: 235-241Crossref PubMed Scopus (28) Google Scholar, 27Lee A.T. Cerami A. Proc. Natl. Acad. Sci. U. S. A. 1987; 84: 8311-8314Crossref PubMed Scopus (91) Google Scholar). To examine a link between the glucose-6-P level and the destabilization of ptsG mRNA, we determined the intracellular levels of glucose-6-P in various strains growing in LB medium containing exogenous glucose. The glucose-6-P level in thepgi strain was about 8 times higher than that in the wild-type strain (Table III, lines 1 and 2). A moderate but still significant accumulation of glucose-6-P was observed in the pfkA strain (Table III, line 3). On the other hand, little or no increase in the glucose-6-P level was observed in other mutant strains (Table III, lines 4–7). A large increase in the glucose-6-P level was observed in the mlc pgi but not in the mlc and the mlc pgi ptsI strains when cells were grown in LB medium with exogenous glucose (TableIV, lines 2, 4, and 8). As expected, the accumulation of glucose-6-P in the mlc pgi strain occurred even when cells were grown in LB medium without exogenous glucose because of trace amounts of glucose in the medium (Table IV, line 3). However, no increase in the glucose-6-P level was observed in thepgi strain when cells were grown in TB medium without exogenous glucose (Table IV, line 5). All of these results are entirely consistent with a view that accumulation of glucose-6-P is a degradation signal for the ptsG mRNA at least in thepgi strain.Table IIIIntracellular levels of glucose 6-phosphateStrainGenotypeGlucose 6-PmmW3110Wild type1.7 ± 0.4TM54Δpgi13.6 ± 1.1TM238pfkA3.4 ± 0.3TM84ΔtpiA2.0 ± 0.4TM61Δpgk1.7 ± 0.5TM81ΔgpmA ΔgpmB1.9 ± 0.6TM154ΔpykAΔpykF2.7 ± 0.8Cells were grown in LB medium containing 1% glucose. Intracellular levels of glucose 6-phosphate were determined as described under "Experimental Procedures." Each value is the average of three independent experiments. Open table in a new tab Table IVGlucose 6-phosphate levels in cells growing in various conditionsStrainGenotypeMediumAdditionGlucose 6-PmmIT1568mlcLBNone0.9 ± 0.11% glucose1.7 ± 0.4TM162mlcΔpgiLBNone9.3 ± 0.61% glucose11.7 ± 0.5TBNone1.5 ± 0.41% glucose9.3 ± 0.2TM145mlcΔpgi ptsILBNone1.9 ± 0.11% glucose2.0 ± 0.1TM109mlcΔpgi pfkA::Tn5TBNone1.0 ± 0.1TB0.1% glucose7.6 ± 0.7TB0.1% fructose 6-P1.5 ± 0.2W3110/pZw