This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental material
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Iida, A.
Right arrow Articles by Horinouchi, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Iida, A.
Right arrow Articles by Horinouchi, S.

 Previous Article  |  Next Article 

Journal of Bacteriology, July 2008, p. 5009-5019, Vol. 190, No. 14
0021-9193/08/$08.00+0     doi:10.1128/JB.00378-08
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

An OmpA Family Protein, a Target of the GinI/GinR Quorum-Sensing System in Gluconacetobacter intermedius, Controls Acetic Acid Fermentation{triangledown} ,{dagger}

Aya Iida, Yasuo Ohnishi, and Sueharu Horinouchi*

Department of Biotechnology, Graduate School of Agriculture and Life Sciences, The University of Tokyo, Bunkyo-ku, Tokyo 113-8657, Japan

Received 15 March 2008/ Accepted 2 May 2008


arrow
ABSTRACT
 
Via N-acylhomoserine lactones, the GinI/GinR quorum-sensing system in Gluconacetobacter intermedius NCI1051, a gram-negative acetic acid bacterium, represses acetic acid and gluconic acid fermentation. Two-dimensional polyacrylamide gel electrophoretic analysis of protein profiles of strain NCI1051 and ginI and ginR mutants identified a protein that was produced in response to the GinI/GinR regulatory system. Cloning and nucleotide sequencing of the gene encoding this protein revealed that it encoded an OmpA family protein, named GmpA. gmpA was a member of the gene cluster containing three adjacent homologous genes, gmpA to gmpC, the organization of which appeared to be unique to vinegar producers, including "Gluconacetobacter polyoxogenes." In addition, GmpA was unique among the OmpA family proteins in that its N-terminal membrane domain forming eight antiparallel transmembrane β-strands contained an extra sequence in one of the surface-exposed loops. Transcriptional analysis showed that only gmpA of the three adjacent gmp genes was activated by the GinI/GinR quorum-sensing system. However, gmpA was not controlled directly by GinR but was controlled by an 89-amino-acid protein, GinA, a target of this quorum-sensing system. A gmpA mutant grew more rapidly in the presence of 2% (vol/vol) ethanol and accumulated acetic acid and gluconic acid in greater final yields than strain NCI1051. Thus, GmpA plays a role in repressing oxidative fermentation, including acetic acid fermentation, which is unique to acetic acid bacteria and allows ATP synthesis via ethanol oxidation. Consistent with the involvement of gmpA in oxidative fermentation, its transcription was also enhanced by ethanol and acetic acid.


arrow
INTRODUCTION
 
Many gram-negative bacteria use N-acylhomoserine lactone (AHL)-dependent quorum-sensing systems to regulate gene expression in concert with cell density (6, 7). Quorum sensing is used to regulate diverse physiological functions and characteristics, including secondary metabolite production, swimming and swarming motility, conjugal plasmid transfer, biofilm formation, and virulence (25, 26). Two important proteins are involved in AHL-dependent quorum sensing: a LuxR-type protein, which is an AHL-dependent transcriptional regulator, and a LuxI-type protein, which directs the synthesis of AHLs from S-adenosylmethionine and diverse β-ketoacyl-acyl carrier protein intermediates in fatty acid biosynthesis. In general, LuxR-type proteins bind their cognate AHLs once the concentration of the AHL reaches a critical level, and the resulting complex activates the transcription of specific target genes by binding the promoter recognition sequences, termed lux boxes.

Acetic acid bacteria are gram-negative, obligatory aerobic bacteria with the ability to oxidize ethanol and sugars into the corresponding organic acids. Acetobacter and Gluconacetobacter species are used commercially to produce vinegar because of their high-level abilities to oxidize ethanol into acetic acid and their strong resistance to acetic acid and ethanol. Two membrane-bound enzymes, alcohol dehydrogenase and aldehyde dehydrogenase, catalyze the oxidation of ethanol. We previously found that several acetic acid bacteria, including Gluconacetobacter intermedius NCI1051, contain a pair of luxI and luxR homologues (10). The LuxI and LuxR homologues, named GinI and GinR, comprise a typical quorum-sensing system, in which GinI directs the synthesis of three AHLs with different acyl chains and GinR serves as a transcriptional regulator using the AHLs as its ligands to control the lux box-containing promoter of ginI. Downstream of ginI, a small protein consisting of 89 amino acids, named GinA, is encoded, and ginA and ginI are cotranscribed. GinA, showing no homology to any known proteins, represses (i) oxidative fermentation, including acetic acid fermentation and gluconic acid fermentation; (ii) growth in the presence of ethanol; and (iii) the antifoam activity of cells. These findings clearly show that the ginI/ginR quorum-sensing system in G. intermedius controls the phenotypes that are characteristic of acetic acid bacteria. Oxidative fermentation and growth in the presence of ethanol are essential for acetic acid bacteria to generate ATP via ethanol oxidation by means of acetic acid fermentation.

In order to reveal how GinA affects the phenotypes characteristic of acetic acid bacteria and what target genes other than ginA are under the control of the ginI/ginR regulatory system, we performed two-dimensional polyacrylamide gel electrophoresis (2D-PAGE) analysis for the comparison of protein profiles of the wild-type strain and mutants deficient in the ginI/ginR regulatory system. As a result, we found an OmpA family protein that was induced in response to the GinI/GinR quorum-sensing system. This OmpA family protein was found to repress oxidative fermentation, including acetic acid and gluconic acid fermentation. The genome sequences of several acetic acid bacteria have revealed the presence of multiple copies of ompA that appeared during evolution through gene duplication in this specific group of bacteria. In fact, in addition to two ompA genes, perhaps playing a structural role in the integrity of the surfaces of gram-negative bacteria by providing a physical linkage between the outer membrane and the underlying peptidoglycan layer, vinegar producers, such as G. intermedius and "Gluconacetobacter polyoxogenes," additionally contain triplicate ompA-like genes in a locus, one of which corresponded to the ompA-like gene that was regulated by the GinI/GinR system. In this report, we describe the identification of the ompA-like gene, named gmpA (Gluconacetobacter outer membrane protein A), as a target of the GinI/GinR quorum-sensing system via GinA and its negative effects on oxidative fermentation, including acetic acid and gluconic acid fermentation. GmpA, an outer membrane protein possibly specific to acetic acid bacteria and members of related genera, appears to play an important role in oxidative fermentation by this unique group of bacteria.


arrow
MATERIALS AND METHODS
 
Bacterial strains and plasmids. G. intermedius NCI1051 and G. polyoxogenes NCI1028 were obtained from stock cultures at the Central Research Institute, Mizkan Group Co., Ltd. An Acetobacter/Gluconacetobacter-Escherichia coli shuttle vector, pMV24, was described previously (5). A ginI-disrupted mutant (ginI::Km), a ginR-disrupted mutant (ginR::Km), and a ginA-disrupted mutant (ginA::Km) were described previously (10). pGinR, pGinI, pGinIA, and pGinA, all of which were pMV24-derived plasmids, contained the intact ginR, ginI, ginI and ginA, and ginA sequences, as described previously (10). pMGinA contained a ginA sequence with a frameshift mutation (10). E. coli JM109 and plasmid pUC19, used for DNA manipulation, were purchased from Takara Bio. Agrobacterium tumefaciens NTL4(pZLR4), obtained from S. K. Farrand, was used to assay AHL production.

Media and culture conditions. YPG medium (pH 6.5) consisted of 5 g of yeast extract (Wako Pure Chemicals), 3 g of polypeptone (Wako Pure Chemicals), and 30 g of glucose in 1 liter of water. The Gluconacetobacter strains were grown at 30°C in YPG medium with 1% (vol/vol) cellulase (Celluclast 1.5L; Novozymes) and with or without 2% (vol/vol) ethanol.

For acetic acid fermentation tests, the Gluconacetobacter strains were first cultured at 30°C for 24 h in 5 ml of YPG medium with 1% (vol/vol) cellulase in a test tube with shaking. A portion (5 ml) of this culture was inoculated into 1.5 liters of YPG medium supplemented with 3% (vol/vol) ethanol, 1% (vol/vol) cellulase, and 0.001% (vol/vol) silicone (KM72; Shin-Etsu Chemical Co., Ltd.) in a 3-liter mini-jar fermentor (Bioneer 300 3L; B. E. Marubishi Co., Ltd.) and cultured at 30°C with agitation at 500 rpm and aeration at a rate of 1.0 liter/min. The ethanol concentration was automatically maintained at 2% (vol/vol) by the addition of ethanol during cultivation. The acetic acid and gluconic acid concentrations in the culture broths were determined by high-performance liquid chromatography with a Shodex RSpak KC-811 column and a Shimadzu CDD-10A conductivity detector. Bacterial growth was monitored by measuring the turbidities of the cultures at 660 nm by a photometer.

E. coli was cultured at 37°C in Luria-Bertani medium. A. tumefaciens NTL4(pZLR4) was cultured at 30°C in AB medium containing 0.2% (wt/vol) glucose, 0.1% (wt/vol) yeast extract, and 5 µg of gentamicin/ml (15). Ampicillin and kanamycin were used at a final concentration of 100 µg/ml, when necessary to maintain the plasmids.

DNA manipulation. Restriction enzymes, T4 DNA ligase, and other DNA-modifying enzymes were purchased from Takara Bio. All DNA manipulations in E. coli were performed as described previously (1, 16). The Gluconacetobacter strains were transformed by electroporation (27). Chromosomal DNA from Gluconacetobacter strains was isolated with the genomicPrep cell and tissue DNA isolation kit (GE Healthcare). Chromosomal DNA from G. intermedius NCI1051 and G. polyoxogenes was used as the templates for PCR. All nucleotide sequences were determined with a CEQ dye terminator cycle sequencer using a quick-start kit (Beckman Coulter).

2D-PAGE. Gluconacetobacter strains were grown at 30°C for 12 h in YPG medium containing 1% (vol/vol) cellulase and 2% (vol/vol) ethanol in a shaking flask. Cells were washed with 10 mM Tris buffer (pH 7.5), resuspended in the same buffer, and then disrupted twice using a French pressure cell at 20,000 lb/in2. The cell lysates were centrifuged at 8,000 x g and 4°C for 10 min, and the supernatants were further ultracentrifuged at 370,000 x g for 1 h. The supernatants were used as soluble fractions. The pellets were resuspended in ReadyPrep Reagent 3 (Bio-Rad Laboratories), and the suspensions were used as membrane fractions. 2D-PAGE was carried out using the PROTEAN isoelectric focusing cell (Bio-Rad Laboratories) with immobilized pH gradients (precast IPG ReadyStrip gel, pH 3 to 10; 11 cm) in the first dimension and sodium dodecyl sulfate-12.5% polyacrylamide gel in the second dimension, according to the instructions of the manufacturer. The protein concentrations were determined with the DC protein assay kit (Bio-Rad Laboratories) using bovine serum albumin as a standard.

N-terminal amino acid sequencing. After 2D-PAGE, the proteins were blotted onto a polyvinylidene difluoride membrane (Millipore) with a semidry blotting system (HorizBlot; ATTO Corporation) and analyzed by Edman degradation on an Applied Biosystems model 492cLC protein sequencer.

Cloning of gmpA, gmpB, and gmpC. A 0.4-kb fragment containing a portion of the GP1736 gene sequence from G. polyoxogenes NCI1028 was amplified by PCR with primers AF and AR (Table 1). The amplified fragment was used for the 32P-labeled probe for Southern hybridization against the restriction enzyme-digested chromosomal DNA from G. intermedius NCI1051. Standard DNA manipulation, including colony hybridization using this 0.4-kb fragment as the probe, gave three DNA fragments, a 3.4-kb SmaI fragment, a 4.0-kb SphI fragment, and a 4.2-kb EcoRV fragment, which revealed the presence of three ompA-like genes, gmpA to gmpC.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Primers used in this study

Disruption of gmpA. The 3.4-kb SmaI fragment cloned in pUC19 was digested by NruI. A SmaI fragment carrying the neomycin-kanamycin resistance gene from Tn5 (2) was then inserted into the NruI site within the gmpA coding sequence. This plasmid was introduced into G. intermedius NCI1051 by electroporation, and kanamycin-resistant colonies were selected as candidates for gmpA-disrupted mutants. Correct gmpA-disrupted mutants were selected by Southern hybridization with a 0.4-kb fragment that had been amplified by PCR with primers AF and AR (Table 1) as the 32P-labeled probe.

Plasmid construction. For the construction of a plasmid for the expression of gmpA alone, a 1.8-kb fragment containing gmpA was amplified by PCR with primers P1F, containing an EcoRI site, and P1R, containing a SmaI site (Table 1), and placed between the EcoRI and SmaI sites of pMV24, generating plasmid pGmpA. Nucleotide sequencing confirmed that no errors occurred during PCR.

For the construction of a mutagenic plasmid, a 0.5-kb fragment containing an upstream region of gmpA was amplified by PCR using primers P1F and MR (Table 1), and a 1.2-kb fragment containing a downstream region of gmpA was amplified by PCR using primers MF and P1R (Table 1). A 1.7-kb fragment was amplified by splice overlap extension-PCR (9) with primers P1F and P1R by using the 0.5-kb fragment and the 1.2-kb fragment as the templates, and the product was placed between the EcoRI and SmaI sites of pMV24, generating plasmid pGmpA{Delta}. Nucleotide sequencing confirmed that no errors occurred during PCR.

S1 nuclease mapping. Total RNA was isolated by the hot-phenol method (20) from cells grown at 30°C for 8 h in YPG medium containing 2% (vol/vol) ethanol and 1% (vol/vol) cellulase in a shaking flask. S1 nuclease mapping was conducted as described previously (12). The hybridization probes were prepared by PCR using pairs of 32P-labeled and nonlabeled primers. The primer pairs were as follows: S1-AF and S1-AR for gmpA, S1-BF and S1-BR for gmpB, S1-CF and S1-CR for gmpC, and S1-IF and S1-IR for ginI (Table 1). Primers S1-AR, S1-BR, S1-CR, and S1-IR were labeled with 32P at their 5' ends by using T4 polynucleotide kinase before PCR. All sequencing reactions were performed using a BcaBEST dideoxy sequencing kit (Takara Bio).

Northern blot analysis. Total RNA was isolated by the hot-phenol method (20) from cells grown at 30°C for 6, 8, 10, 12, 16, and 24 h in YPG medium containing 2% (vol/vol) ethanol and 1% (vol/vol) cellulase in a shaking flask. A 0.4-kb fragment containing part of gmpA was amplified by PCR with primers AF and AR (Table 1) using the 3.4-kb SmaI fragment as the template. The PCR product was verified by nucleotide sequencing. The amplified fragment was used for the 32P-labeled probe for Northern hybridization against the total RNA. The gmpA probe specifically bound the gmpA coding sequence, as determined by Southern hybridization (data not shown).

Reverse transcription (RT)-PCR. Total RNA was isolated by the hot-phenol method (20) from cells grown at 30°C for 8 h in YPG medium containing 2% (vol/vol) ethanol and 1% (vol/vol) cellulase in a shaking flask. RT was performed using the total RNA with SuperScript III reverse transcriptase (Invitrogen). The cDNA was then amplified by PCR using the RT reaction mixture as the template and specific primers (Table 1). The RNA samples were tested with and without reverse transcriptase to verify the complete removal of genomic DNA. PCR using the genomic DNA from G. intermedius NCI1051 as the template was performed to ensure the fidelity for each primer pair.

AHL assay. The Gluconacetobacter strains were grown at 30°C for 24 h in YPG medium containing 1% (vol/vol) cellulase and 2% (vol/vol) ethanol. Cells were removed by centrifugation, and AHLs were extracted from culture supernatants with acidified ethyl acetate, as described previously (23). The AHL production was evaluated by measuring the β-galactosidase activity of an AHL indicator strain, A. tumefaciens NTL4(pZLR4), as described by Luo et al. (15).

Nucleotide sequence accession number. The nucleotide sequence of gmpA to gmpC has been deposited in the DDBJ, EMBL, and GenBank databases under accession number AB426240.


arrow
RESULTS
 
Comparison of protein profiles of the wild-type strain NCI1051, mutant ginI::Km, and mutant ginR::Km. As the first step to identify targets of the GinI/GinR quorum-sensing system in G. intermedius NCI1051, we employed 2D-PAGE to analyze soluble and membrane proteins isolated from wild-type strain NCI1051, the ginI::Km mutant, and the ginR::Km mutant, all of which harbored pMV24 and were grown at 30°C for 12 h in YPG medium containing 2% (vol/vol) ethanol and 1% (vol/vol) cellulase. The comparison of the patterns of membrane protein spots on the 2D gels revealed that the levels of a protein with an observed molecular mass of 35 kDa, named M1, in the ginI::Km and ginR::Km mutants were apparently decreased (Fig. 1B and D) compared to that in the wild-type strain NCI1051 (Fig. 1A). The introduction of the intact ginIA gene on pGinIA into the ginI::Km mutant restored the expression of protein M1 to the wild-type level (Fig. 1C). Similarly, the introduction of ginR on pGinR into the ginR::Km mutant also restored the wild-type expression of protein M1 (Fig. 1E). The N-terminal amino acid sequence of protein M1 was determined by Edman degradation to be TTITGPYVGIGGG. BLAST analysis of the genome sequence of G. polyoxogenes NCI1028, which was determined by a research group at Mizkan Group Co., Ltd. (K. Kondo, personal communication), predicted that one or more of the three products (the GP1736, GP1737, and GP1740 proteins) encoded by three adjacent homologous genes that are clustered at a locus was protein M1. All three proteins, belonging to the OmpA family, contain N-terminal amino acid sequences almost the same as the established sequence of protein M1. Although two additional genes encoding OmpA family proteins exist far from the nested ompA-like gene locus on the chromosome of G. polyoxoxgenes NCI1028, the sequences of the two proteins show no significant similarity to the N-terminal amino acid sequence of M1. As described below, protein M1 was found to be GmpA, because mutant gmpA::Km lacked protein M1 and the introduction of pGmpA into this mutant restored the production of M1.


Figure 1
View larger version (32K):
[in this window]
[in a new window]

  		FIG. 1. 2D-gel images of the membrane fractions prepared from the wild-type (w.t.) strain NCI1051 harboring pMV24 and the ginR::Km mutant strain harboring pMV24 (upper panels) and parts of the 2D gels including GmpA (A to E) and corresponding to the gmpA mutant strains (F to H). The membrane fractions from G. intermedius strains, which were grown at 30°C for 12 h in YPG medium containing ethanol, were prepared as described in Materials and Methods. These proteins were analyzed by 2D-PAGE and stained with Coomassie brilliant blue. The triangles indicate proteins M1 (black), M2 (white), and M3 (gray).

Cloning of gmpA to gmpC. We prepared 32P-labeled probe on the basis of the nucleotide sequence of the GP1736 gene in G. polyoxogenes NCI1028 (Kondo, personal communication). The probe consisted of 0.4 kb corresponding to Leu216 through Gly362 of the GP1736 protein. Southern hybridization with this probe against SmaI-, SphI-, or EcoRV-digested chromosomal DNA from G. intermedius NCI1051 revealed the presence of nucleotide sequences homologous to the probe in strain NCI1051 (data not shown). We used the same probe to clone a 3.4-kb SmaI fragment, a 4.0-kb SphI fragment, and a 4.2-kb EcoRV fragment from G. intermedius NCI1051, all giving a positive signal, by standard DNA manipulation methods including colony hybridization. All three fragments were cloned in pUC19.

The nucleotide sequences of the cloned fragments and the assembly of the fragments indicated the presence of three adjacent homologous open reading frames (ORFs) oriented in the same direction (Fig. 2A), as was found for G. polyoxogenes. We named these three ORFs gmpA, gmpB, and gmpC, because the deduced amino acid sequences of the C-terminal regions of GmpA, GmpB, and GmpC encoded by these ORFs showed similarity to the so-called OmpA motif of E. coli OmpA, the major outer membrane protein (Fig. 2B). PSORTb analysis (http://www.psort.org/psortb/) of GmpA, GmpB, and GmpC predicted that they were outer membrane proteins with a putative 22-amino-acid signal peptide (the most likely cleavage site being between amino acids Ala22 and Thr23) (Fig. 2B), as predicted by SignalP (http://www.cbs.dtu.dk/services/SignalP/). The amino acid sequence from Thr24 to Gly35 (Fig. 2B) just after this cleavage site in GmpA and GmpB completely matched the amino acid sequence of protein M1 determined by N-terminal amino acid sequencing, but that of GmpC did not match. As described below, protein M1 corresponded to GmpA, as determined by its response to the GinI/GinR regulatory system.


Figure 2
View larger version (76K):
[in this window]
[in a new window]

  		FIG. 2. Gene organization of the gmpA-to-gmpC locus in G. intermedius, G. polyoxogenes, and Gluconobacter oxydans (A) and amino acid alignment of the OmpA family proteins (B). (A) The organization of gmpA to gmpC and their neighbors in G. intermedius NCI1051 and of the corresponding regions in G. polyoxogenes NCI1028 and Gluconobacter oxydans 621H is illustrated. gmpA to gmpC are located between orf1 (encoding a putative bifunctional shikimate kinase) and orf2 (encoding a putative alanyl-tRNA synthetase) in G. intermedius. The GP1736, GP1737, and GP1740 genes, encoding OmpA family proteins in G. polyoxogenes, are located between the GP1735 gene, encoding a bifunctional shikimate kinase, and the GP1743 gene, encoding an alanyl-tRNA synthetase. The GOX1787 gene (GenBank gene identification no. 3250051), encoding an OmpA family protein in Gluconobacter oxydans, is located between the GOX1788 gene (identification no. 3250052), encoding a bifunctional shikimate kinase, and alaS (identification no. 3250050), encoding an alanyl-tRNA synthetase. (B) Alignment of the primary sequences of OmpA family proteins. These include GmpA, GmpB, and GmpC from G. intermedius NCI1051; the products of GOX1787 (GenBank accession no. YP_192182) and GOX1260 (GenBank accession no. YP_191677) from Gluconobacter oxydans 621H; the product of Gb0359 (GenBank accession no. YP_744180) from Granulibacter bethesdensis CGDNIH1; the products of Acry_0174 (GenBank accession no. YP_001233321) and Acry_0173 (GenBank accession no. YP_001233320) from Acidiphilium cryptum JF-5; and OmpA (corresponding to GenBank accession no. NP_415477) from E. coli K-12. These amino acid sequences were aligned with ClustalW (http://www.ebi.ac.uk/Tools/clustalw/) and presented by BOXSHADE 3.21 (http://www.ch.embnet.org/software/BOX_form.html). The putative signal peptide comprising Met1 to Ala22 is indicated by a line. The amino acid sequence of GmpA, which matched the N-terminal sequence of protein M1, is indicated by a dashed line. The extra region of GmpA, which is absent from GmpB and GmpC, is indicated by double lines. The OmpA motif is indicated by an arrow. The β-strands (β1 to β8, shown in red) and loops (L1 to L4) of the transmembrane domain of E. coli OmpA, as determined by X-ray crystallography (21), are indicated. The β-strands in GmpA to GmpC, as predicted by Jpred (http://www.compbio.dundee.ac.uk/~www-jpred/), are also shown in red. Part of the fourth β-strand region of GmpA, corresponding to β4 in E. coli OmpA, was predicted to contain an {alpha}-helix (shown in green). Identical amino acids are shaded in black, and similar amino acids are shaded in gray.

GmpA (405 amino acids), GmpB (373 amino acids), and GmpC (364 amino acids) are highly homologous to one another, especially in their C termini containing the OmpA motif. The OmpA motif is believed to interact specifically with the peptidoglycan layer (14). The N-terminal regions of GmpA, GmpB, and GmpC all contain eight β-stranded transmembrane sequences, β1 to β8 (Fig. 2B), as predicted by Jpred (http://www.compbio.dundee.ac.uk/~www-jpred/). The prediction of the structures of these proteins by FUGUE (http://tardis.nibio.go.jp/fugue/prfsearch.html) gave almost the same secondary structures as the prediction by Jpred. Referring to the crystal structure of E. coli OmpA (14, 21), we assume that the N-terminal domains serve as a membrane anchor consisting of eight antiparallel β-barrels, with the N-terminal and C-terminal ends in the periplasm. The spacing patterns of the eight β-strands of E. coli OmpA and the Gmp proteins are almost identical, except that part of the β4 region of GmpA was predicted to contain both a β-strand and an {alpha}-helix (Fig. 2B). GmpA contains an extra hydrophilic sequence in loop 1 in the N terminus compared to GmpB and GmpC (Fig. 2B). Loop 1 is exposed to the surface. As described below, this extra amino acid sequence was found only in vinegar producers and was essential for the function of GmpA.

It should be noted that the primary amino acid sequences of GmpA to GmpC show only 32% identity to that of OmpA from E. coli (corresponding to DNA database accession no. NP_415477) and have no significant similarity to OmpA in their N termini, compared to the level of similarity of their C termini containing the OmpA motif. However, GmpA to GmpC are homologous to the OmpA family proteins found in acetic acid bacteria. GmpA to GmpC show 60 to 64% identity in amino acid sequence to the OmpA family GOX1787 protein (corresponding to GenBank accession no. YP_192182) from Gluconobacter oxydans 621H (22) and about 45 to 48% identity to the OmpA family proteins encoded by Gb0359 (GenBank accession no. YP_744180) from Granulibacter bethesdensis CGDNIH1 (8), GOX1260 (GenBank accession no. YP_191677) from Gluconobacter oxydans 621H (22), and Acry_0174 (GenBank accession no. YP_001233321) and Acry_0173 (GenBank accession no. YP_001233320) from Acidiphilium cryptum JF-5. The OmpA family proteins are grouped into three subfamilies (see Fig. S1A in the supplemental material): (i) so-called OmpA-like proteins containing β-barrels as an outer membrane anchor in their N termini; (ii) rather small proteins, with fewer than 250 amino acid residues, containing no transmembrane domains; and (iii) functionally unknown proteins containing {alpha}-helices instead of β-barrels as a membrane anchor. Figure S1B in the supplemental material is a phylogenetic tree, constructed by a neighbor-joining method, for the entire sequences of the OmpA family proteins and delineates the three classes. Figure S1C in the supplemental material is a phylogenetic tree for the OmpA-like proteins containing β-barrels as a membrane anchor. The evolutionary analysis of the OmpA family proteins gave us no hints to predict or speculate on the function of GmpA, although a GmpA-containing group of proteins distant from the E. coli OmpA (see Fig. S1C in the supplemental material) may have evolved to exert some specific functions in a particular species of bacteria to inhabit a given niche.

In addition to the rather high level of homology in amino acid sequence among the OmpA family proteins, the numbers and organization of the ompA family genes are unique to acetic acid bacteria, especially to those that produce acetic acid at high yields. Among the bacteria, including acetic acid bacteria, whose genome sequences are available, only G. intermedius NCI1051 and G. polyoxogenes NCI1028 (Kondo, personal communication), both of which were isolated from vinegar, contain three adjacent omp genes, like gmpA to gmpC, as a gene cluster. Other Acetobacteraceae, such as Gluconobacter oxydans 621H, Granulibacter bethesdensis CGDNIH1, and Acidiphilium cryptum JF-5 (Fig. 2B), contain one or two ompA-like genes that encode GmpB- and GmpC-type proteins without the N-terminal extra sequence in GmpA. The genes located upstream and downstream of gmpA to gmpC in G. intermedius NCI1051, as well as in G. polyoxogenes NCI1028, and the GOX1787 gene, encoding an OmpA family protein in Gluconobacter oxydans 621H, are conserved (Fig. 2A), which suggests that the duplication of an ancestral omp gene, GOX1787, has occurred twice in strains NCI1051 and NCI1028. Because the GOX1787 protein contains no extra N-terminal sequence in loop 1, gmpA probably acquired the extra sequence during the duplication.

Dependence of gmpA transcription on ginI and ginR. The above-described 2D-PAGE analysis suggested that either gmpA or gmpB or both were under the control of the GinI/GinR quorum-sensing system. We first determined the transcriptional organization of the gmp genes. RNA was isolated by the hot-phenol method from cells grown at 30°C for 8 h in YPG medium containing 1% (vol/vol) cellulase and 2% (vol/vol) ethanol. High-resolution S1 nuclease mapping identified a single transcriptional start point for each of the gmpA, gmpB, and gmpC genes (Fig. 3A), indicating that each gene was transcribed from its own promoter (Fig. 3B). The transcriptional start point of gmpA was, for example, 174 bp upstream of the first letter of the translational start codon. In front of their transcriptional start points, neither probable promoter elements (5'-TATAAT-3' for –10 and 5'-TTGACA-3' for –35) nor a lux box, a 20-bp inverted repeat at approximate position –45 with respect to the transcriptional start point (7, 10), was found.


Figure 3
View larger version (38K):
[in this window]
[in a new window]

  		FIG. 3. Transcriptional analysis of gmpA, gmpB, and gmpC. (A) Determination of the transcriptional start points of gmpA, gmpB, and gmpC by high-resolution S1 nuclease mapping. The resultant DNA fragments in the S1 mapping (left) were analyzed by PAGE with the sequencing ladders (A, C, G, and T) of the probes. The transcriptional start points are indicated by arrows. (B) Nucleotide sequences upstream of the transcriptional start points (+1; corresponding letters are in bold) of gmpA to gmpC. No promoter elements similar to the –35 (5'-TTGACA-3') and –10 (5'-TATAAT-3') elements are present. M, start codon. (C) Agarose gel electrophoresis analysis of RT-PCR products. RT-PCR was performed using total RNA from the wild-type strain NCI1051 (w.t.) and the ginI::Km mutant. The RNA samples were analyzed with (+RT) and without (–RT) reverse transcriptase to verify the absence of genomic DNA. Lane P was a control lane in which the genomic DNA of G. intermedius NCI1051 was amplified with the respective pair of primers. The combinations of the primers, F0 and R1, for example, are given in the form F0R1. The positions of gmpA to gmpC and PCR products are also shown schematically. (D) Transcription of gmpA, gmpB, and gmpC in wild-type strain NCI1051 harboring pMV24, the ginI::Km mutant harboring pMV24, the ginI::Km mutant harboring pGinIA, the ginR::Km mutant harboring pMV24, and the ginR::Km mutant harboring pGinR, as determined by low-resolution S1 mapping. (E) Northern blot analysis of gmpA in wild-type strain NCI1051 harboring pMV24 (lane 1), the ginI::Km mutant harboring pMV24 (lane 2), the ginI::Km mutant harboring pGinIA (lane 3), the ginR::Km mutant harboring pMV24 (lane 4), and the ginR::Km mutant harboring pGinR (lane 5).

RT-PCR analysis to detect read-through transcription using cDNA synthesized by RT from the RNA prepared from the wild-type strain gave a distinct amplified fragment for each gene: fragment F1R1, with 309 bp, synthesized with primers F1 and R1 for gmpA; fragment F2R2, with 349 bp, for gmpB; and fragment F3R3, with 328 bp, for gmpC (Fig. 3C). A gmpA-gmpB-spanning region was also amplified by RT-PCR using primers F1 and R2, yielding fragment F1R2, with 1,866 bp, which indicated the occurrence of read-through transcription from the gmpA promoter into gmpB. In contrast, a gmpB-gmpC-spanning region was not amplified by RT-PCR using primers F2 and R3. The RT-PCR analysis was also performed with the RNA prepared from mutant ginI::Km to show the dependence of the gmpA promoter on the GinI/GinR regulatory system (see below). These high-resolution S1 mapping and RT-PCR analyses showed that each of the gmpA, gmpB, and gmpC genes had its own single promoter, although there was transcriptional read-through from gmpA to gmpB but no read-through from gmpB to gmpC. Despite the finding that gmpA transcription is under the control of the GinI/GinR regulatory system, no lux box-like sequence was present in the promoter regions of gmpA to gmpC. The absence of a lux box is consistent with the fact that gmpA is not directly controlled by GinR but is controlled via GinA (see below).

We next assessed the transcription of gmpA to gmpC in the wild-type strain NCI1051, the ginI::Km mutant, and the ginR::Km mutant, all of which harbored the vector plasmid pMV24, by low-resolution S1 nuclease mapping (Fig. 3D). RNA was isolated by the hot-phenol method from cells grown at 30°C for 8 h in YPG medium containing 1% (vol/vol) cellulase and 2% (vol/vol) ethanol. rRNA was used to check the amount of RNA used. gmpA in strain NCI1051 was actively transcribed, but the levels of transcription in the ginI::Km and ginR::Km mutants were markedly decreased (Fig. 3D). The introduction of pGinIA and pGinR into the ginI::Km and ginR::Km mutants, respectively, restored gmpA transcription to the wild-type level (Fig. 3D), showing that GinI and GinR were essential for gmpA transcription. On the other hand, gmpB and gmpC in the ginI::Km and ginR::Km mutants were transcribed at the same levels as those in the parental strain. Northern blot analysis of gmpA by using RNA isolated from cells grown for 6 to 24 h confirmed the results of the S1 mapping; the intact ginIA and ginR genes on the vector pMV24 complemented the ginI and ginR mutations and restored the wild-type level of gmpA transcription in the mutants (Fig. 3E). The major signal of 1.5 kb represented the gmpA transcript starting from the gmpA promoter, and a minor signal of 3.5 kb represented the read-through transcript from the gmpA promoter into gmpB. The RT-PCR analysis with the RNA from mutant ginI::Km (Fig. 3C) supported these results; only the gmpA promoter depended on the GinI/GinR regulatory system, and there was read-through from the gmpA promoter to gmpB. From these data, we concluded that each of the genes gmpA, gmpB, and gmpC had its own transcription start site and that only the gmpA promoter was positively regulated by the GinI/GinR quorum-sensing system.

These transcriptional analyses suggested that GmpA corresponded to protein M1, which was produced in response to the GinI/GinR quorum-sensing system, as observed by 2D-PAGE (Fig. 1). An additional spot, M2, which was about 5 kDa smaller than M1, was observed at almost the same levels on the 2D gels for all strains (Fig. 1). The N-terminal amino acid sequence of protein M2 also matched the sequences of GmpA and GmpB. Because the calculated molecular mass of GmpB was about 5 kDa smaller than that of GmpA, protein M2 presumably corresponds to GmpB.

Dependence of gmpA transcription on ginA. The absence of any lux box-like sequence in the gmpA promoter excluded the possibility that GinR directly controlled gmpA transcription. Our previous study showed that GinA as a target of the GinI/GinR regulatory system plays an important role in the control of oxidative fermentation, growth in medium containing ethanol, and antifoam activity (10). To determine whether ginA was involved in the control of gmpA transcription, we introduced ginA on pGinA into the ginI::Km mutant and examined gmpA transcription in this mutant by S1 nuclease mapping. ginA on pGinA was transcribed from a heterologous promoter, the E. coli lac promoter on pMV24. The decreased gmpA transcription in the ginI::Km mutant was restored to the wild-type level by the introduction of ginA on pGinA (Fig. 4). The ginI mutation caused no effects on the transcription of gmpB or gmpC, as described above. We next introduced a frameshift mutation into the ginA gene on pGinA, generating pMGinA as described previously (10). The introduction of pMGinA into the ginI::Km mutant gave no restoration of the phenotype of the mutant (Fig. 4), indicating that the GinA protein function was essential for restoring the gmpA transcription in the ginI::Km mutant.


Figure 4
View larger version (34K):
[in this window]
[in a new window]

  		FIG. 4. Dependence of gmpA transcription on ginA. S1 nuclease mapping of gmpA to gmpC in wild-type (w.t.) strain NCI1051 harboring pMV24, the ginI::Km mutant harboring pMV24, the ginI::Km mutant harboring pGinA, the ginI::Km mutant harboring pMGinA, the ginA::Km mutant harboring pMV24, and the ginA::Km mutant harboring pGinA is shown.

To clarify the involvement of ginA in gmpA transcription, we examined gmpA transcription in the ginA::Km mutant. The gmpA transcription in this mutant was greatly reduced (Fig. 4). No such decrease was observed for gmpB or gmpC. The introduction of pGinA into the ginA::Km mutant restored the wild-type level of gmpA transcription (Fig. 4). These data further confirmed that GinA was essential for gmpA transcription. We checked that the introduction of ginA on pGinA into the ginI::Km mutant or the loss of ginA had no significant effects on AHL production, as determined by using the A. tumefaciens reporter system (data not shown). We thus concluded that gmpA transcription was positively regulated by GinA, but in a still unknown manner.

Growth repression by gmpA in the presence of ethanol. To determine the involvement of gmpA in the repression of oxidative fermentation, including acetic acid and gluconic acid fermentation, growth in the presence of ethanol, and the antifoam activity of cells, we disrupted the chromosomal gmpA gene in G. intermedius NCI1051, generating mutant gmpA::Km. The gmpA::Km mutant grew in medium containing no ethanol with the same time course and cell mass as the parental strain (Fig. 5A). In YPG medium containing 2% (vol/vol) ethanol, however, the growth rate and cell mass of mutant gmpA::Km during the exponential growth phase were higher than those of the parental strain (Fig. 5B). This growth profile of mutant gmpA::Km was the same as those of mutants ginI::Km, ginR::Km, and ginA::Km (10). The higher growth rate of the mutant was due to the mutation in gmpA, because the introduction of pGmpA into the mutant decreased the growth rate to the same level as that of the parental strain (Fig. 5B).


Figure 5
View larger version (10K):
[in this window]
[in a new window]

  		FIG. 5. Growth of G. intermedius strains in YPG medium (A) and YPG medium containing 2% (vol/vol) ethanol (B). Wild-type strain NCI1051 harboring pMV24 (•), the gmpA::Km mutant harboring pMV24 ({square}), the gmpA::Km mutant harboring pGmpA ({triangleup}), and the gmpA::Km mutant harboring pGmpA{Delta} ({blacklozenge}) were cultured as described in Materials and Methods. The inset shows an enlarged profile of growth from 6 to 14 h. OD660, optical density at 660 nm.

The importance of the extra region in the N-terminal portion of GmpA was confirmed by constructing plasmid pGmpA{Delta}, in which the sequence encoding the extra region was deleted, and introducing this plasmid into mutant gmpA::Km. The production and localization of the N-terminal mutant GmpA protein GmpA{Delta}, in which the extra region was deleted, were evaluated by 2D-PAGE analysis of the membrane fraction of mutant gmpA::Km harboring pGmpA{Delta} after this mutant was grown at 30°C for 12 h in YPG medium containing 2% (vol/vol) ethanol and 1% (vol/vol) cellulase (Fig. 1F to H). The comparison of the 2D-PAGE patterns of the membrane protein spots revealed that protein M1 (GmpA) disappeared in the gmpA::Km mutant harboring pMV24 (Fig. 1F) but that the introduction of the intact gmpA gene on pGmpA into the gmpA::Km mutant restored the production of GmpA (Fig. 1G). A spot, M3, which was about 2 kDa smaller than GmpA (calculated molecular mass, 41.4 kDa; pI 6.05) and presumably represented the GmpA{Delta} protein (39.3 kDa; pI 5.99), was observed in the membrane fraction of the gmpA::Km mutant harboring pGmpA{Delta} (Fig. 1H). These data showed that the GmpA{Delta} protein was produced and localized in the membrane. The growth rate of mutant gmpA::Km harboring pGmpA{Delta} in medium containing 2% (vol/vol) ethanol was the same as that of the gmpA::Km mutant harboring the vector pMV24 (Fig. 5B). Thus, the extra region in the N terminus of GmpA was important for the protein to exert its function of repressing the growth of strain NCI1051 in the presence of ethanol.

Our previous study showed that GinI/GinR quorum sensing represses antifoam activity; mutants ginI::Km, ginR::Km, and ginA::Km form foams of smaller volumes on the culture broth than those formed by the wild-type strain (10). However, the gmpA::Km mutant formed foams of almost the same volume as those formed by the wild-type strain NCI1051 (data not shown). Therefore, gmpA had no effects on the antifoam activity.

Repression of oxidative fermentation by gmpA. We next measured acetic acid and gluconic acid production by mutant gmpA::Km. The strain was cultured at 30°C in YPG medium containing 3% (vol/vol) ethanol in a 3-liter mini-jar fermentor. Figure 6A and B show the acetic acid and gluconic acid fermentation profiles, respectively, of the wild-type strain NCI1051, the ginI::Km mutant, and the gmpA::Km mutant, all of which harbored the vector pMV24. Like mutant ginI::Km, the gmpA::Km mutant accumulated acetic acid and gluconic acid at higher yields than the parental strain NCI1051. The final yields of acetic acid (mean ± standard deviation, 4.17% ± 0.07% [wt/vol]) and gluconic acid (2.26% ± 0.03%) produced by mutant gmpA::Km were also higher than those (acetic acid, 2.78% ± 0.11%; gluconic acid, 1.89% ± 0.08%) produced by the parental strain NCI1051. The increased production rates and final yields of acetic acid and gluconic acid in the gmpA::Km mutant were restored to almost the wild-type levels by introducing gmpA on pGmpA into this mutant (data not shown).


Figure 6
View larger version (10K):
[in this window]
[in a new window]

  		FIG. 6. Courses of acetic acid (A) and gluconic acid (B) fermentation by G. intermedius strains, together with their growth (C). The growth of the strains was measured by monitoring the optical densities at 660 nm (OD660) of the culture broths. Wild-type strain NCI1051 harboring pMV24 (•), the ginI::Km mutant harboring pMV24 ({blacktriangleup}), the gmpA::Km mutant harboring pMV24 ({square}), and wild-type strain NCI1051 harboring pGmpA ({diamond}) were cultured as described in Materials and Methods.

Consistent with the observation that the gmpA mutation led to the production of acetic acid and gluconic acid at high yields, the introduction of gmpA under the control of both its own promoter and the E. coli lac promoter on the pMV24-derived plasmid pGmpA into the wild-type strain NCI1051 reduced the final yields of acetic acid (2.52% ± 0.03% [wt/vol]) and gluconic acid (1.70% ± 0.02%) below those of the parental strain (Fig. 6A and B). The copy number of pMV24 is about 10 (10). These findings suggested that GmpA, as one of the targets of the GinI/GinR quorum-sensing system in G. intermedius NCI1051, repressed oxidative fermentation, including acetic acid and gluconic acid fermentation, and played an important role in the regulation of oxidative fermentation.

Mutant gmpA::Km continued to produce acetic acid in the stationary phase, although the production of acetic acid by the ginI::Km and ginA::Km mutants ceased in the stationary phase (Fig. 6A) (10). Therefore, the phenotypes of the gmpA::Km mutant were not completely the same as those of the ginI::Km and ginA::Km mutants. This finding raised a possibility that some other target gene(s) of GinA was present, affecting these phenotypes in some unknown way.

Dependence of gmpA transcription on ethanol and acetic acid. Because gmpA affected acetic acid fermentation, we expected that ethanol and acetic acid would affect gmpA transcription. We examined gmpA transcription in the wild-type strain NCI1051 harboring pMV24 by low-resolution S1 nuclease mapping (Fig. 7A). RNA was isolated by the hot-phenol method from cells grown to early exponential, late exponential, and stationary phases at 30°C in YPG medium, YPG medium containing 2% (vol/vol) ethanol, or YPG medium containing 1% (vol/vol) acetic acid (Fig. 7B). rRNA was used to check the amount of RNA used. gmpA transcription was greatly enhanced by the addition of ethanol and acetic acid. On the other hand, the addition of ethanol or acetic acid had no significant effects on the transcription of gmpB, gmpC, or ginI. These data suggested that gmpA transcription was positively regulated not only by the GinI/GinR quorum-sensing system via GinA but also by ethanol and acetic acid in the medium.


Figure 7
View larger version (30K):
[in this window]
[in a new window]

  		FIG. 7. Dependence of gmpA transcription on ethanol and acetic acid. (A) S1 nuclease mapping of ginI and gmpA to gmpC in wild-type strain NCI1051 harboring pMV24 cultured in YPG medium (YPG), YPG medium containing and 2% (vol/vol) ethanol (+ ethanol), and YPG medium containing 1% (vol/vol) acetic acid (+ acetic acid) for the indicated hours. All media contained 1% (vol/vol) cellulase. (B) Growth of wild-type strain NCI1051 harboring pMV24 in YPG medium (•), YPG medium containing 2% (vol/vol) ethanol ({blacktriangleup}), and YPG medium containing 1% (vol/vol) acetic acid ({blacksquare}). All media contained 1% (vol/vol) cellulase. OD660, optical density at 660 nm.

Lack of detectable changes in vesicle formation by or antibiotic susceptibility of the gmpA mutant. Pgm6 and Pgm7, both of which show a high level of similarity to E. coli OmpA in the C-terminal region and form eight-stranded β-barrels in the N-terminal region, are major outer membrane proteins in Porphyromonas gingivalis, one of the most important bacteria in the progression of chronic periodontal disease. Because the heterologously trimeric complex Pgm6-Pgm7 functions as a stabilizer of the cell wall, Pgm6-Pgm7 mutants have a wavy, uneven, discontinuous outer membrane and release more membrane vesicles from cells than wild-type strains (11). These mutants also show increased permeability for solutes compared to wild-type strains (18). These phenotypes of Pgm6-Pgm7 mutants, together with the fact that outer membrane vesicles are found in various gram-negative bacteria, including acetic acid bacteria (3), prompted us to examine the structure of the outer membrane by electron microscopy and the antibiotic susceptibilities of mutants with disruptions in the GinI/GinR regulatory system. We examined mutant ginI::Km, instead of mutant gmpA::Km, because almost no GmpA was produced in mutant ginI::Km (Fig. 1B). However, scanning and transmission electron microscopy showed no apparent difference in membrane structure or in the formation and morphology of membrane vesicles (see Fig. S2 in the supplemental material).

Some OmpA family proteins, located in the outer membrane, form nonspecific diffusion channels that allow the penetration of various solutes, although the rate of solute diffusion through OmpA in E. coli is about 2 orders of magnitude lower than that through the OmpF or OmpC porin (24). To assess the function of GmpA as a diffusion pump, we measured the antibiotic susceptibility of the gmpA mutant and determined MICs of apramycin, erythromycin, gentamicin, novobiocin, polymyxin B, rifampin, and streptomycin by the standard dilution method with a 96-well microtiter plate. Serial dilutions of the corresponding antibiotics were added to YPG medium containing 1% (vol/vol) cellulase with or without 2% (vol/vol) ethanol. After 24 h of incubation with shaking at 30°C, the turbidities of the cultures at 660 nm were measured by a photometer. There was no significant difference in MICs for the wild-type strain NCI1051 and the gmpA::Km mutant (see Table S1 in the supplemental material). Because strong diffusion channels, OmpF and OmpC, are still active in mutant gmpA::Km and because perhaps 99% of the penetration of solutes across the outer membrane would occur through the trimeric OmpF and OmpC porins (19), the difference in susceptibility to antibiotics would be observed only when GmpA serves as a specific, strong diffusion channel. The lack of difference in antibiotic susceptibility between the wild-type strain and mutant gmpA::Km suggested that GmpA did not serve as a special, efficient diffusion channel.


arrow
DISCUSSION
 
We have demonstrated that the GinI/GinR quorum-sensing system in G. intermedius controls the expression of gmpA via GinA, which in turn represses oxidative fermentation, including acetic acid and gluconic acid fermentation, and growth in medium containing ethanol. The OmpA family proteins play a structural role in the maintenance of gram-negative bacterial membrane integrity by embedding their N-terminal transmembrane β-strands into the outer membrane and providing their C-terminal domains in the periplasmic space as a physical linkage between the outer membrane and the peptidoglycan layer (4, 13, 14). The presence of an extra sequence in one of the surface-exposed loops in the N-terminal transmembrane domain, which turned out to be essential for the repression of oxidative fermentation, appears to be specific to so-called vinegar producers that accumulate acetic acid as a waste product in large amounts, as a result of ATP generation by means of ethanol oxidation. In addition, the gene cluster containing three ompA family genes is also specific to G. intermedius and G. polyoxogenes, both of which accumulate acetic acid at high yields. These findings suggest that GmpA plays a specific and important role in acetic acid fermentation. Consistent with this idea, gmpA transcription was regulated not only by the GinI/GinR quorum-sensing system via GinA but also by ethanol as the substrate of the oxidative fermentation and acetic acid as the final product. Although the function of the 89-amino-acid protein GinA, with no known motifs, is unclear at present, GinA may control gmpA transcription by protein-protein interaction with an unknown activator.

Various functions controlled by a quorum-sensing system reflect the needs of a particular species of bacteria to inhabit a given niche. The gene cluster containing three adjacent omp genes, one of which encodes a GmpA-type protein containing an extra region in one of the loops in the N-terminal membrane domain, has been found so far only in G. intermedius NCI1051 and G. polyoxogenes NCI1028, high-yield vinegar producers. Because the substrate, ethanol, and the product, acetic acid, of acetic acid fermentation are both toxic to the cell, acetic acid bacteria synthesize ATP by oxidative fermentation under extremely unfavorable conditions. It is conceivable that acetic acid bacteria have developed the GinI/GinR system to control the GmpA expression via GinA, which in turn controls the rates of growth and acetic acid production in the presence of ethanol, where the bacteria of this group generate ATP by ethanol oxidation via a specific electron transport system. The vinegar producer strains NCI1051 and NCI1028 have probably duplicated the GOX1787 gene twice to adapt to acetic acid fermentation during evolution, which is apparent from the gene organization around the ancestral gmpA gene, GOX1787, in Gluconobacter oxydans 621H and the two vinegar producers (Fig. 2A). Concerning the function of GmpA as a stabilizer of the cell wall, we speculate that GmpA functions as a permeability barrier for environmental factors, such as an organic solvent, ethanol, that is toxic to the cell, although no differences in ethanol resistance between the wild-type strain and mutant gmpA::Km were detectable by a routine test in which both strains were challenged in medium containing various amounts of ethanol (data not shown). The wild-type strain could reduce acetic acid production from ethanol, compared to that in the gmpA::Km mutant, because the outer membrane containing GmpA prevents the ethanol from being imported into the periplasm, where alcohol dehydrogenase and aldehyde dehydrogenase, both of which are anchored to the inner membrane (17) and are required for ethanol oxidation into acetic acid, are present.

The disruption of gmpA resulted in an increase in both the rates of acetic acid and gluconic acid production and the final yields of acetic acid and gluconic acid (Fig. 6A and B), as observed for the ginI::Km, ginR::Km, and ginA::Km mutants (10). However, the gmpA::Km mutant continued to produce acetic acid in the stationary phase (Fig. 6A), unlike the ginI::Km and ginA::Km mutants (10). Thus, the phenotypes of the gmpA::Km mutant were not completely the same as those of the ginI::Km and ginA::Km mutants. Some other target gene(s) of GinA is perhaps present, affecting these phenotypes in some unknown way.

Our previous study showed that GinA represses an antifoam activity, which suggested that ginA has some effects on cell surface hydrophobicity. Because GmpA is an outer membrane protein, we expected that GmpA might be involved in the antifoam activity of cell. However, the disruption of gmpA had no effects on antifoam activity. These findings show that GmpA is not involved in the antifoam activity and that GinA has another unknown target related to the antifoam activity.

In conclusion, we have revealed that GinA, a target of the GinI/GinR quorum-sensing system in the acetic acid bacterium G. intermedius and perhaps G. polyoxogenes, too, represses acetic acid fermentation via an outer membrane protein, GmpA (Fig. 8). In other words, the GinI/GinR quorum-sensing system controls the most important and specific function of acetic acid bacteria, i.e., ATP generation by means of acetic acid fermentation in the presence of ethanol. However, there are several black boxes to be elucidated in the hierarchical regulatory scheme presented in Fig. 8. First, how the 89-amino-acid protein GinA, with no known motifs, activates gmpA transcription should be clarified. Second, how GmpA, with an extra hydrophilic sequence in the surface-exposed loop, represses oxidative fermentation is also unknown. Third, how GinA represses antifoam activity is still unclear. Notwithstanding the unknown steps in the regulatory hierarchy in Fig. 8, our findings hitherto would be useful in strain improvement for industrial vinegar fermentation by acetic acid bacteria. In fact, for example, we successfully increased the final yield of acetic acid by disrupting gmpA in G. intermedius.


Figure 8
View larger version (15K):
[in this window]
[in a new window]

  		FIG. 8. Possible model for the quorum-sensing system in G. intermedius NCI1051. The GinI/GinR quorum-sensing system activates the transcription of ginA, encoding an 89-amino-acid protein, which in turn activates the transcription of gmpA in a still unknown manner. The transcription of gmpA, encoding an OmpA outer membrane family protein, is also activated by ethanol and acetic acid. Thus induced, GmpA represses oxidative fermentation, including acetic acid and gluconic acid fermentation, and growth in the presence of ethanol in an unknown manner. GinA also represses the antifoam activity of cells (10). R, GinR; I, GinI; P, promoter; A, GinA.


arrow
ACKNOWLEDGMENTS
 
We thank Stephen K. Farrand (University of Illinois) for providing the AHL reporter strain. We also thank Shigeru Nakano (Mizkan Group Co., Ltd.) for helpful advice on 2D-PAGE.


arrow
FOOTNOTES
 
* Corresponding author. Mailing address: Department of Biotechnology, Graduate School of Agriculture and Life Sciences, The University of Tokyo, Bunkyo-ku, Tokyo 113-8657, Japan. Phone: 81 3 5841 5123. Fax: 81 3 5841 8021. E-mail: asuhori{at}mail.ecc.u-tokyo.ac.jp Back

{triangledown} Published ahead of print on 16 May 2008. Back

{dagger} Supplemental material for this article may be found at http://jb.asm.org/. Back


arrow
REFERENCES
 
    1
  1. Ausubel, F. M., R. Brent, R. E. Kingstone, D. O. Moore, J. S. Seidman, J. A. Smith, and K. Struhl. 1987. Current protocols in molecular biology. John Wiley & Sons, New York, NY.
    2. 2
  2. Beck, E., G. Ludwig, E. A. Auerswald, B. Reiss, and H. Schaller. 1982. Nucleotide sequence and exact localization of the neomycin phosphotransferase gene from transposon Tn5. Gene 19:327-336.[CrossRef][Medline]
    3. 3
  3. Beveridge, T. J. 1999. Structures of gram-negative cell walls and their derived membrane vesicles. J. Bacteriol. 181:4725-4733.[Free Full Text]
    4. 4
  4. De Mot, R., and J. Vanderleyden. 1994. The C-terminal sequence conservation between OmpA-related outer membrane proteins and MotB suggests a common function in both gram-positive and gram-negative bacteria, possibly in the interaction of these domains with peptidoglycan. Mol. Microbiol. 12:333-334.[CrossRef][Medline]
    5. 5
  5. Fukaya, M., K. Tayama, T. Tamaki, H. Tagami, H. Okumura, Y. Kawamura, and T. Beppu. 1989. Cloning of the membrane-bound aldehyde dehydrogenase gene of Acetobacter polyoxogenes and improvement of acetic acid production by use of the cloned gene. Appl. Environ. Microbiol. 55:171-176.[Abstract/Free Full Text]
    6. 6
  6. Fuqua, C., S. C. Winans, and E. P. Greenberg. 1996. Census and consensus in bacterial ecosystems: the LuxR-LuxI family of quorum-sensing transcriptional regulators. Annu. Rev. Microbiol. 50:727-751.[CrossRef][Medline]
    7. 7
  7. Fuqua, W. C., S. C. Winans, and E. P. Greenberg. 1994. Quorum sensing in bacteria: the LuxR-LuxI family of cell density-responsive transcriptional regulators. J. Bacteriol. 176:269-275.[Free Full Text]
    8. 8
  8. Greenberg, D. E., S. F. Porcella, A. M. Zelazny, K. Virtaneva, D. E. Sturdevant, J. J. Kupko III, K. D. Barbian, A. Babar, D. W. Dorward, and S. M. Holland. 2007. Genome sequence analysis of the emerging human pathogenic acetic acid bacterium Granulibacter bethesdensis. J. Bacteriol. 189:8727-8736.[Abstract/Free Full Text]
    9. 9
  9. Horton, R. M., H. D. Hunt, S. N. Ho, J. K. Pullen, and L. R. Pease. 1989. Engineering hybrid genes without the use of restriction enzymes: gene splicing by overlap extension. Gene 77:61-68.[CrossRef][Medline]
    10. 10
  10. Iida, A., Y. Ohnishi, and S. Horinouchi. 2008. Control of acetic acid fermentation by quorum sensing via N-acylhomoserine lactones in Gluconacetobacter intermedius. J. Bacteriol. 190:2546-2555.[Abstract/Free Full Text]
    11. 11
  11. Iwami, J., Y. Murakami, K. Nagano, H. Nakamura, and F. Yoshimura. 2007. Further evidence that major outer membrane proteins homologous to OmpA in Porphyromonas gingivalis stabilize bacterial cells. Oral Microbiol. Immunol. 22:356-360.[CrossRef][Medline]
    12. 12
  12. Kelemen, G. H., P. Brian, K. Flärdh, L. Chamberlin, K. F. Chater, and M. J. Buttner. 1998. Developmental regulation of transcription of whiE, a locus specifying the polyketide spore pigment in Streptomyces coelicolor A3(2). J. Bacteriol. 180:2515-2521.[Abstract/Free Full Text]
    13. 13
  13. Koebnik, R. 1995. Proposal for a peptidoglycan-associating alpha-helical motif in the C-terminal regions of some bacterial cell-surface proteins. Mol. Microbiol. 16:1269-1270.[CrossRef][Medline]
    14. 14
  14. Koebnik, R., K. P. Locher, and P. Van Gelder. 2000. Structure and function of bacterial outer membrane proteins: barrels in a nutshell. Mol. Microbiol. 37:239-253.[CrossRef][Medline]
    15. 15
  15. Luo, Z. Q., S. Su, and S. K. Farrand. 2003. In situ activation of the quorum-sensing transcription factor TraR by cognate and noncognate acyl-homoserine lactone ligands: kinetics and consequences. J. Bacteriol. 185:5665-5672.[Abstract/Free Full Text]
    16. 16
  16. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.
    17. 17
  17. Matsushita, K., H. Toyama, and O. Adachi. 1994. Respiratory chains and bioenergetics of acetic acid bacteria. Adv. Microb. Physiol. 36:247-301.[Medline]
    18. 18
  18. Nagano, K., E. K. Read, Y. Murakami, T. Masuda, T. Noguchi, and F. Yoshimura. 2005. Trimeric structure of major outer membrane proteins homologous to OmpA in Porphyromonas gingivalis. J. Bacteriol. 187:902-911.[Abstract/Free Full Text]
    19. 19
  19. Nikaido, H. 2003. Molecular basis of bacterial outer membrane permeability revisited. Microbiol. Mol. Biol. Rev. 67:593-656.[Abstract/Free Full Text]
    20. 20
  20. Okamoto-Kainuma, A., W. Yan, S. Kadono, K. Tayama, Y. Koizumi, and F. Yanagida. 2002. Cloning and characterization of groESL operon in Acetobacter aceti. J. Biosci. Bioeng. 94:140-147.[Medline]
    21. 21
  21. Pautsch, A., and G. E. Schulz. 1998. Structure of the outer membrane protein A transmembrane domain. Nat. Struct. Biol. 5:1013-1017.[CrossRef][Medline]
    22. 22
  22. Prust, C., M. Hoffmeister, H. Liesegang, A. Wiezer, W. F. Fricke, A. Ehrenreich, G. Gottschalk, and U. Deppenmeier. 2005. Complete genome sequence of the acetic acid bacterium Gluconobacter oxydans. Nat. Biotechnol. 23:195-200.[CrossRef][Medline]
    23. 23
  23. Ravn, L., A. B. Christensen, S. Molin, M. Givskov, and L. Gram. 2001. Methods for detecting acylated homoserine lactones produced by Gram-negative bacteria and their application in studies of AHL-production kinetics. J. Microbiol. Methods 44:239-251.[CrossRef][Medline]
    24. 24
  24. Sugawara, E., and H. Nikaido. 1992. Pore-forming activity of OmpA protein of Escherichia coli. J. Biol. Chem. 267:2507-2511.[Abstract/Free Full Text]
    25. 25
  25. Waters, C. M., and B. L. Bassler. 2005. Quorum sensing: cell-to-cell communication in bacteria. Annu. Rev. Cell Dev. Biol. 21:319-346.[CrossRef][Medline]
    26. 26
  26. Williams, P. 2006. Quorum sensing. Int. J. Med. Microbiol. 296:57-59.[CrossRef][Medline]
    27. 27
  27. Wong, H. C., A. L. Fear, R. D. Calhoon, G. H. Eichinger, R. Mayer, D. Amikam, M. Benziman, D. H. Gelfand, J. H. Meade, A. W. Emerick, R. Bruner, A. Ben-Bassat, and R. Tal. 1990. Genetic organization of the cellulose synthase operon in Acetobacter xylinum. Proc. Natl. Acad. Sci. USA 87:8130-8134.[Abstract/Free Full Text]

Journal of Bacteriology, July 2008, p. 5009-5019, Vol. 190, No. 14
0021-9193/08/$08.00+0     doi:10.1128/JB.00378-08
Copyright © 2008, American Society for Microbiology. All Rights Reserved.



This article has been cited by other articles:

  • Iida, A., Ohnishi, Y., Horinouchi, S. (2009). Identification and characterization of target genes of the GinI/GinR quorum-sensing system in Gluconacetobacter intermedius. Microbiology 155: 3021-3032 [Abstract] [Full Text]  

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental material
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Iida, A.
Right arrow Articles by Horinouchi, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Iida, A.
Right arrow Articles by Horinouchi, S.

Copyright © 2009 by the American Society for Microbiology.   For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»