Logo of jbacterPermissionsJournals.ASM.orgJournalJB ArticleJournal InfoAuthorsReviewers
J Bacteriol. 2004 Jun; 186(11): 3539–3546.
PMCID: PMC415761

Evidence for an Arginine Exporter Encoded by yggA (argO) That Is Regulated by the LysR-Type Transcriptional Regulator ArgP in Escherichia coli


The anonymous open reading frame yggA of Escherichia coli was identified in this study as a gene that is under the transcriptional control of argP (previously called iciA), which encodes a LysR-type transcriptional regulator protein. Strains with null mutations in either yggA or argP were supersensitive to the arginine analog canavanine, and yggA-lac expression in vivo exhibited argP+-dependent induction by arginine. Lysine supplementation phenocopied the argP null mutation in that it virtually abolished yggA expression, even in the argP+ strain. The dipeptides arginylalanine and lysylalanine behaved much like arginine and lysine, respectively, to induce and to turn off yggA transcription. Dominant missense mutations in argP (argPd) that conferred canavanine resistance and rendered yggA-lac expression constitutive were obtained. The protein deduced to be encoded by yggA shares similarity with a basic amino acid exporter (LysE) of Corynebacterium glutamicum, and we obtained evidence for increased arginine efflux from E. coli strains with either the argPd mutation or multicopy yggA+. The null yggA mutation abolished the increased arginine efflux from the argPd strain. Our results suggest that yggA encodes an ArgP-regulated arginine exporter, and we have accordingly renamed it argO (for “arginine outward transport”). We propose that the physiological function of argO may be either to prevent the accumulation to toxic levels of canavanine (which is a plant-derived antimetabolite) or arginine or to maintain an appropriate balance between the intracellular lysine and arginine concentrations.

Canavanine (CAN) is a plant-derived antimetabolite that, by virtue of its chemical similarity to arginine (Arg), inhibits the growth of bacteria after its competitive misincorporation into polypeptides in place of Arg (40). In Escherichia coli, mutations conferring Canr have accordingly been obtained in the genes argR and argS, which encode the apo-repressor for enzymes of Arg biosynthesis and the arginyl-tRNA synthetase, respectively (19, 28).

Evidence exists for three different periplasmic-binding-protein-dependent Arg transporters in E. coli that presumably also mediate CAN uptake (reviewed in references 19 and 34). The periplasmic proteins for the three transporters are the LAO protein (which binds the basic amino acids lysine [Lys], Arg, and ornithine), the AO protein (which binds Arg and ornithine), and the ArtJ protein (which binds Arg). Canr mutations have been identified in some genes that have been implicated in Arg uptake, including abpS, argP, and argK. It has been proposed that abpS encodes the AO protein (10; also see reference 34) and that the products of argP and argK regulate the activities of the different Arg transporters (11-14, 35, 40).

A molecular analysis of argP undertaken by Celis (12) surprisingly revealed that it is identical to iciA, whose product had previously been implicated as both an inhibitor of chromosomal replication initiation from oriC in vitro (23, 24, 41) and an activator of transcription from the dnaA 1P (29, 30) and nrd (21) promoters. The ArgP (IciA) protein is a member of the LysR family of transcriptional activators (39). The dominant argP55 mutation (argPd) conferring Canr was deduced to cause a Pro-to-Ser substitution at residue 274 (P274S) in the 297-amino-acid protein (12). Based on data from in vitro studies, Celis suggested that wild-type ArgP both represses its own transcription and activates that of argK; he also sought to explain the Canr phenotype associated with the argPd mutation on the assumption that it is a loss-of-function (that is, dominant-negative) allele whose product is unable to efficiently activate the genes involved in Arg uptake, including argK (12).

In theory, perturbations in Arg or CAN efflux may also be expected to affect the CAN tolerance of a bacterial cell. However, an Arg efflux system is not known to exist in E. coli, and indeed only a few amino acid exporters have been characterized for any bacterium (reviewed in references 1, 9, and 17). The amino acid exporters so far identified for E. coli include RhtB and RhtC, for threonine and homoserine (27, 43), and YdeD and YfiK, for cysteine and O-acetylserine (16, 18). The exporter LysE, for both Lys and Arg, has been well characterized in Corynebacterium glutamicum, and the primary sequence of LysE is similar to the product of an E. coli open reading frame called yggA (3, 7, 8, 42). Interestingly, the synthesis of LysE in C. glutamicum is under the transcriptional control of LysG, which in turn is related in its primary sequence to E. coli ArgP (3, 42).

We show in this study that strains with null mutations in either argP or yggA exhibit abnormally increased sensitivities to CAN (Canss) and that ArgP is a transcriptional regulator of yggA that mediates the latter's induction by Arg. Intracellular Lys, on the other hand, mediates a reduction in yggA expression, apparently by abolishing the activating role of ArgP. Furthermore, Canr argPd mutations represent gain-of-function, not loss-of-function, alleles that render yggA expression constitutive. Derivatives with an argPd mutation or with multicopy yggA+ exhibit increased Arg excretion. Our results indicate that yggA encodes an ArgP-regulated Arg exporter in E. coli.


Bacterial strains and growth media.

The genotypes of the E. coli strains used for this study are listed in Table Table1.1. The defined and nutrient media were, respectively, minimal A medium (supplemented with 0.2% glucose and the appropriate auxotrophic requirements) and Luria-Bertani (LB) medium (33), and the growth temperature was 37°C. All amino acids, all dipeptides, and CAN were l-isomers. Unless otherwise indicated, supplementation with Arg, Lys, and the dipeptides was done at a concentration of 1 mM. The concentrations of antibiotics used were described previously (38). Trimethoprim was used in nutrient and defined media at 60 and 30 μg/ml, respectively.

List of E. coli K-12 strainsa

Phages and plasmids.

The transposon vehicle phage λNK1323, used for the generation of transpositions of Tn10dTet, has been described elsewhere (25, 33). Phages λ 471 and λ 472, from the ordered E. coli genomic library of Kohara et al. (26), were obtained from K. Isono.

The plasmid vectors employed were as follows (with salient features in parentheses): (i) pMU575 (IncW based, single copy number, trimethoprim resistant, carries a promoterless lacZ gene for the generation of promoter-lac operon fusions) (2), (ii) pCL1920 (pSC101 based, low copy number, spectinomycin and streptomycin resistant) (31), (iii) pBR329 (pMB9 based; high copy number; ampicillin, tetracycline, and chloramphenicol resistant) (15), and (iv) pBluescript II KS (pMB9 based, very high copy number, ampicillin resistant) (Stratagene, La Jolla, Calif.).

The following plasmids were constructed in this study. Plasmids pHYD915, pHYD933, and pHYD949 are derivatives of pCL1920 carrying the following fragments from the λ phages described by Kohara et al. (26) subcloned into the appropriate sites of the vector: a 1.8-kb SalI fragment containing argP+ from λ 471, a 2.2-kb BamHI fragment containing argK+ from λ 471, and a 1.2-kb PstI-HindIII fragment containing yggA+ from λ 472, respectively. Plasmid pHYD952 carries the 1.2-kb PstI-HindIII fragment containing yggA+ cloned (via an intermediate step of cloning into pBluescript II KS) into the BamHI-HindIII sites of pBR329, and plasmid pHYD954 carries a 1.1-kb PstI-PvuII fragment containing yggA′ (truncated at its 3′ end) cloned into the PstI-EcoRV sites of pBluescript II KS. The latter yggA′ construct was also subcloned from pHYD954 on a PstI-HindIII fragment (HindIII end derived from the multiple-cloning-site region of the vector) into the corresponding sites of plasmid pMU575 to generate pHYD956 (with a yggA-lac operon fusion). Plasmids pHYD926 through pHYD932, with missense mutations in argP conferring Canr that were obtained by nitrosoguanidine mutagenesis of pHYD915, are described below. The 1.8-kb SalI fragment containing the argPd allele from pHYD926 was subcloned into the SalI site of pBR329 to generate plasmid pHYD953.

Tests for CAN tolerance.

The CAN tolerance of strains was tested in glucose-minimal A medium supplemented with various concentrations of CAN, with growth being scored after 24 h at 37°C. In some cases, uracil was added at 40 μg/ml to the medium to enhance the toxicity imposed by a given CAN concentration (19). Typically, the wild-type E. coli strain (Cans) was resistant to 20 μg of CAN/ml but was sensitive to 65 μg of CAN/ml in the presence of uracil. Canss strains were sensitive even to the former concentration, and Canr strains were resistant even to the latter one.

Test for Arg cross-feeding.

Approximately 105 cells of the ΔargH strain SK2226 (or its pBR329 transformant derivative), contained in a 100-μl volume of Luria-Bertani medium, were added to a petri dish with 20 ml of glucose-minimal A agar supplemented with proline, tryptophan, and histidine (that is, all of the auxotrophic requirements of SK2226 except Arg). The agar medium also contained the indicator tetrazolium chloride at 1 μg/ml (33). The test strains were spotted in quadrants on the agar surface, and Arg cross-feeding was visualized as red haloes of syntrophic growth of the auxotrophic strain around the spots.

Isolation and sequence analysis of Canr argP mutants.

To obtain mutations in plasmid pHYD915 (carrying argP+) that conferred a Canr phenotype, we mutagenized the strain MC4100/pHYD915 with nitrosoguanidine according to a previously described procedure (33). A plasmid DNA preparation made from the pool of mutagenized cells was then used to transform an argP202 null mutant, strain GJ4536, and individual transformants were scored for Canr. The DNA sequence of the insert region in each of the mutant plasmids conferring Canr was determined with the aid of one pair of lac primers (recognizing sequences in the vector that flank the insert) as well as another pair of primers internal to argP, namely ARGP1 (5′-GGGCGCGAACTCGCTGAGCGA-3′) and ARGP2 (5′-GAGCAAGTTGTACGAACGCTT-3′).

Localized insertional mutagenesis near serA with Tn10dTet and molecular genetic analysis of mutants.

The wild-type strain MC4100 was subjected to random insertional mutagenesis with the Tn10dTet transposon following infection with λ NK1323, as described elsewhere (25, 33). A phage P1 lysate prepared on a pool of Tetr clones was used to transduce the serA mutant strain GJ4674 simultaneously to the phenotype Ser+ Tetr. Transductants so obtained on double selection plates were expected to have Tn10dTet insertion mutations in the vicinity of the serA locus.

To determine the sites of Tn10dTet insertion in the mutants, we transduced one insertion allele into MC4100 and used the resulting strain, GJ4822, for inverse PCR as follows. A template preparation of circularized fragments of Sau3A1-digested chromosomal DNA from GJ4822 was subjected to PCR with a pair of divergently oriented primers, AH1 and AH2, designed for the ends of Tn10dTet (22), and one of the products so obtained was sequenced with the same primers.

Upon the identification of yggA as the gene disrupted by Tn10dTet in GJ4822, the other Tn10dTet insertions were molecularly characterized and sequenced after PCRs with the aid of (i) two chromosomal primers flanking yggA and its adjacent gene, yggB (YGGAR, 5′-ACCTCTGGATCCAAGCTTAG-3′, and YGGBF, 5′-TCCAGGAATTCAACGCGATCGA-3′; mismatches to the genome sequence [5] are indicated in italics), and (ii) two more previously described primers, namely TetF (5′-TGGTCACCAACGCTTTTCCCGAG-3′) and TetR (5′-CTGTTGACAAAGGGAATCATAG-3′) (36), that read outward from either end of Tn10dTet.

Other techniques.

The procedures for phage P1 transduction (20) and for various recombinant DNA manipulations (37) were done as previously described. β-Galactosidase specific activity measurements in yggA′-lac and argP-lac strains were made as described by Miller (33); the values are reported in Miller units and represent the averages of at least four independent determinations, with the variations between individual values being <20%.


argP null mutants are Canss.

In work to be described elsewhere (M. R. Nandineni, R. Laishram, and J. Gowrishankar, unpublished data), we identified an insertion of the transposon phage λplacMu55(Kan) at codon 35 of the argP structural gene (designated argP202) that was associated with an osmosensitive phenotype for E. coli strains. In light of the report by Celis (12) that a presumptively dominant-negative argP mutant is Canr, we tested the null argP202 mutant strain GJ4536 for its CAN tolerance. Unexpectedly, it was Canss, that is, more sensitive to CAN (with a CAN MIC of approximately 2 μg/ml) than the wild-type strain (Table (Table2).2). A second independent argP insertion mutant was also Canss (data not shown).

CAN tolerance or sensitivity of argP and yggA strains

For complementation studies, plasmid pHYD915 carrying the cloned argP+ gene was constructed as described above. The plasmid was able to reverse the Canss phenotype of the argP202 mutant (Table (Table2),2), indicating that the insertion mutation is recessive to argP+.

Identification of argP mutants conferring Canr.

Our findings above suggested that the loss of argP function is associated with a Canss phenotype, whereas the argP P274S mutant previously described by Celis (12) was Canr. In order to identify additional Canr argP mutants, we mutagenized the argP+ plasmid pHYD915 with nitrosoguanidine by the procedure described above and screened transformants of the argP202 strain GJ4536 for Canr.

Seven of about 800 transformants tested were Canr, and in each of them the mutation was plasmid-borne (data not shown). The plasmids were designated pHYD926 through pHYD932, and a sequence analysis indicated that each carries a GC-to-AT missense mutation at a different site in argP, which was deduced to result in a single amino acid residue alteration in the encoded protein (Table (Table33).

Molecular analysis of plasmid-borne mutations in argP conferring Canr in argP202 strain GJ4536

When the mutant plasmids were introduced into strain MC4100, which carries argP+ chromosomally, all but one conferred a Canr phenotype, indicating that the mutations in them were dominant (Table (Table3).3). For the work described below, the argP S94L mutation in plasmid pHYD926 was used as a prototypic example of a missense argPd mutation. The argP R295C mutation was inferred to be recessive to argP+, since plasmid pHYD931 did not confer Canr to MC4100.

Identification of insertions in yggA conferring Canss.

The argK (also called ygfD) gene is situated 2 kb away from argP (which in turn is situated 2 kb away from serA) (Fig. (Fig.1)1) (5); it was previously suggested both that argK is transcriptionally regulated by argP and that mutations in argK confer Canr (11-13). We accordingly undertook localized transposon (Tn10dTet) mutagenesis of the chromosomal region in the vicinity of serA by the method described above and scored the mutants for altered CAN tolerance. No Canr mutants were obtained, but seven Tetr derivatives were identified that were Canss. In every case, the Tetr insertion was shown (i) to be 100% cotransducible with Canss and (ii) to confer Canss in MC4100 and other strain backgrounds (data not shown). The MIC of CAN for these insertion mutants, at about 0.5 μg/ml, was even lower than that for the argP202 null mutant.

FIG. 1.
Physical map of serA-yggB region of the E. coli genome at 69.5 min (clockwise going from left to right), as annotated in the work of Blattner et al. (5). Positions and transcriptional orientations of individual open reading frames are marked beneath the ...

The CAN phenotype was not complemented by either plasmid pHYD915 (carrying argP+) or pHYD933 (carrying argK+) in any of the mutants (data not shown). As described above, the strategy of inverse PCR (based on primers designed for the ends of Tn10dTet) was then adopted to determine the site of Tn10dTet insertion in one of the Canss mutants. DNA sequence analysis of the resultant PCR products unexpectedly revealed that the insertion was in an open reading frame called yggA (encoding a putative polypeptide of 211 amino acids) situated 8 kb away from argP (and 10 kb from serA) (Fig. (Fig.1).1). Using a pair of primers flanking yggA (and its adjacent gene, yggB) and a second pair of primers designed to read outward from the two ends of Tn10dTet, we performed PCR experiments as described above to establish that the Tetr insertions in all seven Canss mutants were situated in yggA (data not shown). The precise sites of the insertions were determined by sequencing of the PCR products, and the results summarized in Table Table44 indicate that the insertions had occurred in both orientations at different sites in the proximal third of the yggA open reading frame (or in the region immediately upstream of its putative start codon).

Molecular analysis of Tn10dTet insertions in yggA

Plasmids with the cloned minimal yggA+ gene and its regulatory region in a 1.2-kb fragment (pHYD949 [low copy number, pSC101 based] and pHYD952 [high copy number, pBR329 based]) were able to complement a representative yggA insertion mutant, strain GJ4822, for CAN tolerance (Table (Table2).2). A smaller (1.1 kb) fragment, which was expected to encode a YggA′ polypeptide with a truncation of 14 amino acids at its C terminus, was also able to complement the yggA mutant, but only when it was present in a very-high-copy-number plasmid (pHYD954), not in an IncW-derived single-copy-number plasmid (pHYD956) (Table (Table22).

The yggB gene is situated immediately upstream of, and in the same transcriptional orientation as, yggA (Fig. (Fig.1).1). It encodes a mechanosensitive ion channel (32), and we determined that a ΔyggB mutant (kindly provided by Ian Booth) is unaffected in CAN tolerance (data not shown).

Transcriptional regulation of yggA by argP in vivo.

Plasmid pHYD956 also carries the promoterless lacZ gene in the vector region downstream of the yggA′ truncation, and we accordingly employed it as a yggA-lac operon fusion to study the transcriptional regulation of yggA in different strains. The results of these experiments are presented in Fig. Fig.2,2, and the conclusions therefrom are summarized below.

FIG. 2.
Expression of yggA-lac in pHYD956 transformants of the wild-type (WT) strain MC4100 and the following mutant derivatives: argR, GJ4748; argR yggA, GJ4894; argR argP, GJ4895; and argPd, MC4100/pHYD926. Cultures were grown to exponential phase in glucose-minimal ...

(i) In argP+ strains, yggA expression appeared to be induced by Arg, its precursor citrulline, or its analog CAN. An argR mutation that derepressed Arg biosynthesis also served to induce yggA expression, probably by increasing the intracellular Arg pools. The induction by Arg, citrulline, and in particular, CAN was most prominent in an argR yggA mutant strain (which, as discussed below, we believe is blocked for Arg and CAN export).

(ii) The inducing effects of Arg or its related compounds, as well as of the argR mutation, on yggA expression were completely dependent on argP+ and were abolished by the introduction of the null argP202 mutation. (For these experiments, the argP-lac fusion itself encoded by argP202 was inactivated by a Tn10dTet insertion in lacZ.) The addition of Lys also appeared to mimic the effect of the argP null mutation in that it served to nearly completely repress yggA expression, even in the argP+ strains.

(iii) Finally, in an argPd S94L mutant background, yggA-lac expression was rendered high and constitutive by all coeffectors, including Arg, CAN, and Lys.

Evidence for increased Arg excretion in strains with argPd mutation or multicopy yggA+.

The polypeptide deduced to be encoded by yggA shares significant similarity with the C. glutamicum protein LysE, which has been shown to function as an exporter of Lys and Arg for this bacterium (3, 42). We reasoned that the Canss and Canr phenotypes, respectively, of argP null and argPd mutants can be explained on the assumption that YggA also similarly functions in E. coli as an exporter of Arg and its analog CAN.

To test this hypothesis, we constructed test strains with increased levels of YggA protein by the introduction on a multicopy plasmid of either (i) yggA+ (pHYD952) or (ii) the argPd S94L mutation that was shown above to constitutively activate yggA expression (pHYD953). Both test strains also carried an argR mutation in order to increase the intracellular Arg pools, and their phenotypes were compared with that of a control argR mutant derivative that was otherwise wild type (that is, argP+ and haploid yggA+, transformed with the plasmid vector pBR329). Arg excretion was assayed by determining the halo of cross-feeding (that is, syntrophic growth) of the ΔargH strain SK2226/pBR329, which was seeded in an agar medium supplemented with ampicillin and on the surface of which each of the test and control strains had been spotted. The auxotrophic requirement of strain SK2226 (which is blocked in the last step of the Arg biosynthetic pathway) is satisfied only by Arg, not by any of its precursors, such as ornithine or citrulline (6).

By 12 h of incubation, both multicopy yggA+ and argPd were associated with dramatic increases in the Arg cross-feeding ability of the argR mutant (Fig. (Fig.3),3), which was suggestive of increased Arg export from the concerned test strains. In comparison with the argR+ strain MC4100, the argR mutant strain GJ4748 itself exhibited increased cross-feeding of the Arg auxotroph, but only after 40 h of incubation (data not shown). We were unable to introduce the argPd mutation into a strain with multicopy yggA+, suggesting that the massive overexpression of YggA under these conditions is probably lethal, either directly or as a consequence of its physiological effects on intracellular Arg pools.

FIG. 3.
Test for cross-feeding of Arg-auxotrophic strain SK2226/pBR329 by the following strains: argR/vector, GJ4748/pBR329; argR/argPd, GJ4748/pHYD953; argR/m.c. (multicopy) yggA+, GJ4748/pHYD952; and argR yggA/argPd, GJ4894/pHYD953. Cross-feeding was ...

yggA is epistatic to argPd.

If it is true that the Canr and Arg excretion phenotypes of the argPd mutant are a consequence of the increased and constitutive expression of YggA, then the phenotypes must be yggA+ dependent. We found that the introduction of a null mutation in yggA (i) rendered an argR+ argPd strain Canss (Table (Table2)2) and (ii) reversed the increased Arg cross-feeding ability of an argR argPd strain (Fig. (Fig.3).3). Another clue in support of the notion that yggA acts downstream of argP was the finding that multicopy yggA+ on plasmid pHYD952 suppressed the Canss phenotype of the argP null mutant, even as it conferred Canr in the argP+ strain (Table (Table2);2); these observations also suggested that yggA expression from plasmid pHYD952 was driven at least in part from a constitutive vector-borne promoter.

Evidence that yggA expression is affected by both intracellular Arg and Lys.

Although the data in Fig. Fig.22 demonstrated that exogenously supplied Arg or CAN, on the one hand, and Lys, on the other, have opposite effects on yggA-lac transcription, it was unclear whether both effects are primary or only one of them is so. Lys and Arg (and also CAN) share at least one common uptake system (involving the LAO protein, whose synthesis is also repressible by Lys) (19, 34); it was therefore possible that one of the amino acids (either Arg or Lys) affected yggA-lac expression directly by interacting with ArgP in the cytoplasm, while the second did so indirectly by interfering with the uptake of the first. By employing argP and yggA null mutants (in which a possible confounding effect of intracellular Lys on YggA-mediated CAN export does not exist), we did obtain evidence that exogenous Lys interferes with CAN uptake, since the MIC of CAN for the two strains was increased from 1.5 and 0.8 μg/ml in the absence of Lys to 12.5 and 3 μg/ml, respectively, in its presence.

To test the direct effects, if any, of intracellular Arg and Lys on yggA expression and exporter function, we undertook experiments with the dipeptides arginylalanine (Arg-Ala) and lysylalanine (Lys-Ala). The dipeptides were expected to be transported into the cells through uptake systems that were different from those for the amino acids and then to be hydrolyzed within the cells to release the cognate amino acids.

β-Galactosidase expression from the yggA-lac fusion plasmid pHYD956 in an argP+ argR strain was as substantially induced by Arg-Ala (535 Miller units) as it was by Arg (389 Miller units). Likewise, the expression of yggA-lac was reduced by Lys-Ala (33 Miller units) to the same extent that it was by Lys (58 Miller units). Our data are therefore consistent with the model that both Arg and Lys act directly within the cell to control yggA transcription in opposite ways.

Two related observations in support of a direct negative effect of intracellular Lys on yggA expression (and hence on Arg and CAN export) were as follows. (i) As mentioned above, the argR strain GJ4748 was able to cross-feed the ΔargH auxotrophic strain SK2226 for growth on Arg-free medium after 40 h of incubation. This cross-feeding ability was abolished upon supplementation of the medium with either Lys or Lys-Ala but not with a control dipeptide, histidylalanine (data not shown). On the other hand, neither Lys nor Lys-Ala had an effect on the pronounced Arg cross-feeding abilities (as depicted in Fig. Fig.3)3) of the argR derivatives with multicopy yggA+ or argPd, in both of which yggA expression was rendered to be at a high level and constitutive. (ii) The MIC of CAN for the wild-type strain MC4100 was unaffected by histidylalanine, whereas it was sharply reduced by Lys-Ala, from >100 μg/ml in the absence of the latter to 6 μg/ml in its presence. This result also served to rule out the possibility that the effect of Lys-Ala on yggA expression was indirect (for example, through the release of Lys into the medium and subsequent interference with Arg uptake or reuptake), since in that case Lys-Ala would also have been expected to interfere with CAN uptake into (and hence to increase the CAN tolerance of) the wild-type strain.

argP is not autoregulated.

Based on data from in vitro experiments, Celis (12) had suggested that argP transcription may be negatively autoregulated. The argP202 mutation represents a null insertion that also generates a lac operon fusion in the correct orientation with the chromosomal argP regulatory region (unpublished data), and we used strains with this allele to test for argP autoregulation in vivo. The data presented in Table Table55 indicate that the argP promoter is of moderate strength and that its expression in the argP202 strain derivatives is not affected by argP+ or argPd (carried, respectively, by plasmids pHYD915 or pHYD926), nor is it affected by potential coeffectors such as Arg or Lys. Similar results were also obtained for a second independent chromosomal argP-lac fusion (data not shown). We conclude that argP does not regulate its own transcription in vivo.

argP-lac expression in derivatives of argP202 strain GJ4536 with different plasmidsa


The salient findings of this study may be summarized as follows. (i) Null mutants in argP or yggA exhibit a Canss phenotype. (ii) Dominant missense gain-of-function mutations (argPd) that confer a Canr phenotype can occur in argP, but an argPd yggA double mutant is Canss. (iii) The transcription of yggA in vivo is ArgP dependent and is induced by exogenous Arg or Arg-Ala (and also by an argR mutation), whereas the addition of Lys or Lys-Ala phenocopies the effect of an argP null mutation; furthermore, yggA expression is rendered constitutive in an argPd strain. (iv) Finally, argR+ strains with argPd or multicopy yggA+ are Canr, and their argR derivatives (which are additionally derepressed for Arg biosynthesis) show highly increased Arg excretion levels.

YggA is similar to C. glutamicum LysE, which has been shown to export Arg and Lys (3, 42), whereas ArgP is a member of the LysR family of transcriptional activators (which includes LysG, the activator of LysE in C. glutamicum) (39). Therefore, a straightforward interpretation of our results, as further discussed below, is that (i) ArgP is a transcriptional activator of yggA, (ii) ArgP's activator function is enhanced by Arg and inhibited by Lys, and (iii) yggA encodes an Arg (and CAN) exporter in E. coli.

It may be emphasized here that our interpretations are based primarily on indirect lines of genetic evidence and that additional biochemical studies are necessary both to establish the role of ArgP and its coeffectors in yggA transcription and to examine the putative exporter function of YggA. With a native length of just 211 amino acids (from which, furthermore, the C-terminal stretch of 14 residues may be deleted without a complete loss of activity), E. coli YggA is even shorter than C. glutamicum LysE (233 residues) (3, 42); it will be interesting to determine how this class of especially small transport proteins mediates amino acid export and whether an efflux or exchange mechanism is involved in the process (7, 8).

The model proposed by Celis (11, 12) to explain the Canr phenotype of an argPd mutant assumed (i) that the mutation represents a loss-of-function (that is, dominant-negative) allele of argP and (ii) that wild-type ArgP is a transcriptional activator of argK, whose product in turn acts to enhance Arg uptake through two different transport systems. However, our finding in this study that yggA is epistatic to argPd indicates that the Canr phenotype associated with argPd can be entirely accounted for by the ArgP-mediated regulation of yggA itself. Furthermore, even the finding of Celis (12) that argK is a target for transcriptional activation by ArgP may need to be reexamined, given that (i) there is a discrepancy between the expected argK runoff transcript size (approximately 310 bases) determined from the promoter mapping data published by Celis et al. in an earlier paper (13) and that reported by him (215 bases) for ArgP activation experiments (12) and (ii) the promoter (for argK) mapped by his group (13) is within the argK coding region per the genome sequence published by Blattner et al. (5) (GenBank accession number AE000375, wherein argK is annotated ygfD).

Coeffectors of ArgP in yggA regulation.

The data from the yggA-lac expression studies suggest that Arg (or CAN) functions as a coeffector for the apo-regulator ArgP in activating yggA transcription. The ArgP-mediated induction of yggA expression is also observed upon the addition of the Arg biosynthetic precursor citrulline (Fig. (Fig.3)3) or ornithine (data not shown), most likely by virtue of its conversion to Arg within the cells. Similarly, yggA-lac induction by Arg-Ala is most simply explained as the consequence of the release of intracellular Arg after hydrolysis of the dipeptide.

The observation that yggA expression is elevated in an argR mutant raises the formal possibility that ArgR is a second regulator for yggA; however, the fact that the argR effect is also ArgP dependent (that is, it is not observed in an argR argP double mutant [Fig. [Fig.3])3]) indicates once again that it is the increased intracellular Arg concentration in the argR mutant which is responsible for yggA induction. This interpretation is supported by our observation that in a strain that was blocked in the Arg biosynthetic pathway, the introduction of an argR mutation was associated with only a twofold increase in yggA-lac expression, in contrast to a sevenfold increase in an isogenic Arg-prototrophic strain (all cultures were grown with 1 mM Arg) (data not shown).

We also observed that the addition of Lys or the dipeptide Lys-Ala nullified the activator function of ArgP for yggA expression. As explained above, the dipeptide supplementation experiments served to exclude the possibility that exogenously added Lys acts merely to reduce the intracellular Arg pool by competing with the latter's uptake (or reuptake). Two alternative models, therefore, are (i) that intracellular Lys competes with and prevents Arg from binding a single coeffector binding site in ArgP and (ii) that the latter has a second independent Lys-binding site.

It may be noted that the effect of intracellular Lys on E. coli yggA expression (negative) is opposite that on C. glutamicum lysE expression (positive) (3). Furthermore, whereas the latter is reported to be induced in cultures supplemented with histidine (3) or methionine (42), we observed that the expression of the former was reduced approximately 60% in the presence of either of these amino acids (data not shown). These differences in regulation suggest that although the two exporters share substantial similarities in structure and function, their physiological roles in their respective organisms may not be identical.

Likely physiological role(s) of YggA.

As mentioned above, very few amino acid export systems have so far been characterized in bacteria. It has been suggested that the physiological role of these exporters may be either to mediate the secretion of signaling molecules or to avoid the buildup of the substrate compounds to toxic levels in the cytoplasm (1, 9, 17).

In the case of E. coli YggA, three possibilities (not necessarily mutually exclusive) for its physiological function may be postulated. (i) The first possibility is that it acts as a safety valve to prevent the excessive accumulation of Arg following its uptake into the cells and the hydrolysis of nutrient Arg-containing peptides. Eggeling and coworkers showed that, in medium containing Lys-Ala or Arg-Ala, the C. glutamicum lysE mutant accumulates Lys or Arg, respectively, and is consequently growth inhibited (3, 42). We similarly found that, in comparison with an isogenic wild-type E. coli strain, the argP, and more so, the yggA mutants were inhibited for growth on glucose-minimal A medium supplemented with 5 mM Arg-Ala (data not shown). (ii) The second possibility is that YggA serves to maintain a correct balance between intracellular levels of the basic amino acids Arg and Lys, which may then explain why the latter acts to suppress yggA expression. (iii) The third possibility is that YggA has evolved to excrete CAN (which is a natural antimetabolite), which is perhaps supported by our finding that of all the substances that were tested, CAN was the most effective for inducing yggA transcription.

Redesignation of yggA as argO.

Based on the findings described above in support of an Arg exporter function for YggA, we propose that the gene be redesignated argO (for “arginine outward transport”).


We thank Mary Berlyn, Ian Booth, K. Isono, C. G. Lerner, and Jim Pittard for strains and plasmids. The yggA insertion mutants were obtained by Francois Gompel, and Rakesh Laishram performed the experiments with Arg-Ala and Lys-Ala dipeptides. We acknowledge V. Vamsee Krishna and T. Giri Babu for technical assistance, Mehar Sultana for primer synthesis, N. Nagesh for DNA sequencing, and members of the J. Gowrishankar laboratory for advice and discussions.

M.R.N. was a CSIR Research Fellow. J.G. is an honorary faculty member of the Jawaharlal Nehru Centre for Advanced Scientific Research.


1. Aleshin, V. V., N. P. Zakataeva, and V. A. Livshits. 1999. A new family of amino-acid-efflux proteins. Trends Biochem. Sci. 24:133-135. [PubMed]
2. Andrews, A. E., B. Lawley, and A. J. Pittard. 1991. Mutational analysis of repression and activation of the tyrP gene in Escherichia coli. J. Bacteriol. 173:5068-5078. [PMC free article] [PubMed]
3. Bellmann, A., M. Vrljiæ, M. Patek, H. Sahm, R. Krämer, and L. Eggeling. 2001. Expression control and specificity of the basic amino acid exporter LysE of Corynebacterium glutamicum. Microbiology 147:1765-1774. [PubMed]
4. Berlyn, M. K. B. 1998. Linkage map of Escherichia coli K-12, edition 10: the traditional map. Microbiol. Mol. Biol. Rev. 62:814-984. [PMC free article] [PubMed]
5. Blattner, F. R., G. Plunkett III, C. A. Bloch, N. T. Perna, V. Burland, M. Riley, J. Collado-Vides, J. D. Glasner, C. K. Rode, G. F. Mayhew, J. Gregor, N. W. Davis, H. A. Kirkpatrick, M. A. Goeden, D. J. Rose, B. Mau, and Y. Shao. 1997. The complete genome sequence of Escherichia coli K-12. Science 277:1453-1462. [PubMed]
6. Bollon, A. P., and H. J. Vogel. 1973. Regulation of argE-argH expression with arginine derivatives in Escherichia coli: extreme nonuniformity of repression and conditional repressive action. J. Bacteriol. 114:632-640. [PMC free article] [PubMed]
7. Bröer, S., and R. Krämer. 1991. Lysine excretion by Corynebacterium glutamicum. 1. Identification of a specific secretion carrier system. Eur. J. Biochem. 202:131-135. [PubMed]
8. Bröer, S., and R. Krämer. 1991. Lysine excretion by Corynebacterium glutamicum. 2. Energetics and mechanism of the transport system. Eur. J. Biochem. 202:137-143. [PubMed]
9. Burkovski, A., and R. Krämer. 2002. Bacterial amino acid transport proteins: occurrence, functions, and significance for biotechnological applications. Appl. Microbiol. Biotechnol. 58:265-274. [PubMed]
10. Celis, R. T. F. 1981. Chain-terminating mutants affecting a periplasmic binding protein involved in the active transport of arginine and ornithine in Escherichia coli. J. Biol. Chem. 256:773-779. [PubMed]
11. Celis, R. T. F. 1990. Mutant of Escherichia coli K-12 with defective phosphorylation of two periplasmic transport proteins. J. Biol. Chem. 265:1787-1793. [PubMed]
12. Celis, R. T. F. 1999. Repression and activation of arginine transport genes in Escherichia coli K-12 by the ArgP protein. J. Mol. Biol. 294:1087-1095. [PubMed]
13. Celis, R. T. F., P. F. Leadlay, I. Roy, and A. Hansen. 1998. Phosphorylation of the periplasmic binding protein in two transport systems for arginine incorporation in Escherichia coli K-12 is unrelated to the function of the transport system. J. Bacteriol. 180:4828-4833. [PMC free article] [PubMed]
14. Celis, R. T. F., H. J. Rosenfeld, and W. K. Maas. 1973. Mutant of Escherichia coli K-12 defective in the transport of basic amino acids. J. Bacteriol. 116:619-626. [PMC free article] [PubMed]
15. Covarrubias, L., and F. Bolivar. 1982. Construction and characterization of new cloning vehicles. VI. Plasmid pBR329, a new derivative of pBR328 lacking the 482-base-pair inverted duplication. Gene 17:79-89. [PubMed]
16. Daßler, T., T. Maier, C. Winterhalter, and A. Böck. 2000. Identification of a major facilitator protein from Escherichia coli involved in efflux of metabolites of the cysteine pathway. Mol. Microbiol. 36:1101-1112. [PubMed]
17. Eggeling, L., and H. Sahm. 2003. New ubiquitous translocators: amino acid export by Corynebacterium glutamicum and Escherichia coli. Arch. Microbiol. 180:155-160. [PubMed]
18. Franke, I., A. Resch, T. Daβler, T. Maier, and A. Böck. 2003. YfiK from Escherichia coli promotes export of O-acetylserine and cysteine. J. Bacteriol. 185:1161-1166. [PMC free article] [PubMed]
19. Glansdorff, N. 1996. Biosynthesis of arginine and polyamines, p. 408-433. In F. C. Neidhardt, R. Curtiss III, J. L. Ingraham, E. C. C. Lin, K. B. Low, B. Magasanik, W. S. Reznikoff, M. Riley, M. Schaechter, and H. E. Umbarger (ed.), Escherichia coli and Salmonella: cellular and molecular biology, 2nd ed. American Society for Microbiology, Washington, D.C.
20. Gowrishankar, J. 1985. Identification of osmoresponsive genes in Escherichia coli: evidence for participation of potassium and proline transport systems in osmoregulation. J. Bacteriol. 164:434-445. [PMC free article] [PubMed]
21. Han, J. S., H. S. Kwon, J.-B. Yim, and D. S. Hwang. 1998. Effect of IciA protein on the expression of the nrd gene encoding ribonucleoside diphosphate reductase in E. coli. Mol. Gen. Genet. 259:610-614. [PubMed]
22. Higashitani, A., N. Higashitani, S. Yasuda, and K. Horiuchi. 1994. A general and fast method for mapping mutations on the Escherichia coli chromosome. Nucleic Acids Res. 22:2426-2427. [PMC free article] [PubMed]
23. Hwang, D. S., and A. Kornberg. 1990. A novel protein binds a key origin sequence to block replication of an E. coli minichromosome. Cell 63:325-331. [PubMed]
24. Hwang, D. S., B. Thony, and A. Kornberg. 1992. IciA protein, a specific inhibitor of initiation of Escherichia coli chromosomal replication. J. Biol. Chem. 267:2209-2213. [PubMed]
25. Kleckner, N., J. Bender, and S. Gottesman. 1991. Uses of transposons with emphasis on Tn10. Methods Enzymol. 204:139-180. [PubMed]
26. Kohara, Y., K. Akiyama, and K. Isono. 1987. The physical map of the whole E. coli chromosome: application of a new strategy for rapid analysis and sorting of a large genomic library. Cell 50:495-508. [PubMed]
27. Kruse, D., R. Krämer, L. Eggeling, M. Rieping, W. Pfefferle, J. H. Tchieu, Y. J. Chung, M. H. Saier, Jr., and A. Burkovski. 2002. Influence of threonine exporters on threonine production in Escherichia coli. Appl. Microbiol. Biotechnol. 59:205-210. [PubMed]
28. LaRossa, R. A. 1996. Mutant selections linking physiology, inhibitors, and genotypes, p. 2527-2587. In F. C. Neidhardt, R. Curtiss III, J. L. Ingraham, E. C. C. Lin, K. B. Low, B. Magasanik, W. S. Reznikoff, M. Riley, M. Schaechter, and H. E. Umbarger (ed.), Escherichia coli and Salmonella: cellular and molecular biology, 2nd ed. American Society for Microbiology, Washington, D.C.
29. Lee, Y., H. Lee, J. Yim, and D. Hwang. 1997. The binding of two dimers of IciA protein to the dnaA promoter 1P element enhances the binding of RNA polymerase to the dnaA promoter 1P. Nucleic Acids Res. 25:3486-3489. [PMC free article] [PubMed]
30. Lee, Y. S., H. Kim, and D. S. Hwang. 1996. Transcriptional activation of the dnaA gene encoding the initiator for oriC replication by IciA protein, an inhibitor of in vitro oriC replication in Escherichia coli. Mol. Microbiol. 19:389-396. [PubMed]
31. Lerner, C. G., and M. Inouye. 1990. Low copy number plasmids for regulated low-level expression of cloned genes in Escherichia coli with blue/white insert screening capability. Nucleic Acids Res. 18:4631. [PMC free article] [PubMed]
32. Levina, N., S. Totemeyer, N. R. Stokes, P. Louis, M. A. Jones, and I. R. Booth. 1999. Protection of Escherichia coli cells against extreme turgor by activation of MscS and MscL mechanosensitive channels: identification of genes required for MscS activity. EMBO J. 18:1730-1737. [PMC free article] [PubMed]
33. Miller, J. H. 1992. A short course in bacterial genetics: a laboratory manual and handbook for Escherichia coli and related bacteria. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.
34. Reitzer, L., and B. L. Schneider. 2001. Metabolic context and possible physiological themes of σ54-dependent genes in Escherichia coli. Microbiol. Mol. Biol. Rev. 65:422-444. [PMC free article] [PubMed]
35. Rosen, B. P. 1973. Basic amino acid transport in Escherichia coli: properties of canavanine-resistant mutants. J. Bacteriol. 116:627-635. [PMC free article] [PubMed]
36. SaiSree, L., M. Reddy, and J. Gowrishankar. 2000. lon incompatibility associated with mutations causing SOS induction: null uvrD alleles induce an SOS response in Escherichia coli. J. Bacteriol. 182:3151-3157. [PMC free article] [PubMed]
37. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.
38. Saroja, G. N., and J. Gowrishankar. 1996. Roles of SpoT and FNR in NH4+ assimilation and osmoregulation in GOGAT (glutamate synthase)-deficient mutants of Escherichia coli. J. Bacteriol. 178:4105-4114. [PMC free article] [PubMed]
39. Schell, M. A. 1993. Molecular biology of the LysR family of transcriptional regulators. Annu. Rev. Microbiol. 47:597-626. [PubMed]
40. Schwartz, J. H., and W. K. Maas. 1960. Analysis of the inhibition of growth produced by canavanine in Escherichia coli. J. Bacteriol. 79:794-799. [PMC free article] [PubMed]
41. Thony, B., D. S. Hwang, L. Fradkin, and A. Kornberg. 1991. iciA, an Escherichia coli gene encoding a specific inhibitor of chromosomal initiation of replication in vitro. Proc. Natl. Acad. Sci. USA 88:4066-4070. [PMC free article] [PubMed]
42. Vrljic, M., H. Sahm, and L. Eggeling. 1996. A new type of transporter with a new type of cellular function: l-lysine export from Corynebacterium glutamicum. Mol. Microbiol. 22:815-826. [PubMed]
43. Zakataeva, N. P., V. V. Aleshin, I. L. Tokmakova, P. V. Troshin, and V. A. Livshits. 1999. The novel transmembrane Escherichia coli proteins involved in the amino acid efflux. FEBS Lett. 452:228-232. [PubMed]

Articles from Journal of Bacteriology are provided here courtesy of American Society for Microbiology (ASM)
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • Compound
    PubChem Compound links
  • Gene
    Gene links
  • GEO Profiles
    GEO Profiles
    Related GEO records
  • MedGen
    Related information in MedGen
  • Pathways + GO
    Pathways + GO
    Pathways, annotations and biological systems (BioSystems) that cite the current article.
  • Protein
    Published protein sequences
  • PubMed
    PubMed citations for these articles
  • Substance
    PubChem Substance links
  • Taxonomy
    Related taxonomy entry
  • Taxonomy Tree
    Taxonomy Tree

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...