ABSTRACT
In Escherichia coli, argO encodes an exporter for l-arginine (Arg) and its toxic analogue canavanine (CAN), and its transcriptional activation and repression, by Arg and l-lysine (Lys), respectively, are mediated by the regulator ArgP. Accordingly argO and argP mutants are CAN supersensitive (CANss). We report the identification of ybjE as a gene encoding a predicted inner membrane protein that mediates export of Lys, and our results confirm the previous identification with a different approach of YbjE as a Lys exporter, reported by Ueda and coworkers (T. Ueda, Y. Nakai, Y. Gunji, R. Takikawa, and Y. Joe, U.S. patents 7,629,142 B2 [December 2009] and 8,383,363 B1 [February 2013] and European patent 1,664,318 B1 [September 2009]). ybjE was isolated as a multicopy suppressor of the CANss phenotype of a strain lacking ArgO. The absence of YbjE did not confer a CANss phenotype but instead conferred hypersensitivity to the lysine antimetabolite thialysine and led to growth inhibition by the dipeptide lysylalanine, which is associated with elevated cellular Lys content. YbjE overproduction resulted in Lys excretion and syntrophic cross-feeding of a Lys auxotroph. Constitutive overexpression of argO promoted Lys cross-feeding that is indicative of a latent Lys export potential of ArgO. Arg modestly repressed ybjE transcription in an ArgR-dependent manner, and ArgR displayed Arg-sensitive binding to the ybjE promoter region in vitro. Our studies suggest that the reciprocal repression of argO and ybjE, respectively, by Lys and Arg confers the specificity for basic amino acid export by distinct paths and that such cross-repression contributes to maintenance of cytoplasmic Arg/Lys balance. We propose that YbjE be redesignated LysO.
IMPORTANCE This work ascribes a lysine export function to the product of the ybjE gene of Escherichia coli, leading to a physiological scenario wherein two proteins, ArgO and YbjE, perform the task of separately exporting arginine and lysine, respectively, which is distinct from that seen for Corynebacterium glutamicum, where the ortholog of ArgO, LysE, mediates export of both arginine and lysine. Repression of argO transcription by lysine is thought to effect this separation. Accordingly, ArgO mediates lysine export when repression of its transcription by lysine is bypassed. Repression of ybjE transcription by arginine via the ArgR repressor, together with the lysine repression of argO effected by ArgP, is indicative of a mechanism of maintenance of arginine/lysine balance in E. coli.
INTRODUCTION
Bacteria possess membrane exporters for a variety of compounds. Their activities to a large extent are thought to play an adaptive role in mitigating the detrimental effects on bacterial growth caused by the presence of biotic stresses, such as those imposed by antibiotics, heavy metals, and other toxic compounds, in their natural environments. While it is easy to come to terms with the existence of proteins that mediate export of compounds described above, the presence of specific export systems for cellular metabolites such as sugars and amino acids appears somewhat enigmatic. For example, Escherichia coli encodes multiple proteins for the export of sugars, such as glucose and lactose (1), and for arabinose (2, 3). In addition, the occurrence of proteins mediating export of amino acids, such as alanine (4), arginine (5), aromatic amino acids (6), cysteine (7, 8), leucine (9), threonine (10, 11) and valine (12), in E. coli has been reported. While the physiological basis for the existence of amino acid exporters is not clear, their occurrence is relatively widespread in bacteria (13, 14). It is thought that an amino acid exporter may serve to contribute to the fitness of an organism during conditions of metabolic imbalance resulting from excessive levels of its amino acid substrate in the cytoplasm (13, 14). An alternative view could be that export of the amino acid occurs by chance, with the cognate substrate being a naturally occurring structurally related antimetabolite present in the environment. As an example of the latter, one mechanism contributing to the resistance of E. coli to the plant-derived antimetabolite canavanine (CAN), an l-arginine analogue, involves its export by ArgO, the ortholog of the basic amino acid exporter LysE of Corynebacterium glutamicum, following its uptake (5). Harnessing the amino acid export potential of a given bacterium has found widespread application in the commercial production of amino acids (reviewed in references 13 and 14).
In C. glutamicum, LysE mediates export of the basic amino acids l-arginine (Arg) and l-lysine (Lys) (15, 16), whereas its ortholog in E. coli ArgO so far has been thought to promote export only of Arg (5), and evidence for ArgO-mediated Lys export is as yet unavailable. In the present study, we ascribe a Lys export function to the product of ybjE, present on the E. coli chromosome. ybjE was isolated on the basis of its ability to suppress the canavanine-supersensitive (CANss) phenotype of a strain lacking ArgO (5) when present on a multicopy plasmid. A strain lacking YbjE was impaired for growth when challenged with the presence of the lysine antimetabolite thialysine in the medium and displayed reduced fitness under the condition of elevated cytoplasmic Lys content, achieved by the inclusion of a Lys-containing dipeptide in the medium. In addition, experiments on the genetic regulation of ybjE implicated the involvement of Arg in mediating transcriptional repression of ybjE via the arginine repressor ArgR. Lastly, we describe the Lys export potential of ArgO, which establishes a premise for the existence of distinct pathways of basic amino acid export in E. coli, as opposed to the situation in C. glutamicum, where one protein, LysE, promotes the export of both. Our results are consistent with previous reports by Ueda and coworkers, who employed an approach different from ours and have described the identification of YbjE as a Lys exporter in multiple patents (17, 18, 19).
MATERIALS AND METHODS
Growth media, bacterial strains, and plasmids.In this study, LB medium and glucose minimal A medium (MA medium) were routinely used as rich and defined synthetic media, respectively (20). The antibiotics ampicillin (Amp), chloramphenicol (Cm), kanamycin (Kan), tetracycline (Tet), spectinomycin, and trimethoprim and the inducer of the Ptrc promoter isopropyl-β-d-thiogalactoside (IPTG) were used at appropriate concentrations, and antibiotic selection for plasmid maintenance was employed where required. The E. coli K-12 and E. coli B strains employed in this study and their genotypes are listed in Table 1. Strain construction was performed using P1 transduction (20) from appropriate strains of the Keio collection (21). The routine temperature for growth of strains was 37°C. If required, the kanamycin resistance determinant marking the null mutation sourced from the Keio collection was excised by treatment with the plasmid pCP20 as described earlier (22). The molecular characterization of the argO::205Tn10dTet insertion in argO that abolishes ArgO function has been described previously (5). The plasmids used in this study are derivatives of the plasmids pACYC184 (23), pTrc99A (24), pCL1920 (25), and pMU575 (26), and their construction is described in Table 1. Standard procedures for PCR, cloning, and overlap extension PCR-based site-directed mutagenesis were followed for their construction (27). The oligonucleotide primers used in this study are listed in Table S1 in the supplemental material.
E. coli strains and plasmids
To study the transcriptional regulation of ybjE, segments of DNA bearing the 5′ promoter/regulatory region of ybjE and those bearing modifications in the aforementioned DNA sequence, generated by overlap extension PCR, were placed upstream of a promoterless lacZ present on the single-copy promoter probe plasmid pMU575.
Phenotypic tests and other procedures.The levels of tolerance to canavanine (CAN) and thialysine were assessed by spotting 10-fold serial dilutions of cultures of appropriate strains on the surfaces of MA plates containing defined concentrations of CAN and thialysine. Tolerance to CAN was also used as a phenotypic indicator of the functionality of hemagglutinin-tagged ArgO and hexahistidine-tagged ArgR expressed from the plasmids pHYD2835 and pHYD2847, respectively. Presence of the plasmid pHYD2835 complemented the CANss phenotype of an argO-null mutant (5), indicating that the epitope-tagged ArgO retained its normal function. Similarly, an argR argO double mutant is rendered more resistant to CAN than an argO mutant, owing to elevated levels of intracellular Arg in the former strain due to the absence of ArgR (28), which mitigates to some extent the growth-inhibitory effects of CAN. The presence of pHYD2847 in an argR argO double mutant impaired its resistance to CAN in MA medium containing 2 μg/ml CAN, indicating that ArgR encoded on the plasmid pHYD2847 was biologically active.
Growth measurements (see Fig. 3) were undertaken by inoculation of stationary-phase cultures of the chosen strains grown in MA medium at a dilution of 1:100 in the appropriate medium, and the A600 of the cultures was measured over time. Representative curves obtained from two independent experiments are shown. β-Galactosidase specific activities in Miller units in exponential-phase cultures were measured by the method of Miller (20), and values reported herein are means ± standard errors (SE) of values obtained from two independent measurements performed in duplicate. Plasmid clones able to suppress the canavanine-supersensitive (CANss) phenotype of the argO mutant strain GJ4823 were isolated following the transformation of GJ4823 with a plasmid library bearing Sau3AI-digested E. coli K-12 chromosomal DNA inserts present within the BamHI site located in tetA of pACYC184 (23). Transformants exhibiting Cm- and CAN-resistant (CANr) phenotypes were isolated on MA Cm plates containing 2 μg/ml CAN. Following the demonstration that the CANr phenotypes were plasmid borne, junction sequences in the chosen plasmids were determined with a pair of vector-based tet primers, 5′-CGCCGAAACAAGCGCTCATGAGCC-3′ and 5′-CTATGCGCACCCGTTCTCGGAGCAC-3′. The coordinates of chromosomal segments of DNA present in the plasmids isolated from the plasmid library are as per reference 29.
The Lys export phenotype of ybjE was assessed by a Lys cross-feeding assay which involved the construction of a Lys auxotroph of MC4100 bearing the lysA-null mutation, GJ9060, that was also rendered defective for the activities of the three peptide uptake systems due to the presence of null mutations in dppB, tppB, and oppB. Thus, GJ9060 grew on MA agar containing Lys but did not grow in Lys-Ala (1 mM)-supplemented MA agar. Lys export was visualized by the assay of halo formation of GJ9060 bearing the plasmid pHYD3025 (vector), representing its syntrophic growth, when seeded into an MA agar plate containing Amp, 1 mM IPTG, and 1 mM Lys-Ala, on the surface of which 106 cells of MC4100 bearing the plasmid pHYD2836 or the vector control pHYD3025 were spotted. The plates were photographed after 30 h of incubation.
Measurements of cellular and extracellular Lys content.For measurement of cellular Lys content, stationary-phase cultures of the appropriate strains grown in LB medium were washed with MA medium, subcultured in the same medium, and grown until early log phase. The cultures obtained were each inoculated, at an A600 of 0.02, into two flasks containing 30 ml MA medium, one of which was also supplemented with the lysylalanine (Lys-Ala) dipeptide at 1 mM, and the cultures were grown further for 3.5 h. A cell suspension with an adjusted A600 of 7 was generated from these cultures, and the cells were separated from the medium by centrifugation of 1 ml of the cell suspension at 13,000 rpm at room temperature for 5 min, through 300 μl of an organic layer comprising bis-(2-ethylhexyl) phthalate and dibutyl phthalate at a ratio of 1:2 (vol/vol). The supernatant above the organic layer was removed, and the likelihood of contamination with Lys either originating from the growth medium or present in Lys-Ala was minimized by washing the organic layer seven times with water. Following a final centrifugation, the organic layer was removed and the cell pellet was treated with 200 μl of 20% perchloric acid, vortexed, and sonicated in Eppendorf tubes immersed in an ice-water mixture using a Diagenode Bioruptor UCD-200 on a low-wave output power of 160 W using alternate (30-s) on/off cycles for 10 min. Following a centrifugation of this solution at 13,000 rpm at 4°C for 5 min, its supernatant was collected, and its acidity was neutralized by the addition of 10 N KOH. After a clarification by centrifugation, the supernatant was concentrated by vacuum desiccation at 30°C to a volume of 120 μl, and half of it was used to estimate cellular Lys content on the Agilent Technologies (model 1200 series) HPLC system after a precolumn derivatization with phenylisothiocyanate (PITC) using Waters Pico-tag chemistry to detect the product. Appropriate PITC-derivatized standards of Lys monohydrochloride were also processed.
For measurement of extracellular Lys content, stationary-phase cultures of the parent MC4100 bearing the vector pHYD3025 and the plasmid pHYD2836 were washed with MA medium and subcultured in the same medium containing 1 mM IPTG and 1 mM Lys-Ala with the appropriate antibiotic selection and grown for 8 h at 37°C. Three hundred microliters of each culture was passed through an organic layer [30 μl; bis-(2-ethylhexyl) phthalate and dibutyl phthalate in a ratio of 1:2 (vol/vol)]. Two hundred microliters of the supernatant above the organic layer was aspirated and processed for HPLC detection of Lys as described above. Samples that did not contain cells were processed similarly, and the amount of contaminating Lys amounting to 33 ± 1.3 nmol per ml in the preparation of Lys-Ala was determined. The area under the HPLC peak corresponding to contaminating Lys was subtracted from the Lys peaks obtained from grown cultures prior to normalization by A600 of the cultures. The values reported for cellular and extracellular Lys content are means ± SE from two independent experiments.
Purification of ArgR.Strain GJ9048 bearing the plasmid pHYD2847 was inoculated into 5 ml of LB Amp broth. Following overnight growth, the culture was inoculated into 1 liter of LB Amp broth and grown to an A600 of 0.4, after which 1 mM IPTG was added to the culture, and the culture was further incubated with shaking for 6 h. Following centrifugation of the culture at 6,000 rpm for 15 min at room temperature, the cell pellet obtained was resuspended in 20 ml of lysis buffer (20 mM Tris-HCl [pH 8.0], 300 mM NaCl, 10 mM MgCl2, 10 mM imidazole, 1 mM β-mercaptoethanol, and 1 mM phenylmethylsulfonyl fluoride). Cell lysis was performed by subjecting the cell suspension to sonication, and the cell lysate obtained was centrifuged at 13,000 rpm at 4°C for 20 min. The supernatant was loaded onto a nickel-nitrilotriacetic acid (Ni-NTA) (Qiagen) chromatographic column preequilibrated with lysis buffer. After passage of the lysate, the chromatographic column was washed five times with 10 ml of washing buffer (lysis buffer with 30 mM imidazole), and the bound ArgR was eluted with 10 ml of elution buffer (lysis buffer with 250 mM imidazole). The purity of ArgR was determined by SDS-PAGE to be greater than 95%. Purified ArgR was dialyzed against buffer A (lysis buffer with 100 mM imidazole and 10 mM β-mercaptoethanol) at a protein-to-dialysis buffer ratio of 1:200 (vol/vol) for 1 h with the Tube-o-dialyzer system with a 15-kDa cutoff (G-Biosciences). Another dialysis was performed for 1 h against buffer B (buffer A but with 50 mM imidazole). The purified protein solution was subjected to a final dialysis for 12 h against buffer C (buffer A lacking imidazole). Protein concentration was estimated by using the Bradford reagent (Sigma), and the purified ArgR was stored at −80°C in a buffer containing 20 mM Tris-HCl (pH 8.0), 300 mM NaCl, 10 mM MgCl2, 10 mM β-mercaptoethanol, and 45% glycerol.
EMSAs with ArgR.DNA templates for electrophoretic mobility shift assays (EMSAs) for determination of binding of ArgR to the cis regulatory region of ybjE were generated by PCR. In the study described here, the proficiency of binding of ArgR to DNA templates bearing specific deletions and substitutions in the cis regulatory region of ybjE was also tested. In both instances, the desired region was first amplified either from pMU575 bearing the wild type cis regulatory region or from its derivatives (Table 2) bearing the aforementioned modifications with the primer pair JGPMUFP (5′-TCCCCACATCACCAGCAA-3′) and JGPMUGALKRP (5′-CAGAGATTGTGTTTTTTCTTTCAG-3′), which primes DNA synthesis from sites located upstream and downstream of the cloned region present on pMU575 in the various plasmids. The PCR products were gel purified and used as templates for PCR amplification with the primer pair JGLYBJEPFA (5′-ACCCCCGGGCTCTGGCGAAC-3′) and JGLYBJEDRP (5′-TATGCTCTAGAAACAGCCCAGAAAACATGA-3′), a segment of the cis regulatory region of ybjE from −301 to +55 relative to the start site of ybjE transcription that was determined by primer extension. For EMSA reactions, PCR products that were generated as described above were 5′ end labeled with T4 polynucleotide kinase and [γ-32P]ATP, and the EMSA reaction mixtures (20 μl) consisted of 0.5 nM labeled DNA, 1 μg of bovine serum albumin (BSA), and 1 μg of poly(dI·dC) present in the EMSA binding buffer, which comprised 10 mM Tris-HCl (pH 8.0), 50 mM NaCl, 5 mM MgCl2, 5 mM β-mercaptoethanol, and 22.5% glycerol, and various hexamer concentrations of ArgR. Where required, 5 mM Arg (Arg monohydrochloride) was included in the binding reaction mixture. EMSA reaction mixtures of labeled DNA and protein were incubated on ice for 30 min, and DNA-protein complexes were resolved by electrophoresis on a 4% native polyacrylamide gel that was cast with 1× Tris-borate buffer (30) and 1 mM Arg with the electrophoresis buffer comprising 1 mM Arg and 0.5× Tris-borate. The electrophoresis was performed at constant current for 4 to 6 h at 4°C. Radioactive bands following gel drying were visualized with a Fujifilm FLA-9000 scanner.
Effect of Arg and ArgR on ybjE-lac expression from plasmid pHYD2885 and in derivatives bearing deletions or substitutions in the cis regulatory region of ybjE
RESULTS
Identification of ybjE as a multicopy suppressor of the CANss phenotype of an argO mutant.As described above, a previous study showed that a strain lacking ArgO is severely impaired for growth in MA medium containing CAN, yielding a CANss phenotype (5). In order to identify new genes mediating Arg export, we sought multicopy suppressors of the CANss phenotype of an argO mutant. For this purpose, an E. coli genomic library constructed on the backbone of the plasmid pACYC184, with chromosomal DNA inserts present within the BamHI site, located in tetA of pACYC184, was introduced into the argO mutant derivative of MC4100, GJ4823, by transformation. Twenty-three transformants from a total pool of approximately 105 Cm-resistant transformants, exhibiting a CANr phenotype, were isolated on MA agar plates supplemented with Cm and 2 μg/ml CAN. Of these, 20 yielded an argO-specific PCR product when PCR was performed on plasmid DNA isolated using a primer pair specific for argO, indicating that the CANr phenotype they exhibited resulted from the presence and expression of a plasmid-borne wild-type argO. One such plasmid, designated pHYD2834, that upon retransformation into GJ4823 rendered it CANr was saved, and DNA junction sequencing showed that it carried a 3.7-kb insert spanning the chromosomal coordinates 3063431 to 3067219 (Table 1). This showed that the chromosomal insert in pHYD2834 encoded a functional ArgO.
Plasmids designated pHYD2833, pHYD2833.1, and pHYD2833.2, isolated from the remaining three transformants, upon retransformation into GJ4823 also rendered it CANr. DNA junction sequencing showed that they all bore an identical segment of chromosomal DNA extending from chromosomal coordinate 912287 to 914782 bearing a complete ybjE ORF, flanked by two incomplete ybjW and aqpZ ORFs. In addition, the chromosomal insert on the three plasmids was present in the same orientation within the BamHI site of pACYC184 (Fig. 1A and Table 1). Derivatives of GJ4823 bearing the plasmid pHYD2836, which contained a minimal ybjE ORF under the expression control of the Ptrc promoter, were rendered CANr in the presence of 1 mM IPTG, whereas those bearing the vector pHYD3025 remained CANss (data not shown). The CANr phenotype yielded by the presence of the plasmid pHYD2833 was partial in that the CANr phenotype was prominent on MA agar plates containing 2 but not 20 μg/ml CAN (Fig. 1B). Expression of ybjE from the Ptrc promoter of the plasmid pHYD2836 with IPTG also conferred a partial CANr phenotype to GJ4823 (data not shown). Derivatives of GJ4823 bearing the plasmid pHYD2834, however, displayed a robust CANr phenotype in MA agar plates containing CAN up to 20 μg/ml (Fig. 1B). We found that in MC4100 (parent), whereas the absence of ArgO led to a CANss phenotype, the absence of YbjE did not lead to any discernible sensitivity to CAN on MA agar plates containing 20 μg/ml CAN. However, in a medium with a lower CAN concentration (1 μg/ml), the argO ybjE double mutant was less resistant to CAN than the argO mutant (Fig. 2). Lastly, examination of the amino acid sequence of YbjE suggested that it was a protein of a predicted membrane location bearing eight putative transmembrane segments.
Suppression of the CANss growth phenotype of an argO-null mutant by increased dosage of ybjE. (A) Schematic representation (to scale) of the extent of chromosomal DNA borne on plasmids pHYD2833, pHYD2833.1, and pHYD2833.2. The chromosomal coordinates marking the extremities of the insert and the extents of the 3′- and 5′-truncated ybjW and aqpZ genes, respectively, are indicated. (B) Suppression of the CANss growth of an argO mutant by the plasmid pHYD2833. Tenfold serial dilutions of cultures of the parent (MC4100) and its argO::205Tn10dTet derivative GJ4823 (argO) bearing the indicated plasmids were spotted on the surfaces of an MA agar plate (I) and MA agar plates containing CAN at 2 (II) and 20 (III) μg/ml.
Growth phenotype of the ybjE-null mutant in the presence of CAN. Tenfold serial dilutions of cultures of MC4100 (parent), GJ9026 (ΔybjE::Kan ybjE), GJ4823 (argO205::Tn10dTet argO), and GJ9028 (argO205::Tn10dTet ΔybjE::Kan argO ybjE) were spotted on the surface of an MA agar plate (I) and MA agar plates containing CAN at 20 (II) and 1 (III) μg/ml.
Evidence that ybjE mediates Lys export.The aforementioned observation that increased dosage of YbjE mediated resistance to CAN in the absence of a functional ArgO hinted that the mechanism by which it did so may involve enhanced export of CAN or that an increased dosage of ybjE, perhaps by mediating the export of another basic amino acid, competed with CAN uptake. In order to clarify this, we first examined effects of lysylalanine (Lys-Ala) and arginylalanine (Arg-Ala) dipeptides on the growth of MC4100 and its derivatives bearing single or double deficiencies of ArgO and YbjE. Dipeptides provide a means to increase the intracellular level of an amino acid, as their entry into the cytoplasm occurs via the peptide uptake systems. The Lys-Ala dipeptide has previously been employed as a means to increase the intracellular levels of Lys in C. glutamicum (31). We found that the ybjE mutant was impaired for growth in MA medium containing Lys-Ala but not in MA medium containing Arg-Ala or the histidinylalanine (His-Ala) dipeptides (Fig. 3A to D). The argO mutant grew at rates comparable to that of MC4100 in all the above-mentioned media, and the argO mutation did not exacerbate further the impaired growth of the ybjE-null mutant in a medium containing the Lys-Ala dipeptide.
Growth impairment of the ybjE mutant by the Lys-Ala dipeptide and Lys-Ala-mediated elevation of cellular Lys content in the ybjE mutant. The parent MC4100 (open squares), GJ4823 (argO205::Tn10dTet; filled squares), GJ9026 (ΔybjE::Kan; open circles), and GJ9028 (argO205::Tn10dTet ΔybjE::Kan; filled circles) were grown in MA medium (A) and MA medium containing 1 mM Arg-Ala (B), His-Ala (C), and Lys-Ala (D) dipeptides, and the absorbance at 600 nm of the cultures was monitored. (E) Cellular Lys content in exponentially growing cultures of MC4100 (parent), GJ4823 (argO), GJ9026 (ybjE), and GJ9028 (argO ybjE) (white bars) and in their counterparts following their transient exposure to 1 mM Lys-Ala (black bars).
We measured cellular Lys content in MC4100 and its derivatives bearing single or double null mutations in argO and ybjE following their transient exposure to Lys-Ala by HPLC. Upon exposure to Lys-Ala, the ybjE mutant (and its argO derivative) displayed an approximately 3-fold elevation in cellular Lys content in comparison to MC4100 and its argO derivative, and the absence of ArgO did not cause any alteration in cellular Lys levels (Fig. 3E). Assuming that an A600 of 1 corresponds to 109 cells (20) and the cellular volume is 1 μm3 (32, 33), estimates of the intracellular Lys concentrations in the parent and the ybjE mutant in the presence of Lys-Ala can be arrived at, namely, 6.8 and 21 mM, respectively. The level of another basic amino acid, Arg, was comparable in cultures of the above-mentioned strains that were treated with Lys-Ala and those that were untreated (data not shown). In addition, overexpression of ybjE from the plasmid pHYD2836 in MC4100 led to increased cross-feeding of the lysA dppB tppB oppB derivative of MC4100, GJ9060, in comparison to the vector-bearing control, which was indicative of increased ybjE-mediated Lys export (Fig. 4A). Lastly, we estimated Lys in the culture medium following 8 h of growth of MC4100 bearing the vector or pHYD2836 in a medium containing Amp, 1 mM IPTG, and 1 mM Lys-Ala and found that overexpression of ybjE from the plasmid pHYD2836 mediated increased export of Lys in comparison to its haploid ybjE+ (vector-bearing) counterpart (Fig. 4B). From Fig. 4B, the normalized extracellular Lys concentration yielded by cultures of MC4100 bearing the vector or the plasmid pHYD2836 can be estimated to be 92 and 162 μM, respectively.
Visualization by cross-feeding of YbjE-mediated Lys export and extracellular Lys content following heterologous overexpression of ybjE. (A) Cultures of MC4100 bearing the plasmid pHYD3025 (vector) (I) and pHYD2836 (II) were spotted on the surface of an MA agar plate containing 1 mM Lys-Ala, 1 mM IPTG, and 1 μg/ml tetrazolium chloride. In addition, the plate was seeded with cells of strain GJ9060, which is a lysA dppB oppB tppB derivative of MC4100 bearing the vector and imaged after 30 h of incubation. (B) Lysine content in the medium following 8 h of growth of MC4100 (parent) bearing the vector or the plasmid pHYD2836 in MA medium supplemented with 1 mM Lys-Ala and 1 mM IPTG.
Thialysine phenotypes associated with deficiency or overexpression of ybjE.Thialysine (S-aminoethyl-cysteine) is a toxic analogue of Lys, and given the evidence above that YbjE mediates export of Lys, we gauged the growth of MC4100 and its derivatives that were singly and doubly deficient for ArgO and YbjE on MA agar plates containing thialysine and found that severe impairment of growth was seen only in strains bearing the ybjE-null mutation and that the absence of ArgO did not cause any sensitivity to thialysine (Fig. 5A). Furthermore, overexpression of ybjE from the plasmid pHYD2836 conferred increased resistance to thialysine (Fig. 5B).
Hypersensitivity and enhanced resistance to thialysine caused by the absence and overexpression of ybjE. (A) Tenfold serial dilutions of cultures of MC4100 (parent) and its argO205::Tn10dTet (argO), ΔybjE::Kan (ybjE), and argO205::Tn10dTet ΔybjE::Kan (argO ybjE) derivatives, GJ4823, GJ9026, and GJ9028, respectively, were spotted on the surface of an MA agar plate (I) and MA agar plates containing thialysine at 1 (II) and 10 (III) μg/ml. (B) Tenfold serial dilutions of MC4100 (parent) and GJ9026 (ybjE) bearing the plasmids pHYD3025 (vector) and pHYD2836, in which expression of ybjE is under the IPTG inducible Ptrc promoter, were spotted on the surface of an MA agar plate (I) and MA agar plates containing thialysine at 1 (II) and 10 (III) μg/ml. The indicated plates were also supplemented with 1 mM IPTG.
Characterization of the ybjE cis regulatory region.In order to delineate the core promoter elements and associated transcriptional regulation of ybjE, we initially performed primer extension and found that the ybjE message initiated from a G residue located 37 bases upstream of the predicted translational initiation codon of ybjE (Fig. 6A and data not shown). The in vivo expression level of the ybjE promoter was gauged by construction of the plasmid pHYD2885 in which the cis regulatory region of ybjE spanning a region from −456 to +55 was placed upstream of a promoterless lacZ in the single-copy plasmid pMU575 (Table 2). The location of the −10 and −35 promoter elements of ybjE was ascertained by measurements of ybjE-lac activity from derivatives of plasmid pHYD2885 bearing site-specific nucleotide substitutions that abolished ybjE-lac activity, engineered within a pair of hexameric sequences, 5′-TAGTGT-3′ and 5′-TTTACT-3′ (Fig. 6A and data not shown). Another feature of the cis regulatory region of ybjE that presented itself was the occurrence of an inverted repeat extending from −333 to −302, whose removal did not alter ybjE-lac activity (Table 2).
Nucleotide sequence of the cis regulatory region of ybjE and its interaction with ArgR in vitro. (A) The nucleotide sequence of ybjE shown extends from −333 to +55 relative to the start site of the ybjE message (indicated as +1), and the predicted initiation codon of ybjE is underlined. The −10 and −35 sequences of the ybjE promoter are shown as dashed boxes, and an inverted repeat extending from −333 to −302 is marked by a pair of convergent horizontal arrows. The extents of 18-bp deletions in the cis regulatory region of ybjE present in plasmids pHYD2890 (Δ1), pHYD2891 (Δ2), and pHYD2892 (Δ3) are indicated. The T-to-G substitution in the plasmid pHYD2888 and substitutions of 4 bases (4S) in the plasmid pHYD2893 are marked with an upward arrow and asterisks, respectively, and the base substitutions are italicized. Pairs of EMSA of ArgR with the cis regulatory region of ybjE (wild) and with its derivatives bearing the T-to-G base substitution at −58 (B), Δ2 (−64 to −47) (C), substitutions of 4 bases (4S) lying within −42 to −37 of ybjE (D), and Δ3 (−43 to −26) (E). The DNA templates used in the EMSA span a region of ybjE from −301 to +55 with the substitutions and deletions contained within, and the positions of the free DNA and the ArgR DNA complex are indicated with open and filled triangles, respectively. Hexamer concentrations of ArgR are indicated. Arg at 1 mM was incorporated in the electrophoresis buffer and gel, and the binding reaction mixture contained Arg at 5 mM.
ArgR-mediated repression of ybjE expression by Arg.Upon testing the effects of external supplementation of MA medium with Arg and Lys at 10 mM, we found that the presence of Lys did not affect the magnitude of ybjE-lac expression, whereas Arg supplementation led to an approximately 1.5- to 2-fold repression of ybjE-lac (Table 2). ybjE-lac expression was unaffected by the presence of 10 mM Lys-Ala in MA medium, whereas it was repressed 1.6-fold by the presence of 10 mM Arg-Ala (data not shown). Arg repression of ybjE-lac was absent in a strain lacking ArgR, the repressor protein of the Arg regulon (reviewed in references 34 and 35), and was independent of the presence or absence of the inverted repeat located in the cis regulatory region of ybjE (Table 2). The presence of both Arg and Lys in the medium did not alter β-galactosidase activity from the gadA-lac transcriptional fusion present on the plasmid pMU575, indicating that the repression by Arg was specific to the ybjE promoter (data not shown). The magnitude of ybjE-lac was not elevated by the absence of ArgR (Table 2), which is in contrast to that seen for genes of the Arg regulon that are significantly derepressed in an argR mutant (34, 35).
To test whether the cis regulatory region of ybjE contains an ArgR binding site(s), we purified ArgR and showed that in vitro ArgR bound only in the presence of Arg to a DNA fragment extending from −301 to +55 of ybjE, bearing the ybjE core promoter (Fig. 6; also, see Fig. S1 in the supplemental material). Our ArgR preparation also displayed proficient and Arg-dependent binding to a DNA fragment containing the promoter/operator region of argF, known to bear a cognate binding site for ArgR (34, 35) (see Fig. S1 in the supplemental material). ArgR bound with greater avidity to the argF DNA template than the ybjE DNA template, indicating that the binding site for ArgR in the cis regulatory region of ybjE represented a weak ArgR binding site (see Fig. S1 in the supplemental material).
In E. coli ArgR represses the promoters of genes encoding enzymes of Arg biosynthesis, in the presence of Arg by binding in a hexameric state to a pair of 18-bp imperfect palindromes (ARG boxes) separated by a 3-bp spacer and overlapping the promoters of Arg biosynthesis genes to different extents (34, 35). Assuming that the mode of binding of ArgR to the DNA of the cis regulatory region of ybjE was similar to that seen in the case of its cognate targets in promoters of genes involved in Arg biosynthesis, we searched for a pair of 18-bp DNA sequences located in close proximity to the ybjE promoter. We identified an 18-bp sequence extending from −64 to −47 bearing an imperfect match to the consensus sequence of the ARG box (see Fig. S2 in the supplemental material), whose deletion (Δ2) (Fig. 6A), abolished ArgR repression of ybjE-lac (Table 2) and abolished the binding of ArgR to the ybjE promoter DNA fragment (Fig. 6C). An earlier study on ArgR/ARG box interaction has pointed to the importance of a T base at the seventh position (T7), with minor groove occupancy, which is highly conserved in DNA sequences of all ARG boxes, mediating contact with ArgR (36). Furthermore, in all ARG box sequences there is absence of a guanine base at the T7 position (36). Another study has shown that replacement of T7 with G led to loss of ArgR repression at the promoter of hisJ (37). Substitution of the corresponding T7 (at −58) to G in the probable ARG box located within −64 to −47 in ybjE led to loss of ArgR repression of ybjE-lac (Table 2) and impaired the association of ArgR with the ybjE promoter DNA fragment bearing the above-mentioned base substitution (Fig. 6B). Deletion of an 18-bp segment of DNA from −85 to −68 (Δ1) (Fig. 6A), did not alter ArgR repression ybjE-lac (Table 2), indicating that the second ARG box, if present and required for ArgR binding to the ybjE promoter, may lie downstream of −47 in ybjE. We engineered substitutions of 4 bases, A to C, G to T, T to G, and T to G at −37, −40, −41, and −42, respectively, of ybjE (4S) (Fig. 6A) and found that the presence of 4S led to an overall reduction in the magnitude of ybjE-lac and loss of repression by ArgR (Table 2) and impaired the binding of ArgR to the ybjE promoter DNA template bearing 4S (Fig. 6D). The substitution(s) of the 4S cluster that is responsible for mediating the reduction in ybjE-lac and for affecting ArgR binding to the ybjE DNA template remains unknown. Lastly, ArgR displayed impaired binding to a ybjE promoter DNA fragment lacking a region from −43 to −26 (Δ3) (Fig. 6A and E). Since the presence of Δ3 leads to the removal of the −35 sequence, the plasmid pHYD2892 (bearing Δ3) displayed very low levels of ybjE-lac (Table 2).
In addition, we tested the effects of absence of LysR the transcriptional regulator of lysA, whose product mediates the conversion of diaminopimelate to Lys (38), on the expression of ybjE-lac. Levels of ybjE-lac were unaffected by the absence of LysR. Furthermore, the repressive effect of Arg on ybjE-lac and the absence of effect of Lys remained unaltered in the lysR mutant (see Table S2 in the supplemental material for additional details). ybjE-lac expression was similarly unaffected by the absence of ArgP, the Arg/Lys responsive transcriptional regulator of argO (5), and by the presence of two dominant mutations in argP that lead to high constitutive Arg/Lys-insensitive overexpression of argO (5, 39) (see Table S3 in the supplemental material). It may be noted that the repressive effect of Arg on ybjE-lac was not apparent (see Table S3 in the supplemental material) in strains bearing the dominant argP alleles but not the wild-type argP, which may be explained on the basis that constitutive overexpression of argO caused by dominant argP alleles (5, 39) may lead to enhanced cytoplasmic Arg export, thereby alleviating Arg repression of ybjE-lac.
Evidence for a latent Lys export capacity of ArgO.ArgP-argO of E. coli and LysG-lysE from C. glutamicum are thought to constitute orthologous protein-gene pairs (40). ArgP and LysG are transcriptional regulators of argO and lysE, respectively (5, 16), and the two proteins share 35% identity and 53% similarity at the amino acid level. ArgO and LysE have been shown to mediate export of Arg (5) and Arg/Lys (15, 16), respectively, with the two exporters sharing 35% identity and 50% similarity. One intriguing difference between the two orthologous protein-gene pairs is seen at the mode of transcriptional regulation effected by the two regulators. Whereas lysE transcription is activated by both Arg and Lys serving as coeffectors of LysG (16), argO expression is subject to activation by Arg and repression by Lys (5), with repression by Lys occurring via the formation in the presence of Lys (but not Arg) of an ArgP-Esig70 complex, reversibly trapped at the step of promoter clearance, following open complex formation (41). So far, ArgO has been shown to mediate export only of Arg, and in experiments described above, an argO mutant displayed fitness similar to that of its parent when subjected to cytoplasmic Lys stress imposed by Lys-Ala or thialysine (Fig. 3 and 5A).
We tested whether overexpression of argO could mediate export of Lys and alleviate the impaired growth of a ybjE mutant by the imposition of cytoplasmic Lys stress. In this study, we used an ArgP variant bearing a dominant mutation in argP encoding the ArgPP274S, which, among a collection of similar argP mutations described earlier (5), is known to cause the highest magnitude of Lys-insensitive argO expression (39). Expression of ArgPP274S but not ArgP in MC4100 promoted cross-feeding of the lysA dppB tppB oppB quadruple mutant GJ9060 and the ability of ArgPP274S to cross-feed GJ9060 was absent in a strain lacking ArgO (Fig. 7A). In a related experiment, heterologous overexpression of argO from the IPTG inducible Ptrc promoter of the plasmid pHYD2835 rendered the argO ybjE double mutant GJ9028 resistant to thialysine, though the resistance phenotype was best seen at a lower (0.6 μg/ml) concentration of thialysine (Fig. 7B). Lastly, expression of ArgPP274S but not the wild-type ArgP rendered the ybjE mutant GJ9026 partially resistant to thialysine, and the ArgPP274S-mediated resistance phenotype was absent in strain GJ9028 (see Fig. S3 in the supplemental material).
Lys export function of ArgO. (A) Cross-feeding of the lysA dppB oppB tppB mutant GJ9060 by constitutive overexpression of argO elicited by the ArgPP274S variant of ArgP. Cultures of MC4100 (parent) bearing the plasmid pCL1920 (vector) (I) and its derivatives pHYD915, expressing ArgP (II), and pHYD2606, expressing ArgPP274S (III) and of GJ4823 (MC4100 argO205::Tn10dTet) containing the same plasmids (IV, V, and VI) were spotted on the surface of an MA agar plate containing 1 mM Lys-Ala and 1 μg/ml tetrazolium chloride and imaged after 30 h of incubation. In addition, the plate was seeded with cells of the strain GJ9060 bearing the vector. (B) Functional complementation of the thialysine hypersensitive phenotype of an argO ybjE double mutant GJ9028 by heterologous overexpression of argO. Tenfold serial dilutions of cultures of MC4100 (parent) bearing the plasmid pHYD3025 (vector) and GJ9028 (argO ybjE), bearing the vector or the plasmid pHYD2835, which expresses argO from the Ptrc promoter (Ptrc argO), were spotted on the surface of an MA agar plate containing 10 μM IPTG (I) and an MA agar plate containing 10 μM IPTG and 0.6 μg/ml thialysine (II).
DISCUSSION
LysE mediated Arg/Lys export by C. glutamicum represents one of the earliest known examples of microbial amino acid export (42), and LysE belongs to a family of proteins known as the LysE superfamily (43). In E. coli, though, the Lys but not Arg export potential of ArgO is as yet unknown. In the present study, we obtained multiple lines of evidence which indicate that YbjE functions as a separate Lys exporter in E. coli and that under certain conditions, ArgO can mediate Lys export.
YbjE as a Lys exporter.Two phenotypes of the ybjE mutant, namely, its growth inhibition by elevated cytoplasmic levels of Lys attained by Lys-Ala dipeptide feeding (Fig. 3) and its hypersensitivity to thialysine (Fig. 5), fulfill the genetic criteria expected if its product were to mediate export of Lys. Lys excretion in C. glutamicum is thought to represent an example of limited catabolism as a physiological model that has been proposed to explain the biological need for amino acid export, since Lys export is required during growth in a medium containing lysyl-dipeptides, as C. glutamicum does not have the capacity to catabolize Lys (14, 31). Besides export, another way to redress the growth inhibition caused by elevated cytoplasmic content of an amino acid could be via the activation of its corresponding catabolic route. In E. coli, however, the pathway of degradation of Lys (and also of Arg) via its decarboxylation, followed by the export of the decarboxylated product, is fully operational only under conditions of anaerobiosis and low pH in Lys (or Arg)-containing complex medium (44). In addition the catabolism of Arg to glutamate by the arginine succinyltransferase pathway occurs only under conditions of nitrogen limitation (44). The requirement of YbjE function to mitigate the growth-inhibitory effects of a Lys stress imposed by Lys-Ala under the growth condition used for the experiments whose results are shown in Fig. 3D is thus compatible with the limited-catabolism model.
The engagement of amino acid export is thought to place the bacterial cell at a risk of performing an energy-consuming and futile cycle of export and reuptake of the amino acid (14, 45). In some instances, membrane proteins mediating amino acid efflux have been functionally categorized as exporters with very low affinity for their substrate, a property that is thought to mitigate the potential problem of futile cycling and ensure that export occurs only when the cytoplasmic level of the substrate amino acid is very high (reference 45 and references therein). In the current study, high-level expression of ybjE from the Ptrc promoter in the parent did not lead to any discernible Lys requirement in glucose minimal medium (Fig. 5B), which suggests that the Km for Lys export by YbjE may be high. Our studies that ascribe a Lys export function to YbjE are consistent with the previous studies of Ueda and coworkers, who reported in multiple patents the identification of YbjE as an exporter of Lys (17, 18, 19). They identified YbjE based on the observation that its elevated expression conferred resistance to growth-inhibitory concentrations of Lys in the medium, and a strain lacking YbjE displayed significant reduction in growth rate under the same growth condition. Ueda and coworkers reported that besides Lys, elevated expression of ybjE also conferred resistance to high concentrations of Arg, l-ornithine, l-isoleucine, l-glutamic acid, l-threonine, l-histidine, l-proline, l-phenylalanine, and l-cysteine. Among these, ybjE overexpression promoted significant resistance to extracellular l-ornithine and to a lesser extent to l-threonine, whereas for other amino acids, the resistance was marginal. In addition, Ueda et al. also showed that increased expression of ybjE led to an enhancement in the presence of extracellular Lys content, consistent with the notion that YbjE functions as an exporter of Lys in E. coli.
Premise for the identification of ybjE.Although our selection procedure for identification of ybjE points to an added Arg/CAN export potential of YbjE (Fig. 1B), an alternative explanation could be that the enhanced extracellular Lys concentration resulting from ybjE overexpression competitively inhibits CAN uptake through the common Arg, Lys, ornithine, and CAN uptake system (LAO [for “Lys/Arg/ornithine”]) (35). Consistent with this explanation, a previous study had noted that the MIC of CAN for the argO mutant was considerably increased when Lys was present in the medium (5). The observation that absence of YbjE to some degree exacerbated the CanSS phenotype of an argO mutant (Fig. 2) may also be explained by the latter mechanism. It may be noted that overexpression of ybjE has been reported to mediate resistance to high extracellular concentrations of Arg (17, 18, 19); however, the resistance appears to be marginal at best, and an additional role for YbjE as an Arg/CAN exporter seems unlikely. The previously reported resistance of a strain overexpressing ybjE to elevated extracellular concentration of ornithine (17, 18, 19) may also be explained on the basis that enhanced Lys export via YbjE by competing with ornithine uptake through the LAO uptake system mediates resistance to ornithine. Another observation supporting a Lys (but not Arg) export property of YbjE comes from the experiments with dipeptides, wherein it was observed that the ybjE mutant grew at rates comparable to that of the parent in a medium containing the Arg-Ala dipeptide, whereas in a medium with the Lys-Ala dipeptide, its growth was considerably hindered (Fig. 3). It was surprising that in experiments with dipeptides, the argO mutant was not growth inhibited when challenged with the Arg-Ala dipeptide in the medium (Fig. 3), which is in contrast to the scenario in C. glutamicum, wherein the lysE mutant is rendered sensitive to both Arg- and Lys-containing dipeptides (15, 16). It appears that E. coli may bear an additional mechanism(s) to alleviate the potential growth-inhibitory effects of elevated cytoplasmic Arg levels generated by Arg containing dipeptides.
Separation of Arg and Lys export in E. coli.The existence of YbjE-mediated Lys export in E. coli represents a situation that is distinct from that seen in C. glutamicum, wherein a single protein, LysE, performs export of both Arg and Lys (15, 16). In addition, the genome of C. glutamicum does not possess an ortholog of YbjE. The finding that a Lys export potential could be demonstrated for ArgO only under conditions that bypassed the repressive effect of Lys on the expression of its cognate gene (Fig. 7) indicates that the Lys export capacity of ArgO is rendered cryptic, leading to a division of labor in the export of Arg and Lys in E. coli. We suggest that repression of argO by Lys in E. coli perhaps effects this division. Going solely by strength of phenotypes alone, it appears that ArgO possesses a weaker Lys export capacity than YbjE (Fig. 7). Given the absence of an ortholog of YbjE in C. glutamicum, we used its amino acid sequence to query the nonredundant database using the BLAST algorithm, which revealed a wide distribution of YbjE orthologs across Gram-negative bacteria (data not shown). Restricting the BLAST search to Gram-positive bacteria returned only two possible orthologs with significant coverage and E values, in the genomes of Enterococcus gallinarum EGD-AAK12 (100% coverage) and Virgibacillus halodenitrificans (90% coverage).
Repression of Lys exporter expression in E. coli by Arg.Transcription of genes encoding amino acid exporters in many instances is subject to stimulation by its amino acid substrate. Amino acid-mediated expression control is thought to be another means of avoiding futile cycling, since expression would occur only when the level of the amino acid reaches a certain threshold (14, 45). Stimulation of brnFE expression encoding a multi-amino-acid exporter, by methionine and the branched-chain amino acids in conjunction with an Lrp-like transcription factor (45, 46) in C. glutamicum and by Lrp and leucine of the inducible LeuE leucine exporter in E. coli (9), is another example, besides those of lysE and argO (described above), where the amino acid substrate exerts control over the expression of its exporter.
We tested the effects of mutations in genes encoding transcription factors with known roles in Arg/Lys metabolism on expression of ybjE and found that Arg exerted a modest repressive effect on ybjE via ArgR, with the ArgR binding site being likely to exist in an overlap with the ybjE promoter (Table 2 and Fig. 6). It appears therefore that ArgR at the ybjE promoter is likely to exert its repressive effects in a manner analogous to its role as a repressor of genes involved in Arg biosynthesis, that is, by occluding promoter binding of the RNA polymerase holoenzyme (34, 35). Further studies are needed to delineate the mechanism of ArgR repression at the ybjE promoter.
Our studies on the putative ARG boxes within the cis regulatory region of ybjE suggest that of the two boxes, the one absent in the DNA template of ybjE with the Δ2 deletion (from −64 to −47) represents a stronger binding site than the site that is absent in the Δ3 deletion DNA (from −43 to −26). If this is true, then ybjE would represent yet another example where ArgR imparts gene repression by interacting with imperfect low-affinity ARG boxes. The cis regulatory regions of hisJ (37) and the gltBDF operon (47) contain imperfect ArgR binding sites (see Fig. S2 in the supplemental material) located upstream of the core promoter elements. ArgR mediates repression of gene expression at hisJ and gltBDF in mechanistically distinct ways which in both cases are different from its mode of action on genes involved in Arg biosynthesis (37, 47). It may be noted that for the gltBDF operon, hisJ, and genes involved in Arg biosynthesis, the absence of ArgR leads to derepression of gene expression, which is not the case for ybjE (Table 2). Absence of derepression may be explained if in minimal glucose medium ybjE-lac expression is fully derepressed in the first place. Based on estimates of the number of ArgR molecules in the cell and the affinity of ArgR for Arg, it is believed that a fraction of total ArgR exists in the Arg-bound hexameric state capable of DNA binding (34). Since the affinity of the ArgR hexamer for the classical ARG box is very high, hexameric ArgR binds to the ARG box even during growth in minimal medium, and hence its absence causes derepression of expression of genes involved in Arg biosynthesis. Since the ArgR binding sites in the cis regulatory region of ybjE appear to be weak binding sites (see Fig. S1 in the supplemental material), during growth in minimal medium ybjE-lac expression may remain derepressed as a result of titration of hexameric ArgR by other stronger ARG boxes on the chromosome. Consequently, the repressive effect of ArgR would be seen upon increased ArgR hexamer formation, which would occur when cytoplasmic Arg level is elevated during growth in Arg-containing medium. Going by this, it would appear that the ARG boxes in the cis regulatory regions of hisJ and gltBDF may have higher affinities for ArgR than those in the promoter region of ybjE. Overall, though, the repression of the expression of the Arg exporter ArgO by Lys (5) and of the Lys exporter YbjE by Arg in E. coli is indicative of a mechanism for maintenance of an Arg/Lys balance. Lastly, given the evidence herein that ybjE encodes a functional Lys exporter in E. coli, we propose that ybjE be redesignated lysO (for Lys outward permease), and the energetics of Lys export through LysO remains a subject of future study.
ACKNOWLEDGMENTS
We thank J. Gowrishankar, Ivan Matic, H. Mori, and Barry Wanner for providing strains and plasmids used in the study, and we thank members of the Laboratory of Bacterial Genetics, Manjula Reddy, Suhail Yousuf, and Rohan Misra for advice. The assistance of C. Krishna Prasad in performing HPLC measurements is also acknowledged.
A.P. is a recipient of a doctoral research fellowship from the Council of Scientific and Industrial Research, Government of India, and a fellowship from the Centre for DNA Fingerprinting and Diagnostics and is registered under the academic program of the Manipal University. This work was supported by a Centre of Excellence in Microbial Biology research grant from the Department of Biotechnology, Government of India.
FOOTNOTES
- Received 17 November 2014.
- Accepted 1 April 2015.
- Accepted manuscript posted online 6 April 2015.
- Address correspondence to Abhijit A. Sardesai, abhijit{at}cdfd.org.in.
Citation Pathania A, Sardesai AA. 2015. Distinct paths for basic amino acid export in Escherichia coli: YbjE (LysO) mediates export of l-lysine. J Bacteriol 197:2036–2047. doi:10.1128/JB.02505-14.
Supplemental material for this article may be found at http://dx.doi.org/10.1128/JB.02505-14.
REFERENCES
- Copyright © 2015, American Society for Microbiology. All Rights Reserved.
