JOURNAL OF BACTERIOLOGY, Sept. 2004, p. 6059–6069 0021-9193/04/$08.00⫹0 DOI: 10.1128/JB.186.18.6059–6069.2004 Copyright © 2004, American Society for Microbiology. All Rights Reserved.
Vol. 186, No. 18
Two Arginine Repressors Regulate Arginine Biosynthesis in Lactobacillus plantarum Herve´ Nicoloff,1 Florence Arse`ne-Ploetze,1 Ce´dric Malandain,1 Michiel Kleerebezem,2 and Franc¸oise Bringel1* Laboratoire de Dynamique, Evolution et Expression de Ge´nomes de Microorganismes, Universite´ Louis Pasteur/CNRS FRE 2326, Strasbourg, France,1 and Wageningen Centre for Food Sciences, Wageningen, The Netherlands2 Received 29 January 2004/Accepted 9 June 2004
The repression of the carAB operon encoding carbamoyl phosphate synthase leads to Lactobacillus plantarum FB331 growth inhibition in the presence of arginine. This phenotype was used in a positive screening to select spontaneous mutants deregulated in the arginine biosynthesis pathway. Fourteen mutants were genetically characterized for constitutive arginine production. Mutations were located either in one of the arginine repressor genes (argR1 or argR2) present in L. plantarum or in a putative ARG operator in the intergenic region of the bipolar carAB-argCJBDF operons involved in arginine biosynthesis. Although the presence of two ArgR regulators is commonly found in gram-positive bacteria, only single arginine repressors have so far been well studied in Escherichia coli or Bacillus subtilis. In L. plantarum, arginine repression was abolished when ArgR1 or ArgR2 was mutated in the DNA binding domain, or in the oligomerization domain or when an A123D mutation occurred in ArgR1. A123, equivalent to the conserved residue A124 in E. coli ArgR involved in arginine binding, was different in the wild-type ArgR2. Thus, corepressor binding sites may be different in ArgR1 and ArgR2, which have only 35% identical residues. Other mutants harbored wild-type argR genes, and 20 mutants have lost their ability to grow in normal air without carbon dioxide enrichment; this revealed a link between arginine biosynthesis and a still-unknown CO2-dependent metabolic pathway. In many gram-positive bacteria, the expression and interaction of different ArgR-like proteins may imply a complex regulatory network in response to environmental stimuli. tive bacteria, such as B. subtilis (8), Geobacillus stearothermophilus (previously called Bacillus stearothermophilus) (11), and Streptomyces clavuligerus (31). The resolved ArgR structures from E. coli, G. stearothermophilus, and B. subtilis have similar folding patterns, despite only 27% amino acid identity between the enterobacterial and the bacillus arginine repressors (10, 26, 35, 37). The arginine repressor subunit consists of a basic N-terminal DNA binding domain (DBD), which belongs to the winged helix-turn-helix family and is connected through a flexible linker to an acidic C-terminal domain responsible for oligomerization and arginine binding. The active regulator consists of an ArgR hexamer formed by the dimerization of two trimers stabilized by the fixation of six L-arginine molecules at the trimer-trimer interface, which results in the allosteric activation of the regulator (15). The active hexamer recognizes one or more “ARG box” DNA sequences made of 18-bp imperfect palindromes (22–24). The activated transcriptional regulator can act as a transcriptional inhibitor and/or activator, depending on the prokaryote studied. In B. subtilis in the presence of arginine, the arginine repressor AhrC inhibits arginine biosynthesis and activates the arginine catabolic arginase pathway (23). It is also an obligate accessory protein for the cer/Xer site-specific recombination mechanism resolving multimers of ColE1-like plasmids in E. coli (33). Unlike other lactic acid bacteria, most L. plantarum organisms have no known arginine catabolic pathways (1), and no genes of the arginine deiminase (ADI) pathway were found in the genome of the WCFS1 strain (17) or strain CCM 1904 used in this study (27). The genes encoding arginine biosynthetic enzymes were characterized (Fig. 1 shows the gene organization of the
Lactobacillus plantarum, a gram-positive bacterium found in nutritionally rich biotopes like plants and the intestines of mammals, is used in a wide variety of fermentation processes (e.g., meat, vegetables, and dairy products). Among the lactic acid bacteria, L. plantarum is found in the most diverse biotopes. Its metabolic flexibility allows it to adapt to a variety of environments. Genome analysis of L. plantarum revealed a high proportion of regulatory proteins (8.5%), which is a typical feature of bacteria that can be found in diverse environments, such as Listeria monocytogenes and pseudomonads (17, 34). Among the regulatory proteins, two putative arginine repressor genes were identified in the genome of strain WCFS1. This was surprising, in view of the paradigm that arginine repression is mediated by a single arginine repressor, as found in Escherichia coli (22) and Bacillus subtilis (8). Moreover, an analysis of sequenced genomes revealed the frequent presence of several potential arginine regulator genes in the genomes of gram-positive bacteria (5). These observations led us to investigate the putative roles of the two ArgR repressors found in L. plantarum. Arginine-mediated regulation is remarkably well conserved in very divergent bacteria, i.e., gram-negative bacteria, such as E. coli (22), Salmonella enterica serovar Typhimurium (21), Thermotoga (24), and Moritella profunda (38), and gram-posi* Corresponding author. Mailing address: Laboratoire de Dynamique, Evolution et Expression de Ge´nomes de Microorganismes, Universite´ Louis Pasteur/CNRS FRE 2326, 28 rue Goethe, 67083 Strasbourg, France. Phone: 33 3 90 24 18 15. Fax: 33 3 90 24 20 28. E-mail:
[email protected]. 6059
6060
NICOLOFF ET AL.
arginine biosynthetic pathway). The peculiar presence of two arginine repressors in an organism that synthesizes but does not catabolize arginine prompted us to investigate the role of each argR in arginine biosynthesis in L. plantarum. The predicted ArgR1 (153 residues) and ArgR2 (152 residues) have 35% identical amino acids. Thirty-one and 39% of ArgR1 and ArgR2 residues, respectively, are identical to those of the B. subtilis arginine repressor AhrC, and 26 and 21% of the residues are identical to those of the E. coli ArgR protein. We selected spontaneous mutants with altered arginine regulation in order to investigate arginine-dependent regulation in L. plantarum and, in particular, the functions of argR1 and argR2 in this regulation. MATERIALS AND METHODS Bacterial strains and culture conditions. A list of the strains studied is given in Table 1. CCM 1904 is auxotrophic for 11 amino acids and six vitamins but prototrophic for both arginine and pyrimidines. L. plantarum strains were grown at 30°C without agitation on MRS medium (Difco Laboratories) or the defined rich medium DLA (lactobacillus rich defined medium) free of arginine and pyrimidines (3). Nutritional requirements were tested at 30°C on agar plates either unsupplemented or supplemented with 50 g of uracil (DLAU) or of arginine/ml in aerobiosis or CO2-enriched air. Aerobiosis was obtained by incubation in ordinary air. To calibrate the gas phase CO2 concentration at 4%, a water-jacketed CH/P incubator (Forma Scientific) was used. The ability of L. plantarum to grow in the presence of the toxic arginine analogue canavanine in final concentrations ranging from 5 to 200 g/ml was tested in CO2-enriched air in DLAU medium. Liquid cultures were performed in a CO2-enriched atmosphere without agitation in 250-ml Erlenmeyer flasks containing 50 ml of DLAU supplemented with arginine or unsupplemented. For RNA extracts used in reverse transcription (RT)-PCR, cells were cultivated in the water-jacketed CH/P incubator with the CO2 concentration calibrated at 4%. For RNA extracts used in Northern blots, cells were cultivated in Erlenmeyer flasks closed with gas-tight corks to prevent CO2 loss during incubation, and 10.8 ml of pure CO2 gas was injecting through the cork with a sterile syringe to obtain 4% in the gas phase. Positive screening for arginine-deregulated mutants. Spontaneous mutants deregulated in arginine biosynthesis and derived from independent mutational events were isolated from strain FB331. Isolated colonies grown on MRS agar plates from a FB331 frozen stock culture were suspended in 1 ml of sterile physiological water (NaCl, 9 g/liter). Fifty microliters of each suspension was spread on DLA plates or DLA plates supplemented with arginine (50 g/ml) and incubated for 3 days at 30°C in CO2-enriched air. Under these conditions, the wild-type strain FB331 was unable to grow, but a few spontaneous mutants derived from strain FB331 were obtained. A single mutant clone from each plate was transferred to MRS plates, cultivated in MRS medium, and then stored at ⫺80°C in the presence of 20% glycerol. Determining arginine excretion. Arginine excretion was estimated by measuring the arginine present in stationary-phase cell culture supernatants by using a bioassay. The indicator bacterium was L. plantarum HN217 (27). This strain has no arginine-regulated carbamoyl phosphate synthetase (CPS-A) and pyrimidineregulated CPS (CPS-P), so that its growth requires arginine and uracil (Fig. 1A).
J. BACTERIOL. HN217 growth was assessed at 30°C without shaking in liquid DLAU medium supplemented with 4-day-old filter-sterilized (Millipore 0.2-m-pore-size filters) culture supernatants of arginine-excreting strains previously grown in 2 ml of liquid DLAU medium. For each tested strain, 60 and 300 l of this sterilized medium were added to 1.94 and 1.7 ml of new DLAU medium, respectively, and strain HN217 was grown for 4 days. The turbidity (optical density [OD] at 600 nm) of each culture was directly proportional to the arginine concentration in the culture medium. This turbidity was compared to the OD obtained with the sterilized supernatant of the parental arginine-regulated strain FB331 with added arginine concentrations ranging from 0 to 5 g/ml. Two reference curves were obtained by mixing 60 (or 300) l of sterilized used FB331 medium with 1.93 (or 1.69) ml of new DLAU medium and 10 l of new DLAU medium at 0 to 0.1% arginine (data not shown). The minimum detectable arginine concentrations using this method were ⬃0.2 g/ml. DNA techniques. Four genetic loci were PCR amplified and sequenced using the primers described in Table 2. PCR amplifications were done on a Peltier Thermal Cycler PTC-200 (MJ Research) with Taq polymerase (Sigma). The PCR amplifications consisted of denaturation at 95°C for 1 min, followed by 35 cycles of denaturation for 20 s at 94°C, hybridization for 20 s at 56°C, and elongation for 5 min at 72°C. Finally, the PCR was completed by a 10-min postelongation treatment at 72°C. Prior to sequencing of the PCR products, the unincorporated primers and deoxynucleoside triphosphates were eliminated by passage on a Microspin S-400 HR column (Amersham Biosciences). The argR2 gene of L. plantarum CCM 1904 was sequenced after amplification using inverse PCR. In our study of arginine synthesis, we fortuitously found that primer argCP2 not only hybridized to argC but also to the argR2 gene. This primer, in combination with primers argR5 (5⬘-GGTAACGGACCAGCACTAGC-3⬘) and argRd4 (5⬘-CKWGAWAYNGTNGCYTGTGT-3⬘, where K is G or T, W is A or T, Y is C or T, and N is A, C, G, or T) were used to amplify the entire argR2 gene and its flanking regions (data not shown) in a 2,402-bp sequence (GenBank accession no. AF451891). RNA extraction. L. plantarum organisms grown until mid-exponential growth phase (OD at 600 nm, 0.4) were harvested by centrifugation at 4°C. The pellet was suspended in 400 l of suspension solution (10% glucose, 12.5 mM Tris, pH 8, 5 mM EDTA) and 60 l of 0.5 M EDTA. The cell suspension was then transferred to a 1.5-ml Eppendorf tube containing 500 l of acidic phenol and 0.4 g of glass beads (average diameter, between 0.25 and 0.5 mm). The cells were broken by four series of shaking (Retsch apparatus at maximal speed for 30 s) with 1-min pauses at 4°C. The cell debris was pelleted by centrifugation (5 min at 4°C; 10,000 ⫻ g). The supernatant was mixed with 1 ml of Trizol reagent (Invitrogen) and incubated for 5 min at room temperature, and then 100 l of chloroform–isoamyl alcohol (24/1 [vol/vol]) was added for 3 min at room temperature. After centrifugation (5 min at 4°C; 10,000 ⫻ g), the upper phase was treated again with 200 l of chloroform–isoamyl alcohol. To precipitate the nucleic acids in the upper phase, 500 l of isopropanol was added. After centrifugation (15 min at 4°C; 10,000 ⫻ g), the pellet was washed once with 70% ethanol, dried, solubilized in 50 l of Tris-EDTA, and treated for 20 min at 37°C with DNase I (10 U; Amersham Biosciences). The DNase I was heat inactivated (65°C for 10 min). Transcription analysis. RT-PCR was performed with the Invitrogen SuperScript one-step RT-PCR with platinum Taq kit as recommended by the manufacturer. After a 5-min denaturation at 65°C, 25 ng of RNA template was used to detect transcripts specific to the carA, argG, and rrn genes. The argC transcripts were detected using 60 ng of RNA template. RT was performed at 50°C
FIG. 1. Regulation of arginine biosynthesis in L. plantarum. (A) Simplified arginine biosynthesis in L. plantarum FB331. CP is a common precursor between arginine synthesis and the de novo pyrimidine pathway. Wild-type L. plantarum harbors CPS. Strain FB331 has CPS-P, the pyrimidine-regulated CPS, deleted so CP synthesis is entirely dependent on the arginine-regulated CPS (CPS-A) encoded by the arginine-repressed operon carA. The genes coding for enzymes that synthesize citrulline from glutamate, ATP, and CO2 are organized on the divergently transcribed carAB and argCJBDF operons (3, 27). The last two steps that synthesize arginine from citrulline are encoded by the argGH genes. The overlapping argG and argH genes suggest coupled translation and transcription. (B) Model for arginine-dependent repression of arginine biosynthetic operons. The arginine-dependent repression of the three tested operons requires both the argR1 and argR2 gene products, which suggests that the active repressor may be a hexamer composed of ArgR1-ArgR2 molecules, but this hypothesis needs to be experimentally tested. The arrows beneath the genes represent the directions of transcription. (C) Mutation mapping of the Arg operators of the bipolar carA/argC genes. Transcription start sites (⫹1) are indicated with arrows and were determined experimentally for carA (27) and argC (data not shown). The proposed nucleotides of the ⫺10 and ⫺35 promoter boxes are underlined. The nucleotide sequence of the carA-argC intergenic region in mutant FB331-14 (⌬4732 to 4780) is detailed at the top. The additional 48 nucleotides found in the wild-type L. plantarum are listed in the rectangle (EMBL database accession no. X99978). The star indicates the T3C mutation found in strain FB331-13. The dotted boxes represent ARG boxes, with bases that are highly conserved among studied E. coli operators and in operators from other organisms in boldface (24); mismatched bases found in the proposed L. plantarum ARG box are in lowercase.
VOL. 186, 2004
TWO ARGININE REPRESSORS IN LACTOBACILLUS PLANTARUM
6061
6062
NICOLOFF ET AL.
J. BACTERIOL. TABLE 1. L. plantarum strains
Namea
Status of protein
Genotypeb g
ArgR1
ArgR2
CCM 1904 HN217
WT ⌬pyrAaAb ⌬carAB
WT WT
WT WT
FB331
⌬pyrAaAb
WT
WT
Spontaneous derivatives of strain FB331 FB331-1
⌬pyrAaAb argR1 (1291855G3T)
Absentd
WT
FB331-2 FB331-3 FB331-4 FB331-5 FB331-6 FB331-7 FB331-8 FB331-9 FB331-10 FB331-11 FB331-12 FB331-13
⌬pyrAaAb ⌬pyrAaAb ⌬pyrAaAb ⌬pyrAaAb ⌬pyrAaAb ⌬pyrAaAb ⌬pyrAaAb ⌬pyrAaAb ⌬pyrAaAb ⌬pyrAaAb ⌬pyrAaAb ⌬pyrAaAb
A39E S42L D44G A123D ⌬(Q140-D153) Deletede WT WT WT WT WT WT
WT WT WT WT WT WT R6H R43C R43H Chimera f ⌬(L135-H152) WT
FB331-14
⌬pyrAaAb ARG box (⌬4732-4780)
WT
WT
argR1 (1292012C3A) argR1 (1292020C3T) argR1 (1292027A3G) argR1 (1292264C3A) argR1 (1292314C3T) ⌬argR1 argR2 (598G3A) argR2 (708C3T) argR2 (709G3A) argR2 (768insA) argR2 (985T3A) ARG box (4763T3C)
Comments, phenotypec, and reference
Prototroph Deletion of CPS-P and CPS-A; auxotroph for both arginine and pyrimidines (27) Deletion of CPS-P (27)
Mutation in argR1 promoter region; class 3 mutant (This study) Class 3 mutant (This study) Class 3 mutant (This study) Class 3 mutant (This study) Class 3 mutant (This study) Class 3 mutant (This study) Class 3 mutant (This study) Class 3 mutant (This study) Class 3 mutant (This study) Class 3 mutant (This study) Class 3 mutant (This study) Class 3 mutant (This study) Point mutation in intergenic carA-argC region; class 2 mutant (This study) 48-nucleotide deletion in intergenic carAargC region; class 2 mutant (This study)
a
All strains are isogenic to strain CCM 1904, which is equivalent to strain ATCC 8014. Coordinates for mutations in argR1, argR2, and the intergenic carA-argC promoter region refer to sequences with EMBL accession no. NC_004567, AF45189, and X99978, respectively. c Phenotype as defined in Table 4. d Point mutation in the promoter region, strongly suggesting that argR1 transcription regulation is impaired. e The argR1 locus was completely deleted, as demonstrated by lack of PCR amplification and absence of detected bands in Southern hybridization with an argR1 probe (data not shown). f Mutation of residues at positions 63 to 91 and deletion of the last 62 residues. g WT, wild type. b
for 30 min, and the resulting cDNA was PCR amplified. Denaturation at 94°C for 2 min was followed by 30 cycles of denaturation at 94°C for 15 s, hybridization at 54°C for 30 s, and elongation at 68°C for 50 s. The program ended with a postelongation step at 72°C for 10 min. The primers used in RT-PCR are described in Table 2. The RT-PCR products were stained with ethidium bromide and detected by UV after electrophoresis in a gel (2% Nusieve agarose gel; FMC BioProducts) separating small DNA fragments. To control the absence of contaminant DNA in RNA preparations, the RT step was omitted, and no PCR products were amplified. The transcripts of the arginine-regulated operons (carAB, argCJBDF, and argGH) were quantified using Northern hybridization DNA probes specific to carA, argC, and argG, respectively. The probes were obtained after PCR amplification (95°C for 1 min, followed by 35 three-step cycles of 94°C for 40 s, 50°C for 40 s, and 72°C for 2 min; a postelongation at 72°C for 10 min completed the program). The primers are listed in Table 2. The PCR products were digoxigenin (DIG) labeled, and the concentrations of DIG-labeled probes were estimated by comparison with the labeled DNA control from the DIG DNA labeling and detection kit (Roche). For each probe, the optimal amount of RNA to be used for quantification was tested in the range of 0.5 to 10 g of total RNA and was found to be different for the carA (2.5 g), argC and argG (10 g), and rrn (100 to 250 ng) probes. RNAs (25 l) were heat denatured (10 min at 65°C) in the presence of 75 l of denaturation mixture (prepared by mixing 500 l of formamide with 162 l of formaldehyde and 100 l of 10⫻ MOPS [morpholinepropanesulfonic acid]). After 5 min on ice, 100 l of cold 20⫻ SSC (1⫻ SSC is 0.15 M NaCl plus 0.015 M sodium citrate) was added, and the RNAs were transferred onto positively charged Hybond nylon membranes (Amersham Biosciences) using the Bio-Dot SF microfiltration apparatus (Bio-Rad). The air-dried RNAs were fixed on the membranes by UV irradiation. A temperature of 50°C was used for prehybridization and overnight hybridization with 10 ng of DIG-labeled probes/ml. Nucleic acid hybrids were detected with the nonradioactive DIG DNA labeling and detection kit using the alkaline phosphatase chemiluminescent substrate CDP-star (Roche). Image acquisition was performed with the ChemiDoc XRS camera (Bio-Rad), and the signals were quantified with Quantity One software (Bio-Rad). The background level was subtracted from the
measured signal. To calculate the relative signal, the measured signal was divided by the quantity of RNA put on the membrane. This relative signal was then divided by the relative signal obtained with the rrn probe to obtain the relative amount of each arg or car gene. Computer analysis. Sequences were compared using the BLAST program (http://www.ncbi.nlm.nih.gov/BLAST/), and the ClustalX program was used to analyze protein alignments. Three-dimensional visualization of proteins was done with Swiss-Pdb viewer software (14). Nucleotide sequence accession number. The DNA sequence of argR2 from L. plantarum CCM 1904 has been deposited in the GenBank database under accession no. AF451891.
RESULTS In L. plantarum, the two argR-like genes are adjacent to genes also found in other gram-positive bacteria. The complete genome of L. plantarum WCFS1 harbors two argR genes named argR1 and argR2 (17). Since the sequenced strain is auxotrophic for arginine when grown in air (4), the prototroph strain CCM 1904 was chosen to study arginine regulation. CCM 1904 has an argR2 gene identical to the sequenced strain but has three silent mutations in the argR1 locus (1292191C3T, 1292389A3G, and 1292419C3T, according to sequence accession no. NC_004567). The genes adjacent to argR1 and argR2 were also found in other gram-positive bacteria, which highlights synteny group conservation around argR genes (Table 3). Selection and phenotype of spontaneous arginine-deregulated mutants. Arginine biosynthesis-deregulated mutants were obtained from L. plantarum FB331, a strain with CPS-P de-
TWO ARGININE REPRESSORS IN LACTOBACILLUS PLANTARUM
VOL. 186, 2004
6063
TABLE 2. Primers and uses Type of analysis
Regular PCR and sequencing
Genetic locus (EMBL database accession no.)
Primer set
Amplified DNA size (kb)
argR1 (AL935263)
arcT1-argR1r1
0.75
argR2 (AF451891)
argR2-recNr4
0.75
pyrR (Z54240)
N40-N29
1.2
argCP2-carBr4
2.2
carB5-carbr19
2.3
72r1-r207r3
0.79
a
carAB (X99978)
argC-carA (X99978) Additional sequencing primers
carAB
Transcription analysis (RT-PCR and Northern probes)
carA (X99978)
carBr8-r72r3
0.31
argC (X99978)
207r4-r207r8
0.46
argC1-argC2
0.14
argG (AL935263)
p775F-1p7775R
0.37
rrn (AL935263)
907r-530f
0.4
a
Primer name
Sequence (5⬘33⬘)
arcT1 argR1r1 argR2 recNr4 N40 N29 argCP2 carBr4 carB5 carBr19 72r1 r207r3 carB3 carBr6 carBr7 carB7 carB8 carBr10 carB13 carBr8 r72r3 207r4 r207r8 argC1 argC2 1p775F 1p775R 907r 530f
ATTAGGGCAATATTTACGCCGC ATTTATCGAAGGTCGCATCGC GCTTATGAACAATTAAATCAAGC GTCTCACCAGTTAAAACCGTC GTACTAATATTAGGCGACCG TGTGTGCGGGTGCTATTCTC CTAGCGCGACTTGCATATAATC GGTCTTGAGCGAGTTGAACG TACGAACGGCGGCTTTGACG GTAGTCGGTTATCCCGCTTGG GCTCCCGTTTGGCGAATGT TACATGGCTGCCACCGTC CTGCGGGCCACCCACCCTG AATGCTGCCCGGTCTTCAGC CAATGCTGCCGTCATTTAATTC ATCCCGAGACGGTCTCCACC CACCCGACAGATCAGCGGC CAAACTGCGATCAACCTTGC GCCACCAAATTAGCACAACACC AGCTGGTAAGTCTGAGCACC CAAGTACTAACGGCTACC ACGGCTTAGCTGATATGC CCTTGAGCGTCGGCATCC ATACTTTTTGCATGAACATGCAGC TTGGATGCGATCGATCTCGGGATG ATACCCGCTGGTCTCAGCAC ATCAGCTAAGGCCTTCGTCC CCGTCAATTCCTTTGAGTTT GTGCCAGCAGCCGCGG
Overlapping PCRs were done to amplify the carAB sequence.
leted in which the only active CPS is CPS-A (27) (Fig. 1A). Thus, in the presence of arginine, no carbamoyl phosphate (CP) was available for pyrimidine biosynthesis, since the carAcarB gene transcription was arginine repressed (FB331 is arginine inhibited as long as no pyrimidines are present in the medium) (27). This property was exploited in a positive screening to isolate deregulated mutants able to grow in the presence
of arginine in pyrimidine-free medium. Moreover, FB331 is unable to grow in the absence of arginine and pyrimidines in the medium. Apparently, the wild-type CPS-A expression level is not sufficient to provide CP at a level required for both arginine and pyrimidine biosynthesis. This observation was exploited in a second screening procedure to isolate deregulated mutants able to grow in DLA medium without pyrimidines and
TABLE 3. L. plantarum argR1 and argR2 belong to distinct synteny groups Bacterium
Group 1 (argR1 pbp2A) L. mesenteroides E. faecalis Group 2 (yqxC argR2 recN) E. faecalis Listeria innocua L. monocytogenes Oceanobacillus iheyensis G. stearothermophilus B. subtilis L. mesenteroides Streptococcus mutans Streptococcus pneumoniae Streptococcus agalactiae L. lactis
% Identical residues with L. plantarum proteina ArgR1
Pbp2Ab
42 (ZP_00063602) 42 (NP_814427)
55 (ZP_00063603) 47 (NP_814430)
Lp_1603c
63 65 64 52 60 62 64 62 60 62 57
(NP_814718) (NP_470739) (CAC99444) (NP_692797) (unfinished genome) (NP_390306) (ZP_00063476) (NP_721016) (NP_358679) (NP_687529) (NP_267014)
RecNd
ArgR2
45 44 43 43 40 38 36 34 35 32 32
(NP_814719) (NP_470740) (NP_464892) (BAC13831) (O31408) (BAA12578) (ZP_00063477) (NP_721017) (NP_358678) (NP_687530) (NP_267015)
48 45 46 40 43 41 40 40 39 41 40
(NP_814720) (NP_470741) (NP_464893) (NP_692795) (unfinished genome) (NP_390304) (ZP_00063478) (NP_721018) (NP_358677) (NP_687531) (NP_267016)
a Within a group, the microorganisms are ranked by their ArgR relatedness to L. plantarum ArgR1 (group 1) or to L. plantarum ArgR2 (group 2). In parentheses are the GenBank database accession numbers when available. b Putative penicillin-binding protein. c L. plantarum Lp_1603 belongs to the well-conserved FtsJ-like methyltransferase encoded by yqxC. d A DNA repair protein.
6064
NICOLOFF ET AL.
J. BACTERIOL.
TABLE 4. Phenotypes of the prototroph strain (CCM 1904), the selected 44 arginine-deregulated mutants, and their parental strain, FB331 No. of clones selected on:
L. plantarum strain
DLA
Growth ona CO2 in air (%)
DLA ⫹ A
Prototroph FB 331 (⌬CPS-P) Class 1 mutant
9
11
Class 2 mutant
10
2
Class 3 mutant
8
4
⬍0.01 4 ⬍0.01 4 ⬍0.01 4 ⬍0.01 4 ⬍0.01 4
MRS
⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫹
Can MICb (g/ml)
DLA with supplementa None
A⫹U
A
U
⫹ ⫹ ⫺ ⫺ ⫺ ⫹ ⫺ ⫹ ⫺ ⫹
⫹ ⫹ ⫹ ⫹ ⫺ ⫹ ⫹ ⫹ ⫹ ⫹
⫹ ⫹ ⫺ ⫺ ⫺ ⫹ ⫺ ⫹ ⫺ ⫹
⫺ ⫹ ⫺ ⫹ ⫺ ⫹ ⫺ ⫹ ⫺ ⫹
Arginine excretion (g/ml)
10
0
10
⬍0.2
⬎200
2–5
⬎200
0.4–2
⬎200
70–90
a ⫹ and ⫺, good and poor growth, respectively. Phenotypic tests were performed on agar plates of DLA medium complemented or not with arginine (A) and/or uracil (U) at 50 g/ml. b Canavanine (Can) MICs were determined on DLA plus U agar plates.
arginine. All these mutants were isolated under CO2-enriched growth conditions (aerobiosis enriched with 4% CO2), since CPS-A activity depends on CO2 concentrations higher than those found in air (27). Forty-four spontaneous independent mutants were isolated on DLA and DLA-plus-arginine media. The growth of these mutants was then tested under different conditions (aerobiosis and aerobiosis enriched with CO2) and on different media; the results are listed in Table 4. Three different phenotypes were characterized (Table 4, classes 1, 2, and 3). Unlike the parental strain, in pyrimidine-free media all mutants grew in CO2-enriched air. The presence of arginine in the selection process had no detected effect on the mutant phenotypes. All 44 mutants had higher resistance to canavanine (canavanine minimal inhibitor concentrations were ⬎200 g 䡠 ml⫺1 compared to 25 g 䡠 ml⫺1 observed with the parental FB331 strain [Table 4]). The amounts of arginine excreted by the mutants varied from 0.4 to 90 g 䡠 ml⫺1 and were higher than that for the regulated parent (Table 4). Class 1 mutants were the most frequent: 20 out of 44 strains. These mutants displayed the characteristic growth dependency on the CO2 supply; on DLA medium, no growth in aerobiosis was observed even when both arginine and uracil were supplied. Analogous to the parental strain, other mutants grew on DLA complemented with arginine and uracil in aerobiosis. Class 3 mutants (12 mutants) presented the same phenotype as class 2 mutants, except for their extremely high excretion of arginine (averaging ⬃75 g 䡠 ml⫺1) (Table 4). The higher level of arginine excretion suggests that class 3 mutant regulation of arginine biosynthesis was the most impaired. We hypothesized that class 3 mutants were good candidates to search for arginine regulator mutants. Characterization of arginine-deregulated mutants: argR1 and argR2 mutants had the class 3 phenotype. To characterize the mutation loci, the sequences of the parental strain and of its derivative mutant in four genetic loci were compared: (i) the carA-carB genes encoding CPS-A, (ii) the carA-argC interoperonic sequence, (iii) the pyrR gene encoding the pyrimidineregulated transcriptional regulator, and (iv) the argR1 and argR2 genes (see Materials and Methods). The 12 mutants of class 3 (Table 4) contained mutations in either the argR1 or the argR2 gene analyzed (Table 1 and Fig. 2), and no mutations
were found in the other loci analyzed. In mutant FB331-1, a G3T substitution was identified 42 nucleotides upstream of the argR1 initiation codon. The localization of this mutation within the argR1 promoter region suggests that the argR1 transcription is impaired in this mutant. All other class 3 mutations were found in the ArgR coding sequences. Mutations in the ArgR proteins were found in the two functional domains of arginine repressors, the DBD in the N-terminal domain (37) and the C-terminal oligomerization and arginine-binding domain (26). Mutations in the DBD were mapped in ArgR1 (A39E, S42L, and D44G; mutants FB331-2 to -4) and in ArgR2 (R6H, R43C, and R43H; mutants FB331-7 to -9) (Fig. 2). The high homology between the G. stearothermophilus ArgR and L. plantarum ArgR2 proteins (38% identity) suggests that these domains are structurally similar. Such structural similarities are also found for ArgR N- and C-terminal domains of E. coli (35, 37) and G. stearothermophilus, while these proteins show only 28% sequence identity. The corresponding residues mutated in the L. plantarum ArgR1 and ArgR2 proteins were analyzed in view of the G. stearothermophilus ArgR N-terminal region (accession no. 1B4A) (26). By comparison with the three-dimensional representation of the G. stearothermophilus crystallized ArgR, the L. plantarum ArgR mutated residues R43, A39, S42, and D44 are found in the helix ␣3, which has been described as making direct contacts with DNA (data not shown). In mutant FB331-7, the R6H amino acid substitution conferred the fully derepressed phenotype (Tables 1 and 4), confirming that the argR2 translation initiation codon is GTG and not the downstream ATG. Mutation of residue R6 has never been described before (16), although adjacent residues altered E. coli ArgR activity (37). The R6H mutation is located in the ArgR2 ␣1 helix, which is spatially close to the ␣3 helix (data not shown). In this mutant, the loss of a potential hydrogen bond between residues R6 and D44 may be altered. In the C-terminal domain, three mutations were found, including an amino acid substitution (A123D) in ArgR1 and two deletions (⌬Q140-D153 in ArgR1 and ⌬L135-H152 in ArgR2). These deletions in the oligomerization domain of the protein may generate incorrect folding or oligomerization, which may make them nonfunctional repressors. In conclusion, class 3 mutants harbored mutations in one of the argR genes, and these
VOL. 186, 2004
TWO ARGININE REPRESSORS IN LACTOBACILLUS PLANTARUM
6065
FIG. 2. Mapping of the L. plantarum ArgR1 and ArgR2 mutations and alignments with other arginine repressors. Among the aligned repressors, only the functions of the four E. faecalis ArgR-like proteins have not been characterized. Residues conserved (solid background) and similar (shaded background) among most of the proteins are highlighted. ␣ helices (solid rectangles) and  strands (shaded arrows) were positioned according to the G. stearothermophilus ArgR crystal structure (26). In the last  strands, the arginine-binding domains are boxed. Residues important for arginine binding and hexamerization (6, 28, 36) are indicated with open and solid squares, respectively. Important argininebinding residues predicted by crystallography and computational studies (19, 37) are indicated with asterisks. Mutations found in this study in L. plantarum are indicated with diamonds and normal font for ArgR1 and with triangles and italics for ArgR2. A base insertion (768insA according to sequence accession no. AF451891 in the EMBL database) generated a frameshift mutation with the substitution of residues 63 to 91 and deletion of the end of the ArgR2 protein. Mutations 985T3A and 1292314C3T (the coordinates are in reference to the sequences with accession numbers AF451891 and NC_004567, respectively) resulted in deletion of the last 18 and 14 amino acids of ArgR1 and ArgR2, respectively.
mutations appeared to be dominant and to alter arginine regulation. Moreover, arginine regulation depended on functional binding of the repressors to DNA and the corepressor and also on the integrity of the repressor oligomerization domain. Mutations in argR1 or argR2 resulted in loss of arginine repression. The arginine biosynthetic operons (argGH, carAB, and argCJBDF [Fig. 1B]) may be regulated differently by ArgR1 or ArgR2. Thus, the transcription of argG, carA, and argC, which are the first genes of these operons, was analyzed by RT-PCR (Fig. 3) and Northern hybridization (Fig. 4) in cells grown in the presence or absence of arginine in three genetic contexts. Arginine repression was compared in strain FB331, harboring two wild-type argR genes, and the mutants FB331-4 and FB331-10, which harbor single-amino-acid substitutions in ArgR1 (D44G) and ArgR2 (R43H) DBDs, respectively (Fig. 2). In the presence of both wild-type argR alleles, arginine biosynthetic gene transcription was repressed by arginine (Fig. 3 [compare lanes 1 and 2] and 4). Complete arginine repression was observed for the carA and argG genes but not for the argC gene (Fig. 3, argC, compare lanes 1 and 2). This was quantified using Northern hybridization: cultivating the wildtype strain in the presence of arginine drastically inhibited transcription of the carA and argG genes (⬍1% of the nonrepressed level), and only half of the argC transcripts were detected (Fig. 4). On the other hand, in the presence of arginine, carA, argC, and argG transcripts were more abundant in the ArgR1 mutant or the ArgR2 mutant than in the wild-type strain (RT-PCR data [Fig. 3, compare lanes 4 and 6 to lane 2] and Northern blots [Fig. 4]). This demonstrates that a muta-
tion in either ArgR1 or ArgR2 led to the loss of arginine repression of the arginine biosynthetic genes. Since a mutation in the DNA binding site of only one L. plantarum argR gene is sufficient to alter arginine regulation, we propose that both ArgR proteins bind to target DNA sequences, which we refer to as the ARG box by analogy to what has been described in other eubacteria (24). Class 2 mutants with impaired ARG operators. Class 2 mutants harbored wild-type argR1 and argR2 gene sequences (data not shown). Compared to the class 3 mutants, class 2 mutants excreted less arginine and may be the result of mutations present in the operators of the arginine regulons. Putative ARG boxes located in the bipolar carA-argC promoter region have been determined by homology with the E. coli ARG box consensus sequence (3). Among the 10 tested class 2 mutants, two clones presented mutations in the carA-argC intergenic sequence (Table 1), but not within the CPS-A structural genes. Mutant FB331-13 harbored a T3C point mutation located in the previously proposed R9 ARG box, 5⬘-ATTGA AAAAAtATTCACT-3⬘ (accession no. X99978; nucleotides 4773 to 4756) (Fig. 1C). In mutant FB331-14, a deletion of 48 nucleotides including the R9 ARG box was found. Despite the fact that the deletion site is located within the argC promoter, the argC promoter sequence in FB331-14 was unchanged (the T base found in the wild-type ⫺35 box was replaced by another T located upstream of the deletion) (Fig. 1C). The effects of these mutations on arginine-dependent transcription of the bipolar car and arg operons were tested using RT-PCR (Fig. 3) and Northern blotting (Fig. 4). Both mutants had constitutive
6066
NICOLOFF ET AL.
J. BACTERIOL.
FIG. 3. RT-PCR analysis of arginine-dependent regulation of arginine biosynthetic genes. Different arginine-deregulated mutants were compared to the wild-type (wt) strain FB331 (lanes 1 and 2). (A) Mutations in either argR1 (strain FB331-4 [lanes 3 and 4]) or argR2 (strain FB331-10 [lanes 5 and 6]). (B) Mutations in the carA-argC ARG operator with mutant FB331-13 (lanes 7 and 8) and mutant FB331-14 (lanes 9 and 10). RNAs were extracted from cells cultivated in the presence (⫹) or absence (⫺) of arginine (50 g/ml). The size of the amplified DNA was deduced from the molecular mass marker X (Roche Applied Science) (lane M).
carA transcription, but a higher level of transcription was observed with the deletion mutation (in FB331-14) than with the point mutation (in FB331-13) (Fig. 4). Although these mutations had a minor effect on argC transcription compared to carA transcription, the ⌬48-nucleotide deletion clearly impaired argC arginine repression (Fig. 4). DISCUSSION Like many other gram-positive bacteria, L. plantarum harbors several ArgR-like homologues (5). In order to study the roles of ArgR1 and ArgR2 in L. plantarum, arginine biosynthesis mutants were isolated. The 44 independent spontaneous arginine-deregulated mutants had common physiological traits. Compared to the parental strain FB331, all of the mutants (i) were excreting more arginine; (ii) were more resistant to canavanine, a toxic analogue of arginine; and (iii) were able to grow in carbon dioxide-enriched air either in the absence of both uracil and arginine or in the presence of arginine only (Table 4). These common traits suggest that no mutation hindered arginine transport. A mutation in arginine transport would have conferred higher resistance to canavanine and allowed growth in the presence of arginine. However, such a mutation in arginine transport would not explain arginine excretion and growth in the absence of both uracil and arginine. Instead, the observed canavanine resistance would result from deregulation of arginine biosynthesis with overproduction of CP and arginine, as discussed below.
FB331 growth is limited by carAB expression. In the genetic context of strain FB331, which lacks a functional CPS-P, CP synthesis is solely dependent on CPS-A (Fig. 1A). Previous data suggested that CPS-A cannot provide enough CP for both arginine and pyrimidine synthesis, since FB331 is unable to grow in the absence of added uracil (27). This phenotype may be due to low carAB expression. All the selected spontaneous arginine-deregulated mutants acquired the ability to grow on minimal media. Increased expression of the carAB operon was clearly identified in mutants with mutations in the carA operator (transcription studies [Fig. 3 and 4]). As a conclusion, enhanced carAB operon transcription would generate sufficient CPS-A activity to supply CP synthesis for both arginine and pyrimidine biosynthetic pathways. Thus, wild-type carAB transcription is growth limiting in minimal media in L. plantarum strains that lack CPS-P. Evidence of a regulation link between arginine metabolism and another yet-uncharacterized carbon dioxide-dependent metabolism. CPS-A activity is dependent on high concentrations of carbon dioxide (27). CPS-A may have low affinity for its substrate carbon dioxide or its dissolved form, bicarbonate. Another possibility is that other systems regulate CPS-A activity or the CP pool when L. plantarum is grown in CO2-enriched air. In this work, we found that high carAB expression did not alleviate CPS-A dependence on higher CO2 concentrations; mutants with constitutively high carAB expression did not grow in ordinary air in defined medium (DLA) without arginine and uracil (Fig. 3 and 4 [carA probe] and Table 4). Moreover, half
VOL. 186, 2004
TWO ARGININE REPRESSORS IN LACTOBACILLUS PLANTARUM
6067
FIG. 4. Comparison of arginine repression in the wild type (wt) and deregulated mutants using Northern hybridization. Five different genetic contexts were studied: the wild type (strain FB331), two mutants with altered operators in the intergenic region of the bipolar car-arg operons (T3C mutation in strain FB331-13 and a deletion in strain FB331-14), and two argR mutants (argR1 in strain FB331-4 and argR2 in strain FB331-10). These strains were cultivated either in the absence (⫺A) or presence (⫹A) of arginine. For each probe, we expressed the relative amount (defined in Materials and Methods) as a percentage of the relative amount obtained with the wild-type strain cultivated in the absence of arginine. nt, nucleotides.
of the mutants unexpectedly acquired a strict CO2 dependence on DLA (Table 4, class 1 mutants) but not on the rich undefined medium MRS (which also contains arginine and uracil). These mutants had the phenotype of arginine-deregulated mutants, since they (i) resisted high canavanine concentrations, (ii) excreted arginine, and (iii) grew on DLA and DLA complemented with arginine under carbon dioxide-enriched growth conditions. Wild-type argR1 or argR2 genes, as well as the wild-type carA-argC intergenic region, were found in 11 out of 20 class 1 mutants. Thus, even if the mutated genes have not been characterized, the phenotype of these class 1 mutants demonstrates that regulation of CPS-A activity is linked to another metabolism, which is also dependent on carbon dioxide concentrations. Additional physiologic and genetic experiments are required to identify the metabolic pathways that are altered in class 1 mutants. Two arginine repressors control transcription of arginine biosynthetic genes in L. plantarum. One-third of the isolated mutants excrete large amounts of arginine (Table 4, class 3 mutants) and were shown to be mutated in one of the two argR genes present in L. plantarum. Transcriptional studies (Fig. 3 and 4) confirmed that mutation in one of the two regulators resulted in a complete loss of arginine repression, so the activity of any of these proteins is dependent on the active pres-
ence of the other one. Single-amino-acid substitution mutations were obtained for both ArgR copies, which provides valuable information about the domains of these molecules that are involved in arginine repression. (i) Integrity of the DBDs of both ArgR1 and ArgR2 was necessary for arginine repression, so we conclude that both ArgR proteins bind to target DNA sequences. Loss of arginine repression was correlated with mutations found in the ␣3 helix, a structural motif of the DBD in ArgR1 (A39E in FB331-2; S42L in FB331-3; D44G in FB331-4) and in ArgR2 (R43C in FB331-9; R43H in FB331-10). (ii) The last C-terminal residues of both ArgR1 and ArgR2 are essential for arginine repression. Arginine regulation was abolished when the last 18 ArgR1 residues in mutant FB331-6 and the last 14 ArgR2 residues in mutant FB331-12 were deleted. These residues belong to a predicted helix structure (data not shown), despite the fact that their ArgR C-terminal peptides are not homologous. This ␣6 helix may facilitate dimerization of the two repressor trimers. (iii) ArgR1 and ArgR2 may have different corepressor binding sites. The conserved motif GTIAGDDT (Fig. 2) is required for arginine binding and dimerization of the two ArgR trimers of the well-studied single-copy repressors found in E. coli and B. subtilis. Within this motif, residue A123 seems to be essential for ArgR1 activity (mutant FB331-5). This motif is not
6068
NICOLOFF ET AL.
conserved in ArgR2 and other regulators (Fig. 2), including the two Lactococcus lactis repressors (20). This observation suggests a novel arginine regulation mechanism in gram-positive bacteria harboring several argR genes. A sequence resembling an E. coli ARG box mediates carA arginine repression but is not found in the argG operator. The proposed L. plantaraum ARG box consists of two 18-bp imperfect palindromes separated by three A bases (mismatched bases compared to the consensus E. coli ARG box shown in Fig. 1C). The T3C mutation occurred in the highly conserved half-site ARG box that harbored only one mismatch compared to the consensus sequence. When the proposed ARG box is mutated by complete deletion or by a point mutation (mutants FB331-13 and FB331-14), carA transcription is constitutive. These two mutations did not deregulate argC transcription to the same extent as carA transcription (Fig. 3 and 4). This suggests that the proposed ARG box is not common to the bipolar operons. Alternatively, since wild-type arginine repression of argC is weaker than that of carA (Fig. 3 and 4, half versus total repression, respectively), one would expect less effect of mutations impairing the arginine operator in the less tightly regulated operon. The proposed L. plantarum ARG box is localized 5 bp before the argC ⫺35 box and at least 61 bp from the ⫺35 box of the carA promoter (Fig. 1C). This organization is different from that of other reported arginine-repressed bipolar genes. The transcription start sites are adjacent in M. profunda (38), overlap by 13 nucleotides in E. coli (18), and are separated by 162 nucleotides in L. plantarum. Even though the argGH operon was arginine regulated by ArgR1 and ArgR2, no ARG box-like sequence highly similar to known ARG boxes was found in the argG operator. Therefore, ARG boxes highly similar to the E. coli ARG box consensus, like the one found in the bipolar carA-argC operators, appear not to be a prerequisite. Electrophoretic mobility shift assays and footprinting need to be performed to test whether ArgR1, ArgR2, or both repressors bind to the proposed ARG box or other sequences of the bipolar operators. A novel arginine repression mechanism in L. plantarum? In most of the bacteria in which regulation of arginine biosynthesis has been studied in detail, repression of the arginine regulon is exerted by the binding of a hexameric repressor molecule. In fact, the ArgR active form found in E. coli, B. subtilis, and G. stearothermophilus consists of a dimer of homotrimers (10, 26, 35, 37) whose interactions are strengthened by the binding of six L-arginine molecules. This is the first report of two different ArgR transcription regulators that repress arginine biosynthesis operons. The L. plantarum active repressor may be a heterohexamer of ArgR1 and ArgR2 monomers. Mutations in one of those monomers would lead to an inactive repressor form. This mutual dependence and the fact that, compared to the characterized ArgR molecules, L. plantarum ArgR1 and ArgR2 have lost conserved residues of the arginine-binding sites (see above) favor the hypothesis of cooperative binding of ArgR1 and ArgR2 to ARG boxes. Further experiments are needed to elucidate arginine repression in L. plantarum, to clarify whether the functional repressors are constituted of mixtures of ArgR1 and ArgR2 molecules, and to identify the ArgR1-ArgR2 DNA binding to cognate and heterologous operators.
J. BACTERIOL.
What is the physiological impact of argR gene duplication in gram-positive bacteria? More than a single copy of argR is frequently encountered in gram-positive bacteria (5). The number of argR paralogs varies within a given genus, such as within Bacillus (two copies in Bacillus cereus and only one in the other sequenced Bacillaceae), Lactobacillus (two homologs in L. plantarum and none in Lactobacillus johnsonii), Staphylococcus (two homologs in Staphylococcus aureus versus three copies in Staphylococcus epidermidis), and Clostridium (two homologs in Clostridium perfringens versus a single copy in Clostridium tetani). Only Enterococcus faecalis, formerly called Streptococcus faecalis, harbored four paralogs, while the other entirely sequenced Streptococcaceae harbored three copies (5; H. Nicoloff, F. Arse`ne-Ploetze, and F. Bringel, unpublished data). The fact that up to four different ArgR/AhrC-like proteins in some bacteria, and in other prokaryotes a single regulator, regulate both arginine biosynthesis and arginine catabolism raises the question of the roles of multiple regulators in a cell. Gene duplication generates functional redundancy, which is often not advantageous (39). Theoretical population genetics predicts that both duplicates can be stably maintained when some aspects of their functions differ (29), which can occur by subfunctionalization. In this case, each paralog would mutate up to the point at which its total capacity would be reduced to the level of the single-copy ancestral gene, and it would adopt part of the function of the ancestor. In the two cases where the functions of duplicated argR genes were genetically investigated, both gene products were active and required for active transcriptional regulation (reference 20 and this work). The subfunctionalization of the different paralogs can be at the level of protein function or expression. This hypothesis is in agreement with results observed in E. faecalis, where glucose and arginine influenced argR1 and argR2 transcription differently (2), and in L. lactis, where only one argR gene was implicated in the regulation of arginine catabolism (20). Thus, the presence of several argR genes may facilitate microbial response to variations of environmental growth conditions. In L. lactis, mutation in one of its two argR genes conferred improved acid resistance upon enhanced ADI arginine catabolism (30). The ADI system plays a role in the acid resistance of a number of gram-positive bacteria (for a review, see reference 7) and acts as a virulence factor in Streptococcus pyogenes (9). Unlike L. lactis, most L. plantarum organisms do not catabolize arginine, and no difference in the function of L. plantarum argR genes was found under the tested conditions, which does not exclude the possibility that these genes have a different pattern of function or expression. Whether only one ArgR protein is also sufficient to activate the arginine catabolic genes in the few L. plantarum strains described as harboring an ADI pathway (32) remains to be tested. Since genes that are functionally related are sometimes organized in operons, L. plantarum argR gene linkage was examined and compared to that in other organisms. L. plantarum and Leuconostoc mesenteroides both harbor two copies of argR that are found in their genomes adjacent to similar genes. argR1 is adjacent to pbp2A, which encodes a putative membrane carboxypeptidase (penicillinbinding protein) (National Center for Biotechnology Information conserved domain COG0744) (Table 3, group 1). argR2 is bordered by yqxC, which encodes a putative FtsJ-like methyltransferase whose substrate is the 23S rRNA (listed in the
TWO ARGININE REPRESSORS IN LACTOBACILLUS PLANTARUM
VOL. 186, 2004
National Center for Biotechnology Information conserved domains as COG1189), and recN, which encodes a DNA repair protein in E. coli (12) (Table 3, group 2). E. faecalis also contains both synteny groups but contains two additional argR copies (argR1 and argR2), which are linked to the ADI arginine catabolic operon (2). Unlike L. plantarum and L. mesenteroides (13), E. faecalis not only synthesizes arginine (25) but also degrades it. L. lactis harbors two genes with ahrC linked to yqxC-recN (Table 3, group 2) and argR linked to the arginine catabolic arc genes (20). AhrC, but not ArgR, is involved in activation of the arc genes (20). The conservation of argR genes in synteny groups may be correlated with the different physiological roles of ArgR proteins in gram-positive bacteria. We suggest that argR duplication generated complex argR regulation networks in some gram-positive bacteria in response to specific niche adaptation.
17.
18. 19. 20. 21. 22.
ACKNOWLEDGMENTS We thank Jean-Claude Hubert for fruitful discussion about this work and Isabelle Guillouard for providing an RNA extraction protocol. REFERENCES 1. Arena, M. E., F. M. Saguir, and M. C. Manca de Nadra. 1999. Arginine dihydrolase pathway in Lactobacillus plantarum from orange. Int. J. Food Microbiol. 47:203–209. 2. Barcelona-Andres, B., A. Marina, and V. Rubio. 2002. Gene structure, organization, expression, and potential regulatory mechanisms of arginine catabolism in Enterococcus faecalis. J. Bacteriol. 184:6289–6300. 3. Bringel, F., L. Frey, S. Boivin, and J. C. Hubert. 1997. Arginine biosynthesis and regulation in Lactobacillus plantarum: the carA gene and the argCJBDF cluster are divergently transcribed. J. Bacteriol. 179:2697–2706. 4. Bringel, F., and J. C. Hubert. 2003. Extent of genetic lesions of the arginine and pyrimidine biosynthetic pathways in Lactobacillus plantarum, L. paraplantarum, L. pentosus and L. casei: prevalence of CO2-dependent auxotrophs and characterization of deficient arg genes in L. plantarum. Appl. Environ. Microbiol. 69:2674–2683. 5. Bringel, F., H. Nicoloff, F. Arsene-Ploetze, and J. C. Hubert. 2002. Phylogenetic analysis and gene linkage of ArgR/AhrC orthologs revealed groups of synteny in Gram-positive bacteria. Sci. Aliments 22:133–142. 6. Burke, M., A. F. Merican, and D. J. Sherratt. 1994. Mutant Escherichia coli arginine repressor proteins that fail to bind L-arginine, yet retain the ability to bind their normal DNA-binding sites. Mol. Microbiol. 13:609–618. 7. Cotter, P. D., and C. Hill. 2003. Surviving the acid test: responses of grampositive bacteria to low pH. Microbiol. Mol. Biol. Rev. 67:429–453. 8. Czaplewski, L. G., A. K. North, M. C. Smith, S. Baumberg, and P. G. Stockley. 1992. Purification and initial characterization of AhrC: the regulator of arginine metabolism genes in Bacillus subtilis. Mol. Microbiol. 6:267–275. 9. Degnan, B. A., J. M. Palmer, T. Robson, C. E. Jones, M. Fischer, M. Glanville, G. D. Mellor, A. G. Diamond, M. A. Kehoe, and J. A. Goodacre. 1998. Inhibition of human peripheral blood mononuclear cell proliferation by Streptococcus pyogenes cell extract is associated with arginine deiminase activity. Infect. Immun. 66:3050–3058. 10. Dennis, C. C., N. M. Glykos, M. R. Parsons, and S. E. Phillips. 2002. The structure of AhrC, the arginine repressor/activator protein from Bacillus subtilis. Acta Crystallogr. D 58:421–430. 11. Dion, M., D. Charlier, H. Wang, D. Gigot, A. Savchenko, J. N. Hallet, N. Glansdorff, and V. Sakanyan. 1997. The highly thermostable arginine repressor of Bacillus stearothermophilus: gene cloning and repressor-operator interactions. Mol. Microbiol. 25:385–398. 12. Dunman, P. M., L. Ren, M. S. Rahman, V. A. Palejwala, H. S. Murphy, M. R. Volkert, and M. Z. Humayun. 2000. Escherichia coli cells defective for the recN gene display constitutive elevation of mutagenesis at 3,N(4)-ethenocytosine via an SOS-induced mechanism. Mol. Microbiol. 37:680–686. 13. Garvie, E. I. 1986. Genus Leuconostoc van Tieghem 1878, 198AL emend mut. char. Hucker and Pederson 1930, 66AL, p. 1071–1075. In P. H. A. Sneath, N. S. Mari, M. E. Sharpe, and J. G. Holt (ed.), Bergey’s manual of systematic bacteriology, vol. 2. Williams and Wilkins, Baltimore, Md. 14. Guex, N., and M. C. Peitsch. 1997. SWISS-MODEL and the Swiss-PdbViewer: an environment for comparative protein modeling. Electrophoresis 18:2714–2723. 15. Holtham, C. A., K. Jumel, C. M. Miller, S. E. Harding, S. Baumberg, and P. G. Stockley. 1999. Probing activation of the prokaryotic arginine transcriptional regulator using chimeric proteins. J. Mol. Biol. 289:707–727. 16. Karaivanova, I. M., P. Weigel, M. Takahashi, C. Fort, A. Versavaud, G. Van
23.
24.
25.
26. 27.
28. 29. 30. 31. 32. 33. 34.
35. 36. 37. 38.
39.
6069
Duyne, D. Charlier, J. N. Hallet, N. Glansdorff, and V. Sakanyan. 1999. Mutational analysis of the thermostable arginine repressor from Bacillus stearothermophilus: dissecting residues involved in DNA binding properties. J. Mol. Biol. 291:843–855. Kleerebezem, M., J. Boekhorst, R. van Kranenburg, D. Molenaar, O. P. Kuipers, R. Leer, R. Tarchini, S. A. Peters, H. M. Sandbrink, M. W. Fiers, W. Stiekema, R. M. Lankhorst, P. A. Bron, S. M. Hoffer, M. N. Groot, R. Kerkhoven, M. de Vries, B. Ursing, W. M. de Vos, and R. J. Siezen. 2003. Complete genome sequence of Lactobacillus plantarum WCFS1. Proc. Natl. Acad. Sci. USA 100:1990–1995. Krin, E., C. Laurent-Winter, P. N. Bertin, A. Danchin, and A. Kolb. 2003. Transcription regulation coupling of the divergent argG and metY promoters in Escherichia coli K-12. J. Bacteriol. 185:3139–3146. Kueh, R., N. A. Rahman, and A. F. Merican. 2003. Computational docking of L-arginine and its structural analogues to C-terminal domain of Escherichia coli arginine repressor protein (ArgRc). J. Mol. Model. 9:88–98. [Online.] Larsen, R., G. Buist, O. P. Kuipers, and J. Kok. 2004. ArgR and AhrC are both required for regulation of arginine metabolism in Lactococcus lactis. J. Bacteriol. 186:1147–1157. Lu, C. D., J. E. Houghton, and A. T. Abdelal. 1992. Characterization of the arginine repressor from Salmonella typhimurium and its interactions with the carAB operator. J. Mol. Biol. 225:11–24. Maas, W. K. 1994. The arginine repressor of Escherichia coli. Microbiol. Rev. 58:631–640. Miller, C. M., S. Baumberg, and P. G. Stockley. 1997. Operator interactions by the Bacillus subtilis arginine repressor/activator, AhrC: novel positioning and DNA-mediated assembly of a transcriptional activator at catabolic sites. Mol. Microbiol. 26:37–48. Morin, A., N. Huysveld, F. Braun, D. Dimova, V. Sakanyan, and D. Charlier. 2003. Hyperthermophilic Thermotoga arginine repressor binding to fulllength cognate and heterologous arginine operators and to half-site targets. J. Mol. Biol. 332:537–553. Murray, B. E., K. V. Singh, R. P. Ross, J. D. Heath, G. M. Dunny, and G. M. Weinstock. 1993. Generation of restriction map of Enterococcus faecalis OG1 and investigation of growth requirements and regions encoding biosynthetic function. J. Bacteriol. 175:5216–5223. Ni, J., V. Sakanyan, D. Charlier, N. Glansdorff, and G. D. Van Duyne. 1999. Structure of the arginine repressor from Bacillus stearothermophilus. Nat. Struct. Biol. 6:427–432. Nicoloff, H., J. C. Hubert, and F. Bringel. 2000. In Lactobacillus plantarum, carbamoyl phosphate is synthesized by two carbamoyl-phosphate synthetases (CPS): carbon dioxide differentiates the arginine-repressed from the pyrimidine-regulated CPS. J. Bacteriol. 182:3416–3422. Niersbach, H., R. Lin, G. D. Van Duyne, and W. K. Maas. 1998. A superrepressor mutant of the arginine repressor with a correctly predicted alteration of ligand binding specificity. J. Mol. Biol. 279:753–760. Nowak, M. A., M. C. Boerlijst, J. Cooke, and J. M. Smith. 1997. Evolution of genetic redundancy. Nature 388:167–171. Rallu, F., A. Gruss, S. D. Ehrlich, and E. Maguin. 2000. Acid- and multistress-resistant mutants of Lactococcus lactis: identification of intracellular stress signals. Mol. Microbiol. 35:517–528. Rodriguez-Garcia, A., M. Ludovice, J. F. Martin, and P. Liras. 1997. Arginine boxes and the argR gene in Streptomyces clavuligerus: evidence for a clear regulation of the arginine pathway. Mol. Microbiol. 25:219–228. Spano, G., G. Chieppa, L. Beneduce, and S. Massa. 2004. Expression analysis of putative arcA, arcB, and arcC genes partially cloned from Lactobacillus plantarum isolated from wine. J. Appl. Microbiol. 96:185–193. Stirling, C. J., G. Szatmari, G. Stewart, M. C. Smith, and D. J. Sherratt. 1988. The arginine repressor is essential for plasmid-stabilizing site-specific recombination at the ColE1 cer locus. EMBO J. 7:4389–4395. Stover, C. K., X. Q. Pham, A. L. Erwin, S. D. Mizoguchi, P. Warrener, M. J. Hickey, F. S. Brinkman, W. O. Hufnagle, D. J. Kowalik, M. Lagrou, R. L. Garber, L. Goltry, E. Tolentino, S. Westbrock-Wadman, Y. Yuan, L. L. Brody, S. N. Coulter, K. R. Folger, A. Kas, K. Larbig, R. Lim, K. Smith, D. Spencer, G. K. Wong, Z. Wu, I. T. Paulsen, J. Reizer, M. H. Saier, R. E. Hancock, S. Lory, and M. V. Olson. 2000. Complete genome sequence of Pseudomonas aeruginosa PA01, an opportunistic pathogen. Nature 406:959–964. Sunnerhagen, M., M. Nilges, G. Otting, and J. Carey. 1997. Solution structure of the DNA-binding domain and model for the complex of multifunctional hexameric arginine repressor with DNA. Nat. Struct. Biol. 4:819–826. Tian, G., and W. K. Maas. 1994. Mutational analysis of the arginine repressor of Escherichia coli. Mol. Microbiol. 13:599–608. Van Duyne, G. D., G. Ghosh, W. K. Maas, and P. B. Sigler. 1996. Structure of the oligomerization and L-arginine binding domain of the arginine repressor of Escherichia coli. J. Mol. Biol. 256:377–391. Xu, Y., Y. Sun, N. Huysveld, D. Gigot, N. Glansdorff, and D. Charlier. 2003. Regulation of arginine biosynthesis in the psychropiezophilic bacterium Moritella profunda: in vivo repressibility and in vitro repressor-operator contact probing. J. Mol. Biol. 326:353–369. Zhang, J. 2003. Evolution by gene duplication: an update. Trends Ecol. Evol. 18:292–298.
