Inactivation and regulation of the aerobic C(4)-dicarboxylate transport (dctA) gene of Escherichia coli

Abstract

The gene (dctA) encoding the aerobic C(4)-dicarboxylate transporter (DctA) of Escherichia coli was previously mapped to the 79-min region of the linkage map. The nucleotide sequence of this region reveals two candidates for the dctA gene: f428 at 79.3 min and the o157a-o424-o328 (or orfQMP) operon at 79.9 min. The f428 gene encodes a homologue of the Sinorhizobium meliloti and Rhizobium leguminosarum H(+)/C(4)-dicarboxylate symporter, DctA, whereas the orfQMP operon encodes homologues of the aerobic periplasmic-binding protein- dependent C(4)-dicarboxylate transport system (DctQ, DctM, and DctP) of Rhodobacter capsulatus. To determine which, if either, of these loci specify the E. coli DctA system, the chromosomal f428 and orfM genes were inactivated by inserting Sp(r) or Ap(r) cassettes, respectively. The resulting f428 mutant was unable to grow aerobically with fumarate or malate as the sole carbon source and grew poorly with succinate. Furthermore, fumarate uptake was abolished in the f428 mutant and succinate transport was approximately 10-fold lower than that of the wild type. The growth and fumarate transport deficiencies of the f428 mutant were complemented by transformation with an f428-containing plasmid. No growth defect was found for the orfM mutant. In combination, the above findings confirm that f428 corresponds to the dctA gene and indicate that the orfQMP products play no role in C(4)-dicarboxylate transport. Regulation studies with a dctA-lacZ (f428-lacZ) transcriptional fusion showed that dctA is subject to cyclic AMP receptor protein (CRP)-dependent catabolite repression and ArcA-mediated anaerobic repression and is weakly induced by the DcuS-DcuR system in response to C(4)-dicarboxylates and citrate. Interestingly, in a dctA mutant, expression of dctA is constitutive with respect to C(4)-dicarboxylate induction, suggesting that DctA regulates its own synthesis. Northern blot analysis revealed a single, monocistronic dctA transcript and confirmed that dctA is subject to regulation by catabolite repression and CRP. Reverse transcriptase-mediated primer extension indicated a single transcriptional start site centered 81 bp downstream of a strongly predicted CRP-binding site.

FIG. 1

Restriction maps of the dctA-orfQMP region of the E. coli chromosome. The inserts cloned in λ578, λ605, pGS753, pGS754, pGS928, pDctA, pOrfQMP, pDctA::Sp and pOrfQMP::Ap are shown along with E. coli DNA (thick black lines) and the Ap^r and Sp^r resistance cassettes (open bars). Relevant restriction sites are indicated: B, BamHI; Bc, BclI; Bg, BglII; Bs, BsphI; C, ClaI; E, EcoRI; H, HindIII; Hp, HpaI; P, PstI; S, SalI; and Sm, SmaI. Restriction sites within vector DNA are denoted by a v, and hybrid restriction sites no longer recognized by the corresponding enzymes are in parentheses. Solid arrows indicate the positions and polarities of relevant structural genes. The hatched bar represents the DNA fragment used as a hybridization probe, and a strongly predicted stem-loop structure is also indicated. Coordinates are from reference .

FIG. 2

PCR and Southern blot analysis of dctA::spc and orfM::amp mutants. PCR (A and B) and Southern blotting (C and D) were performed, as described in Materials and Methods, with chromosomal DNA from JC7623 (lanes 1), JC7623 dctA::spc (MDO1) (lanes 2), and JC7623 orfM::amp (MDO2) (lanes 3), together with PCR primers or hybridization probes specific for the dctA (A and C) and orfQMP (B and D) regions. The sizes of the PCR products and major hybridizing bands are shown. Analysis of MDO3 gave results similar to those of MDO2 (data not shown).

FIG. 3

Effects of the dctA::spc mutation on growth. Cultures were grown aerobically in M9 salts medium containing 0.4% glucose (A), 50 mM succinate (B), 50 mM malate (C), or 50 mM fumarate: AN387 (dctA⁺) (□), MDO800 (dctA::spc) (▵), AN387(pDctA) (■), and MDO800(pDctA) (▴). OD_{650 nm}, optical density at 650 nm.

FIG. 4

Effects of the dctA::spc mutation on fumarate transport. Cultures were grown aerobically in L broth, harvested, washed in M9 salts solution, and then used to measure the rate of [2,3-¹⁴C]fumarate uptake, in duplicate, as described in Materials and Methods. The strains were AN387 (dctA⁺) (□), MDO800 (dctA::spc) (▵), AN387(pDctA) (■), and MDO800(pDctA) (▴). Standard deviations are shown.

FIG. 5

Expression of a dctA-lacZ transcriptional fusion during aerobic growth in L broth. The β-galactosidase activities (solid symbols) and culture densities (open symbols) of JRG3351 (dctA-lacZ) are shown after growth at 37°C in L broth (◊ and ⧫), L broth with 50 mM succinate (□ and ■), and L broth with 1% glucose (○ and ●). Error bars represent standard deviations for two cultures, each assayed in duplicate. OD_{650 nm}, optical density at 650 nm.

FIG. 6

Effects of various carboxylates and the dcuS mutation on dctA-lacZ expression. Cultures of JRG3351 (dcuS⁺) (A) and JRG3984 (dcuS) (B) were grown aerobically to the postexponential phase in L broth supplemented with carboxylates (50 mM) as indicated. The β-galactosidase activities are shown with standard deviations.

FIG. 7

Expression of dctA-lacZ during aerobic (A) and anaerobic (B) growth. Growth of JRG3351 took place in L broth or M9 minimal medium with or without 50 mM succinate (Suc), malate (Mal), fumarate (Fum), or nitrate or 0.4% glycerol (Gly) or maltose (Malt), or 0.4% glucose (Glu) in minimal medium (M) and 1% glucose in L broth (L). β-Galactosidase activities were assayed in duplicate with samples taken from duplicate cultures at 0.5- to 1-h intervals over the entire growth phase, but only those corresponding to early-logarithmic (open bars) and postexponential (solid bars) growth are shown. Standard deviations are indicated.

FIG. 8

Effects of ArcA, FNR, cAMP, and CRP on dctA-lacZ expression. Cultures were grown anaerobically (A) or aerobically (B to D) at 37°C in L broth plus 0.4% glycerol and 50 mM fumarate (A), in L broth only (B and D), or in L broth with and without 1% glucose (Glu) and 5 mM cAMP (C). The strains were JRG3351 (wild type [wt]), JRG4011 (arcA), JRG4013 (fnr), and JRG4016 (crp) and the corresponding transformants containing plasmids pRB38 (arcA⁺), pCH21 (fnr⁺), and pGS279 (crp⁺). Note that JRG4017 (the corresponding crp⁺ control for JRG4016) gave very similar results to those obtained with JRG3351 (data not shown). Other details are as for Fig. 7.

FIG. 9

Effects of the dctA mutation on dctA-lacZ expression. Cultures of JRG3351 (dctA⁺) (open bars) or JRG4005 (dctA::spc) (solid bars) were grown aerobically at 37°C in M9 salts medium containing 0.4% glycerol with or without 50 mM malate, succinate, or maleate. Other details are as for Fig. 7.

FIG. 10

Analysis of the dctA transcript. (A) Northern blotting. Total RNA was extracted from MC1000 (wild type [wt]), JRG1999 (crp), and JRG1999 transformed with pGS279 (crp⁺). Strains were grown aerobically in L broth with or without 1% glucose (Glu). Following electrophoresis and capillary transfer, RNA was hybridized with a labeled dctA fragment. The arrow on the right indicates the specifically hybridizing dctA transcripts. The size of the dctA mRNA transcript (in kilonucleotides) is indicated. (B) Determination of the dctA transcriptional start site by reverse transcriptase-mediated primer extension. RNA was isolated from MC1000 grown aerobically in L broth. Lane P indicates the primer extension product with primer P2_dctA. Similar results were obtained with primer P1_dctA (results not shown). The sequencing ladder (lanes A, C, G, and T) was generated by using primer P2_dctA and pGS753 as template. The corresponding nucleotide sequence, its complement, and the transcriptional start site (indicated by an arrow) are shown. (C) Nucleotide sequence of the dctA promoter region. Coordinates are from reference . The experimentally determined +1 site is boxed and labeled, as are the deduced −35 and −10 sites and the predicted CRP site. Residues matching the corresponding consensus sequence for the CRP, −35, −10, +1, and Shine-Dalgarno (S-D) sites are in bold and underlined. The positions of the predicted start codon of dctA and termination codon of f651 are in bold.