An energy taxis transducer promotes root colonization by Azospirillum brasilense

Abstract

Motility responses triggered by changes in the electron transport system are collectively known as energy taxis. In Azospirillum brasilense, energy taxis was shown to be the principal form of locomotor control. In the present study, we have identified a novel chemoreceptor-like protein, named Tlp1, which serves as an energy taxis transducer. The Tlp1 protein is predicted to have an N-terminal periplasmic region and a cytoplasmic C-terminal signaling module homologous to those of other chemoreceptors. The predicted periplasmic region of Tlp1 comprises a conserved domain that is found in two types of microbial sensory receptors: chemotaxis transducers and histidine kinases. However, the function of this domain is currently unknown. We characterized the behavior of a tlp1 mutant by a series of spatial and temporal gradient assays. The tlp1 mutant is deficient in (i) chemotaxis to several rapidly oxidizable substrates, (ii) taxis to terminal electron acceptors (oxygen and nitrate), and (iii) redox taxis. Taken together, the data strongly suggest that Tlp1 mediates energy taxis in A. brasilense. Using qualitative and quantitative assays, we have also demonstrated that the tlp1 mutant is impaired in colonization of plant roots. This finding supports the hypothesis that energy taxis and therefore bacterial metabolism might be key factors in determining host specificity in Azospirillum-grass associations.

FIG. 1.

Physical map of the 4,113-bp DNA region encompassing the tlp1 gene (A) and predicted domain architecture of Tlp1 (B). (A) The arrows indicate the direction of transcription. The triangle above the tlp1 gene indicates the insertion of the Km^r cassette. (B) Domains in the predicted Tlp1 protein, as defined by the SMART database (24): HAMP domain (7); MA, methyl-accepting chemotaxis-like domain (23). Black rectangles depict transmembrane regions.

FIG. 2.

A protein domain family exemplified by the N-terminal periplasmic region of Tlp1 from A. brasilense. Multiple alignment of homologous amino acid sequences was constructed with the CLUSTAL_X program (40). The start and end positions (domain boundaries) are shown to the right of each sequence. A consensus for multiple alignment (85% threshold) determined by using the CONSENSUS script (www.bork.embl-heidelberg.de/Alignment/consensus.html) is shown at the bottom. Identical residues are highlighted in black, and chemically similar residues are highlighted in gray. Each sequence in the alignment is identified by its GenBank protein identification number (except for the Tlp1 protein, for which the GenBank accession number of the corresponding DNA region is given) and by the abbreviated name of the organism. Abbreviations: HK, histidine kinase; Abra, A. brasilense; Atum, A. tumefaciens; Bjap, B. japonicum; Bsub, B. subtilis; Cthe, Clostridium thermocellum; Ddes, Desulfovibrio desulfuricans; Gmet, G. metallireducens; Gsul, Geobacter sulfurreducens; Mmag, M. magnetotacticum, Mag, Magnetococcus sp.; Pflu, P. fluorescens; Psyr, Pseudomonas syringae; Rsol, Ralstonia solanacearum; Bfug, Burkholderia fungorum; Sone, Shewanella oneidensis; Syn, Synechococcus; Telo, Thermosynechococcus elongatus; Vcho, Vibrio cholerae; Vpar, Vibrio parahaemolyticus; Vvul, Vibrio vulnificus; Wsuc, Wolinella succinogenes; Xaxo, Xanthomonas axonopodis; Xcam, Xanthomonas campestris.

FIG. 3.

Neighbor-joining tree showing the phylogenetic relationships within the protein domain family exemplified by the N-terminal periplasmic region of Tlp1 from A. brasilense. The tree was constructed from the multiple alignment shown in Fig. 2 with the MEGA phylogenetic package (22). Thick lines mark branches with significant (≥60%) bootstrap support (1,000 replicates). GenBank accession numbers and species abbreviations are the same as defined in the legend to Fig. 2.

FIG. 4.

The tlp1 mutant is deficient in chemotaxis. Chemotaxis of the A. brasilense wild-type strain (Sp7) and the tlp1 mutant was compared by the swarm plate assay. Swarm diameters were measured after incubation at 28°C for 48 h. (A) Representative swarm plate with fumarate as the sole carbon source. (B) The average swarm diameters are expressed as the percentage relative to that of wild-type strain (defined as 100%). Error bars represent standard deviations from the mean calculated from at least six repetitions. Differences in the swarming diameters of the wild type and the tlp1 mutant were found to be statistically significant with the following chemoeffectors: maltose, glycerol, succinate, citrate, fumarate, malate, and pyruvate.

FIG. 5.

The tlp1 mutant is altered for aerotaxis. The ability of the A. brasilense wild-type strain (Sp7) and the tlp1 mutant to form a sharp aerotactic band in a gradient of oxygen formed in a capillary tube was compared. (A) Representative capillary assay for aerotaxis showing the position of the aerotactic band in the gradient for the wild-type strain (Sp7) and the tlp1 mutant. The arrow indicates the position of the meniscus at the air-liquid interface. Fumarate (2.5 mM) was added as the sole carbon and energy source to the suspension of motile cells. Magnification, ×35. (B) The average distances of the aerotactic band to the meniscus expressed as a percentage of the distance for the wild-type strain (defined as 100%). Malate or fumarate was added at 2.5 mM as the sole substrate to the suspension of motile cells, and the distance of the aerotactic band to the meniscus was measured. Error bars represent standard deviations from the mean calculated from three independent experiments. Similar results were obtained with succinate as the sole carbon source.

FIG. 6.

The tlp1 mutant shows altered energy taxis under anaerobic conditions. Tactic responses of the A. brasilense wild-type strain (Sp7) and the tlp1 mutant were measured in the presence of nitrate as electron acceptor and succinate or fructose as electron donor. Plates were incubated anaerobically. Control plates containing no electron acceptor or no electron donor did not support growth. The average swarm diameters are expressed as a percentage of the wild-type swarm diameter (defined as 100%). Error bars represent standard deviation from the mean calculated from three independent experiments. The differences in the swarm diameters of the wild type and the tlp1 mutant were statistically significant.

FIG. 7.

The tlp1 mutant is impaired in colonization of wheat root surfaces. (A) The pattern of root surface colonization was visualized on wheat roots 10 days after inoculation with similar levels of the A. brasilense Sp 7 wild type and the tlp1 mutant. Each strain harbored plasmid pJBA21TC, which constitutively expresses β-glucuronidase, with activity indicated by the blue color. (B) Colonization levels of wheat roots by A. brasilense wild-type strain Sp7 and the tlp1 mutant 10 days after inoculation with similar numbers of cells. The error bars represent the standard deviations about the mean calculated for five plants. Similar results were obtained in three independent experiments.