Chemotaxis of Ralstonia eutropha JMP134(pJP4) to the Herbicide 2,4-Dichlorophenoxyacetate

Ralstonia eutropha JMP134(pJP4) and several other species ofmotile bacteria can degrade the herbicide 2,4-dichlorophenoxyacetate(2,4-D), but it was not known if bacteria could sense and swimtowards 2,4-D by the process of chemotaxis. Wild-type R. eutrophacells were chemotactically attracted to 2,4-D in swarm plateassays and qualitative capillary assays. The chemotactic responsewas induced by growth with 2,4-D and depended on the presenceof the catabolic plasmid pJP4, which harbors the tfd genes for2,4-D degradation. The tfd cluster also encodes a permease for2,4-D named TfdK. A tfdK mutant was not chemotactic to 2,4-D,even though it grew at wild-type rates on 2,4-D.

INTRODUCTION

Top

Introduction

Most motile bacteria can sense and respond to low concentrationsof organic compounds in their environment by the process ofchemotaxis. There is evidence that chemotaxis can enhance biodegradation(11), presumably by rapidly bringing cells into close contactwith degradable substrates. 2,4-Dichlorophenoxyacetate (2,4-D)is a widely used herbicide (Industry Task Force on 2,4-D researchdata [http://www.24d.org]) that can be degraded by a numberof species of motile bacteria (12).

The bacterium Ralstonia eutropha JMP134(pJP4) grows on 2,4-Dby using genes present on a self-transmissible plasmid calledpJP4 (3). Plasmid pJP4 also encodes a permease for 2,4-D namedTfdK (10). The predicted TfdK protein has 12 membrane-spanningregions and is a member of the aromatic acid:H⁺ symporter familyof the major facilitator superfamily of transport proteins (15).It is 33% identical in amino acid sequence to PcaK, a 4-hydroxybenzoatepermease from Pseudomonas putida that is also required for chemotaxisto 4-hydroxybenzoate (2, 7, 13). The similarity of TfdK to PcaKled us to investigate if R. eutropha is chemotactically attractedto 2,4-D and, if so, whether tfdK is required for the chemotacticresponse.

Bacterial strains and experimental procedures.

Top

Bacterial strains and...

Wild-type R. eutropha [strain JMP134(pJP4)], a plasmid pJP4-curedderivative (strain JMP289), and a tfdK mutant [strain JMP134(pJP4::cba79)](10) were obtained from J. R. van der Meer. Strains were grownon mineral salts medium (7) at 30°C with shaking at 250rpm. Analysis by gas chromatography-mass spectrometry as previouslydescribed (17) indicated that the 2,4-D (Sigma-Aldrich, St.Louis, Mo.) used in these studies contained no detectable contaminants.Bacterial transformations were carried out according to themethod of Hanahan (6).

Plasmid DNA was prepared from Escherichia coli with a QIAprepspin miniprep kit (Qiagen Inc., Chatsworth, Calif.). PlasmidDNA was prepared from R. eutropha as described previously (18).DNA fragments were purified from agarose gels with the QIAquickgel extraction kit (Qiagen Inc.).

Plasmid pHAH108, a broad-host-range plasmid designed for expressionof TfdK, was constructed in several steps. First, a 1,388-bpfragment of DNA encompassing the tfdK gene was amplified frompJP4 by PCR and cloned into pAF7 (A. Ferrandez and C. S. Harwood,unpublished data) to form pHAH107. Plasmid pAF7 is a derivativeof pT7-6 (20) that carries the P. putida pcaK gene fused atits predicted N terminus to a hemagglutinin (HA) epitope sequencederived from the X47 virus hemagglutinin (14) behind a transcriptionalenhancer and a ribosome-binding site (RBS).

The PCR-amplified tfdK gene was cloned into the NaeI and BamHIsites of pAF7, replacing the pcaK gene. An EcoRI/HindIII fragmentfrom pHAH107 containing HA-tfdK, the transcriptional enhancer,and the consensus RBS was then subcloned into the broad-host-rangevector pBBR1MCS-5 (9) to create pHAH108. Plasmids were introducedinto R. eutropha by conjugation (1) from Escherichia coli DH5 ${alpha}$ in triparental matings using E. coli CC118(pRK600) (8) to providethe transfer functions. Gentamicin (20 µg per ml), chloramphenicol(100 µg per ml), and ampicillin (100 µg per ml)were added to growth media to select for plasmids in E. coliand R. eutropha. Western blot analysis (20) with anti-HA antiserum(Roche Molecular Biochemicals, Indianapolis, Ind.) was usedto show that HA-TfdK was expressed in R. eutropha JMP134(pJP4::cba79,pHAH108).

R. eutropha JMP134(pJP4) cells are attracted to 2,4-D, and this is a plasmid-encoded, inducible trait.

Top

R. eutropha JMP134(pJP4) cells...

Chemotaxis was tested with a soft agar swarm plate assay anda modified capillary assay. For the soft agar swarm plate assay,an R. eutropha colony was stabbed into the center of a mineralsalts medium plate that contained 0.25 mM 2,4-D or 1.0 mM succinateand was solidified with 0.3% Noble agar (Difco Laboratories,Detroit, Mich.). Wild-type cells had a positive chemotacticresponse to succinate and to 2,4-D, as evidenced by a sharpring of growth that formed when cells responded to the gradientof attractant that was created as they metabolized the carbonsource in the plates (Fig. 1A).

View larger version (54K):
[in this window]
[in a new window]
FIG. 1. Chemotactic response of R. eutropha JMP134(pJP4). (A) Chemotactic rings formed by wild-type cells on soft agar swarm plates containing 0.5 mM succinate (left) or 0.25 mM 2,4-D (right). Plates were inoculated by stabbing motile cells in the center of the plate at a point corresponding to the center of the swarm ring and incubated for 24 to 48 h at 30°C. (B) Attraction of wild-type cells in a modified capillary assay. Cells were grown on 3 mM 2,4-D. Capillaries contained 10 mM 2,4-D, 1% Casamino Acids, or no attractant (chemotaxis buffer). Chemotactic responses are shown in photographic form (top) and illustrative form (bottom). Illustrations for chemotaxis buffer were omitted because no response was observed.

Modified capillary assays allowed qualitative assessment ofchemotaxis with a phase contrast microscope (5). With this method,chemotaxis can be visualized in the absence of metabolism ofthe attractant. Capillaries (1 µl) contained attractantin 1% low-melting-temperature agarose dissolved in chemotaxisbuffer (50 mM potassium phosphate [pH 7.0], 20 µM EDTA,0.05% glycerol). Cells that had been grown in mineral medium(7) containing 3 mM 2,4-D, 10 mM succinate, or 10 mM succinateplus 0.5 mM 2,4-D were harvested in the mid-logarithmic phaseof growth and suspended in chemotaxis buffer to a final A₆₆₀of approximately 0.1. Samples of the cell suspension were placedin a chamber formed by a microscope slide, a glass U-tube, anda cover slip, and the capillary containing the attractant wasinserted into the suspension. Cell behavior around the tip ofthe capillary was observed at a final magnification of x40 andphotographed at 0 min, 5 min, and 10 min. Modified capillaryassays were repeated at least five times for each conditiontested with reproducible results.

R. eutropha wild-type cells formed visible clouds of turbidityas they accumulated around the open tips of capillaries thatcontained either 10 mM 2,4-D or 1% Casamino Acids (Fig. 1B).Wild-type cells also accumulated around the open tips of capillariesthat contained 1 mM 2,4-D, but this response was less dramaticand difficult to photograph (data not shown). Chemotaxis buffer,used as a control, did not elicit a chemotactic response fromR. eutropha in the modified capillary assay. R. eutropha JMP134(pJP4)cells that had been grown in 10 mM succinate did not respondto 2,4-D in modified capillary assays. However, these cellswere attracted to 2,4-D when grown on 10 mM succinate plus 0.5mM 2,4-D (data not shown). This indicates that chemotaxis to2,4-D was induced by growth with 2,4-D. R. eutropha cells thathad been cured of the pJP4 plasmid (strain JMP289) and grownon 10 mM succinate plus 0.5 mM 2,4-D showed little or no attractionto 2,4-D in a modified capillary assay (data not shown). Thisindicates that 2,4-D chemotaxis is a plasmid-encoded trait.

tfdK gene is required for chemoattraction to 2,4-D.

Top

tfdK gene is required...

TfdK is a membrane-spanning protein that catalyzes the activetransport of 2,4-D into cells. The TfdK permease allows cellsto take up very low concentrations of 2,4-D, but it is not essentialfor entry of 2,4-D into cells, as 2,4-D can cross the cell membraneof R. eutropha by simple diffusion (10). When 2,4-D was suppliedat a concentration of 3 mM, the tfdK mutant had a specific growthrate with 2,4-D (0.324 ± 0.064 h^-1) that was nearly identicalto that of the wild type (0.333 ± 0.044 h^-1). A tfdKmutant did not form chemotactic rings in soft agar swarm platescontaining 2,4-D, even after incubation for 2 days (Fig. 2A). Instead,the cells formed a fuzzy, slow-moving ring that reflected randommovement of motile cells.

View larger version (58K):
[in this window]
[in a new window]
FIG. 2. Response of a tfdK mutant to 2,4-D. (A) Chemotactic swarms formed by wild-type cells, the tfdK mutant, and the tfdK mutant expressing the tfdK gene in trans on swarm plates containing 0.25 mM 2,4-D. (B) Modified capillary assays showing the response of tfdK mutant cells and complemented tfdK mutant cells to 2,4-D. Both strains were grown on 3 mM 2,4-D. Results are shown in photographic form (top) and illustrative form (bottom).

tfdK mutant cells that had been grown on 3 mM 2,4-D were attractedto 1% Casamino Acids but not to 10 mM 2,4-D in modified capillaryassays (Fig. 2B). The tfdK mutant sometimes had a slight positiveresponse to 2,4-D, but this response was no greater than thatexhibited by cells of the pJP4-cured strain, JMP289, that hadbeen grown under the same conditions (10 mM succinate plus 0.5mM 2,4-D).

When the TfdK protein was expressed in trans from plasmid pHAH108in the tfdK mutant, cells formed chemotactic rings on 2,4-Dswarm plates (Fig. 2A) and were attracted to 2,4-D in modifiedcapillary assays (Fig. 2B). The vector pBBR1MCS-5 did not complementthe tfdK mutant in chemotaxis assays.

Conclusions.

Top

Conclusions.

The coordinate regulation of 2,4-D chemotaxis with 2,4-D degradationsuggests that chemotaxis may be an integral feature of the biodegradationof 2,4-D by R. eutropha. Various strains of Pseudomonas cansense and swim towards the pollutants benzene, toluene, trichloroethylene,and naphthalene as well as to the aromatic acid 4-hydroxybenzoate(4, 5, 7, 16). Chemotaxis to these compounds requires priorgrowth with toluene (for the benzene, toluene, and trichloroethyleneresponses), naphthalene, or 4-hydroxybenzoate. The coregulationof chemotaxis and degradation often reflects the coordinateinduction of a chemoreceptor gene with degradation genes.

Recently, a methyl-accepting chemotaxis protein (MCP) gene requiredfor chemotaxis to naphthalene was found to be cotranscribedwith meta-cleavage pathway genes present on the NAH7 catabolicplasmid for naphthalene degradation (5). MCPs have been wellstudied in E. coli and Salmonella spp., in which they serveas cell surface receptors for chemoattractants. On binding anattractant, an MCP undergoes a conformational change that initiatessensory signal transduction by altering the activity of CheA,an associated sensor kinase. CheA-P then acts as a phosphodonorfor the response regulator, CheY, which interacts with rotationalswitch proteins in the flagellar motors. This causes a changein swimming behavior so that cells migrate towards chemoattractants(19).

The data presented here suggest that plasmid pJP4 for 2,4-Ddegradation also encodes a chemoreceptor protein that is coordinatelyregulated with 2,4-D degradation genes. However, this protein,TfdK, does not resemble an MCP. Instead, it is homologous toPcaK, a protein that in P. putida functions in the transportof 4-hydroxybenzoate as well as in chemotaxis to 4-hydroxybenzoate(2, 13). TfdK is thus a second example of a major facilitatorsuperfamily transport protein that has dual roles in transportand in chemotaxis.

We do not understand how either PcaK or TfdK functions in chemotaxis.These permeases could play a direct role in signaling and communicatea conformational change in protein structure that occurs onligand binding or transport to a physically associated MCP orother chemosensory protein. Alternatively, they may have a moreindirect role in chemotaxis. Since 2,4-D can diffuse into R.eutropha at rates sufficient to support wild-types rates ofgrowth under the conditions in which 2,4-D chemotaxis was measured,simple accumulation of this aromatic acid within cells doesnot appear to be sufficient for chemotaxis to occur; TfdK mustbe present. A more likely scenario for an indirect role is thatTfdK functions in chemotaxis by delivering a high local concentrationof chemoattractant to chemosensory proteins present on the cytoplasmicside of the cell membrane.

ACKNOWLEDGMENTS

We thank Jan Roelof van der Meer for supplying strains usedin this study, Rebecca Parales for helpful discussions, andJuan Parales for assistance with GC-MS analysis.

This work was supported by Public Health Service grant GM56665from the National Institute of General Medical Sciences. A.C.H.has been supported by a National Science Foundation ResearchTraining Grant (DBI9602247) and by a fellowship from the Universityof Iowa Center for Biocatalysis and Bioprocessing.

FOOTNOTES

* Corresponding author. Mailing address: Department of Microbiology, 3-432 BSB, The University of Iowa, Iowa City, IA 52242. Phone: (319) 335-7783. Fax: (319) 335-7679. E-mail: caroline-harwood{at}uiowa.edu.

REFERENCES

Top