|Home | About | Journals | Submit | Contact Us | Français|
Ni2+ and Co2+ are sensed as repellents by the Escherichia coli Tar chemoreceptor. The periplasmic Ni2+ binding protein, NikA, has been suggested to sense Ni2+. We show here that neither NikA nor the membrane-bound NikB and NikC proteins of the Ni2+ transport system are required for repellent taxis in response to Ni2+.
Escherichia coli cells are repelled by Ni2+ and, with lower sensitivity, Co2+ (21). This response is mediated primarily by the aspartate/maltose chemoreceptor, Tar. A Tar-Tsr chimeric receptor fused at residues 256 and 257 of Tar still senses Ni2+, whereas the reciprocal Tsr-Tar chimera does not (15). The authors of that study concluded that Ni2+ is sensed by the N-terminal periplasmic region of Tar. The fusion joint is actually near the C-terminal end of AS2, the second amphipathic helix of the HAMP domain (4) that couples the transmembrane sensing domain to the cytoplasmic kinase control domain. Thus, a more cautious interpretation of their results is that the ability to sense Ni2+ is conferred by the periplasmic, transmembrane, or HAMP region of Tar.
The five-gene nikABCDE operon encodes an ATP-dependent high-affinity uptake system for Ni2+. This operon is quite similar in its construction to the five-gene operons encoding the oligopeptide (Opp) and dipeptide (Dpp) transport systems (2, 14). Furthermore, the periplasmic binding proteins encoded by the first gene of all three operons are very similar in their folds (23). The DppA protein interacts with the Tap chemoreceptor of E. coli and is the substrate recognition component of the attractant chemotaxis response to dipeptides (1, 8, 16).
The NikA binding protein (19) has been suggested to be the substrate recognition component of repellent chemotaxis to Ni2+ (7). However, there are several problems with this proposal. First, NikA is produced only under conditions of anaerobiosis (7) and Ni2+ limitation (6), but Ni2+ taxis is seen in cells grown aerobically in tryptone broth (18), whether or not NiSO4 is present (9). Second, concentrations of Ni2+ that are needed to see significant responses to up or down step changes are between 10 and 100 μM (22), whereas the dissociation constant (Kd) for Ni2+ binding to NikA is on the order of 0.1 μM (7). Third, the other periplasmic binding proteins of E. coli that are involved in chemotaxis—DppA, the ribose-binding protein (RBP) (3), the galactose/glucose-binding protein (GBP) (13), and the maltose-binding protein (MBP) (12)—all mediate attractant taxis. Thus, NikA would have to evoke a response opposite from those generated by the other binding proteins.
These apparent discrepancies led us to examine whether NikA actually is the Ni2+ sensor in E. coli. We obtained knockout mutations of the nikA, nikB, and nikC genes from the Keio collection (5). These mutations replace the bulk of a given gene sequence with a kanamycin resistance cassette. The knockout mutations were transferred into the chemotactically wild-type strain CV1 (identical to RP437) (20), and the transfer of the mutations was confirmed by PCR analysis. To ensure that we were always working with mutant cells, we left the Kanr cassettes in the disrupted genes. Although the nikA insertion could have a polar effect on nikBCDE and the nikB insertion could have a polar effect on nikCDE, we could still independently assess the effect of knocking out Ni2+ transport while retaining NikA with the nikB and nikC insertions and the effect of eliminating Ni2+ binding protein and transport with the nikA insertion.
The effects of the nikA, nikB, and nikC mutations on chemotaxis were assessed using our recently described microfluidic chemotaxis device (9). In this device, diffusive mixing between two inlet concentrations of a chemoeffector is used to generate a gradient of the chemoeffector. Bacteria entering the device immediately encounter the midpoint of the gradient and are exposed to it for 18 to 21 s before imaging. This assay allows easy and rapid quantification of the chemotactic response.
Figure Figure11 shows the response of wild-type and Δtar and nik mutant cells to gradients of aspartate and NiSO4. High-motility E. coli cells were prepared as described previously (9), except that cells were harvested and washed by centrifugation at 150 × g instead of by filtering. The low-speed centrifugation method produced a higher proportion of fully motile cells. Cells in chemotaxis buffer containing 50 μM l-aspartate or 122.5 μM NiSO4 were introduced at the midpoint of 0 to 50 mM aspartate or 0 to 225 μM NiSO4 gradients. CV1 cells give a very clear response in a 0 to 100 μM gradient of l-aspartate (Fig. (Fig.1A).1A). CV1 cells also show a net migration toward lower concentrations of NiSO4 (Fig. (Fig.1B).1B). Cells of the isogenic Δtar mutant CV4 show no response to the aspartate gradient, but they do seem to be repelled by higher NiSO4 concentrations, although significantly less than CV1 cells.
The responses of CV1 nik mutant cells are shown in Fig. 1D to F. All three nik mutants show a net migration toward lower NiSO4 concentrations comparable to that of strain CV1. Similar results were obtained when the flow rate was lower, but under those conditions the cells were exposed to the gradient for a longer time before imaging, and the average distance migrated was greater (data not shown). Pseudocolored images of the distribution of cells in NiSO4 gradients are shown in Fig. Fig.2.2. These photographs capture one instantaneous distribution, whereas the distributions shown in Fig. Fig.11 are averaged over many images.
The extent of migration in response to the NiSO4 gradient was quantified based on the chemotaxis partition and migration coefficients (CPC and CMC, respectively) (9, 17). The CPC value reflects the direction of migration (i.e., toward or away from a gradient) and quantifies the number of bacteria on either side of the bacterial inlet. For example, a CPC value of −0.30 indicates that 30% more bacteria move to the lower-concentration side than the higher-concentration side. The CMC weights the migration of cells by the distance they move. For example, a cell that moves to the left to the farthest low-concentration position (channel 1) is given a weighting factor of −1, whereas one that moves halfway into the lower concentration side (channel 16) is given a weighting factor of −0.5. CMC values are larger at lower flow rates.
The CPC values for CV1 and all three nik mutants (Table (Table1)1) were similar (−0.21 to −0.39). The CPC value for CV4 cells was −0.08. Cells in a null gradient of NiSO4 (a uniform 122.5 μM across the channel width) showed a slight bias to the right (CPC of 0.04). Such small CPC values are probably not significant because of the difficulty in accurately estimating cell number near the point where they enter the chemotaxis channel (i.e., where the cell density is maximal). The CMC values were also comparable for the wild type and nik mutants (−0.11 to −0.16) and significantly higher than for the CV4 tar mutant in the same gradient (CMC of −0.06). The CMC value for CV1 cells in the null gradient was 0.05. These results show that repellent taxis in response to NiSO4, even in this relatively shallow gradient, does not require NikA, NikB, or NikC. It should be noted that NiSO4 at concentrations of up to 300 μM does not significantly inhibit growth or motility in tryptone broth (9).
Our results clearly show that Ni2+ taxis can occur in the absence of the Nik proteins. The marginal response of CV4 cells to the Ni2+ gradient raises the possibility that Ni2+ is sensed by chemoreceptors other than Tar. The response is so weak that it could have been missed in previous, less-sensitive assays. To test whether the other high-abundance chemoreceptor of E. coli, Tsr, is responsible for the residual NiSO4 taxis, we assayed the responses of strains CV12 (Δtar Δtap trg::Tn10; Tsr as sole chemoreceptor) and CV13 (Δtsr Δtap trg::Tn10; Tsr as sole chemoreceptor) to a 0 to 122.5 μM NiSO4 gradient. Strain CV12 failed to show any significant response, whereas strain CV13 gave a very robust response (Fig. (Fig.33).
An isothermal calorimetric (ITC) analysis shows unequivocally that Ni2+ binds specifically to the isolated periplasmic domain of Tar but not to the isolated periplasmic domain of Tsr (I. Kawagishi, personal communication). That conclusion is consistent with our observation that Ni2+ uptake is not required for repellent taxis in response to Ni2+. The loss of Ni2+ sensing by Tar should give a sufficient difference in behavior to allow for an enrichment, using a variation of our recently developed microfluidic device (9), for tar mutants that are Ni2+ blind but still competent for maltose and/or aspartate taxis (10). In this way, we hope to characterize the Ni2+-binding site in detail and shed more light on the poorly understood mechanism of repellent taxis.
We thank Ikuro Kawagishi for communicating results prior to publication. We are grateful for the Keio strains provided by the Genome Analysis Project and National BioResource Project (NIG, Japan): E. coli.
This work was supported by a grant from the National Science Foundation (CBET 0846453) to A.J. and by the Bartoszek Fund for Basic Biological Science to M.D.M.
Published ahead of print on 12 March 2010.