Research ArticleMicrobiology

A cyclic di-GMP–binding adaptor protein interacts with a chemotaxis methyltransferase to control flagellar motor switching

See allHide authors and affiliations

Science Signaling  18 Oct 2016:
Vol. 9, Issue 450, pp. ra102
DOI: 10.1126/scisignal.aaf7584

Directing the movement of a pathogen

The opportunistic pathogen Pseudomonas aeruginosa is particularly resistant to antibiotic treatment when it forms biofilms on surfaces, such as in lungs or on medical equipment. The formation of P. aeruginosa biofilms requires both random and directed migration, and the bacterial messenger cyclic diguanylate monophosphate (c-di-GMP) is critical for all of these processes. Xu et al. (see also Orr and Lee) found that binding of c-di-GMP to the adaptor protein MapZ suppressed frequent changes in direction (which enables constant adjustment to changing conditions), attachment to surfaces, and biofilm formation by P. aeruginosa. These results suggest that the MapZ-associated chemotaxis pathway could be targeted to prevent the chronic and hard-to-treat infections caused by P. aeruginosa.


The bacterial messenger cyclic diguanylate monophosphate (c-di-GMP) binds to various effectors, the most common of which are single-domain PilZ proteins. These c-di-GMP effectors control various cellular functions and multicellular behaviors at the transcriptional or posttranslational level. We found that MapZ (methyltransferase-associated PilZ; formerly known as PA4608), a single-domain PilZ protein from the opportunistic pathogen Pseudomonas aeruginosa, directly interacted with the methyltransferase CheR1 and that this interaction was enhanced by c-di-GMP. In vitro assays indicated that, in the presence of c-di-GMP, MapZ inhibited CheR1 from methylating the chemoreceptor PctA, which would be expected to increase its affinity for chemoattractants and promote chemotaxis. MapZ localized to the poles of P. aeruginosa cells, where the flagellar motor and other chemotactic proteins, including PctA and CheR1, are also located. P. aeruginosa cells exhibit a random walk behavior by frequently switching the direction of flagellar rotation in a uniform solution. We showed that binding of c-di-GMP to MapZ decreased the frequency of flagellar motor switching and that MapZ was essential for generating the heterogeneous motility typical of P. aeruginosa cell populations and for efficient surface attachment during biofilm formation. Collectively, the studies revealed that c-di-GMP affects flagellar motor output by regulating the methylation of chemoreceptors through a single-domain PilZ adaptor protein.


The intracellular messenger cyclic diguanylate monophosphate (c-di-GMP) mediates various aspects of bacterial physiology in many environmental and pathogenic bacteria. The cellular concentration of c-di-GMP oscillates with environmental changes through the opposing activities of diguanylate cyclases and phosphodiesterases (1, 2). Increased cellular c-di-GMP concentrations are considered to be crucial for the motile-to-sessile switch and formation of surface-attached biofilm (35). c-di-GMP affects bacterial motility and other bacterial behaviors by binding to diverse protein and riboswitch effectors (2, 6). PilZ domain proteins are likely the most prevalent c-di-GMP–binding proteins (710). The 90–130 residue–containing PilZ domains bind to c-di-GMP with a broad range of binding affinities, which presumably allows the cell to control different signaling pathways over a wide c-di-GMP concentration range (11, 12). More than 60% of the 4000 PilZ domain proteins (Pfam family: PF07238) encoded in publicly available bacterial genomes are single-domain PilZ proteins. Although several single-domain PilZ proteins to mediate different cellular functions and bacterial behaviors such as motility, virulence, and extracellular enzyme production (1316), the biological function and mechanism of action for most single-domain PilZ proteins remain elusive. Emerging evidence suggests that the single-domain PilZ proteins function as trans-acting c-di-GMP–binding adaptors to regulate the catalytic or binding properties of other proteins through protein-protein interaction. A single-domain PilZ protein from Xanthomonas campestris pv. campestris interacts with its protein partners in a c-di-GMP–independent manner (17), whereas the single-domain PilZ protein HapZ of Pseudomonas aeruginosa interacts with the histidine kinase SagS to regulate two-component signaling in a c-di-GMP–dependent manner (18). Given the large number of single-domain PilZ proteins encoded by bacterial genomes, elucidation of their cellular functions will substantially advance our understanding of the molecular mechanism of c-di-GMP signaling.

The Gram-negative bacterium P. aeruginosa is notorious for its ability to form multidrug-resistant biofilm on abiotic and biotic surfaces. P. aeruginosa has a highly complex c-di-GMP signaling network that plays a central regulatory role in the formation and dispersal of biofilm (19, 20). The c-di-GMP network of P. aeruginosa consists of more than 30 diguanylate cyclases and phosphodiesterases and a dozen c-di-GMP–binding effector proteins that include seven PilZ domain proteins. The c-di-GMP–binding PilZ proteins include the multidomain protein Alg44, the didomain proteins FlgZ and PA2989, and the four single-domain PilZ proteins HapZ, PA0012, PA4324, and PA4608. c-di-GMP regulates alginate production by binding to Alg44 (21), mediates swarming motility by binding to FlgZ (22), and regulates two-component signaling and gene expression by binding to HapZ (18). To further expand our understanding of c-di-GMP signaling in P. aeruginosa, we analyzed PA4608, which directly interacted with a chemotaxis methyltransferase to control flagellar motor output and bacterial motility. These data unveil a potentially widespread mechanism by which c-di-GMP regulates motility and chemotactic responses through a PilZ adaptor protein.


PA4608 interacts directly with the chemotaxis methyltransferase CheR1

We hypothesized that the single-domain PilZ protein PA4608 functions as a trans-acting adaptor protein by binding to the intended protein target of c-di-GMP. We performed bacterial two-hybrid screening using the BacterioMatch II Two-Hybrid system with PA4608 as the bait to probe for the prey protein encoded by the previously constructed P. aeruginosa PAO1 genomic DNA libraries (18). Most of the colonies from our two-hybrid screening contained DNA fragments from the gene PA3348, whereas the rest of the colonies were false positives that contain shifted open reading frames. Specific two-hybrid binding assay using the full-length PA3348 gene construct showed robust growth of colonies on selective medium (Fig. 1A), indicating that PA4608 interacts with the full-length PA3348. The gene PA3348 codes for CheR1, a chemotaxis methyltransferase that contains a receiver domain and a methyltransferase domain (Fig. 1B) (23). Here, we renamed PA4608 as MapZ, which stands for methyltransferase-associated PilZ protein.

Fig. 1 Identification of CheR1 as the protein partner of PA4608 (MapZ).

(A) Bacterial two-hybrid assay suggests direct interaction between PA4608 and two CheR1 constructs (ΔN12 denotes the construct that lacks the N-terminal 12 residues). LGF2-GallP pair is the positive control. (B) Domain organization of the chemotaxis methyltransferase CheR1. (C) Co-IP of MapZ with CheR1. The cell lysate obtained from PAO1 cells that harbor either pUCP-6×His-CheR1 (lane 1) or pUCP-HA-MapZ/6×His-CheR1 (lane 2) was immunoprecipitated with antibody specific for HA tag protein [the same batch of total cell lysate was also used for co-IP in (D)]. The immunoprecipitated proteins and the total cell lysate used for co-IP were assessed by Western blotting (WB) using antibodies specific for 6×His and HA tags. (D) Co-IP assays were performed to show the effect of c-di-GMP, but not GMP, cGMP, and c-di-AMP, on the CheR1-MapZ interaction. Blots in (C) and (D) are representative of two independent experiments.

c-di-GMP enhances the interaction between CheR1 and MapZ

To validate the CheR1-MapZ interaction and to determine how c-di-GMP affects protein-protein interaction, we performed coimmunoprecipitation (co-IP) assays using P. aeruginosa cell lysate. To facilitate protein detection, we constructed plasmids to express N-terminal 6×His-tagged CheR1 and hemagglutinin (HA peptide YPYDVPDYA)–tagged MapZ, respectively. We found that 6×His-CheR1 coimmunoprecipitated with HA-MapZ that was enhanced at high concentrations of c-di-GMP, but not of GMP, cGMP, or cyclic diadenosine monophosphate (c-di-AMP) (Fig. 1, C and D). c-di-GMP enhanced the interaction in a dose-dependent manner up to 40 to 50 μM c-di-GMP, which is consistent with the binding affinity of MapZ for c-di-GMP, as measured by isothermal titration calorimetry (ITC) (KD = 8.8 ± 1.2 μM) (fig. S1) and fluorescence resonance energy transfer titration (KD = 6.1 ± 0.5 μM) (24).

MapZ inhibits CheR1-catalyzed methylation of the chemoreceptor PctA in the presence of c-di-GMP

The methyltransferase CheR1 methylates the chemoreceptor PctA, which is localized to cell poles and responds to amino acids (23, 25). The CheR1-MapZ interaction suggested that MapZ could be involved in the regulation of PctA-associated chemotaxis by modulating the methylation of PctA. We performed in vitro methylation assays to investigate whether MapZ and c-di-GMP affect the CheR1-catalyzed methylation of PctA. The membrane fraction that contains PctA was prepared and enriched using ultracentrifugation by following established protocols (23). When purified recombinant CheR1 and PctA-containing membrane fraction were incubated in the presence of the methyl donor [3H]Ado-Met (or [3H]SAM), methylation of PctA was observed readily (Fig. 2A, lane 2). Although the methylation of PctA did not seem to be affected by the presence of MapZ, addition of c-di-GMP to the reaction mixture decreased the methylation of PctA in a dose-dependent manner (Fig. 2A, lanes 3 to 8), with the methylation decreasing to below the detection limit at a c-di-GMP concentration of 10 μM. Meanwhile, GMP, cGMP, and c-di-AMP did not seem to affect PctA methylation (Fig. 2B, lanes 5 to 7), suggesting that the effect of c-di-GMP was specific. When MapZ_R13A, a mutant that is incapable of binding c-di-GMP, was used in the assay (Fig. 2C and fig. S1), addition of c-di-GMP did not affect PctA methylation (Fig. 2D). These observations suggest that the binding of c-di-GMP is essential for the inhibition of CheR1 by MapZ and that c-di-GMP likely regulates PctA-associated chemotactic pathways by suppressing the methylation of PctA.

Fig. 2 Effect of MapZ and c-di-GMP on the CheR1-catalyzed methylation of the chemoreceptor PctA.

(A) Effect of c-di-GMP on the methylation of PctA demonstrated by in vitro methyltransferase assay. The membrane fraction containing PctA and [3H]Ado-Met were used as the substrate and cosubstrates of CheR1, respectively. The dose-dependent effect of c-di-GMP on the methylation of PctA was shown using a series of c-di-GMP concentrations. Data are representative of two independent experiments. (B) Effect of c-di-GMP and other nucleotides on the methylation of PctA. Data are representative of two independent experiments. (C) Structure of MapZ in complex with dimeric c-di-GMP (Protein Data Bank code: 2L74). The conserved residue Arg13 for c-di-GMP binding is shown in sticks, and the N-terminal loop is shown in magenta. (D) Effect of c-di-GMP on the methylation of PctA in the presence of MapZ_R13A mutant. Data are representative of two independent experiments.

MapZ is localized to cell poles, and polar localization depends on the presence of c-di-GMP, CheR1, and PctA

As part of the chemoreceptor array, PctA and many other chemotactic signaling proteins are localized to the poles of P. aeruginosa cells (26). If MapZ interacts with CheR1 in vivo, it may also be sequestered to cell poles. To determine the cellular localization of MapZ, we constructed a plasmid to express yellow fluorescence protein (YFP)–tagged MapZ for fluorescence microscopy imaging. The plasmid was transformed into P. aeruginosa PA14 cells but not P. aeruginosa PAO1 cells because of concerns about the stability of the plasmid in PAO1 cells (27). Three-dimensional structured illumination microscopy (3D-SIM) (28), a superresolution microscopic method with increased lateral and axial resolution, revealed that the YFP-MapZ fusion protein displayed a bipolar distribution in PA14 cells (Fig. 3A). In contrast, the YFP-MapZ fusion protein seemed to distribute along the cell membrane in PA14_ΔcheR1 cells and was dispersed in the cytoplasm in PA14_ΔpctA cells. We also found that the YFP-MapZ_R13A fusion protein was no longer localized to the poles of PA14 cells (Fig. 3B), which suggested that the binding of c-di-GMP was important for the sequestration of MapZ at the cell poles.

Fig. 3 Subcellular localization of MapZ in wild-type and various mutant cells.

(A) Polar localization of the YFP-MapZ fusion protein in P. aeruginosa PA14 and two mutant cells as demonstrated by superresolution 3D-SIM fluorescence microscopy (Scale bars, 1 μm). Data are representative of two independent experiments. (B) Subcellular localization of YFP-MapZ_R13A fusion protein in PA14 cells as demonstrated by 3D-SIM fluorescence microscopy (Scale bar, 1 μm). Data are representative of two independent experiments. (C) Polar localization of HA-tagged MapZ in PA14 cells as demonstrated by immunofluorescence imaging. Data are representative of two independent experiments.

To ensure that the observed polar localization of the YFP-tagged MapZ was not an artifact caused by the large fluorescent protein tag, we made another construct that encoded an HA peptide–tagged MapZ protein. Immunofluorescence imaging showed that the HA-tagged MapZ also had a polar distribution (Fig. 3C). In comparison, the HA-tagged MapZ was dispersed in the PA14_ΔcheR1 cells with no fluorescent foci observed at the cell poles. These observations suggest that PctA, CheR, and c-di-GMP are required for polar localization of MapZ, which further supports the in vivo CheR1-MapZ interaction. The results also indicate that MapZ is likely to be colocalized with other chemotactic proteins as part of the chemotactic signaling machinery.

MapZ decreases the frequency of swimming direction reversal

Because MapZ was implicated in chemotaxis based on the above data, we next examined whether MapZ affects chemotactic response in a chemoattractant concentration gradient. Agarose plug–based chemotaxis assays with l-serine, a chemoattractant sensed by PctA, showed that PAO1 and ΔmapZ cells gathered around the serine-soaked agarose plug to form clearly visible zones of cells (fig. S2). In contrast, the chemotactic zone formed by the ΔcheR mutant was hardly visible because of a very low cell density, indicating that most of the ΔcheR cells had lost the ability to respond to the chemoattractant. Likewise, most of the mapZ+ cells (PAO1 cells transformed with the pUCP_mapZ overexpression plasmid) were nonchemotactic because the cell density in the chemotactic zone was also low. We also noticed that mapZ+ cells had a tendency to aggregate and form clumps in the assay, which was consistent with the observation that mapZ+ cells tended to adhere to each other to form insoluble aggregates when cultivated in liquid medium.

P. aeruginosa cells navigate stochastically by alternating periods of forward runs, backward runs, and turns (29). Such stochastic patterns can be masked in the agarose plug chemotaxis assays because of the heterogeneity of a cell population. To further probe the role played by MapZ in bacterial motility, we examined the swimming trajectory of individual cells using a video-recording microscope. Inspection of the swimming trajectories of PAO1 cells revealed that most cells made runs that end in a sharp turn or reversal, with each reversal followed by another forward or backward run in a different direction (Fig. 4A). Reversal is often followed by a change of swimming direction with a turn with acute turn angle. This “run-reverse-turn” pattern agrees with the previously reported swimming behavior of Pseudomonas species, with the reversal caused by a change in the direction of flagellar rotation and turns associated with diffusional cell realignment (Fig. 4B) (23, 30). In contrast to PAO1 cells, ΔcheR1 cells only swam in a single direction and lacked the ability to reverse. Most of the mapZ+ cells also swam in a single direction and resembled the ΔcheR1 cells. The swimming trajectories of ΔmapZ and mapZ_R13A cells were heterogeneous, with some cells exhibiting similar swimming behavior to PAO1 cells with linear runs and reversals, but with others lacking the ability to reverse. In contrast to the differences in direction switching, the mutants did not exhibit significant differences in swimming velocities (Fig. 4C). Comparison of the persistence of runs, which correlates inversely with the frequency of direction reversal, showed that mapZ+ and ΔcheR1 cells had longer runs than PAO1, ΔmapZ, and mapZ_R13A cells (Fig. 4D). The distinct trajectories exhibited by the PAO1 and mutant strains suggest that MapZ, together with c-di-GMP, can affect swimming by decreasing the frequency of direction reversals. Because frequent changes in swimming direction are a hallmark of cells that undergo chemotaxis-dependent random walk (31), the observations suggest that MapZ is an important motility factor that contributes to such basic bacterial behaviors as random walk.

Fig. 4 Effect of MapZ on the frequency of direction reversal in swimming.

(A) Swimming trajectories of PAO1, ΔcheR1, ΔmapZ, mapZ_R13A, and mapZ+ cells showing the lack of direction reversing for the ΔcheR1 and mapZ+ cells. Data are representative of two independent experiments. (B) Illustration of the run-reverse-turn swimming mechanism used by P. aeruginosa cells. (C) Average swimming speed of the wild-type (WT) (PAO1) and mutant strains. Difference between average swimming speed of the WT and each of the mutant strains is not statistically significant (two-tailed t test, P > 0.05 in all cases). Data are means ± SD (n = 30 to 36 cells) from two independent experiments. There was no significant difference in the average speed observed from the amalgamated data (two-tailed t test). (D) Directional swimming persistence of the WT and four mutant strains. Data are means ± SD (n = 30 to 36 cells) from two independent experiments (two-tailed t test, *P < 0.05).

MapZ controls flagellar output by suppressing motor switching

Because monotrichous P. aeruginosa cells reverse the swimming direction by switching the direction of flagellar rotation, we next determined whether MapZ controls flagellar motor switching. To directly observe flagellar rotation, we used the cell tethering assay (29) in which cells are attached to a coverslip using a flagellar protein–specific antibody (Fig. 5A). Cell body rotation was recorded with a camera to measure the speed of rotation and the frequency of switching between clockwise (CW) rotation, counterclockwise (CCW) rotation, and stationary or pausing phases (29, 32). We found that the average rotation speed of PAO1 and the four mutant cell lines did not differ significantly (Fig. 5B), suggesting that MapZ did not affect the speed of flagellar rotation. These results are consistent with the comparable swimming velocities observed for the strains (Fig. 4C). All five strains paused intermittently, and the duration of the pausing phase (less than 1 s) also did not show significant differences (Fig. 5C). Because the duration of the pause determines the turning angle of P. aeruginosa cells in diffusion-driven rotation (29), this observation suggested that MapZ does not affect swimming motility by influencing the diffusional rotation–based direction change. However, ΔcheR1 and mapZ+ cells exhibited considerable heterogeneity in the duration of CW or CCW rotation (see the large error bars in Fig. 5C), with some cells showing CCW and CW duration that is two to four times longer than that of PAO1 cells. Further analysis of the switching frequencies by comparing the amount of time the cells underwent CW and CCW rotation revealed notable differences among the strains. All PAO1 cells switched direction frequently and spent about equal amount of time undergoing CW and CCW rotation (Fig. 5D and movie S1). In contrast, ΔcheR1 cells underwent either CW (42%) or CCW (58%) rotation without switching rotation direction (movies S2 and S3). This finding was consistent with the unidirectional swimming trajectory of the ΔcheR1 cells we (Fig. 4A) and others (23) observed. ΔmapZ cells exhibited heterogeneity in rotational switching, with 18% of the cells rotating in a single direction and 82% of the cells switching direction stochastically (movies S4 to S6). The mapZ+ cells also exhibited some heterogeneity, but unlike ΔmapZ cells, most of the mapZ+ cells (62%) resembled ΔcheR1 cells and rotated in one direction, and 38% of the cells switched direction occasionally (movies S7 to S9). The mapZ_R13A mutant, which was generated by introducing a single mutation (R13A) into the mapZ gene in the PAO1 chromosome, also showed heterogeneity in direction switching (Fig. 5D and movies S10 to S12). These observations suggest that MapZ decreases flagellar switching frequency at high c-di-GMP concentrations and that changes in the c-di-GMP concentration can induce heterogeneous flagellar motor switching for P. aeruginosa.

Fig. 5 Analysis of the role of MapZ in flagellar motor switching by cell tethering assay.

(A) Schematic illustration of the tethering of bacterial cells to allow monitoring the rotation of cell body. (B) Average rotation speed from the WT (PAO1) and mutant cells as measured by cell tethering assay. Differences between average rotation speed of the WT and each of the mutant cells are not statistically significant (two-tailed t test, P > 0.05 in all cases). Data (means ± SD, n = 19 to 33 cells) are representative of two independent experiments. There was no significant difference in the average speed from the amalgamated data (t test). (C) Duration of different rotation modes (CW, CCW, or pause) for the WT and mutant cells. Differences between the pause duration of the WT and each of the mutant cells are not statistically significant (two-tailed t test, P > 0.05 in all cases). The differences between the CCW duration of the WT and the ΔcheR1 and mapZ+ mutants were statistically significant (*P < 0.05). Data (means ± SD, n = 19 to 33 cells) are representative of two independent experiments. There was no significant difference in the average duration observed from the amalgamated data as a result of the large error bars (two-tailed t test). (D) Bar graphs showing the percentage time that the WT and mutants cells underwent CW rotation. The percentage of cells exhibiting bidirectional rotation (CW and CCW) and unidirectional rotation (CCW or CW) is summarized below the bar graphs. The corresponding video for the rotating cells can be found in movies S1 to S12. (E) Analysis of the frequency of flagellar motor switching during chemotactic response to l-serine. Bar graphs show the change of direction switching frequency with time after exposure to l-serine. Sample sizes are 50, 30, 32, and 33 cells for PAO1, ΔmapZ, mapZ+, and mapZ_R13A, respectively. Data are from two independent experiments.

To find out how MapZ affects flagellar switching in a chemoattractant gradient, we performed time-dependent cell tethering assays in a concentration gradient of l-serine. We found that the addition of l-serine to PAO1 cells triggered a rotational bias toward CW or CCW rotation in the first 20 s (Fig. 5E, first row). Such rotational bias agrees with the longer and persistent runs needed to deviate from a random walk behavior. Within 60 s, PAO1 cells showed heterogeneity with different degrees of CW or CCW rotational bias. Unbiased rotation was restored by the end of 60 s with the dissipation of the attractant. A large population of the ΔmapZ and mapZ_R13A cells already exhibited rotational bias before the addition of l-serine, and the addition induced even greater rotational bias in the first 20 s, with about one-third of the cells undergoing single directional rotation only. In contrast, the serine gradient had little effect on the rotation switching of mapZ+ cells because of the already strong preexisting rotational bias before the addition of serine. Together, the data suggest that c-di-GMP and MapZ likely contribute to chemotaxis and bacterial motility by controlling the frequency of flagellar motor switching.

MapZ promotes irreversible surface attachment and biofilm formation

Bacterial motility and chemotaxis are indispensible for P. aeruginosa biofilm formation (23, 33, 34). CheR1 plays a role in biofilm formation by affecting the ability of P. aeruginosa cells to survey and attach to surfaces (23). These findings indicate that methylation-dependent chemotaxis is crucial for the formation of P. aeruginosa biofilm on surfaces. To find out whether MapZ also plays a role in biofilm formation, we monitored the surface attachment and biofilm formation for PAO1 and mutant cells in flow cells (35). PAO1 cells formed typical mushroom-shaped biofilm microcolonies after the 5-day growth, whereas the ΔmapZ mutant formed a fragile and thin biofilm with reduced biomass (Fig. 6, A and B). The ΔmapZ strain was also less efficient at the irreversible surface attachment stage, as suggested by lower surface coverage at the end of day 3 (Fig. 6). In contrast, the mapZ+ strain exhibited significantly increased surface attachment at day 3 to yield a much thicker biofilm and more biomass after 5 days. Neither deletion nor overexpression of mapZ seemed to noticeably affect the formation of the mushroom structures, which suggested that MapZ-related pathways affect biofilm formation mainly by influencing the irreversible surface attachment stage but not the maturation stage.

Fig. 6 Analysis of the role of MapZ in surface attachment and early stage of biofilm formation.

(A) Images of the WT and mutant biofilm were obtained by growing the biofilm in flow cells and monitored with a confocal microscope. Confocal microcopy images show differences in surface attachment and thickness of the biofilm at the end of days 3 and 5 for the two mutants (Scale bars, 20 μm). Data are representative of two independent experiments. (B) Quantification of the biovolume or biomass using IMARIS 8.2.0. For each strain, means ± SD were calculated from the amalgamated data obtained from four biofilm-growing chambers (two-tailed t test, *P < 0.05; **P < 0.001).


Unveiling the function of PilZ domain effectors is essential for understanding the role of c-di-GMP in P. aeruginosa and many other bacterial species. Our results revealed a role for the P. aeruginosa PilZ domain protein MapZ in motility control through its interaction with the chemotaxis methyltransferase CheR1. The dependence of CheR1-MapZ interaction on c-di-GMP suggested that c-di-GMP controlled the motility of P. aeruginosa by modulating the methylation of the chemoreceptor. Our results further unveiled that c-di-GMP and MapZ altered the motility of P. aeruginosa by controlling the switching frequency, but not the speed, of flagellar rotation. This mechanism is distinct from other c-di-GMP–dependent regulatory mechanisms involved in controlling bacterial motility. Moreover, our results showed that the inhibition of CheR1 by MapZ and c-di-GMP correlated with enhanced surface attachment, which suggested a mechanism by which c-di-GMP influenced the early stage of biofilm formation.

Although a common bacterial behavior regulated by c-di-GMP is flagellum-dependent motility, how c-di-GMP controls flagellar output in P. aeruginosa remains to be fully understood. The best-characterized mechanism involves FleQ, a transcriptional regulator that regulates the expression of genes encoding flagellar biosynthetic factors; binding of c-di-GMP to FleQ suppresses gene expression and decreases motility (36, 37). c-di-GMP also regulates flagellar rotation and swarming motility through the binding effector FlgZ, with the binding of c-di-GMP to FlgZ promoting the interaction between FlgZ and the stator protein MotC to reduce motility (22). Apart from P. aeruginosa, c-di-GMP controls flagellar function in several other bacterial species. The best-known mechanism is the YcgR-dependent “backstop brake” mechanism in Escherichia coli, whereby the binding of c-di-GMP to YcgR promotes the binding of YcgR to flagellar switching complex to reduce swimming motility (3840). A similar mechanism operates in Caulobacter crescentus, with increased concentrations of c-di-GMP promoting the interaction between the c-di-GMP–binding DgrA and the motor protein FliL to impede motility (9). In addition to the mechanism that involves the use of a motor brake, c-di-GMP also regulates flagellum-dependent motility through chemotactic pathways. In Azospirillum brasilense, c-di-GMP binds directly to the chemoreceptor Tlp1 through the C-terminal PilZ domain to influence motility (41). In Borrelia burgdorferi, c-di-GMP binds to the PilZ protein PlzA to putatively regulate the flagellum-dependent flexing motility (42) by interacting with a chemotactic protein. The MapZ-dependent mechanism reported here, as featured by the involvement of a trans-acting PilZ-binding adaptor and the regulation of the methylation of chemoreceptors, is distinct from these mechanisms.

The chemotactic machinery of P. aeruginosa has more protein components and is likely more complex than the E. coli chemotactic system. There are four chemotaxis gene clusters and a total of 26 chemoreceptor genes in the genome of P. aeruginosa PAO1. Two of the four gene clusters (Che and Che2) are involved in the regulation of chemotaxis and flagellum-dependent motility (43, 44). The Che system that is essential for chemotaxis includes homologs of the E. coli chemotactic proteins CheA, CheB, CheW, CheY, and CheZ. Although pctA and cheR1 are not localized within the che gene cluster, PctA and CheR1 are believed to function as part of the Che chemotactic system (26). The Che2 system consists of two chemoreceptors (McpA and McpB) and several downstream signaling proteins and plays a role in chemotaxis only when cells enter stationary growth (26, 45). For E. coli, a cell moves up or down a concentration gradient by resetting the CheA autokinase activity to prestimulus amounts after each linear run (46). Such an “adaptation mechanism” relies on the reversible methylation of chemoreceptors catalyzed by the methyltransferase CheR and methylesterase CheB. Methylation decreases the affinity of the chemoreceptor for attractants and stimulates the autokinase activity of CheA, resulting in the phosphorylation of CheY. In E. coli, the methylesterase CheB is regulated by CheA through phosphorylation, and the methyltransferase CheR is believed to be constitutively active and its activity does not appear to be regulated. In P. aeruginosa and some other Pseudomonas species, CheR1 is essential for flagellum-mediated chemotaxis because of its crucial role in methylating chemoreceptors (23, 47, 48). Our data reveal an additional layer of regulation for this system by suggesting that the enzymatic activity of CheR1, and thus the methylation status of its cognate chemoreceptor, can be regulated by cellular c-di-GMP concentrations. A relatively simple chemotactic system, such as the one in E. coli, may be sufficient for the cells to navigate in chemoattractant gradients. It is conceivable that a bacterium with more complex lifestyles such as P. aeruginosa could benefit from having a c-di-GMP regulatory circuit integrated into the chemotactic system to fine-tune flagellar output.

Unlike E. coli cells that swim in a solution using a “run-tumble-run” mechanism (49), monotrichous Pseudomonas cells follow a run-reverse-turn mechanism (29). P. aeruginosa cells move forward (run) or backward (reversal) when the flagellum undergoes CCW or CW rotation. In addition to reversal, the cells can also change swimming direction when the flagellum pauses to allow the cell body to undergo diffusional reorientation (turn). The frequency of flagellar motor switching determines the duration or persistence of the runs. Suppression of motor switching generates more persistent runs to shift the cells from random walk mode to directional swimming mode. Our cell tethering assays showed that the flagellar motor of PAO1 had a high switching frequency in the absence of a chemoattractant and suggested that MapZ abundance and c-di-GMP concentration could substantially affect the switching frequency. Our data support a model whereby c-di-GMP promotes the formation of the CheR1-MapZ complex near the chemoreceptor array at the pole (Fig. 7A). When triggered by environmental cues, an increase in c-di-GMP concentration promotes the formation of the CheR1-MapZ complex to inhibit the methyltransferase activity of CheR1. Inhibition of CheR1 reduces the methylation of PctA to suppress the autokinase activity of CheA. The immediate effect of suppressing the autokinase activity of CheA is unclear because the signaling proteins downstream of CheA are not yet known. In E. coli, low CheA activity results in an increase in the number of unphosphorylated CheY molecules, and the dissociation of unphosphorylated CheY from the flagellar motor induces a CW-to-CCW motor switching. In P. aeruginosa, the role of CheY in motor switching is likely to be different because the knockout of cheY does not prevent the flagellum from rotating in both directions (29). We surmise that the decreased autokinase activity of CheA at high c-di-GMP concentration may lead to the dephosphorylation of another protein (protein X in Fig. 7A). We further postulate that binding of phosphorylated X to flagellar motor proteins is essential for motor switching and that dissociation of dephosphorylated X from flagellar motor decreases motor switching. A lower frequency of flagellar motor switching causes persistent unidirectional rotation and directional swimming. This model implies that the maintenance of a relatively low c-di-GMP concentration near the chemoreceptor array is important for the random walk behavior of P. aeruginosa cells and that increasing the c-di-GMP concentration is crucial for effective chemotactic responses.

Fig. 7 Model for the regulation of flagellar motor switching by c-di-GMP and MapZ.

(A) In cells with low concentrations of c-di-GMP, CheR1 methylates the methyl-accepting chemotaxis protein (MCP) to stimulate the autokinase activity of CheA and increase the phosphorylated population of X. Binding of phosphorylated X to the flagellar motor enables frequent switching between CW and CCW rotation. In cells with high concentrations of c-di-GMP, MapZ blocks the methylation of PctA by CheR1 to lower the autokinase activity of CheA. This decreases the population of phosphorylated X and impedes flagellar motor switching. (P, phosphoryl group; M, methyl group; SAM, S-adenosylmethionine or Ado-Met) (B) Time courses of chemotactic responses for cells with low and high c-di-GMP concentration. Addition of chemoattractant results in a fast decrease in CheA kinase activity, which is followed by a slow CheR1-dependent adaptation mechanism. The heterogeneity in cellular c-di-GMP concentrations gives rise to a range of chemotactic responses.

Heterogeneity in cellular c-di-GMP concentration has been proposed to be required for generating the diversity in motility observed in P. aeruginosa cell populations (27). The diversity in bacterial motility may be crucial for the success and survival of P. aeruginosa because it allows cell populations to survey the surroundings more efficiently. The c-di-GMP effector responsible for generating such diversity or heterogeneity in bacterial motility is unknown. Our results indicate that the c-di-GMP–binding MapZ could be crucial for generating the heterogeneity in bacterial motility, given the considerable heterogeneity in swimming trajectory and flagellar direction switching exhibited by MapZ-deficient cells. As illustrated by the two hypothetical chemotactic response curves, c-di-GMP concentration influences MCP methylation and adaptation time, that is, the time required to restore the CheA autokinase activity to prestimulus amounts (Fig. 7B). As a result, heterogeneity in cellular c-di-GMP concentration can elicit a wide range of chemotactic responses, with cells with high c-di-GMP concentration exhibiting less frequent direction switching and more persistent runs than those with lower c-di-GMP concentration. Because c-di-GMP networks usually include many sensory domain–containing diguanylate cyclases and phosphodiesterases (50), incorporation of a c-di-GMP–binding effector in the chemotactic machinery would allow cellular response to additional environmental or metabolic inputs for the fine-tuning of flagellar output and generation of heterogeneity. Given the advantage of having such a c-di-GMP–dependent circuit embedded in the chemotactic system to fine-tune flagellar output, a similar mechanism may also be used by other non-Pseudomonas species to modulate chemoreceptor methylation and bacterial motility.


Bacterial strains, plasmids, and growth conditions

The bacterial strains and plasmids used in this study are described in table S1. The DNA fragments encoding full-length or truncated proteins were amplified by polymerase chain reaction (PCR) from P. aeruginosa PAO1 genomic DNA and by using primers designed based on the published P. aeruginosa PAO1 genome sequence (51). The primers used for PCR and molecular cloning are summarized in table S2. The MapZ_R13A was generated using the QuikChange Site-Directed Mutagenesis Kit (Agilent Technologies). Some of the P. aeruginosa PAO1 mutant strains used in this study were obtained from the P. aeruginosa PAO1 transposon mutant library (52).

Preparation of the P. aeruginosa mapZ_R13A mutant

The mutant strain was created by using the allelic exchange vector pK18GT-mapZ_R13A (53). Briefly, a DNA fragment of 1216 base pairs containing the mapZ_R13A gene construct was amplified with P. aeruginosa PAO1 genomic DNA as template by overlap extension PCR. The primers used are listed in table S2, with primers 1 and 2 used for upstream fragment amplification, primers 3 and 4 used for downstream amplification, and primers 1 and 4 used for full-length gene amplification. The PCR product was cloned into the vector pK18GT at Bam HI and Hind III sites to generate the suicide plasmid pK18GT-mapZ_R13A. The suicide plasmid was transformed into E. coli DH5α for the subsequent conjugation with PAO1 by triparental mating with the helper plasmid pRK2013. The trans-conjugants were selected on Pseudomonas Isolation Agar (PIA) medium plates (QingDao Hope Biotechnology) containing gentamicin (100 μg/ml). Colonies resulting from the first crossover events were streaked onto LB agar plate supplemented with 10% sucrose, and sucrose-sensitive colonies were selected as positive colonies. Colonies that contained the correct mutation were identified by sequencing of the DNA fragment amplified using primers 5 and 6.

Bacterial two-hybrid screening

The construction of the genomic DNA libraries used for the two-hybrid assays has been described previously (18). The plasmid pBT-MapZ was used as the bait to probe the genomic DNA libraries using the BacterioMatch II Two-Hybrid system (Stratagene). The plasmids extracted from the cells harboring inserted DNA fragments were cotransformed into the provided E. coli reporter strain with the bait plasmid by electroporation. The cotransformed cells were grown on M9+ His-deficient medium containing 5 mM 3-amino-1,2,4-triazole (3-AT) for 48 to 72 hours at 30°C. Colonies that grew on these plates were selected as positive colonies. Positive colonies were picked and restreaked on M9+ His-deficient medium containing 5 mM 3-AT and streptomycin (12.5 μg/ml). The colonies that grew on the double-selection medium were cultured in liquid medium and used for colony PCR with pTRG forward and reverse primer pairs to amplify the insert from the library plasmid. The positive colonies were grown to prepare plasmids for sequencing. For the specific two-hybrid binding assay, the bait and prey plasmids were cotransformed, and the cotransformed mixtures were plated onto selective and nonselective media and incubated at 37°C overnight. Normal growth on the selective screening medium was considered as an indicator of positive protein-protein interaction.

Co-IP assay

The two plasmids, pUCP18-6×His-CheR1 and pUCP18-HA-MapZ/6×His-CheR1, were constructed and transformed into P. aeruginosa PAO1 strain. The strain harboring pUCP18-6×His-CheR1 expressed the protein 6×His-CheR1, and the stain harboring pUCP18-HA-MapZ/6×His-CheR1 produced both HA-MapZ and 6×His-CheR1. The strains were cultured in LB medium that contains carbenicillin (300 μg/ml) with shaking at 37°C overnight. The strains were subsequently subcultured to an optical density at 600 nm (OD600) of 1.0 at 37°C. Cells were harvested by centrifugation at 11,325g for 10 min and lysed by sonication with ice-cold cell lysis buffer [phosphate-buffered saline (PBS) (pH 7.4), 1% NP-40, 10% glycerol, and protease inhibitor cocktail (Roche)]. The total cell lysates were clarified by centrifugation at 75,600g for 1 hour, and the supernatant was filtered by 0.45-μm Minisart high flow syringe filters. Lysates were subsequently immunoprecipitated with the EZview Red Anti-HA Affinity Gel (Sigma-Aldrich). Each IP sample containing 1 ml of lysate and 20 μl of the prewashed beads with different concentrations of nucleotides (c-di-GMP, GMP, cGMP, or c-di-AMP) was incubated for 1 hour at 4°C. The beads were collected by centrifugation and washed six times with lysis buffer at 4°C. The immunoprecipitated proteins were diluted with 2× loading dye and subjected to SDS–polyacrylamide gel electrophoresis (PAGE) and Western blotting with THE His Tag Antibody [HRP] (GenScript) and HA-tag Antibody [HRP] (GenScript).

Expression and purification of recombinant proteins

The plasmids pET28b-MapZ, pET28b-MapZ_R13A, and pET28b-CheR1 were transformed into E. coli BL21(DE3) cell line and single colony of each on LB agar plates containing kanamycin (50 μg/ml). Cultures [20 ml of LB medium containing kanamycin (50 μg/ml)] were grown at 37°C with shaking for 16 hours. The overnight cultures were used to inoculate 2 liters of fresh LB medium supplemented with kanamycin (50 μg/ml). When the culture reached an OD600 of 0.6 to 0.8, the temperature was reduced from 37° to 16°C, and protein expression was induced by adding isopropyl-β-d-thiogalactopyranoside (0.3 mM for CheR1 and 0.4 mM for MapZ and MapZ_R13A). After 20 hours, cells were harvested by centrifugation, and pellets were stored at −80°C until the next step. Cells were resuspended in PBS buffer (pH 7.4) that contained 1 mM dithiothreitol and protease inhibitor (Roche) and were lysed by French press at 1200 bar. After centrifugation at 75,600g for 1 hour at 4°C, the supernatant was filtered by 0.45-μm filter and incubated with Ni-NTA agarose beads (Qiagen). After 1-hour incubation at 4°C with rotation, the beads were sequentially washed with 10 and 20 mM imidazole-containing PBS buffer (pH 7.4). The recombinant protein was eluted using PBS buffer that contained 400 mM imidazole. The fractions that contained the recombinant proteins were desalted using PD-10 Desalting Columns (GE Healthcare) and eluted using PBS buffer (pH 7.4). The proteins were concentrated by Amicon concentrators (5-kDa cutoff for MapZ and MapZ_R13A and 10-kDa cutoff for CheR1), and their concentrations were determined before the proteins were flash-frozen and stored at −80°C.

In vitro methyltransferase assay

Methylation of PctA by CheR1 was examined by performing in vitro methylation assays with 6×His-tagged recombinant proteins with [3H]Ado-Met. Membrane fractions containing PctA and c-di-GMP were prepared as described previously (23, 54). The reaction mixture consisting of 50 μl of membrane fractions (2 mg/ml), 0.1 μM 6×His-CheR1, 50 mM NaH2PO4 (pH 8.0), and 300 mM NaCl was preincubated at 30°C for 10 min. The reaction was initiated by adding 0.625 μM [3H]Ado-Met (specific activity, 15 Ci/mmol; PerkinElmer) and incubated at 30°C for 30 min. The reaction was stopped by adding 2× SDS-PAGE loading dye. Methylation of PctA was visualized by running 12% SDS-PAGE and autoradiography. The effect of MapZ (0.1 μM), c-di-GMP, and other nucleotides on PctA methylation was examined by incubating the protein and nucleotides in the reaction mixture accordingly.

Cellular localization by superresolution 3D-SIM

The YFP-MapZ or YFP-MapZ_R13A gene–containing pUCP18-based plasmids were constructed by PCR and standard molecular cloning technique. The plasmids were used to transform the PA14, ΔmapZ, ΔcheR1, and ΔpctA strains, which were cultured overnight. PA14 and mutant cells that expressed the YFP-MapZ or YFP-MapZ_R13A fusion protein were immobilized using 1% agarose and placed onto glass slides for microscopy imaging. Cell images were captured using the LSM 780 ELYRA PS.1 superresolution structured illumination system (Carl Zeiss) with a 100× plan-apochromatic oil immersion objective lens (numerical aperture, 1.46). YFP was excited using a 488-nm optically pumped semiconductor laser line for observation and data collection. The acquired images were further processed with the Zen 2011 software (Carl Zeiss).

Cellular localization by antibody-based immunofluorescence assay

The pUCP18-based plasmid that expressed the HA-tagged MapZ was constructed by PCR and standard molecular cloning technique. The plasmid was used to transform PA14 and ∆cheR1 strains. Cells expressing HA-tagged MapZ in the midexponential phase were subjected to immunofluorescence staining as described previously, with some modification (55). Briefly, cells were fixed with 4% (w/v) paraformaldehyde, permeabilized with 0.5% (v/v) Triton X-100 in PBS (PBST) solution, blocked with 1% BSA–containing PBST, and then incubated with a primary rabbit antibody that was specific for HA tag (1:1000; Abcam) and then with an antibody specific for rabbit IgG Alexa Fluor 488 (1:2000; Abcam). The stained cells were mounted on adhesion microscope slides (CITOGLAS REF188105), and images were taken using a Leica TCS SP2 confocal microscope.

Agarose plug chemotaxis assay

The assay was performed using a procedure similar to a previously published protocol (56). Cells were grown at 30°C in Murashige and Skoog (MS) medium [10 mM Na2HPO4, 20.6 mM KH2PO4, 25 mM NH4NO3, 0.8 mM MgSO4, and 1.0 ml of trace elements solution (63 mM FeCl3, 3.3 mM CoCl2, 4 mM CaCl2, 1.2 mM Na2MoO4, 1.6 mM H3BO3, 0.7 mM ZnSO4, 0.7 mM CuSO4, and 0.7 mM MnSO4)] containing 10 mM succinate until the cultures reached an OD600 of 0.8. Cells were harvested by centrifugation and washed twice and grown in fresh MS medium until the cultures reached an OD600 of 0.8. Four aliquots (100 μl) of melted low–melting point agarose [2%, (w/v); Sigma-Aldrich] containing 10 mM l-serine (Sigma-Aldrich) were transferred to the petri dish to form small agarose plugs. As negative controls, three agarose plugs containing MS medium were plated on the left side of the petri dish. After the solidification of the agarose, 10 ml of bacterial suspension was added around the agarose plugs, and photos were taken after 10 min to visualize the zone or ring formed by the bacterial cells around the agarose plugs.

Swimming trajectory analysis

Free-swimming P. aeruginosa cells with their intact flagellum were grown to exponential phase (OD600 of 1.0) in LB medium, diluted to 1:500 in 0.9% NaCl containing 3% Ficoll (Sigma-Aldrich) before being loaded in cell chambers with a channel height of 20 μm, and observed in the middle of the chamber depth (away from the coverslip). Cells were visualized under a 40× objective using an inverted microscope (Nikon TE2000-U), and videos of bacteria were taken at 25 frames per second (fps) using a complementary metal-oxide-semiconductor (CMOS) camera (Thorlabs DCC1645). Images with cell outlines were obtained using ImageJ (NIH), and bacterial swimming trajectories were obtained using particle segmentation and tracking algorithms, as described previously (57, 58), and implemented through the software TrackMate plugin in Fiji (59). Briefly, the cell outlines in each image of the time-lapsed video were segmented and identified based on the difference of Gaussians approach (57). The segmented cell outlines were then tracked by linking the individual cell outlines from frame to frame based on the linear assignment problem method (58). The velocities and directional persistence of individual swimming cells were then obtained from the trajectories from the time-lapsed videos.

The directional persistence of individual swimming cells was calculated as follows: the mean squared displacement of a tracked cell at any point in time with two-dimensional coordinates x and y was denoted as r2, where r2 = x2 + y2. The movement of each cell was described with r2α tƔ, where t was the time that has passed since the start of the cell tracking and Ɣ was a cell movement persistence factor. If Ɣ = 1, then r2α t, and the displacement of the cell was proportional to the square root of the elapsed time. In this case, the cell was said to exhibit a random swimming behavior. If Ɣ = 2, then r2α t2, where the displacement of the cell was proportional to the elapsed time, and the cell exhibited directed swimming persistence. Hence, particle movement persistence was quantified using Ɣ.

Flow-cell biofilm formation analysis

The mini-Tn7-eGFP-Gmr cassette was inserted into P. aeruginosa strains to tag them with green fluorescent protein. Biofilms were grown in flow chambers with individual channel dimensions of 1 mm × 4 mm × 40 mm. The flow chambers were inoculated by injecting 350 μl of overnight culture diluted to an OD600 of 0.01 into each flow channel using a small syringe. After inoculation, the flow channels were left without flow for 1 hour, after which point medium flow was started using a Watson-Marlow 205S peristaltic pump. The mean flow velocity in the flow chambers was 0.2 mm s−1. All microscopic observations and image acquisitions were done with a Zeiss LSM 780 confocal laser scanning microscope. Data were analyzed to generate the simulated 3D images using the IMARIS software package (Bitplane AG) (60). The same software was used to quantify the biomass or biovolume.

Cell tethering assay for the analysis of the flagellar rotation

Single colonies of P. aeruginosa PAO1 and ΔcheR1, ΔmapZ, mapZ+, and mapZ_R13A mutant strains were used to inoculate 10 ml of LB broth (BD Biosciences) that was kept at 37°C with shaking at 250 rpm overnight. Cultures were diluted to OD600 of 0.2 using M9 medium and grown at 37°C with shaking at 250 rpm until the cells reached the late exponential growth phase at OD600 of 0.8. Cell cultures were then diluted 10-fold before the microfluidic experiments. Cell tethering assays were carried out by using a microfluidic stagnation flow device precoated with flagellar antibodies (61). Flagella were sheared off by passing the bacterial cells through a 34-gauge blunt-end needle four times. Cells were loaded into the cell chamber of the flow device, and nontethered cells were washed away using M9 broth. Cells were visualized using an inverted microscope (Nikon TE2000-U) under 40× objective. Videos of tethered bacteria were taken at 40 fps for 1 to 5 min using a CMOS camera (Thorlabs DCC1645). l-Serine (10 mM in M9 medium) [0.5% (w/v)] was used as the chemoattractant in the analysis of flagellar rotation in chemical gradient. The M9 solution blank and l-serine solution were rapidly and completely gated to expose the cells to a temporally precise chemoattractant signal. The gating of the chemoattractant was performed through flow deflection between the chemoattractant and buffer inlets (61). Flow deflection ensured that cells tethered in the stagnation flow chamber were shielded from the l-serine before the chemotaxis experiments. The rotation of the cells was monitored and recorded after the reported procedure (62). The time spent by each cell in CW, CCW, or pause phases was tallied in 20-s intervals. Image processing of the tethered cells was carried out as previously described (63), with videos of the tethered cells converted to grayscale and the contrast of the cells adjusted (saturated pixels, 0.4; histogram stack) using ImageJ (NIH). Images of the cells were binarized to isolate them from the background for every frame in the video. The center of mass coordinates of the rotating cell bodies of each cell was measured from the obtained binarized image stack and imported into our in-house analysis program BTAP (bacterial tethering analysis program) in MATLAB.

Isothermal titration calorimetry

ITC measurements were performed at 22°C using MicroCal VP-ITC (MicroCal Inc.). Protein samples were dialyzed into a buffer containing 20 mM tris (pH8.0) and 100 mM NaCl. A sample syringe with stirring speed of 290 rpm was used to inject c-di-GMP (1.18 mM) into a cell containing 50 μM protein. The titration comprised 29 injections of 10 μl each, separated by 240-s equilibration time. The data sets were analyzed using the Origin 7.0 program, fitted to a single-site binding model.

Statistical analysis

The Student’s two-tailed t test was used to test the statistical significance of the data in Figs. 4 to 6. The normality assumption was considered to be sound because, by the central limit theorem, sample means of moderately large samples are well approximated by a normal distribution.


Fig. S1. Measurement of the c-di-GMP–binding affinity for MapZ and the MapZ_R13A mutant by ITC.

Fig. S2. P. aeruginosa PAO1 and mutant cells responding to l-serine in the agarose plug chemotaxis assay.

Table S1. Bacterial strains and plasmids used in this study.

Table S2. DNA primers used in this study for molecular cloning and sequencing.

Movie S1. PAO1 cell undergoing bidirectional rotation.

Movie S2. ΔcheR1 cell undergoing unidirectional CCW rotation.

Movie S3. ΔcheR1 cell undergoing unidirectional CW rotation.

Movie S4. ΔmapZ cell undergoing bidirectional rotation.

Movie S5. ΔmapZ cell undergoing unidirectional CCW rotation.

Movie S6. ΔmapZ cell undergoing unidirectional CW rotation.

Movie S7. mapZ+ cell undergoing bidirectional rotation.

Movie S8. mapZ+ cell undergoing unidirectional CCW rotation.

Movie S9. mapZ+ cell undergoing unidirectional CW rotation.

Movie S10. mapZ_R13A cell undergoing bidirectional rotation.

Movie S11. mapZ_R13A cell undergoing unidirectional CCW rotation.

Movie S12. mapZ_R13A cell undergoing unidirectional CW rotation.

References (6466)


Acknowledgments: We thank T. T. Selão of Nanyang Technological University for assisting us with the preparation of the PctA-enriched membrane fraction for the methyltransferase assay. We thank H. K. Lee for his help with our statistical analysis. We thank P. Greenberg for his valuable comments on our manuscript. Funding: This work was supported by a Tier II ARC grant (to Z.-X.L.) from the Ministry of Education of Singapore. This work was also supported by the National Basic Research Program of China (973 Program, grant no. 2015CB150600) and the Introduction of Innovative R&D Team Program of Guangdong Province (no. 2013S034). Some of the Pseudomonas mutant strains were obtained from the transposon mutant library (supported by NIH grant no. P30 DK089507). Author contributions: Z.-X.L. designed the study and wrote the manuscript with the assistance from other authors. L. Xu constructed the P. aeruginosa genomic DNA library and performed two-hybrid screening and immunofluorescence cellular localization experiments. L. Xu, L.-H.Z., and X.Y. made the mapZ_R13A chromosomal mutant. L. Xin performed co-IP assay, in vitro methyltransferase activity assay, and chemotaxis assays and contributed to cellular localization, cell tethering assay, and ITC studies. Y.Z. and K.-H.C. performed cell tethering assays and swimming trajectory analysis. J.K.H.Y., Y.D., and L.Y. performed biofilm assays and cellular localization microscopy studies by using YFP-tagged proteins. X.T. and L. Xin performed ITC. Q.W.C. and P.V. contributed to the motility assay and co-IP assay. Competing interests: The authors declare that they have no competing interests.
View Abstract

Stay Connected to Science Signaling

Navigate This Article