The Rcs System Contributes to the Motility Defects of the Twin-Arginine Translocation System Mutant of Extraintestinal Pathogenic Escherichia coli

ABSTRACT

Flagellum-mediated bacterial motility is important for bacteria to take up nutrients, adapt to environmental changes, and establish infection. The twin-arginine translocation system (Tat) is an important protein export system, playing a critical role in bacterial physiology and pathogenesis. It has been observed for a long time that the Tat system is critical for bacterial motility. However, the underlying mechanism remains unrevealed. In this study, a comparative transcriptomics analysis was performed with extraintestinal pathogenic Escherichia coli (ExPEC), which identified a considerable number of genes differentially expressed when the Tat system was disrupted. Among them, a large proportion of flagellar biosynthesis genes showed downregulation, indicating that transcription regulation plays an important role in mediating the motility defects. We further identified three Tat substrate proteins, MdoD, AmiA, and AmiC, that were responsible for the nonmotile phenotype. The Rcs system was deleted in the Δtat, the ΔmdoD, and the ΔamiAΔamiC strains, which restored the motility of ΔmdoD and partially restored the motility of Δtat and ΔamiAΔamiC. The flagella were also observed in all of the ΔtatΔrcsDB, ΔmdoDΔrcsDB, and ΔamiAΔamiCΔrcsDB strains, but not in the Δtat, ΔmdoD, and ΔamiAΔamiC strains, by using transmission electron microscopy. Quantitative reverse transcription-PCR data revealed that the regulons of the Rcs system displayed differential expression in the tat mutant, indicating that the Rcs signaling was activated. Our results suggest that the Rcs system plays an important role in mediating the motility defects of the tat mutant of ExPEC.

IMPORTANCE The Tat system is an important protein export system critical for bacterial physiology and pathogenesis. It has been observed for a long time that the Tat system is critical for bacterial motility. However, the underlying mechanism remains unrevealed. In this study, we combine transcriptomics analysis and bacterial genetics, which reveal that transcription regulation plays an important role in mediating the motility defects of the tat mutant of extraintestinal pathogenic Escherichia coli. The Tat substrate proteins responsible for the motility defects are identified. We further show that the Rcs system contributes to the motility suppression. We for the first time reveal the link between the Tat system and bacterial motility, which is important for understanding the physiological functions of the Tat system.

INTRODUCTION

Bacteria utilize various protein transport and secretion systems to deliver proteins that are originally synthesized in the cytoplasm to the subcellular locations beyond the cytoplasm to exert their destined physiological functions (1–3). In many bacteria, a unique protein export system, the twin-arginine translocation (Tat) system, exists to transport a variety of proteins across the cytoplasmic membrane (4). One of the distinguishing features of the Tat system is that it transports fully folded proteins (5). It was originally discovered as a transport system that was closely related to the export of cofactor-containing proteins (6). However, its pleiotropic roles in bacterial physiology have been revealed subsequently (4, 7). A large proportion of them are related to anaerobic respiration, including dehydrogenases, formate dehydrogenases, dimethyl sulfoxide (DMSO) reductases, trimethylamine N-oxide (TMAO) reductase, nitrate reductase, etc. (8). Moreover, the Tat system has been demonstrated to play important roles in the pathogenesis of some important bacterial pathogens, such as pathogenic Escherichia coli, Salmonella, Pseudomonas aeruginosa, and Yersinia pseudotuberculosis (9). In some bacteria, the Tat system is even reported to be essential for viability (10, 11).

Bacterial flagella compose a complex machine that promotes cellular motility, which is important in nutrient acquisition, environmental adaption, and pathogenesis (12–14). The biogenesis of bacterial flagella is under sophisticated transcriptional control. In E. coli, the flhDC operon, which is also known as class I genes, encodes the master regulator that initiates the transcription hierarchy. FlhDC activates the expression of class II genes that encode important regulatory proteins as well as structural components of flagella. Class III genes encode proteins needed in late flagellar assembly, including flagellin FliC. Precise temporal control of flagellar gene expression ensures the production of flagella when needed (15, 16). Apart from this regulatory hierarchy, the expression of bacterial flagellar genes has been reported to be regulated in response to external signals. The Rcs system is an important signal transduction system involved in bacterial envelope stress response and virulence regulation (13). It is composed of three core components, RcsC, RcsD, and RcsB, in which RcsC is a transmembrane histidine kinase that is autophosphorylated when the system is activated and passes the phosphate to RcsD then to RcsB (17). RcsB serves as a response regulator to regulate the expression of downstream genes (17). The Rcs signaling system has been revealed to exert an important regulatory function in controlling bacterial motility as a negative regulator of flagellar biosynthesis (18, 19). In addition, the Rcs system has been reported to be involved in the expression of a variety of genes involved in capsule formation, biofilm formation, and type VI secretion in response to stresses such as envelope stresses, outer membrane damage, and peptidoglycan perturbation (13, 17, 20). Another regulation system involved in the regulation of flagellar expression is the CpxAR two-component system, which is composed of the sensor histidine kinase CpxA and the response regulator CpxR. It can be activated in response to bacterial adhesion to abiotic surfaces, leading to downregulation of fliC expression in enterohemorrhagic E. coli (14). Moreover, appropriate expression of flagellin at different stages of infection has been recently revealed as a new strategy of bacterial pathogenesis (21).

Extraintestinal pathogenic E. coli (ExPEC) is an important bacterial pathogen causing a variety of diseases, including urinary tract infection, septicemia, meningitis, and even death in both humans and farm animals (22–24). Our previous study showed that the Tat system is critical for the virulence of extraintestinal pathogenic E. coli (ExPEC) and revealed that the disruption of the Tat system caused a significant decrease in bacterial motility as reported in other studies (25–27). However, so far, the underlying mechanism of how the Tat system is involved in the regulation of bacterial motility has not been revealed. In this study, we found that in the ExPEC PCN033 strain, deletion of the Tat system caused drastic downregulated expression of flagellar biosynthesis genes. The Tat substrate proteins were further identified that were responsible for the motility defects. We further showed that the Rcs system was a key player that mediates the regulation of motility by the Tat system. Our study for the first time unraveled the link between the Tat system and bacterial motility.

RESULTS

Deletion of the Tat system causes significant alteration in gene expression.

The Tat system has been reported to be involved in a variety of bacterial physiological processes (8). To investigate the influence of the Tat system on gene expression of ExPEC, the transcriptome profiles were compared between ExPEC PCN033 strain (wild type [WT]) and the Tat system-deleted strain (Δtat) by using mRNA sequencing (mRNA-seq). A total of 720 genes were differentially expressed in the Δtat compared with the WT strain [|log₂(fold change)| > 1, P < 0.05], among which 508 genes were upregulated and 212 genes were downregulated, indicating an extensive influence on gene expression due to deletion of the Tat system (see Table S1 in the supplemental material). The KEGG database was then used to perform pathway enrichment analysis. As shown in Fig. 1A, metabolism represented the most enriched level 1 KEGG pathways, in which carbohydrate metabolism, membrane transport, and cell motility were the most enriched level 2 KEGG pathways among the differentially expressed genes. Clusters of orthologous groups (COG) enrichment analysis revealed that the differentially expressed genes belonged to a variety of functional categories (Fig. 1B). Specifically, in the Δtat strain, we found that the genes involved in the biosynthesis of polysialic acid, including the kpsMT and kpsFEDUCS operons, and the genes involved in the biosynthesis of colonic acid, including manB and manC, were significantly upregulated (Table 1), indicating that the capsule biosynthesis-related pathway was activated when the Tat system was disrupted. Several stress response genes, including ydeO (28), yciH (29), ydeP (30), elaB (31), yhcN (32), yggT (33), and bhsA (34), were also upregulated in the Δtat strain (Table 1). Moreover, we noticed that 12 genes belonging to the toxin-antitoxin systems were differentially expressed, among which 6 genes were upregulated and 6 were downregulated (Table 1). These data suggest that deletion of the Tat system caused substantial changes in gene expression.

FIG 1.

Differentially expressed genes (Δtat versus WT). (A) KEGG pathway enrichment of the differentially expressed genes. The differentially expressed genes were analyzed by using the BlastKOALA tool (https://www.kegg.jp/blastkoala/) to assign K numbers which were then used for pathway enrichment by using KEGG Mapper (https://www.genome.jp/kegg/mapper/search.html). The numbers inside the vertical bars are the total numbers of genes belonging to each level 2 KEGG pathway. (B) COG enrichment of the differentially expressed genes. COG enrichment was performed with the EggNOG v5.0 database (http://eggnog5.embl.de/#/app/home).

TABLE 1.

Partial differentially expressed genes in Δtat

No.	Gene ID	Fold change (Δtat/WT)	P value	FDR	Gene name	Description	Function
1	PPECC33_RS16300	115.360	3.15E-33	3.17E-31	kpsT	ABC transporter ATP-binding protein	Capsular polysaccharide biosynthesis
2	PPECC33_RS16295	40.224	4.71E-70	3.48E-67		HAD-IA family hydrolase
3	PPECC33_RS16305	36.504	5.56E-40	8.21E-38	kpsM	ABC transporter permease
4	PPECC33_RS16290	31.779	2.63E-49	7.46E-47		Hypothetical protein
5	PPECC33_RS16280	17.388	1.96E-20	1.06E-18		Hypothetical protein
6	PPECC33_RS16275	16.912	5.47E-35	6.21E-33		Nucleotidyltransferase family protein
7	PPECC33_RS16245	16.912	2.53E-42	4.66E-40	kpsF	KpsF/GutQ family sugar-phosphate isomerase
8	PPECC33_RS16285	11.472	3.79E-29	3.11E-27		Glycosyltransferase family 2 protein
9	PPECC33_RS16250	11.314	1.69E-30	1.47E-28	kpsE	Capsule polysaccharide transporter
10	PPECC33_RS16255	10.056	3.08E-30	2.62E-28	kpsD	Polysaccharide biosynthesis/export family protein
11	PPECC33_RS16270	8.754	9.30E-19	4.48E-17	kpsS	Capsular biosynthesis protein
12	PPECC33_RS16265	7.945	4.89E-31	4.42E-29	kpsC	Capsular polysaccharide biosynthesis protein
13	PPECC33_RS16260	6.364	4.14E-20	2.21E-18	kpsU	3-Deoxy-manno-octulosonate cytidylyltransferase
14	PPECC33_RS10985	2.346	0.000169	0.000959	wzy	O11 family O-antigen polymerase
15	PPECC33_RS10950	2.293	6.53E-07	6.77E-06	manC	Mannose-1-phosphate guanylyltransferase/mannose-6-phosphate isomerase
16	PPECC33_RS10940	2.219	9.50E-08	1.19E-06	cpsG	Colanic acid biosynthesis phosphomannomutase CpsG
17	PPECC33_RS10945	2.071	1.23E-05	9.69E-05	wcaE	Glycosyltransferase
18	PPECC33_RS08185	17.509	9.76E-06	7.82E-05	ydeO	Acid stress response transcriptional regulator YdeO	Stress response
19	PPECC33_RS07220	5.315	0.000105	0.00064	yciH	Stress response translation initiation inhibitor YciH
20	PPECC33_RS08195	3.837	2.65E-07	3.01E-06	ydeP	Acid resistance putative oxidoreductase YdeP
21	PPECC33_RS12175	2.621	1.99E-09	3.33E-08	elaB	Stress response protein ElaB
22	PPECC33_RS17710	2.549	3.38E-12	8.26E-11	yhcN	Peroxide/acid stress response protein YhcN
23	PPECC33_RS16005	2.445	1.14E-06	1.12E-05	yggT	Osmotic shock tolerance protein YggT
24	PPECC33_RS05995	2.219	1.08E-05	8.59E-05	bhsA	Multiple stress resistance protein BhsA
25	PPECC33_RS22830	105.420	1.41E-09	2.47E-08	ghoT	Type V toxin-antitoxin system toxin GhoT	Toxin-antitoxin system
26	PPECC33_RS22825	22.785	3.52E-06	3.12E-05	ghoS	Type V toxin-antitoxin system endoribonuclease antitoxin GhoS
27	PPECC33_RS01555	5.657	0.00019	0.001056	yafO	Type II toxin-antitoxin system mRNA interferase toxin YafO
28	PPECC33_RS03050	3.655	0.009247	0.028521	hokE	Type I toxin-antitoxin system toxin HokE
29	PPECC33_RS01550	2.395	0.004016	0.014354	yafN	Type I toxin-antitoxin system antitoxin YafN
30	PPECC33_RS24555	2.238	0.002638	0.010115	lsoB	Type II toxin-antitoxin system antitoxin LsoB
31	PPECC33_RS07745	0.363	1.37E-08	2.03E-07	hokB	Type I toxin-antitoxin system toxin HokB
32	PPECC33_RS24945	0.409	8.06E-07	8.20E-06		Type I toxin-antitoxin system Hok family toxin
33	PPECC33_RS06820	0.432	6.95E-06	5.78E-05	hokD	Type I toxin-antitoxin system toxin HokD
34	PPECC33_RS00080	0.473	5.55E-07	5.87E-06	mokC	Type I toxin-antitoxin system toxin MokC
35	PPECC33_RS00280	0.480	0.001035	0.004596		Type II toxin-antitoxin system CcdA family antitoxin
36	PPECC33_RS27840	0.490	0.01337	0.038627		Type I toxin-antitoxin system toxin Ldr family protein

Deletion of the Tat system disturbs the expression of flagellar biosynthesis genes.

As we reported previously, the deletion of the Tat system in the ExPEC PCN033 strain led to severe defects in motility (Fig. 2A and reference 25). Also, according to the results of the above KEGG pathway enrichment analysis, cell motility is one of the most influenced pathways in the Δtat strain. Among the first 30 genes with the most downregulated expression, 11 were directly related to bacterial flagellar biosynthesis, including fliC and fliD, which encode the flagellin and the flagellar cap protein, respectively (Table 2). Real-time quantitative PCR (RT-qPCR) was performed, which confirmed that the expression levels fliC and fliD were significantly lower in the Δtat strain than in the WT strain (Fig. 2B). Results of Western blot analysis using anti-FliC antibody also showed a significantly lower level of FliC production in the Δtat strain (Fig. 2C). Therefore, the above-described results suggest that disruption of the Tat system interfered with the expression of flagellar biosynthesis genes.

FIG 2.

Disruption of the Tat system interferes with the expression of the flagellar biosynthesis genes. (A) In the swimming assay, 5 μL of cells of each indicated strain at the mid-log phase were spotted onto the swimming plate (tryptone, 10 g/L; yeast, 5 g/L; NaCl, 10 g/L; agar, 0.2%), which was dried at room temperature for 5 min followed by incubation at 37°C for 7 h. The CΔtat strain is the Δtat strain harboring a plasmid expressing tatABC in trans. The assay was performed in triplicate, and the data shown are representative. (B) RT-qPCR analysis. Total RNA was extracted from cells of each indicated strain grown to the mid-log phase. Genomic DNA was removed, and cDNA was synthesized, which was used as the template for qPCR with a QuantStudio 6 Flex fluorescence quantitative PCR instrument. The expression level of the tested gene was calculated using the 2^−ΔΔCT method and normalized to the housekeeping gene gapA. Data are displayed as the geometric mean ± standard error of the mean (SEM). Statistical significance between the two groups was analyzed using Student’s t test. ***, P  < 0.001. (C) Analysis of flagellin production. Cells of each indicated strain were grown to the mid-log phase in LB at 37°C with shaking. The culture was centrifuged, and the same amount of culture supernatant was collected and concentrated by ultracentrifugation. The same number of bacterial cells was harvested, resuspended in PBS, and lysed by sonication, and these were the whole-cell samples. Samples were normalized and subjected to SDS-PAGE followed by Western blotting using anti-FliC antibody (cat. no. ab93713; Abcam).

TABLE 2.

The first 30 genes with the most downregulated expression in Δtat

No.	Gene ID	Fold change (Δtat/WT)	P value	FDR	Gene namea	Description
1	PPECC33_RS21010	0.001	1.0E-147	4.6E-144	tatA	Sec-independent protein translocase subunit TatA
2	PPECC33_RS21015	0.001	9.10E-66	4.48E-63	tatB	Sec-independent protein translocase subunit TatB
3	PPECC33_RS21020	0.002	4.21E-28	3.27E-26	tatC	Sec-independent protein translocase subunit TatC
4	PPECC33_RS22680	0.011	2.7E-106	6.0E-103	yjcZ	YjcZ-like family protein
5	PPECC33_RS10255	0.024	2.31E-72	2.05E-69	tap	Methyl-accepting chemotaxis protein IV
6	PPECC33_RS10440	0.025	1.48E-39	2.12E-37	fliC*	Flagellin FliC
7	PPECC33_RS10260	0.027	1.73E-41	2.85E-39	tar	Methyl-accepting chemotaxis protein II
8	PPECC33_RS06330	0.033	3.33E-83	4.91E-80	ycgR*	Flagellar brake protein
9	PPECC33_RS12795	0.033	1.38E-48	3.59E-46	flxA	FlxA-like family protein
10	PPECC33_RS10270	0.036	3.73E-43	8.26E-41	cheA	Chemotaxis protein CheA
11	PPECC33_RS10245	0.040	3.04E-58	1.12E-55	cheB	Protein-glutamate methylesterase/protein glutamine deamidase
12	PPECC33_RS10445	0.045	5.80E-38	8.03E-36	fliD*	Flagellar filament capping protein FliD
13	PPECC33_RS10275	0.046	3.14E-66	1.74E-63	motB*	Flagellar motor protein MotB
14	PPECC33_RS10240	0.049	6.55E-53	2.23E-50	cheY	Chemotaxis response regulator CheY
15	PPECC33_RS10235	0.051	2.01E-78	2.23E-75	cheZ	Protein phosphatase CheZ
16	PPECC33_RS10250	0.051	4.46E-63	1.97E-60	cheR	Protein-glutamate O-methyltransferase CheR
17	PPECC33_RS05845	0.054	2.70E-49	7.46E-47	flgK*	Flagellar hook-associated protein FlgK
18	PPECC33_RS24055	0.054	8.65E-53	2.74E-50	yjiH	Nucleoside recognition pore and gate family inner membrane transporter
19	PPECC33_RS05790	0.058	1.33E-59	5.35E-57	flgM*	Anti-sigma-28 factor FlgM
20	PPECC33_RS10280	0.058	1.01E-37	1.36E-35	motA*	Flagellar motor stator protein MotA
21	PPECC33_RS10265	0.061	3.95E-42	7.00E-40	cheW	Chemotaxis protein CheW
22	PPECC33_RS05850	0.066	8.01E-42	1.36E-39	flgL*	Flagellar hook-associated protein FlgL
23	PPECC33_RS20080	0.068	2.30E-33	2.37E-31	uhpT	Hexose-6-phosphate:phosphate antiporter
24	PPECC33_RS19225	0.073	2.41E-40	3.82E-38	pdeH	Cyclic-guanylate-specific phosphodiesterase
25	PPECC33_RS07750	0.075	1.63E-69	1.03E-66	trg	Methyl-accepting chemotaxis protein Trg
26	PPECC33_RS10455	0.090	3.90E-11	8.23E-10	fliT*	Flagella biosynthesis regulatory protein FliT
27	PPECC33_RS10450	0.108	5.03E-24	3.28E-22	fliS*	Flagellar export chaperone FliS
28	PPECC33_RS05785	0.118	1.25E-20	6.93E-19	flgN*	Flagellr biosynthesis chaperone FlgN
29	PPECC33_RS03935	0.135	1.45E-42	2.80E-40	modA	Molybdate ABC transporter substrate-binding protein
30	PPECC33_RS07110	0.137	5.51E-20	2.91E-18	trpE	Anthranilate synthase component I

^aGenes marked with an asterisk (*) are related to flagellar biosynthesis.

Tat substrates MdoD, AmiA, and AmiC are responsible for motility defects.

To determine which Tat substrate was responsible for the loss of motility of the ExPEC Δtat strain, a motility assay using soft agar was performed with the 20 Tat substrate-related mutant strains, which were constructed in our previous study (25). AmiA and AmiC are both Tat substrates and have overlapping functions. Therefore, a double deletion mutant, ΔamiAΔamiC, was constructed and tested (35). As shown in Fig. 3, apart from the Δtat and ΔamiAΔamiC strains, which have been previously shown as motility defective (25), ΔmdoD also showed significantly compromised motility. In trans expression of MdoD in ΔmdoD from the pHSG396-Apra-MdoD plasmid and AmiA and AmiC in ΔamiAΔamiC from the pHSG396-Apra-AmiA/AmiC plasmid restored motility (Fig. 3B). As MdoD was a putative Tat substrate and was reported for the first time to be related to bacterial motility, we tested whether the motility phenotype of ΔmdoD depends on the twin-arginine motif in the N-terminal Tat signal peptide. MdoD^RR-KK in which the characteristic twin-arginine (R3 and R4) in its signal peptide was replaced with twin-lysine was expressed from the pHSG396-Apra-MdoD^RR-KK plasmid in the ΔmdoD strain. It was shown that producing the wild-type MdoD in ΔmdoD restored the motility, while the strain producing MdoD^RR-KK was still defective in motility. Our data suggest that it is the Tat substrate proteins MdoD, AmiA, and AmiC that are responsible for the motility defects of the tat mutant.

FIG 3.

Identification of Tat substrate proteins responsible for loss of motility. (A) For the swimming assay with the Tat substrate-related mutants, 5 μL of cells of each indicated strain grown to the mid-log phase was spotted onto the swimming plate (tryptone, 10 g/L; yeast, 5 g/L; NaCl, 10 g/L; agar, 0.2%) and was dried at room temperature for 5 min followed by incubation at 37°C for 7 h. The assay was performed in triplicate, and the data shown are representative. (B) Swimming assay with ΔamiAΔamiC, ΔmdoD, and the strains in trans expressing the corresponding proteins. The CΔmdoD and CΔmdoD^RR-KK strains are the ΔmdoD strain encompassing a plasmid encoding the wild type and the mutated MdoD in which the twin-arginine in the signal peptide is replaced with a twin-lysine. CΔamiAΔamiC is the ΔamiAΔamiC strain in trans expressing AmiA and AmiC. (C) and (D) RT-qPCR analysis. Total RNA was extracted from cells of each indicated strain grown to the mid-log phase. Genomic DNA was removed, and cDNA was synthesized and was used as the template for qPCR with a QuantStudio 6 Flex fluorescence quantitative PCR instrument. The expression level of the tested gene was calculated using the 2^−ΔΔCT method and normalized to the housekeeping gene gapA. Data are displayed as the geometric mean ± SEM. Statistical significance between the two groups was analyzed using Student’s t test. ***, P < 0.001.

Deletion of rcsDB can partially restore the motility of the tat mutant.

MdoD has been reported as a periplasmic protein involved in glycan biosynthesis, and defects in periplasmic glucan synthesis could induce envelope stress (36–38). AmiA and AmiC are two amidases involved in cell wall remodeling, and their deletion results in cell envelope defects (35). Moreover, several envelope stress response systems, such as the Rcs system and the CpxAR system, have been reported as regulators of bacterial flagella (13, 18, 39). Therefore, a hypothesis was proposed that deletion of the Tat system or the Tat substrates may activate envelope stress response systems which may suppress the expression of flagellar biosynthesis genes. Thus, rcsDB, which encodes the essential components of the Rcs signaling system, was deleted from the Δtat, ΔmdoD, and ΔamiAΔamiC strains, and bacterial motility was tested. In Fig. 4, it is shown that the disruption of the Rcs system completely restored the motility defects of the ΔmdoD strain (P < 0.05) and can partially rescue the motility of the Δtat strain and ΔamiAΔamiC strain (P < 0.05). To test whether the suppressive effect was specific to the Rcs system, another cell envelope stress response system, CpxAR, was deleted in the Δtat strain, which did not rescue the motility phenotype of Δtat (Fig. 4).

FIG 4.

Disruption of the Rcs system restored motility. First, 5 μL of cells of each indicated strain grown the mid-log phase was spotted onto the swimming plate (tryptone, 10 g/L; yeast, 5 g/L; NaCl, 10 g/L; agar, 0.2%) and was dried at room temperature for 5 min followed by incubation at 37°C for 7 h. The assay was performed in 5 replicates for each strain. The representative image is presented on the left. The swimming diameter is measured and shown as the mean ± standard deviation on the right. ***, P < 0.001.

Flagella can be observed in the tat mutant when Rcs was disrupted.

To further confirm that the restored motility phenotype was due to flagellar biosynthesis, transmission electron microscopy analysis was performed to observe bacterial flagella. Consistent with the results of the motility assay, flagella were not seen in Δtat, ΔmodD, or ΔamiAΔamiC but could be observed in the WT, ΔtatΔrcsDB, ΔmodDΔrcsDB, and ΔtatΔrcsDB strains, suggesting that the flagellar biosynthesis was recovered when the Rcs system was deleted (Fig. 5). Western blot analysis also showed that FliC expression and secretion were significantly downregulated in Δtat, ΔmodD, and ΔamiAΔamiC but was restored when Rcs was further disrupted (Fig. 2C). It was noticed that the expression of FliC in the rcsDB-deleted strains was even higher than that in the WT strain. These results suggest that the Rcs system contributes to the disrupted flagellar biosynthesis of the Tat and the Tat substrate mutants.

FIG 5.

Transmission electron microscopy analysis of flagella. Cells of each indicated strain were streaked on an LB agar plate at 37°C for 12 h. A single colony was picked and resuspended in 200 μL of ultrapure water, which was left to stand for 2 h. Then, 10 μL of the bacterial suspension was placed on the grid followed by fixing with 2% phosphotungstic acid staining solution and was imaged using a transmission electron microscope (H-7650; Hitachi, Japan). The scale bar is 1 μm.

The Rcs system was activated when the Tat system was disrupted.

Previous studies have shown that the Rcs system is the negative regulator of the master transcription regulator of flagellar biosynthesis flhDC (18). The loss of motility could be attributed to activation of the Rcs system in the Δtat strain, which downregulated the expression of the flagellar biosynthesis genes. To verify whether the Rcs system was activated when the Tat system was disrupted, we compared the expression of the genes that have been reported to be regulated by the Rcs system between the WT and the Δtat strain, including flhDC (18), rcnB (40), tssB (20), motB (40), and modA (40). As shown in Fig. 6, the expression of rcnB, and tssB was significantly upregulated in the Δtat strain compared with the WT strain. In comparison, the expression of flhD, modA, and motB was significantly downregulated in the Δtat strain, as they have been shown to be negatively regulated by the Rcs system (40). Thus, the data indicate that the Rcs system was activated when the Tat system was disrupted.

FIG 6.

RT-qPCR analysis of the Rcs regulons. Total RNA was extracted from cells of each indicated strain grown to the mid-log phase. Genomic DNA was removed, and cDNA was synthesized and was used as the template for qPCR with a QuantStudio 6 Flex fluorescence quantitative PCR instrument. The expression level of the tested gene was calculated using the 2^−ΔΔCT method and normalized to the housekeeping gene gapA. Data are displayed as the geometric mean ± SEM. Statistical significance between the two groups was analyzed using Student’s t test. ***, P < 0.001.

DISCUSSION

Flagellum-mediated motility is critical for the virulence of several bacterial pathogens in which flagella function at various stages of infection. In enteropathogenic E. coli (EPEC), flagella promote the adhesion of bacteria to epithelial cells (41). In avian pathogenic E. coli (APEC) and Shiga toxin-producing E. coli (STEC), flagella, although not directly involved in adhesion, play important roles in invasion (42–44). Flagella have also been demonstrated to be important for the initial interaction with surfaces, which is critical for the formation of biofilms (45). Moreover, flagella can be recognized by the Toll-like receptor 5, which elicits host immune response, and bacterial pathogens have evolved strategies, such as flagellar expression control, flagellin sequence variations, and flagellin modification, to avoid or evade such immune clearance to establish infection (46). Recently, the Tat system has been reported to play an important role in the pathogenesis of several bacteria (47–49). Interestingly, the compromised motility phenotype has been constantly observed when the Tat system is disrupted in a variety of bacteria, including E. coli (26), Salmonella (27), P. aeruginosa (50), and Yersinia (51). However, so far, the underlying mechanism of how the Tat system is linked to bacterial motility remains unclear.

To understand how the Tat system affects motility, we first compared the mRNA profile between the WT and the Δtat strain. Our data show that a substantial number of genes undergo differential expression, suggesting an important role of the Tat system in the physiology of ExPEC. The results show that the expression of capsular biosynthesis genes and stress response systems is increased in the Δtat strain (Table 1). This is consistent with the findings in a previous report (52). A possible explanation could be that the deletion of the Tat system causes outer membrane defects, which may result in the activation of the stress response systems. Among these envelope stress response systems, the Rcs system has been revealed to positively regulate the expression of capsular polysaccharide biosynthesis (53), and the activation of the Rcs system by the Tat system disruption has been implicated in the previous transcriptomics study (52). In addition, several toxin-antitoxin (TA) systems also showed differential expression in the Δtat strain (Table 1). Among them, ghoT/S showed a drastic upregulation upon Tat deletion, which has been reported to function under stress conditions by reducing metabolism, a possible survival strategy when confronting stresses (54). However, how the Tat system influences these TA systems needs further investigation. The downregulated genes include a substantial amount of flagellar biosynthesis genes, which corresponds with the compromised motility phenotype. This is consistent with a previous study showing that the deletion of the Tat system abolished the synthesis of flagellin in E. coli O157:H7 (26). However, it was not the case in E. coli MC4100, a commensal K-12 strain (52). Whether this difference is attributed to pathotypes remains unclear.

In E. coli, over 30 proteins have been predicted or verified to be transported by the Tat system (8). However, none of them has been revealed to be directly involved in flagellar biosynthesis. To reveal the link between the Tat system and flagellar biosynthesis, we perform a motility phenotype screening with the Tat substrate-related mutants, which identified three Tat substrate proteins, AmiA, AmiC, and MdoD, that are responsible for the loss of motility of the Δtat strain. Deletion of the Tat-dependent amidases AmiA and AmiC simultaneously has been previously shown to cause motility defects in both E. coli and Salmonella (25, 55). We for the first time report that another Tat substrate, MdoD (also known as OpgD), is related to bacterial motility. MdoD is involved in osmoregulated periplasmic glucan (OPG) biosynthesis in which deletion of MdoD affects the glucose backbone structures of periplasmic glucans (56). It has also been reported previously that defects in periplasmic glucan synthesis cause motility defects (36). The osmoregulated periplasmic glucans play important roles in coping with environmental stresses (37, 57). Defects in periplasmic glucans have been reported to induce envelope stress (58, 59). There are several envelope stress response systems that exist in E. coli to cope with envelope stresses (60), among which the Rcs system has been reported to be a negative regulator of flhDC, the master regulator of bacterial flagellar biosynthesis (13, 18). Our result that the disruption of the Rcs system restores the motility of the ΔmdoD strain further demonstrates an important role of Rcs in mediating the motility defects.

It should be noted that the motility was only partially restored for ΔtatΔrcsDB and ΔamiAΔamiCΔrcsDB, suggesting that other factors than the Rcs system may also be involved in the disruption of the motility. AmiA and AmiC are two Tat-exported N-acetylmuramoyl-l-alanine amidases functioning in septum cleavage during cell division (61–63). Moreover, the inactivation of AmiA and AmiC is reported to compromise outer membrane integrity (35). Therefore, peptidoglycan synthesis and outer membrane integrity may be the other factors affecting the motility of the tat mutant.

Conclusions. In conclusion, we show that disruption of the Tat system abolishes motility of extraintestinal pathogenic E. coli and interferes with the expression of a large number of genes, among which, genes involved in flagellar biosynthesis are significantly affected. Tat substrate proteins MdoD, AmiA, and AmiC are identified that are related to the loss of motility. The Rcs system is then demonstrated to play an important role in mediating the loss of motility of the tat mutant. Our study for the first time reveals the underlying link between the Tat system and bacterial motility.

MATERIALS AND METHODS

Strains and plasmid construction.

The strains and plasmids used in this study are all listed in Table 3. All E. coli strains except E. coli χ7213 were grown in lysogeny broth (LB) medium or on LB agar plates at 37°C. E. coli strain χ7213 was cultured in lysogeny broth (LB) medium supplemented with 50 μg/mL of diaminopimelic acid (DAP). The DNA sequences of the primers used in this study are listed in Table S2. The gene deletion strains were constructed according to our previous report (25). Briefly, the pRE112 plasmid containing the homologous arms flanking the target gene to be deleted was used to transform E. coli χ7213 competent cells, which was used as the donor strain for transconjugation. The single exchanged strain was then confirmed by PCR after transconjugation and was incubated on 10% sucrose-containing LB to counterselect the double exchanged strain. The mutant strain was then verified by PCR. To in trans express the Tat substrate proteins in the mutant strains, pHSG396-Apra derivative plasmids were used in which the DNA fragment containing the corresponding coding sequence following a constitutive promoter was inserted into the multiple cloning site. The pHSG396-Apra plasmid was constructed by inserting an apramycin resistance cassette at the NcoI site within the chloramphenicol resistance cassette.

TABLE 3.

Bacterial strains and plasmids used in this study

Name	Description	Reference or source
Strains
ExPEC PCN033	A virulent clinical strain isolated from diseased pig, wild type	72
E. coli χ7213	A diaminopimelic acid autotrophic E. coli strain used for transconjugation	73
Δtat	As ExPEC PCN033, tatABC deleted	25
CΔtat	Δtat strain expressing tatABC from a plasmid
ΔmodD	As ExPEC PCN033, modD deleted
ΔamiA	As ExPEC PCN033, amiA deleted
ΔamiC	As ExPEC PCN033, amiC deleted
ΔefeOB	As ExPEC PCN033, efeOB deleted
ΔsufI	As ExPEC PCN033, sufI deleted
ΔmepK	As ExPEC PCN033, mepK (previously known as ycbK) deleted
ΔfhuD	As ExPEC PCN033, fhuD deleted
ΔmoaA	As ExPEC PCN033, moaA deleted
ΔnrfC	As ExPEC PCN033, nrfC deleted
ΔhybOA	As ExPEC PCN033, hybOA deleted
ΔyagT	As ExPEC PCN033, yagT deleted
ΔyahJ	As ExPEC PCN033, yahJ deleted
ΔhyaA	As ExPEC PCN033, hyaA deleted
ΔfdoG	As ExPEC PCN033, fdoG deleted
ΔwacM	As ExPEC PCN033, wcaM deleted
ΔcueO	As ExPEC PCN033, cueO deleted
ΔnapG	As ExPEC PCN033, napG deleted
ΔydhT	As ExPEC PCN033, ydhT deleted
ΔydhX	As ExPEC PCN033, ydhX deleted
ΔfdnG	As ExPEC PCN033, fdnG deleted
ΔamiAΔamiC	As ExPEC PCN033, amiA and amiC deleted
ΔrcsDB	As ExPEC PCN033, rcsDB deleted	This study
ΔtatΔrcsDB	As Δtat, rcsDB deleted	This study
ΔmdoDΔrcsDB	As ΔmdoD, rcsDB deleted	This study
ΔamiAΔamiCΔrcsDB	As ΔamiAΔamiC, rcsDB deleted
ΔtatΔcpxAR	As Δtat, cpxAR deleted	This study
Plasmids
pRE112	The plasmid used in mutant construction	Lab stock
pHSG396-Apra	The plasmid used for in trans protein expression in ExPEC	This study
pHSG396-Apra-MdoD	As pHSG396-Apra, a DNA fragment containing a tat promoter followed by the coding sequence of mdoD followed by insertion between XbaI and EcoRI	This study
pHSG396-Apra-MdoDRR-KK	As pHSG396-Apra-MdoD, expressing the variant MdoD, where the R3 and R4 residues were both replaced with a lysine	This study
pHSG396-Apra-RcsDB	As pHSG396-Apra, a DNA fragment containing a tat promoter followed by the coding sequence of rcsDB inserted between XbaI and EcoRI	This study
pHSG396-Apra-AmiA/AmiC	As pHSG396-Apra, a DNA fragment containing the coding sequence of rcsDB followed by a tat promoter inserted between kpnI and SacI	This study
pRE112-rcsDB	As pRE112, a DNA fragment containing the upstream 1,000 bp and downstream 1,000 bp of rcsDB inserted between kpnI and SacI	This study
pRE112-cpxAR	As pRE112, a DNA fragment containing the upstream 1,000 bp and downstream 1,000 bp of cpxAR inserted between kpnI and SacI	This study

mRNA-seq.

The ExPEC PCN033 and Δtat strains were subcultured from overnight-grown cultures into fresh LB and grown to the mid-log phase at 37°C with shaking. The cells were harvested and washed with normal saline followed by total RNA extraction using a HiPure bacterial RNA kit (Magen, China) according to the manufacturer’s instructions. RNA quantitation was performed using a Qubit 3.0 device (Thermo Fisher Scientific, Massachusetts, USA). Whole mRNA-seq libraries were constructed by Guangdong Magigene Biotechnology Co., Ltd. (Guangzhou, China), using NEBNext Ultra directional RNA library prep kit for Illumina (New England Biolabs, Massachusetts, USA) following the manufacturer’s recommendations. Briefly, the bacterial 16S rRNA in the total RNA samples were depleted using a Ribo-off rRNA depletion kit (bacteria) (Vazyme, Nanjing, China). Fragmentation was carried out using the RNA first-strand synthesis module (New England Biolabs). The first-strand cDNA was synthesized using a random hexamer primer and M-MuLV reverse transcriptase (New England Biolabs). When synthesizing the second strand of cDNA, a chain-specific library was constructed by replacing dTTP with dUTP to greatly improve the accuracy of the results. The remaining overhangs were converted into blunt ends via exonuclease/polymerase activities. After adenylation of the 3′ ends of DNA fragments, a NEBNext adaptor with a hairpin loop structure was ligated to prepare for hybridization. To select cDNA fragments of preferentially 150 to ∼200 bp in length, the fragments were selected with SpeedBead magnetic carboxylate modified particles (Global Life Sciences Solutions Operations, Buckinghamshire, UK). Then, PCR was performed with Phusion high-fidelity DNA polymerase, universal PCR primers, and index (X) primer. Lastly, PCR products were purified with SpeedBead magnetic carboxylate modified particles, and library insert size was assessed on the Qsep400 high-throughput nucleic acid protein analysis system (Houze Biological Technology Co., Hangzhou, China). The clustering of the index-coded samples was performed on a cBot cluster generation system. After cluster generation, the library was sequenced on an Illumina NovaSeq 6000 platform, and 150-bp paired-end reads were generated.

mRNA-seq data analysis.

The raw RNA-seq data in FASTQ format were processed using Trimmomatic (v0.36) to acquire the clean data (clean reads) (64). Clean reads were mapped to NCBI Rfam databases to remove the rRNA sequences with Bowtie 2 (v2.33) (65). The remaining mRNA sequences were mapped to the reference genome with HISAT 2 (v2.1.0) (66). HTSeq-count (v0.9.1) was used to obtain the read count and function information of each gene according to the result of the mapping (67). To compare the expression level of genes, the fragments per kilobase per million (FPKM) of each gene was calculated. Differentially expressed genes were analyzed using edgeR (v3.16.5) (68). The resulting P value was adjusted using Benjamini and Hochberg’s approach for controlling the false-discovery rate. Genes with an FDR of ≤0.05 and |log₂(fold change)| of ≥1 were taken as differentially expressed genes. To do KEGG pathway enrichment, the differentially expressed genes were assigned a K number using the BlastKOALA tool (69). The K numbers were then used for pathway enrichment by using KEGG Mapper (70). The COG enrichment was performed with the EggNOG v5.0 database (71).

Motility assay.

E. coli cells were subcultured 1:100 from overnight-grown cultures in LB and grown at 37°C with shaking to the mid-log phase. Then, 5 μL of the bacterial culture was spotted onto the swimming plate (tryptone, 10 g/L; yeast, 5 g/L; NaCl, 10 g/L; agar, 0.2%), which was dried at room temperature for 5 min followed by incubation at 37°C for 7 h.

Transmission electronic microscopy.

Cells of each indicated strain were streaked on an LB agar plate at 37°C for 12 h. A single colony was picked and resuspended in 200 μL of ultrapure water which was left to stand for 2 h. Then, 10 μL of the bacterial suspension was placed on the grid followed by fix with 2% phosphotungstic acid staining solution and was imaged using a transmission electron microscope (H-7650; Hitachi, Japan).

Real-time quantitative PCR.

Total RNA was extracted using the RNA isolation kit (Tianmo Biotech, China). The purity and degradation of the RNA were analyzed by measuring the absorbance and agarose gel electrophoresis. HiScript Q select RT SuperMix (+gDNA wiper) (Vazyme, China) was used to remove residual genomic DNA and synthesize (cDNA according to the manufacturer’s instructions. The cDNA was used as the template for qPCR using the TB Green premix Ex Taq II kit (TaKaRa Biomedical Technology, Beijing, China) according to the manufacturer’s instructions with a QuantStudio 6 Flex fluorescence quantitative PCR instrument. The expression level of the tested gene was calculated using the 2^−ΔΔCT method and normalized to the housekeeping gene gapA.

Western blotting.

To detect flagellin production, cells of each indicated strain were subcultured 1:100 from overnight-grown cultures into LB and grown to the mid-log phase at 37°C with shaking. The culture was centrifuged, and the same amount of culture supernatant was collected and concentrated by ultracentrifugation. The same number of bacterial cells were harvested, resuspended in phosphate-buffered saline (PBS), and lysed by sonication, and these were the whole-cell samples. The samples were normalized and separated on 10% SDS-PAGE and were then transferred to polyvinylidene difluoride (PVDF) membranes and probed with anti-FliC antibody (catalog [cat.] no. ab93713; Abcam, Shanghai, China).

Data availability.

Sequence data have been submitted to the NCBI GEO database, and the accession number is GSE181969.