Enteropathogenic and Enterohemorrhagic Escherichia coli Type III Secretion Effector EspV Induces Radical Morphological Changes in Eukaryotic Cells

Abstract

Enteropathogenic Escherichia coli (EPEC) and enterohemorrhagic Escherichia coli (EHEC) are important human pathogens that rely on translocation of type III secretion system (T3SS) effectors for subversion of signal transduction pathways and colonization of the mammalian gut mucosa. While a core set of effectors is conserved between EPEC and EHEC strains, a growing number of accessory effectors that were found at various frequencies in clinical and environmental isolates have been recently identified. Recent genome projects identified espV as a pseudogene in EHEC but a putative functional gene in EPEC strains E110019 and E22 and the closely related mouse pathogen Citrobacter rodentium. The aim of this study was to determine the distribution of espV among clinical EPEC and EHEC strains and to investigate its function and role in pathogenesis. espV was found in 16% of the tested strains. While deletion of espV from C. rodentium did not affect colonization dynamics or fitness in mixed infections, expression of EspV in mammalian cells led to drastic morphological alterations, which were characterized by nuclear condensation, cell rounding, and formation of dendrite-like projections. Expression of EspV in yeast resulted in a dramatic increase in cell size and irreversible growth arrest. Although the role of EspV in infection and its target host cell protein(s) require further investigation, the data point to a novel mechanism by which the T3SS subverts cell signaling.

Enteropathogenic Escherichia coli (EPEC) and enterohemorrhagic E. coli (EHEC) are important human pathogens (14). While EPEC is a major cause of infantile diarrhea in the developing world (9), EHEC is associated with food-borne outbreaks in the developed world and can cause bloody diarrhea, hemorrhagic colitis, and hemolytic uremic syndrome due to the elaboration of a Shiga toxin (14). Citrobacter rodentium causes transmissible colonic hyperplasia in mice and is used as an infection model for EPEC and EHEC (39). The hallmark of EPEC, EHEC, and C. rodentium infection is formation of attaching and effacing (A/E) lesions as a mechanism to colonize the gut mucosa. A/E lesions are characterized by effacement of the brush border microvilli and intimate attachment of the bacteria to the apical plasma membrane of enterocytes (30). Moreover, during infection of cultured cells, EPEC, EHEC, and C. rodentium trigger actin polymerization under attached bacteria with the formation of elevated pedestal-like structures (14). The capacity to trigger A/E lesions and actin polymerization is encoded by the locus of enterocyte effacement (LEE) pathogenicity island (38). The LEE encodes a type III secretion system (T3SS) (27, 18), transcriptional regulators, translocators, chaperones, the outer membrane adhesin intimin (28), and several effector proteins that are translocated into the infected cell (21). Following translocation, the effectors target different cellular compartments and subvert numerous signaling pathways for the benefit of the bacteria.

The first effectors identified in A/E pathogens were those encoded by the LEE, including EspB, EspF, EspG, EspH, EspZ, Map, and Tir (21). However, soon after, many effectors encoded outside the LEE were identified and characterized, including Cif (35); EspJ (15); EspM (2); and NleB, NleC, NleE, and NleH (17). All the LEE-encoded effectors and a number of non-LEE-encoded effectors comprise the core effector set found in all A/E pathogens studied so far (26). Other non-LEE effectors show variable distribution among clinical and environmental isolates, ranging from less than 2% for EspT up to 50% for EspM (1). Indeed, the diversity of the effector protein repertoire has been highlighted by various genome projects, which show that while EPEC strain E2348/69 encodes 21 effectors (26), the incomplete genome sequences of EPEC strains B171, E22, and E110019 encode at least 28, 40, and 24 effectors, respectively. In addition, the EHEC O157:H7 strain Sakai encodes 50 effectors (50), while C. rodentium encodes 30 effectors (44). Functional studies show that the EPEC and EHEC effectors exhibit diverse activities, ranging from modulation of the cell cycle by Cif (35) to subversion of NF-κB by NleE, NleB, and NleH (20, 43); phagocytosis by EspF and EspJ (7, 34, 45); apoptosis by NleH and EspF (11, 25); and actin dynamics by Tir, Map, EspM, EspT, EspG, and EspH (2, 5, 6, 24, 37, 51).

Although the majority of the T3SS effectors of A/E pathogens have been documented, it is likely that some additional effectors are yet to be identified and characterized. For example, espV was recently identified as a pseudogene in EHEC O157:H7 (50), although an intact open reading frame was found in the genome sequence of C. rodentium (44). The aim of this study was to determine the distribution of espV among clinical EPEC and EHEC strains and to investigate its possible function and role in pathogenesis.

MATERIALS AND METHODS

Bacterial strains, growth conditions, and cell culture.

The bacterial strains used in this study are listed in Table 1. Bacteria were grown from single colonies in Luria-Bertani (LB) broth in a shaking incubator at 37°C or maintained on LB plates. Culture media were supplemented with ampicillin (100 μg ml⁻¹) as appropriate.

TABLE 1.

Strains used in the study

Strain	Description	Source/reference
ICC169	C. rodentium	3, 53
ICC313	ICC169 ΔespV	This study
E110019	EPEC O111:H19 wild type	52
E2348/69	EPEC O127:H6 wild type	33
ICC192	E2348/69 ΔescN	22

EPEC cultures were primed prior to infection by growth in Dulbecco's modified Eagle's medium (DMEM) with 1,000 mg liter⁻¹ glucose for 3 h (10) before addition of 1 mM IPTG (isopropyl-β-d-thiogalactopyranoside) to induce protein expression when required.

Plasmids and molecular techniques.

The plasmids used in this study are listed in Table 2, and the primers are listed in Table 3. The genes encoding effector proteins were amplified by PCR from genomic DNA. espV_Cr, amplified from C. rodentium DNA, was cloned into pCX340 with a C-terminal TEM-1 fusion or into pRK5 with an N-terminal Myc tag. espV_E11 was amplified from EPEC E110019 DNA and cloned into pSA10 with a C-terminal hemagglutinin (HA) tag and in the yeast vector pYES2. nleH1 was amplified from EPEC E2348/69 and cloned into pYES2. All constructs were verified by DNA sequencing.

TABLE 2.

Plasmids used in the study

Name	Description	Source/reference
pICC309	pCX340::nleD	36
pICC557	pCX340::espVCr	This study
pICC558	pRK5::myc-espVCr	This study
pSA10	pKK177-3 with Lac	46
pICC559	pSA10::espVE11	This study
pICC560	pYES2::espVE11	This study
pICC561	pYES2::nleH1E69	This study

TABLE 3.

Primers used in this study

Name	Sequence
pCX340-espV-F	5′-GGAATTCCATATGTTTGCAGCAAAGCCCGA-3′
pCX340-espV-R	5′-AAAGAATTCTTGTCTTTGGTTTCAGCTTCG-3′
espV-1	5′-ATGTTTGCAGCAAAGCCCGA-3′
espV-2	5′-TCAATCTTTAGTTTCAGCTTCG-3′
espV-3	5′-ATCTGGAGCGTGGCCTTTGGAA-3′
espV-4	5′-CCTGCATCATAACGAACCTT-3′
espVCr-mut 1	5′-GGTCGATTTTATCCTTCAGTG-3′
espVCr-mut 2	5′-AAAGAATTCAACCACTATTCCTTTATT-3′
espVCr-mut 3	5′-TTTGAATTCACATGTTCTAGTGAAGAAGAC-3′
espVCr-mut 4	5′-GGTTACACCAGCACGAGACAA-3′
pRK5-espVCr-F	5′-TTTGGATCCATGTTTGCAGCAAAGCCCGATAGC-3′
pRK5-espVCr-R	5′-AAAGAATTCTCAGTCTTTGGTTTCAGCTTCGCG-3′
pSA10-espVE11-F	5′-TTTGAATTCATGTTTGCAGCAAAGCCCGA-3′
pSA10-espVE11-R	5′-CCAATGCATTGGTTCTGCAGTCAATCTTTAGTTTCAGCTTCG-3′
pYES2-espVE11-F	5′-TTTGGATCCAACACAATGTCTTTTGCAGCAAAGCCCGA-3′
pYES-espVE11-R	5′-AAAGAATTCTCAATCTTTAGTTTCAGCTTCG-3′
pYES-nleH1E69-F	5′-CGCGGATCCATGTCTCTATCACCATCTTCTG-3′
pYES-nleH1E69-R	5′-TCTAGACTAAATTTTACTTAATACCACAC-3′
2359	5′-ATGTTTGCAGCAAAGCCCGA-3′
2360	5′-TCAATCTTTATGTTCAGCTTCG-3′

β-Lactamase (TEM-1) translocation assay.

HeLa cells were seeded in 96-well plates. Wild-type EPEC E2348/69 and the ΔescN T3SS-null mutant were transformed with the pCX340 vector encoding an EspV_Cr-TEM-1 fusion; an NleD_E69-TEM-1 fusion was used as a positive control. A translocation assay was performed as described previously (8, 41).

Prevalence of espV among EPEC and EHEC strains.

We screened for the presence of espV by colony PCR using two sets of primers derived from the three available espV sequences of C. rodentium, EPEC E110019, and EPEC E22: espV-1 and espV-2, and espV-3 and espV-4 (Table 3). C. rodentium and EPEC E2348/69 were used as positive and negative controls. Representative PCR amplicons were sequenced. For dot blot hybridization, digoxigenin (DIG)-labeled probe was generated by PCR using the primers 2359 and 2360 (Table 3), which amplified the entire espV gene.

Construction of a C. rodentium espV mutant.

The espV gene of C. rodentium was replaced by a kanamycin cassette using lambda red-based mutagenesis (16) adapted as described in reference 12 with primers espVCr-mut 1, espVCr-mut 2, espVCr-mut 3, and espVCr-mut 4.

Mixed-strain infections of mice and determination of the in vivo competitive index.

All animal infection studies were approved by the animal ethics committee at the University of Melbourne, project 0808209. To determine the in vivo competitive index (CI) for mutant derivatives of C. rodentium, overnight cultures of approximately 2 × 10⁹ CFU of the mutant were combined 50:50 with wild-type C. rodentium in 200 μl phosphate-buffered saline (PBS) and used to inoculate 5- to 6-week-old male C57BL/6 or C3H/He mice by oral gavage. At day 7 postinoculation, the mice were killed, their colons were dissected, and colon scrapings were homogenized. To determine the proportion of wild-type to mutant bacteria, dilutions of the inoculum and the recovered bacteria were plated onto LB agar containing nalidixic acid only and onto LB agar containing nalidixic acid and kanamycin. The ability of each mutant to compete with the wild-type strain was analyzed in at least 6 animals, and the CI was calculated as the proportion of mutant to wild-type bacteria recovered from animals divided by the proportion of mutant to wild-type bacteria in the inoculum (40).

Single-strain infections of mice.

Overnight cultures of wild-type C. rodentium and the ΔespV mutant were resuspended in PBS for inoculation, as described above. Seven 5- to 6-week-old C57BL/6 mice per strain were inoculated by oral gavage with approximately 2 × 10⁹ CFU in 200 μl PBS. The viable count of the inoculum was determined by retrospective serial dilution and plating on LB agar containing the appropriate antibiotic. Stool samples were recovered aseptically at various time points after inoculation, and the number of viable bacteria per gram of stool was determined by plating on antibiotic selective media. The limit of detection was 10 CFU g feces⁻¹.

Transfection.

HeLa cells were transfected with the mammalian expression vector pRK5 containing EspV_Cr fused to a Myc tag by Lipofectamine 2000 (Invitrogen) according to the manufacturer's recommendations. The cells were incubated at 37°C in a humidified incubator with 5% CO₂ for 24 h.

Infection of HeLa cells with EPEC E2348/69.

Forty-eight hours prior to infection, HeLa cells were seeded onto glass coverslips at a density of approximately 5 × 10⁵ cells per well and maintained in DMEM 1000 supplemented with 10% fetal calf serum (FCS) at 37°C in 5% CO₂. Three hours before infection, the cells were washed three times with PBS and the medium was replaced with fresh DMEM 1000 without FCS. Overnight cultures of the appropriate bacteria were inoculated at 1:50 into DMEM and primed as described previously (10). Primed bacteria (0.5 ml) were added to each well, and the infection was carried out at 37°C in 5% CO₂ for 1 h. The cells were washed and then incubated with DMEM supplemented with gentamicin at 50 μg/ml for 4 h.

Immunofluorescence staining and microscopy.

Coverslips were washed three times in PBS, fixed with 3% paraformaldehyde for 20 min, and washed three more times in PBS. For immunostaining, the cells were permeabilized for 4 min in 0.5% PBS-Triton X-100, washed three times in PBS, and quenched for 30 min with 50 mM NH₄Cl. The coverslips were blocked for 1 h with PBS-0.5% bovine serum albumin (BSA) before incubation for 1 h with mouse anti-Myc primary antibody (Millipore). The coverslips were washed three times with PBS and then incubated with the secondary donkey anti-mouse IgG antibodies conjugated to a Cy3 fluorophore (Jackson Laboratories) at a 1:200 dilution. Actin was stained using Oregon Green-phalloidin (Invitrogen) at 1:100 dilution, and DNA was stained with DAPI (4′,6-diamidino-2-phenylindole). All dilutions were in PBS-0.5% BSA. Coverslips were mounted on slides using ProLong Gold anti-fade reagent (Invitrogen) and visualized with a Zeiss Axioimager immunofluorescence microscope (100× objective, giving a total magnification of ×1,000) using the following excitation wavelengths: Cy3, 605 nm, and Oregon Green, 525 nm. All images were analyzed using the AxioVision Rel 4.5 software and trimmed to 5 cm² (300 pixels) using Adobe Photoshop.

SEM.

Cells were seeded on glass coverslips and transfected or infected with the appropriate plasmids or strains as described above. The cells were washed three times in phosphate buffer, pH 7.4, and then fixed with 2.5% glutaraldehyde (Agar) in phosphate buffer, pH 7.4, for 15 min. The coverslips were then washed with phosphate buffer, pH 7.4, a further 3 times before being postfixed in 1% osmium tetroxide for 30 min. The cells were then washed 3 times in phosphate buffer before being washed for 15 min in graded ethanol solutions from 50% to 100% to dehydrate the samples. The cells were then transferred to an Emitech K850 Critical Point drier and processed according to the manufacturer's instructions. The coverslips were coated in a gold-palladium mixture using an Emitech Sc7620 minisputter to a thickness of approximately 370 Å. Samples for scanning electron microscopy (SEM) were then examined blindly at an accelerating voltage of 25 kV using a Jeol JSM-6390.

Yeast toxicity and morphology assays.

All yeast transformations were carried out by the standard lithium acetate method. For yeast toxicity assays, Saccharomyces cerevisiae (strain BY4741) was grown on minimal medium plus 2% glucose to repress GAL1-espV_E11 expression. Expression of EspV_E11 or NleH1_E69 was induced by plating yeast on medium with 2% galactose as the carbon source. Transformed yeast colonies were incubated at 30°C for 72 h. To determine yeast morphology, colonies were selected on medium containing glucose and then subcultured into liquid minimal medium containing galactose for 24 h at 30°C. Yeast was visualized using phase-contrast with a Zeiss Axioimager immunofluorescence microscope (40× objective, giving a total magnification of ×4,000), and the area of the yeast was measured using Axiovision Rel 4.5 software. For the SEM analysis, the same method described above was used to visualize yeast bound to the coverslips coated with poly-l-lysine.

Statistics.

All statistical tests were performed using GraphPad InStat version 3.06. All experiments were performed in duplicate and repeated a minimum of three times.

RESULTS

EspV is a novel T3SS effector of A/E pathogens.

EspV, which shares amino acid sequence with the C-terminal part of the T3SS effector AvrA of Pseudomonas syringae, was first described as a putative T3SS effector in EHEC O157:H7 strain Sakai (50). However, espV appeared to be a pseudogene, which was consistent with the lack of T3SS-dependent protein translocation (50). Although many other EPEC and EHEC strains harbor sequences similar to the pseudogene identified in EHEC Sakai, longer espV versions were identified in the genomes of C. rodentium, the rabbit EPEC strain (rEPEC) E22, and the EPEC strain E110019 (44, 50). Although the ∼44-kDa EspV proteins encoded by these strains share 86 to 92% sequence identity, they are encoded in different loci in the three pathogens (Fig. 1A and B). Using the TEM-1-β-lactamase fusion assay (8), we demonstrated that EspV is translocated, in a T3SS-dependent manner, by EPEC E2348/69 (Fig. 1C). The T3SS effector NleD was used as a positive control.

FIG. 1.

(A) Multiple-sequence alignment with hierarchical clustering of EspV from rEPEC E22, EPEC E110019, and C. rodentium and the C-terminal part of AvrA from P. syringae. Residues that are identical or similar are highlighted in dark or light gray, respectively. (B) Graphical representation of the chromosomal loci of espV (shown as black arrows) in C. rodentium, rEPEC E22, and EPEC E110019. (C) Translocation of EspV was quantified using a Fluostar Optima reader with excitation at 410 nm (10-nm band-pass). Emission was detected via 447-nm (blue fluorescence) and 520-nm (green fluorescence) filters; translocation was expressed as the 447/520-nm ratio. EspV-TEM and NleD-TEM, which was used as a positive control, were translocated from EPEC E2348/69 wild type (wt), but not from the T3SS mutant E2348/69 ΔescN. The error bars represent mean standard deviations (SD). Significance, indicated by asterisks, was tested using a one-way analysis of variance (ANOVA) with Bonferroni correction (*, P < 0.01).

Distribution of espV among EHEC and EPEC strains.

In order to determine the distribution of espV among A/E pathogens, we performed a PCR screen using two sets of primers designed to detect espV from C. rodentium, E22, and E110019, but were unable to amplify the espV pseudogenes. Of the 230 EPEC and EHEC isolates tested, espV was found in 37 strains (16%) (Table 4 ). In order to exclude the possibility of false-negative PCR results due to sequence variation in primer-annealing sites, we performed dot blot analysis of 154 strains using a DIG-labeled espV probe. EPEC O26 strain E65/56 served as a positive control, and hybridizations were performed at stringency where the negative control (EHEC O157 strain EDL933) was not detected. All the PCR-positive strains hybridized with the probe, while all the PCR-negative strains did not.

TABLE 4.

Distribution of espV among A/E strains

Pathotype and serotype (no. of strains)a	No. of espVd
EHEC
O15:HNM (1), O15:H- (2), O15:H2 (1b), O15:H14 (1)	3
O76:H7 (1)	1
O98:H- (1b)	1
O145:H- (1b)	1
O147:HNM (1)	1
aEPEC
O26:H- (2), O26:H11 (1)	3
OR:H- (2), Ont:H19 (5), Ont:H5 (1), O55:H7 (1)	1
O177:H-e (1), O45:H19 (1), O15:NT (1)	0
tEPEC
O13:H- (1)	1
O26:H- (1)	1
O45:H- (1)	1
O111:HND (1), O111:H- (1), O111:H2 (2)	1
O153:H- (1), O153:H7 (2), O153:H9 (1)	1
AEECc
O8:HND (1)	1
O15:H- (2)	2
O103:H2 (3)	2
O109:H2 (1)	1
O128:H2 (1)	1

^aespV-negative strains are as follows: EHEC, OND:HND (1), ONT:H- (1), O1:H7 (1), O5:HND (1), O5:H- (3), O26:HND (3), O26:H11 (2), O26:H21 (2), O48:H21 (1), O49:H- (1), O71:H11 (1), O91:H- (2), O91:H14 (2), O111:H- (3), O111:HNM (1), O111ac:H- (1), O113:H7 (1), O113:H21 (8), O116:H21 (1), O118:H16 (1), O121:HNT (1), O121:H19 (1), O128:H2 (2), O130:H11 (1), O146:H2 (1), O157:HND (3), O157:H- (3), O157:H7 (8), O168:H28 (1), and O172:HNM (2); aEPEC, O111:H- (1), O114:H2 (1), O2:H45 (1), Ont:H6 (3), O88:H- (1), OR:H2 (1), O25:H7 (1), Ont:H34 (4), O161:H40 (1), O107:H8 (1), O98:H8 (1), O117:H2 (1), Ont:H19 (5), O119:H21 (3), O176/172:H49 (1), O170/172:H49 (1), O157:H39 (1), O123:H45 (1), Ont:H21 (2), O153:H9 (1), O153:H7 (2), O71:H6 (1), O128:H21 (1), Ont:H8 (1), OR:H40 (1), Ont:R (1), O51:H49 (1), O145:H34 (1), O33:H6 (2), O2:H45 (1), Ont:H31 (1), Ont:H6 (1), O5/71:H31 (1), O98:H8 (1), O28:H45 (1), O55:H7 (2), O51:H49 (1), OR:H- (2), O126:H6 (1), Ont:H5 (1), O139:H14 (1), Ont:H40 (1), O124:H40 (1), O125:H6 (2), OR:H2 (1), O123:H45 (1); tEPEC, O2:H2 (1), O2:H34 (1), O2:H45 (2), O8:H9 (1), O15:HNT (1), O15:H2 (1), O25:H7 (1), O28:H45 (1), O33:H6 (2), O51:H49 (2), O55:HND (1), O55:H6 (1), O55:H7 (2), O71:H6 (1), O86:HNM (1), O98:H8 (1), O114:HNM (1), O114:H2 (2), O117:H2 (1), O118:HR (1), H119:HND (3), O119:H21 (1), O123:H45 (2), O124:H40 (1), O125:H2 (1), O125:H6 (1), O125:H19 (1), O125sc:HND (1), O126:H2 (1), O126:H6 (1), O127:H45 (1), O128:H2 (3), O139:H14 (1), O142:H6 (3), O161:H40 (1), O175:H- (2), O170/172:H49 (2); AAEC, O2:H25 (1), O2K1:H6 (1), O5:HNM (1), O5:H11 (1), O20:H7 (1), O51:H11 (1), O116:H6 (1), O132:H2 (1), O149:H32 (1), O153:H7 (1).

^bSequencing of PCR products showed a fusion between the espV gene and part of the pO55 plasmid.

^cPresence of the Stx toxin has not been tested.

^dNumber of sequenced espV amplicons.

^eNonmotile.

By pathotype, espV was found in 18/81 (22%) atypical EPEC (aEPEC) strains, 5/62 (8%) typical EPEC (tEPEC) strains, and 7/69 (10%) EHEC strains. espV was also found in 7/18 (38%) A/E E. coli (AAEC) strains in which the presence of the Stx toxin was not tested. Sequencing of representative espV amplicons revealed a high level of sequence homology to espV in EPEC E110019 (data not shown).

The role of EspV in C. rodentium colonization in vivo.

To determine if EspV plays a role in gut colonization by C. rodentium, C3H mice were challenged orally with wild-type and espV mutant C. rodentium strains. Stool samples were collected daily over a 10-day period, and the number of CFU per gram of stool was determined by plating them on LB plates containing the appropriate antibiotics. No significant difference in colonization was observed between the wild-type and the espV mutant strains (Fig. 2A). We next measured the competitive index by orally challenging BL6 mice with a mixed inoculum containing approximately 50% wild-type C. rodentium and 50% espV mutant strain. This revealed that espV did not confer a competitive advantage during intestinal colonization by C. rodentium (Fig. 2B).

FIG. 2.

(A) Colonization of C57BL/6 mice with derivatives of C. rodentium. The results are expressed as the mean log₁₀ CFU g feces⁻¹ for 7 mice per group taken at selected time points postinoculation. The mice were infected with approximately 2 × 10⁹ CFU of wild-type C. rodentium or the ΔespV mutant, as indicated. The error bars indicate standard errors. (B) CIs calculated from mixed-strain infections of C3H/He or C57BL/6 mice. Groups of at least 5 animals were inoculated by oral gavage with a 1:1 mixture of the ΔespV mutant ICC313 and wild-type C. rodentium. All mice were sacrificed 7 days postinoculation. The CIs of the different C. rodentium strains in the initial inoculum and the mouse colon at the time of sacrifice were determined by plating serially diluted samples on agar plates containing selective antibiotics. The mean CI of each group is indicated on the graph by a horizontal line.

Ectopic expression of EspV induces drastic alteration in cell morphology in vitro.

Ectopic expression of Myc-tagged espV from the mammalian expression vector pRK5 induced dramatic morphological changes in HeLa cells. After 24 h, transfected cells appeared rounded and exhibited condensed nuclei. Moreover, dendrite-like projections appeared all around the cell circumference in 90% of the transfected cells (Fig. 3). Transfection of Swiss 3T3 cells resulted in an equivalent phenotype (data not shown).

FIG. 3.

Ectopic expression of EspV induces HeLa cell rounding and formation of dendrite-like projections. (A) Myc-tagged EspV was ectopically expressed in HeLa cells for 24 h. Fluorescence microscopy of mock-transfected cells and two examples of EspV-transfected cells are shown with the morphological changes triggered in HeLa cells. Actin was stained with Oregon Green-phalloidin, the Myc tag antibody was detected with monoclonal antibody, and DNA was visualized with DAPI. (B) SEM of mock- or pRK5:espV-transfected HeLa cells. (C) Quantification of the EspV-induced phenotype in mock- and pRK5:espV-transfected cells. One hundred cells were counted in duplicate in three different experiments. The results are displayed as means ± standard errors of the mean. The numbers of rounded cells with dendrites were significantly different (*) in the mock-transfected cells and the cells expressing EspV. Significance was tested using an unpaired t test with a two-tailed P value (*, P < 0. 0001).

EspV translocated from EPEC induces morphological changes on HeLa cells.

Infection of HeLa or Swiss 3T3 cells with wild-type EPEC E110019 and C. rodentium did not reproduce the phenotype seen following ectopic expression of EspV. To further analyze the phenotype induced by EspV in the context of infection, espV was cloned into the bacterial expression vector pSA10. The resulting construct was transformed into the EPEC strain E2348/69, which is naturally espV negative and for which the infection dynamics are well characterized. Infection of HeLa cells with E2348/69 expressing EspV induced morphological changes similar to those seen upon ectopic expression, although the phenotype was less dramatic and was observed in only 50% of the infected cells (Fig. 4). The observed phenotypes were dependent on protein translocation, as they were not seen after infection of HeLa cells with E2348/69 ΔescN expressing EspV (data not shown). Similar to control cells infected with wild-type E2348/69, actin-rich pedestals were observed on rounded cells infected with EPEC overexpressing EspV (Fig. 4). Similar phenotypes were observed upon infection of Swiss 3T3 cells (data not shown).

FIG. 4.

Translocation of EspV from EPEC E2348/69 induces drastic morphological changes in HeLa cells. (A) Fluorescence and SEM of uninfected HeLa cells or cells infected with wild-type EPEC E2348/69 or E2348/69 expressing EspV for 1 h and then incubated with 50 μg/ml of gentamicin for 4 h. Actin was stained with Oregon Green-phalloidin and DNA with DAPI. Rounded cells with dendrite-like projections were observed in cells infected with E2348/69 expressing EspV, but not in the wild-type control. Both strains formed actin pedestals. (B) Quantification of HeLa cells with an EspV-induced phenotype after infection with E2348/69 or E2348/69 expressing EspV and uninfected cells. One hundred cells were counted in duplicate in three different experiments. The results are displayed as means ± standard errors of the mean. The number of cells with dendrites was significantly higher (*) in cells infected with E2348/69 expressing EspV than in uninfected cells and cells infected with wild-type E2348/69. Significance was tested using one-way ANOVA with Bonferroni correction (*, P < 0.0001).

Expression of EspV in yeast inhibits growth.

To test the effect of EspV on yeast, S. cerevisiae BY4741 was transformed with the pYES2 vector, which allows expression of proteins driven off the GAL-1 galactose-inducible promoter. Expression of EspV in yeast grown on plates containing minimal medium with galactose drastically inhibited yeast growth, while in the presence of glucose, growth was only slightly inhibited (Fig. 5A). In contrast, expression of the E2348/69 effector NleH1, which was used as a control, did not affect yeast growth under any conditions, similar to the empty pYES2 vector.

FIG. 5.

EspV is toxic in yeast and induces cell enlargement. (A) S. cerevisiae BY4741 transformed with pYES2, pYES2:espV, or pYES2:nleH1 was streaked on glucose- or galactose-containing minimal medium. BY4741 pYE2:espV growth is inhibited when EspV expression is induced in the presence of galactose. Experiments were performed in triplicate. (B) SEM of BY4741 transformed with pYES2, pYES2:espV, or pYES2:nleH1 grown for 24 h in minimal liquid medium containing galactose. Upon expression of EspV, yeast growth is arrested and the yeast morphology is changed, with enlargement of the cells. (C) Quantification of the area of growth for 24 h in minimal liquid medium containing galactose of BY4741 transformed with pYES2, pYES2:espV, or pYES2:nleH1. The yeasts were visualized using phase-contrast with a Zeiss Axioimager immunofluorescence microscope (40× objective, giving a total magnification of ×4,000), and the area of the yeast was measured using the Axiovision Rel 4.5 software. One hundred yeasts were measured in duplicate in three different experiments. The results are displayed as means ± standard errors of the mean. The average area of the yeast was significantly higher (*) in cells expressing EspV than in yeast harboring pYES2 or expressing NleH1. Significance was tested using one-way ANOVA with Dunnett's post hoc test, with yeast harboring pYES2 as a control (*, P < 0.01).

In order to determine if the EspV-induced growth arrest is reversible, BY4741 strains harboring pYES, pYES:nleH1, or pYES:espV were grown for 24 h in liquid medium in the presence of galactose and then plated on glucose-containing plates. After 72 h of incubation, only yeast containing pYES and pYES:nleH1 grew, while BY4741 containing pYES:espV exhibited no significant growth (data not shown). This result suggests that under these conditions the yeast cells are unable to recover from the EspV-induced growth arrest.

We next used SEM to visualize yeast expressing EspV. Incubation in minimal liquid medium for 24 h (Fig. 5B and C) or 12 h (data now shown) in the presence of galactose showed that yeasts expressing EspV were on average 50% bigger than cells harboring pYES2 or expressing NleH1.

DISCUSSION

T3SS are employed by many bacterial pathogens of plants and animals to inject effector proteins into the eukaryotic cell (19). The large numbers of effectors that have been identified and characterized in recent years reveal broad diversity in terms of both functions and mechanisms. In this study, we have demonstrated that EspV is a novel effector of A/E pathogens. EspV shows significant sequence homology with the C terminus of the P. syringae effector AvrA, which was the first effector found in a plant pathogen (49). Nonetheless, little is currently known about its function in plant pathogenesis. Several other orthologous T3SS effectors of animal pathogens have been found in plant pathogens. For example, the Yersinia pestis effector YopT and the P. syringae effector AvrPpHb are both cysteine proteases (47), EspJ of EPEC and EHEC is a homologue of HopF of P. syringae (15), and EspW of EHEC shows sequence similarity to the C-terminal domain of HopPmaA/HopW (50). These similarities suggest that the effectors target core signaling pathways or cell structures shared by plants and animals.

The genome sequences of A/E pathogens have revealed a conserved set of effectors that are found in all strains. These include all the effectors that are encoded on the LEE and a number of non-LEE effectors, including EspI/NleA, NleB, NleE, NleC, NleD, NleH, and NleG. Importantly, these studies also show that other effectors have variable distributions. For example, EspM, TccP/TccP2, and EspT were found in 51%, 47%, and 1.8% of the tested strains (1, 23). This diversity in the effector protein repertoire suggests the existence of alternative infection strategies and indicates that EPEC and EHEC strains cannot be considered a homogeneous group. In this study, we screened for the presence of espV in a collection of 230 A/E isolates. Overall, we found espV in 16% of the tested strains; in particular, it was found in 22%, 8%, and 10% of the atypical EPEC, typical EPEC, or EHEC strains, respectively.

In order to determine if EspV played a role in colonization, we tested a C. rodentium ΔespV mutant in both single and mixed infections. We found that EspV did not contribute to colonization or fitness in vivo. It is important to note that the C. rodentium model is limited, as it only provides information regarding the role of effectors in colonization. Indeed, in the past, the C. rodentium model failed to reveal a phenotype for a number of important effectors, including NleE, which affects NF-κB signaling, and EspT, involved in bacterial invasion (29, 17). This may be due to functional redundancies with other effectors and does not rule out a novel function for EspV.

To gain insight into the biochemical function of EspV, we expressed the effector in mammalian cells. Two different approaches were used to determine the effect of EspV: ectopic expression and translocation from an EPEC strain, E2348/69, that is naturally espV negative. HeLa and Swiss 3T3 cells transfected with EspV presented a striking phenotype of rounded cells, condensed nuclei, and formation of actin-rich dendrite-like structures. A similar, although less drastic, phenotype was observed upon delivery of EspV from EPEC E2348/69. As far as we know, the phenotype triggered by EspV has not been seen for any other T3SS effector to date. While investigating the signaling pathway targeted by EspV, we found no significant difference upon infection with E2348/69 pSA10:espV of HeLa cells previously transfected with the dominant-negative form of RhoA, Cdc42, or Rac1 (data not shown). These results suggest that the formation of dendrite-like structures by EspV is more likely a result of cell rounding/shrinking than of induction of actin polymerization. Although at present the mechanism by which EspV triggers such a dramatic phenotype is unknown, it somewhat resembles morphological changes seen in HeLa cells during the prophase stage of mitosis. However, expression of EspV led to nuclear condensation distinct from the typical structure of prophase nuclei, and at the moment, the link to mitosis is purely speculative.

To investigate further the function of EspV, we used yeast as an alternative eukaryotic model that is recognized as a powerful system to study the functions of bacterial effectors (32). We found that EspV induces irreversible growth arrest when expressed in the S. cerevisiae strain BY4741. To gain insight into the mechanism of toxicity, we looked at the morphology of yeast cells expressing EspV by SEM. Remarkably, we found that EspV induces enlargement of the yeast before arresting growth. This phenotype has been linked previously to the inactivation/overexpression of certain yeast genes. For example, the aac3 gene encodes a mitochondrial inner membrane ADP/ATP translocator which upon inactivation induces abnormal cell cycle progression and increased cell size (31, 42). Likewise, overexpression of BIK1 and UPC2 triggers similar phenotypes of large and very large cells, respectively (48). BIK1 is a microtubule-associated protein and a component of the interface between microtubules and the kinetochore that is involved in sister chromatid separation (4). UPC2 is a sterol regulatory element (13). The precise pathway modified in yeast by EspV is yet to be elucidated. We have already found that, consistent with the dominant-negative results in mammalian cells, the equivalent of the small-GTPase RhoA pathway in yeast is not essential for EspV activity, as EspV retained its toxicity when expressed in the BY4147 Δrlm1, Δslt2, and Δbck1 mutants of the equivalent Rho1p/MAPK signaling pathway (data not shown). In order to determine the pathway targeted by EspV in yeast, we intend to screen for a suppressor by transforming an array of all the other viable yeast deletion mutants with EspV.

In this work, we identified and characterized a new T3SS effector of the A/E pathogens, EspV. This effector triggers striking morphological changes when expressed in mammalian cells and induces morphological changes and growth arrest in yeast. Elucidating the function and the mechanism of action of EspV is a challenge for future studies.