The enterohemorrhagic Escherichia coli effector protein NleF binds mammalian Tmp21

2.1. Chemicals and antibodies

Chemicals and antibodies were used according to manufacturer’s recommendations. GFP and Tmp21 antibodies were obtained from Cell Signaling. Golgin-97 antibody was obtained from Invitrogen. Calnexin, HA, and His antibodies were obtained from Sigma. NleF antiserum was described previously (Echtenkamp et al., 2008). pEGFP-VSVG was provided by J. Lippincott-Schwartz [Addgene #11912 (Presley et al., 1997)].

2.2. Bacterial strains, cell culture, and infection experiments

The bacterial strains and plasmids used in this study are described in Table 1. HeLa cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum. Cells were transfected using TransPass (New England Biolabs). Bacteria were cultured in Luria-Bertani broth at 37 °C for 18 h without shaking. Overnight LB cultures were diluted 1:10 into DMEM, followed by a further incubation for 3 h at 37 °C, 5% CO₂. Cell culture medium was replaced with DMEM prior to infection and bacteria were added at a multiplicity of infection of 25–50.

2.3. Protein purification

NleF and Tmp21 were cloned into pFLAG-CTC and pET28a, respectively, and expressed in E. coli BL21(DE3). Bacterial cultures were grown to an OD₆₀₀ of 0.3 and then induced with 1 mM IPTG for 2 h. Cells were centrifuged, lysed by sonication, applied to α-FLAG M2 beads and Ni-NTA agarose, respectively, and incubated at 4 °C overnight. After washing, proteins were eluted with either 0.1 M glycine HCl, pH 3.5, or with imidazole and then analyzed on SDS-12% PAGE.

2.4. Yeast two-hybrid assay

NleF was cloned into the yeast two-hybrid GAL4 DNA binding domain vector pGBKT7 to generate the ‘bait’ plasmid. The NleF bait was used to screen a pre-transformed human HeLa cell cDNA library for proteins interacting with NleF according to protocols in the BD Matchmaker Pre-transformed Libraries User Manual (Clontech). Yeast clones containing library plasmids encoding human proteins interacting with NleF were purified by restreaking on selective media and retested for growth phenotypes. β-Galactosidase activity was calculated using Eq. (1) where t refers to the incubation time (min) and v refers to the concentration factor.

2.5. Pull down assay

Purified proteins (~20 μg) were applied to α-FLAG M2 beads and incubated with rotation for 5 h at 4 °C. Beads were washed three times with PBS and resuspended in SDS-PAGE buffer. The samples were interrogated for the presence of His- and FLAG-tagged proteins by immunoblotting.

2.6. Immunoblotting

Cells were lysed in RIPA buffer [150 mM NaCl, 50 mM Tris pH 8.0, 0.5% sodium deoxycholate, 0.1% SDS, 1% Nonident P-40], incubated on ice for 30 min, and centrifuged. Proteins were resolved by SDS-PAGE, transferred to nitrocellulose membranes, blocked in Odyssey blocking buffer (Li-Cor) and then probed with appropriate primary and secondary antibodies. After rinsing in PBS, blots were imaged using an Odyssey infrared imaging system.

2.7. Bimolecular fluorescence complementation

HeLa cells were co-transfected with two BiFC plasmids (250 ng each) representing NleF and Tmp21 sequences cloned as fusions to the N- or C-terminus of Venus eYFP (designated VN and VC). The fluorescence derived from BiFC was visualized using Eclipse 80i fluorescence microscope (Nikon) after 24 h incubation and was quantified using a fluorescence plate reader with appropriate filters (excitation: 500/20 nm; emission: 535/30 nm).

2.8. VSVG trafficking and immunofluorescence microscopy

HeLa cells were grown on glass coverslips in 24-well tissue culture plates. For VSVG experiments, cells were transfected in 4 replicates with pEGFP-VSVG in the presence or absence of NleF-HA. After 24 h, one replicate was left at 37 °C while the other replicates were transferred to 19 °C, 32 °C, or 40 °C. Three hours after temperature shift, cells were rinsed 3 times with PBS, fixed in 3.7% formaldehyde and permeabilized in 0.2% saponin in PBS, blocked with 10% goat serum, and incubated with primary antibodies for 1 h at room temperature. Cells were washed with PBS and probed with Alexa Fluor-conjugated secondary antibodies for 1 h. Coverslips were mounted in Mowiol and samples were visualized using Eclipse 80i fluorescence or Eclipse C1Si confocal microscopes (Nikon).

2.9. Statistical analyses

β-Galactosidase data were analyzed using t-tests. BiFC data were analyzed using one-way ANOVA. VSVG trafficking data were evaluated with Fisher’s exact test. p-Values < 0.05 were considered significant.

3.1. NleF binds to Tmp21

We used a yeast two-hybrid (Y2H) screening assay to determine the mammalian binding partners of NleF. EHEC NleF was cloned as a fusion to the Gal4 DNA-binding domain (DBD) and co-expressed in the yeast reporter strain AH109 with a HeLa cell cDNA library. Co-transformants were plated on synthetic quadruple-dropout (QDO) medium to select for interactions between NleF and a library component that activated the transcriptional reporter. We isolated 153 colonies after 21 days growth on QDO plates, 26 of which had a high level of reporter gene activation as measured in β-galactosidase assays (data not shown).

We identified 3 different human cDNA sequences (Tmp21, CD151, PAIP2) that may encode proteins that interact with NleF. Among these, we further studied Tmp21 (also named p23/p24d) as a putative NleF binding partner. We confirmed the Tmp21–NleF interaction using direct Y2H assays and by quantifying β-galactosidase activity resulting from NleF–Tmp21 co-expression (Fig. 1A).

Tmp21 is a 219 amino acid integral type I transmembrane protein that functions as an integral receptor for the COPI-vesicle coat (Blum et al., 1999). Tmp21 is a member of the p24 (p24/gp25L/emp24/Erp) protein family. These proteins provide cargo receptors to proteins (Anantharaman and Aravind, 2002) and regulate protein packaging into COPI vesicles in concert with a small GTPase, the ADP-ribosylation factor 1 (Arf1). p24 proteins are assembled into heteromeric complexes that cycle between the ER and the Golgi and recruit Arf1 in early stages of vesicle formation (Gommel et al., 2001). p24 proteins thus play active roles in retrograde protein transport from Golgi to ER by facilitating the formation of COPI-coated vesicles (Aguilera-Romero et al., 2008).

By expressing and purifying recombinant forms of NleF and Tmp21 in E. coli BL21(DE3) and then using these proteins in pulldown assays, we confirmed that NleF and Tmp21 bind directly in vitro (Fig. 1B). We also used immunofluorescence microscopy to determine the extent of NleF–Tmp21 colocalization. To do this, we used polyclonal NleF antisera (Echtenkamp et al., 2008) to detect NleF after its translocation into HeLa cells during C. rodentium infection (Fig. 1C). Co-staining for Tmp21 revealed that both proteins colocalized in a perinuclear location (Fig. 1D).

3.2. BiFC

To determine whether NleF and Tmp21 interact when they are co-expressed in mammalian cells, we used bimolecular fluorescence complementation (BiFC) assays. This technology utilizes the reconstitution of two fragments of the enhanced yellow fluorescent protein (eYFP) to generate a fluorescent signal at the site of the protein–protein interaction under investigation (Hu et al., 2002). We generated protein chimeras with split N- and C-terminal fragments (VN and VC, respectively) of eYFP fused to either NleF or Tmp21 (Fig. 2A). Co-expressing eYFP chimeras of Tmp21 and NleF reconstituted YFP fluorescence, to a similar magnitude as the reconstitution of the actin positive control (Fig. 2B). By contrast, transfecting individual plasmids did not reconstitute YFP fluorescence (Fig. 2B), suggesting that NleF binds to Tmp21 in mammalian cells.

The N-terminal luminal domain of Tmp21 mediates cargo uptake into transport vesicles, whereas the KKLIE cytoplasmic tail at the Tmp21 carboxy-terminus mediates COPI-dependent transport vesicle formation (Blum and Lepier, 2008). To map the binding domain of NleF on Tmp21, we carried out a BiFC study with C-terminal deletions of NleF. These experiments revealed that deleting the NleF C-terminus beyond amino acid 84 eliminated NleF binding to full length Tmp21, as indicated by loss of fluorescence with the sequentially truncated constructs (Fig. 2C). A similar BiFC analysis revealed that the C-terminal region of Tmp21, amino acids 63–180, was required for binding to NleF (Fig. 2D). Overall, these data suggest that NleF and Tmp21 associate through their respective C-termini.

3.3. NleF alters VSVG trafficking

We tested the hypothesis that NleF binding to Tmp21 would cause defects in protein trafficking by characterizing the localization of a vesicular stomatitis virus glycoprotein (VSVG)-GFP fusion as a function of NleF expression and of temperature. VSVG localization is commonly used to study mammalian protein trafficking (Wessels et al., 2005). This glycoprotein is essential for viral envelope fusion with the host PM and traffics intracellularly via the ER and Golgi (Lippincott-Schwartz et al., 2000).

VSVG-GFP transport can be manipulated by incubating cell cultures at different temperatures (Toomre et al., 1999). If incubated at 40 °C, VSVG-GFP becomes reversibly misfolded and retained in the ER. If shifted to 19 °C, VSVG-GFP refolds and can be transported to the TGN. A further temperature shift to 32 °C allows subsequent trafficking to the PM (Fig. 3A).

We first transfected HeLa cells with VSVG-GFP and then evaluated VSVG-GFP localization using immunofluorescence microscopy. As expected, when cells were shifted from 37 °C to 40 °C, VSVG-GFP fluorescence became localized primarily in the ER, as shown by its colocalization with the ER protein calnexin (Fig. 3B). When cells were subsequently shifted to 19 °C, VSVG-GFP localization with calnexin was reduced and VSVG-GFP was primarily redistributed to the Golgi, as shown by its colocalization with the Golgi protein golgin-97 (Fig. 3C). Incubating cells at 32 °C also resulted in the expected subsequent shift of VSVG-GFP from the Golgi to the PM.

We predicted that if NleF disrupts Tmp21 function, VSVG-GFP redistribution to the Golgi would either be delayed or blocked by NleF. After transfecting NleF and performing temperature shift experiments, we observed that VSVG-GFP was retained in the ER at 19 °C in 74% of cells examined, rather than redistributing to the Golgi (p < 0.05, Fisher’s exact test; Fig. 3D). Similarly, in the presence of NleF, VSVG-GFP was mislocalized to the Golgi at 32 °C in 58% of cells examined, rather than trafficking to the PM (p < 0.05, Fisher’s exact test; Fig. 3E).