Quantitative Mass Spectrometry Identifies Novel Host Binding Partners

Mar 28, 2016 - Quantitative Mass Spectrometry Identifies Novel Host Binding Partners for Pathogenic Escherichia coli Type III Secretion System Effecto...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/jpr

Quantitative Mass Spectrometry Identifies Novel Host Binding Partners for Pathogenic Escherichia coli Type III Secretion System Effectors Robyn J. Law,†,‡ Hong T. Law,# Joshua M. Scurll,⊥ Roland Scholz,† Andrew S. Santos,†,‡ Stephanie R. Shames,†,‡ Wanyin Deng,† Matthew A. Croxen,† Yuling Li,† Carmen L. de Hoog,∥ Joris van der Heijden,†,‡ Leonard J. Foster,∥ Julian A. Guttman,*,# and B. Brett Finlay*,†,‡,§ †

Michael Smith Laboratories, ‡Department of Microbiology and Immunology, §Department of Biochemistry and Molecular Biology, Centre for High-Throughput Biology, and ⊥Department of Mathematics, University of British Columbia, Vancouver, British Columbia, Canada V6T 1Z4 # Department of Biological Sciences, Simon Fraser University, Burnaby, British Columbia, Canada V5A 1S6 ∥

S Supporting Information *

ABSTRACT: Enteropathogenic and enterohemorrhagic Escherichia coli cause enteric diseases resulting in significant morbidity and mortality worldwide. These pathogens remain extracellular and translocate a set of type III secreted effector proteins into host cells to promote bacterial virulence. Effectors manipulate host cell pathways to facilitate infection by interacting with a variety of host targets, yet the binding partners and mechanism of action of many effectors remain elusive. We performed a mass spectrometry screen to identify host targets for a library of effectors. We found five known effector targets and discovered four novel interactions. Interestingly, we identified multiple effectors that interacted with the microtubule associated protein, ensconsin. Using co-immunoprecipitations, we confirmed that NleB1 and EspL interacted with ensconsin in a region that corresponded to its microtubule binding domain. Ensconsin is an essential cofactor of kinesin-1 that is required for intracellular trafficking, and we demonstrated that intracellular trafficking was severely disrupted during wild type EPEC infections but not during infections with ΔnleB1 or ΔespL mutants. Our findings demonstrate the efficacy of quantitative proteomics for identifying effector−host protein interactions and suggest that vesicular trafficking is a crucial cellular process that may be targeted by NleB1 and EspL through their interaction with ensconsin. KEYWORDS: quantitative mass spectrometry, SILAC, effector proteins, ensconsin, MAP7, enteropathogenic E. coli, enterohemorrhagic E. coli, type III secretion system



INTRODUCTION Enteropathogenic and enterohemorrhagic Escherichia coli (EPEC and EHEC, respectively) are members of a group of virulent bacteria known as attaching and effacing (A/E) pathogens. EPEC causes severe diarrhea in infants that can be fatal,1 and EHEC is associated with hemorrhagic colitis and, in some cases, hemolytic uremic syndrome, which can result in renal failure and death.2 EPEC and EHEC employ a similar pathogenic strategy in which they intimately attach to intestinal surfaces and subvert a variety of host cellular processes while remaining extracellular. Upon host cell contact, EPEC and EHEC use a type III secretion system (T3SS) to deliver a © XXXX American Chemical Society

number of proteins into host cells that co-opt cytoskeletal and signaling molecules as part of their disease process.3,4 These effector proteins are central to pathogenesis because T3SS mutants are often severely hindered in their ability to colonize their hosts. Accordingly, techniques to identify effectors and characterize their interactions within host cells are paramount to understanding the disease mechanisms exploited by A/E pathogens. In particular, intracellular vesicle trafficking has gained recent interest as an important target for Received: January 27, 2016

A

DOI: 10.1021/acs.jproteome.6b00074 J. Proteome Res. XXXX, XXX, XXX−XXX

Journal of Proteome Research



some bacterial and viral pathogens as part of their pathogenic strategy.5 Intracellular trafficking is a critical process involved in maintaining cellular homeostasis, promoting cell survival, and responding to the presence of pathogens. Directed delivery of transport vesicles requires the concerted action of a number of cargo adaptors, molecular motors, and their cytoskeletal tracks, which presents pathogenic microbes with several targets against which to mount an efficient attack. Over the past several years, great strides have been made in identifying the host targets and functions of many T3SS effectors, but due to the redundant and interdependent nature of these proteins,3 a comprehensive understanding of the molecular mechanisms of A/E pathogen infection remains in progress. For this reason, large-scale screens to identify novel host interaction partners for bacterial virulence proteins are of great interest in A/E pathogen research. Quantitative proteomics methods using SILAC (stable isotope labeling by amino acids in cell culture) have made it possible to effectively analyze the type III secretomes of EPEC and the related A/E pathogen Citrobacter rodentium,6,7 but effector identification is only the first step in understanding A/E pathogen-mediated disease. A recently conducted yeast twohybrid screen revealed novel targets for a number of EHEC T3SS effectors.8 However, the use of yeast rather than human cell types may preclude the detection of effector−host protein interactions requiring cofactors or post-translational modifications found only in human cells. Our laboratories first used the SILAC method to identify Cdc42 as a binding partner for the Salmonella T3SS effector, SopB,9 and subsequently used this technique to identify host targets for a library of Salmonella Typhimurium-secreted effectors, many of which have been confirmed through independent analyses.10 The SILAC method uses differential labeling of experimental and control cell populations to distinguish between specific and nonspecific protein−protein interactions. The relative abundances of host proteins in complex with experimental baits and controls are then inferred from the ratios of the differentially labeled peptides from each cell population, thereby allowing the exclusion of nonspecific interactions occurring in equal abundance in both populations.11 Characterization of protein−protein interactions between bacterial effectors and host receptors is a critical step in understanding the physiological significance of T3SS proteins during infection. In this study, we used SILAC paired with liquid chromatography-coupled tandem mass spectrometry (LC−MS/MS) to quantitatively identify host interaction partners for a number of EHEC T3SS effectors. Using this methodology, we were able to generate a strong list of host protein-effector pairs. Our screen confirmed five previously published interactions and identified four novel high-confidence host protein-effector pairs. Interestingly, three of the newly identified interactions involved the microtubule-associated protein (MAP) ensconsin (henceforth referred to as MAP7), which has recently been identified as an essential cofactor of kinesin-1.12,13 We further confirmed and characterized two of the MAP7-effector interactions using co-immunoprecipitations (CoIP) and explored their functions within the host using infection-based analyses. Our data further demonstrate the value of using SILAC-based proteomics as a tool to specifically identify host−pathogen protein−protein interactions and reveal a potentially crucial and previously unknown role for MAP7 during A/E pathogen infections.

Article

EXPERIMENTAL PROCEDURES

Tissue Culture and Infection Conditions

HEK293T cells (ATCC) were maintained in Dulbecco’s Modified Eagle Medium (DMEM) (Thermo Scientific) supplemented with 10% fetal bovine serum (FBS) (Thermo Scientific), 1% Glutamax (Gibco), 1% nonessential amino acids (NEAA) (Gibco), and 0.1 units/L penicillin−streptomycin. Ptk2 cells were maintained in normal growth media (MEM/F12 (1:1)) + 10% FBS (Gibco). For routine cloning and infections, bacteria were grown in lysogeny broth (LB) at 37 °C supplemented with appropriate antibiotics. The final antibiotic concentrations were as follows: ampicillin, 100 μg/mL; kanamycin, 50 μg/mL; streptomycin, 50 μg/mL; and chloramphenicol, 30 μg/mL. Molecular Cloning

The gene sequences encoding the conserved EHEC T3SS effectors Tir, Map, EspB, EspG, EspJ, EspL2 (referred to as “EspL” to correspond to EPEC EspL), EspO, EspW, EspX, EspY, NleA, NleB1, NleB2, NleC, NleD, NleE, and NleG14 were amplified from EHEC O157:H7 strain EDL933 genomic DNA using Elongase Enzyme Mix (Life Technologies). For bacterial expression of effectors, gene sequences were cloned inframe into pGEX-6P-3 (GE Lifesciences) with an upstream triple HA tag (lab vector; pGEX-6P-3::HA3). The map7 gene sequence was amplified from human cDNA clone SC322593 (OriGene Technologies, Inc.) and cloned into pCMVTag4A (Stratagene) to generate pmap7-FLAG. MAP7 truncations were amplified using pmap7-FLAG as a template and inserted into pCMVTag4A. SILAC Labeling and Analysis

For SILAC labeling, the normal growth media for HEK293T cells (DMEM supplemented as described in the Tissue Culture and Infection Conditions section) were replaced with lysineand arginine-free DMEM (Caisson Laboratories Inc.) containing 10% dialyzed FBS, 1% Glutamax, 1% NEAA, 0.1 units/L penicillin−streptomycin, and either 36.5 mg/L 2H4-lysine and 21 mg/L 13C6-arginine (Cambridge Isotope Laboratories) for heavy cells or normal isotopic abundance L-lysine and Larginine (Sigma-Aldrich) for light cells. Cells were maintained in labeling media for at least five cell divisions to ensure complete labeling prior to experimentation. Light- and heavy-labeled cells were collected in NP40 lysis buffer [20 mM Tris−HCl, pH 7.5; 150 mM NaCl; 1% NP40; 10 mM Na−pyrophosphate; 50 mM NaF; 1 mM Na3VO4; and cOmpleteTM Protease Inhibitor Cocktail (Roche)] and centrifuged at 16 000 r.c.f. Protein concentrations were determined using Coomassie Plus Bradford Assay Reagent (Thermoscientific), and 10 mg lysate was transferred to tubes containing Glutathione Sepharose (Sigma-Aldrich) bound to 20 pmol of recombinant GST-HA3-tagged effector or GST-HA3 as a control. Lysates from heavy-labeled cells were mixed with GST-HA3-tagged effector, and lysates from unlabeled “light” cells were mixed with GST-HA3. Tubes were incubated with rotation for 1 h at 4 °C followed by centrifugation at 2100 r.c.f. for 5 min. Beads were washed with cold lysis buffer followed by cold PBS and incubated in 100 μL of 50 mM Tris (pH 9.5) plus 10 mM glutathione overnight at 4 °C with rotation. Tubes were centrifuged at 2100 r.c.f. for 5 min, and supernatants were separated from beads using GELoader Tips (Eppendorf). Supernatants from like samples were pooled, and eluted proteins were precipitated using ethanol−acetate and digested B

DOI: 10.1021/acs.jproteome.6b00074 J. Proteome Res. XXXX, XXX, XXX−XXX

Article

Journal of Proteome Research

Table 1. Effector Proteins and Host Interaction Partners Identified Through SILAC-Based LC−MS/MS and Previously Reported Targets with Source Citations effector

references

NHERF2,a NHERF1, Cdc42, rhophilin-1

EspJ

*WD repeat-containing protein 23 (WDR23), Src, centromere protein H, intraflagellar transport protein 20 homologue, MORF4 familyassociated protein 1-like 1, synembryn-A *ensconsin (MAP7), annexin-2 *ensconsin (MAP7) Sec23A, Sec24B, Disks large homolog 1 (DLG1), Down syndrome critical region protein 4, FERM domain-containing protein 3, proenkephalin-A, protein tyrosine phosphatase type IVA 1, α1 syntrophin, MALS3, PDZK11, SNX27, TCOF1 *ensconsin (MAP7); GAPDH; TRADD; FADD; RIPK1; developmentally regulated GTP-binding protein 2; leucine-rich repeatcontaining protein 18; DNA-directed RNA polymerases I, II, and III subunit RPABC1 P300, p50, p65

EspL EspX NleA NleB1 NleC a

interaction partner(s)

Map

8,22,23,46, 47 8,48 41 8,24−26 8,36,37,49 18,50,51

Bold: interactions identified in our screen. Asterisk: novel interactions identified in our screen.

in-solution as described previously.15 The resulting peptides were acidified and analyzed by liquid chromatography−tandem mass spectrometry using an LTQ-OrbitrapXL. The LTQOrbitrapXL was set to acquire a full-range scan at 60 000 resolution from 350 to 1500 Th in the Orbitrap and to simultaneously fragment the top five multiply charged ions in each cycle in the LTQ. Once fragmented, a precursor was added to the dynamic exclusion list for 180 s. The relative collision energy was 35.0. Charge state screening parameters were set as follows: charge state screening, enabled; monoisotopic precursor selection, enabled; charge state rejection, enabled for 1+ charge state and unknown charge state; all other charge state screening parameters were disabled. Peptides and proteins were identified using Mascot (v2.2, Matrix Science) to search against the human international protein index (IPI) database with common contaminants and bacterial (Salmonella and E. coli) type III effectors added, along with concatenated reversed sequences (v3.69, 154 788 sequences) using the following criteria: electrospray ionization-ion trap fragmentation characteristics and tryptic specificity with up to one missed cleavage; 10 ppm and 0.6 Da accuracy for MS and MS/MS measurements, respectively; cysteine carbamidomethylation as a fixed modification; N-terminal protein acetylation, methionine oxidation, and [2H4]Lys and [13C6]Arg SILAC selected as variable modifications in the Mascot search. Peptide and protein search results were then extracted from the Mascot.dat file using MSQuant.16 Proteins with at least two peptides identified in each biological replicate were inspected further and the SILAC ratios extracted. At this stage, any potential interfering or overlapping isotope envelopes were identified by the further inspection of proteins with a high coefficient of variation among its individual peptide ratios, and peptides with interfering peaks in the spectra were discarded if the interference could not be resolved. At least two unique peptides were required to consider a protein identified. The protein identifications met a 1% false discovery rate, as calculated by the standard target-decoy approach. SILAC ratios were extracted using MSQuant,16 and individual ratio assignments for each peptide from all proteins with a coefficient of variation greater than 50% were inspected manually to identify potential interferences. Ratios for proteins were calculated from the weighted averages of all peptides identified for each protein as implemented in MSQuant.16 Average ratios for each protein were then calculated across all biological replicates for each effector as follows: for EspB, EspO, EspW, EspY, NleA, NleB2, and NleD (two biological replicates); for Map, EspJ, EspL, and EspX (three biological replicates); for EspG and NleC (four

biological replicates); for Tir, NleB1, and NleE (five biological replicates); and for NleG (seven biological replicates). These values were then used as a general screen. From the earlier work on Salmonella and this work, the average ratio for all proteins identified (not including common contaminants that would have any heavy SILAC labels) were roughly normally distributed around 0 (log-transformed values). A value of 2 was chosen as a somewhat arbitrary cutoff, although in all experiments, it exceeded 2σ (two standard deviations). For GST pull-down experiments, proteins had to have an average ratio greater than 2.0 and be observed in at least two of the replicates for a given experiment to be considered further as a potential interactor. All proteins meeting these ratio criteria were further examined to determine which were likely to be specific interactors using the criteria established previously for Salmonella effector interactors.10 Anything that was not the bait itself, that was not a common serum protein, and that was observed in fewer than five different experiments (and was thus unlikely to be a common contaminant) was considered to be a specific interactor and included in Table 1. Heat shock proteins and other chaperones were observed frequently and are very likely to fit the biochemical definition of specific but were not included in Table 1 because they were considered to be mostly likely binding to foreign or improperly folded effector molecules. Notes for all proteins in all experiments can be found in Supplementary Data 1. All raw data, MSQuant output, and Mascot search results have been deposited in ProteomeXchange (submissions PXD002933 and PXD002995). Recombinant Protein Expression and Purification

E. coli BL21 DE3 was transformed with pGEX-6P-3::HA3 to generate a GST-HA3 control or with derivatives of pGEX-6P3::HA3 containing effector sequences and grown with shaking (225 rpm) at 37 °C for ∼3 h. Cells were induced with 1 mM IPTG, and growth was continued for 16 h at 16 °C. Cells were harvested in lysis buffer [phosphate buffered saline (PBS), pH 7.4, 1 mM EDTA, 1 mM DTT, 10 μg/mL DNase, 10 μg/mL RNase, and EDTA-free cOmpleteTM Protease Inhibitor Cocktail (Roche)] and passed through a French pressure cell at 10000 psi. Lysates were clarified by centrifugation for 10 min at 8000 r.c.f. followed by 30 min at 30000 r.c.f. Clarified lysates were added to glutathione−agarose beads (Sigma-Aldrich) and batch-bound with rotation for 1 h at 4 °C. Bead−lysate mixtures were transferred to Bio-Rad Polyprep Chromatography Columns, and columns were washed successively with 20 mL of PBS + 0.1% Triton X-100, 20 mL of PBS + 0.05% Triton X-100 + 0.5 M NaCl, and 20 mL of PBS. Proteins were eluted C

DOI: 10.1021/acs.jproteome.6b00074 J. Proteome Res. XXXX, XXX, XXX−XXX

Article

Journal of Proteome Research

of CO2. Images were acquired with continuous shooting at 200 ms/frame for ∼200 frames using Metamorph v7.8.2 (Molecular Devices) and analyzed with u-track multiple-particle tracking software.21 Prior to detection and tracking, a background estimate was subtracted from each raw movie frame. Each frame was passed through a difference of Gaussians filter, and a suitable intensity threshold was applied to the filtered images. The threshold was at least eight standard deviations in absolute pixel intensity above the mean absolute pixel intensity and guaranteed that no more than the brightest 0.2% of pixels were retained after its application. Tracks were obtained for detected particles in at least three representative cells per infection condition, and particle speeds were calculated for all tracks with at least 10 visible track segments as follows: for any given track with ≥10 visible track segments, the lengths of all visible track segments (in pixels) were added together and divided by the total number of track segments (frames) for which the particle was detected to give an average particle speed for each track in pixels per frame, which was then converted to μm/s. Mean squared displacements (MSDs) were calculated for all particles with tracks persisting ≥10 frames based on five-frame intervals. For each relevant track, the particle’s scalar displacement was calculated for every five-frame interval and then averaged to give the MSD per five frames.

with 10 mM glutathione in 50 mM Tris (pH 9.5) and dialyzed using Slide-A-Lyzer dialysis cassettes (Pierce Scientific). Transfection

HEK293T cells were seeded in 10 cm dishes at a density of ∼1 × 106 cells per dish. The following day, cells were transfected using the calcium phosphate method, and protein expression was allowed to proceed for 48 h. Lysates were then collected using NP40 lysis buffer, and protein concentrations were determined using Coomassie Plus Bradford Assay Reagent (Thermoscientific). Co-immunoprecipitation

HEK293T cells were transiently transfected with pmap7-FLAG constructs as described above. Lysates were collected 48 h post transfection in NP40 lysis buffer and centrifuged at 16000 r.c.f. for 30 min at 4 °C. A total of 70 μL of Protein G Sepharose 4 Fast Flow (GE Healthcare) was used to preclear samples containing 3 mg of total protein mixed with 2 μg of purified effector. Cleared supernatants were incubated with either 1 μg of mouse-α-FLAG M2 (Sigma-Aldrich) or mouse-α-HA (clone 2C16, homemade) and 40 μL of Protein G Sepharose. Beads were pelleted by centrifugation at 4 °C, washed with lysis buffer, and resuspended in 30 μL SDS-PAGE loading buffer.



Western Blotting

Samples were separated on 12% polyacrylamide gels and transferred to Pure Nitrocellulose (Bio-Rad Laboratories, Inc.) using a wet transfer cell. Membranes were blocked with 5% w/v skim milk powder in Tris-buffered saline containing 0.1% Tween 20 (TBST). The same buffer was used to dilute primary and secondary antibodies prior to incubation with nitrocellulose membranes: from a 1 mg/mL stock, rat-α-HA (Roche), 1:2500; mouse-α-FLAG M2 (Sigma-Aldrich), 1:2500; goat-α-rat HRP or goat-α-mouse HRP (Sigma-Aldrich), 1:5000; Membranes were treated with ClarityTM Western ECL Substrate (Bio-Rad Laboratories, Inc.) for chemiluminescent developing.

RESULTS

Identification of Host Proteins Targeted by EHEC Secreted Effectors

We have previously used SILAC-labeled HEK293T cells to identify host proteins targeted by several Salmonella secreted effectors.10 To further capitalize on this method, we used SILAC combined with LC−MS/MS to identify host proteins that specifically bound to 17 purified, recombinant EHEC effectors. Using an existing library of EHEC effector gene sequences cloned into pGEX-6P-3::HA3, each effector protein was purified from E. coli BL21 DE3 and incubated with lysates from HEK293T cells labeled with heavy amino acids while unlabeled cell lysates were incubated with purified GST-HA3 as a control. Experimental and control lysates were mixed and protein complexes were isolated by GST-pulldown prior to trypsinization and LC−MS/MS analysis. Using this strategy, we characterized high-confidence interactions as having an average heavy-to-light (H/L) SILAC ratio of greater than 2.0, and proteins that nonspecifically interacted with the affinity tags or resin displayed equal peak intensities of their heavy or light isotopic forms, yielding H/L ratios smaller than 2.0. Following our analysis of all 17 effectors, we identified nine highconfidence interaction pairs (Table 1 and Supplementary Data 1). Of these high-confidence interactions, five confirmed previously published effector-host protein binding partners (Map-NHERF2, NleA-Sec23A, NleA-Sec24B, NleA-DLG1, and NleC-p300).8,18,22−26 The remaining four novel effector−host protein pairs uncovered two different host targets: the transcriptional regulator, WDR23,27 which interacted with EspJ, and microtubule associated protein 7 (MAP7), which interacted with NleB1, EspL, and EspX (Table 1). Given that three of the four novel interactions we identified involved MAP7, our results suggested that MAP7 may be an important and previously unappreciated target of these enteric pathogens.

Generation of EPEC Effector Deletion Mutants

In-frame deletion of the nleB1, nleB2, and espL genes was carried out using sacB-based allelic exchange and the sucrose selection method. The suicide plasmid, pRE112,17 was used to generate in-frame effector deletion mutants as previously described.18 The pRE112 deletion plasmids were transformed into E. coli donor strain SM10λpir19 or MFDpir20 for conjugation with recipient strain E2348/69 to create ΔnleB1, ΔnleB2, ΔespL, and ΔnleB1ΔnleB2. After sucrose selection, transconjugants that were resistant to sucrose and sensitive to chloramphenicol were screened by colony PCR and verified for chromosomal deletion by sequencing at the Nucleotide and Protein Sequencing (NAPS) Unit at the University of British Columbia (UBC). Live Imaging and Analysis

Ptk2 cells were seeded onto glass coverslips in 12-well plates containing MEM/F-12 (1:1) + 10% FBS (Gibco). Following overnight incubation, cells were transfected with 2 μg of plasmid DNA using jetPEI transfection reagent (Polyplus Transfection), and protein expression was allowed to proceed for 24 h. At 3 h postinfection, coverslips were washed with PBS +/+ (calcium 0.9 mM/magnesium 0.49 mM), spent media were replaced with fresh growth media, and incubation continued for an additional 1.5 h. Samples were imaged with a Leica DMI4000 B microscope connected to a humidified livecell chamber (CU-109) [Chamlide] set at 37 °C and 60 mmHg

Confirmation of Novel Protein−Protein Interactions

To verify the legitimacy of the novel targets identified in our screen, we selected individual hits for confirmation using CoIP. D

DOI: 10.1021/acs.jproteome.6b00074 J. Proteome Res. XXXX, XXX, XXX−XXX

Article

Journal of Proteome Research

Figure 1. Type III secretion system effectors NleB1 and EspL interact with host MAP7. Representative mass spectra for immunoprecipitations with NleB1 (A,B) and EspL (C,D) (filled triangles) or GST-HA3 (A-D) (open triangles). Triangles indicate expected mass/charge (m/z) of light (open) and heavy (filled) forms of individual peptides for carbonyl reductase (A), MAP7 (B,D) and eF1γ (C). Peptide sequences and heavy-to-light (H/L) ratios are shown for each peptide.

Because NleB1, EspL, and EspX were all found to bind MAP7, we prioritized these interactions for further study. Considering that the majority of our experimental models used EPEC rather than EHEC, we chose to continue with NleB1 and EspL, given that these effector families are conserved among EHEC and EPEC, unlike EspX, which is not found among strains of EPEC.28 Additionally, the presence of pathogenicity island O122, which harbors both nleB1 and espL, has been correlated with higher virulence potential in humans.29−31 We therefore focused our investigation on these two effectors. Using the SILAC method, purified effectors were used as bait to enrich for host interacting partners (Figure 1). To confirm these interactions, we performed reciprocal pulldowns using either FLAG-tagged host protein (MAP7-FLAG) or purified recombinant effector as bait to enrich for their corresponding interaction partner. Both NleB1 and EspL were precipitated in greater amounts using α-FLAG antibody directed at MAP7FLAG compared with vector-only controls (Figure 2 and Supplementary Figure 1). Furthermore, the immunoprecipitation of NleB1, using α-HA antibody, showed an enriched amount of MAP7-FLAG precipitating with effector in comparison to an antibody-only control (Figure 2A and Supplementary Figure 1A), thus supporting the predicted interactions from our SILAC screen.

Figure 2. Confirmation of effector-MAP7 interactions. Co-immunoprecipitation of MAP7 and NleB1 or EspL. Lysates of cells transfected with pmap7-FLAG or pCMV4A (control) were incubated with purified GST-HA3-tagged effector followed by IP with antibodies directed at MAP7-FLAG, GST-HA3-NleB1, or GST-HA3-EspL (as indicated). Samples were analyzed by Western blot to probe for MAP7 (αFLAG), GST-HA3-tagged effector (αHA), or calnexin (as a loading control). IP: immunoprecipitation; IB: immunoblot.

Specific Interactions of the Effectors NleB1 and EspL with the Microtubule Binding Region of MAP7

Our SILAC analysis identified MAP7 as an interaction partner for the conserved effectors, NleB1 and EspL. Given that MAP7 contains two separate domains for microtubule binding and the activation of kinesin-1 transport,12 we sought to further characterize the MAP7-effector interaction by identifying the area of effector binding within MAP7. We used pmap7-FLAG as a template to generate truncated versions of MAP7-FLAG and tested the expression of each construct through Western blot analysis (Figure 3). The shortest MAP7-FLAG construct that showed both a high level of expression and interacted with both effectors was MAP7C4 (Figure 3 and Supplementary Figure 2). MAP7C4 comprised amino acids 1−349, which

approximately corresponded to the microtubule binding (MTB) region of MAP7.12,32 On the basis of these results, an additional truncation mutant was made (MAP7N1) containing the proposed “activity” domain of MAP712 but missing the MTB region (Figure 3A). Contrary to what was seen with MAP7C4, CoIP analysis of purified effector with MAP7N1 showed that neither NleB1 nor EspL interacted efficiently with the C-terminal region of MAP7 (Figure 3B,C). These findings E

DOI: 10.1021/acs.jproteome.6b00074 J. Proteome Res. XXXX, XXX, XXX−XXX

Article

Journal of Proteome Research

Figure 3. Interaction between NleB1/EspL and MAP7 is mediated by the microtubule binding region of MAP7. (A) MAP7 constructs used in this study. (B) CoIP of NleB1 and MAP7C4 or Map7N1. (C) CoIP of EspL and MAP7C4 or MAP7N1. Lysates of cells transfected with MAP7 constructs or vector controls were incubated with purified GST-HA3-tagged effector and immunoprecipitated with MAP7-FLAG directed antibodies, followed by immunoblot probing for GST-HA3-tagged effector or calnexin (as a loading control). MTB: microtubule binding; EHR1: ensconsin homology region 1; EHR2: ensconsin homology region 2; IB: immunoblot; FL: full length; aa: amino acids.

in the speeds and mean squared displacements (MSDs) of TfRmCherry particles in cells infected with wild type EPEC or the ΔnleB2 mutant compared to ΔnleB1 or ΔespL (Figure 4B,C). These data show that functional nleB1 and espL genes are required to stall TfR-mCherry vesicle trafficking and that nleB2 is less important for this phenotype. However, other studies have shown that NleB2 retains a low level of activity in comparison to NleB1.36,37 For this reason, we sought to confirm our data through complementation analyses using a ΔnleB1ΔnleB2 double mutant and a ΔespL single mutant. To verify that inhibition of TfR-mCherry trafficking was indeed specific to deletion of nleB1 and espL, we complemented the ΔnleB1ΔnleB2 double mutant with pnleB1 and the ΔespL single mutant with pespL and tested the ability of these strains to perturb TfR-mCherry movement. We found that infection with ΔnleB1ΔnleB2/pnleB1 or ΔespL/pespL significantly reduced TfR vesicle trafficking compared to infection with their uncomplemented mutant strains (Figure 4C). Thus, the complementation of nleB1 and espL in trans partially restored the TfR trafficking inhibition that was seen during infections with wild type EPEC. These data imply that EPEC, through these effectors, not only controls the spatial localization of intracellular TfR-containing vesicles but also the speeds at which they travel during infection. Taken together, our quantitative proteomic interaction studies aided in the discovery of a novel function for the NleB1- and EspLencoding genes during A/E pathogen infections.

confirm that NleB1 and EspL interact specifically with MAP7 and that these effectors interact with an area of the protein that corresponds to its MTB region. EPEC Blocks Intracellular Transferrin Receptor Trafficking in an nleB1-/espL-Dependent Manner

MAP7 is required for kinesin-1-based motility in the neurons and oocytes of Drosophila,12,13 and it has recently been shown to interact with the kinesin-1 family motor, Kif5b, using murine C2C12 myoblasts.33 Identification of the host factors targeted by A/E pathogen-secreted effectors can provide clues to their functions during infection. Given that MAP7 is required for kinesin-1-based trafficking in Drosophila and our findings that NleB1 and EspL interacted with MAP7, we hypothesized that these effectors could be interfering with kinesin-1-based transport during infection. We therefore followed the spatiotemporal movement of the transferrin receptor (TfR), a classical marker for kinesin-based endosomal trafficking,34 during live EPEC infections. We used Ptk2 cells, which are conducive to live imaging analyses and which EPEC readily infects,35 to ectopically express a TfR-mCherry fusion. Cells expressing TfR-mCherry were then infected with EPEC and examined using fluorescent time-lapse imaging to track EPECinduced perturbations in fluorescent vesicle movement. Interestingly, TfR-mCherry movement was drastically reduced throughout the cell during wild type EPEC infections compared to uninfected cells or cells infected with a T3SS mutant (ΔescN) (Figure 4A). We also observed a distinct increase in TfR-mCherry vesicle movement during infections with ΔnleB1 and ΔespL compared to wild type EPEC, similar to what was seen with T3SS-deficient EPEC (Figure 4A). In contrast, cells infected with ΔnleB2 displayed disrupted movement of TfR vesicles similar to wild type EPEC (Figure 4A). Additionally, quantification of these differences showed a marked reduction



DISCUSSION

Many pathogenic bacteria including EPEC, EHEC, and Salmonella utilize T3SSs to facilitate infection of host cells,38−40 and the identification of host proteins targeted by T3SS effectors can provide insight into the molecular strategies F

DOI: 10.1021/acs.jproteome.6b00074 J. Proteome Res. XXXX, XXX, XXX−XXX

Article

Journal of Proteome Research

Figure 4. Enteropathogenic E. coli (EPEC) infection halting TfR-mCherry transport in an nleB1/espL-dependent manner. (A) Tracks of TfRmCherry containing vesicles from representative uninfected or EPEC-infected Ptk2 cells. Ptk2 cells transfected with pTfR-mCherry were infected with EPEC or left uninfected, and fluorescent TfR-containing vesicles were imaged using time-lapse microscopy. U-track multiple-particle tracking software was used to track fluorescent vesicle movement. Particle trajectories persisting ≥10 frames are solid black;