Vibrio effector protein, VopQ, forms a lysosomal gated channel that disrupts host ion homeostasis and autophagic flux

Abstract

Defects in normal autophagic pathways are implicated in numerous human diseases—such as neurodegenerative diseases, cancer, and cardiomyopathy—highlighting the importance of autophagy and its proper regulation. Herein we show that Vibrio parahaemolyticus uses the type III effector VopQ (Vibrio outer protein Q) to alter autophagic flux by manipulating the partitioning of small molecules and ions in the lysosome. This effector binds to the conserved V_o domain of the vacuolar-type H⁺-ATPase and causes deacidification of the lysosomes within minutes of entering the host cell. VopQ forms a gated channel ∼18 Å in diameter that facilitates outward flux of ions across lipid bilayers. The electrostatic interactions of this type 3 secretion system effector with target membranes dictate its preference for host vacuolar-type H⁺-ATPase–containing membranes, indicating that its pore-forming activity is specific and not promiscuous. As seen with other effectors, VopQ is exploiting a eukaryotic mechanism, in this case manipulating lysosomal homeostasis and autophagic flux through transmembrane permeation.

Autophagy is a cellular process by which cells degrade and recycle cytoplasmic contents by encapsulating them within a distinctive double bilayer membrane vesicle for delivery to the degradative lysosome (1). Disruption of normal autophagic pathways is implicated in numerous human diseases, stressing the importance of autophagy and its proper regulation (2). Vibrio parahaemolyticus, a Gram-negative marine bacterium and a major cause of gastroenteritis due to the consumption of contaminated raw or undercooked seafood, induces autophagy during infection (3). V. parahaemolyticus harbors two type 3 secretion systems (T3SSs), molecular syringes that enable the translocation of bacterial proteins, known as effectors, into the eukaryotic host (3). The first T3SS (T3SS1) orchestrates a temporally regulated cell death mediated by the induction of autophagy, followed by cell rounding and resulting in lysis of the host cell (4). T3SS1 effector VopQ, also known as VepA (vp1680), is both necessary and sufficient for the rapid induction of autophagy, even in the presence of known chemical inhibitors of autophagy (5).

VopQ is a 53-kDa protein with no apparent homology to any proteins outside of the Vibrio species. Vibrio homologs of VopQ have no known function or conserved structural domain. Previous work from our laboratory has shown that VopQ is a cytotoxic effector that accelerates host cell death and is essential in protecting V. parahaemolyticus from phagocytic uptake during infection (5). Based on microbial genetic studies, VopQ was shown to be necessary for the formation of an extensive network of autophagic vesicles in host cells within an hour of V. parahaemolyticus infection (5). Strikingly, recombinant VopQ alone is sufficient to induce this massive accumulation of autophagic vesicles, observed within minutes of microinjection of recombinant VopQ (picomolar concentrations) into eukaryotic cells (5). A recent report shows that VopQ interacts with a V_o subunit of the highly conserved eukaryotic vacuolar-type H⁺-ATPase (V-ATPase) and induces lysosomal rupture in vitro and in vivo (6). These authors hypothesized that lysosomes were “suicide bags” that rupture and lead to the release of luminal contents that are responsible for T3SS1-mediated cell death. This dated theory has been shown not to be mechanistically feasible, and would not be predicted to occur within the time frame that VopQ is seen to act (minutes vs. hours) (5, 7). Our current studies show that VopQ does not cause membrane rupture, but rather forms small pores on lysosomal and synthetic membranes, allowing particles of ∼350 Da, but not greater than 3 kDa, to pass through the lipid bilayer. Further analysis of VopQ in lipid bilayers revealed that it forms an ∼18-Å gated, outward-rectifying channel upon the electrostatic association of VopQ with phospholipids. We find that the presence of the V_o domain of the V-ATPase allows the targeting of VopQ to the appropriate acidic compartments under physiological pH during an infection. The molecular mechanism hijacked by VopQ for disrupting lysosomal homeostasis during V. parahaemolyticus infection is similar to that used by the endolysosomal Na⁺ channel regulated by mammalian target of rapamycin (mTOR) and provides insight into the signals that regulate autophagy in host cells (8).

Results

VopQ Interacts with V_o Subunits of the V-ATPase Proton Pump.

The V-ATPase is the main electrogenic proton pump involved in the acidification of many intracellular organelles within the endomembrane system, such as yeast vacuoles and mammalian lysosomes (9). It is made up of two heterologous subunits: The cytoplasmic V₁ domain hydrolyzes ATP, which energizes the translocation of protons through the membrane-bound V_o proton channel, and into the lumen of the acidic organelle. A recent study indicated that the Vibrio effector VopQ (vp1680, VepA) interacts with Vma3p (subunit c) of the highly conserved V-ATPase V_o proton translocation domain (6). Our analyses revealed that His-VopQ interacts with Vma6 (subunit d) of the V_o domain when incubated with highly enriched preparations of yeast vacuoles (Fig. 1A). Using immunoblot analysis, we noted that VopQ interacts with Vma6 and Vph1p V_o domain subunits in vitro, suggesting that VopQ has an affinity for the assembled V_o domain (Fig. 1B). This interaction does not appear to involve the entire V-ATPase, however, as Vma2p, subunit B of the ATP-hydrolyzing V₁ domain, was not detected in this interaction (Fig. 1B). The VopQ–Vma6p interaction was conserved in higher eukaryotes, as the human ortholog, V_o d subunit, coprecipitated from HeLa extracts in a VopQ-dependent manner (Fig. 1C) (9). Therefore, consistent with previous observations, we propose that VopQ has an affinity for the V-ATPase V_o domain, but does not appear to interact with the V₁ domain of the V-ATPase.

Fig. 1.

VopQ interacts with the V-ATPase V_o domain. (A) Two hundred micrograms of vacuoles purified from BJ3505 was incubated with or without 5 µg His₆-VopQ (23 °C, 30 min). VopQ-associated proteins were isolated using nickel beads, and eluates were separated via SDS/PAGE and visualized by Oriole stain. The Vma6p band was excised and digested with trypsin. Tryptic peptides were run through reverse-phase HPLCy/ion trap. Tandem MS/MS files were searched against a National Center for Biotechnology Information nonredundant database. Asterisks denote nonspecific bands. (B) Immunoblot analysis of VopQ-associated proteins from A: Vph1p (V_o), Vma6p (V_o), and Vma2p (V₁). (C) Immunoblot analysis of VopQ-associated proteins from HeLa cell lysates.

VopQ Disrupts Proton Translocation but Not ATPase Activity of the V-ATPase.

Based on these interactions with the V-ATPase, we next investigated the effect of VopQ on the ATPase activity and ATP-dependent proton translocation of the V-ATPase in yeast vacuoles. Because of the modular nature of this enzyme, each of these activities can be measured separately. By using a standard assay that measures the protonation of acridine orange in acidic environments (via loss of absorbance), we find that recombinant VopQ (rVopQ) completely abrogated V-ATPase proton translocation activity in a concentration-dependent manner to levels seen with an inhibitor of the V-ATPase, bafilomycin (Fig. 2A). In contrast to bafilomycin, however, rVopQ does not affect vacuolar ATPase activity (Table S1). Furthermore, 25 nM rVopQ (tagged with either hexahistidine or maltose binding protein; MBP) was sufficient to prevent acidification of vacuoles (Fig. S1A) and deacidify previously acidified vacuoles, as VopQ caused near-immediate proton equilibration when added to vacuoles 20 min after ATP addition (Fig. S1B). These studies support the hypothesis that VopQ has the ability to disrupt proton gradients across acidic organelles.

Fig. 2.

VopQ disrupts acidification in yeast vacuoles and HeLa lysosomes. (A) Proton translocation activity of vacuoles was measured as described (SI Materials and Methods) in the presence of 100 nM bafilomycin or increasing concentrations of His₆-VopQ. (B) HeLa cells expressing GFP-LC3 were microinjected with 10 µM GST or His₆-VopQ, incubated 10 min (37 °C, 5% CO₂), and then incubated with 75 nM LysoTracker Red for 10 min and visualized. (Scale bar, 20 µm.) (C) HeLa cells expressing GFP-LC3 were incubated with or without 25 μM chloroquine for 30 min and then either treated with 100 μM rapamycin or infected with the noted V. parahaemolyticus strains at a multiplicity of infection of 10 for 3 h. Cell lysates were separated via SDS/PAGE and immunoblotted for GFP and tubulin.

VopQ Disrupts Lysosomal Acidification and Autophagic Flux.

To confirm VopQ’s ability to disrupt the acidification of lysosomes in host cells during an infection, we performed V. parahaemolyticus POR3 infections followed by staining with LysoTracker, a pH-sensitive indicator, which accumulates within acidic organelles. POR3 is a derivative of the pathogenic RIMD2210633 strain where only the first T3SS remains active, and thus we can focus on the activity of only T3SS1 effector proteins. During POR3 infection, we see a loss of acidic punctae by 3 h; this observation is VopQ-dependent, as strains lacking VopQ (vopQ∆) maintain acidic organelles over this time frame (Fig. S1C). Therefore, VopQ activity is necessary to cause deacidification of these organelles during an infection. To confirm that VopQ alone is sufficient to induce deacidification of these compartments, we microinjected purified rVopQ or a control protein, glutathione S-transferase (GST), into HeLa cells expressing GFP-light chain 3 (GFP-LC3) and assessed for the presence of acidic vesicles using LysoTracker. Concomitantly, we assayed for the induction of autophagy, the previously identified activity of VopQ (5).

Upon induction of autophagy, cytosolic GFP-LC3-I in these cells is cleaved and lipidated to form GFP-LC3-II, which then becomes localized to autophagosomal membranes, allowing visualization of autophagosomes (10). As expected, microinjection of rVopQ, but not the control protein GST (Fig. 2B), was sufficient to induce autophagosome formation, as measured by the accumulation of GFP-LC3-II–positive vesicles. Furthermore, injection of VopQ, but not GST, completely blocked lysosomal acidification in the same cells within 20 min, as measured by LysoTracker retention within acidic compartments (Fig. 2B). Interestingly, the kinetics of lysosome deacidification by VopQ was nearly identical to the observed accumulation of autophagosomes as measured by the membrane targeting of GFP-LC3-II, all occurring within 20 min of injection (5, 10). Thus, VopQ is necessary and sufficient to not only directly and rapidly induce autophagosome accumulation but also inhibit the acidification of endolysosomal compartments in host cells (5).

VopQ, Like Chloroquine, Inhibits Turnover of Autophagosomes.

Autophagosomes are transient vesicles that are constantly turned over through fusion with the acidic, degradative lysosome. Chloroquine, a lysosomotropic agent, neutralizes acidic compartments, thereby inhibiting the activity of luminal lysosomal proteases and preventing the lysosome-dependent turnover of autophagosomes (11). We hypothesized that VopQ might be disrupting autophagic flux in a similar manner by deacidifying the lysosomes. Therefore, we tested whether VopQ-mediated accumulation of GFP-positive vesicles was due to increased formation of autophagosomes or decreased turnover of autophagosomes concomitant with GFP-LC3-II degradation. To investigate these possibilities, HeLa GFP-LC3 cells were pretreated with chloroquine followed by infection with V. parahaemolyticus POR3 strains or treatment with rapamycin. Rapamycin is a potent inhibitor of mTORC1 activity, which induces the formation of autophagic vesicles through the derepression of ULK1/Atg13/FIP200 activities (12). We hypothesized that if VopQ did not inhibit autophagosome turnover, but rather directly induced an increase in autophagosome formation, similar to rapamycin, an increase in GFP-LC3-II would be observed upon treatment with chloroquine. Alternatively, if VopQ inhibited turnover of autophagosomes, no increase in GFP-LC3-II would be observed upon treatment with chloroquine, because these two activities would be redundant.

Using an immunoblot analysis of cell lysates, we analyzed the conversion of GFP-LC3-I to GFP-LC3-II under specific treatments. As expected, treatment of cells with rapamycin caused an increase in the appearance of GFP-LC3-II (Fig. 2C, lane 2). Furthermore, complete LC3-I conversion was observed with cells cotreated with chloroquine and rapamycin (Fig. 2C, lane 7), consistent with an enhanced accumulation of autophagosomes due to both the blocked turnover of these vesicles by chloroquine treatment and the additional induction of autophagosome formation via mTOR inhibition. When comparing these results to those observed in cells infected with V. parahaemolyticus, we see that conversion of LC3-I to LC3-II appeared essentially complete in cells infected with strains expressing VopQ, but not in the ΔvopQ strain (Fig. 2C, lanes 3–5). The effect on LC3 conversion increased upon treatment with chloroquine of ΔvopQ-infected cells (Fig. 2C, lane 9) but not cells infected with strains expressing VopQ (Fig. 2C, lanes 8 and 10). Further cell biology studies demonstrated that the formation of green punctae representative of autophagosomes is consistent with this biochemical analysis of LC3-I conversion (Fig. S1D), and therefore these observations strongly support the hypothesis that VopQ mimics chloroquine treatments, which disrupt autophagosome–lysosome turnover by disrupting lysosomal acidification.

VopQ Allows for Release of Molecules <350 Daltons, but Does Not Rupture Lysosomes.

The observation that VopQ induces deacidification, coupled with the previous observation that indicates that VopQ has a membrane-disrupting activity (6), led us to further investigate the molecular mechanism of action of T3SS effector VopQ on membranes. Limited bioinformatics on VopQ indicates the presence of two or three transmembrane helices, suggesting that VopQ could have pore-forming activity in host membranes. To test this hypothesis, quenching concentrations of carboxyfluorescein (376 Da) were encapsulated within liposomes constructed of 1-palmytoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) and 1,2-dioleoyl-sn-glycero-3-phospho-L-serine (DOPS) (85:15 molar ratio). Pore formation was measured as a result of fluorescence increase upon dye release and dequenching (13). Similar to the results seen with yeast vacuole deacidification, we noticed that VopQ induces dye release in a concentration-dependent manner. MBP-VopQ, but not MBP, induced leakage of dye within seconds as measured by fluorescence dequenching (Fig. 3A); heat inactivation of VopQ completely abrogated this activity (Fig. 3A, light blue line). To examine whether VopQ is causing the release of only small molecules, rather than complete liposomal rupture, we assessed the size of the potential VopQ pores using quenching concentrations of fluorescein dextrans of differing sizes encapsulated within liposomes. Neither a 3-kDa nor a 10-kDa fluorescein dextran was released from liposomes treated with VopQ, indicating that VopQ likely forms a pore of a defined size (<28 Å diameter), allowing only molecules smaller than 3 kDa to escape the lumens (Fig. 3 B–D).

Fig. 3.

VopQ does not rupture lysosomes, but allows for release of molecules <350 Da (A) Carboxyfluorescein dye release was measured upon addition of increasing concentrations of MBP-VopQ, 100 nM heat-inactivated MBP-VopQ, or 100 nM MBP to carboxyfluorescein-containing liposomes. (B–D) Carboxyfluorescein or FITC dextran release was measured upon addition of 25 nM MBP-VopQ, 25 nM heat-inactivated MBP-VopQ, or 25 nM MBP to quenched liposomes.

A previous report suggested that VopQ causes rupture of lysosomes in vivo and in vitro, resulting in the release of luminal enzymes (6). Because our findings are in direct contrast to this previously published work, we revisited these studies. During a 3-h V. parahaemolyticus infection, we did not observe the release of the ∼33-kDa luminal cathepsin D from lysosomes (Fig. S2A), yet cells treated with the nonionic detergent Triton X-100 showed the complete release of cathepsin D into the cytosolic fraction. Importantly, we see both VopQ-dependent lysosomal deacidification and induction of autophagy within this infection time frame (Fig. S1 C and D), supporting previous results that VopQ is highly active during these conditions. Furthermore, an isolated membrane fraction highly enriched in lysosomes incubated with either recombinant MBP or MBP-VopQ for 30 min did not result in a VopQ-dependent release of cathepsin D (Fig. S2B). Treatment of these lysosomes at 37 °C for 2 h with either recombinant MBP or MBP-VopQ did induce small amounts of cathepsin D release.

Our previous observations suggested that VopQ does not allow release of large dextrans from liposomal membranes. To test this observation in vivo, we loaded HeLa lysosomes with 10-kDa Texas red dextran (TRD). When purified VopQ, but not GST, is injected into these cells, lysosomes release protons, as measured by the loss of LysoTracker accumulation, but not the 10-kDa TRD (Fig. 4). Similarly, in uninfected cells and cells infected with V. parahaemolyticus strains, we observe retention of TRD in intact lysosomes, regardless of the presence of VopQ (Fig. S2C). We also observe that TRD is localized to the lysosomes in cells treated with chloroquine, indicating that retention of TRD in lysosomes is independent of acidification (Fig. S2C). Therefore, we see no evidence for VopQ-mediated lysosomal rupture in vitro or in vivo, but consistently observe VopQ-mediated deacidification of the lysosome coupled with the induction of autophagy.

Fig. 4.

VopQ does not induce lysosomal rupture. HeLa cells were loaded with 1 mg/mL TRD for 6 h and chased overnight. The following day, cells were microinjected with 10 µM GST or His₆-VopQ, incubated 10 min (37 °C, 5% CO₂), and then incubated with 75 nM LysoTracker Green for 10 min and visualized. (Scale bar, 20 µm.)

VopQ Forms a Gated Outward-Rectifying Channel in Lipid Bilayers.

To determine the physical parameters for the potential VopQ channels in membranes, we measured the electrical activities of VopQ inserted in reconstituted lipid bilayers made of POPC/DOPS (85:15 molar ratio) (14). When 25 nM VopQ was introduced to one side of the bilayer, we observed spontaneous channel activities within minutes (Fig. 5A). A voltage ramp elicited an outward-rectifying current between −20 and +40 mV, and beyond this voltage range the channels became inactivated, suggesting an inactivation gate in the pore (Fig. 5B). Similar observations were obtained in the presence of a 10-fold gradient of NaCl across the bilayer (Fig. S3A). The near–0-mV reversal potentials in both cases suggest that the VopQ channel is only slightly selective for Na⁺ over Cl⁻, with a permeation ratio P_Na:P_Cl = 1.4. When single-channel currents were recorded and analyzed at different holding potentials (Fig. 5 C and D), the single-channel conductance of VopQ with equal 100 mM NaCl on both sides of the bilayer was calculated to be 129 pS (Fig. 5E). Using the Nernst–Planck equation, we estimated its pore diameter of ∼18 Å (15), which falls within our previous rough estimate based on the permeability of different dye molecules (Fig. 3). Under the same conditions, VopQ was not able to form active channels in the POPC membranes, suggesting the requirement of negatively charged lipids for VopQ membrane insertion or channel formation (Fig. 5F). We envision that electrostatic interactions between VopQ and the negatively charged lipids may contribute significantly to the free energy change that favors the stable VopQ channel in membrane.

Fig. 5.

VopQ forms a gated channel in membranes. (A) Current recorded after 25 nM MBP-VopQ was incubated with a POPC/DOPS (85:15 molar ratio) bilayer. (Upper) The voltage pulse stimulation is shown. (B) The averaged ramp current after the capacitance components were subtracted. The red trace was a linear fit from 0 mV to +30 mV with a chord conductance of 448.11 ± 1.67 pS. The inactivation of the channel becomes obvious when the voltage is higher than 40 mV or lower than −20 mV. (C) VopQ single-channel activity at different transmembrane voltages. C, close state; O, open state. (D) Histogram of single-channel events at +80 mV. We estimated the pore diameters by applying the Nernst–Planck equation, , for specific ion k. The measured current in our recordings was expressed as the sum of sodium and chloride flows, and the estimated pore diameter is ∼18 Å. (E) The measured single-channel current–voltage relation. The linear fitting (red line) yields a single-channel conductance of 128.9 ± 3.4 pS. (F) VopQ (25 nM) incubated with 100% POPC membranes did not produce any channel activity (80 mV, n = 4). MBP (25 nM) incubated with POPC/DOPS membranes did not produce any channel activity (80 mV, n = 6).

VopQ Forms Channels in V-ATPase–Containing Membranes.

Having determined that VopQ can form a channel in a lipid bilayer, we wondered how this effector specifically recognizes host lysosomal membranes while avoiding V. parahaemolyticus bacterial membranes. Our first clue came from our in vitro liposome dequenching experiments, where we observed pore formation at pH 5.5 but not at the near-physiological pH 7.5 (Fig. 6A). The estimated pI of VopQ is 6, and therefore VopQ would be positively charged at pH 5.5 and attracted to the negatively charged, DOPS-containing liposomes. However, at pH 7.5, negatively charged VopQ would not bind to these negatively charged liposomes. To further investigate the interaction of VopQ with liposomes, we altered the charge of carboxyfluorescein-encapsulated liposomes by changing the POPC:DOPS molar ratio, and hence the negative charge, of these membranes. At pH 5.5, positively charged VopQ binds and induces dye leakage in 10% and above DOPS-containing liposomes, but at not lower DOPS concentrations (Fig. 6B and Fig. S4). As seen before, the presence of V-ATPase subunits was not required for VopQ pore-forming activity in these artificial liposomes. However, we hypothesize that at physiological pH, the presence of the V_o domain of the V-ATPase allows the targeting of VopQ to the appropriate acidic compartments.

Fig. 6.

VopQ is a targeted pore-forming bacterial effector. (A) Carboxyfluorescein (CF) release was measured upon addition of 25 nM MBP-VopQ, 25 nM heat-inactivated MBP-VopQ, or 25 nM MBP to liposomes in buffer with varying pH. (B) Dequenched carboxyfluorescein fluorescence was measured upon addition of 25 nM MBP-VopQ, 25 nM heat-inactivated MBP-VopQ, or 25 nM MBP to liposomes made of increasing concentrations of DOPS (0–20%). (C) Immunoblot analysis of VopQ association with vacuoles isolated from wild-type or V-ATPase mutant yeast strains at pH 5.5 or 7.5. P, vacuole pellet; S, supernatant; T, total reaction. (D) Model for VopQ localization and pore-forming activity. VopQ docks onto the V_o domain of the V-ATPase. VopQ then forms an 18-Å pore in the membrane, allowing the flow of ions and small molecules but not proteins 3 kDa or larger.

In confirmation of this hypothesis, we observed that VopQ associates with yeast vacuolar membrane at pH 5.5, and this association was independent of the presence of the V_o domain of the V-ATPase, which is lacking in vma6∆ vacuoles (Fig. 6C). Strikingly, VopQ only binds to yeast vacuoles at pH 7.5 when the V_o domain of the V-ATPase is present (Fig. 6C). Therefore, we propose that the association of VopQ with a membrane at physiological pH is restricted to membranes containing the V_o domain of the V-ATPase, thereby acting as a protective mechanism against promiscuous pore-forming activity in the bacterium and the host (Fig. 6D).

Discussion

Bacterial pathogens have evolved a number of unique mechanisms to target and manipulate host cell signaling (16). V. parahaemolyticus appears to secrete approximately half a dozen T3SS1 effectors to sequentially induce rapid autophagosome accumulation, followed by cell rounding and lysis. The efficiency of these activities can be attributed in many cases to the directed localization of an effector activity. We observe that VopQ targets the endolysosomal membranes by binding to the V_o complex of the V-ATPase (6, 9). This ensures specificity for VopQ-mediated pore formation in the membranes of acidic compartments where at least protons are released within an extremely rapid time frame (minutes). VopQ antagonizes the activity of the V-ATPase and forms a gated, outward-rectifying channel that disrupts ion homeostasis, resulting in rapid accumulation of autophagosomes. These observations are consistent with rapid autophagosome accumulation resulting from the neutralization of the lysosome, similar to what is observed with the buffering activity of lysosomotropic agents, such as chloroquine (5).

Autophagy is used as a mechanism by which host cells recycle nutrients as well as defend themselves against invasion by bacteria. Because of the extent to which pathogenic microorganisms must manipulate their host cells during infection, multiple studies have used bacterial effectors as a powerful tool to study the regulation of autophagic pathways. Herein we show that disruption of the autophagic flux by VopQ can result in rapid accumulation of these vesicles. As many effectors mimic the activity of host factors, it is reasonable to hypothesize that rapid autophagosomal accumulation could be mediated by a regulated outward-rectifying channel. This channel could open, resulting in lysosomal deacidification and accumulation of autophagosomes. The channel could then close to allow the lysosome to reacidify by proton translocation by the V-ATPase. The accumulated autophagosomes could then be degraded by the lysosome. In support of this hypothesis, Cang et al. recently identified a eukaryotic endolysosomal ATP-sensitive Na⁺ channel that couples endolysosomal function to the cell’s energy state (8). Upon nutrient starvation, mTOR is recruited away from the lysosome and the Na⁺ channel controls lysosomal membrane potential and pH stability. Upon nutrient repletion, the channel is responsive to cellular ATP levels and opens and closes to control lysosomal ion and amino acid homeostasis. It is interesting to note that VopQ forms a pore large enough to disrupt amino acid homeostasis and thereby induce autophagy (17).

In a recent study by Matsuda and colleagues, VopQ (VepA) was implicated in the complete rupture of lysosomes. However, this VopQ-mediated effect was only observed after extensive (hours) incubations and infections (6). Although VopQ may indeed contribute to the eventual lysis of this organelle after an extended time, we find that the molecular activity of VopQ is more refined. We show that VopQ directly forms small pores within minutes, allowing the release of small molecules from acidic compartments, causing a collapse of ion homeostasis inside eukaryotic cells. Previously, the activity of VopQ was implicated in modulating MAPK signaling and inducing IL-8 secretion during V. parahaemolyticus infection (18, 19); researchers have shown that pore-forming toxins can directly regulate these same pathways (20). Therefore, we hypothesize that VopQ could induce IL-8 secretion by forming pores and rapidly disturbing cytosolic ion concentrations. It is reasonable to speculate that other pathogens may alter cellular homeostasis by using targeted effectors that act as channels to disrupt organelles, such as the mitochondria (21).

Bioinformatics analysis indicates that VopQ may encode multiple domains; however, initial deletion analyses of VopQ have resulted in completely inactive proteins. The delivery of the bacterial effector VopQ into cells immediately creates a gated outward-rectifying channel within an acidic compartment’s membrane in the infected cell, disrupting membrane and vacuolar trafficking, ion potentials, and second-messenger signaling. Future studies will involve understanding the biophysics and biochemistry of this soluble 53-kDa protein that efficiently embeds itself into the host lysosomal membrane to form a gated channel.

Materials and Methods

Detailed information is provided in SI Materials and Methods on cell lines and bacterial strains. Protocols for reagent preparation, microinjection, vacuolar pull-down assay, and HeLa lysate pull-down assays are also included. Further information describing VopQ:vacuole membrane interactions, V-ATPase assays, liposome leakage assays, lysosome preparation, and VopQ activities in bilayer lipid membranes is also provided.