Host Serine/Threonine Kinases mTOR and Protein Kinase C-α Promote InlB-Mediated Entry of Listeria monocytogenes

ABSTRACT

The bacterial pathogen Listeria monocytogenes causes foodborne illnesses resulting in gastroenteritis, meningitis, or abortion. Listeria induces its internalization into some human cells through interaction of the bacterial surface protein InlB with the host receptor tyrosine kinase Met. InlB-dependent entry requires localized polymerization of the host actin cytoskeleton. The signal transduction pathways that act downstream of Met to regulate actin filament assembly or other processes during Listeria uptake remain incompletely characterized. Here, we demonstrate important roles for the human serine/threonine kinases mTOR and protein kinase C-α (PKC-α) in InlB-dependent entry. Experiments involving RNA interference (RNAi) indicated that two multiprotein complexes containing mTOR, mTORC1 and mTORC2, are each needed for efficient internalization of Listeria into cells of the human cell line HeLa. InlB stimulated Met-dependent phosphorylation of mTORC1 or mTORC2 substrates, demonstrating activation of both mTOR-containing complexes. RNAi studies indicated that the mTORC1 effectors 4E-BP1 and hypoxia-inducible factor 1α (HIF-1α) and the mTORC2 substrate PKC-α each control Listeria uptake. Genetic or pharmacological inhibition of PKC-α reduced the internalization of Listeria and the accumulation of actin filaments that normally accompanies InlB-mediated entry. Collectively, our results identify mTOR and PKC-α to be host factors exploited by Listeria to promote infection. PKC-α controls Listeria entry, at least in part, by regulating the actin cytoskeleton downstream of the Met receptor.

INTRODUCTION

Listeria monocytogenes is a foodborne pathogen capable of causing serious infections resulting in meningitis or abortions (1). Critical for disease is the ability of Listeria to induce its internalization into nonphagocytic mammalian cells in the intestine, liver, or placenta (2). One of the major pathways of internalization of Listeria into human cells is mediated by interaction of the bacterial surface protein InlB with the host receptor tyrosine kinase Met (3). InlB-dependent entry involves remodeling of the host plasma membrane through actin polymerization (4, 5). Substantial progress in characterizing InlB-mediated uptake has been made, resulting in the identification of several host signal transduction pathways that control actin polymerization and perhaps other physiological processes during entry (4). Despite this progress, a detailed understanding of the molecular mechanism of InlB-dependent internalization of Listeria remains incomplete.

The human type IA phosphoinositide 3-kinase (PI3K) pathway is one of the major signaling networks controlling InlB-mediated entry (4, 6–8). Recent RNA interference (RNAi) screens identified 13 components of the type IA PI3K pathway involved in InlB-dependent internalization of Listeria (9, 10). One host protein demonstrated in this work to play an important role in Listeria entry is mTOR (mammalian target of rapamycin). mTOR is a serine/threonine kinase that regulates cell growth, autophagy, and the actin cytoskeleton in response to growth factor stimulation or the availability of nutrients, energy, or oxygen (11). mTOR exists in two multisubunit protein complexes, termed mTORC1 and mTORC2. These complexes exert distinct biological functions by directly phosphorylating different substrates and indirectly controlling other effectors. For example, mTORC1 stimulates protein synthesis by phosphorylating the translational regulators p70S6K and 4E-BP1 and suppresses autophagy by phosphorylating ULK1 (11, 12). mTORC1 promotes aerobic glycolysis or lipogenesis by upregulating expression of the transcription factor hypoxia-inducible factor 1α (HIF-1α) or SREBP1, respectively. mTORC2 mediates phosphorylation of members of the AGC subfamily of serine/threonine kinases, including Akt and protein kinase C-α (PKC-α). Akt controls several biological events, including cell survival and proliferation (11). PKC-α regulates the actin cytoskeleton and membrane trafficking (13–15).

In this work, we investigated the mechanism by which mTOR controls the InlB-mediated entry of Listeria. Experiments involving the compound Torin 1 (16) showed that entry of Listeria into HeLa cells requires mTOR kinase activity. Infection of cells with Listeria or treatment with InlB protein resulted in increased phosphorylation of mTORC1 or mTORC2 substrates, indicating activation of both mTOR-containing complexes. RNAi-mediated depletion of components specific for mTORC1 or mTORC2 each inhibited Listeria infection, demonstrating roles for these complexes in InlB-dependent uptake. Further RNAi studies identified the mTORC2 substrate PKC-α to be a host factor needed for internalization of Listeria. By performing laser scanning confocal microscopy of HeLa cells depleted for PKC-α or expressing a mutant allele of PKC-α defective in kinase activity, we found that this host protein affected the accumulation of polymerized actin during InlB-mediated internalization. Collectively, our results indicate important roles for mTOR and PKC-α in Listeria infection and identify PKC-α as a regulator of the actin cytoskeleton during entry.

RESULTS

mTOR kinase activity is needed for InlB-mediated entry of Listeria.

In a previous study, we obtained genetic evidence indicating an important role for mTOR in InlB-mediated uptake into HeLa cells (9). Specifically, four different short interfering RNAs (siRNAs) directed against distinct sequences in mTOR mRNA each reduced the level of target protein expression and also inhibited the entry of Listeria. In order to determine if the kinase activity of mTOR is involved in Listeria uptake, we used the chemical inhibitor Torin 1 (16). This compound is an ATP-competitive inhibitor that impairs the kinase activity of both mTORC1 and mTORC2. The 50% inhibitory concentration of Torin 1 is approximately 10 nM, and concentrations of 100 to 250 nM are typically needed to abolish mTORC1 and mTORC2 activity (16–18). Importantly, incubation of HeLa cells with Torin 1 at concentrations ranging from 25 to 250 nM reduced the level of entry of a Listeria strain expressing InlB (Fig. 1Ai). Internalization of Listeria into HeLa cells is dependent almost entirely on InlB, as a bacterial strain deleted for the inlB gene (ΔinlB) enters at about 10% of the level of an isogenic strain expressing InlB (Fig. 1Aii) (19). Treatment of HeLa cells with Torin 1 failed to impair uptake of Listeria deleted for inlB (Fig. 1Aiii), indicating that this compound does not affect the residual entry that occurs in the absence of InlB. The inhibition of InlB-dependent entry by Torin 1 was unlikely due to adverse effects on cell physiology, as the inhibitor did not affect the viability of HeLa cells at most concentrations or diminish the level of entry of Escherichia coli expressing the invasin protein of Yersinia enterocolitica (20) (Fig. 1B and C). Taken together, the data in Fig. 1 provide evidence that mTOR kinase activity participates in InlB-mediated entry.

FIG 1.

Entry of Listeria is impaired by a chemical inhibitor of mTOR. HeLa cells were incubated for 45 min in medium with the indicated concentrations of Torin 1 or the vehicle DMSO (−). Cells were then tested for the ability to support the entry of Listeria strains expressing InlB (inlB⁺) or deleted for the inlB gene (ΔinlB). The effects of Torin 1 on the entry of E. coli strain HB101 expressing invasin (inv⁺ HB101) or on the viability of HeLa cells were also assessed. (A) Effect of Torin 1 on InlB-mediated entry of Listeria. (i) Torin 1 treatment inhibits the entry of Listeria expressing InlB. Data are relative values obtained by normalization to the values for entry in the absence of Torin 1 treatment (−). Results are means ± SEMs from three experiments. (ii) Comparison of entry of isogenic Listeria strains expressing InlB (inlB⁺) or deleted for inlB (ΔinlB). The results are those obtained in the absence of Torin 1 treatment. Data are expressed as the percentage of bacteria internalized and are means ± SEMs from five experiments. (iii) Torin 1 treatment fails to reduce the uptake of a Listeria strain deleted for inlB. Data are relative values obtained by normalization to the values for entry in the absence of Torin 1 treatment. Results are the means ± SEMs from three experiments. (B) Viability of HeLa cells treated with Torin 1, as measured by MTT assays. Data are mean ± SEM values from three experiments. (C) Lack of effect of Torin 1 on uptake of inv⁺ HB101. Data are mean ± SEM values from three experiments. *, P < 0.05 compared to DMSO treatment.

Both mTORC1 and mTORC2 are needed for efficient internalization of Listeria.

mTORC1 and mTORC2 share several common components but are distinguished chiefly by distinct scaffolding proteins that mediate complex assembly and substrate binding (11). The scaffolding protein Raptor is present in mTORC1 and needed for the efficient phosphorylation of mTORC1 substrates (21, 22). Conversely, Rictor is part of mTORC2 and mediates substrate phosphorylation by this complex (23). We used RNAi to investigate the roles of Raptor and Rictor in the entry of Listeria. These two human proteins were each targeted by three different siRNAs in order to minimize the possibility of off-target effects (24). siRNAs were introduced into HeLa cells by transfection. As controls, cells were either mock transfected in the absence of siRNA or transfected with a control nontargeting siRNA that lacks complementarity to any known human mRNA. Importantly, siRNA-mediated depletion of Raptor or Rictor each decreased the level of entry of Listeria into HeLa cells (Fig. 2A and B). We next used RNAi to target human proteins apart from Raptor and Rictor that selectively affect the activity of mTORC1 or mTORC2. The protein mLST8 is a common component of both of these complexes but is needed for the activity of only mTORC2 (11). RNAi-mediated depletion of mLST8 inhibited the internalization of Listeria (Fig. 2C). The activity of mTORC1 is negatively regulated by a heterodimeric complex composed of TSC1 and TSC2 (11). Depletion of TSC2 augmented the entry of Listeria (Fig. 2D). siRNA against Raptor, Rictor, mLST8, or TSC2 did not affect adhesion of bacteria to HeLa cells (see Fig. S1 in the supplemental material), indicating that these host proteins control postbinding steps. Taken together, the results in Fig. 2 indicate that both mTORC1 and mTORC2 participate in the InlB-mediated entry of Listeria.

FIG 2.

mTORC1 and mTORC2 promote InlB-dependent entry of Listeria. HeLa cells were either mock transfected in the absence of siRNA, transfected with a control nontargeting siRNA, or transfected with siRNA against Raptor, Rictor, mLST8, or TSC2. At about 48 h after transfection, cell lysates were prepared for analysis of target gene expression by Western blotting or qPCR or cells were infected with a Listeria strain expressing InlB for assessment of bacterial entry using gentamicin protection assays. (A) Role of the mTORC1 component Raptor in entry. (i) A representative blot showing the depletion of Raptor is displayed. The three different siRNAs used to target Raptor are indicated as 1, 2, and 3. After reaction with anti-Raptor antibodies, the membrane was stripped and probed with antitubulin antibodies to confirm equivalent loading. The adjacent bar graph displays the mean ± SEM values of quantified Western blotting data from three experiments. (ii) Relative entry data are presented as mean ± SEM values from three or four experiments, depending on the siRNA. (B) Function of the mTORC2 component Rictor in bacterial uptake. (i) An image of a representative Western blot and a graph of quantified blotting data from three experiments are shown. The three different siRNAs used to target Rictor are indicated as 1, 2, and 3. (ii) Bacterial entry data from three or four experiments are displayed. (C) Role of the mTORC2 regulator mLST8 in internalization of Listeria. (i) An image of a representative Western blot and a bar graph of quantified blotting data from three experiments are presented. The three siRNAs used to target mLST8 are indicated as 1, 2, and 3. (ii) Bacterial entry data from three or four experiments are shown. The numbers to the right of the blots in panels Ai, Bi, and Ci are molecular sizes (in kilodaltons). (D) Regulation of Listeria uptake by TSC2. (i) qPCR data indicating the siRNA-mediated reduction in the level of expression of TSC2 are shown. The three different siRNAs used to target TSC2 are indicated as 1, 2, and 3. These results are mean ± SEM values from three experiments. (ii) Bacterial entry data from three or four experiments are given. *, P < 0.05 compared to the control siRNA condition.

InlB stimulates mTORC1 and mTORC2 activity.

When used in a soluble form at low nanomolar concentrations, the InlB protein is a potent agonist of several mammalian signal transduction pathways (3, 7, 8, 25–27). For this reason, InlB is often used as a tool to investigate signaling events associated with Listeria entry. We found that treatment of HeLa cells with 4.5 nM InlB for 5 min induced phosphorylation of the mTORC1 substrate p70 S6 kinase (p70S6K) and the mTORC2 substrate Akt (11) (Fig. 3). As expected, phosphorylation of these two substrates was inhibited by siRNA-mediated depletion of Met (Fig. 3Ai and Bi), indicating that mTORC1 and mTORC2 are activated downstream of this human receptor. Phosphorylation or p70S6K or Akt was also abrogated by siRNA-mediated depletion of mTOR or treatment of cells with Torin 1 (Fig. 3Aii and Bii), confirming dependency on mTOR kinase activity. Since the activities of both mTORC1 and mTORC2 are known to require type IA PI3K (11), we verified that the PI3K inhibitor LY294002 impaired InlB-induced phosphorylation of p70S6K or Akt (Fig. 3Aii and Bii). Collectively, the results in Fig. 3 indicate that InlB activates both mTORC1 and mTORC2.

FIG 3.

InlB induces phosphorylation of mTORC1 and mTORC2 substrates. (A) Phosphorylation of the mTORC1 substrate p70S6K. (i) Inhibition of p70S6K phosphorylation by depletion of mTOR. HeLa cells were transfected with siRNAs targeting mTOR or Met. As controls, cells were mock transfected in the absence of siRNA or transfected with a control nontargeting siRNA. At approximately 48 h posttransfection, serum-starved cells were left alone or treated with 4.5 nM soluble InlB protein for 5 min, followed by solubilization in lysis buffer and assessment of phosphorylation of p70S6K by Western blotting with phospho-specific antibodies. After detection of phosphorylated p70S6K, the membranes were stripped and probed with antitubulin antibodies to confirm equivalent loading. Lysates were also probed with antibodies against mTOR or Met to confirm depletion of these proteins. The bar graph on the right displays quantified Western blotting data for p70S6K phosphorylation from three experiments. Values are means ± SEMs. (ii) Reduction of p70S6K phosphorylation by inhibitors of mTOR or PI3K. HeLa cells were treated with 250 nM Torin 1 or 10 μM LY294002, followed by stimulation with 4.5 nM InlB for 5 min and preparation of lysates for Western blotting. Shown is a representative Western blot detecting phosphorylated p70S6K and a bar graph of quantified blotting data from three experiments. (B) Phosphorylation of the mTORC2 substrate Akt. (i) Impaired phosphorylation of Akt upon mTOR depletion. After transfection with siRNAs targeting mTOR or Met, HeLa cells were treated with 4.5 nM InlB protein for 5 min and solubilized for Western blotting. (i) A representative Western blot with anti-phospho-Akt antibodies is shown. Membranes were stripped and probed with anti-Akt or antitubulin antibodies to confirm equivalent loading. Next to the Western blot, a graph of quantified blotting data from three experiments is presented. (ii) Inhibition in Akt phosphorylation by mTOR or PI3K inhibitors. HeLa cells were treated with 250 nM Torin 1 or 10 μM LY294002, incubated with 4.5 nM InlB for 5 min, and solubilized for Western blotting. A representative blot and quantified data from three experiments are displayed. The numbers to the right of the blots are molecular sizes (in kilodaltons). *, P < 0.05.

We next investigated if mTORC1 and/or mTORC2 is activated during InlB-dependent entry. In agreement with a previous report (8), infection of HeLa cells with a Listeria strain expressing InlB (the inlB⁺ strain) resulted in increased phosphorylation of the mTORC2 substrate Akt (Fig. 4A). Importantly, infection with an isogenic Listeria strain deleted for inlB (the ΔinlB mutant) failed to augment Akt phosphorylation, indicating dependency on the InlB uptake pathway. As a second model of InlB-dependent entry, we used latex beads coupled to the InlB protein. These beads are efficiently internalized into human cells (Fig. S2A) and have been extensively used to study signal transduction events during InlB-mediated entry (8, 10, 19, 28–30). As expected, increased phosphorylation of Akt was observed in HeLa cells incubated with InlB-coated beads for 5 to 20 min (Fig. 4B). In contrast, increased Akt phosphorylation was not observed with control beads coupled to glutathione S-transferase (GST). In the case of the mTORC1 substrate p70S6K, phosphorylation increased similarly in cells infected with Listeria expressing or lacking inlB (Fig. 4C). However, incubation of HeLa cells with beads coupled to InlB stimulated p70S6K phosphorylation (Fig. 4D). These results indicate that, while InlB is sufficient to activate mTORC1, bacterial factors in addition to InlB also contribute to activation of this complex. Taken together, the results in Fig. 4 demonstrate activation of mTORC1 and mTORC2 during InlB-mediated internalization.

FIG 4.

InlB-mediated entry is accompanied by phosphorylation of Akt and p70S6K. (A) Stimulation of Akt phosphorylation by Listeria. HeLa cells were either left uninfected (−), infected with a Listeria strain expressing InlB (inlB⁺), or infected with an isogenic strain lacking the inlB gene (ΔinlB) for 5 or 10 min. Cell lysates were prepared and used to detect phosphorylated Akt by Western blotting. (i) Image of a representative Western blot. (ii) Quantified Western blotting data are shown as mean ± SEM values from three experiments. *, P < 0.05 compared to the absence of infection or to infection with the ΔinlB mutant for 10 min. (B) Phosphorylation of Akt induced by InlB-coated beads. HeLa cells were left alone (−) or incubated with latex beads coupled to InlB or GST as a control for the indicated times. Cells were then solubilized for Western blotting. (i) A representative Western blot with anti-phospho-Akt antibodies is shown. (ii) Quantified blotting data from three experiments are presented. *, P < 0.05 compared to incubation without beads or with GST-coated beads. (C) Phosphorylation of p70S6K during infection with Listeria. HeLa cells were left uninfected (−) or infected with Listeria strains expressing InlB (inlB⁺) or lacking the inlB gene (ΔinlB) for the indicated times. Cell lysates were prepared and used for Western blotting to detect phosphorylation of p70S6K. (i) A representative Western blot with anti-phospho-p70S6K antibodies is shown. The membrane was stripped and probed with antitubulin antibodies to verify equivalent loading in the various lanes. (ii) Quantified blotting data from three experiments are displayed. *, P < 0.05 compared to a lack of infection. (D) Increased phosphorylation of p70S6K upon incubation with InlB-coated beads. HeLa cells were incubated with medium alone (−) or with beads coupled to InlB or GST for the indicated times. Cell lysates were made and used for Western blotting with anti-phospho-p70S6K antibodies. (i) A representative Western blot showing p70S6K phosphorylation. (ii) Quantified Western blotting data from three experiments. *, P < 0.05 relative to incubation without beads or with beads coupled to GST. The numbers to the right of all blots are molecular sizes (in kilodaltons).

The mTOR effectors 4E-BP1, HIF-1α, and PKC-α control entry of Listeria.

We investigated known downstream effectors of mTORC1 or mTORC2 for roles in InlB-dependent entry. Previous results indicated that p70S6K or Akt is dispensable for entry of Listeria into HeLa cells (9). We therefore limited our work to the mTORC1 effectors 4E-BP1, HIF-1α, SREBP-1, and ULK1 and the mTORC2 substrate PKC-α (11). In initial studies, single siRNAs were used to target each of these mTOR effectors. siRNA-mediated knockdown of 4E-BP1, HIF-1α, or PKC-α affected the entry of Listeria in a statistically significant fashion (Fig. 5A). Inhibition of the expression of 4E-BP1 augmented the uptake of Listeria, whereas knockdown of HIF-1α or PKC-α decreased the level of bacterial entry. In order to address potential off-target effects of siRNAs, studies were expanded to include three different siRNAs against 4E-BP1, HIF-1α, and PKC-α. Importantly, all siRNAs employed affected the entry of Listeria in a statistically significant manner (Fig. 5B to D). These results indicate that 4E-BP1, HIF-1α, and PKC-α each have bona fide roles in Listeria entry. HIF-1α and PKC-α positively control bacterial uptake, whereas 4E-BP1 is a negative regulator of this process.

FIG 5.

The mTOR effectors 4E-BP1, HIF-1α, and PKC-α control internalization of Listeria. (A) RNAi screen to examine the roles of mTORC1 or mTORC2 substrates in bacterial entry. HeLa cells were mock transfected in the absence of siRNA (none), transfected with a control siRNA, or transfected with siRNAs targeting 4E-BP1, HIF-1α, SREBP-1, ULK1, or PKC-α. After transfection, cells were processed for assessment of target gene expression by qPCR or infected with a Listeria strain expressing InlB for measurement of bacterial entry. (i) qPCR results are shown as mean ± SEM values from three or four experiments, depending on the siRNA. (ii) Bacterial entry data are means ± SEMs from three or four experiments. *, P < 0.05 compared to the control siRNA condition. (B) Regulation of Listeria entry by 4E-BP1. HeLa cells were transfected with three siRNAs targeting 4E-BP1 or subjected to control conditions. (i) The results of qPCR analysis of gene expression are presented. The three different siRNAs used to target 4E-BP1 are indicated as 1, 2, and 3. Data are means ± SEMs from three experiments. (ii) Bacterial entry results are means ± SEMs from three or four experiments. *, P < 0.05 compared to the control siRNA condition. (C) Function of HIF-1α in internalization of Listeria. After transfection with three siRNAs against HIF-1ɑ, target gene expression and the entry of Listeria were assessed. (i) qPCR data of HIF-1α expression are presented as mean ± SEM values from three or four experiments. The three different siRNAs used to target HIF-1α are indicated as 1, 2, and 3. (ii) Bacterial entry results are shown as mean ± SEM values from three to eight experiments. *, P < 0.05 compared to the control siRNA condition. (D) Role of PKC-α in entry of Listeria. HeLa cells transfected with siRNAs targeting PKC-α were examined for effects on target protein expression or internalization of Listeria. (i) A representative Western blot of PKC-α expression is shown. Adjacent to the blot image, a bar graph of quantified Western blotting data from four experiments is presented. The three different siRNAs used to target PKC-α are indicated as 1, 2, and 3. Data are means ± SEMs. The numbers to the right of the blot are molecular sizes (in kilodaltons). (ii) Bacterial entry data are shown as mean ± SEM values from three to eight experiments. *, P < 0.05 compared to the control siRNA condition.

InlB stimulates phosphorylation of PKC-α.

PKC-α is a member of the AGC subfamily of kinases, which also includes Akt (11, 31, 32). mTORC2 phosphorylates serine 657 (S657), located in the hydrophobic motif of PKC-α (23, 33). Phosphorylation at this site is needed for PKC activity (31). Using an antibody recognizing phosphorylated S657 in PKC-α, we found that treatment of HeLa cells with the InlB protein stimulated phosphorylation of this PKC (Fig. 6A). The protein recognized by the phospho-specific antibody was confirmed to be PKC-α by demonstrating decreased immunoreactivity in samples from cells depleted for PKC-α by RNAi (Fig. 6B). Additional RNAi experiments indicated that InlB-induced phosphorylation of PKC-α depended on mTOR and Met (Fig. 6A). Importantly, increased phosphorylation of S657 in PKC-α was also observed in HeLa cells infected for 30 min with Listeria expressing InlB but not in cells infected with the ΔinlB mutant (Fig. 6C). Incubation of cells with InlB-coated beads also augmented PKC-α phosphorylation (Fig. 6D). Collectively, the results in Fig. 6 indicate that InlB stimulates phosphorylation of PKC-α in a manner dependent on Met and mTOR.

FIG 6.

InlB stimulates phosphorylation of PKC-α. (A) mTOR-dependent phosphorylation of PKC-α induced by InlB. HeLa cells were subjected to control conditions or transfected with siRNAs targeting mTOR or Met. At about 48 h posttransfection, cells were left unstimulated (−) or treated with 4.5 nM soluble InlB protein for 10 min. Cell lysates were prepared and used to assess the phosphorylation of PKC-α by Western blotting. (i) A representative Western blot is shown. After reaction with anti-phospho-PKC-α antibodies, the membranes were stripped and probed with antibodies against total PKC-α or tubulin to verify similar loading in the various lanes. Lysates were also Western blotted with antibodies against mTOR or Met to confirm depletion of these proteins. (ii) Quantified data from anti-phospho-PKC-α Western blots are presented. Results are means ± SEMs from three experiments. *, P < 0.05. (B) Confirmation that the anti-phospho-PKC-α antibody recognizes PKC-α. HeLa cells were mock transfected in the absence of siRNA, transfected with a control siRNA, or transfected with an siRNA targeting PKC-α. Cells were then left untreated or stimulated with 4.5 nM soluble InlB for 5 min before preparation of lysates for Western blotting. (i) A representative Western blot is shown. Note that reactivity with the anti-phospho-PKC-α antibody decreased in lysates from cells treated with PKC-α siRNA. (ii) Quantified Western blotting data are shown as mean ± SEM values from three experiments. *, P < 0.05. (C) Effect of bacterial infection on PKC-α phosphorylation. HeLa cells were left alone (−) or infected with strains of Listeria expressing InlB (inlB⁺) or lacking the inlB gene (ΔinlB) for the indicated times. Cell lysates were prepared and Western blotted with anti-phospho-PKC-α antibodies. (i) A representative anti-phospho-PKC-α Western blot is displayed. (ii) Quantified data from three or four experiments are shown as mean ± SEM values. *, P < 0.05. (D) Increased phosphorylation of PKC-α in cells incubated with beads coupled to InlB. HeLa cells were incubated with medium alone (−) or with beads coated with InlB or GST for the indicated times. Cell lysates were prepared for detection of phospho-PKC-α. (i) Representative Western blot obtained with anti-phospho-PKC-α antibodies. (ii) Quantified results from three experiments are presented as mean ± SEM values. *, P < 0.05. The numbers to the right of the blots in panels Ai, Bi, Ci, and Di are molecular sizes (in kilodaltons).

PKC-α kinase activity is needed for efficient InlB-mediated uptake.

We used genetic and pharmacological approaches to determine if the kinase activity of PKC-α is needed for InlB-dependent entry. The genetic approach involved transient expression of a hemagglutinin (HA) epitope-tagged kinase-dead (KD) allele of PKC-α in HeLa cells. As controls, we employed HA-tagged wild-type PKC-α or tagged luciferase, which does not affect Listeria entry (8). InlB-mediated internalization into cells expressing the tagged proteins was assessed using a fluorescence microscopy-based approach, as previously described (8, 34). Beads coated with InlB were used for these experiments, since their efficient uptake makes them amenable to fluorescence microscopy studies (8, 10, 19). Importantly, HeLa cells expressing the tagged kinase-dead allele of PKC-α exhibited decreased levels of entry of beads compared to those for cells expressing tagged wild-type PKC-α or luciferase (Fig. 7A). We next used a chemical inhibitor of PKC enzymes, Go6976, to determine if the entry of Listeria depends on PKC activity. Treatment of HeLa cells with Go6976 caused a dose-dependent inhibition of uptake of the Listeria strain expressing InlB (Fig. 7Bi). Importantly, Go6976 treatment did not affect the viability of HeLa cells or reduce the level of entry of E. coli expressing invasin (Fig. 7B). Taken together, the results in Fig. 7 indicate a role for PKC-α kinase activity in InlB-dependent entry.

FIG 7.

The kinase activity of PKC-α participates in InlB-mediated entry. (A) Inhibition of InlB-dependent uptake by a kinase-dead allele of PKC-α. HeLa cells were transfected with plasmids expressing HA epitope-tagged wild-type PKC-α (PKC-α.wt), kinase-dead PKC-α (PKC-α.KD) containing an arginine substitution of a conserved lysine residue in the kinase domain (41, 42), or luciferase (HA-lucif.) as a control. At about 24 h after transfection, cells were solubilized for Western blotting or incubated with InlB-coated beads for assessment of entry. (i) A Western blot with anti-PKC-α antibodies is presented. A comparison of the extent of PKC-α immunoreactivity in the three lanes indicates that PKC-α.wt and PKC-α.KD are overexpressed relative to the level of expression of endogenous PKC-α, which is visible in lysates of cells expressing HA-luciferase. Adjacent to the Western blot images is a bar graph showing quantified blotting data as mean ± SEM values from three experiments. The numbers to the right of the blots are molecular sizes (in kilodaltons). *, P < 0.05 compared to the HA-luciferase condition. (ii) Entry results for InlB-coated beads are shown. Data are means ± SEMs from five experiments. *, P < 0.05 compared to the HA-luciferase or HA-PKC-α.wt condition. (B) Inhibition of InlB-mediated entry by a chemical inhibitor of PKC-α. HeLa cells were treated with the indicated concentrations of the compound Go6976 or with the vehicle DMSO for 45 min before assessment of bacterial entry or cell viability. (i) The effect of Go6976 treatment on entry of Listeria expressing InlB is shown. Data are means ± SEMs from three experiments. *, P < 0.05 compared to the DMSO-only condition. (ii) Go6976 treatment did not affect the viability of HeLa cells, as measured with MTT assays. Results are mean ± SEM values from three experiments. (iii) Go6976 treatment did not reduce the level of entry of E. coli strain HB101 expressing the invasin protein of Yersinia enterocolitica (inv⁺ HB101). Data are mean ± SEM values from three experiments.

PKC-α impacts remodeling of the actin cytoskeleton during entry.

InlB-coated beads were used to examine the roles of mTOR or PKC-α in modulation of the actin cytoskeleton during uptake. We first performed RNAi experiments to confirm that internalization of beads requires host mTOR and PKC-α, similar to the situation observed with Listeria (Fig. S2B). An siRNA against Met was included as a positive control. We then tested the effects of siRNAs targeting mTOR, PKC-α, or Met on F-actin remodeling during entry of beads. As controls, HeLa cells were mock transfected in the absence of siRNA or treated with a control nontargeting siRNA. As previously reported (8, 10, 19), under control conditions, incubation of HeLa cells for 5 min with InlB-coated beads induced accumulation of F-actin in cup-like structures (Fig. 8A). The degree of accumulation of F-actin around beads was quantified as fold enrichment (FE) values, as previously described (10, 19) (Fig. 8B). FE is defined as the average fluorescence intensity of F-actin around beads normalized to the average intensity throughout the cell. FE values greater than 1.0 indicate enrichment of actin filaments around beads. In cells treated with control siRNA, the mean FE value for F-actin accumulation was approximately 1.7 (Fig. 8B). As expected, Met siRNA substantially reduced F-actin accumulation, resulting in a mean FE value of 1.1. PKC-α siRNA caused a smaller yet statistically significant decrease in the FE value (mean value, 1.4). As previously reported (10), mTOR siRNA did not affect F-actin accumulation in a statistically significant fashion.

FIG 8.

PKC-α affects recruitment of F-actin during InlB-mediated entry. (A) Effect of siRNAs targeting mTOR, PKC-α, or Met on accumulation of F-actin around InlB-coated beads. HeLa cells transfected with siRNAs were incubated with InlB-coated beads for 5 min, fixed, and labeled for extracellular beads and F-actin. Representative images acquired by confocal microscopy are shown. (Left) Merged images of F-actin (green) and beads (red). The areas indicated with arrows are expanded in the remaining three columns of panels. Bars, 5 μm. (B) Quantification of the degree of accumulation of F-actin around beads. Fold enrichment (FE) values were determined as described in the Materials and Methods. Data are displayed as mean FE values ± SEMs from three experiments. *, P < 0.05 compared to the no siRNA or control siRNA conditions.

As a second approach to assess the impact of PKC-α on actin polymerization during InlB-mediated entry, we compared the effects of expression of HA-tagged wild-type or kinase-dead PKC-α on F-actin accumulation around InlB-coated beads (Fig. S3). Importantly, FE values for F-actin accumulation were higher in HeLa cells transiently expressing tagged wild-type PKC-α than in cells expressing kinase-dead PKC-α (Fig. S3A and Bi). Analysis of the mean pixel intensities for the two HA-tagged PKC-α alleles indicated similar levels of expression (Fig. S3Bii). Taken together, the results in Fig. 8 and S3 indicate that PKC-α positively regulates the actin cytoskeleton during InlB-mediated uptake.

DISCUSSION

In this work, we found that both host mTORC1 and mTORC2 control InlB-dependent entry of Listeria into human cells. InlB activated each of these mTOR-containing complexes in a manner that depends on the host receptor tyrosine kinase Met and the activity of PI3K. Importantly, the mTORC2 substrate PKC-α contributed to actin polymerization during InlB-dependent entry.

Although mTOR was needed for full InlB-induced phosphorylation of serine 657 in PKC-α (Fig. 6A), RNAi-mediated depletion of mTOR did not cause a statistically significant effect on the accumulation of F-actin during entry (Fig. 8). A possible explanation for this apparent discrepancy is that the residual phosphorylation of PKC-α in mTOR-depleted cells allows sufficient PKC-α activity to promote actin polymerization. Residual PKC-α phosphorylation could be due to the incomplete depletion of mTOR and/or kinases apart from mTOR that phosphorylate the hydrophobic motif in PKC-α.

How might PKC-α affect F-actin during InlB-mediated internalization? This serine/threonine kinase is known to phosphorylate several substrates that regulate the actin cytoskeleton, including proteins that bundle actin filaments (Filamin A, fascin), proteins that cap actin filaments (adducin), and proteins that link actin filaments to the plasma membrane (MARCKS, ezrin) (15, 35). In future work, it will be important to identify substrates of PKC-α that affect actin cytoskeletal remodeling during InlB-dependent entry.

While our study focused on the function of PKC-α in the InlB-mediated internalization of Listeria, we also identified roles for the mTORC1 effectors 4E-BP1 and HIF-1α in bacterial infection. 4E-BP1 is a translational repressor, and phosphorylation of this protein by mTORC1 induces its dissociation from the translation initiation factor eukaryotic initiation factor 4E (eIF4E) (11). The resulting effect is increased translation of many proteins, including HIF-1α, a transcription factor that promotes expression of several glycolytic enzymes (36). Interestingly, 4E-BP1 negatively controlled Listeria entry, whereas HIF-1α positively affected this process. The role for HIF-1α in the uptake of Listeria is intriguing and raises the possibility that the glycolysis promoted by this transcription factor provides energy for bacterial internalization. The negative role of 4E-BP1 in Listeria uptake could be due to suppressed expression of HIF-1α and/or other cellular proteins important for bacterial entry. Future studies should distinguish between these possibilities and provide a better understanding of how 4E-BP1 and HIF-1α control Listeria infection.

To the best of our knowledge, our work is the first to demonstrate a role for mTOR in the internalization of a bacterial pathogen into human cells. However, two previous reports indicated that mTORC1 promotes intracellular replication of the Gram-negative bacterium Salmonella enterica serovar Typhimurium (37, 38). These studies demonstrated that intracellular Salmonella avoids autophagic killing by stimulating the recruitment and activation of mTORC1 on vacuoles containing bacteria. Given the number and diversity of biological processes controlled by mTOR (11), it seems likely that future work will uncover additional examples of bacterial exploitation of mTORC1- or mTORC2-mediated pathways.

MATERIALS AND METHODS

Bacterial strains, mammalian cell lines, and media.

Listeria monocytogenes strains BUG 947 and BUG 949 were grown in brain heart infusion (BHI; Difco) broth and prepared for infection as described previously (7). BUG 947 contains an in-frame deletion in the inlA gene and is internalized into mammalian cells in a manner dependent on the Listeria protein InlB and its host receptor, Met (3, 8, 39). BUG 949 is isogenic with BUG 947 and has in-frame deletions in both the inlA and inlB genes (39). Escherichia coli strain HB101 containing a plasmid expressing the invasin gene from Yersinia enterocolitica (inv⁺ HB101) (20, 40) was grown and prepared for infection as described previously (10).

Cells of the human epithelial cell line HeLa (ATTC CCL-2) were grown in Dulbecco's modified Eagle's medium (DMEM) with 4.5 g of glucose per liter and 2 mM glutamine (catalog no. 11995-065; Life Technologies) supplemented with 5 or 10% fetal bovine serum (FBS). Cell growth, bacterial infections, incubations with latex beads, and stimulation with the InlB protein were performed at 37°C under 5% CO₂.

Antibodies, inhibitors, and other reagents.

Rabbit antibodies against Akt (catalog no. 9272), mLST8 (catalog no. 3274), mTOR (catalog no. 2972), phospho-Ser473-Akt (catalog no. 9271), phospho-Thr389-p70S6K (catalog no. 9205), phospho-PKC (pan) (catalog no. 9371), Raptor (catalog no. 2280), and Rictor (catalog no. 2140) were all obtained from Cell Signaling Technology. Rabbit polyclonal anti-InlB antibodies were previously described (3). The mouse monoclonal antibodies used were anti-glutathione S-transferase (anti-GST; catalog no. G1160; Sigma-Aldrich), anti-hemagglutinin (anti-HA; catalog no. MMS-101P; Covance), anti-PKC-α (catalog no. sc-8393; Santa Cruz Biotechnology), and antitubulin (catalog no. T5168; Sigma-Aldrich). Horseradish peroxidase (HRPO)-conjugated secondary antibodies were purchased from Jackson Immunolabs. Secondary antibodies or phalloidin coupled to Alexa Fluor 488, Alexa Fluor 555, or Alexa Fluor 647 were obtained from Life Technologies. The inhibitors LY294002, Torin1, and okadaic acid were from Sigma-Aldrich. Go6976 was obtained from Selleckchem.

siRNAs.

The sequences of short interfering RNAs (siRNAs) used in this study are listed in Table S1 in the supplemental material. These siRNAs were obtained from Sigma-Aldrich or Ambion. The negative, nontargeting control siRNA molecule 1 (catalog no. D-001210-01) was purchased from Dharmacon. This siRNA has two or more mismatches with all sequences in the human genome, indicating that it should not target host mRNAs.

Plasmids.

The mammalian expression vectors used were HA-luciferase (34), HA–wild-type PKC-α (PKC-α.wt; a gift of B. Weinstein; catalog no. 21232; Addgene), and HA–kinase-dead PKC-α (PKC-α.KD; a gift of B. Weinstein; catalog no. 21235; Addgene) (41). The plasmids pGEX-3T (GE Healthcare) and pET28a-InlB (7) were used for expression in E. coli of glutathione S-transferase (GST) and 6×His-tagged InlB proteins, respectively.

Transfection.

HeLa cells grown in 24-well plates or on 22- by 22-mm coverslips were transfected with siRNAs or plasmid DNA using the Lipofectamine 2000 reagent (Life Technologies) as previously described (8, 9).

Quantitative PCR (qPCR) analysis.

HeLa cells were processed for RNA isolation at approximately 48 h after siRNA transfection. RNA extraction, cDNA synthesis, and real-time PCR were performed using an ABI 7900 instrument (Applied Biosciences) as previously described (10). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene expression was assessed as an endogenous control. The TaqMan probes from Life Technologies used were Hs00607050_m1 (EIF4EBP1 gene), Hs99999905_m1 (GAPDH gene), Hs00153153_m1 (HIF1A gene), Hs010088691_m1 (SREBF1 gene), Hs01020387_m1 (TSC2 gene), and Hs00177504_m1 (ULK1 gene). Data were analyzed by the comparative threshold cycle (C_T) method, normalizing the C_T values for target gene expression to those for GAPDH. Relative quantity (RQ) values were calculated by the formula RQ = 2^−ΔΔCT. To obtain the relative expression values shown in Fig. 2Di and 5Ai, RQ values in a given experiment were normalized to the values in cells mock transfected in the absence of siRNA (no siRNA condition).

Western blotting.

HeLa cells were solubilized in radioimmunoprecipitation assay (RIPA) buffer (1% Triton X-100, 0.25% sodium deoxycholate, 0.05% SDS, 50 mM Tris-HCl [pH 7.5], 2 mM EDTA, 150 mM NaCl, 1 mM phenylmethylsulfonyl fluoride, 1 mM sodium orthovanadate, 10 mg/liter each of aprotinin and leupeptin). In the cases of experiments involving siRNA transfection (Fig. 2 to 7), HeLa cell lysates were prepared at about 48 h posttransfection. Supernatants remaining after centrifugation at 12,000 rpm were used to quantify the protein concentrations using a bicinchoninic acid kit (Pierce). Equal protein quantities of each sample were migrated on SDS-polyacrylamide gels and transferred to polyvinylidene difluoride (PVDF) membranes. Incubations with primary and secondary antibodies coupled to horseradish peroxidase and detection using enhanced chemiluminescence (ECL) or ECL Plus reagents (GE Health Care) were performed as described previously (3). Film was used for detection of anti-phospho-Akt blots in Fig. 4A, whereas an Odyssey imaging system (Li-Cor Biosciences) was used for detection of all other Western blots.

Western blotting data were quantified using ImageJ (version 1.51e) software. Integrated pixel densities in bands corresponding to proteins of interest were measured, and the background was subtracted. Integrated pixel densities for loading controls were also determined and corrected for the background. For data involving phosphorylated Akt or PKC-α in Fig. 3B, 4A and B, and 6A, C, and D, the loading controls were total Akt or PKC-α. For all other Western blots, tubulin served as the loading control. Background-corrected integrated pixel densities of the protein of interest were normalized to those of the loading control.

Purification and coupling of proteins to latex beads.

Expression in E. coli and purification of 6×His-tagged InlB or glutathione S-transferase proteins were performed as previously described (7, 25).

InlB or GST proteins were coupled to carboxylate-modified latex beads 3 μm in diameter (catalog no. 09850; Polysciences) using either passive binding or covalent linkage. Coupling by passive binding was performed as described previously (8, 10). Covalent coupling using 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDAC) was carried out according to the bead manufacturer's protocol. After coupling, the beads were washed four times in phosphate-buffered saline (PBS) with 1% bovine serum albumin (BSA) and stored at 4°C until use.

Bacterial entry or association assays.

Entry of Listeria or inv⁺ HB101 was assessed using gentamicin protection assays, as previously described (3, 10). Cells were infected for 1 h in the absence of gentamicin and then incubated in DMEM with 20 μg/ml gentamicin (Listeria) or 100 μg/ml gentamicin (inv⁺ HB101) for an additional 2 h. Multiplicities of infection (MOIs) were approximately 30:1 for Listeria and 100:1 for inv⁺ HB101. Assays to measure the association of Listeria with host cells (Fig. S1) were performed by infecting HeLa cells for 30 min in the absence of gentamicin, as described previously (8). In the case of experiments with siRNAs (Fig. 2, 3, and S1), HeLa cells were infected with Listeria at approximately 48 h after transfection. For entry experiments involving chemical inhibitors (Fig. 1 and 7B), cells were pretreated with 10 μM LY29400, 25 to 250 nM Torin1, 25 to 100 nM Go6976, or the vehicle dimethyl sulfoxide (DMSO) about 45 min before bacterial infection. Inhibitors were also included during the 1 h of incubation with bacteria in medium lacking gentamicin but omitted during the following 2 h of incubation in medium with gentamicin. Bacterial entry efficiencies were first expressed as the percentage of the inoculum that survived gentamicin treatment. Association efficiencies were expressed as the percentage of the inoculum recovered in the absence of gentamicin treatment. To obtain relative entry or association values, absolute percent entry values in a given experiment were normalized to the value for cells subjected to control conditions. For experiments with chemical inhibitors (Fig. 1 and 7B), the control condition was DMSO treatment. In the case of all other experiments, the control was mock transfection in the absence of siRNA.

Quantification of internalization of beads.

For the experiments whose results are shown in Fig. 7A and S2, HeLa cells growing on 22- by 22-mm coverslips were transfected with siRNAs for approximately 48 h. For the studies whose results are shown in Fig. 7A, HeLa cells on coverslips were transfected with plasmid DNA allowing expression of HA epitope-tagged luciferase, wild-type PKC-α, or kinase-dead (KD) PKC-α for about 24 h. After transfection, cells were incubated for 30 min at 37°C in 5% CO₂ with beads coupled to the InlB or GST protein. The ratio of particles to human cells was approximately 5:1. Cells were then washed in PBS and fixed in PBS containing 3% paraformaldehyde. Samples were labeled using a previously described approach that distinguishes extracellular or intracellular particles (8, 34). Labeling of beads was performed with polyclonal anti-InlB antibodies or monoclonal anti-GST antibodies. For the experiments whose results are shown in Fig. 7A, HA-tagged proteins were also labeled using anti-HA antibodies after permeabilization of fixed cells in PBS with 0.4% Triton X-100. Only HeLa cells expressing HA-tagged proteins were assessed for uptake of beads. Labeled samples were analyzed for intracellular and extracellular beads using an Olympus BX51 epifluorescence microscope. The data shown in Fig. 7A or S2 are from three experiments. In each experiment, at least 100 intracellular beads were scored for the control conditions involving HA-luciferase (Fig. 7A) or no siRNA (Fig. S2). A similar number of total (intracellular plus extracellular) beads was analyzed for all other conditions. The data were initially expressed as the percentage of total cell-associated beads that were internalized. Results shown in Fig. 7A or S2 are relative internalization values obtained by normalizing the percent internalization values to those obtained under HA-luciferase or no siRNA control conditions.

Measurement of viability of HeLa cells.

HeLa cells were incubated for 45 min in serum-free DMEM with 25 to 1,000 nM Torin 1, 100 to 250 nM Go6976, or 0.1% DMSO as a control. Assays with 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) were performed as previously described (43), except that cells were incubated with 1 mg/ml of MTT for 30 min at 37°C in 5% CO₂, MTT formazan reaction products were solubilized in 100% DMSO, and the absorbance was measured using a wavelength of 600 nm. In order to obtain relative viability values, absorbance values were normalized to those for cells treated with DMSO.

Analysis of phosphorylation of Akt, p70S6K, or PKC-α.

For experiments to assess the phosphorylation of Akt or p70S6K, HeLa cells were serum starved for 3 to 4 h. For studies with the phosphorylation of PKC-α, cells were serum starved for 9 to 10 h. After starvation, cells were incubated at 37°C in 5% CO₂ with 4.5 nM soluble InlB protein, Listeria, or beads coated with InlB or GST. The ratio of Listeria or beads per human cell was approximately 100:1 or 5:1, respectively. After stimulation, the cells were washed in cold PBS and solubilized in RIPA buffer containing 1 μM okadaic acid. Samples were migrated on SDS-polyacrylamide gels, and the phosphorylation of Akt, p70S6K, or PKC-α was detected by Western blotting using phospho-specific antibodies.

Confocal microscopy analysis.

For studies of recruitment of F-actin to beads, HeLa cells were grown on 22- by 22- mm coverslips and transfected with siRNAs for approximately 48 h. Cells were then washed and placed in serum-free DMEM. Beads coupled to InlB or GST were added to the cells at a ratio of about 5 particles per human cell. The cells were centrifuged at 1,000 rpm for 2 min to enhance contact between beads and HeLa cells and then incubated for 5 min at 37°C in 5% CO₂. Cells were washed in PBS, fixed in PBS with 3% paraformaldehyde, permeabilized, and labeled for extracellular beads or F-actin using phalloidin-Alexa Fluor 555 as described previously (10).

Samples for analysis of recruitment of F-actin, the results of which are shown in Fig. 8, were imaged using a Zeiss LSM 710 or Olympus FV1200 laser scanning confocal microscope. Images from serial sections spaced 1.0 μm apart were used to ensure that all cell-associated beads were detected. ImageJ (version 1.51e) software was employed to determine the fold enrichment (FE) values for each cell-associated bead. FE is defined as the mean pixel intensity in a ring-like structure around the bead normalized to the mean pixel intensity throughout the human cell (10, 19). The thresholding function of ImageJ was used to select ring-like structures of F-actin around the beads. This function was also used to measure the mean pixel intensity throughout the cell. Approximately 100 to 250 extracellular, cell-associated beads were analyzed for F-actin recruitment for each condition in each experiment. The data shown in Fig. 8B are mean FE values ± standard errors of the means (SEMs) from three or four experiments.

Experiments measuring F-actin recruitment in HeLa cells transiently expressing HA-tagged PKC-α proteins (Fig. S3) were performed similarly to those described above, except that F-actin was assessed only in cells that detectably expressed the tagged proteins. Gain settings for photomultiplier detectors were kept constant in order to allow analysis of F-actin in cells expressing similar levels of HA-tagged wild-type or kinase-dead PKC-α. Expression of the tagged PKC-α alleles (Fig. 3Bii) was quantified by using ImageJ software to measure the mean pixel intensities of HA fluorescence in the same cells used for analysis of F-actin recruitment.

Statistical analysis.

Statistical analysis was performed using Prism software (version 6.0c; GraphPad Software). In comparisons of data from three or more conditions, analysis of variance (ANOVA) was used. The Tukey-Kramer test was used as a posttest. For comparisons of two data sets, an unpaired Student's t test was used. A P value of 0.05 or lower was considered significant.