S. Typhimurium SL1344 (S. Tm) WT (17), S. Tm ΔinvA, S. Tm ΔsopB, S. Tm ΔsopE/E2, S. Tm ΔavrA and S.Tm eGFP (pFPV25.1) (18) were grown overnight at 37°C with shaking in lysogeny broth (LB) supplemented with 100 μg/ml streptomycin, 100 μg/mL ampicillin, or 50 μg/mL kanamycin, when appropriate.

Mouse embryonic fibroblast (MEFs) WT, MK2 deficient cells transduced with retroviral constructs pMMP–IRES–GFP (vector), MK2 deficient cells transduced with retroviral constructs pMMP–MK2–IRES–GFP (MK2), MK2 deficient cells transduced with retroviral constructs pMMP–MK2^KR–IRES–GFP (MK2-K79R), kinase-dead MK2 mutants in which lysine79 in catalytic subdomain I is replaced with arginine, MK2 deficient HEK293T cells (MK2^-/-), MK2 sufficient HEK293T (MK2^+/+), immortalized bone marrow derived macrophages (iBMDM) from MK2 deficient cells transduced with retroviral constructs pMMP–IRES–GFP (vector), MK2 deficient cells transduced with retroviral constructs pMMP–MK2–IRES–GFP (MK2). The generation of MEF cell line was done by cotransfection of primary MEFs with pSV40Tag encoding simian virus 40 large T-antigen and pREP8 plasmid (Invitrogen) in a 10:1 mixture. Colonies were selected with 3 mM histidinol (Sigma) (19). Immortalized murine macrophages (iBMDMs) were established from bone marrow-derived macrophages (BMDM) that were isolated from mice deficient for MK2 or from wild type mice by retroviral infection with a virus encoding both the v-Raf and the v-Myc oncogenes as described in detail previously (20). Bone marrow derived macrophages (BMDM) were isolated from Mk2^+/+ and Mk2^-/- mice were cultivated as described (21). All cells were cultured in DMEM GlutaMAX (DMEM-Life Technologies) supplemented with 10% (v/v) fetal bovine serum and 1x Penicillin Streptomycin.

Cells as indicated were seeded in 24-well plates (10⁵ cells/well) [for bacterial count and immunofluorescence analysis (IF)] or 6- well plates (4x10⁵ cells/well) [for western blot (WB)] and incubated overnight in 5% CO₂ at 37°C. The next day, culture medium was replaced with DMEM without Penicillin Streptomycin and indicated substance such as DMSO, the pan-p38 inhibitor BIRB-796 (BIRB, 1µM), the MK2 inhibitor PF-3644022 (PF22, 5µM), the autophagy inhibitor Bafilomycin A1 (BafA1, 20nM), the mTOR inhibitor Rapamycin (1µM), the TBK1 inhibitor BX795 (1µM), or TNF-α (20ng/mL) for at least one hour before infection. Cells were then infected with different Salmonella strains at MOI 10, 100, 500, 1000 (as indicated) for 1 hours. For IF and WB, cells were infected at MOI 100. Cells were washed 3 times with PBS and culture medium was replaced with medium supplemented with 100 μg/mL gentamicin (Sigma) and indicated substances as above for 1 hours to kill extracellular bacteria. For 2 hours post infection time point, cells were lysed in PBS containing 1% (v/v) Triton X-100 and 0.1% (v/v) sodium dodecyl sulfate (SDS) (for bacterial counts) or were fixed with 4% PFA (for IF) or lysed in with Laemmli loading buffer (for WB). The cell lysates were then serially diluted in PBS and plated on LB agar for colony-forming unit (CFU) count. For later time points, culture medium was replaced with a medium supplemented with 10 μg/mL gentamicin (Sigma).

Fixed cells were washed 3 times with PBS, and permeabilized with Triton-X100 (0.25%) in PBS for 30 min at RT. Blocking was done using 4% bovine serum albumin (BSA) in PBS (blocking buffer) for 1 h at 4°C. Cells were then incubated with first antibodies in blocking buffer and incubated overnight at 4°C. The following first antibodies were used: LC3 (mbl international), p62 (Abcam), Salmonella O Antiserum Group B (BD Difco), Salmonella Typhimurium (Clone: B2363M, Meridian Life Science), LAMP-1 (eBioscience) and Anti-Ubiquitinylated proteins Antibody, clone FK2 (Sigma). After 3 times washing with PBS, secondary antibodies were added and incubated for 1h at RT. The following secondary antibodies were used: Alexa (Ax) 488- conjugated goat anti-mouse IgG, Ax488-conjugated goat anti-rabbit IgG, Ax555-conjugated goat anti-rabbit IgG, Ax568-conjugated goat anti-mouse IgG, Ax568-conjugated goat anti-rat IgG (Thermofisher Scientific). Phalloidin-iFluor 647 (1:1000) (Abcam) and DAPI (Invitrogen) were applied to visualize F-actin and nuclei, respectively. Images were obtained on a Cytation 1 (BioTek) at 20x magnification in 5-10 fields of view per samples using Gen 11 software (BioTek). Brightness and contrast were adjusted using ImageJ 1.53k software. Gen 11 software (BioTek) was used to quantify median fluorescence intensity (MFI) colocalization of LC3 and p62 with S.Tm eGFP and percentage of colocalization of LAMP-1 and ubiquitinated protein with Salmonella. GFP staining with size of 0.5 – 5 µm was defined as primary mask to identify the bacteria, then the mask was expanded for 1 µm. For p62 and LC3 colocalization with Salmonella, MFI of Ax555 (LC3) and Ax568 (p62) inside the primary mask were measured. The results are defined as the median of Ax555 or Ax568 fluorescence intensity colocalized with STm-GFP. For LAMP-1 and ubiquitin colocalization with Salmonella, MFI of Ax568 inside the primary mask were measure. The threshold was determined from highest MFI of the bacteria completely absence of LAMP-1 or ubiquitin staining. Furthermore, the percentage of LAMP-1 or ubiquitin positive staining were calculated and determined. The results are defined as percentage of LAMP-1 or ubiquitin positive staining to total bacteria.

Cells were directly lysed by boiling in 2x SDS-PAGE loading buffer then separated by SDS-PAGE and transferred to polyvinylidene difluoride (PVDF) membrane. After blocking with 3% milk in PBS with 0.1% Tween 20, membranes were incubated with primary antibody in PBS with 3%BSA and 0.1% Tween 20. The following primary antibodies were used: LC3 (MBL International), p62 (Abcam), p-p62 Ser403 (MBL International), TBK1/NAK, p-TBK1/NAK Ser172, Atg5 (D5F5U), p-MK2 (Phospho-MAPKAPK-2) Thr222, MK2 (MAPKAPK-2) (Cell Signaling Technology), EF-2 (Santa Cruz Biotech), GAPDH (Abcam). Immunoreactive bands were visualized using appropriate secondary antibodies and enhanced chemiluminescence detection reagents. Images were obtained using the ImageQuant LAS 3000 system.

Data were analyzed using Prism V7.0d software (GraphPad). Statistical analysis was done using one-way analysis of variance (ANOVA) followed by Tukey’s multiple comparison test or T-Test as indicated. Graphs display the mean values ± SD, unless stated otherwise.

To investigate the role and detailed mechanisms of p38^MAPK/MK2 pathway in cellular response to Salmonella infection, we first performed the gentamicin protection assay for infection efficacy. Therefore, we infected murine embryonic fibroblast (MEFs) with S. Typhimurium for 1h, treated cells with gentamicin to inactivate extracellular bacteria, lysed cells and counted bacteria numbers at 2 and 6 hours post infection (p.i.). Interestingly, immunofluorescence analysis of individual cells infected with S. Typhimurium expressing GFP showed increased bacterial number after p38^MAPK inhibitor BIRB796 (BIRB) and MAPKAPK2 (MK2) inhibitor PF-3644022 (PF22) treatment at 2h p.i. and 6h p.i. ( Figures 1A, B ). Furthermore, we observed increased S. Typhimurium counts in the presence of the BIRB and the PF22 ( Figure 1C ). To convincingly prove that MK2 contributes to S. Typhimurium control, we performed similar experiments in MK2-deficient MEFs. Consistent with the findings from pharmacological inhibition of p38/MK2 pathway, cells lacking MK2 showed higher intracellular counts of GFP-labelled bacteria compared to cells rescued with retroviral expression of MK2 ( Figures 1D, E ). Furthermore, MK2-deficient MEFs and MEFs expressing a kinase-dead MK2 mutant in which lysine 79 in the ATP-binding catalytic subdomain I is replaced with arginine (19) showed higher bacterial counts compared to wild type Mk2-rescued MEFs at 2 hours and 6 hours p.i. ( Figure 1F ). Interestingly, unlike at later time points, at 90 minutes p.i., MK2-deficient and wild-type Mk2-rescued MEFs showed similar bacterial counts ( Supplementary Figure 1A ). This indicates that there are similar levels of invasion of Salmonella in both cell lines, however the differences arise after cell entry.

Macrophages are important hosts for intracellular Salmonella and hence we also verified our findings in primary bone-marrow-derived macrophages (BMDMs). We used immortalized bone marrow-derived macrophages (iBMDM) with a retroviral MK2 rescue and obtained similar results at 4 hours p.i ( Supplementary Figure 1B ). Furthermore, we generated and compared BMDMs from MK2-deficient and wild-type mice in similar assays and observed higher intracellular Salmonella counts in MK2-deficient BMDMs ( Supplementary Figure 1C ). Moreover, similar results were observed in MK2-deficient HEK293T cells compared to wild-type control cells (MK2 WT) at 2 hours p.i. ( Supplementary Figure 1D ). Thus, our findings show that p38^MAPK/MK2-mediated protein phosphorylation modulates the host cell susceptibility to Salmonella infection in different relevant cell types.

After entry into the host cell, Salmonella resides within a modified phagosomal compartment termed the Salmonella-containing vacuole (SCV) that acquire late endosomal markers, such as LAMP-1. Most intracellular Salmonella remain and replicate in the SCV. However, a small percentage of bacteria can damage the SCV and escape into the host cytosol. Cytosolic Salmonella are rapidly recognized by the ubiquitin E3 ligases resulting in the formation of ubiquitin chain layer surrounding the bacteria, marked by autophagy adaptor molecules for later downstream autophagy activation to eliminate the bacteria. Meanwhile, Salmonella residing in damaged SCVs can also be targeted for autophagic elimination (9). To investigate whether p38^MAPK/MK2 signaling affects subcellular localization of intracellular bacteria, we infected MEFs with S. Typhimurium-GFP and analyzed the samples by co-immunofluorescence microscopy using an anti-LAMP1 antibody. At 2 h p.i., a higher percentage of bacteria colocalized with LAMP-1 (SCV) in MK2 expressing MEFs (93.4%) compared to MK2-deficient MEFs (85.9%) ( Figure 2A ). We also observed a higher percentage of Salmonella colocalized with ubiquitin in MK2^-/- MEFs (11.23%) suggesting that higher abundance of cytosolic Salmonella in Mk2-deficient MEFs compared to Mk2-rescued MEFs (2.33%) ( Figure 2B ).

Subsequently, we used a combination of chloroquine (CHQ) resistance assay and gentamicin protection assay to quantitatively determine cytosolic Salmonella counts. CHQ is a lysosomotropic agent that accumulates only within the SCV and leads to the killing of vacuolar bacteria by a still unknown mechanisms (22). MK2-deficient and MK2^KR -rescued MEFs showed higher viable Salmonella counts compared to wild-type Mk2-rescued MEFs at 2 hours p.i. with or without CHQ treatment ( Figure 2C ), indicating that MEF cells lacking MK2 activity accumulate higher cytosolic Salmonella load.

Autophagy, known for the degradation of cytosolic components in eukaryotic cells, plays an important role in recognizing and protecting cells from invading intracellular pathogens such as Salmonella. To understand the mechanisms of MK2 mediated control of susceptibility to Salmonella infection, we compared the levels of lipidated LC3 (LC3-II) levels between MK2-deficient and expressing MEFs infected with S. Typhimurium. We found that LC3-II levels after infection were lower in Mk2^-/- MEFs compared to Mk2-rescued MEFs ( Figure 3A ). Non-lipidated LC3 (LC3-I) was also upregulated upon infection in both cell lines, with the MK2-rescued MEFs showing slightly increased levels. Mechanisms involved in this upregulation are unknown, but MAPK pathways have been previously reported to regulate LC3-I levels (23). p62 (also known as SQSTM1) is an autophagy receptor and substrate with multiple protein-protein interaction domains, including an ubiquitin-associated (UBA) domain for ubiquitinated cargo binding and an LC3 interaction region (LIR) for binding to LC3. Similar amount of SQSTM1/p62 were detected in the presence or absence of MK2 ( Figure 3A ). Interestingly, p62 phosphorylation at Ser403 was increased in Mk2-rescued MEFs at 2 and 4 h p.i and decreased again at 6 h p.i., even when compared to uninfected MEFs. In contrast, Mk2^-/- MEFs showed lower p62 phosphorylation at Ser403 compared to Mk2-rescued MEFs at all-time points. In addition, no differences were detected in the levels of ATG5, a protein essential for the early steps in autophagosome formation ( Figure 3A ).

We then used immunofluorescence microscopy to specifically look at autophagic clearance of Salmonella. Co-immunofluorescence staining of MEFs infected with Salmonella expressing GFP showed reduced colocalization of Salmonella with LC3 in Mk2^-/- MEFs in comparison to Mk2-rescued MEFs at 2 h p.i. ( Figures 3B, C ). Similar results were observed in the colocalization of Salmonella with the autophagy receptor p62/SQSTM1, known to be essential for efficient autophagy of bacteria ( Supplementary Figures 2A, B ). Inhibition of autophagy with bafilomycin A1 (BafA1) led to increased viable Salmonella counts in MK2-expressing MEFs, comparable to that in untreated MK2-deficient cells. This indicates that the positive role of MK2 in restricting Salmonella is probably via modulation of autophagy ( Figure 4A ). Consistently, there was no further increase in bacterial counts observed upon treating MK2-deficient cells with BafA1. Immunofluorescence analysis of individual cells infected with S. Typhimurium expressing GFP showed increased bacterial numbers after BafA1 treatment at 2h p.i. ( Figures 4B, C ). Furthermore, immunofluorescence staining also showed reduced colocalization of Salmonella with LC3 puncta in BafA1 treated cells ( Figures 4C, D ). Interestingly, activation of autophagy by treatment with the mTOR inhibitor rapamycin was sufficient to reduce the bacterial counts in MK2^-deficient MEFs to levels comparable to that in MK2 expressing cells ( Figure 4E ). This clearly supports a model where p38/MK2 plays essential role in activating selective autophagy of bacterial cells.

It is known that Salmonella injects its effector proteins into the host cell cytosol by type III secretion system (T3SS) to avoid the autophagy pathway and for remodeling of the SCV for intracellular survival (24). The T3SS-1 translocated effectors play an important role to influence the host’s cellular functions to facilitate Salmonella invasion and mediate their persistence in the host cells (7). AvrA is one of the Salmonella effectors secreted by the SPI-1 T3SS and is known to inhibit the host c-Jun N-terminal kinase (JNK)/AP-1 and NF-κB signaling pathways (25) and to suppress autophagy by reducing Beclin-1 protein (26). Therefore, we then infected MEFs with S. Typhimurium ΔavrA and other SPI-1 mutants such as ΔinvA, ΔsopB, ΔsopE/ΔsopE2. MEFs infected with S. Typhimurium ΔavrA showed similar bacterial counts in both Mk2^-/- and Mk2-rescued MEFs at 2h p.i. ( Figure 4F ) whereas, other SPI-1 mutants showed increased S. Typhimurium counts in Mk2-deficient MEFs at 2h p.i. compared to Mk2-rescued MEFs, similar to the S. Typhimurium WT strain infected MEFs ( Figure 4F ). These data support the idea that p38^MAPK/MK2 signaling modulates the host cell susceptibility to Salmonella infection by affecting the autophagy pathways downstream to AvrA.

It has been known for a long time that MK2 is essential for tumor necrosis factor-α (TNF) biosynthesis (13). Since a previous study also showed that TNF-stimulated p62-mediated autophagic activity restricted survival of Shigella and Listeria ActA mutant (27), we infected MEFs with S. Typhimurium after TNF treatment. Upon treatment with exogenous TNF, we observed decreased Salmonella counts in Mk2^-/- MEFs in comparison to untreated Mk2^-/- MEFs ( Figure 5A ). Immunofluorescence analysis in TNF treated S. Typhimurium infected Mk2^-/- MEFs showed reduced bacterial number in comparison to untreated MEFs ( Figures 5B, C ). Hence, addition of exogenous TNF can stimulate cellular response, rescues MK2-deficiency and restricts the survival of S. Typhimurium.

TANK-binding kinase-1 (TBK1), an integral component of Type I interferon induction by microbial infection, is both important for autophagic maturation by coordinating assembly and function of the autophagic machinery (28) and protection of vacuolar integrity during intracellular bacterial infection (29). Catalytically active TBK1 is recruited to cytosol-invading Salmonella and then induce anti-bacterial autophagy via the upstream autophagy regulator WIPI2 (30) and phosphorylation of the autophagic adaptor p62 at Ser-403 (31). Interestingly, IKKϵ and TBK1 are also activated in cells after stimulation with TNF (32). Since we detected reduced p62-S403 phosphorylation in Salmonella infected MK2-deficient cells ( Figure 3A ), we employed western blot to determine whether MK2 signaling affects TBK1 activation. Similar levels of TBK1 were observed in all time points for both genotypes. However, activated TBK1 (pS172-TBK1) was higher in Mk2-rescued MEFs after infection in comparison to Mk2^-/- MEFs ( Figure 6A ). In addition, MK2-K79R MEFs showed a level of pS172-TBK1 similar to Mk2-rescued MEFs after infection indicating that a rescue of the p38^MAPK level by expression of MK2 protein as described in Kotlyarov et al. (33) and Ronkina et al. (19) is sufficient for TBK1 activation ( Supplementary Figure 3 ). Furthermore, treatment with the TBK1 inhibitor BX795 resulted in increased Salmonella counts in MK2 expressing MEFs ( Figure 6B ) indicating a role for TBK1 in restricting Salmonella infection. In agreement with this finding, immunofluorescence microscopy of BX795 treated S. Typhimurium infected MEFs showed increased bacterial number in comparison to untreated S. Typhimurium infected cells ( Figures 6C, D ). These data indicate that a p38^MAPK/MK2-dependent activation of TBK1 during bacterial infection is essential to restrict Salmonella infection and there is a defect in TBK1-regulated bacterial clearance in Mk2^-/- MEFs.