RAW264.7 cells and HEK293T cells were grown in complete Gibco^® RPMI Media 1640 and Gibco^® DMEM (Invitrogen), respectively, supplemented with 10% heat-inactivated fetal bovine serum (Invitrogen) and 1% penicillin/streptomycin (Invitrogen). The pcDNA control and DUSP12 stably overexpressed cell lines were grown in complete RPMI medium containing 500 mg/ml G418 (Sigma). All cells were grown in humidified incubator at 37°C with 5% CO₂.

The pcDNA3.1-DUSP12 and pcDNA3.1 empty vectors were transfected into RAW264.7 cells using Lipofectamine LTX reagents (Invitrogen) following the manufacturer’s protocol. Transfected cells were maintained in complete medium containing 500 mg/ml G418 and the G418 containing medium was replenished for every three days. After two weeks of selection, individual colony was isolated and subjected to screening via quantitative real-time PCR (qRT-PCR) and Western blot to obtain DUSP12 overexpressing cells. Primers used to clone DUSP12 into pcDNA vector: forward primer 5′-TGAAGGAAGTGGGCCAATAG-3′ and reverse primer 5′-GACCACAGGAGCACTGTTCA-3′.

FLAG-tagged constructs were made using the plasmid pXJ40 backbone (26). The following primers were used to construct the FLAG-tagged proteins: forward primer 5′-CCGGATCCATGATGGCGAAAAAGC-3′ and reverse primer5′-GGGAATTCTCATGCATTGTGTGGATTTTTCTTAAAC-3′ for full-length STAP1; forwards primer 5′-GGGGATCCATGGCCACAGCCCTGA-3′ and reverse primer 5′-GGGAATTCTCAGTGTTCCAGTGCCCTCC-3′ for full-length STAP2.

2 × 10⁵ RAW264.7 cells were cotransfected with 50 ng AP-1 reporter construct, 400 ng pcDNA-vector, -DUSP12, and -DUSP12-PM together with 10 ng of pRL-null plasmid using Lipofectamine LTX reagent (Invitrogen). Equal amounts of pcDNA3.1 empty vector were transfected as control. Reporter activity was determined with Promega Dual Luciferase Assay System (Promega). Firefly luciferase values were normalized for transfection efficiency by means of the Renilla luciferase activity that is constitutively expressed by pRL-null.

Total mRNA was extracted using Trizol (Invitrogen) and cDNA was synthesized using ImProm-II™ Reverse Transcription System (Promega). qRT-PCR was performed using CFX connect Real Time System. Relative expression levels of DUSP12, TNF-α, IL-6, IL-1β, MCP-1, and IL-10 were calculated by ΔΔCt method and normalized to β-actin. The following mouse specific primers were used: forward primer 5′-TGAAGGAACTGGGCCAATAG-3′and reverse primer 5′-GACCACAGGAGCACTGTTCA-3′ for DUSP12, forward primer 5′-TCCCAGGTTCTCTTCAAGGGA-3′ and reverse primer 5′-GGTGAGGAGCACGTAGTCGG-3′for TNF-α, forward primer 5′-GATGCTACCAAACTGGATATAATC-3′and reverse primer 5′-TGTACTCCAGGTAGCTATG-3′ for IL-6, forward primer 5′-CAACCAACAAGTGATATTCTCCATG-3′ and reverse primer 5′-ATCCACACTCTCCAGCTGCA-3′ for IL-1β, forward primer 5′-CTCAGCCAGATGCAGTTAACGCCC-3′reverse primer 5′-GGTGCTGAAGACCTTAGGGCAGAT-3′for MCP-1, forward primer 5′-GCTCTTACTGACTGGCATGAG-3′ and reverse primer 5′-CGCAGCTCTAGGAGCATGTG-3′ for IL-10, forward primer 5′-AGCACTGTGTTGGCATAGAGGTC-3′reverse primer 5′-CTTCTTGGGTATGGAATCCTGTG-3′for β-actin.

The concentrations of IL-6 and TNF-α in culture supernatants were determined by IL-6 and TNF-α ELISA kits (BD Pharmingen™). MCP-1 concentration was determined using the MCP-1 ELISA kit (eBioscience).

Proteins were harvested using protein lysis buffer (1 M Tris–Cl, 5 M NaCl, 100% NP-40) containing 1× Halt Protease and Phosphatase Inhibitor Cocktail and 0.5 M EDTA (Thermo scientific). Protein samples were subjected to SDS-PAGE and Western blot was performed using the following antibodies: anti-DUSP12 (Santa Cruz, 1:500 dilution), anti- beta actin (Cell signaling, 1:1,000 dilution), anti-ERK (Cell Signaling, 1:1,000 dilution), antiphosphorylated-ERK(Sigma, 1:1,000 dilution), anti-JNK(Cell Signaling, 1:1,000 dilution), antiphosphorylated-JNK (Sigma, 1:1,000 dilution), anti-p38 (Cell Signaling, 1:1,000 dilution), or antiphosphorlated-p38 (Sigma, 1:1,000 dilution). Primary antibodies were detected with either antigoat (Santa Cruz, 1:4,000 dilution) or anti-rabbit IgG (Cell Signaling, 1:4,000 dilution) horseradish peroxidase-coupled secondary antibodies. Images were captured using Amersham^® Hyperfilm^® ECL™ and MP Autoradiography Films (GE Healthcare).

Flag-tagged or HA-tagged vectors were transiently transfected or cotransfected into 293T cells using Mirus. Proteins were harvested using protein lysis buffer (1 M Tris–Cl, 5 M NaCl, 100% NP-40) containing 1× Halt Protease and Phosphatase Inhibitor Cocktail and 0.5 M EDTA (Thermo scientific) 36 h after transfection. From the protein lysate, 12.5% was used for WCL, while 75% was used for immunoprecipitation. Protein lysates were incubated with anti-Flag M2 beads overnight at 4°C on a shaker. After the protein binding M2 beads had been washed extensively using lysis buffer, 4× loading dye was added to resuspend the M2 beads and elute bound proteins. The samples were subjected to SDS-PAGE and Western blot was performed with respective antibodies.

The cells were washed twice with PBS and infected with L. monocytogenes at the bacteria-to-cell ratio 10:1 in antibiotics-free RPMI medium. After 1 h infection, cells were washed with PBS twice to remove extracellular L. monocytogenes. Cells were incubated in RPMI medium containing 5 μg/ml gentamicin for another 6 or 12 h and lysed by 200 μl 0.2% Triton X-100 in distilled water. The lysates containing L. monocytogenes were plated on Brain Heart Infusion (BHI) agar plates. The BHI agar plates were incubated overnight at 37°C and colony forming units (CFU) were counted.

The macrophages were seeded a day prior to infection at a cell density of 5 × 10⁴ cells/well. Infection was conducted at a multiplicity-of-infection 8 using M. bovis BCG. After an hour of infection, cells were washed with PBS to remove extracellular bacteria and media was replaced with RPMI media containing 10% fetal bovine serum. BCG was harvested at various time-points—days 0, 1, 2, 5, and 7. Cells were lysed with 0.1% Triton X-100 and lysates were plated on Middlebrook’s 7H11 agar plates. The plates were incubated at 37°C for 16 days and colony forming units (CFU) were recorded.

The macrophages were seeded a day prior to infection at a cell density of 1.5 × 10⁶ cells/well. Infection was conducted at a multiplicity-of-infection 8 using heat-inactivated M. tuberculosis. Cells were treated with Trizol at 3 and 6 h post-stimulation to conduct qRT-PCR to assess cytokine expression. Cell supernatant was collected to measure cytokine expression via ELISA.

The cells were washed twice with PBS and infected with L. monocytogenes at the bacteria-to-cell ratio 10:1 in antibiotics-free RPMI medium. After 1 h infection, cells were washed twice to remove extracellular Listeria. Cells were incubated in RPMI medium containing 5 μg/ml gentamicin for another 24 or 48 h. At respective time-points, cells were washed twice with PBS to remove dead floating cells. Cell proliferation was determined by crystal violet staining. Absorbance was measured at 590 nm. Percentage of cell survival was obtained by comparing absorbance of infected to that of uninfected cells.

Data were expressed as the mean ± SE. Data comparisons were made with Student’s t-test or two-way ANOVA using Microsoft Excel and GraphPad PRISM. Differences were considered significant when the P value is <0.05. P-values <0.05, 0.01, and 0.001 were represented by “*,” “**,” and “***,” respectively.

Mitogen-activated protein kinase-binding domain is essential for MAPK interaction and substrate specificity in typical DUSPs. Despite the absence of the binding domain, atypical DUSPs, such as DUSP14, had demonstrated marked ability in dephosphorylating MAPKs (18). Similarly, DUSP12 lacks the MKB domain and thus we examined DUSP12’s ability to interact with MAPKs. Mouse full-length DUSP12 was cloned into pcDNA vectors and transfected into HEK293T cells to examine its ability to bind endogenously to MAPKs. We found that DUSP12 was able to interact with endogenous p38, ERK, and JNK (Figure 1A). To examine the role of DUSP12 in innate immunity, we first examined the expression of DUSP12 in RAW264.7 cells, an immortalized murine macrophage cell line, in response to LPS stimulation. The activation of TLR4 will initiate a downstream MAPK signaling cascade which eventually activates transcription factor, AP-1. MKP5, a typical DUSP, serves as a positive control where its expression and function in macrophages have been examined in an earlier study (27). We found that LPS stimulation resulted in the upregulation of DUSP12 expression in macrophages (Figure 1B), although the upregulation appeared to be weaker as compared to that of MKP5. To examine if DUSP12 regulates MAPK-mediated response in immune cells, we performed a dual-luciferase assay by cotransfecting pcDNA vector carrying murine DUSP12 cDNA sequence and a vector consisting of the promoter construct of transcription factor AP-1 conjoined to the luciferase gene. The DUSP12 construct with a point mutation in the catalytic site serves as a negative control. As shown in Figure 1C, overexpression of functional DUSP12 inhibited AP-1 promoter activity in response to LPS stimulation, whereas the DUSP12 mutant lost the inhibition (Figure 1C). On the other hand, DUSP12 overexpression does not affect NFκB promoter activity in response to LPS stimulation (Figure S1 in Supplementary Material). These evidences indicate that DUSP12 could have utilized its phosphatase domain to regulate MAPK signaling and MAPK-mediated responses. To further investigate the function of DUSP12 in macrophages, we generated stable overexpression of DUSP12 in RAW264.7 macrophage cell line. The overexpression of DUSP12 in selected clones were confirmed at both mRNA and protein levels (Figure 1D). Clone number 5 (D5) with the highest DUSP12 expression was selected for subsequent experiments to elucidate the role of this molecule in innate immunity.

Macrophages are known to secrete proinflammatory mediators including cytokines and chemokines upon activation of their pattern recognition receptors in the event of pathogen infection. To examine whether DUSP12 regulates the expression of these proinflammatory mediators in TLR4 signaling, we stimulated DUSP12 overexpressing (D5) and vector control RAW264.7 cells with LPS and found that mRNA expression of TNF-α, IL-6, IL-1β, and MCP-1 were highly induced in both D5 and control cells by LPS (Figure 2A). Moreover, there was a significant reduction in the expression of these proinflammatory mediators in D5 compared to the control cells. Interestingly, the expression of IL-10, a potent anti-inflammatory cytokine, was enhanced in D5 cells (Figure 2A). Similar phenotype was observed at protein levels for TNF-α, IL-6 and MCP-1, where the proteins detected were suppressed in D5 when compared to the control (Figure 2B). Together, these results indicate the inhibitory role of DUSP12 on the expression of proinflammatory mediators in macrophages in response to TLR activation.

Mitogen-activated protein kinase pathway plays a crucial role in regulating the production of cytokines and chemokines during TLR activation in macrophages. To examine if the inhibition of proinflammatory cytokine production by DUSP12 is due to its regulation on MAPKs, the MAPKs activation profile in D5 and control cells were determined. In response to LPS stimulation, p38, JNK, and ERK were rapidly phosphorylated in both D5 and control cells but to different extents (Figure 2C). Overexpression of DUSP12 demonstrated significant reduction in p38 and JNK activation, while slight reduction in the activation of ERK was observed (Figure 2C). Together, the results demonstrated that DUSP12 inhibited proinflammatory cytokines and chemokines production in macrophages in response to LPS, possibly through inactivation of p38 and JNK.

To validate the function of DUSP12 in regulation of MAPK activation and cytokine expression in response to LPS, we utilized siRNA approach to knockdown the expression of DUSP12 in macrophages. We achieved about 50% knockdown of DUSP12 protein expression using a pool of siRNA targeting 4 regions of Dusp12 gene (Figure S2A in Supplementary Material). Increased activation of p38, JNK, and ERK in response to LPS stimulation was observed in cells with DUSP12 knockdown (Figure S2B in Supplementary Material), which is associated with increased expression of proinflammatory cytokines including TNF-α and MCP-1 (Figure S2C in Supplementary Material). These results confirmed the negative regulation of MAPK activation and proinflammatory cytokine expression in macrophages by DUSP12.

Macrophages serve as a first line of defense during infection and thus, these pathogens evolved a myriad of strategies to evade macrophage-mediated inflammatory responses. It is vital to understand the regulation of these responses to better control bacterial infections. As such, DUSP12 function in regulating inflammatory responses is studied; where intracellular pathogens were used to stimulate macrophages to assess the proinflammatory cytokine production and MAPK activation profile. D5 and control cells were activated by heat-inactivated M. tuberculosis to examine macrophage inflammatory responses. Consistent with the role of DUSP12 in TLR4-mediated cytokine expression, overexpression of DUSP12 suppressed the expression of TNF-α, IL-1β, and MCP-1 (Figure S3A in Supplementary Material). This is also reflected in protein expression of TNF-α and MCP-1, where cytokine levels were significantly reduced in D5 cells as compared to vector control cells (Figure S3B in Supplementary Material).

Thereafter, M. bovis BCG was used as a surrogate for M. tuberculosis in the macrophage stimulation set-ups. D5 and control macrophages were infected at a multiplicity of eight and bacteria growth was assessed by the comparing the number of colony forming units (CFU) at various time-points postinfection. The CFU counts demonstrated that DUSP12 overexpression resulted in higher bacteria burden across all time-points, where the difference is significant in the later time-points as compared to the control cells (Figure 3A). The phenotype observed was correlated with MAPK activation profile and cytokine production. Western blot analysis demonstrated that overexpression of DUSP12 greatly inhibited p38 activation, but not the activation of JNK or ERK upon BCG infection (Figure 3B). The cytokine profile in D5 and control cells upon BCG stimulation revealed that expression of proinflammatory cytokines including TNF-α and MCP-1 was suppressed while anti-inflammatory cytokine, IL10 was upregulated (Figure 3C). Together, these results suggest that DUSP12 negatively regulates p38 activation and proinflammatory cytokine expression in macrophages upon the detection of Mycobacteria.

In the same manner, macrophages play a crucial role in immune responses against L. monocytogenes. To examine the regulatory role of DUSP12 in macrophages in response to Listeria infection, D5 and control cells were infected with L. monocytogenes at a MOI of 10 to assess bacterial load and cell survival. Similar to BCG infection profile, Listeria CFU counts was significantly higher in D5 cells as compared to the control at 6 and 12 h postinfection (Figure 4A). The higher bacteria load had also resulted in greater cell death, where the difference between D5 and control is pronounced at the 48 h postinfection (Figure S4 in Supplementary Material). Furthermore, mRNA and protein expression of proinflammatory cytokines including TNF-α, IL-6, IL-1β, and IFNβ were assessed. In control macrophages, the expression of these cytokines was greatly induced, while D5 cells showed significant suppression at 3 and 6 h postinfection (Figure 4B). In addition, TNF-α protein expression was significantly attenuated in DUSP12 overexpressing cells compared to that in control cells upon Listeria infection (Figure 4C).

Next, the activation of MAPKs in D5 and control cells in response to Listeria infection was examined. All three groups of MAPKs were activated in response to Listeria infection (Figure 4D). Clearly, overexpression of DUSP12 greatly inhibited p38 activation at all time points examined, compared to the control cells. In addition, JNK activation at 6 h postinfection was reduced by DUSP12 overexpression, whereas ERK activation was comparable across all time-points between D5 and control cells (Figure 4D). Together, these results suggest that DUSP12 inhibits inflammatory mediator expression in macrophages during Listeria infection through attenuation of p38, and to a lesser extent, JNK activation.

Next, we selected another overexpressing clone (OXE), which has DUSP12 levels of approximately three to four times that of the control cells (Figure S5A in Supplementary Material), to verify the function of DUSP12 in macrophages in response to intracellular bacterial infection. Similar to D5 (which has eightfolds of DUSP12 overexpression, Figure 1D), OXE cells has reduced inflammatory cytokine expression (Figure S5B in Supplementary Material), associated with reduced MAPK activation but to a lesser extend as D5 (Figure S5C in Supplementary Material), and were less able to control BCG replication compared to control cells (Figure S5D in Supplementary Material). Together, these results confirm the inhibitory function of this protein on macrophage responses to intracellular bacterial infection by targeting MAPKs.

Dual-specificity phosphatase 12 was able to interact with endogenous p38, ERK, and JNK (Figure 1A), with stronger binding to ERK and JNK than to p38. DUSP12 is an atypical DUSP that lacks the MAPK binding domain. Therefore, we postulated that an additional binding agent is recruited for the interaction between DUSP12 and MAPKs. The binding of DUSP12 to p38 is the weakest among the three MAPKs (Figure 1A), yet DUSP12 has the strongest inhibition toward the activation of p38 compared with that of JNK or ERK in macrophages upon TLR activation or intracellular bacterial infection. Therefore, we speculated that scaffold proteins might be present to mediate the interaction between DUSP12 and p38. Studies by Sekine et al. (28) demonstrated that signal-transducing adaptor protein-2 (STAP-2) in macrophages modulates cytokine production in TLR4 signaling pathway and thus, we studied the possible role of STAP proteins (STAP1 and STAP2) in facilitating DUSP12 and p38 interaction. Coimmunoprecipitation of the STAP proteins with DUSP12 revealed that STAP2, and not STAP1, interacted with DUSP12 (Figure 5A). In addition, in the presence of STAP2, DUSP12, and p38 interaction was enhanced (Figure 5B). These results suggest that STAP2 may function as a scaffold protein which mediates the interaction between DUSP12 and p38.

Signal transducing adaptor protein 2 contains an N-terminal pleckstrin homology (PH) domain, a central Src homology 2-like (SH2) domain, and a proline-rich YXXQ motif in the C-terminal. Several studies demonstrated that the SH2 domain is important for binding to other proteins including FAK, IKKs, MyD88, and STAT3/5 (28–30). Hence, several STAP2 constructs were generated (Figure 5C) and the binding potential of these proteins to DUSP12 and p38 were examined. We found that deletion of the SH2 domain (ΔSH2) and C-terminal domain (ΔC) of STAP2 did not affect its interaction with DUSP12, whereas deletion of the PH domain (ΔPH) led to much weaker interaction with DUSP12 (Figure 5D). Similarly, ΔSH2 mutant and ΔC mutant of STAP2 were able to interact with p38, whereas ΔPH mutant failed to interact with p38 (Figure 5E). Together, these results suggest that the PH domain of STAP2 is required for its interaction with both DUSP12 and p38.

Dual-specificity phosphatase 12 contains an N-terminal catalytic domain, a C-terminal zinc finger domain, and an ALDH site (Figure 6A). To test the function of these domains of DUSP12 in its interaction with STAP2, we generated several truncated DUSP12 constructs (Figure 6A) and were used to perform immunoprecipitation with STAP2. We found that deletion of the N-terminal catalytic domain did not affect it interaction with STAP2, whereas deletion of its C-terminal domain, including the zinc finger domain and the ALDH site resulted in weakened interaction with STAP2 (Figure 6B). This suggests that the DUSP12 C-terminal domain including the zinc finger domain and the ALDH site is required for its interaction with STAP2. Furthermore, the lack of C-terminal domain in DUSP12 impaired its interaction with ERK, JNK, and p38 (Figure 6C). Particularly, the interaction between DUSP12 and p38 was completely lost, suggesting that the C-terminal domain of DUSP12 is required for its interaction with p38, likely via STAP2.