DUSP11 Attenuates Lipopolysaccharide-Induced Macrophage Activation by Targeting TAK1

Dual-specificity phosphatases (DUSPs) are a family of protein phosphatases. DUSPs can dephosphorylate both threonine/serine and tyrosine residues in the Thr-X-Tyr (TXY) activation motif of MAPKs (1). The intensity and duration of MAPK activation are negatively regulated by dephosphorylation (2–4). DUSP family phosphatases currently contain 25 members and can be classified into classical or atypical DUSPs based on the presence or absence of the kinase-interacting motif (KIM) (1, 5). Classical DUSPs, also named as MAPK phosphatases (MKPs), contain the KIM motif, an N-terminal Cdc25 homology domain, and a highly conserved C-terminal phosphatase domain. The structure of atypical DUSPs are relatively simpler, contrasting to classical DUSPs; they lack the KIM motif and N-terminal Cdc25 homology domain (5–7).

DUSP 11 (DUSP11, also named RIP1) is an atypical DUSP and is known to be an RNA triphosphatase (8, 9). DUSP11 converts the 5′ triphosphate of microRNA precursors to a 5′ monophosphate and regulates the levels of cellular noncoding RNAs (8–10). During Kaposi sarcoma–associated herpesvirus (KSHV) infection, DUSP11 removes 5′ triphosphates of KSHV vault RNAs and prevents vault RNA recognition by RIG-I, leading to attenuation of KSHV lytic reactivation (11). To date, the functional roles of DUSP11 in the development of immune cells and innate immune responses are still unclear.

TGF-β–activated kinase 1 (TAK1) is a serine/threonine protein kinase and is a member of the MAPK kinase kinase (MAP3K) family (12, 13). TAK1 is activated by various extracellular stimuli such as IL-1β, TNF-α, and TLR signaling. TAK1 transmits the upstream signal from receptor complexes (14) or MAP4Ks (13, 15–17) to the downstream MAP2Ks and IκB kinase (IKK), leading to activation of JNK/p38 and NF-κB.

To investigate the in vivo function of DUSP11 in immune cell signaling and immune responses, a DUSP11-deficient mouse line was generated. In this study, we report that DUSP11 interacts with TAK1 upon LPS signaling in macrophages. Also, DUSP11 deficiency enhances LPS-induced macrophage activation in bone marrow–derived macrophages (BMDMs). DUSP11-deficient mice showed increased susceptibility and cytokine production to LPS-induced endotoxic shock in vivo. In conclusion, our results indicate that DUSP11 inhibits LPS signaling and immune responses by targeting TAK1.

DUSP11 gene deficiency in a 129 mouse embryonic stem cell clone (BB0198) was purchased from the International Gene Trap Consortium. To generate chimeric mice, embryonic stem cells were injected into blastocysts from C57BL/6 mice by the Transgenic Mouse Model Core, National Research Program for Biopharmaceuticals (Taipei, Taiwan). Originally, the extent of DUSP11 deficiency in the mouse line was variable. To solve this problem, mice with efficient DUSP11 deficiency/knockdown were identified and then were used as breeders. The DUSP11-deficient mice in the C57BL/6 background were generated by backcrossing the heterozygous DUSP11 progeny with C57BL/6 mice for more than 10 generations. The genotypes of mice were determined by a PCR analysis with the common 3′ primer 5′-CGCTGCCCAACTGGGAGATAGTCTTT-3′, the 5′ primer specific for wild-type (WT) allele 5′-GTTAGGCTGTTCTGCTATGGC-3′, and the 5′ primer specific for knockout (KO) allele 5′-GATTGTACTGAGAGTGCACC-3′ mixed together in the same reaction. PCR products were analyzed by electrophoresis, and the DNA fragments with 276 and 200 bp represented the WT and KO alleles, respectively. For detection of DUSP11 mRNA levels by quantitative PCR (qPCR), the primer pair forward, 5′-CTACGGAGCAAGTCTACACC-3′ and reverse, 5′-GCCTCTCTGGTAAACACTCTG-3′ were used. Mice used in this study were 5–12 wk of age. Mice were housed under specific pathogen–free conditions under institutional guidelines, and all animal experiments were approved by the Institutional Animal Care and Use Committee at the National Health Research Institutes.

The HEK293T human embryonic kidney cell line and the Raw264.7 mouse macrophage cell line were cultured in DMEM (Life Technologies) containing 10% FCS plus 10 μg/ml streptomycin and 10 U/ml penicillin. Primary BMDMs from WT and DUSP11-deficient mice were differentiated by the standard procedure. In brief, bone marrow cells from femora and tibiae were flushed with cold PBS and cultured with RPMI 1640 with 10% FCS, 1% penicillin/streptomycin (Life Technologies), and 25 ng/ml M-CSF for 7 d. A total of 2 × 10⁶ BMDMs in six-well plates were stimulated with LPSs (0.1, 0.5, or 1 μg/ml) for 72 h. After LPS stimulation, the culture medium was collected and subjected to ELISA.

Thymus, spleen, and lymph nodes were isolated from mice, and cells were mashed and filtered through a 70-μm cell strainer to obtain single-cell suspensions. Cells were depleted of RBCs by RBC lysis buffer. Single-cell suspensions of 1 × 10⁶ cells were stained with the fluorescently labeled anti–CD3-FITC (145-2C11), anti–CD3-PerCP (145-2C11), anti–CD4-Pacific Blue (RM4-5), anti–CD8-allophycocyanin-Cy7 (53-6.7), anti–B220 (RA3–6B2)-PE-Cy7, anti–Gr-1-PE (RB6-8C5), anti–F4/80-PE-Cy7 (BM8), or anti–CD11b-PerCP-Cy5.5 (M1/70) Abs (all from BioLegend) for 15 min at 4°C in 50 μl of staining buffer (1× PBS, 2% FBS, 0.01% NaN₃). The cells were then washed and analyzed on a FACSCanto II flow cytometer (BD Biosciences). The cell populations were analyzed using FlowJo software (BD Biosciences).

The sample preparation and liquid chromatography–mass spectrometry analysis performed were as previously described (18).

Flag-DUSP11 plasmid was constructed by subcloning human DUSP11 cDNA into the vector pCMV6-AC-3DDK (OriGene Technologies). Myc-TAK1 plasmid was constructed by subcloning human TAK1 cDNA into the vector pCMV6-AC-Flag (OriGene Technologies). Flag-DUSP14 plasmid was constructed by subcloning human DUSP14 cDNA into the vector pCMV6-AC-3DDK (OriGene Technologies) as described previously (19). The Abs specific for TGF-β–activated kinase 1–binding protein 1 (TAB1), TAK1, phospho-TAK (pS412), JNK, phospho-JNK (pT183/pY185), IKKβ, and phospho-IKK were purchased from Cell Signaling Technology. The Ab recognizing both human and murine DUSP11 was generated by immunization of a rabbit with peptides (murine DUSP11 epitope ²²⁹TNNKPVKKKPRKNRRGGHL²⁴⁷). Anti-Flag (DDK) Ab was purchased from Origene Technologies. Anti-vinculin Ab was purchased from Merck. The Abs for Myc, tubulin, and actin were purchased from Sigma-Aldrich. LPS (Escherichia coli serotype O1111:B4) and DMSO were obtained from Sigma-Aldrich.

Cell lysate was lysed in lysis buffer (50 mM Tris–Cl [pH 7.4], 150 mM NaCl, 5 mM EDTA, 0.1% Triton X-100, 0.02% NaN₃, 100 mM PMSF, 1 M sodium fluoride, 0.5 mg/ml leupeptin, 1 mg/ml aprotinin, and 5 mg/ml pepstatin A) at 4°C for 30 min. For coimmunoprecipitation, cell lysates were incubated with anti-Myc agarose beads (9E10; Sigma-Aldrich) or anti-Flag agarose beads (M2; Sigma-Aldrich) in 1 ml of lysis buffer at 4°C for 2 h. The immunocomplexes were washed three times with lysis buffer, followed by Western blotting analysis. For Western blotting analysis, the lysates were boiled at 95°C for 5 min and then subjected to SDS-PAGE, and the separated proteins were transferred to a PVDF membrane and blocked with Tris-buffered saline with 0.1% Tween 20 containing 5% BSA for 1 h. The PVDF membrane was incubated with specific primary Abs at 4°C overnight, followed by incubation with HRP-conjugated secondary Abs at room temperature for 1 h. The specific signals were detected by an ECL Substrate Kit (Millipore).

Six- to twelve-week-old DUSP11 WT or DUSP11-deficient mice were i.p. injected with LPS (15 mg/kg body weight) in PBS to induce endotoxic shock and were monitored for survival over 5 d. For measurement the serum cytokine production, blood samples were obtained from mice 6 h after LPS injection. The blood samples were centrifuged at 18,000 × g for 15 min at room temperature. The supernatants were stored at −80°C until they were assayed for cytokines by ELISA.

The levels of TNF-α, IL-1β, and IL-6 in cell supernatants or mouse sera were quantified using ELISA Kits (eBioscience) according to the manufacturer’s instructions.

In situ proximity ligation assay (PLA) analyses were performed as previously described using the Duolink In Situ Red Starter Kit (Sigma-Aldrich) according to the manufacturer’s instructions (20). Briefly, WT and DUSP11 KO BMDMs were stimulated with LPS at the indicated time, and then the cells were deposited on the glass slides and pretreated with fixation and blocking solutions. Cells were then incubated with primary Abs against DUSP11 and TAK1, followed by oligonucleotide (PLA Probe MINUS or PLA Probe PLUS)-conjugated secondary Abs. After ligation and amplification reactions, the slides were washed and mounted using a Duolink In Situ Mounting Medium with DAPI. The PLA signals from paired DUSP11 and TAK1 in close proximity (<40 nm) were visualized as individual red dots by a fluorescence microscope (DM2500; Leica). Red dots represent direct interactions. To analyze the results, the number of red dots and DAPI-positive cells were counted per field. The PLA signal was quantified and was shown as the number of PLA (red) signals per DAPI (blue)-positive cells.

Flag-tagged TAK1 proteins were immunoprecipitated from lysates of HEK293T cells transfected with Flag-TAK1 plasmid. Immunoprecipitated TAK1 proteins were incubated with immunoprecipitated DUSP11 WT or DUSP11 mutant (Cys¹⁵² to Ser¹⁵²) proteins for 2 h. The reaction products were boiled in 5× sample buffer, followed by Western blotting with anti–phospho-TAK1 Ab.

The DUSP11-deficient mice were generated for studying the function of DUSP11 in immune cell development and immune responses. An insertion vector containing a splice acceptor, β-geo (a fusion of β-galactosidase and neomycin resistance genes), and a polyadenylation sequence was inserted into the intron between exon 1 and exon 2 of the targeted DUSP11 allele (Fig. 1A). Successful targeting of DUSP11 was validated by PCR analysis of genomic DNAs from mouse tails (Fig. 1B). qPCR analysis using paired primers targeting DUSP11 at exon 9 confirmed the loss of DUSP11 mRNA expression in the peripheral blood, thymus, spleen, and lymph node of DUSP11-deficient mice (Fig. 1C).

We next examined whether loss of DUSP11 expression might alter the development of immune cells in DUSP11-deficient mice. Flow cytometry analysis showed that DUSP11 deficiency did not alter the development of CD4⁻CD8⁻ double-negative, CD4⁺CD8⁺ double-positive, CD8⁺ single-positive, or CD4⁺ single-positive thymocytes (Fig. 2A). The data also showed similar proportions of CD3⁺ T cells, CD4⁺ T cells, CD8⁺ T cells, and B220⁺ B cells in the spleen and lymph nodes of WT and DUSP11-deficient mice (Fig. 2B, 2C). The frequencies of macrophages (F4/80⁺CD11b⁺) and neutrophils (Gr-1⁺CD11b⁺) in the spleen were also similar in WT and DUSP11-deficient mice (Fig. 2D). These data suggest that DUSP11 is not essential for the immune cell development.

We then searched for the potential target(s) of DUSP11 using DUSP11 immunocomplexes by coimmunoprecipitation and liquid chromatography–mass spectrometry. HEK293T cells were transfected with Flag-DUSP11, and DUSP11-interacting proteins were coimmunoprecipitated with anti-Flag Ab. The immunocomplexes were fractionated by SDS-PAGE. The protein bands were excised and subjected to liquid chromatography–mass spectrometry. We identified TAK1 as a DUSP11-interacting protein by mass spectrometry (Fig. 3A). The interaction of DUSP11 with TAK1 was confirmed by reciprocal coimmunoprecipitation assays using Myc-DUSP11–overexpressing and Flag-TAK1–overexpressing HEK293T cell lysates (Fig. 3B). Furthermore, HEK293T cells were transfected with Flag-DUSP11 or Flag-DUSP14 plasmid. The immunocomplexes were coimmunoprecipitated with anti-Flag Ab and then analyzed by Western blotting. The results demonstrated that DUSP11 interacts with endogenous TAK1 but not endogenous TAB1 proteins (Fig. 3C). In contrast, consistent with our previous findings (19), DUSP14 directly interacted with endogenous TAB1 proteins but not endogenous TAK1 proteins, as detected by coimmunoprecipitation assay (Fig. 3C).

Next, we asked whether DUSP11 dephosphorylates TAK1 in vitro. HEK293T cells were transfected with Flag-TAK1 plasmid, and then TAK1 proteins were purified by immunoprecipitation with anti-Flag Ab. The purified Myc-DUSP11 WT proteins dephosphorylated the purified Flag-TAK1 proteins at Ser412 residue, whereas the phosphatase-dead DUSP11 mutant (Cys¹⁵² to serine) proteins did not dephosphorylate TAK1 using in vitro phosphatase assays (Fig. 3D).

TAK1 is an essential mediator that transmits the TLR4 signal from the receptor complex to downstream effectors IKK and MAPKs, which control the production of inflammatory cytokines important for innate immunity. To determine whether DUSP11 regulates TAK1 activation during LPS signaling, we investigated the potential interaction between DUSP11 and TAK1 upon LPS stimulation in macrophages by coimmunoprecipitation assays and PLA. Raw264.7 macrophages were cotransfected with Flag-DUSP11 plus Myc-TAK1 plasmids and then stimulated with LPS at the indicated time. The results showed that DUSP11 modestly interacted with TAK1 without LPS stimulation when both proteins were overexpressed (Fig. 4A). The interaction between DUSP11 and TAK1 was further enhanced by LPS stimulation in Raw 264.7 macrophages detected by coimmunoprecipitation assays (Fig. 4A). Moreover, the LPS-induced interaction between endogenous DUSP11 and TAK1 proteins in BMDMs were confirmed by PLA (Fig. 4B, 4C). The anti-DUSP11 Ab specificity for PLA signals was validated using DUSP11-deficient BMDMs (Fig. 4B, 4C). Collectively, the data suggest that DUSP11 directly interacts with TAK1 in BMDMs during TLR signaling.

During LPS signaling, the phosphorylation of TAK1 is required for its kinase activity, leading to induction of proinflammatory cytokines. We further characterized the levels of phosphorylated TAK1 in DUSP11-deficient BMDMs upon LPS stimulation. Our results showed that the phosphorylation levels of TAK1 at Ser412 were significantly increased in DUSP11-deficient BMDMs compared with those of WT BMDMs (Fig. 5A). Also, the phosphorylation of IKK and JNK were also enhanced in DUSP11-deficient BMDMs (Fig. 5A). We characterized the proinflammatory cytokine production in DUSP11-deficient BMDMs upon LPS stimulation. Our results showed that levels of IL-1β, TNF-α, and IL-6 were significantly higher in DUSP11-deficient BMDMs than those of WT controls (Fig. 5B). We next examined whether DUSP11 deficiency could increase the susceptibility to LPS-induced endotoxic shock in vivo. WT and DUSP11-deficient mice were i.p. injected with 15 mg/kg body weight LPS and monitored for 120 h. All of the mice that received LPS treatment showed apparent signs of distress such as apathy, fur ruffling, and diarrhea. The data appear to suggest that DUSP11-deficient mice are more susceptible to endotoxic shock (Fig. 5C). In contrast, WT mice did not die until 48 h after LPS injection, and 40% of the WT mice still survived at 120 h (Fig. 5C). Furthermore, serum levels of IL-1β, TNF-α, and IL-6 were significantly higher in DUSP11-deficient mice than those in WT mice at 6 h after LPS injection (Fig. 5D).

DUSP11 has been previously reported as an RNA phosphatase (8, 9). In this study, we demonstrated that in addition to targeting to RNAs, DUSP11 is an important phosphatase that directly interacts with and regulates TAK1 activation in macrophages, leading to the reduction of LPS-induced endotoxic shock. LPS-mediated signaling pathways have been shown to stimulate the production of proinflammatory cytokines, such as TNF-α, IL-6, and IL-1β; the overproduction of these proinflammatory cytokines contributes to the induction of endotoxic shock. The serum levels of TNF-α, IL-6, and IL-1β were significantly elevated in DUSP11-deficient mice compared with those of WT mice upon LPS challenge.

After TLR4 ligand stimulation, the MyD88-dependent pathways were activated, leading to sequential activation of IL-1R–associated kinase (IRAK1), IRAK4, TNFR-associated factor 6 (TRAF6), and TAK1 (14). Furthermore, phosphorylation and ubiquitination of TAK1 and its binding partners TAB1 and TAB2/3 are essential for TAK1 activation (21, 22). TAK1 Ser412 phosphorylation is regulated by X-linked protein kinase (PRKX) and cAMP-dependent protein kinase catalytic subunit α (PKACα) under TLR signaling (23). Conversely, the phosphatase holoenzyme PP1 negatively regulates TAK1 Ser412 phosphorylation in TLR signaling (24). In this report, we showed that DUSP11 directly interacted with and dephosphorylated the TAK1 protein as its substrate, leading to inactivation of TAK1 downstream molecules (IKK and JNK). Another DUSP family member, DUSP14, attenuates activation of the TAB1–TAK1 signaling through directly targeting TAB1 in T cells (18, 19). Unlike DUSP14, DUSP11 bound to TAK1 but not TAB1. DUSP14 phosphatase activity is induced by its protein methylation and ubiquitination under TCR signaling (18, 19, 25); therefore, it would be interesting to study whether and how DUSP11 activity is regulated by its upstream regulators upon TLR signaling. Furthermore, it is important to characterize the structure–function relationships between DUSP11 and TAK1 in immune cells upon TLR signaling. Besides DUSP11 and DUSP14, several DUSP family members play important roles in immune cell activation, inflammatory responses (6, 7, 26–32), or tumorigenesis (33–35). These findings suggest that DUSP family members have individual functions, instead of redundant roles, in regulation of various immune cell signaling pathways.

We thank the Laboratory Animal Center (Association for Assessment and Accreditation of Laboratory Animal Care International accredited) of the National Health Research Institutes for mouse housing. We thank the Transgenic Mouse Model Core of National Research Program for Biopharmaceuticals for blastocyst microinjection and generation of DUSP11 KO mice. We thank the Institute of Biological Chemistry of Academia Sinica for mass spectrometry–based proteomics.

The authors have no financial conflicts of interest.