Visual Abstract

Tristetraprolin (TTP) is an RNA-binding protein and an essential factor of posttranscriptional repression of cytokine biosynthesis in macrophages. Its activity is temporally inhibited by LPS-induced p38MAPK/MAPKAPK2/3–mediated phosphorylation, leading to a rapid increase in cytokine expression. We compared TTP expression and cytokine production in mouse bone marrow–derived macrophages of different genotypes: wild type, MAPKAP kinase 2 (MK2) deletion (MK2 knockout [KO]), MK2/3 double deletion (MK2/3 double KO [DKO]), TTP-S52A-S178A (TTPaa) knock-in, as well as combined MK2 KO/TTPaa and MK2/3 DKO/TTPaa. The comparisons reveal that MK2/3 are the only LPS-induced kinases for S52 and S178 of TTP and the role of MK2 and MK3 in the regulation of TNF biosynthesis is not restricted to phosphorylation of TTP at S52/S178 but includes independent processes, which could involve other TTP phosphorylations (such as S316) or other substrates of MK2/3 or p38MAPK. Furthermore, we found differences in the dependence of various cytokines on the cooperation between MK2/3 deletion and TTP mutation ex vivo. In the cecal ligation and puncture model of systemic inflammation, a dramatic decrease of cytokine production in MK2/3 DKO, TTPaa, and DKO/TTPaa mice compared with wild-type animals is observed, thus confirming the role of the MK2/3/TTP signaling axis in cytokine production also in vivo. These findings improve our understanding of this signaling axis and could be of future relevance in the treatment of inflammation.

Many cytokine mRNAs display adenylate-uridinylate–rich elements (AREs) in their 3′ untranslated region, which allows specific binding of regulatory proteins and a resulting posttranscriptional regulation of cytokine biosynthesis at the level of mRNA stability and translation (1). Tristetraprolin (TTP) is one of such ARE-binding proteins, which destabilizes mRNAs, inhibits their translation, and thereby suppresses cytokine production. Its constitutive deletion in mice leads to increased basal levels of cytokines, such as TNF, and symptoms of cachexia (2). Under steady-state conditions, TTP is ubiquitously expressed at a basal level. Inflammatory stimuli like LPS and various cytokines mediate transcriptional and posttranscriptional induction of TTP expression in hematopoietic cell lineages (2, 3). Rapid upregulation of TTP expression in the early inflammatory response is accompanied by its transient phosphorylation at multiple sites. At least two phosphorylation sites of TTP, S52 and S178, were shown to be phosphorylated directly by the p38MAPK-activated kinases MK2 and MK3 (3, 4). Inactivation of p38MAPK/MK2/3 signaling by p38 or MK2/3 inhibitors or by MK2/3 deletion in mice resulted in a dramatic decrease of TTP-targeted inflammatory cytokines and TTP expression, indicating a transient inhibitory role of p38/MK2/3 signaling on TTP’s suppressive activity (46). Loss of the inhibitory effect of MK2/3-mediated phosphorylation on TTP activity was also observed in knock-in mice, in which the TTP gene was altered to code for a nonphosphorylatable mutant, TTP-S52A-S178A (TTPaa) (2). Similar to MK2/3-deficient (MK2/3 double knockout [DKO]) mice, the TTPaa/aa knock-in mice (TTPaa mice) demonstrate significant reduction of TTP-targeted inflammatory cytokines and TTP expression, reflecting the inability of TTPaa to be inactivated by p38 MAPK/MK2/3 signaling. These observations indicate that transient phosphorylation of TTP inhibits its suppressive function and enables its target cytokines to be expressed and elucidate the inflammatory response before TTP dephosphorylation reactivates its suppressive function, leading to decreased cytokine production resolving inflammation. This is also in agreement with the finding that phospho-TTP can be replaced by the mRNA-stabilizing factor HuR at the AREs (3).

Interestingly, TTPaa protein is induced to lower levels than wild-type (WT) TTP in LPS-stimulated bone marrow–derived macrophages (BMDMs) (2). This can be explained by two distinct phenomena. First, an ARE in TTP’s own mRNA allows negative feedback control of TTP expression (4). Constitutively active TTPaa is therefore able to suppress its own expression more effectively by this feedback mechanism, leading to lower TTPaa levels than WT TTP (2). Second, the phosphorylation of TTP at serines 52 and 178 not only impairs the mRNA-destabilizing activity of TTP, it also protects TTP against proteasome-dependent degradation. Hence, the nonphosphorylatable mutant TTPaa is both highly active and highly unstable (2, 5, 6).

Analyses of TTP phosphorylation revealed many in vitro phosphorylation sites for various protein kinases (7) and several sites phosphorylated in transfected cells (2, 8, 9). However, the above-described signaling pathway targets only two sites of the MK2/3-substrate TTP downstream of the p38MAPK/MK2/3 axis. So far, it is not clear whether other substrates of MK2/3 or downstream to the p38MAPK/MK2/3 axis [such as small Hsps (10), RTN4/NOGO-B (11), SRF (12), RIPK1 (1315), or UBE2J1 (16)] contributes to the complex regulation of cytokine production and whether other phosphorylation sites in TTP are involved. Because a quantitative comparison of the in vivo effects of upstream and downstream mutations in this pathway, per se, is difficult, we decided to combine upstream and downstream mutations in this pathway and compared cytokine production in a combined MK2/3 DKO/TTPaa/aa knock-in mouse model. Using macrophages from this model, we could also answer the question of whether other phosphorylation sites of TTP besides S52 and S178 are relevant for regulation of the inflammatory response. Using these mice in a model of sepsis demonstrates the relevance of MK2/3/TTP signaling in vivo. Taken together, the results revealed that MK2/3 are the only LPS-induced kinases for S52 and S178 of TTP but also showed that the role of MK2/3 in the regulation of TNF biosynthesis is not limited to these phosphorylations.

MK2 knockout (KO) (17), MD2/3 DKO (18), TTPaa (2), MK2/TTPaa, and DKO/TTPaa mice strains were maintained at the animal facility of the Hannover Medical School. To generate BMDMs, bone marrow cells were flushed from the femurs of mice. Cells were cultured on 10-cm dishes in DMEM supplemented with 10% FBS, penicillin/streptomycin, and 50 ng/ml rM-CSF (Wyeth, Boston, MA) under humidified conditions, with 5% CO2 at 37°C for 6 d. On the sixth day, cells were collected by treatment with nonenzymatic cell dissociation solution (C5789; Sigma-Aldrich) for 30–40 min, followed by scraping. After that, cells were washed, counted, and seeded for experiments. On the next day, cells were treated with 1 μg/ml LPS (Escherichia coli 0127:B8; Sigma-Aldrich) dissolved in complete growth medium for the time indicated. Cycloheximide was used at concentration of 5 μg/ml.

Protein extracts were prepared by direct lysis of the cells in a culture plate with 2× Laemmli SDS sample buffer. Protein lysates were separated by SDS-PAGE on 7.5–16% gradient gels and transferred by wet blotting to Hybond ECL nitrocellulose membranes (GE Healthcare). Primary Abs used were anti-eEF2 (2332) and anti-MK2 (3042). Abs against TTP [SAK21B (19) and mTis11 (NE2/1.1) (20)] were used as described previously. Secondary HRP-conjugated Abs (Santa Cruz Biotechnologies) were used. Ag-Ab complexes were detected with homemade ECL detection solution (solution A: 1.2 mM luminol in 0.1 M Tris-HCl [pH 8.6]; solution B: 6.7 mM p-coumaric acid in DMSO; 35% H2O2 solution; ratio of 3333:333:1) using the Luminescent Image Analyzer LAS-3000 (Fujifilm). Secreted factors in tissue culture supernatants were quantified by ELISA, according to the manufacturer’s instructions, or by using multiplex bead capture assays and a Bio-Plex 200 analyzer (Bio-Rad). The mouse TNF-α Ready-SETGo! Kit from eBiosciences (88–7324) was used for TNF ELISA.

For stable isotope labeling by amino acids in cell culture (SILAC), doxycycline-inducible TTP mouse embryonic fibroblasts (20) were cultured at 37°C in a humidified incubator containing 5% CO2 and in media containing either l-arginine and l-lysine (light), l-arginine 13C6 and l-lysine 2H4 (medium), or l-arginine 13C6-15N4 and l-lysine 13C6-15N2 (heavy; Cambridge Isotope Laboratories). To induce TTP expression, cells were induced with 1 μg/ml doxycycline overnight and pretreated with the MK2 inhibitor PF3644022 (catalog no. PZ0188, 10 μM; Sigma-Aldrich) for 60 min before stimulation with anisomycin (1 μg/ml; Sigma-Aldrich) for 60 min. GFP immunoprecipitations were carried out using a magnetic GFP-Trap binding resin (Chromotek). Beads were eluted in 2× Laemmli buffer, and the resulting supernatants were mixed and loaded onto a 4–12% NuPAGE Bis-Tris Gel (Novex; Thermo Fisher Scientific). Further processing of samples and mass spectrometry was performed as described before (21).

Polymicrobial sepsis was induced by cecal ligation and puncture (CLP) under light isoflurane anesthesia in 10- to 12-wk-old mice. A 1-cm ventral midline abdominal incision was made. The cecum was then exposed, ligated with 4-0 silk sutures just distal to the ileocecal valve (comprising 70% of the cecum and sparing the cecal vessels) to avoid intestinal obstruction, and punctured through with a 24-gauge needle. The punctured cecum was gently squeezed to expel a 1- to 2-mm droplet of fecal material and returned to the abdominal cavity. The incision was then closed in layers with 4-0 surgical sutures. Mice were fluid resuscitated with prewarmed normal saline (500 μl) i.p. immediately after the procedure. Blood samples were obtained under light ether anesthesia from the cavernous sinus at the indicated time points. Mice were sacrificed 20 h after CLP, and peritoneal lavage (PL) was performed using 3 ml of PBS. The volume of the collected PL was measured in each sample, and the total WBC count was assessed using Scil Vet ABC Hematology Analyzer (Scil Animal Care Company GmbH, Viernheim, Germany). Polymorphonuclear neutrophils (PMN) in peritoneal effluents were analyzed by flow cytometry using an anti-Gr1 (clone RB6-8C5) mAb (BioLegend, San Diego, CA) and the FACS Canto II cytometer (BD Biosciences, Heidelberg, Germany). Prism 5 software (GraphPad Software) was used for data processing and statistical evaluation. All procedures were carried out at Phenos GmbH (Hannover, Germany), according to the guidelines of the German Society for Animal Science (Gesellschaft fuer Versuchstierkunde), and were approved by the local authorities.

An unpaired, two-tailed Student t test was applied for comparison of two groups.

We first compared cytokine production and TTP expression after LPS stimulation of BMDMs from WT, MK2-deficient (MK2 KO), and TTPaa mice as well as TTPaa mice combined with the MK2-deficient background (MK2/TTPaa). As expected, significant reduction of TTP and TNF expression was observed in MK2 KO, TTPaa, and MK2/TTPaa cells compared with WT cells (Fig. 1). Interestingly, MK2-deficient BMDMs expressed considerably higher levels of TTP than TTPaa or MK2/TTPaa BMDMs (Fig. 1A). This observation could be explained by either residual phosphorylation-mediated stabilization of TTP protein or residual phosphorylation-mediated inactivation of TTP, in the absence of MK2. The residual phosphorylation of TTP at S52/S178 in MK2-deficient cells could occur because of the related protein kinase MK3, which was shown to cooperate with MK2 in TTP repression and cytokine production (22). LPS-induced TNF production is further reduced in MK2/TTPaa cells compared with TTPaa cells, indicating the existence of additional MK2-mediated mechanisms of its regulation, besides TTP phosphorylation at S52/S178 (Fig. 1B).

FIGURE 1.

MK2 deficiency leads to strong but not complete reduction of TTP and TNF production. BMDMs of the different genotypes were treated with 1 μg/ml LPS for the indicated times. TTP protein expression was analyzed by Western blot. EF2 was used as loading control (A). The amount of secreted TNF (B) was measured by ELISA. The experiment was repeated three times. One mouse per each experiment was used, and the cells were seeded and analyzed in triplicates. The representative results are shown, and the error bars represent deviation between triplicates within one experiment. ***p < 0.005. ns, not significant.

FIGURE 1.

MK2 deficiency leads to strong but not complete reduction of TTP and TNF production. BMDMs of the different genotypes were treated with 1 μg/ml LPS for the indicated times. TTP protein expression was analyzed by Western blot. EF2 was used as loading control (A). The amount of secreted TNF (B) was measured by ELISA. The experiment was repeated three times. One mouse per each experiment was used, and the cells were seeded and analyzed in triplicates. The representative results are shown, and the error bars represent deviation between triplicates within one experiment. ***p < 0.005. ns, not significant.

Close modal

Because both MK2 and MK3 were shown to be important for TTP-regulated TNF production (20), we decided to combine TTPaa expression with MK2/3 double deficiency and generated MK2/3 DKO/TTPaa knock-in (DKO/TTPaa) triple mutant mice. We compared TTP and TNF expression in LPS-stimulated WT, MK2 KO, MK2/3 DKO, TTPaa, MK2/TTPaa, and DKO/TTPaa BMDMs. We observed further reduction of TTP and TNF expression in MK2/3 DKO cells compared with MK2 KO cells (Fig. 2). TTP expression levels were comparably reduced in MK2/3 DKO, TTPaa, MK2 KO/TTPaa, and DKO/TTPaa BMDMs, indicating that TTP phosphorylation at S52 and S178 by MK2/3 is necessary and sufficient for full repression of TTP activity and the restoration of TTP expression (Fig. 2A). However, the reduction of TNF production in TTPaa cells was less prominent than in MK2/3 DKO, MK2 KO/TTPaa, or DKO/TTPaa cells (Fig. 2B), indicating that the role of MK2/3 in the regulation of TNF biosynthesis is not restricted to phosphorylation of TTP at S52/S178 but includes independent processes.

FIGURE 2.

MK2/3-mediated phosphorylation of TTP at S52/S178 regulates TTP expression and TTP-dependent TNF production. BMDMs of the different genotypes were treated with 1 μg/ml LPS for the indicated times. TTP and MK2 protein expression were analyzed by Western blot (high and low exposition for TTP), and EF2 expression and ponceau staining were used as loading control (A). The amounts of secreted TNF (B) were measured by ELISA. The experiment was repeated four times. One mouse per each experiment was used, and the cells were seeded and analyzed in triplicates. The representative results are shown, and the error bars represent deviation between triplicates within one experiment. ***p < 0.005. ns, not significant.

FIGURE 2.

MK2/3-mediated phosphorylation of TTP at S52/S178 regulates TTP expression and TTP-dependent TNF production. BMDMs of the different genotypes were treated with 1 μg/ml LPS for the indicated times. TTP and MK2 protein expression were analyzed by Western blot (high and low exposition for TTP), and EF2 expression and ponceau staining were used as loading control (A). The amounts of secreted TNF (B) were measured by ELISA. The experiment was repeated four times. One mouse per each experiment was used, and the cells were seeded and analyzed in triplicates. The representative results are shown, and the error bars represent deviation between triplicates within one experiment. ***p < 0.005. ns, not significant.

Close modal

The above data suggest the existence of MK2/3 downstream substrates (which could be involved in TNF regulation in addition to TTP), further MK2/3-dependent phosphorylation sites on TTP (which have to be phosphorylated for its full inactivation), and/or MK2-regulated interaction partners of TTP that do not depend on phosphorylation at S52/S178. To determine whether any other sites on TTP besides S52/S178 could be phosphorylated via MK2, we first analyzed anisomycin-induced TTP and TTPaa phosphorylation in fibroblasts in the presence or absence of the MK2/3 inhibitor PF3644022 by quantitative SILAC-based mass spectrometry (Fig. 3A, 3B). Anisomycin is a strong activator of p38MAPK in fibroblasts, comparable to LPS in macrophages (23), leading to comparable downstream signaling toward TTP (24). In addition to S52 and S178, four other residues of TTP (T84, S85, T250, and S316, which display MK2-dependent phosphorylation) could be detected (Fig. 3C). Especially, phosphorylation of S316 is of relatively high abundance and displays at least a 4-fold MK2-dependent increase. To this end, a similar trend was observed for TTPaa, but effects were not as clear as for the WT TTP (data not shown). S316 was previously described as a critical residue for the interaction of TTP and CNOT1 (25). Because CNOT1 is the scaffold protein of the CCR4–NOT1 complex involved in TTP-dependent deadenylation and degradation of target transcripts (26, 27), it is suggested that phosphorylation of S316 in TTPaa by MK2/3 might impair the recruitment of this complex and could be responsible for the differences seen in LPS-induced TNF production between MK2/3 DKO and TTPaa. To learn more about interaction partners of TTP and their dependence on the two known MK2 phosphorylations sites, we compared the interactome of TTP and TTPaa in the above-mentioned fibroblasts that were mock treated, anisomycin stimulated, or subjected to the combined anisomycin/MK2 inhibitor PF3644022 treatment (Fig. 3A). Using SILAC, we were able to compare interactions in a quantitative manner and create clusters that describe all MK2-dependent increased and decreased interactions for TTP and TTPaa (Fig. 3D, Supplemental Table I). In total, TTP displays more regulated interactions than TTPaa, whereas TTPaa displays more interactions that are independent of MK2 (Fig. 3E). The total number of MK2-dependent increased interactions is similar for both proteins, whereas TTP shows a significantly higher number of MK2-sensitive decreased protein interactions, as compared with TTPaa. By analyzing these groups of TTP and TTPaa protein interactors, we could identify important well-characterized MK2-regulated TTP interactors, such as 14-3-3 proteins (28) (Fig. 3F) or Pabpc1 (2931). With Cnot1, 2, 8, and 10, we also identified components of the CCR4-NOT complex (26, 32) as TTP interaction partners, but, except for Cnot1, it was not possible to assign them to any cluster in a clear-cut manner. Because we were especially interested in identifying interactors of both TTP and TTPaa that are MK2 dependent but not dependent on the two MK2 sites S52/S178 and, therefore, could potentially contribute to the differences in cytokine production that we observe in genetically modified macrophages, we zoomed into these groups of 30 and 29 proteins (Fig. 3F, 3G). Most strikingly, we identified all three members of the EIF4E2/4EHP-GIGYF1/2 translation inhibitory complex (33, 34) as being MK2-dependent decreased interaction partners of TTP and TTPaa. Together with the described role of this complex in TTP-mediated translation regulation of ARE-containing mRNAs (35, 36), this suggested the possibility that MK2 could regulate the interaction with TTP in a direct or indirect manner, independent of the known MK2 sites S52/S178.

FIGURE 3.

Analysis of MK2-dependent phosphorylation of TTP by mass spectrometry. Next to the interactomes of GFP-TTP and GFP-TTPaa (A and B), anisomycin-induced TTP phosphorylation in mouse embryonic fibroblasts was analyzed in the absence and presence of the MK2 inhibitor PF3644022. Analysis of the SILAC ratios suggested that in addition to the major TTP phospho-sites S52 and S178, other sites such as T84, S85, T250, and S316 are phosphorylated by MK2. The ratios of anisomycin-treated (medium) versus mock-treated (light) and MK2i/anisomycin (heavy) versus mock-treated (light) samples are shown and indicate the MK2-dependent occurrence of TTP phosphorylation sites (C). Clustering (using GProX) of all TTP and TTPaa interaction partners that were increased or decreased, but MK2 dependent (D), revealed that TTP shows a higher degree of regulated interactions than TTPaa (E). Still, both proteins have overlapping regulated binding partners, such as ribosomal proteins (F) or members of the EIF4E2/4EHP-GIGYF1/2 complex (G).

FIGURE 3.

Analysis of MK2-dependent phosphorylation of TTP by mass spectrometry. Next to the interactomes of GFP-TTP and GFP-TTPaa (A and B), anisomycin-induced TTP phosphorylation in mouse embryonic fibroblasts was analyzed in the absence and presence of the MK2 inhibitor PF3644022. Analysis of the SILAC ratios suggested that in addition to the major TTP phospho-sites S52 and S178, other sites such as T84, S85, T250, and S316 are phosphorylated by MK2. The ratios of anisomycin-treated (medium) versus mock-treated (light) and MK2i/anisomycin (heavy) versus mock-treated (light) samples are shown and indicate the MK2-dependent occurrence of TTP phosphorylation sites (C). Clustering (using GProX) of all TTP and TTPaa interaction partners that were increased or decreased, but MK2 dependent (D), revealed that TTP shows a higher degree of regulated interactions than TTPaa (E). Still, both proteins have overlapping regulated binding partners, such as ribosomal proteins (F) or members of the EIF4E2/4EHP-GIGYF1/2 complex (G).

Close modal

Taken together, the mass spectrometry results suggested that sites other than the canonical S52/S178 sites in TTP could be MK2 sensitive and that a group of interacting proteins of TTP are regulated in an MK2-dependent manner, independent of the canonical sites. These interactions can be increased or decreased and might contribute to the differences in cytokine production observed in this study.

We next analyzed whether MK2/3-mediated phosphorylation of TTP on the sites additional to S52/S178 might influence its stability. We compared TTP protein stability in LPS-stimulated and cycloheximide-treated WT, MK2/3 DKO, TTPaa, and DKO/TTPaa BMDMs. Phosphorylation of TTP at S52/S178 by MK2/3 significantly stabilized TTP protein because absence of this phosphorylation in TTPaa cells strongly increased TTP protein degradation (Fig. 4). However, the finding that TTP stability is comparable in TTPaa, MK2/3 DKO, and MK2/3 DKO/TTPaa cells indicates a dominant role of S52/S178 phosphorylation in the regulation of TTP protein stability and makes a role of, for example, S316 rather unlikely.

FIGURE 4.

LPS-induced TTP phosphorylation at Ser52 and Ser178 by MK2/3 has a dominant effect on TTP protein stabilization. BMDMs of indicated genotypes were treated with 1 μg/ml LPS for 2 h followed by the treatment with 5 μg/ml cycloheximide for the time indicated. TTP and MK2 expression were analyzed by Western blot (high and low exposition for TTP). EF2 expression was used as loading control.

FIGURE 4.

LPS-induced TTP phosphorylation at Ser52 and Ser178 by MK2/3 has a dominant effect on TTP protein stabilization. BMDMs of indicated genotypes were treated with 1 μg/ml LPS for 2 h followed by the treatment with 5 μg/ml cycloheximide for the time indicated. TTP and MK2 expression were analyzed by Western blot (high and low exposition for TTP). EF2 expression was used as loading control.

Close modal

We further analyzed the role of MK2/3-mediated phosphorylation of TTP in cytokine production and TTP expression by direct comparison of LPS-stimulated WT, MK2/3 DKO, TTPaa, and DKO/TTPaa BMDMs (Fig. 5). The well-known TTP targets TNF, IL-10, CXCL1, CXCL2, and IL-6 (37) are under stringent control of the p38/MK2/3/TTP axis. Production of these cytokines is strongly inhibited in TTPaa and even further reduced in DKO/TTPaa BMDMs (Fig. 5B). Thus, in the supernatants of LPS-treated MK2/3 DKO or DKO/TTPaa cells, TNF expression varies between 0 and 10% of WT, and IL-10 expression varies between 0 and 3% of WT cells. In TTPaa cells, TNF expression shows ∼17, 10, and 19% of WT after 2, 4, and 8 h of LPS treatment, respectively, and IL-10 expression shows 0, 16, and 35% of WT after 2, 4, and 8 h of LPS treatment, respectively. Obviously, the decisive role of MK2/3 in TNF and IL-10 regulation depends on phosphorylation of TTP at S52 and S178, but it is not restricted to this event. It can be supposed that phosphorylation of TTP at S52/S178 inactivates TTP to 80–90%, but for its full inactivation, TTP also needs to be phosphorylated on other MK2/3-dependent sites, such as S316. However, we also cannot exclude that, in addition to the major role of TTP, other MK2/3 substrates play a supplementary role in the regulation of TNF production. Candidates for such substrates are the endoplasmic reticulum–associated protein Ube2j1, which was shown to be involved in TNF translation at the rough endoplasmic reticulum (16), or the mRNA-binding protein hnRNP A0, which specifically interacts with TNF mRNA (38). Because it is known that deletion of MK2/3 leads to reduced stabilization and activity of p38, as well (39), it could also be possible that there is further involvement of substrates of p38. Interestingly, the TTP antagonist HuR is known to be directly phosphorylated by p38 MAPK at T118 (40). This phosphorylation increases the cytoplasmic localization of HuR and, therefore, could facilitate the replacement of TTP by HuR at the cytoplasmic ARE-bearing mRNAs (3).

FIGURE 5.

The role of MK2/3 in cytokine production is not restricted to phosphorylation of TTP at S52/S178. BMDMs of indicated genotypes were treated with 1 μg/ml LPS for the time indicated. TTP protein expression was analyzed by Western blot, and EF2 expression was used as loading control (A). Quantification of secreted cytokines by multiplex bead capture assays (B). Each experiment was repeated at least three times. One mouse per each experiment was used, and the cells were seeded and analyzed in triplicates. The representative results are shown, and the error bars represent deviation between triplicates within one experiment. *p < 0.05, **p < 0.01, ***p < 0.005. ns, not significant.

FIGURE 5.

The role of MK2/3 in cytokine production is not restricted to phosphorylation of TTP at S52/S178. BMDMs of indicated genotypes were treated with 1 μg/ml LPS for the time indicated. TTP protein expression was analyzed by Western blot, and EF2 expression was used as loading control (A). Quantification of secreted cytokines by multiplex bead capture assays (B). Each experiment was repeated at least three times. One mouse per each experiment was used, and the cells were seeded and analyzed in triplicates. The representative results are shown, and the error bars represent deviation between triplicates within one experiment. *p < 0.05, **p < 0.01, ***p < 0.005. ns, not significant.

Close modal

In contrast to the well-known TTP targets TNF, IL-10, CXCL1, CXCL2, and IL-6, the level of CCL3/MIP-1α was significantly reduced in MK2/3 DKO and TTPaa cells but was not further changed in DKO/TTPaa cells (Fig. 5B). Interestingly, some of the LPS-induced inflammatory cytokines, such as CCL4/MIP-1β and CCL2/MCP-1, demonstrate a continuous upregulation independent of MK2/3 and of the two phosphorylation sites in TTP (Fig. 5B), indicating distinct regulatory mechanisms. This finding is rather unexpected because both CCL2- and CCL4-mRNA contain ARE-motifs to which TTP binding has been unequivocally detected (20, 41).

To analyze whether the above ex vivo findings are also relevant in vivo, we applied the CLP model of systemic inflammation in mice (42). Serum cytokine levels were measured 3, 6, and 20 h after puncture in WT, MK2/3 DKO, TTPaa, and DKO/TTPaa mice (Fig. 6A, see Supplemental Fig. 1 for significance tests). Although TNF was not detectable at these time points, IL-10, CXCL1, IL-6, CXCL2, and CCL3 displayed genotype-specific expression levels similar to the ex vivo situation. Interestingly, CCL2 and CCL4 expression, which were not dependent on the genotype of the BMDMs ex vivo, also displayed a strong genotype dependence in vivo, indicating the involvement of further cell populations, such as neutrophils, in the more-complex in vivo situation of sepsis. To further monitor the septic response at the cellular level, we also counted WBC and PMN in the PL after 20 h (Fig. 6B). Unexpectedly, the TTPaa mutant mice do not display a significant reduction in these cell types in the lavage. In contrast, the lavage of MK2/3 DKO and DKO/TTPaa mice contained significantly reduced cell numbers, probably reflecting a defect in cell migration into the peritoneum. This is in agreement with the already described essential role of MK2/3 for cell migration (39, 43).

FIGURE 6.

TTP is involved in MK2/3-regulated cytokine production but not in MK2/3-regulated neutrophil transmigration in vivo. Mice of different genotypes were subjected to CLP, and blood was obtained at the indicated time points. Quantification of secreted cytokines by multiplex bead capture assays in blood serum (A). WBC counts and PMN counts from PL samples at 20 h after CLP operation (B). Data are expressed as mean ± SE (WT: n = 5; MK2/3: n = 3; TTPaa: n = 12; and DKO/TTPaa: n = 7 mice per group). *p < 0.05, **p < 0.01, ***p < 0.005. ns, not significant.

FIGURE 6.

TTP is involved in MK2/3-regulated cytokine production but not in MK2/3-regulated neutrophil transmigration in vivo. Mice of different genotypes were subjected to CLP, and blood was obtained at the indicated time points. Quantification of secreted cytokines by multiplex bead capture assays in blood serum (A). WBC counts and PMN counts from PL samples at 20 h after CLP operation (B). Data are expressed as mean ± SE (WT: n = 5; MK2/3: n = 3; TTPaa: n = 12; and DKO/TTPaa: n = 7 mice per group). *p < 0.05, **p < 0.01, ***p < 0.005. ns, not significant.

Close modal

The role of TTP phosphorylation in the regulation of inflammatory cytokine production by MK2/3 was analyzed by generation of mice with combined MK2/MK3/TTP-targeted alleles. The comparison of BMDMs of different genotypes for MK2/3/TTP and combinations thereof made it possible to discriminate cytokines with different characteristics of LPS-stimulated TTP-dependent upregulation. TTP phosphorylation is a constitutive event downstream to p38MAPK and MK2/3 activation. The main MAP3K for p38MAPK in macrophages is TAK1, which is activated in an MyD88-dependent manner (44, 45). Therefore, TTP phosphorylation is an MyD88-dependent event in macrophages and proceeds in response to various stimuli that share the MyD88 arm of TLR signaling. Hence, it is unlikely that the different characteristics of LPS-stimulated TTP-dependent upregulation of cytokines by MK2/3 arise from major differences in upstream signaling events. It is more likely that these different characteristics result from differences in downstream signaling, namely from the differences in the ability of TTP to bind to and destabilize the various cytokine mRNAs.

The comparison of mice of different genotypes for MK2/3/TTP in the CLP model of sepsis confirmed the role of this signaling axis for cytokine production also in vivo. It also revealed that cytokine production in vivo is not only the result of activation of macrophages but the concerted action of many cell types, also making CCL2 and CCL4 production dependent on the MK2/3/TTP axis. Although serum levels of the neutrophil chemokines CXCL1 and CXCL2 were strongly reduced in all three genetically modified mouse strains, the accumulation of neutrophils in the peritoneum was reduced only in mice lacking MK2 and MK3. This uncoupling suggests that neutrophil migration to the peritoneum may be driven directly by microbial products or by another TTP-independent chemokine that was not measured in this study. It is also likely that MK2 and MK3 directly regulate neutrophil migration in a manner that does not depend on the phosphorylation of TTP, as has been reported in other cell types (39, 43).

It became clear that MK2/3 are the only relevant protein kinases that phosphorylate TTP at S52 and S178. MK2/3-mediated TTP phosphorylation at S52 and S178 is essential and sufficient for its stabilization during inflammatory response. However, the role of MK2/3 in the regulation of TNF biosynthesis is not restricted to phosphorylation of TTP at S52/S178 but includes some independent processes. Because the additional TTP phosphorylation site S316 demonstrates high abundance and displays at least a 4-fold MK2-dependent increase in quantitative SILAC-based mass spectrometry, we propose that for its full inactivation, TTP needs to be additionally phosphorylated on S316 (Fig. 7).

FIGURE 7.

Schematic representation of the role of MK2/3-mediated phosphorylation of TTP on its activity. Upon activation, MK2/3 phosphorylates TTP at the two major sites S52 and S178 (red) and some additional sites, such as S316 (black). Absence of S52 and S178 phosphorylation in the TTPaa cells leads to significant reduction of TNF production, but for its full inactivation, TTP needs to be phosphorylated at the sites additional to S52 and S178.

FIGURE 7.

Schematic representation of the role of MK2/3-mediated phosphorylation of TTP on its activity. Upon activation, MK2/3 phosphorylates TTP at the two major sites S52 and S178 (red) and some additional sites, such as S316 (black). Absence of S52 and S178 phosphorylation in the TTPaa cells leads to significant reduction of TNF production, but for its full inactivation, TTP needs to be phosphorylated at the sites additional to S52 and S178.

Close modal

We thank Kathrin Laaß (Medical School Hannover) for excellent technical assistance.

The online version of this article contains supplemental material.

Abbreviations used in this article:

ARE

AU-rich element

BMDM

bone marrow–derived macrophage

CLP

cecal ligation and puncture

DKO

double knockout

KO

knockout

PL

peritoneal lavage

PMN

polymorphonuclear neutrophil

SILAC

stable isotope labeling by amino acids in cell culture

TTP

tristetraprolin

TTPaa

TTP-S52A-S178A

WT

wild-type.

1
Anderson
,
P.
2008
.
Post-transcriptional control of cytokine production.
Nat. Immunol.
9
:
353
359
.
2
Ross
,
E. A.
,
T.
Smallie
,
Q.
Ding
,
J. D.
O’Neil
,
H. E.
Cunliffe
,
T.
Tang
,
D. R.
Rosner
,
I.
Klevernic
,
N. A.
Morrice
,
C.
Monaco
, et al
.
2015
.
Dominant suppression of inflammation via targeted mutation of the mRNA destabilizing protein tristetraprolin.
J. Immunol.
195
:
265
276
.
3
Tiedje
,
C.
,
N.
Ronkina
,
M.
Tehrani
,
S.
Dhamija
,
K.
Laass
,
H.
Holtmann
,
A.
Kotlyarov
,
M.
Gaestel
.
2012
.
The p38/MK2-driven exchange between tristetraprolin and HuR regulates AU-rich element-dependent translation.
PLoS Genet.
8
: e1002977.
4
Tchen
,
C. R.
,
M.
Brook
,
J.
Saklatvala
,
A. R.
Clark
.
2004
.
The stability of tristetraprolin mRNA is regulated by mitogen-activated protein kinase p38 and by tristetraprolin itself.
J. Biol. Chem.
279
:
32393
32400
.
5
Brook
,
M.
,
C. R.
Tchen
,
T.
Santalucia
,
J.
McIlrath
,
J. S. C.
Arthur
,
J.
Saklatvala
,
A. R.
Clark
.
2006
.
Posttranslational regulation of tristetraprolin subcellular localization and protein stability by p38 mitogen-activated protein kinase and extracellular signal-regulated kinase pathways.
Mol. Cell. Biol.
26
:
2408
2418
.
6
Hitti
,
E.
,
T.
Iakovleva
,
M.
Brook
,
S.
Deppenmeier
,
A. D.
Gruber
,
D.
Radzioch
,
A. R.
Clark
,
P. J.
Blackshear
,
A.
Kotlyarov
,
M.
Gaestel
.
2006
.
Mitogen-activated protein kinase-activated protein kinase 2 regulates tumor necrosis factor mRNA stability and translation mainly by altering tristetraprolin expression, stability, and binding to adenine/uridine-rich element.
Mol. Cell. Biol.
26
:
2399
2407
.
7
Cao
,
H.
,
R.
Lin
.
2008
.
Phosphorylation of recombinant tristetraprolin in vitro.
Protein J.
27
:
163
169
.
8
Cao
,
H.
,
L. J.
Deterding
,
J. D.
Venable
,
E. A.
Kennington
,
J. R.
Yates
III
,
K. B.
Tomer
,
P. J.
Blackshear
.
2006
.
Identification of the anti-inflammatory protein tristetraprolin as a hyperphosphorylated protein by mass spectrometry and site-directed mutagenesis.
Biochem. J.
394
:
285
297
.
9
Cao
,
H.
,
L. J.
Deterding
,
P. J.
Blackshear
.
2014
.
Identification of a major phosphopeptide in human tristetraprolin by phosphopeptide mapping and mass spectrometry.
PLoS One
9
: e100977.
10
Stokoe
,
D.
,
K.
Engel
,
D. G.
Campbell
,
P.
Cohen
,
M.
Gaestel
.
1992
.
Identification of MAPKAP kinase 2 as a major enzyme responsible for the phosphorylation of the small mammalian heat shock proteins.
FEBS Lett.
313
:
307
313
.
11
Rousseau
,
S.
,
M.
Peggie
,
D. G.
Campbell
,
A. R.
Nebreda
,
P.
Cohen
.
2005
.
Nogo-B is a new physiological substrate for MAPKAP-K2.
Biochem. J.
391
:
433
440
.
12
Ronkina
,
N.
,
M. B.
Menon
,
J.
Schwermann
,
J. S. C.
Arthur
,
H.
Legault
,
J.-B.
Telliez
,
U. S.
Kayyali
,
A. R.
Nebreda
,
A.
Kotlyarov
,
M.
Gaestel
.
2011
.
Stress induced gene expression: a direct role for MAPKAP kinases in transcriptional activation of immediate early genes.
Nucleic Acids Res.
39
:
2503
2518
.
13
Jaco
,
I.
,
A.
Annibaldi
,
N.
Lalaoui
,
R.
Wilson
,
T.
Tenev
,
L.
Laurien
,
C.
Kim
,
K.
Jamal
,
S.
Wicky John
,
G.
Liccardi
, et al
.
2017
.
MK2 phosphorylates RIPK1 to prevent TNF-induced cell death.
Mol. Cell
66
:
698
710.e5
.
14
Dondelinger
,
Y.
,
T.
Delanghe
,
D.
Rojas-Rivera
,
D.
Priem
,
T.
Delvaeye
,
I.
Bruggeman
,
F.
Van Herreweghe
,
P.
Vandenabeele
,
M. J. M.
Bertrand
.
2017
.
MK2 phosphorylation of RIPK1 regulates TNF-mediated cell death.
Nat. Cell Biol.
19
:
1237
1247
.
15
Menon
,
M. B.
,
J.
Gropengießer
,
J.
Fischer
,
L.
Novikova
,
A.
Deuretzbacher
,
J.
Lafera
,
H.
Schimmeck
,
N.
Czymmeck
,
N.
Ronkina
,
A.
Kotlyarov
, et al
.
2017
.
p38MAPK/MK2-dependent phosphorylation controls cytotoxic RIPK1 signalling in inflammation and infection.
Nat. Cell Biol.
19
:
1248
1259
.
16
Menon
,
M. B.
,
C.
Tiedje
,
J.
Lafera
,
N.
Ronkina
,
T.
Konen
,
A.
Kotlyarov
,
M.
Gaestel
.
2013
.
Endoplasmic reticulum-associated ubiquitin-conjugating enzyme Ube2j1 is a novel substrate of MK2 (MAPKAP kinase-2) involved in MK2-mediated TNFα production.
Biochem. J.
456
:
163
172
.
17
Kotlyarov
,
A.
,
A.
Neininger
,
C.
Schubert
,
R.
Eckert
,
C.
Birchmeier
,
H. D.
Volk
,
M.
Gaestel
.
1999
.
MAPKAP kinase 2 is essential for LPS-induced TNF-alpha biosynthesis.
Nat. Cell Biol.
1
:
94
97
.
18
Zaru
,
R.
,
N.
Ronkina
,
M.
Gaestel
,
J. S. C.
Arthur
,
C.
Watts
.
2007
.
The MAPK-activated kinase Rsk controls an acute toll-like receptor signaling response in dendritic cells and is activated through two distinct pathways. [Published erratum appears in 2008 Nat. Immunol. 9: 105.]
Nat. Immunol.
8
:
1227
1235
.
19
Mahtani
,
K. R.
,
M.
Brook
,
J. L.
Dean
,
G.
Sully
,
J.
Saklatvala
,
A. R.
Clark
.
2001
.
Mitogen-activated protein kinase p38 controls the expression and posttranslational modification of tristetraprolin, a regulator of tumor necrosis factor alpha mRNA stability.
Mol. Cell. Biol.
21
:
6461
6469
.
20
Tiedje
,
C.
,
M. D.
Diaz-Muñoz
,
P.
Trulley
,
H.
Ahlfors
,
K.
Laaß
,
P. J.
Blackshear
,
M.
Turner
,
M.
Gaestel
.
2016
.
The RNA-binding protein TTP is a global post-transcriptional regulator of feedback control in inflammation.
Nucleic Acids Res.
44
:
7418
7440
.
21
Tollenaere
,
M. A. X.
,
B. H.
Villumsen
,
M.
Blasius
,
J. C.
Nielsen
,
S. A.
Wagner
,
J.
Bartek
,
P.
Beli
,
N.
Mailand
,
S.
Bekker-Jensen
.
2015
.
p38- and MK2-dependent signalling promotes stress-induced centriolar satellite remodelling via 14-3-3-dependent sequestration of CEP131/AZI1.
Nat. Commun.
6
:
10075
.
22
Ronkina
,
N.
,
A.
Kotlyarov
,
O.
Dittrich-Breiholz
,
M.
Kracht
,
E.
Hitti
,
K.
Milarski
,
R.
Askew
,
S.
Marusic
,
L.-L.
Lin
,
M.
Gaestel
,
J.-B.
Telliez
.
2007
.
The mitogen-activated protein kinase (MAPK)-activated protein kinases MK2 and MK3 cooperate in stimulation of tumor necrosis factor biosynthesis and stabilization of p38 MAPK.
Mol. Cell. Biol.
27
:
170
181
.
23
Cano
,
E.
,
Y. N.
Doza
,
R.
Ben-Levy
,
P.
Cohen
,
L. C.
Mahadevan
.
1996
.
Identification of anisomycin-activated kinases p45 and p55 in murine cells as MAPKAP kinase-2.
Oncogene
12
:
805
812
.
24
Clark
,
A. R.
,
J. L. E.
Dean
.
2016
.
The control of inflammation via the phosphorylation and dephosphorylation of tristetraprolin: a tale of two phosphatases.
Biochem. Soc. Trans.
44
:
1321
1337
.
25
Fabian
,
M. R.
,
F.
Frank
,
C.
Rouya
,
N.
Siddiqui
,
W. S.
Lai
,
A.
Karetnikov
,
P. J.
Blackshear
,
B.
Nagar
,
N.
Sonenberg
.
2013
.
Structural basis for the recruitment of the human CCR4-NOT deadenylase complex by tristetraprolin.
Nat. Struct. Mol. Biol.
20
:
735
739
.
26
Marchese
,
F. P.
,
A.
Aubareda
,
C.
Tudor
,
J.
Saklatvala
,
A. R.
Clark
,
J. L. E.
Dean
.
2010
.
MAPKAP kinase 2 blocks tristetraprolin-directed mRNA decay by inhibiting CAF1 deadenylase recruitment.
J. Biol. Chem.
285
:
27590
27600
.
27
Clement
,
S. L.
,
C.
Scheckel
,
G.
Stoecklin
,
J.
Lykke-Andersen
.
2011
.
Phosphorylation of tristetraprolin by MK2 impairs AU-rich element mRNA decay by preventing deadenylase recruitment.
Mol. Cell. Biol.
31
:
256
266
.
28
Stoecklin
,
G.
,
T.
Stubbs
,
N.
Kedersha
,
S.
Wax
,
W. F. C.
Rigby
,
T. K.
Blackwell
,
P.
Anderson
.
2004
.
MK2-induced tristetraprolin:14-3-3 complexes prevent stress granule association and ARE-mRNA decay.
EMBO J.
23
:
1313
1324
.
29
Bollig
,
F.
,
R.
Winzen
,
M.
Gaestel
,
S.
Kostka
,
K.
Resch
,
H.
Holtmann
.
2003
.
Affinity purification of ARE-binding proteins identifies polyA-binding protein 1 as a potential substrate in MK2-induced mRNA stabilization.
Biochem. Biophys. Res. Commun.
301
:
665
670
.
30
Rowlett
,
R. M.
,
C. A.
Chrestensen
,
M. J.
Schroeder
,
M. G.
Harp
,
J. W.
Pelo
,
J.
Shabanowitz
,
R.
DeRose
,
D. F.
Hunt
,
T. W.
Sturgill
,
M. T.
Worthington
.
2008
.
Inhibition of tristetraprolin deadenylation by poly(A) binding protein.
Am. J. Physiol. Gastrointest. Liver Physiol.
295
:
G421
G430
.
31
Kedar
,
V. P.
,
B. E.
Zucconi
,
G. M.
Wilson
,
P. J.
Blackshear
.
2012
.
Direct binding of specific AUF1 isoforms to tandem zinc finger domains of tristetraprolin (TTP) family proteins.
J. Biol. Chem.
287
:
5459
5471
.
32
Sandler
,
H.
,
J.
Kreth
,
H. T. M.
Timmers
,
G.
Stoecklin
.
2011
.
Not1 mediates recruitment of the deadenylase Caf1 to mRNAs targeted for degradation by tristetraprolin.
Nucleic Acids Res.
39
:
4373
4386
.
33
Morita
,
M.
,
L. W.
Ler
,
M. R.
Fabian
,
N.
Siddiqui
,
M.
Mullin
,
V. C.
Henderson
,
T.
Alain
,
B. D.
Fonseca
,
G.
Karashchuk
,
C. F.
Bennett
, et al
.
2012
.
A novel 4EHP-GIGYF2 translational repressor complex is essential for mammalian development.
Mol. Cell. Biol.
32
:
3585
3593
.
34
Peter
,
D.
,
R.
Weber
,
F.
Sandmeir
,
L.
Wohlbold
,
S.
Helms
,
P.
Bawankar
,
E.
Valkov
,
C.
Igreja
,
E.
Izaurralde
.
2017
.
GIGYF1/2 proteins use auxiliary sequences to selectively bind to 4EHP and repress target mRNA expression.
Genes Dev.
31
:
1147
1161
.
35
Tao
,
X.
,
G.
Gao
.
2015
.
Tristetraprolin recruits eukaryotic initiation factor 4E2 to repress translation of AU-rich element-containing mRNAs.
Mol. Cell. Biol.
35
:
3921
3932
.
36
Fu
,
R.
,
M. T.
Olsen
,
K.
Webb
,
E. J.
Bennett
,
J.
Lykke-Andersen
.
2016
.
Recruitment of the 4EHP-GYF2 cap-binding complex to tetraproline motifs of tristetraprolin promotes repression and degradation of mRNAs with AU-rich elements.
RNA
22
:
373
382
.
37
Stoecklin
,
G.
,
S. A.
Tenenbaum
,
T.
Mayo
,
S. V.
Chittur
,
A. D.
George
,
T. E.
Baroni
,
P. J.
Blackshear
,
P.
Anderson
.
2008
.
Genome-wide analysis identifies interleukin-10 mRNA as target of tristetraprolin.
J. Biol. Chem.
283
:
11689
11699
.
38
Rousseau
,
S.
,
N.
Morrice
,
M.
Peggie
,
D. G.
Campbell
,
M.
Gaestel
,
P.
Cohen
.
2002
.
Inhibition of SAPK2a/p38 prevents hnRNP A0 phosphorylation by MAPKAP-K2 and its interaction with cytokine mRNAs.
EMBO J.
21
:
6505
6514
.
39
Kotlyarov
,
A.
,
Y.
Yannoni
,
S.
Fritz
,
K.
Laass
,
J.-B.
Telliez
,
D.
Pitman
,
L.-L.
Lin
,
M.
Gaestel
.
2002
.
Distinct cellular functions of MK2.
Mol. Cell. Biol.
22
:
4827
4835
.
40
Lafarga
,
V.
,
A.
Cuadrado
,
I.
Lopez de Silanes
,
R.
Bengoechea
,
O.
Fernandez-Capetillo
,
A. R.
Nebreda
.
2009
.
p38 Mitogen-activated protein kinase- and HuR-dependent stabilization of p21(Cip1) mRNA mediates the G(1)/S checkpoint.
Mol. Cell. Biol.
29
:
4341
4351
.
41
Sedlyarov
,
V.
,
J.
Fallmann
,
F.
Ebner
,
J.
Huemer
,
L.
Sneezum
,
M.
Ivin
,
K.
Kreiner
,
A.
Tanzer
,
C.
Vogl
,
I.
Hofacker
,
P.
Kovarik
.
2016
.
Tristetraprolin binding site atlas in the macrophage transcriptome reveals a switch for inflammation resolution.
Mol. Syst. Biol.
12
:
868
.
42
Kumpers
,
P.
,
F.
Gueler
,
S.
David
,
P. V.
Slyke
,
D. J.
Dumont
,
J.-K.
Park
,
C. L.
Bockmeyer
,
S. M.
Parikh
,
H.
Pavenstädt
,
H.
Haller
,
N.
Shushakova
.
2011
.
The synthetic tie2 agonist peptide vasculotide protects against vascular leakage and reduces mortality in murine abdominal sepsis.
Crit. Care
15
:
R261
.
43
Menon
,
M. B.
,
N.
Ronkina
,
J.
Schwermann
,
A.
Kotlyarov
,
M.
Gaestel
.
2009
.
Fluorescence-based quantitative scratch wound healing assay demonstrating the role of MAPKAPK-2/3 in fibroblast migration.
Cell Motil. Cytoskeleton
66
:
1041
1047
.
44
Cheung
,
P. C. F.
,
D. G.
Campbell
,
A. R.
Nebreda
,
P.
Cohen
.
2003
.
Feedback control of the protein kinase TAK1 by SAPK2a/p38alpha.
EMBO J.
22
:
5793
5805
.
45
Shim
,
J.H.
,
C.
Xiao
,
A. E.
Paschal
,
S. T.
Bailey
,
P.
Rao
,
M. S.
Hayden
,
K.Y.
Lee
,
C.
Bussey
,
M.
Steckel
,
N.
Tanaka
, et al
.
2005
.
TAK1, but not TAB1 or TAB2, plays an essential role in multiple signaling pathways in vivo.
Genes Dev.
19
:
2668
2681
.

The authors have no financial conflicts of interest.

Supplementary data