E3 ubiquitin ligase NEDD4L negatively regulates inflammation by promoting ubiquitination of MEKK2

Abstract

Aberrant activation of inflammation signaling triggered by tumor necrosis factor α (TNF‐α), interleukin‐1 (IL‐1), and interleukin‐17 (IL‐17) is associated with immunopathology. Here, we identify neural precursor cells expressed developmentally down‐regulated gene 4‐like (NEDD4L), a HECT type E3 ligase, as a common negative regulator of signaling induced by TNF‐α, IL‐1, and IL‐17. NEDD4L modulates the degradation of mitogen‐activated protein kinase kinase kinase 2 (MEKK2) via constitutively and directly binding to MEKK2 and promotes its poly‐ubiquitination. In interleukin‐17 receptor (IL‐17R) signaling, Nedd4l knockdown or deficiency enhances IL‐17‐induced p38 and NF‐κB activation and the production of proinflammatory cytokines and chemokines in a MEKK2‐dependent manner. We further show that IL‐17‐induced MEKK2 Ser520 phosphorylation is required not only for downstream p38 and NF‐κB activation but also for NEDD4L‐mediated MEKK2 degradation and the subsequent shutdown of IL‐17R signaling. Importantly, Nedd4l‐deficient mice show increased susceptibility to IL‐17‐induced inflammation and aggravated symptoms of experimental autoimmune encephalomyelitis (EAE) in IL‐17R signaling‐dependent manner. These data suggest that NEDD4L acts as an inhibitor of IL‐17R signaling, which ameliorates the pathogenesis of IL‐17‐mediated autoimmune diseases.

NEDD4L functions as a negative regulator of IL‐17‐triggered signaling and downstream cytokine and chemokine expression. NEDD4L participates in the pathogenesis of IL‐17‐mediated inflammation and autoimmune diseases by promoting the ubiquitination‐mediated proteasomal degradation of MEKK2.

Introduction

Inflammation represents an adaptive and dynamic host response against pathogens, physical/chemical injuries, toxins, or allergens. The stimulant triggers leukocyte infiltration, cytokines, and chemokines expression, and eventual subsiding of proinflammatory signaling. Dysregulated or prolonged inflammation has been closely related to a variety of human diseases such as autoimmune diseases, cancer, cardiovascular disease, and metabolic disorders (Coussens & Werb, 2002; Hotamisligil, 2017; Liu et al, 2020). Thus, immune‐regulatory and negative feedback mechanisms are critically required to harness inflammation and prevent excessive immunopathology post‐inflammation (Medzhitov, 2008).

Inflammatory cytokines such as tumor necrosis factor α (TNF‐α), interleukin‐1 (IL‐1), and interleukin‐17 (IL‐17) are key regulators in the inflammation process. By binding IL‐1R, IL‐1 activates the MyD88‐IRAK1‐TRAF6‐TAK1 cascade (Boraschi et al, 2018). TNF‐α exerts its biological effects by binding two different receptors, TNFR1 and TNFR2. Upon activation, TNFR1 recruits adaptor proteins including TRADD, TRAF2, cIAP1, cIAP2, and RIP1. TNFR2 recruits TRAF2 and also TRAF1, TRAF3, cIAP1, and cIAP2, through their binding to TRAF2 (Kalliolias & Ivashkiv, 2016). Although with different biological functions, both IL‐1 and TNF‐α activate downstream MAPK and NF‐κB to induce inflammatory responses. Among the six members of the IL‐17 family (i.e., IL‐17A to IL‐17F), IL‐17A, commonly called IL‐17, is the first identified and also the most intensively studied member (Iwakura et al, 2011). IL‐17F shares the strongest sequence homology with IL‐17A. IL‐17A and IL‐17F activate downstream signaling pathways through a heterodimeric receptor composed of the IL‐17RA and IL‐17RC subunits (Gaffen, 2009). Upon IL‐17 and IL‐17F ligation, IL‐17RA and IL‐17RC recruit TRAF6 through adaptor protein Act1, ultimately activating MAPK, NF‐κB, C/EBPδ, and C/EBPβ to induce proinflammatory cytokines and chemokines expression (Qu et al, 2012).

TNF‐α‐, IL‐1‐, and IL‐17‐induced signaling pathway protectively responds to pathogen invasion, injuries, toxins, and allergens. However, prolonged activated TNF‐α‐, IL‐1‐, and IL‐17‐induced signaling will cause the development of autoimmune diseases such as experimental autoimmune encephalomyelitis (EAE) (Sutton et al, 2009; Rostami & Ciric, 2013; Wolf et al, 2017). As is well demonstrated, deficiency or destruction of the negative regulator of the inflammatory signaling is responsible for the development of autoimmune diseases. As a negative regulator involved in TNF‐α‐ and IL‐1‐induced inflammatory signaling pathway, TNF‐α‐induced protein 3 (TNFIP3, also known as A20) has been reported to be associated with many human inflammation‐related diseases (Catrysse et al, 2014).

Neural precursor cell expressed developmentally down‐regulated gene 4‐like (NEDD4L) is a highly conserved HECT‐type E3 ligase. NEDD4L mediates ubiquitination of epithelial Na⁺ channel (ENaC) and inhibits cell surface expression of ENaC in epithelia (Goulet et al, 1998; Abriel et al, 1999). NEDD4L also limits TGF‐β signal transduction by mediating Smad2/3 poly‐ubiquitination and degradation (Gao et al, 2009). NEDD4L has been reported to be associated with nerve disease in human and mouse models (Charbonnier‐Beaupel et al, 2015; Jewett et al, 2016; Manning & Kumar, 2018). Our recent data demonstrate that NEDD4L promotes Lys‐29‐linked ubiquitination of TRAF3, and thus positively regulates antivirus innate response (Gao et al, 2021). We also showed in a previous study that NEDD4L negatively regulates keratinocyte hyperplasia by promoting Lys‐27‐linked ubiquitination of GP130 and proteasomal degradation in keratinocytes (Liu et al, 2021). However, the physiologic role of NEDD4L in autoimmune diseases remains largely unknown.

In the present study, we identify NEDD4L as a common negative regulator of IL‐17‐, TNF‐α‐, and IL‐1α‐induced signaling. NEDD4L constitutively and directly binds MEKK2 and reduces MEKK2 expression by promoting K27‐linked poly‐ubiquitination of MEKK2. Mekk2 knockdown inhibits IL‐17‐induced p38 and NF‐κB activation, as well as IL‐17‐induced proinflammatory cytokine and chemokine production. NEDD4L inhibits IL‐17‐induced inflammation in a MEKK2‐dependent manner. Further study shows that IL‐17 induces MEKK2 Ser520 phosphorylation, which is not only crucial for the activation of p38 and NF‐κB pathways but also for NEDD4L‐mediated MEKK2 degradation to suppress IL‐17R signaling. Furthermore, Nedd4l‐deficient mice are more susceptible to IL‐17‐induced peritonitis and pneumonia, and IL‐17R‐related EAE. These results suggest that as a negative regulator of MEKK2‐dependent IL‐17R‐mediated inflammation, NEDD4L severs as a pivotal immune modulator for dysregulated inflammation‐mediated autoimmune diseases.

Results

NEDD4L negatively regulates IL‐17‐, TNF‐α‐, and IL‐1α‐induced inflammation

To explore the role of NEDD4L in inflammation, we determined the effect of Nedd4l knockdown on IL‐17‐induced proinflammatory cytokine and chemokine expression. Synthesized Nedd4l‐specific small‐interfering RNA (siNedd4l) was transfected into HeLa cells to inhibit endogenous NEDD4L expression (Fig EV1A). Nedd4l knockdown significantly increased IL‐17‐induced Il‐6, Cxcl2, Ccl20 mRNA expression, and IL‐6 and CXCL2 production (Fig 1A and B). We next investigated the role of NEDD4L in inflammatory responses in embryonic fibroblasts (MEFs), which is well established for the IL‐17R signaling study (Zhong et al, 2012). As shown in Fig 1C and D, IL‐17‐induced Il‐6, Cxcl2, and Ccl20 mRNA expression, and IL‐6 and CXCL2 production were significantly increased in MEFs with Nedd4l homozygous deficiency (Nedd4l ^−/−) compared to the wild‐type (Nedd4l ^+/+) MEFs. Since IL‐17F has similar functions as IL‐17, Nedd4l homozygous deficiency significantly increased IL‐17F‐induced Il‐6, Cxcl2, and Ccl20 mRNA expression in MEFs (Fig 1E). Such significantly increased IL‐17A‐ or IL‐17F‐induced Il‐6, Cxcl2, and Ccl20 mRNA expression were also observed in MEFs with Nedd4l heterozygous deficiency (Nedd4l ^+/−) (Fig EV1B–D). Lung epithelial cells and astrocytes have been also identified to play a crucial role in IL‐17‐induced inflammation and autoimmune diseases (Kang et al, 2010; Zhong et al, 2012). Nedd4l deficiency in these cells resulted in significantly increased IL‐17‐induced Il‐6, Cxcl2, and Ccl20 mRNA expression compared to the wild‐type control cells (Figs 1F and G, and EV1E and F). These results suggest that NEDD4L plays a crucial role in IL‐17‐induced inflammation in vitro.

Figure EV1. Nedd4l deficiency promotes IL‐17‐induced inflammatory cytokine and chemokine expression.

• A, B
  Immunoblot analysis of NEDD4L in HeLa cells silenced with scramble or human Nedd4l‐specific siRNA(A), and in MEFs cells (B).
• C, D
  Real‐time PCR analysis of Il‐6, Cxcl2, and Ccl20 mRNA expression in Nedd4l ^+/+‐ and Nedd4l‐deficient heterozygous (Nedd4l ^+/−) MEFs following stimulation with IL‐17 (C) or IL‐17F (D) (100 ng/ml) (n = 3 technical replicates).
• E, F
  Immunoblot analysis of NEDD4L in Nedd4l ^+/+ and Nedd4l ^−/− primary lung epithelial cells (E), and in Nedd4l ^+/+ and Nedd4l ^+/− primary astrocytes (F).

Data information: Data are shown as the mean ± S.E.M. of technical replicates. All experiments were performed two times, and representative data are shown.

Source data are available online for this figure.

Figure 1. NEDD4L negatively regulates IL‐17‐induced proinflammatory cytokine expression.

• A, B
  HeLa cells were transfected with Nedd4l siRNA (siNedd4l) or control RNA (siNC). Il‐6, Cxcl2, and Cccl20 mRNA expression (A) and ELISA assay of IL‐6 and CXCL2 production (B) in HeLa cells transfected with Nedd4l siRNA (siNedd4l) or control RNA (siNC) and stimulated with IL‐17 (50 ng/ml) for the indicated time (n = 3 technical replicates).
• C, D
  Real‐time PCR analysis of Il‐6, Cxcl2, and Cccl20 mRNA expression (C) and ELISA assay of IL‐6 and CXCL2 production (D) in wild‐type (Nedd4l ^+/+) and Nedd4l‐deficient homozygous (Nedd4l ^−/−) MEFs following stimulation with IL‐17 (100 ng/ml) for the indicated time (n = 3 technical replicates).
• E
  Real‐time PCR analysis of Il‐6, Cxcl2, and Ccl20 mRNA expression in Nedd4l ^+/+ and Nedd4l ^−/− MEFs following stimulation with IL‐17F (100 ng/ml) (n = 3 technical replicates).
• F, G
  Real‐time PCR analysis of Il‐6, Cxcl2, and Cccl20 mRNA in Nedd4l ^+/+ and Nedd4l ^−/− primary lung epithelial cells (F), or in Nedd4l ^+/+ and Nedd4l ^+/− primary astrocytes (G) untreated or stimulated with IL‐17 (100 ng/ml) for the indicated time (n = 3 technical replicates).

Data information: Data are shown as the mean ± S.E.M. of technical replicates. Experiments were performed three times (A‐E) or two times (F, G), and representative data are shown. Significant differences were tested using a two‐tailed Student's t‐test. *P < 0.05, **P < 0.01, ***P < 0. 001. NS, no significance.

Interestingly, increased TNF‐α‐ or IL‐1α‐induced Il‐6, Cxcl2, and Ccl20 mRNA expression was also observed in HeLa cells with Nedd4l knockdown and in MEFs with Nedd4l deficiency (Fig EV2A–D). These data suggest that NEDD4L might be a common negative regulator of TNF‐α‐, IL‐1α‐, and IL‐17‐induced inflammation.

Figure EV2. Nedd4l deficiency promotes TNF‐α‐ and IL‐1α‐induced inflammatory cytokine and chemokine expression.

• A, B
  HeLa cells were transfected with Nedd4l siRNA (siNedd4l) or control RNA (siNC). Il‐6, Cxcl2, and Cccl20 mRNA expression in HeLa cells transfected with Nedd4l siRNA (siNedd4l) or control RNA (siNC) and stimulated with TNF‐α (50 ng/ml) (A) and IL‐1α (10 ng/ml) (B) for the indicated time (n = 3 technical replicates).
• C, D
  Real‐time PCR analysis of Il‐6, Cxcl2, and Ccl20 mRNA expression in Nedd4l ^+/+‐ and Nedd4l‐deficient heterozygous (Nedd4l ^+/−) MEFs following stimulation with TNF‐α (50 ng/ml) (C) and IL‐1α (10 ng/ml) (D) for the indicated time (n = 3 technical replicates).
• E
  Immunoblot analysis of NEDD4L in peritoneal mesothelial cells isolated from Nedd4l ^+/+ and Nedd4l ^+/− mice.

Data information: Data are shown as the mean ± S.E.M. of technical replicates. All experiments were performed two times, and representative data are shown.

Source data are available online for this figure.

NEDD4L inhibits IL‐17‐induced inflammation in vivo

To explore the regulatory role of NEDD4L in IL‐17‐induced inflammation responses in vivo, Nedd4l ^+/+ and Nedd4l ^+/− mice were intraperitoneally injected (i.p.) with IL‐17 or PBS, and Il‐6, Cxcl2, and Ccl20 mRNA expression in peritoneal mesothelial cells were analyzed by real‐time PCR. As shown in Figs 2A and EV2E, Nedd4l deficiency significantly increased IL‐17‐induced Il‐6, Cxcl2, and Ccl20 mRNA expression in peritoneal mesothelial cells. IL‐17‐induced chemokines amplify inflammatory responses by recruiting neutrophils to local tissue, as shown in an inflammation model in which airway administration of IL‐17 caused considerable pulmonary inflammation (Bulek et al, 2011; Zhong et al, 2012). To determine whether NEDD4L affects IL‐17‐induced pulmonary inflammation, Nedd4l ^+/+ and Nedd4l ^+/− mice were intratracheally injected with IL‐17 or PBS, followed by the analysis of Il‐6, Cxcl2, and Ccl20 mRNA expression in lung tissues and production of CXCL2 in the bronchoalveolar lavage fluid (BALF). As shown in Fig 2B, Nedd4l deficiency significantly increased IL‐17‐induced Il‐6, Cxcl2, and Ccl20 mRNA expression in lung tissues. CXCL2 production in BALF from Nedd4l ^+/− mice was also significantly increased compared to that from Nedd4l ^+/+ mice (Fig 2C). Consistent with increased proinflammatory cytokine and chemokine production in Nedd4l ^+/− mice, more CD45⁺ cells were recruited into the BALF of Nedd4l ^+/− mice than that of Nedd4l ^+/+ mice (Fig 2D). In particular, neutrophils (Gr1⁺CD11b⁺) in the BALF of Nedd4l ^+/− mice were increased more than 10 folds compared with that in wild‐type mice (Fig 2E), indicating that Nedd4l deficiency increased IL‐17‐induced pulmonary neutrophil infiltration. Histological analysis of lung tissue showed enhanced immunopathological changes in the lungs of Nedd4l ^+/− mice (Fig 2F). These results suggest that NEDD4L plays an important role in limiting IL‐17‐induced inflammation in vivo.

Figure 2. NEDD4L restricts IL‐17‐induced inflammation.

• A
  Real‐time PCR analysis of Il‐6, Cxcl2, and Cccl20 mRNA expression in peritoneal mesothelial cells isolated from Nedd4l ^+/+ and Nedd4l ^+/− mice treated by intraperitoneal injection of PBS (n = 4 biological replicates) or IL‐17 (n = 6 biological replicates) (0.5 μg in 200 μl PBS).
• B
  Real‐time PCR analysis of Il‐6, Cxcl2, and Cccl20 mRNA expression in lung tissue of Nedd4l ^+/+ and Nedd4l ^+/− mice (n = 3 biological replicates) treated by intratracheal injection of PBS or IL‐17 (2 μg in 30 μl PBS).
• C
  ELISA assay of CXCL2 production in BALF from mice treated as in (B) (n = 3 biological replicates).
• D, E
  Infiltration of total CD45⁺ cells (D) and Gr‐1⁺CD11b⁺ cells (E) in BALF from mice treated as in (B) (n = 3 biological replicates).
• F
  Hematoxylin and eosin staining of lung tissue from mice treated as in (B). Scale bars, 100 μm.

Data information: Data are shown as the mean ± S.E.M. of biological replicates. All experiments were performed two times, and representative data are shown. Significant differences were tested using a two‐tailed Student's t‐test. *P < 0.05, **P < 0.01, and ***P < 0. 001. NS, no significance.

NEDD4L inhibits IL‐17R signaling and MEKK2 expression

IL‐17 induces proinflammatory cytokine and chemokine production through MAPK and NF‐κB pathways. In HeLa cells, Nedd4l knockdown increased IL‐17‐induced phosphorylation of p38 and NF‐κB p65 subunit without affecting IL‐17‐induced phosphorylation of ERK1/2 and JNK1/2 (Fig 3A). In MEFs, Nedd4l deficiency increased IL‐17‐induced p38 and NF‐κB p65 subunit phosphorylation but not ERK1/2 and JNK1/2 phosphorylation (Fig 3B). Accordingly, Nedd4l knockdown in HeLa cells or deficiency in MEFs remarkably increased TNF‐α‐ or IL‐1α‐induced p38 and NF‐κB p65 phosphorylation (Fig EV3A–D). Consistently, Nedd4l deficiency in primary lung epithelial cells and astrocytes significantly promoted IL‐17‐induced p38 and NF‐κB p65 subunit phosphorylation (Fig 3C and D). Overexpression of exogenous wild‐type NEDD4L, but not the E3 enzyme activity mutation NEDD4L‐C942A (NEDD4L‐CA) in HeLa cells, inhibited IL‐17‐induced p38 and NF‐κB p65 subunit phosphorylation compared to the control group (Fig 3E). Similar signal transduction effect mediated by overexpression of exogenous NEDD4L could be reconstituted in Nedd4l‐silenced HeLa cells, suggesting that NEDD4L specifically regulates this signaling (Fig EV3E). These results indicate that NEDD4L plays a negative regulatory role in IL‐17‐induced signaling.

Figure 3. NEDD4L inhibits IL‐17R signaling and MEKK2 expression.

• A–D
  Immunoblot analysis of phosphorylated ERK1/2 (P‐ERK1/2), phosphorylated p38 (P‐P38), phosphorylated JNK1/2 (P‐JNK1/2), phosphorylated p65 subunit of NF‐κB (P‐P65), total P65, MEKK2, TRAF2, TRAF3, TRAF6, NEDD4L, and actin in HeLa cells transfected with Nedd4l siRNA (siNedd4l) or control RNA (siNC) (A), Nedd4l ^+/+ and Nedd4l ^−/− MEFs (B), primary lung epithelial cells (C), or Nedd4l ^+/+ and Nedd4l ^+/− astrocytes (D), treated with IL‐17 for indicated time.
• E
  Immunoblot analysis of phosphorylated p65 subunit of NF‐κB (P‐P65), total P65, phosphorylated p38 (P‐P38), total P38, MEKK2, NEDD4L, and actin in HeLa cells transfected with Myc‐tagged NEDD4L, Myc‐tagged NEDD4L‐CA(C942A), or Myc‐tagged Null plasmids, treated with IL‐17 for the indicated time.
• F
  Immunoblot analysis of MEKK2, TRAF2, TRAF6, NEDD4L, and actin in Nedd4l ^+/+, Nedd4l ^+/−, and Nedd4l ^−/− MEFs.
• G
  Immunoblot analysis of MEKK2 in lysates of Nedd4l ^+/+ and Nedd4l ^−/− MEFs treated with CHX (50 μg/ ml) for the indicated time.
• H
  The intensity values were measured by Image J software to calculate the half‐life of MEKK protein (n = 3 biological replicates).

Data information: Numbers between two blots indicate the densitometry of phosphorylated proteins relative to that of total proteins, respectively. Densitometry of MEKK2 was relative to actin. All experiments were performed three times, and representative data are shown. Significant differences were tested using a two‐tailed Student's t‐test. *P < 0.05 and **P < 0.01. NS, no significance.

Source data are available online for this figure.

Figure EV3. Nedd4l deficiency promotes the IL‐17‐, TNF‐α‐, and IL‐1α‐induced P38, NF‐κB activation.

• A–D
  Immunoblot analysis of phosphorylated p38 (P‐P38), total P38, phosphorylated p65 subunit of NF‐κB (P‐P65), total p65, MEKK2, NEDD4L, and actin in HeLa cells, or Nedd4l ^+/+ and Nedd4l ^−/− MEFs treated with 50 ng/ml TNF‐α (A, C) or 10 ng/ml IL‐1α (B, D) for the indicated time.
• E
  Immunoblot analysis of phosphorylated NF‐κB p65 (P‐P65), total P65, MEKK2, NEDD4L, and actin in HeLa cells transfected with control RNA (siNC) and Nedd4l siRNA (siNedd4l) followed by overexpression of Myc‐tagged Null, Myc‐tagged NEDD4L, or Myc‐tagged NEDD4L‐C942A (Myc‐tagged NEDD4L‐CA), and stimulated with IL‐17 (50 ng/ml) for the indicated time.
• F
  Immunoblot analysis of HA‐tagged MEKK2, Myc‐tagged NEDD4L, and actin in HEK293T cells transfected with the indicated plasmids with anti‐HA, anti‐Myc‐tag, or anti‐actin antibody.

Data information: Numbers between two blots indicate the densitometry of phosphorylated proteins relative to that of total proteins, respectively. All experiments were performed two times, and representative data are shown.

Source data are available online for this figure.

IL‐17RA and IL‐17RC activate MAPKs and NF‐κB pathways through TRAF family members. However, neither Nedd4l knockdown nor Nedd4l deficiency affected expressions of the TRAFs (Fig 3A and B). Given the important role of MEKK family members in MAPK and NF‐κB activation (Zhao & Lee, 1999; Blonska et al, 2004), we set out to investigate the effect of Nedd4l knockdown on MEKKs. Interestingly, Nedd4l knockdown significantly increased MEKK2 expression in HeLa cells (Fig 3A). In MEFs, primary lung epithelial cells, and astrocytes, Nedd4l deficiency also increased MEKK2 expression (Fig 3B–D). Overexpression or reconstitution of exogenous wild‐type NEDD4L, but not NEDD4L‐CA, can inhibit MEKK2 expression compared to the control group (Figs 3E and EV3E). We compared MEKK2 expression in Nedd4l ^+/−, Nedd4l ^−/−, and wild‐type MEFs. As shown in Fig 3F, Nedd4l heterozygous deficiency was sufficient to increase MEKK2 expression in MEFs. Notably, Nedd4l deficiency in MEFs strongly prolonged the half‐life of MEKK2 protein compared to that in wild‐type MEFs (Fig 3G and H). Furthermore, overexpression of NEDD4L reduced MEKK2 expression (Fig EV3F), providing further evidence that NEDD4L inhibits the stability of MEKK2 protein.

NEDD4L mediates MEKK2 ubiquitination

To determine the mechanism through which NEDD4L reduced MEKK2 expression, we investigated the interaction between NEDD4L and MEKK2 in HeLa cells by immunoprecipitation with NEDD4L‐ or MEKK2‐specific antibody. As shown in Figs 4A and EV4A, NEDD4L was co‐immunoprecipitated with MEKK2, but not MEKK3, in both IL‐17‐stimulated and unstimulated cells, suggesting that NEDD4L was constitutively associated with MEKK2 in HeLa cells. In the pull‐down assay, NEDD4L recombinant protein was precipitated with GST‐MEKK2 protein (Fig 4B), suggesting that NEDD4L bound MEKK2 in vitro. To map the domains required for NEDD4L to interact with MEKK2, we constructed a series of plasmids expressing wild‐type and mutant NEDD4L in which C2 (∆C2), WW (∆WW), or HECT (∆HECT) domain was deleted, respectively (Fig EV4B). As shown in Fig 4C, the deletion of the WW domain but not the C2 and HECT domain disrupted the interaction between NEDD4L and MEKK2, demonstrating that the WW domain was necessary for NEDD4L to bind to MEKK2. As an E3 ubiquitin ligase, NEDD4L might inhibit the stability of MEKK2 protein by mediating the ubiquitination of MEKK2. To investigate whether NEDD4L regulates MEKK2 expression through proteasome‐dependent or lysosomal‐dependent pathway, we treated Nedd4l ^−/− and wild‐type control MEFs with a proteasome inhibitor MG132 or a lysosomal degradation inhibitor bafilomycin A1 (Baf A1). Following MG132 treatment but not Baf A1 treatment, MEKK2 expression in wild‐type control cells was elevated to the level comparable with that in Nedd4l ^−/− MEFs (Fig 4D), suggesting that NEDD4L suppresses the stability of MEKK2 protein by mediating MEKK2 ubiquitination in a proteasome‐dependent manner. In further support of our hypothesis, in vitro ligase activity assay demonstrated that it was NEDD4L protein, but not TRAF6 protein, efficiently increased MEKK2 ubiquitination (Fig 4E), suggesting that MEKK2 is a potential substrate of NEDD4L. Furthermore, NEDD4L deficiency reduced total poly‐ubiquitination of MEKK2, without affecting K63‐ and K48‐linked poly‐ubiquitination of MEKK2 in MEFs (Fig 4F). We further determined the effects of wild‐type versus mutant NEDD4L on MEKK2 ubiquitination in HEK293T cells. NEDD4L‐∆WW completely lost the capability to promote MEKK2 ubiquitination. Furthermore, deletion of the HECT domain also largely impaired such capability of NEDD4L. In contrast, deletion of the C2 domain enhanced the ubiquitination of MEKK2, which could be due to the relief of autoinhibitory interaction between its C2 and HECT domains, consistent with the notion that the C2 domain of HECT family ligases, including smurf2, NEDD4, and NEDD4L, inhibits the function of HECT domain (Wang et al, 2010; Malonis et al, 2017; Lorenz, 2018; Fig 4G). To determine the nature of NEDD4L‐mediated MEKK2 ubiquitination, we transfected HEK293T cells with NEDD4L‐ and MEKK2‐expressing vectors in the presence of constructs expressing wild‐type ubiquitin (HA‐Ub) or its mutants. As shown in Fig 4H and I, NEDD4L mainly promoted Lys‐27(K27O)‐linked poly‐ubiquitination of MEKK2 (Fig 4H). Mutation of Lys‐27 (K27R) completely abrogated the poly‐ubiquitination effect of NEDD4L on MEKK2 (Fig 4I), demonstrating that NEDD4L mediates K27‐linked poly‐ubiquitination of MEKK2. It is well established that the ubiquitin chain specificity of the HECT‐type E3 ligases is determined by their C‐terminal amino acids. NEDD4 family ligases, including NEDD4L, exhibit strict specificity toward K63 linkages (Maspero et al, 2013). However, in our study, NEDD4L mediates K27‐linked but not K63‐linked poly‐ubiquitination of MEKK2 in a dosage‐dependent manner (Fig 4J), which could be a consequence of the unique structure of NEDD4L. Further studies will be needed to identify the detailed differences between NEDD4L and other HECT‐type E3 ligases.

Figure 4. NEDD4L ubiquitinates MEKK2 and inhibits the protein stability of MEKK2.

• A
  Immunoblot analysis of NEDD4L, MEKK2, and MEKK3 immunoprecipitated with NEDD4L‐specific antibody from lysates of HeLa cells pre‐treated with 20 μM MG‐132 for 6 h and then treated with IL‐17 for the indicated time.
• B
  Immunoblot analysis of NEDD4L precipitated with GST‐tagged Null or GST‐tagged MEKK2 recombinant protein.
• C
  Immunoblot of HA‐tagged MEKK2 co‐immunoprecipitated with anti‐Myc antibody from lysates of HEK293T cells co‐transfected with plasmids expressing HA‐tagged MEKK2 and Myc‐tagged wild‐type NEDD4L or mutant NEDD4L with C2 (∆C2), WW (∆WW), or HECT (∆HECT) domain deleted.
• D
  Immunoblot analysis of MEKK2 and NEDD4L in Nedd4l ^+/+, and Neddl4 ^−/− treated with DMSO, 20 μM MG‐132, or 0.2 μM bafilomycin A1 (Baf A1) for 6 h. Densitometry of MEKK2 was relative to actin.
• E
  Immunoblot analysis of MEKK2 incubated with NEDD4L or TRAF6 recombinant protein and the full complement of ubiquitination reaction components in vitro.
• F
  Immunoblot analysis of K48‐, K63‐linked, and total ubiquitination of MEKK2 following precipitation of MEKK2 with MEKK2‐specific antibody from lysates of Nedd4l ^+/+ and Nedd4l ^−/− MEFs treated with MG‐132.
• G
  Immunoblot analysis of total ubiquitination of Flag‐tagged MEKK2 following precipitation of MEKK2 with Flag‐tagged specific antibody from lysates of HEK293T cells co‐transfected with plasmids expressing Flag‐tagged MEKK2 and Myc‐tagged wild‐type NEDD4L or mutant NEDD4L with C2 (∆C2), WW (∆WW), or HECT (∆HECT) domain deleted.
• H
  Immunoblot analysis of total ubiquitination of Flag‐tagged MEKK2 following immunoprecipitating MEKK2 with anti‐Flag antibody from lysates of HEK293T cells co‐transfected with plasmids expressing Flag‐tagged MEKK2, Myc‐tagged NEDD4L, and HA‐tagged wild‐type or K6O, K11O, K27O, K29O, K33O, K48O, and K63O mutant Ub.
• I
  Immunoblot analysis of total ubiquitination of Flag‐tagged MEKK2 following immunoprecipitating MEKK2 with anti‐Flag antibody from lysates of HEK293T cells co‐transfected with plasmids expressing Flag‐tagged MEKK2, Myc‐tagged NEDD4L, and HA‐tagged wild‐type or K6R, K11R, K27R, K29R, K33R, K48R, K63R mutant Ub.
• J
  Immunoblot analysis of total ubiquitination of Flag‐tagged MEKK2 following immunoprecipitating MEKK2 with anti‐Flag antibody from lysates of HEK293T cells co‐transfected with plasmids expressing Flag‐tagged MEKK2, increased dosage of Myc‐tagged NEDD4L, and HA‐tagged wild‐type, K27O, or K63O mutant Ub.

Data information: All experiments were performed three times, and representative data are shown.

Source data are available online for this figure.

Figure EV4. Nedd4l deficiency promotes IL‐17‐induced inflammatory cytokine and chemokine expression dependent on MEKK2.

• A
  Immunoblot analysis of NEDD4L and MEKK2 immunoprecipitated with anti‐MEKK2 antibody from lysates of HeLa cells treated with 20 μM MG‐132 for 6 h and then 50 ng/ml IL‐17 for the indicated time.
• B
  Schematic diagram of wild‐type and mutant NEDD4L constructs.
• C
  Immunoblot analysis of phosphorylated NF‐κB p65 (P‐P65), total P65, MEKK2, NEDD4L, Smurf1, and actin transfected with control RNA (siNC), Nedd4l siRNA (siNedd4l), Smurf1 siRNA (siSmurf1) or Nedd4l siRNA plus Smurf1 siRNA (siNedd4l + siSmurf1), and stimulated with IL‐17 (50 ng/ml) for the indicated time.
• D
  Real‐time PCR analysis of Il‐6, Cxcl2, and Ccl20 mRNA expression in HeLa cells treated as in (C) (n = 3 technical replicates).
• E
  Real‐time PCR analysis of Il‐6, Cxcl2, and Ccl20 mRNA expression in HeLa cells transfected with control RNA (siNC) or Nedd4l siRNA (siNedd4l), followed by being pre‐treated with BAY 61–3,606 (10 μM) and stimulated with IL‐17 (50 ng/ml) for the indicated time (n = 3 technical replicates).
• F
  Immunoblot analysis of Ser520 phosphorylation of MEKK2 (P‐MEKK2), total Flag‐tagged MEKK2, and actin in HEK293T cells transfected with Flag‐tagged Null, Flag‐tagged MEKK2, and mutant MEKK2.
• G
  Immunoblot analysis of Ser520 phosphorylation of MEKK2 (P‐MEKK2) immunoprecipitated from HEK293T cells transfected with Flag‐tagged Null, or Flag‐tagged MEKK2, and then incubated with or without calf intestinal alkaline phosphatase time (CIAP) for 30 min.
• H
  Immunoblot analysis of phosphorylated NF‐κB p65 (P‐P65), total P65, MEKK2, and actin in HeLa cells transfected with control RNA (siNC), Mekk2 siRNA (siMekk2) followed by overexpression of Flag‐tagged Null, Flag‐tagged MEKK2, or Flag‐tagged MEKK2‐S520A (Flag‐tagged MEKK2‐SA), and stimulated with IL‐17 (50 ng/ml) for the indicated time.
• I
  Schematic diagram of wild‐type and mutant MEKK2 constructs.
• J
  Immunoblot analysis of Flag‐tagged and mutant MEKK2 co‐immunoprecipitated with anti‐Flag antibody from lysates of HEK293T cells co‐transfected with plasmids expressing HA‐tagged TAK1 and Flag‐tagged wild‐type or mutant MEKK2. Asterisk (*) indicates the target band.

Data information: Data are shown as the mean ± S.E.M. of technical replicates. All experiments were performed two times, and representative data are shown.

Source data are available online for this figure.

MEKK2 is required for IL‐17R signaling

MEKK2 has been reported to participate in TNF‐α‐ and IL‐1‐induced signaling (Chayama et al, 2001; Hammaker et al, 2004). However, it remains unknown whether MEKK2 is also involved in IL‐17R signaling. Thus, our further experiments were focused on the role of MEKK2 in NEDD4L‐mediated negative regulation of IL‐17R signaling. We firstly used synthesized Mekk2‐specific small‐interfering RNA (siMekk2) to inhibit endogenous MEKK2 expression in HeLa cells, and then stimulated the cells with IL‐17 and detected IL‐17‐induced MAPKs and NF‐κB activation. As shown in Fig 5A, Mekk2 siRNA inhibited endogenous MEKK2 expression and IL‐17‐induced p38 and NF‐κB activation. Consistently, Mekk2 siRNA inhibited IL‐17‐induced Il‐6, Cxcl2, and Ccl20 mRNA expression in HeLa cells (Fig 5C). Similarly, Mekk2 knockdown also inhibited IL‐17‐induced p38 and NF‐κB activation and Il‐6, Cxcl2, and Ccl20 mRNA expression in MEFs (Fig 5B–D). These results suggest that MEKK2 is required for IL‐17R signaling.

Figure 5. NEDD4L inhibits IL‐17R signaling by targeting MEKK2.

• A, B
  Immunoblot analysis of phosphorylated p38 (P‐P38), phosphorylated NF‐κB p65 (P‐P65), MEKK2, and actin in HeLa cells (A), and MEFs (B) transfected with Mekk2 siRNA (siMekk2) or control RNA (siNC) and stimulated with IL‐17 (HeLa cells 50 ng/ml, MEFs 100 ng/ml) for the indicated time.
• C, D
  Real‐time PCR analysis of Il‐6, Cxcl2, and Ccl20 mRNA expression in HeLa cells (C) and MEFs (D) transfected with siRNA (siMekk2) or control RNA (siNC) and stimulated with IL‐17 (HeLa cells 50 ng/ml, MEFs 100 ng/ml) for the indicated time (n = 3 technical replicates).
• E
  Immunoblot analysis of phosphorylated p38, phosphorylated NF‐κB p65 (P‐P65), NEDD4L, MEKK2, and actin in HeLa cells transfected with control RNA (siNC), Nedd4l siRNA (siNedd4l), Mekk2 siRNA(siMekk2), or Nedd4l siRNA plus Mekk2 siRNA (siNedd4l + siMekk2), and stimulated with IL‐17 (50 ng/ml) for the indicated time.
• F
  Immunoblot analysis of phosphorylated p38, phosphorylated NF‐κB p65 (p‐p65), NEDD4L, MEKK2, and actin in Nedd4l ^+/+ and Nedd4l ^−/− MEFs transfected with Mekk2 siRNA (siMekk2) or control RNA (siNc), and stimulated with IL‐17 (100 ng/ml) for the indicated time.
• G, H
  Real‐time PCR analysis of Il‐6, Cxcl2, and Ccl20 mRNA expression in HeLa cells (G) and MEFs (H) treated as described in (E) and (F) (n = 3 technical replicates).

Data information: Numbers between two blots indicate the densitometry of phosphorylated proteins relative to that of total proteins, respectively. Data are shown as the mean ± S.E.M. of technical replicates. All experiments were performed two times, and representative data are shown.

Source data are available online for this figure.

To investigate the role of MEKK2 in NEDD4L‐mediated regulation of IL‐17R signaling, we compared the effects of Nedd4l siRNA on IL‐17‐induced p38 and NF‐κB activation in the presence and absence of Mekk2 siRNA. As shown in Fig 5E, transfection of Nedd4l siRNA alone enhanced IL‐17‐induced p38 and NF‐κB activation in HeLa cells. However, in the presence of Mekk2 siRNA, Nedd4l siRNA could not enhance IL‐17‐induced p38 and NF‐κB activation (Fig 5E). Similarly, Mekk2 siRNA eliminated the enhancement of IL‐17‐induced p38 and NF‐κB activation by Nedd4l deficiency in MEFs (Fig 5F). Mekk2 knockdown nearly completely reversed the increase in IL‐17‐induced Il‐6, Cxcl2, and Ccl20 mRNA expression by Nedd4l knockdown in HeLa cells or Nedd4l deficiency in MEFs (Fig 5G and H). It has been identified that Smurf1, another HECT family E3, promotes MEKK2 ubiquitination and degradation in macrophages under the TLR9 activation (Wen et al, 2015). However, no combined effects were observed in Smurf1 and Nedd4l combined silenced HeLa cells (Fig EV4C and D). Syk is known to be a target for NEDD4L in mast cells (Yip et al, 2016). Syk acts as an upstream signaling molecule in IL‐17‐induced Act1‐TRAF6 interaction in keratinocytes, and inhibition of Syk can attenuate CCL20 production (Wu et al, 2015). We, therefore, checked the potential influence of Syk on NEDD4L‐mediated IL‐17 activation by treating HeLa cells with a Syk‐specific inhibitor, BAY 61–3,606. As shown in Fig EV4E, specific inhibition of Syk attenuated IL‐17‐induced Il‐6, Cxcl2, and Ccl20 mRNA expression, but could not eliminate the enhancement of IL‐17‐induced Il‐6, Cxcl2, and Ccl20 mRNA expression by Nedd4l silencing in HeLa cells. These results demonstrate that NEDD4L negatively regulates IL‐17R signaling via MEKK2.

MEKK2 Ser520 phosphorylation is required for IL‐17R signaling and its degradation

MEKK2 Ser520 was reported to be phosphorylated following LPS stimulation and important for MEKK2 activation (Zhang et al, 2006). We prepared a phosphorylated MEKK2 Ser520‐specific antibody to detect MEKK2 phosphorylation in IL‐17‐treated HeLa cells. As shown in Fig 6A, IL‐17 stimulation induced remarkable MEKK2 Ser520 phosphorylation. The antibody against MEKK Ser520 was validated by western blotting analysis of wild‐type MEKK2 and its Ser520 mutants, including MEKK2 Ser520 residue mutating into continuously activated forms asparagine (S520D) and glutamic acid (S520E), or inactivated form alanine (S520A; Figs EV4F and EV4G). As the Ser520 phosphorylation in MEKK2 is regulated by IL‐17R signaling, we, therefore, tested if this phosphorylation is crucial for IL‐17‐induced proinflammatory cytokine and chemokine expression by mutating MEKK2 Ser520 residue into alanine (S520A). As shown in Fig 6B, MEKK2 S520A significantly inhibited the IL‐17‐induced proinflammatory cytokine and chemokine expression in HeLa cells, as well as the phosphorylation of p38 and NF‐κB p65 (Fig 6C). Reconstitution of exogenous wild‐type MEKK2, but not MEKK2‐S520A, can restore the phosphorylation of p38 and NF‐κB p65 in Mekk2‐silenced HeLa cells compared to the control group (Fig EV4H).

Figure 6. IL‐17‐induced MEKK2 Ser520 phosphorylation via TAK1 and promotes NEDD4L‐mediated MEKK2 ubiquitination.

• A
  Immunoblot analysis of Ser520 phosphorylation of MEKK2 immunoprecipitated from HeLa cells stimulated with IL‐17 (50 ng/ml) for the indicated time, and then incubated with or without calf intestinal alkaline phosphatase time (CIAP) for 30 min.
• B
  Real‐time PCR analysis of Il‐6, Cxcl2, and Ccl20 mRNA expression in HeLa cells transfected with HA‐tagged MEKK2 S520A or control vector, and stimulated with IL‐17 (50 ng/ml) for the indicated time (n = 3 technical replicates).
• C
  Immunoblot analysis of phosphorylated p38, phosphorylated NF‐κB p65 (P‐P65), HA, and actin in HeLa cells treated as in (B). Numbers between two blots indicate the densitometry of phosphorylated proteins relative to that of total proteins, respectively.
• D
  Immunoblot analysis of MEKK2 and TAK1 co‐immunoprecipitated with anti‐TAK1 antibody from lysates of HeLa cells treated with 50 ng/ml IL‐17 for the indicated time.
• E
  Immunoblot analysis of Ser520 phosphorylation of MEKK2 immunoprecipitated from HeLa cells transfected with Tak1‐specific (siTak1) or control RNA (siNC), and then stimulated with IL‐17 (50 ng/ml) for the indicated time.
• F
  Immunoblot analysis of Myc‐tagged NEDD4L co‐immunoprecipitated with anti‐HA antibody from lysates of HEK293T cells co‐transfected with plasmids expressing Myc‐tagged NEDD4L and HA‐tagged wild‐type MEKK2 or mutant MEKK2 (S520A).
• G
  Immunoblot analysis of total ubiquitination of HA‐tagged MEKK2 and HA‐tagged MEKK2 (S520A) following precipitation of MEKK2 with anti‐HA antibody from lysates of HEK293T cells co‐transfected with plasmids expressing HA‐tagged MEKK2 or HA‐tagged MEKK2 (S520A) together with Myc‐tagged NEDD4L and Flag‐tagged Ub.
• H
  Immunoblot analysis of HA‐tagged MEKK2, HA‐tagged MEKK2 (S520A), Myc‐tagged NEDD4L, and actin in HEK293T cells transfected with the indicated plasmids with anti‐HA, anti‐Myc, or anti‐actin antibody.

Data information: Data are shown as the mean ± S.E.M. of technical replicates. All experiments were performed two times, and representative data are shown.

Source data are available online for this figure.

We next explored how MEKK2 was activated in IL‐17R signaling. IL‐17 activates MAPK and NF‐κB through the Act1‐TRAF6‐TAK1 cascade. We examined whether MEKK2 interacts with these molecules by immunoprecipitation. TAK1 was found to associate with MEKK2 following IL‐17 stimulation (Fig 6D). We constructed plasmids expressing truncated MEKK2 N351 (a.a.1–351) and MEKK2 C352 (a.a.352–619; Fig EV4I). As shown in Fig EV4J, wild‐type MEKK2, mutant MEKK2 S520A, and MEKK2 C352 could be co‐immunoprecipitated with TAK1. In contrast, MEKK2 N351 could not be co‐immunoprecipitated with TAK1, demonstrating that TAK1 interacted with the C‐terminal kinase domain of MEKK2. We used Tak1‐specific siRNA to inhibit endogenous TAK1 expression (Fig 6E). Tak1 knockdown nearly completely blocked IL‐17‐induced MEKK2 phosphorylation (Fig 6E). These results suggest that MEKK2 interacts with TAK1 and is activated by TAK1 in IL‐17R signaling.

We further investigated the effects of MEKK2 Ser520 phosphorylation on NEDD4L‐mediated MEKK2 degradation. As shown in Fig 6F, NEDD4L was co‐immunoprecipitated with MEKK2 (S520A) as efficiently as with wild‐type MEKK2, indicating that MEKK2 Ser520 phosphorylation was not required for the association between NEDD4L and MEKK2. However, NEDD4L failed to promote MEKK2 (S520A) ubiquitination and degradation (Fig 6G and H), demonstrating that MEKK2 Ser520 phosphorylation is indispensable for NEDD4L‐mediated MEKK2 ubiquitination. Taken together, IL‐17 induces MEKK2 S520 phosphorylation, which is required for downstream p38 and NF‐κB pathway activation. NEDD4L negatively regulates IL‐17R signaling via mediating Ser520‐phosphorylated MEKK2 degradation.

Nedd4l deficiency promotes the pathogenesis of EAE via IL‐17R‐mediated inflammation in mice

It has been reported that IL‐17R signaling in astrocytes plays a crucial role in mouse EAE, a well‐known mouse multiple sclerosis (MS) model (Kang et al, 2010). We then investigated the effect of NEDD4L heterozygous deficiency on the MOG_35‐55‐induced EAE model by subcutaneous immunization of Nedd4l ^+/+ or Nedd4l ^+/− mice with MOG_35‐55 and then intravenously injecting them with pertussis toxin (PTX). As shown in Fig EV5A, Nedd4l ^+/− mice suffered from more severe EAE disease symptoms compared to Nedd4l ^+/+ mice and had higher EAE clinical scores than their counterparts. Moreover, Nedd4l ^+/− mice displayed more inflammatory cell infiltration and more serious demyelination compared to their counterparts (Fig EV5B), as demonstrated by hematoxylin–eosin (H&E) and Luxol fast blue (LFB) staining, respectively. Consistent with these results, flow cytometry analysis of mouse CNS tissues (brains and spinal cords) demonstrated that CNS‐infiltrating CD11b^highGr1⁺ neutrophils were significantly increased in Nedd4l ^+/− mice compared to Nedd4l ^+/+ mice (Fig EV5C–F). We subsequently examined the expression levels of several proinflammatory cytokines and chemokines in the CNS tissues of Nedd4l ^+/+ and Nedd4l ^+/− EAE mice. As shown in Fig EV5G, the gene expression levels of Il‐6, Cxcl2, and Ccl20 were significantly increased in Nedd4l ^+/− EAE mice. These data demonstrate that Nedd4l deficiency promotes the pathogenesis of EAE.

Figure EV5. Nedd4l deficiency promotes MOG_35‐55‐induced EAE.

• A
  The mean clinical scores for Nedd4l ^+/+ and Nedd4l ^+/− mice (Nedd4l ^+/+ n = 5 biological replicates and Nedd4l ^+/− n = 5 biological replicates) induced by being immunized with MOG_35‐55 were assessed according to the standards described in the Materials and Methods from day 0 to day 18.
• B
  Histology of the spinal cord was analyzed by H&E (left panel) or LFB (right panel) staining on day 18 after the second MOG_35‐55 immunization. Scale bars (a whole spinal cord, 100×), 500 μm; scale bars (a portion of the spinal cord, 200X), 200 μm.
• C, D
  Flow cytometric analysis of immune cells (including CD45⁺CD4⁺ T cells, CD45⁺CD8⁺ T cells, CD45⁺CD19⁺ B cells, and CD45⁺CD11b⁺ myeloid cells), infiltrated to the spinal cord and brain of EAE–induced Nedd4l ^+/+ and Nedd4l ^+/− mice at day 18 after immunization (Nedd4l ^+/+ n = 5 biological replicates and Nedd4l ^+/− n = 5 biological replicates). Data are presented as representative plots.
• E
  Summary graph of the percentages of immune cells in the CNS, including CD45⁺CD4⁺ T cells, CD45⁺CD8⁺ T cells, CD45⁺CD19⁺ B cells, CD45⁺CD11b^high, and CD45⁺CD11b^low myeloid cells. CNS‐infiltrating cells isolated from mice treated as in (A) were stained with the appropriate antibodies (Nedd4l ^+/+ n = 5 biological replicates and Nedd4l ^+/− n = 5 biological replicates).
• F
  Summary graph of the percentages of CD11b⁺Gr‐1⁺ neutrophils in the CD45⁺CD11b^high CNS. CNS‐infiltrating cells isolated from mice treated as in (A) were stained with the appropriate antibodies (Nedd4l ^+/+ n = 5 biological replicates and Nedd4l ^+/− n = 5 biological replicates).
• G
  Il‐6, Cxcl2, Cccl20 mRNA, IL‐17A, and IFNγ mRNA in the CNS tissues (including brain and spinal cords) were measured by real‐time PCR on day 14 after the second MOG_35‐55 immunization (Nedd4l ^+/+ n = 5 biological replicates and Nedd4l ^+/− n = 5 biological replicates).

Data information: Data are shown as the mean ± S.E.M. of biological replicates. All experiments were performed two times, and representative data are shown. Significant differences were tested using a two‐tailed Mann–Whitney U test (A) or a two‐tailed Student's t‐test (E, F, G). *P < 0.05 and **P < 0.001. NS, no significance.

We next investigated whether the increased EAE pathogenesis in Nedd4l ^+/− mice is dependent on IL‐17R‐mediated signaling and inflammation. An anti‐IL‐17‐specific blocking antibody was continually i.p. injected into Nedd4l ^+/+ and Nedd4l ^+/− mice during the induction of EAE. As shown in Fig 7A and B, the injection of anti‐IL‐17 antibody resulted in significant inhibition of EAE disease symptoms, including clinical score, inflammation, and demyelination, in Nedd4l ^+/+ and Nedd4l ^+/− mice compared to the anti‐isotype control antibody‐injected group. The Nedd4l ^+/− mice exhibited much more severe EAE disease symptoms compared to Nedd4l ^+/+, which was obliterated after the injection of IL‐17‐blocking antibody (Fig 7A and B). Flow cytometry and real‐time PCR analysis of mouse CNS tissues demonstrated that CNS‐infiltrating CD11b^highGr1⁺ neutrophils and IL‐17‐induced genes, such as Il‐6, Cxcl2, and Ccl20, were significantly reduced in IL‐17‐blocking antibody‐treated Nedd4l ^+/− mice compared to mice treated with anti‐isotype control antibody, which was obliterated after the injection of IL‐17‐blocking antibody (Fig 7C–E). However, continual i.p. injection of a Syk‐specific inhibitor during the induction of EAE, BAY 61‐3606, could not eliminate the EAE phenotype difference between Nedd4l ^+/+ and Nedd4l ^+/− mice, suggesting that Nedd4l deficiency promotes the pathogenesis of EAE independent of Syk‐mediated signaling (Fig 7F and G). Collectively, these data suggest that NEDD4L regulates EAE pathogenesis via IL‐17R‐mediated MEKK2‐dependent signaling.

Figure 7. Nedd4l deficiency inhibits MOG_35‐55‐induced EAE via IL‐17‐dependent manner.

• A
  Nedd4l ^+/+ and Nedd4l ^+/− mice (Nedd4l ^+/+ n = 5 biological replicates and Nedd4l ^+/− n = 4 biological replicates) were treated with intraperitoneal (i.p.) injection of an anti‐IL‐17A antibody (100 μg per mouse each time) or appropriate isotype control anti‐IgG antibody on days 7, 9, 11, and 13 after the second MOG_35‐55 immunization. Mean clinical scores were calculated every other day according to the standards described in the Materials and Methods.
• B
  Histology of the spinal cord was analyzed by H&E (up panel) or LFB (down panel) staining on day 14 after the second MOG_35‐55 immunization. Scale bars, 500 μm.
• C, D
  Summary graph of the percentages of immune cells in the CNS. CNS‐infiltrating cells isolated from mice treated as in (A) were stained with the appropriate antibodies (Nedd4l ^+/+ n = 5 biological replicates and Nedd4l ^+/− n = 4 biological replicates).
• E
  Il‐6, Cxcl2, and Cccl20 mRNA in the CNS tissues from mice treated as in (A) (including brain and spinal cords) were measured by real‐time PCR on day 14 after the second MOG_35‐55 immunization (Nedd4l ^+/+ n = 5 biological replicates and Nedd4l ^+/− n = 4 biological replicates).
• F
  Nedd4l ^+/+ and Nedd4l ^+/− mice (Nedd4l ^+/+ n = 5 biological replicates and Nedd4l ^+/− n = 7 biological replicates) were treated with intraperitoneal (i.p.) injection of Syk‐specific inhibitor BAY 61–3,606 (5 mg/kg each time) on days 7, 9, 11, and 13 after the second MOG_35‐55 immunization. Mean clinical scores were assessed as in (A).
• G
  Histology of the spinal cord was analyzed by H&E (left panel) or LFB (right panel) staining on day 14 after the second MOG_35‐55 immunization. Scale, 500 μm.

Data information: Data are shown as the mean ± S.E.M. of biological replicates. All experiments were performed two times, and representative data are shown. Significant differences were tested using a two‐tailed Mann–Whitney U test (A, F) or a two‐tailed Student's t‐test (C, D, E). **P < 0.01 and ***P < 0.001. NS, no significance.

Discussion

In the present study, we show that endogenous NEDD4L inhibits IL‐17, IL‐1α, and TNF‐α‐induced proinflammatory cytokine and chemokine production. NEDD4L promotes MEKK2 ubiquitination and inhibits the protein stability of MEKK2, which is required for IL‐17‐, TNF‐α‐, and IL‐1α‐induced signaling.

NEDD4L is a highly conserved HECT‐type E3 ligase. It has been reported to be involved with multi‐physiological functions in mice by targeting the degradation of many important signal transduction‐associated proteins (Manning & Kumar, 2018). Here, we identified NEDD4L as a negative regulator of IL‐17‐mediated inflammation by showing that Nedd4l knockdown and Nedd4l deficiency increased IL‐17‐induced proinflammatory cytokine and chemokine production. Even heterozygous deficiency of Nedd4l significantly increased IL‐17‐induced proinflammatory cytokine and chemokine production in vitro and in vivo, emphasizing the importance of NEDD4L in the negative regulation of IL‐17R signaling.

IL‐17 functions mainly by activating NF‐κB and MAPK pathways to induce proinflammatory cytokine and chemokine production. Although NF‐κB and MAPK activations have been demonstrated to be signal events downstream of adaptor protein Act1 and TRAF family members in IL‐17R signaling, the detailed mechanisms by which NF‐κB and MAPK are activated in IL‐17R signaling remain largely unknown. MAP3K family members TAK1 and Tpl2 have been involved in IL‐17‐induced NF‐κB and MAPK activation (Xiao et al, 2014). MEKK2 is involved in signal transduction of TCR, BMP, TGF‐β, FGF‐2, IL‐1α, TNF‐α, and TLR9 agonist CpG (Hammaker et al, 2004; Kesavan et al, 2004; Yamashita et al, 2005; Chang et al, 2011; Wen et al, 2015; Wu et al, 2021). Our results demonstrate that MEKK2 is required for IL‐17‐induced p38 and NF‐κB activation in HeLa cells and MEFs. NEDD4L constitutively and directly binds MEKK2 and inhibits MEKK2 expression by promoting K27‐linked poly‐ubiquitination of MEKK2. Mekk2 knockdown completely reverses Nedd4l deficiency mediated increase in IL‐17‐induced p38 and NF‐κB activation, and also completely reverses the increase in IL‐17‐induced cytokine and chemokine production, clearly demonstrating that NEDD4L negatively regulates IL‐17R signaling by inhibiting MEKK2 expression. In this study, we also show that NEDD4L inhibits TNF‐α‐ and IL‐1α‐induced signaling, which has been reported to require MEKK2 to activate JNK and NF‐κB pathways (Zhao & Lee, 1999; Hammaker et al, 2004).

IL‐17, as well as TNF‐α, are important for host immune defense against certain pathogens. However, excessive IL‐17 activities have a high potential to induce damage to inflamed tissues. Thus, it is not surprising that multiple mechanisms restrict IL‐17R signaling to avoid pathogenic conditions. Several IL‐17R signaling negative regulators have been reported, including TRAF3, USP25, A20, SCF^β‐TrCP, and GSK‐3β (Zhong et al, 2012; Garg et al, 2013) (Shen et al, 2009; Zhu et al, 2010; Shi et al, 2011). We show here that NEDD4L negatively regulates IL‐17R signaling via mediating MEKK2 ubiquitination. Upon IL‐17 stimulation, MEKK2 is phosphorylated at Ser520, which is crucial not only for the activation of p38 and NF‐κB pathways but also for NEDD4L‐mediated MEKK2 degradation that negatively regulates IL‐17R‐induced inflammation.

Excessive IL‐17‐, IL‐1α‐, and TNF‐α‐triggered signaling have been linked to the development of many inflammatory responses or autoimmune diseases. Our data show that Nedd4l deficiency in mice significantly promotes IL‐17‐induced pneumonia and peritonitis, and also MOG_35‐55‐induced EAE in IL‐17R‐signaling dependent manner. NEDD4L has also been identified to regulate phosphorylated‐Syk K48‐linked ubiquitination in mast cells (Yip et al, 2016), and Syk acts as an upstream signaling molecule in IL‐17‐induced Act1‐TRAF6 interaction in keratinocytes (Wu et al, 2015), suggesting that NEDD4L may regulate IL‐17R signaling dependent on Syk. However, inhibition of Syk activation in HeLa cells and during induction of the MOG_35‐55‐induced EAE model did not eliminate the phenotype difference in vitro and in vivo, which means NEDD4L regulates IL‐17R‐mediated inflammation and EAE is not dependent on Syk in our experimental system. Taken together, these data confirm that NEDD4L plays a crucial role in IL‐17R‐mediated inflammation and EAE.

Taken together, the present study identifies MEKK2 as a mediator of IL‐17R signaling, and NEDD4L as a common negative regulator of IL‐17‐, TNF‐α‐, and IL‐1α‐induced signaling by inhibiting MEKK2 expression, providing novel insights into the mechanism of IL‐17‐, TNF‐α, and IL‐1α‐induced signaling. IL‐17 stimulation induces the MEKK2 phosphorylation and degradation, which are required for the induction of proinflammatory cytokine and chemokine. Thus, based on our findings, NEDD4L serves as a functional protein in inflammation and autoimmune diseases. NEDD4L could be a target for the treatment of IL‐17R‐related inflammation and autoimmune disorders.

Materials and Methods

Mice

Nedd4l‐deficient homozygous (Nedd4l ^−/−) and control BALB/cByJ (Nedd4l ^+/+) mice were purchased from JAX® Mice, America. In this study, Nedd4l‐deficient homozygous Nedd4l ^+/− mice were used in vivo diseases research for Nedd4l ^−/− mice show kidney defects and some other problems (Manning & Kumar, 2018). Off‐springs of the mice at the age of 6–8 weeks were employed in experiments. All animal experiments were carried out under the National Institute of Health Guide for the Care and Use of Laboratory Animals, with the approval of the Scientific Investigation Board of Laboratory Animal Center of Zhejiang University.

Reagents

Antibodies for phospho‐ERK1/2 (#4370S), phospho‐JNK (#9251S), phospho‐p38 (#4511S), phospho‐p65 (NF‐κB) (#3033 s), ERK1/2 (#4696), JNK (#9251 s), p38 (#8690), NF‐κB p65 (#8242), and NEDD4L (#4013) were from Cell Signaling Technology (Cell Signaling Technology, Beverly, MA). Antibodies for TRAF2 (#5525–1), MEKK3 (#1673–1), and MEKK5 (#1772–1) were from Epitomics (Epitomics, Burlingame, CA). Antibodies for TAK1 (SC‐7967), MEKK2 (SC‐1088), TRAF3 (SC‐949), and TRAF6 (SC‐7221) were obtained from Santa Cruz Biotechnology (Santa Cruz Biotechnology, Santa Cruz, CA). Phospho‐MEKK2 Ser520‐specific antibody was prepared in Abmart (Shanghai, China). Human or mouse IL‐17A/F (human IL‐17, #200–17; Murine IL‐17A, #210–17; human IL‐17F, #200–25; and murine IL‐17F, #210‐17F) were purchased from PeproTech (PeproTech, Rocky Hill, NJ). E1, UBCH5b, and ubiquitin recombinant proteins were gifts from Prof. Zongping Xia (Zhejiang University).

Plasmids

To construct a plasmid‐expressing Myc‐tagged NEDD4L, cDNA‐encoding NEDD4L was amplified by PCR using mRNA of mouse peritoneal macrophage as a template, and cloned in the pCMV‐Entry plasmid. Flag‐tag in the plasmid was deleted by PCR and the plasmid could express Myc‐tagged NEDD4L. MEKK2 cDNA was amplified by PCR using mRNA of HeLa cells as a template and cloned in pCMV‐HA or pcDNA3.1 plasmid. The vectors for NEDD4L and MEKK2 mutants were subsequently generated by PCR amplification.

IL‐17‐induced peritoneal inflammatory responses

Male age‐matched Nedd4l ^+/+ and Nedd4l ^+/− littermates were treated by intraperitoneal injection of IL‐17 (0.5 μg per mouse) or PBS. Twenty‐four hours later, intraperitoneal leukocytes were removed by washing the peritoneal cavity with 5 ml PBS. And then 5 ml of 0.25% trypsin was injected into the peritoneal cavity. Ten minutes later, the trypsin solution was collected, and then 5 ml DMEM containing 10% FBS was used to wash the peritoneal cavity and residual mesothelial cells were collected. The expression of chemokines in the cells was measured by real‐time PCR analysis (Zhong et al, 2012).

IL‐17‐induced pulmonary inflammation

Male age‐matched Nedd4l ^+/+ and Nedd4l ^+/− littermates were treated by intratracheal injection of IL‐17 (2 μg per mouse, 30 μl) or PBS (30 μl). Eight hours later, BLAF fluid was collected in 1.2 ml ice‐cold PBS through the trachea and centrifuged. Chemokines and cytokines in the supernatants were measured using ELISA assays. Cells in the precipitates were resuspended using 0.1 ml staining buffer and counted as lung‐infiltrating cells. The cells were also analyzed using flow cytometry after staining with anti‐Gr‐1 and anti‐CD11b antibodies. The total RNA of the lung tissues was isolated with 1 ml ice‐cold TriZol for real‐time PCR analysis. Lung tissues were dissected, fixed in 4% paraformaldehyde, embedded in paraffin, sectioned, and then stained with hematoxylin and eosin (Bulek et al, 2011).

Pull‐down assay

GST‐tagged MEKK2 (M0324) and Myc‐tagged NEDD4L recombinant proteins were purchased from Sigma and Origene (Origene, Rockville, MD), respectively. GST‐MEKK2 (0.1 μg) bound to glutathione‐agarose beads were mixed with 0.05 μg Myc‐tagged NEDD4L and incubated at 4°C for 4 h with gentle rotation. The beads were washed three times with cell lysis buffer. Then, bound proteins were extracted with loading buffer and analyzed by immunoblot (Xia et al, 2009).

Ubiquitination assay

For in vivo ubiquitination assay, HEK293T cells were transfected with pCDNA3.1‐HA ubiquitin along with other vectors as indicated and cultured for 36 h. After being treated with MG‐132 for 6 h, the cells were lysed in 1% SDS and then heated for 5 min. After centrifuging, the supernatants were 10‐fold diluted with lysis buffer, followed by immunoprecipitation with the appropriate antibodies. Immunoprecipitants were analyzed by immunoblot with anti‐ubiquitin or antibody specific for K63‐linked ubiquitin/K48‐linked ubiquitin. The in vitro MEKK2 ubiquitination assay was carried out in a reaction mixture containing 0.2 μM E1, 0.1 μM UBCH5b, 0.4 μM Myc‐tagged NEDD4L, and 12.5 μM ubiquitin in buffer B at 30°C for 60 min before termination with SDS‐sample buffer (Xia et al, 2009).

Induction and assessment of EAE

Acute EAE was induced and assessed as previously described (Kang et al, 2010). Briefly, acute EAE was induced by subcutaneous immunization with 300 μg of the MOG_35‐55 peptide (Met‐Glu‐Val‐Gly‐Trp‐Tyr‐Arg‐Ser‐Pro‐Phe‐Ser‐Arg‐Val‐Val‐His‐Leu‐Tyr‐Arg‐Asn‐Gly‐Lys, Sangon Biotech Co.) in CFA containing 5 mg/ml heat‐killed H37Ra strain of Mycobacterium tuberculosis (Chondrex, Inc) in the back region and both sides of the vertebrae. And the immunized mice were i.v. injected with pertussis toxin (List Biological Laboratories, Inc.) at a dose of 250 ng per mouse in PBS on the day of immunization and once more 48 h after the first injection. The clinical score was performed in a double‐blinded manner. Mice were examined every 2–3 days for disease symptoms and were double‐blinded and scored for disease severity using the EAE scoring rulers: 0, no clinical signs; 1, limp tail; 2, paraparesis (weakness and incomplete paralysis of one or two hind limbs); 3, paraplegia (complete paralysis of two hind limbs); 4, paraplegia with fore limb weakness or paralysis; and 5, moribund state or death.

Isolation and analysis of CNS inflammatory cells

Central nervous system tissues, including spinal cords and brains, were homogenized in ice‐cold tissue grinders, filtered through a 70 μm cell strainer (Falcon), and the cells were collected by centrifugation at 160 g for 5 min at 4°C. Cells were re‐suspended in 10 ml of 35% Percoll (GE) and centrifuged onto a 5 ml 70% Percoll cushion in 15‐ml tubes at 280 g for 25 min with a low accelerating or breaking speed. Cells at the 35–70% interface were collected and subjected to flow cytometry. Fluorescence‐conjugated monoclonal antibodies to FVD, CD45, CD4, CD8, CD11b, F4/80, and Gr‐1 were stained together for the surface marker analysis.

Histological analysis

Paraffin‐embedded sections (4 mm thick) were subjected to either hematoxylin and eosin (H&E) or Luxol fast blue (LFB) to evaluate inflammation and demyelination, respectively. H&E and LFB were performed by the Histomorphology Platform, Zhejiang University, with the standard protocol performed according to the manufacturer's instructions.

Real‐time quantitative PCR and ELISA

Total RNA was isolated using TRIzol (TAKARA, Ostushiga, JAPAN) and cDNA was synthesized with a reverse‐transcription kit (TAKARA, Ostushiga, JAPAN). The expression of genes was detected by a LightCycler 480 system with SYBR Premix Ex Tap (TAKARA, Ostushiga, JAPAN). The data were calculated by a standard curve method and normalized to the expression of the gene‐encoding gapdh (human) or β‐actin (mouse). The specific primers for individual genes are in Appendix Table S1. The concentration of cytokines and chemokines in the cell culture medium was detected by ELISA assay (ebioscience, MN, USA and R&D, CA, USA).

Statistical analysis

The statistical analysis was performed using a two‐tailed Mann–Whitney U test or a two‐tailed Student's t‐test. Differences were considered significant at a P value of < 0.05.

Author contributions

Ning Wang: Data curation; software; formal analysis; validation; methodology. Yu Jiang: Resources; funding acquisition; validation. Huazhang An: Conceptualization; resources; funding acquisition; validation; methodology; writing – original draft; writing – review and editing. Haofei Wang: Validation; methodology. Hui Li: Conceptualization; resources; data curation; software; formal analysis; funding acquisition; validation; investigation; methodology; writing – original draft; writing – review and editing. Wenlong Lin: Conceptualization; formal analysis; funding acquisition; investigation; writing – original draft; project administration; writing – review and editing. Wangqian Ma: Resources; funding acquisition. Zengfeng Xin: Resources; funding acquisition. Hao Pan: Resources; funding acquisition. Xiaojian Wang: Conceptualization; resources; formal analysis; supervision; writing – original draft; project administration; writing – review and editing. Ting Zhang: Resources; funding acquisition; project administration.

In addition to the CRediT author contributions listed above, the contributions in detail are:

HL, WL, NW, and HW performed experiments. HL and WL performed the statistical analysis. YJ, ZX, HP, WM, and TZ provided some reagents. HA, XW, WL, and HL designed the study. HL, XW, WL, and HA drafted the manuscript.

Disclosure and competing interests statement

The authors declare that they have no conflict of interest.

Supporting information

Appendix

Expanded View Figures PDF

Source Data for Expanded View

PDF+

Source Data for Figure 3

Source Data for Figure 4

Source Data for Figure 5

Source Data for Figure 6

EMBO reports (2022) 23: e54603

Data availability

No primary datasets have been generated and deposited.