OTUD7B controls non-canonical NF-[kgr]B activation through deubiquitination of TRAF3
Hongbo Hu,George C. Brittain,Jae-Hoon Chang,Nahum Puebla-Osorio,Jin Jin,Anna Zal,Yichuan Xiao,Xuhong Cheng,Mikyoung Chang,Tomasz Zal,Chengming Zhu,Shao-Cong Sun,Yang-Xin Fu,Chengming Zhu
DOI: https://doi.org/10.1038/nature11831
IF: 64.8
2013-01-01
Nature
Abstract:Unlike the canonical NF-κB pathway, which depends on degradation of IκBα, the non-canonical NF-κB pathway depends on the inducible processing of p100, a process that not only generates p52 but also leads to nuclear translocation of the p52–RelB dimer2. The processing of p100 is triggered through its phosphorylation by the NF-κB-inducing kinase (NIK) together with a downstream kinase, IκB kinase-α (IKK-α)4, 5. Steady-state non-canonical NF-κB activity is tightly controlled by TRAF3, which constantly targets NIK to the TRAF2–cIAPs E3 complex for degradation6, 7, 8. Activation of non-canonical NF-κB involves signal-induced TRAF3 proteolysis and NIK accumulation, but how this event is regulated is unknown6. The deubiquitinases (DUBs) CYLD and A20 are vital regulators of the canonical NF-κB pathway, but they are not involved in the regulation of non-canonical NF-κB signalling9, 10. Another DUB, OTUD7B (also called Cezanne), was identified on the basis of its sequence homology with A20 within their ovarian tumour (OTU) domains11. We studied the function of OTUD7B using Otud7b knockout (Otud7b−/−) mice (Supplementary Figs 1 and 2). In contrast to the postnatal lethality of A20-deficient mice12, the Otud7b−/− mice did not show obvious defects in survival, although they had moderately reduced body weight (Supplementary Fig. 3). The development of B and T cells appeared to be also normal in the Otud7b−/− mice (Supplementary Fig. 4). These phenotypes indicate fundamental differences in signalling function between OTUD7B and A20. The OTUD7B deficiency had no appreciable effect on canonical NF-κB activation in mouse embryonic fibroblasts (MEFs), bone-marrow-derived macrophages (BMDMs), or B cells (Supplementary Fig. 5). Although OTUD7B has a related molecule, OTUD7A, OTUD7A was detected mainly in the central nervous system (Supplementary Fig. 6), arguing against the possibility of functional redundancies. We next examined the role of OTUD7B in regulating NF-κB activation by three well-characterized non-canonical NF-κB-stimulating receptors: lymphotoxin-β receptor (LTβR), CD40 and BAFF receptor (BAFFR)1, 2. OTUD7B deficiency greatly enhanced NF-κB activation by agonistic anti-LTβR and anti-CD40 antibodies or BAFF (Fig. 1a–c), a result that was not due to differential expression of the receptors (Supplementary Fig. 7). The heightened NF-κB DNA-binding activity in Otud7b−/− cells was associated with a marked increase in nuclear levels of p52 and RelB and concomitant loss of cytoplasmic p100 (Fig. 1d–f). However, OTUD7B deficiency did not promote the activation of canonical IKK or MAP kinases (Supplementary Fig. 8), indicating a specific role for OTUD7B in non-canonical NF-κB regulation. We also found that OTUD7B expression was induced by the non-canonical NF-κB signals (Supplementary Fig. 9), indicating a negative feedback function. Compared with wild-type cells, Otud7b−/− cells displayed a markedly higher level of TRAF3 degradation and NIK accumulation in response to the non-canonical NF-κB inducers (Fig. 2a, b and Supplementary Fig. 10). This effect was specific for the non-canonical NF-κB pathway, as OTUD7B deficiency did not alter the fate or function of TRAF3 in the type I interferon (IFN-I) pathway (Supplementary Figs 11 and 12)13, 14. The loss of OTUD7B promoted anti-CD40-stimulated TRAF3 ubiquitination (Fig. 2c). Although the recombinant OTU domain of OTUD7B preferentially hydrolyses K11-linked ubiquitin chains in vitro15, we did not detect appreciable levels of K11 ubiquitination of TRAF3 (Supplementary Fig. 13), which might be due to the quality of the anti-K11 ubiquitin antibody or the dominant K48 ubiquitination of TRAF3 (ref. 7). However, K48 ubiquitination of TRAF3 was markedly enhanced in OTUD7B-deficient cells (Fig. 2c). Because OTUD7B also displays K48-specific DUB function in transfected cells16, it is conceivable that the specificity of OTUD7B may be regulated by cofactors in vivo, as implicated for A2010. The OTUD7B-mediated TRAF3 deubiquitination required its DUB activity, as a catalytically inactive OTUD7B mutant, C194S/H358R (CH), failed to deubiquitinate TRAF3 (Fig. 2d, e). Consistently, wild-type OTUD7B, but not OTUD7B(CH), suppressed the activation of non-canonical NF-κB signalling (Fig. 2f). In response to CD40 or LTβR stimulation, OTUD7B was rapidly recruited to TRAF3 (Fig. 3a), and this molecular interaction was coupled with the receptor recruitment of OTUD7B along with TRAF3, TRAF2 and c-IAP2 (Fig. 3b, c). This finding was intriguing, because TRAF3 ubiquitination is mediated by c-IAP following their recruitment to the stimulating receptors2, 7. Although OTUD7B deficiency did not affect the recruitment of TRAF2, TRAF3 and c-IAP (Fig. 3c), TRAF3 knockdown blocked the receptor recruitment of OTUD7B (Supplementary Fig. 14). The ubiquitin-association (UBA) domain, but not the zinc finger (ZF) domain, of OTUD7B was required for its association with TRAF3 and receptor recruitment (Fig. 3d). The UBA domain of OTUD7B was also indispensable for negatively regulating non-canonical NF-κB signalling (Supplementary Fig. 15). These findings indicate that OTUD7B engages TRAF3 in response to cellular stimuli and, thereby, is recruited to the receptor complex, where it may mediate TRAF3 deubiquitination. Our finding that OTUD7B controls signal-induced, but not steady-state, activation of non-canonical NF-κB suggests specific roles for this DUB in the regulation of immune function. Otud7b−/− mice did not show obvious structural abnormalities of secondary lymphoid organs or the thymus, although the medullary thymic epithelial cells of Otud7b−/− mice had elevated expression of the autoimmune regulator (AIRE) and several other related genes, indicating a higher level of maturation (data not shown). Furthermore, Otud7b−/− mice displayed a remarkable phenotype in the mucosal tertiary lymphoid structures, characterized by a marked increase in the size and number of B-cell-containing colonic patches (CLPs) (Fig. 4a and Supplementary Fig. 16). Consistently, Otud7b−/− mice had a significant increase in faecal IgA concentration compared to wild-type mice (Fig. 4b). Intestine is an organ that maintains dynamic host–microbiota homeostasis17, and unlike the secondary lymphoid organs, the development of which is programmed in the fetus, the formation and maintenance of intestinal tertiary lymphoid tissues occur in adults and involve microbial triggers and LTβR signalling18, 19, 20, 21. Indeed, injection of Otud7b−/− mice with a LTβR fusion protein (LTβR–Ig), known to block the interaction of LTβR with its ligands18, suppressed the lymphoid hyperplasia (Fig. 4c and Supplementary Fig. 17). Consistent with the involvement of chemokines in LTβR-mediated induction of lymphoid follicles, the intestines of Otud7b−/− mice had elevated expression of two major chemokines, CXCL12 and CXCL13, and the cell adhesion molecule MADCAM1, which was suppressed upon injection of LTβR–Ig (Fig. 4d). The LTβR-mediated induction of these genes was also enhanced in the Otud7b−/− MEFs (Fig. 4e). The results described above prompted us to examine whether OTUD7B negatively regulates mucosal immunity against infections. Indeed, the Otud7b−/− mice were considerably more resistant to a well-defined intestinal bacterial pathogen, C. rodentium22, as shown by the significantly reduced bacterial load in organs and faeces and the ameliorated crypt hyperplasia in the colon (Supplementary Fig. 18). To compare the C. rodentium-induced lethality in the wild-type and Otud7b−/− mice, we took advantage of the recent finding that depletion of CD4 T cells in wild-type mice renders the mice sensitive to C. rodentium-induced lethality (Y.-X.F. et al., unpublished data). Under these conditions, all of the wild-type mice died of C. rodentium infection by day 14, whereas 75% of the Otud7b−/− mice survived infection (Fig. 4f). Consistently, Otud7b−/− mice also experienced much milder body weight loss and had a significantly reduced faecal bacterial load compared to wild-type mice (Fig. 4g and Supplementary Fig. 19). Collectively, these findings indicate the possibility that inhibition of OTUD7B may enhance mucosal immunity against C. rodentium infection. Deregulated non-canonical NF-κB activation in B cells, as seen in B-cell-conditional Traf3−/− mice, is associated with B-cell hyperplasia and aberrant antibody production23. Consistently, OTUD7B-deficient B cells displayed elevated proliferative and survival ability when stimulated in vitro (Fig. 4h and Supplementary Fig. 20) and were hyper-responsive to antigen-stimulated antibody production when adoptively transferred into Rag1−/− mice along with wild-type T cells (Fig. 4i). We found that Otud7b−/− mice did not show abnormal B-cell homeostasis under non-transferred conditions (Supplementary Fig. 4). This was probably due to attenuated activation of the OTUD7B-deficient T cells (data not shown), a phenotype that has also been observed with TRAF3-deficient T cells24, 25. Notably, upon injection with low doses of recombinant BAFF, the Otud7b−/− mice displayed overt B-cell hyperplasia (Fig. 4j), coupled with accumulation of various antibody isotypes (Fig. 4k), thus further indicating a negative role for OTUD7B in regulating B-cell responses. Our findings establish OTUD7B as a pivotal negative regulator of the non-canonical NF-κB pathway. Although another DUB, DUBA, has been implicated in TRAF3 deubiquitination, DUBA regulates the K63 ubiquitination and non-degradative function of TRAF3 in the pattern-recognition receptor (PRR)-induced IFN-I signalling pathway26. So far, OTUD7B is the only DUB known to regulate TRAF3 degradation and non-canonical NF-κB signalling. OTUD7B is unique in that it specifically regulates signal-induced non-canonical NF-κB activation. These unique functions of OTUD7B make it an attractive candidate to be exploited as a therapeutic target to boost mucosal immunity and treat diseases associated with the non-canonical NF-κB pathway. Otud7b−/− mice (in the C57BL/6x129/sv genetic background) were generated by a conventional gene-targeting approach (Deltagen, Inc.) (Supplementary Fig. 1) and backcrossed for four generations to the C57BL/6 background. Otud7b+/− heterozygous mice were bred to generate age-matched wild-type (+/+) and knockout (−/−) experimental mice. Mice were maintained in specific pathogen-free facility, and all animal experiments were conducted in accordance with protocols approved by the Institutional Animal Care and Use Committee of the University of Texas MD Anderson Cancer Center. HA-tagged OTUD7B (pHR6-Cezanne) was provided by P. Evans28, and OTUD7B C194S/H358R (OTUD7B(CH)), harbouring point mutations in two conserved residues of the catalytic domain, was created by site-directed mutagenesis (Agilent Technologies). OTUD7B, OTUD7B(CH) and OTUD7B(1–810) and OTUD7B(40–843) were cloned into pCLXSN(GFP) vector29 for MEF infections. OTUD7B and OTUD7B(CH) were cloned into pGEX-4T-1 bacterial expression vector for producing glutathione-S-transferase (GST) fusion proteins in Escherichia coli. The HA-tagged TRAF3 and ubiquitin vectors have been described previously6, and the HA–ubiquitin K48 (where all of the lysines, except for lysine 48, were substituted with alanines) was provided by Z. Chen. Murine LTβR antibody (ACH6) was provided by J. Browning (Biogen). Anti-c-IAP2 (H-85) and anti-ubiquitin (P4D1) were from Santa Cruz Biotech, anti-K48 ubiquitin and anti-K11 ubiquitin were from Millipore, and anti-OTUD7B was from Proteintech. Other antibodies for cell stimulation, immunoblot and flow cytometry were as described6, 27. Recombinant LTβR–Ig was described previously30. 5′ppp-dsRNA was from Invivogen, and other agents were as described27. HEK-293 cell culture and transfection, splenic B-cell purification, and the generation of BMDMs and MEFs were as described27. Thymic stromal cells were isolated as reported31. For B-cell proliferation assays, the cells were stimulated with anti-murine CD40 (500 ng ml−1), BAFF (200 ng ml−1), LPS (100 ng ml−1) or anti-mouse IgM (10 μg ml−1) for 40 h and then pulsed for 8 h with 3H-thymidine. Primary MEFs were immortalized by infection with a retroviral vector encoding the adenoviral E1A proto-protein (pCL-E1A). For OTUD7B reconstitution, the immortalized MEFs were infected with pCLXSN(GFP) encoding OTUD7B or its mutants and screened based on GFP expression. For gene induction and signalling studies, BMDMs were stimulated with lipopolysaccharide (LPS) (100 ng ml−1), poly(I:C) (10 μg ml−1), R848 (2.5 μg ml−1), or CpG ODN1668 (1 μg ml−1), B cells were stimulated with anti-murine CD40 (500 ng ml−1), BAFF (200–400 ng ml−1), LPS (5 μg ml−1), or anti-IgM (10 μg ml−1), and MEFs were stimulated with anti-LTβR (500 ng ml−1), LPS (1 μg ml−1), TNF-α (50 ng ml−1), IL-1β (10 ng ml−1), or lipofectamine-transfected poly(I:C) (2.5 μg ml−1) or dsRNA (1 μg ml−1) (for stimulation of RIG-I). The recruitment of OTUD7B and other signalling molecules to CD40 was detected as previously described32. Briefly, M12 cells stably expressing human CD40 (M12-hCD40) (2 × 107) were stimulated with anti-hCD40 (in 1 ml of growth medium) and lysed in 600 μl of a lysis buffer. The CD40 complexes were precipitated by protein G-agarose and analysed by immunoblotting. The recruitment of proteins to LTβR was detected in a similar way, except for the use of MEFs stimulated with anti-LTβR antibody22. TRAF3 was isolated by immunoprecipitation under denaturing conditions27 to inactivate DUBs and disrupt protein complexes, and the ubiquitinated TRAF3 was detected by immunoblotting using pan-ubiquitin or chain-specific ubiquitin antibodies. For transfection models, TRAF3 was transfected into HEK293 cells along with HA–ubiquitin or HA–ubiquitin K48. Following TRAF3-denaturing immunoprecipitation, the TRAF3–ubiquitin conjugates were detected by immunoblotting using anti-HA. In deubiquitination assays, the isolated TRAF3–ubiquitin conjugates were incubated with purified GST–OTUD7B or GST–OTUD7B(CH) recombinant proteins in a deubiquitination buffer33 for 16 h and then subjected to immunoblotting using anti-HA. Cell suspensions were subjected to flow cytometry analyses using a LSRII flow cytometer (BD Biosciences), as described previously34. The data were analysed using FlowJo software. RNA preparation and qPCR assays were as described27 using gene-specific primers listed in Supplementary Table 1. B220+ B cells and CD90.2+ T cells were isolated from the splenocytes of wild-type or Otud7b−/− mice using magnetic beads (Miltenyi Biotec). The isolated cells were >95% pure, as determined by flow cytometry. Wild-type T cells (5 × 106) were mixed with either wild-type or Otud7b−/− B cells (5 × 106) and then injected via a tail vein into Rag1−/− mice. After 16 h, the recipient mice were subjected to immunization with NP-KLH and the sera was collected on day 7, day 14 and day 21. ELISA was used to detect the different isotopes of antibodies (Southern Biotech). For measuring faecal IgA concentration, faeces were taken from age-matched wild-type and Otud7b−/− mice, weighted, and homogenized in PBS. IgA level was detected by ELISA. To block the LTβR signalling, wild-type or Otud7b−/− mice were injected intraperitoneally with 100 μl of PBS or LTβR–Ig (1 μg μl−1) on day 0 and day 5 and killed on day 10. Colons were collected for histological examination. Nuclear extracts were prepared from indicated cells and subjected to EMSA using the following 32P-radiolabelled oligonucleotide probes: NF-κB, 5′-CAACGGCAGGGGAATTCCCCTCTCCTT-3′; NF-Y, 5′-AAGAGATTAACCAATCACGTACGGTCT-3′; OCT1, 5′-TGTCGAATGCAAATCACTAGAA-3′. C. rodentium strain DBS100 (ATCC 51459) was cultured by shaking at 37 °C overnight in LB broth. The concentration of bacteria was assessed by measuring optical density at 600 nm and confirmed by plating serially diluted bacterial cultures for colony-forming unit (c.f.u.) determination. Mice were orally injected with 2–4.5 × 109 C. rodentium in a total volume of 200 μl per mouse. For CD4 T-cell depletion, the mice were also injected intraperitoneally with either a control rat IgG or a rat anti-mouse CD4 antibody (GK1.5, 50 μg per mouse) on day 0, 5 and 10. Mice were allowed access to food and water after the inoculation and killed at the indicated times after infection. Colons were isolated and fixed for haematoxylin and eosin staining to evaluate tissue pathology. Spleen, liver and faeces were collected, weighed and homogenized. The homogenates were serially diluted and plated on MacConkey agar plates, and the plates were incubated at 37 °C for 24 h followed by counting the C. rodentium colonies (pink colonies). Whole-mount immunofluorescence staining of colonic lymphoid follicles was performed essentially as described35. Colons were collected, flushed and opened longitudinally along the mesenteric border. The tissues were incubated with shaking at room temperature for 60 min in Hank’s balanced-salt solution containing 5 mM EDTA, fixed in 4% paraformaldehyde, and rinsed. After blocking nonspecific epitopes with a blocking buffer35 containing 5% BSA, colons were stained with Alexa Fluor 594-conjugated anti-B220 antibody. Pictures were taken by Leica SP5 RS confocal microscope and analysed by SlideBook 5.0 software. The B220+ lymphoid follicles larger than 10,000 μm2 were considered as CLPs. Two-tailed unpaired t-tests were performed using Prism software. P values less than 0.05 were considered significant, and the level of significance was indicated as *P < 0.05, **P < 0.01 and ***P < 0.001. Download references We thank Z. Chen and P. Evans for expression vectors and J. Browning and Biogen for the anti-LTβR antibody. We also thank the personnel from the flow cytometry, DNA analysis, animal facility, and histology core facilities at The MD Anderson Cancer Center for technical assistance. This study was supported by grants from the National Institutes of Health (AI057555, AI064639 and GM84459 to S.-C.S.; CA137059 to T.Z.; T32CA009598 to G.C.B.) and the Sister Institution Network Fund of MD Anderson Cancer Center.