The expression of proinflammatory cytokines and chemokines in response to TCR agonists is regulated by the caspase-recruitment domain membrane-associated guanylate kinase 1 (CARMA1) signalosome through the coordinated assembly of complexes containing the BCL10 adaptor protein. We describe a novel mechanism to negatively regulate the CARMA1 signalosome by the “death” adaptor protein caspase and receptor interacting protein adaptor with death domain (CRADD)/receptor interacting protein-associated ICH-1/CED-3 homologous protein with a death domain. We show that CRADD interacts with BCL10 through its caspase recruitment domain and suppresses interactions between BCL10 and CARMA1. TCR agonist-induced interaction between CRADD and BCL10 coincides with reduction of its complex formation with CARMA1 in wild-type, as compared with Cradd-deficient, primary cells. Finally, Cradd-deficient spleen cells, CD4+ T cells, and mice respond to T cell agonists with strikingly higher production of proinflammatory mediators, including IFN-γ, IL-2, TNF-α, and IL-17. These results define a novel role for CRADD as a negative regulator of the CARMA1 signalosome and suppressor of Th1- and Th17-mediated inflammatory responses.
Proinflammatory pathways are maintained through the assembly of oligomeric signaling complexes known as signalosomes (1). One of them is the caspase-recruitment domain (CARD) membrane-associated guanylate kinase 1 (CARMA1) signalosome employed by multiple immunoreceptors, including the T and B cell Ag receptors (2, 3).
In T cells, Ag receptor (TCR) engagement results in phosphorylation-induced conformational changes in CARMA1 necessary for the recruitment of BCL10, together with the MALT1 paracaspase protein (4–6). These CARD domain-mediated interactions between BCL10 and CARMA1 (7, 8) are required for TCR-coupled expression of cytokines under NF-κB control. While searching in silico for additional proteins bearing CARD and DEATH domains, a novel adaptor protein, designated caspase and receptor interacting protein adaptor with death domain (CRADD) (Mouse Genome Informatics 1336168), or receptor interacting protein-associated ICH-1/CED-3 homologous protein with a death domain (RAIDD), was discovered and characterized for its potential role in apoptosis (9, 10). CRADD was subsequently linked to DNA damage based on its inclusion in a complex with p53-induced death domain protein (PIDD) and caspase-2, termed the PIDDosome (11, 12). However, further studies provided little support for this concept and described other functions for PIDD and caspase-2 in cell cycle control (13–17) and for CRADD/RAIDD, PIDD, and caspase-2 in the cellular responses to lethal heat shock (18). As an alternative model of CRADD function, we hypothesized that CRADD may play a role in TCR signal transduction as an adaptor protein through its CARD domain. In this study, we establish a novel immunoregulatory function for CRADD/RAIDD as a suppressor of the CARMA1 signalosome and as a negative regulator of the T cell-mediated proinflammatory response.
Materials and Methods
Animal work was done in accordance with the Animal Care and Use Committee guidelines for Vanderbilt University. Animals were housed in specific pathogen-free conditions. The Cradd gene (Mouse Genome Informatics 1336168) was disrupted in embryonic stem cells, introduced into the germline (19) (Supplemental Fig. 1), and maintained in a 129SVJ-C57BL/6 background.
RBC-lysed spleen cells from wild-type and Cradd−/− mice were cultured in RPMI 1640 medium with 10% FBS at 37°C in a CO2 incubator. Other cell types were cultured in DMEM medium with 10% FBS.
CD4+ T cells were purified with the CD4+ Isolation Kit II, and naive T cells were purified through the naive cell purification kit according to the manufacturer’s instructions (Miltenyi Biotec). Isolated CD4+ T cells were >95% pure; naive T cells were >92% pure as determined by identifying CD4+CD44loCD62Lhi cells.
Tissues from wild-type and Cradd−/− mice were rinsed in PBS, frozen in liquid N2, resuspended in 10 ml hypotonic buffer (20 mM HEPES-KOH, 10 mM KCl, 1 mM MgCl2, 1 mM EDTA, 1 mM EGTA, and 1 mM DTT [pH 7.5]), and subjected to three freeze/thaw cycles. Splenocytes and cultured cells were lysed in hypotonic buffer plus 0.1% Nonidet P-40. Lysates were precleared with 50 μl protein G-Sepharose beads (Sigma-Aldrich) at 4°C for 1 h and incubated with 5 μg appropriate Abs overnight with protein G beads at 4°C. The beads were washed three times before the proteins were separated by SDS-PAGE.
Preparation of nuclear extracts
Cytoplasmic and nuclear extracts were prepared as described with minor modifications (20). Cells were washed with ice-cold PBS and then lysed by incubation in 200 μl lysis buffer (10 mM HEPES [pH 7.9], 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 0.4% Nonidet P-40, 1 mM DTT, and protease inhibitor mixture [Sigma-Aldrich]) on ice for 5 min. Nuclei were pelleted by centrifugation, supernatant saved as the cytoplasmic extract, and nuclear pellet washed in 1 ml lysis buffer. Nuclear extracts were prepared by resuspension in salt extraction buffer (20 mM HEPES [pH 7.9], 400 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, and protease inhibitor mixture), followed by shaking for 30 min, centrifugation, and removal of genomic DNA.
Mammalian two-hybrid assay
293T cells were transfected with plasmid vectors expressing GAL4 DNA binding domain and VP16 trans-activation domain fusion proteins according to the Matchmaker two-hybrid system (Clontech). Chloramphenicol acetyltransferase (CAT) activities indicative of protein interactions were measured by the FAST CAT assay (Molecular Probes).
Cytokines and chemokines in plasma or tissue culture media were assayed by cytometric bead array (BD Biosciences, San Diego, CA) and analyzed on an FACSCalibur flow cytometer (BD Biosciences). Standard curves were determined for each cytokine.
Results and Discussion
CRADD interacts with BCL10 through CARD domain engagement
To test our hypothesis that CRADD interacts with the CARMA1 signalosome, we performed Western blot analysis using extracts prepared from primary spleen cells. As shown in Fig. 1A, CRADD specifically coprecipitated with BCL10 and with PIDD, which served as a positive control, but not with MALT1, CARMA1, or caspase 2. Endogenous cellular CRADD protein was also recovered in human embryonic kidney 293T cells by using an epitope-tagged BCL10 protein (Fig. 1B), suggesting that the two proteins interact directly. Moreover, in a mammalian two-hybrid assay, CRADD specifically interacted with BCL10 through the CARD domain (Fig. 1C, column 2) and not the DEATH domain (Fig. 1C, column 3). CRADD did not interact with MALT1 (Fig. 1C, column 4), consistent with coimmunoprecipitation data (Fig. 1A).
We then confirmed that the CARD domains of BCL10 and CARMA1 interact in the mammalian two-hybrid assay (Fig. 1D, column 1) and showed that CRADD blocks this interaction in a dose-dependent manner (Fig. 1D, columns 2–6). Thus, we document a novel interaction of CRADD with BCL10 through CARD domain engagement and exclusion of CARMA1 from its union with BCL10.
Agonist-induced CRADD–BCL10 interaction coincides with reduced BCL10–CARMA1 assembly and NF-κB activation in primary cells
We determined whether Cradd deficiency had a measurable impact on CARMA1 signalosome formation. To generate Cradd-deficient mice, we employed embryonic stem cells containing a gene trap mutation (19) (Supplemental Fig. 1). Homozygous mutant animals displayed no obvious developmental abnormalities or altered lymphocyte profiles in the spleen or thymus, as previously reported (13, 14). We then tested the effect of Cradd deficiency on agonist-induced BCL10–CARMA1 interactions. In Cradd-sufficient mouse spleen cells, the amount of CRADD that coprecipitated with BCL10 was enhanced by Con A, a broadly acting T cell agonist and potent inducer of systemic inflammation (Fig. 2A), and was also induced by the T cell-specific stimulus anti-CD3/CD28 in a time-dependent manner (Fig. 2B). Moreover, we analyzed purified naive (CD44loCD62Lhi) CD4+ T cells following anti-CD3/CD28–evoked TCR activation. Similar to whole splenocytes, an interaction between CRADD and BCL10 was induced at 2 h after stimulation (Fig. 2B). Cradd deficiency had a striking impact on CARMA1 signalosome formation by enhancing BCL10 and CARMA1 interaction induced by Con A (Fig. 2C). Notably, the amount of CARMA1 that coprecipitated with BCL10 after Con A treatment was 5-fold (at 30 min) and 3.5-fold (at 2 h) higher in Cradd-deficient as compared with Cradd-sufficient spleen cells. The enhancement of CARMA1 signalosome assembly in CRADD-deficient cells resulted in predicted downstream activation events as demonstrated by elevated levels of nuclear RelA/p65 (p65) in Cradd-deficient purified CD4+ T cells (Fig. 2D).
Cradd deficiency enhances agonist-induced cytokine and chemokine expression in mice, isolated spleen cells, and CD4+ T cells and is reversed by retroviral CRADD gene transfer
Heightened agonist-induced interactions between CARMA1 and BCL10 and enhanced NF-κB activation in Cradd-deficient T cells suggested that CRADD suppresses signaling pathways involved in agonist-induced cytokine and chemokine expression. Confirming this hypothesis, cytokine and chemokine expression was remarkably enhanced in Cradd-deficient mice following i.v. administration of Con A (Fig. 3A). Plasma levels of TNF-α, INF-γ, IL-2, IL-6, and MCP-1, which peaked 8–12 h after Con A injection, were 8–20-fold higher in Cradd-deficient as compared with wild-type animals. Of note, extracts of spleen cells isolated from these animals at 6 h after Con A injection immunoprecipitated with anti-CD3 Ab demonstrated association of CRADD with components of the TCR signaling complex ZAP70, SLP-76, and LAT (Supplemental Fig. 2A, 2B).
Enhanced responses reflected by higher levels of IL-2, IFN-γ, and TNF-α, characteristic of a Th1 response, were also found in purified spleen cells from Cradd-knockout animals after treatment with Con A (Fig. 3B), TCR agonistic anti-CD3/CD28 Abs (Fig. 3C), and superantigen staphylococcal enterotoxin B (not shown). Likewise, higher levels of IL-17, a hallmark of Th17 cells, were induced in splenocytes derived from Cradd-deficient mice. We verified this remarkable enhancement of Th1- and Th17-type cytokine responses in purified CD4+ T cells stimulated with anti- CD3/CD28 (Fig. 3D). By contrast, levels of IL-4 and IL-13, characteristic of Th2 responses, remained below the threshold of detection under the same experimental conditions (not shown).
As a final demonstration of the inhibitory function of CRADD in T cell signaling, CRADD expression was restored in Cradd-deficient spleen cells by retroviral gene transfer. Spleen cells were infected with retroviruses containing the CRADD gene together with the Thy1.1 gene that encodes a cell-surface protein used to immunoselect retrovirally infected cells and epitope tags c-Myc (MC) and AU1 (AU). As expected for the complementation of phenotypes caused by loss of Cradd function, TNF-α (Fig. 4A) and IFN-γ (Fig. 4B) were suppressed to nearly wild-type levels in mutant cells infected with CRADD expression vectors but not in cells infected with an empty vector (TH), consistent with the expression of epitope-tagged CRADD protein in virally infected spleen cells (Fig. 4C). Thus, we showed that the cytokine suppressing activity of CRADD can be restored by its gene transfer in mutant spleen cells and established that the cytokine-enhancing mutant phenotype was caused by Cradd deficiency.
Cumulatively, our results identify a previously unrecognized immunoregulatory function of the CRADD adaptor protein as a suppressor of TCR agonist-evoked signaling. This function is dependent on the CARD domain of CRADD, which binds to BCL10 and interferes with signal-induced targeting of BCL10 to CARMA1, an essential step in formation of the CARMA1, BCL10, and MALT complex required for activation of the NF-κB signaling pathway via TNFR-associated factor 6 and NF-κB essential modulator/IκB kinase γ (21, 22). This new suppressor function of CRADD operates in primary T cells, as judged by: 1) associations between CRADD and BCL10 in spleen cells and CD44hiCD62Llo naive CD4+ T cells stimulated with T cell-specific agonist anti-CD3/CD28 (Fig. 2B); 2) strikingly enhanced cytokine expression in Cradd-deficient spleen cells and purified CD4+ T cells treated with the T cell-specific agonists anti-CD3/CD28 and staphylococcal enterotoxin B and displaying Th1 and Th17 cytokine profiles (Fig. 3); and 3) enhanced recovery of activated T cells from the spleens of Con A-treated Cradd-deficient mice as assessed by CD69 expression (not shown). Given the widespread involvement of BCL10 and CARMA proteins as regulators of NF-κB essential modulator/IκB kinase γ in other immunoreceptor-activated signaling pathways (2, 3), CRADD may also suppress immunoreceptor signaling in other cell types. Moreover, CRADD may also function as a regulatory adaptor in other signaling pathways, which awaits future analysis.
We propose that CRADD provides the first example, to our knowledge, of a TCR-proximal adaptor function that negatively regulates the CARMA1 signalosome, thereby limiting potentially harmful proinflammatory cytokine and chemokine expression under conditions of excessive or inappropriate T cell stimulation. Further understanding of CRADD regulation will inform therapeutic strategies to control the inflammatory responses in human disease.
The authors have no financial conflicts of interest.
We thank Robert Collins for histological analysis, Luc Van Kaer for the initial screen of immune system defects in Cradd-deficient mice, Dean Ballard and Eric Sebzda for helpful discussions, Abudi Nashabi for technical assistance, and Susanna Richards for editorial assistance.
This work was supported by U.S. Public Health Service National Institutes of Health Grants HL687440 (to H.E.R. and J.H.), HL069452, HL085833, and AA015752 (to J.H.), and 1KO8 DK090146 (to D.J.M.) and by a Kleberg Family grant (to H.E.R.). Additional support was provided by the Department of Microbiology and Immunology and by Cancer Center Support Grant P30CA68485.
The online version of this article contains supplemental material.
Abbreviations used in this article:
- caspase-recruitment domain
- caspase-recruitment domain membrane-associated guanylate kinase 1
- chloramphenicol acetyltransferase
- caspase and receptor interacting protein adaptor with death domain
- p53-induced death domain protein
- receptor interacting protein-associated ICH-1/CED-3 homologous protein with a death domain.
- Received May 31, 2011.
- Accepted January 18, 2012.
- Copyright © 2012 by The American Association of Immunologists, Inc.