Report

NF-AT Activation Induced by a CAML-Interacting Member of the Tumor Necrosis Factor Receptor Superfamily

See allHide authors and affiliations

Science  03 Oct 1997:
Vol. 278, Issue 5335, pp. 138-141
DOI: 10.1126/science.278.5335.138

Abstract

Activation of the nuclear factor of activated T cells transcription factor (NF-AT) is a key event underlying lymphocyte action. The CAML (calcium-modulator and cyclophilin ligand) protein is a coinducer of NF-AT activation when overexpressed in Jurkat T cells. A member of the tumor necrosis factor receptor superfamily was isolated by virtue of its affinity for CAML. Cross-linking of this lymphocyte-specific protein, designated TACI (transmembrane activator and CAML-interactor), on the surface of transfected Jurkat cells with TACI-specific antibodies led to activation of the transcription factors NF-AT, AP-1, and NFκB. TACI-induced activation of NF-AT was specifically blocked by a dominant-negative CAML mutant, thus implicating CAML as a signaling intermediate.

We identified proteins that can interact with CAML in a two-hybrid screen (1, 2). To determine if any of these CAML-binding proteins affected signaling in T cells, we examined their ability to modulate activity of the Ca2+-dependent transcription factor NF-AT (3). Overexpression of the two-hybrid clones in Jurkat T cells revealed that expression of one clone (encoding the TACI protein) led to activation of NF-AT, suggesting that TACI may lie in the same signaling pathway as CAML. The deduced amino acid sequence of TACI (Fig.1A) (4) includes a single hydrophobic region (residues 166 to 186) that has features of a membrane-spanning segment. Analysis of the protein sequence (5) predicted extracellular exposure for the NH2-terminus with a cytoplasmic COOH-terminus. Although TACI lacks an NH2-terminal signal sequence, the presence of an upstream stop codon indicates that the complete open reading frame is contained within the clone (6). The predicted cell-surface location of TACI was confirmed in intact Cos-7 cells transfected with an expression plasmid encoding TACI with an NH2-terminal FLAG epitope tag. Staining with monoclonal antibody to FLAG revealed TACI localized to the cell surface (Fig.2A). TACI is therefore a type III transmembrane protein with an extracellular NH2-terminus in the absence of a cleaved signal sequence (7). Inspection of the TACI protein sequence also revealed two repeated regions (residues 33 to 66 and 70 to 104) that are 50% identical. A PROSITE motif search (8) identified this repeated region as a cysteine-rich motif characteristic of the tumor necrosis factor receptor (TNFR) superfamily. Comparison of TACI with other members of TNFR superfamily (Fig. 1B) demonstrates the similarity between these domains, with the best match to cysteine-rich domains of DR3 (also known as Wsl-1, Apo-3, or TRAMP) (9).

Figure 1

Structural features of TACI. (A) Predicted amino acid sequence. Residues composing the proposed transmembrane domain are boxed, and the cysteine-rich TNFR repeats are underlined. (B) Alignment of the cysteine-rich repeats from TACI with those from some members of the TNFR superfamily (9,21). Boxed letters indicate residues from other receptors that share identity with the TACI protein; underlined residues are conservative substitutions. Asterisks denote conserved cysteines and other residues that stabilize the conformation of TNFRI (22).

Figure 2

Cell-surface expression of TACI in some B and T lymphocytes. (A) Cos-7 cells transiently transfected with a plasmid encoding NH2-terminal FLAG-tagged TACI. The NH2-terminal epitope tag appeared to be extracellular, because specific staining with the antibody was observed even in the absence of permeabilizing detergent. (B) Expression of TACI in lymphoid cells. Multiple-tissue Northern (RNA) blot was probed with TACI cDNA. TACI mRNA was not detected in heart, brain, placenta, lung, liver, skeletal muscle, kidney, or pancreas. pb indicates peripheral blood. (C) Shown is flow cytometric analysis of fresh peripheral blood lymphocytes stained with antibodies specific for either B cells (CD19) or T cells (CD3, y axes) and rabbit IgG control antibodies or antibodies to TACI (x axes), as indicated. (D) Induction of TACI expression in a subset of activated T cells. Flow cytometry is shown for peripheral blood T lymphocytes (23) cultured in the absence or presence of PMA (25 ng/ml) and ionomycin (1 μM) for 48 hours, and stained with anti-CD2 (T cell–specific, y axes) and control IgG or anti-TACI (x axes), as indicated. PMA and ionomycin treatment induced surface TACI expression in 54% of CD2-positive T cells. Unstim., unstimulated; iono, ionomycin.

Northern blot analysis of TACI mRNA demonstrated a 1.4-kb transcript expressed in spleen, small intestine, thymus, and peripheral blood lymphocytes, suggesting that a single TACI transcript is present in both T and B lymphocytes (Fig. 2B). Specific antibody staining of peripheral blood cells with a polyclonal antibody to TACI (10) revealed the presence of TACI on the surface of B cells, but not resting T cells (Fig. 2C). Because expression of other TNFR members such as CD30 is increased after activation of T lymphocytes (11) and because TACI appears to be expressed in thymocytes, we examined T cells activated with ionomycin and phorbol ester. Such treatment of T cells induced the synthesis of cell-surface TACI in 54% of CD2-positive cells within 48 hours (Fig. 2D). This subset was equally distributed between CD4 and CD8 cells. Stimulation of interleukin-2 (IL-2)–dependent T cells with antibodies to CD3 and CD28 also induced expression of TACI. A reverse transcriptase–polymerase chain reaction assay revealed TACI message in resting B cells but not in T cells, unless they were activated (6).

Neither TACI mRNA nor protein could be detected in untransfected Jurkat cells expressing the SV40 large T-antigen (TAg), either unstimulated or treated with phorbol myristyl acetate (PMA) and ionomycin (6). To assess the effect of TACI on NF-AT activity in T cells, we transiently expressed the protein in TAg Jurkat cells along with a secreted alkaline phosphatase reporter driven by the NF-AT–binding sequences from the IL-2 promoter (12, 13,14). TACI overexpression could partially replace the requirement for PMA and ionomycin in this assay for maximal activation of the NF-AT reporter. The addition of antibodies to TACI to the cells increased NF-AT activation up to sevenfold (Fig.3A), demonstrating that TACI responds to cross-linking at the cell surface. This affinity-purified antibody to TACI had no effect on control transfected cells. To further verify the specificity of the response, we transfected cells with an NH2-terminal FLAG-epitope–tagged TACI expression plasmid and incubated them with the M2-FLAG monoclonal antibody (15). This treatment gave a similar increase in NF-AT activity (Fig. 3A). The degree of NF-AT activation varied among different experiments because of transfection efficiency, but was typically 40 to 100% of the maximal response to PMA plus ionomycin.

Figure 3

Activation of NF-AT, AP-1, and NFκB-specific transcription in cells with stimulated TACI. (A) Activation of an NF-AT–driven secreted alkaline phosphatase (SEAP) reporter. NF-AT-specific activity was determined in TAg Jurkat cells, cotransfected with the SXNFAT reporter (14) and the expression plasmid pBJ5 (24) containing no insert, cDNA encoding NH2-terminal FLAG-epitope-tagged TACI, or native TACI, in the absence or presence of cross-linked monoclonal antibody to FLAG or polyclonal antibodies to TACI, as indicated. Data were corrected for transfection efficiency, using the maximal response to PMA and ionomycin. Fold-activation ratios were then calculated as compared with the empty vector-transfected, unstimulated control value. (B) Activation of AP-1–specific transcription. TAg Jurkat cells were cotransfected with a mouse metalothionine AP-1–SEAP reporter plasmid (14) and pBJ5 containing no insert (−TACI, left) or TACI cDNA (+TACI, right). Cells were incubated with various amounts of antibodies to TACI. To control for transfection efficiency, we included a plasmid containing a constitutive promoter driving the expression of luciferase (EF-Luc). (C) Activation of an NFκB-specific transcription. The experiment was performed as in (B) except that an NFκB-specific SEAP reporter was used. (D) Inhibition of TACI-mediated activation of NF-AT by CsA. TAg Jurkat cells were transfected with a TACI expression plasmid and reporter plasmids specific for NF-AT, NFκB, or AP-1. Activation after treatment with antibodies to TACI was determined in absence (−) and presence (+) of CsA (100 ng/ml). (E) Requirement of extracellular Ca2+ for NF-AT activation in response to antibody cross-linked TACI. NF-AT activation in TAg Jurkat cells expressing TACI was measured in the presence of EGTA to remove extracellular calcium. The treatments included cross-linked antibody to TACI, transfection with a COOH-terminally truncated, calcium-independent calcineurin A subunit, or activation with the OKT3 antibody to the T cell receptor. All cells were also stimulated with PMA (50 ng/ml).

Activation of NF-AT in T cells requires the activation of both the calcium-dependent protein phosphatase calcineurin and the AP-1 transcription factor (16). Overexpression of CAML in TAg Jurkat cells activates calcineurin, and stimulation with PMA is required to activate NF-AT (12). NF-AT activation by TACI did not depend on additional stimulation with PMA, suggesting that AP-1 may also be activated upon TACI ligation. Reporter assays with the mouse metallothionine AP-1–binding sequence confirmed this (Fig. 3B). Furthermore, ligation of TACI also activated NFκB (Fig. 3C), a transcription factor implicated in the actions of other members of the TNFR superfamily (17). TACI-mediated activation of NF-AT depended on calcineurin, as demonstrated by the loss of NF-AT activity in the presence of immunosuppressive drugs such as cyclosporin A (CsA) (Fig. 3D) or FK506. Depletion of external calcium also blocked TACI-induced NF-AT activation (Fig. 3E).

We determined the regions of TACI and CAML required for their interaction, using deletion mutants with the yeast two-hybrid system (2). The COOH-terminal 126 amino acids of TACI were sufficient to bind to the NH2-terminal 201 amino acids of CAML (Fig.4A). Evidence for the association of TACI with CAML in vivo was provided by experiments in which full-length CAML or the 201 NH2-terminal amino acid residues of CAML coimmunoprecipitated with TACI from lysates of cells overexpressing both proteins (Fig. 4B). We conclude that the cytoplasmic COOH-terminal portion of TACI can physically associate with the NH2-terminal half of CAML.

Figure 4

Interaction of TACI with CAML and inhibition of NF-AT signaling in cells expressing truncated CAML. (A) Yeast two-hybrid interaction. Full-length cDNAs and indicated deletion mutants of TACI and CAML were cloned into the yeast expression plasmids pACT or pAS1, or both, and the indicated combinations were tested for interaction in the yeast two-hybrid system [+, positive interaction (25); −, no interaction; ND, not done]. (B) Coimmunoprecipitation of CAML with TACI. We transfected 293T-cells with the indicated combinations of the expression plasmid pBJ5 containing cDNAs for CAML, TACI with an NH2-terminal FLAG tag, the NH2-terminal 201 amino acids of CAML (CLX91), or no insert. After 48 hours of incubation, the cells were lysed [1% dodecyl maltoside, 20 mM Hepes (pH 7.4), 150 mM NaCl, 10% glycerol, 2 mM MgSO4, 1 mM CaCl2, and 1 mM phenylmethylsulfonyl fluoride], and the lysate was clarified by centrifugation. FLAG-tagged TACI and associated proteins were immunoprecipitated with monoclonal antibody to FLAG conjugated to agarose beads and subjected to protein immunoblotting. The blot was probed with immunoaffinity-purified polyclonal antibodies to CAML (10) followed by chemiluminescent detection (Amersham). Parallel protein immunoblots of each sample confirmed the expected expression of TACI, CAML, or the truncated CAML mutant in all transfections. (C) Inhibition of TACI-induced NF-AT activation in cells overexpressing the NH2-terminal half of CAML. NF-AT activation in cells treated with an antibody to TACI was determined in TAg Jurkat cells transfected with pBJ5 alone, pBJ5-TACI plus a vector control, or pBJ5-TACI with an equivalent amount of CLX91 (top). CLX91 did not alter levels of TACI expression in transfections, as determined by protein immunoblotting with polyclonal antibodies to TACI (bottom). (D) Specificity of the dominant-negative effect of the truncated CAML mutant. Antibody-induced TACI activation of NF-AT, AP-1, and NFκB reporter activities were determined by transient transfection in TAg Jurkat cells. Cells were cotransfected with CLX91 or vector controls in the presence of pBJ5-TACI and reporter plasmids. Data were normalized against values obtained for stimulation with PMA and ionomycin.

CAML is a widely expressed integral membrane protein localized to cytoplasmic vesicles (12). Hydrophobic domains in the COOH-terminal half of the protein are essential for activity, and the hydrophilic NH2-terminal half of the protein may have a regulatory role (18). The NH2-terminal half of the molecule is exposed on the cytoplasmic surface of the vesicle (19). To test whether TACI might signal by directly interacting with CAML, we investigated the possibility that the interacting domain of CAML (residues 1 to 201) could inhibit TACI-induced activation of a transcription factor. Cotransfection of the mutant CAML(1-201) expression plasmid CLX91 inhibited TACI-induced activation of NF-AT by more than 80% (Fig. 4C). There was no inhibitory effect on NF-AT activity induced by PMA and ionomycin, thus ruling out a nonspecific toxic effect of the mutant CAML protein. Coexpression of CAML(1-201) also did not reduce the accumulation of TACI protein, as detected by protein immunoblotting (Fig. 4C), nor did flow cytometric comparison of cells transfected with TACI and CAML(1-201) expression plasmids or with the TACI plasmid and empty vector reveal any differences in surface accumulation of TACI (6). The CAML mutant had little effect on the activation of AP-1 or NFκB in response to ligation of TACI (Fig. 4D).

Overexpression of CAML in T cells contributes to calcium-dependent constitutive activation of NF-AT. We identified a lymphocyte-specific receptor with characteristics of the TNFR superfamily that can activate NFκB, NF-AT, and AP-1. The signal from TACI to NF-AT activation may proceed through direct interaction between TACI, at the cell surface, and CAML, an integral membrane protein located at intracellular vesicles.

  • * To whom correspondence should be addressed. E-mail: richard.bram{at}stjude.org

REFERENCES AND NOTES

View Abstract

Navigate This Article