Tumor Necrosis Factor (TNF) Receptor Superfamily Member TACI Is a High Affinity Receptor for TNF Family Members APRIL and BLyS*

An expression cloning approach was employed to identify the receptor for B-lymphocyte stimulator (BLyS) and identified the tumor necrosis factor receptor superfamily member TACI as a BLyS-binding protein. Expression of TACI in HEK293T cells confers on the cells the ability to bind BLyS with subnanomolar affinity. Furthermore, a TACI-Fc fusion protein recognizes both the cleaved, soluble form of BLyS as well as the membrane BLyS present on the cell surface of a recombinant cell line. TACI mRNA is found predominantly in B-cells and correlates with BLyS binding in a panel of B-cell lines. We also demonstrate that TACI interacts with nanomolar affinity with the BLyS-related tumor necrosis factor homologue APRIL for which no clear in vivo role has been described. BLyS and APRIL are capable of signaling through TACI to mediate NF-κB responses in HEK293 cells. We conclude that TACI is a receptor for BLyS and APRIL and discuss the implications for B-cell biology.

Members of the tumor necrosis factor superfamily of cytokines play diverse roles in the regulation of cell proliferation, differentiation, and survival. Notably, several members of this family play key roles in the regulation of the immune system (1). We and others have previously identified a novel TNF 1related ligand, BLyS (also known as BAFF, TALL-1, THANK, TNFSF20, and zTNF4) which is expressed on monocytes and induces B-cell proliferation and immunoglobulin secretion in vitro and in vivo (2)(3)(4)(5)(6). Like many members of the TNF family, BLyS has activity in vitro as a 152-amino acid soluble molecule and as a 258-amino acid transmembrane form (3). However, the biological significance of these two forms and their relative contributions in vivo remain to be resolved. More recently, transgenic mice that ectopically overexpress BLyS were shown to develop autoimmune-like phenotypes reminiscent of those observed in systemic lupus erythematosus (7)(8)(9). These find-ings suggest that BLyS plays an important role in the regulation of B-cell growth and humoral immunity.
In order to understand the precise mechanism by which BLyS activates B-cells, the range of cell types BLyS may affect, and the potential role of BLyS as a therapeutic agent or target, we have used expression cloning to identify the receptor for BLyS. We have identified the orphan receptor TACI (10), previously characterized as being present on B-cells and a subset of T-cells, as the receptor for BLyS and show that this receptor is capable of mediating NF-B signaling in response to ligand binding. We also show that TACI interacts with another TNF family member, APRIL, which is closely related to BLyS. Parallel work by others has recently shown that TACI and a second TNFR family member, BCMA, are BLyS receptors (9,(11)(12)(13)(14).

EXPERIMENTAL PROCEDURES
Cell Culture and Media-HEK293T cells were cultured in Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal bovine serum and transfected using LipofectAMINE Plus (Life Technologies, Gaithersburg, MD) according to the manufacturer's protocol. For expression cloning screens, cells were attached to plates with poly-D-lysine.
Flow Cytometry-Cells were stained with monoclonal antibodies raised against BLyS at the indicated protein concentrations, with biotinylated BLyS as described previously (2), with recombinant TACI-Fc fusion protein or with recombinant Flag-tagged proteins which were subsequently detected by the M2 anti-Flag monoclonal antibody (Sigma). Flow cytometry was performed using a FACScan instrument and associated CellQuest software (Becton Dickinson, San Jose, CA).
Library Preparation, Screening, and Other DNA Manipulations-All common DNA manipulations such as restriction, ligation, and PCR were as described previously (15). Human tonsillar B-cell cDNA, sizeselected to enrich for potential cDNA clones of Ͼ1.5 kilobase, was ligated into expression vector pCMV-Sport3 (Life Technologies) to generate a library of approximately 5 million independent clones. To facilitate library screening, wells containing 150 clones in a 96-well format were generated, and plasmid DNA was purified from these using the Qiagen biorobot 9600 (Qiagen, Valencia, CA). Pools of approximately 1200 cDNAs (8 wells) were used to transfect HEK293 cells. At 40 -48 h post-transfection, cells were washed once with PBS and incubated for 2 h at 37°C with 300 pM 125 I-radiolabeled BLyS (17-35 Ci/g) in binding buffer (DMEM, 10% fetal bovine serum, 25 mM HEPES, pH 7.4). Subsequently, cells were washed three times with PBS, and fixed with glutaraldehyde (2.5% v/v). The bottoms of the plates were removed as described previously (16) and subjected to autoradiography. Pools that bound radiolabeled BLyS were subsequently partitioned into the individual eight wells of 150 clones that composed the initial positive pool and used for transfection. 600 clones from positive wells of 150 clones were end-sequenced. Full-length TACI was generated by a twostep PCR reaction using the primers ATGAGTGGCCTGGGCCGGAG-CAGGCGAGGTGGCCGGAGCCGTGTGGACCAGG and AAGCTTAG-ATCTGCCACCATGAGTGGCCTGGGCCGGAGC sequentially at the 5Ј end together with the reverse primer GAATTCTCTAGACCCCCATTT-ATGCACCTGG. The PCR product was cut with BglII and XbaI and * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
subcloned to the cytomegalovirus-based expression vector pC4 (Human Genome Sciences, Inc.) Full-length TACI-Fc construction was created by two-step PCR using the following 5Ј primers to sequentially add on the signal sequence of MPIF: CCCTCTCCTGCCTCATGCTTGTTACT-GCCCTTGGATCTCAGGCCATGAGTGGCCTGGGCCGGAGCAGG and GAATTCAGATCTGCCACCATGAAGGTCTCCGTGGCTGCCCTC-TCCTGCCTCATGCTTGTTACTGCC and the 3Ј primer AAGCTTTCT-AGACTGATCTGCACTCAGCTTCAGCCC. Truncated TACI(M31-Q159)-Fc was generated similarly but using the 5Ј primer CTGCCTC-ATGCTTGTTACTGCCCTTGGATCTCAGGCCATGAGATCCTGCCCC-GAAGAGCAG to fuse the 3Ј region of the MPIF signal sequence to TACI. The PCR products were cut with BglII and XbaI and ligated into BamHI-XbaI linearized expression vector pC4-Fc (Human Genome Sciences Inc). Full-length BLyS cDNA was subcloned into the pCDNA3.0 vector (Invitrogen, Carlsbad, CA) and transfected into HEK293F cells (Life Technologies) using LipofectAMINE Plus. Stable transfectants were selected that were resistant to genticin and clones that demonstrated cell surface binding to BLyS antibodies by FACS were used for further study. RANK (17) was also amplified by PCR and subcloned into vector pC4. All constructs were confirmed by DNA sequencing. To monitor NF-B activation, a reporter plasmid that contains 4 tandem repeats of a consensus NF-B-binding site (GGGACTTTCCC) upstream of the pro-SEAP reporter plasmid (CLONTECH, Palo Alto, CA) was employed. HEK293T (1 ϫ 10 5 cells) were transiently co-transfected with the reporter plasmid (0.5 g/ml) together with the indicated amounts of either expression vector alone or the expression vector containing TACI. Ligand was provided either by the addition of recombinant protein or by co-transfection of expression plasmids for soluble BLyS or soluble APRIL. In all cases the total plasmid DNA concentration was adjusted to 2 g/ml with empty vector and transfections performed in duplicate. At 18 h post-transfection, supernatants were collected and SEAP levels determined following the manufacturer's recommendations (Roche Bioscience, Indianopolis, IN) and counted in a Dynax luminometer.
TaqMan Analysis-TACI messenger RNA levels were determined by real time quantitative PCR using an ABI 7700 Taqman sequence detector. Amplification primers and probe were designed to span the region from nucleotide 308 to nucleotide 392 of the human TACI mRNA (Genbank accession number AF023614). Total RNA was prepared from B-cells, primary hematopoietic cells, and B-cell lines and mRNA detected by a one step reverse transcriptase-PCR procedure. For quantitation of TACI mRNA the comparative delta Ct method was used (PerkinElmer Life Sciences user bulletin number 2, 1997) using a 18 S ribosomal RNA probe as endogenous reference. Expression levels are shown relative to expression levels in B-cells.
Protein Purification-BLyS was purified as described previously (2). APRIL was purified by capture on cation exchange resin (HS-50 Poros) and the trimeric form of APRIL isolated by size exclusion chromatography. Full-length TACI(M1-Q159)-Fc or truncated TACI(M31-Q159)-Fc were purified from transiently transfected culture supernatants by capture on Protein A HyperD resin (Life Technologies) and eluted with 0.1 M citrate buffer, pH 3.5. To further enrich for the dimeric TACI-Fc fusion protein the pool was subject to size-exclusion chromatography on a Superdex S200 column (Amersham Pharmacia Biotech, Piscataway, NJ). Proteins were subjected to NH 2 -terminal sequence analysis to verify their NH 2 -terminal sequence using a model ABI-494 sequencer (PerkinElmer Life Sciences). Immunoprecipitations were carried out using the M2 anti-Flag peptide antibody (Sigma). Western blots were decorated with the M1 anti-Flag peptide antibody (Sigma) and detected using an anti-mouse IgG alkaline phosphatase conjugate (Promega, Madison, WI).
BIAcore Analysis-Ligand binding to TACI-Fc was also analyzed by BIAcore analysis. TACI-Fc was covalently immobilized to the BIAcore sensor chip (CM5 chip) via amine groups using N-ethyl-NЈ-(dimethylaminopropyl)carbodiimide/N-hydroxysuccinimide chemistry. A total volume of 50 l of various dilutions of BLyS or APRIL were allowed to flow over the receptor-derivatized flow cells at 15 l/min. The amount of bound protein was determined during washing of the flow cell with HBS buffer (10 mM HEPES, pH 7.4, 150 mM NaCl, 3.4 mM EDTA, 0.005% Surfactant P20). Binding specificity of BLyS or APRIL to TACI was determined by competition with soluble competitor in the presence of 5 g/ml BLyS or APRIL. The flow cell surface was regenerated by displacing bound protein by washing with 20 l of 10 mM glycine-HCl, pH 2.3. For kinetic analysis, the flow cells were tested at different flow rates, and the on-and off-rates were determined using the kinetic evaluation program in BIAevaluation 3 software, using a 1:1 binding model and the global analysis method.
Preparation of Radiolabeled BlyS and Binding Assays-Radioiodination of BLyS was performed using the Iodo-Bead method. Briefly, one Iodo-Bead (Pierce Chemical Co., Rockford, IL) per reaction was pre-washed with PBS and added to 1 mCi of Na 125 I in 80 l of PBS, pH 6.5. The reaction was allowed to proceed for 5 min and then 10 g of BLyS was added and incubated for 5 min at room temperature. Iodinated protein was separated from unbound radioactivity using a G-25 Sephadex quick-spin column previously equilibrated with PBS containing 0.1% bovine serum albumin. Protein concentration and specific radioactivity of 125 I-BLyS were determined by trichloroacetic acid precipitation of pre-column and post-column samples. The specific activity of 125 I-labeled BlyS used in the experiment was 17-35 Ci/g. Binding of 125 I-labeled BLyS to TACI was performed with 1 ϫ 10 6 vector alone or with TACI-transfected HEK293T cells in 100 l of binding buffer containing 0.3 nM 125 I-BLyS (specific activity 34.6 Ci/g), in the absence or presence of increasing concentrations of unlabeled BLyS. The incubation was allowed to proceed for 2 h at room temperature. The cellbound BlyS was separated from unbound free 125 I-BlyS by centrifugation through 200 l of dibutylphthalate/phthalic acid bis(2-ethylhexyl) oil mixture in polyethylene microfuge tubes (Bio-Rad) for 20 s at 12,000 rpm. The tubes were frozen quickly in liquid nitrogen and the bottoms of the tubes were cut off using a tube cutter. Radioactivity in the bottoms of the tubes containing the cell pellet (bound) and in the top (unbound) was determined with a ␥-counter. The saturation binding analysis was performed under similar conditions using varying concentrations of 125 I-labeled BLyS (0.01 to 9 nM) in the absence or presence of 100-fold excess of unlabeled BLyS. The data was analyzed by Prizm software (GraphPad Software, San Diego, CA) to determine dissociation constant (K d ) and number of binding sites.

RESULTS
To identify the BLyS receptor, we developed an expression cloning system using biologically active, radioiodinated BLyS. The labeled BLyS was used to screen an expression library that was generated from human tonsillar B cells previously demonstrated to bind BLyS and proliferate in response to costimulation with BLyS and Staphylococcus aureus Cowan preparation. Library pools of approximately 1200 cDNAs were screened. One pool from 96 screened resulted in reproducible binding of radiolabeled BLyS above background levels. This pool was partitioned into subpools of 150 clones and a positive subpool identified. Clones from a positive subpool of 150 clones were subject to end-sequencing using the HGS high-capacity sequencing facility, and sequences screened for homology to known TNFRs using a Hidden Markow model based on the cysteine-rich motif present in this family (18). This approach identified one clone that contained such a motif. DNA from this single clone was transfected into HEK293T cells and resulted in specific binding of BLyS to the cells that was competed with excess cold ligand (not shown). The insert of the positive clone was sequenced in full to reveal the presence of a 263-amino acid open reading frame which, when compared with known TN-FRs, identified this clone as an amino-terminal truncation of the previously characterized TNFR superfamily member TACI (10). Translation of the TACI protein from the isolated clone must initiate at a downstream ATG (Met-31 in the previously published sequence) to encode a protein which was still capable of being transported to the cell surface and binding ligand. The clone contains both of the cysteine-rich domains which are important for TNF-ligand binding (19). We created a fulllength TACI clone by PCR and used this for further analysis.
BLyS Binds TACI with High Affinity-In order to show that the BLyS-TACI interaction is physiologically relevant we determined the specificity and affinity of BLyS binding to TACI. Saturation binding analysis was performed using various concentrations of 125 I-labeled BLyS on HEK293T cells that had been transfected with full-length TACI. Analysis of binding data revealed that BLyS binds to TACI with high affinity with a K d of 0.1-0.3 nM (Fig. 1A). This is similar to the affinity of BLyS for its receptor on tonsillar B-cells and other B-cell lines (20). The affinity of the truncated TACI was 5-10 lower when compared with the full-length protein as determined by competitive binding assays using various concentrations of unlabeled BLyS (Fig. 1B).

TACI Mediates Binding to Soluble and Membrane
BLyS-We confirmed binding of BLyS to HEK293T transfected with TACI by FACS analysis using biotinylated BLyS; transient transfection of TACI results in a population of cells which bind BLyS in contrast to vector-transfected HEK293T cells. (Fig. 2A). The TNF-related ligand LIGHT/HVEM-L (21,22) failed to bind to the TACI-transfected cells (Fig. 2C). To further explore the TACI-BLyS interaction we created a TACI-Fc fusion protein in which the extracellular domain of TACI is fused to the Fc domain of human IgG 1 . This fusion protein was used to effectively compete for soluble biotinylated BLyS binding to the B-cell-derived line IM-9 (Fig. 3). Because TACI-transfected cells bind soluble biotinylated and radiolabeled BLyS and TACI-Fc competes with IM-9 cells for binding of soluble BLyS, we conclude that TACI is able to interact with the soluble form of BLyS.
To determine if TACI is capable of binding membrane-bound BLyS, a recombinant cell line that expresses cell surface BLyS was generated. HEK293F cells were stably transfected with BLyS and demonstrated to express cell surface BLyS by FACS analysis using a monoclonal antibody derived against BLyS (Fig. 4, A-D). These cells were then tested for their ability to bind TACI-Fc protein using FACS analysis. As demonstrated in Fig. 4, TACI-Fc specifically binds the HEK293-BLyS stable cell line, but not a HEK293-vector cell line. Furthermore, the binding to the HEK293-BLyS cells is inhibited by the addition of soluble BLyS into the binding reaction (not shown). We conclude that TACI is able to bind membrane-bound BLyS.
Binding Specificity-TACI-Fc fusion protein was employed to assess the specificity of the interaction between TACI, BLyS, and other members of the TNF ligand family. We used a panel of 4 conditioned media containing Flag-epitope tagged proteins APRIL (23), LIGHT (21), FasL (24), and BLyS to assess interaction specificity. To evaluate the level of Flag-tagged protein present in the conditioned media, we used anti-Flag antibodies to immunoprecipitate the tagged proteins (Fig. 5A). An equivalent aliquot of the conditioned media was also subject to immunoprecipitation with TACI-Fc (Fig. 5B) or with beads alone (not shown). Immunoprecipitates were detected by Western analysis using anti-Flag antibody. We find that TACI-Fc effectively immunoprecipitates all of the Flag-BLyS present in the conditioned medium (Fig. 5, A and B, lane 3) but does not precipitate either LIGHT or FasL. We also find that TACI-Fc immunoprecipitates approximately 20% of the APRIL present in the conditioned medium, suggesting that TACI-Fc interacts with APRIL (Fig. 5B, lane 1). Interaction between APRIL and TACI-Fc has also been observed with BIAcore analysis (next section).
BIAcore Analysis-We assessed the ability of TACI-Fc to bind BLyS by BIAcore analysis. TACI-Fc was bound to a BIAcore chip and different concentrations of human BLyS were allowed to flow over the cell. There was significant binding to the flow cell of ϳ1000 and 600 relative units at 5 and 2.5 g/ml BLyS, respectively. This binding was specific because BLyS binding to TACI-Fc was competed with increasing concentrations of soluble TACI-Fc (data not shown). Based on the immunoprecipitation data, we also tested the TNF-related ligand, APRIL, for binding to TACI-Fc in BIAcore. APRIL was also found to bind to the TACI-Fc chip and this interaction was also shown to be specific (data not shown). The K d for binding to BLyS and APRIL were determined by analysis of the binding of different concentrations of ligand using the BIAcore biosensor instrument (Fig. 6). The interaction between TACI and APRIL was weaker than that observed for TACI and BLyS; representative sensorgrams for the binding of BLyS and APRIL to TACI-Fc are shown in Fig. 6, A and B, respectively. The association constant, k a , for binding was 6.45 ϫ 10 7 and 9.22 ϫ 10 5 , for BLyS and APRIL, respectively. The dissociation rate, k d , for TACI binding was 1.04 ϫ 10 Ϫ2 and 5.89 ϫ 10 Ϫ3 for BLyS and APRIL, respectively. Thus, the on-rate is faster for BLyS compared with APRIL, whereas, the off-rate for BLyS is faster than that of APRIL. The calculated K d values for the binding of APRIL and BLyS to TACI-Fc were 6.4 and 0.16 nM, respectively. Overall BLyS has a ϳ25-fold higher binding constant than APRIL.
In order to show that the interaction of TACI with APRIL was not an artifactual result of using a TACI-Fc fusion protein, we determined if TACI present on the surface of cells could bind APRIL. We find that TACI-transfected cells are capable of binding Flag-tagged-APRIL or BLyS but not Flag-tagged Fas ligand (Fig. 7, A-F). We also find that APRIL competes for radiolabeled BLyS binding to TACI-transfected cells (Fig. 7G). In agreement with the BIAcore analysis, the relative affinity of TACI for APRIL was again found to be some 10 -20-fold lower than for BLyS. We also observe interaction of TACI-Fc with membrane-bound APRIL on cells transiently transfected with a full-length APRIL construct (not shown).
TACI Expression Correlates with BLyS Binding Capacity-TACI was previously characterized as being present on B-cells and a subset of activated T-cells (10). We extended this analysis initially by Northern analysis on a series of tissues. In agreement with previous observations, we find that TACI is expressed primarily in immune tissues with weaker signals pres- ent in gastrointestinal tissues. Notably, the highest message is present in the B-cell lymphoma cell line RAJI (not shown). We have also refined this analysis by studying TACI mRNA expression using quantitative, real-time PCR (Taqman). In agreement with the above findings, we show that TACI is predominantly expressed on B-cells and B-cell lines; there is a much weaker signal present in cells of the monocytic and T-cell lineage. Furthermore, the weak expression of TACI in T-cells and dendritic cells is unaltered by known modulators of the immune system (Fig. 8A). We and others (2,3) have previously defined the functional presence of BLyS receptor on different cell types and show that among B-cell lines Daudi cells do not effectively bind BLyS. We, therefore, measured TACI expression among different B-cell lines using Taqman analysis and asked if TACI expression correlates with BLyS binding. We find that among B-cell lines Daudi has little or no TACI mRNA expression, while other B-cell lines have good TACI expression (Fig. 8B). Thus TACI expression among B-cell lines correlates with BLyS binding capacity.
Previous work has suggested that TACI is present on the surface of activated T-cells. As we show here that TACI functions as a BLyS receptor we analyzed the capacity of T-cells activated by ionomycin and phorbol 12-myristate 13-acetate to bind biotinylated BLyS. We find that activated T-cells only weakly bind biotinylated BLyS (not shown).
TACI Mediates Signal Transduction in HEK293 Cells-It has previously been demonstrated that several TNF receptor family members mediate NF-B signaling, and that when overexpressed the receptors activate the NF-B pathway independent of ligand binding (25). To determine if TACI is capable of activating the NF-B pathway in HEK293 cells, TACI was transiently overexpressed in this cell type together with an NF-B-SEAP reporter plasmid. As demonstrated in Fig. 9, TACI alone is capable of activating the NF-B signaling pathway. The NF-B response is dependent upon the amount of TACI transfected into the cells. We also show that addition of BLyS or APRIL into this system by co-transfection of a ligand expression plasmid results in significantly increased stimulation of the NF-B reporter (Fig. 9). This demonstrates that both BLyS and APRIL can associate with, and direct TACI mediated signaling.

DISCUSSION
Using an expression cloning protocol we have identified the previously described TNFR family member TACI as a BLySbinding protein that is present predominantly on B-cells. We present several lines of evidence to suggest that TACI is the BLyS receptor; however, the high affinity of the interaction between BLyS and TACI (K d of 0.1 nM), which is typical of physiological receptor-ligand interactions, is probably the most compelling. Additionally, there is good concordance between the TACI expression profile and the BLyS-binding potential of cells. Notably, among different B-cell lines, Daudi cells (which do not bind BLyS) do not express TACI. Significantly, we show that TACI is able to interact with both soluble BLyS and the membrane-bound form of BLyS present on a recombinant cell line, an observation that suggests that TACI is able to mediate signal transduction of both forms of BLyS to B-cells. Previous work has shown that BLyS present on the surface of cells is competent to act as a co-stimulator of B-cells (3); we show here that this activity is probably mediated, at least in part, by TACI.
In parallel work, others have recently shown that TACI as well as another TNFR family member, BCMA (26), act as BLyS-binding proteins (9,(11)(12)(13)(14). Importantly, we show here that the interaction between TACI and BLyS is not exclusive and that TACI is capable of interacting with the TNF-related ligand APRIL, as judged by four independent in vitro methods. The interaction between TACI and APRIL was found to be some 10 -20-fold lower in affinity than that seen for the BLyS-TACI interaction. However, the affinity was still in the nanomolar range (6 nM) and is similar to affinities of other receptorligand interactions in the TNF/TNFR superfamily. The interaction of a TNFR superfamily member with multiple TNF- like ligands is not unprecedented; the receptor HVEM was shown to interact with the ligands LIGHT and lymphotoxin-␣ (21). The in vivo role of APRIL remains elusive; a role as a growth promoting factor has been shown, and a role in tumor cell growth has been suggested (23), while other work has suggested a role in apoptosis (27). However, the expression of APRIL in peripheral blood lymphocytes (23), monocytes, and macrophages (TALL-2 (4)) and its interaction with TACI demonstrated here, make a role in regulation of immune function seem likely.
It has been suggested that TACI, as soluble receptor, may prove useful as a therapeutic agent to antagonize the function of BLyS in autoimmune diseases such as systemic lupus erythematosus (28). Indeed, it has been shown that administration of TACI-Fc soluble receptor results in reduction of proteinuria and prolongation of lifespan in an animal model of systemic lupus erythematosus (9) or inhibition of primary immune responses and germinal center disruption in normal mice (13,11). The interaction between APRIL and TACI demonstrated here has implications for any potential therapeutic utility based on soluble TACI receptor. Since we show that TACI is not specific for BLyS but also binds with high affinity to APRIL, TACI-based therapeutics will almost certainly also antagonize the function of APRIL. The outcome of the previous in vivo studies with TACI-Fc could be the result of TACI-Fc blocking APRIL-mediated B-cell effects, or BLyS-and APRILmediated effects, rather than simply BLyS-mediated effects. While this manuscript was in revision, parallel work was published by others that also demonstrates interaction of TACI and APRIL (29); furthermore, this paper also demonstrates the interaction between APRIL and the alternate BLyS receptor BCMA. Clearly, further work will be required to unravel the relative contributions of BLyS, APRIL, TACI, and BCMA interactions in B-cell regulation.
We show here that TACI mRNA expression profile is predominantly B-cell specific. In contrast, TACI was initially characterized as being present in B-cells and activated T-cells (10). Other cells, such as dendritic cells and monocytes, have much lower levels of TACI. Furthermore, TACI expression is not inducible in these cells by immunomodulatory agents, such as interferon ␥ and interleukin-10. In agreement with the previous report on TACI, we observe weak expression of TACI on T-cells. However, at the mRNA level this expression is at least an order of magnitude lower than that seen on B-cells. In agreement with these findings, we are able to observe weak binding of BLyS to a purified population of stimulated T-cells. This binding, however, is less than 5% of the binding observed on B-cells (not shown). Thus if BLyS plays a role in T-cell regulation as suggested by others (5), it seems unlikely that this role will be a major one. However, we cannot exclude that under normal physiological conditions the interaction of BLyS with T-cells is required for some form of cross-talk between these two compartments of the immune system.
The clone we originally isolated as the BLyS-binding protein represented an amino-terminal truncation of the published TACI sequence. Since TACI is a Type 3 transmembrane protein, it contains no amino-terminal signal sequence; an aminoterminal truncation will not, therefore, interfere with the topological signals present in the internal signal anchor, and the protein will be transported to the cell surface. Surprisingly, this truncated version of TACI is biologically active and has an affinity which is only 5-fold lower than compared with fulllength TACI. This bioactive deletion may prove useful in delineating structure-function relationships between TACI and BLyS or APRIL binding. It may also represent some fraction of the naturally translated protein that is found in vivo.
TACI was initially identified as an orphan receptor in a two-hybrid screen using the signal transduction component CAML as bait and was shown to mediate activation of NF-B, NFAT, and AP-1 in the Jurkat T-cell line (10). We confirm here that TACI is capable of activating the NF-B pathway in HEK293 cells. At the highest levels of TACI transfected, the level of NF-B activation by TACI is comparable to levels observed with RANK (not shown) (17,30), a TNF receptor family member known to activate NF-B suggesting that TACI is a strong activator of the NF-B pathway. Furthermore, this stimulation is augmented by the addition of either BLyS or APRIL. Recent analysis of the TACI intracellular domain has identified TRAF2, TRAF5, and TRAF6 specifically interacting with regions of TACI in a yeast 2-hybrid system and in unstimulated HEK293 cells expressing epitope-tagged components (11). Further work will be required in order to delineate which TRAFs are specifically recruited to TACI following ligand stimulation.