Elsevier

Seminars in Immunology

Volume 18, Issue 5, October 2006, Pages 263-275
Seminars in Immunology

Review
BAFF, APRIL and their receptors: Structure, function and signaling

https://doi.org/10.1016/j.smim.2006.04.006Get rights and content

Abstract

BAFF, APRIL and their receptors play important immunological roles, especially in the B cell arm of the immune system. A number of splice isoforms have been described for both ligands and receptors in this subfamily, some of which are conserved between mouse and human, while others are species-specific. Structural and mutational analyses have revealed key determinants of receptor–ligand specificity. BAFF-R has a strong selectivity for BAFF; BCMA has a higher affinity for APRIL than for BAFF, while TACI binds both ligands equally well. The molecular signaling events downstream of BAFF-R, BCMA and TACI are still incompletely characterized. Survival appears to be mediated by upregulation of Bcl-2 family members through NF-κB activation, degradation of the pro-apototic Bim protein, and control of subcellular localization of PCKδ. Very little is known about other signaling events associated with receptor engagement by BAFF and APRIL that lead for example to B cell activation or to CD40L-independent Ig switch.

Introduction

The TNF family ligands APRIL (CD256, TNFSF13) [1] and BAFF (also known as BLyS, TALL-1, CD257 and TNFSF13B) [2], [3], [4] interact with three TNFR family members, TACI (CD267, TNFRSF13B) [5], BAFF-R (also known as BR3, CD268 or TNFRSF17) [6], [7] and BCMA (CD269, TNFRSF13C) [8]. APRIL and TACI also bind independently to glycosaminoglycan structures such as those present in syndecan-1 (CD138), or other proteoglycans [9], [10], [11].

The interaction pattern between BAFF, APRIL and their receptors is both specific and redundant: BAFF-R binds to BAFF, BCMA binds to APRIL, and TACI binds to BAFF and APRIL. In addition, BCMA binds BAFF with weaker affinity. Despite this complexity, the function of individual members of this subfamily in the immune system is relatively well understood, in great part as a result of the generation of a panel of mouse mutants. Recent papers have reviewed these functions, e.g. [12], [13], [14]. The present review seeks to provide insight into three aspects of BAFF biology: description of the numerous splice variants described for the different members of the BAFF complex; binding specificities of BAFF-R, BCMA and TACI and the structural determinants underlying ligand selectivity; and a summary overview of the fragmentary information available concerning the signaling pathways engaged by BAFF and APRIL, in the context of their known functions.

The TNF family ligand BAFF is a Type II membrane-bound protein, which can be released as a soluble trimeric ligand upon proteolytic processing at a furin consensus site. In the human BAFF gene, exon 1 codes for the transmembrane domain and its flanking regions, exon 2 for the furin processing site, and exons 3–6 for the TNF homology domain (THD), which binds to receptors (Fig. 1, Fig. 2).

At neutral or basic pH, 20 trimers of soluble recombinant human BAFF associate into a 60-mer virus-like structure, which irreversibly dissociates into trimers at acidic pH, or when fused to N-terminal extensions such as a myc tag [15], [16] (Fig. 3C). This association is dependent on an extended loop, known as the “Flap”, that is unique to BAFF in the TNF family [15] (Fig. 3A and B). The physiological importance of the BAFF 60-mer is unclear, but it is a biologically active entity that can bind receptors and is moderately more active than trimers in the in vitro assays [15], [17]. Endogenously produced 60-mer BAFF has recently been detected in supernatants of an histiocytic cell line [17].

The gene for mouse BAFF contains an additional exon encoding a stretch of 30 amino acids located between the furin site and the THD (Fig. 1A). This extension is predicted to prevent 60-mer formation in the mouse, although this has not been formally demonstrated. In both species, alternative splice variants, in which exon 3 (exon 4 in the mouse) is skipped, generate an in-frame deletion of the first β-sheet of the THD to produce a splice variant called ΔBAFF (Fig. 1, Fig. 2, Fig. 3). The structural impact of this deletion is unknown, but does not seem to prevent surface expression and heteromerization of ΔBAFF with BAFF, yet appears to prevent release of the soluble form [18]. Transgenic mice expressing ΔBAFF in myeloid and dendritic cells display reduced B cell number and impaired T-dependent humoral responses, consistent with ΔBAFF being a dominant-negative inhibitor of BAFF [19]. The splicing event in mouse ΔBAFF creates a functional N-linked glycosylation site [18], which is interesting in view of the fact that gain of glycosylation can considerably affect the biological activity of extracellular proteins [20].

Failure to splice intron 1 in human BAFF [18], or alternative splicing to an acceptor site in intron 3 are relatively frequent events, but result in the production of no or prematurely terminated BAFF protein (Fig. 1, Fig. 2).

The architecture of the APRIL gene resembles that of BAFF (Fig. 1). However, APRIL does not form 60-mers, but possesses residues close to the furin processing site in exon 3 that are crucial for binding to glycosaminoglycans (Fig. 2). APRIL binds sulfated glycosaminoglycans at sites independent from those used to bind other receptors [9], [10]. The relevance of this binding is unclear, but may serve to accumulate and/or multimerize APRIL in the extracellular matrix or at the surface of syndecan-positive cells. It thus may facilitate access to its receptor TACI, which also interacts with syndecans [11], or to intracellular BCMA [21] upon syndecan internalization.

A rare splicing event combines exon 1 of human APRIL to an alternative acceptor site in exon 3, thus generating a predicted membrane-bound, uncleavable human APRIL with no binding to glycosaminoglycans (APRIL-δ) (Fig. 1, Fig. 2). Omission of exon 3 yields APRIL-β, an homologue of ΔBAFF that also lacks the first β-sheet of the THD [22], and by analogy with ΔBAFF may regulate APRIL activity in a dominant-negative manner [18] (Fig. 2, Fig. 4). APRIL-γ is generated by splicing of a cryptic intron in exon 6, resulting in a four amino acids C-terminal truncation that is replaced by a single residue, but this isoform has not been further studied [22] (Fig. 2, Fig. 4). No similar sequences have been reported or detected in ESTs of murine APRIL. Mouse APRIL does, however, exist as two variants differing by a single amino acid (Ala120) which are found at similar frequencies. This subtle sequence difference originates from alternative use of two splice acceptor sites only three nucleotides apart at the beginning of exon 4. A similar splicing event was previously reported in another TNF family member, EDA, whose variants differ by two amino acid residues only, yet display distinct receptor specificities [23]. In the case of mouse APRIL, both variants bind TACI and BCMA, and the only difference in receptor specificity was observed in the weak, but detectable, binding of the shorter APRIL variant to mouse BAFF-R (unpublished data) (Fig. 2B). This weak binding is, however, only observed in the mouse system and is unlikely to be physiologically relevant.

In both human and mouse genomes, the APRIL gene is located immediately 3′ of TWEAK, another TNF family ligand. An intergenic splicing event between exon 6 of TWEAK and exon 2 of APRIL generates human TWE-PRIL (Fig. 1A). Although this mRNA has been convincingly shown to exist at abundances close to that of APRIL in T cells and various cell lines [24], it is intriguing that there is currently no EST sequence in the NCBI database harboring this junction, whereas ESTs coding for regular APRIL (in the corresponding region) are abundant (Fig. 2A). However, mouse TWE-PRIL ESTs are present as two variants, with or without Ala120 (Fig. 2B). The genesis of mouse TWE-PRIL is very different from that of its human orthologue, as it uses unique splice sites within exon 7 of TWEAK and within exon 1 of APRIL (Fig. 1A). Despite this difference, both human and mouse TWE-PRIL contain the entire THD of APRIL, and are therefore expected to display the same receptor specificity as APRIL. Although TWE-PRIL has two furin consensus cleavage sites, the overexpressed form of human TWE-PRIL was resistant to cleavage, as was ΔBAFF [18], [24].

BAFF-R and BCMA (and TACI) lack a signal peptide and are therefore classified as Type III membrane proteins. Exon 1 encodes the ligand-binding domain (also called cysteine-rich domain or CRD), exon 2 the transmembrane domain and flanking regions, and exon 3 the intracellular domain (Fig. 1B). The intracellular domains of both BAFF-R and BCMA are strongly homologous over a short sequence of 18 amino acids, which in the case of BCMA contains a TRAF-binding consensus site (P/S/A/T-X-Q/E-E) (Fig. 2). BCMA is indeed known to bind several TRAFs [25], [26]. Although the TRAF signature is not conserved in BAFF-R, this region binds to TRAF3 with high selectivity [27], [28]. In the mutant A/WySnJ mouse, a 4.7 kb gene insertion event disrupts the 3′ end of the BAFF-R gene, including part of this conserved sequence, leading to defective BAFF-R signaling [7]. The use of alternative splice donor and acceptor sites in exons 1 and 3, respectively, give rise to different mouse BAFF-R isoforms. These insertions and deletions occur outside the ligand and TRAF3 binding sites (Fig. 1, Fig. 2). Omission of exon 2 in human BCMA generates a predicted soluble receptor, but in the absence of a signal peptide, this protein is not expected to be secreted (Fig. 2A).

The 3′ genomic organization of TACI is similar to that of BCMA and BAFF-R, but the ligand-binding region is duplicated in an extra 5′ exon. This first ligand-binding domain has a much weaker affinity for BAFF and APRIL than the second one [29]. In human TACI only, an additional 5′ exon encodes a short N-terminal sequence. This often permits skipping of exon 2 to produce a short form of TACI lacking the first ligand-binding domain. This short TACI binds BAFF and APRIL as efficiently as the long form [29] (Fig. 1, Fig. 2). Human TACI is also subject to rarer intergenic splicing events removing the transmembrane and intracellular domains to yield soluble forms of TACI. However, TACI being a Type III protein, these proteins are not predicted to be secreted (Fig. 2A).

In mouse TACI, the initiating methionine and CRD1 are in the same exon. Excision of CRD1 is not possible in mouse TACI as it is in the human protein, because the 5′ non-coding exons do not contain an ATG codon in frame with the exon containing CRD2 (Fig. 1B). However, one sequence of TACI indicates that the reverse event can take place, i.e. deletion of the second CRD after splicing of a cryptic intron within exon 3 (Fig. 1, Fig. 2). Curiously, most mouse TACI transcripts start from within intron 4 and continue with exons encoding the transmembrane and intracellular portions of TACI. However, these transcripts lack initiating methionines and therefore should not encode proteins (Fig. 1, Fig. 2). It is not known if these transcripts regulate TACI expression or fulfill other specific functions. A practical outcome is that studies aimed at evaluating murine TACI expression at the RNA level will be influenced by the choice of RT-PCR primers.

In conclusion, conserved splicing events are relatively few (ΔBAFF and to a certain extent TWE-PRIL). Most other splicing events are species-specific, and therefore unlikely to play major biological roles.

Determining the true affinities of BAFF and APRIL for their receptors is not trivial. Indeed, measures performed with dimeric receptor–Ig proteins are subject to significant but unpredictable avidity effects that increase the apparent affinity. The avidity component is however removed by use of monomeric receptors, and results of such experiments are shown in Table 1. The consensus is that, on the one hand, BAFF binds to BAFF-R and TACI with affinities in the nanomolar range, yet displays two to three order of magnitude weaker binding to BCMA. On the other hand, APRIL binds TACI and BCMA with high affinity (nanomolar range), but not at all to BAFF-R. When receptors are dimerized as Ig fusion proteins, all receptors efficiently bind ligands in the low to subnanomolar range, with the exception of BAFF-R that does not bind to APRIL (Table 1, and reviewed in [30]). In comparison, APRIL binding to heparan sulfate proteoglycans is weak (20–80 μM) [30]. Whether endogenous BAFF can signal through BCMA in vivo is not known, but will certainly depend on avidity effects. Clustering of membrane-bound BCMA, clustering of membrane-bound BAFF or occurrence of multimerized forms of soluble BAFF (60-mer), are potential factors that may affect the avidity of the BAFF–BCMA interaction in vivo.

The molecular determinants of BAFF and APRIL specificity for their receptors have been solved through a number of elegant structural and mutational studies. BCMA and BAFF-R differ from other TNF receptor family members in that they possess a single cysteine-rich domain (CRD), which is used to establish extensive contacts with a single ligand monomer within BAFF or APRIL trimers [29], [31], [32] (Fig. 3, Fig. 4). Although TACI contains two CRDs, only the second one is necessary and sufficient for high affinity binding to BAFF and APRIL, so that the mode of TACI interaction with APRIL and BAFF is essentially the same as that of BCMA or BAFF-R [29]. This is in sharp contrast to other family members that contain up to four CRDs, and establish two main contact areas at the interface between two ligand protomers (reviewed in [33]).

BCMA, TACI and BAFF-R show perfect structural conservation in a β-hairpin structure that fits in a binding pocket of BAFF and APRIL and constitutes the conserved core of the interaction (Fig. 5). The hairpin structure is followed by a helix–loop–helix motive that is strikingly different among receptors: its spatial orientation differs in BCMA and TACI, independently of receptor binding, and BAFF-R contains only the first helix (Fig. 5). In a first approximation, it can be said that the hairpin is crucial for ligand binding, and that the C-terminal domain defines ligand specificity. Systematic mutagenesis studies of the ligand-binding sequence of BAFF-R [34], BCMA [35] and TACI [29] led to the precise identification of residues dictating the ligand specificity. For example, an aromatic residue in the hairpin, present in BCMA and TACI, but not BAFF-R, is absolutely required for APRIL binding, but dispensable for BAFF binding (Fig. 5, F78/Y13/C24). Introduction of this hydrophobic residue in BAFF-R confers binding to APRIL [31]. BAFF-R contains an hydrophobic residue (Leu38) that favors binding to BAFF but is detrimental for APRIL, further explaining the exquisite specificity of BAFF-R for BAFF and not APRIL (Fig. 5, P95/R27/L38). The corresponding position in BCMA is occupied by an arginine, which has the reverse effect, i.e. favoring APRIL binding, but inhibiting the interaction with BAFF. BCMA also lacks an arginine in the hairpin that establishes specific contacts with BAFF but not APRIL, accounting for the weak affinity of BAFF for BCMA (Fig. 5, R84/H19/R30). Human BCMA contains an hydrophobic residue (Ile22) that favors BAFF binding, but is not required for APRIL binding (Fig. 5, I87/I22/V33). Interestingly, mutation of this residue to an arginine abrogates the BCMA–BAFF interaction without significantly affecting binding to APRIL [35]. Finally, TACI does not present residues which are detrimental for APRIL or BAFF binding, explaining its dual specificity [29]. However, the set of residues involved in BAFF or APRIL interaction with TACI is partially different (Fig. 5).

In conclusion, the complex binding specificities of the BAFF and APRIL ligands (Fig. 2) is now very well understood at the molecular level.

Section snippets

BAFF-R and BCMA are involved in B cell survival

B cells enter the spleen at an immature T1 stage and further differentiate through the immature T2 stage to either mature or marginal zone B cells (Fig. 6). During differentiation, the requirement for a functional B cell receptor (BCR) is highlighted in IgβΔC mice [36]. The BCR consists of a surface immunoglobulin, which has no signaling capacity, and of Igβ and Igα chains that signal through their cytoplasmic tails. Deletion of the cytoplasmic tail of Igβ (IgβΔC) abrogates development of B

Concluding remarks

The BAFF subfamily shows a fascinating complexity at the levels of protein expression, ligand receptor interactions, signaling and functional outcomes. The past few years have witnessed considerable progress in our molecular understanding of ligand receptor specificity, and on the function of individual receptors. Although key features of signaling events leading to B cell survival have been highlighted, we are still far away from a complete molecular understanding of these signaling events.

Acknowledgments

We thank Helen Everett for careful reading of the manuscript. This work was supported by grants from the Swiss National Science Foundation, including a grant from the NCCR (National Center of Competence in Research) Molecular Oncology.

References (99)

  • D. Seshasayee et al.

    Loss of TACI causes fatal lymphoproliferation and autoimmunity, establishing TACI as an inhibitory BLyS receptor

    Immunity

    (2003)
  • G. Bonizzi et al.

    The two NF-kappaB activation pathways and their role in innate and adaptive immunity

    Trends Immunol

    (2004)
  • Y. Xu et al.

    Targeted disruption of TRAF3 leads to postnatal lethality and defective T-dependent immune responses

    Immunity

    (1996)
  • G. Liao et al.

    Regulation of the NF-kappaB-inducing kinase by tumor necrosis factor receptor-associated factor 3-induced degradation

    J Biol Chem

    (2004)
  • A.P. Grech et al.

    TRAF2 differentially regulates the canonical and noncanonical pathways of NF-kappaB activation in mature B cells

    Immunity

    (2004)
  • P. Ramakrishnan et al.

    Receptor-specific signaling for both the alternative and the canonical NF-kappaB activation pathways by NF-kappaB-inducing kinase

    Immunity

    (2004)
  • Y. Qian et al.

    Act1, a negative regulator in CD40- and BAFF-mediated B cell survival

    Immunity

    (2004)
  • M.C. Trescol-Biemont et al.

    Regulation of A1/Bfl-1 expression in peripheral splenic B cells

    Biochimie

    (2004)
  • Y. Diaz-de-Durana et al.

    TACI-BLyS signaling via B-cell-dendritic cell cooperation is required for naive CD8 + T-cell priming in vivo

    Blood

    (2006)
  • M.D. Thomas et al.

    c-Myb is critical for B cell development and maintenance of follicular B cells

    Immunity

    (2005)
  • D.D. Tran et al.

    CAML is required for efficient EGF receptor recycling

    Dev Cell

    (2003)
  • D.D. Tran et al.

    CAML is a p56Lck-interacting protein that is required for thymocyte development

    Immunity

    (2005)
  • H. Hase et al.

    BAFF/BLyS can potentiate B-cell selection with the B-cell coreceptor complex

    Blood

    (2004)
  • L. Planelles et al.

    APRIL promotes B-1 cell-associated neoplasm

    Cancer Cell

    (2004)
  • M. Hahne et al.

    APRIL, a new ligand of the tumor necrosis factor family, stimulates tumor cell growth

    J Exp Med

    (1998)
  • H.B. Shu et al.

    TALL-1 is a novel member of the TNF family that is down-regulated by mitogens

    J Leukocyte Biol

    (1999)
  • P. Schneider et al.

    BAFF, a novel ligand of the tumor necrosis factor family, stimulates B cell growth

    J Exp Med

    (1999)
  • P.A. Moore et al.

    BLyS: member of the tumor necrosis factor family and B lymphocyte stimulator

    Science

    (1999)
  • G.U. von Bulow et al.

    NF-AT activation induced by a CAML-interacting member of the tumor necrosis factor receptor superfamily

    Science

    (1997)
  • J.S. Thompson et al.

    BAFF-R, a newly identified TNF receptor that specifically interacts with BAFF

    Science

    (2001)
  • Y. Laabi et al.

    A new gene, BCM, on chromosome 16 is fused to the interleukin 2 gene by a t(4;16)(q26;p13) translocation in a malignant T cell lymphoma

    Embo J

    (1992)
  • K. Ingold et al.

    Identification of proteoglycans as the APRIL-specific binding partners

    J Exp Med

    (2005)
  • J. Hendriks et al.

    Heparan sulfate proteoglycan binding promotes APRIL-induced tumor cell proliferation

    Cell Death Differ

    (2005)
  • D. Bischof et al.

    Selective activation of TACI by syndecan-2

    Blood

    (2005)
  • S.L. Kalled et al.

    The biochemistry and biology of BAFF APRIL and their receptors

    Curr Dir Autoimmun

    (2005)
  • E.A. Zhukovsky et al.

    TNF ligands: is TALL-1 a trimer or a virus-like cluster?

    Nature

    (2004)
  • T.G. Cachero et al.

    Formation of virus-like clusters is an intrinsic property of the tumor necrosis factor family member BAFF (B cell activating factor)

    Biochemistry

    (2006)
  • A.L. Gavin et al.

    deltaBAFF, a splice isoform of BAFF, opposes full-length BAFF activity in vivo in transgenic mouse models

    J Immunol

    (2005)
  • G. Vogt et al.

    Gains of glycosylation comprise an unexpectedly large group of pathogenic mutations

    Nat Genet

    (2005)
  • M.P. Gras et al.

    BCMAp: an integral membrane protein in the Golgi apparatus of human mature B lymphocytes

    Int Immunol

    (1995)
  • K. Kelly et al.

    APRIL/TRDL-1, a tumor necrosis factor-like ligand, stimulates cell death

    Cancer Res

    (2000)
  • M. Yan et al.

    Two-amino acid molecular switch in an epithelial morphogen that regulates binding to two distinct receptors

    Science

    (2000)
  • B. Pradet-Balade et al.

    An endogenous hybrid mRNA encodes TWE-PRIL, a functional cell surface TWEAK-APRIL fusion protein

    Embo J

    (2002)
  • H.B. Shu et al.

    B cell maturation protein is a receptor for the tumor necrosis factor family member TALL-1

    Proc Natl Acad Sci USA

    (2000)
  • A. Hatzoglou et al.

    TNF receptor family member BCMA (B cell maturation) associates with TNF receptor-associated factor (TRAF) 1 TRAF2, and TRAF3 and activates NF-kappa B, elk-1, c-Jun N-terminal kinase, and p38 mitogen-activated protein kinase

    J Immunol

    (2000)
  • L.G. Xu et al.

    TNFR-associated factor-3 is associated with BAFF-R and negatively regulates BAFF-R-mediated NF-kappa B activation and IL-10 production

    J Immunol

    (2002)
  • C.Z. Ni et al.

    Key molecular contacts promote recognition of the BAFF receptor by TNF receptor-associated factor 3: implications for intracellular signaling regulation

    J Immunol

    (2004)
  • S.R. Dillon et al.

    An APRIL to remember: novel TNF ligands as therapeutic targets

    Nat Rev Drug Discov

    (2006)
  • Y. Liu et al.

    Ligand-receptor binding revealed by the TNF family member TALL-1

    Nature

    (2003)
  • Cited by (428)

    • Association between the expression of toll-like receptors, cytokines, and homeostatic chemokines in SARS-CoV-2 infection and COVID-19 severity

      2023, Heliyon
      Citation Excerpt :

      Triggering chemokines at the site of infection plays a key role in the early activation and migration of immune cells, including B cells and T cells. B cell-activating factor (BAFF) and a proliferation-inducing ligand (APRIL) are cytokines that are members of the TNF superfamily and play an important role in producing antibodies, activating B cells, and cell differentiation 11,12. CXCL12 and CXCL13 are potent B cell chemoattractant chemokines, and CCL21 and CCL19 are T cell chemokines [13].

    View all citing articles on Scopus
    View full text