The T cell surface glycoprotein CD6 was first identified in 1981 by Kamoun et al. who for that purpose used a monoclonal antibody 12.1, raised against a human secondary mixed leukocyte reaction, that labeled all normal T cells and many T- and B-cell leukemias and lymphomas (Kamoun et al. 1981). The CD6 gene was later cloned from HPB-ALL cells by the group of Brian Seed using his pioneering cloning strategy, and revealed sequences with high homology with the cysteine-rich domain of the type I scavenger receptor of macrophages (Aruffo et al. 1991). Nevertheless, the obtained cDNA was incomplete such that the sequence corresponding to the cytoplasmic tail only coded for 44 amino acids. Based on comparisons with the mouse CD6 gene, Jane Parnes and colleagues finally published the currently known sequence, which includes a cytoplasmic tail of 244 amino acids (Robinson et al. 1995).
CD6 is highly expressed by mature T cells and medullary thymocytes, increasing steadily from double negative up to maximum expression in single positive thymocytes. Given the distribution pattern of CD6, mAbs against the molecule were used to deplete T cells from bone marrow grafts as a measure to prevent graft-versus-host disease (GvHD) as far back as 1982 (Reinherz et al. 1982). CD6 mAbs have progressively been abandoned in transplantation therapies but have gained importance in the treatment of psoriasis and other autoimmune diseases.
With the discovery in the late 1980s that CD6 could mediate T cell activation when crosslinked with CD3, CD6 became considered as a potent costimulator (Walker et al. 1987). This view was reinforced when it was shown that it localized at the immunological synapse and loosely interacted with the TCR/CD3 complex upon antigen recognition, and moreover, that it established inducible associations with potent intracellular activation transducers such as tyrosine kinases and the adaptor SLP-76 (Gimferrer et al. 2004; Castro et al. 2003; Hassan et al. 2006). Recently, however, it has been shown that CD6 can mediate inhibitory signaling, thus restraining the potency of TCR-mediated activation (Oliveira et al. 2012). These contradictory views on the role of CD6 in T cell activation can be reconciled considering that the receptor may have a dual function depending on the context of activation and cellular development stage, or more likely that CD6 may in fact be a transmembrane adaptor that can transduce signals of opposing nature depending on the sequence of signaling events.
Intracellular CD6 Binding Partners
CD6 is a highly glycosylated transmembrane glycoprotein of 117 kDa, containing an extracellular domain composed of three scavenger receptor cysteine-rich (SRCR) modules and a long cytoplasmic tail containing nine tyrosine residues, which may become phosphorylated upon TCR triggering. Additional activation-relevant consensus sequences include two proline-rich motifs, which are docking sites for SH3 domain-containing proteins, three serine/threonine-rich motifs, three PKC phosphorylation-site motifs, and ten casein kinase-2 phosphorylation sites.
The first intracellular interactions of CD6 were only reported in 2003 when it was shown that rat CD6 could coprecipitate different protein kinases such as Fyn, Lck, ZAP-70, and Itk (Castro et al. 2003). However, it remains to be determined whether these interactions are phosphorylation-dependent, which are the kinases that phosphorylate each of the tyrosine residues, and how signaling progresses downstream of the receptor. The only signaling mediator proven to bind to CD6 through an SH2 domain and in a phosphorylation-dependent manner is the intracellular adaptor protein SLP-76 (Hassan et al. 2006), best known for linking the TCR-CD3/ZAP-70/LAT axis to downstream signaling through connections with a number of other intracellular adaptors and effectors.
CD6 was also described to interact, in a phosphorylation-independent manner, with the PDZ domain-containing adaptor syntenin-1, which is able to bind cytoskeletal proteins and signal transduction effectors (Gimferrer et al. 2005). As CD6 contains a long cytoplasmic tail with several motifs that function as potential docking sites for sequential effector molecules, there is still a long way ahead until all the molecular mechanisms and signaling pathways regulated by CD6 can be fully understood.
Regulation of TCR-Mediated Signal Transduction
The first functional studies on CD6 were performed using specific mAbs and described important changes in cellular behavior. In 1989, Ledbetter and colleagues analyzed the role of 35 heteroconjugates in raising cytoplasmic calcium mobilization and one of such receptors was CD6 (Ledbetter et al. 1989). Some CD6 mAb displayed the most vigorous proliferative responses amongst many other mAb against diverse T cell antigens, and interestingly, co-crosslinking of CD6 was able to restore proliferation even when some anti-CD3 reagents were not effective.
CD6 mAbs can also mediate T cell activation in combination with PMA, a phorbol ester that directly activates PKC, pointing to Ser/Thr phosphorylation as an additional or complementary mechanism to transduce signaling. Two constitutively phosphorylation serine clusters, embedded in CK2 consensus motifs (S480/482/484 and S560/562/565/567/568) were indeed shown to be crucial for CD6-dependent MAPK activation (Bonet et al. 2013). Although PKC activation seems to be necessary for CD6-mediated T cell activation and proliferation, the mechanisms involved are poorly understood.
Altogether, these evidences demonstrate that CD6 has the potential, once directly stimulated, to convey or transduce activation signals, especially since it physically associates with important effectors of T cell signaling. However, all this potential has to be carefully scrutinized given that the new concepts on how T cell activation is triggered attribute to the microenvironment where all the initial steps take place, the immunological synapse, a paramount role.
The IgSF molecule CD166 was identified as a CD6 specific ligand, and this interaction was one of the first reported for an SRCR molecule. Other extracellular interactions have been described for CD6, namely CD6-binding molecules expressed in IFNγ-activated keratinocytes, in synovial fibroblasts, in the HBL-100 breast carcinoma cell line and in cells derived from human thymus, skin, synovium, and cartilage, but no positive identification of these putative ligands has been obtained (Pinto and Carmo 2013).
CD166 has a broad distribution including mononuclear cells in spleen, lymph node, and tonsil; hepatocytes; epithelium in many organs; bone marrow; neurons; and microglial cells in the brain. Despite its wide distribution, CD166 is primarily expressed in subsets of cells involved in dynamic growth and/or migration, including neural development, branching organ development, hematopoiesis, tumor progression, and immune responses.
CD6-CD166 Binding and Functional Effects
Mutational analyses on d3 of CD6 identified eight residues (E293, Y327, S329, F344, N346, N348, Q352, S353) critical for binding to CD166 with the additional contribution of yet five residues (D291, S292, Y295, E298, L349). Conversely, nine residues in CD166 d1 are critical for bond establishment (F26, F40, F43, K28, K48, D54) and three other (M87, T90, E91) also support the interaction. The electrostatic potential of the binding region in both CD6 and CD166 is also likely to contribute for CD6-CD166 binding (Santos et al. 2016).
CD6 binds to CD166 when both molecules, expressed in apposing cellular membranes, are localized in the immunological synapse (Gimferrer et al. 2004). The consequences on the T cell of the engagement of CD6 with CD166 have been addressed through the use of both soluble proteins and monoclonal antibodies targeting CD6 or CD166. Conflicting reports establish CD6 as having a stimulatory or inhibitory role depending on the assay and experimental readout used. The use of soluble proteins or mAb presumably blocking the interaction usually induces a decrease in T cell activation, consistent with CD6 being a costimulator. However, if these reagents do not directly touch the CD6 molecule, the effect may be opposite. A possible explanation is that CD6 mAb or soluble CD166 that bind CD6 induce a direct inhibitory effect, which is different from simply blocking the interaction. Supporting the interpretation, different mAb targeting either CD166-d1 or CD6-d3 are both able to block the interaction; however, anti-CD166-d1 induces T cell proliferation whereas anti-CD6-d3 represses it (Santos et al. 2016).
The Cd6 Knockout Mouse Model
A mouse deficient for the Cd6 gene was recently developed, and it was observed that in the absence of CD6 there is a decrease in the number of CD4+ and CD8+ T cells, but also of Treg cells, and the conjugation of these factors leads to exacerbated responses to autoimmune challenges (Orta-Mascaró et al. 2016). Altogether, the evidence indicates that CD6 can modulate the threshold for thymocyte selection and the generation or function of different peripheral T cell subpopulations.
CD6 Alternative Splicing
The CD6 gene is notorious for the many splice variants produced, originating CD6 molecules that differ in the cytoplasmic tail or that can exclude one or more SRCR extracellular domains. An important splicing-encoded variant, CD6Δd3, which lacks the extracellular d3 (due to exon 5 skipping) and is thus not able to bind to CD166, is markedly upregulated upon T cell activation (Santos et al. 2016). Molecularly, upon activation of T cells, transcription-related chromatin CD6 acetylation and downregulation of the splicing factor SRSF1 (responsible for binding to a regulatory element in CD6 intron 4 leading to exon 5 inclusion) stimulate exon 5 skipping. Functionally, while full-length CD6 targets to the immunological synapse, the alternative splicing-dependent isoform CD6Δd3 may not be able to target to the T cell:APC interface by virtue of the absence of the CD166-binding domain. This may represent a negative feedback mechanism resulting in a loss of adhesiveness between the two cells. The absence of CD6Δd3 from the immunological synapse could additionally mean that signaling mediators that bind to the cytoplasmic tail are effectively removed from the synapse.
The Role of CD6 in Autoimmunity
A good example of the implications of the complex regulation of CD6 is its role on multiple sclerosis (MS). The CD6 gene has been established as a susceptibility gene in MS with a single nucleotide polymorphism (SNP) at the locus rs17828933 within the first large exon being implicated in variations in gene expression (De Jager et al. 2009). The risk-associated allele results, through a still uncomprehended mechanism, in the enrichment of the CD6Δd3 isoform and a corresponding decrease of full-length CD6 (Kofler et al. 2011). Functional assays showed that normal T cells (with the nonrisk allele) were more responsive to TCR triggering and proliferated better than those containing the SNP. It may be puzzling why the disease cells, responsible for autoimmune reactions, may respond less in vitro, but these observations only highlight how complex and misleading it can be to analyze the role of CD6 and its influence of T cell responses in a less-than-perfect context.
CD6 may also have a role in other autoimmune diseases, like psoriasis and rheumatoid arthritis, but in this case there are already developments in the use of therapeutic antibodies with very successful outcomes. Montero and his group explored the effects of anti-CD6 reagents to manipulate responses and while costimulation using soluble ligand (sCD166) led to lymphocyte activation and differentiation, a mAb against d1 of CD6 inhibited T cell proliferation even in the presence of CD166 and excess of IL-2. These results strongly support a CD6-targeted therapy using mAbs that not necessarily prevent the CD6–CD166 interaction (Nair et al. 2010).
While historically CD6 has been viewed as a potent costimulator of T cell activation, recent studies show that CD6 has a clear capacity of attenuating T cell signaling in early and late events (Oliveira et al. 2012). How can these opposing views be reconciled, especially given that it is objective that CD6 associates with positive effectors?
Although insufficiently explored, the combination of alternative splicing-dependent localization with signal transduction mechanisms may engender an intriguing hypothesis to explain the inhibitory properties of CD6: it may be a decoy protein and its removal from the synapse as consequence of alternative splicing results in the exclusion of the positive effectors associated with its cytoplasmic tail. Alternatively, some still unknown CD6-associated proteins can be inhibitory enzymes and with a similar function as observed when they associate with the related receptor CD5.
CD6 can thus have a dual role during T cell activation depending on a combination of factors, including the nature of effectors that bind to the cytoplasmic tail and whether they are cotransported to the synapse, or removed, in different moments of the activation process.