CD81 was originally identified as a Target of an Anti-Proliferative Antibody (TAPA-1) in a study that also defined a new family of transmembrane proteins, later named tetraspanins (Oren et al. 1990). CD81 is expressed on most human cell types; however, Oren et al. have shown that the sensitivity of diverse cell lineages to the anti-proliferative effect of the anti-CD81 antibody differed (Oren et al. 1990). Therefore, subsequent immune-co-precipitation studies were performed to reveal that CD81 associates with different partner proteins in the various cells types. For example, in B cells CD81 associates with CD19, a B cell specific signaling molecule. Similar studies on additional family members confirmed that tetraspanins tend to associate in the membrane with each other and with additional partner proteins. These tetraspanin-enriched microdomains (TEMs) are dynamic membrane entities, which act as signaling platforms (reviewed in (Levy and Shoham 2005)). Indeed, CD81 functions as an organizer and a facilitator of signaling for its associated partners, it is also required for cell fusion and cell–cell interactions, functions that have been subverted by human pathogens.
The discovery that CD81 is a receptor for the hepatitis C virus (HCV) was made more than a decade ago (Pileri et al. 1998), it demonstrated that the viral envelope protein E2 binds CD81. Multiple subsequent studies confirmed the key role of CD81 in the life cycle of HCV (reviewed in (Dubuisson et al. 2008)). Remarkably, HCV is the only known natural ligand for human CD81.
Subsequent studies have shown that CD81 is also required for the life cycle of another major human pathogen, Plasmodium, the malaria-causing pathogen. A study by Silvie et al. has shown that Cd81−/− mice cannot be infected by P. yoelii sporozoites, the first stage of the parasite’s life cycle (Silvie et al. 2003). Follow-up studies demonstrated that anti-human CD81 mAbs blocked the infection of hepatocytes by P. falciparum sporozoites, the human malarial pathogen (Silvie et al. 2003). The mechanism by which CD81 enables the productive invasion of sporozoites into liver cells and their subsequent development into merozoites is not known; however, CD81 is not a receptor for this pathogen, as it does not bind the sporozoites directly.
Additional studies have suggested that CD81 has a role in the life cycle of HIV, especially it has been found to co-localize with this viral proteins in TEMs. For example, CD81 and additional tetraspanins have been shown to associate with the viral Gag and Env proteins and to modulate the release of HIV particles from a chronically infected T cell line. Interestingly, anti-CD81 mAb reduced HIV release and infectivity, whereas silencing of CD81 reduced viral release but increased infectivity (Grigorov et al. 2009).
CD81 also plays a role in fusion of egg with sperm – Cd81−/− female mice are infertile (Rubinstein et al. 2006). Interestingly, this defect became evident only after several backcrosses of the original outbred knockout mice onto inbred mice strains, because of this impairment Cd81−/− mice need to be bred as heterozygotes.
Unlike most other tetraspanins, CD81 is not glycosylated. The molecule contains six juxtamembrane cysteines, which have been shown to be palmitoylated, a modification that is common to tetraspanins. No other modifications have been reported for CD81.
The short cytoplasmic domains of CD81 do not contain known signaling motifs.
CD81-Associated Membrane Partner Proteins
CD81 associates with partner proteins, the latter differ in various cell types. A large number of studies have documented these interactions; few examples highlighting the functional consequences of such interactions are detailed below.
Early biochemical studies aimed at understanding the preferential sensitivity of B cell lines to engagement by an anti-CD81 mAb revealed an association with CD19, a signaling molecule belonging to the immunoglobulin (Ig) superfamily. Subsequently, B cells derived from three independently derived lines of Cd81−/− mice showed reduced CD19 expression. Interestingly, complete lack of CD81 in mice results in a milder effect on CD19 expression than that seen in a human patient diagnosed with a homozygous exon splice mutation in CD81 (Fig. 1), which resulted in complete lack of surface CD19. It is noteworthy that although redundancy in tetraspanins’ function has been suggested frequently, CD19 expression is dependent exclusively on CD81 and not on other tetraspanins. Indeed, re-introduction of CD81 into human or mouse B cells that lack CD81 restores CD19 expression (van Zelm et al. 2010). The absence of normal CD81 expression, both in human and in mouse, results in aberrant glycosylation of CD19. Thus, the absence of CD81 results in impaired trafficking of CD19, as evident by high mannose glycans that are sensitive to digestion by endoglycosidase-H, indicating residence in the endoplasmic reticulum (ER) without further exit through the Golgi to the cell surface.
In T cells CD81 was shown to be associated with CD4, a T cell specific molecule expressed on T helper cells. The cytoplasmic region of CD4 was sufficient for the association with CD81. Interestingly, these molecules associated with each other after removal of the Lck-binding site from the cytoplasmic domain of CD4, and the binding of Lck to CD4 inhibited the association with CD81. These findings led the authors’ suggestion that CD4 exist in two states, one associated with Lck, the other with tetraspanins (Imai et al. 1995).
Another Ig family member that associates with CD81 is EWI-2, an association that is maintained in harsh lysis conditions. This cellular partner of CD81 has been shown to modulate susceptibility to HCV infection. Hepatocytes express the complete EWI-2 and are susceptible to viral infection. However, EWI-2 exists in a form that lacks an Ig domain “without its N-terminus (EWI-2wint)” in some cell types. Cells that express EWI-2wint are not susceptible to HCV infection. For example, EWI-2wint, which is expressed in lymphocytes, blocks HCV entry by inhibiting the interaction of the viral envelope proteins with CD81 (Rocha-Perugini et al. 2008).
Biochemical studies have repeatedly shown that tetraspanins tend to associate with specific integrins in various cell types. This specificity of interaction was demonstrated in cells where only one specific integrin associated with an individual tetraspanin molecule, although the cell contained several integrin molecules. A study that analyzed the functional consequence of the association of CD81 with α4β1 integrin in B cells has shown that CD81 facilitated integrin-dependent adhesion strengthening (Feigelson et al. 2003).
G-Protein-Coupled Receptors (GPCR)
A search for CD81-associated proteins in a teratocarcinoma cell line identified GPR56, an orphan G-protein-coupled receptor. The association with CD81 was also seen in transfected cells where an additional tetraspanin molecule, CD9, was associated with GPR56 (Little et al. 2004). Further studies demonstrated that CD81 was required for the association of GPR56 with Gαq/11 and that engagement of CD81 by an antibody led to the dissociation of the G-protein from the GPCR (Little et al. 2004).
Role of CD81 in Cell Signaling
The presence or absence of CD81 can greatly affect downstream signaling events in a cell-of-origin- dependent manner. For example, while B cell signaling is impaired in the CVID patient shown in Fig. 1, her T cells responded normally to mitogens as measured by proliferation and interferon γ (IFNγ) production. Correspondingly, responses to CD81 engagement are highly dependent on the cell type. The examples below summarize studies that have delineated signaling pathways triggered in response to CD81 engagement either by antibodies or by the HCV envelope proteins. It is noteworthy that although these signaling cascades differ in the various cell types, they all ultimately provide a connection to the actin cytoskeleton.
The engagement of CD81 on human B cells induces tyrosine phosphorylation of a large number of proteins. Analysis of these phosphorylated proteins by mass spectrometry identified ezrin, an actin-binding protein, as the major tyrosine-phosphorylated band (Coffey et al. 2009). Coffey et al. also demonstrated that engagement of CD81 induced the phosphorylation of spleen tyrosine kinase (Syk) and that phosphorylation of Syk was required for the tyrosine phosphorylation of ezrin. Finally, the study demonstrated that the cytoplasmic C-terminal domain of CD81 was required for the induction of activated ezrin. The model emerging from this study is that CD81 interfaces between the cell membrane and the cytoskeleton where its engagement leads to activation of Syk, which in turn activates ezrin, which then recruits filamentous actin (F-actin) to facilitate cytoskeletal organization and cell signaling.
CD81 was demonstrated to be a costimulatory molecule both in human and in mouse T cells, namely, the coengagement of CD81 potentiated the stimulation induced in response to engagement of CD3 on these cells. Interestingly, these studies have also shown that the costimulatory effect induced by engaging CD81 is similar in magnitude to that induced in response to coengagement of CD3 and CD28, the latter being a classical T cell costimulatory molecule. However, while the cytoplasmic domain of CD28 contains tyrosine activation motifs, CD81 lacks such motifs. The detailed mechanism of T cell costimulation by CD81 is yet to be delineated. However, a recent study demonstrated that costimulation of CD3 and CD81 induced phosphorylation of Erk1/2 and activated the actin cytoskeleton (Crotta et al. 2006). An earlier study by the same group has shown that costimulation of T cells by CD81 was mediated by Lck. These and additional studies, which were aimed at understanding the interaction of HCV with human T cells have shown that the costimulatory potential of the envelope protein E2 of HCV (HCV-E2) was comparable to that of the anti-CD81 mAbs.
The hepatocyte cell line, Huh-7, which is susceptible to infection by HCV in cell culture (HCVcc) was used in the following study. Engagement of CD81 in this cell line by a mAb or by HCV-E2 induced the phosphorylation of Erk; moreover, an inhibitor of the Raf/Mek/Erk signaling pathway inhibited viral infection (Brazzoli et al. 2008). The same study also demonstrated that CD81 provided a linkage to the actin cytoskeleton. Briefly, engagement of CD81 induced its movement to the zone of cell–cell contact; in addition, treatment by drugs that inhibit F-actin formation blocked the relocation of CD81 to the cell-to-cell contact areas. Further analysis revealed that CD81 mediated the activation of the actin cytoskeleton by increasing the level of the active GTP-bound Rho family GTPases (Rho-A, Rac1, and Cdc42). Moreover, use of knockdown and of inhibitors of these GTPases reduced viral infectivity (Brazzoli et al. 2008). Thus, HCV subverts key cellular functions of CD81.
The tetraspanin molecule CD81 is a membrane-embedded protein. It functions as an organizer and a facilitator of signaling for its associated protein partners. CD81 associates with different partners in the various cell types. CD81, its partner proteins, and additional tetraspanins form TEMs, which act as signaling platforms connecting the cell membrane to the actin cytoskeleton. The role of CD81 has been subverted by major human pathogens.
- Coffey GP, Rajapaksa R, Liu R, Sharpe O, Kuo CC, Krauss SW, Sagi Y, Davis RE, Staudt LM, Sharman JP, Robinson WH, Levy S. Engagement of CD81 induces ezrin tyrosine phosphorylation and its cellular redistribution with filamentous actin. J Cell Sci. 2009;122:3137–44.CrossRefPubMedPubMedCentralGoogle Scholar