CD43, also called sialophorin or leukosialin, was first identified as a defective molecule of leukocytes and platelets of patients affected with the Wiskott–Aldrich syndrome, yet the molecule responsible for this immunodeficiency turned out to be WASP. It was thought that expression of this heavily glycosylated, mucin-type membrane protein was restricted to hematopoietic cells. However, recent advances in the field evidence that CD43 is present in non-lymphoid tissues, particularly tumor cells. Here we reviewed the most important features about this molecule, highlighting the recent advances that have contributed to our understanding of the roles of CD43.
Gene Expression and Protein Structure
CD43 is a type I cell surface glycoprotein abundantly expressed on almost all hematopoietic cells, except for erythrocytes and resting B cells. Although initially considered an immune cell molecule, recent reports evidence a much broader distribution, as epithelial cells, and a range of tumors of epithelial origin are CD43 positive. Expression of CD43 has also been documented in the normal brain and in the uterus. The gene encoding human CD43 is located in chromosome 16; it has two transcription initiation sites, and it is composed of two exons and a single intron within the untranslated 5′ region, although the entire protein is encoded only by the second exon. Alternative polyadenylation signals generate two mRNAs that differ from each other in the length of the 3′ untranslated region (Pallant et al. 1989).
CD43 Expression Is Tightly Controlled
The levels and relative abundance of CD43 on the cell surface are regulated through several mechanisms: expression regulation, molecular density, or processing, evidencing a regulatory role for CD43 in cell function.
Aside from a tissue-specific gene transcription regulation, the relative abundance of CD43 at a specific location of the cell membrane is tightly controlled. As T lymphocytes and neutrophils migrate to inflammatory lesions, CD43, together with PSGL-1, ICAM-3, and CD44, moves from the leading edge to the rear end (uropod) of the cell through its association with and movement of ERMs, while chemokine receptors relocate at the leading edge. Likewise, and through a similar mechanism, during the encounter of an antigen-specific T cell with the appropriate antigen-presenting cell, CD43 is excluded from the immunological synapse and relocates partially to the distal pole, where it plays a yet to define role. Interestingly, it is excluded from the inhibitor natural killer (NK) cell synapse, but not from the activating one (Reviewed in Ostberg et al. 1998; Aguilar-Delfín et al. 2006).
In addition, CD43 expression is regulated by proteolysis and shedding. Processing of CD43 has been described in human neutrophils and in murine T lymphocytes, as well as in human carcinoma cell lines (Lopez et al. 1998; Andersson et al. 2004; Seo and Ziltener 2009). The enzymatic cleavage of CD43 by a γ secretase-dependent intramembrane processing mechanism leads to the shedding of the extracellular domain, known as galactoglycoprotein and present in high concentrations in normal serum, and to the translocation of the intracytoplasmic domain to the nucleus, where it has been found to protect cells from apoptotic signals, promoting cell survival (Andersson et al. 2004; Seo and Ziltener 2009). Lastly, secretion of intracellularly stored CD43 is yet another mechanism through which CD43 expression levels are regulated: following LPS exposure, the intracellular levels of CD43 present in intestinal epithelial cells have been found to augment and correlate with an important secretion of CD43 to the intestinal lumen, underscoring a role for this mucin as a defense mechanism against infection in the intestinal epithelium (Amano et al. 2001).
A Multifunctional Protein
The abundance of CD43, its elongated structure, and the high sequence homology of its intracytoplasmic domain between species indicate that this molecule transduces information from the cell surface to the intracellular milieu in order to modulate the decisions of the cell. In line with the fact that CD43 has been found to partner with a seemingly ever-growing list of membrane molecules, multiple but often opposing functions have been described for CD43. The capacity of CD43 to transduce activation signals that regulate these multiple functions relies on its intracytoplasmic domain. Depending on the cell, and most probably on the ligand it encounters, CD43 must activate different signaling pathways that ultimately fine-tune the decision-making process of the cell, although little is known about this.
The net negative charge of CD43, resulting from the elevated glycosylation and sialic acid content, led to first propose that the role of this abundantly expressed cell surface mucin was to provide the cells with a repulsive force to limit cell contacts. Further attempts to understand the role of CD43 have been undertaken in vivo with the CD43−/− mice, with different disease models where T lymphocytes are pivotal. In a model of vaccinia virus infection, despite the fact that CD8-dependent cytotoxic response is more robust in the knockout mice, viral clearance is less efficient in these mice as compared to wild-type animals (Manjunath et al. 1995). Infiltration of naïve CD8+ T cells to the brain following intracranial infection with lymphocytic choriomeningitis virus (LCMV) is impaired in CD43−/− mice, resulting in prolonged survival of the animals. Likewise, In an experimental model of autoimmune encephalomyelitis (EAE), CD43−/− mice have a lower incidence rate of the disease and milder symptoms due to a defective migratory capacity of antigen-specific CD4 T cells to the central nervous system. In addition, and consistent with the reduced inflammation, these animals have higher levels of IL-4 rather than of IFNγ. The LCMV model also provides evidence that in addition to regulate the migration of T lymphocytes, CD43 participates in the contraction phase of the response, as antigen-specific CD8+ T-cell numbers remain higher in the knockout mice due to enhanced levels of Bcl-2 and hence decreased apoptosis (Onami et al. 2002; Ford et al. 2003). Another example of the involvement of CD43 in migration is provided by a murine model of elastase-induced abdominal aortic aneurysm (AAA), showing that CD43−/− mice are resistant to the development of such pathological condition due to a reduced recruitment of T cells and macrophages to the aortic wall. Moreover, CD43-mediated susceptibility to AAA depends on augmented IFNγ production by CD43+ CD8+ T cells, thus inducing higher levels of apoptosis in vivo and promoting extracellular matrix degradation (Zhou et al. 2013).
Altogether, this data highlight a positive role of CD43 in regulating the migration of naïve T cells to inflammation sites and in regulating cell number and homeostasis. This in turn determines the development and the outcome of the disease and correlates with the fact that anti-CD43 antibodies prevent the migration of T cells to pancreatic islands and, hence, the development of diabetes (Johnson et al. 1999). However, most of these studies were done in the mixed background BL6.129, and care should be taken in the interpretation of the data, as a mixed genetic background can influence the outcome of the experiments. Also, it should be taken into account that these animals are completely deficient for CD43. It will be interesting to evaluate the function of this mucin in mice where CD43 deficiency is targeted to a specific cell type.
Multiple Ligands for a Multifunctional Protein
The diversity of the physiological ligands identified for CD43 evidences a dynamic role for this cell surface mucin in regulating intercellular interactions and cell function. The fact that ICAM-1 and MHC-1 molecules have been found to act as receptors for CD43 (Reviewed in Pedraza-Alva and Rosenstein 2007), together with the recent finding that CD43 interacts with LFA-1 and CD147 in two distinct complexes (Khunkaewla et al. 2008), indicates that CD43 plays a role in cell–cell adhesion, in concert with, and probably regulating, other adhesion molecules. Consistent with this possibility, splenocytes and thymocytes from CD43-deficient mice exhibit enhanced homotypic adhesion to ICAM-1 and fibronectin (Manjunath et al. 1995). Paradoxically, in contrast with this anti-adhesive function, a number of anti-CD43 mAbs block homotypic cell–cell interactions or lymphoid cell binding to lymph nodes and high endothelial venules (Reviewed in Ostberg et al. 1998 and in Aguilar-Delfín et al. 2006). Furthermore, E-selectin has been found to interact with CD43, favoring Th17 cells and neutrophil and B-cell leukemia cell lines rolling on activated vascular endothelium, further evidencing the participation of CD43 as a pro-adhesive molecule that regulates the infiltrating ability of cells (Matsumoto et al. 2008; Nonomura et al. 2008; Velázquez et al. 2016). CD43 also interacts with Siglec-1 (sialoadhesin), a regulator of lymphoid and myeloid cell adhesion, probably consolidating physical contacts between cells. On the contrary, the putative association of CD43 to human serum albumin inhibits neutrophil function by limiting cell spreading as well as CD43 proteolysis by elastase (Reviewed in Aguilar-Delfín et al. 2006). Altogether, these data support the idea that CD43 triggers signaling pathways that modulate other conventional adhesion mechanisms.
In human dendritic cells, galectin-1–CD43 interaction primes the cells and results in cytokine production, upregulation of metalloproteases, and increased migration, through signaling pathways that depend on calcium fluxes as well as on Syk and Protein kinase C activation. However, through its interaction with galectin-1, CD43 promotes T-cell apoptosis, possibly by recruiting galectin-1 molecules to the cell surface and making them accessible to CD7, which is directly responsible for galectin-1-induced cell death (Hernandez et al. 2006; Fulcher et al. 2009). Interestingly, through the interaction of membrane nucleolin with capped CD43, macrophages recognize apoptotic cells (Miki et al. 2009).
The myriad of ligands underscores a regulatory role for CD43, ultimately favoring homeostasis of the immune system. Since most of these molecules are also ligands for other receptors, and different functions have been associated to each of them, this data underscores a role for CD43 in the fine-tuning of cell fate. However, information about the cellular responses and the intracellular signals that result from the interaction of CD43 with each of its ligands is very fragmentary, and this is undoubtedly a question that needs to be addressed.
CD43 Is a Pathogen Recognition Receptor
The complex glycosylation pattern of CD43 functions as bait for multiple pathogens. During HIV infection, as a result of increased Core2 enzyme activity, more O-Core2 glycans and lactosamine residues decorate CD43, ultimately leading infected cells to apoptosis, presumably as a result of the interaction of galectin-1 with CD43 and CD45. Moreover, it was suggested that this pathway can contribute to the death of infected as well as noninfected cells and thus to a decrease in T-cell counts (Lanteri et al. 2003). What is more, the combination of CD43 and TCR– CD3 complex signals lowers the signaling threshold for HIV LTR-driven gene activity, promoting the activation of NFκB and NFAT, ultimately favoring viral replication (Barat and Tremblay 2002). In addition, the alternative glycosylation of CD43 promotes the production of autoantibodies against hyposialylated isoforms, although their role in HIV pathogenesis is not understood (Reviewed in Pedraza-Alva and Rosenstein 2007).
Influenza A virus (IAV) was also found to interact with polymorphonuclear leukocytes through the specific interaction of CD43 with the IAV hemagglutinin. In addition to lessen the oxidative burst of these cells, IAV modifies the expression level of adhesion-related molecules: while it reduces that of CD43 and L- and P-selectins, it augments that of integrins CD11b and CD11c, subtly modulating the adhesive interactions of polymorphonuclear cells (Hartshorn and White 1999). It is presently not known if this is the result of the signals transduced by the CD43–hemagglutinin A interaction or if it reflects the activation of additional, CD43-independent pathways.
Several pathogenic bacteria also use CD43 as a pathogen recognition receptor (PRR). Recent reports have defined Cpn60.2 (Hsp65, GroEL), a molecular chaperone and an adhesin of the capsule of Mycobacterium tuberculosis, as a ligand for CD43 (Hickey et al. 2010). This is consistent with the fact that on macrophages, CD43 was previously identified as a receptor for M. tuberculosis, taking part in bacillus uptake and limiting its intracellular growth through the production of TNFα (Fratazzi et al. 2000). Additionally, three single-nucleotide polymorphisms (SNPs) in the CD43 gene region found in a Vietnamese population were associated with susceptibility to tuberculous meningitis (TBM) in a recessive model of inheritance. Two of these SNPs are associated with disease severity, and one of them is also associated with decreased survival of patients with TBM, but the underlying mechanism is not known (Campo et al. 2015). Streptococcus gordonii, the causative agent of infective endocarditis, has also been found to specifically interact with CD43 through HSA, another adhesin (Ruhl et al. 2000). More recently, CD43 was reported to be the substrate of a mucinase secreted by enterohemorrhagic Escherichia coli, the enteric pathogen that causes hemorrhagic colitis. The net balance of the proteolytic activity of this metalloprotease is the cleavage of the extracellular domain of CD43 from neutrophils, increased oxidative burst, but impaired neutrophil motility (Szabady et al. 2009).
In addition, CD43 is a counter-receptor for the trans-sialidase of Trypanosoma cruzi, the causative agent of Chagas disease. This enzyme transfers sialic acid from β-galactopyranosil donors to β-galactopyranosil acceptors with α2–3-linkages providing the parasite with a disguise to avoid recognition by the host immune system. Furthermore, it restores the glycosylation status of CD43 to a Core1 glycan expression pattern, similar to that of resting T cells, thus compromising the T-cell-dependent response, ensuring parasite replication. Interestingly, T. cruzi produces large amounts of an inactive and soluble form of the trans-sialidase that is also recognized by CD43, although the biological significance of this is not understood (Freire-de-Lima et al. 2010).
On non-lymphoid cells, CD43 retains the capacity to function as PRR. Together with the glycolipid GM1, CD43 has been described as a receptor for the heat-labile enterotoxin (LT) of enterotoxigenic E. coli in the microvilli of enterocytes (Zemelman et al. 1989). Interestingly, in CaCo2 cells differentiated into enterocytes, CD43 is not exposed on the cell surface, but stored in intracellular granules from where it is released and secreted to the extracellular compartment upon exposure of the cells to bacterial lipopolysaccharides (Amano et al. 2001).
Altogether, these data suggest that through the pathogen receptor function of CD43, PAMPS initiate a signal transduction pathway that promotes the activation of target cells and the onset of an inflammatory response and inflammation-related diseases. A better understanding of the specific role of CD43 in the pathogen recognition and the cell surface association processes will provide a clue to identify key target molecules both on the bacteria and on the host’s cells.
CD43 and Noninfectious Diseases: A Poorly Understood Function
In addition to function as a PRR, CD43 has been associated to diverse diseases. CD43 was first described as the molecule responsible for the Wiskott–Aldrich syndrome (WAS), an immunodeficiency characterized by frequent infections, skin eczema, decreased platelet numbers, and defective T-cell function (Remold-O’Donnell et al. 1984). However, it was later demonstrated that this X-linked disease was caused by mutations in the gene encoding WASP, a protein involved in signal transduction from cell receptors to the actin cytoskeleton. Yet, the link between WASP and CD43 is still unclear.
Although the participation of CD43 in several autoimmune diseases is documented, we are still lacking the full picture regarding the mechanisms of pathogenesis associated to this. Decreased expression of CD43 on the cell surface of synovial fluid polymorphonuclear cells of rheumatoid arthritis patients (Humbria et al. 1994), while an enhanced expression on the T cells infiltrating the synovium of osteoarthritis patients (Sakkas et al. 1998) have been reported. In systemic lupus erythematosus (SLE) patients, anti-CD43 and anti-galectin-1 autoantibodies, together with high levels of soluble galectin-1, were found to correlate with the time of disease evolution and low levels of complement (Montiel et al. 2010). Unusual high intracellular calcium levels, observed in pathological conditions such as systemic lupus erythematosus (SLE), Alzheimer’s disease, and brain trauma, result in capping of CD43. Similar to what happens with apoptotic cells, capped CD43 interacts with the macrophage receptor nucleolin, leading to recognition and phagocytosis of viable cells with elevated intracellular calcium levels by macrophages (Miki et al. 2013).
A role for CD43 in type I diabetes has been suggested, as preventing T-cell migration and possibly that of other CD43+ cells such as monocytes and dendritic cells from bloodstream to inflammation sites (lymph nodes, pancreas, and salivary glands) with exogenous anti-CD43 antibodies inhibits the progression of the disease (Johnson et al. 1999). Thus, reflecting the dual nature of CD43, anti-CD43 antibodies could favor the inflammatory response, exacerbating the disease (SLE and arthritis), or prevent the establishment of the disease (diabetes). Likewise, CD43 functions as a major ligand for E-selectin on activated vascular endothelium, regulating lymphocyte trafficking and favoring the recruitment of CD4+ Th17 cells to the insult site. Particularly, injection of CCL20 or TNFα into an air pouch of CD43−/− mice results in an impaired recruitment of Th17 or Th1 and Th17 cells, respectively. Furthermore, consistent with the fact that EAE is a disease where Th17 cell recruitment into the CNS plays a critical pathogenic role and that CD43−/− mice do not develop EAE, CD43 regulates the recruitment of Ag-specific Th17 cells in the murine model of EAE (Velázquez et al. 2016).
CD43 and Tumor Development: A Productive but Complex Association
CD43 has long been considered as a leukocyte-restricted molecule, and it has been mostly studied in leukocytes. Accordingly, this sialomucin is expressed in more than 90% of T-cell lymphomas and T-cell lymphoblastic lymphomas (Leong et al. 2003). In the B-cell lineage, contrasting with the limited expression of CD43 on normal early B cells and activated peripheral B cells, in transformed cells, CD43 is expressed at other B-cell stages, particularly in B-cell lymphoblastic (70%) and low-grade B-cell (30%) lymphomas (Leong et al. 2003). CD43 expression in B-cell lymphomas is associated to a bad prognosis.
In addition to hematopoietic tumors, non-hematopoietic tumor-derived cells from the breast, lung, colon, bladder, cervix, and prostate express CD43 (Reviewed in Aguilar-Delfin et al. 2006), mostly in the early stages of tumor development. Remarkably, in non-hematopoietic cancer cells, CD43 has mostly an intracellular localization in opposition to the typical membrane-bound one on leukocytes (Andersson et al. 2004; Fu et al. 2013). CD43 was first identified in the colon carcinoma cell lines COLO 205 as a 200 kDa molecule, supporting a role for aberrant glycoforms of CD43 in cancer development. Besides generating unusual epitopes that potentially serve as malignancy markers, CD43 influences the interaction of cancer cells with their microenvironment at different levels, resulting in enhanced cell proliferation, increased ability to invade and metastasize, as well as regulated interactions between the tumor cells and the intra-tumoral immune environment, eventually promoting resistance of the tumor cells to CTL-mediated lysis and the establishment of a microenvironment favorable to tumor growth (Andersson et al. 2004; Fu et al. 2013; Camacho-Concha et al. 2013; Tuccillo et al. 2014; Hasegawa et al. 2016).
Recently, CD43 was shown to participate in the tumor cell–peritoneum interaction, through a metalloprotease-dependent mechanism activated by the interaction with ICAM-1, one of its putative ligands (Alkhamesi et al. 2007). In addition, the proteolytic cleavage of the CD43 cytoplasmic domain (CD43-ct), through a regulated intramembrane proteolysis process (RIP), exposes a nuclear localization signal that promotes the translocation of the truncated protein to the nucleus, where it interacts with β-catenin, resulting in increased expression of the c-Myc and cyclin D1 genes (Andersson et al. 2004, 2005). In the absence of the tumor suppressors p53 and/or ARF, overexpression of CD43 in a human colon carcinoma cell line interferes with FAS expression and thus protects the cells from FAS-mediated apoptosis, ultimately favoring cell proliferation. Interestingly, the full-length molecule can be also found in the nucleus bound to the chromatin, reducing the level of FAS and activating the ARF/p53 pathway (Kadaja et al. 2004; Kadaja-Saarepuu et al. 2008), revealing a complex interplay between the full-length molecule and the CD43-ct peptide. In addition to being modified by SUMO-1 peptides, the cytoplasmic domain of CD43 has been found to localize with PML (promyelocytic) nuclear bodies (Seo and Ziltener 2009), further supporting a role for CD43 in promoting cell survival and carcinogenesis. This data suggests that different peptides perform different functions, ultimately fine-tuning cell function when acting in concert.
Consistent with the fact that CD43 was found to be mostly expressed in early-stage colon carcinomas and that in early stage some tumors accumulate p53, CD43 was reported to prolong the half-life of p53 by inducing its phosphorylation on Ser15, possibly in an ERK1/2-mediated process. This in turn disrupts the binding of p53 with the Mdm2 ubiquitin ligase and triggers p53-dependent cellular responses such as promoting the expression of Mdm2 and p21 cyclin-dependent kinase (Kadaja et al. 2004).
CD43 is a mucin that modulates the response of all cells of the immune system as well as that of other cells. Because of its elongated structure and its abundance on the cell surface, it serves as an antenna that senses the environment and prepares the cells for future actions. CD43 signals help the cell progress toward differentiation and maturation, as it regulates cell migration and cell contact. In addition, through the interaction with its numerous physiological and pathological ligands, CD43 actively participates in the onset of an inflammatory response and of inflammation-related diseases, underscoring the possibility to consider this molecule as a potential therapeutic target. An effort should be undertaken to identify those regions of the extracellular domain that are important for ligand recognition as well as to understand the rules that dictate selectivity of the different isoforms of CD43 for a given ligand. Unraveling the signals that follow interaction with a given ligand will help to untangle the multiple intracellular signals generated through this molecule. Understanding the role of soluble CD43 is also a challenge. A better appreciation of the functions regulated by the interaction of CD43 with each of its multiple ligands in different cells in diverse contexts of activation will provide a sharper picture of the roles this molecule plays in cell physiology in different organs, under normal and pathological scenarios.
- Aguilar-Delfin I, Fierro NA, Rosenstein Y. CD43. UCSD-nature molecule pages. 2006. doi:10.1038/mp.a000001.01.Google Scholar
- Barat C, Tremblay MJ. Engagement of CD43 enhances human immunodeficiency virus type 1 transcriptional activity and virus production that is induced upon TCR/CD3 stimulation. J Biol Chem. 2002;277(32):28714–28724. Epub 2002 Jun 3.Google Scholar
- Camacho-Concha N, Olivos-Ortiz A, Nuñez-Rivera A, Pedroza-Saavedra A, Gutierrez-Xicotencatl L, Rosenstein Y, et al. CD43 promotes cells transformation by preventing Merlin-mediated contact inhibition of growth. PLoS One. 2013;8(11):e80806. doi:10.1371/journal.pone.0080806. eCollection 2013.CrossRefPubMedPubMedCentralGoogle Scholar
- Fratazzi C, Manjunath N, Arbeit RD, Carini C, Gerken TA, Ardman B, et al. A macrophage invasion mechanism for mycobacteria implicating the extracellular domain of CD43. J Exp Med. 2000;192(2):183–92.Google Scholar
- Hasegawa K, Tanaka S, Fujiki F, Morimoto S, Nakano K, Kinoshita H, et al. Glycosylation status of CD43 protein is associated with resistance of leukemia cells to CTL-mediated cytolysis. PLoS One. 2016;11(3):e0152326. doi:10.1371/journal.pone.0152326. eCollection 2016.CrossRefPubMedPubMedCentralGoogle Scholar
- Humbria A, Diaz-Gonzalez F, Campanero MR, Arroyo AG, Laffon A, Gonzalez-Amaro R, et al. Expression of L-selectin, CD43, and CD44 in synovial fluid neutrophils from patients with inflammatory joint diseases. Evid soluble form L selectin synovial fluid Arthritis Rheum. 1994;37(3):342–8.Google Scholar
- Inui M, Hirota S, Hirano K, Fujii H, Sugahara-Tobinai A, Ishii T, et al. Human CD43+ B cells are closely related not only to memory B cells phenotypically but also to plasmablasts developmentally in healthy individuals. Int Immunol. 2015;27(7):345–355. doi:10.1093/intimm/dxv009. Epub 2015 Mar 5.Google Scholar
- Leong AS, Cooper K, Leong FJ. Manual for diagnostic antibodies for immunohistology. 2nd ed. London: Greenwich Medical Media Ltd; 2003.Google Scholar
- Miki Y, Oguri E, Hirano K, Beppu M. Macrophage recognition of cells with elevated calcium is mediated by carbohydrate chains of CD43. Cell Struct Funct. 2013;38(1):43–54. Epub 2013 Feb 9.Google Scholar
- Velázquez F, Grodecki-Pena A, Knapp A, Salvador AM, Nevers T, Croce KJ, et al. CD43 functions as an E-selectin ligand for Th17 cells in vitro and is required for rolling on the vascular endothelium and Th17 cell recruitment during inflammation in vivo. J Immunol. 2016;196(3):1305–16. doi:10.4049/jimmunol.1501171. Epub 2015 Dec 23.CrossRefPubMedGoogle Scholar