Disclosure of potential conflicts of interest is found at the end of this article.
Author contributions: V.M.-S.: conception and design, manuscript writing, financial support, administrative support, final approval of manuscript; R.B.: collection and assembly of data, manuscript writing; E.B.-C.: collection and assembly of data, manuscript writing, final approval of manuscript.
First published online in STEM CELLSEXPRESS January 7, 2011; available online without subscription through the open access option.
CD10 is a remarkable member of the major class of widely expressed cell surface proteins, endopeptidases. First identified in leukemia as a tumor-specific antigen (common acute lymphoblastic leukemia antigen), CD10 has become largely used in cancer diagnosis. However, its function in oncogenesis remains unclear. We previously identified CD10 as a tool to access sphere-forming cells and showed its involvement in mammary stem cell (SC) regulation. We further illustrated that its enzymatic activity is involved, through signaling peptides, in SC maintenance. Therefore, CD10 is not only a cell surface marker in normal and malignant contexts but also affects the extracellular environment and plays a key role in regulation of a number of biological functions and likely in SC. In tumors, the “niche” favors the survival of sheltered cancer SC whose eradication has become the new challenge in oncology. This highlights the importance of understanding the role of CD10 in cancer SC. We will review the characteristics, main functions, and mechanism of action of CD10. Finally, we will review its clinical use and involvement in cancer. STEM CELLS 2011;29:389–396
The key gatekeepers of homeostasis are the stem cells (SCs) and their microenvironment. By providing biochemical and biophysical cues, through the extracellular matrix (ECM), the neighboring cells, the immune system, and soluble factors, the microenvironment is largely involved in the regulation and functioning of adult tissues. Therefore, altering the fine balance of ECM synthesis and homeostasis likely disrupts cell function and induces tumorigenesis. During morphogenesis or tumorigenesis, the tissue microenvironment undergoes extensive remodeling, including changes in deposition, degradation, and structural organization that likely affect all tissue components including SCs. This remodeling involves many enzymes and provides cues for controlling cell survival, proliferation, migration, polarization, and differentiation . ECM-degrading enzymes have raised considerable interest during the past decade, but the involvement in stem cell biology of enzymes that modulate cell signaling through peptides remains to be fully assessed. Neutral endopeptidase, also known as CD10 or enkephalinase, neprilysin, or membrane metalloendopeptidase, and common acute lymphoblastic leukemia antigen (CALLA), is a common zinc-dependent metalloendoprotease that inactivates a number of signaling peptides [2, 3]. This review will discuss the role of CD10 in the biology of normal or cancer SCs in several tissues and tumors. We will review basic and recent developments, in particular (a) CD10 structure, (b) its enzymatic activity, (c) its role in stem cell biology, (d) its potential use as a diagnostic and prognostic marker in oncology, and (e) its role in cell signaling, transduction, and cell mobility.
CD10 MOLECULAR AND PHYSICAL FEATURES
Membrane peptidases are present on the surface of many cell types and their substrates are found in the extracellular space as circulating, secreted, or membrane-linked peptides. Proteases have been implicated in activation or inactivation of several peptides and in ECM remodeling and have been classified according to their enzymatic action. Endopeptidases cleave peptide bonds, whereas exopeptidases liberate N-terminal amino acids . The family of zinc metallopeptidases includes the endothelin-converting enzymes, the Kell blood group protein, and the PEX gene product . CD10 is considered as the prototype of this family of membrane-bound zinc-dependent endopeptidases, which regulate the physiological action of various peptides by lowering their extracellular concentration available for receptor binding .
CD10-related DNA sequences have been found on human chromosome 3, at 3q21-27 . Three types of CD10 resulting from alternative splicing have been identified (Fig. 1A), suggesting that CD10 gene expression may be differentially controlled in a tissue-specific manner . This gene encodes a 90–110-kDa type II transmembrane glycoprotein (Fig. 1B) [3, 4] whose sequence is highly conserved with strong homology (93%) between rat and rabbit and only six nonconservative changes in amino acid sequences between human and rat . Models of CD10-enzyme secondary structure have suggested a three-dimensional representation of its active site involving 400 residues located in a central pocket (Fig. 1C) . Site-directed mutagenesis experiments have demonstrated that a glutamate-active (E646) site is involved in catalysis, whereas two histidin-active (583HExxH687) sites are responsible for binding the zinc cofactor (Fig. 1B) . The enzymatic function of metalloendopeptidases is inhibited by molecules containing biochemical domains such as thiol, carboxyl, hydroxamate, phosphoramide, or phosphonamide with very high affinity for the zinc catalytic domain . In addition, CD10 is sensitive to phosphoramidon, a Streptomyces metabolite first identified as a thermolysin inhibitor, that binds the active enzymatic site of CD10 .
CD10 MAIN BIOLOGICAL FUNCTIONS
CD10 was originally extracted and purified from the kidney (renal brush border of rabbit and hog ) and then found to be widely distributed (lung, male genital tract, intestine, and in some immune cells such as neutrophils, in fibroblasts, and epithelial cells ). In humans, numerous studies have demonstrated that CD10 is transiently present during B-cell maturation at early-B and pre-B lymphoblastic stages  and is expressed in endometrial stromal cells , in the proximal tubules and glomeruli of the human kidney , in the myoepithelial layer of the breast , in the epithelial and few stromal cells of the prostate , in the canaliculi of the liver , and in the epithelial cells of the stomach and colon .
Historical and Major Regulator of the Neural System
CD10 has been found to be expressed in the central nervous system  and to process a number of substrates such as enkephalin, an opioid peptide liberated by neurons in response to pain. Subsequently, other neuropeptides have been identified such as peptides of the tachykinin family (substance P), which possess vasodilatory properties through induction of histamine, serotonin, and bradykinin. CD10 is also known to be an amyloid β-peptide (Aβ) degrading enzyme. Polymorphisms in the CD10 gene increase the risk for Alzheimer's disease . Dysfunction of the CD10-enzyme leads to an accumulation of the insoluble neurotoxic β amyloïd (Aβ42) peptide, inducing premature and definitive neuronal death . CD10 deregulation likely results from different mechanisms such as environmental control by various parameters (somatostatin, estrogen, exercise, environmental enrichment, oxidative stress, hypoxic, and ischemic conditions or aging) . In some cases of Alzheimer's disease, CD10 deregulation originates from the genetic background of the patients who have a CD10 gene polymorphism .
In the Immune System
Several studies have demonstrated that CD10 is also involved in physiological functions outside the nervous system. CD10 is present on the surface of neutrophils and regulates their activation by degradation of inflammatory peptides . Transgenic CD10 knockout mice are almost developmentally normal with only slight differences in lymphoid development and intestinal inflammation. However, CD10−/− mice display a 10-fold increased sensitivity to endotoxin and die 100-fold more rapidly, and CD10+/− mice die 25-fold more rapidly, than wild-type animals . The potentiation of lethal shock is likely caused by the release and complex interplay of CD10-substrate proinflammatory peptides such as endothelin, bradykinin, tachykinins, atriopeptins, and interleukins. These peptides seem to synergistically induce ischemia, hemoconcentration, and vascular permeabilization, contributing to the perfusion defects observed in shock states . These data suggest that CD10 inactivates a proinflammatory peptide involved in immune system regulation and associated to certain autoimmune diseases. For example, rheumatoid arthritis and osteoarthritis are characterized by increased production of the inflammatory cytokines interleukin-β1 (IL-β1) and tumor necrosis factor alpha, which control the expression of Notch and the production of matrix-degrading enzymes. Notch signaling regulates the expression of several genes associated with osteoarthritis, including CD10 which in turn hydrolyzes IL-β1 . Chemical inhibitors, smoke, or allergens have been shown to downregulate CD10 levels in skin or lung, then favoring neurogenic inflammatory responses . A more recent study in melanoma has also demonstrated that the coexpression of CD10 with genes involved in antigen processing and presentation . Altogether, these data implicate CD10 in immune response.
In the Mammary Epithelial System
CD10 is transiently expressed during the development of organs such as lung and breast, suggesting the involvement of some cell surface endopeptidases in tissue morphogenesis [27, 28]. Although aminopeptidase N is expressed by the interlobular and intralobular stroma of the mammary gland, CD10 expression is restricted to myoepithelial cells . CD10 is involved in mammary gland development through control of cell growth and differentiation and in epithelial-mesenchymal morphogenesis by its cleaved peptide tachykinin [29, 30]. Oxytocin, another CD10-cleaved peptide, induces myoepithelial cell contraction through an increase in intracellular calcium levels, allowing milk release by luminal cells within the mammary duct .
CD10 IN STEM CELL BIOLOGY
CD10 has been used as a cell surface marker of the SC in many normal tissues (bone marrow , adipose , lung , and breast [34, 35]). In addition, sorting for CD10 enriches for sphere-forming cells, suggesting that this simple tool can identify SC or progenitor populations in tissues for which lineage studies are not currently possible . CD10 is also involved in many different processes during development such as in breast, an organ that undergoes a number of developmental cycles throughout life. Interestingly, a study in mouse has shown the implication of oxytocin, a peptide cleaved by CD10, in myoepithelial cell differentiation . Moreover, illustrating the dual role of CD10 in this system, we have recently demonstrated the involvement of its enzymatic function in the maintenance of human mammary SC pool . CD10 is also implicated in the differentiation of immature cells in other tissues and in B-cell maturation at early-B and pre-B lymphoblastic stages . Altogether, these data suggest that CD10 and its enzymatic functions might act largely as a SC regulator in a number of cellular compartments (Supporting Information in Supplemental Table 1).
CD10 AND CANCER
CD10 in Hematopoietic Tumors
The diagnostic and prognostic values of CD10 were discovered in acute lymphoblastic leukemia (ALL) with elaboration of a rabbit antisera against leukemic cells . The common ALL (CALLA) antiserum (later named CD10) was further shown to react with leukemic cells from more than 80% of non-T-cell ALL patients but not with normal hematopoietic cells . CD10 is expressed in B-lymphoblastic leukemia/lymphoma and in certain mature B-cell lymphomas (plasma cell myeloma, follicular lymphoma, diffuse large B-cell lymphoma and Burkitt lymphoma) and very rarely in T-cell lymphoma; in some cases CD10 expression has prognostic value . Indeed, CD10 is now used for the diagnosis of B-lineage ALL in combination with other B-lineage markers . CD10 alone can identify patients with higher 5-year survival rates in children with B-lineage ALL and indicates a better prognosis than CD10− . To adapt the chemotherapy to each patient, CD10 is used, in combination with CD19+ and/or CD34+, to assess minimal residual disease (MRD); children with low MRD, which is an indicator of good prognosis, can receive less intensive chemotherapy . In conclusion, in B lymphoma CD10 constitute a good prognostic marker.
CD10 in Solid Tumors
From the very first study, it was obvious that CD10 is not specific to hematopoietic malignancies but is also expressed by normal cells as detected in fetal liver, bone marrow, spleen, and brain  and several solid tumors such as children nephroblastoma and neuroblastoma , in adults with melanoma  and in several carcinomas originating from kidney , lung , skin , pancreas , prostate , liver , breast , stomach , cervix , and bladder .
CD10 As a Diagnosis and Prognostic Marker
Several studies have shown that CD10 might be a good marker for differentiating between primary tumors and derived metastases, and therefore for evaluating tumor progression. In particular, CD10 distinguishes primary hepatocellular carcinoma from secondary liver metastases originating from other organs [49, 54–57]. However, caution remains necessary because CD10 expression was also detected in several tumors initially suspected to be metastatic tumors from other tissues . Conflicting results have been reported in bladder carcinomas, with CD10 downregulation in progressive tumors  or inversely its upregulation associated with invasion and metastasis . CD10 staining can be a good prognostic marker in several carcinomas (cervix and non-small cell lung carcinomas [52, 59]) and inversely indicates poor prognosis in several solid tumors associated with disease progression and metastatic potential (gastric, pancreatic, and colorectal tumors  with a higher risk of liver metastasis , and in melanoma  and other skin tumors such as oral squamous  and basal cell  carcinomas). Interestingly, in conventional clear cell renal carcinoma, specific apical extracellular membrane staining has been associated with better outcome than the cytoplasmic or intracellular membrane staining predominantly observed in poorly differentiated tumors .
In conclusion, even though CD10 might not be used alone, it remains a very useful tool for diagnosis and prognosis, not only in hematopoietic tumors but also in several carcinomas. However, regarding the specific origin of CD10 deregulation for each tissue it is not possible to extrapolate a general interpretation of its expression in cancer and it should remain carefully analyzed in a context-specific fashion.
CD10 Role in Tumorigenesis
In breast cancer, the use of CD10 for diagnosis and prognosis is more complex. A study of 600 tumors from 200 patients with invasive breast cancer, including 54% duct carcinoma in situ (DCIS) and 23% lobular carcinomas, has demonstrated that CD10 overexpression correlates with improved disease-free survival and fewer metastases . A criterion for differentiating invasive from in situ breast carcinoma is the disappearance of myoepithelial CD10-positive cells and the basement membrane . This reflects an alteration of the environment structure by the tumor. Several breast studies have suggested that myoepithelial cells may function as guardians of tissue polarity . The tumor suppressor function of myoepithelial cells is progressively lost during transition from in situ to invasive carcinoma. Conversely, in invasive breast carcinoma CD10 is also abnormally expressed by environmental stromal cells, which contributes to obscure and apparently controversial interpretation. CD10 expression by the stromal cells surrounding the breast tumor is then correlated with poor prognosis, estrogen receptor negativity, and high grade . In prostate cancer, three patterns of CD10 expression have been observed, that is, membranous expression similar to benign epithelium, complete loss of CD10 expression (compared with adjacent benign glands) associated with relapse, and heterogeneous expression . In addition, DNA hypermethylation of the CD10 promoter has been identified in patients with decreased CD10 expression, suggesting that decreased CD10 expression might contribute to progression in localized prostate cancer and that methylation is essential for CD10 silencing . Finally, CD10-peptidase activity, which modulates the accumulation of peptides involved in cell proliferation, is involved in tumor progression as demonstrated for prostate , pancreas , and lung  cancers. Furthermore, the degradation of Met-enkephalin, a substrate of CD10, in colorectal cancer cells accelerates tumor growth and liver metastasis .
Cell Signaling Associated with CD10
In addition to its function through enzymatic activity, CD10 could directly mediate signaling events. Several cell surface markers are anchored in the membrane through glycolipids such as glycosylphosphatidylinositol (GPI; Fig. 1D). GPI complexes seem to define membrane microdomains-containing predominant signaling molecules of the Src family such as phosphotyrosine kinases (PTK) and G-protein. GPI-microdomains of distinct composition are involved not only in signal transduction through GPI-proteins and glycolipids but also in signaling initiation through immune receptors. Some of these immune receptors (CD10, CD90, or CD24) have been described to be anchored in the plasmic membrane through GPI-complexes . Binding of ligands onto these receptors results in signal transduction characterized by a transient elevation of cytoplasmic Ca2+, a tyrosine phosphorylation of cellular substrates, the initiation of effector functions such as degranulation of granulocytes and even proliferation and differentiation [15, 36, 71]. CD10 has been shown to be involved in the activation of focal adhesion kinase (FAK)-promoted cell adhesion [7, 72]. Indeed, CD10 in GPI-microdomains coimmunoprecipitate with Lyn and p85. This protein complex blocks PI3K (phosphatidylinositol 3-kinase) interaction with FAK by competitive binding, leading to decreased FAK phosphorylation and cell migration in the prostatic epithelial model  (Fig. 1D). These findings suggest an indirect negative regulation of cell migration by CD10, independently of its catalytic function . Complementary research from the same team in the same model has shown that enzymatic inactive CD10 maintains its ability to recruit the tumor suppressor phosphatase and TENsin homolog (PTEN) that prevents the activation of the Akt signaling pathway implicated in normal and tumor cell growth . Altogether, these studies suggest that CD10 loss could, like PTEN loss, contribute to the development and progression of prostate cancer. The interaction between CD10 and PTEN has been confirmed and further documented by other authors who have demonstrated the myristoylation of glycine-2-CD10 resulting in a redistribution of the subcellular localization of the CD10 protein. This important post-transcriptional process would contribute to regulating CD10 interactions with other proteins such as PTEN . In addition, CD10 can degrade peptides, such as bombesin and endothelin 1, which stimulate prostate cancer migration and invasion by bombesin-stimulated RhoA-signaling . Moreover, CD10 has been shown to induce the cleavage and inactivation of fibroblast growth factor 2 (FGF2) in vitro in a murine corneal pocket angiogenesis model, resulting in negative regulation of angiogenesis . Finally, CD10 directly binds to ezrin/radixin/moesin proteins, resulting in their decreased binding to the hyaluronan receptor CD44, increased cell adhesion, and decreased migration . In conclusion, CD10 prevents cancer cell migration, invasion, and angiogenesis through a double mechanism, enzymatic peptide degradation, and PTEN-linked signaling.
These data illustrate the difficulties of understanding the role of CD10 in cell biology as it can be mediated by both its extracellular enzymatic activity and its intracellular signaling pathways through crosstalk with other major signaling pathways controlled by hormones, cytokines, or adhesion molecules (Fig. 1D). The CD10-activated extracellular effects or intracellular signaling pathways that control major biological activities appear independent but could be either synergistic or antagonistic in some cases. Altogether, this makes CD10 a key element in the fine-tuned regulation of cell behavior.
CD10 has been identified and largely studied for more than 40 years in different areas of cell biology. It appears to be an important regulator shared by different tissues, from the immune and neural systems to epithelial tissues. CD10 actively participates to the regulation of vital physiological mechanisms (Table 1) and its biological activity is supported by two major mechanisms.
First, CD10, expressed at the surface of either epithelial cells or stromal cells, cleaves peptides through its extracellular enzymatic activity. These processed molecules then become activated or inhibited and contribute to SC regulation either in an autocrine manner or through their surrounding stromal cells. Second, CD10 is involved in an intracellular signaling pathway that interferes with other major signaling pathways controlled by elements present in the microenvironment.
Therefore, CD10 is a key molecule capable to integrate signals from either the cell environment or the intracellular compartment. These signals control major physiological functions and are possibly responsible for important changes in cell/tissue biology and microenvironment remodeling. It is therefore obvious that deregulation of CD10 expression or functions may cause severe disturbances in cells and their environment, as illustrated in neural disorders such as Alzheimer's disease or in cancers such as breast and prostate tumors. Altogether, these data indicate that CD10 enzymatic function and downstream signaling likely constitute key elements in the regulation of the biological functions of normal and malignant SCs and of their environment. On the basis of observations reported in the literature in patients with epithelial cancer (mainly breast or prostate cancer), we propose various scenarios that could happen in SCs, progenitors, or even the surrounding stromal cells for predicting the effects of variations in CD10 enzymatic activity, whether reductions (Fig. 2A) or increases (Fig. 2B). In addition, we will also examine the consequences of the deregulation in intracellular CD10 signaling during the malignant process (Fig. 2C). As CD10 expression is not systematically found to be altered in most epithelial transformed cells, the deregulation of its enzymatic function likely results from previous oncogenic events occurring in (stem) cells which, in turn, modulate the CD10-enzyme in the altered cells or even in the neighboring cellular environment. This complex cascade of events then leads to apparent contradictory results regarding the link between CD10 and cancer, as illustrated by opposite predictive values for CD10 detection in tumors that we previously discussed. Indeed, a decrease in CD10-enzymatic function after the transformation of early common progenitors (ECP) or Progenitors (P) could induce an accumulation of unprocessed peptides in the SC microenvironment, resulting in their lineage commitment and malignant proliferation (Fig. 2A). On the other hand, an upregulation of CD10 enzymatic activity could lead to an accumulation of local CD10-cleaved peptides that inhibit epithelial cell differentiation and maintain cancer SCs (Fig. 2B). This increased CD10-enzymatic activity could simply reflect the proliferation of transformed epithelial cancer SCs expressing CD10. The high number CD10-positive cells could explain the cleavage of the peptide that inhibits the differentiation into smooth muscle actin cells and mature cells. This model then could explain the disappearance in some cases of smooth muscle actin (SMA)–positive cells in invasive carcinoma and the disappearance of the basal membrane (Fig. 2B). This model could also explain the increased expression of CD10 by stromal cells observed in undifferentiated cancer, which could play a cooperative role in SC deregulation and maintenance (Fig. 2B). Finally, CD10 signaling could be modified in cancer progenitors or SCs (ECP/SC), independently of its enzymatic activity. These signaling alterations could block PTEN functions leading to apoptosis inhibition and cell proliferation through the Akt pathway (Fig. 2C). In conclusion, intrinsic CD10 signaling alteration could be considered as an initiating transforming event affecting preferentially immature cells, whereas alterations of CD10-expression/enzymatic activity seem to be only a consequence of the initial transformation.
Additional studies are needed to further explore the role of CD10, in particular, during oncogenesis, and to open new avenues for treatment in different clinical situations like cancer. These studies might especially help to develop new clinical approaches for a better targeting of cancer SCs in their niches.
We thank Elsevier for providing illustrative items. We thank Marie-Dominique Reynaud from the Centre Leon Berard for editing the article. This work was supported by grants from the INSERM and the Ligue Nationale contre le Cancer (Ain, Saone et Loire) to V. M.-S. E.B.-C. is the recipient of a Ph.D. fellowship from the Ligue Nationale contre le Cancer (Rhône) and the ARC (French Association for Cancer Research).
DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
The authors indicate no potential conflicts of interest.