Mouse Cell Surface Antigens: Nomenclature and Immunophenotyping¹

Molecules on the surface of hemopoietic cells play important roles in the development and function of these cells and have permitted us to understand the immune system in increasingly great depth. In recent years, it has become clear that there is a considerable amount of cross-talk between cells of the hemopoietic system and nonhemopoietic cells, with much of this interplay mediated by cell surface molecules. This review includes a discussion of cell surface Ags expressed on both hemopoietic and nonhemopoietic cells. In addition to an update on nomenclature, there will be a discussion of functions of the molecules, when known, and brief comments on the use of particular Ags for immunophenotyping cell subsets. The review will cover molecules included in the cluster of differentiation (CD)³ and lymphocyte Ag (Ly) series as well as some new Ags without current CD or Ly assignments.

As discussed in previous reviews (1, 2), there is a need for unifying mouse and human nomenclature to facilitate communication between researchers studying these species. For mice, the Ly nomenclature was originally devised to classify genes identified through serologic studies of inbred strains; for humans, the CD nomenclature originates from mAb reactivity to human Ags. Human leukocyte differentiation Ag (HLDA) workshops assign each CD based on the same reactivity to one human Ag by at least two mAbs; provisional CDw are sometimes given to clusters not well characterized or represented by only one mAb (3). The Sixth HLDA Workshop, which took place in 1996, resulted in the assignment of novel CDs with new designations spanning CD131 to CD166. mAb submitted to the workshop are tested by laboratories participating in the following sections: T cell, B cell, NK cell, adhesion, endothelial, myeloid, nonlineage, platelet, cytokine receptor, and blind (multilineage panels). More information can be obtained from the HLDA website (http://mol.genes.nig.ac.jp/hlda) or Protein Reviews on the worldwide web (http://www.ncbi.nlm.nih.gov/prow).

Over the years, the Committee on Standardized Genetic Nomenclature for Mice has continued to assign new Ly and CD names to novel genes and Ags. Since the last update (2), many new Ly designations have been assigned; for example, the F4/80 Ag, whose gene has recently been cloned, was given the designation Ly-71. (See Table I for an update of the Ly nomenclature.) The human homologues of a number of mouse Ags or genes, including members of the Ly-6 and Ly-49 families, have not yet been definitively identified. When a mouse Ly Ag is identified as a human CD homologue, the Ly number for the molecule is withdrawn and reassigned the appropriate CD number. If the mouse molecule was encoded by a gene that is assigned a Ly number, that gene name is withdrawn and reassigned a Cd number, unless another gene name was agreed on by the human and mouse nomenclature groups. As one example, the Ly-5 molecule of the mouse, encoded by Ly5, was assigned CD45 in the human nomenclature for Ags and the gene name CD45. The mouse designations were changed to CD45 for the Ag and, initially, Cd45 for the gene. More recently, the human and mouse nomenclature committees adopted the gene Ptprc for the genes encoding CD45 in both species.

Multiple studies, including biochemical analysis, cloning, functional, and immunologic assays, are necessary to confirm homologues between species. In addition to differences in DNA sequence, evolutionary divergence between mice and humans may also be manifested in Ag distribution. Notable examples include CD2, CD90 (Thy-1), and perhaps CD34. Table II reflects several novel mouse CD homologues that have been identified via cloning, Abs, or protein probes, such as the use of ligand-Ig fusion proteins.

Some molecules are particularly useful as phenotyping markers for different cell subpopulations. The large number of well-characterized mAbs has facilitated identifying cell types based on their surface phenotypes. For additional reference, a review of mAbs to human and murine CD Ags has been made available (4). It should be noted that only a few Ags have restricted lineage distributions; most cell surface molecules exhibit a broader distribution than initially reported. Multiparameter immunoanalysis, therefore, is required to isolate different cell types. Below is a summary of relevant Ags associated with cell lineages.

In the mouse, one of the most commonly used pan-B cell markers is identified by the mAb RA3-6B2 (CD45R/B220); however, this epitope is also expressed on activated T and NK cells (5) and NK cell progenitors (6). The CD19 Ag appears to be more restricted to the B cell lineage and is not expressed by NK progenitor cells (6) or by LAK cells (N. Alaverdi, unpublished observation). Hence, Abs to mouse CD19 may be used more reliably to identify B cells. Although the mouse CD20 gene has been cloned (7), no mAb has been reported. While surface IgM and IgD are expressed by both conventional (B-2) and unconventional (B-1) B cells, the expression of CD23 (8), CD5, and CD11b (9) can be used to distinguish these subsets. Other markers identifying the B cell lineage and its subsets include CD138 (Syndecan-1) (10), CD157 (BP-3) (11, 12), CD35 and CD21 (13), CD40 (14), CD72 (15), CD22 (16), Ly-68 (AA4.1) (17), and Ly-78 (RP-105) (18, 19). CD86 (B7-2) and CD80 (B7-1), although broadly expressed on APCs other than B cells, are useful markers for activated B cells.

While mAbs to Thy-1 (CD90) and TCR complex Ags have commonly been used as pan-T cell markers in mice, Thy-1 is not restricted to T cells, while CD3e expression is correlated with T cell maturation, similar to CD27 (20) and CD28 (21). In humans, CD2, CD5, and CD7 are preferred pan-T cell markers, although they are also expressed on subsets of other cell types (3). The mouse CD7 gene has been cloned recently (22); its Ag distribution remains to be determined. The CD8 and CD4 molecules are generally used for identification of mainstream helper and cytotoxic T cells, respectively. A third class of T cells with both helper and cytotoxic properties has been identified using the CD161c (NK1.1, previously Ly-59 or NKR-P1C) (23). Reported memory T cell markers include CD44 (24), CD62L (25), and CD45RB (26). Activated T cell markers include CD26, CD27, CD30, CDw137 (4-1BB), CD152 (CTLA-4), CD154 (gp39), CD134 (OX-40), CD95L (Fas ligand), CD45R/B220, and Ly-6E (TSA, sca-2) (27).

The official mouse and human genetic nomenclature for a substantial number of the surface molecules expressed by NK cells will, most likely, soon be changed by consensus of those expert in the field. The provisional symbol for both the mouse and human nomenclature relating to the lectin-like molecules would be Klr, for killer cell lectin-like receptor. This will be followed by the letters a through e to designate five distinct families and then a number to specify each member of that family: Klra# will be used for the Ly49 family, Klrb# for the NKR-P1 family, Klrc# for the NKG2 family, klrd# for CD94 and related genes, and klre# for the MKAP family. The Klra and Klre families have no human members to date.

The CD161c (NK1.1 Ly-59, NKR-P1C) Ag is the most widely used pan-NK marker in mice. Since its expression is restricted to CE, New Zealand Black, and C57BL/6-related strains, many commonly used strains do not express the Ag. In addition, like many NK cell markers, CD161c is also expressed on a subset of T cells (23). Although CD56 (NCAM (neural cell adhesion molecule)) is used as the human NK cell marker, several existing Abs to mouse CD56 did not react with mouse NK cells (3); however, a novel mAb, DX5, exhibits similar reactivity to anti-CD161c Ab but reacts with NK cells in all strains of mice tested, including NK1.1-negative strains (L. Lanier, unpublished observation). CD122 (IL-2R β-chain), which is constitutively expressed on NK cells and a subset of T cells, can also be used to identify NK cells (V. Kumar, unpublished observation). A multitude of genes encoding NK receptors, such as the newly cloned mouse gene CD94 (L. Lanier, unpublished observation), CD161a, and the Klra family members, have been characterized (28). The MHC class I ligands of these NK receptors and their inhibitory or activating functions are being elucidated. To date, no mouse homologues of the killing inhibitory receptor (KIR) and killing-activating receptor (KAR) Ig superfamily members have been discovered, although two mouse genes related to human KIR, gp49A and gp49B (B1 and B2), have been cloned from mouse mast cells (29). The available data suggest that gp49 is unlikely to be the mouse equivalent of human KIR (29). It is likely that Klra family members serve as KIR and KAR functional equivalents in mice (28).

The anti-Ly-71 (F4/80) mAb has been used extensively in determining the distribution and function of mouse tissue macrophages (30), although eosinophils and dendritic cells (DC) have been reported to react with this mAb (31, 32). To date, no human homologue of Ly-71 (F4/80) has been reported. CD11b (Mac-1), another commonly used marker for the monocyte/macrophage lineage, is expressed on NK cells, granulocytes, a T cell subset, and peritoneal B-1 B cells. CD14 is widely perceived as the best marker for the human macrophage/monocyte population. In mice, the level of CD14, as recognized by an anti-CD14 mAb, rmC5–3, was low or undetectable on resting blood monocytes. It remains to be determined whether other newly generated Abs to mouse CD14 will recognize a distribution similar to that in humans. Other mAbs used for identification of macrophage/monocyte subsets are MOMA-1, MOMA-2 (33), Mac-2, Mac-3, and the macrophage scavenger receptor.

DC display surface phenotype heterogeneity depending on their microenvironment and state of activation. Their ill-defined surface phenotypes in addition to their low numbers in tissues have made the isolation of these cells rather cumbersome. DC express many adhesion and costimulatory molecules and myeloid lineage markers (32). Although several Ags have been reported to be expressed specifically by DC, mAb to these Ags often react with other cell types or only recognize subsets of DC. For separation of DC from other cell types, multicolor analysis or prior enrichment protocols such as plastic adherence methods are necessary. The Ly-75 (DEC-205) Ag (34) and CD11c (35) have proved to be more restricted markers for DC, although they may not be expressed on all DC, and they may also be expressed on other cell types. In humans, CD83 and the Ag identified by mAb CRMF-44 have recently been identified as novel markers for DC (3, 36). Other Abs useful for the identification of DCs include Ly-79 (33D1) (37), 4F7 (38), and Ly-74 (Ep-CAM, gp40), a homologue of human epithelial growth factor (39).

Expression of Ly-6G is reported to be primarily limited to granulocytes; mAbs specific to this molecule, also known as the Gr-1 Ag, have been used successfully to separate mouse granulocyte lineage cells (40).

The Ly-76 Ag detected by the mAb TER-119 has been used to identify cells of the erythroid lineage (41, 42).

MECA-32 appears to be most restricted marker for endothelium (43). Other Ags expressed by endothelial cells include CD106 (vascular cell adhesion molecule-1), CD31, CD34, Ly-73 (Flk-1), and CD105. CD62E is expressed by activated endothelial cells (27, 44).

CD41 (integrin αIIb) is the best marker for platelets; other surface proteins expressed by platelets include CD61 (integrin β3) (45) and CD9 (46). Activated platelets express CD62P (27).

The classic method for identifying mouse stem/pluripotent cells in total bone marrow cells has included the use of a combination of Abs; cells negative for lineage markers CD4, CD8, CD11b (Mac-1), CD45/B220, Ly-6G (Gr-1), and Ly-76 (Ter-119) and positive for CD117 (c-kit), and Ly-6A (Sca-1), which are greatly enriched for capacity to reconstitute hemopoiesis (47, 48). Polyclonal and monoclonal Abs to mouse CD34 (49, 50) have demonstrated that CD34 is expressed on a small subset of bone marrow cells. In humans, the stem cell populations are identified by the expression of CD34. A recent report with one anti-mouse CD34 mAb, 49E8 (RAM34), however, suggests that primitive hemopoietic stem cells capable of long term repopulation are contained in the CD34-negative/low fraction, whereas the CD34-positive cells are committed progenitor cells lacking self-renewal capability (51). Other relevant Ags expressed by progenitor cells include CD25, CD90 (Thy-1), ER-MP12 (52), CD135 (Flk-2/Flt-3) (53), Ly-6E (TSA-1, Sca-2) (54), Ly-51 (BP-1, 6C3), and CD157 (BP-3).

An important feature of cellular activation is the de novo expression of surface molecules or up-regulation of the Ags expressed constitutively. The mode of stimulation, kinetics, and expression pattern of any given marker may imply its role in the immune response. Surface molecules, including CD71 (transferrin receptor), CD98 (4F2), and CD69, are expressed by a wide range of activated cell types, reflecting their general role in cellular proliferation. CD69 is useful as a lymphocyte activation marker because of its expression in the very early stage of activation (55). It is noteworthy that although some markers were initially defined to be restricted to specific types of activated cells, their distributions were subsequently found to be more general. For example, CD25 (IL-2R α-chain), often used as a T cell activation marker, is present on activated T, B, and NK cells and is expressed during ontogeny on pre-B and pre-T cells. CD80 and CD86, initially reported as B cell activation markers, are constitutively expressed by macrophages, DCs, and fibroblasts and can be induced on activated T cells (56, 57). Further, sensitivity of the detection method may be a limiting factor when studying low density Ags. Other molecules reported to be expressed by activated lymphocytes include CD152 (CTLA-4), CDw137 (4-1BB), CD134 (OX-40), DATK44 Ag (TABS), Ly-77 (GL7), CD45R (B220), CD30, CD95 ligand, CD43, Ly-6 family members, CD106, cytokine receptors, and the family of very late Ag adhesion molecules.

In this communication, we provide an update on mouse cell surface molecules, including lists of surface markers that, in combination with multiparameter flow cytometric analysis, can be used to trace cell lineages and activation state. Several novel markers mentioned here may prove useful in mouse immunology research. Most of the reported mAbs and their respective Ags are compiled in the CD and Ly charts. Undoubtedly, many molecules have not been included here. Scientists are encouraged to contact the authors and the Committee on Standardized Genetic Nomenclature for Mice to submit novel molecules for their inclusion in the future reports. In addition to this communication, information on the mouse genome and genetic markers is available on the worldwide web. The Mouse Genome Database can be accessed via http://www.informatics.jax.org. Information on germ-line disruptions of Ly/CD-encoding loci can be obtained at www.bioscience.org/knockout/knochome.htm or TBASE at www.gdb.org/dan/tbase.html.

We thank M. A. Reyes and Drs. C. Shih, K. Holmes, and V. Kumar for their helpful assistance with this manuscript, and B. R. Marshall for excellent editorial assistance.