Every panel starts with the same question: which antigens will stand in for the population you care about. The answer is rarely one antigen, because the useful markers are shared across lineages, several are lost or gained as a cell matures, and a few of the most important populations have no positive marker of their own and are defined by exclusion. This page sets out what each lineage is usually identified by, what the common traps are, and where a marker that works in one species does not transfer.
Which B cell markers identify the lineage
B cells are normally taken by CD19, with CD20 and CD22 beside it and the CD79 heterodimer as the receptor-associated pair. CD19 is the workhorse because it appears early and persists through most of the lineage, but it is a poor sole marker in two situations that matter: after CD19-directed therapy, where the antigen is genuinely absent from cells that are still B cells, and at the plasma cell end, where it dims and CD20 is lost altogether. A panel that carries CD19 and CD20 together tells you where in the lineage you are, which one antigen cannot.
T cell markers, and why one is never enough
CD3 is the closest thing to a defining antigen in immunology, and it still needs company. Surface CD3 identifies mature T cells, but it appears in the cytoplasm before it reaches the surface, so an immature population reads as CD3-negative on a surface stain and CD3-positive when the cell is permeabilised. Past that, CD4 and CD8 divide the compartment and neither is exclusive to it, since CD4 also sits on monocytes and some dendritic cells. Activation antigens are a third axis rather than a lineage one, so treat them as state and not identity.
NK cell markers and the lineage-negative problem
Natural killer cells are the clearest case of a population defined partly by absence: the usual gate is CD3-negative first, then CD56 and CD16 to separate the cytotoxic majority from the cytokine-producing subset. That negative step is not a convenience, it is what keeps the CD56-positive T cells out, and a panel that omits CD3 will report them as NK cells. NKp46 is the closer thing to a positive lineage antigen and is the marker that transfers between species, which is why cross-species work tends to be built on it rather than on CD56.
Dendritic cell markers and plasma cell markers
Both of these lineages are reached by subtraction. Dendritic cells are taken as lineage-negative and HLA-DR-positive, and only then split by CD11c for the conventional subset and CD123 for the plasmacytoid one; without the exclusion step the gate fills with monocytes. Plasma cells run the other way: CD138 and bright CD38 identify them, CD20 has gone and CD19 is dim or absent, so the antigens that found the lineage are the ones that stop working at its end. In both cases the informative part of the panel is the channel you expect to be empty.
Questions people ask about immune cell markers
Can a single marker define a cell type?
Almost never. The useful antigens are shared across lineages and change with maturation and activation, so identity comes from a combination, and often from a combination that includes something the cell does not express.
Why are some populations defined by what they lack?
Because they have no antigen of their own that nothing else carries. NK cells and dendritic cells are both reached by excluding other lineages first, and dropping that step is the commonest reason a gate contains the wrong cells.
Which mouse immune cell markers differ from the human ones?
Enough that a human panel does not transfer. CD56 is not the mouse NK antigen, several CD designations have no mouse equivalent in common use, and clone cross-reactivity has to be checked per antibody rather than assumed from the target name.