OMIP-003: Phenotypic analysis of human memory B cells


Purpose and Appropriate Sample Types

This panel was developed to characterize the phenotypic diversity of human memory B cells, with an emphasis on discriminating cell subsets within both the conventional memory population (CD27+) and the more recently described isotype switched (IgD) population lacking expression of CD27 (1). It has been tested on fresh and cryopreserved peripheral blood mononuclear cells (PBMC), as well as bone marrow aspirates and tonsillar cells (Table 1). The multicolor panel described herein has been used extensively to analyze large numbers of PBMC samples obtained from healthy controls in steady state and in response to infection (HIV, influenza, respiratory syncytical virus) and vaccination (influenza, tetanus) as well as in hundreds of patients with autoimmune diseases (systemic lupus erythematosus (SLE), rheumatoid arthritis, Sjogren's syndrome, psoriatic arthritis and Type 1 diabetes) and conditions characterized by allogeneic immune responses (renal transplantion and chronic graft versus host disease). This panel is also being applied in a longitudinal study in which 150 SLE patients are to be followed quarterly for a period of 2 years.

Table 1. Summary table for OMIP-003
PurposePhenotyping memory B cells
Cell typeFresh or cryopreserved PBMC, bone marrow, and tonsil mononuclear cells
Cross referenceNone
Table 2. Reagents used for OMIP-003
  1. FITC, fluorescein isothiocyanate; PE, R-phycoerythrin; Cy, cyanine; PerCP, peridinin-chlorophyll protein; Aqua, LIVE/DEAD fixable aqua dead cell stain; Qdot, quantum dot; APC, allophycocyanin.

CD24SN3PE-Alexa610Differentiation, epithelial adhesion
CD21B-ly4PE-Cy5Activation, BCR co-receptor for antigen-bound C3d
CD3SP-34-2Pacific BlueExclusion
Dead cells Aqua (Pacific Orange)Exclusion
CD27CLB-27/1Qdot605Differentiation, receptor for CD70 on activated T cells
CD95DX2APCActivation; pro-apoptotic in the presence of activated CD95L T cells
VH4-34- encoded idiotype9G4BiotinAutoreactive B cells


Major human B cell subsets are currently defined by pauci-color flow cytometry protocols that typically include IgD, CD27, CD38, and CD24 to classify the major accepted peripheral blood populations (transitional, naïve, memory, and plasmablast subsets). By and large, the expression of other informative markers (including IgM, CD23, CD10, CD21, and CD95, as well as chemokine receptor expression) is assessed by parallel staining of several sample aliquots with different combinations of the aforementioned markers in four to five color schemes. These approaches suffer from multiple shortcomings including:1) the limited ability of the “defining” markers such as IgD, CD27, and CD38 to properly discriminate major populations;2) the inability to ascertain the actual co-expression of multiple markers in a single population possibly leading to faulty assumptions of extended phenotypes and, by extension, preventing the discovery of new subpopulations; and3) the need for larger number of cells to perform multiple stainings, a major practical limitation when dealing with rare samples. Combined, it seems obvious that limited use of available markers not only fails to differentiate multiple populations within the conventional core subsets, but could potentially lead to erroneous attribution of functional properties. To address these limitations, we have developed several multicolor panels to fully characterize human B cells. These multicolor panels share seven so-called anchor markers. Antibodies against CD19 and CD3, along with the Fixable Aqua Dead Cell Stain, allow the unambiguous identification of live CD19+CD3 B cells. The inclusion of four developmental markers (IgD, CD24, CD27, and CD38) in the same panel makes it feasible to compare and integrate these different classification schemes and provide precise identifications of the core human B cell subpopulations (2).

In addition to these anchor markers, each panel is extended with specific markers for further discrimination of memory, transitional/naïve B cells and plasma cell subsets, respectively. Thus, the incorporation of CD21, CD95, CD45/B220, and CXCR3 in the memory panel (Table 2), as described in this OMIP, provides information regarding the activation status and homing potential of both the CD27+ switched memory and CD27 switched memory B cells (Figure 1). The addition of the rat anti-human Ig idiotype 9G4 antibody completes a 12-color human memory B cell panel, and provides a useful measure of autoreactivity through the identification of B cells expressing autoantibodies encoded by the VH4-34 variable region gene (3).

Figure 1.

Classification of core human B cell subsets and phenotypic characterization of memory B cells in healthy subjects and patients with systemic lupus erythematosus. (A) Lymphocytes gated through the FSC-A vs SSC-A plot were further interrogated by the ratios of height to width in forward scatter and side scatter, as well as their ability to uptake the amine-reactive Aqua fluorescent dye to gate out cell aggregates and dead cells, respectively. Live CD19+CD3 B cells were then selected for analysis as shown in B. (B) The customarily used IgD/CD27 scheme classifies peripheral blood B cells into four core subsets (left panel): naïve (N: IgD+CD27); unswitched memory (Um: IgD+CD27+); IgDCD27+ switched memory (CD27+ Sm); IgDCD27 switched memory B cells (CD27 Sm). IgDCD27++ plasmablasts (PB) are a rare population in steady-state healthy subjects and can be better discriminated as CD27++CD38++ cells in the IgD fraction of a healthy subject who, in this example, received HPV vaccination 11 days before the analysis (middle panel). The rightmost panel shows that autoreactive 9G4+ B cells concentrate within the naïve compartment in healthy subjects (3). The comparisons among different B cell classification schemes made possible by this panel is discussed and illustrated in Supporting Information Figure 3. (C) Comparisons of IgD B cells between a healthy subject and a SLE patient illustrate the value of multicolor panels to reveal informative phenotypic differences in health and disease. As previously reported (2), B220 expression is lost during GC differentiation and thus CD27+ resting memory cells are predominantly B220−. In contrast, IgDCD27 cells, which as in this example may be the dominant memory population in active SLE, are predominantly B220+. This expanded subset is also characterized in SLE by the down-regulation of CD24 and CD21, markers expressed by the majority of PBL B cells in general and CD27+ memory B cells in particular. Loss of CD21 and up-regulation of CD95 have been independently associated with memory B cell activation (4, 5). Accordingly, expansions of activated switched memory subsets (CD27+ and CD27) are evident in active SLE. In contrast, CXCR3 expression (suggestive of migratory potential of activated cells to Th1 areas of systemic inflamed tissues) is concentrated in the IgDCD27+ memory subset. (D) The inclusion of CD21, CD95, and CXCR3 in the same panel allows informative co-localization of these markers. In contrast with CD27+ resting memory cells in the healthy control, a significant fraction of IgDCD27 cells lack expression of CD21, a feature consistent with activation. Yet, the lack of CD95 expression indicates that these two markers are not necessarily correlated. In SLE, CD95+ cells are greatly expanded in both memory subsets and the vast majority of IgDCD27CD95+ cells are also CD21. In contrast, CD95+ cells within the lupus IgDCD27+ memory are almost equally split between CD21+ and CD21, illustrating again that the expression of these markers is not necessarily reciprocal. The CXCR3+ fractions of both IgDCD27+ and IgDCD27 memory subsets exhibit CD21/CD95 expression patterns that reflect the differences generally observed between healthy subjects and SLE patients. As illustrated in this example, the expression of CD95 in IgDCD27CXCR3+ cells is increased in SLE. [Color figure can be viewed in the online issue, which is available at]

Similarity to Published OMIPs

None to date.