Discriminating gene expression profiles of memory B cell subpopulations

Abstract

Morphologically and functionally distinct subpopulations of human memory B (B(Mem)) cells are identifiable by either their expression of CD27 or Fc receptor-like 4 (FCRL4), an immunoglobulin domain containing a receptor with strong inhibitory potential. We have conducted comparative transcriptome and proteome analyses of FCRL4(+) and FCRL4(-) B(Mem) cells and found that these two subsets have very distinctive expression profiles for genes encoding transcription factors, cell-surface proteins, intracellular signaling molecules, and modifiers of the cell-cycle status. Among the differentially expressed transcription factors, runt-related transcription factor 1 (RUNX1) transcript levels were up-regulated in FCRL4(-) cells, whereas RUNX2 transcripts were preferentially detected in FCRL4(+) cells. In vitro evidence for FCRL4 promoter responsiveness and in vivo promoter occupancy suggested that RUNX transcription factors are involved in the generation of these B(Mem) cell subpopulations. A distinctive signature profile was defined for the FCRL4(+) B(Mem) cells by their expression of CD11c, receptor activator for nuclear factor kappaB ligand, and FAS cell-surface proteins, in combination with increased levels of SOX5, RUNX2, DLL1, and AICDA expression. We conclude that this recently identified subpopulation of B(Mem) cells, which normally resides in epithelial tissue-based niches, may serve a unique role in mucosal defense and, conversely, as a target for neoplastic transformation events.

Figure 1.

Transcriptome analysis of FCRL4⁺ and FCRL4⁻ B_Mem cells. The transcripts differentially expressed by FCRL4⁺ and FCRL4⁻ B_Mem from three independent tonsillar cell preparations are displayed. Up- and down-regulated transcripts are indicated in red and green, respectively. The magnitude of expression is depicted by the color bar. Arrowheads indicate transcripts whose expression was validated by quantitative RT-PCR and/or by FACS analysis of other tonsillar samples.

Figure 2.

Genes differentially expressed by FCRL4⁺ and FCRL4⁻ B_Mem cells. (A) RANKL transcript (top) and protein (bottom) levels in FCRL4⁺ and FCRL4⁻ B_Mem cells. RANKL mRNA levels were determined by quantitative RT-PCR normalized to RP-2. Values represent mean ± SEM (n = 3), and statistical significance was determined using the paired Student's t test. RANKL protein was quantitated by examining the mean fluorescence intensity levels for RANKL on FCRL4⁺ and FCRL4⁻ B_Mem cells, normalized to the mean fluorescence intensity levels of isotype control antibodies in the corresponding cell populations. Values represent mean ± SEM (n = 16), and statistical significance was determined using the paired Student's t test. (B) Quantitative RT-PCR analysis of the indicated src family kinases (top), or DLL1 and transcription factor transcripts (bottom) in FACS-sorted FCRL4⁺ and FCRL4⁻ B_Mem cells. Values represent mean ± SEM (n = 4).

Figure 3.

Expression of cell-surface antigens on human B_Mem cell populations. FACS analysis of tonsillar lymphocytes was performed for the indicated markers by gating on the B_Mem cell population (CD19⁺, IgD⁻, CD38⁻). Representative FACS profiles from at least 12 analyzed tonsillar preparations are depicted.

Figure 4.

Cell-cycle phenotype of FCRL4⁺ and FCRL4⁻ B_Mem cells. (A) Quantitative RT-PCR analysis of the indicated transcripts. Values were normalized to RP2 expression and are displayed as the mean ratio of FCRL4⁺/FCRL4⁻ ± SEM (n = 5 independent PCRs on independent tonsil preparations). (B) Staining of FACS-sorted FCRL4⁺ and FCRL4⁻ B_Mem cells for DNA content and Ki67 expression. Shown is a representative example of four independent tonsillar B_Mem cell preparations. Numbers in the top right of each FACS blot indicate the percentage of cells in each quadrant.

Figure 5.

RUNX1 responsiveness of the FCRL4 gene promoter. (A) Analysis of the putative promoter sequence of the FCRL4 gene. Locations are indicated for potential RUNX (open circles) and SOX5 (closed triangle) binding sites. (B) Luciferase reporter gene studies on transiently transfected 293T cells. Cells were transfected with the putative promoter of the FCRL4 gene controlling expression of the luciferase reporter gene and expression vectors containing the indicated transcription factors. RUNX2 Δ5 and RUNX2 Δ5,7 are splice isoforms that lack exon 5 and exons 5 and 7, respectively (see also Fig. S1, available at http://www.jem.org/cgi/content/full/jem.20072682/DC1). Values represent mean ± SD (n = 5). (C) ChIP assay of RUNX1 binding in FACS-sorted FCRL4⁺ and FCRL4⁻ B_Mem cells. Positions of the amplified PCR fragments are indicated by closed rectangles in A. Open bars indicate signals obtained from FCRL4⁺ cells, and shaded bars indicate signals obtained from FCRL4⁻ cells. All values were normalized to the PCR signal obtained from control immunoprecipitates (anti–SHP-1), as well as to unrelated, nonspecifically copurified DNA (CCR4). Values represent mean ± SEM (n = 5). Statistical significance of P < 0.05 is indicated by an asterisk.

Figure 6.

Differentially regulated transcripts in FCRL4⁺ and FCRL4⁻ B_Mem cells in comparison with other stages in B cell differentiation. Analysis of expression levels of the indicated transcripts in naive B, pre–GC B (preGC B), GC B (GC B), FCRL4⁺ and FCRL4⁻ B_Mem, and plasma cells. All values are normalized to RP2 expression. Error bars indicate one SEM (n = 3).