Differences in mouse and human nonmemory B cell pools

Abstract

Identifying cross-species similarities and differences in immune development and function is critical for maximizing the translational potential of animal models. Coexpression of CD21 and CD24 distinguishes transitional and mature B cell subsets in mice. In this study, we validate these markers for identifying analogous subsets in humans and use them to compare the nonmemory B cell pools in mice and humans, across tissues, and during fetal/neonatal and adult life. Among human CD19(+)IgM(+) B cells, the CD21/CD24 schema identifies distinct populations that correspond to transitional 1 (T1), transitional 2 (T2), follicular mature, and marginal zone subsets identified in mice. Markers specific to human B cell development validate the identity of marginal zone cells and the maturation status of human CD21/CD24 nonmemory B cell subsets. A comparison of the nonmemory B cell pools in bone marrow, blood, and spleen in mice and humans shows that transitional B cells comprise a much smaller fraction in adult humans than mice. T1 cells are a major contributor to the nonmemory B cell pool in mouse bone marrow, in which their frequency is more than twice that in humans. Conversely, in spleen, the T1:T2 ratio shows that T2 cells are proportionally ∼ 8-fold higher in humans than in mice. Despite the relatively small contribution of transitional B cells to the human nonmemory pool, the number of naive follicular mature cells produced per transitional B cell is 3- to 6-fold higher across tissues than in mice. These data suggest differing dynamics or mechanisms produce the nonmemory B cell compartments in mice and humans.

Figure 1. CD21 and CD24 co-expression identifies distinct subsets of B cells in mouse and human tissues

A, Diagram of gating strategy to exclude B cell precursors and class-switched memory B cells prior to B cell subset identification using co-expression of CD21 and CD24. Identified are transitional 1 (T1), transitional 2 (T2), naive follicular mature (FM), and marginal zone plus marginal zone precursor (MZ) subsets. B-E, Mononuclear cells from indicated BALB/c mouse and human tissues (spleen, peripheral blood [PB], and bone marrow [BM]) were stained for flow cytometry to detect CD19, IgM, CD21, and CD24. CD19+IgM+ cells falling in lymphocyte light scatter were gated and B cell subsets as diagramed in A were gated. Data shown are representative of mouse spleen: n=4, human spleen: n=4, human PB: n=33, human BM: n=10.

Figure 2. CD1c and CD27 co-expression validate the identity of the CD21/CD24 MZ subset in human spleen

Human splenic cells were co-stained for flow cytometry to detect CD19, IgM, CD24, CD21, and indicated markers. CD19+IgM+ cells falling in lymphocyte light scatter were gated. A, Co-expression of CD1c and CD27 was plotted (left panel). CD1c–CD27– and CD1c+CD27+ populations were gated and their distribution with respect to CD21/CD24 subsets was plotted (right panels). B, CD19+IgM+ cells were gated into CD21/CD24 subsets (left panel). Co-expression of CD1c and CD27 in the CD21CD24 MZ (top right panel) and FM (bottom right panel) subsets is shown. C, CD19+IgM+ cells were gated into CD21/CD24 subsets and expression of IgM and IgD in the MZ and FM subsets is shown in histograms. D, CD24 and CD38 co-expression in gated CD19+IgM+CD27− cells was plotted. Gates for memory, transitional and mature B cell subsets based on the CD24/CD38 identification schema are shown (left panel). Cells in the mature subset were gated and their distribution with respect to CD21/CD24 subsets is plotted (right panels). Data shown are representative of n=4 adult human spleens.

Figure 3. Human markers validate the developmental status of B cell subsets identified by CD21 and CD24 co-expression

A, Diagram of gating strategy to exclude both CD27+ and CD27− IgM+ memory B cells in human tissues. B-C, PB and BM cells were co-stained for flow cytometry to detect IgM, CD27, CD38, CD24, and CD21. IgM+CD27− cells in lymphocyte light scatter were gated as shown to identify memory, transitional and mature B cell subsets based CD24/CD38 expression as shown (left panel). Cells in the mature and transitional subsets were gated and their distribution with respect to CD21/CD24 subsets is shown (right panels). D, IgM+CD27− cells in the CD24/CD38 non-memory gate (dashed oval in panel A) were gated into CD21/CD24 subsets and evaluated for CD38, CD10 and CD5 expression. Data shown are representative of n=27 adult human PB and n=10 adult human BM.

Figure 4. Patterns of differential BAFF-R expression are similar in mouse and human CD21/CD24 B cell subsets

A-C, Mononuclear cells isolated from adult mouse and human spleen, PB, and BM were stained for IgM, CD21, CD24, BAFF-R and in human samples for CD27, CD38 and CD19 as well. IgM+ B cells falling in lymphocyte light scatter in mouse tissues were gated into CD21/CD24 subsets. Human PB, BM, and splenic T1, T2, and FM cells were gated as described in Fig. 3. For MZ cells in human spleen, CD19+IgM+ cells were assessed for CD21/CD24 MZ phenotype and this included both CD27+ and CD27− cells. For each tissue, graphed are the means + SEM of relative BAFF-R expression in each CD21/CD24 subset (normalized to BAFF-R levels in the T1 subset in that tissue). Data are from n=9 mouse spleen, n=7 mouse PB; n=9 mouse BM, n=4 human spleen, n= 15 human PB and n=9 human BM. Statistical differences are shown as *p < .05; **p< .01; *** and p < .001.

Figure 5. Use of CD21/CD24 to compare emerging B cell pools during fetal/neonatal development in mice and humans

Mononuclear cells were isolated from human cord blood (CB), human fetal spleen and pooled mouse fetal spleens. Cells were stained for flow cytometry to detect CD19, IgM, CD21, CD24, and in human samples for CD38, CD10, and CD5 as well. A-C, Cells falling in lymphocyte light scatter and that were CD19+IgM+ or IgM+ were gated into CD21/CD24 subsets. D-E, Subsets in human CB and human fetal spleens were evaluated for CD38, CD10 and CD5 expression. F-H, Graphed are the frequencies ±SD that each of the CD21/CD24 subsets contribute to the composition of the B cell pool. The mean ±SD for subsets in each tissue was derived from: n=10 human CB; n=6 human fetal spleens obtained at 18–23 weeks, (pregnancy date); and pooled mouse fetal spleen from C3H/HeN mice, day 18 of 20 day gestation, assayed in 3 independent experiments. Flow cytometry histogram and dot plots are representative data from indicated tissues.

Figure 6. Transitional B cells are reduced in the non-memory B cell pool in humans as compared to mice

Mononuclear cells were isolated from adult mouse and human spleen, PB, and BM. Cells were stained for IgM, CD21, CD24, and in human samples for CD27, CD38 and CD19 as well. A-C, IgM+ B cells falling in lymphocyte light scatter were gated in mouse tissues. For human spleen, CD19+IgM+ cells falling in lymphocyte light scatter were gated. Human PB and BM were gated as described in Fig. 3. CD21/CD24 gates are as shown. D, of the IgM+ B cell pool with respect CD21/CD24 subsets gated as described in results to include CD27+ and CD27− MZ cells. E-F. Pie graphs showing the composition of the non-memory B cell pool with respect CD21/CD24 subsets. G-I, Graphed are the mean ± SD of the percentages that each CD21/CD24 subset contributes to the IgM+ B cell pool in spleen and the non-memory B cell pool in PB and BM. Graphed are data from n=16 adult mouse spleen; n=6 adult mouse PB; n=9 adult mouse BM; n=4 adult human spleen; n= 33 adult human PB; and n= 11 adult human BM; Mice were adult male BALB/c and non-pregnant female C3H/HeN mice. Statistical differences between analogous subsets are shown as *p < .05; **p< .01; *** p < .001; and **** p< .0001.