Phenotypically distinct B cell development pathways map to the three B cell lineages in the mouse

CD138 Expression Distinguishes the Predominant Neonatal B Cell Development Pathway from the Predominant Pathway in Adults.

To identify the standard subsets in the B cell development pathway, i.e., Fr. A–F (11), we analyzed the stained adult or neonatal BM cells according to the scheme shown in Fig. 1. For this analysis, the BM cells are first gated to include only single cells (no doublets) and to include only live cells [propidium iodide (PI)⁻, Fig. 1]. Next, the single live cells are gated to include B cells and their progenitors, i.e., cells that are B220⁺ and dump⁻ (CD3⁻CD8⁻F4/80⁻Gr-1⁻). Finally, the gated B220⁺ cells are separated by CD43 expression: the CD43⁺ population contains pro-B cells (Fr. A, B, and C/C′) and the CD43⁻ population contains pre-B (Fr. D), immature B (Fr. E), and mature B cells (Fr. F).

Data in Fig. 2A Upper show that CD138 (a surface proteoglycan that binds to the extracellular matrix and is expressed on plasma cells and pre-B cells) (14) is not expressed or is expressed at low levels in the predominant B cell development subsets (Fr. A–F) in neonates (Fig. 2A Upper, shaded histograms). In contrast, the predominant B cell development subsets in adults express relatively high levels of CD138, from Fr. B onward. In Fr. F, which contains a mixture of newly developed B cells and recirculating B cells, approximately half of the cells still express CD138, and some of the others appear to have down-regulated CD138 to different extents (Fig. 2AUpper, unshaded histograms). Consistent with the loss of CD138 on mature B cells, none of the B cells in spleen express CD138 (data for splenic IgM⁺ IgD⁺ are shown in Fig. 2ALower).

Overall, we detect four levels of CD138 expression. Plasma cells, which have been gated out from the data shown here, express the highest CD138 levels, i.e., ≈10-fold above the CD138 levels expressed on CD138-expressing cells during B cell development. Gating on the Fr. A–C (a combined subset in which most of the cells are from Fr. B) shows three levels of CD138 expression (CD138^hi, CD138^int, and CD138⁻) and varying frequencies of cells at these expression levels depending on the source of the cells (Fig. 2B). In adult BM, the frequency of CD138^hi is the highest, and the frequency of CD138^int is greater than CD138⁻. In neonates, the reverse is true: The majority of cells are CD138⁻ cells, although some CD138^int and a few CD138^hi cells are detectable.

These striking differences in CD138 expression, which are fixed early in the B cell development, suggests that the three levels of CD138 expression may distinguish the progenitors for B-1a, B-1b, and B2 cells. Data presented in the next section support this hypothesis.

CD138 Expression as a Marker for B Lineage Progenitors.

Sorting and transfer studies conducted by Montecino-Rodriguez and colleagues (10) have identified subsets of cells in neonates and adults that are each selectively enriched for one of the lineage progenitors. In addition, Hardy and colleagues, in earlier studies (15–17), have shown similar distinctions between sorted B cell progenitors in fetal liver vs. adult BM. Results from our studies indicate that CD138⁻, CD138^int, and CD138^hi, respectively, mark the progenitors for B-1a, B-1b, and B-2 cells.

Montecino-Rodriguez and colleagues (10) demonstrated the presence of B-2 progenitors among AA4.1⁺CD19⁻ cells that express intermediate levels of the “CD45R” (B220-6B2) determinant detected by the B220/RA3–6B2 monoclonal antibody (circular gates, Fig. 3). This subset, which is basically Fr. A with AA4.1⁻ cells excluded and a proportion of CD19^low Fr. B included, is mainly present in adult BM.

In contrast, the B-1a progenitors, which are found mainly in neonatal BM, and the B-1b progenitors, which are found mainly in adult BM, are located in an AA4.1⁺CD19⁺ subset similar to Fr. B with respect to expression of these markers but containing cells that only express minimal levels of the B220-6B2 determinant (designated as B220-6B2^lo/neg). These cells, Montecino-Rodriguez and colleagues (10) show, express other B220 isoforms and hence are committed to the B lineage. Consistent with this commitment, and with the similarity of these subsets to Fr. B, we find that the cells in both subsets express CD24 at levels typical for Fr. B.

With hindsight, the low expression of B220-6B2 on B-1 progenitors is not surprising. Although this determinant is readily detectable on mature B-1 cells, it is usually expressed at lower levels on these cells than on B-2 cells (18, 19). In fact, the B220-6B2 monoclonal antibody was adopted for general use for B cell identification in our laboratory (and elsewhere) specifically because, although its expression overlaps on B-1 and B-2 cells, it provides an opportunity to estimate whether B-1 or B-2 cells predominate in a given B cell population (18).

Results from our studies with the progenitor subsets defined by Montecino-Rodriguez and colleagues (10) strongly support the idea that CD138 expression distinguishes early-stage B lineage progenitors. Data in Fig. 3 show that CD138^hi cells are enriched in the B-2 progenitor subset. CD138⁻ cells, in contrast, constitute ≈90% of the neonatal subset that gives rise to B-1a cells; CD138^int constitute ≈50% of the adult subset that predominantly gives rise to B-1b cells.

Analysis of this latter subset (AA4.1⁺, CD19⁺, and B220-6B2^−/lo BM from adult BM) is complicated by the presence of substantial numbers of cells at all three CD138 levels (Fig. 3). Although approximately half the cells express CD138^int, a quarter is CD138^hi, and the remainder is CD138⁻. The CD138⁻ cells in this subset likely reflect the presence of progenitors for B-1a cells, which this adult subset also reconstitutes (although to a lesser extent that B-1b). The CD138^hi cells, in contrast, likely represent contamination from the previously defined AA4.1⁺, CD19⁺ Fr. B, which expresses intermediate levels of B220-6B2 but does not otherwise differ from the B220-6B2^−/lo subset defined by Montecino-Rodriguez.

Collectively, the data cited above provide good reason to believe that the expression of CD138 on developing B cells (Fr. B–E) will prove to distinguish reliably between the progenitors for each of the B cell lineages. However, in the absence of data from sorting and transfer studies with this marker, these findings must still be considered provisional.

I-A Also Distinguishes the Predominant B Cell Development Pathway in Adults and Neonates.

Previously, I-A expression has been shown to be expressed from Fr. D during B cell development in adults, only to be expressed on mature (IgM⁺ IgD⁺) B cells in neonatal animals (12, 13). In a previous study (20), we detected I-A expression earlier than Fr. D in adults but were unable to obtain a clear view of the expression of this marker on the early subsets. Our current methods resolve these subsets better and show that cells expressing I-A expression become detectable in Fr. B in adults, mainly in the CD138^hi population and increase substantially in frequency in Fr. D (see Fig. 2B Lower). Neonates, in contrast, have virtually no detectable cells expressing I-A in pre-B cells and only a few cells (if any) expressing I-A in pre-B cells (Fig. 2B Upper) Comparison of CD138 and I-A expression in the predominant adult pathway indicates that CD138 is expressed on Fr. B before I-A is expressed, and that both markers are expressed by the majority of the cells in Fr. D (Fig. 2B Lower right).

CD138⁻ Pre-B Cells (Fr. D) from Neonates Differentiate to IgM⁺ Cells in Vitro Without Passing Through a CD138⁺ Stage.

To evaluate whether CD138 is transiently expressed during B-1 development, we sorted and cultured the predominant pre-B cell subsets in adult and neonatal BM, i.e., CD138^hi from adults and CD138⁻ from neonates and determined how/whether CD138 expression changes as these two types of pre-B cells mature (Fig. 4). Over 30% of the cultured pre-B cells were recovered as live cells after 36 h in culture in the absence of stromal or cytokine support; 30–40% of these cells had differentiated to IgM⁺ B cells in all cultures. The IgM⁺ B cells, derived from the sorted neonatal pre-B cells, retained their CD138⁻ phenotype and did not pass through a CD138⁺ expression as they matured. Among the IgM⁺ B cells derived from the adult (CD138^hi) pre-B cells, approximately one-third had down-regulated CD138 expression; the remainder were still CD138^hi, despite the initiation of surface Ig production (Fig. 4).

Preliminary studies with sorted pre-B cells cultured for 90 h indicates that ≈80% of the CD138⁺ pre-B cells sorted from adult BM express IgM and that the majority of these cells down-regulate CD138 expression and up-regulate IgD (data not shown). The expression of IgD, however, does not appear to coordinate with the down-regulation of CD138 (data not shown).

Mice and Tissue Preparation.

BALB/cN and C.B-17 mice were bred and maintained at the Department of Comparative Medicine at Stanford University. BM cells were extruded from femurs of adult or 3- to 4-day-old neonatal mice by using staining medium (RPMI medium 1640 + 3% newborn calf serum)-filled syringes, then washed, filtered through nylon mesh, and resuspended at 25 × 10⁶ cells per ml. Single-cell suspensions of spleens were prepared by first grinding the spleens with frosted slides and then by gentle passage through nylon mesh. The red blood cells were lysed by incubation with chilled 0.14 M NH₄Cl and 20 mM Tris (pH 7.4) for 2 min. These splenic cells were then washed with staining medium and resuspended at 25 × 10⁶ cells per ml.

FACS Analysis and Sorting.

Cell suspensions from BALB/cN or C.B-17 adult and neonate BM or spleen were prestained with anti-CD16/CD32 mAb to block FcγRII/III receptors and stained on ice for 15 min with the following fluorochrome-conjugated antibodies in a ten-color staining combination: FITC-CD19 (1D3) or FITC-I-A^d (AMS-132.1); PE-CD43 (S7); biotin-CD138 (281–2); PECy5.5-CD24 (30F1); PECy7-IgD (11–26); Alexa594-BP-1 (6C3); APC-B220 (RA3–6B2), referred to here as B220–6B2; APCCy5.5-AA4.1; APCCy7-IgM (331); and, as a “dump reagent” with which to gate out unwanted cells, a mixture of cascade blue-conjugated CD4 (GK1.5), CD8a (53–6.7), Gr-1 (8C5), and F4/80. After being stained with this stain set, the cells were washed with staining medium and stained again on ice for 15 min with streptavidin-CyChrome (BD Pharmingen), a second-step reagent that reveals biotin-coupled antibodies. The fluorochrome-conjugated antibodies were either purchased from BD Pharmingen or custom conjugated in the Herzenberg laboratory (Stanford, CA).

After washing away excess reagents, the stained cells were resuspended in staining medium containing 10 μg/ml propidium iodide. Cells that stained with propidium iodide (dead cells) were excluded from analysis and sorting in all cases. Cells were analyzed and sorted on the Stanford shared FACS facility Hi-Dimensional FACS (Hi-D FACS) instruments, either on the FACSAria with DiVa electronics (Becton Dickinson) or on a hybrid instrument (“Flasher II”) in which a FACS II bench is coupled to FACS DiVa electronics. Staining protocols were designed with facsxpert (http://sciencexperts.com), and data were analyzed with flowjo software (http://treestar.com).

Cell Culture.

Between 5 × 10⁴ and 8 × 10⁴ sorted pre-B cells from adult or neonatal BM were cultured in 96-well plates in 300 μl of deficient Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% FCS, 10⁻⁵ M 2-mercaptoethanol, 2 mM glutamine, and 100 μg/ml penicillin and streptomycin. Cells were cultured for 36 or 96 h at 37°C in 7% CO₂ ambient oxygen incubators before FACS analysis. No stromal lines were used, and no cytokines were added.