Mechanisms of Peritoneal B-1a Cells Accumulation Induced by Murine Lupus Susceptibility Locus Sle2

Mice

The production of the B6.NZMC4 congenic mice, now called B6.Sle2, has been detailed previously (19). The chromosome 4 NZM2410-derived congenic 26 cM segment represents the 95% confidence interval flanking Sle2 (17). B6.CD5^−/− mice (23) were a kind gift from Tim Behrens (University of Minnesota, Minneapolis, MN). B6.Sle2.CD5^−/− mice were bred by selection for the NZM2410 alleles of the Sle2 markers D4MIT17, brown, D4MIT9, and D4MIT72 and by CD5-negative staining by flow cytometry on PBLs. C56BL/6 mice (B6), B6.129S7-Rag1^tm1/mo (B6.Rag), and B6.SJL-Ptprc^a Pep3^b/BoyJ (B6.Ly5^a) mice were obtained from The Jackson Laboratory (Bar Harbor, ME) and subsequently bred at the University of Florida. The B6.C20 strain (B6.Igh^a) was originally obtained from G. Bosma (Institute for Cancer Research, Philadelphia, PA). All mice were housed in specific pathogen-free facilities. No sex differences were noted in the expression of any of the reported phenotypes, and both males and females were used. Unless specified, all experimental and age-matched control groups contained three to five mice. All experiments were conducted according to protocols approved by the University of Florida Institutional Animal Care and Use Committee.

Cell preparation

PerC cells were obtained by flushing the peritoneal cavity with RPMI 1640 supplemented with 10% FCS and 1 mM EDTA. PerC B cells were isolated using a two-step process. First, macrophages were depleted with two successive pannings of adherent cells for 2 h at 37°C and 5% CO₂. This was followed by depletion of T cells with Abs to the pan T cell marker Thy1.2 with Dynabeads according to the manufacturer’s instructions (Dynal Biotech, Oslo, Norway). The purified perC B cells were >95% B220⁺. Splenic single cell lymphocyte suspensions were prepared by deleting RBCs with 0.83% NH₄Cl and macrophages by panning, as earlier described.

Flow cytometry

Flow cytometric analysis was performed as previously described (18). Briefly, cells were blocked first with saturating amounts of anti-CD16/CD32 for 15 min in staining buffer (PBS, 5% horse serum and 0.09% sodium azide). Cells were then stained with allophycocyanin-, FITC-, PE-, or biotin-conjugated mAbs, followed by Streptavidin Quantum red conjugate (Sigma-Aldrich, St. Louis, MO). In some cases, Abs directly conjugated to CyChrome (BD Pharmingen, San Diego, CA) were used. mAbs CD5, (53-7.3), B220 (RA3-6B2), IgM^b (AF6-78), IgM^a (DS-1), IgM (II/41), CD11b (M 1/70), CD23 (B3B4), CD43 (S7), CD80 (16-10A1), CD86 (GL1), I-A^b (AF6), CD16/CD32 (2.4G2), CD19 (1D3), and their isotype controls were purchased from BD Pharmingen. The stained cells were analyzed on FACScan or FACSCalibur (BD Biosciences, Mountain View, CA).

CFSE labeling and cell adoptive transfer

Purified perC B cells pooled from 6-mo-old B6 or B6.Sle2 mice were labeled with 10 μM CFSE (Molecular Probes, Eugene, OR), for 30 min at room temperature as previously described (24). Age-matched B6 or B6.Sle2 host mice were exposed to 600 rad whole body sublethal irradiation (25) 1–2 days before i.p. injection of 3–4 × 10⁶ CFSE-labeled B cells in 0.5 ml of PBS with or without 50-μg AffiniPure F(ab′)₂ goat anti-mIgM (Jackson ImmunoResearch Laboratories, West Grove, PA). On day 3 or 7 after transfer, perC cells were analyzed by flow cytometry. Splenic lymphocytes (1 × 10⁷) from 6-mo-old B6 or B6.Sle2 mice were infused into 2-mo-old B6.Rag mice by i.v. injection. Two months after transfer, perC cells were analyzed by flow cytometry.

Proliferation assays

For in vivo assessment, 7-mo-old B6 and B6.Sle2 mice received drinking water containing 0.8 mg/ml BrdU (Sigma-Aldrich) renewed every 2 days for 9 days. PerC cells were stained for surface markers and BrdU incorporation with the FITC-BrdU Flow kit (BD Pharmingen). For in vitro proliferation, purified perC B cells from 3-mo-old B6 and B6.Sle2 mice or splenic B220⁺ B cells from 3-mo-old B6.CD5^−/− and B6.Sle2.CD5^−/− were incubated in duplicate at a density of 2 × 10⁶ cells/ml for 54 h in complete RPMI 1640 medium containing either 0.1–10 μg/ml LPS (Sigma-Aldrich) or 1–30 μg/ml F(ab′)₂ anti-IgM at 37°C and 10% CO₂. The higher CO₂ concentration was used after we found that the resulting slightly lower pH condition improved perC B cell proliferation (our unpublished observations). For the last 18 h of culture, 10 μCi³/ml [H]thymidine (ICN Pharmaceuticals, Costa Mesa, CA) was added.

Apoptosis and irradiation assays

Purified perC B cells (2 × 10⁷/ml) from 7-mo-old B6 and B6.Sle2 mice were cultured in duplicate in complete RPMI 1640 medium with 30 μg/ml F(ab′)₂ anti-IgM at 37°C and 10% CO₂ for 18 h. Cells were stained with surface markers and apoptosis was assessed by flow cytometry with the anti-activated caspase-3 FITC mAb apoptosis kit I (BD Pharmingen) according to the manufacturer’s instructions. For the irradiation assays, splenic and perC cells were exposed in duplicate to increasing doses of gamma radiation (Atomic Energy of Canada, Ottawa, Canada) up to 1200 rad. After overnight incubation (10⁶ cells/ml) in DMEM supplemented with 10% FCS, 200 nM glutamine, 10,000 U/ml each of penicillin/streptomycin, 10 mM nonessential amino acids, and 10⁻⁵ M 2-ME (Sigma-Aldrich), the cells were stained to define B-1a cells as IgM⁺ B220⁺ CD5⁺ and B-2 B cells as IgM⁺ B220⁺ CD23⁺. Just before flow cytometric analysis, the nuclear dye Yo-Pro-1 (Molecular Probes) was added at a final concentration of 10 μM (26). Calculation of percentage of viable cells was based on the percentage of size-gated cells that excluded Yo-Pro-1. Decreased forward scatter confirmed that these were the cells that were Yo-Pro-1-positive.

Preparation of chimeras

Production of mixed bone marrow chimeras was performed at a 1:1 ratio for the two donors. In all other respects, the chimeras were prepared as previously described (27). In brief, mice were lethally irradiated with two doses of 525 rad gamma irradiation (3 h apart) in a Gammacell 40 ¹³⁷Cs apparatus. Donor bone marrow cells were depleted of mature T cells and B cells by a mixture of anti-CD4, anti-CD8, anti-Thy-1.2 (mMT1), and anti-I-A^b (clone D3-137.5), followed by complement (Rabbit Lo-Tox; Jackson ImmunoResearch Laboratories). Ten million cells were given to each mouse by tail vein injection.

Statistics

Unless specified, two-tailed Student’s t test was used, and significance thresholds were indicated in the figures at p < 0.05, p < 0.01, and p < 0.001.

Sle2 results in an early increased number and proportion of perC B-1a cells

We have previously reported that Sle2 resulted in an increased accumulation of B-1a cells first in the perC and later in the spleen of aged mice (16). Since this early characterization, the B6.Sle2 mice have been re-derived in specific pathogen-free conditions. In these mice, a significant increase in number and proportion of B-1a was observed in 2-mo-old B6.Sle2 mice (Fig. 1⇓, A and B). The proportion of lymphocytes with a B-1a phenotype, determined as IgM⁺ B220^low CD5⁺CD23⁻CD43⁺CD11b⁺ (7), was 2-fold higher in B6.Sle2 than in B6 mice at either 2- or 7-mo-old. The number of perC B-1a cells was also 2-fold higher in 2-mo-old B6.Sle2 mice (7.5 × 10⁵±0.4) than in B6 mice (3.4 × 10⁵±0.7), but reached a nearly 10-fold difference by 7 mo of age (12.6 × 10⁶±2.1 vs 1.5 × 10⁶±0.5). A corresponding decrease in the proportion of B-2 cells was observed as shown by the B-1a to B-2 ratios. The total number of perC T cells was significantly higher in B6.Sle2 than in B6 mice at 7 mo of age (9.63 ± 0.04 × 10⁵ vs 3.99 ± 0.04 × 10⁵, p < 0.01). However, the percentage of perC CD3⁺ lymphocytes was lower in B6.Sle2 than in B6 mice, at both 2 mo (1.77 ± 0.06% vs 3.83 ± 0.26%, p < 0.01) and 7 mo of age (7.47 ± 4.78% vs 12.33 ± 8.59%, p < 0.03). As reported previously (16), no difference was obtained for the B-1b CD5⁻ sister population (data not shown). As also reported previously (16), an increased number of B-1a cells was found in the spleen of aged (≥7-mo-old) B6.Sle2 mice, but no difference was observed in 2-mo-old mice (data not shown). These results indicate that the Sle2-mediated accumulation of B-1a cells in the perC is an age-independent primary phenotype, whereas the splenic accumulation appears to be a secondary phenomenon. The early significant difference in number of perC B-1a cells observed between specific pathogen-free B6.Sle2 and B6 mice, but not in the same conventionally housed strains is mostly due to a relative decrease in the numbers and interindividual variability of perC B-1a cells in specific pathogen-free vs conventional B6 (data not shown). This is consistent with a larger pathogen load playing a role in the size of the perC B-1a compartment.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Sle2 results in an increased perC B-1a population. A, Representative lymphocyte-gated perC cells from 2-mo-old B6 and B6.Sle2 mice, showing the B220^low CD5⁺ B-1a and B220^high CD5⁻ B-2 populations. B, Number and percentage of perC B-1a cells, and B-1a to B-2 ratio in B6 (□) and B6.Sle2 (▦) mice at 2 and 7 mo of age (mean ± SD). C, A subset of B-1a cells in B6.Sle2 perC showed decreased levels of IgM (black box), and the percentage of these cells (indicated above the CD5⁺ IgM^low gates) increased with age. All data are shown for five mice per age group in each strain.

Expression of lineage (B220, CD5, CD23, CD43, and CD11b) and activation (CD80, CD86, and MHC class II I-a) surface markers was similar between B6 and B6.Sle2 perC B-1a cells (data not shown). Sle2 was associated, however, with an age-dependent accumulation of an IgM^low B-1a population (Fig. 1⇑C). The higher number and percentage of CD5⁺ IgM^low cells associated with Sle2 is not a mere consequence of the higher number of B-1a cells produced by this locus. The IgM^low to IgM^high ratio among CD5⁺ perC cells was significantly different between the two strains (1.25 + 0.31 vs 0.40 + 0.09, p = 0.016). These cells wereCD11b⁺CD138⁻ and did not have a plasma cell morphology. Cell surface marker expression was otherwise identical with that of the IgM^high population, and most likely reflects an activation-induced down-regulation of BCR expression.

Sle2 effect on the size of the B-1 cell population does not require CD5 expression

We investigated whether expression of CD5, a negative regulator of Ag signaling, was necessary for Sle2-mediated expansion of the perC B-1 population seen on the B6 background. The potential importance of CD5 was also suggested by the fact that expansion of the B1 population was confined to the CD5⁺ B-1a population and did not include the CD5⁻ B-1b cells. As expected (22), B6 and B6.CD5^−/− mice showed equivalent numbers and proportions of IgM^high B220^low CD11b⁺ B-1 cells in the perC (data not shown). More importantly, B6.Sle2 and B6.Sle2.CD5^−/− perC B-1 cells did not show any difference in number or proportion in mice up to 18-mo-old (Fig. 2⇓A). CD5 deficiency in B6.Sle2 mice did not affect either the proliferative response of splenic B220⁺ B cells to LPS or BCR cross-linking (Fig. 2⇓B). As previously shown (16), Sle2 resulted in a significant enhancement of B cell proliferation; CD5 deficiency, however, did not further enhance this proliferative response (Fig. 2⇓B). Except for the absence of CD5, no surface marker differences were seen when perC B-1 cells were compared between B6.Sle2 and B6.Sle2.CD5^−/− mice. Moreover, CD5 deficiency did not result in any difference in Ab production or pathology in B6.Sle2 mice (data not shown), contrary to what has been previously shown in anti-DNA transgenic mice (28). These result show that CD5 expression was not necessary for the Sle2-mediated perC accumulation, and that the absence of CD5 signaling did not produce any major phenotypical change in the Sle2 B-1 cells.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

CD5 deficiency does not alter Sle2 B cell phenotype. A, Lymphocyte-gated perC cells from 12-mo-old B6 and B6.Sle2 mice and 18-mo-old B6.CD5^−/− and B6.Sle2.CD5^−/− mice, showing an increased B220^low CD11b B-1 population (box inset and corresponding percentage) associated with Sle2. Representative samples from 10 mice per strain. B, Proliferative response of splenic B220⁺ B cells in response to LPS (10 μg/ml), anti-IgM F(ab′)₂ (30 μg/ml) or medium alone. Mean ± SD of three mice per group are shown. The differences between B6.Sle2 and B6.Sle2.CD5^−/− were not statistically significant.

Sle2 results in higher intrinsic proliferation of perC B-1a cells

One of the mechanisms by which Sle2 expression could produce higher numbers of the self-renewing B-1a cells is through a greater proliferation. We assessed this possibility both in vivo and in vitro. In vivo BrdU labeling showed that Sle2 B-1a cells spontaneously proliferate significantly more than B6 B-1a cells (Fig. 3⇓A). No difference was observed, however, for the B-2 cells in the same mice. In vitro stimulation of total perC B cells showed a greater response from Sle2 cells to LPS, but not to BCR cross-linking (Fig. 3⇓B). The same results were obtained with various concentrations of both LPS and anti-IgM (data not shown), indicating that receptor cross-linking alone was not sufficient to account for the in vivo results.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Sle2 results in higher proliferation of perC B-1a cells. A, In vivo spontaneous proliferation was significantly higher in B6.Sle2 B-1a cells than in B6, but no significant difference was observed for B-2 cells. Mean and SD of percentage of BrdU⁺ cells in five mice per strain. B, Sle2 perC B cells proliferated in vitro significantly more than B6 perC B cells in response to LPS (10 μg/ml), but not to BCR cross-linking (30 μg/ml anti-IgM F(ab′)₂). Mean ± SD of three mice per group. C, Sle2 (bold histogram) B-1a cells proliferated more than B6 (thin histogram) B-1a cells 7 days after transfer into the perC of either B6 (left) or B6.Sle2 (right) hosts. The percentages of CFSE-diluted proliferating cells in each transfer are indicated. Representative plots from five transfers of CFSE-labeled perC B cells in each group are depicted.

We then determined whether the increased proliferation of Sle2 B-1a cells was intrinsic to these cells by performing reciprocal adoptive transfers. PerC CSFE-labeled B cells from either B6 or B6.Sle2 mice were injected into the perC of either B6 or B6.Sle2 mice. These hosts were sublethally irradiated before transfer, which eliminated >90% of perC lymphocytes (data not shown). Sle2 B-1a cells proliferated more than B6 B-1a cells in both B6 and B6.Sle2 hosts (Fig. 3⇑C). This showed specifically that Sle2 B-1a did not require host factors for increased proliferation, and conversely, that B6.Sle2 host factors did not confer increased proliferation to B6 B-1a cells. Overall these results indicate that in the perC, Sle2 has only a minimal effect, if any, on B-2 proliferation, but that it confers an intrinsic ability for greater proliferation to the B-1a cells.

Sle2 B-1a cells displayed a greater resistance to cell death

Another mechanism by which Sle2 could induce greater numbers of perC B-1a cells was by increasing their resistance to cell death. Spontaneous and activation-induced apoptosis was compared in perC B cell culture between B6 and B6.Sle2 mice (Fig. 4⇓). With or without anti-IgM Ab in the culture, the amount of active caspase-3-positive B-1a cells was significantly reduced in B6.Sle2 mice, whereas no difference was observed for B-2 cells (data not shown). Similarly, Sle2 perC B-1a cells showed better persistence following radiation in vivo. B6 and B6.Sle2 mice were lethally irradiated and reconstituted with allotypically marked B6.Igh^a bone marrow, and the resulting chimeras were analyzed 6 mo later. As expected, host-derived B-2 B cells were absent in the bone marrow and spleen from both B6 and B6.Sle2 chimeras (data not shown). B6.Sle2 chimeras, however, retained a striking B-1a population from host origin (averaging ∼5–10% of the total B cell population), whereas no IgM^b+ cells were found in the perC of B6 chimeras (Fig. 5⇓). These results could indicate that Sle2 B-1a cells were more resistant to radiation-induced apoptosis. In vitro assays, however, failed to show such a difference between B6.Sle2 and B6 B-1a cells (data not shown). Alternatively, Sle2 may confer an increased resistance to radiation-induced apoptosis in a B-1a precursor population, or increased proliferation in a small number of surviving B-1a cells. Nonetheless, these experiments demonstrate Sle2-expressing perC B-1a cells benefit from a survival advantage in various cell death-inducing conditions.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Sle2 results in decreased apoptosis in perC B-1a cells. Sle2 B-1a cells expressed a decreased amount of active caspase-3 after 18 h in culture with or without BCR cross-linking (30 μg/ml anti-IgM F(ab′)₂) as compared with B6. No difference was observed for B-2 cells. Mean ± SD of three mice per group are shown.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

Significant numbers of Sle2 perC B-1a cells were found after lethal irradiation. PerC B-1a cells of B6.Sle2 origin persisted in vivo 6 mo after lethal radiation and reconstitution with B6.Igh^a bone marrow. B6.Sle2, but not B6, chimeras showed expansion of a population of CD5⁺ B cells of host origin (IgM^b).

Sle2 induces a greater production of B-1a cells from the fetal liver

Because B6.Sle2 mice show increased numbers of perC B-1a cells at an early age (Fig. 1⇑), we investigated whether Sle2 was associated with a greater production of B-1a cells by the fetal liver, which is the major source of B-1a cells (29). For this purpose, we produced mixed chimeras in which lethally irradiated B6.Ly5^a mice were reconstituted with an equal mixture of B6.Igh^a and B6.Sle2, or B6.Igh^a and B6 fetal liver cells. In these chimeras, lymphocytes derived from the two paired congenic strains developed in the same environment, and therefore any observed difference in a given cell population resulted from intrinsic properties. One year after transfer, fetal liver cells expressing Sle2 (Fig. 6⇓, right panels) gave rise to a greater number of B-1a cells in the spleen and in the perC than B6 fetal liver (Fig. 6⇓, middle panels). The Igh^a and Igh^b allotypes had different reconstitution efficiencies. Comparison of the B6.Igh^a plus B6.Sle2 to the B6.Igh^a plus B6 chimeras (Fig. 6⇓, A and B, C and D) showed, however, that this allotypic difference could not account for the striking difference in the numbers of IgM^a+ and IgM^b+ B-1a cells in the B6.Igh^a plus B6.Sle2 chimeras. Importantly, B6.Igh^a cells produced equivalent amounts of B-1a cells when developing with either B6 or B6.Sle2 cells. This latter result indicates that the greater production of B-1a cells by Sle2 fetal liver is not due to an extracellular factor, but to a greater intrinsic proliferation capacity of the B-1a cells or their precursors.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

Fetal liver cells of B6.Sle2 origin had an increased capacity to repopulate the B-1a cell compartment in adoptive transfer experiments. B6.Ly5^a mice were lethally irradiated and reconstituted with a combination of fetal liver cells from B6.Igh^a and B6 (A and C) or B6.Igh^a and B6.Sle2 (B and D). One year later, spleen (A and B) and perC (C and D) cells were evaluated by flow cytometry. In the left column, contour plots identify B cells by the combination of CD19 and B220. After gating on this population, contour plots of IgM^a vs CD5 (middle column) and IgM^b vs CD5 (right column) are displayed for representative mice from five chimera per group. CD45.1 staining showed no residual B cells of host origin (data not shown).

Lymphoid tissues from adult B6.Sle2 mice provide a source of perC B-1a cells

A last mechanism by which Sle2 expression could result in a greater number of perC B-1a cells was through a contribution of the adult lymphoid organs to the perC B-1a pool. Although the majority of B-1a cells are fetal-derived and self-renewing, the existence of B-1a precursors in adult spleen (30) or bone marrow (31) has been demonstrated in other models. We investigated the hypothesis that an increased output of perC B-1a cells could occur in adult B6.Sle2 mice.

We first took advantage of the fact that the perC B-1a cell compartment is selectively maintained during homeostatic reconstitution (32, 33). Splenic lymphocytes from either B6 or B6.Sle2 mice were transferred by i.v. injection into B6.Rag mice. The number of transferred splenic B-1a cells was equivalent (∼3%) between the two strains. Eight weeks after transfer, the mice that received B6.Sle2 splenocytes contained a greater number and proportion of B-1a cells in their perC than the mice that received B6 splenocytes (Fig. 7⇓). Interestingly, the percentage of perC B-1a cells and the B-1a to B-2 ratios in both types of chimeras were very similar to the values obtained in unmanipulated 2-mo-old B6.Sle2 and B6 mice (although the absolute numbers were lower). This suggests the existence of an intrinsic homeostatic level of perC B-1a that is affected by the expression of Sle2.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 7.

Splenocytes of B6.Sle2origin showed an increased capacity to reconstitute the perC B-1a population. A, Representative samples of perC lymphocytes from B6.Rag mice 2 mo after adoptive transfers of 1 × 10⁷ RBC-depleted splenocytes from either B6 or B6.Sle2 mice. B, Number and percentage of perC B-1a cells, and B-1a to B-2 ratio in chimera 2 mo after transfer of B6 (□) or Sle2 (▦) splenocytes. The percentage of perC B-1a and B-1a to B-2 ratio in unmanipulated 2 mo B6 and B6.Sle2 mice are also shown in hatched bars. Mean ± SD of five mice per group are shown.

We also determined whether B-1a cells were produced by adult bone marrow in B6.Sle2 mice with bone marrow chimeras in lethally irradiated B6.Igh^a hosts. One year after transfer, all spleen and perC B cells were, as expected, of donor origin (Fig. 8⇓B). As expected, B6 bone marrow produced only a small amount of IgM^b+ B220⁺ CD5⁺ cells. B6.Sle2 bone marrow, however, produced large amounts of B-1a cells that were especially abundant in the perC (Fig. 8⇓A). Overall, these results show that substantial numbers of B-1a cells are produced by adult lymphoid tissues in B6.Sle2 mice.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 8.

Bone marrow of B6.Sle2 origin showed an increased capacity to reconstitute the B-1a subset. A, Lethally irradiated B6.Igh^a mice 1 year after reconstitution with bone marrow from either B6 or B6.Sle2 mice showed increased B-1a cells populations in the perC (right) and to a lesser extent in the spleen (left). B, In the spleen and perC of the B6 and B6.Sle2 chimeras, B cells were of donor origin (IgM^b). These are representative plots from five chimeras per strain.