Cutting Edge: Essential Role of Phospholipase C-γ2 in B Cell Development and Function

Materials and Methods

Generation of MxCre/PLC-γ2^flox/flox mice

A targeting vector to generate PLC-γ2^flox/flox was designed so that the PIP₂ binding site in the PLC-γ2 X domain could be deleted by expression of Cre protein. The PstI-BamHI-digested 1.7-kb fragment of the PLC-γ2 genomic DNA was inserted between two loxP sites for the vector pKSTKNEOLOXP, which has HSV thymidine kinase and loxP-flanked pGK-neo. Then, the PstI-PstI 7.9-kb fragment 5′ upstream and the 1.0-kb fragment 3′ downstream of the exon containing the PIP₂ binding site were inserted (see Fig. 1⇓A).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Inducible targeting of the PLC-γ2 gene. A, Structures of the wild-type PLC-γ2 allele (wt), the targeting vector, and the mutated allele are shown. The X domain is essential for the catalytic activity of PLC-γ2. Open rectangles () and open triangles (▹) represent exons and loxP sites, respectively. The restriction endonuclease cleavage sites are abbreviated as P (PstI) and B (BamHI). B, PCR analysis of genomic DNA using primers a, b, and c as indicated in A. A novel 308-bp band was observed in iPLC-γ2^flox/flox mice (corresponding to Δ). C, Western blot analysis of splenic B cells obtained from iPLC-γ2^flox/+ and iPLC-γ2^flox/flox mice. Cell lysates were immunoprecipitated with anti-PLC-γ2, anti-PLC-γ1, and anti-BLNK Abs and blotted with the same Abs. The PLC-γ2 immunoprecipitate was diluted by 3-fold serial dilutions.

ES cells were transfected with this targeting vector, G-418-resistant clones were screened by PCR, and positive clones were subjected to Southern blot analysis using a 3′ probe (shown in Fig. 1⇑A). Cre protein was transiently expressed in the targeted ES clones to delete the loxP-flanked neo gene. Two types of clones, PLC-γ2 flox and PLC-γ2Δ, were obtained (see Fig. 1⇑A). The PLC-γ2 flox ES clones were used to generate chimeric mice, which successfully contributed to the germline transmission. Mice homozygous for the loxP-flanked PLC-γ2 gene (PLC-γ2^flox/flox) were born at the expected Mendelian ratios and presented no obvious abnormalities. PLC-γ2^flox/+ mice were mated with MxCre mice to generate MxCre/PLC-γ2^flox/+ mice. These mice were further mated with PLC-γ2^flox/flox mice to generate MxCre/PLC-γ2^flox/flox mice.

For analysis of genotypes (see Fig. 1⇑B), ∼0.5 μg of DNA was subjected to 35 cycles of amplification with each cycle consisting of 0.5 min at 94°C, 1 min at 64°C, and 1 min at 74°C followed by an extension of 10 min at 74°C on a thermal cycler.

Results

To inducibly disrupt the PLC-γ2 gene, we crossed two types of mice: one is a MxCre mouse that bears the Cre transgene under the control of type I IFN-inducible Mx promoter (12), and the other is a mouse in which an exon encoding the PIP₂ binding site of the PLC-γ2 gene is flanked by two loxP sites (PLC-γ2^flox/flox) (Fig. 1⇑A; see Materials and Methods for a detailed description in making this mouse). To assess the efficacy of the PLC-γ2 inactivation, newborn mice were injected i.p. with the IFN-inducer poly(I-C) (300 μg) at 2-day intervals after birth without knowledge of their genotype. Genotyping was performed 21 days after birth. Genomic DNA was extracted from various tissues and subjected to PCR analysis using three primers as indicated in Fig. 1⇑A. In the spleen, bone marrow, lymph nodes, peritoneum, and thymus of MxCre/PLC-γ2^flox/flox mice, Cre-mediated deletion resulted in the appearance of a novel 308-bp band corresponding to the PLC-γ2Δ gene. Despite the appearance of the PLC-γ2Δ-specific band, a 365-bp band corresponding to the floxed PLC-γ2 gene was still detectable in these cells, indicating that Cre-mediated deletion did not occur completely at the DNA level (Fig. 1⇑B). In contrast to the high deletion efficiency in the spleen, bone marrow, lymph nodes, and thymus, this efficiency was relatively low in the peritoneum. Western blot analysis of splenic B cells also detected a small amount of residual wild-type PLC-γ2 protein in poly(I-C)-treated MxCre/PLC-γ2^flox/flox mice (Fig. 1⇑C). Poly(I-C)-treated MxCre/PLC-γ2^flox/flox (hereafter iPLC-γ2^flox/flox) mice showed no gross abnormalities compared with iPLC-γ2^flox/+. The expression level of PLC-γ1 in the iPLC-γ2^flox/flox splenic B cells was almost the same as that in iPLC-γ2^flox/+ (Fig. 1⇑C).

T and B cell development can be assessed by flow cytometric analysis of lymphocyte populations stained with Abs to various surface Ags. Staining of thymocytes with Abs to CD4 and CD8 revealed that normal T cell development occurred in both iPLC-γ2^flox/+ and iPLC-γ2^flox/flox (Fig. 2⇓E). The absolute numbers of thymocytes and splenic T cells of iPLC-γ2^flox/flox were comparable to iPLC-γ2^flox/+ (Table I⇓). In contrast, the numbers of B220⁺IgM⁺ mature B cells in the spleen, bone marrow, and lymph nodes of the iPLC-γ2^flox/flox mice showed a 2- to 3-fold reduction compared with those in control littermates (Fig. 2⇓, A–C). Slight reduction of the population of B220⁺CD43⁻IgM⁻ pre-B cells and concomitant increase of the B220⁺CD43⁺IgM⁻ pro-B cells were observed in the bone marrow of the iPLC-γ2^flox/flox mice (Fig. 2⇓A and Table I⇓). In both spleen and lymph nodes (Fig. 2⇓, B and C), the population of the mature B cells characterized as IgM^lowIgD^high was reduced in the iPLC-γ2^flox/flox mice, whereas the population of immature B cells (IgM^highIgD^low) was considerably increased. In addition, the number of the CD5-expressing B1 B cells in the iPLC-γ2^flox/flox mice, usually found in the peritoneum, were 2-fold less than those of control littermates (Fig. 2⇓D).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

FACS analysis of iPLC-γ2^flox/flox and littermate mice. Total cells isolated from various tissues of 7- to 14-wk-old control (iPLC-γ2^flox/+) and age-matched mutant (iPLC-γ2^flox/flox) mice were stained with the indicated Abs. Cells were analyzed by CellQuest (Becton Dickinson) after the appropriate gate settings. In A, the upper panels show the B220 vs IgM profiles of CD43⁻ gated cells, whereas the lower panels show the B220 vs CD43 profiles of IgM⁻ gated cells. In B, the percentages (mean ± SE; n = 4) of IgM^highIgD^low, IgM^highIgD^high, and IgM^lowIgD^high cells in iPLC-γ2^flox/+ and iPLC-γ2^flox/flox were 5.7 ± 2.9 vs 11.4 ± 3.3, 31.9 ± 3.7 vs 27.6 ± 6.6, and 15.1 ± 5.0 vs 3.0 ± 1.0, respectively. In D, the percentages of CD5⁺IgM⁺ cells in iPLC-γ2^flox/+ and iPLC-γ2^flox/flox were 23.3 ± 1.7 vs 10.7 ± 1.5.

The effect of PLC-γ2 on the immune response was analyzed for both T-independent (TNP-Ficoll) and T-dependent Ags (TNP-KLH). As shown in Fig. 3⇓A, in contrast to the control littermates, the iPLC-γ2^flox/flox mice were found to have reduced anti-TNP IgM responses to the T cell-independent Ag, when examined by PFC assays (17). In the T-dependent responses, anti-TNP IgM responses were also reduced, whereas anti-TNP IgG responses were comparable to those of the control littermates. Consistent with these data in the iPLC-γ2^flox/flox mice, these mutant mice exhibited reduced in vitro T-independent and -dependent IgM responses (Fig. 3⇓B).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Humoral responses in iPLC-γ2^flox/flox mice in vivo and in vitro. A, Ab-forming cells were measured by the PFC assay 7 days after the immunization of TNP-Ficoll and 7 days after the secondary challenge of TNP-KLH. iPLC-γ2^flox/+ and iPLC-γ2^flox/flox are shown as • and ○, respectively. B, In vitro PFC responses were analyzed 5 days after the culture with TNP-Ficoll and SRBC; symbols are the same as A.

In vitro B cell functions were further assessed by calcium mobilization and proliferation assays. When stimulated by anti-IgM Ab, splenic B cells from the iPLC-γ2^flox/flox mice mobilized Ca²⁺ to a much lower extent than control B cells (Fig. 4⇓, A and B). In contrast, CD3-mediated Ca²⁺ mobilization from the iPLC-γ2^flox/flox splenic T cells was essentially the same as that from the control T cells (Fig. 4⇓A). The proliferative responses were also perturbed; the iPLC-γ2^flox/flox splenic B cells stimulated via the BCR by treatment with anti-IgM or anti-IgM+IL-4 showed about a 10-fold reduction in proliferative responses. Stimulation with anti-CD40 Ab was also reduced in the mutant cells, while LPS-dependent proliferation was indistinguishable between iPLC-γ2^flox/flox and iPLC-γ2^flox/+ (Fig. 4⇓B). In contrast to abnormalities in B cells, anti-CD3-induced T cell proliferative responses in the iPLC-γ2^flox/flox mice were the same as those in the iPLC-γ2^flox/+ mice (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Calcium mobilization and proliferative responses of B cells in vitro. A, Ca²⁺ mobilization of splenic B and T cells from iPLC-γ2^flox/flox mice. Splenic B cells were stimulated with goat anti-mouse IgM Ab (15 μg/ml), followed by ionomycin (1 μM). Splenic T cells were stimulated with biotinylated anti-CD3 Ab (10 μg/ml) followed by cross-linking with streptavidin (10 μg/ml). Cell surface expression of IgM on splenic B cells used in A was shown in B. C, Splenic B cells of control (iPLC-γ2^flox/+) and mutant (iPLC-γ2^flox/flox) mice were stimulated with the indicated reagents, and the incorporation of [³H]thymidine was measured.

Discussion

In the present study, we have generated and characterized mice lacking PLC-γ2 in an inducible fashion. Consistent with the predominant expression of PLC-γ1 in T lineage cells (8), T cell number, development, and function were apparently normal in iPLC-γ2^flox/flox mice. Thus, as reported previously, poly(I-C) treatment by itself appears to affect no gross abnormalities of lymphocyte development and function (12, 18).

Flow cytometric analysis of iPLC-γ2^flox/flox mice exhibits defects in B cell development at the pre-B cell stage. The transition of B220⁺CD43⁻ pre-B cells from B220⁺CD43⁺ pro-B cells in the bone marrow is perturbed, which, in turn, limits their differentiation into B220^highIgM⁺ mature B cells in the bone marrow. Moreover, more immature B cells are present in the spleen of mutant mice. Hence, in the absence of PLC-γ2, the signals from the pre-BCR and BCR is less efficient in promoting further maturation stages. These developmental deficits in the iPLC-γ2^flox/flox mice are similar to those in BLNK-deficient mice (19, 20, 21, 22), strengthening the previous model that PLC-γ2 and BLNK are components of a common BCR signaling pathway (23, 24). However, the more severe developmental block from the B220⁺CD43⁺ to B220⁺CD43⁻ stage observed in the BLNK^−/− mice may suggest the requirement for additional downstream targets of BLNK in pre-BCR signals.

The iPLC-γ2^flox/flox mice have a phenotype that is also similar to that in mice lacking Btk (13, 14, 15). Both mutant mice show reduced levels of mature B cells in the bone marrow and periphery; a poor humoral response to T-independent Ags; and a poor proliferative response upon BCR stimulation. Thus, these data strongly suggest that the functional roles of Btk and PLC-γ2 lie along the common signaling pathways initiated by BCR. In contrast to Btk^−/− mice, the incomplete blockade of B1 B cell differentiation (Fig. 2⇑D) could be accounted for by the low efficacy of Cre-mediated deletion in the peritoneum (Fig. 1⇑B) and/or the residual B1 B cells before inducible targeting. Another marked difference between iPLC-γ2^flox/flox and Btk^−/− is the in vitro proliferative response via LPS (Fig. 4⇑C); in contrast to the reduced response in Btk^−/− mice, this responses is not significantly affected in the iPLC-γ2^flox/flox mice. The straightforward explanation of these results is that both Btk and PLC-γ2 are associated with BCR signaling, while the signal through LPS uses Btk, but not PLC-γ2. Because a small amount of PLC-γ2 still remains in the iPLC-γ2^flox/flox splenic B cells (Fig. 1⇑C), we cannot exclude the alternative possibility that this residual B cell fraction expressing PLC-γ2 could compensatorily proliferate in response to LPS.