MAPK p38α Is Dispensable for Lymphocyte Development and Proliferation¹

The MAPKs include ERK-1, ERK-2, JNK-1, JNK-2, JNK-3, and the p38 kinases (p38 α, β, γ, and δ) (1). These kinases are broadly expressed and participate in signaling cascades initiated by a host of cellular stimuli, leading to cellular differentiation and the activation of effector responses (2). The activation of specific MAPK family members is determined by the activity of the upstream MAPK kinases (MKKs)³ (3). These dual tyrosine and threonine kinases phosphorylate the MAPK TXY motif required for activation (4). Specific MKKs activate p38, JNK, and ERK kinases, although cross-activation can occur (5, 6, 7, 8).

The p38α MAPK was first identified as a mediator of LPS signaling in B cells and was found to be homologous to the high osmolarity glycerol 1 kinase that is involved in osmotic stress responses in yeast (9). Subsequent studies implicated p38α in mediating signals induced by a wide variety of inflammatory mediators, including TNF-α and IL-1β (9, 10, 11). In addition, p38α is involved in signaling from growth factor and G protein-coupled receptors that promote a wide range of biologic processes, including developmental patterning, proliferation, and tissue differentiation (12, 13, 14, 15, 16). There are three other p38 isoforms (β, δ, and γ) encoded by distinct genes (9, 17, 18, 19). The activity of the p38 MAPKs is regulated by MKK3, MKK4, and MKK6 (7, 8, 20, 21). All of the p38 kinases have a relatively broad tissue distribution, except for p38γ, which appears to be expressed primarily in muscle (17, 22, 23). Although p38α is the major isoform expressed in T lymphocytes, these cells also express p38 β and δ (24, 25).

Four independently generated p38α-deficient mice have been described (26, 27, 28, 29). In each case, p38α deficiency led to embryonic lethality, although there was some variability in the gestational age of death. This embryonic lethality prohibits the assessment of p38α deficiency in adult mouse tissues, including lymphocytes. Nonetheless, p38α has been implicated as an important signaling component in many processes required for T cell development and function. A host of experimental modalities has been used to modulate p38α activity. Pharmacologic inhibitors of p38α, such as the SB203580 compound, can affect the activity of p38 α and β and, at high concentrations, may affect the activity of other kinases (30, 31, 32). Transgenic and retrovirally mediated overexpression of the wild-type or constitutively active forms of MKK3 and MKK6 or a dominant-negative form of p38α has also been used to assess p38α function (33, 34, 35). Together, these studies have implicated p38α in T cell developmental and effector functions, including, but not limited to, signaling from the pre-TCR, TCR, and several cytokine receptors, including IL-2R, IL-7R, and IL-18R (30, 31, 34, 35, 36, 37, 38, 39). Furthermore, as might be expected, analyses using different experimental approaches have at times led to conflicting conclusions (30, 31, 33, 34). In contrast to T cells, few studies have attempted to assess the requirements for p38α in B cell development and function.

We generated a series of low passage p38α^+/+, p38α^+/−, and p38α^−/− embryonic stem (ES) cells through the intercrossing of p38α^+/− mice (28). These ES cells were used to generate chimeric mice by RAG-deficient blastocyst complementation (RDBC) (40). These mice develop normally as cells derived from the RAG-2^−/− blastocyst express p38α. However, the lymphocytes in these mice are derived entirely from the ES cells as the RAG-2-deficient blastocyst is not able to give rise to lymphocytes. These mice were directly analyzed to assess the requirement for p38α in B and T cell development and function.

The p38α^+/− mice were intercrossed, and blastocysts were harvested by flushing oviducts from female mice at 72 h postcoitum. Blastocysts were individually cultured in wells of a 24-well plate containing a mouse embryo fibroblast monolayer in DMEM containing 15% ES-tested FCS (Invitrogen Life Technologies) and 1000 U/ml LIF (Chemicon International), and supplemented with nonessential amino acids and penicillin-streptomycin. After 7–10 days, the cells in individual wells were dispersed with 0.25% trypsin:1 mM EDTA and replated in individual wells of a 24-well plate with a mouse embryo fibroblast monolayer. These cells were redispersed and expanded in culture once ES colonies were visible (usually 7–10 days). ES cell lines were expanded to ∼5 × 10⁶ cells before being frozen for later use in RDBC.

Southern blotting of ES cell genomic DNA was conducted, as previously described (41). The p38α probe is a 1.3-kb fragment from murine p38α intron 2, generated by EcoRI digestion of PCR product generated with forward primer 5′-ttgtgattattggggactgtaggg-3′ and reverse primer 5′-ggacatacacatggacacacatcg-3′.

Single cell suspensions were prepared from thymus, bone marrow, spleen, and lymph nodes, as previously described (40). Cells were stained with FITC- or PE-conjugated Abs or biotinylated Abs, followed by streptavidin-allophycocyanin, and were analyzed by a FACScan (BD Biosciences). Live cell gating was performed by propidium iodide exclusion. The following Abs from BD Pharmingen were used in these studies: CD4 PE, CD8 FITC, B220 PE, CD43 FITC, and IgM biotin. Stained cells were analyzed on a FACSCalibur and plotted using CellQuest software (BD Biosciences).

Thymocyte cell lysates were prepared using 1× Nonidet P-40 supplemented with 1 mM EDTA, 0.5 mM PMSF, 10 μg/ml aprotinin, 1 μg/ml leupeptin, and 1 μg/ml pepstatin. Cell extracts were centrifuged for 15 min at 4°C at 15,000 rpm. Soluble lysates were subjected to SDS-PAGE and immunoblotted with an anti-p38α polyclonal rabbit antisera (C-20; Santa Cruz Biotechnology) or anti-p85 polyclonal rabbit antisera (06-195; Upstate Biotechnology).

Total splenocytes from p38α^+/+, p38α^+/−, p38α^−/− chimeric mice were cultured in complete IMDM. Triplicate samples of 1 × 10⁵ cells were plated in a single well of a flat-bottom 96-well Costar plate and stimulated with plate-bound anti-CD3 (2C11), anti-CD40 (3/23; BD Pharmingen), anti-IgM (Jackson ImmunoResearch Laboratories), or PMA with ionomycin. After 48-h stimulation, cells were pulsed with 2 μCi of [³H]thymidine/well and incubated for an additional 12 h. Proliferation was measured by [³H]thymidine incorporation using a Molecular Devices Micro96 Harvester and counted with a Beckman 6500 scintillation counter.

Serum was collected from 6- to 8-wk-old p38α^+/+, p38α^+/−, p38α^−/− chimeric mice. Goat anti-mouse Ig capture Ab in bicarbonate buffer (pH 8.8) was coated overnight at 4°C in Immulon 96-well plates. Plates were washed with 50 mM Tris (pH 7.6) containing 2% Triton X-100, and blocked in 10% FCS tissue culture medium for 3 h at room temperature. Serum was initially diluted 2000-fold for IgM; 1000-fold for IgG1, IgG2a, and IgG2b; and 500-fold for IgG3 and IgA. Diluted serum was subjected to 3-fold serial dilutions and was plated in triplicate at 4°C overnight. Samples were extensively washed, and HRP-conjugated goat anti-mouse IgM, IgG1, IgG2a, IgG2b, IgG3, and IgA (Southern Biotechnology Associates) were applied for 1 h. Samples were developed with 100 mM sodium citrate, 1 mM ABTS, and 0.016% H₂O₂. OD was read at 405 nm.

A large panel of low passage p38α^+/+, p38α^+/−, and p38α^−/− ES cells was generated through the intercrossing of p38α^+/− mice (28). Blastocysts isolated from these intercrosses were independently cultured for ES cell isolation, as described in Materials and Methods (Fig. 1). A total of 35 ES cell lines was isolated from 77 independently cultured blastocysts, with Southern blot analyses revealing near Mendelian ratios of p38α^+/+, p38α^+/−, and p38α^−/− ES cell lines (Fig. 1). Chimeric mice were generated by RDBC using 3 p38α^+/+, 3 p38α^+/−, and 7 p38α^−/− ES cell lines.

Lymphocyte development in p38α^+/+, p38α^+/−, and p38α^−/− chimeric mice was assessed by flow cytometric analyses of thymocytes and bone marrow cells (Figs. 2 and 3). Surprisingly, chimeric mice generated with p38α^−/− ES cells exhibited similar total thymocyte numbers as compared with chimeric mice generated with either p38α^+/− or p38α^+/+ ES cells (Fig. 2,A). Furthermore, the CD4⁻/CD8⁻ (double-negative), CD4⁺/CD8⁺ (double-positive (DP)), and CD4⁺ or CD8⁺ (single-positive (SP)) thymocyte compartments also appeared intact in p38α^−/− chimeric mice (Fig. 2).

The p38α⁻ allele was generated by insertion of the neomycin resistance gene into exon 3 of the p38α locus (28). Importantly, polyclonal rabbit antisera raised to the C terminus of p38α failed to detect a p38α protein in p38α^−/− thymocytes (Fig. 4). Moreover, analysis of chimeric mice generated from p38α^−/− ES cells in which the p38α⁻ allele was generated through the deletion of eight exons, including the exon encoding the TGY motif, also failed to reveal a perturbation in thymocyte development (data not shown) (27). Together, these data make it unlikely that the robust thymocyte development observed in the p38α^−/− chimeric mice is due to the production of a truncated p38α protein from the p38α⁻ allele.

Flow cytometric analyses of bone marrow from p38α^+/+, p38α^+/−, and p38α^−/− chimeric mice revealed relatively similar fractions of pro-B (B220⁺CD43⁺), pre-B (B220^lowCD43⁻), and immature B (IgM⁺) cells (Fig. 3). Together, these findings demonstrate that, in the absence of p38α, B and T cell development is generally unperturbed.

Flow cytometric analyses revealed similar numbers of mature CD4⁺ and CD8⁺ splenic T cells in p38α^+/+, p38α^+/−, and p38α^−/− chimeric mice (Fig. 5). Furthermore, p38α^−/− αβ T cells express similar TCR levels as compared with p38α^+/+ and p38α^+/− T cells (data not shown). Proliferation of p38α^−/− αβ T cells was assayed upon stimulation of these cells with plate-bound anti-CD3 Abs. The p38α^−/− αβ T cells exhibited robust proliferation over a broad range of anti-CD3 concentrations (Fig. 6). Moreover, this proliferation was similar in magnitude to that observed for p38α^+/+ and p38α^+/− and αβ T cells. Together, these data demonstrate that neither the generation and maintenance of normal numbers of mature peripheral αβ T cells nor the ability of these cells to undergo TCR-driven cellular expansion is dependent on p38α.

Flow cytometric analyses of splenocytes from p38α^+/+, p38α^+/−, and p38α^−/− chimeric mice revealed similar numbers of B220⁺ IgM⁺ mature B cells and similar fractions of IgM⁺ IgD⁺ mature B cells (Fig. 7). When stimulated with either anti-IgM or anti-CD40, mature p38α^−/− B cells exhibited similar levels of proliferation as compared with p38α^+/+ or p38α^+/− B cells (Fig. 8, A and B). Furthermore, serum Ig isotype analyses revealed no significant differences in IgM, IgG1, IgG3, IgG2b, and IgA levels when comparing p38α^+/+, p38α^+/−, and p38α^−/− chimeric mice (Fig. 9). Together, these data demonstrate that p38α is not essential for CD40- or BCR-driven proliferation of mature B cells, nor is it required for the generation and maintenance of normal serum Ig levels.

The p38α was initially identified as a mediator of endotoxin signaling in B cells (9). B lymphocytes undergo proliferation in response to stimulation with endotoxin (42). Surprisingly, proliferation in response to LPS was indistinguishable when comparing p38α^−/−, p38α^+/+, or p38α^+/− B cells (Fig. 8 C). Thus, p38α was also dispensable for the endotoxin-mediated proliferation of B cells.

MAPK p38α deficiency leads to placental defects and embryonic lethality (26, 27, 28, 29). Several different approaches have been used to assess the development and function of lymphocytes in the absence of proteins required for embryonic development. If the embryos survive until a gestational stage at which bone marrow or fetal liver can be harvested, these cells can be used to reconstitute the lymphoid compartment in RAG-deficient mice. Although in one report p38α^−/− mice survived under unique circumstances until a gestational age that permitted fetal liver harvest (29), the other reported p38α^−/− mice, including the one used in this study, did not (26, 27, 28).

RAG-deficient blastocyst complementation is an alternate approach for assessing lymphoid development and function in the absence of proteins required for embryonic development (40). This approach requires the generation of ES cells with homozygous mutations in autosomal genes of interest. Traditionally, this has been done through serial allele targeting or high drug selection approaches. Both of these approaches require long-term ES cell culture that can, on occasion, render these cells incapable of contributing to the lymphoid compartment due to ill-defined effects that are independent of the gene-targeted mutation. To circumvent this potential problem, we generated p38α^+/+, p38α^+/−, and p38α^−/− ES cells through the intercrossing of p38α^+/− mice and expanded these cells in culture for minimal periods of time (<1 mo) before use in RDBC. Approximately 40% of cultured blastocysts gave rise to ES cells, and most of these ES gave robust reconstitution of the lymphoid compartment in chimeric mice generated by RDBC. Thus, the de novo generation of ES cells for use in RDBC is an efficient and effective way to generate chimeric mice for analyses of lymphocyte development and function.

Surprisingly, p38α^−/− chimeric mice generated by RDBC revealed no significant perturbations in T cell development. Total thymocyte cellularity and the number of DP and SP thymocytes were not significantly different when comparing chimeric mice generated with p38α^+/+, p38α^+/−, and p38α^−/− ES cells. These findings do not exclude a potential role for p38α in thymocyte-negative selection (31). Furthermore, other p38 isoforms could compensate for p38α deficiency during lymphocyte development. In this regard, p38δ and p38β, but not p38γ, are expressed in thymocytes (24, 25). Notably, p38 δ and α do not exhibit significant sequence homology and have few known substrates in common (19, 43, 44). In contrast, p38β has significant sequence homology to p38α and shares several known substrates with p38α (18, 19, 43, 44). Thus, it is possible that p38 β and/or δ could compensate for p38α deficiency. Nevertheless, our findings demonstrate that p38α signals per se are not essential for the robust development and maintenance of the different thymocyte subsets.

The generation and maintenance of mature CD4⁺ and CD8⁺ T cells in the periphery were also unaffected by p38α deficiency. Studies using pharmacologic inhibitors have implicated p38α signaling in the TCR-mediated proliferation of mature T cells (45). However, we found no significant differences in the ability of p38α^+/+, p38α^+/−, and p38α^−/− T cells to proliferate in response to anti-CD3 treatment. Together, these findings demonstrate that p38α is not an essential component of TCR or cytokine receptor, such as IL-2R, signaling required for the maintenance and/or proliferation of mature T cells.

Similarly, the development and maintenance of mature B cells were unperturbed in p38α^−/− chimeric mice. The ability of these cells to produce Abs was sufficient to generate normal serum levels of IgM, IgG1, IgG3, IgG2b, and IgA. Proliferative responses of these B cells upon stimulation through the BCR or CD40 were also intact. Furthermore, and surprisingly, p38α^−/− B cells proliferated robustly in response to LPS.

Together, our findings demonstrate that p38α is not an essential component of B and TCR or cytokine receptor signaling pathways required for the robust development and expansion of B and T cells. However, p38α is an essential component of the signaling pathways of inflammatory cytokines. For example, p38α signals are required for the production of IL-6 in response to IL-1 (27). Thus, our findings imply that specific pharmacologic inhibitors of p38α may impact the inflammatory cytokine-mediated responses without impacting lymphocyte development or proliferation.

We thank Dr. John McNeish for providing us with p38α^−/− ES cells, and Dr. John S. Mudgett for p38α^+/− mice.