Cbp Deficiency Alters Csk Localization in Lipid Rafts but Does Not Affect T-Cell Development

Generation of Cbp-deficient mice.The cDNA for Cbp was obtained by reverse transcription-PCR of RNA isolated from mouse spleen and used to probe a mouse 129 genomic DNA library. A phage clone containing the last exon of Cbp was obtained. Restriction enzyme digestion, Southern blotting, and DNA sequencing were used to map the genomic locus. A targeting vector was assembled, linearized by NotI digestion, and transfected into E14.1 embryonic stem (ES) cells. Targeted ES cells were selected by G418 resistance, and homologous recombinants were confirmed by Southern blotting using a 5′ external probe as indicated in Fig. 1. Seven homologous recombinants were obtained, and two of them were injected into C57BL/6 blastocysts to generate chimeric mice for germ line transmission of the inactivated allele. The mutant mice from these two different ES cell clones exhibited comparable phenotypes. All mice used in our experiments have a mixed 129 and C57BL/6 genetic background.

• Open in new tab
• Download powerpoint

FIG. 1.

Inactivation of the cbp gene locus. (A) Partial endonuclease restriction map of the wild-type cbp locus, the targeting construct, and the inactivated locus after homologous recombination. The restriction enzyme sites are as follows: B, BamHI; E, EcoRI; H, HindIII; X, XhoI; and S, SalI. The probe used to screen ES cells and mice, which distinguishes the 8.5-kb wild-type and 5.5-kb targeted alleles, is indicated. (B) Southern blot analysis of EcoRI-digested tail DNA from wild-type (+/+), heterozygous (+/−), and homozygous (−/−) Cbp knockout mice using the external probe. (C) Northern blot analysis of thymocyte RNA prepared from wild-type, heterozygous, and homozygous Cbp knockout mice using the full-length Cbp cDNA as a probe. The glyceraldehyde-3-phosphate dehydrogenase (GAPDH) blot is included as a control for loading and integrity of samples. (D) Western blot analysis of the Cbp protein in wild-type and Cbp knockout mice. Thymocytes were lysed in 0.05% Triton X-100 and subjected to sucrose density gradient centrifugation. Equal volumes of gradient fractions (fractions 2 to 10) were separated by SDS-PAGE and probed for the presence of Cbp using two different antibodies that recognized the C-terminal (panels 1 and 2) and N-terminal (panels 3 and 4) portions of Cbp. A dot blot of GM1 was included to indicate the gradient fractions that correspond to lipid raft fractions. A nonspecific (N.S) band present in both wild-type and mutant cell lysates using the antibody against the N terminus of Cbp is also included.

Flow cytometry analyses of Cbp^−/− mice.Single-cell suspensions were obtained from thymus, spleen, lymph nodes, bone marrow, and peritoneal cavity of mice as described previously (29). Cells were stained with an optimal amount of fluorescein isothiocyanate-, phycoerythrin-, and biotin-conjugated antibodies to cell surface antigens. Biotin-conjugated antibody staining was revealed by a second step, streptavidin-CyChrome staining. The following antibodies, purchased from BD Pharmingen (San Diego, CA), were used in this study: anti-immunoglobulin M (IgM) (R6-60.2), anti-IgD (11-26C.2a), anti-B220 (RA3-6B2), anti-CD3ε (145-2C11), anti-CD4 (L3T4), anti-CD5 (53-7.3), anti-CD8 (53-6.7), anti-CD19 (1D3), anti-CD25 (7D4), anti-CD43 (S7), anti-CD44 (IM7), anti-CD62L (MEL-14), anti-CD69 (H1.2F3), anti-TCRβ (H57-597), anti-TCRVβ6 (RR4-7), and anti-TCRVβ8 (MR5-2). Data were collected on a FACScan (BD, Mountain View, CA) and analyzed using CellQuest software.

Intracellular calcium analysis.Thymocytes (5 × 10⁶) were loaded with Indo-1 AM (2 μM; Molecular Probes) in the loading buffer (1 mM CaCl₂; 1 mM MgCl₂; 1% fetal calf serum in phosphate-buffered saline) for 45 min at 37°C. Cells were stained with phycoerythrin-anti-CD4, fluorescein isothiocyanate-anti-CD8, and biotinylated anti-TCRβ antibodies on ice for 15 min. After 30 min of incubation at room temperature, the cells were stimulated with streptavidin (5 μg/ml) to cross-link the TCRs. Calcium flux was monitored on an LSR II flow cytometer (Becton Dickinson) in real time for 15 min. CD4⁺ CD8⁺ double-positive cells and CD4⁺ CD8⁻ and CD4⁻ CD8⁺ single-positive cells were selectively analyzed for their abilities to flux calcium.

Measurement of T-cell proliferation and cytokine secretion.T cells were purified from lymph nodes of mice using mouse T-cell enrichment columns (R&D system). The purity of the negatively selected T cells was consistently >90% as assessed by flow cytometry using anti-CD3ε antibody. Stimulation of T cells was performed for 48 h at 37°C using 96-well plates that were precoated with various concentrations of monoclonal anti-mouse CD3ε antibodies in the absence and presence of soluble anti-CD28 antibody. To measure proliferation, cells were pulsed with 1.0 μCi [³H]thymidine (Amersham) per well for the last 10 h of incubation. The amount of [³H]thymidine incorporation was measured using a β-scintillation counter (Perkin-Elmer, Life Sciences). To measure cytokine secretion, cell culture supernatant was collected and assayed using the various cytokine enzyme-linked immunosorbent assay (ELISA) kits from BD Pharmingen. All assays were performed in triplicate, and the experiments were repeated at least five times.

Lipid raft purification.Thymocytes (4 × 10⁸) were lysed in 0.05% Triton X-100 in TNEV buffer (150 mM NaCl, 5 mM EDTA, and 25 mM Tris-HCl, pH 7.4), followed by addition of an equal volume of 80% sucrose in lysis buffer and overlaid with 30% and 5% sucrose in the same buffer, respectively. The fractionation was performed in a SW60Ti rotor for 16 to 18 h at 4°C and at 200,000 × g. Eleven fractions were collected, and the lipid raft fractions (corresponding to third, fourth, and fifth fractions) were solubilized in 10 mM octyl glucoside (Sigma).

T-cell stimulation and immunoblotting.Purified lymph node T cells were incubated with 10 μg/ml biotinylated anti-TCRβ antibodies for 30 min at 4°C. Cells were washed and further stimulated with 5 μg/ml streptavidin at 37°C for various times. After stimulation, cells were lysed and analyzed by Western blotting antibodies against various signaling molecules: anti-Cbp (Exbio); anti-phospho-PLCγ1 and anti-phospho-LAT (Tyr191) (Cell Signaling Technology); and anti-phospho-ERK, anti-Csk, anti-ERK2, anti-PLCγ1, and anti-LAT (Santa Cruz). A rabbit polyclonal antibody directed against the N terminus of Cbp was raised using a peptide encompassing amino acids 42 to 56. The glycosphingolipid GM1 was detected with the horseradish peroxidase-conjugated cholera toxin B subunit (Sigma-Aldrich). Immunoblottings were visualized by enhanced chemiluminescence detection (Pierce, Rockford, IL)

Measurement of basal serum Ig, auto- and antigen-specific antibodies.To determine basal serum immunoglobulin levels, sera were obtained from 6- to 8-week-old mice, serially diluted, and added to ELISA plates that were precoated with anti-Igκ antibodies. The levels of serum Ig of various classes were assayed using commercially available kits (BD Pharmingen) according to the manufacturer's instructions. For the detection of autoantibodies, sera were taken from 9- to 18-month-old mice, and the presence of anti-double-stranded DNA (dsDNA) antibodies was measured using poly(dAdT) (Sigma) for coating and horseradish peroxidase-conjugated anti-IgG or biotinylated anti-IgM antibody and streptavidin-conjugated horseradish peroxidase for detection. To examine T-cell-dependent immune response, 10- to 12-week-old mice were immunized intraperitoneally with 100 or 5 μg of alum-precipitated NP₁₇-chicken globulin in a high- or low-dose regime, respectively. Sera were obtained 7, 14, and 28 days after challenge to detect NP-specific antibodies using ELISA. For secondary immune response, the mice were injected with 5 μg of the same antigen 45 days after primary challenge, and sera were collected 7 days later.

Peripheral T-cell tolerance to SEB.Five- to six-week-old mice were injected intraperitoneally with 100 μg of staphylococcal enterotoxin B (SEB) (Toxin Technology) on day 0. Peripheral blood was taken on days 0, 2, 5, 10, and 16 and analyzed by flow cytometry using antibodies against CD4, CD8, TCRβ8, and TCRβ6.

Generation of viable Cbp-deficient mice.Cbp is a transmembrane adaptor protein that recruits Csk to lipid rafts where the latter is postulated to exert its negative regulatory effect on the Src family of tyrosine kinases (1, 16). The inactivation of Csk in the mouse germ line has been shown to impair T-cell development and selection in vivo (9, 23, 24). To determine if Cbp has a similar role in T-cell physiology, we have inactivated the gene in mouse embryonic stem cells by deleting its last coding exon (from amino acid (aa) 311 onwards). This would abolish the critical Tyr-314 residue and the C-terminal VTRL motif, which are required for Cbp's interaction with Csk and EBP-50, respectively (2, 13) (Fig. 1A). Therefore, our targeting strategy would prevent Cbp from binding Csk and EBP-50 in the event that a truncated Cbp protein is ever produced.

The inactivation of the cbp gene in mice was verified by Southern blotting (Fig. 1B). Northern blot analysis further showed that there was no steady-state full-length or truncated cbp mRNA present in the mutant cells (Fig. 1C). To ascertain that we have established a null mutation of cbp in mice, we directly examined whether the Cbp protein was completely disrupted. To this end, we prepared whole-cell lysate from thymocytes of wild-type and mutant animals, separated them into lipid rafts and cytosolic fractions, and subsequently probed them with two different antibodies that recognized the C-terminal (aa 97 to 432) and the N-terminal (aa 42 to 56) regions of Cbp, respectively. As shown in Fig. 1D, both antibodies could detect the 80-kDa Cbp protein in the lipid rafts (fractions 3, 4, and 5) but not in the cytosolic fractions of wild-type thymocytes. As expected, neither a full-length nor a truncated Cbp protein was detected in all cell lysate fractions obtained from mutant cells. This correlated well with the Northern blot analysis of mutant thymocytes, which indicated the absence of either a full-length or truncated cbp mRNA (Fig. 1C). Thus, Cbp has been inactivated in our mutant mice.

Interestingly, Cbp-deficient mice were viable and were born in the normal Mendelian ratio. They also survived to adulthood and beyond with no gross developmental defect. We have maintained Cbp mutant mice up to 20 months of age. This was in contrast to Csk-deficient mice, which suffered from early embryonic lethality (12, 19). Thus, although both Cbp and Csk are ubiquitously expressed and Csk-deficient mice are nonviable, Cbp is apparently not essential for mouse development.

Increased thymocyte population but normal T-cell development in Cbp-deficient mice.Since Csk has been shown to play an important role in T-cell differentiation, we focused our analyses on developing thymocytes and peripheral T cells in the mutant mice. T cells develop from CD4/CD8 double-negative to double-positive and finally to CD4 or CD8 single-positive cells in the thymus (25, 28). As shown in Fig. 2A, the absence of Cbp did not impair the maturation of thymocytes, since flow-cytometry analyses revealed that the different fractions of developing thymocytes were found in similar proportions in both wild-type and mutant mice. There was, however, a consistent and statistically significant overall increase in the total number of thymocytes found in 4- to 5-week-old Cbp^−/− mice compared to age- and sex-matched littermate wild-type mice (Fig. 2B).

• Open in new tab
• Download powerpoint

FIG. 2.

Normal T-cell development in Cbp^−/− mice. Flow cytometry analyses of T-cell populations in the thymus (A), spleen (C), and lymph nodes (D) of wild-type and mutant mice. Numbers indicate percentages of lymphocytes. Data shown are representative of more than five independent experiments. (B) Total number of thymocytes found in wild-type (black circles) and mutant (open circles) mice (4 to 5 weeks old). Each mouse analyzed was depicted as a single data point. Statistical significance was determined by a paired two-tailed Student t test.

Peripheral T-cell populations in Cbp^−/− mice also appeared normal, since equivalent fractions of CD3⁺ and CD4⁺ or CD8⁺ cells were found in the spleen (Fig. 2C) and lymph nodes (Fig. 2D) of these mice compared to control mice. Analyses of older mice (9- to 12-month-old) also suggested that the populations of CD62L⁺ CD4⁺ CD25⁺ regulatory T cells and CD62L^low CD4⁺ CD44^high memory/effector T cells were indistinguishable between wild-type and mutant mice (unpublished data). B-cell development was also not affected in the bone marrow and peripheral lymphoid organs of Cbp^−/− mice (unpublished data). Furthermore, enumeration of cells in the mutant mice also revealed that there was no overt increase in the number of B and T cells in the peripheral lymphoid tissues of these animals (Table 1). Taken together, these data indicate that Cbp deficiency does not affect T-cell development or homeostasis. This, again, is in contrast to the situation found in Csk^−/− mice, which manifest a very early block in T-cell development (9).

Lipid raft-associated Csk is greatly reduced in the absence of Cbp.Given that significant amounts of Csk have been shown to associate with Cbp in lipid rafts (1, 16), we next examined if the amount of Csk in the lipid rafts of resting thymocytes could be altered in the absence of Cbp. Equal volumes of different gradient fractions (fractions 2 to 10) obtained from cell lysates of wild-type and Cbp-deficient thymocytes were blotted with antibodies against Cbp, Csk, and LAT. The presence of LAT denotes the lipid raft fractions of the cell lysates. As shown in Fig. 3A (upper two panels), we were able to detect the presence of Csk in the lipid raft fractions (fraction 4 and to a lesser extent fraction 3) obtained from wild-type thymocytes. However, Csk was hardly detectable in the lipid rafts (fraction 4) of resting Cbp-deficient thymocytes. When equal protein amounts of lipid raft fractions (fraction 3, 4, and 5) were loaded to examine the presence of Csk (Fig. 3A, bottom two panels), the amount of Csk was significantly reduced in the lipid rafts fractions of mutant cells from that of wild-type cells, whereas the amounts of LAT in lipid rafts were comparable between wild-type and Cbp-deficient thymocytes (Fig. 3A, bottom panel). In addition, the amount of Csk in the whole-cell lysate of thymocytes was not affected by the Cbp deficiency (unpublished data). Thus, Cbp is required to localize Csk to lipid rafts in primary resting thymocytes.

• Open in new tab
• Download powerpoint

FIG. 3.

Reduced Csk in lipid rafts but normal TCR-proximal signaling events in Cbp-deficient T cells. (A) The amount of Csk was reduced in the lipid rafts of Cbp-deficient thymocytes. Equal volumes (40 μl) of different gradient fractions (upper two panels) or equal protein amounts (2 μg/lane) of lipid raft fractions (bottom two panels) from thymocytes of wild-type and mutant mice were sequentially probed for the presence of Cbp, Csk, and LAT. (B) Downstream signaling events were intact following TCR activation in Cbp-deficient T cells. Mouse T cells were purified from lymph nodes of wild-type and mutant mice and stimulated for the indicated periods of time at 37°C with biotinylated anti-TCR antibody followed by cross-linking with streptavidin. The activation of LAT, PLCγ1, and ERK, as indicated by their phosphorylation status, was examined by Western blot analyses. (C) Intracellular calcium fluxes in response to TCR stimulation were comparable in wild-type and Cbp-deficient thymocytes. Thymocytes were preloaded with Indo-1 and subsequently stimulated with anti-TCR antibody. Intracellular [Ca²⁺] in CD4⁺ CD8⁺, CD4⁺ CD8⁻, and CD4⁻ CD8⁺ thymocytes is shown as a ratio of Indo-1 violet/blue fluorescence versus time.

TCR-proximal signaling events are normal in absence of Cbp.Since lipid raft-associated Csk was greatly reduced in the absence of Cbp, we next determined if this would affect TCR-proximal signaling events. TCR engagement is known to lead to the phosphorylation of LAT and activation of both PLCγ1 and mitogen-activated protein kinases and to induce intracellular calcium flux (22). As shown in Fig. 3B, the activation of LAT, PLCγ1, and ERK1/2, as indicated by their phosphorylation status, appeared comparable between wild-type and Cbp-deficient T cells upon TCR ligation. We further examined TCR-induced calcium mobilization in both wild-type and Cbp-deficient thymocytes. Thymocytes were preloaded with the calcium indicator dye Indo-1 and were stimulated with biotinylated anti-TCR antibody followed by cross-linking with streptavidin. The intracellular calcium concentration was monitored over time using flow cytometry. As shown in Fig. 3C, TCR-induced calcium fluxes were comparable in either CD4⁺ or CD8⁺ and CD4⁺ CD8⁺ thymocytes obtained from wild-type and mutant mice. Similar results were also attained when a lower dose of streptavidin (2 μg/ml) was used to cross-link TCR (unpublished data). Hence, it appears that the absence of Cbp and the concomitant reduction of Csk in the lipid rafts of resting T cells do not affect the TCR-proximal signaling events.

In vitro T-cell responses are not compromised in the absence of Cbp.Since the initial localization of Csk in resting thymocytes was perturbed in the absence of Cbp but yet T-cell development and TCR-proximal signaling events remained normal in Cbp-deficient mice, we next asked if the absence of Cbp would affect T-cell responses rather than development. Since Csk negatively regulates the Src family of tyrosine kinases, which by themselves are critical for T-cell activation, it is conceivable that the absence of Cbp may lead to exaggerated T-cell responses.

To address this possibility, we examined the ability of mutant T cells to proliferate and secrete cytokines in response to TCR stimulation with and without CD28 costimulation. As shown in Fig. 4A, the engagement of the TCR via stimulation of the signal-transducing CD3 led to the proliferation of wild-type T cells in a dose-dependent manner. This proliferative response was further enhanced by the coengagement of the CD28 molecule. Interestingly, the absence of Cbp neither incapacitated nor exaggerated the proliferative responses of the mutant T cells. Likewise, the ability of T cells to produce cytokine after activation, as shown by the secretion of IL-2 and -4 and gamma interferon (Fig. 4B), was also indistinguishable between wild-type and mutant T cells. Taken together, the data suggest that the absence of Cbp does not affect T-cell responses in vitro.

• Open in new tab
• Download powerpoint

FIG. 4.

T-cell proliferation and cytokine production in the absence of Cbp. Wild-type and mutant T cells were stimulated with various doses of anti-CD3 and anti-CD28 antibodies and examined for their (A) proliferation and (B) secretion of IL-2, IL-4, and IFN-γ. Data shown are representative of at least five independent experiments.

Cbp-deficient mice have normal humoral immune responses.Since Csk, which interacts with Cbp, regulates the activation of the Src family of tyrosine kinases, and the impairment of Lyn, a member of this family, is known to lead to autoimmunity with high serum antibody titers in mice (10), we next assessed if B-cell responses in Cbp^−/− mice would be altered. Examination of the serum antibody levels in Cbp-deficient mice revealed that there were no significant differences in the amounts of immunoglobulins of different classes in the sera of these mice compared to those of control animals (Fig. 5A).

• Open in new tab
• Download powerpoint

FIG. 5.

T-cell-dependent humoral immune responses in Cbp^−/− mice. Basal serum immunoglobulin levels of various classes (A) and primary T-cell-dependent humoral immune responses to a high dose (B) and low dose (C) of antigen were shown for wild-type (black circles) and Cbp^−/− (open circles) mice at various time points after challenge (D, day). Each individual mouse is represented as a single data point.

Antigens that elicit a specific antibody response from B cells can be classified as either T cell independent or T cell dependent according to their dependency on CD4⁺ T-cell help. To determine if Cbp^−/− mice could mount efficient immune responses against exogenous antigens, we first immunized the mice with a T-cell-independent antigen, NP-Ficoll. The antibody response to NP-Ficoll was mainly of the IgM and IgG3 class; again both wild-type and Cbp^−/− mice responded equally well to NP-Ficoll challenge (unpublished data).

For the T-cell-dependent immune response, we first challenged the mice with either a high or low dose of the antigen NP-CG and assessed the magnitude of both the primary and secondary immune responses. The antibody response to NP-CG was mostly of the IgM and IgG1 classes. Again, as shown in Fig. 5B and C, the primary immune responses of Cbp^−/− mice to either a high or low dose of T-cell-dependent antigen were indistinguishable from that of wild-type mice. Similar results were obtained when the mice were rechallenged with the same antigen in a secondary immunization regime (unpublished data). Thus, the lack of Cbp does not alter the ability of T cells to provide help to B cells in the T-cell-dependent humoral immune response.

Absence of Cbp did not affect peripheral B- and T-cell tolerance.To further determine if Cbp plays a role in T-cell activation in vivo, we examined the induction of peripheral T-cell tolerance after SEB challenge in mice. SEB is a superantigen that stimulates the proliferation of a subset of T cells whose TCR contains either a Vβ7, -8, or -17 chain (15). Administration of SEB in adult wild-type mice will result in the initial proliferation of these T cells at early time points, followed by their anergy and deletion at later time points. As shown in Fig. 6A, there were an initial increase and a subsequent decrease in the populations of CD4⁺ Vβ8⁺ and CD8⁺ Vβ8⁺ T cells following the administration of SEB in wild-type mice. Examination of the CD4⁺ Vβ8⁺ and CD8⁺ Vβ8⁺ T-cell populations in Cbp^−/− mice revealed that they were indistinguishable from that of the wild-type T cells. As a control, the populations of CD4⁺ Vβ6⁺ and CD8⁺ Vβ6⁺ T cells, which were not stimulated by SEB, remained fairly constant for the duration of the challenge in both wild-type and Cbp-deficient mice. Taken together, the data indicate that Cbp does not play a role in the induction of T-cell peripheral tolerance to SEB.

• Open in new tab
• Download powerpoint

FIG. 6.

Unaffected B- and T-cell tolerance in the absence of Cbp. (A) Wild-type (black circles) and Cbp^−/− (open triangles) mice were given SEB at day 0, and the fractions of responding CD4⁺ Vβ8⁺ and CD8⁺ Vβ8⁺ and control CD4⁺ Vβ6⁺ and CD8⁺ Vβ6⁺ cells were quantified at various days after antigenic challenge. (B) Serum levels of IgG auto-antibodies against dsDNA were quantified in aged wild-type (black circles), Cbp^−/− (open circles), and control Lyn^−/− (black diamonds) mice. Each data point represents an individual mouse analyzed.

Finally, we assess if Cbp-deficient mice have a propensity to develop autoimmunity by assaying for pathological autoantibodies against dsDNA in the sera of aged Cbp^−/− mice. As shown in Fig. 6B, IgG autoantibodies were not detected in 9- to 12-month-old Cbp^−/− mice, unlike the situation in autoimmune-prone Lyn^−/− mice, which developed autoantibodies by 6 months of age (10). IgG autoantibodies to dsDNA were found in 18-month-old Cbp^−/− mice, but the occurrence was rare and did not differ significantly from that of Cbp^+/+ mice. The similar results were also obtained for IgM autoantibodies (unpublished data). Hence, the absence of Cbp does not lead to the preferential development of autoimmunity in mice.