INTRODUCTION

The human genetic disorder ataxia-telangiectasia (A-T)¹ is characterized by immunodeficiency, neurological abnormality, abnormal development, radiosensitivity, cell cycle anomalies, and cancer predisposition (1-3). Furthermore, a variety of immunological abnormalities characterize this syndrome, including hypogammaglobulinemia, selective deficiency of serum IgA and IgE (4, 5), abnormalities in IgG subclasses (6), depressed blastogenic response (7), faulty development or complete absence of the thymus (8), failure to produce virus-specific histocompatibility-restricted cytotoxic T lymphocytes (9), and an overall poor response to skin test antigens (10).

The gene responsible for the defect in A-T, ATM, has been cloned recently (11) and shown to possess a carboxyl-terminal domain homologous to phosphatidylinositol 3-kinase (PI 3-kinase). This enzyme, a heterodimer composed of a catalytic subunit (p110) and a regulatory subunit (p85), plays a central role in transmitting signals from the cell surface to the nucleus (12). The ATM protein is related to TOR1 and TOR2 proteins of yeast (13) and their mammalian counterparts FRAP (14) and RAFT1 (15) through the PI 3-kinase domain, and to a second group of proteins not only through this domain but also through an adjacent region of weaker homology (16). The latter group includes Mei-41 of Drosophila melanogaster, rad3p, Mec1p and Tel1p of yeast, and DNA-dependent protein kinase from human cells (17-20). These proteins are involved in cell cycle control and response to DNA damage.

It is evident that the ATM gene is involved not only in the response to DNA damage but also in regulating a number of cellular processes important in differentiation and development (3). Evidence for a defect in signal transduction in A-T has been provided in several reports. O'Connor and Scott-Linthicum (21) demonstrated that A-T lymphocytes were defective in the transmission of a mitogen-mediated signal from cytoplasm to nucleus; defective intracellular mobilization of Ca²⁺ in T cells from A-T patients in response to phytohemagglutinin and anti-CD3 antibody has been reported by Kondo et al. (22), and A-T fibroblasts in culture show a greater demand for growth factors (23). These observations raise the possibility that the hyporesponsiveness of the transmembrane signaling, seen primarily in T lymphocytes, may also occur in B cells from A-T patients and account, at least in part, for the immune dysfunction which is also a characteristic of the humoral arm of the response in this syndrome (10).

To investigate this, we have used antibodies to cross-link the B cell antigen receptor (BCR) which triggers the activation of protein tyrosine kinases, resulting in rapid tyrosine phosphorylation of a number of cellular substrates including phospholipase C1 (PLC1) (24). Activation of PLC1 leads to hydrolysis of phosphatidylinositol 4,5-bisphosphate to yield inositol 1,4,5-trisphosphate and diacylglycerol (25-27). Inositol 1,4,5-trisphosphate acts as a second messenger to release ionized Ca²⁺ from intracellular stores and diacylglycerol mediates the activation of protein kinase C. We report here that signaling through the BCR is defective in A-T which is detectable at the level of Ca²⁺ mobilization, PLC1 activation, and PI-3 kinase activation. These results provide evidence for a role for the ATM protein in intracellular signaling.

MATERIALS AND METHODS

All lymphoblastoid cell lines were cultured in RPMI 1640 medium supplemented with 10% fetal calf serum and antibiotics at 37 °C in a humidified atmosphere of 5% CO₂. Nomenclature for control and A-T lymphoblastoid cell lines is the same as that described previously (28), where ATABR represents an A-T cell line established in Australia, Brisbane. AT3LA represents an A-T cell line established in Los Angeles by Dr R. Gatti. Purified rabbit antibodies against PLC1, PI 3-kinase (p85), and Lyn were obtained from Santa Cruz Biotechnology, Inc. Anti-phosphotyrosine monoclonal antibody 4G10 was obtained from Upstate Biotechnology.

Control and A-T cells were washed twice with staining solution (phosphate-buffered saline plus 1% fetal calf serum) and then incubated with rabbit anti-human Ig (detecting IgM, IgG, IgA, and IgG fractions; Silenus, Melbourne, Australia), or anti-human CD45 and CD19 (Becton Dickinson). After the first incubation (30 min, 4 °C), cells were washed twice with staining solution and then incubated for 30 min at 4 °C with fluorescein isothiocyanate-conjugated goat anti-mouse antibody. Cells were then washed twice, resuspended in 0.5 ml of staining solution, and examined by flow cytometry.

B cells from A-T patients and controls were resuspended at a concentration of 5 × 10⁴ cells/ml in RPM1 1640 medium containing 1% fetal calf serum and cultured in 0.2-ml aliquots in flat-bottomed microtiter plates in the presence or absence of anti-Ig (at a final concentration of 5 µg/ml). Cultures were maintained for 24, 48, 72, and 96 h, and the cells were pulsed with [³H]thymidine (2 µCi/well) 4 h before termination of the incubation; [³H]thymidine incorporation was measured using an automated beta liquid scintillation counter (Beckman Instruments). The maximal stimulation index was calculated as counts/min of [³H]thymidine incorporation in anti-Ig-treated cells divided by counts/min of [³H]thymidine incorporation in untreated cells.

The levels of intracellular calcium was measured by flow cytometry using the fluorescent Ca²⁺ indicator indo-1 acetoxymethyl ester. Cells were loaded with 1 µM indo-1 acetoxymethyl ester in loading buffer (150 mM NaCl, 5 mM KCl, 0.41 mM MgCl₂, 0.20 mM Hepes, 10 mM glucose (pH 7.2)) containing 0.1% fetal calf serum with gentle shaking in the dark at 37 °C for 1 h. The relative intracellular calcium concentration was determined using a Becton Dickinson fluorescence-activated cell sorter (FACS Vantage). Intracellular Ca²⁺ was measured, using indo-1, as the ratio of 405 nm/530 nm fluorescence (29). The mean intracellular concentrations of Ca²⁺ were calculated as described by Rabinovitch et al. (29) using spectrofluorimetry (Perkin Elmer Luminescence Spectrometer LS50).

LMP2A expression was determined by immunoblotting. LMP2A antiserum was a kind gift from Professor Muller-Lantzsch (Hamburg, Germany). This antibody recognizes LMP2A expression in cell lines transformed with the B95-8 strain of EBV. Equal amounts of cell extracts of control and A-T cell lines were loaded on 7.5% SDS-polyacrylamide gel electrophoresis (PAGE) gel. Blots were incubated with 1:200 dilution of anti-LMP2A antiserum and then stained indirectly using peroxidase-conjugated goat anti-rabbit antibodies.

B cells were washed and suspended in 200 µl of cold serum-free RPMI 1640 medium. The suspension was preincubated at 37 °C for 1 min and then incubated in the presence of 50 µg of anti-Ig antibody at 37 °C for various periods. The cells were lysed using lysis buffer (50 mM Tris HCl (pH 7.4), 0.5% Triton X-100, 50 mM EDTA, 1 mM sodium o-vanadate, 20 mM NaF, and 10 µg/ml each of leupeptin and aprotinin). Immunoprecipitates from the cell lysates were prepared by preclearing with normal rabbit serum and precipitating with specific antisera (PLC1, Lyn, p85 of PI 3-kinase; 1 µg each) and protein A-agarose. An aliquot of equal volume from each immunoprecipitate was subjected to SDS-PAGE under reducing conditions and immunoblotted with antibody to phosphotyrosine, PLC1, Lyn, and p85.

Cells were lysed in Triton X-100 lysis buffer (same as described above for immunoprecipitation). Equal amounts of cell extracts were loaded on 10% SDS-PAGE gel and transferred to nitrocellulose sheets. The unreacted sites on nitrocellulose blots were blocked by incubation in 1% bovine serum albumin. Blots were incubated overnight with 1 µg/ml of monoclonal 4G10 anti-phosphotyrosine antibody (Upstate Biotechnology) followed by anti-mouse horseradish peroxidase conjugate (Silenus) to detect changes in the phosphorylation status of several proteins using enhanced chemiluminescence (Amersham).

PI 3-kinase was measured as described previously (30). Briefly, Lyn immunoprecipitates were washed twice with the lysis buffer and twice with Hepes buffer (20 mM Hepes (pH 7.4), 100 mM NaCl) and suspended in 25 µl of reaction buffer (20 mM Hepes, 100 mM NaCl, 0.45 mM EGTA) containing 0.2 mg/ml phosphatidylinositol. After incubation at 25 °C for 10 min, the reaction was started by addition of [-³²P]ATP (10 µCi; final concentration, 10 µM) and MnCl₂ (final concentration, 5 mM) and allowed to proceed at 25 °C for 10 min. The reaction was stopped by adding 100 µl of 1 M HCl. The lipid product was extracted, spotted onto a Silica Gel 60 plate (Merck), and developed.

RESULTS

With the cloning and identification of the A-T gene, it seems apparent that the defect is one of cell signaling involving a number of signal transduction pathways. Since immunodeficiency is one of the important characteristics of A-T, we have examined signaling through the BCR. Initially, control and A-T-EBV-transformed B cells were compared for proliferative response after BCR cross-linking. When thymidine incorporation was used as a measure of cell proliferation, it was evident that the control cell lines showed an increase over a 4-day period (Fig. 1), comparable with that observed by Simon et al. (31), also using EBV cell lines. In contrast, in four A-T cell lines there was little or no evidence of a proliferative response after cross-linking of the receptor, providing evidence for a defect in signaling in A-T cells.

In view of the apparent defect in mitogenesis through the BCR, B cell lines derived from A-T patients and controls were initially compared for expression of sIg by immunofluorescence staining. As shown in Fig. 2, levels of sIg detected on A-T cell lines were comparable with those on control cell lines. In addition, no difference in the levels of CD45 and CD19 surface expression or staining intensity was apparent between A-T and control cell lines, indicating that EBV transformation did not select out different B cell populations in the two cell types.

BCR cross-linking induces inositol phospholipid hydrolysis and intracellular Ca²⁺ mobilization in normal B cells manifested as a rapid transient increase in Ca²⁺ concentrations within the cell (25). Preliminary titration studies showed that anti-Ig concentration in the range of 9-36 µg/ml induced a maximal Ca²⁺ mobilization response in normal lymphoblastoid cells within 1 min of cross-linking of the receptor (results not shown). Cross-linking of BCR on lymphoblastoid cells from three different control cell lines resulted in a rapid increase in Ca²⁺, returning almost immediately to basal levels (Fig. 3). Increased Ca²⁺ mobilization after anti-Ig stimulation reflects release from intracellular stores since Ca²⁺ changes were studied in calcium depleted medium, which minimizes any effect of extracellular Ca²⁺ influx. However, as is evident in the case of C2ABR, delayed influx of Ca²⁺ is apparent and is seen for the other two controls (C3ABR and C30ABR) 10 min after stimulation (results not shown on the profile). In addition, the presence of EGTA (50 µM) in the medium did not alter the initial response. The peak concentration changes in Ca²⁺ in control cells ranged between 260 and 300 nM Ca²⁺, as determined by spectrofluorimetry. In contrast, cross-linking of BCR on three A-T lymphoblastoid cell lines from different A-T complementation groups resulted in no response or a very much diminished Ca²⁺ mobilization ranging from 0 to 10 nM (Fig. 3). In all, we screened five A-T lymphoblastoid cell lines and demonstrated that they showed a defect in Ca²⁺ mobilization in response to BCR cross-linking. Previous studies have shown that cross-linking of BCR, in cells transfected with the EBV latent membrane receptor gene LMP2A, failed to lead to changes in intracellular free Ca²⁺ (32, 33). To rule out the possibility that defective Ca²⁺ mobilization in A-T was due to selective overexpression of LMP2A in these cells, we determined expression levels by Western blotting. The results in Fig. 4 show that LMP2A expression in three A-T lines (AT1ABR, AT5ABR, AT3ABR) is comparable with that in controls (C3ABR, C2ABR). It was not possible to detect expression of LMP2A in a fourth A-T cell line, AT3LA. EBV-negative Burkitt's lymphoma cell line DG75 was used as a negative control, and no LMP2A expression was detected in these cells.

Activation of Phospholipase C1

In view of the reduced or absent Ca²⁺ mobilization in A-T cells, it seemed likely that one arm of the cascade activated by BCR cross-linking, namely activation of PLC1 (24), would be defective in A-T cells. Phosphorylation of PLC1 isozyme was assessed in A-T and control B cells by immunoprecipitation with rabbit anti-PLC1 antibodies followed by Western blotting with anti-phosphotyrosine antibody. As is evident from the results in Fig. 5 (top), the extent of tyrosine phosphorylation of PLC1 was reduced in AT1ABR cells incubated with anti-Ig and negligible in two other A-T cell lines, AT5ABR and AT3LA, compared with a control cell line, C3ABR. The difference in response between the A-T cell lines may be explained by the nature of the mutations in the ATM gene. AT1ABR, which showed only a reduced response, has an in-frame deletion of 3 amino acids (11) and is capable of producing a near complete-sized protein.² On the other hand, AT5ABR and AT3LA have mutations that are predicted to give rise to a truncated protein. The reduced response is not a consequence of differences in PLC1 expression in A-T and control cells, since immunoblot analysis revealed that levels of PLC1 protein were similar in A-T and control cell lines (Fig. 5, bottom). Furthermore, when levels of IP3 were measured after anti-Ig stimulation by the specific D-myo-[³H]Ins-1,4,5 P3 assay system using an Amersham kit, an increase was observed in two control cell lines (C2ABR, C3ABR), but two A-T cell lines (AT1ABR, AT3ABR) failed to show an increase (data not shown). These results demonstrate that induction of tyrosine phosphorylation of PLC1, a key event in the activation of this enzyme, after BCR ligation, is impaired in B cells from A-T patients and is consistent with the finding that anti-Ig-induced increases in Ca²⁺ mobilization are markedly reduced or negligible in these cells compared with B cells from controls (Fig. 3).

Activation of PLC1 by cross-linking of BCR.

The results described above would predict that the initial event after cross-linking of the receptor, activation of protein tyrosine kinases, would also be abnormal in A-T cells. Ligation of the receptor in control cells was shown to be associated with increased tyrosine phosphorylation of several proteins, predominantly with molecular masses of 55 and 85 kDa; as determined by immunoblotting with anti-phosphotyrosine antibody (Fig. 6), a comparable pattern of phosphorylations was evident in AT1ABR cells. In contrast, two of the A-T cell lines, AT3LA and AT5ABR, were defective in tyrosine phosphorylation of these protein species. Overall, there was either no phosphorylation evident or a marked reduction in protein phosphorylation for AT5ABR (Fig. 6), and AT3LA was defective for 85-kDa phosphorylation but appeared to have a normal response for the 55-kDa species. Both cell lines are characterized by mutations that give rise to protein truncations (11).² To investigate this defect in tyrosine phosphorylation, we measured tyrosine phosphorylation of the Src-related kinase (35, 36), Lyn, which is expressed in B cells. Lyn was immunoprecipitated at 0, 2, and 5 min after cross-linking and Western blotting of the immunoprecipitates from control cells with antibody to phosphotyrosine revealed a 2-3-fold increase in tyrosine phosphorylation of Lyn (upper doublet) on BCR cross-linking (Fig. 7A). The lower band of this triplet set represents Ig. However, little or no increase of tyrosine phosphorylation of Lyn was observed in AT3LA and AT5ABR (Fig. 7A). The amount of Lyn as determined by immunoblotting with anti-Lyn antibody did not change in any of the cell lines in response to receptor cross-linking as reported previously (Fig. 7B). However, the basal level of Lyn in AT5ABR was reduced compared with the control and other A-T cell lines.

Cross-linking of the BCR also activates the enzyme PI 3-kinase through association with Src kinases (37, 38). This enzyme has dual specificity, phosphorylating both the 3-position of inositol phospholipids (12) and its own regulatory subunit, p85 (39). Previous studies have shown that receptor ligation leads to an increase in PI 3-kinase associated with Lyn (36). The p85 subunit binds to kinases through the interaction between the SH3 (Src homology 3) domains of the kinases and proline-rich sequences in p85 (40). In parallel with the BCR-mediated activation of Lyn kinase, a phosphoprotein of 85 kDa was increased 6-fold after BCR cross-linking of control cells, but no increase was evident in two of the A-T lines, AT3LA and AT5ABR (Fig. 7A, upper band), and the increase was minimal in two other lines, AT1ABR and AT3ABR, compared with control (results not shown). The increase in tyrosine-phosphorylated p85 was attributed, at least in part, to an increased amount of p85 associating with Lyn after receptor cross-linking. Using p85 antibody on blots from Lyn immunoprecipitates, we observed a 3-fold and a 2-fold increase in Lyn-associated p85 at 2 and 5 min, respectively, after cross-linking, and this increase was absent in two of the A-T lines, AT3LA and AT5ABR (Fig. 7C). This also suggests that a proportion of phosphotyrosine-containing PI 3-kinase is associated with Lyn in control cells. This was verified when p85 immunoprecipitates were immunoblotted with anti-phosphotyrosine antibody (Fig. 7D). We observed a 2-3-fold increase in tyrosine phosphorylation of p85 in control cells that was absent in two A-T lines, AT3LA and AT5ABR. Resting levels of p85 in control and A-T cell lines were comparable (data not shown). A 150-kDa phosphoprotein was also observed in the PI 3-kinase immunoprecipitates and showed a 6-fold increase in tyrosine phosphorylation in control cells, but not in A-T cells, after BCR cross-linking (Fig. 7D).

Prior studies have shown that anti-Ig-mediated phosphorylation of the p85 subunit of PI 3-kinase is accompanied by increased enzymatic activity (37). We examined Lyn-associated PI 3-kinase activity at 0, 2, and 5 min after BCR cross-linking of control and A-T cells. The PI 3-kinase activity in anti-Lyn immunoprecipitates was increased by ~5-fold, as determined by densitometry, at 5 min post-cross-linking in control cell extracts (Fig. 8). However, no increase in Lyn-associated PI3-kinase activity was observed in two A-T lines, AT3LA and AT5ABR (Fig. 8), while a reduced response was demonstrated in two other lines, AT1ABR and AT3ABR (results not shown).

DISCUSSION

When the pleiotropic nature of the A-T phenotype is considered, it is evident that the gene involved, ATM, must be a regulator of a variety of cellular events related to differentiation, development, and DNA damage response. The recent cloning and description of this gene (11), with homologies to a family of PI 3-kinases involved in cell signaling pathways (16), supports several previous observations that point to a defect in signal transduction in this syndrome (19, 21, 41-43). We have provided evidence that A-T cells are defective in signaling through the BCR. This is not due to abnormalities in B cell ontogeny since the A-T lymphoblastoid cells used had similar levels of expression of sIg and several coreceptors to those observed in the control cell lines. In this study, we have demonstrated that control EBV-transformed B cells are triggered to mobilize Ca²⁺ in response to BCR cross-linking which is in agreement with previous reports (24, 31, 44). However, other studies have shown that EBV latent membrane protein 2A (LMP2A) expression blocks Ca²⁺ mobilization in B lymphocytes (32, 33). The Ca²⁺ mobilization defect observed here for A-T cells cannot be accounted by selective overexpression of LMP2A in A-T cells.

The data reported here indicate that A-T B-cells are defective in their ability to transduce Ca²⁺ mobilization and tyrosine phosphorylation signals in response to sIg cross-linking. The results suggest that disruption of early signaling events in these cells reflects a defect at a point upstream of tyrosine phosphorylation of PLC1, related to dysfunction of tyrosine kinases such as Lyn. In view of the putative role for Ig and Ig (45), it is possible that failure to phosphorylate Ig and Ig may be directly responsible for reduced activation of PLC1 and Ca²⁺ release in A-T B cells. The data obtained here suggest that at least some of the immunological abnormalities in A-T may reflect impaired ability of lymphocytes to transduce activation signals. While the immunodeficiency profile observed in A-T (10) can be distinguished readily from that seen in other immunodeficiency syndromes, some interesting parallels exist in the nature of the defects involved. Wiskott-Aldrich syndrome is an X-linked recessive disorder characterized by severe thrombocytopenia, eczema, profound immunodeficiency involving both B and T lymphocytes, and increased risk for lymphoid tumors as in A-T (46-48). A gene mutated in Wiskott-Aldrich syndrome, WASP, has been cloned recently and shown to code for a proline-rich protein (49). The protein has been shown to interact with the SH3 domain of the adapter protein Nck (50) and with members of Rho GTPases (Rac and Cdc42) (51), suggesting that Wiskott-Aldrich syndrome results from a defect in signal transduction. It is also noteworthy that defective transmembrane signaling through the BCR has been demonstrated in lymphoblastoid cell lines from these patients (31). These cells had a reduced proliferative response and a markedly decreased mobilization of Ca²⁺ after cross-linking of the BCR. Furthermore, the defects causing three X-linked human immunodeficiencies, agammaglobulinemia, hyper-IgM syndrome, and X-linked severe combined immunodeficiency are also in genes important in lymphocyte development and intracellular signaling (48, 52).

The increased radiosensitivity, chromosomal instability, and defective cell cycle checkpoint control that characterize A-T (3, 41, 53, 54) strongly support a role for the ATM protein in radiation signal transduction. These observations together with the data described here suggest that the ATM protein is involved not only in DNA damage response but also in other aspects of intracellular signaling. In this respect, it compares most closely with the Mei-41 Drosophila mutant which is characterized by sensitivity to DNA-damaging agents, chromosome aberrations, lack of G₂/M checkpoint control in response to X-irradiation as well as defects in oogenesis and embryogenesis (17). Both the ATM and Mei-41 proteins which are related through PI 3-kinase and an adjacent domain, appear to play multiple roles in addition to their functions as DNA damage sensors. Since the protein encoded by the ATM gene is 350.6 kDa in size, based on the complete cDNA sequence (55), it is likely to have other functional domains in addition to the kinase domain. Such a multifunctional protein could account for the varied phenotype observed in A-T (3). It is well established that PI 3-kinase mediates cellular responses to a variety of stimuli including growth factors and hormones to trigger a series of intracellular events (12). The versatility of this signaling system is potentially enhanced by the presence of subtypes of both the catalytic (p110) and regulatory (p85) subunits of PI 3-kinase (12).

The signaling defect described here is not confined to B cells from A-T patients. Kondo et al. (22) demonstrated that the proliferative response and interleukin-2 production of peripheral blood mononuclear cells to T cell mitogens was reduced in A-T patients. In addition, phytohemagglutinin and OKT3 (anti-CD3) only slightly increased intracellular Ca²⁺ in T cells and CD4⁺ cells. They interpreted these data to mean that the functional defect in T cells from A-T patients was caused by a defective Ca²⁺-dependent signal transduction through the CD3 complex. Further support for this proposition comes from the work of O'Connor and Scott-Linthicum (21), who demonstrated that phytohemagglutinin bound to and was internalized by A-T lymphocytes, but it failed to generate a blastogenic response. They suggested that the defect in A-T cells was due to failure to transmit a cytoplasm to nuclear signal, providing early evidence for a defect in signal transduction. The recent localization of ATM to vesicular structures in the microsomes as well as to the nucleus would support such a role in signal transduction.²

ATM may function in the microsomes to detoxify or to respond to reactive oxygen intermediates as part of a redox sensor pathway (56). It has been suggested that reactive oxygen intermediates act as second messengers in the regulation of cellular processes (57) and that antioxidants inhibit the proliferation of activated T cells (58). In addition, hydrogen peroxide is a potent activator of T lymphocyte proliferation and gene expression (59). Recent evidence suggests that src protein kinases, thought to be proximal components in the NF-B signaling cascade, are stimulated by oxidants and inhibited by antioxidants (34). It is possible that oxidants activate membrane-bound tyrosine kinases via the ATM protein present in vesicles associated with the plasma membrane. Such a role for ATM could explain the defective signaling observed in this study after cross-linking of BCR and observed elsewhere with T cells (22).

Defective signal transduction provides an attractive and plausible explanation for impaired B cell function in A-T patients and may underlie other facets of A-T cellular dysfunction. Although it is unclear whether impaired B cell signal transduction in A-T represents a direct or indirect consequence of the primary gene defect, complementation of the defect in A-T B cells should provide valuable insights into the molecular basis for expression of the A-T phenotype. However, generation of full length ATM cDNA (9.9 kilobases) has thus far proved difficult because of the instability at the 5 end of cDNA. The extent to which signaling is altered in other cell lineages of A-T patients remains to be determined. A defect in a signaling protein such as ATM could account for the spectrum of immunological abnormalities observed in this syndrome as well as the other defects.