BAG-1 as a biomarker in early breast cancer prognosis: a systematic review with meta-analyses.

BACKGROUND: The co-chaperone protein Bcl-2-associated athanogene-1 (BAG-1) is overexpressed in breast cancer and has been incorporated in the oncotype DX and PAM50 breast cancer prognostic assays. Bcl-2-associated athanogene-1 exists as multiple protein isoforms that interact with diverse partners, including chaperones Hsc70/Hsp70, Ser/Thr kinase Raf-1 and Bcl-2, to promote cancer cell survival. The BAG-1L isoform specifically binds to and increases the transcriptional activity of oestrogen receptor in cells, and in some, but not all studies, BAG-1 expression is predictive of clinical outcome in breast cancer. METHODS: A systematic review of published studies reporting BAG-1 (mRNA and/or protein) expression and clinical outcome in early breast cancer. The REporting Recommendations for Tumour MARKer and Prognostic Studies (REMARK) criteria were used as a template against which data were assessed. Meta-analyses were performed for studies that provided a hazard ratio and 95% confidence intervals for clinical outcomes including disease-free survival or breast cancer-specific survival from univariate analysis. RESULTS: Eighteen studies used differing methodologies and reported on differing outcomes. Meta-analyses were only possible on results from a subset of reported studies. Meta-analyses suggested improved outcome with high BAG-1 mRNA and high BAG-1 nuclear expression by immunohistochemisty. CONCLUSIONS: Increased levels of BAG-1 are associated with better breast cancer outcomes.

FOXO3 expression during colorectal cancer progression: biomarker potential reflects a tumour suppressor role.

BACKGROUND: In previous studies, the Forkhead/winged-helix-box-class-O3 (FOXO3) transcription factor has displayed both tumour suppressive and metastasis-promoting properties.To clarify its role in human colorectal cancer (CRC) progression, we examined in vivo FOXO3 expression at key points of the metastatic cascade. METHODS: Formalin-fixed paraffin-embedded resection specimens from normal colon, adenomas, primary CRC specimens of different pathological stage and CRC specimens with matched liver metastases were used to generate three separate custom-designed tissue microarray (TMA) representations of metastatic progression. Triplicate cores, immunostained for FOXO3 were scored semiquantitatively by two investigators. RESULTS: The FOXO3 expression is significantly reduced in CRC specimens compared with normal tissue, and progressive FOXO3 downregulation is associated with advancing pathological stage. In addition, recurrent stage I/II primary tumours show a significantly lower FOXO3 expression compared with stage-matched non-recurrent tumours. When stratified according to high and low FOXO3 expression, mean disease-free survival in the low-expressing group was 28 months (95% CI 15.8-50.6) compared with 64 months (95% CI 52.9-75.4) in the high-expressing group. CONCLUSION: We have demonstrated an association between low FOXO3 expression and CRC progression in vivo using purpose-designed TMAs. Forkhead/winged-helix-box-class-O3 may represent a novel biomarker of nodal and distant disease spread with clinical utility in CRC.

The BAG-1 cochaperone is a negative regulator of p73-dependent transcription.

High-level expression of Bcl-2 associated athanogene (BAG-1) protects cancer cells from stress-induced cell death and growth inhibition. These protective effects of BAG-1 are dependent on interactions with the HSC70 and HSP70 chaperones. However, the key stress-response molecules that are regulated by a BAG-1/chaperone mechanism have not been identified. In this study, we investigated the effects of BAG-1 overexpression on the function of p53 family proteins, p53, p63 and p73. Overexpression of BAG-1 isoforms interfered with the transactivating activity of p73 and p63, but had modest and variable effects on p53-dependent transcription. p73 and BAG-1 interacted in intact cells and overexpression of BAG-1 decreased the expression of p73. siRNA-mediated ablation of endogenous BAG-1 increased the activity of a p73-responsive promoter and this was reversed by knock-down of p73. The ability of BAG-1 to modulate p73 activity and expression, and to interact with p73 were dependent on amino acid residues required for the interaction of BAG-1 with HSC70 and HSP70. These results show that BAG-1 inhibits the transactivating functions of p73 and provide new insight into the mechanisms that control the expression of p73. Inhibition of p73 function may be one mechanism that contributes to the pro-survival activity of BAG-1.

Distinct promoters mediate constitutive and inducible Bcl-XL expression in malignant lymphocytes.

Bcl-X(L) is a Bcl-2-related survival protein that is essential for normal development. Bcl-X(L) expression is rapidly induced by a wide range of survival signals and many cancer cells constitutively express high levels. The Bcl-X gene has a complex organization with multiple promoters giving rise to RNAs with alternate 5' non-coding exons. Here we have investigated the mechanisms that control basal and induced expression of Bcl-X(L) in B-lymphoma cells. Antisense experiments demonstrated that Bcl-X(L) was essential for survival of Akata6 B-lymphoma cells. The levels of RNAs containing the IB Bcl-X non-coding exon, derived from the distal 1B promoter, correlated with basal expression of Bcl-X(L) in primary malignant B cells and this promoter was highly active in B-cell lines. The activity of this promoter was largely dependent on a single Ets binding site and Ets family proteins were bound at this promoter in intact cells. CD40 ligand (CD40L)-induced cell survival was associated with increased Bcl-X(L) expression and accumulation of exon IA-containing RNAs, derived from the proximal 1A promoter. Nuclear factor-kappaB (NF-kappaB) inhibition prevented induction of Bcl-X(L) protein and exon IA-containing RNAs by CD40L. Therefore, the distal Bcl-X 1B promoter plays a critical role in driving constitutive expression-mediated via Ets family proteins in malignant B cells, whereas NF-kappaB plays a central role in the induction of Bcl-X(L) in response to CD40 signalling via the proximal 1A promoter.

Pharmacological inhibitors of NF-kappaB accelerate apoptosis in chronic lymphocytic leukaemia cells.

Nuclear factor-kappaB (NF-kappaB) is a transcription factor that plays a critical role in the inappropriate survival of various types of malignant cells. Chronic lymphocytic leukaemia (CLL) is the most common B-cell malignancy in the Western world. Although overexpression and regulation of NF-kappaB has been described in CLL, its function remains unclear. Exposure of CLL cells to BAY117082 or Kamebakaurin, potent pharmacological inhibitors of the NF-kappaB pathway, accelerated apoptosis in approximately 70% of cases. Sensitivity to NF-kappaB pathway inhibitors was not related to the prognostic markers VH status, CD38 or Zap70 expression, or to the levels of nuclear NF-kappaB. Normal peripheral B cells were resistant to the apoptosis-inducing effects of these compounds. Cell death induced by the inhibitors was associated with activation of caspase-9 and -3, and loss of mitochondrial membrane polarization, but did not involve changes in the expression of Bcl-2 or Mcl-1. Inhibitors caused an increase in c-jun NH2-terminal kinase activity in CLL, but this did not appear to be important for apoptosis. Microarray analysis identified some potential novel NF-kappaB target genes, including interleukin-16- and the Bcl-2- related survival protein Bcl-w. These results demonstrate that a substantial proportion of CLL are dependent on NF-kappaB for enhanced survival and suggest that inhibition of NF-kappaB may have therapeutic potential.

Will histone deacetylase inhibitors require combination with other agents to fulfil their therapeutic potential?

Histone deacetylase inhibitors have progressed rapidly from the laboratory to clinical testing. This review highlights the promising data for their combination with a wide range of established and novel anticancer agents and discusses the mechanisms that underpin these interactions.

Ornithine decarboxylase is a mediator of c-Myc-induced apoptosis.

c-Myc plays a central role in the regulation of cell cycle progression, differentiation, and apoptosis. However, the proteins which mediate c-Myc function(s) remain to be determined. Enforced c-myc expression rapidly induces apoptosis in interleukin-3 (IL-3)-dependent 32D.3 murine myeloid cells following IL-3 withdrawal, and this is associated with the constitutive, growth factor-independent expression of ornithine decarboxylase (ODC), a rate-limiting enzyme of polyamine biosynthesis. Here we have examined the role of ODC in c-Myc-induced apoptosis. Enforced expression of ODC, like c-myc, is sufficient to induce accelerated death following IL-3 withdrawal. ODC induced cell death in a dose-dependent fashion, and alpha-difluoromethylornithine (DFMO), an irreversible inhibitor of ODC enzyme activity, effectively blocked ODC-induced cell death. ODC-induced cell death was due to the induction of apoptosis. We also demonstrate that ODC is a mediator of c-Myc-induced apoptosis. 32D.3-derived c-myc clones have augmented levels of ODC enzyme activity, and their rates of death were also a function of their ODC enzyme levels. Importantly, the rates of death of c-myc clones were inhibited by treatment with DFMO. These findings demonstrate that ODC is an important mediator of c-Myc-induced apoptosis and suggest that ODC mediates other c-Myc functions.

E2F-1:DP-1 induces p53 and overrides survival factors to trigger apoptosis.

The E2F DNA binding activity consists of a heterodimer between E2F and DP family proteins, and these interactions are required for association of E2F proteins with pRb and the pRb-related proteins p107 and p130, which modulate E2F transcriptional activities. E2F-1 expression is sufficient to release fibroblasts from G0 and induce entry into S phase, yet it also initiates apoptosis. To investigate the mechanisms of E2F-induced apoptosis, we utilized interleukin-3 (IL-3)-dependent 32D.3 myeloid cells, a model of hematopoietic progenitor programmed cell death. In the absence of IL-3, E2F-1 alone was sufficient to induce apoptosis, and p53 levels were diminished. DP-1 alone was not sufficient to induce cell cycle progression or alter rates of death following IL-3 withdrawal. However, overexpression of both E2F-1 and DP-1 led to the rapid death of cells even in the presence of survival factors. In the presence of IL-3, levels of endogenous wild-type p53 increased in response to E2F-1, and coexpression of DP-1 further augmented p53 levels. These results provide evidence that E2F is a functional link between the tumor suppressors p53 and pRb. However, induction of p53 alone was not sufficient to trigger apoptosis, suggesting that the ability of E2F to override survival factors involves additional effectors.

PI3Kδ inhibition elicits anti-leukemic effects through Bim-dependent apoptosis.

PI3Kδ plays pivotal roles in the maintenance, proliferation and survival of malignant B-lymphocytes. Although not curative, PI3Kδ inhibitors (PI3Kδi) demonstrate impressive clinical efficacy and, alongside other signaling inhibitors, are revolutionizing the treatment of hematological malignancies. However, only limited in vivo data are available regarding their mechanism of action. With the rising number of novel treatments, the challenge is to identify combinations that deliver curative regimes. A deeper understanding of the molecular mechanism is required to guide these selections. Currently, immunomodulation, inhibition of B-cell receptor signaling, chemokine/cytokine signaling and apoptosis represent potential therapeutic mechanisms for PI3Kδi. Here we characterize the molecular mechanisms responsible for PI3Kδi-induced apoptosis in an in vivo model of chronic lymphocytic leukemia (CLL). In vitro, PI3Kδi-induced substantive apoptosis and disrupted microenvironment-derived signaling in murine (Eµ-Tcl1) and human (CLL) leukemia cells. Furthermore, PI3Kδi imparted significant therapeutic responses in Eµ-Tcl1-bearing animals and enhanced anti-CD20 monoclonal antibody therapy. Responses correlated with upregulation of the pro-apoptotic BH3-only protein Bim. Accordingly, Bim-/- Eµ-Tcl1 Tg leukemias demonstrated resistance to PI3Kδi-induced apoptosis were refractory to PI3Kδi in vivo and failed to display combination efficacy with anti-CD20 monoclonal antibody therapy. Therefore, Bim-dependent apoptosis represents a key in vivo therapeutic mechanism for PI3Kδi, both alone and in combination therapy regimes.

IL-10 production by CLL cells is enhanced in the anergic IGHV mutated subset and associates with reduced DNA methylation of the IL10 locus.

Chronic lymphocytic leukemias (CLLs) with unmutated (U-CLL) or mutated (M-CLL) IGHV have variable features of immunosuppression, possibly influenced by those CLL cells activated to produce interleukin 10 (IL-10). The two subsets differ in their levels of anergy, defined by low surface immunoglobulin M levels/signaling capacity, and in their DNA methylation profile, particularly variable in M-CLL. We have now found that levels of IL-10 produced by activated CLL cells were highly variable. Levels were higher in M-CLL than in U-CLL and correlated with anergy. DNA methylation analysis of IL10 locus revealed two previously uncharacterized 'variably methylated regions' (CLL-VMRs1/2) in the gene body, but similarly low methylation in the promoter of both U-CLL and M-CLL. CLL-VMR1/2 methylation was lower in M-CLL than in U-CLL and inversely correlated with IL-10 induction. A functional signal transducer and activator of transcription 3 (STAT3) binding site in CLL-VMR2 was confirmed by proximity ligation and luciferase assays, whereas inhibition of SYK-mediated STAT3 activation resulted in suppression of IL10. The data suggest epigenetic control of IL-10 production. Higher tumor load may compensate the reduced IL-10 production in U-CLL, accounting for clinical immunosuppression in both subsets. The observation that SYK inhibition also suppresses IL-10 provides a potential new rationale for therapeutic targeting and immunological rescue by SYK inhibitors in CLL.

A minimal Bcl-x promoter is activated by Brn-3a and repressed by p53.

The Brn-3a transcription factor stimulates the expression of the anti-apoptotic Bcl-2 and Bcl-x proteins and protects neuronal cells from apoptosis. Here we show that a minimal Bcl-x promoter is activated by Brn-3a and that this stimulation is prevented by the pro-apoptotic p53 protein. Both these effects are mediated via Bcl-x promoter sequences, which are indistinguishable from those required for minimal basal promoter activity. A newly described upstream Bcl-x promoter is also activated by Brn-3a with this activation being prevented by p53. Hence, Brn-3a-mediated activation of two distinct Bcl-x promoters and of the Bcl-2 promoter is blocked by p53 whereas this is not observed for Brn-3a activated promoters derived from genes not involved in inhibiting apoptosis. p53 therefore appears to specifically target the activation by Brn-3a of promoters derived from genes with an anti-apoptotic effect and this may be involved in the pro-apoptotic activity of p53.

Differential effects of estrone and estrone-3-O-sulfamate derivatives on mitotic. Arrest, apoptosis, and microtubule assembly in human breast cancer cells.

There is considerable interest in the potential use of estrogen derivatives for the treatment and prevention of breast cancer. We demonstrated previously that the sulfamoylated estrone derivative 2-methoxyestrone-3-O-sulfamate (2-MeOEMATE) induced G2-M cell cycle arrest and modest levels of apoptosis in breast cancer cells in vitro, whereas the parent estrone derivative, 2-methoxyestrone, did not. 2-MeOEMATE also induced breast tumor regression in vivo in intact rats. To further explore the significance of sulfamoylation on the anticancer activity of estrone derivatives and to elucidate their mechanism of action, we synthesized two additional agents, 2-ethylestrone and 2-ethylestrone-3-O-sulfamate (2EtEMATE). 2-MeOEMATE and 2-EtEMATE inhibited the growth of a panel of estrogen receptor-negative and -positive breast cancer cell lines in vitro, induced mitotic arrest and apoptosis, and suppressed the long-term clonogenic potential of MCF7 and CAL51 breast cancer cells. In each assay, the sulfamoylated estrone derivatives were >10-fold more potent than their parent compounds. The sulfamoylated estrone derivatives were also significantly more potent inhibitors of cell growth than the previously studied endogenous estradiol metabolite 2-methoxyestradiol. 2-MeOEMATE and 2-EtEMATE functioned as antimicrotubule agents and inhibited the ability of paclitaxel to promote tubulin assembly in vitro. Like other antimicrotubule agents, the sulfamoylated estrone derivatives induced BCL-2 and BCL-XL phosphorylation and increased p53 expression. 2-MeOEMATE and 2-EtEMATE are novel antimicrotubule agents that have potent anticancer activity in breast cancer cells in vitro and may be beneficial as anticancer agents in vivo.

BCR-signaling-induced cell death demonstrates dependency on multiple BH3-only proteins in a murine model of B-cell lymphoma.

Genetic recombination during B-cell development regularly results in the generation of autoreactive, potentially pathogenic B-cell receptors (BCRs). Consequently, multiple mechanisms link inappropriate BCR specificity to clonal deletion. Similar pathways remain in malignant B cells, offering the potential for targeting BCR signaling. Recently, small molecule inhibitors have realized this potential and, therefore, a deeper understanding of BCR-induced signaling networks in malignant cells is vital. The BH3-only protein Bim has a key role in BCR-induced apoptosis, but it has long been proposed that additional BH3-only proteins also contribute, although conclusive proof has been lacking. Here, we comprehensively characterized the mechanism of BCR-induced apoptosis in Eµ-Myc murine lymphoma cells. We demonstrate the upregulation of Bim, Bik, and Noxa during BCR signaling in vitro and that intrinsic apoptosis has a prominent role in anti-BCR antibody therapy in vivo. Furthermore, lymphomas deficient in these individual BH3-only proteins display significant protection from BCR-induced cell death, whereas combined loss of Noxa and Bim offers enhanced protection in comparison with loss of Bim alone. Some but not all of these effects were reversed upon inhibition of Syk or MEK. These observations indicate that BCR signaling elicits maximal cell death through upregulation of multiple BH3-only proteins; namely Bim, Bik, and Noxa.

A ribosome-related signature in peripheral blood CLL B cells is linked to reduced survival following treatment.

We have used polysome profiling coupled to microarray analysis to examine the translatome of a panel of peripheral blood (PB) B cells isolated from 34 chronic lymphocytic leukaemia (CLL) patients. We have identified a 'ribosome-related' signature in CLL patients with mRNAs encoding for ribosomal proteins and factors that modify ribosomal RNA, e.g. DKC1 (which encodes dyskerin, a pseudouridine synthase), showing reduced polysomal association and decreased expression of the corresponding proteins. Our data suggest a general impact of dyskerin dysregulation on the translational apparatus in CLL and importantly patients with low dyskerin levels have a significantly shorter period of overall survival following treatment. Thus, translational dysregulation of dyskerin could constitute a mechanism by which the CLL PB B cells acquire an aggressive phenotype and thus have a major role in oncogenesis.

The SF3B1 inhibitor spliceostatin A (SSA) elicits apoptosis in chronic lymphocytic leukaemia cells through downregulation of Mcl-1.

The pro-survival Bcl-2 family member Mcl-1 is expressed in chronic lymphocytic leukaemia (CLL), with high expression correlated with progressive disease. The spliceosome inhibitor spliceostatin A (SSA) is known to regulate Mcl-1 and so here we assessed the ability of SSA to elicit apoptosis in CLL. SSA induced apoptosis of CLL cells at low nanomolar concentrations in a dose- and time-dependent manner, but independently of SF3B1 mutational status, IGHV status and CD38 or ZAP70 expression. However, normal B and T cells were less sensitive than CLL cells (P=0.006 and P<0.001, respectively). SSA altered the splicing of anti-apoptotic MCL-1(L) to MCL-1(s) in CLL cells coincident with induction of apoptosis. Overexpression studies in Ramos cells suggested that Mcl-1 was important for SSA-induced killing since its expression inversely correlated with apoptosis (P=0.001). IL4 and CD40L, present in patient lymph nodes, are known to protect tumour cells from apoptosis and significantly inhibited SSA, ABT-263 and ABT-199 induced killing following administration to CLL cells (P=0.008). However, by combining SSA with the Bcl-2/Bcl-x(L) antagonists ABT-263 or ABT-199, we were able to overcome this pro-survival effect. We conclude that SSA combined with Bcl-2/Bcl-x(L) antagonists may have therapeutic utility for CLL.

Genomic disruption of the histone methyltransferase SETD2 in chronic lymphocytic leukaemia.

Histone methyltransferases (HMTs) are important epigenetic regulators of gene transcription and are disrupted at the genomic level in a spectrum of human tumours including haematological malignancies. Using high-resolution single nucleotide polymorphism (SNP) arrays, we identified recurrent deletions of the SETD2 locus in 3% (8/261) of chronic lymphocytic leukaemia (CLL) patients. Further validation in two independent cohorts showed that SETD2 deletions were associated with loss of TP53, genomic complexity and chromothripsis. With next-generation sequencing we detected mutations of SETD2 in an additional 3.8% of patients (23/602). In most cases, SETD2 deletions or mutations were often observed as a clonal event and always as a mono-allelic lesion, leading to reduced mRNA expression in SETD2-disrupted cases. Patients with SETD2 abnormalities and wild-type TP53 and ATM from five clinical trials employing chemotherapy or chemo-immunotherapy had reduced progression-free and overall survival compared with cases wild type for all three genes. Consistent with its postulated role as a tumour suppressor, our data highlight SETD2 aberration as a recurrent, early loss-of-function event in CLL pathobiology linked to aggressive disease.

Induction of ornithine decarboxylase by IL-3 is mediated by sequential c-Myc-independent and c-Myc-dependent pathways.

Enforced c-Myc expression promotes continuous, growth factor-independent, cell cycle progression and activates expression of the ornithine decarboxylase (ODC) gene and its promoter. c-Myc-responsiveness of murine ODC is mediated by two conserved c-Myc:Max E-boxes in ODC intron 1. c-Myc and ODC are both required for cell growth and their expression is sequentially induced in G0/G1 cells stimulated with mitogens, yet their expression is not modulated by the cell cycle in proliferating cells. Here we demonstrate that regulation of ODC and its promoter by Interleukin-3 (IL-3) in murine myeloid cells is mediated in part by c-Myc, c-Myc induced ODC through the same transcription start site as IL-3 and, in asynchronously growing cells, maximal activity of the ODC promoter required the intronic c-Myc binding sites. However, induction of ODC following IL-3 stimulation of quiescent cells is mediated by at least two pathways. The first phase of this response was independent of the intronic c-Myc:Max E-boxes and de novo protein synthesis. Sustained induction of the ODC promoter however required the c-Myc:Max binding sites and protein synthesis. Accumulation of c-Myc following stimulation of quiescent cells with IL-3 correlated with the delayed phase of the response. Consistent with a two pathway model of ODC regulation, inducible overexpression of dominant negative form of c-Myc (In373-Myc), which specifically inhibits the c-Myc-Max network, inhibited the delayed, but not immediate, induction of ODC promoter activity in response to IL-3. Dominant negative c-Myc protein also effectively suppressed induction of the endogenous ODC gene by IL-3. Therefore, c-Myc functions as a direct and required-regulator of ODC. These results also suggest a model whereby c-Myc's role in regulating its targets may be to convert a transient, immediate-early, activation event into the persistent induction of gene expression.

c-Myc induces apoptosis and cell cycle progression by separable, yet overlapping, pathways.

Enforced c-Myc expression promotes inappropriate cell cycle progression of growth factor deprived cells and triggers concomitant apoptosis. However, it is not clear what role dysregulation of the cell cycle plays in c-Myc-induced apoptosis. Ornithine decarboxylase (ODC) is a transcriptional target of c-Myc and contributes to c-Myc induced apoptosis. Here we have established that high levels of ODC overexpression in interleukin-3 (IL-3)-dependent 32D.3 myeloid cells induces apoptosis at rates comparable to those induced by enforced c-Myc expression. However, ODC-induced apoptosis was not accompanied by dysregulation of cell cycle controls, indicating that cell death was not triggered by inappropriate cell cycle progression. Nonetheless, ODC was required downstream of c-Myc for myeloid cell growth. These results suggested that c-Myc-induced pathways leading to cell cycle progression and apoptosis are separable, yet that they share common mediators. In agreement with this concept, treatment of cells over-expressing c-Myc with the growth inhibitory agent dibutyryl cyclic AMP (Bt2cAMP) arrested these cells G1, without inducing apoptosis. However, c-Myc retained the ability to induce apoptosis of Bt2cAMP-arrested cells following removal of IL-3, demonstrating that Bt2cAMP selectively inhibits c-Myc-induced pathways promoting cell cycle progression but not apoptosis. These results suggest a "multiple effectors' model in which c-Myc regulates the expression of mediators which alone are sufficient to induce apoptosis in the absence of survival factors, yet are required in concert to promote cell cycle progression.

BAX frameshift mutations in cell lines derived from human haemopoietic malignancies are associated with resistance to apoptosis and microsatellite instability.

Bax suppresses tumorigenesis in a mouse model system and Bax-deficient mice exhibit lymphoid hyperplasia suggesting that BAX functions as a tumour suppressor in human haemopoietic cells. We examined BAX expression in 20 cell lines derived from human haemopoietic malignancies and consistent with a potential tumour suppressor function, identified two cell lines, DG75 (a Burkitt lymphoma cell line) and Jurkat (a T-cell leukaemia line), which lacked detectable BAX expression. Apoptosis of DG75 cells induced by low serum or ionomycin was significantly delayed relative to similar Burkitt lymphoma cell lines with normal BAX levels. Although DG75 and Jurkat cells expressed several BAX RNA species including the prototypical BAX alpha RNA, the absence of BAX protein was due to single base deletions and additions in a polyguanine tract within the BAX open reading frame. These frameshift mutations result in premature termination of translation and have recently also been identified in some colon cancers with microsatellite instability. Although mismatch repair defects are not considered a common feature of haemopoietic malignancies, DG75 and Jurkat cells had widespread microsatellite instability and did not express detectable levels of MSH2. In Jurkat cells, lack of MSH2 expression was due to a point mutation in exon 13 of MSH2 resulting in premature termination of translation. Our results suggest that a pathway linking mismatch repair defects, BAX tumour suppressor frameshift mutations and resistance to apoptosis may be a key feature of some lymphomas and leukaemias.