Abstract
Contrary to expectation, we have shown that the absolute number of T cells in the peripheral blood of patients presenting with AML is higher than that of healthy individuals (p=0.0006). Previous publications have demonstrated that in vitro the microenvironment in AML is immunosuppressive to healthy T cell function (Buggins et al. 2001). To analyze the molecular nature of T cell defects in patients with AML, we performed global gene expression profiling of peripheral blood CD4 and CD8 T cells from 10 AML patients compared to that of 21 healthy volunteers. The purity of the CD4 T cells was 86% in AML and 94% in healthy donors and 92% for AML and 87% in healthy donors for CD8 T cells. cRNA was hybridised on Affymetrix U133Plus2 chips using standard protocols. Samples were normalised using the MAS5 algorithm and subsequent analysis using Bioconductor software indicated marked differences between the gene expression profiles of CD4 and CD8 T cells from AML patients compared with normal individuals. Using a false discovery rate of 0.01 and fold change greater than 2, 1407 genes were found to be differentially regulated for CD4 and 779 genes for CD8. There was a large overlap in the two gene lists. The microarray data was validated using real-time quantitative PCR. Initial pathway analysis using Ingenuity software indicated many genes involved in cell proliferation, cell death and gene expression. Molecular defects in T cells from AML patients show marked differences from those we have already noted in T cells from patients with CLL. Only 2% of genes for CD4 and 6% of genes for CD8 are differentially regulated in T cells from both AML and CLL patients. These data suggest a different molecular basis for the T cell defects in these disease types. However, 2 genes downregulated in CD8 T cells in AML patients, ACTN1 and FILIP1 suggested impaired actin cytoskeletal activity may contribute to immune dysfunction in AML T cells as we have already demonstrated in CLL. We therefore tested the ability of AML CD4 and CD8 T cells to form intact immunological synapses (IS) with autologous AML blasts in the presence or absence of superantigen. F-actin was visualised with rhodamine phalloidin and recruitment of activated T cell receptor mediated signalling molecules was detected using an anti-phosphotyrosine antibody. Conjugate and synapse formation and phosphotyrosine signalling was assessed by confocal microscopy with at least 50 random conjugates analyzed. Unlike CLL B cells, AML blasts can act as antigen presenting cells (APCs) for conjugate and IS formation with healthy T cells. Again, in contrast to CLL, AML T cells are able to form conjugates with autologous tumour cells but demonstrate impaired IS formation and, in CD8 T cells, decreased phosphotyrosine signalling at the synapse site. This data indicates that T cells in patients presenting with AML are abnormal in terms of their gene expression profile and although they retain the ability to conjugate with autologous blasts, reduced polarisation of F-actin to form immune synapses is seen. We are currently investigating if these changes are induced directly by tumour cells and the functional consequences of the gene expression changes identified. The ability of AML blasts to function as APCs for IS formation is in keeping with the observed GVL effect in this disease. Understanding the nature of T cell defects in patients presenting with AML is fundamental before successful autologous immunotherapeutic strategies can be implemented.
A gene expression profile for detection of sufficient tumour cells in breast tumour tissue: microarray diagnosis eligibility