Search

For the development of DNA vaccines for B-cell malignancies, three aspects were studied. First, tumour Ig genes of B-cell malignancies were analysed to gain insight into the nature of the tumour antigen and the cell of origin in B-cell tumours. Transcripts of VHDJH-constant region (VH-C) sequences of the rare B-cell malignancy hairy cell leukemia (HLC) were studied. They were found mutated, with a low level of intraclonal heterogeneity, and multiple Ig isotype transcripts with identical VH sequence were identified in single cells. These findings indicated that clonally-related multiple isotypes co-existed in single hairy cells, which were arrested at the point of isotype switch, where RNA processing precedes deletional recombination. Second, DNA vaccines encoding the scFv-FrC fusion gene were prepared from patients with B-cell malignancies selected for clinical trials. Each vaccine was characterised in vitro and in mice prior to human immunization. As a prelude to clinical use for lymphoma, the conformational status of the scFv encoded protein was evaluated in mice by monitoring the ability of the human DNA scFv-FrC vaccines to induce antibodies (Abs) directed against the private Id determinants of the tumour Ig. Non cross-reacting Abs specific to the individual patient's Ig were detected. In each case, there was clear evidence that Abs against private determinants were produced. Ab responses were predominantly driven by Th2 cells. These data demonstrated that DNA scFv-FrC fusion vaccines produced folded protein displaying Id determinants able to induce Ab against natural tumor Ig and set the scene for the application of DNA scFv-FrC vaccines to patients with B-cell malignancies. Third, a DNA based approach was developed with a second tumor antigen for myeloma, MUC1 glycoprotein. Initially, using a panel of anti-MUC1 monoclonal Abs and an RT-PCR technique, MUC1 was confirmed to be expressed by neoplastic myeloma cells. Then, using the same strategy applied to Ig determinants, DNA vaccines encoding human MUC1 protein alone or fused to the FrC were tested in mice for their ability to induce anti-tumor immunity. Fusion of MUC1 to FrC amplified anti-MUC1 Ab responses, but did not appear to increase protection against challenge with MUC1 expressing tumour cells. CD8 T cells seemed to be more important since their depletion in vivo abolished tumour protection. Two peptides that were able to induce cytotoxicity against MUC1 in vitro were identified and will be included in a new generation of DNA vaccines.

Mantle cell lymphoma (MCL) is a mature B-cell lymphoma typically characterised by the expression of cluster of differentiation 5 (CD5) and a t(11;14) translocation resulting in overexpression of cyclins. MCL displays significant morphological and clinical heterogeneity. Like chronic lymphocytic leukaemia (CLL), MCL can be divided into two subsets based upon immunoglobulin heavy chain variable (IGHV) region status: the more common unmutated (U-) IGHV and less common mutated (M-) IGHV subset. It has been suggested that, like in CLL, IGHV status informs different origin and clinical behaviour, with U-MCL being of pre-germinal centre (pre-GC) origin and aggressive clinical behaviour and M-MCL of post-GC origin and indolent course. In CLL, B-cell receptor (BCR) levels and signalling capacity have also helped to define subgroups with important clinical and biological significance. This thesis aimed to extend investigations to the BCR of MCLs and understand if BCR expression and signalling variation in MCL could affect variation in clinical behaviour. A cohort of thirty-six MCL cases were assessed for immunogenetic, phenotypic and functional characteristics including signalling capacity. IGHV3-21 and 4-34 were equally identified in U-MCL whilst IGHV4-34 was most common in M-MCL. Fifty percent (%) had an U-IGHV (>98% homology to germline). Surface immunoglobulin (sIg) M was expressed variably but significantly higher in conventional, nodal MCL (p=0.01) and non-aggressive U-MCL (p=0.03) than in leukaemic non-nodal (LNN) or M-MCL. Surface IgM (sIgM) expression was higher in MCL than in CLL (p < 0.01). Mean signalling capacity was significantly higher in MCL than CLL (p < 0.01), and correlated positively with sIgM expression. A subgroup of MCL had low signalling capacity similar to CLL. This 'CLL-like' signalling group was composed predominantly of LNN MCL and IGHV3-21 U-MCL with low sIgM expression. In contrast to high signalling MCL cases, this group had a CLL-like BCR engaged functional phenotype: recoverable sIgM expression in vitro, high basal proximal signalling kinases and an 'activated' glycosylation profile on the constant Ig region. MCL tissue and matched lymph node and peripheral blood samples were assessed. The lymph node compartment was identified as the likely site of engagement in non-aggressive MCL; sIgM levels were significantly lower in the lymph node than in the peripheral blood in all tissue and paired samples (p=0.03). Signalling capacity was also significantly lower in lymph node tissue samples (p < 0.01). Lower sIgM levels in the lymph node may be indicative of site occurred antigen-induced endocytosis. M-MCL with low sIgM was associated with improved event-free survival (p=0.04). This study reveals that a proportion of MCL have a BCR engaged signature with low sIgM and low signalling capacity. This likely occurs in the lymph node compartment as a result of the influence of putative (super)antigen, with down-modulation of sIgM. This is particularly prominent in LNN M-MCL and may affect the balance between proliferation and anergy. This mirrors other mature B-cell malignancies, particularly CLL, where this interaction influences clinical outcomes.

B-cell receptor (BCR) signalling plays a major role in the pathogenesis of chronic lymphocytic leukaemia (CLL) and other B-cell malignancies. In addition to its ability to promote cell accumulation, BCR signalling also influences the function of other cell surface receptors (receptor cross-talk), including G-protein couple receptors (GPCR). This project addresses two inter-connected areas related to BCR signalling in malignant B cells. First, it explores pathways of cross-talk between the BCR and the GPCRs CXCR4 and EBI2, including the role of kinases and the inositol lipid phosphatase SHIP1. Second, it identifies novel pathways regulated by a chemical SHIP1 activator, AQX-435. Established B-cell lines were selected to investigate BCRCXCR4 to extend previous studies performed using primary CLL cells. This revealed that, similar to CLL cells, anti-IgM-induced CXCR4 down-modulation was dependent primarily on SYK and PI3K, but not BTK. However, in contrast to CLL cells, activation of SHIP1 using AQX-435 was not sufficient to down-modulate CXCR4 in B-cell lines. One aim of these cell line studies was to perform RNAi transfection studies to investigate the role of individual signalling intermediates in BCRCXCR4 cross-talk. However, it was not possible to achieve sufficient target knock-down with RNAi for reliable characterisation. Investigation of cross-talk was extended to EBI2 which, like CXCR4, is involved in positioning B cells during immune responses. Overall, EBI2 was expressed at low levels on CLL cells compared to normal B cells and this low level did not seem to be sufficient to support signalling or survival responses following ligand stimulation. In contrast to CXCR4, there was no strong evidence that EBI2 was subject to regulation either by anti-IgM or AQX-435. Analysis of RNA-seq data demonstrated that AQX-435 and the PI3K inhibitor idelalisib shared the ability to reduce anti-IgM-induced changes in gene expression in CLL cells. However, AQX-435 also triggered a unique transcriptional response driven by activation of liver X receptor (LXR), potentially linked to SHIP1 activation. LXR activation drives cholesterol efflux and functional experiments demonstrated that supplementation of cholesterol using mevalonic acid significantly reduced AQX-435-induced CLL cell apoptosis. Overall, these experiments shed important new light on BCR signalling in malignant B cells, and support the idea that SHIP1 activators, such as AQX-435, are attractive potential therapeutic agents.