Search

The t (l I 19)(p22 q23) translocation, which gives rise to the MLL-ENL fusion protein is commonly found in infant acute leukaemias of both the myeloid and lymphoid lineage. In order to study the molecular mechanism of haematopoietic progenitor cell (HPC) immortalisation by MLL-ENL, a conditional system of MLL- ENL expression was established in primary murine HPCs. This was achieved by delivering the Tet-Off inducible expression system to primary cells using two retroviral expression constructs. Several conditional immortalised myeloid cell lines were generated in vitro which were dependent on continued MLL-ENL expression for their survival and proliferation. The immortalised cells either terminally differentiated or died when MLL-ENL expression was turned off with doxycycline. Since several Hox genes are targets of MLL, the expression profile of all 39 murine Hox genes was analysed in the MLL-ENL immortalised cell lines by real-time quantitative PCR. Loss of MLL-ENL expression resulted in a decrease in the expression of multiple Hoxa genes. By comparing these changes with Hox gene expression in cells induced to differentiate with granulocyte-colony-stimulating factor, we found that reduced Hox expression was specific to loss of MLL-ENL expression and was not a consequence of differentiation. Affymetrix microarray analysis revealed that MLL-ENL may maintain or activate the expression of the transcription factor Sdccag33 and the serine / threonine kinase Pim-2, which confers protection from apoptosis. The analysis also demonstrated that MLL-ENL may repress the expression of apoptosis promoting genes such as Ddb2 and Aatk. In summary, MLL-ENL is required to initiate and maintain the immortalisation of myeloid progenitors and may contribute to immortalisation by aberrantly maintaining the expression of multiple Hoxa genes. The pathways regulated by multiple Hoxa genes and the MLL-ENL target genes identified by Affymetrix analysis represent new possibilities for therapies which may combat these aggressive leukaemias.

Since the first reports of cytogenetic abnormalities in ALL, repeated small and large scale studies have shown that cytogenetic abnormalities detected at presentation continue to be the most important predictor of outcome in ALL, at presentation, on relapse and even in the context of more intensive treatment modalities. The accuracy of diagnosis is much improved by the additional application of molecular techniques including FISH, which is the technique I used to more accurately define the cytogenetic abnormalities in a series of 176 adult patients with ALL. I have studied the incidence of and clinical features associated with three important structural abnormalities, namely BCR/ABL, MLL and ETV6/AML1 abnormalities, as well as alterations in chromosome number and showed that FISH with commercial probes provides a simple and accurate method for the detection of these abnormalities in presentation samples. FISH detected cytogenetically silent abnormalities in twenty-four of the 176 patients. In addition to the patients positive for rearranged genes, I uncovered a rare case of Philadelphia negative BCR/ABL ALL with aberrant insertion ofABL into a morphologically normal chromosome 9q. I went on to study in detail twenty-nine cases with indeterminate cytogenetics, and was able to further characterize the karyotype in fifteen of these. I have shown the applicability of sequential FISH a technique developed in our laboratory, in identifying underlying complex chromosomal abnormalities, even when metaphase number and morphology is very poor. In a number of cases the additional findings appear to influence clinical outcome. Finally by assessing the total cytogenetic information, as well as a number of clinical variables at presentation I developed a simple method of risk stratification for adult patients with ALL which distinguishes three subgroups, including a population accounting for 32% of 159 patients studied who appear to have significantly better outcomes when treated according to the UK ALL XII protocol.

The BCR-ABL negative myeloproliferative disorders (MPDs) are poorly understood at both the clinical and molecular level. Translocations involving chromosome bands 5q33 and 8p11, which cause rearrangements of the PDGFRB and FGFR1 genes respectively, are found in a minority of these patients. In this study, the disease that is characteristic of leukaemia with a rearrangement of 5q31-q33 is defined by a literature review of patient reports. The most common translocation in this group is the t(5;12)(q33;p13) which fuses ETV6 to PDGFRB. We have found this disease has both myeloproliferative and myelodysplastic features, commonly eosinophilia, monocytosis and splenomegaly. The probability of transformation to acute myeloid leukaemia (AML) is approximately 20%, therefore this myelodysplastic syndrome/myeloproliferative disorder (MDS/MPD) is less aggressive than BCR-ABL positive chronic myeloid leukaemia (CML). Using a two colour FISH assay with probes flanking the PDGFRB gene, 5 patients with a translocation involving 5q31-q33 and the MDS/MPD phenotype were identified as having a rearrangement in this gene. This is an important abnormality to identify, as PDGFRB fusions are sensitive to the tyrosine kinase inhibitor imatinib. Four patients with a disease very similar to CML were identified as BCR-ABL negative but had a BCR rearrangement. The t(8;22)(p11;q11) and t(4;22)(q12;q11) translocations were cloned from these cases, and found to result in novel in-frame fusions of BCR to FGFR1 and BCR to PDGFRA respectively. During this study a novel fusion FIP1L1-PDGFRA was described by Cools et al., (2003) in patients with HES. Screening of hypereosinophilic syndrome (HES) and atypical CML patients demonstrated that the FIP1L1-PDGFRA fusion was present in 3/34 patients. Furthermore, screening a cohort of 40 MPD patients, 8 hypereosinophilic syndrome and 10 systemic mastocytosis patients by dHPLC demonstrated that point mutations in KIT, PDGFRA or PDGFRB are uncommon in the MDS/MPD group of disorders. Defining the molecular basis of these diseases is imperative for treatment of patients with the new generation of specific tyrosine kinase inhibitors.

Leukaemia inhibitory factor (LIF) is a cytokine that is active on a wide variety of cells. Multiple LIF transcripts have been described. The transcripts LIF-D and LIF-M encode different signal peptides, which in mouse have been associated with differential localisation of the mature protein. LIF-D is associated with a freely diffusible protein, whereas the LIF-M is associated with the extracellular matrix. The polarity of LIF secretion has yet to be described and could illuminate the mechanisms of LIF localisation. Here the polarised endogenous secretion of human LIF and IL-6 in Caco-2 cells was characterised under normal culture conditions and following induction with IL-1b. Whether the apical or basolateral membrane was stimulated influenced the pattern of secretion (LIF: Unstimulated, 59% basolateral. Dual stimulation, 68% basolateral. Basolateral stimulation, 79% basolateral. Apical stimulation, 53% basolateral). IL-6 displayed a similar dependence on the site of stimulation but was predominantly secreted at the membrane that was stimulated. To determine the effect of the alternate signal peptides on the polarity of LIF secretion, LIF was epitope tagged with FLAG. Epitope-tagging with FLAG was used to separate endogenous from exogenous protein expression. However, despite the normal biological activity of LIF-FLAG and detection of the FLAG in a western blot, detection of the LIF-FLAG under non-reducing conditions was not observed, and therefore it was unsuitable for secretion studies. Untagged LIF was expressed exogenously in Madin-Darby canine kidney (MDCK) cells under the control of a tetracycline response promoter that allowed a variety of LIF expression levels to be tested. Exogenous murine LIF was secreted predominantly from the apical (60%) membrane of MDCK cells irrespective of the signal peptide expressed.

The nucleus contains a variety of morphologically distinct substructures called nuclear bodies, which include the promyelocytic leukemia oncogenic domains (PODs) also known as PML-NDIO. PODs are macromolecular multiprotein complexes that are present in all cultured cell lines as well as in vivo. The major component of PODs is the PML protein, which was originally identified as the fusion partner of retinoic acid receptor alpha (RARa) in the chromosomal translocation t(15;17) in patients with acute promyelocytic leukaemia (APL) (Kakizuka et al, 1991; Lavau et al, 1991; Goddard et al, 1992). The minor capsid proteins L2 of BPV-1, HPV-11 and HPV-33 have been shown to localise to PODs in the absence of other viral components (Day et al., 1998) and coexpression of BPV-1 12 with BPV-1 E2TA recruits E2 to PODs (Lambert et al, 2000). The presence of L2 in PODs also appears to be associated with the recruitment of the major capsid protein LI, the association of PODs with E2 is dependent on L2 but is independent of LI. The effect of HPV-16 L2 on the functions of HPV-16 E2 and the implications of this interaction to the virus life cycle are discussed. This study showed that HPV-16 L2 has a selective effect on the functions of HPV16 E2. L2 was able to down regulate the transcription transactivation function of E2 in HaCaT, U20S and C33a cells. No effect of L2 on E2 mediated DNA replication was observed. L2 was also able to reduce the level of E2 expression in HaCaT and U20S cells but not in C33a cells. The effect of L2 on E2 expression in HaCaT cells was further investigated by examining E2 mRNA levels and protein half-life. No difference in E2 mRNA or protein half-life was detected in the presence of L2. A series of L2 amino and carboxyl terminal deletion mutants were constructed as GST fusion proteins and GST binding assays were performed which showed that the amino terminus ofL2, even just the first 50 amino acids, was capable of binding with E2. GFP fusion forms of each L2 deletion mutant were also constructed and cellular localisation detected by immunofluorescence. GFP-L2 and all C terminal deletion mutants localised and were retained in the nucleus while the N-terminal deletion mutants localised to both the nucleus and the cytoplasm. Investigation of the effect of L2 deletion mutants on the transcription transactivation function of E2 showed that mutants expressing 1-200 and 150-473 amino acids of L2 do not down regulate function in HaCaT and C33a cells. In HaCaT cells, mutants expressing amino acids 1-50 and 1-100 also did not inhibit E2 function indicating that binding to E2 did not correlate with down regulation of transcription transactivation. Furthermore, only full-length L2 was able to reduce the level of E2 expression.