Tumour-associated macrophages (TAM) are a major component of the inflammatory infiltrate that typifies most malignancies. Among them, Burkitt’s lymphoma (BL), a high-grade non Hodgkin lymphoma (NHL) of B-cell origin, represents a characteristic example. Studies from our group have shown that TAM in BL exert pivotal roles that are mainly supportive of tumourigenesis such as maintaining an immunosuppressive microenvironment. In order to unravel the molecular mechanisms underlying TAM functions in BL, solid tumours from a mouse xenograft model of BL have been used to obtain TAM and assess their activation status in vivo. Laser-capture microdissection has been successfully used to procure intact macrophage sections from the tumour site, allowing the production of a pure, in situ gene expression signature of TAM in BL. Tingible-body Mφ from lymph node germinal-centres and resident tissue Mφ from resting lymph nodes of non-tumour bearing mice were chosen for direct comparison with TAM. Whole-genome microarray technology has revealed a distinct TAM gene expression profile, with 454 genes being significantly up-regulated (fc ≥ 2, p<0.05) and 1293 genes being significantly down-regulated (fc ≤ -2, p<0.05) between TAM and either of the two normal Mφ populations. Further bioinformatics analysis of gene functions has highlighted matrix remodeling, phagocytosis, and immune response among the processes most highly enriched in TAM. Importantly, mRNA and tissue expression of selected differentially expressed genes relevant to these processes was validated by real-time qPCR and immunofluorescence labeling respectively. Following the generation of the TAM profile in situ, in vitro experimental approaches were undertaken in order to investigate how specific elements of the BL microenvironment drive the observed TAM signatures. Specifically, the direct role of apoptotic tumour cells, a key component of the BL microenvironment, versus that of viable tumour cells in driving TAM matrix remodelling gene expression was assessed in short-term mouse and human Mφ-NHL cell co-cultures. From the aforementioned cluster, emphasis was given to MMP12 and MMP2 transcripts: mRNA and protein expression of these MMPs was found to be up-regulated in Mφ following viable tumour cell co-cultures and this effect was further enhanced following apoptotic tumour cell co-cultures, implying that apoptotic NHL cells could directly shape TAM matrix remodeling phenotype in BL in vivo. Whereas the mRNA of both MMPs was solely Mφ-derived in this system, MMP12 and MMP2 protein was surprisingly found also to be increased in NHL cells in the apparent absence of increased mRNA. Detailed examination of MMP12 production by NHL cells revealed that it is most likely an apoptosis-dependent process, since apoptotic NHL cells generated through different apoptosis stimuli, as well as apoptotic cell-derived microparticles, showed markedly increased MMP12 protein levels. In conclusion, the data presented in this thesis, provide the first insight into the in vivo activation status of TAM in high-grade NHL, through generation of the TAM gene signature in situ. Upon further in vitro studies, apoptotic NHL cells were shown to directly modulate the matrix remodelling component of the TAM signature as well as to actively produce matrix remodelling mediators themselves, suggesting distinct roles for tumour cell apoptosis within the NHL microenvironment that can profoundly influence the disease outcome.

In high-grade ‘starry-sky’ non-Hodgkin’s lymphoma (NHL), particularly Burkitt’s lymphoma (BL), large numbers of apoptotic tumour cells are engulfed by infiltrating tumour-associated macrophages (TAM). In situ studies suggest that in starry-sky TAM in a xenograft model of BL various tumour-promoting, trophic, angiogenic, tissue remodelling, and anti-inflammatory pathways are activated. Furthermore, apoptotic cells have been shown to activate expression of tumour-promoting matrix metalloproteinases in macrophages. This work investigates the hypothesis that apoptotic cells or factors released from apoptotic cells induce additional aspects of the starry-sky TAM signature, which serve to promote tumour growth. Macrophages at different stages of maturation, cultured in vitro in the presence of large numbers of apoptotic cells, were shown to differ in phenotype, giving credibility to the hypothesis. Less mature mouse bone marrow-derived macrophages (BMDM) were better at migrating towards apoptotic cells, whereas mature BMDM were better at phagocytosing them. Lactoferrin, which is released from cells undergoing apoptosis and inhibits the migration of neutrophils, was selected as a candidate mediator in the activation of macrophages by apoptotic cells. Lactoferrin was shown to bind viable human and murine monocytes and macrophages, however only high concentrations, which are unlikely to be physiologically or clinically relevant, were found to affect expression of starry-sky TAM genes or reduce classically-activated macrophage cytotoxicity. The direct effect of apoptotic cells on macrophage activation was assessed. Mature BMDM were not used for these studies as their development in vitro in a highly apoptotic environment was judged likely to bias their activation state toward that of TAM, therefore macrophages were first classically-activated with IFN-γ and LPS. This reduced the expression of many starry-sky TAM genes, including several genes associated with responses to apoptotic cells. However, classical activation did not inhibit apoptotic cell engulfment, but rather enhanced it. Co-culture with apoptotic cells, but not viable cells, increased the gene expression of Gas6, Mrc1, Cd36, Timp2, and Pparg, and the latter was dependent on direct interaction with macrophages rather than factors released from apoptotic cells. Furthermore, classically-activated macrophages were found to induce apoptosis in lymphoma cells, and although pre-co-culture of the macrophages with apoptotic cells did not reduce their ability to induce apoptosis, it enhanced tumour cell growth. Macrophage deficiency of IL-4Rα or galectin-3 did not affect classically-activated macrophage responses to apoptotic cells. However, classical activation of galectin-3 deficient macrophages appeared to restore the decreased ability of galectin-3 deficient, untreated macrophages to phagocytose apoptotic cells. Because of the unique new method of laser-capture microdissection by which starry-sky TAM signatures were established, direct comparisons with expression databases of tissue and in vitro cultured macrophages were not possible, but indirect comparisons suggest starry-sky TAM activation reflects the activation phenotype of a mixture of tissue macrophages. Furthermore, it highlighted phagocytosis as one of the most important pathways activated by starry-sky TAM. Together the results presented here suggest apoptotic lymphoma cells can shape TAM activation signatures in starry-sky NHL, even when macrophages are pre-activated to induce apoptosis in lymphoma cells. This is important when considering the consequences of anti-cancer therapies that induce apoptosis or aim to redirect phagocyte activation.

Watson-Crick paired B-form DNA is the genetic material in most of the biological systems. Integrity of DNA is of utmost importance for the normal functioning of any organism. Various environmental factors, chemicals and endogenous agents constantly challenge integrity of the genome resulting in mutagenesis. Over the past few decades multiple reports suggest that DNA can adopt alternative conformations other than the right handed double helix. Such structures occur within the context of B-DNA as sequence dependent structural variations and are facilitated by free energy derived from negative supercoiling, which may be generated during physiological processes like transcription, replication, etc. or binding of proteins. Multiple groups have shown that these structures render fragility to the genome owing to single-strandedness (presence of unpaired bases). This conformational polymorphism of the DNA is due to the presence of several repetitive elements across the genome. Some of the common non-B DNA structures include Z-DNA, H-DNA (triplex DNA), cruciform DNA, G-quadruplexes and RNA: DNA hybrid (R-loops). Over the past few decades G-quadruplex structures have gained tremendous importance owing to its role in physiology and pathology. Recently it has been shown that novel sequence motifs, called GNG or bulges can fold into G-quadruplexes, thus increasing the propensity of such structures genome-wide. Neurological diseases, psychiatric diseases and genomic disorders (due to deletions, translocations, duplications and inversions) are some of the consequences of non-B DNA structures in the human genome. Inadvertent genomic rearrangements in human can lead to different diseases including cancer. Immediate consequence of genomic rearrangement includes structural alteration of genome through joining of distant sequences. t(8;14) translocation is the hallmark of Burkitt’s lymphoma, which results in deregulation of c-MYC gene that may contribute to oncogenic transformation. In the present study, we delineate the causes of fragility within the c-MYC gene. In order to do this, breakpoints at the c-MYC locus from Burkitt’s lymphoma patient sequences reported in database were plotted and analysed. Interestingly, unlike many other translocations, breakpoints at c-MYC locus were widespread, except for a cluster of breakpoints downstream to promoter 2 (P2). Previous studies indicate that the translocation breakpoint clusters often correlate with formation of non-B DNA structures. The entire breakpoint cluster downstream of P2 was divided into Region 1, Region 2 and Region 3. Interestingly, in silico analysis of the breakpoint clusters revealed no evidence for predictive classic non-B DNA motifs in Region 2; whereas Region 1 harboured a G-quadruplex motif on the template strand and Region 3 had two short inverted repeats. Intriguingly, as the nontemplate strand of Region 2 was G skewed with a good number of AID binding motifs, we tested the MYC breakpoint Region 2 for its potential to form R-loop due to binding of nascent RNA to template DNA. Our results showed that MYC Region 2 can form RNA-DNA hybrid in a transcription dependent manner in physiological orientation. Observed structure was sensitive to RNase H. We showed Region 2 hindered action of Dpn I upon transcription confirming formation of R-loop structure. Owing to single strandedness, Region 2 R-loop was shown to be sensitive to P1 nuclease as opposed to the untranscribed control. The single strandedness of the Region 2 R-loop was characterized at a single molecule level through bisulfite modification assay. The assay corroborated formation of R-loop along with providing snapshots of various length R-loops formed upon Region 2 transcription. Besides, various biophysical and biochemical assays showed the complementary region (template strand) to be single-stranded in stretches, upon transcription. Length of RNA within the R-loop was within a range of 75 to 250 nt. To delineate the mechanism of R-loop formation we tested the sensitivity of R-loop formation to RNase A during and post transcription; and found that R-loop formation was abrogated in presence of RNase A during transcription suggesting that R-loop formation followed a “thread back model”. Intriguingly we observed that two short regions of the template strand exhibited high degree of single strandedness. To investigate the reason for such unusual single strandedness, oligonucleotides spanning the region was designed and subjected for CD and EMSA studies. EMSA showed robust intramolecular G-quadruplex structure formation in presence of KCl, whereas CD confirmed that both regions formed parallel G-quadruplexes. We also showed the precise involvement of guanines in structure formation through DMS protection assay. Further, the region of interest was cloned into appropriate vectors and primer extension assays were performed in presence of G-quadruplex stabilizing agents like TMPyP4 and KCl. Increasing concentration of these stabilizing agents enhanced the formation of G-quadruplexes in a double stranded context, which hindered polymerase progression. Since these G-quadruplex structures utilized sequences which are deviant to the consensus of G-quadruplex motifs, non-B DNA predicting tools were unable to score them. On closer analysis of the sequences we found that, these G-quadruplexes involve duplex hairpin and GNG motifs during structure formation. Besides, both the G-quadruplexes were highly thermostable and were able to fold back upon renaturation. Till recently, it has been believed that G-quadruplex structures are formed using a minimum of four, 3 guanine tracts, with connecting loops ranging from one to seven. Recent studies have reported deviation from this general convention. One such deviation is the involvement of bulges in the guanine tracts. In the present study, guanines along with GNG motifs have been extensively studied using recently reported HOX11 breakpoint fragile region I as a model template. By strategic mutagenesis approach we show that the core elements of a G-quadruplex are not equally important in structure formation when flanked by GNG motifs. Importantly, the positioning and number of GNG/GNGNG can dictate the formation of G-quadruplexes. In addition to HOX11 fragile region, GNG motifs of HIF1-alpha can fold into intramolecular G-quartet. However, GNG motifs in mutant VEGF sequence could not participate in structure formation, suggesting that the usage of GNG is context dependent. Importantly, we show that when two stretches of guanines are flanked by two independent GNG motifs in a naturally occurring sequence (SHOX), it can fold into an intramolecular G-quadruplex. Interestingly, intra molecular GNG G-quadruplexes were able to fold back after complete denaturation of the oligonucleotides. Besides one of the intra molecular GNG G-quadruplexes was purified and confirmed for parallel conformation. Finally, we show the specific binding of G-quadruplex binding protein, Nucleolin and G-quadruplex antibody BG4 to SHOX G-quadruplex through EMSA studies. Thus, the study provides novel insights into the role of GNG motifs in G-quadruplex structure formation, which may have both physiological and pathological implications. In conclusion, we show formation of transcription dependent R-loop and G-quadruplex structures at the c-MYC gene locus in a mutually exclusive manner. The data presented here, in conjunction with studies from other laboratories suggests that these structures could impart fragility within the c-MYC gene locus during t(8;14) translocation. Besides, we characterised unusual G-quadruplexes harbouring GNG motifs. We find that positioning and number of GNG can dictate the formation of G-quadruplexes and is context dependent.