Voltage-gated sodium channels are crucial determinants of neuronal excitability. Human genetic mutations show that Nav1.7 is critically important for normal pain processing. Here I show that Nav1.7 expression in different subsets of mouse sensory and sympathetic neurons underlies distinct types of pain sensation. Deleting Nav1.7 in all sensory neurons abolishes mechanical pain, inflammatory pain and reflex withdrawal responses to noxious heat. In contrast, heat-evoked pain is retained when Nav1.7 is deleted specifically in Nav1.8-positive nociceptors. Surprisingly, responses to the hotplate test were unaffected by deleting Nav1.7 in all sensory neurons. However, deleting Nav1.7 in both sensory and sympathetic neurons abolishes these pain sensations and recapitulates the pain-free phenotype seen in humans with Nav1.7 loss-of-function mutations. Previously, Nav1.3, 1.7, 1.8 & 1.9 have all been implicated in neuropathic pain through a plethora of different experimental approaches. Here I show that Nav1.3, 1.7, 1.8 & 1.9 have modality and pathology specific roles. Deleting Nav1.3 or 1.7 in all sensory neurons inhibits both cold and mechanical allodynia specifically following chronic constriction injury. While deleting Nav1.8 or 1.9 only inhibits cold but not mechanical allodynia. Importantly, deleting Nav1.7 in adult mice using AdvillCreERT2 reverses CCI mediated pain, thereby further validating Nav1.7 as a viable analgesic drug target. However, deleting the expression of any of these VGSCs in sensory neurons alone does not attenuate L5 spinal nerve transection induced pain. Instead I show that deleting Nav1.7 expression in sympathetic neurons inhibits SNT pain, as well as sympathetic sprouting into the DRG. Surprisingly, pain induced by the chemotherapeutic agent oxaliplatin does not require the presence of Nav1.7, or the ‘classical’ Nav1.8-positive nociceptors. This demonstrates that identical behavioural outcomes in different neuropathic pain models involve distinct molecular and physiological mechanisms. Rational analgesic drug therapy therefore requires patient stratification in terms of mechanisms, not just phenotype.

The Tumour necrosis factor-related apoptosis-inducing ligand (TRAIL) is one of the most recently identified death inducing ligands of the TNF cytokine family. TRAIL induces apoptosis in most cancer cells, whereas the majority of normal cells are resistant. TRAIL receptor agonists are therefore considered to be a promising anti-cancer therapeutic. However, because many cancer cells develop resistance to TRAIL, understanding the mechanisms by which resistance is acquired will be critical for the therapeutic use of TRAIL in cancer therapy. We have discovered that Activated Cdc42-associated kinase 1 (Ack1) is required for TRAIL induced apoptosis in human epithelial cells. Ack1 is a non- receptor tyrosine kinase with numerous protein-protein interaction domains, suggested to have a role in several cellular processes such as trafficking, endocytosis and cell motility. Knockdown of Ack1 in various epithelial cell lines leads to significantly impaired TRAIL induced apoptosis as evident by reduced cleavage of Caspase-8 and -3 and surface exposure of phosphatidylserine. Exploring the underlying mechanism we found that Ack1 knockdown leads to impaired TRAIL induced clustering of TRAIL-R1 and a reduction in the recruitment of Caspase-8 to the DISC complex, essential for death inducing signal transduction. Translocation of the TRAIL receptors to lipid rafts in the plasma membrane have been suggested to be crucial for TRAIL receptor dynamics and downstream signalling following TRAIL ligand binding. In this work we show that Ack1 is required for the translocation of the TRAIL receptors to the lipid rafts. In this thesis, a novel regulatory role of Ack1 in apoptosis, death receptor signalling and lipid raft trafficking is presented, contributing further to the understanding of the molecular regulation of TRAIL receptor signalling.

Myogenesis is highly regulated and its activation in the embryo is controlled by a series of complex transcriptional regulatory networks that ultimately result in the expression of myogenic regulatory factors (MRFs). The MRFS, particularly MyoD and Myf5, are responsible, in concert with a vast range of cofactors, for directing the expression of genes responsible for muscle formation and activity. Several candidate proteins have emerged as being responsible for MRF expression, as well as numerous downstream effectors involved in muscle formation in vivo. Several cis-regulatory elements have been identified for MyoD, but only a handful of factors have been identified that bind these elements. In addition, knockout experiments of these regions do not result in a complete loss of MyoD expression, suggesting a certain level of redundancy and the existence of other yet unidentified cis-regulatory modules. In this study, novel potential regulatory regions within the MyoD upstream genomic locus were identified by comparative genomics. These regions, named ReMos 9, 10 and 11, were conserved in mammals, chick and fish. Reporter assays in C2C12 cells using these regions cloned upstream of the MyoD promoter revealed that they positively enhanced the promoter activity. A synergy was uncovered between ReMo 9 and 10, which have a strong positive effect on promoter activity, but none individually; ReMo 11 seemed to disrupt this synergy. In addition, ReMos 9+10 and the CER enhancer were shown, by double fluorescent RNA in situ hybridisations, to be transcribed and possess cryptic promoter activity. This suggested that these elements acted as alternative promoters and encoded RNAs that regulated MyoD gene expression. Furthermore, the use of a newly engineered database generated predictions of DNA-binding factors interacting with the cis-regulatory regions, as well as protein interaction networks involved in MyoD regulation. These predictions were refined and constrained with biological input data derived by microarrays of single-cells transiently expressing relevant constructs. A list of candidate muscle-specific binding factors was then tested in vitro by siRNA knockdown experiments, and showed that MyoD disrupts the positive synergistic effect of ReMos 9 and 10 on the PRR. In conclusion, this study identified a number of regions that seem to be involved in MyoD regulation, and candidate factors binding to the MyoD cis-regulatory regions. Further in vivo validation will identify their function in MyoD spatio-temporal gene expression.

Kaposi’s sarcoma-associated herpesvirus (KSHV) is causally related to Kaposi’s sarcoma (KS), the most common malignancy in individuals with untreated HIV/AIDS. Several lines of evidence indicate that KS oncogenesis is associated with loss of T cell-mediated control of KSHV-infected cells. However, the adaptive CD8 and CD4 T-cell responses against KSHV have not been fully characterised. Neither the antigenic repertoire nor the immunodominant targets of CD8 and CD4 KSHV-specific T cells are fully understood, and the phenotypes and functions of these cells remain largely unexplored. To investigate the targets of the CD8 and CD4 T-cell responses against KSHV, a novel approach for a large-scale screen of KSHV antigens was proposed that used lentiviral-transduced monocyte-derived dendritic cells (moDCs) expressing a panel of KSHV open reading frames (ORFs). Transduced moDCs naturally process the KSHV gene products and present the resulting antigenic peptides in the context of MHC class I and II. Transduced moDCs were cultured with autologous T cells and the CD8 and CD4 proliferative responses to each KSHV ORF (or pool of ORFs) were assessed. CD8 and CD4 KSHV-specific responses were investigated in 14 KSHV-seropositive individuals. Unexpectedly, both the CD8 and CD4 T-cell responses against KSHV were found to be skewed towards ORFs expressed in the early and late phases of the viral lytic cycle. The most frequently recognised CD8 target was a pool of late lytic KSHV ORFs, [ORF28/ORF36/ORF37]. Identification of novel KSHV CD8 epitopes from within the late lytic ORF pool was attempted. Peptide-MHC binding and denaturation assays identified peptides that had the highest affinity for HLA-A*0201. Recognition of these potential epitopes was tested in clinical samples by IFNγ ELISpot, and compared with recognition of nine previously published HLA-A*0201-restricted KSHV epitopes. Finally, the use of pentamers as tools to investigate the memory phenotypes and functions of virus-specific T cells was explored.

At physiological oxygen concentrations ([O2]) hypoxia-inducible factor-1α (HIF-1α) is constantly hydroxylated and thus prepared for proteosomal degradation through the action of the prolyl hydroxylases (PHDs) (Jiang et al., 1996). In hypoxia, however, the oxygen-sensitive PHDs are inhibited and HIF-1α is stabilised. Other agents, including cytokines and growth factors have been shown to stabilise HIF-1α at physiological [O2] through different mechanisms such as activation of the phosphatidylinositol 3-kinase (PI3K) or mitogen-activated protein kinase pathways (Semenza, 2003). Increased production of reactive oxygen species (ROS) during hypoxia have also been claimed to stabilise HIF-1α (Chandel et al., 1998) and we have now investigated the effect of endogenous ROS on HIF-1α stabilisation. HIF-1α stabilisation and ROS production in human embryonic kidney (HEK 293T) cells were determined by immunoblotting and the use of fluorescent probes, respectively. γGlutamyl cysteine synthetase (γGCS) is the rate limiting enzyme of glutathione (GSH) biosynthesis and therefore a crucial antioxidant. We used small interfering RNA (siRNA) to silence this enzyme and thus impair the capacity of the cells to detoxify ROS. In order to determine whether mitochondria are a major source of ROS we used cells depleted of mitochondrial DNA (Rho0); these were characterised in vitro by monitoring oxygen consumption. RT-PCR was used to determine mitochondrial DNA content and immunoblotting to assess mitochondrial-encoded protein expression. The effects of a Rho0 phenotype were then assessed in relation to HIF-1α stabilisation and ROS production. HIF-1α is stabilised in an oxygen-dependent manner. HIF-1α stabilisation at low [O2] (3%), but not at 0.5% O2 is prevented by treatment with antioxidants. Silencing γGCS augmented free radical production in HEK 293T cells. This was associated with HIF-1α stabilisation at ambient [O2] (21%) and could be prevented by treatment with antioxidants. Rho0 cells produced less ROS than wild-type cells and did not stabilise HIF-1α either at low [O2] (3%) in wild-type cells or at 21% O2 in γGCS silenced cells. The data suggest that HIF-1α can be stabilised by ROS generated by the mitochondria.

c-Met (the Hepatocyte Growth Factor Receptor) is a receptor tyrosine kinase, implicated in various human tumours. Ligand-induced c-Met activation and internalisation was described to play a role in Extracellular signal Regulated Kinase (ERK) phosphorylation and plasma membrane accumulation, with consequent effects on migratory responses. Initial work in this thesis illustrated that this accumulation was prevented using pharmacological agents to disrupt the microtubule network, and more specifically the microtubule motor proteins. To find specific players in this process, RNA interference was used as a tool to silence gene expression, and a high throughput HGF-mediated wound-healing assay was used as a readout. An siRNA library encoding more than 1,400 human genes targeting kinases, phosphatases, motor proteins and genes predicted to influence cell motility, was screened in a lung carcinoma cell line, A549. From a collection of hits identified in the primary screen, 100 were chosen for further validation. These hits fell into three categories: those that had a general effect on motility (but not having any effect on HGF-inducibility), those that markedly reduced the HGF-induced motility and those that increased the effect of HGF on motility. Twenty validated hits were selected for more extensive analysis. Four additional oligos were tested individually and as pools to add to the selection process. The top twelve highly validated hits were selected for more thorough investigation, including effects on cell speed, c-Met expression downregulation, localisation, and signalling. In addition, connection between hits and ERK subcellular localisation was also investigated. The screening approach was validated by the finding that two known players of the HGF-provoked response were identified, namely c-Met and ERK2. Although additional hits do not define the machinery needed to move ERK within cells, they do provide an insight into processes behind HGF-induced cell migration.

Transient receptor potential canonical (TRPC) channels are a family of non-selective cation channels which have a wide range of physiological functions. In this thesis we used genetic approaches to investigate functional roles of TRPC3, TRPC6 We used the Cre/loxP system to delete TRPC3 exclusively from nociceptive neurons, using Nav1.8Cre, and from all dorsal root ganglion (DRG) neurons, using Advillin Cre. Inflammatory pain behavior was attenuated in both TRPC3 conditional null animals but acute mechanical, cold and thermal pain behavior was not affected. Microarray analysis identified a number of dysregulated pathways in the nociceptors of the TRPC3 conditional null mice, providing insights into inflammatory pain pathways. Given the similarity between TRPC channels we used a TRPC3/TRPC6 global double knockout to further investigate TRPC function. The TRPC3/TRPC6 double knockout mice were found to have attenuated mechanosensation specific to light touch, with no deficits in noxious mechanosensation, due to a reduction of rapidly adapting mechanically activated current in the small-medium DRG neurons. The TRPC3/TRPC6 double knockouts had no deficits in acute thermal or cold pain, or in neuropathic pain. The TRPC3/TRPC6 double knockouts were also behaviourally deaf, confirmed by auditory brainstem recordings. This work indicates a role for TRPC3 and TRPC6 in mechanosensation in sensory neurons and outer hair cells.

Voltage-gated sodium (Nav) channels are therapeutic targets for several disorders affecting humans, including epilepsy, neurodegeneration and neuropathic pain. Typically, drugs treating these conditions exert a use- and voltage-dependent inhibition of Na currents, an action attributed to the stabilisation of the slow inactivated state which is formed during prolonged depolarisation. The binding site has been suggested to reside in the channel pore at a site only accessible from the intracellular environment. What gives different chemicals having this action in common selectivity for certain disorders (e.g. neuroprotection versus epilepsy) remains a mystery. Several channel subtypes exist, with types Nav1.2 and Nav1.6 being major isoforms found in the brain, raising the possibility that different subtypes exhibit differential drug sensitivity. To investigate this issue and explore the site of drug action further, the action of several Nav channel modulators, including lidocaine (a local anaesthetic) and sipatrigine (a neuroprotective agent), were studied on Na currents generated by Nav1.2a and Nav1.6 channels stably expressed in cell lines, and by acutely dissociated native cells, using electrophysiological techniques. The findings indicated that different drugs have some selectivity for particular channel subtypes. In addition, to study the slow inactivated state more selectively, fast inactivation was inhibited chemically. Inhibition of this mechanism altered both normal channel function and drug action. Drug effect during external and internal application to cells was also compared. Seemingly contrary to the current hypothesis of an internal site of action, drugs were more potent following external application, suggesting that their site of action may be different from the putative intracellularly-accessible “local anaesthetic receptor”. These experimental results were tested using a mathematical simulation of drug diffusion during whole-cell voltage-clamped electrophysiological recordings, to determine whether variations in the physicochemical properties of these drugs could explain their different potencies on internal and external application.

No description available.

No description available.