No description available.

No description available.

No description available.

Adverse drug reactions represent a major challenge to clinicians, healthcare systems, pharmaceutical companies and academia. With carboxylic acid drugs accounting for the most common class of drugs withdrawn from the market, the carboxylate pharmacophore has received much attention as a potential toxicophore. Direct glucuronidation of the carboxylate group, producing chemically unstable and protein reactive acyl glucuronide (AG) metabolites has received much attention as a bioactivation pathway responsible for generation of these off-target hypersensitivity and hepatotoxicity. It is the chemical instability and protein reactivity of AG metabolites that has led to their hypothesised ability to covalently modify proteins in vivo and subsequently stimulate inappropriate immune responses in susceptible patients. Despite this, whilst the reactivity of AGs has been shown in vitro, their reactivity has never been confirmed in any in vivo system, meaning their association with toxicity may be unjustified. The focus of this thesis was to investigate whether acyl glucuronides could identify covalent adducts to protein in vivo. To address this aim, the thesis first investigates the chemistry of interaction between acyl glucuronides and protein during in vitro investigation. 2mM 1-β diclofenac-AG was found to degrade spontaneously via acyl migration following incubation with 0.1M phosphate buffer pH 7.4 at 37°C with a degradation half-life of 0.78 hours, confirming diclofenac as amongst the most reactive AGs. Further incubations confirmed the action of human serum albumin (HSA) as a mild esterase, and the presence of plasma esterases acting to hydrolyse AGs. The covalent binding of diclofenac-AG to HSA was confirmed using both an alkaline hydrolysis as well as direct mass-spectrometric analyses of modified proteins. Covalent modification of lysine residues was specifically identified, and was found to be concentration and time dependent. Further in vitro incubation experiments revealed for the first time that the 1-β isomer of AGs is responsible for the formation of transacylation adducts, and confirmed previous suggestion that acyl migration is required for the extensive glycation of HSA. Following characterisation of the interaction of diclofenac-AG with HSA, investigations were undertaken in the rat to identify interactions of AGs with circulating rat serum albumin in vivo. In vitro incubations of diclofenac-AG revealed RSA contained fewer binding sites when compared to HSA. Further to this no covalent modification of RSA could be detected in vivo following intravenous administration of 60mg/kg diclofenac-AG. The rapid plasma clearance of diclofenac-AG (67.81 ± 12.83 ml min-1 kg-1) in the rat was shown to be 600 fold faster than that of diclofenac (12.00 ± 2.98 ml min-1 kg-1) following bolus intravenous administration. Use of a continuous intravenous infusion drug delivery system revealed an adaptive change in rats upon continuous infusion of diclofenac, resulting in enhanced plasma elimination of the drug, and induction of the ROS scavenging enzymes catalase and superoxide dismutase-2, without detection of hepatotoxicity. The final experiments in the thesis revealed for the first time the detection of glycation adducts to HSA extracted from volunteer patients receiving chronic diclofenac therapy. These were shown through the detection of glycation adducts in three out of six patients tested. Between 1 and 4 lysine residues were identified in patients, with modifications towards one or all of lysine residues 195, 199, 432 and 525. Transacylation adducts were detected towards lysine residues in all six patient samples analysed. Whilst identification of transacylation adducts reveals bioactivation of the carboxylic acid functional group, it is the identification of glycation adducts to albumin isolated from three of the six patients which reveals, for the first time, definitive evidence for AG reactivity in vivo. This reinforces concerns over the potential of AGs to act as haptens, and re-affirms the carboxylic acid structure as a site of bioactivation forming reactive metabolites.

Membrane transporters are increasingly recognised as being important in determining drug pharmacokinetics at whole body, organ, and cellular levels. At the blood-­‐brain barrier (BBB), membrane transporters determine the passage of drugs into and out of the brain. About 30 % of all patients are classed as non-­‐responders for both epilepsy and schizophrenia. Drug transporters from the adenosine 5'-­‐ triphosphate (ATP)-­‐binding cassette (ABC) transporter family or from the solute carrier (SLC) superfamily may contribute to these drug resistant phenotypes but most have received limited attention. Treatment response to carbamazepine (CBZ) has been associated with genetic polymorphisms in ABCC2, particularly -­‐24C>T, c.1249G>A, and c.3972C>T. However, the results have been conflicting and inconclusive amongst the different studies. A functional and clinical analysis was undertaken to investigate the impact of ABCC2 on CBZ treatment response. In vitro, no ABCC2-­‐mediated CBZ transport could be observed in efflux assays with an ABCC2-­‐transfected human fibrosarcoma cell line (Rht14-­‐10) and a dog kidney cell line (MDCKII). In addition, uptake into inside-­‐out vesicles derived from the Rht14-­‐10 cell line was negative. Clinical analysis of patients from the SANAD (Standard and New Antiepileptic Drugs) trial (assessing the clinical end-­‐points time to first seizure (n = 229) and time to 12-­‐month remission (n = 134)) did not show any significant associations between the three ABCC2 gene polymorphisms, -­‐24C>T, c.1249G>A, c.3972C>T, and clinical outcomes. In an attempt to identify currently unrecognised human drug transporters with potential relevance to epilepsy and schizophrenia, screening of transport of CBZ, lamotrigine (LTG), topiramate (TPM), levetiracetam, valproate, phenytoin, and clozapine (CLP) was undertaken using an immortalised human brain endothelial cell line (hCMEC/D3) as an in vitro model of the BBB. Accumulation of TPM was significantly enhanced by 44-­‐53 % in the presence of the typical ABCC efflux transporter inhibitors MK571 and montelukast. Furthermore, CLP uptake was significantly reduced by 94 % and 83 % in the presence of the typical organic cation transporter inhibitors prazosin and verapamil, respectively. CLP uptake into the hCMEC/D3 cell line followed classical Michaelis-­‐Menten kinetics with Vmax of 3288 (pmol/million cells)/min and Km of 35.93 μM. To identify the exact underlying transporters involved in TPM efflux and CLP uptake, both functional siRNA screening was undertaken and transport was investigated in transfected cell lines. None of the known functional ABCC transporters were shown to transport TPM. In addition, none of the expressed and functionally characterised organic cation transporters from the SLC22A family, as well as transporters from the SLC6A, SLC28A, and SLC29A families, had an effect on CLP accumulation. LTG has recently been identified as a substrate for SLC22A1 (OCT1). Interaction with the human immunodeficiency virus protease inhibitors lopinavir/ritonavir and the antipsychotic CLP was therefore investigated. At clinically relevant concentrations, lopinavir was found to significantly reduce SLC22A1-­‐mediated uptake of LTG by 39 %. In addition, CLP was a potent inhibitor of SLC22A1-­‐mediated LTG uptake yielding an IC50 of 1.8 μM. Similarly low IC50 values were obtained with primary human hepatocytes from two patients (IC50 = 7.9 μM and IC50 = 3.9 μM, respectively) and the hCMEC/D3 cell line (IC50 = 2.0 μM). The clinical consequences of these observations will require further in vivo pharmacokinetic and epidemiological research. In conclusion, the results presented in this thesis demonstrate that membrane transporters can be involved in the passage of AEDs and antipsychotics across the BBB and other membrane barriers. However, currently available in vitro methods proved to be insufficient to identify and characterise the underlying transporters involved and to further evaluate the impact on treatment efficacy. The development of large-­‐scale functional screening methodologies will be crucial for a more systematic and comprehensive understanding of drug transport processes involved in determining access of drugs to the central nervous system. This will help in improving drug efficacy and drug safety, allow prediction of drug-­‐drug interactions, and eventually allowed a more personalised approach to prescribing in diseases such as epilepsy and schizophrenia.

T-cell mediated hypersensitivity reactions (HSRs) to carbamazepine (CBZ), a commonly used anti-epileptic drug, occur only in a small proportion of patients, but can often be severe in nature. As the underlying pathomechanisms are not fully understood, it has proven difficult to predict who may be at risk of developing CBZ-induced HSRs. Recently, specific human leukocyte antigen (HLA) alleles have been identified as susceptibility factors for CBZ hypersensitivity in diverse populations, indicating that HLA molecules may be functionally involved in CBZ-induced T-cell activation. HLA-A*31:01 represents the latest example and has been implicated in CBZ-induced HSRs in Caucasian patients. Thus, the aim of this work was to explore the molecular interactions of CBZ with HLA-A*31:01 and drug-specific T-cells. The HLA restriction pattern of CBZ-reactive T-cells from a patient expressing HLA-A*31:01 was investigated. It was shown that CD8+ T-cells were activated in a HLA-A*31:01 dependent way. Further, HLA-DRB1*04:04 was found to be responsible for the stimulation of CD4+ T-cells, suggesting a common HLA haplotype may be involved in mediating T-cell responses to CBZ in Europeans. Next, in vitro priming of drug-naïve T-cells from HLA-A*31:01+ healthy volunteers against CBZ was attempted. Weak responses to CBZ could be detected in some but not all volunteers, indicating factors additional to HLA- A*31:01 are required to induce a primary stimulation of T-cells to CBZ. Besides, the removal of T-regulatory cells and use of dendritic cells as antigen- presenting cells seemed to generally improve priming conditions. In order to investigate whether CBZ affected the peptide-binding specificity of HLA-A*31:01, in silico and in vitro analysis were performed. In silico modelling provided a possible binding site for CBZ within the HLA peptide-binding cleft. A peptide elution study provided a preliminary indication that binding of CBZ to HLA-A*31:01 may alter the peptide repertoire presented by the allele, which could potentially result in T-cell activation. Most recently, it has been suggested that the T-cell receptor (TCR) may represent an additional predisposing factor for CBZ-induced HSRs. Accordingly, a protocol for the analysis of the TCR Vβ repertoire of drug-reactive T-cells by flow cytometry as well as CDR3 spectratyping was set up using T-cells from healthy donors primed against the model antigen SMX-NO. Both methods showed that antigen stimulation resulted in skewing of common TCR Vβ subtypes among the donors. Combined, the optimised methods will allow assessment of whether specific TCR clonotypes may be implicated in HLA- A*31:01 associated HSRs to CBZ. In summary, the data presented in this thesis provide initial evidence that CBZ is able to interact directly, through a non-covalent binding mechanism, with HLA-A*31:01 causing T-cell activation. However, it cannot be excluded that the stimulation of T-cells in vivo requires the formation of a hapten complex. Further work is needed to define other factors that are involved in predisposing an individual to CBZ hypersensitivity.

Malaria and malnutrition are causes of high morbidity and mortality in developing countries especially in sub-Saharan Africa and Asia. Malnourished children are at higher risk of developing malaria, a problem compounded by the fact that malnutrition affects the metabolism of different antimalarials including chloroquine and quinine. Emerging resistance to chloroquine, which was the drug of choice, has led to the widespread use of artemisinin combination therapy in the population including children. To date, no studies have been undertaken on the pharmacokinetics of artemisinin compounds in malnourished children. This thesis aims to fill this evidence gap by studying patients with uncomplicated malaria of different nutritional status in The Gambia. Analysis of the nutritional status of 97 children in The Gambia with uncomplicated malaria showed that 30% were both underweight and wasting, while 28% were categorised into stunting. This was much higher than the national average which has been estimated to be 17.4%, 9.5% and 23% for underweight, wasting and stunting, respectively, demonstrating a potential relationship between malaria and Protein energy malnutrition. In-vitro studies showed that although pre-treatment of HL-60 cells with the iron chelator (DFO) did not affect the bioactivation of artesunate, there was a 20% increase in cell viability with IC50 increasing from 7.0 ± 4.3 to 33.3 ± 2.9. This is believed to be as a result of DFO chelating the toxic iron generated as a result of artesunate bioactivation which increased from 0.32 ± 0.6 ng/mol in the control incubations to 0.84 ± 0.1ng/mol at 100 μmol artesunate concentration. In light of the fact that iron was important in the mechanism of action of these compounds, and the fact that iron deficiency is commonly in malnourished children, the effect of both PEM and anaemia on plasma drug levels of artemether and DHA was also studied. LC-MS/MS method was optimised and validated for the simultaneous analysis of artemether and DHA in plasma with ≥80% precision and accuracy. Plasma artemether and DHA concentration analysed 2h post first dose was 138.4 ± 80.9 ng/ml and 58.8 ± 43.7 ng/ml respectively. Severely wasted and wasted children had the highest artemether (156.5 ±69.6 ng/ml) and DHA (84.1 ±62.6) plasma concentrations respectively but values were not statistically significant. Anaemic status of children did not have an influence on drug plasma concentration with anaemic children having artemether and DHA plasma concentration of 138.5 ± 73.7 and 57.9 ± 36.6, respectively, and 138.43 ± 85.3 and 59.3 ± 46.8 for non-anaemic children respectively. However, conclusive results were limited by sample size. In conclusion, this thesis has demonstrated a relationship between malaria and Protein energy malnutrition, and highlighted the possible effects pathophysiological changes as a result of protein energy malnutrition can have on drug pharmacology and therapeutic effects in these children. There is a need for further studies in larger cohorts of children with protein energy malnutrition to determine whether therapeutic efficacy of artemisinin combination therapy is affected in an adverse manner, and whether there is a need for changes in dosing recommendations.

Antiepileptic drug (AED) treatment is the first line strategy for seizure control in the majority of individuals with epilepsy but remains challenging, not least because of interindividual variability in efficacy, tolerability and dosing. The studies presented in this thesis set out to explore that variability from a genomic perspective in patients with newly diagnosed epilepsy from across the UK. Single nucleotide polymorphisms (SNPs) in genes encoding drug metabolising enzymes (DMEs) may be associated with the dose of carbamazepine (CBZ) required for seizure control. A cohort of 159 individuals who were seizure-free for 12 months on a stable dose of CBZ monotherapy was genotyped for 51 SNPs across six DMEs. Haplotype analysis identified 8 haplotype blocks across the genes. No single SNPs or haplotype blocks were associated with CBZ dose. Thus, it is unlikely that genetic variability in DMEs accounts for the individual differences in CBZ dose requirement. A splice site SNP (rs3812718) in the SCN1A gene was previously shown to influence maximum doses of AEDs. This SNP was genotyped in 817 patients and tested for association with maximum and maintenance doses of several AEDs. An association was identified between rs3812718 and maximum AED dose, with an interaction analysis suggestive of a drug specific effect. These findings suggest that this SCN1A variant contributes to variability in the limit of tolerability to AEDs. Response to AED treatment is multifactorial and likely to be influenced by multiple genes. Five SNPs previously reported to predict treatment outcome in epilepsy were genotyped in 772 patients and the resulting data, together with data from an Australian cohort, incorporated into a predictive algorithm. The algorithm failed to predict treatment outcome in general but was partially successful in identifying responders to CBZ and valproate. These five SNPs may be relevant to the prognosis of epilepsy, particularly when treated with specific AEDs. Primary generalised epilepsies (PGEs) are highly heritable and believed to be polygenic in origin. Predictive algorithms were employed to explore genetic influences on seizure (absence vs. myoclonus) and epilepsy (PGE vs. focal) type using 1,840 SNP genotypes available from 436 patients with PGE. Although the algorithms failed to distinguish PGE patients on the basis of genetic variants, they showed improved association over univariate methods of analysis. Such an approach may be suitable for future investigations using large genomic datasets. A recent genome-wide association study identified multiple genetic variants that approached genome-wide significance for association with 12 month remission from seizures. Five of these SNPs were genotyped in an independent cohort of 424 patients and tested for association with remission and time to remission. No significant associations were found, questioning the validity of the original observation or the method of replication. Further work is required to understand this outcome. In conclusion, the genetic bases of epilepsy, AED response and AED dose requirement are multigenic and thus far undetectable using traditional association studies in modestly-sized patient cohorts. Further advances in genomic, bioinformatics and statistical methodologies are required before the genetic contribution to heterogeneity in epilepsy-related phenotypes can be translated into improved clinical care.

Deciphering the complex molecular circuitry of the brain is crucial for understanding how processes such as higher cognitive function and behaviour are disrupted in neurological disease. Thus it is imperative to explore further the regulatory mechanisms centred on key transcription factors that orchestrate such processes including REST/NRSF (restrictive element-1 silencing transcription factor/neuron restrictive silencing factor); NRSF targets over 2,000 human genes and plays a central and dynamic role in a myriad of CNS processes. To address the function of NRSF, I employed several research disciplines including bioinformatics, gene association studies for complex polygenic diseases and model systems for understanding the structure and function of several NRSF directed pathways in the CNS. My data demonstrated that disruption of the normal balance of NRSF within the cell may be a fundamental mechanism across a range of common neuropathologies. These included 1) schizophrenia, where NRSF was shown to be capable of directing allele-specific and stimulus-driven expression of MIR137 through identification of a novel promoter in this key schizophrenia genome wide associated gene; 2) cognitive dysfunction, polymorphisms within NRSF and its gene target BDNF influenced memory performance in patients with newly diagnosed epilepsy; 3) mood disorders, NRSF-signalling was identified as a significant pathway regulating cellular processes relevant to mood modification by pharmaceutical challenge and 4) NRSF-mediated regulation of microRNA-137 (miR-137) expression was demonstrated in vivo using a model of cortical spreading depression, consistent with its involvement in associated neuropathologies including epilepsy and ischemia. The analysis was expanded to a common non-neurological disease, breast cancer, where the previous work of others was extended to demonstrate a link between NRSF and miR-137 through the novel promoter identified in this study. Collectively these findings emphasise NRSF as a major contributor to cell pathogenesis, in part by modulation of miR-137, not only in a neuronal context but also in other systems.

Triggering receptor expressed on myeloid cells – 2 (TREM-2) is a receptor expressed mainly in myeloid cells. TREM-2 is involved in the resolution of inflammation through dampening Toll-like receptor (TLR) induced pro-inflammatory cytokine secretion and phagocytic functions. It is increased in inflammatory conditions including rheumatoid arthritis and stroke and also in wound healing. However, little is known about how its expression is regulated. This study analysed the expression of TREM-2 in various myeloid and non-myeloid cell types and investigated the regulation of TREM-2 expression in myeloid cells. TREM-2 was expressed in the Golgi apparatus of microglial cells. Several non-myeloid cell types also expressed TREM-2, including bronchial epithelial cells where it was located on cilia in healthy and diseased lung tissues. In THP-1 cells, the anti-inflammatory cytokines interleukin-4 and transforming growth factor-β1 (TGF-β1) induced TREM-2 expression through phosphoinositide 3-kinase (PI3K) and PI3K/ p38 MAP kinase signalling pathways respectively. TGF-β1-induced TREM-2 also required extracellular signal-regulated kinase 1/2 post-translationally for protein expression. Interestingly, TREM-2 was required for TGF-β1-induced matrix metalloproteinase (MMP)-1 expression, the most characterised MMP in wound healing, suggesting that TREM-2 may be required for the beneficial effects of TGF-β1 on wound healing by regulating MMP-1. An in vitro model of ischemia in stroke was then established to study the mechanisms of TREM-2 regulation in stroke. Oxygen glucose deprivation (OGD) had no direct effect on TREM-2 in N9 microglial cells. However, co-culture with healthy neurons reduced microglial TREM-2 expression, which was abolished in co-culture with OGD neurons, suggesting that in the healthy brain, microglial TREM-2 expression is suppressed by neurons, and this suppression is lost during ischaemia, increasing TREM-2 expression. In conclusion, this study characterised TREM-2 expression in non-myeloid cells and identified novel mediators and signalling pathways that regulate TREM-2 expression, which may be responsible for TREM-2 overexpression in inflammatory and ischemic conditions.