Search

In the UK, stroke is the third most common cause of death after heart disease and cancer. Importantly, most strokes are not fatal meaning stroke is the leading cause of adult disability, with one third of survivors still functionally dependent after one year. Post stroke hyperglycaemia (PSH) predicts poor outcome independent of age, stroke type or severity and occurs in over 60% of patients without a diagnosis of diabetes and in more than 90% of diabetic patients. Around 56% of hyperglycaemic acute stroke patients are subsequently diagnosed with insulin resistance, manifesting as impaired glucose tolerance, impaired fasting glucose or the “metabolic syndrome.” Clinical guidelines recommend routine blood glucose monitoring, and intervention with insulin for hyperglycaemia. However, there is currently no clinical evidence of benefit from insulin use, and recent findings raise concerns about the safety of insulin use in predominantly non-diabetic acute stroke populations. Hyperglycaemia in animal models of cerebral ischaemia is associated with increased infarct size. However, a recent systematic review highlighted uncertainties about the relevance of many previously published preclinical studies to the typical clinical picture of PSH with respect to whether clinically relevant elevations in blood glucose exacerbate ischaemic brain damage in models other than those of type I diabetes, or whether responses are different in animals with and without features of metabolic syndrome. The aim of this thesis was to investigate the differential effects of clinically relevant hyperglycaemia on the acute evolution of damage in animals with and without features of the metabolic syndrome and to investigate the potential mechanistic role of oxidative stress in glucose mediated ischaemic damage. I hypothesised that hyperglycaemia would exacerbate acute lesion evolution and final infarct volume in animals with and without features of the metabolic syndrome and that oxidative stress mechanisms would be involved. Developing suitable animal models for investigating post stroke hyperglycaemia and metabolic syndrome The first aim of this thesis was to develop a series of animal models for the investigation of clinically relevant PSH, metabolic syndrome in the presence and absence of PSH, and an appropriate focal cerebral ischaemia model. A previously described model of metabolic syndrome using the spontaneously hypertensive stroke-prone rat (SHRSP) fed on a 60% fructose diet for two weeks was set up in the laboratory and glucose tolerance, adiposity, plasma insulin, plasma triglycerides, cholesterol were measured and compared to the SHRSP reference strain, the Wistar-Kyoto (WKY). Fructose fed-SHRSP rats exhibited glucose intolerance, hypertriglyceridaemia, hyperinsulinaemia, increase adiposity and reduced HDL cholesterol compared to the WKY controls. This suitably modelled the features of the metabolic syndrome. Clinically relevant levels of PSH were achieved using a bolus intraperitoneal injection of 15% glucose (10ml/kg) 10 minutes prior to permanent middle cerebral artery occlusion (MCAO). Two methods of MCAO were compared; intraluminal filament (ILF) model and a distal diathermy MCAO model. The ILF method produced expansive lesions incorporating the majority of the ipsilateral hemisphere whereas the distal diathermy occlusion produced smaller cortical infarcts which were shown to produce a large volume of penumbra. The diathermy model was selected for all future experiments. Investigating the effects of hyperglycaemia on acute lesion growth after focal cerebral ischaemia in rats with and without components of the metabolic syndrome The second aim of this thesis was to determine the effects of hyperglycaemia on acute ischaemic lesion evolution and infarct volume in rats with and without features of the metabolic syndrome. Hyperglycaemia was induced in WKY or fructose-fed SHRSP rats 10 minutes prior to permanent MCAO. Magnetic resonance imaging (MRI) was used to quantify the expansion of the lesion using ADC maps calculated from diffusion weighted imaging (DWI) carried out over the first 4 hours after MCAO. Acute hyperglycaemia, at clinically relevant levels, exacerbated early ischaemic damage in both normal and metabolic syndrome rats. Hyperglycaemia worsened infarct volume at 24h in normal but not metabolic syndrome rats. These data suggest that management of hyperglycaemia may be most beneficial in the absence of an underlying dysglycaemia. Using a marker of oxidative stress to probe the tissue sections suggested that perhaps hyperglycaemic effects were mediated through oxidative stress mechanisms. Investigation of oxidative stress mechanisms in hyperglycaemia-dependant ischaemic damage. Based on the potential involvement of oxidative stress mechanisms in hyperglycaemia-associated ischaemic damage this was investigated in a proof of concept study using pre-treatment with a SOD/catalase mimetic, EUK-134. The rationale for this study was based on the involvement of increased superoxide production in chronic hyperglycaemia and diabetic complications. Sprague-Dawley rats received the same dose of glucose previously used 10 minutes prior to MCAO by distal diathermy. EUK-134 (2.5mg/kg) or vehicle (saline) was administered 20 minutes prior to MCAO. Infarct volume, calculated from RARE T2-weighted MR images and neurological scoring was measured 24 hours post MCAO. Arterial superoxide levels were measured along with the lipid peroxidation marker malondialdehyde in cerebral tissue. Hyperglycaemia increased infarct volume although no effect of high blood glucose on tissue lipid peroxidation was observed. EUK-134 failed to reduce infarct volume. EUK-134 reduced lipid peroxidation in ipsilateral cortex of normoglycaemic rats but not in hyperglycaemic rats. Thus, hyperglycaemia associated damage was not reduced by attenuating oxidative stress. Generating a threshold for quantitative T2 maps Using the data generated in the previous Chapter a study was carried out to determine if improvements could be made in the analysis of MRI data in pre-clinical rat studies. RARE T2-weighted images, where T2 relaxation time denotes contrast, have been used for infarct quantification 24 hours post MCAO. Generating a lesion volume from these images requires manual tracing of the visually defined infarct from the images. A degree of error is associated with this methodology either as inter-investigator reproducibility or in repeats conducted by the same investigator. The RARE T2 images are not quantifiable as arbitrary values denote contrast and this prevents direct comparison of T2 relaxation times for a particular structure between subjects. Using a MSME T2 sequence, quantitative T2 maps can be generated which calculate absolute T2 relaxation times for each voxel. Using quantitative maps an attempt was made to establish a threshold to discriminate between normal and ischaemically damaged tissue in animals that had undergone MCAO using the distal diathermy model. This threshold aimed to minimise inter- or intra-investigator error. A T2 threshold of abnormality for the quantitative T2 maps was established from the RARE T2 weighted images. Tissue that had a T2 relaxation time of 76 ms or more was deemed to be infarct according to the derived threshold. The suitability of the threshold was then assessed in a separate cohort and was found to produce similar infarct volumes to the manual delineation of RARE T2-weighted images. The 76ms threshold was also applied to animals that had undergone MCAO using the intraluminal filament model. However, the calculated threshold was not suitable for the intraluminal filament model due to interferences from cerebrospinal fluid signal. A more selective method for determining infarct volume in animals subjected to intraluminal filament MCAO needs to be established.

The studies of hypertension, described herein, combined complementary strategies to examine the relevance of human chromosome 5q31.1-qter to blood pressure as a quantitative trait, and hypertension as a qualitative trait. This distal region of chromosome 5 was focused upon due to its implication by other human and animal model studies and the cluster of cardiovascular candidate genes located there. Eleven microsatellites across 55 cM of 5q31.1 to qter were genotyped in 212 hypertensive nuclear families of the Silesian Hypertension Study. Two-point and multipoint analyses showed linkage of a 7 cM region to both hypertensive status and blood pressure phenotypes. A maximal multipoint Z-score of 2.2 for systolic blood pressure (SBP) was obtained proximal to the marker D5s1480. In light of the debate regarding significance in linkage studies, this quantitative trait locus (QTL) was confirmed in a second sample derived from the Scottish population. Genotyping of the same markers was done for 1,469 individuals from sibships of the MIDSPAN Family Study. Two-point and multipoint analyses confirmed a reproducible 7 cM QTL with a maximum multipoint logarithm of odds (LOD) score of 1.8 obtained for mean arterial pressure. Several putative candidate genes were located within the QTL including the b2-adrenergic receptor gene, ADRB2, which is known to have a major role in cardiovascular physiology. Three functional loci, Arg16Gly, Gln27Glu and Thr164Ile, were genotyped within the ADRB2 gene using the Silesian sample. Single locus and haplotype analyses using transmission disequilibrium tests and family-based association methods showed no association to hypertension status (P>0.05), effectively excluding this gene in the Silesian sample. The fibroblast growth factor 1 gene, FGF1, is located at the centre of the QTL. This gene had previously been cited as a putative blood pressure candidate gene in the 5q region but not studied in humans. With roles in endothelial cell proliferation, possible oxidative stress defence and proven direct effects on blood pressure in rodent models, we considered it a possible novel candidate. Coding regions and 3’ UTR were sequenced and family-based association analyses performed for three polymorphisms which were discovered in the 3’ UTR. Association of SBP (P=0.045), pulse pressure (P=0.019) and hypertension status (P=0.038) was demonstrated to the most proximal locus. Association of hypertension to a three-locus haplotype was also shown (P=0.020). The work on essential hypertension demonstrated the merit of confirming linkage in independent populations, implicating a 7 cM region of 5q31.1-q33 linked to blood pressure. Subsequent candidate gene studies utilised the added value of haplotype analyses, excluding the ADRB2 locus but showing interesting data to implicate the novel FGF1 locus as a putative positional candidate. Further fine mapping strategies are now underway to define, more clearly, the haplotype tag across the FGF1 locus.