Characterisation and identification of brown adipose tissue on positron-emission tomography and magnetic resonance imaging Since the first published description of brown adipose tissue (BAT) in 1551 its reputation has changed from that of a mere curiosity of little physiological significance in adult humans, to meriting reclassification as a metabolic organ in its own right. Obesity is a major global public health problem. Modulation of non-shivering thermogenesis through BAT manipulation presents an attractive therapeutic target for inducing weight loss. Testing the efficacy of such pharmacological agents requires the development of a reliable imaging biomarker to enable BAT quantification. In this thesis we have evaluated the effectiveness of positron-emission tomography (PET) and magnetic resonance (MR) imaging in quantifying BAT. Retrospective analysis of 3,295 consecutive PET scans performed at University Hospitals Coventry & Warwickshire NHS Trust (Coventry, UK) in 2007-2012 showed 18F-FDG uptake consistent with BAT in 5.3% of scans. Gender (female), age (younger), BMI (lower), serum glucose (lower), time of day (earlier), and temperature (lower) were all significant predictors of BAT prevalence. Regression analysis showed patients’ age and preceding day’s minimum temperature to correlate most strongly with BAT volume, while the impact of other factors is less clear. We also showed the pattern of BAT uptake within individuals to be consistent across serial PET scans. Quantitative colocalisation techniques showed this degree of colocalisation to be significant in 14/15 patients, implying fixed BAT deposits in adult humans. Concerns over the high ionising radiation dose of PET scans has stimulated research into MR as an alternative means of detecting BAT, with the potential to identify BAT irrespective of its activation state. Using IDEAL FSE sequences acquired on a 3 Tesla clinical MR scanner, we found BAT to have a significantly lower fat fraction than white adipose tissue in rats post mortem and adult humans in vivo. Our efforts to identify BAT prospectively using fat fraction yielded inconsistent results.

No description available.

The morphology and function of the lumbar region is poorly understood. Better understanding of lumbar regional anatomy may enable improved understanding of lumbar stability and may also improve the clinical management of low back pain. Extensive researches have been carried out on the thoracolumbar anatomy and biomechanics. However, these studies lacked detailed anatomical knowledge about the morphology and function of the lumbar region. This study aims to provide a precise and detailed description of the anatomy of the lumbar spine and its supporting structures. A detailed and thorough literature review of background data was undertaken. Gross degenerative features in the lumbar vertebrae were documented. Three dimensional models of the superior and inferior lumbar articular facets were created by Microscribe. This allowed calculation of the facet orientation and surface area by Rhinoceros software. The surface area was increased towards the inferior vertebral levels, while the orientation became less sagittal inferiorly. The investigations suggest that the coronally oriented facet protects and supports the facet joint, while the sagittal orientation may predispose the facet joint to degenerative spondylisthesis. Gross observation of the thoracolumbar fascia documented the superficial myofascial thickenings, decussation and connections. The posterior and middle layers of the thoracolumbar fascia were identified. A three dimensional model enabled visualization of the bilaminar layers of the fascia which was reconstructed in a virtual space. The morphological measurements of the lumbar multifidus, longissimus and iliocostalis muscles were taken. The cross sectional area of the multifidus muscle was increased gradually towards the L5 level. The foot prints of the multifidus, longissimus, iliocostalis lumborum and inter-spinalis muscles enabled the measurement of the surface areas of the attachments of these muscles. The histological study revealed the fibrous enthesis of the iliocostalis muscle and its indirect attachment to the transverse process of the lumbar spine. The multifidus muscle is attached by a fibrocartilaginous enthesis to the articular process and the facet joint capsule. This study suggests that multifidus muscle supports and stabilizes the facet joints. The lumbar enthesis investigation should receive more attention in future studies. The clinical implications of different lumbar structures and functions may provide insight about the lumbar dysfunction. The ability to identify such differences in situ may facilitate varied clinical management of the various types of lumbar disorders.

Diabetic nephropathy is associated with secondary hypertension arising from aberrant sodium reabsorption in the kidney. This thesis characterises a novel human cell line derived from the human cortical collecting duct (HCD) to assess glucoseevoked changes in key elements, such as the serum and glucocorticoid inducible kinase (SGKI) and the epithelial sodium channel (ENaC), involved in the regulation of sodium transport. In addition I have also examined the effects of TGF-f3I and [Ci+]i on SGKI and ENaC expression. RT-PCR, western blot analysis, immunocytochemistry and single cell imaging were employed to determine presence, localisation and function of these elements under various glycaemic conditions. Our data suggest that high glucose, TGF-f3I and [Cl+]i up-regulate both SGKI and [alpha]ENaC protein expression, which in turn stimulates Na+ transport. In pathological conditions associated with aberrant Na + reabsorption, excessive levels of Na + may further exacerbate the state of hypertrophy, a common manifestation associated with diabetic nephropathy. Mechanical stress evoked TRPV4 m~diated changes in [Ca2+]i. Propagation of this Ca2+ signal via the gap junction protein connexin 43 (Cx-43) was enhanced following glucose treatment, as was Cx-43 expression. Under pathophysiological conditions these changes and the increased expression levels of our key signaling elements, may lead to deranged Na+ handling and inhibition of cell volume recovery mechanisms which together may further enhance the condition of diabetic nephropathy in Type 11 diabetes.

Articular cartilage in synovial joints can become damaged due to mechanical loading, trauma or wear and tear. The initiation and progression of damage in cartilage may lead to degenerative changes of the joint. However, links between mechanical loading and the initiation/progression of damage in cartilage remain poorly understood. In this thesis, the damaging effects of loading frequencies representative of normal (1 Hz), above normal (10Hz) and rapid heel-strikes (100Hz) on cartilage/cartilage-on-bone were assessed and compared to test the hypothesis that failure can be influenced by frequency. Bovine cartilage was used as a model for human cartilage. Materials testing machines were used to apply sinusoidally varying loads at different frequencies and altered maximum forces under different loading types. A metal indenter was used to apply cyclic loading on cartilage-on-bone specimens to produce failure on the surface of cartilage-on-bone specimens in compression. Fatigue failure of cartilage-on-bone specimens were determined using cyclic three-point bending. Propagation of an initial crack across the area of cartilage specimens with respect to increasing number of loading cycles were measured and compared under tension. The results from this thesis indicated that failure increases significantly (p < 0.05) in cartilage-on-bone specimens with increasing the loading frequency under compression and bending. Strain experienced by the cartilage specimens at higher frequency, e.g. 100 Hz, caused a greater crack growth under tension. The results from this work have many potential implications in the early onset of osteoarthritis. This is because rapid heel-strike rise times have been implicated in the early onset of osteoarthritis.

The prevalence of muscle dysfunction in elderly populations represents a significant burden on healthcare due to the increased risk of injury, and difficulty in maintaining activities of daily living. This thesis describes the application of advanced computational techniques designed to “learn” the structure of molecular networks to understanding human skeletal muscle ageing. Using this approach we have been able to discover a link between protein translation and age-dependent metabolic reprogramming. Experimental validation using the haploinsufficient eukaryotic initiation factor 6 (eIF6) mouse confirmed this important hypothesis and revealed a substantial molecular reprogramming. The role of eIF6 in skeletal muscle and myoblasts was further investigated, revealing potential up and down-stream signalling mechanisms. The process of angiogenesis is an important step in morphogenesis including systems as diverse as muscle regeneration and tumour growth. This thesis presents the first temporal model of transcriptional alterations in tumour and the surrounding stroma during vascularisation. The application of reverse engineering approaches that are able to integrate perturbation data lead to the hypothesis that modulation of the pro-inflammatory cytokine IL1α may be an important upstream event in angiogenesis.

This thesis investigates the effect of various physical and chemical surface modification methods on the permeation of topically applied pharmaceutical compounds through poly(dimethylsiloxane) (PDMS), a polymer frequently employed as a model barrier in in vitro skin permeation studies. Such studies are essential for safety, risk assessment, and quality control purposes, in addition to assisting in the design and development of efficacious topically applied medicines. The commercial availability, legal status, ease of handling, and the reproducibility of the permeation data associated with polymeric skin mimics renders them an attractive alternative to biological tissue. However, over-predictions of percutaneous absorption observed following the use of such membranes are a significant disadvantage when attempting to obtain quantitative toxicological data. Accordingly, the aims of the work presented in this thesis were to both reduce the permeability of PDMS to pharmaceutical compounds, and to increase correlation between permeation data obtained using the synthetic substitute and data obtained similarly using suitable biological tissue. Primarily, the potential of an air plasma pre-treatment to produce a lamellae-type structure in PDMS, endeavouring to more accurately model the architectural, physical, and chemical properties of the human stratum corneum, was investigated. Reductions in the permeability coefficient of up to 54.4 % were observed, rendering the modified system promising. Correlation analysis revealed an increase in correlation between the data collected using the modified synthetic substitute (R 2 = 0.86) and a selfcollated library of literature-derived epidermal tissue permeability data, relating to eighteen compounds and spanning a range of typical penetrants, compared to similar analysis using data obtained using the native substitute ( R 2 = 0.75), suggesting an increase in the predictive capability. It was hypothesised that an N2 plasma treatment may provide suitable surface functional groups on the PDMS substrate, namely amine groups, for the covalent attachment of biomolecules via an N,N'- dicylohexylcarbodiimide (DCC) coupling reaction, enabling the production of a skin mimic displaying enhanced biorelevance. Therefore, the effect of an N2 plasma pre-treatment on the permeation of a subset of the eighteen compounds investigated. It was found that the N2 plasma pre-treatment was advantageous in terms of offering a greater reduction in permeability, since longer treatment times could be employed i.e. reductions of up to 61.8 % were observed. However, significant surface oxidation was still observed, with only a marginal increase in nitrogen containing functionalities compared with the air plasma analogue i.e. 0.31 %. Furthermore, the treatment did not offer any additional increase in correlation between epidermal-derived data than previously observed. Further chemical methods of biomolecule attachment were pursued for use in the development of a lipidproteinaceous bilayer model, initiated in both cases by surface amination using an alkoxysilane. This was followed by a DCC coupling to an amino acid in the former approach, and use of a glutaraldehyde III linker molecule to attach the same amino acid, namely lysine, in the latter approach. In either case, no further reductions in the permeation of the pharmaceutical compounds tested were observed, with respect to that through plasma treated PDMS. In summary, the air plasma treatment of PDMS was found to be a promising approach to simultaneously reducing the permeability of a silicone skin mimic and increasing correlation with data obtained in similar studies employing biological tissue. Further, the covalent coupling of biomolecules to the surface of PDMS following surface amine group generation, via both plasma and wet chemical methods, appeared not to compromise the integrity the PDMS membranes relating to such applications, rendering the techniques compatible with the production of biorelevant semi-synthetic skin mimics.

The character of the functional output of a motor unit within skeletal muscle has been linked experimentally to the proteins found in the sarcomere, the smallest contractile unit of muscle fibre. Current mathematical models focus on either individual chemical reactions or the bulk properties of muscle with limited reference to the internal processes and structures within the muscle. Without an understanding of those internal properties, the normal function of muscle cannot be simulated and consequently muscular diseases and their treatments cannot be accurately modelled. In this project, a mathematical model has been developed which relates the chemomechanical cycle of individual events (crossbridges) to the transfer of mechanical energy through an actin filament, myosin cofilament and, by incorporating the protein titin, the mechanical properties of the interconnecting proteins in a section of sarcomere. Evaluation and parameterisation of the model were made by comparison with in vitro test data from the published literature at the level of a single crossbridge and single filaments. At the single filament level, the model was evaluated against two conditions: a low load high displacement (concentric contraction) and a high load low displacement (isometric contraction). In isometric loading the peak force level per unit length of actin filament was higher than that observed in vitro, the difference being attributable to the greater compliance in the substrate used in vitro to hold the myosin fragments (~37pN compared to ~12pN). The mean number of concurrent crossbridges was consistent between the model and in vitro data. Under low load the model demonstrated filament movement at speeds comparable to those measured in in vitro motility studies, although longer filaments in the model were required than those in vitro to reach the higher speeds (7μm vs. 2μm for ~8μm/s). By making the pre-lever reaction duration of the crossbridge cycle strain dependent it was possible to obtain long reaction cycles in low load scenarios comparable to those observed for fragments in solution while generating the actin filament speeds observed in vitro. It was necessary to have a distribution of attachment times across the filament in order to generate and maintain filament movement in the model; the variation being governed by the tension distribution in the filament. By applying a passive loading as generated by the titin protein the filaments moved more rapidly, with an increased contribution from each crossbridge to filament movement. Initial results indicate examination of the strain dependency of the post-lever reaction duration may modify filament speeds and will increase the proportion of each crossbridge movement that contributes to the actin filament propulsion (increase crossbridge efficiency). Examination of a selection of the model’s parameters gave an initial evaluation of how the model could be ‘tuned’ to change the number, reaction state and distribution in time of crossbridges to achieve changes in filament contraction speed, isometric force generation and the efficiency with which crossbridges are used; noting that one desired output may conflict with another. The interaction of the passive components in the structure of the sarcomere with the strain dependent reaction cycle at each crossbridge demonstrated the potential limitations of scaling and averaging localised events without consideration of the passive structures present in the fibre and muscle bulk. The model provides a means to examine the mechanisms and parameters of the sarcomere’s function and how those parameters may be adjusted to achieve different output characteristics. The model provides a foundation for the emulation of muscle fibres and a motor unit in health and disease.

The neuromodulator adenosine is involved in both physiological and pathological activity, such as sleep, epilepsy and stroke. However, the complex processes underlying the release, transport and clearance of adenosine from the extracellular space and their interactions are still poorly quantified. In this thesis I develop the �rst detailed model of the dynamics of adenosine in neural tissue, including intracellular and extracellular metabolism, using parameters taken from an extensive search of the literature. This approach also identifies physiological and metabolic parameters that have yet to be experimentally measured. The model provides estimates of the range of influence of adenosine, the distance where the extracellular concentration is greater than that required for half of the maximum inhibition by the dominant type of adenosine receptors in the cortex, and suggests that under physiological conditions the adenosine signal will be highly localised. The model predicts that adenosine concentration profiles are primarily determined by diffusion and that neuronal transport and metabolism are the dominant clearance mechanisms. The model can be used with either experimental or endogenous sources of adenosine, and I apply it to the bath application of adenosine to a tissue slice, (a method used extensively to study the e�ect of adenosine on synaptic transmission). The model is used to predict the effective dose response curve of bath applied adenosine and to compare the effects of transporter blockers. I then turn to the modelling of biosensors, which are used extensively to measure the concentration of various analytes in tissue, including adenosine. Biosensors are often calibrated in a flow injection system with a known concentration of the analyte. Mathematical and computational models are used to compare the response characteristics of biosensors in this free environment with the tortuous environment in which they are used. An estimated correction factor is obtained together with the sensitivity of this factor to the characteristics of the biosensor. This work provides a framework to move from qualitative studies of changes of adenosine in the brain to quantitative analysis of the spatio-temporal dynamics of adenosine signalling and its in uence on networks of neurons.

The ligaments of the wrist are highly variable and poorly described, which is more obvious on the ulnar side of the wrist. Previous studies highlighted the potential differences within the ligaments of the wrist but no consensus has been reached. Poor tissue description and inconsistent use of terminology hindered the reproducibility of the results. Improved understanding of the morphological variations between carpal bones may facilitate improved understanding of the ligamentous structure within the wrist. This study aims to identify the potential variations between carpal bones that could be used to separate palmar ligamentous patterns around the triquetrum-hamate joint into subgroups within the sample population. Investigations were performed following a detailed nomenclature and a clear definition of ligamentous structures to facilitate detailed description and reproducible results. Quantitative analyses were conducted using 3D modelling technique. Histological sections were then analysed to identify the structure of each ligamentous attachment. Variable patterns of ligamentous attachments were identified. Differences were not only obvious between samples but also between the right and left hands of the same person. These identifications suggested that the palmar ligamentous patterns around the triquetrum-hamate joint are best described as a spectrum with a higher affinity of the triquetrum-hamate-capitate ligament and the lunate-triquetrum ligament to be associated with type I lunate wrists on one extreme and type II lunate wrists with the palmar triquetrum-hamate ligament, triquetrum-hamate-capitate ligament and palmar radius-lunate-triquetrum ligament attachments at the other extreme. Histological analyses confirmed pervious established work regarding the mechanical role of ligaments in wrist joint biomechanics. Also, there were no significant differences between the quantitative data obtained from the Genelyn-embalmed and unembalmed specimens (p > 0.05). The current study demonstrated variable ligamentous patterns that suggest different bone restraints and two different patterns of motion. These findings support previous suggestions regarding separating the midcarpal joint into two distinct functional types. Type I wrists were identified with ligamentous attachments that are suggestive of rotating/translating hamate whilst type II wrists identified with ligamentous attachments that are suggestive of flexing/extending hamate motion based upon the patterns of the ligamentous attachments in relation to the morphological features of the underlying lunate type of the wrist. This opens the horizon for particular consideration and/or modification of surgical procedures, which may enhance the clinical management of wrist dysfunction.