The carbon and energy metabolism of the human pathogen Clostridium difficile is poorly understood. Amino acid metabolism by the Stickland reactions has previously been described as a primary source of energy in a number of Clostridium species, especially when grown in a medium containing only amino acids. Deeper insights may be gained by metabolic analyses using liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS) platforms, but such experiments are best performed in well-defined growth media. A number of C. difficile strains have been successfully cultivated on defined media but these media provide an excess of nutrients, particularly in terms of amino acid provision, resulting in undesirable background growth in the absence of glucose. To overcome these challenges, three variants of a defined medium were developed that contain the essential nutrients that support the growth of this bacterium, in the presence of a carbon and energy source such as glucose together with an LC-MS/MS method that will simultaneously measure all twenty amino acids. Since the focus in this study was amino acid utilisation, a comprehensive and most effective technique that will provide as much information as possible for understanding the metabolic requirements of eight Clostridium difficile strains (CD630∆erm, DH196, R20291, EK15, EK28, R12801, L26, O17 Serotype F) and two transposon mutants (CRG-2979, CRG-3887) was required. Due to high water solubility and the range of ionic characteristics of amino acids, an aqueous normal phase chromatographic method was considered to simultaneously separate a mixture of amino acids without derivatisation. Aqueous normal phase chromatographic method represents an important new technology with a capability of the silica-hydride-base stationary phases, offering a distinct advantage for practical application with a high degree of reproducibility and long-term stability of polar and non-polar compounds. The developed methods were subsequently adapted to study amino acid utilisation in the presence or absence of a fermentable carbohydrate and/or selenium. OD determinations were performed by measuring absorbance at 600 nm (or OD600) using a Biomate 3 spectrophotometer. The concentration of individual amino acids remaining in the spent C. difficile culture media was measured after 24 h and/or 48 h using the developed LC-MS/MS method in order to determine which amino acids were being utilised by these organisms and in which order. This analysis should give further insight on the importance of amino acids for the survival of this bacterium in the gut, which may possibly lead to discovery of novel fermentable products and metabolic pathways and/or eventually aid in the successful control of the disease. Data obtained for CD630∆erm, DH196, R20291, EK15, EK28, R12801, L26, O17 Serotype F strains showed that they all grew on the fully defined medium, with different growth profiles in terms of lag phase, growth rate, and maximum OD reached. The LC-MS/MS data generated suggest that cysteine, glutamine, isoleucine, leucine, serine, threonine and tyrosine are preferentially utilised, both in the presence and absence of glucose. Several other amino acids, including asparagine, glycine, phenylalanine, proline and valine were also utilised but to a lesser extent. Notable in this study was the lack of glutamate utilisation, except by strain L26, and the excretion of alanine after its initial uptake by most of the tested stains. The excretion of alanine may be due to the use of pyruvate as an amino acceptor during the degradation of preferentially fermented amino acids, whereas, glutamate is not a substrate for most C. difficile strains. Thus, LC-MS/MS profiling confirmed that these organisms derive most of their carbon and energy from the fermentation of a selected range of amino acids. Given the importance of selenium-dependent Stickland reactions to the growth of this bacterium, further studies were undertaken to evaluate the metabolism of amino acids by two different C. difficile strains (630∆erm, R20291) and two transposon mutants CRG-2979 (defective in hadB, encoding one of the two subunits of hydroxyisocaproyl-CoA dehydratase required for reductive degradation of leucine) and CRG-3887 (defective in selA encoding selenocysteine synthase) in the presence or absence of glucose/selenium. LC-MS/MS data reveal that amino acid utilisation was affected by the presence of selenite, notably proline utilisation, which could be explained by the presence of the enzyme proline reductase and the lack of glycine consumption, known to be selenium-dependent. The non-utilisation of glycine could be explained by the presence of proline which represses the formation of the necessary enzyme systems required for glycine degradation. Data generated for CRG-2979 reveals that this mutant could only thrive in the presence of selenium when glucose is present, possibly due to the presence of proline replacing leucine as the major Stickland acceptor. The results of the transposon mutant CRG-3887, were much of a surprise too, because this mutant was predicted to be deficient in proline and glycine breakdown which is also reliant on selenoenzymes. This suggests the presence of selenium independent proline fermentation pathway although this hypothesis is not supported by the existing literature. Experimental data provided further evidence about the ability of this bacterium to obtain its carbon and energy in the absence of a fermentable carbohydrate; by Stickland reactions and that the presence or absence of certain amino acids could repress the utilisation or biosynthesis of other amino acids.

The aim of this thesis was to investigate the mechanisms that mediate the effects of beta-adrenergic agonists (BA) and Growth Hormone (GH) in sheep, by examining the changes in skeletal muscle transcriptome and blood metabolome in order to identify the predominant metabolic mechanisms by which muscle hypertrophy was mediated. Male lambs (120 days old) were all fed a high protein/energy feed ad-libitum, with the GH group (n=10) receiving a single subcutaneous injection of bovine GH (3.75mg/kg body weight, POSILAC, Monsanto) on day 1; the BA group (n=10) receiving BA (cimaterol) at 10mg/kg feed, whereas the control group (CO, n=11) only had the ad-libitum feed. After 6 days sheep were slaughtered, plasma and samples of the Longissimus dorsi (LD), Supraspinatus (SS) muscles and liver were collected. The effect of treatments on the LD transcriptome was assessed on a subset of samples (n=3 from each treatment) via a cross-species approach using the Affymetrix Human U133+2 GeneChip array (47K human microarray). Verification of identified differentially expressed genes and proteins was by quantitative RT-PCR or western blotting, respectively, on all animals. Metabolomics analysis of plasma samples was carried out by Metabolon Inc. (USA) using GC/MS and LC/MS/MS platforms. BA, but not GH, significantly (P<0.05) increased muscle weights and this was associated transition to large fast-glycolytic muscle fibre types. In GH, but not BA treated animals, there was an increase in liver weights (P<0.001). This was associated with an increase in the whole liver content of glycogen (P<0.001), protein (P<0.01), and lipid (P<0.05) content. Analysis of the LD transcriptome of the treated sheep identified 477 and 316 transcripts were significantly altered (P<0.05 and 1.5 fold change) by BA and GH respectively, relative to controls. This muscle was selected as it is a commercial valuable muscle and is commonly used for muscle biochemical studies therefore this would allow us to make comparisons to other studies, including our own. In addition it is a fast glycolytic muscle fibre type there could be compared against SS muscle (oxidative muscle fibre type). BA decreased the expression of genes involved with oxidative phosphorylation and upregulated those serine biosynthetic pathways. Subsequent qRT-PCR analysis showed a BA induced increase in expression of phosphoglycerate dehydrogenase (PHGDH) (P<0.05) and phosphoserine-aminotransferase (PSAT) (P<0.05) mRNA in both LD and SS but not liver. In LD there was also an increase (P<0.001) in PHGDH protein in muscle from BA treated sheep relative to GH treated sheep. Up-regulation of this pathway has been previously reported in cancer cells which has a tendency to be associated with an increase in gene expression of a specific isoform of the glycolysis enzyme pyruvate kinase (PKM2) which has reduced activity. Total PKM and PKM1 and PKM2 isoforms were increased in the SS and LD of BA treated sheep (P<0.05). Previous studies in cancer cells have suggested that increases in serine synthesis are mediated by changes in PKM2 expression and associated enzyme activity. The lack of a differential increase in PKM2 suggested that the regulation of muscle PK in BA treated animals was not critical to the potential increase in serine synthesis capacity. No clear change in PKM gene expression suggested this was not the mechanism by which the serine synthesis pathway was stimulated. There was an increase (P<0.05) in the expression the mitochondrial form of phosphoenolpyruvate carboxykinase (PCK2) in the LD of BA treated sheep, which might be expected to increase gluconeogenic potential thereby increasing intermediates that could be used for serine synthesis. There was no effect of this gene on sheep treated with GH. An increase in the gene expression of asparagine synthetase (ASNS) was also seen in the muscles of BA but not GH treated sheep (P<0.001) and there was no effect on their livers, which further suggested that BA was influencing the production of nonessential amino acids. Metabolomics analysis showed that products of triacylglycerol breakdown, glycerol and free fatty acids, were all elevated in the plasma of both BA and GH treatments, indicating lipolytic effects but the increase in the free fatty acid profile were more pronounced with GH treatment (P<0.05). Likewise GH rather than BA had a greater impact on elevating plasma glucose and associated metabolites such as pyruvate (P<0.05). There was no effect of either treatment on plasma serine or asparagine concentrations. However there was a decrease in glycine (P<0.05) and glutamine (P<0.05) in GH relative to control, with BA decreasing histidine (P<0.05) and methionine (P<0.01) relative to control. Cell culture experiments were carried out in the myogenic C2C12 cell line to determine if the genes associated with the GH and BA response in sheep were affected during myogenesis and whether there was an effect of des (1-3) IGF-I and dibutyryl cyclic adenosine monophosphate (dbcAMP) that stimulates GH and BA signalling pathways respectively. During differentiation, without treatment, gene expression of PHGDH and PSAT enzymes declined (P<0.05), which might be expected as cells move from a proliferative to a terminally differentiate state. There was no clear effect of treatment on genes associated with the serine synthesis pathway suggesting that the effects of BA, in particular, are on muscle fibres rather than differentiating cells. Of the two growth promoters examined in this thesis BA appears to be the most potent in skeletal muscle. A clear effect of this agent was an increase in the gene expression of the serine biosynthetic pathway, which has been shown to be upregulated in various cancers and, in this pathology, is thought to be a novel mechanism for hyperplastic growth. The associated changes in the expression of genes such as ASNS and PCK2 indicate that their co-ordinated upregulation could be mediated via endoplasmic reticulum stress response mechanisms. Unlike GH, BA does not appear to have a major effect upon the systemic mobilisation of nutrients, but instead seems to targets muscle fibres, activating muscle biosynthetic pathways that potentially provide the substrates required for growth.

Background: Osteoarthritis (OA) is the most common musculoskeletal disease worldwide and a major cause of chronic pain. Treatment of OA pain is still suboptimal due to limited efficacy and considerable side effects of available analgesics. Pain in OA has a significant central component. Cyclooxygenases (COXs) and glial cells in the spinal cord and the periaqueductal gray (PAG) play a significant role in central sensitisation and pain modulation. Vascular endothelial growth factor-A (VEGF-A) is a key molecule in normal and pathological angiogenesis. Serine arginine protein kinase 1 (SRPK1), which phosphorylates serine arginine splice factor 1 (SRSP1), controls VEGF gene alternative splicing. This results in two splice variants; VEGF-A165a, which is pro-angiogenic and pro-nociceptive, and anti-angiogenic VEGF-A165b, which showed anti-nociceptive effects in models of neuropathic and inflammatory pain. Objective: This thesis investigated changes in COX expression and glial activation in the spinal cord and the PAG in the monosodium iodoacetate (MIA) model of OA. It also addressed the effects of VEGF-A165b and SRPK1 inhibitor SPHINX-31 on pain behaviour and joint pathology, as well as COX expression and glial activation in the spinal cord and the PAG in the same model. Hypothesis: VEGF-A165b and SPHINX-31 can prevent and/or reverse enhanced pain behaviour in the MIA model of OA, through involvement of spinal glial cells and COXs in the PAG and spinal cord. Vascular-astrocyte association in the PAG is enhanced in the MIA model of OA, and this effect is reversed by administration of SPHINX-31. Methods: Rats received an intra-articular injection of MIA (1 mg) in the knee. In one study, animals were treated with VEGF-A165b (i.p. 20 ng/g body weight twice weekly) on days 0-13 (VEGF(d0-13) group) or days 14-28 (VEGF(d14-28) group) after MIA injection; in another study, rats received SPHINX-31 (i.p. 0.8 µg/g body weight twice weekly; (MIA/SPHINX group)) for 19 days after induction of the model. Pain behaviour was monitored throughout the studies, at the end of which (day 28) tissues were collected for the assessment of joint histopathology and the evaluation of spinal COX-2 mRNA expression by PCR. In addition, immunofluorescence (IF) was used to assess COX-2 expression and glial activation in the spinal cord, as well as astrocyte activation and vascular-astrocyte association in the PAG. Results: VEGF-A165b significantly attenuated weight bearing asymmetry (%) in MIA rats on day 28 (29.58 ± 1.803 in MIA/VEGF(d0-13) group vs. 22.95 ± 2.088 in MIA/PBS group, p < 0.01; 29.23 ± 1.49 in VEGF(d14-28) vs. 22.95 ± 2.088 in MIA/PBS, p < 0.05). VEGF-A165b reversed mechanical withdrawal thresholds to the naïve level, but without reaching statistical significance. No significant changes in knee joint pathology were observed in VEGF-A165b treated MIA rats compared to the MIA/PBS counterparts. In the MIA/VEGF(d0-13) group, contralateral deep laminae of the dorsal horn had a higher percentage (%) of non-neuronal cells expressing COX-2 than the corresponding superficial laminae (3.26 ±1.16 vs 1.12 ± 0.43, p < 0.05), while no difference was observed in the MIA/PBS group. Administration of VEGF-A165b did not significantly affect spinal microglia and astrocyte activation, nor COX-2 expression in the PAG. SPHINX-31 had no significant effects on pain behaviour, joint pathology or spinal COX-2 expression in the MIA model of OA. On the other hand, MIA/SPHINX group exhibited a higher activation of spinal microglia than MIA controls (% of CD11b +ve cells in MIA/SPHINX-31 vs MIA/vehicle groups: 6.36 ± 0.89 vs 1.72 ± 0.38, p < 0.01). In addition, SPHINX-31 significantly increased astrocyte activation in the ipsilateral dorsolateral (DL) PAG relative to corresponding ventrolateral (VL) PAG (GFAP IF intensity: 12.89 ± 1.52 vs 8.46 ± 0.84, p < 0.05), and it increased vascular-astrocyte association (%) in the contralateral DL PAG relative to corresponding VL PAG (70.35 ± 7.68 vs 38.92 ± 8.19, p < 0.05). Interestingly, naïve rats had a significantly higher astrocyte activation and vascular-astrocyte association in the VL PAG than in DL PAG. Conclusions: VEGF-A165b exerted a significant antinociceptive effect in the MIA model of OA without affecting joint pathology, spinal glial cell activation or COX-2 expression in the PAG. SPHINX-31 did not reverse pain behaviour, but showed a potential effect on astrocyte activation and vascular-astrocyte association in the DL PAG relative VL PAG in the MIA model of OA.

Phosphorus (P) is one of the essential nutrients for plant growth; however, it is usually in low availability to plants in most soils. P deficiency/low-P in the early growth stages of wheat can cause reductions in tiller and head formation, which poses a threat to wheat productivity and global food security. Genetic variation of phosphate use efficiency (PUE) has been documented in wheat. PUE can be improved under P deficiency/low-P by P-stress-responsive adaptation mechanisms that increase P acquisition and/or utilisation. Major regulatory components involved in PUE include Pi itself, microRNAs, hormones and sugars. In addition, several transcriptional factors (TFs) appear to play crucial roles in the regulatory complexity controlling the expression of P-stress-responsive genes. A better understanding of their roles may help to achieve favourable expression patterns of downstream genes and hence potentials to develop wheat cultivars with improved PUE in future crop improvement programme. Using bioinformatic approaches, this study identified two TFs, TaPtf1 and TaMyb67, which may act as key components in PUE in wheat. Their roles in regulating PUE were investigated through molecular and transgenic approaches. Overexpression constructs for TaPtf1 and TaMyb67 were created and subsequently transformed into wheat by Agrobacterium-mediated transformation. Selected transgenic lines were studied for overexpression of these transgenes and their effects on growth-, harvest- and PUE-related plant traits in a soil-pot experiment under different P supply. The phenotypic effects of TaPtf1 in the transgenic lines were implicated to be P-stress responsive and likely associated with plant height, biomass, grain filling and P accumulation in shoots. The results appeared to be consistent with previous studies of TaPtf1 in wild-type wheat suggesting that TaPtf1 has a functional divergence from OsPtf1/ZmPtf1. On the other hand, TaMyb67 was shown to be a likely ecotypic variation of TaPhr1-B1. TaMyb67 transgenic lines gave no clear evidence of phenotypic differences, presumably due to the downstream P-stress-responsive genes regulated by TaMyb67 being unresponsive to high-P and the low level of (or no) overexpression of TaMyb67 under low-P in these lines.

Microglia play a central role in immune surveillance and modulation of neuroinflammation as well as playing a role in neurodevelopment. Microglia involve in the development of pathological pain in adults but not early in life. However little detail is known about the changing phenotype of microglia during development. We examined age-related changes in microglia following activation with pathogen- and damage- associated molecular patterns (PAMPs/DAMPs). Microglial cultures were prepared from neonatal postnatal day (P1) and adult (P40) rat brains and spinal cords. Immunocytochemistry, qRT-PCR and functional assays were used to identify age-related differences. Adult microglia display a pro-inflammatory immune profile characterized by significantly increased IL-1β mRNA levels in response to PAMPs and DAMPs. In contrast, IL-1β mRNA in neonatal microglia showed a slight increase after stimulation with DAMPs. Anti-inflammatory gene expression was significantly increased in neonatal microglia relative to adult microglia. Compared to adult microglia, neonatal cells had increased phagocytic activity when unstimulated and following activation with LPS and ATP. Moreover, the nuclear receptor Nurr1 may play a major role in reducing pro-inflammatory signalling and promoting the anti-inflammatory phenotype in neonatal microglia. Nurr1 isoforms are differentially expressed in neonatal and adult microglia, with the Nurr1a isoform being significantly elevated in neonatal cells. Using lentiviral vector-mediated expression of Nurr1 isoforms, we also show that over-expression of TINUR, a splice variant of Nurr1, in neonatal and adult microglia attenuates inflammation by trans-repression the IL-1β expression and trans-activation the IL-10 gene expression following ATP exposure. Together, these data provide evidence for age-related difference in microglial function during postnatal development. In addition, these findings demonstrate insight into the mechanisms by which Nurr1 might act, and suggest potential therapeutic targets for the treatment of neuro-inflammatory diseases.

Corn Hageman factor inhibitor (CHFI) is a bifunctional serine protease / α-amylase inhibitor protein having 127 residues and a molecular weight of 13.6-kDa. CHFI is selective toward FXIIa without affecting the function of the other coagulation factors. Coagulation FXII is a serine protease recognized to cause kinin generation and blood coagulation, cleaving plasma kallikrein and FXI. Results from FXII-deficient animal models proposed that this protein contributes to stable thrombosis that can cause obstruction of the blood vessels and its subsequent complications such as ischemic stroke. In contrast to other blood coagulation factors, deficiency in FXII is not related with haemorrhage in patients or in animals. These findings propose that specific inhibition of FXII could be an attractive medicine and a new method of anticoagulation to treat or prevent pathological thrombosis that could have a lower risk for bleeding and a safer anticoagulation profile than the currently available anticoagulants. Therefore, the current PhD project aimed at pharmacological investigation into CHFI-FXII interaction and FXII function at molecular level through the following objectives: first, developing an efficient expression and purification system for generating soluble and functional recombinant native type CHFI and establishing an inhibitory activity assay against FXIIa to verify the proper function of the recombinant protein. The second objective was testing the different recombinant variants of CHFI with the desired point mutations guided by a proper prediction study of CHFI-FXII interaction. The third was to investigate into the hypothesis of the tight-binding property of CHFI via different approaches of enzyme inhibition mechanisms and kinetic data analysis. The last objective was to investigate into the function of FXII by examining theeffect of Cys466 and glycosylated peptide remnant from the proline-rich region on the function of the catalytic domain via characterizing the different recombinant variants of the catalytic domain of FXIIa, FXIIc, FXIIac, HISTF-β FXII, and MBP-β-FXIIa. In the current study, an efficient system for soluble expression, single step purification and proper storage of functional, wild type rHIS-GST-CHFI was, for the first time, identified. The fully functional recombinant protein was verified via developing an inhibition test against FXIIa. The established expression, purification and inhibition assays were used as a fundamental guide to both generating and characterizing mutant proteins of interest that were made on the basis of an appropriate docking model of CHFI-FXIIa interaction. For the first time, the current investigation into the question of specificity of CHFI against FXII revealed that the central Arg34 at the very top of the fully exposed region of CHFI inhibition loop play a central role in the inhibition function of CHFI toward FXIIa. In addition, this study identified Trp22 at the N-terminus and Arg43 at the C-terminus of the central inhibition loop as two key interaction residues with FXIIa. It was also observed that, in the preinhibition test, CHFI behaves as a noncompetitive inhibitor. In contrast, it acts as a competitive inhibitor in the acute inhibition test, proposing that CHFI is a competitive inhibitor with slow degree of reversibility due to tightness of binding. Reversibility assay showed that CHFI is an inhibitor with slow degree of dissociation. The tight-binding property of CHFI could be due to a non-active site interaction and or numerous hydrogen bonds between the III key interaction residues and their potential targets on FXIIa. With respect to the investigation into FXII function, It was observed that both Cys340-Cys466 and glycosylated peptide fragment of the proline-rich region have a functional role for the full catalytic activity of FXII protease domain. Cumulatively, the current study identified the key important residues on the exposed surface of CHFI and their potential target residues on the surface of FXIIa that would be highly informative and important factors helping to understand the mechanism of selective and tight binding interaction of CHFI with FXIIa. This project can be considered as an early, necessary approach to design novel, specific and safe anticoagulants for the treatment of thrombosis and its complications.

Tomato (Solanum lycopersicum L.) is one of the most important vegetable crops cultivated worldwide. Rab GTPases are involved in the processes of intracellular membrane traffic, protein secretion and signal responses for targeting of molecules for secretion into the cell wall. Silencing these genes is likely to affect fruit softening by blocking the export of a range of cell wall modifying factors. A cDNA clone from the tomato fruit encoding a protein homologous to rab11/YPT3 was developmentally regulated during fruit ripening. Antisense fruit changed colour but failed to soften normally (Lu et al., 2001). The aim of the project is to investigate the role of Rabs and other elements of intracellular membrane trafficking in secretion of cell wall modifying factors associated with fruit softening. Sixty genes encoding Rab GTPases in tomato were identified by in silico analysis by screening the EST databases and were classified as Rab GTPases according to Release 12 of TIGR Tomato Gene Index. The genes were grouped into 8 different functional classes ranging from Rab1, 2, 5, 6, 7, 8, 11 and Rab18. A phyletic tree was constructed which indicated all the genes were falling into 8 different clades belonging to each class. A comprehensive in silico analysis of other genes involved in intracellular membrane trafficking like Rab GEF, GAP, GDI, and other genes like ARF, ARL, RAN, ROP, SAR (Vernoud et al., 2003)was done by using the same databases. The results showed that there are thirteen ARF GTPases, eighteen ARL GTPases, seven SAR GTPases, eight ROP GTPases, and a total of nine RAN GTPases in tomato. The Rab GTPase interacting partners like GAP, GDI and GEF were also identified. The results revealed that there are twenty three Rab GAPs, two Rab GEFs and a total of six Rab GDIs were found. The information about the genes was updated throughout the course of the present investigation right from the TIGR DFCI TGI release11,12 and 13, SGN Unigene Build 1 and 2 as well the latest release of tomato whole genome sequence database at SGN from SGN ITAG-1 to ITAG-2.4. After the final analysis, a total of 55 Rab GTPases were grouped into eight different sub classes with five Rab1/D, five Rab2/B, four Rab5/F, three Rab6/H, four Rab7/G, five Rab8/E, twenty five Rab11/A and four Rab18/C. Solyc Rab11/A was further classified into 6 sub groups from A1 to A6. The results of classification led to identification of seven RabA1, six RabA2, one RabA3, five RabA4, five RabA5 and one RabA6. The phylogenetic analysis of the tomato Rab GTPases and a comparative analysis with Arabidopsis Rab GTPase confirmed the groups and classes. Similarly four RAN, ten ROP, twelve ARF, four SAR, six ARL GTPases were identified. In case of ARF GTPases seven ARFA, four ARFB, and one ARF C was found. The regulators of small GTPases identified in tomato had one RAB GEF, two ROP GEF, four RABGDI and five ROP GDI, twenty two RabGAP, 2 RAN GAP, nine ROP GAP and nineteen ARF GAP. Comparative phyletic analysis of all the above with Arabidopsis genes confirmed the classification. The molecular and phenotypic characterization confirmed the presence of antisense LeRab11a transgene and the phenotypes were heritable. The textural analysis to characterize the pattern of fruit softening of the wild type and transgenic plants by compression and penetration test from different fruit developmental stages was done and the results showed a sequential pattern of delayed softening and the transgenic fruits were firmer than wild type at the later stages of fruit ripening and had enhanced shelf life and reduced susceptibility to fungal diseases. The relative expression profiles generated by qPCR also showed the pattern of differential expression of Rab11 group of GTPases in wild type and transgenic fruits. Along with this the latest freely available expression database were also used to generated the expression profiles of Rab GTPases during different developmental stages not only in wild type Ailsa Craig (AC++) but also in Heinz cultivar as well as in Solanum pimpinellifolium.

The mouthfeel and texture of food and beverage products play an important role in consumer liking. However, to date these key sensory properties have received limited attention from researchers and our understanding of texture and mouthfeel perception in food remains restricted. For tea, a popular beverage enjoyed around the globe, astringency is the most frequently highlighted mouthfeel attribute. Since black tea is commonly consumed with milk and sugar, research investigating the mouthfeel of tea as impact by these added ingredients is of commercial interest. The aim of this PhD research was to gain a better understanding of the mouthfeel perception of black tea beverages using sensory and instrumental methods. In order to achieve this, a model system consisting of black tea instant powder, sugar, and fat (added as oil-in-water emulsion) was developed which allowed for the independent variation of the design factors (tea, sugar, fat). Firstly, a mouthfeel lexicon was developed using Quantitative Descriptive Analysis (QDA), and the impact of tea, sugar and fat on the perception of selected attributes was assessed by a trained panel (n=10) using a Design of Experiments approach to generate predictive polynomial models from a D-optimal design. Furthermore, the temporal perception of astringency and bitterness over multiple sips of tea was assessed by a trained panel(n=9) using Time Intensity measures. The effect of sweetness and viscosity from sucrose was decoupled using sweeteners and thickeners and the effect of sugar on astringency and bitterness perception was investigated in more detail. Once the mouthfeel perception of black tea beverages was explored, instrumental methods were deployed to reveal possible correlations between instrumental parameters and mouthfeel perception. Two instrumental methods were used: a force plate to measure the frictional and vibrational behaviour of tea samples, whilst an accelerometer was used to measure vibrations related to tongue movements using a technique called “acoustic tribology”. The results showed that besides astringency, fat-driven attributes, such as “thickness”, “slipperiness” and “mouth coating”, play a significant role in the mouthfeel perception of black tea beverages. Predictive polynomial models revealed the complex effects of key ingredients on the mouthfeel of tea. The results showed that tea was not a significant factor (p > 0.05) for the attributes “thickness” and “mouth coating”, whereas for all other attributes all three design factors affected perception significantly (p < 0.001). It was found that astringent and bitter intensity was reduced by the addition of fat and sugar in a similar fashion. However, when evaluated over time, it was found that astringency and bitterness had distinctly different temporal profiles in tea, illustrating the importance of temporal ratings. A build-up in astringency and bitterness intensity with increasing number of sips gave further insight into the perception of both attributes during tea consumption, representing a normal tea drinking behaviour. Furthermore, the data indicated that sugar reduces astringency due to its sweetness and not the rise in viscosity. The results from the force plate experiment showed that the addition of 5% fat significantly reduced the friction coefficient of tea samples (p < 0.0001) and that an addition of 10% fat did not reduce the friction coefficient further (p=0.97). It was also shown that the friction coefficient did not vary significantly between tea levels (p=0.324). The results also revealed that fat-driven attributes were negatively correlated to friction and that astringency was a complex precept, which was difficult to predict using instrumental methods, resulting in a poor correlation between friction coefficient and astringency (R2=0.16). The results from the “acoustic tribology” experiment indicated that mouthfeel was linked to measured vibrations caused by tongue movements. Furthermore, differences in tea composition resulted in different oral vibrations, and it was observed that for example “thick” is positively correlated to low frequency vibrations. The results of this research on the perceived mouthfeel of tea provide much needed insight into this key sensory property of tea. These results will be further useful for product developers interested in producing ready-todrink tea beverages where mouthfeel is likely to be a critical factor for consumer liking and commercial success.

Radial patterning of root hair bearing cells (trichoblasts) in Arabidopsis thaliana (arabidopsis) follows the type 3 root hair patterning mechanism whereby the radial positioning of trichoblasts is coordinated according to the position of underlying cortex cells (Pemberton et al. 2001). Epidermal cells which are positioned over the cleft between two underlying cortex cell files adopt trichoblast identity. Low phosphate stress in arabidopsis roots has been show to result in an increase in the number of cortex cell files from 8 to 12-16, which in turn results in an increase in the number of trichoblast position cells (Zhang et al. 2003, Cederholm and Benfey, 2015). However, little is known about what mechanisms control this proliferation in cortex tissue. Zhang et al. (2003) demonstrate that eir1 (an allele of pinformed2, pin2) mutants are deficient in this response to low P. This finding suggests that PIN2 (an efflux facilitator of directional auxin transport) is required to orchestrate proliferation of cortex cell files under low P. This implies that directional auxin transport and, ultimately, auxin response are required to enable proliferative divisions in cortex tissues under low P. First of all, a deeper anatomical understanding of the cortex cell file number response to low P was developed. That showed that the divisions which lead to the increase in cortex cell file number occur in the first cortex cell next to the quiescent centre. It was also found that the anatomical changes only significantly affect the number of cortex cell and trichoblast cell files, suggesting that it is a cortex specific response to increase radial root hair density. It was further discovered that the root is sensitive to up to 400μM P and responds with increased cortex cell file number within 24 hours. I also showed that this response is independent of iron concentration. It was next hypothesised that the role of directional auxin transport implied that other PIN and AUX/LAX genes may be required as well as activating AUXIN RESPONSE FACTORs (ARFs). These experiments revealed that no other PINs or ARFs play a crucial role in this response. However it was found that PIN2:GFP protein changes its subcellular localisation in response to low P, suggesting that changes in auxin flux direction is required. Based on this, it was hypothesised that PIN2 complementation in the cortex in pin2 would rescue the phenotype. This was indeed the case, but also for PIN2 under the AUX1 promoter, suggesting an additional role for PIN2 in the epidermis and root cap during the low P response. Mathematical modelling suggested that auxin flux out of the cortex, mediated by PIN2, would be required for the response. However, results testing this in vivo with the DII:VENUS auxin reporter were inconclusive. It was suspected that ground tissues patterning regulators, BIRD genes, may also play a role in tissue patterning under low P. jkd (jackdaw) and mgp (magpie) mutants did not show any strong phenotype on low P, but expression analysis using reporter lines and reverse transcription qPCR did suggest changes in regulation. It was then hypothesised that BIRD genes and PIN2 may interact, by crossing pin2 and jkd I found that the two interact to a small degree but it is not evident that they interact directly within the same pathway. It was concluded that this response to low P was controlled by a number of regulatory networks which definitely involves PIN2 mediated auxin transport with a contribution from BIRD genes.

Broiler chicken has been bred for meat production and is characterised with a very fast skeletal muscle growth rate, whilst the layer type have been bred for the production of eggs. This study aimed to understand how intense breeding programmes have developed commercial meat type chickens that have resulted in a fast muscle growth phenotype by a comparative analysis of gene expression between fast-growing broiler and slow-growing layer type chicken. Two chicken trials were carried out. In Trial 1 from fast-growing broiler Ross 308 genotype (R) fast-growing breast Pectoralis major (RPM) and slow-growing leg Peroneus tertius (RPT) muscles were collected at day 14, 36 and 42 post-hatch. In Trial 2, from fast-growing broiler Ross 308 (R) or slow growing layer Hy-Line (H) Pectoralis major (either RPM or HPM) and Peroneus tertius (either RPT or HPM) were collected at day 4, 14, 23, 35 and 42 post-hatch. In Trial 1, there was a muscle type x age interaction in muscle weights with the RPM having the highest value at day 43 (P < 0.001). This effect was also reflected in the protein content with RPM having more protein than RPT. However, RPT had a higher DNA content per unit tissue weight (muscle type x age interaction, P=0.003) with highest value seen at day 43. This suggests that muscle cells of RPM are bigger than those in RPT and that as a result they contain more protein, potentially reflecting RPM hypertrophy compare to RPT. In Trial 1, genes associated with glycolysis, GAPDH and -enolase, were significantly higher in the faster growing RPM (P < 0.05). Genes involved in serine biosynthesis PHGDH, PSAT and PSPH, as well as P70S6K involved in protein translation, also had a significantly higher expression in the RPM when compared to the RPT (P < 0.05) indicating a greater rate of protein synthesis in the faster growing RPM. Expression of genes associated with smaller muscles (myostatin) or protein degradation (calpastatin, the specific endogenous inhibitor of calpain proteinases) were not different between muscles. LIM domain proteins, CSRP3 and FHL2, had a higher expression in the slower growing muscle, indicating the possibility of those two genes being negative regulators of muscle growth, or may be involved in the upregulation of muscle regenerative physiological processes, as a result of the strain on the leg muscles induced by physical activities during locomotion. In Trial 2 there was a three-way interaction between genotype, muscle type and age in muscle growth (P < 0.001) with the highest value seen at day 42 in the RPM. There was a 3-way interaction between genotype, muscle type and age in the expression of -enolase (P=0.036) with the highest value seen at day 42 in the HPM. For -Enolase mRNA expression, there was a separate genotype and a muscle type effect (both P < 0.001) with HPM and HPT been higher than RPM and RPT, and PM muscles having a higher expression than PT muscles in both genotypes. When the serine synthesis pathway was examined, there was a genotype x age interaction (P=0.046) for PSAT and PSPH gene expression, with the highest expression at day 35 in both muscles of the Ross 308 genotype for the former, whilst the latter had the highest expression in the RPM at day 35. For ASNS mRNA expression there were significant genotype (P=0.004) and muscle type effects (P=0.014), with the Ross 308 genotype having the higher expression and the PM being greater than PT. For P70S6K mRNA expression, PM was higher than PT (P=0.012). For P70S6K total protein there was a significant muscle type x age interaction (P=0.049), RPM and HPM being higher than RPT and HPT at days 14 and 35. For calpain activity only at day 35 was there significant genotype x isoform interaction (P < 0.001), with the Hy-Line genotype having a higher micro and milli calpain activity than the Ross 308 genotypes, irrespective of muscle. For trypsin, chymotrypsin and caspase – like activities of the proteasome, there was a significant genotype effects (P < 0.001). In all three assays Ross 308 muscles had a higher activity when compared muscles from Hy-Line at both time points. For the ubiquitin ligases there was a borderline muscle type x time interaction (P=0.05) in the expression of MAFbx mRNA, with the RPT and HPT having a higher expression than the RPM and HPM respectively at day 14. For MuRF1 mRNA there was a significant genotype x age interaction (P < 0.001) with the HPM and HPT having a higher expression than the RPM and RPT at day 14 and 35. For CSRP3 there was a genotype x age interaction (P < 0.001), with the highest expression seen in the HPT at day 42. There was also muscle type effect within genotypes with the RPT and HPT having a higher expression than the RPM and HPM (P < 0.001). Trial 2 indicated that there is an increase in protein synthesis which leads to clear increase in protein accretion in faster growing chicken muscles, thereby supporting the wealth of literature detailing protein turnover in chicken skeletal muscles. There also appears to be an interaction between protein synthesis and degradation, however in most cases protein synthesis seems to be more dynamic and these changes seem to appear around day 35. The novel findings of this study were the observed increase in the expression genes that could limit the synthetic capacity of non-essential amino acid (non-EEA) in fast growing muscles.