School of Botany - Theses
Permanent URI for this collection
Search Results
1 - 10 of 52
-
ItemCharacterisation of cyanogenic glucoside synthesis in Eucalyptus(University of Melbourne, 2012)Cyanogenic glycosides are nitrogen-containing defensive metabolites, releasing toxic hydrogen cyanide following tissue disruption and contact with catabolic enzymes. Within the ecologically and commercially important genus Eucalyptus, approximately 4% of species are cyanogenic, possessing the phenylalanine-derived cyanogenic glucoside prunasin. Cyanogenic Eucalyptus species have been identified as an excellent experimental system in which to investigate defensive ontogenetic regulation as different species have shown variability in the onset and regulation of cyanogenic glucoside biosynthesis through plant development. The overall aim of this PhD thesis is to characterise cyanogenic glucoside synthesis in cyanogenic Eucalytpus species throughout leaf and plant ontogeny by investigating the synthesis, turnover and regulation of cyanogenic glucoside/(s) in E. yarraensis and E. camphora. Using various PCR techniques, four cytochrome P450 genes encoding CYP79 enzymes were detected and isolated from E. yarraensis: CYP79A34, CYP79A35, CYP79A36 and CYP79A37. The CYP79A34 gene was functionally cloned and expressed in Saccharomyces cerevisiae using the USER� cloning technique and a modified pYeDP60 vector, respectively. The recombinantly expressed CYP79A34 was shown to catalyze the conversion of L- phenylalanine into phenylacetaldoxime, verifying its involvement in the first step of prunasin biosynthesis. Using quantitative real-time PCR, CYP79A34 mRNA levels were found to be positively correlated with an increase in prunasin concentration during E. yarraensis ontogeny, suggesting that prunasin synthesis is regulated at the transcript level. Interestingly, CYP79A34 transcript levels also significantly increased in leaves subject to cold temperature and nitrogen application. To further investigate how leaf ontogeny affects the synthesis and possible turnover of prunasin, E. yarraensis coppice branches were pulse-labeled with 14C (via 14C02) and the incorporation of 14C into the nitrile constituent of the cyanogenic glucoside was measured in individual leaves over a 72 h period. Despite the high similarity observed in physical parameters and overall patterns of prunasin allocation across different leaf classes, a surprisingly high level of variability in 14C incorporation was encountered. Prunasin synthesis and turnover was observed in both expanding and fully expanded E. yarraensis leaves, although an accurate measure of turnover was not achieved. Interestingly, diurnal variation in specific activity suggests that the synthesis of prunasin may be influenced by light-dependent plant processes. Whole plant ontogenetic effects on cyanogenic glucoside synthesis were investigated in E. camphora, by identifying, tracking and comparing the concentration and composition of multiple cyanogenic glucosides throughout different ontogenetic stages. Using different analytical methods (LC-MS/MS, GC-MS and NMR) six phenylalanine-derived cyanogenic glucosides were identified, including three cyanogenic diglucosides characterised by unique linkage positions between the two glucose moieties. Two of these cyanogenic glucosides have not previously been reported and were named eucalyptosin B and eucalyptosin C. Interestingly, the different cyanogenic diglucosides could be distinguished by unique MS/MS fragmentation patterns. This knowledge should facilitate analysis and structural elucidation of cyanogenic diglucosides from other plant species. Quantitative and qualitative differences in total cyanogenic glucoside content were observed across different stages of whole plant and tissue ontogeny, as well as within different tissue types. Seedlings of E. camphora were found to produce only the cyanogenic monoglucoside prunasin, but "switch on" prunasin biosynthesis at different times. Once initiated, total cyanogenic glucoside concentration increased throughout plant ontogeny with cyanogenic diglucoside production initiated in saplings and reaching a maximum in flower buds of adult trees. Overall this work uncovered many novel aspects of cyanogenic glucoside synthesis in E. camphora and E. yarraensis, particularly in terms of ontogenetic regulation, and provides a solid platform to further investigate these species and their unique control over cyanogenic glucoside synthesis.
-
ItemGrape and botrytis cinerea proteases : characterization and utilization in winemaking(University of Melbourne, 2012)Grape berries contain high levels of pathogenesis-related (PR) proteins that persist through winemaking and cause commercially unacceptable heat-induced hazes if not removed before bottling. Nearly every commercial winery uses the clay cation exchanger bentonite to remove PR-proteins from wine, but the process results in wine losses, is laborious, and can negatively affect wine sensory properties. An alternative to protein removal by bentonite is hydrolysis of PR-proteins by proteases. However, PR-proteins are highly resistant to enzymatic proteolysis and no commercially viable enzymatic method has been discovered in at least 60 years of attempts. The purpose of the research presented in this thesis was to discover grape and Botrytis cinerea enzymes that could be utilized during winemaking to reduce or eliminate the need for bentonite by degrading PR?proteins. Grapes infected with B. cinerea have been observed previously to have lower levels of PR?proteins than healthy grapes or grapes infected with other pathogens. To investigate Botrytis proteases, knock-out mutants were screened on grape berries and against purified PR-proteins on agar plates. The mutants were no more inhibited by PR?proteins than the wild type, suggesting that either the aspartic proteases serve redundant functions, or their function in vivo is unrelated to grape PR-proteins. A Botrytis protease previously found to be the most highly secreted protein in liquid culture was heterologously produced in Pichia pastoris and used in winemaking trials. The Botrytis protease, BcAP8, significantly reduced the PR-proteins most prone to aggregation, chitinases, but did not significantly reduce,thaumatin?like proteins. BcAP8 could be an effective treatment for reducing bentonite in wines with unstable,chitinases. Using activity-based probes to enrich for proteases followed by mass spectrometry, grape papain-like cysteine proteases were characterized for the first time. One of the proteases is highly glycosylated, contains an extra disulfide bond and several other novel sequence features, and represents a new sub?family of plant proteases. Serine and nepenthesin?like aspartic grape proteases were also discovered by mass spectrometry. The cysteine protease activity of grape berries, most likely attributable to the newly discovered highly glycosylated protease, was associated with the insoluble material of grape juice and was highly heat stable. Those two attributes were exploited to develop a flash pasteurization method that effectively eliminated the need for bentonite in the tested juices.
-
ItemNo Preview AvailableEvaluating the ecological and operational basis of vegetation condition assessments(University of Melbourne, 2011)Globally, market based policies for biodiversity conservation are emerging that aim to achieve better and more cost efficient biodiversity outcomes than traditional command and control mechanisms. Investment accounting, reverse auctions and biodiversity offsets aim to achieve No Net Loss or Net Gain of biodiversity by incorporating biodiversity into the financial decisions of landholders. Market based policies require a proxy commodity for biodiversity that numerically express the capacity of a given site to support biodiversity, and Australia has pioneered the development of vegetation condition indices for this purpose. For every market based decision, there is a chance that the actual biodiversity outcome will be better or worse than expected. Risks of worse than expected outcomes may have significant financial implications and irreversible repercussions for biodiversity. Uncertainty in estimates of current and future vegetation condition may increase these risks. Despite this, market based policies have not been developed within a risk management framework and no systematic analysis of factors that may contribute to uncertainty in biodiversity valuation has previously been conducted. This thesis evaluates the nature of uncertainties in the ecological and operational basis of vegetation condition assessments, and considers the potential effects of uncertainty on the biodiversity outcomes of market based policies. Chapter 2 proposes a framework for identifying epistemic and linguistic uncertainties in current and predicted future estimates of vegetation condition. It contends that uncertainty may arise due to: i) conceptual uncertainty about the factors and processes that are causally related to the definition of biodiversity value; ii) quantitative uncertainty concerning misrepresentation of causal factors and processes by the mathematical constructs of the index; and iii) data uncertainty arising from errors in the input data. Specific occurrences of these uncertainties are discussed for two Australian vegetation condition indices, BioMetric and Habitat Hectares. It is concluded that the nature of uncertainties in estimates of current and future vegetation condition may increase the risk that any given decision will fail to achieve No Net Loss of biodiversity. Chapters 3 to 5 evaluate sources of uncertainty in current estimates of vegetation condition. Chapter 3 uses empirical biodiversity data to evaluate whether, on average, sites with higher vegetation condition scores support a greater diversity of native species than sites with lower scores. The strength of relationships between species composition and condition scores was primarily driven by plant species, which is included as a predictor variable in both BioMetric and Habitat Hectares. Vegetation condition was very poorly correlated with the richness and diversity of native butterflies and native ants, with the possible exception of very low scoring sites for butterflies. The results suggest that patterns in species composition were driven by factors other than the predictor variables included in the indices, and therefore that conceptual uncertainty may be a substantial cause of poor performance of vegetation condition indices. Chapter 4 empirically investigates the magnitude of uncertainty in estimates of current vegetation condition due to observer error in field estimates of predictor variables. Average coefficients of variation in total vegetation condition scores amongst ten independent observers were 15-18%. All observers estimated vegetation condition scores that were substantially different from the group mean on at least some sites. The results indicate that measurement error in field estimates of site attributes may cause vegetation condition to be under- or over-estimated on all but highly degraded sites. Chapter 5 examines the sensitivity of vegetation condition indices to observer error via simulated scenarios. Larger observer errors caused less precision and greater bias in total scores, although compensatory errors generally led to smaller coefficients of variation in total scores than predictor variables. Unexpectedly, unbiased observer errors in the predictor variables underestimated the true vegetation condition of most sites. The underestimation effect was more significant for sites with intact woody or herbaceous features, depending on the index used. Chapters 4 and 5 highlight the fact that quantitative aspects of vegetation condition indices are as important as the spread and bias of raw observer estimates in determining accuracy of vegetation condition assessments. These operational limitations of vegetation condition indices may significantly increase the risk of worse than expected biodiversity outcomes. Chapter 6 examines the implications of uncertainty in estimates of predicted future vegetation condition for achieving No Net Loss in biodiversity offsets. Expert models of change in 12 vegetation attributes were used to simulate change over time for five states of a grassy woodland ecosystem under six (gain and loss) management scenarios. The quantity of gains that would be required to compensate for each loss scenario (i.e. the offset ratio) was calculated using one method that ignored uncertainty and another that accounted for it. Uncertainty increased the offset ratio up to 1400% where the magnitude of both gains and losses were uncertain, but increased the ratio only marginally where minimal losses were incurred on highly degraded site types and offset with maximal gains. Both vegetation condition indices predicted greatest gains in woody vegetation attributes, which may result in landscape wide decline in herbaceous components of vegetation and their associated biodiversity if policies aim to maximise gains in vegetation condition. The results of this research indicate that aspects of the conceptual, quantitative and data models that underpin biodiversity valuation indices are likely to cause estimates of current and future biodiversity value to be unreliable. Outcomes of market based biodiversity conservation policies would greatly benefit from more rigorous procedures for developing and evaluating biodiversity valuation indices, which systematically identify and minimise uncertainties in the underpinning models. Information about uncertainty should be used in a risk-based approach to decision making in market based policies, whereby only low risk decisions are approved. It will be a shame if the pioneering work Australia has conducted in developing market based instruments for biodiversity conservation, and associated vegetation condition indices, is not used to fully evaluate the potential for overcoming market externalities and achieving optimal biodiversity outcomes for society.
-
ItemThe function and evolution of the dinoflagellate mitochondrion(University of Melbourne, 2010)The mitochondrion is a highly evolved, indispensable organelle found in all eukaryotes. This compartment has undergone metabolic and functional changes as cell lineages have diversified and specialised throughout evolution. Alveolata is a well-defined group of single-celled eukaryotes that encompasses related Phyla with extremely diverse lifestyles. Most alveolates belong to one of three main subgroups: predatory ciliates, endoparasitic apicomplexans, and heterotrophic or autotrophic dinoflagellates. Little is known about the biology of the dinoflagellate mitochondrion and studying this compartment offers an opportunity to examine organelle evolution within the alveolate lineage. In this study, I have used the dinoflagellate Karlodinium micrum, as a model to examine mitochondrial evolution. I have investigated the evolution of (1) genes, (2) biochemical functions and (3) protein targeting mechanisms of this organelle. Organelles can replace and gain genes by endosymbiotic gene transfer (EGT, genes derived from an endosymbiont) and lateral gene transfer (LOT, genes derived from an external source). Dinoflagellates have shown a unique propensity to replace their plastids with plastids of other algae during evolution and K. micrum represents a dinoflagellate lineage that has replaced its ancestral plastid with an endosymbiont derived from a haptophyte. In this case, haptophyte endosymbiont plastid genes are located in the dinoflagellate nucleus, providing evidence of EGT in this system. I have assessed if the mitochondrial proteome of K. micrum has been remodelled by EGT and/or LTG. Genes encoding mitochondrial proteins have been identified from a K. micrum expressed sequence tag library and their evolutionary origins inferred by phylogenetics. Several mitochondrial genes are derived from an external source but none originate from the haptophyte endosymbiont, indicating that the K. micrum mitochondrial proteome has been minimally impacted by this endosymbiotic event, but is nevertheless genetically dynamic. Plasmodium falciparum, the disease agent that causes malaria, is a member of Apicomplexa. This protist has a mitochondrion that has been described as metabolically reduced compared to canonical mitochondria, change that has been attributed to the parasitic lifestyle this organism leads. In this thesis, I test whether or not perceived reduction in apicomplexan mitochondrial metabolism is a result of parasitism. A putative metabolic map of the dinoflagellate mitochondrion has been constructed and compared to what is currently known about the mitochondrial biochemistry of closely related apicomplexan parasites and a free living basal alveolate, the ciliate Tetrahymena thermophila. This is the first report of a broad analysis of the mitochondrial metabolism of a dinoflagellate. The mitochondrion of K. micrum shows broad metabolic conservation, having retained pathways implicated in ATP generation by oxidative phosphorylation. Several changes in the metabolism of the P. falciparum mitochondrion were also observed in K. micrum and/or T. thermophila, suggesting that these modifications are not due to parasitism. The presence of most components of the tricarboxylic acid cycle, in addition to what is most likely a functional electron transport chain and ATP synthase complex in both dinoflagellates and P. falciparum indicates that the mitochondrion of the Plasmodium parasite is probably implicated in ATP generation by oxidative phosphorylation. The diversification of dinoflagellates has been accompanied by considerable changes in plastid protein targeting signals, but it is unclear whether or not mitochondrial protein targeting in this lineage has also been modified. In the final experimental chapter of this thesis I have assessed the conservation of mitochondrial protein import mechanisms in dinoflagellates. Genes for K. micrum mitochondrial proteins have been analysed for mitochondrial protein targeting signals using bioinformatic tools, and the function of these signals has been tested using the reporter molecule green fluorescent protein (GFP) and a heterologous yeast expression system. Amino-terminal and internal mitochondrial targeting signals of K. micrum mitochondrial precursors are sufficiently conserved for recognition and import into yeast mitochondria, indicating that dinoflagellate mitochondrial protein targeting signals have been highly conserved, since early in eukaryotic diversification. Overall, my investigations of the dinoflagellate mitochondrion, and this broad comparative analysis of alveolate mitochondria has shown that aspects of mitochondrial biology (ie. mitochondrial gene compliment, mitochondrial biochemistry and function, and mitochondrial protein targeting) have evolved differently during the diversification of alveolates. Alveolate mitochondria are genetically flexible, having experienced gene gains and gene losses. This variation is not always accompanied by functional divergence, and does not necessarily reflect the lifestyle/nutritional requirements of the host. Thus while the mitochondrion is clearly an innovative compartment, its evolutionary behaviour cannot be characterised based on any one aspect of its cell biology alone.
-
ItemThe bryophyte flora of Lord Howe Island: taxonomy, diversity and biogeography( 2018)Before this study the known Lord Howe Island bryophyte flora (mosses, liverworts and hornworts) totalled 173 species, consisting of 131 mosses, 40 liverworts and 2 hornworts. For this study I conducted one month of field studies on the island, during which I collected more than 650 specimens, and also studied collections from the island held in Australian herbaria and other collections available in overseas herbaria. Fourteen moss, 32 liverwort and 1 hornwort species are newly reported from Lord Howe Island, including one liverwort new to science. A further 2 moss and 2 liverwort varieties are also new to the island. Twenty-eight moss and 11 liverwort species are discounted from the island’s flora, as well as 3 moss varieties and 1 liverwort variety. As a result, the known bryophyte flora now totals 178 species, consisting of 117 moss species (122 taxa), 58 liverwort species (60 taxa) and 3 hornwort species. These totals exclude 5 moss and 2 liverwort species whose taxonomic status or presence on the island is considered uncertain. One liverwort variety, Heteroscyphus echinellus var. echinellus, is new to Australia. Fourteen bryophyte species and one variety are endemic to the island. Spiridens muelleri, previously thought to be the same as S. vieillardii from New Caledonia, is shown to be a separate species endemic to Lord Howe Island. Chiloscyphus howeanus is also shown to be a legitimate species endemic to the island. Cololejeunea elizabethae is described as a new species, also endemic to the island. Trachyloma wattsii, considered to be endemic to Lord Howe Island, is supported by a molecular analysis as a legitimate species most closely allied to T. planifolium. Confusion about the correct identities of the two Ptychomitrium species on the island is resolved through a revision of the genus for Australia. A previously unrecorded morphological character of Atrichum androgynum is described from a study of Lord Howe Island plants, and a molecular analysis shows that South American plants previously ascribed to A. androgynum do not belong to that species. Hypnodendron vitiense is shown to be paraphyletic, but not as circumscribed by Touw (1971). The Lord Howe Island plants appear to belong to a morphologically cryptic species distinct from H. vitiense s.str, and substantial genetic variation within H. vitiense subsp. australe as currently circumscribed suggests that it might include more than one taxon. Other molecular investigations clarify the relationship between Lord Howe Island populations and mainland Australian populations of a number of moss species. An original and novel investigation of the potential modes of transport of bryophyte propagules to and from the island is made, and a hypothesis is formed about the origins of its bryophyte flora and the biogeographic relationships to the Australian land mass and other western Pacific islands, including New Zealand and New Caledonia. The nearest region of the Australian mainland is shown to be the most likely origin of most of the island’s bryoflora, with the injection of propagules into the high-level jet stream by storms the most likely dispersal mechanism. The presence of numerous otherwise tropical species on the island is probably a result of dispersal by tropical cyclones moving into the Pacific from north-eastern Australia. Migratory birds are shown to be another potential vector for bryophyte dispersal to the island.
-
ItemRegulation of mucilage production in the Arabidopsis seed coat through MADS-box TF family members( 2018)During Arabidopsis seed development, the epidermal cells of the seed coat produce large quantities of pectinaceous mucilage polysaccharides. Following imbition, mucilage is released from these cells and forms a thick protective layer around the germinating seed that can be visualised with Ruthenium Red staining. The seed coat epidermis is therefore an attractive system to investigate the production of cell wall polysaccharides, particularly pectin. Previously, mutations in the transcriptional regulator LEUNIG_HOMOLOG (LUH) have been shown to be associated with a mucilage extrusion defect that is caused by reduced expression of the β-galactosidase MUCILAGE MODIFIED2 (MUM2). To better understand the role of LUH in regulating MUM2, RNA-Seq analysis was performed on whole seeds and identified genes that were differentially expressed in luh mutants. This transcriptomic analysis not only revealed elevated expression of transcription factors with a known role in epidermal cell differentiation, but also several MADS-box transcription factors that perform roles in floral development and silique shattering such SHATTERPROOF1 (SHP1) and SHP2. This thesis presents evidence that SHP1, SHP2 and SEPALLATA3 (SEP3) are a new class of regulators involved in mucilage polysaccharides modification, as loss of their activity result in mucilage extrusion defects. These mucilage defects are enhanced in shp1 shp2 double mutants and correlate with reduced MUM2 expression. While carbohydrate analysis data failed to show an expected increase in galactose residues attached to pectin in shp1 shp2 and sep3 mutants, the decrease of MUM2 in these mutants is shown to contribute to their mucilage defects as partial rescue of mucilage defects in shp1 shp2 and sep3 was achieved with rescue of MUM2 expression. This is further supported by the use of SRDX repressor motif fusions to SHP2 that not only produced mucilage extrusion defects when introduced into Col-0, but also were correlated with the level of reduced MUM2 levels. Finally, the positive regulation of MUM2 by MADS-box TFs was established through a combination of CArG-box motif mutations in the regulatory sequences of MUM2, demonstrating the requirement of MADS-box TFs for proper MUM2 seed coat expression. This was further supported by luciferase-based transactivation assays, which produced increased MUM2 promoter activity on the addition of SHP1, SHP2 and SEP3. In addition, this positive regulation of MUM2 by SHP1, SHP2 and SEP3 was shown to occur in a complex with LUH, mediated by the co-regulator SEUSS (SEU), in protein-protein interaction assays such as yeast 2-hybrid (Y2H), Y3H, and bimolecular fluorescence complementation (BiFC). Overall, this study provides novel, clear evidence that MADS-box TFs can directly bind to MUM2 and mediate its activation with LUH. Future work includes expanding on and integrating MADS-box TFs into a working model that describes the regulatory pathways controlling mucilage polysaccharides production in the Arabidopsis seed coat.
-
ItemUnderstanding the mechanism of the synthesis of β-D-(1,3;1,4)-glucans, or mixed linkage glucans (MLGs), by the cellulose synthase like (CSL) F6 protein( 2018)The molecular mechanism by which mixed linkage, (1,3;1,4)-β-D-glucan, a non-cellulosic polysaccharide, is synthesised by cellulose synthase-like (CSL) F6 was explored. CSLF6 was shown to be crucial for development in the proposed model grass Brachypodium distachyon through analysis of point mutants. In silico homology modelling of related β-glucan synthases was used to predict domains influencing specificity, informing targeted mutagenesis. CSLF6 from Lolium multiflorum expressed in Escherichia coli was found to produce (1,3;1,4)-β-D-glucan similar to native sources in vitro, allowing biochemical characterisation.
-
ItemMolecular systematics of siphonous green Algae (Bryopsidales, Chlorophyta)( 2018)The evolutionary history of the siphonous green algae (Bryopsidales, Chlorophyta) was investigated using a combination of molecular techniques and phylogenetic inference methods. Analyses of chloroplast genomes of the order revealed the high variability of genome architecture and intron content. Proliferation of nonstandard genes associated with mobile functions (i.e. reverse transcriptase/intron maturase, integrases, etc.) was also observed. Evolutionary relationships of families in the order were investigated by increasing taxon sampling and using chloroplast genome data. The chloroplast phylogenies provided good support for the suborders and most families. Several early branching lineages were also inferred in the Bryopsidineae and Halimedineae. A new classification scheme was proposed for the order, which included the following: establishment of the family Pseudobryopsidaceae fam. nov.; merger of the families Pseudocodiaceae, Rhipiliaceae, and Udoteaceae into a broadly circumscribed Halimedaceae and establishment of tribes for the different lineages found therein; finally, the deep-water genus Johnson-sea-linkia, currently placed in Rhipiliopsis, was reinstated based on the chloroplast phylogenies. Plastid (tufA) and nuclear markers (HSP90) and morphological observations were employed to delimit the Halimeda species found in Western Australia. This facilitated the recognition of Halimeda cuneata and the reinstatement of Halimeda versatilis. Investigation on morphological complexity revealed that simple uniaxial thalli was the ancestral state of the siphonous green algae and was maintained throughout their early evolution. Complex multiaxial thalli evolved afterwards on independent occasions.
-
ItemPhylogeny of Eremophila and tribe Myoporeae (Scrophulariaceae)( 2018)Myoporeae is one of eight tribes recognised in the large, cosmopolitan plant family Scrophulariaceae sensu stricto. Tribe Myoporeae contains seven genera, four of which are endemic to Australia (Calamphoreus, Diocirea, Eremophila, Glycocystis), and the remaining three are distributed in the Caribbean (Bontia); Southern China and Japan (Pentacoelium); and throughout Australia, islands of the Pacific, Hawaii and Mauritius (Myoporum). The largest of these genera, Eremophila, contains over 220 species and is a major component of Australia’s largest biome, the Eremean (arid) zone. A monograph of the tribe was completed just over a decade ago (Chinnock, 2007), which provided an extensive and comprehensive taxonomic framework from which to explore the relationships and evolutionary history of the group. The first phylogenetic study of the Myoporeae (Kelchner, 2003) used two chloroplast markers to better understand generic and species level relationships in the tribe, however, due to a lack of phylogenetic resolution, the results were inconclusive. The aim of this thesis was to generate a comprehensive molecular phylogeny of tribe Myoporeae, utilising the capabilities of high throughput sequencing (HTS) technology. A genome skimming approach was implemented using a custom in-house method of library preparation, to allow for inclusion of the large number of samples required for the study. All three plant genomes (chloroplast, nuclear, mitochondrial) were represented using the genome skimming method, allowing for comparisons to be made between phylogenetic analyses of each genomic dataset. Entire chloroplast genomes (cpDNA) were assembled for 317 taxa, resulting in a well resolved and highly supported phylogeny (see Chapter Three). All allied genera were found to be nested in a paraphyletic Eremophila, with high levels of support. Chinnock’s (2007) sectional classification of Eremophila was only partially supported, with many of the 25 sections scattered throughout numerous clades. For the majority of species included with more than a single sample, a lack of monophyly was observed, which is largely attributed to the effects of introgressive hybridization, incomplete lineage sorting, and/or inappropriate species boundaries. In Chapter Four the entire nuclear ribosomal cistron (nrDNA) was assembled for 355 taxa, then analysed to a produce a moderately supported phylogeny. This phylogeny was largely congruent with the morphology-based taxonomy of the group, though differed markedly from the cpDNA phylogeny of Chapter Three. From a generic perspective, all allied genera were still nested in a polyphyletic Eremophila, while Chinnock’s (2007) sectional classification was better supported by monophyletic lineages (though still in need of revision). An increase in species rank monophyly was also observed relative to the cpDNA analysis, indicating that at least for some species, introgressive hybridization is likely to impact the chloroplast phylogenetic signal. In Chapter Five the mitochondrial genome (mtDNA) was explored, and five regions selected for analysis across a subset of 31 taxa in Myoporeae. The size and prevalence of structural rearrangement within the tribe meant assembly of entire mitogenome(s) was not feasible. Regions selected for analysis displayed low levels of variation, allowing for a moderately well-resolved phylogeny, mostly congruent with the nuclear ribosomal phylogeny of Chapter Four. Overall, construction and comparison of the three genomes in this study allowed for robust interpretation and increased understanding of the complexity in the evolutionary history and phylogenetic relationships of taxa in tribe Myoporeae. Taxonomic revision is needed at generic and sectional levels; however these changes will not be undertaken until further nuclear sequence data allows the relationships of taxa at the basal nodes of the nuclear phylogeny to be resolved. Aside from future taxonomic work, it is anticipated that this study will inform new research on the tribe Myoporeae, including the chemistry of Myoporeae (for pharmacological application); the study of plant:insect interactions between Myoporeae and members of the insect family Miridae; and biogeographic study of Australia’s Eremean zone.
-
ItemMolecular mechanism of the assembly of (1,3;1,4)-β-D-glucan in Italian ryegrass (Lolium multiflorum) suspension-cultured cells( 2018)(1,3;1,4)-β-D-Glucan or mixed linkage β-glucan (MLG) is a major type of non-cellulosic polysaccharide of grass cell walls and is synthesised by the CELLULOSE SYNTHASE-LIKE (CSL) F, H and J proteins with CSLF6 being the predominant isoform. MLG consists of a backbone of unsubstituted and unbranched (1,4)-linked and (1,3)-linked β-glucosyl residues. In cereals, the (1,3)-β-Glc residue always occurs singly, mostly between either two or three (~90%) adjacent (1,4)-β-Glc residues generating characteristic DP3 and DP4 oligosaccharides after digestion with lichenase, an endo-β-glucanase that specifically cleaves MLG. The (1,3)-β-linkages cause molecular “kinks” and their irregular distribution leads to a reduction in molecular alignment between glucan chains, resulting in increased solubility compared to cellulose. The heterogeneity in fine structure of MLG is a key determinant of its solubility in the gastrointestinal tract that imparts its beneficial effects in lowering the risk of several diet-related diseases. The main objective of this project was to understand how grasses make MLG and regulate its fine structure with a view that this knowledge will enable targeted manipulation of MLG to enhance the properties of cereal grains for improved human nutrition and other agro-industrial applications. At the start of this project, it was known that CSLF/H/J proteins were the catalytic subunits of the MLG synthase (MLGS), however, the exact biochemical activity of these proteins had not yet been shown experimentally. It was still unclear how many proteins were required in the MLG biosynthesis and assembly pathway and how they were being regulated. To answer these questions, biochemical, cell biological and molecular approaches were adopted, and primarily applied to our model system, suspension-cultured cells (SCCs) of the grass species Lolium multiflorum (Italian ryegrass). Chapter 1 is a literature review that provides a brief overview of the polysaccharide composition of plant cell walls, the GT2 family of plant polysaccharide synthases to which the CSLF/H/J proteins belong, and ends with a mechanistic model of MLG biosynthesis proposed by Wilson et al (2015), arising, in part, from this study. Chapter 2 is a detailed description of materials and methods used in this project. In Chapter 3, a number of methods were optimised to enable the biochemical characterisation and sub-cellular location of the CSLF6 and CSLH1 proteins to be studied. After evaluating a number of detergents n-dodecyl β-D-maltoside (DDM) was found to effectively solubilise and importantly retain MLGS activity and was also the detergent of choice due to its compatibility with downstream mass spectrometry (MS)-based applications. It was observed that MLGS activity is favoured under alkaline conditions (pH ~9.0) in the presence of both Mg2+ and Ca2+ with the majority of counts incorporated into MLG rather than callose, a (1,3)-β-glucan. Fractionation of microsomal membranes (MM) using sucrose density gradient centrifugation and PEG/Dextran two-phase partitioning showed that most CSLF6 protein is found at the plasma membrane (PM) and not intracellularly, in contrast to CSLH1. The finding of different sub-cellular distributions for these proteins was substantiated by immuno-electron microscopy (Wilson et al., 2015). Western blots of non-denaturing gels on which detergent-solubilised membrane samples were separated revealed that CSLF6 forms higher-order oligomers and an in-gel MLGS activity assay demonstrated that the dimer form of CSLF6 is active. Pre-treatment of membrane proteins with either dithiothreitol (DTT) or hydrogen peroxide (H2O2) favoured CSLF6 monomer or oligomer formation, respectively, indicating that the interactions between CSLF6 polypeptides is likely regulated by redox via the formation of intermolecular disulphide bonds between cysteine (Cys) residues. Also in this chapter, it was discovered that carbonate washing of the PM-enriched (PM-E) fraction increased MLGS specific activity. This was used to good effect in co-immunoprecipitation (co-IP) experiments using the CSLF6-specific antibody to remove contaminant proteins. The combined proteomic analyses of co-IPed proteins and gene network co-expression analyses identified several proteins that likely interact with CSLF6 and impact MLG biosynthesis and/or regulation in some manner: a KORRIGAN-type cellulase (GH9); nucleotide sugar interconverting enzymes relating to UDP-Xyl formation; trafficking-related proteins including RABH1b and ARF-type GTPases, Sec61 and an EPSIN2 protein; cytoskeletal-associated proteins such as a WEB family protein (DUF827); CC1, a CK1 kinase; multiple transferases including an S-adenosyl-L-methionine dependent methyltransferase (DUF248); an O-Fuc transferase-domain containing protein; a CSLD; and multiple unknown proteins. Chapter 4 outlines findings relating to the post-translational modifications (PTMs) of CSLF6 that may be involved in its regulation. A proteomics workflow utilising DTT, IAA and NEM was used to pinpoint which Cys residues were involved in the formation of disulphide bonds and which Cys residues possessed free thiol groups. Only four Cys residues in the LmCSLF6 sequence were detected by MS. C-496 and C-517 were shown to be labelled with IAA, i.e. involved in disulphide bonding while C-246 and C-440 each possessed a free thiol group. A homology model of HvCSLF6 reported by (Schwerdt et al., 2015) and based on the coordinates for a bacterial cellulose synthase (BcsA) (Morgan et al., 2013) offered the possibility of speculating which Cys residues are in close proximity to C-496 and C-517 to potentially form intramolecular disulphide bonds with C-217 or C-643 appearing the most probable. CSLF6 was also identified to be a phospho-protein with three N-terminal phosphorylation sites, Ser-55, Thr-75 and Ser-78, at a similar position within the hypervariable region (HVR1) in which CesA proteins are phosphorylated a short distance downstream of the zinc-finger domain. In addition to observing CSLF6 phospho-sites, many other phospho-proteins were detected including some polysaccharide synthases previously identified to be phosphorylated in other studies, as well as some of the anti-CSLF6 co-IPed proteins identified in Chapter 3 (RABH1b, ARF, EPSIN2, CC1, and an unknown protein). Chapter 5 describes the functional testing of some of the Cys residues and a conserved Thr residue in the gating-loop of CSLF6 using the Nicotiana benthamiana leaf transient expression system. The four Cys residues at the N-terminus of HvCSLF6 were removed by truncation and were shown not be involved in either MLG synthesis or CSLF6 protein dimerisation. Binary vector constructs utilising either the 35S or UBQ10 promoter were tested and it was found that the UBQ10 promoter was preferable as it provided more consistent protein expression of CSLF6 variants. No effect on MLG synthesis or CSLF6 dimer formation was observed when using the UBQ10 promoter to drive the expression of C246S, C496S and C517S mutants. However, T-801 which is highly conserved within all CSLF and CesA proteins was found to be essential for MLG synthesis as no MLG accumulated in infiltrated leaves when it was substituted with Gly. The results of this Chapter indicated that other Cys residues are required to be mutated to elucidate which residues are involved in CSLF6 dimer formation. Chapter 6 briefly summarises the experimental chapters and proposes directions for future experiments in order to fulfil the aims of the research project. In addition, an updated working model of MLG biosynthesis that draws upon the evidence from the experimental chapters and published literature is presented to provoke/guide future experimental approaches.