Doctor of Philosophy in Botany (PhD) 
Cellulose biosynthesis in Arabidopsis seed coat epidermal cells
Agriculture and Agri-food Canada
Molecular Genetic Analysis of Plant Cell Walls Using Seed Coat Mucilage, Targeted Induced Local Lesions in Genomes (TILLING), Seed Coat Differentiation, for details see Haughn lab website.
G+PS regularly provides virtual sessions that focus on admission requirements and procedures and tips how to improve your application.
The seed coat epidermal cells of Arabidopsis thaliana synthesize and secrete large quantities of mucilage, a specialized secondary cell wall which is released from mature seeds upon hydration. The goal of this thesis was to study the biological and biochemical consequences of modifying the structure of cell wall carbohydrates using cell wall degrading enzymes as a tool. To develop seed mucilage as a model, promoters that would only drive expression of genes encoding cell wall degrading enzymes within the mucilage-producing epidermal cells were utilized. In order to demonstrate their ability to promote gene expression sufficiently to modify mucilage, the seed coat-specific promoters from the following genes; TESTA-ABUNDANT2 (TBA2), PEROXIDASE36 (PER36), and MUCILAGE-MODIFIED4 (MUM4), were fused upstream of MUCILAGE-MODIFIED2 (MUM2), a gene that encodes a known mucilage modifying enzyme, which were transformed into a mum2 mutant background (Chapter 3). All three promoters were shown to be able to drive sufficient expression of MUM2, in a spatial and temporal pattern to manipulate mucilage composition and complement the mum2 phenotypes. The strongest of the three promoters, TBA2p, was then used to examine the ability of three previously uncharacterized MUM2 homologs; BGAL11, BGAL16 and BGAL17, to complement the mum2 extrusion and cell wall compositional phenotypes (Chapter 3). It was found that cytological and biochemical complementation of mum2 varied and correlated with the amino acid sequence similarity of the homologous gene products to MUM2. Consistent with the fact that the pectin rhamnogalacturonan-I (RG-I) is the major component of seed mucilage, transgenic plants expressing genes encoding RG-I degrading enzymes driven by TBA2p produced seeds with very little mucilage and greatly reduced levels of sugars comprising RG-I (Chapter 4). Unexpectedly, modifications to a minor component of seed mucilage, homogalacturonan (HG), using HG-degrading enzymes driven by TBA2p, resulted in reduced cell adhesion, no mucilage pocket formation and cell death in developing seed coat epidermal cells (Chapter 5), highlighting HG’s essential role in the cell wall. In summary, the data represented in this thesis has demonstrated the feasibility of manipulating cell wall carbohydrate composition using a genetic engineering approach to explore the relationships between structure and function within the plant cell wall.
Cell-to-cell adhesion is essential for establishment of multicellularity. In plants, cell adhesion ismediated through a middle lamella composed primarily of pectic polysaccharides, but themolecular interactions that promote and regulate such adhesion are not fully understood. InChapter 3, Arabidopsis seed coat mucilage was used as a model system to investigateinteractions between cell wall carbohydrates. Using a forward-genetic approach, we havediscovered a gene encoding a putative galactose oxidase, RUBY PARTICLES IN MUCILAGE(RUBY), that is required for cell-to-cell adhesion in the seed coat epidermis. Cellular andenzymatic analyses support the hypothesis that RUBY facilitates cross-links in the cell walls viathe side-chains of rhamnogalacturonan I (RG-I), a constituent of pectin. These results (Chapter3) provide genetic evidence for oxidative cross-linking in cell walls and assigns a biologicalfunction to the galactose/glyoxal oxidase family of enzymes.To better understand functions of galactose oxidases in plants, Arabidopsis homologues ofRUBY, GALACTOSE OXIDASE-LIKE (GOXL) genes, were studied (Chapter 4). The expressionpatterns of these seven genes suggest that all of the members have functions in specialisedtissues. Phylogenetic analyses suggest that the GOXL family likely has two pairs of GOXLparalogues, GOXL1 and GOXL6, and RUBY and GOXL3. Surprisingly, RUBY and GOXL3 areexpressed in different tissues, whereas GOXL1 and GOXL6 have similar expression patterns,suggesting genetic redundancy. Functional complementation of ruby mutant and qualitativeenzyme assays indicate that GOXL1, GOXL3 and GOXL6 are putative galactose oxidases.Plants with mutations in these genes, apart from RUBY, have no obvious phenotypes. Whenmutations were introduced in both GOXL1 and GOXL6, a collapsed pollen phenotype appeared,indicating that these genes may be redundant. Pollen collapse occurs at the anthesis, when pollengrains are desiccating, suggesting possible roles in pollen wall folding during controlled pollendehydration (harmomegathy). Further genetic analysis is required to confirm that these mutationsare indeed linked to the pollen phenotype. Taken together, putative Arabidopsis galactoseoxidases seem to have specialised roles in tissues that may require mechanical support.
Compositions of plant cell walls are important for morphogenesis and cellular function. Cell wall components are often distributed asymmetrically in the cell wall, but the mechanism for the polar distribution remains unknown. In my research, I used Arabidopsis seed coat epidermal cells as a model system because they deposit large amounts of pectin-rich mucilage in a polar manner to the outer periclinal side of the cell forming a large apoplastic pocket. MUM2 is a β-galactosidase modifying pectin in the mucilage. Using an engineered version of MUM2 fused to a Citrine fluorescent protein (Citrine), distribution pattern of MUM2 in the epidermal cell was determined. MUM2-Citrine is found to preferentially accumulate in the mucilage pocket concomitantly with pectin deposition. The amino acid sequence of MUM2 is not required for the secretion to the pocket. Rather, the polar distribution of MUM2-Citrine is caused by a rearrangement of the secretory pathway that appears to target all secretion to the outer periclinal side of the cell. At the end of mucilage synthesis, the fluorescence of MUM2-Citrine rapidly disappears from the mucilage pocket. The results of western blot analyses shows that the amount of MUM2-Citrine decreases, suggesting that the disappearance of MUM2-Citrine signal is due to protein degradation. Loss-of-function mutations in the genes encoding ASPG1 and RD21A, two proteases that have been detected in mucilage from mature seeds, resulted in a delay of MUM2-Citrine degradation. Also, RD21A, tagged with red fluorescent protein (RFP), accumulated in the vacuole during MUM2 secretion and was translocated to the mucilage pocket at the end of secretion. Taken together, these data suggest that ASPG1 and RD21A are responsible for MUM2 degradation. Mucilage properties were changed in the protease mutants, suggesting that the regulation of distribution of cell wall-modifying enzymes by proteases plays an important role in determining cell wall properties.It has been found that PER36, a peroxidase required for mucilage extrusion, was deposited in a pattern distinct from other mucilage proteins in seed coats. I showed the amino acid sequences of PER36 was required for the unique distribution pattern, suggesting that an unknown distribution mechanism may be involved.
Tomato ringspot virus (ToRSV) replicates in large protein complexes that are associated with modified endoplasmic reticulum membranes. The ToRSV RNA-dependent RNA polymerase (Pol) and integral membrane protein NTB-VPg are essential components of these complexes. Membrane-associated modifications of NTB-VPg (N-glycosylation and a putative signal peptidase cleavage) were previously observed in vitro but were not well characterized. Two forms of the polymerase were detected in infected plants: the full-length Pol, which accumulates at low concentration and the VPg-Pro-Pol' polyprotein, which includes a 15kDa C-terminal truncation of Pol and accumulates to higher levels. My specific objectives were to characterize the signal peptidase cleavage in NTB-VPg and investigate the stability and function of various forms of the polymerase. Using in vitro translation assays and plant transient expression assays, I detected signal peptidase processing in the NTB-VPg of three ToRSV isolates. Using site-directed mutagenesis, I mapped a suboptimal GAAGG cleavage site (Rasp2 isolate) and identified key amino acids that regulate the efficiency of cleavage. Compared to typical signal peptides, the NTB-VPg sequence has an unusually long distance between the end of the hydrophobic region and the cleavage site, indicating that it adopts a unique topology in the membrane. This is the first detailed characterization of signal peptidase cleavage of a plant virus replication protein. This cleavage may alter the conformation of NTB-VPg in the membrane and influence the architecture of the replication complexes. Using agroinfiltration assays, I show that the full-length Pol and VPg-Pro-Pol are unstable when ectopically expressed in N. benthamiana. Truncation of the C-terminal 15 kDa from Pol or VPg-Pro-Pol increased their stability in plants, which is consistent with the accumulation of VPg-Pro-Pol' in infected plants. In spite of repeated attempts, I was unable to establish an in vitro assays to compare the activity of Pol and VPg-Pro-Pol', possibly because an essential plant factor is missing. The instability of VPg-Pro-Pol and Pol may regulate the rate of virus replication, allowing the virus to keep its genome integrity and reducing the chance of being recognized by host defense responses.
Homogalacturonan pectin domains are synthesized in a highly methyl esterified form and can be de-methyl esterified by the cell wall enzyme Pectin Methyl Esterase (PME). The prevalent model for PME mode of action indicates that when PMEs act on a stretch of adjacent galacturonic acid glycosides, they may strengthen the cell wall but when PMEs act on non-adjacent galacturonic acid glycosides they may loosen the cell wall. PME activity can be regulated in planta by the proteinaceous inhibitor, PMEI. I used PME and PMEI to study the importance of methyl esterification in seed development and germination. As a means to identify PMEs involved in seed coat mucilage I identified 7 PMEs expressed in the seed coat. The PME gene HIGHLY METHYL ESTERIFIED SEED (HMS) is highly expressed at 7 Day Post Anthesis (DPA) both in the seed coat and the embryo. Using a hms-1 mutant, I showed that HMS is required for normal levels of PME activity and methyl esterification in the seed, mucilage extrusion and proper embryo cell expansion, rigidity and morphogenesis between 4 and 10 DPA. The mucilage extrusion defect is a secondary effect of the function of HMS in the embryo. I hypothesize that HMS is required for cell wall loosening in the embryo to allow for cell expansion during the accumulation of storage reserves. To evaluate the importance of methyl esterification in germination my collaborators and I first showed that PME activity changed during the different stages of germination: it first increased before testa rupture and decreased during endosperm rupture. Treatment with the hormone abscisic acid (ABA) to increase dormancy prolonged PME activity in the seeds. Inversely when we negatively regulated PME activity in the A. thaliana seed with the overexpression of a PMEI (OE PMEI5), we generated larger seeds with bigger cells. These seeds germinated faster both in presence or absence of ABA. Therefore we hypothesize that the PME(s) inhibited by PMEI5 establishes stronger cell walls that restrict germination. This thesis clearly demonstrates that PME activity is important in the regulation of seed cell wall methyl esterification impacting embryo growth and germination.
Primary plant cell walls are comprised largely of the polysaccharides cellulose,hemicellulose, and pectins, and can also contain up to 10% protein. These cell wall componentsinteract non-covalently and covalently to form a functional cell wall. Interactions betweencellulose and pectins are poorly understood, and are the focus of this research. Arabidopsis seedcoat epidermal cells produce three distinct types of cell walls: an outer primary wall; mucilage, aspecialized wall composed primarily of pectins; and a rigid columella. When seeds are hydrated,mucilage expands rapidly, breaking the outer wall, to form a mucilage halo that surrounds andremains adherent to the seed. The columella appears to be composed primarily of cellulose, andis therefore an excellent model for investigating cellulose biosynthesis. Cell wall biosynthesisand polysaccharide interactions were examined during seed coat development and in mucilageadherence to better understand cell wall assembly and function.Cellulose is synthesized by the CELLULOSE SYNTHASE A (CESA) family ofglucosyltransferases. It has been proposed that at least three different CESAs are required toform a functional Cellulose Synthase Complex (CSC). I investigated the contribution of CESA2,CESA5 and CESA9 in cellulose biosynthesis during seed coat development. Based on seed coatepidermal cell morphology and cellulose quantification, all three CESAs have non-redundantroles in secondary wall biosynthesis, while CESA5 specifically functions in mucilagebiosynthesis. CESA3 is expressed in the seed coat during mucilage biosynthesis and missensemutations in CESA3, isoxaben resistant 1 (ixr1-1 and ixr1-2), result in altered mucilage structureand pectin distribution, and reduced cellulose amounts in seeds.The mechanism of mucilage adherence was examined by comparing two loss of functionmutants that disrupt adherence, cesa5-1 and sos5-2. SOS5 encodes an arabinogalactan protein hypothesized to influence adherence through CESA5. However, the phenotype of each singlemutant differs and a cesa5 sos5 double mutant has an enhanced phenotype. Therefore, it isunlikely that SOS5 promotes mucilage adherence through CESA5. SOS5 may influencemucilage structure through galactans, as it is required for the proper function of the !-galactosidase, MUM2. This demonstrates a role for AGPs in galactan metabolism and cell wallpolysaccharide interactions.
During differentiation, the Arabidopsis seed coat epidermal cells produce copious amounts of mucilage that is extruded from the seed coat upon imbibition. Mucilage is composed primarily of pectin, a polysaccharide that is a main component of the cell wall. For this reason, the Arabidopsis seed coat is a good system for studying the biosynthesis, secretion and modification of pectin. Mutants with mucilage defects can be used to identify genes involved in the production of pectin. Mucilage-Modified mutants, including mum1, mum2 and mum4, were identified using screens of EMS mutagenized plants. Both mum1 and mum2 lack the ability to release the mucilage when mature seeds are imbibed. MUM2 encodes a β-galactosidase that modifies the mucilage structure in the apoplast. I have cloned the MUM1 gene and shown it to encode a putative transcription factor LEUNIG_HOMOLOG (LUH). Cellular localization and transcriptional assay results suggest that LUH/MUM1 is a nuclear-localized, transcriptional activator. LUH/MUM1 is expressed in all the tissues examined including the seed coat. qRT PCR data suggest that LUH/MUM1 is expressed throughout seed coat development, reaching peak expression late in differentiation. MUM2 expression in the luh/mum1 mutant was reduced dramatically, relative to that of wild type. Over-expression of MUM2 could partially rescue the mum1 phenotype. These data suggest that LUH/MUM1 is a positive regulator of MUM2. qRT PCR data revealed a similar expression level of LUH/MUM1 in wild type compared to plants homozygous for mutations in several genes encoding regulators of seed coat mucilage, namely APETALA2, TRANSPARENT TESTA GLABRA1 (TTG1), TTG2 and GLABRA2. Thus the LUH/MUM1-MUM2 regulatory pathway appears to be independent of other transcription factors known to regulate aspects of seed coat mucilage biology. Mutations in the MUM4 gene result in seeds that release little mucilage. A mum4 mutant was mutagenized and resulting M2 progeny screened for modifier mutants. Ten enhancers (mum4 enhancer (men)) and ten suppressors (mum4 suppressors (msu)) mutants were isolated and partially characterized genetically and phenotypically. Further studies are needed to characterize these mutants.
The BLADE-ON-PETIOLE1 (BOP1) and BOP2 genes encode redundant transcription factors involved in morphological patterning in the proximal regions of lateral organs in Arabidopsis thaliana. Loss-of-function bop1 bop2 mutants display several developmental defects including a loss of floral organ abscission. Abscission occurs along specialised cell files, called abscission zones (AZs), which form at the boundary between the leaving organ and main plant body. This dissertation examined the contribution of BOP1 and BOP2 to the known abscission developmental framework and determined that bop1 bop2 flowers lack anatomy associated with AZs. Vestigial cauline leaf AZs are also absent in bop1 bop2 suggesting that BOP proteins are essential to establish AZ anatomy in both leaves and flowers, the first genes identified in Arabidopsis to do so. In support of this hypothesis, BOP1/2 activity is required for both premature floral organ abscission and ectopic cauline leaf abscission promoted by the constitutive expression of INFLORESCENCE DEFICIENT IN ABSCISSION (IDA) gene. In addition, BOP1 and BOP2 were found to promote growth of nectary glands which normally develop off the receptacle at the base of stamens adjacent to AZs. Boundary, nectary and AZ specific-gene expression is relatively unperturbed in bop1 bop2, indicating that positional information is intact. Taken together, these data suggest that BOP1 and BOP2 are key downstream modulators of positional programs operating at the lateral organ-plant body interface. bop1 bop2 flowers also show a subtle loss in floral meristem identity. Genetic analyses revealed a crucial role for BOP1 and BOP2 in suppression of secondary inflorescence identity in early floral primordia only when functions either LEAFY (LFY) or APETALA (AP1) were compromised, suggesting that BOP1 and BOP2 are potentiators of LFY and AP1 function rather than floral meristem identity genes themselves. BOP1 and BOP2 belong to the NPR1-like protein family. A phylogenetic analysis in land plants found both monocot and dicot BOP protein homologues that cluster independently from other NPR1-like proteins. A final contribution examined the subcellular localisation of BOP2 during development and showed that BOP protein-protein interactions in planta showed behaviour consistent with that known for NPR1.
No abstract available.
During differentiation of Arabidopsis thaliana seed coat epidermal cells, dramatic changes occur highlighted by the synthesis and secretion of large amounts of pectinaceous mucilage. This cell type, therefore, provides an excellent molecular-genetic model to study the biosynthesis, secretion and modification of plant cell wall polysaccharides. Here, I describe the development of an experimental tool to aid in studying cell wall components found in mucilage. I sought to identify a promoter that drives gene expression specifically in the mucilage secretory cells to investigate the effect of manipulating different cell wall polymers in the mucilage without detrimental effects to the rest of the plant. The search for such a promoter was initiated by analyzing seed coat microarray data, followed by investigating expression pattern of the candidate genes by Reverse Transcription-Polymerase Chain Reaction (RT-PCR). By fusing the regulatory region of these candidate genes to beta-glucuronidase gene (GUS), one promoter was identified, that of the DIRIGENT PROTEIN1 (DP1) gene, that was able to drive expression specifically in the epidermis and palisade layer of Arabidopsis and Brassica napus seed coats. These results were confirmed using Citrine Yellow Fluorescent Protein (YFP). To verify the potential ability of the DP1 promoter (DP1Pro) to express enzyme-encoding genes and determine putative role(s) of homogalacturonan in mucilage, the ARABIDOPSIS DEHISCENCE ZONE POLYGALACTURONASE2 (ADPG2) encoding a polygalacturonase was expressed under the control of DP1Pro and targeted to the apoplast. Although RT-PCR results showed that the DP1Pro drives expression of this gene in the seed coat, no significant difference was found between transgenic and wild type mucilage or the seed coat epidermal cell morphology.MUCILAGE-MODIFIED4 (MUM4) gene was found to encode a putative UDP-L-rhamnose synthase, required for synthesis of mucilage. While the transcript of MUM4 is found throughout the plant, it is specifically up-regulated during mucilage biosynthesis. To identify the sequences responsible for this up-regulation, a deletion analysis of the MUM4 transcription regulatory region was undertaken. Dissection of the MUM4 promoter region led to identification of two functional domains: one conferring higher level of gene expression in root, cotyledon and silique walls, and one that promotes expression specifically in the seed coat.
The plant cell wall is a complex and dynamic network of polysaccharides and structural proteins, which lies outside the plasma membrane and provides strength and protection. The mechanical properties of the cell wall depend largely on the structure and organization of its components. Pectins form the gel matrix in which all other wall components are embedded and changes in their degree of methylesterification (DM) impact wall strength and adhesion. Low DM pectin molecules can be linked together via calcium bridges to form strong gels, but very few mutants affecting pectin methylesterification have been identified. During my MSc research, I characterized flying saucer 1 (fly1), a novel Arabidopsis thaliana seed coat mutant, which displays primary cell wall detachment, reduced mucilage extrusion, and increased mucilage adherence. These defects appear to result from a lower DM in mucilage, and can be intensified by addition of Ca²⁺ ions or completely rescued by treatment of seeds with cation chelators. The FLY1 gene encodes a protein with multiple transmembrane spans that is targeted to the secretory pathway and contains a RING-H2 domain, which generally facilitates protein-protein interactions. FLY1-YFP fusion proteins localize to small intracellular compartments in seed coat epidermal cells at the stage of mucilage biosynthesis. TUL1, a previously described FLY1 yeast ortholog, is a Golgi-localized E3 ligase involved in the trafficking of membrane proteins. I propose that FLY1 promotes pectin methylesterification in seed coat epidermal cells, potentially through interactions with pectin methyltransferase enzymes in the Golgi apparatus. Co-expression analysis suggests that FLY1 and FLY2, its only paralog, may play partially redundant roles in xylem development. These genes may be regulated by KNAT7, a transcription factor that controls secondary wall biosynthesis in both xylem and seed coat cells. The binding partners of the FLY1 protein and its precise molecular function remain to be determined.