Gotowa bibliografia na temat „Yeast Gal4”

The transcriptional activation function of the Saccharomyces cerevisiae GAL4 protein is modulated by the GAL80 and GAL3 proteins. In the absence of galactose, GAL80 inhibits the function of GAL4, presumably by direct binding to the GAL4 protein. The presence of galactose triggers the relief of the GAL80 block. The key to this relief is the GAL3 protein. How GAL3 and galactose activate GAL4 is not understood, but the long-standing notion has been that a galactose derivative formed by catalytic activity of GAL3 is the inducer that interacts with GAL80 or the GAL80-GAL4 complex. Here we report that overproduction of the GAL3 protein causes constitutive expression of GAL/MEL genes in the absence of exogenous galactose. Overproduction of the GAL1 protein (galactokinase) also causes constitutivity, consistent with the observations that GAL1 is strikingly similar in amino acid sequence to GAL3 and has GAL3-like induction activity. Cells lacking the GAL10-encoded UDP-galactose-UDP-glucose epimerase retained the constitutivity response to overproduction of GAL3, making it unlikely that constitutivity is due to endogenously produced galactose. A galactose-independent mechanism of constitutivity is further indicated by the inducing properties of two newly created galactokinaseless alleles of GAL1. On the basis of these data, we propose a new model for galactose-induced activation of the GAL4 protein. This model invokes galactose-activation of the GAL3 and GAL1 proteins which in turn elicit an alteration of the GAL80-GAL4 complex to activate GAL4. This model is consistent with all the known features of the system and has important implications for manipulating GAL4-dependent transcriptional activation in vitro.

The transcriptional activation function of the Saccharomyces cerevisiae GAL4 protein is modulated by the GAL80 and GAL3 proteins. In the absence of galactose, GAL80 inhibits the function of GAL4, presumably by direct binding to the GAL4 protein. The presence of galactose triggers the relief of the GAL80 block. The key to this relief is the GAL3 protein. How GAL3 and galactose activate GAL4 is not understood, but the long-standing notion has been that a galactose derivative formed by catalytic activity of GAL3 is the inducer that interacts with GAL80 or the GAL80-GAL4 complex. Here we report that overproduction of the GAL3 protein causes constitutive expression of GAL/MEL genes in the absence of exogenous galactose. Overproduction of the GAL1 protein (galactokinase) also causes constitutivity, consistent with the observations that GAL1 is strikingly similar in amino acid sequence to GAL3 and has GAL3-like induction activity. Cells lacking the GAL10-encoded UDP-galactose-UDP-glucose epimerase retained the constitutivity response to overproduction of GAL3, making it unlikely that constitutivity is due to endogenously produced galactose. A galactose-independent mechanism of constitutivity is further indicated by the inducing properties of two newly created galactokinaseless alleles of GAL1. On the basis of these data, we propose a new model for galactose-induced activation of the GAL4 protein. This model invokes galactose-activation of the GAL3 and GAL1 proteins which in turn elicit an alteration of the GAL80-GAL4 complex to activate GAL4. This model is consistent with all the known features of the system and has important implications for manipulating GAL4-dependent transcriptional activation in vitro.

GAL3 gene expression is required for rapid GAL4-mediated galactose induction of the galactose-melibiose regulon genes in Saccharomyces cerevisiae. Here we show by Northern (RNA) blot analysis that GAL3 gene expression is itself galactose inducible. Like the GAL1, GAL7, GAL10, and MEL1 genes, the GAL3 gene is severely glucose repressed. Like the MEL1 gene, but in contrast to the GAL1, GAL7, and GAL10 genes, GAL3 is expressed at readily detectable basal levels in cells grown in noninducing, nonrepressing media. We determined the sequence of the S. cerevisiae GAL3 gene and its 5'-noncoding region. Within the 5'-noncoding region of the GAL3 gene, we found two sequences similar to the UASGal elements of the other galactose-melibiose regulon genes. Deletion analysis indicated that only the most ATG proximal of these sequences is required for GAL3 expression. The coding region of GAL3 consists of a 1,275-base-pair open reading frame in the direction of transcription. A comparison of the deduced 425-amino-acid sequence with the protein data bank revealed three regions of striking similarity between the GAL3 protein and the GAL1-specified galactokinase of Saccharomyces carlsbergensis. One of these regions also showed striking similarity to sequences within the galactokinase protein of Escherichia coli. On the basis of these protein sequence similarities, we propose that the GAL3 protein binds a molecule identical to or structurally related to one of the substrates or products of the galactokinase-catalyzed reaction.

GAL3 gene expression is required for rapid GAL4-mediated galactose induction of the galactose-melibiose regulon genes in Saccharomyces cerevisiae. Here we show by Northern (RNA) blot analysis that GAL3 gene expression is itself galactose inducible. Like the GAL1, GAL7, GAL10, and MEL1 genes, the GAL3 gene is severely glucose repressed. Like the MEL1 gene, but in contrast to the GAL1, GAL7, and GAL10 genes, GAL3 is expressed at readily detectable basal levels in cells grown in noninducing, nonrepressing media. We determined the sequence of the S. cerevisiae GAL3 gene and its 5'-noncoding region. Within the 5'-noncoding region of the GAL3 gene, we found two sequences similar to the UASGal elements of the other galactose-melibiose regulon genes. Deletion analysis indicated that only the most ATG proximal of these sequences is required for GAL3 expression. The coding region of GAL3 consists of a 1,275-base-pair open reading frame in the direction of transcription. A comparison of the deduced 425-amino-acid sequence with the protein data bank revealed three regions of striking similarity between the GAL3 protein and the GAL1-specified galactokinase of Saccharomyces carlsbergensis. One of these regions also showed striking similarity to sequences within the galactokinase protein of Escherichia coli. On the basis of these protein sequence similarities, we propose that the GAL3 protein binds a molecule identical to or structurally related to one of the substrates or products of the galactokinase-catalyzed reaction.

The Saccharomyces cerevisiae GAL5 (PGM2) gene was isolated and shown to encode the major isozyme of phosphoglucomutase. Northern (RNA) blot hybridization revealed that the GAL5 transcript level increased three- to fourfold in response to galactose and was severely repressed in response to glucose. Total cellular phosphoglucomutase activity was likewise responsive to galactose and to glucose, and this responsiveness was found to be due primarily to variation in the activity of the major isozyme of phosphoglucomutase. These results imply that the major and minor isozymes of phosphoglucomutase have distinct roles in yeast cells. The galactose inducibility of GAL5 was found to be under the control of the GAL4, GAL80, and GAL3 genes. In striking contrast to other galactose-inducible genes, the GAL5 gene exhibited an unusually high GAL4-independent basal level of expression. These results have implications for metabolic trafficking.

The Saccharomyces cerevisiae GAL5 (PGM2) gene was isolated and shown to encode the major isozyme of phosphoglucomutase. Northern (RNA) blot hybridization revealed that the GAL5 transcript level increased three- to fourfold in response to galactose and was severely repressed in response to glucose. Total cellular phosphoglucomutase activity was likewise responsive to galactose and to glucose, and this responsiveness was found to be due primarily to variation in the activity of the major isozyme of phosphoglucomutase. These results imply that the major and minor isozymes of phosphoglucomutase have distinct roles in yeast cells. The galactose inducibility of GAL5 was found to be under the control of the GAL4, GAL80, and GAL3 genes. In striking contrast to other galactose-inducible genes, the GAL5 gene exhibited an unusually high GAL4-independent basal level of expression. These results have implications for metabolic trafficking.

The Saccharomyces cerevisiae GCR2 gene affects expression of most of the glycolytic genes. We report the nucleotide sequence of GCR2, which can potentially encode a 58,061-Da protein. There is a small cluster of asparagines near the center and a C-terminal region that would be highly charged but overall neutral. Fairly homologous regions were found between Gcr2 and Gcr1 proteins. To test potential interactions, the genetic method of S. Fields and O. Song (Nature [London] 340:245-246, 1989), which uses protein fusions of candidate gene products with, respectively, the N-terminal DNA-binding domain of Gal4 and the C-terminal activation domain II, assessing restoration of Gal4 function, was used. In a delta gal4 delta gal80 strain, double transformation by plasmids containing, respectively, a Gal4 (transcription-activating region)/Gcr1 fusion and a Gal4 (DNA-binding domain)/Gcr2 fusion activated lacZ expression from an integrated GAL1/lacZ fusion, indicating reconstitution of functional Gal4 through the interaction of Gcr1 and Gcr2 proteins. The Gal4 (transcription-activating region)/Gcr1 fusion protein alone complemented the defects of both gcr1 and gcr2 strains. Furthermore, a Rap1/Gcr2 fusion protein partially complemented the defects of gcr1 strains. These results suggest that Gcr2 has transcriptional activation activity and that the GCR1 and GCR2 gene products function together.

The Saccharomyces cerevisiae GCR2 gene affects expression of most of the glycolytic genes. We report the nucleotide sequence of GCR2, which can potentially encode a 58,061-Da protein. There is a small cluster of asparagines near the center and a C-terminal region that would be highly charged but overall neutral. Fairly homologous regions were found between Gcr2 and Gcr1 proteins. To test potential interactions, the genetic method of S. Fields and O. Song (Nature [London] 340:245-246, 1989), which uses protein fusions of candidate gene products with, respectively, the N-terminal DNA-binding domain of Gal4 and the C-terminal activation domain II, assessing restoration of Gal4 function, was used. In a delta gal4 delta gal80 strain, double transformation by plasmids containing, respectively, a Gal4 (transcription-activating region)/Gcr1 fusion and a Gal4 (DNA-binding domain)/Gcr2 fusion activated lacZ expression from an integrated GAL1/lacZ fusion, indicating reconstitution of functional Gal4 through the interaction of Gcr1 and Gcr2 proteins. The Gal4 (transcription-activating region)/Gcr1 fusion protein alone complemented the defects of both gcr1 and gcr2 strains. Furthermore, a Rap1/Gcr2 fusion protein partially complemented the defects of gcr1 strains. These results suggest that Gcr2 has transcriptional activation activity and that the GCR1 and GCR2 gene products function together.

Abstract The Saccharomyces cerevisiae GAL/MEL regulon genes are normally induced within minutes of galactose addition, but gal3 mutants exhibit a 3-5-day induction lag. We have discovered that this long-term adaptation (LTA) phenotype conferred by gal3 is complemented by multiple copies of the GAL1 gene. Based on this result and the striking similarity between the GAL3 and GAL1 protein sequences we attempted to detect galactokinase activity that might be associated with the GAL3 protein. By both in vivo and in vitro tests the GAL3 gene product does not appear to catalyze a galactokinase-like reaction. In complementary experiments, Escherichia coli galactokinase expressed in yeast was shown to complement the gal1 but not the gal3 mutation. Thus, the complementation activity provided by GAL1 is not likely due to galactokinase activity, but rather due to a distinct GAL3-like activity. Overall, the results indicate that GAL1 encodes a bifunctional protein. In related experiments we tested for function of the LTA induction pathway in gal3 cells deficient for other gene functions. It has been known for some time that gal3gal1, gal3gal7, gal3gal10, and gal3 rho- are incapable of induction. We constructed isogenic haploid strains bearing the gal3 mutation in combination with either gal15 or pgi1 mutations: the gal15 and pgi1 blocks are not specific for the galactose pathway in contrast to the gal1, gal7 and gal10 blocks. The gal3gal5 and gal3pgi1 double mutants were not inducible, whereas both the gal5 and pgi1 single mutants were inducible. We conclude that, in addition to the GAL3-like activity of GAL1, functions beyond the galactose-specific GAL1, GAL7 and GAL10 enzymes are required for the LTA induction pathway.

The yeast GAL1 and GAL10 genes are transcribed at a remarkably low basal level when galactose is unavailable and are induced by over 4 orders of magnitude when it becomes available. Approximately six negative control elements (designated GAL operators GALO1 to GALO6) are located adjacent to or overlapping four binding sites for the transcription activator GAL4 in the GAL upstream activating sequence UASG. The negative control elements contribute to the broad range of inducibility of GAL1 and GAL10 by inhibiting two GAL4/galactose-independent activating elements (GAE1 and GAE2) in UASG. In turn, multiple GAL4-binding sites in UASG are necessary for GAL4 to overcome repression by the negative control elements under fully inducing conditions. When glucose in addition to galactose is available (repressing conditions), the ability of GAL4 to activate transcription is diminished as a result of its reduced affinity for DNA and the reduced availability of inducer. Under these conditions, the negative control elements inhibit transcriptional activation from the glucose-attenuated GAL4 sites, thus accounting at least in part for glucose repression acting in cis. A normal part of transcriptional regulation of the GAL1 and GAL10 genes, therefore, appears to involve a balance between the opposing functions of positive and negative control elements.

The advent and establishment of systems biology has cemented the idea that real understanding of biological systems requires quantitative models, that can be integrated to provide a complete description of the cell and its complexities. At the same time, synthetic biology attempts to leverage such quantitative models to efficiently engineer novel, predictable behaviour in biological systems. Together, these advances indicate that the future understanding and application of biology will require the ability to create, parameterise and discriminate between quantitative models of cellular processes in a rigorous and statistically sound manner. In this thesis we take the regulation of GAL1 expression in Saccharomyces cerevisiae as a test case and look at all aspects of this process: from reporter selection to data acquisition and statistical analysis. In chapter B we will discuss optimal fluorescent reporter selection and construction for investigating transcriptional dynamics, as well as procedures for quantifying and correcting the various sources of error in our microscope system. In chapter 3 we will describe software developed to analyse fluorescent microscopy images and convert them to gene expression data. A number of iterations of the software are tested against manually curated data sets, and the measurement error produced by its imperfections is quantified and discussed. In chapter 4 a method, based on fluctuations in photobleaching, is developed for quantifying both measurement error and the relationship between protein concentration and measured fluorescence. The method is refined and its efficacy discussed. In the last section I make a preliminary application of these methods to investigating the regulatory effect of the GAL10-lncRNA. Interesting phenomena are observed and further investigated using two new strains: genetic variants expressing a fluorescent reporter from the GAL1 promoter, one harbouring a wild type GAL1 promoter and one in which the binding site for the Gal10 noncoding RNA has been removed. The methods developed throughout the thesis are applied and the data analysed. A heterogeneous response, distinguishable between the two strains, is observed and related to cell-to-cell variations in growth rate.

The formation of Saccharomyces cerevisiae cell wall requires the coordinated activity of enzymes involved in the biosynthesis and modification of its components, such as glucans. The β-(1,3)-glucan synthase complexes, that have Fks proteins as putative catalytic subunits, use UDP-glucose as a substrate and catalyse the synthesis and vectorial extrusion of glucan chains into the outer space. Then, β-(1,3)-glucan chains are branched, elongated and remodelled in order to create a robust texture capable of counteracting the high internal pressure and determining cell morphology. β-(1,3)-glucan is the main component of the vegetative cell wall and one of the most abundant polymers of the spore wall. Several enzymes belonging to the family GH72 of glycosyl hydrolases have been identified in fungi. These enzymes are responsible of the lateral elongation of β-(1,3)-glucan, thus contributing to the assembly and organization of the glucan layer. The multigene GAS family of S. cerevisiae is composed of five members, GAS1-5, involved in cell wall maintenance. They share significant similarity with Aspergillus fumigatus GEL1 and GEL2, and with Candida albicans PHR1 and PHR2. Similar to the most extensively characterized member, Gas1p, the remaining Gas proteins are β-(1,3)-glucanosyltransferases involved in cell wall assembly and maintenance. Based on their expression patterns, they appear to play partially overlapping roles throughout the yeast life cycle: GAS1 and GAS5 are expressed during vegetative growth, whereas GAS2 and GAS4 are expressed exclusively during sporulation and required for normal spore wall formation, finally GAS3 is a weakly expressed gene. Thus these enzymes could satisfy the cellular needs to remodel β-(1,3)-glucan in different physiological conditions and in different conformations along the yeast life cell cycle. Moreover, considering its role in yeast cell biology, the GAS enzyme family represents a very promising molecular target for new antifungal drugs. During my PhD thesis I focused my interest on the functional characterization of GAS1, 2, 3 and 4 in various stages of the yeast life cycle: vegetative growth, meiosis, sporulation and spore germination. This study is aimed to understand the biological significance of the developmentally regulated requirement of the specific members of the GAS redundant family in the morphological stages of yeast life cycle. GAS2 and GAS4 genes are specifically induced during sporulation and encode for glycoproteins. The effects of the loss of Gas2 and Gas4 proteins on spore wall morphogenesis are dramatic. Synthesis of all the layers of the spore cell wall occurs, but the accumulation and organization of wall material is abnormal. The lack of the elongase activity of Gas2 and Gas4 proteins in the double mutant might cause the formation of shorter or less branched β(1,3)-glucan chains in the inner layer of the spore wall. Thus, the connection of the outermost layers to a less compact glucan network could make the spore wall more fragile and easily stripped under harsh conditions. These defects cause an increase in spore permeability to exogenous substances, a decrease in refractivity, and a marked decrease in spore viability. The possible execution point for GAS2 and GAS4 could be between the synthesis and organization of β(1,3)-glucan and, more specifically, in the elongation of the β(1,3)-glucan chains. Consistently with their role, during sporulation Gas2 and Gas4 proteins localize at the newly assembling prospore membrane during the meiotic divisions and in mature ascospore the proteins decorate the spore periphery. A slight difference in the protein patterns of fluorescence on the spore suggests that Gas2p and Gas4p final localization could be respectively the spore wall and the prospore membrane. In this work, an extensive study of the localization of the Gas1 protein during the yeast life cycle was performed, taking advantage of a GFP-tagged version of the protein. During vegetative growth Gas1p has a dual localization: in the plasma membrane and at the site of bud emergence, particularly in the neck, in the chitin ring that surrounds the neck region and in the bud scars where Gas1p remains after cytokinesis. At the neck region Gas1p appears to absolve important functions in yeast as a part of the mechanisms that ensure the resistance of the neck region and the morphogenesis of the septum. The size and morphology of the neck region is severely affected both in the gas1Δ and gas1Δ chs3Δ mutant, suggesting an involvement of the protein in the maintenance of the integrity of the mother-bud neck region. The presence of Gas1p in the chitin ring could be part of the mechanism necessary to prevent new incorporation of glucan chains into the neck region or alternatively the protein could be required for a particular type of remodelling necessary for the septum region in preparation to cell division. Additionally, Gas1p could act as landmark protein for the choice of the site of bud emergencee. As to Gas1p localization at the plasma membrane, our study supports the validity of Gas1p-GFP as a marker to follow the dynamics of lipid raft. At the induction of sporulation, GAS1 mRNA levels steadily decrease and by 10h it is completely declined. Surprisingly, Gas1p levels are roughly constant during the entire sporulation processs and the protein is very stable, being detectable also at 43h after the induction of sporulation. During spore development, a translocation event occurs through which at the completion of meiosis II, Gas1p, synthesized during vegetative growth, is removed from the plasma membrane and internalized. Later, Gas1p is detected associated to the nascent prospore membrane surrounding the nuclear lobes and finally in mature spores it localizes at the spore periphery. This translocation event suggests that Gas1p delivery to the spore surface is not part of the developmentally reprogramming of the secretory pathway from the trans Golgi to the prospore membrane, whereas it involves at least in part the endocytic pathway. We demonstrated that END3-mediated endocytosis is one of the mechanisms required for the removal of the Gas1p from the plasma membrane and its efficient re-localization at the prospore membrane. Moreover in a sps1Δ mutant, Gas1p remains localized at the plasma membrane and fails to reach the spore surface. Sps1p is a member of the Ste20 protein kinase family and regulates the trafficking to the prospore membrane of enzymes involved in spore wall synthesis, such as the glucan synthase Fks2p and chitin synthase Chs3p. Thus Sps1p could regulate the traffic of Gas1p most likely in an indirect way by interacting and modifying the components of the intracellular trafficking machinery. Gas1p translocation during sporulation To test a possible involvement of Gas1p in spore wall formation, in this study we tried to characterize the sporulation phenotype of a gas1Δ mutant. Unfortunately our analysis was complicated by the mutant reduced cell viability when grown in presence of a poor carbon source such as acetate. gas1Δ sporulation defect could rely in a unsatisfied energetic request as the cell wall perturbations, typical of a gas1Δ mutant, enhance carbon and energy mobilization to efficiently combat cell wall weakening and the metabolism of acetate as the sole carbon source could be not sufficient to satisfy this energetic request. Moreover the addition of sorbitol to the sporulation medium only partially rescues gas1Δ defective phenotype during spore development. Even though sorbitol can mitigate the gas1Δ cell wall damages, it has no buffering effect on the gas1Δ energetic request, thus the mutant cells remained substantially unable to sporulate. Consequently, gas1Δ sporulation defective phenotype appears to be reminiscent of the mutant defects during vegetative growth, even worsened in a poor carbon source. Even though we cannot exclude a role for Gas1p during spore morphogenesis, it is our consumption that the protein translocation to the spore represents a “storage”mechanisms to ensure the presence of the Gas1p during spore germination. At 3h after the shift to a rich medium, Gas1p exhibits a highly polarized distribution, decorating exclusively half of the germinating spore in its growing pole. The protein localization is consistent with its role in glucan layer remodelling of the cell wall at the growing portion of the germinating cell. Besides gas1Δ germinating spore inability to support the elongation during the polarized growth of the cell suggests that Gas1p is required for a very early step in germination. Besides the protein is involved in a post-germination stage to support the polarized growth of the newly emerging bud. Finally, in this study we reported the preliminary results about the functional characterization of GAS3. The gene is expressed at a very low level during the vegetative growth in glucose and acetate. Consistently with the GAS3 expression pattern, Gas3p appears as a highly polydispersed glycoprotein of high molecular weight that is present in vegetative growing cells and along the sporulation process. EndoH treatment reduces the size and the aspect of the protein to a sharp band, suggesting that Gas3p is a heavily N-glycosylated protein. The experiments indicated that neither the overexpression nor the deletion of the GAS3 gene, alone or in combination with GAS2 and GAS4, lead to relevant differences in sporulation with respect toh the wild type or with the defective phenotype of the gas2 gas4 null mutant strain . The construction of a tagged version of the Gas3 protein to determine its localization will be a useful tool to understand the function ofl Gas3p during yeast life cycle.

Genomic and proteomic investigations have yielded, and continue to produce, a large amount of information about genes and their protein products. In contrast, the evidence bearing on physiological roles of specific proteins is much more scarce. To address the functional part of biological inquiry, one would like to perturb, at will and selectively, the function of any protein of interest in vivo and to analyze the resulting phenotypic effects, thereby probing the protein’s role in a cell. Ideally, a method for doing so should be applicable both to individual gene products and to a large collection of them. Gene knockouts, a powerful tool to study gene function, have limitations in the study of development when the early phenotypes are cell- or organismal- lethal. Conditional mutants are particularly useful for analysis of genes whose functions are essential for the organism’s viability. A conditional mutant retains the function of a gene under one set of conditions, called permissive, and shows an inactive phenotype under a different set of conditions, called nonpermissive; the latter must be still permissive for the wild type (wt) allele of a gene. Conditional mutants make possible the analysis of physiological changes that follow controlled inactivation of a gene or gene product and can be used to address the function of any gene. Temperature sensitive (ts) mutants are an important class of conditional mutants whose phenotype is similar to that of wt at lower (permissive) temperature, but show low or reduced level of activity above a certain temperature called restrictive temperature, while the wt gene shows a similar phenotype at both the temperatures. Ts mutants provide an extremely powerful tool to study gene expression in vivo and in cell culture. They provide a reversible mechanism to lower the level of a specific gene product simply by changing the temperature of growth of the organism. Ts mutants are typically generated by random mutagenesis; either by ultraviolet light, a chemical mutagen or by error-prone PCR followed by often laborious screening procedures. Therefore, they are cumbersome to make, especially in the case of organisms with long generation times. Keeping in view the importance of ts mutants in biology, Varadarajan et al. 1996, had developed an algorithm to predict ts mutants at predicted, buried sites of a globular protein from its amino acid sequence. Experimental tests of the algorithm were carried out on the CcdB toxin of Escherichia coli to further refine and improve the method (Chakshusmathi et al. 2004). Based on this result simple rules for the design of ts mutants were suggested. This thesis aims at validating and improving on these rules and to find out if ts mutants of a protein can also be generated by perturbing functionally important residues. In addition, it is currently unclear with what frequency ts mutants of a protein isolated in one organism will show a ts phenotype in a completely different organism. This thesis makes preliminary efforts to address this issue. The model system chosen to carry out these studies is a protein called Gal4, which is a yeast transcriptional activator. This protein is biologically relevant as it has been used for ectopic gene expression in diverse organisms including yeast, fruitflies, zebrafish, mice and frogs (Ornitz et al. 1991; Brand and Perrimon 1993; Rahner et al. 1996; Andrulis et al. 1998; Scheer and Camnos-Ortega 1999; Hartley et al. 2002). The introductory chapter (Chapter 1) discusses the importance of ts mutants and our understanding and progress in this field so far, relevant for the work reported in this thesis. Chapter 2 describes generation of ts mutants of Gal4 in yeast. Full length Gal4 (fGal4) is an 881-aa protein. To simplify the construction of ts Gal4, we have designed a functional truncated Gal4 (miniGal4 or mGal4) of 197 residues. Five residues (9, 10, 15, 18 and 23) of the Gal4 DNA binding domain, which are in close contact with the DNA, were randomized in mGal4. Based on average hydrophobicity and hydrophobic moment, 68, 69, 70, 71, and 80 are the only residues in the region 1-150 that are predicted to be buried at the 90% confidence level. Of these five sites, residues 68, 69 and 70 were chosen for mutagenesis. At these three sites, four stereochemically diverse substitutions (Lys, Ser, Ala and Trp) were made. In a separate set of experiments each predicted, buried residues were also individually randomized in both mini and in full length Gal4 (fGal4). In all cases, we have been successful in isolating ts mutants in more than one position. At both permissive and restrictive temperatures, the activity of the Gal4 ts mutants is substantially lower than the wt. However, at the restrictive temperature, the activity of the ts Gal4 is lowered below the threshold required for reporter gene expression. This view of how ts mutants function is quite different from the general notion that the ts and wt behave similarly at permissive temperatures. Chapter 3 deals with transferability of two of the ts constructs mutated at DNA binding residues (R15W and K23P) to Drosophila. Two fGal4 encoding DNA fragments carrying the mutations were cloned into P element vectors under control of Elav and GMR promoters and several transgenic Drosophila lines were generated. These were crossed to various UAS reporter lines and progeny were characterized for reporter gene expression as a function of temperature. We show that both of these yeast ts mutants also show a ts phenotype in Drosophila. We have compared our ts Gal4 system with a popularly used system (TARGET) (McGuire et al. 2003) used for conditional gene expression in Drosophila. Our ts Gal4 mutants appear to provide tighter control at the restrictive temperature and a more uniform and rapid induction of gene expression upon shifting from the restrictive to the permissive temperature than the TARGET system with the promoters and the reporters we have used. Although cold sensitive (cs) mutants are often more useful than ts mutants, for reasons currently unclear, cs mutants are much more difficult to isolate than ts mutants. In Chapter 4, we have attempted to convert the ts phenotypes observed with Gal4 mutants in Drosophila and CcdB mutants in E. coli (Chakshusmathi et al. 2004) to cs phenotypes by increasing the expression level of these mutant proteins selectively at higher temperature. Several ts mutants of CcdB have been previously reported (Chakshusmathi et al. 2004). For converting the ts phenotype observed by E. coli toxin CcdB mutants (Chakshusmathi et al. 2004) to a cs phenotype, the arabinose inducible plasmid pBAD24CcdB and its mutant derivatives were used. By inducing expression of the mutant protein at higher temperature with arabinose, while keeping the basal level of expression without arabinose at lower temperature, we have been able to show cold sensitive behavior by these CcdB ts mutants in E. coli. For producing a cs phenotype with Gal4 mutants in Drosophila, we have used a P element vector where the GMR element is placed in-between hsp70 binding sites. This driver results in enhanced expression of downstream genes at 30 relative to 18°C because of the presence of the hsp elements (Kramer and Staveley 2003). Ts mutants at DNA binding and buried residues of fGal4 were cloned into this vector and several transgenic lines for each construct were obtained. The Gal4 mutants at exposed DNA binding residues but not at buried residues show a cs phonotype when they were crossed to various UAS-reporters lines. The buried residue mutants are likely to be destabilized and their degradation pathway might differ in yeast and in Drosophila. Because of this, these mutants might not be showing the desired cs phenotype in Drosophila. Although mGal4 and fGal4 have very similar activities in yeast, it was necessary to examine if they also had identical activities in Drosophila. Determining their relative activities in Drosophila is the aim of Chapter 5. To this end, mGal4 was cloned into P element vectors under control of hsp70 or GMRhs promoters and transgenic flies were generated. The transgenic lines were crossed to various UAS-reporters and reporter gene activities in the progeny were characterized. Although mGal4 and fGal4 showed similar activity in yeast, in Drosophila for reasons that are currently unclear, mGal4 was considerably less active than fGal4. As some of the fGal4 mutants showed a cs phenotype under GMRhs driver as shown in the earlier chapter (Chapter 4), several ts mutants of mGal4 in yeast in buried and as well as at the DNA binding residues were transferred to Drosophila under hs and GMRhs promoter. The transgenic lines obtained were tested for cold sensitivity by crossing with various UAS-reporter lines. However, in all cases mutant mGal4 showed an inactive phenotype in Drosophila. We suggest that this is because the intrinsic activity of these mGal4 mutants is substantially weaker than wt mGal4 even at permissive temperature in yeast. The further lowering of activity in Drosophila pushes the activity below the threshold required for reporter gene expression resulting in an inactive phenotype. The concluding chapter (Chapter 6) summarizes the conclusions drawn from this entire study and provides insights into possible mechanisms responsible for ts and cs phenotypes. The mutant phenotypes of Gal4 in yeast and in Drosophila suggest that ts phenotypes appear to result from a threshold effect. Such mutations lower the activity and/or level of the protein relative to the wt at all temperatures. Since maximal stability temperatures are rarely in excess of room temperature, with an increase in temperature, the activity of an already marginally active mutant can fall below the threshold required for function resulting in a temperature sensitive phenotype. The strategies we used for producing ts mutants have several advantages over standard approaches of generating ts alleles by random mutagenesis. We anticipate that conclusions of this study would be useful for generation of ts mutants of other globular proteins in diverse organisms. We also show that exposed, functional residues involved in protein: ligand or protein: protein interactions appear to be attractive candidate sites for generating ts mutants that are transferable between organisms. In addition, the active site mutants of fGal4 in Drosophila, which show ts and cs phenotypes depending on the Drosophila promoter chosen for expression, can be used for conditional and reversible expression of a number of other genes using the Gal4-UAS system (Brand and Perrimon 1993).

Genomic and proteomic investigations have yielded, and continue to produce, a large amount of information about genes and their protein products. In contrast, the evidence bearing on physiological roles of specific proteins is much more scarce. To address the functional part of biological inquiry, one would like to perturb, at will and selectively, the function of any protein of interest in vivo and to analyze the resulting phenotypic effects, thereby probing the protein’s role in a cell. Ideally, a method for doing so should be applicable both to individual gene products and to a large collection of them. Gene knockouts, a powerful tool to study gene function, have limitations in the study of development when the early phenotypes are cell- or organismal- lethal. Conditional mutants are particularly useful for analysis of genes whose functions are essential for the organism’s viability. A conditional mutant retains the function of a gene under one set of conditions, called permissive, and shows an inactive phenotype under a different set of conditions, called nonpermissive; the latter must be still permissive for the wild type (wt) allele of a gene. Conditional mutants make possible the analysis of physiological changes that follow controlled inactivation of a gene or gene product and can be used to address the function of any gene. Temperature sensitive (ts) mutants are an important class of conditional mutants whose phenotype is similar to that of wt at lower (permissive) temperature, but show low or reduced level of activity above a certain temperature called restrictive temperature, while the wt gene shows a similar phenotype at both the temperatures. Ts mutants provide an extremely powerful tool to study gene expression in vivo and in cell culture. They provide a reversible mechanism to lower the level of a specific gene product simply by changing the temperature of growth of the organism. Ts mutants are typically generated by random mutagenesis; either by ultraviolet light, a chemical mutagen or by error-prone PCR followed by often laborious screening procedures. Therefore, they are cumbersome to make, especially in the case of organisms with long generation times. Keeping in view the importance of ts mutants in biology, Varadarajan et al. 1996, had developed an algorithm to predict ts mutants at predicted, buried sites of a globular protein from its amino acid sequence. Experimental tests of the algorithm were carried out on the CcdB toxin of Escherichia coli to further refine and improve the method (Chakshusmathi et al. 2004). Based on this result simple rules for the design of ts mutants were suggested. This thesis aims at validating and improving on these rules and to find out if ts mutants of a protein can also be generated by perturbing functionally important residues. In addition, it is currently unclear with what frequency ts mutants of a protein isolated in one organism will show a ts phenotype in a completely different organism. This thesis makes preliminary efforts to address this issue. The model system chosen to carry out these studies is a protein called Gal4, which is a yeast transcriptional activator. This protein is biologically relevant as it has been used for ectopic gene expression in diverse organisms including yeast, fruitflies, zebrafish, mice and frogs (Ornitz et al. 1991; Brand and Perrimon 1993; Rahner et al. 1996; Andrulis et al. 1998; Scheer and Camnos-Ortega 1999; Hartley et al. 2002). The introductory chapter (Chapter 1) discusses the importance of ts mutants and our understanding and progress in this field so far, relevant for the work reported in this thesis. Chapter 2 describes generation of ts mutants of Gal4 in yeast. Full length Gal4 (fGal4) is an 881-aa protein. To simplify the construction of ts Gal4, we have designed a functional truncated Gal4 (miniGal4 or mGal4) of 197 residues. Five residues (9, 10, 15, 18 and 23) of the Gal4 DNA binding domain, which are in close contact with the DNA, were randomized in mGal4. Based on average hydrophobicity and hydrophobic moment, 68, 69, 70, 71, and 80 are the only residues in the region 1-150 that are predicted to be buried at the 90% confidence level. Of these five sites, residues 68, 69 and 70 were chosen for mutagenesis. At these three sites, four stereochemically diverse substitutions (Lys, Ser, Ala and Trp) were made. In a separate set of experiments each predicted, buried residues were also individually randomized in both mini and in full length Gal4 (fGal4). In all cases, we have been successful in isolating ts mutants in more than one position. At both permissive and restrictive temperatures, the activity of the Gal4 ts mutants is substantially lower than the wt. However, at the restrictive temperature, the activity of the ts Gal4 is lowered below the threshold required for reporter gene expression. This view of how ts mutants function is quite different from the general notion that the ts and wt behave similarly at permissive temperatures. Chapter 3 deals with transferability of two of the ts constructs mutated at DNA binding residues (R15W and K23P) to Drosophila. Two fGal4 encoding DNA fragments carrying the mutations were cloned into P element vectors under control of Elav and GMR promoters and several transgenic Drosophila lines were generated. These were crossed to various UAS reporter lines and progeny were characterized for reporter gene expression as a function of temperature. We show that both of these yeast ts mutants also show a ts phenotype in Drosophila. We have compared our ts Gal4 system with a popularly used system (TARGET) (McGuire et al. 2003) used for conditional gene expression in Drosophila. Our ts Gal4 mutants appear to provide tighter control at the restrictive temperature and a more uniform and rapid induction of gene expression upon shifting from the restrictive to the permissive temperature than the TARGET system with the promoters and the reporters we have used. Although cold sensitive (cs) mutants are often more useful than ts mutants, for reasons currently unclear, cs mutants are much more difficult to isolate than ts mutants. In Chapter 4, we have attempted to convert the ts phenotypes observed with Gal4 mutants in Drosophila and CcdB mutants in E. coli (Chakshusmathi et al. 2004) to cs phenotypes by increasing the expression level of these mutant proteins selectively at higher temperature. Several ts mutants of CcdB have been previously reported (Chakshusmathi et al. 2004). For converting the ts phenotype observed by E. coli toxin CcdB mutants (Chakshusmathi et al. 2004) to a cs phenotype, the arabinose inducible plasmid pBAD24CcdB and its mutant derivatives were used. By inducing expression of the mutant protein at higher temperature with arabinose, while keeping the basal level of expression without arabinose at lower temperature, we have been able to show cold sensitive behavior by these CcdB ts mutants in E. coli. For producing a cs phenotype with Gal4 mutants in Drosophila, we have used a P element vector where the GMR element is placed in-between hsp70 binding sites. This driver results in enhanced expression of downstream genes at 30 relative to 18°C because of the presence of the hsp elements (Kramer and Staveley 2003). Ts mutants at DNA binding and buried residues of fGal4 were cloned into this vector and several transgenic lines for each construct were obtained. The Gal4 mutants at exposed DNA binding residues but not at buried residues show a cs phonotype when they were crossed to various UAS-reporters lines. The buried residue mutants are likely to be destabilized and their degradation pathway might differ in yeast and in Drosophila. Because of this, these mutants might not be showing the desired cs phenotype in Drosophila. Although mGal4 and fGal4 have very similar activities in yeast, it was necessary to examine if they also had identical activities in Drosophila. Determining their relative activities in Drosophila is the aim of Chapter 5. To this end, mGal4 was cloned into P element vectors under control of hsp70 or GMRhs promoters and transgenic flies were generated. The transgenic lines were crossed to various UAS-reporters and reporter gene activities in the progeny were characterized. Although mGal4 and fGal4 showed similar activity in yeast, in Drosophila for reasons that are currently unclear, mGal4 was considerably less active than fGal4. As some of the fGal4 mutants showed a cs phenotype under GMRhs driver as shown in the earlier chapter (Chapter 4), several ts mutants of mGal4 in yeast in buried and as well as at the DNA binding residues were transferred to Drosophila under hs and GMRhs promoter. The transgenic lines obtained were tested for cold sensitivity by crossing with various UAS-reporter lines. However, in all cases mutant mGal4 showed an inactive phenotype in Drosophila. We suggest that this is because the intrinsic activity of these mGal4 mutants is substantially weaker than wt mGal4 even at permissive temperature in yeast. The further lowering of activity in Drosophila pushes the activity below the threshold required for reporter gene expression resulting in an inactive phenotype. The concluding chapter (Chapter 6) summarizes the conclusions drawn from this entire study and provides insights into possible mechanisms responsible for ts and cs phenotypes. The mutant phenotypes of Gal4 in yeast and in Drosophila suggest that ts phenotypes appear to result from a threshold effect. Such mutations lower the activity and/or level of the protein relative to the wt at all temperatures. Since maximal stability temperatures are rarely in excess of room temperature, with an increase in temperature, the activity of an already marginally active mutant can fall below the threshold required for function resulting in a temperature sensitive phenotype. The strategies we used for producing ts mutants have several advantages over standard approaches of generating ts alleles by random mutagenesis. We anticipate that conclusions of this study would be useful for generation of ts mutants of other globular proteins in diverse organisms. We also show that exposed, functional residues involved in protein: ligand or protein: protein interactions appear to be attractive candidate sites for generating ts mutants that are transferable between organisms. In addition, the active site mutants of fGal4 in Drosophila, which show ts and cs phenotypes depending on the Drosophila promoter chosen for expression, can be used for conditional and reversible expression of a number of other genes using the Gal4-UAS system (Brand and Perrimon 1993).

Conditional gene expression and conditional mutants provide a means to modulate the expression of specific genes and to control the activity of their protein products in vivo, so as to be able to study their effects on the cell. Conditional mutants are functional under one set of conditions, termed ‘permissive’, while under other conditions that are ‘restrictive’, they become non-functional. The wild-type (Wt) is functional under both these conditions. Conditional mutants are especially useful for studying essential or lethal genes in an organism. In the case of temperature-sensitive (ts) and cold-sensitive mutants (cs), by using temperature shift as a condition, the target gene function can be modulated easily, rapidly, reversibly and selectively, at any stage in the life cycle of an organism. Cs mutants are less common and molecular determinants of cs phenotypes are poorly understood.This thesis presents a method for rational elicitation of cold sensitive phenotypes which involves design of partial loss-of-function mutants based solely on amino acid sequence, and coupling of such mutants to a heat responsive promoter to result in cs phenotypes. This approach has been validated for different proteins (CcdB, Gal4, Ura3 and Trp1) in different organisms (E. coli, S. cerevisiae and D. melanogaster). This is a straightforward approach which does not involve complex temperature-dependent, mutational effects.Additional characterization of purified Gal4 mutants by measuring protein thermal stability and DNA binding affinity, as well as measurements of transcript levels by qPCR were carried out, to understand the molecular basis of the cs phenotype. The pBAD series of vectors, containing the PBAD promoter and the araC regulator-activator, are very convenient for cloning and expression purposes. They have been widely for cloning and for graded expression of cloned genes. However, there have been reports of non-uniform gene expression across cells at sub-saturating concentrations of arabinose. This thesis also covers studies on this issue of heterogeneity of expression from PBAD promoters at the single cell level, using stable and degradable GFPs as reporters, in a variety of conditions such as constitutive versus autocatalytic expression of arabinose transporter, presence and absence of arabinose metabolising araBAD genes in the host, and varying time periods of induction. Several single amino acid substitutions which cause folding defects in the protein are seen to give rise to cs or ts phenotypes. The key to understanding the basis of such phenotypes caused by folding defective mutants lies in the folding pathway of the protein. To this end, the folding kinetics of E. coli CcdB were studied, which is prerequisite for understanding the folding defects in various ts/cs mutants of CcdB. This study looks at the folding of a dimeric protein, which is of great interest as it involves conformational changes as well as association steps. The results from various experimental observations, ligand binding studies and simulations lead to the conclusion that CcdB folds via parallel pathways, each involving an unstructured dimeric intermediate, to arrive at its native state. In summary, the thesis covers the research carried out ontuning conditional gene expression, rationally designing conditional mutants and understanding their folding mechanisms.

The Hereditary ataxias (HAs) are a group of heterogenous neurological disorders associated with multiple genetic etiologies and encompassing a wide spectrum of phenotypes, where ataxia is the prominent feature. HAs are characterized by degeneration of Purkinje cell and/or spinocerebellar connections, often associated with defects in additional brain structures, and all patterns of inheritance may occur. Similar to other fields of medical genetics, Next Generation Sequencing (NGS) has entered the HA scenario widening our genetic and clinical knowledge of this condition, but routine NGS applications still miss genetic diagnosis in about two third of patients. In this doctoral study, we applied multi-gene panels to define the molecular basis in 259 patients with a clinical diagnosis of HA and negative to tests for pathological expansion in SCA1, 2, 3, 6, 7, 8, 12, 17 and FXN. We found a positive molecular diagnosis in 25% of patients, whereas a similar number of patients had an uncertain diagnosis due to the presence of either variants of uncertain significance or lack of biological samples to determine segregation among family members. Hence despite a higher positive diagnostic rate compared to similar studies described in literature, a half of patients lacked any indication of the genetic cause of their disease. Using exome sequencing as a second-tier approach in some families, refractory to multi-gene panel analysis, did not significantly improved our diagnostic yield. On the other hand, NGS analysis in our cohort indicated that familial cases were more easily diagnosed rather than sporadic cases, and also that combining massive sequencing with detailed clinical information and family studies increases the likelihood to reach a molecular diagnosis. Among positive patients, we could expand clinical and allelic information in a subgroup of genes offering original description of new mutations and corroborating genetic findings with functional investigations that took advantage of different in vitro or in vivo platforms. In particular, through functional studies in SPG7 knock-down models of Drosophila melanogaster, we remarked that SPG7, whose mutations cause spastic paraplegia type 7, has a critical role in neurons more than in skeletal muscle. The high frequency of p.Ala510Val mutation in SPG7 observed in our cohort as well in similar studies performed elsewhere moved us to develop a humanized knock-in fruit fly model harboring that specific mutation and prepare preliminary characterizations. Similar studies in fruit fly were performed silencing AFG3L2, the gene causing SPAX5 in a child in association with an unusual, relatively milder phenotype. Furthermore, combination of skin fibroblasts and Saccharomyces cerevisiae as models was employed in the genetic characterization of new mutations in a novel recessive HARS-related phenotype whereas primary human cells, yeast and Danio rerio models were used to functionally characterize new HA-related mutations in COQ4. Finally, we could expand the clinical presentation of rare causes of HAs describing new dominant mutations in STUB1 and biallelic variants in RFN216, COQ8A, and ATP13A2. Altogether, studies performed during this doctoral work further underlined the usefulness of NGS in HAs and highlighted how NGS technologies rely on the integrated use of family and clinical studies and different in vitro/in vivo platforms to substantiate molecular findings. The latter platform will be also a tool for future investigations to dissect pathogenesis and to improve therapies.

Many phytopathogenic bacteria use a type III protein secretion system (T3SS) to inject type III effectors into plant cells. In the experiments supported by this one-year feasibility study we investigated type III effector function in plants by using two contrasting bacterial pathogens: Pseudomonas syringae pv. tomato, a necrotrophic pathogen and Pantoea agglomerans, a tumorigenic pathogen. The objectives are listed below along with our major conclusions, achievements, and implications for science and agriculture. Objective 1: Compare Pseudomonas syringae and Pantoea agglomerans type III effectors in established assays to test the extent that they can suppress innate immunity and incite tumorigenesis. We tested P. agglomerans type III effectors in several innate immunity suppression assays and in several instances these effectors were capable of suppressing plant immunity, outputs that are suppressed by P. syringae effectors. Interestingly, several P. syringae effectors were able to complement gall production to a P. agglomerans pthGmutant. These results suggest that even though the disease symptoms of these pathogens are dramatically different, their type III effectors may function similarly. Objective 2: Construct P. syringae mutants in different combinations of type III-related DNA clusters to reduce type III effector redundancy. To determine their involvement in pathogenicity we constructed mutants that lack individual and multiple type III-related DNA clusters using a Flprecombinase-mediated mutagenesis strategy. The majority of single effector mutants in DC3000 have weak pathogenicity phenotypes most likely due to functional redundancy of effectors. Supporting this idea, Poly-DNAcluster deletion mutants were more significantly reduced in their ability to cause disease. Because these mutants have less functional redundancy of type III effectors, they should help identify P. syringae and P. agglomerans effectors that contribute more significantly to virulence. Objective 3: Determine the extent that P. syringae and P. agglomerans type III effectors alter hormone levels in plants. Inhibition of auxin polar transport by 2,3,5-triiodobenzoic acid (TIBA) completely prevented gall formation by P. agglomerans pv. gypsophilae in gypsophila cuttings. This result supported the hypothesis that auxin and presumably cytokinins of plant origin, rather than the IAA and cytokinins secreted by the pathogen, are mandatory for gall formation. Transgenic tobacco with pthGshowed various phenotypic traits that suggest manipulation of auxin metabolism. Moreover, the auxin levels in pthGtransgenic tobacco lines was 2-4 times higher than the control plants. External addition of auxin or cytokinins could modify the gall size in gypsophila cuttings inoculated with pthGmutant (PagMx27), but not with other type III effectors. We are currently determining hormone levels in transgenic plants expressing different type III effectors. Objective 4: Determine whether the P. agglomerans effectors HsvG/B act as transcriptional activators in plants. The P. agglomerans type III effectors HsvG and HsvB localize to the nucleus of host and nonhost plants and act as transcription activators in yeast. Three sites of adjacent arginine and lysine in HsvG and HsvB were suspected to act as Nuclear localization signals (NLS) domains. A nuclear import assay indicated two of the three putative NLS domains were functional NLSs in yeast. These were shown to be active in plants by fusing HsvG and HsvB to YFP. localization to the nucleus was dependent on these NLS domains. These achievements indicate that our research plan is feasible and suggest that type III effectors suppress innate immunity and modulate plant hormones. This information has the potential to be exploited to improve disease resistance in agricultural crops.

Blue mold of apple caused by Penicilliumexpansumis a major postharvest disease. Selection for postharvest disease resistance in breeding programs has been ignored in favor of fruit quality traits such as size, color, taste, etc. The identification of postharvest disease resistance as a heritable trait would represent a significant accomplishment and has not been attempted in apple. Furthermore, insight into the biology of the pathogenicity of P. expansumin apple could provide new approaches to postharvest decay management. Hypothesis: Postharvest resistance of apple to P. expansumcan be mapped to specific genetic loci and significant quantitative-trait-loci (QTLs) can be identified that account for a major portion of the population variance. Susceptibility of apple fruit to P. expansumis dependent on the ability of the pathogen to produce LysM effectors that actively suppress primary and/or secondary resistance mechanisms in the fruit. Objectives: 1) Identify QTL(s) and molecular markers for blue mold resistance in GMAL4593 mapping population (‘Royal Gala’ X MalussieversiiPI613981), 2) Characterize the transcriptome of the host and pathogen (P. expansum) during the infection process 3) Determine the function of LysM genes in pathogenicity of P. expansum. Methods: A phenotypic evaluation of blue mold resistance in the GMAL4593 mapping population, conducted in several different years, will be used for QTL analysis (using MapQTL 6.0) to identify loci associated with blue mold resistance. Molecular markers will be developed for the resistance loci. Transcriptomic analysis by RNA-seq will be used to conduct a time course study of gene expression in resistant and susceptible apple GMAL4593 genotypes in response to P. expansum, as well as fungal responses to both genotypes. Candidate resistance genes identified in the transcriptomic study and or bioinformatic analysis will be positioned in the ‘Golden Delicious’ genome to identify markers that co-locate with the identified QTL(s). A functional analysis of LysM genes on pathogenicity will be conducted by eliminating or reducing the expression of individual effectors by heterologous recombination and silencing technologies. LysMeffector genes will also be expressed in a yeast expression system to study protein function. Expected Results: Identification of postharvest disease resistance QTLs and tightly-linked genetic markers. Increased knowledge of the role of effectors in blue mold pathogenic

The original objectives of the research were: i. To study the role of GA in anther development, ii. To manipulate GA and/or GA signal transduction levels in the anthers in order to generate male sterility. iii. To characterize the GA signal transduction repressor, SPY. Previous studies have suggested that gibberellins (GAs) are required for normal anther development. In this work, we studied the role of GA in the regulation of anther development in petunia. When plants were treated with the GA-biosynthesis inhibitor paclobutrazol, anther development was arrested. Microscopic analysis of these anthers revealed that paclobutrazol inhibits post-meiotic developmental processes. The treated anthers contained pollen grains but the connective tissue and tapetum cells were degenerated. The expression of the GA-induced gene, GIP, can be used in petunia as a molecular marker to: study GA responses. Analyses of GIP expression during anther development revealed that the gene is induced only after microsporogenesis. This observation further suggests a role for GA in the regulation of post-meiotic processes during petunia anther development. Spy acts as a negative regulator of gibberellin (GA) action in Arabidopsis. We cloned the petunia Spy homologue, PhSPY, and showed that it can complement the spy-3 mutation in Arabidopsis. Overexpression of Spy in transgenic petunia plants affected various GA-regulated processes, including seed germination, shoot elongation, flower initiation, flower development and the expression of a GA- induced gene, GIP. In addition, anther development was inhibited in the transgenic plants following microsporogenesis. The N-terminus of Spy contains tetratricopeptide repeats (TPR). TPR motifs participate in protein-protein interactions, suggesting that Spy is part of a multiprotein complex. To test this hypothesis, we over-expressed the SPY's TPR region without the catalytic domain in transgenic petunia and generated a dominant- negative Spy mutant. The transgenic seeds were able to germinate on paclobutrazol, suggesting an enhanced GA signal. Overexpression of PhSPY in wild type Arabidopsis did not affect plant stature, morphology or flowering time. Consistent with Spy being an O-GlcNAc transferase (OGT), Spy expressed in insect cells was shown to O-GlcNAc modify itself. Consistent with O-GlcNAc modification playing a role in GA signaling, spy mutants had a reduction in the GlcNAc modification of several proteins. After treatment of the GA deficient, gal mutant, with GA3 the GlcNAc modification of proteins of the same size as those affected in spy mutants exhibited a reduction in GlcNAcylation. GA-induced GlcNAcase may be responsible for this de-GlcNAcylation because, treatment of gal with GA rapidly induced an increase in GlcNAcase activity. Several Arabidopsis proteins that interact with the TPR domain of Spy were identified using yeast two-hybrids screens. One of these proteins was GIGANTEA (GI). Consistent with GI and Spy functioning as a complex in the plant the spy-4 was epistatic to gi. These experiments also demonstrated that, in addition to its role in GA signaling, Spy functions in the light signaling pathways controlling hypocotyl elongation and photoperiodic induction of flowering. A second Arabidopsis OGT, SECRET AGENT (SCA), was discovered. Like SPY, SCA O-GlcNAc modifies itself. Although sca mutants do not exhibit dramatic phenotypes, spy/sca double mutants exhibit male and female gamete and embryo lethality, indicating that Spy and SCA have overlapping functions. These results suggest that O-GlcNAc modification is an essential modification in plants that has a role in multiple signaling pathways.

Gram-negative plant pathogenic bacteria employ specialized type-III secretion systems (TTSS) to deliver an arsenal of pathogenicity proteins directly into host cells. These secretion systems are encoded by hrp genes (for hypersensitive response and pathogenicity) and the effector proteins by so-called dsp or avr genes. The functions of effectors are to enable bacterial multiplication by damaging host cells and/or by blocking host defenses. We characterized essential hrp gene clusters in the Stewart's Wilt of maize pathogen, Pantoea stewartii subsp. stewartii (Pnss; formerly Erwinia stewartii) and the gall-forming bacterium, Pantoea agglomerans (formerly Erwinia herbicola) pvs. gypsophilae (Pag) and betae (Pab). We proposed that the virulence and host specificity of these pathogens is a function of a) the perception of specific host signals resulting in bacterial hrp gene expression and b) the action of specialized signal proteins (i.e. Hrp effectors) delivered into the plant cell. The specific objectives of the proposal were: 1) How is the expression of the hrp and effector genes regulated in response to host cell contact and the apoplastic environment? 2) What additional effector proteins are involved in pathogenicity? 3) Do the presently known Pantoea effector proteins enter host cells? 4) What host proteins interact with these effectors? We characterized the components of the hrp regulatory cascade (HrpXY ->7 HrpS ->7 HrpL ->7 hrp promoters), showed that they are conserved in both Pnss and Fag, and discovered that the regulation of the hrpS promoter (hrpSp) may be a key point in integrating apoplastic signals. We also analyzed the promoters recognized by HrpL and demonstrated the relationship between their composition and efficiency. Moreover, we showed that promoter strength can influence disease expression. In Pnss, we found that the HrpXY two-component signal system may sense the metabolic status of the bacterium and is required for full hrp gene expression in planta. In both species, acyl-homoserine lactone-mediated quorum sensing may also regulate epiphytic fitness and/or pathogenicity. A common Hrp effector protein, DspE/WtsE, is conserved and required for virulence of both species. When introduced into corn cells, Pnss WtsE protein caused water-soaked lesions. In other plants, it either caused cell death or acted as an Avr determinant. Using a yeast- two-hybrid system, WtsE was shown to interact with a number of maize signal transduction proteins that are likely to have roles in either programmed cell death or disease resistance. In Pag and Pab, we have characterized the effector proteins HsvG, HsvB and PthG. HsvG and HsvB are homologous proteins that determine host specificity of Pag and Pab on gypsophila and beet, respectively. Both possess a transcriptional activation domain that functions in yeast. PthG was found to act as an Avr determinant on multiple beet species, but was required for virulence on gypsophila. In addition, we demonstrated that PthG acts within the host cell. Additional effector genes have been characterized on the pathogenicity plasmid, pPATHₚₐg, in Pag. A screen for HrpL- regulated genes in Pnsspointed up 18 candidate effector proteins and four of these were required for full virulence. It is now well established that the virulence of Gram-negative plant pathogenic bacteria is governed by Hrp-dependent effector proteins. However; the mode of action of many effectors is still unresolved. This BARD supported research will significantly contribute to the understanding of how Hrp effectors operate in Pantoea spp. and how they control host specificity and affect symptom production. This may lead to novel approaches for genetically engineering plants resistant to a wide range of bacterial pathogens by inactivating the Hrp effectors with "plantabodies" or modifying their receptors, thereby blocking the induction of the susceptible response. Alternatively, innovative technologies could be used to interfere with the Hrp regulatory cascade by blocking a critical step or mimicking plant or quorum sensing signals.