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).