UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The mechanism of action of inhibitory domain 1 of GAL4 Hentschel, Claudia Perelli 1995

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1995-0638.pdf [ 7.34MB ]
Metadata
JSON: 831-1.0087015.json
JSON-LD: 831-1.0087015-ld.json
RDF/XML (Pretty): 831-1.0087015-rdf.xml
RDF/JSON: 831-1.0087015-rdf.json
Turtle: 831-1.0087015-turtle.txt
N-Triples: 831-1.0087015-rdf-ntriples.txt
Original Record: 831-1.0087015-source.json
Full Text
831-1.0087015-fulltext.txt
Citation
831-1.0087015.ris

Full Text

THE MECHANISM OF ACTION OF INHIBITORY DOMAIN 1 OF GAL4 b y CLAUDIA PERELLI HENTSCHEL B.Sc.(H.) (Biochemistry), University of British Columbia, 1993 A THESIS SUBMITTED LN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE THE FACULTY OF GRADUATE STUDIES (Department of Biochemistry and Molecular Biology) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA September, 1995 © Claudia Perelli Hentschel In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of The University of British Columbia Vancouver, Canada Date QMM*_ i0_ DE-6 (2/88) A B S T R A C T The mechanisms involved in the regulation of gene expression are of major interest in molecular biological research. Simpler models of lower eukaryotes such as yeast have been classically employed in such studies as a basis for the further characterization of more complex organisms. The galactose catabolism pathway of Saccharomyces cerevisiae provides an ideal model for research on gene regulation. The carbon source available to the yeast dictates the expression of the GAL genes involved in the utilization of galactose in budding yeast. The key regulator in this pathway is the transcriptional activator GAL4. GAL4 is active in the presence of galactose and absence of glucose, the preferred carbon source. When glucose is present, though, GAL4 is rendered inactive by several glucose repression mechanisms. One of these mechanisms acts through the central region of the GAL4 protein itself, which is composed of a glucose-responsive domain and at least three inhibitory domains, one being inhibitory domain 1. The activity of the inhibitory domains has been suggested to be governed by the glucose-responsive domain . With biochemical evidence, I suggest that the mechanism of transcriptional inactivation by inhibitory domain 1 occurs through multimerization of the GAL4 protein. Multimerization disrupts the formation of GAL4 dimers, preventing the transcriptional activator from binding to D N A and rendering the protein inactive. i i T A B L E O F C O N T E N T S Abstrac t ii List of Tables vi List of Figures vii List of Abbreviations ix Acknowledgements x INTRODUCTION 1 1. Eukaryotic Gene Regulation 1 2. Galactose Catabolism in S. cerevisiae 3 3. Induction by Galactose 4 4. Repression by Glucose 6 5. G A L 4 Structure and Function 10 a. D N A Binding 10 b. Transcriptional Activation 13 c. G A L 8 0 Interaction 14 d. G A L 4 Phosphorylation 15 e. G A L 4 Central Region 15 i i i I 6. A i m of Research 17 M A T E R I A L S A N D M E T H O D S 18 1. Ribonuclease Protection Assay 18 2. (3-Galactosidase Assays 19 3. D N A Binding Assays 19 4. Chemical Crosslinking Experiments 22 5. Experiments with E. coli Histidine-tagged G A L 4 Derivatives 23 RESULTS 26 1. Glucose Inhibits G A L 4 Through Its Central Region 26 2. ID1 Is a Strong Inhibitory Domain in Yeast 28 3. ID1 Impairs D N A Binding 31 4. ID1 Prevents Dimerization 42 5. ID1 Promotes Multimerization 46 6. ID1 Appears to Act by Itself 54 7. ID1 is Homologous to Other Proteins 66 DISCUSSSION 71 1. Glucose Inhibits G A L 4 Through Its Central Region 71 i v 2. ID1 Is a Strong Inhibitory Domain Homologous to Other Proteins 72 3. ID1 Impairs D N A Binding 73 4. ID1 Prevents Dimerization and Promotes Multimerization 74 5. ID1 Does Not Require Other Cellular Factors 76 6. Comparison of ID1 to Other Known Inhibitory Domains 76 7. Model of the Mechanism of Action of ID1 77 REFERENCES 80 v LIST OF TABLES Table 1. fi-Galactosidase Assay of G A L 4 Derivatives 30 Table 2. Proteins Homologous to ID1 of G A L 4 68 v i LIST OF FIGURES Figure 1. Model of Transcriptional Activation 2 Figure 2. Galactose Catabolism Pathway in S. cerevisiae 5 Figure 3. Mechanisms of Glucose Repression 7 Figure 4. Functional Domains of G A L 4 and Their Positioning 11 Figure 5. Glucose Represses G A L 4 Through Its Central Region 27 Figure 6. G A L 4 Derivatives 29 Figure 7. Expression of G A L 4 Derivatives in E. coli 32 Figure 8. Expression of G A L 4 Derivatives in Wheat Germ Extracts 33 Figure 9. ID1 Impairs D N A Binding in E.coli in E M S A 35 Figure 10. ID1 Impairs D N A Binding of In Vitro Translated G A L 4 in E M S A 37 Figure 11. ID1 Prevents D N A Binding in E. coli to a G A L 4 Biotinylated Oligo 40 Figure 12. ID1 Prevents D N A Binding in Wheat Germ Extracts to a G A L 4 Biotinylated Oligo 41 Figure 13. ID1 Prevents the Formation of Homodimers 43 Figure 14. In Vitro Co-translation of G A L 4 Derivatives in Wheat Germ Extracts 45 v i i Figure 15. ID1 Prevents the Formation of Heterodimers Figure 16. Limited Crosslinking Reveals that ID1 Causes the Formation of Multimers in Wheat Germ Extracts Figure 17. Limited Crosslinking Reveals that ID1 Causes the Formation of Multimers in E. coli Figure 18. Batch Purification of Histidine-tagged G A L 4 Derivat ives Figure 19. Column Purification of the G A L 4 Derivative his-DBD Figure 20. Column Purification of the G A L 4 Derivative h i s - D B D - A R 2 Figure 21. Column Purification of the G A L 4 Derivative h i s - D B D - I D l - A R 2 Figure 22. E M S A of Batch Purified Histidine-tagged G A L 4 Derivat ives Figure 23. E M S A of Column Purified Histidine-tagged G A L 4 Derivat ives Figure 24. Alignment of Protein Sequences Homologous to ID1 of G A L 4 Figure 25. Proposed Mechanism of Action of ID1 of G A L 4 v i i i LIST OF ABBREVIATIONS B S A bovine serum albumin j B M V brome mosaic virus DEPC diethylpyrocarbonate DMSO dimethylsulfoxide DSP dithiobis(succinimidylproprionate) D T T dithiothreitol E C L enhanced chemiluminescence E D T A ethylenediaminetetraacetate E M S A electrophoretic mobility shift assay Hepes 4-(hydroxyethyl)- l -piperazinethanesulfonic acid I P T G isopropylthiogalactoside m A b monoclonal antibody N P - 4 0 NonidetP-40 P A G E polyacrylamide gel electrophoresis P M S F phenylmethylsulfonyl fluoride SDS sodium dodecylsulfate T B P T A T A - b i n d i n g protein ix A C K N O W L E D G E M E N T S I would first like to thank my parents and family for their constant support and encouragement. I am grateful to my husband Steve for always being there for me. I thank Todd Schindeler very much for making the DSP reagent for me, and Logan Donaldson for . the help with the computer homology search and sequence alignment. I also sincerely thank all the members of the Sadowski lab for their help, advice and friendship throughout. I would also like to thank my supervisor, Ivan Sadowski, for his help and guidance. Finally, I would like to acknowledge the help of my committe members including Michel Roberge, Louise Glass, and Ivan Sadowski, for their suggestions on my research and my thesis. x I N T R O D U C T I O N 1. E u k a r y o t i c T r a n s c r i p t i o n a l Regula t ion A n interesting aspect of molecular biological research focuses on how genes are variably expressed in response to extracellular or developmental signals. These signals generate complex biochemical changes to transduce the signal to the nucleus, and to transcriptionally activate or repress one or more genes. In turn, the change in expression of a gene, primarily at the level of transcription, then allows an appropriate response to the signal. Transcriptional activators and repressors are mediators between the signal and the response, controlling the level of gene expression. A model of how transcriptional activation occurs is depicted in figure 1. The function of transcriptional activators is to recognize and bind to specific D N A sites called enhancer elements or upstream activating sequences (UAS) found upstream of a gene. Once bound, transcriptional activators interact with proteins of the general transcription initiation complex found at the T A T A box, allowing R N A polymerase II to transcribe D N A into R N A (64, 65, 66, 79). Traditionally, it was thought that in all eukaryotes the basal transcription complex was assembled at the promoter in a stepwise fashion (11). Recently, though, it has been shown that in the yeast Saccharomyces cerevisiae, this complex is largely preassembled and termed the R N A polymerase II holoenzyme (42). Koleske and Young (1994) showed that the holoenzyme consists of R N A polymerase II or B, suppressors of R N A polymerase B or SRB's , and the general factors TFIIB, TFIIH, TFIIF. This protein complex then joins TFIIA, TFIID and TFIIE to form the complete general transcription initiation machinery at the promoter. Transcriptional activators function by contacting components of the initiation complex. For example, 34 carboxy-terminal amino acids of the transcription factor G A L 4 have been shown to interact with the T A T A - b i n d i n g protein (TBP) and with other coactivators known as SUG1 and A D A 2 (58). In addition to the general initiation complex components, other proteins called 1 RNA POLYMERASE II INITIATION COMPLEX URS TATA BOX Figure 1. Model of Transcriptional Activation Transcriptional activation, turning the expression of a gene from off to on, requires several steps. A transcriptional activator binds to its U A S (upstream activating sequence) site on the D N A upstream of a specific gene, changing the conformation of the promoter. The R N A polymerase II general initiation complex is recruited to the T A T A box site. The transcriptional activator and the R N A polymerase II general initiation complex interact to allow transcriptional activation of a specific gene. 2 adaptors or coactivators are also found to be essential for transcription, such as the protein G A L 1 1 (31). The mechanism of transcriptional repression is not clear. However, transcriptional repressors have been found to bind D N A at repression sequences or upstream repression sequences (URS) in the promoter of a gene and block transcriptional activation by a largely uncharacterized mechanism (21). Transcriptional activators and repressors have also been shown to play an important role in changing chromatin structure (1, 13, 15), an important aspect in allowing the necessary stimulatory protein-protein interactions to occur (64). Transcriptional activators are highly modular in nature, allowing researchers the flexibility to swap domains between them while retaining protein function also in other organisms (65, 66, 80). Transcription factors usually contain at least three functional domains to allow for D N A binding, transcriptional activation and regulation activities of the protein. Several structural D N A binding motifs have been identified in both prokaryotic and eukaryotic transcriptional regulators such as zinc finger, leucine zipper, helix-loop-helix and homeodomain motifs (33). Similarly, there are several categories of activating domains such as glutamine rich, proline rich or acidic domains, the latter containing a high proportion of negatively charged residues (65). The regulation of transcription factors usually occurs by either protein-protein interactions or post-translational modification, such as phosphorylation. For example, the S. cerevisiae transcription factor G A L 4 , which responds to galactose and glucose as extracellular signals, contains a zinc cluster D N A binding motif (57), and two acidic activating domains (56). G A L 4 is regulated in part by phosphorylation (71), by the binding of the negative regulator G A L 8 0 (55), and by a glucose-responsive domain which appears to govern the activity of its inhibitory domains (78). 2. Galactose Catabolism in S. cerevisiae The galactose catabolism pathway of S. cerevisiae is an excellent model for studying the mechanisms of gene regulation. 3 Yeast mutants unable to utilize galactose were first isolated almost forty years ago (69). Since then, the pathway involved and the key players have been defined (17, 18) and continue to be studied intensely. The galactose catabolism pathway, depicted in figure 2, allows yeast to utilize galactose or melibiose, a galactose-glucose disaccharide, as carbon sources, converting them to glucose in order to access glycolysis. The enzymes involved in this conversion are encoded by genes called the GAL genes of the Leloir pathway whose expression, with the exception of GAL5, is controlled by the transcriptional activator GAL4 (18, 40, 41). GAL4 itself is subject to further regulation by the presence or absence of glucose or galactose. In the presence of galactose and the absence of glucose, the preferred carbon source, GAL4 is active and the GAL genes are transcribed at high levels; conversely, in the absence of galactose or in the presence of glucose, GAL4 is inactive and the GAL genes are repressed. 3. Induction by Galactose As mentioned, GAL4 activity is induced in the presence of galactose but strongly repressed in the presence of glucose. Transcriptional induction of the galactose catabolic pathway genes occurs within minutes of addition of galactose to a yeast culture (17). The actual induction mechanism is not yet fully understood, but it is known to require the interplay of the gene products GAL2, GAL3, GAL1, GAL4 and GAL80. GAL2 acts as the galactose permease, allowing galactose to enter the yeast cell from the medium (59). Once galactose is inside the cell, the GAL3 protein plays an important role in the induction pathway. In fact, yeast strains carrying gal3 mutants are either slow inducers or non-inducers of the galactose catabolism pathway (4, 6). GAL1, which has considerable sequence and functional similarity to GAL3, is capable of replacing GAL3 function (4, 6). GAL3 has been proposed to catalyze the conversion of galactose inside the cell to an inducer molecule, whose target would be the GAL4-GAL80 protein complex. GAL80 binds to GAL4 and acts as a negative regulator to keep GAL4 inactive; thus, 4 c h 2 o h 4L^L Glu. Gal. (out) GAL2 permease \ MEL1 a-galactosidase HO Gal. (in) GAL1 kinase HO Gal-1-P GAL7 transferase HU^ I^ O- UDP H O UDP-Glu. , o H o H 0 UDP-Gal. Glu-1-P GAL5 mutase u- I H O ^ L ^ O H OH Glu-6-P >^_^Jr O C H 2 H O > L H B J ' O H O H Melibiose GAL10 epimerase Glycolysis Figure 2. Galactose Catabol ism Pathway in S. cerevisiae The enzymes in this pathway are alpha-galactosidase ( M E L 1 ) , galactose permease ( G A L 2 ) , galactokinase ( G A L 1 ) , galactose- l -phosphate ur idyl t ransferase ( G A L 7 ) , u r id ine d iphosphoglucose-4-epimerase ( G A L 1 0 ) , and phosphog lucomutase ( G A L 5 ) . 5 perturbation of this protein complex inactivating G A L 8 0 would allow G A L 4 to be activated under inducing conditions (4, 6). In fact, gal80~ strains do not require induction. However, since the inducer molecule has not been identified, this model has been modified to consist of a galactose-dependent conversion of G A L 3 and G A L 1 to forms able of relieving G A L 8 0 negative regulation of G A L 4 (5). Without G A L 8 0 inhibition, G A L 4 becomes able to activate transcription of the GAL genes. 4. Repression by Glucose As found in many organisms from prokaryotes to eukaryotes, glucose elicits a stringent repression of several sets of genes in budding yeast (70). One of the sets of genes subject to glucose repression is the GAL gene ensemble (83). This . regulation is energetically advantageous to the cell since glucose can enter glycolysis directly as a carbon source, whereas galactose requires an additional pathway and enzyme activities to first convert it into glucose in order to be utilized as a carbon source. Glucose repression of the GAL genes is a slow process compared with induction, perhaps because the G A L 4 protein is quite stable (34). It is also a more complex process, requiring the interplay of many proteins and signals. A summary model of these glucose repression mechanisms is shown in figure 3. There appear to be four main glucose repression mechanisms (36, 78). The first acts at the level of induction. The expression of both GAL2 and GAL3 is repressed in glucose. The G A L 2 permease is also inactivated by glucose through proteolysis (83). Also, the negative regulator G A L 8 0 appears to work better in the presence of glucose, perhaps as a consequence of lack of potential induction by G A L 3 (43). This mechanism acts at the binding sites for G A L 4 called upstream activating sequences for galactose ( U A S G ) of galactose-inducible genes, and shows that even if galactose is present, glucose is used preferentially. 6 Galactose O Glucose Repression w Mechanism GAL2 H-0 UAS G URS G Figure 3. Mechanisms of Glucose Repression The glucose repression mechanisms that affect the G A L genes are numbered 1 to 4. Mechanism 1 affects induction, by inhibiting G A L 2 and G A L 3 and enhacing G A L 8 0 activity. Mechanism 2 inhibits transcription of G A L 4 by allowing the MIG1-SSN6-TUP1 repression complex to bind to the U R S G sequence found in the G A L 4 promoter. Mechanism 3 inhibits transcription of G A L 4 regulated genes, like G A L 1 shown above, by allowing the MIG1-SSN6-TUP1 repression complex to the U R S G sites found upstream of G A L 4 regulated genes. Mechanism 4 inhibits G A L 4 activity through the central region of G A L 4 . 7 Glucose also represses GAL gene expression by reducing the level of the G A L 4 activator itself. The GAL4 promoter contains G C rich, upstream repression sequences called U R S G - The MIG1 protein binds to the U R S G (28) and attracts the SSN6-TUP1 repressor complex (82), inhibiting the expression of the GAL4 gene. This inhibition accounts for a powerful and long-term, slow glucose repression effect. Powerful in that G A L 4 binds cooperatively to its U A S G binding sites (25, 38), and thus a slight change in the amount of G A L 4 protein in the cell causes a pronounced effect. Long-term and slow in that the G A L 4 protein is quite stable (36), and thus would remain in the cell even without the synthesis of new G A L 4 molecules. A third mechanism by which glucose stringently regulates G A L gene expression is by inhibiting G A L 4 function through U R S G elements found at the promoters of galactose inducible genes, such as the GAL1 promoter (21). This recruits the binding of the MIG1 protein as is the case for the GAL4 promoter, and, in concert with the SSN6-TUP1 repression complex, these proteins are able to inhibit transcription by an as of yet uncharacterized mechanism. The GAL1 promoter is believed to contain two and possibly three such U R S G elements (22). This mechanism is quite rapid, occurring within minutes of glucose addition (36). The last mechanism of glucose repression acts through the G A L 4 protein itself. Stone and Sadowski (1993) showed that the central part of G A L 4 contains a glucose-responsive domain termed glucose response domain or G R D that appears to control the activity of three inhibitory domains, also in the central region of G A L 4 , which prevent G A L 4 activity. These observations were made by removing all other known glucose repression mechanisms and still observing 10-fold repression by glucose on G A L 4 activity in only twenty minutes. Also, when fused to the chimeric activator L e x A - V P 1 6 , the G A L 4 central region can confer glucose responsiveness to this activator, which is otherwise unresponsive to glucose. Furthermore, if the G R D is removed and only the inhibitory domains of G A L 4 are 8 fused to L e x A - V P 1 6 , inhibition becomes constitutive, regardless of carbon source. This concurs with the idea that glucose controls the G R D , and that, in turn, the G R D controls the activity of the inhibitory domains. Despite these four main mechanisms, there are several other proteins that play a role in glucose repression of GAL gene expression in an intricate and not yet well defined signalling system. Much of this work has been through genetic findings, but more biochemical work is underway to uncover the details involved. These other proteins include: H X K 2 , GRR1, R E G 1 , G A L 8 2 , G A L 8 3 , SNF1, and N G G 1 . Mutations in H X K 2 , a hexokinase, were found to result in a loss of glucose repression (54). GRR1 is believed to act on the glucose transporters, acting in a different class of glucose repression genes than R E G 1 , G A L 8 2 and G A L 8 3 (19); R E G 1 appears to act upstream of G A L 8 2 and G A L 8 3 , in an inhibitory way towards these proteins (19). The serine, threonine protein kinase SNF1, in contrast, is required for expression of glucose repressible genes, suggesting a role for phosphorylation in this intricate regulation (14). SNF1 is believed to act downstream of G A L 8 2 and G A L 8 3 and upstream of the MIG1, SSN6, T U P I repressor complex, but its substrates have not yet been defined. It is interesting to note that this repressor complex has been found to contain a D N A binding component, M I G 1 , which directly interacts with an adaptor, SSN6, which in turn interacts with T U P I , the component believed to carry out the actual repression (84). Also, MIG1 has been found to be phosphorylated when isolated from cells grown in glucose. The MIG1 kinase, though, is still unknown, as is the significance of the phosphorylation of MIG1 in glucose (82). Finally, another protein involved in glucose repression is N G G 1 , which was shown genetically to be required for glucose repression (9) and through two-hybrid data to interact with parts of G A L 4 (8). Once again, though, the exact mechanism involved is still unknown. Glucose repression of the galactose-regulated genes, then, involves a series of mechanisms and players to achieve stringent regulation of gene expression. 9 5. G A L 4 Structure and Function As mentioned, transcriptional activators are highly modular. G A L 4 is certainly no exception as can be seen in figure 4. GAL4 was first cloned and analyzed in the early 1980's (37, 44), and subsequently further characterized by revealing its D N A sequence, a 2.8 kilobase polyadenylated R N A transcript, a predicted 881 amino acid encoded protein with a molecular weight of approximately 100 kilodaltons (KDa) (45). Since then, many groups have unveiled a more detailed description of the structure and function of this transcriptional activator. G A L 4 is now widely used as a tool in methods such as the two-hybrid system of protein-protein interaction (20), directional binding and selectable expression of fusion proteins, as well as many other techniques (72, 73). a. DNA Binding Much work has been carried out to define the D N A binding domain of G A L 4 . G A L 4 was determined to bind to U A S G sequences found upstream of galactose-inducible genes. For example, the divergent GAL1-10 promoter bears four such U A S G elements, each consisting of a 17 base pair sequence with dyad symmetry and the following conserved motif 5 ' - C G G N l l C C G - 3 ' (7, 27). As suggested by the dyad symmetry of the 17 mer, G A L 4 binds to each site as a dimer (12). The region of D N A binding recognition and specificity was mapped to amino acids 1-65 of G A L 4 , whereas a dimerization function is provided by amino acids 65-94 (12). Commonly, the first 147 amino acids, which also contain the nuclear localization signal (residues 1-74) (76), are used as the entire G A L 4 D N A binding domain, lacking transcriptional activation (39). The D N A binding motif of G A L 4 has been classified as a Z n 2 C y s 6 cluster with six cysteine residues coordinating two zinc ions in order to bind D N A . Recently, the structure of a 65 residue, N-terminal fragment of G A L 4 and its binding site has been solved 1 0 GAL4 1 100 200 I 1 300 400 500 600 700 1 1 1 1 1 800 900 1 1 1 1 1 1 1 1 1 i 1 m m ID1 ID2 ID3 GRD mm DNA Binding Spedficity — Dimerizatiori — - : _ J -Cooperative DNA Binding Transcriptional Activation GAL80 Interaction -Phosphorylation sites Glucose Response Inhibition Inhibitory Domain 1 _ i.— h Figure 4. Functional Domains of G A L 4 and Their Positioning The locations of the functional domains of G A L 4 with respect to the amino acid sequence of the protein are represented by the shaded blocks beside the specified function. The abbreviations used in the linear representation of the G A L 4 protein stand for: D N A binding domain (DBD), activating region 1 and 2 (AR1 and AR2) , inhibitory domain 1, 2, and 3 (ID1, ID2, and ID3), and glucose response domain (GRD). 1 1 by X-ray crystallography (57). This study reconfirmed that G A L 4 binds as a dimer, and showed that the conserved C C G triplet at each end of the D N A site is recognized in the major groove by the Z n 2 C y s 6 cluster. The two fold symmetry is explained by a coiled-coil weak dimerization region perpendicular to the minor groove. The dimerization region is connected by a linker region, which in the case of G A L 4 spans 11 base pairs over the minor groove. The findings of this study were confirmed by the results obtained by an N M R solution of the D N A binding domain of G A L 4 binding to half of the U A S G site as a monomer (2). The linker region in the G A L 4 D N A binding domain is important in the in vivo recognition of the activator's specific sites. In fact, other fungal activators also contain the same zinc requiring D N A binding motif, such as L A C 9 , the lactose/galactose regulator of Kluveromyces lactis which binds to the same site, PPR1, which regulates the expression of genes involved in pyrimidine biosynthesis in S. cerevisiae and binds to a C C G triplet separated by 6 base pairs instead of 11, and P U T 3 , which is responsible for controlling genes involved in proline utilization and has a linker region of 10 base pairs (57, 67). Another important consideration about G A L 4 D N A binding is that it does so in a cooperative fashion (25, 38); the exact location of the cooperativity domain of G A L 4 is currently under study, but is suggested to be in the C-terminus of the protein (32). This is especially important for a rapid and potent induction response of G A L 4 , a weakly expressed protein (29, 45). A final aspect to consider about G A L 4 binding to D N A is whether it is always bound at its site, regardless of carbon source available, or whether it is unbound when repressed by glucose. The evidence on this matter is contradictory, with experiments pointing to both results (26, 51). These results may be complicated by the fact that the expression of the GAL4 gene itself is glucose repressed, 1 2 and thus, in the presence of glucose, G A L 4 may simply be depleted from the cell. b. Transcriptional Activation D N A binding is a necessary but insufficient step for transcriptional activation by G A L 4 , as was noted by Keegan et al. (1986). In fact, G A L 4 contains two acidic domains that are able to activate transcription when fused to the D N A binding domain of the protein; these are termed activating region 1 (AR1, residues 149-238) and activating region 2 (AR2, residues 768-881) (56). These activation domains were characterized through a series of deletions of G A L 4 . In addition, there is a third, small activation domain termed activating region 3 (AR3, residues 94-106) that contributes to transcriptional activation in vitro (49). G A L 4 ' s activating regions belong to a family of acidic activating domains that many other transcription factors also possess; these all share a high proportion of acidic residues. Mutants of G A L 4 have been characterized which indicate that the high acidic content is important in the potency of the activation. (23, 24). These acidic residues also appear to be arranged in an a helical conformation, although some studies contradict these findings by showing that these residues need not be acidic and may not be a helical but may form (3 sheet conformations (47, 85). Nevertheless, these domains are presumed to contact one or more of the general initiation factors, anchored near the promoter by the D N A binding component of the transcription factor. In fact, it was recently shown that the carboxy-terminal 34 amino acids of G A L 4 can interact in vitro with T B P and other general yeast coactivators, such as SUG1 and A D A 2 (58). Furthermore, another mechanism by which the activating domains are involved in enhancing transcription is by increasing the ability of basal transcription factors to occupy the promoter region, allowing preinitiation complexes to form after nucleosome assembly (86). Another important aspect of transcriptional activation potency was 1 3 observed recently to be an interplay between additional binding sites for the transcription factor and the additional amounts of transcriptional activation domains in an activator, at least in vitro (61). Finally, as was observed with the positioning of binding sites and respective D N A binding domains of transcription factors in other organisms, the acidic activating regions of G A L 4 are also functional in other eukaryotic systems, such as mammalian cells (24). A conserved mechanism of transcriptional activation thus exists throughout eukaryotes, allowing the flexibility to use the activating regions of G A L 4 as research tools (72, 73). c. GAL80 Interaction It was noticed early on that G A L 8 0 is a negative regulator of GAL gene expression (17, 18), but it was not until later that G A L 8 0 was excluded from being one of the mediators of glucose repression (81). GAL80 was cloned in the early 1980's (37, 44) and was determined to be the negative regulator that acts upon G A L 4 itself, a GAL80 null mutation resulting in constitutive GAL expression in the absence of glucose (55). By deletion experiments, M a and Ptashne (1987) mapped the region responsible for G A L 8 0 binding to the 30 carboxy-terminal amino acids of the G A L 4 protein. Furthermore, it was found that G A L 4 and G A L 8 0 coimmunoprecipitate and copurify in an equimolar ratio, further strengthening the notion that these two proteins bind (16, 53, 88). Three main functional domains were found in the G A L 8 0 protein which allow it to function as a negative regulator: a repression domain, an inducer interaction domain, and a nuclear localization domain (60). Interestingly, GAL80 expression is G A L 4 -induced in the presence of galactose, the GAL80 promoter bearing one U A S . G , but appears to be at a basal level in glucose or glycerol, unlike other G A L 4 regulated genes (50). This regulation is probably necessary for the stringent inactivation of G A L 4 by G A L 8 0 in non-inducing conditions. 14 Opposing views exist on the matter of whether G A L 8 0 is always bound to G A L 4 , the G A L 4 - G A L 8 0 complex being conformationally altered upon induction by galactose, or whether G A L 8 0 simply dissociates from G A L 4 under such conditions. Earlier views favoured the dissociation mechanism (55, 63), with G A L 8 0 masking an essential part of G A L 4 ' s A R 2 , but more recent results suggest that the conformational change mechanism is correct in vivo (46, 62). d . G A L 4 P h o s p h o r y l a t i o n G A L 4 is phosphorylated on several serine residues (71), phosphorylation being a possible regulatory mechanism of protein activity. Unfortunately, the mechanisms involved and the possible kinase or kinases are not yet clearly defined. It is known, however, that, unlike most transcription factors, G A L 4 phosphorylation is acquired as a consequence of activity after induction and not as a requirement for transcriptional activation (62, 74). Also, the general transcriptional regulator G A L 11 was observed to augment G A L 4 dependent-transcription by maintaining a phosphorylated form of G A L 4 (52). A n important phosphorylation site appears to be Ser 699, which, when mutated to A l a renders G A L 4 a very slow inducer (71). Phosphorylation of Ser 699 appears to prevent G A L 8 0 from repressing G A L 4 (71). e. G A L 4 C e n t r a l Region Surprisingly, until 1993, no function had been assigned to the large central region of G A L 4 (CR, residues 239-767), which comprises over 60% of the entire protein. Through a series of deletion experiments and fusion protein expression and function studies, Stone and Sadowski (1993) were able to show that this region is comprised of at least three inhibitory domains, termed inhibitory domain 1 (ID1, residues 320-412), inhibitory domain 2 (ID2, residues 412-478) and inhibitory domain 3 (ID3, residues 554-585), and of a 1 5 glucose-responsive domain ( G R D , residues 600-767), as discussed above. The conclusions about the G A L 4 C R originated from the observations that either AR1 or A R 2 , when fused to the D N A binding domain, can activate transcription in the absence of the C R (23, 24, 56), but transcription is severely impaired when either AR1 or A R 2 is deleted, suggesting an inhibitory, regulatory role for the C R of G A L 4 (56, 74, 78). These findings were confirmed by deletion experiments of G A L 4 as well as with experiments employing the heterologous transcriptional activator L e x A - V P 1 6 (78). When the central region of G A L 4 is fused to LexA-VP16 , this activator becomes responsive to glucose when it normally is unaffected by the carbon source available. Also, by deleting G R D and fusing the rest of the central region to L e x A - V P 1 6 , the inhibition becomes constitutive, regardless of carbon source, a result which also seen with the deletion of the G R D from G A L 4 itself. The above results suggest two conclusions. They reconfirm the idea that G A L 4 contains a glucose-responsive domain, G R D , which has indeed been shown to be responsible for one of the G A L 4 glucose repression mechanisms. They also suggests a model whereby the glucose controlled G R D is in turn responsible for regulating the inhibitory domains, which should be active in glucose and inactive otherwise. The exact mechanism by which this regulation occurs is still unknown, but it was suggested that the G R D and the inhibitory domains interact, whether directly or through other intermediary protein or proteins. This interaction would allow the G R D to keep the inhibitory domains from functioning in the absence of glucose. Upon glucose addition, though, this interaction could be disrupted, the inhibitory domains becoming active to prevent transcriptional activation (77, 78). This model is supported by several missense mutants that have been found between residues 320-520 of G A L 4 , the inhibitory region, which inactivate the protein (35). This inactivation can be explained by the disruption of GRD-inhibitory domains interaction, rendering inhibition constitutive. Exactly how 1 6 inhibition is achieved is largely unknown, but disruption of D N A binding seems to be a likely mechanism, the strongest evidence presented so far being that L e x A derivatives bearing G A L 4 C R fusions bind more weakly to D N A when extracted from yeast grown in glucose as compared to other carbon sources (77). No homology has been found between the three inhibitory domains, and some preliminary results suggest that they have different mechanisms of action (32). O f the three inhibitory domains, ID1 was found to be the strongest and to be quite flexible in terms of its location in a fusion protein while still retaining activity (77). ID1 is also the site of three point mutants at amino acids 322, 331 and 352 which, along with several D N A binding mutants and one at residue 511, can severely hinder G A L 4 activity (35). These mutants suggest that ID1 is an important regulatory region. This domain, residues 320-412, also contains a high degree of homology with other fungal transcription factors involved in the regulation of metabolic pathways such as L A C 9 , PUT3, PPR1, L E U 3 , NIRA, THI1 and NIT4, as well as other proteins such as C D C 6 , involved in initiation of D N A replication, and PDR3, involved in pleiotropic drug resistance. The significance of this homology is not yet understood, but it suggests the interesting possibility of a conserved functional mechanism by which inhibition can be achieved. 6. A im of Research The aim of my research was to determine the mechanism of action of inhibitory domain 1 (ID1) of G A L 4 . With biochemical evidence, I suggest that the transcriptional inhibition activity posessed by ID1 occurs through a protein multimerization mechanism. This protein multimerization prevents efficient formation of G A L 4 dimers, and thus impairs the D N A binding ability necessary for the activity of the G A L 4 transcriptional activator. 1 7 M A T E R I A L S A N D M E T H O D S 1. R i b o n u c l e a s e P r o t e c t i o n Assay R N A Extractions: Yeast strain YT6::171 derived from Y T 6 (MATa, gal4, gal80, ura3, his3, ade2, adel, lys2, trpl, aral, leu2, met) (31) with a wild-type GALl-lacZ reporter gene bearing both U A S G and U R S G (87) was used for R N A extractions. YT6::171 was transformed by the lithium acetate method with A R S - C E N plasmids yCPG4trp , yCPG4242 and ycplac22 (constructed by I. Sadowski) expressing wild type G A L 4 , G A L 4 without the central region and a vector control, respectively, from G A L 4 ' s own promoter. 20 ml yeast cultures were grown from these transformants in minimal medium lacking tryptophan and uracil and having glycerol, lactic acid and ethanol as carbon sources. At OD600=l-5, 2% glucose was added to half of the cultures, and all cultures were grown for another hour. 10 ml of each culture (either with or without glucose) was for R N A extraction. R N A was extracted by pelleting the cells, washing with 1 ml sterile, deionized water, resuspending in 500 u L Ultraspec R N A (Biotecx) and adding 400 u L glass beads before vortexing 40 minutes at 4 ° C . 100 j iL chloroform was then added. The samples were incubated for 5 minutes on ice, centrifuged 5 minutes at 4 ° C , and the top aqueous layer removed for isopropanol freezer precipitation overnight. The pellet was then recovered by centrifugation, washed with 70% ethanol, and resuspended in water treated with diethylpyrocarbonate (DEPC) . A sample was taken to determine the concentration by A 2 6 0 reading. Ribonuclease Protection Assay: The lacZ and actin probes used in this assay were derived from linearized plasmids pIS009 and pIS015 (constructed by I. Sadowski). Each template was transcribed in vitro with SP6 R N A polymerase and gel purified as described in the manual for the RPAII kit sold by Ambion. The rest of the assay was also performed as described in the RPAII kit manual, using 1 pg R N A per reaction. 1 8 2. (3-Galactosidase A s s a y s P-galactosidase assays were performed on strain YT6::171. This strain was transformed with high copy 2|i plasmids p M A 2 1 0 , pMA241 , pMA236 , pMA200 (contructed by J. Ma), and p M A I D l (derived from pMA210 , substituting the Xho I-Mlu I fragment with that of plasmid pkwlO(Bam)A312:A421 constructed by G . Stone). These plasmids express wild type G A L 4 , G A L 4 1-147 D N A binding domain (DBD), G A L 4 D B D fused to AR2, a vector control, and G A L 4 D B D fused to ID1 fused to A R 2 , respectively, from an ADH1 promoter. The transformants were grown in histidine - , uracil" dropout medium. The assays were performed as described by Himmelfarb et al. (1990), the results expressed as Miller units (1 unit= 1 pmol o-nitrophenol consumed/ min/ mg protein). 3. D N A B i n d i n g Assays Expression in E. coli: BL21 E. coli cells which express T7 R N A polymerase were transformed with plasmids p T M C 1 4 7 , p T M C l d 7 6 8 , p T M C l - 1 4 7 - I D l - I I , p T M C l d 6 8 3 , and pRB451 (contructed by I. Sadowski, I. Sadowski, I. Sadowski, T. Kang, and R. Brent, respectively) expressing G A L 4 ' s D B D , D B D fused to A R 2 , D B D fused to ID1 fused to A R 2 , D B D fused to G R D fused to AR2 , and a vector control, respectively, from a tac promoter. 50 ml of L B broth medium with 1 mg/mL ampicillin was inoculated with overnight cultures of the individual transformants, and grown until O D 5 5 0 was 0.7. The cultures were then induced with 1 m M isopropylthiogalactoside (IPTG) for 3 more hours. Crude extracts were subsequently made by spinning down the cells, washing them with cold E. coli extraction buffer (20 m M 4-(hydroxyethyl)-l-piperazinethanesulfonic acid or Hepes at p H 7.5, 0.2 M NaCl , 10% glycerol, 0.1 m M dithiothreitol or D T T , 1 m M phenylmethylsulfonyl fluoride or P M S F ) , resuspending the cells in 2 m L of extraction buffer, sonicating for a total time of 3 minutes, centrifuging the sonicates for 20 minutes at 12100xg, aliquotting and flash freezing 19 the supernatants. The crude extracts were stored at - 7 0 ° C until time of use. Cloning for Expression In Vitro: Plasmids pG147+AR2 and p G I D l for expression in vitro of G A L 4 ' s D B D fused to A R 2 and D B D fused to ID1 fused to A R 2 were subcloned from the pGEM3(- ) derivative p G W T (contructed by I. Sadowski). pG147+AR2 was made by replacing the Xho I-Mlu I fragment of p G W T with that of pMA236 (constructed by J. Ma); the Xho I-Mlu I of pkwlO(Bam)A312:A421 (constructed by G . Stone) replaced that of p G W T to make p G I D l . P C R reactions were set up using 0.01 u\g of each Nde I linearized plasmid (pG147+AR2 and p G I D l ) to make D N A templates of defined sizes to be used for in vitro transcription reactions. Vent D N A Polymerase was used in standard 100 p L P C R reactions using 55 pmol of each primer CP2 (5' - C A C A G T T G A A G T G A A C T T G C - 3 ' ) and SP6P/P (5'-C A T A C G A T T T A G G T G A C A C T A T A G - 3 ' ) . The P C R products were purified using Promega's W i z a r d ^ M P C R * kit and resuspended in 100 p L DEPC-treated water. Plasmid pG241 (constructed by I. Sadowski), derived from pGEM3(-) and expressing G A L 4 ' s D B D , was also used as a template for in vitro transcription reactions by linearizing it with Ndel. In Vitro Transcription Reactions: The templates described above were used to synthesize G A L 4 derivative R N A in vitro. The 50 p L reactions were performed as described in Promega's "Transcription In Vitro" technical publication and contained 10 p L of the P C R products described above or 2 jxg of Nde I linearized pG241 as templates and SP6 R N A polymerase as enzyme. In Vitro Translation Reactions and Protein Normalization: The reactions were carried out in the wheat germ extract system sold by Promega by adding 5 p L of the in vitro translated products described above to a standard translation reaction with 6 p C i [35S]-methionine label as described in Promega's "Translation In Vitro" technical publication. For co-translation reactions, 5 u.L of each two chosen types of R N A were added instead of a single type in normal 20 translation reactions. To normalize the amount of protein of interest subsequently used in further experiments, 2 jxL samples of each completed reaction were run on a 12% S D S - P A G E gel. The gel was then Coomassie blue stained, fixed, enhanced with E N ^ H A N C E (Dupont), swollen with water and 10% glycerol, dried and exposed to X-ray film overnight. The bands of interest were cut out of the gel and counted by liquid scintillation. The number of counts obtained was then normalized to mmol of protein per methionine residue by using the specific activity of the [35s]-methionine label (1175 Ci/mmol), from which an equimolar ratio of protein was calculated, knowing that 1 C i = 2 . 2 x l 0 1 2 dpm. Electrophoretic Mobility Shift Assays: These assays were done using either crude E. coli extracts, purified E. coli produced proteins (described below) or wheat germ extracts, all expressing specific proteins. Each reaction contained equal amounts of specific protein as previously determined by Western blotting, S D S - P A G E Coomassie Blue staining or liquid scintillation counting. Binding reactions were performed in Ficoll buffer (20 m M Hepes p H 7.5, 5 m M M g C l 2 , 100 fig/mL bovine serum albumin or B S A , 2.5% ficoll, 10 u,M ZnSo4, 50 m M N a C l , 0.1 m M D T T , 0.1 m M ethylenediaminetetraacetate or E D T A ) for 30 minutes on ice with 4 pmol of [32p]_i abelled double stranded GS3/GS4 G A L 4 oligo ( 5 ' - T C G A C G G A G G A C A G T A C T C C G - 3 ' ) prior to analysis on 4.5% non-denaturing gels as described by Carey et al. (1989). G A L 4 Biotinylated Oligo Binding Assay: These assays were done using proteins expressed either in E. coli as crude extracts or in vitro in wheat germ extracts. For in vitro produced proteins, the experiment was carried out as follows: equimolar amounts of protein (as predetermined by liquid scintillation counting) were incubated with 25 u\L of pre-washed streptavidin agarose beads (Gibco-BRL) at 4 ° C on a rocker for 20 minutes as a preclearing step. The supernatants were then incubated with 100 pmol of G A L 4 biotinylated oligo (5' - b i o t i n - C G G A G T A C T G T C C T C C G - 3 ' ) , 150 p L of buffer 2xBIO (50 m M Hepes p H 7.5, 10 m M M g C l 2 , 300 m M KC1, 0.2 2 1 m M E D T A , 20% glycerol, 0.2 m M D T T , 2 m M P M S F , 20 p g / m L aprotinin, 2 p M leupeptin), 25 pg casein, and 20 pg poly-dldC in a total 300 p L reaction for 30 minutes on a rocker at room temperature. 100 p L of washed streptavidin agarose beads was then added, and the mixture further incubated for 30 minutes on a rocker at 4 ° C . The incubated beads were then washed four times with 1 m L of buffer l x B I O , transferring tubes on the last wash. The beads were then resuspended in 10 p L of 2xSDS sample buffer, boiled for 5 minutes and centrifuged for 5 minutes. The supernatant was then loaded on a 12% SDS-polyacrylamide gel which was then subjected to electrophoresis, Coomassie blue staining, fixing, enhancing with E N ^ H A N C E , and swelling with water and 10% glycerol, before being dried and exposed to X-ray film. For E. coli produced proteins, the assay was performed similarly except that equal amounts of protein (as predetermined by Western blotting) were added directly to the 300 p L reaction (without a preclearing step) which lacked poly-dldC and whose 2xBIO buffer contained 50 m M Hepes pH7.5, 10 m M M g C l 2 , 100 m M KC1, 0.2 m M E D T A , 20% glycerol, 0.2 m M D T T , 2 m M P M S F , 20 p g/mL aprotinin, and 2 p M leupeptin. The remaining incubations were as described above as were the four washes in l x B I O buffer for E. coli. There was no tube transfer at the end of wash four. The supernatant derived from the resuspension, boiling and centrifuging of the beads in 2xSDS sample buffer was also loaded on a 12% gel which was subjected to electrophoresis and then transferred onto nitrocellulose for Western blotting with m A b 5c8-12 to G A L 4 ' s D B D , and E C L detection as described by Ambion. 4. C h e m i c a l C r o s s l i n k i n g E x p e r i m e n t s The thiol cleavable, homobifunctional chemical crosslinker dithiobis(succinimidylpropionate) or D S P (Pierce chemicals) was used in all chemical crosslinking reactions. The stock reagant was prepared in dry dimethylsulfoxide (DMSO) by Todd Schindeler ( U . B . C . Chemistry Department) to a concentration of 25 m M and stored at 4 ° C . Equimolar amounts of protein of interest from either E. coli or wheat germ extracts were added to a 100 p i total volume reaction 22 consisting of D S P reaction buffer (20 m M sodium phosphate p H 7.5, 0.15 M NaCl) and, unless otherwise stated, 1 m M D S P crosslinker. The reaction was allowed to proceed at room temperature for 30 minutes and then quenched with 50 m M Tris, p H 8.0 for a further 30 minutes. For crosslinking with E. coli proteins, the reactions were then divided in half; 2xSDS sample buffer containing (3-mercaptoethanol was added to one half of the sample, and 2 X S D S sample buffer lacking P-mercaptoethanol was added to the other half. A l l reactions were then boiled and subjected to S D S - P A G E , transferred onto nitrocellulose and Western blotted with G A L 4 m A b 5c8-12 to G A L 4 ' s D N A binding region. For in vitro synthesized proteins, the crosslinked reactions were diluted to 500 J I L in l x R I P A buffer (100 m M Tris pH8.0, 100 m M NaCl , 1 m M E D T A , 1% Nonidet-P-40, 0.5% sodium deoxycholate, 0.1% SDS, I m M P M S F , 10 ug/ml aprotinin, 1 p M leupeptin) following crosslinking, and then immunoprecipitated. The immunoprecipitations were performed by first preclearing with 50 uX 10% formalin fixed Staphylococcus aureus (S.A.) for 30 minutes at 4 ° C , then incubating the diluted reactions with 20 p L mAb 5c8-12 for one hour on ice, adding 50 p L rabbit-anti-mouse ( R A M ) coated 10% S.A. and incubating for one hour on a rocker at 4 ° C . After centrifugation, the pellet was washed four times with 1 m L l x R I P A buffer, the pellet being split into two equal portions upon the last wash. One half of the pellet was then resuspended in 2xSDS sample buffer containing P-mercaptoethanol and the other half in 2xSDS sample buffer lacking P-mercaptoethanol. The suspensions were incubated 10 minutes at 3 7 ° C , centrifuged for 5 minutes and the supernatants subjected to S D S - P A G E . The gels were then Coomassie blue stained, fixed, enhanced ( E N ^ H A N C E ) and swollen with water and 10% glycerol before being dried and exposed to X-ray film. 5. Experiments with E. coli Histidine-tagged G A L 4 Derivatives Cloning and Expression: The histidine-tagged G A L 4 derivative expression plasmids were subcloned from the p B I D O M A I N plasmid (expressing G A L 4 ' s D B D fused to A R 2 and tagged with six histidine 23 residues) constructed by W . Crosby in p R S E T B (Invitrogen). The p B I D O M A I N plasmid was digested with EcoR I and Hind III, treated with Klenow and religated to give p H I S D B D , expressing a histidine-tagged version of G A L 4 ' s D B D . To obtain pHISIDl , a histidine-tagged version of G A L 4 ' s D B D fused to ID1 fused to AR2 , the p B I D O M A I N Xho I-Mlu I fragment was replaced with that of pkwlO(Bam)A321:A412 (constructed by G . Stone). The clones were transformed into NM522 E. coli cells ([proAB+lacIqZAM15] supE thi-1 A(lacproAB)' Ahsd ( r m j Ar d e o R + ) (donated by D . Kilburn) for expression. Expression of the histidine-tagged fusion proteins was achieved as follows: 250 m L of fresh L B medium with 1 mg/mL ampicillin was inoculated with a 1/100 dilution of an overnight culture. The cells were grown to OD600 of 0.3, and then I m M I P T G was added. At OD600=0.6, M 1 3 T 7 - R N A P o l phage expressing T7 R N A polymerase (obtained from W . Crosby) was added to a multiplicity of infection of 5 per cell. The cultures were further induced for 4 hours. The cells were harvested by centrifugation. The pellet was washed with X A 9 0 lysis buffer (10 m M Tris p H 8.0, o.5 M NaCl , 10% glycerol, 10 m M P-mercaptoethanol, and 0.1% Tween-20), and then resuspended in 4 m L of the same buffer before sonicating for 4 minutes. The supernatant was cleared by centrifugation (15 minutes, 27200xg), aliquoted, flash frozen and stored at - 7 0 ° C until time of use. Samples were routinely screened by Western blotting for protein expression levels. Batch Purification: Crude E. coli extracts expressing G A L 4 histidine-tagged derivatives were purified in a batch procedure as follows. The extracts were incubated with 500 p L of ProBond Ni^+-N T A agarose beads (Invitrogen) equilibrated in buffer A (20 m M Tris p H 8.0, 100 m M KC1, 20% glycerol, 30 m M imidazole) for 30 minutes on a rocker at 4 ° C . The beads were then washed 4 times with 5 m L cold buffer A . The proteins were then eluted with 400 p L of buffer B (20 m M Tris p H 8.0, 100 m M KC1, 20% glycerol, 250 m M imidazole). Samples of the purified proteins were then analyzed by S D S - P A G E and Western blotting. This procedure was adapted from Reece, Rickles and Ptashne (1992). 24 Column Purification: A 2 m L ProBond resin column equilibrated in buffer A (composition described above) was used to purify the histidine-tagged G A L 4 derivatives from crude E. coli extracts (68). A n Econo Sytem purification apparatus (Bio-Rad) was used at 4 ° C , with a flow rate of 0.5 ml/min. 1 m L fractions were collected while monitoring A280- The elution gradient ranged from buffer A described above at 30 m M imidazole to buffer B described above at 500 m M imidazole. The duration of the elution gradient ranged from 1 hour for p H I S D B D extracts, to 20 minutes for p B I D O M A I N extracts, to 40 minutes for pHISIDl extracts. Samples of the column fractions were analyzed by S D S - P A G E and Coomassie blue staining. 25 R E S U L T S 1. Glucose Inhibits G A L 4 T h r o u g h Its C e n t r a l Reg ion As described by Stone and Sadowski (1993), the central region of G A L 4 contains a glucose-responsive domain called G R D and three inhibitory domains (ID1, 2, 3). To further characterize the effect of glucose on the G A L 4 protein, I performed a ribonuclease protection assay. The analysis of m R N A production instead of protein activity in response to glucose was necessary since G . Stone showed with nuclear run-on transcription assays that glucose inhibition of G A L 4 through its central region has rapid kinetics, peaking within 20 minutes, the G A L 4 protein recovering with time (77). Yeast R N A was extracted from cultures grown in glycerol with or without the addition of 2% glucose for one hour prior to extraction. The yeast strain used was YT6::171 which contains U A S G and U R S G sites upstream of a GAL1 -lacZ reporter gene. The strain expressed either wild type G A L 4 protein (plasmid yCPG4trp), G A L 4 protein lacking the central region (CR, residues 320-767; plasmid yCPG4242) or a vector control (ycplac22) from GAL4\ wild type promoter. 1 pg of each R N A extracted, as measured by optical density, was hybridized to lacZ and actin probes. These probes were used to detect transcriptional activation of the reporter gene as well as production of an internal control to quantitate more accurately the amount of R N A in each sample. The results of the ribonuclease protection assay are depicted in Figure 5. As can be seen, both wild type G A L 4 and G A L 4 C R deletion could activate transcription in yeast growing in glycerol. But, when yeast were subjected to 2% glucose for one hour, wild type G A L 4 became completely inactive whereas G A L 4 C R deletion was not affected by the presence of glucose, and, actually, even showed about 10 fold induction. The vector control behaved as expected, showing no transcriptional activation of the lacZ reporter gene in either glycerol or glucose. 26 glycerol glucose lacZ m f f actin - m *» • 1 2 3 4 5 6 7 8 Figure 5. Glucose Represses GAL4 Through Its Central Region Ribonuclease protection assay performed with 1 pg total yeast RNA probed with lacZ and actin probes. The yeast RNA was extracted from strain YT6::171 expressing, from a wild type GAL4 promoter, either wild type GAL4 (lanes 3 and 6), GAL4 lacking its central region (lanes 4 and 7) or a vector control (lanes 5 and 8). The yeast cultures were grown in glycerol until OD600=l-5, then separated into two, one half receiving 2% glucose for one hour before extraction (lanes 6-8), the other half receiving no glucose (lanes 3-5). Lanes 1 and 2 were positive yeast RNA controls probed with actin or lacZ probes alone, respectively. This experiment, as well as all others, were repeated with reproducible results at least twice, except where indicated. 27 This ribonuclease protection assay result, then, reconfirmed that the G A L 4 protein is indeed rapidly affected by glucose through its central region as described by Stone and Sadowski (1993). Also, this result establishes glucose inhibition of the G A L 4 protein as an important and relevant glucose repression mechanism, unlike what Johnston et al. (1994) stated. This glucose repression mechanism remains to be further characterized by determining how the signal from glucose is relayed to the central region of G A L 4 and how the G R D controls the activity of the inhibitory domains as has been proposed (78). 2 . I D 1 Is a Strong Inhibitory Domain in Yeast The next step of my research involved reconfirming that G A L 4 ' s ID1 (residues 320-412) is a strong inhibitory domain in vivo in yeast cells, as observed by Stone and Sadowski (1993). Again, I used the yeast strain YT6::171 with a GALl-lacZ reporter gene to assay transcriptional activation by different G A L 4 derivatives in the form of p-galactosidase activity in dropout, selective medium. The G A L 4 derivatives, depicted in figure 6, consisted of wild type G A L 4 , G A L 4 D N A binding domain (DBD, residues 1-147), G A L 4 D B D fused to A R 2 (DBD-AR2) , G A L 4 D B D fused to ID1 fused to A R 2 (DBD-ID1-AR2), as well as a vector control (plasmids pMA210, pMA241 , pMA236 , p M A I D l , and pMA200 , respectively). A l l of these derivatives were overexpressed from an A D H 1 promoter contained in high copy 2 p plasmids. The results of the P-galactosidase assays are shown in table 1. As can be seen, the wild type G A L 4 and no G A L 4 controls showed the expected high activity and no activity, respectively. Also as expected, D B D was inactive as it can only bind D N A , lacking transcriptional activation activity which was seen in the D B D - A R 2 construct. On the other hand, transcriptional activation activity was lost with the D B D - I D 1 - A R 2 construct. 28 1 100 200 300 400 500 600 700 800 900 wild type GAL4 ID1 ID2ID3 GRD m m rfY/y/yy/y/A DBD DBD-AR2 DBD -M DBD-ID1-AR2 ID1 Figure 6. G A L 4 Derivatives The makeup of three G A L 4 derivatives called D B D ( D N A binding domain), D B D - A R 2 ( D N A binding domain fused to activating region 2) and D B D - I D 1 - A R 2 ( D N A binding domain fused to inhibitory domain 1 fused to activating region 2) is depicted in a linear form with respect to wild type G A L 4 . The scale at the top numbers the amino acid residues in the proteins. The dashed line (—) joining the different parts of the G A L 4 derivatives represents a direct fusion between these parts. These derivatives were expressed in yeast, E. coli and wheat germ extracts. 29 G A L 4 Derivative p-Galactosidase Activity (Miller units ± SD) Wild Type GAL4 1505 ±168 No GAL4 <1 DBD <1 DBD-AR2 1373 ± 4 3 DBD-ID1-AR2 <1 Table 1. p-Galactosidase Assay of G A L 4 Derivatives The (3-galactosidase activities were assayed in the yeast strain YT6::171 transformed with plasmids over-expressing the G A L 4 derivatives listed above. The results are expressed as Mil ler units plus or minus the standard deviation (SD) of the duplicate samples. 30 The (3-galactosidase assays in table 1 show that ID1 is responsible for strong in vivo inhibition of G A L 4 in yeast, as was observed by Stone and Sadowski (1993). Stone and Sadowski suggested that G A L 4 ' s D N A binding ability was affected by the inhibitory action of the C R but had not elucidated a clear mechanism of action of the inhibitory domains. They also noted that ID1 was the strongest inhibitory domain, and that it contained homology to other fungal transcription factors. For these reasons, I set out to characterize the mechanism of action of ID1 in my research plans. 3. I D 1 Impairs D N A B i n d i n g The first question to be answered was whether ID1 affects G A L 4 D N A binding as was suggested (78), or whether it impairs G A L 4 functioning while G A L 4 is bound to D N A . D N A binding of a transcriptional activator is a necessary but insufficient step to achieve transcriptional activation. I used two approaches to resolve this D N A binding question: electrophoretic mobility shift assays ( E M S A ) and G A L 4 biotinylated oligo assays. I performed E M S A experiments with G A L 4 derivatives expressed in E. coli as crude extracts, seen in figure 7, and with G A L 4 derivatives expressed in vitro in wheat germ extracts, seen in figure 8. The amounts of each protein used were standardized to comparable amounts by Western blotting for E. coli expression, and to exact amounts by liquid scintillation counting for in vitro expression. Even though Western blotting was performed with a G A L 4 monoclonal antibody, some background bands were consistently observed, even with the no G A L 4 control extract (figure 7, lane 4), suggesting that these bands result from some cross-reactivity of the antibody with components of the E. coli crude extract. These background bands were observed particularly when the amount of the crude extract tested was increased to normalize the protein of interest, such as lanes 3, 4, and 5 in figure 7, as opposed to lanes 1 and 2, which contained less crude extract. 3 1 Figure 7. Expression of GAL4 Derivatives in E. coli Western blot of crude E. coli extracts separated on a 12% SDS-PAGE gel, probed with mAb 5C8-12 to the GAL4 D N A binding domain, and detected by E C L . The GAL4 derivatives expressed in the extracts are DBD (lane 1, position a), DBD-AR2 (lane 2, position b), DBD-ID1-AR2 (lane 3, position c), no GAL4 vector control (lane 4), and DBD-GRD-AR2 (lane 5, position c). 3 2 Figure 8. Expression of GAL4 Derivatives in Wheat Germ Extracts Autoradiogram of a 12% SDS-PAGE gel showing [35s]-methionine labelled G A L 4 derivatives translated in vitro in wheat germ extracts. The GAL4 derivatives are: DBD (lane 1, position a), DBD-AR2 (lane 2, position b), and DBD-ID1-AR2 (lane 3, position c). Luciferase protein (lane 4, position d) was also expressed in wheat germ extracts as a control. 3 3 As can be seen in figure 9, the E. coli G A L 4 derivative consisting of D B D showed the expected strong binding ability to its G A L 4 labelled binding site. Similarly, D B D - A R 2 , a transcriptionally active G A L 4 derivative as shown above by (3-galactosidase assay, also bound D N A . But, with the D B D - I D 1 - A R 2 fusion protein, D N A binding ability was severely impaired when compared to all other derivatives. In this case, no D N A binding was observed even with the use of three times the amount of protein compared to the other derivatives, as can be seen by the increasing breakdown product, also present with the other G A L 4 derivatives (Figure 9, position a). I speculate that this breakdown product, seen consistently in all E M S A experiments done with E. coli extracts, consists of a minimal D N A binding G A L 4 derivative, thus the band shift. A transcriptionally active control G A L 4 derivative with a similar molecular weight to DBD-ID1-AR2 (figure 7), consisting of G A L 4 ' s D B D fused to G R D , instead of ID1, fused to A R 2 ( D B D - G R D - A R 2 ) (77), did not show altered D N A binding. This result suggests that it was not the addition of any G A L 4 sequence placed between D B D and A R 2 that caused the D N A binding impairement, but a direct result of the presence of ID1. Other controls were also tested. A baculovirus produced wild type G A L 4 extract (made by T. Kang) bound D N A as expected, whereas an E. coli extract expressing no G A L 4 did not bind to the oligo, showing specificity of the observed D N A binding. E M S A results with G A L 4 derivatives produced in vitro in wheat germ extracts shown in figure 10 support the E M S A results obtained with E. coli produced G A L 4 derivatives shown in figure 9. Equimolar amounts of the different derivatives were used in the binding reactions. As seen with E. coli G A L 4 fusions, in vitro synthesized D B D and D B D - A R 2 were able to bind to the G A L 4 specific oligo, whereas D B D - I D 1 - A R 2 was not. D N A binding of the ID1 containing derivative was impaired at equimolar amounts of protein compared to the other two derivatives as well as with twice the amount of D B D - I D 1 - A R 2 protein. A wheat germ extract containing no R N A upon translation as well as one containing Brome Mosaic Virus ( B M V ) R N A (provided by Promega), which translates into five proteins, showed no D N A 34 Figure 9. ID1 Impairs D N A Binding in E. coli in E M S A E M S A reactions were performed with equal amounts of crude E. coli extracts containing G A L 4 derivatives and a GAL4-specif ic oligo. The G A L 4 oligo alone is shown in lane 1. The G A L 4 derivatives added to the binding reactions are: D B D (lane 2, position b), D B D - A R 2 (lane 3, position c), D B D - I D 1 - A R 2 at IX, 2X and 3X concentrations (lanes 4, 5, 6, respectively), no G A L 4 vector control (lane 7), and D B D - G R D - A R 2 (lane 9, position d). Wi ld type G A L 4 expressed from baculovirus was also used as a positive control (lane 8, position e). Position a points out the protein breakdown product forming a G A L 4 minimal D N A binding domain seen in all reactions containing G A L 4 derivatives. 3 5 36 Figure 10. ID1 Impairs D N A Binding of In Vitro Translated G A L 4 in E M S A E M S A reactions were performed with equimolar amounts of [ ^ S ] -methionine labelled G A L 4 derivatives expressed in wheat germ extracts to a G A L 4 specific oligo. The G A L 4 oligo alone is shown in lane 1. The G A L 4 derivatives added to the binding reactions were: D B D (lane 2, position a), D B D - A R 2 (lane 3, position b), and D B D - I D 1 -A R 2 at I X and 2X concentrations (lanes 4 and 5). Lanes 6 and 7 were negative controls of wheat germ extract expressing no protein or B M V proteins, respectively. 3 7 1 2 3456 7 38 binding, suggesting that the D N A binding seen with D B D and D B D -A R 2 derivatives was specific. To further confirm the D N A binding results suggested by the E M S A data, I performed experiments with a biotinylated G A L 4 U A S G oligo. Any G A L 4 derivative bound to this oligo was isolated with the use of streptavidin agarose beads, whereas all unbound material was washed away. Once again, I tested both E. coli and in vitro expressed G A L 4 fusion proteins with this D N A binding assay. As shown in figure 11, the E. coli made D B D , D B D - A R 2 and D B D -G R D - A R 2 G A L 4 derivatives (expression shown in figure 7) bound to the G A L 4 biotinylated oligo. On the other hand, the D B D - I D 1 - A R 2 derivative, whose molecular weight is very similar to the D B D - G R D -A R 2 derivative, was not retained by the beads, implying that this G A L 4 fusion had not bound to the G A L 4 oligo. The D N A binding that was observed, for example with D B D , was specific to the G A L 4 oligo, in that if the G A L 4 biotinylated oligo was omitted from the reaction, no D B D protein was retained. Once again, the presence of ID1 appeared to impair the protein's ability to bind D N A . Similarly, G A L 4 biotinylated oligo assay performed with G A L 4 fusions translated in vitro in wheat germ extracts, displayed in figure 12, also confirmed the results obtained with E. coli expressed proteins (figure 11). Both the D B D and the D B D - A R 2 derivatives were specifically retained on the oligo bound to the beads, but the D B D - I D 1 - A R 2 derivative was not. If the oligo was omitted from the reactions, no derivative bound, as expected. As a control for specificity, luciferase ( R N A supplied by Promega) in vitro translated protein did not bind to the G A L 4 oligo, also as expected. The results obtained with E M S A and G A L 4 biotinylated oligo assays with both E. coli and wheat germ in vitro expressed proteins, then, suggest that the presence of ID1 severely impairs D N A binding of G A L 4 derivatives, probably to the order of 100-fold. This result suggests that ID1 inhibits G A L 4 from binding to D N A instead of 3 9 oligo 55^™ 36— 26— + + • -r - + + + C b 1 2 3 4 5 6 7 Figure 11. ID1 Prevents D N A Binding in E. coli to a G A L 4 Biotinylated Ol igo Western blot of a D N A binding assay performed with G A L 4 derivatives expressed in E. coli to a G A L 4 specific biotinylated oligo isolated with streptavidin agarose beads. A l l unbound material was washed away with four washes of 1 X B I O buffer. The G A L 4 derivatives used in this D N A binding assay are: D B D (lanes 1 and 2, position a), D B D - A R 2 (lane 3, position b), D B D - I D 1 - A R 2 (lanes 4 and 5), no G A L 4 vector control (lane 6) and D B D - G R D - A R 2 (lane 7, position c). The biotinylated oligo was omitted from the reactions i n lanes 1 and 4 as a control for binding specificity. 4 0 oligo + 63 _ 53 — 36 — 32 — - + - + - + -1 2 3 4 5 6 7 8 Figure 1 2 . ID1 Prevents DNA Binding in Wheat Germ Extracts to a GAL4 Biotinylated Oligo Autoradiogram of a DNA binding assay performed with [^ ^S]-methionine labelled, in vitro translated GAL4 derivatives to a GAL4 specific biotinylated oligo isolated with streptavidin agarose beads. All unbound material was washed away with four washes of 1XBIO buffer. The GAL4 derivatives used in this DNA binding assay are: DBD (lanes 1 and 2, position a), DBD-AR2 (lanes 3 and 4, position b) and DBD-ID1-AR2 (lanes 5 and 6). Luciferase protein translated in vitro in wheat germ extract was also used as a negative control (lanes 7 and 8). The GAL4 biotinylated oligo was either added to or omitted from the binding reactions as indicated as a control for binding specificity. 4 1 inhibiting G A L 4 while bound to D N A . The next step to be resolved was how ID1 was affecting D N A binding. 4. ID1 Prevents D i m e r i z a t i o n A possibility of how ID1 may impair D N A binding is by preventing dimerization of G A L 4 molecules. I set out to examine this possibility by chemical crosslinking experiments using in vitro translated G A L 4 derivatives D B D , D B D - A R 2 and D B D - I D 1 - A R 2 from wheat germ extracts. The crosslinker used for these experiments was the thiol cleavable, homobifunctional D S P reagent. With this reagent, I examined both homodimer and heterodimer formation of G A L 4 derivatives. For homodimer analysis, equimolar amounts of each in vitro synthesized G A L 4 fusion (a typical in vitro synthesis is shown in figure 8) were added to the crosslinking reaction. The G A L 4 species were then isolated by immunoprecipitation and analyzed by S D S -P A G E , either in a crosslinked state using a sample loading buffer lacking P-mercaptoethanol or in an uncrosslinked state using a sample loading buffer containing P-mercaptoethanol. The results of this experiment are depicted in figure 13. Figure 13 shows that both D B D and D B D - A R 2 formed homodimers when crosslinked, whereas D B D - I D 1 - A R 2 was unable to form homodimers. Instead, D B D - I D 1 -A R 2 appeared to form large protein complexes that were retained at the top of the gel. Since G A L 4 binds as a dimer to its 17 mer dyad symmetrical site (12), lack of homodimer formation caused by the presence of ID1 would necessarily impair D N A binding, a result concurring with my previous observation about the mechanism of action of ID1. I also tested whether ID1 would prevent the formation of heterodimers. First, all paired combinations of the three G A L 4 derivatives D B D , D B D - A R 2 , D B D - I D 1 - A R 2 were co-translated in vitro in wheat germ extracts, as shown in figure 14. Subsequently, these co-translated proteins were subjected to D S P crosslinking, 42 Figure 13. ID1 Prevents the Formation of Homodimers Autoradiogram of a crosslinking experiment examining the formation of homodimers of [35s]-methionine labelled G A L 4 derivatives expressed in wheat germ extracts. Equimolar amounts of G A L 4 derivatives were crosslinked with D S P , immunoprecipitated with m A b 5C8-12 to the D N A binding domain of G A L 4 and separated by 10% S D S - P A G E either in the crosslinked state without the addition of P-mercaptoethanol (P-Me) to the sample loading buffer (lanes 1-3), or in the uncrosslinked state with the addition of P-mercaptoethanol to the sample loading buffer (lanes 4-6). The G A L 4 derivatives used were: D B D (lanes 1 and 4), D B D - A R 2 (lanes 2 and 5), and DBD-ID1-A R 2 (lanes 3 and 6). The species indicated by the arrows represent: D B D monomers and dimers (positions a and b, respectively), D B D - A R 2 monomers and dimers (positions c and e, respectively), and D B D - I D 1 -A R 2 momoners (position d). 4 3 p-Me 200—1 116—' 97— 66— 45 + + + 29 1 2 3 4 5 6 44 53 36; 32 c b 1 2 3 Figure 14. In Vitro Co-translation of GAL4 Derivatives in Wheat Germ Extracts Autoradiogram of a 12% SDS-PAGE gel showing the [35s]-methionine labelled, co-translated GAL4 derivatives DBD (position a), DBD-AR2 (position b) and DBD-ID1-AR2 (position c). The GAL4 derivatives were doubly translated in wheat germ extracts as: DBD plus DBD-AR2 (lane 1), DBD plus DBD-ID1-AR2 (lane 2), and DBD-AR2 plus DBD-ID1-AR2 (lane 3). 4 5 immunoprecipitation and S D S - P A G E as for singly translated species, the results shown in figure 15. Again the crosslinker was either released or retained by the presence or absence of (3-mercaptoethanol in the sample loading buffer, respectively. A heterodimer form was observed amongst D B D and D B D - A R 2 derivatives, but no heterodimers were observed when one of the two co-translated G A L 4 derivatives contained ID1. Even when co-translated, D B D and D B D - A R 2 still formed homodimers, as expected, yet D B D - I D 1 - A R 2 formed neither homodimers or heterodimers. ID1, then, appears to prevent the dimerization function of G A L 4 , impairing its D N A binding ability. 5. ID1 Promotes Multimerization Since G A L 4 derivatives containing ID1 do not appear to dimerize but instead tend to aggregate as large complexes, I set out to determine the makeup of these complexes. I employed a strategy that would potentially form a ladder of partially formed protein complexes. To accomplish this, I added an approximate 10 fold higher concentration of in vitro translated D B D - I D 1 - A R 2 protein than usually used in crosslinking reactions, while gradually decreasing the concentration of D S P crosslinker used. These reactions were then immunoprecipitated and analyzed by S D S - P A G E , again either including or excluding p-mercaptoethanol from the sample loading buffer. Figure 16 demonstrates the results of the limited D S P crosslinking experiment with wheat germ extracts. The uncrosslinked lanes displayed the D B D - I D 1 - A R 2 monomer of apparent molecular weight of approximately 48 K D a , as observed in figures 13 and 15 as well. On the other hand, as the concentration of crosslinker was increased in the crosslinked lanes, a gradual disappearance of the excess 48 K D a monomer was observed along with the gradual increase of fully formed large protein complexes at the top of the gel. But, the important observation that could be seen in the crosslinked lanes of figure 16, was the appearance of three additional bands that 46 Figure 15. ID1 Prevents the Formation of Heterodimers Autoradiogram of a crosslinking experiment examining the formation of heterodimers of [35s]-methionine labelled G A L 4 derivatives co-translated in vitro in wheat germ extracts. The co-translated G A L 4 derivatives were crosslinked with DSP, immunoprecipitated with mAb 5C8-12 to the D N A binding domain of G A L 4 and separated by 10% S D S - P A G E either in the crosslinked state without the addition of P-mercaptoethanol (P-Me) to the sample loading buffer (lanes 1-3), or in the uncrosslinked state with the addition of P-mercaptoethanol to the sample loading buffer (lanes 4-6). The co-translated G A L 4 derivatives used were: D B D - A R 2 plus D B D - I D 1 - A R 2 (lanes 1 and 4), D B D plus DBD-ID1-AR2 (lanes 2 and 5), and D B D plus D B D - A R 2 (lanes 3 and 6). The species indicated by the arrows represent: D B D monomers and dimers (positions a and b, respectively), D B D - A R 2 monomers and dimers (positions c and f, respectively), D B D - I D 1 - A R 2 momoners (position d), and D B D plus D B D - A R 2 heterodimers (lane 3, position e). 4 7 p-Me . . . + + + 1 2 3 4 5 6 48 Figure 16. Limited Crosslinking Reveals that ID1 Causes the Formation of Multimers in Wheat Germ Extracts Autoradiogram of a crosslinking experiment with limited D S P crosslinker examining the formation of protein multimers of the [35s]-methionine labelled, wheat germ expressed G A L 4 derivative D B D - I D 1 - A R 2 . D B D - I D 1 - A R 2 was crosslinked with DSP, immunoprecipitated with mAb 5C8-12 to the D N A binding domain of G A L 4 and separated by 7.5% S D S - P A G E either in the crosslinked state without the addition of p-mercaptoethanol (P-Me) to the sample loading buffer (lanes 1-6), or in the uncrosslinked state with the addition of p-mercaptoethanol to the sample loading buffer (lanes 6-12). A gradient of DSP crosslinker concentration was used in the reactions: 0.06 m M , 0.125 m M , 0.25 m M , 0.5 m M , 0.75 m M and 1 m M (lanes 1 to 6 and lanes 7 to 12). The species indicated by the arrows represent: D B D - I D 1 - A R 2 monomers (position a), dimers (position b), trimers (position c), and tetramers (position d). This experiment was only performed once. 4 9 1 2 3 4 5 6 7 8 9 10 11 12 50 were not detected in the uncrosslinked lanes. I suggest that these three bands of apparent increasing molecular weights of approximately 100 K D a , 141 K D a , and 186 K D a represent the dimer, trimer, and tetramer forms, respectively, of the D B D - I D 1 - A R 2 protein. The calculated molecular weights of the dimer, . trimer and tetramer forms of the 48 K D a ID1 containing G A L 4 derivative are 96 K D a , 144 K D a and 192 K D a , very similar to the molecular weights of the three protein bands immunoprecipitated with G A L 4 m A b 5C8-12 observed in figure 16. The intensity of these three protein bands is rather faint compared to that of the intensity of the monomer bands and to that seen for the bands at the top of the gel. This faintness implies that the higher order multimer forms caused by ID1 are favoured, preferentially forming a large multimer complex. The formation of protein multimers by the D B D - I D 1 - A R 2 derivative was also observed with E. coli expressed G A L 4 derivatives in a limited D S P crosslinking experiment, similar to that for in vitro translated proteins. The results of the limited crosslinking experiment performed with E. coli expressed G A L 4 constructs are shown in figure 17. The dimer, trimer, tetramer and higher multimer forms of D B D -ID1-AR2 appear to form naturally in solution as they were also observed when in vitro expressed D B D - I D 1 - A R 2 was immunoprecipitated with G A L 4 antibody and separated on a gel without p-mercaptoethanol in the sample buffer and in the absence of any D S P crosslinker (data not shown). The same result was also observed with E. co//-expressed D B D - I D 1 - A R 2 (data not shown). These protein multimers probably do not exist normally in solution with wild type G A L 4 in yeast as ID1 is kept inactive by the G R D in the absence of glucose, but may form in the presence of glucose when ID1 is active. The formation of only a small amount of ID1 containing dimers would impair the ability of this protein to bind D N A . This latter observation agrees with the results of the experiments detailed above. The mechanism of action of ID1 of G A L 4 , then, appears to be 5 1 Figure 17. Limited Crosslinking Reveals that ID1 Causes the Formation of Multimers in E. coli Autoradiograms of a crosslinking experiment with limited D S P crosslinker examining the formation of protein multimers of the G A L 4 derivative D B D - I D 1 - A R 2 expressed in E. coli. D B D - I D 1 - A R 2 was crosslinked with DSP and separated by 7.5% S D S - P A G E either in the crosslinked state without the addition of P-mercaptoethanol (p-Me) to the sample loading buffer (panel A) , or in the uncrosslinked state with the addition of P-mercaptoethanol to the sample loading buffer (panel B). A gradient of D S P crosslinker concentration was used in the reactions: 0.06 m M , 0.125 m M , 0.25 m M , 0.5 m M , 0.75 m M and 1 m M (lanes 1 to 6 in each panel). The species indicated by the arrows represent: D B D - I D 1 - A R 2 monomers (position a), dimers (position b), trimers (position c), and tetramers (position d). 5 2 A -P-Me b + [DSP] f ^ 3 1 2^3 A 5 6 B [DSP] + (3-Me 1 2 3 4 5 6 the formation of protein multimers, preventing dimer formation and impairing D N A binding. 6. ID1 A p p e a r s to A c t by Itself M y results suggest that ID1 causes protein multimerization to prevent D N A binding of G A L 4 , inhibiting G A L 4 function. The limited D S P crosslinking experiment that elucidated the multimerization mechanism suggested that the complexes are homogenous judged from the apparent molecular weights of the multimer bands. Also, G . Stone (1992) had attempted to titrate out a cellular factor or factors that possibly bound to ID1 to inhibit G A L 4 activity by overexpressing an ID1 protein fragment in yeast. This titration attempt proved unsuccessful, further suggesting that ID1 imposes inhibition without the aid of other factors. I set out to resolve this aspect of ID1 functioning. I expressed histidine-tagged versions of G A L 4 derivatives in E. coli. These derivatives consisted of his-DBD, h i s - D B D - A R 2 and his-D B D - I D 1 - A R 2 . The six histidine residues tag expressed N-terminal to the G A L 4 fusions allowed these proteins to be purified with the use of N i 2 + - N T A -agarose beads. I purified each G A L 4 fusion protein using either a batch or a column purification procedure, looking for a protein or proteins that would be purified along with h i s - D B D - I D l -A R 2 but not with the other G A L 4 fusions. If such proteins existed, it would suggest that ID1 requires these conserved cellular proteins that bind to ID1 and contribute to inhibition. The batch purifications are displayed in figure 18, whereas the purified column fractions can be found in figures 19, 20, and 21. Both the batch purification attempt and the column purification procedure, which yielded purer protein, failed to produce one or more proteins that copurified solely with h i s - B D B - I D l - A R 2 . Furthermore, I performed E M S A experiments with comparable amounts of either batch purified (figure 22) or column purified (figure 23) histidine-tagged G A L 4 derivatives to determine whether 54 Figure 18. Batch Purification of Histidine-tagged G A L 4 Derivatives Coomassie blue staining of S D S - P A G E gels, 15% (panel A) or 7.5% (panel B), of histidine-tagged G A L 4 derivatives expressed in E. coli and purified in a batch procedure. E. coli extracts expressing different G A L 4 derivatives were incubated with N i 2 + - N T A - a g a r o s e beads, washed four times with buffer A , at 30 m M imidazole, then eluted with buffer B , at 250 m M imidazole. Lane 1 in both panels represent the batch purification of the vector control p R S E T B , lane 2 in both panels that of G A L 4 his -DBD (position a), lane 3 in both panels that of G A L 4 h i s -DBD-AR2 (position b), and lane 4 in both panels that of G A L 4 h i s - D B D - I D l - A R 2 (position c). 5 5 56 Figure 19. Column Purification of the G A L 4 Derivative h i s -DBD Coomassie blue staining of the column fractions 1-24 (panels A , B , and C) containing the purified G A L 4 derivative h i s -DBD separated on a 10% S D S - P A G E gel. The his-DBD protein was expressed in E. coli and loaded on a 2 m L column of Ni2+-NTA-agarose beads. The column was run on an Econo System apparatus at 4 ° C with a flow rate of 0.5 mL/min with an elution gradient of buffer A at 30 m M imidazole to buffer B at 500 m M imidazole over 1 hour. Samples of the 1 m L column fractions that were collected are shown above. The h i s -DBD G A L 4 derivative is indicated by the arrow a. 57 1 2 3 4 5 6 7 8 910111213141516 1718192021222324 58 1 2 3 4 5 6 7 8 Figure 20. Column Purification of the GAL4 Derivative his-DBD-AR2 Coomassie blue staining of the column fractions 1-8 containing the purified GAL4 derivative his-DBD-AR2 separated on a 10% SDS-PAGE gel. The his-DBD-AR2 protein was expressed in E. coli and loaded on a 2 mL column of Ni2+-NTA-agarose beads. The column was run on an Econo System apparatus at 4°C with a flow rate of 0.5 mL/min with an elution gradient of buffer A at 30 mM imidazole to buffer B at 500 mM imidazole over 20 minutes. Samples of the 1 mL column fractions that were collected are shown above. The his-DBD-AR2 GAL4 derivative is indicated by the arrow a. 59 Figure 2 1 . Column Purification of the G A L 4 Derivative h i s - D B D - I D l - A R 2 Coomassie blue staining of the column fractions 1-16 (panels A and B) containing the purified G A L 4 derivative h i s - D B D - I D l - A R 2 separated on a 10% S D S - P A G E gel. The his-DBD-ID 1 - A R 2 protein was expressed in E. coli and loaded on a 2 m L column of N i ^ + - N T A -agarose beads. The column was run on an Econo System apparatus at 4 ° C with a flow rate of 0.5 mL/min with an elution gradient of buffer A at 30 m M imidazole to buffer B at 500 m M imidazole over 40 minutes. Samples of the 1 m L column fractions that were collected are shown above. The his -DBD-ID 1-AR2 G A L 4 derivative is indicated by the arrow a. 60 97 66 45 29 1 2 3 4 5 6 7 8 97 66 45 29 910111213141516 6 1 Figure 22. E M S A of Batch Purified Histidine-tagged G A L 4 Derivatives Equal amounts of the batch purified E. coli derivatives h i s -DBD (lane 2, position b), h i s -DBD-AR2 (lane 3, position c) and h i s - D B D - I D l - A R 2 (lanes 4-6, position d) were added to E M S A D N A binding reactions with a G A L 4 specific oligo. Arrow a displays the protein breakdown product forming a G A L 4 minimal D N A binding domain. The G A L 4 oligo alone is shown in lane 1. A 100X excess of unlabelled, specific competitor (sp. competitor) G A L 4 oligo was added in the reaction in lane 5, whereas the reaction in lane 6 contained a 100X excess of an unlabelled, non-specific competitor (n. sp. competitor) human R B F protein oligo. 62 sp. competitor n. sp. competitor Figure 23. E M S A of Column Purified Histidine-tagged G A L 4 Derivatives Equal amounts of the column purified E. coli derivatives h i s - D B D (lane 2, position b), h i s -DBD-AR2 (lane 3, position c) and h i s -DBD-ID 1-AR2 (lanes 4, position d) were added to E M S A D N A binding reactions with a G A L 4 specific oligo. Arrow a displays the protein breakdown product forming a G A L 4 minimal D N A binding domain. The G A L 4 oligo alone is shown in lane 1. The column fractions used were fraction 11 for his -DBD, fraction 8 for h i s - D B D - A R 2 , and fraction 11 for h i s - D B D - I D l - A R 2 , samples of which can be seen in figures 19, 20 and 21, respectively. 64 1 2 3 4 6 5 the purer ID1 containing protein would behave differently than a more impure form. A difference in DNA binding behaviour would suggest the loss of a necessary factor that perhaps could not be detected by the protein purification attempts. As had been previously observed in other EMSA experiments, such as figure 9, the ID1 containing derivative was severely impaired, to the order of about 100-fold, in binding ability compared to the other GAL4 fusions; however, no difference was found in his-DBD-ID 1-AR2 DNA binding behaviour in relation to purity. This lack of difference again implies that ID1 impairs DNA binding ability without the help of cellular proteins. The small amount of DNA binding observed in figures 22 and 23 with his-DBD-ID 1-AR2 appears to be specific by the competition reactions performed in figure 22. The amount of his-DBD-ID 1-AR2 E. coli derivatives added to the EMSA reactions of figures 22 and 23 was much higher than that added to similar EMSA reactions with non-histidine tagged E. coli versions of DBD-ID1-AR2 (figure 9) which showed no binding ability at all; I believe that the higher the GAL4 fusion protein concentration, the higher the amount of ID1 containing dimers versus multimers that can be formed, thus the minimal DNA binding observed. The purification results and the EMSA results with the purified GAL4 fusions agree with the limited DSP multimerization data that ID1 appears to be able to inhibit GAL4 derivatives without the aid of any additional cellular factors. However, how the multimerization mechanism occurs in terms of the exact location of the multimerization contacts and whether the multimers are aberrant are still unresolved questions. Nevertheless, it appears that ID1 inhibits GAL4 activity by forming homogenous protein multimers, preventing dimerization and thus impairing DNA binding, a necessary step to achieve transcriptional activation. 7. ID1 is Homologous to Other Proteins As Stone and Sadowski (1993) originally noted, the short 320-412 amino acid stretch found in GAL4 known as ID1 is homologous to 66 other proteins, mainly fungal transcriptional activators, but also to other proteins such as C D C - 6 and PDR3. Table 2 summarizes some of the proteins homologous to ID1 as well as their respective function and species of origin. The list of proteins having homology to ID1 is constantly growing as the genome sequencing projects of many organisms are progressing, and new genes are discovered. A computer aided sequence alignment ( B L I T Z search) between ID1 of G A L 4 and some of its homologous proteins is displayed in figure 24. The majority of the conserved residues seen in figure 24 are hydrophobic or polar. On the other hand, there are also several conserved charged residues as well as a few particular amino acids that are conserved in all proteins. Three residues in particular may be especially important for the functioning of ID1 or its regulation. These three residues, marked with an asterix in figure 24, were found to be point mutations (Ser 322 to Phe, Leu 331 to Pro, and Ser 352 to Phe) that inactivate G A L 4 , by Johnston and Dover (1988). These particular residues, especially Leu 331 and Ser 352 which show good homology to the other proteins and may thus serve conserved roles, could be responsible for protein-protein interactions that either govern the regulation of ID1 through contact with the G R D , or mediate protein multimerization promoted by ID1. Either way, mutations at these residues cause unusual amino acid function that render G A L 4 inactive. Since ID1 has been shown to be modular and to function in different fusion protein locations, it is possible that ID1 forms a globular domain, perhaps consisting of the conserved hydrophobic residues interacting with each other and exposing the conserved charged or polar residues. These conserved charged or polar residues may be involved in protein-protein interactions, forming the protein multimers observed with ID1. Conversely, the conserved hydrophobic residues may be important for forming protein-protein interactions. On the other hand, ID1 may not be globular but unstructured while still able to form protein multimers. Nevertheless, there is a strong possibility requiring further investigation that the sequence similarity between ID1 and other proteins may underlie structural and functional 67 P R O T E I N P R O T E I N F U N C T I O N S P E C I E S L A C 9 Lactose/galactose transcription factor K. lactis P U T 3 Proline utilization transcription factor S. cerevisiae L E U 3 Leucine anabolism transcription factor S. cerevisiae PPR1 Pyrimidine pathway transcription factor S. cerevisiae T H I 1 Thiamine repressible genes regulator S. pombe N I T 4 Nitrogen assimilation transcription factor N. crassa N I R A Nitrogen assimilation transcription factor A. nidulans PDR3 Pleiotropic drug resistance regulator S. cerevisiae CDC6 Initiation of D N A replication S. cerevisiae Y H X 8 Putative transcription factor S. cerevisiae YINO Putative transcription factor S. cerevisiae YJ16 Putative transcription factor S. cerevisiae YCZ6 Putative transcrition factor s. cerevisiae YBOO Putative transcription factor s. cerevisiae YE14 Putative transcripiton factor s. cerevisiae Table 2. Proteins Homologous to ID1 of G A L 4 The proteins listed above with their respective protein function and species of origin contain a high degree of sequence homology to ID1 of G A L 4 according to a computer homology search (BLITZ) . 68 GAL 4 PDR3 PPRl LEU3 NIRA NIT4 LAC 9 PUT 3 THI1 CDC 6 SGSIILV TALHLLSRYTQWRQ KTNTSYNFHSFSIRMAISLGLNRDLPSS FSD KETIYLILR LFDLC YEHLIQ GCISISNPLENYLQKIK QTPTTTASASL PTSPA NSQLPLLHRELFLKK YFEPIY GPWNPNIALASDQTGINSAFEIPIT SAFSA HTEPK ASVYSV QAFLLYT FWPPLT SSLSADTSWNTIGTAMFQALRV GLNCAGFSKEYASAN NSKLCTV QALALMS VRE A GCGREGKGWVYSGMSFRMAFDLGLN LESSSLRDLSE MS VRE A GCGREAKGWVYSGMSFRMAQDIGLN LDIGSL DE GSTDLTI ALILLTHYVQKMH I EVLLLYAFFLQ VA DY LI GLYLQSTIYE IYSI QAIFMMTIFLQCSA ! !## ## #####! !## # KPNTAWSLIGLCSHMATSL GLHRDL PN TLASYFYFGQALRTCLIL GLHVDSQS DTLSR KSSFAYFGLAIKFAVAL GLHKNSDDPSL TQN NLKACYSFIGIALRAALKE GLHRRSSIVGPTPIQ !!#!# !! #!#!!##+ ## #GL!+ !#!# !## 6AL4 PDR3 PPRl LEU3 NIRA NXT4 LAC 9 PUT3 THI1 CDC 6 SSILEQRRRIW W SVYSWEIQLSLLY G RSI QL SQNTISFPSS PLSNDLVIS VIHQLPQPFIQSITGFTTTQLIENLHDSFSMFRIVTQMYAQHRKRFAEF RENVTEKID VCSSVDVPWYDT WETSQKVN SELVNEQIRTWICCNWSQTVASSFGFPAYVSFDYLV EEIDARRITFWGCFLFDKCW KEVDARRITFWGCFVFDKCW S RRV LW YEIEHHR RLW SKELRNRL LW DET KKRL FW !-## ++# #W#! W W ! #- !W S NYLG S NYLG TIYCTGCDLSLE TG TVYMFERMLSSK AG SVFCIDRFVSMT TG SVYKLDLYMNC ILG !#!!#- !#! !#! MRPIVELPTKFHIPYF F ISSIR VPN SKSQVDIPNE RQPQF TTANTSVS RLPQLPKNTYN R LPLSFTDYTIS RR PSIPLEC ISIP #S #R ## # ! !!#!##!# Figure 24. Alignment of Protein Sequences Homologous to ID1 of G A L 4 The amino acid sequences of the selected proteins were aligned by a B L I T Z search ( B L I T Z search scores of 60 or higher). The aligned residues were scored as either conserved hydrophobic residues (#), conserved polar uncharged residues (!), conserved acidic residues (-) or conserved basic residues (+) if 50% or more of the residues belonged to one of the categories. If 80% or more of the residues was the same amino acid, the one letter code for this amino acid was indicated. Three point mutants (*) in ID1 of G A L 4 were found to inactivate G A L 4 protein function. 69 similarity. ID1 may prove to have a conserved mechanism of protein-protein interactions, resulting in multimerization and regulation of protein function. 70 DISCUSSION 1. Glucose Inhibits GAL4 Through Its Central Region M y ribonuclease protection assay fully supports the evidence described by Stone and Sadowski (1993) that an important glucose repression mechanism occurs through the central region of G A L 4 . This central region is composed of a glucose-responsive domain or G R D , which is able to sense the presence of glucose, arid of three inhibitory domains (ID1, 2, 3) which carry out the inhibition of G A L 4 activity. The current model is that, in the absence of glucose, the G R D antagonizes the inhibitory domains, perhaps by a direct interaction with these domains. But, upon addition of glucose, a signal is generated which disrupts this interaction, allowing the inhibitory domains to function. This mechanism appears to work rather quickly, as I have demonstrated and as Stone and Sadowski have shown (1993). There are still many aspects to be clarified about this glucose repression mechanism. It should be established how the signal is relayed from glucose to G R D , whether there is a signal transduction pathway or whether a factor like glucose itself or a metabolite contact G R D directly. In fact, G . Stone (1992) has shown by a titration experiment that an unknown glucose responsive factor interacts with residues 693-767 of G A L 4 , within the G R D . Furthermore, it would be interesting to determine whether the phosphorylation sites that lie within G R D , such as Ser 699, are involved in the glucose signal transduction and the interaction with this unknown factor. Also, a detailed mechanism of how G R D and the inhibitory domains interact is necessary to understand how this regulation is achieved. Some missense mutations of G A L 4 within residues 320-520, the inhibitory region, have been found to inactivate G A L 4 (35). It is possible that these mutations disrupt the interaction between the G R D and the inhibitory domains, resulting in constitutive inhibition. Further characterization of these mutants 7 1 may lead to an informative part of the mechanism of action of the G A L 4 central region. 2. ID1 Is a Strong Inhibitory Domain Homologous to Other P r o t e i n s Stone and Sadowski (1993) had established that residues 320-412 of G A L 4 , a region termed ID1, had a strong inhibitory effect on transcriptional activation. M y in vivo results in yeast reconfirm this function for ID1. The mechanism of action of ID1, though, was still unknown. This mechanism appears to be evolutionarily conserved, since I have shown that ID1 functions not only in yeast, but also in E. coli, in wheat germ extracts, in rabbit reticulosyte lysates (data not shown), and even in mammalian cells (73). Furthermore, the homology that exists between ID1 and other fungal transcription factors as well as other proteins, seen in figure 24, suggests that the ID1 sequence may form a conserved structural motif with a possible important conserved mechanism of action. There appear to be many conserved hydrophobic and polar residues, as well as some charged and specific amino acids between ID1 and the other proteins. The homology to ID1 is found mostly in other fungal transcription factors such as L A C 9 , PUT3, L E U 3 , PPR1, N I R A , NIT4 and THI1 that are also involved in regulating a metabolic response to a particular signal, but also in other proteins such as C D C -6, involved in initiation of D N A replication, and PDR3, involved in pleiotropic drug resistance. If indeed the ID1 motif is conserved not only in sequence but also in function, "ID1" in these other fungal transcription factors may also work to promote protein multimerization, inhibiting the particular transcription factor under non-inducing conditions. As for how "ID1" may be regulated in the other transcription factors, it is plausible that each has its own "GRD-like" domain that masks inhibition when the inducing signal is present, but releases "ID1" when the signal is absent. It will be interesting to determine 72 whether these other transcription factors are regulated similarly to G A L 4 as far as the "ID1" mechanism of action is concerned. As for the similarity to proteins like C D C - 6 , which is not a transcription factor, the "ID 1-like" domain may equally serve a regulatory function to inactivate C D C - 6 through multimerization once D N A replication is no longer needed during the cell cycle. 3. I D 1 Impairs D N A Binding A transcription factor must first bind D N A to be able to activate transcription. M y E M S A and G A L 4 biotinylated oligo assay data suggest that ID1 severely impairs the D N A binding ability of G A L 4 . By preventing D N A binding, ID1 can inactivate G A L 4 . This ID1 inhibition mechanism agrees with the observation Stone and Sadowski (1993) had made that L e x A - V P 1 6 - I D l fusions extracted from yeast grown in glucose showed decreased D N A binding abilities to six L e x A operator sites in E M S A experiments as compared to fusions extracted from yeast grown without glucose. As well, Hirst and Sadowski (unpublished) observed that ID1 containing fusions expressed in E. coli and assayed on a A, phage repressor reporter displayed no D N A binding ability; this was not true, however, of derivatives containing ID2 and ID3, suggesting that the three inhibitory domains impose inhibition differently. Three independent modes of inhibition as opposed to a single one carried out by three separate domains may increase the stringency and rapidity of the regulation. The mechanisms of action of ID2 and ID3 still remain uncharacterized. The fact that ID1 can impair D N A binding has strong implications with respect to the fate of G A L 4 during glucose repression. As mentioned, G A L 4 is stringently repressed in the presence of glucose by several mechanisms. One of the mechanisms acts through the central region of G A L 4 itself and occurs rapidly. A quick way to ensure that G A L 4 is inactive in the presence of glucose would be by forcing G A L 4 off its D N A binding site, preventing G A L 4 from further activating transcription of the GAL genes. In the 73 meantime, the other glucose repression mechanisms would also ensure G A L 4 inactivation. The U R S Q binding repression complex would rapidly act on any G A L 4 dimers that are still bound to D N A , while further induction and GAL4 gene transcription are prevented, and G A L 8 0 mediated repression is re-established. A n interplay of all these mechanisms would make certain that no GAL gene transcription occurs in the presence of glucose. M y results regarding the impairement of D N A binding exerted by ID1 on G A L 4 also suggest another important result. Giniger et al. (1985) have suggested by in vivo footprinting experiments that G A L 4 is not bound to D N A in glucose; however, this result was complicated by the fact that GAL4 gene expression is repressed in glucose, questioning the amount of G A L 4 protein left in the cell to bind D N A . On the other hand, Lohr et al. (1985) argued that G A L 4 remains bound to D N A , regardless of carbon source, by D N A s e I hypersensitivity experiments. M y results, those of Stone and Sadowski (1993), and of Hirst and Sadowski (unpublished), then, would support the in vivo footprinting data of Giniger et al. (1985), and would argue against the results of Lohr et al. (1985). Since glucose repression upon G A L 4 through its central region occurs rapidly, the uncertainty of whether any G A L 4 protein remained in the cell is lifted. M y results, then, strongly support the evidence that G A L 4 is found bound to its U A S G site in the absence of glucose, but is removed from the D N A in the presence of glucose. 4. ID1 Prevents Dimer i za t ion and Promotes M u l t i m e r i z a t i o n Carey et al. (1989) showed that G A L 4 binds to its 17 mer dyad symmetrical site as a dimer. Without dimerization, which occurs partly through residues 65-94, G A L 4 cannot bind D N A . M y chemical crosslinking results suggest that the presence of ID1 in in vitro synthesized G A L 4 derivatives prevents the formation of homodimers. The lack of dimerization would impair D N A binding, as was observed with the D N A binding assays I performed. The mechanism seems specific to the ID1 region because ID1 containing 74 derivatives not only cannot homodimerize, but also cannot heterodimerize. The dimerization contact found in the D N A binding domain of G A L 4 is known to be somewhat weak. It is thought that there exists another dimerization domain in the C-terminus of the G A L 4 protein (32). Nevertheless, ID1 seems able to disrupt dimer formation. Alternatively, ID1 may promote aberrant dimers and multimers that cannot bind D N A . The mechanism of disruption of dimer formation by ID1 appears to be the formation of G A L 4 multimers, as indicated by the limited chemical crosslinking experiments I performed. By the molecular weights of the multimer species formed, these G A L 4 multimers seem to be homogenous. Some dimers were observed with the D B D - I D 1 - A R 2 G A L 4 derivative, and some E M S A results with histidine-tagged derivatives indeed suggest that this derivative has a very limited D N A binding potential, probably provided by the small proportion of normal dimers it can form. Nevertheless, it appears that higher order multimers and even monomers are favoured. It is still unclear, though, where the multimer contacts form and whether the multimers formed are aberrant with respect to the normal dimerization of G A L 4 molecules. Assuming that no other factors are involved from the molecular weights of the multimers, my chemical crosslinking results allow the possibility that ID1 multimerizes by contacting intermolecularly either G A L 4 ' s D B D , A R 2 or ID1 itself. A n important control to try would be a DBD-ID1 fusion. If DBD-ID1 also did not dimerize but formed multimers, it would imply that multimerization does not involve ID1-AR2 contacts. On the other hand, Stone and Sadowski (1993) observed that if either A R 1 or A R 2 is deleted from G A L 4 , G A L 4 is rendered inactive. In this case, it may be possible that ID1, a rather basic region, may be able to interact with either one of the two acidic activating regions, leaving one of them free to function. Upon deletion of either one of the activating regions, the second, being masked by ID1, may be unable to activate transcription. Finally, yet another observation comes from my heterodimerization experiments. Since D B D - I D 1 - A R 2 75 was unable to form heterodimers with either D B D nor D B D - A R 2 , this result raises the possibility that the multimerization contacts are intermolecular LD1-ID1 contacts. Further experimentation is required to resolve this aspect of ID1 functioning. 5. ID1 Does Not Require Other Cel lu la r Factors From the molecular weights of the multimers obtained with the limited chemical crosslinking experiments, it appeared that ID1 formed multimers that were homogenous, I pursued this issue further by purifying histidine-tagged G A L 4 fusions from E. coli to determine whether a protein or proteins copurified with the ID1 containing derivative but not with the other derivatives. Several proteins copurified with H I S - D B D - I D 1-AR2, but none were unique to this G A L 4 derivative purification. Furthermore, the purified D B D -ID1-AR2 derivatives behaved similarly in E M S A experiments, regardless of their level of purity, suggesting that no protein necessary for ID1 functioning was lost. The purification procedures, then, also support the chemical crosslinking results with the idea that ID1 does not require other proteins to function. Another possibility is that ID1 requires a cellular factor other than a protein to cause inhibition. This possibility was discredited by G . Stone (1992) since an attempt at titrating a factor binding to the ID1 region by overexpressing ID1 failed. The models I proposed above for direct multimer contacts between ID1 and D B D or A R 2 or ID1 itself, then, seem valid. 6. Compar i son of ID1 to Other K n o w n Inhibi tory Domains Many transcriptional activators have been shown to contain regions that lack sequence homology to ID1 but that also cause transcriptional inhibition. The mechanisms of action of these other inhibitory domains are somewhat similar yet different from G A L 4 ID1 mechanism of multimerization, impairing D N A binding. For example, the inhibitory P O Z domain found in poxvirus transcriptional activators such as ZID, Ttk and G A G A has been described to also 76 prevent D N A binding by promoting protein-protein interactions without the aid of other cellular factors (3). These P O Z domain promoted protein-protein interactions are believed to interfere with the interaction of the Cys2-His2 zinc finger D N A binding domain with D N A . The P O Z domain mechanism of inhibition is highly similar to that of ID1, but it is still unclear what part of protein activity ID1 multimerization interferes with. On the other hand, the inhibitory domain of transcription factors Ets-1 and Ets-2 causes inhibition of protein function by interfering with D N A binding directly and not homodimerization of Ets-1 and Ets-2 (30), unlike ID1. The opposite is true of the inhibitory domain found in the transcriptional activator E12, which interferes with homodimerization of E12 but not with D N A binding of MyoD-E12 heterodimers (75). Finally, other transcription factors such as Oct-2 and c-Fos contain inhibitory domains that unlike ID1 require other cellular factors to carry out inhibition by titration experiments (10, 48); c-Fos inhibitory domain 1, however, contains some important basic residues as also seems true for G A L 4 ID1. ID1 mechanism of action, then, is similar to that of other inhibitory domains, yet unique, representing a new mode of inhibit ion. 7 . Model of the Mechanism of Action of I D 1 Figure 25 depicts a summary model of the proposed mechanism of action of ID1 of G A L 4 . In the absence of glucose, the inhibitory domains of G A L 4 are believed to be kept inactive by the interaction, whether direct or through some intermediary factor, with the G R D . In the presence of glucose, a signal is relayed to the G R D , possibly in the form of the titratable glucose-responsive factor mentioned above. The interaction between the G R D and the inhibitory domains is thus disrupted, allowing the inhibitory domains to function. In the case of ID1, this inhibitory domain would cause protein multimers to form, either through ID 1-DBD, ID1-AR2 or ID1-ID1 contacts. These multimers are unable to bind D N A , rendering G A L 4 transcriptionally inactive. 77 Figure 25. Proposed Mechanism of Action of ID1 of G A L 4 In the absence of glucose (- G L U C O S E ) , the glucose response domain (GRD) interacts with the inhibitory domains (ID1, ID2, and ID3), either directly or through an intermediary factor, and antagonizes their function. G A L 4 is bound to D N A through its D N A binding domain (DBD) and can activate transcription through its two activating regions (AR1 and AR2) . Upon addition of glucose (+ G L U C O S E ) , a signal is generated which causes a glucose-responsive factor (G.R.F) to bind to G R D , freeing the inhibitory domains. ID1 can now act by causing the formation of G A L 4 protein multimers, by contacting either D B D , ID1 or A R 2 intermolecularly (possibility 1, 2 or 3, respectively). These protein multimers are unable to bind to D N A , rendering G A L 4 transcriptionally inactive. 7 8 - GLUCOSE A D O GRD ID3 INTERACTION D 1 i r DBD AR1 ID1 + GLUCOSE GLUCOSE DBD AR1 ID1 ID2 ID3 GRD AR2 GAL4 PROTEIN MULTIMERS ID1-DBD ID1-ID1 ID1-AR2 CONTACTS CONTACTS CONTACTS R E F E R E N C E S 1. Axel rod, J . D. , M . S. Reagan and J . Majors . 1993. G A L 4 disrupts a repressing nucleosome during activation of GAL1 transcription in vivo. Genes & Development 7: 857-69. 2. Baleja, J . D. , T. M a u and G . Wagner. 1994. Recognition of D N A by G A L 4 in solution: use of a monomeric protein-DNA complex for study by N M R . Biochemistry 33: 3071-8. 3. Bardwel l , V . J . and R. Treisman. 1994. The P O Z domain: a conserved protein-protein interaction motif. Genes and Development 8: 1664-1677. 4. Bhat, P . J . and J . E . Hopper. 1991. The mechanism of inducer formation in gal3 mutants of the yeast galactose system is independent of normal galactose metabolism and mitochondrial respiratory function. Genetics 128: 233-9. 5. Bhat, P . J . and J . E . Hopper. 1992. Overproduction of the G A L 1 or G A L 3 protein causes galactose-independent activation of the G A L 4 protein: evidence for a new model of induction for the yeast GAL/MEL regulon. Mol . Cell. Biol. 12: 2701-7. 6. Bhat, P. J . , D . Oh and J . E . Hopper. 1990. Analysis of the G A L 3 Signal Transduction Pathway Activating G A L 4 Protein-Dependent Transcription in Sac char omyces cerevisiae. Genetics 125: 281-291 . 7. B ram, R. and R. Kornberg. 1985. Specific protein binding to far upstream activating sequences in polymerase II promoters. Proc. Natl. Acad. Sci. U S A 82: 43-47. 8. B rand l , C . unpublished. 80 9. B r a n d l , C . J . , A . M . Furlanetto, J . A . Martens and K . S. H a m i l t o n . 1993. Characterization of NGG1, a novel yeast gene required for glucose repression of GAL4p-regulated transcription. E M B O Journal 12: 5255-65. 10. B r o w n , H . J . , J . A . Sutherland, A . Cook , A . J . Bannister and T . Kouzar ides . 1995. A n inhibitor domain in c-Fos regulates activation domains containing the H O B 1 motif. E M B O Journal 14: 124-31. 11. Buratowski , S. 1994. The basics of transcription by R N A Polymerase II. Cel l 77: 1-3. 12. C a r e y , M . , H . K a k i d a n i , J . Leatherwood , F . M o s t a s h a r i and M . Ptashne. 1989. A n Amino-Terminal Fragment of G A L 4 Binds D N A as a Dimer. J. Mol . Biol. 209: 423-432. 13. Cava l l i , G . and F . T h o m a . 1993. Chromatin transitions during activation and repression of galactose-regulated genes in yeast. E M B O Journal 12: 4603-13. 14. Celenza, J . L . and M . Car l son . 1986. A yeast gene that is essential for release from glucose repression encodes a protein kinase. Science 233: 1178-1180. 15. C h a n g , C . and J . D . G r a l l a . 1994. A critical role for chromatin in mounting a synergistic transcriptional response to G A L 4 - V P 1 6 . M o l . Cell . Biol. 14: 5175-5181. 16. C h a s m a n , D . I. and R. D . K o r n b e r g . 1990. G A L 4 protein: purification, association with G A L 8 0 protein, and conserved domain structure. M o l . Cell . Biol . 10: 2916-23. 17. Douglas, H . D . and D . C . Hawthorne. 1964. Enzymatic expression and genetic linkage of genes controlling galactose utilization in Saccharomyces. Genetics 49: 837-844. 8 1 18. Douglas, H. D. and D. C. Hawthorne. 1966. Regulation of genes controlling synthesis of the galactose pathway enzymes in yeast. Genetics 54: 911-916. 19. Erickson, J. R. and M. Johnston. 1994. Suppressors reveal two classes of glucose repression genes in the yeast Saccharomyces cerevisiae. Genetics 136: 1271-8. 20. Fields, S. and O. Song. 1989. A novel genetic system to detect protein-protein interactions. Nature 340: 245-246. 21. Flick, J. S. and M. Johnston. 1990. Two Systems of Glucose Repression of the GAL1 Promoter in Saccharomyces cerevisiae. M o l . Cell . Bio l . 10: 4757-4769. 22. Flick, J. S. and M. Johnston. 1992. Analysis of U R S G -mediated glucose repression of the GAL1 promoter of Saccharomyces cerevisiae. Genetics 130: 295-304. 23. Gill, G. and M. Ptashne. 1987. Mutants of G A L 4 Protein Altered in an Activation Function. Cell 51: 121-126. 24. Gill, G., I. Sadowski and M. Ptashne. 1990. Mutations That Increase the Activity of a Transcriptional Activator in Yeast and Mammalian Cells. Proc. Natl. Acad. Sci. U S A 87: 2127-2131. 25. Giniger, E. and M. Ptashne. 1988. Cooperative D N A Binding of the Yeast Transcriptional Activator G A L 4 . Proc. Natl. Acad. Sci. U S A 85: 382-386. 26. Giniger, E., S. Varnum and M. Ptashne. 1985. Specific D N A binding of G A L 4 , a positive regulatory protein of yeast. Cell 40: 767-774. 82 27. Gin iger , E . , S. M . V a r n u m and M . Ptashne. 1985. Specific D N A Binding of G A L 4 , a Positive Regulatory Protein of Yeast. Cell 40: 767-774 . 28. Griggs , D . W . and M . Johnston. 1991. Regulated expression of the G A L 4 activator gene in yeast provides a sensitive genetic switch for glucose repression. Proc. Natl. Acad. Sci. U S A 88: 8597-601. 29. Griggs , D . W . and M . Johnston. 1993. Promoter elements determining weak expression of the GAL4 regulatory gene of Saccharomyces cerevisiae. Mol . Cell . Biol. 13: 4999-5009. 30. H a g m a n , J . and R. Grosschedl . 1992. A n inhibitory carboxyl-terminal domain in Ets-1 and Ets-2 mediates differential binding of E T S family factors to promoter sequences of the mb-1 gene. Proc. Natl. Acad. Sci. U S A 89: 8889-93. 31. H i m m e l f a r b , H . J . , J . Pearlberg , D . H . L a s t and M . Ptashne . 1990. G A L 1 1 P : a Yeast Mutation that Potentiates the Effect of Weak GAL4-Der ived Activators. Cel l 63: 1299-1309. 32. H ir s t , M . and I. Sadowski . unpublished. 33. Johnson, P. F . and S. L . M c K n i g h t . 1989. Eukaryotic Transcriptional Regulatory Proteins. Annu. Rev. Biochem. 58: 799-839. 34. Johnston, M . 1987. A Model Fungal Gene Regulatory Mechanism: the GAL Genes of Saccharomyces cerevisiae. Microb. Rev. 51: 458-476. 83 35. Johnston, M . and J . Dover. 1988. Mutational Analysis of the GAL4-Encoded Transcriptional Activator Protein of Saccharomyces cerevisiae. Genetics 120: 63-74. 36. Johnston, M . , J . S. F l i ck and T . Pexton. 1994. Mult iple mechanisms provide rapid and stringent glucose repression of GAL gene expression in Saccharomyces cerevisiae. M o l . Cell . Biol . 14: 3834-41 . 37. Johnston, S. A . and J . E . Hopper . 1982. Isolation of the yeast regulatory gene GAL4 and analysis of its dosage effects on the galactose/melibiose regulon. Proc. Natl. Acad. Sci. U S A 79: 6971-5. 38. K a n g , T . , T . M a r t i n s and I. Sadowski. 1993. W i l d type G A L 4 binds cooperatively to the GAL1-10 U A S G in vitro. J. Biol . Chem. 268: 9629-35. 39. Keegan, L . , G . G i l l and M . Ptasnhe. 1986. Separation of D N A binding from the transcription-activating function of a eukaryotic regulatory protein. Science 231: 699-704. 40. K l a r , A . J . and H . O . Halvorson. 1974. Studies on the positive regulatory gene, GAL4, in regulation of galactose catabolic enzymes in Saccharomyces cerevisiae. M o l . Gen. Genet. 135: 203-12. 41. K l a r , A . J . and H . O . Halvorson. 1976. Effect of G A L 4 gene dosage on the level of galactose catabolic enzymes in Saccharomyces cerevisiae. J. Bacteriol. 125: 379-81. 42. Koleske, A . J . and R. A . Young . 1994. A n R N A polymerase II holoenzyme responsive to activators. Nature 368: 466-9. 43. L a m p h i e r , M . S. and M . Ptashne. 1992. Mult iple mechanisms mediate glucose repression of the yeast GAL1 gene. 84 Proc. Natl. Acad. Sci. U S A 89: 5922-6. 44. L a u g h o n , A . and R . F . Gesteland. 1982. Isolation and preliminary characterization of the G A L 4 gene, a positive regulator of transcription in yeast. Proc. Natl. Acad. Sci. U S A 79: 6827-31. 45. L a u g h o n , A . and R . F . Gesteland. 1984. Primary structure of the Saccharomyces cerevisiae GAL4 gene. Mol . Cell . Biol . 4: 260-267. 46. Leuther , K . K . and S. A . Johnston. 1992. Nondissociation of G A L 4 and G A L 8 0 in vivo after galactose induction. Science 256: 1333-5. 47. Leuther , K . K . , J . M . Salmeron and S. A . Johnston. 1993. Genetic evidence that an activation domain of G A L 4 does not require acidity and may form a beta sheet. Cell 72: 575-85. 48. L i l l y c r o p , K . A . , S. J . Dawson, J . K . Estr idge, T . Gerster , P . Matthias and D . S. L a t c h m a n . 1994. Repression of a herpes simplex virus immediate-early promoter by the Oct-2 transcription factor is dependent on an inhibitory region at the N terminus of the protein. M o l . Cell . Biol . 14: 7633-42. 49. L i n , Y . S., M . F . Carey , M . Ptashne and M . R . Green . 1988. G A L 4 derivatives function alone and synergistically with mammalian activators in vitro. Cell 54: 659-64. 50. L o h r , D . 1993. Chromatin structure and regulation of the eukaryotic regulatory gene G A L 8 0 . Proc. Natl. Acad. Sci. U S A 90: 10628-32. 51. L o h r , D . and J . E . Hopper . 1985. The relationship of regulatory proteins and DNase I hypersensitive sites in the yeast GAL1-10 genes. Nucleic Acids Research 13: 8409-8423. 8 5 52. L o n g , R . M . , L . M . M y l i n and J . E . Hopper . 1991. G A L 11 (SPT13), a transcriptional regulator of diverse yeast genes, affects the phosphorylation state of G A L 4 , a highly specific transcriptional activator. M o l . Cell . Biol . 11: 2311-4. 53. L u e , N . F . , D . I. Chasman, A . R . Buchman and R . D . K o r n b e r g . 1987. Interaction of G A L 4 and G A L 8 0 Gene Regulatory Proteins In Vitro. M o l . Cell . Biol . 7: 3446-3451. 54. M a , H . and D . Botstein. 1986. Effects of null mutations in the hexokinase genes of Saccharomyces cerevisiae on catabolite repression. M o l . Cell . Biol . 6: 4046-4052. 55. M a , J . and M . Ptashne. 1987. The carboxy-terminal 30 amino acids of G A L 4 are recognized by G A L 8 0 . Cell 50: 137-142. 56. M a , J . and M . Ptashne. 1987. Deletion analysis of G A L 4 defines two transcriptional activating segments. Cel l 48: 847-853. 57. M a r m o r s t e i n , R . , M . C a r e y , M . Ptashne and S. C . H a r r i s o n . 1992. D N A recognition by G A L 4 : structure of a protein-D N A complex. Nature 356: 408-414. 58. Melcher , K . and S. A . Johnston. 1995. G A L 4 interacts with T A T A - b i n d i n g protein and coactivators. M o l . Cell . Biol . 15: 2839-48. 59. Nehl in , J . O . , M . C a r l b e r g and H . Ronne. 1989. Yeast galactose permease is related to yeast and mammalian glucose transporters. Gene 85: 313-9. 60. Nogi , Y . and T . Fukasawa. 1989. Functional domains of a negative regulatory protein, G A L 8 0 , of Saccharomyces cerevisiae. M o l . Cell . Biol . 9: 3009-3017. 86 61. O h a s h i , Y . , J . M . B r i c k m a n , E . F u r m a n , B . M i d d l e t o n and M . C a r e y . 1994. Modulating the potency of an activator in a yeast in vitro transcription system. M o l . Cell . Biol . 14: 2731-2739. 62. P a r t h u n , M . R . and J . A . Jaehning. 1992. A transcriptionally active form of G A L 4 is phosphorylated and associated with G A L 8 0 . M o l . Cell . Biol . 12: 4981-7. 63. Per lman, D . and J . E . Hopper . 1979. Constitutive synthesis of the G A L 4 protein, a galactose pathway regulator in Saccharomyces cerevisiae. Cell 16: 89-95. 64. Ptashne, M . 1986. Gene regulation by proteins acting nearby and at a distance. Nature 322: 697-701. 65. Ptashne, M . 1988. How eukaryotic transcriptional activators work. Nature 335: 683-689. 66. Ptashne, M . 1989. How gene activators work. Scientific American. J a n u a r y 1989: 23-29. 67. Reece, R . J . and M . Ptashne. 1993. Determinants of binding-site specificity among yeast C6 zinc cluster proteins. Science 261: 909-11 . 68. Reece, R . J . , R . J . Rickles and M . Ptashne. 1993. Overproduction and single-step purification of G A L 4 fusion proteins from Escherichia coli. Gene 126: 105-7. 69. Rob ichon-Szu lmajs ter , H . D . 1958. Induction of enzymes of the galactose pathway in mutants of Saccharomyces cerevisiae. Science 127: 28-29. 70. Ronne, H . 1995. Glucose repression in fungi. Trends in Genetics 11: 12-7. 87 71. Sadowski , I. unpublished. 72. Sadowski , I. 1995. Uses for G A L 4 expression in mammalian cells. Genetic Engineering 17: 119-148. 73. Sadowski, I., B . Bel l , P. B r o a d and M . Holl is . 1992. G A L 4 fusion vectors for expression in yeast or mammalian cells. Gene 118: 137-41. 74. Sadowski , I., D . Niedbala , K . W o o d and M . Ptashne. 1991. G A L 4 is phosphorylated as a consequence of transcriptional activation. Proc. Natl. Acad. Sci. U S A 88: 10510-4. 75. Shirakata , M . and B . M . Paterson. 1995. The E12 inhibitory domain prevents homodimer formation and facilitates selective heterodimerization with the M y o D family of gene regulatory factors. E M B O Journal 14: 1766-72. 76. Silver, P. A . , L . P. Keegan and M . Ptashne. 1984. A m i n o terminus of the yeast G A L 4 gene product is sufficient for nuclear localization. Proc. Natl. Acad. Sci. U S A 81: 5951-5. 77. Stone, G . and I. Sadowski. 1992. G A L 4 is regulated by a glucose-responsive functional domain. M . Sc. Thesis, U . B . C . 78. Stone, G . and I. Sadowski. 1993. G A L 4 is regulated by a glucose-responsive functional domain. E M B O Journal 12: 1375-85. 79. S truhl , K . 1989. Molecular mechanisms of transcriptional regulation in yeast. Annu. Rev. Biochem. 58: 1051-1077. 80. T j i a n , R . and T . Maniat i s . 1994. Transcriptional activation: a complex puzzle with few easy pieces. Cell 77: 5-8. 81. T o r c h i a , T . E . , R . W . Hamil ton , C . L . Cano and J . E . H o p p e r . 1984. Disruption of regulatory gene GAL80 in 88 Saccharomyces cerevisiae: effects on carbon-controlled regulation of the galactose/melibiose pathway genes. M o l . Cel l . Biol . 4: 1521-1527. 82. Treitel , M . A . and M . Carlson. 1995. Repression by SSN6-TUP1 is directed by M I G 1 , a repressor/activator protein. Proc. Natl. Acad. Sci. U S A 92: 3132-3136. 83. Trumbly, R. J . 1992. Glucose repression in the yeast Saccharomyces cerevisiae. Molecular Microbiology 6: 15-21. 84. Tzamarias, D . and K . Struhl . 1994. Functional dissection of the yeast C y c 8 - T u p l transcriptional co-repressor complex. Nature 369: 758-761. 85. VanHoy, M . , K . K . Leuther, T. Kodadek and S. A . Johns ton . 1993. The acidic activation domains of the G C N 4 and G A L 4 proteins are not alpha helical but form beta sheets. Cel l 72: 587-594 . 86. Workman , J . L . , I. C . Taylor and R. E . Kingston. 1991. Activation domains of stably bound G A L 4 derivatives alleviate repression of promoters by nucleosomes. Cell 64: 533-44. 87. Yocum, R., S. Hanley, R. West and M . Ptashne. 1984. Use of lacZ fusions to delimit regulatory elements of the inducible divergent GAL1-GAL10 promoter in Saccharomyces cerevisiae. M o l . Cell . Biol . 4: 1985-1998. 88. Y u n , S. J . , Y . Hiraoka , M . Nishizawa, K . Takio , K . Ti tani , Y . Nogi and T. Fukasawa. 1991. Purification and characterization of the yeast negative regulatory protein G A L 8 0 . J. Biol . Chem. 266: 693-7 . 89 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0087015/manifest

Comment

Related Items