we , l., W t~;-w,--.“m:3;-v:: ' t 151;; *gflu 3a; ‘ p5: W; M ‘3' Wang r - .V H 32‘; ' ' x 3“ MMY'K' I '. 4 I f" :L} .- z- “a; , “9751 ti: ‘ ,.-- ‘J u "'5- “.1 : - . I » :‘L '7‘ 3. 1‘“ :3 P m... : wu'mé‘. ' . ch --". -. amps“ a» 9-... " v- . ' ' . , S- Ty-z this, .' - ‘v 33.513 ' ' - a, “Egg-{ma A f." u- <7 “. ‘21 1.... '- -. ‘ . .. , . " “1”! “i . . I __fi}d£,}.i ; . " 'V 5 7- ' ‘53:" LV.“_ . ' ’ _ ‘ .17 “MS- is; s. .' . g‘n» 3.“. . 1‘ ~_ .g; v: ”1‘“ :‘L‘ ..'- .’ . , 2‘ , . “.§L.,‘l§;§hfi ' o 3 xi -.. . . .m- *. = ‘. :~ ~ . 113W”: ‘3? V. . a . t w \ “i " ' l 91:. ‘3’- : km a f w r?;.‘uc3 .131 :2 g?! Fir; C i s we: . . wiptvu‘féf 3;} «4.15;. -. .." ‘ . fi 1 -"," m 7 ~I~ . Mg? v. a: l ' i 'i . q . ski? "' r}! . ‘3’; I . ._.. d. 1.27;»— IV «a -v- "' . n“,- .'.‘..’:u— v ' -. ,..;..:g,.j- . ' 13‘...‘ ~ ..... .. a. . . Lynn!”- “4,- ‘ x g. u , :b‘.‘ ' -3.§ .‘vt 'li‘J ~. 4 nun-:1 - .2-l~— . - * “h 5t": J“ 1‘ 9 3 If,“ . . ”5-37 igzjaig‘fitét - ~ - .(I a “1:32! {1.112511 ‘ ¥ 1 i1; ‘5 - v -o.-u.n . 4. -. v . 2. ”m " L.- ‘ . .. J .. ~. MC; . .u‘nn -"l-." , 5' ijé’; I: . . , . g - ‘ 1 “if J " .U' ‘ 'l‘h Q" I . “ ' ‘TL‘ 1 V —.‘ ll 1“". ‘ ' . -. I! ., .‘31:‘,‘~H- {€55 migmggz mg: 23 . L 5%,.“ 9:1? 1 n; . .1 ‘ ‘il‘iA 1 9‘ 1‘“? a- x -, ., :34; ~ . .rii‘ . F 4" "1;, . I ' —' flux» .-.-.—‘- $3,333-. w “(fig-git '49 .shififii'fifi- ,~ ,5. 173mg“) egaz 5?: a: v... "I '35: D--- -41 . w-‘~ ‘ s. . .2 r, ‘ 'r fin ‘Cgh .v. “’,....-n_ H. " :1. u. 1" "’ "'1: a. 59- ' “m a .1: “0. u- .mfl—‘_ vfi mr VT. .« .1319 1 . " '\ . '3 3- H i “M. :‘v 333%?!“ rt‘ 33?: it“; A. L ... M? R 'c C a»: -.... '2; m a-. .w0 ”I. u vA-.-J"’ ‘ ’0 , * .. v us ‘1'." ‘t ::.::r- w~v< .133”- )? .3- ... . .... ’ i “”7 ..- M.“ mu“ a ‘ --w* "at? $25.», .J-n ‘-““' .5} “3 1:! v a. ‘0 {c 1.! ’1' , "35‘ (- rw‘ FEE. 7 3 u . "‘ ,- . _.,_.. M ,, A, . O @ %§153 x..::;"".——— --’ " Ana - > v... n 5’ Q‘ I ~ ‘12 .4 ‘v‘ri‘ .1 5..- .. m... minimum; 93! l ,2 z 0:. LlfiRARY Michigan state University I'll/III!!!(lillll/Nl/l/IIWI 2074 1926 This is to certify that the thesis entitled ANALYSIS OF A TRANSCRIPTIONAL REPRESSOR LN DROSOPHLLA MELANOGASTEK presented by Y1 FAN MAO has been accepted towards fulfillment of the requirements for M. S. degree in BIOCHEMISTRY QMM Major professor 6/3/50 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINE-3 return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 11m WWW-p.14 ANALYSIS OF A TRANSCRIPTIONAL REPRESSOR IN DROSOPHILA MELANOGASTER Yifan Mao A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Biochemistry 2000 ABSTRACT ANALYSIS OF A TRANSCRIPTIONAL REPRESSOR IN DROSOPHILA MELANOGASTER By YIFAN MAO The repressor protein Knirps regulates gene expression in Drosophila embryonic development. I have been using an in viva repression assay to define the repression domains of the Knirps protein. A panel of mutant knirps gene constructs were generated and transformed into Drosophila embryos by microinjection. The repression activities monitored in embryos suggested that two portions of the Knirps protein can mediate repression. One region contains the binding site for dCtBP, a cofactor previously shown to be important for Knirps function. In addition, a more N-terminal portion of Knirps is able to repress transcription without the dCtBP binding sites. Thus, Knirps may fimction via two different pathways. Residues 202-358 and residues 139-330 have been mapped to the minimal C-terminal repressor and N-terminal repressor respectively. By the comparing the repression activities of repression domain with different DNA binding domains, the Knirps DNA binding domain is proposed to exert negative interference on the repression domain. To Mum and Dad ACKNOWLEDGMENTS I am grateful to those who have helped me on this project. First of all, I want to thank my advisor, Dr. David Arnosti, for his teaching, patience and support. During the three years I spent in Dr. Arnosti’s lab, I have enjoyed the cooperative environment provided by Dr. Arnosti’s group and have grown into a mature scientist. I want to thank all the current and previous Amosti lab members for their help and fiiendship, especially Dr. Scott Keller for his analysis of the N-terminal repressor Knirps 75-331. I also want to thank Dr. Zachary Burton, Donald Jump for being my committee and for the guidence they gave me over these years. I thank Dr. Laura McCabe for helping me with gel shift experiments. I also thank Dr. Steve Small from New York University for providing the Casper 22FAKE plasmid and Dr. Nibu fi'om University of California at Berkeley for providing the Gal4-dCtBP fly lines. I also thank Kirsten F ertuck for reading my thesis. Finally, I want to give my special thanks to my family, for their support and faith. TABLE OF CONTENTS LIST OF FIGURES .................................................................. viii CHAPTER 1. LITERATURE REVIEW .......................................... 1 Overview of the transcriptional regulation in eukaryotes ..................... 2 Transcriptional regulation in Drosophila embryonic development .......... 4 Introduction of the knirps gene and the Knirps protein ........................ 5 The transcriptional corepressor of Knirps, dCtBP .............................. 10 CHAPTER 2. THE REPRESSION ACITIVITIES OF KNIRPS ............... 18 Short-range repression activity of the Knirps protein .......................... 19 Gal4-Knirps, the model for the study of Knirps ................................ 20 Mutational analysis of the Gal4-Knirps chimera ............................... 23 CHAPTER 3. THE MINIMAL REPRESSION DOMAIN OF KNIRPS . 27 Introduction ......................................................................... 28 Materials and methods ............................................................ 28 Plasmids ........................................................................... 28 P-element transformation whole-mount in situ hybridization of embryos, and crosses to reporter lines .................................................... 30 Assays of in vivo repression activity .......................................... 30 Results .............................................................................. 33 202-358 is a minirml repression domain of Knirps protein ............... Activity of the dCtBP repressor protein fi'om a distal enhancer position CHAPTER 4. THE ANALYSIS OF THE N-TERMINAL REPRESSOR OF KNIRPS .............................................................................. Introduction ........................................................................ Materials and methods ............................................................ N-terminal deletions ............................................................ P-element transformation whole-mount in situ hybridization of embryos, and crosses to reporter lines ................................................... Assays of in vivo repression activity .......................................... Results .............................................................................. Discussion and conclusion ....................................................... CHAPTER 5. Knirps DNA binding domain, the inhibitory element in repression. . . ............................................................................................... Introduction ........................................................................ Materials and methods ............................................................ C-terminal deletions ............................ . ................................ 33 43 47 48 48 49 51 51 51 52 59 60 61 61 P-element transformation who le-mount in situ hybridization of embryos, and crosses to reporter lines ......................................................... Assays of in vivo repression activity .......................................... vi 62 62 Results .............................................................................. 62 Discussion and conclusion ....................................................... 63 CHAPTER 6. DETECTION OF THE CHIMERIC GAL4-KNIRPS REPRESSOR PROTEINS ............................................................................................. . 74 Introduction ........................................................................ 75 Materials and methods ............................................................ 75 Whole mount in situ antibody staining ....................................... 75 Embryo nuclear protein extraction ............................................ 78 Gel mobility shilfis .............................................................. 79 Results .............................................................................. 80 Detections of 11181011 proteins by gel mobility shift assay .................. 8O 202-358AE, the non-functional protein, is stably expressed in the transgenic embryos ........................................................................... 83 Discussion ........................................................................... 90 Conclusion ........................................................................... 91 Surmnary and perspectives ........................................................... 92 Bibliography ........................................................................... 94 vii LIST OF FIGURES Images in this thesis are presented in color. Chapter 1 Figure l. The schematic representation of the Knirps protein ...................... 7 Figure 2. The P-DLS-K motif is essential for Knirps-mediated repression ...... 11 Figure 3. Expression pattern of synthetic reporter genes in the dCtBP mutant embryo .................................................................................... 14 Chapter 2 Figure 4. Gal4-Knirps mimics the activity of endogenous Knirps protein on a rhomboid enhancer ..................................................................... 21 Figure 5. dCtBP interaction region of Knirps is dispensable for activity of the C- terrninal truncated proteins ............................................................ 24 Chapter 3 Figure 6. The making of a transgenic fly ............................................ 31 Figure 7. The injection vector pTwiggy ............................................. 34 Figure 8. Representative repressed embryos show the repression of eve stripe 2 by Gal4-Knirps ............................................................................. 37 Figure 9. The schematic structures and in vivo activities of the Gal4-Knirps mutant proteins ................................................................................. 39 Figure 10. A central region of Knirps requires dCtBP binding motif for acitivity 41 Figure 11. The Gal 4-dCtBP acts as an independent repressor within an enhancer 45 Chapter 4 Figure 12. The schematic structures of the amino-terminal Knirps mutants. . .. 53 Figure 13. Sequence comparison of Knirps-homologous proteins fiom different species .................................................................................... 56 viii Chapter 5 Figure 14. The schematic structures of the carboxyl-terminal Knirps mutants and their repression activities ............................................................ 64 Figure 15. The schematic representation of model I: dCtBP helps Knirps dimerization to perform effective repression ...................................... 67 Figure 16. The schematic representation of model 11: The DNA binding domain interferes with the repression of Knirps ........................................... 69 Figure 17. The schematic representation of model 111: The repression domain interferes with the DNA binding of Knirps ....................................... 72 Chapter 6 Figure 18. Gal4-Knirps 189-358 were detected by antibody staining ......... 76 Figure 19. Gal4-Knrips 202-358 protein binds to its cognate sites specifically 81 Figure 20. Gal4-Knirps l89-358-flag, 202-358-flag and 248-358-flag were stably expressed in transgenic embryos ................................................... 84 Figure 21. 202-358AE-5 and 202-358AE-10 were stably expressed in transgenic embryos ................................................................................. 88 CHAPTER 1 LITERATURE REVIEW Overview of the transcriptional regulation in eukaryotes Eukaryotic transcription is carried out by three DNA-dependent RNA polymerases (reviewed by Young 1991): RNA polymerase I, II and III. RNA polymerase I synthesizes ribosomal RNA; RNA polymerase II synthesizes messenger RNA and some small nuclear RNA; RNA polymerase II synthesizes the SS rRNA, tRNA and some small nuclear RNA. Each of these three RNA polymerases is composed of 8-14 polypeptides. Most eukaryotic Pol II promoters that recognized by RNA polymerase 11 contain cormnon elements such as a TATA box, GC rich sequences or CAAT boxes. To initiate transcription, multiple transcriptional factors (TFs) are required for accurate initiation fiom a promoter DNA sequence. These transcriptional factors include TBP (TATA-binding protein) or TFIID, TFIIA, TFIIB, TFIIF, TFIIE and TFIIH (Zawel and Reinberg, 1993, Roeder, 1996, Orphanides et al., 1996). Similar TFs have been identified in yeast, Drosophila, rat and human systems. These TFs along with RNA polymerase II are oflen referred to as the general transcriptional machinery. In eukaryotes, DNA was packaged into chromatin and DNA sequences within the chromatin structure are generally inaccessible to transcription factors and Pol II. Transcriptional activators can relieve this repression caused by chromatin structure by direct or indirect covalent modification of the histones to loosen the packaging of the nucleosomal DNA. Activators also possess the ability to facilitate transcription by acting on the general machinery to increase the efficiency of the process. Like transcriptional activation, transcriptional repression is important for appropriate gene expression in vivo by turning off transcription (Hanna Rose and Hansen, 1996). Repressors play a major role in many biological processes. Mutations in repressor genes may cause tumor growth, developmental disorders or disregulation of the cell cycle (Wang et al., 1999). There are four different models proposed regarding the mechanisms of repressors in eukaryotes. The first one is called the competition model; in which the DNA binding sites of the repressor overlap with those of the activator. The repressor binds to the enhancer and prevents the activator fiom binding to DNA (F oulkes et al., 1991). The second model is called the quenching model. Both the activator and the repressor are capable of binding to DNA simultaneously, but the repressor keeps the activator fi'om functioning by masking the activation surface and preventing it from making prOper contact with the basal machinery. Some repressors can repress multiple enhancers from long distances (over 500 bp). The third model of repression is that the repressor somehow prevents the formation of the transcription complex, thus blocking transcription. The Drosophila protein Even-skipped is an example of a repressor that uses this mechanism. This protein represses transcription by binding to TBP and blocking the TFIID-TATA box interaction (Li et al., 1998). The fourth model is the squelching model. In this model, the repressor interacts neither with DNA nor the activator protein, but represses transcription by sequestering factors required for the fimction of the activator. The estrogen receptor (ER) represses the transcriptional activation by progesterone receptor (PR) in this manner (Etienne et al., 1989) Transcriptional regulation in Drosophila embryonic development Transcription is regulated by factors that result in highly selective gene expression in organisms. In Drosophila, enhancers integrate both positive and negative regulatory information to direct sharp patterns of gene expression in the embryo. These enhancers contain tightly clustered binding sites for both transcriptional activators and repressors (J fickle 1992) and lead to the cross and hierachical gene regulation cascade. Activators, such as Bicoid (Driever et al., 1989), turn on gene expression, while repressors, such as Knirps (Amosti et al., 1996), either reduce or turn off gene expression completely. Thus, transcriptional regulation of one gene in the Drosophila embryo usually involves multiple transacting factors. For example, in Drosophila melanogaster, even-skipped stripe 7 is activated by Tailless and repressed by Knirps. The interplay of these multiple transcriptional regulators causes the specific developmental pattern of Drosophila embryos. Drosophila segmentation also depends on the successive activities of several classes of genes; The maternal gene products Bicoid, localized in the anterior pole of the embryos, and Nanos, localized in the posterior pole of the embryo, set up the anterior- posterior axis of body pattern formation. Those maternal gene products provide the position information for the zygotic genes, including the gap genes, pair-rule genes and segment-polarity genes (for reviews, see Niisslein-Volhard et al., 1987). In the hierarchical gene regulation process involved in embryonic development, the gap genes respond to the regulation of matemal gene products. In turn, the gap genes regulate expression of the pair-rule genes and the pair-rule genes regulate their downstream genes including the segment polarity genes (for reviews, see Pankratz and Jackle, 1993). This sequential regulation of gene expression determines the fate of the embryonic development of Drosophila. Transcriptional repression is a complicated issue which people are addressing from difl‘erent respects. Drosophila is an ideal system to study transcriptional repression because of several advantages: Drosophila is an eukaryotic system; it has a relatively short generation time; it is relatively easy to maintain and manipulate; Drosophila genetics has been well studied. Most importantly, the same types of proteins are found in other higher eukaryotes. Since Knirps is a transcriptional repressor that plays a major role in the embryonic development of Drosophila melanogaster, the study of Knirps will have significance in both developmental biology and the study of transcription mechanisms. Introduction to the knirps gene and the Knirps protein The knirps gene is one of the gap genes involved in early embryonic development of Drosophila melanogaster. It encodes a transcriptional repressor that regulates correct spatial and temporal gene expression in the early Drosophila embryo. The knirps gene was identified in a screen for early developmental defects, when a type of knirps mutant embryo was found to form only two abdominal segments instead of the normal eight segments in the embryo (N iisslein-Volhard and Wiesclmus, 1980). Knirps protein is expressed in the anterior-ventral region and the posterior region of the blastoderm embryo, which will give rise to the abdomen between nuclear division cycle 11 and 14 (Figure 1) (Rothe et al., 1992). It is known that the products of other genes including activators such as Kriippel, Caudal, Bicoid (Rivera-Pomar, et al., 1995) and repressors such as Giant and Tailless, set the boundaries of expression of Knirps (Pankratz, et al., 1993) Knirps has been well studied genetically and a series of mutant alleles have been identified: 1. a number of null alleles that causes the embryo to lack six out of eight abdominal segments; 2. an intermediate hypomorphic allele that result in the loss of three or four abdominal segments; and 3. a weak hypomorphic allele that leads to the loss of one or two abdominal segments (Nauber et al., 1988). The knirps gene has been cloned and located on the left arm of the third chromosome, based on the evidence that certain chromosomal rearrangements are associated with a knirps mutation (N auber et al., 1988). Knirps protein is a member of the nuclear hormone receptor superfamily (N auber et al., 1988). A CysZCysZ type zinc-finger DNA binding domain (residues 1-74) and a repression domain (residues 75-429) (Figure 1) functional when separated (Gerwin 1994, Amosti et al., 1996). The “Kni box” between residue 52 and 93 is highly conserved in the Knirps-like proteins of Drosophila melanogaster, Drosophila virilis and the more distantly related Musca domestica. Knirps protein possesses a zinc-finger DNA binding domain and binds to DNA sequence 5’-ACTGAACTAAATCCGG-3’. Knirps is capable of binding to DNA as a monomer (Gerwin et al., 1994). Therefore, its DNA binding properties can be distinguished from those of the known CysZCysZ type nuclear receptors, which require homodirner or heterodirner formation. In vivo, Knirps acts as a repressor to define the posterior boundary of even-skipped stripe 3 (Small et al., 1992) and to define the hairy stripe 7 in the anterior boundary (Pankcratz, M. and Jackle, H., 1993). Mutating knirps gene in embryos results in the expanding of the expression Figure l. The schematic representation of the Knirps protein (top) and the expression pattern of the endogenous Knirps protein (bottom). Images in this thesis are presented in color. 5905 3:5. msocomoucm EmEoc uses.“ $6.. guesses.“ nonczcu . E «SE! mue 5m 3." a: ma Eon oEEe patterns of those stripes due to the loss of repression (Small et al., 1996), eventually causing embryonic lethality (Niisslein-Volhard, C. and \Vieschaus. 1980). There are two classes of repressors: long-range repressor versus short-range repressor. The Drosophila long-range repressor Hairy is capable of working over 1 kb away fi'om the upstream activator to repress the rhomboid NEE enhancer (Barolo et al., 1997). Knirps is a member of the short-range class of repressors, which includes Smil (Gray et al., 1994), Krfippel (Gray et al., 1996) and Giant (Hewitt et al., 1999). These factors act on nearby DNA sequences to repress gene expression. For example, Knirps represses an even-skipped stripe 2 enhancer element, when it is bound less than 100 bp away from this enhancer (Amosti et al., 1996). Knirps can also fimction as a dominant repressor, blocking multiple enhancers, such as the twist enhancers and the rhomboid enhancer, when it is bound 75 bp or less away from the transcription start site (Amosti et al., 1996). It has been shown that Knirps protein contains several functional domains which are arranged in a modular fashion. The alanine-rich part of the Knirps protein (encoded by codons 189-254) acting as a functional repressor in Drosophila Schneider cells (Gerwin et al., 1994). The transcriptional corepressor of Knirps, dCtBP Co-factor mediated repression has been commonly discovered in a variety of systems. In mammalian cells, the transcriptional repression by the methyl-CpG-binding protein MeCP2 requires a corepressor complex containing the transcriptional repressor mSin3A and histone deacetylase (Nan et al., 1998). During an investigation of the mechanisms of Knirps repression function, dCtBP, the Drosophila homologue of the mammalian C-terminal binding protein, was shown to be involved in the repression activities of Knirps protein (Nibu et al., 1998b). The mammalian protein, CtBP (C- terminal binding protein) interacts with the adenovirus E 1 a protein and decreases the transformation activity of the Ela protein (Schaeper et al., 1995). This interaction was suggested to be critical for adenovirus replication and oncogenic tranformation. A specific sequence motif P-DLS-K, highly conserved among ElA proteins of various human and animal adenovirus, is required for the CtBP-BIA interaction (Schaeper et al., 1995). dCtBP protein is found to be maternally deposited into Drosophila embryos, involving in developmental processes (N ibu et al., 1998b). dCtBP protein recognizes the same motif P-DLS-K in Knirps protein, and alanine substitution mutation in the DLS sequence abolished interaction between dCtBP and Knirps in in vitro glutathione-S transferase pull-downs (N 1bu et al., 1998b). The evidence as following supports the hypothesis that the P-DLS-K motif is essential for Knirps-mediated repression: 1. St2- Knirps was constructed by fusing the even-skipped stripe 2 enhancer to the knirps gene and was transformed into embryo. The transgenic embryo displays an ectopic Knirps expression pattern localized as a stripe in the anterior of the embryo (Figure 2A). The products of this St2-knirps repress the even-skipped stripe 3 (F igure2B). The repression is not seen when a st2-knirps mutant gene with the alanine substitutions in the dCtBP- binding motif was transformed into the embryo (Figure 2C) (Nibu et al. 1998a). 2. The results from analyzing the expression of synthetic transgenes in dCtBP mutant embryos. The transgene contains a modified form of 700 bp rhomboid lateral stripe enhancer (NEE) and directs strong expression in both lateral and ventral regions. The wild type 10 Figure 2. The P-DLS-K motif is essential for Knirps-mediated repression. Cellularizing embryos were hybridized with mixtures of a digoxigenin-labeled knirps antisense RNA (red) and a fluorescein- labeled eve antisense RNA (black). They are orientated with anterior to the left and dorsal up. (A) Double staining pattern in a wild-type embryo. eve is expressed in a series of seven stripes, while knirps is expressed at the anterior pole and antero-ventral regions, as well as in a broad posterior band which encompasses eve stripe 4 and 5. (B) Same as (A) except that the embryo contains a transgene with the full-length knirps coding region was mutagenized to disrupt the P-DLS-K motif (PMDLSMK to AAAASMA). The ectopic knirps stripe does not cause an obvious change in the eve pattern; in particular, stripe 3 pattern, suggesting that the mutant Knirps protein retains weak repressor activity I] st2-kni C st2-mutkn w . A embryo, where dCtBP protein is maternally deposited, contains a modified enhancer NEE with two synthetic Knirps—binding sites positioned within 50 bp of the NEE activators. Repression of the reporter gene expression in the ventral region was observed due to the repression activity of the Knirps protein (Figure 3); whereas the dCtBP mutant embryo without contains the same NEE enhancer and the synthetic Knirps binding site. No repression was observed (Figure 3) (Zhang et al., 1998), suggesting that dCtBP is necessary for Knirps repression. 3. Alanine substitution mutations in PMDL sequence in the dCtBP-binding domain eliminate the repression activity of the Gal4-Knirps 254-429 in transgenic embryo assays (N ibu et al., 1998b). 4. Gene dosage assays also suggest that Knirps and dCtBP interact in viva. Embryos that are heterozygous for the knirps 9 null mutation exhibited occasional defects in the eve expression pattern. Combining the dCtBP and knirps mutations resulted in more severe disruptions in the eve pattern (N ibu et al., 1998b). Both CtBP and dCtBP have been reported as corepressors for an increasing number of transcriptional factors. CtBP is reported involved in the mechanism of Rb/pl 3O transcriptional repression (Meloni et al., 1999). CtBP interacts with zinc finger/homeodomain protein ZEB and its Drosophila homologue zih-l . CtBP 1 and CtBP 2 are two CtBP proteins with slight sequence variations that were discovered to interact with mouse SEF 1 protein in a yeast two hybrid screen (F urusawa et al., 1999). dCtBP protein mediates transcriptional repression by Knirps, Krfippel and Snail in the Drosophila embryo, and interacts with Drosophila protein Hairy in the embryo (Poortinga et al., 1998). dCtBP is also suggested to interact with polycomb group proteins 13 Figure 3. Loss of Knirps-mediated repression in a dCtBP mutant embryo. LacZ-white reporter genes were introduced into dCtBP mutant embryos and stained after in situ hybridization with a white antisense RNA probe. Cellularized embryos are oriented with anterior to the lefi and dorsal up (A and B). white staining patterns in a wild-type (A) and dCtBP mutant embryo (B). The reporter gene contains a rhomboid lateral stripe enhancer (NEE) and two synthetic Knirps-binding sites flanking the four Dorsal activator sites (see diagram beneathe the embryos) (From Nlbu et al., 1998a). 14 New. 22; .311. .L.. 5.90 on? 2.3 15 (Sewalt 1999). Recent studies suggest that ectopic expression of the native Krfippel protein causes patterning defects in early embryos, which are reversed when the P-DLS- H motif is mutagenized (Zhang et al. 1999). These results suggest that Krfippel mediated repression also depends on the recruitment of the dCtBP corepressor. Although both CtBP and dCtBP have been shown to interact with a number of transcriptional regulator proteins, the mechanisms through which they affect developmental processes are unclear. Both CtBP and dCtBP were reported to display significant homology to the D isomer 2-hydroxy acid dehydrogenase family (Schaper et a1, 1995, Nibu et al., 1998b). Members of this family have so far been identified only fi'om bacteria, plant and lower fungi (Arthur et al., 1991, Taguchi et al., 1991, Chow et al., 1993). However, neither CtBP nor dCtBP shows any dehydrogenase enzyme activity or NAD binding activity (Schaper et al., 1995). The mammalian CtBP protein has been shown to interact with the histone deacetylase HDACl (Sundqvist et al., 1998). Meanwhile, it was also reported that the SV 40 promoter, shown to be repressed by CtBP, is relatively insensitive to histone deacetylase inhibitors (Meloni et al., 1999), suggesting that this is unlikely to be the primary mechanism of CtBP mediated repression. Previous reports showed that dCtBP interacts with the repression domain of Knirps protein though the conserved motif P-DLS-K Meanwhile, it is a common phenomenon that one repressor could contain multiple repression domains which can independently exert repression function on its target genes. One example is the Drosophila Engrailed protein, in which two separate repression activities have been identified. One domain provides most of the activity in embryonic assays and on 16 integrated genes, and binds the Groucho corepressor, a protein that also interacts with the Hairy, Runt, and Dorsal repressors. Another region of Engrailed contains an activity that appears to be primarily active on transiently transfected reporter genes and does not bind Groucho (Tolkunova et al., 1998). The Hairy protein has also been found to contain multiple activities; Hairy also interacts with the Groucho cofactor via a C terminal WRPW motif, and a separate “Orange” domain mediates antagonistic activities toward Scute protein. Recent work shows that Hairy can also interact with the dCtBP protein via a third region of the protein (Poortinga et al., 1998), although it is not clear whether this interaction contributes to repression (Zhang et al., 1999). Krfippel has also been found to contain two separable repression domains, both of which are active in cell culture assays; these two domains have been reported to have different activator specificity (Hanna-Rose et al., 1997). Knirps repressor has been suggested to possess at least two independent repression domains. These two repression domains function autonomously to repress transcription through distinct mechanisms (Mao et al., manuscript in preparation). One is the C-terminal portion of the knirps protein, acting through a dCtBP dependent mechanism, and the other repression domain is located in the N-terminal region. The N- terminal repressor apparently functions through a dCtBP independent mechanism although the specific residues involved and molecular mechanisms of this repression activity are not clear. 17 CHAPTER 2 THE REPRESSION ACITIVITIES OF KNIRPS 18 Short-range repression activity of the Knirps protein Short-range repression activities have been observed in the behavior of a number of repressors, in which repression occurs only when the targets are near the repressor. Krilppel and Snail are examples of short-range repressors. Knirps represses transcription within a certain distance fi'om either the promoter or the enhancer. Previous studies indicated that endogenous Knirps was capable of blocking transcription when the binding sites were situated at —55 bp, but not at —130 bp from a heterologous promoter (Amosti et al., 1996). Knirps can also repress a heterologous enhancer over a short distance (Amosti et al., 1996). To determine the exact distance dependence of Knirps repression, and test whether the loss of activity with increasing distance represents a “step fimction” or a gradual tapering off of activity, transgenes were designed with tandem binding sites situated at -55, -70, -75, -100, -130 or —180 bp. These genes were introduced into Drosophila by P-element mediated germ line transformation, and the embryonic expression patterns of the transgenes were analyzed by in situ hybridization. Strong repression was observed in genes with Knirps binding sites whose 3’ edges were situated at -55bp, -70, or -75 bp. The Knirps binding sites used in this gene have been shown to confer repression in a Knirps-dependent manner (Amosti et al., 1996), and the repression was observed only in the presumptive abdomen and ventral anterior regions, where the knirps gene is expressed (J fickle 1992). Repression was less efl‘ective in the —100 bp and —130 bp constructs, and almost undetectable in the construct with the sites at -180 bp. No differences were observed in the pattern of genes with repressor sites situated at -70 or - 19 75bp, indicating phasing effects are not important on this reporter (Mao et al., manuscript in preparation). Gal4-Knirps, the model for the study of Knirps Mutagenesis of the endogenous knirps gene leads to embryonic lethality. To study the mechanisms of the repression behavior of Knirps protein in embryos without interfering with the normal embryonic development, various recombinant knirps genes were constructed by replacing the DNA binding domain codons 1-74 of Knirps with the DNA binding domain codons 1-93 of the yeast Gal4 protein. Two reporter systems were constructed and subjected to the in viva repression assay, in which a lac Z reporter gene was under the control of two enhancer elements. One was the even-skipped stripe 2 and stripe 3 enhancers and the Gal 4 cognate sites (U AS) are located adjacent to stripe 2 enhancer. The other was the even skipped stripe 3 and rhomboid enhancers with the Gal 4 recognition sites located within the rhomboid enhancer. In both cases, the proximal enhancers, stripe 2 in the first reporter and rhomboid in the second reporter, were repressed in the region where knirps was present. These results indicate that the Gal4- Knirps chimeric proteins are causing the same efl‘ect as the endogenous protein and Gal4- Knirps is a good model for study the repression mechanisms of Knirps protein in Drosophila embryos (Figure 4) (Mao et al., manuscript in preparation). 20 Figure 4. Gal4-Knirps chimeric repressors mimic the activity of endogenous Knirps portein on a rhomboid enhancer element. (A) Expression pattern of the eve stripe 3/rho lacZ reporter gene showing robust ventral expression directed fiom a rho enhancer element lacking endogenous Snail binding sites (Gray et al., 1996), and a central stripe fi'om the eve stripe 3 enhancer. (B) Repression in ventral regions mediated by Gal4-Knirps 75-429 repressor protein expressed in ventral regions of the embryo under control of a twist promoter construct. (C) Repression mediated by the Gal4-Knirps 75-332 chimera, lacking the dCtBP interaction motif. (D) Repression mediated by the Gal4-Knirps 202—358 chimera, which contains the dCtBP interaction motif. Gal4 binding sites (marked UAS in the figure were introduced in a 600 bp rho enhancer at the positions previously used for targeting Knirps protein to this gene complex (Amosti et al., 1996). Ventrolateral views are shown, with anterior to the left (From Mao et al., manuscript in preparation). knl 202-358 mm rm: Iacz UAS UAS 22 Mutational analysis of the GaM-Knirps chimera. A series of tnmcation mutations were introduced in the Gal4-Knirps gene to identify residues critical for repression. Genes encoding residues 75-332, 75-254, 75- 189, 75-189 + 254-429, 188-429, or 189-254 fused to the Gal4 DNA binding domain were expressed in ventral regions of transgenic embryos under control of a twist promoter element. These chimeric repressors were tested on lacZ reporter genes activated by eve stripe enhancers (Figure 5A) (Mao et al., manuscript in preparation). A full-length 75- 429 chimera showed effective repression of the proximal stripe 2 enhancer, while, consistent with the short-range of Knirps activity, the distal stripe 3 enhancer was not affected (Figure SB) (Mao et al., manuscript in preparation). A C-terminal truncation removing residues 333-429 did not compromise activity (Figure 5C) (Mao et al., manuscript in preparation), although it does remove most of the dCtBP interaction region of Knirps (see below). More extensive C-terminal truncations starting at residues 254 or 189 were inactive, suggesting that a region of the protein from 254-332 might be necessary for repression (Figure 5D—E). The N terminal region 75-187 was dispensable for activity, as was the region of the protein from 189-254 (Figure 5F,G). The chimera containing only residues 189-254 was not active (Figure 5H), unlike the result obtained for a similar construct in transient transfection assays (Gerwin et al., 1994). Constructs were also tested on the eve stripe 3/ rho reporter gene as in Figure 3 and a stripe 2 lacZ gene (Amosti et al., 1996), with identical results (data not shown). All constructs scored as active showed at least 20% of the embryos were repressed in blinded scoring assays, 23 Figure 5. dCtBP interaction region of Knirps is dispensible for repression activity in C-terminal truncated proteins. (A) Expression pattern of the eve stripe 2/stripe 3 reporter gene, with UAS binding sites shown adjacent to the stripe 2 element. (B) Expression of the Gal4-Knirps 75-429 gene represses stripe 2, but not stripe 3, in ventral regions. (C) Gal4-Knirps 75- 332 chimera, lacking an intact dCtBP binding motif (gray box), is a functional repressor. (D, E) Truncation at codon 254 or 189 leads to loss of activity. (F) Deletion of residues 7 5-1 87 does not compromise repression activity. (G) Deletion of residues 189-254 does not compromise repression activity. Genes shown here were produced by introducing stop codons into the gene at appropriate sites, thus the overall transcript size and 3 ’UTR is conserved . Embryos are shown with anterior to left, dorsal side up (Mao et al., manuscript in preparation). 24 25 while inactive repressor genes and the reporter construct alone had less than 2% of the embryos scored as repressed (Mao et al., manuscript in preparation). One of the Gal4-Knirps chimeric proteins, Gal4-Knirps 75-330, which contains the residues 75-330 of Knirps, and Gal4-Knirps 75-332, which contains the residues 75- 332, do not have the dCtBP-binding motif but do show in vivo repression activity (Dr. Scott Keller unpublished results). In addition, Gal4-Knirps 75-332 protein does not interact with dCtBP in GST pull down assays (Dr. Carla Margulies unpublished results). Thus, binding to dCtBP in vitro depends on an intact interaction motif. However, this interaction is not a requirement for the activity of the 75-3 32 protein. These results suggest that dCtBP-binding is dispensable for the repression fimction of the N-terminal portion of Knirps protein, between residues 75 and 332. 26 CHAPTER 3 THE MINIMAL REPRESSION DOMAIN OF KNIRPS 27 Introduction Like transcriptional activators (Mitchell and Tjian, 1989), many transcriptional repressors are modular factors since the effector domain can be separated from the DNA binding dormin. Repression domains in rmny repressors can be classified according to the amino acid sequence, such as charged, alanine-rich and proline-rich domains (Gray et al., 1996). Previous work has demonstrated that a region between residue 75 and 429 is suflicient to confer transcriptional repression when separated from the DNA binding domain (Amosti et al., 1996). To further identify the repression domains within this region, various recombinant knirps genes were constructed by replacing the DNA binding domain codons 1-74 of Knirps with a DNA binding domain to identify residues and regions critical for repression. Materials and methods Plasmids Minimal repression domains: Fragments of the knirps open reading fi'ame were generated by PCR from the pCarnegie 20 vector pN741 containing the knirps cDNA (G. Struhl, unpublished) with Pfu DNA polymerase (Stratagene), using different pairs of primers. The products were subcloned into the KpnI and XbaI sites of the pTwiggy vector (Arnosti et al., 1996) containing twi enhancer element 2xPEe-Et and twist basal promoter. The constructs and the corresponding primers used were: 28 202-358-fl_ag: oligo 1, 5’-CGGGGTACCGCTGCCGCTGCAGCGGCTTCTGCTGCCGATGCCGCT-3’; 202-358-flag: oligo 2, 5’-GGGGAATCTAGACTAACTAATTACTACTTGTCATCGTCATCCTTGTAATC CACCTCCACTTCTTGATCCTCGGA-3’; 211-358-flag: oligo 1, 5’- CGGGGTACCGATGCCGCTTACCGGCAGGAGATGTACAAGCACCGC-3’; 211858-ng oligo 2, 5’-GGGGAATCTAGACTAACTAATTACTACTTGTCATCGTCATCCTTGTAATC CACCTCCACTTCTTGATCCTCGGA-3’; The amplified fiagments were digested with ClaI and XbaI and used to replace the C101- XbaI fi'agment of the knirps cDNA clone in pBluescript(SK+). The KpnI-Xbal fi‘agments of these clones were then inserted in pTwiggy (Amosti et al., 1996). The final constructs encode Knirps amino acids followed by an eight amino-acid sequence including the FLAG epitope, DYKDDDDK. 248-358-flag: oligo 1, 5’-GGG TCGGTACCGCAGCCTCGGCCCGCCAGTCGCCCATCGAT-3’ 248-358-flag: oligo 2, 5’-GGGGAATCTAGACTAACTAATTACTACTTGTCATCGTCATCCTTGTAATC CACCTCCACTTCTTGATCCTCGGA-3’ 202-429-flag: oligo 1, 5’-CGGGGTACCGCTGCCGCTGCAGCGGCTTCTGCTGCCGATGCCGCT-3’ 202-429-flag: oligo 2, 29 GTAATCGACACACACGAATATTCCCCTCAT-3 ’ 211-429-flag: oligo 1, 5’- CGGGGTACCGATGCCGCTTACCGGCAGGAGATGTACAAGCACCGC-3’ 211-429-flag: oligo 2, 5’-GGGGAATCTAGACTAACTAATTACTACTTGTCATCGTCATCGTCATCCTT GTAATCGACACACACGAATATTCCCCTCAT-3’ P-element transformation, whole-mount in situ hybridization of embryos, and crosses to reporter lines. P-element transformation vectors were introduced into the Drosophila germline by injection of y W67 embryos as described (Small et al., 1992) (Figure 6). For each gene construct, at least three separate lines were tested, and similar results were obtained in each case. In situ hybridizations were performed as described (Srmll et al., 1992) using digoxigenin-UTP labeled antisense RNA probes to IacZ. The Gal4-dCtBP chimeric lines used in Figure 11 were kindly supplied by Dr. Yutaka Nibu and Dr. Michael Levine. Assays of in vivo repression activity. Transformant ofi'springs were crossed with reporter lines containing one of three reporters: I) even-skipped stripe 2 linked to eve- lacZ (Amosti et al., 1996), 2) even-skipped stripe 2 and stripe 3 enhancers linked to eve- lacZ (N ibu et al., 1998b). The stripe 2 and stripe 3 enhancers of even-skipped were placed 1.6 kb apart from each other, and two copies of the UAS sequence were introduced adjacent to the proximal enhancer, the stripe 2 enhancer. or 3) eve stripe 3 and rhomboid enhancers linked to the transposase-lacZ fUSiOIl gene (Hewitt et al., 1999). 30 Figure 6. The making of a transgenic fly. 31 <7:— cmfimug 95: Ease. 2. 3 3.3.5 n1 .U 2.: a 3.53 omcuwmflab 8E3 2.: 22.2 d .5228 «sum-5.5 3.33 95982. 55.59: .EEmaE 32.5.:— omameamgh O m omega—man: k 3.1.... e: 28:56 3 .z a 3:83. 32 Embryos generated by the crosses were fixed and stained by in situ hybridizations. To quantitate repression activity of Gal4-Knirps chimeras, heterozygous effector lines were crossed to homozygous reporter lines, and several hundred embryos aged to 2-4 hours fiom each cross were collected at room temperature, fixed and stained as described in Small et al., 1992. After mounting on microscope slides, embryos were visually scored in a blinded experiment for evidence of repression. Most fimctional repressors completely abolished ventral staining in the eve stripe 2 region; embryos exhibiting weakened but not complete repression were scored in a separate category. Typically, a larger proportion of older embryos exhibited repression, presmnably because of the lag between the activation of the eve or rho enhancer and the production of adequate amounts of the Gal4-Knirps protein after its gene’s transcription under the control of the twist enhancer. In this assay, the maximum percentage of embryos exhibiting repression in the assay would be 50%, because only halfof the fertilized embryos receive the Gal4-Knirps effector gene from the heterozygous male parent. The actual percentages are lower than this because all embryos showing reporter gene expression were counted, including younger embryos in which repressors had not yet reached appreciable levels. Results: 202-358 is a minimal repression domain of Knirps protein Genes encoding portions of the knirps open reading flame and the Gal4DNA binding domain were expressed in ventral regions of transgenic embryos under control of a twist promoter element (Figure 7). These chimeric repressors were tested on a IacZ 33 Figure 7. The injection vector pTwiggy contains a twist enhancer element, which drives the expression of Gal4-Knirps proteins in the ventral region of the embryo, and a white gene, which serves as a marker of transgenics and the inverted repeats of the P element, which can mediate the transposition of the injection construct into the germline of Drosophila embryos. 34 y: 5:5 0 38:83 u _ 6a-: :5 32 52555 «$5 3.53 m: .L! 25w 25:: ESE: «e 55.52: 35 reporter gene activated by eve stripe 2 and stripe 3 enhancers, with Gal4 binding UAS sequences adjacent to the stripe 2 element. An alanine-rich sequence is located within the region from residue 202 to 211. Constructs Gal4-Knirps 202-429, Gal4-Knirps 211—429, Gal4-Knirps 202-358, Gal4-Knirps 211-358 were designed to test the whether the alanine rich region is related to repression function of Knirps protein. A PvuII restriction site, located at codon 189 within the cDNA sequence of knirps, was used to generate the construct Gal4-knirps 189-358. Gal4-Knirps 189-358, Gal4-Knirps 202-429, Gal4-Knirps 211-429, Gal4-Knirps 202-358, Gal4-Knirps 211-358 showed effective repression of the proximal stripe 2 enhancer, while, consistent with the short-range of Knirps activity, the distal stripe 3 enhancer was not affected. (Representative examples of repressed and non- repressed embryos from in viva repression assays are shown in Figure 8). In construct Gal4-knirps 248-358, the similar repression was observed in a smaller proportion of embryos. Tint is to say, the ratio of the numbers of the repressed embryos to that of the non-repressed embryo is six times lower than knirps 189-358, knirps 202-358 etc., indicating that amino acids 248-358 may encode a weak repressor (Figure 9). To further identify residues important for repression, Drs Keller and Foley tested Gal4 fusion proteins containing residues 202-358 and derivatives in transgenic assays. Their results have shown that deletions of residues 202-210 (202-358 AA), (alanines previously identified in a minimal repression dormin in cell culture assays), 220-227 (202-358 AB), 228-251 (202-358 AC), or 292-313 (202-358 AD) did not impair repression activity, while deletions of residues 330-343 (202-358 AE) abolished repression completely (Figure 10). The region afl‘ected in 202-358 AE includes the residues P-DLS-K, recently shown to mediate Knirps interaction with the dCtBP protein 36 Figure 8. Representative embryos show the repression of eve stripe 2 by Gal4-Knirps. Transgenic embryos show the absence of stripe 2 expression in the ventral sides (A and B), while embryos lacking the Gal4-knirps transgene show equally well-expressed stripe 2 an stripe 3 (C and D). 37 38855 2: 9m 25 «av 8%.an 3.5.30 98330.. a «.9 «areas omuowmuauh 33> 3.5:?» 33> .933..— 38 Figure 9. The schematic structures and the in vivo repression activities of the Gal4-Knirps mutant proteins. Three transgenic lines of each construct were assayed except that only one line of 189-358 and one line of 248-358 were generated and assayed. 39 Activities of the C-terminal Knirps mutants amino aci . 75 189 254 331 358 29 :I 3:. Knirps In vivo Zn kni alanine dCtBP repression fingers box rich binding activity 75429 I H I l + 189 ; 189-358- E: It-fi-I + Flag flag tag 248 358 248-358--\/m:1:l weak Flag . ,l w 429 202-429- -\/i a + Flag 429 211-429- E I + \f Flag 202-358 + 21 211-358 -\/J:Ejl + 202 358 202.353- n-a-l + Flag E El [I E] I Gal4 Zinc Alanine- dCtBP Flag Kni (1'93) finger rich binding tag box 40 Ratio of repressed 31% 20% 4% 30% 24% 28% 29% 32% Figure 10. A central region of Knirps requires the dCtBP binding motif for activity. Genes were generated by PCR amplification of regions of the knirps gene, followed by site-directed mutagenesis (AA-AE) to remove internal conserved residues (Gerwin et al., 1994) and non- conserved residues , AC. The dCtBP binding motif is removed in AB. The activities of the genes are indicated to the right of the construct. 41 _ _ amt—:Vflo Tana: 4.. . : Ms... TIL jawaflamu i. I gear . a; m4 + H _ MLSMUUr ; gal; a: + a E Raga—:3 a + _ E . . u. a. + — a , i. L... a. a _ a 42 (Nibu et al., 1998a, 1998b). The lack of repression activity of protein 202-358 AE could be caused by the lack of important repression residues, but also could be due to the lack of protein expression in embryos. To address this question, gel mobility shifts experiments have been carried out and the mobility shift complexes formed by the mutant protein 202-358 AE and the Gal4 binding site-containing probe were detected in the polyacrylamide gel (see Chapter 6), indicating that the non-functional protein is expressed. Alanine scanning mutations affecting these residues have been shown to compromise the repression activity of a chimeric Gal4-Knirps protein containing residues 255-429 (Nibu et al., 1998a) and by ectopically expressed Knirps (Nibu et al., 1998b). Our results indicate that the minimal 202-358 repression construct is also dependent on dCtBP for activity. A minimal construct containing only residues 248-291 and 314-358 was also active, consistent with earlier reports that residues N terminal to 255 were dispensable for activity (N ibu et al., 1998a). Our result was consistent with Nibu et al., who found that the dCtBP-binding is important for the repression function of knirps. On the other hand, two other constructs, Gal4-knirps 75-332, Gal4-knirps 75-330 lack the dCtBP-binding motif and display robust repression activity in viva. Given the fact that Knirps 75-332 do not bind to dCtBP in vitro in the glutathione S-transferase pull down assay (Dr. Carla Margulies unpublished results), our results indicate that the dCtBP- binding motif is dispensible in the N-terminal portion between residues 75 and 330. Activity of the dCtBP repressor protein from a distal enhancer position. Our results suggest that for the N-terminal repressor, residues 75-3 30, binding of the dCtBP repressor protein to Knirps is not required for repression. We considered the 43 possibility that dCtBP may merely affect an allosteric alteration of the Knirps protein, allowing the Knirps protein’s own repression domain to contact a target in the transcription machinery. In this case, dCtBP would not directly mediate transcriptional repression. Previous experiments have established that tethering dCtBP or a murine homologue, CtBP2, to a promoter inhibited gene expression, suggesting that the cofactor is itself capable of repression (Turner and Crossley, 1998; Nibu et al., 1998). However, these assays used promoter-proxirnal Gal4 binding sites that may be subject to steric hindrance effects. Therefore, we tested whether the Gal4-dCtBP protein was active from a distal enhancer site within the rho enhancer (Figure 11). Expression of a Gal4-dCtBP chimera under the control of a Krfippel promoter inhibited activity of the rha enhancer, but not the eve stripe 3 enhancer that is far away from the Gal 4 binding sites, demonstrating that this cofactor is capable of independent action as a repressor when targeted to a specific enhancer. Thus, the central 202-358 portion of Knirps is likely to act through dCtBP, and the N terminal portion of Knirps that does not bind to dCtBP may repress via an alternative pathway. Figure 11. The Ga14-dCtBP chimeric protein acts as an independent repressor within an enhancer. (A) Expression pattern of the eve stripe /rha lacZ reporter gene, showing robust ventral expression directed from a rha enhancer element lacking Snail binding sites (1), and a central stripe from the eve stripe 3 enhancer. (B) Repression of rho enhancer activity in the central portion of the embryo mediated by the Gal4- dCtBP chimera. The repressor is driven by a Kriippel promoter (25). Gal4 binding sites were introduced in a 600 bp rho enhancer at the positions previously used for targeting Knirps protein to this gene complex (1). Ventrolateral views are shown, with anterior to the left. 45 w‘hitel Istripe3| Ii rho il IIacZ UAS UAS 46 CHAPTER 4 THE ANALYSIS OF THE N-TERMINAL REPRESSOR OF WRPS 47 Introduction Many repressor proteins have multiple repression donnins. Engrailed, Hairy and Krfippel are examples (Tolkunova et al., 1998, Poortinga et al., 1998, Light et al., 1994). Our data suggest that Knirps protein also possesses more than one repression donmin. As previous work has pointed out, the residues 331-337 PMDLSMK is the dCtBP-binding motif and disruption of this motif will abolish dCtBP-Knirps interaction. The mutant proteins, which have this particular motif PMDLSMK, are potent repressors in embryos, for example, the Ga14-Knirps 202-358. Meanwhile the mutant protein Gal4-Knirps 75- 330 is a functional repressor although it does not contain the dCtBP-binding motif, which indicates that there is another repression domain located in the N-terminal region between residue 75 to 330 and this N-terrninal region confers transcriptional repression independent of recruiting dCtBP protein. To reveal the possible mechanisms involved and to define the important amino acids or motifs involved in repression for the N- terminal repressor, the polymerase chain reactions with corresponding primers were used to generate a series of N-terminal truncation Gal4-Knirps mutants (See Materials and Methods). These mutants contain the cDNA sequence of knirps, starting at various positions and ending at codon 330. They are constructs 94-330, 124-330, 139-330, 169- 330, 189-330, 200-330 and 75-304. Materials and Methods: 48 N-terminal deletions: Fragment of the knirps open reading were generated from the pCarnegie 20 vector pN741 containing the knirps cDNA (G. Struhl, unpublished) with Pfil] DNA polymerase (Stratagene) with polymerase chain reactions, using different pairs of primers. The products were subcloned into the KpnI and XbaI sites of the pTwiggy vector containing twist enhancer element 2xpEe-Et and twist basal promoter. The constructs and the corresponding primers were: 94-330: primer 1, 5’-GGGTCGGTACCCACGAACAGGCCGCCGCAGCGGCGGGCAAG-3’ 94-330: primer 2, 5’-GGGGAATCTAGACTAACTAATTACTACTTGTCATCGTCATCCTTGTAATC TCCTTCTTGAGCGGAAACGGTGGG-3’ 124-330: primer 1, 5’-GGGTCGGTACCGCCGCAGCGGGCTCGCCACACACTCCCGGATITGGG—3’ 124-330: primer 2, 5’-GGGGAATCTAGACTAACTAAITACTACTTGTCATCGTCATCCTTGTAATC TCCTTCTTGAGCGGAAACGGTGGG-3’ 136-330: primer 1, 5’-GGGTCGGTACCCACCACCATCATCAGCAGCAGCAGCAC-3’ 136-330: primer 2, 5’-GGGGAATCTAGACTAACTAATTACTACTTGTCATCGTCATCCTTGTAATC TCCTTCTTGAGCGGAAACGGTGGG-3’ 169-330: primer 1, 49 5’-GGGTCGGTACCGCCGCAGCGTCCGCCGCCCTGCCCTTCTTCAGC-3’ 169-330: primer 2, 5’-GGGGAATCTAGACTAACTAATTACTACTTGTCATCGTCATCCTTGTAATC TCCTTCTTGAGCGGAAACGGTGGG-3’ 189-330: primer 1, 5 ’-GGGTCGGTACCCTGCCCCCACACCTCCTCTTCCCAGGCTAC-3 ’ 189-330: primer 2, 5’-GGGGAATCTAGACTAACTAATTACTACTTGTCATCGTCATCCTTGTAATC TCCTTCTTGAGCGGAAACGGTGGG-3’ 200-300: primer 1, 5 ’-GGGTCGGTACCGCAAGTGCTGCCGCTGCAGCGGCTTCTGCT-3 ’ 200-300: primer 2, 5’-GGGGAATCTAGACTAACTAATTACTACTTGTCATCGTCATCCTTGTAATC TCCTTCTTGAGCGGAAACGGTGGG-3’ 75-304: primer 1, 5’- GGGTCGGTACCGCAGCCTCGGCCCGCCAGTCGCCCATCGAT-3’ 75-304: primer 2, 5’-GGGGAATCTAGACTAACTAATTACTACTTGTCATCGTCATCCTTGTAATC GGGCGACAAGCTCTGCATCTTGGC-3’ 75-304 plasmid was constructed by polymerase chain reaction and subcloned into the Clal and Xbal site of 75-429 in Twiggy plasmid. 50 DNA sequencing reactions confirmed all the constructs except 75-304 and 94- 330. Both construct 75-304 and construct 94-330 contain a point mutation from base T to base C within residue 107. P-element transformation, whole-mount in situ hybridization of embryos, and crosses to reporter lines. The methods are the same as described in Chapter 3 Assays of in viva repression activity. The methods are the same as described in Chapter 3. Resu Its: The mutant knirps genes 94-330, 124-330, 139-330, 169-330 189-330, 200-330 and 75-304 were constructed as described in the materials and methods. 124-330, 169- 330 and 200-330 were constructed but to be transformed into embryos. Construct 75-304 was constructed and transformed into embryos but to be assayed for repression activity. 94-330, 139-330 and 189-330 were transformed into yw67 fly embryos. 19, 4 and 10 transgenic lines were generated from construct 94-330, 139—330 and 189-330 respectively. The 94-330 transgenic lines were divided into three pools by mixing the transgenic males from individual lines with each approximately 50 flies per pool. 4 lines of 139-330 were mixed into one pool, and 10 lines of 189-330 were mixed into one pool to obtain sufficient number of male flies. The reporter lines contain the stripe 2 and stripe 3 enhancer and two copies of UAS sequences (the Gal4-Knirps binding sites) adjacent to the stripe 2 enhancer. Crossing the males from the transgenic lines of 94-3 30, 139—330 51 and 189-330 respectively with the virgin females fi'om the reporter lines was performed. Embryos fiom these crosses were stained and looked for the expression pattern of the lac Z gene under the control of stripe 2 and stripe 3 enhancers (See Materials and Methods). Gal4 Knirps 94-330 and Gal4-Knirps 139-330 repress the proximal enhancer stripe 2 but not the distal stripe 3, while Gal4-Knirps 189-330 repress neither stripe 2 nor the stripe 3 (Fig 12). These results suggest that amino acids 139 to 189 contain a motif necessary for repression. Discussion and conclusion The repressed embryo ratio in 94-330 (12%-18%) is higher than that in 139-330 (5%-7%). This phenomenon can be explained by either 1) The repression activity of the N-terminal repressor is additive, i. e. An auxiliary element located between residue 94 and 139 is facilitating the repression in addition to residues 139-189. Or 2) Position efl'ect altered observed repression activities in P element mediated transformations. When a transposon is translocated into the locus of a heterochromatin region, the transgene will be silenced. Therefore, even a transgene capable of encoding a potent repressor may not express repression activities in some transgenic lines. The repression assays presented in this chapter are done by mixing several transgenic lines together due to the insufficient number of male flies present in each individual line. The embryos scored for repression in this study are derived fiom the pool of lines with different sites of integration. Thus, the observed repression activities monitored by the average repressed embryo ratios are expected to be small if a particular transgene is particularly sensitive to position effects. 52 Figure 12. The schematic structures of the amino-terminal Knirps mutants. The Knirps DNA binding domain is indicated in green, while the Gal4 DNA binding domain is indicated in red. Gal4 fiision proteins containing residues 75-330, 94-330, 139-330 showed effective in vivo repression activities, whereas the Gal4-fusion protein 189-330 does not possesses repression activity. 53 75 330 429 Knirps E a law I Knirps DBD dCtBP-binding motif Repressed Repression embryos activity 75 Kni box 330 75-33' ‘_ 10%-16% '1' Gal 4 DBD 94 330 Pm“! 139-330 5%'7% + 189-330 Lg | 0% - 54 The lack of region from amino acid 139-189 inactivates the protein 189-330. This region contains a LGYPSY box that is conserved in Knirps derived from several different insect species: Drosophila melanogaster, Drosophila virilis and Drasphila damestica (Figure 13). In addition, the region between amino acid 139 and 149 contains multiple charged residues of histidine and glutamine and is also well conserved (Figure 13). Many transcriptional regulator proteins function through protein-protein interactions, for example activator VP 1 6 interacts with TAF (the TBP associated factors) and Drosophila eve protein interact with TBP. Regulations of transcription appear to be the result of multiple interactions between the regulator proteins and the targets. It was previously reported a glutamine-rich region constitute the activation domain of VPl6 and Spl . Therefore, it is likely that the glutamine-rich region present in Knirps is also important for repression activity. These predictions are consistent with the observations that the multiple glutamine-rich regions spanning protein 94-330 and 139-330 contribute to repression while the region 189-330 does not (Figure 12). The LGYPSY box contained in the functional proteins 94-330 and 139-330 but not in the non-functional 189-330 constitutes the repression domain. Praline has the tendency of introducing turns into the protein secondary structure. It is likely that proline residue in the LGYPSY box contributes to the repression function of Knirps by helping the protein maintain proper conformation. Although the Kni box is conserved in all the three species: Drosophila melanogaster, Drosophila virilis and Drasphila damestica, the lack of the Kni box in the 55 Figure 13. Sequence comparison of Knirps-homologous proteins from diflerent species: Drosophila melanogaster, Drosophila virilis and Drosophila damestica. Conservation of amino acids between the different proteins are indicated by a grey background. Positions 75-93 are located as the absolutely conserved Kni box. Other conserved protein motifs are marked by boxes as well, such as LGYPSY box. The first N-terminal Knirps residues of recombinant proteins 94-330, 139-330 and 189-339 are indicated by the arrows. (From Gerwin et al., 1994) 56 EZHIIA— HIM .85.‘ .5 335.9 “a 2.3.8— §.HOI.Q HIM .5.‘ «a .NflbAu .8 .HII.Q §.HII.D HIE .86.: .5 335.0 .5 .HII.O 57 protein 139-330 does not impair the repression activity, it suggests that this motif is not necessary for repression. The conclusions and discussions above were based on the presumption that all the three recombinant proteins, 94-330, 139-330 and 189-330 are expressed in embryos. 189- 330 is the only one that does not display activity in vivo, thus, further experiments need to be done to detect the protein expression level of the construct 189-330 in embryos as describes in Chapter 6. The results indicate that the region between 139-189 containing the LGYPSY and poly histidines and poly glutamines is important for the function of the amino terminal repressor and Knirps 139-330 contains the minimal N-terminal repression domain. 58 CHAPTER 5 DNA BINDING DOMAIN, THE INHIBITORY ELEMENT IN KNIRPS REPRESSION 59 Introduction A full-length wild type Knirps protein was misexpressed in the same location as the even-skipped stripe 2 by experimental manipulations in previous experiments. Knirps diffused into the stripe 3 region and binds to its cognate sites within the stripe 3 enhancer, eliminating the expression of even-skipped gene in the area of stripe 3. Meanwhile it has been observed that alanine substitution mutations in the dCtBP-binding motif, (PIVfl)L were mutated into AAAA) caused the derepression of eve stripe 3 (N ibu et al., EMBO, 1999) (Figure 2). The conclusions fi'om this experiment were that dCtBP-binding motif is required for full-length knirps to repress even-skipped stripe 3. However, it was observed that the Gal4-Knirps 75-330, which contains a heterologous DNA binding domain and lacks the C-terminal tail residues 331-429, including the dCtBP binding motif, repressed transcription when assayed on the lacZ reporter controlled by eve stripe 2 and stripe 3 enhancers. The differences between the repression caused by the full-length Knirps and that caused by the Gal4-Knirps 75-330 are as follows: 1) the presence or absence of the C-terrninal stretch of amino acids, 2) the DNA binding domain, 3) the enhancers which drive the ectopic Knirps expression (eve stripe 2 versus twist), 4) the target genes (stripe 3 of endogenous eve enhancer versus a stripe 2 and stripe 3 LacZ chimera enhancers). Based on the difference 1) and 2), we can infer that either the zinc finger DNA binding domain of Knirps or the C-terminal tail abolishes the protein fucntion. Thus, a reasonable hypothesis is that there maybe an inhibitory element present in either the zinc finger DNA binding domain of Knirps or in the C-terrninal tail. I chose to test the C-terrninal inhibition hypothesis first. Therefore, 60 the following constructs were designed. Knirps cDNAs starting from 75 and ending with different residues in the C-terminal region between residue 330 and 429 with a mutated dCtBP-binding motif (N ibu et al., 1998a) fused to Gal4 DNA binding domain were constructed and tested for in viva repression activities. These constructs are: Gal4 knirps 75-429*, Gal4-knirps 75-394* and Gal4-knirps 75-364* (* indicate the mutations in the dCtBP-binding motif) (See materials and methods). Materials and Methods: C-terminal deletions: Fragments of the knirps open reading were generated from the plasmid Casper 22FAKE (N ibu et al., 1998a ) containing the alanine substitution of the dCtBP-binding motif in the knirps cDNA with pfuI DNA polymerase (Stratagene) with polymerase chain reactions, using different pairs of primers. The products were subcloned into the Clal and Xbal sites of the 75-429 vector containing twist enhancer element 2xPEe-Et and twist basal promoter. The constructs and the corresponding primers used were: 75-429*: oligol, 5’-GGGTCGGTACCGCAGCCTCGGCCCGCCAGTCGCCCATCGAT-3’ 75-429“: olig02, 5’-GGGGAATCTAGACTAACTAATTACTACTTGTCATCGTCATCCTTGTAATC GAAC'ITCCGGCGCGGAGCCACCTC-3’ 75-394": oligol, 5’-GGG TCGGTACCGCAGCCTCGGCCCGCCAGTCGCCCATCGAT-3’ 75-394": olig02, 61 5 ’-GGGGAATCTAGACTAACTAATTACTACTTGTCATCGTCATCCTTGTAATC GAACTTCCGGCGCGGAGCCACCTC- 3’ 75-364“: oligol, 5’-GGG TCGGTACCGCAGCCTCGGCCCGCCAGTCGCCCATCGAT-3’ 75-364*: olig02, 5’-GGGGAATCTAGACTAACTAA’ITACTACTTGTCATCGTCATCCTTGTAATC GAACTTCCGGCGCGGAGCCACCTC-3’ P element transformation, whole-mount in situ hybridization of embryos, and crosses to reporter lines. The methods are the same as described in Chapter 3. Assays of in viva repression activity. The methods are the same as described in chapter 3. Resu Its: Three constructs encode the Gal4-Knirps proteins with the mutated dCtBP- binding motif, Gal4-knirps 75-429", Gal4-knirps 75-394*and Gal4-knirps 75-364“ were transformed into fly embryos and assayed on the reporter lacZ gene under the control of the stripe 2, 3 enhancers for repression activities. Three transgenic lines of Gal4 knirps 75-429* were tested and an average of 21-26% embryos showed robust repression of the eve stripe 2 enhancer in the region where Gal4-Knirps 75-429* is expressed, indicating that Gal4-Knirps75-429“ is a functional repressor (Figure 14). The difference in activity 62 between the functional repressor, Gal4 Knirps 75-429*, with the non-functional protein, full length mutant Knirps protein (Nibu et al., EMBO 1998, also see the beginning of Chapter 5) could rely on the different DNA binding domains. Our result indicates that replacement of the Knirps DNA binding domain with the Gal4 DNA binding domain rescued the function of the mutated 75-429 portion (dCtBP-binding motif is mutated). Similarly, one of two Gal4-Knirps 75-394* lines assayed and two of three Gal4—Knirps 7 5364* lines assayed also showed repression activity and elimination of the staining of the even-skipped stripe 2 in the ventral portion of the embryos were observed (Figure 14). However, one line of each construct failed to show repression. These variations in activity between lines of the same constructs may reflect position effects, i.e. the chromosomal position of the integrated P element causes variation in the level of expression of the transgenes. Discussion and Conclusion: Mutagenizing of the dCtBP-binding motif in the Gal4 Knirps 75-429“, Gal4-Knirps 75-394‘ and Gal4-Knirps 75-364* did not affect the repression activity, suggesting that the dCtBP-binding is not a requirement for repression by these proteins. This result not only confirms the result that protein encoded by construct Knirps 75-330 in which the dCtBP-binding motif is completely deleted showed effective repression (Mao et al., Inanuscript in preparation), but also provides a hint as to the role of Knirps DNA binding domain on Knirps repression function. Mutation in the dCtBP-binding motif abolished the repression activity of Knirps protein with the Knirps DNA binding domain, but did 63 Figure 14. The schematic representation of the carboxy terminal mutants of Knirps protein and their repression activities. Blue boxes represent the Knirps DNA binding dornainand the red boxes represent the Gal4 DNA binding domain. With the same alanine mutations, Gal4 DNA fusion proteins containing residues 75-429, 75-394 and 75-364 show effective repression whereas when linked to the Knirps DNA binding donnin, residues 75—429 showed no repression activity. Three transgenic lines of 75-429, two transgenic lines of 75-394 and two transgenic lines of 75-364 showed repression. The embryos fi'om lines that showed repression were scored and average ratios of the repressed embryos for each constructs were indicated. + ease + §o-§m gm + £08-05 fl N + gem fl -3 - sec 368 worsen a? commmoaom commoaou we 23m Em mw 65 not afiect activity of the knirps protein with a Gal4 binding domain. In addition Gal4- dCtBP chimera shows good repression. These observations help to clarify the repression mechanisms of Knirps protein. Regarding what are the roles of the knirps DNA binding domain and binding to dCtBP in repression activity of knirps, I propose three models to explain the observations above: 1) dCtBP promotes dirnerization of the Knirps protein with the Knirps binding domain to exert efl‘ective repression. As a member of the nuclear hormone receptor superfarnily, Knirps is very likely to repress transcription as a dimer. As a matter of fact, in our transgenic assays, the Knirps binding sites consists of 16 nucleotides with inverted repeats (Amosti et al., 1996). However, there is also evidence that Knirps can form a monomer as well as a dimer in gel retardation assays (Gerwin et al., 1994). In the embryo, whether Knirps exert repression as a monomer or as a dimer is not clear. dCtBP is known to share significant homology with the D-isomer 2— hydroxy acid dehydrogenases (Schaeper et al., 1995). Members of this particular family of dehydrogenase have been shown to form homodirners. It is possible that the homology between dCtBP and the dehydrogenase family constitutes the preservation of structural rather than enzymatic features and dCtBP-Knirps interaction facilitates Knirps dirnerization and leads to efl‘ective repression (Figure 15). 2) The Knirps DNA binding domain prevents the Knirps repression domain from fimctioning. Mutations in the dCtBP-binding motif disrupt the repression fimction of the full-length Knirps protein, but not the Gal4-Knirps protein, It is likely that the Knirps DNA binding domain acts as an inhibitory element in addition to its role of DNA binding. The recruitment of dCtBP relieves this inhibition possibly due to allosteric changes and restores activity of the repression domain (Figure 16). 3) The repression domain with the mutated dCtBP-binding motif interferes with DNA binding by the Knirps DNA binding domain. Binding to dCtBP causes conforrmtional changes that disrupt this interference (Figure 17). 67 Figure 15. The schematic representation of the model I. dCtBP helps Knirps dirnerization to perform effective repression. The green circles represent the Knirps DNA binding domain. The black boxes represent the repression domain(s) of Knirps protein. “AAA” represents the alanine substitution mutation within the repression domain. 68 29 23 tL commmofiom o>uootm man; he .eefee nommmoaom on Ho #33 . e 0 ee 0 a he 69 Figure 16. Schematic representation of model II. The DNA binding domain interferes with the function of the repression domain. “AAA” represents the alanine substitution mutation in the dCtBP binding motif. (A) The mutation in the dCtBP binding motif inactivates Knirps through the inhibitory element within the Knirps DNA binding domain (in blue). (B) Wild-type Knirps represses transcription through recruiting dCtBP (in yellow). (C) Mutations in the dCtBP ginding motif do not affect the repression activity of the Gal- Knirps function. Eanw Ems—es 523