Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A fuctional [sic] analysis of Cis-elements contributing to developmental regulation of gene expression… Neustaedter, David A. 1997

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

Item Metadata

Download

Media
831-ubc_1997-251284.pdf [ 10.19MB ]
Metadata
JSON: 831-1.0088336.json
JSON-LD: 831-1.0088336-ld.json
RDF/XML (Pretty): 831-1.0088336-rdf.xml
RDF/JSON: 831-1.0088336-rdf.json
Turtle: 831-1.0088336-turtle.txt
N-Triples: 831-1.0088336-rdf-ntriples.txt
Original Record: 831-1.0088336-source.json
Full Text
831-1.0088336-fulltext.txt
Citation
831-1.0088336.ris

Full Text

A Fuctional Analysis of C/s-Elements Contributing to Developmental Regulation of Gene Expression by the 4CL1 Promoter in Transgenic Tobacco Plants. By David A. Neustaedter B.Sc.(Honours) The University of Calgary, 1989. A Thesis Submitted in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy in The Faculty of Graduate Studies Department of Botany. W e accept this thesis as conforming to the required standard. The University of Brit ish Co lumbia Apr i l , 1997 © D a v i d A . Neustaedter, 1997. 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 13/7 " / ^ * - / • The University of British Columbia Vancouver, Canada Date DE-6 (2/88) A b s t r a c t It has been widely noted that elucidating the mechan isms by which eukaryot ic organisms control the timing and location of the express ion of particular genes is one of the central problems of developmental biology. This thesis is intended to contribute to an understanding of the mechanisms by which one promoter, that of the parsley 4-coumarate:CoA ligase 1 (4CL1) gene, directs expression of at tached protein-coding sequences in developing tobacco plants. In tobacco, 227 base-pairs of the parsley 4CL1 promoter are suff icient to direct a complex pattern of developmenta l ly- regulated gene express ion (Hauffe et. al., 1991). Within this region, six sequences have been identified as possible sites of prote in-DNA interaction in parsley cel l culture (Ibid). In order to test whether these sequences are c/ 's-elements contributing to the developmental regulation of 4CL1-directed gene expression, these sequences were mutated and the effects on express ion of a reporter gene examined in different organs and cel ls of t ransgenic tobacco plants. Resul ts d iscussed in this thesis strongly indicate the presence of at least four developmental ly-important e lements within the tested sequences . Whi le mutation of each element effected express ion quantitatively, each element exhib i ted different ce l l - and organ-speci f ic i t ies in its inf luence on gene express ion. The identified elements include both positive and negative regulators of gene express ion. Express ion patterns directed by Vr\e4CL1 promoter when mutated at multiple sites suggest that the examined elements are in some cases interdependent, requiring other e lements in order to function. One element, the footprint 5/6 element, may play a particularly important role in these interactions. Taken together, these observat ions identify developmental ly- important cis-elements in the 4CL 7-promoter and provide insights into the mechan isms by which they function. Table of Contents Abstract i i Table of Contents i i i List of Tables v i List of Figures v i Acknowledgements viii 1. Introduction. 1.1 Regulat ion of Gene Expression 1.1.1 Overv iew 1 1.1.2 Role of Transcript ion Factors in Regulat ing Spat ia l and Temporal Aspects of Gene Expression 3 1.1.3 The Role of C/s-Act ing Sequences in Transcr ipt ional Regu la t ion 5 1.1.4 Prote in-Prote in Interactions in Transcr ip t iona l Regu la t ion 9 1.2 Phenylpropanoid Metabol ism 1.2.1 Introduction 11 1.2.2 Enzymes of the Genera l Phenylpropanoid Pathway 1 2 1.2.3 Phenylpropanoid Branch Pathways: Products, Function and Occurance 1 4 1.2.4 Induction of Phenylpropanoid Metabol ism 21 1.3 Phenylpropanoid Gene Expression 1.3.1 Introduction 2 2 1.3.2 Chalcone Synthase 2 3 IV 1.3.3 Phenylalanine Ammonia-Lyase 2 6 1.3.4 4 CoumarateiCoA Ligase 3 2 1.3.5 Co-ordinate regulation of Phenylpropanoid G e n e s 3 7 1.4 Thes is Object ives 3 9 2. Methods and Materials. 2.1 Product ion of Mutant Promoter/Reporter Fus ions 41 2.2 Product ion and Sampl ing of Transgenic Plants 4 4 2.3 Wounding 4 5 2.4 Analys is of G U S Expression in Transgenic Plants 4 6 2.5 In Vitro Examinat ion of Pro te in -DNA Interactions 4 7 2.6 Addi t ional Exper imentat ion 4 9 3. Extension of Previous Observations on Parsley 4CL1/GUS Expression in Tobacco. 3.1 Introduction 5 0 3.2 Quantitat ive Fluorometr ic Ana lys is of Reporter Exp resson 51 3.3 Effect of Light Level Fluctuation on 4CL1 Regulat ion 5 5 3.4 Extension of Prev ious Histochemical Observat ions 5 6 3.5 4CL Express ion in the Nectaries 6 0 3.6 D iscuss ion 61 4. Characterization of the 5/6 Element. 4.1 Introduction 7 0 4.2 Quantitative Analys is of the In Vivo S igni f icance of Footprints 5 and 6 7 2 4.3 Histochemical Ana lys is of the Qualitative Role of F P 5/6 7 6 4.4 The Role of F P 5/6 in Wound-Induced Expression 7 9 4.5 Sequences Involved in Protein Binding to 5/6 81 4.6 D iscuss ion 8 8 5. Participation of Other In Vivo Footprints in Regulation of Gene Expression In Planta. 5.1 Introduction 9 7 5.2 Footprint 1 9 8 5.3 Footpr ints 2/3 100 5.4 Footprint 4 104 5.5 Footprints 5/6 106 5.6 3" Delet ions 109 5.7 D iscuss ion 112 6. Conclusions and Future Research. 126 B i b l i o g r a p h y 130 V I List of Tables Table 2.1 Ol igonucleot ides used in site-directed mutagenesis 4 0 Table 4.1 The role of F P 5/6 in wound-induction by 4CL1 7 0 List of Figures Figure 1.1 Genera l phenylpropanoid pathway 1 3 Figure 1.2 Putative mataboli tes of c i nnamoykCoA 1 6 Figure 2.1 Exper imental constructs for tobacco transformation 4 3 Figure 3.1 Reporter gene expression in different organs of t ransgenic tobacco 5 3 Figure 3.2 Organ-specif ic i ty of position effects in 4CL1/GUS t ransgenic tobacco 5 5 Figure 3.3 Extens ions of previous histochemical obsrevat ions on 4CL1/GUS expression in tobacco 5 7 Figure 3.4 Phenylpropanoid products in tobacco flowers 5 9 Figure 4.1 Effects of mutating F P 5/6 on 4CL 7-directed exp ress ion in leaf t issues of transgenic tobacco 71 Figure 4.2 Effects in the floral organs of mutating F P 5/6 7 3 Figure 4.3 Histochemical detection of express ion directed by 810m5/6 and 405m5/6 7 8 Figure 4.4 Effect of deoxycholate on 4CL 7-nucleoprotein complex f o r m a t i o n 8 4 Figure 4.5 T issue- /organ-spec i f ic retardation of the 4CL1 promoter 8 6 Figure 4.6 Effect of speci f ic and non-specif ic competit ion on the binding of ovary nuclear proteins 8 7 Figure 5.1 The effects of mutating FP1 on 4CL 7-directed reporter e x p r e s s i o n 9 9 Figure 5.2 His tochemica l ly -detectab le reporter express ion di rected by the 4CL1 promoter after mutation of F P 1, 2/3 or 4. 101 Figure 5.3 The effects of mutating F P 2/3 on 4CL 7-directed reporter e x p r e s s i o n 102 Figure 5.4 The effects of mutating F P 4 on 4CL 7-directed reporter e x p r e s s i o n 105 Figure 5.5 Retardation of a F P 4 probe by tobacco nuclear extracts 107 Figure 5.6 The effects of mutating F P 5/6 in combinat ion with other 4CL1 c / s - e l e m e n t s 108 Figure 5.7 Reporter express ion directed by reporter/promoter fus ions miss ing TATA-prox imal 4CL1 sequences 110 Figure 5.8 His tochemica l ly -detectab le reporter express ion di rected by 4CL deletion constructs 111 Figure 5.9 Median decrease in reporter expression which resulted from each mutation of a single footprint 113 Figure 6.1 Schemat ic summary of some observat ions regarding the role of each tested element 127 A c k n o w l e d g m e n t s I think that this work is most indebted to my thesis supervisor, Dr. Car l Douglas. In addition to being a knowledgeable and judicious scient ist , throughout the severa l years during which the thesis research and thesis writing took place, he was consistently a support ive, considerate boss. There was always a sense in the lab that good sc ience and a healthy life aren't mutually exclusive. I am also indebted to essential ly all the other members of the lab for producing a co-operat ive and congenial environment. I don't think that I have ever benefited as much from a co-worker as I did from Diana Lee. In our many, many conversat ions across the lab bench and across the park bench, she informed me, counseled me, humoured me, and occasional ly, comforted me. Col laborat ing with Dr. Steve Lee was generative and educat ional . In the last couple of years of my thesis, the energy and optimism of Jenny Jones helped provide a springboard to catapult me into a larger world. I will remain appreciat ive of my friends and family for their participation in my graduate student exper ience, as wel l . Super -imposed on the sc ience, I was of course learning about life - learning to cook, and to kayak and to behave like some sort of professional. I found those events to be surprisingly inseparable from my thesis work. Thanks, Wil l , and Mom and Dad; Ute and Dominik, Lane, Zora, Natalie, Spike, Bev and Dion, Li l , Diana and Jenny again, you guys from the Haughn lab, Tamara and Andrea. Thanks Sarah, Gretchen, Tracy. And of course Darren. You guys made a world of difference. 1 1 In t roduct ion 1.1 Regulation of Gene Expression 1.1.1 O v e r v i e w Over the last decade-and-a-hal f , an enormous body of literature has been contributed towards a largely coherent meme address ing the mechan isms by which an organism directs its own development. Ev idence generated by investigations of Drosophi la development and supported by research in numerous other organisms indicates that a c a s c a d e of transcription factors is establ ished early in development. These transcript ion factors regulate the concentrat ions of each other and of an increasing number of downstream transcription factors in a cornucopian cascade which is maintained throughout the development of the organism [worthwhile reviews include (Lawrence and Struhl , 1996; Nuss le in-Vo lhard , 1996; St. Johnson and Nuss le in-Vo lhard , 1992)]. Communicat ion across cell membranes often requires mediat ion by a s ignal t ransduct ion pathway involving molecules which don't directly regulate transcript ion. The regulatory cascade del ineates success i ve l y finer div is ions of the proliferating cell mass of the develop ing organ ism. Concomitant ly with propagat ing itself, the regulatory c a s c a d e addit ional ly spec i f ies levels of structural proteins and enzymes . A s a result, the different regions express different protein prof i les, physical ly dist inguishing the ce l ls , t i ssues and developmental s tages in which they occur, and articulating the complex, emergent organism (Alberts et a l . , 1994). In both mammals (Lewin, 1980), and plants (Goldberg, 1986), it has been est imated that approximately two thirds of detectable m R N A s display restricted express ion patterns. Th is pattern is potentially consis tent with a model wherein temporospatial regulation of gene express ion is respons ib le for variat ion within the organism, between ce l ls , t i ssues and organs. Even closely related cell populations are dist inguishable by consistent , different patterns of gene express ion , whether within the mammal ian blood system (Dzierzak and Medvinsky, 1995) or a flower petal (Drews et a l . , 1992). S o m e of the questions relevant to this model of development relate to the molecular mechanisms by which patterns of gene express ion are ultimately ach ieved. At the level of an individual gene, the spat ial and temporal pattern of gene express ion within the developing organism is most often determined primarily by the interact ions of speci f ic transcription factors (also known as trans-acting factors) with short recognit ion e lements (c /s -ac t ing e lements ) embedded in the non-coding sequences proximal to the protein-coding s e q u e n c e s (Alberts et. al., 1994). In animal systems, the regulatory regions of a number of genes which are expressed in complex, d e v e l o p m e n t a l ^ regulated patterns have been relatively well character ized. These include the ADH, immunoglobulin and globin genes (Maniatis et a l . , 1987). While genes encoding numerous plant transcription factors have recently been c loned (Katagiri and C h u a , 1992) and tremendous progress has been made in elucidating mechan isms of light regulation (Donald and Cashmore , 1990), few r igorous examinat ions of the molecular regulation of complex, developmental ly regulated expression patterns have yet emerged in the plant literature. This thesis is intended to contribute to the understanding of one potentially informative model sys tem, the 4CL1 promote r . 1.1.2 Role of Transcription Factors in Regulating Spatial and Temporal Aspects of Gene Expression It has frequently been noted that one of the central problems of eukaryot ic molecular biology is to elucidate the control mechan isms directing speci f ic genes to be expressed at particular p laces and t imes within a developing organism [e.g. (Maniatis et a l . , 1987; Sa l inas et a l . , 1992). Whi le regulation of gene expression can occur at many levels, inc luding m e s s a g e stabil i ty, t ranslat ional regulat ion, protein stabi l i ty and protein activation (Griffiths et a l . , 1993), studies of numerous genes suggest that transcript ional regulation is likely the most common (Alberts et a l . , 1994; Beardsley, 1991). The developmental regulation of gene express ion is currently best understood in early development. Genet ic analysis of early development in Drosophila and other organisms has identified numerous regulators of pattern formation. In Drosophila, genes encoding components of severa l success i ve tiers of the developmental regulatory cascade have been identified and c loned (St. Johnson and Nussle in-Volhard, 1992). Their roles range from establ ishing the polarity of the egg to del ineat ing individual segments and speci fy ing organ identit ies. Almost all of these early regulators of Drosophila deve lopment are transcription factors, although some, such as bicoid may affect gene expression through more than one mechanism (Dubnau and Struhl, 1996; R ivera-Pomar et a l . , 1996). Others, such as nanos and decapentaplegic effect gene express ion by mechan isms including translat ional regulation and communicat ing a signal across the cell membrane (St. Johnson and Nuss le in-Volhard, 1992). A substantial fraction of the transcript ion factors regulating early Drosophila deve lopment are homeodomain proteins, possess ing characterist ic, highly conserved 60 amino acid homeodomain DNA-binding regions (Branden and Tooze , 1 9 9 1 ) . In both Drosophila and mouse, it has become apparent that speci f icat ion of t issue- and cel l- type is frequently determined by combinat ions of b H L H (basic hel ix- loop-hel ix) transcript ion factors. In a number of c a s e s , the combinat ion of a particular differentially-local ized b H L H protein and the ubiquitous Drosophi la bHLH protein, D A U G H T E R L E S S (or its murine homologue) have proven to be determinat ive in t issue differentiation (Jan and J a n , 1993; Weintraub, 1993). It is bel ieved these dimers of b H L H transcription factors act ivate distinct sets of genes in the effected ce l ls , directing branching in the developmental cascade and establ ishing distinct cel l l i neages . How conserved are the mechanisms of developmental regulation between plants and other organisms? Many of the components of the developmental regulatory cascade appear to be widely conserved, both with respect to structure and function. Homeodomain proteins have important developmental roles in a wide range of organisms. In plants, members of this family appear to play important roles in developmental decis ions both very early, [e.g. knotted 1 (Sinha et a l . , 1993)] and late [e.g. Athb-1 (Aoyama et a l . , 1995)] in development. Other c l asses of transcription factors such as the M Y B family, z inc finger, leucine zipper and M A D S - b o x proteins are also conserved in eukaryot ic organisms including plants (Coen, 1991; Katagiri and C h u a , 1992). S o m e components of the signal transduction machinery a lso appear to be partially conserved between plants and many other organisms (Chen et a l . , 1995). Taken together, these results suggest that many components of the regulatory system are likely substant ial ly conserved between plants and other eukaryotes. However, there are likely to be substant ial di f ferences between the regulatory sys tems of plants and organisms in other kingdoms. For example, there appear to be approximately 100 myb-re lated genes in Arab idops is , while only 1 myb-\\ke gene has been identified in Drosophila (Solano et a l . , 1995). Whi le the temporospatial regulation of gene express ion is best understood in early development, the regulation of genes encoding eukaryot ic metabol ic enzymes is probably best understood for pigment biosynthet ic pathways, such as the eye-colour pigments of Drosophila or f lavonoid pigments of plants. The regulation of studied pigment biosynthetic genes is less comprehensively understood than for some early regulatory genes. However, existing evidence suggests that the temporospat ia l regulation of pigment genes is accompl i shed similar ly, by a network of transcription factors (Holton and Corn ish , 1995; J a c k s o n et a l . , 1992). At least some of these factors belong to famil ies of transcript ion factors whose members also play important roles early in development. The regulation of some f lavonoid biosynthetic genes is descr ibed in greater detail in Sect ion 1.3. 1.1.3 The Role of Cis-Acting Sequences in Transcriptional Regulation. A supplemental perspective on gene expression to the one yielded by studying a particular transcription factor is that provided by promoter ana lys is . Deletion or mutation of protein-binding sequences within a promoter can be used to identify and examine the roles of cis -e lements within the promoter. This is normally accompl ished by monitoring the expression of a reporter gene fused to the promoter of interest, in t ransgenic cel ls or organisms [e.g. (Schel l , 1987)]. Whi le such promoter analysis is a valuable tool in studying the regulation of any gene, it is particularly valuable in studying genes for which a s imple sc reen for regulatory mutants affecting express ion of the gene is not avai lable. It is also a strategy particularly well suited to the study of plant gene express ion, due to the relative ease of transforming plants with mutated promoter constructs (Benfey and C h u a , 1989; Sche l l , 1987). However, in very few cases have the contributions of any promoter 's c/ 's-elements to developmental regulation been r igorously examined in plants. The vagar ies of 'posit ion effect'-related variat ion in levels of gene express ion contribute to the immensity of the task (Benfey and C h u a , 1989; Weis ing et a l . , 1988). Furthermore, even the most intensively studied and potential ly s implest deve lopmenta l ly -regulated promoters are complex and likely still possess many secrets . For example, the insulin promoter directs expression in a single mammal ian cell l ineage and has been intensively studied, yet the mechan i sms by which 5'-f lanking sequences direct its ce l l -spec i f ic express ion pattern are descr ibed in the current literature as being largely unknown (Naya et al . , 1995). Many of the events typical of transcriptional initiation by R N A polymerase II have been elucidated (Alberts et a l . , 1994; Draphin et a l . , 1993). In order for R N A polymerase II to be recruited by a promoter, TFIID, TFIIB and TFIIE must sequential ly bind to the TATA-box located about 25 bp upstream from the site of transcriptional initiation. Whi le these factors appear to be sufficient to tether polymerase II to the promoter, the participation of TFIIH appears necessary for transcript ion to begin. In many invest igated promoters, addit ional c is -e lements have been identif ied which either facilitate or inhibit transcription by polymerase II. In most cases , these regulatory e lements appear to act as sequence-spec i f ic binding sites for t ranscr ipt ion factors (P tashne, 1988). Many transcript ion factors which promote transcript ion appear to act by facil i tating TATA-b ind ing by the general transcription factors TFIID and/or TFIIB (Draphin et. al., 1993; Alberts et. al., 1994). The mechanisms by which repressors inhibit transcript ion are less well understood, and likely util ize a number of mechan isms. Although the impact of mutating a single gene encoding a transcript ion factor is often very dramatic in terms of express ion of target genes , b iochemical ev idence suggests that most eukaryot ic promoters bind many proteins (Maniatis, 1987; Henderson and Ca lame , 1995). Th is is consistent with regulation of gene express ion occurr ing through compl icated mechanisms. A 1991 estimate of the average number of regulatory elements per eukaryotic gene was five or more [Beardsley, 1991 #145]. A current estimate would likely be higher, as the number of identif ied regulatory e lements in the best -character ized promoters cont inues to grow (Henderson and Ca lame, 1995). Crysta l lographic studies of prote in-DNA contacts suggest that the space that a transcription factor occupies on D N A general ly falls in a range of approximately 10 - 40 bp (Boul ikas, 1994). However, the number of f ra/ is-act ing factors in a regulatory complex is not directly l imited by this constraint, as protein-protein interactions have the potential to recruit factor binding in the absence of prote in-DNA contact (Ptashne, 1988). Compos i te promoters, possess ing a number of protein-binding si tes, have the potential to respond to a combination of developmental s ignals ; generat ing a variable transcriptional response based on the integration of multiple regulatory inputs (Boul ikas, 1994). In addit ion to the protein-binding capabi l i t ies of a particular promoter, the orientation and spac ing of elements with respect to each other and the T A T A - b o x is an important factor in gene expression directed by some e lements , suggest ing that requisite interactions between trans-acWng factors are facil i tated by promoter architecture (Maniat is et a l . , 1987). The s igni f icance of posit ioning is likely more complex than s imply reflecting a requirement that interacting proteins bind to the same face of the DNA. For example, a center-to-center distance of 45 bp between serum response elements is important for chicken a-actin promoter function (Lee et a l . , 1991). Torsional deformation of the promoter resulting from co-operat ive binding to the serum response e lements may be signif icant. The posit ioning of transcription factors along regulatory sequences can be important in the formation of large mult i -protein-DNA complexes (Rousseau et a l . , 1993). Of genes exhibit ing complex, developmental ly regulated expression patterns, the alcohol dehydrogenase (ADH) and the immunoglobulin promoters are well studied examples from the animal literature (Henderson and Ca lame, 1995; Jackson and Benyajat i , 1992). In plants, l ight-inducible promoters such as those of the chlorophyll a/b-binding (cab) (Ha and An , 1988), rubisco small sub-unit (rbcS) (Donald and Cashmore , 1990) and chalcone synthase (CHS) (Block et al. , 1990) genes are best understood. The developmental regulation of promoters including B-phaseolin (Bustos et al . , 1991), phenylalanine ammonia lyase (PAL) (Bustos et al . , 1991) and the viral C a M V 35S promoter (Benfey et a l . , 1990) have been investigated in relative detai l , in plants. The structure and regulation of these and other plant promoters appears to be general ly consistent with the paradigms d iscussed above. The structure and function of the CHS and PAL promoters will be d iscussed more extensively in Sect ion 1.4.2. 1.1.3 Protein-Protein Interactions in Transcr ipt ional Regulat ion An accumulat ing body of evidence indicates that many transcript ion factors physical ly interact with each other, in addit ion to D N A . In some cases , these protein-protein interactions have been shown to be important in the target-specif ici ty of the transcript ion factors. For example , target-specif ici ty of the prototypical M Y C transcript ion factor appears to be modulated via dimerization with Max or Mad (Ayer et a l . , 1993; Ferre-D'Amare et al . , 1993). The M Y C - M a x heterodimer d isp lays altered DNA-b ind ing specif icity from the M Y C homodimer. S o m e combinat ions of these proteins promote transcript ion, while others can favour repression (Prendergast and Ziff, 1991). Two c lasses of domain, helix-loop-helix and leucine zipper, have been impl icated in the dimerizat ion of numerous transcript ion factors as homo- or heterodimers. These same domains can also result in the formation of larger complexes. In mammals, M Y C homologues are known to form tetramers through dimerization of their H L H and leucine z ipper domains (Fisher et a l . , 1991). That these domains are relatively common suggests the potential frequency of these interactions as a component of gene regulation. In Drosophi la and mouse, the emerging paradigm of organ- and t issue- specif icat ion is that a particular combinat ion of ubiquitous and region-speci f ic H L H transcript ion factors interact to speci fy spat ial identity, within an exist ing range of fates. The frequent occurrence of H L H and leucine zipper domains in putative plant transcript ion factors suggests that these interact ions are important in plants (Katagiri and C h u a , 1992), as well as animals. The dimerizat ion of different combinat ions of M Y B - and M Y C - l i k e transcription factors has been proposed as an important mechan ism to generate diversity in plant development (Wissenbach et a l . , 1993). Genet ic and biochemical evidence indicates that in many cases , inc luding the speci f icat ion of t issue-identi ty by b H L H transcript ion factors descr ibed above, interactions between the transcript ion factors are important in determining their activity and target promoter specificity. In other examples: the E X T R A D E N T I C L E homeodomain protein of Drosophi la has been shown to interact with each U L T R A B I T H O R A X and A B D O M I N A L - A to regulate their target specif icity through co-operative DNA binding (van Dijk and Murre, 1994); BICOID and H U N C H B A C K are believed to co-operate for both DNA-binding and interaction with the transcript ional complex in the early act ivat ion of many anterior genes (Arnosti et a l . , 1996). Comp lexes regulating in vivo gene expression may potentially be quite large. A 12-sub-unit DNA-bind ing protein complex has recently been isolated from plants (Chamovitz et a l . , 1996). Members of this complex are already known to be important in light-regulation of gene express ion . One model of transcriptional activation suggests that the 11 core transcript ional machinery is st imulated by a crit ical m a s s of activator proteins (Carey et a l . , 1990). Consistent with this possibi l i ty, po lymers of binding si tes often exhibit dramat ical ly st ronger transcr ipt ional act ivat ing capaci t ies than s ingle cop ies (Ondek et a l . , 1987; Sch i rm et a l . , 1987). In this model , certain critical t ranscr ipt ion factors could function in establ ishing the stabil i ty of prote in-DNA complexes. In hepatic cel ls, the ubiquitous factor N F - Y has been observed to permit stable complex formation involving the albumin promoter and hepat ic-specif ic C R P factors (Milos and Zaret, 1993). In a transient blood cell l ineage, binding of P A X to D N A appears to facil itate proximal binding of three alternate members of the ETS-1 family of transcription factors (J . Hagman, personal communicat ion) . Not all influential members of a regulatory complex necessar i ly interact with D N A directly or are involved in formation of the initial complex. The viral V P 16 factor and the vertebrate ret inoblastoma susceptibi l i ty protein Rb are known to form ternary complexes with bound transcription factors (Mudryj et a l . , 1991; O'Hare and God ing , 1988). The Oct cofactor O C A - B has been shown to contribute to the t issue-spec i f i c i ty of immunoglobul in promoter activity through contact with OCT-1 and O C T - 2 (Strubin et al . , 1995). O C A - B does not appear to directly contact target DNA, and has been proposed to act as a bridge between OCT-1 and -2 and the basal transcription machinery. 1.2 Phenylpropanoid Metabolism 1.2.1 Introduction Natural products derived from a phenylpropane skeleton have been demonstrated or suggested to serve a large number of roles in higher plants, including U V protectants, structural molecu les , phytoalex ins, and signal ing molecules in plant development and in interactions with symbiot ic insects and microbes. The diverse structures of these compounds are achieved by the action of multiple branch pathways acting downstream of a core group of three general phenylpropanoid reactions (Figure 1.1). Enzymes of the branch pathways are present in varying temporal and spatial patterns bel ieved to reflect the funct ions of their products. For example, enzymes speci f ic to lignin b iosynthesis are primarily assoc ia ted with developing vasculature. In contrast, the activit ies of the general phenylpropanoid enzymes must be present at all s i tes where downstream pathways are active, in order to supply metabol ic precursors for these pathways (Hahlbrock and Schee l , 1989). 1.2.2 Enzymes of the General Phenylpropanoid Pathway The general phenylpropanoid pathway consists of the enzymes phenylalanine ammonia lyase (PAL) , cinnamate 4-hydroxylase (C4H) and 4 -coumara te :CoA l igase (4CL), which act in that order (Figure 1.1), (reviewed in Hahlbrock and Schee l , 1989). P A L deaminates L-phenylalanine, forming cinnamic ac id. P A L has been character ized from a large number of plant spec ies (Jones, 1984) and appears to be a major control point in the diversion of carbon from primary metabol ism into phenylpropanoid metabol ism (Bate et a l . , 1994). C 4 H hydroxylates the 4' carbon of c innamic acid to yield coumaric ac id. Further modif ication of 13 [ FLAVONOIDS~~) [ISOFLAVANOIDS ) | COLIMARINS \ [ SOLUBLE ESTERS) G E N E R A L PHENYLPROPANOID METABOLISM COOH COOH 1 COOH 1 COSCoA |^NH2 ) i i P A L ^fS C4H f \ ACL V O H O H Phenyl-alanine Cinnamic Acid 4-coumaric Acid 4-coumaroyl-CoA(R-R--H) I LIGNIN: j ( SUBERIN \[ OTHER WALL-BOUND PHENQLICS } | STILBENES^ Figure 1.1. General Phenylpropanoid Pathway. Reactions in the centre box comprise the general phenylpropanoid pathway. Surrounding boxes indicate natural products resulting from distinct branch pathways. PAL, phenylalanine ammonia lyase; C4H cinnamate 4-hydroxylase; 4CL, 4-coumarate:CoA ligase. (Derived from Douglas et. al., 1994.) the 6-carbon ring by hydroxylation and methylation often occur in the early s tages of phenylpropanoid metabol ism. 4 C L , the final enzyme in the general phenylpropanoid pathway, ut i l izes var iously hydroxylated and methylated moiet ies of 4-coumar ic acid in the formation of activated Co-enzyme A esters (reviewed in Hahlbrock and Schee l , 1989). These esters act as substrates for subsequent react ions in branch pathways synthesiz ing l ignin, f lavonoids, and other products. The amino acid sequence of 4 C L shows loca l ized similarity to severa l other enzymes which util ize the hydrolysis of A T P to activate aromatic carboxyl ic ac ids (Becker -Andre et a l . , 1991) One group of these enzymes, the bacterial peptide synthetases , uti l izes the phenylpropanoid-precursor, phenyla lan ine. 4 C L isoforms with distinct substrate preferences have been observed in some plant spec ies . It has been postulated that these different 4 C L isoforms may contribute to channel ing of carbon from the general phenylpropanoid pathway into specif ic branch pathways by preferentially supplying the C o A ester of a moiety util ized by a particular pathway (Knobloch and Hahlbrock, 1975; Ranjeva et a l . , 1976). However, in some spec ies , including parsley and tobacco (Lee and Douglas, 1996; Lozoya et al . , 1988), there does not appear to be differential substrate preference between identif ied 4 C L isoforms, suggest ing that this is not a mechanism regulating carbon flow in these p lan ts . 1.2.3 Phenylpropanoid Branch Pathways: Products, Function and Occurrence A number of the possible metabolic fates of intermediates in the general phenylpropanoid pathway are illustrated in Figure 1.2. The spat ial and temporal distribution of these natural products is complex. Pheny lpropano id metabol ism is highly integrated into plant deve lopment , with speci f ic products accumulat ing in part icular organs and cel l- types at a given developmental stage (Hahlbrock and Schee l , 1989; Wiermann, 1981). Which products are synthesized is determined by the co-ordinate action of the general pathway and of branch pathways speci f ic to the observed products. For example, in parsley leaves, PAL express ion is assoc ia ted primarily with epidermal cel ls and oil-duct epithelial cel ls (Jahnen and Hahlbrock, 1988). CHS express ion is observed primarily in the epidermis, while bergapto l -O-methy l - t rans ferase (BMT ) expression is noted primarily in the oil-duct epithelial ce l ls . C H S is the first committed enzyme of f lavonoid b iosynthes is , while B M T is speci f ic to furanocoumarin b iosynthesis . The observed patterns of CHS and BMT expression correlate well with PAL express ion and the localization of f lavonoids and furanocoumarins in pars ley seedl ings. The best understood branch pathway downstream of 4 C L is the f lavonoid biosynthetic pathway. Entry into this pathway is ca ta lyzed by C H S , which cata lyzes the condensat ion of 4 -coumaroy l :CoA with 3 molecu les of ma lonykCoA, forming naringenin cha lcone. Structural elaborat ion results in different c l asses of f lavonoids, such as anthocyanins, f lavonols and f lavones [e.g. (Hahlbrock, 1981; Holton and Corn ish , 1995)]. Compounds in this group are frequently credited with roles in attracting poll inators and as UV protectant pigments. They have a lso been implicated as anti-microbial compounds (Hahn et a l . , 1985), as microbial attractants (Long, 1989), as regulators of auxin COOH Chain elongation ^ x {MaioftytCoA " Cttm -0-» • HOR o c^s C » A Cinnamoyi CoA Cinnamoyt o-pyfoo«» onyl CoA *** COj S t i l b t n i t 0 ~ ' Cinnamoyi -aldthydu •NADPH Malonyl CoA HO 0 -CHi-H~SC«A 1 0 0 C-CHi-H^-S-CoA < • L i; 0 8-c* . co, ^ O H Xanthonts MaJonyll_CoA I 0 " r - » CH,-8~S CoA Ac«toph««o««» i 0 H-S-CoA / \ R Ph«nol» R ft Btnioic »«ld» B*A*«ttftftyd«» CM,OM 8«ntalcohots Figure 1.2. Putative metabolites of cinnamoyi:CoA. Solid arrows represent steps which have enzymatically demonstrated. Dashed arrows indicate hypothetical reactions. From Zenk. 1979. transport (Jacobs and Rubery, 1988) and, in a number of recent papers, as regulators of pollen germination and growth (Taylor and Jo rgensen , 1992; Yl is t ra et a l . , 1994; Yl ist ra et a l . , 1992). For a recent review of funct ions assoc ia ted with f lavonoids, see Shir ley (1996). F lavono ids can utilize a substantial share of carbon diverted to secondary metabol ism. In parsley, leaves and seeds can each contain f lavones and f lavonol g lycos ides to several percent of dry weight (Hahlbrock, 1981). In leaves, f lavonoids are most often local ized to the epidermis. This local ization has been observed in mustard (Sinapis alba), on i on (Allium spp.), and parsley (Petroselenium hortense) (d iscussed in Wiermann, 1981). In the roots of C. arietinum and other members of the Fabaceae, two c losely related isof lavonoids accumulate to high concentrat ions (Gierse, 1975). One is local ized primarily to the rhizodermis, and the other to the cortex. In f lowers and fruit, anthocyanin f lavonoid pigments are most often assoc ia ted with the epidermis (Stafford, 1990). In petunia (Petunia hybrida) f lowers, f lavonols have been local ized to the surfaces of ovules and neighbouring maternal t issue, to pol len, and to the stigmatic sur face (Yl istra et a l . , 1994). Over la id upon the spatial control of f lavonoid accumulat ion are speci f ic temporal constraints. In parsley leaves and coty ledons maximum synthesis is evident very early in organogenesis (Hahlbrock and Gr i sebach , 1974). Temporal control of f lavonoid accumulat ion has also been demonstrated with respect to anthocyanins in petals (Jackson et a l . , 1992) and f lavonols in ovules and pollen (Ylistra et a l . , 1994). St i lbenes, while much less common than f lavonoids, are formed by a mechan ism closely related to the formation of naringenin cha lcone. C o u m a r y h C o A is similarly condensed with 3 molecules of malony l :CoA; but in this case , the two benzene rings in the resulting molecule are separated by an ethane ring (Figure 1.2). Sti lbene synthase, displays very high homology with chalcone synthase, suggest ing that the two enzymes diverged relatively recently (Schroder et a l . , 1988). St i lbenes have most frequently been identified in the xylem of woody spec ies , but are frequently a lso pathogen inducible in other t issues (Hart, 1981). Study of these compounds has largely focused on their anti-fungal activity (Dercks et a l . , 1994). Lignin b iosynthesis is another important phenylpropanoid branch pathway which util izes the products of 4 C L activity (Gross, 1985). C o A esters of hydroxycinnamic acids are metabol ized to produce monol ignols in t racel lu lar^ , then transported ac ross the p lasma membrane. Polymerizat ion of monolignols by a free radical mechan ism results in a complex, heterogeneous matrix with valuable structural propert ies [reviewed by (Whetten and Sederoff, 1995)]. Lignin b iosynthesis is restricted to the secondary wal ls of spec ia l i zed ce l l -types such as the tracheary elements of the xylem, and sc lerenchyma. Xy lem is initially developed as a continuous network throughout all organs during formation of the plant body. However, in many plants, the activity of the vascular cambium subsequent ly generates secondary xylem which also undergoes lignification (Essau , 1977). S o m e sc le renchyma cel ls (generally fibers) are commonly observed as groups of ce l ls assoc ia ted with the vasculature. Others (sclereids) are more often isolated and distributed among other cell types ( Ibid ). The biosynthesis of f lavonoids, st i lbenes and lignin are general ly accepted to involve hydroxycinnamic ac id: C o A thioesters and thus require the participation of 4 C L . The biosyntheses of a number of other c l asses of phenol ic natural products, such as coumarins, l ignans, benzoic ac ids, and cinnamic acid esters or amides have been suggested to proceed v ia C o A ester intermediates; but the ev idence for this is often less conclusive than in the preceding instances (Zenk, 1979). The assert ion has often been based on thermodynamic arguments, or on substi tut ion patterns consistent with known 4 C L substrates (Gross , 1981; Zenk, 1966; Zenk, 1979). Experimental reports both support ing and contradict ing the involvement of C o A intermediates in the biosynthesis of some of these compounds have continued to emerge over the three decades since the issue was raised. S ince different results are often obtained for different plants, some of the incons is tenc ies are likely due to inter-specif ic variat ion (Yalpani et a l . , 1993). Other factors, such as whether the subject plants are under pathogen attack, may also play a role ( Ibid ). A brief introduction to some of these compounds is included here because of the uncertainty regarding the involvement of 4 C L in their synthesis. Furthermore, the complex pattern of 4 C L expression has not been entirely l inked to part icular natural products and may partially reflect the b iosynthesis of one or more of these c lasses of phenylpropanoids. Much debate has focused on the origin of C 6 - C 1 compounds, such as benzoic ac id , chlorogenic ac id, and sal icyl ic ac id. After showing by in vivo feeding exper iments that vanil l in is derived from ferulic ac id , Zenk (1966) suggested a synthetic pathway involving an oxidative shortening of the C 3 tail of an activated coumaric a c i d - C o A thioester. Whi le many subsequent experiments failed to show a requirement for C o A in the synthesis of C 6 - C 1 compounds, a recently improved assay system in Lithospermum shows that, in this system, the C o A ester is a detectable intermediate in the synthesis of p-hydroxybenzoate (Loscher and Heide, 1994). Much current interest in C 6 - C 1 molecules focuses on their roles as signals or intermediates in the plant defense process (Leon et a l . , 1993). Whi le there is little direct ev idence for the involvement of 4 C L in the synthes is of coumar ins, the strong correlation between elicitor-induced furanocoumarin biosynthesis and the induction of 4CL expression in parsley cel ls has been taken to support a role for the C o A ester in coumarin biosynthesis (Brown, 1985; Tietjen and Matern, 1983). For example, the addition of cell wall extracts from the fungus Phytopthera megasperma to parsley cell culture results in the production of large amounts of coumarin. Prior to the accumulat ion of coumar in, the induction of genes encoding all members of the general pheny lpropanoid pathway is observed. In addition to being elicit ible, coumar ins are developmental ly synthes ized in floral and vegetat ive organs of a wide variety of plants (Brown, 1985). Var ious reports also support a model where other c innamic acid esters can be derived from the C o A ester (Gross, 1981; Zenk, 1979). If accurate, this could implicate 4 C L in a great deal more synthetic activity than is currently recognized. In a wide variety of plants, most low molecular weight phenol ics appear to be conjugated as esters of g lycos ides (Harborne, 1979). Amides represent another prevalent c lass of phenol ic conjugates. If, as has been proposed, phenolic amides are also derived from the C o A ester (Gross, 1981; Zenk, 1979), it would be of specia l s igni f icance in the flower. Hydroxycinnamic acid amides appear to be the largest phenol ic component of reproductive organs in a wide range of plants (Mart in-Tanguay et a l . , 1978). Because of the timing of accumulat ion, and correlat ions between def ic iencies in these compounds and the occur rence of sterile f lowers, hydroxycinnamoyl amides have been proposed as regulators of floral development (Mart in-Tanguay et a l . , 1982; Ponchet et a l . , 1982). Interestingly, these compounds are a lso induced by wounding and infection (Cabanne et al . , 1977). A s the research for this thesis focuses on tobacco plants, it s e e m s worthwhile to highlight what is known regarding the accumulat ion of phenylpropanoids in this spec ies . Anthocyanins are prominent in the col lars of tobacco f lowers. In healthy tobacco plants, lignin consti tutes approximately 3 .5% of the dry weight of leaves (Chortyk, 1972). Tobacco anthers are known to accumulate coumaroy l ty ramine and d icoumaroylputresc ine and -spermid ine, while ovar ies accumulate very large amounts (2 micromoles/ gram fresh weight) of caffeoylputrescine and -spermidine (Cabanne et a l . , 1977). The coumar in , esculet in, is detectable in the organs of all four whorls of tobacco flowers (Watanabe and Wender, 1965). Whi le little free benzo ic ac id is detected, large amounts (100 micrograms/gram fresh weight) are likely present in conjugated forms (Yalpani et a l . , 1993). Chlorogenic acid is also detectable in most, if not all organs of tobacco (Watanabe and Wender, 1965). 1.2.4 Induction of Phenylpropanoid Metabolism The induction of phenylpropanoid metabol ism is a common response of plants subjected to mechanical wounding or pathogen attack [reviewed by (Dixon et a l . , 1993)]. Character ist ical ly , phenylpropanoid metabol ism is dramatical ly act ivated at the site of injury. Products that accumulate locally often include smal l phenylpropanoid molecules such as l ignans, bel ieved to act as phytoalexins, as well as wal l -bound phenol ic esters. Taxonomica l ly -character is t ic phenylpropanoid phytoalexins, such as the isof lavonoids produced by legumes, are often observed, as well . The phenylpropanoid-der ived molecule, sal icyl ic ac id , is often produced at greater d istance from the site of injury and is likely to play a role in systemic res is tance against pathogens. Phenylpropanoid metabol ism is also activated by UV-l ight and a variety of other s t resses such as nutrient starvat ion, temperature shock or cel lular dilution. The role of f lavonoid synthesis in response to UV-l ight seems relatively c lear because of the UV-absorb ing propert ies of f lavonoids (see Shir ley, 1996), but the s igni f icance of phenylpropanoid metabol ism in many stress responses is poorly understood (Dixon et a l . , 1993). 1.3 Regulation of Phenylpropanoid Gene Expression 1.3.1 Introduction The regulation of phenylpropanoid gene expression provides an attractive model for the study of developmental regulation of plant gene expression. Many of the genes encoding enzymes from both the general pathway and some of the branch pathways have been isolated, often from multiple spec ies (Douglas et a l . , 1992). Promoters of PAL, 4CL and CHS genes have been shown to convey complicated patterns of developmental regulation upon reporter gene express ion. Where examined, these expression patterns have proven to be correlated with expression of the endogenous genes (Bevan er. al., 1989; Liang et. al., 1989; Koes et. al., 1990; Schmid, 1990; Reinold et. al., 1993; Lee et. al., 1995) suggest ing that transcript ional regulation by promoter sequences is an important component of developmental regulation of phenylpropanoid gene expression. Several of these studies have been performed in tobacco and Arabidopsis , using promoters from heterologous plants, suggest ing evolutionary conservat ion of the regulatory circuits directing these express ion patterns in plants. Bec ause of the complexit ies of the pathways under regulation, and the variation in gene number and expression characterist ics in the gene famil ies encoding these enzymes, investigations in this area a lso contribute to an understanding of co-ordinate regulat ion, both within gene famil ies and metabol ic pathways. 1.3.2 Chalcone Synthase Of all phenylpropanoid genes, the regulation of chalcone synthase (CHS) expression is certainly understood best. While much of the study of CHS has addressed its environmental induction by light and biotic s t ress, it has also been well character ized with respect to its developmenta l regulation. A s noted earlier, C H S express ion is primarily assoc ia ted with the epidermis: in f lowers, fruit, and in young vegetat ive t i ssues [reviewed in (Stafford, 1990)]. In petunia, there appear to be eight complete CHS genes. Of these, two [CHS A and CHS J) are expressed in healthy floral t issues, in a l ight-dependent manner (Koes et a l . , 1989). Ana lys is of reporter express ion directed by the promoters of five of these CHS genes shows that only the CHS A and CHS J promoters are capable of directing reporter express ion in petunia flowers (Koes et a l . , 1990). Express ion is apparent in the epidermis of the petals, ovules, and placenta, and throughout the anthers. Reporter expression is for the most part temporal ly and spatially consistent with CHS m R N A accumulat ion patterns in both anthers and petals. However, in the corol la of the petal , endogenous transcripts accumulate almost exclus ively in the epidermis, while CHS A and J both direct expression in parenchyma as wel l . This result suggests that a post-transcript ional mechan ism of regulation may be involved in spatial restriction of express ion in this context. Interestingly, CHS A and CHS J direct proper reporter express ion in the tube of the petal, indicating partially distinct mechan isms of regulation between these two cont iguous regions of the same organ and t issue. Whi le the petunia C H S A and CHS J genes show identical patterns of express ion , their regulation appears to be achieved by partially distinct mechan isms. In the corol la, expression of CHS J, but not CHS A, requires functional copies of the An 1, An 2 and An 11 regulatory genes (Quattrochio et al . , 1993). In the anther, CHS J, but not CHS A, requires An 4. It has recently been demonstrated that the CHS J promoter can be act ivated in tobacco by a myb-related transcription factor normally expressed in petunia petals (Solano et a l . , 1995). Similarly, CHS J (but not CHS A) expression is activated by the maize myb and myc homologues C1 and R (Quattrochio et al. , 1993) . A 67 bp promoter fragment of the petunia CHS A promoter is suff ic ient to direct weak, l ight-dependent, f lora l -speci f ic exp ress ion in petunia (van der Meer et a l . , 1990). Signif icantly, this fragment omits the highly conserved Box 1 and G-box sequences observed in many C H S genes (Weisshaar et a l . , 1991). Further analysis indicates that 14 bp of this fragment, containing a direct repeat of the sequence T A C P y A T (-67 to -53 of the promoter), acts as a negative element with respect to both organ- and cel l - specif ic expression (van der Meer et al . , 1992). When fused to the C a M V 35S promoter 'B ' domain, the T A C P y A T repeats substantial ly reduce vegetat ive express ion and limit 35S-d i rec ted express ion within the anther. In exper iments where the CHS A promoter is subjected to 5' deletion analysis, deletion of the T A C P y A T repeat el iminates detectable reporter express ion , suggest ing an addit ional role as an activator (van der Meer et al . , 1990). Sequences of an Antirrhinum CHS promoter fragment between -197 and -39 are necessary for floral expression in t ransgenic tobacco, while s e q u e n c e s TATA-prox ima l to -39 are sufficient for vegetat ive expression (Fritze et a l . , 1991). The T A C P y A T repeat noted in the petunia CHS A promoter is present within Antirrhinum CHS p romo te r sequences required for floral express ion. Furthermore, a natural 5 bp deletion in the T A C P y A T repeat resulted in a 6 5 % reduction of in CHS transcript levels (Sommer et a l . , 1988). This result agrees with the observat ions from petunia suggest ing that the T A C P y A T element acts as an activator of floral expression. It does not appear to have been noted whether the T A C P y A T also acts as a negative element in Antirrhinum. However, the delila mutant of Antirrhinum d i s p l a y s ectopic CHS expression in petal mesophyl l and vascular cel ls (Luo et a l . , 1991).[De//'/a has subsequently been identified as a myc homologue (Goodr ich et a l . , 1992)]. Together, the loss of t issue-restr icted express ion in petunia and Antirrhinum suggests that the molecular mechan isms restricting C H S express ion to epidermal cel l - types may be relatively easy to separate from other regulatory character is t ics. A bean CHS 15 promoter fragment has been shown capable of direct ing appropriate reporter express ion in t ransgenic tobacco (Schmid et a l . , 1990). Deletion of promoter sequences between -82 and -74 (including a G-box) abol ished expression in petals of the t ransgenic tobacco (Dixon et a l . , 1993). G-boxes have been implicated in environmental ly-regulated expression of a number of other genes , such as rbcS (Donald and Cashmore, 1990) and ADH (McKendree et a l . , 1990). Putat ive bZIP transcription factors have been identified which bind to the G-box of promoters including CHS 15, but it is not yet clear what role these factors play in developmental ly-regulated express ion (Feldbrugge et a l . , 1994). A small CHS 15 promoter fragment containing the G-box and another element designated an H-box (Loake et a l . , 1992) appears sufficient to establ ish express ion in the petal epidermis and root tips of t ransgenic tobacco . Furthermore, mutation of an H-box sequence in an Antirrhinum CHS promoter decreases petal express ion(Coen et al . , 1986; Dixon et al . , 1993). It has been proposed that the G - and H-boxes act in combination to activate deve lopmenta l ly - regu la ted CHS expression (Will iams et a l . , 1992; Y u et a l . , 1993). However, as developmental ly-regulated expression of CHS is l ight-dependent, and the G - and H-boxes have been implicated in environmental ly- induced express ion of CHS genes (Dixon et a l . , 1993; Loake et a l . , 1992), they may participate in developmental regulation through a complex mechanism. 1.3.3 Phenylalanine Ammonia Lyase In all cases investigated thus far, the PAL and 4CL gene famil ies have been found to be co-ordinately regulated (Hahlbrock and Schee l , 1989; Logemann et a l . , 1995). This pattern has been most extensively studied in parsley, where the complete PAL and 4CL gene famil ies have been c loned (Logemann et al . , 1995). It has been suggested that co-ordinate regulation of PAL and 4CL genes is ach ieved through conserved c/s-e lements, common to the promoters of both gene famil ies. Thus , observat ions regarding the regulation of PAL expression are of part icular interest in relation to 4CL. Most functional information about PAL regulatory sequences has come from reporter expression directed by the bean PAL2 and 3 promoters in transgenic Arabidopsis, tobacco and potato ( Bevan et. al., 1989; L iang et. al., 1989; Leyva et. al., 1992; Shufflebottom et. al., 1993; Hatton et. al, 1995,1996). In all three spec ies , the PAL2 promoter directs express ion in xylem (Bevan et. al., 1989; Liang et. al., 1989; Shuff lebottom et. al., 1993). In potato and tobacco, where secondary growth also occurs, reporter staining is also present in the ray parenchyma of secondary xylem. The PAL3 promoter is not active in xy lem, instead directing express ion in a ring of cel ls surrounding the vasculature. In the roots, this cel l layer is well character ized as the suber in ized Caspar ian strip, suggest ing that this express ion domain may be representative of endogenous phenylpropanoid metabol ism (Shuff lebottom et a l . , 1993). Both promoters direct express ion in root cort ical cel ls and root exodermis of tobacco and potato. In Arab idops is , root express ion is limited to the vasculature. In all three spec ies , both promoters direct epidermal express ion ( Ibid ). In f lowers, express ion patterns again vary with both spec ies and promoter (Shufflebottom et a l . , 1993). PAL2, which directs petal express ion in bean, directs express ion in petal epidermal cel ls of tobacco and potato. PAL3, which is not detectably expressed in bean petals, directs low levels of expression in tobacco and potato petal epidermal ce l ls . In unpigmented Arab idops is petals, neither promoter directs reporter express ion. PAL2 was observed to direct express ion in pollen of all three spec ies and in the fracture zone of tobacco anthers and Arab idops is si l iques ( Ibid ). PAL3-6irected express ion was observed in the pollen of only tobacco. Both promoters direct express ion in the petal absc iss ion zone of Arab idops is f lowers. These results suggest that reporter express ion directed by the bean PAL promoters in heterologous spec ies represent partial integrations into host patterns of phenylpropanoid gene express ion . In all three plants, either one or both promoters directed express ion in severa l cel l - types known to accumulate phenylpropanoid metabol i tes: xy lem cel ls , pol len, and the stem epidermis. Other aspects of reporter express ion were specif ic to the host spec ies . In Arab idops is , the absence of petal express ion correlates with the absence of anthocyanins [although other f lavonoids are present in Arab idops is petals (Shirley et a l . , 1995)]. A lso in Arabidopsis , there was no reporter express ion in root tips or exodermis, or in wounded leaves. Whi le s o m e of the spec ies-spec i f i c di f ferences in express ion pattern may represent an accurate reflection of the variation in the hosts' phenylpropanoid program, others likely do not. Although both PAL2 and PAL3 directed wound responses in tobacco and potato, neither produced an observab le response in Arabidopsis (Shufflebottom et a l . , 1993). It is likely that this reflects an evolutionary d ivergence between host and source signal transduction pathways for phenylpropanoid gene activat ion rather than dif ferences in phenylpropanoid metabol ism, as both Arab idops is PAL (Ohl et a l . , 1990) and ACL (Lee et al . , 1995) m R N A s undergo wound-induct ion. The expression patterns descr ibed above can be conferred by 310 bp and 460 bp promoter fragments of bean PAL2 and 3 , respectively (Shuff lebottom et a l . , 1993). Within these promoter fragments, a number of smal ler sequence motifs within these fragments have been postulated or demonstrated to play roles in developmental regulation by the PAL promoters. In the PAL 2 promoter, sequences between -289 and -74 are necessary for xylem expression, but not other aspects of t issue-speci f ic express ion, such as petal expression (Leyva et a l . , 1992). Th is is particularly interesting because deletion of this region results in strong phloem express ion, while abol ishing xylem express ion . Examinat ion of the express ion directed by a number of internally-deleted promoter constructs leads to the conclusion that sequences between -135 and -119 act as an activator of xylem and a repressor of phloem express ion . Xy lem activat ion/phloem repression was postulated to reside in the sequence C C A ( A / C ) C ( A / T ) A A C ( C / T ) C C , because of similar sequences present in promoters of a variety of phenylpropanoid genes (Leyva et al . , 1992; Lois et a l . , 1989; Ohl et a l . , 1990). Th is element was first l inked to UV- and el ic i tor- induced expression of the parsley PAL1 gene, and was observed to be the site of U V - and el ic i tor- induced footprints in parsley cell culture (Lois et a l . , 1989). This element, now referred to as AC-r ich Box II, has subsequent ly been disrupted by site-directed mutation in the context of a -254 promoter fragment (Hatton et a l . , 1995). When all twelve nucleot ides are subjected to t ransversions, xy lem express ion is decreased and strong phloem expression is observed. Net stem express ion is reduced approximately two-fold. When only the 3' five nucleot ides are mutated, no significant effect is noted. These results suggest that the same portion of the element is important for both phenotypes and that only one portion of the consensus sequence is important for these phenotypes. The AC-II mutation a lso reduces petal express ion approximately 3-fold. S e q u e n c e s directing posit ive regulation in phloem were observed to reside between -480 and -289, in contradict ion of earl ier observat ions (Leyva et a l . , 1992) . The AC-I sequence present in PAL2 and PAL3 and the AC-MI sequence present in PAL2 exhibit substantial similarity to the AC-II box (Hatton et al . , 1995). These sequences and the G-box have been mutated individually and in combinat ion, in the PAL2 promoter in order to examine their functional s igni f icance (Hatton et a l . , 1995; Hatton et a l . , 1996). A s with AC-II , mutation of AC-I results in ectopic express ion in phloem cel ls and reduced xylem expression in transgenic tobacco. This suggests that both AC-I and AC-II function similarly and are both required for phloem repression. In petals, the three-fold reduction upon mutating AC-I was again reminiscent of results with AC-II . However, mutation of AC-I has the additional phenotype of el iminating reporter express ion in root tips, suggest ing that the roles of these elements are overlapping, but not identical. When AC-I and -II are mutated in combinat ion, a further substantial decrease is observed in xy lem and petals and, interestingly, phloem. When the third AC- r i ch sequence , AC-MI, is mutated, the effect seems similar to results for AC-I and AC-II in petals, but negligible in the vasculature (Hatton et. al., 1995). This result again points to over lapping but partially distinct roles for the AC- r i ch elements. AC-I and -II were subsequent ly mutated in the PAL3 promoter (Hatton et a l . , 1996). In this context, mutation of either e lement c a u s e s a several- fold reduction in express ion in the endodermis (PAL3 does not direct xylem expression) and the root tip. Express ion in the petal, is again effected. Unlike the observed result for PAL2, the mutation of both elements together causes no further dec rease in express ion in either stem or petals. The G-box in PAL2 is located centrally in the 120-bp interval between AC- I and -II (Hatton et a l . , 1995). Mutation of the G-box has very little effect on express ion in the stem. In the petal however, approximately a three-fold reduction in express ion is observed. Taken together, these experiments are informative with respect to the roles of particular cis -sequences of the bean P A L 2 and 3 promoters in directing developmental ly-regulated express ion . The results suggest that the AC-I and AC-II elements present in both promoters are important in positive regulation of xylem and petal express ion and negative regulation of phloem express ion. These elements are to a large degree functionally redundant, and the effect of altering these sequences in combination appears to be additive in some cases , but not in others. The AC- l l l -box and G-box appear to play relat ively l imited roles in developmenta l ly- regulated exp ress ion , but contr ibute quantitat ively to petal express ion . Appropr iate patterns of express ion are observed when the entire -245 to -75 region of PAL2 is deleted in the context of a larger promoter, suggest ing that addit ional functional redundancy is likely present in the native promoter (Hatton et a l . , 1995). Cand ida te transcript ion factors responsib le for mediat ing activation from the A C - and G-boxes have been noted (Sablowski et. al., 1994; Hatton et. al., 1995). Leucine zipper proteins have been shown to bind to G-box sequences matching the one found in PAL 2 (Feldbrugge et a l . , 1994). All of the d iscussed AC-r ich sequences match preferred binding sites of the maize MYB- l i ke protein P (Grotewald et a l . , 1994; Sab lowsk i et a l . , 1994). AC- l -b ind ing activity in tobacco petal nuclear extracts is targeted by an t i -MYB305 anti-bodies, adding support to the above model. Furthermore, the M Y B 3 0 5 and M Y B 3 4 0 proteins, which are expressed in Antirrhinum f lowers, are each able to activate express ion of PAL and other phenylpropanoid genes in yeast and plant protoplasts (Moyano et a l . , 1996; Sablowski et al . , 1994). However, the transcriptional regulation of PAL genes likely involves addit ional f rans-act ing factors. In animal sys tems, M Y B proteins have been determined to bind heterologous partners, somet imes in multiple (Amati and Land, 1994). In plants, putative Z n -finger proteins have been isolated through their affinity to an AC- I -homologous sequence known as the H-box in the bean CHS15 promoter (Yu et a l . , 1993). Interestingly, it has been demonstrated that the combinat ion of the bean H/AC- l -box and G-box is capable of activating reporter express ion in tobacco protoplasts, in a manner facil i tated by co-express ion of M Y B 3 0 5 (Sablowski et a l . , 1994). 1.3.4 4-Coumarate:CoA Ligase In parsley, there are two 4CL genes, which are approximately 9 8 % identical for over 400 bp 5' of the genes' TATA-boxes (Douglas et a l . , 1987). O n this bas is , and because of similarit ies in developmental (Wu and Hahlbrock, 1992) and stress- induced expression patterns (Douglas et a l . , 1987), it is bel ieved that each of the two genes is responsive to the full complement of s ignals inducing 4CL express ion. Exper iments in which sequences of the parsley 4CL1 promoter were used to direct express ion of G U S in transgenic plants suggest that 4CL1 promoter sequences are adequate to direct complex expression patterns in tobacco (Hauffe et al . , 1991) and Arabidopsis (Lee et a l . , 1995). Resul ts from histochemistry, northern analys is and in situ h y b r i d i z a t i o n suggest that reporter expression patterns directed by the parsley 4CL1 promoter in tobacco and Arabidopsis are very similar to that of endogenous 4CL genes (Reinold et. al., 1993; Lee et. ai, 1995). Thus, as with PAL, transcript ional control appears to represent a major determinant of developmental ly-regulated 4CL gene express ion, and the mechan isms regulating this transcriptional control seem to have been highly conse rved , e v o l u t i o n a r y . In tobacco, there are at least 4 genes encoding 4CL (Lee and Doug las , 1996). The 4 identified genes exhibit similar express ion patterns, with m R N A levels highest in the stem, but also high in severa l floral organs and roots. In situ hybridization showed high 4CL transcript levels in st igmatic papi l lae and sub-epidermal ce l ls , in the stomium and endothecium of anthers at differing stages, and in the sur faces of the ovules, pollen and carpel wall (Reinold et a l . , 1993). Hybr idizat ion was also evident in petals, primarily in epidermal ce l l -layers, but throughout the base of the organ at later s tages of floral development. In sepa ls , hybridization was only evident in the v a s c u l a t u r e . Temporal patterns of expression were also examined in tobacco floral organs (Reinold et. al., 1993). Whi le expression directed by 4CL1 in the petal epidermis, the st igma epidermis, pol len, ovules and nectar ies were all observed in open flowers (stages 5 and 6, ibid), some dif ferences in the timing of expression at these sites were noted. Express ion in the st igma epidermis and nectary were detected at the earl iest developmental stage examined (stage 1, 13-14mm buds, ibid). Express ion in the petal became evident latest in development (stage 5, opening flower, ibid). Expression in nectaries and ovules appeared to decl ine in fully open f lowers, while express ion in the other t i ssues remained strong at the latest stage investigated. Thus, the 4CL1 promoter directs a pattern of express ion in t ransgenic tobacco which var ies temporaly as well as spatial ly. Whi le reporter gene expression directed by the 4CL1 promoter reflects patterns of endogenous tobacco transcript accumulat ion in most cases , an exception was noted at the base of the petal (Reinold et. al., 1993). Al though endogenous 4CL transcript levels were high in cel ls throughout the base of the petal, neither histochemical staining nor in situ hybridizat ion was able to detect reporter express ion in t ransgenic plants. Express ion of a complete parsley 4CL1 gene in tobacco petals was similar to endogenous genes, suggest ing that coding or intron sequences are likely important in this particular aspect of express ion ( Ibid ). While in situ hybridization was not performed on other parts of the plant, h is tochemica l examinat ion of plants t ransformed with 4CL 7/reporter fusions suggested that expression occurs at si tes in addition to those noted above. Reporter expression was observed in nectar ies, sub-ap ica l cel ls of the root, root hairs, in xy lem throughout the plant and in ray parenchyma of old stems (Hauffe et a l . , 1991). Northern ana lys is supports a correlation with endogenous transcript ion in roots and stems (Lee and Douglas, 1996; Reinold et al . , 1993). Most of these si tes of reporter express ion also correlate with known patterns of accumulat ion of phenylpropanoid natural products, as descr ibed earl ier (Section 1.2). Reporter gene expression directed by bean PAL promoters in tobacco correlates with 4CL 7/reporter express ion in xylem and roots, providing additional ev idence that the express ion at these sites is representative of native phenylpropanoid gene express ion (Shufflebottom et. ai, 1993). Two previous studies have examined the role of sequences within the 4CL1 promoter in mediating developmental ly-regulated gene express ion (Hauffe et a l . , 1993; Hauffe et a l . , 1991). The first study reported the effects of 5' deletions within the promoter sequences . A 227 bp 4CL1 promoter fragment (-210 to +17 relative to the t ranscr ipt ion init iation site) d i rected organ- and ce l l -spec i f i c express ion identical to the larger, 600 bp promoter, suggest ing that the 227 bp promoter contains all the sequences necessary for correct temporospat ia l regulation by the 4CL1 promoter. 5' deletion to -174 abol ishes detectable expression in vegetative organs and petals. Fragments deleted to -120 or -78 did not provide detectable express ion in any transgenic plants, suggest ing that sequences between -120 and -210 are quali tat ively or quantitatively necessary for all developmental aspects of expression and that there are at least two important domains within this 90 bp region. In vivo footprinting analys is of the 4CL1 promoter in parsley cel ls provided an interesting correlat ion with the observed importance of these sequences . Whi le in the TATA-prox ima l 200 bp six (Hauffe et a l . , 1991) or eight (Becker-Andre et a l . , 1991) protected sequences are observed, no others are evident in the TATA-d is ta l 500 bp of flanking sequence. Due to the correlat ion with regions of the promoter important for developmental ly- regulated gene express ion , the footprinted s e q u e n c e s (designated F P 1-8) were considered strong candidates for cis-elements important in developmental regulation by the 4CL1 p romote r . Subsequent analysis examined a large number of internally-deleted promoter constructs, and initiated in vitro analysis of the tobacco nuclear proteins binding to 4CL1 promoter sequences (Hauffe et. al., 1993). One of the most striking observat ions from the addit ional analys is was that deletion of sequences between -166 and -126 resulted in reporter express ion in phloem cel ls. This suggests that sequences within this region may be functionally equivalent to the A C -II element of the bean PAL2 promoter. Because of sequence similarity to AC-I I , in vivo F P 4 (situated at -127 to -129) is the most likely candidate for this role. This observation supports the model that footprinted sequences from parsley cel l culture may have functional importance in developmental regulation of gene express ion in planta. Resul ts from deletion analysis also indicated that, as with PAL2, there was functional redundancy present within promoter sequences . For example , the deletion of sequences including the -210 to -120 region could be compensated for by replacement with the -597 to -244 4CL1 promoter fragment. Sequences very proximal to T A T A are likely a lso important in direct ing t issue-spec i f ic express ion , as a construct miss ing -52 to +17 was only able to direct weak pollen express ion from a C a M V 35S 46 bp minimal promoter, in spite of directing moderate levels of expression in tobacco seedl ings and parsley cell culture (Hauffe et. al., 1993 ). Ge l retardation experiments using the -210 to -27 4CL1 promoter fragment as a probe indicated that tobacco shoot nuclear extracts p o s s e s s the capaci ty to form a number of sequence-spec i f i c protein-D N A complexes with this promoter (Hauffe et al., 1993). Tobacco ovary and seedl ing nuclear extracts yielded a very similar retardation pattern. When different fragments of the promoter were used to compete for complex formation, it became clear that only sequences between -120 and -77 could compete with the -210 to -27 promoter fragment for formation of the detected complexes . This result suggested the possibil i ty that sequences between -120 and -77 were involved in interactions with prevalent 4CL 7-speci f ic D N A - b i n d i n g proteins. In vivo F P s 5 and 6 are located within this region ( at -107 to -109 and -97 to -99, respectively) and were cons idered possib le si tes of the prote in-DNA interactions (Hauffe et. al., 1993). Cons idered as a whole, the empirical data on the 4CL1 promoter indicate: i) 227 bp of this promoter can direct a complex pattern of express ion in t ransgenic tobacco which is largely consistent with endogenous 4CL expression patterns, ii) the presence of several promoter modules within this fragment which can effect developmental ly-regulated express ion, iii) the presence of 6 to 8 sites of potential pro te in-DNA interaction within this fragment, iv) the presence of an element between -166 and -126 with some sequence and functional similarity to the bean PAL2 AC-II box, v) the -120 to -78 region is essent ia l for the formation of a number of readily detectable comp lexes with tobacco nuclear proteins. 1.3.5 Co-ordinate Regulation of Phenylpropanoid Genes The elaborat ion of complex molecules can require co-ordinate express ion of a number of genes encoding the requisite enzymes. In phenylpropanoid metabol ism, co-ordinate regulation has been suggested for biosynthet ic pathways elaborat ing diverse molecu les including furanocoumar ins, l ignins, and f lavonoids (Hahlbrock and S c h e e l , 1989; Jackson et a l . , 1992; Whetten and Sederoff, 1995). For example, the biosynthesis of anthocyanins from phenylalanine and ace ty l -CoA in Antirrhinum majus petal epidermal cel ls requires the action of approximately 14 enzymes (Jackson et a l . , 1992). Most, if not al l , of the genes encoding these enzymes are developmental ly regulated, requiring co-ordinate expression of these genes, possibly as a result of shared regulatory mechan isms (Holton and Corn ish , 1995; Jackson et a l . , 1992 ) . The co-ordinate regulation of anthocyanin biosynthetic genes has been most rigorously investigated in maize (Zea mays), snapdragon [Antirrhinum majus) and petunia (Petunia hybrida) [ d iscussed in (Holton and Corn ish, 1995)]. In Antirrhinum this analysis has been extended to show co-ordination at the level of individual cel ls (Jackson et a l . , 1992). Whi le there appears to be a high degree of co-ordination of anthocyanin biosynthetic genes, in petunia and Antirrhinum, genes early in the pathway appear to be regulated by partially distinct mechan isms from genes late in the pathway (Holton and Corn ish , 1995; Jackson et a l . , 1992). Furthermore, even members of the same gene family may be regulated by distinct mechan isms. At least two of the genes regulating petunia C H S J exhibit little influence over CHS A express ion, even though the two CHS genes are expressed in a very similar manner (Quattrochio et a l . , 1993). Severa l of the genes cloned thus far from both monocots and dicots as regulators of anthocyanin biosynthetic genes display homology to genes encoding the mammalian myb and myc t ranscr ip t ion factors (Holton and Corn ish , 1995). Recent experiments suggest that genes encoding enzymes of the general phenylpropanoid pathway are likely a lso regulated by M Y B - and MYC- l i ke transcription factors. Sab lowsk i et. al., (1994) demonstrated that ant ibodies raised against Antirrhinum M Y B . 3 0 5 recognize tobacco nuclear proteins which bind the bean PAL 2 and CHS 15 promoters. Furthermore, these binding sites have been implicated in petal expression of PAL 2 (Hatton et. al., 1995) and CHS 15 (Loake et. al., 1992). MYB.305 is also able to activate express ion from bean P A L 2 in tobacco protoplasts. While no b iochemical ev idence has been reported which demonstrates interaction of a 4CL promoter with a M Y B protein, numerous authors have noted putative c/'s-elements in the parsley 4CL1 promoter with homology to MYB-bind ing sites in bean P A L 2 (Lois et. al., 1989; Ohl et. al., 1990; Loake et. al., 1992; Douglas, 1996). Two putative elements in the 4CL1 promoter, F P s 4 and 8, are particularly similar to the suspected MYB-b ind ing sites (Logemann et. al., 1995) and are cons idered strong candidates to mediate co-ordinate regulat ion. Two other elements in the 4CL1 promoter have been noted for their homology to c/s-elements in parsley PAL promoter sequences (Logemann et. al., 1995). However, it has been noted in the mammal ian literature that a 13 bp consensus sequence allowing up to five mismatches would posit a binding site every 82 bases throughout the genome (Mikkelsen, 1993). Given the numerous claims regarding conserved cis -elements in phenylpropanoid promoters, it seems judic ious to await further functional ev idence before attempting to evaluate their accuracy . 1.4 Thes is Objectives The over-al l objective of my thesis research was to contribute to an understanding of the mechanisms by which temporospatial regulation of 4CL1 transcription is achieved in the course of development. Toward that end, I tested the hypothesis that sites of p ro te in -DNA interaction within the 4CL1 promoter identif ied in pars ley cel l culture by in vivo footprinting represented cis -e lements important to developmenta l regulation in planta. In order to test this hypothes is , I produced si te-directed mutations at these putative elements in a number of combinations, and examined expression of the result ing constructs in t ransgenic tobacco plants. The effects of the mutations on reporter gene expression were gauged in two ways: quantitative analys is of express ion in different organs or t issues, and qualitat ive ana lys is of cel l -speci f ic express ion . In an effort to extend these observat ions to the level of the protein-DNA interactions, I ut i l ized gel retardation assays to examine organ-speci f ic complex formation and to look at the effects of mutated cis -e lements on complex formation. Except where indicated, the exper iments descr ibed in this thesis were performed by me. 2. Methods and Materials 2.1 Production of Mutant Promoter/Reporter Fusions The promoter/reporter fusions created for transformation into tobacco plants are summar ized in Figure 2.1. Standard procedures of molecular biology used in generating the constructs were accompl ished as per Sambrook et. al. (Sambrook et a l . , 1989). All constructs in the 810 ser ies were derived from the descr ibed 810 construct (Hauffe et a l . , 1991). The -210 to +17 sequences of the parsley 4CL1 promoter were exc ised from this plasmid as a Hind III - Bam HI and sub-c loned into these sites in the pBluescript I K S plasmid (Stratagene, LaJo l la , C A ) . Single-stranded DNA for mutagenesis was produced according to the B io -Rad M u t a - G e n e ^ M Phagemid In Vitro Mutagenesis Kit Instruction Manual (B io-Rad Laborator ies, R ichmond, C A ) . Mutagenes is of the 810 construct was accompl ished via the Kunkel method, using the ol igonucleot ides shown in Table 2.1. Mutants were selected on the bas is of introduced restriction si tes. Each se lected mutated promoter was sequenced through the entire promoter fragment, using a T7 Sequenc ing Kit (Pharmacia , P iscataway, NJ) . Double mutants were generated using single-stranded D N A s of the single mutants, through the use of the same procedures. Mutants of the internally deleted constructs were generated from the Hind Ill-Bam HI promoter fragment of the 423 construct (Hauffe et a l . , 1993). The 5/6 footprint was mutated as descr ibed above. The 405m5/6 construct was produced by switching the Sac l-Bam HI promoter fragments of 423m5/6 and 405. The -210 to -27 3' promoter deletion (Figure 2.1) was generated by P C R amplification using the 810 construct as a template, and the D A N 7 (Table 2.1) and C D 5 (Hauffe et al . , 1993) primers. The P C R product was sequenced through the entire promoter fragment. The other 3' deletion construct was produced by sub-c loning the -210 to -52 4CL1 promoter sequences as a Hind I l l -Sac I fragment from construct 450 (Hauffe et a l . , 1993). Table 2.1: Oligonucleotides used in site-directed mutagenesis. Underlining indicates sequences which differ from wild type. A and v indicate hypo- and hyper-methylated nucleotides in vivo in parsley cell culture (Hauffe et. al., 1991). O l i g o R e g i o n , S i t e A d d e d O l i c r o S e a u e n c DAN 3 V f t 1, P v u I I 5 ' P - GATGAAGCAGCAGCTGATGGCGAGA-3' T DAN 7 f t 2+3, X b a l A A v v v 5'P-GTGTTAGGTGATGAAGCCTAAATCTAGACTAAGATGAAGCAGC-3' DAN 5 f t 4 , N c o l A v v 5'P-GAATTAAATTATGAGTGGCCATGGTGAGAAAAATATTG-3' DAN 6 f t 5 + 6, M l u l v w A w 5'P-GAATATGAGGTTTGGTTACGCGTAAAATCCCGAATTAAATTATGAG-3 DAN 4 TATA p r o x . , S a c I 5 1 -TGAGCTCAATGTGGAGGGTTTG-3 ' Mutated promoter fragments of the 810, 423 and 405 ser ies were sub-c loned into the vector p R T - 9 9 - G U S - J D / K o z a k (JDD) (Schu lze-Lefert et a l . , 1989) as Hind Ill-Bam HI fragments, in order to produce fusions with the B-glucuronidase (GUS) reporter gene (Jefferson et a l . , 1987) Hind lll-Eco Rl fragments containing each p romoter /GUS fusion were then c loned into the binary vector BIN 19 (Bevan, 1984). Creat ion of the 3'-deletion constructs required an addit ional step. In order to 810 Series (227 base-pair fragment) Footprint -210 1 2 3 4 5 6 +17, G U S 8io ) m — m WMUM T A T A 810m1 810m23 810m4 810m56 810m1,23 I I I n r 810m1,4 C 810m1,56 I M 1 BH i"'T 810m23,4 I m kt.ii m 810m23,561 m I I I aa — ' Internally Deleted Series -210 -174 -120 423 Figure 2.2. Experimental constructs for tobacco transformation. Solid boxes represent in vivo footprints. Lightened boxes indicate a mutation introduced within the footprint. obtain desi rable restriction site arrangements, the 4CL1 promoter f ragments and the C a M V 35S promoter fragments were sub-c loned into p S L 1180 (Pharmacia), before sub-cloning into J D D . At each step of sub-c lon ing , putative posi t ives were subjected to d iagnost ic restr ict ion d igest ion or P C R ampl i f icat ion. 2.2 Production and Sampling of Transgenic Plants BIN 19 p lasmids containing promoter /GUS fusions of interest were introduced into Agrobacterium tumefaciens strain LBA4404 by electroporation using the B io -Rad Gene P u l s e r ^ M according to the manufacturer 's instructions for Agrobacterium t ransformation. Leaf d iscs of SR1 tobacco were transformed, and plants regenerated according to standard methods (Horsch et al . , 1988). Kanamycin and Carbenic i l l in concentrat ions of 100 mg/ml and 300 mg/ml, respect ively, were used in select ion. P C R amplif ication products from plant D N A minipreps (Edwards et a l . , 1991) were subjected to d iagnost ic restriction digests to verify the genotype of at least one plant t ransformed with each construct. Most plants were grown in greenhouses, under natural light. The inf luence of f luctuating natural light levels on t ransgene express ion was examined as descr ibed in Sect ion 3.4. Sampl ing for f luorometric ana lys is was performed on e leven primary transformants for each construct. Plants were sampled while f lowering. Sampl ing of a particular plant was normally accompl ished during a single sampl ing sess ion , in order to avoid the effects of wound-induction upon the 4CL promoter. Figure 2.2 il lustrates the t issues sampled . T i ssues inc luded: petal col lar, anthers, st igma, ovary, leaf vein and minimally ve inous leaf t issue. In initial sampl ing, sepa ls were also taken. Three floral samp les and two vegetative samples were taken for each t issue. Al l f lowers were between developmental stages 9 and 11 as descr ibed by Kol tunow et. al. (Koltunow et al . , 1990), or stages 4 and 5 as descr ibed by Reinold et. al. (1993). Thus, petals had started to become pink, but the st igma was no more than visible and anthers had not yet deh isced. Samp led leaves were between 10 and 20 cm in length and still growing. The developmental age of each flower or leaf taken from a plant was ranked for subsequent ana lyses. All samples were frozen immediately after exc is ion, and stored at -80C until assayed . T i plants used in analysis of root expression were grown axenical ly on Murashige and Skoog (MS) medium (G ibco /BRL, Bethesda, MD.) supplemented with 100 ug/ml Kanamycin and 0.8% agar. Seeds were steri l ized by a one minute wash with 70% ethanol. Plants were grown for 30 days, and roots were physically separated from as much agar as possib le. 2.3 Wounding A s with plants used to examine expression in roots, T i plants to be wounded were grown axenically on M S medium (Gibco/BRL) supplemented with 100 ug/ml Kanamycin and 0.8% agar. Plants were grown for 30 days; typically developing four true leaves and two coty ledons. Leaves and cotyledons were injured by pressing firmly between the arms of serrated tweezers. Les ions were produced approximately every 3 mm such that about 50 per cent of each leaf was marred in this way. Initial assays , while not conc lus ive, suggested that 48 hours between wounding and harvest al lowed maximum response with respect to G U S expression. This period was used in the reported a s s a y s . Harvest ing was accompl ished by cutting the stem just under the coty ledons. 2.4 Analysis of GUS Expression in Transgenic Plants Fluorometr ic analys is of G U S reporter express ion in plant samp les was accompl ished according to standard methods (Gal lagher, 1992; Jefferson et a l . , 1987). Plant samples were ground in 400 ul of extraction buffer: 50mM N a H P 0 4 (pH 7), 1 mM EDTA, 0 .1% Triton X-100, 10mM DTT (freshly added). Grinding was carried out in an 1.5 ml microcentri fuge tube containing a smal l amount of sand , using a hand-held dril l. Extracts were centr i fuged for ten minutes at approximately 10 OOOg, and 250 microL of supernatant transferred to a new tube. The protein concentrat ion of each extract was determined according to the Bradford assay , using a B IO-Rad Protein Assay kit. G U S activity was assayed in extraction buffer modified by the addit ion of 4-methylumbell i feryl B-D-glucuronide (MUG) to a 1 mM concentrat ion. Methanol was added to a final proportion of 20 per cent of reaction vo lume. Y ie ld of the f luorescent compound, 7-hydroxy-4-methylumbel l i ferone (MU) was determined using the L S - 5 0 spectrof luoromoter (Perk in-E lmer , Beacons f ie ld , England) . Typ ica l ly , 100 uL of plant extract was assayed in a reaction volume of 1 mL. 100 uL of a reaction would be stopped in 900 uL of 0.2M N a 2 C 0 3 at time zero and one or more subsequent time-points. Excitation and emiss ion wavelengths were 365 and 443 nm, respectively. A number of initial character izat ions were performed in order to verify that activity was within a linear range under the assay condit ions used. G U S concentrat ions yielding M U to approximately 50 000 pmoles /min /mg protein provided approximately l inear results for periods of up to a few hours. For lower G U S concentrat ions, reactions were observed to be approximately linear over eight hours, but decreased by an average of 20 per cent over a 16 hour assay period. Whi le protein concentrat ions decl ined substant ial ly in extracts refr igerated overnight, G U S activity decl ined minimally. It was a lso noted that refrigeration of stopped assay samples prior to reading substant ial ly increased background f luorescence. A s a result of these observat ions , protein concentrat ions were determined immediate ly after extraction, and G U S assays were normally performed the next day. Stat ist ical analys is of reporter express ion is d i scussed in Sect ion 3.6. H is tochemica l local izat ion of G U S activity was performed as descr ibed by Hauffe et. al. (1993) for hand-sect ioned samples . Only leaf and s t igma samp les were typically c leared with ethanol prior to o b s e r v a t i o n . 2.5 In Vivo Examination of Protein-DNA Interactions Except for the experiments using deoxy-cholate to examine complex composi t ion, organ-speci f ic nuclear extracts were generated accord ing to the nuclear protein miniprep method (Sablowski , 1994, and by personal communicat ion). All procedures were carr ied out at 4 C . Approximately 100 mg of fresh t issue was typically used as a source of nuclei . For leaf vein t issue, more starting material was required. In each case , t issue was ground in a 1.5 ml microcentrifuge tube containing sand and 400 ul of pH 6.5 nuclear extraction buffer (NEB) : twelve per cent hexylene glycol, 20 mM P I P E S - K O H , 10 mM MgC l2 , 0.25 per cent Triton X-100 and freshly added 0.5 mM P M S F , 0.3 ml/L 2-mercaptoethanol and 5 ug/ml leupeptin. The ground slurry was centr i fuged (500xg for five minutes) through a mini-fi ltration dev ice into 500 ml N E B containing 30 per cent Percol l (Sigma) volume/volume. The mini-filtration device cons is ted of microfuge tube with the bottom cut off, p laced into a whole tube, with two layers of membrane between them. The upper layer had a 100 micron pore s ize, while the bottom had 20 micron. After centrifugation, the Percol l cushion was careful ly removed. The remaining nuclei were re-suspended in 50 ul of pH 7.5 storage buffer (STB): 25 mM Hepes, 40 mM K C L , 0.5 mM EDTA, 2 0 % glycerol and freshly added 5 mM DTT and 5 ug/ml leupeptin. Nuclei were then lysed by adding one tenth volume of saturated ammonium sulphate, pH 7.5, for a 30 minute incubation. The suspension was then subjected to five minutes of centrifugation at approximately 10 OOOxg, and the supernatant retained. In speci f ied c a s e s , proteins from the supernatant were precipitated by 30 minute incubation with from 0.4 to 1.0 volume of saturated ammon ium sulphate. After five minutes of centr i fugation (10 OOOxg), the supernatant was de-sal ted by three minutes centri fugation (500xg) through five cm of Sephadex G-25 in S T B . Nuclear extracts were either used immediately, or stored at -80C . Ge l retardation assays were performed as descr ibed by Hauffe et. al. (1993). Nuclear extracts used in the deoxy-cholate experiments were prepared by S . Lee, as described (Hauffe et. al., 1993). Lic\. 50 2.6 Addit ional Experimentation Thin- layer chromatography was performed on s i l ica gel , accord ing to Markham et. al. (1989). Solvent mixes of acetone-chloroform-water (8:2:0.5) and ethyl acetate-methanol -water (63:12:9) were used. Samples for T L C analysis were generated by incubating d issec ted floral t issues in methanol overnight. Fol lowing gr inding, insoluble components were pelleted for two minutes at approximately 10000xg, and the supernatant retained. F luorescence microscopy utilized a Ze iss Ax ioscope ( Car l Ze i ss , Oberkochen, Germany) to examine hand-sect ioned material under exitation of 365 nm to 425 nm. 51 3. Extension of Previous Observations on Parsley 4CL1/GUS Expression in Tobacco. 3.1 Introduction A s d iscussed in Sect ion 1.3.4, fragments of the parsley 4CL1 promoter from -597 to +17 and -210 to +17 direct an express ion pattern very similar to endogenous tobacco 4 C L genes (Hauffe et. al., 1991; Reino ld et. al., 1993). Further deletions, either from the 5' end, or internally, abo l ished var ious aspects of h is tochemical ly detectable express ion and quantitatively decrease express ion in tobacco seedl ings (Hauffe et. al., 1991, 1993). One major limitation of previous studies in this a rea is the difficulty in dist inguishing quantitative and quali tat ive changes in gene expression. For example, if a promoter deletion abo l ishes detectable express ion in the petal, but not in the ovary, is this because the deleted sequences are of specif ic importance to petal express ion, or because expression in the ovary is simply higher? Such analys is is further compl icated by the fact that a cis -element could play both t issue-speci f ic and genera l , quantitative roles. A s a prelude to further studies defining the roles of particular elements in expression directed by the 4CL1 promoter, I wished to establ ish a quantitative understanding of the relative levels of reporter express ion in different tobacco organs. The smal ler of the two promoter f ragments previously shown to direct appropriate express ion was se lected as being more amenable to further d issect ion. This fragment (-210 to +17, construct 810, Hauffe et. al., 1991) directs reporter gene express ion in xylem cel ls of veins throughout t ransgenic tobacco, in sub-apical root cel ls and root hairs, and in a number of f loral cel l - types: in epidermal cel ls of the st igma, ovules, p lacenta and the coloured portion of the petal, in the anthers and pollen cel ls , in ovary wal ls and in nectar ies. In addit ion, two internally deleted promoter constructs were included in the initial character izat ion. These constructs (constructs 405 and 423, Hauffe et, al., 1993) were se lec ted because of their relatively high expression and inclusion of a domain which appears to be important for interactions between the 4CL1 promoter and nuclear proteins (Hauffe et a l . , 1993) Another experiment included in this chapter add resses the effect of seasona l f luctuations in light levels on transgene express ion. Other exper iments attempt to identify phenylpropanoid compounds correlated with reporter express ion in the nectar ies. 3.2 Quantitative Fluorometric Analysis of Reporter E x p r e s s i o n Resu l ts of quantitat ive f luorometric ana lys is of reporter express ion in several tobacco organs and t issues are shown in Figure 3.1. Except where noted, each point represents a separate transformant and was derived from the average of multiple samples for each transformant. Resul ts are depicted using a logarithmic sca le , because of the high degree of variation between transformants. G U S activity was observed to be highest in the anthers, ovaries and st igmas of plants t ransformed with each of the three promoter/reporter fus ions, indicating the highest reporter express ion in these organs. G U S activity in plants t ransformed with the promoter less reporter construct was 53 Figure 3.1 Reporter gene expression in different tissues of transgenic tobacco. The structure of each construct transformed into plants is indicated in A). Each panel in B) represents one independent transformant and is averaged from multiple samples per transformant. An horizontal bar indicates the median level of expression directed by each construct. highest in anthers and roots. In roots, this background value represents a high proportion of activity directed by 810. The median level of expression directed by each promoter construct could consistent ly be ranked, with 810 highest and 423 lowest. Furthermore, while levels of expression varied by t issue or organ, the relative express ion levels directed by the three constructs appeared to be similar in each t issue/organ. For example, in severa l t i ssues/organs, the median express ion level directed by 810 is c lose to 10-fold higher than that directed by 405 (Figure 3.1). In leaf vein and st igma, where relative express ion levels of 810 and 405 are farthest from this va lue, it may be in large part due to the small sample s izes and uneven distributions of express ion data. The expression levels directed by 423 a lso appear to be conserved relative to the other constructs. One except ion to this trend is observed, in the st igma. In this organ, express ion levels directed by 423 are more similar to levels directed by 810 or 405 than in other organs. The relative ranking of individual transformants are shown for severa l t issues/organs in Figure 3.2. Each individual transformant is given a number based on its rank in expression level relative to other plants transformed with the same construct. The same number is uti l ized to present the rank of that transformant in other organs. Express ion rank in one organ appears to be a moderately good predictor of rank in each other organ/t issue. Summar ized quantitatively, rank in one organ provides a predictive accuracy for each other organ, ranging from 0.9 to 3.5 (data not shown in tabulated form), depending on the construct examined and the particular combination of organs examined. For example, the rank of an 810 transformant in vein expression predicts its rank in anther expression with an average error of 810 = ZJT A T A GUS 405 = C llj TATA [ii GUS 423 c | "c "3 o Q. U) .£ 3 E • o E Q. o (0 10 5 10* 10 3 102 101 10° 5 6 8 g 7 10 11 1 ?1 9 V 11 10 2 7 4 6 8 10 1 9 6 11 2 5 7 10 a) i- n » 5 O >, a. 810 N D a> > a> 1 2 1 4 i .2 6 3 l 4 6 » ° 7 8 | 1 1 3 91011 7 1 0 4 , 6 3 1 5 110 8 2 9 1 6 10 7 a 11 c < ? g » O >. o. a) 405 N D 10 11 > 4 3 r '10 11 N D 423 1 2 6 4 3 5 11 10 i l l O >» 0. 5) In cases where the data set is close to background, rankings are not included. Figure 3.2 Organ-specificity of position effects in 4CL1/GUS transgenic tobacco. Each individual transformant was designated by a distinct number, for each of the three constructs. For example "4" in the 810 panel refers to the same transformant in all five organs. "ND" indicates cases where the data set is close to background levels and rankings are thus not included. The structure of each construct is indicated at top. approximately one rank posit ion. In a random distribution, the average predictive error would be approximately 3.5, the maximum observed in these data. For all three constructs, the ranking of each t issue/organ was least predictive of express ion in the st igma. In almost all other cases , ranking displayed moderate to high predictive accuracy. The highest predictive accurac ies were all observed in 423 transformants. In anther-ovary, anther-st igma or ovary-st igma compar i sons , the predictive error of a transformant's express ion ranking was less than 1.0. 3.3 Effect of Light Level Fluctuation on 4CL1 Regulation Because of the role of 4 C L in the production of l ight-induced f lavonoids, and because transformed plants would be grown under natural light, it seemed worthwhile to address the effect of seasona l variation in light levels on 4CL1/GUS expression in tobacco. Ten plants which showed substant ial levels of G U S activity in initial sampl ing were cut back and al lowed to shoot and flower again. Plants were initially sampled in the summer, under long days/high light levels. Plants were re-sampled two to four months later, during lower natural light regimes. Multiple samples were col lected and assayed as per initial sampl ing . Reporter express ion in initial/high light samp les and later/low light samples were compared (data not shown). There was no dist inguishable trends toward either reduced or increased G U S activity in any of the six organs/ t issues examined at the two t imes. Whi le light levels may play a quantitative role in 4CL1/GUS regulation, any Figure 3.3 Extensions of previous hi stochemical observations of 4CL1/GUS expression in tobacco: a) longitudinal sect ion through petal and nectary in plane of sketch; b) c r o s s - s e c t i o n through base of ovary (plane B in "d)"), c) longitudinal sect ion through unfer t i l ized st igma; d) sketch of longitudinally sectioned tobacco f lower, "p" indicates petal; "n", nectary. contribution appears to be relatively smal l , or saturated at low light l e v e l s . 3.5 Extension of Previous Histochemical Observations While the expression patterns of 4CL1 /reporter constructs have previously been descr ibed (Hauffe et. al., 1991; Reinold et. al., 19 93), observat ions in the course of this thesis research provided some addit ional insights. In contrast to the finding of Reinold et. al. (1991) that reporter expression was not observed in the base of the petal, this phenotype was observed in numerous sections (e.g. Figure 3.3a). In order to address the relative levels of G U S expression in coloured and white port ions of the petal. The petals of five t ransgenic plants were d issected into coloured and white portions and subjected to f luorometric analys is . In each case , levels of G U S activity in the white portion were well above background, approximately one quarter of the level observed in the pink portion of the same flower (data not shown). Thus , two measures of reporter express ion show signif icant levels of express ion in the white tube of tobacco petals. Resul ts from thin-layer chromatography (Figure 3.4a) suggest the presence of f lavonoids in both the white and pink portions of the petal. An addit ional site where reporter express ion was frequently detected histochemical ly was at the junction between ovary wall and petal (Figure 3.3b). This expression appeared to be highest near where petal and ovary separated (data not shown). In the st igma, express ion had previously been reported in the papil lae and some sub-epidermal cel ls (Reinold et. al., 1991). In numerous sect ions examined in this invest igat ion, express ion was frequently observed to extend well into Figure 3.4. Phenylpropanoid products in the tobacco flowers. A) Thin-layer chromatography of dissected floral organs in organic solvent mix, fumed with NH3 under UV: rutin, standard; n, nectary, o, ovary except nectary, w, white portion of petal; p, pink portion of petal. B) Longitudinally-sectioned pistil under UV: p, petal; n, nectary; o, ovules; v, vasculature. the conduct ing t issue of unferti l ized st igmas (Figure 3.3c). Interestingly, h is tochemica l staining for G U S corre lated wel l spat ia l ly with photosynthet ical ly active ce l ls . Reporter express ion in the ovar ies and ovary walls appeared to follow an axial progression in development , with staining occurr ing most proximal to the style in youngest ovar ies (data not shown). Cumulat ive observat ions during numerous h is tochemical sect ions examined suggest an organ-specif ic bias in the wound response regulated by 4CL1. Sect ions from the leaf, the pedicel of the flower and the petal d isplayed a high frequency of local ized staining at exc is ion s i tes. The ovary was never observed to exhibit wound-speci f ic reporter express ion, and the frequency was very low in the st igma (data not shown). Cel l -speci f ic i ty a lso appeared to vary, with epidermal cel ls being the most likely to display wound-related express ion . 3.6 4CL Expression in the Nectaries His tochemica l detect ion of reporter express ion indicated high levels of 4CL 7-directed express ion in the nectary. A s previous studies had not local ized expression of endogenous 4CL genes beyond the level of the ovary, an effort was made to determine if nectary-speci f ic express ion was representative of endogenous patterns of 4CL express ion . Nectar ies were exc ised from tobacco ovaries and protein extracts from the two portions compared in western analys is using ant i -bodies to parsley 4 C L . Resul ts indicated that nectar ies contain high levels of 4 C L protein and suggest that 4 C L was more abundant in the nectaries than in other ovary tissue (D. Lee and D.N., data not shown). Thus 4CL 1 -directed reporter expression in the nectary correlates with expression of endogenous 4 C L . A s previous studies had local ized f lavonoids to the ovules of petunia (Ylistra et. al., 1994) and tobacco (Reinold and Douglas, unpubl ished results), nectary t issue was examined for f lavonoids. In agreement with Watanabe and Wender (1966), thin-layer chromatography fai led to detect f lavonoids in ovary extracts (Figure 3.4a). Similar ly, h istochemical staining in response to the f lavonoid-speci f ic reagent diphenyl boric acid 2-amino ester was low in the nectary, relative to petals (data not shown). F luorescence microscopy indicates that under UV-l ight, intense blue-green f luorescence is observed in the nectary (figure 3.4b), possibly from the accumulat ion of UV-f luorescent phenylpropanoid natural products. 3.7 D iscuss ion Comparisons with expression of endogenous 4CL It has previously been demonstrated that the parsley 4CL1 promoter directs reporter express ion which is quali tat ively s imi lar to the endogenous pattern of tobacco 4CL expression (Reinold et. al., 1993). The quantitative data reported here provide a measure of the levels of 4CL 1 -directed expression in discrete organs and t issues of t ransgenic tobacco. Observat ions here are general ly consistent with publ ished results examining relative levels of 4CL m R N A levels in the organs of tobacco. Results reported by Reinold et. al. (1993) and Lee and Douglas (1996) suggest that endogenous 4CL m R N A levels in tobacco f lowers are highest in ovar ies, then st igma, then petals (anthers were not examined) . In this study, relative express ion levels speci f ied by 4CL 7/reporter fusions in different floral organs fol lowed the same pattern, suggest ing that the relative organ-speci f ic express ion levels directed by the 4CL1 -promoter are largely appropriate. At least one quantitative difference is apparent between reporter express ion directed by the 810 and 405 constructs, and endogenous 4CL. While 4CL transcripts are several t imes more prevalent in the petal tube than in the collar (Lee, 1996), 4CL 7-directed G U S express ion was approximately four t imes higher in the col lar. Peta l -spec i f ic d i f fe rences in either t ranscr ipt ional act ivat ion or post - t ranscr ip t ional regulation of promoter/reporter fusions and endogenous genes might explain this d iscrepancy. If the d iscrepancy is due to dif ferences in t ranscr ipt ional act ivat ion, it might result either from the absence of pertinent c /s -e lements in the promoter/reporter fus ions, or funct ional dif ferences between the parsley and tobacco 4CL p romote rs . Interpreting variation between data sets One major considerat ion in quantitative compar isons of t ransgene express ion is the wide variation in express ion between individual transformants due to the influence of "position effect". The mechan isms by which position effect inf luences gene express ion are not well understood, but are generally assumed to relate to the chromosomal location of transgene insertion (Benfey and C h u a , 1989; Weis ing et a l . , 1988). Compar isons between different groups of t ransgenic plants are compl icated by the fact that few t ransgenic populat ions exhibit normal distributions of t ransgene express ion . Express ion levels of 4CL 7/reporter fusions are presented as logari thmic transformations in Figure 3.1. Fol lowing this transformation, most of the data sets presented satisfy a modif ied ch i -square test for normality (Campbel l , 1989; data not shown). However , compar isons between different data sets will util ize median rather mean values, as this is a more conservative measure (Campbel l , 1989 ) . The quantitative expression data reported in Figure 3.1 may be compared between constructs, or between organs/ t issues. The effect of posit ion effect on compar isons between constructs is maximal , because the compar ison is between independent transformants. This is consistent with the large variation observed in express ion directed by each construct. Posit ion effect can also influence the t issue-specif ic i ty of t ransgene express ion . Therefore, posit ion effect a lso has ramif icat ions for compar isons between organs/ t issues of the same transformants. However, data presented in Figure 3.2 suggests that express ion level in one t issue/organ is general ly a good predictor of express ion in other domains of 4CL7-directed express ion in the same plant. Thus , posit ion effect is much less signif icant for compar isons between organs/ t issues of the same 4CL t - t ransformants than in compar isons of independent transformants. Resu l ts from the animal literature appear to indicate that promoters are very variable in their level of autonomy from neighbour ing cis -elements. The Slh promoter of Drosophila is very resistant to changes in expression when enhancer elements are inserted nearby (Merli et a l . , 1996). Express ion patterns directed by other Drosophila promoters are typically altered by the same enhancer e lements (Ibid). The promoters of other genes, such as the chicken B-globin gene are l ikewise suscept ib le to ectopic express ion resulting from insertion c lose to heterologous regulatory elements (Chung et a l . , 1993). In the case of the 4CL1 promoter, the profile of cel l - types exhibi t ing h is tochemical ly detectable express ion is extremely consistent, suggest ing either a low frequency of host cis -e lements capab le of effecting transgene express ion, or a substantial level of autonomy for the 4CL1 promoter. This observation is supported by cons is tenc ies in the ranking of individual plants with respect to express ion level in different organs (Figure 3.2). There are clearly except ions to this pattern, however. Transformant 9 of the 810 construct d isp lays large di f ferences between relative express ion level in leaf vein and all floral t issues. Transformant 1 of the 405 construct presents dramatical ly different express ion levels in the s t igma relative to other floral t issues. Relat ive express ion levels were most consistent, by far, in the organs of plants transformed with 423. One poss ib le explanat ion is that the greater simplicity of the smal ler promoter fragment reduces the number of components which can interact with the inf luences which cause posit ion effect. Functional comparisons of the TATA-dista l promoter sequences of 810, 405 and 423. Median expression levels directed by the 810 and 423 constructs were the highest and lowest, respectively, for the three constructs, in all six organs/ t issues tested. This order of ranking correlates well with these constructs ' previously noted express ion levels in pars ley protoplasts and whole tobacco seedl ings (Hauffe et. al., 1993). Thus , 4CL1 sequences between -210 and -120 act as a stronger activator than sequences from -597 to -244, which in turn are a stronger activator than the -210 to -174 fragment, when each is paired with the -120 to +17 fragment. The signif icance of this observat ion depends on whether the activating capaci t ies of the upstream fragments of 405 and 423 are effected by their spatial posit ioning relative to the -120 to +17 fragment. In investigations of the c/'s-elements of a number of promoters, it has been shown that some elements function effectively in a range of spatial relat ionships with basal promoter e lements, while others require a particular posit ioning and orientation in order to function (Ptashne, 1988; Alberts et. al., 1994). Thus , the dramatic quantitative dif ferences in expression directed by 810 and 423 may result from the absence of sequences between -174 and -120, the altered posit ioning of sequences between -210 and -174, or both. The TATA-d is ta l sequences of 405 and 423 are clearly able to posit ively contribute to reporter gene express ion, as the isolated -120 to +17 fragment is not able to direct detectable expression (Hauffe et. al., 1991 ) . Interestingly, the relative express ion levels directed by the three constructs are similar in a large number of cases . Each construct directs highest levels of express ion in anther, ovary, and st igma; lower levels in leaf vein and petal; lowest levels in leaf blade. This suggests a high degree of functional redundancy between the three upstream regions tested in conjunction with the -120 to +17 fragment. Whi le quantitative activation from each upstream region var ied, the pattern of activation was very similar at the organ level. Two models would explain this result. One possibil i ty is that all sequences required to direct the observed organ-speci f ic express ion pattern reside between -120 and +17 and that each upstream fragment acts primarily quanti tat ively. Alternat ively, funct ional redundancy between s e q u e n c e s from -210 to -174 and -597 and -244 may be more complex, with each f ragment contr ibut ing quanti tat ively and qual i tat ively. Whi le express ion directed by the three constructs is qualitat ively simi lar at the organ-speci f ic level, a major difference is apparent at the level of cel l -speci f ic express ion. It has been noted that 423 directs ectopic express ion in phloem (Hauffe et. al., 1993). This suggests that a missing element or elements between -174 and -120 represses phloem express ion and is thus required for correct spatial regulation at the level of cel l- type, in the vasculature. The ectopic phloem express ion is much lower than expression in the xylem, explaining why the qualitat ive effect of the deletion is only dist inguishable at the cel lu lar l e v e l . Function of 4CL expression in floral organs With all three 4CL1/GUS fusions, expression was highest in the ovar ies, anthers and st igma. The activity of 4 C L in the flower has primarily been l inked to f lavonoid synthesis. F lavonoid synthesis in petals and anthers has been studied extensively (Stafford, 1990). St igmat ic f lavonoids have been l inked to the fertility of plant spec ies including petunia and maize (Taylor and Jorgensen, 1992). A lso in petunia, f lavonoids have been identified in the ovaries (Quattrochio et. al., 1993). Al l of these sites correlate with 4CL1/GUS express ion in t ransgenic tobacco f lowers. However, in tobacco at least, the strongest 4CL 1 -directed reporter express ion in ovaries is assoc ia ted with the nectar ies, with weaker expression in ovules (Hauffe et. al., 1991). This pattern of 4CL 7-directed reporter express ion suggests that 4 C L may participate in the synthesis of phenylpropanoid natural products other than f lavonoids in the ovary. Western analysis indicates that endogenous 4 C L is present at high levels in tobacco nectaries (D. Lee and D.N., unpublished results). This indicates that 4CL1/GUS expression in this t issue is representative of endogenous express ion patterns and supports the possibil i ty that pheny lpropanoid biosynthet ic activity is occurr ing at this si te. Th in-layer chromatography (Watanabe and Wender, 1965; Figure 3.4a, this thesis) suggests that f lavonoids are much less abundant in ovar ies than in the petal, in spite of the much higher 4CL expression in the ovar ies (Reinold et. al., 1993; Lee and Douglas, 1996). These results suggest a major role other than f lavonoid synthesis for 4 C L in the ovary. Consis tent with this hypothesis, reporter express ion directed by CHS promoters in petunia has been reported only at the surface of ovules and the surrounding placenta (Ylistra et. al., 1994). One possible phenylpropanoid end-product at this site would be coumar ins, previously identified in tobacco ovar ies by T L C , and consistent with the blue-green f luorescence noted in the nectaries (Figure 3.3 b). Whi le no direct demonstrat ion of the involvement of 4 C L in the synthesis of coumarins has been made, it has previously been suggested due to a strong correlat ion between fungal elicitation of furanocoumar in synthesis and the induction of 4 C L expression (Tietjen and Matern, 1983; Brown, 1985; Hahlbrock and Schee l , 1989). The anti-microbial activity frequently assoc ia ted with coumar ins might play a va luable role in floral nectar ies. Other possib le products which might be related to 4 C L activity in the nectar ies include hydroxycinnamoyl amides, and phenol ic ac id esters. A s d iscussed in section 1.2.3, both c lasses of molecules have been suggested as downstream products of 4 C L activity and hydroxycinnamoyl amides have been identified as the most abundant phenol ic consti tuents of tobacco ovaries (Cabanne et a l . , 1977). Es ters of c innamic ac ids are identifiable in T L C analys is of tobacco pistils (Watanabe and Wender, 1965). and might also account for the observed f luorescence in sect ioned ovar ies. Tobacco plants with dramatical ly reduced 4 C L levels have been generated through the use of ant i -sense technology (Lee, 1996). These plants may prove useful in evaluating the b iochemica l and biological s igni f icance of 4 C L activity in the nectar ies and other floral t i ssues. Express ion of 4CL1/GUS was observed to follow axial development in the ovary. Yl istra et. al. (1994) observed the same pattern in the ovules of petunia plants expressing a CHS /GUS fusion. Stain ing with diphenyl boric acid 2-amino ester suggests that in both tobacco and petunia, f lavonols occur on the surface of the ovules (Ylistra et. al., 1994; Reinold and Douglas, unpubl ished results). F lavonols thus seem to be a likely end-product of 4 C L activity at this site. Whi le the axial progression in 4CL1/GUS express ion presumably reflects developmenta l stage specif ici ty, it is not c lear what s igni f icance the timing of expression has. Ovule f lavonols do not appear to be essent ia l for ferti l ization, as petunia mutants devoid of detectable f lavonols in the ovary are fertile (R. Koes , personal commun ica t i on ) . The 4CL1 promoter also directs reporter gene expression at the junctions of petal and ovary (Figure 3.3b). Phenylpropanoid products in the anther have previously been implicated in tapetal breakdown (Wiermann, 1981). A s express ion at the petal-ovary junction dec l ines both above and below the point of their separat ion, it may be that downstream products of 4 C L activity are involved in breaking contact between the two groups of cel ls . 4CL1/GUS fusions tested in the course of this thesis were expressed in both the tube and collar of tobacco petals. This result shows that the spatial pattern of gene expression directed by the 4CL1 promoter in the petal is consistent with the endogenous pattern of 4CL express ion, in spite of a previous report to the contrary (Reinold et. al., 1993). Recent results suggest that expression of endogenous genes are much higher in the tube and base, than in the collar (Lee and Douglas, 1996). Cons ider ing that the opposite pattern of expression has been observed for tobacco C H S (Drews et. al., 1992), this would suggest that 4 C L expression in the base of the petal may again be associated with the b iosynthesis of non-f lavonoid products. 4. Characterization of the 5/6 Element 4.1 Introduction In previous investigations of the 4CL1 promoter, it was determined that sequences between -210 and -78 were necessary for h is tochemical ly detectable express ion in most organs/ t i ssues of tobacco (Hauffe et. al., 1991, 1993). Furthermore, sequences between -120 and -78 proved to be a major site of complex formation during interaction between tobacco nuclear proteins and the 4CL1 promoter (Hauffe et. al., 1993). Because of these observat ions, it was decided to focus a spec ia l effort on analyzing protein-DNA interactions in the -120 to -78 region, and on evaluating their role in regulating 4CL1-directed gene express ion. In parsley cell culture, two guanine repeats within this region ( -97 to -99 and -107 to -109) had been shown to be protected in vivo in two separate studies (Becker-Andre et. al., 1991; Hauffe et. al., 1991). Because the guanine repeats were only separated by 7 nucleot ides, the two footprints (designated F P 5 and F P 6) were mutagenized as a single unit, introducing a total of five t ransvers ions (Table 2.1). A s deletion analysis had previously suggested that upstream regions of the promoter were functionally redundant to the region containing F P 5/6 (Hauffe et. al., 1993), we chose to mutate F P 5/6 in each of three promoter contexts, the wild type 810 construct, and the internally deleted 405 and 423 constructs (Figure 4.1). The mutated promoter constructs were introduced into tobacco as G U S fusions and their express ion ana lyzed in direct compar ison with express ion directed by the unmutated versions of the same constructs 71 A) • 405 = C • 810 423 = ^ TATA ^ ^ TATA ^ O pBG GUS GUS GUS B) 10 5 central 5/11 median —•»»- — central 9/11 background Figure 4.1 Effects of mutating 5,6 on 4CL expression in leaf tissues of transgenic tobacco: A) Sequence changes introduced are indicated next to wild type sequences. B) Levels of expression directed by wild-type and mutant promoter sequences. Each point represents one independant transformant, and is averaged from multiple samples per transformant. A horizontal bar indicates a median value and a vertical bar represents the fold-decrease in median activity in the mutant transformants. (Chapter 3). In addition to in vivo analysis of the role of the putative element(s) in regulating the temporospat ial pattern of reporter gene exp ress ion , o rgan- and t issue-spec i f ic p ro te in -DNA interact ions within this region were investigated in vitro. Much of this in vitro research was in col laboration with Dr. Steven Lee and specif ied exper iments performed by Dr. Lee will be d iscussed in some detail in order to synthes ize as complete an understanding as possible of the s igni f icance of F P 5/6. 4.2 Quantitative Analysis of the In Vivo Significance of Footprints 5 and 6. The quantitative and qualitative effects of the introduced mutations in F P 5/6 were tested by compar ison with levels of G U S express ion directed by the unmutated constructs in six t i ssues/organs of t ransgenic tobacco (Figures 4.1 and 4.2). Five of these organs/ t issues were chosen as displaying high express ion levels at the direction of 4CL1 (Section 3.2). The sixth (leaf blade) was chosen in order to include one t issue where minimal reporter gene express ion is observed under direction of W T 4CL1 promoter sequences . Figures 4.1 and 4.2 indicate the level of G U S activity observed in tobacco plants transformed with each construct, in each of the six organs/ t issues tested. Each point represents the average activity from multiple samp les per organ/t issue of an individual transformant. For each construct, e leven independent transformants were tested. For each set of e leven expression values the median value is indicated by a horizontal bar. Levels of G U S activity observed in the organs of plants 73 GUS activity (pmol 4-MU/mg protein/min) rV > fD o < fi) 1 • < • — •tt x » i - i i 3 0 4* • • • • * u is-i tt | H rV M «9 •o fi) 5 3 D CO (5' 3 fi> !» ' H> • o 8 8 II I I t ransformed with a promoterless G U S construct (pBG) are indicated as grey boxes and a dashed black line. The black line indicates the median G U S activity in these plants. The dark and light grey boxes represent, respectively, the median five and nine G U S activities observed in the e leven p B G transformants tested (Section 3.2). One transformant containing the 405 construct mutated at F P 5/6 exceeded express ion levels of the next highest plant by approximately 1000-fold, and was d iscarded as an out-lier (not shown). It is apparent from the many instances of minimal overlap between express ion levels directed by the unmutated promoters and those mutated at F P 5/6 (m5/6) (Figures 4.1 and 4.2) that F P 5/6 plays an important quantitat ive role in deve lopmenta l ly - regula ted act ivat ion of reporter gene transcription by the 4CL1 promoter. In each organ invest igated, the mutation of F P 5/6 in the 810 construct results in a clear reduction in G U S expression levels. For the 405 and 423 constructs, mutating F P 5/6 resulted in a discernible decrease in express ion levels in every case where the level of express ion directed by the unmutated construct is above background. However in some cases involving each of the three constructs, the quantitative effect of mutating F P 5/6 is likely greater than indicated (Figures 4.1 and 4.2), as the median express ion level specif ied by the mutated construct was well within the range of background G U S activity present in plants t ransformed with the promoter less G U S construct. Resul ts from expression of 810 and 810m5/6 in the leaf vein strongly suggest that F P 5/6 is important in determining the level of 4CL 7-directed vascular express ion (Figure 4.1b). However, it is difficult to determine the absolute or relative reduction in the level of vein express ion for any of the three mutated constructs, because in each case mutant expression levels are completely or largely within background levels. This was also observed in leaf blade and petal express ion data (Figures 4.1, 4.2). In both leaf and petal, the apparent effect of mutating F P 5/6 is again largest in the 810 context. However, because of the low expression levels directed by the other constructs even when unmutated, it can not be ruled-out that mutating F P 5/6 caused a major reduction in expression directed by these constructs. Mutating F P 5/6 in the 810 context reduced median express ion levels by at least 70-fold in the leaf vein, and by at least 25-fold in the petal . In severa l floral organs the expression of the unmutated constructs was very high and even greater effects of mutating F P 5/6 were observed. In the case of the 810 construct, this decrease was greatest in the ovary. Median expression decreased approximately 800-fold in the ovary. F P 5/6 is also important to the level of ovary express ion directed by the 423 and 405 constructs. Upon mutation of F P 5/6 in the 405 context, median G U S expression in the ovary decl ined approximately 100-fold. G U S express ion directed by 423m5/6 fell to within background levels, suggest ing that the true effect of F P 5/6 mutation in this context may be greater than the 200-fold decrease in express ion which is apparent. While F P 5/6 appears to play an important role in ovary express ion directed by all three constructs, the role of F P 5/6 in anther and st igma expression appears be more dependent on sequence context. In the 810 context, the m5/6 mutation resulted in decreases in median G U S express ion of 140-fold in the anther, and 18-fold in the st igma. Thus, in the 810 promoter context, the st igma appears to be the organ in which F P 5/6 has the smal lest role. In contrast, mutating F P 5/6 in the 405 construct resulted in decreases in median G U S expression of 270-fold in the st igma, and only 8-fold in the anther. In the 423 construct, the m5/6 mutation decreased median G U S express ion by at least 200-fold in the st igma, and at least 36-fold in the anther. Fluorometric data from the roots of seedl ings grown on agar suggest that F P 5/6 is also important in root-specif ic express ion (data not shown). However, due to the high background expression and low level of 810-der ived express ion (Section 3.2) in roots, it was impossib le to quantitate the effect of the m5/6 mutation on root-spec i f i c exp ress ion . In summary, F P 5/6 was shown to play an important positive role in reporter expression directed by the 810 4CL1/GUS fusion in at least five organs of transgenic tobacco. Due to overlap with background G U S activity, the effects of mutating F P 5/6 could not be quantitatively d iscerned in some organs. However, the quantitative contribution of F P 5/6 to express ion in different tobacco organs was observed to vary. In the 810 promoter context, the m5/6 mutation had the greatest d iscernible effect in the ovary, and the smal lest effect in the st igma. In the 405 context, the largest discernible effect of the m5/6 mutation was in the st igma. 4.3 Histochemical Analysis of the Qualitative Role of Region 5,6. A s indicated in the previous sect ion, F P 5/6 plays an important quantitative role in transcriptional activation by the 4CL1 promoter in severa l organs of t ransgenic tobacco plants. In order to determine if mutation of F P 5/6 altered the pattern of t issue- and cel l -speci f ic express ion in these organs, transformants containing the mutated and unmutated vers ions of 810 and 405 were examined histochemical ly for G U S activity. Al l reported results reflect observat ions in multiple independent transformants. No dif ferences were observed in ce l l - and t issue-speci f ic i ty between 810 and 405 and express ion patterns d isp layed by the mutated versions of these two constructs were l ikewise indist inguishable from one another. Whi le 810 and 405 transformants displayed clear G U S express ion in the xylem cel ls of leaf vasculature, express ion directed by 810m5/6 or 405m5/6 at this site was very weak and rarely observed (not shown). In stems, where endogenous 4CL genes are expressed at high levels (Lee and Douglas, 1996), histochemical staining of the vasculature, while weak, was more discernible. Staining was observed in the primary xy lem of 810m5/6 and 405m5/6 transformants, but was strongest in ray parenchyma cells (Figure 4.3 a and b). Vascu lar staining was also clearly observed in the ovary and petal (Figure 4.3 e). Thus , the m5/6 mutation decreases 4CL 7-directed express ion in the vascu la ture, but does not appear to alter the cel l-specif ic i ty of express ion in the leaf or s tem. In the ovary, the effect of mutating F P 5/6 was clearly greatest in the nectary. While ovary expression directed by 810 and 405 are highest in the nectary (Figure 4.3 f), ovary expression directed by m5/6 mutants was highest at the petal-ovary junctions (Figure 4.3 e). Faint express ion was occasional ly observed in nectary cel ls and ovules of m5/6 mutants (not shown). Thus, all cel l-types encompassed in the ovary express ion pattern of the unmutated constructs were seen to exhibit G U S activity in m5/6 mutants. However , a ce l l - type-speci f ic role for F P 5/6 was suggested by the preferential loss of express ion in nectary cel ls of m5/6 plants. 7 8 • 1 g) n) Figure 4.3 Histochemical detection of expression directed by 810m56 and 405m56. Cross-sections of a) 810m56 stem, b) 405m56 stem; longitudinal sections of c) 810m56 stigma, d) 405m56 stigma; cross-sections of e) 810m56 ovary, f)405 ovary; g) 810m56 pollen, h) 405m56 pollen, n, nectary; p, petal. In the st igma, expression was observed in papil lae and in neighbouring sub-epidermal cel ls and conduct ing t issue of m5/6 plants (Figure 4.3 c and d). This pattern is identical to st igma express ion directed by 810 and 405 (Figure 3.3c; Hauffe et. al., 1991, 1993). St igma express ion directed by 810m56 appeared to be stronger than that directed by 405m5/6, in agreement with data from f luorometric analys is (Figure 4.2). Express ion in petals was faint in 810m56 and 405m56, but could occasional ly be d iscerned in a petal epidermis or at the base of a petal (not shown). While it was impossible to be certain that no changes in cel l-specif ici ty - had occurred in petals of m5/6 plants, none were evident. In pollen, staining was observed in a distinct fraction of pol len grains from plants transformed with mutated and unmutated constructs (Figure 4.3 g), potentially consistent with Mendel ian segregat ion of one or more t ransgenes from wild-type chromosomes. In plants transformed with the promoterless G U S gene (pBG), histochemical staining was not observed in any organs. This is most noteworthy in the pollen (Figure 4.3 h), where fluorometric analys is indicated that p B G plants display substantial ly higher G U S activity than untransformed plants (Sect ion 3.2). 4.4 The Role of FP 5/6 in Wound-Induced Expression. Because the 227 bp 4CL-1 promoter fragment has been shown to direct wound- induced reporter express ion in tobacco (El lard- lvy and Doug las , 1996) in addition to directing complex developmental regulat ion, we wished to determine the quantitative role of the 5,6 element in wound induction. Leaves of four-week-old seedl ings were wounded as descr ibed (Section 2.3) and harvested 48 hours later. The results of f luorometric analys is of G U S activity in those seedl ings is d isplayed in Table 4 .1. Each result is the average of two trials. Table 4.1 The role of FP 5/6 in wound-induction by 4CL1. The level of GUS activity in unwounded (U) and wounded (W) tobacco seedlings containing different promoter constructs. "I" indicates the fold induction observed following wounding. Each pair of data points represents an independent transgenic line. Construct U W I Construct U W I 91 507 5.5X 292 576 2 OX 810 17 703 47X 405 435 1178 2 7X 60 170 2 . 8X 203 795 3 9X 88 198 2 . 6X 1121 3381 3 OX 94 1005 11X 85 87 1 OX 428 1090 2 4X 8 . 6 44 5 . OX . 2.4 3 . 5 1 5X 810m5/6 8.5 39 4.5X 405m5/6 2 . 8 3.4 1 2X 3 . 8 94 22X 11 17 1 5X 4.0 20 5. OX 3.3 3.9 1 2X Wound inducibility was calculated as the ratio of express ion in wounded and unwounded seedl ing sets of the same line. The degree of wound- inducibi l i ty var ied substant ial ly in seedl ing l ines contain ing the 810 and 810m5/6 constructs, possibly due to position effect. Whi le induction of G U S expression varied substantially, the range and average wound-induct ion observed in 810 and 810m5/6 seedl ing sets were similar, suggest ing that F P 5/6 plays no signif icant role in wound-induction in the context of the 810 promoter. The effect of the m5/6 mutation on wound- induced expression is less clear when the mutation is in the 405 promoter context. Because of the low degree of wound-induction of GUS expression observed in the 405 seedl ing l ines, more l ines of both 405 and 405m5/6 would have to be examined in order to determine whether F P 5/6 is important to wound- induced gene express ion directed by 405. It does appear likely that both 405 and 405m56 are less wound-inducible than 810 and 810m56. This would suggests that sequences between -210 and -120 may contribute to wound-regulat ion of the 4CL1 p romoter . G U S levels in plants containing the mutated constructs were much lower than in plants containing the unmutated constructs. This suggests that F P 5/6 plays an important role in developmental ly-regulated 4CL/-directed expression in seedl ings, as was observed in mature plants. Interestingly, G U S expression levels in unwounded 405 seedl ings were typically several t imes higher than those in 810 seedl ings. This data suggests that sequences between -597 and -244 act as a stronger activator of expression in seedl ings than sequences between -210 and -120. In no other developmental context has the -597 to -244 fragment proven to be a stronger activator of express ion than the -210 to -174 fragment. 4.5 Sequences Involved in Protein Binding to the 5,6 Domain Ge l retardation analys is using nuclear extracts from tobacco shoots identif ied multiple prote in-DNA complexes involving the -210 to -27 fragment of the 4CL1 promoter (Hauffe et. al., 1993). Furthermore, exper iments using extracts from tobacco leaves, shoots and ovar ies showed that a 42 bp domain spanning -120 to -78 and containing F P 5/6 was required for formation of these complexes . Unpublished collaborative results of Dr. Steven Lee. In vitro footprint analysis of the -210 to -27 fragment of the 4CL1 promoter using tobacco shoot nuclear extracts produced a very c lear sequence-spec i f i c footprint containing in vivo footprints 5 and 6. The footprint appears to constitute a single protein-binding cis-element, as competit ion with a 31 bp, double-stranded ol igonucleot ide spanning from -120 to - 90 was shown to compete efficiently for footprint format ion, while o l igonucleot ides with mutat ions at either footprint 5 or 6 dramatical ly reduced compet i t iveness. The mutations in the ol igonucleot ide competitors were the same nucleotide changes .which were examined for effects on in vivo expression using the m56 vers ions of the 810, 405 and 423 constructs. Scann ing mutagenesis of the ol igonucleot ide showed that only approximately 15 nucleot ides, conta in ing in vivo footprints 5 and 6 , are required for the binding which es tab l i shes the in vitro 5/6 footprint. The in vitro footprinting experiments also resulted in a very unexpected observat ion, that in vitro F P 5/6 was correlated with an addit ional footprint from -55 to -85 on the top strand and -62 to -98 on the bottom strand. A weak in vivo footprint, designated footprint 7, has previously been noted at -78, within this region (Hauffe et. al, 1991; Becker -Andre et. al., 1991), in parsley cell culture. Compet i t ion with cold ol igonucleot ide corresponding to the wild type 5,6 D N A -binding domain removed the observed protection of the F P 7 region, while competit ion with the m6 ol igonucleotide does not. This data suggests that the protein(s) bound to 5,6 are able to facilitate the formation of addit ional prote in-DNA complexes . Partial characterization of FP 5/6-binding complexes. Although multiple sequence-spec i f i c comp lexes of differing mobility were formed in gel retardation assays using shoot nuclear extracts to retard the 4CL1 promoter, all of the observed complexes required the 5/6 element for formation and no competitor was able to compete for a sub-set of the complexes (Hauffe et. al., 1993; S . Lee , unpubl ished results). One possible explanation of these observat ions is that alternate complexes are formed, with each complex dependent on protein-protein interactions with a factor bound to 5/6. In this model , disrupt ion of protein-protein interact ions might reduce the concentrat ion of a sub-set of the observed complexes. In an attempt to dist inguish between protein-DNA complexes formed with the 4CL1 promoter, the non-ionic detergent deoxycholate was added to incubations containing the -210 to -27 fragment of 4CL1 and tobacco shoot nuclear extracts prepared by Dr. Lee. Deoxycholate has previously been uti l ized to disrupt protein-protein interactions (Baeuer le and Balt imore, 1988; Despres et a l . , 1995). Complexes formed in these react ions were subsequent ly ana lyzed in gel retardation assays . C h a n g e s in banding pattern resulted from the addition of deoxycholate in three repetit ions of this experiment. A representat ive autoradiogram is shown in Figure 4.4. At deoxycholate concentrat ions between 0 .035% and 0 .1%, complexes which formed in the absence of the detergent (C1-7) were diminished and at least two bands previously not observed became apparent (T1 and T2). The concentrations of T1 and T2 are reduced at deoxycholate concentrations of 0.25%. Thus, very low concentrat ions of detergent disrupt all of the 5/6-dependent comp lexes observed in the absence of the detergent. The novel complexes which form in the presence of deoxycholate may represent intermediates in the formation of complexes observed in the absence of deoxycholate. 84 I r e e 0 o b o e t 1 2 3 4 5 6 7 — C7 — 0 6 — C5 — C4 — T2 - C3 - C2 - — T l Cl probe Figure 4.4. tiffed of deoxycholate on 4CI. 1 - nucleoprotein complex formation. A n end-label led -210 to -27 fragment of the 4CL1 promoter was incubated with tobacco shoot nuclear proteins prepared by a macroprep method, and deoxycholate in a range of concentrations. Lane 1, free probe; lane 2, shoot nuclear protein, no deoxycholate; lanes 3 - 8 shoot protein with deoxycholate declining from .2 5% to .01%w/v. Solid arrows indicate retarded bands. C1-C7 indicate bands present in the absence of deoxycholate. T l and T2 indicate bands observed only in the presence of deoxycholate. deoxycholate concentration 85 Organ-specific complex formation with the 4CL1 promoter. To invest igate whether organ- / t issue-speci f ic levels of 4CL1-directed express ion could be correlated with the formation of part icular prote in-DNA complexes , nuclear extracts were made from the organs/ t issues which had been examined in the in vivo express ion studies, using a nuclear mini-prep method (Section 2.5). When extracts from these sources were incubated with the end- labeled -210 to -27 fragment of the 4CL1 promoter, organ-speci f ic patterns of complex formation were observed in gel retardation analysis (Figure 4.5). Quali tat ively, two main banding patterns were observed; that apparent in reactions with ovary and st igma extracts (lanes 6 and 10) and that present with all other extracts (lanes 2,4,8,12 and 14). These patterns were observed in a number of replicates. The major bands observed are designated M C 1 - 4 (the distinct numbering system is introduced to dist inguish nuclear extracts used in this exper iments from those descr ibed in the previous experiment, which were prepared by a different method). Bands MC1 and M C 2 , while specif ic to ovary and st igma samp les , are not substantial ly decreased by addition of unlabeled F P 5/6 oligonucleotide. Bands at M C 3 and M C 4 are dramatical ly reduced upon competit ion by 100-fold excess of un labeled F P 5/6 oligonucleotide (e.g. lane 6 vs. , lane 7; lane 8 vs. lane 9). One dist inction between the two patterns is the intensity of the major upper band (MC4), which is only strong in the ovary and st igma samples . At M C 3 , the greatest intensity of banding is slightly higher in the ovary and st igma lanes (6 and 10) than in other reactions (e.g. 4,8 and 10). 86 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 Figure 4.5. Tissue- /organ-specif ic retardation of the 4CL1 promoter. A n end-label led -210 to -27 fragment of the 4CL1 promoter was incubated with nuclear proteins from seven different tobacco tissues/ organs prepared by the nuclear miniprep method. Alternate samples were incubated with 100-fold excess of cold competitor corresponding to FP 5/6. I.ane 1, free probe; lanes 2 and 3, shoot protein +/- oligo 5/6; lanes 4 and 5, anther protein +/- oligo 5/6; lanes 6 and 7, ovary protein +/- oligo 5/6; lanes 8 and 9, petal protein +/- oligo 5/6; lanes 10 and 1 1, stigma protein +/- oligo 5/6; lanes 12 and 13, vein protein +/-oligo 5/6; lane 14, leaf protein +/- oligo 5/6. MC 1-4 indicate retarded bands observed in the presence of the nuclear miniprep extracts. The large arrow indicates the free probe. 8 7 1 2 3 4 5 6 7 8 9 Figure 4.6. Effect of specific and non-specific competition on the binding of ovary proteins to the 4CI. 1 promoter. A n end-label led - 210 to -27 fragment of the 4CI.1 promoter was incubated with tobacco ovary nuclear proteins prepared by the nuclear miniprep method and unlabelled non-specific competitor or unlabelled sequences from the 4CL1 promoter. Lane 1, free probe; lane 2, ovary extract, no competitor; lane 3, 250-fold salmon sperm DNA; lane 4, 2000-fold salmon sperm DNA; lane 5, 100-fold oligo 5/6; lane 6, 100-fold (-210 to -27); lane 7, 100-fold 5' fragment (-210 to -126); lane 8, 100-fold 3' fragment (-70 to +17); lane 9, 100-fold oligo FP 4. MC 1-4 indicate retarded bands observed in the presence of the nuclear miniprep extracts. The large arrow indicates the free probe. Complexes formed by the -210 to -27 4CL1 probe and ovary nuclear extracts were subjected to competit ion by different sequences from the 4CL1 promoter (Figure 4.6). Competit ion with a large molar excess of salmon sperm D N A reduced MC1-4 by similar degrees (lanes 3 and 4). Of 4CL1 sequences tested as competitors, only the F P 5/6 ol igonucleot ide and the unlabeled fragment corresponding to the probe were able to compete efficiently for M C 3 and M C 4 formation (lanes 5 and 6). Fragments 5' (lane 7) or 3' (lane 8) of F P 5/6 or an ol igonucleot ide corresponding to F P 4 were not effective competi tors for formation of these complexes. This result suggests that F P 5/6 is required for formation of complexes between tobacco ovary nuclear proteins and the 4CL1 promoter, as has been demonstrated for complexes with shoot proteins (Hauffe et. al., 1993; S . Lee, unpubl ished r e s u l t s ) . 4.6 D iscuss ion Resul ts from in vivo and in vitro study of F P 5/6 suggest that this sequence plays a critical quantitative role in developmental ly-regulated gene expression directed by the 4CL1 promoter, and suggest some insights into possible mechanisms by which this role is accompl i shed . Four types of assay contributed to information regarding the function of F P 5/6. Fluorometric analysis measured the quantitative contr ibutions of F P 5/6 to organ-speci f ic gene express ion of 4CL1/GUS, while histochemical analys is examined the contr ibutions of F P 5/6 to the cell-specif icity of 4CL1. D N a s e I footprinting ana lys is performed by Dr. Steven Lee and gel retardation assays performed by both Dr. Lee and myself provided a partial characterizat ion of the prote in-DNA complexes formed in a F P 5/6-dependent manner. FP 5/6 contributes to activation of gene expression in different tobacco organs. In vivo results strongly suggest that F P 5/6 contributes to the developmental regulation of gene expression by the 4CL1 promoter. In all five tobacco organs tested, substantial decreases in express ion resulted from mutation of F P 5/6. In the context of the cont iguous -210 to +17 promoter, decreases in median expression level ranged from approximately 20-fold, in the st igma, to several hundred-fold, in the ovary. The magnitude of these decreases is especial ly excit ing in the context of related literature. For example, disruption of the AC- I , A C -II, AC-I 11 or G-box elements of the bean PAL2 promoter caused no more than a few-fold dec rease in express ion in either stem or petal (Hatton et. al., 1995). In contrast, F P 5/6 appears critical to express ion directed by the 810 construct in ovary, anther and xylem. However, in at least some 810m56 transformants, express ion was still h is tochemical ly (and in anthers and ovary, f luorometrically) detected in these si tes. It is not clear whether this is due to incomplete abrogation of binding to F P 5/6, or is completely independent of this c/ 's-element. However, previous results make it c lear that F P 5/6 is insufficient to speci fy the patterns of express ion for which it is necessary , s ince 4CL1 sequences from -120 to +17 do not direct h is tochemical ly -detectable express ion (Hauffe et. al., 1991). The role of F P 5/6 in establ ishing the spatial pattern of 4CL1-directed expression appears to be complex. As d iscussed above, it is not c lear whether or not F P 5/6 is absolutely required for express ion in any region of tobacco. However, two lines of evidence indicate that the role of F P 5/6 is not uniform throughout the plant. First; the effect of mutating F P 5/6 was quantitatively different in different organs. Second ly ; within the nectary, F P 5/6 was much more important to nectary express ion than to expression at the petal-ovary junct ions. The role of F P 5/6 in achieving the observed pattern of 4CL 7 -d i rec ted express ion may be less like a binary switch acting in particular groups of ce l ls , than as a volume dial with a different range of amplif ication in each spatial domain of 4CL1 express ion. The effect of promoter context on FP 5/6 function. The effect of mutating F P 5/6 varied according to the promoter context in which it was mutated. Mutating F P 5/6 in the context of the 405 promoter resulted in a much greater reduction in express ion in the st igma, and a much smaller reduction in expression in the anther, relative to the effect of the m5/6 mutation in the 810 context. The 423 construct behaved similarly to 405 with respect to the effect of the m5/6 mutation on st igma expression, but not anther express ion. One model for the relatively smal l role of F P 5/6 in st igma express ion , in the 810 promoter context, would involve functional redundancy. An important aspect of this model is that the postulated functional ly-redundant elements do not act additively. If that were the case , the effects of mutating F P 5/6 should be roughly constant, as a relative effect. For example, the effects of mutating either of the distal S P 1 binding sites of the S V 4 0 early promoter were minor relative to the effects of mutating both (Courey and Tjian, 1992). This appeared to result from functional redundancy in these binding si tes. Accord ing to this model , sequences functioning in the 810 context, but either miss ing or non-functional in 405 and 423, would be functionally redundant with F P 5/6 specif ical ly in the st igma. In an intriguing correlat ion with this model , both 405 and 423 are missing sequences from -174 to -120. This region might harbour the hypothetical functional ly redundant element. Alternatively, it may be that F P 5/6 in the native 810 context is simply less important to st igma express ion. In this model, the role of F P 5/6 in st igma express ion directed by 405 and 423 would reflect promiscuous interactions result ing from the altered spatial relat ionships between c / s - e l e m e n t s in these internal ly-deleted constructs. It is likely that many transcript ion factors are able to participate in promiscuous interactions under some c i rcumstances (Ptashne, 1988). In plants, it was shown that random rearrangements of the viral C a M V 35S promoter were able to direct reporter gene expression in novel patterns not exhibited by wild type promoter sequences. These novel domains of express ion may have reflected promiscuous interactions between cognate f rans-ac t ing factors. Another instance where the role of F P 5/6 depended on promoter context was in the anther. In this organ, the effect of mutating F P 5/6 was smal lest in the 405 promoter context. This may reflect a functional ly redundant activator of anther express ion upstream of -210. Whi le the arrangement of c/s-elements in 810 is native and therefore presumably more representative of endogenous mechan isms of gene regulation, the different roles of F P 5/6 in other promoter contexts may contribute to an understanding of the mechan isms by which F P 5/6 acts. Resul ts here suggest two models for the different organ-spec i f ic roles of F P 5/6 in different promoter contexts. Al tered arrangements of 4CL1 c/'s-elements may permit proteins binding to F P 5/6 to contr ibute promiscuously to enhanced transcript ional act ivat ion (e.g. s t igma express ion in the 423 context). Alternatively, these results may reflect the presence of additional c/ 's-elements which are funct ional ly redundant with F P 5/6 in different, partially over lapping sub-sets of the 4CL1 expression pattern (e.g. -174 to -120 in the s t igma, upstream of -210 in the anther). Further experimentat ion will be required in order to dist inguish between these possibi l i t ies. Role of FP 5/6 in wound-induced gene expression. While mutation of F P 5/6 resulted in a greater than 90 per cent reduct ion developmental ly-regulated express ion in all organs examined , any role which F P 5/6 plays in modulating wound-inducible express ion appears to be very minor. A s with developmental activation of 4CL1/GUS express ion, the fold-induction of G U S express ion following wounding var ied substant ial ly between independent t ransformants, suggest ing the influence of position effect. The range of increases in reporter activity which were observed were similar between 810 and 810m5/6 transformants, in spite of much lower levels of activity in 810m5/6 plants. This result suggests that the integrity of F P 5/6 has little inf luence on wound- induced express ion in the 810 promoter context, while reducing developmenta l ly - responsive express ion in tobacco seed l ings . It is less c lear whether F P 5/6 contributes to the wound-respons iveness of the 405 promoter construct. While the increase in reporter activity fol lowing wounding was very smal l in the examined 405m5/6 l ines, the low levels of wound- induced G U S express ion in seed l ings t ransgenic for the unmutated 405 construct make it difficult to assert that F P 5/6 is important to wound- respons iveness directed by 4CL1 in this promoter context either. It a lso appears possible that the 405 construct is less wound-responsive than the 810 construct. To resolve either of these ambiguit ies will require the examinat ion of a greater number of transgenic l ines. However, given the clear ev idence for the participation of F P 5/6 in developmental regulation v ia the 4CL1 promoter, the cumulative data on wound-induced reporter express ion in 810, 810m5/6, 405 and 405m5/6 plants suggests that F P 5/6 is much less important to wound- induced regulation. An addit ional interesting observat ion resulted from the ana lys is of reporter expression in the unwounded seedl ings. In 5 organs of f lowering tobacco plants, the 810 construct had consistent ly directed higher median expression levels than did 405 (Section 4.2). In contrast, median express ion in the four-week-old seedl ings was severa l t imes higher in the 405 transformants, than in 810 transformants. This suggests that 4CL1 sequences upstream of -210 may have a previously unsuspected role in seedl ing express ion. FP 5/6 is a binding site for multiple protein-DNA c o m p l e x e s . At least four protein-DNA complexes which formed between the -210 to -27 4CL1 fragment and tobacco shoot nuclear extracts were shown to require intact F P 5/6 sequences (Hauffe et. al., 1993). When the -210 to -27 fragment was used as a probe to test the binding of nuclear proteins prepared by a mini-prep method from various tobacco organs / t i ssues , multiple F P 5/6-dependent 4CL / - n u c l e o p r o t e i n complexes were observed for each organ (Figure 4.5). Certain complexes appeared to be organ-specif ic, while others appeared to occur in multiple extracts, possib ly reflecting the presence of broadly-distr ibuted nuclear factors, or famil ies of nuclear factors. Cons ider ing the importance of F P 5/6 to 4CL / -d i rected gene express ion , the observed complexes would seem to be strong candidates for regulatory factors contributing to patterns of 4CL1 express ion in tobacco. O b s e r v e d organ-spec i f ic variat ion in complex- format ion may reflect organ-spec i f ic factors involved in achiev ing different levels of express ion . However, it is unlikely that the observed dif ferences in complex formation are sufficient to explain di f ferences in gene express ion between the organs/ t issues studied. Whi le two of the h ighest-expressing organs (ovary and stigma) exhibit a pattern of complex formation distinct from other organs, a third (anther) does not. Furthermore, complex formation was indist inguishable, at the level of this assay , using leaf blade extracts and extracts from a number of organs and t issues which exhibit higher levels of 4CL/-directed express ion ; including leaf vein. Thus, dif ferences between nuclear extracts which are not detectable with assay condit ions used here must be postulated in order to explain the complex spatial pattern of 4CL/-directed express ion . The observat ion that a single element is required for formation of a number of the different complexes observed raises quest ions as to the relat ionships between the complexes. How many polypept ides contribute to each complex? Do the different complexes share some sub-uni ts? If so , are some bands intermediates in the formation of other bands? The detergent deoxycholate has somet imes proven useful in address ing sub-unit composi t ion (Bauerle and Balt imore, 1988; Despres et. al., 1995). S ince any complexes including the labeled probe will generate autoradiographic s igna ls , disruption of mult imeric protein complexes with low concentrat ions of detergent can be informative if substi tuents retain DNA-b ind ing capaci ty. In the c a s e of complexes formed between shoot nuclear extracts and the 4CL1 promoter, at least two novel bands are observed at detergent concentrat ions over- lapping with the loss of native complexes . These bands may thus represent intermediates in the formation of larger, natural ly-occurr ing complexes . If so, the number of nuclear proteins involved in native complexes is likely larger than is apparent from the number of bands. At no concentration of detergent are any of the native complexes observed to increase in intensity. Thus, no native complex can be identified as an intermediate in the formation of another. The nature of the relat ionships between these complexes remains unclear. However , these results suggest that the compl icated interactions involved in formation of the observed 5,6-binding complexes are amenab le to further in vitro s t udy . A model for the mode of action of F P 5/6 ar ises from an unpub l i shed in vitro footprinting ana lys is of synerg is t ic interact ions between F P 5/6 and 7 within the 4CL1 promoter (Lee and Douglas, unpubl ished results). S ince formation of a DNase I footprint near in vivo F P 7 was signif icantly d iminished by competit ion with a F P 5/6 ol igonucleot ide, it seems likely that proteins binding to F P 5/6 facilitate protein binding to at least one other site in the promoter. In this model , binding to F P 5/6 would be functionally analogous to the action of M C M 1 (Bruhn et al. , 1992) or S R F (Grueneberg et al . , 1995), where binding to D N A facilitates recruitment of a second factor to a nearby element. While no targeted mutagenesis of F P 7 has been carried-out, it may be that F P 7 specif ies expression in one or more t issues/organs in which F P 5/6 up-regulates express ion. It seems noteworthy that the 5' end of the 5/6 in vitro footprint extends to within approximately five bp 3' of in vivo footprint 4 (Lee and Douglas, unpubl ished results). It is possible that F P 5/6 plays a similar role in stimulating binding to F P 4. Summary F P 5/6 appears to be a critical c /s-e lement for deve lopmenta l , but not wound- induced, activation of 4CL 7-regulated transcr ipt ion. The contribution of F P 5/6 to up-regulation of 4CL / / repor ter exp ress i on was observed to vary signif icantly in different organs, and in different promoter contexts. In vitro analysis of the binding of tobacco nuclear proteins to the 4CL1 promoter suggest that certain F P 5/6-dependent complexes vary by organ, suggest ing that the variable role of F P 5/6 in organ-spec i f ic express ion may reflect the part icipation of proteins binding to F P 5/6 in organ-speci f ic complexes. 5. Participation of Other In vivo Footprints in Regulation of Gene Expression by the 4CL1 Promoter. In t roduct ion Prev ious investigations indicate that there are severa l regions of the 4CL1 promoter which effect express ion in t ransgenic tobacco plants (Hauffe et. ai, 1991, 1993). In order to investigate the importance of in vivo footprints 1 to 4 in expression directed by the 4CL1 promoter, these putative c/s-e lements were mutated, as descr ibed in Chapter 2 (Figure 2.1). Three nucleotide changes were introduced into each footprint 1 and 4, separately. Footprints 2 and 3 were mutated as a single unit, with a total of five nucleotide changes introduced (Figure 2.1). The footprints were also mutated in severa l combinat ions (Figure 2.3). Mutated promoters were fused to the B-glucuronidase (GUS) reporter gene and expression of the reporter gene in t ransgenic tobacco plants was analyzed both histochemical ly and f l u o r o m e t r i c a l l y . In a previous study, 4CL1 promoter fragments which did not include TATA-prox imal sequences were found to be incapable of direct ing h is tochemical ly-detectable express ion in any t issues of t ransgenic tobacco except pollen (Hauffe et. al., 1993). In an effort to define TATA-prox ima l sequences required for express ion in other organs, sequences extending from -210 to -27 of the 4CL1 p romo te r were fused to the CaMV 35S 46 bp and 90 bp minimal promoters (Benfey and C h u a , 1990). A -210 to -52 4CL1 promoter fragment was a lso fused to the CaMV 90 bp promoter. Expression directed by these hybrid promoters was assayed as above. 98 Footprint 1 Levels of reporter gene expression directed by the F P 1 mutant and each of the double mutants which include mutant F P 1 are presented in Figure 5.1. Reporter expression was measured f luorometr ical ly in e leven independent transformants for each construct. For each organ of each transformant, the express ion level presented represents the average of 2-3 separate assays . In each of the organs examined, mutation of F P 1 resulted in about a 5- to 40-fold decrease in median level of reporter express ion, relative to levels directed by the unmutated 810 promoter. The observed decrease in expression was smallest in the anthers and largest in the vein and ovary. When comparison includes double mutants, it is c lear that the m l , m2/3 double mutant directed higher express ion levels than the m l mutant in all examined organs. This pattern was most distinctive in petal and st igma, where median express ion levels directed by the m l , m2/3 double mutant were higher than those directed by 810 (see also Figure 5.3). Median reporter express ion in m l , m4 double mutants was also higher than in m l single mutants, although not by as much as was observed in compar ison with m l , m2/3 mutants. In contrast with results for the other double mutants, reporter express ion in m l , m5/6 mutants was much lower than in m l single mutants, except in the vein and petal, where expression levels in F P 1 mutants were c lose to background. Reporter express ion was examined histochemical ly in at least three plants transformed with each m l single or double mutant 99 GUS activity (pmol 4-MU/mg protein/min) 53 2 • *4f • • • » 00 0 BOO 0 0 1 1 , 00 0 GD 9 Op© 0 | 00 dp op 98 © i 1 1 1 • 4 M M t 1 O O O dfocD O 1 o O | © © © | 1 jVn i 9 0 0 O I 0 I • i I 3 0 I 8 O d§ 00 ococo o tj>0b o <9 o o < • 9 djfbo GO O O 3 & 3 pa 3 construct. In each case , the ovary, st igma, pollen, leaf and stem were examined. In m l single mutant and m.1, m4 double mutant t ransformants, no consistent di f ferences in express ion pattern were observed relative to 810 transformants. Reporter express ion was histochemical ly observed in the nectar ies, ovary walls (Figure 5.2a), ovules, placental epidermis and pollen (data not shown). Although express ion in the leaf vasculature was c lose to background levels, h is tochemical staining was evident in xylem cel ls of the stem (Figure 5.2b) In m l , m2/3 double mutants, express ion appeared to include all cel l - types in the petal and ovary (Figure 5.2c). Express ion in cel ls which normally exhibit 4CL1/GUS expression tended to be highest. In the s t igma, normal express ion was observed at the stigmatic sur face, in addition to weak expression throughout (data not shown). In the leaf and s tem, detectable expression remained limited to the xylem (Figure 5.2d). In m l , m5/6 double mutants, h is tochemical ly-detectable express ion was very weak, but appeared to reflect wild type patterns of express ion (data not shown). Footpr ints 2/3 In spite of two efforts to produce plants transformed with the F P 2/3 mutant construct, no plants were recovered which could be demonstrated to p o s s e s s the intact promoter/reporter fus ion. Whi le fifteen kanamycin resistant plants were regenerated, 4CL1 promoter sequences amplif ied by P C R from several transformants fai led to d isplay a d iagnost ic restriction site included in the m2/3 mutation. Sequenc ing of one P C R product indicated that mutations were present in each F P 2 and F P 3, but that at least one additional mutation was 101 Figure 5.2 Histochemically detectible reporter expression directed by the 4CL1 promote after mutation of FP 1, 2/3, or 4. Sections a, b - mutant FP 1; c, d - mutant FPs 1,2/3; e, f - mutant FPs 2/3, 4; g - mutant 2/3, 5/6; h, i, j - mutant FP 4. Sections b, d, h, i are cross-sections; a, c, e, f, g, j are longitudinal sections, "n" indicates nectary, "rp" ray parenchyma, "p" phloem, "x" xylem."ip". internal phloem Figure 5.3 Histochemically detectible reporter expression directed by 4CL deletion constructs. Sections a, b, are (-210 to -27)/CaMV 90; c, d, e, are (-210 to -27)/CaMV 46. Sections a and c are cross-sections; sections b and d, longitudinal sections, "n" indicates nectary, "p" petal, "w" wound. present near F P 5, and that the mutations in F P 2 and F P 3 were not as engineered (data not shown). Double mutant constructs in which the m2/3 mutation was combined with mutated F P 1, F P 4 or F P 5/6 were successfu l ly expressed in plants. The integrity of the mutated promoter/reporter constructs was verif ied for each construct by diagnost ic digest of promoter sequences amplif ied by P C R from transformed plants. Thus , reliable express ion data address ing the function of F P 2/3 is only avai lable from double mutants. Reporter expression directed by each of the F P 2/3 double mutants is presented in Figure 5.3. For purposes of compar ison, data pertaining to related single mutants is included. In m l , m2/3 and m2/3, m4 double mutant transformants, expression is higher than in mutant m l or m4 transformants, in each organ examined. In the vein and petals, the dif ference between m l , m2/3 mutant plants and m l mutants is greater than that between m2/3, m4 and m4 mutants. In the anthers, the opposi te is true. Interestingly, m2/3, m5/6 mutants display very similar express ion levels to F P 5/6 mutants in all organs examined, with the poss ib le except ion of the vein. The pattern of h istochemical ly-detectable express ion in m l , m2/3 double mutant transformants was d iscussed above. In the flower, m2/3, m4 double mutants displayed a pattern of express ion similar to that of m l , m2/3 double mutants (Figure 5.2 e, f). In the leaf and stem, express ion was observed in the xy lem, with express ion at lower levels also evident in the phloem (data not shown, see F P 4 data). Express ion patterns for mutant F P 2/3, 5/6 transformants were difficult to determine because of much lower levels of express ion. However, express ion appeared to follow the pattern of the other F P 2/3 mutants, with faint, but extended expression evident in ovary (Figure 5.2g), petal, and st igma (not shown). Footprint 4 Reporter expression directed by the m4 mutant and each of the double mutants which include mutant m4 is presented in Figure 5.4. In each of the organs examined , mutation of F P 4 resulted in about a 30-to 300-fold decrease in median level of reporter express ion , relative to levels directed by the unmutated 810 promoter. The observed decrease in express ion was largest in the ovary. When transformants of the single mutant are compared to double mutants, it is clear that the m l , m4 and m2/3, m4 double mutants directed higher expression levels in all floral organs (petal data not shown). In all organs except the leaf, the median expression level directed by each of these double mutants was at least 20-fold higher than observed with the m4 single mutant. Whi le in most organs, the m2/3, m4 mutant plants showed equivalent or stronger express ion than m l , m4 mutant plants, the opposite was observed in the ovary. In contrast with results for the other double mutants, reporter express ion levels in m4, m5/6 mutants were similar to those of m4 single mutants. The h is tochemical ly-detectable reporter express ion pattern of m2/3, m4 mutants is d iscussed above. In plants transformed with constructs incorporating the m4 mutation, staining was often observed in the phloem (Figure 5.2h). Staining of the phloem was weaker than staining observed in the xy lem. In other respects, m4 single mutants d isp layed reporter express ion patterns similar to those directed by unmutated 4CL1 sequences (e.g. Figure 5.2i,j). While expression 105 GUS activity (pmol 4-MU/mg protein/min) < CO 5' • 0 Of O O O dfydCD O © © > cgBjfoo CD <5» j h t 4 <*> • © 3 0) directed by F P 4, 5/6 mutants was very weak, it appeared to follow the same pattern (data not shown). Unpubl ished experiments suggest that an ol igonucleot ide probe corresponding to F P 4 and flanking sequences is able to bind parsley nuclear proteins ( C . J . Douglas). An ol igonucleotide probe compris ing sequences from -144 to -120 of the 4CL1 promoter was util ized in gel retardation assays as a ligand for tobacco nuclear proteins (Figure 5.5). Extracts from several tobacco organs including ovary, petal, anther and st igma produced two distinct complexes. In some cases a third, fainter band was also observed (lane 9 and data not shown). In order to determine the sequence-speci f ic i ty of binding to the F P 4 o l igonucleot ide, the compet i t iveness of o l igonucleot ides conta in ing wild type and mutant F P 4 sequences were examined. The mutation was identical to the mutation introduced in the express ion exper iments above. Compet i t ion with a 100-fold molar excess of unlabel led ol igonucleot ide reduced the concentration of each observed complex severa l - fo ld , while the m4 ol igonucleot ide was a less effective competi tor for proteins present in ovary and petal extracts. Intriguingly, an ol igonucleot ide corresponding to F P 5/6 competed strongly for binding of tobacco ovary nuclear proteins to the F P 4 ol igonucleot ide, while an m5/6 was a much weaker competitor. Footpr ints 5/6 The effects of the F P 5/6 mutation on reporter express ion were examined in each of three promoter contexts, as d iscussed in Chapter 4. The mutation introduced in the 810 promoter context is here compared to double mutants in the same context (Figure 5.6). In most organs, 107 I r e e P r o b e 0 O O P P + + + I P 4 I P m 4 I P 4 + F P m 4 o o o + + + ( I p 5 / () I p m 5 / (> 3 4 2 1 0 2 7 ) 7 8 9 10 11 12 Figure 5.5. Retardation of a FP 4 oligonucleotide by tobacco nuclear extracts. A rad io labe led 25 base-pair oligonucleotide corresponding to 4CL1 sequences from -144 to -120 (containing FP 4) was incubated with nuclear proteins prepared from different tobacco floral organs by the nuclear miniprep method. Incubations were performed either without any specific competitor, or with a 100-fold excess of wi ld type or mutated sequence-specific competitor, o indicates ovary; p, petal; a, anther; s, stigma. OFP 4 and OFP 5/6 indicate oligonucleotide competitors corresponding to those footprints.m indicates the mutated version of the oligonucleotide competitor. FC 1-3 indicate retarded bands observed in the presence of tobacco nuclear proteins. The large arrow indicates the free probe. 108 m TI 8 £ •a 0 g cn CD CD CD CD O 5T o 0) =r. CQ 3 CD CQ G) Tt C "0 Co cn o> S? o _ < 5 " 8 0 3 CT 1 a. ~ o cn £ . —K ,—». O 3 a S> 3 3 » 3 o CD r-ca - 1 CD O £ »• s £ T3 CD CD g 2, * CD 5T CB • o CD 3 Q. o' 0) CD Q. CD CT O < CD GUS activity (pmql 4-MU/mg protein/min) o S o ^ o 4} • • » rV mm m b M M o e fv p o 4* o o o o S® 1ft co c |% ^ % • • <5 8o^ 3 $ 3 CD i r 1 • , 1 - 1 1 •• «>#• § P i o o 0^9° © 3 oi 1 o© * 3 Ul CD I» ro OJ © © © © OI < ro 5' > 3 O < m ca' 3 m express ion levels directed by the single mutant were similar to, or slightly lower than, express ion directed by the double mutants. Furthermore, in most organs, the expression levels directed by each double mutants were quite constant relative to each other; i.e. the F P 4, 5/6 double mutants showed slightly higher expression than the F P 1, 2/3 and F P 2/3, 4 double mutants. However, in the st igma, the F P 1, 5/6 mutants exhibited lower express ion than the single mutant or the other double mutants. The histochemical ly-detectable express ion patterns observed for each 5/6 mutant have been d iscussed above. 3' Deletions In order to determine the importance of TATA-prox ima l sequences to 4CL 7-regulated expression, a -210 to -27 4CL1 promoter fragment was fused to CaMV 35S 46 or 90 bp minimal promoters and then to G U S coding sequences (Figure 2.1). A third chimeric construct fused 4CL1 -210 to -52 with the CaMV 35S 90 bp minimal promoter and G U S . Reporter expression levels directed by these three chimeric promoters are depicted in Figure 5.7, along with expression levels directed by the unmutated -210 to +17 (810) 4CL1 promoter fragment. Each of the three chimeric promoters was able to direct express ion above background levels in multiple organs. However, expression directed by each of the chimeric promoters was lower than that directed by the 810 construct, in all organs examined. In each organ, the median express ion level directed by the (-210 to -21)1 CaMV 46 construct was weakest . Reporter gene expression levels directed by the (-210 to -27)1 CaMV 90 and (-210 to - 5 2 ) / C a W 90 constructs were a | £ » 0{3" <• =r CD CD r- c CD * 0 —• g-XI ® co ^ 3 ° ^ cn " O CD CD ° ^ Q . UJ =; CD -> <° -2 , CO / - . ( Q CD _ _ CO i.' =J 3 © co CQ 05 3 5 CD » 3 <D " ' c5" r « S ' S.co CD = • CD Q . cr O O CQ CD S -*• 3 CD n » O 2 « (£> CO 2-~ - cr® i»-a S 7 = 1 < B 3 3 i < if 3 I ^3 3 co 5 - Q . - 0 S.-a 9: O S »' §. | c ? i l £ Q . CO — ^vj ^ CD O g CD CO g.o 3 =• - 3- • ,2 < 0) CD CTCQ 0) O i o m < CD . o . 3 ) CD CO co O _ i ro i ^ - CT O I X 3 ® "O m 3 3 CD CO 2 D. — ® CD CD - r ; CD . CQ — CD CO CD O " X CO CD -a to CD cn ^ CD • -9 o 52, co cn c m O CO ro CD CD  CD CO 83 GUS activity (pmol 4-MU/mg protein/min) CD Q> < CD > 3 (D O < Q> 73 a 0) CQ" 3 0) I 1 I _L 9 DI Li fV • V 4 i it • 14 7 i 1 • •1 ! J L G D l | l | f | 1 1 1 m TI 8 £ 3" ^ TJ <" o cn 5' bo CD TJ —i CD cn CD 3 =3- O CD 3 JB 4^ 0) -* g CD ?:"§ >< ^ CD o -> O X 3 T3 CD =• cn B> cn 3 o ' !2, ^ l a 3 c SB 3^5-CD CQ CQ CD 3 g >< TJ CD CD OJ O cn CD OJ 3 3 Q. o' S. CD Q. OJ CT O < CD GUS activity (pmol 4-MU/mg proteln/mln) I cb o ^oo 00 OCD 4 % —ft — •4 •• V o • I T 4 3 » ' 5 e J I • o ' 4— to —i I * 1 •« ooo o coo oo o 0 o» 1 oo I o»o oo % m 4* • • • • ©o k r » • • • o s < i . 5 _L 1 ,oo ooofooojo oi o 85 oo »^ ©9 • jjj^ 0 00 p ST r 1 1 o or 00 © % • • • o so c lO 0 • 0 »• o J I L o oo 0>#» • • % 4 ^ 1 cn s I 3 0) 3 similar, al though the -210 to -27 fragment directed higher median levels of express ion in vein and st igma. Cel l -speci f ic i ty of reporter gene express ion directed by each of the chimeric promoters appeared to be similar to that directed by the 810 fragment (Figure 5.8 a, b, c, d). In plants transformed with each of three constructs, express ion was observed in appropriate cel l - types of s tem, pol len, ovary, petal and st igma. Non-speci f ic staining of t issue from plants t ransformed with the chimer ic promoters seemed to be more prevalent than in other plants (not shown), suggest ing that low levels of non-speci f ic expression may have been occurr ing. Sect ions from plants t ransformed with each chimer ic promoter construct d isp layed local ized staining at excis ion sites (Figure 5.8 e), suggest ing that c /s-e lements necessary for wound- induced express ion are present in all three constructs. D i s c u s s i o n Effects of mutating individual elements Analys is of reporter gene expression directed by mutated 4CL1 promoter sequences in t ransgenic tobacco suggests that in vivo footprints 1, 2/3, 4 and 5/6 each contribute to developmental regulation of attached coding sequences . The effects of each individual mutation (except 2/3) are summar ized in Figure 5.9. The mutation of F P 1, 4 or 5/6 caused at least a 5-fold decrease in the median level of reporter gene expression in each organ. Mutation of each element dec reased the median expression level in at least one organ by 25-fold or more, suggest ing that each of these footprints is a c / ' s -e lement Figure 5.9 Fold decrease in reporter expression in single mutants. Peaked bars indicated mutations where the median expression value directed by the mutated promoter is well within the background range, suggesting that the fold decrease may be even greater than indicated. which contr ibutes to posit ive developmental regulation of gene express ion by the 4CL1 promoter. In other investigations of promoter funct ion, it has similarly been observed that multiple c / ' s -e l emen ts within the same promoter can contribute to positive developmenta l regulation [e.g. (Crabtree et a l . , 1992; Hatton et a l . , 1995; Hatton et a l . , 1996)]. In the bean PAL 2 promoter, four c/'s-elements have been identif ied which contribute to posit ive regulation of reporter gene express ion in petals of t ransgenic plants (Hatton et. al., 1995). Interestingly, mutation of any one of the four elements resulted in an average reduction of only about three-fold. The 4CL1 F P 5/6 element appears to quantitatively contribute most to 4CL1 -directed regulat ion in each investigated organ except the st igma. However, as it is not known to what degree each mutation effects binding of cognate trans-factors in vivo, it is not certain that the relative roles of each element are qualitat ively reflected by reporter express ion data reported here. Furthermore, as mutation of any one of F P 1, 4, or 5/6 appeared to reduce reporter expression several-fold in a number of organs, it appears that all three elements are essent ial in order to maintain levels of express ion directed by the unmutated promoter. F P s 4 and 5/6 exhibit a similar organ bias in affecting express ion in the different organs examined. Each appears to be most important quantitatively in the ovaries and in the anthers; in that order. This could potentially reflect a common mechanism of act ion, such as shared participation in an activation complex. Tobacco shoot nuclear proteins binding to the 5/6 element provide DNase protection to a region extending to within approximately five nucleot ides from in vivo footprint 4 (S. Lee and C . Douglas, unpublished results), indicating that proteins binding to the two sites are in c lose proximity and thus have opportuni ty to interact. Whi le 4CL/-directed reporter expression is highest in the ovar ies, anthers and st igmas of t ransgenic tobacco plants, the introduced mutations in F P 4 and F P 5/6 have much less effect on express ion in the st igma than in the other two organs. This suggests that other elements in the 4CL1 promoter are likely important for express ion in the st igma. When the 5/6 element was mutated in the context of either of two internally deleted promoter constructs, larger dec reases in expression were observed (Chapter 4), support ing this hypothesis and suggest ing that additional sequences important to s t igma express ion are located between -210 and -120. While e lements 4 and 5/6 d isplay similar specif ic i t ies at the organ level , distinct di f ferences were noted in ce l l -spec i f ic express ion patterns in transformants of the mutant constructs. Within the ovary wal l , the mutant 5/6 construct was noted to reduce express ion in nectary cel ls to a much greater degree than in parenchyma cel ls at ovary/petal junctions. There was no observable bias in which of these cel l - types was effected by mutation of F P 4. In the xy lem, mutation of F P 4 resulted in ectopic reporter expression in phloem cel ls. Thus, F P 4 appears to act as a negative regulator in one vascular cel l- type, while acting as a positive regulator of xylem express ion. Mutation of 5/6 was not observed to result in phloem express ion. However, the possibil i ty that mutation of 5/6 derepresses phloem expression can not be ruled-out because vascular expression in 5/6 mutants may have been too low to observe this phenotype. FP 4 functions similarly to putative MYB-binding sites in the bean PAL 2 promoter. Footprint 4 has been previously noted as a candidate c/ 's-e lement to part icipate in the co-ordinate developmental regulation of phenylpropanoid genes (Douglas, 1996; Logemann et al . , 1995; Sablowsk i et a l . , 1994). F P 4 and similar sequences present in the promoters of a number of phenylpropanoid genes match the consensus binding site of a maize MYB-protein (P) (Grotewald et. al., 1994). One such sequence in the bean P A L 2 promoter (the AC-I 11 element) has been reported to bind a MYB- l i ke protein present in tobacco petal nuclear extracts (Sablowski et. al., 1993). F P 4 of the parsley 4CL1 promoter d isplays greater sequence homology (8/8 bp in the P-box binding consensus sequence) to the AC-II element of bean PAL 2. Furthermore, there is evidence of functional homology between AC-II and F P 4. Mutation of AC-II results in ectopic reporter express ion in phloem cel ls of t ransgenic tobacco plants (Leyva et. al., 1992; Hatton et. al., 1995). Earl ier results indicated that a deletion spanning 4CL1 F P 4 causes ectopic express ion in tobacco phloem cel ls (Hauffe et. al., 1993), while results presented here indicate that this phenotype speci f ical ly results from alteration of sequences present in F P 4. Mutation of AC-II has also been noted to decrease reporter expression in the petals and xylem (Hatton et. al., 1995), as is evident upon mutation of F P 4 (Figure 5.4). Thus , the effects of mutating F P 4 are consistent with a model in which F P 4 of the parsley 4CL1 promoter and AC-II of the bean P A L 2 promoter serve c losely related functions, possibly in mediat ing the co-ordinate regulation of phenylpropanoid gene express ion , potentially through the binding of identical or related MYB- l i ke transcript ion factors. One observat ion makes the apparent conservat ion of function between AC-II and F P 4 surprising. Mutation of the half of AC-II most-c losely related to 4 C L footprint 4 has very little effect on express ion directed by P A L 2 (Hatton et a l . , 1995). Tobacco nuclear proteins which were observed here to bind to a rad io labe led F P 4 ol igonucleotide represent a candidate or candidates for cognate transcription factors. The reduced affinity of the observed protein(s) for an ol igonucleotide containing the m4 mutation supports the sequence-speci f ic i ty with which these proteins bind to F P 4. Interestingly, while the F P 4 ol igonucleotide did not compete effectively for binding to the 5/6 element (S. Lee, D. N., unpubl ished results), the ol igonucleotide corresponding to F P 5/6 appears able to compete for binding to F P 4 sequences. A 7/8 match for the F P 4 P-box is present in F P 5/6 (-98 to -91). One possible explanation is that the F P 5/6 sequences have greater affinity for the observed nuclear proteins. However, we have been unable to show sequence-spec i f ic binding to 4CL1 promoter sequences when F P 5/6 is mutated (S. Lee, D. N., data not shown), suggest ing that F P 4 is not effective at binding tobacco nuclear proteins in the context of other promoter sequences . Taken as a whole, these observations indicate that F P 4 and F P 5/6 are each important to developmental regulation of 4CL1 and suggest the possibi l i ty that the two elements function co-operat ively. Further experimentat ion will be required in order to explore the latter possibi l i ty and to determine whether the observed F P 4/nucleoprotein complexes are truly specif ic to F P 4 and thus represent strong candidates for in vivo mediators of F P 4's role in 4 C L 7-re la ted e x p r e s s i o n . FP 2/3 acts as a repressor in ovary and petals. While no transgenic plants with the 4CL1 promoter mutated at F P 2/3 alone were produced, plants containing three different double mutants containing an altered F P 2/3 were analyzed. Of these, the m l , m2/3 and m2/3, m4 transformants exhibited strong enough reporter express ion to exhibit c lear patterns of express ion. In plant of both types, cel l - type specif ic i ty of express ion was partially lost and expanded expression was evident in ovaries, petals, and to a lesser degree, st igmas. Ectopic reporter expression was clear in parenchyma throughout the ovary and petal. Expression also appeared to be occurr ing in all ovary epidermal cel ls . Express ion in m2/3, m5/6 double mutants was faint, but appeared to occupy a similar range of ce l l -t ypes . In the stem and leaf, reporter expression directed by each of the nri2/3 double mutants was only histochemical ly detectable in xylem and ray parenchyma cel ls . This is the same pattern of expression directed by the unmutated promoter. Thus, based on the expression patterns directed by m l , m2/3 and m2/3, m4 double mutant promoters, F P 2/3 appears to act as a negative regulator of expression in parenchyma and epidermal cel ls of the ovary, petal and st igma. It remains a possibil i ty that mutation of F P 2/3 derepresses reporter expression in leaf and stem parenchyma and epidermal cel ls, but that the express ion remains below detectable levels. This possibil i ty seems especia l ly val id because F P 1, 2/3 and 2/3, 4 double mutants displayed higher expression in leaf veins than the F P 1 and F P 4 single mutants did (Figure 5.3). Comparisons between single and double mutants. Many eukaryot ic transcript ion factors part icipate in multi-protein comp lexes assemb led v ia protein-protein interact ions (Ptashne, 1988; Alberts et. al., 1994). One method of identifying putative si tes of such interact ions between transcription factors is to look for synerg ism in the function of c /s-e lements. For example, the T A T A -proximal and distal SP1 binding sites of the S V 4 0 early promoter are only weak activators when present individually, but together provide st rong t ranscr ipt ional act ivat ion. This reflects synerg is t ic interactions between bound SP1 molecules (Courey and Tjian, 1992). In hepatocyte-speci f ic regulation by the mouse albumin promoter, interactions between the HNF-I and N F - Y transcription factors were suggested when mutation of the HNF-I binding site only disrupted transcription if the neighbouring N F - Y binding site was a lso mutated (Crabtree et. al., 1992). Potential s i tes of physical interaction between proteins bound to 4CL1 c /s-elements were revealed by the expression patterns of promoter mutants descr ibed in this thesis. In m l , m2/3 and m2/3, m4 double promoter mutants, reporter expression was higher than for the respect ive m l or m4 single mutants, in all organs. This observat ion is consistent with a possible role for F P 2/3 as a negative regulatory element. However, in most organs, expression levels directed by the m2/3, m5/6 double mutant were very similar to levels directed by the m5/6 single mutant (Figure 5.3). This could be explained by postulating that the activity of F P 2/3 as a negative regulatory element is largely dependent on the integrity of F P 5/6. According to this model , F P 2/3-binding factors would require protein-protein interact ions with factors binding F P 5/6, either to stably bind F P 2/3, or in order to function as a t ranscr ip t iona l rep ressor . It a lso appears that the m5/6 mutation limits the effect of the m l and m4 mutations on 4CL/-directed expression in most organs. Express ion levels in the vein and petal are difficult to interpret because expression levels directed by each construct except the m4, m5/6 double mutant were similar to the range of background G U S activity (Figure 5.6). However, in the anther, ovary and st igma, median express ion levels directed by the m5/6 construct were higher than background G U S activities. In each of these organs the m4, m5/6 construct directed median express ion levels slightly higher than those directed by the m5/6 single mutant promoter construct. This contrasts sharply with the observed effect of the m4 mutation when F P 5/6 is not mutated. In each of those three organs, the m4 mutation resulted in a decrease in 4CL/-directed reporter gene expression (Figure 5.4). In the ovary, the decrease in median expression of G U S was greater than 300-fold. W h e n , median expression directed by the m4, m5/6 construct in anther, ovary and st igma is contrasted with that directed by the m5/6 construct, there is no further decrease due to m4. This result strongly suggests that the function of F P 4 in these organs requires that F P 5/6 remain intact. This conclusion adds support to the hypothesis stated earl ier (Section 4.6) that nuclear proteins binding to F P 4 and F P 5/6 interact in a functionally significant manner. Because the effects of the single m l mutation on 4CL/-directed reporter express ion were less dramatic than those of m4 (Figure 5.3 vs. Figure 5.4) it is more difficult to judge whether m l exacerbates the effect of the m5/6 mutation when the two are paired in the m l , m5/6 double mutation. However, in each organ except the st igma, the m l , m5/6 double promoter mutant directs express ion equivalent to or higher than the m5/6 promoter construct. This suggests that m l , too, requires an intact F P 5/6 in order to effect 4CL 7-directed exp ress i on (at least in organs other than the stigma). Thus, all three investigated elements appear to require an intact F P 5/6 in order to contribute to expression in the context of the 4CL1 promoter. This ev idence supports a critical role for the F P 5/6 element in 4CL1 exp ress ion . The roles of F P 1 and F P 4 also appear to be connected to each other. W h e n mutated individually, m l and m4 each resulted in a decrease in expression in all organs (Figures 5.1, 5.4). However, in the anther, ovary and st igma, median expression levels directed by the m l , m4 double mutant were approximately 20- to 100-fold higher than those directed by the m4 single mutant (Figure 5.4). Thus, mutating F P 1 has opposite effects, depending on whether or not F P 4 is also mutated. This result is not readily consistent with a model in which F P 4 function is simply dependent on F P 1 function, as that model would predict that express ion levels directed by the m l , m4 double mutant would be similar to those directed by the m4 single mutant. One possib le explanation is that a repressor of 4CL7-directed g e n e express ion requires the function of at least one of F P 1 or F P 4. In this model , disruption of both F P 1 and F P 4 prevents function of this repressor, and results in an increase in expression relative to express ion levels directed by m4 single mutant transformants. F P 2/3-binding proteins would be an obvious candidate for the postulated repressor. F P 2/3 is located between F P 1 and F P 4, and has been demonstrated to act as a repressor in 1m, m2/3 and m2/3, m4 double promoter mutants. However, both double mutants of F P 2/3 exhibited ectopic floral expression which was not observed in m l , m4 double mutants. A more complex explanation would therefore be required to explain the involvement of F P 2/3 in the m l , m4 double mutant phenotype. An alternative model for the differences between m4 and m l , m4 expression levels is that cognate binding proteins of F P 1 and F P 4 interact in a manner that alters their functional roles. In the case of regulat ion of transcript ional activation by the mammal ian plfG element, it has been determined that the glucocort icoid receptor (GR) acts as a repressor in the absence of AP-1 [Yamamoto, 1992 #172]. In the presence of A P - 1 , the factors interact with the result that G R functions as a positive regulator. Thus, mutation of the plfG GR-b ind ing site has opposite effects, depending on the presence or absence of AP-1 (Yamamoto et a l . , 1992). Accord ing to this model, a factor or factors binding to F P 4 would transform an F P 1-binding factor(s) from a repressor into an activator. If this were the case , the decrease in express ion observed in m4 transformants might result from this switch. Upon subsequent additional mutation of F P 1, repression would be removed. Addit ional research will be required to determine if either of these models is consistent with the mechanism by which m l effects F P 4 function in vivo. Cons ide red together, experiments examining express ion directed by vers ions of the 4CL1 promoter mutated at different combinat ions of c/ 's-elements suggest a large number of functional interactions between these elements. It has recently been shown that the C O P 9 regulatory complex of Arab idops is consis ts of at least twelve sub-units and conta ins transcript ion factors important to l ight-regulated gene express ion (Chamovitz et. al., 1996). The simplest model to explain the mechan ism by which the integrity of F P 5/6 effects the function of each F P 1, F P 2/3 and F P 4 would postulate a complex of nuclear proteins which spans from at least -195 (FP 1) to -96 (FP 5/6) of the 4CL1 promoter. Proteins binding to F P 5/6 might play a critical role in facil i tating the binding of other proteins or in interaction with the basa l transcript ional apparatus or other factors. An experiment showing that F P 5/6 is essent ial for the binding of nuclear proteins to sequences between F P 5/6 and -55 (S. Lee and C. Douglas, unpublished results; d i scussed in sect ion 4.5) provides biochemical ev idence that F P 5/6-binding proteins do participate in a smal ler protein complex. Thus if the larger, postulated complex would likely encompass sequences from -195 to -55 of the 4CL1 promoter. 3' Deletion Mutants All three chimer ic promoter constructs which were tested were able to direct appropriate patterns of cel l -speci f ic express ion , in each t issue/organ examined. This suggests that sequences between -210 and -52 of the 4CL1 promoter possess all c/ 's-elements necessary to direct the qualitative pattern of reporter gene express ion directed by larger fragments. The -210 to -52 fragment had previously been tested as a fusion to the CaMV 46 minimal promoter and had only directed detectable express ion in pollen (Hauffe et. al., 1993). The difference between that observat ion and those reported here is consistent with the contrast between expression directed by the (-210 to -27)1 CaMV 90/GUS and (-210 to -27)1 CaMV 46/GUS constructs (Figure 5.8). In most organs, the CaMV 90 fusion directed expression that was severa l -fold higher than that directed by the CaMV 46 fusion, without an apparent dif ference in cel l -specif ic i ty. Thus , the levels of express ion directed by the (-210 to -52)/CaMV 46/GUS (Hauffe et. al., 1993)would likely be several- fold lower than observed for the (-210 to -52)/CaMV 90IGUS fusion tested here (Figure 5.8) and would likely not be h i s t o c h e m i c a l l y - d e t e c t a b l e . It is interesting to note that median reporter express ion directed by the (-210 to -21)1 CaMV 46 hybrid promoter is 20- to 200-fold lower than that directed by the -210 to +17 4CL1 fragment, even though all sequences between -210 and the 4CL1 TATA-box are present in the former construct. Either of two alternate explanat ions would explain this observat ion. Approximately one additional turn of the D N A is introduced between upstream c/'s-elements and the TATA-box in the CaMV 46 fusion. Spat ial relationships between c/'s-elements are often important in determining activation (e.g. Ptashne, 1988) and might be signif icantly altered in the CaMV 46 fusion. Alternatively, sequences 3' of T A T A in the two constructs might be responsible for the quantitative dif ference in express ion levels directed by the two constructs. Hatton et. al. (1995) have previously suggested that 5' untranslated sequences of the bean PAL 2 gene may quantitatively contribute to the regulation of an attached reporter gene. Summary Taken together, experiments d iscussed in this chapter establ ish roles in 4CL1 regulation for a number of sequences which had been previously been shown to be footprinted in parsley cel ls and were thus cons idered candidate c/'s-elements (Hauffe et. al., 1991; Becker -Andre et. al., 1991). Mutation of each of the investigated in vivo footprints effected developmental regulation by the 4CL1 promoter in several organs of t ransgenic plants. In addition to quantitative effects at the organ level, two of the elements (FP 2/3 and 4) were also shown to effect patterns of cel l- type specif ici ty when mutated. Exci t ingly, the m5/6 mutation appeared to limit the effects of mutating each F P 1, F P 2/3, and F P 4. This result supports a model of protein-protein interactions between cognate binding factors of these e lements. Another experiment suggests that all sequences necessary for the qualitative pattern of expression directed by 4CL1 in tobacco are present between -210 and -52, though qualitatively important sequences may extend 3' of the 4CL1 TATA-box . Conclusions and Future Research Resul ts presented in this thesis address the mechan isms by which the parsley 4CL1 promoter directs developmenta l ly - regula ted gene express ion in transgenic tobacco. Quantitative ana lyses of the roles of four putative c/ 's-elements indicate that all four s e q u e n c e s contribute to regulation of reporter gene expression and that some of these elements play complex roles. A number of results point to interact ions between c/'s-elements within the 4CL1 promoter; which would likely be mediated by protein-protein interactions between f fans-act ing factors that bind to these elements. A number of experiments can be suggested which might develop the observat ions reported in this thesis. The construct ion of careful ly des igned promoter constructs might elucidate some of the interactions involved in 4CL-directed express ion . For example, different combinat ions of 4CL1 c/'s-elements could be fused to a minimal promoter to determine which elements are sufficient for part icular aspec ts of express ion and whether the synergist ic effects observed in this thesis depend only on the identified elements. In order to address the possibi l i ty of protein-protein interact ions that contribute signif icantly to 4CL7-directed exp ress ion , it will a lmost certainly prove necessary to clone cognate binding factors of the identif ied e lements. A southwestern screening approach might prove fruitful, but might be limited by the inferred protein-protein interactions, if they are important to DNA binding. Another approach would be to attempt to biochemical ly purify and character ize some of the factors involved. Resu l ts from gel-shift exper iments indicate that some of the tobacco nuclear proteins binding to 4CL1 promoter 127 -the single mutant was not successfully tested -the 1,2/3 and 2/3,4 double mutants suggest that this element is a negative regulator and in the flower at least, is important in cell-specificity -in the FP 5/6 background, this mutation has very little effect T A T A 2 3 - weakest activator of tested elements -there appears to be a complex functional interaction between this element and FP4, in planta. 5 6 • T -not tested • -quantitatively the most important element tested -mutation reduces 4CL1-directed expression to near background in several organs/tissues -displays some cell-specificty in ovary -likely binds multiple proteins -required for binding to FP 7 in vitro -required for function of FP 2/3 in planta -putative MYB-binding site -appears to be an important activator in several organs -represses phloem expression while promoting xylem expression -the FP 4 oligonucleotide is bound by nuclear proteins of unclear specificity -appears to influence the role of FP 1, in planta Figure 6.1 Schematic summary of some observations regarding the role of each tested element. sequences are stable and present at high concentrat ion. Furthermore, this approach has recently been employed successfu l ly in character iz ing a number of nuclear factors participating in a light-dependent regulatory complex (Chamovi tz et. al., 1996). An alternate approach to identifying transcription factors regulating 4CL 7 -d i rec ted gene express ion would exploit growing evidence that MYB- re la ted transcript ion factors part icipate in the regulation of pheny lpropanoid genes (e.g. Sablowsk i et. al., 1993; Douglas, 1996). Current exper iments in Dr. Douglas ' lab are directed toward the cloning of MYB- l i ke proteins putatively participating in 4CL1 regulation. Yet another strategy would exploit the co-ordinate regulation of phenylpropanoid genes to look for regulators of 4CL. Many mutants in the regulation of f lavonoid synthesis have been identified (reviewed in Holton and Corn ish , 1995). S o m e of these are likely to participate in the regulation of 4CL genes (e.g. An 10, Koes et. al., 1993), while others might be expected to belong to shared fami l ies of transcript ion factors. If factors which bind the 4CL1 promoter in vivo are successfu l ly identified, a number of experiments could be suggested. Reducing in vivo levels of each transcription factor by sense or ant i -sense suppress ion might be used to determine the in vivo s igni f icance of each factor. Co -exp ress ion experiments in yeast and plant cel l culture might also be employed to investigate the role of each factor in activating or repressing expression from the 4CL1 and other phenylpropanoid promoters. Co-precipi tat ion and yeast two-hybrid exper iments could be used to invest igate protein-protein interact ions between identif ied f r ans -ac t i ng fac tors . Another potentially valuable extension of this thesis would be a r igorous investigation of phenylpropanoid metabol ism in the nectary. Whi le developmenta l ly-d i rected 4CL gene expression is very strong in tobacco nectar ies, it is highly uncertain what natural products result and with what biological consequence . Chemica l methods would likely be effective in elucidating which phenylpropanoid compounds are most prevalent in nectar ies. Exist ing tobacco ant i -sense transformants with greatly reduced 4 C L activity (Lee, 1996) might al low identif ication of downst ream products of 4 C L activity in this t issue and ultimately al low an identif ication of assoc ia ted biological funct ion(s). B i b l i o g r a p h y Alberts, B., Bray, D., Lewis, J . , Raff, M., Roberts, K., and Watson, J . D. (1994). Molecular Biology of the Cell, 3rd Edition (Hamden, CT: Garland Publishing). Amati, B., and Land, H. (1994). Myc-Max-Mad: a transcription factor network controlling cell cycle progression, differentiation and death. Curr. Op. Genet. Dev. 4, 102-108. Aoyama, T., Dong, C.-H., Wu, Y., Carabelli, M., Sessa, G., Ruberti, I., Morelli, G., and Chua, N.-H. (1995). Ectopic expression of the Arabidopsis transcriptional activator athb-1 alters leaf fate in tobacco. Plant Cell 7, 1773-1785. Arnosti, D. N., Barolo, S., Levine, M., and Small, S. (1996). The eve stripe 2 enhancer employs multiple modes of transcriptional synergy. Development 122, 205-214. Ayer, D. E., Kretzner, L., and Eisenman, R. N. (1993). Mad: a heterodimeric partner for Max that antagonizes Myc transcriptional activity. Cell 72, 211-219. Baeuerle, P. A., and Baltimore, D. (1988). IkB: A specific inhibitor of the NF-kB transcription factor. Science 242, 540-546. Bate, N. J . , Orr, J . , Ni, W., Meromi, A., Nadler-Hassar, T., Doerner, P. W., Dixon, R. A., Lamb, C. J . , and Elkind, Y. (1994). Quantitative relationship between phenylalanine ammonia-lyase levels and phenylpropanoid accumulation in transgenic tobacco identifies a rate-determining step in natural product synthesis. Proc. Nat. Acad. Sci. USA 91, 7608-7 6 1 2 . Beardsley, T. (1991). Smart genes. Sci. Am., 86-95. I ! 130 Becker-Andre, M., Schulze-Lefert, P., and Hahlbrock, K. (1991). Structural comparison, modes of expression and putative c/s-acting elements of the two 4-coumarate:CoA ligase i genes in potato. J . Biol. Chem. 266, 8551-8559. Benfey, P. B., and Chua, N.-H. (1989). Regulated genes in transgenic plants. Science 244, 174-181. Benfey, P. N., and Chua, N.-H. (1990). The Cauliflower Mosaic Virus 35S promoter: combinatorial regulation of transcription in plants. Science 250, 959-966. Benfey, P. N., Ren, L, and Chua, N.-JH. (1990). Tissue-specific expression from CaMV 35S I enhancer subdomains in early stages of plant development. EMBO J. 9, 1677-1684. ! Bevan, M. (1984). Binary Agrobacterium tumefaciens vectors for plant transformation. Nuc. Acids Res. 12, 8711-8721. ! i i Block, A., Dangl, J . L, Hahlbrock, K.i, and Schulze-Lefert, P. (1990). Functional borders, i genetic fine structure, and distance requirements of cis elements mediating light responsiveness of the parsley chalcone synthase promoter. Proc. Nat. Acad. Sci. USA 87, 5387-5391. Boulikas, T. (1994). A compilation and classification of DNA binding sites for protein i transcription factors from vertebrates.' Crit. Rev. Euk. Gene Expres. 4, 117-321. i Branden, C , and Tooze, J . (1991). Introduction to Protein Structure (New York: Garland Publishing, Inc.). Brown, S. A. (1985). Recent advances in the biosynthesis of coumarins. In The biochemistry of plant phenolics, C. I. vSumere and P. J . Lea, eds. (Oxford: Clarendon Press). Bruhn, L., Hwang-Shum, J . J . , and Sprague, G. F. J . (1992). The N-terminal 96 residues of MCM1, a regulator of cell type-specific genes in Saccharomyces cerevisiae, are i suffucuent for DNA binding, transcription activation, and interaction with alpha 1. Mol. Cel l . Biol. 12, 3563-72. , | Bustos, M. M., Begum, D., Kalkan, F. |A., Battraw, M. J . , and Hall, T. C. (1991). Positive and negative cis -acting DNA domains arej required for spatial and temporal regulation of gene I expression by a seed storage protein promoter. EMBO J . 10, 1469-1479. Cabanne, F., Martin-Tanguay, J . , and (Martin, C. (1977). Phenolamines associees a I'induction florale et a I'etat reproducteur du Nicotiana tabacum var. Xanthi. Physiol. Veg. 15, 429-443. Campbell, R. C. (1989). Statistics forj Biologists. (Cambridge: Cambridge University Press) . Carey, M., Lin, Y. -S. , Green, M. R., and Ptashne, M. (1990). A mechanism for synergistic activation of a mammalian gene by GAL4 derivatives. Nature 345, 361-367. Chamovitz, D. A., Wei, N., Osterlund, M. T., von Arnim, A. G., Staub, J . M., Matsui, M., and Deng, X.-W. (1996). The COP9 complex, a novel multisubunit nuclear regulator involved in light control of a plant developmental switch. Cell 86, 115-121. i Chen, R.-H., Miettinen, P. J . , Maruoka, E. M., Choy, L., and Derynck, R. (1995). A WD-domain protein that is associated with jand phosphorylated by the type II TGF-B receptor. I Nature 377, 548-552. I I : 132 i i i Chortyk, O. T. (1972). High Molecular Weight Materials of Tobacco. In The Chemistry of Tobacco and Tobacco Smoke., I. Schmeltz, ed. (New York: Plenum Press). Chung, J . , Whiteley, M., and Felsenfield, G. (1993). A 5' element of the chicken B-globin i domain serves as an insulator in human erythroid cells and protects against position effect in Drosophila. Cell 74, 505-514. I Coen, E., Carpenter, R., and Martin, C. (1986). Transposable elements generate novel spatial patterns of gene expression in Antirrhinum majus. Cell 47, 285-296. i I Coen, E. S. (1991). The role of homeofic genes in flower development and evolution. Ann. Rev. Plant Physiol. Plant Mol. Biol. 42, 241-279. I i Courey, A. J . , and Tjian, R. (1992). Mechanisms of Transcriptional Control as Revealed by Studies of Human Trabscription Factor SP1 . In Transcriptional Regulation, S. L. McKnight and K. R. Yamamoto, eds. (New York:;Cold Spring Harbor Laboratory Press). Crabtree, G . R., Schibler, U., and Scott, M. P. (1992). Transcriptional Regulatory Mechanisms in Liver and Midgut Morphogenesis of Vertebrates and Invertebrates. In Transcriptional Regulation, S. L. McKnight and K. R. Yamamoto, eds. (New York: Cold Spring i Harbor Laboratory Press). ] Dercks, W., Creasy, L. L., and Luczka, -. B., C.J . (1994). Stilbene phytoalexins and disease i resistance in Vitis. In Handbook of Phytoalexin Metabolism and Action., M. Daniel and R. P. Purkayastha, eds. (New York: Marcel-Dekker Inc.). i i i i I Despres, C , Subramaniam, R., Matton, D. P., and Brisson, N. (1995). The activation of the potato PR-10a gene requires the phosphorylation of the nuclear factor PBF-1. Plant Cell 7, 5 8 9 - 5 9 8 . Dixon, R. A., Bhattacharyya, M. K., Harrison, M. J . , Faktor, O., Lamb, C. J . , Loake, G., Ni, W., Oommen, A., Paiva, N., Stermer, B., and Yu, L. M. (1993). Transcriptional Regulation of Phytoalexin Biosynthetic Genes., F. W. Nester and D. P. S. Verma, eds. Donald, R. G., and Cashmore, A. (1990). Mutation of either G box or I box sequences profoundly affects expression from the Arabidopsis rbcS-1A promoter. EMBO J . 9, 1717-1726 . Douglas, C , Hoffman, H., Schulz, W., and Hahlbrock, K. (1987). Structure and alicitor or u.v.-light-stimulated expression of two 4-coumarate:CoA ligase genes in parsley. EMBO J . 6, 1189-1195. Douglas, C. J . (1996). Phenylpropanoid metabolism and lignin biosynthesis: from weeds to trees. Trends Plant Sci . 1, 171-178. Douglas, C. J . , Ellard, M., Hauffe, K. D., Molitor, E., Moniz de Sa, M., Reinold, S., Subramaniam, R., and Williams, F. (1992). General Phenylpropanoid Metabolism: Regulation by Environmental and Developmental Signals. In Phenolic Metabolism in Plants, H. A. Stafford and R. K. Ibrahim, eds. (New York: Plenum Press). Draphin, R., Merino, A., and Reinberg, D. (1993). Regulation of RNA polymerase II transcription. Curr. Opin. Cell Biol. 5, 469-476. Drews, G. N., Beals, T. P., Bui, A. Q., and Goldberg, R. B. (1992). Regional and Cell-Specific Gene Expression Patterns During Petal Development. Plant Cell 4, 1383-1404. 1 34 Dubnau, J . , and Struhl, G. (1996). RNA recognition and translational regulation by a homeodomain protein. Nature 379, 694-699. Dzierzak, E., and Medvinsky, A. (1995). Mouse embryonic hematopoiesis. Trends in Genetics and Development 11, 631-638. Edwards, K., Johnstone, C , and Thompson, C. (1991). A simple and rapid method for the preparation of plant genomic DNA for PCR analysis. Nucleic Acids Res. 19, 1349. Ellard-lvey, M., and Douglas, C .J . (1996). Role of Jasmonates in the Elicitor- and Wound-Inducible Expression of Defense Genes in Parsley and Transgenic Tobacco. Plant Physiol. 112, 183-192. Essau, K. (1977). Anatomy of seed plants. (Toronto: Wiley and Sons). Feldbrugge, M., Sprenger, M., Dinkelbach, M., Yazaki, K., Harter, K., and Wesisshaar, B. (1994). Functional analysis of a light-responsive plant bZIP transcriptional regulator. Plant Cell 6, 1607-1621. Ferre-D'Amare, A. R., Prendergast, G. C , Ziff, E. B., and Burley, S. K. (1993). Recognition by Max of its cognate DNA through a dimer b/HLH/Z domain. Nature 363, 38-43. Fisher, D., Carr, C , Parent, L , and Sharp, P. (1991). TFEB has DNA-binding and oligomerization properties of a unique helix-loop-helix/zipper family. Genes Dev. Fritze, K., Staiger, D., Czaja, I., Walden, R., Schell, J . , and Wing, D. (1991). Developmental and UV light regulation of the snapdragon chalcone synthase promoter. Plant Cell 3, 893-905. 1 35 Gallagher, S. R. (1992). Quantitation of GUS Activity by Fluorometry. In GUS Protocols, S. R. Gallagher, ed. (San Diego, CA: Academic Press, Inc.). Gierse, H.-D. (1975). Dissertation: University of Freiburg). Goldberg, R. B. (1986). Regulation of Plant Gene Expression. Phil. Trans. R. Soc. Lond. fi, 3 4 3 - 3 5 3 . Goodrich, J . , Carpenter, R., and Coen, E. S. (1992). A common gene regulates pigmentation pattern in diverse plant species. Cell 68, 955-964. Griffiths, A. J . F., Miller, J . H., Suzuki, D. T., Lewontin, R. C , and Gelbart, W. M. (1993). An Introduction to Genetic Analysis, 5th Edition (New York: W. H. Freeman and Co.). Gross, G. (1985). Biosynthesis and metabolism of phenolic acids and monolugnols. In Biosynthesis and Biodegradation of Wood Components., T. Higuchi, ed. (Toronto: Academic Press) . Gross, G. G. (1981). Phenolic acids. In Secondary Plant Products, E. E. Conn, ed. (Toronto: Academic Press), pp. 301-316. Grotewald, E., Drummond, B., Bowen, B., and Peterson, T. (1994). The Myb -homologous P gene controls phlobaphene pigmentation in maize floral organs by directly activating a subset of flavonoid biosynthetic genes. Cell 76, 543-553. Grueneberg, D. A., Simon, K. J . , Brennan, K., and Gilman, M. (1995). Sequence-specific targeting of nuclear signal transduction pathways by homeodomain proteins. Mol. Cell. Biol. 1555, 3318-3326. 1 36 Ha, S. B., and An, G. (1988). Identification of upstream regulatory elements involved in the developmental expression of the Arabidopsis thaliana cab1 gene. Proc. Nat. Acad. Sci. USA 85. Hahlbrock, K. (1981). Flavonoids. In Secondary Plant Products, E. E. Conn, ed. (Toronto: Academic Press). Hahlbrock, K., and Grisebach, H. (1974). . In The Flavonoids, J . B. Harborne, T. J . Mabry and H. Mabry, eds. (London: Chapman and Hall). Hahlbrock, K., and Scheel, D. (1989). Physiology and molecular biology of phenylpropanoid metabolism. Ann. Rev. Plant Physiol. Plant Mol. Biol. 40, 347-369. Hahn, M. G., Binhoff, A., and Grisebach, H. (1985). Quantitative localization of the phtyoalexin glyceollin I in relation to fungal hyphae in soybean roots infected with Phytophthora megasperma f. sp. glycinea. Plant Physiology 77, 591-601. Harborne, J . B. (1979). Variation in and Functional Significance of Phenolic Conjugation in Plants. Recent Adv. Phytochem. 12, 457-474. Hart, J . H. (1981). Ann. Rev. Phytopathol. 19, 437-458. Hatton, D., Sablowski, R., Yung, M.-H., Smith, C , Schuch, W., and Bevan, M. (1995). Two classes of cis sequences contribute to tissue-specific expression of a PAL2 promoter in transgenic tobacco. Plant J . 7, 859-876. Hatton, D., Smith, C , and Bevan, M. (1996). Tissue-specific expression of the PAL3 promoter requires the interaction of two conserved cis sequences. Plant Mol. Biol. 31, 3 9 3 - 3 9 7 . Hauffe, K. D., Lee, S. P., Subramaniam, R., and Douglas, C. J . (1993). Combinatorial interactions between positive and negative cis -acting elements control spatial patterns of 4CL-1 expression in transgenic tobacco. Plant J . 4, 235-253. Hauffe, K. D., Paszkowski, U., Schulze-Lefert, P., Hahlbrock, K., Dangl, J . L., and Douglas, C. J . (1991). A parsley 4CL-1 promoter fragment specifies complex expression patterns in transgenic tobacco. Plant Cell 3, 435-443. Henderson, A. J . , and Calame, K. L. (1995). Lessons in transcriptional regulation learned from studies on immunoglobulin genes. Crit. Rev. in Euk. Gene Express. 5, 255-280. Holton, T. A., and Cornish, E. C. (1995). Genetics and biochemistry of anthocyanin biosynthesis. Plant Cell 7, 1071-1083. Horsch, R. B., Fry, J . , Hoffman, N., Neidermeyer, J . , Rogers, S. G., and Fraley, R. T. (1988). Leaf Disc Transformation. In Plant Molecular Biology Manual, S. B. Gelvin and R. A. Schilperoort, eds. (Boston: Kluwer Academic Publishers). Jackson, D., Roberts, K., and Martin, C. (1992). Temporal and spatial control of expression of anthocyanin biosynthetic genes in developing flowers of Anirrhinum majus. Plant J . 2, 4 2 5 - 4 3 4 . Jackson, J . R., and Benyajati, C. (1992). In vivo stage- and tissue-specific DNA-protein interactions at the D. melanogaster alcohol dehydrogenase distal promoter and adult enhancer. Nucl. Acid Res. 20, 5413-5422. 1 38 Jacobs, M., and Rubery, P. H. (1988). Naturally occurring auxin transport inhibitors. Science 241, 346-349. Jahnen, W., and Hahlbrock, K. (1988). Differential regulation and tissue-specific distribution of enzymes of phenylpropanoid pathways in developing parsley seedlings. Planta 173, 453-458. Jan, Y. N., and Jan, L. Y. (1993). HLH Proteins, Fly Neurogenesis, and Vertebrate Myogenesis. Cell 75, 827-830. Jefferson, R. A., Kavanagh, T. A., and Bevan, M. W. (1987). GUS fusions: B-glucuronidase as a sensitive and versatile gene fusion marker in higher plants. EMBO J . 6, 3901-3907. Jones, D. H. (1984). Phenylalanine ammonia lyase: regulation of its induction, and its role in plant development. Phytochemistry 23, 1349-1359. Katagiri, F., and Chua, N.-H. (1992). Plant transcription factors: present knowledge and future challenges. Trends in Genetics 8, 22-32. Knobloch, K.-H., and Hahlbrock, K. (1975). Isozymes of p-coumarate:CoA ligase from cell suspension cultures of Glycine max. Eur. J . Biochem. 52, 311-320. Koes, R. E., Spelt, C. E., and Mol, N. M. (1989). The chalcone synthase multigene family of Petunia hybrida (V30): Differential, light-regulated expression during flower development and UV light induction. Plant Mol. Biol. 12, 213-225. Koes, R. E., van Blokland, R., Quattrochio, R., van Tunen, A. J . , and Mol, J . N. (1990). Chalcone synthase promoters in petunia are active in pigmented and unpigmented cell types. Plant Cell 2, 379-392. Koltunow, A. M., Truettner, J . , Cox, K. H., Wallroth, M., and Goldberg, R. B. (1990). Different temporal and spatial gene expression patterns occur during anther development. Plant Cell 2, 1201-1224. Lawrence, P. A., and Struhl, G. (1996). Morphogens, compartments, and patterns: lessons from Drosophila? Cell 85, 951-961. Lee, D., and Douglas, C. (1996). Two divergent members of a tobacco (Nicotiana tobacum) 4-coumarate:CoA ligase (4CL) gene family: cDNA structure, gene inheritance and expression, and properties of recombinant proteins. Plant Physiol. 112, 193-205. Lee, D., Ellard, M., Wanner, L. A., Davis, K. R., and Douglas, C. J . (1995). The Arabidopsis thaliana 4-coumarate:CoA ligase (4CL) gene: stress and developmentally regulated expression and nucleotide sequence of its cDNA. Plant Mol. Biol. 28, 871-884. Lee, T. C , Chow, K. L., Fang, P., and Schwartz, R. J . (1991). Activation of skeletal alpha-actin gene transcription: the cooperative formation of serum response factor-binding complexes over positive c/s-acting promoter serum response elements displaces a negative-acting nuclear factor enriched in myoblasts and nonmyogenic cells. Mol. Cell Biol. 11, 5 0 9 0 - 6 0 0 0 . Leon, J . , Lawton, M., and Raskin, I. (1993). Phenolic Acids and the Biosynthesis of Defense-Related Compounds. In Plant Signals in Interactions with Other Organisms., J . C. Schultz and I. Raskin, eds. (Rockville, Maryland: American Society of Plant Physiologists), pp. 65-78. 1 40 Lewin, B. (1980). Gene Expression: Eucaryotic Chromosomes. Leyva, A., Liang, X., Pintor-Toro, J . A., Dixon, R. A., and Lamb, C. J . (1992). C/s-element combinations determining phenylalanine ammonia-lyase gene tissue-specific expression patterns. Plant Cell 4, 263-271. Loake, G. J . , Faktor, O., Lamb, C. J . , and Dixon, R. A. (1992). Combination of H-box and G-box c/'s-elements is necessary for feedforward stimulation of a chalcone synthase promoter by the phenylpropanoid pathway intermediate p-coumaric acid. Proc. Nat. Acad. Sci. USA 89, 9 2 3 0 - 9 2 3 4 . Logemann, E., Parniske, M., and Hahlbrock, K. (1995). Modes of expression and common structural features of the complete phenylalanine ammonia-lyase gene family in parsley. Proc. Nat. Acad, Sci. USA 92, 5905-5909. Lois, R., Dietrich, A., Hahlbrock, K., and Schulz, W. (1989). A phenylalanine ammononia-lyase gene from parsley: structure, regulation and identification of elicitor and responsive cis -acting elements. EMBO J . 8, 1641-1648. Long, S. (1989). Rhizobium - legume nodulation: Life together in the underground. Cell 56, 2 0 3 . Loscher, R., and Heide, L. (1994). Biosynthesis of p-Hydrxybenzoate from p-coumarate and p-coumaroyl-coenzyme A in cell-free extracts of Lithospermum erythrorhizon cell cultures. Plant Physiol. 106, 271-279. Lozoya, E., Hoffman, H., Douglas, C , Schulz, W., Scheel, D., and Hahlbrock, K. (1988). Primary structures and catalytic properties of isoenzymes encoded by the two 4-coumarate;CoA ligase genes in parsley. Eur. J . Biochem. 176, 661-667. Luo, D., Coen, E. S., Doyle, S., and Carpenter, R. (1991). Pigmentation mutants produced by transposon mutagenesis in Antirrhinum majus. Plant J . 1, 59-69. Maniatis, T., Goodbourn, S., and Fischer, J . A. (1987). Regulation of Inducible and Tissue-Specific Gene Expression. Science 236, 1237-1244. Martin-Tanguay, J . , Cabanne, F., Perdrizet, E., and Martin, C. (1978). The distribution of hydroxycinnamic acid amides in flowering plants. Phytochemistry 18, 1921-25. Martin-Tanguay, J . , Perdrizet, E., Prevost, J . , and Martin, C. (1982). Hyroxycinnamic acid amides in fertile and cytoplasmic male sterile lines of maize. Phytochem. 21, 1939-1 9 4 5 . McKendree, W. L , Paul, A. L , DeLisle, A. J . , and Ferl, R. J . (1990). In vivo and in vitro characterization of protein interactions with the dyad G-box of the Arabidopsis Adh gene. Plant Cell 2, 207-214. Merli, C , Bergstrom, D. E., Cygan, J . A., and Blackman, R. K. (1996). Promoter specificity mediates the independent regulation of neighbouring genes. Genes Dev. 10, 1260-1270. Mikkelsen, T. (1993). Interpreting sequence motifs: a cautionary note. Trends Genet. 9, 159. Milos, P. M., and Zaret, K. S. (1993). A ubiquitous factor is required for C/EBP-related proteins to form stable transcription complexes on an albumin promoter segment in vitro. Genes Dev. 6, 991-1004. Moyano, E., Martinez-Garcia, J. F., and Martin, C. (1996). Apparent redundancy in myb -gene function provides gearing for the control of flavonoid biosynthesis in Antirrhinum flowers. Plant Cell 8, 1519-1532. Mudryj, M., Devoto, S. H., Hiebert, S. W., Hunter, J., Pines, J., and Nevins, J . R. (1991). Cell cycle regulation of the E2F transcription factor involves an interaction with cyclin A. Cell 65, 1243-1252. Naya, F. J., Stellrecht, C. M. M., and Tsai, M. J. (1995). Tissue-specific regulation of the insulin gene by a novel basic helix-loop-helix transcription factor. Genes Dev. 9, 1009-1019 . Nusslein-Volhard, C. (1996). Gradients That Organize Embryo Development. Scientific American, 54-61. O'Hare, P., and Goding, C. R. (1988). Herpes simplex virus regulatory elements and the immunoglobulin octamer domain bind a common factor and are both targets for virion transactivation. Cell 52, 435-443. Ohl, S., Hedrick, S. A., Chory, J., and Lamb, C. J. (1990). Functional properties of a phenylalanine ammonia-lyase promoter from Arabidopsis. Plant Cell 2, 837-848. Ondek, B., Shepard, A., and Herr, W. (1987). Discrete elements with the SV40 enhancer region display different cell-specific enhancer activities. EMBO J. 6, 1017-1028. Ponchet, M., Martin-Tanguay, J . , Marais, A., and Martin, C. (1982). Hydroxycinnamoyl acid amides and aromatic amines in the inflorescences of some Araceae species. Phytochem. 21, 2865-2869. Prendergast, G. C , and Ziff, E. B. (1991). Association of Myn, the murine homolog of max, with c-Myc stimulates methylation-sensitive DNA binding and ras cotransformation. Cell 65, 395-407. Ptashne, M. (1988). How eukaryotic transcriptional activators work. Nature 335, 683-689 . Quattrochio, R., Wing, J . F., Leppen, H., Mol, J . N. M., and Koes, R. E. (1993). Regulatory genes controlling anthocyanin pigmentation are functionally conserved among plant species and have distinct sets of target genes. Plant Cell 5, 1497-1512. Ranjeva, R., Boudet, A. M., and Faggion, R. (1976). Phenolic metabolism in petunia tissue. Biochimie 58, 1255-1262. Reinold, S., Hauffe, K. D., and Douglas, C. J . (1993). Tobacco and parsley 4-coumarate:CoA ligase genes are temporally and spatially regulated in a cell type-specific manner during tobacco flower development. Plant Physiol. 101, 373-383. Rivera-Pomar, R., Niessing, D., Schmidt-Ott, U., Gehring, W. J . , and Jackie, H. (1996). RNA binding and translational suppression by bicoid. Nature 379, 746-749. Rousseau, S., Asselin, M., Renaud, J . , and Ruiz-Carrillo, A. (1993). Transcription of the histone H5 gene is regulated by three differentiation-specific enhancers. Mol. Cell Biol. 13, 4904 -1 7. Sablowski, R. W. M., Moyano, E., Culianez-Macia, F. A., Schuch, W., Martin, C , and Bevan, M. (1994). A flower-specific Myb protein activates transcription of phenylpropanoid biosynthetic genes. EMBO J . 13, 128-137. Salinas, J . , Oeda, K., and Chua, N.-H. (1992). Two G-Box Related Sequences Confer Different Expression Patterns in Transgenic Tobacco. Plant Cell 4, 1485-1493. Sambrook, J . , Fritsch, E. F., and Maniatis, T. (1989). Molecular Cloning: A Laboratory Manual. (Cold Spring Harbor, NY: Cold Spring Harbor Press). Schell, J . T. (1987). Transgenic plants as tools to study the molecular organization of plant genes. Science 237, 1176-1183. Schirm, S., Jiricny, J . , and Schaffner, W. (1987). The SV40 enhancer can be dissected into multiple segments each with a different ceel type specificity. Genes Dev. /, 65-74. Schmid, J . , Doerner, P. W., Clouse, S. D., Dixon, R. A., and Lamb, C. J . (1990). Developmental and environmental regulation of a bean chalcone synthase promoter in transgenic tobacco. Paint Cell 2, 619-631. Schroder, G., Brown, J . S., and Schroder, J . (1988). Molecular analysis of resveratrol synthase. J . Biochem. 172, 161-169. Schulze-Lefert, P., Dangl, J . L , Becker-Andre, M., Hahlbrock, K., and Schulz, W. (1989). Inducible in vivo footprints define sequences necessary for UV-light activation of the parsley chalcone synthase gene. EMBO J . 8, 651-656. Shirley, B., Kubasek, W. L , Storz, G., Bruggeman, E., Koorneef, M., Ausubel, F., and Goodman, H. M. (1995). Analysis of Arabidopsis mutants deficient in flavonoid biosynthesis. Plant J . 8, 659-671. Shufflebottom, D., Edwards, K., Schuch, W., and Bevan, M. (1993). Transcription of two members of a gene family encoding phenylalanine ammonia- lyase leads to remarkably different cell specificities and induction patterns. Plant J . 3, 835-845. Sinha, N. R., Williams, R. E., and Hake, S. (1993). Overexpression of the maize homeobox gene, KNOTTED-1, causes a switch from determinate to indeterminate cell fates. Genes Dev. 7, 787-795. Solano, R., Nieto, C , Avila, J . , Canas, L , Diaz, I., and Paz-Ares, J . (1995). Dual DNA binding specificity of a petal epidermis-specific MYB transcription factor (MYB.Ph3) from Petunia hybrida. EMBO J . 14, 1773-84. Sommer, H., Bonas, U., and Saedler, H. (1988). Transposon-induced alterations in the promoter region affect transcription of the chalcone synthase gene in Antirrhinum majus. Mol. Gen. Genet. 202, 429-434. St. Johnson, D., and Nusslein-Volhard, C. (1992). The Origin of Pattern and Polarity in the Drosophila Embryo. Cell 68, 201-219. Stafford, H. (1990). Flavonoid Metabolism (Boca Raton: C R C Press). Strubin, M., Newell, J . W., and Mathias, P. (1995). OBF-1, a novel B-cell-specific coactivator that stimulates immunoglobulin promoter activity through association with octamer-binding proteins. Cell 80, 497-512. Taylor, L. P., and Jorgensen, R. (1992). Conditional male fertility in chalcone synthase--deficient petunia. Journal of Heredity 83, 11-17. Tietjen, K. G., and Matern, U. (1983). Differential response of cultured parsley cells to elicitors from two non-pathogenic strains of fungi. Eur. J . Biochem. 131, 409-413. van der Meer, I. M., Brouwer, M., Spelt, C. E., Mol, J . N. M., and Stuitje, A. (1992). The TACPyAT repeats in the chalcone synthase promoter of Petunia hybrida act as a dominant negative cis -acting module in the control of organ-specific expression. Plant J . 2, 525-5 3 5 . van der Meer, I. M., Spelt, C. E., Mol, J . N. M., and Stuitje, A. R. (1990). Promoter analysis of the chalcone synthase (chs A ) gene of Petunia hybrida : a 67 bp promoter region directs flower-specific expression. Plant Mol. Biol. 15, 95-109. van Dijk, M. A., and Murre, C. (1994). Extradenticle raises the DNA binding specificity of homeotic selector gene products. Cell 78, 617-624. Watanabe, R., and Wender, S. H. (1965). Flavonoid and certain related phenolic compounds in parts of the tobacco flower. Archives of Biochemistry and Biophysics 112, 111-114. Weintraub, H. (1993). The MyoD family and myogenesis: redundancy, networks and thresholds. Cell 75, 1241-1244. Weising, K., Schell, J . , and Kahl, G. (1988). Foreign genes in plants: transfer, structure, expression, and applications. Ann. Rev. Genet. 22, 421-477. Weisshaar, B., Block, A., Armstrong, G. A., Herrmann, A., Schulze-Lefert, P., and Hahlbrock, K. (1991). Regulatory elements required for light-mediated expression of the Petroselenium crispum chalcone synthase gene. Symp. Soc. Exp. Biol. 45, 191-210. Whetten, R., and Sederoff, R. (1995). Lignin Biosynthesis. Plant Cell 7, 1001-1013. Wiermann, R. (1981). Secondary plant products and cell and tissue differentiation. In Secondary plant products., E. E. Conn, ed. (Toronto: Academic Press). Williams, M. F., Foster, R., and Chua, N.-H. (1992). Sequences flanking the hexameric G-box core C A C G T G affect the specificity of protein binding. Plant Cell 4, 485-496. Wissenbach, M., Uberlacker, B., Vogt, F., Becker, D., Salamini, F., and Rohde, W. (1993). Myb genes from Hordeum vulgare : tissue specific expression of chimaeric Myb promoter/Gus genes in transgenic tobacco. Plant J . 4, 411-422. Wu, S. C , and Hahlbrock, K. (1992). In situ localization of phenylpropanoid-related gene expression in different tissues of light-grown and dark-grown parsley seedlings. Z. Naturforsch C 47, 591-600. Yalpani, N., Leon, J., Lawton, M. A., and Raskin, I. (1993). Pathway of salicylic acid biosynthesis in healthy and virus-inoculated tobacco. Plant Physiol. 103, 315-321. Yamamoto, K. R., Pearce, D., Thomas, J . , and Miner, J . N. (1992). Combinatorial Regulation at a Mammalian Composite Response Element. In Transcriptional Regulation, S. L. McKnight and K. R. Yamamoto, eds. (New York: Cold Spring Harbor Laboratory Press). Ylistra, B., Busscher, J., Franken, J . , Hollman, P. C. H., Mol, J . N. M., and van Tunen, A. J . (1994). Flavonols and fertilization in Petunia hybrida: localization and mode of action during pollen tube growth. Plant Journal 6, 201-212. Ylistra, B., Touraev, A., Moreno, M. B., Stoger, E., van Tunen, A. J . , Vicente, O., Mol, J. N. M., and Heberle-Bors, E. (1992). Flavonols stimulate development, germination and tube growth of tobacco pollen. Plant Physiology 100, 902-907. Yu, L. M., Lamb, C. J . , and Dixon, R. A. (1993). Purification and biochemical characterization of proteins which bind to the H-box cis -element implicated in transcriptional activation of plant defense genes. Plant J . 3, 805-816. Zenk, M. H. (1966). . In Biosynthesis of Aromatic Compounds, H. Wagner and L. Horhammer, eds. (Oxford: Pergammon), pp. 45-60. Zenk, M. H. (1979). Recent work on cinnamoyi coA derivatives. Recent Adv. Phytochem. 12, 1 3 9 - 1 7 6 . 

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.831.1-0088336/manifest

Comment

Related Items