u k .gwaamflufia THE!“ This is to certify that the dissertation entitled COMPONENTS INVOLVED IN THE DEGRADATION OF RNA IN PLANTS. presented by Crispin Barwick Taylor has been accepted towards fulfillment of the requirements for PH.D. degreein GENETICS Wafiw Majcflrofessor Date gig g2 43 Mrn.‘~...un ,- 1 - n1 -n . . . - 0-12771 Jllilillllillllili“ LIBRARY Michigan State University PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE MSU is An Affirmative Action/Equal Opponunity Institution czblmw‘educma- COMPONENTS INVOLVED IN THE DEGRADATION OF RNA IN PLANTS By Crispin Barwick Taylor A DISSERTATION Submitted to Michigan State University in partial fulfilment of the requirements for the degree of DOCTOR OF PHILOSOPHY Graduate Program in Genetics 1993 ABSTRACT COMPONENTS INVOLVED IN THE DEGRADATION OF RNA IN PLANTS By Crispin B. Taylor As a first step towards the elucidation of fundamental principles governing the control of gene expression at the level of mRNA stability in plants, two types of molecular components that may be involved in the process have been identified: ribonuclease genes in Arabidopsis, and sequence elements that target transcripts for rapid degradation in tobacco. Three Arabidopsis RNase genes, RN37, RN82 and RN83 were identified by virtue of their homology to a class of fungal RNases. The RNS RNases are also related to the S-RNases, components of a plant self-recognition response termed self-incompatibility. The expression of the RN32 gene was found to be induced during senescence, and in response to phosphate starvation in Arabidopsis, implying a role for RN52 in the remobilization of phosphate from RNA. Sequence comparisons demonstrated that RN82 falls into a subclass within the T2 family that is distinct from that formed by the S-RNases. Further, using a system developed to measure mRNA half-lives in stably transformed tobacco cell lines, two sequence elements have been found to confer instability on reporter transcripts. When the first, termed DST, was present in two copies in the 3' untranslated regions of two different reporter transcripts, it caused a pronounced decrease in the half-lives of those transcripts. The second element, termed the AUUUA repeat, also caused a marked destabilization of reporter transcripts. In both cases, the effect of the elements on transcript stability in the cell lines was mirrored by a marked decrease in the accumulation of the corresponding transcripts in transgenic plants. Finally, a novel differential screening strategy has been employed to identify eight genes with unstable transcripts (GUTs) in tobacco, solely on the basis of the instability of their mRNAs. Future identification of endogenous instability determinants within the GUTs should lead to a better understanding of the different mechanisms by which unstable transcripts may be recognized and targeted for rapid degradation in plants. Taken together, this work should provide the foundation for future efforts aimed at elucidating mRNA decay pathways in plants. To Cyn and to Emma who have, quite simply, made it all worthwhile. ACKNOWLEDGEMENTS I would like to take this opportunity to thank the many individuals who have contributed directly and indirectly to the work described in this Dissertation. I thank Pam Green for her unfailing and enthusiastic support of my endeavors, and Tom Friedman, Natasha Raikhel, Mike Thomashow and Steve Triezenberg, for their patience, and for generously dispensing advice and encouragement whenever it has been solicited. Thanks are also due to each of the rotation students, undergraduates, post-docs and graduate students who have enriched my "PhD experience". It has truly been a pleasure to work in Pam’s lab, not least because of the atmosphere created by Pam and all of these individuals. In particular, I want to thank André Dandridge, who has diligently provided expert technical assistance and good company almost from the start, and my co-authors Pauline Bariola, Tom Newman, Masaru Ohme-Takagi, whose individual contributions to this work are indicated throughout. I would also like to acknowledge the contribution of all the people who work behind the scenes to keep the PRL and its constituent labs functioning. Without past and present office staff, and individuals such as Marlene Cameron, Barb Hoeffler, Bob Keever, Elliot Light and Kurt Stepnitz, completing my PhD would have been much harder. Finally, I must recognize the contribution of my family. Emma has accepted the ephemeral presence of her father at home with grace, and Cynthia has supported, guided, encouraged and loved. Without their unfailing . friendship I could never have done this. Thank you. TABLE OF CONTENTS LIST OF TABLES ........................................................................ ix LIST OF FIGURES ....................................................................... x CHAPTER 1: Introduction ............................................................ 1 CHAPTER 2: Genes with homology to fungal and S-gene ribonucleases are expressed in Arabidopsis thaliana ............................................ 14 ABSTRACT ....................................................................... 15 INTRODUCTION ................................................................ 16 MATERIALS AND METHODS .............................................. 17 RESULTS AND DISCUSSION ............................................... 20 LITERATURE CITED ........................................................... 30 CHAPTER 3: RN82: a senescence-associated RNase of Arabidopsis that diverged from the S-RNases prior to speciation ................................................... 32 ABSTRACT ....................................................................... 33 INTRODUCTION ................................................................ 34 MATERIALS AND METHODS .............................................. 36 RESULTS AND DISCUSSION ............................................... 39 LITERATURE CITED ........................................................... 57 vi CHAPTER 4: DST sequences, highly conserved among plant SAUR genes, target reporter transcripts for rapid degradation in tobacco .................................... 60 ABSTRACT ....................................................................... 61 INTRODUCTION ................................................................ 62 MATERIALS AND METHODS .............................................. 67 RESULTS ......................................................................... 76 DISCUSSION .................................................................... 99 LITERATURE CITED ........................................................... 107 CHPATER 5: The effect of sequences with high AU content on transcript stability in tobacco ................................................................................. 113 ABSTRACT ....................................................................... 114 INTRODUCTION ................................................................ 115 MATERIALS AND METHODS .............................................. 118 RESULTS ......................................................................... 119 DISCUSSION .................................................................... 131 LITERATURE CITED ........................................................... 139 CHAPTER 6: The identification of genes with unstable transcripts (GUTs) in tobacco ................................... 142 ABSTRACT ....................................................................... 143 INTRODUCTION ................................................................ 144 MATERIALS AND METHODS .............................................. 147 RESULTS AND DISCUSSION ............................................... 150 CONCLUSIONS AND FUTURE PROSPECTS ........................... 169 vii CHAPTER 6 (cont) LITERATURE CITED ........................................................... 171 CHAPTER 7: Conclusions ............................................................ 173 viii LIST OF TABLES Table 2-1: Amino acid homologies among the RNS gene products and related RNases .............................. 24 Table 4-1: Inhibition of mRNA synthesis in NT-I cells ..................... 77 Table 4-2: mRNA half-lives in NT-1 cells stably transformed with pMON505 ...................................... 82 Table 4-3: Effect of DST sequences on mRNA stability in NT—1 cells ............................................................ 92 Table 5-1: Effect of the AUUUA repeat on GUS mRNA half-lives in NT-1 cells ............................................... 125 Table 6-1: GUT and NOTGUT mRNA levels and half-lives ................. 155 Table 6-2: Structural characteristics of GUTs and NOTGUTs ............ 162 Figure 2-1: Figure 2-2: Figure 2-3: Figure 3-1: Figure 3-2: Figure 3-3: Figure 3-4: Figure 3-5: Figure 3-6: Figure 4-1: Figure 4-2: Figure 4-3: Figure 4-4: LIST OF FIGURES Sequence alignment of the RNS gene products .............. 22 Southern analysis of the A. thaliana RNS genes ............. 25 Northern analysis of the A. thaliana RNS transcripts ....... 27 Alignment of S- and S-like RNase amino acid sequences ............................................... 43 Expression of RN82 in yeast ........................................ 45 Gene genealogy of the RNase superfamily ..................... 47 Expression of RN82 in different organs of Arabidopsis ............................................... 50 Induction of the RN82 transcript during senescence in Arabidopsis ......................................... 51 Effect of phosphate starvation on RN82 expression ...................................................... 54 Location and structure of DST elements ........................ 66 Representation of the CATand GU8 genes following integration of pMON505 into the tobacco genome ........................................... 79 Decay of transcripts in tobacco (NT-1) cells stably transformed with pMON505 ...................... 81 Disappearance of transcripts encoded by CAT and GU8 genes expressed under the control of copies of the Cab-1 promoter in transgenic tobacco ................ 86 Figure 4-5: Representation of the test and reference genes following integration into the tobacco genome .............. 88 Figure 4-6: Destabilization of the GU8 transcript by DST sequences in stably transformed NT-1 cells .................. 91 Figure 4-7: Effects of DST and control inserts on the stability of GU8 or globin test transcripts ................................. 94 Figure 4—8: Mapping of the 3’ ends of the GU8 transcripts ............... 96 Flgure 4-9: Reduced globin mRNA accumulation caused by DST sequences in transgenic plants ............................ 98 Figure 5-1: Constructs used to test the effect of AU-containing sequences ......................................... 122 Figure 5-2: Rapid mRNA decay caused by the AUUUA repeat in tobacco cells ........................................................ 124 Figure 5-3: Histograms comparing the effect of the AU-containing inserts on GU8 and globin mRNA stability .................... 127 Figure 5-4: SI nuclease protection analysis of the 3’ ends of the test transcripts ................................................ 130 Figure 5-5: Effect of the AUUUA repeat in transgenic plants ............ 132 Figure 6-1: Preliminary characterization of potential GUTs ................ 153 Figure 6-2: Time-course experiments to determine GUT transcript half-lives ............................................ 159 Figure 6-3: Sequence similarities of GUT and NOTGUT cDNAs ......... 166 xi CHAPTER 1 INTRODUCTION 2 In eukaryotes, each of the steps between transcription of a gene and degradation of the corresponding protein are potential targets for regulation of expression. Of these, transcriptional control mechanisms are the best understood, both from the perspective of basic components of the transcriptional apparatus, and with respect to the mechanisms that regulate specific genes or sets of genes. However, it has become apparent that many eukaryotic genes are controlled to some extent by post-transcriptional events (Peltz et al., 1991). The contribution of mRNA processing, transport, and stability, and of translational and post-translational mechanisms to the regulation of gene expression has therefore received increasing attention recently (Green, 1993; Gallie, 1993; Sachs, 1993; Belasco and Brawerman, 1993). Of these, regulation at the level of mRNA stability is among the most widely evoked, but the mechanisms by which such control is exerted remain, on the whole, poorly understood. Unstable transcripts, with half-lives on the order of minutes, are the exception rather than the norm in higher eukaryotes. Average transcript half- lives are on the order of several hours (Siflow and Key, 1979; Brook and Shapiro, 1983), and it has therefore been suggested that the destabilization of specific transcripts is an active process (Hunt, 1988). The most extensively studied unstable transcripts in mammalian systems often encode proteins involved in. the regulation of cellular growth and differentiation, such as c-myc (Jones and Cole, 1987), granulocyte/macrophage - colony stimulating factor 3 (GM-CSF) (Shaw and Kamen, 1986), and others. The proteins encoded by these genes are required only transiently in the cell, and their aberrant expression can have catastrophic consequences, such as cellular transformation and oncogenesis (Eick et al., 1986). Rapid changes in the levels of these proteins is thought to be facilitated by the instability of the corresponding transcripts, as unstable transcripts will more quickly attain new steady-state levels (both higher and lower) following changes in the transcription rate of their genes (Peltz, et al., 1991). The rate of synthesis of an individual transcript is regulated via the interactions between components of the transcriptional machinery with specific sequences in the promoter of the corresponding gene. Following this paradigm, it seems likely that the degradation rate of that transcript will largely be defined via the interactions between specific sequence elements within that transcript and cellular factors that mediate it’s degradation. Indeed, a number of sequence determinants of mRNA instability have been identified in mammalian systems. Among the first instability sequences to be described was an AU-rich element (ARE) in the 3’ untranslated region (UTR) of GM-CSF. When this element was placed in the 3’UTR of the otherwise stable B—globin transcript, it caused the transcript to be rapidly degraded (Shaw and Kamen, 1986). Other characterized instability sequences include, for example, an element in the 3’UTR of the transferrin receptor mRNA, which facilitates the destabilization of this transcript in response to increases in cellular iron concentration (Klausner 4 et al., 1993), and a coding-region portion of the c-fos transcript which only appears to destabilize this transcript in serum-stimulated cells (Shyu et al., 1989). In all of these cases, one or more factors have been identified that specifically interact with each instability sequence. Presumably some or all of these factors are involved in targeting the transcripts for rapid degradation. Rapid changes in gene expression are likely to be particularly important in plants (eg. Theologis, 1986). As sessile organisms plants cannot move away from adverse stimuli but must respond by altering their metabolism, often quite quickly, if they are to survive (Green, 1993). Plants are therefore likely to have a large number of genes whose expression must change quickly in response to environmental stimuli and, accordingly, mRNA stability may figure prominently in the regulation of many plant genes (Green, 1993). As yet, however, there are only a few reports of unstable mRNAs in plants, and none in which the mechanisms that mediate a transcripts’ instability are defined. In addition to sequence elements that confer instability on transcripts, RNases are also likely to figure prominently in the degradation pathways of all mRNAs, in animals as well as plants. A large number of RNases have been described from various plant species, and have been classified into four major groups on the basis of characteristics such as molecular mass, pH optima, homopolymer specificity and cation requirements (Wilson, 1982). The levels of some of these RNases, or their activities, are known to respond to a variety of environmental stimuli and developmental cues, such as senescence (Matile 5 and Winkenbach, 1971; Blank and McKeon, 1991). However, the RNases that play a role in mRNA degradation have not been differentiated from those with other functions in any eukaryote. Without knowing which enzymes are involved, it is difficult to elucidate the mechanisms by which mRNAs are recognized and degraded. This Dissertation addresses efforts to identify sequence elements that confer instability on reporter transcripts in tobacco, and also focusses on the initial isolation and characterization of RNase genes in Arabidopsis. The bases for each of the major research projects described in the Dissertation are outlined below. ArgQiQQQsjfi RN§§§§ and RN§§§ QQDQS One approach toward the identification of plant RNases that participate in the degradation of messenger RNA is to biochemically characterize as many RNases as possible from a model plant species. Using this work as a basis, it should then be possible to screen for mutants deficient in any of the RNases identified. These mutants can then be used to address the function of the corresponding RNase. To this end, initial studies have led to the identification of some 16 bands of RNase activity in extracts of Arabidopsis tissue using a gel-based RNase assay (Yen and Green, 1991). Among these are 6 representatives of at least three of the four major classes of RNase previously described in plants (Wilson, 1982; Yen and Green, 1991). More recently a number of Arabidopsis mutants have been identified that exhibit differences from the wild type profile of RNase activities on these gels (Abler, M.A., Danhof, L. and Green, P.J. unpublished). By examining these mutants for defects in the degradation of specific mRNAs, it may be eventually be possible to identify those mutants affected in the activity of a RNase involved in mRNA degradation. A complementary approach toward elucidating the function of RNases in Arabidopsis is to identify RNase genes in this species. Until recently, the only plant RNases that had been cloned were the S-RNases (McClure et al., 1989b). The S-RNases, which are associated with self-incompatibility (SI) in the Solanaceae, were identified as RNases through their homology to a class of fungal RNases typified by RNase T2 of Aspergi/Ius oryzae (Kawata et al., 1988; McClure et al., 1989b). The S-RNases exhibit many of the properties expected for cellular recognition molecules, but their precise roles in mediating the SI response are still not clear (Haring et al., 1990). Nonetheless, the sequence conservation between the S-RNases and the fungal RNases lent itself to a PCR approach toward detecting related RNase genes in Arabidopsis. The application of this strategy, and the identification of three Arabidopsis RNase genes, RN81, RN82, and RN83, are described in Chapter 2 of this Dissertation. Chapter 3 includes a discussion of the evolutionary relationships among the 7 T2/S class of RNases. The further characterization of RN82 expression in Arabidopsis, and the potential functions of the RN82 protein are also discussed in this Chapter. As discussed above, unstable transcripts are likely to be targeted for rapid degradation via interactions between specific sequences within that transcript and components of the cellular RNA degradation machinery, such as RNases. The identification of sequences that confer instability on transcripts is thus a prerequisite to understanding how mRNA decay rates are controlled. One approach that has been widely employed in mammalian systems is to examine the effect of putative instability determinants on the half-lives of otherwise stable reporter transcripts (reviewed in Peltz, et al.,) In many of these experiments, reporter transcript half-lives are measured by blocking transcription with inhibitors, such as Actinomycin D (ActD). The subsequent disappearance of the reporter transcript is followed by, for example, northern blotting. Chapter 4 of this Dissertation describes the development of an analogous system for measuring transcript half-lives in stably transformed plant cells. The system makes use of a tobacco cell line, NT-1, that can be stably transformed (An, 1985) with reporter constructs containing putative instability 8 determinants, and that is also amenable to ActD treatment. The NT-1 cell system has been used to identify two very different sequence elements that function to destabilize transcripts in tobacco. The first, termed DST, is derived from the Small Auxin Up BNAs (8AURs), which were initially characterized in soybean (McClure et al., 1989a). This element was chosen to be tested for its effect on transcript stability because the 8AURs appear to be among the most inherently unstable plant transcripts yet characterized, with half lives estimated to be on the order of 10 - 50 minutes (McClure and Guilfoyle, 1989; Franco et al., 1990). Also, the DST element is a short (~40 bp) sequence that is conserved in the 3’UTRs of all the SAUR genes whose sequences have been reported to date (McClure et al., 19893; Yamamoto et al., 1992; Gil, F. Liu, Y., Orbovic, V., Verkamp, E., Poff, K.L. and Green, P.J., submitted). The conservation of this element in a non-coding region supports its functional significance; its location is similar to that of several mammalian instability determinants, which suggested that it may mediate the instability of the 8A UR transcripts (McClure et al., 19893). When present in two copies in the 3’UTRs of reporter transcripts, the DST element caused a pronounced decrease in the half-lives of those transcripts in stably transformed NT-1 cells. The same transcripts also accumulated to markedly reduced levels in transgenic tobacco plants, implying that the DST element can target transcripts for rapid degradation in plants as well as NT-1 cells. The experiments to analyze the effect of the DST element on transcript stability in 9 NT-1 cells, and on transcript accumulation in transformed tobacco are also described in Chapter 4. The DST element differs from a prominent class of mRNA instability sequences that have been shown to function in mammalian cells. Sequences in the 3’UTRs of a number of proto-oncogene, Iymphokine and cytokine transcripts are AU-rich and contain multiple copies of an AUUUA motif. The AREs from GM-CSF (Shaw and Kamen, 1986) and c-fos (Shyu et al., 1989) are known to target reporter transcripts for rapid degradation in mammalian cells. To test whether similar AU-rich sequences may also destabilize transcripts in plants, two sequence elements with identical A + U content, but different primary structures, were inserted into the 3’UTRs of reporter transcripts. The effect of these elements on reporter transcript half-life in NT-1 cells was analyzed as outlined above for the DST element. Of the sequences tested, only an element containing 11 overlapping copies of the AUUUA sequence (termed the AUUUA repeat) conferred instability on the reporter transcripts. This element also caused a marked decrease in the accumulation of the reporter transcript in transformed tobacco plants. These experiments are described in Chapter 5 of this Dissertation. The identification of the DST and AUUUA elements as instability determinants in plants is an important step toward the goal of understanding the mechanisms controlling mRNA stability in plants. However the examination of the role of the DST and AUUUA elements was prompted by previous 1O evidence suggesting that they may affect transcript stability. They are therefore unlikely to be representative of the repertoire of instability sequences that may function in endogenous plant transcripts. An initial step toward the identification of additional plant instability sequences is to first identify unstable transcripts, before determining which portions of those transcripts mediate their instability. We have recently used a novel differential screening strategy to search for genes with unstable transcripts (GUTs) in tobacco. One major advantage of this approach is that it facilitates the identification of GUTs solely on the basis of the instability of their transcripts. The screen is based on the rationale that levels of unstable mRNAs will rapidly fall in cells treated with ActD. PolylA)+ RNA isolated from ActD-treated cells will therefore be depleted of unstable transcripts, relative to the levels of those transcripts in control, untreated cells. The differential screen, and the preliminary characterization of the eight GUTs we have isolated using this strategy, are the subjects of Chapter 6 of this Dissertation. In Chpater 7, some of the recent work that has had as its impetus the studies performed in this Dissertation is summarized, and suggestions for future research efforts are outlined. 1 1 LITERATURE CITED An, G. (1985) High efficiency transformation of cultured tobacco cells. Plant Physiol. 79: 568-570 Belasco, J., and Brawerman, G. (1993) eds. Control of messenger RNA stability. (San Diego, California: Academic Press) Blank, A., and McKeon, TA. (1991) Expression of three RNase activities during natural and dark-induced senescence of wheat leaves. Plant Physiol. 97: 1409-1413 Brock, ML, and Shapiro, DJ. (1983) Estrogen stabilizes vitellogenin mRNA against cytoplasmic degradation. Cell 34: 207-214 Eick, D., Piechaczyk, M., Henglein, B., Blanchard, J.M., Traub, B., Kofler, E., Wiest, S., Lenoir, GM, and Bornkamm, G.W. (1986) Aberrant c-myc RNAs of Burkitt’s lymphoma cells have longer half-lives. EMBO J. 4: 3717-3724 Franco, A.R., Gee, M.A., and Guilfoyle, T.J. (1990) Induction and super- induction of auxin-responsive mRNAs with auxin and protein synthesis inhibitors. J. Biol. Chem. 265: 15845-15849 Gallie, DR. (1993) Post-transcriptional regulation of gene expression in plants. Annu. Rev. Plant Physiol. Mol. Biol. 44: 77-106 Green, P.J. (1993) Control of mRNA stability in plants. Plant Physiol. 102: 1065-1070 Haring, V., Gray, J.E., McClure, B.A., Anderson, M.A., and Clarke, A.E. (1990) Self-incompatibility: a self-recognition system in plants. Science 250: 937-941 Hunt, T. (1988) Controlling mRNA lifespan. Nature 38: 567-568 Jones, T.R., and Cole, MD. (1987) Rapid cytoplasmic turnover of c-myc mRNA: requirement of the 3’ untranslated sequences. Mol. Cell. Biol. 7: 4513-4521 Kawata, Y., Sakiyama, F., and Tamaoki, H. (1988) Amino acid sequence of ribonuclease T2 from Aspergi/lus oryzae. Eur. J. Biochem. 176: 683-697 12 Klausner, R.D., Rouault, T.A., and Harford, J.B. (1993) Regulating the fate of mRNA: the control of cellular iron metabolism. Cell 72, 19-28 Matile, P., and Winkenbach, F. (1971) Function of lysosomes and lysosomal enzymes in the senescing corolla of the Morning Glory (Ipomoea purpurea). J. Exp. Bot. 22: 759-771 McClure, B.A., and Guilfoyle, T. (1989) Rapid redistribution of auxin-regulated RNAs during gravitropism. Science 243: 91-93 McClure, B.A., Hagen, G., Brown, C.S., Gee, M.A., and Guilfoyle, T.J. (1989a) Transcription, organization, and sequence of an auxin-regulated gene cluster in soybean. Plant Cell 1: 229-239 McClure, B.A., Haring, V., Ebert, P., Anderson, M.A., Simpson, R.J., Sakiyama, F., and Clarke, A.E. (1989b) Style self-incompatibility gene products of Nicotiana alata are ribonucleases. Nature 342: 955-957 Peltz, S.W., Brewer, G., Bernstein, P., Hart, P.A., and Ross, J. (1991) Regulation of mRNA turnover in eukaryotic cells. In Critical Reviews in Eukaryotic Gene Expression, G.S. Stein, J.L. Stein, and J.B. Lian, eds (Boca Raton: CRC Press), pp. 99-126 Sachs, AB. (1993) Messenger RNA degradation in eukaryotes. Cell 74, 413- 421 Shaw, G., and Kamen, R. (1986) A conserved AU sequence from the 3’ untranslated region of GM-CSF mRNA mediates selective mRNA degradation. Cell 46: 659-667 Shyu, A.-B., Greenberg, ME, and Belasco, JG. (1989) The c-fos transcript is targeted for rapid decay by two distinct mRNA degradation pathways. Genes and Devel. 3: 60-72 Siflow, OD. and Key, J.L. (1979) Stability of polysome-associated polyadenylated RNA from soybean suspension culture cells. Biochemistry 18: 1013-1018 Theologis, A. (1986) Rapid gene regulation by auxin. Annu. Rev. Plant Physiol. 37: 407-438 Wilson, CM. (1982) Plant nucleases: biochemistry and development of multiple molecular forms. in MG. Rattazzi, S.C. Scandalios, G.S. Whitt, eds, lsozymes. Current Topics in Biological and Medical Research, Vol. 6. 13 Alan R. Liss, Inc, New York, pp 33-54 Yamamoto, K.T., Mori, H., and Imaseki H. (1992) cDNA cloning of indole-3- acetic acid-regulated genes: Aux22 and SAUR from mung bean (Vigna radiata) hypocotyl tissue. Plant Cell Physiol. 33: 93-97 Yen, Y., and Green, P.J. (1991) Identification and properties of the major ribonucleases of Arabidopsis thaliana. Plant Physiol. 97: 1487-1493 CHAPTER 2 GENES WITH HOMOLOGY TO FUNGAL AND S-GENE RNASES ARE EXPRESSED IN Arabidopsis thaliana This Chapter was originally published in Plant Physiology. Reference: Taylor, CB. and Green, P.J. (1991 ). Genes with homology to fungal and S-gene RNases are expressed in Arabidopsis thaliana. Plant Physiol. 96: 980-984. 14 1 5 ABSTRACT The only RNase genes that have been cloned from higher plants are those expressed in species that are known to exhibit self-incompatibility, such as the S—genes of Nicotiana alata. In this report, we investigated whether the expression of this particular type of RNase gene is restricted to self— incompatible species, or if similar genes are expressed in other plants. Using a polymerase chain reaction approach we have identified a set of three putative RNase genes in the self-compatible plant Arabidopsis thaliana IL.) Heynh. ecotype Columbia. These A. thaliana genes, designated RNS1, RN82, and RNS3, do not cross—hybridize to each other, and their products are homologous to both the S-gene RNases of N. alata and a set of related fungal enzymes. The A. thaliana RNSI , RN32 and RN83 genes encode transcripts of 1.1, 1.3 and 1.1 kb respectively, and of the three genes, RNS2 is the most highly expressed in whole plants. The identification of the RNS genes in a self- compatible species suggests that this class of RNases Is of general significance in RNA catabolism in higher plants. 16 INTRODUCTION RNA catabolism is an important component of many cellular processes including, for example, the post—transcriptional regulation of gene expression through mRNA decay. However, little is known about the control of RNA degradation in any eukaryote. This is especially true in higher plants where, despite the biochemical characterization of numerous RNase activities [18, 6], the only RNase genes that have been cloned are those coding for the style- specific products of the S-genes, components of the self-incompatible response in species such as Nicotiana alata [2]. The S-genes had been cloned for some time before comparison of deduced amino acid sequences [12] showed that their products were related to a group of fungal RNases consisting of RNase T2 of Aspergil/us oryzae [10], RNase Rh of Rhizopus niveus [8] and RNase M of Aspergil/us saitoi [9]. Style extracts and purified S-gene products of N. alata were shown to have RNase activity in vitro [12], but the precise roles of the S-gene products, termed S-RNases [12], in arresting the growth of incompatible pollen tubes remains unclear. One model is that the S-RNases produced in the style enter incompatible pollen tubes and attack their rRNA, hence inhibiting their growth [13, 7]. We are interested in identifying RNases that have fundamental roles in RNA catabolism in higher plants. This interest prompted us to ask whether the expression of RNases similar to the S—RNases is only associated with the self- 17 incompatible response, or if related enzymes are also expressed in self- compatible species. The latter would suggest that these RNases have more widespread functions in higher plants. Our approach was to use DNA primers, corresponding to amino acids conserved among fungal and S-RNases [12], for PCR2 amplification of an Arabidopsis thaliana IL.) Heynh. ecotype Columbia cDNA library. Following PCR amplification of the A. thaliana cDNA library, we have identified three putative RNase genes, termed RN81, RN82 and RN83. These genes are all expressed in the self-compatible species A. thaliana, and their products are strikingly similar to the N. alata S-RNases and the fungal RNases typified by RNase T2. This suggests that RNases of this type may be expressed in many plant species, and therefore are likely to be involved in other aspects of RNA turnover beyond those associated with self-incompatibility. MATERIALS AND METHODS QDNA Library An Arabidopsis thaliana IL.) Heynh. ecotype Columbia cDNA library constructed from total green tissue using the Lambda Zap vector (Stratagene) was kindly provided by N.-H. Chua and A. Gasch. The library was ’zapped’ into plasmid form [4], and DNA representing 180,000 recombinants was amplified using 18 PCR, as described below. P R lnin The sequences of the primers used for PCR were: a n I Y achpn rcnwis 5' GAATTCAT can TTN TGG ccu GA 3' 5' crccacr ncc ATG TTT TTT ATA TTC 3- EcoRI c c XhoI c 0 AAC cc c c c Nucleotide sequences were derived from previously published deduced amino acid sequences [2, 3, 8, 9, 10] indicated in bold type, where N represents all four possible nucleotides. The second primer was the complement of the sequence coding for the amino acids indicated. Amplification of sequences in the library flanked by these primers was carried out using two sequential PCR experiments. The initial amplification consisted of 25 cycles of 94°C for 1 min, 42°C for 2 min, and 72°C for 3 min. The reaction, in a volume of 25uL, contained 0.3,ug of plasmid library DNA linearized with EcoRI, 42 pmol of each primer, 50mM KCI, 10mM Tris-HCI, pH 8.3, 1.5mM MgClz, 0.01% (w/v) gelatin, 200pM of each dNTP and 0.6 units of Taq DNA polymerase (Perkin Elmer). The products of the reaction were separated on a low melting temperature agarose gel and the major product (175-225bp) was excised and used for a second round of PCR under the same conditions as above, except that the 42°C step was carried out at 50°C. The major PCR product was excised, blunt-end ligated Into a plasmid vector, and 19 sequenced using the dideoxy method of Sanger et al [16]. Southern and Ngrthern Analyses Genomic DNA was isolated from all above-ground tissues of mature, five-week old soil-grown A. thaliana, using the method of Dellaporta et al. [5]. After digestion with restriction endonucleases, DNA was separated on an agarose gel and blotted onto BioTrace RP (Gelman Scientific). Prehybridization was for five hours at 52°C in a buffer containing 5XSSC, 10X Denhardt’s solution [15], 0.1% SDS, 0.1M potassium phosphate, pH6.8 and 100ug/mL denatured salmon sperm DNA. Hybridization was for 16 hours at 52°C in the same buffer, except that 10% (w/v) dextran sulphate and 30% (v/v) formamide were added, and the SDS was omitted. Total RNA was isolated from similar tissues essentially as described by Puisant and Houdebine [14]. PolyA+ RNA was purified by oligo dT—cellulose affinity chromatography as recommended by the supplier (Pharmacia). PolyA+ and polyA'(the flow-through from the oligo dT-cellulose column) RNAs were denatured and separated on formaldehyde/agarose gels. Following transfer to BioTrace RP, the blots were treated as described above. 20 RESULTS AND DISCUSSION Am lifi ion f P a iv RN n The primers for PCR amplification described above were designed to code for two of the longest stretches of amino acid homology among fungal and S- RNases [12]. These blocks of amino acid homology have subsequently been found in the S-gene products of a number of other plant species that exhibit self-incompatibility, including Petunia hybrida [3], Petunia inf/ata [1] and Solanum tuberosum [7], as well as in additional N. alata S-alleles [11]. In between these blocks of homology, McClure et al [12] also noted a number of individual amino acid residues that were absolutely conserved among three 8- RNases (82, S3 and SB) and two fungal RNases (T2 and Rh). The presence of these residues in the PCR products from A. thaliana would provide additional confirmation that the corresponding polypeptides were related to the aforementioned RNases. When the primers described above were used to amplify hybridizing sequences from an A. thaliana cDNA library, the major PCR product was approximately 200bp. This could encode some 67 amino acids, which was within the range of sizes expected for the region between the primers from the deduced amino acid sequences of the RNases discussed above. 21 Relationship of the RNS Gene Preducts to Other RNases Full or partial (single nucleotide) sequencing of 21 individual recombinants obtained following cloning of the major PCR product demonstrated that there are at least three putative RNases in A. thaliana related in deduced amino acid sequence to the fungal and S-RNases shown in Figure 2-1. These have been designated RN81, RN82 and RN83. 0f the 21 clones that were sequenced, and two additional clones analyzed by dot blot hybridization (not shown), twelve were identical to RN81, five to RN82 and three to RN83. The three remaining clones showed no similarity to the RNases shown in Figure 2-1. In Figure 2-1 the amino acid residues that are absolutely conserved among the RNS gene products of A. thaliana and the 82, 83, 86, T2 and Rh RNases are boxed. These identical residues include the blocks of amino acid homology used for design of the PCR primers. Moreover, all of the residues between these blocks that were shown to be completely conserved among the fungal and S- RNases in Figure 2-1 [12], are also found in each of the three RNS clones. This strongly suggests that the RNS gene products are indeed RNases. This argument is strengthened by the fact that two of the identical amino acids that fall within the regions covered by the PCR primers, His 1 and His 59 in Figure 2-1, are known to be part of the active site of RNase T2 [10] and RNase M [9]. Two other amino acids that are identical among all the RNases shown in Figure 2-1, Trp 38 and Pro 39, have been suggested to contribute to the base non-specific character of RNase T2 [10]. Another residue that is .omfito ._\ ..o.< “oboe: .t ..:.m “Em-R .2 ..m.z 88$me 8‘ 5.4. 832.8 mm Em 05m: 9: E to. 9: 9. 95:555.? 9:. .EmEcmzm 05 32530 3 3269:: 3mm 98:05 3:me .F wzm Lo EoEmm: 60:26 of E Eon 05:5 3.2: of Lou— P 53> 9:568 6938:: Em $0533 of. 6963 So 332$ 65. Co :6 E 52:52 mum 55 3563. AN 2 .E B 9200.). +0 55 co comma mm; 295.95 8.533 .8. «Sea: .m 3 5. 9.62m 6cm 5: omeo i we N... ommzm AN 5 83m .2 .6 mommzmm 65 53> mecca mzm 6:865 ..\ 9: .5 305353 Eon oEEm vooauoc Co EmEcm=< .mommzm-m new .35: 53> $2665 memo mzm 05 Lo EoEcmzm moccacom .7N 2:9”. 2 2 905 23 H mm 3% Illlwwoo lllll Umwz m3 >v§2|>mA A MBImUOmOAHABHmemmmMn—I U Ihdlwwmgoz Qm3A0m NB .0.4 EOE m3 m mm 3% llll>zzz IIIII 002m m3 >BAEImZM A mmaxm¥IIH>mQDHMUI U IIIhZABBmNVZ szAUm mm "0.2 80mg Z» W EVA 3% >OEIWHUOEIIHBAD m3 Imolloll A QMIEQQIIIQAwamammn—I U IIIVZASBm>z onASAUm mm .m.m .50va >> m QM 3h AOMIDAmQKXBIIMfiAO m3 Immlloll A OZvagglllBhSZumxmmol U III>ZAZBBSZ ogAUm mm .muz .60va H0 m EB 3h llégzllmmlIOmAn—t m3 mmOAIQmE A Om IIIIIII >mAQIOm—Immomz U IZOQISUOBMV DQZAUm mmzm .u_a 90mg >0 ”w :0 3m mOVAIUUZUmmmmUUmAm m3 NMQAUQIE A Em IIIIIII HMmelmIn—mmwl U Ulmml3mwoz> Qm3AUm Nmzm .974 .50va Hummlmm 3h mwwmollllmmllofifi m3 mvmvdzlmBA A 0m IIIIIII Hlfimmommxmdol H Izmmlwfigv; hum—SAGE Hmzm Jain. om om ov on ON CA A 23 identical in all the sequences in Figure 2-1, Cys 16, has been found (in RNase T2) to form a disulfide bond with a Cys located just outside the region shown in Figure 2-1 [10]. In addition to the identical amino acids, there is significant homology between the RNS gene products and the fungal and S-RNases throughout the cloned region. When the percentages of functionally identical amino acids are compared (Table 2-1), the homology between individual RNS sequences and the S-RNases is between 34 and 48%, while that between the RNS sequences and the fungal RNases ranges from 46 to 57%. The highest degree of homology (between 65 and 72%) is found when the individual RNS sequences are compared to each other. These extensive homologies are also reflected in hydrophobicity plots of this region of the RNases (data not shown). r r and Ex re ion f h RN nes Analysis of DNA sequence data showed that the percentages of identical nucleotides for the RNS genes of A. thaliana were 59% for RN81 and RN82, 57% for RN81 and RNS3, and 61% for RN82 and RN83 (data not shown). This suggested that the RNS sequences would not cross-hybridize, which was confirmed by Southern analysis (Figure 2-2). As shown in Figure 2-2, each RNS clone hybridizes to a unique set of restriction fragments in A. thaliana genomic DNA, indicating that RN81, RN82 and RN83 correspond to different genes. It is likely that the RNS genes are not members of large gene families 24 Table 2-1. Amino Acid Homologies Among RNS Gene Products and Related RNases RN32 RN83 N.a.SZ N.a.S3 N.a.S6 R.n.Rh A.oT2 RN81 67 72 48 4O 34 52 57 RN82 65 44 45 38 46 48 RN83 47 37 35 46 51 Deduced amino acid sequences were aligned as shown in Figure 2-1, and the number of functionally identical amino acids for each pair of gene products was totalled. Percent homologies were calculated by dividing the number of functionally identical amino acids by the average total number of amino acids for each pair of sequences. Functionally identical amino acids are grouped as: A,S,T; N,0; D,E; l,L,M,V; H,K,R; and F,W,Y. Abbreviations are as described in the legend to Figure 2-1. 25 RNS1RNS2 RN83 123456789 .” ' — 6-6 _ — 4-4 ,.‘ if A u . I H — 2'3 - - — 06 Figure 2-2. Southern analysis of the A. thaliana RNS genes. 10pg of genomic DNA was cut with EcoRV (lanes 1,4,7), EcoRV and Hindlll (lanes 2,5,8) or Hindlll (lanes 3,6,9) and separated on a 1% agarose gel. Blots were hybridized with the [32Pl—labelled RNS probes indicated above the lanes. 26 because, with one exception, each clone hybridizes to a single band in genomic Southerns. The two bands in lanes 1 and 2 of Figure 2-1 are expected because of the presence of a site for the restriction endonuclease EcoRV within the cloned fragment of RN81. Analysis of polyA+ RNA shows that each of the RNS genes of A. thaliana is expressed in above-ground tissue (Figure 2-3). The sizes of the RNS gene transcripts are similar to those shown for transcripts of the S-genes of P. hybrida, Le. 1100 bases for RN81 and RNS3, and 1300 bases for RN82, compared to 900 bases for the P. hybrida S-allele 3A [3]. These transcripts have the capacity to code for proteins of molecular mass up to 40kD, which is within the size range of the major RNase activities recently identified in A. thaliana using a substrate-based gel assay [Y. Yen and P.J.G., submitted]. At present it is unknown whether the RNS gene products correspond to any of these RNases. Nevertheless, the expression of genes highly homologous to the S-genes in the self-compatible species A. thaliana suggests that the corresponding RNases may be involved in more general aspects of RNA catabolism in higher plants. Interestingly, expression levels for each of the three A. thaliana genes differ markedly in this tissue sample. The northern blots with RN81 and RN83 probes had to be exposed several times longer than that with an RN82 probe to produce signals of similar intensities (Figure 2-3). These data obtained using Poly A+ RNA, and other experiments performed with total RNA (not shown) 27 RNS 1 RNS 2 RNS 3 1 2 3 4 5 6 kb _ I'8 — I'5 . — 1'3 . o —O.8 Figure 2-3. Northern analysis of the A. thaliana RNS transcripts. 6pg of poly A+ (lanes 2,4,6) and 20pg of poly A' (lanes 1,3,5) RNA isolated from above-ground tissue were denatured and separated on a 1% agarose/2% formaldehyde gel. Blots were hybridized with the [32PI-labelled RNS probes indicated above the lanes. Lanes 3 and 4 were exposed for 1 day; lanes 1,2,5 and 6 for 7 days. 28 clearly indicate that RN82 is the most highly expressed of the RNS genes. However, the relative transcript levels in the tissues analyzed in the northern blots do not correlate with the frequency with which the individual RNS clones were isolated by PCR amplification of the cDNA library (see above). This discrepancy may reflect differences in primer homology to each of the RNS cDNAs. Alternatively, it may be due to a difference between the tissue samples used for preparation of the cDNA library and those used to prepare the RNA for northern analysis. lmplieetiene end Fetere Preepeete Our data indicate that the RNS gene products belong to a class of enzymes that have a homologous active site. This class includes the S-RNases of N. alata, and several fungal RNases. The expression of the RNS genes in a self- compatible plant strongly argues for a function of this class of RNases beyond their role in self-incompatibility, perhaps in stress related responses or senescence [6, 18]. It remains to be determined whether the expression patterns of the RNS genes are the same as those of the S-genes and whether the corresponding enzymes share any common functions. The cloning of the RNS genes from A. thaliana should facilitate the characterization of their functions via antisense RNA techniques, or the use of insertional mutagenesis in the future [17]. Further studies on the temporal and spatial regulation of RNS gene expression in A. thaliana are in progress. These approaches, in conjunction with a detailed biochemical study of the RNases of A. thaliana, will help to define the roles of the RNS gene products in RNA mm J.L.... ._ 2...... .__........_._ 29 catabolism, and may also provide insight into the roles of related proteins in the self-incompatible response. ACKNOWLEDGEMENTS We are greatful to Drs. Nam-Hai Chua and Alex Gasch for providing the A. thaliana cDNA library, and to Jim Dombrowski and Robin Steinman for zapping the library into plasmid form. We thank Dr. Natasha Raikhel and members of the Green lab for helpful comments on the manuscript. [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] 30 LITERATURE CITED Ai Y, Singh A, Coleman CE, loeger TR, Kheyr-Pour A, Kao TH (1990) Self-incompatibility in Petunia inf/ata. Isolation and characterization of complementary DNA encoding three S-allele-associated proteins. Sex. Plant Reprod. 3: 130-138 Anderson MA, Cornish EC, Mau S-L, Williams EG, Hoggart R, Atkinson A, Bonig l, Grego B, Simpson R, Roche PJ, Haley JD, Penshow JD, Niall HD, Tregear G, Coghlan JP, Crawford RJ, Clarke AE (1986) Cloning of cDNA for a stylar glycoprotein associated with expression of self- incompatibility in Nicotiana alata. Nature 321: 3844 Clark KR, Okuley JJ, Collins PD, Sims TL (1990) Sequence variability and developmental expression of S-alleles in self-incompatible and pseudo self-compatible Petunia. Plant Cell 2: 815-826 Delauny AJ, Verma DPS (1990) Isolation of plant genes by heterologous complementation in Escherichia coli. In 88 Gelvin, RA Schilperoort, DPS Verma eds, Plant Molecular Biology Manual. Kulwer Academic, Dordrecht, pp. A14/1-A14/23 Dellaporta SL, Wood J, Hicks JB (1983) A plant DNA minipreparation: Version ll. Plant Mol. Biol. Rep. 1: 19-21 Farkas GL (1982) Ribonucleases and ribonucleic acid breakdown. In D Boulter, B Parthier, eds, Nucleic Acids and Proteins in Plants, Encyclopedia of Plant Physiology, Vol 16. Springer-Verlag, Berlin. pp 224-262 Haring V, Gray JE, McClure BA, Anderson MA, Clarke AE (1990) Self- incompatibility: a self-recognition system in plants. Science 250: 937- 941 Horiuchi H, Yanai K, Takagi M, Yano K, Wakabayashi E, Sanda A, Mine 8, Ohgi K, Irie M. (1988) Primary structure of a base non-specific ribonuclease from Rhizopus niveus. J. Biochem (Tokyo) 103: 408-418 lrie M, Watanabe H, Ohgi K, Harada M (1986) Site of alkylation of the major ribonuclease from Aspergillus saitoi with iodoacetate. J. Biochem (Tokyo) 99: 627-633 Kawata Y, Sakiyama F, Tamaoki H (1988) Amino-acid sequence of ribonuclease T2 from Aspergillus oryzae. Eur. J. Biochem. 176: 683-697 [11] I12] I13] I14] [15] [16] [17] [18] 31 Kheyr-Pour A, Bintrin SB, loerger TR, Remy R, Hammond SA, Kao TH (1990) Sequence diversity of pistil S-proteins associated with gametophytic self-incompatibility in Nicotiana alata. Sex. Plant Reprod. 3: 88-97 McClure BA, Haring V, Ebert PR, Anderson MA, Simpson RJ, Sakiyama F, Clarke AE (1989) Style self-incompatibility gene products of Nicotiana alata are ribonucleases. Nature 342: 955-957 McClure BA, Gray JE, Anderson MA, Clarke AE (1990) Self- incompatibility in Nicotiana alata involves degradation of pollen rRNA. Nature 347: 757-760 Puissant C, Houdebine L-M (1990) An improvement of the single-step method of RNA isolation by acid guanidinium thiocyanate-phenol- chloroform extraction. BioTechniques 8: 148-149 Sambrook J, Fritsch EF, Maniatis T (1989) Molecular Cloning: A Laboratory Manual 2"° Edition. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY Sanger F, Nicklen S, Coulsen AR (1977) DNA sequencing with chain- terminating inhibitors. Proc. Natl. Acad. Sci. USA 74: 5463-5467 Somerville C (1989) Arabidopsis blooms. Plant Cell 1: 1131-1135 Wilson CM (1982) Plant nucleases. Biochemistry and development of multiple molecular forms. In MC Ratizzi, JG Scandalios and GS Whitt, eds, Isozymes: Current Topics in Biological and Molecular Research, Vol 6. Alan R. Liss, New York, pp 33-54 CHAPTER 3 RNS2: A SENESCENCE-ASSOCIATED RNASE OF ARABIDOPSIS THAT DIVERGED FROM THE S-RNASES BEFORE SPECIATION This Chapter was originally published in Proc. Natl. Acad. Sci. Reference: Taylor, C.B., Bariola, P.A., delCardayré, S.B., Raines, RT. and Green, P.J. (1993) RN82: A senescence-associated RNase of Arabidopsis that diverged from the S-RNases prior to speciation. Proc. Natl. Acad. Sci. 90: 5118-5122 32 33 ABSTRACT Several self-compatible species of higher plants, such as Arabidopsis thaliana, have recently been found to contain S-like ribonucleases (RNases). These S-Iike RNases are homologous to the S-RNases that have been hypothesized to control self-incompatibility in Solanaceous species. However, the relationship of the 8- like RNases to the S—RNases is unknown, and their roles in self-compatible plants are not understood. To address these questions, we have investigated the RN82 gene, which encodes an S-like RNase (RN82) of Arabidopsis. Amino acid sequence comparisons indicate that RN82 and other S-Iike RNases comprise a subclass within an RNase superfamily, which is distinct from the subclass formed by the S-RNases. RN82 is most similar to RNase LE [Jost, W., Bak, H., Glund, K., Terpstra, P., Beintema, J.J. (1991) Eur. J. Biochem. 198, 1-6.], an S-like RNase from Lycopersicon esculentum, a Solanaceous species. The fact that RNase LE is more similar to RN82 than to the S-RNases from other Solanaceous plants indicates that the S-like RNases diverged from the S-RNases prior to speciation. Like the S-RNase genes, RN82 is most highly expressed in flowers, but unlike the S-RNase genes, RN82 is also expressed in roots, stems and leaves of Arabidopsis. Moreover, the expression of RN82 is increased in both leaves and petals of Arabidopsis during senescence. Phosphate starvation can also induce the expression of RN82. Based on these observations we suggest that one role of RN82 in Arabidopsis may be to remobilize phosphate, particularly when cells senesce, or when phosphate becomes limiting. 34 INTRODUCTION Fundamental insights into the relationship between protein structure and function and gene evolution have been gained from the study of members of the pancreatic ribonuclease (RNase) superfamily typified by RNase A [1]. Another RNase family has recently been identified in the plant kingdom [2]. RNases in this family are not homologous to the pancreatic RNase superfamily, but rather share homology with a class of fungal RNases that includes RNase T2 of Aspergillus oryzae [3]. The plant members of this family include the S- RNases, proteins associated with a self-recognition response known as self- incompatibility (SI) in certain species of higher plants. In Nicotiana alata and other members of the Solanaceae, pollen carrying a particular allele at the S- locus, which controls SI, are unable to fertilize plants carrying the same S-allele. The mechanism by which S-RNases may participate in the rejection of incompatible pollen is unknown; however, their genes co-segregate with the S- locus [4]. Initially it seemed plausible that the S-RNases were highly specialized polymorphic enzymes, without homologs in self-compatible species. However, several recent studies have shown that this is not the case. Among these are data obtained from PCR experiments performed with cDNA from the self- compatible crucifer, Arabidopsis thaliana. Using primers directed against the regions most highly conserved between the fungal and S-RNases, amplification products corresponding to three Arabidopsis RNase genes, RN81, RN82, and 35 RN83, were identified [5]. Further evidence for S-RNase homologs in self- compatible plants is indicated by the protein sequences of RNase LE, isolated from cultured cells of Lycopersicon esculentum (tomato) [6], and RNase MC, isolated from seeds of Momordica charantia (bitter gourd) [7]; both share homology with fungal RNases and S-RNases. It has also been shown that two RNases isolated from tomato fruit, Tf1 and Tf2, have a number of biochemical properties in common with the S-RNases [8]. The finding of S-RNase homologs ("S-like" RNases) in self-compatible species indicates that enzymes related to the S-RNases play a fundamental role in RNA catabolism in higher plants. Plants contain a large number of RNase activities that are regulated in response to a variety of stimuli [9, 10]. Several plant RNases have been purified and characterized biochemically, but nothing is known about the corresponding genes. It is therefore difficult to assess whether previously described RNase activities, such as those induced during senescence [11, 12], are encoded by S-like RNase genes. In an effort to gain a better understanding of the role of S—Iike RNases in plants and their relationships with the S-RNases, we have investigated RN82, the most highly expressed of the Arabidopsis RNS genes [5]. 36 MATERIALS AND METHODS llinn ninofRNZDNA The Arabidopsis cDNA library in Lambda Zap [13] that was initially used to detect the RNS genes by PCR amplification [5] was used to screen for RN82 cDNA clones by plaque hybridization [14] using the RN82 PCR product as a probe. Positive plaques were purified, the cDNA clones converted into plasmid form [15] and sequenced with Sequenase (United States Biochemical), using the chain-termination method [16]. Expression ef RN82 in Yeast A yeast expression system developed previously for the expression of RNase A (delCardayré, 8.8. Quirk, D.J., Rutter, W.J., and Raines, R.T., submitted) was used in conjunction with a substrate-based gel assay to confirm the RNase activity of RN82. Briefly, an RN82 cDNA covering the entire coding region including the signal sequence was inserted between the yeast PH05 promoter and GADPH terminator [17], and the resulting PH05-RN82-GADPH gene was then cloned into the yeast shuttle vector pWL. The detailed structure of pWL will be published elsewhere (delCardayré, 8.B. Quirk, D.J., Rutter, W.J., and Raines, R.T., submitted). Transformants of Saccharomyces cerevisiae strain BJ2168 (MAT a pro 1-407 prb1-1122 pep4—3leu2 trpl ura3-52) harboring these plasmids were grown in minimal dextrose liquid lacking Trp and containing 0.2 mM KH2P04 (which induces expression from the PH05 37 promoter) [18], or 1 1 mM KH2P04 (which does not induce the PH05 promoter), for 2 days at 30°C. Following clarification of the cultures by two centrifugation steps (both for two minutes in a microfuge), 5 [IL of the resulting supernatant was separated on RNase activity gels [19]. The gels were run, and subsequently treated to visualize RNase activities, essentially as described previously [19]. Multiple Segpenee Alignment The amino acid sequence of RN82 deduced from the cDNA clones was aligned with the amino acid sequences of fifteen Solanaceous S-RNases [20-25], four fungal RNases [3, 26-28], and two S-Iike RNases for which protein sequence data are available [6, 7], using the University of Wisconsin Genetics Computer Group program PILEUP with a gap weight of 3.0 and a gap length weight of 0.1 [29]. The RNase sequences were aligned from the positions corresponding to the presumptive mature N-termini of the N. alata S-RNases [4]. The gene genealogy was also generated using PILEUP, which produces a similarity score for each possible pair of sequences. The genealogy is a representation of these similarity scores, which are used to order the alignment. The horizontal branch distances in the genealogy are proportional to the similarities between the sequences. 38 Expreeeipn Anelyeee Arabidopsis thaliana (L.) Heynh. ecotype RLD was grown under conditions of 12 hours light - 12 hours dark with a relative humidity of 50% at 20°C . Total RNA was isolated from roots, stems, leaves and flowers of 4-5 week old Arabidopsis plants using the method of Puissant and Houdebine [30]. For the senescence experiments, flowers were staged based on the morphological characteristics as defined in Smyth et al. [31]. Senescing leaves were those showing visible signs of senescence, including chlorosis at the leaf margins and wilting. To starve Arabidopsis for phosphate, 1-week-old etiolated seedlings were removed from solid AGM medium (4.39/L MS salts (Sigma), 309/L sucrose, 2mg/L glycine, 100mg/L myoinositol, 0.5mg/L pyridoxine, 0.5mg/L nicotinic acid, 0.1mg/L thiamine-HCI, buffered with 2.5mM 2(N-morpholino ethanesulfonic acid), pH5.7) and shaken in liquid media with or without 1.25 mM KH2P04 as described in [32], for 12 hours in the dark. RNA was separated on formaldehyde-agarose gels and transferred to Biotrace HP membrane (Gelman). The northern blots were hybridized as described previously [5] to a 0.7kb gene—specific probe for RN82, corresponding to the EcoRI - Xbal fragment of the longest RN82 cDNA clone. This probe, which includes 15 bp of the 5’ untranslated region and the coding region up to nucleotide position 696, does not hybridize to RN81 and RN83, and produces an identical pattern of hybridization on genomic Southern blots to that reported for the PCR fragment of RN82 ([5]; C.B.T., P.J.G., results not shown). For use as an internal standard in the senescence and phosphate starvation experiments, a 39 probe for the ubiquitous, highly expressed [33] translation initiation factor elF4A was generated using PCR. The 380bp EIF4A probe was isolated following PCR amplification of the same Arabidopsis cDNA library, using oligonucleotide primers corresponding to amino acids 197 - 203 (5’ GGN THC/T) AA(A/G) GAN CA(A/G) AT 3') and to amino acids 317 - 323 (reverse complement: 5’ TC NIC/TIITlG) CAT lA/G/TIAT NIA/GllC/T I lA/GITC 3’) of Nicotiana tabacum elF4A2, which are highly conserved in N. tabacum elF4A3 and related gene products [33]. The PCR reaction was performed using the conditions previously described for the amplification of the RNS sequences [5]. The major product of the reaction was blunt-end ligated into a plasmid vector and sequenced [16]. Between the primers used for PCR amplification, the deduced amino acid sequence of the major PCR product shows 96.5% identity with the same region of N. tabacum elF4A2, confirming its identification as an Arabidopsis elF4A sequence (C.B.T., P.J.G., unpublished observations). RESULTS AND DISCUSSION RN82 was initially identified as a PCR product amplified from an Arabidopsis cDNA library using primers corresponding to the regions most conserved between the S-RNases and a class of fungal RNases [5]. This PCR product was used as a hybridization probe to isolate RN82 cDNA clones from the same library. The nucleotide and deduced amino acid sequences of the longest clone 40 containing a full-length coding region have been deposited in GENBANK under Accession Number M98336. Analysis of several independent cDNA clones showed that transcripts from the RN82 gene can be polyadenylated at three different sites, a feature common to many plant genes [34]. The 19 amino acids at the N-terminus of the RN82 protein (see legend to Figure 3-1) are typical of a eukaryotic signal sequence, with Ala and Leu making up over half of the residues in this region. This indicates that RN82 is targeted to the secretory pathway in Arabidopsis, as are the S-RNases of the Solanaceae, which are secreted enzymes. To confirm that RN82 is indeed an RNase, the coding region of the RN82 cDNA, including the putative signal sequence was expressed in S. cerevisiae under the control of the PH05 promoter which is induced (derepressed) in medium low in phosphate [18]. RNase activity secreted into the culture medium by yeast transformed with the vector control (CON) or the RN82 expression construct (RN82), was then detected following electrophoresis on RNase activity gels [19], (see Figure 3-2). Under inducing conditions, two bands of RN82 activity were observed that have apparent molecular weights of 28 to 33 kDa, slightly higher than the predicted molecular weight of 27 kDa (if the RN82 signal sequence is cleaved as indicated in the legend to Figure 3- 1). This slight difference in molecular weight and the presence of two RN82 bands may result from processing of the RNS2 signal sequence at multiple sites [35] or differences in glycosylation [36] that are known to affect the mobility of heterologous proteins produced in yeast. It should also be noted that the 41 RNase activity gels are run under non-reducing conditions [19] which may also contribute to the differences. However, both RNase bands are specific to the RN82 clone and correlate with the induction of the PH05 promoter as expected. This demonstrates that the RN82 gene encodes an active RNase. m ri n f RN 2 R l RN To compare the similarity of the deduced amino acid sequence of RN82 with those of related plant RNases, the alignment shown in Figure 3-1 was generated as described in Materials and Methods. The alignment demonstrates that the similarity of RN82 to the S-RNases is dispersed throughout the coding region (light shading in Figure 31). Moreover, each of the five regions most conserved among the S-RNases (numbered C1 - C5 by Kao and co-workers [21, 37], and boxed in Figure 3-1), is also evident in RN82. In contrast to the extensive similarity of RN82 with the S-RNases, its similarity to the related fungal enzymes is highest in the central region of the protein (positions 37 - 110), but is markedly less pronounced towards the N- and C-termini (see asterisks above sequence in Figure 3-1). Twenty-five residues are absolutely conserved among all the RNases of this superfamily. The most prominent of these are the His residues at positions 41 and 106 in Figure 3-2 that have been shown to be important for catalysis in RNase T2 [3]. Others include the pairs of Cys residues at positions 56 and 109, and 174 and 214 that have been , shown to form disulfide bonds that are critical for maintaining the structure of RNase M of Aspergillus saitoi [28]. Trp 80, which has been proposed to 42 Legend to Figure 3-1. Alignment of S- and S-like RNase Amino Acid Sequences The alignment was performed as described in Materials and Methods using the U.W.G.C.G. package [29]. Sequences are of RN82 of Arabidopsis, LE from L. esculentum [6], MC from M. charantia [7], N. alata S-alleles 2, 3 and 6 (sequences labelled 2Nic, 3Nic and 6Nic [20]) and 1, F1 1 and 2 (sequences labelled 1Nic, F1 1 and Z [21]), P. inf/ata S-alleles 1, 2 and 3 (sequences labelled 1Pet, 2Pet, 3Pet [22], 8. chacoense S-alleles 2 and 3 (sequences labelled 280i and 3Sol [23]), P. hybrida S- alleles 1, 2 and 3 (sequences labelled Ps1 B, PsZA, Ps3A [24]), 8. tuberosum 81 allele (sequence labelled 18th [25]). The RNases are aligned from the presumptive mature N-termini of the N. alata S-alleles, as in [4]. Light shading indicates residues in any sequence that are identical or functionally identical to a residue at the same position in the RN82 sequence, ellipses denote residues that are identical or functionally identical in the three S-like RNases, RN82, LE and MC, and heavy shading denotes residues that are identical or functionally identical in at least twelve of the fifteen S- RNases. Conserved regions CI - C5 of loerger et al. [37] are boxed. Asterisks above the RN82 sequence denote residues in the fungal RNases T2 [3], Rh [26], M [27] and Trv [28] that are identical or functionally identical to the RN82 residues. Functionally identical residues are grouped as: A,S,T; l,L,M,V; H,K,R; F,W,Y; D,E; 0,N. The deduced amino acid sequence of RN82 includes the following residues that precede the sequence shown in the alignment: MASRLCLLLLVACIAGAFAIGDVIELNRSQR. The arrow indicates the most likely site for cleavage of the signal sequence, based on a statistical comparison of amino acid sequences in the vicinity of known cleavage sites [38]. The sequence of the full length RN82 cDNA clones were obtained by Pauline Bario/a DJ AM moocmaamm Eo< oEE< ommzm 8:5 ucm -w Ho EoEcm=< . Tm 2:2“— omwemmHemo MonHmeHm H..Mom>HHotzHM>m. MMmchmszmoaHem ome>momMHHm HHnemwx...H chmomw M>HmHmezmo<:o......H3m =UMH>MHMmHHm zHozz..aHoszmm>MzHHOMHmmomzmHHHMszHHm HHHM HM ..H zzaHzM> HMeHoxeoxmzo..... HzmxmoxH>zo¢omo MoH=m>oo. Hme>HzoHzmmammmeumeHMMZHHHM HHHMMMM ..H zdeom» mnz>oxze mxo...... axmzmoaH>zx>Hmo MoHuHmsz MH. Mm....ueHzmMoo. HmHMHezZHzo>a> whommH.MzMHmm HHQMMHM ..H HzHHoMH HMeHozme «Mo. .....aMmMmooH>zx>xoo MUHHHHOHH so. M2....oeHzmHox. eHezzHMo>aH mxemHHomzmmmx HHoHMaM...H HzeHoMH HMzHoozmzx¢emo mon>m2Hm .ooMe....omHzm>oM. HHexHezzHMo>HH MHoMHHMMzMHme HHHHMMM...H «MHHMMH maxHomzozmxo .... HzmamooH>zx>zmo MonHHHHm zomma....omHzmHuo. HHHMHHmzHMo>HH mzomHHmmeHmH HHQMMMM...H HzeHoHH mmxHxx. ozz:o...... woeMmom>2>M¢mma HoHuHmsz >onm. onoMHmm. >mHmfiezH HmomszMzMHew HHQMMMM...H MZHHQHH MMMHHM..¢eMoo .... Momomommzmoazma MUHHHHHHH HHUMH. >moMH. =m>oMHomo>oze>exHHomemzma: xeomeHmmeMHHH HHamMaM. ..H MHHHHMH MHHHzxmoemoo..... Hymnzomymzoddmz moHuHHZHm o>oMH. =m>quomo>oze>eMHoMH. =m>oxHaHoMzz exHHMonoMH: MeomemeeMHHe HHomxox ..> MZHHHHH oMMHHMemmHmao .... HzmmmomommocoMH. zm>ozHomozzmeexHaonmuMe: MaomHHemeMHHH HHHHMMM...H HzHHHMH MMMHeoemoemoo. .... Mxmmmoozmmoacmz MUHMHHZHH M>oMH. zm>oMHoHoxzze>exHHMonome: xeomevaeMHHH HHHHMMM...H MHHHHMH MMHHeoewoemoo..... .Mwmmmoozemoadmz MonHHZHm x>oMH. zm>oxHamoMzze>exH>o>H> m>zeom HMUMHoHHMo.MMzzmz o>¢me> MammmHmeoo.Mo...oHMHmomHHmmomoo >oH>>0HHo m.zom >zoonHeHo.HcmxHHzMHz>HonHoaHMHazuxmomMMmaHoomo>o.ommxmaxu moHMHH..o ....HHQMMU>>HHHHH.M<20H49>HoonHMmmzm<>>o.. H>oe>zmxHHH.zHeeHm> M i H i e i «M i « « s H H _ _ _ _ _ . _ . _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ cem cm~ cmm oH~ oom omH omH oeH ooH omH oeH OHH omH «momM.>HH . HHOHMMHmamHHHMOHH. HeHmmzwmemeHoxHMeHemHmo HzHomHemo .zmH.....>mmmo...ox>xu>om HMMH>Ha3mom m ozeHmm.. ozz. HHHMMzo mzHom>emz .MHH.......Mmo...mszUH2a H30H>szwmm e Home.HoH Huoxuzm>aomxm>Hx.. on. HMHmezo mzHomHemz .mmm.......mHo.m..mxzommH mzeH>HaHzmH a ommaH.xom ooeomz >HmoszH02HHH < omozH.mHH 0096mm MHHHmzmoMonoooMo. .HHHomswooHozoxMoanHxMMMomo..on Hmem>zo HzHozHemz .HHH.......mmo.z..:>zoM4H HzHH>Hazwmm z o>on.Hmm ouaumM Hummmamzoexzooaxm.. MHHosz>>HonmmoezHszmomxo .on. HHHm>zo mzHoxHHmz .mmH.......zMo.m..Hemumea H30H>H02HHM 4 oeHzH.mMm museum zwmzszHHMHHUQHMH. HeHoszooHoxxxxononmumoHmo .on. Hzem>zo mzHomHemz .zmH.. ..... MM0.HmmeHMQMHH m30H>Hoz>mH 4 «moxM.mzH momomm HmmzszuomeuHHHo..HMHoszMoneHHMMHHHHHymomzemxzo mzHomHemz zmHm ..... .mxmo...oH<>oMHM mzflm>HmHHmM o HzozH.Mxo oomozx HHonzmmamHmowoom.. MMHOZMMMQHQZszoH. M..Hmaemzmxoz=. Hmex>zo HzHuxHHmz .zm<......Hmmu..:z<.>ommM HMHH>H0HHHH a ozozH.MmH_Quao:M owoomzHMozmHomamz. HszzzmmoHooHHmmMOH..Hz>>xmmeomo Hmom>xo HzHozHemz .xm<......HMmu..MzMH>oMmM HzHH>Hozon n zmo. .mxzo>3MMmHHa>szan. .ememooHomm. HmmoeHm mzHozHemz .MHM.......MMo. Hzxmmmommm H3HH>HOHHHM z 3mmmHoz>HmoMHm. Mm>mxooaoam Hmmzmxm mzHoMHeMz .zmH....... Mao. HszMHUMmM mzaH>HoHHQM z «MooH. Hzx puHomm zumzmszaMezHHHHo..mmonzMMmH>zHHzommH HMMHMQoHon H. emxxm mzHquemz .zmm.......mou. szm. Henna HMHH>Ham>oH z emoom. Hzx MuHomm szzmszMMHmHHMHo. MmonzMMmH>zHHonM.. HmmeouHon Hmmzzzx HzHomHemz .zmx....... M20. mzxm. HUMHM H3HH>HOM>QM 2 «Mon». HzM MUHumm zumzoszaMzHHzoMM... HMmmxoomomm Hammxzm mzHomHeMz .zmm.......moo. Msz. Human HzHH>HaH>oH z <402M.emm ezmxmsm.......oozzHMH>szH92Hom0H=mww.xHHQmmwo oomoz mewmo.o mzHomHeme MHomom.......omomxommo>¢m Hz©o>mzmmom . <=ozHH>mH Hamzmsmam.....owomHUHHHmazoozmmHHo ..moo zmaoz Hemozzz HzHomHomo HHHMoeemH....o.omoxeoo>mw Hzoo>MHMHoM a zwmozm>mm mammozmmoxuozommmmoomHmHMHMHHoozHamem..memnm. Mxoommzmwozxo mzHomHaHo amaomomou Figure 4-2. Representation of the CA Tand GUS genes following integration of pMON505 into the tobacco genome. The plant transformation vector pMON505-70 (Fang et al., 1989) contains CA T and GUS genes controlled by cepies of the full-size (-940 to + 9) 355 promoter (358). E9 3’ and 3C 3’ represent the 3’ ends of the pea rch-E9 and rch-SC genes, respectively. Five kilobases of vector DNA separates the two genes. The solid arrow represents the right T-DNA border. It is located 2 kilobases from the 3C 3’ end. 80 Legend to Figure 4-3: Decay of transcripts in tobacco (NT-1) cells stably transformed with pMON505—70. (A) RNA was isolated from samples harvested every 30 min after transcription was inhibited with 100 pg/mL ActD. RNA gel blots containing 10 pg of total RNA per lane were probed for the transcripts indicated on the left. (8) Signals from (A) were quantitated using a B—scanning device (Betagen), normalized to the zero time point, and subjected to linear regression analysis. Data is from a single experiment but is representative of the experiments listed in Table 4-2. These experiments were performed by Tom Newman 81 A CAT ...... GUS ... in». “‘ ...... ATPase maul 10_ : VCAT _ OGUS — lB-ATPase _ DHSP70 U) .E _ .5 (B E . l _ ID . - h - Cl < _ z - o: _ E i 01 . 1 . 1 . 1 1 4 0 3O 60 90 120 150 Minutes Figure 4-3. Decay of transcripts in tobacco (NT-1) cells stably transformed with pMON505-70. 82 Table 4-2. Messenger RNA Half-Lives in NT-l Cells Stably Transformed with pMON505-70 Absolute half- life (mins) Cell GUS CAT GUS/ Line Expt. Blot -3C -E9 CAT A 1 1 95 85 1.12 2 95 100 0.95 2 1 62 60 1.03 3 1 90 70 1.29 4 1 80 57 1.40 2 73 57 1.28 B 1 1 85 80 1.06 2 1 83 61 1.36 2 80 65 1.23 3 1 80 65 1.23 2 65 54 1.20 Mean° 80.4 68.9 1.18 2*: SD 10.3 13.1 0.14 °Values obtained from duplicate blots from the same experiment were averaged prior to calculation of the mean .Th ese experiments were perform ed by Tom Newman. 83 (Boutry and Chua, 1985) or heat shock protein 70 (Wu et al., 1988), neither transcript decayed appreciably over the 2.5 hr time course and are likely to have half-lives of greater than 4 hr. To assess the reproducibility of the data, a large number of independent half-life experiments were performed on each of two stably transformed lines as shown in Table 4—2. These data showed that the half-lives of GUS-SC and CA T-E9 were 80.4 :I: 10.3 and 68.9 :I: 13.1 min, respectively. Although the standard deviation in these determinations was less than 20%, the minimum variation between experiments was observed when the decay rate of GUS was expressed relative to that of CAT (1 .18 :l: 0.14). These data suggested that it should be possible to test putative instability sequences for function by inserting them into copies of one of these reporter genes, while using the other as an internal reference. It was also of interest to address whether the decay characteristics of the CA T—E9 and GUS-3C transcripts in NT-1 cells are reflective of what occurs in transgenic plants. For these experiments, a transformation vector containing CA T-E.9 and GUS-30 genes under the control of the wheat chlorophyll a/b binding protein (Cab-1) promoter (Nagy et al., 1987) was introduced into tobacco and transgenic plants were regenerated. Cab-1 is regulated by an endogenous circadian clock in transgenic tobacco, such that the promoter is activated just before dawn and then shuts off about midday (Nagy et al., 1988a). If the CAT and GUS transcripts decay with similar half-lives in transgenic plants, as the NT-1 cell experiments predict, then they would be 84 expected to disappear with approximately the same kinetics during the afternoon and evening when produced in transgenic seedlings under Cab-1 control. This is exactly what we observed as shown in Figure 4—4. Because the Cab-1 promoter is relatively weak compared to the 358 promoter used in NT-1 cells, we monitored the transcript levels in transgenic plants using a quantitative S1 nuclease protection assay (Kuhlemeier et al., 1987; Fang et al., 1989) to increase the sensitivity of detection (Figure 4-4A). When the transcript levels from 12:00 to 17:00 hr were subjected to linear regression analysis (Figures 4-48 and 4-4C), both CATand GUS transcripts disappeared with apparent half-lives of approximately 100 min. These values were similar to, but slightly longer than, the absolute half-lives of the CA T-E.9 and GUS-3C transcripts measured in NT—1 cells (70-80 min). The slight discrepancy may be due to minor differences in overall metabolic rates between the two systems and/or to the Cab-1 promoter not being completely shut-off during the period when the disappearance rates were measured (Giuliano et al., 1988; Millar and Kay, 1991). Nevertheless, the apparent correlation between the NT-1 cell and the transgenic plant data indicates that cultured NT-1 cells should provide an effective and expedient model system to investigate the effect of putative instability determinants such as DST sequences, on mRNA stability in plants. 85 Legend to Figure 4-4 Disappearance of transcripts encoded by CA Tand GUS genes expressed under the control of copies of the Cab-1 promoter in transgenic tobacco. (A) RNA was isolated from 2 week - old transgenic tobacco seedlings (T1) grown on kanamycin-containing medium at the indicated times. Forty micrograms of total RNA was hybridized to single - stranded DNA probes corresponding to the 5’ regions of GUS and CA Tand subjected to S1 nuclease analysis. Protected products corresponding to the GUS and CA Ttranscripts are indicated to the left of the gel. Positions of the undigested probe (CP, CAT probe; GP, GUS probe) are indicated to the right. WT-SR1 represents RNA isolated from untransformed wild-type SR1 seedlings. (B) The CAT protection products from (A) were quantitated and normalized to the 9:00 time point. Points from 12:00 to 17:00 were subjected to linear regression analysis. (C) Same as for (B) but the GUS protection products were quantitated. A we §§§§§§§§S§§ - _s._---'CP .- GP CAI ------ man"- '10E l‘ vCAT g 1 .. 1’ .5 f V v v I! E 2 v < V V E 0.1 ' E i 0.01I. 1 I 1 I 1 41_ 1 9:00 11 :00 13:00 15:00 17:00 Time 10 OGUS g _ E "E . 00 “I t E . 0 i- h g P m 0.11;- E g 1 t 0.01 I I 4 4 1 I 1 9:00 11 :00 13:00 15:00 17:00 Time Figure 4-4. Disappearance of transcripts encoded by CAT and GUS genes expressed under the control of copies of the Cab-1 promoter in transgenic tobacco. 87 Effoot of DST soooonoos on mRNA W To test directly the hypothesis that DST sequences can target transcripts for rapid decay in NT-1 cells, we inserted copies of a synthetic DST sequence into the 3’UTR of a GUS "test" gene. Each modified GUS gene was inserted into a pMON505 derivative containing an unmodified CATgene (designated p847 as described in Methods). The unmodified CAT gene on each vector was designed to serve as an internal reference. The basic structure of the final constructs shown in Figure 4-5 was nearly identical to pMON505-70 (Figure 4- 2) used in our pilot studies except that the SC 3’ end was attached to CA Tand the E9 3’ end was attached to GUS to facilitate cloning. Because many isolated sequence motifs function better when they are multimerized (Kuhlemeier et al., 1987; Green et al., 1988; Herr and Clarke, 1986), GUS genes containing one (DST x1) and two copies (DST x2) of the DST sequence were made. As discussed below, a second set of test genes was made with the human B—globin gene, using an unmodified GUS reference gene as the internal standard, as shown in Figure 4-5. Each construct was introduced into NT-1 cells and stably transformed cell lines were isolated. To examine the effects of the inserts on mRNA stability, suspension cultures were established from the transformants, and the degradation of the GUS and CATtranscripts were monitored on RNA blots following the inhibition of transcription with ActD as outlined above. The most marked effects on GUS mRNA stability resulted from the DSTx2 88 Insert Test gene —-[ ass [ GUS or Globin £9 3' Reference gene C—j 358 [ CAT or GUS l 3c 3' }-—> Figure 4-5. Representation of the test and reference genes following integration into the tobacco genome. GUS and globin test genes used to investigate the effects of DST inserts, and CAT and GUS reference genes used as internal standards are illustrated. To facilitate their transfer into the tobacco genome in the relative orientations shown, the test and reference genes were cloned into the polylinker and Hpal sites, respectively, of the plant transformation vector pMON505 (Rogers et al., 1987) as described in Methods. 89 insert that caused a rapid degradation of the GUS transcript relative to the no insert control. A representative set of experiments is shown in Figure 4-6. In the transformed cell line expressing the no insert control, the CA Tand the GUS transcripts decayed at approximately the same rate over a two-hour time course. In contrast, the GUS transcript containing two copies of the synthetic DST sequence (DST x2) was degraded much more rapidly than the CAT reference and the no insert control. The blot in Figure 4-68 was overexposed relative to that in Figure 4—6A because the abundance of the DST x2 transcript is also diminished. From the graphs in Figures 4-6C and 46D, the half-life values for the GUS transcripts containing no insert and DST x2 were calculated to be 120 and 37 min respectively. mRNA stability was also assayed in additional cell lines transformed with these and other constructs described below. The results of these measurements are shown in Table 4-3 and Figure 4-7. The decreased stability caused by the DSTx2 insert can be seen by comparing the absolute half-life (4.4 - fold decrease), or the relative half-life (3.6 - fold decrease) of GUS-DSTx2 to that of the GUS-no insert control. However, as expected based on our pilot studies with pMON505-70, the relative half-life values in Table 4-3 showed the least experimental variation because they were normalized to the decay of the internal reference transcripts. That is, experimental variations generally affect the test and reference mRNA half-lives coordinately. Therefore, in most cases, the SD values are lower and more consistent for the relative half-life values, than for the the absolute half- 90 Legend to Figure 4-6 Destabilization of the GUS transcript by DST sequences in stably transformed NT-1 cells. RNA was isolated from samples harvested from stably transformed cell lines every 30 min for a period of 2 hr after the addition of ActD as described for Figure 4-3. Gel blots containing 10pg of RNA per lane were probed with GUS (test) and CAT (reference) probes, and signals were quantitated using a B- scanning device (Phosphorlmager). (A) Gel blot of RNA isolated from a cell line transformed with a construct containing a GUS-no insert test gene and the CAT reference gene. (8) Gel blot of RNA isolated from a cell line transformed with a construct containing a GUS-DSTx2 test gene and the CAT reference gene. (C) and (D) Quantitative representation of the GUS test and CAT reference transcripts from (A) and (B) are shown in (C) and (B). respectively. The data are derived from single experiments with individual cell lines transformed with each construct, but are representative of the data summarizing multiple half-life measurements for independent cell lines shown in Table 4-3 and Figure 4—7. The blot in (B) was overexposed to allow the GUS-DST x2 transcript, which was present in reduced amounts, to be visualized. These experiments were performed by Masaru Takagi and Tom Newman 91 120’ A 0 120' B 0 GUS-nolnsert . ' ' ' . cusosrxz . . v CAT-REF . . . . . CAT-REF W O GUS-DST x 2 V CAT-REF O GUS-no Insert v CAT-REF IIIII IIII mRNA remaining 1 I / I L 1 l l l I 0 1 i l x l I l_ . .2 o 30 so 90 120 ' o 30 60 90 120 Minutes Minutes 0.1 Figure 4-6. Destabilization of the GUS transcript by DST sequences in stably transformed NT-1 cells. tags .52 Eek hem .533 .255 Emma: .3 6.250th .29: .9 26553.... 085 .P Co o2m> m nmcmfimm 2:9:an we; :02? .6580 tmmcmoc mcficonmotoo or: we $5 3 «027.50 some .2 ...mmbmok cam-2 ecu mc_~__mELo: >3 63230.8 995 327:2 m>_um_mmo .5538. cam—2 2: Co 53230.8 S .25 ummm6>m 99$ EmEtoaxm 0E8 9: E0: 303 33.39 :5: 35950 mo:_m> .uatomcmb Com: oocefiom 9: Co S5 .3 62:26 mm; Stomach “we... 65 Co 95.22 9: .55 Ecztoc comm Lon... div 2an E mm:_m> 35.2 3233 cums. 65 5.. costume“. we 63230.8 295 new .ow H 9522 9a mm:_m> .uozbmcoo some .8 mac: :8 cmELoCmcmb >33.» “concoaocg e>z 3 95 Co some :5: mucmEtogxm SE Eoctoc 95 ammo. pm Eo: 69th 9c; Ema. modfi 3.0 3.0”." and mmH nm N H mm om-m30 Nx kwo-z_mo._0 Nodflmmd modfimné mHNm mflm: om-m30 «x 803925040 wodH 55.0 m Conn 34 3 How manly: Um-m30 Fx 508925040 hodflwod mFdHoNé orflmm FNHNOF Um-m30 Fx Pmo-z_mo._0 2 m P.0Hooé ¢NdH¢wé mHow emu." m: Um-w:o tome 8-25040 9 3.0ku.0 modflmmd FmHNmF mHNm Om-._._um_mm iombmmk cam—2 <2mE Com .4sz amok mcmm Com mcom amok .AmEE. 8:-22 95.8% cams. poabmcou 2.8 52 5 3:53 203 degradation complex. Genes and Devel. 6. 1927-1939. Seeley, K.A., Byrne, D.H., and Colbert. J.T. (1992). Red light-independent instability of oat phytochrome mRNA in vivo. Plant Cell 4, 29-38. 111 Shaw, G., and Kamen, R. (1986). A conserved AU sequence from the 3’ untranslated region of GM-CSF mRNA mediates selective mRNA degradation. Cell 46. 659-667. Shirley, B.W., and Meagher, 8.8. (1990). A potential role for RNA turnover in the light regulation of plant gene expression: ribulose-1, 5-bisphosphate carboxylase small subunit in soybean. Nucleic Acids Res. 18, 3377- 3385. Shyu, A.-B., Greenberg, M.E., and Belasco, J.G. (1989). The c-fos transcript is targeted for rapid decay by two distinct mRNA degradation pathways. Genes and Devel. 3, 60-72. Siflow, CD. and Key, J.L. (1979) Stability of polysome-associated polyadenylated RNA from soybean suspension culture cells. Biochemistry 18.1013-1018. Stern, D.B., Jones, H., and Gruissem, W. (1989). Function of plastid mRNA 3’ inverted repeats: RNA stabilization and gene-specific protein binding. Taylor, CB. and Green, P.J. (1991). Genes with homology to fungal and S- gene ribonucleases are expressed in Arabidopsis thaliana. Plant Physiol. Theologis, A. (1986). Rapid gene regulation by auxin. Annu. Rev. Plant Physiol. 37. 407-438. Thompson, W.F. and White, M.J. (1991). Physiological and molecular studies of light-regulated nuclear genes in higher plants. Annu. Rev. Plant Physiol. Plant Mol. Biol. 42, 423-466. van der Zaal, E.J., Droog, F.N.J., Boot, C.J.M., Hensgens, L.A.M., Hoge, J.H.C., Schilperoort, R.A., and Libbenga, K.R. (1991). Promoters of auxin-induced genes from tobacco can lead to auxin-inducible and root tip-specific expression. Plant Mol. Biol. 16, 983-998. Walling, L., Drews, G.N., and Goldberg, RB. (1986). Transcriptional and post- transcriptional regulation of soybean seed protein mRNA levels. Proc. Natl. Acad. Sci. USA 83. 2123-2127. Wu, C.H., Caspar, T., Browse, J., Lindquist, S., and Somerville, C. (1988). Characterization of an HSP70 cognate gene family in Arabidopsis. Plant Physiol. 88. 731-740. 112 Yamamoto, K.T., Mori, H., and Imaseki H. (1992). cDNA cloning of indole-3- acetic acid-regulated genes: Aux22 and SAUR from mung bean (Vigna radiata) hypocotyl tissue. Plant Cell Physiol. 33, 93-97. CHAPTER 5 THE EFFECT OF SEOUENCES WITH HIGH AU CONTENT ON mRNA STABILITY IN TOBACCO This Chapter is in press in Proc. Natl. Acad. Sci. Reference: Ohme-Takagi, M., Taylor, C.B., Newman, T.C. and Green, P.J. (1993). The effect of sequences with high AU-content on mRNA stability in tobacco. P.N.A.S. (in press). 113 1 14 ABSTRACT Little is known about the mechanisms that target transcripts for rapid degradation in plants. In mammalian cells, sequences with a high A+ U content and multiple AUUUA motifs have been shown to cause mRNA instability when present in the 3’ untranslated regions of several transcripts. This precedent, coupled with the poor accumulation of AU-rich foreign transcripts in plants (e.g., BT-toxin mRNAs), prompted us to test whether AU sequences could destabilize transcripts in tobacco. To address this question, we made a set of constructs containing sequences with high A+U content inserted into the 3' untranslated regions of reporter genes. The stability of the corresponding transcripts was then assayed in stably transformed cell lines of tobacco. These experiments showed that a 60-base sequence containing 11 copies of the AUUUA motif ("AUUUA repeat") markedly destabilized a B-glucuronidase reporter transcript, compared to a no-insert control or a 60-base spacer sequence (GC-control). Another sequence with an identical A + U content had little effect. The same results were obtained when each sequence was assayed within the 3’ untranslated region of a B-globin reporter transcript. In regenerated transgenic plants, the AUUUA repeat decreased the accumulation of the B—globin transcript by about 14 fold, compared to the GC-control. Taken together, our results indicate that the AUUUA repeat is recognized as an instability determinant in plant cells and that the effect is due to the sequence of the element, not simply to the high A+U content. 1 15 INTRODUCTION An important aspect of the regulation of gene expression in all organisms is the control of cytoplasmic mRNA levels. In mammalian, plant, and yeast cells it is clear that for many transcripts, differences in mRNA stability contribute greatly to the establishment of steady-state mRNA levels and the speed at which those levels are achieved after a change in transcription rate [1-3]. In particular, proteins that are required by the cell only transiently are often encoded by very unstable mRNAs, generally with half-lives of less than 60 min. In mammalian cells, these include mRNAs encoding regulatory proteins such as c-myc, c-fos and granulocyte-monocyte colony stimulating factor (GM-CSF) [1,2]. The instability of these mRNAs allows for their rapid disappearance from the cytoplasm after a decrease in transcription. Efforts to elucidate mechanisms of mRNA turnover in mammalian cells have led to the identification of sequences that mediate the selective decay of specific transcripts. For some transcripts, such as those encoding histone H4 [4], B—tubulin [5], and transferrin receptor [6], the mechanisms that control mRNA stability are highly specialized and therefore involve unique regulatory sequences. In contrast, there is a common sequence characteristic among many of the Iymphokine, cytokine, and proto-oncogene mRNAs that are rapidly degraded. The 3’ untranslated regions (UTRs) of these transcripts generally contain a 30 to 80-base sequence that is rich in A and U residues and includes 1 16 copies of the pentamer motif AUUUA [7, 8]. When the AU domains of the GM- CSF [7] or c-fos [9] 3’ UTRs were inserted into the otherwise stable B—globin mRNA, the modified transcripts were rapidly degraded. Substitution mutations in natural and synthetic sequences indicate that AUUUA repeats in some of these transcripts can function as important determinants of mRNA instability in mammalian cells [10, 11]. However, there does not appear to be any clear relationship between the number of AUUUAs and the instability conferred by endogenous AU domains, perhaps due to differences in context [12]. AU domains in the 3’ UTR may also contribute to mRNA instability in yeast, but AUUUA sequences do not appear to be critical determinants [13]. The apparent lack of recognition of AUUUA sequences in a molecular genetic system presently limits the approaches that can be taken to address the mechanisms of AUUUA- mediated mRNA instability. A number of plant transcripts contain AU motifs that may contribute to post-transcriptional control. For example, it has been suggested that the par transcript of tobacco [14] may be relatively short lived because of the presence of three AUUUA sequences in the 3’ UTR, but this possibility has not been tested. Another AU motif, AUUAA, is repeated six times in the UTRs of the soybean AUX22 mRNA [15]. The stability of this transcript has not been directly measured, but its level decreases appreciably when excised tissue is incubated in auxin-free medium for 6 h [16]. Evidence suggesting that AU-rich transcripts may be rapidly degraded in plants has also resulted from 117 experiments aimed at engineering plants to express bacterial insecticidal protein genes. Several groups have demonstrated that the mRNAs encoded by insecticidal protein (BT-toxin) genes of Bacillus thuringiensis fail to accumulate in plants even when strong plant promoters are used to drive gene expression [e.g., 17, 18]. These BT-toxin transcripts are generally AU-rich and contain multiple copies of both AUUUA and AUUAA motifs that may contribute to mRNA instability, but other mechanisms limiting transcript accumulation (e.g. anomalous RNA processing or a block in transcriptional elongation) have not been ruled out. Thus, direct evidence linking sequences having high AU content with mRNA instability in plants is lacking. The only sequences that have been demonstrated to selectively target transcripts for rapid decay in plants are DST sequences [19], elements highly conserved in the 3’ UTRs of plant Small Auxin Up RNA (SAUR) transcripts [20]. These sequences are not AU-rich, nor do they contain conserved AUUUA or AUUAA motifs [19, 20]. In this study, we compare the effect of AU-containing and control sequences on mRNA stability in plants by direct measurement of mRNA decay rates in stably transformed cells of tobacco. Our results demonstrate that a 60 base sequence consisting of 11 overlapping repeats of the AUUUA motif can target transcripts for rapid degradation in tobacco cells and decrease transcript accumulation in transgenic plants. Furthermore, we show that the effect of AU-containing sequences on mRNA stability is sequence-specific and not simply due to high A+U content. 1 18 MATERIALS AND METHODS n n r i n Oligonucleotides were synthesized using an Applied Biosystems DNA Synthesizer, subcloned, and sequenced before insertion into the test genes as described in Figure 5-1. GUS test genes were inserted into a pMON505 derivative containing a 355-CAT-3C reference gene (p847; [19]; see Figure 5- 1). Globin test genes were inserted into a pMON505 derivative containing a 358-GUS-3C reference gene (p851; [19]). All constructs were introduced into Agrobacterium tumefaciens GV3111SE, as described previously [19]. T tran f rm i n An established line of Nicotiana tabacum (NT-1) cells [23] was transformed as described by An [23], as modified by Newman et al. [19]. After 3 to 5 weeks, individual stably transformed calli were transferred to fresh medium (NTKC; [19]), and the expression of the test and reference genes was confirmed by Northern blots [19]. For regeneration of transgenic plants, constructs were introduced into Nicotiana tabacum cv. Xanthi by leaf disc transformation [19]. Plants were grown and leaves were harvested for RNA isolation as described previously [19]. 119 RNA i l i n measur men f mRNA il' n 1 m In To establish suspension cultures of NT-cells for mRNA half-life measurements, stably transformed calli were transferred to liquid NT-medium and grown as described [19]. Three to four days after sub-culture, Actinomycin D (ActD) was added to a concentration of 100 pg/ml and 10 ml samples were withdrawn every 30 min for 2.5 h. Cells were harvested, frozen in liquid nitrogen, and RNA was isolated and analyzed on Northern blots [19, 24]. The blots were hybridized to random-primed probes corresponding to the coding regions of the appropriate test and reference genes [19]. Signals were quantitated with a Phosphorlmager and linear regressions were plotted to calculate the half-lives of the individual mRNAs [19]. S1 nuclease protection analyses were performed as described previously [25], except that the annealing reactions contained 70% formamide and 40 pg of total RNA. RESULTS We were interested in determining whether sequences with high A+ U content could selectively target transcripts for rapid decay in plants and, if so, whether the sequence of the element was important or simply the high AU-content. Testing the function of AUUUA sequences in plants was of particular interest because these elements can cause mRNA instability in mammalian cells, but 120 have not been shown to function in other systems. To this end, we prepared the series of constructs illustrated in Figure 5-1. Beginning with a 35S-GUS-E9 test gene (designated "no insert"), three derivatives were made, each containing a different 60-bp fragment inserted into the 3’ UTR between the GUS coding region and the E9 3'end. As shown in Figure 5-18, two of the constructs, the AUUAA repeat and the AUUUA repeat, contained inserts with identically high A+U content, but different primary sequences. The AUUAA repeat was designed to test whether AUUAA sequences, such as those repeated in the UTRs of the soybean AUX22 gene [15], could destabilize transcripts in plants. To investigate specifically the function of AUUUA sequences in plants, a 60 base insert containing the maximum possible number of AUUUA motifs (11 overlapping pentamers) was made and designated the AUUUA repeat. As a size control, an insert interspersed with Gs and Cs was synthesized and designated the GC-control [7]. Each of the GUS test genes was incorporated into a transformation vector containing an identical 35S-CAT- 3C reference gene that served as an internal standard to facilitate comparisons between experiments. These constructs were introduced into a tobacco cell line in which mRNA decay rates could be directly measured. The NT-1 cell line was chosen for these experiments because it can be grown as a suspension culture in liquid media and can be stably transformed with foreign genes via A. tumefaciens [23]. In addition, ActD, a transcription inhibitor commonly used to facilitate the 121 Legend to Figure 5-1 Constructs used to test the effect of AU-containing sequences. (A) The vectors used for plant transformation included a test gene to test the insert of interest and a reference gene to serve as an internal standard. In the first series of constructs, the test gene consisted of the B-glucuronidase (GUS) gene under the transcriptional control of the 358 promoter (-941 to + 8) from cauliflower mosaic virus (358) as described previously [19]. Polyadenylation signals were provided by the 3’ end of the RBCS-E9 gene (E9 3’) [19]. Inserts were cloned into a unique BamHI site between the GUS coding region and the E9 3’ end of the test gene. A 35S-GUS-E9 test gene without an insert was designated the "no insert" control. All vectors containing GUS test genes contained an identical reference gene consisting of a chloramphenicol acetyltransferase (CAT) gene flanked by the 358 promoter and RBCS- 3C 3’ end (3C 3’) [19] as shown. The test and reference genes were inserted into the polylinker (between Sacl and Clal), and the Hpal site of the binary vector pMON505 [21], respectively. A second set of constructs (not shown) was made in the same manner but contained a 35S-globin-E9 test gene and a 358-GUS-3C reference gene. Globin sequences were derived from a human ,B-globin cDNA clone [22] modified to remove the BamHI site within the coding region [19]. (8) Synthetic sequences inserted into the test genes contained a Bglll site and a BamHI site at the 5’ and 3’ ends, respectively; the GATC sequences within these sites are underlined. The pentamer motifs, AUUAA and AUUUA, found within the AUUAA repeat (rpt) and AUUUA repeat (rpt) are indicated by arrows. 122 .mmocoaaom mcEfiEooj< B “cote 9: “m3 3 com: muoacumcoo . Tm 2:2“. me0 wozwmmuwm *\\\\%§ A... Mushaowm.» m «02 on \m ta «8 Ema Al cameefiaeameeamessays:sedeaeMaeenaeemaeenaaefie3:9536159mm .EL <:::< Al Al All All AIII oalImIo63999959953afieefieafieefieeieefie525529955545.36 ”a. <<==< .8232 «xx OB‘UOEBBBUBBBOGawafiflmvBBflBflBUflBUBGBflflwBOBdOdBflflUn—vdfigawflaofiflw 20.5200 00 % wzwc ...mw... m < 123 measurement of mRNA decay rates in cultured mammalian cells (e.g., [7, 12]), is an effective inhibitor in NT-1 cells [19]. To measure decay rates in NT-1 cells, stably transformed cell lines containing each of the constructs in Figure 5-1 were grown in suspension culture and treated with ActD. RNA was isolated from samples removed from the cultures at various times as described in Materials and Methods. Degradation of the test and reference transcripts was monitored on Northern blots by hybridization to GUS and CAT probes, as described [19]. For each construct, two to four independent stably transformed cell lines were analyzed and four to eight Northern blot experiments were performed. Among all of the sequences tested, only the AUUUA repeat was found to affect significantly the stability of the GUS test transcript in NT-1 cells. As the Northern blots in Figure 5-2 indicate, the GUS transcript containing the AUUUA repeat (GUS-AUUUA) was degraded considerably faster than the GUS transcript containing the GC-control (GUS-GC). Table 5-1 shows the mRNA half-life measurements for the test (GUS) and reference (CAT) transcripts in cell lines transformed with the GUS-AUUUA and GC-control constructs. We also normalized the half-life of test transcript to that of the reference transcript to calculate the relative half-life in each case. While the decreased stability of the GUS-AUUUA transcript can be observed by comparing either value, we have found previously [19] that the relative half-life calculations most effectively normalize for cell—line to cell-line variations that are periodically observed (e.g., .50<0k-0Et0 50003 >0 .00EBE0Q 00>. .9 :0Et0QX0 00.05 .>_0>_~00000._ 65:00:05 0:0m EmE 00:220.. 0:0 0.000. 05 “000.00 3 0000.5 :060: @5000 .56 0:0 mDO 9. 00325? 0.03 00.05 3:05.00: 00.04. .0000 00E; 000.0065 05 0 000.0050: 0:00 :5: 000.062 <2: a: 0.0. 00:_0E00 0:0. :00m 0:00 T._.z 00:05.5:me >305 :_ :6 030:. *0 N 0:: :00 “(533.4039 000:0. 800 06. 05000:. 9. 000: 3:05.580 53 E055: 0209:0350”. 0:00 000029. :_ 000:0: <333< 05 >0 000300 >0000 _0>_000000. .000_.00:0.0 00:0.000. 0:0 0000 0:0 000000 00 0000.: 0.050. 05000 mac 0:0 £003 00 005000.»: 0.0.5 0005 .0:0E.000:0.0 0:00:00005 :0 E90 <2: 00 a: 00 00:00:00 0:0_ :00m .000:.00:00 (3302-556 .0 00:30.0 0:0 0:_000.0x0 00:03 2:000:00. 00 0002 E0: 0000.00_ 00.5 <2: 00:03 2:000:00 E 00000. <333< 0:0 00 000.05 .m-m 0.00.”. _ 0:00.0u D A 2.0000 8.353. . I a.‘.. n 2530'. ....II1II... I .. , . 1 .. . ._ ,. .. fl. -... . .0: 000 Y .D 0.... i. ......D D D . ‘._D.A.mm 0:0 _ h 240 min half-life exhibited by a GC control in mammalian cells [11]. In another example, insertion of sequences from the GM-CSF 3’UTR into a globin reporter transcript decreased its half-life to <30 min [7]. The GC control in this case had a half life of >120 min. Therefore it appears that the degradation kinetics of globin transcripts containing the AUUUA repeat in plants are similar to those of transcripts containing AU-rich instability determinants in mammalian cells. Our results demonstrated that the instability caused by the AUUUA repeat in tobacco is not the result of altered polyadenylation because all the transcripts have the expected size and 3’ ends. Therefore, the increase in mRNA degradation observed in stably transformed tobacco cell lines is due to the direct effect of the element on GUS and globin mRNA decay. Because the AUUUA repeat also decreases mRNA levels in leaves of transgenic plants, it is highly likely that the element acts as an mRNA instability determinant in intact 134 plants as well. At present it is unknown why the magnitude of the effect was greater when mRNA accumulation was measured in transgenic plants (14 fold decrease) than when mRNA half-lives were measured in cultured cells (3.4 fold decrease). This difference is not due to an inherent difference between NT-1 cells and plants because in NT-1 cells the accumulation of the globin-AUUUA repeat mRNA is also decreased more than 10 fold compared to that of the globin-GC control transcript (M.L. Sullivan and P.J.G., unpublished). A similar discrepancy (of up to 10 fold) between mRNA accumulation and mRNA half- lives was observed recently when an AU instability determinant was investigated in stably transformed mammalian cells [26]. in the latter study, it was suggested that mRNA half-lives measured in the presence of ActD may have been over estimated if recognition of the AU sequence was dependent on ongoing transcription. A similar hypothesis may also explain the discrepancy between mRNA levels and half-life measurements in tobacco. Alternatively, the difference could indicate that the AUUUA repeat limits transcript accumulation by more than one mechanism. In order to differentiate between these possibilities, we are presently developing methods for direct measurement of mRNA decay rates in intact transgenic plants and cultured cells using regulated promoters, rather than transcriptional inhibitors. Although it is significant that both plants and mammalian cells have mechanisms to rapidly degrade transcripts containing AUUUA repeats, it should also be noted that all AUUUA sequences do not cause mRNA instability. 135 Several years ago Shaw and Kamen [7] and Caput et al. [8] pointed out that AUUUA sequences were common to the AU-rich 3’ UTRs of many unstable Iymphokine and protooncogene transcripts, but no correlation has been established between the degree of mRNA instability and the number of AUUUA sequences that are present in these transcripts. Moreover, AUUUA sequences are not excluded from stable mammalian mRNAs [12]. Another consideration is that even when an AUUUA containing sequence is found to cause mRNA instability, it may not do so constitutively. For example, an AU-rich sequence in the 3’UTR of the B—interferon transcript mediates hormone-induced mRNA destabilization [27], and the GM-CSF 3’UTR does not cause mRNA instability in all mammalian cell lines [28]. It has been proposed that the arrangement or context of AUUUA sequences may control differential recognition of mammalian AUUUA sequences [12]. Nonfunctional AUUUA repeats are also likely to exist in plants because we have observed that some sequences that contain AUUUA motifs, albeit fewer and in a different context than the AUUUA repeat, do not cause mRNA instability in tobacco (M.O.-T., C.B.T., G. Coruzzi, and P.J.G., unpublished results). Our strategy in constructing the AUUUA repeat as a potential instability determinant was to maximize the number of elements within the insert (because multiple elements work better for some transcription factors [29, 30] and the DST instability determinant in plants [19]), and to place them in an overlapping arrangement as found in some natural mammalian elements 136 [7]. Although this strategy was effective, a systematic study will be necessary in both mammalian and plant cells to identify the critical parameters for AUUUA function in the two systems. Another aspect that should be considered is whether secondary or higher order structure of AUUUA repeats can influence their function. Higher order structure is difficult to determine experimentally and that of the transcripts investigated in this study is unknown, but the secondary structure potential in each of the 60 base inserts has been analyzed (unpublished results). Only the AUUAA repeat insert has sufficient internal complementarity to form a stable stem-loop structure, and it is possible that this structure contributes to the lack of function of this insert. Recognition of the AUUUA repeat as an instability determinant in plant cells most likely involves one or more sequence-specific RNA-binding factors or ribonucleases. In mammalian systems, several RNA-binding factors have been identified, via UV crosslinking or gel retardation assays, that interact with AUUUA multimers or U-rich sequences. These include AUBF from Jurkat cells [31, 32]; the AU-A, AU-B, and AU-C proteins from human T cells [33, 34]; a 32-kDa AU-binding protein from HeLa cells [11, 35] that may be identical to AU-A [34]; and a set of U-rich sequence-binding proteins (URBPs) from mouse 3T3 cells [36]. A cytosolic factor from human erythroleukemia K562 cells has also been identified that binds to U-rich sequences such as the 3’ end of c-myc mRNA [37]. The latter activity has been reported to destabilize c-myc mRNA 137 in vitro [37]. While the in viva role of all of these RNA-binding activities remains unknown, several are inducible and therefore may have a regulatory function [33, 37]. AUUUA motifs have been reported to destabilize spliced and unspliced transcripts [11], and some of the RNA-binding proteins that interact with the element are nuclear [11, 33, 35]. This indicates that AUUUA elements may be recognized in both the nucleus and the cytoplasm. The identification of the AUUUA repeat as an instability determinant in tobacco should facilitate the biochemical investigation of cellular factors that interact with the element in plants. The properties of the plant factors can then be compared with AUUUA-binding proteins identified in mammalian cells. The most significant result of this study is the demonstration that mechanisms that recognize AUUUA-containing instability determinants extend beyond mammalian cell systems, and more importantly, that such mechanisms are present in organisms that are more amenable to molecular genetics. Genetic approaches have led to major breakthroughs in our understanding of other pathways in RNA metabolism such as splicing [38, 39], but have not been applied to the study of AUUUA recognition because of the technical limitations posed by mammalian systems and the apparent lack of function of the element in yeast. However, in model plants such as Arabidopsis thaliana [40], it should be feasible to search for genetic determinants that mediate decay of transcripts containing AUUUA repeats. For example, it may be possible to engineer selectable marker genes with AUUUA repeats and then 138 select for mutants that lose the ability to degrade rapidly the corresponding transcripts. If the mechanisms for degradation of AUUUA-containing transcripts are similar in plants and mammalian cells, the characterization of such mutants should have broad significance. In addition, the ease of generating transgenic plants should allow the metabolism of AUUUA containing transcripts to be studied in fully intact organisms throughout development. ACKNOWLEDGEMENTS We are grateful to Dr. Nam-Hai Chua for synthetic Oligonucleotides, and Drs. Michael Sullivan and Natasha Raikhel for comments on the manuscript. This work was supported by grants from the United States Department of Agriculture, the Department of Energy, and the Midwest Plant Biotechnology Consortium and matching funds to P.J.G. M.O.-T. was supported in part by a fellowship from the Human Frontier Science Program. 10. 11. 12. 13. 14. 15. 16. 1 39 LITERATURE CITED Atwater, J.A., Wisdom, R. & Verma, I.M. (1990) Ann. Rev. Genet. 24, 519-541. Peltz, S.W., Brewer, G., Bernstein, P., Hart, P.A. & Ross, J. (1991) Crit. Rev. Eukaryotic Gene Expression 1 , 99-126. Green, P.J. (1993) Plant Physiol. 102, 1065-1070. Pandey, N.B. & Marzluff, W.F. (1987) Mol. Cell. Biol. 7, 4557-4559. Yen, T.J., Machlin, P.S. & Cleveland, D.W. (1988) Nature (London) 334, 580-585. Miiller, E.W. & Kiihn, LC. (1988) Cell 53, 815-825. Shaw, G. & Kamen, R. (1986) Cell 46, 659-667. Caput, D., Beutler, B., Hartog, K., Thayer, R., Brown-Shiner, S. & Cerami, A. (1986) Proc. Natl. Acad. Sci. USA 83, 1670-1674. Shyu, A.-B., Greenberg, M.E. & Belasco, J.G. (1989) Genes Dev. 3, 60- 72. Shyu, A.-B., Belasco, J.G. & Greenberg, ME. (1991) Genes Dev. 5, 221-231. Vakalopoulou, E., Schaack, J. & Shenk, T. (1991) Mol. Cell. Biol. 11, 3355-3364. Greenberg, M.E. & Belasco, J.G. (1993) in Control of Messenger RNA Stability, eds. Belasco, J.G. & Brawerman, G. (Academic, New York), pp. 199-218. Muhlrad, D. & Parker, R. (1992) Genes Dev. 6, 2100-2111. Takahashi, Y., Kuroda, H., Tanaka, T., Machida, Y., Takebe, I. & Nagata, T. (1989) Proc. Natl. Acad. Sci. USA 86, 9279-9283. Ainley, W.M., Walker, J.C., Nagao, R.T. & Key, J.L. (1988) J. Biol. Chem. 263. 1058-1066. Walker, J. & Key, J. (1992) Proc. Natl. Acad. Sci. USA 79, 7185-7189. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 140 Vaeck, M., Reynaerts, A., H6fte, Jansens, S., De Beuckeleer, M., Dean, C., Zabeau, M., Van Montagu, M. & Leemans, J. (1987) Nature (London) 328, 33-37. Fischoff, D.A., Bowdish, K.S., Perlak, F.J., Marrone, P.G., McCormick, S.M., Niedermeyer, J.G., Dean, D.A., Kusano-Kretzmer, K., Mayer, E.J., Rochester, D.E., Rogers, S.G. & Fraley, R.T. (1987) Biotechnology. 5, 807-813. Newman, T.C., Ohme-Takagi, M., Taylor, C.B. & Green, P.J. (1993) Plant Cell 5, 701-714. McClure, B.A., Hagen, G., Brown, 0.8., Gee, M.A. & Guilfoyle, T.J. (1989) Plant Cell 1. 229-239. Rogers, S.G., Klee, H.J., Horsch, R.B. & Fraley, R.T. (1987) Methods Enzymol. 153, 253-277. Lang, K.M. & Spritz, R.A. (1987) Mol. Cell. Biol. 7, 3428-3437. An, G. (1985) Plant Physiol. 79, 568-570. Taylor, C.B. & Green, P.J. (1991) Plant Physiol. 96, 980-984. Fang, R.-X., Nagy, F., Sivasubramaniam, S. & Chua, N.-H. (1989) Plant Cell 1, 141-150. Savant-Bhonsale, S. & Cleveland, D.W. (1992) Genes Dev. 6, 1927- 1939. Peppel, K., Vinci, J. & Baglioni, C. (1991) J. Exp. Med. 173, 349-355. Schuler, G.D. & Cole, MD. (1988) Cell 55, 1115-1122. Kuhlemeier, C., Fluhr, R., Green, P.J. & Chua, N.-H. (1987) Genes Dev. 1, 247-255. Green, P.J., Yong, M.-H., Cuozzo, M., Kano-Murakami, Y., Silverstein, P. & Chua, N.-H. (1988) EMBO J. 7, 3635-3642. Malter, J.S. (1989) Science 246, 664-666. Gillis, P. & Malter, J.S. (1991) J. Biol. Chem. 266, 3172-3177. 33. 34. 35. 36. 37. 38. 39. 40. 141 Bohjanen, P.R., Petryniak, B., June, C.H., Thompson, C.B. & Lindsten, T. (1991) Mol. Cell. Biol. 11, 3288-3295. Bohjanen, P.R., Petryniak, B., June, C.H., Thompson, C.B. & Lindsten, T. (1992) J. Biol. Chem. 267. 6302-6309. Meyer, V.E., Lee, S.|. & Steitz, J.A. (1992) Proc. Natl. Acad. Sci. USA 89. 1296-1300. You, Y., Chen, C.-Y.A. & Shyu, A.-B. (1992) Mol. Cell. Biol. 12, 2931- 2940. Brewer, G. (1991) Mol. Cell. Biol. 11, 2460-2466. Ruby, S.W. & Abelson, J. (1991) Trends in Genet. 7, 79-85. Guthrie, G. (1991) Science 253, 157-163. Somerville, C. (1989) Plant Cell 1, 1131-1135. CHAPTER 6 IDENTIFICATION AND CHARACTERIZATION OF GENES WITH UNSTABLE TRANSCRIPTS (Ms) IN TOBACCO 142 1 43 ABSTRACT Plants and other higher eukaryotes have the ability to recognize and target specific transcripts for rapid decay from among the majority of relatively stable mRNAs present within cells. However, little is known about the nature of unstable transcripts in plants, nor the mechanisms that facilitate their rapid degradation. As a first step towards understanding how plants distinguish between unstable and stable transcripts, a novel differential screen was used to identify cDNAs for Genes with unstable Iranscripts (GUTs), solely on the basis of the instability of their mRNAs. cDNA probes were prepared from tobacco cells that had been depleted of highly unstable mRNAs by treatment for 90 mins with a transcriptional inhibitor, and from untreated control, untreated cells. GUT clones were selected on the basis of weak hybridization to the former probe relative to the latter probe. Half-life measurements performed on the mRNAs hybridizing to eight GUT clones indicated that each was unstable, with a half-life on the order of an hour or less. All eight of the cDNAs corresponded to new tobacco genes, and four showed sequence similarity with genes from other species, including the eukaryotic family of DNAJ homologs, a tomato wound-inducible protein, and histone H3. In addition to providing information about the types of transcripts that are inherently unstable in plants, the GUT clones should provide excellent tools for the identification of and cis- and trans-acting determinants of mRNA instability. 144 INTRODUCTION Control at the level of mRNA stability is an important component in the regulation of many eukaryotic genes [1,2,3]. This is because mRNA decay rates define the rapidity with which new steady-state transcript levels can be achieved following changes in the transcription rate of a gene. Most mRNAs in higher eukaryotes appear to be relatively stable, with half-lives on the order of hours [4]. Unstable transcripts, with half-lives on the order of minutes, tend to encode proteins that are required only transiently in cells, such as those with critical roles in the regulation of cellular growth and differentiation. Rapid alterations in the synthesis of these proteins is thought to be facilitated by the instability of the corresponding mRNAs [3]. A number of sequence elements have been identified within unstable transcripts that mediate the targeting of those mRNAs for rapid degradation, presumably via their interactions with specific binding factors [1, 2]. One well characterized example is a sequence element termed the AU-rich element (ARE), which consists of a stretch of 40 - 60 A and U bases, including one or more copies of an AUUUA pentamer motif. AREs have been shown to play an important role in the rapid degradation of certain Iymphokine, cytokine and proto-oncogene transcripts (reviewed in [2]). Recently, a number of proteins have been identified in mammalian cells that specifically bind to AREs (reviewed in [2]). Some of these proteins may be involved in the recognition of ARE-containing unstable transcripts and their 145 targeting for rapid degradation. Control of gene expression at the level of mRNA stability is likely to be particularly important in plants [1]. As sessile organisms plants are unable to move away from adverse stimuli, and are obliged to respond by altering endogenous gene expression, often very quickly. In addition to genes with a role in growth and development, genes involved in rapid plant responses are therefore apt to encode unstable transcripts. The functions of the few plant proteins that are known to be encoded by unstable transcripts generally support this contention. These include PvPRP1, which participates in the response of bean cells to pathogen attack, (the half-life of its transcript appears to decrease markedly following the treatment of bean cells with an elicitor [5]), phytochrome, which mediates many light regulated responses [6]. and the Small _A_uxin _U_p fiNAs (SAURs) of soybean, which encode proteins of unknown function, but whose expression properties are consistent with a role in auxin- induced cell elongation [7, 8]. However, the mechanisms that target these and other unstable transcripts for rapid degradation in plants remain to be elucidated. Initial insights have been provided by recent studies examining the effect of specific sequence elements on reporter transcript stability in transformed tobacco cell lines [9]. In these experiments, transcription was halted by the addition of Actinomycin D (ActD), and the subsequent decay of the reporter transcripts was monitored by Northern blotting [9]. Using this approach two 146 very different sequence determinants have been identified that can confer instability on mRNAs in plants. One, termed DST, is highly conserved within the 3’ UTRs of all the SAURs reported to date, it’s position and sequence conservation suggesting that it may mediate post-transcriptional control of SAUR expression [10]. When present in two copies in the 3'UTRs of reporter genes, the DST element leads to a pronounced reduction in mRNA half-life [9]. The other sequence, consisting of 11 copies of the AUUUA motif, was also found to markedly decrease reporter transcript half-lives [11]. These observations demonstrate that sequence-specific recognition of unstable transcripts occurs in plants, but it is unlikely that the DST and AUUUA elements are representative of the repertoire of instability determinants present in plant transcripts. To obtain direct information about the complexity of different mechanisms by which specific endogenous transcripts are selected for rapid degradation, the identification and study of a variety of unstable plant transcripts will be necessary. As a first step towards this goal, we report here the application of a novel differential screening approach to identify Genes with Unstable Iranscripts (GUTs) in tobacco. The GUTs, which were identified solely on the basis of the instability of their transcripts, provide tools for the further characterization of mRNA decay pathways, and have yielded additional information concerning the kinds of proteins that are encoded by unstable transcripts in plants. 147 MATERIALS AND METHODS Pr r in fRNA n DNAPr Culture conditions for the Nicotiana tabacum cv bright yellow 2 (NT-1) cells [12] used in this work have been described elsewhere [9]. To generate cDNA probes for the differential screen, duplicate NT—1 cultures at 3 days post sub- culture were left untreated (control cells). or treated for 90 minutes with the transcriptional inhibitor ActD at 100pg/mL (this concentration of ActD was previously shown to inhibit incorporation of[3H]-uridine into poly (A)" RNA by 94% [9]). Cells were harvested by centrifugation at 700 x g and frozen in liquid nitrogen. Total RNA was isolated essentially as described [13], except that the extraction solution was buffered with 80mM Tris at pH7.5 (in the place of 25mM sodium citrate, pH 7.0) and a second CsCI cushion, at 2.8M, was included in the gradient. PolylA)+ RNA was prepared from these samples using oligo dT columns (5 Prime - 3 Prime, lnc., Boulder, Co.), and 1119 of each Poly(A)+ RNA was used as template for the preparation of random-primed first- strand cDNA probes, as described [14]. l ol i n f T I n The NT-1 cell cDNA library used for screening for GUTs (the generous gift of Dr. G. An) was prepared from RNA isolated from NT-1 cells 3 days after subculture, and cloned into AZAP ll (Stratagene) with EcoRl-Notl adaptors. The 148 library was plated at low density (5,000 plaques per 150 mm plate). Duplicate lifts were prepared, using standard protocols [14], and were prehybridized and hybridized as described [15]. Promising clones were plaque purified and zapped into plasmid form [16]. For Northern and Southern blot analyses, the GUT inserts were labelled with [32F] by random priming [17]. Half-Life Determinations Determinations of GUT mRNA half-lives were performed as described previously [9]. Briefly, NT-1 cell lines expressing a 3SS-GUS-E9 reference gene [9] were treated with ActD at 100ug/mL, and RNA was isolated from aliquots of NT-1 cells harvested from these cultures every 30 mins thereafter. Total RNA was subjected to Northern blotting and hybridized to GUT and GUS probes, as described [9]. Ouantitation of GUT and GUS mRNA half-lives were performed using a phosphorimager as described previously [9]. At least 2 Northerns were performed on at least two independent cell lines with each GUT probe. In some cases, absolute half-life determinations were performed using untransformed cells. W DNA was isolated from Nicotiana tabacum, strain SR1 using an SDS/ phenol extraction procedure [18]. DNA was dissolved in dde, and aliquots were digested with the restriction enzymes EcoRl, EcoRV or Hindlll. 10pg of 149 digested DNA was separated on 1% agarose gels and transferred to Biotrace HP membrane, as recommended by the manufacturer (Gelman). Prehybridization, hybridization and washing of the Southern blots was performed using the same conditions as for the Northern blots. mm Sequencing was performed at the Plant Biochemistry Facility at the Plant Research Laboratory. Plasmid DNA from each GUT cDNA clone (or subclone) was prepared using Magic miniprep columns (Promega), and sequenced using Taq cycle sequencing and the Applied Biosystems Inc. 373A automatic sequencer, as recommended by the manufacturer. Nucleotide sequences of 250-300 bp from each end of the cloned GUT fragments were compared to sequences in the protein databases using the BLASTX algorithm under default parameter settings [19]. The nucleotide sequences of any GUT clones for which no database matches were found at the amino acid level, were compared to the nucleotide databases using the BLASTN algorithm [19]. To search for any conserved motifs among the GUTs, the GUT cDNA sequences were compared to one another using DNAsis (Hitachi). 1 50 RESULTS AND DISCUSSION lenifiainof TDNA In We have found previously that treatment of NT-1 cells with actinomycin D (ActD) effectively inhibits RNA polymerase II transcription [9], and have used this inhibitor as a tool to measure the half-lives of reporter transcripts in the absence of mRNA synthesis [9, 11]. In an analogous fashion, the levels of endogenous unstable transcripts should also decrease rapidly over time following the treatment of NT-1 cells with ActD. For example, a transcript with a half-life of 20 mins would be present at only 6% of it’s initial level after 90 mins of ActD treatment. In contrast, the levels of an average transcript, with a half-life of several hours [3, 4] would remain essentially unchanged after 90 mins of ActD treatment. This wide discrepancy between levels of unstable transcripts in 90 min ActD-treated and in control cells formed the basis for the differential screen described in this report. cDNA probes prepared using PolylA)+ RNA from ActD-treated cells will be essentially depleted of sequences corresponding to highly unstable transcripts, whereas control cDNA probes will contain a full complement of NT-1 mRNA sequences. The control probe should therefore hybridize to each cDNA clone in the library, but the 90 min ActD probe should hybridize weakly or not at all to plaques harboring cDNA clones derived from unstable transcripts. A total of approximately 300,000 plaques of an NT-1 cell cDNA library 151 were screened with control and 90 min ActD cDNA probes, as described in Materials and Methods. To minimize the selection of false positive clones, the first lift from each library plate was hybridized to the 90 min ActD first-strand cDNA probe, and the second lift to the control probe. 150 plaques with significantly decreased hybridization to the 90 min ActD probe, relative to that of the control probe, were isolated and rescreened. Of these, 7 were designated potential GUTs, as the plaques continued to show a diminished hybridization signal with the 90 min ActD probe. Purified plaques of potential GUTclones were zapped into plasmid form, and the plasmid DNA was digested with the restriction endonuclease EcoRl. In cases where there was more than one EcoRI fragment in a clone, the fragments were subcloned. This was because, in most cases, Southern analyses indicated that the fragments were derived from a fusion of two or more cDNAs in one A clone (see below). Further evaluation of the potential GUT clones was performed by examining transcript levels in the control and ActD-treated NT-1 cells. The EcoRI fragments from the potential GUT clones were labelled with [32P]-dCTP to similar specific activities, and hybridized to Northern blots of total RNA from control and ActD treated NT-1 cells (the same RNAs used for the isolation of poly(A)+ RNA for the generation of cDNA probes for the library screening). Ouantitation of the Northern blots shown in Figure 6-1 demonstrated that from the original 7 A clones, a total of 8 clones (3-2, 7-2A, 7-2C, 8-1, 8-2A, 8-2B, 8-3, 15) could be designated GUTs as they exhibited a 2 2.5 fold decrease in 152 Legend to Figure 6-1: Preliminary Characterization of Potential GUTs Northern blots of 10pg total RNA isolated from control, untreated NT-1 cells (0), and from cells treated for 90 minutes with ActD (90’) or with ActD and cycloheximide (CX) were hybridized to the indicated GUTand N0 TGUT probes. Transcript sizes were calculated relative to RNA molecular weight markers. GUT and NOTGUT transcript levels under each experimental condition were quantitated using a phosphorimager (see Table 6-1). 0 90' CX GUT 3-2 1.1-v. w. GUT 7-2a 1.8-» g ' GUT 7-2c “+000 O 90’ CX NOT GUT 7-2b 1.0 —> . 153 0 90' CX GUT 8-2a 1.4+... GUT 8-2b 1'0+.§. GUT 15 1.9 -> n.'.' O 90' CX NOT GUT 10-3 1.0 -> ". Figure 6-1. Preliminary Characterization of Potential GUTs 154 Legend to Table 6-1 ‘ The fold increase (1) or decrease (I) in GUT and NOTGUT mRNA in NT—1 cells treated for 90 minutes with Actinomycin D (ActD) or with ActD plus cycloheximide (CHX) was calculated relative to the mRNA levels in control NT-1 cells using the Northern blots in Figure 6-2. 2The absolute mRNA half-lives are averages :l: standard error of at least three half-life determinations for each GUT and NOTGUT, with one exception (GUT 15 half-life has been determined once). In addition to measurements made in the transformed NT-1 cell lines, these include half-life determinations in untransformed NT-1 cells. 3Relative half-lives were calculated by measuring GUT or NOTGUT and GUS mRNA half-lives, as described in the legend to Figure 6-3, in NT-1 cell lines transformed with a GUS reference construct used previously as an internal standard [9]. Values are average ratios of GUT or NOTGUT half-life to GUS half-life, :I: standard error, calculated from at least three different Northern blots for each GUT and NOTGUT. Table 6-1: GUT and NOTGUT mRNA Levels and Half-Lives ActD CHX Absolute Relative GUT Effect Effect Half-Life Half-Life (Fold)‘ (Fold)‘ (GU712 (GUT/GUSI3 L 3-2 3.0) 1.5? 65’ :I: 30’ 1.26 :l: 0.30 7-28 2.81 1.41‘ 74 :l: 19’ 0.92 :t 0.13 7-2c 2.2) 1.51 56’ a: 13' 0.84 :l: 0.19 II 8-1 4.0.) 2.01 44’ :I: 16' 0.77 :l: 0.20 8-2a 2.6l 1.1) 49’ :I: 20’ 0.79 :I: 0.10 l 8-2b 2.8I 2.0+ 55' :I: 16' 0.86 :I: 0.39 8-3 3.0l 1.0 I 52' :l: 06' 0.79 :I: 0.16 15 i) 18.0) 6.0I 43’ 0.36 I ll) 5.5I 6.5f 62’ 0.52 N0 TGU T ##— 7-2b 1.1I 2.3‘I >3hr >3 | 10-3 1.1l 3.1I >3hr >3 II 156 transcript abundance during the 90 mins of ActD treatment (see Table 6-1). This corresponds to a theoretical half-life of 72 mins or less. Two clones (7-2B and 10-3) that showed little difference in transcript levels in the cells treated for 90 mins with ActD (Table 6—1) were designated NOTGUTs, and used as controls for subsequent experiments. Cycloheximide Effect It has been suggested that treatment with the protein synthesis inhibitor cycloheximide (CHX) tends to "stabilize" unstable transcripts in yeast [20]. These data have been interpreted as either implying a role for active translation in the degradation of unstable mRNAs, or the participation of an labile factor in the degradation of these transcripts [20; reviewed in 2]. We were therefore interested in examining the effect of CHX on the accumulation of GUT and NOTGUTtranscripts in NT-1 cells. The blots used for the initial characterization of GUT transcript levels in control and 90 min ActD-treated cells also included total RNA isolated from NT-1 cells that had been treated for 90 mins with both ActD and the protein synthesis inhibitor cycloheximide (CHX; Figure 6-1). In contrast to the data from yeast [20], our data suggest that CHX treatment does not affect all unstable transcripts similarly in NT-1 cells. Although most of the GUT and NOTGUT mRNAs showed a modest increase in mRNA levels in the presence of ActD + CHX, for the larger of the two transcripts identified by the GUT 15 cDNA probe, and NOTGUT10-3 there was a clear decrease in mRNA 157 levels in cells treated with ActD + CHX, relative to the control cells (Figure 6- 1; quantitation in Table 6-1). Conversely, the accumulation of the smaller GUT 15 transcript in the ActD + CHX treated cells increased markedly, more than that of any other GUT (Figure 6-1). It is interesting to speculate on the causes of the different CHX effects on the two GUT 15 transcripts. We have preliminary evidence that there are two different classes of GUT1 5 cDNA (CBT and Green, P.J., unpublished) and it is possible that the CHX effect is specific for one of these classes. However, there is some evidence that CHX can stimulate the transcription of some genes [21], and such an effect cannot be ruled out for GUT 15 (despite the presence of ActD concentrations known to halt transcription). Nonetheless, whether the CHX effect is mediated post- transcriptionally, or transcriptionally, it will be most interesting to identify any sequence differences between the two classes of GUT 15 that may cause this effect. GUT mRNA Half-Lives To confirm that the GUTs did indeed encode unstable transcripts, GUT mRNA decay rates were measured by Northern blotting of RNAs isolated from NT-1 cells at 30 minute intervals after the addition of ActD to the cultures (see Materials and Methods). Representative Northern blots of GUT and NOTGUT mRNA decay are shown in Figure 6-2. To calculate the half-lives of their transcripts, the signals of each GUT and NOTGUT probe were quantitated as 158 Legend to Figure 6-2: Time-Course Experiments to Determine GUT transcript Half-Lives Northern blots of 10pg total RNA, isolated from NT-1 cells at the indicated times (in minutes) after the addition of 100pg/mL ActD to the cultures. Blots were hybridized to the indicated GUT and NOTGUT probes. Half-life measurements from these and additional experiments were determined following quantitation of the signals using a phosphorimager, and are summarized in Table 6-1. 159 30 60 90 120150 GUT3-2 s m»..... 30 60 90 120150 30 60 90 120150 GUT7£a a, GUTss 1.3.. ~ ' 07+ DO... "3 30 60 90 120150 30 60 90 120150 GUT 7-20 30 60 90 120150 NOT GUT 7-2b 1.0+ ..... Figure 6-2. Time-Course Experiments to Determine GUT transcript Half-Lives 160 described in Materials and Methods. These experiments show that the GUT mRNAs all fall into the category of unstable transcripts [3, 4] as their absolute half-lives are on the order of an hour or less (Table 6-1). We have found previously that comparison of the half-lives of a number of different transcripts is most reliably achieved by expressing them relative to half-lives of a single reference transcript, such as GUS [9]. This procedure is also an effective way of normalizing for cell line - to - cell line variations in the absolute half-lives of each transcript [9, 11]. To normalize GUT half-lives to those of GUS, and thus optimize comparisons between the half-lives of the different GUT mRNAs, each time-course Northern blot was stripped of the GUT probe and hybridized to a GUS probe. GUS mRNA half-lives were quantitated and ratios of GUT/GUS half-lives (relative half-lives) were calculated as described [9]. Table 6-1 shows that, as expected, the rank order of GUT mRNA stabilities is similar, whether they are expressed as absolute or relative half-lives. The two NOTGUT transcripts did not degrade appreciably over the 150 min timecourse (Figure 6-2 and Table 6-1). rcrlhrriifh TnNTT To investigate the number of genes that hybridize to the individual GUT probes, Southern analyses were performed as described in Materials and Methods. As summarized in Table 6-2, in most cases the GUT and NOTGUT cDNA probes hybridized to between 2 and 5 genomic DNA fragments in each restriction 161 digest (data not shown), demonstrating that they are encoded by single or low copy genes. The most complex pattern was obtained with the GUT 8-2B probe, which encodes a tobacco histone H3 cDNA (see below). This probe hybridized to >20 fragments in each digest. The copy numbers for GUTs 7- 2A, B and C cannot be reliably determined at present from Southern blots, as these Southerns were performed with a probe from the original GUT 7-2 A clone, which includes EcoRl fragments corresponding to each of the three cDNAs (CBT and Green, P.J.), unpublished observations). In general, the sizes of the GUT and NOTGUT cDNA clones correlate with the sizes of the corresponding transcripts. With the exception of GUT 3-2, all of the GUT and NOTGUT cDNA probes are smaller than, or the same size as, the major transcripts they detect on Northern blots (see Table 6-2). Nonetheless, in some cases (eg GUTs 3-2, 7-2c and 15), it is clear that an individual GUT cDNA probe hybridizes to more than one transcript in NT-1 cells (see Figures 6-1 and 6-2). The origins of these smaller transcripts are unknown at present. One intriguing possibility is that some may represent GUT mRNA degradation products. However, it is also possible that they are full-length mRNAs derived from genes closely related to the GUTs. The presence of two different cDNAs in a single GUT probe cannot be ruled out, especially for GUT 3-2, where the cloned insert is about three times larger than the major transcript it detects on Northern blots (Table 6-2). The Northern blot experiments in Figures 6-1 and 6-2 also serve to demonstrate that the basal 162 Table 6-2: Structural Characteristics of GUTs and NOTGUTs "l Transcript cDNA Size Estimated Sequence GUT Size (Kb) (Kb) Gene Number‘ Similarityz 3-2 1 .1 2.8 2-5 none 7-2a 1.8 0.75 <10 DNA J 7-2c 0.7 0.35 <10 W.l.P 8-1 1.0 1.0 2-5 none 8-2a 1.4 1.4 1-2 E2 8-2b 1.0 0.6 5-10 Histone H3 II II 8-3 0.7 0.6 2-5 none 15 i) 1.9 ii) 1.7 1.7 1-2 none I NOTGUT I 7-2b 1.0 0.4 <10 85 protein 10-3 1.4 1.1 1-2 Histone H1 1Gene numbers were estimated from Southern blots of tobacco genomic DNA. 2Summarized from Figure 6-3. 163 expression levels of the GUTs differ markedly from one and other, despite the instability of all their transcripts. For example, GUTs 3-2 and 8-3 are expressed at very high levels in control NT-1 cells, but GUTs 7-2a and 15 are expressed at levels that are barely detectable on Northern blots (Figures 6-1 and 6-2). Moreover, the sizes of the GUT transcripts range from 1.9 kb for the larger GUT 1 5 mRNA, to 0.7 kb for GUTs 7-2c and 8-3, indicating that transcript size per se is unlikely to influence transcript half-life. T u n imil ri ie Information about the possible functions of protein products of the GUTs can be obtained by identifying similarities between the sequences of the GUTs and known protein sequences. This information may suggest reasons why the corresponding transcript is unstable. A similarity search of deduced amino acid sequences from the ends of each GUT and NOTGUT cDNA clone against protein sequences in the databases using the BLAST algorithm [19] reveals that four of the GUTs (7-2A, 7-2C, 8-2A, and 8-23) and both of the NOTGUTs (7- 28 and 10-3) are similar to previously characterized proteins. These similarities are discussed further below. For the remaining GUTs (3-2, 8-1, 8-3, and 15), there were no database matches at the amino acid level, suggesting that these GUTs represent novel cDNAs. A subsequent similarity search of the nucleotide sequences of these GUT cDNAs against the nucleotide databases confirmed this. 164 As shown in Figure 6-3A, sequence from one end of the GUT 7-2A cDNA suggests that this GUT is a novel plant member of the eukaryotic family of E. coli DNAJ homologs. These proteins, which have been identified in yeast, Drosophila and human cells, are thought to interact with HSP70 proteins and play roles in protein sorting [22]. GUT 7-2A possesses each of the residues that define the N-terminal "DNAJ domain" of this family of proteins [22]. The C-terminal domains of the DNAJ homologs are generally less conserved, and are thought to modulate the specificity of their interactions with HSP70 proteins. It is not surprising, then, that sequence from the other end of the GUT 7-2A clone appears to have no similarity with previously characterized DNAJ proteins. Deduced amino acid sequence of GUT 7-2C (sequence from each end of the 350 bp EcoRI fragment overlaps in the middle) is similar to a fragment of a tomato wound-inducible protein (see Figure 6-3B). Although the kinetics of induction of the homologous tomato transcript following wounding have not been described, in general wound-inducible proteins are rapidly induced at the mRNA level in response to wounding in plants [e.g. 23], and therefore conform to our predictions as to the type of genes that are likely to have unstable transcripts. GUT 8-2A has restricted similarity to the extreme N- terminus of the ubiquitin conjugating enzyme E2 of Drosophila melanogaster (63% similarity over 44 amino acids - Figure 6-3C), but it’s similarity to plant E23 is weaker. Whether or not the E2 enzymes have unstable transcripts has not been evaluated. 165 Legend to Figure 6-3 Nucleotide sequence data were obtained from both ends of each GUT and NOTGUT cDNA clone and compared to sequences on databases using the default parameters of the BLASTX algorithm as described in Materials and Methods. Shown here are the alignments produced from the BLAST similarity searches. (A) GUT 7-2A; (B) GUT 7-2C; (C) GUT 8-2A; (D) GUT 8-2B; (E) NOTGUT 7-2B; and (E) NOTGUT 10-3. 166 >SP:DNAJ_ECOLI DNAJ PROTEIN. Length 8 376 Score I 248 (122.7 bits). Identities I 44/88 (50‘). Positives I 69/88 (78!) GUT 7-2a: 19 VPKGASDEQIKRAYRKLALKYHPDKNPGNEEANTKFAEINNAYEVLSDSEKKNIYDRYGEEGLKQHBASGGGRGAGHNIQDIFSQFFG 282 V+K A + +I++AY++LA+KYHPD+N 6+ EA++KF EI++AYEVL+DS+K+ YD+Y6 + Q + +666 6+6 + DIP++ P6 DNA J: 12 VSKTAEERBIRKAYKRLAMKYHPDRNQGDKEAEAKEKBIKEAYEVLTDSQKRAAYDQYGHAAFEQGGNGGGGPGGGADPSDIEGDVFG 99 >PIR:519773 Hound-induced protein - Tomato (fragment). Length I 76 Score I 77 (39.5 bits) Score I 62 (31.8 bits) Score I 53 (27.2 bits) Identities I 15/16 (93‘) Identities I 11/20 (551) Identities I 13/25 (52\) Positives I 16/16 (100\) Positives I 15/20 (75‘) Positives I 17/25 (68‘) GUT 7-2c: 72 WIVAASIGAVEALKDQ 119 120 GPARNNYALRSLHHYAKTNL 179 225 ASAADISGlxlnxTBESLNKVHGLS 299 HIVBAS+6AVBALKDQ 6 RNNY+LRSL + +K N+ +8+ +2! K+BBSL KVH LS W.I.P.: 1 WIVAASVGAVEALKDQ 16 18 GLCRWNYPLRSLAQHTKNNV 37 47 SSSITTKSBXNBXSEBSLRKVNYLS 71 >PIR:519157 *Ubiquitin-conjuqstinq enzyme - Fruit fly (Drosophila melanogaster). Length I 147 Score I 80 (39.7 bits). Identities I 15/44 (34\). Positives I 28/44 (63\). our s-2a: 298 AVKRILQEVKEHQSNPSDDPHSLPLEENIFEWQFGIRGPRDSBF 429 A+KRI 3+ ++ +1» + + 9+ +++r we I op 05 + Drama :2: 2 m1NKILQDLGRDPPAQCSAGPVGDDLFHWQATIMGPPDSP! 45 >up:905203:ss_nxzs HISTONB 33. Length - 136 Score I 668 (315.4 bits). Identities I 135/136 (99t). Positives I 135/136 (99‘) GUT 8-2b: 92 HARTKQTARKSTGGRAPRKQLATKAARKSAPATGGVKKPHRFRPGTVALREIRKXQXSTBLLIRNLPP 161 HARTKQTARKSTGGKAPRKQLATKAARKSAPATGGVKKPHRFRPGTVALREIRXXQKSTELLIRKLP? Hinton. H3: 1 HARTKQTARKST66RAPRKQLATKAARNSAPATGGVRKPHRFRPGTVALREIRXXQKSTELLIRKLPP 69 GUT 8’2b: 162 QRLVRBIAQDFXTDLRIQSSAVAALQEAAEAYLVGLPEDTNLCAIHAKRVTIHPKDIQLPRRIRGERA 499 QRLVREIAQDFKTDLRFQSSAVAALQEAAEA!LVGLFBDTNLCAIHAKRVTIHPKDIQL RRIRGERA Histone 83: 70 QRLVRBIAQDFXTDLRFQSSAVAALQEAAEAYLVGLPEDTNLCAIHAKRVTIHPKDIQLARRIRGERA 136 >SP:RSS_RAT 40$ RIBOSOHAL PROTEIN 55. Length I 204 Score I 217 (111.0 bits). Identities I 45/49 (91‘). Positives I 48/49 (97‘). NOT GUT 7-2b: 19 TGARESAFRNIKTIAECLADELINAAKGSSNSYAIKKKDEIERVRKANR 165 TGARB+AFRNIKTIAECLADELINA KGSSNSYBIKKKDE+BRVAK+NR Rat SS: 156 TGAREAAFRNIKTIAECLADELINARXGSSNSYAIKKKDELERVAKSNR 204 >SP:H11_ARATH HISTONE H1.1. Length I 274 Score I 198 (98.6 bits). Identities I 39/52 (75‘). Positives = 45/52 (86!). NOT GUT 10-3: 234 PAAPRKKTASSHPPCFEHISDAIVTLKERTGSSQYAITKFIEDKQKNLPPNF 389 +AAP+K+T SSH? EMI DAIVTLKERTGSSQYAI KFIE+K+K+LPP F Histone Hl-l: 51 AAAPKKRTVSSHPTYEEHIKDAIVTLKERTGSSQYAIQKFIEEKRKELPPTF 102 >SP:H1_PEA HISTONE al. >PIR:500033 Histone 81b - Garden pea. Length I 265 Score I 96 (49.4 bits). Identities = 20/38 (529). Score I 62 (31.9 bits). Identities I 12/20 (60%). Positives I 27/38 (71‘). Positives I 18/20 (90‘) NOT GUT 10-3: 364 RTTPSRKAAPXAAPAKKTPARSVKSPVKKSSTRRGGRK 251 394 KPAKVARTATRTTPSRKAAP 335 + ++ +K A+K P K A+SVKSPVKK S +RGGRK RPAKVA+T+ +TTP++K 3+ Histone 81-1: 228 KVAAVKKVAAKKVPVKSVKAKSVKSPVKKVSVKRGGRK 265 212 KPAKVARTSVKTTPGXKVAA 231 Figure 6-3. Sequence similarities of GUT and NOTGUT cDNAs 167 The strongest similarity between a GUT sequence and one on the databases is that of GUT 8-2B which, over the 600bp cloned fragment, is 99% identical to maize histone H3 (Figure 6-3D). Rapid changes in the levels of cell- cycle regulated histone genes is thought to be facilitated by the instability of their transcripts [24]. Indeed, the histone H3 transcript has previously been characterized as unstable in mammalian cells [25], so it is not surprising the corresponding tobacco gene was identified in our screen for GUTs. The sequences mediating the rapid degradation of animal cell-cycle regulated histone transcripts are thought to reside in the extreme 3’ stem-loop structure [26]. It is not clear whether this is also the case for plant histone mRNAs, which lack strong sequence conservation in their 3’UTRs and, unlike animal histone transcripts, are polyadenylated [27]. Both NOTGUTs are also similar to proteins in the databases. NOTGUT 7- 23 is 97% similar to rat 40$ ribosomal protein S5 over a stretch of 49 amino acids, and NOTGUT 10-3 is similar, but not identical, to Arabidopsis histone H1 (Figure 6-3 E and F, respectively). The stability of the NOTGUT10-3 (histone H1) transcript is in contrast to the instability of the GUT 8-2B (histone H3) transcript in the same cells. One possibility that might explain this observation is that NOTGUT10-3 encodes a "variant" or "replacement" form of histone H1 . Unlike the cell-cycle regulated forms, the variant histones are expressed constituitively [28], and therefore mRNA instability would provide no particular advantage. However, sequence conservation among H1 histones is generally 168 less pronounced than is the case with other histones [27], making it difficult to differentiate between cell-cycle regulated and variant forms based on sequence information alone. Nonetheless the stability of the NOTGUT 10-3 transcript is consistent with a "housekeeping" role for the corresponding protein. A similar conclusion can be drawn by correlating the stability of the NOTGUT 7-28 transcript, with ribosomal protein 85, which it encodes. Beyond their similarities to sequences in the databases, there are a number of other interesting features of the GUT cDNA sequences. First, a comparison of identifiable GUT 3’UTRs shows that there are no obvious conserved motifs. There are occasional AUUUA motifs, but these do not occur in highly AU-rich regions resembling the AREs, nor are they restricted to GUT 3’UTRs, occurring also in the NOTGUT sequences. While a contribution of these AUUUA elements in the instability of the GUT transcripts cannot be ruled out, previous studies have shown that the mere presence of an AUUUA element is insufficient to cause mRNA instability (e.g. see [11]). Secondly, GUT 8-1, which gives a diffuse signal on Northern blots, with consistently high background signals, has two curious repeats which are quite close together. An (AC)8 motif (which, if translated, would correspond to a Threonine-Histidine repeat) is found some 50 bases upstream of a stretch of 13 C residues (potentially encoding four Prolines). The significance of these sequences is not clear, but interestingly, a search of the nucleotide databases with the GUT 8-1 sequence showed that the former motif is related to a human genomic 169 minisatellite sequence. It is not surprising that we found no sequence conservation within the GUT 3' UTRs. With the possible exception of the AREs in mammals, little such conservation has been found between different unstable transcripts in animals. Recently, an extensive search of the nucleotide databases for broadly conserved elements in the non-coding regions of a large number of eukaryotic genes was performed [29]. Interestingly, the only similarities that were identified in this search were between pairs of homologous genes from different species, not between groups of otherwise unrelated genes [29]. An alternative reason for the apparent lack of sequence conservation among the GUTs is that the relatively small number of different GUT clones that were isolated in this initial screen may preclude an exhaustive analysis. The fact that only a single cDNA for each different GUTwas identified suggests that the current collection of GUTs is not complete, and that many additional GUTs are likely to exist in tobacco cells. CONCLUSIONS AND FUTURE PROSPECTS The data presented in this paper demonstrate that the differential screening strategy is a promising approach for the isolation of GUTs in tobacco, which may be broadly applicable to the identification of genes with unstable 170 transcripts in other species. Using this approach we have identified ten previously uncharacterized tobacco cDNAs. These include eight GUTs, with mRNA half-lives in the range of an hour or less, on the same order as those of transcripts characterized as unstable in mammalian cells [3, 4]. Four of the GUTs represent completely novel cDNAs, whereas the other four GUTs, and both of the NOTGUTs, encode proteins with similarity to proteins in the databases. These similarities conform to our predictions concerning the kinds of proteins likely to be encoded by unstable or stable transcripts. The GUTs should constitute excellent tools for evaluating the mechanisms by which unstable transcripts are recognized and targeted for rapid degradation in plants. In particular, it will be most informative to delineate the sequences within the GUT transcripts that mediate their instability. Moreover, it should also be possible to investigate the roles of processes such as translation, deadenylation and protein binding in the mRNA decay pathways of a number of individual unstable transcripts. These efforts will substantially increase our understanding of the range of different mechanisms by which unstable transcripts are targeted for rapid degradation in plants. ACKNOWLEDGEMENTS We would like to thank Dr. Gynheung An for the NT-1 cell cDNA library, Steven Reiser, and Pat Tranel for isolating the GUT 15 clone, and helping with its sequencing, respectively, Dr. Michael Sullivan for maintaining the wild-type and transformed NT-1 cell lines, and André Dandridge and Linda Danhof for expert technical assistance. This work was supported by grants from the USDA (901 167) and the DOE (FG02-90ER- 20021) to P.J.G. 10. 11. 12. 13. 14. 15. 16. 1 71 LITERATURE CITED Green, P.J. (1993) Plant Physiol., 102, 1065-1070. Sachs, AB. (1993) Cell, 74, 421-427. Peltz, S.W., Brewer, G., Bernstein, P., Hart, P.A. and Ross, J. (1991) Crit. Rev. Euc. Gene Expr., 1, 99-126. Brock, ML. and Shapiro, D.J. (1983) Cell, 34, 207-214. Zhang, S., Sheng, J., Liu, Y. and Mehdy, MC. (1993) Plant Cell, 5, in press. Seeley, K.A., Byrne, D.H., and Colbert, J.T. (1992) Plant Cell, 4, 29-38. McClure, BA. and Guilfoyle, T. J. (1989) Science, 243, 91-93. Gee, M.A., Hagen, G. and Guilfoyle, T.J. (1991) Plant Cell, 3, 419-430. Newman, T.C., Ohme-Takagi, M., Taylor, CB. and Green, P.J. (1993) Plant Cell, 5, 701-714. McClure, B.A., Hagen, G., Brown, C.S., Gee, M.A. and Guilfoyle, T.J. (1989) Plant Cell, 1, 229-239. Ohme-Takagi, M., Taylor, C.B., Newman, T.C. and Green, P.J. (1993) Proc. Natl. Acad. Sci, USA, in press. An, G. (1985) Plant Physiol., 79, 568-570. Nagy, F., Kay, S.A. and Chua, N.-H. (1990) in Gelvin, S.B., Schilperoort, R.A. and Verma, D.P.S. (eds), Plant Molecular Biology Manual. Kluwer, Dordrecht, pp B4/13 - B4/14. Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, Plainview. Taylor, OB. and Green. P.J. (1991) Plant Physiol., 96, 980-984. Delauny, A.J. and Verma, D.P.S. (1990) in Gelvin, S.B., Schilperoort, R.A. and Verma, D.P.S. (ed.s), Plant Molecular Biology Manual. Kluwer, Dordrecht, pp A14/1 - A14/23. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 172 Feinberg, AP. and Vogelstein, B. (1983) Annal. Biochem., 132, 6-13. Dellaporta, S.L., Wood, J. and Hicks, J.B. (1983) Plant Mol. Biol. Rep., 1, 19-21. Altshul, S.F., Gish, W., Miller, W., Meyers, E.W. and Lipman, D.J. (1990) J. Mol. Biol., 215, 403-410. Herrick, D., Parker, R. and Jacobson, A. (1990) Mol. Cell. Biol., 10, 2269-2284 Ballas, N., Wong, L.-M. and Theologis, A. (1993) J. Mol. Biol., in press. Silver, P.A. and Way, J.C. (1993) Cell, 74, 5-6. Rohrmeir, T. and Lehle, L. (1993) Plant Mol. Biol., 22, 783-792. Pandey, N.B. and Marzluff, W.F. (1987) Mol. Cell. Biol., 7, 4557-4564. Schuler, G.D. and Cole, MD. (1988) Cell, 55, 1115-1127. Harris, M.E., Bfihni, R., Schneiderman, M.H., Ramamurthy, L., Schiimperli, D. and Marzloff, W.F. (1991) Mol. Cell. Biol., 11, 2416- 2424. Nakayama, T. and Iwabuchi, M. (1993) Crit. Rev. Plant Sci, 12, 97- 110. Wu, R.S. and Bonner, WM. (1981) Cell, 27, 321-330. Duret, L., Dorkeld, F. and Gautier, C. (1993) Nucl. Acids Res., 21, 2315. CHAPTER 7 CONCLUSION 173 174 At its inception, one of the initial goals of the research project described in this Dissertation was to identify genes encoding RNases in Arabidopsis. This has been achieved with the identification of PCR products corresponding to three different Arabidopsis RNase genes, RNSI, RN82, and RNS3, by virtue of their homology to fungal and S-gene RNases (Chapter 2). Each of the RNS genes is expressed in Arabidopsis, albeit to different levels, and each is present in one or two copies in the Arabidopsis genome (Chapter 2). With a view towards identifying functions for the RNS RNases, the expression properties of RN82, the most highly expressed of the RNS genes, were explored (Chapter 3). These experiments showed that, like the S-RNase genes to which it is related, RN82 is most abundantly expressed in flowers of Arabidopsis. However, unlike the S-RNase genes, RN82 is also expressed at some level in all other organs that were tested (Chapter 3). The most informative data pertaining to the identification of potential functions for the RN82 enzyme, came from experiments demonstrating that RNSZ expression increased markedly during senescence, in both petals and leaves of Arabidopsis (Chapter 3). This suggests that RN82 may play a role in the remobilization of nutrients, such as phosphate, from RNA. The increased expression of RN82 in Arabidopsis seedlings starved for phosphate supports this contention (Chapter 3). The RN82 PCR product was also used as a probe to identify full length RN82 cDNAs. These clones were used in a heterologous expression system to demonstrate that RN82 encodes an active RNase (Chapter 3). Furthermore, 175 comparison of the deduced amino acid sequence of RNSZ to those of related fungal and plant RNases suggested that there are two distinct subclasses of RNases within the plant lineage of the T2/S RNase superfamily. These are the S-RNases, which are involved in self-incompatibility in the Solanaceae, and the S-like RNases (including RN32), which are expressed in self-compatible species. The placing of RNase LE (from tomato) within the S-like RNase lineage suggests that the S-like and S-RNases diverged before speciation (Chapter 3). As sequences of more S-like RNases have become available (eg RNS1, RN83, and a second tomato enzyme, RNase LX), this conclusion has been extended, and it now seems likely that divergence within the S-like lineage also preceded speciation. This is because RNS1 appears to be more similar to LE, and RNS3 to LX, than either is to the other, or to RNSZ (Bariola, P.A. and Green, P.J., unpublished). RN82 was among the first senescence-ind ucible and phosphate starvation- inducible genes identified for which a specific function (phosphate remobilization) could be proposed. The RN82 cDNA therefore constitutes a novel molecular probe for the study of senescence in plants, and for the characterization of plant responses to nutrient limitation. For example, it should be possible to identify regions of the RN82 promoter that mediate the induction of this gene under phosphate limiting conditions, or during senescence. Understanding plant responses to nutrient limitation is of fundamental importance to agriculture. Ultimately, the study of the roles of RNS2, and of 176 RNSI, which is even more strongly induced by phosphate starvation (Bariola, P.A. and Green, P.J., unpublished) may lead to increased understanding of the roles of RNases in these responses. In the long term, this work is directed towards the identification of those RNases with a specific role in mRNA catabolism. Based on their sequences, RNS1 and RNS3 are thought to be extracellular (Bariola, P.A. and Green, P.J., unpublished), and RN82 may be targeted to the vacuole in Arabidopsis (Chapter 3). These presumed localizations of the RNS enzymes are difficult to reconcile with a role in the critical initial steps in mRNA degradation, which are presumably cytoplasmic. However, small fragments of RNA have been found in plant vacuoles [Abel, S., Blume, B. and Glund, K. (1990), Plant Physiol. 94, 1163-1171], so it is possible that later steps in mRNA catabolism may occur in that organelle. In the future, it should be possible to adapt the PCR approach initially used to identify RNSI, RN82, and RNS3, to explore the possibility that cytoplasmic S-like RNases exist in Arabidopsis. Such enzymes would be of great interest, as very few intracellular RNases have been identified in eukaryotes. Cytoplasmic RNases are likely to be pivotal components in the decay pathways of highly unstable transcripts in plants, either directly involved in the recognition of such transcripts, or as the means for their rapid degradation. In the later case, RNases may be recruited to unstable transcripts via their interactions with factors bound to specific sequence elements within those transcripts. 177 The search for, and identification of, sequence determinants that confer instability on transcripts in tobacco was described in Chapters 4, 5 and 6 of this Dissertation. An essential initial step was the development of a system for the accurate measurement of reporter transcript half-lives in stably transformed tobacco (NT-1) cells (Chapter 4). That mRNA half-lives measured in the NT-1 cells were likely to reflect those in green plants was suggested by experiments demonstrating that control transcripts disappeared with similar kinetics when expressed in transgenic plants under the transcriptional control of a regulated promoter (Cab-1), as they did in NT-1 cells (Chapter 4). Using the NT-1 cell system, two very different sequences that confer instability on reporter transcripts were identified. Repeats of the DST element (Chapter 4), and the AUUUA element (Chapter 5), were each found to markedly destabilize mRNAs in the NT-1 cells. Moreover, each of these sequences caused a pronounced decrease in the accumulation of the reporter transcript in the leaves of transgenic tobacco, suggesting that both also function as instability determinants in plants, as they do in NT-1 cells. (Chapters 4 and 5). Having identified the DST and AUUUA sequences as potent determinants of mRNA instability, it is of interest to begin to address the mechanisms by which they function. An important first step is to identify and characterize components of the RNA degradation machinery that specifically recognize these elements. One approach to this end is to search for factors that bind to the AUUUA or DST elements in vitro. This strategy has proven particularly 178 effective in the identification of a number of mammalian ARE-binding proteins. However, it is also necessary to identify the specific sequences within each instability determinant that are necessary for them to confer instability on transcripts. Once this is achieved, it should then be possible to correlate the binding of proteins to active or inactive elements with the impact of those same elements on the half-lives of reporter transcripts. Genetic approaches toward the identification of factors involved in the recognition of unstable transcripts are also available in plants. For example, it may be possible to identify mutant Arabidopsis plants that are unable to recognize a particular instability determinant. One way this could be achieved is to place an instability determinant within the transcript of a selectable marker gene. Arabidopsis plants transformed with this construct should be sensitive to the selectable marker, owing to the instability of the corresponding transcript. In contrast, mutant progeny of these plants in which the instability determinant is no longer recognized as such, should be able to survive selective conditions. The major advantage of this strategy is that it may lead to the identification of factors that are not directly bound to the unstable transcripts, but still participate in their recognition and/ or degradation. Moreover, genetic approaches are not yet feasible in mammalian cells, and their application in plants may therefore be of particular relevance to the characterization of factors that interact with the AUUUA repeat. This element is similar to the AREs that destabilize transcripts such as c-myc and c-fos in mammalian cells. In the long 179 run, it will be interesting to address the specificity of factors identified using these approaches by examining their interactions with other instability determinants. These experiments will help to determine the extent to which mechanisms targeting a variety of unstable transcripts for rapid degradation in plants are shared or unique. Future identification of endogenous instability determinants within the GUT transcripts (Chapter 6) will facilitate these analyses. Concluding Remarks Using the tools that have been described above, it may well be found that many features of mRNA degradation pathways are shared between plants and animals. Exploration of these potential similarities can only benefit from the application of genetic approaches toward identifying components of mRNA degradation pathways in Arabidopsis. Nonetheless, given their more fundamental requirement for rapidly regulated gene expression, it is also likely that plants will have developed unique and fascinating mechanisms to facilitate the regulation of gene expression at the level of mRNA stability. The work described in this Dissertation provides the foundation for future efforts aimed at elucidating mRNA decay pathways in plants. Ultimately, the analysis of sequence determinants of stability will converge with efforts to 180 characterize plant RNases, as those RNases with a specific role in the degradation of mRNAs are identified. This may be achieved via the isolation of cytoplasmic S-RNase homologs in Arabidopsis, or through the continuing characterization of cellular factors that bind to specific sequence elements. Taken together, these approaches should provide fundamental information about the mechanisms that target transcripts for rapid degradation in plants.