\ ‘NHlWIllllWllUI‘iHlWNl‘WlWNHwUHI —-\ 4:. N 161 THS r “DDADV 1 h Micmga . State University J This is to certify that the dissertation entitled CONSERVATION OF THE LOW TEMPERATURE TRANSCRIPTOMES AND CBF REGULONS BETWEEN SOLANUM AND ARABIDOPSIS presented by Marcela Alejandra Carvallo-Pinto has been accepted towards fulfillment of the requirements for the PhD degree in BIOCHEMISTRY AND MOLECULAR BIOLOGY Kiwi Mi Major Professcf’s Signature I 2 » IS" 20 02 Date MSU is an Afiinnative Action/Equal Opportunity Employer PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 5/08 K;IProj/Acc&Pres/ClRC/DateDue.indd CONSERVATION OF THE LOW TEMPERATURE TRANSCRIPTOMES AND CBF REGULONS BETWEEN SOLANUM SPECIES AND ARABIDOPSIS By Marcela Alejandra Carvallo-Pinto A DISSERTATION Submitted to Michigan State University In partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Biochemistry and Molecular Biology 2009 ABSTRACT CONSERVATION OF THE LOW TEMPERATURE TRANSCRIPTOMES AND CBF REGULONS IN SOLANUM SPECIES AND ARABIDOPSIS By Marcela A. Carvallo-Pinto Plants from tropical regions have no freezing tolerance whereas plants from temperate regions can survive freezing after a period of cold acclimation (exposure to low nonfreezing temperature). In Arabidopsis the AP2 transcriptional activators CBF 1, CBF2 and CBF3 have an important role in cold acclimation. They are quickly induced in response to low temperature followed by expression of the CBF regulon, which results in an increase in freezing tolerance. Little is known about the conservation of low temperature transcriptomes and CBF regulons in different plant species. Solanum tuberosum (common potato) (St) and its wild close relative S. commersonii (So) are two closely related species with different levels in freezing tolerance, therefore they constitute an excellent model to study conservation of the cold transcriptomes and CBF regulons. The work in this dissertation focused on the identification and comparison of the low temperature transcriptomes of So and St, and also their CBF regulons. Using the St 10K cDNA array. the cold- and CBF-transcriptomes of these species were analyzed, and by identification of putative orthologous groups between St and Arabidopsis, the transcriptomes of So and St were compared to that of Arabidopsis. With the criteria used (2FC, p<0.05) there was more than 50% overlap between cold transcriptomes of the two Solanum species, suggesting that there are species specific cold regulated genes. However, no obvious differences could be identified between Sc and St cold- transcriptomes that explain their differences in freezing tolerance. Only around 10% of the cold regulated genes in Solanum species, that have Arabidopsis orthologs, were also identified as cold regulated in Arabidopsis. This indicates significant differences between the two Solanum species and Arabidopsis cold transcriptomes. The Sc and St CBF regulons were identified, as genes that are ' differentially expressed by cold treatment and by CBF overexpressionfracture. About 48% of the Sc CBF regulon is also part of the St CBF regulon, suggesting that the genes that are members of the CBF regulon in each of these two Solanum species have evolved different cis-acting DNA regulatory elements. When compared to Arabidopsis, only 14% of the Sc and St CBF regulons identified in this study are also part of the Arabidopsis CBF regulon, indicating that there are important differences between these CBF regulons. The identification of low temperature transcriptomes of the two Solanum species provides a start point to the study of these two closely related species with different levels in freezing tolerance. Future analysis of the sequenced potato genome will provide the bases for novel strategies to expand our knowledge of these plants freezing stress mechanism. PREFACE In chapter 2, the S. tuberosum and S. commersonii plant growth and treatments were conducted by Jeff Skinner and Zoran Jeknic from Tony Chen’s laboratory at Oregon State University. 358::AtCBF3 S. tuberosum and S. commersonii transgenic plants were obtained from Maria Teresa Pino at Tony Chen’s laboratory, Oregon State University. Expression profiles generated by microarray hybridizations and real time PCR, as well as data analysis was conducted by the author of this thesis. The groups of putative orthologs between Arabidopsis and S. tuberosum were generated by Cheng Zou from Shinhan Shiu’s laboratory at Michigan State University. The list of cold regulated genes in Arabidopsis was generated by Colleen Doherty. Cold data analysis and comparison across species was performed by the author of this thesis. In chapter 3, the S. tuberosum and S. commersonii plant growth and treatments were conducted by Zoran Jeknic from Tony Chen laboratory at Oregon State University. Real time PCR and data analysis was conducted by the author of this thesis. iv TABLE OF CONTENTS LIST OF TABLES ................................................................................. vi LIST OF FIGURES ................................................................................ vii KEY TO ABREVIATIONS ........................................................................ ix CHAPTER 1 Literature Review ................................................................ 1 Freezing Damage and Cold Acclimation ........................................... 2 Cold responsive CBF pathway in Arabidopsis thaliana ........................ 4 CBF regulation in response to low temperature .................................. 6 CBF regulation in response to other environmental cues ..................... 8 CBF independent pathways in cold acclimation ................................. 11 Cold responsive CBF pathway in other plant species ......................... 12 CHAPTER 2 Transcriptome profiles of Solarium species with different levels of freezing tolerance ............................................................................... 17 Introduction ............................................................................... 17 Results ..................................................................................... 22 Discussion ................................................................................. 58 Materials and Methods ................................................................ 67 CHAPTER 3 Regulation of conserved transcription factors by the circadian clock, cycloheximide treatment and mechanical agitation ..................................... 74 Introduction ................................................................................ 74 Results ...................................................................................... 78 Discussion ................................................................................ 94 Materials and Methods ................................................................ 98 APPENDIX ....................................................................................... 100 LITERATURE CITED ......................................................................... 104 LIST OF TABLES Table 2.1 Primers used on real time PCR .................................................. 71 vi LIST OF FIGURES “Images in this thesis/dissertation are presented in color” Figure 2.1 Hierarchical clustering and expression profiles of Sc and St EST clones at 2, 24 and 168h of cold treatment (2°C) ................................ 24 Figure 2.2 Comparison of cold-induced ESTs (2FC, p<0.05) in S. commersonii (Sc) and S. tuberosum (St) ........................................................... 28 Figure 2.3 Comparison of cold-repressed ESTs (2FC, p<0.05) in S. commersonii (Sc) and S. tuberosum (St) ........................................ 30 Figure 2.4 Putative orthologous groups (pOGs) between Arabidopsis and S. tuberosum ................................................................................ 33 Figure 2.5 Comparison of cold transcriptomes of Arabidopsis (At), 8. commersonii (Sc) and S. tuberosum (St) ......................................... 37 Figure 2.6 Transcript accumulation of CAF1a-like and CDF3-like in S. commersonii (Sc) and S. tuberosum (St) .......................................... 40 Figure 2.7 Transcript accumulation in response to low temperature of Myb73- like, ZAT10-Iike, CZF 1 -like, ZAT12-Iike, RAV1 -Iike and CBF1 Solanum genes ....................................................................................... 45 Figure 2.8 AtCBF3 transgene accumulation in St and So 358::AlCBF3 transgenic lines ......................................................................... 49 Figure 2.9 Hierarchical clustering and expression profiles of So and St EST clones at 2, 24 and 168h of cold treatment (2°C), in combination with 35S::AtCBF3 Sc and St lines ........................................................ 50 Figure 2.10 S. commersonii and S. tuberosum CBF regulons ..................... 53 vii Figure 2.11 Comparison of CBF regulons among 8. commersonii (Sc), 8. tuberosum (St), and Arabidopsis (At) based on putative orthologous groups (pOGs) ........................................................................... 57 Figure 2.12 Representation of putative orthologous groups (pOGs) identification ............................................................................. 73 Figure 3.1 Transcript accumulation of conserved cold-induced transcription factors in response to mechanical agitation ...................................... 82 Figure 3.2 Transcript accumulation of conserved cold-induced transcription factors in response to cycloheximide treatment ................................. 83 Figure 3.3 Circadian clock regulation of CBF1 ....................................... 86 Figure 3.4 Circadian clock regulation of RAV1 ....................................... 88 Figure 3.5 Circadian clock regulation of CZF1 ....................................... 90 Figure 3.6 Circadian clock regulation of ZAT10 ...................................... 93 viii KEY TO ABREVIATIONS CAC = Cold Acclimation Capacity CHX = Cycloheximide DE = Differentially expressed EST = Expressed Sequence Tags FC = Fold Change G0 = Gene Ontology NAFT = Non acclimated Freezing Tolerance PUT = Putative Unique Transcripts Sc = Solanum commersonii St = Solanum tuberosum (common potato) TIGR = The Institute for Genomic Research ZT = Zeitgeiber time ix CHAPTER ONE Literature review Due to the sedentary nature of plants, they have evolved strategies to adapt to different environmental changes. Plants that grow in different climates exhibit differences in cold tolerance. Many plants from temperate regions, such as Arabidopsis, wheat, rye, barley and canola survive freezing temperatures and are able to cold acclimate, process whereby plants increase in freezing tolerance after exposure to low non-freezing temperatures (1,2). For instance, non-acclimated wheat plants are killed at freezing temperatures of about -5°C, but cold acclimated wheat can increase its freezing tolerance and survive to about -20°C (3). In contrast, plants that grow in tropical or subtropical regions, including crop species such as rice, maize, tomato and potato are freezing sensitive and generally do not cold acclimate (4-7). Improving the tolerance of crop species to lower temperatures would increase the land where the crops could be grown and would also lengthen the growing season, improving the food supply for a growing world population. Freezing Damage and Cold Acclimation When plants are exposed to freezing temperatures, ice formation occurs in the extracellular space due to the extracellular fluid having a higher freezing point than the intracellular fluid. Freezing of the extracellular fluid increases the solute concentration outside the cell. This high osmotic potential draws out water from the cell causing dehydration (1,8). Freezing-induced dehydration can cause a series of cellular injuries including protein denaturation and precipitation of molecules, and membrane damage. Freezing-induced dehydration can cause different types of membrane lesions. At freezing temperatures between -2°C and -4°C the freezing-thaw cycles can cause expansion-induced cell lysis; at lower temperatures, between -4°C and -10°C, the most common form of membrane injury is the phase transition of bilayer lipids from lamellar to hexagonal ll (an interbilayer event that involves fusion of cellular membranes); at temperatures below -10°C severe dehydration occurs and causes fracture jump lesions, an alteration in membrane ultrastructure that is manifested. as localized deviations of the plasma membrane fracture plane to subtending Iamellae (2,9,10). The gradual exposure to low non freezing temperatures during fall allows plants to increase their freezing tolerance during the winter. This cold acclimation process involves adjustment of metabolism and cellular functions to the constraints imposed by low temperature and the induction of freezing tolerance (1). Cold acclimation induces changes in membrane lipid composition, increasing levels of fatty acid desaturation in the membrane phospholipids (11). It also prevents expansion-induced lysis and the formation of hexagonal M phase lipids in the plasma membrane (9). Additionally, there is accumulation of small cryoprotective molecules such as soluble sugars and proline during cold acclimation (12,13), and it has been suggested that through interaction with proteins and membranes by hydrogen bonding these could prevent protein denaturation and stabilize membranes (14). Another event that occurs during cold acclimation is the accumulation of certain hydrophilic polypeptides that help to stabilize membranes against freeze-induced damage. Among these polypeptides are the COR (cold- regulated) proteins such as COR6.6, COR15a, COR47 and COR78 (10,15,16). COR47 is a member of the group II late embryogenesis abundant (LEA) type proteins, also known as dehydrins (10,17). The role of these hydrophilic polypeptides has been elusive for many years but they are thought to be E]DDCICICJDDZIZIJCICIDClnClCJ'CiilDDIIDUBUDJDDDJD'DDDJDDDDUDEU been shown to increase freezing tolerance of isolated protoplast due to a decrease in incidence of lamellar to hexagonal ll phase transitions. These occurred in regions where the plasma membrane comes into close proximity with the chloroplast envelope upon freeze-induced dehydration (18). Microarray studies have revealed induction of many genes by low temperature. Among these, numerous genes encode proteins that share the COR proteins property of being highly hydrophilic but their functions are still unknown (19,20). Cold responsive CBF pathway in Arabidopsis thaliana Discovery of CBF The process of cold acclimation involves changes in gene expression that accounts for specific biochemical changes that are thought to contribute to the increase in freezing tolerance. The COR gene transcripts accumulated after 4h of cold treatment, and it has been observed that they can stay induced for 2 weeks in the cold. COR gene transcripts come back to their warm levels as soon as 4h after transfer to warm (deacclimation) (21). A cis-acting DNA regulatory element present in the COR gene promoters was identified to be responsible for their cold-induction, the C-repeat/Dehydration Responsive Element (CRT/DRE) (core sequence = CCGAC) (22). The CBF1 (CRT/DRE Binding Factor 1) transcription factor was found to bind to this CRT/DRE element and activate transcription of CRT/DRE reporter gene fusions in yeast (23). CBF proteins are members of the AP2/ERBP family of transcription factors (24). In the model plant Arabidopsis, there are 6 members of the CBF family, three of which are cold-induced: CBF1, CBF2 and CBF3, also known as DREB 1B, 1C, and 1A respectively (25,26). CBF1-3 are the major regulators of cold acclimation in Arabidopsis. CBF transcripts accumulate soon after exposure to 4°C. They are detectable by Northern hybridization within 15 minutes of exposure to low temperature and they peak at around 2h (27). Constitutive expression of any of the three cold-inducible CBFs in Arabidopsis leads to induction of the CBF target genes at warm temperatures and results in the ability of these plants to be freezing tolerant without the requirement of a period of cold acclimation (26,28,29). CBF regulon and its predominant role in configuring the low temperature responses in Arabidopsis Microarray technology has allowed the identification of hundred of genes that are responsive to low temperature in Arabidopsis. The COR gene transcripts accumulate in the cold soon after CBF transcript accumulation (27). Other genes that accumulate by low temperature include enzymes involved in synthesis of protective sugars such as sucrose synthase and galactinol synthase. By overexpression of each of the three CBFs in Arabidopsis and comparison to the cold regulated genes, about 100 genes have been identified as members of the CBF regulon in Arabidopsis, but beside those, several hundred cold-induced genes fall outside CBF regulation, which implies that additional transcription factors play a role in the process of cold acclimation (19,20). Despite the presence of additional cold-responsive pathways, the CBF pathway plays a predominant role in cold acclimation. Among the genes that are cold responsive, the most highly induced ones are members of the CBF regulon (20). It is known that overexpression of any of the CBF proteins in Arabidopsis leads to constitutive expression of the CBF regulon and enhanced freezing tolerance without cold acclimation (26,28-30). Besides the large changes in gene expression caused by CBF overexpression, changes in metabolite profiles have also been studied. Metabolite profiling has demonstrated that 79% of the metabolites that increase in response to low temperature in Arabidopsis Wassilewskija-2 (Ws-2), also increase in non-acclimated plants by AtCBF3 overexpression. Moreover, the Arabidopsis Cape Verde Islands-1 (Cvi), which is less tolerant to freezing, expressed less CBF1-3 and CBF target genes in response to the cold, and the low temperature metabolome of Cvi-1 plants was depleted in metabolites affected by CBF3 overexpression (31). CBF regulation in response to low temperature Given the predominant role of CBF in freezing tolerance, much effort has been put into identifying regulators of its induction. Inducer of CBF expression 1 (ICE1) is a MYC-like bHLH transcriptional activator that has been identified as a positive regulator of CBF3 in Arabidopsis. ICE1 binds specifically to the Myc recognition site in the CBF3 promoter. A point mutation in ICE1 (ice1 mutant) almost completely abolished expression of the endogenous CBF3 gene, but CBF1 and CBF2 expression are only reduced at 1h of cold treatment and reach similar levels to wild type after 6h of cold. Many CBF target genes have decreased expression in the ice1 mutant after cold treatment, which leads to a reduction in plant chilling and freezing tolerance. Overexpression of ICE1 enhances the expression of CBF2, CBF3, and the CBF regulon in the cold and improves freezing tolerance (32). However, ICE1 overexpression is not able to induce any of the three CBF transcripts at warm temperatures, indicating that ICE1 alone is not sufficient to induce CBF expression (32). Maybe ICE1 needs to have a modification that only occurs in the cold, or alternatively there are other factors needed to activate CBF. ICE1 is expressed constitutively, and cold induces the degradation of ICE1 through the E3 ligase, HOS1, a negative regulator of cold acclimation, that targets ICE1 for ubiquitination (33). SIZI, a SUMO E3 ligase, mediates SUMO (small ubiquitin-related modifier) conjugation of ICE1 during cold acclimation, reducing its polyubiquitination and leading to an enhanced cold induction of CBF and COR genes and increased freezing tolerance (34). Given that there is little effect on CBF1 and CBF2 expression in the ice1 mutant, it is thought that the regulation of these three CBF genes may be independent. Another transcription factor has been recently identified as a positive regulator of CBF2 expression, the CAMTA3. This protein belongs to the CAMTA family of calmodulin-binding transcription factors that has six members in Arabidopsis. A camta3 single knock out mutant had 50% reduction of CBF2 transcript and 40% reduction of CBF1 transcript under low temperature compared to WT. CAMTA3 binds to the conserved motif 2 (CM2) present in the CBF2 promoter. The double camta1/camta3 mutant is impaired in freezing tolerance after cold acclimation, indicating that CAMTA1 and CAMTA3 are both needed to attain full levels of freezing tolerance (35). Negative regulation of CBF1-3 gene expression has also been identified. The null cbf2 mutant has more CBF1 and 3 transcripts in warm and cold conditions, suggesting that CBF2 could be a negative regulator of CBF1 and CBF3 (36). Two other transcription factors repress CBF1-3 accumulation. The overexpression of Myb15 (a R2R3 type Myb transcription factor) and ZAT12 (a zinc finger transcription factor) reduce CBF1-3 cold accumulation (20,37). Myb15 and ZAT12 transcripts are cold-induced. The Myb15 protein binds to the Myb recognition sequences in the promoters of CBF1-3 genes. A knock out mutation of Myb15 causes increased expression of CBF genes under low temperature. However, overexpression or knock out of Myb15 does not change the transcripts of CBF regulon genes such as COR15 or RDZ9a genes. All these studies suggest that the regulation of CBFs is very complex. CBF regulation in response to other environmental cues CBF genes are not only responsive to low temperature signals but also to other environmental changes. It has previously been shown that Arabidopsis CBFs are induced in response to mechanical agitation and inhibition of protein synthesis (cycloheximide treatment) (20,27,38). The CBF2 promoter has two sequences, lCEr1 and ICEr2 (Induction of CBF expression region 1 and 2) that impart cold-regulated gene expression and also stimulate transcription in response to mechanical agitation and the protein synthesis inhibitor, cycloheximide. It is possible that there is a regulatory link between these different responses that it is yet to be discovered. In addition to these responses Arabidopsis CBF3 has also been shown to be regulated by the circadian clock (39). Harmer et al. (2000) showed that at warm temperature, CBF3 transcripts undergo circadian cycling, with a peak at ZT 4 (Zeitgeiber Time, hours after dawn) and a trough at ZT 16 (39). Given this circadian regulation of CBF3 at warm temperatures, the question was raised whether the circadian clock also gated the expression of CBF1-3 in response to low temperature. Fowler et al. (2005) showed that indeed it did. The circadian clock has a gating effect on the low temperature induction of CBF1-3 genes. When plants are shifted to low temperature at ZT 4 (4h after dawn), which coincides with the peak of CBF3 circadian expression in the warm, the cold induction of CBF1-3 is higher than when transferred to cold at the trough of CBF3 expression (ZT 16) (40). Furthermore, disruption of the circadian clock by overexpression of Circadian Clock Associated 1 (CCA1), a Myb-related transcription factor member of the Arabidopsis clock (41,42), also disrupted the cycling of CBF1-3 (40). Circadian gating of CBFs has also been suggested in tomato. When Solanum lycopersicum (tomato) and its wild relative Solanum pimpinellifolium were entrained under a 16:8 L:D photoperiod, CBF1 expression cycled, reaching higher responsiveness during the light period in both species. When plants were shifted to constant dark or constant light CBF transcripts continued to cycle, however the peaks and troughs did not correspond to those observed in plants grown under a normal 16:8 L:D photoperiod (43). Another environmental factor that has been observed to be involved in CBF regulation is light quality. More CBF1-3 transcripts accumulate at 16°C under a low R/FR ratio compared to a high R/FR, and this increase is dependent on the circadian clock. Plants grown at 16°C in low R/FR light are more freezing tolerant than plants treated with high R/FR light, indicating that this light quality-dependent increase in CBF expression is sufficient to confer freezing tolerance at higher temperatures than those required for cold acclimation (4°C). In nature, a decrease in R/FR light occurs during twilight periods. It is reasonable to think that low temperature, shorter day length and longer twilight periods during fall will trigger CBF expression to confer freezing tolerance before the winter comes (44). Recently, the transcription factor PIF7 (Phytochrome Interacting Factor 7) was demonstrated to bind specifically the G-box of CBF1 and CBF2 promoters. PIF7 is a basic helix-loop-helix (bHLH) transcription factor that interacts with the far red light-absorbing Pfr form of phytochrome B (phyB) (45). Transactivation experiments showed that PIF7 acts as a transcriptional repressor for CBF2 expression and this activity is mediated by two PIF7- interacting factors, T001 and PhyB (components of the circadian clock and the red light photoreceptor respectively). PIF7 is localized to the nucleus and it is expressed under warm conditions in rosette leaves. After entrainment by the clock in a 12:12h L:D photoperiod, the pif7 mutant showed no repression of 10 CBF1 and CBF2 under continuous light, which translated in no cycling of these genes, indicating that PIF7 functions as a transcriptional repressor of CBF1 and CBF2 under circadian control (46). The integration of knowledge about regulation of CBF under different environmental signals will enable a better understanding of CBF regulation under low temperature stress. CBF independent pathways in cold acclimation Some recent evidence suggests that, besides the CBF pathway, there are CBF-independent pathways that contribute to freezing tolerance. Microarray studies have revealed that there are hundreds of genes that are cold-regulated but are not affected by CBF overexpression in warm grown plants, even though constitutive expression of CBF is sufficient to increase freezing tolerance in warm conditions (19,20). These genes may be part of a CBF independent pathway in the cold or may require one or more additional factor(s) that are only present in the cold. Moreover, when CBF1, 2 or 3 is overexpressed, cold acclimated plants have an increase in freezing tolerance compared to warm non acclimated plants (26,28). This additional freezing tolerance may be due to CBF independent pathways, additional components that act in concert with CBF to confer freezing tolerance or may just be a quantitative effect due to more CBF transcript present in the cold as a result of endogenous CBF accumulation at .low temperature. 11 Mutational analysis has supported the idea of CBF independent pathways involved in freezing tolerance. For instance, the hos10 mutation in the R2R3-type Myb transcription factor, produced a mutant that is extremely sensitive to freezing and is unable to cold acclimate; however the cold-induction of CBF1-3 transcripts and CBF target genes (COR15a, COR78) is not altered (47). Similarly, Gigantea (GI), a protein involved in developmental regulation of flowering in response to day length and circadian clock, may be involved in CBF-independent cold acclimation. GI transcript has been shown to be cold- induced. The gi-3 mutation shows increased sensitivity to freezing stress with no changes in the transcript accumulation of CBF1-3 or CBF target genes (COR15a, coma, KIN1) (48). Cold responsive CBF pathway in other plant species Considerable evidence suggests that the CBF cold responsive pathway is present in a wide variety of plant species and that it functions in the development of freezing tolerance in many of them. CBF proteins are highly conserved and are not limited only to cold acclimating plants. Cold-inducible CBF genes have been identified in B. napus, wheat, rye, barley, rice, maize, Populus, tomato, and potato among others (3,43,49-52). The region of highest amino acid sequence identity among the CBF proteins is within the AP2/EREBP (Apetala2/Ethylene Responsive Element Binding Protein) DNA binding domain (3). This domain is common to several transcription factors in plants known as 12 the AP2/EREBP proteins (53). CBF proteins form a subset of this group and are characterized by having two regions flanking the AP2/EREBP DNA binding domain. These two regions (PKK/RPAGRxKFxETRHP upstream and DSAWR downstream of the DNA binding domain) are called the “signature sequences” and are very well conserved in CBF-like proteins from B. napus, wheat, rye, tomato, and barley; and less conserved in pepper, rice, and maize. Several studies have shown that CBFs from species other than Arabidopsis also bind to the CRT/DRE DNA binding motif. In B. napus, the CRT/DRE element is critical to the low temperature response of the Bn115 gene (54). BnCBF5 and BnCBF17 (B. napus homologs of the Arabidopsis CBF proteins) are able to bind to this element in vitro and are able to trans activate promoter regions containing CRT/DRE elements fused to a lacZ reporter gene in yeast (55). The cold-inducible HvCBF1 from barley is able to bind to the GCCGAC motif and is involved in the regulation of cold-responsive genes from barley (56). The rice cold-inducible OsDREB1A (ortholog of AtCBF) has been ' shown to bind efficiently to the GCCGAC CRT/DRE elements (51). Another example is the CBF maize ortholog ZmDREB1A that is also cold-inducible and able to bind the DRE motif (52). Interestingly, it has been shown that CBFs confer freezing tolerance in other plant species. Overexpression of any of the three Arabidopsis CBF genes increase freezing tolerance of B. napus in non-acclimated conditions (3). When maize ZmDREB1A (an ortholog of CBF) is overexpressed in Arabidopsis there is constitutive expression of AtCOR15a and other cold-induced genes and this 13 results in an increase in freezing and drought tolerance (52). Overexpression of AtCBF1 increases freezing tolerance of Populus (50). CBFs have also been shown to have an effect in freezing sensitive species. Overexpression of wheat CBF2 in transgenic tobacco increase freezing tolerance (57). Overexpression of AtCBF3 increased chilling tolerance in tobacco and resulted in a small increase in freezing tolerance in potato (58,59), while AtCBF1 overexpression has been demonstrated to increase freezing tolerance in transgenic potato plants (50,60). In tomato, a freezing and chilling sensitive non acclimating plant, there are three orthologs of the CBF genes, LeCBF 1, 2 and 3, which, as in Arabidopsis, are present in tandem array in the genome. However, only LeCBF1 is induced by low temperature (61). Overexpression of LeCBF1 or AtCBF3 in Arabidopsis leads to induction of the CBF target genes and an increase in freezing tolerance. This indicates that LeCBF1 encodes a functional CBF protein. However, overexpression of the LeCBF1 or AtCBF3 genes in tomato plants do not increase freezing tolerance and they regulate very few genes (61), indicating that AtCBF3 overexpression is not sufficient to induce cold acclimation in tomato. This suggests that tomato and Arabidopsis have critical differences, resulting in an absence of response to AtCBF in tomato. It is possible that there is some co-activator, a protein that works in concert with CBF that is not present in tomato. Alternatively the CBF target genes do not have CRT/DRE elements in their promoters, which would explain why CBF 14 does not induce many genes in tomato. To date, it is unknown why tomato is freezing sensitive. Another level of conservation in the CBF pathway may lay upstream of CBF. Recently, two wheat ICE (Inducer of CBF expression) genes have been identified: TalCE41 and TalCE87. Both genes are expressed constitutively as is the Arabidopsis ICE1 (62). TalCE41 and TalCE87 bind to different MYC elements in the wheat TaCBFIVd-BQ promoter, and both TalCE proteins can activate TaCBFlVd-BQ transcription when transiently transformed in N. benthamiana plants. As observed with the AtICE1 in Arabidopsis (32), overexpression of either TalCE41 or TalCE87 genes in Arabidopsis increased freezing tolerance only after cold acclimation, suggesting that other factors induced by low temperature are required for ICE activity. All these studies suggest that the CBF cold-responsive pathway is conserved in diverse plant species. However, little is known about what the differences are between freezing tolerant and freezing sensitive species. Microarray technology has provided the opportunity to study gene expression changes under cold stress at a whole genome level. For instance, wheat, one of the most freezing tolerant crop plants, has been used to compare cold transcriptomes among cultivars with different levels of freezing tolerance. The highly cold tolerant winter wheat cultivar CDC Clair was compared to the less tolerant spring cultivar, Quantum (63). It was found that a large number of genes had altered levels of expression in each cultivar and there were significant differences in expression between the two cultivars. After 6 hours of 15 cold, the number of up regulated genes was higher in the spring cultivar; however throughout the time course (up to 14 days of cold acclimation) the number of up regulated genes was higher in the winter wheat. In the future, it would be very interesting to find what the differences are between freezing sensitive and freezing tolerant species. Is there one gene or many genes? Are there any differences in the cis elements that drive cold expression? Are these factors transferable from freezing tolerant to freezing sensitive species? Answers to these questions would not only be important for our basic knowledge of plants responses to environmental changes, but also to improve crop production. 16 CHAPTER TWO TRANSCRIPTOME PROFILES OF SOLANUM SPECIES WITH DIFFERENT LEVELS OF FREEZING TOLERANCE INTRODUCTION Cold acclimation is the process whereby plants increase their level of freezing tolerance by exposure to low non-freezing temperatures. The CBF (CRT/DRE Binding Factor) family of transcription factors (CBF1, CBF2 and CBF3, also known as DREB 1B, 1C, and 1A respectively) is a major regulator of cold acclimation in Arabidopsis. The CBF genes are induced soon after exposure to 4°C and CBF transcripts are detectable by Northern hybridization within 15 minutes (27). Arabidopsis CBF proteins bind the C-repeat/Dehydration Responsive Element (CRT/DRE) (core sequence = CCGAC) present in the promoters of many cold responsive (COR) genes to induce their expression (20,23). Transcripts for the COR genes accumulate soon after CBF transcript accumulation in response to cold (27). About 100 genes have been identified as CBF target genes in Arabidopsis, but beside these, several hundred fall outside CBF regulation, which implies that additional transcription factors play a role in the process of cold acclimation (19,20). Considerable evidence suggests that the CBF cold response pathway is present in a wide variety of plant species and that it functions in the 17 development of freezing tolerance in many of them. Genes that may encode CBF orthologs have been found in many species; they are highly conserved and not limited to cold acclimating plants. For instance, orthologs of CBF genes from B. napus, wheat, rye, tomato, barley, pepper, grape, rice, and maize are induced in response to low temperature (3,49,51,52,64,65). Furthermore, CBF proteins are highly conserved. The region of highest amino acid sequence identity among the CBF proteins is within the AP2/EREBP (Apetala2lEthylene Responsive Element Binding Protein) DNA binding domain (3,52,66). Overexpression of any of the CBF genes in Arabidopsis leads to constitutive expression of the CBF regulon and enhanced freezing tolerance (26,28-30). Additionally, it has been shown that CBF overexpression increases freezing tolerance in other plant species. Overexpression of any of the three Arabidopsis CBF genes increases freezing tolerance of B. napus, in non- acclimating conditions (3). Overexpression of AtCBF1 has also been shown to increase freezing tolerance under non-acclimating conditions in populus and potato, and overexpression of AtCBF3 has a small increase in freezing tolerance in potato (50,51 ,59,60). Despite the increasing evidence for conservation of the CBF pathway, little is known about the differences and similarities between plants with different levels of freezing tolerance. It has been previously shown that tomato, a freezing and chilling sensitive non acclimating plant, encodes three orthologs of the CBF genes, LeCBF 1, 2 and 3, which, as in Arabidopsis, are present in 18 tandem array in the genome. However, only LeCBF1 is induced by low temperature (61 ). LeCBF proteins are 70-84% identical to each other and 51-59% identical to those of Arabidopsis CBF proteins. The 3 tomato CBF proteins have the two conserved “signature sequences” that distinguishes the CBF proteins from other AP2/ERBP proteins (PKKPAGR and DSAWR). Overexpression of LeCBF1 or AtCBF3 in Arabidopsis leads to induction of the CBF target genes and an increase in freezing tolerance. This indicates that LeCBF1 encodes a functional CBF protein. However, overexpression of the LeCBF1 or AtCBF3 genes in tomato plants does not increase freezing tolerance (61), indicating that AtCBF3 overexpression is not sufficient to induce cold acclimation in tomato. This suggests that tomato and Arabidopsis have critical differences, resulting in an absence of response to AtCBF in tomato. Recently, it was found that Solanum tuberosum (common potato) (St) and its wild relative Solanum commersonii (Sc) may have conserved parts of the CBF cold responsive pathway. So is able to cold acclimate and has a moderate level of freezing tolerance; it is killed at -4.5°C and after cold acclimation is able to survive down to — 11.5°C. On the other hand, St does not cold acclimate and is freezing sensitive; it is killed at -3°C before and after cold acclimation (7). Sc has four CBFs genes and St has five. Both these species also have CBF1-3 genes in tandem array in their genomes; and each of these species has two cold-induced CBFs (CBF1 and CBF4) (43). The Sc and St CBFs are highly similar to those of Arabidopsis (54-64% identity), and they also 19 contain the signature sequences. It has also been demonstrated that overexpression of AtCBF1 or AtCBF3 increases freezing tolerance of St by 2°C, but AtCBF2 does not increase freezing tolerance; and AtCBF1 increases freezing tolerance of So by 4°C (59,60). The low temperature transcriptomes of So and St are as yet unknown. It is not clear if they are similar to each other or what portion of them is regulated by CBF. Because these are very closely related species with different levels of freezing tolerance, they constitute an excellent model to study conservation of the cold transcriptomes and CBF regulons. Variation in gene expression can result in phenotypic differences. Studies of comparative transcriptomes are still not very prevalent. Some studies have focused on expression variation, for instance, between species of primates and between yeast species (67,68). In plants, a few recent studies have attempted to compare differences in gene expression between species challenged by stress (6369-71). The main goal of the experiments described in this chapter was to identify and compare the low temperature transcriptomes of So and St, and also their CBF regulons. Using the St 10K cDNA array, the cold- and CBF- transcriptomes of these species were analyzed, and by identification of putative orthologous groups between St and Arabidopsis the transcriptomes of Sc and St were compared to that of Arabidopsis. Hierarchical clustering analysis of the low temperature transcriptomes of So and St indicates that, in general, they are very similar. With the criteria used (2FC, p<0.05) there was more than 50% 20 overlap between cold transcriptomes of the two Solanum species. In general, there are no obvious differences between the Sc and St cold transcriptomes that could account for their differences in freezing tolerance. Only around 10% of the cold regulated ESTs in Solanum species, that have Arabidopsis orthologs, were also identified as cold regulated in Arabidopsis. This indicates significant differences between the two Solanum species and Arabidopsis cold transcriptomes. The Sc and St CBF regulons were identified. There are significant differences between the genes that are regulated by AtCBF3 overexpression in So and St. About 48% of the Sc CBF regulon is also part of the St CBF regulon, suggesting that these CBF regulons are not very well conserved. When compared to Arabidopsis, only 14% of the Sc and St CBF regulons identified in this study are also part of the Arabidopsis CBF regulon, indicating that there are important differences among these CBF regulons. 21 RESULTS The transcriptomes of both 8. commersonii and S. tuberosum are significantly altered in response to low temperature 8. commersonii (Sc) and S. tubemsum (St) have very different tolerances to freezing. Moreover, So is able to cold acclimate but St is not. It was hypothesized that differences in their low temperature gene expression could result ultimately in their differences in freezing tolerance. Therefore, the low temperature transcriptomes of these two Solanum species were identified and compared. The TIGR potato cDNA array (10K, version 4) was used to compare the low temperature transcriptomes of Sc and St. The array represents about 10,000 of the ~70,000 Putative Unique Transcripts (PUT) available at PlantGDB. Plants were grown for three weeks at 25°C and then transferred to 2°C for 2h, 24h and 168h. RNA was isolated from plants at the various time points and their transcriptomes determined. First, the log ratios of all the Expressed Sequence Tags (ESTs) in the array were hierarchical clustered to compare the general patterns of cold regulated kinetics between Sc and St low temperature transcriptomes. This hierarchical cluster was done including all flagged spots and is prior to statistical selection. Fig 2.1 shows this hierarchical cluster done with average log ratios of three biological replicates per time point. A large part of the ESTs spotted on the array showed cold regulation. The highest up regulated cluster (A) 22 correspond to ESTs cold—induced in Sc and St at 2h and St at 24h of cold. Only some of the ESTs in this cluster show up regulation in So at 24h of cold. This indicates that the ESTs in this cluster are only transiently induced by cold in So, but they stay at least until 24h of cold in St. This different kinetic between Sc and St in cluster A could be a reason why these species are different in freezing tolerance, but there is no enough evidence at this stage to support that. Another major cluster is B; this corresponds to a group of ESTs that are cold-induced only at 168h in both Sc and St. A small cluster (C) corresponds to ESTs that are induced in response to low temperature at all time points tested in both species. A fourth cluster is D; the pattern of cold induction in this cluster indicates up regulated ESTs at 24h and 168h in both So and St. The highest cold down regulated cluster (E) has ESTs from So 168h and St 168h. The hierarchical cluster analysis of all the data revealed that the Sc and St transcriptomes are largely changed by low temperature exposure, with very similar responses in Sc and St at the different cold time points tested and possibly with some kinetic differences. However, the analysis described here serves the purpose of a general overview of all EST clones present in the array, lacking statistical significance. 23 St 168h SC 168h Cluster A - Cluster B Cluster C .9 “'5 Cluster D (I N C) O _I Cluster E Fig 2.1: Hierarchical clustering and expression profiles of Sc and St EST clones at 2, 24 and 168h of cold treatment (2°C). Sc: 8. commersonii, St: S. tuberosum. Data showed as average log ratio from 3 biological replicates. The figure shows all spots on the array (including bad flagged spots) prior to statistical selection. 24 After the previous general analysis was done, the idea was to identify a core set of cold responsive genes in each of the two Solanum species. To do that, a list of ESTs that were reproducibly cold regulated were obtained, employing linear models (Limma package, (72)) as a statistical tool to rank the ESTs in order of evidence (p value) of differential expression (DE), thus addressing any variability between biological replicates. After ranking the ESTs based on p value, a p<0.05 and 2 fold change (F0) was used as the cutoff. A list of cold-regulated ESTs for So and St can be accessed at http://www.jmmsu.edu/Facultypages/NSF MFT Site/gatahtml. About 13% of the ESTs on the array were cold-induced and 5% cold-repressed at one or more of the time points in So. Similarly, about 10% of the ESTs were cold- induced and 6% cold-repressed in St. Given that without applying selection criteria, a large part of the ESTs in the array showed cold regulation, but after applying the criteria (2FC, p<0.05) only small percentage of ESTs were cold regulated, this indicates that many ESTs were lowly expressed and there was large technical or biological variability. Similarities and differences in the cold-induced gene sets of So and St A total of 1532 and 1084 ESTs were cold-induced in So and St, respectively (totals were determined by combining the results from all three cold-treated time points). Around 50% of the cold-induced ESTs in So were also 25 induced in St (Fig 2.2a). Early in the cold (2h and 24h) there was a large overlap between cold-induced ESTs in So and St, and there were more ESTs induced apparently only in So (Fig 2.2b). However, late in the cold (168h) (Fig 2.2c) there was a large group of ESTs that were cold-induced apparently only in St. These results suggest that the cold-regulation of these genes could be different between Sc and St. The results presented above indicated that even when statistics are applied to the data, the overlap between the cold-induced transcriptomes of So and St is still considerable. The differences could be real or apparent due to the arbitrary criteria used to define cold-induced genes. Changing the criterion to make it more stringent or more relaxed increased the overlap between cold- induced ESTs in So and St. Moreover, with the criteria of two-fold change and p<0.05, there were no cold-induced ESTs for the 2h St RNA, but by relaxing the criteria to two-fold and p<0.07, the number of ESTs that were cold-induced at 2h was 675. These findings indicate that the overlaps detected between the Sc and St transcriptomes were minimal estimates of conservation. The functions encoded by the genes that were significantly cold-induced at early and late time points in both So and St were compared to determine whether the functional categories of the genes changed with time of exposure to low temperature (Fig 2.2d). The results indicated that the categories did not change much between the early and late samples, but that there were some differences. Earlier in the cold there were more cold-induced ESTs annotated as transcription factors than later in the cold. Later in the cold, there were more 26 ESTs annotated as structural molecules (like ribosomal proteins) and translation factors being induced. These results suggest protein synthesis is activated later in the cold. 27 a. Total Sc St b. Early cold-induced c. Late cold-induced Sc St Sc St d. 60 Cold induced in both Solanum I Early 50 I Late 40 30 20 10 0 III- IllIILJ‘ l..9:\ _. .I.-__ Jul <19 \ t s “eds“; \ «9 (55859 «8:98 6&¢&¢(b§6&<¢§9 e$ée°°teeéo°eqo6<8 396$: (span? (5} (96> (flog 0‘69 oosxboo V“; 9". “\d‘ i as?" S \3’0 6&1), \6‘920‘0 o @669 \2} $39 49%) 49 \000 t; gig; gs ‘ \0" 02° &6\Q& 65" $5 (9&0 $650 0 0c? 9&0 05°63 '6 Fig 2.2: Comparison of early and late cold-induced ESTs (2FC, p<0.05) in both Solanum species (a) Number of total ESTs cold-induced (determined by combining the results from all three cold treated time points). (b) Number of ESTs cold-induced early (2h and 24h at 2°C) and (c) late (7 days at 2°C). (d) The ESTs that were induced either early or late in both Solanum species were classified according to their functional categories. Each category is shown as a percentage of the total 651 early cold-induced or total 160 late cold-induced ESTs 28 Similarities and differences in the cold-repressed gene sets of So and St A total of 530 and 688 ESTs were cold-repressed in Sc and St, respectively (totals include results from the three cold-treated time points) (Fig 2.3a). Around 70% of the cold-repressed ESTs in So were also down regulated in St. The percentage of overlap between cold-repressed ESTs was similar between early and late cold treatments (Fig 2.3b, and c). Functional analysis of the ESTs cold-repressed in both So and St indicated that early in the cold there were more down regulated genes annotated as transferases and oxygen binding proteins than later in the cold (Fig 2.3d). Together, these results indicate that the transcriptomes of So and St are significantly altered in response to low temperature. Even though some small differences can be identified between Sc and St cold transcriptomes, there were not dramatic differences at a global level that could account for their differences in freezing tolerance, but rather similarities in the patterns of gene expression were identified. 29 Total Sc St b. Early cold-repressed 0. Late cold-repressed Sc St Sc St (I. 45 Cold repressed in both Solanum 4° - Early 35 I Late 30 25 20 1 5 10 5 H ,i.. i... I n .. l. .i . _ L, \ $9 8M9 90 to‘ so 0‘ 6“ ‘° \0‘ s9 «:0 s96" 0‘ ‘6‘ 418°- °°" 06: web Gcbtécbb «2:39 9:0: ‘o 9:19 9°50" «0° «9 $66 <§$§°e affléébecb 5"? 6° K00 c9“ (59 v“ v-Qo o" o“ 8909.3; @6945 0° \5 0 9‘9 «09-00 '9 3°29 Q5: sage,» s s” 0 '\\ 0°~9°’\"‘-e° at“ ‘ *‘° 6:;2‘5‘60 (5,0 0 ex 06‘“ @0 68‘Q $.34? o ‘59 4° ra‘ @{oo o o“? 9 09° ‘9 Cl 6% Fig 2.3: Comparison of early and late cold-repressed ESTs (2FC, p<0.05) in both Solanum species (a) Number of total ESTs cold-repressed (determined by combining the results from all three cold treated time points). (b) Number of ESTs cold-repressed early (2h and 24h at 2°C) and (0) late (7 days at 2°C). (d) The ESTs that were repressed either early or late in both Solanum species were classified according to their functional categories. Each category is shown as a percentage of the total 56 early cold-repressed or total 351 late cold-repressed ESTs. 30 Comparison of the low temperature transcriptomes between the two Solanum species and Arabidopsis The major goal is to understand to which extent cold responsive pathways are conserved in plant species. To explore conserved aspects of the cold response pathways, experiments were conducted to determine the degree to which the cold-regulated genes in the two Solanum species were also cold- regulated in Arabidopsis. To conduct this comparison, a list of putative orthologous gene groups (pOG) between S. tuberosum (St) and Arabidopsis was obtained from Shinhan Shiu’s laboratory. This list was generated using all the Arabidopsis protein sequences (T AIR) and all the potato protein sequences predicted from the ~70,000 PUT (Putative Unique Transcripts) available at PlantGDB (see methods). This resulted in the identification of 8,714 pOGs between Arabidopsis genes and potato PUTs (Fig 2.4). A pOG may have more than one Arabidopsis gene and/or more than one potato PUT. The potato genome is not known therefore the ESTs that are spotted in the potato array are only small portions of potato genes. There are many ESTs that cannot be assembled into transcripts (PUTs), because there are not more ESTs known for those genes, and a single EST could be just a small portion of a transcript therefore not enough sequence to identify its Arabidopsis ortholog gene. That is why, only the ESTs that could be assigned to transcripts (PUTs) were used to identify their Arabidopsis ortholog. From the 11,366 ESTs represented in the potato array, most of them (9,900 ESTs) can be assigned to 31 a PUT, and only those were considered in the pOGs identification. From these 9,900 ESTs only 3,934 ESTs belong to pOGs with Arabidopsis members (within the 8,714 pOGs). The rest of pOGs between At and St include PUTs that are not represented in the potato array (Fig 2.4). 32 5, 956 ESTs r 9:900 5873 do not have At Assigned to orthologs 1 ~12,000 PUT r (of the 70K 3,944 ESTs 8,714 pOGs 11.366 ESTs c PUT k"°W”)J belong to pOGs With At and St Potato array with At members members f 4 . L J 1’ Egg-[3:98 Cannot be Not assigned to 3.3“?“ to ra ido sis PUT p L J PUTs that are not present on potato array (~58,000 PUT) Fig 2.4: Putative orthologs groups (pOGs) between Arabidopsis and S. tuberosum. 8,714 pOGs were identified between Arabidopsis proteins and the 70,000 PUT (potato unique transcripts). From the 11,366 ESTs present in the potato array, 9,900 can be assigned to potato PUTs, and only those were considered for the pOGs identification. Only 3,944 ESTs present in the potato array belong to pOGs that have Arabidopsis members. The rest of pOGs between Arabidopsis and potato includes PUTs that are not present in the potato array. 33 The cold-regulated ESTs that were conserved between Sc and St were first analyzed to identify how many of them had Arabidopsis putative orthologs. From the 790 ESTs that were identified as cold-induced in both So and St above, 278 ESTs (35%) were represented in the list of orthologous groups and correspond to 244 pOGs (Fig 2.5a). From the 383 ESTs that were identified as cold—repressed in both So and St above, 174 ESTs (45%) were represented in the list of orthologous groups and correspond to 129 pOGs (Fig 2.5a). These results indicate that a big percentage of the cold-regulated ESTs in the two Solanum species do not have Arabidopsis orthologous genes under the criteria used. Of the 8,714 pOGs identified between St and Arabidopsis genes, only 2,944 pOGs were determined to have at least one Arabidopsis gene present on the ATH1 Affymetrix chip, and at least one potato EST present on the potato array. These pOGs represented in both arrays were considered for the following analysis (Fig 2.5b). To study the conservation of cold responsive pathways, the low temperature transcriptomes of So and St were compared to a list of Arabidopsis cold-regulated genes generated in our laboratory (unpublished data). This list was obtained from the AtGeneexpress website, from experiments done in 16:8h L:D photoperiod with cold treatments of 4°C for different times. The criteria of DE genes selection was 2FC and p<0.05, and the lists consisted of 1,151 cold- induced and 1,095 cold-repressed Arabidopsis genes. 34 Based on the 2,944 pOGs represented in both arrays, the overlap between the cold-induced transcriptomes of the three species was determined (Fi92.5c). Numbers of pOGs with at least one cold induced gene from each species were identified. As the potato array does not include all potato genes— likely well less than half—these values are minimum estimates of pOGs that include cold-regulated genes; that is, it is possible that a given pOG includes multiple genes, one or more of which is cold-induced, but the EST on the array is one that is not cold-induced. Forty pOGs with at least one cold-induced gene in each of the three species were identified (Fig 2.50). Thus, only 9% and 13% of the pOGs with cold-induced genes in So and St, respectively, are also cold-induced in Arabidopsis. Given that the Arabidopsis cold-induced list of genes comes from a microarray that represents almost its entire genome, this result likely indicates a real difference between cold-induced transcriptomes between the two Solanum species and Arabidopsis. Forty four percent of the 197 pOGs that are only induced in both Solanum species but are not induced in Arabidopsis are pOGs with genes of unknown molecular function; many are pOGs with genes encoding proteins with catalytic activity (10%), hydrolases (8%), transporters (6%) and kinases (5%). The 40 pOGs common to all three species (Table A1) include genes that are thought to have protective roles against freezing and drought such as LEA14 (Iate embryogenesis abundant) (73), ERD10 (early responsive to desiccation) and ERD14 (73,74), as well as ELIP (early light inducible protein), 35 which is thought to have a protective role against photooxidative damage (75). There were also conserved cold-induced transcription factors in these three species: Agamous-like 20, also called Suppressor of overexpression of CO (8001); NACO19; R026 (responsive to dessication 26); ADOF1; and Heat shock factor 8 (HSFA8). Thirteen pOGs with at least one cold-repressed gene in each of the three species were identified (Fig 2.5d). Thus, only 8% and 6% of the pOGs with cold-repressed genes in So and St, respectively, also have Arabidopsis cold repressed genes. The 13 pOGs that were cold-repressed in the three species (Table A2) include chloroplast metabolic genes such as carbonic anhydrase 1, glucose-6-phosphate dehydrogenase, and a cell wall metabolic gene, xyloglucan endotransglycosylase. In addition to the conservation in cold-regulated genes identified by pOGs present in both microarrays, pOGs with Arabidopsis and potato members that were represented in only one of the two arrays were also identified. Among the around five thousand present only in the Arabidopsis array, many were cold-regulated in Arabidopsis (Fig 2.5b). Therefore, from the total 1,145 pOGs cold-regulated in At that have putative potato orthologs, 65% (478 pOGs up and 268 pOGs down) cannot be compared to Solanum species, because they are not present on the potato array. Given this, the number of genes that are cold- regulated in the two Solanum species could be larger and therefore the conservation with Arabidopsis could be underestimated in these experiments. 36 Cold-induced Cold-repressed 278 EST (35%) 174 EST( 4)5% . / 3,714 pOG J between At and St . __ Present in both arrays Used for transcriptoms comparison 1/ i Y Li\\ c_ Cold-induced d, Cold-repressed At At 260 pOGs 139 pOGs 441 :CQGS 308 ptOGs 166 pOGs 204 pOGs Fig 2.5: Comparison of cold transcriptomes of Arabidopsis (At), 8. commersonii (Sc) and S. tuberosum (St). (a) Cold-induced and cold-repressed ESTs in So and St, showing the percentage that have Arabidopsis putative orthologous genes. (b) Putative orthologous groups (pOG) distribution on the Arabidopsis ATH1 Chip and potato cDNA array. Venn diagram shows pOGs with at least one Arabidopsis gene on the ATH1 chip (green), at least one potato clone on the potato cDNA array (brown) or at least one gene from each species in both arrays. Among the pOGs that are present only in one of the two arrays, the number of cold-induced or cold-repressed pOGs in the corresponding species is shown. From the pOGs present in both arrays, overlaps of (c) cold-induced or (d) cold-repressed pOGs are shown. 37 Together these results suggest that a large percentage (around 60%) of the cold-regulated ESTs in the two Solanum species do not have Arabidopsis orthologs. From the ones that do have, a large percentage (around 90%) is not cold regulated in Arabidopsis. This suggests differences in the evolution of low temperature transcriptomes between the two Solanum species and Arabidopsis. Conservation of cold-regulated transcription factors A major goal is to determine the extent to which cold regulatory pathways are conserved in plants. It was hypothesized that differences in freezing tolerance between Sc and St could be explained by differences in the cold-induction of transcription factors. In the previous section, 27 conserved orthologous groups (pOGs) were identified as cold-induced in the two freezing tolerant species Arabidopsis and So, but not in the freezing sensitive St (Table A3). Among them, only two were classified as transcription factors: cycling DOF factor 1 and 3 (CDF1, CDF3, both in the same pOG); and Short Hypocotyl 2 (SHY2) also called lAA3. Additionally, two genes were classified as regulators of transcription: BTB AND TAZ domain protein 4 (8T4), and a gene involved in RNA modification, CCR4 Associated Factor 1a (CAF1a). The array results showed that these genes were not cold-induced in St. In order to confirm this result, the expression of two of them was tested by quantitative real time PCR (qRT-PCR) (Fig 2.6). The CDF3-like and CAF1a-like genes are cold-induced in 38 both St and So, with similar kinetics of expression in both species. Therefore these genes are false negatives in the microarray for St. These results expand the list of conserved cold-induced TFs in the three species. 39 ScCDFa-Ilke c C .2 ,2 2.0 - In a g g 1.5 « g g 1.0 g g 0.5 ~ ta ta 3 '8 0.0 1 9‘ 2 8 24 72 168 0‘ hours of treatment ScCAF1-Ilke g 1.2 4 § 1.2 '1 '3 1.0 1 g 1.0 ~ 0’ g 0.8 « g 0.8 g 06 4 3 0.6 g 0.4 . g 0.4 E 0.2 4 ii 3 0.2 do 00 . I 4.1 .T iii, a! . fi,l I 0.0 2 8 24 72 168 hours of treatment 1 l 1 StCDF3-Ilke 8 StCAF1-llke 1 ii , his, -. 8 24 72 hours of treatment 24 hours of treatment 72 168 I... 168 Fig 2.6: Transcript accumulation of CAF1a-Iike and CDF3-Iike in S. commersonii (Sc) and S. tuberosum (St). Sc and St wild type plants were grown for 3 weeks under a 16:8h L:D photoperiod. Eight hours after dawn, plants were either transferred to 2°C (black) or kept at 25°C (grey). Tissue was collected at the different times shown. qRT-PCR analysis was used to determine the transcript levels of So and St genes. Average values of three different experiments are shown. Relative expression levels of each transcript were normalized using the potato 60$ gene (clone STMCK67) as an internal reference. Relative expression of the 2h cold sample was set to 1. Error bars indicate SE. Cold samples were significantly different from warm samples (ANOVA, p<0.0001, n=3). 4O In Arabidopsis, there are a number of transcription factors that are quickly induced in response to low temperature in addition to CBF1-3. The potato array does not represent the whole genome, nor the whole set of PUTs known to date, and does not include orthologs of known rapidly cold-induced transcription factors including Myb73, CZF1, ZAT10, ZAT12, and RAV1 (20). Based on the pOG list, PUT sequences of the putative potato orthologous genes were selected, and primers were designed for qRT-PCR based on those St sequences. The results are shown in Figure 2.7. The expression levels for all gene transcripts were compared between Sc and St, given that the same primers were used for both species. If some differences in hybridization efficiency of the primers were to happen, the primers could have had less homology to the Sc genes (the Sc genes are unknown) and that could result in less hybridization efficiency, but in almost all the cases the Sc gene transcript was higher than the St, suggesting that primers hybridized to Sc transcripts as well as to St transcripts. In the pOG list, only one potato PUT was identified as a putative ortholog of AtMyb73 (PUT_69025). The ScMyb73-Iike transcript accumulates in the cold; the cold samples are significantly different from the warm sample (ANOVA, p=0.005, n = 3). However, the StMyb73-like transcript accumulation by cold was not statistically significant (ANOVA, p = 0.2, n = 3) (Fig 2.7a). Additionally, the Sc transcript accumulates to a higher level than the St transcript. The ScMyb73-like transcript accumulation kinetics is similar to the one previously observed in Arabidopsis (20). 41 Three CZF1 putative potato orthologs were identified (PUT12601, PUT25814, and PUT25815). Given the high identity between PUT25814 and 25815, primers that target both sequences were tested by qRT-PCR (Fig 2.70). The transcript accumulation of these genes in response to cold was significant in both So and St (ANOVA, p<0.0001, n = 3), and the transcript levels were very similar in both species. The Sc and St transcripts had kinetics very similar to those observed for the Arabidopsis CZF1 gene (20). The other CZF1-like gene was PUT12601. The transcript accumulation of PUT12601 by cold was significant in St (ANOVA, p = 0.02, n = 3), but not in So (ANOVA, p = 0.3, n = 3) (Fig 2.7d). The St transcript accumulates to a higher level than the Sc transcript. The kinetics of PUT12601, however, are different to those previously observed for At CZF 1. Four potato EST contigs were identified as putative orthologs of AtZAT10. PUT22120, PUT22122, PUT32825, and PUT45213. Two of them (PUT32825 and PUT45213) were highly identical, so primers that target both sequences were used in qRT-PCR (Fig 2.7e). Both Sc and St transcript were highly cold induced at 2h reaching similar levels in both species. Cold samples were significantly different from the warm sample in both So and St (ANOVA, p< 0.0001, n = 3). Their kinetics of transcript accumulation were similar to those observed for Arabidopsis ZAT10 (20). The other two PUTs (PUT22122 and PUT 22120) were highly identical, so primers that target both sequences were used (Fig 2.7f). These genes are significantly induced by cold in So (ANOVA, 42 p= 0.005, n = 3) but not in St (ANOVA, p = 0.3, n = 3). These transcripts have different kinetics t6 the AtZA T10. Only one EST contig was identified as a putative ortholog of AtZAT12: PUT68089. The expression of this gene, even though very low, was detected in St (Fig 2.7b), but it could not be detected in So. The gene was significantly induced by cold (ANOVA, p=0.02, n = 3). The expression kinetics of St transcript was similar to the one observed for AtZA T12, but in the latter case the expression goes down at 24h (Vogel et al, 2005). Two EST contigs were identified as putative potato orthologs of AtRA V1: PUT3404 and PUT3405. Given their high identity, primers that target both sequences were designed (Fig 2.7h). Transcripts for these genes accumulate in response to cold in both So and St; in both cases cold samples were significantly different than the warm sample (ANOVA, p<0.0001 for So and p = 0.05 for St, n = 3). However, the Sc transcript reached higher levels compared to the St transcript. Their kinetic pattern is similar to that of AtRAV1 (Vogel et aL,2005) The transcript accumulation of ScCBF1 and StCBF1 is also shown (Fig 2.79). The CBF genes (5 in St and 4 in So) are highly similar, therefore primers that primarily, but not exclusively, amplify ScCBF1 and StCBF1 were used. Both genes are highly cold-induced at 2h, however ScCBF1 reached higher levels than StCBF1. It can be concluded that most of the putative potato orthologs to the Arabidopsis early cold-induced transcription factors, are cold-induced too in 43 both Solanum species. Only Myb73-like transcript was significantly cold induced in So but not in St. However, it appears that some of these gene transcripts reach higher levels of accumulation in Sc than in St, suggesting a quantitative rather than qualitative difference in their cold response between Sc and St. Together these results expand the list of conserved cold-induced transcription factors in So, St and Arabidopsis. 44 a Myb73-Iike b ZAT12-like (PUT59025) (PUT68089) 5 3.0 4 . ' 2.0 ' 2 . . 1.0 w2 2 8 24 72168 W2 2 8 24 72168 C CZF1-like d CZF1-like (PUT25814/25815) (PUT12601) 1.5 6 1 , 4 ' , C 0.5 , 2 . . % oahflllh Gaul-Illa! U) 9 W2 2 8 24 72168 w2 2 3 24 72168 8 .2 e ZAT10-like f ZAT10-like % (PUT32825/45213) (PUT22120/22122) n: 1.5 10 1.0 5 a 0.5 . :- 00 ...El‘-h-- ._ 0 -_I_LHEL W2 2 8 24 72168 W2 2 8 24 72168 g CBF1 h RAV1-like 1.5 (PUT3404/3405) 10 1.5 05 g 1.0 E- - 1. , 0.5 ' 0.0 -I'l'---—--— 0.0 ——- LL”... w2 2 8 24 72168 W2 2 8 24 72168 Hours of cold treatment Fig 2.7: Transcript accumulation in response to low temperature of Myb73-like, ZAT10-like, CZF1-like, ZAT12-like, RAV1-like, and CBF1 Solanum genes. Sc and St wild type plants were grown for 3 weeks under 16:8h L:D photoperiod. Eight hours after dawn, plants were either transferred to 2°C for 2, 8, 24, 72 and 168h or kept at 25°C for the same periods of time. Since, for each gene, all warm samples reached same levels, only samples kept at 25°C for 2h are shown as control (W2). qRT—PCR analysis was performed to determine the transcript levels of Sc and St genes. Average values of three different experiments are shown. Relative expression level of each transcript was normalized using the potato 6OS gene (clone STMCK67) as an internal reference. Relative expression of the Sc 2h cold sample was set to 1. Error bars indicate SE. 45 S. commersonii and potato CBF regulons CBF overexpression is sufficient to increase freezing tolerance in Arabidopsis (26,28-30). In order to explore if differences in the CBF regulon of So and St could account for their differences in freezing tolerance, the next objective was to determine how much of the cold transcriptomes are regulated by CBF in these species, and how do they compare to each other. To identify the CBF regulons of these two Solanum species microarray hybridizations were conducted, using Sc and St transgenic lines expressing AtCBF3 under the constitutive CaMV35S promoter (obtained from Tony Chen’s laboratory). These lines were first tested for AtCBF3 transgene expression (Fig 2.8). The Sc transgenic lines expressed higher levels of AtCBF3 transgene than the St lines did. RNA from 35S::AtCBF3 Sc and St transgenic lines was hybridized to the potato cDNA array (see methods). Sc and St WT plants were used as reference samples in each case. Figure 2.9a shows expression profiles as average log ratio of the three 358::AtCBF3 Sc lines (So ox 3) and two 35S::AtCBF3 St lines (St ox 2). The data indicated that AtCBF3 overexpression produced many changes in gene expression in both So and St, and only a small number of those changes (highly up or down regulated) are similar in both species. There were some clusters where there was induction by CBF overexpression in both So and St and also some clusters where there was repression by CBF overexpression in both So and St. However there were some clusters where there was either induction or 46 repression by CBF overexpression only in one of the two species. Interestingly, there were clusters where there was induction in one species and repression in the other one. The next step was to determine which ESTs were not only regulated by AtCBF3 overexpression but also by cold. For that, the expression profiles from Sc and St WT cold treated plants at different time points were compared to that of 35S::AtCBF3 Sc and St lines (Fig 2.9b). Many of the up or down regulated ESTs in transgenic lines were also up or down regulated by cold at 168h to similar levels. This result suggests that CBF is responsible mainly for the induction or repression of ESTs late in the cold. There was a cluster (F) where the majority of the genes were induced by CBF overexpression in both So and St (Sc and St lines) and also induced by cold at 24h and 168h in both So and St. Another cluster of ESTs (H) from Figure 2.9b was identified as having the majority of its ESTs repressed in So and St lines and also repressed by cold at 24h and 168h. There were a few clusters where there was high activation by CBF overexpression in both So and St, but not in the cold treated WT samples (Fig 2.9b). This could be explained by the fact that CBF overexpression produces a stunted phenotype in Arabidopsis (28); therefore it is possible that CBF is altering expression of some developmentally regulated genes that are not necessarily affected by cold. The cluster with the highest number of ESTs up regulated by cold at 2h in So and St and 24h in St (Cluster G) was not up regulated by CBF 47 overexpression (or just induced to a low level) either in So or in St. It is reasonable to assume that if CBF itself reaches its peak of expression at around 2h of low temperature (3,38), the ESTs being induced by cold at 2h will not be regulated by CBF and, more likely, CBF target ESTs will be in the second wave of induction. These data suggest that other TFs are responsible for the early cold-induction of ESTs in Sc and St. 48 144 i c -l g 1-2 c c ,1, AtCBF3 a 1.0 '. 1155-41! I $08—1 ?3 iii *5 a: . I l . g‘ .. l" 7.. 0.6 . ' I. :3 ‘. .2 . .. x - a , b b u .- _~ ,‘ g 73 0'4 i i3 '5 ? PI“? ‘51": It -« ~‘ is i % f : ‘ = 0'2 a l i - if 1 a i . 0.0 a“. ___.Hw___rw_ 11.1. ".2131 . ,. 1211-. 16 -,_ basalts T I LEELJB,.22T0-.-,ii:-....‘-.111_ .1 SCWT 8021 802 $023 StWT St28 8135 Fig 2.8: AtCBF3 transgene accumulation in St and Sc 358::AtCBF3 transgenic lines. The AtCBF3 transgene accumulation was tested in St and So 35S::AtCBF3 transgenic lines in warm conditions. qRT—PCR was performed using 100ng of RNA for each sample. Relative expression was calculated using the potato 608 gene (clone STMCK67) as an internal reference. ScWT, S. commersonii wild type; StWT, S. tuberosum wild type. Relative expression of So 21 was set to 1. The letters a, b and 0 indicate statistically significant differences (ANOVA, p<0.05, n=2). 49 Stox2 Scox3 5 | Cluster F .Q ‘01 0: N 0 I Cluster G 01 o _l Cluster H -7 Fig 2.9: Hierarchical clustering and expression profiles of Sc and St EST clones at 2, 24 and 168h of cold treatment (2°C), in combination with 35S::AtCBF3 Sc lines and 35S::AtCBF3 St lines. (a) 35S::AtCBF3 Sc lines (Sc ox 3) and 35S::AtCBF3 St lines (St ox 2)". (b) Same as (a) plus WT cold treated. Data shown as average log ratio (lines/wild type) of 3 Sc lines and 2 St lines, and 3 biological replicates for each time of cold treatment. The figure shows all spots on the array (including bad flagged spots) prior to statistical selection. Sc 02h, Sc 024h, and So C168h correspond to Sc Wl' cold treated for 2h, 24h and 168h respectively. St C2h, St 024h, and St C168h correspond to St Wl' cold treated for 2h, 24h and 168h respectively. 50 To identify ESTs that were likely to be part of the CBF regulon in Sc and St the data were selected using the following criteria. The Sc CBF regulon was defined by ESTs being DE in the three 35S::AtCBF3 Sc transgenic lines compared to WT and also being DE by cold in the three biological replicates at any time point by 2FC and p<0.05 (Limma package, (72)). Similarly, the St CBF regulon was defined by ESTs being DE in the two 35S::AtCBF3 St transgenic lines compared to WT and also being DE by cold in the three biological replicates at any time point. The criteria used were 2FC and p<0.05; however, since we had only two St transgenic lines, these p-values were generated by Limma using the average of these two transgenic lines as a third replicate. Fifteen and thirty percent of the cold regulated ESTs in So and St, respectively, can be assigned to their CBF regulons under the criteria used (p<0.05, 2FC). The CBF regulon lists can be accessed at mtg/wwwprl.mStfigg/Facultvpages/NSF MFT Site/datahtml. A total of 160 ESTs in Sc and 170 ESTs in St were members of the CBF regulon of induced ESTs (Fig 2.10). The overlap between the Sc and St CBF regulons of induced EST is not very large (around 30%). A total of 137 ESTs in So and 364 ESTs in St were members of the CBF regulon of down regulated ESTs (Fig 2.10). There is a large number of ESTs that are members of the St CBF regulon of repressed ESTs that apparently are not members of the Sc CBF regulon. Together these data indicate that among the most likely CBF regulon members defined by the 2FC and p<0.05 criterion, the overlap between Sc and 51 St ESTs is not very big. These differences between Sc and St CBF regulons could account for the differences in their freezing tolerance. 52 S. commersonii S. tuberosum CBF regulon: induced ESTs CBF regulon: repressed ESTs Fig 2.10: S. commersonii and S. tuberosum CBF regulons. CBF regulated genes were selected as being DE in 35S::AtCBF3 transgenic lines compared to WT and also DE in 3 WT cold treated biological replicates (all times cold considered) with p<0.05 and 2FC cutoff. 53 Comparison between Sc, St and Arabidopsis CBF regulons Given that CBF pathway plays a predominant role in freezing tolerance in Arabidopsis, it is important to know to which extent the CBF regulons are conserved in different plant species. To address this issue, the Sc and St CBF regulons were compared to that of Arabidopsis. For this analysis, the Arabidopsis-potato pOG list was used. From the 54 ESTs that were identified as CBF regulon ESTs that were up regulated in both So and St, 28 ESTs (52%) were represented in the list of 8,714 putative orthologous groups (pOGs) and correspond to 22 pOGs (Fig 2.11a). From the 91 ESTs that were identified as down regulated CBF regulon ESTs in both So and St, 38 ESTs (42%) were represented in the list of pOGs between Arabidopsis and St and correspond to 31 pOGs (Fig 2.11a). These results indicate that a large percentage of the CBF regulon ESTs that are conserved in So and St do not have Arabidopsis orthologs based on the criteria used in this study. The CBF regulons of Sc and St were compared to that of Arabidopsis. The Arabidopsis CBF regulon used in this comparative study corresponds to the group of genes that were DE by 2FC cut off (p<0.05) in 358::CBF2 Arabidopsis transgenic lines (20) and also DE in the Arabidopsis cold data sets described above. Figure 2110 shows the overlap of CBF regulon up regulated genes between the three species. Seven pOGs were identified that have at least one CBF regulon up regulated member from each species. The Arabidopsis 54 members of these seven pOGs are: early light-induced protein 2 (ELIP2); LEA14; ERD10; COR47; ADOF1; responsive to desiccation 26 (R026); sucrose synthase 1 (SUS1); and invertase/pectin methylesterase inhibitor family protein. Among the CBF regulon activated pOGs, only 15% of the pOGs with CBF regulon genes in Sc and 14% of the pOGs with St CBF regulon genes also have Arabidopsis CBF regulon members. Given that the Arabidopsis array represents almost its entire genome, these Solanum CBF regulon members that are not members of Arabidopsis CBF regulon represent a real difference between these CBF regulons, suggesting differences in the evolution of CBF upregulated genes. Figure 2.11d shows the overlap of CBF regulon down regulated genes between the three species. Two pOGs were identified that have at least one CBF regulon down regulated member from each species. The Arabidopsis members of these 2 pOGs are a phosphoethanolamine N-methyltransferase; and two beta galactosidases (BGAL1 and BGAL4). Among the CBF regulon repressed pOGs, only 4% and 2% of the pOGs with CBF regulon members from So and St respectively also have Arabidopsis CBF regulon members. With these results it was concluded that there is more conservation between the CBF regulons of the two Solanum species than between the two freezing tolerant species (Sc and Arabidopsis). Despite these differences, the results reported here are minimal estimates of conservation, given that the potato array used does not represent the complete potato genome; therefore the number of CBF regulon genes 55 conserved between these species could be an underestimate. As was determined for the cold transcriptomes, CBF-regulon pOGs that are present in only one of the two arrays (either Arabidopsis or potato) were also identified, these therefore escape this conservation analysis (Fig 2.11b). Of the approximately 5,316 pOGs present on the Arabidopsis array, but not in the potato array, many had members of the Arabidopsis CBF regulon. There were 69 pOGs with CBF regulon up regulated members and 15 pOGs with CBF regulon down regulated members (58% of the total pOGs with Arabidopsis CBF regulon members). These pOGs cannot be compared to those of the Solanum species because the putative St orthologs are not represented in the potato array. Given this, the number of CBF regulon genes that are conserved between the two Solanum species and Arabidopsis could be largely underestimated in these experiments. From these results it was concluded that the overlap identified between CBF regulons from Arabidopsis and the two Solanum species is small. This suggests that CBF is regulating different genes in Arabidopsis and in the two Solanum species, suggesting a significant difference in the evolution of these CBF regulons. 56 a. CBF regulon up CBF regulon down 28 EST (52%) 38 EST ()42% 1 8,714 pOG i between At and St ._i__ both arrays Used for transcriptoms comparison c_ CBF regulon up // \‘d, CBF regulon down At At 42 pOGs 19 pOGs Sc St Sc St 48 pOGs 50 pOGs 42 pOGs 110 pOGs Fig 2.11: Comparison of CBF regulons among S. commersonii (Sc), S. tuberosum (St) and Arabidopsis (At) based on putative orthologous groups (pOGs). (a) CBF regulon up and CBF regulon down ESTs in So and St, showing the percentage that have Arabidopsis putative orthologous genes. (b) Putative orthologous groups (pOGs) distribution on the Arabidopsis ATH1 Chip and potato cDNA array. The venn diagram shows pOGs with at least one Arabidopsis gene on the ATH1 chip (green), at least one potato clone on the potato cDNA array (brown) or at least one gene from each species in both arrays. Among the pOGs that are present only in one of the two arrays, the number of CBF regulon induced or CBF regulon repressed pOGs in the corresponding species is shown. (c) Overlaps of CBF regulon induced or (d) CBF regulon repressed pOGs from pOGs present in both arrays. 57 DISCUSSION To date, there is no explanation for the differences in freezing tolerance between Sc and St. It is known that the CBF pathway is the only known pathway to play a predominant role in freezing tolerance in Arabidopsis. Therefore, the goal of this study was to identify and compare the low temperature transcriptomes and CBF regulons of two closely related Solanum species that have different levels of freezing tolerance. The model used was two closely related potato species, 8. tuberosum (common potato, tetraploid) that is freezing sensitive and does not cold acclimate, and its wild relative S. commersonii (diploid) that is freezing tolerant and can cold acclimate. The low temperature transcriptomes of these two species were identified. In both, hundreds of ESTs were cold-induced and cold-repressed. A global view of the cold regulated ESTs in the two Solanum species showed that the changes in gene expression are very similar between Sc and St (Fig 2.1). The results suggest that these two Solanum species share between 50 to 70% of their cold- regulated ESTs. The ESTs regulated in both species might be involved in tolerance to chilling temperatures or may be part of the general response to stress damage. From the transcriptome analysis done in this study, there is nothing immediately obvious to conclude why Sc and St have differences in their freezing tolerance. Presumably the specific differences in the kinetics of gene cold induction or quantitative differences in the low temperature transcriptomes 58 of So and St could account for their differences in freezing tolerance, but there is not enough evidence at this stage. There is a group of transcription factors that are candidates to configure the low temperature transcriptome in Arabidopsis. The potato orthologs of these genes were not represented in the potato array, therefore they were tested for cold induction in So and St by qRT- PCR. Results revealed that most of these genes were cold induced in both So and St, moreover with similar kinetics than the Arabidopsis orthologs (Fig 2.7). However, in some cases the Sc genes were induced at a higher level compared to the St genes. The role of these transcription factors in freezing tolerance it is not known, then whether these quantitative differences are responsible for the differences in freezing tolerance between Sc and St is yet to be discovered. In the future it would be interesting to study the role of these conserved transcription factors in freezing tolerance. It is also possible that the freezing tolerance difference between Sc and St comes from differences at a post transcriptional level, possibly due to differences in protein levels, protein modifications, or metabolite levels. Another group (76) has previously studied other potato species under cold stress: 8. phureja CHS (diploid), S. tuberosum cv. Desiree (tetraploid) and S. tuberosum PS3 (dihaploid). Electrolyte leakage experiments demonstrated that S. phureja CHS was the most cold tolerant at a constitutive level (LT50 of - 9.6 under non acclimating conditions and -11 after 3 weeks of cold acclimation). Transcriptome analysis at one and three days of cold treatment revealed significant changes in expression of genes related'to amino acid metabolism 59 and carbohydrate metabolism in the three potato species. There are no immediately obvious differences between these three species transcriptomes. However, they found that S. phureja CHS had higher constitutive levels and higher accumulation upon cold exposure of protective sugars such as sucrose, galactose, trehalose, galactinol, raffinose, and glucose. Another study (77) also looked at S. tuberosum transcriptomes under cold stress. This study found that cold stress resulted in a large number of DE ESTs (2,318, p<0.01) (77). Using the same cut off criteria they used, 1,128 ESTs (50%) can be identified as cold regulated in St in this study. The differences observed between Rensink et al and this study could be due to differences in the growth and experimental conditions. Rensink et al grew plantlets on magenta boxes, and their cold treatments were done by transferring the magenta boxes to ice just after dawn. In our study we transferred the plants to cold 8h after dawn. It is suspected that some potato genes might be under circadian regulation, as is the case with Arabidopsis CBFs (40) and tomato CBF1 (43); therefore the results could be different depending on the time in the photoperiod when the plants were transferred to cold. They collected leaf and root samples at different time points, and they also mentioned that their cold treated plants collapsed after 3h of cold treatment, and later they recover. This could have had an effect on gene expression. Given that whole genome duplication and gene duplication and retention following duplication has been very extensive in plants (78), orthologous genes are not only in a one-to-one relationship, but rather organized in groups of 60 orthology. In this study a list of groups of putative orthologous genes between Arabidopsis and potato were used to compare their low temperature transcriptomes. First, many of the ESTs that were cold-regulated in both So and St did not have Arabidopsis orthologous genes according to the criteria used (Fig 2.5a). This could indicate an important difference between the cold responses of the two Solanum species and Arabidopsis. Thirteen percent (1,466) of the ESTs present in the potato array have not been assigned to a PUT, and therefore were not compared to Arabidopsis proteins (Fig 2.4). Given that this is a small percentage, it is not believed that having more sequence information for these genes would affect largely the results. Second, the cold regulated list of genes in the two Solarium species could be underestimated, given that many Arabidopsis genes that are cold regulated, have potato orthologs that are not present in the potato array (Fig 2.5b). Therefore, the overlap between the two Solanum species and Arabidopsis could be larger (FigZ.5c and d). This is supported by the qRT-PCR results of known early cold induced transcription factors in Arabidopsis whose potato orthologs were not represented in the potato array. The expression of these potato orthologs was also induced in both So and St (Fig 2.7). However, given that the number of cold regulated genes in Arabidopsis likely reflects reality (the array represent almost its entire genome), then the number of pOGs that are only cold-regulated in the two Solanum species, but not in Arabidopsis (Fig 2.50 and d) could be the same or even larger. Therefore this result indicates another level of difference between Solanum and Arabidopsis low 61 temperature transcriptomes. This suggests significant differences in the evolution of these low temperature responses, a result that is not surprising considering the evolutionary distance between Arabidopsis and potato (around 125 Mya) (79). To date, there is no large scale comparison of low temperature transcriptomes between distantly related species. Only one study has been published comparing stress regulated genes (including cold stress) of rice and Arabidopsis (80). By a combination of rice cDNA microarray (1,700 rice cDNAs) and northern analysis, they identified 73 genes as stress inducible in rice, 36 of which were induced by cold. Fifty percent of these stress inducible rice genes were identified as having similar functions or gene names with already reported Arabidopsis stress inducible genes. Therefore this study is the first large scale comparison of low temperature transcriptomes between distant related species such as potato and Arabidopsis. The CBF regulons of So and St were identified in this study. With the criteria used (2FC and p<0.05), the overlap between the Sc and St CBF regulons is not very big (Fig 2.10). This suggests that CBF turns on different genes in So and St. When these CBF regulons are compared to that of Arabidopsis, there is more conservation between the two Solanum CBF regulons than to that of Arabidopsis. Many genes that are members of CBF regulon in the two Solarium species are not in Arabidopsis. Considering that the Arabidopsis array represents almost its entire genome, these genes likely represent a real difference between these CBF regulons. This result suggests 62 that there has been divergence in the evolution of CBF regulons. It could be that CBF is binding to different sites in these different species or there could have been loss of cis-acting DNA regulatory sequences in some genes of one species, therefore CBF no longer binds to the same targets. Examples of cis-acting regulatory mutations that cause evolutionary changes have been observed in different species (81). In yeast for example, it has been shown that transcription factor binding sites are evolving at a fast rate and could be the major cause of divergence between related species. The transcription factor binding sites for the two pseudohyphal regulators Ste12 and Tec1 were studied in three species of Saccharomyces and it was found that these binding sites have diverged across species. Most target genes were bound in only one or two of the three Saccharomyces species studied. This group identified many examples where a species-specific loss of binding and/or loss of cis-acting sequence had occurred (82). A previous study has shown conservation between the Populus and Arabidopsis CBF regulon (50). They searched the Poplar cDNA array (POP1 13K) for homologs of previously identified CBF responsive genes in Arabidopsis. They found that 12 (32%) of the 38 CBF3 regulon up-regulated members in Arabidopsis (83) had orthologous in the poplar array, and of those 12, 7 were up regulated in their AtCBF1 overexpressing poplar lines. They do not mention if those 7 genes were also cold regulated. They identified 22 Populus genes as being up regulated by cold and by AtCBF1 overexpression. Given that only 7 genes that had Arabidopsis orthologs were also regulated by 63 AtCBF1 overexpression in Populus, the majority of these 22 Populus CBF regulon genes either did not have Arabidopsis orthologs or were not members of the Arabidopsis CBF3 regulon, indicating important differences between Arabidopsis and Populus CBF regulons. In their CBF regulon comparison, they are using a very limited Arabidopsis CBF regulon list of genes. The Arabidopsis CBF3 regulon from Maruyama et al (2004) only surveyed 8,000 genes, and found 38 genes cold up regulated and CBF3-upregulated. In the present study, the Arabidopsis CBF regulon used was much larger set of genes (169 up- and 58 down-regulated genes). The CBF regulon consisted of genes being regulated by CBF2 overexpression (20) and by cold in different experiments (Atgeneexpress) using the ATH1 Affymetrix array that represent around 22K genes (almost its entire genome). Benedict, Skinner et al. 2006 also mention that their Populus CBF regulon had greater disagreement with the AtCBF2 regulon identified in Vogel, Zarka et al. 2005. It has been reported that overexpression of AtCBF1, AtCBF2 or AtCBF3 regulates very similar sets of genes in Arabidopsis suggesting that there are no CBF-specific regulon differences (26). Therefore, the differences between Populus and the Arabidopsis CBF2 regulon (20) suggest differences between species. All this suggests that the Populus and Arabidopsis CBF regulons are not as strongly conserved as they claimed. Preliminary studies suggest that overexpression of AtCBF3 in So did not increased freezing tolerance of So lines in non acclimating conditions. However, after a period of cold acclimation, the freezing tolerance of 358::AtCBF3 Sc 64 transgenic lines increases 4°C compared to WT cold acclimated plants (unpublished data). This suggests that AtCBF3 alone is not sufficient to impart freezing tolerance in So, but it may act in concert with other factors that are present in the cold. Overexpression of AtCBF3 in St has been reported to have a small increase (about 2°C) in freezing tolerance under non-cold acclimating conditions (59), suggesting that CBF could be sufficient to impart freezing tolerance in St. These experiments need to be repeated in parallel to have a better understanding of the effect of AtCBF3 overexpression in So and St. Pino, et al. 2008 reported that AtCBF1 overexpression in So increases freezing tolerance in non acclimating conditions by about 4°C, suggesting that AtCBF1 is sufficient to increase freezing tolerance in So; and overexpression of AtCBF1 in St had a small increase in freezing tolerance (about 2°C) in non acclimating conditions. In the future it would be interesting to test by microarrays which genes are differentially expressed by AtCBF1 overexpression. It would be also interesting to know if endogenous overexpression of ScCBF1 and StCBF1 in So and St respectively, had any effect on freezing tolerance. In some species there is genetic evidence for the role of CBF in freezing tolerance. In wheat, a cluster of eleven CBF genes have been mapped to the Frost resistance-2 (Fr-Am2) locus (84). This locus was mapped at the peak of two overlapping quantitative trait loci (QTL), one for frost survival and the other for differential expression of the cold regulated gene COR14b. Similarly, in barley, a cluster of six HvCBFs genes mapped to the Fr-H2 cold tolerance QTL (85). These are evidence that support the important role of CBFs in freezing 65 tolerance. However, to date, there is no genetic evidence that proves CBFs to be in a loci associated with freezing tolerance in Solanum species. There is genetic evidence that suggests that there is independent genetic control of non- acclimated freezing tolerance (NAFT) and cold acclimation capacity (CAC) in Solanum species (86). Two wild diploid Solarium species, 8. commersonii (freezing tolerant and able to cold acclimate) and S. cardiophyllum (freezing sensitive and unable to cold acclimate) were crossed. By analysis of the two segregating backcross populations, it was observed that the NAFT and CAC were not correlated in any of the two backcross segregating populations. Two QTLs for NAFT and two for CAC have been identified. The QTLs for NAFT and CAC were found at separate genomic regions (87). The genes in these QTL responsible for these two traits have not been identified yet. Future experiments will be crucial to understand the role of CBF in freezing tolerance in Solanum species. As this thesis was written, the potato whole genome became available (nt_tp://www.potatogenome.net/index.pho/Main Page). With this, many more future studies can be done to comprehend better the differences between the Sc and St cold transcriptomes. For instance, a more rigorous analysis of the complete cold transcriptomes from So and St can be done using deep sequencing techniques such as Illumina or 454 sequencing. Additionally, the promoters of the cold regulated ESTs identified in this study can be searched for enriched motifs using different bioinforrnatic tools. Once identified, these potential motifs can be tested in vivo for cold regulation. Cold-responsive motifs 66 from Sc and St could be fused to a reporter gene and transformed into Arabidopsis to test the conservation of their cold-response. 67 MATERIALS AND METHODS Plant growth, cold treatments and RNA extraction Solanum commersonii and Solanum tuberosum cv. Umatilla wild type plants were maintained under greenhouse conditions (Pino et al. 2007). Wild type plants and transgenic lines used in experiments were transferred to a Percival model MB6OB growth chamber (Percival Scientific, Inc, Perry, IA) under a 16h photoperiod, 350 pmol rn'2 s‘1 PAR at 25°C. Three biological replicates for each. wild type species were grown for 3 weeks under these conditions (each biological replicate consisted on three plants). Eight hours after dawn, wild type plants were transferred to an environmentally controlled cold room maintained at 2°C, under a 16h photoperiod with 50 umol m'2 s'1 light intensity and leaf tissue was harvested after 2, 8, 24, 72, and 168 h. Warm controls were maintained at 25°C under normal growth photoperiod and tissue was harvested at 2, 8, and 24h in the light for use as reference control samples. In the case of the 72 and 168h cold samples the 24h warm control was used as reference since 72 and 168h warm plants were already flowering. The leaf tissue of transgenic lines was collected eight hours after dawn. Total RNA was isolated from leaf tissue using RNeasy Plant Mini Kits (Qiagen, Valencia, CA). For real time PCR experiments, samples were treated with RNAse-free DNAse (Qiagen, Valencia, CA) using the on-column DNAse digestion method provided by the manufacturer. 68 RNA labeling and hybridization of potato microarrays cDNA microarray experiments were conducted using the 10K potato cDNA microarray (TIGR, h_ttp://www.icvi.org[potato/sol ma microarravsshtml). 20 pg of RNA were labeled by the indirect labeling aminoallyl method. The slides were hybridized using the indirectly labeled aminoallyl probes hybridization method (hltpzllwwchvi.orgjpotato/sol ma fltocolsshtml). To avoid bias due to dye-related differences, labeling dyes for each sample pair (cold/warm or transgenic line/wild type) were swapped in one of the three independent hybridizations (three biological replicates for cold treatments and two or three transgenic lines for the CBF regulon experiment). Data processing and analysis The TIFF images were quantified using Genepix 3.0 (Axon Instruments, Union City, CA). The software automatically flags spots that cannot be found in one of the channels. Spots with aberrant shapes were checked manually and flagged as bad. Spots with lower signal intensity than the background (spots with >=55% of the pixels with lower signal intensities than background) were also flagged as bad. All these “bad” flagged clones were excluded from further analysis. The data were normalized using the print tip loess method in the Limma package (Smyth 2004). A list of differentially expressed (DE) clones were 69 ranked based on their p value, using a false discovery rate to correct for multiple testing. The results were selected by p<0.05 or else as indicated in the text. Average fold change (PC) was calculated for the two duplicates of each clone on the array. In cases where one of the duplicates didn’t pass the p value out off or the duplicate was flagged as a “bad quality spot", the value of the other duplicate was used. Hierarchical clustering was performed using Cluster software (88), using normalized log ratios. Gene Onthologies (GO) for St EST clones were obtained from TIGR. Quantitative real time PCR (qRT-PCR) RNA, using an amount that fell within the linear range for all genes tested (generally 100-250ng), was reverse transcribed using a reverse transcription system (Promega, Madison, WI) according to the manufacturer’s instructions. The 20pL final reaction was diluted to 200pL. A 1 pL aliquot of each cDNA was used in a real time PCR (qRT-PCR) reaction, with the addition of 0.4 pM of each primer and Fast SYBR Green master mix (Applied Biosystems, Foster City, CA) to make a final reaction volume of 10p.L. The qRT—PCR reactions were performed using a FAST 7500 Real Time PCR System (Applied Biosystems, Foster City, CA). The primers used are listed in Table 2.1. 70 Data were analyzed by ANOVA using SAS program 9.1 (SAS Institute Inc., Cary, NC) with mixed procedures; when appropriate, least significant difference was used for multiple comparisons. TABLE 2.1: Primers used in real time PCR. Name Sequence Target CL_RT_15 GGCCTTGTATAATCCCTGATGAATAAG At Ubiquitin 10 CL_RT_16 AAAGAGATAACAGGAACGGAAACATAGT At Ubiquitin 10 MC__122 TGCCCATAAACCCTTTI’TGT St 608 (clone STMCK67) MC_123 AACAATGGCGGCTAAGAAGA St 60s (clone STMCK67) MC_132 TGGGCTCATAATCTCGAATC St CAF1a (clone smooss) MC_133 GCTTGAAAACAACACCAGGAA St CAF1a (clone smooss) MC_201 GATCAGATCGAAACGACCTCGTA St ZAT10 (PUT32825/45213) MC_202 ATCTTCGGAAATAA‘I‘I'GGTTGTTGT St ZAT10 (13073232045213) MC_203 TGCGCGTGAC'ITI'GACCTAA St ZAT10 (PUT22122/22120) MC_204 AGTCAATGGTCAGATCCAATTGG St ZAT10 (PUT22122/22120) MC_205 GGTI'CCGAGGTCGATCTGGTA St Myb73 (PUT69025) MC_206 AAGCACGTATGCTCCACTTGAG St Myb73 (PUT69025) MC_207 CCCACCACAATTA‘I‘I’CAAACGA St ZAT12 (PUT68089) MC_208 GAAGGTGCTAGTAGTGGTGAATTGG St ZAT12 (PUT68089) MC_212 GCTTGATGCTTCTGCTGATGTG St CZF1 (PUT12601) MC_213 AGATCAGGTCAACAGCTCGTTTC St CZF1 (PUT12601) MC__214 TCACCCCACCTGCATTACAG St CZF1 (PUT25814/25815) MC_215 CCCGAGCGTTI’AGAGATGTCTT St CZF1 (PUT25814/25815) MC_216 TTAGTCTGGAAAATGACTTGTGATTCA St CDF3 (clone STMCY67) MC_217 GAGACGACCACCGGAAGTATCT St CDF3 (clone STMCY67) MC_323 GG‘l‘l'TGGTl'AGGCACATTl’AACG St RAV1 (PUT3404/3405) MC__324 GGCCGCGACGTCGTAA St RAV1 (PUT3404/3405) SJG_80 TTCCGTCCGTACAGTGGAAT At CBF3 SJG_81 AACTCCATAACGATACGTCGTC At CBF3 Inference of putative orthologous groups (pOGs) A list of pOGs between the putative unique transcripts (PUT) sequences assembled from Expressed Sequence Tags of St (PlantGDB, http://www.plantgdb.org[, version 157a) and Arabidopsis protein sequences was obtained from Shinhan Shiu’s laboratory, Plant Biology Department, Michigan State University. 71 pOGs have been previously established using protein sequences from four plant species with complete genomes (Arabidopsis thaliana: TAIR6, Populus tn'chocarpa; v1.1, Oryza sativa japonica; version 2, Physcomitrella patens; version 1.1) (89). From all these 4 genomes, a best matching protein for PUT was identified by Blast, using only matches with E-values lower than 10’s. A PUT was assumed to be in the pOG of its best matching protein from these four species, only if the evolutionary distance between PUT and the rice member(s) and PUT and the Arabidopsis member(s) in the pOG is less than the distance between rice and the Arabidopsis members. The evolutionary distances of all sequences were calculated using the protdist program in the P_h)Llogeny Inference Package (PHYLIP) with Gamma correction (PHYLIP version 3.6, Felsenstein, J., 2005, distributed by the author, Department of Genome Sciences, University of Washington, Seattle). A simplified version of the process is shown in Fig 2.12. First, based on blast search, reciprocal best matches (double pointed arrows) or single match (one way arrow) are obtained between Arabidopsis proteins and potato translated PUTs. In the figure example based on the blast results, two possible scenarios for phylogenetic trees can be described. Then protdist calculated the phylogenetic distances (d) and based on those, phylogenetic trees are generated. All the genes subsequent to a speciation event (red dot) will belong to the same pOG. 72 PhylogeneticTree Evolution Blast _ Based on sequence distances Possible scenario 1: P1 1 P1 0 L POG 1 /, P1 . ‘ A1 ' . A1 lfd between _ _//P2 If: P2 _ A2 ' A2and A3 is < )(\ . P2 : thand to P2 “—_‘I'\ ‘\ A1 A3 A2 nos 2 \s A2 ; \ A3 A3 . Possible scenario 2: | . P1 ; pOG 1 A1 /P1 A2 I pOG 2 / P2 0 _ lfd between P lost P2 and A3 is gm ’ < than d to A2 / P2 -- \ . \ pOG 3 \ \ A3 \\ A2 ~ = best blast A3 reciprocal match I = speciation event d = phylogenetic distance Fig 2.12: Representation of putative orthologous groups (pOGs) identification. pOGs between Arabidopsis proteins (A) and potato translated PUTs (P). Pointed line, branch of the tree leading to a gene that has been lost in evolution (P lost). 73 CHAPTER THREE REGULATION OF CONSERVED TRANSCRIPTION FACTORS BY THE CIRCADIAN CLOCK, CYCLOHEXIMIDE TREATMENT AND MECHANICAL AGITATION INTRODUCTION CBFs are a group of AP2/ERF transcription factors that are quickly induced in response to low temperature (27). CBF1-3 transcription factors regulate the expression of more than a hundred genes in Arabidopsis known as the CBF regulon (19,20,83). Expression of these genes leads to an increase in freezing tolerance (26,28,29). Solanum tuberosum (St) (common potato) is a freezing sensitive species and it is unable to cold acclimate. Its wild relative, Solanum commersonii (So), is able to cold acclimate and has a moderate level of freezing tolerance. These species have conserved CBF genes. There are 4 So CBFs and 5 St CBFs genes; each species has two cold-induced CBFs (CBF1 and CBF4) (43). In Arabidopsis, there is a group of transcription factors that are early cold induced with similar kinetics to CBF1-3 genes (20). These genes are ZAT12, ZAT10, RAV1, CZF1 and MYB73. These transcription factors are likely candidates to configure the low temperature transcriptome of Arabidopsis. It has previously been shown that in addition to their low temperature response, 74 CBFs and these other factors are also induced in response to mechanical agitation and inhibition of protein synthesis (cycloheximide treatment) (20,27,38). Moreover, the CBF2 promoter and some of these transcription factors have two regulatory sequences lCEr1 and ICEr2 (Induction of CBF expression region) that are involved in gene induction by cold, mechanical agitation and cycloheximide treatment. It is thought that there is a regulatory link between these different responses, but it is yet to be discovered in Arabidopsis (38). In addition to these responses Arabidopsis CBF3 has also been shown to be regulated by the circadian clock (39). The Arabidopsis circadian clock is an internal oscillator that maintains an endogenous period of 24h. Its signaling networks enhance the plant’s responses to its rhythmic environment. Environmental signals can regulate the phase and period of the circadian clock. A consequence of circadian control is that stimuli of the same strength applied at different times of the day can result in responses of dlfferent intensities, phenomenon known as “gating” (42,90). Harmer et al., showed that under warm temperature, AtCBF3 transcripts undergo circadian cycling, with a peak at ZT4 (Zeitgeiber Time, hours after dawn) and a trough at ZT 16 (39). Given this circadian regulation of AtCBF3 under warm temperatures, the question was raised whether the circadian clock also gated the expression of CBF1—3 in response to low temperature. Fowler et al. showed that indeed it did (40). When Arabidopsis plants were entrained in 12:12 h photoperiod and then transferred to continuous light, the accumulation 75 of each CBF1- 3 after exposure to 4°C cycles depending on the time during the subjective day or night that the plants were transferred to cold. Furthermore, disruption of the circadian clock by overexpression of Circadian Clock Associated 1 (CCA1), a Myb-related transcription factor member of the Arabidopsis clock (41,42), also disrupted the AtCBFs’ cycling (40). In addition, two other Arabidopsis early cold-induced transcription factors were also gated by the clock: RAV1, which encodes an AP2/B3 domain transcription factor (91) and ZAT12, which encodes a zinc-finger domain transcription factor (92). RAV1 showed cycling with the same phase as CBF and ZAT12 had the opposite phase (40). Circadian gating of CBFs has also been suggested in tomato. When Solanum chopersicum (tomato) and its wild relative Solanum pimpinellifolium were entrained under a 16:8 L:D photoperiod, CBF1 expression cycled, reaching higher responsiveness during the light period in both species. When plants were shifted to constant dark or constant light CBF transcripts continued to cycle, however the peaks and troughs did not correspond to those observed in plants grown under a normal 16:8 L:D photoperiod (43). Given that the Arabidopsis early cold induced transcription factors are also cold induced in Sc and St, the main goal of the experiments described in this chapter was to test if there was conservation in the response of these genes to these other environmental cues or if there were differences in the regulation of these genes between Sc and St that could account for their differences in freezing tolerance. We wanted to test if these transcription factors 76 were also responsive to mechanical agitation and cycloheximide treatment in So or St. We also wanted to know if the cold induction of these genes was gated by the circadian clock in So or St. Four early cold-induced transcription factors with orthologs in all 3 species were selected: CBF1, ZAT10, RAV1, and CZF1. The expression of the Sc and St genes in response to mechanical agitation and cycloheximide (CHX) treatment was tested by qRT-PCR. In addition, circadian experiments were conducted to test if these transcription factors were cold gated by the circadian clock in So or St. ScCBF1 and ScZAT10 were induced by mechanical agitation in So, and the four genes tested were induced by cycloheximide treatment in So. In the case of St, only StZAT10 was responsive to mechanical agitation and only StCZF1 was responsive to cycloheximide treatment, but both St genes were induced to much lower levels than the ones observed in the Sc genes. The circadian experiments suggest that these genes are cold-gated by the circadian clock in Sc and St. Together these results suggest that the regulatory link between cold, mechanical agitation and cycloheximide treatment appears to be conserved in So, but is not so clear if there is conservation to these other treatments in St. The circadian clock gating of these transcription factors appears to be conserved in the two Solanum species. 77 RESULTS Regulation of conserved transcription factors by mechanical agitation and inhibition of protein synthesis. From the previous chapter, some transcription factors were identified to be conserved in their response to low temperature. Their transcripts accumulated early upon cold treatment in Arabidopsis, Sc and St. The question was raised whether these transcription factors were also conserved in their response to other environmental perturbations in Sc and St, or if there were differences in the regulation of these Sc and St genes. Four transcription factors that are early cold-induced in Arabidopsis as well as in So and St were selected: CBF, RAV1, ZAT10, and CZF1. The Arabidopsis genes are also induced by mechanical agitation and inhibition of protein synthesis (CHX treatment). In order to test if the St and So orthologs of these four transcription factors were also responsive to these treatments, they were tested for their response to mechanical agitation and CHX treatment. The putative potato orthologs with cold-induced kinetics closer to that observed for the Arabidopsis orthologs were chosen (See Fig 2.7, chapter 2). Primers that target the St PUT sequences (Putative Unique Transcript, PIantGDB website) were generated and used to detect St and So transcript by real time PCR (qRT- PCR). 78 Given the high similarity of some orthologs, primers that target both PUT3404 and 3405 (RAV1), both PUT32825 and 45213 (ZAT10), and both PUT25814 and 25815 (CZF1) were designed. The CBF genes (5 in St and 4 in So) are highly similar, therefore primers that primarily, but not exclusively, amplify ScCBF1 and StCBF1 were used (Table 2.1, chapter 2). To test mechanical agitation responsiveness, plants grown in magenta boxes were dropped approximately 6” every 2 seconds for 15 minutes, and then incubated with no agitation for 0, 15 or 30 minutes. Tissue was collected from 2 plants at each time point and the experiment was repeated once. Figure 3.1 shows results of mRNA accumulation in response to mechanical agitation. The four gene tested showed higher transcript levels in So than in St. Only ScCBF1 and ScZAT10 showed statistically significant (ANOVA, p<0.05, n=2) accumulation by mechanical agitation in Sc. ScCBF1 mRNA accumulation reached its peak just after the treatment has been stopped (0 minutes after mechanical agitation) (5 fold compared to no shake control) and then declined after 15 minutes (3 fold), going back to the levels of no shake control after 30 minutes of incubation with no agitation. The same kinetics was observed for ScZA T10, but the induction was only 3 fold. The only St gene that had statistically significant (ANOVA, p<0.05, n=2) transcript accumulation upon mechanical agitation treatment was StZA T10 (Fig 3.1). The kinetics of transcript accumulation of StZAT10 was similar to that of ScZAT10, reaching 3 fold at 0 minutes after treatment. All the other genes showed no statistically significant difference between samples. 79 To test responsiveness to inhibition of protein synthesis, three week old plants were treated with 10 ug/mL of CHX. Tissue was harvested after 0, 1, 2, 4, 8, and 24h of treatment (2 plants per time point and the experiment was repeated once). Fig 3.2 shows the results of mRNA accumulation under CHX treatment. All 4 gene transcripts had higher levels in So compared to St. The four Sc gene transcripts accumulated significantly (ANOVA, p<0.05, n=2) in response to CHX treatment. ScCZF1, ScZAT10, and ScRAV1 accumulated slowly after CHX treatment reaching their peak accumulation after 4 hours of treatment by 66, 72, and 7 FC respectively. ScCBF1, however, reached its peak (20 FC) quicker after 1 h of treatment, declining to 11 FC after 4h of treatment. All of them reached their normal (no treatment control) levels after 8 h of treatment. In the case of St genes, only StCZF1 showed statistically significant differences between CHX treatment and control samples (ANOVA, p<0.05, n = 2) (Fig 3.2). The induction of StCZF1 was lower (3FC at 4h) as compared to ScCZF1 (66FC at 4h). All the other genes showed no statistically significant differences between CHX treatment and the control. Together, the results of this section indicate that the four genes that are conserved in terms of their cold induction between Arabidopsis and the two Solanum species (CBF 1, ZAT10, RAV1, and CZF1), are also conserved in their response to CHX treatment in So. Only one of these four genes, StCZF1, was significantly induced by CHX treatment in St; however it accumulated to a smaller level compared to ScCZF 1. 80 Only two of the four cold-induced conserved transcription factors, ScCBF1 and ScZAT10, are significantly induced in So in response to mechanical agitation. Only one of the four St counterparts, StZAT10, is significantly induced in St. It can be concluded that, besides the conservation in their cold response, it appears to be some degree of conservation in the responses of these transcription factors to mechanical agitation and CHX treatment between Arabidopsis and Sc. However, the conservation in St is not so clear. 81 1.2 1 1.2 7 1 ' 1 i 0.8 - 0.8 0.6 -« i 0.6 0.4 ~ 1?. ~ 0.4 Ti . C 0.2 ~ 17“ 191‘ 0.2 :1 ’5 i 0 iii 1 g 0 .2.EL_22L_ 0 1M2 9 control control a. X 8 g 2 CZF1-like 13 - RAV1-like 713 .l 0: 1'5 0.8 ~ 1 . 0.6 1 0.5 *1 h 3': '4 0 22.1-2- 22 O .1 control 0 15 30 control 0 15 30 Time (minutes) after mechanical agitation Fig 3.1: Transcript accumulation of conserved cold-induced transcription factors in response to mechanical agitation. Three weeks old plants grown in magenta boxes under continuous light were dropped approximately 6” every 2 seconds for 15 minutes. Tissue was collected from 2 plants each time, 0, 15, and 30 minutes after the mechanical agitation treatment. qRT-PCR analysis determining the transcript levels of Sc (grey) and St (black) genes. Average values of two different experiments are shown. Relative expression level of each transcript were normalized using the potato 60$ gene (clone STMCK67) as internal reference. Relative expression of So 0 minute sample was set to 1. Error bars indicate SE. 82 CBF1 5 “ ZAT10-like I Sc 4 I St 3 1 I. l 2 ”i '1' c 1 ' "r 2“? 1” _8 “1— ..2i‘-TJI—- _! O 1 .- .41.}_ 1 1131‘. in. ._ r“ 2.— 3 0 1 2 4 8 24 0 1 2 4 8 24 O. X G) (D .2 g 7 j CZF1-Ii e 5 “ RAV1-like 0: 6 t": 5 ': 5 l 4 1 4 1 2'; 3 .. 3 ‘ “3 1 1 ; {Us 5,31 1 ‘ 1811:? 11 0 1 2 4 8 24 0 1 2 4 8 24 h of CHX treatment Fig 3.2: Transcript accumulation of conserved cold-induced transcription factors in response to cycloheximide treatment. Plants were grown under continuous light on liquid MS medium. After 3 weeks, cycloheximide was added to a final concentration of 10 pg/mL. Tissue was collected from 2 plants each time after 0, 1, 2, 4, 8, and 24h of cycloheximide treatment. qRT-PCR analysis determining the transcript levels of So and St genes. Average values of two different experiments are shown. Relative expression level of each transcript were normalized using the potato 60S gene (clone STMCK67) as internal reference. Relative expression of So 1h sample was set to 1. Error bars indicate SE. 83 Regulation of conserved transcription factors by the circadian clock Fowler, et al., 2005 have shown that in Arabidopsis, the low temperature response of the CBFs is gated by the circadian clock. It was also found that RAV1, other early cold—induced transcription factor, was gated by the clock, with the same phase than that of the CBFs. In order to explore in more detail the conservation in the regulation of these cold-induced transcription factors, circadian experiments were conducted with So and St plants to determine if the four conserved genes described in the previous section, were also gated by the circadian clock in these species. Sc and St plants were grown for 3 weeks under a 12:12 h (L:D) photoperiod and then switched to continuous light at dawn (ZTO). Plant tissue was harvested at different times after dawn (ZT) under warm conditions. Additionally, plants were transferred to 4°C at the same ZT, and were maintained in the cold for 1 or 4h. Two biological replicates were done for each experiment. The same primers for ScCBF1 and StCBF1 described in the previous section, were used here to test transcript accumulation by qRT-PCR. Fig 3.3a shows that ScCBF1 transcript accumulation cycles in warm conditions (grey line) (ANOVA, p<0.05,n=2). The cycling pattern can be observed only from the second day and on. There are peaks at ZT 28 and ZT 52, during subjective day and troughs around ZT 16 and ZT 40, during subjective night. The same pattern of transcript accumulation was observed when Sc plants were transferred to the 84 g.- cold for 1h (Fig 3.3a, black line). The pattern of cycling after 4h of cold (Fig 3.3b, black line) is similar, with peaks during subjective day and troughs during subjective night, however the peak during the second day was observed later, at ZT 34. The transcripts accumulated to a higher level when transferred to cold for 1h or 4h compared to warm. Given that the cold stimulus applied at different times of the day produces different intensities of ScCBF1 transcript accumulation, indicates that the cold response in ScCBF1 is gated by the circadian clock. The transcript level of StCBF1, like ScCBF1, cycles under warm conditions (Fig 3.30). There is a statistically significant difference between samples at different ZTs (ANOVA, p<0.05, n=2). Peaks are observed at ZT 10, 34 and 52 (all during the subjective day) and troughs at ZT 22 and 46 (at the end of the subjective night). This suggests that StCBF1 is under circadian regulation in the warm. When plants are transferred to cold for 1h, there is higher level of StCBF1 transcript compared to warm, however, the peaks at ZT1O and 52 disappear (Fig 3.30). In the case of plants that have been transferred to the cold for 4h (Fig 3.3d), the transcript levels are similar to those observed at warm temperatures and there is no statistically significant difference between peaks and troughs. The cycling observed at 1h cold during the second day suggests that StCBF1 may also be gated by the circadian clock. Given that the low temperature responses of AtCBFs are gated by the circadian clock, the results shown here suggest conservation in the CBF cold response in the two Solanum species studied. 85 a C 6 Sc CBF1 20 ; StCBF1 5 . 4 . 15 « 3 ~ ._w 10 j _..w 2 % —1nc 5 i —1hC c 1 l g 0 1 ' ‘ i O ' I ‘K I' "'l a ZT(h)4 1O 16 22 28 34 40 46 52 ZT(h) 4 1016 22 28 34 40 46 52 6 I: III-I m =5 b d g 10 Sc CBF1 3.5 St CBF1 a: 8 i 3.0 I 2.5 l 6 2.0 4 “W 1.5 ”W 2 , —4nc 1.0 . -4hC ; 0.5 O1 ‘ i l“ l 0.01/TII l‘! I lA'U’WIH') ZT (h) 4 10 16 22 28 34 40 46 52 ZT(h) 4 1016 22 28 34 40 46 52 l:-:-:| m: Fig 3.3: Circadian clock regulation of CBF1. Wild type S. commersonii (Sc) and S. tuberosum (St) plants were grown under 12:12 photoperiod for 3 weeks and then switched to continuous light at ZTO. Plant tissue was harvested at the different ZT shown (W: warm temperature). Additionally, plants were transferred to 4°C at the same ZT, and were maintained at 4°C for 1h (a, c) and 4h (b, d). qRT-PCR analysis determining the transcript levels of Sc (a-b) and St (c—d) genes. Average values of two different experiments are shown. Relative expression level of each transcript were normalized using the potato 60$ gene (clone STMCK67) as internal reference. Relative expression of ZT 4 4hC sample was set to 1. Error bars indicate SE. White box indicates subjective day and black box subjective night. 86 To test ScRAV1 and StRAV1 transcript accumulation under circadian conditions, primers that target PUT3404 and 3405 were used. The level of ScRAV1 transcript is higher at 1h and 4h of cold compared to warm (fig 3.4a and b). This transcript shows cycling at warm, 1h cold and 4h cold. However, the cycling at 1h cold was not statistically significant (ANOVA, p>0.05, n=2). In the three cases, the peaks occur roughly at the end of the subjective day period, and the troughs during the subjective night which is similar to that observed for AtRAV1 (40). These results suggest that ScRAV1, like AtRAV1, is also gated by the clock. StRA V1 also shows cycling under warm conditions (Fig 3.4c). with peaks during the subjective day and troughs during the subjective night, however there was not a statistically significant difference between the peaks and troughs (ANOVA, P>0.05, n=2). When St plants were transferred to cold for 1h, there was higher accumulation of StRAV1 transcript, but there was no statistically significant cycling (Fig 3.4c). When plants were transferred to cold for 4h, there is no cold-induction of this transcript, but there was a statistically significant difference between the peaks during subjective day and troughs during subjective night, indicating that after 4h of cold this transcript is cycling (ANOVA, p<0.05, n=2) (Fig 3.4d). There is no sufficient data to conclude whether the cold-induction of StRAV1 is gated by the clock or not. 87 2-01 ScRAV1 j “W 41 —1hC C .2 j 5 am) 4 1016 22 28 34 4o 46 52 a: 0 .2. b d % 2.5 1 ScRAV1 a: 2.0 1 1.5 ' -—--w 1-0 ! —4nc 0.5 '1 0.0 J ~.- -—~ ZT(h) 4 1016 22 28 34 40 46 52 30 2.0 4 fl“ ""‘C 1.0 , .4, 0.0 +-r~~~- ~ r—~~---—— _- zr (h) 4 1016 22 28 34 4o 46 52 3.0 > St RAV1 2.5 2.0 -i 1.5 4: 1.0 3 0.5 i 0.0 4444—44-4 —--——-—~4 -~ -—. --W —4h C ZT(h) 4 1016 22 28 34 40 46 52 l:-:-::l Fig 3.4: Circadian clock regulation of RAV1. Wild type S. commersonii (Sc) and S. tuberosum (St) plants were grown under 12:12 photoperiod for 3 weeks and then switched to continuous light at ZTO. Plant tissue was harvested at the different ZT shown (W: warm temperature). Additionally, plants were transferred to 4°C at the same ZT, and were maintained at 4°C for 1h (a, c) and 4h (b, d). qRT-PCR analysis determining the transcript levels of Sc (a-b) and St (c-d) genes. Average values of two different experiments are shown. Relative expression level of each transcript were normalized using the potato 60$ gene (clone STMCK67) as internal reference. Relative expression of ZT 4 4hC sample was set to 1. Error bars indicate SE. White box indicates subjective day and black box subjective night. 88 Preliminary studies suggest that AtCZF1 and AtZAT10 may also be gated by the clock in Arabidopsis (Dong, unpublished data). To determine whether CZF1 and ZAT10 were cold-gated or not in Sc and St, their transcript accumulation was tested in the circadian experiment. ScCZF1 and StCZF1 transcript accumulation was also tested under circadian conditions (Fig 3.5). Primers that target PUT32825 and 45213 were used for qRT-PCR. The ScCZF1 transcript is cold-induced at 1h and 4h (Fi93.5a and b). The ScCZF1 transcript levels cycles in warm and cold (1h and 4h) conditions (ANOVA, p<0.05, n=2) (Fig 3.5a and b). In all cases there were peaks during the subjective day and troughs during the subjective night, a pattern that is similar to that observed for ScRAV1. These results indicate that the low temperature induction of ScCZF1 is gated by the circadian clock. In the case of StCZF1, the induction by 1h and 4h of cold was only around two fold. The cycling of this transcript was statistically significant only under warm conditions and 4h cold (ANOVA, p<0.05, n=2) (Fig 3.5d). In these two cases, the phase of the cycling was the same as the one observed for ScCZF1. Given these results, it is possible that the lack of StCZF1 cycling at 1h cold may be due to an error in the handling of the St 1h cold samples, rather than a real absence of cycling just after 1h cold (Fig 3.5C). Therefore, these results suggest that StCZF1 low temperature induction is also gated by the clock. 89 4.0 6 a 3.0 . 6 ‘ 2.0 / “W 4 1 R —1nc : 1.0 fl-+\‘1--- >/ "‘1‘\ 2 O \‘1’ “3...; 2 ZT(h) 4 1o 16 22 26 34 4o 46 52 ZT(h) 4 1016 22 28 34 40 46 52 & EZ-I [II (D o .2 33 b ‘0 1 Sc CZF1 d 10 StCZF1 g 8 4 8 6 W 6 l “w *W 4 —4hC 4 zjx -4hC 2 2 0 ‘11“; 1"”"T ECETi‘é 0 ‘ /l I r r I ifi 7 1 ZT(h)4 1o 16 22 26 34 4o 46 52 ZT(h)4 1o 16 22 26 34 4o 46 52 l:-:-:l [II-I Fig 3.5: Circadian clock regulation of CZF1. Wild type S. commersonii (Sc) and S. tuberosum (St) plants were grown under 12:12 photoperiod for 3 weeks and then switched to continuous light at ZTO. Plant tissue was harvested at the different ZT shown (W: warm temperature). Additionally, plants were transferred to 4°C at the same ZT, and were maintained at 4°C for 1h (a, c) and 4h (b, d). qRT—PCR analysis determining the transcript levels of Sc (a-b) and St (c—d) genes. Average values of two different experiments are shown. Relative expression level of each transcript were normalized using the potato 60$ gene (clone STMCK67) as internal reference. Relative expression of ZT 4 4hC sample was set to 1. Error bars indicate SE. White box indicates subjective day and black box subjective night. 90 Finally, the ScZAT10 and StZAT10 genes were tested for circadian regulation. Primers that target PUT32825 and 45213 were used for qRT-PCR. ScZAT10 transcript accumulates to a higher level in 1h and 4h cold treated samples compared to warm samples (Fig 3.6a and b). This transcript showed cycles with peaks during subjective day and troughs during subjective night, the same phase as ScRAV1 and ScCZF1. Even though the cycling at 1h cold is not statistically significant (ANOVA, p>0.05, n=2), peaks and troughs can also be observed at the same ZT as the ones at warm and 4h cold. These results suggest that the ScZA T10 low temperature response is gated by the clock. In the case of StZAT10, there is also cold-induction at 1h and 4h, and there is cycling with the same phase as ScZAT10, that is peaks during subjective day and troughs during subjective night (Fig 3.60 and d). Also in this case, the cycling was statistically significant only under warm and 4h cold (ANOVA, p<0.05, n=2). These results suggest that StZAT10 is also gated by the clock. From these results it was concluded that the low temperature responses of three of the four conserved early cold-induced genes selected (CBF1, CZF1, and ZAT10) are gated by the circadian clock in St and Sc. It can also be concluded that the cold gating of RAV1 is conserved between Sc and Arabidopsis. In the case of CBF1, their cold—gating response is conserved in these two Solanum species as well as in Arabidopsis. There are no obvious differences between the cold gating of these transcription factors between Sc and St that could account for their differences in freezing tolerance. Rather, it 91 appears that the imput of the circadian clock in cold stress response is conserved in the two Solanum species. 92 a c 2-0 - Sc ZAT10 2.0 - StZAT10 1.5 ~ 1.5 1.0 ....w 1.0 -‘ --w c 0.5 V —1nc 0.5 —1hC .9 {/T 5V” g o 0 ' 0 0 " T ‘l i l l '1"! 1 '9- zrm) 4 1016 22 26 34 4o 46 52 ZT(h) 4 1o 16 22 26 34 4o 46 52 a, :-_—_'-:I 0 .2 d 5 b3'0 3 Sc ZAT10 ‘2 StZAT10 2.0 J 0.8 ~ 1.5 4 ._..w 0.6 « _..w 1.0 J -4nc 0.4 —4hC o 5 o 2 " o 0 b 1 “ ‘1 T“ _1 0 0 H F’ ’1" '1' T i ZT(h)4 1016 22 26 34 4o 46 52 ZT(h) 4 1016 22 26 34 4o 46 52 IE III-I Fig 3.6: Circadian clock regulation of ZAT10. Wild type S. commersonii (Sc) and S. tuberosum (St) plants were grown under 12:12 photoperiod for 3 weeks and then switched to continuous light at ZTO. Plant tissue was harvested at the different ZT shown (W: warm temperature). Additionally, plants were transferred to 4°C at the same ZT, and were maintained at 4°C for 1h (a, c) and 4h (b, d). qRT-PCR analysis determining the transcript levels of Sc (a-b) and St (c-d) genes. Average values of two different experiments are shown. Relative expression level of each transcript were normalized using the potato 60$ gene (clone STMCK67) as internal reference. Relative expression of ZT 4 4hC sample was set to 1. Error bars indicate SE. White box indicates subjective day and black box subjective night. 93 DISCUSSION In order to further explore the conservation of early cold induced transcription factors in Sc, St and Arabidopsis, the main goal of this chapter was to study the responses of four transcription factors (CBF1, RAV1, ZAT10, and CZF1) to mechanical agitation, inhibition of protein synthesis, and circadian control. Given that these four transcription factors are responsive to mechanical agitation and CHX treatment (inhibition of protein synthesis) in Arabidopsis, accumulation of the St and Sc transcripts upon the different treatments was tested by qRT-PCR. Given that the same primers were used to test transcript accumulation of both species, it can be concluded that the gene transcript levels were always higher for the Sc transcripts than to those of St. If some differences in hybridization efficiency of the primers were to happen, the primers could have had less homology to the Sc genes (the Sc genes are unknown) and that could result in less hybridization efficiency, but given that in all the cases the Sc transcripts were higher than the St transcripts, this suggests that primers hybridized to Sc transcripts as well as to St transcripts. In this study the transcript of ScCBF1, ScZAT10 and StZAT10 were shown to significantly accumulate after mechanical agitation. The lack of significant accumulation of the other gene transcripts tested suggests that the mechanical agitation response is not that well conserved between Arabidopsis and the Solanum species studied. However, this experiment was done only 94 twice therefore more replicates of this experiment should give more statistically significant results. To test transcript accumulation in response to inhibition of protein synthesis, CHX treatment was performed. The four gene transcripts tested were significantly induced in Sc. Only StCZF1 was significantly induced by CHX treatment in St; however to a smaller level. StCBF1 showed high levels of transcripts in no treatment controls (0, Fig 3.2). It is known that the levels of StCBF1 are almost undetectable in non-inducing (warm) conditions. In this experiment the level of StCBF1 in the no treatment control is higher than any other time point of CHX treatment. Given that some induction of StCBF1 can be observed by mechanical agitation (Fig 3.1) it is possible that undesired agitation of no treatment control plants is responsible for this high level of StCBF1. It is unknown if inhibition of protein synthesis by CHX treatment occurs with the same efficiency in both So and St. It is possible that the Sc plants had better adsorption of the CHX than St. This was not tested. It has been previously observed that unstable transcripts like CBF1-3, which have a half life of 7.5 min at warm temperatures (38), are associated with a mechanical agitation response and with clock-controlled genes (93). Given that CBF1-3 and RAV1 cold-induction are gated by the circadian clock in Arabidopsis (40), the low temperature responses of the four conserved early- induced genes were tested to determined if they were also gated by the circadian clock in Sc and St. 95 The results presented in this study indicate that three (CBF1, CZF1, and ZAT10) of the four transcription factors that are early cold-induced in St, Sc and Arabidopsis are gated by the circadian clock in the two Solanum species studied. The gating of RAV1 cold-induction is conserved in Sc and Arabidopsis. ScCBF1 cold-induction is gated by the circadian clock. The lack of a peak at ZT 4 may be because plants are in the transition from 12:12 photoperiod to continuous light. In the first 12 h into the continuous light, plants don’t sense yet they are in continuous light. It is possible they are still behaving as in 12:12h photoperiod and it is unknown whether in this condition ScCBF1 peaks at ZT 4 or earlier. If it peaks earlier, the peak would have been missed in this experiment. It has been observed that, for instance, CCA1 peaks at ZTO under a 12:12h photoperiod, but when is transferred to continuous light it peaks later (94). The StCBF1 transcript is circadian regulated in warm conditions. When transferred to cold for 1h the peak and trough can only be observed during the second day (peak at ZT 34 and trough at ZT46). After 4h of cold there is apparent cycling, however the peaks (ZT 10 and ZT 34) and troughs (ZT 22 and ZT 40) are not statistically significant. This suggests that StCBF1 may also be gated by the circadian clock. More replicates of the experiment should give more statistically significant results. Previously, Pennycooke et al. (2008) showed that S. chopersicon (common tomato) and S. pimpinellifolium (wild tomato), two Solanum species that are chilling and freezing sensitive, have CBF1 genes that are regulated by 96 light and the circadian clock (43). Therefore, it would not be unexpected that StCBF1, another Solanum species that is freezing sensitive, would also be gated by the clock. The other three transcription factors studied (RAV1, ZAT10, and CZF 1) also showed gating of their low temperature response in So, however in St the gating of RAV1 cannot be concluded. There are peaks during the subjective day and troughs during the subjective night for all of these cold gated genes. ScRAV1 has a cycling pattern that has the same phase as AtRAV1 (40), which suggest that the gated cold response of RA V1 is conserved. The cold-induction of StCZF1 was only around two fold (Fig 3.5c and d). However, in chapter 2 (Fig 2.7) there was much more high cold induction of this transcript. This difference could be due to the fact that in the experiments of chapter 2, plants are grown in 16:8h photoperiod and then transferred to cold at ZT 8, but in this circadian experiment the plants were entrained in 12:12h photoperiod and then transferred to continuous light, where the temperature is dropped at the different Us This different growth conditions might have had this effect on the cold-induction of this gene. It is unknown if Arabidopsis ZAT10 and CZF1 cold responses are gated by the circadian clock. Preliminary data suggest that they might be (Dong, unpublished data). If these results are confirmed, it will suggest conservation in the cold-induction gating of these genestoo. 97 "1 MATERIALS AND METHODS Plant growth and experimental conditions Single node cuttings of Solanum tuberosum cv. Umatilla and Solanum commersonii were grown on half strength MS medium supplemented with ZOg/L sucrose and 7g/l Plant Agar (Sigma), pH 5.6 in Magenta GA7 vessels with six plants per vessel. Plants were grown in an Enconair growth chamber ("Bigfoot" GC-20) at 25°C under a 12:12h L:D photoperiod (light intensity 100 u mol rn'1 s"), for 3 weeks before sampling. Two plants were randomly selected for each treatment and each treatment was repeated once for both species. The top portions (approximately 5 cm) of plants were collected and the lower sections and root systems were discarded. Samples were placed in 15 ml falcon tubes, immediately submerged in liquid nitrogen and stored in a -80°C freezer. For collections during the dark period, samples were collected with very minimal indirect light. Samples that coincided with the transition from light to'dark or dark to light were collected immediately prior to the transition. For the circadian experiments, three week old plants grown in a 12:12 L:D photoperiod were transferred to continuous light (100 pmol rn'1 s"). Two replicates from both species were collected at 4, 10, 16, 22, 28, 34, 40, 46 and 52h after the beginning of continuous light (warm samples). In addition, 3 week old plants were also transferred to continuous light at ZTO. In this case, plants were transferred to a 4°C walk-in cold room with continuous light (100 pmol m'1 98 s") at ZT4, 10, 16, 22, 28, 34, 40, 46 and 52. After 1h and 4h of cold, two replicates were collected for each ZT. For the mechanical agitation treatment, 3 week old plants of both species grown under continuous light in Magenta vessels were secured together in a cardboard box and dropped approximately 6” every 2 seconds for 15 minutes. Samples were collected at O, 15 and 30 minutes after mechanical treatment. Two replicates of two plants from each species were randomly selected from each vessel and frozen in liquid nitrogen immediately. A control without agitation was included. For the inhibition of protein synthesis experiment, plants were grown under continuous light on liquid half strength MS medium supplemented with 209/l sucrose, pH 5.6 on a filter paper bridge held above the level of the liquid medium. Capillary action maintained a constant supply of culture medium to the plants on the filter paper. After 3 weeks of growth, cycloheximide (CHX) was mixed into the existing liquid medium to give a final concentration of 10pg/ml. Two replicates of two plants from each vessel were collected and frozen in liquid nitrogen for each treatment. Samples were collected after 0, 1, 2, 4, 8 and 24 hours after the beginning of CHX treatment. A control without CHX in the medium was included. RNA extraction and qRT-PCR RNA extraction and qRT-PCR were done as described in chapter two. Primers used are listed in Table 2.1. 99 APPENDIX Table A1: 40pOG cold-induced in S. commersonii (Sc) 8. tuberosum (St) and Arabidopsis (At). St up Sc up At up At description STMDJ69 STMGE83 ATZG18900 transducin family protein / WD-40 repeat family STMDJ69 protein STMIW06 STMIW06 AT4625990 chloroplast import apparatus CIA2-Iike STMIU11 STMIU11 AT5660680 unknown protein unknown protein AT2628400 STMCG§2 STMGV17 AT2G45660 AGL20 (Agamous-Iike 20) STMCG§2 STMCN22 STMCN22 ATSG65280 GCL1 (GCR2-Iike 1); catalytic STMEK16 STMIQ63 AT3G12670 EMBZ742 (embryo defective 2742) STMEK16 STMIV71 STMIV71 AT1GZ776O interferon-related developmental regulator family STMIY51 STMlY51 protein / IFRD protein family STMJG77 STMIU74 ATSGO1880 zinc finger (C3HC4-type RING finger) family protein STMIU74 STMEP26 STMEP26 AT5G26920 calmodulin binding STME027 STME027 AT2633210 chaperonin putative HSP60 (Heat shock protein 60) AT3623990 STMIU32 STMIU32 AT3G$3230 cell division cycle protein 48 putative (CDC48) STMIDZ4 STMIDZ4 STMIQZS STMIQZ6 AT1GO1470 LEA14 (Late embryogenesis abundant 14) STMJI56 STMJI56 STMGA34 STMGA34 STMDH66 STMJL22 AT4635940 unknown protein STMJL22 STMEI36 STMEI36 AT1G31660 unknown protein STMEQSS STMET41 STMET41 AT1625400 unknown protein STMGF95 STMGF95 AT1GS1700 ADOF1 (Arabidopsis dof zinc finger protein 1) STMIY82 STMIY82 AT1652890 ANACO19 (Arabidopsis NAC domain containing AT4627410 protein 19) R026 (responsive to dessication 26) STMDOBG STMDS75 AT3616810 APUM24 (Arabidopsis pumilio 24) STMDOBS STMEW81 STMEW81 AT5G62360 invertase/pectin methylesterase inhibitor family STMCBQO STMCBQO AT5662350 protein invertase/pectin methylesterase inhibitor family protein (DC 1.2 homolog) STMHE19 STMHE19 AT5620830 SUS1 (sucrose synthase 1) STMDP77 STMDP77 AT4629780 unknown protein 100 Table A1 continued STMGH65 STMJJ17 STMHA92 STMHSZQ STMGG79 STMCX87 STMDU38 STMHO64 STMIH78 STMGL16 STMEDSO STMGR56 STMIP59 STMHT66 STM|O48 STMGU17 STMHS17 STMJ029 STMJO47 STMIX48 STMHO88 STMHT73 STMGJ81 STMH634 STMJJ17 STM HA92 STMHSZQ STMGG79 STMCX87 STMHN39 STMDU38 STMHO64 STMIH78 STMGL16 STMED50 STMGR56 STMIP59 STMHT66 STM|O48 STMGU17 STMHS17 STMJ029 STMJO47 STMIX48 STMHO88 STMHT73 STMGJ81 AT4630290 AT5648070 AT1 667970 AT1 642440 AT3G5551 0 AT561601 0 AT168027O AT2617270 AT4627940 AT4633905 AT2614860 AT4628450 AT1632860 AT4G31 140 AT4600640 AT4612000 AT361 141 0 AT1 607430 AT1620450 AT1620440 AT1 6761 80 AT3622840 AT4614690 AT5607990 AT1653645 ATXTH19 ATXTH20 (Xyloglucan endotransglucosilase hydrolase 19 and 20) AT-HSFA8 (Arabidopsis thaliana heat shock transcription factor A8) unknown protein unknown protein 3-oxo-5-alpha-steroid 4-dehydrogenase family protein I steroid 5-aIpha-reductase family protein DNA-binding protein putative mitochondrial substrate carrier family protein mitochondrial substrate carrier family protein peroxisomal membrane protein 22 kDa putative peroxisomal membrane protein 22 kDa putative transducin family protein l WD-40 repeat family protein glycosyl hydrolase family 17 protein glycosyl hydrolase family 17 protein unknown protein unknown protein ATPPZCA (Arabidopsis protein phosphatase ZCA) protein phosphatase 2C putative ERD10/LTI45 (early responsive to dehydration 10) COR47 (cold regulated 47) ERD14 (early responsive to dehydration 14) ELIP1 (early light-inducible protein) chlorophyll binding ELIP2 (early light-inducible protein 2) chlorophyll binding TT7 (transparent testa 7) hydroxyproline-rich glycoprotein family protein 101 Table A2: 13pOG cold-repressed in S. commersonii (Sc), S. tuberosum (St), and Arabidopsis (At). St down So down At down At description STMDV46 STMDV46 AT1609750 chloroplast nucleoid DNA-binding protein-related STMIV36 STMIV36 AT2G39470 PPL2 (PSBP-like protein 2) STMCX38 STMCX38 ATSG16150 L-asparaginase putative STMER63 STMDBS7 AT3G23730 xyloglucan:xyloglucosyl transferase putative STMER63 STMDB$7 STMGOZ3 STMGOZ3 AT4G14540 CCAAT—box binding transcription factor subunit 8 (NF- YB HAP3 STMCR16 STMCR16 AT1G70410 car)bgnic an1'1ydrase putative; carbonate dehydratase STMCL01 STMCL01 AT3GO1500 putative CA1 (carbonic anhydrase 1) STMCV75 STMCV75 STMIV24 STMIV24 STMCK44 STMC K44 STMEP82 STMCSBQ AT1G48600 phosphoethanolamine N-methyltransferase 2 putative STMCSBQ (NMT2) STMCD65 STMCDBS AT5G56870 BGAL4 (beta-galactosidase 4) STMGX24 STMGX24 AT5G35790 G6PD1 (glucose-6-phosphate dehydrogenase 1) STMJ H69 STMJ H69 AT3G15840 PIFI (post-illumination chlorophyll fluorescence increase STMIM55 STMIM55 AT1632080 membrage protein putative 3TM3378 STMC055 AT1G73330 ATDR4 (Arabidopsis thaliana drought-repressed 4) TM 55 STMJD18 STMJD18 AT4625260 invertase/pectin methylesterase inhibitor family protein AT4G12390 PME1; pectinesterase inhibitor 102 Table A3: 27 pOG cold-induced in S. commersonii (Sc) and Arabidopsis (At), but not in S. tuberosum (St). Sc up At up At description STMEF80 AT3G49320 unknown protein STMEF80 STMDJ58 AT5G39410 binding / catalytic STMCN38 AT3608950 electron transport SCO1/SenC family protein STM|U24 AT3G16720 ATL2 (Arabidopsis Toxicos en Levadura 2) STMCM56 AT3627880 unknown protein AT1GZ3710 AT1G70420 STMDH61 AT4G37090 unknown protein STMDH61 STMCN51 AT1G13930 unknown protein STMCSGG STMEQZS STMCSZS STMIN87 AT2G14560 unknown protein STMHSG7 AT3G46460 UBC13 (Ubiquitin-conjugating enzyme 13) STMESBQ AT1G73630 calcium-binding protein, putative STMCY67 AT5G62430 CDF1 (cycling dof factor 1), CDF3 (cycling dof factor 3) AT3G47500 STMEU37 AT4G25470 CBF2 (freezing tolerance QTL 4) DREB1A (dehydration AT4G25480 response element B1A) STMDE93 AT3G13940 DNA binding / DNA-directed RNA polymerase STMDE93 STMDP46 AT1G04240 SHY2 (short hypocotyl 2) STMDDSQ AT3G44260 CCR4-NOT transcription complex protein, putative STMEQSB AT2G28720 histone H28, putative STMJN45 AT5667480 BT4 (BTB AND TAZ domain protein 4) STMJF94 AT3655120 A1 1/CFl/TT 5 (transparent testa 5) STMCV31 AT5620180 ribosomal protein L36 family protein STMGD36 STMIQ74 AT4G27520 plastocyanin-like domain-containing protein STMEY77 AT3G55430 glycosyl hydrolase family 17 protein / beta-1,3-glucanase STMDZ38 AT3G54030 protein kinase family protein STMEY96 AT2625625 unknown protein STMCG80 AT3GS6090 ATFER3 (ferritin 3) STMEY27 AT4G18530 unknown protein STMGX29 AT1G76930 ATEXT4 (extensin 4) STMHS45 AT2G39130 amino acid transporter family protein 103 10. 11. 12. 13. 14. 15. l6. l7. LITERATURE CITED Guy, C. (1990) Annu. Rev.Plant Physiol. Plant Mol. Biol. 41, 187-223 Thomashow, M. F. (1999) Annu Rev Plant Physiol Plant Mol Biol 50, 571- 599 Jaglo, K. R., Kleff, S., Amundsen, K. L., Zhang, X., Haake, V., Zhang, J. Z., Deits, T., and Thomashow, M. F. (2001) Plant Physio! 127, 910-917 Oh, S. J., Song, S. 1., Kim, Y. S., Jang, H. J., Kim, S. Y., Kim, M., Kim, Y. K., Nahm, B. H., and Kim, J. K. (2005) Plant Physiol 138, 341-351 Foyer, C., Vanacker, H., Gomez, L., and Harbinson, J. (2002) Plant Physiol and Biochem 40 Kamps, T., Isleib, T., Herner, R., and Sink, K. (1987) HortScience 22, 1309- 1312 Chen, H. H., and Li, P. H. (1980) Plant Physio! 65, 1146-1148 Shame, P., Sharma, N., and Deswal, R. (2005) Bioessays 27, 1048-1059 Uemura, M., Joseph, R. A., and Steponkus, P. L. (1995) Plant Physio! 109, 15-30 Thomashow, M. F. (1998) Plant Physiol 118, 1-8 Uemura, M., and Steponkus, P. L. (1994) Plant Physiol 104, 479-496 Strauss, G., and Hauser, H. (1986) Proc Nat! Acad Sci U S A 83, 2422-2426 Rudolph, A. S., and Crowe, J. H. (1985) Cryobiology 22, 367-377 Hoekstra, F. A., Golovina, E. A., Tetteroo, F. A., and Wolkers, W. F. (2001) Cryobiology 43, 140-150 Gilmour, S. J., Lin, C., and Thomashow, M. F. (1996) Plant Physiol 111, 293-299 Lin, C., and Thomashow, M. F. (1992) Plant Physiol 99, 519-525 Campbell, S., and Close, T. (1997) New Phytol 137, 61-74 104 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. Steponkus, P. L., Uemura, M., Joseph, R. A., Gilmour, S. J., and Thomashow, M. F. (1998) Proc Nat! Acad Sci U S A 95, 14570-14575 Fowler, S., and Thomashow, M. F. (2002) Plant Cell 14, 1675-1690 Vogel, J. T., Zarka, D. G., Van Buskirk, H. A., Fowler, S. G., and Thomashow, M. F. (2005) Plant J41, 195-211 Hajela, R. K., Horvath, D. P., Gilmour, S. J., and Thomashow, M. F. (1990) Plant Physio! 93, 1246-1252 Yamaguchi-Shinozaki, K., and Shinozaki, K. (1994) Plant Cell 6, 251-264 Stockinger, E. J., Gilmour, S. J., and Thomashow, M. F. (1997) Proc Natl Acad Sci U S A 94, 1035-1040 Riechmann, J. L., and Meyerowitz, E. M. (1998) Biol Chem 379, 633-646 Shinwari, Z. K., Nakashima, K., Miura, S., Kasuga, M., Seki, M., Yamaguchi-Shinozaki, K., and Shinozaki, K. (1998) Biochem Biophys Res Commun 250, 161-170 Gilmour, S. J., Fowler, S. G., and Thomashow, M. F. (2004) Plant Mo! Biol 54, 767-781 Gilmour, S. J., Zarka, D. G., Stockinger, E. J., Salazar, M. P., Houghton, J. M., and Thomashow, M. F. (1998) Plant J 16, 433-442 Gilmour, S. J., Sebolt, A. M., Salazar, M. P., Everard, J. D., and Thomashow, M. F. (2000) Plant Physio! 124, 1854-1865 Jaglo-Ottosen, K. R., Gilmour, S. J., Zarka, D. G., Schabenberger, 0., and Thomashow, M. F. (1998) Science 280, 104-106 Liu, Q., Kasuga, M., Sakuma, Y., Abe, H., Miura, S., Yamaguehi-Shinozaki, K., and Shinozaki, K. (1998) Plant Cell 10, 1391-1406 Cook, D., Fowler, S., Fiehn, 0., and Thomashow, M. F. (2004) Proc Nat! Acad Sci U S A 101, 15243-15248 Chinnusamy, V., Ohta, M., Kanrar, S., Lee, B. H., Hong, X., Agarwal, M., and Zhu, J. K. (2003) Genes Dev 17, 1043-1054 Dong, C. H., Agarwal, M., Zhang, Y., Xie, Q., and Zhu, J. K. (2006) Proc Natl Acad Sci U S A 103, 8281-8286 105 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. Miura, K., Jin, J. B., Lee, J., Yoo, C. Y., Stirm, V., Miura, T., Ashworth, E. N., Bressan, R. A., Yun, D. J., and Hasegawa, P. M. (2007) Plant Cell 19, 1403-1414 Doherty, C. J., Van Buskirk, H. A., Myers, S. J., and Thomashow, M. F. (2009) Plant Cell 21, 972-984 Novillo, F., Alonso, J. M., Eeker, J. R., and Salinas, J. (2004) Proc Nat! Acad Sci U S A 101, 3985-3990 Agarwal, M., Hao, Y., Kapoor, A., Dong, C. H., Fujii, H., Zheng, X., and Zhu, J. K. (2006) J Biol Chem 281, 37636-37645 Zarka, D. G., Vogel, J. T., Cook, D., and Thomashow, M. F. (2003) Plant Physiol 133, 910-918 Harmer, S. L., Hogenesch, J. B., Straume, M., Chang, H. S., Han, B., Zhu, T., Wang, X., Kreps, J. A., and Kay, S. A. (2000) Science 290, 2110-2113 Fowler, S. 6., Cook, D., and Thomashow, M. F. (2005) Plant Physio! 137, 961-968 Wang, Z. Y., Kenigsbuch, D., Sun, L., Harel, E., Ong, M. S., and Tobin, E. M. (1997) Plant Cell 9, 491-507 McClung, C. R. (2008) Curr Opin Plant Biol 11, 514-520 Pennycooke, J. C., Cheng, H., Roberts, S. M., Yang, Q., Rhee, S. Y., and Stockinger, E. J. (2008) Plant Mol Biol 67, 483-497 Franklin, K. A., and Whitelam, G. C. (2007) Nat Genet 39, 1410-1413 Leivar, P., Monte, E., Al-Sady, B., Carle, C., Storer, A., Alonso, J. M., Ecker, J. R., and Quail, P. H. (2008) Plant Cell 20, 337-352 Kidokoro, S., Maruyama, K., Nakashima, K., Imura, Y., Narusaka, Y., Shinwari, Z. K., Osakabe, Y., Fujita, Y., Mizoi, J., Shinozaki, K., and Yamaguchi-Shinozaki, K. (2009) Plant Physiol Zhu, J., Verslues, P. E., Zheng, X., Lee, B. H., Zhan, X., Manabe, Y., Sokolchik, I., Zhu, Y., Dong, C. H., Zhu, J. K., Hasegawa, P. M., and Bressan, R. A. (2005) Proc Nat! Acad Sci U S A 102, 9966-9971 Cao, S., Ye, M., and Jiang, S. (2005) Plant Cell Rep 24, 683-690 Choi, D. W., Rodriguez, E. M., and Close, T. J. (2002) Plant Physiol 129, 1781-1787 106 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. Benedict, C., Skinner, J. S., Meng, R., Chang, Y., Bhalerao, R., Huner, N. P., Finn, C. E., Chen, T. H., and Hurry, V. (2006) Plant Cell Environ 29, 1259-1272 Dubouzet, J. G., Sakuma, Y., Ito, Y., Kasuga, M., Dubouzet, E. G., Miura, S., Seki, M., Shinozaki, K., and Yamaguchi-Shinozaki, K. (2003) Plant J33, 751-763 Qin, F., Sakuma, Y., Li, J., Liu, Q., Li, Y. Q., Shinozaki, K., and Yamaguchi-Shinozaki, K. (2004) Plant Cell Physio! 45, 1042-1052 Okamuro, J. K., Caster, B., Villarroel, R., Van Montagu, M., and Jofuku, K. D. (1997) Proc Natl Acad Sci U S A 94, 7076-7081 Jiang, C., In, B., and Singh, J. (1996) Plant Mo! Bio! 30, 679-684 Gao, M. J., Allard, G., Byass, L., Flanagan, A. M., and Singh, J. (2002) Plant Mo! Biol 49, 459-471 Xue, G. P. (2002) Biochim Biophys Acta 1577, 63-72 Takumi, S., Shimamura, C., and Kobayashi, F. (2008) Plant Physiol Biochem 46, 205-211 Kasuga, M., Miura, S., Shinozaki, K., and Yamaguchi-Shinozaki, K. (2004) Plant Cell Physiol 45, 346-350 Pino, M. T., Skinner, J. S., Park, E. J., Jeknic, Z., Hayes, P. M., Thomashow, M. F., and Chen, T. H. (2007) Plant Biotechnol J 5, 591-604 Pino, M. T., Skinner, J. S., Jeknic, Z., Hayes, P. M., Soeldner, A. H., Thomashow, M. F., and Chen, T. H. (2008) Plant Cell Environ 31, 393-406 Zhang, X., Fowler, S. G., Cheng, H., Lou, Y., Rhee, S. Y., Stockinger, E. J., and Thomashow, M. F. (2004) Plant J39, 905-919 Badawi, M., Reddy, Y. V., Agharbaoui, Z., Tominaga, Y., Danyluk, J., Sarhan, F., and Houde, M. (2008) Plant Cell Physiol 49, 1237-1249 Monroy, A. F., Dryanova, A., Malette, B., Oren, D. H., Ridha Farajalla, M., Liu, W., Danyluk, J., Ubayasena, L. W., Kane, K., Scoles, G. J., Sarhan, F., and Gulick, P. J. (2007) Plant Mo! Biol 64, 409-423 Kim, S., An, C. S., Hong, Y. N., and Lee, K. W. (2004) Mo! Cells 18, 300-308 107 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. Xiao, H., Siddiqua, M., Braybrook, S., and Nassuth, A. (2006) Plant Cell Environ 29, 1410-1421 Xiong, Y., and Fei, S. Z. (2006) Planta 224, 878-888 Caceres, M., Lachuer, J., Zapala, M. A., Redmond, J. C., Kudo, L., Geschwind, D. H., Lockhart, D. J., Preuss, T. M., and Barlow, C. (2003) Proc Natl Acad Sci U S A 100, 13030-13035 Brem, R. B., Yvert, G., Clinton, R., and Kruglyak, L. (2002) Science 296, 752-755 Jiao, Y., Ma, L., Strickland, E., and Deng, X. W. (2005) Plant Cell 17, 3239- 3256 Walia, H., Wilson, C., Ismail, A. M., Close, T. J., and Cui, X. (2009) BMC Genomics 10, 398 Gong, Q., Li, P., Ma, S., Indu Rupassara, S., and Bohnert, H. J. (2005) Plant J 44, 826-839 Smith, G. (2004) Appl Genet Mol Biol 3 Cheong, Y. H., Chang, H. S., Gupta, R., Wang, X., Zhu, T., and Luan, S. (2002) Plant Physio! 129, 661-677 Kiyosue, T., Yamaguchi-Shinozaki, K., and Shinozaki, K. (1994) Plant Cell Physiol 35, 225-231 Hutin, C., Nussaume, L., Moise, N., Maya, 1., Kloppstech, K., and Havaux, M. (2003) Proc Natl Acad Sci U S A 100, 4921-4926 Oufir M., L. S., Nicot N., Van Moer K., Hoffmann L., Renaut J., Hausman J., Evers D. (2008) Plant Science 175, 839-852 Rensink, W. A., Iobst, S., Hart, A., Stegalkina, S., Liu, J., and Buell, C. R. (2005) Funct Integr Genomics 5, 201-207 Sterck, L., Rombauts, S., Vandepoele, K., Rouze, P., and Van de Peer, Y. (2007) Curr Opin Plant Biol 10, 199-203 Wikstrom, N., Savolainen, V., and Chase, M. W. (2001) Proc Biol Sci 268, 2211-2220 108 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. Rabbani, M. A., Maruyama, K., Abe, H., Khan, M. A., Katsura, K., Ito, Y., Yoshiwara, K., Seki, M., Shinozaki, K., and Yamaguchi-Shinozaki, K. (2003) Plant Physiol 133, 1755-1767 Wray, G. A. (2007) Nat Rev Genet 8, 206-216 Borneman, A. R, Gianoulis, T. A., Zhang, Z. D., Yu, H., Rozowsky, J., Seringhaus, M. R. ,Wang, L. Y. ,Gerstein, M., and Snyder, M. (2007) Science 317, 815-819 Maruyama, K., Sakuma, Y., Kasuga, M., Ito, Y., Seki, M., Goda, H., Shimada, Y., Yoshida, S., Shinozaki, K., and Yamaguchi-Shinozaki, K. (2004) Plant J 38, 982-993 Knox, A. K., Li, C., Vagujfalvi, A., Galiba, G., Stockinger, E. J., and Dubcovsky, J. (2008) Plant Mol Bio! 67, 257-270 Tondelli, A., Francia, E., Barabaschi, D., Aprile, A., Skinner, J. S., Stockinger, E. J., Stanca, A. M., and Pecchioni, N. (2006) T heor Appl Genet 112, 445-454 Stone, J. M., Palta, J. P., Bamberg, J. B., Weiss, L. S., and Harbage, J. F. (1993) Proc Natl Acad Sci U S A 90, 7869-7873 Vega, S., Rio, A., Geunhwa, J., Bamberg, J., and Palta, J. (2003) Amer J of Potato Res 80, 359-369 Eisen, M. B., Spellman, P. T., Brown, P. 0., and Botstein, D. (1998) Proc Natl Acad Sci U S A 95, 14863-14868 Hanada, K., Zou, C., Lehti—Shiu, M. D., Shinozaki, K., and Shiu, S. H. (2008) Plant Physio! 148, 993-1003 Hotta, C. T., Gardner, M. J., Hubbard, K. E., Baek, S. J., Dalchau, N., Suhita, D., Dodd, A. N., and Webb, A. A. (2007) Plant Cell Environ 30, 333- 349 Kagaya, Y., Ohmiya, K., and Hattori, T. (1999) Nucleic Acids Res 27, 470- 478 Rizhsky, L., Davletova, S., Liang, H., and Mittler, R. (2004) J Biol Chem 279, 11736-11743 Gutierrez, R. A., Ewing, R. M., Cherry, J. M., and Green, P. J. (2002) Proc NatlAcad Sci USA 99, 11513-11518 109 94. Farre, E. M., Harmer, S. L., Harmon, F. G., Yanovsky, M. J., and Kay, S. A. (2005) Curr Biol 15, 47-54 110