| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
First published online July 25, 2002; 10.1105/tpc.003483 American Society of Plant Biologists
Arabidopsis Transcriptome Profiling Indicates That Multiple Regulatory Pathways Are Activated during Cold Acclimation in Addition to the CBF Cold Response PathwayMichigan State University–Department of Energy Plant Research Laboratory and Department of Crop and Soil Sciences, Michigan State University, East Lansing, Michigan 48824-1312 1 To whom correspondence should be addressed. E-mail thomash6{at}msu.edu; fax 517-353-9168
Many plants, including Arabidopsis, increase in freezing tolerance in response to low, nonfreezing temperatures, a phenomenon known as cold acclimation. Previous studies established that cold acclimation involves rapid expression of the CBF transcriptional activators (also known as DREB1 proteins) in response to low temperature followed by induction of the CBF regulon (CBF-targeted genes), which contributes to an increase in freezing tolerance. Here, we present the results of transcriptome-profiling experiments indicating the existence of multiple low-temperature regulatory pathways in addition to the CBF cold response pathway. The transcript levels of  8000 genes were determined at multiple times after plants were transferred from warm to cold temperature and in warm-grown plants that constitutively expressed CBF1, CBF2, or CBF3. A total of 306 genes were identified as being cold responsive, with transcripts for 218 genes increasing and those for 88 genes decreasing threefold or more at one or more time points during the 7-day experiment. These results indicate that extensive downregulation of gene expression occurs during cold acclimation. Of the cold-responsive genes, 48 encode known or putative transcription factors. Two of these, RAP2.1 and RAP2.6, were activated by CBF expression and thus presumably control subregulons of the CBF regulon. Transcriptome comparisons indicated that only 12% of the cold-responsive genes are certain members of the CBF regulon. Moreover, at least 28% of the cold-responsive genes were not regulated by the CBF transcription factors, including 15 encoding known or putative transcription factors, indicating that these cold-responsive genes are members of different low-temperature regulons. Significantly, CBF expression at warm temperatures repressed the expression of eight genes that also were downregulated by low temperature, indicating that in addition to gene induction, gene repression is likely to play an integral role in cold acclimation.
Many plants increase in freezing tolerance in response to low, nonfreezing temperatures, a phenomenon known as cold acclimation. Interest in understanding the molecular basis of cold acclimation is driven by both a desire to understand the mechanisms that plants have evolved to tolerate environmental stresses and the prospect that such knowledge might provide new strategies to improve the environmental stress tolerance of agriculturally important plants. Toward these ends, a number of important insights have been gained.
One is that the membrane systems of the cell are a primary site of freeze-induced injury and that this injury results largely from the severe dehydration associated with freezing (Levitt, 1980
Multiple mechanisms appear to be involved, including changes in lipid composition and the accumulation of compatible solutes with cryoprotective properties, such as Suc and Pro (Thomashow, 1999
A recent important insight into the cold acclimation process was the discovery of the Arabidopsis CBF cold response pathway (Shinozaki and Yamaguchi-Shinozaki, 2000
Three members of the CBF/DREB1 family, CBF1, CBF2, and CBF3 (or DREB1b, DREB1c, and DREB1a, respectively), are induced within 15 min of transferring plants to cold temperatures, followed at
Overexpression of the CBF/DREB1 proteins in Arabidopsis results in an increase in freezing tolerance at the whole-plant level in both nonacclimated and cold-acclimated plants (Jaglo-Ottosen et al., 1998 The CBF cold response pathway is one low-temperature gene network that contributes to cold tolerance in Arabidopsis. A fundamental question is whether other low-temperature gene networks contribute significantly to freezing tolerance or other aspects of growth and development at low temperature. To address this issue, we used Affymetrix GeneChip arrays to analyze the Arabidopsis transcriptome at multiple times after transfer of plants from warm to cold temperature and in warm-grown plants that constitutively express CBF1, CBF2, or CBF3. The results indicate that a dynamic series of changes in the Arabidopsis transcriptome is set in motion upon transfer of plants from warm to cold temperature, including changes in cold-regulatory gene networks in addition to the CBF cold response pathway. The results identify >250 newly described cold-responsive genes that offer explanations for certain biochemical changes that occur during cold acclimation, identify candidate polypeptides with roles in cold tolerance, and indicate that gene repression may play an integral role in the cold acclimation response.
Transfer of Arabidopsis to Low Temperature Produces Waves of Changes in the Transcriptome Affymetrix GeneChip arrays, which contain 8297 DNA oligonucleotide probe sets representing 8000 genes per chip, were used to assay changes in the Arabidopsis transcriptome in response to exposure of plants to low temperature. Transcript levels were analyzed in duplicate biological samples harvested just before plants were transferred from 22 to 4°C (warm sample) and then at 0.5, 1, 4, 8, and 24 h and 7 days after transfer. Fold change values were calculated for the duplicate samples harvested at 4°C compared with each of the two warm samples, generating four comparisons for each time point.
A gene was designated as being upregulated at a given time point if the signal intensity was greater than background ("present") for both duplicate cold samples, if there was a difference call of "increase" for all four comparisons, and if the fold increase value was
Control experiments indicated that using a cutoff of threefold would effectively exclude the possibility of a gene being inappropriately designated as cold responsive as a result of a technical error. When a single RNA sample from warm-grown plants was used to prepare two probes that were hybridized to two different GeneChips, only 3 of the 8297 total probe sets were found to have a difference call of increase or decrease and a fold change value of Using these criteria, 330 probe sets were found to represent cold-responsive genes, corresponding to 4% of the total probe sets. The number of cold-responsive probe sets increased to a maximum of 182 at 24 h and then decreased to 97 at 7 days (Figure 1) . Analysis of these data indicated that 306 genes (some genes are represented by more than one probe set) were cold responsive at a minimum of one time point during the course of the experiment (Figure 2 and supplemental data online). Of these, 218 genes were scored as being upregulated in response to low temperature, and another 88 were scored as being downregulated (Figure 2).
As expected, among the cold-regulated genes were members of the CBF cold response pathway. In particular, the transcript levels for CBF1, CBF2, and CBF3 increased within the first hour of plants being exposed to low temperature, followed closely (within 4 h) by expression of known CBF target genes, including COR6.6, ERD10, COR47, and COR78 (Figure 3) . These results indicated that the tissue samples used in this experiment underwent a typical cold acclimation response and that the GeneChips replicated results obtained previously by RNA gel blot analysis (Gilmour et al., 1998  ; Liu et al., 1998; Shinwari et al., 1998).
Hierarchical clustering of the entire set of 218 upregulated genes revealed that blocks of genes were induced in multiple waves after transfer of plants to low temperature (Figure 4A) . In addition, it was evident that some genes were expressed transiently, whereas others were induced and remained activated for the entire 7-day experiment. Overall, the kinetic pattern observed did not fit a simple two-step cascade profile of the CBF cold response pathway—rapid cold induction of the CBF transcriptional activators (during the first hour) followed by expression of the CBF regulon (by 4 h)—suggesting that multiple regulatory pathways were activated in response to the temperature decrease. A similar picture emerged with the set of 88 downregulated genes (Figure 4B).
Long-Term Upregulated Genes Encode Multiple Transcription Factors and Proteins with Potential Roles in Freezing Tolerance Transfer of plants from warm to cold temperature triggers the cold acclimation response, which includes expression of COR and other genes that remain upregulated for extended periods at low temperature. Of the 218 genes that were determined to be upregulated in response to cold (i.e., that were upregulated at least 3-fold at one or more time points during the course of the experiment), 64 genes remained upregulated at least 2.5-fold at 7 days. These were considered long-term upregulated genes (Figures 5A and 6 and supplemental data online). The lower threshold of 2.5-fold was used for this specific designation because we considered it inappropriate to assert that an upregulated gene showing expression at this level at 7 days had returned to a prestress level of expression. A search of the literature revealed that 50 of these long-term upregulated genes had not been reported previously to be cold regulated in Arabidopsis (supplemental data online).
Hierarchical clustering indicated that the 64 long-term upregulated genes were induced at different times after transfer of plants to low temperature (Figures 5A and 6). This was visualized most easily using a "binary" hierarchical clustering format (Figure 6). In this case, genes that were upregulated 2.5-fold or more were considered to be induced, and time points at which this occurred were colored red, whereas those that were not changed by 2.5-fold were considered unchanged, and these time points were colored black (thus, all time 0 values were black). The presentation reveals that blocks of genes were induced at each time point after transfer of plants to low temperature, suggesting that multiple regulatory pathways likely were involved in their induction.
Consistent with this notion was the finding that in addition to CBF2, eight other long-term cold-responsive genes that encoded either known or putative transcription factors were induced at various times during the course of the experiment (Figure 6 and supplemental data online). Two of these genes were induced rapidly in response to low temperature in parallel with CBF2, RAV1 (Kagaya et al., 1999
Six additional long-term upregulated genes were found to encode known or putative transcription factors: a second putative zinc finger protein (At4g38960; Mayer et al., 1999 The fact that these six genes were induced after the initial wave of CBF2, RAV1, and ZAT12 induction raised the possibility that one or more of them might be induced by one of these transcription factors. Inspection of the promoter region of RAP2.1 indicated that it contained two copies of the CCGAC core sequence of the CRT/DRE elements, suggesting that it might be a target of the CBF activators. Indeed, RNA gel blot analysis indicated that the transcript levels of RAP2.1 did not increase until 4 to 8 h after transfer of plants to low temperature (Figure 7A) and that they were increased in transgenic Arabidopsis plants that constitutively overexpressed CBF1, CBF2, or CBF3 (Figure 7B). Thus, the CBF regulon presumably includes a "subregulon" controlled by RAP2.1.
In addition to long-term upregulated genes encoding transcription factors, 55 genes encoded proteins with diverse known or proposed functions (Figure 6 and supplemental data online). The largest group encoded COR/LEA proteins, polypeptides thought to play roles in cryoprotection (Thomashow, 1999 All of these COR/LEA proteins have different sequences, but they have in common the biochemical property of being highly hydrophilic. Interestingly, 5 of the 11 long-term upregulated unknown proteins also encoded polypeptides that were highly hydrophilic (Figure 8) . We designated these COR/LEA-like polypeptides COR8.5 (At2g23120), COR8.6 (At2g05380), COR12 (At4g33550), COR18 (At2g43060), and COR28 (At4g33980), which encode polypeptides of 8.5, 8.6, 12.3, 17.8, and 28.1 kD, respectively. The N-terminal end of COR12 is predicted to encode a signal sequence that would result in the secretion of a mature hydrophilic polypeptide of 10.1 kD.
Sugars, including Suc and raffinose, accumulate during cold acclimation in Arabidopsis (Wanner and Junttila, 1999 ; Gilmour et al., 2000). Thus, genes that encode proteins with roles in sugar metabolism might be expected to be cold responsive. We found previously that transcripts encoding Suc synthase accumulated in response to low temperature (Gilmour et al., 2000), a finding that is confirmed here (Figure 6 and supplemental data online). Additionally, transcripts for three different genes encoding putative galactinol synthases (At1g09350, At2g47180, and At1g60470), the enzyme that catalyzes the first committed step in the synthesis of raffinose, were found to accumulate in response to low temperature. In one case, induction was >160-fold at 24 h (Figure 6 and supplemental data online). Transcripts encoding a putative sugar transporter also accumulated in response to low temperature.
Among the most rapid and highly induced genes was that encoding ELIP1 (Early Light-Induced Protein 1) (Figure 6 and supplemental data online). ELIP1, as well as ELIP2, are nuclear-encoded thylakoid membrane proteins that are expressed in response to light stress (Moscovici-Kadouri and Chamovitz, 1997
Finally, an intriguing finding was that transcripts for GIGANTEA (GI) were found to increase 5- to 10-fold in response to low temperature (Figure 6 and supplemental data online). The GI gene encodes a protein that has no homology with any proteins of known function in the databases (Fowler et al., 1999
Transient Upregulated Genes Encode Proteins with a Wide Range of Functions and Suggest a Period of Oxidative Stress after Transfer of Plants to Low Temperature
Indeed, of the 156 transiently cold-induced genes, 34 (22%) corresponded to known or putative transcription factors (Figure 9 and supplemental data online). In addition, 16 genes (10%) encoded known or putative proteins involved in signal transduction or cellular communication, including response regulators, protein kinases, and phosphatases. In total, 33% of the transiently expressed genes potentially had roles in gene regulation.
The transient nature of the changes in transcript levels suggested that the abrupt temperature decrease might have resulted in a short-lived "shock" response followed by an adjustment to the new environmental conditions. Indeed, low temperature can cause a decrease in the turnover rate of photosystem II components, causing an increase in photosystem II excitation pressure, or "excess excitation energy," and the generation of damaging reactive oxygen species, including hydrogen peroxide (Huner et al., 1998
An indication that such a response occurred in our experiments was the fact that among the transiently expressed genes were three known or putative glutathione S-transferases, which are known to be involved in the detoxification of toxic metabolites arising from oxidative damage caused by excess excitation energy (Marrs, 1996
The production of ethylene also has been associated with cold stress (Ciardi et al., 1997
However, two other known ethylene-inducible genes, AtERF-1 and basic chitinase, were not induced (these genes are represented by probe sets on the GeneChip). This and the fact that cold induction of AtERF-4 and AtERF-5 can occur independent of ethylene (Fujimoto et al., 2000 Of the total transiently expressed genes, 24% encoded unknown proteins (Figure 9 and supplemental data online). Of these, 11 were novel polypeptides that, like LEA/COR proteins, were unusually hydrophilic. The probe sets corresponding to these genes were At2g41010, At1g23710, At2g28400, At4g35320, At2g22860, At4g04330, At1g10410, At2g46970, At2g32210, At2g36220, and At1g11960 (supplemental data online). In addition, transcript levels for genes encoding three previously described LEA/COR proteins (LEA M17, LEA D113, and RD29B) were found to be expressed transiently (Figure 9 and supplemental data online). These results indicate that low temperature generated a short-lived signal that induces the expression of LEA/COR and LEA/COR-like genes through a pathway that is independent of the CBF cold response pathway.
Genes Encoding Proteins with Diverse Functions Are Downregulated during Cold Acclimation
Like the genes that were upregulated in response to low temperature, the downregulated genes encoded proteins with a wide range of functions, including transcription, signaling, cell wall biogenesis, and defense. Four of the long-term downregulated transcripts encoded proteins with known or predicted roles in energy production: two light-harvesting proteins, Lhca2*1 and Lhcb4*3; a putative 5-kD photosystem II protein; and a ferredoxin precursor. The downregulation of these genes may have resulted from the decreased light levels that the plants were exposed to during the cold treatment. However, it has been reported that the transfer of warm-grown plants to chilling temperature (4°C) leads to a rapid inhibition of photosynthesis followed by a reduction in transcript levels for genes encoding photosynthetic proteins (Krapp and Stitt, 1995
A striking difference between the downregulated and upregulated genes was that few genes were downregulated during the first 4 h of plant exposure to cold temperature. The full significance of this apparent delay is uncertain, because any decrease in transcript levels must take into account the turnover rate of the transcript. Given that the half-life of the average plant transcript is on the order of several hours (Abler and Green, 1996
Genes Regulated by the CBF Cold Response Pathway Are a Subset of Total Cold-Responsive Genes
A gene was designated as being a member of the CBF regulon if the signal intensity was greater than background (present) for all three transgenic lines, if there was a difference call of increase for all six comparisons, and if the fold increase value was Of the cold-induced genes, 60 were not upregulated in any of the CBF-expressing plants; thus, they were designated as being independent of the CBF cold response pathway (Figure 10A) . Of these, the large majority, 50 genes (or 83%), were upregulated transiently upon exposure to low temperature; the remaining 10 (17%) were long-term upregulated (Figure 10B).
Analysis of the CBF-expressing plants indicated that 41 genes were upregulated in all three transgenic lines (Figure 10A and supplemental data online). Included in this group were genes reported previously as being CBF targets, such as COR6.6, COR78, COR47, P5CSb, and ERD10 (Jaglo-Ottosen et al., 1998 ; Gilmour et al., 2000; Seki et al., 2001). Of the 41 CBF-regulated genes, 30 were found to be upregulated in response to low temperature in the experiments described above (Figure 10A and supplemental data online). Thus, these 30 genes were considered genuine members of the CBF regulon. The 11 CBF-responsive genes that were not cold induced presumably were downstream consequences of CBF expression related to the altered growth phenotypes displayed by plants that overexpressed CBF (Liu et al., 1998; Kasuga et al., 1999; Gilmour et al., 2000). Of the 30 genes that were members of the CBF regulon, 19 were in the long-term cold-responsive group and the remaining 11 were induced transiently (Figure 10B). Among these genes was RAP2.6, which encodes an AP2 protein and thus is another likely transcription factor that controls the expression of a subregulon of the CBF regulon. Finally, of the 88 genes found to be downregulated during cold acclimation, 8 also were downregulated by CBF overexpression: At4g22690, At3g57260, At1g75040, At2g14560, At1g21270, At2g43570, At1g69490, and At4g14400 (Figures 10C and supplemental data online). These results indicate that the CBF cold-response pathway not only acts to activate gene expression but also is involved in repressing the expression of certain genes.
Central goals in cold acclimation research include identifying cold-responsive genes, determining how they are regulated, and understanding their roles in plant life at low temperature. Most of the studies to date have been with individual or small numbers of genes. However, with the development of genomic technologies, including methods for gene expression profiling, these issues can be addressed on a more global scale.
Seki et al. (2001)
This report expands on the findings of Seki et al. (2001)
Both increases and decreases in transcript levels occurred, some of which were transient and others of which were long lived, being sustained for up to 7 days of cold treatment. The affected genes encompassed a wide range of functions, including transcription, signaling, metabolism, cellular biogenesis, and cell rescue and defense. The results indicate that the expression of as much as 4% of the genome may be affected by exposing plants to low temperature. Thus, if the probe sets used in our experiments generally are representative of the entire genome, then
Fifteen of the 19 genes that Seki et al. (2001)
Four additional novel cold-regulated genes identified by Seki et al. (2001)
Finally, four of the genes designated as being cold regulated by Seki et al. (2001)
Previous studies have established that the CBF cold response pathway is an integral component of the cold acclimation response (Shinozaki and Yamaguchi-Shinozaki, 2000 Our results provide direct evidence for cold-regulatory pathways in addition to the CBF cold response pathway. Of the 306 cold-responsive genes, 106 were affected over the long term (i.e., transcript levels were upregulated or downregulated at least 2.5-fold at 7 days). Among these genes were members of the CBF cold response pathway. However, in addition, eight other genes encoding known or putative transcription factors were found to be long-term cold responsive: ZAT12, RAV1, AtMYB73, ATHB-12, H-protein binding factor 2a, RAP2.1, a zinc finger protein (At4g38960), and RAP2.7 (Figure 6 and supplemental data online).
One of these, RAP2.1, proved to be a target of CBF transcription factors (Figure 7B)
and presumably regulates the expression of a subregulon of genes within the larger CBF regulon. However, two of the transcription factors, ZAT12 and RAV1, were found to be induced in parallel with the CBF transcriptional activators (i.e., transcript levels were increased more than threefold at 30 min). It is possible that the low-temperature regulation of ZAT12 and RAV1 involves the action of the same regulatory proteins that activate CBF expression. Gilmour et al. (1998)
Regardless, the parallel induction of CBF2, ZAT12, and RAV1, well before the induction of known CBF target genes such as COR47, COR6.6, and COR78, argues against their being members of the CBF regulon. Consistent with this notion was the fact that ZAT12 transcript levels were unaffected in CBF1-, CBF2-, or CBF3-overexpressing plants. A comparison of the transcriptomes of CBF-overexpressing plants and control plants indicated that at least 60 cold-induced genes are independent of the CBF cold response pathway, including at least 15 transcription factors. Investigations to date have focused on genes that are upregulated in response to low temperature. The results presented here, however, indicate that downregulation of gene expression also may be an important component of the adaptation to low temperature. Indeed, the downregulation response was extensive. Transcript levels for 88 genes were found to decrease either transiently or long term in response to low temperature. These genes, like those that were upregulated, encompassed a wide range of functions, including transcription, signaling, cell wall biogenesis, defense, and photosynthesis, almost all of which (to the best of our knowledge) have not been shown previously to be cold repressed. Some of the changes in transcript levels, especially those associated with photosynthesis, might have been caused by the lower light conditions used during the cold treatment. However, this would not appear to be true for all of the down-regulated genes. In particular, eight genes that were downregulated in response to low temperature also were downregulated at warm temperature in response to the constitutive expression of CBF1, CBF2, and CBF3. These data indicate that the CBF cold response pathway includes downregulation of the expression of certain genes and suggest genes whose expression might be incompatible with the enhancement of freezing tolerance. Previous studies have identified 14 genes as members of the CBF regulon. Here, we expand that number to 45 (37 upregulated and 8 downregulated). Among the newly described members of the CBF regulon are genes encoding two putative transcription factors, RAP2.1 and RAP2.6, that presumably control the expression of subregulons of the larger CBF regulon. In this regard, it is interesting that a number of the genes that are activated by CBF expression do not have the core CCGAC sequence of the CRT/DRE element within 1 kb of the start of transcription. These are candidate members of a CBF subregulon controlled by RAP2.1, RAP2.6, or a yet-to-be-discovered transcription factor(s) induced by the CBF activators.
Other new members of the CBF regulon include genes that encode a putative sugar transporter, water channel proteins, and a new hydrophilic polypeptide that potentially acts as a cryoprotectant. In addition, there are three genes that putatively encode galactinol synthase, which catalyzes the first committed step in raffinose synthesis. In fact, while we were preparing this article, Taji et al. (2002) Given that we sampled approximately one-third of the genome in our experiments, there could be as many as 100 genes that are part of the CBF regulon. Again, this is probably an underestimate because the criteria used here to assign a gene to the CBF regulon included it being induced threefold or more in each CBF overexpression line. However, it is not known whether CBF1, CBF2, and CBF3 control completely overlapping sets of genes. We observed differences in the set of genes that were upregulated in the CBF1, CBF2, and CBF3 transgenic plants tested here. However, we do not know whether these differences were attributable to biological variation, differences in the level of CBF expression, or bona fide differences in the activities of the CBF proteins. Additional experiments will be required to address this issue definitively.
One intriguing finding is that among the 280 newly described cold-responsive genes was GI. GI transcript levels increased fivefold to eightfold after 24 h of cold treatment and remained increased after 7 days. GI is a novel protein with roles in the promotion of flowering by photoperiod and circadian clock function (Fowler et al., 1999
More than half of the cold-responsive genes were upregulated or downregulated transiently in response to low temperature. It seems likely that the expression of a significant number of these genes was caused by the abrupt change in temperature used in our experiments. We transferred plants directly from 22 to 4°C, a protocol that is used commonly in the study of cold acclimation. Transferring plants from warm to cold temperature in the light (as was done in our experiments) can result in excess light energy, which can lead to the production of hydrogen peroxide and other reactive oxygen species (Huner et al., 1998
A number of genes known to be responsive to hydrogen peroxide were upregulated transiently in our experiments, including glutathione S-transferases, indicating that the plants at least transiently experienced oxidative stress. Thus, some of the transient cold-responsive genes probably do not play direct roles in life at low temperature per se but instead play critical roles in enabling plants to adjust to quickly fluctuating environmental conditions, including protection against conditions that result in excess light (Huner et al., 1998
Finally, it is relevant that when Escherichia coli (and other bacteria) are subjected to an abrupt temperature decrease of The results of this transcriptome study demonstrate the highly complex nature of plant adaptation to low temperature. The results indicate that the expression of hundreds of genes is affected upon exposure of plants to low temperature and that this involves the action of multiple cold-regulatory pathways, including the activation of regulons within regulons. Through the further application of genomic approaches, it should be possible to construct a diagram of the low-temperature "gene circuitry" in plants and to determine the roles of the regulatory networks in cold tolerance.
Plant Material and Growth Conditions Arabidopsis thaliana ecotype Wassilewskija-2 and transgenic plants constitutively expressing CBF1 (G5, G6, and G26), CBF2 (E2, E8, and E24; S.J. Gilmour and M.F. Thomashow, unpublished results), or CBF3 (A28, A30, and A40; Gilmour et al., 2000 ) in the Wassilewskija-2 background were used in these experiments. Seeds were surface-sterilized and then spread on Petri plates containing Gamborg's B-5 medium (Life Technologies, Gaithersburg, MD) solidified with 0.8% phytagar (Life Technologies). Immediately after plating, the seeds were stratified for 4 days at 4°C to ensure uniform germination.
Plants were grown in controlled-environment chambers at 22°C under continuous illumination of 100 µmol·m-2·s-1 from cool-white fluorescent lights for 11 days. For cold treatments, plates containing the plants were transferred to 4°C under continuous light (20 to 40 µmol·m-2·s-1) as described (Gilmour et al., 1998 Upon request, all novel material described in this article will be made available in a timely manner for noncommercial research purposes. No restrictions or conditions will be placed on the use of any materials described in this article that would limit their use for noncommercial research purposes.
RNA Isolation and Probe Labeling
Affymetrix GeneChip Hybridization and Data Collection
Data Analysis The software also was used to determine whether expression of each gene was present or absent (absolute call) and whether the fold change value represented a genuine change in expression (difference call). Fold change values were calculated for each sample harvested at 4°C compared with each of the samples harvested before transfer to 4°C, generating four measurements for each gene at each time point during the cold treatment. Fold change values also were calculated for the three CBF transgenic samples compared with each of the wild-type samples, generating six measurements for each gene. Probe sets that met the following criteria were selected for further analysis. Those determined to be upregulated by the cold treatment were selected as having, at any time point, an absolute call of present in both cold samples, difference calls of increase, and fold changes of at least 3.0 for all four fold change comparisons. Those determined to be downregulated by the cold treatment were selected as having, at any time point, an absolute call of present in both warm samples, difference calls of decrease, and fold changes of at least -3.0 for all four fold change comparisons. Cold-regulated genes were determined to be long-term upregulated if, at the 7-day time point, both absolute calls were present and the four fold change comparisons were 2.5-fold or greater associated with four difference calls of increase. Similarly, cold-regulated genes were determined to be long-term downregulated if, at the 7-day time point, the fourfold change comparisons were -2.5-fold or greater associated with four difference calls of decrease. Genes upregulated by CBF overexpression were selected as having an absolute call of present in all three CBF transgenic samples, difference calls of increase, and a fold change of at least 3.0 for all six fold change comparisons between wild-type and CBF transgenic lines. Those determined to be downregulated by CBF expression were selected as having an absolute call of present in both wild-type samples, difference calls of decrease, and fold changes of at least -3.0 for all six fold change comparisons between wild-type and CBF transgenic lines. Cold-regulated genes that were independent of CBF expression were selected as having a difference call of no change in all six fold change comparisons between wild-type and CBF transgenic lines. Microsoft Access database management software (Microsoft, Redmond, WA) was used to manage and filter the GeneChip data, and Genespring 4.0.4 (Silicon Genetics, Redwood City, CA) was used to generate hierarchical gene clusters using a Pearson correlation (separation ratio of 0.5, minimum distance of 0.001). Before clustering, all data points that were associated with a difference call of no change were converted to 1. Binary hierarchical clusters were generated by altering the data used to generate the clusters such that data points that fulfilled a particular set of criteria were converted to 2, whereas all other points were assigned a value of 1.
The false-positive rate was calculated as the number of probe sets changed significantly as a percentage of probe sets on the array (Lipshutz et al., 1999
RNA Gel Blot Hybridization Analysis
We thank Dan Zarka for preparing the RNA samples used in the GeneChip experiments; Dan Zarka and Jonathon Vogel for performing the RNA gel blot analysis shown in Figure 7; Steve Triezenberg and Sarah Gilmour for critical comments on the manuscript; Michael Gieseg, Li Tian, and Annette Thelen for helpful discussions and advice; and Marlene Cameron for help with figures. GeneChip hybridization and data collection were performed at the Genomics Technology Support Facility (Michigan State University). This work was supported by grants from the National Science Foundation (Grant DBI 0110124), the Department of Energy (Grant DEFG0291ER20021), and the Michigan Agricultural Experiment Station.
Online version contains Web-only data.  Article, publication date, and citation information can be found at www.plantcell.org/cgi/doi/10.1105/tpc.003483. Received March 28, 2002; accepted May 7, 2002.
Abler, M.L., and Green, P.L. (1996). Control of mRNA stability in higher plants. Plant Mol. Biol. 32, 63–78.[CrossRef][ISI][Medline] Adamska, I. (1997). ELIPs: Light induced stress proteins. Physiol. Plant. 100, 794–805.[CrossRef] Baker, S.S., Wilhelm, K.S., and Thomashow, M.F. (1994). The 5'-region of Arabidopsis thaliana cor15a has cis-acting elements that confer cold-, drought- and ABA-regulated gene expression. Plant Mol. Biol. 24, 701–713.[CrossRef][ISI][Medline]
Beator, J., Pötter, E., and Kloppstech, K. (1992). Coordinated circadian regulation of mRNA levels for light-regulated genes and of the capacity for accumulation of chlorophyll protein complexes. Plant Physiol. 100, 1780–1786. Ciardi, J.A., Deikman, J., and Orzolek, M.D. (1997). Increased ethylene synthesis enhances chilling tolerance in tomato. Physiol. Plant. 101, 333–340.[CrossRef]
Desikan, R., A.-H.-Mackerness, S., Hancock, J.T., and Neill, S.J. (2001). Regulation of the Arabidopsis transcriptome by oxidative stress. Plant Physiol. 127, 159–172. Eimert, K., Wang, S.-M., Lue, W.-L., and Chen, J. (1995). Monogenic recessive mutations causing both late floral initiation and excess starch accumulation in Arabidopsis. Plant Cell 7, 1703–1712.[Abstract] Fowler, S., Lee, K., Onouchi, H., Samach, A., Richardson, K., Morris, B., Coupland, G., and Putterill, J. (1999). GIGANTEA: A circadian clock-controlled gene that regulates photoperiodic flowering in Arabidopsis and encodes a protein with several possible membrane-spanning domains. EMBO J. 18, 4679–4688.[CrossRef][ISI][Medline]
Fujimoto, S.Y., Ohta, M., Usui, A., Shinshi, H., and Ohme-Takagi, M. (2000). Arabidopsis ethylene-responsive element binding factors act as transcriptional activators or repressors of GCC box–mediated gene expression. Plant Cell 12, 393–404.
Gilmour, S.J., Sebolt, A.M., Salazar, M.P., Everard, J.D., and Thomashow, M.F. (2000). Overexpression of the Arabidopsis CBF3 transcriptional activator mimics multiple biochemical changes associated with cold acclimation. Plant Physiol. 124, 1854–1865. Gilmour, S.J., Zarka, D.G., Stockinger, E.J., Salazar, M.P., Houghton, J.M., and Thomashow, M.F. (1998). Low temperature regulation of the Arabidopsis CBF family of AP2 transcriptional activators as an early step in cold-induced COR gene expression. Plant J. 16, 433–442.[CrossRef][ISI][Medline]
Hajela, R.K., Horvath, D.P., Gilmour, S.J., and Thomashow, M.F. (1990). Molecular cloning and expression of cor (cold-regulated) genes in Arabidopsis thaliana. Plant Physiol. 93, 1246–1252.
Heddad, M., and Adamska, I. (2000). Light stress-regulated two-helix proteins in Arabidopsis thaliana related to the chlorophyll a/b-binding gene family. Proc. Natl. Acad. Sci. USA 97, 3741–3746. Heintzen, C., Melzer, S., Fischer, R., Kappeler, S., Apel, K., and Staiger, D. (1994). A light- and temperature-entrained circadian clock controls expression of transcripts encoding nuclear proteins with homology to RNA-binding proteins in meristematic tissue. Plant J. 5, 799–813.[CrossRef][ISI][Medline] Huner, P.A., Öquist, G., and Sarhan, F. (1998). Energy balance and acclimation to light and cold. Trends Plant Sci. 3, 224–230.[CrossRef] |
TOP
Abstract
INTRODUCTION
3 for all four comparisons. Similarly, a gene was designated as being downregulated at a given time point if the transcript levels produced a hybridization intensity greater than background for both duplicate warm samples, if there was a difference call of "decrease" for all four comparisons, and if the fold decrease value was 
-amylase (FL5-90/At4g17090), also was found by us to be cold regulated. 