Abstract
Many plants, including Arabidopsis, show increased resistance to freezing after they have been exposed to low nonfreezing temperatures. This response, termed cold acclimation, is associated with the induction of COR (cold-regulated) genes mediated by the C-repeat/drought-responsive element (CRT/DRE) DNA regulatory element. Increased expression of ArabidopsisCBF1, a transcriptional activator that binds to the CRT/DRE sequence, induced COR gene expression and increased the freezing tolerance of nonacclimated Arabidopsis plants. We conclude that CBF1 is a likely regulator of the cold acclimation response, controlling the level of COR gene expression, which in turn promotes tolerance to freezing.
Studies of the molecular basis of plant tolerance to freezing have focused primarily on the cold acclimation response, the process by which plants increase their tolerance to freezing in response to low nonfreezing temperatures (1). Cold acclimation is associated with biochemical and physiological changes and alterations in gene expression (1,2). Studies of genes stimulated by low temperature have revealed that many, including the Arabidopsis COR genes, encode hydrophilic polypeptides that potentially promote tolerance to freezing (1-3). Indeed, constitutive expression of COR15a(which encodes the chloroplast-targeted polypeptide COR15am) in transgenic Arabidopsis plants improves the freezing tolerance of chloroplasts frozen in situ and of protoplasts frozen in vitro (4). Unlike cold acclimation, however,COR15a expression has no discernible effect on the survival of frozen plants (2, 5).
Genetic analyses indicate that multiple genes are involved in cold acclimation in plants (6). Several COR genes are coordinately stimulated along with COR15a in response to low temperature (2, 7), which suggests thatCOR15a might act in concert with other COR genes to enhance tolerance to freezing in plants. If so, expression of the entire battery of COR genes would have a greater effect on freezing tolerance than COR15a expression alone. To test this hypothesis, we attempted to induce expression of theCOR gene “regulon” with the Arabidopsistranscriptional activator CBF1 (CRT/DRE binding factor 1) (8), a putative COR gene regulator. CBF1 binds to the cis-acting CRT (C-repeat)/DRE (drought-responsive element) sequence (9, 10), a DNA regulatory element that stimulates transcription in response to both low temperature and water deficit (9). The element is present in the promoters of multipleCOR genes including COR15a, COR78(also known as RD29A and LTI78), and COR6.6(10-12). Expression of CBF1 in yeast (Saccharomyces cerevisiae) activates expression of reporter genes that have the CRT/DRE as an upstream activator sequence (8).
We created transgenic Arabidopsis plants that overexpress CBF1 by placing a cDNA encoding CBF1 under the control of the strong cauliflower mosaic virus (CaMV) 35SRNA promoter and transforming the chimeric gene intoArabidopsis ecotype RLD plants (13). Initial screening gave rise to two transgenic lines, A6 and B16, that accumulated CBF1 transcripts at high concentrations. Southern blot analysis indicated that the A6 plants had a single DNA insert and the B16 plants had multiple inserts. Examination of fourth generation homozygous A6 and B16 plants indicated that amounts ofCBF1 transcript were higher in nonacclimated A6 and B16 plants than they were in nonacclimated RLD plants (Fig.1A). Quantities of CBF1transcript were greater in the A6 plants than in the B16 plants (Fig.1A).
Expression of CBF1and COR genes in RLD and transgenic Arabidopsisplants. (A) CBF1 and CORtranscripts. Leaves from nonacclimated and 3-day cold-acclimated plants (20) were harvested and total RNA was prepared and analyzed for CBF1 and COR transcripts by RNA blot analysis with 32P-radiolabeled probes (21). The autoradiograms for CBF1 resulted from 3-day film exposure and those for COR6.6 and COR15a were from a 3-hour exposure (the 32P-radiolabeled probes were of similar specific activity). (B) COR15am proteins. Total soluble protein (100 μg) was prepared from leaves of the nonacclimated RLD (RLDw), 4-day cold-acclimated RLD (RLDc4d), 7-day cold-acclimated RLD (RLDc7d), and nonacclimated A6 and B16 plants; the amounts of COR15am were determined by immunoblot analysis with antiserum raised against the COR15am polypeptide (22). No reacting bands were observed with preimmune serum.
CBF1 overexpression induced COR gene expression without a low-temperature stimulus (Fig. 1A). Specifically, greater than normal amounts of COR6.6, COR15a,COR47, and COR78 transcripts were detected in nonacclimated A6 and B16 plants. In nonacclimated A6 plants,COR transcript concentrations approximated those found in cold-acclimated RLD plants. In nonacclimated B16 plants, they were less than in cold-acclimated RLD plants. Immunoblot analysis indicated that the amounts of the COR15am (Fig. 1B) and COR6.6 polypeptides were also elevated in the A6 and B16 plants, with a higher level of expression in A6 plants. We were unable to identify the CBF1 protein in either RLD or transgenic plants (5). Overexpression of CBF1 did not affect transcript concentrations of eIF4A(eukaryotic initiation factor 4A) (14), a constitutively expressed gene that is not responsive to low temperature (Fig. 1A), and had no obvious effects on plant growth and development.
Two additional transgenic lines, K16 and 1-11, that overexpressCBF1 have recently been identified. Northern blot analysis of nonacclimated T2 generation plants indicated that, in both of these lines, COR gene expression is also higher than that in nonacclimated RLD plants.
CBF1 overexpression increased the tolerance of plants to freezing (Fig. 2), as determined by the electrolyte leakage test (15). Detached leaves were frozen to various subzero temperatures and, after thawing, cellular damage (due to freeze-induced membrane lesions) was estimated by measuring ion leakage from the tissues. Leaves from nonacclimated A6 and B16 plants were more tolerant to freezing than those from nonacclimated RLD plants (Fig. 2). The freezing tolerance of leaves from nonacclimated A6 plants exceeded that of leaves from nonacclimated B16 plants (Fig. 2A), which had lower levels of CBF1 andCOR gene expression (Fig. 1A). T8 transgenic plants (4), which constitutively express only COR15a(under control of the CaMV 35S RNA promoter) (Fig. 1A), were less freezing tolerant than A6 plants (Fig. 2B).
Freezing tolerance of leaves from RLD and transgenic Arabidopsis plants. Leaves from nonacclimated RLD (RLDw) plants, 10-day cold-acclimated RLD (RLDc) plants, and nonacclimated A6, B16, and T8 plants were frozen at the indicated temperatures and the extent of cellular damage was estimated by measuring electrolyte leakage (23). Error bars indicate standard deviations.
A comparison of EL50 values (the freezing temperature that results in release of 50% of tissue electrolytes) of leaves from RLD, A6, B16, and T8 plants is presented in Table1. Data from multiple experiments indicate that the freezing tolerance of leaves from nonacclimated A6 and B16 plants was greater than that of leaves from nonacclimated RLD and T8 plants and that leaves from nonacclimated A6 plants were more freezing tolerant than leaves from nonacclimated B16 plants.
Comparison of EL50 values of leaves from RLD and transgenic Arabidopsis plants. EL50values were calculated and compared by analysis of variance (25). EL50 ± SE (n) are presented on the diagonal line for leaves from nonacclimated RLD (RLDw), cold-acclimated (7 to 10 days) RLD (RLDc), and nonacclimated A6, B16, and T8 plants. P values for comparisons of EL50values are indicated in the intersecting cells.
The enhancement of freezing tolerance in A6 plants was apparent in whole plant survival tests (Fig. 3). Nonacclimated A6 plants displayed variable, but greater, freezing tolerance than nonacclimated RLD plants (Fig. 3). No difference in plant survival was detected between nonacclimated B16 and RLD plants and nonacclimated T8 and RLD plants.
Freezing survival of RLD and A6Arabidopsis plants. Nonacclimated (Warm) RLD and A6 plants and 5-day cold-acclimated (Cold) RLD plants were frozen at −5°C for 2 days and then returned to a growth chamber at 22°C (24). A photograph of the plants after 7 days of regrowth is shown.
Our results demonstrate that constitutive overexpression of theArabidopsis transcriptional activator CBF1 induces expression of Arabidopsis COR genes and increases the freezing tolerance of nonacclimated plants. These results are consistent with CBF1 having a role in regulating COR gene expression and further link the COR genes to plant cold acclimation. The increase in freezing tolerance brought about by expressing the battery of CRT/DRE-regulated COR genes was greater than that brought about by overexpressing COR15aalone, which implicates additional COR genes in freezing tolerance. Whether CRT/DRE-containing COR genes are involved in bringing about the full array of biochemical and physiological changes that occur with cold acclimation (1, 2) remains to be determined.
Freezing temperatures greatly limit the geographical distribution of native and cultivated plants and often cause severe losses in agricultural productivity (16). Traditional plant breeding approaches have met with limited success in improving the freezing tolerance of agronomic plants (6). The freezing tolerance of the best wheat varieties today is essentially the same as the most freezing-tolerant varieties developed in the early part of this century. Biotechnology, however, may offer new strategies. Here we show that the freezing tolerance of nonacclimated Arabidopsisplants is enhanced by increasing the expression of theArabidopsis regulatory gene CBF1. The CRT/DRE DNA regulatory element we have targeted here is not limited toArabidopsis (17) and thus may provide a way to improve the freezing tolerance of crop plants.
↵* To whom correspondence should be addressed. E-mail: thomash6{at}pilot.msu.edu
