Plant Physiol. (1993)103: 1047-1053

Regulation of Arabicfopsis thaliana 1. (Heyn) cor78 in Response to Low Temperature’ David P. Horvath*, Brett K. Mclarney, and Michael F. Thomashow*

Department of Crop and Soil Sciences (M.F.T.), Program in Genetics (D.P.H., M.F.T.), and Department of Biochemistry (B.K.M.), Michigan State University, East Lansing, Michigan 48824

Changes in gene expression occur during cold acclimation in a variety of plants including Arabidopsis thaliana 1. (Heyn). Here we examine the cold-regulated expression of A. thaliana cor78. The results of gene-fusion experiments confirm the finding of Yamaguchi-Shinozaki and Shinozaki ([1993] MOICen Cenet 2 3 6 331340) that the 5’ region of cor78 has cis-ading regulatory elements that can impart cold-regulated gene expression. Further, histochemical staining experiments indicated that this cold-regulatory element(s) was active at low temperature throughout much of the plant including leaves, stems, roots, flower petals, filaments, and sepals. Time-course experiments indicated that the adivity of the cor78 promoter in cold-acclimated plants was down-regulated quickly in response to noninducing temperatures and that the halflife of the cor78 transcripts was only about 40 min at normal growth temperature. Fusion of the entire transcribed region of cor78 to the cauliflower mosaic virus 35s promoter resulted in a chimeric gene that was constitutively expressed and displayed little if any posttranscriptional regulation in response to low temperature.

Many plants increase in freezing tolerance in response to low nonfreezing temperatures, a process known as cold acclimation. Over the past few years, it has been established that cold acclimation is associated with changes in gene expression (Guy, 1990; Thomashow, 1993).Detennining how these changes are regulated is now a major goal of cold acclimation research. Toward this end, genes that are responsive to low temperature have been isolated from a variety of plants and their regulation has been studied (Guy, 1990; Thomashow, 1993). The results indicate that the transcript levels for most cold-regulated genes increase dramatically within a few hours of transfening plants to low nonfreezing temperatures, that the levels of the transcripts remain high for as long as the plants are kept in the cold, and that the levels retum to basal values within a few hours of returning cold-acclimated plants to normal growth temperatures. It has also been found that many cold-regulated genes are responThis work was supported by grants from the U.S. Department

of Agriculture-National Research Initiative Competitive Grants Program (90-37264-5450and 92-37100-7531)and the Michigan Agri-

sive to both ABA and drought, but the relationships between cold-, ABA-, and drought-regulated expression of these genes is poorly understood. A number of cold-regulated genes have been isolated from Arabidopsis thaliana (Hajela et al., 1990; Kurkela and Franck, 1990; Nordin et al., 1991; Gilmour et al., 1992; Kurkela and Borg-Franck, 1992; Ling and Palva, 1992; Lin and Thomashow, 1992). One of these, altematively designated cor78 (Horvath, 1993), M78 (Nordin et al., 1993), and rd29A (Yamaguchi-Shinozaki and Shinozaki, 1993), encodes a 78-kD hydrophilic polypeptide of unknown function. cor78, like many other cold-regulated genes, is responsive to both ABA and drought (Hajela et al., 1990; Nordin et al., 1993; Yamaguchi-Shinozaki and Shinozaki, 1993). An examination of cor78 transcript levels in the abi mutants of A . thaliana (Koomneef et al., 1984)has suggested that cold-regulated expression of cor78 (as well as cor6.6 and cor47) does not require the action of ABA (Gilmour and Thomashow, 1991; Nordin et al., 1991). Thus, for these genes, cold- and ABA-regulated expression appear to involve separate signal transduction pathways. Further, Nordin et al. (1991) have examined drought-regulated expression of lti78 in the abi mutants and have concluded that the regulation of this gene by dehydration stress and ABA can occur through independent pathways. Here we present additional-analysesregarding the regulation of cor78 transcription. The results of gene-fusion experiments confirm the finding of Yamaguchi-Shinozaki and Shinozaki (1993) that the 5’ region of cor78 has regulatory elements that can impart cold-, ABA-, and drought-regulated gene expression. In addition, the data presented indicate that the promoter of cor78 is cold induced in most but not a11 plant tissues, that its activity is quickly down-regulated in response to normal growth temperature, and that the halflife of the cor78 transcript is only about 40 min at normal growth temperature. Finally, fusion of the entire transcribed region of cor78 to the constitutive CaMV 35s promoter resulted in the production of transcripts that accumulated to near equal levels in control and cold-treated plants. Abbreviations: CaMV, cauliflower mosaic virus; cor, cold-regulated; gus, gene encoding P-glucuronidase; GUS, 6-glucuronidase; Iti, low-temperature-induced;PCI, pheno1:chloroform:isoamyl alcohol (25:24:1);rd, responsive to desiccation; Tween-20, polyoxyethylenesorbitan monolaurate 20.

culture Experiment Station. Present address: U.S. Department of Agriculture-Agricultura1 Research Service, P.O. Box 5677,State University Station, Fargo, ND

58105. * Correspondingauthor; fax 1-517-353-5174. 1047

Horvath et al.

1048 MATERIALS AND METHODS Plant Material aiid Treatments

Arabidopsis thaliana L. (Heyn) ecotypes RLD and Landsberg erecta were generally grown in pots for 2 to 3 weeks in controlled environment chambers at 21OC under constant light (approximately 100 pmol m-' s-') as previously described (Gilmour et al., 1988). For most cold-regulation experiments, plantj were transferred to a controlled environment chamber aí: 5OC (constant light) for various lengths of time. For ABA, drought, and some cold treatments, plants were grown in F'etri plates on germination medium supplemented with 8-5 vitamins (Valvekens et al., 1988) at room temperature (about 25OC) with a 16-h day/8-h night schedule until they reached the four- to six-leaf stage (about 2 weeks). ABA treatment consisted of plants being sprayed to runoff with 100 p~ ABA (mixed isomers, Sigma) in 0.02% (v/v) Tween-20; control plants were sprayed with a solution of 0.02% (v/v) Tween-20. Treated plants were covered with the Petri dish lids to slow evaporation and placed on the lab bench for 4 h. Drought-stress treatments consisted of removing the lid of the Petri dish and allowing the plants to dry ovemight in the growth chamber, at which point they had become visibly wilted. In the time-course cold-regulation studies, A. thaliana ecotype RLD was grown on Petri plates as described above. Plates were then transferred to a cold room at 3OC with constant light (approximately50 pmol m-' s-') for 24 h arid then retumed to room temperature for various lengths of time. The half-life of the cor78 transcripts was calculated a:; described (Kabnick and Housman, 1988). lsolation and Arialysis of Nucleic Acids

Plant material was frozen in liquid nitrogen, pulverized using a mortar and pestle, and stored at -8OOC prior to extraction of nucleic acids. Total and poly(A') RNA were isolated as previously described (Hajela et al., 1990). Plasmid DNA was prepared from Escherichia coli using standard protocols (Sambrook et al., 1989). Northem transfers were prepared and hybridized with 32P-labeledprobes as described previously (Hajela et al., 1990). In Vitro Transcription/Translation Reactions

Plant Physiol. Vol. 103, 1,993

cor78 Constructs and Transgenic Plants

Details on the construction of 78P-gus, 78Pl-gus, and 35scor78 are presented elsewhere (Horvath, 1993). Briefly. the 78P-gus gene, canied on plasmid pDH78P, contains the promoter region of cor78 (bp -808 to +5 relative to the site of transcription initiation) fused to the gus reporter gerie in pBIIOl.1 (Jefferson, 1987). The 78PI-gus gene, camed on plasmid pDH78P1, contains cor78 bp -808 to +250 fused in frame to the gus gene in pBI101.2 (Jefferson, 1987). The 35scor78 gene, camed on pDH78T, contains the entire transcribed region of cor78 from bp -11 to +2715 fused to the CaMV 35s promoter (Benfey and Chua, 1990) in pEiI121 I (Jefferson, 1987). Each of the constructs was transformed into A . thaliana RLD using the Agrobacterium-mediated root transformation protocol (Valvekens et al., 1988) with minor modificeition: the kinetin was omitted from the callus-inducing medium and IAA was omitted from the shoot-inducing medium.lTwo independent transgenic lines were obtained carrylng 78F'-gus, three carrying 78PI-gus, and two canying 35s-cor78. For the experiments reported, each relevant transgenic line carrying a given construct was tested and gave similar resultsl The transgenic A. thaliana RLD line carrying the CaMVl 35s promoter fused to gus (35s-gus) was obtained from Stokes Baker (S. Baker, K. Wilhelm, M. Thomashow, unpubliished results). RNase Protection Assays

I

RNA probes were made by in vitro transcription using a Maxi-Script Kit (Ambion,Inc:) and 3ZP-labeledUTP according to the manufacturer's protocols. The template for the reaction was MluIISalI-digestedpDH15K that had been gel purified (Sambrook et al., 1989); pDHl5K contains cor78 bp -808 to +250 cloned into pBluescript SK- (Horvath, 1993). The transcribed fragment contained cor78 sequences from bp +250 to -113 downstream from the T7 promoter of pBluescript SK-. Full-length, radiolabeled transcripts were purified on ai1 8 M urea/5% acrylamide gel. Purified RNA probe (5 X 1041cpm) was hybridized to 10 pg of total RNA isolated from control and transgenic plants. RNase protection assays were performed using an RPA I1 Kit (Ambion, Inc.) according to the manufacturer's protocols. Protected fragments were reslolved on a 7 M urea/6% acrylamide sequencing gel and visualized by autoradiography. I I I

I

Histochemical Staining for GUS Activity

RNA corresponding to the sense strand of cor78 was synthesized from pBM1, a cDNA clone containing the entire coding sequence of cor78 inserted into pBluescript SK- (Stratagene) (B. McLamey, unpublished data). The plasmid was digested with BamHI and the linearized DNA was transcribed in vitro using T7 RNA polymerase (Promega). The resulting transcripts were extracted with PCI and precipitated with ethanol. Transcriptionproducts from these reactions and total RNA isolated from cold-treated and control A. thaliana RLD were translated in vitro using the rabbit reticulocyte lysate system (Promega) containing [35S]Met.Boiling-soluble polypeptides were prepared as described (Lin et al., 1990), fractionated by SDS-PAGE (Laemmli, 1970) on 15% (w/v) polyacrylamide gels, and detected by autoradiography.

Whole plants or plant parts were placed in GUS staining solution (100 m~ NaP04, pH 7.0, 3 m~ K3[Fe(CN),], 10 m~ EDTA, 0.1% Triton X-100, and 2 m~ 5-bromo-4-chlciro-3indolyl-P-D-glucopyranoside), vacuum infiltrated for at least 20 min, and then incubated at 37OC ovemight (Gallagher, 1992). Chl was removed from the tissue by washing with 50% ethanol and incubating ovemight in 95% ethanol! RESULTS

I

cor78 Encodes a 78-kD Polypeptide That Migrates wilh an Apparent Mass of 160 kD on SDS-PAGE

Hybrid-arrest/in vitro translation experiments have' indicated that the cor gene represented by cDNA clone pHH28, I

1049

Regulation of Arabidopsis thaliana cor78

S 2

Mol. Wt. (kDa)

<*160

I 47

24

Figure 1. In vitro transcription/translation of the cor78 coding sequence. pBMI was transcribed and translated in vitro, and the polypeptide products that remained soluble upon boiling were analyzed (lane marked pBMI). Also shown are the boiling-soluble polypeptides synthesized by in vitro translation of total RNA isolated from A. thaliana ecotype Landsberg erecta grown at 22°C (Warm) and cold treated at 5°C for 24 h (Cold). A negative control in which no RNA was added to the in vitro translation reaction is shown (Water). Radioactive polypeptides were separated by SDS-PACE and visualized by autoradiography.

The 5' Region of cor78 Has cis-Acting Elements That Impart Cold-Regulated Gene Expression in Leaves, Stems, Roots, and Certain Flower Parts

Yamaguchi-Shinozaki and Shinozaki (1993) reported the results of gene-fusion experiments indicating that the 5' region of rd29A has a ds-acting element(s) that can impart cold-, drought-, and ABA-regulated gene expression. The construct used in the study contained rd29A sequences between positions —880 and +81 (relative to the site of transcription initiation) fused to the gus reporter gene. Similarly, we found that cor78-gus fusions containing cor78 sequences between either positions —808 and +250 (78P1gus) or -808 and +5 (78P-gus) were cold regulated (Fig. 2, A and B); the latter fusions were also shown to be responsive to ABA and drought (Fig. 3). Histochemical staining of plants transformed with the 78PI-gus construct indicated that the gene was expressed in most, but not all, tissues of cold-treated plants. Relatively strong GUS staining occurred in the leaves (Fig. 4E), stems (Fig. 4E), roots (Fig.

GUS

78P-gus

Nos

Cor78 -808

78PI-gus designated corlBO, encodes a 160-kD polypeptide that remains soluble upon boiling in aqueous buffer (Thomashow et al., 1992). Nordin et al. (1991) isolated a cDNA for the same gene, which they designated Itil40, and presented hybrid-select/in vitro translation experiments indicating that the gene encoded a 140-kD polypeptide. The corlBO and UH40 transcripts, however, are only about 2.5 kb in length (Hajela et al., 1990; Nordin et al., 1991). Such transcripts would be expected to encode a polypeptide of approximately 80 kD. Indeed, Yamaguchi-Shinozaki and Shinozaki (1993), Nordin et al. (1993), and Horvath and McLarney (Horvath, 1993; GenBank accession numbers L22567 and L22568) recently determined the nucleic acid sequence of the gene, which indicated that it encodes a 78-kD hydrophilic polypeptide; it was designated rd29A, Iti78, and cor78, respectively. The question raised by these results was whether the previously described 160-kD 'boiling-soluble' polypeptide corresponded to the cor78 (IH78, rd29A) gene product. To address this issue, the cDNA insert cloned in pBMI (this insert contains the complete coding sequence for cor78) was transcribed and translated in vitro and the polypeptide products were fractionated on SDS-PAGE. The results indicated that the cor78 transcript did indeed encode a boiling-soluble polypeptide that migrated with an apparent mass of about 160 kD (Fig. 1).

I Cor78 -808

+ 250

35S 78P2 78PI3 WC WC WC

GUS

i

cor 78

Figure 2. Expression of 78P-gus, 78PI-gus, and 355-gus in control and cold-treated plants. A, The 355-gus, 78P-gus, and 78PI-gus promoter fusions are indicated. The dark box in the 78PI-gus construct indicates the position of the first intron in cor78. B, Northern analysis of transgenic plants carrying 355-gus (35S), 78Pgus (78P2), and 78PI-gus (78PI3). Total RNA was isolated from plants grown at 22°C (W) or cold treated at 5°C for 24 h (C). Northern blots were prepared and hybridized with 32P-labeled inserts from the gus gene carried on pBI121 (GUS) or the -808 to +250 region of cor78 (cor78).

Horvath et al.

1050

D W C A

GUS

cor78 Figure 3. Expression of 78P-gus in control, cold-, drought-, and ABA-treated plants. Transgenic plants carrying 78P-gus were grown at 22°C (W), cold treated at 5°C for 24 h (C), drought stressed (D), or treated exogenously with ABA (A) (see "Materials and Methods" for details). Total RNA was isolated from the plants, northern blots were prepared, and the transfers were hybridized with 32P-labeled inserts from the gus gene carried on pBI121 (GUS) or the -808 to +250 region of cor78 (cor78).

Plant Physiol. Vol. 103, 1993

4F), and flower sepals (Fig. 4G) of cold-treated plants, and weak but clearly discernible staining was observed in flower petals and anther filaments from cold-treated plants (Fig. 4G). However, no GUS activity was detected in the anthers, ovaries, stigmas, or styles of flowers from coldtreated plants (Fig. 4G). The lack of GUS activity in the anthers, styles, and stigmas was not an artifact of the staining procedure, because these tissues did stain positively for GUS activity in transgenic plants that had been transformed with the CaMV 355-gus gene (not shown). The significance of the lack of staining in the ovaries of transgenic plants carrying the 78Pl-gus gene is uncertain since these tissues also lacked GUS activity in transgenic plants carrying the 355-gus gene (not shown). Expression of the 78Pl-gus gene was either very low or undetectable in almost all tissues of nonacclimated plants (Fig. 4, A, B, and C). However, relatively high levels of expression were often detected in the trichomes and associated cells of control plants (Fig. 4, A and D). YamaguchiShinozaki and Shinozaki (1993) obtained the same result

Figure 4. Histochemical staining for GUS activity in transgenic plants carrying the 78PI-gus gene. Plants were grown at 22°C for approximately 2 weeks and then either stained directly for GUS activity (control plants: A, B, C, D, H) or first cold acclimated at 5°C for 24 h and then stained for GUS activity (cold-treated plants: E, F, G). A, Control plant; B, roots from control plant; C, flower from control plant; D, close-up view of leaves from control plant; E, cold-treated plant; F, root from cold-treated plant; G, flower from cold-treated plant; H, close-up of leaf from control plant that had been kept moist by misting with water.

Regulation of Arabidopsis thaliana cor78

with their rd29A-gus construct. These data suggested that the 78Pl-gus gene might be expressed constitutively in these cells. This, however, was not the case, because expression of the gene was not detected when the plants were keep moist by misting with water (Fig. 4H). Thus, expression of the 78PI-gus gene in the trichomes and associated cells in control plants appeared to result from a localized response to water stress.

1051

Temperature (OC ) 3 14

21

23

I

I

24 I

90

120

24

21

24

I

I

I

Kinetics of cor78 Temperature Regulation

Time-course experiments indicated that the activity of the cor78 promoter was quickly down-regulated in response to normal growth temperature. Specifically, 78Pgus transcripts began to decrease at a rapid rate about 15 min after cold-acclimated plants were transferred to nonacclimating temperatures (Fig. 5A). At that time, the plants were at about 14OC. Previous experiments indicated that the threshold temperature at which cor78 transcripts accumulate is between 10 and 12OC (Thomashow et al., 1990). Thus, the time-course data indicated that the cor78 promoter was strongly down-regulated within a few minutes of the plants reaching "noninducing" temperatures. This same conclusion was suggested by northern analysis of endogenous cor78 transcripts; the level of cor78 transcripts also began to decrease at a rapid rate about 15 min after cold-acclimated plants were transferred to nonacclimating temperatures (Fig. 5B). These latter experiments also indicated that the cor78 transcripts had a maximum half-life of only about 40 min at normal growth temperature.

O

60

30

150

180

Time in Minutes

210

2 O

Temperature (OC ) 15

21

24

25

25

25

45

60

75

90 105 1 2 0 135

25

f l4

c

2

b aJ

.-> 2 CI

(d

3

Fusion of cor78 to the CaMV 35s Promoter Results in Constitutively Expressed Transcripts

ol O

O

Nuclear run-on experiments have suggested that the accumulation of cor78 transcripts in response to low temperature is regulated largely at the posttranscriptional level (Hajela et al., 1990). Such regulation might result from differential stability of cor78 transcripts at cold and normal growth temperatures or temperature-induced alterations in cor78 transcript processing (cor78 has three introns). To test for such regulation, the entire transcribed region of cor78 was fused to the CaMV 35s promoter (Fig. 6A), the construct was transformed into Arabidopsis, and accumulation of 35s-cor78 transcripts in control and cold-treated transgenic plants was determined by RNase protection assays (Fig. 6B). The CaMV 35s promoter was used in these experiments because its activity, unlike that of cor78, does not appear to be dramatically affected by low temperature (Fig. 2; S.S. Baker and M.F. Thomashow, unpublished results). The results of the RNase protection assays indicated that there was little if any difference between the levels of the 35s-cor78 transcripts in control and coldtreated plants. In contrast, the levels of the endogenous cor78 transcripts were at least 50-fold higher in the coldtreated plants than they were in the control plants.

15

30

Time in Minutes Kinetics of cor78 temperature regulation. Wild-type or transgenic plants carrying 78P-gus were grown on agar medium, cold acclimated at 3°C overnight, and returned to room temperature. The temperature of the plants at various times after transfer to warm temperature was estimated by determining the temperature of the agar and air using a thermocouple. Total RNA was also isolated at various times and northern blots were prepared and hybridized with either a 32P-labeledprobe for the gus gene (the Smal/Sacl fragment of gus in pBI101) (A) o r a "P-labeled probe for cor78 transcripts (the cor78 insert in pDH15K) (B). Hybridization was quantified on a Betascope 3000 (Betagen).The highest level of Figure 5.

radioactivity in a given experiment was arbitrarily designated 100 and the other values were adjusted accordingly. The log of the relative transcript level is plotted as a function of time.

DISCUSSION

The promoter-fusion experiments reported here indicate that cor78 has a cis-acting regulatory element(s) between nucleotides -808 and +5 that can impart strong coldregulated gene expression (Fig. 2). These results confirm the finding of Yamaguchi-Shinozaki and Shinozaki (1993)

1052

Horvath et al.

35S-COr78 |CaMV Promoter | Cor?8 Transcribed Region

Plant Physiol. Vol. 103, 1993

transduction pathway responds either directly to temperature or to a quickly induced manifestation of it, rather than to a physiological adjustment associated with cold

acclimation such as altered osmotic potential. Whether the \CACGCGGC GGCACTCTAGA CTGTTTTAC -11 -1 B

RLD W

78T2

C W

C

35S-cor78

cor78

If Figure 6. Transcript levels of 35S-cor78 in control and cold-treated plants. A, DMA sequences contained in the 35S-cor78 gene fusion. The gene contains the CaMV 35S promoter from pBI121 (the first 9 nucleotides of the 35S transcript are indicated in italics), 11 nucleotides from the pBI121 polylinker (indicated by standard letters), and sequences from —11 to about +3000 of cor78 (nucleotides -11 to +1 of cor78 are indicated in bold letters). B, Total RNA was isolated from A. thaliana RLD (RLD) and transgenic plants carrying the 35S-cor78 gene (78T2) grown at either 22°C (W) or cold treated at 5°C for 24 h (C). Expression of cor78 and 35S-cor78 was determined by RNase protection assays using an antisense RNA that corresponded to positions —113 to +250 of cor78. Transcription of the 11 nucleotides of cor78 (-11 to +1 of cor78) in the 355cor78 gene fusion allowed it to be distinguished from the endogenous cor78 gene. Radioactivity was quantified using a Betascope 3000.

that rd29A has a cold-regulatory element between positions -880 and +81. Presumably, this element(s) acts at the transcriptional level. If, alternatively, it acted at a posttranscriptional level, it would mean that the first five nucleotides of the cor 78 transcript were sufficient to impart coldregulated accumulation of the hybrid transcript. Although this is a formal possibility, it seems unlikely. Indeed, fusion of the entire transcribed region of cor78 to the CaMV 35S promoter resulted in a chimeric gene that was constitutively expressed and displayed little if any posttranscriptional regulation in response to low temperature (Fig. 6). Time-course experiments indicate that the cor78 promoter can respond quickly to temperature (Fig. 5). In particular, cor78 promoter activity was found to be downregulated within minutes of cold-acclimated plants being exposed to noninducing temperatures. These data suggest that in down-regulation, the cor78 promoter responds directly to temperature per se, not to a change in physiological state imposed by low temperature. That is, it seems more likely that a DNA binding protein or factor in a signal

same situation holds for up-regulation of the cor78 promoter is less clear because northern analyses suggest that cold induction of the cor78 promoter might require as much as 1 h of exposure to low temperature (Hajela et al., 1990; Nordin et al., 1991; Nordin et al., 1993; Yamaguchi-Shinozaki and Shinozaki, 1993; D.P. Horvath, unpublished results). Future studies will be directed at determining whether this is actually the case, and if so, what the molecular basis is for the delay. The time-course experiments indicate that the maximum half-life of the cor78 transcripts is only about 40 min at normal growth temperature. This places the cor78 transcripts among the most rapidly degraded RNAs that have been described in plants (Green, 1993). It will be of interest to determine whether there is a specific element(s) within the transcript analogous to the DST sequences of SAUR (small auxin-up RNAs) genes (Newman et al., 1993) that target the message for quick degradation, or whether the instability of the transcript relates to more general properties of the RNA. The 78Pl-gus chimeric gene was found to be expressed in many but not all plant tissues (Fig. 4). The molecular basis for this tissue specificity is not known. It is possible that the cor78 cold-regulatory element(s) can function independently from other enhancers and that the tissue specificity observed reflects that of the cold-regulatory element itself. Alternatively, the cold-regulatory element may need to work in

concert with tissue-specific enhancers. In this case, the tissue specificity observed may reflect either the absence of enhancers from the cor78 promoter or their deletion in the 78PIgus construct. It should be possible to dishnguish between these possibilities once the cold-regulatory element(s) is defined. The existence of a strong transcriptional cold-regulatory element in the cor78 promoter was not predicted from the results of nuclear run-on experiments (Hajela et al., 1990). Instead, these experiments suggested that the cor78 promoter was active at normal growth temperature and that the rates of transcription increased only about 2-fold in cold-treated plants. The reason for the apparent discrepancy is not obvious. One possible explanation is raised by the finding of Yamaguchi-Shinozaki and Shinozaki (1993) and Nordin et al. (1993) that there is a gene located immediately upstream from cor78, designated rd29B and Iti65, respectively, that has a high degree of nucleic acid sequence identity with cor78. Perhaps it is this second gene that has a constitutive promoter and is regulated primarily at the posttranscriptional level. For such a situation to explain the nuclear run-on results, however, the probe used in the experiments, pHH28, would have had to detect the rd29B/lti65 transcripts. The results of Nordin et al. (1993) suggest that this would not have been the case (a cDNA probe containing only the 3' end of Iti78/cor78, like pHH28, was specific for the Iti78/cor78 transcripts), although direct experiments will be required to rule out the possibility. Another possible explanation for the nuclear

Regulation of Arabidopsis thaliana cor78 run-on experiments is t h a t t h e constitutive expression observed for cor78 was a n artifact. Whether s u c h a n artifact h a s a trivial explanation, s u c h as t h e promoter being induced b y t h e procedure used t o isolate t h e nuclei, o r has a more profound explanation remains t o b e determined. It should b e noted, however, t h a t t h e results of t h e nuclear run-on experiments (Hajela e t al., 1990) indicated t h a t cor25n h a s a cold-regulated promoter and t h a t this result h a s been corroborated b y promoter fusion studies (Baker and Thomashow, 1992; S.S. Baker, K.S. Wilhelm, a n d M.F. Thomashow, unpublished results). ACKNOWLEDCMENTS

We thank Sarah Gilmour and Stokes Baker for critical readings of the manuscript. Received May 14, 1993; accepted August 15, 1993. Copyright Clearance Center: 0032-0889/93/103/1047/07 LITERATURE ClTED

Baker SS, Thomashow MF (1992) Transcriptional regulation of cor15, a cold-regulated gene of Arabidopsis thaliana (abstract No. 194). Plant Physiol99: S-33 Benfey PN, Chua NH (1990) The cauliflower mosaic virus 3 5 s promoter: combinatorial regulation of transcription in plants. Science 250: 959-966 Gallagher SR (1992) Gus Protocols. Academic Press, San Diego, CA Gilmour SJ,Artus NN, Thomashow MF (1992) cDNA sequence analysis and expression of two cold-regulated genes of Arabidopsis thaliana. Plant Mo1 Biol 18: 13-21 Gilmour SJ, Hajela RK, Thomashow MF (1988) Cold acclimation in Arabidopsis thalinna. Plant Physiol87: 745-750 Gilmour SJ, Thomashow MF (1991) Cold acclimation and coldregulated gene expression in ABA mutants of Arabidopsis thaliana. Plant Mo1 Biol 17: 1233-1240 Green PJ (1993) Control of mRNA stability in higher plants. Plant Phvsiol102: 1065-1070 Guy -CL (1990) Cold acclimation and freezing stress tolerance: role of protein metabolism. Annu Rev Plant Physiol Plant Mo1 Biol41: 187-223 Hajela RK, Horvath DP, Gilmour SJ, Thomashow MF (1990) Molecular cloning and expression of cor (cold-regulated) genes in Arabidopsis thaliana. Plant Physiol93 1246-1252 Horvath DP (1993) Cloning, characterization and regulation of cold-induced genes from Arabidopsis thaliana. PhD thesis. Michigan State University, East Lansing Jefferson RA (1987) Assaying dimeric genes in plants: the GUS gene fusion system. Plant Mo1 Biol Rep 5: 387-405 Kabnick KS, Housman DE (1988) Determinants that contribute

1053

to cytoplasmic stability of human c-fos and P-globin mRNAs are located at severa1 sites in each mRNA. Mo1 Cell Biol 8: 3244-3250 Koornneef M, Reuling G, Karssen CM (1984) The isolation and characterization of abscisic acid-insensitive mutants of Arabidopsis thaliana. Physiol Plant 61: 377-383 Kurkela S, Borg-Franck M (1992) Structure and expression of kin2, one of two cold- and ABA-induced genes of Arabidopsis thaliana. Plant Mo1 Biol 19: 689-692 Kurkela S, Franck M (1990) Cloning and characterization of a cold- and ABA-inducible Arabidopsis gene. Plant Mo1 Biol 1 5 137-144 Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680-685 LIng V, Palva ET (1992) The expression of a rab-related gene, rabld, is induced by abscisic acid during the cold acclimation process of Arabidopsis thaliana (L.) Heynh. Plant Mo1 Biol 20: 951-962 Lin C, Guo WW, Everson E, Thomashow MF (1990) Cold acclimation in Arabidopsis and wheat. A response associated with expression of related genes encoding ‘boiling-stable’ polypeptides. Plant Physiol94: 1078-1083 Lin C, Thomashow MF (1992) DNA sequence analysis of a complementary DNA for cold-regulated Arabidopsis gene cor15 and characterization of the COR15 polypeptide. Plant Physiol 99: 519-525 Newman TC, Ohme-Takagi M, Taylor CB, Green PJ (1993) DST sequences, highly conserved among plant SAUR genes, target reporter transcripts for rapid decay in tobacco. Plant Cell 5: 701-714 Nordin K, Heino P, Palva ET (1991) Separate signal pathways regulate the expression of a low-temperature-induced gene in Arabidopsis thaliana (L.) Heynh. Plant Mo1 Bioll6 1061-1071 Nordin K, Vahala T, Palva ET (1993) Differential expression of two related, low-temperature-induced genes in Arabidopsis thaliana (L.) Heynh. Plant Mo1 Biol21: 641-653 Sambrook J, Fritsch EF, Maniatis T (1989) Molecular Cloning. A Laboratory Manual, Ed 2. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY Thomashow MF (1993) Genes induced during cold acclimation in higher plants. Zn PL Steponkus, ed, Advances in Low Temperature Biology, Vol 2. JAI Press, London, pp 183-210 Thomashow MF, Gilmour SJ, Hajela R, Horvath D, Lin C, Guo W (1990) Studies on cold acclimation in Arabidopsis thaliana. In AB Bennett, SD ONeill, eds, Horticultural Biotechnology. Wiley-Liss, New York, pp 305-314 Thomashow MF, Gilmour SJ, Lin C (1992) Cold-regulated genes of Arabidopsis thaliana. In PH Li, L Christersson, eds, Advances in Plant Cold Hardiness. CRC Press, Boca Raton, FL, pp 32-44 Valvekens DM, Van Montagu M, Van Lijsebettens M (1988) Agrobacterium tumefaciens-mediatedtransformation of Arabidopsis thaliana root explants by using kanamycin selection. Proc Natl Acad Sci USA 85: 5536-5540 Yamaguchi-Shinozaki K, Shinozaki K (1993) Characterization of the expression of a desiccation-responsiverd29 gene of Arabidopsis thaliana and analysis of its promoter in transgenic plants. Mo1 Gen Genet 236: 331-340