INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

High soil salinity, which is caused typically by NaCl, results in ion toxicity and hyperosmotic stress leading to numerous pathologies including generation of reactive oxygen species (ROS) and programmed cell death (Niu et al., 1995; Zhu et al., 1997; Hasegawa et al., 2000b). Salt tolerance determinants are categorized either as effectors that directly modulate stress etiology or attenuate stress effects, or as regulatory molecules that are involved in stress perception, signal transduction, or modulation of effector function. Enzymes that catalyze rate-limiting steps in the biosynthesis of compatible solutes or osmoprotectants, e.g. sugar alcohol, quaternary ammonium, and tertiary sulfonium compounds, are categorical examples of osmotic stress tolerance effectors (Hanson et al., 1994; Ishitani et al., 1997; Yoshiba et al., 1997; Nelson et al., 1998; Hasegawa et al., 2000b). Other effectors include proteins that protect membrane integrity, control water or ion homeostasis, and ROS scavenging (Bray, 1994; Ingram and Bartels, 1996; Hasegawa et al., 2000b). Determinant function of some effectors has been confirmed because expression in transgenic plants enhances salt tolerance sufficiency (Hasegawa et al., 2000b).

Regulatory determinants include transcription factors and signal pathway intermediates that posttranscriptionally activate effectors (Hasegawa et al., 2000b). Basic Leu zipper motif, MYB and MYC, and zinc finger transcription factors, including rd22BP1 (MYC), AtMYB2 (MYB), DREB1A, and DREB2A (AP2 domain), and ALFIN1 (zinc finger), interact with promoters of osmotic-regulated genes (Abe et al., 1997; Liu et al., 1998; Hasegawa et al., 2000b). The osmotic stress tolerance function of DREB1A in Arabidopsis (Kasuga et al., 1999) and ALFIN1 in alfalfa (Medicago sativa; Winicov, 2000) has been confirmed by ectopic expression in transgenic plants. Regulatory intermediates that modulate plant salt stress responses include SOS3 (Ca²⁺-binding protein), SOS2 (Suc non-fermenting-like) kinase, Ca²⁺-dependent protein kinases, and mitogen-activated protein kinases (Sheen, 1996; Halfter et al., 2000; Kovtun et al., 2000). Additional signal intermediates have been implicated in the plant response to salt (Hwang and Goodman, 1995; Mizoguchi et al., 1996; Mikami et al., 1998; Piao et al., 1999; Hasegawa et al., 2000b).

The Zhu laboratory recently has pioneered identification of salt tolerance determinants using forward genetics in the plant model Arabidopsis (Hasegawa et al., 2000a; Sanders, 2000; Zhu, 2000). This effort has identified three complementation groups of ion hypersensitive (salt overly sensitive [sos1-sos3]) mutants. Genetic and physiological data indicate that SOS3, SOS2, and SOS1 are components of a signal pathway that regulates ion homeostasis and salt tolerance and their functions are Ca²⁺ dependent. Positional cloning revealed that SOS1 encodes a putative plasma membrane Na⁺/H⁺ antiporter, SOS2 encodes a Suc non-fermenting-like (SNF) kinase, and SOS3 encodes a Ca²⁺-binding protein with sequence similarity to the regulatory subunit of calcineurin and neuronal Ca²⁺ sensors (Liu and Zhu, 1998; Liu et al., 2000; Shi et al., 2000). Molecular interaction and complementation analyses indicate that SOS3 is required for activation of SOS2 that regulates SOS1 transcription (Halfter et al., 2000; Shi et al., 2000), further confirming that the order of the signal pathway is SOS3  SOS2  SOS1 (Hasegawa et al., 2000a; Sanders, 2000; Zhu, 2000).

Herein is described a functional dissection of the plant salt stress response by analysis of gene expression controlled by the SOS signal pathway. Differential subtraction and northern analysis identified 84 salt-regulated genes, the majority of which have not been annotated to be salt responsive in Arabidopsis. Comparison of salt regulation in wild type and sos3 identified six genes that are controlled by the SOS signal pathway. Transcription of one gene (VSP2) is controlled as an output from the SOS3/SOS2 pathway, similar to SOS1 (Shi et al., 2000), whereas regulation of five other genes (AD06C08, CCR1, SAMT, COR6.6/KIN2, and STZ) is dependent on both SOS3 and SOS2 as well. Because SAMT, COR6.6/KIN2, and STZ are induced by NaCl, negative control by SOS3 indicates that the SOS pathway functions to reduce the numerous signals induced by salt to those that function more specifically, to mediate processes like ion homeostasis.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Identification of Expressed Sequence Tags (ESTs) Corresponding to Salt-Regulated Genes

A combination of approaches was used to identify stress-regulated transcriptional outputs from the SOS pathway on the premise that these are salt tolerance determinants. A population of ESTs representing salt-regulated transcripts was identified by screening, through three rounds, a subtracted cDNA library prepared from wild-type (Col-0 gl-1) seedlings treated for 4 h without or with 160 mM NaCl. Differential dot-blot hybridization, and sequence and northern-blot analyses identified unique ESTs (salt-regulated ESTs [SREs]) that detected salt-regulated transcripts (Table I). The second and third rounds of screening eliminated highly repetitive clones from the initial subtracted library, including those ESTs identified during the first round of screening. The differential subtractive chain selection protocol was used in the third round (Luo et al., 1999) to further enrich for less abundant cDNAs in the library. The different rounds of screening led to identification of 84 nonredundant ESTs that hybridize to salt-regulated transcripts based on northern-blot analysis (Table II). The SREs represent the greatest number that has been used in one study to define the transcriptional response of plants to salt stress. These results identify genes whose expression is most likely controlled by transcriptional activation, although other factors such as salt stress-dependent mRNA stability might contribute also to steady-state transcript abundance.

View this table: [in this window] [in a new window]   Table II.   Arabidopsis (Col-0 and gl1) ESTs corresponding to salt-responsive genes determined by northern-blot analysis Total RNA was isolated from seedlings transferred to medium without or with 160 mM NaCl for 4 h SREs were subjected to Blastn analysis and illustrated are annotations based on the information from the Arabidopsis genome sequence project: chromosome no., section on chromosomes 2 and 4, or BAC clone identification no. (for chromosomes 1, 3, and 5), and gene/protein accession nos. Highlighted in bold are SREs that detected transcripts differentially regulated in gl1 and sos3. Transcript abundance is rated based on the scale illustrated.

Database comparisons of SREs using Blast programs determined that the corresponding encoded proteins included those involved in primary metabolism, cell wall synthesis or degradation, other cellular functions, transport or nutrient assimilation, signaling, and defensive responses (Table II). The SREs were compared with Arabidopsis genome data using Blastn. Blastp analysis was performed on any SRE ORF, without predicted function, that was identified in an Arabidopsis database. Blastx/Blastp analysis was performed on SRE sequences that were unannotated as an ORF.

Several of the salt-responsive genes identified in this evaluation encode components of octadecanoid signaling through jasmonic acid (Table II, lipid signaling responses). The plant hormone appears to be derived from hydrolysis of membrane phospholipids (Koiwa et al., 1997). Triacyl glycerol lipase (AD03B08) can release free linolenic acid from phospholipids that is then oxidized by lipoxygenase (AD04G12) and cyclized by allene oxide synthase and allene oxide cyclase (AD04D07) to 12-oxo-phytodienoic acid. Two SREs were annotated previously as jasmonic acid regulated, VSP2 (AD05E05) and AD06F05. Furthermore, the genes of a number of SREs are involved in plant defense and may be regulated by the octadecanoid signal pathway. Some of these genes have been shown to express upon dehydration in tomato (Lycopersicon esculentum; Reymond et al., 2000). Several abscisic acid (ABA)-responsive SREs are included in Table II and the plant hormone is a potentiator of octadecanoid signaling.

Genes Differentially Regulated by Salt in Wild Type and sos3

The salt regulated expression profile of SRE transcripts, as well as that of previously characterized stress-regulated genes (ACP1, RD22BP1, MYB2, STZ, and PAL), revealed that most are controlled similarly in wild type and sos3. Six of the salt-responsive genes, including STZ (Lippuner et al., 1996), were differentially regulated in wild type and sos3 (Fig. 1). Transcript abundance of two was lower (AD06C08/unknown and AD05E05/vegetative storage protein2 [VSP2]) and of four was higher (encoding AD05C10/cold-circadian rhythm-RNA binding1 [CCR1], STZ/salt tolerance zinc finger [not shown], AD05B11/S-adenosyl-L-Met: salicyclic acid carboxyl methyltransferase [SAMT], AD03D05/cold regulated/cold inducible [COR6.6/KIN2]) in the salt-sensitive mutant. SAMT and COR6.6/KIN2 transcript abundance was slightly elevated in sos3 but the steady-state mRNA levels were hyper-induced by salt treatment. CCR1 was the only gene for which transcript abundance is reduced by salt treatment.

View larger version (72K): [in this window] [in a new window]   Figure 1.   Salt-responsive gene expression that is dependent on the SOS pathway in Arabidopsis. Genes that are differentially regulated in wild-type (Col-0 gl1) and sos3. Total RNA (40 µg) from seedlings grown in liquid culture (14 d) and treated without or with 160 mM NaCl for 4 h. The northern blot was hybridized with 32P-labeled probe corresponding to: COR6.6/KIN2, SAMT, CCR1, VSP2, and AD06C08 (unknown). AtGAPDH (glyceraldehyde-3-phosphate-dehydrogenase) is the control.

Salt regulation of VSP2 is similar in wild type and sos1 indicating that transcriptional activation is not dependent on SOS1 (Fig. 2). Methyl jasmonate (MeJA) induces VSP2 transcript abundance in wild type and sos3 (not shown). The SOS3/2 pathway and the hormone seem to regulate VSP2 independently. Signal pathways often converge to regulate transcription of key effectors involved in cellular adaptation to environmental perturbation (Rep et al., 2000). VSP2 is a member of a two-gene family (87% nucleotide sequence identity over the coding region) that encodes a protein with similarity to soybean VSPs (Berger et al., 1995; Utsugi et al., 1998), which are vacuolar-localized glycoproteins with acid phosphatase activity (Mason and Mullet, 1990). These proteins are presumed to be amino acid sinks during water deficit but are important reduced nitrogen sources after stress relief (Mason and Mullet, 1990).

View larger version (72K): [in this window] [in a new window]   Figure 2.   Comparative expression of genes dependent on the SOS pathway in wild type (Col-0 and gll) and sos1, sos2, and sos3. Illustrated is the northern blot of steady-state mRNA levels of COR6.6/KIN2, SAMT, CCR1, and VSP2 in plants without (0 h) or 160 mM NaCl (24 h). Ethidium bromide staining was used to monitor RNA loading.

CCR1, STZ, SAMT, COR6.6/KIN2, and AD06C08/unknown have similar expression profiles in sos1, sos2, and sos3, implicating these as transcriptional outputs requiring all components of the SOS pathway. CCR1 encodes a Gly-rich RNA-binding protein implicated in posttranscriptional regulation. CCR1 and CCR2 comprise a two-gene family, and their expression is regulated by a diurnal circadian clock (Carpenter et al., 1994; Heintzen et al., 1997; Kreps and Simon, 1997). CCR1 or CCR2 expression negatively regulates either gene and this feedback loop presumably facilitates diurnal oscillation controlled by the master circadian clock (Heintzen et al., 1997). CCR1 and CCR2 steady-state mRNA levels are induced by cold but CCR1 transcript is negatively regulated by ABA and dehydration, whereas CCR2 expression is induced by dehydration (Carpenter et al., 1994). CCR1 transcript abundance was down-regulated in sos3 indicating that the SOS pathway is at least another negative regulator that controls CCR1 expression downstream of the circadian rhythm clock (Heintzen et al., 1997; Kreps and Simon, 1997). These results indicate that control of circadian oscillations may be required during the salt stress response.

COR6.6//KIN2 and SAMT are both implicated in plant stress responses. COR6.6/KIN2 is linked in tandem to its homolog KIN1 (95% nucleotide sequence identity in the coding region) and both encode hydrophilic peptides that are boiling soluble but of unknown function (Thomashow, 1999). KIN1 and 2 transcript abundance is cold and ABA induced (Kurkela and Franck, 1990; Kurkela and Borg-Franck, 1992; Thomashow, 1999) but KIN2 and not KIN1 expression is modulated positively by drought and salt (Kurkela and Borg-Franck, 1992). SAMT catalyzes the formation of methylsalicylate from salicylic acid using S-adenosyl-L-Met as the methyl donor (Ross et al., 1999). The volatile ester is implicated as a pollinator attractant and a signal in plant defense mediated by salicylic acid (Ross et al., 1999; Dudareva et al., 2000).

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Many salt-regulated genes are responsive also to other biotic or abiotic perturbations, indicating that these stresses have common etiologies, e.g. water deficit, and cold are both osmotic stresses, occur simultaneous in the environment or elicit similar pathologies (Bohnert et al., 1995; Shinozaki and Yamaguchi-Shinozaki, 1997; Zhu et al., 1997; Hasegawa et al., 2000b). Furthermore, osmotic and ionic stresses induce secondary cellular perturbations that arise from ROS, elicitors from degradation of cell wall and plasma membrane macromolecules, or wounding, which initiate signal transduction pathways that modulate other plant defensive processes (Hasegawa et al., 2000b). In fact, many of the proteins encoded by the 84 salt-responsive genes identified in this study can be categorized as functional outputs from these different signal cascades (Table II). Most of these genes were not known previously to be modulated as a part of the Arabidopsis salt response. It is interesting that even after subtractive hybridization only approximately 13% of the ESTs (84/614) were unique and detected salt-regulated transcripts.

Six of 89 genes examined were differentially responsive to NaCl in wild type and sos3, implicating the SOS pathway in their transcriptional regulation. AD06C08/unknown and VSP2 are induced and CCR1, STZ, SAMT, and COR6.6/KIN2 are controlled negatively by SOS3. CCR1 expression was lower in wild type after NaCl treatment, whereas the message abundance of the others was salt induced. From these results, a model for the SOS pathway regulation of these genes is illustrated in Figure 3 (Zhu, 2000). Salt regulated expression of VSP2 is the same in wild type and sos1 defining this gene as a transcriptional output from the SOS pathway that does not require SOS1. This supports the premise that SOS3 and SOS2 are signal intermediates and SOS1 is an effector of Na⁺ homeostasis (Shi et al., 2000). However, salt regulation of AD06C08/unknown, CCR1, STZ, SAMT, and COR6.6/KIN2 is dependent on SOS2 and SOS1, perhaps implicating a signaling function for SOS1. Genes encoding an enzyme catalyzing the penultimate step in Pro biosynthesis (P5CS) and a putative transcription factor (AtMYB) are hyper-induced by salt in sos1 compared with wild type (Liu and Zhu, 1997). Furthermore, some transport proteins include sensor domains or function in association with sensors (Ozcan et al., 1998; Heinisch et al., 1999; Sabirov et al., 1999) as could the putative Na⁺/H⁺ antiporter SOS1 (Shi et al., 2000).

View larger version (16K): [in this window] [in a new window]   Figure 3.   Illustrated is a model depicting the SOS pathway regulation of salt responsive genes. Hypersaline conditions activate the SOS (SOS3  SOS2  SOS1) signal pathway (Zhu, 2000) and transcript abundance of AD0608 (unknown), VSP2 (vegetative storage protein 2, AD05E05), SAMT (S-adenosyl-L-Met:salicylic acid carboxyl methyltransferase, AD05B11), COR6.6/KIN2 (cold regulated 6.6/inducible 2, AD03D05 [COR6.6/KIN2], and STZ [salt tolerance zinc finger], Lippuner et al., 1996) increase and of CCR1 (circadian rhythm-RNA binding1, AD05C10) decreases in Arabidopsis seedlings. Positive () or negative () regulation by the SOS pathway is indicated.

The experimental evidence presented here indicates that the SOS pathway controls expression of only a few salt stress-specific tolerance determinant genes among the numerous genes (six of 89 in this study) that are regulated in the plant response to NaCl treatment (Zhu et al., 1997). This is similar to the paradigm that has been established recently for the salt stress response of the unicellular eukaryote yeast (Saccharomyces cerevisiae). Genome-wide array analysis determined that osmotic upshock causes a rapid and multi-fold increase in mRNA of between 186 and 1,359 genes and reduced transcript abundance of more than 100 genes depending on the severity of osmotic shock, the osmotic agent (NaCl or sorbitol), and time after treatment (Posas et al., 2000; Rep et al., 2000; J. Yale and H.J. Bohnert, unpublished data). Salt-induced expression of most is either partially or completely controlled by the high osmolarity glycerol and mitogen-activated protein kinase pathway. The yeast calcineurin pathway is analogous to the Arabidopsis SOS pathway, controls ion homeostasis and is essential for salt tolerance in yeast (Mendoza et al., 1994, 1996), and affects expression of many fewer osmotic responsive genes (T.K. Matsumoto, unpublished data). Like the SOS pathway, calcineurin regulates expression of genes that encode salt tolerance effectors such as Na⁺ efflux transporters (Mendoza et al., 1994; Shi et al., 2000; Matusmoto et al., unpublished data).

It is conceivable that signal transduction through the SOS pathway that mediates salt tolerance may have a substantial component that involves posttranscriptional activation of salt tolerance effectors, particularly over the time span (minimum of 4 h) of the salt treatment used in experiments reported here. Plant survival in severe stress likely requires very immediate cellular responses, whereas transcriptional regulation may be sufficient for stress recovery and adaptation. Notwithstanding, salt induces transcriptional activation of genes in yeast within minutes (Posas et al., 2000; Rep et al., 2000). Genes that are transiently induced or weakly expressed further complicate inference of function from expression profile analysis. The majority of yeast genes induced by mild salt shock exhibit transient expression (Posas et al., 2000). Determinant gene transcript abundance differences in wild type and sos3 may be insufficient for the resolution limits of the subtraction protocols, yet are biologically meaningful to salt stress adaptation.

It is interesting that the SOS pathway negatively controls the expression of four salt-regulated genes (SAMT, C0R 6.6/KIN2, STZ, and CCR1), three (SAMT, C0R6.6/KIN2, and STZ) of which are induced by NaCl treatment. So, salt tolerance determinants include genes that must be repressed, at least temporally, during the plant stress response. Negatively regulated genes may include those that contribute to growth arrest necessary during the period of adjustment or ameliorate other etiologies that occur coincidentally with salt in the native environment of the organism. CCR1 and COR6.6/KIN2 are cold induced, whereas SAMT is implicated in plant defense against pathogens indicating their principal function is not in salt adaptation. Together, this suggests that a function of the SOS pathway is to discriminate against the myriad of stress signals that are elicited by salt and to focus the capacity of the plant to cope with the principal etiology; in this instance, ion dis-equilibrium. Furthermore, the SOS pathway may coordinate temporal gene expression to focus the availability of effectors, as required, during stress perception, amelioration, or adaptation. Confirmation that the genes identified in this study encode tolerance determinants awaits molecular genetic confirmation by loss- or gain-of-function experimentation.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Plant Material

Arabidopsis (ecotype Columbia-0 gl1) and sos1, sos2, and sos3 were wild-type and salt-hypersensitive genotypes, respectively (provided by Dr. Jian-Kang Zhu, University of Arizona). Seeds were surface disinfected and stratified for 2 d at 4°C. Seeds were germinated in liquid medium (Murashige and Skoog salts [Murashige and Skoog, 1962] and 3% [w/v] Suc [pH 5.8]) in 250-mL flasks on a gyratory shaker (80-100 rpm) under low-intensity Cool White fluorescent illumination (light/dark:16/8 h daily) at 22°C to 24°C. After 14 d, seedlings were transferred to fresh medium without or with 160 mM NaCl for the time interval indicated. Seedlings were harvested, frozen in liquid nitrogen, and stored at 80°C.

Construction of Subtraction Libraries

Total RNA was isolated as described by Gong et al. (1997). mRNA was isolated using the Poly(A)⁺ RNA purification kit (CLONTECH, Palo Alto, CA). The PCR-Select cDNA subtraction kit (K1804-1, CLONTECH) was used to obtain subtracted cDNA libraries.

Subtraction of cDNAs Obtained from mRNA of Salt-Treated sos3 and gl1

Subtractive hybridization was used to identify cDNAs corresponding to salt regulated genes differentially expressed in sos3. Both forward and reverse subtractive hybridizations were performed with salt-treated (160 mM NaCl, 4 h) seedlings. The forward subtraction used tester cDNA obtained from mRNA of gl1 and driver cDNA from sos3. In the reverse subtraction, the tester cDNA was obtained from sos3 and driver cDNA from gl1. Driver cDNAs are the reference and are targets for elimination during subtraction leaving unique tester cDNAs. Forward and reverse subtractive hybridization was meant to identify salt-responsive genes that are specifically regulated in sos3.

Subtraction of cDNAs Obtained from gl1 after Treatment without or with NaCl

This subtraction was intended for identification of genes differentially regulated by salt in gl1. The forward subtraction tester cDNA was obtained from mRNA of gl1 seedlings 4 h after transfer to medium with 160 mM NaCl and driver cDNA from gl1 plants grown in medium without salt. In the reverse subtraction, tester and driver cDNA was obtained from gl1 grown without and with 160 mM NaCl, respectively.

The subtracted libraries were subjected to two rounds of PCR amplification, the second using nested primers for adaptors 1 and 2R (CLONTECH). The PCR products were ligated into pT-Adv (CLONTECH), and transformed into Escherichia coli. The white colonies were isolated and inserts amplified by PCR.

Enrichment of Unique Salt-Regulated cDNAs in the Subtraction Library

A modification of the differential subtraction chain method (Luo et al., 1999) was used to enrich the subtracted library for unique salt-regulated cDNAs. The driver for this subtraction (driver 2) is a mixture of second PCR product from reverse subtraction (100 µL; tester DNA:gl1 without salt, driver DNA:gl1 with salt) and PCR products amplified from highly repetitive salt-induced cDNAs isolated from previous rounds of screening (0.5 µL of each). To remove adaptor sequences from driver 2, the mixture was digested at the restriction sites of adaptors 1 (SmaI and RsaI) and 2R (EagI and RsaI) by consecutive restriction digestion using SmaI, and then EagI + RsaI. The digested DNA was extracted with phenol and phenol/chloroform, and then precipitated with ethanol. The precipitate was redissolved in 100 µL of water and passed through the Microcon YM-30 column (Amicon, Beverly, MA) to separate the DNA from Adaptor fragments. Driver 2 DNA was recovered in 20 µL of water and used as driver in the following PCR subtraction. About 10 times excess amount of driver 2 DNA (2 µL) was mixed with 1µL of forward subtraction library (driver and tester = gl1 grown without and with 160 mM NaCl, respectively) in hybridization buffer {5 mM HEPES [4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid]-HCl (pH 8.3), 12 mM NaCl, and 0.05 mM EDTA} in a total volume of 12 µL. The hybridization solution was denatured at 94°C for 5 min, and incubated at 72°C for 12 h.

After hybridization, the mixture was precipitated with ethanol, and digested with mung bean nuclease (Promega, Madison, WI) to remove adaptor sequence from products of the driver-tester hybridization. Adaptor sequences in tester-tester hybridization products are aligned with matching ends and these are not digested by mung bean nuclease. The mung bean nuclease digested mixture was passed through the Microcon column and purified DNA was rehybridized at 72°C. The mung bean nuclease digestion and DNA purification procedures were repeated. The DNA was amplified by PCR using nested primers for 1 and 2R, 11 cycles. The cloning of PCR products was as described as above.

Dot-Blot Analysis

For dot-blot analysis, cDNA inserts of forward subtraction library clones were individually amplified by PCR and 2 µL of PCR product was mixed with 2 µL of 0.6 M NaOH. Two microliters of the mixture was blotted onto each of two duplicate nylon membrane filters. Probes for dot-blot analysis were either forward or reverse subtraction products or cDNA of RNA from plants, as indicated. The products of forward and reverse subtraction were digested with SmaI, RsaI, and EagI to remove the adapter sequences and labeled with ³²P using the Ready-To-Go kit (Amersham Pharmacia Biotech, Piscataway, NJ). The labeled forward probes were hybridized to one membrane and the reverse probes to the duplicate.

Template Preparation, DNA Sequencing, and Data Analysis

Plasmid templates were prepared from selected bacterial colonies by 96-well alkaline lysis minipreps according to the manufacturer's instructions (Edge BioSystems, Inc., Gaithersburg, MD).

DNA sequencing reactions were conducted using DyeDeoxy "Terminator PRISM" mix (Perkin-Elmer-ABI, Foster City, CA) according to the manufacturer's instructions in a multiplate thin-wall 96-well microplate on an MJ Research PTC-100-96 (MJ Research, Watertown, MA) programmable thermal controller using the following profile: 96°C for 30 s, 45°C for 15 s, and 60°C for 4 min for 49 cycles. Unincorporated dye terminators were removed by passing reactions over a 96-well gel filtration block (Edge BioSystems). Recovered sequencing reaction products were analyzed on either an ABI 373A-XL Stretch or an ABI 3700 capillary array automated DNA sequencing system (Perkin-Elmer Applied BioSystems). Raw sequence data was analyzed using PHRED (Ewing and Green, 1998; Ewing et al., 1998) and Cross  match to removal vector sequences. Additional vector sequence removal and editing was done manually using FACTURA software (Perkin-Elmer Applied BioSystems). Polished EST sequence files were assembled into singleton and contig files using PHRAP (P. Green, unpublished data). EST identities were determined by sequence comparison to the nonredundant GenBank database using BLASTN (BLAST 2.0) using default parameters (Altschul et al., 1997). In instances where an unannotated match was obtained, BLASTX searchers were conducted and sequence homology information was used to assign putative identities. All EST sequences reported here have been deposited in dbEST and can be browsed and retrieved from the NCBI website (http://www.ncbi.nlm.nih.gov) under accession numbers BE844684 through BE845405.

Northern-Blot Analysis of Putative Clones

Total RNA was isolated from seedlings and 40 µg from each sample was separated on 1.2% (w/v) agarose formaldehyde gels and transferred to Hybond-N nylon membranes (Amersham) as previously described (Gong et al., 1997). The cDNA insert of each clone was amplified by PCR using nested primers that hybridize to adaptors 1 and 2R, and purified from agarose gels using the Qiagen Gel Purification kit. The probes for RD22BP1, MYB2, PAL, STZ, and ACP1 were obtained by PCR amplification using cDNA obtained from mRNA of salt-treated gl1 as template; RD22BP1 (AB000875), forward primer: 5'-ATGACGCTGT-TGATGAGGAG-3' and reverse primer: 5'-TTTCGGATT-CTGGGTCTGAG-3' (0.56 kb); MYB2 (D14712), forward primer: 5'-GAAATGGAAGATTACGAGCG-3' and reverse primer: 5'-TTAATTATACGAATACGATGTC-3' (1.0 kb); PAL (L33677), forward primer: 5'-ATGGAGATT-AACG-GGGCACAC-3' and reverse primer: 5'-ACGT-TCACCG-TTGGGACCAG-3' (1.1 kb); STZ (X95573), ORF; and ACP1 (AF009228), forward primer: 5'-CAA-AAGCCATTTTT-CAAATTTCAAACTCAG and reverse primer 5'-GTTTT-CAATGATAGTGAAGAAAGATG-TAC-AAC (0.83 kb). The purified PCR products were labeled using ³²P dCTP using the Ready-To-Go kit. Blot-blot hybridization and washes were as described (Gong et al., 1997). The blots were stripped by boiling in 0.5% (w/v) SDS solution for 3 min and were rehybridized with another probe.