INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Cold acclimation is a process that occurs in many types of organisms in response to a decrease in temperature (Levitt, 1980; Graumann and Marahiel, 1996; Hughes and Dunn, 1996; Huner et al., 1998). The physiological, biochemical, and molecular processes involved in the attainment of cold acclimation have been studied extensively, but a complete understanding of the functions of the various genes induced by low temperature is still lacking (Thomashow, 1999). However, the role of some of these proteins has been determined, and they act as transcription or elongation factors, antifreeze proteins, and proteins involved in stabilizing membrane architecture (Jones and Inouye, 1994; Nishida and Murata, 1996; Chun et al., 1997; Gilmour et al., 1998).

At low temperatures, organisms have two primary difficulties. The first problem is maintaining membranes in a fluid state that allow them to resist subzero temperatures (Nishida and Murata, 1996). Membrane integrity and cell survival can also be compromised because of both intracellular and extracellular ice formation (Steponkus, 1984; Thomashow, 1999; Yu and Griffith, 1999). The second problem is encountered by photosynthetic organisms and is related to the thermodependency of photosynthetic electron transport and carbon fixation, which are slowed at low temperature (Guy, 1990; Huner et al., 1998). However, the primary photochemical reactions of light absorption by the light-harvesting antennae and the transfer of excitation energy to the photosynthetic reaction centers occur at rates that are essentially independent of temperature. Primary charge separation and exciton transfer from the antennae pigments to the reaction centers can still occur at the temperature of liquid N₂ (196°C; Butler, 1978). Thus, changes in temperature can result in the inability of the organism to use the absorbed light energy, leading to the over-reduction of the electron transport chain (Huner et al., 1998). This can, in turn, lead to photoinhibition of photosystem II (PSII) and increased production of oxygen radicals (Asada, 1994). In chloroplasts, there is a dynamic equilibrium between PSII damage and repair. When the rate of damage is greater than the rate of repair, photoinhibition occurs and is reflected in a decrease in the measurable F_V/F_M. Thus, avoidance of photoinhibition can occur either by decreased rates of damage or increased rates of repair (Melis, 1999). Cold-acclimated cereals have been shown to be less sensitive to photoinhibition, and interestingly, this appears to correlate with the maximum freezing tolerance of the plants (Hurry and Huner, 1992; Gray et al., 1997). This resistance has in part been explained by decreased rates of damage due to increased photosynthetic capacity (Hurry and Huner, 1992).

Previous screening of a wheat (Triticum aestivum) cold-acclimated cDNA library allowed the identification of a novel cold-regulated gene called Wcs19 that encodes a protein of unknown function (Chauvin et al., 1993). The mRNA of this gene was shown to be expressed exclusively in photosynthetic tissues, and its transcript accumulation was found to be correlated with the level of PSII excitation pressure (Gray et al., 1997). Therefore, this gene was not regulated by temperature per se but rather by a complex interaction of temperature and light. Sequence comparisons have shown the existence of closely related genes in barley (Hordeum vulgare; cor14b; Crosatti et al., 1999) and wheat (Wcor14a and Wcor14b; Tsvetanov et al., 2000), suggesting that Wcs19 may be part of a small family of genes with similar or overlapping functions and regulation. In an effort to gain a better insight into the structural and functional features of this gene family, we first set out to identify and characterize Wcs19 homologs and orthologs in rye (Secale cereale) and wheat. Second, given the correlation between Wcs19 mRNA accumulation and PSII excitation pressure, we developed antibodies to the WCS19 protein, and used them to examine its accumulation and subcellular localization in rye and wheat in response to excitation pressure and cold acclimation. These molecular and biochemical analyses revealed that this new gene family encodes chloroplastic proteins related to group 3 LEA (late embryogenesis abundant) proteins. Last, in an attempt to begin determining the function of this protein family, Arabidopsis was transformed with the Wcs19 gene, a representative member of this family, under the control of a constitutive promoter. The effects of the constitutive expression of WCS19 on the freezing tolerance and resistance to photoinhibition of transgenic Arabidopsis leaves are described and discussed with respect to the predicted structural characteristics of the WCS19 protein and its orthologs.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Identification of Wcs19 Homologous Genes

Sequence analyses have indicated that WCS19 shares two regions of homology with BCOR14b protein from barley (Crosatti et al., 1999) and WCOR14a protein from wheat (Tsvetanov et al., 2000). The first conserved region (I) lies at the N-terminal end of these proteins, whereas the second conserved region (II) lies at the C-terminal end of these proteins (Fig. 1). Between these conserved regions exists a variable region (V) that shows little identity (Fig. 1). These analyses suggested that WCS19 might be part of a small family of related proteins. The first strategy allowed us to identify three new genes, Rep14, Rep13, and Wcor14c (Fig. 1) that are expressed during growth at high levels of PSII excitation pressure and during cold acclimation. Based on their sequences and their degree of homology, these small proteins and the previously isolated WCOR14a, BCOR14b, and WCS19 can be classified into two groups. The first group contains WCOR14a, WCOR14c, REP13, and BCOR14b (Fig. 1), and members of this group show identities ranging from 63% to 95%. The second group contains WCS19, REP14; and BF625247 (Fig. 1); these proteins show identities ranging from 89% to 95%. However, when proteins from the first group are compared with proteins from the second group, the percent identity falls to between 46% and 55%.

View larger version (72K): [in this window] [in a new window]   Figure 1.   Alignment of three groups of proteins from wheat, rye, and barley sharing identities with WCS19. ClustalW alignment of wheat WCS19 (accession no. L13437), rye REP14 (accession no. AF491840), barley BF625247 (deduced from cv Morex; EST accession no. BF625247), wheat WCOR14a (accession nos. AF207545 and AF491838) and WCOR14c (accession no. AF491837), rye REP13 (accession no. AF491839), and barley BCOR14b (accession no. M60732) and BG369977n. ESTs from barley cv Morex, BG369977, BG369422, BE454426, and BE196464 were used to generate the sequence BG369977n of 999 bases. The sequence contained an open reading frame with three possible start codons and one stop codon encoding a protein of 293 amino acids. Residues within motifs resembling the consensus 11-mer repeat characteristic of group 3 LEA proteins are shown and numbered above the BG369977n protein. { distinguishes the groups of proteins. I, The conserved region coding for the putative signal peptide; II, the conserved C-terminal region; V, the variable region. The arrow indicates the putative cleavage site of the signal peptide determined with ChloroP. The boxed amino acids in region I highlight a sequence resembling the 14-3-3 recognition motif. Shaded amino acids in the largely -helical mature proteins represent regions predicted to adopt a coil conformation by PELE, a program that uses eight different algorithms to study secondary structure. , Gaps introduced to maximize alignment; *, identical residues; :, highly conserved residues; ., conserved residues.

Results from our second strategy revealed that both groups of proteins share significant homologies with a new barley protein BG369977n reconstituted from several overlapping expressed sequence tags (ESTs; Fig. 1). This large 293-amino acid protein was found to be more homologous to BCOR14b (58% identity and 79% similarity) than to BF625247 (45% identity and 68% similarity). The main difference is the variable region that extends for 191 amino acids in BG369977n compared with 50 in BCOR14b (Fig. 1). Moreover, sequence comparisons suggest that this large variable region shares a homology with several proteins that belong to group 3 LEA proteins. Careful analyses of the BG369977n protein shows that it could have as many as 14 imperfect copies of the 11-mer repeat that characterize group 3 LEA proteins (Dure, 1993; Fig. 1). Furthermore, the identification of two partial wheat ESTs homologous to BG369977n may suggest the existence of a third group of related proteins in the three grass species. Based on sequence analyses of the three groups and their relation to group 3 LEA proteins, they are named LEA3-L1 (WCOR14a, WCOR14c, REP13, and BCOR14b), LEA3-L2 (WCS19, REP14, and BF625247), and LEA3-L3 (BG369977n). LEA3-L stands for LEA group 3 protein-like.

Arabidopsis genome was also found to encode two large proteins, T10644 (266 amino acids) and BAB10116 (331 amino acids) that have several 11-mer repeats and, thus, may represent Arabidopsis homologs of the LEA3-L3 BG369977n protein. These two proteins share 31% to 36% identity and 59% similarity with the LEA3-L3 protein BG369977n from barley. Higher identities between the C-terminal regions of these two Arabidopsis proteins and the LEA3-L3 BG369977n were found (44%-50%) suggesting that this region may have important conserved properties for the function of these related proteins. Therefore, only the alignment of the conserved C-terminal regions from members of the three LEA3-L groups present in grass species and the two Arabidopsis proteins is presented in Figure 2A. In addition, four other proteins were found to have this highly conserved C-terminal region and were included in the alignment (Fig. 2A).

View larger version (35K): [in this window] [in a new window]   Figure 2.   Alignment and structure analysis of the conserved C-terminal region from different plants. A, ClustalW alignment and Multicoil analysis of the conserved C-terminal region (II) of proteins from the LEA3-L1, -L2, and -L3 groups with six proteins from other plants sharing a homology in this region. Arabidopsis proteins of 331 amino acids (BAB10116) and 266 amino acids (T10644) were deduced from genomic sequencing. PM32 (AF166485) is a 173-amino acid maturation protein identified in soybean (Glycine max; Chow et al., 1999). L42465 is a 190-amino acid LEA protein identified in Picea glauca (Dong and Dunstan, 1996). AU089534n is a 165-amino acid protein from Lotus japonicus that was reconstituted from three ESTs: AU089534, AU089020, and AU089581. BE592220n is an incomplete protein (102 amino acids) from sorghum (Sorghum bicolor) reconstituted from ESTs BE592220 and BE592752. The consensus sequence represents the compilation of amino acids conserved in at least 11 out 14 proteins. To the right of the alignment is a summary of the probabilities for the C-terminal region of the 14 proteins to form a coiled coil motif. The results were obtained using Multicoil scores based upon pair wise interactions for residues at distances 2, 3, and 4 apart with the dimeric table and at distances 3, 4 and 5 apart with the trimeric table (Wolf et al., 1997). Segment and minimum probability (%), High scoring segments with minimum 25 residues are defined by numbering relative to the first residue in the C-terminal region alignment. In this segment, the lowest total probability for a residue or subsegment to form a coiled coil motif is given as a percentage. Maximum total probability, The maximum probability for a residue or subsegment in the previously defined 25 residues segment. Trimeric oligomerization ratio, The trimeric score divided by the total score in the C-terminal region. To analyze L42465 from P. glauca, the Pro (a helix breaker) at position 10 was changed to Ala the most common amino acid at that position. B, Helical wheel projection of the consensus sequence generated from region II of the 14 proteins. Amino acids and the number of times they were repeated in at least 11 proteins are indicated for each position. Circled positions indicate single residues conserved in at least 11 proteins. *, Position 1 in Figure 2A.

Identification of Chloroplast and Mitochondria Sorting Signals

To determine the properties of the N-terminal end (region I) of the LEA3-L1, -L2, and -L3 polypeptides and their relationship with other proteins presented in Figure 2A, all polypeptides were analyzed with TargetP, Predotar, ChloroP, and Mitoprot softwares. The output from TargetP incorporates a measure of accuracy in the prediction, which was shown to vary in test cases from 99% accuracy with reliability class (RC) = 1% to 55% with RC = 5 (Emanuelson et al., 2000). Using this program, the LEA3-L1 (RC = 3) and -L2 (RC = 2) groups were predicted to be targeted to the chloroplast, whereas the LEA3-L3 protein (RC = 4) was predicted to be targeted to the mitochondria. Predictions using Predotar follow the same pattern with localization probabilities for the LEA3-L1 group (chloroplast P = 0, 469), L2 group (chloroplast P = 0, 924) and L3 protein (mitochondrion P = 0, 708). Because chloroplast and mitochondria sorting signals share common properties (Emanuelson et al., 2000), programs such as ChloroP and Mitoprot predicted, respectively, a chloroplast and mitochondria localization for these three groups of proteins. Analysis of the Arabidopsis, P. glauca, and L. japonicus proteins (Fig. 2A) also revealed that they are likely targeted to the chloroplast and/or the mitochondria. This analysis suggests that the proteins presented in Figure 2A share an additional important feature, which is a sorting signal predicted to target them to organelles.

Secondary and Tertiary Structure Predictions for the Mature Proteins

The mature proteins of the LEA3-L1 and L2 groups (Fig. 1) share several characteristics such as similar size (8.5-9.6 kD) and acidic pI (4.1-4.5). The analyses of these proteins by PELE reveals that the variable region V is predicted to contain segments in the coil conformation (Fig. 1; shaded amino acids). These segments may represent breaks in the largely -helical proteins and, thus, induce slightly different properties in both groups. However, biochemical evidence is needed to support these predictions.

To get a better idea of the structural features or properties of these proteins, a consensus helical wheel projection was generated for the conserved -helical region II of all proteins presented in Figure 2A. Analysis of the projection in Figure 2B reveals a 140^o hydrophobic side that is bordered on either side (40 and 60^o) by basic or polar amino acids. In the 40^o border region, only one protein in 14 contained an acidic amino acid, whereas in the 60^o region, two proteins contained an acidic amino acid. It is important to note that the majority of positions showing a high identity in the consensus C-terminal region are concentrated to one side of the amphipatic -helix.

The strict conservation of amphipatic character and identity in certain amino acid positions in the C-terminal end raised the possibility that this region may be involved in some sort of tertiary interaction. One type of structure that is known to require an amphipatic property is the coiled coil motif (Lupas et al., 1991; Berger et al., 1995; Wolf et al., 1997). Thus, proteins containing the conserved C-terminal end were analyzed using the program Multicoil (Wolf et al., 1997). This program helps in identifying the coiled coil motif in proteins and the oligomerization states. The results of this analysis are presented in Figure 2A and show that seven out 14 proteins have segments with minimum total probability scores ranging from 28% to 50%, with most of the probability coming from the trimeric score indicating a trimeric oligomerization. Overall, this analysis reveals that homologs from different species share distinct properties such as an amphipatic character and a conserved C-terminal region that may be involved in oligomerization.

Expression and Chromosomal Localization of LEA3-L1 and LEA3-L2 Genes

The results in Figure 3A illustrate the RNA hybridization blot for the 600-bp transcript corresponding to the rye LEA3-L2 (Rep14) gene. The blot for the rye LEA3-L1 (Rep13) gene gave identical results (result not shown). Both genes appear to be regulated by PSII excitation pressure because they exhibited higher expression in rye plants grown at 20/800 for 14 d or 5/250 for 40 d compared with plants grown at 20/250 or 5/50 (Fig. 3A). The expression and regulation of rye LEA3-L1 and -L2 by PSII excitation pressure are consistent with previous reports for the wheat Wcs19 gene (Chauvin et al., 1993; Gray et al., 1997).

View larger version (19K): [in this window] [in a new window]   Figure 3.   Expression analysis and accumulation of LEA3-L2 proteins in rye and wheat. A, Transcript accumulation of rye LEA3-L2 gene using Rep14 as probe. Equal amounts of total RNA (5 µg) were separated by agarose gel electrophoresis in the presence of formaldehyde and transferred to a nitrocellulose membrane. 20/800, 20/250, and 20/50 represent rye plants grown at 20°C and at 800 µmol m2 s1, 250 µmol m2 s1, or 50 µmol m2 s1, respectively. 5/250 and 5/50 represent rye plants grown at 5°C and at 250 µmol m2 s1 or 50 µmol m2 s1, respectively. The size of Rep14 transcript is indicated on the right in bases. B, Boiling solubility of LEA3-L2 proteins. NA and A, Soluble proteins (5 µg) from leaves of rye plants (cv Musketeer) grown at 20/250 (24 d) and 5/250 (40 d), respectively. In addition, soluble proteins from 5/250 leaves were heated at 100°C for the time indicated. Insoluble (I) and boiling soluble (S) fractions were analyzed by immunoblotting as described in "Materials and Methods." The size of the mature REP14 in kilodaltons is indicated at right. C, Accumulation kinetics of the rye LEA3-L2 protein during high-light exposure. Rye plants grown at 20/250 for 24 d were shifted to 20/800 conditions for the number of hours (h) and days (d) indicated. Soluble proteins (5 µg) were analyzed by immunoblotting. To show uniform loading, the Coomassie Blue-stained Rubisco subunit (55 kD) is shown. D, Accumulation kinetics of the rye LEA3-L2 protein during low-temperature acclimation. Plants grown at 20/250 for 24 d were shifted to 5/250 conditions for the time indicated, and soluble proteins (5 µg) were analyzed by immunoblotting. E, Accumulation of LEA3-L2 proteins in different wheat and rye cultivars cold acclimated for 49 d. Immunoblot analysis was done with soluble proteins (5 µg) from leaf tissues of spring rye and wheat cv Gazelle (Gaz), Glenlea (Glen), Manitou (Man), and Chinese Spring (CS) and from winter rye and wheat cv Puma (Puma), Besostoya (Bes), Cheyenne (CNN), and Ulian (Ulian).

The ditelocentric series of cv Chinese Spring wheat in which one homologous pair of chromosome arms is missing in each line, was used to determine which chromosome arms carry members of the LEA3-L1 and LEA3-L2 groups (results not shown). Results using this series have shown that: (a) wheat LEA3-L1 and L2 genes reveal a different pattern of genomic organization indicating that these two genes are sufficiently different to be used as probes for localization; (b) the wheat LEA3-L1 genes were mapped to the long arms of chromosome 2, as was Bcor14b (Vàgujfalvi et al., 2000); and (c) the wheat LEA3-L2 genes could not be mapped with the available genetic stocks.

Accumulation of LEA3-L2 Protein (WCS19) during Exposure to High Light and Cold Temperature

To determine whether the LEA-L2 (REP14) protein could remain soluble upon boiling, protein extracts from 5/250 rye grown leaves were boiled for 30 min. Immunoblot results in Figure 3B indicate that the REP14 protein remained soluble as found in all LEA proteins. Although the anti-WCS19 antibody was raised against a LEA3-L2 protein, we cannot rule out the possibility that this immune serum also cross-reacts partially with proteins from the LEA3-L1 group. However, the anti-WCS19 antibody does not seem to cross-react with LEA3-L3 proteins because no additional bands were detected.

The immune serum was also used to measure the accumulation kinetics of proteins during exposure to high-excitation pressure conditions induced by high light (20/800) and low temperature (5/250). When rye plants grown at 20/250 were shifted to high-excitation pressure conditions, the anti-WCS19 antibody detected a 14-kD protein that accumulated gradually and reached a maximum level near 12 d at 20/800 or 10 d at 5/250 (Fig. 3, C and D). The 14-kD protein, which may represent the mature protein, is not in agreement with the calculated molecular mass (9.6 kD) deduced from the mature rye REP14 protein. Such differences have been observed previously with different plant stress proteins (Houde et al., 1992a).

In addition, the accumulation level of LEA3-L2 proteins during low temperature exposure in a number of cultivars differing in their capacity to cold acclimate was evaluated. The results in Figure 3E show that LEA3-L2 proteins accumulate to a higher level in cultivars with an enhanced capacity to develop freezing tolerance. The more freezing tolerant wheat cv Bes, CNN, and Ulian (LT₅₀ of 16.4, 19, and 19.5°C, respectively) maintained a higher level of WCS19 protein compared with the less freezing tolerant cv Glen (5.5°C), Man (6.2°C), and CS (9.4°C). A similar result was obtained with the freezing tolerant rye cv Puma (24.8°C) compared with rye cv Gaz (6.5°C). These results support the concept that LEA3-L2 proteins are associated with the plant capacity to develop freezing tolerance.

Localization and Immunolocalization of LEA3-L2 Proteins

The results in Figure 4A show that the wheat LEA3-L2 protein (WCS19) accumulates specifically in the leaves of wheat during cold acclimation. No signal was detected in the roots nor in the crown, suggesting that WCS19 accumulates exclusively in the photosynthetic tissue. In addition, analysis of total chloroplast extracts and stromal and thylakoid fractions from cold-acclimated wheat shows that the WCS19 protein accumulates specifically in the stroma (Fig. 4B).

View larger version (23K): [in this window] [in a new window]   Figure 4.   Wheat LEA3-L2 (WCS19) tissue distribution and subcellular localization. A, Immunoblot analysis of soluble leaf proteins (5 µg) present in different tissues of wheat (cv Fredrick) grown under low-temperature conditions. NA, Nonacclimated plants; A, plants cold acclimated for 2 weeks. B, Immunoblot analysis of proteins (5 µg) present in different chloroplast compartments. Legend as in A.

To further confirm whether LEA3-L2 proteins were truly localized to the chloroplast, we used immunocytochemical localization. When sections of leaf tissues from control rye plants grown at 20/250 were incubated with the anti-WCS19 antibody and with gold-conjugated antiserum to rabbit immunoglobulins, little deposition of gold particles was observed in the chloroplast (0.49 ± 0.09 gold particles µm² for n = 10 chloroplasts counted; Figs. 5C and 6). On the other hand, leaf sections from rye plants grown at 5/250 revealed the strongest deposition of gold particles in the chloroplast (1.38 ± 0.13 gold particles µm² for n = 10 chloroplasts counted; Figs. 5A and 6). In addition to being present in the chloroplast stroma, the rye LEA3-L2 protein (REP14) was also found to be associated with the periphery of the thylakoid membranes but not membrane bound (result not shown). Finally, leaf sections from plants grown at 20/800 revealed an intermediate level of deposition of gold particles in the chloroplast (0.82 ± 0.09 gold particles µm² for n = 10 chloroplasts counted; Figs. 5B and 6).

View larger version (87K): [in this window] [in a new window]   Figure 5.   Electron microscopy of rye leaf sections incubated with anti-WCS19 antibody. A, Plants grown at low temperature for 40 d (5/250). B, Plants grown at high light for 14 d (20/800). C, Nonacclimated plants grown at 20/250 for 24 d. The magnification of each image is 22,000×. The images shown are representative of typical WCS19 labeled chloroplasts from plants grown under the given conditions.

View larger version (17K): [in this window] [in a new window]   Figure 6.   Density of the immunogold labeling obtained with the anti-WCS19 antibody in chloroplasts of rye plants exposed to three growth conditions. The number of gold particles per square micrometer of the chloroplasts was determined using the Northern Eclipse Image Analysis software package. This allowed the analysis of digitized images of representative chloroplasts from WCS19 labeled sections of rye leaves grown under the indicated conditions. Bars represent mean ± SE, n = 10.

To confirm the specificity of chloroplast labeling obtained with the anti-WCS19 antibody, we conducted similar experiments using sections of 20/800 grown leaves, which were immunolabeled with antibodies raised against the LHCII polypeptides of spinach (Spinacia oleracea) or the Rubisco holocomplex of rye (results not shown). Results confirmed the integrity of the fixed tissue because it was possible to observe that the Rubisco immunogold label was primarily associated with the stromal compartment, whereas the LHCII immunogold label was primarily associated with the thylakoid membranes of the chloroplast. Additional control tests including the use of preimmune serum or distilled water did not reveal any deposition of gold particles.

Thus, it appears that LEA3-L2 is either localized to the chloroplast stroma or loosely associated with thylakoid membranes, which confirms the results of the biochemical fractionation. Therefore, members from both groups, rye and wheat LEA3-L2 and barley LEA3-L1 protein (BCOR14b; Crosatti et al., 1999), are now known to specifically accumulate in the stroma. This suggests that both groups of proteins have a functional role in the chloroplast during acclimation to growth at high-excitation pressures induced by high light and/or low temperature.

Effects of the Constitutive Expression of the Wheat LEA3-L2 (WCS19) on Cold Acclimation and Freezing Tolerance in Arabidopsis

Although it is well documented that the accumulation of LEA3-L2 mRNAs and proteins is increased in response to cold temperature, it is still not known what function these proteins might have in the cold acclimation process. Using Arabidopsis plants that have one representative of the LEA3-L2 genes (Wcs19) inserted under the control of the cauliflower mosaic virus (CaMV) 35S promoter, we examined the freezing tolerance of plants grown at 20°C and plants that had been grown at 20°C but shifted to low temperature (5°C) for 7 d. To ensure that the protein was being expressed, soluble leaf protein extracts of the three transgenic lines (B, C₄₈, and C₇₁) were analyzed by immunoblots. It can be observed from the data in Figure 7 that, whereas the level of the WCS19 protein present in Arabidopsis transgenic plants is lower than in cold-acclimated (5/250) grown rye or wheat, the protein is present in transgenic plants but absent in wild-type plants. The freezing tolerance of Arabidopsis leaves was determined using the electrolyte leakage technique and reveals that in nonacclimated plants, the LT₅₀ values for the transgenic plants were similar to wild-type plants, indicating no increase in the freezing tolerance. The LT₅₀ of wild-type leaves was 4.0 ± 0.4°C compared with 3.5 ± 0.3°C in transgenic leaves (line B), 4.0 ± 0.5°C for line C₄₈, and 4.6 ± 1.4°C for line C₇₁ (Fig. 8A). However, when the freezing tolerance of the cold-acclimated leaves (plants shifted from 20°C to 5°C for 1 week) was compared, the transgenic lines were significantly more freezing tolerant. The LT₅₀ of leaves from wild-type plants shifted to 5°C for 1 week was 5.9 ± 0.4°C compared with 7.2 ± 0.5°C for line B, 7.9 ± 0.3°C for line C₄₈ and 8.1 ± 0.3°C for line C₇₁ (Fig. 8B). These differences were statistically significant (P = 0.02; 0.0018 and 0.0013 two-tailed Student's t test respectively for line B, C₄₈ and C₇₁). Moreover, when the LT₅₀ values of the three transgenic lines are combined, the transgenic group remained statistically more freezing tolerant (7.7 ± 0.39 P = 0.008) compared with the wild type. Thus, it appears from the results obtained that the WCS19 protein may play a role in the cold acclimation process and that its constitutive expression could enhance freezing tolerance.

View larger version (35K): [in this window] [in a new window]   Figure 7.   Accumulation of the wheat LEA3-L2 (WCS19) protein in transgenic Arabidopsis grown at 20°C/100 µmol m2 s1. Soluble proteins (5 µg) from rye leaves (5/250) were used as positive control. The immunoblot was overexposed to show that the anti-WCS19 antibody does not recognize any endogenous proteins in wild-type (WT) Arabidopsis plants. B, C48, and C71 represent the transgenic lines. NA, Nonacclimated; A, plants cold acclimated for 1 week.

View larger version (17K): [in this window] [in a new window]   Figure 8.   Effect of the constitutive expression of wheat LEA3-L2 (WCS19) on the freezing tolerance of Arabidopsis leaves. A, Electrolyte leakage curves from leaves of plants grown at 20°C. B, Electrolyte leakage curves from leaves of plants shifted from 20°C to 5°C for 1 week. Wild-type () and transgenic lines B, C48, and C71 (, , and , respectively). From the electrolyte leakage curves, the LT50 was determined as the temperature at which 50% of the ions leaked out the leaf and used as an estimate of freezing tolerance. The experiments were done four times (n = 3 leaves for each temperature) for each transgenic lines and values represent mean ± SE.

Effects of the Constitutive Expression of the Wheat LEA3-L2 (WCS19) on the Tolerance of Arabidopsis Plants to Photoinhibition

Cold acclimation is correlated quite strongly with increased tolerance to photoinhibition, and maximum LT₅₀ is dependent upon low temperature and high-light levels (Öquist et al., 1993; Gray et al., 1997; Pocock et al., 2001). Therefore, because wheat and rye LEA3-L2 proteins are localized in the chloroplast, and their transcript abundance is regulated in response to the redox-state of PSII, we examined whether the presence of the WCS19 protein in the leaves of transgenic Arabidopsis plants resulted in increased resistance to photoinhibition.

Detached leaves of Arabidopsis plants that had been grown at 20°C or grown at 20°C and transferred to 5°C were exposed to 1,600 µmol photons m² s¹ for 3 h. F_V/F_M was determined at 30-min intervals to estimate the level of photoinhibition of PSII. The presence of the WCS19 protein in the leaves of plants grown at 20°C had no apparent effect on the tolerance to photoinhibition when wild-type and transgenic lines are compared (Fig. 9A). The leaves of both wild-type and transgenic plants were photoinhibited to the same degree, the dark adapted F_V/F_M after 3 h at photoinhibitory conditions was reduced to 35.5% ± 1.9% of the initial value for wild-type plants compared with 31.4% ± 2.3% for line B, 34.8% ± 1.7% for line C₄₈, and 37% ± 1.6% for line C₇₁. However, when the plants had been shifted from 20°C to 5°C for 1 week, the presence of the WCS19 protein did appear to indicate minimal but significant protection (P = 0.03, two-tailed Student's t test) of the leaves of the C₇₁ line only from photoinhibition after 3 h (Fig. 9B). In leaves of wild-type plants, F_V/F_M was decreased to 31% ± 2.1% compared with 36.3% ± 5.4% for line B, 37.3% ± 3.8% for line C₄₈ and 43.5% ± 3.9% for line C₇₁.

View larger version (15K): [in this window] [in a new window]   Figure 9.   The effect of constitutive expression of the wheat LEA3-L2 (WCS19) protein on photoinhibition of Arabidopsis leaves. A, Photoinhibition of detached leaves from plants grown at (20°C). B, Photoinhibition of detached leaves from plants shifted from 20°C to 5°C for 1 week. Symbols as in Figure 8. The experiments were done four times (n = 3 leaves for each time) for each transgenic lines and values represent mean ± SE. Where not visible, the error bars are smaller than the symbols.

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Results presented in this report show that grass species such as rye, wheat, and barley contain at least three different groups of LEA3-like proteins that can be classified as small (LEA3-L1 and LEA3-L2 groups) and large (LEA3-L3 group) proteins. These three groups and their related proteins in other plants share several common characteristics. An important feature of these proteins is the presence of a sorting signal that is predicted to target them either to the chloroplast, mitochondria or both. Although a few cases are known where one sorting signal can route with similar efficiency a protein to chloroplast and mitochondria (Creissen et al., 1995; Chow et al., 1997; Akashi et al., 1998; Menand et al., 1998), such dual targeting signals are predicted to be quite rare (Emanuelson et al., 1999). Detailed analysis of the sorting signals present in the LEA3-L1, -L2, and -L3 groups of proteins (Fig. 1) revealed that the LEA3-L2 proteins contain a sequence sharing a high identity with the 14-3-3 recognition motif (R S X S/T X P; May and Soll, 2000; Fig. 1). On the other hand, this sequence is less conserved in the LEA3-L1 group and nonexistent in the LEA3-L3 group. It was shown that when this recognition motif was phosphorylated, it interacts with 14-3-3 proteins (May and Soll, 2000). This complex-bound precursor was much more efficient for import into isolated chloroplasts than was the free precursor protein (May and Soll, 2000). The high identity between these three groups sorting signals suggest that they may have diverged from a common ancestor rather recently, and/or some primary sequence is being conserved because of some function other than just directing import to the chloroplast. The fact that, to date, most of the newly identified members of this new class of LEA3-like proteins are predicted to be targeted to organelles is intriguing and suggests that properties and functions inferred from their sequences may have evolved to work in a specific organelle environment. Under stress conditions, it will be critical to protect the chloroplasts and mitochondria to maintain cellular energy production. Thus, these groups of proteins could be very important for stress survival of the cell and/or organism.

These proteins differ on at least one aspect, which is the length of the sequence between their sorting signal and the conserved C-terminal region. This variable region (V) shares a similarity with group 3 LEA proteins, and allowed us to classify these proteins as distantly related to this group. A main characteristic of group 3 LEA proteins is the presence of several copies of an 11-mer repeat (Dure, 1993). Proteins containing several copies of this repeat have been proposed to function by scavenging ions during desiccation (Dure, 1993). It should be noted that at least another function for group 3 LEA proteins has been inferred, because it was shown that a protein containing a sufficient number of the 11-mer repeat had some cryoprotective activity (Honjoh et al., 2000). Because the smaller LEA-like proteins identified in this study contain little if any of the 11-mer repeats, it is doubtful that they are involved in scavenging ions. This raises an interesting dilemma concerning their evolution, suggesting that they evolved new functions or that there is a common, yet unknown, function of group 3 LEA proteins that relies on their amphipatic character or other properties. Although it was not studied in great detail, it was noted during the analysis with Multicoil that several of the large LEA-like proteins identified in this study contained segments in their variable regions that showed high probabilities (up to 80%) of forming dimeric or trimeric coiled coils (results not shown). The properties of these hypothetical coiled coils and their relationship with the oligomerization-forming potential of the C-terminal end is unknown but stresses the importance of continuing the characterization of this new class of protein and group 3 LEA proteins in general. As an alternative, these amphipatic proteins may function by stabilizing and/or protecting partially denaturated proteins or membranes through their hydrophobic interactions. Such a function has been proposed for the amphipatic -helices present in group 2 LEA proteins (Close, 1996).

For the moment, several indications suggest that the small chloroplastic LEA-like proteins (LEA3-L1 and -L2) evolved from larger proteins in group LEA3-L3 through a deletion of their variable region, and that this may have occurred independently in several plants after the divergence of monocots and dicots. This statement is based on the observation that Arabidopsis genomic sequences and EST databases have only representatives of the large proteins (T10644, BAB10116). Initially, we believed that the Arabidopsis protein COR15a (Lin and Thomashow, 1992) was a homolog of proteins from the LEA3-L1 and -L2 groups even if it shared little homology. This was based on common properties such as similar size, amino acid composition, chloroplast localization, and amphipatic character. On the other hand, this study suggests that COR15a does not contain the conserved C-terminal region and shares little identity with the large Arabidopsis proteins. However, we cannot rule out the possibility that COR15a is a functional homolog of both LEA3-L1 and L2 groups of proteins since the properties responsible for the function of COR15a, LEA-L1 and L2 proteins are still unknown. The results obtained by Artus et al. (1996) have shown that COR15am improves the freezing tolerance of chloroplast frozen in situ and of protoplast frozen in vitro by inhibiting the formation of the hexagonal II phase, which is a major cause of membrane damage.

Ongoing genomic and EST sequencing projects will help to determine the exact number and characteristics of this new class of proteins in plants. Increased sequence data may allow a correlation to emerge between the appearance of the small proteins like LEA3-L1 and -L2 and a phenotypic adaptation. It was rather surprising that from the 243,000 rice (Oryza sativa), maize (Zea mays), and sorghum ESTs present in GenBank, only two related ESTs were found. However, previous Southern-blot experiments using Wcs19 as probe did not detect any signal in the cold-sensitive species corn and rice (result not shown). This suggests that cold-sensitive species may not have evolved or else may have lost homologs of LEA3-L1 and -L2 groups.

To investigate the function of LEA3-L2 group of proteins, Wcs19 was constitutively expressed in Arabidopsis to determine whether it has a discernable effect on tolerance to freezing and photoinhibition. The results presented in this study indicate that the observed increase in freezing tolerance of cold-acclimated transgenic plants was statistically significant. However, it was not clear why the WCS19 protein only improved the freezing tolerance of cold-acclimated plants. It is possible that WCS19 needs other chloroplast components or endogenous COR proteins that only accumulate in response to low temperature to accomplish its function, because it acts as part of a complex or in synergy with these components. WCS19 may interact with specific lipids and/or proteins in the membrane and/or stroma. Therefore, if the(se) other component(s) accumulates little, if not at all, in 20°C grown plants, it could explain the absence of phenotypic effect on freezing tolerance as a result of the increased level of WCS19.

As a alternative, the WCS19 that is constitutively expressed at 20°C could be non-functional. It may only become functional when activated in response to a temperature shift (phosphorylation, pH-induced conformational change, etc.). As mentioned previously, one of the most important characteristics of the LEA3-L1, -L2, and -L3 groups of proteins is in the conserved C-terminal sequence. This region is predicted to contain two different windows composed of several heptad repeats that may be involved in a tertiary interaction called a trimeric coiled coil. However, it should be noted that a more precise designation of the oligomerization ratio would have to await further work because it is known that Multicoil can detect other types of tertiary structures such as a four -helical bundle (33% total probability) and a tetrameric and pentameric coiled coils (54%; Wolf et al., 1997). This oligomerization potential raises the possibility that homo-and hetero-oligomerization could create a vast array of polypeptide complexes with different or overlapping properties. In addition, the prediction that residues in the conserved C-terminal sequence can occupy two different positions in a heptad repeat raises the additional possibility that the segment may have the ability to shift to a second state based on environmental cues. Such a shift could induce a conformational change in the rest of the polypeptide and hence, change the functional properties of the protein complex. Although no experimental data exists to support this suggestion, it is known that a drop in pH is instrumental in driving a region of influenza hemaglutinin to adopt a coiled coil structure provoking a conformational change (Carr and Kim, 1993). Thus, changes in chloroplast stromal pH or other modifications that occur during exposure to low temperature may alter the structure of WCS19 and help to increase the freezing tolerance of plants.

In rye and wheat, increased freezing tolerance (LT₅₀) was correlated with resistance to photoinhibition (Pocock et al., 2001) and with the ability of plants to maintain Q_A in a more oxidized state (Öquist and Huner, 1992; Öquist et al., 1993). Because the wheat LEA3-L2 protein (WCS19) was associated with an increase in freezing tolerance of the transgenic Arabidopsis leaves and was localized to the chloroplast, we also determined whether the presence of WCS19 was associated with an increased resistance to photoinhibition. The preliminary results show that although the C₇₁ line exhibited statistically (P < 0.05) increased resistance to photoinhibition, the transformants shifted to low temperature exhibited generally minimal changes in susceptibility to photoinhibition compared with wild type. Given that rye plants exhibited greater resistance to photoinhibition and exhibited much higher levels of WCS19 than the Arabidopsis transformants, it is possible that Arabidopsis lines with higher levels of WCS19 expression may be required to observe larger differences in resistance to low-temperature photoinhibition. Regardless, the photoinhibition results are consistent with the LT₅₀ data and confirm that there must be some additional low temperature-induced factor required for WCS19 to increase resistance to photoinhibition.

In summary, we have shown that the wheat LEA3-L2 (WCS19) is a stromal protein that belongs to a new class of organelle-targeted group 3 LEA proteins. The constitutive expression of the WCS19 protein in Arabidopsis was shown to protect cold-acclimated leaves from freeze-induced damage. Despite the observed cryoprotective activity of LEA3-L2 proteins, their exact roles are not clear. Further studies are required to determine their mode of action, to determine whether they act as part of a complex, and to identify the chloroplast component(s) protected by this family of proteins during stress conditions.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Plant Material and Growth Conditions

Seeds of winter rye (Secale cereale L. cv Musketeer) and wheat (Triticum aestivum L. cv Fredrick and Norstar) were germinated in coarse vermiculite and grown at temperatures of either 20/16°C or 5/5°C (day/night) with a 16-h photoperiod in controlled environment chambers (Conviron, Manitoba, Canada) as described previously (NDong et al., 2001). Growth irradiance was adjusted to 50 or 250 µmol m² s¹ at 5°C (5/50 and 5/250, respectively) and 50, 250, or 800 µmol m² s¹ at 20°C (20/50, 20/250, or 20/800, respectively).

Growth conditions and LT₅₀ (temperature at which 50% of the plants are killed) determinations for the wheat and rye cultivars were as described previously (Limin and Fowler, 1988). After 49 d of acclimation, the LT₅₀ values were as follows: spring wheat cv Glenlea, 5.5°C; cv Manitou, 6.2°C; and cv Chinese Spring, 9.4°C. Winter wheat cv Fredrick, 15.6°C; cv Besostoya, 16.4°C; cv Cheyenne, 19°C; cv Ulian, 19.5°C; and cv Norstar, 21.2°C. Rye cv Gazelle, 6.5°C; and cv Puma, 24.8°C.

Seeds from Arabidopsis wild-type plants (ecotype Columbia) and transgenic Wcs19 lines (F3 populations) were germinated and grown in a mix composed of three parts Promix soil-less mix (Premier Brands, Rivière-du-Loup, Canada) and one part vermiculite. The plants were exposed to an 8-h photoperiod with a growth irradiance of 100 µmol m² s¹. Plants that had reached the fourth leaf stage were transferred to individual pots and grown until the appearance of 12 leaves at constant temperature of 20°C before determining resistance to photoinhibition and freezing-induced cell damage. To examine the effects of low-temperature acclimation on both resistance parameters, plants that had been grown at 20°C were transferred to 5°C for 7 d.

Plant Transformation

The Wcs19 cDNA was excised using SmaI and EcoRV that cut in the polylinker region of pBluescript. The insert was ligated into the BamHI-SacI restricted and Klenow-treated pBI121 vector between the CaMV 35S promoter and the nopaline synthase (NOS) terminator. The chimeric construct 35SCaMV-Wcs19-NOS with the correct orientation was introduced into Agrobacterium tumefaciens strain GV3101 (Koncz and Schell, 1986). The floral dip transformation protocol of Clough and Bent (1998) was used to transform Arabidopsis with A. tumefaciens carrying the construct.

Selection of putative transformants was performed as described by Clough and Bent (1998) with slight modifications. To select for transformants, sterilized seeds were suspended in 0.1% (w/v) sterile agarose and plated on 50 µg mL¹ kanamycin and 500 µg mL¹ cefotaxime. Three transgenic lines (F0 population) that grew on kanamycin were transferred to pots and moved into a growth chamber. A small leaf sample from these transgenic plants was tested for the presence of the WCS19 protein by western analysis. These three lines were advanced to the F3 population and were found to be phenotypically uniform for kanamycin resistance and constitutive production of WCS19 protein.

Assessment of Freezing Tolerance and Resistance to Photoinhibition

The freezing tolerance of the transgenic Arabidopsis lines was estimated using the ion leakage technique as described previously (Gray et al., 1997; Pocock et al., 2001). Whole leaves were harvested from the plant, wrapped in moist cheesecloth, chilled rapidly to 2°C, and left for 1 h. Ice nucleation was then induced with an ice chip. The leaf samples were then further cooled at a rate of 2°C per hour to a minimum of 20°C and sampled at each 2°C interval. The percent ion leakage was calculated as the ratio of the conductivity before and after boiling. The percent ion leakage was plotted versus temperature, and a sigmoidal function was fitted to the data using the Microcal Origin software package (Microcal Software Inc., Northampton, MA). The temperature at which 50% of the total ion leakage occurred was determined from the sigmoidal curve and used to estimate the LT₅₀.

Photoinhibition of photosynthesis was induced by exposing detached leaves of Arabidopsis to 1,600 µmol m² s¹ at 5°C. The photoinhibitory treatment occurred under ambient O₂ and CO₂ conditions, and the leaves were kept moist to minimize the effects of desiccation during the high-light exposure. Estimates of PSII photochemistry (F_V/F_M) were used to monitor susceptibility to photoinhibition. To measure F_V/F_M, the leaves were dark adapted for 10 min at room temperature and the F₀ and F_M values were determined using a Plant Stress Meter (PSM Chlorophyll Fluorometer, Biomonitor S.C.I. AB, Umeå, Sweden; Öquist and Wass, 1988).

Identification and Characterization of Rye and Wheat Genes Sharing Identities with Wcs19

Recent analysis has revealed that nucleotides 1 to 285 of the original Wcs19 (Chauvin et al., 1993) are 98% identical to an unrelated EST and, thus, represent a fusion product during cDNA library construction. The GenBank file has been corrected to reflect this new information.

Two strategies were used to identify genes related to Wcs19. As part of our first strategy, three approaches were used to physically isolate homologous genes. First, the complete Wcs19 cDNA (Chauvin et al., 1993) was used to screen 100,000 plaques from the rye cDNA library prepared from plants grown at 20°C and an irradiance of 800 µmol m² s¹ (20/800; NDong et al., 2001). Five plaques showing a strong hybridization signal were selected and purified using standard molecular biology techniques (Sambrook et al., 1989). A clone of approximately 600 bases, Rep14 (for rye excitation pressure) was sequenced, and the deduced amino acid sequence is presented in Figure 1. The gene was found to encode the rye ortholog of Wcs19.

Second, the rye 20/800 cDNA library was differentially screened with both a 183-bp EcoRI-PstI fragment of Rep14 (encoding a putative chloroplast transit peptide) and a 430-bp PstI-XhoI fragment. Eight plaques showing a specific hybridization signal with the 183-bp fragment were selected, purified, and analyzed by terminal sequencing. This revealed that five clones were identical and that only these clones showed a significant homology with the chloroplast signal peptide. One of these, Rep13 was sequenced, and the deduced protein sequence is presented in Figure 1. This Rep13 gene was found to encode the rye ortholog of Bcor14b, Wcor14a, and Wcor14b (Crosatti et al., 1999; Tsvetanov et al., 2000).

Finally, PCR was used to search for other wheat genes containing the putative chloroplast transit peptide. Poly(A⁺) RNA was isolated from 2-d cold-acclimated wheat cv Fredrick as described (Danyluk and Sarhan, 1990) and reverse transcribed with the first strand cDNA synthesis kit from Roche Molecular Biochemicals (Summerville, NJ) using the primer 5'-GGCCAAGCTTATCGATCC(T)₁₇-3'. PCR was performed using Taq DNA polymerase (Amersham Pharmacia Biotech, Uppsala) with the following primers: 5'-GATGGCTTCTTCTTCCGTGCTGCTCG-3' and 5'-GGCCAAGCTTATCGATCC-3'. The PCR products were cloned into the pSTBlue-1 vector (Novagen, Madison, WI) using the Perfect Blunt Cloning kit (Novagen). Twenty-seven clones with inserts were sequenced with the dye terminator sequencing kit (Beckman Coulter, Inc., Fullerton, CA) and run on a Beckman CEQ 2000 sequencer. Sequencing revealed that 19 of them showed identities with Wcor14a and Wcor14b, two were identical to Wcs19, and six were false positives and did not show any homology with the transit peptide. Clones identical to Wcor14a and similar to Wcor14b were sequenced. Their deduced protein sequences are presented in Figure 1 under the names of WCOR14a and WCOR14c. This WCOR14c polypeptide differs from WCOR14b (Tsvetanov et al., 2000) by six mismatches and does not contain the one-nucleotide deletion that causes a frame shift mutation in the protein coding sequence.

In our second strategy, the nucleotide and the amino acid sequences of isolated genes were subdivided into three segments (Fig. 1) and used to search the National Center for Biotechnology Information non-redundant, EST, and high-throughput genomic sequence databases with different BLAST programs (Altschul et al., 1997). Sequences of genes or ESTs showing high scores were downloaded and analyzed in more detail. In the case of ESTs, several identical but overlapping ESTs were used to generate a more complete sequence for analysis.

Sequence analysis was done at sites such as the Canadian Bioinformatics Resource (http://www.cbr.nrc.ca/), Biology Workbench (http://workbench.sdsc.edu/), and Expasy (http://www.expasy.ch/). Alignment tools used were ClustalW from BCM launcher analyses service (Baylor College of Medicine search launcher; http://searchlauncher.bcm.tmc.edu) and Gap (http://www.cbr.nrc.ca/); programs for predicting subcellular localization were TargetP (Emanuelson et al., 2000), Predotar (http://www.inra.fr/internet/produits/predotar), ChloroP (Emanuelson et al., 1999), and Mitoprot (Claros and Vincens, 1996); and programs for analyzing secondary and tertiary structures were PELE (http://workbench.sdsc.edu/) and Multicoil (Wolf et al., 1997). Northern- and Southern-blot analyses were done as described previously (Limin et al., 1997; NDong et al., 1997).

Production and Purification of WCS19 Antibodies

The Wcs19 cDNA was digested with MboI and subcloned into the BamHI site of pTrcHisC (Invitrogen, Carlsbad, CA). The clone with the correct orientation was used to express Wcs19 as a N-terminal His-tagged fusion product in E. coli. The WCS19 protein was purified by affinity chromatography on a His-Bind resin (Novagen) and then separated on a 12% (w/v) SDS-polyacrylamide gel. The expressed protein was excised and electroeluted for 3 h. Immune serum from rabbits injected with the WCS19 protein was found to cross-react with bacterial proteins. As a first step in purifying the specific WCS19 antibodies, the immune serum was first depleted of these cross-reacting antibodies. For this purpose, proteins from non-transformed bacteria eluting in the wash buffer (His-bind manual, Novagen) were concentrated using Centricon 10 (Amicon Inc., Beverly, MA), dialyzed against 120 mM HEPES (pH 7.5), and coupled at a concentration of 2 mg mL¹ with Affi-gel-10 beads in the presence of 80 mM CaCl₂ (Bio-Rad manual, Hercules, CA). The immune serum was passed repeatedly on this column until monitoring showed no cross-reaction with bacterial proteins. The resulting immune serum was further purified on an affinity column containing Affi-gel-10 beads that were previously coupled to the WCS19 protein at a concentration of 2 mg mL¹ in the presence of 80 mM CaCl₂. Specific anti-WCS19 antibodies were eluted with 0.1 M Gly (pH 2.5) and neutralized immediately with 1 M KPO₄ (pH 8.0), concentrated using a Centricon 10 (Amicon Inc.), and frozen.

Protein Extraction and Immunoblot Analysis

Soluble proteins were extracted from frozen plant tissue as described (Houde et al., 1992b). To determine the boiling solubility, a fraction of the supernatant was boiled for 10, 20, and 30 min and boiling soluble proteins were recovered by precipitation with 5 volumes of acetone and centrifugation at 12,000g for 10 min. Proteins were separated on a 15% (w/v) SDS-polyacrylamide gel and transferred electrophoretically for 1 h at 100 V to a 0.45-µm nitrocellulose membrane (Hybond-C; Amersham Pharmacia Biotech) without SDS in the transfer buffer. Immunoblotting was performed as described previously (Danyluk et al., 1998) with the anti-WCS19 antibody diluted at 1:10,000 and the secondary antibody diluted at 1:25,000.

Chloroplast Preparation and Fractionation

Chloroplasts were prepared using the method of Kunst et al. (1988) with slight modifications. Wheat leaves (10 g) were ground in 100 mL of extraction buffer (0.45 M sorbitol, 20 mM Tricine KOH, pH 8.4, 2.5 mM EDTA, and 5 mM MgCl₂). The extract was filtered through two layers of Miracloth (Calbiochem, San Diego) and centrifuged at 270g for 90 s, and the pellet was resuspended in buffer A (0.3 M sorbitol, 20 mM Tricine KOH, pH 7.6, 5 mM MgCl₂, and 2.5 mM EDTA). The chloroplast suspension was then layered on a Percoll gradient previously prepared by centrifuging 50% (v/v) Percoll in buffer A at 43,000g for 30 min in a SW41 Ti rotor. The gradients were centrifuged at 13,000g for 6 min. Intact chloroplasts, which formed a band near the bottom of the gradient, were recovered, and an aliquot (whole chloroplasts) was mixed with 2× Laemmli buffer for analysis. The remaining intact chloroplasts were diluted with 1 volume of buffer A and pelleted at 3,000g for 90 s. The chloroplasts were then resuspended in buffer A without sorbitol and centrifuged at 3,000g for 5 min. The pellet (whole thylakoids) was resuspended in 1× Laemmli buffer for analysis. The supernatant constitutes the soluble fraction (stroma) and was precipitated with 5 volumes of acetone, and the resulting pellet was suspended in 1× Laemmli buffer for analysis.

Immunocytochemistry and Electron Microscopy

Discs 1.5 mm in diameter were cut from the fourth fully developed leaf of plants grown at 20/250, 20/800, or 5/250, were fixed on ice using 0.5% (w/v) glutaraldehyde and 1.5% (w/v) paraformaldehyde in 0.2 M cacodylate buffer (pH 6.8) for 1 h, and were post-fixed in 2% (w/v) osmium tetroxide. The samples were then rinsed with water and stained with 3% (w/v) uranyl acetate for 30 min. The leaf discs were dehydrated in a graded ethanol series and embedded in LR White resin.

Silver-gold sections were cut from the polymerized blocks using a Sorvall MT2-B ultramicrotome equipped with a diamond knife and mounted on nickel grids (400 mesh). The specimens were then floated sample side down on droplets of the appropriate solutions for immunolabeling. The sections were etched with saturated sodium periodate for 8 min, washed with water, and treated with 0.1 N HCl for 10 min to further increase the availability of antigenic sites (Craig and Goodchild, 1984). The sections were then blocked using PTBN (0.02 M NaPO₄ pH 7.4, 0.15 M NaCl, 0.1% [w/v] bovine serum albumin, and 0.05% [w/v] Tween 20), and incubated overnight in anti-WCS19 antibodies at a dilution of 1:100 in PTBN. After washing in PTBN, the sections were incubated with 15-nm gold-labeled goat-anti-rabbit secondary antibodies (Cedar Lane Laboratories) diluted 1:50 in PTBN for 30 min. The specimens were then rinsed with water and post-stained with 3% (w/v) uranyl acetate before viewing. Photographic prints were made of representative chloroplasts from each treatment group. These were then scanned into a digital format, and the number of gold particles per square micrometer of chloroplast was calculated using the northern Eclipse Image Analysis software package (v5.0, Emplix Image Inc., Mississagua, ON, Canada).

For the images presented, the 15-nm gold particles were enlarged using silver enhancement (Oliver, 1999). Enhancement was carried out for 5 min at room temperature in the dark. The samples were then rinsed thoroughly with water, post-stained with 3% (w/v) uranyl acetate, washed with water, and then viewed.

To ensure the affinity and specificity of the WCS19 antibodies to the pure WCS19 protein, aliquots of the protein were placed on formvar-coated nickel grids (400 mesh), allowed to dry, blocked with PTBN, and then incubated with the WCS19 antibody, the rabbit preimmune serum (both at 1:100 dilutions), or distilled water for 10 min. The samples were then washed with PTBN, incubated with the gold-labeled goat-anti-rabbit secondary antibody for 10 min, rinsed with water, and viewed.

To further ensure the specificity of our antibodies, sections of the 20/800 grown plants were treated as described above, but the WCS19 antibody was substituted with either preimmune serum (1:100) or distilled water. Thus, ensuring that the rabbit had not been presensitized to chloroplast-localized proteins, and that the secondary antibody was not binding directly to the specimens.