INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Both eukaryotes and prokaryotes respond to high temperatures by synthesizing heat shock proteins (HSPs) when environmental temperatures were elevated 5°C to 10°C above normal growth temperatures (Parsell and Lindquist, 1993). The synthesis of HSPs occurs rapidly, whereas the expression of normally active genes is substantially repressed in responding to the onset of heat shock stress. Accumulation of HSPs is closely related to tolerance of extreme temperatures. HSPs are grouped into families based on sequence homology. Low-molecular mass (LMM) HSPs or small HSPs (sHSPs), ranging in size from 15 to 30 kD, are more abundant in higher plants than in other organisms.

The conserved amino acid sequence within divergent LMM HSPs is generally restricted to the C-terminal region, which is termed "-crystallin domain" due to homologies to the -crystallin of vertebrate eye lens. This domain can be further divided into two specific regions, consensus I and consensus II, separated by a varying length hydrophilic linker sequence (Yeh et al., 1994; Waters et al., 1996). Consensus I is highly conserved among all eukaryotic LMM HSPs discovered so far (Yeh et al., 1994). Some studies on -crystallin and LMM HSPs have shown that this conserved -crystallin domain is important for their distinct oligomerization as well as for chaperone activity (Merck et al., 1993; Andley et al., 1996; Lindner et al., 1998; Young et al., 1999).

The term "molecular chaperone" is usually applied to describe the function of HSPs because they bind to and stabilize the unstable conformers of proteins and facilitate correct fate of substrate proteins in vivo (Parsell and Lindquist, 1993). There is ample evidence showing that LMM HSPs from mammalian and plant sources as well as related -crystallin proteins have chaperone activities (Merck et al., 1993; Lee et al., 1995; Ehrnsperger et al., 1997; Lee et al., 1997; Yeh et al., 1997; Sharma et al., 1998). Results from studies of purified murine (Mus musculus) HSP25 (Ehrnsperger et al., 1997) and pea (Pisum sativum) HSP18.1 (Lee et al., 1997) revealed that LMM HSPs could trap unfolding proteins on their surface in an ATP-independent manner and maintain them in a folding-competent state. Therefore, LMM HSPs act as a reservoir of nonnative proteins in preventing them from irreversible aggregation until ATP-dependant chaperones, such as HSP70 and HSP60 (GroE), restore the refolding of thermally denatured proteins to native physiological conditions (Veinger et al., 1998; Lee and Vierling, 2000).

It is known that a higher degree of multimerization of LMM HSPs is necessary for their full stress protective activities (Arrigo and Landry, 1994). Recent studies showed that most LMM HSPs required intact N-terminal domains to form these multimeric complexes (Leroux et al., 1997; Berengian et al., 1999; Lambert et al., 1999). Plant LMM HSPs generally form multimeric protein complexes ranging in size from 200 to 310 kD (Jinn et al., 1995; Lee et al., 1995; Yeh et al., 1995), as shown in rice (Oryza sativa) Oshsp16.9 and pea HSP18.1 consisting of 15 to 18 and 12 subunits, respectively (Jinn et al., 1995; Lee et al., 1995; Yeh et al., 1995). In contrast, LMM HSP complexes from mammalian sources vary considerably and exhibit a larger size (300-800 kD) than their plant counterparts (Arrigo and Landry, 1994). The crystal structure of HSP16.5 from Methanococcus jannaaschii shows this complex to be a hollow sphere built from 24 subunits (Kim et al., 1998). The other characterized quaternary structure of LMM HSP is the 150-kD trimer of trimers formed by Mycobacterium tuberculosis HSP16.3 (Chang et al., 1996).

In our previous study on a rice class I 16.9-kD LMM HSP (Oshsp16.9), we showed that a specific domain in the consensus II region was responsible but not sufficient for Oshsp16.9 function (Yeh et al., 1997). In the current report, we generated 13 variants of Oshsp16.9, and compared the survival rates of Escherichia coli cells harboring these Oshsp16.9 variants under thermal stress. Our results allow us to identify two domains necessary for chaperone activity of class I LMM HSPs in which one located near the N terminus and the other in the consensus II region of the -crystallin domain.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Construction of Oshsp16.9 Variants for Functional Domain Studies

Our previous studies on Oshsp16.9 suggested that a specific domain located within the consensus II region could play a role in providing thermoprotection (Yeh et al., 1997). To determine which part of the Oshsp16.9 consensus II region was critical for thermoprotection, we constructed seven Oshsp16.9 derivatives truncated at the end of consensus II hydrophobic regions and assessed their capacity for thermoprotection (Fig. 1). The pGST-N76, pGST-N74, pGST-N72, pGST-N61, pGST-N58, and pGST-N42 cells produced GST fusion proteins with Oshsp16.9 N-terminal 76, 74, 72, 61, 58, and 42 amino acid residues, respectively. The pGST-FL59-78 cells produced a fusion protein without the consensus II region (amino acid residues 59 through 78) of Oshsp16.9.

View larger version (22K): [in this window] [in a new window]   Figure 1.   Summary of all Oshsp16.9 variant constructs. Top, Hydropathy profile of Oshsp16.9 as analyzed by the method of Kyte and Doolittle (1982). Bottom, Schematic representation of all Oshsp16.9 derivatives in this study, drawn approximately to scale. Two site-directed constructs, N74E73K and N74E74K, are the same length as N74. The regions of Oshsp16.9 corresponding to consensus regions I and II (Yeh et al., 1994; Waters et al., 1996) are also indicated.

It has been suggested that the N-terminal domain of HSP16-2 (Caenorhabditis elegans) is critical for multimerization and chaperone activity (Leroux et al., 1997). To test the importance of N terminus for the function of Oshsp16.9, we created four N-terminal deletion mutants of Oshsp16.9, pGST-N781-11, pGST-N781-21, pGST-N781-29, and pGST-N781-36 cells (Fig. 1). The growth curves of these 11 Oshsp16.9 variants were similar to those of the pGST-N78 and pGST-C108 cells in the presence or absence of isopropyl -D-thioglactopyranoside (IPTG; data not shown).

Thermotolerance of E. coli Cells Expressing Recombinant Proteins of Oshsp16.9 Variants

We have shown that E. coli cells producing the GST-Oshsp16.9 (full-length) protein or GST-N78 protein were more thermotolerant than E. coli cells producing only the GST protein or GST-C108 protein (Yeh et al., 1997). Therefore, we used pGST-N78 cells as a positive control and pGST-C108 cells as a negative control in the thermoprotection assays reported herein. Figure 2A showed, after 2 h of IPTG induction, that the amount of accumulated GST-FL59-78, GST-N72, GST-N74, and GST-N76 was similar to that of GST-C108 or GST-N78. By incubating cells at 47.5°C for 30 min, both pGST-N74 and pGST-N76 cells had survival rates close to those of the positive control pGST-N78 cells, whereas only about 10% of pGST-N72 or pGST-FL59-78 cells survived (Fig. 2B). After 1 h of incubation, both pGST-N74 and pGST-N76 cells were capable of achieving approximate 30% survival rates, 11-fold higher than those of pGST-N72 cells or 50-fold higher than those of pGST-FL59-78 cells. Cells producing GST-N42, GST-N58, and GST-N61, like the negative control pGST-C108 cells, did not survive during 1 h of heat treatment (data not shown). These results suggested that the amino acid residues 73 and 74 in consensus II of Oshsp16.9 played an important role in providing thermotolerance.

View larger version (44K): [in this window] [in a new window]   Figure 2.   Effects of Oshsp16.9 consensus II region on thermotolerance of E. coli. Transformed cells were cultured as described and subjected to 47.5°C for up to 60 min. A, Coomassie Blue-stained SDS-PAGE (13.75% [w/v] acrylamide) of proteins (30 µg) from transformants after 2 h of IPTG induction. Samples shown in lanes 1 through 6 were from pGST-C108, pGST-N78, pGST-FL59-78, pGST-N72, pGST-N74, and pGST-N76 cells, respectively. Fusion proteins are indicated (). B, Survival (%) of pGST-FL59-78 (), pGST-N72 (), pGST-N74 (), and pGST-N76 cells () were shown. Cell lines pGST-C108 () and pGST-N78 (; Yeh et al., 1997) were used as negative and positive controls, respectively. Values represent means ± SD of at least three independent experiments.

To determine the effects of N-terminal deletions on thermotolerance, survival rates of four different E. coli cell lines (pGST-N781-11, pGST-N781-21, pGST-N781-29, and pGST-N781-36 cells) were measured. The levels of accumulated GST-N781-11, GST-N781-21, GST-N781-29, and GST-N781-36 after 2 h of IPTG induction were similar to those of GST-N72 or GST-N78 (Fig. 3A). As shown in Figure 3B, the survival rates of pGST-N781-11, pGST-N781-21, and pGST-N781-29 cells (more than 10% at 1 h) were significantly higher than those of pGST-N781-36 cells (about 3% at 1 h). Compared with pGST-N72, pGST-N781-11, pGST-N781-21, and pGST-N781-29 cells had 5-fold higher survivability after 1 h of high temperature exposure, whereas pGST-N781-36 cells had only slightly higher survival rate. This implied that deletion of the Oshsp16.9 N-terminal residues 30 through 36 led to suppression of its chaperone activity.

View larger version (42K): [in this window] [in a new window]   Figure 3.   Effects of Oshsp16.9 N-terminal region on thermotolerance of E. coli. Transformed cells were cultured as described and subjected to 47.5°C for up to 60 min. A, Coomassie Blue-stained SDS-PAGE (13.75% [w/v] acrylamide) of proteins (30 µg) from transformants after 2 h of IPTG induction. Samples shown in lanes 1 through 6 were from pGST-N78, pGST-N72, pGST-N781-11, pGST-N781-21, pGST-N781-29, and pGST-N781-36 cells, respectively. Fusion proteins are indicated (). B, Survival (%) of pGST-N781-11 (), pGST-N781-21 (), pGST-N781-29 (), and pGST-N781-36 cells () were shown. Cell lines pGST-N72 () and pGST-N78 (; Yeh et al., 1997) were used as negative and positive controls, respectively. Values represent means ± SD of at least three independent experiments.

GST-N72 Is Stable under Heat Stress But Cannot Prevent Thermal Aggregation of Citrate Synthase (CS)

To provide thermotolerance, the fusion protein has to be stable itself to execute its function. To assess the stability of the recombinant fusion proteins, we examined the change in levels of light scattering of purified recombinant proteins at 47.5°C as described in "Materials and Methods." As shown in Figure 4A, incubation at high temperature only led to slightly rise in the level of light scattering of GST-N78, GST-N72, and GST-N781-29, indicating that they were stable under thermal stress. However, the relative level of light scattering of GST-C108 and GST-N781-36 increased significantly, as would be expected if they were unstable during heat treatment. This may explain why pGST-C108 and pGST-N781-36 cells are not thermotolerant. It is interesting that the GST-N72 fusion protein is stable at 47.5°C (Fig. 4A), but the pGST-N72 cells are not thermotolerant (Figs. 2B and 3B). We speculated that the GST-N72 protein was unable to confer thermotolerance because of the disability to prevent aggregation of cellular proteins. To test this hypothesis, we used the model substrate CS (Lee et al., 1995) to measure the chaperone activity of the recombinant proteins. The results in Figure 4B showed that the light scattering by CS aggregation could be detected at 43°C; however, it was effectively suppressed in the presence of GST-N78 (80% reduction). Neither GST-N72 nor GST-N781-36 were able to act efficiently as a molecular chaperone to prevent CS thermal aggregation (40% and 25% reduction, respectively). We further determined that GST-N78 could compensate for loss of chaperone activity of GST-N72 and GST-N781-36. When GST-N78 was added to the preheated reaction solution of CS with GST-N72, prevention of thermal aggregation of CS was as effective as with GST-N78 alone (Fig. 4B). However, no significant difference was detected when GST-N78 was added to the mixture of GST-N781-36 and CS (Fig. 4B). Under these conditions at 43°C, GST-N72, GST-N78, or GST-N781-36 did not increase in light scattering (data not shown). To understand if the increase of light scattering in CS and GST-N781-36 mixture involved in the protein aggregates formed by unfolded CS alone or by unfolded CS/GST-N781-36 complex, we therefore examined the precipitated proteins by SDS-PAGE. It was found that CS and GST-N781-36 were in the pellet (data not shown). The above results indicated that GST-N72 could not block the binding of GST-N78 to CS, whereas GST-N781-36 was able to compete with GST-N78 for the binding to unfolded CS.

View larger version (21K): [in this window] [in a new window]   Figure 4.   Chaperone activity of GST-Oshsp16.9 fusion proteins. A, Stability of 2 µM purified GST-C108 (), GST-N78 (), GST-N72 (), GST-N781-29 (), and GST-N781-36 () was monitored by measuring apparent light scattering (320 nm) at 47.5°C. B, CS (0.45 µM) in the absence or presence of purified GST-Oshsp16.9 proteins (3 µM) was incubated at 43°C for up to 30 min. Samples were taken for light scattering at 320 nm. Curves shown are CS alone (- - -), CS incubated with GST-N78 (), GST-N72 (), or GST-N781-36 () and CS incubated with GST-N72 () or GST-N781-36 () for 3 min followed by addition of GST-N78.

Mutation of the Glu-73 or Glu-74 Leads to Abolish Chaperone Activity of GST-N74

The data shown in Figures 2B and 4B demonstrated that the 73rd and 74th amino acid residues were important for the chaperone activity of Oshsp16.9. To confirm this aspect, two site-directed mutants, pGST-N74E73K and pGST-N74E74K cells, of which the 73rd and 74th glutamates of Oshsp16.9 were replaced by Lys, were generated. The survival rates of these two mutants and chaperone activity of the two fusion proteins were characterized. The levels of accumulated GST-N74E73K and GST-N74E74K after 2 h of IPTG induction were similar to those of GST-N78 or GST-N72 (Fig. 5A). After heat treatment, the survival rates of pGST-N74E73K and pGST-N74E74K cells dropped to a level comparable with those of pGST-N72 cells (Fig. 5B). As for chaperone activity, neither GST-N74E73K nor GST-N74E74K fusion protein was able to suppress CS thermal aggregation (Fig. 5C). The 73rd or 74th glutamates of Oshsp16.9 was also replaced by Leu, and the survival rates of the mutants were similar to those of pGST-N74E73K or pGST-N74E74K cells at 47.5°C (data not shown). These results suggested that the 73rd and 74th Glu were critical for Oshsp16.9 to interact with its substrate and to confer thermotolerance.

View larger version (25K): [in this window] [in a new window]   Figure 5.   Effects of Glu-73 and Glu-74 on providing chaperone activity of Oshsp16.9. Two-point mutants of pGST-N74, pGST-N74E73K, and pGST-N74E74K cells were cultured as described and subjected to 47.5°C for up to 60 min. A, Coomassie Blue-stained SDS-PAGE (13.75% [w/v] acrylamide) of proteins (30 µg) from E. coli transformants after 2 h of IPTG induction. Samples shown in lanes 1 through 4 were from pGST-N78, pGST-N72, pGST-N74E73K, and pGST-N74E74K cells, respectively. Fusion proteins are indicated (). B, Survival (%) of pGST-N74E73K () and pGST-N74E74K cells () were shown. Cell lines pGST-N72 () and pGST-N74 () as well as pGST-N78 (; Yeh et al., 1997) were used as negative and positive controls, respectively. Values represent means ± SD of at least three independent experiments. C, CS (0.45 µM) in the absence or presence of purified proteins (3 µM) was incubated at 43°C for up to 30 min. Samples were taken for light scattering at 320 nm every 3 min. Curves shown are CS alone (-), incubated with GST-N78 (), GST-N72 (), GST-N74E73K (), or GST-N74E74K ().

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

In contrast to the highly conserved HSP70s (Lindquist and Craig, 1988), LMM HSPs are characterized by a low degree of homology but share a conserved domain known as the -crystallin domain. There is increasing evidence showing that these proteins can function as molecular chaperones in vitro to suppress protein thermal aggregation and as a reservoir of unfolded protein to facilitate refolding in cooperation with other molecular chaperones (Ehrnsperger et al., 1997; Veinger et al., 1998; Lee and Vierling, 2000). In our previous results, expression of a plant LMM HSP, Oshp16.9, allowed E. coli cells to survive at high temperature (Yeh et al., 1997). Basically, the chaperone behavior of plant LMM HSPs is ATP independent, different from the ATP-dependent HSP70, and lacks the specificity for substrate proteins (Jinn et al., 1995; Lee et al., 1995). However, the function of plant LMM HSPs in vivo remains unclear. For studying the function of HSPs in vivo, null mutants or organisms without equivalent HSPs were transformed with a specific HSP gene to assay the function of the transferred gene (Mehlen et al., 1993; Lee et al., 1994; Forreiter et al., 1997). To study the function of plant LMM HSPs, we transformed E. coli cells with plasmids bearing the rice class I LMM HSP gene, Oshsp16.9, and determined the number of cells capable of surviving at lethal temperature (Yeh et al., 1997). In this study, we expanded the scope to test the functions of N terminus as well as the -crystallin domain of Oshsp16.9 in vivo.

Many studies on -crystallin and Drosophila melanogaster LMM HSPs have shown that the C-terminal half of the conserved -crystallin domain is important for the ability of the proteins to prevent other polypeptides from thermal aggregation (Mehlen et al., 1993; Takemoto et al., 1993). It is generally accepted that hydrophobic regions in LMM HSPs are important for substrate binding (Arrigo and Landry, 1994). Recently, the substrate-binding sites of pea HSP18.1 (Lee et al., 1997), -crystallin (Sharma et al., 1998), and E. coli IbpB (Shearstone and Baneyx, 1999) have been identified and are located within the N-terminal half of the conserved -crystallin domain, which is the equivalent region of Oshsp16.9 consensus II domain. Our previous studies on two Oshsp16.9 variants have shown that E. coli cells (pGST-N78 cells) producing N-terminal 78 amino acids of Oshsp16.9 (includes the consensus II region of plant LMM HSPs) can survive at high temperature, whereas E. coli cells (pGST-C108 cells) producing C-terminal 108 amino acids of Oshsp16.9 (includes both consensus I and II regions of plant LMM HSPs) cannot mitigate the detrimental effects of high temperature (Yeh et al., 1997). Similar results were found if the pUC8 expression vector in place of the GST expression system was used to express Oshsp16.9 and its deletion mutant (Young et al., 1999). In the current study, we introduced truncations of Oshsp16.9 into E. coli using pGEX-2T expression vector and compared the survivability of these transformants with those of pGST-N78 (thermotolerant) and pGST-C108 cells (thermosensitive) under thermal stress after IPTG induction. Our major aim is to find which domain(s) is/are necessary for conferring cellular thermotolerance.

In pGST-N76 and pGST-N74 cells, production of the N-terminal 76 or N74 amino acid residues of the Oshsp16.9 protein, respectively, were able to help cells survive at a temperature that was lethal to pGST-C108 cells (Yeh et al., 1997). The survival rates of pGST-N76 and pGST-N74 cells were slightly less than those of pGST-N78 cells although they decreased gradually. However, expression of N-terminal 72 amino acid residues of Oshsp16.9 protein (pGST-N72 cells) or a fusion protein without consensus II region of Oshsp16.9 (pGST-FL59-78 cells) had no effect on surviving of cells under heat treatment. These results clearly indicate that amino acid residues 73 through 78 (EEGNVL) of Oshsp16.9, especially the 73rd and 74th amino acid residues, are essential for providing thermotolerance. This conclusion is further supported by the observation that neither pGST-N74E73K nor pGST-N74E74K cells show thermotolerance. On the other hand, in the N-terminal amino acid residue deletion variants, expression of GST-N781-29 was more effective than GST-N781-36 in conferring cellular thermotolerance. A similar conclusion was obtained from in vitro studies on suppression of thermal aggregation of E. coli-soluble proteins (data not shown) or CS. Thus, the regions of residues 73 through 78 (EEGNVL) and residues 30 through 36 (PATSDND) may be indispensable for Oshsp16.9 providing thermoprotection.

To better understand the roles of the two domains, we compared the characteristics of the Oshsp16.9 mutants at 47.5°C. It is interesting that the GST-N72 protein was as heat stable as GST-N78 and GST-N781-29 proteins and that most of the pGST-N72 cells were killed under the same condition. A different result was observed for GST-N781-36 and GST-C108 proteins, where lacking residues 30 through 36 led to loss of heat stability in correlation with decreased survival rates of their transformants. The data obtained here showed these two regions had different functions in chaperone properties of LMM HSPs. The 1,1'-bi(4-anilino)naphthalene-5,5'-disul-fonic acid-binding sites of LMM HSPs have been proposed as the binding sites where LMM HSPs trap and stabilize partially unfolded proteins (Ehrnsperger et al., 1997; Lee et al., 1997; Lambert et al., 1999). Pair-wise analysis of Oshsp16.9 with pea HSP18.1 have shown that the region of residues 73 through 78 (EE/D- -VL) may be closely related to the 1,1'-bi(4-anilino) naphthalene-5,5'-disulfonic acid binding site of pea HSP18.1. In the proceeding study, we showed that GST-N78 could fully prevent thermal aggregation of CS in the presence of GST-N72, but only decreased by about one-half the thermal aggregation of CS that has been pre-incubated with GST-N781-36. It was found that the longer the pre-incubation of CS with GST-N781-36 at high temperature, the less effect of GST-N78 on prevention of thermal aggregation of CS (data not shown). These data suggested that GST-N78 and GST-N781-36 had the same CS binding site and competed for interacting with unfolded CS. On the other hand, substitution of Glu-73 or Glu-74 impaired the chaperone activity of GST-N74. Thus, we confirmed the idea that the substrate-binding site of Oshsp16.9 was on the region of residues 73 through 74.

Many reports have shown that multimerization of LMM HSPs is perquisite for their chaperone/thermoprotection function (Lee et al., 1995; Rogalla et al., 1999). From native protein analysis, recombinant proteins from thermotolerant transformants (pGST-FL and GST-N78 cells) and thermosensitive transformants (pGST-N72 and pGST-N781-36 cells) could form oligomeric structures (data not shown). These results suggested that the oligomerization of LMM HSPs was necessary but not sufficient for their thermoprotection. Taken together, these observations suggest that two domains are necessary for Oshsp16.9 to confer thermoprotection. The domain located at the N-terminal region between residues 30 and 36 is involved in the stability for LMM HSP oligomer itself. Without this region, the LMM HSP may lose its stability under thermal stress, and thus has no chaperone activity. The other domain in the consensus II region between residues 73 and 74 is responsible for the substrate binding. Although deletion of this region or replacement with neutral or basic amino acids still allows the HSP to form a stable oligomer at high temperature, it cannot provide thermoprotection in vivo or in vitro because of lacking interactions between LMM HSP and its substrates. Both of these domains are necessary for chaperone activities in vivo of Oshsp16.9.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Preparation of Oshsp16.9 Deleted and Site-Directed Mutants

The full-length construct encoding the entire Oshsp16.9 coding region and two deletion constructs (N78 and C108) encoding the N-terminal 78 amino acids and the C-terminal 108 amino acids of Oshsp16.9 were previously subcloned into the BamHI-EcoRI site of pGEX-2T expression vector (Amersham-Pharmacia Biotech, Uppsala) for producing fusion proteins (Yeh et al., 1997). For expressing variant protein mutants, DNA fragments encoding N-terminal 42, 58, 61, 72, 74, 76, 12 through 78, 22 through 78, 30 through 78, and 37 through 78 amino acids of Oshsp16.9 (N42, N58, N61, N72, N74, N76, N781-11, N781-21, N781-29, and N781-36) were prepared by PCR and subcloned into the BamHI-EcoRI site of pGEX-2T. To generate a construct (FL59-78) lacking most consensus II region of Oshsp16.9, we first synthesized two DNA fragments (I and II) encoding the N-terminal 58 amino acids and the C-terminal 72 amino acids by PCR using primer sets of 5'-GC-GGATCCATGTCGCTGGTG-3'/5'GGCCTTGAAGACGTG-CGA-3' and 5'-GTGATCAGCGGGCAGCGC-3'/5'-CC-GAATTCCTTAACCCGGAGATCTC-3', respectively (BamHI and EcoRI restriction sites are underlined). These two DNA fragments were then ligated, gel purified, and subcloned into BamHI and EcoRI sites of pGEX-2T. The GST-N74 site-directed mutants (GST-N74E73K and GST-N74E74K) in which Glu-73 (GAG) or Glu-74 (GAA) was changed to Lys (AAG) separately were also generated by PCR using primer sets of 5'-GCGGATCCATGTCGCTGGTG-3'/5'-ATGAATTCCCTTCTCCACCTCCACCTTCAC-3' and 5'G-CGGATCCATGTCGCTGGTG-3'/5'-ATGAATTCCTTCCT-TCACCTCCACCTTCAC-3', respectively (BamHI and EcoRI restriction sites are underlined). The PCR products were gel purified and subcloned into BamHI and EcoRI sites of pGEX-2T. Transformation was performed according to standard protocol (Sambrook et al., 1989). Escherichia coli (XL1-blue) strains harboring the vector alone and these recombinant plasmids were designated as pGST, pGST-FL, pGST-N42, pGST-N58, pGST-N61, pGST-N72, pGST-N78, pGST-N781-11, pGST-N781-21, pGST-N781-29, pGST-N781-36, pGST-C108, pGST-FL59-78, pGST-N74E73K, or pGST-N74E74K cells. Expressed in E. coli, the plasmids above could produce fusion proteins of N-terminal GST and variant C-terminal protein fragments of Oshsp16.9. The expression and purification of these recombinant proteins were described previously (Yeh et al., 1995).

Analysis for Thermotolerance of Transformants

For thermotolerance assays, Oshsp16.9 variants were cultured as described previously (Yeh et al., 1997). IPTG (1 mM) was added to induce the production of fusion proteins. The cultures were then diluted to 5.0 × 10⁶ cells mL¹. One milliliter of the culture was subjected to heat shock treatment at 47.5°C. Samples (100 µL) were taken at the time indicated and serial dilutions of the cells were plated in triplicate onto Luria-Bertani broth plates containing ampicillin (100 µg mL¹). To determine the percentage of survivors, the plates were incubated overnight at 37°C prior to counting colony formation.

Thermal Aggregation Analysis

Thermal aggregation suppression analysis was performed basically according to the method of Lee et al. (1995) using porcine heart CS (Sigma, St. Louis) as substrate. CS (450 nM; monomer concentration) was incubated in the absence or presence of 3 µM purified GST-N78, GST-N72, and GST-N781-36 (monomer concentration) in 50 mM potassium phosphate, pH 7.5 (800 µL total), preheated to 43°C. Samples were monitored for light scattering at 320 nm in a U3200 spectrophotometer (Hitachi, Tokyo) with a thermostatted cell compartment. The proteins were quantified with the Bradford assay (Bradford, 1976).

Electrophoresis, Immunoblotting Analysis

SDS-PAGE was performed according to the method of Laemmli (1970) using 13.75% (w/v) polyacrylamide gels. Non-denaturing PAGE was performed on a 5% to 20% (w/v) linear polyacrylamide gel as described (Anderson et al., 1972). The conditions for western blots were performed with anti-Oshsp16.9 antibodies as previously described (Yeh et al., 1995).