Plant Physiol. (1998) 116: 1097-1110

Heat-Stress Response of Maize Mitochondria¹

Adrian A. Lund², Paul H. Blum, Dinakar Bhattramakki³, and Thomas E. Elthon^*

School of Biological Sciences and the Center for Biotechnology, University of Nebraska, Lincoln, Nebraska 68588-0118

	ABSTRACT

Top Abstract Introduction Methods Results Discussion References

We have identified maize (Zea mays L. inbred B73) mitochondrial homologs of the Escherichia coli molecular chaperones DnaK (HSP70) and GroEL (cpn60) using two-dimensional sodium dodecyl sulfate-polyacrylamide gel electrophoresis and immunoblots. During heat stress (42°C for 4 h), levels of HSP70 and cpn60 proteins did not change significantly. In contrast, levels of two 22-kD proteins increased dramatically (HSP22). Monoclonal antibodies were developed to maize HSP70, cpn60, and HSP22. The monoclonal antibodies were characterized with regard to their cross-reactivity to chloroplastic, cytosolic, and mitochondrial fractions, and to different plant species. Expression of mitochondrial HSP22 was evaluated with regard to induction temperature, time required for induction, and time required for degradation upon relief of stress. Maximal HSP22 expression occurred in etiolated seedling mitochondria after 5 h of a +13°C heat stress. Upon relief of heat stress, the HSP22 proteins disappeared with a half-life of about 4 h and were undetectable after 21 h of recovery. Under continuous heat-stress conditions, the level of HSP22 remained high. A cDNA for maize mitochondrial HSP22 was cloned and extended to full length with sequences from an expressed sequence tag database. Sequence analysis indicated that HSP22 is a member of the plant small heat-shock protein superfamily.

	INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

The effect of environmental stress on agronomic plants has been a major focus of research. Plant productivity is related to the ability of plants to respond to and adapt to environmental stress (Sachs and Ho, 1986). The proteins produced by higher plants in response to stress have been well characterized (Key et al., 1981; Cooper and Ho, 1983; Sachs and Ho, 1986). Many stress proteins have recently been found to be chaperones, a class of proteins involved in the folding of newly synthesized proteins (Ellis and van der Vies, 1991; Gething and Sambrook, 1992; Craig et al., 1993). The chaperones have been proposed to function during stress in the binding of partially denatured proteins, thus preventing their degradation, and in the refolding of these proteins into their native structure in an ATP-dependent manner after relief of stress (Rochester et al., 1986; Ellis and Hemmingsen, 1989; Hendrick and Hartl, 1993; Schröder et al., 1993). The two most extensively studied classes of chaperones are HSP70 homologs and cpn60 homologs.

HSP70 homologs have been found in higher-plant cytoplasm (Giorini and Galili, 1991), the ER (Denecke et al., 1991), chloroplasts (Marshall et al., 1990; Ko et al., 1992; Marshall and Keegstra, 1992; Madueño et al., 1993; Wang et al., 1993), and mitochondria (Watts et al., 1992; Neumann et al., 1993). The chloroplast and mitochondrial forms of HSP70 in plants are similar to those of cyanobacteria and purple bacteria, respectively (Boorstein et al., 1994). Mitochondrial HSP70 has been shown to undergo calcium-stimulated autophosphorylation, suggesting possible regulation in vivo (Miernyk et al., 1992a, 1992b). Genes for mitochondrial HSP70 are nuclearly encoded and have been isolated from pea (Pisum sativum L.) (Watts et al., 1992) and from potato (Solanum tuberosum) and tomato (Lycopersicon esculentum) (Neumann et al., 1993). A multigene family for the higher-plant ER HSP70 homolog has been isolated from tobacco (Nicotiana tabacum) (Denecke et al., 1991). Genes for chloroplast forms of HSP70 have been isolated from pea (Ko et al., 1992; Marshall and Keegstra, 1992) and spinach (Spinacia oleracea) (Wang et al., 1993). Specific genes for constitutively and inducibly expressed cytosolic forms of HSP70 have been identified in higher plants (Bates et al., 1994). Cytosolic forms of HSP70 lack organelle-targeting peptides or ER retention signals and are less similar to the prokaryotic HSP70 proteins. A multigene family for cytosolic HSP70 was found in Arabidopsis thaliana (Wu et al., 1988), and genomic clones for two inducible HSP70s were found in maize (Zea mays L.) (Rochester et al., 1986), but the subcellular destination of the gene products was not firmly established.

The cpn60s are a group of ubiquitous proteins with a subunit size of approximately 60 kD that share a functional and structural similarity to the tetradecameric Escherichia coli GroEL complex (Gatenby, 1992). Eukaryotic representatives of this group include the chloroplast Rubisco subunit-binding protein (Hemmingsen and Ellis, 1986; Hemmingsen et al., 1988; Martel et al., 1990; Madueño et al., 1993) and the mitochondrial cpn60 protein (Prasad and Hallberg, 1989; Tsugeki et al., 1992). The maize and A. thaliana mitochondrial cpn60 genes have been isolated and found to be encoded in the nucleus (Prasad and Stewart, 1992). The maize cpn60 protein was hypothesized to aid in the assembly of new mitochondrial protein complexes during the rapid organelle biogenesis of seedling germination and heterotrophic growth (Prasad and Stewart, 1992). Mitochondrial cpn60 cDNAs have also been isolated from pumpkin (Cucurbita pepo) cotyledons (Tsugeki et al., 1992). Genes for the chloroplast GroEL homologs have been isolated from Brassica napus and A. thaliana (Martel et al., 1990). There is no evidence to date for any cytoplasmic cpn60 homologs in eukaryotes. However, several cytosolic chaperones, which do not appear to be related to the chaperonins, have been observed, including the mammalian TCP-1 (Gupta, 1990), TF55 from Sulfolobus shibitae (Trent et al., 1992), and a molecular chaperone from rabbit reticulocyte lysate (Gao et al., 1992).

Another group of HSPs, which is more diverse and abundant in plants than other organisms, are the low-molecular-mass (17-30 kD) HSPs. Some of the low-molecular-mass HSPs contain a C-terminal protein domain similar to a domain found in the mammalian eye lens -crystallin proteins and are called the sHSPs (Waters et al., 1996). Recent reports have established that the cytosolic forms of plant sHSPs can function as molecular chaperones in vitro (Lee et al., 1995). Lenne and Douce (1994) identified a mitochondrial, matrix-localized, low-molecular-mass HSP, HSP22. Pea leaf mitochondrial HSP22 is conditionally expressed only at high temperatures and the protein level remained high for at least 3 d after heat stress (Lenne and Douce, 1994). A cDNA for pea mitochondrial HSP22 has been identified and establishes this protein as a member of the sHSP superfamily (Lenne et al., 1995). cDNAs for mitochondrial sHSPs have also been characterized in soybean (LaFayette et al., 1996), A. thaliana (Willett et al., 1996), and Chenopodium rubrum (Lenne et al., 1995; Waters et al., 1996).

It has been suggested for E. coli that the chaperones DnaK and GroEL, along with GroES, DnaJ, and GrpE, act in concerted succession to fold nascent polypeptides (Langer et al., 1992) and to repair heat-induced protein damage (Schröder et al., 1993). Lacking in vitro chaperone activity themselves, the GroES, DnaJ, and GrpE proteins have been termed co-chaperones because of their significant impact on the process as a whole. Although there are abundant findings for DnaK and GroEL homologs in plants, the identification of the plant co-chaperones has been more elusive. A cDNA was recently isolated from Atriplex nummularia for a higher-plant DnaJ homolog that could complement a DnaJ mutant in yeast (Zhu et al., 1993).

In this paper we evaluate the heat-stress response of maize mitochondria in planta. We have found that levels of HSP70 and cpn60 do not change to any extent during the stress treatments. In contrast, levels of mitochondrial HSP22 increase dramatically during stress and decrease after the stress is relieved. We have identified HSP22 as a member of the mitochondrial sHSP superfamily, the first identification, to our knowledge, of a mitochondrion-localized sHSP from a heat-tolerant plant. These results suggest that mitochondrial HSP22 may be the protein that effectively protects mitochondria during heat stress.

	MATERIALS AND METHODS

Top Abstract Introduction Methods Results Discussion References

Isolation of Mitochondria, Cytoplasmic Proteins, and Chloroplasts

Maize chloroplasts were isolated using a combination of procedures from Leegood and Walker (1979) and Mourioux and Douce (1981). Maize seeds were allowed to imbibe, planted in vermiculite, and watered as needed for 2 weeks in a growth room with a 12-h photoperiod under fluorescent lighting (130-160 µmol m² s¹). The ambient temperature was 21/13°C (light/dark). Fifty grams of leaves was cut transversely with a razor blade into 1-cm segments, placed into 150 mL of semifrozen-grinding medium (330 mm mannitol, 10 mm EDTA, 5 mm MgCl₂, 0.2% [w/v] sodium d-isoascorbate, and 30 mm Mops, pH 7.6), and homogenized with three 2-s bursts at full speed in a Waring Blendor. The brei was squeezed through two layers of cheesecloth and allowed to drip through eight layers of muslin wetted with grinding medium. The homogenate was centrifuged in an SS-34 rotor (Sorvall) at 6,000g for 90 s. Crude chloroplast pellets were resuspended in 5 mL of 1× Percoll-gradient buffer (330 mm mannitol, 2 mm EDTA, and 50 mm Mops, pH 7.8) and layered on top of two 25-mL 50% Percoll (12.5 mL of 2× Percoll-gradient buffer and 12.5 mL of Percoll each) gradients that had been precentrifuged for 2 h at 10,000g in an SS-34 rotor. The crude chloroplasts were then centrifuged on the gradients for 10 min at 5,000g and the intact chloroplasts were collected from the rapidly sedimenting diffuse green band. The purified chloroplasts were diluted by adding 2 volumes of 1× Percoll-gradient buffer and pelleted in a microfuge at 3,500g for 90 s. The supernatant was aspirated and the chloroplasts were resuspended in a minimal volume of 30 mm Mops, pH 8.0, and stored at 80°C for subsequent gel analysis.

One-Dimensional and 2D Gel Electrophoresis

Polyclonal Antibodies, MAbs, and Immunoblotting

For the production of the HSP70, cpn60, and -ATPase subunit MAbs, female BALB/C mice were immunized with whole maize mitochondrial proteins. Mice producing HSP22 polyclonal antisera were used for the HSP22 monoclonal line development. Hybridomas were produced according to Elthon et al. (1989) except that growth medium contained 20% (v/v) fetal calf serum, 2 mm l-Gln, 25 mg/L ampicillin, 100 mg/L streptomycin sulfate, and 0.1% (w/v) amphotericin B in a base culture medium of 1× Dulbecco's modified Eagle's medium (Sigma). Hybridomas secreting useful antibodies were selected using immunoblots of mitochondrial proteins. Culture supernatant containing the MAbs was stored at 80°C and used at a 1:10 dilution. For immunoblots, protein gels were transferred to nitrocellulose and probed with antibodies according to Hayes et al. (1991). Goat anti-mouse IgG and anti-rabbit IgG antibodies conjugated with alkaline phosphatase were purchased from Sigma. Proteins transferred to nitrocellulose were reversibly visualized by staining with 0.2% (w/v) Ponceau S in 3% (w/v) TCA for 2 min followed by rinsing with distilled H₂O. Blots were fully destained before antibody probing by washing with PBS containing 0.3% (v/v) Tween 20.

Mitochondrial HSP22 Protein N-Terminal Sequencing

View larger version (101K): [in this window] [in a new window]   Figure 3. 2D Coomassie blue-stained gel of heat-shocked seedling mitochondrial proteins. Three-day-old etiolated maize seedlings grown at 29°C were heat shocked for 4 h at 42°C and the mitochondria were isolated. Three hundred micrograms of mitochondrial protein was run on a 2D gel and stained with Coomassie blue. The two spots that are indicated by arrows are HSP22A (acidic) and HSP22B (basic). HSP22 protein approximate molecular mass is indicated to the left (in kilodaltons).

Cloning and Sequencing of HSP22 cDNA

Protein and Nucleotide Sequence Analysis and Comparison

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Identification and MAb Production to Mitochondrial HSP70 and cpn60

View larger version (65K): [in this window] [in a new window]   Figure 1. Identification of E. coli DnaK (HSP70) and cpn60 protein homologs in maize mitochondria using 2D SDS-PAGE and immunoblots. 2D gels were prepared with approximately 300 µg of maize mitochondrial protein and were either stained with Coomassie brilliant blue R-250 or blotted onto nitrocellulose. The top panel is a 2D immunoblot probed with polyclonal antisera against E. coli DnaK. The center panel is a similar 2D gel stained with Coomassie blue. The bottom panel is a 2D immunoblot probed with polyclonal sera against maize cpn60. The positions of mitochondrial HSP70 and cpn60 are indicated by brackets in the center panel. Approximate molecular mass markers are on the left (in kilodaltons).

A set of MAbs produced against maize mitochondrial fractions was screened for antibodies that bound to the identified HSP70 and cpn60 proteins. Three MAbs were found that bound to HSP70 (A-C) and three were found for cpn60 (A-C). A 2D immunoblot of MAb HSP70B (Fig. 2, top) revealed the major HSP70 spots, as well as some lower-molecular-mass minor spots, that are presumably degradation products. HSP70C yielded a similar pattern on a 2D immunoblot but HSP70A did not bind to the lower-molecular-mass spots (data not shown). The cpn60A MAb (Fig. 2, bottom) and MAbs cpn60B and cpn60C (data not shown) are all very specific for the major 64.5-kD proteins identified using the polyclonal antisera.

View larger version (82K): [in this window] [in a new window]   Figure 2. Identification of maize mitochondrial HSP70 and cpn60 MAbs with 2D immunoblots. 2D immunoblots were prepared with approximately 300 µg of maize mitochondrial proteins and probed with MAbs. The top panel was probed with MAb HSP70B and the bottom panel with MAb cpn60A. Approximate molecular masses are indicated to the left (in kilodaltons).

To prove that the identified proteins were cpn60, we partially purified cpn60 using the same procedure that Prasad and Hallberg (1989) used to initially isolate cpn60 from maize mitochondria. Their procedure purifies cpn60 on a Suc gradient based on the high molecular mass of the multimeric cpn60 complex. Mitochondria were subfractionated as described in ``Materials and Methods'' to yield a fraction consisting of high-molecular-mass soluble protein complexes. When these complexes were separated on a Suc gradient, we observed that the cpn60 complex sedimented to approximately 23% Suc (data not shown), which is similar to values published by Prasad and Hallberg (1989). Immunoblots of similar gels proved that the MAbs were to cpn60. In our preparations we have also observed that proteins with molecular masses of 73 and 63 kD co-sediment with cpn60, with the 63-kD protein band being of very low abundance. The HSP70A MAb was also used to probe an immunoblot of the Suc gradient fractions, and the results showed that HSP70 moved into the gradient as a more slowly sedimenting band located at 16.8% Suc, which was similar to the sedimentation of the F₁ ATPase protein complex (data not shown). This may indicate that HSP70 can be part of a higher-molecular-mass complex present in the mitochondria.

The Heat-Shock Response of Maize Mitochondria

Because increased expression of the HSP22s was the major protein response of plant mitochondria to heat stress, we injected mice independently with the two HSP22 proteins. Regardless of which protein was injected, polyclonal serum that bound to both proteins was obtained. All but one of the independent antibodies obtained recognized cpn60 in addition to HSP22 when used to probe 2D immunoblots of mitochondrial proteins from heat-shocked seedlings (Fig. 4, top). To determine if there was antigenic cross-specificity between HSP22 and cpn60 we affinity purified the HSP22 polyclonal antibodies using 2D gel-purified HSP22 protein. The affinity-purified HSP22 polyclonal antibodies were found to have no specificity for cpn60 (Fig. 4, center). A MAb was obtained that was specific for HSP22 (Fig. 4, bottom). The affinity-purified polyclonal antibodies and MAbs were both observed to bind to two basic proteins of low abundance with molecular masses of approximately 30 kD (Fig. 4, center and bottom).

View larger version (36K): [in this window] [in a new window]   Figure 4. Mitochondrial HSP22 polyclonal antibodies, affinity-purified polyclonal antibodies, and MAb. The top panel is a 2D immunoblot of heat-shocked maize mitochondrial proteins probed with polyclonal antiserum to HSP22. A similar unprobed blot was stained with Ponceau S to reveal the proteins, the spots corresponding to HSP22 were cut out, destained, and incubated with the HSP22 polyclonal antisera, and the antibodies were eluted with a low-pH wash. The center panel is a 2D immunoblot probed with the affinity-purified polyclonal antibodies. The lower panel is a similar 2D immunoblot probed with the MAb generated to the HSP22 proteins.

The six MAbs (to HSP70 and cpn60) were isotyped and the cross-reactivity of the HSP70 and cpn60 MAbs to rat mitochondria and mitochondria of several plant species was evaluated using one-dimensional immunoblots of proteins isolated from unstressed tissues. The HSP70 and cpn60 MAbs did not cross-react to rat mitochondrial proteins, but cross-reactivity to other plant species was common (Table I). The isotype of the HSP22 MAb was found to be IgG₁. Because the HSP22 protein is not expressed constitutively, we have only begun to characterize the cross-reactivity of the maize HSP22 MAb. We have determined that it cross-reacts to heat-inducible, low-molecular-mass proteins from mature leaf whole-protein extracts of Arabidopsis thaliana ecotype Columbia (grown at 21°C and heat stressed at 42°C for 4 h).

View this table: [in this window] [in a new window]   Table I. MAb isotypes and cross-reactivities to other species Immunoblots were prepared using mitochondria isolated from several species (20 µg each) and probed with the three HSP70 and three cpn60 MAbs. The species and tissues that were used for mitochondrial isolation were Rattus rattus (Rat) liver, Z. mays L. inbred B73 (Corn) etiolated shoots, Arum italicum Mill (Arum) spadix, Solanum tuberosum L. cv Russet (Pot) tuber, Phaseolus vulgaris cv Sprite (Bean) shoot, Sauromatum guttatum Schott (Saur) spadix, Beta vulgaris L. (Beet) root, Triticum aestivum cv Arapaho (Wheat) etiolated shoot, and Brassica oleracea L. (Caul) inflorescence. The relative degree of antibody binding was evaluated visually on the immunoblots and designated as follows: +, high-level binding; w, weak binding; and , no binding.

Cross-Reactivity of the HSP70, cpn60, and HSP22 MAbs to Chloroplast and Cytoplasmic Subcellular Fractions

View larger version (94K): [in this window] [in a new window]   Figure 5. Distinct subcellular fractions from maize seedlings. Immunoblots were prepared from the chloroplastic, cytosolic, and mitochondrial fractions (20 µg/lane) and probed with antibodies to known marker enzymes. The chloroplast (Chlpt) proteins were isolated from 2-week-old maize seedlings as described in ``Materials and Methods''. The cytoplasmic (Cyto) and mitochondrial (Mito) fractions were isolated from 3-d-old etiolated shoots grown at 29°C. The chloroplast marker is the 45-kD NADP-malate dehydrogenase protein (NADP-MDH), left; the 50- and 52-kD enolase proteins represent the cytoplasm marker, center; and the 35- and 36-kD proteins of the alternative oxidase (Alt. Ox.) represent the mitochondrial marker, right.

The three subcellular fractions were then probed with the HSP70 and cpn60 MAbs (Fig. 6). MAb HSP70A recognized proteins only in the chloroplast and mitochondrial fractions, indicating that it is specific for organellar forms of HSP70 (top left). MAbs HSP70B and HSP70C recognized proteins in all three subcellular fractions, including high-molecular-mass bands that are present in the mitochondrial fraction (top center and right). This may indicate the presence of a high-molecular-mass aggregate containing HSP70 that does not enter the resolving gel. The cpn60 MAbs did not recognize any proteins in the cytoplasmic fraction (bottom). In the chloroplast fraction, MAb cpn60A reacted the strongest, cpn60C weakly, and cpn60B very weakly (bottom). MAb cpn60A thus reacts with organellar cpn60, whereas MAb cpn60B reacts almost exclusively with the mitochondrial form.

View larger version (83K): [in this window] [in a new window]   Figure 6. HSP70 and cpn60 MAb cross-reactivity to different subcellular fractions of maize. Immunoblots similar to those in Figure 5 were prepared and probed with the MAbs to HSP70 and cpn60. The top panels show blots probed with MAbs HSP70A (A), HSP70B (B), and HSP70C (C), which identify the 70-kD species. The bottom panels show blots probed with cpn60A (A), cpn60B (B), and cpn60C (C), which identify a 64.5-kD species. Labels are as in Figure 5.

Because the sHSPs are also widely distributed in the various compartments of the plant cell, similar cellular fractionation experiments were performed and analyzed with the HSP22 MAb. Control and heat-shocked plants and seedlings were fractionated into chloroplast, cytoplasm, and mitochondrial fractions. Proteins from each fraction were separated by SDS-PAGE and either stained with Coomassie blue (Fig. 7, top) or immunoblotted with the HSP22 MAb (Fig. 7, bottom). These results indicate that the HSP22 MAb is specific to proteins in the mitochondrial fraction.

View larger version (46K): [in this window] [in a new window]   Figure 7. Subcellular cross-reactivity of the HSP22 MAb. The top panel is a Coomassie blue-stained SDS-PAGE gel of subcellular fractions (20 µg/lane) isolated from heat-shocked and control maize tissue. The chloroplast (Chlpt) proteins were isolated from 2-week-old maize seedlings treated for 4 h at 42°C in the dark (HS) or left at the 21°C growth temperature (Con). The cytoplasmic (Cyto) and mitochondrial (Mito) fractions were isolated from 3-d-old etiolated shoots grown at 29°C (Con) and heat shocked for 4 h at 42°C (HS). The bottom panel is an immunoblot of a similar gel probed with the MAb to HSP22. Approximate molecular mass markers are indicated to the left (in kilodaltons).

Submitochondrial Distribution of HSP70, cpn60, and HSP22

Temperature Required for Induction of HSP22 Expression

View larger version (47K): [in this window] [in a new window]   Figure 8. The effect of temperature on the induction of mitochondrial cpn60, HSP70, and HSP22 proteins. Three-day-old etiolated maize seedlings were treated for 4 h at 29 (control), 33, 37, or 42°C. The mitochondria were isolated and analyzed by SDS-PAGE and immunoblots. The top panel is a Coomassie blue-stained SDS-PAGE gel loaded with 20 µg of mitochondrial protein per lane. Approximate molecular mass markers are on the left (in kilodaltons). The four other panels are immunoblots of similar gels probed with the cpn60B MAb (center left), the HSP70B MAb (center right), the -ATPase D MAb as a control (bottom left), and the HSP22 MAb (bottom right).

Time Course of Induction and Decay of Mitochondrial HSP22

View larger version (67K): [in this window] [in a new window]   Figure 9. SDS-PAGE and immunoblot analysis of the time course of induction for maize mitochondrial HSP70, cpn60, and HSP22 proteins. Three-day-old etiolated maize seedlings grown at 29°C were placed at 42°C (0 h at 42°), samples were removed after 0.5, 1, 2, 3, or 4 h of heat shock, and the mitochondria were immediately isolated. Two trays of seedlings that received 4 h of heat shock were returned to the 29°C incubator and allowed to recover for 24 or 50 h before mitochondrial isolation. The top panel is a Coomassie blue-stained SDS-PAGE gel loaded with 20 µg of mitochondrial protein per lane. Approximate molecular mass markers are on the left (in kilodaltons). The three bottom panels are immunoblots of similar gels probed with the MAbs HSP70A, cpn60B, and HSP22.

View larger version (94K): [in this window] [in a new window]   Figure 10. SDS-PAGE and immunoblot analysis of the time course of HSP22 decay from heat-shocked maize mitochondria. Three-day-old etiolated maize seedlings were grown at 29°C and heat shocked at 42°C for 4 h and then returned to 29°C to recover for 3, 6, 9, 12, 15, 18, or 21 h. Mitochondria were isolated from samples taken just before and after heat shock (0 and 4 h at 42°C) and immediately after the recovery times. The top panel is a Coomassie blue-stained SDS-PAGE gel loaded with 20 µg of the mitochondrial isolations per lane. Approximate molecular mass markers are on the left (in kilodaltons). The bottom panel is an immunoblot of a similar gel probed with the HSP22 MAb.

Because no change in HSP70 or cpn60 levels could be observed after 4 h of heat stress (Fig. 9), it was unclear if a long-term heat stress would be sufficient to induce a change in HSP70 or cpn60 levels. It was also unclear if 4 h of exposure produced the maximum expression of HSP22 and if the HSP22 level would be maintained if the plants were not allowed to recover. To address these questions we then evaluated HSP70, cpn60, and HSP22 levels at different time intervals under a continuous heat shock. Other than HSP22, we observed no significant change in the mitochondrial protein profile (Fig. 11, top) or the level of HSP70 and cpn60 (Fig. 11, center), even after 44 h of heat stress. Two hours of heat shock yielded significant induction of HSP22, with maximum expression occurring between 4 and 6 h (Fig. 11, bottom). After maximal induction, the levels remained high until the experiment was terminated after 44 h of heat shock.

View larger version (66K): [in this window] [in a new window]   Figure 11. SDS-PAGE and immunoblot analysis of the effect of continuous heat shock on the levels of maize mitochondrial HSP70, cpn60, and HSP22 proteins. Three-day-old etiolated maize seedlings grown at 29°C were placed at 42°C, samples were removed after 0, 2, 4, 6, 8, 12, 16, 24, or 44 h of heat shock, and the mitochondria were immediately isolated. The top panel is a Coomassie blue-stained SDS-PAGE gel loaded with 20 µg of mitochondrial protein per lane. Approximate molecular mass markers are on the left (in kilodaltons). The three bottom panels are immunoblots of similar gels probed with the MAbs HSP70B, cpn60B, and HSP22.

HSP22 N-Terminal Amino Acid Sequencing and cDNA Characterization

View larger version (59K): [in this window] [in a new window]   Figure 12. Complete cDNA nucleotide sequence and protein translation for maize mitochondrial HSP22. A partial cDNA clone for HSP22, ZmHSP22p8, was isolated by screening a  phage cDNA expression library using the HSP22 MAb. The library was constructed using mRNA from heat-stressed etiolated maize seedlings. The nucleotide sequence was completed using homologous overlapping sequences identified in the Pioneer Hi-Bred (Johnston, IA) Expressed Sequence Tagged database. The mitochondrial transit peptide sequence and the 5 untranslated region of the cDNA were added from sequences CHSSH24R and CTSCG49R from the Heat Shock Recovery Seedling (8 h) and Tassel Shoot Expressed Sequence Tagged libraries, respectively (underlined nucleotide sequence). The putative translational start is at position 79. The complete mitochondrial transit peptide is encoded from positions 79 to 213. The mature HSP22 protein sequence is from 214 to 735 (end). The N terminus of the mature HSP22 protein was confirmed by Edman degradation of the HSP22 polypeptides from 2D SDS-PAGE gels of total mitochondrial proteins that were transferred to PVDF. The amino acid sequence from spot HSP22B is shown in boldface. The amino acid sequence from spot HSP22A is identical to the first 13 residues of the sequence from spot HSP22B (boldface and underlined). The identity of the 18th residue (Ser-63) from spot HSP22B could not be determined during the Edman degradation. The 3 untranslated region is from 736 to 1028 plus a 15-nucleotide polyadenylated tail.

To identify the gene encoding the mitochondrial HSP22 protein we prepared a cDNA library from heat-shocked etiolated maize seedlings using the UniZAP XR phage  expression vector. This expression library was screened with the HSP22 MAb and one positive plaque was obtained, which was selected after two additional rounds of consecutive screening. After in vivo excision of the selected phage, the plasmid (ZmHsp22P8) was sequenced from both directions and this sequence was found to contain the entire mature HSP22 protein coding sequence (bases 214-735, Fig. 12), the 3 untranslated region (bases 736-1028, Fig. 12), and a 15-bp polyadenylated tail (not shown). The N-terminal sequence identified for spots HSP22A and HSP22B matched this sequence exactly and identified HSP22 spot B, residue 18, as Ser (Fig. 12). In an effort to obtain the full cDNA sequence, including the transit peptide, we used the zmhsp22P8 clone as a probe to screen the library again. Twenty-four additional clones of various lengths were obtained after screening 3.2 × 10⁵ plaques. All clones were sequenced and found to contain sequence identical to that of the P8 clone but none contained the full transit peptide (data not shown). Comparison of the incomplete HSP22 cDNA sequence to the Pioneer Hi-Bred Maize Expressed Sequence Tagged allowed identification of three clones that contained sequences identical to the N-terminal region of the mature HSP22 protein. Two of these clones (CTSCG49R and CHSSH24R) extended the sequence to include a putative N-terminal transit peptide and 78 nucleotides of the 5 untranslated region (underlined nucleotides, Fig. 12). The predicted protein sequence for the entire coding portion of the putative HSP22 precursor cDNA and the cDNA nucleotide sequence are shown in Figure 12. The predicted mass of the entire 218-amino acid sequence is 23,816 D.

Comparison of Maize Mitochondrial HSP22 to Other sHSPs

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

In this paper we report the characterization of the proteins involved in the response of plant mitochondria to heat stress. The mitochondrial homologs of HSP70 and cpn60 were identified on 2D immunoblots using polyclonal antibodies to E. coli DnaK and maize mitochondrial cpn60, respectively. On 2D Coomassie blue-stained gels, protein levels of HSP70 and cpn60 did not change significantly under the heat-stress conditions evaluated. In contrast, HSP22 protein levels increased dramatically during stress, and decreased upon relief of stress. A cDNA for HSP22 was identified and found to be similar to that of mitochondrial members of the plant sHSP superfamily. These results suggested that expression of HSP22 may be the effective response to heat stress in plant mitochondria.

Three MAbs were developed to HSP70, three to cpn60, and one to HSP22. The MAbs developed for HSP70 and cpn60 exhibited varied cross-reactivities, which would suggest that each of the three MAbs for HSP70 and each of the three for cpn60 bind to different epitopes (Table I). The cross-reactivity of the MAbs to other species shows that the MAbs will be valuable for investigating chaperone function and expression in a number of systems.

Because of their varied affinity for different subcellular forms of HSP70, the HSP70 MAbs will be useful for investigating defined sets of the large family of HSP70s present. It is interesting that the three MAbs for HSP70 that we have identified have been shown by Mooney and Harmey (1996) to be unable to recognize HSP70 homologs in the intermembrane space of cauliflower mitochondria. Mooney and Harmey (1996) stated that this evidence would suggest that the mitochondrial intermembrane HSP70 is more similar to the cytosolic HSP70s than to the matrix HSP70s. Analysis of the data presented in Table I shows that there is high epitope variability between the plant mitochondrial HSP70 proteins in different species, which suggests that the inability of the MAbs to detect intermembrane HSP70 in cauliflower may or may not apply to other species.

HSP70B and HSP70C MAbs can detect lower-molecular-mass species in whole maize mitochondria (Fig. 6, top center and right). However, HSP70A did not bind to the lower-molecular-mass proteins (Fig. 6, top right), indicating that it binds to an epitope removed early in the degradation process. Oster et al. (1995) found that a C-terminal 35-kD fragment of HSP70 could be detected in whole-tissue extracts of Arabidopsis seeds, fruits, and flowers but not in leaves. It is likely that the lower-molecular-mass species that we are detecting are similar fragments present in the mitochondrial fraction. The amount and pattern of the lower-molecular-mass species detected by the HSP70B antibody does not seem to change during prolonged heat stress of the maize seedling mitochondria (Fig. 11). This may indicate that mitochondrial HSP70 is not highly proteolyzed during heat stress.

There are conflicting reports on the effect of heat stress on the amount of HSP70 that is present in plant mitochondria. Watts et al. (1992), using a polyclonal antibody raised to a mtHSP70 peptide, found that the protein was not induced in mitochondria isolated from pea leaves after a 30-min, 15°C up-shift. Neumann et al. (1993) found a 2- to 3-fold increase in mitochondrial HSP70 (HSP68) using immunoblots and immunogold labeling from whole tomato leaf samples treated with two 15-min, 15°C up-shifts 2 h apart. Our maize data support and extend the findings of Watts et al. (1992) in that the level of mitochondrial HSP70 is not significantly increased in maize after short- or long-term exposure to heat stress as judged by Coomassie blue-stained 2D gels (Fig. 1, center, versus Fig. 3) and immunoblot analysis (Fig. 11). Because it is known that maize mitochondria do not synthesize any HSPs (Nieto-Sotelo and Ho, 1987) we believe that all of the HSPs present in the mitochondria are translated in the cytosol and imported into the mitochondria during heat stress. Experiments to determine protein synthesis during heat stress have shown de novo synthesis of 70-kD mitochondrial HSPs as a result of heat stress (Cooper and Ho, 1987), but because of the high constitutive level of HSP70, we feel that short-term additional accumulation in the mitochondria is negligible.

Results of the analysis of mitochondrial cpn60 seem to parallel the findings for HSP70. The constitutive mitochondrial cpn60 protein level does not appear to be affected by short- or long-term heat stress. This finding is in disagreement with that of Prasad and Stewart (1992), who saw induction of mitochondrial cpn60 in maize (cv Black Mexican Sweet) seedlings. During purification of cpn60 we observed proteins that co-sedimented with the cpn60 complex. It is possible that the co-sedimenting proteins could be associated with the cpn60 protein complex or they could simply be proteins in another large complex. The cpn60A MAb appears to have epitope specificity that is moderately conserved between the plastid and the mitochondrial forms of the GroEL homolog (Fig. 6, bottom left). As expected, we did not observe any immunologically similar proteins in the cytoplasmic fractions of the plant material tested.

The sHSPs that are present in several plant subcellular fractions share significant structural homology and have been identified in a number of plant species. Recombinant forms of the cytosolic class of sHSPs have been characterized as molecular chaperones in vitro (Lee et al., 1995; Waters et al., 1996). We have identified and characterized two 22-kD HSPs that are heat inducible and appear to be expressed constitutively at very low levels. Using Edman degradation techniques, we have obtained N-terminal sequences for the mature HSP22A and HSP22B polypeptides. Characterization of the HSP22 proteins was aided by the production of polyclonal antibodies and MAbs. The HSP22 MAb does not recognize any proteins in the chloroplastic or cytosolic fractions isolated from stressed or nonstressed maize plants. In addition to binding HSP22, the polyclonal antibodies and MAbs were observed to bind to minor proteins of about 30 kD (Fig. 4, center and bottom). These spots may represent precursor forms of HSP22. If the 30-kD species is pre-HSP22 then it would apparently have an approximately 8-kD transit peptide, which would be significantly larger than the transit peptide identified for the pea mitochondrial HSP22 protein precursor (Lenne et al., 1995). Using the MAb raised to maize mitochondrial HSP22 we have identified a nearly full-length cDNA for maize HSP22 (Fig. 12). The 5 sequence of the gene was extended to complete the cDNA using three homologous sequences from the Pioneer Hi-Bred Maize Expressed Sequence Tagged database. As has been demonstrated for the other members of the mitochondrial sHSP group (Waters et al., 1996), the maize HSP22 protein is more similar to other mitochondrial sHSPs than to the other maize sHSPs. In contrast to the sHSPs present in soybean (LaFayette et al., 1996), the maize mitochondrial HSP22 appears to have greater similarity to the HSP17.1 class I cytosolic protein than to the class III HSP26 chloroplastic form (Table II). The N-terminal processing site for the putative HSP22 precursor sequence is similar to the sequence identified for pea mitochondrial HSP22 (Lenne et al., 1995).

Our investigations have indicated that HSP22 expression during heat stress is the primary response that occurs in plant mitochondria. We have observed that mitochondrial HSP22 expression closely follows the onset and relief of heat stress. Our findings show that the half-life of HSP22 in the mitochondria of maize seedlings upon recovery from stress is approximately 4 h. This is in stark contrast to the findings of Lenne and Douce (1994), who showed that the mitochondrial HSP22 protein was present in the total protein extracts of pea leaves for at least 2 d after the heat stress with almost no loss of protein. The disparity in mitochondrial HSP22 protein stability observed between pea and maize plants may be attributable to differences in the developmental stages and/or the tissue type analyzed. The pea mitochondrial HSP22 has a stability that is similar to that which has been found for the pea chloroplastic HSP21 and pea cytoplasmic HSP18.1, which had half-lives of 52 ± 12 h (Chen et al., 1990) and 37.7 ± 8 h (DeRocher et al., 1991), respectively, after a 4-h heat stress at 16°C above the control growth temperature. The persistence of the pea mitochondrial sHSP and other plant sHSPs has been hypothesized to play a role in the plant's ability to establish thermotolerance by providing a memory of the heat stress that occurred on previous days (Lenne et al., 1995).

In this work we have identified the first mitochondrial sHSP, to our knowledge, to be characterized in a species known to be heat tolerant. From this research focused at the protein level, it appears that expression of HSP22 is the major response of plant mitochondria during long- and short-term exposure to heat stress. Our data suggest that maize mitochondrial HSP22 may have a protective role against immediate heat stress. The molecular chaperones characterized to date all are produced at relatively high concentrations because the chaperone processes require physical contact between the chaperone and the unfolded protein. This could explain why the molecular chaperone homologs of HSP70 and cpn60 are present in the mitochondria at a high constitutive level and do not appear to change despite severe and prolonged exposure to heat stress. These proteins may be highly expressed for constitutive folding of proteins or they may be expressed in excess to afford protection for stressful events. Mitochondrial HSP22 is apparently not necessary for constitutive protein folding, and appears to be expressed only during stress. This expression pattern suggests that it may be actively functioning as an inducible molecular chaperone, which is minimizing and/or repairing the damage caused by the heat stress, potentially to augment or replace the capabilities of HSP70 and cpn60. It is also possible that HSP22 levels are indicative of the physiological state of the mitochondria and are perhaps involved in organellar signaling of heat-stress damage.

	FOOTNOTES

   Received September 2, 1997; accepted November 26, 1997.
   The accession number for the 2mHSP22 sequence is AF035460.

	ABBREVIATIONS

Abbreviations: 2D, two-dimensional. HSP, heat-shock protein. MAb, monoclonal antibody. sHSP, small HSP.

	ACKNOWLEDGMENTS

We thank Dr. Gautham Sarath (University of Nebraska Center for Biotechnology Protein Sequencing Core Facility, Lincoln) for his N-terminal amino acid sequencing of the HSP22 protein from 2D blots.

	LITERATURE  CITED

Top Abstract Introduction Methods Results Discussion References

Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25: 3389-3402 [Abstract/Free Full Text]

Barent RL, Elthon TE (1992) Two-dimensional gels: an easy method for large quantities of proteins. Plant Mol Biol Rep 10: 338-344

Bates EEM, Vergne P, Dumas C (1994) Analysis of the cytosolic hsp70 gene family in Zea mays. Plant Mol Biol 25: 909-916 [CrossRef][ISI][Medline]

Boorstein WR, Ziegelhoffer T, Craig EA (1994) Molecular evolution of the HSP70 multigene family. J Mol Evol 38: 1-17 [ISI][Medline]

Chen Q, Lauzon LM, DeRocher AE, Vierling E (1990) Accumulation, stability, and localization of a major chloroplast heat-shock protein. J Cell Biol 110: 1873-1883 [Abstract/Free Full Text]

Cooper P, Ho THD (1983) Heat shock proteins in maize. Plant Physiol 71: 215-222 [Abstract/Free Full Text]

Cooper P, Ho THD (1987) Intracellular localization of heat shock proteins in maize. Plant Physiol 84: 1197-1203 [Abstract/Free Full Text]

Craig EA, Gambill BD, Nelson RJ (1993) Heat shock proteins: molecular chaperones of protein biogenesis. Microbiol Rev 57: 402-414 [Abstract/Free Full Text]

Denecke J, Goldman MHS, Demolder J, Seurinck J, Botterman J (1991) The tobacco lumenal binding protein is encoded by a multigene family. Plant Cell 3: 1025-1035 [Abstract/Free Full Text]

DeRocher AE, Helm KW, Lauzon LM, Vierling E (1991) Expression of a conserved family of cytoplasmic low molecular weight heat shock proteins during heat stress and recovery. Plant Physiol 96: 1038-1047 [Abstract/Free Full Text]

Dunbar B, Elthon TE, Osterman JC, Whitaker BA, Wilson SB (1997) Identification of plant mitochondrial proteins: a procedure linking two-dimensional gel electrophoresis to protein sequencing from PVDF membranes using a FastBlot cycle. Plant Mol Biol Rep 15: 46-61

Edwards GE, Huber SC (1981) The C₄ pathway. In PK Stumpf, EE Conn, eds, The Biochemistry of Plants: Photosynthesis, Vol 8. Academic Press, New York, pp 237-281

Ellis RJ, Hemmingsen SM (1989) Molecular chaperones: proteins essential for the biogenesis of some macromolecular structures. Trends Biochem Sci 14: 339-342 [CrossRef][ISI][Medline]

Ellis RJ, van der Vies SM (1991) Molecular chaperones. Annu Rev Biochem 60: 321-347 [CrossRef][ISI][Medline]

Elthon TE, McIntosh L (1986) Characterization and solubilization of the alternative oxidase of Sauromatum guttatum mitochondria. Plant Physiol 82: 1-6 [Abstract/Free Full Text]

Elthon TE, Nickels RL, McIntosh L (1989) Monoclonal antibodies to the alternative oxidase of higher plant mitochondria. Plant Physiol 89: 1311-1317 [Abstract/Free Full Text]

Gao Y, Thomas JO, Chow RL, Lee GH, Cowan NJ (1992) A cytoplasmic chaperonin that catalyzes -actin folding. Cell 69: 1043-1050 [CrossRef][ISI][Medline]

Gatenby AA (1992) Protein folding and chaperonins. Plant Mol Biol 19: 677-687 [CrossRef][Medline]

Gething M, Sambrook J (1992) Protein folding in the cell. Nature 355: 33-45 [CrossRef][Medline]

Giorini S, Galili G (1991) Characterization of HSP-70 cognate proteins from wheat. Theor Appl Genet 82: 615-620

Gupta RS (1990) Sequence and structural homology between a mouse t-complex protein TCP-1 and the chaperonin family of bacterial (GroEL 65 kDa heat-shock antigen) and eukaryotic proteins. Biochem Int 20: 833-841 [Medline]

Hayes MK, Luethy MH, Elthon TE (1991) Mitochondrial malate dehydrogenases from corn. Plant Physiol 97: 1381-1387 [Abstract/Free Full Text]

Hemmingsen SM, Ellis RJ (1986) Purification and properties of ribulosebisphosphate carboxylase large subunit binding protein. Plant Physiol 80: 269-276 [Abstract/Free Full Text]

Hemmingsen SM, Woolford C, van der Vies SM, Tilly K, Dennis DT, Georgopoulos CP, Hendrix RW, Ellis RJ (1988) Homologous plant and bacterial proteins chaperone oligomeric protein assembly. Nature 333: 330-334 [CrossRef][Medline]

Hendrick JP, Hartl FU (1993) Molecular chaperone functions of heat-shock proteins. Annu Rev Biochem 62: 349-384