Chaperone activity of cytosolic small heat shock proteins from wheat Eman Basha 1, *, Garrett J. Lee 1, †, Borries Demeler 2 and Elizabeth Vierling 1 1 Department of Biochemistry & Molecular Biophysics, University of Arizona, Tucson, AZ, USA; 2 Department of Biochemistry, University of Texas, San Antonio, TX, USA Small Hsps (sHsps) and the structurally related eye lens a-crystallins are ubiquitous stress proteins that exhibit ATP- independent molecular chaperone activity. We studied the chaperone activity of dodecameric wheat TaHsp16.9C-I, a class I cytosolic sHsp from plants and the only eukaryotic sHsp for which a high resolution structure is available, along with the related wheat protein TaHsp17.8C-II, which represents the evolutionarily distinct class II plant cytosolic sHsps. Despite the available structural information on TaHsp16.9C-I, there is minimal data on its chaperone activity, and likewise, data on activity of the class II pro- teins is very limited. We prepared purified, recombinant TaHsp16.9C-I and TaHsp17.8C-II and find that the class II protein comprises a smaller oligomer than the dodecameric TaHsp16.9C-I, suggesting class II proteins have a distinct mode of oligomer assembly as compared to the class I proteins. Using malate dehydrogenase as a substrate, TaHsp16.9C-I was shown to be a more effective chaperone than TaHsp17.8C-II in preventing heat-induced malate dehydrogenase aggregation. As observed by EM, mor- phology of sHsp/substrate complexes depended on the sHsp used and on the ratio of sHsp to substrate. Surprisingly, heat-denaturing firefly luciferase did not interact signifi- cantly with TaHsp16.9C-I, although it was fully protected by TaHsp17.8C-II. In total the data indicate sHsps show substrate specificity and suggest that N-terminal residues contribute to substrate interactions. Keywords:sHsps;a-crystallins; protein folding; protein aggregation; luciferase. Small Hsps (sHsps) and the structurally related eye lens a-crystallins are ubiquitous stress proteins that exhibit ATP- independent molecular chaperone activity [1]. sHsps are defined by a conserved C-terminal domain of  90 amino acids, called the a-crystallin domain, which is flanked by a short C-terminal extension and a variable length, noncon- served N-terminal arm [2]. sHsps range in size between 15 and 40 kDa and form high molecular mass oligomers of 9–32 subunits, depending on the sHsp. sHsps are very efficient at binding denatured proteins, and current models propose that they function to prevent irreversible protein aggregation and insolubilization, thereby increasing the stress resistance of cells [1]. Plants are unusual among eukaryotes in that they express multiple sHsp gene families that appear to have evolved after the divergence of plants and animals [3–5]. While in other organisms sHsps are found in the cytosol, plants express both cytosolic sHsps and specific isoforms targeted to intracellular organelles. There are at least two types of sHsps in the cytosol, referred to as class I and class II proteins, which share only  50% identity in the a-crystallin domain and are estimated to have diverged over 400 million years ago [6]. Five separate gene families encode mitochon- drion, plastid, peroxisomal, nuclear and endoplasmic reticulum-localized sHsps, each with appropriate organelle targeting signals [3,4]. The evolutionary expansion of the plant sHsp family may be the result of selection pressure for tolerance to the many types of stresses encountered by plants when they made the transition to growth on land. In what way these sHsp families may serve specialized functions is unknown. High resolution structures of two sHsp oligomers are now available: the class I plant sHsp, Triticum aestivum (wheat) TaHsp16.9 C-I, and an sHsp from a prokaryotic archeaon, Methanococcus jannaschii MjHsp16.5. Although TaHsp16.9C-I is a dodecameric disk [7], and MjHsp16.5 forms a sphere composed of 24 subunits [8], both sHsp oligomers are built from a conserved dimer structure, and similar contacts between dimers stabilize the oligomer. Although the oligomer is the dominant species at optimal temperature for the organism, sHsp oligomers are in rapid equilibrium with dissociated species as revealed by subunit exchange [7,9–12], and some sHsps dissociate to a stable suboligomeric species at the heat stress temperatures at which they are predicted to be most active [7,13]. These dynamic properties are likely to be important for sHsp function. The mechanism of sHsp chaperone action is an area of active research. In vitro studies have shown that sHsps Correspondence to E. Vierling, Department of Biochemistry, University of Arizona, Tucson, AZ 85721–0106, USA. Fax: +1 520 621 3709, Tel.: + 1 520 621 1601, E-mail: vierling@u.arizona.edu Abbreviations: Hsp(s), heat shock protein(s); sHsp(s), small heat shock protein(s); MDH, porcine mitochondrial malate dehydrogenase; Luc, firefly luciferase; SEC, size exclusion chromatography. Present addresses: *Department of Botany, Tanta University, Tanta, Egypt; Monsanto, Co., 800 N. Lindbergh Blvd., St. Louis, MO 63167, USA. (Received 16 December 2003, revised 29 January 2004, accepted 6 February 2004) Eur. J. Biochem. 271, 1426–1436 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04033.x bind partially unfolded substrate proteins in an ATP-independent fashion [1]. Current models suggest that it is an sHsp dimer or other suboligomeric species that is the active substrate-binding unit [7,14]. However, sHsp/ substrate complexes are typically significantly larger than sHsp oligomers, consistent with some kind of reassocia- tion of sHsp subunits after substrate binding. A few studies have identified regions of sHsps that potentially interact with substrate [15–17], and sHsp/substrate inter- actions are proposed to involve hydrophobic contacts [1]. The inactive, non-native substrate that is associated with the sHsp can be refolded by ATP-dependent chaperones, primarily the Hsp70/DnaK system, although under some conditions, Hsp100/ClpB or GroEL may also be required [14,15,18–20]. To better define sHsp chaperone function and the potential differences in function between the divergent cytosolic class I and II plant sHsps, we initiated in vitro studies of the chaperone activity of class I wheat TaHsp16.9C-I in comparison to a wheat class II protein, TaHsp17.8C-II. As mentioned above, TaHsp16.9C-I is the only eukaryotic sHsp for which a high resolution structure is available, but no significant characterization of its chaper- one activity has been performed. Wheat TaHsp17.8C-II is  33% identical overall to TaHsp16.9C-I. Only two previ- ous studies have examined the chaperone activity of this classofsHsp,andnoplantclassIIsHsphasbeentested for the ability to support substrate refolding [21,22]. Both TaHsp16.9C-I and TaHsp17.8C-II are undetectable in vegetative plant tissues, but accumulate dramatically during heat stress and are also expressed during seed development (E. Basha & E. Vierling, unpublished observation). In vivo studies indicate that plant class I and II sHsps, although both present in the cytosol, do not coassemble into mixed oligomers, suggesting they have distinct functions in the cell [23]. We prepared purified, recombinant TaHsp16.9C-I and TaHsp17.8C-II and found that the class II protein compri- ses a smaller oligomer than the dodecameric TaHsp16.9C-I, suggesting class II proteins have a distinct mode of oligomer assembly as compared to the class I proteins. Using malate dehydrogenase (MDH) as a substrate, TaHsp16.9C-I was shown to be a much more effective chaperone than TaHsp17.8C-II in preventing heat-induced MDH aggrega- tion. Surprisingly, heat-denaturing firefly luciferase (Luc), a commonly used sHsp substrate, did not interact significantly with TaHsp16.9C-I, although it was fully protected by TaHsp17.8C-II. In total, the data indicate these sHsps show substrate specificity and suggest that the divergent sHsp N-terminal arm contributes significantly to substrate inter- actions. Materials and methods Bacterial expression and purification of Ta Hsp16.9C-I and Ta Hsp17.8C-II Triticum aestivum TaHsp16.9C-I (AZ 369) and Ta- Hsp17.8C-II (Accession number: AF350423) [24], were produced as recombinant proteins in Escherichia coli BL21 cells using the pJC20 expression plasmid [25]. Cells were grown in Luria–Bertani broth with 200 lgÆmL )1 carbeni- cillin at 37 °C (for TaHsp16.9C-I) or 32 °C(for TaHsp17.8C-II), induced by the addition of isopropyl thio-b- D -galactoside to 1 m M andthengrownforafurther 6 h. Purification of the recombinant sHsp from the soluble cell fraction was performed essentially as described in Lee and Vierling [25] with the following modifications. TaHsp16.9C-I was enriched in the 55–90% (w/v) ammo- nium sulfate fraction, while TaHsp17.8C-II was more concentrated in the 40–70% (w/v) fraction. For TaHsp17.8C-II, DEAE chromatography (diethylamino- ethyl-Sepharose Fast Flow resin; Sigma) was performed in 3.2 M urea (2.8 M urea for TaHsp16.9C-I). After DEAE chromatography, fractions containing sHsps were dialyzed into 25 m M Tris/HCl, 1 m M EDTA, pH 7.5 (T25E1 buffer) and applied to an hydroxyapatite column equilibrated in 10 m M Na/P i buffer, pH 7.5. The columns were eluted using 10–400 m M Na/P i buffer, pH 7.5. Fractions containing sHsps were pooled and dialyzed against T25E1 and concentrated, if necessary, to 1–2 mgÆmL )1 with an Amicon filter (YM10 membrane). Protein concentration was deter- mined using the Bio-Rad protein assay with BSA as a standard. Concentrations for the sHsp are expressed in terms of subunit molecular mass, which is 16721.96 Da for TaHsp16.9C-I and 17649.40 Da for TaHsp17.8C-II, as calculated without the start Met residue, which is removed in vivo in E. coli. Yields were  30 mgÆL )1 bacterial culture. Protein was stored at )80 °C. Gel electrophoresis SDS/PAGE was performed on 14.5% (w/v) acrylamide gels using standard procedures. Non-denaturing pore exclusion PAGE was performed on 4–18% (w/v) acrylamide gradient gels as described by Helm et al. [23,26]. Gels were stained with Coomassie Blue. Electron microscopy Proteins were applied to carbon-coated 200 mesh copper grids (Ted Pella, Inc., Redding, CA, USA) in 50 m M phosphate buffer, pH 7.5 at 6.0 l M subunits and negat- ively stained with 2% (w/v) uranyl acetate. The sHsp/ substrate complexes were obtained by incubating either TaHsp16.9C-I or TaHsp17.8C-II with MDH under condi- tions described for sHsp/substrate complex formation assays (below). Grids were viewed in a Philips 420 transmission electron microscope (Philips Electronics, Ein- dhoven, the Netherlands) and micrographs were taken at 82 000· magnification. Sedimentation velocity experiments Analytical ultracentrifugation was performed with a Beck- man Optima XL-A ultracentrifuge. Samples (450 lL) were centrifuged for 3.5 h at 4 °C and 40 000 r.p.m. in an AN 60 TI rotor using double sector epon centerpieces. Measure- ments were taken at 230 and 280 nm using a 0.001 cm radial step size in continuous measurement mode. Data were analyzed with ULTRA SCAN II version 6.2 for Unix (http://www.ultrascan.uthscsa.edu/), using the van Holde– Weischet method [27] and finite element analysis as described previously [28]. Ó FEBS 2004 Small Hsp chaperones (Eur. J. Biochem. 271) 1427 Thermal aggregation protection assays Thermal aggregation protection assays were performed with MDH essentially as described by Lee et al. [15] using 0.6 l M MDH and purified sHsps from 0.18 to 3.0 l M (monomer) in 50 m M NA/P i buffer, pH 7.5. Samples were incubated in 1 mL quartz cuvettes in a thermostated water bath at 45 °C. To quantify changes in light scattering, absorbance at 320 nm was taken before heating began and monitored throughout heating every 5 min. Bovine IgG (reagent grade, Sigma) at a final concentration equivalent to the weight of 1.8 l M monomer TaHsp16.9C-I or 3.0 l M monomer in thecaseofTaHsp17.8C-II was added to 0.3 l M MDH as a negative control. Aggregation protection of firefly luciferase (Luc) (Pro- mega) was assessed as follows. Luc at 1 l M was heated with 12 l M TaHsp16.9C-I or TaHsp17.8C-II subunits in 50 m M Na phosphate, pH 7.5 (denaturation buffer) for 15 min at 42 °C in siliconized 0.65 mL microcentrifuge tubes. After heating, samples were centrifuged for 15 min at 16 250 g and the supernatant fractions removed. The supernatant and pellet fractions were treated with SDS sample buffer and the entire amount analyzed by SDS/PAGE and Coomassie blue staining. sHsp/substrate complex formation and size exclusion chromatography Purified TaHsp16.9C-I and TaHsp17.8C-II were analyzed by size exclusion chromatography (SEC) on a Rainen HPLC using a Toso-Haas TSK G4000 SWXL column, in a mobile phase containing 250 m M Na/P i ,pH7.3and 200 m M NaCl. For analysis of sHsp/MDH complexes, 6.0 l M of TaHsp16.9 C-I or TaHsp17.8C-II subunits were incubated with different MDH concentrations in 50 m M Na/P i buffer, pH 7.5 for 30 or 90 min at 45 °C. After incubation, samples were cooled on ice for 2 min and centrifuged at 16 000 g for 15 min. NaCl was added to the supernatant to a final concentration of 200 m M .Samples were size-fractionated on the SEC column in a mobile phase containing 250 m M Na/P i , pH 7.3 and 200 m M NaCl. Analysis of sHsp/Luc complexes was performed similarly, except the concentration of Luc was 1 l M and the concentration of sHsps was 12 l M subunits. Samples were heated for 15 min at 42 °C as described by Lee et al.[15]. SEC was performed as above except the mobile phase was 200 m M Na/P i , 100 m M NaCl, pH 7.3. Firefly luciferase reactivation assays Luc was heat-inactivated at 42 °C in the presence of sHsp as described for formation of sHsp/Luc complexes above. Luc reactivation in reticulocyte lysate was assayed as described previously [18]. The sHsp/Luc mixture was diluted to 25 n M Luc in 50% rabbit reticulocyte lysate (Green Hectares, Oregon, WI, USA) in refolding buffer and incubated at 30 °C and assayed as described previ- ously. Luc activity was determined over time by adding 2.5 lL of the reticulocyte lysate reaction to 50 lLof Luciferase Assay Mix (Promega) and monitoring light emission in a Turner 20/20 luminominer. Activity is plotted as a percentage relative to that of an equivalent amount of native Luc measured prior to the heating step. As a negative control, 0.11 lgÆlL )1 bovine IgG was substituted for sHsp (equivalent weight) in the initial heat- inactivation step. Data points and error bars reflect the mean and standard deviation of three replicates. Results Comparison of TaHsp16.9 C-I and TaHsp17.8 C-II To produce recombinant wheat class I and II sHSPs for these studies, we utilized the wheat class I cDNA, TaHsp16.9C-I (Accession number, S21600), corresponding to the sHsp for which the high resolution structure (2.65 A ˚ ) has been described [7], and a new wheat class II cDNA, TaHsp17.8 C-II (Accession number, AAK51797) [24]. Amino acid sequence alignment illustrates the conserved and divergent regions of these two sHsps (Fig. 1). TaHsp16.9C-I and TaHsp17.8C-II have an overall identity of only  33%, but regions corresponding to secondary Fig. 1. Amino acid sequence alignment of TaHsp16.9C-I (Accession number S21600) and TaHsp17.8C-II (Accession number AAK51797). Identical residues are indicated with * and highly conservative replacements indicated with colons or periods under the alignment. Regions of secondary structure in TaHsp16.9C-I [7] are indicated above the alignment. The a-helices are displayed as open bars; b-strands as lines. The conserved a-crystallin domain extends from b-strand 2 to b-strand 9. Regions in gray shaded boxes correspond to consensus regions within the a-crystallin domain that show particularly high conservation between plant sHsps. Residues in the N-terminal region shown in bold correspond to sequences conserved in all class I or class II proteins, respectively. Residues in the C-terminus in bold correspond to the conserved Basic-X-I/V-Q-I/V motif identified by de Jong et al. [29]. Underlined residues in the C-terminus of TaHsp17.8C-II correspond to a conserved motif of class II proteins [6]. The alignment was performed using the CLUSTAL - W program (European Bioinformatics Institute; http://www.ebi.ac.uk/clustalw/index.html). 1428 E. Basha et al. (Eur. J. Biochem. 271) Ó FEBS 2004 structure in the TaHsp16.9C-I a-crystallin domain along with the conserved C-terminal Ôbasic-X-I/V-Q-I/VÕ motif identified by de Jong et al. [29] show very high similarity. In contrast, the N-terminal arms show very little similarity, as is typical for sHsps [2,29], and each protein contains an N-terminal consensus unique to the class I or class II plant sHsps [6]. TaHsp17.8C-II also has a C-terminal motif containing ProProPro that is typical of class II plant sHsps. Thus, although these proteins would be predicted to have a similar fold in the a-crystallin domain and to utilize the hydrophobic residues of the basic-X-I/V-Q-I/V motif for oligomer assembly, differences in the N-terminal arms and flexibility of the C-terminal extension suggest that their overall oligomeric structure may differ. TaHsp16.9C-I, as reported previously [7], and TaHsp17.8C-II were purified to greater than 98% homo- geneity from E. coli cells, and the purified recombinant proteins migrated as a single species at the expected monomer mass on SDS/PAGE (Fig. 2A). Non-denaturing pore exclusion PAGE and size exclusion chromatography (SEC) were then utilized to compare the native structure of the two sHsps. By both methods, although TaHsp16.9C-I has a smaller monomeric size than TaHsp17.8C-II, TaHsp16.9 C-I appears to exist as a larger oligomeric structure than the class II sHsp. On nondenaturing PAGE TaHsp16.9C-I has an estimated mass of 284 kDa, while TaHsp17.8C-II migrates at 242 kDa. Similarly, on SEC the TaHsp16.9C-I peak eluted at 10.32 min while TaHSP17.8C-II eluted later at 10.65 min (Fig. 2C). Compared to TaHsp16.9C-I, TaHsp17.8C-II always exhibited a fairly broad elution profile, which could result from a variety of factors, including oligomeric instability, nonuniformity of oligomer size or interaction with the column matrix. As TaHsp16.9C-I is a 12-subunit oligomer [7], these results suggest the TaHsp17.8C-II oligomer is composed of fewer than 12 subunits. Size of the recombinant sHsp oligomers Although nondenaturing PAGE and SEC indicated the class I and II sHsps have different oligomeric structures, neither of these techniques are primary methods for size determination. Therefore, to better understand the differ- ence in subunit organization of these sHsps, we compared them by EM using negative staining and by sedimentation velocity centrifugation analysis. As shown in Fig. 3, purified preparations of either sHsp appear as mostly uniform, roughly spherical particles. The TaHsp16.9C-I particles have a diameter of approximately 11 nm, consistent with the crystal structure [7]. They are clearly larger than the TaHsp17.8C-II particles, which have an estimated diameter of only 9 nm. Therefore, the relative sizes of the two oligomers are consistent with their behavior on nondenaturing PAGE and SEC. Their appearance is also similar to what has been observed for class I and II sHsps from Pisum sativum (pea) [21], suggesting conservation of the subunit stoichiometry of class I and II oligomers. In sedimentation velocity experiments, the sedimen- tation distribution profile of both TaHsp16.9C-I and TaHsp17.8C-II indicate that both proteins are associated in a higher order structure (Fig. 4). The fact that these sHsps exhibit nearly identical sedimentation coefficient distribu- tions, but have different monomer molecular masses, is consistent with the interpretation that TaHsp17.8C-II contains either fewer subunits than TaHsp16.9C-I, or has a more nonglobular shape. Finite element analysis of the data [28] estimates a molecular mass of 201 kDa for TaHsp16.9C-I and 173 kDa for TaHsp17.8C-II. These data are consistent with a more extended shape for TaHsp16.9C-I than for TaHsp17.8C-II and with a dodecameric organiza- tion for TaHsp16.9C-I and an oligomer of TaHsp17.8C-II containing 9–10 monomer units. These data are in good agreement with the results from the other methods. In total, the data indicate these two plant cytosolic sHsps have a different oligomeric organization. The wheat sHsps prevent heat-induced aggregation of MDH Although the TaHsp16.9C-I structure is known, there is only one published experiment concerning its chaperone Fig. 2. Purified TaHsp16.9C-I and TaHsp17.8C-II form high molecular mass homo-oligomers. Purified TaHsp16.9C-I and TaHsp17.8C-II (5 lg) were separated by SDS/PAGE (A), nondenaturing pore exclu- sion PAGE (9 lg) (B), or size-exclusion HPLC (10 lg) (C). Gels were stained with Coomassie blue. (C) Elution time in min is shown relative to protein absorbance at 220 nm. Approximate elution positions of molecular mass standards are indicated. Asterisk indicates a peak arising from buffer absorbance. Ó FEBS 2004 Small Hsp chaperones (Eur. J. Biochem. 271) 1429 activity, in which it was shown to form complexes with the model substrate MDH when the two proteins are heated together [7]. We undertook a more detailed examination of the activity of TaHsp16.9C-I and, in comparison, TaHsp17.8C-II. A well-established assay for sHsp chaper- one activity is the ability to prevent heat-induced protein aggregation as measured by light scattering [25]. In this assay, the ratio of sHsp to substrate that is required to suppress light scattering can be used as a measure of effectiveness of the chaperone. Using this assay we tested the relative activity of the two wheat proteins in suppressing aggregation of MDH. As shown in Fig. 5, both wheat sHsps were effective in preventing MDH aggregation as assessed by suppression of an increase in light scattering over time at 45 °C. TaHsp16.9C-I achieved maximum protection of MDH at a ratio of two to three subunits of Fig. 3. Visualization of negatively stained TaHsp16.9C-I (left) and Ta Hsp17.8C-II (right) by electron microscopy. Both proteins appear as homogenous, roughly spherical particles, but TaHsp17.8C-II is smaller. Bar indicates 50 nm for both. Fig. 4. Sedimentation velocity analysis of TaHsp16.9C-I and TaHsp17.8C-II. Shown is the integral distribution plot from the van Holde–Weischet analysis of the data obtained for TaHsp16.9C-I (s) and TaHsp17.8C-II (d). For both proteins, the majority of the sample sedimented between 8.6 s and 9.2 s, with a small amount of smaller association states (less than 6% of the total concentration) sedimenting at S-values between 1 s and 8 s. Both samples also display some amount of slightly higher order association states (< 10%) sedi- mentingbetween9sand10s. Fig. 5. Wheat sHsps suppress heat-induced aggregation of MDH. Relative light scattering (320 nm) is plotted vs. time at 45 °Cfor 0.6 l M MDH (monomer) incubated with the indicated concentration (in monomers) of either TaHsp16.9C-I (top) or TaHsp17.8C-II (bot- tom). Numbers in parentheses indicate the ratio of sHsp monomer to MDH monomer. For the IgG control, IgG was added at a weight equivalent to 1.8 l M TaHsp16.9C-I or 3.0 l M TaHsp17.8C-II. Points represent average and standard deviation of three replicates. 1430 E. Basha et al. (Eur. J. Biochem. 271) Ó FEBS 2004 sHsp to one subunit of MDH. In contrast, TaHsp17.8C-II required four to five subunits of the sHsp per MDH subunit to achieve the same level of protection. Note that no aggregation protection is seen with the control protein IgG, even when used at the same concentration as the highest sHsp concentration (on a weight basis). Thus, while both sHsps are effective chaperones, with this substrate TaHsp16.9C-I is approximately twice as active on a subunit (or weight) basis. At the lowest concentrations of sHsp used (0.18 l M TaHsp16.9C-I and 0.60 l M TaHsp17.8C-II) the extent of light scattering was actually higher than in the absence of the sHsp. This may reflect the formation of very large aggregates of MDH that also include the sHsp. At low sHsp concentrations the sHsp could be bound to the MDH, but not be abundant enough to prevent interaction of unfolded MDH with itself. Analysis of sHsp/MDH complexes The differences in effectiveness of aggregation protection between the two wheat sHsps suggests that the way in which the sHsp and denatured substrate interact may be different. To characterize sHsp/MDH complexes, SEC analysis was performed after heating either TaHsp16.9C-I or TaHsp17.8C-II (6.0 l M ) with 1–4 l M MDH, yielding sHsp/MDH ratios comparable to those used in the light- scattering assays. MDH does not interact with either sHsp when the proteins are incubated together at 22 °C (Fig. 6A); the proteins elute at the predicted position based on their individual native molecular masses. We have noted that TaHsp17.8C-II consistently yields a lower absorbance (A 220 )thanTaHsp16.9C-I on column chromatography. We attribute this to either irreversible interaction of the protein with the column, or presence of aggregates too large to enter the column, but too small to be removed by brief centrifugation prior to loading. MDH incubated alone at 45 °C becomes insoluble and does not enter the column (Fig. 6B). However, a higher molecular mass species becomes visible after heating sHsps and MDH together at 45 °C for 30 or 90 min (Fig. 6C). At the ratio of sHsp:substrate of 6 : 1, TaHsp16.9C-I gives full substrate protection; the size of the complex peak increases between 30 and 90 min, and after 90 min the MDH peak is completely depleted (Fig. 6C, upper panels; compare to MDH peak in A). A majority of the MDH is in complexes after the first 30 min, consistent with the rate of aggregation protection measured by light scattering. The same phenomenon occurs at a ratio of TaHsp16.9C-I to MDH of 3 : 1, where full protection from aggregation is also observed. However, as the ratio of TaHsp16.9C-I to MDH is further decreased, the complex peak, while higher at 30 min, actually decreases over time, and aggregated MDH is now found in the sample pellet prior to column loading (not shown). All of the 30 min samples show a detectable complex peak that elutes earlier (7 min), which may represent some kind of intermediate in complex assembly. As the vast majority of complexes elute in the column void volume, differences in complex size at the different ratios cannot be estimated. Complex formation with TaHsp17.8C-II reveals the reduced capacity of this sHsp to protect MDH compared to TaHsp16.9C-I (Fig. 6C, lower panels). Although com- plete protection is observed at the sHsp:MDH ratio of 6 : 1, already at a 3 : 1 ratio the TaHsp17.8C-II/MDH complex peak does not increase after 30 min. This result is consistent with the light-scattering data, which showed the sHsp was unable to fully protect MDH at this ratio. As the amount of sHsp to substrate is further decreased, most of the MDH is no longer found in complexes, but rather is aggregated and removed by centrifugation prior to sample loading on the column (not shown). Interestingly, the sHsp itself does not appear to be complexed with the insoluble MDH at the 2 : 1, sHsp/MDH ratio; after 90 min the free sHsp is still all accounted for in the peak at 10.65 min. However, some sHsp is clearly lost to the insoluble fraction, which is not loaded on the column, when the ratio is only 1.5 : 1. This is consistent with the maximum light-scattering values observed for TaHsp17.8C-II/MDH at a 1 : 1 ratio, which were higher than those for MDH alone. A potential intermediate-sized species of complex is also evident at  6.5 min in most of the samples. In total, as observed by light scattering, substrate denaturation and aggregation are time-depend- ent, the sHsps can be saturated with substrate, and TaHsp17.8C-II is less effective in protecting MDH compared to TaHsp16.9C-I. To visualize the sHsp substrate complexes directly, samples incubated as for the SEC analysis at 45 °Cfor 30 min were observed by EM and negative staining (Fig. 7). Note that because samples were centrifuged prior to application to the grid, only soluble material was observed. Two consistent observations arose from this analysis. First, complexes formed at an sHsp to substrate ratio that was sufficient, or higher, than that required for full protection (as determined in the light-scattering experiments) were the most uniform. At the ratio required for full protection, complexes formed with TaHsp16.9C-I (3 : 1 ratio) had an average diameter of  54 nm, while complexes formed with TaHsp17.8C-II (6 : 1 ratio) were somewhat larger (60 nm). Second, complex regularity decreased as the amount of sHsp to substrate decreased below the level of full protection for either sHsp, with the irregular complexes looking more like aggregates composed of smaller particles ( 40– 46 nm), as seen in the 2 : 1 and lower ratio mixtures. MDH heated alone and applied to the grid before centrifugation was a large amorphous mass, while after centrifugation no proteinaceous material could be observed (not shown). Ta Hsp17.8 C-II, but not Ta Hsp16.9 C-I, suppresses aggregation of Luc The above data indicate that TaHsp16.9C-I is more effective in preventing aggregation of MDH than is TaHsp17.8C-II. To test if this difference in chaperone activity is the same with another heat sensitive substrate, aggregation protection of firefly luciferase (Luc) was examined. A simple differential centrifugation assay was employed to determine if either wheat sHsp could prevent insolubilization of Luc during heating. This assay was employed in place of the spectrophotometric assay used for MDH because of difficulties with adhesion of Luc to Ó FEBS 2004 Small Hsp chaperones (Eur. J. Biochem. 271) 1431 the cuvette walls. As shown in Fig. 8A, Luc incubated with TaHsp17.8C-II at 42 °C for 15 min was recovered almost exclusively in the soluble fraction, indicating that TaHsp17.8C-II was able to protect Luc from heat-induced insolubilization. A similar weight of IgG gave no protec- tion (not shown). Surprisingly, when Luc was incubated with TaHsp16.9C-I, virtually all of the Luc was found in the pellet fraction, while the sHsp remained soluble. Thus, in contrast to results with MDH, TaHsp16.9C-I is a less effective chaperone with Luc than is TaHsp17.8C-II. Note that a higher ratio of sHsp to substrate is required to protect Luc as compared to MDH. In parallel to these observations, TaHsp17.8C-II formed a complex when heated with Luc, which could be observed by SEC (Fig. 8B). No such complex formed with TaHsp16.9C-I (not shown). Thus, these two sHsps do not behave equivalently with all substrates. Denatured Luc bound to Ta Hsp17.8C-II can be reactivated in a cell free lysate The effectiveness of sHsp chaperone activity can also be assessed by the ability of an sHsp to maintain substrate in a state from which it can be refolded by ATP-dependent Fig. 6. TaHsp16.9C-I is more effective in forming complexes with MDH than is TaHsp17.8C-II. All samples were separated by SEC, and absorbance (220 nm) monitored over elution time. Samples were centrifuged to remove insoluble material prior to loading on the column. (A) sHsps (6 l M )andMDH (2 l M ) incubated together at room tempera- ture. (B) MDH heated alone. (C) High molecular mass complexes formed between sHsps and MDH after heating at 45 °Cfor 30 or 90 min. Concentrations were 6 l M sHsp subunits for all samples, with from 1 to 4 l M MDH as indicated by the sHsp/MDH ratio. Asterisk indicates a buffer peak. 1432 E. Basha et al. (Eur. J. Biochem. 271) Ó FEBS 2004 chaperones [18]. To test if the Luc protected by TaHsp17.8C-II was in a conformation capable of reacti- vation, the TaHsp17.8C-II/Luc complexes, or Luc heat- denatured in the presence of TaHsp16.9C-I or an equivalent weight of IgG, were incubated with reticulocyte lysate plus or minus ATP (Fig. 8C). Reactivation of Luc bound to TaHsp17.8C-II was highly efficient, achieving up to 70% reactivation in less than 1 h in the presence of ATP. As expected, TaHsp16.9C-I supported less than 15% reactivation of Luc because most of the Luc was insoluble and not associated with the sHsp. Controls using IgG, or in the absence of ATP, showed 5% or less reactivation. Thus, formation of TaHsp17.8C-II/Luc com- plexes is correlated with ability to support substrate reactivation. Discussion Our data provide the first detailed analysis of the in vitro chaperone activity of TaHsp16.9C-I, the only eukaryotic sHsp for which a high resolution structure is available. Surprisingly, although this sHsp effectively protects MDH from insolubilization, it did not interact with a second substrate, Luc, under the conditions tested. In parallel, we analyzed a related wheat sHsp, TaHsp17.8C-II, which proved to be less effective in protecting MDH, but interacted well with Luc, both preventing aggregation and supporting refolding. Thus, these results document the first clear example of apparent substrate specificity for sHsps. TaHsp16.9C-I and TaHsp17.8C-II represent two distinct classesofcytosolicsHspsfromplants(classIandclassII), Fig. 7. MDH/sHsp complexes visualized by electron microscopy and negative staining. TaHsp16.9C-I or TaHsp17.8C-II (6 l M sub- units) heated with different concentrations of MDH ranging from 6 to 1 subunit sHsp: 1 subunit substrate. Complexes were formed at 45 °C for 30 min then centrifuged and loaded on the EM grids for visualization at magni- fication of 820 000. Bar indicates 110 nm for all. Ó FEBS 2004 Small Hsp chaperones (Eur. J. Biochem. 271) 1433 estimated to have diverged at least 400 million years ago [6]. Comparing these two wheat proteins, amino acid sequence identity is 33% overall, and 46% for the a-crystallin domain. In addition to sequence differences, our analysis of the purified recombinant proteins indicates that they assemble into different quaternary structures. Solution methods, from this work and previous studies, and a crystal structure [7] demonstrate that native TaHsp16.9C-I is dodecameric. In contrast, by EM, SEC and sedimentation velocity experiments, the class II TaHsp17.8C-II was found to form regular, but smaller oligomers, with an estimated nine to ten subunits. Members of these same two sHsp classes have also been characterized from Pisum sativum (pea), PsHsp18.1C-I and PsHsp17.7C-II [21]. EM pictures of the purified pea oligomers are remarkably similar to those of the wheat proteins, with the class I sHsp having a diameter of 10–11 nm and the class II protein a slightly smaller diameter, despite the larger subunit size. PsHsp18.1C-I was also found to be dodecameric by sedimentation equilibrium analysis, like the homologous wheat TaHsp16.9C-I (amino acid sequence identity/simi- larity 68/75% throughout, and 80/86% in the a-crystallin domain). However, sedimentation equilibrium analysis of PsHsp17.7C-II estimated an oligomer of 11.3 ± 0.5 sub- units [21], larger than our estimate for the wheat class II protein. Therefore, it is unclear whether the stoichiometry of oligomeric assembly is the same for all plant class II proteins. However, we would predict that the assembly should comprise an even number of subunits, based on the dimeric building block of TaHsp16.9C-I, which involves features conserved in the class II proteins as well [1,6]. Regardless of absolute subunit numbers, class I and II sHsp oligomers clearly have distinct modes of assembly, as also reflected in the fact that these two classes of sHsps do not coassemble into mixed oligomers in vivo or in vitro, although class I or II sHsps will coassemble into normal oligomers when mixed with class I or II sHsps, respectively, from different plant species ([7,23], and E. Basha & E. Vierling, unpublished observation). Distinct assemblies of different sHsps in the same cell have also been observed in humans and bacteria [30,31], suggesting there are different, conserved roles for specific sHsps. Both TaHsp16.9C-I and TaHsp17.8C-II were able to suppress the heat-dependent aggregation of MDH. How- ever, TaHsp16.9C-I suppresses MDH aggregation com- pletely at a stoichiometry of 2–3 subunits sHsp to 1 subunit MDH. In contrast, complete suppression of MDH aggregation by TaHsp17.8C-II required the higher ratio of 4–5 sHsp subunits:1 MDH subunit. From previous work, PsHsp18.1C-I was found to be somewhat more effective in the aggregation protection of MDH than either of the wheat sHsps, suppressing MDH aggregation at a ratio of 2 : 1, sHsp subunit:MDH [15]. The pea class II protein was not tested with MDH, but when tested with citrate synthase, it was more than sixfold less effective than the PsHsp18.1C-I [21]. Thus, using these in vitro assays with two different substrates, class II proteins have proven to be less effective as chaperones than class I proteins, consistent with some type of substrate specificity for these two classes of proteins. At the lowest concentrations of sHsp used (0.18 l M TaHsp16.9C-I and 0.6 l M TaHsp17.8C-II to 1 l M MDH) the extent of light scattering was actually higher than in the absence of the sHsp. This may reflect the formation of very large aggregates of MDH that also include the sHsp, as evidenced by the loss of sHsp from the SEC profile under these conditions. At this low sHsp concentration, the sHsp might be bound to the MDH but not be abundant enough to prevent extensive interaction of unfolded MDH with itself. Bova et al. [32] noticed such an effect using aB-crystallin containing the R120G mutation linked to desmin-related myopathy. One of the authors’ interpreta- tions for the effect was a possible change in the availability of substrate binding sites resulting in a less efficient chaperone. When we used a low concentration of wheat sHsps, therefore providing fewer binding sites, we may have Fig. 8. TaHsp17.8C-II, but not TaHsp16.9C-I, maintains Luc in a soluble form during heating and facilitates Luc reactivation. (A) Coomassie blue stained SDS/PAGE of soluble (S) and pellet (P) fractions prepared after heating 12 l M TaHsp16.9 or 17.8 with 1.0 l M Luc at 42 °C for 15 min. (B) SEC analysis of 12 l M TaHsp17.8C-II plus 1.0 l M Luc either before (22 °C) or after heating at 42 °Cfor15min(42°C). Approximate elution times of molecular mass markers are indicated. (C) Time course of Luc reactivation in reticulocyte lysate. (d) TaHsp17.8C-II + ATP; (j) TaHsp17.8 C-II – ATP; (m) TaHsp16.9C-I + ATP; (h)Hsp16.9C-I–ATP;(s)IgG+ATP. 1434 E. Basha et al. (Eur. J. Biochem. 271) Ó FEBS 2004 imitated the same effect of the R120G mutation in aB-crystallin. SEC analysis showed that the complexes formed between the two wheat sHsps and MDH are quite large. Working with PsHsp18.1C-I, Lee et al. [15] found complexes with MDH were much smaller than those formed with the wheat sHsps, although the size observed by SEC was dependent on the substrate concentration as well as the denaturation temperature. The less efficient aggregation protection obtained with the wheat sHsps (on a molar basis of sHsp: substrate) compared to PsHsp18.1C-I suggests that the MDH aggregates more rapidly than it can form stabilizing interactions with the wheat sHsps. It is interesting that there isalwaysafreepeakofsHsponSEC,evenwhensome of the substrate has precipitated. The free sHsps could still have a role in protection, by cycling on and off the aggregates, as suggested by both Lindner et al.[33]and Friedrich et al. [12]. The decrease in SEC complex peak height and the eventual loss of sHsp at the highest substrate concentrations is due to the insolubility of the sHsps bound to excess substrate (as indicated by SDS PAGE; not shown). Transition of sHsps to an insoluble fraction is observed in vivo in many organisms [15,34,35], and may also result from overloading of the sHsp with substrates. Experiments in E. coli suggest that the chaperone ClpB is necessary to resolubilize sHsp/substrate complexes in vivo [36], and in vitro, protein aggregates containing sHsps are more effective ClpB substrates than aggregates without sHsps [19]. Therefore, even when complexed in an insoluble fraction, the sHsps may confer an advantage for recovery of protein activity in the cell. We also observed by EM that at sHsp:substrate ratios sufficient for complete substrate protection, sHsp/substrate complexes had dimensions of  56 and 60 nm for TaHsp16.9C-I and TaHsp17.8C-II, respectively. As repor- ted previously, and consistent with the SEC comparisons, PsHsp18.1C-I complexes with MDH were smaller on average, being frequently 16 to 20 nm [15]. Complex morphology also changed with decreasing sHsp to substrate ratio, with much more heterogeneous particles and aggre- gates of particles observed. These results are at odds with a report by Stromer et al. [37] in which complex morphology was reported to be dictated by substrate identity and independent of the identity of the sHsp, although different sHsp:substrate ratios were not observed by EM. It is interesting that  40 nm particles, termed heat shock granules, are found after heat stress in plants in vivo [34,38]. To what extent the in vitro-formed complexes resemble in vivo heat shock granules remains to be determined. Surprisingly, while TaHsp16.9C-I was more effective than TaHsp17.8C-II in protecting MDH, TaHsp16.9C-I showed no ability to protect Luc under the conditions tested. In contrast, TaHsp17.8C-II protected Luc from aggregation and formed high molecular mass complexes with Luc. We also showed that TaHsp17.8C-II supported Luc refolding using rabbit reticulocyte lysate as a source of ATP-dependent eukaryotic chaperones. However, the inability of TaHsp16.9C-I to protect Luc is not true for all classIsHsps.PsHsp18.1C-I has been shown to protect Luc with the same effectiveness as TaHsp17.8C-II and to support Luc refolding [12,15,18]. This fact indicates that the differences in sHsp–substrate interactions must be more subtle than the differences between class I and II sHsps in primary sequence or quaternary structure. TaHsp16.9C-I and PsHsp18.1C-I show 80% identity and 86% similarity in the conserved C-terminal a-crystallin domain. In contrast they show only 41% identity and 50% similarity in the N-terminal arm, suggesting substrate specificity is deter- mined by the N-terminal arm. The N-terminus of PsHsp18.1.C-I was also implicated in substrate interactions in bis-ANS binding experiments [15]. As it is proposed that substrate binding and protection involves oligomer dissoci- ation and some type of reassociation to form the large sHsp/ substrate complexes [7,13], it must also be considered that overall differences in oligomer stability and/or the kinetics of oligomer dissociation, rather than specific sequence differences, dramatically affect sHsp interactions with different substrates. Although to date sHsps have been ascribed little substrate specificity, it is clear from this study and previous work [21,37] that the effectiveness of substrate protection, on a molar basis, by different sHsps can vary significantly under the same conditions. A difference in effectiveness is obvious to the extreme with TaHsp16.9C-I, which fails to interact with Luc. It should be considered that minor differences in the ratio of sHsp/substrate required for maximal substrate protection are potentially functionally important differences in the cellular environment and are essentially an indication of substrate specificity. Full understanding of sHsp substrate interactions will require not only considera- tion of substrate binding sites and binding interactions, but also the dynamics of the sHsp oligomer and the kinetics of substrate aggregation. Acknowledgements This work was supported by National Institutes of Health grant RO1- GM42762, USDA-NRICGP and University of Arizona Experiment Station Funds, and American Cancer Society Faculty Research Award #FRA-420 to E. V. G. J. L. was a recipient of a National Institutes of Health Postdoctoral Fellowship. B. D. was supported by NSF BB1- 9974819. We thank Drs Kim Giese and Kenneth Friedrich for critical reading of the manuscript. References 1. van Montfort, R.L.M., Slingsby, C. & Vierling, E. (2002) Structure and function of the small heat shock protein/a-crystallin family of molecular chaperones. In Protein Folding in the Cell (Horwich, A.L., ed.), pp. 105–156. Academic Press, New York. 2. deJong, W.W., Leunissen, J.A. & Vooter, C.E. (1993) Evolution of the a-crystallin/small heat-shock protein family. Mol. Biol. Evol. 10, 103–126. 3. Waters, E.R., Lee, G.J. & Vierling, E. (1996) Evolution, structure and function of the small heat shock proteins in plants. J. Exp. Bot. 47, 325–338. 4. Scharf, K D., Siddique, M. & Vierling, E. (2001) The expanding family of small Hsps and other proteins containing an a-crystallin domain. Cell Stress Chaperones 6, 225–237. 5. Waters, E.R. & Vierling, E. (1999) Chloroplast small heat shock proteins: Evidence for atypical evolution of an organelle-localized protein. Proc.NatlAcad.Sci.USA96, 14394–14399. Ó FEBS 2004 Small Hsp chaperones (Eur. J. Biochem. 271) 1435 [...]... 33 34 35 36 37 38 heat- shock protein in Arabidopsis, can enhance osmotolerance upon overexpression Plant J 27, 407–415 Helm, K.W., Lee, G.J & Vierling, E (1997) Expression and native structure of cytosolic class II small heat shock proteins Plant Physiol 114, 1477–1485 Basha, E & Vierling, E (2003) Cloning and sequencing of a new cytosolic class II small heat shock protein from wheat Egypt J Biotechn... The diversification of plant cytosolic small heat shock proteins preceded the divergence of mosses Mol Biol Evol 16, 127–139 7 van Montfort, R.L.M., Basha, E., Friedrich, K.L., Slingsby, C & Vierling, E (2001) Crystal structure and assembly of a eukaryotic small heat shock protein Nat Struct Biol 8, 1025–1030 8 Kim, K.K., Kim, R & Kim, S.H (1998) Crystal structure of a small heat- shock protein Nature... E & Robinson, C.V (2002) Subunit exchange of multimeric protein complexes – Real-time monitoring of subunit exchange between small heat shock proteins by using electrospray mass spectrometry J Biol Chem 277, 38921–38929 10 Bova, M.P., Mchaourab, H.S., Han, Y & Fung, B.K.K (2000) Subunit exchange of small heat shock proteins – Analysis of oligomer formation of aA-crystallin and Hsp27 by fluorescence... Expression, purification, and molecular chaperone activity of plant recombinant small heat shock proteins Methods Enzymol 290, 350–365 Helm, K.W., LaFayette, P.R., Nagao, R.T., Key, J.L & Vierling, E (1993) Localization of small heat shock proteins to the higher plant endomembrane system Mol Cell Biol 13, 238–247 Van Holde, K & Weischet, W.O (1978) Boundary analysis of sedimentation-velocity experiments... Determination of molecular parameters by fitting sedimentation data to finite-element solutions of the Lamm equation Biophys J 74, 444–454 De Jong, W.W., Caspers, G.-J & Leunissen, J.A.M (1998) Genealogy of the a-crystallin -small heat- shock protein superfamily Int J Biol Macromol 22, 151–162 Studer, S & Narberhaus, F (2000) Chaperone activity and homoand hetero-oligomer formation of bacterial small heat shock proteins. .. stress-denatured proteins for subsequent refolding by a multichaperone network J Biol Chem 273, 11032–11037 21 Lee, G.J., Pokala, N & Vierling, E (1995) Structure and in vitro molecular chaperone activity of cytosolic small heat shock proteins from pea J Biol Chem 270, 10432–10438 22 Sun, W.N., Bernard, C., Van de Cotte, B., Van Montagu, M & Verbruggen, N (2001) At-HSP17.6A, encoding a small 23 24 25... & Horwitz, J (2002) Subunit exchange, conformational stability, and chaperone- like function of the small heat shock protein 16.5 from Methanococcus jannaschii J Biol Chem 277, 38468–38475 12 Friedrich, K.L., Giese, K., Buan, N.R & Vierling, E (2004) Interactions between small heat shock protein subunits and substrate in small heat shock protein/substrate complexes J Biol Chem 279, 1080–1089 13 Haslbeck,... of the mammalian lowmolecular-weight heat shock protein Mol Cell Biol 8, 5059– 5071 Mogk, A., Deuerling, E., Vorderwulbecke, S., Vierling, E & ¨ Bukau, B (2003) Small heat shock proteins, ClpB and the DnaK system form a functional triade in reversing protein aggregation Mol Microbiol 50, 585–595 Stromer, T., Ehrnsperger, M., Gaestel, M & Buchner, J (2003) Analysis of the interaction of small heat shock. .. (1999) Hsp26: a temperature-regulated chaperone EMBO J 18, 6744–6751 14 Ehrnsperger, M., Graber, S., Gaestel, M & Buchner, J (1997) ¨ Binding of non-native protein to Hsp25 during heat shock creates a reservoir of folding intermediates for reactivation EMBO J 16, 221–229 15 Lee, G.J., Roseman, A.M., Saibil, H.R & Vierling, E (1997) A small heat shock protein stably binds heat- denatured model substrates and... reactivate a heatdenatured protein Plant Physiol 122, 189–197 19 Mogk, A., Schlieker, C., Friedrich, K.L., Schofeld, H.J., Vierling, ¨ E & Bukau, B (2003) Refolding of substrates bound to small Hsps relies on a disaggregation reaction mediated most efficiently by ClpB/DnaK J Biol Chem 278, 31033–31042 20 Veinger, L., Diamant, S., Buchner, J & Goloubinoff, P (1998) The small heat- shock protein IbpB from Escherichia . Chaperone activity of cytosolic small heat shock proteins from wheat Eman Basha 1, *, Garrett J. Lee 1, †, Borries. ATP- independent molecular chaperone activity. We studied the chaperone activity of dodecameric wheat TaHsp16.9C-I, a class I cytosolic sHsp from plants and the only