Biochemical Journal

Research article

The dramatically increased chaperone activity of small heat-shock protein IbpB is retained for an extended period of time after the stress condition is removed

Wangwang Jiao, Weizhe Hong, Pulin Li, Shihu Sun, Jing Ma, Mengding Qian, Mengdie Hu, Zengyi Chang


sHSP (small heat-shock protein) IbpB (inclusion-body-binding protein B) from Escherichia coli is known as an ATP-independent holding chaperone which prevents the insolubilization of aggregation-prone proteins by forming stable complexes with them. It was found that the chaperone function of IbpB is greatly modulated by the ambient temperature, i.e. when the temperature increases from normal to heat-shock, the chaperone activity of IbpB is dramatically elevated to a level that allows it to effectively bind the aggregation-prone client proteins. Although it is generally believed that the release and refolding of the client protein from the sHSPs depends on the aid of the ATP-dependent chaperones such as Hsp (heat-shock protein) 70 and Hsp100 when the ambient temperature recovers from heat-shock to normal, the behaviour of the sHSPs during this recovery stage has not yet been investigated. In the present study, we examined the behaviour and properties of IbpB upon temperature decrease from heat-shock to normal. We found that IbpB, which becomes functional only under heat-shock conditions, retains the chaperone activity for an extended period of time after the heat-shock stress condition is removed. A detail comparison demonstrates that such preconditioned IbpB is distinguished from the non-preconditioned IbpB by a remarkable conformational transformation, including a significant increase in the flexibility of the N- and C-terminal regions, as well as enhanced dynamic subunit dissociation/reassociation. Intriguingly, the preconditioned IbpB displayed a dramatic decrease in its surface hydrophobicity, suggesting that the exposure of hydrophobic sites might not be the sole determinant for IbpB to exhibit chaperone activity. We propose that the maintenance of the chaperone activity for such ‘holdases’ as sHSPs would be important for cells to recover from heat-shock stress.

  • chaperone activity
  • inclusion-body-binding protein B (IbpB)
  • molecular chaperone
  • small heat-shock protein (sHSP)
  • structural flexibility


Living organisms have evolved ways to protect themselves under stress conditions. Stress proteins, or molecular chaperones, have been identified and studied extensively as the type of protein that function to prevent the irreversible aggregation of the unfolded proteins under stress conditions such as heat-shock [1,2]. The stress proteins have been generally categorized into five families, Hsp (heat-shock protein) 100, Hsp90, Hsp70, Hsp60 and sHSP (small heat-shock protein), in accordance with the molecular mass of their subunits. Some of them (e.g. sHSP) mainly function under stress conditions by binding to the unfolded proteins, whereas others (e.g. Hsp60 and Hsp70) also assist the folding and assembly of the nascent polypeptides [24]. It is generally believed that the members of the five families of molecular chaperones work together as a network to exert quality control on proteins under normal or stress conditions [59].

sHSPs, ubiquitously and abundantly present in various living organisms [2,4,10], function as ATP-independent ‘holdase’ chaperones [2,11], whose defect has been related to a variety of human diseases, including desmin-related myopathy and cataract [12]. Previous observations demonstrate that sHSPs appear to play roles in preventing aging [13,14]. A unique feature of sHSPs is that they exist as dynamic homo-oligomers with the dissociation of the oligomers apparently being a prerequisite for them to exhibit chaperone activity [1517]. Successful determination of the crystal structures of the substrate-free forms of the archaeal Hsp16.5 and the plant Hsp16.9 revealed that the N-terminal regions are highly flexible in all of the subunits for the former or in half for the latter, whereas their C-terminal extensions extrude out from one subunit to interact with its neighbouring subunit [18,19]. More recent studies indicate that the N-terminal regions, which are highly variable in sequence and length, seem to play major roles in both the substrate binding and self-oligomerization for sHSPs [2022].

Ibp (inclusion-body-binding protein) A and IbpB, as two sHSP homologues in Escherichia coli, were originally identified as a component present in inclusion bodies [23], with the deletion of their encoding genes (ibpA/ibpB) reported to result in a low yield for producing recombinant proteins [24]. Both of them were demonstrated to exhibit chaperone activities under in vitro conditions [25,26]. Our previous studies revealed that the structure and chaperone activity of IbpB are highly temperature-responsive, with the activity detectable only at elevated temperatures, and that its terminal regions are highly flexible and involved in both the assembly of the higher-order oligomers as well as the chaperone activity of the protein [21]. Although it is believed that the client proteins bound to sHSPs would be released and refolded with the aid of the ATP-dependent chaperones (Hsp100 and Hsp70) after the heat-shock condition is removed, the structural property and behaviour of sHSP during the recovery stage have rarely been examined. The present study demonstrates that the enhanced chaperone activity of IbpB is retained for an extended period of time after preconditioning at the physiologically relevant heat-shock temperatures. Structural analysis of the preconditioned IbpB revealed a remarkable conformational transformation, as partially reflected by a significant increase in the flexibility for its N- and C-terminal regions, as well as an enhancement in dynamic subunit dissociation/reassociation of its oligomers. The biological implication of such retention of the chaperone activity for sHSPs is discussed.



The wild-type and mutant IbpB were purified as described previously [21]. Insulin, DTT (dithiothreitol) and trypsin were purchased from Sigma. The protein concentration was determined by the BCA (bicinchoninic acid) assay (Pierce) according to the manufacturer's instructions.

Preconditioning of IbpB

IbpB samples in 0.5 ml tubes were first incubated in water baths with temperatures at 25, 30, 37, 45, 53 and 60 °C for 20 min, before being transferred to a water bath at 25 °C for another 20 min.

Assay of chaperone activity

The relative chaperone activity was assayed basically following the methods described previously [15,21,27]. Briefly, the capacity for IbpB to suppress the DTT-induced aggregation of insulin B chains at specific temperatures (with insulin present at 0.5 mg/ml and DTT at 20 mM) was monitored at 360 nm using a spectrophotometer (Amersham Biosciences), and the relative chaperone activity was calculated as the percentage of the aggregation of DTT-induced insulin that was inhibited by the presence of IbpB at 25 °C: Embedded Image The spectrophotometer was equipped with a water-cycled machine which maintains the cuvette at the specific temperature. For the stability of measurement, the insulin sample was freshly prepared. For better solubility of insulin, a small amount of HCl was added into the insulin stock solution.

Analytical size-exclusion chromatography

The size-exclusion chromatography was performed on an AKTA Purifier operation system (Amersham Biosciences) at room temperature (25 °C) using a self-packed Superdex 200 XK 16/70 column (see Figure 2A) or a prepacked Superose HR 10/30 column (see Figure 6) with phosphate buffer (50 mM sodium phosphate and 50 mM NaCl, pH 7.3) as the eluent. The flow rate was set at 1 ml/min for the former and 0.4 ml/min for the latter. The standard markers used for the calibration were Blue Dextran (2 MDa), thyroglobin (660 kDa), ferritin (440 kDa), catalase (232 kDa), BSA (68 kDa), ovalbumin (45 kDa) and lysozyme (14 kDa).

CD spectroscopy

Far-UV CD measurements were performed using a J-715 spectropolarimeter (Jasco). A 300 μl volume of preconditioned or non-preconditioned IbpB (0.2 mg/ml) in a 2-mm-pathlength quartz cuvette was analysed at room temperature. The spectra represent the average of 16 scans corrected for buffer absorbance.

Tryptophan intrinsic fluorescence spectroscopy

The tryptophan fluorescence was measured using a Hitachi F4500 fluorescence spectrophotometer at an excitation wavelength of 280 nm and slit widths set at 5 nm. The fluorescence intensity is in arbitrary units. Emission spectra were plotted between 300 and 400 nm and fitted with a ninth-order polynomial equation.

Limited digestion of IbpB by trypsin

When the protein digestion by protease is limited by means of the low concentration of protease and non-optimal conditions, cleavage occurs only on the flexible and surface regions of proteins [28]. In the present study, the proteolysis of IbpB was performed at 25 °C for 30 min, and the digestion was stopped by boiling with supplemented SDS loading buffer. The concentration of IbpB was 1 mg/ml, with trypsin at 0.005, 0.01, 0.015, 0.02, 0.05 or 0.1 mg/ml. The protein fragments were separated by Tricine SDS/PAGE (16.5% gels) [29] and visualized with Coomassie Brilliant Blue R250.

Subunit-exchange studies by FRET (fluorescence resonance energy transfer)

The subunit-exchange study was performed mainly as described previously [30]. Briefly, IbpB was covalently labelled with the fluorescent probes AIAS {4-acetamido-4′-[(iodoacetyl)amino]stilbene-2,2′-disulfonic acid} or LYI (Lucifer Yellow iodoacetamide) (Molecular Probes), separately by incubating the protein samples (1 mg/ml) in 20 mM Mops buffer (pH 7.9) containing 100 mM NaCl. LYI and AIAS were at final concentrations of 8.5 and 3.2 mM respectively, and the reactions were allowed at proceed at room temperature in the dark for 12 and 24 h respectively. Covalently bound probe was completely separated away from the free probe on a Superdex 200 HR 16/30 column. To initiate the subunit-exchange reaction, 30 μM AIAS- and 30 μM LYI-labelled IbpB were mixed together in a cuvette at 25 °C. The sample was subsequently excited at 335 nm, and the fluorescence emission spectra were recorded from 360 to 600 nm. The emission intensity at 415 nm was calculated, and the changes of the donor fluorescence intensity as a function of time, F(t), from that at zero time, F(0), were plotted. All fluorescence spectra were recorded using a Hitachi F4500 fluorimeter.

ANS (8-anilinonaphthalene-1-sulfonic acid) fluorescence assay

ANS and bis-ANS [1,1-bis(4-anilino)naphthalene-5,5′-disulfonic acid], the commonly used probes for detecting the hydrophobic surfaces on proteins [20,21,3134], were purchased from Molecular Probes. The ANS-binding fluorescence to preconditioned or non-preconditioned IbpB was monitored with the excitation at 396 nm and the emission recorded from 425 to 575 nm, using a Hitachi F4500 fluorescence spectrophotometer at room temperature. Bis-ANS, instead of ANS, can be covalently photo-incorporated into the proteins. For photo-incorporation, bis-ANS was incubated with each protein sample in phosphate buffer, with the protein final concentration at 0.1 mg/ml and bis-ANS at 50 μM. Such mixtures were then subjected to cross-linking reactions as induced by exposure to UV light in a UVC 200 UV cross-linker (Amersham Biosciences) (with power at 120000 μJ/cm2) for 20 min at 25 °C.


IbpB retains a high level of chaperone activity after the heat-shock stress condition is removed

Our previous work demonstrated that IbpB exhibits little chaperone activity at room temperature, but high activity under heat-shock conditions [21]. In the present study, we asked how much and how long such activity would be retained when the heat-shock condition is removed. To address this, IbpB was pre-heated at various temperatures (30–60 °C) and then cooled to 25 °C before the chaperone activity, i.e. the capacity of such preconditioned IbpB to suppress DTT-induced aggregation of insulin B chains, was measured. Figure 1(A) demonstrates that IbpB retained a high level of chaperone activity at room temperature after being pre-heated to 37 °C or to higher temperatures.

Figure 1 IbpB is able to continually bind the aggregation-prone substrate protein after experiencing heat-shock conditions

(A) The relative chaperone activities of IbpB measured at 25 °C after being preconditioned to temperatures up to 60 °C. The relative chaperone activities were taken as the percentage of DTT-induced insulin aggregation inhibited by the presence of 0.4 mg/ml IbpB (for details, see the Materials and methods section). (B) Time-dependent light-absorbance curves for the insulin B chain aggregation performed at 25 or 45 °C in the presence or absence of 0.5 mg/ml IbpB. Curve 0′ represents the time-dependent turbidity of the insulin B chain in the presence of non-preconditioned IbpB at 25 °C, whereas curve 0 is that in the presence of IbpB heated at 45 °C. The turbidity curves for insulin alone at 25 and 45 °C are also shown. For curves 1–5, IbpB was first preconditioned at 45 °C for 20 min, and then cooled to 25 °C for a specific period of time (20, 90, 170, 240 and 1440 min for curves 1–5 respectively). The chaperone activity determinations were done with duplicate samples, and one group of curves is shown in (B). (C) Shown is the quantitative description for the relative chaperone activities of the IbpB after experiencing heat-shock conditions shown in (B). The chaperone activity of non-preconditioned IbpB was set as zero. It should be noted that its actual activity is much lower than zero, since the non-preconditioned IbpB even increases the aggregation turbidity of the insulin B chain. Points 0′ and 0–5 corresponds to the curves 0′ and 0–5 in (B) respectively.

The duration within which the chaperone activity can be retained for IbpB after pre-heating treatment was then examined by conducting an activity measurement after keeping the treated sample at room temperature for the indicated period of time. Figures 1(B) and 1(C) demonstrate that the heat-shock-preconditioned IbpB (at 45 °C) exhibited a significantly high level of activity hours after being withdrawn from the treatment (curves 1–4 in Figure 1B, and the corresponding scheme in Figure 1C showing the levels of chaperone activity), with approx. 10% of the relative chaperone activity retained even after the sample was kept at 25 °C for 24 h (curve 5 in Figure 1B). In contrast, the untreated IbpB exhibited hardly any chaperone activity at 25 °C (Figures 1B and 1C).

The conformation of heat-preconditioned IbpB is significantly transformed

The overall structural transformation in IbpB after the pre-heating treatment was then examined. A significant conformational transformation for the preconditioned protein was clearly demonstrated by results of the following analyses: size-exclusion chromatography on the oligomeric sizes (Figure 2A); intrinsic fluorescence measurements of the micro-environment of the aromatic residues (Figure 2B); far-UV CD of the secondary structure (Figure 2C); and protein-binding-dependent ANS fluorescence on the surface hydrophobicity (Figure 2D). Remarkably, data from all these various analyses consistently demonstrate that the IbpB samples pre-treated at a temperature between 30 and 60 °C fall into two categories: those preconditioned at 30 °C or lower possess a conformational state that exhibited little enhancement in chaperone activity, whereas those preconditioned at 37 °C or higher, possess a conformational state that exhibited a greatly enhanced chaperone activity (compare Figures 1A and 2). Figure 2 demonstrates significant structural alterations for IbpB that has experienced heat-shock preconditioning at 37 °C or higher, as indicated by the following observations: a significant decrease in the oligomeric size (Figure 2A); the intrinsic tryptophan fluorescence intensity (Figure 2B), or ANS fluorescence intensity (Figure 2D); and a slight loss of β-sheet content and increase of disordered structure, as shown by having a wavelength shift in minimum CD spectra from 211 to 213 nm (Figure 2C). It should be pointed out that the significant decrease in ANS-binding fluorescence for IbpB after experiencing the heat-shock preconditioning was not expected, in view of their significant increase in chaperone activity (see Figure 1A) and the general belief that a higher chaperone activity is correlated with a higher surface hydrophobicity [35].

Figure 2 The conformational transformation of IbpB after heat-shock preconditioning as monitored at room temperature by various techniques

IbpB was incubated at the indicated temperatures for 20 min, transferred to 25 °C for 20 min, and then the determinations for the curves were made all at room temperature. (A) The oligomeric state analysed by analytical size-exclusion chromatography. The peak positions of the eluted molecular standards of the indicated molecular masses are indicated along the top. mAu, milli-absorbance units. (B) Intrinsic fluorescence spectra. (C) Far-UV CD spectra. (D) ANS-binding fluorescence spectra. The curves in (B, C and D) are the average results from multiple scanning.

The N- and C-termini of the preconditioned IbpB become more susceptible to proteolytic cleavage

We have demonstrated previously that the N- and C-termini of IbpB are highly flexible, even at room temperature, and are involved in forming oligomers of this protein [21]. We then tried to find out whether there is any flexibility change for such termini after the heat-shock preconditioning, by subjecting the protein samples to limited proteolytic cleavage analysis [21,28]. Figure 3 demonstrates that the flexible termini of the preconditioned IbpB became more susceptible to trypsin digestion than the non-preconditioned protein. These results also demonstrate that the rest of IbpB (labelled ΔNC in Figure 3; having 11 residues removed from both the N-terminus and C-terminus as characterized previously via both MS and N-terminal sequencing [21]) exhibits the same level of susceptibility to trypsin digestion for both the heat-shock-preconditioned and the untreated proteins. These observations apparently suggest that the conformational transformation for the preconditioned IbpB mainly occurs in the N- and C-termini.

Figure 3 The terminal regions of IbpB become more susceptible to the limited proteolysis after experiencing heat-shock conditions

Dosage effect of trypsin cleavage on the preconditioned and non-preconditioned IbpB. The preconditioning temperature was 45 °C. Proteolysis was performed at 25 °C for 30 min. Concentrations for the non-preconditioned and preconditioned IbpB were kept constant at 1 mg/ml, whereas trypsin was present at 0.005 (lanes 1 and 1′), 0.01 (lanes 2 and 2′), 0.015 (lanes 3 and 3′), 0.02 (lanes 4 and 4′), 0.05 (lanes 5 and 5′) or 0.1 (lanes 6 and 6′) mg/ml. Protein fragments were separated by Tricine SDS/PAGE and visualized with Coomassie Brilliant Blue 250. Marked on the left are the positions of intact IbpB (IbpB) or IbpB lacking N- and C-termini (ΔNC) [21].

IbpB with truncated N- and C-termini exhibits neither conformational transformation nor chaperone activity after heat-shock preconditioning

To characterize further whether the flexible N- and C-termini are solely responsible for the conformational transformation and activity retention described above, we applied the ΔNC mutant IbpB in which both the flexible N- and C-termini were truncated [21] to similar heat-shock-preconditioning analysis. Figures 4(A) and 4(B) demonstrate that the apparent conformational transformation observed for wild-type IbpB was not detectable for the mutant protein after experiencing the same heat-shock preconditioning. These results strongly support that the flexible N- and C-termini are mainly responsible for the above-described conformational transformation for IbpB after being subjected to heat-shock preconditioning. Consistent with this, the ΔNC IbpB exhibited hardly any chaperone activity after experiencing such heat-shock preconditioning (Figure 4C).

Figure 4 Effects of heat-shock preconditioning on the conformation and chaperone activity for the ΔNC mutant IbpB

ΔNC mutant IbpB, with both N-terminal 11 and C-terminal 11 amino acids removed [21], was used. (A and B) Respective intrinsic fluorescence spectra and ANS-binding fluorescence spectra of the wild-type and mutant IbpB after experiencing preconditioning at the indicated temperatures. (C) Time-dependent light absorbance of the aggregates formed from the insulin B chain as a result for DTT-induction, performed at 25 °C, in the presence of preconditioned or non-preconditioned wild-type or ΔNC mutant IbpB (all at 0.5 mg/ml). The preconditioning temperature was 45 °C.

Dynamic dissociation/reassociation between IbpB subunits is enhanced after the heat-shock preconditioning

The sHSP oligomers were found to undergo dynamic dissociation/reassociation, which was in turn demonstrated to be a prerequisite for them to exhibit chaperone activity [15,17,30,36,37]. In view of our previous revelation that the flexible N- and C-terminal regions are involved in the oligomerization of IbpB [21] and the current observation that such termini become even more flexible with a remarkable retention of a high level of chaperone activity after heat-shock preconditioning (Figures 1 and 3), we then asked whether the dynamic oligomeric dissociation/reassociation of IbpB was also enhanced after such heat-shock. This would be reflected by an increase in the rate of subunit exchange, a parameter measurable using FRET [30,36].

For this purpose, the IbpB was labelled separately with two different fluorophores: either AIAS (as the donor) or LYI (as the acceptor). Exchange of subunits between these two sets of labelled IbpB would result in the coexistence of the two fluorophores within one IbpB oligomer, so that the fluorescence energy of the donor (AIAS) will be resonantly transferred to the acceptor (LYI) at ∼415 nm, being the emission wavelength of the former and the excitation wavelength of the latter. The level of FRET between the donor and acceptor can thus be estimated by measuring the decrease of fluorescence intensity at 415 nm. It follows that extent of dynamic disassociation/reassociation of IbpB would be reflected by the value of slope of the time-dependent emission fluorescence-intensity curve for AIAS at 415 nm (see the Materials and methods section for details).

Figure 5 demonstrates that the heat-shock-preconditioned IbpB is able to exchange subunits at a rate significantly higher than that for the non-preconditioned protein, with the former reaching the equilibrium state in approx. 100 min (with a half-life of ∼25 min), and the latter reaching equilibrium in approx. 200 min (with a half-life of ∼110 min). These results strongly suggest that the dynamic subunit dissociation/reassociation for IbpB was greatly enhanced due to the heat-shock preconditioning.

Figure 5 FRET between subunits of heat-shock-preconditioned or non-preconditioned IbpB

IbpB was labelled with AIAS (as the fluorescence donor) and LYI (as the fluorescence acceptor). For the preconditioning proteins, AIAS–IbpB and LYI–IbpB were preconditioned at 45 °C, cooled to 25 °C for 20 min, and then mixed for the subunit-exchange reaction. The rate of subunit exchange is reflected by the time-dependent change for the emission fluorescence intensity of AIAS recorded at 415 nm [k=dF415(t)/dt]. The starting fluorescence intensity was normalized to 1.

Consistent with the FRET results presented in Figure 5, Figure 6 reveals that, although the size of the oligomers for the non-preconditioned IbpB remain largely unchanged at various protein concentrations (Figure 6A shows that all of the protein samples were eluted around the void volume), the size of the oligomers for the heat-shock-preconditioned IbpB displayed a concentration-dependent alteration, becoming smaller at a lower concentration (Figure 6B). This observation supports the conclusion that the oligomers of IbpB become much more dynamic after the heat-shock preconditioning, with the interconversion between the large oligomeric (with dimers as the building blocks) and dimeric forms became even more effective.

Figure 6 Examination of the concentration effect on the size distribution of the oligomers of the polydisperse IbpB

Elution curves of the non-preconditioned (A) and heat-shock-preconditioned (B) IbpB of various concentrations (140, 35 and 9 μg in 100 μl of solution for 1, 1/4 and 1/16 samples respectively), as examined by size-exclusion chromatography performed at room temperature using a Superose 6 HR 10/30 column. mAu, milli-absorbance units. The peak positions of the eluted molecular standards of the indicated molecular masses are indicated along the top.

Heat-shock-preconditioned IbpB exhibits a decrease in the level of surface hydrophobicity

The unexpected observations presented in Figure 2(D), indicating a decrease instead of an increase in the level of surface hydrophobicity for the heat-shock-preconditioned IbpB, which was shown to exhibit an increase in the level of chaperone activity, prompted us to perform a more systematic investigation of the properties of such preconditioned IbpB.

The heat-shock-preconditioned IbpB was first subjected to an ANS titration as well as a Scatchard analysis [32,38] in an attempt to estimate the number of ANS molecules bound per IbpB. The titration data presented in Figure 7(A) indicate that the ANS molecule binds to the preconditioned IbpB in a remarkably less effective manner than to the non-preconditioned protein, although both eventually become saturated. Scatchard analysis (inset of Figure 7A) of such binding curves allowed us to estimate the number of molecules of ANS bound per subunit of the preconditioned and non-preconditioned IbpB to be approx. 3.5 and 5.3 respectively.

Figure 7 The level of hydrophobic surface is not quantitatively correlated with the chaperone activity for IbpB

(A) Saturation curves for the ANS fluorescence intensities at 490 nm when a constant amount of preconditioned or non-preconditioned IbpB was titrated with an increasing amount of ANS. The preconditioning temperature for IbpB is 45 °C. The inset is the Scatchard analysis plot for the binding of ANS to IbpB, which indicates 3.6 and 5.3 molecules of ANS bound to preconditioned and non-preconditioned IbpB respectively. (B) Light-absorbance curves for the DTT-induced insulin B chain aggregation occurring at 25 °C in the presence of the same amount of preconditioned 0.5 mg/ml IbpB with or without bis-ANS incorporated. The aggregation turbidity was assayed by the absorbance at 488 nm (instead of 360 nm) because of the presence of bis-ANS.

Nevertheless, the lack of correlation between the level of chaperone activity and that of exposure of hydrophobic surfaces does not necessarily mean that the hydrophobic surface is not important for the chaperone function, since our data demonstrate that the chaperone activity of heat-shock-preconditioned IbpB becomes greatly decreased when its hydrophobic surfaces were irreversibly blocked by bis-ANS incorporation (Figure 7B). It should be noted that the bis-ANS incorporation hardly affected the structure of IbpB as examined by CD and size-exclusion chromatography (results not shown). This observation indicates that, although there is a lack of correlation between the level of chaperone activity and that of hydrophobic surfaces, the exposed hydrophobic surfaces are essential for IbpB to exhibit chaperone activity. Similar lack of correlation between the increase of hydrophobic exposure and that of chaperone activity was also observed for Hsp16.3, where the electrostatic interaction was hinted to play an important role for Hsp16.3 to bind to the client proteins [39]. An isothermal titration calorimetric study on α-crystallin also suggests that electrostatic interaction may be involved in binding client proteins [40]. These observations indicate that surface hydrophobicity is not solely responsible for sHSPs to bind to the aggregation-prone forms of the client proteins.


Recent studies by us and others have revealed that certain stress proteins are able to exhibit immediate conformational transformations accompanied by a dramatic increase in chaperone activity for responding to stress conditions, and such instant activation of the stress proteins was proposed to provide the essential immediate protection for the cells to survive under stress conditions such as heat-shock and acidity [15,16,21,31,41]. IbpB, a sHSP from E. coli, responds to the ambient temperatures by elaborately modulating its chaperone activity. The present study was performed in an attempt to examine the behaviour of activated IbpB upon the removal of the stress condition. Our results demonstrate that the chaperone activity of IbpB is effectively retained for an extended period of time, even after the stress condition is removed (Figure 1), implying that IbpB, which is activated by the heat-shock stress condition [21], is able to continue to bind to the unfolded client proteins, even after the stress conditions are removed. These observations strongly suggest that sHSPs such as IbpB are able to play a protective role not only under stress conditions, but also after the stress condition is removed.

Retention of the chaperone activity for heat-shock-preconditioned IbpB was found to be accompanied by a remarkable structural transformation (Figure 2). In particular, the flexible N- and C-termini that were previously proposed to be involved in protein oligomerization were found to become even more flexible after experiencing such stressful conditions (Figures 3 and 4) and the dynamic dissociation/reassociation of the protein oligomers is also increased (Figures 5 and 6). An unexpected observation is that the retention of a high level of chaperone activity by IbpB after experiencing the heat-shock conditions was correlated with a decrease, instead of an increase, of exposure for its hydrophobic surfaces, despite the fact that the hydrophobic sites are indeed important for the protein to bind to the aggregation-prone client proteins (Figure 7).

Similar retention of chaperone activity after experiencing stress conditions has also been observed for other sHSPs, such as α-crystallin [35,40], Hsp26 [42] and Hsp16.3 [43]. Such properties of the sHSPs would be advantageous for serving at least two roles: first to release the unfolded client proteins only when the cellular conditions become optimal after experiencing the stress conditions; and, secondly, to allow the sHSPs to bind to the unfolded client proteins in a more effective manner when facing a subsequent exposure to the stress conditions. The importance of such properties of the sHSPs might be better realized in view of the fact that they act as ‘holdases’, instead of ‘foldases’ [7,44]. It is generally believed that the release of the aggregation-prone client proteins from sHSPs as well as the subsequent refolding of them have to be aided by the ‘foldase’ chaperones such as Hsp100, Hsp70 and Hsp60, whose action requires a supply of ATP [1,3]. Releasing the bound client proteins from the ‘holdases’ before the stress-devastated cells recover at least to a state that resumes ATP generation would probably lead to an unwanted immediate aggregation of the released client proteins. Previous observations indicate that IbpB is partially degraded in the cells during the recovery stage after heat-shock preconditioning [5,45], but the detailed fate of IbpB has not yet been fully characterized. It is conceivable that free IbpBs (i.e. those that did not get a chance to bind to any unfolded client proteins) which have experienced such a stress condition, but were not degraded would somehow function in a more effective way when facing another immediate subsequent stress condition.

Currently, the relationships between the oligomeric status, conformational change and chaperone function for sHSPs are still not well understood. On the one hand, the dissociation of the oligomers was suggested as a prerequisite for sHSPs to exhibit chaperone activity [2,15,27], and on the other hand, increase of chaperone activity was also reported with the oligomeric sizes unchanged [16,42,43] or even increased [46,47]. For IbpB, the size of the oligomers apparently decreased after the heat-shock preconditioning (Figures 2A and 6). Our data presented here and elsewhere [16,17,43] apparently demonstrate the following concepts. First, it is not the oligomeric size itself, but rather the increased dissociation of the subunits, that contributes to the increased chaperone activities for sHSPs in general upon exposure to the stress conditions. Secondly, both subunit dissociation of sHSP and sHSP–substrate complex formation (i.e. binding affinity) are needed for sHSPs to exhibit effective chaperone activities that were observed to form large complexes between the sHSPs and the client proteins. The conformational transformation detected here for heat-shock-preconditioned IbpB might act as the driving force for the increases of both the dynamic dissociation and the substrate binding affinity, which are correlated with the retaining of chaperone activity.

Another point worthy of contemplation is the observation that the flexible N- and C-termini of IbpB become even more flexible after experiencing a heat-shock preconditioning, which correlates with the retention of the chaperone activity for an extended period of time after the stress conditions were removed. Our previous studies have demonstrated that these flexible termini play a major role in allowing IbpB to undergo effective stress-induced disassociation of IbpB oligomers, leading to a dramatic increase in the chaperone activity [20,21]. In view of our previous observations that the presumed flexible N-terminal plays a dual role, being involved not only in protein oligomerization, but also in binding aggregation-prone client proteins [20,48], it should be considered whether the N-terminal region is also involved in binding client proteins of IbpB. Direct involvement of the flexible termini in protein oligomerization has also been implicated for some other sHSPs, including Hsp16.5 and Hsp26 [49,50]. Recent studies on Hsp18.1 and Hsp16.9 also revealed that their flexible N-terminal regions are responsible for binding the client proteins [51]. High flexibility has been considered to be a general feature for molecular chaperones to bind to their client proteins [52]. The importance of conformational flexibility for chaperone proteins to bind to their unfolded client proteins is probably best demonstrated by our recent observation on HdeA, which exhibits chaperone activity only when it is transformed into a globally disordered conformation, with the well-folded structure exhibiting no activity [31]. The present study demonstrates once again that oligomeric dissociation/reassociation and structural flexibility of the N- and C-terminal regions are key features for sHSPs to exhibit the ‘holdase’ chaperone activities.


This work was supported by grants from the National Key Basic Research Foundation of China (No. 2006CB806508 and No. 2006CB910304) and National Natural Science Foundation of China (30570355 and 30670022). We appreciate Dr Xinmiao Fu, Dr Yan Qin, Dr Xuefeng Zhang and Dr Chong Liu (Peking University, Beijing) for the critical comments on this work, and Dr Yongzhang Luo, Dr Haimeng Zhou and Ms Xiaolan Ding (Tsinghua University, Beijing) for providing technical assistance and expert help in spectroscopic analysis.

Abbreviations: AIAS, 4-acetamido-4′-[(iodoacetyl)amino]stilbene-2,2′-disulfonic acid; ANS, 8-anilinonaphthalene-1-sulfonic acid; bis-ANS, 1,1-bis(4-anilino)naphthalene-5,5′-disulfonic acid; DTT, dithiothreitol; FRET, fluorescence resonance energy transfer; Hsp, heat-shock protein; IbpB, inclusion-body-binding protein B; LYI, Lucifer Yellow iodoacetamide; sHSP, small heat-shock protein


View Abstract