Chaperone and antichaperone activities of trigger factor Guo-Chang Huang, Jia-Jia Chen, Chuan-Peng Liu and Jun-Mei Zhou National Laboratory of Biomacromolecules, Institute of Biophysics, Academia Sinica, Beijing, China Reduced denatured lysozyme tends to aggregate at neutral pH and competition between productive folding and aggregation substantially reduces the efficiency of refolding. Trigger factor, a folding catalyst and chaperone can, depending on the concentration of trigger factor and the solution conditions, cause either a substantial increase (chaperone activity) or a substantial decrease (antichaperone activity) in the recovery of native lysozyme as compared with spontaneous refolding. When trigger factor is working as a chaperone, the reactivation rates of lysozyme are decelerated and aggregation decreases with increasing trigger factor concentrations. Under conditions where antichaperone activity of trigger factor dominates, the reactivation rates of lysozyme are accelerated and aggregation is increased. Trigger factor and lysozyme were both released from the aggregates on re-solubilization with urea indicating that trigger factor participates directly in aggregate formation and is incorporated into the aggregates. The apparently dual effect of trigger factor toward refolding of lysozyme is a consequence of the peptide binding ability and may be important in regulation of protein biosynthesis. Keywords: chaperone; antichaperone; protein folding; trig- ger factor. Molecular chaperones assist protein folding by binding unfolded or misfolded chains and preventing or reversing misfolding or aggregation [1]. However, in certain cases, chaperones may also be involved in formation of aggregates [2–6]. This so-called Ôantichaprone activityÕ, or incitement to aggregate by a molecular chaperone, has been studied in most detail for protein disulfide isomerase (PDI) [7–11]. With aggregation-prone substrates and at substoichiometric concentrations, PDI promotes substrate aggregation ham- pering productive folding. PDI is involved directly in aggregate formation and is detected within the aggregates [7,8,11]. Antichaperone activity has also been observed for other chaperones, such as heavy chain-binding protein (BiP) [9]. Similar to PDI, low stoichiometries of BiP induces lysozyme aggregate formation. Furthermore, the aggregates formed may act as the intermediates that lead to amyloid diseases [12]. The participation of chaperones in aggre- gate formation may present an important physiological phenomenon [11]. The multifunctional Escherichia coli trigger factor was originally identified as being involved in the maintenance of a translocation-competent conformation of the precursor protein proOmpA (outer member protein A) in a cell free translation system [13] and stoichiometric complexes of trigger factor and proOmpA were isolated and studied [14,15]. Trigger factor was subsequently identified as a peptidyl-prolyl cis–trans isomerase [16,17] and was detected in the 50S subunit of functional ribosomes known to contain the peptidyl transferase center, which covers the exit domain of the nascent polypeptide chain [17]. Cooperation of enzymatic and chaperone functions makes trigger factor more effective than cyclophilins (CyPs), FK506 binding proteins (FKBPs) and the parvulin family in the catalysis of prolyl limited protein folding [18]. The groups of Luirink and Bukau have successfully cross-linked presecretory and nonsecretory proteins to trigger factor while still associated with the ribosome [17,19]. Further, trigger factor has been shown to be an important cofactor in GroEL-dependent protein degradation in E. coli and to promote binding of GroEL to unfolded proteins [20,21]. Trigger factor may also be a rate-limiting component in the degradation of abnor- mal proteins. Recently, trigger factor from Bacillus subtilis was reported to catalyze in vitro protein folding and to be necessary for viability under starvation conditions [22]. Trigger factor from Streptococcus pyogenes contributes post-transcriptionally to the secretion and processing of secreted cysteine proteinase (SCP) [23]. There is ample evidence that trigger factor plays an important and multi- functional role during protein synthesis in vivo and further facets to its role remain to be investigated. We reported that trigger factor could, as a molecular chaperone, inhibit aggregation and increase the reactivation yield of D -glyceraldehyde-3-phosphate dehydrogenase (GAPDH) [24]. A model for trigger factor assisted refolding of GAPDH and the conformational states that are prefer- entially recognized by trigger factor were proposed [24,25]. In order to further investigate how trigger factor influences the partitioning of an unfolded protein between folding and aggregation, here we examine the trigger factor assisted folding of reduced denatured lysozyme, in which folding is affected by buffer conditions. Lysozyme is a particularly appropriate substrate to study the chaperone activity in isolation from the isomerase activity of trigger factor, because the two prolyl bonds (Pro70 and Pro79) are both trans in native lysozyme, thus involvement of isomerization of prolyl bond during refolding is negligible as prolyl bonds Correspondence to J M. Zhou, National Laboratory of Biomacromolecules, Institute of Biophysics, Academia Sinica, 15 Datun Road, Chaoyang district, Beijing 100101, China. Fax: + 86 10 64872026, Tel.: + 86 10 64889859, E-mail: zhoujm@sun5.ibp.ac.cn Abbreviations: PDI, protein disulfide isomerase; CyP, cyclophilin; GSH, glutathione; GSSG, glutathione disulfide; GdnHCl, guanidine hydrochloride. (Received 19 March 2002, revised 19 July 2002, accepted 26 July 2002) Eur. J. Biochem. 269, 4516–4523 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.03145.x are also predominately trans in the unfolded chains [26]. The results show that in redox phosphate buffer in which spontaneous refolding of lysozyme is poor, trigger factor acts as a molecular chaperone that increases the reactivation yield and decelerates refolding rates. However, in redox Hepes buffer in which lysozyme refolds well spontaneously, low concentrations of trigger factor reduce the reactivation yield significantly and facilitate the formation of aggregates, behavior that has been described as ÔantichaperoneÕ [7–9]. In the aggregates, lysozyme is extensively cross-linked by intermolecular disulfide bonds and trigger factor partici- pates specifically in the mixed aggregates as an integral component. The dual effects of trigger factor toward refolding of lysozyme may be important in regulation of protein biosynthesis. MATERIALS AND METHODS Materials Hen egg white lysozyme was purchased from Serva, and GSSG and GSH were from Fluka. Bovine serum albumin (BSA), ovalbumin, Micrococcus lysodeikticus cell walls and dithiothreitol were obtained from Sigma. Hepes was from Merck. Guanidine hydrochloride (GdnHCl) was a product of ICN Biomedicals (Cosa Mesa, CA, USA), and urea was purchased from Nacalai tesque Inc. (Kyoto, Japan). Reagents for gel electrophoresis were from Bio- Rad. All other chemicals were local products of analytical grade. Trigger factor was expressed and purified as described previously [16]. Final trigger factor preparations were typically > 90% homogeneous as judged by SDS/PAGE. An absorbance coefficient of e 280nm ¼ 15 930 M )1 Æcm )1 , calculated using the procedure of Gill and von Hippel [27], was used for protein concentration determination. Cyclo- philin (CyP) was prepared from porcine kidney according to Kofron et al. [28]. The specific constant of the final product is about 1.9 · 10 7 M )1 Æs )1 when assayed using the chymo- trypsin-coupled method [29]. Reduction and denaturation of lysozyme Lysozyme at 20 mgÆmL )1 was completely reduced and denatured by incubation at room temperature for 4 h in 100 m M sodium phosphate buffer, pH 8.0, containing 8 M GdnHCl and 150 m M dithiothreitol. The reaction mixture was adjusted to pH 2.0 with 6 M HCl, and then dialyzed at 4 °C, first against 10 m M HCl and then against 100 m M acetic acid. The 200 l M reduced and denatured lysozyme was divided into aliquots and stored at )20 °C. Refolding of lysozyme Oxidative refolding of reduced and denatured lysozyme was achieved by dilution in various buffers as specified with or without different concentrations of trigger factor or CyP at 25 °C. The Hepes buffer, 0.1 M , pH 7.0, contained 2 m M EDTA, 5 m M MgCl 2 and 20 m M NaCl, and the phosphate buffer, 0.1 M , pH 7.5, contained 2 m M EDTA. If not otherwise specified, the refolding buffer contained 1 m M GSSG and 2 m M GSH (as the ratio of GSH to GSSG has been determined to be 2 in the endoplasmic reticulum) [30]. The final concentration of lysozyme for refolding was 10 l M . When GSSG and GSH were not present, the system was referred to as a nonredox buffer. Recovery of activity was complete 5 h after dilution and no further change was observed for at least 24 h. Lysozyme activity was deter- mined at 30 °C by following the lysis of Micrococcus lysodeikticus [7,31]. The decrease in A 450 of a 0.25 mgÆmL )1 cell suspension in 67 m M sodium phosphate buffer, pH 6.2, containing 100 m M NaCl was measured in a Shimadzu UV-1601 spectrophotometer. Protein concentra- tions were determined by measuring A 280 using absorbance coefficients of 36 636 M )1 Æcm )1 for native lysozyme and 33 014 M )1 Æcm )1 for denatured lysozyme. The time course of reactivation of lysozyme was followed by determining activities of samples withdrawn at the indicated times. The half-times were determined by fitted to a single-exponential function. Lysozyme itself is stable when subjected to the same treatment without denaturant. Aggregation of lyso- zyme upon dilution was monitored at 25 °Cby90° light scattering at 500 nm in a Hitachi F-4500 spectrofluorimeter. All measurements were repeated several times and the rate constants obtained were highly reproducible. Aggregate resolubilization The insoluble aggregates formed during refolding of lyso- zyme in the presence of 5 l M trigger factor in Hepes buffer were isolated according to the procedure described by Sideraki and Gilbert [11] as follows: aggregates were collected by centrifugation in a bench top centrifuge (6000 g for 8 min). After washing twice with Hepes refolding buffer, the pellets were resuspended in various concentrations of urea in buffer containing 0.1 M Hepes, pH 7.0, with or without 150 m M dithiothreitol. After four rounds of vortex mixing, the solution was incubated overnight at room temperature. After incubation with urea, the residual insoluble materials were separated from the supernatant by centrifugation in a bench top centrifuge (6000 g for 10 min) and then the proteins in the supernatant were quantified by reducing SDS/PAGE. In another set of experiments, SDS sample buffers with or without 2- mercaptoethanol were used to re-solubilize the pellets instead of urea and samples were examined on both reducing and nonreducing gels, respectively. RESULTS Refolding of lysozyme in phosphate buffer The spontaneous refolding of reduced and denatured lysozyme (10 l M ) in phosphate buffer with no redox component is only 1.4% (Fig. 1) and shows extensive aggregation as the native disulfide bonds of lysozyme cannot form. The presence of trigger factor at a concentra- tion within the range 5 l M to 20 l M (molecular ratios of 0.5–2) has no effect on lysozyme refolding in terms of reactivation yield (Fig. 1) or extent of aggregation under nonredox conditions. At pH 7.5, 25 °C, the spontaneous refolding of reduced denatured lysozyme in a glutathione redox phosphate buffer (1 m M GSSG, 2 m M GSH) is relatively rapid (t 1/2 ¼ 20.4 min, see later), but only 2.8% of the lysozyme folds productively (Fig. 1). Upon dilution of reduced denatured lysozyme in the presence of trigger Ó FEBS 2002 Chaperone and antichaperone activities of trigger factor (Eur. J. Biochem. 269) 4517 factor, the recovery of lysozyme activity increases with increasing molecular ratios of trigger factor to lysozyme until at 15 l M trigger factor, 17% of the lysozyme is productively folded (Fig. 1). Control experiments show that trigger factor neither affects the lysozyme activity assay directly nor do trigger factor preparations exhibit any apparent lysozyme activity. The amount of lysozyme activity recovered does not increase further during the 24 h after activity determination. Therefore, the partial recovery of lysozyme activity is due to irreversible misfold- ing and/or aggregation rather than a biphasic or kinetically incomplete reaction. The reduced and denatured lysozyme in the absence of trigger factor aggregated rapidly and to a significant degree upon dilution, as monitored by light scattering (Fig. 2). In the absence of trigger factor, light scattering started to increase within 10 min of dilution and approached a constant value at about 1 h. Accompanying the increase in reactivation yield (Fig. 1), the extent of lysozyme aggregation was inhibited markedly by increasing concentrations of trigger factor (Fig. 2). CyP, another peptidyl-prolyl cis–trans isomerase, was used as a compari- son to dissect out the isomerase and chaperone activities of trigger factor. Increasing concentrations of CyP showed no effect on either the extent of lysozyme reactivation (Fig. 1) or nonproductive aggregation (Fig. 2). There is essentially the same amount of native lysozyme recovered when refolding is performed in the absence of CyP (Fig. 1). The kinetics of reactivation of lysozyme in the presence of different concentrations of trigger factor or CyP is com- pared in Fig. 3. The half times (t 1/2 ) of lysozyme reactivation in the presence of trigger factor, as was found earlier for trigger factor assisted refolding of GAPDH [24], increase with increasing concentrations of trigger factor. The t 1/2 of Fig. 2. Effect of trigger factor on lysozyme aggregation in redox phos- phate (j) and redox Hepes (d)buffers.Aggregation of lysozyme upon dilution was monitored at 25 °Cby90° light scattering at 500 nm. Final levels of light scattering were determined 1 h after dilution. The concentration of lysozyme was 10 l M . A CyP control in redox phos- phate buffer is indicated as (h). Fig. 1. Effect of trigger factor or CyP on lysozyme re-activation in phosphate buffer. Refolding of lysozyme was initiated by a 20-fold dilution into 0.1 M phosphate buffer, pH 7.5, containing 2 m M EDTA. The reactivation mixtures were kept at 25 °Cfor5hbeforesamples were taken for assay of activity. Data are presented as the percentage of lysozyme refolded with respect to nondenatured lysozyme otherwise treated in exactly the same way. The final concentration of lysozyme for refolding was 10 l M (m)and(j) represent lysozyme in nonredox and redox buffers, respectively, in the presence of trigger factor. (h) represents lysozyme reactivation in the redox buffer in the presence of CyP. The data are fitted to an arbitrary curve. Fig. 3. Dependence of half times of lysozyme reactivation in redox phosphate (j) and redox Hepes (d) buffer, respectively, on the con- centrations of trigger factor. The refolding was followed by the regain of enzyme activity at a final concentration of lysozyme of 10 l M at 25 °C. The kinetic data were analyzed by fitted to a single-exponential function. The data shown are fitted to an arbitrary curve. CyP is used as a control at redox phosphate (h)andredoxHepes(s)buffer, respectively. 4518 G C. Huang et al.(Eur. J. Biochem. 269) Ó FEBS 2002 lysozyme reactivation in the presence of 20 l M trigger factor was 57.5 min, 2.8 times longer than that determined for spontaneous refolding. Stoichiometric concentrations of CyP had no effect on the kinetics of lysozyme reactivation. Thus, in a redox phosphate buffer, trigger factor behaves like a molecular chaperone that prevents denatured lyso- zyme from partitioning towards a nonproductive folding pathway. The chaperone-like activity of trigger factor is specific, other proteins such as BSA or ovalbumin, at comparable concentrations, have no effects on lysozyme refolding under the same conditions (data not shown). Refolding of lysozyme in Hepes buffer As in phosphate buffer, refolding of reduced denatured lysozyme in nonredox Hepes buffer showed almost no reactivation either in the presence or absence of trigger factor (Fig. 4). However, the spontaneous refolding of lysozyme in a glutathione redox Hepes buffer results in a recovery of activity of 42% (Fig. 4). Intriguingly, the reactivation of lysozyme decreases significantly with increasing trigger factor concentration at low molecular ratios. The reactivation yield of lysozyme decreased from 42% in the absence of trigger factor to 6.6% in the presence of 5 l M trigger factor (Fig. 4), indicating that trigger factor shows antichaperone activity in lysozyme refolding, similar to that observed for PDI [7–11]. When the concentration of trigger factor was greater than 5 l M , the reactivation curve of lysozyme shows a slow upward turn although the reactivation yields were still much lower than that of spontaneous refolding (Fig. 4), suggesting that antichaper- one and chaperone activities of trigger factor operate in competition to one another. We examined whether the decrease in lysozyme refolding yield is accompanied by aggregation by monitoring light scattering. As shown in Fig. 2, the extent of aggregation was found to increase with increasing trigger factor when the trigger factor concentra- tion was below 5 l M . Trigger factor alone, in control experiments, showed no scattered light under the same conditions. Further increase in trigger factor concentration resulted in a decrease in light scattering, indicating that chaperone activity begins to dominate under these condi- tions. Unlike trigger factor, CyP, as observed in phosphate buffer, has no influence in either recovery of native lysozyme (Fig. 4) or the extent of aggregation (Fig. 2). The kinetics of reactivation of reduced denatured lyso- zyme in redox Hepes buffer and in the presence of different concentrations of trigger factor or CyP were also investi- gated and the results are shown in Fig. 3. The half time of spontaneous reactivation was 48 min, which is slower than in phosphate buffer. While stoichiometric quantities of CyP show no effect on lysozyme refolding, trigger factor at substoichiometric concentrations accelerates the reactiva- tion rates to about 1.9 times that of spontaneous refolding at a molecular ratio of 0.25. However, the accelerated reactivation results in decreased recovery of activity of lysozyme (Fig. 4). When the molar ratio of trigger factor in the refolding buffer is increased above 0.5, the reactivation yields begin to increase (Fig. 4) and the extents of aggre- gation to decrease (Fig. 2). This is accompanied by a change from acceleration of the reactivation rates to deceleration (Fig. 3). The above results are similar to those reported by Puig and Gilbert [7,9] for antichaperone and chaperone activities of PDI except that refolding of lysozyme is catalyzed by PDI regardless of whether it is the chaperone or the antichaperone activity that predominates. Effects of NaCl and ethylene glycol on trigger factor-assisted lysozyme refolding Phosphate and Hepes buffers differ greatly in ionic strength [10], and the different effects of trigger factor on lysozyme refolding in the two kinds of redox refolding buffers prompted us to investigate the effects of the refolding buffer, in terms of ionic strength and hydrophobicity. As shown in Fig. 5, the spontaneous reactivation of lysozyme in redox Hepes buffer is dramatically affected by addition of 100 m M NaCl, decreasing from 42% to about 6%. On addition of trigger factor, the recovery of activity increases gradually with increasing trigger factor concentration and above 5 l M trigger factor is essentially the same as in buffer without added salt. It is interesting to note that under these conditions where the yield of spontaneous folding is low, trigger factor shows no detectable antichaperone activity. When 5% ethylene glycol instead of NaCl was added to the refolding system, the spontaneous reactivation of lysozyme reached a maximum of 47%, which is slightly higher than that in the absence of ethylene glycol. On addition of trigger factor the reactivation falls dramatically reaching a mini- mum of 4.8% at a trigger factor concentration of 5 l M .This value is slightly lower than in the absence of ethylene glycol. These small differences are highly reproducible. This indicates that addition of ethylene glycol causes a slight enhancement of the antichaperone effect. It seems that whether it is the antichaperone or the chaperone activity of trigger factor that dominates may be determined by the Fig. 4. Effects of trigger factor or CyP on lysozyme reactivation in Hepes buffer. The refolding was carried out in 0.1 M Hepes buffer, pH 7.0, containing 2 m M EDTA, 5 m M MgCl 2 and 20 m M NaCl. All other details were the same as for Fig. 1. (m)and(d) represent lyso- zyme in nonredox and redox buffers, respectively, in the presence of trigger factor. (s) represents lysozyme in redox buffer in the presence of CyP. Ó FEBS 2002 Chaperone and antichaperone activities of trigger factor (Eur. J. Biochem. 269) 4519 effect of the solution conditions on folding of the substrate itself. Composition of trigger factor accelerated aggregates As shown in Fig. 2, maximum formation of insoluble aggregates during lysozyme refolding under redox condi- tions occurs at a molecular ratio of trigger factor to lysozyme of 0.5 (10 l M lysozyme, 5 l M trigger factor) indicating that aggregate formation is accelerated by trigger factor. To understand the mechanism of aggregate forma- tion, the isolated aggregates were incubated in various concentrations of urea with or without 150 m M dithiothre- itol and the re-solubilized proteins were analyzed by reducing SDS/PAGE. As shown in Fig. 6 Aa,b, the proteins in aggregates were re-solubilized by urea and the total amount of soluble protein increased with increasing urea concentration. In each urea concentration, lysozyme and trigger factor were solubilized to the same extent and in the same ratio as the original reaction mixture within experimental error. This suggests that trigger factor and lysozyme coaggregates and is not present as separate aggregates. Aggregates formed under conditions that allow disulfide formation are highly cross-linked, which makes the trigger factor-lysozyme aggregates more resistant to extrac- tion with urea unless dithiothreitol is added (Fig. 6A,b). It should be noted that no covalent bonds between trigger factor and lysozyme can form, as trigger factor itself contains no cysteine residues. In order to understand whether trigger factor, when acting as an antichaperone, is integrated specifically into the mixed aggregates or only coprecipitates with rapidly formed lysozyme aggregates, we carried out experiments using BSA as a control. When lysozyme (10 l M ) was diluted into the Hepes buffer containing 5 l M BSAaswellas5l M trigger factor, the refolding was not affected either in recovery of activity or aggregation formation compared with in the presence of trigger factor alone (data not shown). The soluble and insoluble fractions formed in the presence of trigger factor and BSA were isolated by centrifugation and the aggregates were then incubated in SDS sample buffer with or without 200 m M 2-mercaptoethanol, respect- ively. The resolubilized proteins were measured by reducing or nonreducing SDS/PAGE, respectively. As shown in Fig. 6B,a, in contrast to a clear band of trigger factor resolubilized together with lysozyme, there is no visible BSA band on the gel. Clearly, BSA is not present in the mixed aggregates. This indicates that trigger factor does not coprecipitate with aggregated lysozyme in a nonspecific manner. In addition, comparison of electrophoresis under reducing (Fig. 6B,a) and nonreducing (Fig. 6B,b) condi- tions shows that aggregated lysozyme is highly cross-linked by disulfide bonds preventing re-solubilization of lysozyme from aggregates in the absence of 2-mercaptoethanol. After treatment in SDS sample buffer containing no 2-mercapto- ethanol, cross-linked lysozyme, although partially soluble, is present only as a high molecular weight species, indicating that intermolecular crosslinking has occurred (not shown). Under nonreducing conditions, there is also no BSA detected in the coprecipitated aggregates, although this result is most clearly seen under reducing conditions where BSA and trigger factor do not comigrate. It is clear that the coprecipitation of trigger factor with lysozyme is specific and is related to its antichaperone function. DISCUSSION Upon dilution into refolding buffer, the reduced denatured lysozyme faces two alternative fates: productive folding to form active enzyme or aggregation. The relative size of the populations that partition between productive folding and aggregation depends to a considerable degree on the solution conditions, of which the ionic strength and the nature of the redox reagents are significant factors [10]. In the absence of GSSG, the spontaneous reactivation of lysozyme in both phosphate and Hepes buffers is very low because disulfide formation cannot proceed efficiently in the absence of redox reagents. In this case, trigger factor shows no effect on the reactivation yield, indicating that correct disulfide formation is essential for productive folding of lysozyme. Increasing the ionic strength in redox phosphate or Hepes buffers by the addition of 100 m M NaCl causes a marked decrease in the spontaneous reactivation yield due to increased population of aggregation-prone intermediates of lysozyme. However, increasing the hydrophobicity of the refolding buffer by including ethylene glycol, thereby decreasing the hydrophobic interaction between aggrega- tion-prone intermediates, results in a slight increase of spontaneous refolding yield (Fig. 5). Depending on the conditions, trigger factor shows apparently opposite effects on lysozyme refolding: as a chaperone, the productive refolding is enhanced (Fig. 1); or as an antichaperone, the productive refolding is inhibited (Fig. 4). As a chaperone In redox phosphate buffer, trigger factor hinders the incorrect association of aggregation-prone species and thus favors the pathway to formation of native lysozyme, Fig. 5. Effects of trigger factor on lysozyme reactivation in redox Hepes buffer (d)orthesamebuffercontaining100m M NaCl (,)or5% ethylene glycol (r). The experiments were carried out as described in the legend to Fig. 4. 4520 G C. Huang et al.(Eur. J. Biochem. 269) Ó FEBS 2002 improving reactivation yield but without being a part of the final functional structure (Figs 1 and 2). In this case, the rates of assisted refolding of lysozyme are reduced by trigger factor compared to the rate of spontaneous refolding (Fig. 3). Typical chaperone behavior, exhibited by various members of the stress proteins, involves the interaction of the nonspecific peptide-binding site of the chaperone with a denatured protein in such a way as to inhibit aggregation. Trigger factor possesses a nonspecific peptide/protein-bind- ing site and a comparison with CyP as a reference foldase suggests that the high affinity toward unfolded protein chains is a requisite for the high efficiency of trigger factor in assisting protein folding [23]. In its efficient binding to unfolded proteins, trigger factor resembles a chaperone. Our previous work indicates that trigger factor shows chaperone activity for GAPDH and strong binding of GAPDH intermediates appears to decelerate their dissociation from trigger factor, thus resulting in a decrease in the rate constant of refolding [24]. Likewise, under conditions where trigger factor improves the recovery yield of native lysozyme, it also decreases the rate constant for the folding reaction. An increase in refolding yields and slowing down of refolding rates may be a common characteristic of molecular chaperones [24]. As an antichaperone In redox Hepes buffer, the productive refolding of lysozyme is substantially lower in the presence of trigger factor than in its absence (Fig. 4). At the same time, trigger factor accelerates the conversion of the denatured lysozyme into large, disulfide cross-linked aggregates (Fig. 6A,a and b). As a substantial proportion of the lysozyme would fold productively in the absence of trigger factor, trigger factor must intervene early in the folding process to redirect most of the substrate along an alternative nonproductive pathway to aggregation. Such behavior of trigger factor is reminis- cent of PDI, for which chaperone and antichaperone activities in lysozyme refolding have also been observed [7–11]. As trigger factor is integrated specifically into the Fig. 6. (A) Composition of trigger factor accelerated aggregates and (B) reducing (a) and nonreducing (b) SDS/PAGE (15%) analysis of the mixed aggregates. (A) The aggregates that formed during refolding of 10 l M lysozyme in the presence of 5 l M trigger factor were separated and then re-solubilized in increasing concentrations of urea with (a) or without (b) 150 m M dithiothreitol. Re-solubilized materials were analyzed on reducing SDS/PAGE (15%). Lanes labelled 0–7 represent the molar concentration of urea used. L and T indicate native lysozyme and native trigger factor, respectively, loaded in a molar ratio consistent with the reaction conditions. (B) Insoluble aggregates were formed in redox Hepes buffer at a lysozyme to trigger factor ratio that ensured maximal aggregation (10 l M lysozyme, 5 l M trigger factor) in the presence or absence of 5 l M BSA. The isolated aggregates were then incubated in SDS sample buffer with (a) or without (b) 200 m M 2-mercaptoethanol before analysis by SDS/ PAGE. 1, native trigger factor; 2, native BSA; 3, re-solubilized aggregates formed in the presence of BSA; 4, re-solubilized aggregates formed in the absence of BSA; 5, native lysozyme. Ó FEBS 2002 Chaperone and antichaperone activities of trigger factor (Eur. J. Biochem. 269) 4521 mixed aggregates (Fig. 6A), the binding of trigger factor with folding intermediates must be an essential step in the antichaperone activity. It is quite possible that antichaper- one activity is not determined by trigger factor’s third active site, but probably depends on the folding pathways of the substrate and the stability and relative populations of different intermediates, both of which could be dependent on the solution conditions, such as the ionic strength and redox state of the solution. As a chaperone and an antichaperone The spontaneous folding rate of lysozyme in the redox Hepes buffer is significantly slower than that in the redox phosphate buffer (Fig. 3). The intermediates recognized by trigger factor may differ in conformation in each case and thus differ in their ability to bind to trigger factor. It has been reported that trigger factor, in accord with its location at the ribosome in vivo, binds most strongly to early folding intermediates which lack compact structure [25,32]. Dena- tured lysozyme folds more slowly in Hepes buffer than in phosphate buffer and there may be more availability of loosely structured intermediates allowing tight binding to trigger factor. Meanwhile, the strong binding of folding intermediates also appears to decelerate their dissociation from trigger factor hence slowing the rate of folding. In Hepes buffer and at low concentrations of trigger factor where trigger factor behaves as an antichaperone, each trigger factor molecule attracts multiple lysozyme mole- cules to compete for the same peptide/protein-binding site, thus indirectly facilitating spatial contact between folding intermediates to form intermolecular disulfide cross-links. At the same time, decelerated dissociation resulting from strong binding of trigger factor provides folding interme- diates with enough time to weave a large, disulfide cross- linked insoluble network involving trigger factor as an integral component. The observed relative increase in folding rate in this region may reflect that it is only the fastest folding fraction of the population which escapes interaction with trigger factor and so can fold instead of aggregating. As the molecular ratio of trigger factor to lysozyme increases, contact between aggregation-prone intermediates of lysozyme is prevented resulting in suppression of aggregation and an up-turn in the amount of activity recovered (Fig. 4). The balance between apparent chaper- one and antichaperone functions of trigger factor in lysozyme refolding is controlled by the surrounding envi- ronment and the relative amount of trigger factor to lysozyme. An apparently similar switch from chaperone to antichaperone activity was observed in trigger factor assisted GAPDH refolding when the trigger factor concen- tration was very high. It seems therefore that the antichap- erone activity of trigger factor is actually the same as its chaperone activity, not a distinct function in addition to its isomerase and chaperone activities. It is a consequence of the ability of trigger factor to bind folding intermediates with non-native conformations, depends on the same peptide-binding site as the chaperone activity and is closely related to the folding properties of the substrate as controlled by the conditions. The antichaperone activity of trigger factor is not specific to lysozyme, as indicated by our findings with GAPDH [24] and an observation that trigger factor can also significantly decrease refolding yields of creatine kinase under conditions where full regain activity is obtained in spontaneous refolding (C. P. L and J. M. Z, unpublished results). ACKNOWLEDGEMENTS The present study was supported in part by the 973 Project of the Chinese Ministry of Science and Technology (G1999075608) and the exchange program between the Max-Plank Society and Chinese Academy of Sciences. The authors would like to thank Dr S. Perrett of this department for a critical reading of this paper and helpful suggestions. REFERENCES 1. Beissinger, M. & Buchner, J. (1998) How chaperones fold pro- teins. Biol. Chem. 379, 245–259. 2. Carrio, M.M., Corchero, J.L. & Villaverde, A. (1998) Dynamics of in vivo protein aggregation: building inclusion bodies in recombinant bacteria. FEMS Microbiol. Lett. 169, 9–15. 3. Arias, A.E., Velez-Granell, C.S., Torres-Ruiz, J.A. & Bendayan, M. (1994) Involvement of molecular chaperones in the aberrant aggregation of secretory proteins in pancreatic acinar cells. Exp. Cell Res. 215, 1–8. 4. Bova, M.P., Yaron, O., Huang, Q., Ding, L., Haley, D.A., Stewart, P.L. & Horwitz, J. (1999) Mutation R120G in alphaB- crystallin, which is linked to a desmin-related myopathy, results in an irregular structure and defective chaperone-like function. Proc. Natl Acad. Sci. USA 96, 6137–6142. 5. Kim, P.S., Bole, D. & Arvan, P. (1992) Transient aggregation of nascent thyroglobulin in the endoplasmic reticulum: relationship to the molecular chaperone, BiP. J. Cell Biol. 118, 541–549. 6. Cotner, T. & Pious, D. (1995) HLA-DR beta chains enter into an aggregated complex containing GRP-78/BiP prior to their deg- radation by the pre-Golgi degradative pathway. J. Biol. Chem. 270, 2379–2386. 7. Puig, A. & Gilbert, H.F. (1994) Protein disulfide isomerase exhibits chaperone and anti-chaperone activity in the oxidative refolding of lysozyme. J. Biol. Chem. 269, 7764–7771. 8. Puig, A., Lyles, M.M., Noiva, R. & Gilbert, H.F. (1994) The role of the thiol/disulfide centers and peptide binding site in the chaperone and anti-chaperone activities of protein disulfide isomerase. J. Biol. Chem. 269, 19128–19135. 9. Puig, A. & Gilbert, H.F. (1994) Anti-chaperone behavior of BiP during the protein disulfide isomerase-catalyzed refolding of reduced denatured lysozyme. J. Biol. Chem. 269, 25889–25896. 10. Song, J.L., Quan, H. & Wang, C.C. (1997) Dependence of the anti-chaperone activity of protein disulphide isomerase on its chaperone activity. Biochem. J. 328, 841–846. 11. Sideraki, V. & Gilbert, H.F. (2000) Mechanism of the anti- chaperone activity of protein disulfide isomerase: facilitated assembly of large, insoluble aggregates of denatured lysozyme and PDI. Biochemistry 39, 1180–1188. 12. Kelly, J.W. (1998) The environmental dependency of protein folding best explains prion and amyloid diseases. Proc. Natl Acad. Sci. USA 95, 930–932. 13. Crooke, E. & Wickner, W. (1987) Trigger factor: a soluble protein that folds pro-OmpA into a membrane-assembly-competent form. Proc. Natl Acad. Sci. USA 84, 5216–5220. 14. Crooke, E., Guthrie, B., Lecker, S., Lill, R. & Wickner, W. (1988) ProOmpA is stabilized for membrane translocation by either purified E. coli trigger factor or canine signal recognition particle. Cell 54, 1003–1011. 15. Guthrie, B. & Wickner, W. (1990) Trigger factor depletion or overproduction causes defective cell division but does not block protein export. J. Bacteriol. 172, 5555–5562. 4522 G C. Huang et al.(Eur. J. Biochem. 269) Ó FEBS 2002 16.Stoller,G.,Ru ¨ cknagel, K.P., Nierhaus, K.H., Schmid, F.X., Fischer, G. & Rahfeld, J.U. (1995) A ribosome-associated pepti- dyl-prolyl cis/trans isomerase identified as the trigger factor. EMBO J. 14, 4939–4948. 17. Hesterkamp, T., Hauser, S., Lu ¨ tcke, H. & Bukau, B. (1996) Escherichia coli trigger factor is a prolyl isomerase that associates with nascent polypeptide chains. Proc. Natl Acad. Sci. USA 93, 4437–4441. 18. Scholz, C., Stoller, G., Zarnt, T., Fischer, G. & Schmid, F.X. (1997) Cooperation of enzymatic and chaperone functions of trigger factor in the catalysis of protein folding. EMBO J. 16, 54–58. 19. Valent, Q.A., Kendall, D.A., High, S., Kusters, R., Oudega, B. & Luirink, J. (1995) Early events in preprotein recognition in E. coli: interaction of SRP and trigger factor with nascent polypeptides. EMBO J. 14, 5494–5505. 20. Kandror, O., Sherman, M., Rhode, M. & Goldberg, A.L. (1995) Trigger factor is involved in GroEL-dependent protein degrada- tion in Escherichia coli and promotes binding of GroEL to unfolded proteins. EMBO J. 14, 6021–6027. 21. Kandror, O., Sherman, M., Moerschell, R. & Goldberg, A.L. (1997) Trigger factor associates with GroEL in vivo and promotes its binding to certain polypeptides. J. Biol. Chem. 272, 1730–1734. 22. Go ¨ thel, S.F., Scholz, C., Schmid, F.X. & Marahiel, M.A. (1998) Cyclophilin and trigger factor from Bacillus subtilis catalyze in vitro protein folding and are necessary for viability under star- vation conditions. Biochemistry 37, 13392–13399. 23. Lyon, W.R., Gibson, C.M. & Caparon, M.G. (1998) A role for trigger factor and an rgg-like regulator in the transcription, secretion and processing of the cysteine proteinase of Strepto- coccus pyogenes. EMBO J. 17, 6263–6275. 24. Huang, G.C., Li, Z.Y., Zhou, J.M. & Fischer, G. (2000) Assisted folding of D -glyceraldehyde-3-phosphate dehydrogenase by trig- ger factor. Protein Sci. 9, 1254–1261. 25. Huang, G.C., Li, Z.Y. & Zhou, J.M. (2000) Conformational specificity of trigger factor for the folding intermediates of alpha- lactalbumin. Biochim. Biophys. Acta 1480, 77–82. 26. Schmid, F.X., Mayr, L.M., Mu ¨ cke, M. & Scho ¨ nbrunner, E.R. (1993) Prolyl isomerases: role in protein folding. Adv. Protein Chem. 44, 25–66. 27. Gill, S.C. & von Hippel, P.H. (1989) Calculation of protein extinction coefficients from amino acid sequence data. Anal Bio- chem. 182, 319–326. 28. Kofron, J.L., Kuzmie ` ,P.,Kishore,V.,colo ´ n-Bonilla, E. & Rich, D.H. (1991) Determination of kinetic constants for peptidyl prolyl cis-trans isomerases by an improved spectrophotometric assay. Biochemistry 30, 6127–6134. 29.Fischer,G.,Bang,H.&Mech,C.(1984)Determinationof enzymatic catalysis for the cis-trans-isomerization of peptide binding in proline-containing peptides. Biomed.Biochim.Acta43, 1101–1111. 30. Hwang, C., Sinskey, A.J. & Lodish, H.F. (1992) Oxidized redox state of glutathione in the endoplasmic reticulum. Science 257, 1496–1502. 31. Goldberg, M.E., Rudolph, R. & Jaenicke, R. (1991) A kinetic study of the competition between renaturation and aggregation during the refolding of denatured-reduced egg white lysozyme. Biochemistry 30, 2790–2797. 32. Scholz,C.,Mcke,M.,Rape,M.,Pecht,A.,Pahl,A.,Bang,H.& Schmid, F.X. (1998) Recognition of protein substrates by the prolyl isomerase trigger factor is independent of proline residues. J. Mol. Biol. 277, 723–732. Ó FEBS 2002 Chaperone and antichaperone activities of trigger factor (Eur. J. Biochem. 269) 4523 . of trigger factor, as was found earlier for trigger factor assisted refolding of GAPDH [24], increase with increasing concentrations of trigger factor. The t 1/2 of Fig. 2. Effect of trigger factor. presence of CyP. Ó FEBS 2002 Chaperone and antichaperone activities of trigger factor (Eur. J. Biochem. 269) 4519 effect of the solution conditions on folding of the substrate itself. Composition of trigger. dissociation from trigger factor hence slowing the rate of folding. In Hepes buffer and at low concentrations of trigger factor where trigger factor behaves as an antichaperone, each trigger factor molecule