Báo cáo khoa học: The loss of tryptophan 194 in antichymotrypsin lowers the kinetic barrier to misfolding

The loss of tryptophan 194 in antichymotrypsin lowers the kinetic barrier to misfolding Mary C. Pearce, Lisa D. Cabrita, Andrew M. Ellisdon and Stephen P. Bottomley

Department of Biochemistry and Molecular Biology, Monash University, Clayton, Australia

Keywords antichymotrypsin; protein aggregation; protein misfolding; serpin; stability

Correspondence S. P. Bottomley, Department of Biochemistry and Molecular Biology, PO Box 13D, Monash University, 3800, Australia Fax: +61 3 99054699 Tel: +61 3 99053703 E-mail: steve.bottomley@med.monash.edu.au

(Received 21 November 2006, revised 17 May 2007, accepted 23 May 2007)

doi:10.1111/j.1742-4658.2007.05897.x

Antichymotrypsin, a member of the serpin superfamily, has been shown to form inactive polymers in vivo, leading to chronic obstructive pulmonary disease. At present, however, the molecular determinants underlying the polymerization transition are unclear. Within a serpin, the breach position is implicated in conformational change, as it is the first point of contact for the reactive center loop and the body of the molecule. W194, situated within the breach, represents one of the most highly conserved residues within the serpin architecture. Using a range of equilibrium and kinetic experiments, the contribution of W194 to proteinase inhibition, stability and polymerization was studied for antichymotrypsin. Replacement of W194 with phenylalanine resulted in a fully active inhibitor that was desta- bilized relative to the wild-type protein. The aggregation kinetics were sig- nificantly altered; wild-type antichymotrypsin exhibits a lag phase followed by chain elongation. The loss of W194 almost entirely removed the lag phase and accelerated the elongation phase. On the basis of our data, we propose that one of the main roles of W194 in antichymotrypsin is in pre- venting polymerization.

change conformational

The serpins (serine proteinase inhibitors) are a large family of metastable proteins with a highly conserved tertiary structure. Metastability of the serpin native state facilitates the rapid structural changes required for efficient inhibition of a range of serine and cysteine proteinases [1]. During inhibition, the exposed flexible reactive center loop (RCL), which acts as bait for the proteinase (Fig. 1), becomes incorporated into the A b-sheet following cleavage by a proteinase. The A b-sheet is composed of five strands, which are referred to as s1A, s2A, etc. This structural rearrangement results in translocation of the proteinase from one pole of the molecule to the other, deforming the covalently bound proteinase against the base of the serpin and thus preventing efficient hydrolysis of the ester bond linking the proteinase and inhibitor [2–4].

Abbreviations ACT, antichymotrypsin; a1-AT, antitrypsin; bis-ANS, bis-8-anilinonaphthalene-1-sulfonate (4,4¢-dianilino-1,1¢-binaphthyl-5,5¢-disulfonic acid dipotassium salt); muACT, murine homolog of antichymotrypsin; PAI-1, plasminogen activator inhibitor-1; RCL, reactive center loop; SI, stoichiometry of inhibition; TEM, transmission electron microscopy; TUG, transverse urea gradient.

FEBS Journal 274 (2007) 3622–3632 ª 2007 The Authors Journal compilation ª 2007 FEBS

3622

that is considerably more thermodynamically stable than its initial native state [5]. This energetically fav- orable can, however, also occur in the absence of a proteinase to form a vari- ety of misfolded states [6]. In some serpins, the RCL becomes completely incorporated into the central b-sheet while remaining intact, with concomitant dis- ruption of the C b-sheet, which results in the stable, inactive latent state [7]. Several serpins have also been shown to form inactive polymers when the RCL of one serpin inserts into a b-sheet of another [8,9]. This can lead to accumulation of the aggrega- ted serpin both within the cells producing the protein and in the circulation. Polymerization of the plasma serpins, antitrypsin (a1-AT), antichymotrypsin (ACT) and antithrombin III, can be promoted by single point mutations and lead to many human diseases [10]. Insertion of the RCL into the body of the ser- pin, during proteinase inhibition, creates a molecule

M. C. Pearce et al.

W194 protects against antichymotrypsin aggregation

Results

The serpin architecture can be split into regions that either deform or remain intact during proteinase inhi- bition and misfolding [11]. One such critical region that undergoes deformation during inhibition is the ‘breach’, an area that encompasses the top of the A b-sheet and the N-terminus of the RCL (Fig. 1). The breach region is the first point of insertion for the RCL during inhibition, and therefore has evolved to readily accommodate this conformational change.

The ‘breach’ has been identified as a region possessing high structural flexibility in serpins, and many of the residues occupying this area are strongly conserved [12,16]. In addition, several disease states have been linked to serpins that contain mutations within this region [10]. In this study, we examined the role of the highly conserved breach residue, W194, in maintaining the structural and functional properties of ACT. ACTW194F was generated using site-directed mutagen- esis to replace the intrinsic tryptophan at this location (wt)ACT and with phenylalanine. Both wild-type ACTW194F were expressed and purified as previously described [17]. Gel filtration chromatography was used to monitor the aggregation status of the protein after storage and before each experiment. In all the experi- ments described below, only active, monomeric protein was used. We attempted expression of ACT containing different mutations at position 194, such as Gly, Ile, Lys, Asp and Asn; in all cases, we were unable to pur- ify any monomeric protein.

Characterization of the inhibitory properties of ACTW194F

Fig. 1. Schematic representation of murine ACT homolog, Serpina3n. Ribbon diagram of native ACT, Protein Data Bank 1YXA [1], with W194 (black) and surrounding residues identified (gray). The inset is a close-up of the region around W194. A b-sheet is high- lighted in gray.

FEBS Journal 274 (2007) 3622–3632 ª 2007 The Authors Journal compilation ª 2007 FEBS

3623

Located at the top of the breach, on a loop connect- ing s3A and s3C, is the residue W194, which is present in over 95% of serpin sequences [12]. The structural and functional roles of this residue have been studied in various serpins, and the results indicate that repla- cing it with a phenylalanine has variable effects on the overall inhibitory or thermodynamic properties of the protein [13–15]. In plasminogen activator inhibitor-1 (PAI-1), the replacement of this highly conserved tryp- tophan residue reduced the propensity of the protein to adopt the latent state. This phenomenon has not been observed for any other member of the family. Taken together, these studies do not suggest any rea- son for the high level of conservation of tryptophan at this position within the family. Therefore, we reasoned that it might play a role in serpin polymerization, an area that has not been previously investigated for pro- teins carrying this mutation. We have mutated W194 to phenylalanine in ACT, and monitored the effect of this mutation upon inhibition, stability and polymer- ization. Our data indicate that W194 plays a critical role in maintaining the stability of the breach region and preventing polymerization of ACT. The stoichiometry of inhibition (SI) values for inhibi- tion of chymotrypsin by wtACT and ACTW194F were determined by titration of the proteinase with known concentrations of the serpin. The SI for ACTW194F was 1.4, slightly higher than that observed for wtACT for (Table 1). The association rate constant (kass)

M. C. Pearce et al.

W194 protects against antichymotrypsin aggregation

Table 1. Inhibitory and stability properties of wtACT and ACTW194F. SI, association rate constant (kass) and apparent association rate constant (kapp) were determined as described in Experimental pro- cedures. The rates of unfolding (kN fi I and kI fi U) in 4 M guanidine hydrochloride were determined as described in Experimental proce- dures. Each value represents the average of five separate experi- ments; actual value and standard error are listed.

Protein

wtACT

ACTW194F

)1Æs)1) M )1Æs)1)

1.1 ± 0.02 8.70 ± 0.58 9.59 2.2

14.45 ± 0.2

3.6 ± 0.008

1.4 ± 0.03 6.39 ± 0.35 9.22 0.8 ND 3.8 ± 0.005

SI kapp (· 105 kass (· 105 M TUG transition (M urea) kN fi I (s)1) kI fi U (s)1)

Fig. 2. SDS-PAGE analysis of complex formation between chymo- trypsin and ACT. ACT and ACTW194F were mixed with chymotrypsin in a 2 : 1 molar ratio and allowed to incubate for 30 min at 37 (cid:2)C, following which the complexes were analyzed by 10% (v ⁄ v) SDS- PAGE. Arrows indicate the position of native ACT, cleaved ACT released by the proteinase, and ACT still in complex with chymo- trypsin.

Fig. 3. Spectral properties of wtACT and ACTW194F. In each figure, wtACT is represented by squares, and ACTW194F is represented by circles. (A) Far-UV CD spectrum of native (empty symbols) and guanidine hydrochloride-unfolded (filled symbols) ACT proteins. (B) Intrinsic tryptophan fluorescence emission spectra for native (empty symbols) and unfolded (filled symbols) ACT proteins. (C) Bis-ANS fluorescence emission spectra for wtACT and ACTW194F in the native conformation. Protein was present at 500 nM for each experiment, and each spectrum represents the average of five scans.

inhibition of chymotrypsin by ACTW194F was similar to that of wtACT (9.22 · 105 m)1Æs)1 and 9.59 · 105 m)1Æs)1, respectively; Table 1). The ability of ACTW194F to form SDS-stable complexes with chymotrypsin was also examined by incubating 4 lm serpin with 2 lm chymotrypsin for 30 min at 37 (cid:2)C. The resulting mixture was then analyzed by SDS ⁄ PAGE under reducing conditions (Fig. 2). Both wtACT and ACTW194F formed SDS-stable complexes with chymotrypsin, as indicated by the presence of a larger species (Fig. 2, labeled complex). In addition, a small amount of cleaved serpin was present (Fig. 2, labeled cleaved), which is consistent with the SI data. Taken together, these data suggest that the replacement of W194 with phenylalanine had no significant effect upon the inhibitory properties of the protein. structural differences may

Characterization of the spectral properties of ACTW194F

FEBS Journal 274 (2007) 3622–3632 ª 2007 The Authors Journal compilation ª 2007 FEBS

3624

determine what exist between them (Fig. 3). Far-UV CD spectra were obtained for both proteins in the folded and unfolded (achieved by incubation in 6 m guanidine states hydrochloride) at 25 (cid:2)C and pH 7.8. The native indicating that both proteins spectra were identical, Both fluorescence and far-UV CD spectra were in order to obtained for wtACT and ACTW194F,

M. C. Pearce et al.

W194 protects against antichymotrypsin aggregation

The role of W194 in the stability of ACTW194F

bands representing possess a similar fold and that W194 does not con- tribute to the far-UV spectrum of ACT (Fig. 3A). The unfolded spectra for both ACT proteins were also similar, implying that the mutation had no effect the unfolded state of ACT on the structure of (Fig. 3A). These data indicate that the replacement of W194 with phenylalanine did not alter the secondary structure of the variant.

The stability of wtACT and its variant was determined by transverse urea gradient (TUG) PAGE analysis (Fig. 4). Both proteins demonstrated the characteristic unfolding profile seen for other serpins; in addition, some polymeric ACT were observed in higher denaturant concentrations. The unfolding transition for ACTW194F occurred at a signi- ficantly lower urea concentration (0.8 m) than that of wtACT (2.2 m) (Table 1). Owing to the formation of polymers in high urea concentrations, no spectroscopic equilibrium studies were performed.

The unfolding kinetics of both wtACT and ACTW194F were determined using stopped-flow fluores- cence. In order to directly compare the two proteins, which have different intrinsic fluorescence properties, we followed the unfolding reaction using the change in bis-ANS fluorescence. wtACT unfolded with a biphasic transition that corresponded to the presence of a fold- ing intermediate, which is consistent with previous findings on this and other serpins [19]. bis-ANS fluor- escence intensity increased during adoption of the intermediate ensemble, and this was followed by a decrease in fluorescence intensity as the rest of the

ACT has three intrinsic tryptophan residues, one of which was mutated to phenylalanine in ACTW194F. Fluorescence emission spectra were obtained for both wtACT and ACTW194F in their folded and unfolded states (Fig. 3B). In both cases, the fluorescence spectra represent the average of all the tryptophan residues in the protein. There was no difference in emission max- ima (kmax) between wtACT and ACTW194F in the native state (kmax ¼ 335 nm for both). There was, however, a significant difference in the emission inten- sity of both proteins. The emission intensity at the spectral peak for ACTW194F was 75% that of the wtACT value, indicating that in the native state W194 contributes 25% to the total tryptophan fluorescence signal of the wild-type form (Fig. 3B). Both proteins were unfolded in 5 m guanidine thiocyanate, as this denaturant is strong enough to completely unfold all structure within ACT [17]. There was no significant difference in the kmax for either unfolded protein. The fluorescence emission intensity of unfolded ACTW194F was, however, two-thirds the intensity of wtACT, which could be attributed to the loss of one of the three intrinsic tryptophan residues through the intro- duction of the mutation. Both unfolded proteins showed a dramatic decrease in fluorescence signal in comparison to the native protein. This is most likely attributable to the nature of the interaction between the protein and guanidine thiocyanate, as this denatu- rant readily quenches tryptophan fluorescence in solu- tion.

the relative

Fig. 4. TUG PAGE analysis of ACT proteins. Both wtACT and ACTW194F were applied to a 0–5 M urea gradient gel, increasing from left to right. (A) wtACT and (B) ACTW194F formed polymers (P) in high denaturant concentrations. The final protein concentration was 10 lM, and each figure shows the best representative of three gels run for each protein.

FEBS Journal 274 (2007) 3622–3632 ª 2007 The Authors Journal compilation ª 2007 FEBS

3625

to larger The fluorescence emission spectrum of bis-8-anili- (4,4¢-dianilino-1,1¢-binaph- nonaphthalene-1-sulfonate thyl-5,5¢-disulfonic acid dipotassium salt) (bis-ANS) when bound to proteins can provide information exposure of hydrophobic regarding pockets ⁄ surfaces [18]. Figure 3C shows the emission spectra of bis-ANS in the presence of wtACT and ACTW194F. The emission intensity of bis-ANS in the presence of ACTW194F was two times greater than that of wtACT. This could be due to either a small conformational change within ACTW194F that altered the binding affinity of bis-ANS for ACT, for or a more accessible hydrophobic core, due, example, in conformational fluctuations ACTW194F.

M. C. Pearce et al.

W194 protects against antichymotrypsin aggregation

Fig. 5. Stopped-flow unfolding kinetics of ACT proteins. wtACT (squares) and ACT W194F (circles) were unfolded in 4 M guanidine hydrochloride, at a final protein concentration of 1 lM in the pres- ence of bis-ANS. The change in bis-ANS fluorescence was monit- ored over time. The data were analyzed using either single or double exponential equations, using the software provided by the manufacturer. Each dataset represents the average of ten separate traces.

and 3.6 s)1, 14.5 s)1

structure was lost (Fig. 5). The rates of unfolding, in 4 m guanidine hydrochloride, corresponding to the N to I (kN fi I) and I to U (kI fi U) transitions for respectively wtACT, were (Table 1). When ACTW194F was unfolded under the same conditions, only a single exponential decay was observed (Fig. 5). The rate of decay was 3.8 s)1, sim- ilar to that observed for the second phase of wtACT unfolding. Taken together, these data indicate that the loss of W194 destabilizes ACT and increases the rate at which the protein unfolds.

Fig. 6. Polymerization characteristics of wtACT and ACTW194F. Both wtACT (squares) and ACTW194F (circles) were heated at 48 (cid:2)C, and changes in various parameters were monitored. (A) Light scatter. (B) Bis-ANS fluorescence changes. (C) Loss of inhibitory activity towards chymotrypsin. Light scatter data were fitted to an expo- nential equation that includes a term describing the initial plateau in the data. Bis-ANS and activity loss data were fitted to a single exponential equation. Protein was present in each experiment at 1 lM, and each figure represents the average of five separate experiments.

Characterization of the polymerization behavior of ACTW194F

The mechanism of ACT polymerization has been char- acterized previously in two studies [20,21]. A key fea- ture is the presence of a lag phase that can be abolished by addition of preformed polymers, a situ- ation that does not occur for other serpins studied so far. ACT polymerization was initiated by incubating the protein at 48 (cid:2)C and monitoring the changes in light scatter, bis-ANS fluorescence and activity loss (Fig. 6). This temperature was chosen because experi- ments performed at 45 (cid:2)C and 50 (cid:2)C were either too fast or too slow to allow monitoring of the entire pro- cess for both the wild-type and mutant protein (data not shown).

FEBS Journal 274 (2007) 3622–3632 ª 2007 The Authors Journal compilation ª 2007 FEBS

3626

at 48 (cid:2)C. Both proteins demonstrated a lag phase prior to observation of any signal change. Although both proteins showed a similar initial signal, the lag phase for ACTW194F (1.2 min) was approximately five-fold shorter than that observed for wtACT (7.2 min). The aggregation phase was also faster for ACTW194F, as the reaction was observed to be complete in less than 10 min, whereas it was still continuing after 60 min for wtACT (Fig. 6A). The rate of ACTW194F polymer- 8.71 ± 0.42 · 10)3 s)1, approximately ization was than that observed for wtACT 14 times faster Figure 6A shows the changes in light scatter at 405 nm as both the wtACT and ACTW194F aggregated

M. C. Pearce et al.

W194 protects against antichymotrypsin aggregation

Scheme 1. Kinetic analysis of wtACT and ACTW194F aggregation. N refers to the native state, M* to the polymerogenic intermediate state, and P to the polymer. These data were collected through a combination of techniques that included light scatter, bis-ANS fluorescence and activity loss. Each value represents the average of five separate experiments; actual value and standard error are listed.

Fig. 7. Transmission electron micrographs of ACT polymers. Poly- mers were formed by heating ACT proteins at 48 (cid:2)C for 30 min at pH 9.0. Samples were stained negatively with 1% (w ⁄ v) uranyl acetate, and viewed with a magnification of · 100 000. The scale bar represents 200 nm. (A) ACT polymers. (B) ACTW194F polymers.

(0.60 ± 0.007 · 10)3 s)1; Scheme 1). These experi- ments were performed over a 10-fold concentration range, and the results were consistent with those of previous studies, where both phases were found to be concentration dependent [20,21].

approximately 20 times to 13.9 ± 0.8 · 10)3 s)1 for ACTW194F (Scheme 1). These experiments were also performed at a range of protein concentrations, and no concentration-dependent change could be observed for either protein (data not shown).

The morphology of the endpoint polymers formed by ACT and ACTW194F were analyzed using transmis- sion electron microscopy (TEM; Fig. 7). No significant structural difference could be determined when the polymers formed by either ACT or ACTW194F were compared, with both proteins adopting the characteris- tic linear filaments seen in previous TEM studies of serpins [23,25]. The width of each polymer chain, formed by either protein, ranged from 8 to 10 nm, and the chains had a strong tendency to clump together to form larger aggregates.

The role of W194 in a1-AT time as

FEBS Journal 274 (2007) 3622–3632 ª 2007 The Authors Journal compilation ª 2007 FEBS

3627

rate of activity loss Monitoring serpin polymerization through changes in bis-ANS fluorescence intensity has been previously used to observe formation of the intermediate species, M*, which displays greatly enhanced fluorescence [22,23]. Here, both ACT proteins demonstrated an increase in bis-ANS fluorescence upon heating at 48 (cid:2)C (Fig. 6B). There was no detectable lag phase present, indicating that the conformational change that bis-ANS is indica- ting occurs during the lag phase of the reaction. These experiments were performed at a range of protein con- centrations, and the rate of M* formation was found to be independent of protein concentration, as previously described (data not shown) [20]. Figure 6B highlights the dramatic difference between the rates of M* forma- tion between ACT and ACTW194F. Introducing the mutation at position 194 accelerated the rate of M* formation approximately 35-fold, with rates of 2.04 ± 0.002 · 10)3 s)1 for wtACT and 70 ± 2 · 10)3 s)1 for ACTW194F (Scheme 1). The rate of polymerization can also be monitored through changes in bis-ANS fluores- cence; however, ACTW194F formed polymers too quickly and readily precipitated, precluding further experiments. Another means of monitoring serpin polymerization involves observing the loss of inhibitory activity as the sample is heated over time. As serpins are incorporated into the growing polymer, their inhibitory activity is abolished [24]. Both wtACT and ACTW194F lost activ- ity over incubated at 48 (cid:2)C they were (Fig. 6C). Whereas all inhibitory activity was elimin- ated within 10 min for ACTW194F, there still appeared to be some active wtACT remaining after 60 min (Fig. 6C). The for wtACT was 0.69 ± 0.15 · 10)3 s)1, which was accelerated The data described above clearly demonstrate that the presence of W194 plays a fundamental role in prevent- ing the misfolding of ACT. This raises the following question: does W194 play this role in other serpins? To address this, we expressed and purified a1-ATW194F,

M. C. Pearce et al.

W194 protects against antichymotrypsin aggregation

A

In addition,

which is a fully active inhibitor, and unlike its ACT counterpart, has the same stability as the wild-type protein [15]. We analyzed the ability of a1-ATW194F to bind bis-ANS, and found that there was no fluores- cence emission enhancement as compared to a1-AT (Fig. 8A). the rate of unfolding of a1-ATW194F in 4 m guanidine hydrochloride was the for a1-AT (a1-AT ¼ 9.0 ± 0.16 s)1 and same as a1-ATW194F ¼ 9.4 ± 0.12 s)1; Fig. 8B). The rate of aggregation was also unchanged upon mutation of W194 (Fig. 8C and Scheme 1). Cumulatively, these data indicate that W194 does not perform the same role in a1-AT as it does in ACT.

Discussion

B

Serpins are marginally stable proteins that respond to certain stimuli by undergoing conformational change to a more stable conformation [1]. This can be produc- tive as in protease inhibition, or nonproductive as in misfolding. The serpin structure can maintain this bal- ance through the presence of a vast range of highly conserved residues that are critical in regulating the many complicated changes that must occur for these proteins to perform their key role, that of inhibiting proteinases.

C

functional,

Fig. 8. Effects of W194 replacement within a1-AT. The effects of replacing W194 with a phenylalanine in a1-AT was analyzed by (A) bis-ANS binding, (B) rate of unfolding in 4 M guanidine hydrochlo- ride, and (C) rate of aggregation monitored by light scatter. a1-AT (squares) and a1-ATW194F (diamonds) were present at the same concentration (1 lM) for all the experiments.

During folding, misfolding and function, the breach region is proposed to be critical in maintaining this balance [11]. The exact role of this region is, however, poorly understood. Many highly conserved residues reside in the breach, one of these being W194. The area surrounding this residue contains many large side chains; however, few form hydrogen bonds with W194. This residue has been mutated to phenylalan- ine in a range of serpins, with varied results. In a1- AT, this substitution has no detectable effect on the protein, with all stability and folding parameters being maintained [15]. One serpin that does result in a difference is PAI-1, a protein that readily adopts the latent state. The loss of W194 slows down this transition, indicating that the con- tacts made within the breach in PAI-1 maintain a slightly destabilized state that readily allows for inser- tion of the RCL [13].

FEBS Journal 274 (2007) 3622–3632 ª 2007 The Authors Journal compilation ª 2007 FEBS

3628

study, where tion experiments performed in this removal of this residue resulted in formation of the polymerogenic intermediate state, M*, occurring far When W194 is mutated in ACT, the protein is fully active; however, the native state is destabilized and more susceptible to misfolding, yet the misfolded struc- tures adopted are identical to those formed by the wild-type protein. Therefore, the contacts formed by W194 in ACT are critical in maintaining the structure of the breach in order to reduce the chances of aggre- gation should the environment become more hostile for the protein. This is highlighted in the polymeriza-

M. C. Pearce et al.

W194 protects against antichymotrypsin aggregation

more rapidly. With the equilibrium shifted towards the intermediate state, the concentration-dependent phase of aggregation is also accelerated, as evidenced by the light scatter and activity loss data. No appreciable change in the structure of the native state was detected that could explain a conformational reason behind this; however, a detailed comparison of the murine homolog of ACT (muACT) with other serpins reveals some interesting points that may indicate why this occurs.

Fig. 9. Overlay of the native crystal structures of muACT, a1-AT and PAI-1. The conserved tryptophan at position 194 is highlighted in blue, and the RCL is highlighted in red. Insets show close-up of the breach region for each protein. Those residues that form hydrogen bonds within the area are in cyan. Protein Data Bank identifiers are as follows: 1YXA for murine ACT (muACT) [1], 1QLP for a1-AT [2], and 1B3K for PAI-1 [3].

FEBS Journal 274 (2007) 3622–3632 ª 2007 The Authors Journal compilation ª 2007 FEBS

3629

A structural comparison was performed on the crys- tal structures of human a1-AT, PAI-1 and muACT (all in the native state; Fig. 9). Although there a human ACT crystal structure has been published [26], it does not represent the best structure for comparison, for several reasons: first, the dataset is not in the Protein Data Bank, which makes a detailed comparison diffi- cult; second, the molecule contains the RCL of a1-AT; and third, the refinement of this dataset was incom- plete, which means that errors could be present in the interpretation of the data, and that model bias is a possibility. The most significant differences between the native structures of the three proteins presented in Fig. 9 are within the breach region. The RCL of muACT is partially inserted into the central b-sheet, with residues 366–368 forming hydrogen bonds within

M. C. Pearce et al.

W194 protects against antichymotrypsin aggregation

the family; however, it is crucial to the breach of this partially inserted, native structure.

Experimental procedures

Site-directed mutagenesis and expression

Mutagenesis of ACT was performed using the Stratagene (La Jolla, CA, USA) site-directed mutagenesis kit, and veri- fied using DNA sequencing. Production of the mutant form of ACT studied here, ACTW194F, involved replacing the intrinsic tryptophan at position 194 with a phenylalanine residue. Both wtACT and ACTW194F were purified as previ- ously described [17]. a1-AT and a1-ATW194F was expressed and purified as previously described [15].

this sheet [27]. In both PAI-1 and a1-AT, the corres- ponding region of the RCL is exposed from the sheet and is part of the loop as the bulk of the RCL winds over the body of the protein [28,29]. The backbone of the W194 residue forms a hydrogen bond within this portion of the RCL, which does not occur for the other proteins. The tryptophan side chain occupies the same plane in each protein; however, there is a slight shift in the backbone, which results in slight differences in hydrogen bonding pattern between this loop and s2B of the B b-sheet. The residues that pack against the tryptophan residue in each protein appear to be conserved.

Determination of the inhibitory and stability characteristics this region of

Inhibitory activity assays including SI and calculation of the association rate constant, kass, were determined for the interaction between ACT and the enzyme chymotrypsin, in order to determine whether the mutation had altered the interaction between the two proteins. Each serpin was mixed with chymotrypsin at a variety of concentration ratios, and the SI and kass were calculated as previously described [31]. TUG PAGE was performed in order to esti- mate the stability of each protein used in the study. The gels and solutions used were as previously described by Goldenberg [33], although the final percentage of acryla- mide used was 7.5% (v ⁄ v), and the urea concentration was 5 m [32,33]. The midpoint of each unfolding trace was determined by visual observation of each gel. Each experi- ment was performed four times.

From this comparison, it can be seen that the key difference between these proteins lies in the partial insertion of the RCL into the central A b-sheet, which forms hydrogen bonds with the W194 back- bone. These bonds are most likely critical in main- taining the stability of the protein. Mutation of this residue in ACT may alter the bond- ing pattern within this area, which in turn would have dramatic effects on the stability of the breach region. In a1-AT and PAI-1, this area is not disrup- ted by the presence of a partially inserted RCL, which in turn would make the area more robust to the presence of mutations. Differences in the hydro- gen bonding pattern of PAI-1 in comparison to that of a1-AT may also be responsible for the effect seen on latency in this protein.

Spectral studies

Fluorescence spectra were obtained on a Perkin-Elmer (Waltham, MA, USA) LS-50B Luminescence Spectropho- tometer at 25 (cid:2)C. Samples were prepared at a protein con- centration of 500 nm in 50 mm Tris ⁄ HCl (pH 7.4), either alone or with 5 m guanidine thiocyanate. For bis-ANS fluor- escence spectra, samples were prepared in 50 mm Tris ⁄ HCl (pH 7.4), at a protein concentration of 500 nm, along with bis-ANS at a five-fold molar excess over the protein present. Spectra were obtained with slit widths of 3 nm, each spec- trum representing the average of five scans between 300 and 400 nm, with excitation wavelengths of 280 nm for trypto- phan fluorescence and 390 nm for bis-ANS fluorescence, all at 25 (cid:2)C. CD spectra were obtained on a Jasco (Nippon, Bunkon Co., Japan) 810 spectropolarimeter at 25 (cid:2)C. Each protein was present at a concentration of 4 lm in 50 mm Tris ⁄ HCl (pH 7.4), except for those unfolded samples that also contained 4 m guanidine hydrochloride.

in Another well-known breach mutation results expression of Z a1-AT (Glu342 to Lys), which behaves in a similar manner to ACTW194F, as it is fully active, yet susceptible to misfolding and aggregation [30]. This mutation results in the loss of a critical salt bridge that anchors the top of strand s5A with a strand above the sheet. Loss of this interaction demonstrates the vulner- ability of the breach to detrimental mutations, which do not largely affect native fold or function, while still playing a major role in aggregation and subsequent disease.

FEBS Journal 274 (2007) 3622–3632 ª 2007 The Authors Journal compilation ª 2007 FEBS

3630

Together, these findings highlight a significant role for the breach region in maintaining overall native sta- bility, rather than in regulating inhibitory function in ACT. Single point mutations such as those seen here can destabilize the area without causing large, detect- able structural changes, resulting in a native-like relatively mild protein that will polymerize under conditions. Previous studies investigating the role of W194 in other serpins have reported little or no effect on the inhibitory capacity or stability of the serpin [13,15]. It is most likely that the variability of the breach region between different serpins means that W194 does not play a global role in all members of

M. C. Pearce et al.

W194 protects against antichymotrypsin aggregation

adsorbed onto carbon-coated grids and stained with 1% (w ⁄ v) uranyl acetate. A magnification of · 100 000 was used.

Unfolding kinetics

Acknowledgements

such that upon mixing

the

This work was supported by grants from the National Health and Medical Research Council and the Austra- lian Research Council to S. P. Bottomley. S. P. Bot- tomley is a Senior Research Fellow of the NHMRC. L. D. Cabrita is a CJ Martin Fellow of the NHMRC.

References

Protein (10 lm) was rapidly mixed at a 10 : 1 ratio with denaturant, so that the final concentrations of protein and guanidine hydrochloride were 1 lm and 4 m, respectively. Both protein and denaturant solutions also contained 5 lm bis-ANS, concentration remained constant. Samples were mixed and read in an Applied Photophysics (Leatherhead, UK) SV18 stopped- flow fluorimeter at 22 (cid:2)C. An excitation wavelength of 390 nm was used, with slit widths of 2 mm used. Fluores- cence was monitored using a cut-off filter removing all sig- nal < 455 nm. Data were fitted to a double or triple exponential equation, using the manufacturer’s software.

1 Cabrita LD & Bottomley SP (2004) How do proteins avoid becoming too stable? Biophysical studies into metastable proteins. Eur Biophys J 33, 83–88.

2 Huntington JA, Read RJ & Carrell RW (2000) Struc- ture of a serpin–protease complex shows inhibition by deformation. Nature 407, 923–926.

3 Stratikos E & Gettins PG (1999) Formation of the

covalent serpin–proteinase complex involves translocation of the proteinase by more than 70 A and full insertion of the reactive center loop into beta-sheet A. Proc Natl Acad Sci USA 96, 4808–4813.

4 Tew DJ & Bottomley SP (2001) Intrinsic fluorescence changes and rapid kinetics of proteinase deformation during serpin inhibition. FEBS Lett 494, 30–33.

5 Bruch M, Weiss V & Engel J (1988) Plasma serine pro- teinase inhibitors (serpins) exhibit major conformational changes and a large increase in conformational stability upon cleavage at their reactive sites. J Biol Chem 263, 16626–16630.

6 Chow MK, Lomas DA & Bottomley SP (2004) Promis- cuous beta-strand interactions and the conformational diseases. Curr Med Chem 11, 491–499.

7 Carrell RW, Stein PE, Fermi G & Wardell MR (1994) Biological implications of a 3 A structure of dimeric antithrombin. Structure 2, 257–270.

8 Dunstone MA, Dai W, Whisstock JC, Rossjohn J, Pike RN, Feil SC, Le Bonniec BF, Parker MW & Bottomley SP (2000) Cleaved antitrypsin polymers at atomic reso- lution. Protein Sci 9, 417–420.

For each polymerization assay, the buffer comprised 50 mm Tris ⁄ HCl (pH 7.4) and 50 mm NaCl, preheated to 48 (cid:2)C. Bis-ANS fluorescence assays also contained bis-ANS dye at a five-fold excess over the protein concentration present in the sample. Each run was followed over time using the time-drive function on a Perkin-Elmer LS-50B Lumines- cence Spectrophotometer, with slit widths of 2.5 nm, using an excitation wavelength of 390 nm, and monitoring the change in fluorescence at an emission wavelength of 480 nm. Runs were terminated once the signal dropped severely, which indicated protein precipitation. Activity loss assays were performed by removing aliquots of heated pro- tein and quenching the reaction by diluting the sample in ice-cold buffer. Diluted samples were then incubated with chymotrypsin at a 1 : 1.5 molar ratio, and the level of enzyme activity remaining was assessed as described above. The remaining inhibitory activity was normalized and plot- ted as a function of duration of incubation at 48 (cid:2)C. Sam- ples monitored by light scatter were monitored in a Perkin- Elmer LS-50B Luminescence Spectrophotometer, with slit widths being chosen individually in order to obtain the best signal for each experiment. An excitation wavelength of 395 nm was chosen, with an emission wavelength of 405 nm. As for bis-ANS fluorescence assays, runs were ter- minated once the signal dropped severely due to protein precipitation. In each case, the experiments presented were performed at least five times.

Polymerization assays

9 Sivasothy P, Dafforn TR, Gettins PG & Lomas DA (2000) Pathogenic alpha 1-antitrypsin polymers are formed by reactive loop-beta-sheet A linkage. J Biol Chem 275, 33663–33668.

10 Stein PE & Carrell RW (1995) What do dysfunctional serpins tell us about molecular mobility and disease? Nat Struct Biol 2, 96–113.

Samples for TEM were prepared by diluting either ACT or ACTW194F in 50 mm Tris ⁄ HCl, 100 mm NaCl (pH 9.0) at a concentration of 1 lm. Samples were then incubated at 48 (cid:2)C for 30 min, and the reaction was quenched by pla- cing the samples on ice. TEM images were obtained using a Hitachi (Tokyo, Japan) H7500 transmission electron micro- scope. The acceleration voltage was 80 kV. Samples were

11 Whisstock JC, Pike RN, Jin L, Skinner R, Pei XY, Car- rell RW & Lesk AM (2000) Conformational changes in serpins. II. The mechanism of activation of antithrom- bin by heparin. J Mol Biol 301, 1287–1305.

FEBS Journal 274 (2007) 3622–3632 ª 2007 The Authors Journal compilation ª 2007 FEBS

3631

TEM

M. C. Pearce et al.

W194 protects against antichymotrypsin aggregation

cence spectroscopy. Arch Biochem Biophys 356, 296–300.

23 Devlin GL, Chow MK, Howlett GJ & Bottomley SP

12 Irving JA, Pike RN, Lesk AM & Whisstock JC (2000) Phylogeny of the serpin superfamily: implications of patterns of amino acid conservation for structure and function. Genome Res 10, 1845–1864.

13 Blouse GE, Perron MJ, Kvassman JO, Yunus S,

(2002) Acid denaturation of alpha1-antitrypsin: charac- terization of a novel mechanism of serpin polymeriza- tion. J Mol Biol 324, 859–870.

Thompson JH, Betts RL, Lutter LC & Shore JD (2003) Mutation of the highly conserved tryptophan in the ser- pin breach region alters the inhibitory mechanism of plasminogen activator inhibitor-1. Biochemistry 42, 12260–12272.

14 Meagher JL, Beechem JM, Olson ST & Gettins PG

24 Devlin GL, Parfrey H, Tew DJ, Lomas DA & Bottom- ley SP (2001) Prevention of polymerization of M and Z alpha1-antitrypsin (alpha1-AT) with trimethylamine N-oxide. Implications for the treatment of alpha1-at deficiency. Am J Respir Cell Mol Biol 24, 727–732. 25 Lomas DA, Evans DL, Finch JT & Carrell RW (1992)

The mechanism of Z alpha1-antitrypsin accumulation in the liver. Nature 357, 605–607.

(1998) Deconvolution of the fluorescence emission spec- trum of human antithrombin and identification of the tryptophan residues that are responsive to heparin bind- ing. J Biol Chem 273, 23283–23289.

26 Wei A, Rubin H, Cooperman BS & Christianson DW (1994) Crystal structure of an uncleaved serpin reveals the conformation of an inhibitory reactive loop. Nat Struct Biol 1, 251–258.

15 Tew DJ & Bottomley SP (2001) Probing the equilibrium denaturation of the serpin alpha(1)-antitrypsin with sin- gle tryptophan mutants; evidence for structure in the urea unfolded state. J Mol Biol 313, 1161–1169. 16 Whisstock JC, Skinner R, Carrell RW & Lesk AM

(2000) Conformational changes in serpins. I. The native and cleaved conformations of alpha(1)-antitrypsin. J Mol Biol 295, 651–665.

27 Horvath AJ, Irving JA, Rossjohn J, Law RH, Bottom- ley SP, Quinsey NS, Pike RN, Coughlin PB & Whis- stock JC (2005) The murine orthologue of human antichymotrypsin: a structural paradigm for clade A3 serpins. J Biol Chem 280, 43168–43178.

17 Pearce MC, Cabrita LD, Rubin H, Gore MG & Bot- tomley SP (2004) Identification of residual structure within denatured antichymotrypsin: implications for serpin folding and misfolding. Biochem Biophys Res Commun 324, 729–735.

28 Sharp AM, Stein PE, Pannu NS, Carrell RW, Berken- pas MB, Ginsburg D, Lawrence DA & Read RJ (1999) The active conformation of plasminogen activator inhib- itor 1, a target for drugs to control fibrinolysis and cell adhesion. Structure Fold Des 7, 111–118.

18 Stryer L (1965) The interaction of a naphthalene dye

29 Elliott PR, Abrahams JP & Lomas DA (1998) Wild- type alpha1-antitrypsin is in the canonical inhibitory conformation. J Mol Biol 275, 419–425.

with apomyoglobin and apohemoglobin. A fluorescent probe of non-polar binding sites. J Mol Biol 13, 482– 495.

30 Lomas DA (1994) Loop-sheet polymerization: the struc- tural basis of Z alpha1-antitrypsin accumulation in the liver. Clin Sci (Lond) 86, 489–495.

19 Pearce MC, Rubin H & Bottomley SP (2000) Conform- ational change and intermediates in the unfolding of alpha1-antichymotrypsin. J Biol Chem 275, 28513– 28518.

20 Crowther DC, Serpell LC, Dafforn TR, Gooptu B &

31 Bottomley SP & Chang WS (1997) The effects of react- ive centre loop length upon serpin polymerisation. Bio- chem Biophys Res Commun 241, 264–269.

32 Dafforn TR, Pike RN & Bottomley SP (2004) Physical characterization of serpin conformations. Methods 32, 150–158.

Lomas DA (2003) Nucleation of alpha1-antichymotryp- sin polymerization. Biochemistry 42, 2355–2363. 21 Devlin GL, Carver JA & Bottomley SP (2003) The

33 Goldenberg DP (1989) Analysis of protein conformation by gel electrophoresis. In Protein Structure: a Practical Approach (Creighton TE, ed.), pp. 225–250. IRL Press, New York, NY.

selective inhibition of serpin aggregation by the molecu- lar chaperone, alpha-crystallin, indicates a nucleation- dependent specificity. J Biol Chem 278, 48644–48650. 22 James EL & Bottomley SP (1998) The mechanism of alpha1-antitrypsin polymerization probed by fluores-

FEBS Journal 274 (2007) 3622–3632 ª 2007 The Authors Journal compilation ª 2007 FEBS

3632