Logo of prosciprotein sciencecshl presssubscriptionsetoc alertsthe protein societyjournal home
Protein Sci. 2001 Jul; 10(7): 1446–1453.
PMCID: PMC2374102
PMID: 11420446

Cavities of α1-antitrypsin that play structural and functional roles

Cheolju Lee,1,3 Jin-Soo Maeng,1,4 Jean-Pierre Kocher,2,5 Byungkook Lee,2 and Myeong-Hee Yu1
Author information Article notes Copyright and License information Disclaimer

Abstract

The native form of inhibitory serine protease inhibitors (serpins) is strained, which is critical for their inhibitory activity. Previous studies on stabilizing mutations of α1-antitrypsin, a prototype of serpins, indicated that cavities provide a structural basis for the native strain of the molecule. We have systematically mapped the cavities of α1-antitrypsin that play such structural and functional roles by designing cavity-filling mutations at residues that line the walls of the cavities. Results show that energetically unfavorable cavities are distributed throughout the α1-antitrypsin molecule, and the cavity-filling mutations stabilized the native conformation at 8 out of 10 target sites. The stabilization effect of the individual cavity-filling mutations of α1-antitrypsin varied (0.2–1.9 kcal/mol for each additional methylene group) and appeared to depend largely on the structural flexibility of the cavity environment. Cavity-filling mutations that decreased inhibitory activity of α1-antitrypsin were localized in the loop regions that interact with β-sheet A distal from the reactive center loop. The results are consistent with the notion that β-sheet A and the structure around it mobilize when α1-antitrypsin forms a complex with a target protease.

Keywords: α1-antitrypsin, cavity-filling mutations, conformational stability, native strain, molecular packing

The native forms of some proteins exist in a strained state rather than their thermodynamically most stable state. The native structure of serine protease inhibitors (serpins) (Huber and Carrell 1989), the spring-loaded structure of the fusion protein of influenza virus (Carr and Kim 1993; Bullough et al. 1994), and the structure of heat shock factors (Orosz et al. 1996) are typical examples. The native strain of these proteins plays an important role in regulating their biological functions (Wiley and Skehel 1987; Huber and Carrell 1989; Stein and Carrell 1995; Stefansson and Lawrence 1996; Carr et al. 1997). Serpins include various protease inhibitors found in plasma, such as α1-antitrypsin (α1AT), antithrombin III, α1-antichymotrypsin, plasminogen activator inhibitor-1, α2-antiplasmin, and C1-inhibitor (Huber and Carrell 1989). Ovalbumin and angiotensinogen are not protease inhibitors but are included in the serpin family because their amino acid sequences and tertiary structures are homologous to other members of the family (Huber and Carrell 1989). The inhibition process of serpins has been described as a suicide substrate mechanism (Rubin et al. 1990; Wright and Scarsdale 1995) in which the serpin, on binding to the target protease, is cleaved and then either is released as a free, cleaved molecule (substrate pathway) or forms a stable serpin–enzyme complex (inhibitory pathway). The stoichiometry of inhibition (the number of moles of inhibitors required to completely inhibit one mole of a target protease) is given by 1 + ksubstrate/kinhibition, in which ksubstrate and kinhibition are the rate constants for the substrate and inhibitory pathways, respectively. On binding with cognate target proteases, most serpin molecules partition into the complex formation, bringing the stoichiometry of inhibition close to one. However, interactions with noncognate proteases or some mutations in serpin molecules induce more partitioning into the substrate pathway, resulting in the increase in the stoichiometry of inhibition.

The crystal structure of a serpin–protease complex was reported recently (Huntington et al. 2000). When a serpin forms a complex with its target protease, the reactive center loop of the serpin is cleaved and inserted into the major β-sheet, sheet A (Lawrence et al. 1995; Shore et al. 1995), while the target protease is still attached to the cleaved loop of the serpin as an acyl intermediate (Lawrence et al. 1995; Wilczynska et al. 1995; Huntington et al. 2000). This conformational switching prevents catalytic deacylation due to the distortion of the protease active site (Huntington et al. 2000), which results in trapping the stabilized complex. The strain of the native serpin structure is probably required for this conformational switching. Indeed, recent stop-flow measurements showed that retarding the loop insertion into β-sheet A increased the stoichiometry of inhibition (Lawrence et al. 2000; Lee et al. 2000).

To elucidate the structural basis of the native strain of serpins, we previously screened and characterized amino acid substitutions that stabilized α1AT (Lee et al. 1996; Ryu et al. 1996; Im et al. 1999; Seo et al. 2000). Many stabilizing mutations of α1AT decreased the size of the side chains inside the molecule, whereas some others increased the size of side chains. This is unusual, because the hydrophobic side chains in the interior of common globular proteins are tightly packed as in small organic molecule crystals (Richards 1974), and both size reduction and incorporation of bulkier residues are usually destabilizing (Lee and Vasmatzis 1977; Karpusas et al. 1989). Structural examination of the mutation sites of α1AT, however, revealed that side chains are overpacked wherever large-to-small substitutions occurred (Lee et al. 1996; Ryu et al. 1996; Im et al. 1999; Seo et al. 2000), whereas cavities were observed near the small-to-large substitutions (Im et al. 1999; Seo et al. 2000). These studies indicated that the inhibitory serpin molecules are folded suboptimally, and that the side chain overpacking and the presence of cavities are parts of the structural features that facilitate conformational switching during complex formation.

In the present study, we designed cavity-filling mutations by analyzing the X-ray crystal structure of the wild-type molecule (Elliott et al. 1998) to identify systematically the cavities that are important for strain and inhibitory activity of α1AT. Small hydrophobic residues were selected from various residues that line the walls of cavities. We chose only sites where larger residues occupy equivalent positions in the sequence of ovalbumin, which is a more stable member of the serpin family, because many of the substituted residues stabilizing α1AT identified previously are ones found in the ovalbumin sequence (Lee et al. 1996; Im et al. 1999; Seo et al. 2000). We replaced the small hydrophobic residues at the target sites with larger hydrophobic residues. Our results show that cavities playing a role in the native strain are distributed throughout the α1AT molecule, but those that regulate inhibitory activity are highly localized at one end of the molecule that interacts with β-sheet A.

Results

Effect of cavity-filling mutations on the stability and activity of α1AT

There were 23 cavities inside the crystal structure of the wild-type α1AT, and the total cavity volume was about 850 Å3. We designed cavity-filling mutations of α1AT at the residues that met the following criteria: (1) residues whose side chain atoms form a part of the cavity wall; (2) small hydrophobic residues, alanine and valine; and (3) residues smaller than those at the equivalent positions in the sequence of ovalbumin. With such criteria, 10 target sites were chosen for cavity-filling mutations, which are shown and described in Figure 1 ▶ and Table 1​1,, respectively. They are located at α-helix A, hydrophobic core (α-helix B and β-sheet B), loops interacting with β-sheet A at the distal end from the reactive center loop, and β-sheet C. Each alanine target residue was replaced with valine, isoleucine, or leucine. Each valine target residue was replaced with isoleucine or leucine. Mutational effects on the conformational stability were measured by equilibrium unfolding in the presence of guanidine (Fig. 2A ▶). The changes in the conformational stability (ΔΔG) of the mutant proteins are summarized in Table 2​2.. Among the 10 target sites, stabilizing substitutions were found at 8 sites. At the other two sites, Val 55 and Ala 250, stabilizing mutations were not obtained. For the sites where all the substitutions increased the stability (Ala 31, Ala 183, and Ala 248), additional substitutions with phenylalanine were made. The value of ΔΔG varied from 0.2 to 3.8 kcal/mol for the stabilizing mutations. Generally, the increase in stability was saturated by a single (Val→Ile) or double (Ala→Val) methylene group insertion, except for the Ala 31 site, where stability increased steadily by replacement with bulkier hydrophobic residues up to phenylalanine. At almost all sites, isoleucine variants were more stable than leucine variants (for the Ala 31 site, the leucine variant was as stable as the isoleucine variant). For the Ala 183 and Ala 248 sites, where phenylalanine substitutions were also made, phenylalanine variants were more stable than leucine variants but less stable than isoleucine variants (Table 2​2).). Inhibitory activities of the stable α1AT variants were examined by determining the stoichiometry of inhibition against porcine pancreatic elastase (Fig. 2B ▶). Many of the stabilizing substitutions did not decrease the inhibitory activity significantly, but V145I, V173I, and V321I did (Table 2​2,, bold numbers).

Table 1.

Target cavities of α1AT for cavity-filling mutations

LocationaCavity IDbVolume (Å3)Residues involvedc
α-Helix A2145A31 (CB), A34, F35, H269, I272, T273, L276, L373, I375, K380, P382
Hydrophobic core524V55 (CG2), L99, L103, E376, S381, L383
1734F190, A248 (CB), M374, L383, F384
2212A250 (CB), I251, F252, F372, M374
Loops interacting with β-sheet A713L64, T68, L318, V321 (CG2)
1113V145 (CB, CG2), N146, A153, Q156
1413I157, V173 (CG2), A183 (CB), V185, K331
β-sheet C2143F227, I229, L240, A284 (CB), L286, V364 (CB, CG2)

a Location of the cavities was designated as described (Huber and Carrell 1989).

b The cavity identification number is arbitrary. Cavities are numbered consecutively as they are found in our cavity-searching algorithm.

c A one-letter code for amino acids was used. Bold-faced residues are the target sites of cavity-filling mutations in this study. Side-chain atoms that line the walls of each cavity are listed in parentheses.

Table 2.

Stability and inhibitory activity of α1AT variants

α1AT variantsaOvalbumin sequencebLocationcΔΔGd (kcal/mol)ΔΔG/CH2 (kcal/mol)Relative activitye
A31VfMhA0.6 ± 0.10.30.97
A31I0.9 ± 0.10.31.00
A31L0.9 ± 0.10.30.98
A31P2.5 ± 0.10.41.04
V55IIhB−0.2 ± 0.1N.D.g
V55L−0.3 ± 0.2N.D.
V145IIs1A0.2 ± 0.10.20.91
V145L−0.7 ± 0.2N.D.
V173ILthFs3A0.5 ± 0.10.50.83
V173L−0.6 ± 0.1N.D.
A183VfVs3A3.8 ± 0.21.90.95
A183I1.8 ± 0.10.60.96
A183L1.4 ± 0.10.5N.D.
A183F1.5 ± 0.10.3N.D.
A248VfMs3B2.3 ± 0.11.20.97
A248I2.2 ± 0.10.71.05
A248L1.5 ± 0.10.50.95
A248F1.8 ± 0.10.30.99
A250VMs3B−0.4 ± 0.1N.D.
A250I−0.4 ± 0.2N.D.
A250L−1.0 ± 0.2N.D.
A284VIs2C0.8 ± 0.10.40.99
A284I0.0 ± 0.20.97
A248L−0.8 ± 0.2N.D.
V321IIthI1s5A0.6 ± 0.10.60.85
V321L−1.2 ± 0.2N.D.
V364IfFs1C0.4 ± 0.10.41.01
V364L−0.3 ± 0.1N.D.

a A one-letter code for amino acids was used.

b Ovalbumin sequences were found in Huber and Carrell (1989).

c The nomenclature of Huber and Carrell (1989) was followed.

d Change in the conformational stability by each mutation (ΔΔG) was calculated using the equation equation M2, where ΔCm is the difference between the value of Cm, equilibrium transition midpoint, for wild-type and mutant protein, and ≤m> is the average of the "m-value," 7.9 kcal/mol per M. We preferred this method to an alternative one using ΔG (H2O) values determined by linear extrapolation to 0 M denaturant (Pace et al. 1989), because the m-values for the mutant α1AT proteins varied from 7.2 to 8.6 kcal/mol per M (the same variation was observed for the wild-type α1AT from experiment to experiment), whereas the Cm values were highly reproducible (standard deviations were within 0.02 M). The errors in ΔΔG were derived from the experimental errors of ΔCm.

e Relative activity is indicated as the inverse of the ratio of the stoichiometry of inhibition toward porcine pancreatic elastase of each mutant type over that of the wild type (1.66 ± 0.06).

f The same mutations were obtained previously from screening (Im et al. 1999; Seo et al. 2000).

g Not determined. Stoichiometries of inhibition of the destabilized proteins were not determined because of their higher heat lability. Unstable α1AT often brought about increased stoichiometry of inhibition, not because of the alteration in the partitioning between substrate and inhibitory pathways, but because of the decrease in the active concentration of α1AT during the assay.

An external file that holds a picture, illustration, etc. Object name is 19f1_C4TT.jpg

A stereo diagram of α1AT showing the target sites for cavity-filling mutations. Side chain heavy atoms and Cα of the 10 target residues are rendered in a space-filling model. The residue sites where cavity-filling mutations decreased the inhibitory activity are colored green. The reactive center loop is colored yellow. The figure was prepared using MOLSCRIPT (Kraulis 1991).

An external file that holds a picture, illustration, etc. Object name is 19f2_L1TT.jpg

Molecular properties of variant α1AT carrying cavity-filling mutations. (A) Guanidine-induced equilibrium unfolding transition of representative α1AT variants. (Filled circles) Wild type; (open circles) V173I; (squares) A183V; (triangles) A248V; (inverted triangles) A250V; (diamonds) A284V. The equilibrium unfolding of α1AT was monitored at 25°C by fluorescence spectroscopy with excitation at 280 nm and emission at 360 nm. Native α1AT was preequilibrated in the guanidine solution containing 10 mM phosphate, 50 mM NaCl, 1 mM EDTA (pH 6.5). Curves are the best fits to a two-state unfolding model. (B) Stoichiometry of inhibition of representative α1AT variants against porcine pancreatic elastase. (Filled circles) Wild type; (open circles) V145I; (triangles) V173I; (squares) A284V; (inverted triangles) V321I; (diamonds) V173I/A183V. Porcine pancreatic elastase (100 nM) was reacted with various α1AT concentrations at molar ratios of α1AT to porcine pancreatic elastase ([AT]/[PPE]) indicated at the abscissa. The residual protease activity was measured with 1 mM of N-succinyl-(Ala)3-p-nitroanilide. The lowest value of [AT]/[PPE] giving 100% inhibition is the determined stoichiometry of inhibition.

Effect of the double mutations in a single cavity

The side chains of Ala 183 and Val 173 form parts of a single common cavity (Table 1​1,, Cavity 14), and Ala 284 and Val 364 form parts of another single cavity (Table 1​1,, Cavity 21). To examine specific pairwise interactions between these residues, double mutations were constructed. From the stability increase for each single mutation and the double mutation, the interaction energy (ΔΔGint) between V173I and A183V was calculated to be +1.0 kcal/mol, and that between A284V and V364I to be −1.1 kcal/mol (Table 3​3).). Relative inhibitory activities toward porcine pancreatic elastase for the double mutants, V173I/A183V and A284V/V364I, were 0.53 and 0.96, respectively.

Table 3.

Stability and inhibitory activity of α1AT variants carrying double mutations

α1AT variantsΔΔG (kcal/mol)aΔΔGint (kcal/mol)bRelative activityc
V173I/A183V5.3 ± 0.3+1.00.53
V173I0.5 ± 0.10.83
A183V3.8 ± 0.20.95
A284V/V364I0.1 ± 0.1−1.10.96
A284V0.8 ± 0.10.99
V364I0.4 ± 0.11.01

a Change in the conformational stability by each mutation (ΔΔG) was calculated as in Table 2​2.. Stability increase of single mutations was the same as in Table 2​2.

b The interaction energy (ΔΔGint) between the two mutation sites was calculated from the equation equation M3, where ΔΔGXY is the stability increase of the double mutant, and ΔΔGX and ΔΔGY are the stability increases of the corresponding single mutants.

c Relative activity was determined from the stoichiometry of inhibition as in Table 2​2.. Relative activity of single mutations was the same as in Table 2​2.

Discussion

Cavity-filling mutations of α1AT

Many protein crystal structures contain internal cavities (Williams et al. 1994), but there are only a few examples in which the protein stability has been increased successfully by cavity-filling mutations (Akasako et al. 1997; Ohmura et al. 2001). The probable reason is that the bulkier residue required to fill a cavity often causes unfavorable van der Waals overlaps (Karpusas et al. 1989; Daopin et al. 1991). In contrast, cavity-filling mutations at 8 out of 10 target sites of α1AT increased the stability of the molecule. Even the unsuccessful substitutions at the other two sites showed only mild destabilization effects (Table 2​2).). Despite such a high frequency of successful stabilization by the cavity-filling of α1AT, only a modest individual stabilization effect was shown by most substitutions. The magnitudes of the contribution made by a single methylene group to the stability gain varied from 0.2 to 1.9 kcal/mol, but most were in the range of 0.3–0.6 kcal/mol (Table 2​2).). This value is lower than the average value of 1.3 kcal/mol reported by Pace from the cavity-creating mutation studies (Pace 1992), the value of 1.1 kcal/mol for cavity-filling mutations at the hydrophobic core of RNase HI (Akasako et al. 1997), and the value of 0.9–1.7 kcal/mol for cavity-creating mutations in T4 lysozyme (Eriksson et al. 1992). Thus, our small-to-large core mutations of α1AT can be characterized by two features: high frequency of successful stabilization but modest individual stabilization effect.

Structural basis of the cavity-filling stabilization of α1AT

In general, the stabilizing effect caused by the cavity-filling mutations of α1AT was saturated by the addition of one or two methylene groups, and isoleucine variants were more stable than leucine variants at most sites (Table 2​2).). The results indicate that there is a limitation in accommodating size increases and steric adjustments in the interior of α1AT. Studies on the cavity-filling mutations of hen egg white lysozyme indicated that the sum of free energy change caused by the hydrophobicity and the cavity size correlated well with the stability change (Ohmura et al. 2001). The cavity size or the hydrophobicity of the newly introduced residue, however, does not seem to be the major factor that determines the stabilizing effect of cavity-filling mutations of α1AT, because there is no correlation between the cavity size and the stability changes in the same kinds of substitutions. It appears that the stabilization effect of the cavity-filling mutations of α1AT depends not only on the size and shape of the cavities but also on the flexibility of their local environment, as has been suggested (Eriksson et al. 1992; Lee 1993). Consistent with this, Val 55 and Ala 250, the two sites in the hydrophobic core where cavity-filling mutations show destabilization effects, show lower B factors than Ala 248, the other site in the hydrophobic core, mutations of which stabilize the protein (side chain average B of 20 and 13 vs. 41).

The stabilization effect of cavity-filling mutations also appears to depend on the existence and distribution of other cavities nearby. The stabilization effect of A183V is 1.9 kcal/mol per methylene group, the greatest reported so far (Pace 1992; Akasako et al. 1997). Ala 183 is located on the exposed face of sheet A but is underneath helix F (Fig. 1 ▶). The cavity volume is 13 Å3, large enough to accommodate perhaps one methylene group. Interestingly, however, there are several small cavities nearby, especially at the interface between helix F and sheet A (Fig. 3A ▶). Structural adjustment, including rearrangement of other cavities, could result in better packing. Structural flexibility in this region was also noticeable in the double mutation analysis. The side chains of Val 173 and Ala 183 form the walls of a single cavity (Table 1​1;; Fig. 3A ▶) but do not interact directly (the distance between the two closest atoms is 6.6 Å). However, the stabilization obtained by the double mutation, V173I/A183V, is more than the sum of those of single mutations by 1.0 kcal/mol (Table 3​3).). This indicates that the stabilizing effect of A183V mutation is greater in the V173I background than in the wild-type background. The cavity distribution near Ala 183 is quite a contrast to that near Ala 284 and Val 364. Ala 284 and Val 364 form a single isolated cavity of 43 Å3 in the core next to sheet C (Fig. 3B ▶). Unlike the substitutions at Ala 183, A284V and V364I stabilized α1AT only by 0.4 kcal/mol per methylene group (Table 2​2).). In addition, the double mutations of A284V/V364I showed a negative effect (Table 3​3),), indicating that the stabilization effects of the two mutations are compensatory. This may be because of steric clash of the two side chains. These results indicate that the structure can adjust at the interface between sheet A, helix F, and the following loop, but that the adjustment in the interior of sheet C is rather limited.

An external file that holds a picture, illustration, etc. Object name is 19f3_F1TT.jpg

Stereo diagram showing cavities around Val 173 and Ala 183 (A) and Ala 284 and Val 364 (B). The cavity is represented by drawing a bead at each lattice point where a probe of 1.4 Å can be laid without creating any van der Waals contact with the surrounding atoms of protein body. The lattice points are the same as those used in calculating cavity volumes (see Materials and Methods). The size of beads is arbitrary. Residues at the mutation sites are depicted with thick sticks. The figures were prepared with RasMol (Sayle and Milner-White 1995).

Cavities that regulate inhibitory activity

If the native strain of the serpin molecule is used for functional regulation, increasing stability should affect the inhibitory activity. Among the stabilizing substitutions at eight sites of α1AT, only the substitutions at Val 145, Val 173, and Val 321 significantly affected the inhibitory activity (Table 2​2).). These sites are localized in the tertiary structure of α1AT at the region interacting with sheet A distal from the reactive center loop (Fig. 1 ▶, colored green). Val 173, located in the loop connecting helix F and sheet A, is involved in Cavity 14 along with residues in helix F (Ile 157) and sheet A (Ala 183, Val 185, and Lys 331). This region appears to be critical for the complex formation with a target protease because our previous studies of the stabilizing mutations at Gly 117 (Lee et al. 2000) and Lys 335 (Im and Yu 2000) also showed clearly that the strain in this region contributes to the inhibitory activity. Gly 117 is located in strand 2 of sheet A and interacts with Tyr 160 of helix F, and Lys 335 is located in strand 5 of sheet A and interacts with residues at the tip of helix F and the following loop that cover sheet A. It is likely that helix F and the following loop of α1AT have to be lifted to open β-sheet A for the facile loop insertion that occurs during complex formation with a target protease. Consistent with this, our double mutation analysis in this region (Table 3​3)) indicates strongly the structural flexibility of the region.

Val 145 and Val 321 are located in the loop region at the bottom of the molecule (Fig. 1 ▶). Val 145 is involved in Cavity 11 along with residues of helix F (Ala 153, Gln 156; Table 1​1),), and Val 321 is involved in Cavity 7 along with residues of helix B in the hydrophobic core (Leu 64, Thr 68; Table 1​1).). The functional importance of these sites has not been suggested explicitly in previous reports. In the recently reported α1AT–trypsin complex structure (Huntington et al. 2000), the reactive center loop of the serpin is inserted fully into A sheet, and the protease is located at the bottom of the molecule, slightly skew of the central axis of the serpin molecule. Our mutations at Val 145 and Val 321, however, do not appear to influence the contact interactions between the serpin and the protease because the mutation sites are in the interior of the serpin molecule and are apart from the final protease docking site. Rather, the cavity-filling by V145I and V321I substitutions might strengthen the interactions between sheet A and the surrounding secondary structures. In the complex structure (PDB code ID: 1ezx), the cavity interactions and their distribution near Val 145 and Val 321 are rearranged substantially and the cavities near Val 173 disappeared completely, supporting that the cavity interactions near these residues in the native form are to be mobilized during the complex formation. Our results are consistent with the notion that a substantial opening and mobilization of β-sheet A and the interacting secondary structures are accompanied during the complex formation.

Implications of the α1AT structure

The total cavity volume and average packing density computed in the α1AT structure are 2% and 0.67, respectively. These values are not very different from those of common globular proteins (Richards 1974; Rashin et al. 1986). However, previous mutational studies strongly indicated that the native form of α1AT is suboptimally folded and that it can be stabilized by large-to-small mutations at many sites as well as by small-to-large mutations at other sites (Lee et al. 1996; Ryu et al. 1996; Im et al. 1999; Seo et al. 2000). The structural features that stabilize a protein molecule by a small-to-large substitution will be different from those that allow stabilization upon a large-to-small substitution. Stabilization by large-to-small substitutions must arise from relieving a preexisting strain in an overpacked, inflexible region. Stabilization by small-to-large substitutions, on the other hand, would normally require an efficient repacking, because the shape of the new larger side chains will rarely fit the existing cavity perfectly (Karpusas et al. 1989; Daopin et al. 1991). Such efficient repacking can be obtained only when the structure is flexible. It is also not unreasonable to expect that the stabilization by a large-to-small mutation will generally be larger in magnitude than that by a small-to-large mutation because the former involves relieving a preexisting strain whereas the latter involves finding a suitable new arrangement within the given structure. This explains why the magnitude of the stabilization observed for most of the cavity-filling mutations of the present study tends to be small. The fact that stabilization could be obtained at so many sites, both by small-to-large and large-to-small mutations, indicates that the α1AT structure is made of a "mosaic" of many over- and under-packed regions.

Conclusions

We systematically mapped the cavities of α1AT that play roles in the stability and function of the molecule. Although energetically unfavorable cavities are found throughout the molecule of α1AT, cavity-filling mutations that affect the inhibitory activity appear to occur only at the sites that need to be mobilized during complex formation with a target protease. Our results on the stabilizing mutations of α1AT provide insight into the unusual packing nature of the molecule that appears to be the structural basis of the strain of its native form. Our results also indicate that a substantial portion of the structural components interacting with sheet A of α1AT, the major β-sheet, is designed to be flexible and underpacked to allow complete opening of sheet A during complex formation with a target protease.

Materials and methods

Materials

The plasmid for α1AT expression in Escherichia coli and the purification of recombinant proteins were described previously (Kwon et al. 1994). Mutations were induced by oligonucleotide-directed mutagenesis and confirmed by sequencing. Ultrapure guanidine hydrochloride was purchased from ICN. Porcine pancreatic elastase and N-succinyl-(Ala)3-p-nitroanilide were purchased from Sigma. All other reagents were of analytical grade.

Cavity mapping

The atomic coordinates for the crystal structure of wild-type α1AT were taken from the PDB protein databank (code ID: 2psi, now updated to 1qlp) (Elliott et al. 1998). Cavities were identified (Alard and Wodak 1992) and their volumes calculated analytically using the program SurVol (Alard 1991). The atomic radii used were those of Chothia (1975). The probe radius was set to 1.4 Å. The volumes reported throughout this paper are molecular volumes (total empty space), which include the volumes of the probe as well as the excluded volumes (Lee and Richards 1971; Richards 1977).

Equilibrium unfolding and determination of stability change

Conformational stability of α1AT variants was determined by guanidine-induced equilibrium unfolding in 10 mM phosphate, 50 mM NaCl, 1 mM EDTA (pH 6.5) at 25°C. Native α1AT was diluted into the appropriate guanidine solution and incubated for about 4 h at 25°C. The final protein concentration was 5 μg/mL. The equilibrium unfolding of α1AT was monitored by fluorescence spectroscopy, with excitation at 280 nm and emission at 360 nm (slit width, 5 nm for both). Experimental data of the fluorescence measurement were fitted to a two-state unfolding model, details of which were described previously (Kwon et al. 1994). Changes in the free energy of unfolding (ΔΔG) of the mutant proteins were determined according to Pace et al. (1989). Briefly, it was calculated with the fitted thermodynamic parameters and the equation equation M1, where ΔCm is the difference between the value of Cm, equilibrium transition midpoint, for wild-type and mutant protein, and <m> is the average of the "m-value," a measure of the dependence of the free energy of unfolding (ΔG) on denaturant concentration.

Determination of stoichiometry of inhibition

The stoichiometry of inhibition was determined as described (Rubin et al. 1990). Various amounts of α1AT variants were incubated with 100 nM porcine pancreatic elastase in 50 μL of assay buffer (30 mM phosphate, 160 mM NaCl, 0.1% PEG 8000, 0.1% Triton X-100 at pH 7.4). After incubation for 10 min at 37°C, the reaction mixture was diluted 10-fold with the same buffer, and the residual protease activity was determined using N-succinyl-(Ala)3-p-nitroanilide as a substrate. The activity inhibition was extrapolated to yield the minimum molar ratio of α1AT to the protease giving 100% inhibition.

Acknowledgments

We thank Drs. H. Im and S.-H. Park for helpful comments. We also thank Dr. E.J. Seo for graphic work. This research was supported by National Creative Research Initiatives grant from the Korean Ministry of Science and Technology.

The publication costs of this article were defrayed in part by payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 USC section 1734 solely to indicate this fact.

Abbreviations

  • serpin, serine protease inhibitor

  • α1AT, α1-antitrypsin

Notes

Article and publication are at http://www.proteinscience.org/cgi/doi/10.1101/

References

  • Akasako, A., Haruki, M., Olbatake, M., and Kanaya, S. 1997. Conformational stabilities of Escherichia coli RNase HI variants with a series of amino acid substitutions at a cavity within the hydrophobic core. J. Biol. Chem. 272 18686–18693. [PubMed] [Google Scholar]
  • Alard, P. 1991. "Calculs de surface et d'energie dans le domaine des macromolecules." Ph.D. thesis, University of Libre de Bruxelles.
  • Alard, P. and Wodak, S.J. 1992. Detection of cavities in a set of interpenetrating spheres. J. Comput. Chem. 12 918–922. [Google Scholar]
  • Bullough, P.A., Hughson, F.M., Skehel, J.J., and Wiley, D.C. 1994. Structure of influenza haemagglutinin at the pH of membrane fusion. Nature 371 37–43. [PubMed] [Google Scholar]
  • Carr, C.M. and Kim, P.S. 1993. A spring-loaded mechanism for the conformational change of influenza hemagglutinin. Cell 73 823–832. [PubMed] [Google Scholar]
  • Carr, C.M., Chaudhry, C., and Kim, P.S. 1997. Influenza hemagglutinin is spring-loaded by a metastable native conformation. Proc. Natl. Acad. Sci. 94 14306–14313. [PMC free article] [PubMed] [Google Scholar]
  • Chothia, C. 1975. Structural invariants in protein folding. Nature 254 304–308. [PubMed] [Google Scholar]
  • Daopin, S., Alber, T., Baase, W.A., Wozniak, J.A., and Matthews, B.W. 1991. Structural and thermodynamic analysis of packing of two α-helices in bacteriophage T4 lysozyme. J. Mol. Biol. 221 647–667. [PubMed] [Google Scholar]
  • Elliott, P.R., Abrahams, J.P., and Lomas, D.A. 1998. Wild-type α1-antitrypsin is in the canonical inhibitory conformation. J. Mol. Biol. 275 419–425. [PubMed] [Google Scholar]
  • Eriksson, A.E., Baase, W.A., Zhang, X.J., Heinz, D.W., Blaber, M., Baldwin, E.P., and Matthews, B.W. 1992. Response of a protein structure to cavity-creating mutations and its relation to the hydrophobic effect. Science 255 178–183. [PubMed] [Google Scholar]
  • Huber, R. and Carrell, R.W. 1989. Implications of the three-dimensional structure of α1-antitrypsin for structure and function of serpins. Biochemistry 28 8951–8966. [PubMed] [Google Scholar]
  • Huntington, J.A., Read, R.J., and Carrell, R.W. 2000. Structure of a serpin–protease complex shows inhibition by deformation. Nature 407 923–926. [PubMed] [Google Scholar]
  • Im, H. and Yu, M.-H. 2000. Role of Lys335 in the metastability and function of inhibitory serpins. Prot. Sci. 9 934–941. [PMC free article] [PubMed] [Google Scholar]
  • Im, H., Seo, E.J., and Yu, M.-H. 1999. Metastability in the inhibitory mechanism of human α1-antitrypsin. J. Biol. Chem. 274 11072–11077. [PubMed] [Google Scholar]
  • Karpusas, M., Baase, W.A., Matsumura, M., and Matthews, B.W. 1989. Hydrophobic packing in T4 lysozyme probed by cavity-filling mutants. Proc. Natl. Acad. Sci. 86 8237–8241. [PMC free article] [PubMed] [Google Scholar]
  • Kraulis, P.J. 1991. MOLSCRIPT. A program to produce both detailed and schematic plots of protein structure. J. App. Crystallogr. 24 946–950. [Google Scholar]
  • Kwon, K.-S., Kim, J., Shin, H.S., and Yu, M.-H. 1994. Single amino acid substitutions of α 1-antitrypsin that confer enhancement in thermal stability. J. Biol. Chem. 269 9627–9731. [PubMed] [Google Scholar]
  • Lawrence, D.A., Ginsburg, D., Day, D.E., Berkenpas, M.B., Verhamme, I.M., Kvassman, J.-O., and Shore, J.D. 1995. Serpin–protease complexes are trapped as stable acyl–enzyme intermediates. J. Biol. Chem. 270 25309–25312. [PubMed] [Google Scholar]
  • Lawrence, D.A., Olson, S.T., Muhammad, S., Day, D.E., Kvassman, J.-O., Ginsburg, D., and Shore, J.D. 2000. Partitioning of serpin–proteinase reactions between stable inhibition and substrate cleavage is regulated by the rate of serpin reactive center loop insertion into β-sheet A. J. Biol. Chem. 275 5839–5844. [PubMed] [Google Scholar]
  • Lee, B. 1993. Estimation of the maximum change in stability of globular proteins upon mutation of a hydrophobic residue to another of smaller size. Prot. Sci. 2 733–738. [PMC free article] [PubMed] [Google Scholar]
  • Lee, B. and Richards, F.M. 1971. The interpretation of protein structures: Estimation of static accessibility. J. Mol. Biol. 55 379–400. [PubMed] [Google Scholar]
  • Lee, B. and Vasmatzis, G. 1997. Stabilization of protein structures. Curr. Opin. Biotech. 8 423–428. [PubMed] [Google Scholar]
  • Lee, C., Park, S.-H., Lee, M.-Y., and Yu, M.-H. 2000. Regulation of protein function by native metastability. Proc. Natl. Acad. Sci. 97 7727–7731. [PMC free article] [PubMed] [Google Scholar]
  • Lee, K.N., Park, S.D., and Yu, M.-H. 1996. Probing the native strain in α1-antitrypsin. Nat. Struct. Biol. 3 497–500. [PubMed] [Google Scholar]
  • Ohmura, T., Ueda, T., Ootsuka, K., Saito, M., and Imoto, T. 2001. Stabilization of hen egg white lysozyme by a cavity-filling mutation. Prot. Sci. 10 313–320. [PMC free article] [PubMed] [Google Scholar]
  • Orosz, A., Wisniewski, J., and Wu, C. 1996. Regulation of Drosophila heat shock factor trimerization: Global sequence requirements and independence of nuclear localization. Mol. Cell. Biol. 16 7018–7030. [PMC free article] [PubMed] [Google Scholar]
  • Pace, C.N. 1992. Contribution of the hydrophobic effect to globular protein stability. J. Mol. Biol. 226 29–35. [PubMed] [Google Scholar]
  • Pace, C.N., Shirley, B.A., and Thomson, J.A. 1989. Measuring the conformational stability of a protein. In Protein structure: A practical approach (ed. T.E. Creighton), pp. 311–330. IRL, New York.
  • Rashin, A.A., Iofin, M., and Honig, B. 1986. Internal cavities and buried waters in globular proteins. Biochemistry 25 3619–3625. [PubMed] [Google Scholar]
  • Richards, F.M. 1974. The interpretation of protein structures: Total volume, group volume distributions and packing density. J. Mol. Biol. 82 1–4. [PubMed] [Google Scholar]
  • ———. 1977. Areas, volumes, packing and protein structure. Ann. Rev. Biophys. Bioeng. 6 151–176. [PubMed] [Google Scholar]
  • Rubin, H., Wang, Z.M., Nickbarg, E.B., McLarney, S., Naidoo, N., Schoenberger, O.L., Johnson, J.L., and Cooperman, B.S. 1990. Cloning, expression, purification, and biological activity of recombinant native and variant human α1-antichymotrypsins. J. Biol. Chem. 265 1199–1207. [PubMed] [Google Scholar]
  • Ryu, S.-E., Choi, H.-J., Kwon, K.-S., Lee, K.N., and Yu, M.-H. 1996. The native strains in the hydrophobic core and flexible reactive loop of a serine protease inhibitor: Crystal structure of an uncleaved α1-antitrypsin at 2.7 Å. Structure 4 1181–1192. [PubMed] [Google Scholar]
  • Sayle, R. and Milner-White, E.J. 1995. RasMol: Biomolecular graphics for all. Trends Biochem. Sci. 20 374–376. [PubMed] [Google Scholar]
  • Seo, E.J., Im, H., Maeng J.-S., Kim, K.E., and Yu, M.-H. 2000. Distribution of the native strain in human α1-antitrypsin and its association with protease inhibitor function. J. Biol. Chem. 275 16904–16909. [PubMed] [Google Scholar]
  • Shore, J.D., Day, D.E., Francis-Chmura, A.M., Verhamme, I., Kvassman, J., Lawrence, D.A., and Ginsburg, D. 1995. A fluorescent probe study of plasminogen activator inhibitor-1: Evidence for reactive center loop insertion and its role in the inhibitory mechanism. J. Biol. Chem. 270 5395–5398. [PubMed] [Google Scholar]
  • Stefansson, S. and Lawrence, D.A. 1996. The serpin PAI-1 inhibits cell migration by blocking integrin αvβ3 binding to vitronectin. Nature 383 441–443. [PubMed] [Google Scholar]
  • Stein, P.E. and Carrell, R.W. 1995. What do dysfunctional serpins tell us about molecular mobility and disease? Nat. Struct. Biol. 2 96–113. [PubMed] [Google Scholar]
  • Wilczynska, M., Fa, M., Ohlsson, P.-I., and Ny, T. 1995. The inhibition mechanism of serpins: Evidence that the mobile reactive center loop is cleaved in the native protease–inhibitor complex. J. Biol. Chem. 270 29652–29655. [PubMed] [Google Scholar]
  • Wiley, D.C. and Skehel, J.J. 1987. The structure and function of the hemagglutinin membrane glycoprotein of influenza virus. Ann. Rev. Biochem. 56 365–394. [PubMed] [Google Scholar]
  • Williams, M., Goodfellow, J.M., and Thornton, J.M. 1994. Buried waters and internal cavities in monomeric proteins. Prot. Sci. 3 1224–1235. [PMC free article] [PubMed] [Google Scholar]
  • Wright, H.T. and Scarsdale, J.N. 1995. Structural basis for serpin inhibitor activity. Prot. Struct. Funct. Genet. 22 210–225. [PubMed] [Google Scholar]
Articles from Protein Science : A Publication of the Protein Society are provided here courtesy of The Protein Society