The Z Mutation Alters the Global Structural Dynamics of α1-Antitrypsin

α1-Antitrypsin (α1AT) deficiency, the most common serpinopathy, results in both emphysema and liver disease. Over 90% of all clinical cases of α1AT deficiency are caused by the Z variant in which Glu342, located at the top of s5A, is replaced by a Lys which results in polymerization both in vivo and in vitro. The Glu342Lys mutation removes a salt bridge and a hydrogen bond but does not effect the thermodynamic stability of Z α1AT compared to the wild type protein, M α1AT, and so it is unclear why Z α1AT has an increased polymerization propensity. We speculated that the loss of these interactions would make the native state of Z α1AT more dynamic than M α1AT and that this change renders the protein more polymerization prone. We have used hydrogen/deuterium exchange combined with mass spectrometry (HXMS) to determine the structural and dynamic differences between native Z and M α1AT to reveal the molecular basis of Z α1AT polymerization. Our HXMS data shows that the Z mutation significantly perturbs the region around the site of mutation. Strikingly the Z mutation also alters the dynamics of regions distant to the mutation such as the B, D and I helices and specific regions of each β-sheet. These changes in global dynamics may lead to an increase in the likelihood of Z α1AT sampling a polymerogenic structure thereby causing disease.


Introduction
The misfolding and subsequent polymerization of members of the serpin superfamily leads to a variety of diseases collectively known as the Serpinopathies [1]. The most common serpinopathy is a 1 -antitrypsin (a 1 AT) deficiency, which affects approximately 1 in 2000 people [2]. The serpin, a 1 AT, is synthesized by hepatocytes and released into the circulation where it protects the lung from the action of neutrophil elastase. Over 70 mutations have been identified that lead to a 1 AT deficiency. The most common pathological variant, accounting for 95% of all clinical cases, is the Z variant [3,4,5] in which Glu342, which is located at the junction between the top of s5A and the base of the reactive center loop (RCL), is replaced by a Lys (Fig. 1a). The presence of this mutation results in the removal of both a salt bridge to Lys290 and a hydrogen bond to Thr203. The loss of these interactions brings about misfolding and polymerization of the protein within the endoplasmic reticulum of hepatocytes resulting in a lack of secretion and is characterized by a reduction in plasma levels to 10-15% of normal [6]. The polymerized Z a 1 AT damages the hepatocytes and predisposes the carrier to liver disease. The decreased plasma levels give rise to severe early onset emphysema.
The molecular basis of Z a 1 AT polymerization is not completely understood. The structure, stability and polymerization characteristics of native Z a 1 AT have been studied using a range of biochemical and biophysical techniques [4,7,8,9]. These data reveal that Z a 1 AT, in contrast to wild type a 1 AT (M a 1 AT), polymerizes rapidly when incubated at physiological temperatures [4,6,10]. The crystal structure of Z a 1 AT has not yet been determined however it is still an efficient proteinase inhibitor indicating that it possesses the serpin fold [9,10]. In support of this the equilibrium unfolding of Z a 1 AT has been studied and shown to be the same as M a 1 AT suggesting that compensating interactions are formed in Z a 1 AT to counteract for the loss of the two native state interactions [8,11]. Two additional pieces of experimental evidence suggest that there are substantial differences within the native state of Z a 1 AT. First, recent spectroscopic data using mutants of M and Z a 1 AT have shown that there are structural differences between the proteins [8,11]. Secondly, kinetic unfolding studies indicated that in the three state unfolding reaction the transition from the native state to a partially folded intermediate state proceeds almost two times faster for Z a 1 AT than for M a 1 AT [4]. Therefore, we speculated that the native state of Z a 1 AT may be more dynamic than M a 1 AT and that it is this change which renders the protein prone to polymerization. To examine this hypothesis we have used hydrogen/deuterium exchange combined with mass spectrometry (HXMS) to determine the structural and dynamic differences between native Z and M a 1 AT and to reveal the molecular basis of Z a 1 AT polymerization.

Results
M and Z a 1 AT appear to possess similar thermodynamic stability [4,8], yet native Z a 1 AT, incubated at physiological temperatures (37-41uC), readily polymerizes whereas M a 1 AT does not [6,8,9]. One potential explanation for the rapid polymerization of Z a 1 AT is that it is in, or can access more readily, a non-native, yet active, conformation [12,13]. In order to examine this possibility we compared the native state dynamics of both M and Z a 1 AT using HXMS coupled with pepsin digestion, to measure the flexibility of specific regions within these serpins [14].
Both M and Z a 1 AT were expressed in P. pastoris and purified as previously described [10]. The H/D exchange of the M a 1 AT   used in this study is in excellent agreement with our previous study using M a 1 AT produced in E. coli [15]. Tandem mass spectrometry experiments were carried out and 132 overlapping peptic fragments were identified from both M and Z a 1 AT. A comparison of H/D exchange of native M and Z a 1 AT was performed at pD 8 and 25uC followed by pepsin digestion and HPLC-MS to quantify the mass of each peptic fragment. Analysis of the pepsin digest of undeuterated M and Z a 1 AT under the rapid HPLC gradient required for the H/D experiments identified 19 peptic fragments, with good signal to noise ratio (Table 1). These fragments cover 79% of the entire a 1 AT molecule and are well distributed throughout the sequence; the only significant gaps in coverage encompass regions around helices A and H (Fig. 1a,b).
Using previously established procedures in our laboratory [15,16] we were able to measure the kinetics of deuterium incorporation for the 19 peptic fragments (Table 1), from both M and Z a 1 AT over a period of 2000 sec. The first experimental point measured was 10 sec after isotope exchange was initiated. Deuterium labelling was performed at 25uC, with deuteration times ranging from 10 to 2000 s. Under these experimental conditions, both M and Z a 1 AT remained in a monomeric form during the deuterium labelling time (Fig. 2).
Twelve pairs of peptides from M and Z a 1 AT displayed similar kinetics and extent of deuterium incorporation. These data therefore suggest that the Z mutation had minimal structural or dynamic effects on the serpin in these regions which are spread throughout the molecule (Fig. 3).  Six peptides showed a significant enhancement of deuterium exchange in Z a 1 AT compared to M a 1 AT (Fig. 4a-f). The extent of labelling was increased for peptic fragments 38-62 (hA-hB), 62-77 (hB-hC), 101-119 (b2A), 297-303 (b6A-hI), 339-353 (b5A-Linker) and 372-384 (b4B-b5B). Results for these peptic fragments were mapped onto the crystal structure of M a 1 AT with the peptides showing an increase in exchange of the peptide in Z a 1 AT compared to M a 1 AT coloured red (Fig. 4a-f).
Only one peptide, 191-212, (the loop connecting b3A and b4C) showed a significant reduction of deuterium exchange in Z a 1 AT compared to M a 1 AT (Fig. 5a,b). Results for this peptide was mapped onto the structure of M a 1 AT with the peptide showing an increase in exchange in M a 1 AT compared to Z a 1 AT coloured in blue (Fig. 5b).
Three peptic fragments, 127-142 (hE-b1A), 191-212 (b3A-b4C) and 252-272 (b3B-hG-hH), displayed greater than two-fold protection in Z a 1 AT in comparison to M a 1 AT after only 10 seconds of deuterium labelling ( Fig. 6a). Under these conditions, amides in unfolded regions of the molecule will undergo nearly complete exchange, while hydrogens in folded regions remain largely unexchanged. This type of pulse labelling has been shown to be an effective tool for monitoring site specific folding and unfolding in proteins [17]. The hydrogens in these 3 peptides are less labile due to decreased flexibility or a different conformation of the peptide in Z a 1 AT. It is also clear that there is considerable exchange in peptic fragments 38-62 (b6B-hB), 62-77 (hB-hC), 101-119 (hD-b2A), 215-227 (b3C) 297-303 (b6A-hI), 325-338 (b5A), 339-353 (b5A-Linker) and 372-384 (b4B-b5B) (Fig. 6B). These data suggest that regions covered by these peptides are either partially unfolded or marginally stable in Z a 1 AT.
Significant differences in deuterium incorporation are also observed at longer exchange times and suggest that globally Z a 1 AT is more dynamic than M a 1 AT. To better represent the data we have grouped the deuterium exchange into classes depending on the exchange at 2000 seconds. Class 1 peptides exchange rapidly in the native state with greater than 80% exchange in 2000 seconds. Peptic fragments 62-77 encompassing the helices B-C show rapid exchange in Z a 1 AT only, suggesting a lack of stable secondary structure leading to a more dynamic molecule. The rapid exchange of the peptic fragments 339-352 and 352-372 corresponding to the RCL show an enhanced exchange suggesting less interactions in Z than WT a 1 AT.
Class 2 peptides show moderate exchange (30-80%) at 2000 seconds in the native state and are shown in yellow. Residues within areas of high a-helical and b-sheet content are expected to exchange more slowly than those of turns and loops and make up the majority of peptides seen for class 2 [18]. Peptic fragments 38-51, 127-142, and 304-317 associated with helix D, E, G and I respectively show only a 60% exchange at 2000 seconds in M and Z a 1 AT. Residues 191-212 (the loop connecting b3A and b4C) also fall into the category although this is the only peptide that shows a reduction in exchange in Z a 1 AT.
In M a1AT, peptic fragments that are protected from exchange include the top of hD and b2A, , b1B-b2B,, b6A-hI and the loop connecting b4B to b5B, (101-119, , 227-240, 297-303 and 372-384) previously attributed to the hydrophobic core [15,19] and are described as showing class 3 exchange in yellow in figure 7. Z a1AT shares with M a1AT only residues (227-240) that show significant protection and differently from M protein displays residues 191-212 having high protection as discussed before, both peptides belonging to class 3.

Discussion
The structural integrity of a protein generally relies on its ability to adopt and maintain a unique native state. For members of the serpin superfamily the integrity of the native state must also allow local motions that facilitate proteinase inhibition. However, these motions can be highjacked and used to promote disease causing polymerization. In the case of Z a 1 AT we have a protein whose fold and apparent thermodynamic stability are similar to M a 1 AT, yet it polymerizes from the native state much more rapidly [4]. Using HDX we have examined the global and local changes that arise in the natively folded ensembles of Z a 1 AT, this study shows that previous studies may have underestimated the effect of the E242K substitution on the molecule and the effects are not just localised at the site of mutation but extend to distant regions of the structure.
The HDX results presented here reveal that the structural effects due to the E342K mutation are not distributed uniformly throughout the structure, but are instead localized in specific regions. Exchange at 10 seconds indicates partial loss of structure in several regions, the most dramatic being b2A and the top of hD (Fig. 6a). Compared with M a 1 AT, Z a 1 AT has lost ,8 hydrogen bonds in this region, suggesting significant disruption of interactions between b2A and the surrounding structural elements. Previous molecular dynamics simulations support the idea that the effects of the Glu342Lys substitution can propagate to this region. While significant disruption of b2A was not observed on the 50 ns timescale to the simulations, a large change in the conformation of the hD-b2A loop was observed, consistent with our HDX results [11]. The top of helix F remains highly dynamic as previously seen in M a 1 AT [15,19]. Deuterium levels at 10 s also indicate that the region covered by residues 339-353 has lost ,3 hydrogen bonds, suggesting a loss of structure at the top of b-sheet A that is an important site in the early stages of RCL insertion. Additionally, there is disruption of hydrogen bonds between the central portion of b3A and the adjacent b2A and b5A. Loss of hydrogen bonds in these regions, together with smaller but still significant losses in helices A, B, and C, clearly demonstrates that the E342K mutation disrupts native structure in areas both distant from and close to the mutation site. In addition to the loss of hydrogen bonds, deuterium uptake at 10 seconds also indicates the formation of additional hydrogen bonds in regions spanned by residues 127-142, 191-212  and 252-272, in Z a 1 AT compared to M. These regions correspond to hE-b1A, b3A-b4C and hG-hH respectively. However, the added hydrogen bonds do not appear to be stable, as the additional protection against deuterium uptake in Z a 1 AT is lost within 100 seconds (for peptides 191-212 and 352-372) to 1000 seconds (for peptide 191-212). Taken together these results on deuterium uptake at 10 seconds clearly indicate that Z a 1 AT exists in an altered native conformation compared to M a 1 AT and that there is significant disruption of hydrogen bonding in much of b-sheet A which is in agreement with our previously published data using site single point mutations and molecular dynamic simulations [11] [8].
Significant differences in the extent of deuterium exchange at longer labeling times were found within 7 peptides (Fig. 4 and 5), indicating dynamic and structural differences between the two proteins. One of the peptides (residues 339-353, the top of s5A and the RCL) includes the mutation site, Glu 342; this peptide was observed to be more mobile in Z a 1 AT (Fig. 4e). Also in this region was peptide 191-212 (b3A-b4C) which displayed decreased deuterium uptake indicating that this region contains additional hydrogen bonds and is more rigid in Z a 1 AT (Fig. 5). This increased rigidity may be due to stabilizing interactions between Lys342 and Glu199. Trp194 is located in this region, and the increased rigidity may appear to be at odds with previous results showing differences in Trp fluorescence between M and Z a 1 AT. We note, however, that while the region covered by the peptide containing Trp194 shows decreased exchange at short times, the top of b5A, which is immediately adjacent to Trp194, shows increased exchange, indicating a more dynamic local environment. We therefore conclude that there is no inconsistency between the fluorescence and H/D exchange data. These changes in deuterium uptake suggests that the interactions within the vicinity of the mutation are altered by the removal of the salt bridge between Glu342 and K290, which allows this region to sample a conformation in which the top of s5A is open. This open conformation is maintained by new interactions formed between Lys342 and Val200, Thr203 present within peptide 191-212 [11].
There are several regions, distant from the mutation site, whose structure and stability depend upon the residues they pack against such as helix A, B and H which are affected by the Z mutation  [20,21]. We observe a significant increase in the flexibility of peptic fragments corresponding to the helix B in Z a 1 AT (Fig. 4a,  b). Peptide 38-51 show a comparable behavior in both M and Z a 1 AT (Fig. 3a) whereas an increase in exchange is seen for residues 38-62 (Fig. 4a) suggesting that the increase in exchange can be attributed to the B. The flexibility in this region suggests that the amide hydrogen bonds in these peptides are less stable and the packing around the helix is loosened in Z relative to M a 1 AT and may explain the loss of helical structure seen in the CD spectra of Z a 1 AT [4,9].
What is apparent from the experimental data is reduced protection in the hydrophobic core of Z a 1 AT (Figs. 4 and 6). In fact, regions showing significantly increased exchange in Z a 1 AT form a nearly contiguous group that encompasses much of the core of the molecule (Figure 8). The exchange resistant core of M a 1 AT has previously been shown to consist of b-sheet rich regions [15,16]. In M a 1 AT, peptic fragments that are protected from exchange include b2A, b3A-b4C, b2B-b3B and b6A-hI, (101-119, 191-212, 227-240, 297-303) and are described as showing class 3 exchange in figure 7. For Z a 1 AT only residues (227-240) show significant protection with residues corresponding to b2A, b3A-b4C, b2B-3B and b6A-hI demonstrating class 2 behaviour (Fig. 7). The peptide covering residues 227-240 are heavily protected from exchange in both M and Z a 1 AT with less than 10% of the hydrogen available for exchange exchanging in the experimental time frame. This peptide, which has been identified in several previous studies as being resistant to chemical denaturation [22,23] and has been proposed to play a role as folding initiator [23], remains unaffected by the Z mutation. The increased deuterium exchange seen for the 'core' peptides in Z a 1 AT may allow the molecule to sample conformations that on the folding pathway and lead to accumulation of the polymerogenic folding intermediate that the open sheet intermediate is on-pathway and the loss of the salt bridge leads to enhanced lability and ability to switch to a polymerogenic conformation.
In conclusion, our data clearly demonstrates that the single mutation Glu342Lys results in global dynamic changes to the serpin fold. This in turns leads to an increase in the probability of Z a 1 AT sampling an open sheet structure capable of polymerisation.

Materials and Methods
Expression and purification of M and Z a 1 AT M and Z a 1 AT were expressed and purified from P. Pastoris as previously described [10].

Peptide mapping by high performance liquid chromatography (HPLC)-Tandem mass spectrometry
Peptide mapping was carried out as previously described [15]. In brief, a total of 5 mg (0.1 nmol) of purified M or Z a 1 AT in 100 mL of 50 mM Tris (pH 8) and 50 mM NaCl was mixed with 95 mL of 100 mM NaH 2 PO 4 (pH 2.4) followed by the addition of 5 mg of porcine pepsin dissolved in 0.05% (v/v) TFA and H 2 O for pepsin digestion. M or Z a 1 AT was digested for 5 min on ice. The digested sample was then injected into a micropeptide trap (Michrom Bioresources) connected to a C18 HPLC column (5 cm61 mm, Alltech) coupled to a LTQ linear ion-trap mass spectrometer (ThermoElectron). Peptic fragments were eluted using a gradient of acetonitrile (Burdick and Jackson) at a flow rate of 50 mL/min for a tandem mass spectrometry experiment to sequence each peptic fragment. Peptic fragments were identified by using the search algorithm SEQUEST (ThermoElectron) and manual inspection. Hydrogen/Deuterium Exchange A sample containing 5 mg (0.1 nmol) of M or Z a 1 AT in 50 mM Tris (pH 8) and 50 mM NaCl was diluted 24-fold with 50 mM Tris and 50 mM NaCl dissolved in D 2 O (Cambridge Isotope Laboratories) at 25uC to label the sample. The deuteration reaction was quenched at different time points by adding an equal volume of 100 mM NaH 2 PO 4 (pH 2.4) and quickly frozen in a dry ice2ethanol bath. Samples were stored at 280uC until use.

Isotope Analysis by HPLC2Electrospray Ionization Mass Spectrometry (ESI-MS)
The frozen sample was quickly thawed and digested with 5 mg of pepsin on ice for 5 min followed by immediate injection into a micropeptide trap connected to a C18 HPLC column coupled to a Finnigan LCQ quadrupole ion-trap mass spectrometer. Peptic peptides were eluted in 12 min using a gradient of 10-45% acetonitrile at a flow rate of 50 mL/min. The micropeptide trap and C18 HPLC column were immersed in ice to minimize back exchange. Because the mass of a peptic fragment increases by one for every amide hydrogen atom exchanged with deuterium, the amount of deuterium in each peptic fragment can be determined by comparing the mass of a labelled peptic fragment with the mass of the same peptide without the label. The centroid mass of each peptic fragment was determined using the software package MagTran. To correct for the back-exchange reaction of hydrogen atoms during pepsin digestion and HPLC2MS, a fully deuterated sample was prepared by incubating 5 mg of M or Z a 1 AT in 6 M guanidine deuterochloride, 50 mM Tris (pH 8) and 50 mM NaCl for 60 min at 25uC. The deuterium incorporation of each peptic fragment, corrected for the back exchange, was calculated using the following equation: D/N = [(m2m 0% )/(m 100% 2m)] where m is the mass of deuterated peptic fragment, m 0% and m 100% are the mass of the unlabeled and fully deuterated peptic fragments, respectively, N is the total number of exchangeable amide hydrogen atoms in each peptic fragment, and D is the number of amide hydrogen atoms incorporated in each peptic fragment.