Alpha₁-antitrypsin is a glycoprotein of the serpin (serine protease inhibitor) protein superfamily (Silverman et al. ²⁰⁰¹). It is synthesized within hepatocytes and secreted into the circulatory system, where it is the most abundant human plasma proteinase (Gooptu and Lomas ²⁰⁰⁹; Travis and Salvesen ¹⁹⁸³). The primary function of α₁-antitrypsin is the inhibition of neutrophil elastase. Naturally occurring point mutations within the protein such as the Z (Glu342Lys) variant cause it to undergo conformational rearrangement and self-associate into polymer chains (Lomas et al. ¹⁹⁹²). The accumulation of α₁-antitrypsin polymers within the endoplasmic reticulum (ER) is associated with cellular toxicity and development of liver diseases such as cirrhosis and hepatocellular carcinoma (Eriksson et al. ¹⁹⁸⁶; Perlmutter ²⁰⁰²). Polymerisation abolishes functional activity of α₁-antitrypsin and so allows dysregulated elastase activity within the lung, causing emphysema (Elliott et al. ¹⁹⁹⁸). Moreover, extracellular α₁-antitrypsin polymers are found within the lung in patients with α₁-antitrypsin deficiency related emphysema, and are likely to have pro-inflammatory effects. The combination of loss- and gain-of-function consequences of α₁-antitrypsin polymerisation mean there is a need for a strategy that prevents polymerisation and maintains enzyme inhibitory activity (Gooptu et al. ²⁰⁰⁹). While some small molecules have shown polymerisation blocking effects in vitro, so far they have all abolished the functional activity of α₁-antitrypsin (Mahadeva et al. ²⁰⁰²; Mallya et al. ²⁰⁰⁷). Moreover, there is a lack of structural data defining interactions of the protein with promising lead compounds, preventing their rational modification. Solution NMR is well-suited to provide a residue-specific characterisation of this conformationally heterogeneous protein and to probe its interactions with putative polymerisation blockers. We report here the ¹H, ¹⁵N and ¹³C backbone assignments of 45 kDa α₁-antitrypsin as a crucial step not only for detailed characterization of this highly dynamic molecule but also towards the application of drug discovery techniques to α₁-antitrypsin deficiency.

cDNA coding for 6xHis-tagged α₁-antitrypsin within a pQE31 plasmid was transformed into BL21-Gold (DE3) E. coli (Stratagene). Cells were grown in M9 minimal medium (H₂O or D₂O) at 37°C with 1 g/l of ¹⁵NH₄Cl (Spectra Stable Isotopes) and 2 g/l of either glucose or ¹³C-glucose (Sigma-Aldrich) as the sole nitrogen and carbon sources, respectively. Following induction of protein expression at OD₆₀₀ = 0.6 using isopropyl β-D-1-thiogalactopyranoside (IPTG), cells were incubated at 30°C for 8 h. The cells were collected by centrifugation, resuspended in 10 mM sodium phosphate buffer, pH 8.0, 0.5 M NaCl, 20 mM imidazole, and lysed using a cell disruptor. The supernatant of the cell lysate was loaded onto a HisTrap Crude FF column (5 ml; GE Healthcare) and after washing to baseline with the same buffer, the bound protein was eluted with an imidazole gradient (20 mM – 200 mM). The fractions containing α₁-antitrypsin were collected, dialyzed 10 mM Tris, pH 8.0, 1 mM EDTA, and 1 mM β-mercaptoethanol and loaded onto a HiTrap-Q Sepharose column (5 ml; GE Healthcare). Bound protein was then eluted with a 0–0.5 M NaCl gradient. The fractions containing α₁-antitrypsin were pooled, dialysed against 25 mM Na₂HPO₄, 50 mM NaCl and 1 mM EDTA at pH 8.0. For complete back exchange of buried backbone amides in ²H/¹⁵N/¹³C labeled α₁-antitrypsin, the protein (1 mg/ml) was unfolded in storage buffer containing 8 M urea for 1 h at room temperature. It was subsequently refolded by dropwise addition to 5 L storage buffer with continuous stirring at 4°C. Precipitated material was removed by centrifugation, and the conformational integrity of the refolded protein was confirmed by non-denaturing and SDS-PAGE and inspection of its ¹⁵N TROSY-HSQC spectrum.

NMR spectra were collected from α₁-antitrypsin at a typical concentration of 225 μM in 25 mM Na₂HPO₄, 50 mM NaCl, 1 mM EDTA and 0.001% 4,4-dimethyl-4-silapentane-1-sulfonic acid (DSS) at pH 8.0 or pH 7.0 (10% D₂O/90% H₂O). Spectra from a sample prior to refolding at pH 8.0 were recorded on a Bruker Avance III 700 MHz spectrometer equipped with an HCN cryoprobe. TROSY (Pershuvin et al. ¹⁹⁹⁷) versions of HNCO, HN(CA)CO, HNCA, HN(CO)CA, HNCACB and HN(CO)CACB spectra (Eletsk et al. ²⁰⁰¹; Salzmann et al. ¹⁹⁹⁸, ¹⁹⁹⁹) were acquired with constant-time chemical shift evolution in the ¹⁵N dimension and ²H decoupling (1 kHz WALTZ-16) (Shaka et al. ¹⁹⁸³) while magnetisation was transverse on CA/CB. Spectra from refolded samples at pH 8.0 and pH 7.0 were acquired on Varian Inova 600, 800 and 900 MHz spectrometers, all equipped with HCN cold-probes. TROSY-HNCA (Yang and Kay ¹⁹⁹⁹) spectra were acquired with ²H decoupling as above and constant-time chemical shift evolution in the ¹³C indirect dimension. All spectra were processed using NMRPipe (Delaglio et al. ¹⁹⁹⁵) and analysed using the CCPN software suite (Fogh et al. ²⁰⁰²).

The sequential backbone resonance assignments of α₁-antitrypsin at pH 8.0 and pH 7.0 have been deposited in the Biological Magnetic Resonance Data Bank (http://www.bmrb.wisc.edu) under the accession number 17804. The ¹⁵N-HSQC spectrum of 45 kDa α₁-antitrypsin shows excellent cross-peak dispersion (Fig. 1). With 394 residues, α₁-antitrypsin is to date one of the largest proteins for which the backbone has been assigned to near completion (Supp Fig. 1). The high molecular weight rendered the backbone assignment a challenging task and this was therefore carried out in successive stages. Initially (stage 1), native ²H/¹⁵N/¹³C-labelled antitrypsin was used, followed by the use of refolded (see Methods) ²H/¹⁵N/¹³C antitrypsin samples at pH 8.0 (stage 2) and pH 7.0 (stage 3). The majority of the secondary-structure elements were assigned in stage 1, except for resonances corresponding to the core β-sheet B and also the loop regions (Supp Fig. 2a), which were assigned during stages 2 (Supp Figs. 2b and 3a) and 3 (Supp Figs. 2c and 3b), respectively. Aside from the additional expected cross-peaks, the ¹⁵N-TROSY-HSQC spectra before and after refolding were identical.

82.8 and 79.3% of backbone amide resonances were assigned at pH 8.0 and pH 7.0, respectively. 84.9, 88.3 and 75.1% of the CO, CA and CB resonances, respectively, were assigned at pH 8.0. At pH 7.0, 82.5% of the CA resonances were assigned. The secondary structure of a1-antitrypsin was predicted from the assigned chemical shifts at pH 8.0 using DANGLE (Cheung et al. ²⁰¹⁰). The measured CA and CB chemical shifts were first corrected for deuterium isotope shifts (Venters et al. ¹⁹⁹⁶; Gardner et al. ¹⁹⁹⁷). The predicted secondary structure is broadly consistent with that observed crystallographically (Elliott et al. ²⁰⁰⁰; Fig. 2).

M. P. Nyon, J. Kirkpatrick contributed equally.

We are grateful for the use of the Biological NMR Centre at UCL, the Henry Wellcome Building Biomolecular NMR Facility, University of Birmingham and the MRC Biomedical NMR Centre (Mill Hill). This work was supported by funding from the Wellcome Trust, the Medical Research Council (UK), the Human Frontiers Science Programme, the Biotechnology and Biological Sciences Research Council, Papworth Hospital NHS Trust, and the Birkbeck College Faculty Research Fund.