New variants of alpha-1-antitrypsin: structural simulations and clinical expression.

Background: Alpha-1 antitrypsin deficiency (AATD) is characterized by reduced serum levels of the AAT protein and predisposes to liver and lung disease. The characterization at structural level of novel pathogenic SERPINA1 mutants coding for circulating AAT could provide novel insights into the mechanisms of AAT misfolding. The present study aimed to provide a practical framework for the identification and analysis of new AAT mutations, combining structural simulations and clinical data.

Methods: We analysed a total of five mutations (four not previously described) in a total of six subjects presenting moderate to severe AATD: Gly95Alafs*18, Val210Glu, Asn247Ser, Pi*S + Asp341His and Pi*S + Leu383Phe + Lys394Ile. Clinical data, genotyping and phenotyping assays, structural mapping, and conformational characterization through molecular dynamic (MD) simulations were developed and combined.

Results: Newly discovered AAT missense variants were localized both on the interaction surface and the hydrophobic core of the protein. Distribution of mutations across the structure revealed Val210Glu at the solvent exposed s4C strand and close to the "Gate" region. Asn247Ser was located on the accessible surface, which is important for glycan attachment. On the other hand, Asp341His, Leu383Phe were mapped close to the "breach" and "shutter" regions. MD analysis revealed the reshaping of local interactions around the investigated substitutions that have varying effects on AAT conformational flexibility, hydrophobic packing, and electronic surface properties. The most severe structural changes were observed in the double- and triple-mutant (Pi*S + Asp341His and Pi*S + Leu383Phe + Lys394Ile) molecular models. The two carriers presented impaired lung function.

Conclusions: The results characterize five variants, four of them previously unknown, of the SERPINA1 gene, which define new alleles contributing to the deficiency of AAT. Rare variants might be more frequent than expected, and therefore, in discordant cases, standardized screening of the S and Z alleles needs complementation with gene sequencing and structural approaches. The utility of computational modelling for providing supporting evidence of the pathogenicity of rare single nucleotide variations is discussed.