Antithrombin Glasgow II: alanine 382 to threonine mutation in the serpin P12 position, resulting in a substrate...

A female with recurrent thrombosis was found to have a functional abnormality of antithrombin, with a ratio of functional to immunological activity in plasma of approximately 50%. Crossed immunoelectrophoresis in the presence of heparin was normal, indicating an abnormality of the reactive site, rather than the heparin binding domain. Accordingly, the antithrombin was isolated by heparin-Sepharose chromatography: this produced a mixture of normal and variant antithrombin, as the patient was heterozygous for the abnormality. To remove the normal component, the antithrombin was passed through a column of thrombin-Sepharose. On sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE), prior to its application to thrombin-Sepharose, the antithrombin migrated as a single band with identical mobility to that of normal antithrombin. After thrombin-Sepharose, the purified variant component was proteolysed, and migrated as two components, one with a reduced and one with enhanced mobility under non-reducing conditions. This demonstrated that the variant was unable to form stable inhibitor-thrombin complexes and was cleaved in a substrate reaction with thrombin. One site of cleavage was unambiguously ascertained to be the Arg 393-Ser 394 reactive site bond, by NH2 terminal sequencing of the cleaved variant antithrombin: 10 steps beginning at the P1' position, Ser-Leu-Asn-Pro-Asn-Arg,..., were clearly identified. The mutation responsible for this defect was studied by polymerase chain reaction (PCR) amplification of exon 6 of the antithrombin gene and direct sequencing of the amplified product. The presence of both a G and A in the first position of codon 382, identified the mutation GCA to ACA, which results in the substitution of Ala 382 to Thr. This is identical to that reported for antithrombin Hamilton (Devraj-Kizuk et al, 1988), although antithrombin gene polymorphism analysis suggests that the antithrombin Glasgow II mutation has arisen independently. We have recently shown (Caso et al, 1991) that mutation at a nearby position, Ala 384 to Pro, also transforms another variant, antithrombin Vicenza/Charleville, into a substrate for thrombin. The present results with antithrombin Glasgow II suggest that all the alanine residues at the base of the reactive site loop in positions P12-10 may be important for the formation of a stabilized inhibitor-thrombin complex.