846 Thromb Haemost 2002; 87: 846 –53 © 2002 Schattauer GmbH, Stuttgart Prothrombin G20210A is a Bifunctional Gene Polymorphism Angela M. Carter 1 , Mythily Sachchithananthan 2 , Stan Stasinopoulos 2 , Fabienne Maurer 3 , Robert L. Medcalf 2 1 Academic Unit of Molecular Vascular Medicine, University of Leeds, United Kingdom; 2 Monash Department of Medicine, Box Hill, Victoria, Australia; 3 Department of Medical Genetics, Central Hospital of the University of Vaud, Lausanne, Switzerland Keywords Prothrombin, G20210A, mRNA stability, thrombosis Summary The G20210A polymorphism has been shown to alter the efficiency of prothrombin mRNA processing. Here we show that the G20210A mutation also alters prothrombin mRNA stability. Three-fold more prothrombin protein and mRNA were produced in NIH-3T3 cells trans- fected with the prothrombin cDNAs containing the 20210A variant compared to cells expressing the 20210G variant. mRNA stability assays using chimeric globin transcripts harboring the G or A variant of the 97 nt prothrombin 3’-UTR indicated that the 20210G variant conferred greater instability to the globin reporter transcript than the A variant in transfected HepG2 cells. Both variants of the prothrombin 3’-UTR were shown to provide binding sites for a number of cellular proteins including HuR, an RNA binding protein associated with mRNA stability. Our results indicate that the G20210A is a bifunc- tional polymorphism, as it not only alters the efficiency of mRNA processing, but also the decay rate of prothrombin mRNA. Introduction The prothrombin G20210A polymorphism is located in the 3’-un- translated region (3’UTR) of the prothrombin mRNA, proximal to the poly (A) addition site (1). The 20210A allele has been associated with increased risk for venous thrombosis in a number of studies, and is the most common genetic risk factor for venous thrombosis in the Spanish population (2). In addition, there is a significant association of G20210A with plasma levels of prothrombin activity and antigen, with the highest levels reported in individuals homozygous for the 20210A allele and the lowest levels in those homozygous for the 20210G allele. In the Leiden Thrombophilia Study, the association of the 20210A allele with thrombosis was accounted for entirely by its association with plasma prothrombin levels (1). Soria et al. (3) confirmed an asso- ciation of the 20210A allele with a 3-fold increased risk of thrombosis and an association with plasma prothrombin. Furthermore, using a combined linkage/disequilibrium analysis, they demonstrated that the G20210A polymorphism accounted for the majority of variation in plasma prothrombin that can be attributed to the prothrombin gene locus, supporting the suggestion that it is a functional polymorphism and not in linkage disequilibrium with an, as yet unidentified, poly- morphism elsewhere in the prothrombin gene locus. The location of this polymorphism at the last residue of the pro- thrombin 3’UTR suggests that it may be influencing prothrombin levels via a mechanism related to prothrombin mRNA stability, mRNA pro- cessing or protein translation. A recent study by Gehring et al. (4) has indicated that this mutation does in fact alter the efficiency of pro- thrombin mRNA processing. In this current study, we sought to deter- mine whether the G20210A polymorphism also influences prothrombin protein secretion and mRNA stability independently of changes in prothrombin mRNA processing. Methods Cell Culture Mouse NIH-3T3 cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM: Life Technologies), supplemented with 10% (v/v) heat inactivated foetal calf serum (HI-FCS), 2 mM glutamine, and streptomycin and penicillin, in a humidified atmosphere at 37° C with 5% CO 2 as previously described (5) HepG2 cells were also maintained in DMEM containing 10% fetal calf serum, but were supplemented with 1X non-essential amino acids (Sigma, MO, USA), and 1 mM sodium pyruvate. Plasmids To assess the influence of the prothrombin G20210A polymorphism on prothrombin secretion in vitro, expression vectors containing full-length wild-type (20210G) and mutant (20210A) prothrombin cDNA variants were generated. The prothrombin cDNA (6) inserted into the Pst I site of pBR 322 is missing part of exon 1 which includes 31 bp of 5’UTR and the ATG codon of the prothrombin lead sequence. An extended oligonucleotide including this missing sequence (5’ – TTA CTA GTC GAC AAG CTT AAT TCC TCA GTG ACC CAG GAG CTG ACA CAC TAT GGC GCA CGT CCG AGG CTT GCA GCT GCC TGG CTG CC – 3’) was used in conjunction with oligonucleotides directed to the 3’UTR with the relevant base substitution (underlined residue): 20210G: 5’ – ATG TAT ACG CGG CCG CCG CTG AGA GTC ACT TTT ATT G – 3’; 20210A: 5’ – ATG TAT ACG CGG CCG CTG CTG AGA GTC ACT TTT ATT A – 3’, in order to amplify each variant of full-length pro- thrombin cDNA using Vent polymerase (New England Biolabs, MA, USA). The prothrombin PCR fragments (both variants) were inserted into the multiple cloning site (5’ – Sal I; 3’ – Not I) of the pCI neo mammalian expression vector (Promega, Madison, WI, USA) to produce vectors pCMV-PT-WT and pCMV-PT-Mut (see Fig. 1 panel A). The prothrombin cDNA inserts were veri- fied by sequencing. These vectors express prothrombin mRNA using the SV40 polyadenylation site in the pCI-neo vector, hence the prothrombin mRNA produced contains additional sequence at the 3’end due to the use of this poly- adenylation site and is also cleaved and processed downstream of this position. Using this approach, the two prothrombin mRNA variants undergo identical cleavage and polyadenylation reactions. This approach was taken to allow us to study the effect of the polymorphism on prothrombin mRNA stability inde- pendent of any additional effects this mutation may have on mRNA processing. Correspondence to: Dr. Robert Medcalf, Monash Department of Medicine, Clive Ward Centre, Box Hill Hospital, Box Hill, Victoria, Australia – Tel.: +61 3 9895 0318; Fax: +61 3 9895 0332; E-mail: robert.medcalf@med. monash.edu.au