Thrombomodulin Gene Mutations Associated With Myocardial Infarction
Background Thrombomodulin is an important receptor for thrombin on the endothelial cell surface of most blood vessels, including those of the heart. Thrombin-bound thrombomodulin activates protein C, which inhibits thrombin generation by degrading factors Va and VIIIa. The aim of this study was to analyze the 5′ region of the thrombomodulin gene to determine whether mutations contribute a risk for myocardial infarction.
Methods and Results We screened the promoter region of the thrombomodulin gene by single-stranded conformation polymorphism analysis in 104 patients with diagnosed myocardial infarction. Five mutations (three distinct) were identified (GG−9/−10AT, G−33A, and C−133A). The dinucleotide mutation GG−9/−10AT was identified in 3 individuals (2 heterozygous, 1 homozygous). Only one of the three different mutations was identified in 104 patient control subjects matched for age, sex, and race (G−33A in a single individual). All mutations identified were in close proximity to consensus sequences for transcription control elements within the thrombomodulin gene. In contrast, no difference was observed between patients and control subjects for the allelic frequency of a previously identified neutral polymorphism GCC/GTC coding for Ala/Val455, with 3 individuals homozygous for GTC (Val) in both groups.
Conclusions The findings suggest that mutations in the promoter region of the thrombomodulin gene may constitute a risk for arterial thrombosis.
Endogenous anticoagulant mechanisms regulate thrombin generation both by direct inhibition of thrombin (via the inhibitor antithrombin) and by inhibition of the thrombin positive-feedback mechanism (via the protein C inhibitory pathway).1 2 3 4 5 6 One protein of the protein C pathway not comprehensively studied in terms of its contribution to either venous or arterial thrombosis is thrombomodulin, an endothelial cell receptor for thrombin.
Natural deficiencies or variants of thrombomodulin have been difficult to detect to date by use of phenotypic analysis because of its endothelial-cell–membrane location. Direct screening of the gene in the absence of phenotypic information, however, has been used to identify three thrombomodulin gene mutations in patients with venous thrombosis.7 8 A region of the thrombomodulin gene 5′ to the coding region (promoter region) has several potential regulatory sequence motifs; a TATA box occurs 190 bp upstream of the methionine initiation codon.9 10 Farther upstream (83 bp) is a possible CAAT box and four possible transcription-factor Sp1 binding sites.9 Two studies have identified regions within the thrombomodulin gene promoter that have an associated loss of reporter gene–transcription activity when deleted.10 11 Because of the existing body of information regarding potentially important elements involved in the regulation of transcription of the thrombomodulin gene, it was of interest to investigate this region in a clinical study of arterial thrombosis (MI). The results of this first investigation are reported here.
We analyzed 104 patients admitted to west London hospitals with chest pain who were subsequently diagnosed by use of WHO criteria as having sustained an MI. Mean age of the patients was 60.5 years (range, 30 to 79 years). Nineteen patients were Asian, one was Afro-Caribbean, and the rest were white. Each patient was matched by sex, race, and age to a control subject. Control samples were collected from patients attending the outpatient department. The only criterion used for selection of patients as control subjects was that they had not suffered any thrombotic episodes. Written consent was given by these individuals after the study had been fully explained to them. The ethics committees of West Middlesex and Charing Cross Hospitals gave clearance for both the study of the patients after MI and patients for use as control subjects.
Genomic Amplification and Sequencing
Primers used for amplification of genomic DNA for screening and sequencing are shown in the Table⇓ along with details of polymerase chain reaction and sequencing reactions. The thrombomodulin promoter region was amplified in four overlapping fragments of ≈150 bp in length for screening of the patient group (tm1 through tm4). Fragments of this size previously have been shown to be optimal for mutation detection by SSCP (97% of mutations detected12 ). Two fragments, tm2 and tm3 (see the Table⇓), were amplified for the control group.
Analysis was performed with the use of a Phast System and premade 20% polyacrylamide gels (Pharmacia). A nondenaturing native buffer system was used and electrophoresis performed at 15°C. Gels were visualized with the use of a silver stain. Additional information is shown in the Table⇑. All potential band shifts were analyzed by direct sequencing of a separately amplified fragment (see the Table⇑). Any mutations found in patient samples were then sought in the control group, also by SSCP.
We calculated an estimate of relative risk using an odds ratio with the use of the Statistical Package for the Social Sciences for Microsoft Windows, release 6.1.
Band shifts were observed and mutations identified in the promoter region in 5 (4.8%) of the 104 individuals within the MI group. The 5 patients were between the ages of 53 and 70 years when they had their first MI. Interestingly, 4 of the 5 mutations identified were in 19 Asians (21% of Asians). One mutation was the substitution of two adjacent bases at positions −9/−10 (GG to AT), which was identified in 3 individuals (2 heterozygous [1 Asian, 1 white] and 1 homozygous [Asian]). In the latter case, we cannot rule out the possibility that only one allele was present, because there are no other informative polymorphic sites in the regions analyzed. The next occurred at position −33 (G to A, heterozygous) and the fourth at position −133 (C to A, heterozygous). The mutations were all present in fragments tm2 and tm3 (Table⇑). SSCP band shifts and sequencing autoradiographs for the different mutations are shown in Fig 1⇓. Only one of these mutations was identified in the control group (G−33A).
The risk of MI was therefore ≈5 times increased in patients with a mutation (odds ratio, 5.2; 95% confidence interval, 0.6 to 45.3). Because the prevalence of mutations was only 1% in the control subjects and 5% in patients with MI, a larger sample group would be required to achieve a more precise estimate of the relative risk. Similarly, comparisons by tests of association (for example, by χ2) would not be expected to reach statistical significance with the low prevalence of mutations.
We also screened for a previously identified common polymorphism within the coding region (C/T coding for Ala455Val13 ). With the use of SSCP, the polymorphic site in tm18 coding for Ala455Val was identified as a band shift in 33 of 104 individuals within both the patient and the control groups. Both alleles were detected in whites and Asians. Band shifts suggesting a homozygous T allele (Val) were sequenced in each case. Three individuals in both the MI and control groups were homozygous for GTC rather than GCC.
All of the 5′ region thrombomodulin gene mutations identified were very close to previously proposed transcription control elements (Fig 2⇓). Naturally occurring single-base substitutions within the short sequence from the transcription start site upstream to −40 have been identified in other genes coding for coagulation proteins, protein C, and factor IX and have been shown to be the cause of reduced protein levels.14 15 Two of the three different mutations in the present report were identified in the corresponding region of the thrombomodulin gene. Furthermore, reporter gene assays using deletion mutants of the thrombomodulin gene promoter have already demonstrated that the sequence −33 to −70 is important for transcription of CAT.10 One of the mutations identified in the current study in both a patient and a control subject occurs at the boundary of this region (G−33A) and is adjacent to the TATA box. Both cases of this mutation occurred in Asians, suggesting a polymorphic site within this population. In another deletion mutant study of reporter gene activity, the second and third potential Sp1 binding sites of thrombomodulin were shown to be important for transcription of CAT.11 The identified mutation at position −133 lies two bases from the second Sp1 binding site. The first and second Sp1 binding sites have only six intervening bases, and the mutation at position −133 also lies five bases from the first Sp1 binding site. There is currently no experimental evidence linking the −9/−10 GG to AT substitution to reduced gene transcription, but its proximity to the TATA box can be noted. The finding of this mutation in three patients subsequent to an MI but in none of the control subjects suggests that this mutation may constitute a particular risk for arterial thrombosis. The high odds ratio, calculated for all mutations, suggests an association between mutation and an increased risk of MI. However, because the confidence interval of the odds ratio is wide, this estimate of the relative risk is uncertain. Given the low prevalence of mutation, a larger study group is required to firmly establish and extend our findings.
The results demonstrate that mutation of the 5′ region of the thrombomodulin gene appears to be associated with MI. They suggest that in certain individuals, abnormal thrombomodulin gene expression may impair normal functioning of the protein C anticoagulant pathway, allowing increased thrombin to be generated in the myocardial circulation. This suggestion will require verification with appropriately designed investigation of coagulation system activation and regulation in carriers of mutations. Furthermore, direct experimental evidence for reduced transcription activity of mutant thrombomodulin constructs is desirable. It is interesting to note that Asians in particular appear to carry thrombomodulin promoter-region mutations. Individuals of south Asian descent who have a western lifestyle are known to have an increased risk of ischemic heart disease. A number of acquired and potentially genetic contributory factors (smoking, hypertension, and diabetes) have been suggested,16 17 and it is proposed that interactions between impaired coagulation regulation and these factors may determine clinical outcome.
Very recently, a study was reported of the C/T dimorphism (predicting the 455Ala to Val substitution) in survivors of MI.18 It was found that the C allele was significantly more frequent among patients than among control subjects, which raised the possibility that the C/T dimorphism may be a factor in the pathogenesis of MI. The results presented above, which show no difference between control subjects and patients matched carefully for age, sex, and race, do not support this proposal.
Selected Abbreviations and Acronyms
|CAT||=||chloramphenicol acetyl transferase|
|Sp1||=||stimulatory protein 1|
|SSCP||=||single-stranded conformation polymorphism|
This work was supported by grants from the British Heart Foundation and from the Special Trustees of Charing Cross Hospital and Medical School.
- Received February 12, 1997.
- Revision received April 14, 1997.
- Accepted April 15, 1997.
- Copyright © 1997 by American Heart Association
Esmon CT, Esmon NL, Harris KW. Complex formation between thrombin and thrombomodulin inhibits both thrombin-catalyzed fibrin formation and factor V activation. J Biol Chem. 1982;257:7944-7947.
Esmon NL, Carroll RC, Esmon CT. Thrombomodulin blocks the ability of thrombin to activate platelets. J Biol Chem. 1983;258:12238-12242.
Esmon CT. The regulation of natural anticoagulant pathways. Science. 1987;235:1348-1352.
Bourin MC, Ohlin AK, Lane DA, Stenflo J, Lindahl U. Relationship between anticoagulant activities and polyanionic properties of rabbit thrombomodulin. J Biol Chem. 1988;263:8044-8052.
Tsiang M, Lentz SR, Sadler JE. Functional domains of membrane-bound human thrombomodulin: EGF-like domains four to six and the serine/threonine-rich domain are required for cofactor activity. J Biol Chem. 1992;267:6164-6170.
Ohlin AK, Marlar RA. The first mutation identified in the thrombomodulin gene in a 45-year-old man presenting with thromboembolic disease. Blood. 1995;85:330-336.
Ohlin AK, Marlar RA. Mutations in the thrombomodulin gene associated with thromboembolic disease. Thromb Haemost. 1995;73:1096. Abstract.
Jackman RW, Beeler DL, Fritze L, Soff G, Rosenberg RD. Human thrombomodulin gene is intron depleted: nucleic acid sequences of the cDNA and gene predict protein structure and suggest sites of regulatory control. Proc Natl Acad Sci U S A. 1987;84:6425-6429.
Yu K, Morioka H, Fritze LM, Beeler DL, Jackman RW, Rosenberg RD. Transcriptional regulation of the thrombomodulin gene. J Biol Chem. 1992;267:23237-23247.
Tazawa R, Hirosawa S, Suzuki K, Hirokawa K, Aoki N. Functional characterization of the 5′-regulatory region of the human thrombomodulin gene. J Biochem Tokyo. 1993;113:600-606.
Spek CA, Greengard JS, Griffin JH, Bertina RM, Reitsma PH. Two mutations in the promoter region of the human protein C gene both cause type I protein C deficiency by disruption of two HNF-3 binding sites. J Biol Chem. 1995;270:24216-24221.
Reijnen MJ, Sladek FM, Bertina RM, Reitsma PH. Disruption of a binding site for hepatocyte nuclear factor 4 results in hemophilia B Leyden. Proc Natl Acad Sci U S A. 1992;89:6300-6303.
McKeigue PM, Miller GJ, Marmot MG. Coronary heart disease in South Asians: a review. J Clin Epidemiol. 1989;42:579-609.