Are thrombophilia more multifactorial than we thought: report of mosaicism for FII G20210A and novel FII T20061C gene variants


Valentina Djordjevic, Institute of Molecular Genetics and Genetic Engineering, Vojvode Stepe 444A, PO Box 23, 11010 Belgrade, Serbia.
Tel.: +381 11 3976658; fax: +381 11 3975808.

Mosaicism implies the presence of more than one genetically distinct cell line in a single organism. Somatic mosaicism has been implicated in more than 30 monogenetic disorders, but its possible role in multifactorial diseases is still unresolved [1,2]. Thrombophilia is a multifactorial disorder, involving both genetic and acquired risk factors that lead to increased tendency to thrombosis. The most frequent thrombophilic genetic risk factors are FV Leiden and FII G20210A gene variants [3]. The FII G20210A (c.*97G>A) variant is associated with elevated plasma prothrombin levels and an almost 3-fold increased risk of venous thrombosis [3,4].

Here we report a 48-year-old female patient diagnosed with recurrent pulmonary thromboembolism (PTE) and pulmonary artery hypertension. She was admitted to hospital because of chest pain, tachycardia, dyspnea and fatigue 2 months after introduction of hormonal replacement therapy. After a thorough diagnostic examination, including CT angiography and perfusion scintigraphy, she was diagnosed with PTE and anticoagulant therapy was introduced. Two weeks after discharge, the patient suffered another episode of PTE, with subtherapeutic INR at admission. During the second hospitalization she developed pulmonary artery hypertension. The patient had three uncomplicated pregnancies and myoma uteri diagnosed at the age of 42. The family history was markedly positive: her father had two ischemic strokes and died of myocardial infarction (MI), her grandfather and father’s brother also died of MI. Routine thrombophilia screening, with the exception of protein C and protein S levels (due to vitamin K antagonists therapy), was performed 6 months after thromboembolic episodes. All tested parameters, including antithrombin activity, FVIII level, lupus anticoagulant, anticardiolipin antibodies IgG and IgM and antibeta2 glycoprotein I antibodies IgG and IgM, as well as homocysteine level, were within the normal range. Two samples (peripheral blood and buccal swab) were referred for genetic testing for thrombophilia (FV Leiden and FII G20210A). Mutations were detected by the PCR-RFLP method, using the routine procedure [5]. Analysis for FV Leiden mutation was negative in both samples, but contradictory results for FII G20210A gene variant were obtained. To explore the inconclusive results for G20210A testing, new samples from different tissues (peripheral blood, buccal swab and hair roots) were provided and analyzed by the PCR-RFLP method and capillary electrophoresis using a DNA 500 Assay kit on an Agilent 2100 Bioanalyzer (Agilent Technology, Santa Clara, CA, USA). We also performed sequencing of the 715 bp segment of the 3′end of the prothrombin gene (primers: 5′-GGAAACGAGGGGATGCCTGT-3′ and 5′-GTGAGAG GAAAGATGGCAGG-3′) using the BigDye Terminator Version 3.1 Ready Reaction Kit (Applied Biosystems, Foster City, CA, USA) on a 3130 Genetic Analyzer (Applied Biosystems). The presence of the heterozygous G20210A variant in buccal swab cells shown by PCR-RFLP analysis, was confirmed by sequencing (Fig. 1A2, B2). On the other hand, no sequence abnormalities were detected in blood (Fig. 1A1,B1) and hair bulb (Fig. 1A3,B3) as both samples carried the normal G20210 gene variant. Sequencing data also revealed the presence of a new T20061C gene variant (c.1817T>C), which was present only in the buccal swab sample (Fig. 1C2). This novel variant is located in the last exon of the prothrombin gene, and leads to the replacement of valine for alanine at position 606 in the protein (nomenclature according P00734 UniProt). We performed genotyping using an AmpF STR Identifiler Kit (Applied Biosystems) according to the manufacturer’s instructions on a 3130 Genetic Analyzer (Applied Biosystems). Analyses of 15 microsatellite loci and amelogenin marker by GeneMapper software excluded the possibility of contamination of the buccal swab sample and confirmed that the tested samples were derived from the same person.

Figure 1.

 PCR-RFLP analysis by capillary electrophoresis (A), and part of a sequence with FII G20210A (B) and FII T20061C polymorphisms (C): (1) peripheral blood samples, (2) buccal swab samples, (3) hair root samples. Green, lower marker; black, sample; purple, upper marker. See figure online for colour.

We cloned the 715 bp PCR fragment of the 3′ end of the prothombin gene, which included both T20061C and G20210A gene variants (primers: 5′-GGAAACGAGGGGATGCCTG T-3′ and 5′-GTGAGAGGAAAGATGGCAGG-3′) in the pG EM-T Easy vector (Promega, Madison, WI, USA) according to manufacturer’s instructions. Sequencing of these clones (primer: 5′-GGAAACGAGGGGATGCCTGT-3′), using the BigDyeTM Terminator Version 3.1 Ready Reaction Kit (Applied Biosystems) on a 3130 Genetic Analyzer (Applied Biosystems), revealed that both variants are located on the same DNA strand.

The obtained results suggested the presence of somatic mosaicism for G20210A and a novel T20061C gene variant in our patient. We also tested peripheral blood and buccal swab samples from available family members (mother, brother and two of the patient’s children) and neither of these two gene variants was detected.

To the best of our knowledge this is the first report of mosaicism for the FII G20210A gene variant. Our finding raises the question of interpretation of genetic testing data and the patient’s future medical care, as well as more broad-spectrum questions regarding the role of mosaicism in the pathogenesis of thrombophilia and multifactorial diseases in general.

Published studies, including ours, report that carriers of the FII G20210A gene variant have an increased risk of developing isolate PTE [6,7]. Nevertheless, our patient is also a carrier of a novel T20061C gene variant. The frequency of this variant, its potential role in the pathogenesis of thrombophilia and interplay with the G20210A variant are unknown and require additional studies on a larger cohort of patients.

If we had only tested a peripheral blood sample from our patient, we would not have been able to detect mosaicism. Negative results for genetic testing in multifactorial diseases are common, and do not lead to the testing of new samples. Could the choice of sample type influence the result of testing; for example, should genetic testing for the FII G20210A variant include only samples with the same embryonic origin as hepatocytes, because the human prothrombin gene is predominantly expressed in liver [8]? This is not the case with peripheral blood leukocytes, which are one of the most commonly used samples for DNA analysis [9].

Current knowledge about the potential role of mosaicism in the pathogenesis of multifactorial diseases other than cancer is very limited and this area requires further studies. Recently, Thibodeau et al. [10] reported a potential role for somatic mosaicism in the pathogenesis of lone atrial fibrillation. There are also data regarding an association between a certain type of neurodegeneration and mosaicism [11]. Gottlib et al. [12] proposed a hypothesis in which selection of pre-existing mosaic gene variants in certain disease-susceptible tissues, due to changes in the microenvironment, can lead to onset of common multifactorial diseases, such as cardiovascular disorders or diabetes.

Could somatic mosaicism be one of the missing pieces from the complex genotype/phenotype puzzle in multifactorial diseases? We speculate that this might be the case, but future studies are needed to investigate this hypothesis. The large-scale DNA analysis systems (next generation sequencing or ultra-deep pyrosequencing) available nowadays, open the possibility to detect genetic mosaicism by sampling a large number of cellular genomes from the same proband and could provide the answer to the question regarding the role of mosaicism in the pathogenesis of multifactorial disease, including thrombophilia.


This work was supported by grant 173008 from The Ministry of Education and Science, Republic of Serbia.

Disclosure of Conflict of Interests

The authors state that they have no conflict of interests.