Genotype–phenotype gradient of <i>SERPINC1</i> variants in a single family reveals a severe compound antithrombin deficiency in a dead embryo

Abstract

Antithrombin is a key endogenous anticoagulant.1 Congenital antithrombin deficiency, mainly but not only due to serpin family C member 1 (SERPINC1) variants, is a major thrombophilia significantly associated to early onset venous thromboembolism (VTE).2 The fact that total absence of antithrombin is lethal in animal models,3 suggests that the control of thrombin might be crucial during embryonic development. However, the role of antithrombin deficiency in embryonic states in humans has only been suggested by indirect observations. The analysis of natural mutants of antithrombin constitutes the ideal platform to investigate the potential association of antithrombin deficiency and embryonic lethality in humans. From a large cohort of 346 unrelated subjects with antithrombin deficiency (plasma antithrombin activity <80%) recruited by our group over 21 years (1998–2019), we identified a thrombophilic family with antithrombin deficiency and a strong history of fetal and neonatal deaths. Plasma antithrombin was analysed by functional anti-activated Factor X (FXa) and immunological methods (enzyme-linked immunosorbent assay, crossed-immunoelectrophoresis, and Western blot). Sanger sequencing of SERPINC1 was performed in peripheral blood mononuclear cells, as well as in donated tissue from an early miscarriage. The clinical and molecular characterisation of the family is shown in Fig 1. The index case was a 32-year-old woman (II-3) with VTE and two miscarriages (G4:P2:A2:L2), who had type I (quantitative) antithrombin deficiency, but traces of aberrant disulphide-linked dimers in plasma. Antithrombin deficiency was caused by an intronic variant (c.1154-14G>A) of paternal origin, who also had severe type I deficiency and recurrent VTE despite oral anticoagulation. Her sister, also with antithrombin deficiency, died from a venous thrombotic event at the age of 22 years. The pathogenic mechanism of this variant was suggested by Jochmans et al.4 (1994) and fully characterised by our group in unrelated patients carrying the same pathogenic variant. The intronic point mutation created a new cryptic acceptor sequence that led to an in-frame insertion of four amino acids at the loop connecting hI and s5A, which induced polymerisation of variant antithrombin.5 The second partner of the patient (II-4) was an asymptomatic heterozygous carrier of the Budapest 3 variant (p.Leu131Phe), a type II (qualitative) deficiency affecting the heparin binding site of antithrombin (type II/HBS). This family branch also had interesting information concerning the clinical consequences of this mutation. Whilst heterozygous carriers of p.Leu131Phe had no history of VTE (n = 7), a homozygous Budapest 3 newborn (III-7) had severe antithrombin deficiency and died from spontaneous aortic thrombosis. Finally, the proband had reported two early miscarriages. Adequate evaluation and management of these high-risk pregnancies was performed. Notably, additional adverse obstetric conditions, as well as the lack of adherence to therapeutic-dose low-molecular-weight heparin, were excluded. Interestingly, genotypic characterisation of the second dead embryo (III-4), the only available for analysis, revealed a double heterozygosis for the type I (c.1154-14G>A) and the type II/HBS (Budapest 3) pathogenic variants. Although antithrombin deficiency is considered the most severe thrombophilia, it certainly constitutes a highly heterogeneous disorder. Heterozygous type I mutations, associated to no circulating variant antithrombin, are normally associated to the classic major thrombophilic phenotype, with early onset, idiopathic and recurrent VTE.6 On the other hand, heterozygous type II genetic variants, leading to the secretion of defective proteins, are more heterogeneous, but they tend to cause milder phenotypes, especially in case of type II/HBS deficiency. The best example is antithrombin Budapest 3 (p.Leu131Phe), which is compatible with nearly normal function in progressive activity assays (with no heparin) and has been shown to have null effect in the beta glycoform, a N-glycan minority fraction of antithrombin that however has higher heparin affinity.7 These observations explain the mild risk of VTE associated to this variant in the heterozygous state, as well as the existence of homozygous subjects.8 Moreover, few additional homozygous cases have been described carrying antithrombin Toyama (p.Arg79Cys), another type II/HBS deficiency,2 or the p.Phe261Leu, a type II pleiotropic variant whose unstable conformation makes it prone to polymerisation.9 Remarkably, these cases mostly have very severe thrombotic phenotype, with very early and recurrent events.2 Exceptionally, three viable cases of combined type I and type II/HBS mutations have been reported worldwide in two families, all again with early and life-threatening events.10 However, no patient homozygous for type I variants has been described to date, suggesting the 100% embryonic lethality associated to this condition. Unfortunately, we were not able to determine antithrombin levels in the dead embryo but, based on previous reports, we can assume that the combination of these type I and type II/HBS variants may lead to an extremely severe deficiency. The combination identified in the present study might be aggravated by a potential dominant-negative effect of the type I variant on the type II/HBS one through disulphide-bond-mediated polymerisation. Animal models also contribute to evidence this genotype–phenotype gradient concerning antithrombin defects and show the key role of antithrombin during embryonic development. Thus, antithrombin knockout mice die at 16·5 gestational days because of intravascular coagulation.3 Additionally, disruption of antithrombin locus in zebrafish results in spontaneous venous thrombosis in larvae, and although homozygous mutants survive into early adulthood, they eventually succumb to massive intracardiac thrombosis.11 Interestingly, homozygous mice for antithrombin Toyama develop spontaneous thrombosis, occurring as early as the day of birth, while heterozygous mice may require additional triggering factors to develop thrombosis.12 Our present study provides new evidence in humans that extremely severe antithrombin deficiency might cause embryonic lethality. The potential functions of antithrombin during the embryonic state, which may be also beyond haemostasis, should be further characterised. In conclusion, we have characterised an exceptional thrombophilic family, with an excellent genotype–phenotype gradient that covers, and perfectly illustrates, the whole antithrombin deficiency continuum, that is, the clinical gradient that may be found among carriers of different types and combinations of SERPINC1 defects (Fig 2): from mild heterozygous type II/HBS deficiency (mostly asymptomatic), to severe type I and very severe homozygous type II/HBS deficiencies (with early and neonatal thrombosis respectively) and, finally, to the extremely severe double heterozygosis for these type I and II/HBS variants, which may have led to embryonic lethality. Thus, we provide new evidence in humans that a severe antithrombin deficiency might cause embryonic lethality and highlight the role of antithrombin in embryonic development. A. Palomo and L. Entrena provided samples and clinical data. C. Bravo-Pérez, M. E. de la Morena-Barrio, B. de la Morena-Barrio, J. Padilla, A. Miñano, E. Navarro and R. Cifuentes performed experiments, analysed data, made figures and wrote the paper. J. Corral and V. Vicente designed and supervised the work, discussed the results and wrote the paper. All authors revised the manuscript. The authors declare no conflict of interests. The study was funded by PI18/00598 (ISCIII&FEDER) and 19873/GERM/15 (Fundación Séneca). M. E. de la Morena-Barrio has a postdoctoral contract from University of Murcia. B. de la Morena-Barrio has a pre-doctoral fellowship from Fundación Séneca and R. Cifuentes has a PFIS pre-doctoral fellowship from ISCIII&FEDER.

Topics & Concepts