First report of inherited protein S deficiency caused by paternal PROS1 mosaicism

Inherited protein S (PS) deficiency is a thrombophilic disorder with an autosomal dominant mode of inheritance. ¹ In plasma, 60% of PS exists in a complex with C4b binding protein and the remaining 40% in free form; only free PS functions as a cofactor for the activated protein C. Inherited PS deficiency is classified into types I, II, and III according to the level of PS activity or free and total PS antigen.² Types I and III have different phenotypes but they are considered to be caused by the same genetic abnormality. According to the Human Gene Mutation Database Pro (https://my. qiagendigitalinsights. com/bbp/view/hgmd/pro/all/php), 453 variants localized in the protein S gene (PROS1 gene) have been reported worldwide as of January 2021. Among these, missense/nonsense variants account for 61% (277 variants). However, genetic variants have been reported as undetectable in approximately half of the families with PS deficiency.³ Based on the Suita study, which examined the prevalence of PS deficiency in the Japanese general population, the prevalence is 1.12%;⁴ about 5-10 times higher than that in Europe and America (0.16–0.21%).⁵ PS deficiency is the most common inherited thrombophilia in Japan.^6,7 In this study, we identified a novel PROS1 variant and paternal mosaicism in a family with inherited PS deficiency in which both parents are asymptomatic. This is the first report of inherited thrombophilia due to parental mosaicism and highlights the importance of including information about potential mosaicism in genetic counseling.

The proband was a 34-year-old Japanese woman who was hospitalized with histiocytic necrotizing lymphadenitis. On admission to the hospital, total PS antigen (NSauto total protein S; NASCA Co., Ltd., Yokohama, Japan), free PS antigen (NSauto free protein S; NASCA Co., Ltd.), and PS activity levels (HemosIL PS-clot; I.L.Japan Co.,Ltd., Tokyo, Japan) were 55%, 25%, and <10%, respectively (Figure 1A). In the proband, acute inflammation caused by histiocytic necrotizing lymphadenitis may have increased vascular permeability, resulting in a decrease in free PS and a significant decrease in PS activity. The D-dimer concentration was significantly increased to 25.3 mg/mL (reference range, <1 mg/mL), but contrast-enhanced computed tomography showed no obvious thrombosis. In addition, the proband’s sister had already been diagnosed with PS deficiency after developing deep vein thrombosis in her leg at 25 years old and is on warfarin therapy (Figure 1A). Based on the above, the proband was diagnosed with type I PS deficiency. Both parents have normal levels of PS activity and antigen (Figure 1A), and the father had suffered a stroke at 62 years old and was taking edoxaban but the mother had no history of thrombosis.

This study was approved by the Ethics Committee for Human Genome and Gene Analysis Research at Kanazawa University (approval no. 450-5) and by the National Cerebral and Cardiovascular Center (approval no. M14-026). After obtaining informed consent from the proband and her family members, genomic DNA was extracted from peripheral blood leukocytes using a GENERATION Capture Column kit (QIAGEN, Tokyo, Japan). All exons and the exon-intron boundaries of PROS1 were polymerase chain reaction (PCR)-amplified using GoTaq G2 Hot Start Master Mixes (Promega, Tokyo, Japan). PCR products were sequenced using a BigDye Terminator v3.1 Cycle Sequencing kit (Thermo Fisher Scientific, Tokyo, Japan) and an Applied Biosystems 3500xL Genetic Analyzer (Thermo Fisher Scientific). Sanger sequencing indicated a novel heterozygous nonsense variant in exon 3 (NM_000313.4: c.246T>G; p.Tyr82*) in the proband and her sister. However, neither parent showed this variant by Sanger sequencing (Figure 1B). Since the probability of the same de novo variant occurring in the proband and her sister was considered virtually zero, this case was attributed to another cause such as the parents’ genetic mosaicism. We therefore attempted to quantify the variant allele ratio. First, digital PCR (dPCR) was performed using QuantStudio 3D Digital PCR Master Mix v2 (Thermo Fisher Scientific) and the QuantStudio 3D Digital PCR system (Thermo Fisher Scientific) (Figure 2). The variant allele ratio was 48.5% for the proband, 52.9% for the sister, 15.2% for the father, and 0.1% for the mother (Figure 2B).

Figure 1.Protein S antigen and activity levels of a protein S-deficient family and the results of Sanger sequencing. (A) Levels of total protein S (PS) antigen, free PS antigen, and PS activity in PS-deficient family members. The type of PS deficiency of the proband and sister were shown. (B) Pedigree and sequencing results. Arrow indicates the proband, and filled symbols indicate affected patients. The proband and her sister were heterozygous for c.246T>G (p.Tyr82*) whereas the parents seemed to have no causative variants. Data from the father showed a very small peak of G (black) with a normal peak of T (red). Abbreviations: RR: reference range; NA: not applicable; *, warfarin therapy.

Figure 2.Quantification of the variant allele ratio of blood-derived DNA using digital PCR. (A) Sequences of the primers and probes. (B) Blue indicates FAM fluorescence for the target variant allele; red indicates VIC fluorescence for the normal allele; green indicates FAM+VIC; and yellow indicates no amplification.

Second, we quantified the variant allele ratios using pyrosequencing with PyroMark Q24 (QIAGEN) to confirm the reproducibility of paternal mosaicism using a different principle of measurement (Figure 3). The results of pyrosequencing indicated that the variant allele ratios were 50% for the proband, 49% for the sister, 15% for the father, and 1% for the mother (Figure 3B). The variant allele ratios were consistent between dPCR and pyrosequencing for all family members, suggesting the presence of somatic mosaicism in the father. Based on this result, a careful review of the Sanger sequencing data of the father revealed the presence of a very small G peak with a normal T peak (Figure 1B).

Third, we evaluated the ratio of variant alleles in saliva- derived DNA to confirm the reproducibility of the somatic mosaicism presented in the blood-derived DNA of the father. The saliva-derived DNA was extracted and purified using Oragene Purifier (Kyodo International). The results of pyrosequencing showed that the variant allele ratios of the saliva-derived DNA were 50% for the proband, 50% for the sister, 17% for the father, and 2% for the mother (Figure 3C), confirming the reproducibility of the father’s mosaicism. Generally, the detection limits of dPCR and pyrosequencing are said to be more sensitive than that of Sanger sequencing: Sanger sequencing 15-20%;^8,9 pyrosequencing 5%;^8,10 dPCR 0.1-1%.^11,12 We performed pyrosequencing of five control subjects and an average variant allele ratio was 1% (data not shown). In addition, the detection limit of our pyrosequencing assay system calculated at 3d was 2% (data not shown). However, we could not examine the control subjects with our dPCR assay system. Based on the above, the variant allele ratios detected in the mother, 0.1% for dPCR and 1-2% for pyrosequencing, were below the detection limits of their respective assay system and were, therefore, considered to be background. In this case, the paternal mosaicism was observed in the blood cells (mesoderm origin) and salivary gland cells (ectoderm origin). This may suggest that the genetic variant might have occurred early in the fetal development process. However, the variant allele ratio of the two types of DNA collected in this study happened to be about 15%-17%, but the variant allele ratios in other organs have not yet been verified; this makes it difficult to determine when the variant occurred from the variant allele ratios. It is also assumed that the father's germ cells carry the variant allele, but the semen could not be collected and analyzed due to lack of consent. For a definitive diagnosis of gonadal mosaicism, it is necessary to prove the presence of mosaicism in the father's semen, but in this case, since the two children are heterozygous for the same variant, it is probable that the father has somatogonadal mosaicism.¹³

Figure 3.Quantification of variant allele ratios in blood- and saliva-derived DNA using pyrosequencing. (A) Sequences of the primers. The results of variant allele ratios quantified by pyrosequencing of blood-derived DNA (B) and saliva-derived DNA (C) from the proband and family members. The gray area shows the allele ratio of A and C at the variant locus. Since pyrosequencing was performed using reverse primers, the allele ratio of C represents the variant allele ratio.

In recent years, some cases of various diseases that were originally considered to be caused by de novo variants have actually been attributed to parental somatic or gonadal mosaicism.^13-15 Since the sensitivity of variation detection in Sanger sequencing, considered the standard for evaluating PROS1 variants, is limited to approximately 15-20%,^8,9 heterozygous variants can be detected but mosaicisms are likely to be undetectable. Thus, another more sensitive method, such as dPCR or pyrosequencing, is needed for the genetic evaluation of parental mosaicism. In this study, we detected a mosaic variant in the father of two heterozygous sisters with PS deficiency. This is the first case of parental mosaicism in an inherited PS-deficient family, suggesting that parental mosaicism may exist among some patients who have been thought to have de novo variants of inherited thrombophilia. It has been reported that approximately 20% of sporadic cases of hemophilia A, as determined by traditional Sanger sequencing methods, have an asymptomatic mosaic mother.¹⁵ The percentage of parental mosaicism present in sporadic cases of inherited PS deficiency needs to be determined by accumulating more cases in the future.

In conclusion, in cases of autosomal-dominant inherited thrombophilia, when the causative variant is detected only in the proband and a de novo variant is suspected, it is necessary to investigate the presence of parental mosaicism. For this purpose, it is very informative to quantify the variant allele ratio using multiple readily accessible specimens, such as blood, saliva, hair follicles, and urine to confirm the parental mosaicism. Furthermore, semen analysis should be performed as genetic screening for germline mosaicism in fathers of patients with de novo variants, and this analysis could be useful for prenatal counseling. In sporadic cases of inherited diseases, even if neither parent has the same variant using traditional sequencing methods, confirming the mosaicism of the parents is very important regarding the recurrence risk for the patient’s siblings and is essential information for genetic counseling.

Footnotes

• Received June 23, 2021
• Accepted October 1, 2021

Correspondence

*SN and KM contributed equally as co-first authors

ERIKO MORISHITA - eriko86@staff.kanazawa-u.ac.jp

Disclosures: no conflicts of interest to disclose.

Contributions: SN: investigation, visualization and writing, original draft; KM: investigation, visualization and writing, original draft; SN and KM are co-first authors. AW: methodology and supervision; MH: investigation; YI: investigation; YH: resources; SH: methodology, investigation, formal analysis; KK: investigation, formal analysis, funding acquisition, supervision; EM: conceptualization, funding acquisition, supervision, writing-review and editing.

Funding

this work was partially supported by Grants-in-Aid from the Ministry of Health, Labour and Welfare of Japan to EM (grant number 20FC0201) and KK (grant number 20FC1024).