- Email: [email protected]
Multiple HABP2 variants in familial papillary thyroid carcinoma: Contribution of a group of “thyroid-checked” controls
European Journal of Medical Genetics xxx (2017) 1e7
Contents lists available at ScienceDirect
European Journal of Medical Genetics journal homepage: http://www.elsevier.com/locate/ejmg
Multiple HABP2 variants in familial papillary thyroid carcinoma: Contribution of a group of “thyroid-checked” controls pin a, Isabelle Szuster a, Benjamin Kern a, Lucie Coppin a, Pauline Romanet b, Michel Cre de ric Fre nois d, Jean-Louis Wemeau e, Florence Renaud c, Emmanuelle Leteurtre c, Fre f e Bruno Carnaille , Catherine Cardot-Bauters , Christine Do Cao e, Pascal Pigny a, * a Service de Biochimie « Hormonologie-M etabolisme-Nutrition & Oncologie », Centre de Biologie Pathologie, Centre Hospitalier R egional & Universitaire, F59037 Lille Cedex, France b ^pital de la Conception, AP-HM, 13385 Marseille Cedex 05, France Laboratoire de Biologie Mol eculaire, Ho c Service d’Anatomie Pathologique, Centre de Biologie Pathologie, Centre Hospitalier R egional & Universitaire, F-59037 Lille Cedex, France d Universit e de Lille 2, CHRU, Equipe RADEME, EA 7364, Centre Hospitalier R egional & Universitaire, F-59037 Lille Cedex, France e ^pital Claude Huriez, Centre Hospitalier R Service d’Endocrinologie, Ho egional & Universitaire, F-59037 Lille Cedex, France f ^pital Claude Huriez, Centre Hospitalier R Service de Chirurgie Endocrinienne, Ho egional & Universitaire, F-59037 Lille Cedex, France
a r t i c l e i n f o
a b s t r a c t
Article history: Received 20 September 2016 Received in revised form 15 December 2016 Accepted 7 January 2017 Available online xxx
A heterozygous germline variant in the HABP2 gene c.1601G > A (p.Gly534Glu), which negatively impacts its tumor suppressive activity in vitro, has been described in 4e14% of kindreds of EuropeanAmerican ancestry with familial papillary thyroid carcinoma (fPTC). But it is also found in z4% of Europeans and European/Americans from public databases that, however, did not provide information on the thyroid function of the controls. To get unbiased results, we decided to compare HABP2 genotypes of patients with fPTC with those of “thyroid-checked” controls. A control group consisting of 136 European patients who were thyroidectomised for medullary thyroid carcinoma and devoid of any histologically detectable PTC or follicular-deriving carcinoma was built. In parallel we recruited 20 patients with fPTC from eleven independent European kindreds. The entire coding region of HABP2 was analyzed by Sanger sequencing the germline DNAs of patients. Nucleotide variants were searched for by Snap Shot analysis in the controls. Two variants, c.1601G > A (p.Gly534Glu) and c.364C > T (p.Arg122Trp), were found in 2 and 3 patients at the heterozygous level respectively (minor allele frequency (MAF): 5.0% and 7.5%, respectively). In controls, the MAF was either similar for the c.1601G > A HABP2 variant (2.94%, ns) or significantly lower for the c.364C > T variant (0.73%, p ¼ 0.016). The Arg122 residue lies in the EGF-3 domain of HABP2 which is important for its activation but, however, superposition of the predicted 3D structures of the wild type and mutated proteins suggests that this variant is tolerated at the protein level. In conclusion, our data do not support the pathogenicity of the HABP2 c.1601G > A variant but highlight the existence of a new one that should be more extensively searched for in fPTC patients and its pathogenicity more carefully evaluated. © 2017 Elsevier Masson SAS. All rights reserved.
Keywords: Thyroid cancer Familial cancer Papillary HABP2 Novel variant
1. Introduction HABP2 encodes a hyaluronic acid-binding protein, also named Factor VII-activating protease (FSAP), which is known to play roles in coagulation and fibrinolysis through the activation of Factor VII and pro-urokinase. Recently HABP2 has been described as a new
culaire, Centre de * Corresponding author. Institut de Biochimie et Biologie Mole Biologie Pathologie, CHRU, F-59037 Lille cedex, France. E-mail address: [email protected] (P. Pigny).
susceptibility gene for familial papillary thyroid carcinoma (fPTC) based on the description of a heterozygous variant c.1601G > A (p.Gly534Glu) in one kindred with fPTC (Gara et al., 2015). The authors provided three arguments supporting the causal role of this variant in carcinogenesis and its classification as a mutation: (i) functionally, HABP2 has a tumor suppressor effect in epithelial cell lines which is reversed by the mutation (dominant negative effect); (ii) this mutation co-segregates with PTC in 6 patients on 2 generations; (iii) its prevalence is significantly higher in patients with apparently sporadic PTC (4.7%) versus controls (0.7%) from a
http://dx.doi.org/10.1016/j.ejmg.2017.01.001 1769-7212/© 2017 Elsevier Masson SAS. All rights reserved.
Please cite this article in press as: Kern, B., et al., Multiple HABP2 variants in familial papillary thyroid carcinoma: Contribution of a group of “thyroid-checked” controls, European Journal of Medical Genetics (2017), http://dx.doi.org/10.1016/j.ejmg.2017.01.001
2
B. Kern et al. / European Journal of Medical Genetics xxx (2017) 1e7
multiethnic database (Gara et al., 2015). However, the causal role of the HABP2 c.1601G > A mutation in fPTC has been quickly challenged by different studies that reported higher frequencies of this variant in controls from different databases, ranging from 2.9 to 4.6% in the European population (Zhou et al., 2015; Sponziello et al., 2015; Sahasrabudhe et al., 2016; Weeks et al., 2016; Tomsic et al., 2015, 2016; Zhao et al., 2015) thus invalidating the third criteria mentioned above. However, as recently stated by both the American College of Medical Genetics and Genomics, and the Association for Molecular Pathology (Richards et al., 2015), “population databases cannot be assumed to include only healthy individuals”. Indeed, databases never provide extensive information on any
putative disease or phenotype that may be present in population selected as controls. Interestingly, the prevalence of occult (clinically indolent) PTC varies from 3.7% to 11% in most autopsy studies in Europe (Seehofer et al., 2005; Tanriover et al., 2011). Therefore, in this specific context, the comparison of the prevalence of a nucleotide variant between patients with fPTC and “thyroid-unchecked” controls may be especially inaccurate and biased. To clarify this issue, we decided to retrospectively search for HABP2 variants in several kindreds with fPTC by nucleotide sequencing the 13 exons of the gene, and then to look for these variants by SnapShot in a group of controls who underwent thyroidectomy and were free of any PTC based on the pathological
Fig. 1. Trees of the 2 kindreds with germline HABP2 variants. Squares denote male members, circles denote female members, hatched symbols denote patients with PTC, slashed symbols denote deceased members, and grey symbols denote family members without the variant and without any thyroid cancer at ultrasonography. Empty symbols denote unexplored family members. HABP2 genotype is shown in blue caracters (upper right). TNM classification of the thyroid tumor is shown in black caracters (lower right). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Please cite this article in press as: Kern, B., et al., Multiple HABP2 variants in familial papillary thyroid carcinoma: Contribution of a group of “thyroid-checked” controls, European Journal of Medical Genetics (2017), http://dx.doi.org/10.1016/j.ejmg.2017.01.001
B. Kern et al. / European Journal of Medical Genetics xxx (2017) 1e7
3
Table 1 Comparative analysis of the frequency of HABP2 variants in controls and fPTC patients. HABP2, p.R122W, c.364C > T
European (non finnish) population
Our Controls without any PTC
fPTC patients
allele C frequency allele T frequency Number of studied alleles
99,7% 0,30%a 66,710
99.26% 0.73% ns 272
92.5% 7.5% *,# 40
HABP2, p.G534E, c.1601G > A Healthy controls from UK, Italy, Netherlands and Germany (from 4 published studies) Our Controls without any PTC fPTC patients allele G frequency allele A frequency Number of studied alleles
97,10% 2,9%b 7218
97.06% 2.94% ns 272
95% 5% ns 40
ns, non significant vs any control group. ns, non significant between the 2 control groups by Fisher exact test. *, p ¼ 0,016 by Fisher exact test vs our controls. #, p < 0.001 by Fisher exact test vs E/EA population. a Frequency observed in 33,355 European subjects from the ExAC 60,000 exomes database (Lek et al., 2016). b Mean frequency (CI 95%: 1,0e4,8%) obtained by merging data from n ¼ 810 controls from Italy (Circulation, 2003; 107:667), n ¼ 2076 from UK (Thromb Haemost, 2004; 92:986), n ¼ 441 from Holland (Blood, 2005; 105:4898) and n ¼ 282 from Germany (Thromb Haemost, 2006; 96:465).
Table 2 Clinical features of controls with variants of HABP2. HABP2, p.R122W, c.364C > T
Age at thyroidectomy
MTC sporadic or familial
Pathological report (Nature of the thyroid lesion in addition to MTC if any)
C1-male C2-female
46 70
sporadic sporadic
macrovesicular adenoma macrovesicular adenoma
72 79 70 68 67 67 55 48
sporadic sporadic sporadic sporadic sporadic sporadic Sporadic sporadic
microvesicular adenoma macrovesicular goiter macrovesicular goiter macrovesicular goiter lymphocytic thyroiditis none None none
HABP2, p.G534E, c.1601G > A C3-female C4-male C5-female C6-female C7-male C8-female C9-female C10-female
report. 2. Material A total of 11 independent kindreds with fPTC (4 families with 3 and 7 with 2 individuals) were recruited in this study. This represents a total of 20 patients with fPTC who gave a written consent allowing genetic analysis and were screened for HABP2 mutations by Sanger sequencing. None had any clinical evidence for other tumors. The control group consists of 136 subjects (88 females and 48 males) who underwent total thyroidectomy for medullary thyroid carcinoma (MTC) and who gave an informed consent allowing germline DNA analysis. Based on initial RET germline mutation screening, 111 subjects had a sporadic MTC whereas 25 had a familial MTC. The pathological report, which was available for each removed thyroid specimen, did not indicate the presence of a PTC. However, occurrence of a benign thyroid lesion i.e. multinodular goiter, adenoma (micro or macro vesicular) or thyroiditis on the thyroid specimen was not an exclusion criteria. Both patients with fPTC and controls were from European descent. 3. Methods 3.1. HABP2 analysis 3.1.1. Sanger sequencing Genomic DNA was extracted from peripheral blood cells. Each of the thirteen exons of HABP2 was amplified by PCR using previously published primers (Gara et al., 2015), except for exons 1 and 2 for which we redesigned specific primers using the online Applied Biosystems tool (www.thermofisher.com/order/genome-database/
edit). Primers sequences were as follows: exon 1: Forward: TCTCCCCAGAAGAGAAACCACTTCA; Reverse: TCCAAGATGGCTCAAGTCAGACA; exon 2: Forward: CCATCAGAAAGCACGCAGTG; Reverse: AACCAGTTGAACCCTGAGGTC. 3.1.2. SnapShot Multiplex PCR was performed on genomic or somatic DNA to coamplify exons 5 and 13 of HABP2 in a 25 mL reaction volume containing 1 Buffer, MgCl2 2.5 mM, 2.5 mM for each dNTP, 5pMol of each primer, 1U of AmpliTaq Gold DNA-polymerase (Applied BioSystems/Life Technologies, Foster City, CA, USA). Multiplex amplification was performed as follows: 95 C for 12 min; 95 C for 30 s, 60 C (minus 1 C per cycle) for 90 s, 72 C for 90 s for 10 cycles; 95 C for 30 s, 55 C for 90 s, 72 C for 90 s for 27 cycles and 72 C for 5 min for 1 cycle. The PCR primers were as follows: Exon 13: Forward: AGTCACCCCACCCTCAAACA; Reverse: ATCAGATGCATCTGGCCTTC. Exon 5: Forward: CCCTGACACCCCCTGGAGAG; Reverse: GCTCTGGAGGTGTCCATTGT. All amplicons were controlled by electrophoresis on 2% agarose gel and purified by centrifugation through Bio-Gel P10 polyacrylamide gel (BIORAD, Marnes la Coquette, France). Specific primers were designed to analyze the c.364C > T (dbsnp: rs78201625) and c.1601G > A (dbsnp:rs7080536) variants by SNaPshot Applied BioSystems/Life Technologies with sequences as follows: c.364C > T: Forward: *N*N CAAGGACAACCCATGTGGC; Reverse: *N*N GTAATGAGACATTGGCCCC; c.1601G > A: Forward: *N GCTGGGGCCTGGAGTGTG; Reverse: *N TGTAGACCCCTGGCCTCTTC, where *N is TGGTTAGATG, a rare ubiquitous 10 nucleotides sequence allowing to differentiate the amplicon size during migration. Two primer extension reactions [25 cycles: 96 C for 10 s, 50 C for 5 s, 60 C for 30 s ]were performed on each purified
Please cite this article in press as: Kern, B., et al., Multiple HABP2 variants in familial papillary thyroid carcinoma: Contribution of a group of “thyroid-checked” controls, European Journal of Medical Genetics (2017), http://dx.doi.org/10.1016/j.ejmg.2017.01.001
4
B. Kern et al. / European Journal of Medical Genetics xxx (2017) 1e7
PCR sample, one with the forward specific primers mix and the other with the reverse specific primers mix. Extension products were purified by adding Thermosensitive Shrimp Alkaline Phosphatase (PROMEGA Madison WI USA) for 1 h at 37 C followed by 15 min at 75 C for enzyme inactivation. After denaturation by Hi Di Formamide the products were analyzed on a 3130 XL Applied Biosystems automatic DNA sequencer suing the GeneMapper software V4.1 (Applied BioSystems, Foster City, CA, USA). 3.1.3. 3D structure predictions Prediction of the human native HABP2 protein and the human Arg122Trp mutated HABP2 protein was carried out with the Phyre2 server (Protein Homology/analogY Recognition Engine V2.0) using the deduced amino-acid sequence of each proteins. The predicted 3D structure of both proteins was compared with the 3D resolved structure of the Ca2þ bound human prothrombin (33% identity sequence with the human native HABP2 protein and 32% with the human Arg122Trp mutated HABP2 protein) using the molecular visualization system, RasMol 2.7.5.1 (Herbert J. Bernstein). The topology of proteins is almost conserved in the 3D predicted structures as illustrated on Fig. 3. To determine whether both proteins were structurally homologous, the protein SuperPose server v.1.0 (Rajarshi Maiti, Gary Van Domselaar, Haiyan Zhang, and David Wishart) was used to measure the root-mean-square-deviation (*RMSD): the average distance between the alpha carbons atoms (the backbone atoms) of superimposed proteins. The RMSD value on the domain of amino acids 94 to 141 where the Arg122Trp mutation is located was equal to 0,48 Å. If the RMSD is below 1,5 Å, two 3D structures or domains whose sequence alignment is over 30% can be considered as almost homologous. The most significant result is correlated with the highest number of residues aligned. Regarding both HABP2 proteins, there is structural homology between the 3D predicted structures of the human native HABP2 protein and the human Arg122Trp mutated HABP2 protein, especially on the domain of amino acids 94 to 141. 3.1.4. Expression analysis mRNA from frozen tumor tissues (12 samples) were extracted as previously described (Renaud et al., 2014). Then HAPB2 and GAPDH mRNA were quantified by RT-qPCR using SsoFastTM EvaGreen® supermix on the CFX96™ Real-Time System (Bio-Rad) with the €e et al. (2010). Semi-quantitative RTprimers described by Altma PCR was performed on two tumors samples positive for HABP2 expression (T1 and T2) using Multiplex PCR Qiagen Master Mix (Qiagen). Primers used for HABP2 amplification were as follows: F: 50 AGGAAGAGAACACCAGTAGCA 30 and R: 50 TAGTAGGGAGGACTCTGGGTA 3’. 18S was used as a loading control and amplified with the following primers F: 50 GGACCAGAGGCAAAGCATTTGCC30 and R: 50 TCAATCTCGGGTGGCTGAACGC 3’. 3.1.5. Statistical analysis Minor allele frequencies (MAF) were compared by the Fisher exact test using the GraphPad Prism 6 software. P < 0.05 was considered as statistically significant. 4. Results Kindred #1 (Fig. 1) includes three sisters who underwent thyroid surgery between the ages of 48 and 51 years for one or several micro PTCs associated with chronic thyroiditis in 2 cases. Two sisters (II.1 & II.3) had the HABP2 c.1601G > A variant at the heterozygous state whereas patient II.2 did not. Since the micro PTCs of the 3 sisters were undistinguishable morphologically, we decided to analyze all exons of HABP2 by Sanger sequencing. We identified in patient II.2 a new heterozygous variant in exon 5 of HABP2 i.e.
c.364C > T that leads to a p.Arg122Trp change at the protein level. This new variant was absent in her two sisters while her mother (I.1) had both variants on different alleles and no detectable thyroid cancer at ultrasonography at 79 yrs (Fig. 1). Based on the existence of these 2 variants, we decided to analyze all HABP2 exons in the 10 remaining fPTC families. Another kindred (#6, Fig. 1) was positive for the c.364C > T variant. The 2 affected members were operated on at 42 (II.2) or 58 yrs (I.1). Four additional subjects from this family were analyzed. The variant was absent in the 4 subjects who had no ultrasonographic detectable thyroid cancer. Additionally, we found 3 recurrent synonymous variants, p.His61His in 10 patients, p.Lys319Lys in 15 patients and p.Ala350Ala in 16 patients. All these variants correspond to Single Nucleotide Polymorphism since their minor allele frequency (MAF) in the ExAC database was 41%. Finally, another variant (c.95G > T, p.Ser32Ile) was found in one patient negative for the c.364C > T and the c.1601G > A variants. In the ExAC database, its MAF was 0.59%. Moreover, bioinformatics analysis using SIFT software predicts that this variant, which concerns a poorly conserved amino-acid, is tolerated. Overall, the prevalence of the 2 most frequent HABP2 variants i.e. c.1601G > A and c.364C > T in the current study were 2/20 (10%) and 3/20 patients with fPTC (15%) respectively corresponding to minor allele frequencies (MAF) of 5% (allele A, c.1601G > A variant) and 7.5% (allele T, c.364C > T variant) (Table 1). The control group consists of 136 patients who underwent total thyroidectomy for sporadic or familial medullary thyroid carcinoma, and who were free from any histologically proven PTC or follicular cell-deriving carcinoma. A Snap Shot assay was set up to specifically identify these 2 events. Two controls had the c.364C > T variant, whereas eight controls had the c.1601G > A variant in both cases at the heterozygous level. None had both variants. Clinical features of the positive controls are detailled in Table 2. The MAF were 0.73% (allele T, c.364C > T variant) and 2.94% (allele A, c.1601G > A variant) in controls. In Table 1, we compared the MAF of each variant between patients with fPTC, our control group and other european control populations deriving either from the ExAc catalogue (Lek et al., 2016) (c.364C > T) or from 4 studies that included healthy controls (c.1601G > A). The MAF of the c.364C > T variant was significanly higher in patients with fPTC than in both control groups, which were very similar. Regarding the c.1601G > A variant, the MAF was not different between patients and any control group and also between the 2 groups of controls. The Arg122 residue belongs to the EGF-3 domain of HABP2 which is important for its activation by polyanionic molecules (Muhl et al., 2009) and is a conserved site in mammalian homologs of HABP2 except in mouse and rat. In the mutated protein, the polar and positively charged arginine is replaced by a less polar tryptophan. However, there is structural homology between the 3D predicted structures of the human native HABP2 protein and the human Arg122Trp mutated HABP2 protein, especially on the domain of amino acids 94 to 141 (Fig. 2). We studied HABP2 expression in thyroid tissues by RT-qPCR using 12 frozen PTC tumors samples coming from a different cohort of patients (Renaud et al., 2014). We found 8 positive samples with Ct values inferior to 40 and with a range of 20e22 for GAPDH endogenous control whereas 4 samples did not express HABP2 (the Ct value was undetermined for HABP2 while in the range of 20e22 for GAPDH endogenous control). Moreover, we selected two tumors samples positive for HABP2 expression by RTqPCR (T1: Ct value ¼ 37.6 and T2: Ct value ¼ 38.4) and performed a semi-quantitative RT-PCR to visualize HABP2 transcripts. In both cases we observed an amplicon whose intensity is inversely correlated with the Ct value thus confirming our qPCR data.
Please cite this article in press as: Kern, B., et al., Multiple HABP2 variants in familial papillary thyroid carcinoma: Contribution of a group of “thyroid-checked” controls, European Journal of Medical Genetics (2017), http://dx.doi.org/10.1016/j.ejmg.2017.01.001
B. Kern et al. / European Journal of Medical Genetics xxx (2017) 1e7
5
Fig. 2. Three-dimensional predicted structure of the human native HABP2 protein superimposed with the three-dimensional predicted structure of the human Arg122Trp mutated HABP2 protein. Ribbon diagrams showing the superimposition of the 3D structure prediction of the human native HABP2 protein (colored in GreenTint) with Arg122 shown as sticks and balls and colored in green and the 3D structure prediction of the human Arg122Trp mutated HABP2 protein (colored in Orange) with the Trp122 shown as sticks and balls and colored in red. The domain of amino acids 94 to 141 is delimited by a red circle. 3D structures are visualized with RasMol 2.7.5.1 (Herbert J. Bernstein). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
5. Discussion The p.Gly534Glu variant of HABP2, also known as the Marburg polymorphism, results in a 50e80% decrease in the proteolytic activity of the protein towards various substrates including growth factor PDGF-BB (Sedding et al., 2006) and has been associated with an increased risk of thromboembolic events and carotid restenosis. VEGF which regulates angiogenesis has also been described as a substrate of HABP2 (Jeon et al., 2006). The fact that the p.Gly534Glu variant negatively impacts HABP2 proteolytic activity (Sedding et al., 2006) and tumor suppressor effect (Gara et al., 2015) is in favor of a deleterious mutation. By contrast, the fact that the minor allele of this variant had a similar frequency in patients with familial PTC and in controls from different databases (Zhou et al., 2015; Sponziello et al., 2015; Sahasrabudhe et al., 2016) questions its pathogenicity. Here we demonstrated that controls with a MTC and a benign follicular thyroid lesion but without any PTC may frequently have the HABP2 c.1601G > A (p.Gly534Glu) variant. Therefore the MAF did not differ between those controls (2.94%) and patients with fPTC (5.0%). These results obtained on “thyroidchecked” controls do not support the pathogenicity of the c.1601G > A, at least in the European population, in accordance with a recent study (Sahasrabudhe et al., 2016). Regarding the new c.364C > T (p.Arg122Trp) variant, data are rather scarce. It has not been previously described at the somatic level in different tumor types according to the data available in the COSMIC or TCGA databases (query date: 04/22/2016). Our data
demonstrate (i) a higher MAF in patients with fPTC than in controls of the same ethnic origin devoid of any PTC (pathogenicity criteria PS4 of Richards et al., 2015) and (ii) co-segregation of the variant with PTC in one kindred (pathogenicity criteria PP1). The combination of pathogenicity criteria PS4 and PP1 is not sufficient to classify the c.364C > T variant as likely pathogenic, and we should consider it as of uncertain significance. Indeed computional analysis by 3 predictive in silico algorythms together with the superposition of the 3D structure of the wild type and mutated proteins suggest that this variant is tolerated at the protein level. Literature data are quite controversial about HABP2 mRNA expression in thyroid tumors. Gara et al. (2015) found a higher expression in thyroid tumor tissue in comparison with normal thyroid tissue whereas Tomsic et al. (2016) demonstrated that tumorous and normal thyroid tissue did not express these transcripts. In this study, we demonstrated that HABP2 mRNA are frequently expressed in PTC sample. So deleterious mutation occurring in this gene could contribute to the pathogenesis of PTC by generation of non-functional protein. In conclusion, we demonstrated that the c.364C > T variant of HABP2 has a significantly higher prevalence in patients with fPTC. However a better understanding of the role of this variant in fPTC needs a collaborative effort allowing the study of a larger number of fPTC kindreds from different ethnic origins, preferentially including at least 3 cases of PTC to avoid any a fortuitous aggregation of sporadic cases (Cao et al., 2016). In vitro data on experimental models are also required to evaluate the consequences of the
Please cite this article in press as: Kern, B., et al., Multiple HABP2 variants in familial papillary thyroid carcinoma: Contribution of a group of “thyroid-checked” controls, European Journal of Medical Genetics (2017), http://dx.doi.org/10.1016/j.ejmg.2017.01.001
6
B. Kern et al. / European Journal of Medical Genetics xxx (2017) 1e7
Fig. 3. HABP2 gene expression in thyroid tumors. (A) HABP2 expression was quantified by RT-qPCR in cDNA extracted from 12 PTC samples and GAPDH was used as an internal control. All samples express GAPDH and 8 out of 12 tumor samples express HABP2 with cycle threshold (Ct) below 40. Amplification plot on the right shows the Ct values for each tumor sample and Caco-2 cell line (used a positive control: Ct value ¼ 21 for HABP2). Each sample was tested in duplicate. (B) Two tumors samples positive for HABP2 expression by RT-qPCR (T1: Ct value ¼ 37.6 and T2: Ct value ¼ 38.4) were analyzed by semi-quantitative RT-PCR to visualize HABP2 transcripts. In both cases we observed an amplicon whose intensity is inversely correlated with the Ct value thus confirming our qPCR data. M corresponds to the DNA ladder.
p.R122W variant on HABP2 function and to determine whether it may have a dominant negative tumor suppressor effect as previously proposed (Gara et al., 2015). Funding This research did not receive any specific grant from funding agencies in the public, commercial or non-for-profit sectors. References €e, S., Martínez-Conejero, J.A., Salumets, A., Simo n, C., Horcajadas, J.A., StavAltma reus-Evers, A., 2010. Endometrial gene expression analysis at the time of embryo implantation in women with unexplained infertility. Mol. Hum. Reprod. 16, 178e187. Cao, J., Chen, C., Chen, C., Wang, Q.L., Ge, M.H., 2016. Clinicopathological features and prognosis of fPTC-a large scale, matched, case-control study. Clin. Endocrinol. 84, 598e606. Gara, S.K., Jia, L., Merino, M.J., Agarwal, S.K., Zhang, L., Cam, M., Patel, D., Kebebew, E., 2015. Germline HABP2 mutation causing familial non medullary thyroid cancer. N. Engl. J. Med. 373, 448e455. Jeon, J.W., Song, H.S., Moon, E.J., Park, S.Y., Son, M.J., Jung, S.Y., Kim, J.T., Nam, D.H., Choi-Miura, N.H., Kim, K.W., Kim, Y.J., 2006. Anti-angiogenic action of plasma hyaluronan binding protein in human umbilical vein endothelial cells. Int. J. Oncol. 29, 209e215. Lek, M., Karczewski, K.J., Minikel, E.V., Samocha, K.E., Banks, E., Fennel, T., et al., 2016. Analysis of protein-coding genetic variation in 60,706 humans. Nature 536, 285e291. Muhl, L., Hersemeyer, K., Preissner, K.T., Weimer, T., Kanse, S.M., 2009. Structurefunction analysis of factor VII activating protease (FSAP): sequence
determinants for heparin binding and cellular functions. FEBS Lett. 583, 1994e1998. pin, M., Coppin, L., Ramdane, N., Renaud, F., Gnemmi, V., Devos, P., Aubert, S., Cre Bouchindhomme, B., d'Herbomez, M., Van Seuningen, I., Do Cao, C., Pattou, F., meau, J.L., Leteurtre, E., 2014. MUC1 expression in Carnaille, B., Pigny, P., We papillary thyroid carcinoma is associated with BRAF mutation and lymph node metastasis; the latter is the most important risk factor of relapse. Thyroid 24, 1375e1384. Richards, S., Aziz, N., Bale, S., Bick, D., Das, S., Gastier-Foster, J., Grody, W.W., Hegde, M., Lyon, E., Spector, E., Voelkerding, K., Rehm, H.L., 2015. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American college of medical genetics and genomics and the Association for Molecular Pathology. Genet. Med. 17, 405e424. Sahasrabudhe, R., Stulz, J., Williamson, J., Lott, P., Estrada, A., Bohorquez, M., Palles, C., Polanco-Echeverry, G., Jaeger, E., Martin, L., Echeverry, M.M., 2016. Tomlinson I on behalf of TUCKIN, Carvajal-Carmona LG: the HABP2 G534E variant is an unlikely cause of familial non medullary thyroid cancer. J. Clin. Endocrinol. Metab. 101, 1098e1103. Sedding, D., Daniel, J.M., Muhl, L., Hersemeyer, K., Brunsch, H., Kemkes-Matthes, B., Braun-Dullaeus, R.C., Tillmanns, H., Weimer, T., Preissner, K.T., Kanse, S.M., 2006. The G534E polymorphism of the gene encoding the factor VII activating protease is associated with cardiovascular risk due to increased neointima formation. J. Exp. Med. 203, 2801e2807. Seehofer, D., Rayes, N., Klupp, J., Nüssler, N.C., Ulrich, F., Graef, K.J., Schindler, R., Steinmüller, T., Frei, U., Neuhaus, P., 2005. Prevalence of thyroid nodules and carcinomas in patients operated on for renal hyperparathyroidism : experience with 339 consecutive patients and review of the litterature. World J. Surg. 29, 1180e1184. Sponziello, M., Durante, C., Filetti, S., 2015. HABP2 mutation and non medullary thyroid cancer. N. Engl. J. Med. 373, 2085. Tanriover, O., Comunoglu, N., Eren, B., Comunoglu, C., Turkmen, N., Dogan, M., Gündogmus, U.N., 2011. Occult PTC: prevalence at autopsy in Turkish people. Eur. J. Cancer Prev. 20, 308e312.
Please cite this article in press as: Kern, B., et al., Multiple HABP2 variants in familial papillary thyroid carcinoma: Contribution of a group of “thyroid-checked” controls, European Journal of Medical Genetics (2017), http://dx.doi.org/10.1016/j.ejmg.2017.01.001
B. Kern et al. / European Journal of Medical Genetics xxx (2017) 1e7 Tomsic, J., He, H., de la Chapelle, A., 2015. HABP2 mutation and nonmedullary thyroid cancer. N. Engl. J. Med. 373, 2086. Tomsic, J., Fultz, R., Liyanarachchi, S., He, H., Senter, L., de la Chapelle, A., 2016. HABP2 G534E variant in papillary thyroid carcinoma. PLoS One 11. Weeks, A.L., Wilson, S.G., Ward, L., Goldblatt, J., Hui, J., Walsh, J.P., 2016. HABP2 germline variants are uncommon in familial nonmedullary thyroid cancer. BMC
7
Med. Genet. 17, 60. Zhao, X., Li, X., Zhang, X., 2015. HABP2 mutation and nonmedullary thyroid cancer. N. Engl. J. Med. 373, 2084. Zhou, E.Y., Lin, Z., Yang, Y., 2015. HABP2 mutation and non medullary thyroid cancer. N. Engl. J. Med. 373, 2084.
Please cite this article in press as: Kern, B., et al., Multiple HABP2 variants in familial papillary thyroid carcinoma: Contribution of a group of “thyroid-checked” controls, European Journal of Medical Genetics (2017), http://dx.doi.org/10.1016/j.ejmg.2017.01.001
Contents lists available at ScienceDirect
European Journal of Medical Genetics journal homepage: http://www.elsevier.com/locate/ejmg
Multiple HABP2 variants in familial papillary thyroid carcinoma: Contribution of a group of “thyroid-checked” controls pin a, Isabelle Szuster a, Benjamin Kern a, Lucie Coppin a, Pauline Romanet b, Michel Cre de ric Fre nois d, Jean-Louis Wemeau e, Florence Renaud c, Emmanuelle Leteurtre c, Fre f e Bruno Carnaille , Catherine Cardot-Bauters , Christine Do Cao e, Pascal Pigny a, * a Service de Biochimie « Hormonologie-M etabolisme-Nutrition & Oncologie », Centre de Biologie Pathologie, Centre Hospitalier R egional & Universitaire, F59037 Lille Cedex, France b ^pital de la Conception, AP-HM, 13385 Marseille Cedex 05, France Laboratoire de Biologie Mol eculaire, Ho c Service d’Anatomie Pathologique, Centre de Biologie Pathologie, Centre Hospitalier R egional & Universitaire, F-59037 Lille Cedex, France d Universit e de Lille 2, CHRU, Equipe RADEME, EA 7364, Centre Hospitalier R egional & Universitaire, F-59037 Lille Cedex, France e ^pital Claude Huriez, Centre Hospitalier R Service d’Endocrinologie, Ho egional & Universitaire, F-59037 Lille Cedex, France f ^pital Claude Huriez, Centre Hospitalier R Service de Chirurgie Endocrinienne, Ho egional & Universitaire, F-59037 Lille Cedex, France
a r t i c l e i n f o
a b s t r a c t
Article history: Received 20 September 2016 Received in revised form 15 December 2016 Accepted 7 January 2017 Available online xxx
A heterozygous germline variant in the HABP2 gene c.1601G > A (p.Gly534Glu), which negatively impacts its tumor suppressive activity in vitro, has been described in 4e14% of kindreds of EuropeanAmerican ancestry with familial papillary thyroid carcinoma (fPTC). But it is also found in z4% of Europeans and European/Americans from public databases that, however, did not provide information on the thyroid function of the controls. To get unbiased results, we decided to compare HABP2 genotypes of patients with fPTC with those of “thyroid-checked” controls. A control group consisting of 136 European patients who were thyroidectomised for medullary thyroid carcinoma and devoid of any histologically detectable PTC or follicular-deriving carcinoma was built. In parallel we recruited 20 patients with fPTC from eleven independent European kindreds. The entire coding region of HABP2 was analyzed by Sanger sequencing the germline DNAs of patients. Nucleotide variants were searched for by Snap Shot analysis in the controls. Two variants, c.1601G > A (p.Gly534Glu) and c.364C > T (p.Arg122Trp), were found in 2 and 3 patients at the heterozygous level respectively (minor allele frequency (MAF): 5.0% and 7.5%, respectively). In controls, the MAF was either similar for the c.1601G > A HABP2 variant (2.94%, ns) or significantly lower for the c.364C > T variant (0.73%, p ¼ 0.016). The Arg122 residue lies in the EGF-3 domain of HABP2 which is important for its activation but, however, superposition of the predicted 3D structures of the wild type and mutated proteins suggests that this variant is tolerated at the protein level. In conclusion, our data do not support the pathogenicity of the HABP2 c.1601G > A variant but highlight the existence of a new one that should be more extensively searched for in fPTC patients and its pathogenicity more carefully evaluated. © 2017 Elsevier Masson SAS. All rights reserved.
Keywords: Thyroid cancer Familial cancer Papillary HABP2 Novel variant
1. Introduction HABP2 encodes a hyaluronic acid-binding protein, also named Factor VII-activating protease (FSAP), which is known to play roles in coagulation and fibrinolysis through the activation of Factor VII and pro-urokinase. Recently HABP2 has been described as a new
culaire, Centre de * Corresponding author. Institut de Biochimie et Biologie Mole Biologie Pathologie, CHRU, F-59037 Lille cedex, France. E-mail address: [email protected] (P. Pigny).
susceptibility gene for familial papillary thyroid carcinoma (fPTC) based on the description of a heterozygous variant c.1601G > A (p.Gly534Glu) in one kindred with fPTC (Gara et al., 2015). The authors provided three arguments supporting the causal role of this variant in carcinogenesis and its classification as a mutation: (i) functionally, HABP2 has a tumor suppressor effect in epithelial cell lines which is reversed by the mutation (dominant negative effect); (ii) this mutation co-segregates with PTC in 6 patients on 2 generations; (iii) its prevalence is significantly higher in patients with apparently sporadic PTC (4.7%) versus controls (0.7%) from a
http://dx.doi.org/10.1016/j.ejmg.2017.01.001 1769-7212/© 2017 Elsevier Masson SAS. All rights reserved.
Please cite this article in press as: Kern, B., et al., Multiple HABP2 variants in familial papillary thyroid carcinoma: Contribution of a group of “thyroid-checked” controls, European Journal of Medical Genetics (2017), http://dx.doi.org/10.1016/j.ejmg.2017.01.001
2
B. Kern et al. / European Journal of Medical Genetics xxx (2017) 1e7
multiethnic database (Gara et al., 2015). However, the causal role of the HABP2 c.1601G > A mutation in fPTC has been quickly challenged by different studies that reported higher frequencies of this variant in controls from different databases, ranging from 2.9 to 4.6% in the European population (Zhou et al., 2015; Sponziello et al., 2015; Sahasrabudhe et al., 2016; Weeks et al., 2016; Tomsic et al., 2015, 2016; Zhao et al., 2015) thus invalidating the third criteria mentioned above. However, as recently stated by both the American College of Medical Genetics and Genomics, and the Association for Molecular Pathology (Richards et al., 2015), “population databases cannot be assumed to include only healthy individuals”. Indeed, databases never provide extensive information on any
putative disease or phenotype that may be present in population selected as controls. Interestingly, the prevalence of occult (clinically indolent) PTC varies from 3.7% to 11% in most autopsy studies in Europe (Seehofer et al., 2005; Tanriover et al., 2011). Therefore, in this specific context, the comparison of the prevalence of a nucleotide variant between patients with fPTC and “thyroid-unchecked” controls may be especially inaccurate and biased. To clarify this issue, we decided to retrospectively search for HABP2 variants in several kindreds with fPTC by nucleotide sequencing the 13 exons of the gene, and then to look for these variants by SnapShot in a group of controls who underwent thyroidectomy and were free of any PTC based on the pathological
Fig. 1. Trees of the 2 kindreds with germline HABP2 variants. Squares denote male members, circles denote female members, hatched symbols denote patients with PTC, slashed symbols denote deceased members, and grey symbols denote family members without the variant and without any thyroid cancer at ultrasonography. Empty symbols denote unexplored family members. HABP2 genotype is shown in blue caracters (upper right). TNM classification of the thyroid tumor is shown in black caracters (lower right). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Please cite this article in press as: Kern, B., et al., Multiple HABP2 variants in familial papillary thyroid carcinoma: Contribution of a group of “thyroid-checked” controls, European Journal of Medical Genetics (2017), http://dx.doi.org/10.1016/j.ejmg.2017.01.001
B. Kern et al. / European Journal of Medical Genetics xxx (2017) 1e7
3
Table 1 Comparative analysis of the frequency of HABP2 variants in controls and fPTC patients. HABP2, p.R122W, c.364C > T
European (non finnish) population
Our Controls without any PTC
fPTC patients
allele C frequency allele T frequency Number of studied alleles
99,7% 0,30%a 66,710
99.26% 0.73% ns 272
92.5% 7.5% *,# 40
HABP2, p.G534E, c.1601G > A Healthy controls from UK, Italy, Netherlands and Germany (from 4 published studies) Our Controls without any PTC fPTC patients allele G frequency allele A frequency Number of studied alleles
97,10% 2,9%b 7218
97.06% 2.94% ns 272
95% 5% ns 40
ns, non significant vs any control group. ns, non significant between the 2 control groups by Fisher exact test. *, p ¼ 0,016 by Fisher exact test vs our controls. #, p < 0.001 by Fisher exact test vs E/EA population. a Frequency observed in 33,355 European subjects from the ExAC 60,000 exomes database (Lek et al., 2016). b Mean frequency (CI 95%: 1,0e4,8%) obtained by merging data from n ¼ 810 controls from Italy (Circulation, 2003; 107:667), n ¼ 2076 from UK (Thromb Haemost, 2004; 92:986), n ¼ 441 from Holland (Blood, 2005; 105:4898) and n ¼ 282 from Germany (Thromb Haemost, 2006; 96:465).
Table 2 Clinical features of controls with variants of HABP2. HABP2, p.R122W, c.364C > T
Age at thyroidectomy
MTC sporadic or familial
Pathological report (Nature of the thyroid lesion in addition to MTC if any)
C1-male C2-female
46 70
sporadic sporadic
macrovesicular adenoma macrovesicular adenoma
72 79 70 68 67 67 55 48
sporadic sporadic sporadic sporadic sporadic sporadic Sporadic sporadic
microvesicular adenoma macrovesicular goiter macrovesicular goiter macrovesicular goiter lymphocytic thyroiditis none None none
HABP2, p.G534E, c.1601G > A C3-female C4-male C5-female C6-female C7-male C8-female C9-female C10-female
report. 2. Material A total of 11 independent kindreds with fPTC (4 families with 3 and 7 with 2 individuals) were recruited in this study. This represents a total of 20 patients with fPTC who gave a written consent allowing genetic analysis and were screened for HABP2 mutations by Sanger sequencing. None had any clinical evidence for other tumors. The control group consists of 136 subjects (88 females and 48 males) who underwent total thyroidectomy for medullary thyroid carcinoma (MTC) and who gave an informed consent allowing germline DNA analysis. Based on initial RET germline mutation screening, 111 subjects had a sporadic MTC whereas 25 had a familial MTC. The pathological report, which was available for each removed thyroid specimen, did not indicate the presence of a PTC. However, occurrence of a benign thyroid lesion i.e. multinodular goiter, adenoma (micro or macro vesicular) or thyroiditis on the thyroid specimen was not an exclusion criteria. Both patients with fPTC and controls were from European descent. 3. Methods 3.1. HABP2 analysis 3.1.1. Sanger sequencing Genomic DNA was extracted from peripheral blood cells. Each of the thirteen exons of HABP2 was amplified by PCR using previously published primers (Gara et al., 2015), except for exons 1 and 2 for which we redesigned specific primers using the online Applied Biosystems tool (www.thermofisher.com/order/genome-database/
edit). Primers sequences were as follows: exon 1: Forward: TCTCCCCAGAAGAGAAACCACTTCA; Reverse: TCCAAGATGGCTCAAGTCAGACA; exon 2: Forward: CCATCAGAAAGCACGCAGTG; Reverse: AACCAGTTGAACCCTGAGGTC. 3.1.2. SnapShot Multiplex PCR was performed on genomic or somatic DNA to coamplify exons 5 and 13 of HABP2 in a 25 mL reaction volume containing 1 Buffer, MgCl2 2.5 mM, 2.5 mM for each dNTP, 5pMol of each primer, 1U of AmpliTaq Gold DNA-polymerase (Applied BioSystems/Life Technologies, Foster City, CA, USA). Multiplex amplification was performed as follows: 95 C for 12 min; 95 C for 30 s, 60 C (minus 1 C per cycle) for 90 s, 72 C for 90 s for 10 cycles; 95 C for 30 s, 55 C for 90 s, 72 C for 90 s for 27 cycles and 72 C for 5 min for 1 cycle. The PCR primers were as follows: Exon 13: Forward: AGTCACCCCACCCTCAAACA; Reverse: ATCAGATGCATCTGGCCTTC. Exon 5: Forward: CCCTGACACCCCCTGGAGAG; Reverse: GCTCTGGAGGTGTCCATTGT. All amplicons were controlled by electrophoresis on 2% agarose gel and purified by centrifugation through Bio-Gel P10 polyacrylamide gel (BIORAD, Marnes la Coquette, France). Specific primers were designed to analyze the c.364C > T (dbsnp: rs78201625) and c.1601G > A (dbsnp:rs7080536) variants by SNaPshot Applied BioSystems/Life Technologies with sequences as follows: c.364C > T: Forward: *N*N CAAGGACAACCCATGTGGC; Reverse: *N*N GTAATGAGACATTGGCCCC; c.1601G > A: Forward: *N GCTGGGGCCTGGAGTGTG; Reverse: *N TGTAGACCCCTGGCCTCTTC, where *N is TGGTTAGATG, a rare ubiquitous 10 nucleotides sequence allowing to differentiate the amplicon size during migration. Two primer extension reactions [25 cycles: 96 C for 10 s, 50 C for 5 s, 60 C for 30 s ]were performed on each purified
Please cite this article in press as: Kern, B., et al., Multiple HABP2 variants in familial papillary thyroid carcinoma: Contribution of a group of “thyroid-checked” controls, European Journal of Medical Genetics (2017), http://dx.doi.org/10.1016/j.ejmg.2017.01.001
4
B. Kern et al. / European Journal of Medical Genetics xxx (2017) 1e7
PCR sample, one with the forward specific primers mix and the other with the reverse specific primers mix. Extension products were purified by adding Thermosensitive Shrimp Alkaline Phosphatase (PROMEGA Madison WI USA) for 1 h at 37 C followed by 15 min at 75 C for enzyme inactivation. After denaturation by Hi Di Formamide the products were analyzed on a 3130 XL Applied Biosystems automatic DNA sequencer suing the GeneMapper software V4.1 (Applied BioSystems, Foster City, CA, USA). 3.1.3. 3D structure predictions Prediction of the human native HABP2 protein and the human Arg122Trp mutated HABP2 protein was carried out with the Phyre2 server (Protein Homology/analogY Recognition Engine V2.0) using the deduced amino-acid sequence of each proteins. The predicted 3D structure of both proteins was compared with the 3D resolved structure of the Ca2þ bound human prothrombin (33% identity sequence with the human native HABP2 protein and 32% with the human Arg122Trp mutated HABP2 protein) using the molecular visualization system, RasMol 2.7.5.1 (Herbert J. Bernstein). The topology of proteins is almost conserved in the 3D predicted structures as illustrated on Fig. 3. To determine whether both proteins were structurally homologous, the protein SuperPose server v.1.0 (Rajarshi Maiti, Gary Van Domselaar, Haiyan Zhang, and David Wishart) was used to measure the root-mean-square-deviation (*RMSD): the average distance between the alpha carbons atoms (the backbone atoms) of superimposed proteins. The RMSD value on the domain of amino acids 94 to 141 where the Arg122Trp mutation is located was equal to 0,48 Å. If the RMSD is below 1,5 Å, two 3D structures or domains whose sequence alignment is over 30% can be considered as almost homologous. The most significant result is correlated with the highest number of residues aligned. Regarding both HABP2 proteins, there is structural homology between the 3D predicted structures of the human native HABP2 protein and the human Arg122Trp mutated HABP2 protein, especially on the domain of amino acids 94 to 141. 3.1.4. Expression analysis mRNA from frozen tumor tissues (12 samples) were extracted as previously described (Renaud et al., 2014). Then HAPB2 and GAPDH mRNA were quantified by RT-qPCR using SsoFastTM EvaGreen® supermix on the CFX96™ Real-Time System (Bio-Rad) with the €e et al. (2010). Semi-quantitative RTprimers described by Altma PCR was performed on two tumors samples positive for HABP2 expression (T1 and T2) using Multiplex PCR Qiagen Master Mix (Qiagen). Primers used for HABP2 amplification were as follows: F: 50 AGGAAGAGAACACCAGTAGCA 30 and R: 50 TAGTAGGGAGGACTCTGGGTA 3’. 18S was used as a loading control and amplified with the following primers F: 50 GGACCAGAGGCAAAGCATTTGCC30 and R: 50 TCAATCTCGGGTGGCTGAACGC 3’. 3.1.5. Statistical analysis Minor allele frequencies (MAF) were compared by the Fisher exact test using the GraphPad Prism 6 software. P < 0.05 was considered as statistically significant. 4. Results Kindred #1 (Fig. 1) includes three sisters who underwent thyroid surgery between the ages of 48 and 51 years for one or several micro PTCs associated with chronic thyroiditis in 2 cases. Two sisters (II.1 & II.3) had the HABP2 c.1601G > A variant at the heterozygous state whereas patient II.2 did not. Since the micro PTCs of the 3 sisters were undistinguishable morphologically, we decided to analyze all exons of HABP2 by Sanger sequencing. We identified in patient II.2 a new heterozygous variant in exon 5 of HABP2 i.e.
c.364C > T that leads to a p.Arg122Trp change at the protein level. This new variant was absent in her two sisters while her mother (I.1) had both variants on different alleles and no detectable thyroid cancer at ultrasonography at 79 yrs (Fig. 1). Based on the existence of these 2 variants, we decided to analyze all HABP2 exons in the 10 remaining fPTC families. Another kindred (#6, Fig. 1) was positive for the c.364C > T variant. The 2 affected members were operated on at 42 (II.2) or 58 yrs (I.1). Four additional subjects from this family were analyzed. The variant was absent in the 4 subjects who had no ultrasonographic detectable thyroid cancer. Additionally, we found 3 recurrent synonymous variants, p.His61His in 10 patients, p.Lys319Lys in 15 patients and p.Ala350Ala in 16 patients. All these variants correspond to Single Nucleotide Polymorphism since their minor allele frequency (MAF) in the ExAC database was 41%. Finally, another variant (c.95G > T, p.Ser32Ile) was found in one patient negative for the c.364C > T and the c.1601G > A variants. In the ExAC database, its MAF was 0.59%. Moreover, bioinformatics analysis using SIFT software predicts that this variant, which concerns a poorly conserved amino-acid, is tolerated. Overall, the prevalence of the 2 most frequent HABP2 variants i.e. c.1601G > A and c.364C > T in the current study were 2/20 (10%) and 3/20 patients with fPTC (15%) respectively corresponding to minor allele frequencies (MAF) of 5% (allele A, c.1601G > A variant) and 7.5% (allele T, c.364C > T variant) (Table 1). The control group consists of 136 patients who underwent total thyroidectomy for sporadic or familial medullary thyroid carcinoma, and who were free from any histologically proven PTC or follicular cell-deriving carcinoma. A Snap Shot assay was set up to specifically identify these 2 events. Two controls had the c.364C > T variant, whereas eight controls had the c.1601G > A variant in both cases at the heterozygous level. None had both variants. Clinical features of the positive controls are detailled in Table 2. The MAF were 0.73% (allele T, c.364C > T variant) and 2.94% (allele A, c.1601G > A variant) in controls. In Table 1, we compared the MAF of each variant between patients with fPTC, our control group and other european control populations deriving either from the ExAc catalogue (Lek et al., 2016) (c.364C > T) or from 4 studies that included healthy controls (c.1601G > A). The MAF of the c.364C > T variant was significanly higher in patients with fPTC than in both control groups, which were very similar. Regarding the c.1601G > A variant, the MAF was not different between patients and any control group and also between the 2 groups of controls. The Arg122 residue belongs to the EGF-3 domain of HABP2 which is important for its activation by polyanionic molecules (Muhl et al., 2009) and is a conserved site in mammalian homologs of HABP2 except in mouse and rat. In the mutated protein, the polar and positively charged arginine is replaced by a less polar tryptophan. However, there is structural homology between the 3D predicted structures of the human native HABP2 protein and the human Arg122Trp mutated HABP2 protein, especially on the domain of amino acids 94 to 141 (Fig. 2). We studied HABP2 expression in thyroid tissues by RT-qPCR using 12 frozen PTC tumors samples coming from a different cohort of patients (Renaud et al., 2014). We found 8 positive samples with Ct values inferior to 40 and with a range of 20e22 for GAPDH endogenous control whereas 4 samples did not express HABP2 (the Ct value was undetermined for HABP2 while in the range of 20e22 for GAPDH endogenous control). Moreover, we selected two tumors samples positive for HABP2 expression by RTqPCR (T1: Ct value ¼ 37.6 and T2: Ct value ¼ 38.4) and performed a semi-quantitative RT-PCR to visualize HABP2 transcripts. In both cases we observed an amplicon whose intensity is inversely correlated with the Ct value thus confirming our qPCR data.
Please cite this article in press as: Kern, B., et al., Multiple HABP2 variants in familial papillary thyroid carcinoma: Contribution of a group of “thyroid-checked” controls, European Journal of Medical Genetics (2017), http://dx.doi.org/10.1016/j.ejmg.2017.01.001
B. Kern et al. / European Journal of Medical Genetics xxx (2017) 1e7
5
Fig. 2. Three-dimensional predicted structure of the human native HABP2 protein superimposed with the three-dimensional predicted structure of the human Arg122Trp mutated HABP2 protein. Ribbon diagrams showing the superimposition of the 3D structure prediction of the human native HABP2 protein (colored in GreenTint) with Arg122 shown as sticks and balls and colored in green and the 3D structure prediction of the human Arg122Trp mutated HABP2 protein (colored in Orange) with the Trp122 shown as sticks and balls and colored in red. The domain of amino acids 94 to 141 is delimited by a red circle. 3D structures are visualized with RasMol 2.7.5.1 (Herbert J. Bernstein). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
5. Discussion The p.Gly534Glu variant of HABP2, also known as the Marburg polymorphism, results in a 50e80% decrease in the proteolytic activity of the protein towards various substrates including growth factor PDGF-BB (Sedding et al., 2006) and has been associated with an increased risk of thromboembolic events and carotid restenosis. VEGF which regulates angiogenesis has also been described as a substrate of HABP2 (Jeon et al., 2006). The fact that the p.Gly534Glu variant negatively impacts HABP2 proteolytic activity (Sedding et al., 2006) and tumor suppressor effect (Gara et al., 2015) is in favor of a deleterious mutation. By contrast, the fact that the minor allele of this variant had a similar frequency in patients with familial PTC and in controls from different databases (Zhou et al., 2015; Sponziello et al., 2015; Sahasrabudhe et al., 2016) questions its pathogenicity. Here we demonstrated that controls with a MTC and a benign follicular thyroid lesion but without any PTC may frequently have the HABP2 c.1601G > A (p.Gly534Glu) variant. Therefore the MAF did not differ between those controls (2.94%) and patients with fPTC (5.0%). These results obtained on “thyroidchecked” controls do not support the pathogenicity of the c.1601G > A, at least in the European population, in accordance with a recent study (Sahasrabudhe et al., 2016). Regarding the new c.364C > T (p.Arg122Trp) variant, data are rather scarce. It has not been previously described at the somatic level in different tumor types according to the data available in the COSMIC or TCGA databases (query date: 04/22/2016). Our data
demonstrate (i) a higher MAF in patients with fPTC than in controls of the same ethnic origin devoid of any PTC (pathogenicity criteria PS4 of Richards et al., 2015) and (ii) co-segregation of the variant with PTC in one kindred (pathogenicity criteria PP1). The combination of pathogenicity criteria PS4 and PP1 is not sufficient to classify the c.364C > T variant as likely pathogenic, and we should consider it as of uncertain significance. Indeed computional analysis by 3 predictive in silico algorythms together with the superposition of the 3D structure of the wild type and mutated proteins suggest that this variant is tolerated at the protein level. Literature data are quite controversial about HABP2 mRNA expression in thyroid tumors. Gara et al. (2015) found a higher expression in thyroid tumor tissue in comparison with normal thyroid tissue whereas Tomsic et al. (2016) demonstrated that tumorous and normal thyroid tissue did not express these transcripts. In this study, we demonstrated that HABP2 mRNA are frequently expressed in PTC sample. So deleterious mutation occurring in this gene could contribute to the pathogenesis of PTC by generation of non-functional protein. In conclusion, we demonstrated that the c.364C > T variant of HABP2 has a significantly higher prevalence in patients with fPTC. However a better understanding of the role of this variant in fPTC needs a collaborative effort allowing the study of a larger number of fPTC kindreds from different ethnic origins, preferentially including at least 3 cases of PTC to avoid any a fortuitous aggregation of sporadic cases (Cao et al., 2016). In vitro data on experimental models are also required to evaluate the consequences of the
Please cite this article in press as: Kern, B., et al., Multiple HABP2 variants in familial papillary thyroid carcinoma: Contribution of a group of “thyroid-checked” controls, European Journal of Medical Genetics (2017), http://dx.doi.org/10.1016/j.ejmg.2017.01.001
6
B. Kern et al. / European Journal of Medical Genetics xxx (2017) 1e7
Fig. 3. HABP2 gene expression in thyroid tumors. (A) HABP2 expression was quantified by RT-qPCR in cDNA extracted from 12 PTC samples and GAPDH was used as an internal control. All samples express GAPDH and 8 out of 12 tumor samples express HABP2 with cycle threshold (Ct) below 40. Amplification plot on the right shows the Ct values for each tumor sample and Caco-2 cell line (used a positive control: Ct value ¼ 21 for HABP2). Each sample was tested in duplicate. (B) Two tumors samples positive for HABP2 expression by RT-qPCR (T1: Ct value ¼ 37.6 and T2: Ct value ¼ 38.4) were analyzed by semi-quantitative RT-PCR to visualize HABP2 transcripts. In both cases we observed an amplicon whose intensity is inversely correlated with the Ct value thus confirming our qPCR data. M corresponds to the DNA ladder.
p.R122W variant on HABP2 function and to determine whether it may have a dominant negative tumor suppressor effect as previously proposed (Gara et al., 2015). Funding This research did not receive any specific grant from funding agencies in the public, commercial or non-for-profit sectors. References €e, S., Martínez-Conejero, J.A., Salumets, A., Simo n, C., Horcajadas, J.A., StavAltma reus-Evers, A., 2010. Endometrial gene expression analysis at the time of embryo implantation in women with unexplained infertility. Mol. Hum. Reprod. 16, 178e187. Cao, J., Chen, C., Chen, C., Wang, Q.L., Ge, M.H., 2016. Clinicopathological features and prognosis of fPTC-a large scale, matched, case-control study. Clin. Endocrinol. 84, 598e606. Gara, S.K., Jia, L., Merino, M.J., Agarwal, S.K., Zhang, L., Cam, M., Patel, D., Kebebew, E., 2015. Germline HABP2 mutation causing familial non medullary thyroid cancer. N. Engl. J. Med. 373, 448e455. Jeon, J.W., Song, H.S., Moon, E.J., Park, S.Y., Son, M.J., Jung, S.Y., Kim, J.T., Nam, D.H., Choi-Miura, N.H., Kim, K.W., Kim, Y.J., 2006. Anti-angiogenic action of plasma hyaluronan binding protein in human umbilical vein endothelial cells. Int. J. Oncol. 29, 209e215. Lek, M., Karczewski, K.J., Minikel, E.V., Samocha, K.E., Banks, E., Fennel, T., et al., 2016. Analysis of protein-coding genetic variation in 60,706 humans. Nature 536, 285e291. Muhl, L., Hersemeyer, K., Preissner, K.T., Weimer, T., Kanse, S.M., 2009. Structurefunction analysis of factor VII activating protease (FSAP): sequence
determinants for heparin binding and cellular functions. FEBS Lett. 583, 1994e1998. pin, M., Coppin, L., Ramdane, N., Renaud, F., Gnemmi, V., Devos, P., Aubert, S., Cre Bouchindhomme, B., d'Herbomez, M., Van Seuningen, I., Do Cao, C., Pattou, F., meau, J.L., Leteurtre, E., 2014. MUC1 expression in Carnaille, B., Pigny, P., We papillary thyroid carcinoma is associated with BRAF mutation and lymph node metastasis; the latter is the most important risk factor of relapse. Thyroid 24, 1375e1384. Richards, S., Aziz, N., Bale, S., Bick, D., Das, S., Gastier-Foster, J., Grody, W.W., Hegde, M., Lyon, E., Spector, E., Voelkerding, K., Rehm, H.L., 2015. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American college of medical genetics and genomics and the Association for Molecular Pathology. Genet. Med. 17, 405e424. Sahasrabudhe, R., Stulz, J., Williamson, J., Lott, P., Estrada, A., Bohorquez, M., Palles, C., Polanco-Echeverry, G., Jaeger, E., Martin, L., Echeverry, M.M., 2016. Tomlinson I on behalf of TUCKIN, Carvajal-Carmona LG: the HABP2 G534E variant is an unlikely cause of familial non medullary thyroid cancer. J. Clin. Endocrinol. Metab. 101, 1098e1103. Sedding, D., Daniel, J.M., Muhl, L., Hersemeyer, K., Brunsch, H., Kemkes-Matthes, B., Braun-Dullaeus, R.C., Tillmanns, H., Weimer, T., Preissner, K.T., Kanse, S.M., 2006. The G534E polymorphism of the gene encoding the factor VII activating protease is associated with cardiovascular risk due to increased neointima formation. J. Exp. Med. 203, 2801e2807. Seehofer, D., Rayes, N., Klupp, J., Nüssler, N.C., Ulrich, F., Graef, K.J., Schindler, R., Steinmüller, T., Frei, U., Neuhaus, P., 2005. Prevalence of thyroid nodules and carcinomas in patients operated on for renal hyperparathyroidism : experience with 339 consecutive patients and review of the litterature. World J. Surg. 29, 1180e1184. Sponziello, M., Durante, C., Filetti, S., 2015. HABP2 mutation and non medullary thyroid cancer. N. Engl. J. Med. 373, 2085. Tanriover, O., Comunoglu, N., Eren, B., Comunoglu, C., Turkmen, N., Dogan, M., Gündogmus, U.N., 2011. Occult PTC: prevalence at autopsy in Turkish people. Eur. J. Cancer Prev. 20, 308e312.
Please cite this article in press as: Kern, B., et al., Multiple HABP2 variants in familial papillary thyroid carcinoma: Contribution of a group of “thyroid-checked” controls, European Journal of Medical Genetics (2017), http://dx.doi.org/10.1016/j.ejmg.2017.01.001
B. Kern et al. / European Journal of Medical Genetics xxx (2017) 1e7 Tomsic, J., He, H., de la Chapelle, A., 2015. HABP2 mutation and nonmedullary thyroid cancer. N. Engl. J. Med. 373, 2086. Tomsic, J., Fultz, R., Liyanarachchi, S., He, H., Senter, L., de la Chapelle, A., 2016. HABP2 G534E variant in papillary thyroid carcinoma. PLoS One 11. Weeks, A.L., Wilson, S.G., Ward, L., Goldblatt, J., Hui, J., Walsh, J.P., 2016. HABP2 germline variants are uncommon in familial nonmedullary thyroid cancer. BMC
7
Med. Genet. 17, 60. Zhao, X., Li, X., Zhang, X., 2015. HABP2 mutation and nonmedullary thyroid cancer. N. Engl. J. Med. 373, 2084. Zhou, E.Y., Lin, Z., Yang, Y., 2015. HABP2 mutation and non medullary thyroid cancer. N. Engl. J. Med. 373, 2084.
Please cite this article in press as: Kern, B., et al., Multiple HABP2 variants in familial papillary thyroid carcinoma: Contribution of a group of “thyroid-checked” controls, European Journal of Medical Genetics (2017), http://dx.doi.org/10.1016/j.ejmg.2017.01.001