Clin Genet 2015 Printed in Singapore. All rights reserved
© 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd CLINICAL GENETICS doi: 10.1111/cge.12612
Short Report
Mutation analysis in Norwegian families with hereditary hemorrhagic telangiectasia: founder mutations in ACVRL1 Heimdal K., Dalhus B., Rødningen O.K., Kroken M., Eiklid K., Dheyauldeen S., Røysland T., Andersen R., Kulseth M.A. Mutation analysis in Norwegian families with hereditary hemorrhagic telangiectasia: founder mutations in ACVRL1. Clin Genet 2015. © John Wiley & Sons A/S. Published by John Wiley & Sons Ltd, 2015 Hereditary hemorrhagic telangiectasia (HHT, Osler–Weber–Rendu disease) is an autosomal dominant inherited disease defined by the presence of epistaxis and mucocutaneous telangiectasias and arteriovenous malformations (AVMs) in internal organs. In most families (∼85%), HHT is caused by mutations in the ENG (HHT1) or the ACVRL1 (HHT2) genes. Here, we report the results of genetic testing of 113 Norwegian families with suspected or definite HHT. Variants in ENG and ACVRL1 were found in 105 families (42 ENG, 63 ACVRL1), including six novel variants of uncertain pathogenic significance. Mutation types were similar to previous reports with more missense variants in ACVRL1 and more nonsense, frameshift and splice-site mutations in ENG. Thirty-two variants were novel in this study. The preponderance of ACVRL1 mutations was due to founder mutations, specifically, c.830C>A (p.Thr277Lys), which was found in 24 families from the same geographical area of Norway. We discuss the importance of founder mutations and present a thorough evaluation of missense and splice-site variants. Conflict of interest
The authors declare no conflicts of interest.
K. Heimdala , B. Dalhusb,c , O.K. Rødningena , M. Krokena , K. Eiklida , S. Dheyauldeend , T. Røyslandd,e , R. Andersenf and M.A. Kulsetha a Department of Medical Genetics, Oslo University Hospital, Oslo, Norway, b Department for Medical Biochemistry, University of Oslo, Oslo, Norway, c Department for Microbiology, d Department for Otorhinolaryngology, Oslo University Hospital, Oslo, Norway, e Department for Otorhinolaryngology, Innlandet Hospital, Gjøvik, Norway, and f Department of Radiology and Nuclear Medicine, Oslo University Hospital, Oslo, Norway
Key words: ACVRL1 – ENG – founder mutation – mutation – genetic testing – hereditary hemorrhagic telangiectasia – HHT – Osler–Weber–Rendu Corresponding author: Ketil Heimdal, MD, PhD, Department of Medical Genetics, Oslo University Hospital, Oslo Norway. Tel.: +47 2307 5701; fax: +47 2307 5590; e-mail: [email protected] Received 9 February 2015, revised and accepted for publication 11 May 2015
Hereditary hemorrhagic telangiectasia (HHT) is a rare disease defined by epistaxis, the presence of mucocutaneous telangiectasias and arteriovenous malformations (AVMs) in internal organs. Most patients suffer from spontaneous recurrent epistaxis starting in childhood or adolescence. AVMs are most commonly found in lungs (30–50% of patients), central nervous system (10%) and liver (40–75%) (1). Although complications from internal organs are potentially fatal, most patients have a near normal life expectancy. The condition is inherited as an autosomal dominant trait with near full penetrance at the age of 65 (2). Severity is variable both between and
within families. Diagnosis is clinical according to the Curacao criteria (3). A score of 3–4 points (i.e. Curacao criteria positive) has a high sensitivity and specificity for HHT and for detecting mutations in the relevant genes. Detection of a mutation might be helpful when the clinical criteria are equivocal (score of 2 equal to a ‘possible’ diagnosis of HHT) (4). In most families, HHT is caused by mutations in either ENG (HHT1, 39–59% of cases) or ACVRL1 (HHT2, 25–57%) and for a small minority, in SMAD4 (1–2%) or GDF2 (HHT5, A (p.Trp217*), c.830C>A (p.Thr277Lys) and c.1450C>T (p.Arg484Trp) were found in 36 of 97 mutation positive families (37.1%). Altogether, 104 of 234 individuals (44.4%) with HHT with documented mutations in Norway have one of these three variants. Twenty-two of 24 families with the ACVRL1 c.830C>A could be traced to the Rana district in Nordland County. The mutation dominates HHT in this geographical area. All tested families (13 families) with c.830C>A shared a common core haplotype (Table S3). Similarly, ACVRL1 c.651G>A detected in seven families from South-Eastern Norway and families with origin in West Sweden, and all five families with the ACVRL1 c.1450C>T share common haplotypes. One of the latter families had another likely pathogenic ACVRL1 variant, c.11delG located in cis. The most frequent mutation in ENG, c.277C>T (p.Arg93*), is reported repeatedly in other populations (10–12) and in this study found in five families from various locations in South-Eastern Norway. In Norway, this mutation has apparently occurred repeatedly and on different haplotypes (data not shown). Most probably, c.277C>T represents a hot spot for mutations in the ENG gene. Evaluation of the variants found in ENG
Results
In 105 of 113 HHT families, a sequence variant in class 3–5 was identified (Fig. 1a,b and Table S1a,b). Three of these families had an unclassified variant in ENG and five in ACVRL1, giving a detection rate of likely pathogenic or pathogenic variants of 97/113 (85.8%). Forty-two (39+3VUS) families had a variant in ENG and 63 (58+5VUS) in ACVRL1. Sixty unique sequence variants were identified (33 in ENG and 27 in ACVRL1) including 32 novel variants (19 in ENG and 13 in ACVRL1). One family had two ACVRL1 variants in cis (c.1450C>T and c.11delG). Thirty variants in ENG and 23 in ACVRL1 (including c.11delG) were evaluated as likely pathogenic or pathogenic. The distribution according to variant type was as expected with more missense variants in ACVRL1 (45/63 families→71.4%) compared to ENG (8/42 families→19.0%). There were three gross alterations in ENG and none in ACVRL1. In eight families, no mutation in ACVRL and ENG could be identified.
2
Nearly half of the DNA variants found in ENG were nonsense or small out-of-frame indels detected in one or two families each (Table S1a and Fig. 1a). These variants are evaluated as likely pathogenic because haploinsufficiency of ENG has been shown to be an underlying cause of HHT (5). Similarly, the three gross alterations are expected to lead to loss of protein and to be pathogenic. We identified nine splice-site variants. Five of these were located in the invariant splice-site sequence GU (+1 +2) or AG (−1 −2), and expected to cause aberrant splicing. The variants c.67+5G>A, c.68−3C>A, c.360+3G>C and c.361−11T>A were analyzed by reverse transcription polymerase chain reactions (RT-PCR). Complete missplicing was confirmed for c.67+5G>A and partial missplicing was found for c.360+3G>C and c.361−11T>A. Missplicing was found for c.68−3C>A; however, we were unable to determine the degree of missplicing (Fig. S1). We identified seven missense variants in ENG, of which three were novel (p.Cys30Arg, p.Leu80Arg and
HHT in Norway ENG
2
Arg171Pro
3
4
Arg93* Gln110*
Gly331Ser
Arg437Gln
5 6 7 8
9 10
Arg529His
11 12
13
14
Lys585*
Lys237* Tyr120*
nonsense
1
Leu80Arg
Trp261*
c.166dup
c.647del
c.1128dup
c.1385dup
indels /frameshift
Cys30Arg
missense
Arg197Pro
c.1680dup c.1452dup
c.1120_1123del
c.68-3C>G
c.67+1G>A c.67+5G>A
c.360+3G>C
splice site mutations
c.772_774delTACinsA c.817-1G>T
c.360+1G>A
c.361-11T>A
c.1134+1G>A
c.1686+1G>A
del/dup Cys344Tyr
ACVRL1
His328Tyr His314Gln
Ala400Val
1
2
Arg411Trp
Gly211Cys
3
4
5
Arg411Gln
6
7
Trp217* c.11del c.203del
Pro424Arg
8
Val442Ala
9
Arg484Trp
10
Arg479*
Gln272*
c.353_360dup
c.1041del
c.444_445del
c.1042del
c.1431_1432insCACCCGCT
nonsense
Glu236Lys Gly79Arg
missense
Arg374Gln
Thr277Lys
indels /frameshift
c.264del c.314-3C>G c.313+1G>A
splice site mutations
Fig. 1. Sequence variants found in the Norwegian cohort of HHT patients. (a) ENG. (b) ACVRL1. Deletions are shown as a blue bar, duplication as a red bar. Dark yellow = pathogenic variant. Light yellow = likely pathogenic. White = Variant of Uncertain Significance (VUS).
p.Arg171Pro). These variants are located in the orphan N-terminal domain of endoglin, which show no significant homology to any known family of proteins (13) The pathogenic effect of these variants is uncertain. Evaluation of the variants found in ACVRL1
Seven of the novel variants found in ACVRL1 were nonsense, small out-of-frame indels or located in the
consensus splice-site position +1, and are thus (likely) pathogenic (Table S1b and Fig. 1b). RT-PCR confirmed that the novel splice-site mutation c.314-3 located in intron 3 caused missplicing of ACVRL1 mRNA (Fig. S1). These novel variants are likely pathogenic because haploinsufficiency of ACVRL1 is an underlying cause of HHT (5). We identified 14 missense variants in ACVRL1 in our families, 4 of these were novel to this study. One was
3
Heimdal et al. located in the extracellular ligand binding domain and 13 in the cytoplasmic serine/threonine kinase domain. ALK1 belongs in the protein kinase superfamily and an alignment of paralogous proteins represents a valuable tool in variant evaluation. Amino acids that are functionally equivalent across the different protein family members can be identified. Variants known to be pathogenic in one family member can be used to evaluate an equivalent variant in another family member (14). A paralog alignment of protein kinase subfamily I and II is shown in Fig. S3. None of the missense variants identified in ACVRL1 are listed in the databases for normal variation (dbSNP, 1000Genome and ESP). The affected amino acids in the cytoplasmic domain are evolutionary conserved from human to zebrafish (Fig. S2). None of the nucleotide substitutions are predicted to effect ACVRL1 mRNA splicing. An evaluation of each variant was performed in a three-dimensional structure model of ALK1 (Fig. S4). A summary of the evaluation is found in Table S2 and the evaluation of each variant is given in Appendix S1. Discussion
This is the first report of HHT related mutations from Norway. The findings illustrate the importance of founder mutations in our population. We have showed by haplotyping, supported by the geographical location of the oldest affected family members, that ACVRL1 c.651G>A, c.830C>A and c.1450C>T are most probably founder mutations. These mutations are responsible for HHT in more than one third (36/97) of the families. Knowledge of local founder mutations has been used to simplify diagnostic gene testing in specific populations. Founder mutations may also have a different phenotype than that of the whole group of mutation carriers. It remains to be shown that this is the case in HHT. Finally, founder families may be advantageous to identify modifying factors (15). The number of persons with demonstrated mutations cannot be used directly to establish reliable prevalence figures for HHT in Norway. Taking this into account, our data compare well with the recent report from Denmark indicating that the prevalence in the two countries is about the same (1:6500) (4). We identified a pathogenic/likely pathogenic variant in 86% (97/113) of HHT families, which is in accordance with the findings of others. Due to founder mutations, we have a preponderance of ACVRL1 mutations (58 ACVRL1/39 ENG = 60%). Counting the number of unique mutations, the figure is almost reversed; 22 ACVRL1/30 ENG (42%, ACVRL1 c.11del not included). In 21 individuals with Curacao score = 2, we identified a likely or clearly pathogenic variant, enabling us to set the HHT diagnosis. In several of these individuals, epistaxis and telangectasias may have been present, but not typical for HHT. In addition, in some families, we found a pathogenic variant in an individual presenting with pulmonary AVM, but not admitting typical nosebleeds or displaying telangiectasias. We agree with Tørring et al. (4) that it should be considered to supplement the Curacao criteria with results of genetic testing.
4
Evaluation of DNA variants, especially missense, for pathogenicity may be difficult and must rely on combining information from several sources. Validation was performed based on reports of the frequency of the variant in normal populations, ortholog and paralog conservation, identified pathogenic variants in paralog proteins and predicted effect on protein stability. Of the 14 missense variants identified in ACVRL1, 4 were classified as pathogenic, 6 as likely pathogenic and 4 remained uncertain. Two putative splice-site mutations were identified in ACVRL1 and nine in ENG. One ACVRL1 variant and five ENG variants affected the invariant splice-site sequences (AG–GT). The pathogenicity of variants located elsewhere in the splice sites was validated by mRNA analyses and found to cause various degrees of aberrant splicing. The exact effect of observed partial aberrant splicing in lymphocytes might be difficult to interpretate. However, it should be considered likely that a variant showed to affect splicing of a particular mRNA in lymphocytes from a patient suffering from a disease, known to be caused by the particular gene, represents the pathogenic cause. By thorough evaluation of variants in ENG and ACVRL1, we have identified the cause of HHT in 97 Norwegian families. One of three founder mutations in ACVRL1 was found in approximately one third of the families. Supporting Information Additional supporting information may be found in the online version of this article at the publisher’s web-site.
Acknowledgements The staff at the Department of Medical Genetics in Bergen, Trondheim, and Tromsø have carried out the genetic counseling and diagnostics of the several probands and family members and submitted patient records to the HHT centre in Oslo. The staff at the Department of Medical Genetics in Oslo has carried out the genetic testing for variants in ENG and ACVRL1. Genetic testing for variants in SMAD4 was carried out in the laboratory at the Department of Medical Genetics, St. Olav’s Hospital, Trondheim, Norway. The project was supported by the Regional Core Facility for Structural Biology and Bioinformatics at the South-Eastern Norway Regional Health Authority (grant no. 2012085). All help and support are greatly appreciated.
References 1. Bayrak-Toydemir P, Mao R, Lewin S, McDonald J. Hereditary hemorrhagic telangiectasia: an overview of diagnosis and management in the molecular era for clinicians. Genet Med 2004: 6 (4): 175–191. 2. Plauchu H, Dechadarevian JP, Bideau A, Robert JM. Age-related clinical profile of hereditary hemorrhagic telangiectasia in an epidemiologically recruited population. Am J Med Genet 1989: 32 (3): 291–297. 3. Shovlin CL, Guttmacher AE, Buscarini E et al. Diagnostic criteria for hereditary hemorrhagic telangiectasia (Rendu-Osler-Weber syndrome). Am J Med Genet 2000: 91 (1): 66–67. 4. Torring P, Brusgaard K, Ousager L, Andersen P, Kjeldsen A. National mutation study among Danish patients with hereditary haemorrhagic telangiectasia. Clin Genet 2014: 86: 123–133.
HHT in Norway 5. McDonald J, Damjanovich K, Millson A et al. Molecular diagnosis in hereditary hemorrhagic telangiectasia: findings in a series tested simultaneously by sequencing and deletion/duplication analysis. Clin Genet 2011: 79 (4): 335–344. 6. The HHT Mutation Database, http://arup.utah.edu/database/hht/. 7. Hanks SK. Genomic analysis of the eukaryotic protein kinase superfamily: a perspective. Genome Biol 2003: 4 (5): 111. 8. David L, Mallet C, Mazerbourg S, Feige JJ, Bailly S. Identification of BMP9 and BMP10 as functional activators of the orphan activin receptor-like kinase 1 (ALK1) in endothelial cells. Blood 2007: 109 (5): 1953–1961. 9. Moller P, Heimdal K, Apold J et al. Genetic epidemiology of BRCA1 mutations in Norway. Eur J Cancer 2001: 37 (18): 2428–2434. 10. Brusgaard K, Kjeldsen AD, Poulsen L et al. Mutations in endoglin and in activin receptor-like kinase 1 among Danish patients with hereditary haemorrhagic telangiectasia. Clin Genet 2004: 66 (6): 556–561.
11. Gedge F, McDonald J, Phansalkar A et al. Clinical and analytical sensitivities in hereditary hemorrhagic telangiectasia testing and a report of de novo mutations. J Mol Diagn 2007: 9 (2): 258–265. 12. Olivieri C, Pagella F, Semino L et al. Analysis of ENG and ACVRL1 genes in 137 HHT Italian families identifies 76 different mutations (24 novel). Comparison with other European studies. J Hum Genet 2007: 52 (10): 820–829. 13. Ali BR, Ben Rebeh I, John A et al. Endoplasmic reticulum quality control is involved in the mechanism of endoglin-mediated hereditary haemorrhagic telangiectasia. PLoS One 2011: 6 (10): e26206. 14. Ware JS, Walsh R, Cunningham F, Birney E, Cook SA. Paralogous annotation of disease-causing variants in long QT syndrome genes. Hum Mutat 2012: 33 (8): 1188–1191. 15. Goldstein AM, Chaudru V, Ghiorzo P et al. Cutaneous phenotype and MC1R variants as modifying factors for the development of melanoma in CDKN2A G101W mutation carriers from 4 countries. Int J Cancer 2007: 121 (4): 825–831.
5
© 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd CLINICAL GENETICS doi: 10.1111/cge.12612
Short Report
Mutation analysis in Norwegian families with hereditary hemorrhagic telangiectasia: founder mutations in ACVRL1 Heimdal K., Dalhus B., Rødningen O.K., Kroken M., Eiklid K., Dheyauldeen S., Røysland T., Andersen R., Kulseth M.A. Mutation analysis in Norwegian families with hereditary hemorrhagic telangiectasia: founder mutations in ACVRL1. Clin Genet 2015. © John Wiley & Sons A/S. Published by John Wiley & Sons Ltd, 2015 Hereditary hemorrhagic telangiectasia (HHT, Osler–Weber–Rendu disease) is an autosomal dominant inherited disease defined by the presence of epistaxis and mucocutaneous telangiectasias and arteriovenous malformations (AVMs) in internal organs. In most families (∼85%), HHT is caused by mutations in the ENG (HHT1) or the ACVRL1 (HHT2) genes. Here, we report the results of genetic testing of 113 Norwegian families with suspected or definite HHT. Variants in ENG and ACVRL1 were found in 105 families (42 ENG, 63 ACVRL1), including six novel variants of uncertain pathogenic significance. Mutation types were similar to previous reports with more missense variants in ACVRL1 and more nonsense, frameshift and splice-site mutations in ENG. Thirty-two variants were novel in this study. The preponderance of ACVRL1 mutations was due to founder mutations, specifically, c.830C>A (p.Thr277Lys), which was found in 24 families from the same geographical area of Norway. We discuss the importance of founder mutations and present a thorough evaluation of missense and splice-site variants. Conflict of interest
The authors declare no conflicts of interest.
K. Heimdala , B. Dalhusb,c , O.K. Rødningena , M. Krokena , K. Eiklida , S. Dheyauldeend , T. Røyslandd,e , R. Andersenf and M.A. Kulsetha a Department of Medical Genetics, Oslo University Hospital, Oslo, Norway, b Department for Medical Biochemistry, University of Oslo, Oslo, Norway, c Department for Microbiology, d Department for Otorhinolaryngology, Oslo University Hospital, Oslo, Norway, e Department for Otorhinolaryngology, Innlandet Hospital, Gjøvik, Norway, and f Department of Radiology and Nuclear Medicine, Oslo University Hospital, Oslo, Norway
Key words: ACVRL1 – ENG – founder mutation – mutation – genetic testing – hereditary hemorrhagic telangiectasia – HHT – Osler–Weber–Rendu Corresponding author: Ketil Heimdal, MD, PhD, Department of Medical Genetics, Oslo University Hospital, Oslo Norway. Tel.: +47 2307 5701; fax: +47 2307 5590; e-mail: [email protected] Received 9 February 2015, revised and accepted for publication 11 May 2015
Hereditary hemorrhagic telangiectasia (HHT) is a rare disease defined by epistaxis, the presence of mucocutaneous telangiectasias and arteriovenous malformations (AVMs) in internal organs. Most patients suffer from spontaneous recurrent epistaxis starting in childhood or adolescence. AVMs are most commonly found in lungs (30–50% of patients), central nervous system (10%) and liver (40–75%) (1). Although complications from internal organs are potentially fatal, most patients have a near normal life expectancy. The condition is inherited as an autosomal dominant trait with near full penetrance at the age of 65 (2). Severity is variable both between and
within families. Diagnosis is clinical according to the Curacao criteria (3). A score of 3–4 points (i.e. Curacao criteria positive) has a high sensitivity and specificity for HHT and for detecting mutations in the relevant genes. Detection of a mutation might be helpful when the clinical criteria are equivocal (score of 2 equal to a ‘possible’ diagnosis of HHT) (4). In most families, HHT is caused by mutations in either ENG (HHT1, 39–59% of cases) or ACVRL1 (HHT2, 25–57%) and for a small minority, in SMAD4 (1–2%) or GDF2 (HHT5, A (p.Trp217*), c.830C>A (p.Thr277Lys) and c.1450C>T (p.Arg484Trp) were found in 36 of 97 mutation positive families (37.1%). Altogether, 104 of 234 individuals (44.4%) with HHT with documented mutations in Norway have one of these three variants. Twenty-two of 24 families with the ACVRL1 c.830C>A could be traced to the Rana district in Nordland County. The mutation dominates HHT in this geographical area. All tested families (13 families) with c.830C>A shared a common core haplotype (Table S3). Similarly, ACVRL1 c.651G>A detected in seven families from South-Eastern Norway and families with origin in West Sweden, and all five families with the ACVRL1 c.1450C>T share common haplotypes. One of the latter families had another likely pathogenic ACVRL1 variant, c.11delG located in cis. The most frequent mutation in ENG, c.277C>T (p.Arg93*), is reported repeatedly in other populations (10–12) and in this study found in five families from various locations in South-Eastern Norway. In Norway, this mutation has apparently occurred repeatedly and on different haplotypes (data not shown). Most probably, c.277C>T represents a hot spot for mutations in the ENG gene. Evaluation of the variants found in ENG
Results
In 105 of 113 HHT families, a sequence variant in class 3–5 was identified (Fig. 1a,b and Table S1a,b). Three of these families had an unclassified variant in ENG and five in ACVRL1, giving a detection rate of likely pathogenic or pathogenic variants of 97/113 (85.8%). Forty-two (39+3VUS) families had a variant in ENG and 63 (58+5VUS) in ACVRL1. Sixty unique sequence variants were identified (33 in ENG and 27 in ACVRL1) including 32 novel variants (19 in ENG and 13 in ACVRL1). One family had two ACVRL1 variants in cis (c.1450C>T and c.11delG). Thirty variants in ENG and 23 in ACVRL1 (including c.11delG) were evaluated as likely pathogenic or pathogenic. The distribution according to variant type was as expected with more missense variants in ACVRL1 (45/63 families→71.4%) compared to ENG (8/42 families→19.0%). There were three gross alterations in ENG and none in ACVRL1. In eight families, no mutation in ACVRL and ENG could be identified.
2
Nearly half of the DNA variants found in ENG were nonsense or small out-of-frame indels detected in one or two families each (Table S1a and Fig. 1a). These variants are evaluated as likely pathogenic because haploinsufficiency of ENG has been shown to be an underlying cause of HHT (5). Similarly, the three gross alterations are expected to lead to loss of protein and to be pathogenic. We identified nine splice-site variants. Five of these were located in the invariant splice-site sequence GU (+1 +2) or AG (−1 −2), and expected to cause aberrant splicing. The variants c.67+5G>A, c.68−3C>A, c.360+3G>C and c.361−11T>A were analyzed by reverse transcription polymerase chain reactions (RT-PCR). Complete missplicing was confirmed for c.67+5G>A and partial missplicing was found for c.360+3G>C and c.361−11T>A. Missplicing was found for c.68−3C>A; however, we were unable to determine the degree of missplicing (Fig. S1). We identified seven missense variants in ENG, of which three were novel (p.Cys30Arg, p.Leu80Arg and
HHT in Norway ENG
2
Arg171Pro
3
4
Arg93* Gln110*
Gly331Ser
Arg437Gln
5 6 7 8
9 10
Arg529His
11 12
13
14
Lys585*
Lys237* Tyr120*
nonsense
1
Leu80Arg
Trp261*
c.166dup
c.647del
c.1128dup
c.1385dup
indels /frameshift
Cys30Arg
missense
Arg197Pro
c.1680dup c.1452dup
c.1120_1123del
c.68-3C>G
c.67+1G>A c.67+5G>A
c.360+3G>C
splice site mutations
c.772_774delTACinsA c.817-1G>T
c.360+1G>A
c.361-11T>A
c.1134+1G>A
c.1686+1G>A
del/dup Cys344Tyr
ACVRL1
His328Tyr His314Gln
Ala400Val
1
2
Arg411Trp
Gly211Cys
3
4
5
Arg411Gln
6
7
Trp217* c.11del c.203del
Pro424Arg
8
Val442Ala
9
Arg484Trp
10
Arg479*
Gln272*
c.353_360dup
c.1041del
c.444_445del
c.1042del
c.1431_1432insCACCCGCT
nonsense
Glu236Lys Gly79Arg
missense
Arg374Gln
Thr277Lys
indels /frameshift
c.264del c.314-3C>G c.313+1G>A
splice site mutations
Fig. 1. Sequence variants found in the Norwegian cohort of HHT patients. (a) ENG. (b) ACVRL1. Deletions are shown as a blue bar, duplication as a red bar. Dark yellow = pathogenic variant. Light yellow = likely pathogenic. White = Variant of Uncertain Significance (VUS).
p.Arg171Pro). These variants are located in the orphan N-terminal domain of endoglin, which show no significant homology to any known family of proteins (13) The pathogenic effect of these variants is uncertain. Evaluation of the variants found in ACVRL1
Seven of the novel variants found in ACVRL1 were nonsense, small out-of-frame indels or located in the
consensus splice-site position +1, and are thus (likely) pathogenic (Table S1b and Fig. 1b). RT-PCR confirmed that the novel splice-site mutation c.314-3 located in intron 3 caused missplicing of ACVRL1 mRNA (Fig. S1). These novel variants are likely pathogenic because haploinsufficiency of ACVRL1 is an underlying cause of HHT (5). We identified 14 missense variants in ACVRL1 in our families, 4 of these were novel to this study. One was
3
Heimdal et al. located in the extracellular ligand binding domain and 13 in the cytoplasmic serine/threonine kinase domain. ALK1 belongs in the protein kinase superfamily and an alignment of paralogous proteins represents a valuable tool in variant evaluation. Amino acids that are functionally equivalent across the different protein family members can be identified. Variants known to be pathogenic in one family member can be used to evaluate an equivalent variant in another family member (14). A paralog alignment of protein kinase subfamily I and II is shown in Fig. S3. None of the missense variants identified in ACVRL1 are listed in the databases for normal variation (dbSNP, 1000Genome and ESP). The affected amino acids in the cytoplasmic domain are evolutionary conserved from human to zebrafish (Fig. S2). None of the nucleotide substitutions are predicted to effect ACVRL1 mRNA splicing. An evaluation of each variant was performed in a three-dimensional structure model of ALK1 (Fig. S4). A summary of the evaluation is found in Table S2 and the evaluation of each variant is given in Appendix S1. Discussion
This is the first report of HHT related mutations from Norway. The findings illustrate the importance of founder mutations in our population. We have showed by haplotyping, supported by the geographical location of the oldest affected family members, that ACVRL1 c.651G>A, c.830C>A and c.1450C>T are most probably founder mutations. These mutations are responsible for HHT in more than one third (36/97) of the families. Knowledge of local founder mutations has been used to simplify diagnostic gene testing in specific populations. Founder mutations may also have a different phenotype than that of the whole group of mutation carriers. It remains to be shown that this is the case in HHT. Finally, founder families may be advantageous to identify modifying factors (15). The number of persons with demonstrated mutations cannot be used directly to establish reliable prevalence figures for HHT in Norway. Taking this into account, our data compare well with the recent report from Denmark indicating that the prevalence in the two countries is about the same (1:6500) (4). We identified a pathogenic/likely pathogenic variant in 86% (97/113) of HHT families, which is in accordance with the findings of others. Due to founder mutations, we have a preponderance of ACVRL1 mutations (58 ACVRL1/39 ENG = 60%). Counting the number of unique mutations, the figure is almost reversed; 22 ACVRL1/30 ENG (42%, ACVRL1 c.11del not included). In 21 individuals with Curacao score = 2, we identified a likely or clearly pathogenic variant, enabling us to set the HHT diagnosis. In several of these individuals, epistaxis and telangectasias may have been present, but not typical for HHT. In addition, in some families, we found a pathogenic variant in an individual presenting with pulmonary AVM, but not admitting typical nosebleeds or displaying telangiectasias. We agree with Tørring et al. (4) that it should be considered to supplement the Curacao criteria with results of genetic testing.
4
Evaluation of DNA variants, especially missense, for pathogenicity may be difficult and must rely on combining information from several sources. Validation was performed based on reports of the frequency of the variant in normal populations, ortholog and paralog conservation, identified pathogenic variants in paralog proteins and predicted effect on protein stability. Of the 14 missense variants identified in ACVRL1, 4 were classified as pathogenic, 6 as likely pathogenic and 4 remained uncertain. Two putative splice-site mutations were identified in ACVRL1 and nine in ENG. One ACVRL1 variant and five ENG variants affected the invariant splice-site sequences (AG–GT). The pathogenicity of variants located elsewhere in the splice sites was validated by mRNA analyses and found to cause various degrees of aberrant splicing. The exact effect of observed partial aberrant splicing in lymphocytes might be difficult to interpretate. However, it should be considered likely that a variant showed to affect splicing of a particular mRNA in lymphocytes from a patient suffering from a disease, known to be caused by the particular gene, represents the pathogenic cause. By thorough evaluation of variants in ENG and ACVRL1, we have identified the cause of HHT in 97 Norwegian families. One of three founder mutations in ACVRL1 was found in approximately one third of the families. Supporting Information Additional supporting information may be found in the online version of this article at the publisher’s web-site.
Acknowledgements The staff at the Department of Medical Genetics in Bergen, Trondheim, and Tromsø have carried out the genetic counseling and diagnostics of the several probands and family members and submitted patient records to the HHT centre in Oslo. The staff at the Department of Medical Genetics in Oslo has carried out the genetic testing for variants in ENG and ACVRL1. Genetic testing for variants in SMAD4 was carried out in the laboratory at the Department of Medical Genetics, St. Olav’s Hospital, Trondheim, Norway. The project was supported by the Regional Core Facility for Structural Biology and Bioinformatics at the South-Eastern Norway Regional Health Authority (grant no. 2012085). All help and support are greatly appreciated.
References 1. Bayrak-Toydemir P, Mao R, Lewin S, McDonald J. Hereditary hemorrhagic telangiectasia: an overview of diagnosis and management in the molecular era for clinicians. Genet Med 2004: 6 (4): 175–191. 2. Plauchu H, Dechadarevian JP, Bideau A, Robert JM. Age-related clinical profile of hereditary hemorrhagic telangiectasia in an epidemiologically recruited population. Am J Med Genet 1989: 32 (3): 291–297. 3. Shovlin CL, Guttmacher AE, Buscarini E et al. Diagnostic criteria for hereditary hemorrhagic telangiectasia (Rendu-Osler-Weber syndrome). Am J Med Genet 2000: 91 (1): 66–67. 4. Torring P, Brusgaard K, Ousager L, Andersen P, Kjeldsen A. National mutation study among Danish patients with hereditary haemorrhagic telangiectasia. Clin Genet 2014: 86: 123–133.
HHT in Norway 5. McDonald J, Damjanovich K, Millson A et al. Molecular diagnosis in hereditary hemorrhagic telangiectasia: findings in a series tested simultaneously by sequencing and deletion/duplication analysis. Clin Genet 2011: 79 (4): 335–344. 6. The HHT Mutation Database, http://arup.utah.edu/database/hht/. 7. Hanks SK. Genomic analysis of the eukaryotic protein kinase superfamily: a perspective. Genome Biol 2003: 4 (5): 111. 8. David L, Mallet C, Mazerbourg S, Feige JJ, Bailly S. Identification of BMP9 and BMP10 as functional activators of the orphan activin receptor-like kinase 1 (ALK1) in endothelial cells. Blood 2007: 109 (5): 1953–1961. 9. Moller P, Heimdal K, Apold J et al. Genetic epidemiology of BRCA1 mutations in Norway. Eur J Cancer 2001: 37 (18): 2428–2434. 10. Brusgaard K, Kjeldsen AD, Poulsen L et al. Mutations in endoglin and in activin receptor-like kinase 1 among Danish patients with hereditary haemorrhagic telangiectasia. Clin Genet 2004: 66 (6): 556–561.
11. Gedge F, McDonald J, Phansalkar A et al. Clinical and analytical sensitivities in hereditary hemorrhagic telangiectasia testing and a report of de novo mutations. J Mol Diagn 2007: 9 (2): 258–265. 12. Olivieri C, Pagella F, Semino L et al. Analysis of ENG and ACVRL1 genes in 137 HHT Italian families identifies 76 different mutations (24 novel). Comparison with other European studies. J Hum Genet 2007: 52 (10): 820–829. 13. Ali BR, Ben Rebeh I, John A et al. Endoplasmic reticulum quality control is involved in the mechanism of endoglin-mediated hereditary haemorrhagic telangiectasia. PLoS One 2011: 6 (10): e26206. 14. Ware JS, Walsh R, Cunningham F, Birney E, Cook SA. Paralogous annotation of disease-causing variants in long QT syndrome genes. Hum Mutat 2012: 33 (8): 1188–1191. 15. Goldstein AM, Chaudru V, Ghiorzo P et al. Cutaneous phenotype and MC1R variants as modifying factors for the development of melanoma in CDKN2A G101W mutation carriers from 4 countries. Int J Cancer 2007: 121 (4): 825–831.
5
