J Cancer Res Clin Oncol (2009) 135:1463–1470 DOI 10.1007/s00432-009-0594-4

ORIGINAL PAPER

APC mutation spectrum of Norwegian familial adenomatous polyposis families: high ratio of novel mutations Per Arne Andresen · Ketil Heimdal · Kristin Aaberg · Kristin Eklo · Sarah Ariansen · Alexandra Silye · Olav Fausa · Lars Aabakken · Stefan Aretz · Tor J. Eide · Tobias Gedde-Dahl Jr.

Received: 15 September 2008 / Accepted: 21 April 2009 / Published online: 15 May 2009 © Springer-Verlag 2009

Abstract Introduction Familial adenomatous polyposis (FAP) is an autosomal dominantly inherited disease caused by mutations in the adenomatous polyposis coli (APC) gene. Massive formation of colorectal adenomas, of which some will inevitably develop into adenocarcinomas, is the hallmark of the disease. Characterization of causative APC mutations allows presymptomatic diagnosis, close followup and prophylactic intervention in families. To date more than 900 diVerent germline mutations have been characterized worldwide demonstrating allelic heterogeneity. Purpose The germline mutation spectrum of APC identiWed in 69 apparently unrelated Norwegian FAP families are presented and discussed with reference to clinical phenotype and novel mutation rate. This paper is dedicated to late Professor Tobias Gedde-Dahl (deceased in March 2006). P. A. Andresen (&) · S. Ariansen · A. Silye · T. J. Eide Pathology Division, University Hospital of Oslo-Rikshospitalet, N0027 Oslo, Norway e-mail: [email protected]

Methods DiVerent methods have been used over the years. However, all mutations were conWrmed detectable by an implemented denaturing high-performance liquid chromatography screening approach. Multiplex ligationdependent probe ampliWcation analysis was employed for potential gross rearrangements. Results Fifty-three distinctive mutations were detected, of which 22 have been detected in Norway exclusively. Except for two major deletion mutations encompassing the entire APC, all mutations resulted in premature truncation of translation caused by non-sense (31%) or change in reading frame (69%). Conclusion A high ratio of novel APC mutations continues to contribute to APC mutation heterogeneity causing FAP. This is the Wrst comprehensive report of APC germline mutation spectrum in Norway. Keywords Adenomatous polyposis coli · Familial adenomatous polyposis · Novel mutations · Germline mutations · DHPLC

Introduction K. Heimdal Department of Medical Genetics, University Hospital of Oslo-Rikshospitalet, N0027 Oslo, Norway K. Aaberg · K. Eklo Department of pathology, University Hospital of Northern Norway, N9037 Tromsø, Norway O. Fausa · L. Aabakken Department of medicine, University Hospital of Oslo-Rikshospitalet, N0027 Oslo, Norway S. Aretz Institute of Human genetics, University of Bonn, 53111 Bonn, Germany

Familial adenomatous polyposis (FAP) (MIM#175100) is a rare autosomal dominant precancerous disease characterized by hundreds to thousands of polyps. In nearly 100% of the cases some of the polyps develop into colorectal cancer if left untreated. The vast majority of FAP is caused by germline mutations in the adenomatous polyposis coli (APC) tumor suppressor gene (Groden et al. 1991; Kinzler et al. 1991). Based on number of polyps and age of onset FAP is divided into two main groups; classical FAP (or only FAP) and attenuated FAP (AFAP). Classical FAP is characterized by the development of hundreds to more than

123

1464

thousand colorectal adenomas early in life, normally before puberty (Bülow 1986). Attenuated FAP has a less overt phenotype with a smaller number of adenomas (<100) generally by their fourth or Wfth decade, and hence a later age of onset of colorectal cancer (mean age 55 years) (Knudsen et al. 2003). The APC gene occurs in multiple forms and the most discussed transcript relates to a protein of 2,843 amino acids. This isoform consists of 14 small 5⬘exons (range 84–379 bp) and a large 3⬘ exon 15 (6,571 bp) representing 75% of the entire coding sequence. The majority of APC mutations result in COOH-terminally truncated proteins (98%). These are dominated by frameshift mutations due to small deletions or insertions (20 bp) and non-sense mutations. More than 900 diVerent mutations have been registered in The Human Gene Mutation Database (www.hgmd.cf.ac.uk/ac). The apparently arbitrarily distribution of the mutations throughout the gene (codons 99– 2,504) requires an eYcient screening approach. Mutation screening by denaturing high-performance liquid chromatography (DHPLC) has proven successful and is more sensitive than screening by single-strand conformational polymorphism, denaturing gradient gel electrophoresis, protein truncation test and other widely used, but less eYcient and reliable methods (Bennett et al. 2001; Wu et al. 2001). The majority of the mutations in the present study was originally detected by one of these methods and have all later been conWrmed detectable by the presented DHPLC approach modiWed after Wu et al. (2001). Here we report for the Wrst time the comprehensive germline mutation spectrum of APC in Norwegian FAP families of which 22 (41.5%) so far have been identiWed exclusively in Norway.

Materials and methods

J Cancer Res Clin Oncol (2009) 135:1463–1470

PCR conditions Primers were according to Wu et al. (2001) except for primers used to amplify exons 5, 6, 13 and 15-7 which were replaced to improve detection sensitivity. New primers were: exon 5-forward: 5⬘-CTGATTAACGTAAATACAA GATATTGATACT-3⬘; exon 5-reverse: 5⬘-CAGAGCTGT AATTCATTTTATTCCT-3⬘; exon 6-forward: 5⬘-TGAAAGAACATCTA TATTTCAATGGTG-3⬘; exon 6-reverse: 5⬘-TAATATTATTAATAAAAACATAACTAAT TAGG TTTC-3⬘; exon 13-forward: 5⬘-AGCAACTAGTAT GATTTTATGTATAAATTAATC T-3⬘; exon13-reverse: 5⬘-AAGCCAAACATGAAATTCATATTATAGTACT-3⬘; exon 15-7-forward: 5⬘-TTCAGGAGACCCCACTCAT-3⬘ and 15-7-reverse: 5⬘-CATTTGATTCTTTAG GCTGCT CTGATT-3⬘. An error, a missing T (underlined) in the reverse 15-7 primer (c.4613–4639) published by Wu et al. was corrected. Exons were ampliWed from 20 ng DNA template in detergent-free HF® buVer with Phusion® DNA Polymerase (Finnzymes, Espoo, Finland) according to the manufacturer’s instructions. The products were ampliWed by an initial denaturizing step at 98°C for 30 s, Wve cycles at 98°C for 20 s, 63°C for 30 s and 72°C for 20 s, followed by 30 similar cycles, but with annealing temperature lowered to 58°C. A Wnal extension for 1 min at 72°C ended the ampliWcation. Denaturing high-performance liquid chromatography Screening for heterozygous fragments was carried out by DHPLC analysis on a semi-automated 3500 High Throughput Wave System (Transgenomics WAVE™, UK). Elution gradient and oven temperatures were predicted by WAVEMAKER Software version 4.1 (Transgenomics).

Patients Sequencing The majority of FAP patients were originally enrolled in a national screening program initiated in 1978 (Gedde-Dahl Jr et al. 1988). Over the years additional patients have been referred for molecular genetic analyses through gastroenterologists and medical geneticists. Referral for analysis was based on clinical detection of numerous of colorectal adenomas (range 20–1,000), or an unambiguous family history of FAP. All patients provided an informed consent; 5 ml of EDTA-blood samples was collected for genetic analysis. DNA puriWcation DNA was puriWed from 200
l whole blood using the Qiagen blood DNA mini-kit™ according to the manufacturer’s instructions (Qiagen Inc.).

123

PCR products indicating heterozygosity were puriWed (Exo-Sap IT, GE Healthcare Europe GMBH, Germany), and sequenced in both directions according to manufacturer’s instruction (ABI BigDye® Terminator Cycle Sequencing Kit, version 3.1, on the ABI3130x; Applied Biosystems). Sequencing primers were the same as for the ampliWcation reactions prior to DHPLC. Nomenclature of the characterized mutations is according to recommendations given by the Human Genome Variation Society (www.hgvs.org). If any of the detected mutations have been registered in the Human Genome Mutation Database (HGMD; www.hgmd.cf.ac.uk/ac), HGMD is the only given reference for recurrent mutations (Table 2) if not discussed in any further detail.

J Cancer Res Clin Oncol (2009) 135:1463–1470

Multiplex ligation-dependent probe ampliWcation Search for larger deletions was conducted by multiplex ligation-dependent probe ampliWcation (MLPA). We employed an MLPA kit (no. P043, MRC Holland, Amsterdam, The Netherlands) dedicated to deletion analysis of the APC locus 5q21-q22.

Results The spectrum of 53 mutations distributed among 69 apparently unrelated families is shown in Tables 1 and 2 presenting 22 novel and 31 recurrent mutations, respectively. All patients had a conWrmed or suspected FAP diagnosis. DeWning novel mutations as original and exclusively detected in Norway, the ratio between novel and recurrent mutations appears to be 22:31 (0.71), representing a novel mutation frequency of 41.5%. Overall, small insertions or deletions of <20 bp dominated (56.6%), followed by single base substitutions resulting in non-sense mutations (28.3%). Two families had combined small insertions and deletions (indels), and one family had a gross (35 bp) insertion. Two mutations with the potential of aVecting splicing were detected, both segregating with FAP and also identiWed as cause of FAP in a previous Israeli (c.1312 + 1G > A; Gavert et al. 2002) and German study (c.1957A > G; Aretz et al. 2004), respectively. The intronic c.1312 + 1G > A transition aVects splicing of exon 9 predicting production of truncated APC—the Wrst aVected amino acid being at codon 312 before premature stop (p.Val312CysfsX16). Likewise, the other splice-site aVection is associated with c.1957A > G within exon 14 suggesting skipping and splicing between exon 13 and 15 directly. Consequently, amino acids are aVected from codon 543 (p.Val543AlafsX20), ahead of the actual mutation at codon 653. Functional analysis of this particular variant supports this scenario (Aretz et al. 2004). The last mutation to have a deleterious eVect on APC was the deletion of exons 11–13 allowing splicing directly between exon 10 and 14. This predicts erroneous amino acids from codon 471 and premature stop 19 codons downstream (p.Gly471IlefsX19). Taken together, 51 of the 53 (96%) identiWed mutations predicted truncated APC proteins contributing to variable degrees of dominant negative interaction with wild-type allele. The remaining two families represent major deletions indicated through MLPA analysis and supported by haplotyping. One of these was visible through cytogenetic analysis in a patient with classical FAP and mental retardation. No further characterizations were carried out in these families at this point. No mutations were detected in exons 1, 2, 4, 7, 8 or 12. The most 5⬘ mutation was a recurrent single nucleotide deletion (c.417delA) in codon 139 resulting in

1465

p.Glu140ArgfsX30, while the most 3⬘ was a recurrent twobase pair deletion (c.4733_4734delGT) resulting in p.Cys1578TyrfsX12. Nearly 70% (69.9%) of the families representing 67.9% of the mutations were located within the Wrst half of exon 15. The PCR/DHPLC fragments 15-4 (codon 1,026–1,163) and 15-6 (codon 1,258–1,389) alone, represented 44% of the mutations. Five apparently unrelated families (7.3%) had the frequent 5 bp deletion of codon 1,061, while eight families (11.6%) had the 5-bp deletion of codon 1,309. In our series of families putative attenuated FAP were seen in four families. One was associated to the most 5⬘ mutation in codon 139, although the Wrst amino acid change is in codon 140. Another was uncommonly associated to a deletion of the entire APC, and the remaining two to the most 3⬘ frameshift mutations in codons 1,552 and 1,578, respectively. Classical phenotypes were seen associated to mutations from codon 182 to codon 1,549. The various mutations and corresponding phenotypes were plotted and discussed against a genotype phenotype correlation scale (Fig. 1) anchored in previous publications reviewing such correlations (Galiatsatos and Foulkes 2006; Nieuwenhuis and Vasen 2007).

Discussion Novel mutations Fifty-three dissimilar mutations were disclosed among 69 apparently unrelated Norwegian FAP families. Twenty-two were deWned as novel, 31 as recurrent (Tables 1, 2, respectively). Two of the novel mutations have been described in a previous report (Norheim Andersen et al. 1999). As a reference base for validation of newly discovered mutations, HGMD (www.hgmd.cf.ac.uk/ac) is useful, but not necessarily adequate. Four mutations among our FAP families were grouped as recurrent, even not registered in HGMD (Table 2). As a consequence, the proportion of novel mutations reached 41.5% (22/53), not unlike the proportion (44.3%) reported in a recent Swedish study (Kanter-Smoler et al. 2008). Some of the recurrent mutations, like c.3151delA, cannot be excluded to have originated in other countries. This mutation was reported as novel in a previous Swedish report (Björk et al. 2001). Another mutation (c.1312 + 1G > A) had a FAP history dating back to midEuropean ancestors, although an identical mutation has been reported in an Israeli study only (Gavert et al. 2002). Apart from the frequent 1,061 and 1,309 5 bp deletions, Wve other mutations were shared by apparently unrelated families; p.Gln203X, p.Arg213X, p.Arg564X and p.Tyr935X and p.Arg414ThrfsX5. The various non-sense mutations were all categorized as recurrent. DiVerent origins

123

1466

J Cancer Res Clin Oncol (2009) 135:1463–1470

Table 1 Mutations categorized as novel (N = 22) Family

Exon(s)

Mutationa

Predicted consequence

FAP phenotype, extracolonic tumor/disease

Ref

01, 02

9

c.1239dupA

p.Arg414ThrfsX5

Classical, hepatocellular ca, pancreatic cancer, duodenal cancer

Current

03

10

c.1364_1365delAA

p.Lys455ThrfsX4

Classical

04

11

c.1473_1474ins35b

p.His492TyrfsX18

Classical, duodenal cancer

Current

05

15-2

c.2379_2380delAA

p.Tyr796TrpfsX2

Classical, duodenal cancer

Norheim Andersen et al. (1999)

06

15-2

c.2404_2407delGACA

p.Asp802ProfsX17

Classical, carcinoids in colon

Current

07

15-2

c.2482_2483dupAC

p.Thr829LeufsX14

Classical, adv. duodenal adenomatosis

Current

08

15-3

c.2815_2816delAA

p.Lys939ValfsX6

Classical, adv. duodenal adenomatosis

Current

09

15-3

c.2819C > A

p.Ser940X

Classical, adv. duodenal adenomatosis

Current

10

15-4

c.3082delA

p.Ser1028ValfsX9

Classical, osteomas, duodenal cancer, hepatoblastoma

Current

11

15-4

c.3213_3217delAAGTA/ c.3213insT

p.Gln1071HisfsX54

Classical, desmoids

Current

12

15-4

c.3345_3347delGGG/ c.3345insTTTGT

p.Gly1116LeufsX11

Classical

Current

13

15-4

c.3408delA

p.Asp1137MetfsX28

Classical, adrenal tumor

Current

14

15-4

c.3466G > T

p.Glu1156X

Classical

Current Current

15

15-6

c.4041dupC

p.Arg1348GlnfsX6

Classical, duodenal cancer

16

15-6

c.4132C > T

p.Gln1378X

Classical, rectal cancer

17

15-6

c.4152dupT

p.Ser1385X

Classical, osteomas

Current

18

15-7

c.4393_4394dupAG

p.Ser1465ArgfsX9

Classical

Norheim Andersen et al. (1999)

19

15-7

c.4415delT

p.Val1472GlufsX35

Classical, rectal cancer

Current

20

15-7

c.4463T > G

p.Leu1488X

Classical

Current

21

15-8

c.4646_4647delAA

p.Gln1549ArgfsX9

Classical, osteoma, exostosis, desmoid

Current

22

DelAPC

MLPA indicated cyt.; Del5q21-22

No details available

Classical, muscular atrophy, mental retardation

Current

23

DelAPC

MLPA indicated

No details available

Attenuated

Current

a b

According to the NCBI database transcript NM_000038.3 (gi: 21626462) Insertion of 5⬘-TACAGTATTACACTAAGTATTACACTACAGTATTA-3⬘

of the p.Gln203X (c.607C > T) mutation detected in two apparently unrelated families is not unlikely, as an identical mutation has been identiWed somatically in a non-FAP colorectal tumor (Olschwang et al. 1997). Lack of relationship between the two cases of p.Arg213X is reasonable, because it has earlier been described in other reports (Miyoshi et al. 1992; Giarola et al. 1999). Apparently, both families presented with the mutations de novo. The same goes for p.Arg564X and p.Tyr935X (Fodde et al. 1992; Wallis et al. 1999). Finally, a mutation categorized as novel (c.1239dupA) was seen in two families from the same geographical region. A common origin of the mutation cannot be excluded unless haplotyping shows otherwise.

123

Mutation spectrum The huge spectrum of mutations in APC demands sensitive and versatile methods of detection as oVered by DHPLC. All base substitutions, minor deletions and/or insertions— the largest in this study being a 35 bp insertion in exon 11 (c.1473_1474ins35)—were proven detectable by DHPLC. Three large deletions were indicated by MLPA after having appeared negative by DHPLC. Two encompassed the entire APC coding sequence. The third was limited to deletion of exons 11–13. Direct bridging between exons 10 and 14 was conWrmed by sequencing of cDNA. An identical mutation (c.1409-?_1743 + ?del) was previously reported in a Swiss

c.4733_4734delGT

c.4655_4656delAG

c.4390_4393delGAGA

c.4384_4385delAA

c.4348C > T

c.4099C > T

c.4016delG

c.3927_3931delAAAGA

c.3523C > T

c.3471_3474delGAGA

c.3366_3369delTCAA

c.3329C > G

c.3202_3205delTCAA

c.3183_3187delACAAA

c.3151delA

c.3063dupA

c.2805C > A

c.2677G > T

c.2493dupA

c.1957A > G

c.1863_1866delTTAC

c.1775T > A

c.1690C > T

c.1409-?_1743 + ?del

c.1312 + 1G > A

c.1100_1101delCT

c.646C > T

c.637C > T

c.607C > T

c.543_546delAACA

c.417delA

Mutationa

p.Cys1578TyrfsX12

p.Glu1552GlyfsX6

p.Glu1464ValfsX8

p.Lys1462GlufsX6

p.Arg1450X

p.Gln1367X

p.Gly1339ValfsX76

p.Glu1309AspfsX4

p.Gln1175X

p.Glu1157AspfsX7

p.Asn1122LysfsX3

p.Ser1110X

p.Ser1067GlyfsX57

p.Gln1062X

p.Arg1051AspfsX5

p.Asp1022ArgfsX7

p.Tyr935X

p.Glu893X

Attenuated, osteoma, desmoid

Attenuated, osteoma, desmoid, adrenal tumor

Classical, osteoma, desmoid

Classical, desmoid, duodenal cancer

Classical

Classical

Classical, duodenal cancer

Classical, duodenal cancer, desmoids, renal cell cancer

Classical, duodenal cancer

Classical, duodenal cancer

Osteoma, epidermoid cysts, hepatocellular ca, adv. duodenal adenomatosis

Classical

Classical

Classical, desmoids

Classical, desmoids

Classical

Classical, osteoma

Classical, duodenal cancer

Classical

Variable

p.Arg653Gly (p.Val543AlafsX20)b p.Pro832ThrfsX12

Classical, adv. duodenal adenomatosis

Classical, adv. duodenal adenomatosis

p.Tyr622GlyfsX7

p.Leu592X

Classical

Classical, duodenal cancer

p.Arg564X

Classical

(p.Val312CysfsX16)b

Classical

Classical

Classical, duodenal cancer, desmoids

Classical

Classical

Attenuated, ovarial cancer

FAP phenotype, extracolonic tumor/disease

(p.Gly471IlefsX19)b

p.Ser367CysfsX10

p.Arg216X

p.Arg213X

p.Gln203X

p.Thr182IlefsX2

p.Glu140ArgfsX30

Predicted consequence

HGMD: www.hgmd.cf.ac.uk/ac; SW Swedish origin, NL Dutch origin a According to the NCBI database transcript NM_000038.3 (gi: 21626462) b Prediction of functional consequences c Identical sequence variant detected at Laboratory for Diagnostic Genome Analysis (LDGA), Leiden University Medical Center: www.lumc.nl/4080/dna/APC_appendix1.pdf

15-8

15-7

66

69

15-7

65

15-7

15-6

64

15-8

15-6

63

68

15-6

55; 56; 57; 58; 59; 60; 61; 62

67

15-4

52

15-5

15-4

51

54

15-4

50

53

15-4

15-4

45; 46; 47; 48; 49

15-4

15-4

44SW

15-3

41, 42

43

15-2

15-3

14 (del)

38

40

14

37

39

14

36

9

31

13

6

30

34; 35

5

28; 29

9 (del)

5

26; 27

11-13 (del)

5

25

32NL

3

24

33

Exon(s)

Family

Table 2 Mutations categorized as recurrent (N = 31)

HGMD

HGMD

HGMD

Leidenc

HGMD

HGMD

Kanter-Smoler et al. (2008)

HGMD

HGMD

HGMD

HGMD

HGMD

HGMD

HGMD

HGMD

HGMD

HGMD

HGMD

HGMD

HGMD

HGMD

Friedl and Aretz (2005)

HGMD

HGMD

HGMD

HGMD

HGMD

HGMD

Kanter-Smoler et al. (2008)

HGMD

HGMD

Ref

J Cancer Res Clin Oncol (2009) 135:1463–1470 1467

123

1468 Fig. 1 Norwegian APC mutations positioned related to functional domains (above) and codon positions correlating to reported phenotype (below)

J Cancer Res Clin Oncol (2009) 135:1463–1470

Homology domain Armadillo region oligomerisation

β -catenin binding 15 aa repeat

β -catenin binding & downregulation

20 aa repeat

axin binding

EB1 basic hDLG domain binding binding

RECURRENT NOVEL Attenuated

Classical

0

600 800 1000 1200 1400 1600 1800 2000 2200 2400 2600 2800

200

400

Profuse

Desmoid

Attenuated

SYMBOL MUTATIONS: Recurrent: Nonsense and small deletion/insertion: Recurrent: Functional consequence, splicing affection: large partial deletion: Novel: Nonsense and small deletion/insertion mutations:

study (Andreutti-Zaugg et al. 1999). Whether identical at genome level, remains to be seen. The frequent Wve base pair deletions at codons 1,061 and 1,309 detected in 5 (7.3%) and eight families (11.6%), respectively, is in accordance with other reports (Miyoshi et al. 1992; Vandrovcova et al. 2004), although signiWcantly higher and lower proportions have been reported (Nagase and Nakamura 1993; De Rosa et al. 2004; GarcíaLozano et al. 2005). Since the frequent 5 bp deletions of 1061 and 1309 are straightforwardly detected irrespective of methods being applied, the ratio between these two should represent solid indicators of ethnic variations in mutation spectra. When related to other mutations, the sensitivity of methods could have inXuenced on the reported frequencies. Variations in mutation diversity, de novo mutation susceptibility, and the ratio of novel mutations, appear to be associated to ethnic and/or environmental diVerences. Five of the eight (62.5%) apparently unrelated families with 1309 deletions were classiWed as de novo mutations, contributing to both high de novo mutation frequencies and overrepresentation compared to other mutations. Fifteen base substitutions, of which eight (53%) were C > T transitions, were identiWed. Nearly two-thirds of the chain terminating mutations caused by base substitutions and minor deletions and/or insertions were located in the Wrst part of exon 15. This is in line with results from other comprehensive studies (Wallis et al. 1999; Friedl and Aretz 2005). Genotype–phenotype correlation The existence of APC genotype and FAP phenotype correlations has been well documented. Figure 1 illustrates an overview of such correlations based on previous reports

123

(for reviews Galiatsatos and Foulkes 2006; Nieuwenhuis and Vasen 2007). When classiWed into classical and attenuated FAP, associated genotype could be instrumental to guidance of follow-up and treatment. However, if divided further into profuse or severe polyposis, and sparse or intermediate FAP, associated genotypes do not hold the consistency for being clinically useful. Instead it might be more relevant to pay attention to family history. Desmoid tumors are typical traits of FAP. Even infrequent (15%) and locally secluded, they represent the second leading cause of death in FAP patients apart from colorectal cancer (Arvanitis et al. 1990). Risk has been associated to mutations between codons 1,445 and 1,580 (Galiatsatos and Foulkes 2006). In our study, Wve out of nine families with mutations beyond codon 1,445 reported incidents of desmoid tumors. These lesions were also seen associated to mutations at codons 213, 1,051, 1,071, and the frequent 5 bp deletions of 1,061 and 1,309 (Tables 1, 2). This clearly underscores the necessity of approaching risk assessment independent of genotype. Likewise, risk of duodenal adenomas and cancer have been tried pinched to speciWc regions of APC (for review Nieuwenhuis and Vasen 2007). In our study, advanced duodenal aVection and death of duodenal cancer were reported in 18 unrelated families with mutations spanning from codon 213 through codon 1,464. This is a relatively high frequency indicating that risk of duodenal adenomas is likely to be inXuenced by factors additional to APC, as have been suggested by others (Enomoto et al. 2000; Matsumoto et al. 2002). A variable FAP phenotype was seen in a family representing the most 5⬘end mutation (c.417delA) in this study, within a region associated to attenuated FAP (Friedl et al. 2001). Apparently a de novo mutation, the patient was diagnosed as classical FAP based on hundreds of polyps

J Cancer Res Clin Oncol (2009) 135:1463–1470

and colorectal cancer at the age of 57. An aVected oVspring presented with <100 polyps around the age of 20. Age of onset and diagnosis suggest that this family as a whole should have been categorized as AFAP. Attenuated phenotypes caused by early 5⬘ mutations have been associated to an alternative translational start signal (ATG) at codon 184 (Heppner Goss et al. 2002). A more distinct variation in phenotype was associated to a transition of the second last nucleotide (c.1957A > G) of exon 14. In one branch of the family, members presented with hundreds of polyps, while in another, members had consistently <100 polyps at a comparable age. In contrast, the family described by Aretz et al. was associated to classical FAP. Functional analysis revealed that the aberrant transcript appeared to be dominant of the normally spliced variant. Interestingly, a non-synonymous variant at the same codon, but substituting arginine with lysine (p.Arg653Lys) caused by transition from G to A in the last nucleotide of exon 14 (AG|G > AA|G; vertical line denotes exon 14–15 boundary) has been described (Azzopardi et al. 2008). The case was associated to more than 100 colorectal adenomas and discussed as non FAP and solely in the context of the amino acid substitution, not related to splicing distortion. Patients with frameshift mutations in codons 1552 and 1578, the two most 3⬘ end mutations in this study, were phenotypically categorized as atypical or attenuated FAP. Both mutations are proximal to the region normally associated to attenuated FAP, but within a region representing a borderline between classical and attenuated FAP and where an intermediate phenotype have been discussed (Nieuwenhuis and Vasen 2007). Because both these 3⬘ mutations occurred de novo it cannot be excluded that deviations from predicted genotype can be attributed to the presence of somatic mosaicism. The present report is the Wrst comprehensive presentation of APC germline mutations in Norway. Even if more than 900 diVerent germline mutations have been reported to date, this study conWrms that a relatively high proportion of novel mutations in national studies continue to contribute to expand heterogeneity of APC mutations. Acknowledgments This work was supported by the Norwegian Cancer Society. Highly appreciated is the contribution of late Professor Tobias Gedde-Dahl Jr who initiated the Norwegian Polyposis Project and the establishment of a national register at the Norwegian Cancer Register dedicated to Norwegian FAP families (www.kreftregisteret.no/en/The-Registries/Clinical-registries). ConXict of interest statement

None.

References Andreutti-Zaugg C, Couturier A, Chappuis P, Hutter P (1999) Detection of protein truncating mutations in exons 1–14 of the

1469 APC gene using an in vivo fusion protein assay. Hum Mutat 13:170–171. doi:10.1002/(SICI)1098-1004(1999)13:2<170:: AID-HUMU13>3.0.CO;2-4 MIB online no. 214 Aretz S, Uhlhaas S, Sun Y, Pagenstecher C, Mangold E, Caspari R, Moslein G, Schulmann K, Propping P, Friedl W (2004) Familial adenomatous polyposis: aberrant splicing due to missense or silent mutations in the APC gene. Hum Mutat 24:370–380. doi:10.1002/humu.20087 Arvanitis ML, Jagelman DG, Fazio VW, Lavery IC, McGannon E (1990) Mortality in patients with familial adenomatous polyposis. Dis Colon Rectum 33:639–642. doi:10.1007/BF02150736 Azzopardi D, Dallosso AR, Eliason K, Hendrickson BC, Jones N, Rawstorne E, Colley J, Moskvina V, Frye C, Sampson JR, Wenstrup R, Scholl T, Cheadle JP (2008) Multiple rare nonsynonymous variants in the adenomatous polyposis coli gene predispose to colorectal adenomas. Cancer Res 68:358–363. doi:10.1158/00085472.CAN-07-5733 Bennett RR, den Dunnen J, O’Brien KF, Darras BT, Kunkel LM (2001) Detection of mutations in the dystrophin gene via automated DHPLC screening and direct sequencing. BMC Genet 2:17 Epub Oct 17 Björk J, Akerbrant H, Iselius L, Bergman A, Engwall Y, Wahlstrom J, Martinsson T, Nordling M, Hultcrantz R (2001) Periampullary adenomas and adenocarcinomas in familial adenomatous polyposis: cumulative risks and APC gene mutations. Gastroenterology 121:1127–1135. doi:10.1053/gast.2001.28707 Bülow S (1986) Clinical features in familial polyposis coli. Results of the Danish Polyposis Register. Dis Colon Rectum 29:102–107. doi:10.1007/BF02555389 De Rosa M, Dourisboure RJ, Morelli G, Graziano A, Gutiérrez A, Thibodeau S, Halling K, Avila KC, Duraturo F, Podesta EJ, Izzo P, Solano AR (2004) First genotype characterization of Argentinean FAP patients: identiWcation of 14 novel APC mutations. Hum Mutat 23:523–524. doi:10.1002/humu.9237 Enomoto M, Konishi M, Iwama T, Utsunomiya J, Sugihara KI, Miyaki M (2000) The relationship between frequencies of extracolonic manifestations and the position of APC germline mutation in patients with familial adenomatous polyposis. Jpn J Clin Oncol 30:82–88. doi:10.1093/jjco/hyd017 Fodde R, van der Luijt R, Wijnen J, Tops C, van der Klift H, van Leeuwen-Cornelisse I, GriYoen G, Vasen H, Khan PM (1992) Eight novel inactivating germ line mutations at the APC gene identiWed by denaturing gradient gel electrophoresis. Genomics 13:1162– 1168. doi:10.1016/0888-7543(92)90032-N Friedl W, Aretz S (2005) Familial adenomatous polyposis: experience from a study of 1164 unrelated german polyposis patients. Hered Cancer Clin Pract 3:95–114 Friedl W, Caspari R, Sengteller M, Uhlhaas S, Lamberti C, Jungck M, Kadmon M, Wolf M, Fahnenstich J, Gebert J, Moslein G, Mangold E, Propping P (2001) Can APC mutation analysis contribute to therapeutic decisions in familial adenomatous polyposis? Experience from 680 FAP families. Gut 48:515–521. doi:10.1136/gut.48.4.515 Galiatsatos P, Foulkes WD (2006) Familial adenomatous polyposis. Am J Gastroenterol 101:385–398. doi:10.1111/j.1572-0241.2006.00375.x García-Lozano JR, Cordero C, Fernández-Suárez A, Encarnación M, Pizarro A, Núñez-Roldán A (2005) APC germ-line mutations in southern Spanish patients with familial adenomatous polyposis: genotype–phenotype correlations and identiWcation of eight novel mutations. Genet Test 9:37–40. doi:10.1089/gte.2005.9.37 Gavert N, Yaron Y, Naiman T, Bercovich N, Rozen P, Shomrat R, Legum C, Orr-Urtreger A (2002) Molecular analysis of the APC gene in 71 Israeli families: 17 novel mutations. Hum Mutat 19:664 MIB#508 Gedde-Dahl T Jr, Heim S, Lothe R, Bye R, Brevik K, Geitvik GA, Kyrkjebo HT, Hognestad J, Nygaard K, Bergan A, Olaisen B,

123

1470 Kildal S, Svanes KH (1988) The polyposis project. Tidsskr Nor Laegeforen 108:2465–2468 Article in Norwegian Giarola M, Stagi L, Presciuttini S, Mondini P, Radice MT, Sala P, Pierotti MA, Bertario L, Radice P (1999) Screening for mutations of the APC gene in 66 Italian familial adenomatous polyposis patients: evidence for phenotypic diVerences in cases with and without identiWed mutation. Hum Mutat 13:116–123. doi:10.1002/(SICI)1098-1004(1999)13:2<116::AID-HUMU3>3.0. CO;2-2 Groden J, Thliveris A, Samowitz W, Carlson M, Gelbert L, Albertsen H, Joslyn G, Stevens J, Spirio L, Robertson M, Sargeant L, Krapcho K, WolV E, Burt R, Hughes JP, Warrington J, McPherson J, Wasmuth J, Le Paslier D, Abderahim H, Cohen D, Leppert M, White R (1991) IdentiWcation and characterization of the familial adenomatous polyposis coli gene. Cell 66:589–600. doi:10.1016/ 0092-8674(81)90021-0 Heppner Goss K, Trzepacz C, Tuohy TM, Groden J (2002) Attenuated APC alleles produce functional protein from internal translation initiation. Proc Natl Acad Sci USA 99:8161–8166. doi:10.1073/ pnas.112072199 Kanter-Smoler G, Fritzell K, Rohlin A, Engwall Y, Hallberg B, Bergman A, Meuller J, Grönberg H, Karlsson P, Björk J, Nordling M (2008) Clinical characterization and the mutation spectrum in Swedish adenomatous polyposis families. BMC Med 6:10. doi:10.1186/1741-7015-6-10 Kinzler KW, Nilbert MC, Su LK, Vogelstein B, Bryan TM, Levy DB, Smith KJ, Preisinger AC, Hedge P, Mc Kechnie D, Finniear R, Markham A, GroVen J, Boguski MS, Altschul SF, Horii A, Ando H, Miyoshi Y, Miki Y, Nishisho I, Nakamura Y (1991) IdentiWcation of FAP locus gene from chromosome 5q21. Science 253:661–665. doi:10.1126/science.1651562 Knudsen AL, Bisgaard ML, Bülow S (2003) Attenuated familial adenomatous polyposis (AFAP). A review of the literature. Fam Cancer 2:43–55. doi:10.1023/A:1023286520725 Matsumoto T, Lida M, Kobori Y, Mizuno M, Nakamura S, Hizawa K, Yao T (2002) Genetic predisposition to clinical manifestations in familial adenomatous polyposis with special reference to

123

J Cancer Res Clin Oncol (2009) 135:1463–1470 duodenal lesions. Am J Gastroenterol 97:180–185. doi:10.1111/ j.1572-0241.2002.05434.x Miyoshi Y, Ando H, Nagase H, Nishisho I, Horii A, Miki Y, Mori T, Utsunomiya J, Baba S, Petersen G, Hamilton SR, Kinzler KW, Vogelstein B, Nakamura Y (1992) Germ-line mutations of the APC gene in 53 familial adenomatous polyposis patients. Proc Natl Acad Sci USA 89:4452–4456. doi:10.1073/pnas.89.10.4452 Nagase H, Nakamura Y (1993) Mutations of the APC (adenomatous polyposis coli) gene. Hum Mutat 2:425–434. doi:10.1002/humu.1380020602 Nieuwenhuis MH, Vasen HFA (2007) Correlations between mutation site in APC and phenotype of familial adenomatous polyposis (FAP): A review of the literature. Crit Rev Oncol Hematol 61:153–161. doi:10.1016/j.critrevonc.2006.07.004 Norheim Andersen S, Lovig T, Fausa O, Rognum TO (1999) Germline and somatic mutations in exon 15 of the APC gene and K-ras mutations in duodenal adenomas in patients with familial adenomatous polyposis. Scand J Gastroenterol 34:611–617. doi:10.1080/003655299750026083 Olschwang S, Hamelin R, Laurent-Puig P, Thuille B, De Rycke Y, Li YJ, Muzeau F, Girodet J, Salmon RJ, Thomas G (1997) Alternative genetic pathways in colorectal carcinogenesis. Proc Natl Acad Sci USA 94:12122–12127. doi:10.1073/pnas.94.22.12122 Vandrovcova J, Stekrova J, Kebrdlova V, Kohoutova M (2004) Molecular analysis of the APC and MYH genes in Czech families aVected by FAP or multiple adenomas: 13 novel mutations. Hum Mutat 23:397 MIB#695 Wallis YL, Morton DG, McKeown CM, Macdonald F (1999) Molecular analysis of the APC gene in 205 families: extended genotypephenotype correlations in FAP and evidence for the role of APC amino acid changes in colorectal cancer predisposition. J Med Genet 36:14–20 Wu G, Wu W, Hegde M, Fawkner M, Chong B, Love D, Su LK, Lynch P, Snow K, Richards CS (2001) Detection of sequence variations in the adenomatous polyposis coli (APC) gene using denaturing high-performance liquid chromatography. Genet Test 5:281–290. doi:10.1089/109065701753617408