Acta Paadiatr 81: 522-6: 1992
Genetic analysis of cystic fibrosis in Denmark. Implications for genetic counselling, carrier diagnosis and prenatal diagnosis M Schwartz, NJ Brandt, C Koch, S Lanng and PO Schiratz' Section of Clinical Genetics. Department of Pediatrics, Rigshospitalet, Copenhagen and Department of Pediatrics'. University Hospital of Arhus. Arhus, Denmark
Schwartz M, Brandt NJ, Koch C, Lanng S, SchiratzPO. Genetic analysis of cystic fibrosis in Denmark. Implications for genetic counselling, carrier diagnosis and prenatal diagnosis. Acta Pzdiatr 1992;81: 522-6. Stockholm. ISSN 0803-5253 Cystic fibrosis is the most common, severe, inherited disease in the Caucasian population. As a consequence, the demand for genetic counselling of patients with cystic fibrosis and their families is large. In Denmark the incidence of cystic fibrosis is 1 :4700, which is quite low compared to other European countries. We have investigated 268 Danish cystic fibrosis patients with respect to DNA markers (haplotypes) and the most common mutation AF508. The AF508 mutation is found on 88%of all cystic fibrosis chromosomes, the highest frequency reported so far. This has had an important impact on genetic counselling, prenatal diagnosis and eventually population screening. In the Danish population 78% of all couples at risk will be informative for AF508 and will be identifiable by simple screening methods. 0 Cysticfibrosis. AF508, genetic counselling, mutation, prenatal diagnosis, screening M Schwartz, Section of Clinical Genetics, Rigshospitalet 4062, Blegdamsvej 9, D K 2100 Copenhagen, Denmark
Cystic fibrosis (CF) is the most common severe recessive genetic disorder in the Caucasian population. Its occurrence varies amongst the European populations studied, being highest in France, Italy and Germany with an incidence of approximately 1 :2500 (I), and lowest in Finland with an incidence of 1 : 14000 (2). In Denmark the incidence has recently been estimated to be 1 :4700, giving a carrier frequency of 1 :34 (3). In 1985 linkage between DNA markers and the C F gene was established and the gene localized to the long arm of chromosome 7 (4, 5). In 1987 DNA probes localized very close to the gene were identified (6) and linkage disequilibrium was found between the C F mutation and restriction fragment length polymorphisms (RFLPs) at locus D7S23 (7, 8). The gene responsible for C F has recently been cloned and is predicted to code for a protein of approximately 1480 amino acids, called the cystic fibrosis transmembrane regulator protein (CFTR). The most common C F mutation (AF508) is a 3-bp deletion resulting in a protein missing the amino acid phenylalanine at position no. 508 (9-1 1). As predicted, this mutation is in strong linkage disequilibrium with the so-called B haplotype determined by the DNA markers XV2c and KM 19 (12). However, so far at least 150 other mutations in the CFTR gene have been identified, most of which are found on chromosomes with the B haplotype (1 3). The possibility of a direct detection of the mutation responsible for the disease creates new perspectives for genetic counselling, prenatal diagnosis and eventually population screening for carriers of CF.
We have investigated almost all known living Danish patients with CF, their parents, siblings and other relatives with respect to haplotypes ( X V ~ CKM , 19) and the frequency of AF508 and other mutations in the CFTR gene.
Materials and methods Patients
The study group consisted of 254 unrelated C F patients of Danish origin, their parents, healthy and affected siblings and more distant relatives. The patients studied account for more than 90% of the known living C F cases in Denmark. The majority of the C F patients included were attending the C F center at Rigshospitalet, Copenhagen and k h u s Kommune Hospital. The diagnostic criteria for C F were the presence of at least two of the following three signs: a positive sweat test; chronic obstructive pulmonary disease; or an exocrine pancreatic insufficiency. D N A analyses DNA was prepared from EDTA-stabilized blood samples by conventional methods. RFLP analysis was performed as reported previously (14). The DNA probes for the XV2c and KM19 polymorphisms have
A C T A PRDIATR
523
Genetic analysis of cysticjbrosis in Denmark
XI (1992)
Exon 10 AF508
A/N
bP
A/N
A/N
A/A
N/N
1
Fig. 1. Detection of AFS08 in a family. Polyacrylamide gel stained with ethidium bromide. Lane 1: Molecular weight marker; lanes 2, 3, 4: mother, father and a healthy sibling, respectively, all heterozygous for F508 and showing heteroduplex bands; lane 5: sample from the C F patient, homozygous for F508; lane 6: sample from a normal homozygous control.
Table 1. Frequency of the various haplotypes detected in the Danish C F population.
XV2c
KM 19
Haplotype
CF chromosomes
Normal chromosomes
I 2 1 2
A
18 (3.5)
B
451 (89.9) 22 (4.3) 17 (3.3)
66 (33.3) 22 ( I 1.1) 81 (40.9) 29 (14.6)
Table 2. Distribution of mutations on chromosomes with different haplotypes.
Haplotype AF508 Unknown G551 D G542X 621 + IG-rT N1303K A
I 1
2 2
C D
been described by Estivill et al. (6). For each polymorphism investigated, the allele 1 corresponds to the larger restriction fragment and the allele 2 to the smaller fragment.
PCR analysis Taq polymerase was purchased at Perkin-Elmer Cetus and the amplification was performed as recommended. The AF508 mutation was identified by electrophoresis in a 12% polyacrylamide gel as described by Rommens et al. (15) (Fig. 1). Other mutations were detected as described in previous publications: G55 I D by Cutting et al. (16); 621 + G + T by Zilienskyet al. (17); N1303K by Osborne et al. (18); G542X by Kerem et al. (19).
Results Frequency of AF508 The results of haplotype and AF508 analysis in 254 independent C F patients of Danish origin are shown in
B C D Total (%)
8 428 4 9
10 14 18 8
449 50 (88.3%) (9.7%)
0 2 0 0
0 3 0 0
0 2 0 0
2 0 0
2
3
2
2
0
Table 3. Carrier testing. Identification of carriers among close relatives to C F patients. In families not informative for AFS08 this was done by haplotyping. The individuals totalled 145 males and 175 females.
Observed
Expected
320 161 75 89
179 73 87
Family members tested Carriers identified Carrier males Non-carrier females
Table 4 . Prenatal diagnosis of cystic fibrosis.
Method
No.
Predicted affected
M. villar enz (34) Linked markers AF508/AF508 AF508/G542X
14 36
11
15 1
2 I
4
Confirmed 3CF I 1 CF -
I carrier
524
M Schwarrz el al.
Tables 1 and 2. These are an extension of previously reported data (20). The data clearly show the homogeneity of the Danish CF population. Eighty-eight percent of all C F chromosomes had the AF508 mutation. Hence, according to the Hardy-Weinberg equation, 77.4% of the patients should be homozygous for this mutation, 21.2% of the patients heterozygous and 1.4% should have another mutation on both chromosomes. These frequencies were confirmed by our data: among 254 C F patients we identified 198 homozygous for AF508 (77.9%), 53 heterozygous for AF508 (20.8%) and three without AF508 (1.2%). Haplotypes defined by the DNA probes XV2c and KM 19 (Tables 2 and 3) are in strong linkage disequilibrium with the AF508 mutation. The B haplotype is found on 95.1% of all chromosomes with the AF508 deletion. However, the B haplotype is still the most common C F haplotype even when chromosomes without AF508 are considered. Twenty-three of 59 C F chromosomes without AF508 have the haplotype B which is significantly different from the frequency of 1 1 Yo for the B haplotype on non-CF chromosomes ( p < 0.005).
ACTA PEDIATR 81 (1992)
AF508 and found three couples at risk. At their request two of these couples had prenatal diagnosis performed. The last couple is not yet pregnant. Prenatal diagnosis Prenatal diagnosis has been offered to CF parents since 1985, first by microvillar enzyme determination (22), then by more or less distant RFLP markers (23) and in the past year by a direct mutation test. We have performed 66 prenatal diagnoses for CF since 1985 and the numbers, methods and outcome are shown in Table 4. DNA from all aborted fetuses (predicted affected) has retrospectively been tested for confirmation of the diagnosis by AF508 test. In one case the microvillar enzyme test predicted an affected fetus which was later proven to be heterozygous healthy.
Discussion
The frequency of AF508 varies considerably among the populations studied. So far the Danish population has the highest frequency of AF508 (88%). In the pooled data from Europe the frequency of AF508 is 68% (12). The strong linkage disequilibrium between a specific Other mutations haplotype (B) determined by the DNA markers XV2c In the past year several other mutations in the CFTR and KM 19 in all populations studied led previously to gene have been found. More than 150different mutations the conclusion that most CF cases might be the result of have been identified (1 3). Most of these have been found a single or a few predominant ancestral mutation(s). only on a single C F chromosome or in special ethnic However, after the isolation of the gene responsible for groups. We have tested 59 CF chromosomes without CF (CFTR) it has been shown that many different AF508 for several other mutations. Among the mutations exist in this gene although one (AF508) is mutations reported we chose to test for those which have clearly more frequent. been found on more than one C F chromosome (13). We We have investigated almost the entire living Danish have identified the splice mutation 621 + I G + T in CF population and have shown that the Danish populaintron 4 (17) on two chromosomes, the missense tion is extremely homogeneous with a frequency of mutation G551D in exon 1 1 (16) on two chromosomes, AF508 of 88%. Furthermore, only few other known the stop mutation G542X in exon 1 1 (19) on three mutations have been identified on the Danish C F chromosomes and the missense mutation N1303K in chromosomes. The first nucleotide-binding fold of the exon 21 (18) on two chromosomes. All the mutations CFTR protein seems to have a cluster of CF mutations were found on chromosomes with the B haplotype (16). It is noteworthy that only two of these (G542X and (Table 2). G551D) have been found in our material. These two mutations, accounting for less than 1% of all our C F chromosomes, are by far the most common, and account for l l % , on average, in other populations Currier testing of CF relatives and their partners studied (13). In the Scottish population, which should be The high frequency of the AF508 mutation in the considered related to the Danish population, these two Danish population has made it possible to identify mutations account for 9.5% (24). The mutation 1717carriers with certainty in most families. We have tested IG+A was not found on any of the 59 non-AF508 CF 117 healthy siblings (a priori risk of 66%), and 203 chromosomes investigated, although it is relatively relatives with an a priori risk of being a carrier between common in Scotland (24). 12.5 and 50%. In contrast to other data (21), we did not Two C F chromosomes with the 621 + I G + T intron 4 find any surplus of carriers or any surplus of males splicing mutation were identified, whereas the two among the carriers or females among the non-carriers in relatively common mutations, R117H and D110H (25) the group tested (Table 3). We have tested 185 persons in exon 4 were not. without any known risk of being carriers. We have However, despite the few mutations found on the identified six individuals as being heterozygous for non-AF508 C F chromosomes in the Danish population,
Generic analysis of cysticfibrosis in Denmark
ACTA PRDIATR 81 (1992)
Table 5. Probability that a given haplotype without AF508 defines a C F chromosome.
A B C D No haplotyping
Denmark
UK(29)
1 in 1595
1 in 457
I in 125 I in 863 1 in 796 I in 566
I in 43.5 I in 375 1 in 155.5 1 in 182
Danish values are compared with values calculated for the British population (British: d=0.78 and q = 1/40).
our detection rate, including all known mutations, is close to 90% (90.3Y0)which definitely justifies analysis of partners to known CF carriers. The counselling of partners of CF relatives has previously been based on linkage disequilibrium data for the haplotypes determined by XV2C and KM 19. Quite good estimates can be calculated from these figures. A surplus of carriers among siblings or other close relatives has been proposed to explain the high incidence of CF (I). However, the identification of the most common mutation allows a direct carrier testing of partners to identified carriers, dependent of the detection rate of the disease mutations. The detection rate (d) of CF chromosomes in the Danish population is 0.88, and therefore 77.4% (0.88 x 0.88) of at-risk couples can be detected. The risk that a AF508 negative person is a carrier (the false negative rate) can be calculated from the expected frequency q of the CF mutation and the detection rate 2pq( 1 -d)/p* + 2pq( 1 - d) (26), and therefore varies greatly with the values of both d and q. T o illustrate this we have compared the Danish population with the British population (27). In the Danish population the gene frequency q is 1 :68 and d = 0.88, and the risk of a AF508 negative person being a carrier equals 1 :284, whereas in the UK the carrier risk is I :99, since q = 1 :40 and d =0.78. Because chromosomes with the haplotype B are more likely to carry a yet unknown CF mutation, it is possible to give a more accurate risk estimate by including haplotyping in the risk calculation. We have compared the calculated risk for a “Danish” chromosome (without AF508) and a specific haplotype to be a CF chromosome with the values published for the British population (Table 5). Without including haplotype the risk of a AF508 negative person being a carrier is lowered from 1 :34 to 1 :284 (8.3 times) compared to the British situation where the risk is lowered 4.5 times. By including haplotypes the risk of a person without the B haplotype is not larger than 1 in 400, compared to a person with the genotype BB where the risk is 1 in 63. The direct test has affected the group of couples to whom prenatal diagnosis can be offered. Previously it was a requirement that the index case be alive in order to establish haplotypes. Several couples were in the situation that their first child had died. The direct mutation
525
test will allow prenatal diagnosis in approximately 80% of all couples and will also allow a postmortem diagnosis in cases where the child died without a diagnosis. The identification of various mutations will eventually also help in the direct diagnosis of patients suspected of having CF. Eighty percent of all Danish CF patients will be homozygous for AF508 and a mutation test is much quicker, cheaper and more reliable than a sweat test. It can be performed easily on small blood samples (3 pl) or on dried blood from a Guthrie card. Furthermore, a prognostic evaluaton will also be possible following knowledge of the different mutations causing CF. In a previous report of 235 patients (28) we found that significantly more of the patients homozygous for AF508 had onset of symptoms before the age of six months, were diagnosed earlier, had poorer lung function, required greater pancreatic substitution and had a greater yearly mortality rate than patients heterozygous for AF508. Since only few other mutations have been identified it is too early to subdivide the patients that are heterozygous for AF508 into different groups. The present investigation serves to illustrate the importance of genetic testing of CF patients in the population. It is very important to obtain reliable information about the incidence of CF and the frequency of the most common mutations in order to be able to give accurate genetic counselling in extended families. This is even more important when population screening is considered. Acknow1edgemenfs.-The authors would like to thank all Danish cystic fibrosis families for their great help and interest, and Gitte Vedel Christensen, Merete Steensgaard-Hansen, Anne-Lise Dyrbye Palle, Birgitte Wermer Nielsen and Bodil Mogensen for excellent technical assistance. Novo’s Fond, P Carl Petersens Fond and Direkterr Jacob Madsens & Hustrus Fond are thanked for financial support.
References 1. Romeo G, Devoto M, Galietta JLV. Why is the C F gene so frequent? Hum Genet 1989;84:1-5 2. Kere J, Norio R, Savilahti E, Estivill X, Chapelle A de la. Cystic fibrosis in Finland: a molecular and genealogical study. Hum Genet 1990:83:20-5 3. Nielsen OH, Thomsen BL, Green A, Andersen PK, Hauge M, Schistz PO. Cystic fibrosis in Denmark 1945 to 1985. Acta Paediatr Scand 1988;77:83641 4. Wainwright BJ, Scambler PJ, Schmidtke J, et al. Localization of cystic fibrosis locus to human chromosome 7 cen-q22. Nature 1985;318~384-5 5. Knowlton RG, Cohen-Haguenauer 0, Nguyen VC, et al. A polymorphic DNA marker linked to cystic fibrosis is located on chromosome 7. Nature 1985;318:380-2 6. Estivill X, Farrall M, Williamson R, et al. A candidate for the cystic fibrosis locus isolated by selection for methylation free islands. Nature 1987;326:840-5 7. Estivill X, Farrall M, Williamson R, et al. Patterns of polymorphism and linkage disequilibrium for cystic fibrosis. Genomics 1987;1:257-63 8. Serre JL, Simon-Bouy B, Mornet E, et al. Studies of RFLPclosely
526
M Schwartz et al.
linked to the cystic fibrosis locus throughout Europe lead to new considerations in population genetics. Hum Genet 1990;8444954 9. Rommens JM, Iannuzzi MC, Kerem B, et al. Identification of the cystic fibrosis gene: chromosome walking and jumping. Science 1989;245:1059-65 10. Riordan JR, Rommens JM, Kerem B-S, et al. Identification ofthe cystic fibrosis gene: cloning and characterization of complementary DNA. Science 1989;245:1066-73 I 1. Kerem B, Rommens JM, Buchanan JA, et al. Identification of the cystic fibrosis gene: genetic analysis. Science 1989;245:1073-80 12. European Working Group on C F Genetics (EWGCFG). Gradient of distribution in Europe of the major C F mutation and of its associated haplotype. Hum Genet 1990;85:43641 13. Tsui L-C. Cystic fibrosis genetic analysis consortium 14. Schwartz M, Rosenberg T, Niebuhr E, et al. Choroideremia: further evidence for assignment of the locus to Xq13-q21. Hum Genet 1986;78:156-60 15. Rommens J, Kerem B-S,Greer W, Chang P, Tsui L-C, Ray i. Rapid nonradioactive detection of the major cystic fibrosis mutation. Am J Hum Genet 1990;46395-6 16. Cutting GR, Kasch LM, Rosenstein BJ, et al. A cluster of cystic fibrosis mutations in the first nucleotide-binding fold of the cystic fibrosis conductance regulator protein. Nature 1990;346:366-9 17. Zielenski J, Bozon D, Kerem B, Markiewicz D, Rommens JM, Tsui L-C. Identification of mutations in exon 1 through 8 of the cystic fibrosis transmembrane conductance regulator (CFTR) gene. Genomics 1991;10:229-35 18. Osborne L, Knight RA, Santis G, Hodson M. A mutation in the second nucleotide binding fold of the cystic fibrosis gene. Am J Hum Genet 1991;48:608-12 19. Kerem BS, Rommens JM. Buchanan JA, et al. Identification of
ACTA PRDIATR 81 (1992)
mutations in regions corresponding to the putative nucleotide (ATP)-binding folds in the cystic fibrosis gene. Proc Natl Acad Sci USA 1991;87:8447-51 20. Schwartz M, Johansen HK, Koch C, Brandt NJ. Frequency of the AF508 mutation on cystic fibrosis chromosomes in, Denmark. Hum Genet 1990;85:427-8 21. Kitzis A. Unusual segregation of cystic fibrosis alleles. Nature 1988;336:316 22. Brock DJH, Clarke HAK, Baron L. Prenatal diagnosis of cystic fibrosis by microvillar enzyme assay on a sequence of 258 pregnancies. Hum Genet 1988;78:271-5 23. Schwartz M, Super M, Schmidtke J, et al. Prenatal diagnosis of cystic fibrosis using linked DNA probes. Prenat Diagn 1988; 8:619-24 24. Shrimpton AE, McIntosh I, Brock D. The incidence of different cystic fibrosis mutations in the Scottish population: effect on prenatal diagnosis and genetic counselling. J Med Genet 1991;28:in press 25. Dean M, White MB, Amos J, et al. Multiple mutations in highly conserved residues are found in mildly affected cystic fibrosis patients. Cell 1990;61:863-70 26. Lemna WK, Feldman GL, Kerem B, et al. Mutation analysis for heterozygote determination and the prenatal analysis of cystic fibrosis. N Engl J Med 1990;322:291-6 27. Watson E, Williamson R, Brueton L, Winter R. Genetic counselling for cystic fibrosis based upon mutation/haplotype analysis. Lancet 1990;2:190-1 28. Krogh Johansen H, Nir M, Heiby N, Koch C, Schwartz M. Severity of cystic fibrosis in patients homozygous and heterozygous for the AF508 mutation. Lancet 1991;337:631-4 Received May 13, 1991. Accepted Sep. 12, 1991
Genetic analysis of cystic fibrosis in Denmark. Implications for genetic counselling, carrier diagnosis and prenatal diagnosis M Schwartz, NJ Brandt, C Koch, S Lanng and PO Schiratz' Section of Clinical Genetics. Department of Pediatrics, Rigshospitalet, Copenhagen and Department of Pediatrics'. University Hospital of Arhus. Arhus, Denmark
Schwartz M, Brandt NJ, Koch C, Lanng S, SchiratzPO. Genetic analysis of cystic fibrosis in Denmark. Implications for genetic counselling, carrier diagnosis and prenatal diagnosis. Acta Pzdiatr 1992;81: 522-6. Stockholm. ISSN 0803-5253 Cystic fibrosis is the most common, severe, inherited disease in the Caucasian population. As a consequence, the demand for genetic counselling of patients with cystic fibrosis and their families is large. In Denmark the incidence of cystic fibrosis is 1 :4700, which is quite low compared to other European countries. We have investigated 268 Danish cystic fibrosis patients with respect to DNA markers (haplotypes) and the most common mutation AF508. The AF508 mutation is found on 88%of all cystic fibrosis chromosomes, the highest frequency reported so far. This has had an important impact on genetic counselling, prenatal diagnosis and eventually population screening. In the Danish population 78% of all couples at risk will be informative for AF508 and will be identifiable by simple screening methods. 0 Cysticfibrosis. AF508, genetic counselling, mutation, prenatal diagnosis, screening M Schwartz, Section of Clinical Genetics, Rigshospitalet 4062, Blegdamsvej 9, D K 2100 Copenhagen, Denmark
Cystic fibrosis (CF) is the most common severe recessive genetic disorder in the Caucasian population. Its occurrence varies amongst the European populations studied, being highest in France, Italy and Germany with an incidence of approximately 1 :2500 (I), and lowest in Finland with an incidence of 1 : 14000 (2). In Denmark the incidence has recently been estimated to be 1 :4700, giving a carrier frequency of 1 :34 (3). In 1985 linkage between DNA markers and the C F gene was established and the gene localized to the long arm of chromosome 7 (4, 5). In 1987 DNA probes localized very close to the gene were identified (6) and linkage disequilibrium was found between the C F mutation and restriction fragment length polymorphisms (RFLPs) at locus D7S23 (7, 8). The gene responsible for C F has recently been cloned and is predicted to code for a protein of approximately 1480 amino acids, called the cystic fibrosis transmembrane regulator protein (CFTR). The most common C F mutation (AF508) is a 3-bp deletion resulting in a protein missing the amino acid phenylalanine at position no. 508 (9-1 1). As predicted, this mutation is in strong linkage disequilibrium with the so-called B haplotype determined by the DNA markers XV2c and KM 19 (12). However, so far at least 150 other mutations in the CFTR gene have been identified, most of which are found on chromosomes with the B haplotype (1 3). The possibility of a direct detection of the mutation responsible for the disease creates new perspectives for genetic counselling, prenatal diagnosis and eventually population screening for carriers of CF.
We have investigated almost all known living Danish patients with CF, their parents, siblings and other relatives with respect to haplotypes ( X V ~ CKM , 19) and the frequency of AF508 and other mutations in the CFTR gene.
Materials and methods Patients
The study group consisted of 254 unrelated C F patients of Danish origin, their parents, healthy and affected siblings and more distant relatives. The patients studied account for more than 90% of the known living C F cases in Denmark. The majority of the C F patients included were attending the C F center at Rigshospitalet, Copenhagen and k h u s Kommune Hospital. The diagnostic criteria for C F were the presence of at least two of the following three signs: a positive sweat test; chronic obstructive pulmonary disease; or an exocrine pancreatic insufficiency. D N A analyses DNA was prepared from EDTA-stabilized blood samples by conventional methods. RFLP analysis was performed as reported previously (14). The DNA probes for the XV2c and KM19 polymorphisms have
A C T A PRDIATR
523
Genetic analysis of cysticjbrosis in Denmark
XI (1992)
Exon 10 AF508
A/N
bP
A/N
A/N
A/A
N/N
1
Fig. 1. Detection of AFS08 in a family. Polyacrylamide gel stained with ethidium bromide. Lane 1: Molecular weight marker; lanes 2, 3, 4: mother, father and a healthy sibling, respectively, all heterozygous for F508 and showing heteroduplex bands; lane 5: sample from the C F patient, homozygous for F508; lane 6: sample from a normal homozygous control.
Table 1. Frequency of the various haplotypes detected in the Danish C F population.
XV2c
KM 19
Haplotype
CF chromosomes
Normal chromosomes
I 2 1 2
A
18 (3.5)
B
451 (89.9) 22 (4.3) 17 (3.3)
66 (33.3) 22 ( I 1.1) 81 (40.9) 29 (14.6)
Table 2. Distribution of mutations on chromosomes with different haplotypes.
Haplotype AF508 Unknown G551 D G542X 621 + IG-rT N1303K A
I 1
2 2
C D
been described by Estivill et al. (6). For each polymorphism investigated, the allele 1 corresponds to the larger restriction fragment and the allele 2 to the smaller fragment.
PCR analysis Taq polymerase was purchased at Perkin-Elmer Cetus and the amplification was performed as recommended. The AF508 mutation was identified by electrophoresis in a 12% polyacrylamide gel as described by Rommens et al. (15) (Fig. 1). Other mutations were detected as described in previous publications: G55 I D by Cutting et al. (16); 621 + G + T by Zilienskyet al. (17); N1303K by Osborne et al. (18); G542X by Kerem et al. (19).
Results Frequency of AF508 The results of haplotype and AF508 analysis in 254 independent C F patients of Danish origin are shown in
B C D Total (%)
8 428 4 9
10 14 18 8
449 50 (88.3%) (9.7%)
0 2 0 0
0 3 0 0
0 2 0 0
2 0 0
2
3
2
2
0
Table 3. Carrier testing. Identification of carriers among close relatives to C F patients. In families not informative for AFS08 this was done by haplotyping. The individuals totalled 145 males and 175 females.
Observed
Expected
320 161 75 89
179 73 87
Family members tested Carriers identified Carrier males Non-carrier females
Table 4 . Prenatal diagnosis of cystic fibrosis.
Method
No.
Predicted affected
M. villar enz (34) Linked markers AF508/AF508 AF508/G542X
14 36
11
15 1
2 I
4
Confirmed 3CF I 1 CF -
I carrier
524
M Schwarrz el al.
Tables 1 and 2. These are an extension of previously reported data (20). The data clearly show the homogeneity of the Danish CF population. Eighty-eight percent of all C F chromosomes had the AF508 mutation. Hence, according to the Hardy-Weinberg equation, 77.4% of the patients should be homozygous for this mutation, 21.2% of the patients heterozygous and 1.4% should have another mutation on both chromosomes. These frequencies were confirmed by our data: among 254 C F patients we identified 198 homozygous for AF508 (77.9%), 53 heterozygous for AF508 (20.8%) and three without AF508 (1.2%). Haplotypes defined by the DNA probes XV2c and KM 19 (Tables 2 and 3) are in strong linkage disequilibrium with the AF508 mutation. The B haplotype is found on 95.1% of all chromosomes with the AF508 deletion. However, the B haplotype is still the most common C F haplotype even when chromosomes without AF508 are considered. Twenty-three of 59 C F chromosomes without AF508 have the haplotype B which is significantly different from the frequency of 1 1 Yo for the B haplotype on non-CF chromosomes ( p < 0.005).
ACTA PEDIATR 81 (1992)
AF508 and found three couples at risk. At their request two of these couples had prenatal diagnosis performed. The last couple is not yet pregnant. Prenatal diagnosis Prenatal diagnosis has been offered to CF parents since 1985, first by microvillar enzyme determination (22), then by more or less distant RFLP markers (23) and in the past year by a direct mutation test. We have performed 66 prenatal diagnoses for CF since 1985 and the numbers, methods and outcome are shown in Table 4. DNA from all aborted fetuses (predicted affected) has retrospectively been tested for confirmation of the diagnosis by AF508 test. In one case the microvillar enzyme test predicted an affected fetus which was later proven to be heterozygous healthy.
Discussion
The frequency of AF508 varies considerably among the populations studied. So far the Danish population has the highest frequency of AF508 (88%). In the pooled data from Europe the frequency of AF508 is 68% (12). The strong linkage disequilibrium between a specific Other mutations haplotype (B) determined by the DNA markers XV2c In the past year several other mutations in the CFTR and KM 19 in all populations studied led previously to gene have been found. More than 150different mutations the conclusion that most CF cases might be the result of have been identified (1 3). Most of these have been found a single or a few predominant ancestral mutation(s). only on a single C F chromosome or in special ethnic However, after the isolation of the gene responsible for groups. We have tested 59 CF chromosomes without CF (CFTR) it has been shown that many different AF508 for several other mutations. Among the mutations exist in this gene although one (AF508) is mutations reported we chose to test for those which have clearly more frequent. been found on more than one C F chromosome (13). We We have investigated almost the entire living Danish have identified the splice mutation 621 + I G + T in CF population and have shown that the Danish populaintron 4 (17) on two chromosomes, the missense tion is extremely homogeneous with a frequency of mutation G551D in exon 1 1 (16) on two chromosomes, AF508 of 88%. Furthermore, only few other known the stop mutation G542X in exon 1 1 (19) on three mutations have been identified on the Danish C F chromosomes and the missense mutation N1303K in chromosomes. The first nucleotide-binding fold of the exon 21 (18) on two chromosomes. All the mutations CFTR protein seems to have a cluster of CF mutations were found on chromosomes with the B haplotype (16). It is noteworthy that only two of these (G542X and (Table 2). G551D) have been found in our material. These two mutations, accounting for less than 1% of all our C F chromosomes, are by far the most common, and account for l l % , on average, in other populations Currier testing of CF relatives and their partners studied (13). In the Scottish population, which should be The high frequency of the AF508 mutation in the considered related to the Danish population, these two Danish population has made it possible to identify mutations account for 9.5% (24). The mutation 1717carriers with certainty in most families. We have tested IG+A was not found on any of the 59 non-AF508 CF 117 healthy siblings (a priori risk of 66%), and 203 chromosomes investigated, although it is relatively relatives with an a priori risk of being a carrier between common in Scotland (24). 12.5 and 50%. In contrast to other data (21), we did not Two C F chromosomes with the 621 + I G + T intron 4 find any surplus of carriers or any surplus of males splicing mutation were identified, whereas the two among the carriers or females among the non-carriers in relatively common mutations, R117H and D110H (25) the group tested (Table 3). We have tested 185 persons in exon 4 were not. without any known risk of being carriers. We have However, despite the few mutations found on the identified six individuals as being heterozygous for non-AF508 C F chromosomes in the Danish population,
Generic analysis of cysticfibrosis in Denmark
ACTA PRDIATR 81 (1992)
Table 5. Probability that a given haplotype without AF508 defines a C F chromosome.
A B C D No haplotyping
Denmark
UK(29)
1 in 1595
1 in 457
I in 125 I in 863 1 in 796 I in 566
I in 43.5 I in 375 1 in 155.5 1 in 182
Danish values are compared with values calculated for the British population (British: d=0.78 and q = 1/40).
our detection rate, including all known mutations, is close to 90% (90.3Y0)which definitely justifies analysis of partners to known CF carriers. The counselling of partners of CF relatives has previously been based on linkage disequilibrium data for the haplotypes determined by XV2C and KM 19. Quite good estimates can be calculated from these figures. A surplus of carriers among siblings or other close relatives has been proposed to explain the high incidence of CF (I). However, the identification of the most common mutation allows a direct carrier testing of partners to identified carriers, dependent of the detection rate of the disease mutations. The detection rate (d) of CF chromosomes in the Danish population is 0.88, and therefore 77.4% (0.88 x 0.88) of at-risk couples can be detected. The risk that a AF508 negative person is a carrier (the false negative rate) can be calculated from the expected frequency q of the CF mutation and the detection rate 2pq( 1 -d)/p* + 2pq( 1 - d) (26), and therefore varies greatly with the values of both d and q. T o illustrate this we have compared the Danish population with the British population (27). In the Danish population the gene frequency q is 1 :68 and d = 0.88, and the risk of a AF508 negative person being a carrier equals 1 :284, whereas in the UK the carrier risk is I :99, since q = 1 :40 and d =0.78. Because chromosomes with the haplotype B are more likely to carry a yet unknown CF mutation, it is possible to give a more accurate risk estimate by including haplotyping in the risk calculation. We have compared the calculated risk for a “Danish” chromosome (without AF508) and a specific haplotype to be a CF chromosome with the values published for the British population (Table 5). Without including haplotype the risk of a AF508 negative person being a carrier is lowered from 1 :34 to 1 :284 (8.3 times) compared to the British situation where the risk is lowered 4.5 times. By including haplotypes the risk of a person without the B haplotype is not larger than 1 in 400, compared to a person with the genotype BB where the risk is 1 in 63. The direct test has affected the group of couples to whom prenatal diagnosis can be offered. Previously it was a requirement that the index case be alive in order to establish haplotypes. Several couples were in the situation that their first child had died. The direct mutation
525
test will allow prenatal diagnosis in approximately 80% of all couples and will also allow a postmortem diagnosis in cases where the child died without a diagnosis. The identification of various mutations will eventually also help in the direct diagnosis of patients suspected of having CF. Eighty percent of all Danish CF patients will be homozygous for AF508 and a mutation test is much quicker, cheaper and more reliable than a sweat test. It can be performed easily on small blood samples (3 pl) or on dried blood from a Guthrie card. Furthermore, a prognostic evaluaton will also be possible following knowledge of the different mutations causing CF. In a previous report of 235 patients (28) we found that significantly more of the patients homozygous for AF508 had onset of symptoms before the age of six months, were diagnosed earlier, had poorer lung function, required greater pancreatic substitution and had a greater yearly mortality rate than patients heterozygous for AF508. Since only few other mutations have been identified it is too early to subdivide the patients that are heterozygous for AF508 into different groups. The present investigation serves to illustrate the importance of genetic testing of CF patients in the population. It is very important to obtain reliable information about the incidence of CF and the frequency of the most common mutations in order to be able to give accurate genetic counselling in extended families. This is even more important when population screening is considered. Acknow1edgemenfs.-The authors would like to thank all Danish cystic fibrosis families for their great help and interest, and Gitte Vedel Christensen, Merete Steensgaard-Hansen, Anne-Lise Dyrbye Palle, Birgitte Wermer Nielsen and Bodil Mogensen for excellent technical assistance. Novo’s Fond, P Carl Petersens Fond and Direkterr Jacob Madsens & Hustrus Fond are thanked for financial support.
References 1. Romeo G, Devoto M, Galietta JLV. Why is the C F gene so frequent? Hum Genet 1989;84:1-5 2. Kere J, Norio R, Savilahti E, Estivill X, Chapelle A de la. Cystic fibrosis in Finland: a molecular and genealogical study. Hum Genet 1990:83:20-5 3. Nielsen OH, Thomsen BL, Green A, Andersen PK, Hauge M, Schistz PO. Cystic fibrosis in Denmark 1945 to 1985. Acta Paediatr Scand 1988;77:83641 4. Wainwright BJ, Scambler PJ, Schmidtke J, et al. Localization of cystic fibrosis locus to human chromosome 7 cen-q22. Nature 1985;318~384-5 5. Knowlton RG, Cohen-Haguenauer 0, Nguyen VC, et al. A polymorphic DNA marker linked to cystic fibrosis is located on chromosome 7. Nature 1985;318:380-2 6. Estivill X, Farrall M, Williamson R, et al. A candidate for the cystic fibrosis locus isolated by selection for methylation free islands. Nature 1987;326:840-5 7. Estivill X, Farrall M, Williamson R, et al. Patterns of polymorphism and linkage disequilibrium for cystic fibrosis. Genomics 1987;1:257-63 8. Serre JL, Simon-Bouy B, Mornet E, et al. Studies of RFLPclosely
526
M Schwartz et al.
linked to the cystic fibrosis locus throughout Europe lead to new considerations in population genetics. Hum Genet 1990;8444954 9. Rommens JM, Iannuzzi MC, Kerem B, et al. Identification of the cystic fibrosis gene: chromosome walking and jumping. Science 1989;245:1059-65 10. Riordan JR, Rommens JM, Kerem B-S, et al. Identification ofthe cystic fibrosis gene: cloning and characterization of complementary DNA. Science 1989;245:1066-73 I 1. Kerem B, Rommens JM, Buchanan JA, et al. Identification of the cystic fibrosis gene: genetic analysis. Science 1989;245:1073-80 12. European Working Group on C F Genetics (EWGCFG). Gradient of distribution in Europe of the major C F mutation and of its associated haplotype. Hum Genet 1990;85:43641 13. Tsui L-C. Cystic fibrosis genetic analysis consortium 14. Schwartz M, Rosenberg T, Niebuhr E, et al. Choroideremia: further evidence for assignment of the locus to Xq13-q21. Hum Genet 1986;78:156-60 15. Rommens J, Kerem B-S,Greer W, Chang P, Tsui L-C, Ray i. Rapid nonradioactive detection of the major cystic fibrosis mutation. Am J Hum Genet 1990;46395-6 16. Cutting GR, Kasch LM, Rosenstein BJ, et al. A cluster of cystic fibrosis mutations in the first nucleotide-binding fold of the cystic fibrosis conductance regulator protein. Nature 1990;346:366-9 17. Zielenski J, Bozon D, Kerem B, Markiewicz D, Rommens JM, Tsui L-C. Identification of mutations in exon 1 through 8 of the cystic fibrosis transmembrane conductance regulator (CFTR) gene. Genomics 1991;10:229-35 18. Osborne L, Knight RA, Santis G, Hodson M. A mutation in the second nucleotide binding fold of the cystic fibrosis gene. Am J Hum Genet 1991;48:608-12 19. Kerem BS, Rommens JM. Buchanan JA, et al. Identification of
ACTA PRDIATR 81 (1992)
mutations in regions corresponding to the putative nucleotide (ATP)-binding folds in the cystic fibrosis gene. Proc Natl Acad Sci USA 1991;87:8447-51 20. Schwartz M, Johansen HK, Koch C, Brandt NJ. Frequency of the AF508 mutation on cystic fibrosis chromosomes in, Denmark. Hum Genet 1990;85:427-8 21. Kitzis A. Unusual segregation of cystic fibrosis alleles. Nature 1988;336:316 22. Brock DJH, Clarke HAK, Baron L. Prenatal diagnosis of cystic fibrosis by microvillar enzyme assay on a sequence of 258 pregnancies. Hum Genet 1988;78:271-5 23. Schwartz M, Super M, Schmidtke J, et al. Prenatal diagnosis of cystic fibrosis using linked DNA probes. Prenat Diagn 1988; 8:619-24 24. Shrimpton AE, McIntosh I, Brock D. The incidence of different cystic fibrosis mutations in the Scottish population: effect on prenatal diagnosis and genetic counselling. J Med Genet 1991;28:in press 25. Dean M, White MB, Amos J, et al. Multiple mutations in highly conserved residues are found in mildly affected cystic fibrosis patients. Cell 1990;61:863-70 26. Lemna WK, Feldman GL, Kerem B, et al. Mutation analysis for heterozygote determination and the prenatal analysis of cystic fibrosis. N Engl J Med 1990;322:291-6 27. Watson E, Williamson R, Brueton L, Winter R. Genetic counselling for cystic fibrosis based upon mutation/haplotype analysis. Lancet 1990;2:190-1 28. Krogh Johansen H, Nir M, Heiby N, Koch C, Schwartz M. Severity of cystic fibrosis in patients homozygous and heterozygous for the AF508 mutation. Lancet 1991;337:631-4 Received May 13, 1991. Accepted Sep. 12, 1991
