The New England Journal of Medicine
e-mail icon  FREE NEJM E-TOC    HOME   |   SUBSCRIBE   |   CURRENT ISSUE   |   PAST ISSUES   |   COLLECTIONS   |    Advanced Search
Sign in | Get NEJM's E-Mail Table of Contents — Free | Subscribe
 
Original Article
PreviousPrevious
Volume 336:905-911 March 27, 1997 Number 13
NextNext

Identification of a Genetic Locus for Familial Atrial Fibrillation
Ramon Brugada, M.D., Terry Tapscott, B.S., Grazyna Z. Czernuszewicz, M.S., A.J. Marian, M.D., Anna Iglesias, B.S., Lluis Mont, M.D., Josep Brugada, M.D., Josep Girona, M.D., Anna Domingo, M.D., Linda L. Bachinski, Ph.D., and Robert Roberts, M.D.

 

This Article
-Abstract
- PDF

Commentary
-Letters

Tools and Services
-Add to Personal Archive
-Add to Citation Manager
-Notify a Friend
-E-mail When Cited

More Information
-PubMed Citation
ABSTRACT

Background Atrial fibrillation, the most common sustained cardiac-rhythm disturbance, affects over 2 million Americans and accounts for one third of all strokes in patients over 65 years of age. The molecular basis for atrial fibrillation is unknown, and palliative therapy is used to control the ventricular rate and prevent systemic emboli. We identified a family of 26 members of whom 10 had atrial fibrillation that segregated as an autosomal dominant disease. We subsequently identified two additional families in which the disease was linked to the same locus.

Methods We screened the human genome with 300 polymorphic dinucleotide-repeat markers using an unconventional strategy of pooling the DNA samples into two groups (affected and unaffected), which reduced the sample size by approximately 90 percent, before performing linkage analysis to map the locus. This made it possible to identify potential loci within a few weeks.

Results The lod scores for markers D10S569 and D10S607, located at 10q22–q24, were 3.60 in Family 1. The disease locus in Families 2 and 3 was also linked to the same markers, with lod scores of 6.02 and 5.35 for markers D10S569 and D10S607, respectively, when data on all three families were combined. Haplotype analysis of the three families showed that the locus was between D10S1694 and D10S1786, an interval of 11.3 centimorgans.

Conclusions Identification of the gene for familial atrial fibrillation will help to elucidate the molecular basis of the disease and provide insights into acquired forms. The strategy of pooling DNA samples for analysis is more time and cost effective than conventional screening and should accelerate the process of gene mapping in the future.


Atrial fibrillation, the most common sustained cardiac-rhythm disturbance,1 affects more than 2 million Americans,2 with an overall prevalence of 0.89 percent. The prevalence increases rapidly with age to 2.3 percent between the ages of 40 and 60 years and to 5.9 percent over the age of 65.2 Although the initial course of atrial fibrillation is often paroxysmal, it almost invariably progresses to a chronic sustained rhythm disturbance with manifestations ranging from palpitations to cardiac failure. The most dreaded complication is stroke; atrial fibrillation accounts for one third of all strokes in patients over the age of 65.3 Since there is currently no effective means of preventing or eliminating atrial fibrillation, therapy consists of controlling the ventricular rate and preventing systemic emboli with antiplatelet and anticoagulant therapy.4 The emotional and medical burdens imposed by long-term medical therapy and, in many cases, subsequent stroke are immense, with a total annual cost of $9 billion.

The molecular basis of atrial fibrillation has yet to be determined. One approach, which has been successful for metabolic disorders, is to identify the gene responsible for a familial form of the disease. Familial atrial fibrillation is probably very uncommon,5 but we located a small family in Spain in which atrial fibrillation segregates as an autosomal dominant trait. Only 10 living family members were affected; thus, to map the chromosomal locus by conventional genetic-linkage analysis would probably require analyzing DNA markers every 5 to 10 centimorgans (cM) throughout the human genome. The arduous task of analyzing DNA from each of the 26 family members for 300 to 600 markers would also be very costly. Accordingly, we adopted an unconventional strategy of pooling the DNA of the affected family members and comparing the results of the DNA analysis with those of an analysis of pooled DNA from unaffected family members at each marker locus to detect differences that would suggest segregation of a particular allele with the disease. This reduced the number of samples to be analyzed by more than 90 percent and enabled us to identify four potential loci within a few weeks.

Methods

Study Families and Diagnosis

We identified a family (Family 1) in which atrial fibrillation appears to be segregating as an autosomal dominant disease with high penetrance. The family consists of 26 living members spanning three generations, of whom 10 have atrial fibrillation. Two additional family members are known to have died of complications of the disease. We subsequently identified two other small families with a total of 17 living members, 9 of whom are affected. In these two families (Families 2 and 3) the defect mapped to the same locus as in the original family. The three pedigrees are shown in Figure 1.


View larger version (6K):
[in this window]
[in a new window]
 
Figure 1. Pedigrees of Three Families with Familial Atrial Fibrillation.

Circles denote female family members, squares male family members, solid symbols affected family members, symbols with a slash deceased family members, and the symbol with a question mark a family member whose disease status has not been determined.

 
Clinical Evaluation

After providing informed consent according to the guidelines of Baylor College of Medicine and Methodist Hospital, the subjects were evaluated by a detailed history taking, physical examination, 12-lead electrocardiography, and two-dimensional echocardiography. Criteria for diagnosis were based on the electrocardiographic findings. Other likely causes of atrial fibrillation in this population — such as hypertension, valvular disease, and thyroid disease — were ruled out at the time of diagnosis.

Preparation of DNA and Pooled Samples

Blood was collected from each member of the three families, DNA was extracted by the salting-out procedure,6 and lymphocytes were isolated for the development of transformed cell lines.7 For the pooled-sample analysis we used blood only from Family 1. Equimolar amounts of DNA from 10 affected members (Subjects II-2, II-3, III-1, III-3, III-5, III-11, IV-6, IV-7, IV-8, and IV-9) were combined into a single sample, as was DNA from 10 unaffected family members (Subjects II-1, II-4, II-6, III-4, III-6, III-8, III-9, III-10, III-12, and IV-5). Whenever possible, unaffected siblings and parents of affected subjects were chosen to provide samples for the unaffected pool in order to minimize the allelic differences between the two groups.

Marker Analysis

A total of about 300 polymorphic dinucleotide — (CA)n — repeat markers located approximately 15 to 20 cM apart were selected on the basis of their polymorphic information content primarily from the genetic maps of Genethon8 or the National Institutes of Health–Centre d'Etude du Polymorphisme Humain.9 For each microsatellite marker, primers annealing to the sense strand were end-labeled with [32P]{gamma}-deoxyadenosine triphosphate with polynucleotide kinase (Pharmacia), and the DNA was amplified by the polymerase chain reaction under standard conditions for 30 cycles consisting of denaturation at 94°C for 45 seconds, annealing at 55°C for 30 seconds, and extension at 72°C for 30 seconds. Some markers required a higher annealing temperature. Initial denaturation was carried out at 95°C for five minutes. Each 50-µl reaction contained 150 pmol of specific primers, 0.3 U of Taq polymerase (Pharmacia), 200 µM 4-deoxynucleoside triphosphate, and 200 ng of genomic DNA for the analysis of either individual or pooled DNA samples. The amplified DNA products were analyzed by electrophoresis through 6 percent denaturing polyacrylamide–urea sequencing gels as previously described.10

After the 300 chromosomal markers were amplified from the two pooled DNA samples, the products were loaded side by side on a polyacrylamide gel and then subjected to electrophoresis. The electrophoretic pattern exhibited on the autoradiograph by each marker allele amplified from the pooled sample from the affected group was visually inspected, and the results were compared with those of the pooled sample from the unaffected group. The presence of a unique band or a band of greater intensity in the sample from the affected group suggested cosegregation of the marker allele with the disease allele.

Linkage Analysis

Two-point linkage analysis was carried out on a personal computer with version 5.2 of the Linkage program.11 Multipoint linkage analysis was conducted on a Vax computer with Fastlink. An autosomal dominant pattern of inheritance was assumed, and penetrance was set at 99 percent, on the basis of the observed high frequency of affected persons in sibships at risk in Family 1. The frequencies of the disease allele and the normal allele were assumed to be 0.0001 and 0.9999, respectively, and the allele frequencies of microsatellite markers were arbitrarily assigned a value of 1/n, where n refers to the number of alleles observed. Since the use of incorrect values for linkage parameters can cause false positive results, positive lod scores were tested for robustness with respect to variations in penetrance (from 60 to 99 percent), variations in the prevalence of phenocopies (from 0 to 5 percent), and marker-allele frequency with the use of published frequencies, when available, or the frequencies observed in this family.

Results

Clinical Characteristics

In Family 1 there were 10 living affected subjects (6 male and 4 female) whose age at diagnosis ranged from 2 to 35 years (average, 17.8). Nine family members had chronic atrial fibrillation, and one (Subject IV-6) had paroxysmal atrial fibrillation. Three had dyspnea on exertion (Subjects IV-6, II-2, and III-5), and seven were asymptomatic. Subject III-1 had pericarditis and atrial fibrillation at the time of diagnosis, but since the atrial fibrillation has persisted for 10 years, pericarditis was not considered the cause. Subject II-8 died of a cerebrovascular accident at the age of 68. Subject III-2, who was given a diagnosis of paroxysmal atrial fibrillation at the age of 20, died suddenly at the age of 36; no autopsy was performed. Electrocardioversion was attempted unsuccessfully in three of those with chronic atrial fibrillation. Subjects II-2 and III-5 had increased left ventricular internal diameters and ventricular ejection fractions of 51 and 54 percent, respectively. The remainder had no abnormalities on echocardiography, and the average ejection fraction was 69 percent.

In Families 2 and 3, all nine affected subjects had chronic atrial fibrillation, were asymptomatic, and had no echocardiographic abnormalities. The age at diagnosis in these two families ranged from 2 to 46 years.

DNA Analysis and Genetic Linkage

The 300 chromosomal markers used to screen the human genome were amplified from the pooled DNA samples from affected and unaffected subjects; the pooling procedure reduced the number of samples to be analyzed from 26 to 2. The electrophoretic pattern of the sample from the affected group differed from that of the unaffected group for 50 of the 300 marker alleles. These differences, which suggested potential loci for the gene responsible for atrial fibrillation, were identified within a few weeks. A typical example of a difference in a marker allele between the pooled samples is shown in Figure 2. In lanes 1 and 2 and lanes 3 and 4 of Figure 2, the electrophoretic pattern is similar, indicating that the two marker alleles are segregating at random with respect to the disease allele. In contrast, the intensity of the uppermost allele band in lane 6, showing the sample from the affected group, is greater than the intensity of the corresponding allele in lane 5, showing the sample from the unaffected group. A marker allele with greater intensity in the affected group, presumably due to an increased number of copies of the allele, suggests cosegregation with the disease.


View larger version (16K):
[in this window]
[in a new window]
 
Figure 2. Results of Polyacrylamide-Gel Electrophoresis of Four DNA-Marker Loci Amplified from Pooled DNA.

In lanes 1 and 2 and lanes 3 and 4 the electrophoretic patterns are identical for the markers amplified from the pooled DNA samples from the unaffected and affected groups. The samples shown in lanes 5 and 6 and lanes 7 and 8 show typical differences in band intensity observed between the two groups. The intensity of the lowermost band in lane 8 is much greater than that of the corresponding band in lane 7, reflecting an increase in the number of copies of that allele in the affected group that is consistent with the occurrence of cosegregation of the marker allele and the disease allele.

 
To confirm or rule out cosegregation, genetic-linkage analysis was necessary. This required genotyping of DNA from each of the 26 members of Family 1 for each of the 50 marker alleles. The resulting lod scores were at least -2 for all but four loci, for which low, positive scores were obtained that were consistent with the occurrence of cosegregation of the marker allele and the disease allele. Additional markers located in these four regions were amplified and analyzed, and the results ruled out three of these regions (4q, 7q, and 13q), but marker alleles (D10S569 and D10S607) in the region 10q22 cosegregated with the disease allele, with maximal two-point lod scores of 3.60, indicating genetic linkage. Lod scores for these markers remained above 3.0 despite variations in penetrance from 60 to 99 percent.

Since the estimated prevalence of atrial fibrillation in the general population below the age of 40 is less than 0.04 percent2 and the nonfamilial causes of atrial fibrillation at this young age are almost exclusively related to valvular heart disease or congenital defects, a phenocopy prevalence of even 1 percent is probably excessive. Nevertheless, lod scores for these markers remained significant even with a phenocopy prevalence of 5 percent. Similarly, lod scores were not sensitive to variations in the estimated frequencies of marker alleles.

We also typed a number of markers in Families 2 and 3 and determined that they had the same haplotype on the affected chromosome as Family 1 for all markers from D10S188 to D10S219, suggesting that the three families are distantly related. Markers outside this interval (D10S581, D10S1694, D10S1786, and D10S1686) were associated with different alleles in the other two families. A total of 27 markers were examined in an effort to define the precise boundaries of the region, but not all these markers were informative. Lod scores for the four most informative markers are presented in Table 1. Multipoint analyses resulted in a peak lod score of 3.60 between markers D10S569 and D10S607 for Family 1 alone and 6.17 for all three families combined and contributed no new information with respect to the location of the gene for atrial fibrillation.

View this table:
[in this window]
[in a new window]
 
Table 1. Two-Point Lod Scores for Four Informative Markers Linked to 10q22–q24.

 
Figure 3, Figure 4, and Figure 5 show the haplotypes for 11 informative markers in all three families and the positions of the recombination events that helped determine the location of the gene. In Family 1 crossovers were observed between the disease allele and D10S1786 in Subject III-1 and between the disease allele and D10S581 in Subject IV-6. Thus, on the basis of haplotype analysis, the gene causing atrial fibrillation in Family 1 lies between D10S581 and D10S1786, an interval of approximately 20.2 cM, on chromosome 10q22–q24. Family 3 had no crossovers in this region. In Family 2, either Subject II-2 or Subject II-3 had a crossover between D10S1694 and the disease allele, further limiting the region to one measuring 11.3 cM. Affected members of all three families had identical alleles for all markers tested from D10S188 through D10S219. Because the haplotype was conserved throughout such a large region, we conclude that these three families must be distantly related and have the same mutation at this locus, supporting the conclusion that the disease gene lies proximal to D10S1786 and distal to D10S1694, an interval of 11.3 cM.


View larger version (42K):
[in this window]
[in a new window]
 
Figure 3. Haplotypes and Recombination in Family 1.

Alleles are shown as base-pair sizes in centromere-to-telomere order for markers D10S581, D10S537, D10S1694, D10S188, D10S556, D10S569, D10S607, D10S1677, D10S219, D10S1786, and D10S1686. Alleles in parentheses represent inferred genotypes. For genotypes represented by question marks, no data were available. The most highly conserved haplotypes are assumed, and the chromosomal region cosegregating with the disease locus in each family is represented by the black bar. No attempt has been made to show crossover on the nondiseased chromosome. Circles denote female family members, squares male family members, solid symbols affected family members, and symbols with a slash deceased family members.

 

View larger version (32K):
[in this window]
[in a new window]
 
Figure 4: Haplotypes and Recombination in Family 2.

See the legend to Figure 3 for an explanation of the symbols. The symbol with a question mark denotes a family member whose disease status has not been determined.

 

View larger version (29K):
[in this window]
[in a new window]
 
Figure 5: Haplotypes and Recombination in Family 3.

See the legend to Figure 3 for an explanation of the symbols.

 
Discussion

We identified a small family in whom atrial fibrillation segregated as an autosomal dominant trait; this family consisted of 26 members, 10 of whom were affected and alive at the time of the study. Genetic-linkage analysis indicated that the gene responsible for atrial fibrillation in this family is located on chromosome 10q in the region of 10q22–q24, with two markers having maximal lod scores of 3.60. We identified regions containing potential loci for this gene within a few weeks by comparing the electrophoretic patterns exhibited by markers amplified from pooled DNA samples from the affected group with those observed in the pooled DNA samples from the unaffected parents and siblings, rather than by genotyping each marker in all 26 family members. This procedure reduced the number of samples in the genomic screen by 90 percent. Individual DNA analyses were performed for only 50 markers (a reduction of approximately 75 percent), resulting in the identification of the locus within a few weeks as opposed to several months to a year. The strategy gives rapid results and should be particularly useful in small families with a high penetrance of the disease gene. Subsequently, we identified two other families with atrial fibrillation that was also linked to 10q22–q24 (combined lod score for all three families, 6.02). The odds of genetic linkage of the disease in these families to the region of 10q22–q24 remained significant despite the wide range of values used for the variables in the linkage analyses — namely, a phenocopy prevalence of up to 5 percent and penetrance ranging from 60 to 99 percent. This analysis provides the first essential step in the identification of the responsible molecular defect.

The pooled-sample approach is based on the fact that each DNA marker has a unique and characteristic distribution of allele frequencies in the general population. When the DNA is pooled into two groups (affected and unaffected), each sample contains many alleles (rather than two, as is the case for a single subject), and thus, the intensity of the electrophoretic bands is dependent on the frequencies with which those alleles are represented in the pool. If a marker allele is segregating at random with respect to a disease, its distribution in the affected and unaffected subjects will be identical. If, on the other hand, a particular marker allele is cosegregating with the disease allele, then all the members of the affected pool will share this allele and DNA analysis will show a band of greater intensity than in the sample from the unaffected subjects or a unique band. The strategy of using pooled DNA samples to look for differences in allele distribution has been used to identify the locus of a recessive disease in a highly inbred population.12 Its lack of use in autosomal dominant diseases is probably due to the expectation that the inordinately high incidence of differences between affected and unaffected samples would represent false positive results. We minimized false positive results in our study by choosing as unaffected subjects in the pooled sample only parents and siblings, who were more likely to share alleles with the affected subjects.

Candidate genes that must be considered in the search for a genetic cause of atrial fibrillation include genes encoding channel or pore proteins and genes encoding the sympathetic or parasympathetic system that will influence cardiac conduction or automaticity, such as the {beta}-adrenergic receptor or signaling proteins. The genes for the {beta}-adrenergic receptor (ADRB1)13 and {alpha}-adrenergic receptor (ADRA2)14 are located on 10q23–q26, as is the gene for G-protein–coupled receptor kinase (GPRK5),15 which interacts with adrenergic receptors.

Mapping of the locus for atrial fibrillation provides the information necessary to identify the gene. Although atrial fibrillation is commonly associated with acquired heart disease, a substantial proportion of patients, particularly those with an early onset of disease, have no other forms of heart disease. It remains to be determined what percentage of the cases of atrial fibrillation in these patients is due to genetic defects. Thus, identification of the gene and elucidation of the molecular pathogenesis of atrial fibrillation should not only clarify the mechanism responsible for the familial form, but also shed light on the common acquired forms and eventually lead to more effective therapy. Identification of the gene and the responsible mutation is required to establish causality. The incidence of familial atrial fibrillation and the percentage of cases attributable to the 10q22–q24 locus remain to be determined. Nevertheless, families with familial atrial fibrillation can now be screened rapidly for genetic linkage to the 10q22–q24 region. For families in which the disease does not map to 10q22–q24, ruling out this locus will turn the search to other loci.

Supported in part by grants from the National Heart, Lung, and Blood Institute Specialized Centers of Research (P50-HL313-01); the National Heart, Lung, and Blood Institute Training Center in Molecular Cardiology (T32-HL-07706); and the American Heart Association Texas Affiliate (95R-1191).

We are indebted to Debora Weaver, Anna Zamora, and Che Rivers for their assistance in the preparation of the manuscript.


Source Information

From the Department of Cardiology, Baylor College of Medicine, Houston (R.B., T.T., G.Z.C., A.J.M., A.I., L.L.B., R.R.); the Cardiac Arrhythmia Service, Department of Cardiology, Hospital Clinic, University of Barcelona, Barcelona, Spain (L.M., J.B.); and the Department of Cardiology, Hospital Materno–Infantil Vall d'Hebron, Barcelona, Spain (J.G., A.D.).

Address reprint requests to Dr. Roberts at Baylor College of Medicine, 6550 Fannin, MS SM677, Houston, TX 77030.

References

  1. The National Heart, Lung, and Blood Institute Working Group on Atrial Fibrillation. Atrial fibrillation: current understandings and research imperatives. J Am Coll Cardiol 1993;22:1830-1834. [Abstract]
  2. Feinberg WM, Blackshear JL, Laupacis A, Kronmal R, Hart RG. Prevalence, age distribution, and gender of patients with atrial fibrillation: analysis and implications. Arch Intern Med 1995;155:469-473. [Abstract]
  3. Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation: a major contributor to stroke in the elderly: the Framingham Study. Arch Intern Med 1987;147:1561-1564. [Abstract]
  4. Albers GW. Atrial fibrillation and stroke: three new studies, three remaining questions. Arch Intern Med 1994;154:1443-1448. [Abstract]
  5. Beyer F, Paul T, Luhmer I, Bertram H, Kallfelz HC. Familiäres idiopathisches Vorhofflimmern mit Bradyarrhythmie. Z Kardiol 1993;82:674-677. [Medline]
  6. Miller SA, Dykes DD, Polesky HF. A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res 1988;16:1215-1215. [Free Full Text]
  7. Neitzel H. A routine method for the establishment of permanent growing lymphoblastoid cell lines. Hum Genet 1986;73:320-326. [CrossRef][Medline]
  8. Weissenbach J, Gyapay G, Dib C, et al. A second-generation linkage map of the human genome. Nature 1992;359:794-801. [CrossRef][Medline]
  9. NIH/CEPH Collaborative Mapping Group. A comprehensive genetic linkage map of the human genome. Science 1992;258:67-86. [Free Full Text]
  10. Durand JB, Bachinski LL, Bieling LC, et al. Localization of a gene responsible for familial dilated cardiomyopathy to chromosome 1q32. Circulation 1995;92:3387-3389. [Free Full Text]
  11. Ott J. Analysis of human genetic linkage. Rev. ed. Baltimore: Johns Hopkins University Press, 1991.
  12. Scott DA, Carmi R, Elbedour K, Yosefsberg S, Stone EM, Sheffield VC. An autosomal recessive nonsyndromic-hearing-loss locus identified by DNA pooling using two inbred Bedouin kindreds. Am J Hum Genet 1996;59:385-391. [Medline]
  13. Hoehe M, Berrettini W, Leppert M, et al. Genetic mapping of adrenergic receptor genes. Am J Hum Genet 1989;45:Suppl:A143-A143.abstract 
  14. Yang-Feng TL, Xue FY, Zhong WW, et al. Chromosomal organization of adrenergic receptor genes. Proc Natl Acad Sci U S A 1990;87:1516-1520. [Free Full Text]
  15. Bullrich F, Druck T, Kunapuli P, et al. Chromosomal mapping of the genes GPRK5 and GPRK6 encoding G protein-coupled receptor kinases GRK5 and GRK6. Cytogenet Cell Genet 1995;70:250-254. [Medline]

 

This Article
-Abstract
- PDF

Commentary
-Letters

Tools and Services
-Add to Personal Archive
-Add to Citation Manager
-Notify a Friend
-E-mail When Cited

More Information
-PubMed Citation

Related Letters:

Familial Atrial Fibrillation
MacRae C. A., Roberts R.
Extract | Full Text  
N Engl J Med 1997; 337:350, Jul 31, 1997. Correspondence

This article has been cited by other articles:



HOME  |  SUBSCRIBE  |  SEARCH  |  CURRENT ISSUE  |  PAST ISSUES  |  COLLECTIONS  |  PRIVACY  |  HELP  |  beta.nejm.org

Comments and questions? Please contact us.

The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved.