|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||
Background Cystic fibrosis is a life-limiting autosomal recessive disorder with a highly variable clinical presentation. The classic form involves characteristic findings in the respiratory tract, gastrointestinal tract, male reproductive tract, and sweat glands and is caused by loss-of-function mutations in the cystic fibrosis transmembrane conductance regulator (CFTR ) gene. Nonclassic forms of cystic fibrosis have been associated with mutations that reduce but do not eliminate the function of the CFTR protein. We assessed whether alteration in CFTR function is responsible for the entire spectrum of variant cystic fibrosis phenotypes.
Methods Extensive genetic analysis of the CFTR gene was performed in 74 patients with nonclassic cystic fibrosis who had been referred by 34 medical centers. We evaluated two families that each included a proband without identified mutations and a sibling with nonclassic cystic fibrosis to determine whether there was linkage to the CFTR locus and to measure the extent of CFTR function in the sweat gland and nasal epithelium.
Results Of the 74 patients studied, 29 had two mutations in the CFTR gene, 15 had one mutation, and 30 had no mutations. A final genotype of two mutations was more common among patients who had been referred after screening for common cystic fibrosis–causing mutations identified one mutation than among those who had been referred after screening had identified no such mutations (26 of 34 patients vs. 3 of 40 patients, P<0.001). Comparison of clinical features and sweat chloride concentrations revealed no significant differences among patients with two, one, or no CFTR mutations. Haplotype analysis in the two families revealed no linkage to CFTR. Although each of the affected siblings had elevated sweat chloride concentrations, measurements of cyclic AMP–mediated ion and fluid transport in the sweat gland and nasal epithelium demonstrated the presence of functional CFTR.
Conclusions Factors other than mutations in the CFTR gene can produce phenotypes clinically indistinguishable from nonclassic cystic fibrosis caused by CFTR dysfunction.
To further our understanding of the role of CFTR dysfunction in the development of the cystic fibrosis phenotype, we extensively analyzed CFTR genes in 74 patients with nonclassic cystic fibrosis who were referred for confirmatory genetic diagnosis. Each patient had had either one or no mutations identified on screening for common cystic fibrosis–causing alleles. We also studied linkage to the CFTR locus and CFTR function in two families that each included two siblings with nonclassic cystic fibrosis.
Methods
Patient Population
All studies were approved by the Johns Hopkins Joint Committee on Clinical Investigation, and written informed consent was obtained from all participants or their parents. Seventy-four patients were referred to the Cystic Fibrosis Foundation Genotyping Center from 1998 to 2002 by 34 centers that specialize in the care of patients with cystic fibrosis. Patients with a nonclassic phenotype were enrolled if they had either one mutation or none (but not two mutations) identified by a screening test that can detect at least 80 percent of common cystic fibrosis–causing alleles.4 A nonclassic phenotype was defined by the presence of two or three (but not all four) of the following: sweat-gland dysfunction (as evidenced by a sweat chloride concentration of more than 40 mmol per liter in a sample of at least 100 mg of sweat induced by means of pilocarpine iontophoresis3); cystic fibrosis–related respiratory disease, defined by the presence of sputum cultures positive for Pseudomonas aeruginosa, bronchiectasis, chronic productive cough, documented pneumonia, reactive airway disease, abnormal results on pulmonary-function tests, chronic sinusitis, and nasal polyposis, alone or in combination; malformation of the vas deferens on the basis of semen analysis, physical examination, or both; and gastrointestinal disease, as defined by fecal-fat values in excess of 7 g per day during a 72-hour period, rectal prolapse, or recurrent episodes of pancreatitis. Gastrointestinal disease was diagnosed on the basis of the finding of reduced malabsorption and steatorrhea after pancreatic enzyme supplementation in two patients and by the finding of reduced immunoreactive serum trypsinogen levels and clinical evidence of malabsorption in one patient.
CFTR Genotyping
Point mutations make up approximately 98.5 percent of reported mutations of the CFTR gene, and genomic rearrangements account for the remaining 1.5 percent of mutations.14 Thus, to detect mutations in CFTR that were not identified by the standard screening test, we used DNA sequencing optimized for the detection of heterozygous point mutations. For each patient, 31 polymerase-chain-reaction (PCR) products were generated from genomic DNA covering all 27 exons of the gene and 2 intronic mutations known to cause abnormal splicing (3849+10kbC
T and 1811+1.6kbAG). Bidirectional dye-primer labeled-DNA sequencing was performed on each amplicon (Applied Biosystems), and the resulting sequences were analyzed by automated capillary electrophoresis (ABI-310 or ABI-3100 Genetic Analyzer, Applied Biosystems). All exonic and bordering intronic sequences (25 bp before each exon and 15 bp after each exon) were analyzed independently by two reviewers who were unaware of subjects' identities, using Factura software (Applied Biosystems) and the Sequencher analysis program (Gene Codes). The chromatographs were visually inspected if mutations were not identified by the preceding analyses. Samples from 41 of the 45 patients who did not have two mutations identified after this analysis underwent bidirectional sequencing of the CFTR basal promoter (290 bp upstream of the translation starting site) to identify potential regulatory mutations.
Southern blotting was performed to detect rearrangements in genomic DNA in 25 of the 45 patients who did not have two detectable mutations after DNA sequencing. Total genomic DNA was digested with SstI, NheI, KpnI, or BglI; separated on 0.5 percent agarose gels; transferred to nitrocellulose filters; and hybridized with probes randomly labeled with phosphorus-32. Probes were derived from the pBQ4.7 plasmid containing CFTR complementary DNA (cDNA).
To determine the validity of our sequencing method, we also evaluated nine subjects whose genotypes had previously been determined (five normal subjects, two carriers of cystic fibrosis, and two patients with classic cystic fibrosis) in a blinded fashion and seven unrelated patients with classic cystic fibrosis (four had one previously identified common cystic fibrosis allele and three had no common cystic fibrosis alleles).
Family Studies
Haplotype studies were performed in two families (Families 1 and 2) that each included two affected siblings to determine whether the CFTR locus was linked to the disease phenotype. Five microsatellite markers (D7S523, D7S486, IVS17bTA, D7S480, and D7S490) in a 5-cM region encompassing the CFTR locus were typed.15 Markers were amplified by PCR with one unlabeled primer and one labeled primer (6-FAM [6-carboxyfluorescein]). Sizes were estimated against an internal standard (TAMRA-500 [6-carboxytetramethylrhodamine]) with capillary electrophoresis (ABI-310, Applied Biosystems) and GeneScan software (Applied Biosystems).
CFTR Functional Studies
Sweat rates for each of the four affected siblings in Families 1 and 2 were calculated at the Johns Hopkins Hospital.16 Sweat rates were measured in the forearms of each subject over a 20-minute period after the subcutaneous injection of 0.3 ml of Ringer's solution containing isoproterenol (50 µmol per liter) and aminophylline (5 mmol per liter to increase intracellular cyclic AMP), and atropine sulfate (140 µmol per liter) to block cholinergic stimulation. We also compared results to the mean (±SD) values in 30 healthy volunteers and in 30 patients with cystic fibrosis.16
The nasal potential difference was measured in the two affected siblings in each family at Children's Hospital Medical Center in Cincinnati and Children's Hospital of the University of Alabama at Birmingham with the use of procedures described by Knowles et al.17 Mean (±SD) values were calculated in 10 healthy control subjects and 10 patients with cystic fibrosis who were evaluated at Johns Hopkins Hospital.18 These values were similar to those previously reported17 and to values recorded for healthy controls and patients with cystic fibrosis at the centers in Cincinnati and Birmingham.
Analysis of CFTR Transcripts
Epithelial cells from the inferior nasal turbinates were collected with a brush (CytoSoft, Medical Packaging) from Subject II-2 in Family 1 and Subjects II-2 and II-4 in Family 2. RNA was extracted with the use of RNAzol B (Tel-Test) according to the manufacturer's directions, and cDNA was synthesized with the use of reverse-transcriptase (ProSTAR Ultra HF RT-PCR system, Stratagene). CFTR cDNA was amplified with the use of five sets of CFTR-specific primers and sized with the use of agarose-gel electrophoresis to check for splice variants. The entire CFTR cDNA from each patient was sequenced in both directions to identify mutations introduced after transcription and to confirm the presence of polymorphic variants in the cDNA.
Statistical Analysis
Chi-square tests were used to compare the frequencies of recurrent cough and wheezing, pseudomonas infection, nasal polyposis, steatorrhea, and congenital absence of the vas deferens according to the genotype. The sweat chloride concentrations, the forced expiratory volume in one second (FEV1), age at diagnosis, and age at referral were compared according to genotype with the use of one-way analysis of variance. The JMP statistical package (version 3.2.2, SAS Institute) was used for all statistical analyses.
Results
The clinical presentation, sweat chloride concentrations, and the number of mutations identified by screening for common CFTR mutations at referral are shown in Table 1 for the 74 patients with nonclassic cystic fibrosis. Approximately half the patients had elevated sweat chloride concentrations (
40 mmol per liter) and clinical manifestations limited to the respiratory tract, and about one third had clinical symptoms in more than one organ system. There was no difference in the number of organ systems involved or in the mean sweat chloride concentrations between patients with one mutation and those with no mutations at referral (Table 1).
|
|
|
|
|
|
Discussion
Although classic cystic fibrosis is a monogenic disorder resulting from sequence variants in the CFTR gene,4,7,11 mutations in CFTR have also been found in patients with nonclassic cystic fibrosis phenotypes.12,13,19,20 The identification of mutations in CFTR in a large fraction of our patients confirms the involvement of this gene in the nonclassic phenotype. Indeed, each of the 29 patients with two identified CFTR mutations had at least one mutation that was predicted to be associated with residual CFTR function, consistent with the hypothesis that nonclassic cystic fibrosis can be caused by at least one mild mutation.21 Those with only one CFTR mutation may harbor a second mutation that was not detectable with the methods that we used. Alternatively, a single CFTR mutation in combination with specific alleles of other genes or an unfavorable environment may produce a nonclassic phenotype. This possibility is supported by numerous studies indicating that persons with a single mutation in CFTR are at higher risk than persons in the general population for a clinical feature of cystic fibrosis.22,23,24,25,26,27
A surprising result of our study was the large proportion of patients with symptoms of cystic fibrosis who did not have CFTR mutations. Although some mutations may have been missed, several lines of evidence suggest that undetected mutations would not explain our findings. First, the analytic validity of the DNA-sequencing technique for the detection of mutations in the heterozygous state was 100 percent in our blinded sample from nine subjects and in samples from the seven subjects with classic cystic fibrosis. We observed the same level of sensitivity when analyzing 35 carriers of 
-thalassemia, and 51 carriers of X-linked adrenoleukodystrophy.28 Second, if mutations were missed, we would expect these unidentified alleles to be distributed equally between patients who were referred with a common cystic fibrosis–causing mutation and those who did not have such a mutation. Instead, the rates of detection of mutations were five times as high in patients with a common mutation. Finally, the nonclassic phenotype was not linked to the CFTR gene in Families 1 and 2, and the sweat rates and nasal potential-difference values were consistent with normal CFTR function.
Taken together, these data suggest that factors other than CFTR dysfunction can cause the nonclassic phenotype. A previous report described a patient with a nonclassic phenotype in whom no CFTR mutations could be found.29 Although we cannot rule out the intriguing possibility of an environmental cause, our identification of affected siblings with similar clinical and electrophysiological features suggests that a genetic basis is likely. Of course, the ultimate proof of a genetic cause is the identification of the causative gene or genes.
Patients with CFTR mutations could not be distinguished from those without CFTR mutations on the basis of the sweat chloride concentration or clinical presentation, highlighting the need for tests that differentiate patients with CFTR dysfunction from those with cystic fibrosis from other causes. Screening for common CFTR mutations can be a valuable tool in evaluating the role of CFTR in the nonclassic phenotype. Our data suggest that the presence of a single common mutation predicts the presence of a second disease-causing mutation and that the absence of a common mutation predicts normal CFTR status. There were, however, several exceptions to these generalizations (Table 3), suggesting that great care must be taken to ensure that extensive genotypic analysis is conducted before the diagnosis of nonclassic cystic fibrosis without CFTR mutations is made. Methods of measuring CFTR function other than pilocarpine iontophoresis (e.g., with the use of nasal potential-difference values and cyclic AMP–mediated sweat rate) may also be helpful in distinguishing CFTR dysfunction from other causes of the nonclassic phenotype.
The recognition and characterization of different causes of cystic fibrosis should facilitate the discovery of new pathways important for disease pathophysiology. This strategy may be particularly informative for the study of cystic fibrosis, since there is considerable controversy regarding the exact mechanism by which defective CFTR causes disease.30,31,32 Of particular interest is whether the defective pathways in these patients involve CFTR or act independently of CFTR. Further study of our patients may reveal additional pathways contributing to cystic fibrosis phenotypes, and these pathways may ultimately prove more amenable to therapeutic intervention than CFTR itself.
Supported by grants from the Cystic Fibrosis Foundation and the National Institutes of Health (to Dr. Cutting) and by a Clinical Research Center Grant from the National Institutes of Health (RR00052, to Dr. Zeitlin).
We are indebted to J.P. Clancy and V. Kociela for assistance with nasal potential-difference measurements; to L. Brass-Ernst and S. Watts for assistance with the calculation of the sweat rates; to M. Boyle, C. Merlo, R. Enke, and M. Couch for assistance with nasal brushings; to C. Yurk, J. Kirk, and the Johns Hopkins DNA Diagnostic Laboratory for technical assistance; to B. Karczeski for assistance with patient enrollment and the collection of clinical data; to P. Campbell, B. Rosenstein, T. Holtzman, W. Guggino, R. McWilliams, and A. Hamosh for their critique of the manuscript; to the clinicians and staff of the cystic fibrosis centers who referred patients; and most of all, to the patients and families for their willingness to participate in this study.
Source Information
From the McKusick–Nathans Institute of Genetic Medicine and Cystic Fibrosis Foundation Genotyping Center (J.D.G., M.E.M., G.R.C.), the Predoctoral Training Program in Human Genetics (J.D.G.), and the Department of Pediatrics (P.L.Z., G.R.C.), Johns Hopkins University School of Medicine, Baltimore; and the Department of Pediatrics, Saint Louis University School of Medicine, St. Louis (R.W.W.).
Address reprint requests to Dr. Cutting at Childrens Medical Surgical Center 9-125, 600 N. Wolfe St., Baltimore, MD 21287, or at gcutting{at}jhmi.edu.
References
F508 homozygous cystic fibrosis patients. Am J Respir Crit Care Med 1998;157:484-490.
| |||||||||||||||||||||||||||||||||||||||||||||||||||
Related Letters:
Variant Cystic Fibrosis Phenotypes in the Absence of CFTR Mutations
Couper R. T.L., Groman J. D., Cutting G. R.
Extract |
Full Text |
PDF
N Engl J Med 2002;
347:1892-1893, Dec 5, 2002.
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 © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
