PEDIATRICS Vol. 107 No. 2 February 2001, pp. 280-286

Comprehensive Mutation Screening in a Cystic Fibrosis Center

Jeffrey J. Wine, PhD^*, Eugene Kuo, BS^*, Gregory Hurlock, BS^*, and Richard B. Moss, MD

From the ^* Cystic Fibrosis Research Laboratory, Stanford University, Stanford, California, and  Department of Pediatrics, Stanford University School of Medicine, Stanford, California.

	ABSTRACT

Top Abstract MaterialsMethods Results Discussion References

Objectives and Background.  The identities of a cystic fibrosis (CF) patient's CFTR mutations can influence therapeutic strategies, but because >800 CFTR mutations exist, cost-effective, comprehensive screening requires a multistage approach. Single-strand conformation polymorphism and heteroduplex analysis (SSCP/HA) can be an important part of mutation detection, but must be calibrated within each laboratory. The sensitivity of a combined commercial-SSCP/HA approach to genotyping in a large, ethnically diverse US center CF population has not been established.

Study Design.  We screened all 27 CFTR exons in 10 human participants who had an unequivocal CF diagnosis including a positive sweat chloride test and at least 1 unknown allele after commercial testing for the 70 most common mutations by SSCP/HA. These participants were compared with 7 participants who had negative sweat tests but at least 1 other CF-like symptom meriting complete genotyping.

Results.  For the 10 CF participants, we detected 11 of 16 unknown alleles (69%) and all 4 of the known alleles (100%), for an overall rate of 75% inpatients not fully genotyped by conventional 70 mutation screen. For 7 participants with negative sweat tests, we confirmed 1 identified mutation in 14 alleles and detected 3 additional mutations. Mutations detected in both groups included 7 missense mutations (S13F, P67L, G98R, S492F, G970D, L1093P, N1303K) and 9 deletion, frameshift, nonsense or splicing mutations (R75X, G542X, F508, 451-4588 bp, 5T, 663T, exon 13 frameshift, 1261+1GA and 3272-26AG). Three of these mutations were novel (G970D, L1093P, and 451-4588 bp¹). Thirteen other changes were detected, including the novel changes 1812-3 ins T, 4096-278 ins T, 4096-265 ins TG, and 4096-180 TG.

Conclusion.  When combined with the 70 mutation Genzyme test, SSCP/HA analysis allows for detection of >95% of the mutations in an ethnically heterogeneous CF center population. We discuss 5 possible explanations that could account for the few remaining undetected mutations.  Key words:  CFTR, alleles, SSCP, heteroduplex analysis.

The ability to detect disease-causing mutations rapidly and with high sensitivity is transforming many areas of medicine. With regard to cystic fibrosis (CF), a common human genetic disease, identifying the mutations that cause disease can have therapeutic consequences, because a number of allele-specific therapies are in development. Identification of mutations is particularly important in carrier studies and prenatal diagnosis. Genetic counseling can be exact and confident when mutations are known, allowing family members to make fully informed decisions based on a known risk for carrier status or disease inheritance, respectively. In the absence of mutation identification risk can only be estimated, leading to considerable uncertainty, anxiety, and potentially mistaken conclusions and actions. However, mutation detection is not trivial because >800 mutations in a single gene have been reported to cause CF.

CF is a recessive disease that affects ~1/2000 Caucasians and smaller proportions of all other human populations, each of which have at least some distinctive mutations. CF is caused by mutations in CFTR, a gene that comprises ~250 kb on chromosome 7 encompassing 27 exons. CFTR codes for an integral membrane protein of 1480 amino acids, the cystic fibrosis transmembrane conductance regulator (CFTR).^2-4 CFTR is an anion channel expressed primarily in the apical membranes of wet epithelia, where it participates in fluid secretion and salt absorption. CFTR may be more susceptible to mutations than an average protein because of its large size and sensitivity to being misprocessed.

The large and growing number of CFTR mutations requires a multiple-step procedure for efficient detection of mutations. In most populations a few alleles account for the majority of mutations, and simple tests exist for their detection. Patients negative for the most common mutations can then be screened with a powerful commercial assay that detects the 70 most common mutations in the North American population (see "Methods"). Screening for the remaining 700+ mutations and for novel mutations requires methods that can detect mutations solely based on the physical properties of DNA.

A sensitive and widely used method for such detection is single strand conformation polymorphism and heteroduplex analysis (SSCP/HA).⁵ Although SSCP/HA is extremely sensitive when optimized, different laboratories have obtained different results when using SSCP/HA, and it was therefore necessary to quantify the accuracy of the method as used in the Stanford CF Center. The method used here is a modified version⁶ of a method optimized for detecting CFTR mutations.⁷ In this report, we calibrate the sensitivity of our SSCP/HA method by screening participants known or suspected of having CF. Denaturing gradient gel electrophoresis (DGGE) is considered to be the most sensitive method, but is more laborious than SSCP/HA. Our results indicate the method is as sensitive as DGGE. When used to resolve mutations not detected by simpler methods, the combination can detect 96% of the mutations in our Center population. We discuss possible reasons why the remaining 4% of mutations cannot be detected with this approach.

	MATERIALS AND METHODS

Top Abstract MaterialsMethods Results Discussion References

Participants were 263 Stanford CF Center patients and 7 other individuals referred for testing who exhibited 1 or more symptoms associated with CF, but had negative or borderline sweat chloride values. Diagnosis of CF was made based on a pilocarpine iontophoresis sweat chloride test performed according to National Committee for Clinical Laboratory Standards, Inc. (NCCLS) standards, with values >60 mM chloride considered as diagnostic for CF in patients with consistent clinical presentation.

All participants were screened with the Genzyme Genetics Assay Genzyme70 test according to the testing laboratory's instructions. After Genzyme70 testing, 84 alleles in 78 patients remained unidentified. From this group we selected 10 participants for SSCP/HA genotyping. Several participants were selected at random, but 6 participants were tested with SSCP/HA because both alleles remained unknown after Genzyme70 testing.

Blood for SSCP/HA genotyping was obtained by venipuncture after informed consent. This study was approved by the Human Subjects Panel (institutional review board) of Stanford University Medical Center. For on-site SSCP/HA analysis, genomic DNA was purified using a commercially available kit (Puregene, Gentra Inc., Minneapolis, MN) from a 2-ml blood sample.

Polymerase Chain Reactions (PCR)

All 27 exons of CFTR together with some flanking intronic sequence were amplified using conventional PCR. Exon 13 was amplified as 2 overlapping PCR products, while the remaining exons were amplified as single products ranging from 168 to 562 bp in length. Reactions were conducted as 10 µL reaction mixtures containing 40 ng DNA, 0.2 unit Taq polymerase, 2.5 pmol of each primer, 50 mM KCl, 2.5 mM MgCl₂, 10 mM Tris (pH 8.3), 200 µM each of dATP, dGTP, dTTP dCTP; 0.5 µCi of ³²PdCTP (3000 Ci/mM). Amplification parameters were: denature 6 minutes at 94°C, then 30 cycles of: denature for 30 seconds at 94°C, anneal 30 seconds at 55°C, extend 1 minute at 72°C; followed by 7 minutes at 72°C. The 28 pairs of primers used were described previously.⁶

Screening

All 27 exons of CFTR were screened using SSCP/HA. Enough of the adjacent intronic sequence was included to detect most splice mutations. For detection, 10 µL of a denaturing mixture consisting of 95% Formamide, 20 mM EDTA, 0.05% bromophenol blue, 0.05% xylencyanol, and 20 mM NaOH was added to the PCR reaction mixture. The mixture was then heated to 95°C for 2 minutes, snap cooled in an ice/water bath, and 1 to 2 µL loaded in 1 well of a 96 lane, polyacrylamide gel. The gel formulation was MDE (FMC BioProducts, Rockland, ME) plus 10% glycerol.⁸ For each exon, a control (non-CF) DNA sample was run in an adjacent lane and a third lane was a mixture of the control and sample DNA to detect heteroduplexes caused by homozygous changes. Thus each gel contained at least 28×3 = 84 lanes. In some cases extra controls were run for common genetic variations such as GATT_6/7 or M470V. Candidate sequence variations were detected as shifts in the DNA migration pattern. Gels were run in a 4°C cold room for 4 to 8 hours at 55 W, dried and autoradiographed.

Sequencing

PCR-amplified DNA was gel-purified and then run through QIAquick columns (QIAGEN, Chatsworth, CA) followed by cycle sequencing. For sequencing, 30 to 60 ng PCR product, 0.4 pmol primer, and 8.0 µL Ready Reaction Mix (Perkin Elmer, Foster City, CA) were mixed to a final volume of 20 µL. Thermal cycling of the reaction mix consisted of 25 cycles of 96°C for 10 seconds, 50°C for 5 seconds, and 60°C for 4 minutes. Extension products were purified with Centri-Sep spin columns (Princeton Separations, Adelphia, NJ) and sequenced on an ABI 373 automated sequencer. All exons were sequenced in both directions.

	RESULTS

Top Abstract MaterialsMethods Results Discussion References

The Stanford CF mutation database contains information on 526 alleles. After screening of most patients with the Genzyme70 assay, 84% of the mutations were identified. (Twenty-seven patients who have not yet undergone genetic testing were not included in this analysis, and 6 unknown alleles occur in participants who have not yet been screened with the Genzyme70 assay.) Of total alleles identified in this way, 353 (80%) are F508 and 89 (20%) are other identified mutations. For the entire population the figures are 67% F508, 17% other identified mutations, and 16% unknown mutations (Table 1). Thus, excluding F508, the Genzyme70 assay identified ~50% of the alleles in this population. The ethnicity of our presently followed population is shown in Table 2, it is representative of the total population included in Table 1.

From the patients with unknown alleles, 10 patients with positive sweat tests were selected for SSCP/HA analysis based on clinical status, ethnicity, and previous screening with the Genzyme70 assay. Of the 20 alleles represented by this group, only 4 (20%) had been identified with the Genzyme70 assay. Full-gene scanning of each of these patients typically revealed shifts in 3 to 5 exons. In some cases, the identity of common DNA changes was obvious simply by comparing the SSCP/HA patterns with appropriate controls. In all other cases, the exons showing a shift were sequenced. The results are summarized in Table 3 and Fig 1. Eleven of the 16 unknown mutations (69%) were detected. Figure 1 shows the SSCP patterns associated with each of the mutations detected, and the accompanying chromatograms. When our results for this sample are extrapolated to the full Stanford CF population, they predict a mutation detection rate of 95% for combined screening with Genzyme70 followed by SSCP/HA.

View larger version (46K): [in this window] [in a new window]   Fig. 1.   SSCP/HA autoradiographs and sequence chromatograms for 10 mutations not detected with the Genzyme70 assay. A. Autoradiographs. Each panel contains 2 side-by-side lanes. The first lane is the control and the second lane is the sample with the mutation. Mutation 663T was homozygous: the third lane shows the mixture of the control and 663T samples. For most mutations a bottom panel is included to show heteroduplexes. Scans picked up arrows marked on some of the original gels. B. Chromatograms from ABI370 sequencing of double-strand DNA are shown for each of the corresponding gel shifts observed in A. Arrows indicate mutation sites. All mutations were heterozygous except for 663T. Novel mutations 451-458 and G970D were reported separately1; novel mutation L1093P has been submitted for publication. For each mutation a forward and reverse sequence was obtained, but only one of those is shown (usually the cleanest sequence was selected).

A second group of 7 individuals with negative sweat tests was screened. This heterogeneous group included otherwise healthy individuals with chronic sinusitis, parents of CF children who themselves displayed some clinical signs of CF, and patients with various kinds of lung disease. One patient was pancreatic insufficient despite a normal sweat test. For this group the results were quite different. Only 1 mutation in 14 chromosomes was detected by the Genzyme70 test in this group. After screening with SSCP/HA, 2 mutations, P67L and 1261+1GA, were detected in another patient who also had the highest sweat chloride value in this group (48 mM). No other mutations were found with SSCP/HA in the entire group (Table 4).

For both groups, SSCP/HA detected all mutations that had been detected with the Genzyme70 test, and in no case did we detect a mutation included in the Genzyme70 test but not detected by that test.

	DISCUSSION

Top Abstract MaterialsMethods Results Discussion References

The dual purposes of this research were to calibrate our SSCP/HA procedures while concomitantly identifying unknown CFTR mutations in participants known or suspected to have CF.

Sensitivity of the Modified SSCP/HA Method

The method we used is similar to most SSCP techniques except that 10% glycerol was added to an MDE gel formulation, and a single condition was used for all mutations. This method is much simpler than DGGE, which detects virtually all mutations.^9,10 How well does our present method compare with DGGE? In our small sample of participants with positive sweat tests, we detected 69% of mutations that were not included in the Genzyme 70 mutation test. At first glance this may seem to indicate insensitivity. However, certain mutations, such as deletions of 1 or more exons, are not detected by either method. Therefore, the apparent sensitivity of the method will be affected to the extent that such mutations occur within the population being tested. We propose that such differences in populations can explain much of the reported variation in sensitivity of the 2 methods.

As shown in Table 4, the general utility of the sweat test in differentiating CF is confirmed, in that the majority of patients referred for suggestive symptoms but with negative sweat tests were without detectable CFTR mutations. It also illustrates the converse point that several such patients (SN1, SN6, SN7) do have mutations missed by commercial genotyping. Finally, the finding of CFTR polymorphisms (genetic variations that are supposedly inconsequential) in such patients suggests that in the presence of unidentified modifying genetic or environmental influences, these changes may play some role in pathophysiology.

Several studies have used DGGE, either alone or in combination with other methods, to screen for CF mutations in particular populations. As shown in Table 5, the total percentage of mutations detected ranged from 60% to 96%. Analysis of these studies shows how the proportion of detected mutations changes according to sample characteristics and assay used. In the largest study using DGGE, Tzetis et al detected 85.6% of the mutations in 500 chromosomes from a Greek population of CF patients.¹¹ Of the detected mutations, 384 (76.8%) are detectable by the Genzyme70 test (261 were F508), leaving 116 mutations that would have been unknown after Genzyme70 screening. From the set of 116 mutations not detectable by the Genzyme70 test, the DGGE method used by Tzetsis et al detected 44 additional mutations (38%). Similar analyses were applied to the other studies with the results shown in Table 5. For the 6 studies, the Genzyme70 test detects from 21% to 87.5% of mutations, with lower values in populations for which the test was not optimized, such as Pakistanis and Tunisians. For the remaining, unidentified mutations, DGGE detected between 24% to 40% of mutations^11-14 (Table 5). In our study using SSCP and in a study of Pakistanis using SSCP + DGGE,¹⁵ 69% to 80% of mutations not included in the Genzyme70 test were detected. We do not propose that this difference indicates a superiority of SSCP/HA. Instead, our interpretation is that the 2 methods are each capable of detecting virtually all point mutations and small indels within the coding region of CFTR, but that some populations harbor a larger proportion of individuals whose disease is caused by mutations that are not detectable with either SSCP/HA or DGGE.

What can explain the missing mutations? There are at least five possibilities. (1) Mutations that lie outside the coding region and the immediate flanking exons are not detected by typical SSCP or DGGE procedures, which do not target deep intronic regions. Examples of such mutations are 3849+10kb CT¹⁶ and 1811+1.6kbAG.¹⁷ (2) Large deletions of entire exons or multiple exons will also be missed, because in patients heterozygous for such mutations the only indication might be a lighter than normal banding pattern on the SSCP/HA gel. Seven examples of such large mutations have been reported to the Cystic Fibrosis Genetic Analysis Consortium, and their incidence is likely to be underestimated. It is interesting that several large deletions have been reported for individuals originally from Turkey, Cyprus, or Iran. (3) Changes considered to be inconsequential genetic variations may actually affect CFTR function adversely, perhaps in conjunction with other features of the DNA sequence that are as yet unrecognized as being important. For example, some evidence suggests that M470V, a very common DNA change, is actually a mild mutation that decreases the open probability of the CFTR ion channel and reduces splicing efficiency.¹⁸ Two of our CF participants for whom a second mutation could not be identified with SSCP/HA are homozygous for M470V (Table 4). (4) Syndromes sharing some features of CF might arise in rare cases for reasons other than CFTR mutations, or because of an interaction between one CFTR mutation and other aspects of a person's physiology, including mutations in other genes. For example, CFTR mutations are significantly elevated in cases of sarcoidosis and of disseminated bronchiectasis of unknown cause.¹⁹ (5) Finally, some mutations within coding regions are almost certainly missed by DGGE and SSCP/HA. However, because of the contribution of the other 4 factors, we propose that the proportion of point mutations within the coding region that are not detected by these methods is much smaller than the total percentage of missing mutations.

Mutation Detection and Allele-Specific Therapeutics

How can detailed information about a patient's alleles benefit the patient? Mutations in CFTR cause CF disease by distinct mechanisms,²⁰ some of which may be corrected with allele-specific approaches. Stop mutations account for about 7% of the patients at this center, and some of these mutations may be amenable to correction by encouraging read-through with agents such as aminoglycosides.^21,22 The mutation G551D, which interferes with efficient adenosine triphosphate gating of CFTR, accounts for ~4% of our patients, and experiments indicate that the activity of G551D-CFTR can be increased with the isoflavone genistein.²³ Whereas these kinds of mutations are relatively rare, mutations that lead to improper folding of CFTR, such as F508, are responsible for the great majority of CF.²⁴ The molecular basis for these processes is under intense investigation,²⁵ and strategies for inducing an increased proportion of properly folded, functional molecules at the plasma membrane have achieved success in vitro^26-28. In addition, symptoms of CF disease caused by any mutation that gives rise to some functional CFTR might be improved simply by increasing the production of CFTR.²⁹

In conclusion, our results show that ~96% of mutations in a heterogeneous American CF population can be efficiently detected by a 2-step procedure in which patients are screened for 70 common mutations, and then by SSCP/HA, if necessary. In our population the procedure could be made even more economical by first screening for F508 in-house, because 46.6% of our patients are homozygous for the F508 mutation. We confirm the well-established point that the results of sweat testing are highly predictive for the eventual detection of CFTR mutations.

	ACKNOWLEDGMENTS

This work was supported by the Cystic Fibrosis Foundation, the Ross Mosier Fund, and by gifts from Ronald and Kay Presnell and Patricia Bresee.

We thank our colleagues Ann Harkins, Judy Palmer, Jeffrey Riker, Terry Robinson, and John Wagner for directing our attention to certain patients and for help in collection of DNA. Al Smith and the staff of the Protein and Nucleic Acid facility, Center for Molecular Genetics and Medicine, helped with DNA sequencing. Sequence analysis was conducted with programs made available on the internet by the Human Genome Center, Department of Molecular and Human Genetics, Baylor College of Medicine.

	FOOTNOTES

The Genzyme Genetics Assay (referred to throughout as "Genzyme70" is a diagnostic procedure that detects the 70 most common mutations in the North American CF population. A table of the mutations detected by the procedure is available at http://www.genzyme.com/prodserv/genetics/molgen/cystic.htm

Received for publication Feb 2, 2000; accepted Jun 12, 2000.

Address correspondence to Richard B. Moss, MD, Stanford University School of Medicine, Department of Pediatrics, Pulmonary Division, 701 Welch Rd, Suite 3328, Palo Alto, CA 94304-5786. E-mail: rmoss{at}leland.stanford.edu

	ABBREVIATIONS

CF, cystic fibrosis; CFTR, cystic fibrosis transmembrane conductance regulator; SSCP/HA, single strand conformation polymorphism and heteroduplex analysis; DGGE, denaturing gradient gel electrophoresis; PCR, polymerase chain reactions.

	REFERENCES

Top Abstract MaterialsMethods Results Discussion References

1. Wagner JA, Vassilakis A, Yee K, Two novel mutations in a cystic fibrosis patient of Chinese origin. Hum Genet 1999; 104:511-515 [CrossRef][Medline]
2. Rommens JM, Iannuzzi MC, Kerem B, Identification of the cystic fibrosis gene: chromosome walking and jumping. Science 1989; 245:1059-1065 [Abstract/Free Full Text]
3. Riordan JR, Rommens JM, Kerem B, Identification of the cystic fibrosis gene: cloning and characterization of complementary DNA. Science. 1989; 245:1066-1073 [Abstract/Free Full Text]
4. Kerem B, Rommens JM, Buchanan JA, Identification of the cystic fibrosis gene: genetic analysis. Science 1989; 245:1073-1080 [Abstract/Free Full Text]
5. Orita M, Iwahana H, Kanazawa H, Hayashi K, Sekiya T Detection of polymorphisms of human DNA by gel electrophoresis as single-strand conformation polymorphisms. Proc Natl Acad Sci U S A 1989; 86:2766-2770 [Abstract/Free Full Text]
6. Wine JJ, Glavac D, Hurlock G, Genomic DNA sequence of Rhesus (M. mulatta) cystic fibrosis (CFTR) gene. Mamm Genome. 1998; 9:301-305 [CrossRef][Medline]
7. Ravnik-Glavac M, Glavac D, Dean M Sensitivity of single-strand conformation polymorphism and heteroduplex method for mutation detection in the cystic fibrosis gene. Hum Mol Genet. 1994; 3:801-807 [Abstract/Free Full Text]
8. Ravnik-Glavac M, Glavac D, Komel R, Dean M Single-stranded conformation polymorphism analysis of the CFTR gene in Slovenian cystic fibrosis patients: detection of mutations and sequence variations. Hum Mutat. 1993; 2:286-292 [Medline]
9. Audrezet MP, Mercier B, Guillermit H, Identification of 12 novel mutations in the CFTR gene [published erratum appears in Hum Mol Genet 1993 Apr;2(4):496]. Hum Mol Genet. 1993; 2:51-54 [Abstract/Free Full Text]
10. Macek M Jr, Mercier B, Mackova A, Sensitivity of the denaturing gradient gel electrophoresis technique in detection of known mutations and novel Asian mutations in the CFTR gene. Hum Mutat. 1997; 9:136-147 [CrossRef][Medline]
11. Tzetis M, Kanavakis E, Antoniadi T, Doudounakis S, Adam G, Kattamis C Characterization of more than 85% of cystic fibrosis alleles in the Greek population, including five novel mutations. Hum Genet. 1997; 99:121-125 [Medline]
12. Hughes DJ, Hill AJ, Macek M, Jr, Redmond AO, Nevin NC, Graham CA. Mutation characterization of CFTR gene in. Northern Irish CF families: thirty mutations, including two novel, account for approximately 94% of CF chromosomes. Hum Mutat. 1996;206;8:340-347
13. Messaoud T, Verlingue C, Denamur E, . Distribution of CFTR mutations in cystic fibrosis patients of Tunisian origin: identification of two novel mutations. Eur J Hum Genet. 1996; 4:20-24 [Medline]
14. Plouvier E, Cougoureux E, Sardet A, Tournier G, Aymard P, Feldmann D CFTR gene analysis in cystic fibrosis patients: detection of 91% of molecular defects and identification of the novel mutation D979V. Ann Genet. 1997; 40:185-188 [Medline]
15. Malone G, Haworth A, Schwarz MJ, Cuppens H, Super M Detection of five novel mutations of the cystic fibrosis transmembrane regulator (CFTR) gene in Pakistani patients with cystic fibrosis: Y569D, Q98X, 296+12(T>C), 1161delC and 621+2(T>C). Hum Mutat. 1998; 11:152-157 [CrossRef][Medline]
16. Highsmith WE, Burch LH, Zhou Z, A novel mutation in the cystic fibrosis gene in patients with pulmonary disease but normal sweat chloride concentrations. N Engl J Med. 1994; 331:974-980 [Abstract/Free Full Text]
17. Chillon M, Dork T, Casals T, A novel donor splice site in intron 11 of the CFTR gene, created by mutation 1811+1.6kbA->G, produces a new exon: high frequency in Spanish cystic fibrosis chromosomes and association with severe phenotype. Am J Hum Genet. 1995; 56:623-629 [Medline]
18. Cuppens H, Lin W, Jaspers M, Polyvariant mutant cystic fibrosis transmembrane conductance regulator genes. The polymorphic (Tg)m locus explains the partial penetrance of the T5 polymorphism as a disease mutation. J Clin Invest. 1998; 101:487-496 [Medline]
19. Bombieri C, Benetazzo M, Saccomani A, Complete mutational screening of the CFTR gene in 120 patients with pulmonary disease. Hum Genet. 1998; 103:718-722 [CrossRef][Medline]
20. Welsh MJ, Smith AE Molecular mechanisms of CFTR chloride channel dysfunction in cystic fibrosis. Cell 1993; 73:1251-1254 [CrossRef][Medline]
21. Bedwell DM, Kaenjak A, Benos DJ, Suppression of a CFTR premature stop mutation in a bronchial epithelial cell line. Nat Med. 1997; 3:1280-1284 [CrossRef][Medline]
22. Howard M, Frizzell RA, Bedwell DM Aminoglycoside antibiotics restore CFTR function by overcoming premature stop mutations [see comments]. Nat Med. 1996; 2:467-469 [CrossRef][Medline]
23. Illek B, Zhang L, Lewis NC, Moss RB, Dong JY, Fischer H Defective function of the cystic fibrosis-causing mutation G551D is recovered by genistein. Am J Physiol. (Cell Physiol) 1999; 277:C833-C839 [Abstract/Free Full Text]
24. Cheng SH, Gregory RJ, Marshall J, Defective intracellular transport and processing of CFTR is the molecular basis of most cystic fibrosis. Cell 1990; 63:827-834 [CrossRef][Medline]
25. Kopito RR Biosynthesis and degradation of CFTR. Physiol Rev. 1999; 79:S167-S173
26. Denning GM, Anderson MP, Amara JF, Marshall J, Smith AE, Welsh MJ Processing of mutant cystic fibrosis transmembrane conductance regulator is temperature-sensitive [see comments]. Nature. 1992; 358:761-764 [CrossRef][Medline]
27. Sato S, Ward CL, Krouse ME, Wine JJ, Kopito RR Glycerol reverses the misfolding phenotype of the most common cystic fibrosis mutation. J Biol Chem. 1996; 271:635-638 [Abstract/Free Full Text]
28. Brown CR, Hong-Brown LQ, Welch WJ Strategies for correcting the delta F508 CFTR protein-folding defect. J Bioenerg Biomembr. 1997; 29:491-502 [CrossRef][Medline]
29. Rubenstein RC, Zeitlin PL A pilot clinical trial of oral sodium 4-phenylbutyrate (Buphenyl) in deltaF508-homozygous cystic fibrosis patients: partial restoration of nasal epithelial CFTR function. Am J Respir Crit Care Med. 1998; 157:484-490 [Abstract/Free Full Text]