Identification of mutations in exons 1 through 8 of the cystic fibrosis transmembrane conductance regulator (CFTR) gene

GENOMICS

10,

229-235

(19%)

Identification of Mutations in Exons 1 through 8 of the Cystic Fibrosis Transmembrane Conductance Regulator (CFTR) Gene JULIAN ZIELENSKI,*

DOMINIQUE BOZON,* BAT-SHEVA KEREM,*,’ DANUTA MARKIEWICZ,* JOHANNA M. ROMMENS,* AND LAP-CHEE TSUI*~

PETER DURIE,t

*Department of Genetics, The Hospital for Sick Children, Toronto, Ontario M5G 1X8, Canada; *Departments of Molecular and Medical Genetics, University of Toronto, Toronto, Ontario M5.S lA8, Canada; and tDepartment of Pediatrics, University of Toronto and Division of Gastroenterology, The Hospital for Sick Children, Toronto, Ontario M5G lA8, Canada Received

October

4, 1990;

revised

January

10, 1991

1991). The polypeptide predicted from completely sequenced cDNA consists of 1480 amino acids and it is named cystic fibrosis transmembrane conductance regulator (CFTR). On the basis of its remarkable sequence and structural similarity to many ATP-dependent transport proteins, it has been generally assumed that CFTR contains two highly hydrophobic transmembrane regions, each followed by two nucleotide-binding folds (NBFs) (Riordan et al., 1989). A deletion of 3 bp in exon 10 of the CFTR gene, which results in the loss of a phenylalanine residue at amino acid position 508 (AF508), is the most common mutation causing CF, accounting for approximately 70% of the CF chromosomes (Kerem et al., 1989b; Lemna et al., 1990). A large number of other mutations have also been identified in the CFTR gene (Cutting et al., 1990a,b; Dean et al., 1990; Kerem et uZ., 1990a; Vidaud et al., 1990; Guillermit et al., 1990; White et al., 1990), but most of them are rare in the population. These include missense, nonsense, frameshift, and RNA splicing mutations. Since the searches have been concentrated mostly on the two NBFs, many of the known mutations are also located in these regions. We have initiated a systematic analysis of all the CF chromosomes in our cohort of Canadian families (Tsui et al., 1986) and have reported mutations in the two NBFs (Kerem et al., 1990b). Since each of the CF chromosomes is characterized by extensive DNA marker haplotype (Kerem et al., 1989b), they may serve as excellent references for future population analysis. To continue our identification of disease-relevant mutations for delineation of the function of CFTR, we have screened the first eight exons of the gene by direct sequence analysis of genomic DNA amplified by the polymerase chain reaction (PCR; Saiki et al., 1988). In this communication, we report five (possibly six) apparent mutations within exons 1 through 8 of the CFTR gene.

Five different mutations have been identified in the gene causing cystic fibrosis (CF) through sequencing regions encompassing exons 1-8, including the 5’ untranslated leader. Two of these apparent mutations are missense mutations, one in exon 3 (Gly to Glu at position 85; GSBE) and another in exon 5 (Gly to Arg at 178; G178R), both causing significant changes in the corresponding amino acids in the encoded proteincystic fibrosis transmembrane conductance regulator (CFTR). Two others affect the highly conserved RNA splice junction flanking the 3’ end of exons 4 and5(621+1G+T,711+1G+T),resultinginaprobable splicing defect. The last mutation is a single-basepair deletion in exon 4, causing a frameshift. These five mutations account for the 9 of 31 non-AF508 CF chromosomes in our Canadian CF family collection and they are not found in any of the normal chromosomes. Three of the mutations, 621+ 1G + T, 711+ lG+T,andG85E,arefound in the French-Canadian population, with 621 + 1G --t T being the most abundant (5/7). There are two other sequence variations in the CFTR gene; one of them (129G --* C) is located 4 nucleotides upstream of the proposed translation initiation codon and, although present only on CF chromosomes, it is not clear whether it is a disease-causing mutation; the other (R75Q) is most likely a sequence variation within the coding region. o 1991 Academic press, IDC.

INTRODUCTION

The gene causing cystic fibrosis (CF) in humans has been recently identified (Rommens et al., 1989; Riordan et al., 1989; Kerem et al., 1989b). The 27 exons and the immediately flanking (intron) sequences of this gene have been determined (Zielenski et al., Sequence data from this article have been deposited with the EMBL/GenBank Data Libraries under Accession Nos. M55499M55504. 1 Present address: Department of Genetics, Hebrew University, Jerusalem 91904, Israel.

229

o&3%7543/91

$3.00

Copyright 0 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.

230

ZIELENSKI

MATERIALS

Patient

AND

ET

AL.

TABLE

METHODS

Samples

The 46 Canadian CF families used in this study had been previously described (Tsui et al., 1986). Extensive DNA marker haplotype information for the CF chromosomes in these families was available (Kerem et al., 1989b). Peripheral blood samples were also collected from additional CF patients and family members registered at the CF clinic of the Hospital for Sick Children. Informed consent was obtained for each participating individual, and blood samples were collected according to procedures approved by the HSC Human Subjects Review Committee. The pancreatic status of patients was defined as described (Kerem et al., 1990b). Data for one of the families (Toronto 17) reported here were deposited at the Human Genetic Mutant Cell Repository at Camden, New Jersey (GM1076). The criteria for diagnosis of pancreatic status were described (Gaskin et aZ., 1984; Durie et al., 1986; Kerem et al., 1989a).

Methods

Direct sequencing of PCR products was performed as described (Winship, 1989; Kerem et al., 1990a). The PCR primers B115-B, lOD, Zi-5, 3i-3, 4i-3, 5i-3, 6Ai-3,6Ci-3, ?i-5,8i-3, as described in the accompanying paper (Zielenski et al., 1991), were used for sequencing. An internal primer for exon 1 (primer A; 5’-CAAGTGCATAGTAGCGTACT-3’) was used. Allele-Specific Oligonucleotide (ASO) Hybridization The oligonucleotide hybridization conditions have been previously described (Kerem et al., 1989b). The normal and mutant-specific oligonucleotides are listed in Table 1. The hybridization was performed at 37°C and the washings were done twice with 5~ SSC

Method

Hintl digestion

G85E

556delA

621 + 1G +

T

711 + 1G +

T

G178R

129G

+ C

R75Q

DNA Sequence Determination

Mutations and Variations l-8 of CFTR

Name

Polymerase Chain Reactions DNA sequences spanning individual exons were amplified by PCR with oligonucleotide primers located in the respective flanking introns. The general procedures and the specific oligonucleotide primers used for the amplification of individual exon sequences have been described (Zielenski et at., 1991). Oligonucleotides were either purchased from the Hospital for Sick Children DNA Biotechnology Service Center or synthesized with the use of a Cruachem PS250 machine. PCR amplification of genomic DNA from cultured lymphoblasts or peripheral blood was performed in a Perkin-Elmer/Cetus DNA thermal cycler as described (Kerem et al., 1990a), except for exons 1 and 2 for which a modified protocol was used (Zielenski et al., 1991).

for Detecting in Exons

1

Normal: 32 + 172 + 105 bp Mutant: 32 + 277 bp (destruction of site) BglI digestion Normal: 438bp Mutant: 288 + 149 bp (creation of site) MseI digestion Normal: 35 + 71 f 299 + 33 bp Mutant: 35 f 71 f 245 + 54 + 33 bp (creation of site) AS0 hybridization; washing at 40°C Normal: GGTACATACTTCATCA Mutant: GGTACATAATTCATCA AS0 hybridization; washing at 37°C Normal: AAGTA’I’TGGACAACTT Mutant: AAGTATTAGACAACTT AS0 hybridization; was?ling at 49’C Normal: GCCCAGAGACCATGCA Mutant(?): GCCCAGACACCATGCA AS0 hybridization; washing at 51°C Common: AACATCGCCGAAGGG Variant: AACAmGCCGAAGGG

for 10 min each at room temperature followed by twice with 2X SSC for 30 min each at 45-57°C; the exact temperature was determined empirically for each AS0 as indicated. RESULTS

As an ongoing effort to screen for mutations in the CFTR gene, 18 unrelated individuals (including 4 patients and 14 obligate carriers) carrying 20 nonAF508 CF chromosomes were used as the primary panel for DNA sequence analysis. These chromosomes were chosen because their haplotypes, consisting of 23 DNA markers spanning the entire gene, were available for comparative analysis (Kerem et al., 1989b). In the present study, direct DNA sequencing of PCR-amplified product was performed on the first eight exons of the CFTR gene, which included the region extending from 5’ of the presumptive promoter of the gene to sequences flanking the 3’ end of exon 8. A total of seven nucleotide alterations, compared to the published cDNA sequence, was noted (Table 2). Each of the changes was verified by repeated sequencing and confirmed with a specific diagnostic test (as indicated below) in other family members. G85E A missense mutation was found in exon 3, in a region corresponding to the predicted transmembrane

CF

231

MUTATIONS

TABLE Mutations

and Sequence

Variations

2

in Exon

Regions

1 to 8 of the CFTR Number

Gene

of chromosomes

screened

CF Name

Amino

None Gln + Arg at 75 Gly * Glu at 85 Frameshift Splice mutation Splice mutation Gly + Arg at 178

129G + C R75Q G85E 556delA 621 + 1G + T 711+ 1G * T G178R Note.

mutation; parentheses

acid change

Nucleotide

change

G --) C at 129 T --* C at 356 G + A at 386 1-bp deletion G-*Tat621+1 G * Tat 711 + 1 G -+ A at 664

Mut.

1 3 3 4 Intron Intron 5

3

Number of chromosomes screened refers to the number of parental Mut., number of chromosomes with the indicated change; nonmut., indicates the number of chromosomes within the same haplotype

domain of CFTR. It was detected on paternal CF chromosome from family 26, a French-Canadian family classified as pancreatic insufficient (PI). As shown in Fig. 1, the patient was heterozygous for this mutation, which involved a G to A transition at nucleotide position 386, according to the numbering system of Riordan et al. (1989). The maternal CF chromosome was found to carry the 621 + 1G --) T mutation as described below. The G to A transition in G85E would predict a change in the encoded amino acid from Gly to Glu at amino acid position 85 of CFTR. G85E was found to be associated with the group IIb haplotype, as defined by Kerem eCal. (1989b) (see Fig. 2). The mutation destroyed a HinfI site which could T A T A T \

Exon

G85E

T A c T T \

de1556A

1 1 5 1 1

4 5

Normal Nonmut.

Mut.

68 (10) 41 (23) 54 (8) 31@) 33 (18) 54 (9) 61 (1)

Nonmut. 51 28 50 30 29 49 54

5

(43) (11) (44) (16) (13) (41) (4)

CF and normal chromosomes examined for the indicated number of chromosomes without the change; the number in group (see Fig. 2 and Ref. (10)).

be used for rapid screening of G85E-upon cleavage with this enzyme, the normal PCR-amplified DNA would generate three fragments of 172,105, and 32 bp in size, and the mutant would give a novel 277 bp (Table 1). Screening of 54 CF chromosomes (8 from group II) and 50 normal chromosomes (44 from group II) did not detect another example of G85E. 556delA

A frameshift mutation caused by the deletion of a single nucleotide (A) at position 556 of the coding region was found in exon 4 (Table 2). This mutation (556delA) was detected in a PI patient (Toronto fam-

C A T\

621+1G+T

71 l+lG+T

E\

G176R

i T GT A

G

A

T

C

G T A G T T T A A A

A’

FIG. 1. DNA sequence analyses of detected CF mutations (from left to right): All DNA samples were amplified by PCR, and sequences were determined with

G85E, de1556A, 621+ 1G + T, 711+ 1G + T, and G178R. primers described under Materials and Methods.

232

ZIELENSKI

ET

AL.

1

1

2

1 2

2

2

Ub)

1 1

1 1

2 2

1 2 1 2

2 2

2 2

(Ib) Ob)

1 1

1 1

2 2

1 2 1 -

2

2 -

Ub) Ub)

621+1&T 621+1G+T

2 2

1 2 1 2

1 1

2 2

1 1

1 1

1 1

1 1

1 1

2 2

2 2

1 2 1 2

1 1

2 2

621+lG+T 621+1G+T

2 2

2 1 1 1

1 1

2 2

1 1

1 1

-

1 1

1 1

2 2

2 2

1 1

2 2

1 1

2 2

621+1G+T

2

1 2

1

2

1

1

1

1

1

2

-

1

2

1

2

129G4!

21212112221---212111212

@a)

G85E

1111211-221121212112222

OW

711+1G+T

1

556delA

21212222-12212112221222

G178R

1

1 1

1 -

1

1

2

2

1

1

1

1

2

2

1

1 2

1 2

1

2

1

1

2

2

1

1

(IIb) (In@

1

2 (2) (2) (1)

-

-

-

-

-

-

1

-

-

-

1

1

(Va)

FIG. 2. DNA marker haplotypes associated with the CF chromosomes. The DNA markers (with the locus names underlined), revealing the polymorphism, haplotypes, and group classifications (in parentheses) have been defined previously (10).

ily 17, GM1076) heterozygous for this deletion; overlapping ladders were thus observed on the sequencing gel from the point of the deletion (Fig. 1). The 556delA deletion was thought to be disease-causing because it would result in a shift of translation reading frame leading to a premature termination of CFTR synthesis. The mutant chromosome was classified as a member of the IIIb haplotype group, and the paternal CF chromosome in this family was found to have the AF508 mutation (Kerem et al., 198913). Since 556delA created a novel BgZI site in exon 4, the PCR-amplified product could be digested with BgZI to distinguish the mutant sequence, which would be cleaved by this enzyme into two fragments (287 and 150 bp), from the normal (438 bp) (Table 1). The mutation was not found in 31 other CF chromosomes (9 from group IIIb) and 30 normal chromosomes (16 from IIIb) . 621 i- 1G --+ T A putative mutation was found in the splice donor site flanking the 3’ end of exon 4; the highly conserved GT at this location was changed to TT (Fig. 1). This mutation (621 + 1G + T), which was expected to result in aberrant splicing, was detected in five French-Canadian non-AF508 CF chromosomes belonging to haplotype group I (one each in Toronto families 22, 23,26,36, and 53) but not in 33 other CF chromosomes (18 from the same group) and 29 normal chromosomes (13 from the same group). The

enzyme

other CF chromosomes in all five families were also known: the ones for families 23 and 36 were AF508, that for family 53 was A455E (Kerem et aZ., 1990a), that for family 26 was G85E (as described above), and that for family 22 was 711+ 1G + T (see below). The patients in these families, except those in family 53, were all documented to be PI. Since a novel We1 site was created by this G to T transversion, the mutant allele could be distinguished from the non-621 + 1G + T sequence with this enzyme; whereas four fragments of 33,35,71, and 299 bp in size would be generated by MseI digestion of the normal PCR-amplified product, the 299-bp fragment in the mutant would be cleaved by the enzyme to give two fragments of 54 and 245 bp instead. G178R A missense mutation (G178R) was detected in exon 5, in a region predicted to be intracellular, and between the first and second membrane-spanning segments. The nucleotide at position 664 was changed from G to A (Fig. l), resulting in the replacement of Gly178 by Arg in the presumptive CFTR. The mutation was found to belong to haplotype group V, on the maternal CF chromosome in family 50. The patients in this family were classified as PI, and the paternal CF chromosome was AF508. G178R could be identified with AS0 hybridization, as described in Table 1. This G to A transition was thought to be a mutation causing CF because the en-

CF

233

MUTATIONS

coded amino acid residue switched from neutral to basic. No other sequence alterations were found in the coding region of this allele. In addition, the sequence variation was not detected in 54 normal chromosomes (including 5 in the same haplotype group); it was, in fact, found only once after screening of 61 CF chromosomes (including 2 others from the same group).

711 + 1G + T This G to T change was found to be located at the highly conserved consensus sequence at the splice junction (nucleotide position 711+ 1) at the 3’ flanking region of exon 5 (Fig. 1). As a consequence, it was expected that this change (GT + TT) would result in aberrant splicing. The 711 + 1G --* T mutation was found on the maternal CF chromosome of a FrenchCanadian family with PI (family 22) from Chicoutimi; this chromosome was included in haplotype group 1% The 711 + 1G --) T mutation could be detected by AS0 hybridization as described in Table 1. We examined 54 CF chromosomes (9 from haplotype group II) and 49 normal chromosomes (41 group II), but did not see another example.

129G + C While all of the presumptive CF mutations identified thus far were located in the coding region and RNA splice junctions (Kerem et al., 1989b, 1990a; Cutting et al., 1990a,b; Dean et al., 1990; White et al., 1990), one candidate mutation outside these regions was found at a nucleotide position 4 bp upstream of the presumptive initiation codon. The sequence alteration was a G to C transversion (129G + C). It was first found on the paternal CF chromosome in Toronto family 14 and associated with haplotype IIa. The CF children in this family were documented to be PS (pancreatic sufficient) and the maternal CF chromosome was AF508. The 129G + C mutation could be detected by AS0 hybridization as described in Table 1. Haplotype analysis indicated that haplotype IIa (the chromosome containing this mutation) was common among the normal chromosome population (Kerem et al., 1989b), but the G to C change was not found in 51 normal chromosomes screened (43 from group II). To obtain further evidence that this sequence alteration might associate with the disease, additional chromosomes were examined. Among 68 other CF chromosomes (including 10 from haplotype group II) analyzed, two additional examples were found in two unrelated PS patients. The other chromosomes in these patients were AF508 but, unfortunately, their haplotype was not determined.

R75Q An alternative nucleotide was noted at position 356 in exon 3, where a T nucleotide was found in place of the reported C (data not shown); the corresponding amino acid would be Gln instead of Arg at amino acid position 75. The T nucleotide was found in 5 of 16 normal chromosomes of haplotype IIa but not in 17 normal chromosomes of other haplotypes nor in 41 CF chromosomes (23 from IIa), suggesting that this mutation represents an amino acid sequence polymorphism. DISCUSSION

Upon sequence analysis of exons 1 through 8, we have identified five, possibly six, different mutations in the CFTR gene. Although one of them, 556delA (the single-basepair deletion), which is expected to create a translational frameshift, is clearly a defective allele, the other five require further justification that they are disease-causing mutations. The identification of this frameshift mutation, together with G85E, confirms that the previously predicted initiation codon is in fact utilized, as the next available methionine codon is at amino acid position 150 toward the end of exon 4. Two of these apparent mutations (G85E and G178R) would lead to drastic amino acid substitutions (change of type). Since Gly85 is predicted to be in the first transmembrane segment of CFTR, replacement with a negatively charged residue (Glu) would probably cause a severe consequence to the localization of the mutant protein in the membrane. The result of substituting Arg for Glyl78 is not as obvious, as there is no information about a possible function of this presumed intracellular region. However, since no other sequence alterations have been detected in the coding regions of these alleles and since no normal chromosomes have been found to harbor these missense mutations, G85E and G178R are most probably disease-causing mutations. Formal proof that these sequence alterations are bona fide mutations awaits functional testing of each of them in transfection assays. In this regard, the transfection systems recently developed (Drumm et al., 1990; Rich et al., 1990) should be useful for the study. Two of the sequence changes (621 + 1G + T and 711 + 1G + T) are expected to cause RNA splicing defects. The unprocessed or aberrantly processed RNA may be highly unstable and rapidly degraded, resulting in low or no CFTR synthesis. It is unclear, however, whether alternative splicing might occur at cryptic sites. Total omission of exon 4 or 5 is expected to create a shortened polypeptide, as both of these exons contain a number of nucleotides in multiples of

234

ZIELENSKI

3, but the resulting proteins might not be functional. Direct examination of the RNA or protein derived from these mutant alleles should provide a better understanding of these mutations. The sequence change (129G --f C) located 4 nucleotides upstream of the proposed translation initiation codon is perplexing; it is unclear whether it represents a disease-causing mutation, especially since a C residue at position -4 of the initiation codon is preferred in experimental systems (Kozak, 1986). The mutation has, however, so far been found on 3 CF chromosomes but not on any of the 51 normal chromosomes examined, including 43 from the same haplotype group (IIa). It is possible that there is a mutation elsewhere in this allele, but no other change has been found after sequencing of the entire coding region and immediately flanking intron regions. Alternatively, 129G + C may reduce expression of CFTR at the transcriptional level. Further analysis at the expression level may prove this point. On the basis of DNA marker haplotype analysis, we previously showed that all CF chromosomes carrying AF508 appear to have descended from a single mutational event (Kerem et al., 198913). A similar conclusion may be drawn for the 621 + 1G + T mutation described in this study; all the CF chromosomes carrying this mutation belong to haplotype group Ib (Fig. 2). In fact, the CF patients with 621 + 1G + T all come from the Saguenay-Lac St. Jean region in Quebec, confirming the common origin of this mutation. It is probable that the mutation was introduced into this historical settlement by a small number of founders (Bouchard, 1987). In this regard, it is of interest to note that another mutation, A455E, found in this French-Canadian population is also associated with haplotype Ib (Kerem et aZ., 1990a). Since group I haplotype is rare among normal chromosomes, it is likely that 621+ 1G --* T, A455E, and perhaps AF508 (Kerem et aZ., 1989a) all originated in a unique geographic location. Subsequent selection or drift could increase the frequencies of these mutations in the population. The two other mutations (G85E and 711 + 1G + T) found in the French-Canadian population, although less frequently detected, also share a similar haplotype (group IIb, Fig. 2). This similarity may reflect their common ancestral relationship, but it is more likely to be a coincidence, as group II is the most common haplotype among normal chromosomes (Kerem et aZ., 1989b). We previously suggested that CF mutations could be classified as seuere and mild alleles with respect to their influence on pancreatic function (Kerem et al., 1989a,b). The AI?508 deletion has been included as one of the severe alleles, and patients homozygous for this mutation are almost exclusively PI (Kerem et al., 1989b, 199Ob). The hypothesis also predicts that pa-

ET

AL.

tients homozygous for any combination of the severe alleles will be PI. Under this assumption, a number of other mutations could be classified as severe alleles for their association with PI patients (Kerem et al., 1990a). Similarly, five of the mutations reported here, i.e., G85E, G178R, 556delA, 621 + 1G + T, and 711 + 1G --* T, could also be classified as severe alleles. Definition of a mild allele is less straightforward because the patient could carry one or two copies of a mild mutation(s). Two mutations (A455E and P574H) that are most likely mild alleles have been described; both of them are found in patients carrying AF508 (severe allele) as their other CF chromosome (Kerem et al, 1990a). In the present study, one of these mild alleles, A455E, is found in combination with 621 + 1G + T (a severe allele as defined above) in a PS patient, thus in complete agreement with our hypothesis. Further, the CFTR gene with the 129G --t C mutation probably constitutes a mild allele, because all three patients found to have this mutation are PS and carry AF508 on their other CF chromosome. In addition to furthering our understanding of the functional properties of CFTR and the clinical consequence of the different types of mutant alleles, the knowledge of CF mutations at the DNA level increases our ability to detect these mutations in the population. A cohort of 94 CF chromosomes that have been well characterized with respect to DNA marker haplotypes (Kerem et al., 1989b) has been used in our primary mutation detection study. The majority (631 94) of these chromosomes carry the AF508 mutation, at a frequency similar to the world average (68%). After screening of over 90% of the coding region of CFTR, 15 additional mutations have been detected in this cohort (Kerem et al., 1990a; this study). Together, these mutations account for over 90% (85/94) of CF chromosomes in this population. Knowledge of the distribution of these mutations in other populations awaits further data collection through the CFTR Genetic Analysis Consortium. It is anticipated that there will be substantial variation in the frequency of each of the mutant alleles in different geographic locations.

ACKNOWLEDGMENTS The authors thank Dara Kennedy for her expert technical assistance, and Mary Fujiwara and Richard Rozmahel for helpful discussions. The availability of the CF patient database at the Hospital for Sick Children in Toronto is also gratefully acknowledged. The research was supported by grants from the National Institutes of Health (U.S.A.) (DK-34944-53, the Cystic Fibrosis Foundation (U.S.A.), and the Canadian Cystic Fibrosis Foundation. L.-C.T. is a recipient of the Scientist Award from the Medical Research Council of Canada.

235

CF MUTATIONS

REFERENCES 1. BOUCHARD, G. (1987). The population history and the gene pool of the north-eastern populations of the province of Quebec. In “Proceedings, SOREP International Symposium, Chicoutimi,” pp. 11-17. 2. CUTTING, G. R., KASH, L. M., ROSENSTEIN, B. J., ZIELENSKI, J., TSUI, L.-C., ANTONARAKIS, S. E., AND KAZAZIAN, H. H. (1990a). A cluster of cystic fibrosis mutations in the first nucleotide-binding fold of the cystic fibrosis conductance regulator protein. Nature 346: 366-369. 3. CUTTING, G. R., KASH, L. M., ROSENSTEIN, B. J., TSUI, L.-C., KAZAZIAN, H. H., JR., AND ANTONAFZAKIS, S. E. (1990b). Two cystic fibrosis patients with mild pulmonary disease and nonsense mutations in each CFTR gene. New Engl. J. Med. 323: 1685-1689. 4. DEAN, M., WHITE, M. B., AMOS, J., GERRARD, B., STEWART, C., KHAW, K.-T., AND LEPPERT, M. (1990). Multiple mutations in highly conserved residues are found in mildly affected cystic fibrosis patients. Cell 61: 863-870. 5. DRUMM, M. L., POPE, H., CLIFF, W. H., ROMMENS, J. M., MARVIN, S. A., TSUI, L.-C., COLLINS, F. S., FRIZZELL, R. A., AND WILSON, J. M. (1990). Correction of the cystic fibrosis defect in vitro by retrovirus-mediated gene transfer. Cell 62: 1227-1233. 6. DURIE, P. R., FORSTNER, G. G., GASKIN, K. J., MOORE, D. J., CLEGHORN, G. J., WONG, S. S., AND COREY, M. L. (1986). Age-related alterations in immunoreactive pancreatic cationic trypsinogen in sera from cystic fibrosis patients with and without pancreatic insufficiency. Pediatr. Res. 20: 209213. 7. GASKIN, K. J., DURIE, P. R., LEE, L., HILL, R., AND FORSTNER, G. G. (1984). Colipase and lipase secretion in childhood-onset pancreatic insufficiency: Delineation of patients with steatorrhoea secondary to relative colipase deficiency. Gastroenterology 86: l-7. 8. GUILLERMIT, H., FANEM, P., AND FEREC, C. (1990). A 3’ splice site consensus sequence mutation in the cystic fibrosis gene. Hum.

Genet.

85: 450-453.

9. KEREM, B., BUCHANAN, J. A., DURIE, P., COREY, M. L., LEVISON, H., ROMMENS, J. M., BUCHWALD, M., AND TSUI, L.-C. (1989a). DNA marker haplotype association with pancreatic sufficiency in cystic fibrosis. Am. J. Hum. Genet. 44: 827-834. 10. KEREM, B., ROMMENS, J. M., BUCHANAN, J. A., MARKIEWICZ, D., Cox, T. K., CHAKRAVARTI, A., BUCHWALD, M., AND TSUI, L.-C. (1989b). Identification of the cystic fibrosis gene: Genetic analysis. Science 245: 1073-1080. 11. KEREM, B., ZIELENSKI, J., MARKIEWICZ, D., BOZON, D., GAZIT, E., YAHAV, J., KENNEDY, D., RIORDAN, J., COLLINS, S. F., ROMMENS, J., AND TSUI, L.-C. (1990a). Identification of mutations in regions corresponding to the 2 putative nucleotide (ATP)-binding folds of the cystic fibrosis gene. Proc. N&Z. Acad. Sci. USA 87: 8447-8451. 12. KEREM, E., COREY, M., KEREM, B., ROMMENS, J., MARKIEWICZ, D., LEVISON, H., TSUI, L.-C., AND DURIE, P. (199Ob). Association between the most common mutation (AF508) and phenotypes in cystic fibrosis. New Engl. J. Med. 323: 1517-1522. 13. KOZAK, M. (1986). Point mutations define a sequence flank-

14.

15.

16.

17.

18.

ing the AUG initiator codon that modulates translation by eukaryotic ribosomes. Cell 44: 283-292. LEMNA, W. K., FELDMAN, G. L., KER.EM, B., FERNBACH, S. D., ZEVKOVICH, E. P., O’BRIEN, W. E., COLLINS, F. S., TSUI, L.-C., AND BEAUDET, A. L. (1990). Mutation analysis for heterozygote detection and the prenatal diagnosis of cystic fibrosis. New Engl. J. Med. 322: 291-296. MIZUSAWA, S., NISHIMURA, S., AND SEXLA, F. (1986). Improvement of the dideoxy chain termination method of DNA sequencing by use of deoxy-7-deazaguanosine triphosphate in place of dGTP. Nucleic Acids Res. 14: 1319-1324. RICH, D. P., ANDERSON, M. P., GREGORY, R. J., CHENG, S. H., PAUL, S., JEFFERSON, D. M., MCCANN, J. D., KLINGER, K. W., SMITH, A. E., AND WELSH, M. J. (1990). Expression of cystic fibrosis transmembrane conductance regulator corrects defective chloride channel regulation in cystic fibrosis airway epithelial cells. Nature 347: 25&265. RIORDAN, J. R., ROMMENS, J. M., KEREM, B., ALON, N., RozMAHEL, R., GRZELCZAK, Z., ZIELENSKI, J., LOK, S., PLAVSIC, N., CHOU, J-L., DRUMM, M., IANNUZZI, M. C., COLLINS, S. F., AND TSUI, L-C. (1989). Identification of the cystic fibrosis gene: Cloning and characterization of complementary DNA. Science 246: 1066-1073. ROMMENS, J. M., KEREM, B., GREER, W., CHANG, P., TSUI, L.-C., AND RAY, P. (1990). Rapid nonradioactive detection of the major cystic fibrosis mutation. Am. J. Hum. Genet. 45: 337-339.

19. ROMMENS, J. M., IANNUZZI, M. C., KEF~EM, B., DRUMM, M. L., MELMER, G., DEAN, M., ROZMAHEL, R., COLE, J., KENNEDY, D., HIDAKA, N., ZSIGA, M., BUCHWALD, M., RIORDAN, J. R., TSUI, L-C., AND COLLINS, F. S. (1989). Identification of the cystic fibrosis gene: Chromosome walking and jumping. Science 245: 1059-1065. 20. SAIKI, R. K., GELFAND, D. H., STOFFEL, S., SHARF, S. J., HIGUCHI, R., HORN, G. T., AND ERLICH, H. A. (1988). Primerdirected enzymatic amplification of DNA with a thermostable DNA polymerase. Science 239: 487-491. 21. SANGER, F., NICKLEN, S., AND COULSON, A. R. (1977). DNA sequencing with chain terminating inhibitors. Proc. Natl. Acad.

Sci. USA

74: 5463-5467.

22. TSUI, L.-C., ZENGERLING, S., WILLARD, H. F., AND BUCHWALD, M. (1986). Mapping of the cystic fibrosis locus on chromosome 7. Cold Spring Harbor Symp. Quant. Biol. 51: 325335. 23. VIDAUD, M., FANEN, P., MARTIN, J., GHANEM, N., NICOLAS, S., AND GOOSSENS, M. (1990). Three mutations in the CFTR gene in French cystic fibrosis patients: Identification by denaturing gradient gel electrophoresis. Hum. Genet. 85: 434-435. 24. WHITE, M., AMOS, J., Hsu, J. M.-C., GERRARD, B., FINN, P., AND DEAN, M. (1990). A frameshift mutation in the cystic fibrosis gene. Nature 344: 665-667. 25. WINSHIP, P. R. (1989). An improved method for directly sequencing PCR amplified material using dimethyl sulphoxide. Nucleic Acids Res. 17: 1266. 26. ZIELENSKI, J., ROZMAHEL, R., BORON, D., KEREM, B., GRZELCZAK, Z., RIORDAN, J. R., ROMMENS, J. M., AND TSUI, L.-C. (1991). Genomic DNA sequence of the cystic fibrosis transmembrane conductance regulator (CFTR) gene. Genomics 10: 214-228.