Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
2~12~3
. .
INTRONS AND EXONS OF THE CYSTIC FIBROSIS GENE
AND MUTATIONS AT VARIOUS POSITIONS OF THE GENE
- FIELD OF THE INVENTION
The present invention relates generally to the
cystic fibrosis (CF) gene, and, more particularly to the
identification, isolation and cloning of the DNA
sequence corresponding to the normal and a mutant of the
CF gene, as well as their transcripts, gene products and
genetic information at exon/intron boundaries. The
; lO present invention also relates to methods of screening
for and detection of CF carriers, CF diagnosis, prenatal
CF screening and diagnosis, and gene therapy utilizing
recombinant technologies and drug therapy using the
information derived from the DNA, protein, and the
metabolic function of the protein.
BACKGROUND OF THE INVENTION
Cystic fibrosis (CF) is the most common severe
autosomal recessive genetic disorder in the Caucasian
population. It affects approximately 1 in 2000 live -~
births in North America [Boat et al, The Metabolic
''! Basis of Inherited Disease, 6th ed, pp 2649-2680, McGraw
i Hill, NY (1989)]. Approximately 1 in 20 persons are
carriers of the disease.
Although the disease was first described in the
25 late 1930's, the basic defect remains unknown. The -
major symptoms of cystic fibrosis include chronic
pulmonary disease, pancreatic exocrine insufficiency,
and elevated sweat electrolyte levels. The symptoms are
consistent with cystic fibrosis being an exocrine
disorder. Although recent advances have been made in
the analysis of ion transport across the apical membrane
of the epithelium of CF patient cells, it is not clear
that the abnormal regulation of chloride channels
represents the primary defect in the disease. Given the
- 35 lack of understanding of the molecular mechanism of the
disease, an alternative approach has therefore been
'' :
~' ; ~.'
20~2~3
taken in an attempt to understand the nature of the
molecular defect through direct cloning of the
responsible gene on the basis of its chromosomal
location.
However, there is no clear phenotype that directs
an approach to the exact nature of the genetic basis of
the disease, or that allows for an identification of the ;~
cystic fibrosis gene. The nature of the CF defect in
relation to the population genetics data has not been
10 readily apparent. Both the prevalence of the disease -~
and the clinical heterogeneity have been explained by ~
several different mechanisms: high mutation rate, ~ -
heterozygote advantage, genetic drift, multiple loci, ;
and reproductive compensation. ~ -
Many of the hypotheses can not be tested due to the
lack of knowledge of the basic defect. Therefore,
alternative approaches to the determination and
' characterization of the CF gene have focussed on an
attempt to identify the location of the gene by genetic
analysis.
Linkage analysis of the CF gene to antigenic and
protein markers was attempted in the 1950's, but no
positive results were obtained [Steinberg et al Am. J.
Hum. Genet. 8: 162-176, (1956); Steinberg and Morton Am. ~; -
` 25 J. Hum. Genet 8: 177-189, (1956); Goodchild et al J.
Med. Genet. 7: 417-419, 1976.
More recently, it has become possible to use
RFLP's to facilitate linkage analysis. The first
linkage of an RFLP marker to the CF gene was disclosed -~
30 in 1985 [Tsui et al. Science 230: 1054-1057, 1985] in
which linkage was found between the CF gene and an
- uncharacterized marker DOCRI-917. The association was
found in an analysis of 39 families with affected CF
children. This showed that although the chromosomal
35 location had not been established, the location of the ~-
~.
: :
., '
.;,f`,- ~ ' ' ' " . ' .
~ 2~112~3
`
disease gene had been narrowed to about 1% of the human
genome, or about 30 million nucleotide base pairs.
The chromosomal location of the D0CRI-sl7 probe was
established using rodent-human hybrid cell lines
containing different human chromosome complements. It
was shown that DOCR1-917 (and therefore the CF gene)
j maps to human chromosome 7.
- Further physical and genetic linkage studies were
} pursued in an attempt to pinpoint the location of the CF
gene. Zengerling et al [Am. J. Hum. Genet. 40: 228-236
(1987)] describe the use of human-mouse somatic cell
hybrids to obtain a more detailed physical relationship
between the CF gene and the markers known to be linked
with it. This publication shows that the CF gene can be
assigned to either the distal region of band q22 or the
proximal region of band q31 on chromosome 7.
Rommens et al [Am. J. Hum. Genet. 43: 645-663,
(1988)] give a detailed discussion of the isolation of
,., :
; many new 7q31 probes. The approach outlined led to the
20 isolation of two new probes, D7S122 and D7S340, which
are close to each other. Pulsed field gel
electrophoresis mapping indicates that these two RFLP -~
' markers are between two markers known to flank the CF
gene, MET [White, R., Woodward S., Leppert M., et al.
25 Nature 318: 382-384, (1985)] and D7S8 [Wainwright, B.
J., Scambler, P. J., and J. Schmidtke, Nature 318: 384-
385 (1985)], therefore in the CF gene region. The
discovery of these markers provides a starting point for
chromosome walking and jumping.
Estivill et al, [Nature 326: 840-845(1987)]
disclose that a candidate cDNA gene was located and
partially characterized. This however, does not teach
the correct location of the CF gene. The reference
' discloses a candidate cDNA gene downstream of a CpG
island, which are undermethylated GC nucleotide-rich
regions upstream of many vertebrate genes. The
.
:~
. ~
.
- 20~2~
chromosomal localization of the candidate locus is
identified as the XV2C region. This region is described
in European Patent Application 88303645.1. However,
that actual region does not include the CF gene.
A major difficulty in identifying the CF gene has
been the lack of cytologically detectable chromosome
rearrangements or deletions, which greatly facilitated
all previous successes in the cloning of human disease
genes by knowledge of map position.
lo Such rearrangements and deletions could be observed
cytologically and as a result, a physical location on a
particular chromosome could be correlated with the
particular disease. Further, this cytological location -
could be correlated with a molecular location based on
known relationship between publicly available DNA probes
and cytologically visible alterations in the
chromosomes. Knowledge of the molecular location of the
gene for a particular disease would allow cloning and
sequencing of that gene by routine procedures,
particularly when the gene product is known and cloning
success can be confirmed by immunoassay of expression
products of the cloned genes.
In contrast, neither the cytological location nor
the gene product of the gene for cystic fibrosis was
known in the prior art. With the recent identification
of MET and D7S8, markers which flanked the CF gene but
did not pinpoint its molecular location, the present
inventors devised various novel gene cloning strategies
to approach the CF gene in accordance with the present
invention. The methods employed in these strategies
include chromosome jumping from the flanking markers,
cloning of DNA fragments from a defined physical region
with the use of pulsed field gel electrophoresis, a ;
combination of somatic cell hybrid and molecular cloning
35 techniques designed to isolate DNA fragments from -~
undermethylated CpG islands near CF, chromosome
.` :
- 20~ ~2~
microdissection and cloning, and saturation cloning of a
large number of DNA markers from the 7q31 region. By
means of these novel strategies, the present inventors
were able to identify the gene responsible for cystic
fibrosis where the prior art was uncertain or, even in
one case, wrong.
The application of these genetic and molecular
cloning strategies has allowed the isolation and cDNA
cloning of the cystic fibrosis gene on the basis of its ~;
chromosomal location, without the benefit of genomic
rearrangements to point the way. The identification of
the normal and mutant forms of the CF gene and gene
products has allowed for the development of screening
and diagnostic tests for CF utilizing nucleic acid
probes and antibodies to the gene product. Through
interaction with the defective gene product and the
pathway in which this gene product is involved, therapy
through normal gene product supplementation and gene
manipulation and delivery are now made possible.
The gene involved in the cystic fibrosis disease
process, hereinafter the "CF gene" and its functional
equivalents, has been identified, isolated and cDNA
cloned, and its transcripts and gene products identified
and sequenced. A three base pair deletion leading to
i 25 the omission of a phenylalanine residue in the gene
product has been determined to correspond to the
mutations of the CF gene in approximately 70% of the
patients affected with CF, with different mutations
involved in most if not all the remaining cases. This ~-
subject matter is disclosed in co-pending United States
patent application S.N. 396,894 filed August 22, 1989
and its related continuation-in-part applications S.N.
399,945 filed August 24, 1989 and S.N. 401,609 filed
August 31, 1989.
'4
~'
~,
':j
:-~ 2~2S3
SUMMARY OF THE INVENTION
According to this invention, other base pair
deletions leading to the omission of amino acid residues
in the gene product have been determined.
With the identification and sequencing of the
mutant gene and its gene product, nucleic acid probes
and antibodies raised to the mutant gene product can be
used in a variety of hybridization and immunological
assays to screen for and detect the presence of either
the defective CF gene or gene product. Assay kits for
such screening and diagnosis can also be provided. The
genetic information derived from the intron/exon
boundaries is also very useful in various screening and
diagnosis procedures.
Patient therapy through supplementation with the
normal gene product, whose production can be amplified
using genetic and recombinant techniques, or its
~ functional equivalent, is now also possible. Correction
- or modification of the defective gene product through
; 20 drug treatment means is now possible. In addition,
cystic fibrosis can be cured or controlled through gene
therapy by correcting the gene defect ln situ or using
recombinant or other vehicles to deliver a DNA sequence
capable of expression of the normal gene product to the
cells of the patient.
According to another aspect of the invention, a
purified mutant CF gene comprises a DNA sequence
encoding an amino acid sequence for a protein where the
protein, when expressed in cells of the human body, is
associated with altered cell function which correlates
with the genetic disease cystic fibrosis.
According to another aspect of the invention, a
i~ purified RNA molecule comprises an RNA sequence i~
corresponding to the above DNA sequence.
According to another aspect of the invention, a DNA
~;
.'
.
~ .
20112~3
molecule comprises a cDNA molecule corresponding to the
above DNA sequence.
According to another aspect of the invention, a ~-
DNA molecule comprises a DNA sequence encoding mutant
CFTR polypeptide having the sequence according to the
following Figure 1 for amino acid residue positions 1 to
1480. The sequence is further characterized by a base
pair mutation which results in the deletion of an amino
acid at an amino acid residue position selected from the
group consisting of 455, 507, 542, 549 and 560.
According to another aspect of the invention, a DNA
molecule comprises a cDNA molecule corresponding to the
above DNA sequence.
According to another aspect of the invention, the
cDNA molecule comprises a DNA sequence selected from the
group consisting of:
(a) DNA sequences which correspond to the mutant
DNA sequence selected from the group of mutant positions
of 455, 507, 542, 549 and 560 and which encode, on
expression, for mutant CFTR polypeptide;
(b) DNA sequences which correspond to a fragment
of the selected mutant DNA sequence, including at least
twenty nucleotides;
' (c) DNA sequences which comprise at least twenty
nucleotides and encode a fragment of the selected mutant
CFTR protein amino acid sequence;
(d) DNA sequences encoding an epitope encoded by
at least eighteen sequential nucleotides in the selected
mutant DNA sequence.
According to another aspect of the invention, a DNA
sequence selected from the group consisting of:
, (a) DNA sequences which correspond to portions of
`~ DNA sequences of boundaries of exons/introns of the
genomic CF gene;
(b) DNA sequences of at least eighteen sequential
~'. ,
'~'
.~t, . ~
~ - 2~2~3
- nucleotides at boundaries of exons/introns of the
genomic CF gene depicted in Figure 18; and
(c) DNA sequences of at least eighteen sequential
nucleotides of intron portions of the genomic CF gene of
Figure 18.
According to another aspect of the invention, a
purified nucleic acid probe comprises a DNA or RNA
nucleotide sequence corresponding to the above noted
selected DNA sequences of groups (a) to (c).
According to another aspect of the invention,
purified RNA molecule comprising RNA sequence corres-
ponds to the mutant DNA sequence selected from the group
of mutant positions consisting of 455, 507, 542, 549 and
560.
~ 15 A purified nucleic acid probe comprising a DNA or
- RNA nucleotide sequence corresponding to the mutant
sequences of the above recited group.
According to another aspect of the invention, a
recombinant cloning vector comprising the DNA sequences
' 20 of the mutant DNA and fragments thereof selected from
the group of mutant positions consisting of 455,
` 507,542, 549 and 560 is provided. The vector, according
to an aspect of this invention, is operatively linked to
an expression control sequence in the recombinant DNA
-~ 25 molecule so that the selected mutant DNA sequences for
,! the mutant CFTR polypeptide can be expressed. The ~-~
expression control sequence is selected from the group
consisting of sequences that control the expression of
genes of prokaryotic or eukaryotic cells and their
viruses and combinations thereof.
According to another aspect of the invention, a
method for producing a mutant CFTR polypeptide
comprises the steps of:
(a) culturing a host cell transfected with the
~;~ 35 recombinant vector for the mutant DNA sequence in a
~- medium and under conditions favorable for expression of
:
20112~3
g
the mutant CFTR polypeptide selected from the group of
:~ mutant CFTR polypeptides at mutant positions 455, 507,
542, 549 and 560; and
(b) isolating the expressed mutant CFTR
polypeptide.
According to another aspect of the invention, a
purified protein of human cell membrane origin comprises
an amino acid sequence encoded by the mutant DNA
sequences selected from the group of mutant positions of
10 455, 507, 542, 549 and 560 where the protein, when
present in human cell membrane, is associated with cell
function which causes the genetic disease cystic
. fibrosis.
According to another aspect of the invention, a
method is provided for screening a subject to determine
; if the subject is a CF carrier or a CF patient
~ comprising the steps of providing a biological sample of
the subject to be screened and providing an assay for
detecting in the biological sample, the presence of at
least a member from the group consisting of:
~ (a) mutant CF gene selected from the group of
? mutant positions 455, 507, 542, 549 and 560;
- (b) mutant CF gene products and mixtures thereof;. (c) DNA sequences which correspond to portions of
DNA sequences of boundaries of exons/introns of the
. genomic CF gene;
(d) DNA sequences of at least eighteen sequential
nucleotides at boundaries of exons/introns of the
; genomic CF gene depicted in Figure 18; and
. 30 (e) DNA sequences of at least eighteen sequential
. nucleotides of intron portions of the genomic CF gene of.i .
Figure 18.
According to another aspect of the invention, a kit
for assaying for the presence of a CF gene by
immunoassay techniques comprises:
.,
~.1
. .
,, .
:,
., ., - ~ ,
~ .~
-` 20112~3
(a) an antibody which specifically binds to a gene
product of the mutant DNA sequence selected from the
group of mutant positions 455, 507, 542, 549 and 560;
(b) reagent means for detecting the binding of the
antibody to the gene product; and
(c) the antibody and reagent means each being
present in amounts effective to perform the immunoassay.
According to another aspect of the invention, a kit
for assaying for the presence of a mutant CF gene by
hybridization technique comprises:
(a) an oligonucleotide probe which specifically
binds to the mutant CF gene having a mutation at a
position selected from the group consisting of 455, 507,
542, 549 and 560;
(b) reagent means for detecting the hybridization
of the oligonucleotide probe to the mutant CF gene; and
(c) the probe and reagent means each being present
in amounts effective to perform the hybridization assay.
1 According to another aspect of the invention, an
; 20 animal comprises an heterologous cell system. The cell
system includes a recombinant cloning vector which
includes the recombinant DNA sequence corresponding to
`11 the mutant DNA sequence which induces cystic fibrosis
symptoms in the animal.
According to another aspect of the invention, in a
` polymerase chain reaction to amplify a selected exon of
O a cDNA sequence of Figure 1, the use of oligonucleotide
primers from intron portions near the 5' and 3'
boundaries of the selected exon of Figure 18.
j~ 30 BRIEF DESCRIPTION OF THE DRAWINGS
^~ Figure 1 is the nucleotide sequence of the CF gene
3 and the amino acid sequence of the CFTR protein amino
acid sequence with ~ indicating mutations at the 507 and ^~
508 protein positions.
Figure 2 is a restriction map of the CF gene and
.
~ 2011253
11
the schematic strategy used to chromosome walk and jump
to the gene.
-~ Figure 3 is a pulsed-field-gel electrophoresis map
of the region including and surrounding the CF gene.
Figures 4A, 4B and 4C show the detection of
conserved nucleotide sequences by cross-species
hybridization.
Figure 4D is a restriction map of overlapping
segments of probes E4.3 and Hl.6.
~- 10 Figure 5 is an RNA blot hybridization analysis,
using genomic and cDNA probes. Hybridization to
fibroblast, trachea (normal and CF), pancreas, liver,
;~ HL60, T84, and brain RNA is shown.
;~
Figure 6 is the methylation status of the E4.3
cloned region at the 5' end of the CF gene.
~, Figure 7 is a restriction map of the CFTR cDNA
~, showing alignment of the cDNA to the genomic DNA
t 't fragments.
Figure 8 is an RNA gel blot analysis depicting
hybridization by a portion of the CFTR cDNA (clone 10-1)
to a 6.5 kb mRNA transcript in various human tissues.
Figure 9 is a DNA blot hybridization analysis
depicting hybridization by the CFTR cDNA clones to
genomic DNA digested with EcoRI and Hind III.
~; 25 Figure 10 is a primer extension experiment
~.
characterizing the 5' and 3' ends of the CFTR cDNA.
Figure 11 is a hydropathy profile and shows
predicted secondary structures of CFTR.
;;; Figure 12 is a dot matrix analysis of internal
-;30 homologies in the predicted CFTR polypeptide.
Figure 13 is a schematic model of the predicted
~ CFTR protein.
sFigure 14 is a schematic diagram of the restriction
fragment length polymorphisms (RFLP's) closely linked
~?35 to the CF gene where the inverted triangle indicates the
location of the F508 3 base pair deletion. ~ -
'.:
, . I .
~.
-` 20112~3
12
Figure 15 represents alignment of the most
conserved segments of the extended NBFs of CFTR with
comparable regions of other proteins.
Figure 16 is the DNA sequence around the F508
deletion.
Figure 17 is a representation of the nucleotide
sequencing gel showing the DNA sequence at the F508
deletion.
Figure 18 is the nucleotide sequence of the
portions of introns and complete exons of the genomic CF
~ gene for exons 4 and 6 to 24 of cDNA sequence of Figure
! Figure 19 shows the results of amplification of
genomic DNA using intron oligonucleotides bounding exon
10;
Figure 20 shows the separation by gel
electrophoresis of the amplified genomic DNA products of
a CF family.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
1. DEFINITIONS
~3 In order to facilitate-review of the various
~j embodiments of the invention and an understanding of -
various elements and constituents used in making the
JI invention and using same, the following definition of
, 25 terms used in the invention description is as follows~
3 CF - cystic fibrosis
CF carrier - a person in apparent health whose
i chromosomes contain a mutant CF gene that may be
transmitted to that person's offspring.
CF patient - a person who carries a mutant CF gene
, on each chromosome, such that they exhibit the clinical
, symptoms of cystic fibrosis.
¦ CF gene - the gene whose mutant forms are ~-
associated with the disease cystic fibrosis. This
35 definition is understood to include the various sequence ~-~
polymorphisms that exist, wherein nucleotide
`'' '~ .' ~
'' : ".
-:
.,
~ -` 20 l~2~3
13
substitutions in the gene sequence do not affect the
essential function of the gene product. This term
primarily relates to an isolated coding sequence, but
can also include some or all of the flanking regulatory
elements and/or introns.
Genomic CF gene - the CF gene which includes
flanking regulatory elements and/or introns at
boundaries of exons of the CF gene.
CF - PI - cystic fibrosis pancreatic insufficient,
the major clinical subgroup of cystic fibrosis patients,
characterized by insufficient pancreatic exocrine
function.
CF - PS - cystic fibrosis pancreatic sufficient, a
clinical subgroup of cystic fibrosis patients with
sufficient pancreatic exocrine function for normal
digestion of food.
CFTR - cystic fibrosis transmembrane conductance
regulator protein, encoded by the CF gene. This
definition includes the protein as isolated from human
or animal sources, as produced by recombinant organisms,
and as chemically or enzymatically synthesized. This
definition is understood to include the various
polymorphic forms of the protein wherein amino acid
substitutions in the variable regions of the sequence
does not affect the essential functioning of the
protein, or its hydropathic profile or secondary or
~' tertiary structure.
DNA - standard nomenclature is used to identify the
bases.
Intronless DNA - a piece of DNA lacking internal
non-coding segments, for example, cDNA.
IRP locus sequence - (protooncogene int-l related),
a gene located near the CF gene.
Mutant CFTR - a protein that is highly analagous to
CFTR in terms of primary, secondary, and tertiary
; structure, but wherein a small number of amino acid
. .
.~ .~ .. . . . .
. - . . ~ . . . - . .
2 0 ~ 3
.
14
substitutions and/or deletions and/or insertions result
in impairment of its essential function, so that
organisms whose epithelial cells express mutant CFTR
rather than CFTR demonstrate the symptoms of cystic
- 5 fibrosis.
mCF - a mouse gene orthologous to the human CF gene
NBFs - nucleotide (ATP) binding folds
ORF - open reading frame
PCR - polymerase chain reaction
Protein - standard single letter nomenclature is
~ used to identify the amino acids
;~ R-domain - a highly charged cytoplasmic domain of
i the CFTR protein
RSV - Rous Sarcoma Virus
SAP - surfactant protein
RFLP - restriction fragment length polymorphism
~l 507 mutant CF gene - the CF gene which includes a
DNA base pair mutation at the 506 or 507 protein
position of the cDNA of the CF gene
507 mutant DNA sequence - equivalent meaning to the
507 mutant CF gene
507 mutant CFTR protein or mutant CFTR protein
amino acid sequence, or mutant CFTR polypeptide - the ~;
mutant CFTR protein wherein an amino acid deletion
occurs at the isoleucine 506 or 507 protein position of
the CFTR.
~ 2. ISOLATING THE CF GENE
;l Using chromosome walking, jumping, and cDNA
hybridization, DNA sequences encompassing > 500 kilobase
` 30 pairs (kb) have been isolated from a region on the long
arm of human chromosome 7 containing the cystic
fibrosis (CF) gene. This technique is disclosed in ~
detail in the aforemention co-pending United States -
patent applications. For purposes of convenience in
understanding and isolating the CF gene and identifying
other mutations, such as at the 455, 507, 542, 549 and
.,
.
:', '~
., :
20~ 1253
560 amino acid residue positions, the technique is
reiterated here. Several transcribed sequences and
conserved segments have been identified in this region.
One of these corresponds to the CF gene and spans
approximately 250 kb of genomic DNA. Overlapping
complementary DNA (cDNA) clones have been isolated from
epithelial cell libraries with a genomic DNA segment
containing a portion of the cystic fibrosis gene. The
nucleotide sequence of the isolated cDNA is shown in
Figure 1. In each row of the respective sequences the
lower row is a list by standard nomenclature of the
nucleotide sequence. The upper row in each respective
row of sequences is standard single letter nomenclature
for the amino acid corresponding to the respective
codon.
Accordingly, the isolation of the CF gene provided
a cDNA molecule comprising a DNA sequence selected from
the group consisting of:
(a) DNA sequences which correspond to the DNA
sequence of Figure 1 from amino acid residue position 1
to position 1480;
(b) DNA sequences encoding normal CFTR polypeptide
having the sequence according to Figure 1 for amino acid
residue positions from 1 to 1480;
(c) DNA sequences which correspond to a fragment
of the sequence of Figure 1 including at least 16
sequential nucleotides between amino acid residue
positions 1 and 1480;
(d) DNA sequences which comprise at least 16
nucleotides and encode a fragement of the amino acid
sequence of Figure 1; and
(e) DNA sequences encoding an epitope encoded by
at least 18 sequential nucleotides in the sequence of
Figure 1 between amino acid residue positions l and
1480.
~ 20112~
16
According to this invention, the isolation of
another mutation in the CF gene also provides a cDNA
molecule comprising a DNA sequence selected from the
group consisting of:
a) DNA sequences which correspond to the DNA
sequence encoding mutant CFTR polypeptide characterized
by cystic fibrosis-associated activity in human
epithelial cells, or the DNA sequence of Figure 1 for
the amino acid residue positions 1 to 1480 yet further
characterized by a base pair mutation which results in
the deletion of an amino acid from residue positions
455, 507, 542, 549 and 560;
b) DNA sequences which correspond to fragments of ~ -
the mutant portion of the sequence of paragraph a) and ~ -
; 15 which include at least sixteen nucleotides;
c) DNA sequences which comprise at least sixteen
nucleotides and encode a fragment of the amino acid
sequence encoded for by the mutant portion of the DNA -~
sequence of paragraph a); and
d) DNA sequences encoding an epitope encoded by
at least 18 sequential nucleotides in the mutant portion
of the sequence of the DNA of paragraph a).
I Transcripts of approximately 6,500 nucleotides in
- size are detectable in tissues affected in patients with
CF. Based upon the isolated nucleotide sequence, the
- predicted protein consists of two similar regions, each
containing a first domain having properties consistent
with membrane association and a second domain believed
to be involved in ATP binding.
A 3 bp deletion which results in the omission of a
phenylalanine residue at the center of the first
predicted nucleotide binding domain (amino acid position
508 of the CF gene product) was detected in CF
patients. This mutation in the normal DNA sequence of
~' 35 Figure 1 corresponds to approximatel~ 70% of the
mutations in cystic fibrosis patients. Extended
. . .
. :,
" ~:
2~ 125~
17
haplotype data based on DNA markers closely linked to
the putative disease gene suggest that the remainder of
the CF mutant gene pool consists of multiple, different
mutations. This is now exemplified by this invention at
the 506 or 507 protein position. A small set of these
; latter mutant alleles (approximately 8~ may confer
residual pancreatic exocrine function in a subgroup of
patients who are pancreatic sufficient.
2.1 CHROMOSOME WALKING AND JUMPING
Large amounts of the DNA surrounding the D7S122 and
D75340 linkage regions of Rommens et al supra were
searched for candidate gene sequences. In addition to
conventional chromosome walking methods, chromosome
jumping techniques were employed to accelerate the
search process. From each jump endpoint a new
bidirectional walk could be initiated. Sequential walks
halted by "unclonable" regions often encountered in the
mammalian genome could be circumvented by chromosome
~ Umping .
The chromosome jumping library used has been
described previously [Collins et al, Science 235, 1046
(1987); Ianuzzi et al, Am. J. Hum. Genet. 44, 695
(1989)]. The original library was prepared from a
preparative pulsed field gel, and was intended to
contain partial EcoRl fragments of 70 - 130 kb;
subsequent experience with this library indicates that
smaller fragments were also represented, and jumpsizes
of 25 - 110 kb have been found. The library was plated
on sup~ host MC1061 and screened by standard techniques,
[Maniatis et al]. Positive clones were subcloned into
pBR~23Ava and the beginning and end of the jump
`~ identified by EcoRl and Ava 1 digestion, as described in
Collins, Genome analysis: A practical approach (IRL,
London, 1988), pp. 73-94) . For each clone, a fragment
from the end of the jump was checked to confirm its
location on chromosome 7. The contiguous chromosome
.,, .,~.
.
2~12~
18
region covered by chromosome walking and jumping was
about 250 kb. Direction of the jumps was biased by
careful choice of probes, as described by Collins et al
and Ianuzzi et al, supra. The entire region cloned,
including the sequences isolated with the use of the CF
gene cDNA, is approximately 500 kb.
The schematic representation of the chromosome
walking and jumping strategy is illustrated in Figure 2.
CF gene exons are indicated by Roman numerals in this
Figure. Horizontal lines above the map indicate walk
steps whereas the arcs above the map indicate jump
steps. The Figure proceeds from left to right in each
of six tiers with the direction of ends toward 7cen and
7qter as indicated. The restriction map for the enzymes
EcoRI, HindIII, and BamHI is shown above the solid line,
spanning the entire cloned region. Restriction sites
indicated with arrows rather than vertical lines
indicate sites which have not been unequivocally
positioned. Additional restriction sites for other
enzymes are shown below the line. Gaps in the cloned
region are indicated by ¦¦. These occur only in the
portion detected by cDNA clones of the CF transcript.
These gaps are unlikely to be large based on pulsed
field mapping of the region. The walking clones, as
indicated by horizontal arrows above the map, have the
direction of the arrow indicating the walking progress
obtained with each clone. Cosmid clones begin with the
letter c; all other clones are phage. Cosmid CF26
proved to be a chimera; the dashed portion is derived
from a different genomic fragment on another chromosome.
Roman numerals I through XXIV indicate the location of
exons of the CF gene. The horizontal boxes shown above
the line are probes used during the experiments. Three
of the probes represent independent subcloning of
j 35 fragments previously identified to detect polymorphisms
in this region: H2.3A corresponds to probe XV2C (X.
:
2~12~3
19
Estivill et al, Nature, 326: 840 (1987)), probe E1
corresponds to KMl9 (Estivill, supra), and probe E4.1
corresponds to Mp6d.9 (X. Estivill et al. Am. J. Hum.
Genet. 44, 704 (1989)). G-2 is a subfragment of E6
which detects a transcribed sequence. R161, R159, and
; R160 are synthetic oligonucleotides constructed fromparts of the IRP locus sequence [B. J. Wainwright et al,
EMBO J., 7: 1743 (1988)], indicating the location of
this transcript on the genomic map.
As the two independently isolated DNA markers,
D7S122 (pH131) and D7S340 (TM58), were only
approximately 10 kb apart (Figure 2), the walks and
jumps were essentially initiated from a single point.
The direction of walking and jumping with respect to MET
and D7S8 was then established with the crossing of
several rare-cutting restriction endonuclease
recognition sites (such as those for Xho I, Nru I and
Not I, see Figure 2) and with reference to the long
range physical map of J. M. Rommens et al. Am. J. Hum.
Genet., in press; A. M. Poustka, et al, Genomics 2, 337
(1988); M. L. Drumm et al. Genomics 2, 346 (1988).
; The pulsed field mapping data also revealed that the
Not I site identified by the inventors of the present
invention (see Figure 2, position 113 kb) corresponded
to the one previously found associated with the IRP
locus (Estivill et al 1987, supra). Since subsequent
genetic studies showed that CF was most likely located
between IRP and D7S8 [M. Farrall et al, Am. J. Hum.
Genet. 43, 471 (1988), B.S. Kerem et al. Am. J. Hum.
Genet. 44, 827 (1989)], the walking and jumping effort
was continued exclusively towards cloning of this
interval. It is appreciated, however, that other coding
regions, as identified in Figure 2, for example, G-2,
CF14 and CF16, were located and extensively
35 investigated. Such extensive investigations of these ~--
other regions revealed that they were not the CF gene -
, ~ ~
, .,
.'' ' ~
20112~3
based on genetic data and sequence analysis. Given the
lack of knowledge of the location of the CF gene and its
characteristics, the extensive and time consuming
examination of the nearby presumptive coding regions did ;
not advance the direction of search for the CF gene.
However, these investigations were necessary in order to
rule out the possibility of the CF gene being in those
regions.
Three regions in the 280 kb segment were found not
to be readily recoverable in the amplified genomic
libraries initially used. These less clonable regions
were located near the DNA segments H2.3A and X.6, and
just beyond cosmid cW44, at positions 75-100 kb, 205-225
kb, and 275-285 kb in Figure 2, respectively. The
recombinant clones near H2.3A were found to be very
unstable with dramatic rearrangements after only a few
passages of bacterial culture. To fill in the resulting
gaps, primary walking libraries were constructed using
special host-vector systems which have been reported to
allow propagation of unstable sequences [A. R. Wyman, L.
B. Wolfe, D. Botstein, Proc. Nat. Acad. Sci. U. S. A.
. 82, 2880 (1985); K. F. Wertman, A. R. Wyman, D.
Botstein, Gene 49, 253 (1986); A. R. Wyman, K. F.
Wertman, D. Barker, C. Helms, W. H. Petri, Gene, 49, 263
~ 25 (1986)]. Although the region near cosmid cW44 remains
`l to be recovered, the region near X.6 was successfully
- rescued with these libraries.
2.2 CONSTRUCTION OF GENOMIC LIBRARIES
Genomic libraries were constructed after
procedures described in Manatis, et al, Molecular
Cloninq: ~ Laboratory Manual (Cold Spring Harbor
Laboratory, Cold Spring Harbor, New York 1982) and are
listed in Table 1. This includes eight phage
, libraries, one of which was provided by T. Maniatis
[Fritsch et al, Cell, 19:959 (1980)]; the rest were
i constructed as part of this work according to procedures
:'
, ,.. ~ , . : , . . . : , . . . .. ~ ~ . ,
2 ~ 3
21
described in Maniatis et al, supra. Four phage
libraries were cloned in ~DASH (commercially available
from Stratagene) and three in ~FIX (commercially
available from Stratagene), with vector arms provided b~
the manufacturer. One ~DASH library was constructed
from Sau 3A-partially digested DNA from a human-hamster
hybrid containing human chromosome 7 (4AF/102/K015)
[Rommens et al Am. J. Hum. Genet 43, 4 (1988)], and
other libraries from partial Sau3A, total BamHI, or
total EcoRI digestion of human peripheral blood or
lymphoblastoid DNA. To avoid loss of unstable
sequences, five of the phage libraries were propagated
on the recombination-deficient hosts DB1316 (recD~), CES
200 (recBC~) [Wyman et al, supra , Wertman et al supra,
Wyman et al supra]; or TAP90 [Patterson et al Nucleic
Acids Res. 15:6298 (1987)]. Three cosmid libraries were
then constructed. In one the vector pCV108 [Lau et al
~- Proc. Natl. Acad. Sci USA 80:5225 (1983)] was used to
clone partially digested (Sau 3A) DNA from 4AF/102/K015
[Rommens et al Am.J. Hum. Genet. 43:4 (1988)]. A second
cosmid library was prepared by cloning partially
` digested (Mbo I) human lymphoblastoid DNA into the
- vector pWE-IL2R, prepared by inserting the RSV (Rous
Sarcoma Virus) promoter-driven cDNA for the
25 interleukin-2 receptor ~-chain (supplied by M. Fordis
`- and B. Howard) in place of the neo-resistance gene of
n pWE15 [Wahl et al Proc. Natl. Acad. Sci. USA 84:2160
(1987)]. An additional partial Mbo I cosmid library was
prepared in the vector pWE-IL2-Sal, created by inserting
30 a Sal I linker into the Bam HI cloning site of pWE-EL2R
(M. Drumm, unpublished data); this allows the use of the -
partial fill-in technique to ligate Sal I and Mbo I ~-
ends, preventing tandem insertions [Zabarovsky et al
- Gene 42:19 (1986)]. Cosmid libraries were propagated in
35 E. coli host strains DH1 or 490A [M. Steinmetz, A.
Winoto, K. Minard, L. Hood, Cell 28, 489(1982)].
,
:,, '
2~1~2~
22
TABLE 1
' OENaMrC ~:
Vector Source of human DNA Host Complexit~ Ref
5 ~ Charon HaeII/AluI-partially LE392 1 x 106 Lawn
` 4A digested total human (amplified) et al
liver DNA 1980
pCV108Sau3a-partially digested DKl 3 x lo6
DNA from 4AF/K015 (amplified)
~dash Sau3A-partially digested LE392 1 x 1o6
DNA from 4AF/KO15 (amplified)
-~ 15 ~dashSau3A-partially digested DB1316 1.5 x 106
total human peripheral
blood DNA
~dashBamHI-digested total DB1316 1.5 x 106
r~ 20human peripheral blood
DNA
~,
~dashEcoRI-partially digested DB1316 8 x 1o6
~ total human peripheral
!~. 25 blood DNA
~FIXMboI-partially digested LE392 1.5 x 106
; human lymphoblastoid DNA
30 ~FIXMboI-partially digested OE200 1.2 x 106
human lymphoblastoid DNA
!' ~FIXMboI-partially digested TAP90 1.3 x 106
human lymphoblastoid DNA
~`` pWE-IL2R MboI-partially digested 490A 5 x 105
~! human lymphoblastoid DNA
pWE-IL2R- MboI-partially digested 490A 1.2 x 10
40 Sal human lymphoblastoid DNA
Ch3A EcoRI-partially digested MC1061 3 x 106 Collins
` ~lac (24-110 kb) et al
(jumping) human lymphoblastoid DNA supra and
` 45 Iannuzzi
`-~ et al
, supra
:r ~ ~ ~
~
. . .
.. ''',, .
2~125~
Three of the phage libraries were propagated and
amplified in E. coli bacterial strain LE392. Four
subsequent libraries were plated on the recombination-
deficient hosts DB1316 (recD~) or CES200 (rec BC-)
[Wyman 1985, supra; Wertman 1986, suPra; and Wyman 1986,
supra] or in one case TAP90 [T.A. Patterson and M. Dean,
Nucleic Acids Research 15, 6298 (1987)].
Single copy DNA segments (free of repetitive
elements) near the ends of each phage or cosmid insert
were purified and used as probes for library screening
to isolate overlapping DNA fragments by standard
procedures. (Maniatis, et al, supra).
1-2 x 106 phage clones were plated on 25-30 150 mm
petri dishes with the appropriate indicator bacterial
15 host and incubated at 37~C for 10-16 hr. Duplicate
"lifts" were prepared for each plate with
nitrocellulose or nylon membranes, prehybridized and
hybridized under conditions described [Rommens et al,
1988, supra]. Probes were labelled with 32p to a
specific activity of >5 x 1o8 cpm/~g using the random
priming procedure [A.P. Feinberg and B. Vogelstein,
Anal. Biochem. 132, 6 (1983)]. The cosmid library was
spread on ampicillin-containing plates and screened in a
similar manner.
DNA probes which gave high background signals --
could often be used more successfully by preannealing
the boiled probe with 250 ~g/ml sheared denatured
placental DNA for 60 minutes prior to adding the probe
to the hybridization bag.
; 30 For each walk step, the identity of the cloned DNA ~ ~
fragment was determined by hybridization with a somatic - -
cell hybrid panel to confirm its chromosomal location,
and by restriction mapping and Southern blot analysis to
`i confirm its colinearity with the genome.
The total combined cloned region of the genomic DNA
sequences isolated and the overlapping cDNA clones,
.'', : ~.
.~ ~
201~3
24
extended >500 kb. To ensure that the DNA segments
isolated by the chromosome walking and jumping
procedures were colinear with the genomic sequence,
each segment was examined by:
(a) hybridization analysis with human-rodent
somatic hybrid cell lines to confirm chromosome 7
localization,
(b) pulsed field gel electrophoresis, and
~ (c) comparison of the restriction map of the cloned
- 10 DNA to that of the genomic DNA.
Accordingly, single copy human DNA sequences were
~ isolated from each recombinant phage and cosmid clone
; and used as probes in each of these hybridization
analyses as performed by the procedure of Maniatis, et
al supra.
While the majority of phage and cosmid isolates
represented correct walk and jump clones, a few resulted
from cloning artifacts or cross-hybridizing sequences
from other regions in the human genome, or from the
hamster genome in cases where the libraries were derived
from a human-hamster hybrid cell line. Confirmation of
correct localization was particularly important for
clones isolated by chromosome jumping. Many jump clones ~-
were considered and resulted in non-conclusive
1 25 information leading the direction of investigation away
from the gene.
2.3 CONFIRMATION OF THE RESTRICTION MAP
Further confirmation of the overall physical map of
the overlapping clones was obtained by long range ~;~
30 restriction mapping analysis with the use of pulsed ~
` field gel electrophoresis (J. M. Rommens, et al. Am. J. ~ ~ ;Hum. Genet. in press, A. M. Poustka et al, 1988, supra
M.L. Drumm et al, 1988 supra).
- Figures 3A to 3E illustrates the findings of the
long range restriction mapping study, where a schematic
representation of the region is given in Panel E. DNA
~ .
: .
. .
'`,',".`.: '' ~ . , :
- 2~2~
from the human-hamster cell line 4AF/102/K015 was
digested with the enzymes (A) Sal I, (B) Xho I, (C) Sfi
I and (D) Nae I, separated by pulsed field gel
electrophoresis, and transferred to Zetaprobe'~(BioRad).
For each enzyme a single blot was sequentially
hybridized with the probes indicated below each of the
panels of Figure A to D, with stripping of the blot
between hybridizations. The symbols for each enzyme of
Figure 3E are: A, Nae I; B, Bss HII; F. Sfi I; L, Sal I;
M, Mlu I; N, Not I; R, Nru I; and X, Xho 1. C
corresponds to the compression zone region of the gel. -~
DNA preparations, restriction digestion, and crossed
field gel electrophoresis methods have been described ~-
(Rommens et al, in press, supra). The gels in Figure 3
were run in 0.5X TBE at 7 volts/cm for 20 hours with
switching linearly ramped from 10-40 seconds for (A),
(B), and (C), and at 8 volts/cm for 20 hours with
switching ramped linearly from 50-150 seconds for (D).
Schematic interpretations of the hybridization pattern -~
are given below each panel. Fragment lengths are in
kilobases and were sized by comparison to oligomerized
bacteriophage ~DNA and Saccharomyces cerevisiae
- chromosomes.
H4.0, J44, EGl.4 are genomic probes generated from
the walking and jumping experiments (see Figure 2).
J30 has been isolated by four consecutive jumps from
D7S8 (Collins et al, 1987, supra; Ianuzzi et al, 1989,
supra; M. Dean, et al, submitted for publication). 10-
1, B.75, and CEl.5/1.0 are cDNA probes which cover
different regions of the CF transcript: 10-1 contains
exons I - VI, B.75 contains exons V - XII, and CEl.5/1.0
contains exons XII - XXIV. Shown in Figure 3E is a
composite map of the entire MET - D7S8 interval. The
open boxed region indicates the segment cloned by
walking and jumping, and the closed arrow portion
indicates the region covered by the CF transcript. The
- 2~2!~3
26
CpG-rich region associated with the D7S23 locus
(Estivill et al, 1987, supra) is at the Not I site shown
in parentheses. This and other sites shown in
parentheses or square brackets do not cut in
4AF/102/K015, but have been observed in human
lymphoblast cell lines.
2.4 IDENTIFICATION OF CF GENE
Based on the findings of long range restriction
mapping detailed above it was determined that the entire
CF gene is contained on a 380 kb Sal I fragment.
Alignment of the restriction sites derived from pulsed
field gel analysis to those identified in the partially
overlapping genomic DNA clones revealed that the size of
the CF gene was approximately 250 kb.
The most informative restriction enzyme that served
to align the map of the cloned DNA fragments and the
long range restriction map was Xho I; all of the g Xho l
sites identified with the recombinant DNA clones
appeared to be susceptible to at least partial cleavage
in genomic DNA (compare maps in Figures 1 and 2).
Furthermore, hybridization analysis with probes derived
from the 3' end of the CF gene identified 2 SfiI sites
and confirmed the position of an anticipated Nae I site.
These findings further supported the conclusion
G5 that the DNA segments isolated by the chromosome walking
and jumping procedures were colinear with the genuine
sequence.
2.5 CRITERIA FOR IDENTIFICATION
A positive result based on one or more of the
following criteria suggested that a cloned DNA segment
may contain candidate gene sequences:
(a) detection of cross-hybridizing sequences in
other species (as many genes show evolutionary
conservation),
(b) identification of CpG islands, which often mark
the 5' end of vertebrate genes [A. P. Bird, Nature, 321,
, ~ :
' ,
2~1~253
27
209 (1986); M. Gardiner-Garden and M. Frommer, J. Mol.
Biol. 196, 261 (1987)],
(c) examination of possible mRNA transcripts in
tissues affected in CF patients,
(d) isolation of corresponding cDNA sequences,
(e) identification of open reading frames by direct
sequencing of cloned DNA segments.
Cross-species hybridization showed strong sequence -~
conservation between human and bovine DNA when CF14,
E4.3 and Hl.6 were used as probes, the results of which
are shown in Figures 4A, 4B and 4C. - -
Human, bovine, mouse, hamster, and chicken genomic ~-
DNAs were digested with Eco RI (R), Hind III (H), and
Pst I (P), electrophoresed, and blotted to Zetabind~
(BioRad). The hybridization procedures of Rommens et
al, 1988, supra, were used with the most stringent wash
at 55C, 0.2X SSC, and 0.1% SDS. The probes used for
hybridization, in Figure 4, included: (A) entire
cosmid CF14, (B) E4.3, (C) H1.6. In the schematic of
Figure (D), the shaded region indicates the area of
;~ cross-species conservation.
The fact that different subsets of bands were -
detected in bovine DNA with these two overlapping DNA
` segments (H1.6 and E4.3) suggested that the conserved
sequences were located at the boundaries of the
overlapped region (Figure 4(D)). When these DNA
segments were used to detect RNA transcripts from a
;l ` variety of tissues, no hybridization signal was
detected. In an attempt to understand the cross-
: 30 hybridizing region and to identify possible open reading
frames, the DNA sequences of the entire Hl.6 and part of
the E4.3 fragment were determined. The results showed
that, except for a long stretch of CG-rich sequence
containing the recognition sites for two restriction
enzymes ~Bss HII and Sac II), often found associated
with undermethylated CpG islands, there were only short
,
., '
201~2~3
28
open reading frames which could not easily explain the
strong cross-species hybridization signals.
To examine the methylation status of this highly
CpG-rich region revealed by sequencing, genomic DNA
samples prepared from fibroblasts and lymphoblasts were
digested with the restriction enzymes Hpa II and Msp I
and analyzed by gel blot hybridization. The enzyme Hpa
II cuts the DNA sequence 5'-CCGG-3' only when the
second cytosine is unmethylated, whereas Msp I cuts this
sequence regardless of the state of methylation. Small
DNA fragments were generated by both enzymes, indicating
that this CpG-rich region is indeed undermethylated in ~-
genomic DNA. The gel-blot hybridization with the E4.3 ~ -
segment (Figure 6) reveals very small hybridizing
fragments with both enzymes, indicating the presence of
; a hypomethylated CpG island. ;~
The above results strongly suggest the presence of
a coding region at this locus. Two DNA segments (E4.3
and Hl.6) which detected cross-species hybridization
signals from this area were used as probes to screen
cDNA libraries made from several tissues and cell types.
~` cDNA libraries from cultured epithelial cells were
prepared as follows. Sweat gland cells derived from a
~` non-CF individual and from a CF patient were grown to
first passage as described [G. Collie et al, In Vitro
Cell. Dev. Biol. 21, 592,1985]. The presence of
outwardly rectifying channels was confirmed in these
:~ cells (J.A. Tabcharani, T.J. Jensen, J.R. Riordan, J.W.
Hanrahan, J. Memb. Biol., in press) but the CF cells
were insensitive to activation by cyclic AMP (T.J.
Jensen, J.W. Hanrahan, J.A. Tabcharani, M. Buchwald and
J.R. Riordan, Pediatric Pulmonoloqy, Supplement 2, 100,
1988). RNA was isolated from them by the method of J.M.
Chirgwin et al (Biochemistrv 18, 5294, 1979). Poly
~l 35 A+RNA was selected (H. Aviv and P. Leder, Proc. Natl.
`, Acad. Sci. USA 69, 1408, 1972) and used as template for ~;
. .~ .
, . ~,
.;~
29 2~ ~ 2~3
the synthesis of cDNA with oligo (dT) 12-18 as a primer.
The second strand was synthesized according to Gubler
and Hoffman (Gene 25, 263, 1983). This was methylated
with Eco RI methylase and ends were made flush with T4
DNA polymerase. Phosphorylated Eco RI linkers were
ligated to the cDNA and restricted with Eco RI.
Removal of excess linkers and partial size
fractionation was achieved by Biogel A-50
chromatography. The cDNAs were then ligated into the
Eco RI site of the commercialy available lamdba ZAP.
Recombinants were packaged and propagated in E. coli
BB4. Portions of the packaging mixes were amplified and
the remainder retained for screening prior to
amplification. The same procedures were used to
construct a library from RNA isolated from preconfluent
cultures of the T-84 colonic carcinoma cell line
(Dharmsathaphorn, K. et al. Am. J. Physiol. 246, G204,
1984). The numbers of independent recombinants in the
three libraries were: 2 x 106 for the non-CF sweat
gland cells, 4.5 x 106 for the CF sweat gland cells and
3.2 x 106 from T-84 cells. These phages were plated at
50,000 per 15 cm plate and plaque lifts made using nylon
membranes ~Biodyne) and probed with DNA fragments
labelled with 32p using DNA polymerase I and a random
mixture of oligonucleotides as primer. Hybridization
- conditions were according to G.M. Wahl and S.L. Berger
(Meth. Enzymol. 152,415, 1987). Bluescript~ plasmids
were rescued from plaque purified clones by excision
with M13 helper phage. The lung and pancreas libraries
were purchased from Clontech Lab Inc. with reported
sizes of 1.4 x 106 and 1.7 x 106 independent clones.
After screening 7 different libraries each
containing 1 x 105 - 5 x 106 independent clones, 1
single clone (identified as 10-1) was isolated with Hl.6
from a cDNA library made from the cultured sweat gland
epithelial cells of an unaffected (non-CF) individual.
;' :
: :
. :
::
. ,~
. . k
20~12~3
DNA sequencing analysis showed that probe 10-1
contained an insert of 920 bp in size and one potential,
long open reading frame (ORF). Since one end of the
sequence shared perfect sequence identity with H1.6, it
5 was concluded that the cDNA clone was probably derived
from this region. The DNA sequence in common was,
however, only 113 bp long (see Figures 1 and 7). As
detailed below, this sequence in fact corresponded to
the 5'-most exon of the putative CF gene. The short
t 10 sequence overlap thus explained the weak hybridization
signals in library screening and inability to detect
transcripts in RNA gel-blot analysis. In addition, the
orientation of the transcription unit was tentatively
established on the basis of alignment of the genomic DNA
15 sequence with the presumptive ORF of 10-1.
Since the corresponding transcript was estimated to
be approximately 6500 nucleotides in length by RNA gel-
blot hybridization experiments, further cDNA library
screening was required in order to clone the remainder
20 of the coding region. As a result of several
successive screenings with cDNA libraries generated from
the colonic carcinoma cell line T84, normal and CF sweat
'`! gland cells, pancreas and adult lungs, 18 additional
; clones were isolated (Figure 7, as subsequently
discussed in greater detail). DNA sequence analysis
revealed that none of these cDNA clones corresponded to
the length of the observed transcript, but it was
~` possible to derive a consensus sequence based on
, overlapping regions. Additional cDNA clones
- 30 corresponding to the 5' and 3' ends of the transcript -
were derived from 5' and 3' primer-extension
experiments. Together, these clones span a total of
about 6.1 kb and contain an ORF capable of encoding a
polypeptide of 1480 amino acid residues (Figure 1).
It was unusual to observe that most of the cDNA
` clones isolated here contained sequence insertions at
. .
~,
`.~`i
i
201 ~ 25~
31
various locations of the restriction map of Figure 7.
- The map details the genomic structure of the CF gene.
Exon/intron boundaries are given where all cDNA clones
isolated are schematically represented on the upper half
of the figure. Many of these extra sequences clearly
corresponded to intron regions reversely transcribed
during the construction of the cDNA, as revealed upon
alignment with genomic DNA sequences.
Since the number of recombinant cDNA clones for the
CF gene detected in the library screening was much less
than would have been expected from the abundance of
transcript estimated from RNA hybridization experiments,
it seemed probable that the clones that contained
aberrant structures were preferentially retained while
15 the proper clones were lost during propagation. -
Consistent with this interpretation, poor growth was
observed for the majority of the recombinant clones
isolated in this study, regardless of the vector used.
The procedures used to obtain the 5' and 3' ends of
the cDNA were similar to those described (M. Frohman et
al, Proc. Nat. Acad. Sci, USA, 85, 8998-9002, 1988).
For the 5' end clones, total pancreas and T84 poly A +
RNA samples were reverse transcribed using a primer,
` (lOb), which is specific to exon 2 similarly as has been
described for the primer extension reaction except that
radioactive tracer was included in the reaction. The
fractions collected from an agarose bead column of the
first strand synthesis were assayed by polymerase chain
reaction (PCR) of eluted fractions. The
;{ 30 oligonucleotides used were within the 10-1 sequence (145
nucleotides apart) just 5' of the extension primer. The
~ earliest fractions yielding PCR product were pooled and
!-, concentrated by evaporation and subsequently tailed with
terminal deoxynucleotidyl transferase (BRL Labs.) and
dATP as recommended by the supplier (BRL Labs). A
second strand synthesis was then carried out with Taq -~
.,, : ~
.: ~
? ~: ~
`: :
2~1253
32
Polymerase (Cetus, AmpliTaq'~) using an oligonucleotide
containing a tailed linker sequence
5'CGGAATTCTCGAGATC(T)123 .
Amplification by an anchored (PCR) experiment
using the linker sequence and a primer just internal to
the extension primer which possessed the Eco RI
restriction site at its 5' end was then carried out.
Following restriction with the enzymes ~co RI and Bgl II
and agaros~ gel purification size selected products were
cloned into the plasmid Bluescript KS available from
Stratagene by standard procedures (Maniatis et al,
supra). Essentially all of the recovered clones
;~ contained inserts of less than 350 nucleotides. To
obtain the 3' end clones, first strand cDNA was prepared
15 with reverse transcription of 2 ~g T84 poly A + RNA
using the tailed linker oligonucleotide previously
described with conditions similar to those of the primer
extension. Amplification by PCR was then carried out
with the linker oligonucleotide and three different
20 oligonucleotides corresponding to known sequences of
clone T16-4.5. A preparative scale reaction (2 x 100
ul) was carried out with one of these oligonucleotides
with the sequence 5'ATGAAGTCCAAGGATTTAG3'.
This oligonucleotide is approximately 70
25 nucleotides upstream of a Hind III site within the known
sequence of Tl6-4.5. Restriction of the PCR product
with Hind III and Xho l was followed by agarose gel
~, purification to size select a band at 1.0-1.4 kb. This
product was then cloned into the plasmid Bluescript KS
30 available from Stratagene. Approximately 20~ of the
obtained clones hybridized to the 3' end portion of T16-
3 4.5. 10/10 of plasmids isolated from these clones had
identical restriction maps with insert sizes of approx.
1.2 kb. All of the PCR reactions were carried out for
35 30 cycles in buffer suggested by an enzyme supplier.
. .
.. :
20~ 12~3
33
An extension primer positioned 157 nt from the
5'end of 10-1 clone was used to identify the start point
of the putative CF transcript. The primer was end
labeled with ~[32P]ATP at 5000 Curies/mmole and T4
polynucleotide kinase and purified by spun column gel
filtration. The radiolabeled primer was then annealed
with 4-5 ug poly A + RNA prepared from T-84 colonic
carcinoma cells in 2X reverse transcriptase buffer for 2
hrs. at 60 C. Following dilution and addition of AMV
reverse transcriptase (Life Sciences, Inc.) incubation
at 41 C proceeded for 1 hour. The sample was then
adjusted to 0.4M NaOH and 20 mM EDTA, and finally
neutralized, with NH40Ac, pH 4.6, phenol extracted,
ethanol precipitated, redissolved in buffer with
formamide, and analyzed on a polyacrylamide sequencing
gel. Details of these methods have been described
(Meth. EnzYmol. 152, 1987, Ed. S.L. Berger, A.R. Kimmel,
Academic Press, N.Y.).
Results of the primer extension experiment using an
extension oligonucleotide primer starting 157
nucleotides from the 5' end of 10-1 is shown in Panel A
of Figure 10. End labeled ~X174 bacteriophage digested
with Hae III (BRL Labs) is used as size marker. Two
major products are observed at 216 and 100 nucleotides.
The sequence corresponding to 100 nucleotides in 10-1
corresponds to a very GC rich sequence (11/12)
suggesting that this could be a reverse transcriptase
pause site. The 5' anchored PCR results are shown in
panel B of Figure 10. The 1.4% agarose gel shown on
the left was blotted and transferred to Zetaprobe~
membrane (Bio-Rad Lab). DNA gel blot hybridization with
radiolabeled 10-1 is shown on the right. The 5'
extension products are seen to vary in size from 170-280
~, nt with the major product at about 200 nucleotides. The
PCR control lane shows a fragment of 145 nucleotides.
; It was obtained by using the test oligomers within the
~ ~ ,
2 ~ ~
34
lO-l sequence. r~he size markers shown correspond to
sizes of 154, 220/210, 298, 344, 394 nucleotides (lkb
ladder purchased from BRL Lab).
The schematic shown below Panel B of Figure 10
outlines the procedure to obtain double stranded cDNA
used for the amplification and cloning to generate the
clones PA3-5 and TB2-7 shown in Figure 7. The anchored
PCR experiments to characterize the 3'end are shown in
panel C. As depicted in the schematic below Figure lOC,
three primers whose relative position to each other were
known were used for amplification with reversed
transcribed T84 RNA as described. These products were
separated on a 1% agarose gel and blotted onto nylon
membrane as described above. DNA-blot hybridization
15 with the 3' portion of the T16-4.5 clone yielded bands
of sizes that corresponded to the distance between the
specific oligomer used and the 3'end of the transcript.
These bands in lanes 1, 2a and 3 are shown schematically
below Panel C in Figure 10. The band in lane 3 is weak
as only 60 nucleotides of this segment overlaps with the
probe used. Also indicated in the schematic and as
shown in the lane 2b is the product generated by
restriction of the anchored PCR product to facilitate
cloning to generate the THZ-4 clone shown in Figure 7.
DNA-blot hybridization analysis of genomic DNA
digested with EcoRI and HindIII enzymes probed with
portions of cDNAs spanning the entire transcript suggest
that the gene contains at least 26 exons numbered as
Roman numerals I through XXVI (see Figure 9). These
correspond to the numbers 1 through 26 shown in Figure
7. The size of each band is given in kb.
In Figure 7, open boxes indicate approximate
positions of the 24 exons which have been identified by
the isolation of >22 clones from the screening of cDNA
libraries and from anchored PCR experiments designed to
clone the 5' and 3' ends. The lengths in kb of the Eco
' ~
,*.......... . ..
~. .. . . :
, j ~ :,. . , ~ ,
20~ ~253
-
RI genomic fragments detected by each exon is also
indicated. The hatched boxes in Figure 7 indicate the
presence of intron sequences and the stippled boxes
indicate other sequences. Depicted in the lower left by
the closed box is the relative position of the clone
H1.6 used to detect the first cDNA clone 10-1 from among
106 phage of the normal sweat gland library. As shown in
Figures 4(D) and 7, the genomic clone H1.6 partially
overlaps with an EcoRI fragment of 4.3 kb. All of the
cDNA clones shown were hybridized to genomic DNA and/or
were fine restriction mapped. Examples of the
restriction sites occurring within the cDNAs and in the
corresponding genomic fragments are indicated.
With reference to Figure 9, the hybridization
analysis includes probes; i.e., cDNA clones 10-1 for
panel A, T16-1 (3' portion) for panel B, T16-4.5
(central portion)~for panel C and T16-4.5 (3' end
portion) for panel D. In panel A of Figure 9, the cDNA
probe 10-1 detects the genomic bands for exons I through
VI. The 3' portion of T16-1 generated by NruI
,
restriction detects exons IV through XIII as shown in
~'Panel B. This probe partially overlaps with 10-1.
Panels C and D, respectively, show genomic bands
detected by the central and 3' end EcoRI fragments of
25 the clone T16-4.5. Two EcoRI sites occur within the -~
cDNA sequence and split exons XIII and XIX. As
indicated by the exons in parentheses, two genomic EcoRI
bands correspond to each of these exons. Cross ;
hybridization to other genomic fragments was observed.
30 These bands, indicated by N, are not of chromosome 7 ;~
origin as they did not appear in human-hamster hybrids
containing human chromosome 7. The faint band in panel
D indicated by XI in brackets is believed to be caused
`,by the cross-hybridization of sequences due to internal ~-
35 homology with the cDNA. ~
.
. ., ~ ~ -:
;::
2011253
36
Since 10-1 detected a strong band on gel blot
hybridization of RNA from the T-84 colonic carcinoma
cell line, this cDNA was used to screen the library
constructed from that source. Fifteen positives were
obtained from which clones T6, T6/20, T11, T16-1 and
T13-1 were purified and sequenced. Rescreening of the
same library with a 0.75 kb Bam HI-Eco RI fragment from
the 3' end of T16-1 yielded T16-4.5. A 1.8kb EcoRI
fragment from the 3' end of T16-4. 5 yielded T8-B3 and
T12a, the latter of which contained a polyadenylation
signal and tail. Simultaneously a human lung cDNA
library was screened; many clones were isolated
including those shown here with the prefix `CDL'. A :
pancreas library was also screened, yielding clone
CDPJ5.
To obtain copies of this transcript from a CF
patient, a cDNA library from RMA of sweat gland
epithelial cells from a patient was screened with the
0.75 kb Bam HI - Eco RI fragment from the 3' end of T16-
1 and clones C16-1 and C1-1/5, which covered all but
exon 1, were isolated. These two clones both exhibit a
j 3 bp deletion in exon lO which is not present in any
other clone containing that exon. Several clones,
including CDLS26-1 from the lung library and T6/20 and
T13-1 isolated from T84 were derived from partially
processed transcripts. This was confirmed by genomic
hybridization and by sequencing across the exon-intron
boundaries for each clone. T11 also contained
additional sequence at each end. T16-4.5 contained a
small insertion near the boundary between exons 10 and
11 that did not correspond to intron sequence. Clones
CDLS16A, lla and 13a from the lung library also
contained extraneous sequences of unknown origin. The
clone C16-1 also contained a short insertion
~orresponding to a portion of the ~-transposon of E.
coli; this element was not detected in the other clones.
. , .
:;
201 t 2~3
37
The 5' clones PA3-5, generated from pancreas RNA and
TB2-7 generated from T84 RNA using the anchored PCR
technique have identical sequences except for a single
nucleotide difference in length at the 5' end as shown
in Figure 1. The 3' clone, THZ-4 obtained from T84 RNA
contains the 3' sequence of the transcript in
concordance with the genomic sequence of this region.
A combined sequence representing the presumptive
coding region of the CF gene was generated from
overlapping cDNA clones. Since most of the cDNA clones
were apparently derived from unprocessed transcripts,
further studies were performed to ensure the
authenticity of the combined sequence. Each cDNA clone
was first tested for localization to chromosome 7 by
hybridization analysis with a human-hamster somatic
cell hybrid containing a single human chromosome 7 and
by pulsed field gel electrophoresis. Fine restriction
enzyme mapping was also performed for each clone. While
overlapping regions were clearly identifiable for most ~-
of the clones, many contained regions of unique
restriction patterns.
To further characterize these cDNA clones, they -~
were used as probes in gel hybridization experiments
with EcoRI-or HindIII-digested human genomic DNA. As
shown in Figure 9, five to six different restriction
fragments could be detected with the 10-1 cDNA and a
, similar number of fragments with other cDNA clones,
; suggesting the presence of multiple exons for the
putative CF gene. The hybridization studies also ;~
identified those cDNA clones with unprocessed intron
sequences as they showed preferential hybridization to a
subset of genomic DNA fragments. For the confirmed cDNA
clones, their corresponding genomic DNA segments were
isolated and the exons and exon/intron boundaries
sequenced. As indicated in Figure 7, at least 28 exons
have been identified which includes split exons 6a, 6b,
''~
38 2 0~1 25 ~
6c, 14a, 14b and 17a, 17b. Based on this information
and the results of physical mapping experiments, the
gene locus was estimated to span 250 kb on chromosome 7.
2.6 THE SEOUENCE
Figure 1 shows the nucleotide sequence of the
cloned cDNA encoding CFTR together with the deduced
amino acid sequence. The first base position
~,! corresponds to the first nucleotide in the 5' extension
clone PA3-5 which is one nucleotide longer than TB2-7.
Arrows indicate position of transcription initiation
site by primer extension analysis. Nucleotide 6129 is
i
;~ followed by a poly(dA) tract. Positions of exon
junctions are indicated by vertical lines. Potential
membrane-spanning segments were ascertained using the
algorithm of Eisenberg et al J. Mol. Biol. 179:125
(1984). Potential membrane-spanning segments as analyzed
and shown in Figure 11 are enclosed in boxes of Figure
1. In Figure 11, the mean hydropathy index [Kyte and
Doolittle, J. Molec. Biol. 157: 105, (1982)] of 9
residue peptides is plotted against the amino acid
number. The corresponding positions of features of
secondary structure predicted according to Garnier et
al, [J. Molec. Biol. 157, 165 (1982)] are indicated in
' the lower panel. Amino acids comprising putative ATP-
binding folds are underlined in Figure 1. Possible
sites of phosphorylation by protein kinases A (PKA) or C
(PKC) are indicated by open and closed circles,
respectively. The open triangle is over the 3bp (CTT)
which are deleted in CF (see discussion below). The
cDNA clones in Figure 1 were sequenced by the dideoxy
chain termination method employing 35S labelled
nucleotides by the Dupont Genesis 2000~ automatic DNA
sequencer.
The combined cDNA sequence spans 6129 base pairs
` 35 excluding the poly(A) tail at the end of the 3'
untranslated region and it contains an ORF capable of
.
, "
''
20~12~
39
encoding a polypeptide of 1480 amino acids (Figure 1).
An ATG (AUG) triplet is present at the beginning of this
ORF (base position 133-135). Since the nucleotide
sequence surrounding this codon (5'-AGACCAUGCA-3') has
the proposed features of the consensus sequence (CC)
A/GCCAUGG(G) of an eukaryotic translation initiation
site with a highly conserved A at the -3 position, it is
highly probable that this AUG corresponds to the first
methionine codon for the putative polypeptide.
To obtain the sequence corresponding to the 5' end
of the transcript, a primer-extension experiment was
performed, as described earlier. As shown in Figure
10A, a primer extension product of approximately 216
nucleotides could be observed suggesting that the 5'
end of the transcript initiated approximately 60
nucleotides upstream of the end of cDNA clone 10-1. A
modified polymerase chain reaction (anchored PCR) was
then used to facilitate cloning of the 5'-end sequences
(Figure 10b). Two independent 5'-extension clones, one
from pancreas and the other from T84 RNA, were
characterized by DNA sequencing and were found to differ
by only 1 base in length, indicating the most probable
~` initiation site for the transcript as shown in Figure 1.
Since most of the initial cDNA clones did not
contain a polyA tail indicative of the end of a mRNA,
anchored PCR was also applied to the 3' end of the
transcript (Frohman et al, 1988, supra). Three 3'-
extension oligonucleotides were made to the terminal ~ ~
portion of the cDNA clone T16-4.5. As shown in Figure ~;
10c, 3 PCR products of different sizes were obtained.
All were consistent with the interpretation that the end
I of the transcript was approximately 1.2 kb downstream
of the HindIII site at nucleotide position 5~27 (see
Figure 1). The DNA sequence derived from representative
clones was in agreement with that of the T84 cDNA clone
... .
~; .
'.:
:
~ 2011253
T12a (see Figure 1 and 7) and the sequence of the
corresponding 2.3 kb EcoRI genomic fragment.
; 3.0 MOLECULAR GENETICS OF CF
I 3.1 SITES OF EXPRESSION
i~, 5 To visualize the transcript for the putative CF
j' gene, RNA gel blot hybridization experiments were
;;- performed with the 10-1 cDNA as probe. The RNA
hybridization results are shown in Figure 8.
RNA samples were prepared from tissue samples
obtained from surgical pathology or at autopsy according
to methods previously described (A.M. Kimmel, S.L.
Berger, eds. Meth. Enzymol. 152, 1987). Formaldehyde
gels were transferred onto nylon membranes (Zetaprobe
:! TM; BioRad Lab). The membranes were then hybridized
with DNA probes labeled to high specific activity by the
~ random priming method ~A.P. Feinberg and B. Vogelstein,
`! Anal. Biochem. 132, 6, 1983) according to previously
published procedures (J. Rommens et al, Am. J. Hum.
Genet. 43, 645-663, 1988). Figure 8 shows hybridization
by the cDNA clone 10-1 to a 6.5kb transcript in the ~ -
tissues indicated. Total RNA (10 ~g) of each tissue,
and Poly A+ RNA (1 ~g) of the T84 colonic carcinoma cell
, line were separated on a 1% formaldehyde gel. The
positions of the 28S and 18S rRNA bands are indicated.
Arrows indicate the position of transcripts. Sizing was
established by comparison to standard RNA markers (BRL
- Labs). HL60 is a human promyelocytic leukemia cell line,
and T84 is a human colon cancer cell line.
Analysis reveals a prominent band of approximately
6.5 kb in size in T84 cells. Similar, strong
hybridization signals were also detected in pancreas and
primary cultures of cells from nasal polyps, suggesting
that the mature mRNA of the putative CF gene is
approximately 6.5 kb. Minor hybridization signals,
probably representing degradation products, were
detected at the lower size ranges but they varied
,~ .
.
:
.. ,, '~,;.:
2~112~3
41
between different experiments. Identical results were
obtained with other cDNA clones as probes. Based on the
; hybridization band intensity and comparison with those
detected for other transcripts under identical
experimental conditions, it was estimated that the
~ putative CF transcripts constituted approximately 0.01%
;~ of total mRNA in T84 cells.
`~ A number of other tissues were also surveyed by RNA
~ gel blot hybridization analysis in an attempt to
; 10 correlate the expression pattern of the 10-1 gene and
... .
the pathology of CF. As shown in Figure 8, transcripts,
l all of identical size, were found in lung, colon, sweat
; glands (cultured epithelial cells), placenta, liver, and
~~ parotid gland but the signal intensities in these
; 15 tissues varied among different preparations and were
generally weaker than that detected in the pancreas and ~;
nasal polyps. Intensity varied among different
preparations, for example, hybridization in kidney was
not detected in the preparation shown in Figure 8, but
can be discerned in subsequent repeated assays. No
hybridization signals could be discerned in the brain or
adrenal gland (Figure 8), nor in skin fibroblast and
lymphoblast cell lines.
In summary, expression of the CF gene appeared to
occur in many of the tissues examined, with higher
levels in those tissues severely affected in CF. While
this epithelial tissue-specific expression pattern is in
good agreement with the disease pathology, no
significant difference has been detected in the amount
or size of transcripts from CF and control tissues,
consistent with the assumption that CF mutations are
,
i subtle changes at the nucleotide level.
; . . .
. .,
.,.~ .
.. ~
, ........................................................................... .
~ .
.~'~'.
~ ' Y . . ~
; ~ `
2~1~2~3
42
3.2 THE MAJOR CF MUTATION
Figure 16 shows the DNA sequence at the F508
deletion. On the left, the reverse complement of the
sequence from base position 1649-1664 of the normal
sequence (as derived from the cDNA clone T16). The
nucleotide sequence is displayed as the output (in
arbitrary fluorescence intensity units, y-axis) plotted
' against time (x-axis) for each of the 2 photomultiplier
tubes (PMT#1 and #2) of a Dupont Genesis 2000TM DNA
analysis system. The corresponding nucleotide sequence
is shown underneath. On the right is the same region
from a mutant sequence (as derived from the cDNA clone
C16). Double-stranded plasmid DNA templates were
prepared by the alkaline lysis procedure. Five ~g of
plasmid DNA and 75 ng of oligonucleotide primer were
used in each sequencing reaction according to the
protocol recommended by Dupont except that the annealing ~-
was done at 45C for 30 min and that the
elongation/termination step was for 10 min at 42C. The
unincorporated fluorescent nucleotides were removed by
precipitation of the DNA sequencing reaction product
with ethanol in the presence of 2.5 M ammonium acetate
~, at pH 7.0 and rinsed one time with 70% ethanol. The
primer used for the T16-1 sequencing was a specific
oligonucleotide 5'GTTGGCATGCTTTGATGACGCTTC3' spanning
base position 1708 - 1731 and that for C16-1 was the
universal primer SK for the Bluescript vector
~` (Stratagene).
Figure 17 also shows the DNA sequence around the
F508 deletion, as determined by m?nual sequencing. The
normal sequence from base position 1726-1651 (from cDNA
T16-1) is shown beside the CF sequence (from cDNA C16-
l). The left panel shows the sequences from the coding
strands obtained with the B primer
(5'GTTTTCCTGGATTATGCCTGGCAC3') and tne right panel
those from the opposite strand with the n primer
'
:
2 0 1 ~
43
(5'GTTGGCATGCTTTGATGACGCTTC3'). The brackets indicate
~` the three nucleotides in the normal that are absent in
CF (arrowheads). Sequencing was performed as described
in F. Sanger, S. N~cklen, A. R. Coulsen, Proc. Nat.
Acad. Sci. U. S. A. 74: 5463 (1977).
The extensive genetic and physical mapping data
have directed molecular cloning studies to focus on a
small segment of DNA on chromosome 7. Because of the
lack of chromosome deletions and rearrangements in CF
and the lack of a well-developed functional assay for
the CF gene product, the identification of the CF gene
required a detailed characterization of the locus itself
l and comparison between the CF and normal (N) alleles.
- Random, phenotypically normal, individuals could not be
included as controls in the comparison due to the high
` frequency of symptomless carriers in the population. As
a result, only parents of CF patients, each of whom by
definition carries an N and a CF chromosome, were
suitable for the analysis. Moreover, because of the
-l 20 strong allelic association observed between CF and some
~~ of the closely linked DNA markers, it was necessary to
, ~
exclude the possibility that sequence differences
detected between N and CF were polymorphisms associated
with the disease locus.
3.3 IDENTIFICATION OF RFLPs AND FAMILY STUDIES
To determine the relationship of each of the DNA
~"l segments isolated from the chromosome walking and
jumping experiments to CF, restriction fragment length
`~ polymorphisms (~FLPs) were identified and used to study ~ -
families where crossover events had previously been
detected between CF and other flanking DNA markers. As
~ shown in Figure 14, a total of 18 RFLPs were detected in
i`~ the 500 kb region; 17 of them (from E6 to CEl.0) listed
in Table 2; some of them correspond to markers
;~ 35 previously reported. ~ ;~
20112~
44
Five of the RFLPs, namely 10-lX.6, T6/20, Hl.3 and
CE1.0, were identified with cDNA and genomic DNA probes
derived from the putative CF gene. The RFLP data are
presented in Table 2, with markers in the MET and D7S8
regions included for comparison. The physical distances
between these markers as well as their relationship to
the MET and D7S8 regions are shown in Figure 14.
.
.
,: ' ;' '
, . :
.,~ .' ':
. ': ~'.
.-
~ ~ .
.. ,: ~ ~
'``
'
20112~3
r` ~ 1
h ~ ~1 0
o ~ In ~O t~
~1 o o o o ., '
. ~1 o o o o o
:,
~ O
1~1 o o o o o
:1 L
, 1~ ~ .
~ o
I
I
l ~
.~ ~ O ~-1
.1 I
~1 I ~ ~ .
1 ~ " o
d ~
,.,
~ ..
~: ~
.. ~
--- 20112~3
46
.a~ r~
. ~ ~ ~ 1 ~
, ,~ 1~ ~ o ~ ~ ~ ~ gl~
. ~ ~g,
~ m ~Ix ~ ~ x ~ ,i x ~
~, ~ O ~, ~, ~I
; o o O o o o
.'
o O o O o O
~5
:, ~ .
:~ l ~ ~ o o --l ~ co co o ,~
,. j O O ~ -I ~ ~i ~ ~ ~ 'I ~ ' - "': .
H
."'` -
,'
-" 201~253
47
~1 ~ ~1 N N In ~1
O O O O O O O O
æ ~ ~ ~
~, O O O O O O O O
. .
:.
o u~ ~ ~ o In o ~ I
i' N : :~
O
N N 10 ~7 ~ d ' ~ ~ ~1 ~1 ~ 15) N In
H
~ ;;
~-` 201~2~3
48
, H ' --` 3 `~ H
~ ~IG' ';~
i O O O o
O O
. I .
.~ .
''~ O O O ~ -~
1 :
`,L t~l - ,':: .
,'.
:"~
'-i : :
~ ~ ,
"` O :,
; ' U) ~` U~ N co
.~ H . . -.
`, .
1 . :
~'~1 0 ~I . ::
~,
20112~3
49
NOTES FOR TABLE 2
(a) The number of N and CF-PI (CF with pancreatic
insufficiency) chromosomes were derived from the
parents in the families used in linkage analysis
[Tsui et al, Cold Sprinq Harbor Symp. Ouant. Biol.
51:325 (1986)].
~, ,
(b) Standardized association (A), which is less
i influenced by the fluctuation of DNA marker allele
distribution among the N chromosomes, is used here
for the comparison Yule's association coefficient
A=(ad-bc)/(ad+bc), where a, b, c, and d are the
number of N chromosomes with DNA marker allele 1,
'~ CF with 1, N with 2, and CF with 2 respectively.
Relative risk can be calculated using the
~; relationship RR = (l+A)/(l-A) or its reverse.
(c) Allelic association (*), calculated according to
A. Chakravarti et al, Am. J. Hum. Genet. 36:1239,
(1984) assuming the frequency of 0.02 for CF
-~ chromosomes in the population is included for
comparison.
Because of the small number of recombinant families
available for the analysis, as was expected from the
, close distance between the markers studied and CF, and
the possibility of misdiagnosis, alternative approaches -~
were necessary in further fine mapping of the CF gene.
¦ 3.4 ALLELIC ASSOCIATION ~ -
Allelic association (linkage disequilibrium) has ~-
l been detected for many closely linked DNA markers.
'! While the utility of using allelic association for
-¦ measuring genetic distance is uncertain, an overall
-¦ correlation has been observed between CF and the
, 35 flanking DNA markers. A strong association with CF was
j noted for the closer DNA markers, D7S23 and D7S122, ~
." :
. .
:,
2~1~253
whereas little or no association was detected for the
more distant markers MET, D7S8 or D7S424 (see Figure 1).
As shown in Table 2, the degree of association
between DNA markers and CF (as measured by the Yule's
association coefficient) increased from 0.35 for metH
and 0.17 for J32 to 0.91 for 10-lX.6 (only CF-PI
patient families were used in the analysis as they
appeared to be genetically more homogeneous than CF-PS).
The association coefficients appeared to be rather
constant over the 300 kb from EG1.4 to Hl.3; the
fluctuation detected at several locations, most notably
~ at H2.3A, E4.1 and T6/20, were probably due to the
: variation in the allelic distribution among the N
- chromosomes (see Table 2). These data are therefore -~
consistent with the result from the study of recombinant
families (see Figure 14). A similar conclusion could ~-
1 also be made by inspection of the extended DNA marker ~-
haplotypes associated with the CF chromosomes (see ~-~
below). However, the strong allelic association
detected over the large physical distance between EGl.4
and Hl.3 did not allow further refined mapping of the CF
gene. Since J44 was the last genomic DNA clone
isolated by chromosome walking and jumping before a cDNA
clone was identified, the strong allelic association
detected for the JG2E1-J44 interval prompted us to
search for candidate gene sequences over this entire
interval. It is of interest to note that the highest
degree of allelic association was, in fact, detected
; between CF and the 2 RFLPs detected by 10-lX.6, a region -
near the major CF mutation.
Table 3 shows pairwise allelic association between
DNA markers closely linked to CF. The average number of
chromosomes used in these calculations was 75-80 and
' only chromosomes from CF-PI families were used in
scoring CF chromosomes. Similar results were obtained
when Yule's standardized association (A) was used.
: :
:; :
2 ~ 5 3
51
CF chromosorn~s
E ~ ~ 4 ~ ~ E ~
o c~ p ~ ~ ~ p ~o ~ ~ g ~ p~
~ 5 g P ~
~P ~ fi ~ t
g ~ ~ ~ E Ç
.P P ~ fi ~ e ~ iæ z
P ~ fi ~ p ~ ~ I
} i t ~ Ç
P~ t ~ g
,P~ ~ P ~ x
E ~ P Ç a; P
; g ' Ç ~ ~ P ~ Ç ~ ~ U S
o, g P~ Ç
:
.~ . ~ . . .
..i
,. . .
~-............ .
.. .
~ . . ..
--- 20112~3
52
Strong allelic association was also detected among
subgroups of RFLPs on both the CF and N chromosomes. As
shown in Table 3, the DNA markers that are physically
close to each other generally appeared to have strong -
association with each other. For example, strong (in
some cases almost complete) allelic association was
detected between adjacent markers E6 and E7, between
pH131 and W3Dl.4 between the AccI and HaeIII polymorphic
`, sites detected by 10-lX.6 and amongst EG1.4, JG2E1,
E2.6(E.9), E2.8 and E4.1. The two groups of distal
markers in the MET and D7S8 region also showed some
degree of linkage disequilibrium among themselves but
they showed little association with markers from E6 to
CE1.0, consistent with the distant locations for MET and
D7S8. On the other hand, the lack of association
between DNA markers that are physically close may
indicate the presence of recombination hot spots.
Examples of these potential hot spots are the region
between E7 and pHl31, around H2.3A, between J44 and the
' 20 regions covered by the probes 10-lX.6 and T6/20 (see
~ Figure 14). These regions, containing frequent
':`! recombination breakpoints, were useful in the subsequent
analysis of extended haplotype data for the CF region.
' 3.5 HAPLOTYPE ANALYSIS
~; 25 Extended haplotypes based on 23 DNA markers were
generated for the CF and N chromosomes in the collection
~ of families previously used for linkage analysis.
`I Assuming recombination between chromosomes of different
haplotypes, it was possible to construct several
lineages of the observed CF chromosomes and, also, to
predict the location of the disease locus.
~ 1 .
. , '' ~
~ .
--- 20~ 12~3
53
To obtain further information useful for
understanding the nature of different CF mutations, the
F508 deletion data were correlated with the extended DNA
marker haplotypes. As shown in Table 4, five major ~.
5 groups of N and CF haplotypes could be defined by the ~ :
` RFLPs within or im~.ediately adjacent to the putative CF
gene (regions 6-8).
. ~.
' ~:
'
.
'~ :
~1
54 2011253
T~blo 4 DN~ m~k hJp~c~ ~p~ C~s loc~
~n PS~
. I ~t .~ C ~ t ~ o ~_
. t ) A A A "~ 10 1 .
A ~ 3 . . . .
I . . .. .
A .~ ~ A A A . . . o I
A ~ ~ ~A ~ 10 l
.~ A A ~ ~ . . . .
I . . . .
C I . . .
A ~ ~ . . . .
B ~ . A AA A A ~ 1 . . .
B ~ A ~ ~ t .
B ~ . t . . . _
g A A A . J~ . A I . . . _
A ~ ~ A I . . . _
D ~ A ~ I . . .
I . . _
B ~ A ~ I . . . _
C ~ A A ~ A A A ~ . . .. _
B ~ A .~ A . A I . . . _
D ~ ~ A . ~ I . . . _
D ~ ~ A A A A A ~ I . . . _
~ A . ~ A A A A ~ t . . . .
C ~ A A A A ~ I . . . .
: J~ D ~ J~ A A A A ~ I . . . .
D C A A ~ A ~ A I . . .
A D A J~ . A ~ A ~ I . . . .
D D ~ A A ~ 8 . o . . I
.
.
. A A A A ~ I I - .
A ~ ~ . . . .
C ~ I . . . .
C ~ C ~ I .
C 8 A A ~,9 8 i 1
C . ~ A ~ . . . I .
C ~ J~ A ~ A A . . . ~ . .
D C A A J~ A A 8 . . ~ . .
D C ~ A ~ J~ A A D . . . . I
C ~ . . I .
C J~ A A A A A . . 3 .
tc) ~ C A ~ C J~ A D A . . . . I
C J~ C A A D 8 . . ~ . .
~ C A C A A D . . . . t
`` 11 A ~ C A J~ D . . . . I
C ~ ~ D ~ . . . . l
8 C A J~ D . . . . I
D A 8 C ~ . D C . . . . I
D A . . . . I
...
~` ~
: `
.. . .
`. , .. , ,, . ..... . - ~
, .
.
. .,., ~ , .
TABIE 4 (corltinued) 5 5 2 01 1 2 5 3
. .
. ~ ~ .
._ . ._ _
IIL (~ ~ C B ~ A C 11 ~ B I . . . .
2: ^ ~ C ~ ~ .
~ C . ~ ~ C . ,~ . . . .. i
C ~ C A . ~ : :
D C i~ ~ ~ C
C A ~ . . 2
D C ~ C ~ A ~1 ., . . . 2
D ~ C ~ . . I
(c) J~ C ~ ~ i . ~ ~ i
.. . I 0 7 ~ 17 ,
. _ .
A ~ ~ CC C
D ~ C C . . . t
_ . O O O I ~. ' .:
... _ ~ _ -
V.~ C C ~ C ~
O ~ C ~ C ~ C ~ . . . I .
C ~ J C C ~ C D . . . . I
i . . O o a I ~ .
.
~ C ~ . . . . . I
C g ~ S ~
A ~ C ~ . . . . I
D ~ C ~ . . . .
~ C ~ C ~ .... I .. ~
~'
~ .
, -
~, . . . . . .
- . . ....
~; ~
TABIE 4 ( c~ntinu~d) ~ 6 2 0 ~ 1 2 5 3
(4 D a ~ ~ A A A C ~ . . .
a c ~ c ~ A ~ C ~ . . ~ .
~ . ~ ~_ .
- _ ... . _
IL ( ~ B ~~ ~ C 11 . . I .
C . . I
B A ~ 8 11 A ~ ., . . t
: ~ B 11 ~ ~8 ~ C . . I . .
~ ~ a ~ C ~ . O . . 3
.~ C ~ C .~ . . . . l
~ C ~ . C A . . . . I
~ c ~ ~ e A C ~ . . . . I
C ~ A C . . . . I
C ~ . C C . . . . I
C ~ C . . . I .
C ~ A C . . . .
~ c a ~ c ~ . . r ~ l
11 C 1~ C D . . . . I
C ~ C . . ._ . I
~ C B C . . . . I
D C ~ C . . . .
D ~ A C ~ . . . .
~ C ~ A C 11 . . . . I
C C ~ ~~ A C . . . . l
~ A 1~ ~ A C . . . . I
A C . . . . I
? A . ~ C ~ . . . . I
H ~ C . . . . l
A C . . . . l
A ~ C . . . . l
? D A ~ ~~ A C ~ . . . . I
C D A ~ ~~ A C ~ . . . . I
~ D ~ C J~ . . . . I
~ C A ~~ ~ C ~ . . . . ~,
A C ~ A C . . . . t
C ~ G ~ . . . . I
C A ~ ~ A C C . . . .
C A ~ C . . . . l
D ~ . . . . S
C C ~. C ~ . . . I
D ~ A ~C . . . .
D ~ C . . . .
J~ O ~C .~ . . . .
C ~ A U~ ~ . : . . l
J~ C ~ C . . I .
~ C ~ C ~ . . . . I
A C C . . . . l
D A ~ ~ A C C . . . I .
C 1~ . . . . l
C ~ -~ C . . . .
J~ ~ A ~ - ac A C ~ . . . . I
~ C ~ . . I
" ~ C . ~ . . I . .
D ~ . . . . I
."
.
i
~' :
T~BIE 4 ( continued)
a a ~ D B O ~ ~ l 2 01 12
_~ ., .
. _ .
......... _ ~1l02l~ ~
Tolal: ~ 9~ _
(a) The extended haplot~pe data are derived fran ~e CF faruilies used
in previous lirJcage studies (see footrY~te (a) of Table 3) with
a~ditional ~-PS fanilies collected subs~uently (Kerem et al, ~ . J.
Genet. 44:827 (1989)). me data are shawn in gmup (regions) to reduoe
space. The regio3~ are assigne~ ~r~marily accc~ding to pairwise
assoc~ation data sha~n in Tab~e3 with regior~ 6-8 spanning the putative
CF locus (the F508 deletion i3 between regions 6 and 7). A dash (-) is
sh~n at the region where the haplotype has not been dete~mined due to
inc~nplete data ar inability to establi~h phase. Alternative haplotype
assignnent~ are also given where data are incomplete. Unclas~ified
includes those dhromcccn#- with more than 3 unknown assignmentQ. The
haplotype definitions for eadh of the 9 regicns are:
~ . ~.
A I I I
C
D
1~ Q I a -
P ~
~L~
S
C
D
P ~
H ~ I a
A ~
R~. ECIA I~OIA Ja~l
.
...... . .
. ,
., . . ~ ..
. .
`;. ~ , .
..
.
.. . .
,-- . , ' - ' ' '
.
T~ 4 (^ontinued) 5 8
B 201125~)
C 1- 2 2 2
D ~ I I t
l a
~ L~L~
A 1 l 2
~ . l a
c . a 2
Rc~ ~ J4~ 141X.
1 2
B ~ 2 1 2
C
D
~ . ~ a ~
P 2 ~ I
B a
R~ 1~. HW C~lt.O
~ .
a
C
J3a ~1
~eLY-L~
A I I I
~ . ~ a :~ -
D
It
O
N o
... .
- ~ .
,:; . :
. .,~ , . ... . . , . ",: .
., :
2~12~
59
It was apparent that most recombinations between
haplotypes occurred between regions 1 and 2 and between
regions 8 and 9, again in good agreement with the
relatively long physical distance between these regions.
Other, less frequent, breakpoints were noted between
short distance intervals and they generally corresponded
to the hot spots identified by pairwise allelic
association studies as shown above. It is of interest
to note that the F508 deletion associated almost
exclusively with Group I, the most frequent CF
haplotype, supporting the position that this deletion
constitutes the major mutation in CF. More important,
; while the F'508 deletion was detected in 89% (62/70) of
the CF chromosomes with the AA haplotype (corresponding
to the two regions, 6 and 7) flanking the deletion, none
was found in the 14 N chromosomes within the same group
(X2 = 47.3, p <10-4). The F508 deletion was therefore
not a common sequence polymorphism associated with the
core of the Group I haplotype (see Table 5).
Together, the results of the oligonucleotide
hybridization study and the haplotype analysis support
the fact that the gene locus described here is the CF
gene and that the 3 bp (F508) deletion is the most
common mutation in CF.
;~ 25 3.6 INTRON/EXON BOUNDARIES
Genomic CF gene includes all of the regulatory
genetic information as well as intron genetic
~; information which is spliced out in the expression of
the CF gene. It has been found that portions of the
introns at the intron/exon boundaries for the exons of
` the CF gene are very helpful in not only locating
i mutations in the CF gene, but are also very useful in
PCR analysis. Such intron information can be employed
in PCR analysis for purposes of CF screening which will
be discussed in more detail in a later section. As set
out in Figure 18 with the headings "Exon 4 and 6 through
.... . ~ ~
~;
~. ,
:,~.' . ;
: ,,: , ,
20112~3
~ Exon 24", there are portions of the bounding introns
- which are particularly preferred in PCR exon
amplification.
In sequencing the intron portions, it has been
determined that there are at least 28 exons instead of
the previously reported 24 exons in applicants'
aforementioned co-pending applications. Exons 6, 14 and
17, as previously reported, are found to be in segments
and are now named exons 6a, 6b, 6c and 14a and 14b and
exons 17a and 17b.
The intron portions, which have been used in PCR
amplification, are identified by arrows in Figure 18.
The portions identified by the arrows are preferred, but
it is understood that other portions of the intron
sequences are also useful in the PCR amplification
; technique. For example, for exon 9 the relevant genetic
information which is preferred in PCR is noted by
` reference to the 5' and 3' ends of the sequence. The
intron section is identified with an "i". Hence in
Figure 18 for exon 9, the preferred portions are
` identified by the arrows 9i-5 and 9i-3. They include
; the sequence TAA... TGT for 9i-5 and for 9i-3 TGT... CGT.
j~ Similarly, the preferred portions of the intron for exon
10 is identified as lOi-5 and lOi-3 and similarly for
` 25 exons 11 through 24. Similar regions are also
identified and/or can be extracted from the bordering
intron regions for exons 4 and 6 through 8 of Figure 18.
` These oligonucleotides, as derived from the intron
~`` sequence, assist in amplifying by PCR the respective ~
30 exon, thereby providing for analysis for DNA sequence -
alterations corresponding to mutations of the CF gene.
The mutations can be revealed by either direct sequence
determination of the PCR products or sequencing the
products cloned in plasmid vectors. The amplified exon
can also be analyzed by use of gel electrophoresis in
~- the manner to be further described. It has been found
` that the sections of the intron for each respective exon
:' ''
20~12~
- are of sufficient length to work particularly well with
PCR technique to provide for amplification of the
relevant exon.
3.7 CF MUTATIONS - ~I506 OR AI507
The association of the F508 deletion with 1 common
and 1 rare CF haplotype provided further insight into
the number of mutational events that could contribute to
the present patient population. Based on the extensive
haplotype data, the original chromosome in which the
F508 deletion occurred is likely to carry the haplotype
- AAAAAAA- (Group Ia), as defined in Table 4. The other
Group I CF chromosomes carrying the deletion are
probably recombination products derived from the
original chromosome. If the CF chromosomes in each
haplotype group are considered to be derived from the
` same origin, only 3-4 additional mutational events would
be predicted (see Table 4). However, since many of the
CF chromosomes in the same group are markedly different
from each other, further subdivision within each group
....
is possible. As a result, a higher number of
independent mutational events could be considered and
the data suggest that at least 7 additional, putative
mutations also contribute to the CF-PI phenotype (see
Table 3). The mutations leading to the CF-PS subgroup
are probably more heterogeneous.
`;3~ The 7 additional CF-PI mutations are represented by
the haplotypes: -CAAAAAA- (&roup Ib), -CABCAAD- (Group
Ic), ---BBBAC- (Group IIa), -CABBBAB- (Group Va).
Although the molecular defect in each of these mutations
has yet to be defined, it is clear that none of these
`~ mutations severely affect the region corresponding to
.
` the oligonucleotide binding sites used in the
PCR/hybridization experiment.
~i One CF chromosome hydridizing to the ~F508-ASO
probe, however, has been found to associate with a
different haplotype (group IIIa). It appeared that the
.
:~,
.`.
~ ¢,- , ~
2~112~
62
~I508 should have appeared in both haplotypes, but with
the discovery of ~I507, it is not discovered that it is
not. Instead, the QI507 is in group I, whereas the
~I507 is in group IIIa. None of the other CF nor the
5 normal chromosomes of this hoplotype group have shown
hybridization to the mutant (~F508) ASO [B. Kerem et al,
Science 245:1073 (1989)]. In view of the group Ia and
IIIa haplotypes being distinctly different from each
R", other, the mutations harbored by these two groups of CF
10 chormosomes must have originated independently. To
- investigate the molecular nature of the mutation in this
` group IIIa CF chromosome, we further characterized the
region of interest through amplification of the genomic
DNA from an individual carrying the chromosome IIIa by
15 the polymerase chain reaction (PCR).
These polymerase chains reactions (PCR) were
performed according to the procedure of R.k. Saiki et al
Science 230:1350 (1985). A specific DNA segment of 491
bp including exon 10 of the CF gene was amplified with
~ 20 the use of the oligonucleotide primers lOi-5 (5'-
q GCAGAGTACCTGAAACAGGA-3') and lOi-3
(5'ATGGGTAAGCTACTGTGAATG-3')1Ocated in the 5' and 3'
flanking regions, respectively, as shown in Figure 18.
, Both oligonucleotides were purchased from the HSC DNA
25 Biotechnology Service Center (Toronto). Approximately
500 ng of genomic DNA from cultured lymphoblastoid cell
lines of the parents and the CF child of Family 5 were
used in each reaction. The DNA samples were denatured
~` at 94C for 30 sec., primers annealed at 55C for 30
`~ 30 sec., and extended at 72C for 50 sec. (with 0.5 unit of
Taq polymerase, Perkin-Elmer/Cetus, Norwalk, CT) for 30
sycles and a final extension period of 7 min. in a
Perkin-Elmer/Cetus DNA Thermal Cycler.
Hydridization analysis of the PCR products from ;~
three individuals of Family 5 of group IIIa was
performed. The carrier mother and father are
~'
20~12~3
63
represented by a half-filled circle and square,
respectively, and the affected son is a filled square in
- Figure l9a. The conditions for hybridizaton and washing
have been previously described (Kerem et al, supra).
There is a relatively weak signal in the father's PCR
product with the mutant (oligo QF508) probe. In Figure
l9c, DNA sequence analysis of the clone 5-3-15 and the
PCR products from the affected son and the carrier
father. The arrow in the center panel indicates the
` 10 presence of both A and T nucleotide residue in the same
position; the arrow in the right panel indicates the
points of divergence between the normal and the ~I507
sequence. The sequence ladders shown are derived from
the reverse-complements as will be described later.
Figure l9b shown the DNA sequences and their
corresponding amino acid sequences of the normal ~I507
~, and ~I508 alleles spanning the mutation sites are
j shown. With reference to Figure l9a, the PCR-amplified
DNA from the carrier father, who contributed the group
IIIa CF chromosome to the affected son, hybridized less
, efficiently with the ~F50~ ASO than that from the mother
who carried the group Ia CF chromosome. The difference
` became apparent when the hybridization signals were
compared to that with the normal ASO probe. This result
therefore indicated that the mutation carried by the
group IIIa CF chromosome might not be identical to
~F508.
To define the nucleotide sequence corresponding to
the mutant allele on this chromosome, the PCR-amplified
product of the father's DNA was excised from a
polyacrylamide-electrophoretic gel and cloned into a -
sequencing vector.
The general procedures for DNA isolation and
purification for purposes of cloning into a sequencing
vector are described in J. Sambrook, E.F. Fritsch, T.
Maniatis, Molecular Cloninq: A Laboratory Manual, 2nd
.,
.~,
; - 20~12~3
64
ed. (Cold Spring Harbor Press, N.Y. 1989). The two
homoduplexes generated by PCR amplification of the
paternal DNA were purified from a 5% non-denaturing
polyarylamide gel (30:1 acrylamide:bis-acrylamide). The
appropriate bands were visualized by staining with
ethidium bromide, excised and eluted in TE (lo mM Tris-
; HCl; lmM EDTA; pH 7.5) for 2 to 12 hours at room
temperature. The DNA solution was sequentially treated
with Tris-equilibrated phenol, phenol/CHC13 and CHC13.
~-~ 10 The DNA samples were concentrated by precipitation in
ethanol and resuspension in TE, incubated with T4
polynucleotide kinase in the presence of ATP, and
ligated into diphosphorylated, blunt-ended Bluescript
KS~ vector (Stratagene, San Diego, CA). Clones
containing amplified product generated from the normal
parental chromosome was identified by hybridization with ~
the oligonucleotide N as described in Kerem et al ~-
`I~ su~ra.
Clones containing the mutant sequence were
identified by their failure to hybridize to the normal
ASO (Kerem et al, supra). One clone, 5-3-15 was
isolated and its DNA sequence determined. The general
~; protocol for sequencing cloned DNA is essentially as
` described [J.R. Riordan et al, Science 245:1066 (1989)]
I 25 with the use of an U.S. Biochemicals Sequenase~ kit. To
-` verify the sequence and to exclude any errors introduced
by DNA polymerase during PCR, the DNA sequences for the
PCR products from the father and one of the affected
children were also determined directly without cloning.
This procedure was accomplished by denaturing 2
pmoles of gel-purified double-stranded PCR product in
0.2 M NaOH/0.2 mM EDTA (5 min. at room temperature),
neutralized by adding 0.1 volume of 2 M ammonium acetate
(pH 5.4) and precipitated with 2.5 volumes of ethanol at
-70C for 10 min. After washing with 70% ethanol, the
DNA pellet was dried and redissolved in a sequencing
.
.
- 20~1~53
reaction buffer containing 4 pmoles of the
oligonucleotide primer lOi-3 of Figure 18,
dithiothreitol (8.3 mM) and [~-35S]-dATP (0.8 ~M, 1000
Ci/mmole). The mixture was incubated at 37C for 20
min., following which 2 ~1 of labelling mix and then 2
units of Sequenase were added. Aliquotes of the
reaction mixture (3.5 ~1) were transferred, without
delay, to tubes each containing 2.5 ~1 of ddGTP, ddATP,
ddTTP and ddCTP solutions (U.S. Biochemicals Sequenase
kit) and the reactions were stopped by addition of the
stop solution.
The DNA sequence for this mutant allele is shown in
` Figure l9b. The data derived from the cloned DNA and
direct sequencing of the PCR products of the affected
child and the father are all consistent with a 3 bp
~ deletion when compared to the normal sequence (Figure
i! l9c). The deletion of this 3 bp (ATC) at the I506 or
I507 position results in the loss of an isoleucine
residue from the putative CFTR, within the same ATP-
binding domain where ~F508 resides, but it is notevident whether this deleted amino acid corresponds to
the position 506 or 507. Since the 506 and 507
positions are repeats, it is at present impossible to
~ determine in which position the 3 bp deletion occurs.
; 25 For convenience in later discussions, however, we refer
to this deletion as QI507.
The fact that the ~I507 and ~I508 mutations occure
in the same region of the presumptive ATP-binding domain
of CFTR is surprising. Although the entire sequence of
~I507 allele has not been examined, as has been done for
~F508, the strategic location of the deletion argues
that it is the responsible mutation for this allele.
This argument is further supported by the observation
that this alteration was not detected in any of the
` 35 normal chromosomes studied to date tKerem et al, supra).
The identification of a second single amino acid
:. ...
20~12~3
66
deletion in the ATP-binding domain of CFTR also provides
information about the structure and function of this
protein. Since deletion of either the phenylalanine
residue at position 508 or isoleucine at position ~I507
is sufficient to affect the function of CFTR, it is
suggested that these residues are involved in the
folding of the protein but not directly in the binding
of ATP. That is, the length of the peptide is probably
more important than the actual amino acid residues in
this region. In support of this hypothesis, it has been
found that the phenylalanine residue can be replaced by
a serine and that isoleucine at position 506 with
valine, without apparent loss of function of CFTR. ~-
When the nucleotide sequence of ~I507 is compared
to that of ~F508 at the ASO-hybridizing region, it was
noted that the difference between the two alleles was
only an A T change (Figure l9c). This subtle
difference thus explained the cross-hybridization of
the ~F508-ASO to AI507. These results therefore ~ -
exemplified the importance of careful examination of
both parental chromosomes in performing ASO-based
genetic diagnosis. It has been determined that the
~F508 and ~I507 can be distinguished by
increasing the stringency of oligonucleotide
hybridization condition or by detecting the unique
mobility of the heteroduplexes formed between each of
these sequences and the normal DNA on a polyacrylaminde
gel. The stringency of hybridization can be increased
by using a washing temperature at 45 D C instead of the
prior more mild 39C in the presence of 2XSSC (150 mM
NaCL/15 mM Na citrate).
Identification of the ~I507 and ~I508 alleles by
polyacrylamide gel electrophoresis is shown in Figure
20. The PCR products were prepared from the three
` 35 family members and separated on a 5% polyacrylamide gel
as described above. A DNA sample from a known
: 2~11253
67
heterozygous ~Iso8 carrier is included for comparison.
With reference to Figure 20, the banding pattern of the
PCR-amplified genomic DNA from the father, who is the
carrier of ~I507, is clearly distinguishable from that
of the mother, who is of the type of carriers with the
~F508 mutation. In this gel electrophoresis test, there
were actually three individuals (the carrier father and
the two affected sons in Family 5) who carried the ~I507
~` deletion. Since they all belong to the same family,
they only represent one single CF chromosome in our
population analysis [Kerem et al, supra] The two
patients who also inherited the ~F508 mutation from
~; their mother showed typical symptoms of CF with
pancreatic insufficiency. The father of this family was
the only parent who carries this ~I507 mutation; no
other CF parents showed reduced hybridization intensity
; signal with the ~F508 mutant oligonucleotide probe or a
peculiar heteroduplex pattern for the PCR product (as
defined above) in the retrospective study. In addition,
two representatives of the group IIIb and one of the
group IIIc CF chromosomes from our collection [Kerem et
al, supra] were sequenced, but none were found to
contain QI507. Since the electrophoresis technique
eliminates the need for probe-labelling and
hybridization, it may prove to be the method of choice
for detecting carriers in a large population scale.
, The present data also indicate that there is a
strict correlation between DNA market haplotype and
mutation in CF. The ~F508 deletion is the most common
CF mutation that occurred on a group Ia chromosome
background [Kerem et al, supra]. The QI507 mutation is,
however, rare in the CF population; the one group IIIa
CF chromosome carrying this deletion is the only example -
in our studied population (1/219) [J. M. Rommens et al, ;
Am. J. Hum. Genet. in press (1990)]. Since the group
III haplotype is relatively common among the normal -
~,.,.:
'. ' ' '
~'A'~.?
2011253
68
chromosomes (17/198), the ~Iso7 deletion probably
occurred recently. Additional studies with larger
populations of different geographic and ethnic
backgrounds should provide further insight in
understanding the origins of these mutations.
3.8 ADDITIONAL CF MUTATIONS
Fol~owing the above procedures, other mutations in
the CF gene have been identified. The following brief
description of each identified mutation is based on the
previously described procedures using PCR procedures to
locate the mutation.
3.8.1 MU~ATIONS IN EXON 9
A mutation in exon 9 is a change of alanine (GCG) ~
to glutamic acid (GAG) at amino acid position 455 ~ -
15 (A455 ~ E). Two of the 38 non-~F508 CF chromosomes
examined carries this mutation; both of them are from
patients of a French-Canadian origin, which we have
identified in our work as families #27 and #53, and they
belong to haplotype group Ib. The mutation is
detectable by allele-specific oligonucleotide (ASO)
hybridization with PCR-amplified genomic DNA sequence.
The PCR primers are 91-5 (5'-TAATGGATCATGGGCCATGT-3')
and 9i-3 t5'-ACAGTGTTGDDTGTGGTGCA-3'). The ASOs are 5'-
GTTGTTGGCGGTTGCT-3' for the normal allele and 5'-
GTTGTTGGAGGTTCGT-3' for the mutant. The oligonucleotide
hybridization is as described in Kerem et al (1989)
supra at 37C and the washings are done twice with 5XSSC
for 10 min each at room temperature followed by twice
with 2 X SSC for 30 min each at 52C. Although the
alanine at position 455 (Ala455) is not present in all
; ATP-binding folds across species, it is present in all
known members of the P-glycoprotein family. Further,
A455 ~ E is believed to be a mutation rather than a
sequence polymorphism because the change is not found in
16 non-~F508 CF chromosomes and three normal chromosomes
carrying the same group I haplotype.
' ,.
'~r,!': ~ : , -
20~2~
69
3.8.2 MUTATIONS IN EXON 11
The glycine codon (GGA) at amino acid position 542
is changed to a stop codon (TGA) (G542 ~ Stop). The
single chromosome carrying this mutation is of
Ashkenazic Jewish origin (family A) and has the B
haplotype (XV2C allele l; KM.l9 allele 2). The mutant
sequence can be detected by hybridization analysis with
allele-specific oligonucleotides (ASOs) on genomic DNA
amplified by PCR with the lli-5 and lli-3
oligonucleotide primers. The normal ASO is 5'-
ACCTTCTCCAAGAACT-3' and the mutant ASO, 5'-
ACCTTCTCAAAGAACT-3'. The oligonucleotide hybridization
condition is as described in Kerem et al (1989) supra
and the washing conditions are twice in 5 x SSC for 10
min. each at room temperature followed by twice in 2 X
SSC for 30 min. each at 45C. The mutation is not
detected in 52 other non-~F508 CF chromosomes, 11 of
which are of Jewish origin (three have a B haplotype),
nor in 13 normal chromosomes.
The highly conserved serine residue at position 549
is changed to arginine (S549 ~ R); the codon change is
AGT ~ AGG. The CF chromosome with this mutation is
carried by a non-Ashkenazic Jewish in Morocco (family
i B). The chromosome also has the B haplotype. Detection
!, 25 of this mutation may be achieved by ASO hybridization or
allele-specific PCR. In the ASO hybridization
procedure, the genomic DNA sequence is first amplified
by PCR with the lli-5 and lli-3 oligonucleotides; the
, ASO for the normal sequence is 5'~ACACTGAGTGGAGGTC-3'
and that for the mutant is 5'-ACACTGAGGGGAGGTC. The
oligonucleotide hybridization condition is as described
by Kerem et al (1989) supra and the washings are done
twice in 5 x SSC for 10 min. each at room temperature
followed by twice in 2 x SSC for 30 min. eachat 56C.
- 35 In the allele-specific PCR amplification, the
~ oligonucleotide primer for the normal sequence is
~ ' ~
. ~
~ -~` 20112~3
.
` 70
5'TGCTCGTTGACCTCCA-3', that for the mutant is
5'TGCTGCTTGACCTCCC-3' and that for the common, outside
sequence is lli-5. The reaction is performed with 500
ng of genomic DNA, 100 ng of each of the
oligonucleotides and 0.5 unit of Taq polymerase. The
DNA template is first denatured by heating at 94C for 6
min., followed by 30 cycles of 94 for 30 sec, 55 for
~, 30 sec and 72 for 60 sec. The reaction is completed by
-~ a 6 min heating at 72 for 7 min. This S549 ~ R
mutation is not present in 52 other non-~F508 CF
chromosomes, 11 of which are of Jewish origin (three
have a B haplotype), nor in 13 normal chromosomes.
The arginine (AAG) at amino aid position 560 is
changed to threonine (AAC). The individual carrying
this mutation (R560 ~ T) is from a family we have
identified in our work as family #32 and the chromosome
is marked by haplotype IIIb. The mutation creates a
Maell site which cleaves the PCR product of exon 11
(generated with primers lli-5 and lli-3) into two
fragments of 214 and 204 bp in size. None of the 36
' non-QF508 CF chromosomes (seven of which have haplotype
~ IIIb) or 23 normal chromosomes (16 have haplotype IIIb)
! carried this sequence alteration. The R560 ~ T mutation
, is also not present on eight CF chromosomes with the
-~ 25 ~F508 mutation.
: I .
4.0 CFTR PROTEIN
~i As discussed with respect to the DNA sequence of
Figure 1, analysis of the sequence of the overlapping
cDNA clones predicted an unprocessed polypeptide of 1480
amino acids with a molecular mass of 168,138 daltons.
`~ As later described, due to polymorphisms in the protein,
. the molecular weight of the protein can vary due to
possible substitutions or deletion of certain amino
;~ acids. The molecular weight will also change due to the
addition of carbohydrate units to form a glycoprotein.
It is also understood that the functional protein in the
.,; , :
; . :
:
- 20112~3
71
cell will be similar to the unprocessed polypeptide, but
may be modified due to cell metabolism.
Accordingly, purified normal CFTR polypeptide is
characterized by a molecular weight of about 170,000
daltons and having epithelial cell transmembrane ion
conductance activity. The normal CFTR polypeptide,
which is substantially free of other human proteins, is
encoded by the aforementioned DNA sequences and
according to one embodiment, that of Figure 1. Such
polypeptide displays the immunological or biological
activity of normal CFTR polypeptide. As will be later
discussed, the CFTR polypeptide and fragments thereof
may be made by chemical or enzymatic peptide synthesis
or expressed in an appropriate cultured cell system.
The invention provides purified 507 mutant CFTR
polypeptide which is characterized by cystic fibrosis-
, associated activity in human epithelial cells. Such 507
mutant CFTR polypeptide, as substantially free of other
human proteins, can be encoded by the 507 mutant DNA
sequence.
4.1 STRUCTURE OF CFTR
The most characteristic feature of the predicted
~, protein is the presence of two repeated motifs, each of ~;
which consists of a set of amino acid residues capable -~
of spanning the membrane several times followed by
sequence resembling consensus nucleotide (ATP)-binding
folds (NBFs) (Figures 11, 12 and 15). These -~-
characteristics are remarkably similar to those of the
mammalian multidrug resistant P-glycoprotein and a
` 30 number of other membrane-associated proteins, thus
implying that the predicted CF gene product is likely to
, be involved in the transport of substances (ions) across
the membrane and is probably a member of a membrane
~3 protein super family.
Figure 13 is a schematic model of the predicted
. CFTR protein. In Figure 13, cylinders indicate membrane
.. ~
''. ' ,
' 1
.
201~253
72
spanning helices, hatched spheres indicate NBFs. The
stippled sphere is the polar R-domain. The 6 membrane
spanning helices in each half of the molecule are
depicted as cylinders. The inner cytoplasmically
oriented NBFs are shown as hatched spheres with slots to
indicate the means of entry by the nucleotide. The
large polar R-domain which links the two halves is
represented by an stippled sphere. Charged individual
amino acids within the transmembrane segments and on the
R-domain surface are depicted as small circles
containing the charge sign. Net charges on the internal
-' and external loops joining the membrane cylinders and on
regions of the NBFs are contained in open squares.
Sites for phosphorylation by protein kinases A or C are
shown by closed and open triangles respectively.
K,R,H,D, and E are standard nomenclature for the amino
~, acids, lysine, arginine, histidine, aspartic acid and
glutamic acid respectively.
Each of the predicted membrane-associated regions
l 20 of the CFTR protein consists of 6 highly hydrophobic
:! segments capable of spanning a lipid bilayer according
to the algorithms of Kyte and Doolittle and of Garnier
et al (J. Mol. Biol. 120, 97 (1978) (Figure 13). The
¦ membrane-associated regions are each followed by a large -~
¦ 25 hydrophilic region containing the NBFs. Based on
sequence alignment with other known nucleotide binding
proteins, each of the putative NBFs in CFTR comprises at
least 150 residues (Figure 13). The 3 bp deletion at
position 507 as detected in CF patients is located
between the 2 most highly conserved segments of the
first NBF in CFTR. The amino acid sequence identity
between the region surrounding the isoleucine deletion -;
, and the corresponding regions of a number of other
proteins suggests that this region is of functional
importance (Figure 15). A hydrophobic amino acid,
usually one with an aromatic side chain, is present in
::
:' :
~3~- - -
201~53
73
most of these proteins at the position corresponding to
I507 of the CFTR protein. It is understood that amino
acid polymorphisms may exist as a result of DNA
' polymorphisms. Similarly, mutations at the other
positions in the protein suggested that corresponding
regions of the protein are also of functional
importance. Such additional mutations include
substitutions of:
i) glutamic acid (TGA) for alanine (GCG) at amino
position 455;
ii) stop codon (TGA) for glycine (GGA) at amino
acid position 542;
iii) arginine (AGG) for serine (AGT) at amino acid -
` position 549;
iv) threonine (ACC) for arginine (AAG) at amino
acid position 560.
Figure 15 shows alignment of the 3 most conserved
segments of the extended NBF's of CFTR with comparable
regions of other proteins. These 3 segments consist of
20 residues 433-473, 488-513, and 542-584 of the N-terminal
l half and 1219-1259, 1277-1302, and 1340-1382 of the C-
-:! terminal half of CFTR. The heavy overlining points out
~ the regions of greatest similarity. Additional general
j' homology can be seen even without the introduction of
`' 25 gaps.
' Despite the overall symmetry in the structure of
the protein and the sequence conservation of the NBFs,
,`, sequence homology between the two halves of the
` predicted CFTR protein is modest. This is demonstrated ~
; 30 in Figure 12, where amino acids 1-1480 are represented -
on each axis. Lines on either side of the identity
; diagonal indicate the positions of internal
similarities. Therefore, while four sets of internal
sequence identity can be detected as shown in Figure
' 35 12, using the Dayhoff scoring matrix as applied by -~
Lawrence et al. [C. B. Lawrence, D. A. Goldman~ and R.
, ~
20~12~
74
T. Hood, Bull Math Biol. 48, 569 (1986)], three of
these are only apparent at low threshold settings for
standard deviation. The strongest identity is between
sequences at the carboxyl ends of the NBFs. Of the 66
residues aligned 27% are identical and another 11% are
functionally similar. The overall weak internal
homology is in contrast to the much higher degree (>70%)
in P-glycoprotein for which a gene duplication
hypothesis has been proposed (Gros et al, Cell 47, 371,
1986, C. Chen et al, Cell 47, 381, 1986, Gerlach et al,
ature, 324, 485, 1986, Gros et al, Mol. Cell. Biol. 8,
2770, 1988). The lack of conservation in the relative
positions of the exon-intron boundaries may argue
against such a model for CFTR (Figure 2).
Since there is apparently no signal-peptide
sequence at the amino-terminus of CFTR, the highly
~ charged hydrophilic segment preceding the first
`~ transmembrane sequence is probably oriented in the
cytoplasm. Each of the 2 sets of hydrophobic helices
20 are expected to form 3 transversing loops across the ~
membrane and little sequence of the entire protein is ~ ;
expected to be exposed to the exterior surface, except
the region between transmembrane segment 7 and 8. It is
of interest to note that the latter region contains two
potential sites for N-linked glycosylation.
Each of the membrane-associated regions is followed
by a NBF as indicated above. In addition, a highly
charged cytoplasmic domain can be identified in the
middle of the predicted CFTR polypeptide, linking the 2
halves of the protein. This domain, named the R-domain,
; is operationally defined by a single large exon in which
69 of the 241 amino acids are polar residues arranged in
alternating clusters of positive and negative charges.
Moreover, 9 of the 10 consensus sequences required for
phosphosphorylation by protein kinase A (PKA), and, 7 of
, ~
.' ~
- 201~2~3
the potential substrate sites for protein kinase C (PKC)
found in CFTR are located in this exon.
4 . 2 E'UNCTION OF CFTR
, Properties of CFTR can be derived from comparison
to other membrane-associated proteins (Figure 15). In
addition to the overall structural similarity with the
.f mammalian P-glycoprotein, each of the two predicted
domains in CFTR also shows remarkable resemblance to the
' single domain structure of hemolysin B of E. coli and -
- 10 the product of the White gene of Drosophila. These
latter proteins are involved in the transport of the
lytic peptide of the hemolysin system and of eye pigment
molecules, respectively. The vitamin B12 transport
system of E. coli, BtuD and MbpX which is a liverwort
chloroplast gene whose function is unknown also have a
similar structural motif. Furthermore, the CFTR protein ~;
shares structural similarity with several of the
periplasmic solute transport systems of gram negative
bacteria where the transmembrane region and the ATP-
~- 20 binding folds are contained in separate proteins which
function in concert with a third substrate-binding
`~if polypeptide.
.. . .
The overall structural arrangement of the
' transmembrane domains in CFTR is similar to several
-i 25 cation channel proteins and some cation-translocating
ATPases as well as the recently described adenylate
cyclase of bovine brain. The functional significance of
l this topological classification, consisting of 6
`~ transmembrane domains, remains speculative.
Short regions of sequence identity have also been
detected between the putative transmembrane regions of
`! CFTR and other membrane-spanning proteins.
Interestingly, there are also sequences, 18 amino acids
in length situated approximately 50 residues from the
carboxyl terminus of CFTR and the raf serine/threonine
.. :
..~
,. ' .
.~ :
.. - , ~.. : :::: . ~
2~253
76
kinase protooncogene of Xenopus laevis which are
identical at 12 of these positions.
Finally, an amino acid sequence identity (10/13
conserved residues) has been noted between a hydrophilic
segment (position 701-713) within the highly charged R-
domain of CFTR and a region immediately preceding the
first transmembrane loop of the sodium channels in both
rat brain and eel. The charged R-domain of CFTR is not -~
shared with the topologically closely related P-
glycoprotein; the 241 amino acid linking-peptide is
apparently the major difference between the two
proteins.
In summary, features of the primary structure of
the CFTR protein indicate its possession of properties
suitable to participation in the regulation and control
of ion transport in the epithelial cells of tissues
affected in CF. Secure attachment to the membrane in
two regions serve to position its three major
intracellular domains (nucleotide-binding folds 1 and 2
and the R-domain) near the cytoplasmic surface of the
cell membrane where they can modulate ion movement
through channels formed either by CFTR transmembrane
segments themselves or by other membrane proteins.
In view of the genetic data, the tissue-
specificity, and the predicted properties of the CFTRprotein, it is reasonable to conclude that CFTR is
directly responsible for CF. It, however, remains
unclear how CFTR is involved in the regulation of ion
conductance across the apical membrane of epithelial
cells.
It is possible that CFTR serves as an ion channel
itself. As depicted in Figure 13, 10 of the 12
transmembrane regions contain one or more amino acids
with charged side chains, a property similar to the
brain sodium channel and the GABA receptor chloride
channel subunits, where charged residues are present in
`: :
,
.. :
... . . . .
2011253
77
4 of the 6, and 3 of the 4, respective membrane-
associated domains per subunit or repeat unit. The
amphipathic nature of these transmembrane segments is
believed to contribute to the channel-forming capacity
of these molecules. Alternatively, CFTR may not be an
ion channel but instead serve to regulate ion channel
activities. In support of the latter assumption, none
of the purified polypeptides from trachea and kidney
that are capable of reconstituting chloride channels in
lipid membranes [Landry et al, Science 224:1469 (1989)]
appear to be CFTR if judged on the basis of the
molecular mass.
In either case, the presence of ATP-binding domains
in CFTR suggests that ATP hydrolysis is directly
involved and required for the transport function. The
I high density of phosphorylation sites for PKA and PKC -
¦ and the clusters of charged residues in the R-domain may
both serve to regulate this activity. The deletion of a -~
phenylalanine residue in the NBF may prevent proper
binding of ATP ~r the conformational change which this
normally elicits and consequently result in the
observed insensitivity to activation by PKA- or PKC-
' mediated phosphorylation of the CF apical chloride
conductance pathway. Since the predicted protein
'25 contains several domains and belongs to a family of
proteins which frequently function as parts of multi-
component molecular systems, CFTR may also participate
'in epithelial tissue functions of activity or
regulation not related to ion transport.
With the isolated CF gene (cDNA) now in hand it is
possible to define the basic biochemical defect in CF
and to further elucidate the control of ion transport
pathways in epithelial cells in general. Most
important, knowledge gained thus far from the predicted
structure of CFTR together with the additional
information from studies of the protein itself provide a
'
J ~ j . ~ . .
20~12~3
78
basis for the development of improved means of treatment
of the disease. In such studies, antibodies have been
, raised to the CFTR protein as later described.
5.0 CF SCREENING
5.1 DNA BASED DIAGNOSIS
Given the knowledge of the 455, 507, 542, 549 and ~-~
560 mutation as disclosed herein, carrier screening and
prenatal diagnosis can be carried out as follows. -
The high risk population for cystic fibrosis is
Caucasians. For example, each Caucasian woman and/or
man of child-bearing age would be screened to determine
if she or he was a carrier (approximately a 5%
probability for each individual). If both are carriers,
they are a couple at risk for a cystic fibrosis child.
Each child of the at risk couple has a 25% chance of
being affected with cystic fibrosis. The procedure for
i determining carrier status using the probes disclosed
herein is as follows.
i For purposes of brevity, the discussion on
screening by use of one of the selected mutations is
directed to the I507 mutation. It is understood that
screening can also be accomplished using one of the
other mutations or using several of the mutations in a
,l screening process or mutation detection process of this
.
section on CF screening involving DNA diagnosis and
mutation detection.
One major application of the DNA sequence
~ information of the normal and 507 mutant CF gene is in
i~ the area of genetic testing, carrier detection and
prenatal diagnosis. Individuals carrying mutations in
the CF gene (disease carrier or patients) may be
detected at the DNA level with the use of a variety of
`~ techniques. The genomic DNA used for the diagnosis may
be obtained from body cells, such as those present in
35 peripheral blood, urine, saliva, tissue biopsy, surgical -
specimen and autopsy material. The DNA may be used
,
.
. ~
~ , ,, ~, .
~,. - .
201~253
79
directly for detection of specific sequence or may be
amplified enzymatically in vitro by using PCR [Saiki et
al. Science 230: 1350-1353, (1985), Saiki et al. Nature
324: 163-166 (1986)] prior to analysis. RNA or its cDNA
form may also be used for the same purpose. Recent
reviews of this subject have been presented by Caskey,
[Science 236: 1223-8 (1989) and by Landegren et al
(Science 242: 229-237 (1989)].
The detection of specific DNA sequence may be
achieved by methods such as hybridization using specific
oligonucleotides [Wallace et al. Cold S~ring Harbour
Symp. Ouant. Biol. 51: 257-261 (1986)], direct DNA
sequencing [Church and Gilbert, Proc. Nat. Acad. Sci. U.
S. A. 81: 1991-1995 (1988)], the use of restriction
15 enzymes [Flavell et al. Cell 15: 25 (1978), Geever et al
Proc. Nat. Acad. Sci. U. S. A. 78: 5081 (1981)],
discrimination on the basis of electrophoretic mobility
in gels with denaturing reagent (Myers and Maniatis,
Cold Sprinq Harbour Sym. Ouant. Biol. 51: 275-284
(1986)), RNase protection (Myers, R. M., Larin, J., and
T. Maniatis Science 230: 1242 (1985)), chemical cleavage
(Cotton et al Proc. Nat. Acad. Sci. U. S. A. 85: 4397-
- 4401, (1985)) and the ligase-mediated detection
procedure [Landegren et al Science 241:1077 (1988)].
Oligonucleotides specific to normal or mutant
sequences are chemically synthesized using commercially
available machines, labelled radioactively with isotopes
(such as 32p) or non-radioactively (with tags such as
biotin (Ward and Langer et al. Proc. Nat. Acad. Sci. U.
30 S. A. 78: 6633-6657 (1981)), and hybridized to
individual DNA samples immobilized on membranes or other
solid supports by dot-blot or transfer from gels after
electrophoresis. The presence or absence of these
specific sequences are visualized by methods such as
autoradiography or fluorometric (Landegren et al, 1989,
supra) or colorimetric reactions (Gebeyehu et a. Nucleic
`,' .
20~253
Acids Research 15: 4513-4534 (1987)). An embodiment of
this oligonucleotide screening method has been applied
in the detection of the I507 deletion as described
herein.
Sequence differences between normal and mutants may
be revealed by the direct DNA sequencing method of
Church and Gilbert (supra). Cloned DNA segments may be
used as probes to detect specific DNA segments. The
- sensitivity of this method is greatly enhanced when
combined with PCR [Wrichnik et al, Nucleic Acids Res.
15:529-542 (1987); Wong et al, Nature 330:384-386
(1987); Stoflet et al, Science 239:491-494 (1988)]. In
the latter procedure, a sequencing primer which lies
within the amplified sequence is used with double-
stranded PCR product or single-stranded template
` generated by a modified PCR. The sequence determination
is performed by conventional procedures with
radiolabeled nucleotides or by automatic sequencing
procedures with fluorescent-tags.
Sequence alterations may occasionally generate
fortuitous restriction enzyme recognition sites which
are revealed by the use of appropriate enzyme digestion ~-
l followed by conventional gel-blot hybridization
`` (Southern, J. Mol. Biol 98: 503 (1975)). DNA fragments
carrying the site (either normal or mutant) are detected
by their reduction in size or increase of corresponding
restriction fragment numbers. Genomic DNA samples may
also be amplified by PCR prior to treatment with the
~' appropriate restriction enzyme; fragments of different
sizes are then visualized under UV light in the presence
of ethidium bromide after gel electrophoresis.
`~ Genetic testing based on DNA sequence differences
may be achieved by detection of alteration in
electrophoretic mobility of DNA fragments in gels with
or without denaturing reagent. Small sequence deletions
and insertions can be visualized by high resolution gel
.~ .
l,
.
x , .. .
~:`
---` 20~ ~25~
:
81
electrophoresis. For example, the PCR product with the
3 bp deletion is clearly distinguishable from the normal
sequence on an 8% non-denaturing polyacrylamide gel.
DNA fragments of different sequence compositions may be
distinguished on denaturing formamide gradient gel in
which the mobilities of different DNA fragments are
retarded in the gel at different positions according to -
, their specific "partial-melting" temperatures (Myers,
supra). In addition, sequence alterations, in
particular small deletions, may be detected as changes
in the migration pattern of DNA heteroduplexes in non-
denaturing gel electrophoresis, as have been detected
for the 3 bp (I507) mutation and in other experimental
` systems [Nagamine et al, Am. J. Hum. Genet, 45:337-339
(1989)]. Alternatively, a method of detecting a
- mutation comprising a single base substitution or other
small change could be based on differential primer
length in a PCR. For example, one invariant primer
could be used in addition to a primer specific for a
mutation. The PCR products of the normal and mutant
genes can then be differentially detected in acrylamide
gels.
Sequence changes at specific locations may also be
revealed by nuclease protection assays, such as RNase
(Myers, supra) and Sl protection (Berk, A. J., and P. A.
Sharpe Proc. Nat. Acad. Sci. U. S. A. 75: 1274 (1978)),
'!~
the chemical cleavage method (Cotton, supra) or the
~', ligase-mediated detection procedure (Landegren supra).
In addition to conventional gel-electrophoresis and
blot-hybridization methods, DNA fragments may also be
' visualized by methods where the individual DNA samples
are not immobilized on membranes. The probe and target
sequences may be both in solution or the probe sequence
~ may be immobilized [Saiki et al, Proc. Natl. Acad. Sci
7`'`~`, 35 USA, 86:6230-6234 (1989)]. A variety of detection
~ methods, such as autoradiography involving
" ~:
, ~
20112~3
radioisotopes, direct detection of radioactive decay (in
the presence or absence of scintillant),
spectrophotometry involving colorigenic reactions and
- fluorometry involving fluorogenic reactions, may be used
to identify specific individual genotypes.
Since more than one mutation is anticipated in the
CF gene such as I507 and F508, a multiples system is an
ideal protocol for screening CF carriers and detection
of specific mutations. For example, a PCR with
multiple, specific oligonucleotide primers and
hybridization probes, may be used to identify all
possible mutations at the same time (Chamberlain et al.
Nucleic Acids Research 16: 1141-1155 (1988)). The
procedure may involve immobilized sequence-specific
oligonucleotides probes (Saiki et al, su~ra).
5.2 DETECTING THE CF 507 MUTATION
,
These detection methods may be applied to prenatal
diagnosis using amniotic fluid cells, chorionic villi
biopsy or sorting fetal cells from maternal circulation.
The test for CF carriers in the population may be
incorporated as an essential component in a broad-
. scale genetic testing program for common diseases.
According to an embodiment of the invention, theportion of the DNA segment that is informative for a
mutation, such as the mutation according to this
` embodiment, that is, the portion that immediately
surrounds the I507 deletion, can then be amplified by
using standard PCR techniques [as reviewed in Landegren,
Ulf, Robert Kaiser, C. Thomas Caskey, and Leroy Hood,
DNA Diagnostics - Molecular Techniques and Automation,
in Science 242: 229-237 (1988)]. It is contemplated
that the portion of the DNA segment which is used may be
a single DNA segment or a mixture of different DNA
segments. A detailed description of this technique now
~` 35 follows.
:` ~
i~,. ~'
` . ~ .: ~ ' '.i` ~;,
~ ' " ~ ~ ' ' ' ,
20112~3
: 83
A specific region of genomic DNA from the person or
fetus is to be screened. Such specific region is
defined by the oligonucleotide primers C16B
(5'GTTTTCCTGGATTATGCCTGGCAC3') and C16D
;` 5 (5'GTTGGCATGCTTTGATGACGCTTC3') or as shown in Figure 18
-- by primers lOi-5 and lOi-3. The specific regions using
lOi-5 and lOi-3 were amplified by the polymerase chain
reaction (PCR). 200-400 ng of genomic DNA, from either
cultured lymphoblasts or peripheral blood samples of CF
10 individuals and their parents, were used in each PCR
~ with the oligonucleotides primers indicated above. The
; oligonucleotides were purified with Oligonucleotide
` Purification Cartridges~ (Applied Biosystems) or
NENSORB~ PREP columns (Dupont) with procedures
15 recommended by the suppliers. The primers were annealed
at 55 C for 30 sec, extended at 72 C for 60 sec (with 2
units of Taq DNA polymerase) and denatured at 94 C for
~` 60 sec, for 30 cycles with a final cycle of 7 min for
extension in a Perkin-Elmer/Cetus automatic thermocycler
'!,i',,, 20 with a Step-Cycle program (transition setting at 1.5
m min). Portions of the PCR products were separated by
~ electrophoresis on 1.4% agarose gels, transferred to
il Zetabind~; (Biorad) membrane according to standard
~; procedures.
The normal and ~I507 oligonucleotide probes of
` Figure 19 (10 ng each) are labeled separately with 10
units of T4 polynucleotide kinase (Pharmacia) in a 10 ~1
reaction containing 50 mM Tris-HC1 (pH7.6), 10 mM MgC12,
~`i 0.5 mM dithiothreitol, 10 mM spermidine, 1 mM EDTA and
30-40 ~Ci of ~[32p] - ATP for 20-30 min at 37C. The
unincorporated radionucleotides were removed with a
Sephadex G-25 column before use. The hybridization
conditions were as described previously (J.M. Rommens et
~`~ al Am. J. Hum. Genet. 43,645 (1988)) except that the
temperature can be 37 C. The membranes are washed
twice at room temperature with 5xSSC and twice at 39~C
. , . ~
:
--- 2~2~
84
with 2 x SSC (1 x SSC = 150 mM NaCl and 15 mM Na
citrate). Autoradiography is performed at room
; temperature overnight. Autoradiographs are developed to
show the hybridization results of genomic DNA with the 2
specific oligonucleotide probes. Probe C normal detects
the normal DNA sequence and Probe C ~I507 detects the
mutant sequence.
Genomic DNA sample from each family member can, as
explained, be amplified by the polymerase chain reaction
using the intron sequences of Figure 18 and the products
separated by electrophoresis on a 1.4~ agarose gel and
then transferred to Zetabind (Biorad) membrane according
to standard procedures. The 3bp deletion of ~I507 can
` be revealed by a very convenient polyacrylamide gel
electrophoresis procedure. When the PCR products
generated by the above-mentioned lOi-5 and lOi-3 primers
` are applied to an 5% polyacrylamide gel,
electrophoresed for 3 hrs at 20V/cm in a 90mM Tris-
borate buffer (pH 8.3), DNA fragments of a different
mobility are clearly detectable for individuals without
the 3 bp deletion, heterozygous or homozygous for the
deletion.
As already explained with respect to Figure 20, the
PCR amplified genomic DNA can be subjected to gel
electrophoresis to identify the 3 bp deletion. As shown
in Figure 20, in the four lanes the first lane is a
~, control with a normal/~F508 deletion. The next lane is
-l the father with a normal/~I507 deletion. The third lane
is the mother with a normal/~F508 deletion and the
fourth lane is the child with a QF508/~I507 deletion.
The homoduplexes show up as solid bands across the base
3 of each lane. In lanes 1 and 3, the two heteroduplexes
show up very clealy as two spaced apart bands. In lane
2, the father's ~I507 mutation shows up very clearly,
whereas in the fourth lane, the child with the adjacent
507, 508 mutations, there is no distinguishable
,^,~
:
:. ,
2011253
heteroduplexes. Hence the showing is at the homoduplex
line. Since the father in lane 2 and the mother in lane
3 show heteroduplex banding and the child does not,
indicates either the child is normal or is a patient.
This can be futher checked if needed, such as in
embryoic analysis by mixing the 507 and 508 probes to
determine the presence of the ~I507 and ~F508
mutations.
Similar alteration in gel mobility for
heteroduplexes formed during PCR has also been reported
.
for experimental systems where small deletions are
, involved (Nagamine et al supra). These mo~ility shifts
may be used in general as the basis for the non-
radioactive genetic screening tests.
5 . 3 CF SCREENING PROGR~MS
It is appreciated that approximately 1% of the
carriers can be detected using the specific QI507 probes
of this particular embodiment of the invention. Thus,
if an individual tested is not a carrier using the ~I507 ~ -
probes, their carrier status can not be excluded, they
i may carry some other mutation, such as the ~F508 as
;' previously noted. However, if both the individual and
, the spouse of the individual tested are a carrier for
the ~I507 mutation, it can be stated with certainty that
' 25 they are an at risk couple. The sequence of the gene as
' disclosed herein is an essential prerequisite for the
determination of the other mutations.
i Prenatal diagnosis is a logical extension of
carrier screening. A couple can be identified as at
risk for having a cystic fibrosis child in one of two
ways: if they already have a cystic fibrosis child,
they are both, by definition, obligate carriers of the
defective CFTR gene, and each subsequent child has a 25%
chance of being affected with cystic fibrosis. A ma~or
advantage of the present invention eliminates the need
for family pedigree analysis, wher~as, according to this
~ '
.; :
2~112~
86
invention, a gene mutation screening program as outlined
above or other similar method can be used to identify a
genetic mutation that leads to a protein with altered
function. This is not dependent on prior ascertainment
of the family through an affected child. Fetal DNA
- samples, for example, can be obtained, as previously
mentioned, from amniotic fluid cells and chorionic villi
specimens. Amplification by standard PCR techniques can
then be performed on this template DNA.
i` 10 If both parents are shown to be carriers with the
~IS07 deletion, the interpretation of the results would
be the following. If there is hybridization of the
fetal DNA to the normal probe, the fetus will not be
affected with cystic fibrosis, although it may be a CF
carrier (50% probability for each fetus of an at risk
couple). If the fetal DNA hybridizes only to the ~I507
deletion probe and not to the normal probe, the fetus
will be affected with cystic fibrosis.
It is appreciated that for this and other mutations
in the CF gene, a range of different specific
procedures can be used to provide a complete diagnosis
for all potential CF carriers or patients. A complete
description of these procedures is later described.
The invention therefore provides a method and kit
for determining if a subject is a CF carrier or CF
patient. In summary, the screening method comprises the
steps of:
providing a biological sample of the subject to be
' screened; and providing an assay for detecting in the
biological sample, the presence of at least a member
from the group consisting of a 507 mutant CF gene, 507
mutant CF gene products and mixtures thereof.
The method may be further characterized by
including at least one more nucleotide probe which is a
different DNA sequence fragment of, for example, the DNA
of Figure 1, or a different DNA sequence fragment of
.. .:
:,-,
,' ~,.
201~2~3
,
87
human chromosome 7 and located to either side of the DNA
sequence of Figure 1. In this respect, the DNA
fragments of the intron portions of Figure 2 are useful
in further confirming the presence of the mutation.
Unique aspects of the introns at the exon boundaries may
be relied upon in screening procedures to further
confirm the presence of the mutation at the I507
position or othe mutant positions.
A kit, according to an embodiment of the invention,
suitable for use in the screening technique and for
assaying for the presence of the mutant CF gene by an
immunoassay comprises:
(a) an antibody which specifically binds to a gene
product of the mutant CF gene having a mutation at one
of the positions of 455, 507, 542, 549 and 560;
(b) reagent means for detecting the binding of the
antibody to the gene product; and
(c) the antibody and reagent means each being
present in amounts effective to perform the immunoassay.
The kit for assaying for the presence for the
mutant CF gene may also be provided by hybridization
techniques. The kit comprises:
(a) an oligonucleotide probe which specifically
binds to the mutant CF gene having a mutation at one of
the positions, 455, 507, 542, 549 and 560;
` (b) reagent means for detecting the hybridization
of the oligonucleotide probe to the mutant CF gene; and
` (c) the probe and reagent means each being present -~
i in amounts effective to perform the hybridization assay.
5.4 ANTIBODIES TO DETECT MUTANT CFTR
As mentioned, antibodies to epitopes within the
` mutant CFTR protein at positions 455, 507, 542, 549 and
560 are raised to provide extensive information on the `~
characteristics of the mutant protein and other valuable
information which includes:
:- .
- 2 ~ 2 ~ ~
88
1. The antibodies can be used to provide another
technique in detecting any of the other CF
mutations which result in the synthesis of a
- protein with an altered size.
2. Antibodies to distinct domains of the mutant
protein can be used to determine the topological
arrangement of the protein in the cell membrane.
This provides information on segments of the
I protein which are accessible to externally added
'I 10 modulating agents for purposes of drug therapy.
3. The structure-function relationships of
portions of the protein can be examined using
specific antibodies. For example, it is possible
to introduce into cells antibodies recogniæing each
of the charged cytoplasmic loops which join the
transmembrane sequences as well as portions of the
nucleotide binding folds and the R-domain. The
~ influence of these antibodies on functional
!~ parameters of the protein provide insight into cell
regulatory mechanisms and potentially suggest means
of modulating the activity of the defective protein
in a CF patient.
4. Antibodies with the appropriate avidity also
enable immunoprecipitation and immuno-affinity
purification of the protein. Immunoprecipitation
will facilitate characterization of synthesis and ,¢
post translational modification including ATP
binding and phosphorylation. Purification will be
required for studies of protein structure and for
reconstitution of its function, as well as protein
based therapy.
In order to prepare the antibodies, fusion proteins
containing defined portions of anyone of the mutant CFTR
polypeptides can be synthesized in bacteria by
expression of corresponding mutant DNA sequence in a
suitable cloning vehicle. Smaller peptide may be
~.,
,.
- 2011~3
89
synthesized chemically. The fusion proteins can be
purified, for example, by affinity chromatography on
glutathione-agarose and the peptides coupled to a
carrier protein (hemocyanin)/ mixed with Freund's
adjuvant and injected into rabbits. Following booster
injections at bi-weekly intervals, the rabbits are bled
and sera isolated. The developed polyclonal antibodies
in the sera may then be combined with the fusion
proteins. Immunoblots are then formed by staining with,
for example, alkaline-phosphatase conjugated second
antibody in accordance with the procedure of Blake et
al, Anal. Biochem. 136:175 (1984).
Thus, it is possible to raise polyclonal antibodies
specific for both fusion proteins containing portions of
15 the mutant CFTR protein and peptides corresponding to -~;
short segments of its sequence. Similarly, mice can be
injected with KLH conjugates of peptides to initiate
the production of monoclonal antibodies to corresponding
segments of mutant CFTR protein.
As for the generation of monoclonal antibodies,
immunogens for the raising of monoclonal antibodies
(mAbs) to the mutant CFTR protein are bacterial fusion
proteins [Smith et al, Gene 67:31 (1988)] containing
portions of the CFTR polypeptide or synthetic peptides
corresponding to short (12 to 25 amino acids in length)
segments of the mutant sequence. The essential
methodology is that of Kohler and Milstein [Nature 256: ;
495 (1975)]-
Balb/c mice are immunized by intraperitoneal
injection with 500 ~g of pure fusion protein or
synthetic peptide in incomplete Freund's adjuvant. A
second injection is given after 14 days, a third after
21 days and a fourth after 28 days. Individual animals
so immunized are sacrificed one, two and four weeks
following the final injection. Spleens are removed,
their cells dissociated, collected and fused with Sp2/O-
~ :
., ~'.
2~2~
Agl4 myeloma cells according to Gefter et al, somaticCell Genetics 3:231 (1977). The fusion mixture is
distributed in culture medium selective for the
propagation of fused cells which are grown until they
are about 25% confluent. At this time, culture
supernatants are tested for the presence of antibodies
reacting with a particular CFTR antigen. An alkaline
, phosphatase labelled anti-mouse second antibody is then
¦ used for detection of positives. Cells from positive
culture wells are then expanded in culture, their
supernatants collected for further testing and the cells
stored deep frozen in cryoprotectant-containing medium.
i To obtain large quantities of a mAb, producer cells are
injected into the peritoneum at 5 x 106 cells per
animal, and ascites fluid is obtained. Purification is
by chromotography on Protein G- or Protein A-agarose
according to Ey et al, Immunochemistry 15:429 (1977).
` Reactivity of these mAbs with the mutant CFTR
protein can be confirmed by polyacrylamide gel
, 20 electrophoresis of membranes isolated from epithelial
cells in which it is expressed and immunoblotted
[Towbin et al, Proc. Natl. Acad. Sci. USA 76:4350
' (1979)]-
i In addition to the use of monoclonal antibodies
specific for the particular mutant domain of the CFTR
~ protein to probe their individual functions, other mAbs,
; which can distinguish between the normal and mutant
forms of CFTR protein, are used to detect the mutant
protein in epithelial cell samples obtained from
patients, such as nasal mucosa biopsy "brushings" [ R.
De-Lough and J. Rutland, J. Clin. Pathol. 42, 613
(1989)] or skin biopsy specimens containing sweat
glands.
Antibodies capable of this distinction are obtained
by differentially screening hybridomas from paired sets
of mice immunized with a peptide containing, for
,'s ' ~"
20~12~3
~: 91
- example, the isoleucine at amino acid position 507 (e.g.
GTIKENIIFGVSY) or a peptide which is identical except
for the absence of I507 (GTIKENIFGVSY). mAbs capable
i of recognizing the other mutant forms of CFTR protein
present in patients in addition or instead of I507
deletion are obtained using similar monoclonal antibody
production strategies.
Antibodies to normal and CF versions of CFTR
protein and of segments thereof are used in ~-~
diagnostically immunocytochemical and immunofluorescence
light microsc~py and immunoelectron microscopy to
demonstrate the tissue, cellular and subcellular
distribution of CFTR within the organs of CF patients,
carriers and non-CF individuals.
Antibodies are used to therapeutically modulate by
! promoting the activity of the CFTR protein in CF
patients and in cells of CF patients. Possible modes of
`' such modulation might involve stimulation due to cross-
linking of CFTR protein molecules with multivalent
antibodies in analogy with stimulation of some cell
~ surface membrane receptors, such as the insulin receptor
- [O'Brien et al, Euro. Mol. Biol. Organ. J. 6:4003
(1987)], epidermal growth factor receptor [Schreiber et
al, J. Biol. Chem. 258:846 (1983)] and T-cell receptor-
associated molecules such as CD4 [Veillette et al
` Nature, 338:257 (1989)].
Antibodies ars used to direct the delivery of
therapeutic agents to the cells which express defective
CFTR protein in CF. For this purpose, the antibodies
are incorporated into a vehicle such as a liposome
[Matthay et al, Cancer Res. 46:4904 (1986)] which
`l carries the therapeutic agent such as a drug or the
normal gene.
, 5.5 RFLP ANALYSIS
`~ 35 DNA diagnosis is currently being used to assess
~ whether a fetus will be born with cystic fibrosis, but
~,
~` !
.:
,. . . ~ . -
- .. ~ ' . . - .. . . : ~ .: . .: . . . .
L 2 ~ ~
,
historically this has only been done after a particular
set of parents has already had one cystic fibrosis child
which identifies them as obligate carriers. However, in
combination with carrier detection as outlined above,
DNA diagnosis for all pregnancies of carrier couples
will be possible. If the parents have already had a
cystic fibrosis child, an extended haplotype analysis
can be done on the fetus and thus the percentage of
false positive or false negative will be greatly
reduced. If the parents have not already had an
affected child and the DNA diagnosis on the fetus is
being performed on the basis of carrier detection,
haplotype analysis can still be performed.
Although it has been thought for many years that
there is a great deal of clinical heterogeneity in the
cystic fibrosis disease, it is now emerging that there
are two general categories, called pancreatic
sufficiency ( CF-PS ) and pancreatic insufficiency ( CF-
PI). If the mutations related to these disease
categories are well characterized, one can associate a
particular mutation with a clinical phenotype of the
disease. This allows changes in the treatment of each
patient. Thus the nature of the mutation will to a
certain extent predict the prognosis of the patient and
indicate a specific treatment.
6.0 MOLECULAR BIOLOGY OF CYSTIC FIBROSIS
The postulate that CFTR may regulate the activity
of ion channels, particularly the outwardly rectifying
Cl channel implicated as the functional defect in CF,
- 30 can be tested by the injection and translation of full
length in vitro transcribed CFTR m~NA in Xenopus
oocytes. The ensuing changes in ion currents across the
oocyte membrane can be measured as the potential is
clamped at a fixed value. CFTR may regulate endogenous
35 oocyte channels or it may be necessary to also introduce ~-~
epithelial cell RNA to direct the translation of channel ~
:. :
~03l~2~3
93
proteins. Use of mRNA coding for normal and for mutant
CFTR, as provided by this invention, makes these
experiments possible.
Other modes of expression in heterologous cell
S system also facilitate dissection of structure-function
relationships. The complete CFTR DNA sequence ligated
- into a plasmid expression vector is used to transfect
cells so that its influence on ion transport can be
assessed. Plasmid expression vectors containing part of
- 10 the normal CFTR sequence along with portions of modified
sequence at selected sites can be used in vitro
mutagenesis experiments performed in order to identify
those portions of the CFTR protein which are crucial for
regulatory function.
6.1 EXPRESSION OF THE MUTANT DNA SEOUENCE
The mutant DNA sequence can be manipulated in
studies to understand the expression of the gene and its
product, and, to achieve production of large quantities
of the protein for functional analysis, antibody
production, and patient therapy. The changes in the
sequence may or may not alter the expression pattern in
terms of relative quantities, tissue-specificity and
` functional properties. The partial or full-length cDNA
sequences, which encode for the subject protein,
, 25 unmodified or modified, may be ligated to bacterial
` expression vectors such as the pRIT (Nilsson et al. EMBO
.~ J. 4: 1075-1080 (1985)), pGEX (Smith and Johnson, Gene
` 67: 31-40 (1988)) or pATH (Spindler et al. J. Virol. 49:
132-141 (1984)) plasmids which can be introduced into E.
coli cells for production of the corresponding proteins
` which may be isolated in accordance with the previously
i~ discussed protein purification procedures. The DNA
sequence can also be transferred from its existing
~ context to other cloning vehicles, such as other
"! 35 plasmids, bacteriophages, cosmids, animal virus, yeast
1 artificial chromosomes (YAC)(Burke et al. Science 236~
. :~ :: :
~-
~ -: , : .
~ ^:
--- 2~12~
94
806-812, (1987)), somatic cells, and other simple or
complex organisms, such as bacteria, fungi (Timberlake
and Marshall, Science 244: 1313-1317 (1~89),
invertebrates, plants (Gasser and Fraley, Science 244:
1293 (1989), and pigs (Pursel et al. Science 244: 1281-
1288 ~1989)).
For expression in mammalian cells, the cDNA
~- sequence may be ligated to heterologous promoters, such
- as the simian virus (SV) 40, promoter in the pSV2 vector
~- 10 [Mulligan and Berg, Proc. Natl. Acad. Sci USA, 78:2072-
2076 (1981)] and introduced into cells, such as monkey
COS-l cells [Gluzman, Cell, 23:175-182 (1981)], to
achieve transient or long-term expression. The stable
integration of the chimeric gene construct may be
maintained in mammalian cells by biochemical selection,
such as neomycin [Southern and Berg, J. Mol. APpln.
Genet. 1:327-341 (1982)] and mycophoenolic acid
[Mulligan and Berg, supra].
DNA sequences can be manipulated with standard
procedures such as restriction enzyme digestion, fill-in
with DNA polymerase, deletion by exonuclease, extension
by terminal deoxynucleotide transferase, ligation of
synthetic or cloned DNA sequences, site-directed
~~I sequence-alteration via single-stranded bacteriophage
-~, 25 intermediate or with the use of specific
,; oligonucleotides in combination with PCR.
` The cDNA sequence (or portions derived from it), or
a mini gene (a cDNA with an intron and its own promoter)
. is introduced into eukaryotic expression vectors by ~-
: .
30 conventional techniques. These vectors are designed to ~;
permit the transcription of the cDNA in eukaryotic cells
by providing regulatory sequences that initiate and
enhance the transcription of the cDNA and ensure its
: 1 .
proper splicing and polyadenylation. Vectors containing
! 35 the promoter and enhancer regions of the simian virus
(SV)40 or long terminal repeat (LTR) of the Rous Sarcoma
i ~
:~`
-` 20112~3
virus and polyadenylation and splicing signal from SV
40 are readily available [Mulligan et al Proc. Natl.
Acad. Sci. USA 78:1078-2076, (1981); Gorman et al Proc
Natl. Acad. Sci USA 79: 6777-6781 (1982)].
Alternatively, the CFTR endogenous promoter may be used.
The level of expression of the cDNA can be manipulated
with this type of vector, either by using promoters that
have different activities (for example, the baculovirus
pAC373 can express cDNAs at high levels in S.
frunqiPerda cells [M. D. Summers and G. E. Smith in,
Genetically Altered Viruses and the Environment (B.
Fields, et al, eds.) vol. 22 no 319-328, Cold Spring
Harbour Laboratory Press, Cold Spring Harbour, New York,
1985] or by using vectors that contain promoters
amenable to modulation, for example the glucocorticoid-
` responsive promoter from the mouse mammary tumor virus
[Lee et al, Nature 294:228 (1982)]. The expression of
, the cDNA can be monitored in the recipient cells 24 to
72 hours after introduction (transient expression).
In addition, some vectors contain selectable
markers [such as the a~t [Mulligan et Berg supra] or neo
[Southern and Berg J. Mol. Appln. Genet 1:327-341
! (1982)] bacterial genes that permit isolation of cells,
by chemical selection, that have stable, long term
expression of the vectors (and therefore the cDNA) in
the recipient cell. The vectors can be maintained in
the cells as episomal, freely replicating entities by
, using regulatory elements of viruses such as papilloma
[Sarver et al Mol. Cell Biol. 1:486 (1981)] or Epstein-
Barr (Sugden et al Mol. Cell Biol. 5:410 (1985)].
Alternatively, one can also produce cell lines that have
integrated the vector into genomic DNA. Both of these
types of cell lines produce the gene product on a
~i~ continuous basis. One can also produce cell lines that
have amplified the number of copies of the vector (and
therefore of the cDNA as well) to create cell lines that ~-~
.. ~.
20~12~3
96
can produce high levels of the gene product [Alt et al.
J. Biol. Chem. 253: 1357 (1978)].
The transfer of DNA into eukaryotic, in particular
human or other mammalian cells is now a conventional
technique. The vectors are introduced into the
recipient cells as pure DNA (transfection) by, for
example, precipitation with calcium phosphate [Graham
and vander Eb, Virology 52:466 (1973) or strontium
phosphate [Brash et al Mol. Cell Biol. 7:2013 (1987)],
10 electroporation [Neumann et al EMBO J 1:841 (1982)],
lipofection ~Felgner et al Proc Natl. Acad. Sci USA
84:7413 (1987)], DEAE dextran [McCuthan et al J. Natl
Cancer Inst. 41:351 1968)], microinjection [Mueller et
al Cell 15:579 1978)], protoplast fusion [Schafner, Proc
15 Natl. Aca. Sci USA 72:2163] or pellet guns [Klein et al,
Nature 327: 70 (1987)]. Alternatively, the cDNA can be
introduced by infection with virus vectors. Systems are
developed that use, for example, retroviruses [Bernstein
et al. Genetic Enaineerinq 7: 235, (1985)], adenoviruses
20 [Ahmad et al J. Virol 57:267 (1986)] or Herpes virus
`i [Spaete et al Cell 30:295 (1982)].
;; These eukaryotic expression systems can be used for
many studies of the mutant CF gene and the mutant CFTR
product, such as at protein positions 455, 507, 542, 549
and 560. These include, for example: (1) determination
that the gene is properly expressed and that all post-
; translational modifications necessary for full
biological activity have been properly completed (2)
` identify regulatory elements located in the 5' region of
the CF gene and their role in the tissue- or temporal-
regulation of the expression of the CF gene (3)
~ production of large amounts of the normal protein for
;, isolation and purification (4) to use cells expressing
the CFTR protein as an assay system for antibodies
generated against the CFTR protein or an assay system totest the effectiveness of drugs, (5) study the function
', ;'' ~ ~':':
2011253
97
of the normal complete protein, speci~ic portions of the
protein, or of naturally occurring or artificially
produced mutant proteins. Naturally occurring mutant
proteins exist in patients with CF while artificially
produced mutant protein can be designed by site directed
sequence alterations. These latter studies can probe
the function of any desired amino acid residue in the
protein by mutating the nucleotides coding for that
amino acid.
Using the above techniques, the expression vectors
containing the mutant CF gene sequence or fragments
thereof can be introduced into human cells, mammalian
cells from other species or non-mammalian cells as
desired. The choice of cell is determined by the
purpose of the treatment. For example, one can use
monkey COS cells [Gluzman, Cell 23:175 (1981)], that
produce high levels of the SV40 T antigen and permit the
- replication of vectors containing the SV40 origin of
replication, can be used to show that the vector can
l 20 express the protein product, since function is not
¦ required. Similar treatment could be performed with
Chinese hamster ovary (CHO) or mouse NIH 3T3 fibroblasts
or with human fibroblasts or lymphoblasts.
; The recombinant cloning vector, according to this ~;~25 invention, then comprises the selected DNA of the DNA -~
sequences of this invention for expression in a suitable ~-~
host. The DNA is operatively linked in the vector to an
expression control sequence in the recombinant DNA
molecule so that normal CFTR polypeptide can be
expressed. The expression control sequence may be
selected from the group consisting of sequences that
` control the expression of genes of prokaryotic or
eukaryotic cells and their viruses and combinations
' thereof. The expression control sequence may be
specifically selected from the group consisting of the
lac system, the trp system, the tac system, the trc
. :
:` :
' ~
~.12~
98
system, major operator and promoter regions of phage
lambda, the control region of fd coat protein, the early
and late promoters of SV40, promoters derived from
polyoma, adenovirus, retrovirus, baculovirus and simian
virus, the promoter for 3-phosphoglycerate kinase, the
promoters of yeast acid phosphatase, the promoter of the
yeast alpha-mating factors and combinations thereof.
The host cell, which may be transfected with the
vector of this invention, may be selected from the group
consisting of E. coli, Pseudomonas, Bacillus subtilis,
Bacillus stearothermophilus or other bacili; other
bacteria; yeast; fungi; insect; mouse or other animal;
or plant hosts; or human tissue cells.
It is appreciated that for the mutant DNA sequence -~
15 similar systems are employed to express and produce the -
j mutant product.
6.2 PROTEIN FUNCTION CONSIDERATIONS
To study the function of the mutant CFTR protein,
it is preferable to use epithelial cells as recipients, ~ -
since proper functional expression may require the
' presence of other pathways or gene products that are
i only expressed in such cells. Cells that can be used
include, for example, human epithelial cell lines such
~; as T84 (ATCC #CRL 248) or PANC-l (ATCC # CLL 1469), or
the T43 immortalized CF nasal epithelium cell line
[Jettan et al, Science (1989)] and primary [Yanhoskes et
al. Ann. Rev. Resp. Dis. 132: 1281 (1985)] or
transformed [Scholte et al. EXP. Cell. Res. 182:
559(1989)] human nasal polyp or airways cells,
~ 30 pancreatic cells [Harris and Coleman J. Cell. Sci. 87:
; 695 (1987)], or sweat gland cells [Collie et al. In
Vitro 21: 597 (1985)] derived from normal or CF
subjects. The CF cells can be used to test for the
functional activity of mutant CF genes. Current
35 functional assays available include the study of the ~-
movement of anions (Cl or I) across cell membranes as a
.
' :
::~i?: :
2~1~2~3
function of stimulation of cells by agents that raise
intracellular AMP levels and activate chloride channels
[Stutto et al. Proc. Nat. Acad. Sci. U. S. A. 82: 6677
(1985)]. Other assays include the measurement of changes
in cellular potentials by patch clamping of whole cells
or of isolated membranes [Frizzell et al. Science 233:
558 (1986), Welsch and Liedtke Nature 322: 467 (1986)]or
, the study of ion fluxes in epithelial sheets of
confluent cells [Widdicombe et al. Proc. Nat. Acad. Sci.
82: 6167 (1985)]. Alternatively, RNA made from the CF
gene could be injected into Xenopus oocytes. The oocyte
will translate RNA into protein and allow its study. As
other more specific assays are developed these can also
be used in the study of transfected mutant CFTR protein
function.
.
"Domain-switching" experiments between mutant CFTR
and the human multidrug resistance P-glycoprotein can
also be performed to further the study of the mutant
' CFTR protein. In these experiments, plasmid expression
; 20 vectors are constructed by routine techniques from
; fragments of the mutant CFTR sequence and fragments of
the sequence of P-glycoprotein ligated together by DNA ~ -
ligase so that a protein containing the respective
portions of these two proteins will be synthesized by a
host cell transfected with the plasmid. The latter
approach has the advantage that many experimental
parameters associated with multidrug resistance can be
measured. Hence, it is now possible to assess the
ability of segments of mutant CFTR to influence these
- 30 parameters.
These studies of the influence of mutant CFTR on
ion transport will serve to bring the field of
epithelial transport into the molecular arena.
6.3 THERAPIES
- 35 It is understood that the major aim of the various
biochemical studies using the compositions of this
.,
. .1. 2 ~ ~
100
; invention is the development of therapies to circumvent
or overcome the CF defect, using both the
pharmacological and the "gene-therapy" approaches.
In the pharmacological approach, drugs which
circumvent or overcome the CF defect are sought.
Initially, compounds may be tested essentially at
random, and screening systems are required to
discriminate among many candidate compounds. This
invention provides host cell systems, expressing various
` 10 of the mutant CF genes, which are particularly well
suited for use as first level screening systems.
~ Preferably, a cell culture system using mammalian cells
: (most preferably human cells) transfected with an
expression vector comprising a DNA sequence coding for
CFTR protein containing a CF-generating mutation, for
example the I507 deletion, is used in the screening
process. Candidate drugs are tested by incubating the
cells in the presence of the candidate drug and
measuring those cellular functions dependent on CFTR,
especially by measuring ion currents where the
transmembrane potential is clamped at a fixed value. To
accommodate the large number of assays, however, more
convenient assays are based, for example, on the use of
ion-sensitive fluorescent dyes. To detect changes in
` 25 Cl~ion concentration SPQ or its analogues are useful.
Alternatively, a cell-free system could be used.
Purified CFTR could be reconstituted into articifial
membranes and drugs could be screened in a cell-free
assay [Al-Aqwatt, Science, (1989)].
At the second level, animal testing is required.
It is possible to develop a model of CF by interfering
with the normal expression of the counterpart of the CF
gene in an animal such as the mouse. The "knock-out" of
this gene by introducing a mutant form of it into the
germ line of animals will provide a strain of animals
with CF-like syndromes. This enables testing of drugs
." '
:-`` 2~112~3
101
which showed a promise in the first level cell-based
screen.
As further knowledge is gained about the nature of
the protein and its function, it will be possible to
predict structures of proteins or other compounds that
interact with the CFTR protein. That in turn will allow
for certain predictions to be made about potential drugs
that will interact with this protein and have some
~ effect on the treatment of the patients. Ultimately -~
;~ 10 such drugs may be designed and synthesized chemically on
the basis of structures predicted to be required to
interact with domains of CFTR. This approach is
reviewed in Capsey and Delvatte, Genetically Enqineered
Human Therapeutic Drugs Stockton Press, New York, 1988.
These potential drugs must also be tested in the
screening system.
6.3.1 PROTEIN REPLACEMENT THERAPY
I Treatment of CF can be performed by replacing the
;~~ defective protein with normal protein, by modulating the
function of the defective protein or by modifying
another step in the pathway in which CFTR participates
in order to correct the physiological abnormality.
` To be able to replace the defective protein with
the normal version, one must have reasonably large
amounts of pure CFTR protein. Pure protein can be
obtained as described earlier from cultured cell
systems. Delivery of the protein to the affected
airways tissue will require its packaging in lipid-
containing vesicles that facilitate the incorporation of
30 the protein into the cell membrane. It may also be ~ -
feasible to use vehicles that incorporate proteins such
as surfactant protein, such as SAP(Val) or SAP(Phe) that
performs this function naturally, at least for lung
, alveolar cells. (PCT Patent Application WO/8803170,
Whitsett et al, May 7, 1988 and PCT Patent Application
W089/04327, Benson et al, May 18, 1989). The CFTR-
.~ ~
. .:
2~1 12~3
102
containing vesicles are introduced into the airways by
inhalation or irrigation, techniques that are currently
used in CF treatment (Boat et al, supra).
- 6.3.2 DRUG THERAPY
Modulation of CFTR function can be accomplished by
. .
- the use of therapeutic agents (drugs). These can be
identified by random approaches using a screening
program in which their effectiveness in modulating the
defective CFTR protein is monitored ln vitro. Screening
programs can use cultured cell systems in which the
defective CFTR protein is expressed. Alternatively,
drugs can be designed to modulate CFTR activity from
` knowledge of the structure and function correlations of
CFTR protein and from knowledge of the specific defect
15 in the CFTR mutant protein (Capsey and Delvatte,
supra). It is possible that the mutant CFTR protein
will require a different drug for specific modulation.
It will then be necessary to identify the specific
mutation(s) in each CF patient before initiating drug
20 therapy.
Drugs can be designed to interact with different
aspects of CFTR protein structure or function. For
` example, a drug (or antibody) can bind to a structural
fold of the protein to correct a defective structure.
25 Alternatively, a drug might bind to a specific
functional residue and increase its affinity for a
substrate or cofactor. Since it is known that members
of the class of proteins to which CFTR has structural
homology can interact, bind and transport a variety of
30 drugs, it is reasonable to expect that drug-related
therapies may be effective in treatment of CF. -
; A third mechanism for enhancing the activity of an
' effective drug would be to modulate the production or
~- the stability of CFTR inside the cell. This increase
` 35 in the amount of CFTR could compensate for its defective -
v function.
i`~l
~,
.,, ~
:. :
.r `
2~1~2~3
103
Drug therapy can also be used to compensate for the
- defective CFTR function by interactions with other
j components of the physiological or biochemical pathway
, necessary for the expression of the CFTR function.
¦ 5 These interactions can lead to increases or decreases in
; the activity o~ these ancillary proteins. The methods
- for the identification of these drugs would be similar
to those described above for CFTR-related drugs.
In other genetic disorders, it has been possible to
correct for the consequences of altered or missing
normal functions by use of dietary modifications. This
has taken the form of removal of metabolites, as in the
case of phenylketonuria, where phenylalanine is removed
from the diet in the first five years of life to prevent
mental retardation, or by the addition of large amounts
of metabolites to the diet, as in the case of adenosime
deaminase deficiency where the functional correction of
the activity of the enzyme can be produced by the
addition of the enzyme to the diet. Thus, once the
details of the CFTR function have been elucidated and
the basic defect in CF has been defined, therapy may be
achieved by dietary manipulations.
~, The second potential therapeutic approach is so-
called "gene-therapy" in which normal copies of the CF
gene are introduced in to patients so as to successfully
code for normal protein in the key epithelial cells of
affected tissues. It is most crucial to attempt to
achieve this with the airway epithelial cells of the
respiratory tract. The CF gene is delivered to these
cells in form in which it can be taken up and code for
. .
sufficient protein to provide regulatory function. As a
~ result, the patient's quality and length of life will be
'! greatly extended. Ultimately, of course, the aim is to
deliver the gene to all affected tissues.
,.~
. . .
: 2~12~
- 104
~- 6.3.3 GENE THERAPY
One approach to therapy of CF is to insert a normal
version of the CF gene into the airway epithelium of
affected patients. It is important to note that the
respiratory system is the primary cause of mordibity and
mortality in CF; while pancreatic disease is a major
; feature, it is relatively well treated today with enzyme
` supplementation. Thus, somatic cell gene therapy [for a
-~ review, see T. Friedmann, Science 244:1275 (1989)]
targeting the airway would alleviate the most severe
problems associated with CF. `
A. Retroviral Vectors. Retroviruses have been
.. .
considered the preferred vector for experiments in
somatic gene therapy, with a high efficiency of
infection and stable integration and expression [Orkin
et al rog. Med. Genet 7:130, (1988)]. A possible
drawback is that cell division is necessary for
retroviral integration, so that the targeted cells in
the airway may have to be nudged into the cell cycle
prior to retroviral infection, perhaps by chemical
means. The full length CF gene cDNA can be cloned into
a retroviral vector and driven from either its
endogenous promoter or from the retroviral LRT (long
terminal repeat). Expression of levels of the normal
protein as low as 10% of the endogenous mutant protein
` in CF patients would be expected to be beneficial, since
this is a recessive disease. Delivery of the virus
could be accomplished by aerosol or instillation into
the trachea. ~ ~`
B. Other Viral Vectors. Other delivery systems
,~ which can be utilized include adeno-associated virus
[AAV, McLaughlin et al, J. Virol 62:1963 (1988)],
vaccinia virus [Moss et al Annu. Rev.Immunol, 5:305,
- 1987)], bovine papilloma virus [Rasmussen et al, Methods
Enzvmol 139:642 (1987)] or member of the herpesvirus
, group such as Epstein-Barr virus (Margolskee et al Mol.
,, ,:
~- 201~ 2~3
~.
105
Cell. Biol 8:2937 (1988)~. Though much would need to
be learned about their basic biology, the idea of using
a viral vector with natural tropism for the respiratory
track (e.g. respiratory syncytial virus, echovirus,
Coxsackie virus, etc.) is possible.
! C. Non-viral Gene Transfer. Other methods of
inserting the CF gene into respiratory epithelium may
also be productive; many of these are lower efficiency
and would potentially require infection in vitro,
selection of transfectants, and reimplantation. This
would include calcium phosphate, DEAE dextran,
electroporation, and protoplast fusion. A particularly
attractive idea is the use of liposome, which might be
~ possible to carry out ln vlvo [Ostro, Liposomes, Marcel-
- 15 Dekker, 1987]. Synthetic cationic lipids such as DOTMA
[Felger et al Proc. Natl. Acad.Sci USA 84:7413 (1987)3
may increase the efficiency and ease of carrying out
this approach.
6.4 CF ANIMAL MODELS
' 20 The creation of a mouse or other animal model for
CF will be crucial to understanding the disease and for
testing of possible therapies (for general review of
~ creating animal models, see Erickson, Am. J. Hum.Genet
`3 43:582 (1988)]. Currently no animal model of the CF
exists. The evolutionary conservation of the CF gene
(as demonstrated by the cross-species hybridization
blots for E4.3 and Hl.6), as is shown in Figure 4,
indicate that an orthologous gene exists in the mouse
(hereafter to be denoted mCF, and its corresponding
protein as mCFTR), and this will be possible to clone in
mouse genomic and cDNA libraries using the human CF gene
~`.! , probes. It is expected that the generation of a
specific mutation in the mouse gene analogous to the
I507 mutation will be most optimum to reproduce the
;l 35 phenotype, though complete inactivation of the mCFTR
gene will also be a useful mutant to generate.
~i
'.~
2 ~ 3
106
A. Mutaaenesis. Inactivation of the mCF gene can
be achieved by chemical [e.g. Johnson et al Proc. Natl.
Acad. Sci. USA 78:3138 (1981)] or X-ray mutagenesis
[Popp et al J. Mol. Biol. 127:141 (1979)] of mouse
gametes, followed by fertilization. Offspring
heterozygous for inactivation of mCFTR can then be
identified by Southern blotting to demonstrate loss of
one allele by dosage, or failure to inherit one parental
allele if an RFLP marker is being assessed. This
approach has previously been successfully used to
identify mouse mutants for ~-globin [Whitney et al Proc.
Natl. Acad. Sci. USA 77:1087 (1980)], phenylalanine
hydroxylase [McDonald et al Pediatr. Res 23:63 (1988)],
and carbonic anhydrase II [Lewis et al Proc. Natl. Acad.
Sci. USA 85:1962, (19~8)].
B. Transgenics A mutant version of CFTR or mouse
CFTR can be inserted into the mouse germ line using now
` standard techniques of oocyte injection [Camper, Trends
in Genetics (1988)]; alternatively, if it is desirable
to inactivate or replace the endogenous mCF gene, the
homologous recombination system using embryonic stem
(ES) cells [Capecchi, Science 244:1288 (1989)] may be
applied.
1. Oocyte In~ection Placing one or more
copies of the normal or mutant mCF gene at a random
location in the mouse germline can be accomplished by
microinjection of the pronucleus of a just-fertilized
mouse oocyte, followed by reimplantation into a pseudo-
pregnant foster mother. The liveborn mice can then be
screened for integrants using analysis of tail DNA for
the presence of human CF gene sequences. The same ~
` protocol can be used to insert a mutant mCF gene. To ~;
~ generate a mouse model, one would want to place this
`~ transgene in a mouse background where the endogenous mCF
~i 35 gene has been inactivated, either by mutagenesis (see
; above ) or by homologous recombination (see below). The
201~ 2~3
:''
107
transgene can be either: a) a complete genomic
sequence, though the size of this (about 250 kb) would
require that it be injected as a yeast artificial
chromosome or a chromosome fragment; b) a cDNA with
; 5 either the natural promoter or a heterologous promoter;
c) a "minigene" containing all of the coding region and
~`~ various other elements such as introns, promoter, and 3'
flanking elements found to be necessary for optimum
~, expression.
3 lo 2. Retroviral Infection of Early Embryos.
~J This alternative involves inserting the CFTR or mCF
, gene into a retroviral vector and directly infecting
~,1
-~ mouse embroyos at early stages of development generating
a chimera [Soriano et al Cell 46:19 (1986)]. At least
some of these will lead to germline transmission.
3. ES Cells and Homologous Recombination.
The embryonic stem cell approach (Capecchi, supra and
Capecchi, Trends Genet 5:70 (1989)] allows the
possibility of performing gene transfer and then
;~ 20 screening the resulting totipotent cells to identify the
;~ rare homologous recombination events. Once identified,
these can be used to generate chimeras by injection of
mouse blastocysts, and a proportion of the resulting
, mice will show germline transmission from the
;~ 25 recombinant line. There are several ways this could be
~l useful in the generation of a mouse model for CF:
'?`'~ aj Inactivation of the mCF gene can be
conveniently accomplished by designing a DNA fragment
, which contains sequences from a mCFTR exon flanking a
~'~ 30 selectable marker such as neo. Homologous recombination
~: .
will lead to insertion of the neo sequences in the
~- middle of an exon, inactivating mCFTR. The homologous
-~ recombination events (usually about 1 in 1000) can be
recognized from the heterologous ones by DNA analysis of
individual clones [usually using PCR, Kim et al Nucleic
Acids Res. 16:8887 (1988), Joyner et al Nature 338:153
',
. ' .
` i` '- , .
~".. " ' ., .
'~ ' ' ' ~ ' ~ ' '; ' " ' ' ' ' ' ` ' -
2~112~3
108
(1989); Zimmer et al supra, p. 150] or by using a
negative selection against the heterologous events [such
as the use of an HSV TK gene at the end of the
construct, followed by the gancyclovir selection,
Mansour et al, Nature 336:348 (1988)]. This inactivated
mCFTR mouse can then be used to introduce a mutant CF
gene or mCF gene containing, for example, the I507
abnormality or any other desired mutation.
b) It is possible that specific mutants of mCFTR
cDNA be created in one step. For example, one can make
a construct containing mCF intron 9 sequences at the 5'
end, a selectable neo gene in the middle, and intro 9 +
exon 10 (containing the mouse version of the I507
mutation) at the 3' end. A homologous recombination
event would lead to the insertion of the neo gene in
intron 9 and the replacement of exon 10 with the mutant
i. , ,
,. verslon.
,
c) If the presence of the selectable neo marker in
the intron altered expresson of the mCF gene, it would
20 be possible to excise it in a second homologous
recombination step.
d) It is also possible to create mutations in the
f mouse germline by injecting oligonucleotides containing
the mutation of interest and screening the resulting
25 cells by PCR.
t This embodiment of the invention has considered
primarily a mouse model for cystic fibrosis. Figure 4
shows cross-species hybridization not only to mouse DNA,
but also to bovine, hamster and chicken DNA. Thus, it
30 is contemplated that an orthologous gene will exist in
many other species also. It is thus contemplated that
it will be possible to generate other animal models
<l using similar technology.
Although preferred embodiments of the invention
35 have been described herein in detail, it will be
understood by those skilled in the art that variations
. :
, , ~. ,
20112~3
,
. .
.,..`
`~,.
] 109
~1 may be made thereto without departing from the spirit of
-~ the invention or the scope of the appended claims.
..,
.~
,. . .
:~ :
~j
!
' :
' .,
~' . .:
'
~';~ '
1 '~
,1
;, ~
~j :
;.,~; ' ~'
- ~':
.` ~, ' ,,
'.
:
.,.
. ' ....... _. .. , '
