Note: Descriptions are shown in the official language in which they were submitted.
37~
UTSD:048
METHODS AND COMPOSITIONS FOR THE DETECTION
OF FAMILIAL ~YPERCHOLESTEROLEMIA
The present invention is directed towards methods and
compositions useful in the diagnosis of a genetic predis-
position towards the development of hypercholesterolemia,
atherosclerosis, and eventually, heart disease. More
particularly, the present invention is directed towards
recombinant DNA molecules which serve as useful probes in
detecting the presence of mutant low density lipoprotein
(LDL) receptor genes in individuals suspected of having
familial hypercholesterolemia (FH).
Half of all deaths in the U.S. are caused by athero-
sclerosis, the disease in which cholesterol~ accumulating
in the wall of arteries, forms bulky plaques that inhibit
the flow of blood until a clot eventually forms, obstruct-
ing an artery and causing a heart attack or a stroke. The
cholesterol of atherosclerotic plaques is derived ~rom
particles called low-density lipoprotein (LDL) that circu-
.
`
' ~.
.
,
- . . : '
13~
--2--
late in the bloodstream. The more LDL there is in the
blood, the more rapidly atherosclerosis develops.
Epidemiologic data reveal the surprising fact that
more than half of the people in Western industrialized
societies, including the U.S., have a lsvel of circulating
LDL that puts them at high risk for developing athero-
sclerosis. Because such concentrations are so prevalent,
they are considered "normal," but clearly they are not
truly normal. They predispose to accelerated atheroscler-
osis and heart attacks or strokes.
Some answers as to why the LDL levels are so danger-
ously high in many Americans are emerging from studies of
specialized proteins, called LDL receptors. These recep-
tors project from the surface of animai cells. The recep-
tors bind LDL particles and extract them from the fluid
that bathes the cells. The LDL is taken into the cells
and broken down, yielding its cholesterol to serve each
cell's needs. In supplying cells with cholesterol the
receptors perform a second physiological function, which
is critical to the prevention of atherosclerosis: they
remove LDL from the bloodstream.
The number of receptors displayed on the surface of
cells varies with the cells' demand for cholesterol. When
the need is low, excess cholesterol accumulatesi cells
make fewer receptors and take up LDL at a reduced rate.
This protects cells against excess cholesterol, but at a
high price: the reduction in the number of receptors
decreases the rate at which LDL is removed from the circu-
lation, the blood level of LDL rises and atherosclerosis
is accelerated.
It has been proposed that the high level of LDL in
many Americans is attributable to a combination of ~actors
,: .
~2~3~37~3
--3--
that diminish the production of LDL receptors. Recogni-
tion of the central role of the receptors has led to a
treatment for a severe genetic form of atherosclerosis,
and it has also shed some light on the continuing contro-
versy over the role of diet in atherosclerosis in thegeneral population.
LDL is a large spherical particle whose oily core is
composed of some 1,500 molecules of the fatty alcohol
cholesterol, each attached by an ester linkage to a long-
chain fatty acid. This core of cholesterol esters is
enclosed in a layer of phospholipid and unesterified cho-
lesterol molecules. The phospholipids are arrayed so that
their hydrophilic heads are on the outside, allowing the
LDL to be dissolved in the blood or intercellular fluid.
Embedded in this hydrophilic coat is one large protein
molecule designated apoprotein B-lO0.
It is apoprotein B-lO0 that is recognized and bound
by the LDL receptor, a glycoprotein (a protein to which
sugar chains are attached). The receptor spans the thick-
ness of the cell's plasma membrane and carries a binding
site that protrudes from the cell surface. Binding takes
place when LDL is present at a concentration of less than
10 9 molar, which is to say that the receptor can pick out
a single LDL particle from more than a billion molecules
of water. The receptor binds only lipoproteins carrying
apoprotein B-100 or a related protein designated apopro-
tein E.
In 1976 it was discovered that the LDL receptors are
clustered in specialized regions where the cell membrane
is indented to form craters known as coated pits (because
the inner surface o~ the membrane under them is coated
with the protein clathrin). Within minutes of their form-
ation the pits pouch inward into the cell and pinch off
' ' -
78
-4-
from the surface to form membrane-bounded sacs called
coated vesicles; and LDL bound to a receptor is carried
into the cell. Receptor-mediated endocytosis, the term
applied to this process of uptake through coated pits and
vesicles, is now being recognized as a general mechanism
whereby cells take up many large molecules, each having
its own highly specific receptor.
Eventually the LDL is separated from the receptor
(which is recycled to the cell surface) and is delivered
to a lysosome, a sac filled with digestive enzymes. Some
of the enzymes break down the LDL's coat, exposing the
cholesterol ester core. Another enzyme clips off the
fatty acid tails of the cholesterol esters, liberating
unesterified cholesterol, which leaves the lysosome. All
cells incorporate the cholesterol into newly synthesized
surface membranes. In certain specialized cells the
cholesterol extracted from ~D~ has other roles. In the
adrenal gland and in the ovary it is converted into
respectively the steroid hormones cortisol and estradiol;
in the liver it is transformed to make bile acids, which
have a digestive function in the intestine.
The central role of the LDL receptor in
atherosclerosis was first appreciated in 1974 when it was
shown that absence of the receptor was responsible for the
severe disease called familial hypercholesterolemia (FH).
Much earlier, in 1939, Carl Muller of the Oslo Community
Hospital in Norway identified the disease as an inborn
error of metabolism causing high blood cholesterol levels
and heart attacks in young people; he recognized that it
is transmitted as a dominant trait determined by a single
gene. In the 1960's two forms of the disease were
delineated, a heterozygous form and a more severe
homozygous form. Heterozygotes, who inherit one mutant
gene, are ~uite common: about one in 500 people in most
-5-
ethnic groups. Their plasma LDL level is twice the normal
level (even before birth) and they begin to have heart
attacks by the time they are 35; among people under 60 who
have heart attacks, one in 20 has heterozygo~ls FH.
If two FH heterozygotes marry (one in 250,000 mar-
riages), each child has one chance in four of inheriting
two copies of the mutant gene, one from each parent. Such
FH homozygotes (about one in a million people) have a
circulating LDL level more than six times higher than
normal; heart attacks can occur at the age of two and are
almost inevitable by the age of 20. It is notable that
these children have none of the risk factors for athero-
sclerosis other than an elevated LDL level. They have
normal blood pressure, do not smoke and do not have a high
blood glucose level. Homozygous FH is a vivid experiment
of nature. It demonstrates une~uivocally the causal rela-
tion between an elevated circulating LDL level and athero-
sclerosis.
Heterozygotes with familial hypercholesterolemia can
often be suspected at birth because blood plasma from the
umbilical cord contains a two- to three-fold increase in
the concentration of LDL-cholesterol. The elevated levels
of plasma LD1 persist throughout life, but symptoms typi-
cally do not develop until the third or fourth decade.
The most important clinical feature is premature and ac-
celerated coronary atherosclerosis. Myocardial infarc-
tions begin to occur in affected men in the third decade,
showing a peak incidence in the fourth and fifth decades.
By age 60, approximately 85% have experienced a myocardial
infarction. In women the incidence of myocardial infarc-
tion is also elevated, but the mean age of onset is
delayed lO years in comparison to men. Heterozygotes for
familial hypercholesterolemia constitute about 5% of all
patients who have a myocardial infarction.
~13C3 3~
--6--
Xanthomas of the tendons are the second major clini-
cal manifestation of the heterozygous state. These
xanthomas are nodular swellings that typically involve the
Achilles and other tendons about the knee, elbow, and
dorsum of the hand. They are formed by the deposition of
LDL-derived cholesterol esters in tissue macrophages
located in interstltial spaces. The macrophages are swol
len with lipid droplets and form foal cells. Cholesterol
is also deposited in the soft tissue of the eyelid, pro-
ducing xanthelasma, and within the cornea, producing arcuslipoides corneae. Whereas tendon xanthomas are essen-
tially diagnostic of familial hypercholesterolemia,
xanthelasma and acrus lipoides corneae are not specific.
The latter abnormalities also occur in many adults with
normal plasma lipid levels. The incidence of tendon
xanthomas in familial hypercholesterolemia increases with
age. Eventually, about 75% of affected heterozygotes
display this sign.
Homozygote individuals have marked elevations in the
plasma level of LDL from birth. A unique type of planar
cutaneous xanthoma is often present at birth and always
develops within the last six years of life. These
cutaneous xanthomas are raised, yellow plaque like lesions
that occur at points of cutaneous trauma, such as over the
knees, elbows, and buttocks. Xanthomas are almost always
present in the interdigital webs of the hands, particu-
larly between the thumb and index finger. Tendon
xanthomas, arcus lipoides corneae, and xanthelasma are
also characteristic. Coronary artery atherosclerosis
frequently has its clinical onset in homozygotes before
age lO, and myocardial infarction has been reported as
early as 18 months of age. In addition to coronary
atherosclerosis, cholesterol is fre~uently deposited in
the aortic valve~ producing symptomatic aortic stenosis.
3~7~
--7--
Homozygotes usually succumb to the complications of myo-
cardial infarction before age 30.
One in 500 persons in most populations has a mutation
in the LDL receptor gene that destroys the function of the
~ene and produces the clinical syndrome of heterozyyous
familial hypercholesterolemia. Many of these mutant genes
fail to produce any detectable receptors. Other mutant
genes produce a small number of receptors; still other
mutant genes produce essentially normal numbers of
defective receptors that do not bind LDL properly. In
other affected individuals, an LDL receptor protein of
abnormal length is produced. Whereas the normal precursor
form of the receptor displays an apparent molecular weight
of approximately 120,000 daltons when measured by gel
electrophoresis, aberrant forms of the protein encoded by
mutant genes have been identified that migrate at apparent
molecular weights of 100,000, 135,000 and 170,000 daltons.
It is possible that such mutations observed in the LDL
receptor protein may be the result of deletion or
insertion mutations in the gene responsible for LDL
receptor production. All of the above described mutations
are felt to reside in or near the gene for the LDL
receptor. Thus, a means of identifyin~ directly those
individuals who carry a mutant LDL receptor gene would
greatly facilitate our ability to identify those
individuals with a genetic predisposition towards
developing atherosclerosis and heart disease. These
mutations could be identified if a complementary DNA
(cDNA) for the receptor gene were discovered.
3~
--8--
Figure 1 is a schematic representation of the cDNA
cloning strategy used in cloning the human LDL receptor.
Human ~hl is a partial genomic clone correspondin~ to the
3' end of the LDL receptor gene. Recombinant plasmids
plO1 and p203 contain partial cDNA's which are comple-
mentary to the human LDL receptor mRNA. Plasmid pLDLR-2
is a fusion construct of plO1 and p203 and contains a
nearly full-length cDNA for the human LDL receptor mRNA.
The numbers in parentheses refer to the lengths of the DNA
inserts in a given clone.
Figure 2 is a structural representation of recombin-
ant plasmid pLDLR-2 which contains a nearly full-length
cDNA for the human LDL receptor gene. The coding region
(hatched area) encompasses nucleotides 1 to 2580. This
fusion plasmid was constructed by joining the cDNA inserts
of p203 (nucleotides 1 to 323) and plOl (nucleotides 324
to 5144) via overlapping Hgi A1 sites. The solid areas in
pLDLR-2 denote regions in the cloning vector that contain
S~40 sequences, including the origin of replication, 16S
and l9S donor and acceptor splicing sites, and polyadenyl-
ation signals.
Figure 3 displays the nucleotide sequence of the cDNA
corresponding to the human LDL receptor mRNA and the pre-
dicted amino acid sequence of the receptor protein. The
nucleotides are numbered on the right-hand side in the
5'-to-3' direction; nucleotide 1 is the A of the ATG codon
that encodes the initiator methionine; negative numbers
refer to the 5' untranslated region. The amino acids are
numbered underneath the sequence; residue l is the alanine
found at the NH2 terminus of the mature protein; negative
numbers refer to the cleaved signal sequence. The signal
sequence (21 residues) and the membrane-spanning sequence
(22 residues) are indicated by single solid underlines.
37~3
g
The sites to which N-linked carbohydrate could be attached
(Asn-X-Ser or Asn-X-Thr) are indicated by double solid
underlines. Cysteine residues are circled. A stretch o~
48 residues rich in serines and threonines t~ which 0-
linked carbohydrate could be attached is indicated by thedotted underlines. The Alu sequences in the 3' untrans-
lated region of the cDNA are boxed; the direct repeats
associated with the first Alu sequence are indicated by
dotted arrows above the sequence. Three potential poly-
adenylation signals in the 3' untranslated region areindicated by overlines and underlines.
Figure 4 is an analysis of Xba I restriction digests
of genomic DNA from a normal subject and an individual
with familial hypercholesterolemia ~designated FH274) with
DNA probes from different regions of the LDL receptor
cDNA. Figure 4A is a diagram of the mRNA for the human
LDL receptor which is shown with AUG and UGA indicating
the beginning and end of the translated region,
respectively. The sizes and locations of the 32P-labeled
cDNA probes used to map the LDL receptor gene are shown by
the closed bars and are numbered 1-9. Figure 4B is an
autoradiograph of a Southern blot of either normal or
mutant (FH 274) genomic DNA following hybridization with
either probe 1 (left) or probe 8 (right). The Xba I
restriction fragments are designated A-D along the right
side of each blot. Molecular size standards were
generated by Hind III cleavage of bacteriophage DNA and
are indicated to the left of each blot. The Xba I
fragments detected by probes 2-9 in the normal subject and
FH 274 are indicated at the bottom of Figure 4A.
Figure 5 is a Southern blot hybridization of XbaI-
cleaved genomic DNA from a normal subject, ~H 274 and his
parents. Genomic DNA (5 ug) isolated from cultured
fibroblasts from the indicated subject was digested ~ith
,,: '
~10-
XbaI, electrophoresed, transferred to nitrocellulose, and
hybridized with 32P~probe 1. The four relevant Xb I
restriction fragments are designated A-D along the right
side of the blot. Molecular size standards were generated
by HindIII cleavage of bacteriophage ~DNA.
Figure 6 is a comparison of restriction maps of the
3' end of the normal LDL receptor gene and the deletion-
bearing gene from FH 274. The scale at the bottom indi-
cates the length of genomic DNA in kilobases. The organization of the normal LDL receptor gene is shown in the
diagram at the top. Exons are indicated by solid segments
and upper case letters; intervening sequences (IVS) are
indicated by open segments and lower case letters. The
Alu repetitive sequences in IVS c and Exon E are indi-
cated. Restriction enzyme recognition sites used to de-
fine the gene deletion in FH 274 are shown.
37~
The present invention discloses a technique suitable
for the construction of recombinant DNA transfer vectors
which contain a cDNA sequence corresponding to the mRNA
sequence of the human Low Density Lipoprotein (LDL~ recep-
tor gene. In addition, the present invention discloses
recombinant DNA transfer vectors that contain DNA inserts
which are complementary to various portions of either the
human LDL receptor gene or the mRNA transcript of that
gene.
Recombinant DNA transfer vectors disclosed by the
present invention need not necessarily contain the entire
human LDL receptor gene in order to be useful in the prac-
tice of the invention. Similarly, the recombinant trans-
fer vector need not contain a DNA fragment which is com-
plementary to the entire mRNA transcript for that gene.
Recombinant transfer vectors may be constructed of smaller
subfragments which are complementary to either the human
LDL receptor gene or the mRNA for that gene. In fact, the
use of different subfragments of the gene as probes may be
necessary for detailing specific mutations in certain FH
individuals. A11 that is required is that these fragments
be of sufficient length to form a stable duple~ or hybrid.
Such fragments are said to be "hybridizable" in that they
are capable of stable duplex formation. Generally, DNA
fragments at least fourteen nucleotides in length are
capable of forming stable duplex's (i.e. - a
tetradecamer).
Individuals with Familial Hypercholesterolemia (FE)
are diagnosed using the present invention by determining
the presence of a mutation in the gene which codes or the
LDL receptor. DNA isolated from the recombinant cDNA
clones has been used to diagnose one patient with familial
hypercholesterolemia who has a deletion in the gene. ~he
'
~213~37~
-12-
cDNA, or fragments thereof, should also be useful in
diagnosing other mutations, including those resulting from
single nucleotide changes (point mutations).
In general, the method consists of fragmentiny the
DNA from cells of an individual who is suspected o~ having
a mutation followed by separating the DNA fragments into a
pattern according to some physiochemical property of the
DNA, for example, molecular weight or size of the DNA
fragments. The separated DNA is then hybridization probed
with labeled LDL receptor DNA in order to identify those
fragments of DNA from the individual which correspond to
the LDL receptor gene. Then, by comparing the pattern of
LDL receptor gene fragments of the suspected individual to
a similar ragment pattern from a normal individual, it
can be determined whether the suspected individual dis-
plays a mutation. If the pattern of ~DL receptor gene
fragments identified in the suspected individual exhibits
an alteration relative to the control pattern, a gene
mutation has been detected.
One method which has proved particularly useful in
fragmenting the DNA utilizes restriction enzyme digestion.
However, other methods, including chemical cleavage of the
DNA, could also be used providing that such methods are
capable of reproducibly cleaving genomic DNA into the same
discrete ragments.
The fragmented DNA can be separated into a recogniz-
able pattern using various methods, the most useful owhich take advantage of the varying sizes of the discrete
DNA. For example, DNA fragments can be separated accord-
ing to molecular weight by velocity sedimentation through
a density gradient, or, by molecular size by gel exclusion
chromatography. However, for the purposes o~ the present
invention, the preferred techni~ue is to separate the DNA
~L21~03~
-13-
fragments by electrophoresis through an agarose or poly-
acrylamide gel matrix.
The cloned LDL receptor DNA can be conveniently
labeled with radioactive nucleides which allow ~or ready
visualization of the corresponding genomic LDL receptor
DNA fragment pattern after hybridization and autoradio-
graphy. Other labeling techniques, including for example,
heavy isotopes, would be possible but would prove cumber-
some in practice as a means of identifying the correspond-
ing genomic sequences.
In addition to the use of cloned DNA fragments
diagnosis can in principle be made with chemically
synthesized oligonucleotides that correspond to portions
of the cDNA that are disclosed herein. Genomic DNA from
individuals with single base substitutions in the LDL
receptor gene will hybridize to such oligonucleotides less
strongly than does DNA from a normal individual. Such
weakened hybridization will therefore be a method of
diagnosis of many patients with FH in both the
heterozygous and homozygous forms.
~3633~
-14-
The low density lipoprotein (LDL) receptor is a cell
surface protein that plays a central role in the metabo-
lism of cholesterol in humans and animals. Through the
process of endocytosis, the LDL receptor is responsible
for binding serum cholesterol and making it available for
cellular metabolism. This is made possible by interlla.li-
zation of the receptor~cholesterol complex, the choles-
terol then being liberated by catabolism of the inter-
nalized complex. The liberated cholesterol regulates, viaa feedback mechanism, the rate of synthesis of the LDL
receptor. The increased demand for cholesterol in certain
steroidogenic tissues, such as the adrenal cortex and the
ovarian corpus luteum, is met by an increased number of
LDL receptors. A foremost distinguishing feature of the
LDL receptor is that mutations affecting its structure and
function give rise to one of the most prevalent human
genetic diseases, familial hypercholesterolemia.
Recombinant DNA technology provides one approach to
detecting the presence of mutations in an individual
suspected of having FH. By using a probe consisting of a
purified human LDL receptor gene, or a hybridizable
subfragment thereof, abnormalities present in a particular
individual's LDL receptor gene can be identified and that
individual can then be targeted for other types of therapy
aimed at addressing the symptoms of FH. That is, once FH
individuals are identified through processes disclosed by
the present invention, these individuals can be targeted
for therapy, including both diet modification, pharma-
cologic approaches, and surgery, all aimed at reducing the
levels of circulating cholesterol in these individuals.
In addition, knowledge concerning the genetic structure of
the LD~ receptor gene in EH individuals could eventually
lead to more dramatic therapeutic approaches to the
~36337~3
-15-
disease, including somatic gene replacement or
modification.
The first step in understanding and identiying the
underlying genetic abnormalities in FH individuals is
through the development of suitable probes, using genetic
engineering techniques, by which both normal and abnormal
LDL receptor genes may be studied. The most ideal genetic
probe for studying the structure of the human LDL receptor
gene would be a cloned human LDL receptor gene or a cDNA
prepared to the receptor mRNA. However, there are
difficulties in approaching this problem directly in that
it would require ~solation of the probe from some human
source. This source would preferably be an adrenal or
ovarian source where the receptor, and its mRNA, are in
greater abundance. This approach is somewhat impractical
in that it is difficult to obtain sufficient amounts of
the appropriate human tissues. The present invention
describes a fortuitous approach whereby the LDL receptor
cDNA is cloned from a bovine source and the cloned bovine
LDL receptor cDNA is then used to isolate part of the
human gene from a gene library. The part of the human
gene is then used to isolate nearly full-length human cDNA
clones. This was fortuitous in that it was not apparent
until after the human cDNA was cloned and sequenced, that
the bovine sequences which were used as hybridization
probes could be used to correctly probe for the human LDL
receptor gene. In retrospect, the homology between the
bovine and human gene was sufficient enough to allow for
the cloning approach detailed herein.
The isolation of human recombinant DNA clones, bear-
ing copies of the human LDL receptor mRNA, facilitated the
development of an assay whereby LDL receptor gene
mutations could be detected. To illustrate the utility of
this assay, a case study of an individual afflicted with
378
-16-
familial hypercholesterolemia (hereinafter designated FH
274) was undertaken. Alth~ugh the case study described
herein goes into extreme detail as to the structure of
that individuals specific genetic defect, it is not con-
templated by the present invention that such measures willbe necessary as a diagnostic approach. However, this
material is included herein to illustrate the power of
these genetic techniques which are made available by the
present invention. Moreover, they provide a means whereby
many other defects in the LDL receptor gene may be
diagnosed.
Thus, the present invention provides a method whereby
not only may the presence of a genetic defect in the LDL
receptor gene be identified but, in addition, the particu-
lar genetic defect may be detailed with great specifica-
tion.
The present inventors feel that these techniques
provide a method whereby segments of the population as a
whole may be screened for the presence of genetic defects
in the LDL receptor gene. Once these individuals are
identified through the screening procedures detailed
herein, the specific abnormalities exhibited by the mutant
LDL receptor gene can then be studied in great detail.
Therefore, the present Applicants fael that the present
invention will lead to a greater future understanding o~
the genetic events which give rise to the serious and
prevalent human genetic disease, Familial
Hypercholesterolemia.
3~
-17-
EXAMPLE I
CLONING OF THE BOVINE LDL RECEPI'OR GENE cD~~A
A Bovine LDL receptor gene cDNA clone, hereina~ter
designated pLDLR-l, was isolated using a combination of
polysome immunopurification and oligonucleotide hybridi-
zation. Generally, the technique proceeds through five
steps. These steps consist of (1) isolation of the bovine
receptor protein, (2) generation of a polyclonal antibody
capable of reacting with the bovine LDL receptor protein,
13) specific immunoprecipitation of those bovine polysomes
which are actively engaged in translation of the bovine
recep~or mRNA, followed by enriching for the receptor mRNA
isolated from the precipitated polysomes, (4) preparation
of a cDNA clone bank from the enriched mRNA, (5) screening
of the cDNA clone bank to isolate a representative clone
bearing a ~ovine LDL receptor cDNA insert. These steps
are described in detail as follows.
Bovine ~eceptor Isolation and Generation
Of A Polvclonal Antibody
Homogeneous LDL receptor protein was isolated from
bovine adrenal cortex as described by Schneider, et al.,
J. Biol. Chem., 257:2664-2673 (1982). A polyclonal
antibody against the bovine adrenal LDL receptor was
raised in rabbits and purified on staphylococcal
protein A-sepharose as described by Tolleshaug, et al.
3~ Cell, 30:715-724 (1982). This antibody and its corres-
ponding non-immune rabbit IgG were free of gross RNase
contamination as shown by their failure to alter -the
sedimentation behaviour of polysomes or sucrose gradients.
3~
-18-
Polysome Immunoprecipitation of Bovine ~DL ReceDtor mRNA
Polysomes enriched in mr~NA for the LDL receptor were
prepared as ollows. Bovine tissue was frozen in liquid
nitrogen within 5 min of slauyhter. Adrenal glands were
powdered in liquid nitrogen in a Waring ~lender and stored
at -70C prior to polysome isolation.
Ten-gram aliquots of powdered adrenals were homo-
genized with a Brinkmann Polytron in 42 ml of 25 mM Tris-
HCl, pH 7.5/25 mM NaCl/5 mM MgC12/2% (vol/vol) Triton X-
100/0.3 mg of heparin per ml/l ug of trichodermin per
ml/60 ug of phenylmethylsulfonyl fluoride per ml. Poly-
somes were isolated from the homogenate by MgC12 precipi-
tation as described by Palmiter, Biochemistry, 13:3606-
3615 (1974~, and stored at -70 . Twenty-five A260 units
of polysomes were obtained per gram of adrenal powder.
.. . .
On linear sucrose gradients approximately 70% of the
A260 material sedimented as polysomes; the remaining ab-
sorbance was present in 80S monosomes. Polysomes (1,000
A260 units) were clarified with a 10 minute centrifugation
at 20,000 x g, then diluted to 15 A260/ml in a buffer con-
25 taining 25 mM Tris-HC1 at pH 7.5, 150 mM NaCl, 5 mM MgC12
0.1% Nonidet P-40, heparin at 0.2 mg/ml, and trichodermin
at 1 ug/ml and incubated with 6.25 mg of anti-receptor IgG
or non-immune IgG for 1 hour with stirring at 4C.
The polysome/antibody slurry was then passed twice
through a column of protein A-Sepharose (0.7 x ~3 cm)
equilibrated in the above dilution buffer at a flow rate
of 8-10 ml/hour at 4C. The column was washed overnight
with 120 ml of dilution buffer. Bound polysomes were
eluted at a maximal flow rate with 20 ml of 25 mM Tris-
HCl, pH 7.5/20 mM EDTA. The eluted fraction was heated S
37~
--19--
minutes at 65~C, brought to 0.5 M NaCl and 0.2% NaDodS04
cooled to 24C, and passed through a column of oligo(dT)-
cellulose (0.8 x 2.3 cm) equilibrated in 10 mM Tris-HC1,
pH 7.5/0.5 M NaCl. The column was washed with 20 ml of
this buffer and poly(A) RNA was eluted With 5 ml of 1~ mM
Tris-HC1, pH 7.5. Yeast carrier tRNA (50 ug) was added,
and the RNA was precipitated twice with NaOAc and ethanol.
This immunopurified poly(A) RNA was resuspended in 20 ul
of water and stored at -70C.
The immunoselected poly(A) RNA was assayed for the
presence of LDL receptor mRNA by in vitro translation in a
reticulocyte lysate system. The lysate system is de-
scribed as follows.
Aliquots of poly(A) mRNA were incubated with 2.5 mM
CH3HgOH for 10 min at 4C and then translated in rabbit
reticulocyte lysates prepared as described by Pelham and
Jackson, Eur. J. ~iochem., 67:247-256 (1970), and supple-
mented with 80 mM KOAc, 1 mM Mg(OAc)2, 19 amino acids
(excluding methionine) at 16 uM each, and [35S]methionineat 0.2 mCi/ml ~1 Ci = 3.7 x 10 Bq). The final concentra-
tion of CH3HgOH in the translation reaction was 0.3 mM.
Translation products were analyzed by electrophoresis on
NaDodSO4/7% polyacrylamide gels.
Total adrenal gland poly(A) RNA directed the synthe-
Si5 of many proteins, as determined by NaDodS04 gel elec-
trophoresis and fluorography of the synthesized products.
Poly(A) RNA derived from the immunopurified polysomes
directed synthesis of several of the same protein bands
plus one clear addition: a protein that migrated with a
Mr of approximately 120,000. This protein was not
demonstrable after translation of poly(A) RNA selected
from polysomes with nonimmune IgG. Biosynthetic studi~s
on the LDL receptor from hamsters and rabbits have show~
3~
-20-
th2t the receptor is initially made as an apparent 120,000
Mr precursor (where Mr stands for molecular weight in
daltons) that undergoes a series of posttranslational
glycosylation events during transport to the cell surface,
resul~ing in a mature protein with an apparent Mr f
160,000. Thus, the size of the enriched protein seen
after translation of the immunoselected poly(A) RNA was
consistent with that of the LDL receptor precursor.
Preparation And Screening Of The Bovine cDNA Clone Bank
The immunoselected poly(A) RNA was used to construct
a cDNA library by the method of Okayama and Berg, Mol.
Cell. Biol., 2:161-170 (1982) from
poly(A)+ RNA derived from 2,000 A260 units
of polysomes. In the cloning reactions, which employed
enzymes obtained from Life Sciences and P-L Biochemicals,
1.~ ug of dT-tailed vector primer and 0.52 pmol of dG-
tailed linXer were used.
Portions of the cDNA library were used to transformEscherichia coli RRl to ampicillin resistance by the
CaC12-shock procedure described in Maniatis, et al.
Molecular Cloning (Cold Springs Harbor Laboratory, Cold
Spring Harbor, N.Y., page 250).
~olonies were plated at high density on
nitrocellulose filters, and two replica filters were
prepared for hybridization (Maniatis, su~ra, p. 316). To
reduce nonspecific background, baked filters were washed
overnight in 50 mM Tris-HCl, pH 8/i ~ EDTA/l M NaCl/0.1%
NaDodSO4 at 37 or a2C and then incubated at 65C for 3 hr
in 4 x SSC (1 x SSC = 0.15 M ~aCl/15 mM sodium citrate),
10 X Denhardt's solution (lX = 0.02%
polyvinylpyrrolid~ne/0.02% bovine serum albumin/0.02%
.~,...
- '
~8~,3~
-21-
Ficoll) (Maniatis, supra, p. 327), and sonicated and
denatured E. coli DNA at lO00 ug/ml.
Hybridization was performed overnighk in the latter
solution containing 32p_5'-end-labeled oligonucleotide
mixtures (6 X 106 cpm/pmol at 1 pmol/ml) prepared a~
described below. Hybridization temperature for a given
oligonucleotide probe corresponded to the minimum melting
temperature, tm, calculated from the empirical formula tm
= 2C x (number of dA dT bp) + 4C x (number of dG dC bp),
in which bp is base pairs. Filters were washed three
times in 4 x SSC at the hybridization temperature for 30
minutes per wash, dried at room temperature, and subjected
to autoradiography. Positive clones were picked from the
master plate and purified through several rounds of
screening.
The hybridization probes used in the screening proto-
col detailed above were prepared based on se~uence infor-
mation obtained from peptide subfragments of the LDLreceptor protein as follows.
The purified LDL receptor was digested with CNBr, an
internal CNBr fragment was isolated by high pressure
liquid chromatography (HPLC), and its partial amino acid
sequence was determined by automated Edman degradation.
CNBr fragments were generated from two different prepara-
tions of reduced and [3H] carboxymethylated receptor (1.6
and 1.8 mg of protein) and fractionated by reverse-phase
HPLC on a Brownlee (Santa Clara, CA) RP 300 column. The
CNBr peptide described here was subjected to two separate
runs on an automated Beckman 890C sequencer using a 0.25 M
Quadrol program and the nonprotein carrier Polybrene.
Yields of the NH2-terminal residue of the C~Br peptide
were 400 and 1,100 pmol for the two runs. Repetitive
-22-
yields, calculated on the basis of recovery of the phenyl-
thiolhydantion of [3H] cysteine, averaged 91%.
Two families of synthetic oligonucleotide probes that
corresponded to all possible codons specifyiny the
sequence of amino acids in two neighbori.ng regions of this
CNBr fragment were synthesized. One family of oligonuc-
leotides, designated A in Table I consisted of 32
tetradecamers encoding (Met)-Ala-Glu-Asn-Leu. The exis-
tence of a methionine residue at the amino terminus ofthis sequence was inferred from the fact that the peptide
was generated by CNBr digestion. A second family of
tetradecamers, designated B and B in Table I, encoded the
sequence Pro-Glu-(Asp)-Ile-Val. The assignment of the Asp
residue in this sequence was provisional because it was
observed in only one of two sequenator runs. The B/B
oligonucleotide family consisted of a total of 48 members
that were synthesized as two subfamilies of 24 each, dif-
fering only in the codons used to specify the Pro residue
(CCT in B and CCG in B ).
-23- ~2~378
TABLE I
CNBr Peptide Sequence and Oligonucleotide Synthesis
Met Ala Glu Asn Leu Leu Ser Pro Glu (Asp) Ile Val
C C C
T A C C T A C T
ATGGC GA AA T CC GA GA AT GT
A G T T A G T A
G G
Oligo Family A Oligo Families B and B
32 x 14 mers 2 x 24 x 14 mers
Thirty hybridization positive cDNA clones were
identified by screening the above cDNA library with
oligonucleotide family B. When these clones were probed
separately with the subfamilies B or B , 16 clones
hybridized strongly with oligonucleotide mixture B, but
not with B . Twelve of the 30 clones were positive only
*
with mixture B . These 28 positive clones were then
screened with oligonucleotide mixture A, and two plasmids,
both from the latter group of 12, hybridized with this
probe. These two clones were considered to contain cDNAs
for the receptor and were chosen ~or further study.
Plasmid DNAs from the two clones that hybridized to
both the B and A oligonucleotide probes were subjec-ted to
restriction endonuclease mapping, and the results indi-
cated that these two clones were identical. Therefore,
these clones were considered to be representative bovine
LDL receptor cDNA clones. One of these clones, designated
-24- ~ 7~
pLDLR-l, was chosen in order to confirm that it repre-
sented a true LDL receptor clone.
To confirm the identity of pLDLR-l, total poly(A~
RNA was extracted from bovine adrenal glands and liver and
analyzed in blotting experiments using nick-translated
32P-labeled plasmid as a probe. RNA blottin~ experiments
were performed as follows. Total RNA was isolated by
treatment of tissues or cells with guanidinium thiocyanate
(Maniatis, sUPra, page 196). Poly(A) RNA was purified by
oligo(dT)-cellulose chromatography, denatured with
glyoxal, size-fractionated by electrophoresis (20 volts
for 16 hours) on 1.5~ agarose gels containin~ 40 mM 3-N-
morpholinopropanesulfonic acid (pH 7.0), and then trans-
ferred to Zeta Probe membranes (Bio-Rad) by capillary
blotting in 20 x SSC. Prehybridization and hybridization
were carried out as described by Maniatis, ~pra, page
320.
Increasing amounts of adrenal gland RNA yielded a
progressively stronger hybridization signal corresponding
to a mRNA of approximately 5.5 kb. Densitometric scanning
showed that the signal obtained with a given amount of
adrenal RNA was 9-fold more intense than that obtained
with the same amount of liver RNA. Previous studies have
shown that functional hDL reGeptor activity is about one
order of magnitude more abundant in bovine adrenal than in
bovine liver, a finding that coincides with the difference
in the abundance of the mRNAs discussed above.
The number of LDL receptors can be markedly reduced
when cultured cells are grown in the presence of choles-
terol or related sterols. Poly(A) RNA was isolated from
human A-431 cells grown in the absence of sterols (recep-
tor-induced) and presence of sterols (receptor-suppressed)
and analyzed by blotting with pLDLR-l. A strong hybridi-
3~3
-25-
zation signal from a mRNA of approximately 5.5 kb was
detected in the induced RNA and this signal was reduced by
more than 90% in the suppressed RNA.
These results indicate that pLDLR-l contains a cDMA
copy of at least a portion of the bovine LDL receptor
gene. This clone has been deposited as ATCC #39965.
Fortunately, there is sufficient homology between the
bovine receptor gene and the human gene to allow the use
of the bovine se~uence as a probe in isolating the human
gene. These procedures are disclosed in Example II.
EXAMPLE II
Cloning and Characterization of the
Human LDL Receptor Gene
The strategy used to obtain a full-length cDNA for
the human LDL receptor is outlined in Figure 1. The par-
tial bovine cDNA (pLDLR-l) was used to screen a human
genomic library cloned in bacteriophage by the procedure
of Lawn, et al., Cell, 15:1157-1174 (1978)
as follo~s. Approximately l x 106
bacteriophages containing human genomic DNA inserts were
screened with 32P-labeled pLDLR-l. Hybridization was
performed under conditions of reduced stringency: 30%
formamide, 5 x SSC, 5 x Denhardt's solution, 0.1% SDS
(Sodium Dodecyl Sulfate), 100 ug/ml salmon sperm DNA and 1
ug/ml poly(A) at 42C. Filters were washed twice at 22C
for 10 minutes in 2 x SSC, 0.1% SDS (sodium dodecyl
sulfate), and once at 54C for 60 minutes in the same
solution.
From this human genomic clone library, a single
clone, hl, ~as identified. It contained an 11 kb (kilo-
base pair) insert encoding the 3' end of the human LDL
37~3
-26-
receptor gene. ~ hl was used to generate a 236 bp Pvu II
fragment that served as a unique probe for the COOH-termi-
nal end of the human l,DL receptor cDNA.
A cD~A library was constructed from human fetal
adrenal poly(A) RNA by the method of Okayama and Berg,
Mol. Cell, Biol., 2:161-170 (1982).
In the cloning reactions, we used commercially
obtained enzymes, 2 ug of poly (A) RNA, 1.5 ug of dT-
tailed pcDVl vector-primer, and 0.52 pmole of dG-tailed
linker. Following transformation into E. coli ~BlOl,
plasmid cDNAs were isolated from approximately 3 X 105
transformants and enriched for longer cDNAs (6 to 10 kb)
by the sublibrary method of Okayama and Berg, Mol. Cell.
Biol., 3:280-289 (1983).
Human LDL receptor cDNAs were identified using two
probes, a 5'-specific oligonucleotide family consisting of
32 heptadecamers derived from the sequence Asn-Glu-Phe-
Glu-Cys-Gln, present at the MH2 ter~inus of the bovine LDL
receptor protein and the 3'-specific 236 base pair Pvu II
fragment containing a 158 bp COOH-terminal exon that was
derived from the human genomic clone~ hl (described
above). Oligonucleotides were synthesized by the phospho-
ramidite method and provided by Mark Zoller (Cold SpringHarbor Laboratory) and Ray MacDonald (University of Texas
Health Science Center at Dallas). Replica filters were
screened, and five colonies out of 2396 recombinants were
positive with both the 5'- and 3'-specific probes. The
plasmid with the longest cDNA insert in these positive
clones (~.9 kb) was designated plOl.
The nucleotide sequence of the cDNA insert of plOl
revealed that it did not contain the 5' end of the coding
region of the LDL receptor mRNA. Nucleotide sequence
-analysis of this cDNA revealed that the 5'-oligonucleo~ide
-27~ 37~
probe was not hybridizing to the region corresponding to
the extreme NH2 terminus of the protein. Rather, the
probe was hybridizing to an imperfect repeat of the NH2-
terminal sequence that occurred within ~he coding reyion,
The open reading frame continued to the extr~me 5' end o
the cDNA insert in plO1, and there was no evidence of a
predicted signal sequence or an initiator methionine
codon. Therefore, plO1 did not contain the entire coding
region.
To obtain the rest of the coding region, an oligo-
nucleotide corresponding to a sequence near the 5' end of
the cDNA insert in plO1 was prepared and primer extension
using human fetal adrenal poly~A) RNA as a template was
performed. A synthetic oligonucleotide of 20 bases com-
plementary to the mRNA strand and originating 63 nucleo-
tides from the 5' end of the cDNA insert of plO1 was used
to construct a primer-extended cDNA library from human
fetal adrenal poly(A) RNA in pBR322. This library was
screened with a second oligonucleotide of 20 bases that
originated 8 nucleotides from the 5' end of the cDNA
insert of plO1. Of the 1044 recombinants screened, one
plasmid, designated p203, was identified whose cDNA insert
overlapped that of plOl for ~3 nucleotides and extended to
near the approximate 5' end of the human LDL receptor
mRNA.
To construct a nearly full-length cDNA, it was
necessary to ligate the appropriate portions of plO1 and
p203 (Figure 2). Owing to a paucity of convenient
restriction enzyme sites, this ligation required several
partial digests and the preparation of two intermediate
plasmids as follows.
A cDNA containing the entire translated region of the
human LDL receptor mRNA, was constructed by joining the
-28- ~2~378
ineserts o p203 and plO1 via overlapping Hgi A1 sites
(Figure 2). The construction involved three partial di-
gestions, three multifragment ligations, and two inter-
mediate plasmids. p203 was partially digested with Pst 1
and then digested completely with Pvu II to yield a fray-
ment of 341 bp. A Hind III-Pst 1 (518 bp) frayment was
purified from pL 1, which contains the early region
promoter and splicing signals of the SV40 virus (see
Figure 2 and Okayama and Berg, 1983). These two fragments
were ligated and cloned into the Hind III-Pvu II site of
p~R322. This intermediate plasmid was designated pHP1. A
39a bp Hgi AI-Eco R1 fragment from the 5' end of plOl was
mixed with a 105 bp Pvu II-Hgi A1 fragment from the 3' end
of p203 and ligated into Eco R1-Pvu II-digested pBR322.
This intermediate plasmid was designated pEP1. pEP1 was
partially digested with Pvu II and then completely
digested with Eco R1 to yield a 499 bp Pvu II-Eco R1
ragment. This DNA fragment was ligated with the 859 bp
Hind III-Pvu II insert from pHPl and with the 7 kb Hind
III-Eco R1 fragment of plO1 corresponding to the 3'
eighty-five percent of the cDNA and the cloning vector.
The final 8.4 kb plasmid designated p~DLR-2, contained a
full-length cDNA copy of the human LDL receptor Iinked to
SV40 sequences (Figure 2).
The result of this genetic engineering was pLDLR-2, a
plasmid containing a 5.3 kb cDNA insert corresponding to
the entire coding region, the entire 3' untranslated
region, and at least a portion of the 5' untranslated
region of the human LDL receptor mRNA (ATCC #39966).
The cDNA insert of pLDLR-2 was sequenced by the
method of Maxam and Gilbert, Meth. Enzymol., 65:499-500
(1980)and Sanger et al, Proc. Natl. Acad. Sci. USA, 74:
5463-5467 (1977). The nucleotide sequence determined for the
-29~ 37~
human LDL receptor cDNA is displayed in Figure 3 along
with the predicted amino acid sequence of the
corresponding receptor protein. On the 5' side of the 5'
end of the sequence shown in Figure 3, plasmid pLDLR-2
contains extraneous DNA that appeared to arise from the
formation of a hairpin loop during the c~oning reactions.
The presence of this extraneous DNA in no way affects the
utility of this cDNA for the diagnostic purposes outlined
in this application.
EXAMPLE III
ISOLATION OF GENOMIC RECOMBINANT CLONES
CORRESPONDING TO THE NORMAL_LDL RECEPTOR GENE
Genomic clones spanning the 3' end of the human LDL
receptor gene were isolated from a genomic library con-
structed in the bacteriophage vector,~ Charon 4A. The
library was provided us by T. Maniatis, Harvard
University. Recombinant phage were screened with 32p_
labeled pLDLR-l and pLDLR-2,~ cDNA probes for the bovine
and human LDL receptors (see Examples I and II,
respectively). Hybridization was performed under
stringent conditions: 50% formamide, 5 x SSPE (l x SSPE =
0.18 M NaCl/lO mM NaH2P04, pH 7.4/2 mM EDTA~, 5 x
Denhardt's solution , 0.1% SDS, 100 ug/ml salmon sperm
DNA, and 1 ug/ml poly(A) at 42 for 16 hours. Filters
were washed twice at 22C for lO minutes in 2 x SSC (1 x
SSC = 0.15 M NaCl/0.015 M sodium citrate), 0.1% SDS and
once at 60C for greater than or equal to 2 hours in 0.1 x
SSC and 0.1% SDS. Washed filters were air-dried and sub-
jected to autoradiography at -70C with Kodak XAR-5 film
and Dupont Cronex Lighting Plus Intensifying screens.
The 3' end of the normal LDL receptor gene shown in
Figure 6 is contained on three of the~ clones isolated in
* Trade Marks
_30_ ~ 337~
the above manner: ~ 33-2, ~33-l, and~ hl. Note that the
restriction map of the normal LDL receptor gene shown in
Figure 6 also displays both "exons" and IVS regions
(intervening seguences or "introns"). The term "exons"
refers to domains of a gene that are "transcribed~' into
mRNA and eventually appear in the cytoplasm as mature
m~NA. Any one gene may have numerous exons, which
together, c~mprise the structural gene itself. Exons are
separated within the gene by regions referred to as
"intervening sequences" or IVS regions. The IVS regions
are transcribed into the initial RNA transcript of the
DNA. However, unlike exons, the IVS regions are processed
out of the initial RNA transcript and are therefore never
expressed in the ultimate protein product.
The DNA insert in ~33-2, approximately 12 kb, encodes
Exons A, B, and C; the insert in ~33-l, approximately ll
kb, encodes Exons D, E, and F; and the insert in~ hl,
approximately ll kb, encodes Exons E and F. The 32P-cDNA
probes shown in Figure 4A correspond to the following
regions of the LDL receptor gene: probes 2-5 hybridize to
exons in the 5' end of the LDL receptor gene denoted as
approximately ~0 kb of DNA in Figure 6; probe 6 hybridizes
to Exon C; probe 7 hybridizes to the 5' half of Exon F;
and probes 8 and 9 both hybridi~e to the 3' half of Exon
F. Exon C encodes the 0-linked sugar region of the recep-
tor (amino acid residues 693 to 749); Exon D encodes the
region between the 0-linked sugar domain and the mem-
brane-spanning domain plus 8 of the 22 amino acids com-
prising the membrane-spanning domain (residues 750 to
775); Exon E encodes the remainder of the membrane-span-
ning domain and 39 of the 50 amino acids comprising the
cytoplasmic domain (residues 776 to 828); Exon F encodes
the terminal ll amino acids of the cytoplasmic domain of
the receptor protein(residues 829 to 839) plus all of the
3' untranslated region of the mRNA.
~31- ~8~
E~AMPLE IV
CASE STUDY OF AN INDIVIDUAL WITH
FAMILIAL HYPERCHOLESTEROLEMIA
The present example discloses the use of the recom-
binant clones disclosed in Example's II and III ko charac-
terize a mutation in the structural gene for the LDL
receptor in a family with FH. The index case is a young
man (B.H.), hereafter designated FH 274, who has all of
the clinical features of homozygous FH. Previous
functional studies revealed that cultured fibroblasts rom
F~ 274 bound about one-third of the normal amount of
125I-labeled LDL. However, the receptors in FH 274 did
not cluster in coated pits and hence did not transport
their bound LDL into the cell. Thus, FH 274 was
categorized functionally as an "internalization-defective"
mutation. Studies of fibroblasts from the relatives of FH
274 revealed that he had inherited two different mutant
alleles; the allele encoding the internalization-defective
receptor was inherited from his mother and a null (or
silent) allele that produced no functional receptor
protein was inherited from his father.
Preliminary biosynthetic studies of cultured fibro-
blasts from FH 274 revealed that the LDL recèptor protein
encoded by the internalization-defective allele is about
10,000 daltons smaller than the normal receptor. Accord-
ingly, a study of this mutation ~as initiated by analyzing
the genomic DNA from FH 274 and his family members. As
described below, we found that the mutant gene has under-
gone a large deletion that eliminates two exons completely
and one exon partially.
The deletion mutation revealed by practice of the
present invention with respect to FH 274 was shown to
result from a recombination between two repetitive DNA
-32- ~2~ 78
elements: an Alu element in the intervening sequence
(IVS) that precedes the exon encoding the membrane-span-
ning region of the receptor and an Alu element in the exon
encoding the 3'-untranslated region of the gene. Alu
seguences are human DNA sequences which display a highly
repetltive character. It is thought that due to their
highly repetitive nature, Alu sequences may be responsible
for a number of genetic mutations: an Alu sequence from
one region of a gene may cross-hybridize with Alu
sequences from another region, forming a "loop" in the
DNA. Thus, the deletion occurs when this "loop" is
processed out of the gene, leaving an incomplete gene.
The resulting mutant gene, in the case of FH 274,
produces a truncated LDL receptor that lacks a membrane-
spanning region and a cytoplasmic domain. Most of these
truncated receptors are secreted from the cell, but some
of them remain associated with the outer surface of the
cell. In this position they can bind LDL, but the lack of
a cytoplasmic domain renders these receptors incapable of
clustering into coated pits and carrying LDL into the
cell.
Southern Blot Analysis of FH 274 DNA
Relative To Normal DNA
-
The preferred mode contemplated by the present inven-
tors for displaying a deletion mutation in an F~ indivi-
dual involves the use of a well-known technique known as
Southern blotting. Briefly, Southern blotting is a proce-
dure whereby ~enomic DNA from an individual is first iso-
lated and fragmented into discrete fragments and separated
electrophoretically on an agarose gel. The pattern of DNA
from the gel may then be "imprinted" onto a stable matrix.
The pattern of those fragments which correspond to the LDL
_33~ 3~8
receptor gene may then be visualized by hybridizing a
labeled LDL receptor cDNA or genomic probe with the
imprinted matrix and visualizing the gene pattern by means
of the label.
In performing the initial DNA fragmentation, the DMA
is preferrably restriction endonuclease digested into
smaller DNA fragments. However, the only requirement is
that the method chosen should be able to cleave the
genomic DNA reproducibly into the same fragment pattern.
In this manner, identically cleaved gene fragments will
exhibit a reproducible pattern when separated, for
instance, on the basis of fragment length. Thus, LDL
receptor gene fragments from test individuals might be
compared to the corresponding fragments from control
individuals to detect a shift in the respective LDL gene
pattern. A shift in the pattern of LDL receptor gene
fragments from a test individual relative to a control
pattern would be indicative of a mutation in the gene.
The present inventors have determined that the
restriction endonuclease Xba I was capable of displaying
the genetic mutation exhibited by FH 274. However, it is
contemplated that in future case studies, it may become
necessary to use other restriction enzymes. Thus, a bat-
tery of enzymes may be useful in certain instances to find
the correct enzyme for that particular defect. Those of
skill in the art will recognize that such a battery of
digestions may be necessary.
After fragmenting the DNA, the fragments produced are
then separated into a pattern whereby indi~idual fragments
are separated from one another. The preferable means for
separating the DNA fragments is to separate the fragments
according to size by subjecting the DNA to electrophoresis
in an agarose gel matrix. Agarose gel electrophoresis is
_3~ 37~
a procedure well-known in the art. However, other types
of separation techniques may be used, including, for
example, column chromatogarphy or density gradient
centrifugation. The gel electrophoresis technique is
useful in that it allows the separation o~ fragments in a
manner which allows for precise determination of the
apparent size of the separated fragments and allows for
easy handling of the frayments so separated. Furthermore,
the LDL receptor gene fragments which are present in the
gel matrix may be directly visualized by the Southern
blotting technique described more fully below.
Figure 4B shows Southern blots of genomic DNA from
normal cells and from FH 274, after digestion with Xba I,
and hybridization with several cDNA probes. More particu-
larly, the Southern blot hybridizations were performed as
follows. Genomic DNA (4 ug) was isolated from cultured
fibroblasts from the indicated subject (Maniatis, supra.),
digested with XbaI (New England Biolabs), electrophoresed
in 1% agarose containing buffer A (40 mM Tris-acetate, 3
mM Na2EDTA, 20 mM NaOAc, and 18 mM NaCl at pH 8.15), and
transferred to nitrocellulose paper by osmotic diffusion
(Maniatis, supra). The paper was incubated for 16 hours
at 42C with the appropriate 32P-labeled cDNA probe (2-4 x
cpm/ml) in 50% formamide, 1% SDS, 5x Denhardt's solu-
tion, 5xSSPE, and 100 ug/ml E. coli DNA after prehybridi-
zation for 1 hour at 42C in the same solution without the
32P-labeled probe. After hybridization, the paper was
washed in 1% SDS plus 2xSSC for 15 minutes at 23C and
then in 1% SDS plus 0.1xSS~ (for probe 1 hybridizations)
or in 1% SDS plus 0.5xSSC ~for probes 2-9~ for 4 hours at
68C. Filters were exposed ~o X-ray film with an intensi-
fying screen for 24 hours at -70C. Two representative
blot hybridizations using probes 1 and 8 are shown in
Figure 4B.
_35_ ~ 7~
The sizes and locations of 2P-labeled cDNA probes
used to map the LDL receptor gene are shown in Fi~ure ~A
by the closed bars and are numbered 1 to 9. Probe 1
~double-stranded DNA) was a mixture of a 2.1-kb Eco~I-SmaI
fragment and three different O 9-kb BamHI-XhoI fragments
from plO1, which together spanned most of the translated
region of the gene. The fragments were purified by
polyacrylamide gel electrophoresis and electroelution and
then labeled with 32p by random hexanucleotide prirning as
described by ~einberg and Vogelstein, Analyt. Biochem,
132:6-13 (1983). Probes 2 through 9 were prepared ~rom
M13 subclones of pLDLR-2 as a single-stranded, uniformly
32P-labelled DNA approximately 100 nucleotides in length
by the method of Church and Gilbert, Proc. Natl. Acad. Sci.
. .
USA, 81:1991-1995 (1984). AlI probes had a specific
radioactivity of at least 5 x 108 cpm/ug.
Briefly, to prepare probe 2 a DNA fragment
encompassing nucleotides 267 to 1081 (Figure 3) was cloned
into the bacteriophage M13mp9 vector as described by
Messing, Meth. Enzymol., 101: 20-78 (1983)- `
Single-stranded, uniformly 32P-labeled
DNA--probes--approximately 100 bases long were
prepared from the resulting clone by the method of Church
and Gilbert, supra, using an M13 universal primer and the
Klenow fragment of ~NA polymerase I to extend the primer
in the presence of three unlabeled deoxynucleotides and
one alpha-32P-labeled deoxynucleotide. The resulting
~adioactive primer extension product was denatured from
the template by boiling, size-fractionated on a denaturing
acrylamide gel, electroeluted, and then used directly as a
probe. Probes 3-9 were prepared in a similar manner,
except M13 clones containing different regions of the
sequence in Figure 3 were used as templates.
-36~ 3~
Referring to Figure 3, probe 3 encompasses
approximately nucleotides number 719 to ~544; Probe 4
encompasses approximatley nucleotides 267 to 1078; Probe 5
encompasses approximately nucleotides 1573 to 3486; Probe
6 encompasses approximately nucleotides 2154 to 2544;
Probe 7 encompasses approximately 2545 to 3948; and Probes
8 and 9 encompass approximately 4508 to 4962.
Figure 4B shows Southern blots of genomic DNA from
normal cells and from FH 274 after digestion with XbaI and
hybridization with several cDNA probes. Probe 1 was a
mixture of cDNA fragments that spanned most of the trans-
lated region of the LDL receptor mRNA ~Figure 4A). When
this probe was hybridized to the XbaI-digested genomic DNA
from the normal subject, three bands were observed, of
23-, 10-, and 7-kb, designated A, C, and D, respectively
(Figure 4B). All of these normal bands plus one addi-
tional band of 13-kb, designated B, were present in the
genomic DNA of FH 274. These findings suggest that one of
the mutant alleles in FH 274 has a normal restriction
pattern whereas the other allele gives rise to band B.
To localize the DNA segment that gives rise to band
B, the XbaI digests were probed with short DNA fragments
that corresponded to discrete regions of the LD~ receptor
cDNA (probes 2-9, Figure 4A). Probes 2 to 5 hybridized to
identical bands in normal and FH 274 DNA. Probes 6, 8 and
9 (but not probe 7) hybridized to the abnormal 13-kb band
B in the FH 274 DNA. Inasmuch as band B does not hybrid-
ize with probe 7 but does hybridize with probes on eitherside of probe 7 (that is, probes 6, 8, and 9), this frag-
ment appears to result from a deletion of DNA that in-
cludes the region encoding the mRNA encompassed by probe
7. This deletion would presumably involve the removal of
at least one XbaI site with fusion of the adjacent DNA
~37~ 63~37~
se~uences into a single XbaI fragment of 13 kb, namely
band B.
The specific mutation exhibited by FH 274 should in
no way be construed as the only type of mutation that will
be found in other FH individuals. Therefore, although
probes 2 through 5 failed to demonstrate an altered LDL
receptor gene fragment pattern in the case of FH 274,
these probes will be useful in detecting other mutations
in other FH individuals. Probes 2 through 5 hybridize to
the 5' half of the receptor gene. Therefore, mutations
which occur in this region of the receptor gene will be
detectable using probes 2 through 5. Similarly, since
probe 7 hybridizes to band D (Figure 4A), it will be
useful in detecting mutations which occur in this gene
region.
To determine whether band B originated from the
maternal (internalization-defective) or paternal (null)
allele, we performed XbaI digests of genomic DNA from the
two parents. Figure 5 shows that band B was present in
the DNA from the mother, but not the father, indicating
that the deletion was present on the internalization-
defective allele.
Detailed Characterization of the
Mutation Exhibited by FH 274
The following set of experiments demonstrate the
utility of the present invention in being able to detect
the presence of a mutation in the LDL receptor gene and in
providing the means whereby such mutation may be
specifically detailed and identified.
:
-38- ~B~3~
By using ~enomic recombinant clones and Southern
blottin~ techni~ues with LDL receptor cDNA probes, a
detailed restriction map of an FH individual can be
generated. These gene mapping techniques are now well-
known to those of skill in the art. The particular mapgenerated for FH 274 is displayed in Figure 6. Figure 6
is a comparison of restriction maps of the 3' end of the
normal LDL receptor gene and the deletion-bearing gene
from FH 274. The scale at the bottom indicates the length
of genomic DNA in kilobases. The organization of the
normal LDL receptor gene is shown in the diagram at the
top. Exons are indicated by solid segments and upper case
letters; intervening sequences (IVS) are indicated by open
segments and lower case letters. The Alu repetitive
sequences in IVS c and Exon F are indicated. Restriction
enzyme recognition sites used to define the gene deletion
in FH 274 are shown.
The restriction map of the normal receptor gene was
generated from studies of three genomic clones (~33-2,
~33~ hl) (Example III). The map of the gene in FH 274
was generated from ~FH 274-lO, which contains XbaI frag-
ment B (Figures 4B and 5) (see below~. Solid bars above
or below the restriction maps denote segments of the nor-
mal and mutant genes that were used for DNA sequencing.
To obtain ~FH 274-lO, we prepared 570 ug of genomic
DNA from fibroblasts of FH 274 and digested it with 1700
units XbaI. The digested DNA was extracted with
phenol/chloroform and then chloroform, preci~itated with
70% ethanol and 86 mM sodium acetate, and dissolved in 200
ul of buffer B (lO mM Tris-chloride and 1 mM Na2EDTA at pH
7.5). The DNA (80 ug) was redigested with 100 units & I,
loaded onto a 1% low-gelling temperature agarose gel
(Bethesda Research Laboratories) containing buffer A, and
electrophoresed at 40 V for 72 hr at 4~C. After electro
-39~ 7~
phoresis, ten 2-mm slices of the gel containing DNA frag-
ments ranging from 9 to 23 kb were extracted, concen-
trated, and dissolved in 20 ul of buffer B. One aliquot
(4 ul) of each DNA fraction, 5 ug of XbaI-digested genomic
DNA from FH 274, and size marker fragments were loaded
onto individual lanes of a 0.8% agarose gel containing
buffer A and electrophoresed at 35 V for 16 hours at ~3C,
The DNA was transferred to nitrocellulose paper and
hybridized with 32P-labeled probe 1 as described above.
The resulting autoradiogram identified the fraction that
contained the abnormal 13-kb XbaI fragment (fragment B,
Figure 4). The remaining DNA from this fraction (100 ng)
was mixed with 500 ng of XbaI-digested arms of ~Charon 35
and incubated with 490 units of T4 DNA ligase (New England
Biolabs) for 72 hours at 14C. The ligated material was
packaged into ~ phage particles in vitro (Amersham) to
yield a total of 6.7 x 103 plaque forming units. This
library was screened with 32P-labeled probe 8 (Figure 1),
which was expected to detect only the abnormal fragment
(13 kb), since the corresponding normal fragment ( 7 kb)
was too small to generate viable recombinant phage. One
recombinant clone was identified (~FH 274-10) and isolated
after an additional cycle of plaque purification. The
13-kb insert in ~FH 274-10 was isolated from purified
DNA, subcloned into pSP65 (Promega Biotec), and used for
restriction endonuclease mapping.
Figure 6 demonstrates that the PvuII site in IVSc and
the SstI site in Exon E were separated by approximately
5.9-kb in the cloned fragments of the normal gene but only
by approximately 0.6-kb in the cloned fragment of the FH
274 gene. Moreover, several restriction enzyme sites
between the PvuII site and the SstI site are shown to be
missing from the cloned FH 274 gene. These data confirmed
the diagnosis made from the genomic Southern blots
~' '
,
-40~
discussed above and further suggested that 5 kb of DNA was
deleted rom the FH 274 gene. The deletion included the
3' end of IVS c, all of Exons D and E and the IVS's
separating them, and the 5' end of Exon F.
To locate precisely the 5' and 3' breakpoints and the
structure at the deletion joint in FH 27~, we determined
the nucleotide sequences of the cloned portions of the
genes delimited by the bars in Figure 6. These sequences
revealed that the deletion joint occurred between two
repetitive elements of the Alu family that were oriented
in opposite directions. The 5' side of the deletion joint
was derived from an Alu sequence in IVS c. The 3' side of
the deletion joint was derived from an oppositely-oriented
Alu sequence in Exon F.
A test performed on a second FH individual using the
present invention failed to reveal a deletion in the LDL
receptor gene of that individual. It is felt that the
defect in the LDL receptor gene of this individual is due
to a point mutation. Defects which are not due to a
deletion mutation can be detected by modification and
extentions of the present invention which will be
described in future applications.
Although recombination between repetitive DNA
sequences has been postulated to be a cause of deletions,
to the knowledge of the present inventors such rearrange-
ments hava not previously been reported in eukaryotic
cells. In the most well characterized set of mammalian
deletion mutations, i.e., those that occur in the human
alpha- and beta-globin genes, one of the deletion
breakpoints frequently occurs within an Alu sequence but
the other breakpoint thus far has always occurred in a
nonrepetitive sequence of DNA.
-41- ~2~3~
* * *
The present invention has been disclosed in terms of
examples considered by the inventors to be the preferred
methods for practicing the invention. However, they are
in no way meant to be the only modes of practicing this
invention. For example, although the restriction enzyme
XbaI was useful in exhibiting the particular mutation
carried by FH ~74, it is contemplated that the use of
other restriction protocols may be necessary to display
the specific mutation in other FH individuals. Similarly,
although the present inventors feel that the Southern blot
technique represents the best mode for displaying the
pattern of LDL receptor gene fragments, other techniques
should also work and should be considered as included
within the scope of the present invention. For example,
it would be possible to separate the fragments by column
chromatography and assaying for the presence of gene frag-
ments as they elute from the column. Similarly, although
the probes used in the practice of the present invention
have been radiolabeled, it is not considered that
radiolabeling is the only technique whereby the probes may
be rendered detectable. For example, DNA hybridization
probes may be labeled with heavy isotopes or,
alternatively, by binding a ligand such as biotin. In
general, any specifically bindable ligand that is capable
of being independently detected can be used. These and
all other changes should be considered within the scope of
the appended claims.
Although the particular case disclosed was a mutation
resulting from a deletion, further extensions of the
preseni invention can also detect mutations resulting from
other types of events, including single nucleotide
changes. The latter can be detected directly on Southern
gel blots when they lead to the loss or acquisition of a
337~3
- ~2 -
restriction enzyme cleavage site. Single nucleotide changes can
also be detected by their reduced hybridization with short segments
of the cDNA for chemically synthesized oligonucleotides
corresponding to the sequence of the cDNA. Weakened hybridization
can be detected by the "melting" or dissociation of the mutant and
normal DNA sequences after they have been hybridized to each other.
Such abnormal melting behavior can be detected as an increased
sensitivity to heat, lower ionic strength, or chemicals such as urea
and formamide that dissociate chains of hybridized DNA. Thus, the
present invention will be useful for the diagnosis of all potential
mutations in the LDL receptor gene.
The following are the culture deposits which have been made
with the American Type Culture Collection (ATCC) in connection with
the disclosure of the within application:
ATCC NO. DEPOSIT DATE DESCRIPTION
39965 12/20/84 E. coli bearing plasmid pLDLR-l
39966 12/20/84 E. coli bearing plasmid pLDLR-2
40147 12/28/84 Bacteriophage lamda 33-1
40148 12/28/84 Bacteriophage lamda 33-2
40149 12/28/84 Bacteriophage lamda hl
:
