Note: Descriptions are shown in the official language in which they were submitted.
7~ ;~3
A DIAGNOSTIC ASSAY FOR THE PRESENCE OF
APOLIPOPROTEINS ASSOCIATED WIT~ PLASMA
HIGH DENSITY LIPOPROTEINS
Desc_iption
Technical Field of the Invention
The present invention relates to epitope-
specific reagents that bind apolipoproteins, and
particularly to monoclonal receptors ~hat form
immunoreactants with apolipoprotein A thereby
permitting a determination of the immunochemical
heterogencity of lipoproteins.
Back~ound of the Invention
A. Atherosclerosis and Lipo~roteins
Atherosclerosis is the disease in which
cholesterol and other lipids, accumulating on the
walls of arteries, form bulky plaques that inhibit
the flow of blood and may lead to the formation of a
clot, obstructing an artery and causing occlusive
thrombotic or embolic disease such as a heart attack
or strokeO Up to 50 percent of all deaths in the
United States are caused by atherosclerosis and its
secondary complications~
Human atherosclerosis is defined as the
accumulation of selected lipids, including
cholesterol, and cells in the walls of arteries and
with time produces occlusive lesions. Although the
etiology of atherosclerosis is multi-factorial, a
large body of clinical, pathologic, genetic and
experimental evidence suggests that abnormalities of
lipoprotein metabolism can contribute to the
development of atherosclerosis. These lipids are
carried in the blood stream as lipid-protein
complexes called lipoproteins~
~7~
Atherosclerosis, and particularly that form
known as coronary artery disease (CAD), is a major
health problem. Atherosclerosis and its related
vascular diseases acounted for 983,000 deaths in
1983; and CAD alone accounts for more deaths annually
than all forms of cancer combined. In the United
States, more than 1 million heart attacks occur each
year and more than five hundred thousand people die
as a result of this disease. In direct health care
costs, CAD costs the United States more than $60
billion a year. This enormous toll has focused
attention on ways to identify particular populations
at risk for CAD so that the disease can be controlled
~ith diet, behavioral modification (exercise), and
specific therapeu~ic agents.
Four major classes of cholesterol-associated
plasma lipoprotein particles have been defined, and
have their origin in the intestine or liver. These
particles are invGlved in the transport of the
neutral lipids including cholesterol and
triglycerides. All classes of plasma lipoproteins
have apolipoproteins associated with the
lipid-protein complex; and the apolipoproteins play
requisite roles in the function of these lipoproteins.
The first class is the chylomicrons. They
are the largest of the lipoproteins and are rich in
triglycerides. The site of origin of the
chylomicrons is the intestine.
While apolipoproteins are a quantitati~ely
minor proportion of the mass of chylomicrons,
apolipoproteins A-I, A~II and A-IV are significantly
associated with chylomicrons, and intestinal
synthesis of these A apolipoproteins has been found.
Much of the chylomicron complement of A
apolipoproteins is lost, and C and E apolipoproteins
are acquired when chylomicrons are exposed to plasma
or HDL ln vitro. Intestinal production of the A
apolipoproteins (apo-A) may be regulated by factors
other than fat absorption and chylomicron formation.
The next class of lipoproteins is the very
low density lipoproteins, VLDL. The VLDL particle is
involved in tr glyceride metabolism and transport of
these lipids from the liver. The apolipoproteins,
apo-B and apo-E are the major constituents of the
VLDL particle.
The third lipoprotein is called low density
lipoprotein (LDL), and is a specific product of the
catabolism of VLDL. The predominant apolipoprotein
in the LDL particle is apolipoprotein B, or apo-B.
Analytical techiques have revealed that apo-B is also
the specific apolipoprotein associated with
chylomicrons and VLDL.
The results of the now classic Framingham
study (1971) showed a clear correlation between risk
for CAD and serum cholesterol levels. This study
also demonstrated tha~ elevated levels of low density
lipoprotein (LDL) choles~erol are associated with
increased risk of CAD. Recently, a study conducted
by the Lipid Research Clinics Coronary Primary
Prevention Trial tl984) has demonstrated that plasma
levels of cholesterol and LDL cholesterol can be
reduced by a combined regime of diet and drugs, and
that this reduction of plasma cholesterol results in
reduction of the incidence of CAD mortality.
The cholesterol of atherosclerosis plaques
is derived in part~if not mostly from low-density
lipoprotein (LDL). LDL is a large spherical particle
whose oily core is composed of about 1500 molecules
of cholesterol, each attached by an ester linkage to
a long chain fatty acid. This core of cholesterol is
enclosed by a layer of phospholipid and unesterified
cholesterol molecules. The phospholipids are arrayed
so that the hydrophilic heads are on the outsid-e,
allowing the LDL to be in hydrated suspension in the
blood or extracellular fluids.
The cholesterol delivered to, and liberated
from LDL particles taken up by cells, controls cell's
cholesterol metabolism. An accumulation of
intracellular cholesterol modulates three processes.
First, it reduces the cell's ability to make
its own cholesterol by turning off the synthesis of
an enzyme, HMG CoA reductase, that catalyzes a step
in cholesterol's biosynthetic pathway. Suppression
of the enzyme leaves the cell dependent on external
cholesterol derived from the receptor-mediated uptake
of LDL.
Second, the incoming LDL-derived cholesterol
promotes the storage of cholesterol in the cell by
activating an enzyme denominated lipoprotein
acyltransferase. That enzyme esterifies fatty acids
to excess cholesterol molecules, making cholesteryl
esters that axe deposited in storage droplets.
Third, and most significant, the
accumulation of cholesterol within the cell drives a
feedback mechanism that makes the cell stop
synthesizing new LDL receptors. Cells thereby adjust
their complement of external receptors so that enough
cholesterol is brought into the cells to meet the
cells' varying demands but not enough to overload
them. For example, fibroblasts that are actively
dividing, so that new membrane material is needed,
maintain a maximum complement of LDL receptors of
about 40,000 per cell. In cells that are not
growing, the incoming cholesterol begins to
accumulate, the feedback system reduces receptor
manufacture and the complement of receptors is
reduced as much as tenfold.
On the other hand, it has been shown that
another circulating lipoprotein, high density
lipoprotein (HDL) is implicated in a state of
elevated cholesterol associated with lowered risk of
atherosclerosis. Apolipoprotein A is a ligand of the
HDL particle. The amount of HDL provides an inverse
correlation with the predicted incidence of
atherosclerosis.
High density lipoprotein (HDL) contains two
major apolipoproteins, apo-A-I and apo-A-II. Apo-A-I
is the major protein component of all primate HDL.
All HDL particles contain apo-A-I, and therefore
immuno ~uantification of HDL has usually involved the
quantitation of apo-A-I. HDL particles containiny
only apo-A-II have not been described.
One function of apo-A-I is the activation of
the plasma en~yme, lecithin-cholesterol
acyltransferase (LCAT). This enzyme is required for
the esterification of free cholesterol for transport
to the liver~ In the absence of apo-A-I, cholesterol
in the blood is not esterified and thus cholesterol
is not cleared from the blood. The specific role in
HDL metabolism served by apo-A-II has not been
defined.
Many studies have shown that elevated HDL
levels correlate with a reduced incidence of CAD.
Some authors have speculated that HDL removes
cholesterol from peripheral sites, such as the
arterial wall, therefore attributing anti-atherogenic
properties to HDL. Higher concentrations of HDL
cholesterol are correlated with a lower incidence of
and/or a decreased severity of cardiovascular
disease, while elevated levels of LDL cholesterol are
associated with an increased risk of CAD. For the
proper management of patients with hyperlipidemia
(excess lipids in the blood) and those patients at
special risk for CAD, it is desirable to frequently
determine levels of LDL and HDL cholesterol.
To date, assays of HDL cholesterol have been
cumbersome and inaccurate in determining blood levels
of HDL. It would therefore be beneficial to provide
an assay that is easy to use and accurately
determines HDL blood levels.
B. Lipoprotein Structure and Function
It is important to understand that
cholesterol does not exist free in plasma but is
transported to tissue sites in the body by
llpoproteins. Cholesterol can be obtained from
directed cellular synthesis or by diet. However,
cholesterol can be removed from the host only by the
liver, where it is converted to bile acids and
excreted.
Very low density lipoprotein (VLDL) carries
cholesterol and tri~lycerides to the liver for
subsequent excretion, whereas, LDL delivers
cholesterol to extrahepatic tissues, including the
coronary arteries. Hence, the "bad" lipoprotein,
LDL~apo-B, is involved in the deposition of
cholesterol in peripheral tissue. Conversely, the
"good" lipoprotein HDL/apo-A, removes cholesterol
from the tissues and returns cholesterol to the liver
for excretion.
Historically, many systems have been
developed to isolate and to characterize
lipoproteins. These techniques are usually based
upon the physicochemical properties of the
lipoprotein particles. The two most frequently used
~2~7~
--7--
techniques are ultracentrifugation and
electrophoresis.
Differential density gradient
ultracentrifugation takes advantage of the fact that
the lipoproteins are lighter or less dense, than
other plasma proteins, and it is easy to separate the
chylomicrons (the lightest lipoproteins), VLDL, LDL
and HDL from each other Electrophoretic techniques
have been useful for the classification of patients
with hyperlipidemias. However, these techniques are
not easily carried out in an ordinary clinical
laboratory.
One can also see that the simple
quantitation of blood cholesterol or triglycerides
does not provide the physician with the specific
information about which lipoproteins are carrying
these lipids and their quantitation.
C. The Plasma Lipoproteins
Four major classes of plasma lipoproteins;
i.e., chylomicrons, VLDL, ~RL and HDL, have been
defined, and subclasses within these undoubtedly
exist. All lipoproteins have their origin in the
intestine or liver, or both, and appear to have a
pseudomicellar structure. Neutral lipids, and
particularly, cholesterol esters and triglycerides,
are maintained in the lipoproteins in a soluble and
stable form through interactions with the
apolipoproteins and phospholipids, which are more
polar.
Unesterified cholesterol is also present in
these complexes. Its polarity lies between that of
the neutral lipids (cholesteryl esters and
triglycerides) and that of the more polar
apolipoproteins and phospholipids.
1247~Z3
An outer surface consisting of
apolipoproteins, unesterified cholesterol, and
phospho1ipids surrounds a water-insoluble core of
cholesteryl esters and triglycerides, protecting the
apolar lipids from the aqueous environment. This
general structural concept has been supported by
low-angle x-ray scattering studies and by other
physical methods in which a variety of probes have
been used to explore the structure of the
lipoproteins. An important function of the plasma
lipoproteins is thus the solubilization and transport
of the neutral plasma lipids.
D. The APolipoproteins
Apolipoproteins are the lipid-free protein
components of the plasma lipoproteins obtained by
treating intact lipoproteins with organic solvents,
detergents, or chaotropic agents. Not all proteins
captured with lipoproteins necessarily have a role in
lipid transport. A pertinent example is the recent
~; 20 recognition that the serum amyloid A proteins, acute
phase reactants, are transported in plasma bound to
HDL. These low molecular weight proteins may
comprise up to 30 percent of apo-HDL in inflammatory
states, but it is doubtful that they have specific
lipid transport roles.
1. APolipoproteins A-I and A-II
Two o the apolipoproteins of interest in
the present invention are apolipoprotein A-I
(apo-A-I) and apolipoprotein A-II (apo-A-II). These
are discussed below.
Apo-A-I is the major protein component of
all primate HDL. It consists of a single chain of
243 to 245 residues; does not contain cystine,
cysteine, leucine, or carbohydrate; and exists in
several isoforms. Apo-A-I has an alpha helical
.
Z3
content of about 55 percent in the lipid-free state,
which increases to about 75 percent upon binding
phospholipid. Repeating cycles of 11 helical
residues have been identified in this
apolipoprotein. It has been suggested that these
units represent a single ancestral chain which, by
gene duplication, has generated a 22-residue repeat
unit. These units have close sequence homology and
are believed to represent the lipid-binding regions
of the protein.
Apo-A-I is potent activator of LCAT, a
plasma enzyme that catalyzes the conversion of
cholesterol and phosphatidylcholine to cholesteryl
ester and lysophosphatidylcholine, respectively.
Specific lipid-binding regions of apo-A-I have been
found to activate LCAT, and this activity has been
associated with the property of lipid binding. As
already noted, liver and intestine synthesize
apo-A-I, but their relati~e contributions to the
total plasma content and the factors modulating
apo-A-I production are not well defined. Typically,
more than about 90 percent of plasma apo-A-I is
associated with HDL, less than about 1 percent with
VLDL and LDL, and ~bout 10 percent or less is
associated with the lipoprotein-free fraction of
plasma.
Apo-A-II is also a major constituent of
human HDL, accounting for about one-third of the
total protein and about 15 percent of HDL mass. It
exists as a dimer of two identical chains of 77
residues, which are linked covalently at the cysteine
of position 6 from the amino-terminus by a disulfide
bond, and its primary structure is known. Both the
monomeric and dimeric forms of apo-A-II are capable
of reassembling with phospholipifl. The alpha helix
~.Z~7~3
--10--
content of apo-A-II increases from about 40 to 65
percent on interaction with egg lecithinl and
specific lipid binding segments have been identified
and synthesized.
The specific role of apo-A-II in lipid
transport has not been identified, and it is a
quantitatively minor ~DL apolipoprotein in most lower
species. The bulk of plasma apo-A-II is found in
HDL, with less than about 5 percent in other density
classes.
2. Clinical Importance of
Apo-A Lipo~roteins
Measurement of the major protein constituent
of HDL, apo-At is clinically important. The results
lS of a number of studies have demonstrated that apo-A-I
levels are decreased in subjects with CAD. This
observation stresses the protective role of plasma
apo-A-I in this patient groupO
The results of several studies suggests that
by measuring the apo-A-I and apo A-II levels
accurately, it may be possible to predict an
individual's prognosis for atherosclerosis,
specifically for CAD.
Brief Summary of the Invention
The present invention contemplates
monoclonal receptors that immunologically binding
with apolipoprotein A, but are free from
immunoreaction with and binding to apolipoproteins B,
C, D and E. Particularly preferred monoclonal
receptors are monoclonal antibodies.
A method of preparing a monoclonal antibody
that immunologically binds with an apolipoprotein A
constitutes another aspect of the invention. In
accordance with that method, a host animal such as a
mammal is immunized with an human apolipoprotein A
~7~3
such as HDL or VLDL. Antibody~producing cells of the
immunized host are collected as by removing the
host's spleen and preparing a suspension of
splenocytes. The antibody-producing cells so
collected are fused with cells of a myeloma cell
line, preferably of the same animal species as the
immunized host, and typically in the presence of a
cell fusion promoter to Eorm hybridoma cells. The
hybridoma cells are diluted and cultured in a medium
that does not support growth of unfused myeloma cells
such as HT or HAT media. Such dilution and culturing
are typically carried out at an initial concentration
of about one hybridoma cell per cell growth wellO
The monocolonal antibodies produced by the culture
hybridomas are thereafter assayed for the ability to
immunologically bind with apolipoprotein A. A
hybridoma whose monoclonal antibodies
immunochemically bind with apolipoprotein A is
selected and cloned, and is thereafter recovered.
The particularly preferred monoclonal
antibodies are produced by hybridomas denominated
P3x63Ag8 ~ATCC T lB9), MPC~ ATCC CRL 167),
S/P 2-O-Agl4 (ATCC CRL 1581), and P3x63Ag8.653 (ATCC
CRL 1580).
The above-described method of preparing
monoclonal antibodies can include culturing the
hybridoma in vitro in a suitable medium and
__
recovering the antibody from the hybridoma
supernatant, i.e., a cell culture system. The above
method can include injecting the hybridoma into an
animal host and recovering the antibodies from
ascites fluid of the host.
The present invention also includes the
monoclonal antibodies produced by any oE the
-12-
above-described methods, and the above-denominated
hybridomas.
A method for assaying the presence of an
apolipoprotein A such as HDL constitutes another
aspect of the present invention. Here, a monoclonal
receptor such as a whole antibody of this invention
is provided, and a known amount is admixed with an
aliquot of a sample to be assayed for the presence of
an apolipoprotein A to form an admixture. The
admixture is maintained for a period of time
sufficient for the receptor to immunologically bind
with an apolipoprotein A present in the sample and
form an immunoreactantO The amount of receptor bound
in the sample is determined, thereby determining the
presence and quantity o~ the apolipoprotein A such as
HDL in the sample.
The methods o~ the present invention enable
the practioner to assay for total HDL present in the
sample, as well as for independently assaying for
apo-A I and apo A-II. The methods also enable the
assay of subsets of apo-A-l that are immunologically
bound by each of the specific monoclonal receptors of
the invention.
The invention further contemplates a
diagnostic system such as a kit that includes at
least one package containing as an active ingredient
an effective amount of the monoclonal receptor
(epitope-specific reagent) of this invention which,
when introduced into a sample to be assayed (for
example, serum), immunologically binding with an
apolipoprotein-A such as apo-A-I or apo-A-I~, but
does no~ react with other classes of apolipoproteins
including apolipoproteins B, C, D and E or
non-apolipoproteins, i.e.~ it is specific.
~7~Z3
-13-
Brief Description of the Drawings
Figure l is a photograph of a polyacrylamide
gel electrophoresis (PAGE) separation showing -
apolipoprotein chain specificity of a mouse
monoclonal antibody of this invention for VLDL, LDL,
HDL, apo-A-I, and apo-A-II at concentrations of 30,
20, 20, 10, and 10 micrograms, respectively. The
samples were electrophoresed in 7.5-20 percent
polyacrylamide gradient slab gels containing 0.1
percent sodium dodecyl sulfate (SDS). The top left
is a photograph of a Coomassie Brilliant Blue R-250
protein-stained gel before electrophoretic transfer
of the apolipoproteins to nitrocellulose. The
abbreviations are: B, apo-B; HSA, human serum
albumin; E, apo-E; C, apo-C; A-II apo-A-I; A-II(D),
apo-A-II dimer~; A-II(M), apo-A-II monomers. The
remaining photographs are 24-hour autoradiographs of
identical nitrocellulose paper transfers after
incubation with the individual hybridoma ascites
20 fluids and 125I-goat anti-mouse Ig [0.5
micro-Curies/milliliter (micro Ci/ml)]O The antibody
numbers above each autoradiograph refer to the
apolipoprotein specificity as determined by this
procedure. The monoclonal receptors denominated
A-II-l, A-I-~, A-I-7, and A-I-~ were used at
dilutions of 1:5000, 1:2000, 1:3000, and 1:2000,
respectively.
Figure 2 is a graph of data illustrating
maximum binding capacity of mouse ascites fluids
containing human apo-A-I- and apo-A-II-specific
monoclonal receptors (antibodies). The upper portion
of the figure shows binding of 125I-HDL. The lower
portion of the figure is a graph showing data of
binding of 125I_apo-A-I or l25I-apo-A-II- The
fluid-phase RIAs were incubated for 18 hours at 4
-14-
degrees C and contained I-HDL, 1 5I_apO-A~I~
or 125I-apo-A-II at final concentrations of 66.7,
33.3, and 33.3 nanograms per milliliter (ng/ml)-,
respectively. The coefficient of variation for all
data points was less than 10 percent.
Figure 3 is a photograph of a 24 hour
autoradiograph aftef gel electrophoresis for
determining apolipoprotein composition of the
antibody bound and unbound portions of 125I-HDL.
The 1 5I-HDL was contacted and maintained in
contact (incubated) with monoclonal antibodies A-I-7,
A-II-l, and A-I-4 in antibody excess for 18 hours at
4 degrees C followed by precipitation of the
monoclonal antibody with optimal proportions of goat
anti-mouse Ig antiserum. The precipitates containing
the bound fractions (B) and the supernatants
containing the unbound fraction~ (UB) were recovered,
dissolved in 1 percent sodium dodecyl sulfate (SDS),
and electrophoresed along with the starting
20 1 I-~DL (S) on a 7O5-20 percent po~yacrylamide
gradient pore gel in the presence of 0.1 percent
SDS~ The notations are as follows: HSA represents
human serum albumin; (D~ represents apo-A-II dimer
and (M) represents apo-A II monomer.
Figure 4 is a photograph of a 24-hour
radioimmunoassay following polyacrylamide gel
electrophoresis of the five HDL subfractions isolated
by density gradient ultracentrigugation in KBr. A
fraction of plasma brought to a density of 1.063
grams per milliliter (g/ml~ with KBr was centrifuged
for 48 hours at 10 degrees C in a Beckman 60 Ti
ultracentrifuge rotor at 54,000 rpm. The gradient
was fractionated from the top, and the 4.0-ml
fractions were dialyzed into 0.15 molar (M) NaCl
containing 0.1 percent ethylenediaminetetracetic acid
~7~'~3
--15--
(EDTA). The upper photograph shows the
protein-staining pattern of the fractions after
electrophoresis on a 7.5-20 percent acrylamide
gradient gel in the presence of 0.1 percent SDS to
visualize the apolipoproteins. The bottom photograph
shows the protein-staining pattern of the same
density fractions after electrophoresis on a 4-30
percent acrylamide gradient gel in the absence of a
denaturant for separation of intact lipoprotein
particles on the basis of size. The total
cholesterol of HDL fractions 1 through 5 was ~98,
321, 231, 194, and 222 micrograms per milligram
tug/mg) protein, respectively.
Figure 5 is a bar graph representing the
relative expression of apo-A-I and apo-A-II epitopes
in HDL subpopulations. The upper portion shows HDL
fractions 1 through 5 obtained by density gradient
ultracentrifugation. The lower portion shows HDL
fractions from the PBE 94 chromatofocusing column.
Data shown were obtained from logit-transformation
analysis of the competitive RIAs. The relative
epitope expression in each density or
chromatofocusing HDL subfraction was obtained by
assigning for each antibody a value of 1.0 to the
fraction that required the least amount of protein
(ug/ml) for 50 percent inhibition of antibody
binding. For each antibody, other density or
chromatofocusing fractions were then expressed as a
fraction of that value.
Figure 6 is a photogxaph of an
autoradiograph of representative HDL chromatofocusing
fractions following polyacrylamide gel
electrophoresis. A fraction of plasma having a
density of l.n63-1.21 g/ml (20 mg protein in 2 ml)
was dialyzed into piperazine HCl, having a pH value
~7~
of 5.8, chromatographed on a PBE 94 column, and
eluted with Polybuffer 74. The Polybuffer was
removed from selected 4-ml column fractions (11, 18,
27, 32, 34, and 37) by chromatography on Sephadex
G-75 with 0.15 M NaCl, 1 millimolar (mM) EDTA, and
0.02 percent NaN3 (having a pH value of 7.4) as
eluant. The top photograph is the protein staining
pattern after electrophoresis on a 7.5-20 percent
acrylamide gradient gel in the presence of 0.1
percent SDS to identify the particle size
distribution of each fraction. The bottom photograph
is of the same HDL fractions after electrophoresis on
a 4-30 percent polyacrylamide gradient pore gel to
identify the particle size distribution. Total
cholesterol of HDL fractions 11 through 37 above was
294, 233, 183, 174, 190, 185; and 93 micrograms/mg
protein, respectively.
Figure 7 is a photograph of an
autoradiograph of a Western blot analysis performed
to delineate apo-A-I isoforms. Increasing amounts of
each radiolabeled apolipoprotein or lipoprotein were
paired with decreasing amounts of homologous
non-radioiodinated antigen so that a constant amount
of total antigen was added to each RIA to insure that
radioiodination of the ligands did not interfere with
antibody binding. Varying proportions of labeled and
nonlabeled HDL (Figure 7A) or soluble apolipoprotein
(Figure 7B) were incubated with each monoclonal
antibody for 18 hours at 4 degrees C. Constant
concentrations of 133.3 nanograms per milliliter
(ng/ml) HDL (125I-HDL plus homologous HDL), and
33.3 ng/ml apo-A-I or apo-A-II (125I-apO-A-I plus
homologous apo-A-I, or 125I-apo-A-II plus
homologous apo-A-II) were maintained~
In panel B, monoclonal antibody A~ l was
incubated with apo-A-II, and monoclonal antibodies
A-I-4, A-I-7 and A-I-9 were incubated with apo-A-I.
Results were plotted as the mean counts per minute
(cpm) recovered in the precipitate after reaction
with an optimal proportion of goat anti-mouse Ig
antiserum versus the percent of 1 5I-antigen
added. For the linear regression, correlation
coefficients were equal to or greater than 0.995 for
all antigen and antibody combinations shown. The
linearity and concordance indicated by the high
correlation coefficients (r greater than or equal to
0.995) identified that each antibody reacted with
each labeled and nonlabeled antigen pair with the
same apparent aEfinity.
In separate studies, 125I-HDL labeled by
two different methods, namely with either the
Bolton-Hunter reagent or by the lactoperoxidase
procedure were found to be equivalent in their
reactivity for each antibody. In addition, a 1:16
dilution of a polyvalent antisera obtained from a
rabbit hyperimmunized with human HDL precipitated 100
percent of 100 ug/ml of 125I-HDL. Therefore,
radioiodination of the antigens did not interfere
with the immunoreactivity or account for the
inability of these antibodies to bind 100 percent of
the labeled antigen.
Figure 8 is a graph showing radioiodination
of 1 I-HDL and 1 I-apolipoproteins. The
antigens were subjected to mild dissociating
conditions that included heat and exposure to
detergents to insure that the epitopes recognized by
these antibodies were exposed and available for
reaction with antibody. Limiting amounts of
monoclonal antibodies (Mab) A-I-4, A-I-7, A-I-9 and
~'7~`~3
A-II-l were added to I-HDL (final concentration
133 ug/ml) that had been incubated at either 4
degrees C or 52 degrees C. The isolated
apolipoproteins, 125I-apo-A-I and 125I-apo-A-II
were similarly heated before exposure to antibody.
Again, no significant increases in antibody binding
were observed. In fact, the binding of Mab A-I-7 to
125I-apo-A-I and antibody Mab A-II-l to
1 5I-apo-A-II was reduced by heating.
To determine if higher temperatures during
rather than before antibody exposure would increase
binding, reaction mixtures containing 125I-HDL and
antibody were incubated at 4 degrees C, 24 degrees C,
37 degrees C, or 52 degrees C for up to 18 hours. In
no instance did incubation at 37 degrees C or 52
degrees C increase binding above that observed at 4
degrees C or 24 degrees C and, as noted above, the
binding of Mab A-I-7 and Mab A-II-l to I-HDL was
reduced at 52 degrees C.
Figure 9 is a photograph of an
autoradiograph showing the polyacrylamide gel
electrophoresis of the individual 125I-HDL
ligands. Apolipoproteins and the molecular weights
indicated in Figure 9 were obtained from other lanes
of the same gels containing molecular weight markers
and unlabeled HDL from a pooled plasma source after
staining the gels for protein with Coomassie
Brilliant Blue ~ 250. The sex of the donor, and the
specific activity and acid precipitability of the
125I-HDL ligands, respectively were: LK, female,
5.6 disintegrations per minute per picogram (dpm/pg)
and 99.2 percent; AD, female, 6.1 dpm/pg and 99.2
percent; AL, ~emale, 7.1 dpm/pg and 99.2 percent; JR,
female, 6.2 dpm/pg and 99.3 percent, EW, male, 6.0
dpm/pg and 99.1 percent; PM, male, 5.8 dpm/ug and
~llf~
--19--
99.0 percent; GM, male~ 4.5 dpm/pg and 98.7 percent;
and CW, male 6.6 dpm/pg and 99.3 percent.
Figure 10 is a graph showing the binding
capacities of the apo-A antibodies for 1 5I-HDL
ligands obtained from eight unrelated individuals.
The percent of l25~-HDL that was maximally bound
from a pooled HDL source was consistently greatest
with Mab A-I-7 (greater than 50 percent), less with
antibody A-I-4 (40-50 percent) and lowest with Mab
A-I-9 (30-40 percent). This pattern of reactivity
was duplicated with the eight 125I-HDL ligands
(final concentration, 15 ng/ml) from the individual
donors (Figure 10A, B and C). Each line represents a
different 125I-HDL ligand. Data from female donors
is indicated with solid lines, males in hatched
lines. No consistent sex differences were noted.
Figure 10A, RIA with antibody A-I-4; Figure 10B, RIA
with antibody A-I-7; Figure 10C, RIA with antibody
A-I 9; 10D, RIA with antibodies A-I-4, A-I-7 and
A-I-9 at a ratio of 1:16:8; Figure 10E, RIA with
antibody AII-l; and Figure 10F, RIA with antibodies
A-I-4, A-I-9 and A-II-l at a ratio of 1:16:8.
Detailed Description of the Invention
I GENERAL DISCUSSION
The term "receptor" as used herein is meant
to indicate a biologically active molecule that
immunologically binds to (or with) an antigen. Such
binding typicaly occurs with an affinity of about
105 liters per mole and is specific interaction of
the epitope of the antigen with the Fab portion of
the receptor.
A receptor molecule of the present invention
is an intact antibody protein, substantially intact
antibody or an idiotype-containing polypeptide
portion of an antibody (antibody combining site) in
z~
-20-
subtantially pure form, such as in ascites fluid or
serum of an immunized animal. The terms "receptor"
and "monoclonal receptor" are used interchangeably
herein in a generalized sense for a molecular entity
that contains the antibody combining site of a
monoclonal antibody of 'his invention. The terms
"antibody", "monoclonal antibody" and "Mab" are
utilized interchangeably herein for a whole antibody
of this invention.
The term "ligand" as used herein is meant to
indicate a molecule that contains a structural part
that is immunologically bound by a specific receptor
to form an immunoreactant. A ligand used in the
present invention is an apolipoprotein A-containing
entity such as a radioiodinated HDL antigen adhered
to a solid matrix as described in the
radioimmunoassay described hereinafter.
Biological activity of a receptor molecule
is evidenced by the immunologic reaction o the
receptor to its antigenic ligand upon their admixture
in an aqueous medium to form an immunoreactant, at
least at physiological pH values and ionic
strengths. Preferably, the receptors also bind to
the antigenic ligand within a pH value range of about
5 to about 9, and at ionic strengths such as that of
distilled water to that of about one molar sodium
chloride.
Idiotype-containing polypeptide portions
(antibody combining sites3 of antibodies are those
3~ portions of antibody molecules that contain the
idiotype and bind to the ligand, and include the Fab,
Fab and F(ab )2 portions of the antibodies. Fab
and F~ab')2 portions of antibodies are well known
in the art, and are prepared by the proteolytic
reaction of papain and pepsin, respectively, on
7~J ~3
substantially intact antibodies by methods that are
well known. See for example, ~.S. Patent No.
4,342,566 to Theofilopolous and Dixon. Fab' antibody
portions are also well known and are produced from
F(ab')2 portions followed by reduction of the
disulfide bonds linking the two heavy chain portions
as with rnercaptoethanol, and then alkylation of the
resulting protein mercaptan with reagent such as
iodoacetamide. Intact antibodies are preferred, and
will be utiliz~d as illustrative of the receptor
molecules of this invention.
A "~onoclonal receptor" (Mab) is produced by
clones of a single cell called a hybridoma that
produces (secretes) but one kind of receptor
molecule. "Polyclonal" antibodies (Pab) are
antibodies produced by clones derived from different
cells that secrete different antibodies that bind to
a plurality of epitopes of the immunogenic molecule.
The preparation of Pab is discussed hereinafter as
part of the production of Mabs.
The hybridoma cell is fused from an
antibody producing cell and a myeloma or other
sel~-perpetuating cell line. Such receptors were
first described by Kohler and Milstein, Nature, 256,
495 (1975). Monoclonal receptors are typically
obtained from the supernatants of hybridoma cell
cultures, or, alternatively, from ascites fluid or
other body fluids obtained from non-human, warm
blooded host animals into which the hybridoma cells
were introduced.
Antibodies are secreted by specialized cells
called plasma cells and to a quantitatively lesser
degree by their precursor B cells (bone
marrow-derived lymphocytes). Each B cell or plasma
~2~7~ 3
-22-
cell secretes one type of antibody having a single
specificity, so various antibodies of different
specificites are each secreted by different B cells
and their derivative plasma cells. These B cells may
be cloned to provide a source of single antibodies.
However, these cells die in a few days in culture
media and must be made relatively "immortal" so that
a supply of the desired antibodies may be obtained.
This is accomplished by removing the B cells and
plasma cells from the animal, typically from the
spleen, fusing them with a cancerous or myeloma cell
to form a somatic cell hybrid thybridoma), and then
cloning and propagating the hybridoma.
The antibody-producing cells that are
employed may be obtained from a non-human host animal
immuni~ed by injection of an immunogen, in this
instance a human apolipoprotein A, typically followed
by one or more booster injections with the same
immunogen. The spleen is isolated after a sufficient
time period has elapsed for the host to produce
antibodies, this is typically about one month to
about three months after the first immunization.
Non-human, warm blooded animals usable in
the present invention as hosts may include poultry
(such as a chicken or a pigeon), a member of the
ratitae bird group (such as an emu, ostrich,
cassowary or moa) or a mammal (such as a dog, cat,
monkey, goat, pig, cow, horse, rabbit, guinea pig,
rat, hamster or mouse). Preferably, the host animal
is a mouse or rabbit.
It is preferred that a myeloma cell line be
from the same species as the antibody-producing
cells. Therefore, fused hybrids such as mouse-mouse
hybrids [Shulman et al., Na _re, 276, 269 (19783 or
rat-rat hybrids (Galfre et al., Nature, 277, 131
~47~;~3
-23-
(1979)] are typically utilized. However, some
rat-mouse hybrids have also been successfully used in
forming hybridomas [Goding, "Production of Monoclonal
Antibodies by Cell Fusion", in Antibody as a Tool,
Marchalonis et al. eds., John Wiley & Sons Ltd.,
273-289 (1982)~ hereinafter Marchalonis et al.].
Suitable myeloma lines for use in the present
invention include MPC-ll (ATCC CRL 167),
P3X63-Ag8.653 (ATCC CRL 1580), Sp 2/0-Agl4 (ATCC C~L
1581), P3X63Ag8U.1 (ATCC CRL 1597), Y3-Agl.2.3
(deposited at Collection Nationale de Cultures de
Microorganisms, Paris, France, number I-078 and
P3X63Ag8 (ATCC TIB9). Myeloma line P3X63-Ag8.653 is
preferred for use in the present invention.
Monoclonal anti-apolipoprotein A receptors
were formed as described herein from murine (mouse)
splenocytes fused with murine myeloma cells. The
polyclonal anti-apolipoprotein A antibodies
described were formed from rabbits. The hybridomas
that produce the monoclonal anti-apo-A I and
anti-apo-A-II receptors of this invention were given
the following designations for reference purposes and
were deposited with the American Type Culture
Collection (ATCC), Rockville, Maryland on March 6,
1985 under the following ATCC accession numbers.
ATCC Accession
Hybridoma Mab Number
HA61 H112F3.lAll A-II-1 HB 8743
611 AV63C2.lFl A~I-4 HB 8744
HA60 HA22G7.5F8 A-I-7 HB 8745
HA62 HA227A2.7D3 A-I-9 HB 8741
-24-
Receptors are t~pically utilized along with
an indicator labeling means or "indicating group" or
a "labeln. The indicating group or label is utilized
in conjunction with the receptor as a means for
determining the extent of a reaction between the
receptor and the antigen.
The terms "indicator labeling means",
"indicating group" or "label" are used herein to
include single atoms and molecules tha~ are linked to
the receptor or used separately, and whether those
atoms or molecules are used alone or in conjunction
with additional reagents. Such indicating groups or
labels are themselves well-known in immunochemistry
and constitute a part of this invention only insoEar
as they are utilized with otherwise novel receptors,
methods and/or systems.
The indicator labelling means can be a
fluorescent labelling agent that chemically binds to
antibodies or antigens without denaturing them to
form a fluorochrome (dye) that is a useful
immunofluorescent tracer. Suitable fluorescent
labelling agents are fluorochromes such as
fluorescein isocyanate (FIC), flourescein
isothiocyanate (FITC), dimethylamino-naphthalene-
S-sulphonyl chloride (DANSC), tetramethylrhodamine
isothiocyanate (TRITC), lissamine, rhodamine B200
sulphonyl chloride (RB 200 SC) and the like. A
description of immunofluorescence analysis techniques
is found in Marchalonis et al., "Immunofluorescence
30 Analysis'l, 189-231, supra.
The indicating group may also be an enzy~e,
such as horseradish peroxidase (HRP) or glucose
oxidase, or the like. Where the principal indicating
group is an enzyme such as HRP or glucose oxidase,
~47~23
-25-
additional reagents are required to visualize the
fact that a receptor-ligand complex has formed~ Such
additional reagents for EIRP include hydrogen peroxide
and an oxidation dye precursor such as
diaminobenzidine. An additional reagent useful with
glucose oxidase is 2,2'-azino-di-(3-ethyl-
benzthiazoline-6-sulfonic acid) (ABTS).
An exemplary radiolabelling agent is a
radioactive element that produces gamma ray
emissions. Elements which themselves emit gamma
rays, such as 124I~ 125I 128I 131I 132I
and 51Cr represent one class of gamma ray
emission-producing radioactive element indicating
groups. Particularly preferred is 125I. ~nother
class of useful indicating groups are those elements
h llC 13F 150 and 13N which themselves
emit positrons. The positrons so emitted produce
gamma rays upon encounters with electrons present in
the animal's body. Also useful is a beta ray
emitter, such as lllindium.
A preferred radioactively labeled monoclonal
receptor may be prepared by culturing hybridoma cells
in a medium containing radioactive amino acids, as is
well known, as well as by isolating the monoclonal
receptor and then labelling the monoclonal receptor
with one of the above radioactive elements as
described in U.S. Patent No. 4,3~1,292 to Bieber and
Howard.
Specific indicating means linked to reagents
that react with the receptors of this invention are
discussed hereinafter.
Four previously identified monoclonal
antibodies that bind to apolipoproteins A of human
plasma HDL were obtained from their respective
hybridomas and characterized. Each of these
,: ',''''~
`J
antibodies was specific for the apolipoproteins of
human HDL, based on binding ~o delipidated and
isolated apolipoproteins of ~IDL after transfer to
nitrocellulose and binding of the soluble
apolipoproteins in fluid phase.
The vast majority of the monoclonal
antibodies obtained by immunization of mice with
native human HDL were specific for human apo-A-I,
suggesting greater immunogenicity of human apo-A-I
for BALB/c mice. This difference in immunogenicity
between apolipoproteins A-I and A-II was observed
also when the isolated apolipoproteins were used as
immunogens. Thus, of the four antibodies
characterized in this study, three were specific for
three separate epitopes on apo-A-I. Only a single
apo~ specific antibody was obtained and
characterized.
Human plasma H~L of density 1.063-1.21 g/ml
represents a heterogenous mixture of HDL particles
that differ with respect to both lipid and protein
composition. Using the hybridoma-produced antibodies
of this invention that define apo-A-I-specific and
apo-A-II-specific epitopes, immunochemical
heterogeneity of HDL was clearly evident.
Solid-phase immunoassays permitted analysis
of antibody specificity, but with fluid-phase assays
it was possible to analyze the heterogeneity of
molecules with respect to expression of individual
epitopes. It was found that not all HDL particles
expressed the defined apo-A-I and apo-A-II epitopes
that could be bound by a given apo-A-I-specific
antibody. The unbound HDL contained apo-A-I and HDL,
indicating heterogeneity of epitope display by
apo-A-I on different particles (Figure 2).
-27-
The existence of at least two types of HDL;
i.e., particles containing apo-A-I and apo-A-II, and
particles containing apo-A-I but no apo-A-II r was
verified with monoclonal antibody A-II-l. Whereas
this apo-A-II-specific antibody bound only a subset
of total HDL, it did bind all apo-A-II (Figure 2).
The unbound HDL was devoid of detectable apo-A-II,
appearing to contain only apo-A-I (Figure 3). Thus,
all HDL particles possessing an apo-A-II chain
expressed this A-II epitope.
Heterogeneity of epitope expression by
isolated apo-A-I was readily evident. None of the
apo-A-I-specific antibodies was able to bind all
apo-A-I molecules, either as HDL or soluble apo-A-I.
The inability of each of the anti-apo-A-I antibodies
to identify its complementary epitope on all A-I
apolipoprotein chains was examined. First, technical
issues were excluded such as affinity, quantity of
available antibody, or radioiodination of the
ligands. Second, it was demonstrated that the
antibodies did not selectively bind different apo-A-I
isoforms. Third, the use of dissociating conditions
(e.g~, heat and nonionic detergents) designed to
mobilize and expose cryptic epitopes of the
apolipoprotein on either HDL or the isolated soluble
state did not result in a significant increase in the
capacity of antibody to bind all molecules.
Immunochemical heterogeneity of epitope
expression by apo-A-I organized on HDL was further
supported by the demonstration that the combination
of three apo-A-I-specific antibodies could bind a
greater relative proportion of HDL than any single
antibody (Table I, hereinafter). Thus, the three
apo-A-I epitopes recognized by the anti-apo-A-I
3~
t7~z3
-2~-
differed, and some particles existed that expressed
only one or another of these three epitopes.
Apo-A-I occurs predominantly, if not
virtually entirely, associated with lipid. The
heterogeneity of epitope expression was determined by
lipoprotein-associated apo-A-I~
The possibility that these antibodies
distinguished individual epitopes of apo-A-I that
were present or not on the basis of allotypic
(genetically determined individual differences) or
sex differences in apo-A-I was examined. For those
studies, HDL was isolated from single individual
normo-lipidemic subjects. Compared with pooled
plasma, HDL isolated from the plasmas of four
unrelated normal donors of each sex had similar
patterns of heterogeneous epitope expression.
Be~ause these antibodies did not appear to
identify allotypic or sex differences between
individuals in their apo-A-I molecules, the existence
of multiple apo-A-I genes was considered since
differential gene regulation, or differential sites
or rates of metabolism might account for the observed
heterogeneity of apo-A-I. It appears that there is a
single apo-A-I gene (Karathanasis et al., Proc. Natl.
25 Acad. Sci, VSA, 80, 6147-61Sl (1983). Alternative
splicing oE the gene has not been described but has
not been examined.
Also, each of the epitopes is expressed by
the different apo-A-I charge isoEorms. Thus, it was
determined whether the apo-A-I antibodies selectively
distinguished apo-A-I on HDL that was derived from
different known major synthetic sources such as the
liver and the intestine. Thoracic duct lymph was
used as an enriched source of intestinal apo-A-I.
~7~Z3
-29-
The medium from human Hep G2 hepatoma cultures
provided a source of pure hepatic apo-A-I.
Both the hepatic and intestinal apo-A-I
contained molecules expressing epitopes bound by
antibodies A-I-7 and A-I-9; i.e., A-I-7 and A-I-9
epitopes. In addition, plasma VLDL fractions that
provide a source of both hepatic ~nd intestinal
(chylomicron remnant) apo-A-I expressed only the
epitope bound by antibody A-I-9. Therefore, the
results are not consistent with epitope differences
based on different synthetic sources.
The hypothesis that the three apo-A-I
epitopes distinguish between molecules differently
organized on different HDL particles was in part
substantiated by separation of HDL on the ~asis of
the physical properties of density and charge.
However, because density and chromatofocusing
~ractions dif~ered quantitatively but not absolutely
in the expression of individual apo-A-I epitopes,
these methods did not entirely resolve the
responsible subsets of HDL. Rather, they facilitate
only enrichment or relative depletion of particles
expressing individual apo-A-I epitopes.
Physical fractionation of native HDL is
unlikely to result in complete segregation of
specific apo-A-I epitopes expressed by apo-A-I on
HDL, since HDL particles appear not to exist that
exclusively contain only apo-A-I organized in a
single conformational formatO However,
immunochemical separation may provide new
information. Recent studies of immunopurified HDL
have shown that ultracentrifugation can alter HDL
structure and suggest that additional studies of the
immunochemical properties of HDL should be directed
at the HDL particle as it exists in plasma ~McVicar
-30-
et al., Proc. Natl. Acad. Sci. USA, _ , 356-1360
(1984)].
There is no reason to assume that
conformational variation will be identical for
lipid-free and lipid-associated apo-A-I. For
example, protein-protein interactions resulting in
the formation of soluble oligomers of lipid-free
apo-A-I have been observed in preliminary studies to
influence the degree of expression of epitope A-I-4;
whereas protein-lipid or lipoprotein interactions
appear to have a similar influence. Studies of the
HDL density and chromatofocusing subfractions
demonstrate that apo-A-I is not organized the same on
different HDL particles.
The lighter, larger cholesterol-rich HDL
(~IDL2-like) that are enriched in apo-A-I relative
to other apolipoproteins are rich in apo-A-I that
express predominantly the A-I-9 epitope. In
contrast, the more dense~ smaller, cholesterol-poor
HDL which contain apo-A-II and other minor
apolipoproteins are rich in apo-A-I that express
predominantly the A-I-7 epitope. Because these two
types of HDL particles may represent different
metabolic states of HDL, ~he different apo-A-I
conformations on HDL may serve to direct HDL
particles to their proper enzymatic or cellular sites.
Some methods of quantitative analysis of
plasma HDL have employed immunologic assays for
apolipoproteins A-I or A-II. The immunochemical
properties of these apolipoproteins as evident from
analysis with polyclonal antibodies have indicated
the existence of unusual and distinctive properties.
The reactivity of apo-A-II-specific antisera is for
the most part comparable for apo-A-II whether in free
-31-
solution or associated with HDL [Mas et al.,
Biochemistry, 14, 4127-4131 (1975)].
- ~owever, the HDL density class is composed
of at least two types of HDL particlesi i.e., those
possessing both apo-A-I and apo-A-II, and those
containing apo-A-I, but no apo-A-II. Because all HDL
particles appear to contain apo-A-I, immunologic
analyses of apo-A-I have been herein used in
quantitating total plasma HDL. A caveat is the
difference in the ability of various antisera to
detect all apo-A-I in HDL or plasma. The reasons
offered for this discrepancy have centered around the
hypothesis that some apo-A-I epitopes on native HDL
are sterically occult.
As noted before, hybridoma cell lines that
secrete human HDL-binding monoclonal antibodies were
prepared to examine this molecular aberration, to
determine if the apparent immunochemical
heterogeneity of HDL and its apolipoproteins is
valid, and to obtain precise immunochemical reagents
tha~ permit quantitation of all HDL particles in
plasma as well as defined subsets of HDL.
Three mouse monoclonal antibodies (Mab7s)
specific for human apolipoprotein (apo) A-I and one
specific for human apo-A-II that were prepared have
been highly characterized and their binding of high
density lipoprotein (HDL) particles in solution was
determined. The apo-A-II-specific antibody bound 85
percent of 1 5I-HDL and 100 percent of soluble
I-apo-A-II. However, none of the
apo-A-I-specific antibodies bound more than 60
percent of either HDL or soluble apo-A-I.
These results suggested the existence of
intrinsic immunochemical heterogeneity of apo-A-I
both as organized on HDL as well as in free apo-A I
-32-
in solution. The validity of this observed
heterogeneity was supported by demonstrating that (i)
increased binding of HDL occurred when each of the
apo-A-I antibodies was combined with another ~o form
an oligoclonal antibody mixture, and (ii)
approximately 100 percent binding of HDL occurred
when any two apo-A-I antibodies (antibodies
denominated A-I-4 and A-I-7; i.e., Mab A-I-4, of
hybridoma 611 AV63C2.lFl (ATCC HB 8744) and Mab
A-1-7, of hybridoma HA60 HA22G7.5F8 (ATCC HB 8745)
were combined with the single apo-A-II antibody Mab
A-II-1 produced by hybridoma HA61 H112F.31All (ATCC
HB 8743).
To understand the basis for the
heterogeneity of the expression of apo-A-I epitopes
on HDL, two hypotheses were examined. The first
hypothesis that these apo-A-I antibodies
distinguished apo-A-I molecules from different
synthetic sources was not substantiated. Two of the
antibodies bound epitopes on apo A-I molecules in
both thoracic duct lymph as an enriched source of
intestinal HDL and the culture supernatants of the
hepatic cell line Hep G2 as a source of hepatic HDL.
From the assays of this invention, it has
been shown that the monoclonal antibodies identified
differences in the expression of apo-A-I on HDL
subpopulations that were distinguished on the basis
of size or net particle charge; i.e., organizational
heterogeneity appeared to provide the best available
explanation for the immunochemical heterogeneity of
apo-A-I in HDL.
Relative differences in the expression of
three distinct apo-A-I epitopes were demonstrated in
HDL subpopulations obtained by either density
gradient ultracentrifugation or chromatofocusing. In
light of these studies, it is concluded that there is
7~
-33-
intrinsic heterogeneity in the expression of
intramolecular loci representing the apo-A-I epitopes
identified by the monoclonal antibodies of this
invention. Such heterogeneity must be considered in
analysis of the biology and physiology of apo-A-I and
lipoprotein particles bearing this chain as well as
any attempt to immunologically quantitate or
characterize HDL.
II. ASSAY METHODS
The monoclonal receptor molecules of the
present invention are particularly useful in methods
for assaying the presence and amount of an
apolipoprotein A such as that of HDL in a sample to
be assayed such as blood, serum or plasma. As noted
hereinafter, the presence and amount of HDL and
soluble apolipoproteins A may also be assayed in
other body fluids such as ~ymph, and in ln vitro
materials such as hepatic cell cultures and the like.
Useful solid and liquid phase assay methods
are discussed hereinafter. However, the invention is
not so limited. Further, while the particularly
described assay methods utilize a radioactive element
and determination of receptor bound in apolipoprotein
A/receptor-containing immunoreactants
(radioimmunoassay; RIA), the present invention is
also not specifically limited to such assays.
Additional assay methods are described hereinbelow
with particular emphasis on sol d phase immunoassay
methods.
Solid phase assay methods are comprised of
an antigen or a receptor of this invention affixed to
a solid matrix as a solid support.
The antigen or receptor is typically affixed
to the solid matrix by adsorption from an aqueous
medium, although several modes of adsorption, as well
~2~ 3
34-
as other modes of affixation, well known to those
skilled in the art may be used. Exemplary of such
modes are the reaction of the receptor or antigen
with the reactive carboxyl functionality produced by
the reaction of cyanogen bromide with
glucose-containing matrices such as cross-linked
dextrans or cellulosics, glutaraldehyde linking as
discussed hereinafter in conjunction with latex
particles, and the like.
Useful solid matrices are well known in the
art. Such materials include the cross-linked dextran
available under the trademark S~PHADEX from Pharmacia
Fine Chemicals (Piscataway, NJ); agarose; beads of
glass; polystyrene beads about 1 micron to about 5
millimeters in diameter available from Abbott
Laboratories of North Chicago, IL; polyvinyl
chloride, polystyrene, cross-linked polyacrylamide,
nitrocellulose or nylon-based webs such as sheets,
strips or paddles; or tubes, plates or the wells of a
microtiter plate such as those made from polystyrene
or polyvinylchloride.
Latex particles us~ful in agglutination-type
assays are also useful solid matrices. Such
materials are supplied by the Japan Synthetic Rubber
Company of Tokyo, Japan, and are described as
carboxy-functional particles dispersed in an anionic
soap. Typical lots of such particles have an average
diameter of 0.308 microns, and may have an average
carboxy-functional group distribution of about 15 to
about 30 square Angstroms per carboxy group.
Prior to use, the particles are reacted with
a diamine such as 1,3-diamino-2-propanol to form a
plurality of amide bonds with the particle carboxy
groups while maintaining free amine groups. The free
amines are thereafter reacted with a dialdehyde such
23
-35-
as glutaraldehyde and the receptor or antigen to form
Schif base reaction products. The Schiff base
reaction products are thereafter reduced with a
water-soluble reductant such as sodium borohydride to
provide a useful solid support.
Those skilled in the art will understand
that there are numerous methods for solid phase
immunoassays that may be utilized herein. Exemplary,
useful solid phase assays include enzyme-linked
immunosorbant assays (ELISA) and fluorescence immune
assays (FIA), in addition to the specifically
discussed RIA. However, any method that results in a
signal imparted by the reaction of apolipoprotein A
with a receptor of this invention is considered.
Each of those assay methods may employ single or
double antibody techniques in which an indicating
means is utilized to signal the im~unoreaction, and
thereby the binding of an apolipoprotein A that is to
be assayed with a receptor of this invention.
Exemplary techniques may be found explained in
Maggio~ Enzyme Immunoassay, CRC Press, Cleveland, OH
(1981); and in Goldman, Fluorescent Antibody Methods,
Academic Press, New York; NY (1980).
Broadly, the presence of an apolipoprotein A
such as that of human HDL in a sample to be assayed
includes the following steps.
(a) An effective amount of a monoclonal
receptor of this invention whose antibody combining
site immunoreacts with and binds to human
apolipoprotein A, but is free from immunoreaction
with and binding to human apolipoproteins B, D, D,
and E, or other known proteins or ligands is
provided. The receptor is also free from
immunological binding with any other protein or
ligand found in plasma or serum of normal
v`~3
-36-
individuals. This is typically accomplished by using
an aliquot of an appropriate hybridoma supernatant or
ascites.
The effective amount of receptor will
differ, inter alia, with the particular receptor
used, and with the particular assay method utilized,
as is well known. ~lso ~Jell known is the ease ~ith
which the effective amount may be determined using
standard laboratory procedures by one skilled in
preparing such assays.
(b) A known amount of the receptor is
admixed with aliquot of a sample to be analyzed for
the presence of an apolipoprotein A such as that of
human HDL, to form an admixture. The admixture so
lS formed may be a liquid admixture as in the liquid
phase RIA desceibed hereinafter, or that admixture
may be a solid/liquid admixture as where a solid
support is utilized.
(c) In either event, the admixture 50
formed is maintained for a predetermined period of
time from minutes ~o hours, such as about 90 minutes
to about 16-20 hours at a temperature of about 4
degrees to about 45 degrees C that is sufficient for
the receptor to immunoreact with and bind to apo-A
present in the sample, and form an immunoreactant.
(d) The amount of receptor bound in the
immunoreactant is then determined to thereby
determine the amount of apo-A as in HDL present in
the sample. That amount may be zero, thereby
indicating that no apo-A is present in the sample,
within the limits that may be detected.
Individual receptors of this invention may
be utilized or the individual receptor molecules may
be admixed for use~ The particular receptor or
combination to use for assaying for the presence of a
~7~3
particular apo-A-containing molecule may be
determined from the data of the RESULTS section (IV)
that follows. Thus, one may select a receptor that
immunoreacts with and binds to apolipoprotein A-I, or
A-II, or both A-I and A-II.
For example, if it is desirable to analyze
only apo-A-lI molecules, the receptor of choice
(A-II-l) is that produced by the hybridoma
denominated HA61 ~112F3.lAll (ATCC HB 8743). If only
apo A-I subsets are desired, then each of the three
different receptors (A-I-4, A-I-7 or A-I-9) provide a
reagent for each subset defined by these receptors.
Where the total HDL present in a sample is desired, a
mixture containing A-II-l receptors plus receptors
produced by any two of the other three hybridomas of
this invention, i.e.~ receptors denominated A-1-4,
A-1 7, or A-1-9 (from hybridomas 611 AV63C.21Fl, ATCC
HB 8744; HA60 HA22GF.5F8, ATCC HB 8745; or
HA62 HA227A2.7D3~ ATCC HB 8741; respectively).
In one embodiment o~ the above, general
method, an apolipoprotein A that is bound by the
receptor used in the method such as human HDL is
provided affixed to a solid matrix as a solid support
antigen. The admixture of step (c), above, is
present as a liquid admixture, and is admixed with
the solid support to form a solid/liquid phase
admixture. That solid/liquid phase admixture is
maintained for a predetermined time period such as
about 16-18 hours at 4 C that is sufficient for the
receptor molecules in the liquid admixture to
immunoreact with and bind to the antigen and form an
immunoreactant. The solid and liquid phases are
separated, and the solid phase is usually rinsed to
remove non-specifically bound receptor molecules.
~'~4~ 3
-3~-
The amount of receptor molecules bound (specifically)
in the immunoreactant is then determined.
Where the sample is free from apo-A
molecules, the amount of receptor in the solid phase
immunoreactant is relatively high. Conversely, where
there is a relatively large amount of apo-A molecules
as where there is a large amount of human HDL present
in the sample, the amount of bound receptor is
relatively lower. Quantitative comparison of the
result obtained with separately obtained control
results provides quantitation of the amount of apo-A
in the sample.
The determination of the amount of receptor
bound may be by means of an indicating
means-containing reagent that reacts with the bound
receptor but does not react with the solid support
antigen such as 125I-labeled goat anti-mouse Ig,
where the receptors are mouse antibodies. The
receptor may itself include a linked indicating means
such as a radioactive element or an enzyme that
signals the formation of an immunoreactantl or an
added ligand specific for another indicating
receptor.
In another embodiment of the general method,
the sample to be assayed may be affixed to a solid
matrix as a solid support antigen prior to forming
the admixture described in the general method in
step (b), above. It is understood that while several
entities from the sample may become affixed to the
solid support, the useful solid support antigen
includes those entities such as HDL that contain
apolipoprotein A.
~ he sample may be affixed in several ways as
are known, and described previously. One exemplary
~Z~7~3
-39-
method is by adsorption as is discussed in connection
with the solid phase RIA described hereinafter.
When the sample is affixed to the solid
support prior to formation of the admixture of
step (b), the admixture formed in that step is a
solid/liquid admixture in which the solid phase is
the solid support antigen and the liquid phase is the
aqueous composition that includes a receptor of this
invention. The solid/liquid phase admixture is
maintained as already described, and i~ separated
prior to determining the amount of receptor that is
bound in the immunoreactant. The separated solid
phase is typically rinsed prior to that determination
being made, as discussed before.
lS A convenient way to determine the amount of
receptors bound in the above-described method
utilizes an indicating means-containing reagent that
reacts with the bound receptors to form a bound
reaction product, but does not bind to the solid
support antigen. The indicating means of the reagent
signals the presence of the bound receptor.
A known amount of a liquid composition
including such a reagent is admixed with the
separated solid phase to form a second solid/liquid
admixture. That admixture is maintained for a
predetermined period of time sufficient for the
reagent to react with the bound receptor of the
immunoreactant and form a bound reaction product.
The solid and liquid phases are thereafter
separated as described before and the amount of bound
reaction product is determined.
In the case of the specifically disclosed
RIA, the reagent was goat anti-mouse antibodies that
immunoreact with and bind to the mouse-derived
receptor molecules. That reagent included linked
7~:23
-4~-
iodine-125 atoms (indicator) whose gamma radiation
provided the signal that bound receptor was present
in the solid phase, and consequently that an human
apolipoprotein A was present in the sample.
The indicating means may also be an enzyme
or a fluorescent molecule that is linked to the
reagent for use in an enzyme-linked immunosorbent
assay (ELISA) or fluorescence immunoassay ~FIA),
respectively.
For an ELISA, typically used enzymes linked
to the reagent as a signalling means include
horseradish peroxidase, alkaline phosphatase and the
like. Each of those enzymes is used with a
color-forming reagent or reagents (substrate) such as
hydrogen peroxide and o-phenylenediamine; and
~-nitrophenyl phosphate, respectively.
Enzyme-linked antibody (conjugate) reagents
of one animal raised to the antibodies of another
animal such as peroxidase-linked rabbit anti-goat and
goat anti-mouse antibodies, as well as
phosphatase-linked rabbit anti-goat, and rabbit
anti-mouse antibodies are commercially available from
several suppliers such as Sigma Chemical Company of
St. Louis, MO. Those indicating means-containing
reagents may be used where the receptor utilized has
an Fc portion of the "other animal", e.g., goat and
mouse.
Similar assays may also be carried out using
fluorochrome dyes linked to an antibody as an
indicating means-containing reagent to signal the
presence of receptors bound in an immunoreaction
product The fluorochrome dye is typically linked by
means of an isothiocyanate group to form the
conjugate. Exemplary fluorochrome dyes include
fluorescein isothiocyanate (FITC), rhodamine B
-41-
isothiocyanate (RITC~ and tetramethylrhodamine
isothiocyanate (TRITC). Conjugates such as
FITC-linked rabbit anti-mouse, goat anti-mouse, goat
anti-rabbit and sheep anti-mouse antibodies are
commercially available from several sources such as
Sigma Chemical Company.
In addition to the RIA, ELISA and FIA
techniques for determining the presence of receptors
of this invention bound to an antigen in an
immunoreactant, other well known techniques are also
available. In one technique, protein A of
Staphylococcus aureus linked to a signalling means
~ ~r
such as 1 ~I is utilized to determine the presence
of the receptors bound to the solid support.
In another technique, biotin linked to an
antibody reagent is utilized to signal the presence
of the immunoreactant in conjunction with avidin that
is itself linked to a signalling means such as
horseradish peroxidase. Biotin-linked antibody
conjugates such as biotin-linked goat anti-rabbit,
goat anti-mouse and rabbit anti-goat IgG's are
commercially available from Polysciences, Inc. of
Warrington, PA. Avidin-FITC, avidin-RITC,
avidin-peroxidase and avidin-alkaline phosphatase are
also available commercially from Polysciences, Inc.
for use with the biotin-linked antibody conjugates to
provide the signal. Still other techniques are well
known to those skilled in this art.
In a still further embodiment of the
3~ before-described method, the admixture formed in step
(b) is a liquid admixture; i.e., the sample to be
assayed and the receptors are admixed in a liquid
composition that is typically aqueous. That
admixture includes a known amount of a radiolabeled
~7~3
-42-
apoliprotein A-containing competitive antigen such as
HDL, or free human apo-A-II or apo-A-I.
Where such a liquid phase admixture is used,
the amount of receptor bound in the immunoreactant
may be determined by admixing an excess of an
antibody that immunoreacts with, binds to and
precipitates the receptors with the liquid phase
admixture, to form a second liquid phase admixture.
The precipitating antibody so used does not
immunoreact with, bind to or precipitate the
apolipoprotein being assayed for or the competitive
antigen. An exemplary antibody is the 125I-goat
anti-mouse Ig used in a ~IA described hereinafter.
The second liquid phase admixture is
maintained for a predetermined period of time
sufficent for the admixed antibody to immunoreact
with, bind to and precipitate the receptors of the
immunoreactant, and form a precipitate and a
supernatant.
~he precipitate and supernatant are
separated; and the radioactivity present in ~he
precipitate is measured. That measurement, when
compared to control valves obtained with known
amounts of assayed apolipoprotein A, radioactive
competitive antigen, receptor and precipitating
antibody, may be used to provide the amount of
receptor bound in the immunoreactant, and thereby the
amount of apolipoprotein A present in the sample
assayed.
A still further aspect of the invention
contemplates the use of the before-mentioned latex
particles as a solid matrix of a solid support. In
an exemplary method, a receptor of this invention is
affixed to the latex particles, as described before,
;23
-43-
prior to the admixture of step (b) of the previously
described, general method.
The sample to be assayed is admixed in an
aqueous medium with those particles to form a
solid/liquid phase admixture that is a dispersion of
solid latex particles in an aqueous medium. The
admixture is maintained for a time period sufficient
for an immunoreactant to form, which formation causes
the latex particles to agglutinate.
The time required for the latex particles to
agglutinate is measured. That measurement provides a
determination of the amount of receptor bound in an
immunoreactant, and thereby the presence and relative
amount of apolipoprotein A present in the sample by
comparison with values obtained with controls.
Similar agglutination methods may be
performed with red blood cells (hemayglutination) or
with other agglutinatable particles or cells
following the above steps.
Still further assay methods within the
before-described general method may also be
employed. Each of those methods differs from those
previously described by the manner in which the
amount of immunochemical binding is determined
~5 One group of such methods utilizes optical
measurements for that determination. In one
exemplary procedure, a liquid admixture is formed in
before-described steps (b) and (c) and the turbidity
of the liquid admixture is measured and compared to
control values. In another embodiment, the change in
light scattering after step (c) is compared to
control values.
A still further method utilizes the direct
precipitation of the immunoreactant formed. The
amount of binding may also be determined by noting
7~3
-44-
changes in electrophoretic mobility of the liquid
admixture of step (c) under non-denaturing
conditions.
Yet another method utilizes a receptor of
this invention affixed to a soid matrix such as
SEPHAROSE beads as an a~finity sorbant. Here, the
admixture formed in step (b) is a solid/liquid
admixture that physically separates the
immunoreactant from the liquid portion of the
admixture. The liquid portion is thereafter
subjected to electrophoretic separation and compared
to a similar separation using another aliquot of the
sample to determine whether an apolipoprotein A was
present in the sample.
It is to be noted that values obtained from
appropriate controls are stated as being utilized in
several o~ the methods. It is to be understood that
such control values are obtained separately, and may
be so obtained before, during or after the recited
steps.
III. DIAGNOSTIC SYSTEMS
The present invention also contemplates
diagnostic systems, preferably in kit form. Several
embodiments of a diagnostic system are contemplated.
However, each diagnostic system comprises at least
one package that contains a known amount of a
monoclonal receptor of this invention that
immunoreacts with and binds to human apolipoprotein
A, but is free from immunoreaction with and binding
to apolipoproteins B, C, D and E.
Exemplary packages include glass and plastic
such as polyethylene and polypropylene bottles or
vials; plastic, plastic-metal foil, and plastic-metal
foil-paper envelopes, and the like. The receptor may
be packaged in an aqueous liquid form as in ascites
~7~;~3
-45-
or buffer, but preferably, the receptor is supplied
in dried form such as that provided by lyophilization.
A known amount of the receptor is provided.
That amount is at least enough to carry out one
assay. The provided receptor is typically supplied
in a form and amount that is designed to be diluted
to a prescribed volume with water, saline or a buffer
such as phosphate-buffered saline at pH 7.3-7.5.
In another embodiment, the system includes a
second package that includes a known amount of an
apolipoprotein A with which the receptor immunoreacts
and binds to form an immunoreactant. The
apolipoprotein A is provided affixed to a solid
matrix as a solid support antigen.
Useful solid matrices are as already
described. Preferably, however, the solid matrix is
the well of a microtiter plate. The microtiter plate
forms the package for the wellr but may also be
separately enclosed in a paper envelope or plastic
film to avoid contamination of the wells.
In a Eurther embodiment, the receptor is
provided affixed to a solid matrix as a solid
support. Exemplary of such a solid support are
receptor-affixed latex particles that are dispersed
in an aqueous medium as previously described.
Additional packages may also be included in
the system. Such packages may contain (i) buffer
salts in dry or liquid form, (ii) substrates such as
hydrogen peroxide and o-phenylenediamine, (iii) an
indicating means-containing reagent such as
peroxidase-linked goat anti-mouse antibodies in a
liquid or dry Eorm, and the like.
It is also noted that the receptor that is
required for a diagnostic system of this invention
may be any individual receptor of this invention or
~2'~7~'2~3
-46-
may be a mixture that contains the antibody-combining
sites (idiotype polypeptide portions) of two or more
such receptors.
IV. RESULTS
A. Apo~rotein SPecificity
Each of the four monoclonal antibodies
(designated A-I-4 antibody from 611 AV63C2.lFl
hybridoma; A-I-9 antibody from HA62 HA227A2.7D3
hybridoma; A-I-7 antibody from HA60 HA22G7.5F8
hybridoma; and A-II-l ~rom HA61 H112F3.LAll
hybridoma) was selected on the basis of its capacity
to bind intact HDL. Three were selècted by screening
for antibodies that reacted with the immobilized
immuni~ing antigen using a solid-phase RIA. The
fourth (A-II-l) was selected on the basis of indirect
precipitation of soluble 1 5I-HDL in a fluid-phase
assay. In addition to the immunizing antigen, the
antibodies produced by each of the four hybridomas
bound to immobilized human HDL in a solid-phase RIA,
suggesting that each of these antibodies was specific
for one of the apolipoproteins of human HDL.
Antibody specificity was determined by
Western blotting of the electrophoretically separated
apolipoproteins of human VLDL, LDL, and HDL, as well
as isolated apo-A-I and apo-A-II. Antibodies A-I-4,
A-I-7, and A-I-9 bound completely to apo-A-I of HDI,
and isolated apo-A-I. Some of the antibodies
identified trace amounts of what appeared to be
contaminating apo-A-I in both LDL and isolated
apo-A-II; i.e., proteins that were marginally visible
in the stained gel.
The one exception to this pattern of
reactivity was antibody A-II-l. This antibody bound
to isolated human apo-A-II dimers and apo-A-II
~7~23
-47-
monomers as well as the apo-A-II dimers and monomers
o~ human HDL ~Figure l).
In addition, this antibody bound a trace
VLDL protein of apparent molecular weight of 52,000
daltons that was not readily observed in the
protein-stained gel. This protein, which appeared to
be present also in HDL, had a mobility that was
intermediate between apo-E and albumin, and may have
been an apo-E-A-II dimer as described by Weisgraber
and Mahley, J. Biol. Chem., 253~ 6281-6288 (1978)o
Thus, three of the monoclonal antibodies
were specific for apo-A-I, and the fourth was
specific for apo-A-II. The numerical antibody
designations shown in Figure l reflect this
apolipoprotein specificity. In addition, each of the
apo-A-I antibodies bound multiple apo-A-I isoforms
includin~ A-I-l, A-I-2, and pro-A-I from either HDL
or isolated apo-A-I, after separation of those
isoforms in isoelectric focusing gels.
2~ B. Lipoprotein Specificit~
To characterize the reactivity of these
antibodies for native HDL, binding of the antibodies
to l25I-HDL was studied in a fluid-phase
double-antibody RIA. Antibody binding was measured
at a final antigen concentration of 66.7 ng of
l25I HDL/ml. Maximum binding of l25I-HDL by each
of the four antibodies in antibody excess varied from
18 to 56 percent for the apo-A-I-specific antibodies
and was 87 percent for the apo-A-II-specific antibody
(Figure 2). It was notable that lOO percent binding
of l I-HDL was uniformly expressed by the
apolipoprotein chains as organized on all HDL
particles.
As reported by Chung and Albers, Lipid
Res. 23, 747-753 (1982) r HDL of density equal to
-4~-
1.063 to 1.21 contains two types of particles: (i)
particles that contain apo-A-I and apo-A-II in an
approximate 2:1 molar ratio; and (ii) particles that
contain apo-A-I but no apo-A-II. Therefore, it was
not surprising that the apo-A-II antibody did not
bind 100 percent of HDL. However, if all HDL
particles contained at least apo-A-I, other
explanations must exist for the inability of any one
of the three apo-A-I antibodies to bind all HDL.
Because each of the antibodies bound all isoforms of
the isolated apolipoprotein after electrophoresis in
SDS, the ability of these antibodies to recognize the
isolated apolipoprotein in a fluid-phase RIA also was
examined.
Antibody A-II-l bound 100 percent of
125I-apo-A-II (Figure 2). Thereforet this protein
chain appeared to be immunochemically homogeneous in
that all apo-A-II molecules expressed the epitope
defined by the A-II-l antibody.
~owever, none of the apo-A-I-specific
antibodies bound 100 percent of soluble
125I-apo-A-I (Figure 2). In antibody excess,
antibodies A-I-4, A-I-7, and A-I-g bound 55, 60, and
13 percent of I-apo-A-I, respectively.
To determine if there was a difference in
the apolipoprotein composition of 125I-HDL
particles bound by each antibody, as opposed to those
particles that were not bound by antibody,
precipitates and supernatants formed in the presence
of high concentrations of monoclonal antibody (and a
slight excess of precipitating antibody to fully
precipitate all monoclonal antibody) were dissolved
in SDS and electrophoresed on SDS-PAGE. A
representative autoradiograph of the bound
(precipitate) and unbound (supernatant) fractions of
~,~ 47~23
-49-
1 5I-HDL after reaction with antibodies A-I-7,
A-II-l, and A-I-4 is shown in Figure 3.
All apo-A-I-specific antibodies, including
antibodies A-I-4 and A-I-7, bound I-HDL
particles that contained both apo-A-I and apo-A-II,
and the bound fractions were indistinguishable from
either the starting I-HDL or the unbound
125I-HDL; i.e., the unbound 125I-HDL contained
nonprecipitable apo-A-I. In contrast, antibody
A-II-l appeared to bind most if not all of the
125I-HDL that contained apo-A-II, because the
unbound supernatant fraction from this reaction
mixture was free of demonstrable apo-A-II dimers or
monomers ~Figure 3)O Thus, antibody A-II-l bound all
HDL particles that contained apo-A-II, whereas none
of the A-I-specific antibodies were capable of
binding all HDL particles that contained only apo-A-I.
C. Incomplete Binding of Antigen
To explain the inability of the
apo-A-I-specific antibodies to bind to and facilitate
total precipitaton of either I-HDL, soluble
I-HDL or soluble 125I-apo-A-I, two general
possibilities were con~idered: (i) heterogeneity of
apo-A-I with respect to expression of epitopes; and
(ii) nonoptimal conditions of analysis of binding.
In initial studies, the optimum time and temperature
was determined for the maximum binding of antibodies
A-I-4, A-I-7, A-I-9, and A-II-l to 125I-HDL in
fluid phase. For each of these antibodies, maximal
30 binding was observed within 18-20 hours at either 4
or 24 degrees C. The quantity of 125I-HDL that was
bound by each antibody was maximal and independent of
the amount of antibody added under conditions of
antibody excess. In additional studies, it was shown
that (a) antibody binding was independent of the
7~2~3
-50-
amount of antigen added; i.e., antibody affinity; (b)
radioiodination of apo-A-I or HDL did not interfere
with antibody binding; (c) mild antigen dissociating
conditions such as heating and detergents did not
expose additional antigen epitopes; and (d)
individual allytypic differences in apo-A-I did not
account for the incomplete binding of HDL.
Because none of the above manipulations led
to complete binding of HDL, the alternative
possibility was considered that there may be
heterogeneity of apo-A-I. It was hypothesized that
all apo-A-I molecules in plasma were not absolutely
identical; i.e., all molecules of apo-A-I did not
uniformly express the epitopes defined by the three
apo-A-I-specific antibodies. If each apo-A-I
antibody bound a different epitope on apo-A-I, and if
all HDL particles contained an apo-A-I expressing one
or more of these epitopes, then complete binding of
all 125I-HDL particles would be observed by
combining the three apo-A-I-specific antibodies~
When all possible combinations of two or three
apo-A-I-specific antibodies were analyzed for
binding, only incomplete binding of 125I HDL was
observed as shown in Table 1, below.
7~
TABLE 1
All 125I-HDL Bound By One Apo-A-II-
Specific And Three Apo-A-I-Specific Antibodies
- i25
Antibody I-HDL bound 1% of maximum)
Alone In combination
10 A-I-4, A-I-7 44+1; 61+3 80 + 3
A-I-4, A-I-9 44; 32+2 63 + 2
A-I-7, A-I-9 61; 32 76 + 4
A-I-4, A-I-7, A-I-9 44, 61, 32 83 + 3
A-II-l, A-I-4 67 + 6, 44 92 + 2
A~ l, A-I-7 67; 61 93 + 1
A~ l, A-I-9 67; 32 87 + 3
20 A-II-l, A-I-4~ A I-7 67, 44, 61 100 + 2
A-II-l, A-I-4, A I-9 67, 44, 32 98 + 3
A-II-l, A-I-7~ A I-9 67, 61, 32 99 + 1
125I-HDL was used at 66.7 ng/ml in the
fluid-phase RIA
In view of the before-discussed results~
each of the apo-A-I-specific antibodies must bind a
different epitope because as each additional antibody
was added, additional apolipoprotein A-I was bound,
although all antibodies were present in excess. All
combinations of the A-I antibodies were present in
excess. All combinations of the A-I antibodies bound
more HDL than any single A-I antibody, and the
oligoclonal mixture of the three apo-A-I antibodies,
~24'7~23
-52-
and the oligoclonal mixture of the three apo-A-I
antibodies most closely approached complete binding
of 125I-HDL. These results suggest that HDL
particles may exist that either do not contain
apo-A-I or contain apo-A-I molecules that are not
recognized by any of these apo-A-I-specific
antibodies.
Because complete binding of 125I-HDL could
not be obtained with any combination of the three
apo-A-I-specific antibodies, binding all 125I-HDL
was examined by combining each of the
apo-A-I-specific antibodies individually with the
apo-A-II-l antibody (Table 2).
TABLE 2
The Percent Of 1 5I-H~L Bound By Each Antibody Was
Independent Of The Amount Of 125I-HDL Added
125 125
I-HDL Added I-HDL Bound (Percent of Maximum)
tng/ml) A-I-4 A-I-7 A-I-9
1031.8 + 2.8 43.7 + 3.1 30.2 + 0.5
3030.6 + 4.1 42.0 + 7.7 29.0 + 2.3
10033.3 + 1.1 44.4 + 0,6 29.2 ~ 1.1
30034.7 ~ 3.2 ~4.4 ~ 1.7 29.1 + 1.2
To insure that each of the apo-A-I specific
antibodies identified all apo-A-I isoforms, HDL and
isolated apo-A-I were separated by isoelectric focusing in
polyacrylamide gels and were Western blotted to
nitrocellulose for reaction with antibody. The left panel
of Figure 1 is a photograph of a Coomassie Brilliant Blue
R250 stained gel before electrophorectic transfer of the
-53-
apolipoproteins to nitrocelluloseO The remaining three
panels are 2~ hour autoradiographs of identical
nitrocellulose paper transfers after incubation with each
of the individual antibodies and 125I-goat-anti mouse Ig
(0.5 milliCi/ml). As shownl each apo-A-I antibody bound
multiple apo-A-I bands, suggesting that none of the
antibodies distinguished among the various isoforms. No
combination of one apo-A-I-specific antibody with antibody
A-II-l resulted in 100 percent binding of 1 5I-HDL.
However, the combination of any two
apo-A-I-antibodies with the single apo-A-II-specific
antibody resulted in 100 percent binding of 125I-HDL
(Table 1). Those results confirm that all HDL particles
express at least one of the three apolipoprotein epitopes
defined by antibodies A-II-l, A-I-4, and A-I-7; A-II-l,
A-I-4, and A-I-9; or A-II-l, A-I-7, and A-I-9, and thus
establish limits on the degree of heterogeneity.
B. HDL and Apoprotein Affinity
Because complete binding of 125I-HDL could be
achieved with an oligoclonal mixture of monoclonal
antibodies, the feasibility of using these antibodies to
accurately quantitate total plasma HDL was further
analyzedO The quantity of apo-A-I measured in HDL and
apo-HDL with polyclonal antisera has often been different,
suggesting that the affinities of antibodies might differ
for soluble apolipoproteins as compared with the same
apolipoproteins when they are organized on HDL.
Competitive RIAs with 1 5I-HDL were used in which the
ability of HDL and the isolated apolipoprotein to compete
for binding of 125I-HDL was analyzed to identify
differences in antibody affinities for HDL and apo HDL.
Slope analysis of the logit-transformed competitive
curves indicated that two of the antibodies, A-I-7 and
A-II-l, had the same affinity for the isolated
apolipoprotein and for that apolipoprotein when
~4~3
organized on HDL, whereas the other two antibodies,
A-I-4 and A-I-9 differed. For both antibodies A-I-4
and A-I-9, the affinities were less for free apo-A-I
than for apo-A-I organized on HDL.
C. Expression of Apo-A-I and Apo-A-II
Epitopes by Apoproteins of ~ifferent
Biosynthetic Origin
The apo-A-I and apo-A-II epitopes defined by
the antibodies of this invention were examined to
determine if apolipoproteins from different
biosynthetic sites differed in epitope expression.
Included in this analysis were (a) conditioned
culture medium from the hepatic cell line ~ep G2; ~b)
human lymph collected by thoracic duct drainage; and
(c) unfractionated whole plasma, lipoprotein-depleted
plasma, VLDL, and HDL from the same pooled plasma
source. Each of those samples was examined for
epitope expression by competitive inhibition
immunoassay for each monoclonal antibody using
125I-HDL as the ligand. Inhibition was based on
total protein added. The 125I-~DL used in each
immunoassay was obtained from a pooled plasma source.
Each apolipoprotein source was analyzed at
three levels. First it was determined whether the
protein competitvely inhibited antibody binding to
125I-HDL. Second, if inhibition was observed, the
affinity of the antibody for the competing protein
was determined by slope analysis and compared with
the affinity for HDL. Third, if similar affinities
were observed, the quantitative expression of the
epitope by the competing protein (based on total
protein added) was compared with that expressed by
either HDL or plasma. If the affinities were not the
same as detexmined by slope analysis, no quantitative
conclusions could be drawn.
~Z~7~:~3
-55-
The epitope defined by antibody ~-I-4, which
bound a subset of apo-A-I present on 40-50 percent of
125I-HDL, was not expressed by apo-A-I of VLDL or
by hepatocyte-derived apo-A-I present in culture
medium from the Hep G2 cells This epitope was
expressed in lipoprotein~deficient plasma (LPDP), but
the affinity for the epitope in LPDP was less than
the epitope expressed by apo-A-I organized on HDL.
As demonstrated previously, the affinity of this
antibody for isolated Apo-A-I was less than its
affinity for HDL, indicating differences in the
defined epitope. This suggested that the majority of
the apo-A-I in LPDP was not associated with
lipoprotein particles.
In contrast this epitope was expressed by
apo-A-I in normal human plasma (NHP) and in thoracic
duct l~mph (lymph) with an affinity that was
indistinguishable from that for plasma-derived HDL,
suggestin~ that the apo-A-I of NHP and lymph was
associated with lipid. On the basis of total protein
added, more A-I-4 epitope was detected in lymph than
in NHP.
Antibody A-I-7, which identified a major
apo-A-I epitope denominated A-I-7, also did not bind
apo-A-I of plasma VLDL. The A-I-7 epitope expressed
by molecules in the Hep G2 culture medium and lymph
interacted with a higher affinity compared with HDL,
whereas the A-I-7 epitope expressed by molecules in
LPDP and NHP interacted with its antibody with an
affinity that was indistinguishable from that of the
same epitope expressed on HDL.
As demonstrated abo~e, the affinity of this
antibody for isolated apo-A-I and apo-A-I in HDL is
the same. Thus, this antibody did not appear to
distinguish between free or lipid-associated
2~
apo-A-I. The difference in affinity of the antibody
for the A-I-7 epitope in lymph and hepatocyte medium
indicated a modification of apo-A-I in these
sources. On a quantitative basis, the amount of
apo-A-I in LPDP by reference to the A-I-7 epitope was
11 percent of the apo-A-I present in normal human
plasma.
The third apo-A I epitope identified by
antibody A-I-9 was distinguished from epitope bound
by antibodies A-I-4 and A-I-7 by its expression by
apo-A-I on VLDL, although it was not expressed by
molecules in LPDP. The A-I-9 epitope was expressed
by apo-A-I from Hep G2 cells in culture.
Compared with HDL, the A-I-9 epitope of
apo-A-I of lymph had higher affinity for antibody,
that of VLDL had lower affinity, and the same
affinity was observed for Hep G2 culture medium and
NHP. The epitope bound by antibody A-I-9 was thus
sub]ect to fine differences in structure on different
apo-A-I.
The apo-A-II epitope identified by antibody
A-II-l was present in all samples s~udied. Compared
with HDL, the A-II-l epitope was expressed with the
same affinity by molecules in Hep G2 culture medium,
LPDP, and NHP. This apo-A-II epitope interacted with
epitopes that appeared the same for isolated apo-A-II
and apo-A-II organized in HDL. The apo-A-II of LPDP
represented 3.4 percent of the apo-A-II present in
NHP. Surprisingly, the binding affinity of this
antibody for apo-A-II in VLDL and lymph was slightly
greater than its affinity for HDL.
~4~ 3
D. Expression of Apo-A-I and Apo-A-II
Epitopes in HDL Subfractions
Epitope expression by HDL subpopulatians
separated by density gradient ultracentrifugation and
chromatofocusing was examined to determine if HDL
subpopulations differing in apo-A-I and apo-A-II
epitopes could be distinguished on the basis of
particle size or composition. Five HDL density
subfractions were isolated from a single plasma
source that was pooled from three donors. The
apoprotein composition of the subfractions was
characterized by SDS-PAGE (sodium dodecyl
sulfate-polyacrylamide gel electrophoresis), and the
particle size distribution of particles present in
these HDL subfractions was characterized by PAGE
(Figure 4).
HDL density subfraction 1 (the lowest
density H~L) was distinguished by SDS-PAGE from
fractions 2 through 5 by the presence of apo-B,
greater quantities of apo-E, and relatively small
amounts of apo-A-II (dimer and monomer) and apo-D.
Because these HDL subfractions were obtained from a
single ultracentrifugation to minimize apoprotein
loss and potential perturbation of the HDL particles,
they also contained small amounts of other plasma
proteins (Figure 4, top).
Electrophoresis of the HDL density
subfractions on 4-30 percent polyacrylamide gradient
(PAGE) pore gels in the absence of SDS or other
dissociating agents demonstrated the presence of
~arying proportions of HDL particles of at least two
sizes. Predominantly large HDL particles were
present in density subfractions 1 and 2, and small
HDL particles predominated in subfractions 4 and 5
(Figure 41 bottom)~ In addition, the HDL
7~;~3
-58-
subfractions differed with respect to their total
cholesterol content. The light HDL fractions (1 and
2) contained the largest amount of free and
esterified cholesterol/mg of total protein.
When epitope expression by each ~IDL
subfraction was analyzed by competitive inhibition,
complete inhibition of the binding of each antibody
could be achieved, confirming that each of the
defined epitopes was present in each HDL fraction.
When the competitive inhibition profiles were
analyzed by logit transformation to compare the
qu~litative epitope expression by each subfraction,
slope analysis indicated that the affinity of each
epotope for its antibody did not significantly differ
one from another (p less than or equal to 0~2).
Thus, a relative assessment of the quantitative
expression of each epitope in each HDL subfraction
was feasible.
From the competitive inhibition regression
line, the protein required for 50 per~ent inhibition
of antibody binding was determined (Table 3~. On a
quantitative basis, epitopes A-I-4 and A-I-9 were
expressed at highest conce~tration by subfraction 2,
whereas epitope A-I-7 was most highly expressed by
subfractions 4 and 5O Epitope A-II-l was most
abundant in subfraction 4 as set forth in Table 3
below.
~'7~
-59-
TABLE 3
~uantitative Expression Of Apolipoprotein
A-I And A-II Epitopes And HDL Subfractions
Competitor concentration
per antibo~ya
A-I~4 A-I-7 A-I-9 A-II-l
HDL densit~ fractions
1 (light) 137 20.519.5 1.08
2 46 9.5 8.4 0.44
3 78 7.414.8 0O39
4 78 6.421.9 0.32
5 (heavy) 84 6.432.0 0.49
HDL chromatofocusing
fractionsC
11 (pH 5.0) 72 3.3 5.6 2.70
18 59 3.0 5.4 1.75
27 23 1.5 7.4 0.82
32 20 1.8 8.2 0.53
34 23 1~610.3 0.40
37 (pH 4.4) 109 6.0 27.0 1.38
~5 _ _
a Concentration of competing protein required
to exhibit 50 percent inhibition of antibody binding
expressed in milligrams of total protein per
milliliter.
b Fractions were obtained by density gradient
ultracentrifugation. Apoprotein compcsitions and the
size distributions are shown in Figure 4O
Unfractionated homologous HDL was used as the
radiolabeled ligand and was used at a final
concentration of 66.7 ng/ml. Mean slopes and minimum
~3
-60-
correlation coefficients of the logit-transformed
inhibition curves by all HDL density subfractions
with antibodies A-I-4, A-I-7, A-I-9, and A-II-l were:
-3.01 + 0.30, r greater than or equal to 0.995;
-2.94 + 0.13, r ~reater than or equal to 0.996;
-2.27 ~ 0.14, r greater than or equal to 0.992; and
-3.16 -~ 0.15, r greater than or equal to 0.997,
respectively~ Therefore, no differences were
observed in the affinity of each antibody for its
epitope in each HDL fraction.
c Fractions were obtained by column
chromatography as described in the Materials and
Methods section (V). Unfractionated homologous HDL
was used as the radioiodinated ligand and was added
at a final concentration of 66.7 ng/ml. Mean slopes
and minimum correlation coefficients of the
logit-transformed inhibition curves by all HDL
chromatofocusing fractions with antibodies A-I-4,
A-I-7, A-I-9, and A-II-l were:
-3.11 + 0.2~, r greater ~han or equal to 0.995;
2.87 + 0.21, r greater than or equal to 0.994;
-2.06 ~ 0.17, r greater than or equal to 0.997; and
-3.36 + 0.13, r greater than or equal ~o 0.998,
respectively.
~ comparison of the quantitative expression of
the four apoprotein epitopes in the five HDL
subfractions is illustrated in Figure 5 (top).
Relative epitope expression for each antibody was
calculated from Table 3 by assigning a value of 1.0
to the HDL subfraction that contained, on a protein
basis, the greatest quantity of the epitope. All
other HDL subfractions were then expressed
fractionally. As shown, the relative epitope
expression varied for each density subfraction with
~Z~7~-Z3
-61-
epitopes A-I~4 and A-I-9 predominating in the light
HDL subfractions and epitopes A-I-7 and A-II-l
predominating in the heavy HDL subfractions.
This same epitope analysis was performed on
another set of HDL subfractions separated on the
basis of net charge by chromatofocusing. A single
plasma source pooled from three donors was used to
isolate a d=1.063-1.21 g/ml HDL fraction by
ultracentrifugation that was then chromatographed on
a Polybuffer-Exchange 94 column. The apoprotein
composition of representative subfractions as
determined by SDS-PAGE, and the particle size
distribution of the same subfractions as determined
by PAGE are illustrated in Figure 6. Fractlons 11
and 18 were distinguished from subfractions 32 and 34
by the smaller ~uantity of apo-A~ Figure 6, top).
PAGE demonstrated the presence of varying proportions
of at least two particle sizes. Subfractions 11 and
18 contained predominantly large cholesterol-rich
HDL, whereas subfractions 34 and 37 contained
predominately small cholesterol-poor HDL (Figure 6,
bottom).
When epitope analysis was performed,
complete competitive inhibition of the binding of
each antibody was observed in excess antigen. The
affinity of each antibody for its complementary
epitope expressed by particles in each
chromatofocusing subfraction was equivalent (p equals
0.2). From the competitive inhibition regression
lines, the total protein required for 50 percent
inhibition was determined (Table 3). On a
quantitative basis, epitope A-I-9 was most abundant
in subfractions 11 and 18, epitopes A-I-4 and A-I-7
were most abundant in fractions 27, 32, and 3~,
whereas epitope A-II-l was present in fraction 34
7~23
-62-
(Figure 6, bottom). The most striking feature of the
distribution of each of the apoprotein epitopes was
the predominance of epitope A-l-9 in HDL particles
eluted at pH 5.0 that appeared to contain only
apo-A-I (Figure 6).
V. MATERIALS AND METHODS
_
A. Lipoproteins
During the course of these studies,
lipoproteins were isolated from nine different plasma
pools, each made up of three or more individual
fasting donors. The isolated lipoproteins, including
LDL, density equal to 1.006 g/ml; LDL, density equal
to 1.019 to 1.0~3 g/ml; and HDL density equal to
1.063 to 1.21 g/ml, were dialyzed against lipoprotein
buffer (LLB) containing 150 mM NaCl, 1 mM EDTA, 0.005
percent alpha-tocopherol~ and 5 mM benzamidine, and
were stored under sterile conditions for no more than
21 days. In selected studies to identify potential
allelic differences, the HDL (density equal to 1.063
to 1.21 g/ml) was isolated from plasmas o~tained from
individual normolipidemic donors and treated in the
same manner.
HDL density subfractions were obtained from
a single pooled plasma source by isopycnic density
gradient ultracentrifugation~ After removal of the
lipoprotein of density less than or equal to 1.063
g/ml by a single 18-hour run at 200,000xg, the
infranatant plasma fraction (20 ml) was increased to
a density of 1.21 g/ml and centrifuged at 10 degrees
C through 20 ml of 1.21 g/ml KBr for about 4 to about
8 hours at 200,000xg. Five 4-ml fractions were
collected beginning at the top of the tube, and were
dialyzed into LLB for further analysis.
HDL chromatofocusing fractions were obtained
from a separate pooled plasma source essentially as
~7g?~3
--63--
described by Nestrock et al., Biochem. Biophys. Act,
753, 65-73 tl983). The HDL (density equal to 1.063
to 1.21 g/ml) was isolated by ultracentrifugation and
dialyzed into 25 mM piperazine hydrochloride, having
a pE~ value of 5.8. Forty mg of protein were applied
to a 1.6x30 cm column oE Polybuffer -Exchanger 94
(Pharmacia Fine Chemicals, Piscataway, NJ;
hereinafter Pharmacia) equilibrated with 25 mM
piperazine HCl, pH 5.8, and the HDL was eluted with
PolybufferTM 74 (Pharmacia) diluted 1:15 with
H2O, having a pH value of 4Ø The effluent was
monitored for absorbance at 280 nanometers (nm) and
for p~ value. Six HDL subpopulations corresponding
to those described by Nestrock et al., (supra~, and
eluting at pH maximal values of 5.0, 4.9, 4.8, 4.7,
4.5, and 4.4, respectively, were collected and
desalted by chromatography on SephadexTM G-75
equilibrated with LLB.
B. Isolation of Apoproteins A-I and A-II
Apoproteins A-I and A~II were isolated from
ether/ethanol-delipîdated HDL by chromatography on
DE}~E-cellulose in deionized 6 M urea as described
below and by Blaton et al., Biochemistry; 16,
2157-2163 (1977). The isolated apolipoproteins were
25 stored in dilute solution in 0.1 percent sodium
bicarbonate at -20 degrees C.
C. Lipoprotein Characterization
Lipoproteins were analyzed for protein by a
modification of the method of Lowry [Lowry et al.,
30 J. Biol. Chem. 193, 265-275 (1951)] in the presence
of SDS using a bovine alb~unin standard. Lipoprotein
concentrations were expressed as the mass of
protein. Total and free cholesterol were measured by
the enzymatic fluorometric method. Esterified
35 cholesterol was taken as the difference between total
~4'7~Z3
-64-
and free cholesterol. Results were expressed as
micrograms of cholesterol/mg of total protein.
The apolipoprotein composition of the
lipoproteins was analyzed by polyacrylamide slab gel
electrophoresis in the presence of 0.1 percent SDS as
described by Curtiss et al., J._Biol. Chem., 257,
15213-15221 (1982). The running gels contained a
linear 7.5-20 percent acrylamide gradient. The
apo-A-I isoforms were separated by isoelectric
focusing in a 6 percent polyacrylamide gel containing
8 M urea and 2 percent Ampholine (1 percent having a
pH value between 4-6 and 1 percent having a pH value
between 5-8) as described by Weisgraber et al., J.
Lipid Research, 21, 316-325 (1980). Lipoproteins
were delipidated by boiling for 3 minutes in 1
percent SDS before electrophoresis, and the gels were
stained after electrophoresis with 0.1 percent
Coomassie Brilliant Blue R-250 in 50 percent
trichloracetic acid. Gels containing radioiodinated
lipoproteins were visualized by au~oradiography.
Lipoprotein particle size distributions were
determined by lipoprotein polyacrylamide gradient
pore gel electrophoresis using the system of [Blanche
et al. ~Biochem. Biophys-Acta, 665; 408-419
(1981))]. Samples containing 10-20 micrograms of
protein in 0.008-0.010 ml aliquots were
electrophoresed for 24 hours at 130 volts (constant
voltage) in 4-30 percent acrylamide gradient slab
gels. The high molecular weight calibration kit
(Pharmacia) was used for molecular weight
determinations. The gels were fixed, stained and
destained, and where appropriate, visualized by
autoradiography.
~7¢~3
--65--
D. Generation of Monoclonal Antibodies
The four monoclonal antibodies were obtained
from three separate fusiQnS of splenocytes from
immunized Balb/c mice (Scripps Clinic and Research
5 Foundation Vivarium, La Jolla, CA), using standard
fusion protocols discussed herein. Culture
supernatants were collected and screened by either
solid-phase or fluid-phase radioimmunoassay as
described below. All hybridomas were cloned at least
twice by limiting dilution, and were stored frozen in
liquid nitrogen.
Briefly, Balb/c mice were immunized
intraperitoneally with native human HDL or apo-VLDL
as immunoyen in complete Freund's adjuvant. A
booster injection of immunogen in incomplete Freund's
adjuvant was administered appro~imately 3 to 4 weeks
ollowing the first injection. Three days prior to
harvesting of the mouse spleen, a final booster of
immunogen in normal saline was in~ected intravenously.
The animals so treated were sacrificed, and
the spleen of each mouse was harvested. ~ spleen
cell suspension was then prepared. Spleen cells were
then extracted from the spleen cell suspension by
centrifugation for about 10 minutes at lOnO r.p.m.,
at 23 degrees C. Following removal of supernatant,
the cell pellet was resuspended in 5 ml. cold N~4Cl
lysing buffer, and was incubated for about 10 minutes.
To the lysed cell suspension were added 10
ml Dulbecco's Modified Eagle Medium ~DMEM) (Gibco)
and HEPES [4-(2-hydroxyethyl)-1-
piperidineethanesulfonic acid] buffer, and that
admixture was centrifuged for about 10 minutes at
1000 r.p.m. at 23 degrees C.
The supernatant was decanted, the pellet
resuspended in 15 ml of DMEM and HEPES, and was
~7~23
-66-
centrifuged for about 10 minutes at 1000 r.p.m. at 23
degrees C. The immediately preceding procedure was
repeated.
The pellet was then resuspended in 5 ml DMEM
and HEPES. An aliquot of the spleen cell suspension
was then removed for counting.
Fusions were accomplished in the following
manner using mouse myeloma cell line P3x63Ag8 for
ATCC HB 8744 and line P3x63Ag8.653 for the remaining
hybridomas. Using a myeloma to spleen cell ratio of
about 1 to 10 or about 1 to 5 (the most preferred
myeloma to spleen cell ratio being 1 to 5), a
sufficient quantity of myeloma cells were centrifuged
to a pellet, washed once in 15 ml DMEM and HEPES, and
centrifuged for 10 minutes at 1000 r.p.m. at 23
degrees C. Spleen cells and myeloma cells were
combined in round bottom 15 ml tubes (Falcun). The
cell mixture was centrifuged for 7 minutes at 800
r.p.m. at 23 degrees C, and the supernatant was
removed by aspiration. The remaining cell pellet was
then gently broken into large chunks. Thereafter,
20G microliters of 30 percent aqueous polyethylene
glycol (w/v) (PEG) (ATCC Baltimore, MD) at about 16
degrees C were added, and the mixture was gently
mixed for between 15 and 30 seconds. The cell
mixture was centrifuged 4 minutes at 600 r.pOm. At
about 8 minutes from the time of adding the PEG, the
supernatant was removed.
Then 5 ml DMEM plus HEPES buffer was added
to the pellet, allowed to set for 5 minutes, and was
followed by gently breaking the pellet into large
chunks. This mixture was centrifuged 7 minutes at
600 r.p.m. The supernatant was decanted, 5 ml of ~IT
~hypothanthine/thymidine) media were added to the
pellet and left undisturbed for 5 minutes. The
-67-
pellet was then broken into larye chunks and the
final cell suspension was placed into T75 flasks (2.5
ml per flask) into which 7.5 ml HT media had been
placed previously. The resulting cell suspension was
incubated at 37 degrees C to grow the fused cells.
Three days after fusion the fused cells were plated
out and treated as described below.
In an alternate procedure, the spleens of
the two mice were removed, suspended in complete HT
medium containing 0.1 millimolar azaguanine
[formulated according to Kennett et al., Curr. Top.
Microbiol. Immunol., 81, 77 (1978)], pooled to ~ield
3.2x108 total cells, and fused with mouse myeloma
cells in the presence of a fusion promoter [e.g., 30
percent (weight per volume) polyethylene glycol-1000
to about 4000; ATCC] at a ratio of 10 myeloma cells
per spleen cell as described in Curtiss et al., J.
Biol. Chem., _ , 15213-15221 (1982).
Three days after fusion, viable cells were
plated out in 96-well tissue culture plates at
2x104 viable cells per well (768 total wells) in
HAT (hypothanthine, aminopterin, thymidine) buffer
medium as described in Kennett et al., supra). The
cells were fed seven days after fusion with HT medium
and at approximately 4-5 day intervals thereafter as
needed. Growth was followed microscopically and
culture supernatants that contained antibodies were
collected on day 14 for assay of antigen-specific
antibody production by solid phase radioimmunoassay
(RIA).
The hybridomas so prepared were screened,
assayed, and their viabilities were determined.
The ~ybridomas were given the following
designations for reference purposes and were
deposited on March 6, 1985 with the American Type
)23
-68-
Culture Collection, Rockville, Maryland under the
following ATCC of accession numbers.
Hybridoma ATCC Accession No.
HA62 HA227A2.7D3 HB 8741
HA61 H112F3.1All HB 8743
HA60 HA22GF.5F8 HB 8745
611 AV63C2.2Fl HB 8744
Immunoglubolin heavy and light chains of the
antibodies secreted by the cloned hybridomas were
typed using the Mono AB-ID EIA Kit A (Zymed Labs
Inc., San Francisco, CA). The assays were performed
with hybridoma culture supernatants as described by
the manufacturer~ Those results were as sh/ wn
below.
ATCC Accession No. Isot~pe
HB 8741 IgGl kappa
20 ~B 8743 IgM kappa
HB 8745 IgGl kappa
HB 8744 IgGl kappa
E. Monoclonal Antibody Production
Once the desired hybridoma had been selected
and cloned, the resultant monoclonal antibody
(receptor) was produced in one of two ways. The more
pure monoclonal antibody is produced by in vitro
culturing of the desired hybridoma in a suitable
medium for a suitable length of ~ime, followed by
recovery of the desired antibody from the
supernatant. Suitable media and length of culturing
time are well known in the art, and may be readily
determined. The in vitro technique produces
essentially monospecific monoclonal antibodies that
;23
-69-
are substantially free from other specific
antibodies. There is often a small amount of other
antibodies present since usual media contain
exogenous serum (e.gO, ~etal calf serum). However,
this ln vitro method may not produce a sufficient
quantity or concentration of antibody for some
purposes.
To produce a much greater concentration o~
slightly less pure monoclonal antibody, the desired
hybridoma may be injected into mice, preferably
syngenic or semi-syngenic mice as described
hereinbelow. The hybridoma causes formation of
antibody-producing tumors after a suitable incubation
time, which result in a relatively high concentration
of the desired antibody in the bloodstream and
peritoneal exudate (ascites) of the host mouse.
Although these host mice also have normal antibodies
in their blood and ascites, the concentration of
these normal antibodies is typically only about 5
percent of the monoclonal antibody concentration.
Ascites fluids containing the antibodies
were obtained from 10-week-old Balb/c mice (Scripps
Clinic and Research Foundation), which had been
primed with 0.3 ml of mineral oil and injected
25 intraperitoneally with 3-50x105 hybridoma cells.
Alternatively, antibodies were produced by
injecting Balb/c mice intraperitoneally with 0.3 ml
Pristane (2,6,10,14 tetramethylpentadecane) (Sigma
Chemical Co., St. Louis MO; hereinafter Sigma).
30 Seven to ten days later, 1-5x106 hybridoma cells in
log phase growth were injected intraperitoneally into
the same mice. Following a 7-14 day incubation
period, ascites fluid was removed from the mice~ The
concentration of antibody in the ascites fluid was
within the range of about 1 to about 10 mg/ml.
~47C~;23
-70-
F. A-I Vesicle Formation
Lipid-protein complexes were prepared from
cholesterol. The lipid-protein complexes were formed
into vesicles, purified by Bio Gel P 4
chromatography. ~arge particles or vesicles were
collected Eor the radioimmunoassay setforth
hereinbelow.
The reagents for this procedure were
prepared in accordance with the method set forth by
Selinger and Lapidot, J. Lipid Res., 7, 174 (1966).
Vesicle formation, the formation of a liquid protein
complex, was performed in accordance with the method
of Pownall et al., Biochem. Biophys. Acta, 713,
494-503 (1982). The lipid protein complex was
utilized in the isolation of various densities of
liquid proteins including VLDL, density of less than
1.005 g/ml; LDL, density equal to 1.019 to 1.063
g/ml; and HDL, density equal to 1.063 to 1.21 g/ml.
G. Enzymatic Cholesterol Assa~
_
The enzymatic cholesterol assay was used to
obtain a standard against which the efficacy of the
assay was tested. A free cholesterol standard was
prepared by serial dilution utilizing cholesterol
(U.S.P.) at original concentration of 1 mg/mlr and
diluted in 95 percent ethanol to give a Einal 6
standard points ranging from 1000 ng/15 mCi to 31.25
ng/mCi.
A cholesteryl oleate standard was prepared in
the same manner as the free cholesterol standard
(Gibco).
~ssay solutions were prepared, and the assay
of the plasma samples obtained was performed against
the above standards utilizing a fluorometer at 325
nanometers, in accordance with the method set forth
by Gamble et al., J. Lipid Res., _ ,1068-1070 (1978)
23
and Heider and Boyett, J. Lipid Res., 19,514 581
~1978).
H. Chromatofocusing
Chromatofocusing was performed as a technique
to separate ~DL from the admixture of lipoproteins
found in the plasma pool analyzed in accordance with
the method of this invention. Chromatofocusing was
performed in accordance with the following method.
The HDL was isolated by ultracentrifugation
and dialyzed into 25 mM piperazine hydrochloride
having a pH value of 5.8. Forty milligrams of
protein was applied to a 1.6 by 30 centimeter column
of PolybufferTM exchanger 94 (Pharmacia),
equilibrated with 25 mM piperazine HCl having a pH
value of 5.8, and the HDL was eluted with
PolybufferTM 74 (Pharmacia) diluted 1 to 15 with an
aqueous solution having a pH value of 4Ø Six HDL
subpopulations corresponding to those described by
Nestrock et al., Biochem. Biophys. Actat 753, 65-73
20 (1980) in eluting a p~ maxima value of 5.0, 4.9, 4.8,
~.7, 4.5, and 4.4, respectively were collected and
desalted by chromatography on SephadexTM G-75
~Pharmacia) that had been equilibrated with LLB.
I. Iodination of Immunoaffinity Purified
Goat Anti-Mouse Immunoglobulin
_
Iodination was performed utilizing the
EnzymobeadTM iodination procedure and
EnzymobeadsTM obtained from Biorad, (Burlingame,
CA). The EnzymobeadTM iodination was utilized to
characterize the antigens and antibodies for the
solid phase radioimmunoassay as discussed later
herein.
The solid phase radioimmunoassay was performed
utilizing a quantitative aliquot of dilute antibodies.
The antibody dilution curve was prepared by
the following method. In a series of glass
.,' L _`, .~
~47C~3
disposable tubes, the following were added in 0.100
ml aliquots: Il 5 antigen plus 9 percent BSA in
barbital buffer; competitor in barbital buffer, and
first antibody in a 1:40 diluted normal mouse serum
or optimum dilution in barbital buffer; 1:40 normal
mouse serum in barbital without antibody was added to
control tubes. The aliquots were admixed and
incubated for four hours at four degrees C.
The tubes were placed on ice and 0.100 ml. of
second antibody and barbital buffer, normal goat
serum or 100 percent TCA was added. 0.100 Ml of 100
percent trichloroacetic acid were placed in the
control tubes in lieu of normal goat serum. The
admixture was then incubated on ice for four hours
and 2.0 ml of barbital buffer were added at 4
degrees C. The admixtures were then spun for 30
minutes at 2700 r.p.m. (1500 g) at 4 degrees C. The
supernatant was aspirated and discarded and counts of
the I125 gamma emissions were measured. Values for
the ratio o~ bound antibodies to maximum available
antibody binding (B/Bo) were calculated as:
B/Bo = (X CPM) - (PPT CPM)
(TCA CPM) -(PPT CPM)
where X is the iodinated sample; PPT is the
protein precipitate; TCA is the maximum
trichloroacetic acid precipitated radioactivity; and
CPM is counts per minute.
J. Solvent Delipidization of Lipoproteins
The lipoprotein to be analyzed was dialyzed
against 0.01 percent EDTA having a pH value of 7.5
overnight (approximately 18 hours).
The resulting sample was dialyzed against
0.003 percent EDTA for approximately 12 hours, and
~Z~L7~3
-73-
was then lyophilized at 10 to 20 milligrams of
protein per tube. To each tube was added 35 ml of
1:1 absolute ethanol:anhydrous ether at 4 degrees C.
This solution was mixed.
Following mixture, the solution was
incubated for 20 minutes at -20 degrees C. The
solutions were then spun for 30 minutes at 2000
r.p.m. at 0 degrees C, and the supernatant was poured
off.
The ethanol ether extraction as described
above was performed twice for a total of three
extractions. Then 35 ml anhydrous ether at 4 degrees
C was added to the sample and incubated for 30
minutes at -20 degrees C. The sample was spun at
2000 r.p.m. for 30 minutes at -20 degrees centigrade,
and the supernatant poured off and discarded.
Pellets were dried using nitrogen gas.
K. Protein_Transfers
Proteins w~re transferred ~rom the
polyacrylamide gel (Biorad) using a transfer
cassette. Proteins were electrophoresed from the
polyacrylamide gel to nitrocellulose (Biorad) . The
process employed utilized a 2-hour electrophoreses at
400 milliamperes.
Following the transfer, active sites were
blocked utlizing a blocking buf~er solution of 24 mM
Tris, 192 rnM glycine and 20 percent methanol.
Incubation of the protein was performed in a two step
incubation at 4 degrees C; one incubation of about 6
hours and the second incubation of about 18 hours.
The proteins were then stained according to
manufacturers directions using Coomassie Brilliant
Blue-250 (Sigma), destained with 10 percent acetic
acid, and dried.
7~3
A dilution of antibody was then prepared.
Antibody was diluted in blocking buffer (as prepared
above). Gel membrane fragments, prepared by the
transfer process set forth hereinabove, were then
incubated with antibody dilutions ~or 6 hours at 4
degrees C. The incubated gel membrane fragments were
washed in a solution of 0.05 percent Tween-20
[polyoxyethylene (20) sorbitan monolaurate] (Sigma),
3.0 percent BSA (Sigma), 3 percent normal goat serum
(Sigma) in phosphate buffered-saline for about 30
minutes followed by a LiCl-SDS wash in a solution of
0.5 M LiCl, 0.1 percent SDS in water, for about 10
minutes. This was followed with another wash in the
Tween-20 wash solution (as set forth above) for
about 30 minutes and blocking with the above blocking
buffer additionally containing Tween-20 (Tween
Blocking Buffer) for about 18 hours.
Then, a 0.5 micro Ci/ml dilution of I
immuno-purified goat anti-mouse IgG was prepared in
Tween-Blocking Buffer. The membranes were incubated
in this solution at minimum volume for 4 hours at
degrees C. on a horizontal rotator as follows:
Tween-20T~ wash for about 30 minutes; LiC1-SDS wash
for 5-10 minutes at 20-22 degrees C; Tween-20
wash for 30 to 60 minutes; incubation in
Tween-Blocking Buffer at minimum volume for about 18
hours and Tween-20TM wash for about one hour.
Membranes were then air dried on absorbent paper for
at least two hours.
L Radioimmunoassays (RIA)
.
Solid-phase RIAs were performed in polyvinyl
chloride microtiter plates (Falcon, Becton-Dickenson
Rutherford, NJ) as solid supports. The plates were
coated with antigen at about 1 microgram per well in
50 microliter aqueous solutions in phosphate-buffered
saline (PBS) at pH 7.3. The plates were then
~ . , .
-75-
maintained for 3 hours at 37 degrees C. The solution
was removed, and the wells were washed 3-4 times with
PBS. Non-specific binding sites were then blocked.
The antigen-coated plates were admixed with
50 microliter dilutions of mouse serum, hybridoma
culture supernatants or ascites fluids and the
admixtures were maintained for about 16-18 hours at 4
degrees C. The solid and liquid admixtures were
separated, and the wells were rinsed. Antibody
binding was detected by a second admixture following
a maintenance period of about 4 hours at 4 degrees C
using 10 ng/well of I-goat anti-mouse Ig (4-4
micro-Ci/microgram as the indicating means.
Fluid-phase RIAs were performed in
triplicate in 12x75-mm glass tubes. To 0.1 ml of
radioiodinated antigen (human HDL, apo-A-I, or
apo-A-II) were admixed 0.1 ml of buffer or competing
antigen i present, and 0.1 ml of varying dilutions
of mouse hybridoma antibody diluted in 1:60 normal
mouse serum. All buffers also contained 5 percent
dextran (MW, 40,000). The admixtures were maintained
for a time period of 18 hours at 4 or 24 degrees C,
at the end of which time 0.1 ml of precipitating
second antibody (goat anti-mouse Ig serum) was
added. The second antibody was diluted to give a
slight antibody excess and complete precipitation of
mouse immunoglobulin. That admixture was maintained
for a time period of 4 hours, after which time, 2 ml
of cold borate buffer was added, and the tubes were
centrifuged at 2000xg for 30 minutes. Supernatants
were removed by aspiration, and the 125I content of
the pellet was determined in a gamma radiation
counter.
Maximum precipitable radioactivity was
determined by replacing the goat anti-mouse Ig serum
23~
-76-
with 100 percent trichloroacetic acid. The minimum
precipitable radioactivity or zero binding control
(B) was determined by replacing the specific
hybridoma antibody with an irrelevant hybridoma
antibody of the same heavy chain class.
The minimum precipitable radioactivity or
zero binding control (B) was determined by replacing
the specific hybridoma antibody with an irrelevant
hybridoma antibody of the same heavy chain class
Data were calculated as either total counts
bound or as percent of 125I-antigen bound =
X - B x 100,
TCA - B
where X = mean radioactivity precipitated in the
presence of a given amount of specific antibody, and
TCA is the maximum trichloroacetic acid-precipitable
radioactivity. Competitive radioimmunoassays were
analyzed by logit-transformation to compare
~ualitative and quantitative epitope expression. The
variance of the slopes of the competitive
inhibitition dose titration regression lines was
compared using the Student's t test.
M. ~
Radioiodination of HDL, apo-A-I, apo-A-II,
and immunochemically purified goat anti-mouse Ig was
performed enzymatically using immobilized
lactoperoxidase and glucose oxidase Enzymobeads,
(Biorad). For selected studies, HDL was labeled also
with the Bolton-Hunter reagent. The specific
activity of 125I in each preparation of 125I-HDL
was trichloroacetic acid-precipitable, and 5 percent
of the radioactivity was extractable into organic
solvent~ Greater than 99 percent of the
~47~23
-77-
radioactivity of 125I-apo-A~ 25I-apo-A-
ranged from 20.9 to 25.5 micro-Ci/microgram.
The preceding description of the invention
is set forth by way of example and not of
limitation. Others skilled in the art may discern
additional applications of the invention that are
fully within the scope and spirit of the invention
set forth herein.
