 |
Description  |
|
|
FIELD OF THE INVENTION
The present invention relates generally to liposome conjugates, and more
particularly to liposome-protein conjugates which have an enhanced
agglutination ability, can rapidly and sensitively agglutinate cells such
as erythrocytes, lymphocytes, and leukocytes, and which are useful in
applications such as blood typing.
BACKGROUND OF THE INVENTION
Liposomes are now well recognized as useful for delivery of therapeutic
agents, such as cytotoxic drugs or other macromolecules capable of
modifying cell behaviour, to in vivo sites. For example, U.S. Pat. No.
3,993,754, inventors Rahman, et al, issued Nov. 23, 1976, discloses an
improved method for chemotherapy of malignant tumors in which an antitumor
drug is encapsulated within liposomes and the liposomes containing the
encapsulated drug are injected into an animal or man.
It has been suggested that target, or in vivo site, specificity might be
conferred on liposomes by their association with specific antibodies or
lectins. Methods of associating antibodies with liposomes have been
described, and may be generally divided into two groups-nonspecific
association and covalent attachment.
Nonspecific association appears to rely upon the affinity of the Fc portion
of the antibody for the hydrophobic region of the lipid bilayer. However,
nonspecific association appears incapable of associating more than about
15-30 microgram per micromole of lipid. Also, nonspecific association has
little practical value because the liposomes are rendered more permeable
than their encapsulated contents and protein aggregation is produced
during formation of the nonspecifically associated liposome-protein.
Prior to preparation of the covalently attached protein of coupled-protein
species described in U.S. patent application Ser. No. 129,654, attempts to
covalently attach protein to liposomes had been unsatisfactory. For
example, some of the prior attempts had involved modifications of the
proteins which tended to denature the protein, and thus a substantial loss
of biological activity had ensued. Other attempts to covalently attach
protein to liposomes had produced very small amounts of specific
attachment.
By contrast, activated liposomes in accordance with U.S. patent application
Ser. No. 129,654 are readily and efficiently covalently bound to a variety
of biologically active proteins with at least about 40 microgram of
protein per micromole of lipid. For example, use of the activated
liposomes has achieved coupling of up to about 200 microgram of
F(ab').sub.2 per micromole of lipid; further, such coupled
liposome-protein species were shown to have an improved hemagglutinating
titre by comparison to the original, soluble antibody from which they were
derived.
Very recently, another efficient method for coupling protein to liposomes
has reported coupling of up to 600 microgram of Fab' per micromole of
phospholipid via a disulfide interchange reaction. Martin, et al.,
Biochemistry, 20, pages 4429-4238 (July, 1981).
Meanwhile, agglutination methods are known and useful for applications such
as blood typing. However, many such methods have had to be performed
indirectly, or have been of relatively low sensitivity. For example, the
Coombs test is an indirect agglutination method in the sense that a
secondary, or intermediate, antibody must be used. Further, detection of
antibodies which do not produce positive agglutination (e.g. visible
agglutination) when combined with their specific antigens has presented
difficulties in applications such as blood crossmatching. Such
serologically "incomplete" antibodies are believed to be fully functional
bivalent IgG molecules, but they are unable to bridge two cells and hence
do not produce positive agglutination.
SUMMARY OF THE INVENTION
Accordingly, it is an object of the present invention to provide a
hemagglutination reagent having an improved capacity to agglutinate
erythrocytes.
It is a further object of the present invention to provide an agglutination
method, useful for assaying erythrocyte antigens, which is rapid and
sensitive.
It is yet a further object that human erythrocyte surface antigens, which
are not normally capable of producing visible hemagglutination, may be
agglutinated and the clumps subject to detection.
In one aspect of the present invention, an agglutination method useful for
assaying cell surface antigens comprises providing a quantity of
liposome-protein conjugates where the protein thereof supplies an antigen
binding capacity for at least a majority of the liposome-protein
conjugates, contacting the liposome-protein conjugates with substantially
unagglutinated cells to form a mixture, and examining the mixture for cell
agglutination. The liposome-protein conjugates have at least about 40
micrograms of protein per micromole of lipid.
In another aspect of the present invention, a reagent, useful for
hemagglutination, comprises liposome-protein conjugates having an antibody
covalently bound to the liposomes and an antigen binding capacity for
erythrocytes bearing a surface antigen for which the antibody is specific.
The reagent has a hemagglutinating activity which is improved with respect
to the hemagglutination activity of the original antibody from which the
liposome-protein conjugates are derived.
The liposome-protein conjugates and method of the present invention provide
considerably improved sensitivity for agglutination assays. For example,
agglutination of erythrocytes by use of the liposome-protein conjugates
can occur in seconds with large, clearly visible agglutinated clumps of
cells. By contrast, agglutination assays using the original, soluble
antibody typically require minutes to produce agglutination visible to the
naked eye, and the clumps of cells are much smaller. Thus, the ability of
liposome-protein conjugates in accordance with the present invention to
produce larger, more visible clots suggests the possibility of simple,
visual spot tests which need not require special optical equipment for
observation, and for use in a variety of diagnostic applications.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
Broadly, one aspect of the present invention is a diagnostic method wherein
liposome-ligand conjugates are contacted with ligand-binding molecules to
form a mixture, and the mixture is examined for combinations, or
interactions, between the liposome-ligand conjugates and the
ligand-binding molecules. The ligand-binding molecules are carried by at
least one surface, more preferably carried by a plurality of surfaces
defined by discrete particles, and most preferably carried by cell
surfaces. The ligand of the liposome-ligand species is in an amount of at
least about 40 micrograms per micromole of lipid, and preferably is a
protein having an antigen binding capacity. For example, when the protein
is an antibody and the ligand-binding molecules are antigens carried by
cell surfaces, then the mixture examination following the contacting step
typically includes determining cell agglutination mediated by combinations
between the coupled antibody and immunological partners, or specific
antigens, carried on the cells. Where the ligand is protein, the
conjugates will hereinafter sometimes be referred to as liposome-protein
conjugates.
The at least one surface carrying the ligand-binding molecules may be a
natural multivalent antigen or may also be formed by inanimate, synthetic
materials. For example, among suitable non-cellular materials are
polyacrylamide beads about 5-10 microns in diameter having immobilized, or
covalently bound, immunoglobulin on the bead surfaces, which are
commercially available from Bio-Rad Laboratories under the trademark
"Immunobead", and polystyrene spheres about 1 micron in diameter, which
may be coupled along their surfaces with immunologlobulin such as IgG,
commercially available from Covalent Technology Corp. under the trademark
"Covaspheres".
As used herein, ligand and ligand-binding molecules mean moieties which can
interact specifically but non-covalently with each other. One type of such
moiety pairings in an antigen-antibody interaction, another is a
hormone-receptor interaction, and yet another is a carbohydrate-lectin
interaction.
Liposome-protein conjugates useful in accordance with the present invention
may be prepared in various ways. For example, one suitable preparation is
via activated liposome precursors where the precursor liposomes, before
being covalently bound, are activated by means of an oxidizing reagent. A
modification of this preparation is wherein the lipid, such as
gangliosides, is first oxidized and then formed into the precursor
liposomes. Another suitable preparation is via disulfide bonds, as
described by Martin et al, supra.
In any event, it is necessary that the protein be then covalently bound to
the liposomes precursors in an amount of at least about 40 micrograms per
micromole of lipid. A suitable preparation and properties of
lipsome-protein conjugates will now be more fully described.
PREPARATION OF LIPOSOME-PROTEIN CONJUGATES VIA ACTIVATED LIPOSOME
PRECURSORS
Activated liposomes may be prepared from starting vesicles which are
generally characterized either as unilamellar vesicles or multilamellar
vesicles. Either liposomal structure is suitable. A particularly preferred
preparation is by the reverse-phase evaporation vesicle (REV) method,
which is disclosed by U.S. Pat. No. 4,235,871, and as is described in
Proc. Natl. Acad. Sci. U.S.A., Volume 75, No. 9, pp. 4194-4198 (1978),
entitled "Procedure For Preparation of Liposomes With Large Internal
Aqueous Space And High Capture By Reverse-Phase Evaporation", Szoka, Jr.
and Papahadjopoulos, which disclosures are incorporated herein by
reference.
As is known to the art, a wide variety of materials may be encapsulated, if
desired, by the precursor liposomes. For example, the precursor liposomes
can encapsulate cytotoxic drugs, can encapsulate nucleic acids, and can
encapsulate various proteins.
In any event, the precursor liposomes suitable for the present invention
may be formed from either phosphatidylglycerol (hereinafter also referred
to as "PG"), which has an oxidisable group at the polar head region, as
the sole lipid, or may be formed from a mixture of two or more different
lipids.
When formed from two or more different lipids, at least one of the lipids
contains oxidisable groups, such as vicinal amino or vicinal hydroxyl
groups, along the polar head region of the lipid molecule. For example, in
the instance of vicinal amino groups, a glycolipid having galactosamine or
glucosamine residue is a suitable oxidisable lipid. More usually, at least
one of the lipids will have vicinal hydroxyl groups at the polar head
region. Particularly preferred as one of the lipids (that is, the
oxidisable lipid) in a lipid mixture are the glycolipids such as lactosyl
ceramide, galactocerebroside, gangliosides, and trihexosyl ceramide, and
the phospholipids, such as phosphatidylclycerol and phosphatidylinositol.
The amount of such lipids having oxidisable groups (generally herein
referred to as "oxidisable lipids") may vary with respect to the total
lipids forming the precursor liposomes; however, it is preferred that the
mole percent of oxidisable lipids be in an amount of at least about 10
mole percent with respect to a total of the mixture of lipids.
Particularly preferred amounts of oxidisable lipids with respect to the
total lipids are illustrated by Table I, below.
TABLE I
______________________________________
Mole of Oxidisable Lipid
Oxidisable Lipid To Total Lipid Mixture
______________________________________
Lactosylceramide About 10
Trihexosylceramide
About 10
Galactocerebroside
About 20
Phosphatidylglycerol
About 33-40
Phosphatidylinositol
About 20
Gangliosides About 10
______________________________________
The structures of the preferred oxidisable lipids are well known; however,
for clarity FIG. 1, below, illustrates PG as representative of the general
structures of the oxidisable lipids having the polar head regions and the
region of non-polar tails.
##STR1##
FIG. 1 is generally representative of all of the lipids which may be mixed
to form the precursor liposomes in defining the polar head region and the
non-polar tails. The FIG. 1 structure is more particularly generally
representative of the oxidisable lipids which have vicinal hydroxyl groups
at the polar head region thereof.
When a mixture of lipids, including the oxidisable lipid, is utilized to
form the precursor liposomes, then the remaining lipid or lipids may
include any of the amphiphilic substances known to produce liposomes. A
particularly preferred lipid for combination with the oxidisable lipids is
phosphatidylcholine (hereinafter also referred to as "PC"), sphingomyelin
or mixtures thereof.
As is known, the above discussed mixtures of lipid molecules form precursor
liposomes with the lipid molecules being arranged in either one
bimolecular layer (unilamellar) or a plurality of bimolecular layers
(multilamellar). In any event, the most outward bimolecular layer forms an
outer surface for the liposome. In an aqueous solution, the polar head
regions of the lipid molecules are exposed, or extend into, the aqueous
system in a generally radially outward orientation with respect to the
outer surface. The non-polar tails extend radially inwardly with respect
to the outer surface and form a substantially continuous hydrocarbon phase
of the bimolecular layer. This substantially continuous hydrocarbon phase
is relatively impermeable, and acts to encapsulate the materials inside
the precursor liposomes.
Nevertheless, some mixtures of lipids forming the precursor liposomes may
tend to be permeable to small molecules, and cholesterol is a desirable
addition to some of these lipid mixture for reducing the permeability of
the precursor liposomes. The cholesterol tends to orientate within the
bimolecular layer. Other components may be utilized in place of
cholesterol to reduce the liposome permeability. For example, a
phosphatidylcholine having the fatty acid saturated aliphatic chain, or
non-polar tails, of a length of 18 (rather than the usual unsaturated 16
to 18 carbon chain obtainable from egg yolks) may be utilized. However,
when sphingomyelin is mixed with the oxidisable lipid, the precursor
liposomes thereof are inherently quite impermeable to small molecules.
A solution of precursor liposomes may thus be provided as has been
described above. This solution is preferably a polar solution, such as an
aqueous solution, but may also be a non-polar solution. The precursor
liposomes are contacted with a sufficient amount of a relatively mild
oxidizing reagent to produce activated liposomes. Where the lipids to be
used for liposomes are in a non-polar solvent, the oxidizing reagent may
be lead tetraacetate. In the preferred polar solution, the oxidizing
reagent of the contacting step is a periodate reagent, usually sodium
periodate, which cleaves the vicinal amino or hydroxyl groups at the polar
head regions of the oxidisable lipids.
Where the solution is polar and the oxidizing agent is a periodate reagent,
the pH and osmolarity of the liposome solution and an added amount of
periodate reagent should be substantially the same. The pH is typically
about 6.0 to about 8.5. The oxidizing reagent produces activated liposomes
by oxidizing the oxidisable groups, such as the vicinal hydroxyl or amino
groups of the oxidisable lipid, to yield aldehyde moieties at the polar
head regions of the oxidisable lipids. A sufficient quantity of periodate
reagent will usually be a molar ratio with respect to the total of lipid
molecules of from about 1.5:1 to about 6:1. The oxidation reaction of the
contacting step is typically left to proceed for about one-half hour at
room temperature, although the reaction may be permitted to proceed for up
to about one hour on ice. The periodate reagent is then preferably removed
by gel filtration through a column of dextran polymeric beads having an
exclusion limit of about 75,000 daltons.
Reaction Schemes I, II and III diagrammatically illustrate the activation
of precursor liposomes, with the oxidisable lipids being
phosphatidylglycerol, phosphatidylinositol and lactosylceramide
respectively.
##STR2##
Substantially all of the material which is interior the outer surfaces of
the activated liposomes remains encapsulated during the above-described
oxidation with periodate reagent. As illustrated by Reaction Schemes
I-III, the aldehyde moieties which are formed by the oxidation, or
modification, of the oxidisable lipids at the polar head regions thereof
define covalent binding sites for the protein to be bound, or coupled.
A wide variety of proteins may be attached, or coupled, to the activated
liposomes. The mechanism of coupling is believed to occur between the
primary or secondary amino group along the protein and the aldehyde moiety
of the activated liposomes so as to form a Schiff-base, for example, with
the primary amino group of a lysyl moiety. Such a mechanism is
diagrammatically represented by Reaction Scheme IV, which for simplicity
illustrates only the terminal galactose (after modification) of
lactosylceramide.
##STR3##
The coupling is driven to completion by a mild reducing agent, preferably
sodium cyanoborohydride, so that a stable, covalent bond is formed between
the protein and the activated liposome. For example, addition of a
sufficient amount of sodium cyanoborohydride drives the Schiff-base of
reaction Scheme IV, above, to completion, as is generally illustrated by
Reaction Scheme V, below.
##STR4##
Reaction Schemes IV and V, above, diagrammatically illustrate coupling of
a protein with lactosylceramide, where the lactosylceramide has been
modified by oxidation to include aldehyde moieties. Use of the other
oxidisable lipids proceeds by an analogous manner. In the instance of
modified lactosylceramide, the secondary amine moiety which is covalently
binding the protein to the activated liposome may further proceed, in the
presence of sodium cyanoborohydride, into an even more stable, tertiary
amine form.
Although sodium cyanoborohydride is the preferred reducing agent for
coupling proteins to the activated liposomes, other reducing agents may be
utilized, depending upon the particular circumstances. For example,
borohydride may be utilized; however, the coupling reaction would usually
then be conducted at a relatively alkaline pH, which may tend to denature
the protein being coupled.
Suitable proteins for adequate coupling will have at least one primary or
secondary amino group, and preferably a plurality of primary or secondary
groups. Proteins having at least about 20 lysyl moieties per molecule are
more preferred. IgG, with about 60 lysyl moieties, has been found to be
particularly well coupled; another preferred antibody for coupling with
the activated liposome is F[ab'].sub.2.
Four aqueous solutions of liposome, each containing from about 10 to about
40 micromoles of lipid per mililiter, were activated as previously
described. The precursor liposomes had been prepared by the REV procedure
and had been extruded through a polycarbonate filter to yield liposomes
having a diameter of about 0.2 micron. The solutions were buffered at a pH
of from about 6.0 to about 8.5. A fifth liposome solution, wherein the
oxidisable lipid was not oxidized, was prepared as a control. These four
activated liposome solutions in accordance with the present invention and
the fifth control solution were as illustrated by Table II.
TABLE II
______________________________________
Solution
Lipid Composition Total Lipid
Vol.
No. Molar Ratio (micromole)
(ml.)
______________________________________
1 PC/Lactosylceramide, 10:1
9.21 2.8
2 PC/Trihexosylceramide, 10:1
16.44 4.5
3 PC/PG, 1:1 9.5 3.1
4 Galactocerebroside/
15 0.6
PC/Cholesterol, 2:4:5
5* PC/Lactosylceramide, 10:1
7.11 2.8
______________________________________
*Control solution, liposomes not activated
The five solutions as in Table II were treated as follows. 5 to 10
millligrams of IgG in the same buffer as the liposome solutions were added
to the respective liposome solutions (the activated liposomes were
suspended in the solutions with substantially no clumping). Sufficient
sodium cyanoborohydride was added to give a concentration of about 20
millimolar, and the solutions were left for about 2 to about 3 hours at
room temperature. The liposomes having covalently bound IgG thereon were
then purified by conventional methods, such as column gel filtration or
centrifugation. The amount of coupling is illustrated by Table III, below
(the number of molecules per vesicle was estimated on the assumption that
the vesicles were 0.2 micron in diameter, with about 1.8.times.10.sup.12
vesicles/micromole lipid).
TABLE III
______________________________________
Liposome
IgG (mg) Protein: Lipid Ratio
Molecules
Solution
added (.mu.g/mole)
(.mu.g/mg lipid)
IgG/vesicle
______________________________________
1 20 112 147 251
2 14 57 75 128
3 14 47 62 106
4 10 96 126 216
5* 20 11 14 25
______________________________________
*Control solution, liposomes not activated.
BINDING OF PROTEIN TO LIPOSOMES
IgG coupling to activated liposomes has typically resulted in the binding
of from about 50 to about 200 micrograms of IgG per micromole of lipid.
Substantially no coupling is observed in control liposomes. Nonspecific
binding of proteins to activated liposomes was below the limits of the
protein assay utilized in determining coupling.
The proteins which may be covalently bound, or coupled, to the activated
liposomes retain a significant amount of biological activity. This is
illustrated by use of immunopurified rabbit antifluorescein antibody, as
follows.
Antifluorescein IgG binds specifically to fluorescein isothiocyanate and
carboxyfluorescein. Upon binding to the antibody, the fluorescence of the
fluorescein is abolished, and this was used to measure the binding
activity of the antibody. Successive additions of antibody to a solution
of carboxyfluorescein reduced the fluorescence due to quenching of the
fluorophore upon binding to the antibody. The antigen binding capacity of
liposome-bound antibody was compared by correlating the percentage
reduction in fluorescence for a variety of specified protein
concentrations in linear ranges where quenching was proportional to the
protein concentration as illustrated by Table IV, below (wherein the
original antibody, or control, linear range was from about 78/1 to about
30/4; unbound antibody was from about 85/1 to about 40/5; and coupled
antibody-activated liposomes was linear over the entire range
illustrated).
TABLE IV
______________________________________
Fluorescent Intensity/Antibody conc. .times. 10.sup.8 (.mu.mole/ml)
Original Antibody Coupled Antibody-
(Control) Unbound Antibody
Activated Liposome
______________________________________
78/1 85/1 95/1
60/2 75/2 90/2
40/3 62/3 85/3
30/4 50/4 78/4
18/5 40/5 70/5
12/6 35/6 65/6
8/7 30/7 60/7
8/8 20/8 58/8
15/9 50/9
______________________________________
As illustrated by Table IV, above, the fluorescent quenching of
carboxyfluorescein by the original antibody preparation (control) and the
antibody that was recovered from the coupling process may be compared to
antibody bound to the activated liposomes. If the activity of the
original, control preparation is set at 100%, then the activity of the
activated liposome bound antibody is about 33%, and of the recovered
antibody is about 70%. Antigen binding capacity is, therefore, only
partially inhibited by the inventive coupling process, and the coupled
protein displays, or retains, a significant amount of antigen binding
capacity.
HEMAGGLUTINATION
Liposome-protein conjugates in accordance with the present invention have
the ability to agglutinate erythrocytes. Activated liposomes, prepared as
described above, were conjugated with antigen-antibody, and incubated with
erythrocytes conjugated with fluorescein-isothiocyanate. This resulted in
the agglutination of the erythrocytes and the hemagglutinating titre
(expressed as the minimum concentration observed to cause agglutination)
was improved by a factor of about 1.5 with respect to the original,
soluble antibody from which the liposome-protein conjugates were derived.
Lower values of the hemagglutinating titre indicate more effective
agglutination capacity. This is illustrated by Table V, below.
TABLE V
______________________________________
Hemagglutination of FITC-Human Erythrocytes
by Rabbit Antifluorescein IgG
Preparation Titre (microgram/ml)
______________________________________
Liposomes Bound Antibody
1.22
(500 molecules/vesicle)
Untreated Antibody
1.92
______________________________________
The improved ability of liposome-protein conjugates in accordance with the
present invention to agglutinate erythrocytes is also demonstrated in
another preparation of liposome-protein conjugates via the activated
liposome precursors, as follows.
10 mg of F[ab'].sub.2 was added to 10 .mu.mole of vesicle lipid (after
oxidation and desalting) in 1 ml of borateasline (pH 8.4), and then added
10 .mu.l of 1M sodium cyanoborohydride. After 18 hours at room
temperature, the vesicles were separated from unbound protein by flotation
on discontinuous dextran gradients (0 to 20 percent, weight to volume).
For quantitation of binding to cells the vesicles contained .sup.3
H-labeled dipalmitoylphosphatidylcholine (DPPC; 10 .mu.Ci/.mu.mole) and
.sup.14 C-labeled sucrose (1 .mu.Ci/.mu.mole). "Targeted" vesicles were
those having coupled rabbit F[ab'].sub.2 to human erythrocytes, whereas
"control" vesicles were coupled to F[ab'].sub.2 prepared from rabbit gamma
globulin by pepsin digestion and absorption to a Staphylococcus aureus
suspension.
The targeted vesicles and the control vesicles were then separately
incubated with 10.sup.6 to 10.sup.8 human erythrocytes in 0.2 ml of
phosphate-buffered saline (PBS) at pH 7.4 for 1 hour at 37.degree. C.
Cells were washed to remove unbound vesicles and were either taken up
directly in 10 ml of Triton-toluene scintillant and counted for [.sup.3
H]DPPC (10.sup.6 to 10.sup.7 cells) or extracted after washing 10.sup.8
cells; the chloroform phase was evaporated and counted for [.sup.3 H]DPPC
content, and the aqueous phase was incubated overnight at 60.degree. C. to
remove methanol and counted for [.sup.14 C]sucrose content.
When vesicles (with from about 1 to 500 nmoles of lipid) were incubated
with 10.sup.8 cells there was a marked difference between targeted and
control binding, with 80% of the vesicles binding at about 20 to 500
.mu.mole of lipid. Control binding with vesicles conjugated to nonspecific
F[ab'].sub.2 was very low (<1 percent) and did not appreciably increase
between 100 and 500 nmole of lipid. Both the vesicle lipids and the
encapsulated sucrose bound to the cells in nearly identical proportions,
indicating that cell binding caused no loss of vesicle contents and that
the antibody-conjugated liposome preparation was reasonably homogeneous
with respect to lipid, encapsulated aqueous marker, and antibody. When
vesicles were incubated with 10.sup.7 erythrocytes there was a marked
difference in binding between the targeted and non-targeted samples.
Although a small fraction of the total available vesicles became bound,
the number of vesciles that bound to each cell was increased. Vesicles
incubated with 10.sup.6 cells also exhibited binding specificity (not
shown), with 6 nmole of specific antibody-bearing vesicles and 2 nmole of
nonspecific vesicles being bound when 100 nmole of lipid was incubated
with the cells. The addition of serum (25 percent fetal calf) during
incubation had no substantial effect on the binding.
The vesicle preparations contained 143 molecules per vesicle, assuming
F[ab'].sub.2 has a molecular weight of 90,000 and that the vesicle
preparations contained 1.8.times.10.sup.12 vesicles per micromole (for
unilamellar vesicles of 0.2 .mu.m diameter). Antibodies not purified
immunologically, such as those used here, may contain only 1 to 5 percent
of molecules that are specifically reactive to the cell antigens. The
preparation therefore probably contained approximately one to five
specific molecules per vesicle, so that most vesicles were specific for
the target cells. The use of nonimmunopurified antibodies with coupling
method that bind only a few antibody molecules per vesicle would result in
many vesicles having no specificity for the target.
The association of 400 nmole of lipid with 10.sup.8 human erythrocytes
constitutes a lipid mass three times greater than the lipid content of the
cell membranes. If one assumes that the vesicles are 0.2 .mu.m in diameter
and are unilamellar, the number of vesicles bound per cell is 8000 and
their encapsulated volume is 0.33 of the cell volume. Thus, about 80% of
the targeted vesicles associated with the human erythrocytes.
The hemagglutinating titre of the F[ab'].sub.2 bound to vesicles was
measured and compared to the original antibody preparation. The
nonspecific soluble F[ab'].sub.2 and the control vesicles derived from it
produced no hemagglutination at concentrations up to 1 mg of F[ab'].sub.2
per milliliter. The soluble antibody to human erythrocyte F[ab'].sub.2 had
a hemagglutination titre of 4 .mu.g/ml, and the titre of the
liposome-protein conjugates derived from the original, soluble antibody
was 1.5 .mu.g/ml, for an improvement factor of about 2.7.
The extent of improvement, or enhancement, of hemagglutinating activity for
suitable liposome-protein conjugates is greater than appears from the
data, for example in Table V, since during conjugation some of the bound
antibody is partially inactivated. In a further experiment, performed with
antifluorecein conjugated liposomes prepared via the activated method, a
variety of antibody preparations were obtained by mixing the
immunopurified antifluorescein with normal rabbit IgG to vary the extent
of antifluorescein substitution. The activity of the antifluorescein was
calculated by fluorescence quenching, as previously described and
illustrated by Table IV, and this value was used to calculate the number
of active antibody molecules per liposome of the liposome-antibody
conjugates and the corrected minimum hemagglutinating concentration. The
liposome-protein conjugates were contacted with unagglutinated
erythrocytes which had been coated with fluorescein. (The uncorrected
minimum hemagglutinating concentration (MHC) is calculated from the total
protein concentration.) The data is illustrated by Table VI, below.
TABLE VI
______________________________________
Active antibody
MHC corrected MHC
Improvement
molecules/liposome
(.mu.g/ml)
(.mu.g/ml) factor
______________________________________
67 5.10 1.63 2.3
100 1.60 0.70 5.6
186 0.57 0.31 12.6
soluble antibody
3.90 3.90 --
______________________________________
The improvement factor, illustrated by Table VI, above, compares the
corrected MHC of lipsome-protein conjugates to the MHC of the original,
soluble antibody from which the liposome-protein conjugates were derived.
As may be seen, the improvement factor varied from about 2.3 to about
12.6, depending upon the number of active antibody molecules which were
covalently bound per liposome.
In another preparation of liposome-protein conjugates the liposome-protein
conjugates had 50 .mu.g of antihuman erythrocyte Fab' fragments per
.mu.mole liposomal phospholipid (about 500 antihuman Fab' fragments per
liposome). The precursor liposomes were formed from PC:cholesterol:PDP-PE
and conjugated to Fab' by the procedure of Martin and Papahadjopoulos, J.
Biol. Chem., (1981-In Press). The minimum hemagglutinating concentration
(MHC) for soluble antibody was 5.2 .mu.g/ml, whereas the MHC for
liposome-protein conjugates was 0.17 .mu.g/ml. That is, the agglutination
improvement factor was about 30.
BINDING INHIBITION
Another batch of liposome-protein conjugates via activated liposome
precursors was prepared and tested with soluble antibody for binding
inhibition as follows.
The precursor liposomes were prepared from a mixture of
phosphatidylcholine:cholesterol:oxidized ganglioside (5:5:1) which
contained trace amount of .sup.3 H dipalmiolylphosphatidylcholine to give
2000 counts per minute (cpm) per nanomole lipid. These vesicles were then
conjugated to monoclonal mouse anti H2K.sup.k antibody by reductive
amination with sodium cyanoborohydride. The resultant liposome-protein
conjugates had an antibody:lipid ratio of 60 .mu.mole.
Meanwhile, 5.times.10.sup.6 L929 fibroblasts in confluent monolayers in 6
cm petri dishes were incubated for 30 minutes at 37.degree. C. with 0.2 ml
phosphate buffered solution (PBS) containing 50% serum and 20 nmole lipid
to which was conjugated 1.2 .mu.g antibody. The incubation mixture also
contained variable amounts of soluble antibody, as indicated in Table VII,
below. After incubation, the cells were washed four times with phosphate
buffered saline, trypsinized to remove them from the monolayer and taken
up in scintillant for counting.
In a similar manner, 2.times.10.sup.6 R1.1 T-lymphoma cells were suspended
in 0.2 ml PBS containing 50% serum, 20 nmole lipid conjugated to 1 .mu.g
antibody and various amounts of soluble antibody as indicated in Table
VII, below. After 30 minutes they were washed four times by centrifugation
and resuspension of the cells in 5 ml portions of PBS. The cells were
finally resuspended in 0.5 ml and taken up in scintillant.
The antibody-anti H2K.sup.k reaction which occurred between the
liposome-conjugates and the cells (and, to a lesser extent between the
soluble antibody and the cells) illustrates a reaction with the H2K of
certain mouse strains. This protein is a membrane antigen present at high
levels in most mouse tissues. The L929 fibroblast and the R1.1 T-lymphoma
are cultured cell lines derived from mice which express the H2K.sup.k
antigen.
TABLE VII
______________________________________
INHIBITION OF TARGETED LIPOSOME
BINDING BY SOLUBLE ANTIBODY
Soluble Antibody
Percent Control Binding
Per Sample (.mu.g)
L929 R1.1
______________________________________
0.1 100% 100%
0.3 86% 97%
1 97% 81%
3 100% 62%
10 95% 44%
30 66% 31%
______________________________________
For example, as illustrated by Table VII above, a ratio of soluble to
liposome-bound antibody of at least 10:1 was required to achieve about 50%
binding inhibition with the R1.1 T-lymphoma cells. These data demonstrate
that liposome-protein conjugates in accordance with the present invention
bind with greater functional affinity to th | | |
