This invention relates to methods for obtaining antibodies and assays
for using such antibodies. More specifically, the invention relates to methods of
obtaining antibodies which specifically bind to naturally occurring forms of PrPSc.
Background of the Invention
Prions are infectious pathogens that cause central nervous system
spongiform encephalopathies in humans and animals. Prions are distinct from bacteria,
viruses and viroids. The predominant hypothesis at present is that no nucleic acid
component is necessary for infectivity of prion protein. Further, a prion which
infects one species of animal (e.g., a human) will not infect another (e.g., a mouse).
A major step in the study of prions and the diseases that they cause
was the discovery and purification of a protein designated prion protein ("PrP")
[Bolton et al.,Science218:1309-11 (1982); Prusiner, et al.,Biochemistry21:6942-50 (1982); McKinley, et al., Cell35:57-62 (1983)].
Complete prion protein-encoding genes have since been cloned, sequenced and expressed
in transgenic animals. PrPC is encoded by a single-copy host gene [Basler,
et al., Cell46:417-28 (1986)] and is normally found at the outer
surface of neurons. Prion diseases are accompanied by the conversion of PrPC
into a modified form called PrPSc. However, the actual biological or
physiological function of PrPC is not known.
The scrapie isoform of the prion protein (PrPSc) is necessary
for both the transmission and pathogenesis of the transmissible neurodegenerative
diseases of animals and humans. See Prusiner, S.B., "Molecular biology of prion
disease," Science252:1515-1522 (1991). The most common prion diseases
of animals are scrapie of sheep and goats and bovine spongiform encephalopathy (BSE)
of cattle [Wilesmith, J. and Wells, Microbiol. Immunol.172:21-38 (1991)].
Four prion diseases of humans have been identified: (1) kuru, (2) Creutzfeldt-Jakob
Disease (CJD), (3) Gerstmann-Strassler-Scheinker Disease (GSS), and (4) fatal familial
insomnia (FFI) [Gajdusek, D.C.,Science197:943-960 (1977); Medori
et al., N. Engl. J. Med.326:444-449 (1992)]. The presentation of
human prion diseases as sporadic, genetic and infectious illnesses initially posed
a conundrum which has been explained by the cellular genetic origin of PrP.
Most CJD cases are sporadic, but about 10-15% are inherited as autosomal
dominant disorders that are caused by mutations in the human PrP gene [Hsiao et
al.,Neurology40:1820-1827 (1990); Goldfarb et al., Science258:806-808
(1992); Kitamoto et al., Proc. R. Soc. Lond. (In press) (1994)]. Iatrogenic
CJD has been caused by human growth hormone derived from cadaveric pituitaries as
well as dura mater grafts [Brown et al., Lancet340:24-27 (1992)].
Despite numerous attempts to link CJD to an infectious source such as the consumption
of scrapie infected sheep meat, none has been identified to date [Harries-Jones
et al., J. Neurol. Neurosurg. Psychiatry51:1113-1119 (1988)] except
in cases of iatrogenically induced disease. On the other hand, kuru, which for many
decades devastated the Fore and neighboring tribes of the New Guinea highlands,
is believed to have been spread by infection during ritualistic cannibalism [Alpers,
M.P., Slow Transmissible Diseases of the Nervous System, Vol. 1, S.B. Prusiner
and W.J. Hadlow, eds. (New York: Academic Press), pp. 66-90 (1979)].
The initial transmission of CJD to experimental primates has a rich
history beginning with William Hadlow's recognition of the similarity between kuru
and scrapie. In 1959, Hadlow suggested that extracts prepared from patients dying
of kuru be inoculated into non-human primates and that the animals be observed for
disease that was predicted to occur after a prolonged incubation period [Hadlow,
W.J., Lancet2:289-290 (1959)]. Seven years later, Gajdusek, Gibbs
and Alpers demonstrated the transmissibility of kuru to chimpanzees after incubation
periods ranging form 18 to 21 months [Gajdusek et al., Nature209:794-796
(1966)]. The similarity of the neuropathology of kuru with that of CJD [Klatzo et
al., Lab Invest.8:799-847 (1959)] prompted similar experiments with
chimpanzees and transmissions of disease were reported in 1968 [Gibbs, Jr. et al.,
Science161:388-389 (1968)]. Over the last 25 years, about 300 cases
of CJD, kuru and GSS have been transmitted to a variety of apes and monkeys.
The expense, scarcity and often perceived inhumanity of such experiments
have restricted this work and thus limited the accumulation of knowledge. While
the most reliable transmission data has been said to emanate from studies using
non-human primates, some cases of human prion disease have been transmitted to rodents
but apparently with less regularity [Gibbs, Jr. et al., Slow Transmissible Diseases
of the Nervous System, Vol. 2, S.B. Prusiner and W.J. Hadlow, eds. (New York:
Academic Press), pp. 87-110 (1979); Tateishi, et al., PrionDiseases of
Humans and Animals, Prusiner, et al., eds. (London: Ellis Horwood), pp. 129-134
The infrequent transmission of human prion disease to rodents has
been cited as an example of the "species barrier" first described by Pattison in
his studies of passaging the scrapie agent between sheep and rodents [Pattison,
I.H., NINDB Monograph 2, D.C. Gajdusek, C.J. Gibbs Jr. and M.P. Alpers, eds.
(Washington, D.C.: U.S. Government Printing), pp. 249-257 (1965)]. In those investigations,
the initial passage of prions from one species to another was associated with a
prolonged incubation time with only a few animals developing illness. Subsequent
passage in the same species was characterized by all the animals becoming ill after
greatly shortened incubation times.
The molecular basis for the species barrier between Syrian hamster
(SHa) and mouse was shown to reside in the sequence of the PrP gene using transgenic
(Tg) mice [Scott, et al., Cell59:847-857 (1989)]. SHaPrP differs
from MoPrP at 16 positions out of 254 amino acid residues [Basler, et al.,
Cell46:417-428 (1986); Locht, et al.,Proc. Natl. Acad. Sci. USA83:6372-6376 (1986)]. Tg(SHaPrP) mice expressing SHaPrP had abbreviated incubation
times when inoculated with SHa prions. When similar studies were performed with
mice expressing the human, or ovine PrP transgenes, the species barrier was not
abrogated, i.e., the percentage of animals which became infected were unacceptably
low and the incubation times were unacceptably long. Thus, it has not been possible,
for example in the case of human prions, to use transgenic animals (such as mice
containing a PrP gene of another species) to reliably test a sample to determine
if that sample is infected with prions. The seriousness of the health risk resulting
from the lack of such a test is exemplified below.
More than 45 young adults previously treated with HGH derived from
human pituitaries have developed CJD [Koch, et al., N. Engl. J. Med.313:731-733 (1985); Brown, et al., Lancet340:24-27 (1992);
Fradkin, et al.,JAMA265:880-884 (1991); Buchanan, et al.,
Br. Med. J.302:824-828 (1991)]. Fortunately, recombinant HGH is now
used, although the seemingly remote possibility has been raised that increased expression
of wtPrPC stimulated by high HGH might induce prion disease [Lasmezas,
et al.,Biochem. Biophys. Res. Commun.196:1163-1169 (1993)]. That
the HGH prepared from pituitaries was contaminated with prions is supported by the
transmission of prion disease to a monkey 66 months after inoculation with a suspect
lot of HGH [Gibbs, Jr., et al., N. Engl. J. Med.328:358-359 (1993)].
The long incubation times associated with prion diseases will not reveal the full
extent of iatrogenic CJD for decades in thousands of people treated with HGH worldwide.
Iatrogenic CJD also appears to have developed in four infertile women treated with
contaminated human pituitary-derived gonadotrophin hormone [Healy, et al.,
Br. J. Med.307:517-518 (1993); Cochius, et al., Aust. N.Z. J.
Med.20:592-593 (1990); Cochius, et al., J. Neurol. Neurosurg. Psychiatry55:1094-1095
(1992)] as well as at least 11 patients receiving dura mater grafts [Nisbet, et
al., J. Am. Med. Assoc.261:1118 (1989); Thadani, et al.,
J. Neurosurg.69:766-769 (1988); Willison, et al., J. Neurosurg.
Psychiatric54:940 (1991); Brown, et al., Lancet340:24-27
(1992)]. These cases of iatrogenic CJD underscore the need for screening pharmaceuticals
that might possibly be contaminated with prions.
Recently, two doctors in France were charged with involuntary manslaughter
of a child who had been treated with growth hormones extracted from corpses. The
child developed Creutzfeldt-Jakob Disease. (See New Scientist, July 31, 1993,
page 4). According to the Pasteur Institute, since 1989 there have been 24 reported
cases of CJD in young people who were treated with human growth hormone between
1983 and mid-1985. Fifteen of these children have died. It now appears as though
hundreds of children in France have been treated with growth hormone extracted from
dead bodies at the risk of developing CJD (see New Scientist, November 20,
1993, page 10.) Prior attempts to create PrP monoclonal antibodies have been unsuccessful
(see Barry and Prusiner, J. of Infectious Diseases Vol. 154, No. 3, Pages 518-521
(1986). Thus there is a need for an assay to detect compounds which result in disease.
Specifically, there is a need for a convenient, cost-effective assay for testing
sample materials for the presence of prions which cause CJD. The present invention
offers such an assay.
Summary of the Invention
Antibodies of the invention will specifically bind to a native prion
protein (i.e., native PrPSc) in situ with a high degree of binding
affinity. The antibodies can be placed on a substrate and used for assaying a sample
to determine if the sample contains a pathogenic form of a prion protein. The antibodies
are characterized by one or more of the following features (1) an ability to neutralize
infectious prions, (2) will bind to prion proteins (PrPSc)
in situ i.e., will bind to naturally occurring forms of a prion protein in
a cell culture or in vivo and without the need to treat (e.g., denature)
the prion protein, and (3) will bind to a high percentage of the PrPSc
form (i.e. disease form) of prion protein in a composition e.g., will bind to 50%
or more of the PrPSc form of the prion proteins. Preferred antibodies
are further characterized by an ability to (4) bind to a prion protein of only a
specific species of mammals e.g., bind to human prion protein and not prion protein
of other mammals.
An important object is to provide antibodies which bind to native
prion protein (PrPSc).
Another object is to provide antibodies which specifically bind to
epitopes of prion proteins (PrPSc) of a specific species of animal and
not to the prion protein (PrPSc) of other species of animals.
Another object is to provide monoclonal antibodies which specifically
bind to prion proteins (PrPsc) associated with disease, (e.g., human PrPSc)
which antibodies do not bind to denatured PrP proteins not associated with disease
(e.g., human PrPC).
Still another object is to provide specific methodology to allow others
to generate a wide range of specific antibodies characterized by their ability to
bind one or more types of prion proteins from one or more species of animals.
Another object of the invention is to provide an assay for the detection
of PrPSc forms of PrP proteins.
Another object of the invention is to provide an assay which can specifically
differentiate prion protein (PrPSc) associated with disease from PrPSc
not associated with disease.
Another object is to detect prions which specifically bind to native
PrPSc of a specific species such as a human, cow, sheep, pig, dog, cat
An advantage of the invention is that it provides a fast, efficient
cost effective assay for detecting the presence of native PrPSc in a
A specific advantage is that the assay can be used as a screen for
the presence of prions (i.e., PrPSc) in products such as pharmaceuticals
(derived from natural sources) food, cosmetics or any material which might contain
such prions and thereby provide further assurances as to the safety of such products.
Another advantage is that the antibodies which can be used with a
protease which denatures PrPc thereby providing for a means of differentiating
between infectious (PrPSc) and non-infectious forms (PrPSc)
Yet another advantage of the invention is that antibodies of the invention
are characterized by their ability to neutralize the infectivity of naturally occurring
prions e.g., neutralize PrPSc.
Another advantage is that antibodies of the invention will bind to
(PrPSc) prion proteins in situ, i.e., will bind to naturally occurring
(PrPSc) prions in their natural state in a cell culture or
in vivo without requiring that the prion proteins be particularly treated,
isolated or denatured.
Another advantage is that the prion proteins of the invention will
bind to a relatively high percentage of the infectious form of the prion protein
(e.g., PrPSc) -- for example bind to 50% or more of the PrPSc
form of prion proteins in a composition.
An important feature of the invention is that the methodology makes
it possible to create a wide variety of different prion protein antibodies with
the same or individually engineered features which features may make the antibody
particularly suitable for uses such as (1) prion neutralization to purify a product,
(2) the extraction of prion proteins and (3) therapies.
A feature of the invention is that it uses phage display libraries
in the creation of the antibodies.
Another feature of the invention is that the phage are genetically
engineered to express a specific binding protein of an antibody on their surface.
These and other objects, advantages, and features of the invention
will become apparent to those persons skilled in the art upon reading the details
of the chimeric gene, assay method, and transgenic mouse as more fully described
Brief Description of the Drawings
Figure 1 is a schematic view of a portion of PrP proteins showing the differences
between a normal, wild-type human PrP protein and a normal, wild-type mouse PrP
Figure 2 shows the amino acid sequence of mouse PrP along with specific differences
between mouse PrP and human PrP;
Figure 3 shows the amino acid sequence of mouse PrP and specifically shows differences
between mouse PrP and bovine PrP;
Figure 4 shows the amino acid sequence of mouse PrP and specifically shows differences
between mouse PrP and bovine PrP;
Figure 5 is a bar graph of serum dilution vs optical density at 405 nm for the
mouse (D7282) for serum against denatured mouse PrP 27-30;
Figure 6 shows the amino acid sequences of selected (A) heavy chain and (B)
light chain variable regions generated by panning an IgG1 library from mouse D7282
against denatured MoPrP 27-30 rods;
Figure 7 shows the deduced amino acid sequences for some of the phage clones
obtained in one panning against PrP;
Figures 8A-8H show photos of histoblots 8A, 8B, 8C, 8D, 8E, 8F, 8G and 8H showing
staining of SHaPrP 27-30 and denatured SHaPrP 27-30;
Figure 9 is a graph showing the ELISA reactivity of purified Fabs against prion
protein SHa 27-30;
Figure 10 is a graph of the ELISA reactivity of purified Fabs against denatured
prion protein SHa 27-30;
Figure 11 is a photo showing amino precipitation of SHaPrP 27-30 with recombinant
Fabs of the invention; and
Figure 12 is a photo showing amino precipitation of SHaPrP 27-30 with purified
Fabs of the invention.
Detailed Description of Preferred Embodiments
Before the present antibodies, assays and methods for producing an
using such are disclosed and described, it is to be understood that this invention
is not limited to particular antibodies, assays or method as such may, of course,
vary. It is also to be understood that the terminology used herein is for the purpose
of describing particular embodiments only, and is not intended to be limiting, since
the scope of the present invention will be limited only by the appended claims.
Unless defined otherwise, all technical and scientific terms used
herein have the same meaning as commonly understood by one of ordinary skill in
the art to which this invention belongs. Although any methods and materials similar
or equivalent to those described herein can be used in the practice or testing of
the present invention, the preferred methods and materials are now described. All
publications mentioned herein are incorporated herein by reference to disclose and
describe the methods and/or materials in connection with which the publications
The terms "PrP protein", "PrP" and the like are used interchangeably
herein and shall mean both the infectious particle form PrPSc known to
cause diseases (spongiform encephalopathies) in humans and animals and the non-infectious
form PrPC which, under appropriate conditions is converted to the infectious
The terms "prion", "prion protein" and "PrPSc protein"
and the like used interchangeably herein to refer to the infectious PrPSc
form of a PrP protein and is a contraction of the words "protein" and "infection"
and the particles are comprised largely if not exclusively of PrPSc molecules
encoded by a PrP gene. Prions are distinct from bacteria, viruses and viroids. Known
prions include those which infect animals to cause scrapie, a transmissible, degenerative
disease of the nervous system of sheep and goats as well as bovine spongiform encephalopathies
(BSE) or mad cow disease and feline spongiform encephalopathies of cats. Four prion
diseases known to affect humans are (1) kuru, (2) Creutzfeldt-Jakob Disease (CJD),
(3) Gerstmann-Strassler-Scheinker Disease (GSS), and (4) fatal familial insomnia
(FFI). As used herein prion includes all forms of prions causing all or any of these
diseases or others in any animals used - and in particular in humans and in domesticated
The term "PrP gene" is used herein to describe genetic material which
expresses proteins as shown in Figures 2-4 and polymorphisms and mutations such
as those listed herein under the subheading "Pathogenic Mutations and Polymorphisms."
The term "PrP gene" refers generally to any gene of any species which encodes any
form of a prion protein. Some commonly known PrP sequences are described in Gabriel
et al., Proc. Natl. Acad. Sci. USA89:9097-9101 (1992) which is incorporated
herein by reference to disclose and describe such sequences. The PrP gene can be
from any animal including the "host" and "test" animals described herein and any
and all polymorphisms and mutations thereof, it being recognized that the terms
include other such PrP genes that are yet to be discovered. The protein expressed
by such a gene can assume either a PrPC (non-disease) of PrPSc
The terms "standardized prion preparation", "prion preparation", "preparation"
and the like are used interchangeably herein to describe a composition containing
prions (PrPSc) which composition is obtained from brain tissue of mammals
which contain substantially the same genetic material as relates to prions, e.g.,
brain tissue from a set of mammals which exhibit signs of prion disease which mammals
(1) include a transgene as described herein; (2) have an ablated endogenous prion
protein gene; (3) have a high copy number of prion protein gene from a genetically
diverse species; or (4) are hybrids with an ablated endogenous prion protein gene
and a prion protein gene from a genetically diverse species. The mammals from which
standardized prion preparations are obtained exhibit clinical signs of CNS dysfunction
as a result of inoculation with prions and/or due to developing the disease due
to their genetically modified make up, e.g., high copy number of prion protein genes.
The term "artificial PrP gene" is used herein to encompass the term
"chimeric PrP gene" as well as other recombinantly constructed genes which when
included in the genome of a host animal (e.g., a mouse) will render the mammal susceptible
to infection from prions which naturally only infect a genetically diverse test
mammal, e.g., human, bovine or ovine. In general, an artificial gene will include
the codon sequence of the PrP gene of the mammal being genetically altered with
one or more (but not all, and generally less than 40) codons of the natural sequence
being replaced with a different codon - preferably a corresponding codon of a genetically
diverse mammal (such as a human). The genetically altered mammal being used to assay
samples for prions which only infect the genetically diverse mammal. Examples of
artificial genes are mouse PrP genes encoding the sequence as shown in Figures 2,
3 and 4 with one or more different replacement codons selected from the codons shown
in these Figures for humans, cows and sheep replacing mouse codons at the same relative
position, with the proviso that not all the mouse codons are replaced with differing
human, cow or sheep codons. Artificial PrP genes can include not only codons of
genetically diverse animals but may include codons and codon sequences not associated
with any native PrP gene but which, when inserted into an animal render the animal
susceptible to infection with prions which would normally only infect a genetically
The terms "chimeric gene," "chimeric PrP gene", "chimeric prion protein
gene" and the like are used interchangeably herein to mean an artificially constructed
gene containing the codons of a host animal such as a mouse with one or more of
the codons being replaced with corresponding codons from a genetically diverse test
animal such as a human, cow or sheep. In one specific example the chimeric gene
is comprised of the starting and terminating sequence (i.e., N- and C-terminal codons)
of a PrP gene of a mammal of a host species (e.g. a mouse) and also containing a
nucleotide sequence of a corresponding portion of a PrP gene of a test mammal of
a second species (e.g. a human). A chimeric gene will, when inserted into the genome
of a mammal of the host species, render the mammal susceptible to infection with
prions which normally infect only mammals of the second species. The preferred chimeric
gene disclosed herein is MHu2M which contains the starting and terminating sequence
of a mouse PrP gene and a non-terminal sequence region which is replaced with a
corresponding human sequence which differs from a mouse PrP gene in a manner such
that the protein expressed thereby differs at nine residues.
The term "genetic material related to prions" is intended to cover
any genetic material which effects the ability of an animal to become infected with
prions. Thus, the term encompasses any "PrP gene", "artificial PrP gene", "chimeric
PrP gene" or "ablated PrP gene" which terms are defined herein as well as modification
of such which effect the ability of an animal to become infected with prions. Standardized
prion preparations are produced using animals which all have substantially the same
genetic material related to prions so that all of the animals will become infected
with the same type of prions and will exhibit signs of infection at about the same
The terms "host animal" and "host mammal" are used to describe animals
which will have their genome genetically and artificially manipulated so as to include
genetic material which is not naturally present within the animal. For example,
host animals include mice, hamsters and rats which have their PrP gene ablated i.e.,
rendered inoperative. The host is inoculated with prion proteins to generate antibodies.
The cells producing the antibodies are a source of genetic material for making a
phage library. Other host animals may have a natural (PrP) gene or one which is
altered by the insertion of an artificial gene or by the insertion of a native PrP
gene of a genetically diverse test animal.
The terms "test animal" and "test mammal" are used to describe the
animal which is genetically diverse from the host animal in terms of differences
between the PrP gene of the host animal and the PrP gene of the test animal. The
test animal may be any animal for which one wishes to run an assay test to determine
whether a given sample contains prions with which the test animal would generally
be susceptible to infection. For example, the test animal may be a human, cow, sheep,
pig, horse, cat, dog or chicken, and one may wish to determine whether a particular
sample includes prions which would normally only infect the test animal.
The terms "genetically diverse animal" and "genetically diverse mammal"
are used to describe an animal which includes a native PrP codon sequence of the
host animal which differs from the genetically diverse test animal by 17 or more
codons, preferably 20 or more codons, and most preferably 28-40 codons. Thus, a
mouse PrP gene is genetically diverse with respect to the PrP gene of a human, cow
or sheep, but is not genetically diverse with respect to the PrP gene of a hamster.
The terms "ablated PrP protein gene", "disrupted PrP gene", and the
like are used interchangeably herein to mean an endogenous PrP gene which has been
altered (e.g., add and/or remove nucleotides) in a manner so as to render the gene
inoperative. Examples of non-functional PrP genes and methods of making such are
disclosed in Büeler, H., et al "Normal development of mice lacking the neuronal
cell-surface PrP protein" Nature 356, 577-582 (1992) and Weisman (WO 93/10227).
The methodology for ablating a gene is taught in Capecchi, Cell 51:503-512 (1987)
all of which are incorporated herein by reference. Preferably both alleles of the
genes are disrupted.
The terms "hybrid animal", "transgenic hybrid animal" and the like
are used interchangeably herein to mean an animal obtained from the cross-breeding
of a first animal having an ablated endogenous prion protein gene with a second
animal which includes either (1) a chimeric gene or artificial PrP gene or (2) a
PrP gene from a genetically diverse animal. For example a hybrid mouse is obtained
by cross-breeding a mouse with an ablated mouse gene with a mouse containing (1)
human PrP genes (which may be present in high copy numbers) or (2) chimeric genes.
The term hybrid includes any offspring of a hybrid including inbred offspring of
two hybrids provided the resulting offspring is susceptible to infection with prions
with normal infect only a genetically diverse species. A hybrid animal can be inoculated
with prions and serve as a source of cells for the creation of hybridomas to make
monoclonal antibodies of the invention.
The terms "susceptible to infection" and "susceptible to infection
by prions" and the like are used interchangeably herein to describe a transgenic
or hybrid test animal which develops a disease if inoculated with prions which would
normally only infect a genetically diverse test animal. The terms are used to describe
a transgenic or hybrid animal such as a transgenic mouse Tg(MHu2M) which, without
the chimeric PrP gene, would not become infected with a human prion but with the
chimeric gene is susceptible to infection with human prions.
By "antibody" is meant an immunoglobulin protein which is capable
of binding an antigen. Antibody as used herein is meant to include the entire antibody
as well as any antibody fragments (e.g. F(ab')2' Fab', Fab, Fv) capable
of binding the epitope, antigen or antigenic fragment of interest.
Antibodies of the invention are immunoreactive or immunospecific for
and therefore specifically and selectively bind to a PrPSc protein. Antibodies
which are immunoreactive and immunospecific for natural or native PrPSc
are preferred. Antibodies for PrPSc are preferably immunospecific --
i.e., not substantially cross-reactive with related materials. Although the term
"antibody" encompasses all types of antibodies (e.g., monoclonal) the antibodies
of the invention are preferably produced using the phage display methodology described
By "purified antibody" is meant one which is sufficiently free of
other proteins, carbohydrates, and lipids with which it is naturally associated.
Such an antibody "preferentially binds" to a native PrPSc protein (or
an antigenic fragment thereof), i.e., does not substantially recognize and bind
to other antigenically-unrelated molecules. A purified antibody of the invention
is preferably immunoreactive with and immunospecific for a PrPSc protein
of specific species and more preferably immunospecific for native human PrpSc.
By "antigenic fragment" of a PrP protein is meant a portion of such
a protein which is capable of binding an antibody of the invention.
By "binds specifically" is meant high avidity and/or high affinity
binding of an antibody to a specific polypeptide i.e., epitope of a PrPSc
protein. Antibody binding to its epitope on this specific polypeptide is preferably
stronger than binding of the same antibody to any other epitope, particularly those
which may be present in molecules in association with, or in the same sample, as
the specific polypeptide of interest e.g., binds more strongly to PrPSc
than denatured fragments of PrPC so that by adjusting binding conditions
the antibody binds almost exclusively to PrPSc and not denatured fragments
of PrPC. Antibodies which bind specifically to a polypeptide of interest
may be capable of binding other polypeptides at a weak, yet detectable, level (e.g.,
10% or less of the binding shown to the polypeptide of interest). Such weak binding,
or background binding, is readily discernible from the specific antibody binding
to the compound or polypeptide of interest, e.g. by use of appropriate controls.
In general, antibodoies of the invention which bind to native PrPScin situ with a binding affinity of 107 mole/l or more, preferably
108 mole/liters or more are said to bind specifically to PrPSc.
In general, an antibody with a binding affinity of 106 mole/liters or
less is not useful in that it will not bind an antigen at a detectable level using
conventional methodology currently used.
By "detectably labeled antibody", "detectably labeled anti-PrP" or
"detectably labeled anti-PrP fragment" is meant an antibody (or antibody fragment
which retains binding specificity), having an attached detectable label. The detectable
label is normally attached by chemical conjugation, but where the label is a polypeptide,
it could alternatively be attached by genetic engineering techniques. Methods for
production of detectably labeled proteins are well known in the art. Detectable
labels may be selected from a variety of such labels known in the art, but normally
are radioisotopes, fluorophores, paramagnetic labels, enzymes (e.g., horseradish
peroxidase), or other moieties or compounds which either emit a detectable signal
(e.g., radioactivity, fluorescence, color) or emit a detectable signal after exposure
of the label to its substrate. Various detectable label/substrate pairs (e.g., horseradish
peroxidase/diaminobenzidine, avidin/streptavidin, luciferase/luciferin)), methods
for labelling antibodies, and methods for using labeled antibodies are well known
in the art (see, for example, Harlow and Lane, eds. (Antibodies: A Laboratory
Manual (1988) Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY)).
The terms "treatment", "treating" and the like are used herein to
generally mean obtaining a desired pharmacologic and/or physiologic effect. The
effect may be prophylactic in terms of completely or partially preventing a disease
or symptom thereof and/or may be therapeutic in terms of a partial or complete cure
for a disease and/or adverse effect attributable to the disease. "Treatment" as
used herein covers any treatment of a disease in a mammal, particularly a human,
(a) preventing the disease from occurring in a subject which may be predisposed
to the disease but has not yet been diagnosed as having it;
(b) inhibiting the disease, i.e., arresting its development; or
(c) relieving the disease, i.e., causing regression of the disease. The invention
is directed toward treating patients with infectious prions and is particularly
directed toward treating humans infected with PrPSc, resulting in a disease
of the central nervous system such as bovine spongiform encephalopathy; Creutzfeldt-Jakob
Disease; fatal familial insomnia or Gerstmann-Strassler-Scheinker Disease.
Abbreviations used herein include:
CNS for central nervous system;
BSE for bovine spongiform encephalopathy;
CJD for Creutzfeldt-Jakob Disease;
FFI for fatal familial insomnia;
GSS for Gerstmann-Strassler-Scheinker Disease;
Hu for human;
HuPrP for a human prion protein;
Mo for mouse;
MoPrP for a mouse prion protein;
SHa for a Syrian hamster;
SHaPrP for a Syrian hamster prion protein;
Tg for transgenic;
Tg(SHaPrP) for a transgenic mouse containing the PrP gene of a Syrian hamster;
Tg(HuPrP) for transgenic mice containing the complete human PrP gene;
Tg(ShePrP) for transgenic mice containing the complete sheep PrP gene;
Tg(BovPrP) for transgenic mice containing the complete cow PrP gene;
PrPSc for the scrapie isoform of the prion protein;
PrPC for the cellular contained comon, normal isoform of the prion
MoPrPSc for the scrapie isoform of the mouse prion protein;
MHu2M for a chimeric mouse/human PrP gene wherein a region of the mouse PrP
gene is replaced by a corresponding human sequence which differs from mouse PrP
at 9 codons;
Tg(MHu2M) mice are transgenic mice of the invention which include the chimeric
MHu2MPrPSc for the scrapie isoform of the chimeric human/mouse PrP
PrPCJD for the CJD isoform of a PrP gene;
Prnp0/0 for ablation of both alleles of an endogenous prion protein
gene, e.g., the MoPrP gene;
Tg(SHaPrP+/0)81/Prnp0/0 for a particular line (81) of
transgenic mice expressing SHaPrP, +/0 indicates heterozygous;
Tg (HuPrP) /Prnp0/0 for a hybrid mouse obtained by crossing a mouse
with a human prion protein gene (HuPrP) with a mouse with both alleles of the endogenous
prion protein gene disrupted;
Tg(MHu2M)/Prnp0/0 for a hybrid mouse obtained by crossing a mouse
with a chimeric prion protein gene (MHu2M) with a mouse with both alleles of the
endogenous prion protein gene disrupted.
FVB for a standard inbred strain of mice often used in the production
of transgenic mice since eggs of FVB mice are relatively large and tolerate microinjection
of exogenous DNA relatively well.
General Aspect of the Invention
The core of the invention is an antibody which specifically binds
to a PrPSc protein and preferably binds to a native non-denatured PrPSc
protein in situ with an affinity of 107 moles/liter or more, preferable
108 moles/liter or more of a single species (e.g., human) and more preferably
binds only to human PrPSc and not denatured fragments of human PrPC).
The antibody may bind to all proteins coded by the different mutations and/or polymorphisms
of the PrP protein gene. Alternatively, a battery of antibodies (2 or more different
antibodies) are provided wherein each antibody of the battery specifically binds
to protein coded by a different mutation or polymorphism of the PrP gene. The antibody
can be bound to support surface and used to assay a sample in vitro for the
presence of a particular type of human PrPSc. The antibody can also be
bound to a detectable label and injected into an animal to assay in vivo
for the presence of a particular type of native PrPSc.
Although there are known procedures for producing antibodies from
any given antigen practice has shown that it is particularly difficult to produce
antibodies which bind to certain proteins e.g., PrPSc. The difficulty
with obtaining antibodies to PrPSc relates, in part, to its special and
unknown qualities. By following procedures described herein antibodies which bind
native PrPScin situ have been obtained and others may follow
the procedures described here to obtain other antibodies to PrPSc and
to other proteins for which it is difficult to generate antibodies.
To produce antibodies of the invention it is preferable to begin with
inoculating a host mammal with prion proteins i.e., infectious PrPSc.
The host mammal may be any mammal and is preferably a host mammal of the type defined
herein such as a mouse, rat, guinea pig or hamster and is most preferably a mouse.
The host animal is inoculated with prion proteins which are endogenous to a different
species which is preferably a genetically diverse species. For example a mouse is
inoculated with human prion proteins. Preferably, the host mammal is inoculated
with infectious prion proteins of a genetically diverse mammal. For example, a mouse
is inoculated with human PrPSc. Using a normal host mammal in this manner
it is possible to elicit the generation of some antibodies. However, when a hosts
animal includes a prion protein gene and is inoculated with prions from a genetically
diverse species antibodies will, if at all, only be generated for epitopes which
differ between epitopes of the prion protein of the host animal and epitopes of
the genetically diverse species. This substantially limits the amount of antibodies
which might be generated and decreases the ability to find an antibody which selectively
binds to an infectious form of a prion protein and does not bind to denatured fragments
of a non-infectious form. Thus, unless one is attempting to generate antibodies
which differentiate between prion proteins of different species it is preferable
to begin the antibody production process using a mammal which has an ablated prion
protein gene i.e., a null PrP gene abbreviated as Prnp0/0. Accordingly,
the invention is generally described in connection with the use of such "null" mammals
and specifically described in connection with "null mice."
Antibodies are produced by first producing a host animal (e.g., a
mouse) which has its endogenous PrP gene ablated, i.e., the PrP gene rendered inoperative.
A mouse with an ablated PrP gene is referred to as a "null mouse". A null mouse
can be created by inserting a segment of DNA into a normal mouse PrP gene and/or
removing a portion of the gene to provide a disrupted PrP gene. The disrupted gene
is injected into a mouse embryo and via homologous recombination replaces the endogenous
The null mouse is injected with prions in order to stimulate the formation
of antibodies. Further, injections of adjuvants and prions are generally used to
maximize the generation of antibodies.
The mouse is then sacrificed and bone marrow and spleen cells are
removed. The cells are lysed, RNA is extracted and reversed transcribed to cDNA.
Antibody heavy and light chains (or parts thereof) and then amplified by PCR. The
amplified cDNA library may be used as is or after manipulation to create a range
of variants and thereby increase the size of the library.
An IgG phage display library is then constructed by inserting the
amplified cDNA encoding IgG heavy chain and the amplified cDNA encoding a light
chain into a phage display vector (e.g., a pComb3 vector) such that one vector contains
a cDNA insert encoding a heavy chain fragment in a first expression cassette of
the vector, and a cDNA insert encoding a light chain fragment in a second expression
cassette of the vector.
Ligated vectors are then packaged by filamentous phage M13 using methods
well known in the art. The packaged library is then used to infect a culture of
E. coli, so as to amplify the number of phage particles. After bacterial
cell lysis, the phage particles are isolated and used in a panning procedure.
The library created is panned against a composition containing prions.
Antibody fragments which selectively bind to PrPSc e.g., human PrPSc
are then isolated.
Specifics of a PrP Protein
The major component of purified infectious prions, designated PrP
27-30, is the proteinase K resistant core of a larger native protein PrpSc
which is the disease causing form of the ubiquitous cellular protein PrPC.
PrPSc-is found only in scrapie infected cells whereas PrpC
is present in both infected and uninfected cells implicating PrPSc as
the major, if not the sole, component of infectious prion particles. Since both
PrPC and PrPSc are encoded by the same single copy gene, great
effort has been directed toward unraveling the mechanism by which PrPSc
is derived from PrPC. Central to this goal has been the characterization
of physical and chemical differences between these two molecules. Properties distinguishing
PrPSc from PrPC include low solubility (Meyer, et al 1986
PNAS), poor antigenicity (Kascack, J. Virol 1987; Serban D. 1990) protease
resistance (Oesch, et al 1985 Cell) and polymerization of PrP 27-30 into
rod-shaped aggregates which are very similar, on the ultrastructural and histochemical
levels, to the PrP amyloid plaques seen in scrapie diseased brains (Prusiner,
et al Cell 1983). By using proteinase K it is possible to denature PrPC
but not PrPSc. To date, attempts to identify any post-transitional chemical
modifications in PrPC that lead to its conversion to PrPSc
have proven fruitless (Stahl, et al 1993 Biochemistry). Consequently, it
has been proposed that PrPC and PrPSc are in fact conformational
isomers of the same molecule.
Conformational description of PrP using conventional techniques has
been hindered by problems of solubility and the difficulty in producing sufficient
quantities of pure protein. However, PrPC and PrPSc are conformationally
distinct. Theoretical calculations based upon the amino acid sequences of PrPs from
several species have predicted four putative helical motifs in the molecule. Experimental
spectroscopic data would indicate that in PrPC these regions adopt α-helical
arrangements, with virtually no β-sheet (Pan, et al PNAS 1993). In dramatic
contrast, in the same study it was found that PrPSc and PrP 27-30 possess
significant β-sheet content, which is typical of amyloid proteins. Moreover,
studies with extended synthetic peptides, corresponding to PrP amino acid residues
90-145, have demonstrated that these truncated molecules may be converted to either
α-helical or β-sheet structures by altering their solution conditions.
The transition of PrPC to PrPSc requires the adoption of
β-sheet structure by regions that were previously α-helical.
In general, scrapie infection fails to produce an immune response,
with host organisms being tolerant to PrPSc from the same species. Polyclonal
anti-PrP antibodies have though been raised in rabbits following immunization with
large amounts of SHaPrP 27-30 (Bendheim, et al PNAS 1985, Bode, et al J. Gen. Virol.
1985). Similarly, a handful of anti-PrP monoclonal antibodies have been produced
in mice (Kascack, et al, J. Virol. 1987, Barry, et al, J. Infect. Dis. 1986). These
antibodies are able to recognize native PrPC and denatured PrPSc
from both SHa and humans equally well, but do not bind to MoPrP. Unsurprisingly,
the epitopes of these antibodies were mapped to regions of sequence containing amino
acid differences between SHa- and MoPrP (Rogers, et al, J. Immunol. 1993).
It is not entirely clear as to why antibodies of the type described
in the above cited publications will bind to PrPC but not to PrPSc.
Without being bound to any particular theory it is suggested that such may take
place because epitopes which are exposed when the protein is in the PrPC
conformation are unexposed or partially hidden in the PrPSc configuration
-- where the protein is relatively insoluble and more compactly folded together.
It is pointed out that stating that an antibody binds to PrPC but not
to PrPSc is not correct in absolute terms (but correct in commonly accepted
terms) because some minimal binding to PrPSc may occur. For purposes
of the invention an indication that no binding occurs means that the equilibrium
or affinity constant Ka is 106 l/mole or less. Further, binding
will be recognized as existing when the Ka is at 107 l/mole
or greater preferably 108 l/mole or greater. The binding affinity of
107 l/mole or more may be due to (1) a single monoclonal antibody (i.e.,
large numbers of one kind of antibodies) (2) a plurality of different monoclonal
antibodies (e.g., large numbers of each of five different monoclonal antibodies)
or (3) large numbers of polyclonal antibodies. It is also possible to use combinations
Preferred antibodies will bind 50% or more of the PrPSc
in a sample. However, this may be accomplished by using several different antibodies
as per (1)-(3) above. It has been found that an increased number of different antibodies
is more effective in binding a larger percentage of the PrPSc in a sample
as compared to the use of a single antibody. For example, the use of six copies
of a single antibody "Q" might bind 40% of the PrPSc in a sample. Similar
results might be obtained with six copies of antibody "R" and "S". However, by using
two copies each of "Q", "R" and "S" the six antibodies will bind over 50% of the
PrPSc in a sample. Thus, a synergistic effect can be obtained by combining
combinations of two or more antibodies which bind PrPSc i.e., by combining
two or more antibodies which have a binding affinity Ka for PrPSc
of 107 l/mole or more. Thus combination of D4, R2, 6D2, D14, R1 and R10
and/or related antibodies can provide synergistic results.
Antibody/Antigen Binding Forces
The forces which hold an antigen and antibody together are in essence
no different from non-specific interactions which occur between any two unrelated
proteins i.e., other macromolecules such as human serum albumin and human transferrin.
These intermolecular forces may be classified into four general areas which are
(1) electrostatic; (2) hydrogen bonding; (3) hydrophobic; and (4) Van der Waals.
Electrostatic forces are due to the attraction between oppositely charged ionic
groups on two protein side-chains. The force of attraction (F) is inversely proportional
to the square of the distance (d) between the charges. Hydrogen bonding forces are
provided by the formation of reversible hydrogen bridges between hydrophilic groups
such as -OH, -NH2 and -COOH. These forces are largely dependent upon
close positioning of two molecules carrying these groups. Hydrophobic forces operate
in the same way that oil droplets in water merge to form a single large drop. Accordingly,
non-polar, hydrophobic groups such as the side-chains on valine, leucine and phenylalanine
tend to associate in an aqueous environment. Lastly, Van der Waals are forces created
between molecules which depend on interaction between the external electron clouds.
Further information regarding each of the different types of forces
can be obtained from "Essential Immunology" edited by I.M. Roitti (6th Edition)
Blackwell Scientific Publications, 1988. With respect to the present invention useful
antibodies exhibit all of these forces. It is by obtaining an accumulation of these
forces in larger amounts that it is possible to obtain an antibody which has a high
degree of affinity or binding strength to the PrP protein and in particular an antibody
which has a high degree of binding strength to PrPScin situ.
Measuring Antibody/Antigen Binding Strength
The binding affinity between an antibody and an antigen can be measured
which measurement is an accumulation of a measurement of all of the forces described
above. Standard procedures for carrying out such measurements exist and can be directly
applied to measure the affinity of antibodies of the invention for PrP proteins
including native PrPScin situ.
One standard method for measuring antibody/antigen binding affinity
is through the use of a dialysis sac which is a container comprised of a material
which is permeable to the antigen but impermeable to the antibody. Antigens which
are bound completely or partially to antibodies are placed within the dialysis sac
in a solvent such as in water. The sac is then placed within a larger container
which does not contain antibodies or antigen but contains only the solvent e.g.,
the water. Since only the antigen can diffuse through the dialysis membrane the
concentration of the antigen within the dialysis sac and the concentration of the
antigen within the outer larger container will attempt to reach an equilibrium.
After placing the dialysis sac into the larger container and allowing for time to
pass towards reaching an equilibrium it is possible to measure the concentration
of the antigen within the dialysis sac and within the surrounding container and
then determine the differences in concentration. This makes it possible to calculate
the amount of antigen which remains bound to antibody in the dialysis sac and the
amount which disassociates from the antibody and diffuses into the surrounding container.
By constantly renewing the solvent (e.g., the water) within the surrounding container
so as to remove any antigen which is diffused thereinto it is possible to totally
disassociate the antibody from antigen within the dialysis sac. If the surrounding
solvent is not renewed the system will reach an equilibrium and it is possible to
calculate the equilibrium constant (K) of the reaction i.e., the association and
disassociation between the antibody and antigen. The equilibrium constant (K) is
calculated as an amount equal to the concentration of antibody bound to antigen
within the dialysis sac divided by the concentration of free antibody combining
sites times the concentration of free antigen. The equilibrium constant or "K" value
is generally measured in terms of liters per mole. The K value is a measure of the
difference in free energy (deta g) between the antigen and antibody in the free
state as compared with the complexed form of the antigen and antibody. When using
the phage display methodology described below the antibodies obtained have an affinity
or K value of 107 mole/liter or more.
As indicated above the term "affinity" describes the binding of an
antibody to a single antigen determinate. However, in most practical circumstances
one is concerned with the interaction of an antibody with a multivalent antigen.
The term "avidity" is used to express this binding. Factors which contribute to
avidity are complex and include the heterogeneity of the antibodies in a given serum
which are directed against each determinate on the antigen and the heterogeneity
of the determinants themselves. The multivalence of most antigens leads to an interesting
"bonus" effect in which the binding of two antigen molecules by an antibody is always
greater, usually many fold greater, than the arithmetic sum of the individual antibody
links. Thus, it can be understood that the measured avidity between an antiserum
and a multivalent antigen will be somewhat greater than the affinity between an
antibody and a single antigen determinate.
Null PrP Mice to make Antibodies
The present invention circumvents problems of tolerance and more efficiently
generates panels of monoclonal antibodies capable of recognizing diverse epitopes
on Mo and other PrPs in part using mice with both alleles of the PrP gene (Prnp)
are ablated (Prnp0/0) (Bueler, et al, 1992). These PrP-deficient
mice (or null mice), are indistinguishable from normal mice in their development
and behavior. These null mice are resistant to scrapie following intracerebral inoculation
of infectious MpPrPSc (Bueler, et al, 1993 Cell; Prusiner,
et al, PNAS 1993). In addition Prnp0/0 mice will develop IgG serum
titers against Mo-, SHa and human PrP following immunization with relatively small
quantities of purified SHaPrP 27-30 in adjuvant (Prusiner, et al, PNAS 1993).
After allowing sufficient time to generate antibodies the immunized Prnp0/0
mice were sacrificed for hybridoma production in the conventional manner. Fusions
derived from these mice did secrete PrP specific antibody. However, these hybridomas
would not secrete PrP specific antibodies for more than a few hours. In view of
the somewhat limited success a different approach was taken.
Combinatorial antibody library technology, i.e., antigen based selection
from antibody libraries expressed on the surface of M13 filamentous phage, offers
a new approach to the generation of monoclonal antibodies and possesses a number
of advantages relative to hybridoma methodologies which are particularly pertinent
to the prion problem (Huse, et al. 1989; Barbas, et al, 1991; Clackson, et al,
1991; Burton and Barbas, 1994). The present invention uses such technology to
provide PrP-specific monoclonal antibodies from phage antibody libraries prepared
from MoPrP-immunized Prnp0/0 mice. The invention provides the first monoclonal
antibodies recognizing MoPrP in situ and demonstrates the application of
combinatorial libraries for cloning specific antibodies from null mice. The general
methodologies involved in creating large combinatorial libraries using phage display
technology is described and disclosed in U.S. Patent 5,223,409 issued June 29, 1993
which patent is incorporated herein by reference to disclose and describe phage
The invention is largely described herein with respect to null mice
i.e., FVB mice with both alleles of the PrP gene ablated. However, other host animals
can be used and preferred host animals are mice and hamsters, with mice being most
preferred in that there exists considerable knowledge on the production of transgenic
animals. Possible host animals include those belonging to a genus selected from
Mus (e.g. mice), Rattus (e.g. rats), Oryctolagus (e.g. rabbits), and Mesocricetus
(e.g. hamsters) and Cavia (e.g., guinea pigs). In general mammals with a normal
full grown adult body weight of less than 1 kg which are easy to breed and maintain
can be used.
The genetic material which makes up the PrP gene is known for a number
of different species of animals (see Gabriel et al., Proc. Natl. Acad. Sci. USA89:9097-9101 (1992)). Further, there is considerable homology between the
PrP genes in different mammals. For example, see the amino acid sequence of mouse
PrP compared to human, cow and sheep PrP in Figures 2, 3 and 4 wherein only the
differences are shown. Although there is considerable genetic homology with respect
to PrP genes, the differences are significant in some instances. More specifically,
due to small differences in the protein encoded by the PrP gene of different mammals,
a prion which will infect one mammal (e.g. a human) will not normally infect a different
mammal (e.g. a mouse). Due to this "species barrier", it is not generally possible
to use normal animals, (i.e., animal which have not had their genetic material related
to PrP proteins manipulated) such as mice to determine whether a particular sample
contains prions which would normally infect a different species of animal such as
a human. The present invention solves this problem by providing antibodies which
bind to native PrPSc proteins of any species of animal for which the
antibody is designed.
Pathogenic mutations and polymorphisms
There are a number of known pathogenic mutations in the human PrP
gene. Further, there are known polymorphisms in the human, sheep and bovine PrP
genes. The following is a list of such mutations and polymorphisms:
The DNA sequence of the human, sheep and cow PrP genes have been determined
allowing, in each case, the prediction of the complete amino acid sequence of their
respective PrP proteins. The normal amino acid sequence which occurs in the vast
majority of individuals is referred to as the wild-type PrP sequence. This wild-type
sequence is subject to certain characteristic polymorphic variations. In the case
of human PrP, two polymorphic amino acids occur at residues 129 (Met/Val) and 219
(Glu/Lys). Sheep PrP has two amino acid polymorphisms at residues 171 and 136, while
bovine PrP has either five or six repeats of an eight amino acid motif sequence
in the amino terminal region of the mature prion protein. While none of these polymorphisms
are of themselves pathogenic, they appear to influence prion diseases. Distinct
from these normal variations of the wild-type PrP proteins, certain mutations of
the human PrP gene which alter either specific amino acid residues of PrP or the
number of octarepeats have been identified which segregate with inherited human
In order to provide further meaning to the above chart demonstrating
the mutations and polymorphisms, one can refer to the published sequences of PrP
genes. For example, a chicken, bovine, sheep, rat and mouse PrP gene are disclosed
and published within Gabriel et al., Proc. Natl. Acad. Sci. USA89:9097-9101 (1992). The sequence for the Syrian hamster is published in
Basler et al.,Cell46:417-428 (1986). The PrP gene of sheep is published
by Goldmann et al., Proc. Natl. Acad. Sci. USA87:2476-2480 (1990).
The PrP gene sequence for bovine is published in Goldmann et al., J. Gen. Virol.72:201-204 (1991). The sequence for chicken PrP gene is published in Harris
et al., Proc. Natl. Acad. Sci. USA88:7664-7668 (1991). The PrP gene
sequence for mink is published in Kretzschmar et al., J. Gen. Virol.73:2757-2761 (1992). The human PrP gene sequence is published in Kretzschmar
et al., DNA5:315-324 (1986). The PrP gene sequence for mouse is published
in Locht et al., Proc. Natl. Acad. Sci. USA83:6372-6376 (1986). The
PrP gene sequence for sheep is published in Westaway et al., Genes Dev.8:959-969 (1994). These publications are all incorporated herein by reference
to disclose and describe the PrP gene and PrP amino acid sequences.
"Strains" of Human Prions
Studies in rodents have shown that prion strains produce different
patterns of PrPSc accumulation [Hecker et al., Genes & Development6:1213-1228 (1992); DeArmond et al., Proc. Natl. Acad. Sci. USA90:6449-6453 (1993)]; which can be dramatically changed by the sequence of
PrPSc [Carlson et al., Proc. Natl. Acad. Sci. USAin press
(1994)]. The molecular basis of prion diversity has for many years been attributed
to a scrapie specific nucleic acid [Bruce et al., J. Gen. Virol.68:79-89 (1987)] but none has been found [Meyer et al., J. Gen. Virol.72:37-49 (1991); Kellings et al., J. Gen. Virol.73:1025-1029
(1992)]. Other hypotheses to explain prion strains include variations in PrP Asn-linked
sugar chains [Hecker et al., Genes & Development6:1213-1228 (1992)]
and multiple conformers of PrPSc [Prusiner, S.B., Science252:1515-1522
(1991)]. The patterns of PrPSc in Tg(MHu2M) mice were remarkably similar
for the three inocula from humans dying of CJD.
The patterns of PrPSc accumulation in the brains of inoculated
Tg(MHu2M) mice were markedly different for RML prions and Hu prions. However, RML
prion inocula containing MoPrPSc stimulated the formation of more MoPrPSc
while Hu prion inocula containing HuPrPCJD triggered production of MHu2MPrPSc.
The distribution of neuropathological changes characterized by neuronal vacuolation
and astrocytic gliosis is similar to the patterns of PrPSc accumulation
in the brains of Tg(MHu2M) mice inoculated with RML prions or Hu prions.
Standardized Prion Preparation
Standardized prion preparations may be produced in order to test assays
of the invention and thereby improve the reliability of the assay. Although the
preparation can be obtained from any animal it is preferably obtained from a host
animal which has brain material containing prions of a test animal. For example,
a transgenic mouse containing a human prion protein gene can produce human prions
and the brain of such a mouse can be used to create a standardized human prion preparation.
Further, in that the preparation is to be a "standard" it is preferably obtained
from a battery (e.g., 100; 1,000, or more animals) of substantial identical animals.
For example, 100 mice all containing a very high copy number of human PrP genes
(all polymorphisms and mutations) would spontaneously develop disease and the brain
tissue from each could be combined to make a useful standardized prion preparation.
Standardized prion preparations can be produced using any of modified
host mammals of the type described above. For example, standardized prion preparations
could be produced using mice, rats, hamsters, or guinea pigs which are genetically
modified so that they are susceptible to infection with prions which prions would
generally only infect genetically diverse species such as a human, cow, sheep or
horse and which modified host mammals will develop clinical signs of CNS dysfunction
within a period of time of 350 days or less after inoculation with prions. The most
preferred host mammal is a mouse in part because they are inexpensive to use and
because a greater amount of experience has been obtained with respect to production
of transgenic mice than with respect to the production of other types of host animals.
Details regarding making standardized prion preparation are described in U.S. Patent
application entitled "Method of Detecting Prions in a Sample and Transgenic Animal
Used For Same" filed August 31, 1995, Serial No. 08/521,992 and U.S. Patent application
entitled "Detecting Prions In A Sample And Prion Preparation And Transgenic Animal
Used For Same", Attorney Docket No. 06510/056001, filed July 30, 1996, both of which
applications are incorporated herein by reference.
Once an appropriate type of host is chosen, such as a mouse, the next
step is to choose the appropriate type of genetic manipulation to be utilized to
produce a standardized prion formulation. For example, the mice may be mice which
are genetically modified by the insertion of a chimeric gene of the invention. Within
this group the mice might be modified by including high copy numbers of the chimeric
gene and/or by the inclusion of multiple promoters in order to increase the level
of expression of the chimeric gene. Alternatively, hybrid mice of the invention
could be used wherein mice which have the endogenous PrP gene ablated are crossed
with mice which have a human PrP gene inserted into their genome. There are, of
course, various subcategories of such hybrid mice. For example, the human PrP gene
may be inserted in a high copy number an/or used with multiple promoters to enhance
expression. In yet another alternative the mice could be produced by inserting multiple
different PrP genes into the genome so as to create mice which are susceptible to
infection with a variety of different prions, i.e., which generally infect two or
more types of test animals. For example, a mouse could be created which included
a chimeric gene including part of the sequence of a human, a separate chimeric gene
which included part of the sequence of a cow and still another chimeric gene which
included part of the sequence of a sheep. If all three different types of chimeric
genes were inserted into the genome of the mouse the mouse would be susceptible
to infection with prions which generally only infect a human, cow and sheep.
After choosing the appropriate mammal (e.g., a mouse) and the appropriate
mode of genetic modification (e.g., inserting a chimeric PrP gene) the next step
is to produce a large number of such mammals which are substantially identical in
terms of genetic material related to prions. More specifically, each of the mice
produced will include an identical chimeric gene present in the genome in substantially
the same copy number. The mice should be sufficiently identical genetically in terms
of genetic material related to prions that 95% or more of the mice will develop
clinical signs of CNS dysfunction within 350 days or less after inoculation and
all of the mice will develop such CNS dysfunction at approximately the same time
e.g., within ± 30 days of each other.
Once a large group e.g., 50 or more, more preferably 100 or more,
still more preferably 500 or more of such mice are produced. The next step is to
inoculate the mice with prions which generally only infect a genetically diverse
mammal e.g., prions from a human, sheep, cow or horse. The amounts given to different
groups of mammals could be varied. After inoculating the mammals with the prions
the mammals are observed until the mammals exhibit symptoms of prion infection e.g.,
clinical signs of CNS dysfunction. After exhibiting the symptoms of prion infection
the brain or at least a portion of the brain tissue of each of the mammals is extracted.
The extracted brain tissue is homogenized which provides the standardized prion
As an alternative to inoculating the group of transgenic mice with
prions from a genetically diverse animal it is possible to produce mice which spontaneously
develop prion related diseases. This can be done, for example, by including extremely
high copy numbers of a human PrP gene into a mouse genome. When the copy number
is raised to, for example, 100 or more copies; the mouse will spontaneously develop
clinical signs of CNS dysfunction and have, within its brain tissue, prions which
are capable of infecting humans. The brains of these animals or portions of the
brain tissue of these animals can be extracted and homogenized to produce a standardized
The standardized prion preparations can be used directly or can be
diluted and tittered in a manner so as to provide for a variety of different positive
controls. More specifically, various known amounts of such standardized preparation
can be used to inoculate a first set of transgenic control mice. A second set of
substantially identical mice are inoculated with a material to be tested i.e., a
material which may contain prions. A third group of substantially identical mice
are not injected with any material. The three groups are then observed. The third
group, should, of course not become ill in that the mice are not injected with any
material. If such mice do become ill the assay is not accurate probably due to the
result of producing mice which spontaneously develop disease. If the first group,
injected with a standardized preparation, do not become ill the assay is also inaccurate
probably because the mice have not been correctly created so as to become ill when
inoculated with prions which generally only infect a genetically diverse mammal.
However, if the first group does become ill and the third group does not become
ill the assay can be presumed to be accurate. Thus, if the second group does not
become ill the test material does not contain prions and if the second group does
become ill the test material does contain prions.
By using standardized prion preparations of the invention it is possible
to create extremely dilute compositions containing the prions. For example, a composition
containing one part per million or less or even one part per billion or less can
be created. Such a composition can be used to test the sensitivity of the antibodies,
assays and methods of the invention in detecting the presence of prions.
Prion preparations are desirable in that they will include a constant
amount of prions and are extracted from an isogeneic background. Accordingly, contaminates
in the preparations will be constant and controllable. Standardized prion preparations
will be useful in the carrying out of bioassays in order to determine the presence,
if any, of prions in various pharmaceuticals, whole blood, blood fractions, foods,
cosmetics, organs and in particular any material which is derived from an animal
(living or dead) such as organs, blood and products thereof derived from living
or dead humans. Thus, standardized prion preparations will be valuable in validating
purification protocols where preparations are spiked and reductions in teeter measured
for a particular process.
As indicated above and described further below in detailed examples
it is possible to use the methodology of the invention to create a wide range of
different antibodies. i.e., antibodies having different specific features. For example,
antibodies can be created which bind only to a prion protein naturally occurring
within a single species and not bind to a prion protein naturally occurring within
other species. Further, the antibody can be designed so as to bind only to an infectious
form of a prion protein (e.g., PrPSc) and not bind to a non-infectious
form (e.g., PrPC). A single antibody or a battery of different antibodies
can then be used to create an assay device. Such an assay device can be prepared
using conventional technology known to those skilled in the art. The antibody can
be purified and isolated using known techniques and bound to a support surface using
known procedures. The resulting surface having antibody bound thereon can be used
to assay a sample in vitro to determine if the sample contains one or more types
of antibodies. For example, antibodies which bind only to human PrPSc
can be attached to the surface of a material and a sample can be denatured via proteinase
K. The denatured sample is brought into contact with the antibodies bound to the
surface of material. If no binding occurs it can be deduced that the sample does
not contain human PrPSc.
Antibodies of the invention are also characterized by their ability
to neutralize prions. Specifically, when antibodies of the invention are allowed
to bind to prions the infectivity of the prion is eliminated. Accordingly, antibody
compositions of the invention can be added to any given product in order to neutralize
any infectious prion protein within the product. Thus, if a product is produced
from a natural source which might contain infectious prion proteins the antibodies
of the invention could be added as a precaution thereby eliminating any potential
infection resulting from infectious prion proteins.
The antibodies of the invention can be used in connection with immunoaffinity
chromatography technology. More specifically, the antibodies can be placed on the
surface of a material within a chromatography column. Thereafter, a composition
to be purified can be passed through the column. If the sample to be purified includes
any prion protein which binds to the antibodies those prion proteins (PrPSc)
will be removed from the sample and thereby purified.
Lastly, the antibodies of the invention can be used to treat a mammal.
The antibodies can be given prophylactically or be administered to an individual
already infected with infectious prion proteins such infection having been determined
by the use of the assay described above. The exact amount of antibody to be administered
will vary depending on a number of factors such as the age, sex, weight and condition
of the patient. Those skilled in the art can determine the precise amount by administering
antibodies in small amounts and determining the effect and thereafter adjusting
the dosage. It is suggested that the dosage can vary from 0.01 mg/kg to about 300
mg/kg, preferably about 0.1 mg/kg to about 200 mg/kg, more preferably about 0.2
mg/kg to about 20 mg/kg in one or more dose administrations daily, for one or several
Preferred is administration of the antibody for 2 to 5 or more consecutive days
in order to avoid "rebound" of prion infectivity occurring.
The following examples are put forth so as to provide those of ordinary
skill in the art with a complete disclosure and description of how to make and use
the chimeric genes, transgenic mice and assays of the present invention, and are
not intended to limit the scope of what the inventors regard as their invention.
Efforts have been made to ensure accuracy with respect to numbers used (e.g. amounts,
temperature, etc.) but some experimental errors and deviations should be accounted
for. Unless indicated otherwise, parts are parts by weight, molecular weight is
weight average molecular weight, temperature is in degrees Centigrade, and pressure
is at or near atmospheric.
Construction of phage display antibody libraries expressing antibodies
Construction of phage display libraries for expression of antibodies,
particularly the Fab portion of antibodies, is well known in the art. Preferably,
the phage display antibody libraries that express antibodies are prepared according
to the methods described in U.S. Patent 5,223,409, issued June 29, 1993 and U.S.
Patent Application Serial No. 07/945,515, filed September 16, 1992, incorporated
herein by reference. Procedures of the general methodology can be adapted using
the present disclosure to produce antibodies of the present invention.
Isolation of RNA encoding prion-specific antibodies
In general, the phage display anti-PrP antibody libraries are prepared
by first isolating a pool of RNA that contains RNA encoding anti-PrP antibodies.
To accomplish this, an animal (e.g., a mouse, rat, or hamster) is immunized with
prion of interest. However, normal animals do not produce antibodies to prions at
detectable or satisfactorily high levels. This problem is avoided by immunizing
animals in which the (PrP) gene has been ablated on both alleles. Such mice are
designated Prnp0/0 and methods for making such mice are disclosed in
Büeler, Nature (1992) and in Weismann Publication WO 93/10227, published
May 27, 1993. Inoculation of "null" animals with prions results in production of
IgG serum titers against the prion (Prusiner et al. PNAS 1993). In one preferred
embodiment, the animal selected for immunization is a Prnp0/0 mouse described
by Büeler and Weismann.
Generally, the amount of prion necessary to elicit a serum antibody
response in a "null" animal is from about 0.01 mg/kg to about 500 mg/kg.
The prion protein is generally administered to the animal by injection,
preferably by intraperitoneal or intravenous injection, more preferably by intraperitoneal
injection. The animals are injected once, with at least 1 to 4 subsequent booster
injections, preferably at least 3 booster injections. After immunization, the reactivity
of the animal's antisera with the prion can be tested using standard immunological
assays, such as ELISA or Western blot, according to methods well known in the art
(see, for example, Harlow and Lane, 1988, Antibodies: A Laboratory Manual,
Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY). Animals having prion-binding
antisera may be boosted with an additional injection of prion.
Serum antibody levels are predictive of antibody secretion, and therefore
of levels of specific mRNA in lymphocytes, particularly plasma cells. Detection
of serum antibodies, particularly relatively high levels of serum antibodies, is
thus correlated to a high level of lymphocytes such as plasma cells producing mRNA
encoding those serum antibodies. Thus, plasma cells isolated from the prion protein-immunized
mice will contain a high proportion of lymphocytes (e.g., plasma cells) producing
prion-specific antibody, particularly when the plasma cells are isolated from the
mice within a short time period after the final injection boost (e.g., about 2 to
5 days, preferably 3 days). Immunization of the mice and the subsequent injection
boosters thus serve to increase the total percentage of anti-PrP antibody-producing
plasma cells present in the total population of the mouse's plasma cells. Moreover,
because the anti-PrP antibodies are being produced at or near peak serum levels,
then anti-PrP antibody-producing plasma cells are producing anti-PrP antibodies,
and thus mRNA encoding these antibodies at or near peak levels.
The above correlation between serum levels of antigen-specific antibodies,
the number of lymphocytes producing those antigen-specific antibodies, and the amount
of total mRNA encoding the antigen-specific antibodies provides a means for isolating
a pool of mRNA that is enriched for the mRNA encoding antigen-specific antibodies
of interest. Lymphocytes, including plasma cells are isolated from spleen and/or
bone marrow from the prion-immunized animals according to methods well known in
the art (see, for example, Huse et al. Science 1989). Preferably the lymphocytes
are isolated about 2 to 5 days, preferably about 3 days after the final immunization
boost. The total RNA is then extracted from these cells. Methods for RNA isolation
from mammalian cells are well known in the art (see, for example, Sambrook et al.,
1989, Molecular Cloning: A Laboratory Manual, 2nd Ed., Cold Spring Harbor
Laboratory Press, Cold Spring Harbor, NY).
Production of cDNA encoding antibodies from lymphocyte mRNA
cDNA is produced from the isolated RNA using reverse transcriptase
according to methods well known in the art (see, for example, Sambrook et al., supra),
and cDNA encoding antibody heavy chains or light chains is amplified using the polymerase
chain reaction (PCR). The 3' primers used to amplify heavy chain or light chain-encoding
cDNAs are based upon the known nucleotide sequences common to heavy chain or light
chain antibodies of a specific antibody subclass. For example, one set of primers
based upon the constant region of the IgG1 heavy chain-encoding gene can be used
to amplify heavy chains of the IgG1 subclass, while another set of primers based
upon the constant portion of the IgG1 light chain-encoding gene is used to amplify
the light chains of the IgG1 subclass. The '5 primers are consensus sequences based
upon examination of a large number of variable sequences in the data base. In this
manner, DNA encoding all antibodies of a specific antibody class or subclass are
amplified regardless of antigen-specificity of the antibodies encoded by the amplified
DNA. The entire gene encoding the heavy chain or the light chain can be amplified.
Alternatively, only a portion of the heavy or light chain encoding gene may be amplified,
with the proviso that the product of PCR amplification encodes a heavy or light
chain gene product that can associate with its corresponding heavy or light chain
and function in antigen binding i.e., bind selectively to a prion protein. Preferably,
the phage display product is a Fab or Fv antibody fragment.
The antibody encoding cDNA selected for amplification may encode any
isotope and preferably encode a subclass of IgG. Exemplary mouse IgG subclasses
include IgG1, IgG2a, IgG2b, and IgG3. The selection of the specific antibody subclass-encoding
cDNA for amplification will vary according to a variety of factors, including, for
example, the animal's serum antibody response to the antigen. Preferably, the antibody
subclass-encoding cDNA selected for PCR amplification is that antibody subclass
for which the animal produced the highest titer of antibody. For example, if the
titers of serum IgG1 are higher than any other subclass of IgG detected in the serum
antibody response, then cDNA encoding IgG1 is amplified from the cDNA pool.
Preferably, the heavy and light chains are amplified from the plasma
cell cDNA to produce two separate amplified cDNA pools: 1) a cDNA pool containing
heavy chain cDNA amplimer products, where the heavy chain is of a specific antibody
subclass; and 2) a cDNA pool containing light chain cDNA amplimer products, where
the light chain is of a specific antibody subclass.
Antibodies From Transgenic Animals
In addition to obtaining genetic material which encodes antibodies
by infecting an animal with an antigen and thereafter extracting cells (and their
DNA) responsible for antibody production it is possible to obtain the genetic material
by producing a transgenic animal or by using the above described technology and
transgenic animal technology in order to produce chimeric mouse/human or fully human
antibodies. The technology for producing a chimeric or wholly foreign immunoglobins
involves obtaining from cells of transgenic animals which have had inserted into
their germ line a genetic material which encodes all or part of an immunoglobin
which binds to the desired antigen. Wholly human antibodies can be produced from
transgenic mice which have had inserted into their genome genetic material which
encodes human antibodies. The technology for producing such antibodies from transgenic
animals is described within PCT Publication No. WO 90/04036, published April 19,
1990. Further, see Goodhartd, et al, Proc. Natl. Acad. Sci. U.S.A. Vol. 84,
pages 4229-4233, June 1987 and Bucchine, et al, Nature, Vol. 326, pages 409-411,
March 26, 1987, all of which are incorporated herein by reference to disclose and
describe methods of producing antibodies from transgenic animals.
Vectors for use with phage display antibody libraries
The heavy chain-encoding cDNAs and the light chain-encoding cDNAs
are then each inserted into separate expression cassettes of an appropriate vector.
Preferably the vector contains a nucleotide sequence encoding and capable of expressing
a fusion polypeptide containing, in the direction of amino- to carboxy-terminus,
1) a prokaryotic secretion signal domain, 2) an insertion site for DNA encoding
a heterologous polypeptide (e.g., either the heavy or light chain-encoding cDNA),
and in the expression cassette for the heavy chain cDNA 3) a filamentous phage membrane
The vector includes prokaryotic or mammalian DNA expression control
sequences for expressing the fusion polypeptide, preferably prokaryotic control
sequences. The DNA expression control sequences can include any expression signal
for expressing a structural gene product, and can include 5' and 3' elements operatively
linked to the expression cassette for expression of the heterologous polypeptide.
The 5' control sequence defines a promoter for initiating transcription, and a ribosome
binding site operatively linked at the 5' terminus of the upstream translatable
sequence. The vector additionally includes an origin of replication for maintenance
and replication in a prokaryotic cell,. preferably a gram negative cell such as
E. coli. The vector can also include genes whose expression confers a selective
advantage, such as drug resistance, to a prokaryotic or eukaryotic cell transformed
with the vector.
The filamentous phage membrane anchor is preferably a domain of the
cpIII or cpVIII coat protein capable of associating with the matrix of a filamentous
phage particle, thereby incorporating the fusion polypeptide onto the phage surface.
The secretion signal is a leader peptide domain of a protein that targets the protein
to the periplasmic membrane of gram negative bacteria. Such leader sequences for
gram negative bacteria (such asE. coli) are well known in the art (see, for
example, Oliver, In Neidhard, F.C. (ed.), Escherichia coli andSalmonella
typhimurium, American Society for Microbiology, Washington, D.C.,
Filamentous phage membrane anchors for use in the phage display vector
Preferred membrane anchors for the vector are obtainable from filamentous
phage M13, f1, fd, and equivalent filamentous phage. Preferred membrane anchor domains
are found in the coat proteins encoded by gene III and gene VIII. The membrane anchor
domain of a filamentous phage coat protein is a portion of the carboxy terminal
region of the coat protein and includes a region of hydrophobic amino acid residues
for spanning a lipid bilayer membrane, and a region of charged amino acid residues
normally found at the cytoplasmic face of the membrane and extending away from the
membrane. In the page f1, gene VIII coat protein's membrane spanning region comprises
the carboxy-terminal 11 residues from 41 to 52 (Ohkawa et al., J. Biol. Chem.,
256:9951-9958, 1981). An exemplary membrane anchor would consist of residues
26 to 40 to cpVIII. Thus, the amino acid residue sequence of a preferred membrane
anchor domain is derived from the M13 filamentous phage gene VIII coat protein (also
designated cpVIII or CP 8). Gene VIII coat protein is present on a mature filamentous
phage over the majority of the phage particle with typically about 2500 to 3000
copies of the coat protein.
The amino acid residue sequence of another preferred membrane anchor
domain is derived from the M13 filamentous phage gene III coat protein (also designate
cpIII). Gene III coat protein is present on a mature filamentous phage at one end
of the phage particle with typically about 4 to 6 copies of the coat protein. Detailed
descriptions of the structure of filamentous phage particles, their coat proteins,
and particles assembly are found in the reviews by Rached et al., (Microbiol.
Rev., 50:401-427, 1986) and Model et al. (In:The Bacteriophages: Vol.
2, R. Calendar, ed., Plenum Publishing Co., pgs. 375-456, 1988).
Preferably, the filamentous phage membrane anchor-encoding DNA is
inserted 3' of the cDNA insert in the library vector such that the phage membrane
anchor-encoding DNA can be easily excised and the vector relegated without disrupting
the rest of the expression cassettes of the vector. Removal of the phage membrane
anchor-encoding DNA from the vector, and expression of this vector in an appropriate
host cell, results in the production of soluble antibody (Fab) fragments. The soluble
Fab fragments retain the antigenicity of the phage-bound Fab, and thus can be used
in assays and therapies in the manner that whole (non-fragmented) antibodies are
The vector for use with the present invention must be capable of expressing
a heterodimeric receptor (such as an antibody or antibody Fab). That is, the vector
must be capable of independently containing and expressing two separate cDNA inserts
(e.g., the heavy chain cDNA and the light chain cDNA). Each expression cassette
can include the elements described above, except that the filamentous phage anchor
membrane-encoding DNA is present only in the expression cassette for the heavy chain
cDNA. Thus, when the antibody or Fab is expressed on the surface of the phage, only
the heavy chain polypeptide is anchored to the phage surface. The light chain is
not directly bound to the phage surface, but is indirectly bound to the phage via
its association with the free portion of the heavy chain polypeptide (i.e., the
portion of the heavy chain that is not bound to the phage surface).
Preferably, the vector contains a sequence of nucleotides that allow
for directional ligation, i.e., a polylinker. The polylinker is a region of the
DNA expression vector that operatively links the upstream and downstream translatable
DNA sequence for replication and transport, and provides a site or means for directional
ligation of a DNA sequence into the vector. Typically, a directional polylinker
is a sequence of nucleotides that defines two or more restriction endonuclease recognition
sequence, or restriction sites. Upon restriction enzyme cleavage, the two sites
yield cohesive termini to which a translatable DNA sequence can be ligated to the
DNA expression vector. Preferably, the two cohesive termini are non-complementary
and thereby permit directional insertion of the cDNA into the cassette. Polylinkers
can provide one or multiple directional cloning sites, and may or may not be translated
during expression of the inserted cDNA.
Preferably, the expression vector is capable of manipulating in the
form of a filamentous phage particle. Such DNA expression vectors additionally contain
a nucleotide sequence that defines a filamentous phage origin of replication such
that the vector, upon presentation of the appropriate genetic complement, can replicate
as a filamentous phage in single stranded replicative form, and can be packaged
into filamentous phage particles. This feature provides the ability of the DNA expression
vector to be packaged into phage particles for subsequent isolation of individual
phage particles (e.g., by infection of and replication in isolated bacterial colonies).
A filamentous phage origin of replication is a region of the phage
genome that defines sites for initiation of replication, termination of replication,
and packaging of the replicative form produced by replications (see, for example,
Rasched et al.,Microbiol. Rev.,50:401-427, 1986; Horiuchi,
J. Mol. Biol., 188:215-223, 1986). A preferred filamentous phage origin
of replication for use in the present invention is an M13, f1, or fd phage origin
of replication (Short et al., Nucl. Acids Res.,16:7583-7600, 1988).
Preferred DNA expression vectors are the expression vectors pCOMB8, pCKAB8, pCOMB2-8,
pCOMB3, pCKAB3, pCOMB2-3, pCOMB2-3' and pCOMB3H.
The pComb3H vector is a modified form of pComb3 in which (i) heavy
and light chains are expressed from a single Lac promoter as opposed to individual
promoters and (ii) heavy and light chains have two different leader sequences (pg1B
and ompA) as opposed to the same leader sequence (pHB). Reference for pComb3H Wang,
et al (1995) J. Mol. Biol., Inpress. The principles of pComb3H are basically the
same as for pComb3.
Production of the phage display antibody library
After the heavy chain and light chain cDNAs are cloned into the expression
vector, the entire library is packaged using an appropriate filamentous phage. The
phage are then used to infect a phage-susceptible bacterial culture (such as a strain
of E. coli), the phage allowed to replicate and lyse the cells, and the lysate
isolated from the bacterial cell debris. The phage lysate contains the filamentous
phage expressing on its surface the cloned heavy and light chains isolated from
the immunized animal. In general, the heavy and light chains are present on the
phage surface as Fab antibody fragments, with the heavy chain of the Fab being anchored
to the phage surface via the filamentous phage membrane anchor portion of the fusion
polypeptide. The light chain is associated with the heavy chain so as to form an
antigen binding site. Method of producing chimeric antibodies are described within
U.S. Patent 4,816,567, issued March 28, 1989 to Cabilly, et al which is incorporated
herein by reference to disclose and describe such procedures. Further, See Bobrzecka,
et al,Immunology Letters, 2, pages 151-155 (1980) and Konieczny, et al,
Haematologia 14 (1), pages 85-91 (1981) also incorporated herein by reference.
Selection of prion-antigen specific Fabs from the phage display antibody
Phage expressing an antibody or Fab that specifically binds a prion
antigen can be isolated using any of a variety of protocols for identification and
isolation of monoclonal and/or polyclonal antibodies. Such methods include, immunoaffinity
purification (e.g., binding of the phage to a columna having bound antigen) and
antibody panning methods (e.g., repeated rounds of phage binding to antigen bound
to a solid support for selection of phage of high binding affinity to the antigen).
Preferably, the phage is selected by panning using techniques that are well known
in the art.
After identification and isolation of phage expressing anti-PrP antibodies,
the phage can be used to infect a bacterial culture, and single phage isolates identified.
Each separate phage isolate can be again screened using one or more of the methods
described above. In order to further confirm the affinity of the phage for the antigen,
and/or to determine the relative affinities of the phage for the antigen, the DNA
encoding the antibodies or Fabs can be isolated from the phage, and the nucleotide
sequence of the heavy and light chains contained in the vector determined using
methods well known in the art (see, for example, Sambrook et al., supra).
Isolation of soluble Fabs from phage selected from the phage display
Soluble antibodies or Fabs can be produced from a modified display
the same dicistronic vector by excising the DNA encoding the filamentous phage anchor
membrane that is associated with the expression cassette for the heavy chain of
the antibody. Preferably, the DNA encoding the anchor membrane is flanked by convenient
restriction sites that allow excision of the anchor membrane sequence without disruption
of the remainder of the heavy chain expression cassette or disruption of any other
portion of the expression vector. The modified vector without the anchor membrane
sequence then allows for production of soluble heavy chain as well as soluble light
chain following packaging and infection of bacterial cells with the modified vector.
Alternatively, where the vector contains the appropriate mammalian
expression sequences the modified vector can be used to transform a eukaryotic cell
(e.g., a mammalian or yeast cell, preferably a mammalian cell (e.g., Chinese hamster
ovary (CHO) cells)) for expression of the Fab. Where the modified vector does not
provide for eukaryotic expression, preferably the vector allows for excision of
both the heavy and light chain expression cassettes as a single DNA fragments for
subcloning into an appropriate vector. Numerous vectors for expression of proteins
in prokaryotic and/or eukaryotic cells are commercially available and/or well known
in the art (see, for example Sambrook et al., supra).
Examples 14-18 below and specifically Example 17 show the isolation
of an antibody which specifically binds to PrPSc without any denaturation.
A sample containing PrP proteins (i.e., PrPC and PrPSc) can
be subjected to denaturation by the use of protease K (PK) digestion. The use of
such will digest PrPC but not PrPSc. Thus, after carrying
out the digestion the sample is contacted with the antibody (e.g., R2) as per Example
17 under suitable binding conditions. Preferably, the antibody is bound to a substrate
and can be positioned such that the sample can be easily contacted with the substrate
material having the antibody bound thereon. If material binds to the antibodies
on the substrate the presence of infectious PrPSc is confirmed.
In commercial embodiments of the invention it may be desirable to
use antibodies of the invention in a sandwich type assay. More particularly, the
antibody of the invention may be bound to a substrate support surface. The sample
to be tested is contacted with the support surface under conditions which allow
for binding. Thereafter, unreacted sites are blocked and the surface is contacted
with a generalized antibody which will bind to any protein thereon. The generalized
antibody is linked to a detectable label. The generalized antibody with detectable
label is allowed to bind to any PrPSc bound to the antibodies on the
support surface. If binding occurs the label can be made to become detectable such
as by generating a color thereby indicating the presence of the label which indirectly
indicates the presence of PrPSc within the sample. The assay can detect
prions (PrPSc) present in an amount of 1 part per million or less, even
one part per billion or less. The PrPSc may be present in a source selected
from the group consisting of (a) a pharmaceutical formulation containing a therapeutically
active component extracted from an animal source, (b) a component extracted from
a human source, (c) an organ, tissue, body fluid or cells extracted from a human
source, (d) a formulation selected form the group consisting of injectables, orals,
creams, suppositories, and intrapulmonary delivery formulations, (e) a cosmetic,
and (f) a pharmaceutically active compound extracted from a mammalian cell culture.
Such source materials can also be treated to remove or neutralize PrPSc
protein by adding an antibody of the invention. The invention also includes a method
of treating, comprising administering to a mammal in need thereof a therapeutically
effective amount of an antibody which selectively binds PrpSc protein
which antibody is characterized by its ability to neutralize PrPSc protein
Antibodies of the invention could be obtained by a variety of techniques.
However, the general procedure involves synthesizing a library of proteins (i.e.,
antibodies or portions thereof) on the surface of phage. The library is then brought
into contact with a composition which includes PrP proteins and in particular is
a naturally occurring composition which includes PrPSc. The phage which
bind to PrP protein are then isolated and the antibody or portion thereof which
binds the PrP protein is isolated. It is desirable to determine the sequence of
the genetic material encoding the antibody or portion thereof. Further, the sequence
can be amplified and inserted, by itself, or with other genetic material into an
appropriate vector and cell line for the production of other antibodies. For example,
a sequence encoding a variable region which binds PrPSc can be fused
with a sequence which encodes a human constant region of an antibody producing a
constant/variable construct. This construct can be amplified and inserted within
a suitable vector which can be inserted within a suitable cell line for the production
of humanized antibodies. Procedures such as this are described within U.S. Patent
4,816,567, issued March 28, 1989 to Cabilly, et al which is incorporated herein
by reference to disclose and describe such procedures. Further, See Bobrzecka, et
al, Immunology Letters, 2, pages 151-155 (1980) and Konieczny, et al,
Haematologia 14 (1), pages 85-91 (1981) also incorporated herein by reference.
When the genetic material encoding an antibody or portion thereof
which binds a PrP protein is isolated it is possible to use that genetic material
to produce other antibodies or portions thereof which have a greater affinity for
binding PrP proteins. This is done by site directed mutagenesis technology or by
random mutagenesis and selection. Specifically, individual codons or groups of codons
within the sequence are removed or replaced with codons which encode different amino
acids. Large numbers of different sequences can be generated, amplified and used
to express variations of the antibody or portions thereof on the surface of additional
phage. These phage can then be used to test for the binding affinity of the antibody
to PrP proteins.
The phage library can be created in a variety of different ways. In
accordance with one procedure a host animal such as a mouse or rat is immunized
with PrP protein and preferably immunized with PrPSc. The immunization
may be carried out along with an adjuvant for the formation of larger amounts and
types of antibodies. After allowing for sufficient time for the generation of antibodies,
cells responsible for antibody production are extracted from the inoculated host
mammal. RNA is isolated from the extracted cells and subjected to reverse transcription
in order to produce a cDNA library. The extracted cDNA is amplified by the use of
primers and inserted into an appropriate phage display vector. The vector allows
the expression of antibodies or portions thereof on the phage surface. It is also
possible to subject the cDNA to site directed mutagenesis prior to insertion into
the display vector. Specifically, codons are removed or replaced with codons expressing
different amino acids in order to create a larger library (i.e., a library of many
variants) which is then expressed on the surface of the phage. Thereafter, as described
above, the phage are brought into contact with the sample and phage which bind to
PrP protein are isolated.
The following examples are put forth so as to provide those of ordinary
skill in the art with a complete disclosure and description of how to make and use
the recombinant anti-PrP antibodies and assays of the present invention, and are
not intended to limit the scope of what the inventors regard as their invention.
Efforts have been made to ensure accuracy with respect to numbers used (e.g. amounts,
temperature, etc.) but some experimental errors and deviations should be accounted
for. Unless indicated otherwise, parts are parts by weight, molecular weight is
weight average molecular weight, temperature is in degrees Centigrade, and pressure
is at or near atmospheric.
EXAMPLE 1Purification of MoPrP 27-30
Purified MoPrP 27-30 rods were prepared from the brains of clinically
ill CD-1 mice inoculated with RML prions (Chandler scrapie isolate (Chandler R.L.
1961 Lancet, 1378-1379)). Prion rods were recovered from sucrose gradient fractions
as previously described (Prusiner, McKinley 1983 Cell). Briefly, the fractions containing
prion rods, which sediment in 48-60% (wt/vol) sucrose, were diluted 2:1 in distilled
water and centrifuged at 100,000 x g for 6 h at 4°C. The pellet was resuspended
in water, centrifuged again, and the rods resuspended at 1 mg/ml in Ca/Mg-free phosphate
buffered saline (PBS) containing 0.2% Sarcosyl. PrP 27-30 was the major protein
as determined by SDS-PAGE and silver staining analysis. Protein quantitation was
performed by bicinchonic acid dye binding, with a known amount of bovine serum albumin
as the protein concentration standard.
EXAMPLE 2Immunization of Prnp0/0 mice
Prnp0/0 mice, in which both alleles of the PrP gene (Prnp)
is ablated, were immunized with the purified MoPrP 27-30 rods, which were isolated
as described in Example 1. Prnp0/0 mice and methods for making this strain
are well known in the art (Büeler, et al. 1992). Prnp0/0 mice, which
are indistinguishable from normal mice in their development and behavior, are resistant
to scrapie following intracerebral inoculation of infectious MoPrPSc
(Büeler, et al. 1993 Cell; Prusiner et al. PNAS 1993), and will develop IgG serum
titers against Mo-, SHa, and human PrP following immunization with relatively small
quantities of purified SHaPrP 27-30 in adjuvant (Prusiner et al. PNAS 1993).
Three (3) six week old Prnp0/0 mice were immunized by intraperitoneal
injection of 100 µg of MoPrP 27-30 rods fully emulsified in complete Freund's adjuvant.
Subsequently mice were boosted 2 times at 2-week intervals with incomplete Freund's
adjuvant containing in the first instance 100 µg, then 50 µg of rods. Four days
after the second boost, the reactivity of each mouse's serum against prion proteins
was analyzed as described below in Example 3. Those mice having anti-PrP reactive
antisera received a third injection boost of 50 µg prion rods in incomplete Freund's
adjuvant 14 days after the second boost.
EXAMPLE 3Serum reactivity of Prnp0/0mice immunized with MoPrP 27-30
A primary prognostic indicator for success in isolating a specific
antibody from combinatorial libraries is serum antibody reactivity with the antigen(s)
to be studied (Burton and Barbas, Adv. Immunol. 1994). Serum antibody levels are
predictive of antibody secretion and therefore predictive of the levels of specific
mRNA in plasma cells. It is this latter factor that ultimately dictates the composition
of the antibody-encoding cDNA library.
Four days after the second boost, the Prnp0/0 mice immunized
with MoPrP 27-30 as described in Example 2 were bled from the tail, and the antisera
stored at -20°C for subsequent immunological analysis. The reactivity of the immunized
mouse serum (IgG1, IgG2a, IgG2b and IgG3 antibody subclasses) was measured against
denatured and non-denatured Mo- and SHaPrP 27-30 in ELISA. ELISA wells were coated
overnight at 4°C with 50 µl of PrP rods at 40 µg/ml in 100 mM sodium bicarbonate
pH 8.6. Where denatured PrP rods were used as the antigen in the ELISA, 50 µl of
6M guanidinium isothiocyanate was added to the well for 15 min at room temperature,
after which the wells were washed 6 times with Ca/Mg-free PBS. All wells were then
blocked with Ca/Mg-free PBS containing 3% BSA. The antisera was serially diluted
in PBS, and incubated with the wells for one hour at 37°C. Excess antisera was removed
by washing 10 times with PBS 10.05% Tween 20 and bound antisera detected using labeled
goat anti-mouse antibody that specifically binds either IgG1, IgG2a, IgG2b or IgG3
All 3 mice produced anti-PrP IgG antibodies. Serum reactivity from
one of the mice, designated D7282, is illustrated in Figure 5 as exemplary of the
antibody responses of the immunized mice. The highest serum titers against Mo- and
SHaPrP antigens were of the IgG1 and IgG2b subclasses. In contrast, the IgG2a and
IgG3 anti-PrP titers were close to the background levels of reactivity seen for
all IgG subclasses in the serum of non-immunized Prnp0/0 mice. Antibody
titers were greater against denatured rods than non-denatured rods. The similar
serum reactivity against Mo- and SHa denatured rods is likely reflective of the
high amino acid sequence homology between the two proteins. However, although there
was considerable serum reactivity against non-denatured Mo- rods (approximately
40-50% of the level of that for denatured MoPrP 27-30), reactivity with non-denatured
SHa rods was at the level of background.
EXAMPLE 4Isolation of mRNA encoding anti-PrP antibodies and construction of
antibody phage display libraries
Three days after the final injection boost, the D7282 mouse was sacrificed
and RNA prepared from bone marrow and splenic tissues. Total RNA from mouse spleen
was prepared according to methods well known in the art (Huse, et al Science 1989).
RNA was prepared from bone marrow tissues by first removing the tibia and fibula
from both rear legs of the mice. The bones were then cut through close to each end,
and their contents flushed out by injection of guanidinium isothiocyanate into the
bone cavity using a 27 gauge needle. RNA preparation was then continued as described
for the mouse spleen.
The RNA preparations were then pooled, and cDNA generated from the
mRNA using reverse transcriptase according to methods well known in the art. Two
cDNA libraries were independently constructed from the D7282 mouse mRNA: 1) an IgG1
library; and 2) a IgG2b library. For each of these libraries, cDNAs encoding heavy
chains and cDNA light chains were separately amplified by PCR from separate fractions
of the pooled cDNA. The oligonucleotide 5' and 3' primers employed for PCR amplification
of DNA fragments encoding murine light (κ) chains and heavy (α1
or α2b) chains of the IgG1 subclass wee those used by Huse, et al (Science
1989) and additional heavy chain primers as presented in Table 1 and heavy chain
polymers which are presented in Table 1. Primers used for amplification of cDNAs
encoding heavy chain fragments.
PCR was performed using a Perkin Elmer 9600 with 35 rounds of amplification; denaturation
at 94°C for 30 sec, hybridization at 52°C for 60 sec and extension at 72°C for 60
The resulting amplified cDNAs encoding heavy chains of the IgG1 and
IgG2b subclasses and light chains were cloned into the vector pComb3. The preparation
of Fab antibody libraries displayed on the surface of a filamentous phage using
the pComb3 vector have been described (Williamson et al. PNAS, 1993; Barbas et al.
PNAS 1991). Briefly, the IgG1 or IgG2b phage display library is constructed by inserting
the amplified cDNA encoding IgG1 or IgG2b heavy chain and the amplified cDNA encoding
light chain into the pComb3H vector such that each vector contains a cDNA insert
encoding a heavy chain fragment in one expression cassette of the vector, and a
cDNA insert encoding a light chain fragment into the other expression cassette of
the vector. The resulting IgG1 library contained approximately 9 x 106
individual clones, while the resulting IgG2b library contained approximately 7 x
106 individual clones.
The ligated vectors were then packaged by the filamentous phage M13
using methods well known in the art (see, for example, Sambrook et al, supra). The
packaged library is then used to infect a culture of E. coli, so as to amplify
the number of phage particles. After bacterial cell lysis, the phage particles are
isolated and used in the panning procedure that follows. Aliquots of the phage library
are stored for future amplification and use. Separate aliquots of the phage libraries
are isolated and stored for future amplification and use.
EXAMPLE 5Screening of the phage display antibody library for binding to PrP
Antigen binding phage were selected for binding to denatured MoPrP
27-30 rods against PrP antigen bound to ELISA wells through a panning procedure
described in (Burton, et al PNAS 1991, Barbas Lerner Methods in Enzymol 1991). Briefly,
ELISA wells were coated overnight at 4°C with 50 µl of MoPrP 27-30 rods at 40 µg/ml
in 100 mM sodium bicarbonate pH 8.6. The PrP rods were then denatured by incubation
with 50 µl of 6M guanidinium isothiocyanate for 15 min at room temperature, after
which the wells were washed 6 times with Ca/Mg-free PBS. The wells were then blocked
with Ca/Mg-free PBS containing 3% BSA.
Aliquots of antibody phage were applied to separate PrP coated ELISA
wells. A total of approximately 1 x 1010 antibody phage were added per
well in the panning experiment.
The phage were incubated with the well-bound MoPrP antigen for 2 hrs
at 37°C. Unbound phage were removed by washing 10 times with PBS 0.5% TWEEN 20.
Bound phage were then removed from the wells by acid elution, pooled, reamplified
and subjected to a second round of panning.
The IgG1 library was selected through 5 rounds of panning. A 40-fold
amplification of PrP-specific antibody phage, as determined by the number of phage
eluted from PrP-coated ELISA wells, was measured from the first to the fifth round.
EXAMPLE 6Soluble Fab production from selected antibody-producing phage
Soluble Fabs were produced from phage clones eluted from the fourth
and fifth rounds of panning. DNA from the selected phage clones was isolated, and
the phage coat protein III (the filamentous phage membrane anchor) was removed from
the pComb3H vector using the appropriate restriction enzymes. The DNA was self-ligated
to yield a vector capable of expressing soluble Fab (the procedure for production
of soluble Fabs is detailed in (Barbas et al. PNAS 1991)). The vectors were then
separately used to transform bacteria for expression of the Fabs, and isolated transformants
Fab expression was induced in an overnight bacterial culture using
isopropyl β-D-thiogalactopyranoside. The bacteria were centrifuged, and the
resulting bacterial pellet was either sonicated or frozen and thawed three times
to release Fab from the bacterial periplasmic space. The bacterial Fab supernatants
were then tested for reactivity against PrP in ELISA.
EXAMPLE 7ELISA analysis of anti-PrP Fabs binding to PrP antigens
The binding of soluble Fabs produced in Example 6 to denatured and
non-denatured PrP antigens as well as to synthetic PrP peptides was determined using
the ELISA assay described in Example 3. Synthetic PrP peptides were produced using
conventional peptide synthesis protocols well known in the art.
Of the Fab clones taken from the fourth round of the panning against
denatured MoPrP rods, less than 5% were reactive with denatured PrP, while approximately
50% of the clones taken from the fifth round of the same panning recognized PrP
antigens. In ELISA all of the reactive clones from this panning were able to bind
specifically to denatured Mo and SHa rods, but not to non-denatured rods from either
species. In addition, all the anti-PrP Fabs failed to recognize synthetic peptides
spanning residues 90-145 of Mo and SHa PrP, suggesting the antibodies bind between
residues 146 and 231 of the prion protein.
EXAMPLE 8Analysis of selected anti-PrP antibody (Fab) binding to prion-infected
and uninfected rodent brain tissue
The reactivity of the antibodies identified by panning of the phage
display antibody library was tested by SDS/PAGE of prion-infected rodent brain tissue
and Western blot analysis using the selected Fabs. Protein from brain tissues of
prion-infected and uninfected mice was used as the antigen against which immunoreactivity
was tested. The antigen was prepared by disrupting rodent brain tissue in Ca/Mg-free
PBS by passage 5 times through a 20 gauge needle, followed by passage 10 times through
a 22 gauge needle. The 10% (wt/vol) homogenate was then centrifuged at 1600 x
g for 5 min at 4°C. Aliquots of the supernatant protein were diluted to a
final concentration of 1 mg/ml in Ca/Mg-free PBS containing 0.2% Sarcosyl. This
dilution was mixed with an equal volume of non-reducing 2 x SDS/PAGE sample buffer
and boiled for 5 min, before SDS/PAGE (Laemmli. U.K. (1970) Nature (London) 227,
Immunoblotting was performed as previously described (Pan et al, PNAS 1993) with
primary mouse IgG antiserum (Pierce) diluted 1:1000.
EXAMPLE 9Nucleic Acid Sequencing
The nucleotide and amino acid sequences of the variable domains of
the antibody light and heavy chains were determined for several of the PrP specific
clones. Nucleic acid sequencing was performed with a model 373A automated DNA sequencer
(Applied Biosystems) using a Taq fluorescent dideoxynucleotide terminator
cycle sequencing kit (Applied Biosystems). Primers for the elucidation of antibody
light-chain sequence were primers MoSeqKb [5'-CAC GAC TGA GGC ACC TCC-3'] and OmpSeq
[5'-AAG ACA GCT ATC GCG ATT GCA G-3'] hybridizing to the (-)-strand and for the
heavy chain MOIgGGzSeq [5'-ATA GCC CTT GAC CAG GCA TCC CAG GGT CAC-3'] binding to
the (+)-strand and PelSeq [5'-ACC TAT TGC CTA CGG CAG CCG-3'] binding to the (-)-strand.
The deduced amino acid sequences for some of the phage clones obtained
in one panning against denatured PrP are provided in Figures 6 and 7. Figure 6 shows
the amino acid sequences of selected (A) heavy chain and (B) light chain variable
regions generated by panning an IgG1 library from mouse D7282 against denatured
MoPrP 27-30 rods. The sequences are very similar but contain a number of heterogeneities
which are likely the result of somatic mutation following repeated exposure of the
mouse to PrP antigen. All of the heavy chain sequences examined in these clones
contained very similar sequences. In particular, the heavy chain complementarity
determining region 3(HCDR3) was identical at the nucleotide level in all the Fab
clones examined. Small differences were observed in the CDR1, CDR2, framework (FR)
3 and FR4 of the heavy chain. These differences are too numerous to be attributable
to PCR or sequencing errors and have probably accrued during rounds of somatic mutation
as the mouse was repeatedly boosted with antigen. The light chain sequences were
also very similar, but with localized heterogeneity throughout the variable domain,
again probably resultant of somatic mutation.
EXAMPLE 10Selection of anti-prion antibodies following masking of epitopes
with existing antibodies
Panning of the IgG1 library against denatured PrP produced a series
of related antibodies, presumably somatic variants of a clone directed to a single
epitope (Example 9). To access antibodies to other epitopes, a' prototype antibody
from the above series was added to denatured PrP in ELISA wells prior to panning
in the normal way. The masking antibody was used in all subsequent panning steps.
Using this procedure, antibodies were derived of different sequence which reacted
with denatured PrP in ELISA. These antibodies are likely directed to different epitopes
on PrP. The masking procedure was carried out as described in Ditzel, et al (1995)
J. Immunol. Masking could also be carried out with molecules other than antibodies
which interacted with PrP.
EXAMPLE 11Selection of phage particles expressing anti-PrP antibodies specific
A phage display antibody library similar to that described in the
Examples above is subjected to panning experiments to identify phage clones that
bind to PrPSc, but not to PrPC. PrPSc antigen and
PrPC antigen are bound to separate wells of a microtiter dish as described
above for the ELISA assay. The phage display antibody library is first panned over
the PrPC ELISA wells. Unbound phage are retrieved from the wells and
pooled. Phage that binds to the PrPC antigen are removed from the wells
and either discarded or pooled for later analyses. The pooled unbound phage are
then again added to PrPC ELISA wells, with selection again being based
upon lack of binding to the PrPC. After several repeated selections on
the PrPC antigen, the phage are pooled and panned on the ELISA wells
containing the PrPSc antigen. The panning is repeated for several rounds,
with the phage that binds to the PrPSc antigen being the phage that is
selected for further rounds of panning. After 5 to 10 rounds of panning on the PrPSc
antigen, the phage are isolated one from another. The ability of the PrPSc-specific
phage or isolated Fab to bind PrPC antigen can be double-checked by ELISA
with the PrPC antigen. The resulting selected phage are those that bind
PrPSc, but do not bind PrPC.
EXAMPLE 12Selection of phage particles expressing anti-PrP antibodies to identify
PrPSc regardless of isoform
A phage display antibody library is prepared as described above from
lymphocyte RNA from a mouse immunized with several PrPSc isoforms, or
from a pool of lymphocyte RNA from several mice immunized with different PrPSc
isoforms. The phage are then panned with several different wells containing antigens
from different isoforms of PrPSc. The phage are panned over each PrPSc
isoform with the selection being for phage that bind the isoform at each stage.
The phage are panned for a total of about 5 to 10 rounds on each PrPSc
isoform. The phage that remain after all stages of panning against all the isoforms
tested are then isolated. The immunoreactivity of each selected phage or isolated
Fab is tested by ELISA or Western blot or histochemistry against each of the various
PrPSc isoforms, as well as for cross-reactivity with PrPC.
EXAMPLE 13Selection of phage particles expressing anti-PrP antibodies specific
for isoforms of PrPSc
A phage display antibody library prepared from lymphocyte RNA of a
mouse immunized with a specific PrPSc isoform is prepared according to
the Examples above. The resulting phage are then selected for their ability to bind
only one specific PrPSc isoform by panning. The panning uses several
different wells containing antigens from different isoforms of PrPSc,
including one set of wells containing antigens from the specific PrPSc
isoform against which specific antibodies are desired. The phage are first panned
over the undesirable PrPSc isoforms, with the selection being for phage
that do not bind the antigen. Panning continues for a total of about 5 to 10 rounds
on each of the PrPSc isoforms. The phage that did not bind the undesirable
PrPSc isoforms are then panned for about 5 to 10 rounds against the desirable
PrPSc isoform, with selection for antigen binding. The phage that remain
after all rounds of panning are isolated. These selected phage are those that express
antibodies with binding specificity for only the specific PrPSc isoform
desired. The immunoreactivity of each selected phage or isolated Fab is tested by
ELISA or Western blot against each of the various PrPSc isoforms, as
well as for cross-reactivity with PrPC.
EXAMPLE 14Generation and Characterization Of Serum Reactivity Against PrPSc
In PrP% Mice
Experimentation per the above Examples established that the primary
prognostic indicator for success in isolating a specific antibody from combinatorial
libraries with the size range of 107 pfu/ml is the serum reactivity with
the antigen to be studied, and it is this factor which will ultimately dictate the
composition of the library. Although Prnp0/0 mice elucidated a strong
immune response upon immunization with either mouse (Mo) or Syrian hamster (SHa)
prion rods composed of PrP 27-30 proteins, the highest serum titers were seen in
the IgG1 and IgG2b subclasses. The IgG2a and IgG3 anti-PrP titers were close to
the background levels of reactivity seen for all IgG subclasses in the serum of
non-immunized mice. In an attempt to increase the immune response and augment the
immune repertoire against PrPSc, Prnp0/0 (94% FVB) female
mice were immunized with liposomes containing SHaPrP 27-30. To further increase
the immune response diversity, mice were immunized using both short and long term
protocols. In contrast to immunization with SHa prion rods immunization with liposomes
containing SHaPrP 27-30 resulted in antiserum titer which includes all four IgG
EXAMPLE 15PrP-immunized Sera Reactivity Against Histoblots
To further investigate the properties of the IgG anti-SHaPrP 27-30
found in the sera from mice immunized with liposomes containing SHaPrP 27-30, we
tested the sera in situ with histoblotting techniques, in which cryostat
sections of normal and scrapie infected SHa brain were transferred onto nitrocellulose
membranes. Although both sera showed some nonspecific reactivity against proteinase
K (PK)-treated normal SHa brain sections, only the sera from the long term immunized
mice showed increased reactivity against PK-treated SHa scrapie infected brain sections.
This reactivity was also evident in sera dilution to 1/1000 (results not shown).
Both sera showed typical reactivity against SHa scrapie infected brain sections
which were first PK-treated and then exposed to 3M GdnSCN for 10 minutes. Sera from
non-immunized Prnp0/0 (94% FVB) female mice did not show any immune reactivity
against normal scrapie infected SHa brain sections.
Staining of SHaPrP 27-30 and Denatured SHaPrP 27-30 in Histoblots
of Scrapie Infected SHa Brain
Histoblots were treated with proteinase K to remove PrPC
from the brain of normal, uninoculated control SHa and SHa showing clinical signs
of scrapie following inoculation with Sc237 prions. To denature SHaPrP 27-30, histoblots
were treated with 3M GdnSCN for 10 minutes. Blots were incubated overnight at 4°C
with sera diluted 1/200 from the short and the long term immunized mice. The results
described here show clear positive reactivity of an antiserum with non-denatured
infectious prions i.e., native PrPSc.
Figure 8 shows eight different stained histoblots of scrapie infected
SHa brain. The histoblots were treated with proteinase K to remove PrPC
from the brain of normal, non-inoculated control SHa(A, C, E and G) and SHa showing
clinical signs of scrapie following inoculation with Sc 237 prions (B, D, F and
H). To denature the SHaPrP 27-30, the histoblots were treated with 3M GdnSCN for
10 minutes (C, D, G and H). The blots were incubated overnight at 4°C with sera
diluted 1/200 from the short (A-D) and the long (E-H) term immunized mice. The results
clearly show the ability of the antibodies of the invention to bind to native, non-denatured
infectious prions i.e., bind to native PrPSc.
EXAMPLE 16Generation Of Monoclonal Antibodies From Immunized Mice Of Example
Overall, eight phage Fab display libraries were constructed: IgG1k,
IgG2ak, IgG2bk and IgG3k from mRNA extracted from the short and long term immunized
mice. To overcome difficulties with the isolation of phage expressing anti-PrP Fab
by panning against prion rods containing PrP 27-30, a panning system was used where
libraries are panned against biotinylated SHa 27-30, dispersed into liposomes, and
bound to streptavidin-coated microtiter plates. After five rounds of panning,E.
Coli extracts from more than 50 clones reacted with biotinylated SHa 27-30,
SHa 27-30 rods and 90-231 recombinant SHa in ELISA. Since these clones also react
with recombinant rPrP corresponding to SHaPrP residues 90-231, Melhorn,I., et al,
High-level Expression and Characterization of a Purified 142-residue Polypeptide
of the Prion Protein. Biochemistry 35, 5528-2237 (1996), all eight libraries
were panned against this antigen to successfully isolate more distinct clones from
virtually all the libraries. Upon DNA sequencing of the plasmid region coding for
the IgG heavy chain, 30 Fabs were identified as distinct clones.
EXAMPLE 17Characterization Of Monoclonal Antibodies
Initial ELISA with E. Coli extracts from positive clones
suggested that the Fabs, in contrast to the monoclonal 3F4 antibody, Kascsak, R.J.,
et al, Mouse Polyclonal and Monoclonal Antibody to Scrapie Associated Fibril Proteins,
J. Virol. 61, 3688-3693 (1987), bind to PrP 27-30 in a native state, i.e.,
without a denaturation step. To characterize quantitatively the novelty of these
Fabs, we purified them and produced 3F4 Fab from the monoclonal 3F4 by enzymatic
cleavage. Standard ELISA for the detection of SHaPrP was performed using different
concentrations of the purified Fabs. In contrast to 3F4 which showed characteristic
SHa PrP binding properties (basal binding to prion rods and strong reactivity against
SHaPrP 27-30 after treatment with 3M non-denaturant GdnSCN), the newly isolated
Fabs reacted against prion rods without any denaturation step. The half-maximal
binding to non-denatured prion rods occurs at a Fab concentration of approximately
0.5 pg/ml, indicating that the antibody has an apparent binding affinity of approximately
Figure 9 is a graph showing the ELISA reactivity of purified Fabs
against prion protein SHa 27-30. The antibody 3F4 and recombinant antibodies were
examined at different concentrations for binding to ELISA wells which were coated
with 0.2 µ/g of sucrose purified infectious SHa prion rods. The results clearly
show that all of the recombinant antibodies of the invention have substantially
higher degrees of binding to prions as compared to the antibody-3F4.
Protocol For ELISA Reactivity Of Purified Fabs Against Denatured
Prion Protein SHa 27-30
Purified 3F4 Fab and recombinant Fabs were examined at different concentrations
for binding to ELISA wells coated with 0.2 µg of sucrose purified SHa prion rods
either native or denatured in the ELISA well with 3M GdnSCN for 10 min.
Figure 10 is a graph showing the results of ELISA reactivity purified
Fabs against denatured prion protein SHa 27-30. Figure 10 is interesting as compared
to Figure 9 in that the recombinant antibodies of the invention as per Figure 9
show a higher degree of affinity for the prion rods as compared to 3F4 whereas all
of the recombinant antibodies but for R1 show a lower degree of affinity against
EXAMPLE 18Characterization Of Monoclonal Antibody By ImmunoprecipitationImmunoprecipitation of SHaPrP 27-30
To confirm the anti-PrP 27-30 activity of the Fabs as well as to confirm
the in-ability of 3F4 to bind nondenatured SHaPrP 27-30, an immunoprecipitation
method was developed using liposomes containing SHa 27-30.E. Coli extracts
from Fab producing clones immunoprecipitated 40-50% of the SHaPrP 27-30 present
in the solution, while 3F4 in dilution of 1/500 immunoprecipitated only trace amounts
of SHaPrP. Fab concentrations in bacterial supernates are typically on the order
of 1-10 pg/ml. This implies that the affinity for antigen are high (on the order
of 107-108 moles/liter or more). The antibody 3F4 was obtained
as an ascetic fluid and is expected to have a concentration of approximately 1 µg/ml
at the dilution used in the immunoprecipitation experiment. The ability of the new
Fabs to immunoprecipitate SHaPrP 27-30 in comparison to 3F4 was determined quantitatively
with purified Fab mAbs D4 and R2. Fab 2R immunoprecipitated SHaPrP 27-30 strongly
at concentrations as low as 0.1 pg/ml (50 ng in 500 pl) indicating an affinity on
the order of greater than 108M-1 (i.e., 108mole/liter).
Fab 2R was less potent but clearly immune precipitated antigen more efficiently
than 3F4. Note that D4, R2, 6D2, D14, R1, and R10 all refer to antibodies of the
Immunoprecipitation of SHaPrP 27-30 with Recombinant Fabs
The ability of 3F4 diluted 1/500 and 100 µl ofE. Coli extracts
containing Fab to immunoprecipitate SHaPrP 27-30 was monitored by western blotting.
All lanes except lane 14 are from immunoprecipitations containing goat anti-mouse
IgG Fab and protein A agarose. 10 µl of liposomes containing SHa PrP 27-30 were
added to lanes 1, 3, 5, 7, 9, 11, 13. 100 µl of E. Coil extracts from
different clones diluted 1/500 were added as follows: lanes 2-3, 6D2; lanes 4-5,
D14; lanes 6-7, R1; lanes 8-9, R10; lanes 10-11, D4; lanes 12-13, 3F4. Lane 14 was
loaded with S volume of liposomes used for immunoprecipitations.
The results described above are shown within the photograph of Figure
11. The photo clearly shows higher degrees of immunoprecipitation when using the
recombinant antibodies of the invention.
Figure 12 is a photo showing the immunoprecipitation of SHaPrP 27-30
with purified Fabs of the invention (2R and 4D) as well as 3H4. The ability to immunoprecipitate
the antigen is monitored by western blotting. All of the lanes shown in Figure 12
but for lane 14 are immunoprecipitations containing goat anti-mouse IgG Fab and
protein Agarose. To obtain the results 10 µl of liposomes containing SHaPrP 27-30
were added to all lanes except for lanes 5, 9 and 13. Each of the lanes are marked
with the indicated amounts of purified Fabs (nanograms) which were added to lanes
2-13. Lane 14 was loaded with one-half volume of liposomes used for the immunoprecipitation.
The results clearly show a dramatically higher degree of precipitation when using
the antibodies 2R and 4D of the invention as compared to 3F4.
The ELISA data (Figure 9) clearly show a number of Fabs with a saturable
binding to non-denatured PrP 27-30 and a half-maximal binding at around 0.5 µg/ml.
This corresponds to an apparent affinity constant at 108 M-1
(MW of Fab = 50,000). At the same time, 3F4 shows insignificant binding out to 2
µg/ml. Moving to denatured PrP 27-30, Figure 10, the recombinant Fabs now bind to
a higher level but with a similar apparent affinity. This suggests denaturation
has revealed more antigenic sites but their affinities are the same. Significantly,
3F4 is now binding comparably to the recombinant Fabs with an apparent affinity
of the order of 108 M-1. Comparison of the 3F4 data in Figures
9 and 10 strongly suggests the integrity of Prp 27-30 in the non-denatured form.
Thus it could have been argued that the recombinant Fabs were reacting with a fraction
of denatured PrP present in the PrP 27-30 preparation. The lack of reactivity of
3F4 with non-denatured PrP 27-30 coupled with its strong reactivity with denatured
PrP 27-30 refutes this interpretation and strongly suggests the recombinant Fabs
recognize non-denatured rods with high affinity.
The immunoprecipitation data are confirmatory of the ELISA data. Low
concentrations of recombinant Fabs as found in crude bacterial supernates (typically
1-10 µl/ml) are highly effective at immunoprecipitating PrP 27-30 (Figure 11). This
implies an affinity on the order of 107-108 M-1.
Under comparable concentration conditions, 3F4 does not produce significant precipitation.
A more quantitative analysis (Figure 12) shows that Fab R2 immunoprecipitates PrP
27-30 highly effectively with some titration in the range 0.1-0.2 µg/ml implying
a binding affinity on the order of 108 M-1. Fab 4D has a lower
affinity and 3F4 immunoprecipitates very weakly indeed. From this particular experiment
one could argue that the affinity of 3F4 is considerably less than 5 x 107
M-1 and probably less than107 M-1.
Overall, the data indicates that the recombinant Fabs have affinities
in the range of 107-108 M-1.
The instant invention is shown and described herein in what is considered
to be a most practical and preferred embodiments. It is recognized, however, that
departures may be made from which are within the scope of the invention and that
modifications will occur to one who is skilled in the art upon reading this disclosure.
Use of an antibody, characterized by its ability to bind to native PrPScin
situ with a binding affinity Ka of 107 l/mol or more,
in an assay for detection of PrPSc.
A method, comprising:
contacting a material with an antibody characterized by its ability to
bind to native PrPScin situ with a binding affinity of Ka
of 107 l/mol or more; and
determining whether the antibody binds specifically to PrPSc in the
The use of claim 1 or the method of claim 2, wherein the antibody binds to native
PrPScin situ with a binding affinity Ka of 108
l/mol or more.
The use of claim 1 or the method of claim 2, wherein the antibody specifically
binds to PrPSc of a mammal chosen from a human, a cow, a sheep, a horse,
a pig, a dog, a chicken and a cat.
The use of claim 1 or the method of claim 2, wherein the antibody is bound to
a detectable label.
The use of claim 1 or the method of claim 2, wherein the antibody ischaracterized
by an ability to bind to 50% or more of PrPSc in a liquid flowable
The use of claim 1 or the method of claim 2, wherein a plurality of different
antibodies are contacted with the material and the antibodies have a Ka
of 107 l/mol or more relative to PrPSc.