Note: Descriptions are shown in the official language in which they were submitted.
. .,, ~- ` t -' 21 631 25
W094t28026 ~ PCT~S94/06090
GENETICALLY ENGINEERED IMMUNOGLOBULINS
Field of the Invention
We have developed a new method for the presentation of
immunogenic epitopes to cytotoxic T lymphocytes (CTL). The
new method is based on antibody antigenization, a process
whereby one or several loops of an immunoglobulin molecule
are re-engineered to encompass the sequence of selected
portions of pathogens (virus and parasites), self antigens
and tumor antigens.
The present invention may utilize in its preferred
embodiments, the use of recombinant DNA technology to
genetically engineer natural or synthetically-derived
immunoglobulin molecules, imparting therein novel epitopes,
so as to create novel entities that can be employed ln
vitro and in vivo in a variety of means, such as to
immunize against pathogens, and for example, build
tolerance to antigens. In preferred embodiments, the
epitopes are inserted into the so-called heavy or light
chain variable domain of a given immunoglobulin molecule.
Thus, known recombinant DNA technologies come to bear in
the present invention, helping create novel immunoglobulin
entities that retain functionality by localizing to
particular cell types mechanistically via the so-called
constant domains but otherwise functionally exploited to
provide a novel localization of a particular antigenic
determinant or epitope.
W094/28026 ~ 2 1 ~ 3 1 2 5 ~CT~594/06090 ~
Back~round of the Invention
Recombinant DNA technology has reached the point currently
of being capable, in principle, of providing the
methodology sufficient to identify, isolate and
characterize DNA sequences, configure them for insertion
into operative expression vectors and transfect those
vectors variously into recombinant hosts such that those
hosts are harnessed in their ability to produce the
polypeptide encoded by the DNA sequence. Obviously, many
variations attend the methodology associated with
recombinant DNA technology, and particular means are not
without inventive faculty. Nonetheless, methods are
generally known in the published literature enabling
requisite mental equipment for the art skilled to practice
recombinant DNA technology in the production of
polypeptides from a given recombinant host system.
Immunoglobulins (Igs) are the main effectors of humoral
immunity, a property linked with their ability to bind
antigens of various types. In view of the myriad numbers
of antigens to a particular host organism, it can be
appreciated that there are a like number or more of
immunoglobulins that contain antigenic determinants or
epitopes against particular such antigens. Immunoglobulin
molecules are unique in their functionality of being
capable of localizing to certain cell types, probably by
means of mutual recognition of certain receptors that are
located on the cell membrane. Immunoglobulins demonstrate
a second general property whereby they act as endogenous
modulators of the immune response. Igs and their idiotypic
determinants have been used to lmml~nlze at the B- and/or T-
cell level against a variety of exogenous antigens. In
many cases, the immunity they evoke is comparable with that
induced by the antigen itself. Although the principle
underlying this phenomenon is understood, little is known
about the molecular basis and the minimal structural
~ W094/28~6 ~ 2163125 PC1~594/06090
requirements for the immunogenicity of Igs molecules and
the interaction between those regions which may be
responsible for such immunogenicity and the regions that
are thought to provide the localization of a given
immunoglobulin molecule with a particular cell/receptor
type.
In the last many years, much progress has been made in
endeavors to understand the immunogenic properties,
structure and genetics of immunoglobulins. See Jeske, et
al., Fundamental Immunoloqy, Paul, ed., Raven Press, New
York (1984), p. 131 and Kabat, Journal Immunoloqy 141:525
(1988). Initially, the antigenicity of the so-called
variable (V) domain of antibodies was demonstrated. Oudin,
et al., Academy of Sciences D 257:805 (1963) and Kunkel et
al., Science 140:1218 (1963). Subsequently, further
research pointed out the existence of discrete areas of
variability within V regions and introduced the notion of
hypervariable (HV) or complementarily-determining regions
(CDR). Wu, et al., J. Exp. Med. 132:211 (1970). Many
studies since have indicated that the immunogenic property
of Ig molecules is determined presumably primarily by amino
acid sequence contained in the CDRs. Davie, et al., Ann.
Rev. Immunol. 4:147 (1986).
The basic immunoglobulin or antibody structural unit is
well understood. The molecule consists of heavy and light
chains held together covalently through disulfide bonds.
The heavy chains are also covalently linked in a base
portion via disulfide bonds and this portion is often
referred to as the so-called constant region which is
thought responsible for a given immunoglobulin molecule
being mutually recognizable with certain sequences found at
the surface of particular cells. There are five known
major classes of constant regions which determine the class
of the immunoglobulin molecule and are referred to as IgG,
IgM, IgA, IgD and IgE. The N-terminal regions of the so-
2 1 6 3 1 2 5
W094/28026 PCT~S94/06090
called heavy chains branch outwardly in a pictorial sense
so as to give an overall Y-shaped structure. The light
chains covalently bind to the Y branches of the two heavy
chains. In the regions of the Y branches of the heavy
chains lies a domain of approximately 100 amino acids in
length which is variable, and therefore, specific for
particular antigenic epitopes incidental to that particular
immunoglobulin molecule.
It is to the Y branches containing the variable domains
harboring the antigenic epitopes to which the particular
attention is directed as a predicate of the present
invention.
Prior researchers have studied and manipulated entire CDRs
o~ immunoglobulins, producing chimeric molecules that have
reported functionality. Exemplary attention is directed to
Jones, et al., Nature 321:522 (1986) reporting on a V-
region mouse-human chimeric immunoglobulin molecule. This
research thus amounted to a substantially entire CDR
replacement as apparently does the research reported by
Verhoeyen, et al., Science 239:1534 (1988); Riechmann, et
al., Nature 332:323 (1988); and by Morrison, 5cience
229: 1202 (1985) . See also European Patent Application
Publication No. 125023A, published 14 November 1984.
Bolstered by the successful research summarized above that
resulted presumably in functional chimeric molecules, the
goal of the present research was to explore further the
variable region contained in the N-terminus Y branches. It
was a goal of the present research to manipulate these
variable regions by introduction or substitution of novel
determinants or epitopes so as to create novel
immunoglobulin molecules that would possibly retain the
localization functionality and yet contain functional
heterologous epitopes. In this manner, the novel
immunoglobulin molecules hereof could be employed ~or use
c ~ 2 1 6 3 1 2 5
W094/~8026 PCT~S94/06090
within the organism at foreign sites, thereby imparting
immunity characteristics in a novel site-directed manner.
A problem facing the present researchers at that time lay
in the fact that epitopes are found in a region of the Y
branch. Therefore, it was difficult to envision whether
any manipulation of the variable region would be possible
without disrupting the interaction of heavy chain with the
corresponding light chain, and if that proved
inconsequential, whether the resultant molecule would
retain its functionality, with respect to the novel
epitope, in combination with the constant region of the
basic immunoglobulin molecule. Thus, even hurdling the
problem of where to experiment, it was not possible to
predict whether one could successfully produce such novel,
bifunctional immunoglobulin molecules.
Recognition of antigen peptides by T lymphocytes is
restricted by the major histocompatibility complex (MHC)
gene products and is mediated by the T-cell receptor (TCR)
recognition structure.
CD4+ T helper lymphocytes recognize antigen peptides
presented in the context o~ class I I MHC molecules (unanue,
Curr. OPin. Immunol. 4:63 (1992)) while CD8+ cytotoxic T
lymphocytes (CTL) require that antigen peptides be
presented by class I MHC molecules (Braciale, Curr. Opin.
Immunol. 4 :59 (1992) ) . In both instances, antigens need to
be processed into small peptides, 9 residues for peptides
that bind class I molecules ~Rotzschke et al., Nature
348:252 (1990)) and 13-17 residues for peptides that bind
class II molecules (Chicz et al., Nature 358:764 (1992);
Rudensky et al., Nature 353:622 (1991)). Antigens
presented to CD4' T cells derive primarily ~rom exogenous
proteins that are processed into peptides that bind class
II MHC molecules in the endocytic compartment (Unanue,
Curr. Opin. Immunol. 4 :63 (1992)), albeit proteins from the
cytosol can also bind class II MHC molecules (Malnati et
xr~ ~ ~ 6 3 ~ 2 5
W094l28026 PCT~S9~/06090
al., Nature 357:702 (1992)). Antigens presented to CD8+ T
cells derive mostly from proteins processed in the cytosol
and bind to class I MHC molecules in the endoplasmic
reticulum (ER) (Braciale, Curr. Opin. Immunol. 4:59
(1992)). Prototype antigens for this type of presentation
are viral proteins generated by intracellular replication
of an infectious virus (Long et al., Immunol. Today 10:45
(1989)).
While cytosolic proteins are fragmented into peptides and
translocated across the ER membrane by proteasomes,
intracellular polypeptides that map to the class II MHC
complex (Glynne et al., Nature 353:357 (1991)), endogenous
proteins destined for secretion are synthesized by
ribosomes attached to the rough ER. In both instances
peptides from endogenous proteins complex with class I MHC
molecules in a pre-Golgi compartment.
B lymphocytes are specialized antigen presenting cells
(APC) (Lanzavecchia, Immunol. Today 10:157 (1989)) that
express both class II and class I MHC molecules. As such,
B cells constitutively present immunoglobulin peptides
within the MHC molecules expressed at their surface in the
context of class II molecules (Rudensky et al., Nature
359:429 (1992). However, there exists only indirect
evidence that endogenous Ig peptides are presented in class
I MHC molecules (Shinohara et al., Nature 336:481 (1988);
Weiss et al., Cell 64:767 (1991); Yamamoto et al., Eur. J.
Immunol. 17:719 (1987)) and no endogenous Ig peptide has
beer isolated from B cells that mediates CTL function.
It was predicted that the translated polypeptide encoded by
antigenized antibody gene would follow the secretory
pathway. There is large indirect evidence in favor of this
assumption. First, is the demonstration that unassembled
immunoglobulin ~ chains accumulate in the ER where they
undergo processing (Sitia et al., 1990). In the absence of
W094/2~026 ~ .3l ~ 2 1 6 3 1 25 PCT1'594/06090
any L chain, the H chain is retained in the ER bound to the
H chain-binding protein (BiP) (Bole et al., 1986). By
analogy, processing of the ~2 315 L chain and generation of
an idiotype peptide presented in class II molecules was
shown to occur in the ER (Weiss et al., 1989). Recently,
it was demonstrated that the ER possess an enzyme necessary
to cleave the leader sequence from a nascent class I
molecule.
One important issue in current immunology is to be able to
program the immune system towards preselected T-cell
epitopes, whether these be restricted by class I or class
II MHC molecules. This is significant in defensive immune
responses against pathogens, and, in particular,
intracellular pathogens for eliciting CTL specific for
protective epitopes. A CTL response can be achieved
through a number of strategies like immunization with
synthetic peptides (Aichele et al., 1990), recombinant
proteins (Kleid et al., 1981), vaccinia virus constructs
(Mackett et al., 1985), soluble proteins osmotically
vehicled to the cytosol (Carbone et al., 1990) and cells
pulsed in vi tro with synthetic peptides (external loading)
(De Bruijn et al., 1991). While synthetic peptides are
scarcely immunogenic and vaccinia vectors have drawbacks in
previously vaccinated individuals (Lane et al., 1968),
external loading of peptide provides a limited availability
of empty class I MHC molecules at the cell surface.
The present research and invention is based upon the
successful threshold experiment, producing model, novel
immunoglobulin molecules found to be fully functional by
virtue of their ability to localize on certain
cell/receptor sites and elicit reactivity to the antigens
spec fic for the introduced novel antigenic determinant or
epitope. This invention demonstrates a new method for the
engineering of cellular vaccines that can be used for the
in vivo or in vi tro induction of CTL.
W094/~026 ' ~ 2 1 6 3 1 2 5 PCT~S9~/06090 ~
A plasma cell can secrete about 103 molecules of Ig
cell/sec, Ig can be an extraordinary source of endogenous
peptides and B cells efficient APCs for presentation of
peptide epitopes in the context of class I MHC molecules.
In this invention, it is demonstrated that one can use B
cells as APC to process and present a peptide from an
endogenous Ig heavy (H) chain to a class I MHC restricted
CTL clone specific for Ig peptide.
Sl ~y of the Invention
The present invention is based upon the successful
production of novel immunoglobulin molecules having
introduced into the N-terminus variable region thereof a
novel epitope not ordinarily found in the immunoglobulin
molecule used as a starting molecule.
Successful model systems of the foreign molecules include
the hydrophilic tetrapeptide Asn-Ala-Asn-Pro (NANP) of
Plasmodium falciparum circumsporozite protein, the
tripeptide Arg-Gly-Asp (RGD) involved in the interaction of
a variety of adhesive proteins, and oligopeptide epitopes
of the human CD4 HIV binding domain.
More particularly, this invention relates to introduction
of oligopeptide epitopes of a nucleoprotein (NP) peptide of
influenza virus for expression within the third fold of an
immunoglobulin molecule. This virus peptide is recognized
by CTL in the context of the H-2 Db allele. It is shown
that the NP peptide engineered in the H chain: 1) mediates
killing of B cell lymphomas by a CTL clone specific for
that peptide restricted by the Db molecule, and 2) could be
purified from the H-2 Db molecules at the cell surface.
This study formally demonstrates that peptides from the
hypervariable loops of Ig are presented by class I MHC
molecules and validates a role for the processing and
presentation of self immunoglobulin V regions to CD8+ T
cells in the regulation of the immune response.
~ W094/23026 ~ 2 1 6 3 1 2 5 PCT~S94/06090
This invention demonstrates the possibility to program
class I-restricted presentation of intracellular pathogens
peptides using antigenized antibody genes as non-infectious
"replicating" material .
The present invention is thus directed to novel
immunoglobulin molecules having at least one novel
heterologous epitope contained within the N-terminus
variable domain thereof, said novel immunoglobulin molecule
having retained functionality with respect to its C-
terminus constant domain of the heavy chain specific for aparticular cell/receptor type, and having novel, specific
epitope in vitro and in vivo reactivity.
The present invention is further directed to pharmaceutical
compositions containing, as essential pharmaceutical
principal, a novel immunoglobulin hereof, particularly
those in the form of an administrable pharmaceutical
vaccine.
The present invention is further directed to methods useful
~or building tolerance to certain antigens, including those
associated with autoimmune diseases, or for down-regulating
hypersensitivity to allergens, or for providing active or
passive immunity against certain pathogenic antigens, by
administering to an individual in perceived need of such,
a novel immunoglobulin molecule as defined above.
The present invention is further directed to novel
recombinant means and methods useful for preparing,
identifying and using the novel immunoglobulin molecules
hereof including DNA isolates encoding them, vectors
operatively harboring such DNA, hosts transfected with such
vectors, cultures containing such growing hosts and the
methods useful for preparing all of the above recombinant
aspects.
W094/28026 w . ~ 2 1 ~ 3 1 2 5 PCT~S94/06090 ~
Detailed Description of the Invention
The present invention is described herein with particular
detail for the preparation of model, novel immunoglobulin
entities. This description is provided, as it was
conducted, using recombinant DNA technology. Further
detail herein defines methods by which one can test a given
immunoglobulin to assure that it exhibits requisite
functionality common to its starting material
immunoglobulin and specially as to its novel epitopic
antigenic activity. Given this information with respect to
the particular novel immunoglobulin molecules described
herein, coupled with general procedures and techniques
known in the art, the art skilled will well enough know how
to configure recombinant expression vectors for the
preparation of other novel immunoglobulin molecules falling
within the general scope hereof for use as herein
described. Thus, having described the threshold experiment
of the successful preparation of a novel immunoglobulin
molecule, one skilled in the art need not follow the exact
details used for reproducing the invention. Instead, the
art skilled may borrow from the extant, relevant art, known
techniques for the preparation of still other novel
immunoglobulin molecules falling within the general scope
hereof.
1. Fiqure Leqends
Figure 1 is a diagram illustrating the construction of the
pN~lNANP expression vector.
Figure 2 is an SDS-PAGE of the ~lNANP and WT recombinant
Ig.
Figure 3 shows the binding of 12;I-labelled monoclonal
antibody Sp3-B4 to engineered antibody ~lNANP.
~ W094/28026 ~ y~,~s~ ~; 2 1 63 1 25 PCT~'S94/06090
1~ .
Figure 4 is a Western blot binding of 12~-labelled antibody
Sp3-B4 to engineered antibody rlNANP and localization of
the engineered (NANP) 3 epitope in the H chain.
.
Figure 5 shows results of cross-inhibition of l25I-labelled
antibody Sp3-84 binding to synthetic peptide (NANP) (panel
A) or engineered antibody ~lNANP (panel 8) by ~lNANP Ig or
peptide (NANP) 3 .
Figure 6 is a diagram of pN~lNP expression vector and
general strategy of transfection.
Figure 7 depicts specificity of target cell recognition by
cold target competition. The inset shows the dose response
of killing of B6-2 HNP transfectants. 5lCr-labeled B6-2.503
cells (2.5 x 105 cells/ml) were mixed with CTL clone 34
cells at an E:T ratio of 10:1, 1:1, 0.1:1 or 0:1. Percent
cytotoxicity was calculated 4 hours later from triplicate
cultures as described.
Figure 8 shows lack of interference of soluble ~lNP with
external loading of peptide and lysis of target cells by a
CTL clone.
Figure 9 shows inhibition of lysis of B6-2 HNPtransfectants
by a monoclonal antibody to Db.
Figure 10 shows presentation of processed NP peptide by
engineered cells is restricted by H-2b.
Figure 11 shows elution of influenza virus NP peptide from
B6-2 HNP tranfectants. (a) HPLC profile of the synthetic
peptide ASNENMETM of influenza virus (100 ~g). (b)
Cytotoxic assay using single HPLC fractions from the
- experiment shown in a. (c) HPLC profile of the peptide
mixture eluted from the Db molecules purified from B6-2 HNP
transfectants. (d) Cytotoxic assay using single HPLC
-
W094/28026 ~ 2 1 6 3 1 2 5 PCT~S94/06090 ~
12
fractions from the experiment shown in c.
Figure 12 shows that addition of exogenous peptide does not
increase lysis of cells engineered with the HNP gene.
2. General Methods and Definitions
"Expression vector" includes vectors which are capable of
expressing DNA sequences contained therein, where such
sequences are operatively linked to other sequences capable
of effecting their expression. It is implied, although not
always explicitly stated, that these expression vectors may
be replicable in the host organisms either as episomes or
as an integral part of the chromosomal DNA. "Operative,"
or grammatical equivalents, means that the respective DNA
sequences are operational, that is, work for their intended
purposes. In sum, "expression vector" is given a
functional definition, and any DNA sequence which is
capable of effecting expression of a specified DNA sequence
disposed therein is included in this term as it is applied
to the specified sequence. In general, expression vectors
of utility in recombinant DNA techniques are often in the
form of ~plasmids" referred to as circular double stranded
DNA loops which, in their vector form, are not bound to the
chromosome. In the present specification, "plasmid" and
"vector" are used interchangeably as the plasmid is the
most commonly used form of vector. However, the invention
is intended to include such other forms of expression
vectors which serve equivalent functions and which become
known in the art subsequently hereto.
Apart from the novelty of the present invention involving
the introduction of novel epitopes by means of
repositioning or augmentation of a parent immunoglobulin,
it will be understood that the novel immunoglobulins of the
present invention may otherwise permissively differ from
the parent in respect of a difference in one or more amino
acids from the parent entity, insofar as such differences
~ W094l28026 ~ 2 1 6 3 1 2 5 PCT~S94/06090
13
do not lead to a destruction in kind of the basic activity
or bio-functionality of the novel entity.
"Recombinant host cells" refers to cells which have been
transfected with vectors defined above.
Extrinsic support medium is used to support the host cells
and includes those known or devised media that can support
the cells in a growth phase or maintain them in a viable
state such that they can perform their recombinantly
harnessed function. See, for example, ATCC Media Handbook,
Ed. Cote et al., American Type Culture Collection,
Rockville, MD (1984). A growth supporting medium for
m~mm~l ian cells, for example, preferably contains a serum
supplement such as fetal calf serum or other supplementing
component commonly used to facilitate cell growth and
division such as hydrolysates of animal meat or milk,
tissue or organ extracts, macerated clots or their
extracts, and so forth. Other suitable medium components
include, for example, transferrin, insulin and various
metals.
The vectors and methods disclosed herein are suitable for
use in host cells over a wide range of prokaryotic and
eukaryotic organisms.
"Heterologous" with reference herein to the novel epitope
for a given immunoglobulin molecule refers to the presence
of (at least one) such epitope in the N-terminus domain of
an immunoglobulin that does not ordinarily bear that
epitope(s) in its native state. Hence, that chain contains
heterologous epitope sequence(s). Such heterologous
epitope sequences shall include the classic antigenic
epitopes as well as receptor like binding domains or
binding regions that function as receptor sites, such as
the human CD4 binding domain for HIV, hormonal receptor
binding ligands, retinoid receptor binding ligands and
W094l28026 ~ '` ~ 2 i 6 3 i 2 5 PCT~S94/06090
14
ligands or receptors that mediate cell adhesion.
"Chimeric" refers to immunoglobulins hereof, bearing the
heterologous epitope (s), that otherwise may be composed of
parts taken from immunoglobulins of more than one species.
Hence, a chimeric starting immunoglobulin hereof may have
a hybrid heavy chain made up of parts taken from
corresponding human and non-human immunoglobulins.
In addition to the above discussion and the various
references to existing literature teachings, reference is
made to standard textbooks of molecular biology that
contain definitions and methods and means for carrying out
basic techniques encompassed by the present invention.
See, for example, Maniatis, et al, Molecular Cloninq: A
LaboratorY Manual, Cold Spring Harbor Laboratory, New York,
1982 and the various references cited therein, and in
particular, Colowick et al., Methods in Enzvmoloqy Vol.
152, Academic Press, Inc. (1987). All of the herein cited
publications are by this reference hereby expressly
incorporated herein.
The foregoing description and following experimental
details set forth the methodoloqy employed initially by the
present researchers in identifying and characterizing and
preparing particular immunoglobulins. The art skilled will
recognize that by supplying the present information
including the wherewithal of the location and makeup of the
epitope containing domain of a given immunoglobulin, and
how it can be manipulated to produce the novel
immunoglobulins hereof. Therefore, it may not be necessary
to repeat these details in all respects in their endeavors
to reproduce this work. Instead, they may choose to employ
alternative, reliable and known methods, for example, they
may synthesize the underlying DNA sequences encoding a
particular novel immunoglobulin hereof for deployment
within similar or other suitable, operative expression
~ W094/28026 ~ ~ 2 1 63 1 25 PCT~S94/06090
vectors and culture systems. Thus, in addition to
supplying details actually employed, the present disclosure
serves to enable reproduction of the specific
immunoglobulins disclosed and others, and fragments
thereof, such as the individual chains for in vitro
assembly, using means within the skill of the art having
benefit o~ the present disclosure. All of such means are
included within the enablement and scope of the present
invention.
3. DescriPtion of ParticularlY Preferred Embodiments
Protein engineering was used to introduce a foreign epitope
into the CDR3 of the H chain of a mouse/human chimeric
antibody (C~162). This epitope consists of three copies of
the tetrapeptide Asn-Ala-Asn-Pro (NANP). The tetrapeptide
occurs naturally as a 37 tandem repeat in the Plasmodium
falciparum circumsporozoite (CS) protein, interspersed with
four repeats of the variant sequence Asn-Val-Asp-Pro [Dame
et al., Science 229:593 (1984)]. In the construct
described here, the epitope is flanked by Val and Pro
residues at each end [VP (NANP) 3 VP] . The experiment
verified that the (NANP) 3 epitope could be inserted in the
NV region of a host H chain (VH) without altering the
framework folding of the Ig molecule, i.e., its molecular
assembly with the light (L) chain and it determined that
the antigenic and immunogenic properties of the recombinant
Ig molecule were expressed. It is known that the CDR3 of
VH regions of antibody is often the structural correlate of
an immunodominant idiotope [Davie, et al., Ann. Rev.
Immunol. 4:147 (1986)], which indicates that the CDR3 is at
the surface of the molecule. Moreover, it is well
established that because of recombination of the variable-
diversity-joining (VDJ) regions, as well as N-addition
mechanisms [Tonegawa, Nature 302:575 (1983); Miller et al.,
Immunol. TodaY 7:36 (1986)], the CDR3 may vary considerably
in length (from 3 to 19 amino acids) [Kabat, et al.,
W094/28026 ~ .~ t ~ 2 ~ 63 1 25 PCT~TSg~06090 ~
16
Proteins of Immunoloqical Interest, U.S. Dept. of Health
and Human Service NIH (1987)], implying a high degree of
plasticity at the structural level. Second, the (NANP) 3
epitope selected for this study is relatively short,
repetitive and of proven immunogenicity in mice and humans
[Good et al., Ann. Rev. Immunol. 6:663 (1988)l.
Particularly developed is a new method for the presentation
of immunogenic epitopes to cytotoxic T lymphocytes (CTL).
The new method is based on antibody antigenization, a
process whereby one or several loops of an immunoglobulin
molecule are re-engineered to encompass the sequence of
selected portion of pathogens (virus and parasites), self
antigens and tumor antigens.
Antibodies may be antigenized by inserting immunogenic
epitopes in any of the three CDR regions of each heavy
chain and any of the three CDR regions of the light chain.
A preferred site of engineering an immunogenic epitope is
the third CDR region of the heavy chain.
An immunogenic epitope may be inserted into one or more of
the 9iX CDRs, thus generating an antibody antigenized with
between one and six epitopes. In a preferred embodiment,
one immunogenic oligopeptide sequence is engineered within
the third complementarity-determining region (CDR3) of the
heavy chain of the immunoglobulin.
Immunogenic epitopes may be engineered within any or all of
the CDRs by inserting a nucleic acid sequence encoding the
epitope at a site unique to the CDR and absent from the
nucleic acid sequence of the immunoglobulin chain wherein
the epitope sequence is to be inserted. Insertion may be
accomplished, for example, using a restriction enzyme
capable of recognizing the unique sequence in the CDR.
- 094/2~026 ~; `~i~i`` 2 1 6 3 1 2 5 PCT~S94/06090
Mature lymphocytes of the CD8 phenotype recognize antigen
in conjunction with class I MHC molecules. The best
studied systems relate to CTL that recognize virally-
infected cells (Long et al., Immunol. Today 10:45, 1989).
CTL's function requires active replication of the virion
within the cell. Few examples do however, exist to
indicate that inactivated (non-replicating) virus (Wraith
et al., J. Gen. Virol. 66:1327, 1985) or soluble proteins
(Moore et al., Cell 54:777 (1988); Staerz et al., Nature
329:449 (1987)) can also be presented to class I-restricted
T cells provided that they reach the inside of the cell.
To further demonstrate the invention, an antibody was
engineered to encompass the oligopeptide sequence
ASNENMETMESSTL representing a CTL epitope of influenza
virus nucleoprotein (NP) (Bastin et al., J. Exp. Med.
165:1508, 1987; Townsend et al., Cell 44:959 (1986)). This
epitope has been characterized as a short nonglycosylated
protein sequence which is recognized as a target by NP-
specific CTL clones in a MHC-restricted way. More
recently, it has been directly proven that this peptide is
indeed responsible for targeting CTL on in~luenza virus-
infected cells as it could be eluted from the MHC class I
molecule (Rotzschke et al., Nature 348:2 (199O)).
4. Examples
Exam~le I
A. Construction of the pN~lNANP expression vector
The production of hybridoma 62 and BlOH2, and the
purification of mAb 62 and 109.3 (anti-2,4-dinitrophenol)
have been described previously [Zanetti et al., J. Immunol.
131:2452 (1983) and Glotz et al., J. Immunol. 137:223
(1986)].
A DNA library was constructed from size-selected 2-2.5-kb
Eco RI fragments from hybridoma 62 genomic DNA. Fragmen~s
were eluted from low melting point agarose and ligated into
W094/28026 ~ 2 1 ~ 3 1 2~ PCT~S9~/06090
18
the AgtlO vector [Huynh et al., DNA Cloninq Techniques 1:49
(1985)]. After ligation and packaging, 5 x 104 plaque-
forming units were screened by replicate hybridization with
the JH [Sakano et al., Nature 286:676 (1980)] and pSAPC15
[Brodeur et al., Eur. J. Immunol. 14:922 (1984)] probes.
Four clones were isolated and plaque purified; the 2.3-kb
EcoRI insert form one of them was subcloned into pEMBL18
vector [Dente et al., DNA Clonina Techniques l:101 (1985)].
The VHBlOH2 coding se~uence was determined by cloning the
cDNA from the parental hybridoma by primer extension of the
poly(A)+ RNA with a synthetic oligonucleotide
(5'-GGGGCCAGTGGATAGAC-3') that anneals at the 5' end of the
CHl region. The same oligonucleotide was used as a probe
for screening the library after 5' end-labeling by kinase
with 32P-ATP. The nucleotide sequence of both clones was
determined by dideoxy method on both strands after
subcloning suitable restriction fragments into the pEMBL18
vector.
Plasmid pNr162 containing DNA encoding Cl, 62 antibody was
constructed by subcloning in the proper orientation the
2.3-kb EcoRI DNA fragment carrying the VH62 rearrangement
into the unique EcoRI site of the PN~1 vector [Sollazo et
al., Focus 10: 64 (1988)] (a PSV derived vector harboring an
human ~1, gene). This vector encodes a human r1 gene
downstream from the EcoRI site. It also carries a neomycin
resistance gene under the control of the SV40 promoter for
the selection of stable transformant cells. Transfectoma
cells were constructed by introducing the plasmids pN~162
and pN~lCHA, a chimeric construct encoding an antibody
lacking Id62 and Ig binding into J558L mouse by
electroporation. This cell line is an H chain-defective
variant of myeloma J558 [Morrison et al., Science 229:229
(1985)] and carries the rearrangement for a A1 light (L)
chain. Briefly 3 x lC6 cells in 1 ml of Dulbecco's
modified minimum essential medium (DMEM) containing 10 ~g
of supercoiled plasmid DNA were pulsed for 17 ms at 650
094/28026 ~ 2 1 6 3 1 2 5 PCT~S94/06090
19
V/cm in a Cell Porator apparatus (Bethesda Research
Laboratories, Bethesda, MD). After pulsing, the cells were
resuspended in 10 ml of DMEM supplemented with 10 mM Hepes
buffer, 2 mM L-glutamine, penicillin (50 ~g/ml),
streptomycin (50 ~g/ml) and 10~ fetal calf serum (cDMEM),
and incubated for 48 h at 37OC in a 10~ CO2 atmosphere.
The cells were then resuspended in 20 ml of cDMEM and an
aliquot (2 ml) was diluted into 20 ml of cDMEM containing
1.2 mg/ml of G418 (Gibco, Grand Island, NY), plated on a
96-well microtiter plate and cultured for 14 days. The
supernatants of neomycin-resistant colonies (stable
transformants) were tested by solid-phase radioimmunoassay
(RIA) and enzyme-linked immunosorbent assay (ELISA).
The presence of Id62 in the supernatant of J558L cells
transfected with pN~l62 vector was tested by competitive
inhibition in ELISA. This measures the inhibition
(percent) of the binding of horseradish peroxidase (HP)-
conjugated mAb 62 (ligand) to anti-Id62 antibody coated on
96-well polyvinyl microtiter plate (Dynatech, Alexandria,
VA) [Zanetti et al., J. Immunol. 131:2452 (1983)]. The
supernatant of J558L cells transfected with pN~lCHA plasmid
and purified mAb 62 and 109.3 (an IgGl, x anti-2,4-
dinitrophenol) served as controls [Zanetti et al., J.
Immunol. 131:2452 (1983)]. A second method to test for
Id62 expression was by Western blot [Towbin et al., Proc.
Natl. Acad. Sci. USA 71:4350 (1979)]. Briefly,
approximately 5 ~g of antibody C~l62 purified by affinity
chromatography on an anti-human Ig Sepharose 4B column
(Pharmacia, Uppsala, Sweden) was electrophoresed on a 10~
sodium dodecyl sulfate polyacrylamide gel electrophoresed
on a 10~ sodium dodecyl sulfate polyacrylamide gel
electrophoresis under reducing conditions. The gel was
then blotted onto 0.45 M nitrocellulose paper (Millipore,
- Bedford, MA) and probed withl2~I-labelled affinity-purified
syngeneic anti-Id62 antibody [Zanetti et al. J. Immunol.
135:1245 (1985)]. Antibodies 62 and 109.3 served as
W094/28026 i~ T f~ 2 ~ 63 1 2~ PCT~S91/06090
positive and negative control, respectively. The filter
was exposed a first time for 24 h at -70C with intensifier
screen. To demonstrate the co-expression of the human C
region on the H chain of the chimeric C~162 antibody, the
nitrocellulose paper was re-probed with l25I-labelled goat
anti-human Ig antibody and exposed for 2 h at 70C.
Sequence data is publicly available from EMBL/Gene Bank
Data Library under Accession No. Y00744.
The rlNANP antibody carrying the malarial CS immunodominant
B-cell epitope NANP in the CDR3 of its H chain was
engineered as follows:
Figure 1 is a diagram illustrating the construction of the
pN~lNANP expression vector. In panel A: (a) The
productively rearranged VH gene of the hybridoma cell line
62 isolated from a size-selected lambda gtlO library and
subcloned into pBluescript (publicly available from
Stratagene, San Diego, CA) is described infra.; (b) The
restriction site KpnI/Asp718 of the polylinker region was
deleted by Kpn I digestion, filled in with T4 polymerase
and ligated, yielding the plasmid pH62~k; (c) pH62~k was
used as a template for site-directed mutagenesis to
introduce a unique Asp718 restriction site in CDR3 of the
VH gene. The synthetic oligonucleotide
(5-'CAAGAAAGGTACCCTACTCTC-3'), which encodes a 3 bp
insertion (TAC), was annealed to the uracylated single-
stranded complementary template and elongated; (d)
Complementary synthetic oligonucleotides
(5'-GTACCCAATGCAAACCCAAATGCAAACCCA~ATGCAAACCCA-3'
3'- GGTTACGTTTGGGTTTACGTTTGGGTTTACGTTTGGGTCATG-5')
were annealed and subcloned into the unique Asp718 site
of pH62k. The construction was verified by sequence
analysis by using a 15m'r primer corresponding to the 5'
end of VH62 gene (5'-GACGTGAAGCTGGTG-3'); (e) The 2.3-
~ 094/~8026 2 1 6 3 1 2 5 PCT~S94/06090
21
kb Eco RI fragment carrying the engineered VHNANP genewas subcloned upstream from the human yl C region into
the 15-kb pNyl vector. The pN~lNANP construct was
electroporated into J558L cells subsequently cultured
in the presence of G418. Resistant clones were
- screened for Ig production by a sandwich enzyme-linked
immunosorbent assay (ELISA) using goat anti-human
antibodies immobilized on microtiter wells as the
capturing antibodies and horseradish peroxidase (HP)
conjugated goat anti-human Ig (Sigma) as the revealing
antibodies. Clones producing >2-5 ~g Ig/ml of protein
106 cells were expanded and the antibody purified from
culture supernatants. Sequence modifications
illustrated in panel A are shown in detail in panel B.
Abbreviations used: Asp - Asp 718; B - BamHI; RI -
EcoRI; FR - framework region; CDR - complementarily-
determining region; neo - neomycin (G418) resistance;
amp - ampicillin resistance.
The restriction fragment encoding the VH gene of a
murine monoclonal antibody to thyroglobulin (mAb 62)
was modified as shown in Figure l. A double-stranded
synthetic DNA fragment encoding three copies of the
NANP tetramer (NANP) 3 and carrying Asp718 protruding
ends was inserted in frame between Pro 95 and Tyr 96 of
VH62k coding region. The pH62NANP construct was
verified by dideoxy sequencing. The Eco RI restriction
fragment encoding the engineered Vg was subcloned into
the pNYl expression vector upstream from the human yl
constant (C) region to obtain the pNr1NANP construct.
This plasmid was electroporated into the murine J558L
cell line, a H chain-defective variant of myeloma J558L
that carries the rearrangement for a lambda-1 L chain
[Morrison et al., Science 229:1202 (1985)].
Transfectoma cells were cultured, subcloned and
screened for secretion of the engineered Ig molecule
~ A~ 2 ~ ~ 3 1 2 5
W094/~8026 PCT~S94/06090
22
using a sandwich enzyme-linked immunosorbent assay
(ELISA) with goat anti-human Ig antibodies. Clones
producing 2-5 ~g/ml of protein 106 cells were selected
and expanded, and the chimeric protein was purified by
means of affinity chromatography on a Sepharose 48-
Protein-A column. The purified Ig molecule was
analyzed by SDS-PAGE under reducing and nonreducing
conditions.
Figure 2 is an SDS-PAGE of the ~lNANP and WT
recombinant Ig. Five ~g of Protein A-purified antibody
were loaded on a 7.5~ polyacrylamide gel under
nonreducing conditions. The gel was stained with
Comassie blue. The inset shows the resolution into
heavy (H) and light (L) chains of engineered antibody
~lNANP electrophoresed on a 10~ polyacrylamide gel
under reducing (5~ ~-mercaptoethanol) conditions.
Figure 2 shows that the nonreduced ~lNANP chimeric
antibody has an apparent molecular weight of 160 kD,
suggesting a proper H2L2 assembly to form a tetrameric
Ig protein. When the ~lNANP antibody was compared with
the wild-type (WT) Ig, a chimeric antibody lacking the
(NANP) 3 insert, purified from culture supernatant fluid
of J558L cells transfected with pNyl62, a slight
difference in size was observed due to the presence of
the inserted epitope. However, the molecular weight of
the ~lNANP antibody is well in the range of a
tetrameric complex. Both preparations also showed a
smear in the region below the 160 kD band, suggesting
some degradation and/or noncorrectly assembled protein
products. Under reducing conditions, the engineered
~lNANP antibody was appropriately resolved into an H
and na L chain (Figure 2, inset). As determined by
ELISA of NP-40 lysates, transfectoma cells secreting
the ~lNANP antibody had approximately the same
cytoplasmic levels of H chains as cells producing the
~V~94/28026 ~ 2 1 6 3 1 2 5 PCT~S94/06090
23
WT Ig. Collectively, these results indicate that the
insertion of 15 amino acids into the CDR3 of VH62 did
not appreciably alter the interaction between VH and VL
polypeptide chains nor the assembly and secretion of
the tetrameric (H2L2) Ig molecule.
B. Bindinq of monoclonal antibody Sp-3-B4 to
enqineered antibodY ~lNANP
To determine if the engineered ~lNANP antibody indeed
expresses the (NANP) 3 epitope in an immunological
accessible form, solid-phase radioimmunoassay (RIA) and
Western blot techniques were used and a murine
monoclonal antibody (Sp3-B4) generated against P.
falciparum and specific the NANP epitope.
Figure 3 shows the binding of 125I-labelled monoclonal
antibody Sp-3-B4 to engineered antibody ~lNANP. Murine
monoclonal antibody (mAb) Sp3-B4, an IgG2a,k antibody
produced by immunization with the P. falci~arum
parasite and reacting with the repetitive epitope NANP.
Specific for the NANP epitope, any antimalainal
2 0 antibody could be so used as a tool and generated via
analogous techniques. Polyvinyl microtiter wells were
coated by drying at 37C with 5 ~g/ml solution in 0.9~
NaCI of purified ~lNANP Ig (solid diamonds), WT (solid
triangles), (NANP) 3 synthetic peptide (solid squares),
a 16mersynthetic peptide (YYCARKAYSHGMDYW) encompassing
the CDR3 of the VH region of prototype antibody 62
(open squares), and the 15mer synthetic peptide
YPQVTRGDVFTMPED of the cell-adhesive molecule
vitronectin (open diamonds). The 125I-labelled antibody
Sp3-B4 (20 X 104 cpm/50~1) was incubated overnight at
+4C. After extensive washing, the bound radioactivity
was counted in a gamma counter. The test was done in
triplicate.
W094/28026 ~ tl~ 2 ~ 6 3 ~ 25 PCT~S91/06090
24
The results of the direct RIA binding (Figure 3) showed
that 125I-labelled mAb Sp3-B4 bound both the synthetic
peptide (NANP) 3 and the recombinant ~lNANP antibody
immobilized on microtiter wells. However, the binding
to antibody ~lNANP can be considered more efficient; in
molar terms, the estimated ratio of peptide to antibody
was about 50 to 1, assuming that the antibody expresses
two copies of the (NANP~ 3 epitope per Ig molecule. No
binding occurred to either the WT Ig or two irrelevant
synthetic peptides, one corresponding to the CDR3
sequence of prototype V~62 and the other to residues
YPQVTRGDVFTMPED of vitronectin.
Figure 4 is a Western blot binding of 25I-labelled
antibody Sp3-B4 to engineered (NANP) 3 epitope in the H
chain. Ten ~g of purified ~lNANP Ig, recombinant WT
Ig, native monoclonal antibody 62, and polyclonal human
gamma globulins (HGG) (Cohn fraction II, Miles) were
loaded onto a 10~ SDS-PA~E and electrophoresed at 150
V under nonreducing (left panel) and reducing (right
panel) conditions. Resolved proteins or polypeptide
chains were transferred from the gel to 0.45-~m
nitrocellulose paper. After blotting, the filter was
blocked with 10~ solution of dry milk in 0.9~ NaCI for
two hours at room temperature. The sheet was then
incubated overnight at +4C by rocking with 125I-
labelled antibody Sp3-B4 (40 x 104 cpm/ml) in
phosphate-buffered saline, pH 7.3, containing 1~ bovine
serum albumin and 1~ Tween 20. After incubation, the
filter was washed extensively, dried and exposed to
Kodak XAR-5 film at -700C for 18 hours. Binding to
~lNANP Ig, recombinant WT Ig, antibody 62 and HGG in
RIA by the same 12sI-labelled probe (105 cpm/50~1) was
10,560; 420; 360; and 330 cpm, respectively.
Western blot analysis (Figure 4) showed that l2'I-
labelled mAb Sp3-B4 specifically bound antibody rlNANP
~ 094/28026 -~ T ,~ 2 t 63 1 25 PCT~S91/06090
in both the nonreduced (left panel) and reduced (right
panel) forms. In the latter, as expected, binding
occurred on the H- but not the L-chain, confirming that
the engineered ~lNANP antibody bears the (NANP) 3
epitope on the H chain. No binding occurred to
controls for the H and L chain and the human C region.
C. Efficiency of enqineered ~lNANP antibody in
ex~ressinq the (NANP) 3 epitope
A cross-inhibition assay was employed to assess the
engineered ~lNANP antibody's relative efficiency in
expressing the (NANP) 3 epitope. The synthetic peptide
(NANP) 3 and antibody ~lNANP were used to inhibit the
binding of 125I-labelled mAb Sp3-B4 to either the
(NANP) 3 peptide or the ~lNANP antibody immobilized on
microtiter plates.
Figure 5 shows results of cross-inhibition of l25I-
labelled antibody Sp3-B4 binding to synthetic peptide
(NANP) 3 (panel A) or engineered antibody ~lNANP (panel
8) by ~lNANP Ig or peptide (NANP) A fixed amount of
l25I-labelled antibody Sp3-B4 (probe) was mixed vol/vol
with decreasing amounts o~ the various inhibitors
diluted in phosphate-buffered saline, pH 7.3,
containing l~ bovine serum albumin and l~ Tween 20.
The mixture was incubated at +4C overnight by rocking.
Fifty ~l of each mixture were incubated on individual
polyvinyl microtiter wells coated with either synthetic
peptide (NANP) 3 (panel A) or purified engineered ~lNANP
Ig (panel B). The conditions of coating are as
detailed in the legend to Figure 4. The following
inhibitors were used: purified ~lNANP Ig, WT Ig, and
synthetic peptides (NANP) 3, CDR3 and vitronectin. The
percentage of inhibition was calculated as ~ollows:
[(average binding of the probe alone) - (average
binding of the probe incubated in the presence of
W094l28026 ~ S ~l ~ 3 ~ 2 5 PCT~S94/06090
26
inhibitor)]/(average binding of the probe alone) x 100.
Tests were done in duplicate.
Figure 5 shows that both the peptide and the engineered
antibody efficiently inhibited the binding to both
physical forms of the ~NANP) 3 epitope, i.e., synthetic
peptide and antibody borne. However, whereas the
~lNANP antibody was about four times more effective
than the peptide itself (panel A) in inhibiting binding
to the synthetic peptide, it was approximately 150
times more effective than the peptide in inhibiting
binding to the engineered Ig (panel B). The WT Ig and
control peptides (CDR3 and vitronectin) caused no
inhibition. Thus, when compared with the synthetic
peptide it appears that the (NANP) 3 epitope borne on
the ~lNANP antibody assumes a three-dimensional
configuration that in immunological terms more closely
mimics that of the active CS protein.
D. Induction in vivo of anti-NANP antibodies by
recombinant ~lNANP antibody
To determine whether the recombinant ~lNANP antibody
could be used to induce anti-NANP antibodies, in vivo
experiments were performed in rabbits. Two rabbits
were immunized with the engineered ~lNANP antibody, and
two controls receive the WT Ig. As indicated in Table
I, infra., as early as 30 days after the first
immunization, both rabbits immunized with the rlNANP
antibody produced anti-NANP antibodies detectable by
ELISA and RIA. After booster immunizations, the titer
rose in both rabbits; the maximal titer was 1/3200 on
day 70. Importantly, this antiserum was positive when
tested by indirect immunofluorescence on P. falci~arum
sporozoite showing that the epitope expressed by the
~lNANP Ig is indeed mimicking the native antigen. Sera
from control rabbits immunized with the WT Ig did not
~ 094/28026 ~ t ~ 2 1 6 3 1 2 5 PCT~S94/06090
react with the (NANP) 3 peptide immobilized on
microtiter wells nor with the parasite. Rabbits of
both groups produced an anti-human response as
determined by agglutination of red cells coated with
human gamma globulin. Rabbits' antisera were tested by
direct immunofluorescence on P. falciparum (strain
Indochina III) dried onto glass slides in the presence
of 10~ fetal bovine serum.
The observation that the VH region of an antibody
molecule can be engineered to express 15 amino acid
residues containing an epitope of an unrelated molecule
shows that the VH/CH polypeptide chain containing the
foreign epitope is properly assembled with the
endogenous L chain to form a (H2L2) tetramer, so it
appears that the insertion of this epitope in the CDR3
was tolerated and did not affect the overall Ig
framework folding. Based upon the present research, as
long as the recombinant epitope is stereochemically
compatible with contiguous CDR residues, it can be
inserted or substituted for a CDR and can be expected
to be exposed at the sur~ace o~ the molecule, although
it cannot be ruled out that the results reported here
may be due to the nature of the epitope itself. In the
construct described here, the (NANP) 3 sequence is
flanked on both sides by the amino acids Val and Pro.
Possibly, this helps stabilize the inserted epitope by
anchoring it at each end. The large ramification at
the C~ atom and the C~-methyl group of the Val residue
may hinder the main chain by decreasing its
flexibility; the side chain of Pro by curling back to
the main chain seizes it, leading to the formation of
an almost rigid side chain.
Studies in vitro using the binding site of a NANP-
specific monoclonal antibody as a probe for the
protein-surface interaction and in vivo demonstrating
W094/28026 ~ 2 1 ~ 3 1 2 5 PCT~S94/06090 ~
28
that rabbits immunized with the engineered Ig molecule
produce anti-NANP antibodies that react with the
plasmodium antigen show that the (NANP) 3 epitope
expressed by the engineered Ig is both antiyenic and
immunogenic. In other terms, neither the molecular
environment nor the globular folding of Ig modified the
immunologic structure of the tNANP) 3 epitope. From a
biological standpoint, the (NANP) 3 epitope engineered
into an Ig molecule can be viewed as an idiotope a la
carte built into the CDR3 of a host VH domain. Based
on what is known of the immunogenicity of idiotypes and
the predictable events that follow induction of
immunity via the idiotype network [Jerne Ann. Immunol.
(Paris) 125:373 (1974); Cozenave et al., PNAS 74:5122
(1977); Urbain et al., PNAS 74:5126 (1977); Bona et
al., J. Exp. Med 153: 951 (1981)], these results imply
that an immune response of predetermined epitope
specificity can be dictated in molecular terms and
predicted in vitro. This strategy can be exploited to
render a B-cell epitope T-independent, proving its
utility not only for analyses of the structure and
function of epitopes and Igs but also for the
development of new antibody vaccines, for example, as
an alternative to peptide based vaccines. Preparation
of vaccines may be accomplished using extant
methodology, already developed for immunoglobulins as
such.
2163125
094l28026 ~ PCT~S94/06090
29
TABLE I
Induction of Anti-NANP Antibodies in Rabbits
Immunized with the Engineered
~lNANP Antibodya
Rabbit Immunogen Days After Immunization
No.
0 30* 40 60* 70
44 WT 0 ND 0 0 0
WT 0 ND 0 0 0
49 ~lNANP 0 1/100 1/400 1/400 1/3200
~lNANP 0 0 1/400 1/200 1/1600
aAdult white rabbits were immunized subcutaneously in
several points of the back with 50 ~g of recombinant
~lNANP or the WT antibody emulsified in complete
Freund's adjuvant tCFA). Booster injections of 50 ~g
of the same immunogen in incomplete Freund's adjuvant
were given at monthly intervals (denoted by an
asterisk). Sera were collected on the days indicated
and tested for reactivity with the synthetic (NANP) 3
peptide by solid-phase ELISA and RIA. Briefly, serial
two-fold dilutions of individual sera in phosphate-
buffered saline, pH 7.3, containing 1~ bovine serum
albumin and 1~ Tween 230 were incubated overnight at
+4C on microtiter plates coated with the (NANP) 3
peptide at 5 ~g/ml in 0.9~ NaCI. After the incubation,
the plates were washed and incubated with either a
horseradish peroxidase conjugated goat anti-rabbit Ig,
or 12sI-labelled Protein A (Amersham) for one hour at
room temperature. Next, the plates were washed and the
bound antibodies determined by using a Bio-Rad
(Richmond, CA) ELISA reader or a gamma counter. The
binding of the preimmune sera was considered the
reference background value. The titer was determined
from the mean binding of triplicate samples after
subtracting the background binding values and is
expressed as the reciprocal serum dilution.
2 ~ ~ 3 ~ .2~
W094/28026 PCT~S9~/06090
ExamPle II
Materials and Methods
Monocl onal an tibodi es
The murine monoclonal antibody 28-14-8S (~2a, k)
specific for H-2b (Db allele) was purchased from the
American Tissue Type Collection (ATCC No. HB27).
Fluorescein-conjugate murine monoclonal antibody AMS-
32.1 reacting with I-Ad was purchased from Pharmigen
(San Diego, CA).
~y~ the ti c Pep ti des
A/PR/8/34 influenza virus nucleoprotein synthetic
peptide ASNENMETM (amino acid residues 366-374) was
synthesized on an ABI 430-A automated synthesizer
(Applied Biosystems, Inc., Foster City, CA).
Cells
B6-2 is a nonsecreting murine B cell hybridoma (H-2d~)
originally established by fusing C57Bl/6 (B6) (H-2b)
splenic B cell with M12.4.1 lymphoma cells of BALB/c
(H-2d) origin, and were kindly obtained from Dr. R. Abe
(National Institute of Health, Bethesda, MD). J558L is
a murine myeloma of BALB/c (H-2d) origin and is a H-
chain defective variant of J558 myeloma carrying the
rearrangement for a ~l light (L)-chain (Morrison,
Science 229:1202 (1985)). J558L cells lack
constitutive Ig secretion, but they secrete a H2L2 Ig
molecule when transfected with a H-chain gene. The
non-secreting Sp2/0 myeloma (H-2d) was obtained through
passage from ATCC No. CRL 1581. CD8+ murine CTL clone
34 (Vitiello et al., J. Immunol. 143:1512 (1989);
Vitiello et al., J. Immunol. 131:1635 (1983)) is
specific for the monopeptide ASNENMETM (residues 366-
374) of the nucleoprotein (NP) antigen of A/PR8
influenza virus (Bastin et al., J. Ex~. Med. 165:1508
(1987); Rotzschke et al., Nature 348:252 (1990);
2 1 6 3 1 2 5
~'094/28026 PCT~S94/06090
31
Townsend et al., Cell 44:959 (1986)), and is restricted
by the class I histocompatibility Db gene product. The
clone was maintained in culture by stimulation at
weekly intervals with irradiated syngeneic spleen cells
pulsed with the ASNENMETM synthetic peptide.
~ngineering techniques
The D region of the parental VH gene (KAYSHG; residues
93-98) was mutagenized (Sollazzo et al., Eur. J.
Immunol. 19:453 (1989)) to introduce a single
KpnI/Asp718 site to yield the intermediate sequence
KVPYSHG (residues 93-99). The amino acid 94A was
deleted and substituted by the VP doublet encoded by
the nucleotide sequence of the Asp718 cloning site.
Subsequently, complementary oligonucleotides 5' GTA CCC
GCT TCC AAT GAA AAT ATG GAG ACT ATG GAA TCA AGT ACA CTT
3', 5' GTA CAA GTG TAC TTG ATT CCA TAG TCT CCA TAT TTT
CAT TGG AAG CGG 3' coding for residues 366-379 of the
influenza nucleoprotein (NP) (ASNENMETMESSTL) were
introduced between 94V and 95P of the mutagenized VH
region. The engineered VHNP coded by the 2.3 kb EcoRI
~ragments was cloned upstream from a human rl constant
(C) region gene contained in the 12.8 kb vector pN~1
(Sollazzo et al., Eur. J. Immunol. 19:453 (1989)).
Thirty ~g of the DNA construct pN~lNP were
electroporated in B6-2, J558L and Sp2/0 cells (2 x 107)
using a field strength of 625 V/cm. Transfected cells
were cultured in RPMI 1640 supplemented with 10~ fetal
calf serum (FCS), 4 mM glutamine, 0.1 mM non-essential
amino-acids, 1 mM sodium pyruvate, 0.1 mM HEPES, 100
U/ml penicillin, 100 ~g/ml streptomycin and 0.5 ~M ~-
mercaptoehanol for 24 hours, and then selected in the
presence of neomycin (0.8 mg/ml) (G418; Gibco-BRL).
~ ~ r 2 1 6 3 1 25
W094/28026 PCT~S9~/06090
32
Esterase Release Assay
B6-2 HNP transfectants cells were screened for
presentation of the NP peptide by using an
immunoenzymatic method that measures the release of
esterase by CTL upon specific peptide antigen
recognition (Kane et al., Mol. Immunol. 26:759 (1989);
Pasternack et al., Nature 322:740 (1986)). Briefly, 105
effector cells (CTL clone 34) and 104 cells from each
transfectoma were coincubated in a final volume of 100
~l of culture medium in 96-well flat-bottom plates.
Untransfected B6-2 cells and B6-2 cells pulsed with the
NP peptide (5 ~g/ml) served as negative and positive
controls, respectively. Spontaneous and maximum
esterase release were assessed on effector cells
incubated in medium alone or in the presence of 1
Triton 100-X. After 4 hours of incubation at 37 ) C,
the plates were centrifuged of 400 rpm for 2 min and
the supernatants collected. Twenty five microliters of
each supernatant were transferred to 96-well flat-
bottom plates to which were added 175 ~l of phosphate
buffered saline (PBS) containing 2 x 10-4 M N-
benzyloxycarbonyl-L-lysine thiobenzil ester and 2.2 x
10-4 M 5,5'-dithiobis(2-nitrobenzoic acid) (Sigma, St
Louis, MO). After an incubation period of 30 min. at
25 C the absorbance at 412 nm was determined in an
ELISA plate reader. The results were expressed in
percent secretion as follows: [sample secretion -
spontaneous secretion / maximum secretion - spontaneous
secretion] x 100.
Cytotoxicity Assay
Cytotoxicity was tested in a 4 hours s1Cr release assay.
Briefly, target cells were labeled with NaslCrO4 (150
~Ci/1 x 106 cells) for 1 hour at 37 C in an atmosphere
of 5~ of CO2 with or without NP peptide (10 ~g/ml or as
specified), then washed and resuspended in culture
medium supplemented with 10~ FCS. One hundred ml of
~ 094/28026 ~ A ~ 2 1 6 3 1 2 5 PCT~S94/06090
33
5lCr-labeled target cells (2.5 x 105 cells/ ml) were
mixed with 100 ~l of CTL clone 34 (effector cells) at
an effector:target cells (E:T) ration of 10:1, or as
specified. The plates were incubated for 4 hours at 37
C in 5~ CO2, then centrifuged at 500 g for 4 minutes.
One hundred microliters of supernatant was removed and
counted in a gamma counter. Spontaneous and maximal
5lCr release were determined by incubating target cells
in medium alone or in the presence of 1~ Triton X-100,
respectively. The cytotoxic activity was calculated
from triplicated wells as follows: [experimental
release - spontaneous release / maximal release -
spontaneous release] x 100. Cold target competition
was done by mixing 50 ~1 of 5lCr-labeled B6-2.503 cells
(5 x 105 cells/ml) with 50 ~1 of EI-4 or B6-2 cells
pulsed with NP peptide (10 ~g/ml) at a cold:hot cell
ratio of 0:1, 5:1, 25:1, and 50:1. Then, 100 ~l of CTL
clone 34 (effector cells) were added at a E:T ratio of
10:1. Percent cytotoxicity was calculated 4 hours
later as described above. ~lNP, or ~lNANP as negative
control, was added at a final concentration of 100
~g/ml either during the pulsing or the cytotoxicity
phase. The murine monoclonal antibody 28-14-8S (~2a/K)
specific for Dbq or a mouse monoclonal antibody of the
same isotype but of unrelated specificity as control
was added during the cytotoxic assay at a final
concentration of 50, 5, and 0.5 ~g/ml, respectively.
HPLC fractions were tested for their capacity to pulse
B6-2 cells as follows: 100 ~1 of 5lCr-labeled B6-2
cells (2.5 x 105 cells/ml) were mixed with 5 ~l of each
fraction. After 1 hour incubation at 37 C, 100 ~1 of
CTL clone 34 (effector cells) were added at a E:T ratio
of 10:1. The NP peptide (10 ~g/ml) was used as
positive control.
W094/28026 ~i ~ a ~ ~ 2 1 6 3 1 2 5 PCT~S9~/06090
34
Isolation of NP peptide from the class
histocompatibility D surface molecule
The NP peptide was isolated from B6-2 HNP transfectants
by acid elution. Briefly, Db-specific monoclonal
antibody 28-14-8S (~2a,k) was immobilized on Protein-A
beads at a ratio of 500 ~1 of beads:3 mg of 28-14-8S
antibody for 1 hour at 4 C. Bulk cultures of 10~
stable B6-2 HNP transfectants (clone 514 or 503) were
pelleted and resuspended at 2 x 108 cells/ml in lysis
buffer (20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.5~ N~-
40, 5mM EDTA) containing a freshly-made cocktail of
proteases inhibitors [A-Protinin (5 ~g/ml), Leupeptin
(10 ~g/ml), Pepstatin-A (10 ~g/ml), and PMSF 1 mM] for
30 min. at 4 C. Nuclei were pelleted by
centrifugation at 3,000 g for 15 min. Lysates were
then mixed with antibody 28-14-8 S/Protein-A
immunosorbent by rocking for 1 hour at 4 C. Protein-A
beads conjugated with the influenza peptide/Db
complexes were then pelleted by centrifugation at 1,000
g for 5 min., washed three times with wash buffer (50
mM Tris, pH 7.5, 150 mM NaCl, 0.5~ NP40, 5 mM EDTA)
followed by a wash with PBS and HPLC-grade H2O. The NP
peptides were acid-extracted by washing the Protein A-
class I MHC molecules beads twice with 0.2
trifluoroacetic acid (TFA). The low molecular weight
material was separated by filtration through a
Centricon 10 titer (Amicon) with a molecular weight cut
off of 10,000 kD. The filtrate was lyophilized and
kept at -20 C until used.
HPLC analysis of peptides
Low molecular weight material containing peptides
purified from Db+ B6-2-HNP transfectants (clones 514 cr
503) were analyzed by reverse phase HPLC using a SMART
System unit (Pharmacia) and a mRPC C2/C18 2.1 /iO
column (Pharmacia). Peptides were eluted using 0.1~
TFA in H2O (v/v) (solution A) and 0.08~ TFA in
~ 094/28026 ~ `'` 2 ~ 6 3 1 25 PCT~S94/06090
acetonitrile (solution B). The flow rate was 100
~l/min and fractions of 100 ~1 were collected. The
following gradient conditions were used: 0-61 min. a
linear increase to 60~ B; 61-66 min. 6090 B; 66-71 min.
increase to lOO9o- B; 71-76 min. decrease to 0~ B. One
hundred micrograms of synthetic NP peptide (residues
366-374) was purified using the same conditions and
served as a reference.
A. Generation of B cells that present the influenza
virus NP peptide (residues 366-379) to a CTL clone
s~ecific for the same viral peptide
A H chain gene was engineered to encompass in the third
complementarity-determining region (CDR3) a nucleotide
sequence encoding for the amino acid sequence
ASNENMETMESSTL (residues 366-379) of influenza virus NP
antigen through the process of antibody antigenization
(Zanetti, Nature 355:466 (1992)). The engineered H
chain gene (HNP) was used to transfect B6-2 (H-2db),
SP2-0 (H-2d) and J558L (H-2d) cells, respectively.
J558L cells carry the gene for the A1 light chain and
served to produce a H2L2 antibody molecule.
The H chain plasmid is the product of the fusion of a
human ylC region with a murine VH engineered to express
the NP sequence 366-379 in CDR3. The coding strand of
the CDR3 region is shown in bold, with the NP-coding
sequence underlined. The amino acid sequence of the
influenza peptide 356ASNENMETMESSTL379 is shown in bold.
B, BamHI; RI, EcoRI; Neo, neomycin (G418) resistance;
Amp, ampicillin resistance. The DNA construct (pN~lNP)
was electroporated in the murine B6-2 (H-2db) B cell
hybridoma to generate target cells. The stable
transfectants were initially screened using a serine-
esterase release assay to select clones that could
activate CTL clone 34 to release serine esterase used
as a cellular probe (peptide presentation). The
selected clones were then tested in a conventional s;Cr-
W094/~80~6 ~''i ~CT~S9~/06090
release assay; clones that confirmed positive were
expanded. Alternatively, the DNA construct was
electroporated into murine myeloma cell line J558L (H-
2d), a H chain-defective variant of myeloma J558
5 carrying the rearrangement for a A1 L chain.
Supernatants of neomycin-resistant colonies (stable
tranfectants) were tested by ELISA for Ig production.
The final product is a H2L2 molecule ~lNP (antigenized
antibody).
Neomycin-resistant B6-2 hybrid transfectants were
screened for their ability to present the NP peptide to
CTL clone 34 specific for the 9m'r ASNENMETM sequence.
To simultaneously screen a large number of
transfectants, a serine esterase-release assay
(Pasternack et al., Nature 322: 740 (1986) ) was used.
This assay is a readily-detectable indicator of
secretory granules exocytosis triggered in the CTL
clone by specific recognition of the NP peptide/class
I MHC molecule complex on B6-2 HNP transfectants. Five
20 out of 120 (4~) transfectants induced a release of
esterase equivalent to that of B6-2 cells pulsed with
a synthetic NP peptide (Table 2) . B6 -2 cells
transfected with a wild type H chain gene (H~) -lacking
the NP epitope- and B6-2 cells alone served as negative
25 controls. To confirm the results of the esterase-
release assay a conventional 51Cr-release assay was
used. This assay showed an absolute correlation
between the two tests (Table 2). In order to select
transfectants with stable integration of the H~P gene
30 primary transfectants were subcloned and retested using
the same strategy. It is worth noting that
transfectants negative by both esterase- and 51Cr-
release assays remained negative throughout subsequent
tests. On the other hand, positive transfectants
3 5 maintained their ability to present the NP epitope
during a twelve month period.
~O 94/28026 ~ f PCT/US94/06090
37
tD
O
c~ z
~ --
tD 8
._ -- tD
Cl. D ~
z a~ ~ O O
~D C~i U ) S ~ C
D Z
`D ~ ~ D
~D
C ~D 0U~ --
~D ~ a~ D
~) ~ U) I ~t ~, ~ ~ Q
m s .c ~ Z ~,
~- ~ I Z a~ E
a~
m -- -' E
~ E _ {_
a: ~n ~D , o
m ~D 2g
~ Q
tD ~D ~"
C ~ ~ n - ~D
~D ~D * ~ o tn w
C ~n ~n . C, - _ ~
c ~ m E
* * _ ~
W094/28026 ; ~ 2 1 6 3 ~ 2 5 PCT~S9~/06090 ~
38
B. Specificitv of the killinq of enqineered cells by
the CTL clone
To ascertain the specificity of lysis of B6-2 HNP
transfectants by the CTL clone cold target competition
experiments were performed using B6-2 and EL-4 cells
pulsed in vitro with the synthetic NP peptide (10
~g/ml) as the cold inhibitor.
5lCr-labeled B6-2. 503 cells t5 x 105 cells/ml) were
mixed with EI-4 or B6-2 cells pulsed or not with NP
peptide (10 ~g/ml) at a cold:hot target cell ratio of
0:1, 5:1, 25:1 and 50:1, respectively. Then, CTL
(effector cells) clone 34 were added at a E:T ratio of
10:1. Percent cytotoxicity was calculated 4 hours
later from triplicate wells as described. Maximum and
minimum slCr release were 48, 197tl,177 and 6,137+93
cpm, respectively.
As shown in Figure 7 in both instances there was
complete inhibition of cytolysis at a competitor:
target ratio of 50:1. This demonstrates that killing
of B6-2 lymphoma cells engineered with the HNP gene was
specific for the NP peptide/class I MHC molecule
complex.
C. Soluble antiqenized antibody e.-pressing the NP
epito~e in CDR3 does not inter-ere with
presentation of processed pept-de from the
endoqenous HNP chain
Although B6-2 is a nonsecreting B cell lymphoma, it is
not unprecedented that transfection with a H chain gene
reactivates a latent L chain gene. For instance,
threshold amounts of antibody that may be undetectable
in our assay (< 1 ng/ml) could have been endocytosed,
processed and presented. To rule out this possibility,
it was verified whether or not soluble antigenized
antibody could mediate lysis of B6-2 cells.
~ 094l28026 ~ 2 1 6 3 1 2 5 PCT~S94/06090
39
s1Cr-labeled EI-4 cells (5 x 105 cells/ml) were mixed
with NP peptide and/or ~lNP (or ~lNANP as control) at
a final concentration of 10 or 100 ~g/ml, respectively.
Then, CTL clone 34 (effector cells) was added at an E:T
ratio of 10:1. Intact ~lNP molecule (or rlNANP as
control) was added during the cytotoxic phase of the
assay at a final concentration of 100 ~g/ml. Percent
cytotoxicity was calculated 4 hours later from
triplicate wells as described.
As shown in Figure 8 untransfected B6-2 cells pulsed
with ~lNP were not lysed. Moreover, the antigenized
antibody added to HNP transfectants during pulsing with
peptide did not affect lysis nor did it modify the
percent of lysis when added during the lytic phase.
These results demonstrate that lysis of B6-2 HNP
transfectants is the result of presentation of a
processed peptide derived from the endogenously
synthesized HNP chain. It should be pointed out that
lack of direct influence on CTL lysis suggests that
the whole antibody does not function as an anti-
receptor antibody.
5 D. LYsis of B6-2 HNP transfectants is restricted by
the Db allele
The role of the Db allele in the presentation of the NP
peptide to the CTL clone by the engineered B cells wa
analyzed using a twofold approach. First, it was
ascertained that a murine monoclonal antibody
(28.14.8S) specific for the Db allele could block
cytotoxicity.
s1Cr-labeled B6-2.503 cells (2.5 x 10~ cells/ml) were
mixed with CTL clone 34 (effector cells) at an E:T
ratio of 10:1 in the presence of various doses of the
murine monoclonal antibody 28.14 ( K, ~2a) specific for
W094/28026 ~ 2 1 6 3 1 2 5 PCT~S94/06090
Dbq. A mouse monoclonal antibody of the same isotype,
but of unrelated specificity, was used as control.
Percent cytotoxicity was calculated 4 hours later from
triplicate wells as described. Figure 9 shows a dose-
dependent inhibition by antibody 28.14.8S, but not byan isotype-matched control antibody.
Second, a series of HNP transfectants carrying the H-2d
haplotype (Sp2/0 and J558L cells) were analyzed. 51Cr-
labeled B6-2 (H-2b~d), Sp2/0 (H-2d) or J557L (H-2d) cells
pulsed with NP peptide (10 ~g/ml) or transfected with
HNP or HWT (2.5 x 105 cells/ml), were mixed with CTL
clone 34 (effector cells) at an E:T ratio of 10:1.
Percent cytotoxicity was calculated 4 hours later from
triplicate wells as described. As illustrated in
Figure 10 lysis of these cells did not exceed that of
control H~ transfectants. Thus, the NP peptide
resulting from the proteolytic fragmentation of the
endogenous HNP chain is presented in association with
D~ allele as if it were generated from an intracellular
replicating virus.
E. Purification of viral NP PePtide from B6-2 cells
transfected with the HNP gene
It was important to demonstrate that the NP peptide
could be purified from surface class I Db molecules and
that it could be used to pulse untransfected B6-2 cells
and mediate lysis by CTL clone 34. A lysate of 109 B6-
2 HNP transfectants was mixed with Protein-A Sepharose
beads coated with monoclonal antibody 28.14.8S. MHC-
bound peptides were extracted by acid elution using
0.2~ TFA, and the peptides were separated from class I
MHC molecules by centrifugation on a low molecular
weight Centricon filter. The low molecular weight
material was fractionated by reverse-phase HPLC. The
elution profile of a representative experiment is shown
,~ ?t~ 2 1 6 3 1 2 5 ~
094/28026 : PCT~S94/06090
41
in Figure ll (A and C). The HPLC profile of the
control gmer synthetic peptide ASNENMETM is shown as a
comparison. Individual fractions were used to pulse
untransfected 5lCr-labelled B6-2 cells to identify the
fraction(s) containing the NP peptide. Figure ll
(panel B and D) shows that in both instances the active
peptide was eluted in fractions l9 and 20, suggesting
that the peptide purified from the B6-2 HNP
transfectants has physicochemical and biological
characteristics similar to the 9mer synthetic peptide
ASNENMETM. Thus, a proteolytic fragment of the
endogenously-synthesized HNP chain bound the Db allele,
was transported at the cell surface and mediated lysis
by the NP-specific CTL clone.
F. Ouantitat-ve analysis of ~rocessinq and
presentat-on of the NP peptide in B6-2 cells
enqineere~ with the HNP
That B6-2 and HNP transfectants were killed efficiently
by a specific CTL clone and the NP peptide could be
eluted from the cell surface of these cells prompted a
quantitative analysis of this phenomenon. First, the
effect of exogenous addition of synthetic peptide NP on
the lysis of B6-2 HNP transfectants was verified.
Cr-labeled B6-2 or B6-2.503 cells were pulsed with
O.l, l or lO ~g of NP peptide for l hour at 37 C, then
mixed with CTL clone 34 (effector cells) at an E:T
ratio of lO:l. Percent cytotoxicity was calculated 4
hours later from triplicate wells. As shown in Figure
12 excess amounts (lO ~g/ml) of NP peptide added at the
beginning of the cytotoxicity assay failed to induce an
increase of lysis, hence implying that occupancy of MHC
class I molecules by processed peptide from the
endogenous HNP chain was already maximum.
The foregoing description details specific methods that
can be employed to practice the present invention.
2 1 6 3 ~ 25
W094/28026 PCT~S94/06090
42
Having detailed specific methods initially used to
identify, isolate, characterize, prepare and use the
immunoglobulins hereof, and a further disclosure as to
specific model entities, the art skilled will well
enough know how to devise alternative reliable methods
for arriving at the same information and for extending
this information to other intraspecies and interspecies
related immunoglobulins.
For example, antigen sequences can be engineered at any
restriction site unique to the CDR sequence within
which the antigen sequence is to be inserted, and
absent from the sequence of the immunoglobulin chain
wherein the CDR is located. Unique sequences in the
six CDRs can be identified and located using a
combination of known immunoglobulin nucleic acid
sequences and cleavage sites of restriction enzymes.
Further, a desired unique restriction site may be
introduced into the CDR wherein the antigenic
determinant is to be inserted using molecular
techniques well known to those skilled in the art. In
addition, it is well within the knowledge of those
skilled in the art to modify the present invention by,
for example, engineering an antigen within any of the
six complementarity-determining regions of an
immunoglobulin.
Thus, however detailed the foregoing may appear in
text, it should not be construed as limiting the
overall scope hereof; rather, the ambit of the present
invention is to be governed only by the lawful
construction of the appended claims.
