U.S. patent application number 11/173653 was filed with the patent office on 2005-11-03 for expression of functional antibody fragments.
This patent application is currently assigned to Genentech, Inc.. Invention is credited to Carter, Paul J..
Application Number | 20050244929 11/173653 |
Document ID | / |
Family ID | 25064638 |
Filed Date | 2005-11-03 |
United States Patent
Application |
20050244929 |
Kind Code |
A1 |
Carter, Paul J. |
November 3, 2005 |
Expression of functional antibody fragments
Abstract
Methods for the high yield production of antibody Fv-containing
polypeptides, especially Fab' and F(ab').sub.2 antibody fragments
are provided. Expression of heavy and light chain Fv in a microbial
secretory system is followed by recovery of Fv from the periplasm
under conditions that maintain a cysteine residue as a free thiol.
The free thiol is reacted with free thiol of an antibody fragment
of the same or differing specificity, or with agents such as
diagnostic labels or therapeutic moieties. The products offer
advantages of homogeneity and purity not available through the use
of known methods for preparing such derivatives.
Inventors: |
Carter, Paul J.; (San
Francisco, CA) |
Correspondence
Address: |
MERCHANT & GOULD PC
P.O. BOX 2903
MINNEAPOLIS
MN
55402-0903
US
|
Assignee: |
Genentech, Inc.
South San Francisco
CA
|
Family ID: |
25064638 |
Appl. No.: |
11/173653 |
Filed: |
July 1, 2005 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
11173653 |
Jul 1, 2005 |
|
|
|
08199268 |
Feb 28, 1994 |
|
|
|
08199268 |
Feb 28, 1994 |
|
|
|
PCT/US92/07986 |
Sep 18, 1992 |
|
|
|
08199268 |
Feb 28, 1994 |
|
|
|
07762292 |
Sep 19, 1991 |
|
|
|
Current U.S.
Class: |
435/69.1 ;
435/252.3; 435/320.1; 530/387.1; 536/23.53 |
Current CPC
Class: |
C07K 16/065 20130101;
C07K 16/468 20130101; C07K 2317/24 20130101; C07K 16/00 20130101;
C07K 16/32 20130101; C07K 2317/73 20130101; C07K 2319/034 20130101;
C07K 2317/10 20130101; Y10S 435/972 20130101; C07K 2319/00
20130101; C07K 2317/732 20130101; C07K 2317/77 20130101 |
Class at
Publication: |
435/069.1 ;
435/320.1; 435/252.3; 530/387.1; 536/023.53 |
International
Class: |
C12P 021/06; C07H
021/04; C07K 016/18; C12N 001/21; C12N 015/74 |
Claims
1-21. (canceled)
22. A method for high yield production of an immunoglobulin
polypeptide comprising culturing a host cell transformed with
nucleic acid encoding an immunoglobulin polypeptide under the
transcriptional control of an inducible promoter/operator system
wherein the promoter/operator system is subsequently induced,
thereby resulting in polypeptide levels in the cell culture of
greater than about 1 gram of polypeptide per liter of cell
culture.
23-24. (canceled)
Description
FIELD OF THE INVENTION
[0001] This invention relates to the production of functional
antibody fragments in a microbial host.
BACKGROUND OF THE INVENTION
[0002] Naturally occurring antibodies (immunoglobulins) comprise
two heavy chains linked together by disulfide bonds and two light
chains, each light chain being linked to one of the heavy chains by
disulfide bonds. Each chain has an N-terminal variable domain
(V.sub.H or V.sub.L) and a constant domain at its C-terminus; the
constant domain of the light chain is aligned with and disulfide
bonded to the first constant domain of the heavy chain, and the
light chain variable domain is aligned with the variable domain of
the heavy chain. The heavy chain constant region includes (in the
N- to C-terminal direction) the C.sub.H1 and hinge regions. The
light chain also contains a hinge domain. Particular amino acid
residues are believed to form an interface between and disulfide
bond the light and heavy chain variable domains, see e.g. Chothia
et al., J. Mol. Biol. 186:651-663 (1985); Novotny and Haber, Proc.
Natl. Acad. Sci. USA 82:4592-4596 (1985); Padlan et al., Mol.
Immunol., 23(9): 951-960 (1986); and S. Miller, J. Mol. Biol., 216:
965-973 (1990).
[0003] The constant domains are not involved directly in binding
the antibody to an antigen, but are involved in various effector
functions, such as participation of the antibody in
antibody-dependent cellular cytotoxicity and complement dependent
cytotoxicity. The variable domains of each pair of light and heavy
chains are involved directly in binding the antibody to the
antigen. The domains of natural light and heavy chains have the
same general structure, the so-called immunoglobulin fold, and each
domain comprises four framework (FR) regions, whose sequences are
somewhat conserved, connected by three hyper-variable or
complementarity determining regions (CDRs) (see Kabat, E. A. et
al., Sequences of Proteins of Immunological Interest, National
Institutes of Health, Bethesda, Md., (1987)). The four framework
regions largely adopt a .beta.-sheet conformation and the CDRs form
loops connecting, and in some cases forming part of, the
.beta.-sheet structure. The CDRs in each chain are held in close
proximity by the framework regions and, with the CDRs from the
other chain, contribute to the formation of the antigen binding
site.
[0004] Antibodies can be divided into a variety of antigen-binding
fragments. The F.sub.V fragment is a heterodimer containing only
the variable domains of the heavy chain and the light chain. The
Fab fragment also contains the constant domain of the light chain
and the first constant domain (C.sub.H1) of the heavy chain. Fab'
fragments differ from Fab fragments by the addition of a few
residues at the carboxy terminus of the heavy chain C.sub.H1 domain
including one or more cysteines from the antibody hinge region.
Fab'-SH is the designation herein for Fab' in which the cysteine
residue(s) of the constant domains bear a free thiol group.
F(ab').sub.2 antibody fragments originally were produced as pairs
of Fab' fragments which are between hinge cysteines.
[0005] Recombinant production of antibodies and antibody fragments
facilitates the engineering of antibodies with enhanced antigen
binding affinities, modified immunogenicity, and also of
bifunctional antibodies. The first expression systems reported with
which functional antibodies could obtained were for mammalian
cells. The U.S. patent of Cabilly et al., U.S. Pat. No. 4,816,567,
for example, teaches a method of co-expressing at least the
variable region sequences of light and heavy chains in host cells.
Other researchers in the field have reported baculovirus expression
systems (Haseman et al, Proc. Natl. Acad. Sci. USA 87:3942-3946
(1990), yeast systems (Horwitz et al., Proc. Natl. Acad. Sci. USA,
85:8678-8682 (1988), combinatorial libraries in phage lambda (Huse
et al., Science 246:1275-1281 (1989), and work with filamentous
phage (McCafferty et al., Nature 348:552-554 (1990).
[0006] The production of antibodies and antibody fragments in
bacterial systems have been pursued by workers in the field,
particularly in E. coli expression systems. There are several
advantages to E. coli expression systems, including a well-studied
and convenient gene technology which permits constructs to be made
easily and directly expressed, and the relatively convenient and
economical large-scale production of product made possible by the
fast growth of E. coli and its comparatively simple fermentation.
The large-scale production of functional antibody fragments in E.
coli would be valuable for research as well as commercial
applications.
[0007] The expression of antibody genes in bacteria was reported by
Cabilly et al., Proc. Natl. Acad. Sci. USA 81:3273-3277 (1984),
Boss et al., Nucleic Acids Res. 12:3791-3806 (1984); these reports
show cytoplasmic expression and rather variable yields were
reported. Zemel-Dreasen et al., Gene 315-322 (1984) report the
secretion and processing of an immunoglobulin light chain in E.
coli. Pluckthun et al., Cold Spring Harbor Symposia on Quantitative
Biology, Volume LII, pages 105-112 (1987, Cold Spring Harbor
Laboratory) disclose expression of a cytoplasmic hybrid protein, a
potentially exportable hybrid protein, and expression and
periplasmic transport of V.sub.L, V.sub.H, V.sub.LC.sub.L, and
V.sub.HC.sub.H chains as fusions with an alkaline phosphatase or
.beta.-lactamase signal sequences. Skerra and Pluckthun, Science
240:1038-1041 (1988) report the periplasmic secretion and correct
folding in vivo of the variable domains of an antibody to the E.
coli periplasm; a similar strategy and results were reported by
Better et al., Science 240:1041-1043 (1988) for expression of a
murine Fab fragment.
[0008] Bird et al., Nature 332:323-327 (1988) report the linkage of
the light and heavy chain fragment of the Fv region via an amino
acid sequence, and production of the complex as a single
polypeptide in E. coli; see also Ladner et al., U.S. Pat. No.
4,946,778. Huston et al., Proc. Natl. Acad. Sci. USA 85:5879-5883
(1988) report similar work. Ward et al., Nature 341:544-546 (1989)
report the production in and secretion of "single-domain"
antibodies (isolated heavy chain variable domains) from E. coli.
Condra et al., Journal of Biological Chemistry, 265(4):2292-2295
(1990) disclose the expression of cDNAs encoding antibody light and
heavy chains in E. coli and their renaturation into Fab fragments.
Better and Horwitz, Methods in Enzymology, 178:476-496 (1989),
describe the expression, and secretion of functional Fab fragments
from E. coli and Saccharomyces cerevisiae.
[0009] Pluckthun and Skerra describe techniques for the expression
of functional antibody Fv and Fab fragments in E. coli in Methods
In Enzymology 178:497-515 (1989). According to their strategy, in
the cytoplasm, the precursor proteins for V.sub.L and V.sub.H, each
fused to a bacterial signal sequence, are synthesized in reduced
form. After translocation through the inner membrane into the
periplasm, the signal sequences are cleaved, the domains fold and
assemble, and the disulfide bonds form. They teach that expression
of the Fab fragment according to their strategy is analogous.
Similar expression strategies are found elsewhere in the
literature. See also Pluckthun, Biotechnology, 9:545-551 (1991) for
a review of E. coli expression of antibody fragments.
[0010] Cabilly (Gene, 85:553-557 [1989]) teaches that, in E. coli
cells growing at reduced temperatures (21.degree. C. or 30.degree.
C., rather than at 37.degree. C.), a single expression plasmid
coding for kappa-chains and truncated heavy chains (Fd fragments)
gives rise to high yields of functional Fab fragments. Cabilly
discusses that the Fab fragments seem to exist in the E. coli
cytoplasm as non-covalently linked dimers, but that soluble Fab
fragments isolated from E. coli appear as covalent dimers, formed
by air oxidation following cell rupture.
[0011] It is known in the literature that in the presence of low
concentrations of a mild reductant such as cysteamine the bivalent
F(ab').sub.2 antibody fragment dissociates into two Fab' fragments.
This dissociation is reversible by mild oxidation. The production
of Fab and F(ab').sub.2 antibody fragments has also been shown by
partial reduction and limited proteolysis of intact antibodies, see
e.g. Parham, in Cellular Immunology (E. M. Weir, Ed., Blackwell
Scientific, CA) 4th edition, vol. 1 chapter 14 (1983), however with
these methods it is difficult to control the precise nature and
proportions of the antibody fragment recovered. Bivalent antibodies
are those which contain at least two epitopic combining sites
(which sites may be on the same or different antigens).
[0012] Bispecific antibodies are bivalent antibodies capable of
binding two epitopes not shared by a single antigen. Bispecific
monoclonal antibodies (BsMAbs) with dual specificities for
tumor-associated antigens on tumor cells and for surface markers in
immune effector cells have been described (see, e.g. Liu et al.,
Proc. Natl. Acad. Sci. USA 82:8648 (1985); Perez et al., Nature
316:354 (1985)). These BsMAbs have been shown to be effective in
directing and triggering effector cells to kill tumor cell targets
(Fanger et al., Immunol. Today 12:51 (1991)). One approach to the
production of BsMabs involves the fusion of two monoclonal
antibody-producing hybridomas to form quadromas (hybrid hybridomas)
which secrete BsMab in addition to undesirable chain combinations
including parental MAbs (Milstein, C. and Cuello, A. C., Nature,
305:537 [1983]) However, for production of bispecific humanized
antibodies and antibody fragments, other techniques would be
preferred.
[0013] Nisonoff and Mandy (Nature 4826:355-359 (1962)) describe the
digestion of rabbit antibodies and subsequent recombination of the
antibody fragments; they disclose that antibody molecules of dual
specificity can be obtained by combining univalent fragments of
pepsin-treated antibodies of different specificities. See also
Hammerling et al., Journal of Experimental Medicine 128:1461-1469
(1968); Parham, Human Immunology 12:213-331 (1985); Raso and
Griffin, Cancer Research 41:2073-2076 (1981); and Paulus (U.S. Pat.
No. 4,444,878).
[0014] Another approach utilizes directed chemical coupling of
bispecific Fab' fragments from two different MAbs to assemble a
BsMAb, in this case a F(ab').sub.2, with the desired specificities
(See e.g., Nolan et al., Biochimica et Biophysica Acta 1040:1
(1990). See also R. A. Maurer's Ph.D. Thesis, Harvard University
(1978), and Brennan et al., Science 229:81-83 (1985) for
chemistries for the directed coupling of dithionitrobenzoate
derivatives of Fab' fragments. Brennan et al. also teach the use of
use sodium arsenite to cross-link two proximate cysteines, however
this reaction involves highly toxic compounds. (Glennie et al., J.
Biol. Chem., 141(10): 3662-3670 [1985] and J. Immunol.,
139:2367-2375 [1975]) teach the preparation of bispecific
F(ab').sub.2 antibody fragments containing thioether linkages.
These chemistries would also be applicable for the coupling of
identical Fab' fragments.
[0015] Lyons et al., Protein Engineering 3(8)703-708 (1990) teach
the introduction of a cysteine into an antibody (there the C.sub.H1
domain of a heavy chain) and the site-specific attachment of
effector or reporter molecules through the introduced cysteine.
[0016] Despite the advances in E. coli expression of functional
antibody fragments shown in the literature, there remains a need
for efficient and economical techniques for the production of
bivalent antibodies, particularly F(ab').sub.2 molecules, and for
methods which permit the tailoring of bivalent and bispecific
F(ab').sub.2 molecules. It would be desirable to produce stable
Fab'-SH polypeptides which may be conveniently coupled in vitro to
form bivalent Fv or F(ab').sub.2 molecules.
[0017] It is therefore an object of this invention to provide
methods for the preparation of polypeptides comprising Fv domains,
particularly Fab', Fab'-SH and F(ab').sub.2 antibody fragments, in
or derived from bacterial cell culture in high yield.
[0018] It is a further object of this invention to provide methods
for the efficient preparation of homogenous bivalent and bispecific
F(ab').sub.2 antibody fragments.
[0019] It is another object of this invention to provide Fab'
antibody fragments having at least one hinge region cysteine
present as a free thiol (Fab'-SH). It is a related object to
obviate the inherent problems in generating Fab'-SH from intact
antibodies: differences in susceptibility to proteolysis and
non-specific cleavage, low yield, as well as partial reduction
which is not completely selective for the hinge disulfide bond(s).
It is another object of the present invention to prevent
intra-hinge disulfide bond formation without resorting to the use
of highly toxic arsenite to chelate vicinal thiols, or other
inefficient and undesirable methods.
[0020] Other objects, features, and characteristics of the present
invention will become apparent upon consideration of the following
description and the appended claims.
SUMMARY OF THE INVENTION
[0021] The principal embodiments of this invention are based on the
surprising identification of cysteinyl free thiol in cysteinyl
residues located outside of the light-heavy chain interface of
recombinant microbial periplasmic antibody fragments, and the
surprising discovery that Fv variants can be produced which contain
only a single cysteinyl residue in the free thiol form. This
facilitates the preparation of homogeneous recombinant F(ab').sub.2
and other Fv-containing bivalent polypeptides. Accordingly, in one
embodiment this invention comprises expressing and secreting into
the periplasm of a recombinant microbial cell culture a Fv
polypeptide containing an immunoglobulin heavy chain Fv region and
an immunoglobulin light chain Fv region, said light or heavy chain
also comprising an unpaired cysteinyl residue as a free thiol, and
recovering said polypeptide under conditions that substantially
maintain said cysteinyl residue as the free thiol.
[0022] It will be understood that either of the Fv light or heavy
chains optionally is fused to a polypeptide sequence which contains
one or more cysteinyl residues, provided that at least one of such
cysteinyl residues located in the domain C-terminal to either of
the light or heavy chain Fv is present as the free thiol in the
periplasm. Suitable polypeptide sequences include but are not
limited to sequences derived from immunoglobulins, carrier
proteins, receptors, growth factors, nutrient substances, cellular
adhesive molecules, tissue-specific adhesion factors, enzymes,
toxins and the like. Typically, an unpaired cysteinyl residue is
substituted at the C-terminus of either the light or the heavy
chain Fv domain or at the C-terminus of the light chain or heavy
chain CH1 domain. The fused polypeptide sequence may comprise only
the cysteinyl residue, or the cysteinyl residue can be present in a
polypeptide fused to the C-terminus of (1) one of the Fv domains,
(2) the light chain constant region or (3) the CH1 domain of a
heavy chain. In the latter instance, the Fv-CH1 domain is fused to
a hinge region bearing a single free thiol cysteinyl residue, and
preferably a hinge region amino acid sequence variant containing
only a single cysteinyl residue. However, any other sequence
containing a single free thiol cysteinyl residue is useful in place
of the immunoglobulin hinge region.
[0023] In other embodiments, the Fv-CH1 domain is fused to a hinge
region bearing more than one, typically two or three, free thiol
residues, preferably a hinge region amino acid sequence variant
containing two or more cysteinyl residues. For example, Fab'-SH
amino acid sequence variants which have been modified to contain
three cysteinyl residues may be produced recombinantly in a
suitable host cell, for example CHO cells or E. coli, and are
conveniently coupled in vivo by the host cell to form F(ab').sub.2
with three disulfide bonds connecting the heavy chains.
[0024] More particularly, the objects of this invention are
accomplished by a method for the production of a Fab' antibody
polypeptide having at least one hinge region cysteine present as a
free thiol (Fab'-SH), comprising the steps of:
[0025] a. expressing nucleic acid encoding an immunoglobulin
presequence comprising Fab' in a microbial host cell culture
transformed with a vector comprising said nucleic acid operably
linked to control sequences recognized by the host cell transformed
with the vector, under conditions suitable for the secretion of
Fab' to the periplasmic space of the host cell and formation of
Fab'-SH; and
[0026] b. recovering Fab'-(SH).sub.n from said host cell, where n
is greater than or equal to one.
[0027] Additional embodiments of the method of this invention
comprise the subsequent recovery (including purification) of the
Fab'-SH under conditions suitable for maintaining the hinge
cysteinyl thiol(s) in protonated form. In certain embodiments, a
metal ion chelating agent and/or a protease inhibitor is present
during the culturing of the transformed cell or in the recovery of
the Fab'-SH.
[0028] Optionally, the Fab'-SH is released from the host by
freeze-thawing the host cell, subjecting it to osmotic shock,
preparing a cell paste and purifying the Fab'-SH from the cell
paste. Optionally, release of Fab'-SH from the host cell is
facilitated by enzymatic digestion of the cell e.g., using lysozyme
or physical disruption, e.g., by sonication or by use of a French
press.
[0029] The Fv or polypeptide fused to the Fv (typically the hinge
sequence) is modified in certain embodiments so that only one
cysteinyl residue is present which, in the periplasm, exhibits free
thiol. Thus, for example, the Fab amino acid sequence is modified
by deleting or substituting all of the hinge sequence cysteinyl
residues C-terminal to the first cysteine. Similarly, to obtain
embodiments where, for example, three cysteinyl residues are to be
present as the free thiol form, all of the hinge sequence cysteinyl
residues C-terminal to the third cysteine are modified by deletion
or substitution.
[0030] This invention also provides Fv, Fab', Fab'-SH and
F(ab').sub.2 polypeptides which do not have a disulfide bond
between the light and heavy chains. These polypeptides are referred
to herein as "linkless". For example, the amino acid sequence of
the polypeptide is modified by substituting or deleting one or,
preferably, both of the two cysteines which form the inter-chain
(heavy-light) disulfide bond. Typically, these cysteines are
replaced with serines, although it is within the scope hereof to
covalently modify one or both of the cysteine side chain so as to
be incapable of forming a disulfide bond. In certain embodiments,
there is a very strong interaction between the light and heavy
chains such that eliminating the interchain disulfide does not
result undesirable levels of dissociation between the light and
heavy chains. The linkless Fab'-SH polypeptides advantageously
allow homogeneous F(ab').sub.2 to be prepared by chemical
coupling.
[0031] In another aspect, a method for the preparation of a
polypeptide comprising F(ab').sub.2 comprises the steps of:
[0032] a. expressing nucleic acid encoding an immunoglobulin
presequence comprising a first Fab' in a microbial host cell
culture transformed with a vector comprising the nucleic acid
operably linked to control sequences recognized by the host cell
transformed with the vector, under conditions suitable for the
secretion of said first Fab' to the periplasmic space of the host
cell and formation of Fab'-SH, said first Fab' being capable of
binding a first epitope;
[0033] b. expressing nucleic acid encoding an immunoglobulin
presequence comprising a second Fab' in a microbial host cell
culture transformed with a vector comprising the nucleic acid
operably linked to control sequences recognized by the host cell
transformed with the vector, under conditions suitable for the
secretion of said second Fab' to the periplasmic space of the host
cell and formation of Fab'-SH, said second Fab' being capable of
binding a second epitope;
[0034] c. recovering said first and second Fab'-SH from said host
cells; and
[0035] d. forming covalent bonds between the free thiol cysteinyl
residues of said first and second Fab'-SH to form bivalent
F(ab').sub.2.
[0036] In embodiments of the invention, efficient directed
disulfide bond formation occurs in vitro, utilizing directed
covalent coupling methods known in the art whereby bispecific
bivalent antibodies are produced. In particularly preferred
embodiments, the disulfide bond formation between the first and
second Fab'-SH comprises the following steps:
[0037] a. reacting the first Fab'-SH with (i) 5,5'-dithiobis
(2-nitrobenzoic acid) (DTNB) to form a thionitrobenzoate derivative
Fab'-TNB or (ii) a bifunctional maleimide;
[0038] b. directly coupling said first Fab'-TNB or maleimidated
Fab' to the second Fab'-SH to form a F(ab').sub.2; and
[0039] c. recovering said F(ab').sub.2.
[0040] Novel F(ab').sub.2 compositions are provided by the methods
of this invention. Such compositions are
[0041] a. essentially free of F(ab').sub.2 Fv regions containing
cysteinyl residues with derivatized sulfhydryl groups, except for
any native disulfide bond found in the native Fv regions,
[0042] b. entirely free of F(ab').sub.2 having hinge region
intrachain disulfide bonds,
[0043] c. entirely free of contaminating arsenite, and
[0044] d. entirely homogenous as to the heavy chain C-terminal
amino acid residue.
[0045] In another embodiment, high level expression of
immunoglobulins or fragments thereof in recombinant microbes is
achieved by a method comprising culturing a host cell transformed
with nucleic acid encoding an immunoglobulin polypeptide under the
transcriptional control of an inducible promoter/operator system
whereby expression of the polypeptide is repressed prior to
induction sufficient to permit post-induction polypeptide levels in
the cell culture of greater than about 1 gram of polypeptide per
liter of cell culture. Typically, this is accomplished by using a
powerful promoter such as phoA in a low copy number vector or in a
host engineered to express levels of repressor sufficient to fully
occupy all phoA operator sites and fully repress the operon.
BRIEF DESCRIPTION OF THE DRAWINGS
[0046] FIG. 1 shows the plasmid pA19 as used for the co-secretion
of huMAb4D5-8 .kappa. light chain and heavy chain Fd' fragment from
E. coli. The Fab' expression unit is dicistronic with both chains
under the transcriptional control of the PhoA promoter (C. N. Chang
et al., Gene 44: 121 (1986)) which is inducible by phosphate
starvation. The humanized variable domains (huV.sub.L and
huv.sub.H, P. Carter et al., Proc. Natl. Acad. Sci. U.S.A.,
89(10):4285-9 (1992)) are precisely fused on their 5' ends to a
gene segment encoding the heat stable enterotoxin II (stII) signal
sequence (R. N. Picken et al., Infect. Immun. 42: 269 (1983)) and
on their 3' side to human .kappa..sub.1 (C.sub.L, W. Palm and N.
Hilschmann, Z. Physiol. Chem. 356: 167 (1975)) and IgG1 (C.sub.H1,
J. W. Ellison et al., Nucleic Acids Res. 10: 4071 (1982)) constant
domains respectively. The coding regions are separated by 83 base
pairs and each is preceded by a ribosomal binding site (Picken et
al., supra) to enable efficient initiation of translation. The Fab'
expression unit was cloned into the Eco RI site of pBR322 (F.
Bolivar et al., Gene 2: 95 (1977)) previously modified by removal
of the Sal I and Sph I sites without changing the amino acid
sequence of the tetracycline resistance gene product. Construction
of different Fab' variants was facilitated by installing unique Sal
I and Sph I sites towards the end of the C.sub.H1 gene and
immediately preceding the bacteriophage .lambda. t.sub.0
transcriptional terminator (S. Scholtissek, et al., Nucleic Acids
Res. 15: 3185 (1987)).
[0047] FIG. 2 shows the purification of the Fab, Fab' and
F(ab').sub.2 (Cys Ala Ala variant) fragments of huMAb4D5-8.
Proteins were analyzed by SDS-PAGE on a 4 to 20% gel under
non-reducing conditions with Coomassie brilliant blue (R250).
Samples shown are protein molecular weight markers (lane 1), Fab
(lane 2); Fab'-SH before (lane 3) and after (lane 4) a mock
coupling reaction; Fab'-TNB before (lane 5) and after (lane 6) a
mock coupling reaction; Fab'-SH coupled with Fab'-TNB (lane 7),
F(ab').sub.2 purified by gel filtration (lane 8) and F(ab').sub.2
derived from limited pepsin digestion (E. Lamoyi, et al., Methods
Enzymol. 121: 652 (1986)) of full length huMAb4D5-8 expressed in
mammalian cells (P. Carter et al., Proc. Natl. Acad. Sci. U.S.A.
(1992) supra (lane 9). Antibody fragments (2 .mu.g per sample) were
reacted with 4 mM iodoacetamide prior to electrophoresis. The
huMAb4D5-8 Fab fragment was purified from fermentation supernatants
by centrifugation to remove cell debris, DEAE sepharose ion
exchange chromatography and then affinity purified using protein A
CL 4B or protein G sepharose. HuMAb4D5-8 Fab'-SH was purified by
thawing 15 g cell paste in the presence of 21 ml of 100 mM sodium
acetate (pH 3.5), 10 mM EDTA, 0.2 mM PMSF, 5 .mu.M pepstatin, 5
.mu.M leupeptin, 2.4 mM benzamidine. Cell debris was removed by
centrifugation (40,000 g, 10 min, 4.degree. C.). The resultant
supernatant (pH 5.0) was passed over DEAE sepharose and loaded on
to a 2 ml protein G sepharose column. Protein eluted with 100 mM
sodium acetate (pH 3.5), 10 mM EDTA was adjusted to pH 4.0 in the
presence of 1.5M (NH.sub.4).sub.2SO.sub.4 and loaded on to 2 ml
phenyl Toyopearl column. Fab'-SH was eluted with 20 mM sodium
acetate (pH 4.0), 0.8 M (NH.sub.4).sub.2SO.sub.4, 10 mM EDTA,
reduced in volume by ultrafiltration (Centriprep-10, Amicon) and
buffer exchanged into 10 mM sodium acetate (pH 5.0), 10 mM EDTA by
G25 gel filtration. The Fab thionitrobenzoate derivative was
prepared in a similar manner except that the DEAE flow through was
adjusted to 5 mM DTNB and pH 7.5. The total concentration of
huMAb4D5-8 Fab and Fab' variants was determined from the measured
absorbance at 280 nm and the extinction coefficient determined by
amino acid composition analysis (.epsilon..sup.0.1%=1.56). The free
thiol content of huMAb4D5-8 Fab'-SH was determined by analysis with
DTNB as described (T. E. Creighton, Protein Structure, a Practical
Approach (IRL Press, Oxford, UK, 1990), p. 157), whereas the TNB
content of huMAb4D5-8 Fab'-TNB was determined by the yield upon
reduction with dithiothreitol. Equimolar quantities of huMAb4D5-8
Fab'-TNB (by TNB content) and Fab'-SH (by -SH content) were coupled
at a combined concentration of .gtoreq.1.4 mg/ml in the presence of
100 mM tris-HCl (pH 7.5) and 10 mM EDTA for 1 hour at 37.degree. C.
HuMAb4D5-8 F(ab').sub.2 was isolated from the coupling reaction by
S100-HR gel filtration (Pharmacia) in the presence of
phosphate-buffered saline. The F(ab').sub.2 samples were passed
through a sterile 0.2 .mu.m filter and stored either at 4.degree.
C. or flash frozen in liquid nitrogen and stored at -70.degree.
C.
DETAILED DESCRIPTION OF THE INVENTION
Definitions
[0048] In general, the following words or phrases have the
indicated definitions when used in the description, examples, and
claims:
[0049] The term Fv is defined to be a covalently or
noncovalently-associated heavy and light chain heterodimer which
does not contain constant domains.
[0050] The term Fab' is defined herein as a polypeptide comprising
a heterodimer of the variable domain and the first constant domain
of an antibody heavy chain, plus the variable domain and constant
domain of an antibody light chain, plus at least one additional
amino acid residue at the carboxy terminus of the heavy chain
C.sub.H1 domain including one or more cysteine residues.
F(ab').sub.2 antibody fragments are pairs of Fab' antibody
fragments which are linked by a covalent bond(s).
[0051] The Fab' heavy chain may include a hinge region. This may be
any desired hinge amino acid sequence. Alternatively the hinge may
be entirely omitted in favor of one or more cysteine residues or,
preferably a short (about 1-10 residues) cysteine-containing
polypeptide. In certain applications, a common naturally occurring
antibody hinge sequence (cysteine followed by two prolines and then
another cysteine) is used; this sequence is found in the hinge of
human IgG.sub.1 molecules (E. A. Kabat, et al., Sequences of
Proteins of Immunological Interest 3rd edition (National Institutes
of Health, Bethesda, Md., 1987)). In other embodiments, the hinge
region is selected from another desired antibody class or isotype.
In certain preferred embodiments of this invention, the C-terminus
of the C.sub.H1 of Fab' is fused to the sequence Cys X X. X
preferably is Ala, although it may be any other residue such as
Arg, Asp, or Pro. One or both X amino acid residues may be
deleted.
[0052] The "hinge region" is the amino acid sequence located
between C.sub.H1 and C.sub.H2 in native immunoglobulins or any
sequence variant thereof. In the case of the humanized 4D5 antibody
described infra, the hinge region is located between residues 224
(asp in . . . Cys Asp Lys . . . ) and 233 (Pro in . . . Cys Pro
Ala). Analogous regions of other immunoglobulins will be employed,
although it will be understood that the size and sequence of the
hinge region may vary widely. For example, the hinge region of a
human IgG1 is only about 10 residues, whereas that of human
IgG.sub.3 is about 60 residues.
[0053] The term Fv-SH or Fab'-SH is defined herein as a Fv or Fab'
polypeptide having at least one cysteinyl free thiol. Preferably
the free thiol is in the hinge region, with the light and heavy
chain cysteine residues that ordinarily participate in inter-chain
bonding being present in their native form. In the most preferred
embodiments of this invention, the Fab'-SH polypeptide composition
is free of heterogenous proteolytic degradation fragments. In
certain embodiments, the Fab'-SH polypeptide is also substantially
(greater than about 90 mole percent) free of Fab' fragments wherein
heavy and light chains have been reduced or otherwise derivatized
so as not to be present in their native state, e.g. by the
formation of aberrant disulfides or sulfhydryl addition products.
In alternative embodiments, the Fab'-SH has heavy and light chains
which are not covalently coupled.
[0054] A humanized antibody for the purposes herein is an
immunoglobulin amino acid sequence variant or fragment thereof
which is capable of binding to a predetermined antigen and which
comprises a FR region having substantially the amino acid sequence
of a human immunoglobulin and a CDR having substantially the amino
acid sequence of a non-human immunoglobulin or a sequence
engineered to bind to a preselected antigen.
[0055] The term "control sequences" refers to DNA sequences
necessary for the expression of an operably linked coding sequence
in a particular host organism. The control sequences that are
suitable for prokaryotes, for example, include a promoter,
optionally an operator sequence, a ribosome binding site, and
transcriptional terminators. Particularly preferred are highly
regulated inducible promoters that suppress Fab' polypeptide
synthesis at levels below growth-inhibitory amounts while the cell
culture is growing and maturing, for example, during the log
phase.
[0056] Nucleic acid is "operably linked" when it is placed into a
functional relationship with another nucleic acid sequence. For
example, DNA for a presequence or secretory leader is operably
linked to DNA for a polypeptide if it is expressed as a preprotein
that participates in the secretion of the polypeptide; a promoter
or enhancer is operably linked to a coding sequence if it affects
the transcription of the sequence; or a ribosome binding site is
operably linked to a coding sequence if it is positioned so as to
facilitate translation. Generally, "operably linked" means that the
DNA sequences being linked are contiguous and, in the case of a
secretory leader, contiguous and in same reading frame. However
enhancers do not have to be contiguous. Linking is accomplished by
ligation at convenient restriction sites. If such sites do not
exist, then synthetic oligonucleotide adaptors or linkers are used
in accord with conventional practice.
[0057] An "exogenous" element is defined herein to mean a nucleic
acid sequence that is foreign to the cell, or homologous to the
cell but in a position within the host cell nucleic acid in which
the element is ordinarily not found.
[0058] As used herein, the expressions "cell" and "cell culture"
are used interchangeably and all such designations include progeny.
Thus, the words "transformants" and "transformed cells" include the
primary subject cell and cultures derived therefrom without regard
for the number of transfers. It is also understood that all progeny
may not be precisely identical in DNA content, due to deliberate or
inadvertent mutations. Mutant progeny that have the same function
or biological activity as screened for in the originally
transformed cell are included. Different designations are will be
clear from the contextually clear.
[0059] "Plasmids" are designated by a lower case p preceded and/or
followed by capital letters and/or numbers. The starting plasmids
herein are commercially available, are publicly available on an
unrestricted basis, or can be constructed from such available
plasmids in accord with published procedures. In addition, other
equivalent plasmids are known in the art and will be apparent to
the ordinary artisan.
[0060] "Recovery" or "isolation" of a given fragment of DNA from a
restriction digest means separation of the digest on polyacrylamide
or agarose gel by electrophoresis, identification of the fragment
of interest by comparison of its mobility versus that of marker DNA
fragments of known molecular weight, removal of the gel section
containing the desired fragment, and separation of the gel from
DNA. This procedure is known generally. For example, see Lawn et
al., Nucleic Acids Res., 9: 6103-6114 (1981), and Goeddel et al.,
Nucleic Acids Res. 8:4057 (1980).
[0061] "Preparation" of DNA from cells means isolating the plasmid
DNA from a culture of the host cells. Commonly used methods for DNA
preparation are the large and small scale plasmid preparations
described in sections 1.25-1.33 of Sambrook et al., (Molecular
Cloning: A Laboratory Manual New York: Cold Spring Harbor
Laboratory Press, 1989). DNA preparations are purified by methods
well known in the art (see section 1.40 of Sambrook et al.,
supra).
Suitable Methods for Practicing the Invention
[0062] Typically a cultured microbial host cell is transformed with
a vector comprising Fab'-encoding nucleic acid (i.e., nucleic acid
encoding heavy chain Fd fragment and light chain) operably linked
to control sequences recognized by the host cell transformed with
the vector (hereafter, "Fab" will be referred to as a specific
embodiment, but it will be understood that any Fv-containing
antibody fragments or fusion/sequence derivative thereof can be
used provided that a cysteine residue capable of forming a
periplasmic free thiol is present in the Fv region or sequence
fused thereto). The cells are cultured under conditions suitable
for the secretion of Fab' into the periplasmic space of the host
cell and formation of the free thiol. In general, a dicistronic
operon is used to direct the co-expression of corresponding light
and heavy chain fragments. Alternatively, separate chains are
expressed from separate promoters on the same or different
plasmids. Secondly, the Fab chains are preceded by signal sequences
to direct secretion into the periplasmic space, where it is
believed that the redox environment favors disulfide bond formation
for assembly of the light and heavy chain fragments but not
disulfide bond formation between hinge cysteine residues. In
particularly preferred embodiments, the expression control sequence
is the E. coli phoA promoter (C. N. Chang et al., Gene 44: 121
(1986)) inducible by phosphate starvation and the signal sequence
is the heat-stable enterotoxin II signal sequence (R. N. Picken et
al, Infect. Immun. 42: 269 (1983)).
[0063] It is currently preferred that the antibody fragments be
expressed in bacterial cells grown at high cell density in a
fermentor. Suitable fermentation conditions are described in the
Example below.
[0064] The polypeptides containing free thiol are recovered from
the fermentation media and/or recovered from freeze-thawed cells
(typically by osmotic shock) and subsequently purified. Recovery
(including purification) is most successful if the Fab'-SH is
maintained in the protonated form. Other conditions for maintaining
the protonated form include the use of organic solvents or other
agents for shifting the pKa of dissociation of -SH. This is
conveniently accomplished at acid pH, i.e., preferably 2 or more pH
units below the pKa of the hinge or unpaired cysteinyl thiol.
Alternatively, the Fab'-SH is reacted with a protective group such
as TNB or p-methoxybenzyl in order to maintain the Fab' in a
homogenous state suitable for further reaction. In yet another
alternative, pyridine disulfide is added to the Fab'-SH to form a
mixed disulfide; this stabilizes the free sulfhydryl until it is
deprotected for coupling or other processing. In still another
alternative embodiment, the free sulfhydryl is not protected but is
reduced prior to coupling or further processing. Suitable
protecting groups, known in the art, are described in E. Gross
& J. Meiemhofer, The Peptides: Analysis, Structure, Biology Vol
3: Protection of Functional Groups in Peptide Synthesis (Academic
Press, New York, 1981).
[0065] Affinity purification, such as on streptococcal protein G
sepharose or staphylococcal protein A at acidic pH (typically,
about pH 4 to pH 6, preferably about pH 5.0) is preferred.
Alternatively, two-phase liquid extraction may be used. Small
amounts of contaminating proteolytic fragments are readily removed
by hydrophobic interaction chromatography, using, for example,
silica gel and/or alkyl or aryl-substituted chromatography resins
such as phenyl Toyopearl. It is preferred that a cocktail of
protease inhibitors be used (such as phenylmethylsulfonyl fluoride
(PMSF), leupeptin, pepstatin and/or benzamidine) to inactivate or
inhibit host proteases, and to include a metal ion chelating agent
such as EDTA in the culture and recovery procedures. The chelating
agent is selected and included in an amount to prevent metal ion
catalysis of disulfide bond formation.
[0066] In certain embodiments, Fab'-(SH).sub.n (where n is equal to
or greater than one) form F(ab').sub.2 in vivo during recombinant
production. In these embodiments, the Fab'-(SH) amino acid sequence
has been modified to contain preferably more than one cysteinyl
residue. For example, F(ab').sub.2 having three disulfide bonds
between heavy chains are recovered directly from an E. coli cell
paste, following the culture in E. coli of Fab'-SH.sub.3.
[0067] In other embodiments, F(ab').sub.2 is prepared from purified
Fab'-(SH).sub.n (where n is equal to or greater than one) by in
vitro chemical coupling using cross-linking agents or adventitious
oxidants such as dissolved oxygen. In the latter instance, purified
Fab' will form F(ab').sub.2 due simply to air oxidation. In
addition, bispecific F(ab').sub.2 is produced by the use of Fab'-SH
derivatives which direct the coupling of discrete Fab' and prevent
the formation of bivalent monospecific F(ab').sub.2. One suitable
chemistry for mono- or bi-specific F(ab').sub.2 is that of Brennan
et al., supra. A Fab'-TNB derivative is prepared in a similar
manner to Fab'-SH, except that Fab'-SH released from freeze-thawed
cells by osmotic shock is adjusted to about pH 7.5 in the presence
of excess DTNB. Equimolar quantities of Fab'-SH and Fab'-TNB are
coupled together efficiently to form the F(ab').sub.2 fragment by a
disulfide exchange reaction in the presence of EDTA.
[0068] The linkless Fv, Fab', Fab'-SH and F(ab').sub.2 polypeptides
of this invention do not have a disulfide bond between the light
and heavy chains. Typically, the polypeptide is modified as the
result of direct expression of a modified polypeptide, or by
chemical or enzymatic means. Typical amino acid sequence
modifications of the polypeptide involve substituting or deleting
one or, preferably, both of the two cysteines which form the
inter-chain (heavy-light) disulfide bond. Typically, these
cysteines are replaced with serines, although it is within the
scope hereof to covalently modify one or both of the cysteine side
chain so as to be incapable of forming a disulfide bond.
[0069] One or both of the cysteines may be modified with fatty
acids or other chemical groups and rendered incapable of forming a
disulfide bond by a variety of known methods, including but not
limited to the following methods: (1) covalent modification using
dehydrating or activating agents such as
N,N'-dicyclohexylcarbodiimide (DCC) or ethyoxy-ethoxycarbonyl-dih-
ydroquinoline (EEDQ); (2) acylation, using ketenes, anhydrides,
isothiocyanates, or beta-lactones; (3) carbamoylation using
cyanates; (4) hemimercaptal or hemimercaptol formation using
aldehydes and some keto acids; (5) alkylation and arylation, by
addition to activated double bonds (using N-ethylmaleimide), by
reactions with quinones, by reaction with haloacids and their
amides (using iodoacetic acid, or alpha-bromo-hexadecanoic acid),
by methylation reactions (such as with dimethylsulfate), by
sulfoalkylation, by arylation (with nitrobenzene compounds), or by
reaction with diazo compounds; (6) reaction with metal ions such as
silver and organic mercury compounds such as mercuric chloride; (7)
reaction with arsenic compounds; (8) reaction with sulfites; (9)
oxidation reactions; and (10) reaction with sulfenyl halides to
form mixed disulfides.
[0070] The polypeptide may be modified within a host cell, or as a
post-translational modification to the recombinantly produced
polypeptide. It is currently preferred that any post-translational
modifications take place within 24, and preferably within a few
hours of recovery from a host cell or microorganism.
[0071] Methods for detecting the presence of modifications to the
cysteines of the polypeptides of this invention lung surfactant
protein are commonly known, such as through the analysis of the
mass spectra of a sample of the polypeptide. Alternatively, thin
layer chromatography (TLC) is performed on a sample. While this
approach to peptide analysis is common in the field, see e.g.
Stuart and Young, Solid Phase Peptide Synthesis, pp 103-107, and
118-122, (Pierce Chem. Co., 2d. ed., 1984).
[0072] In certain embodiments, there is a very strong interaction
between the light and heavy chains such that eliminating the
interchain disulfide does not result undesirable levels of
dissociation between the light and heavy chains. The linkless
Fab'-SH polypeptides advantageously allow homogeneous F(ab').sub.2
to be prepared by chemical coupling.
[0073] The Fab sequences of this invention are obtained from
conventional sources of antibodies. Polyclonal antibodies to an
antigen generally are raised in animals by multiple subcutaneous
(sc) or intraperitoneal (ip) injections of the antigen and an
adjuvant. It may be useful to conjugate the antigen or a fragment
containing the target antigen amino acid sequence to a protein that
is immunogenic in the species to be immunized, e.g., keyhole limpet
hemocyanin, serum albumin, bovine thyroglobulin, or soybean trypsin
inhibitor using a bifunctional or derivatizing agent, for example,
maleimidobenzoyl sulfosuccinimide ester (conjugation through
cysteine residues), N-hydroxysuccinimide (through lysine residues),
glutaraldehyde, succinic anhydride, SOCl.sub.2, or
R.sup.1N.dbd.C.dbd.NR, where R and R.sup.1 are different alkyl
groups.
[0074] The route and schedule of immunization of the animal or
cultured antibody-producing cells therefrom are generally in
keeping with established and conventional techniques for antibody
stimulation and production. While mice are frequently employed as
the test model, it is contemplated that any mammalian subject or
antibody-producing cells obtained therefrom can be employed.
[0075] After immunization, monoclonal antibodies are prepared by
recovering immune lymphoid cells--typically spleen cells or
lymphocytes from lymph node tissue--from immunized animals and
immortalizing the cells in conventional fashion, e.g. by fusion
with myeloma cells or by Epstein-Barr (EB)-virus transformation and
screening for clones expressing the desired antibody. The hybridoma
technique described originally by Kohler, G. and Milstein, C.,
Nature 256: 52-53 (1975) has been widely applied to produce hybrid
cell lines that secrete high levels of monoclonal antibodies
against many specific antigens. Hybridomas secreting the desired
antibody are identified by conventional methods. Routine methods
are then employed to obtain DNA from hybridomas that encode the
heavy and light chains of the selected antibody. Alternatively, one
extracts antibody-specific messenger RNAs from B-cells of an
immunized animal, reverse transcribes these into complementary DNA
(cDNA), and amplifies the cDNA by PCR or by cloning it in a
bacterial expression system. Another technique suitable for
obtaining source heavy and light chain sequences uses a
bacteriophage lambda vector system (which contains a leader
sequence that secretes the expressed Fab protein into the
periplasmic space) together with the generation and screening of
great numbers of functional antibody fragments for those which
demonstrate the desired activity. This system is commercially
available.
[0076] Amino Acid Sequence Variants
[0077] This invention also encompasses amino acid sequence variants
of the native Fab polypeptide sequences. These variants are
prepared by introducing appropriate nucleotide changes into the DNA
encoding the Fab or by in vitro synthesis of the desired Fab. Such
variants include, for example, humanized variants of non-human
antibodies, as well as deletions from, or insertions or
substitutions of, residues within particular amino acid sequences.
Any combination of deletion, insertion, and substitution can be
made to arrive at the final construct, provided that the final
construct possesses the desired characteristics. The amino acid
changes also may alter post-translational processing of the target
polypeptide, such as changing the number or position of
glycosylation sites, introducing a membrane anchoring sequence into
the constant domain or modifying the leader sequence of the native
Fab.
[0078] In designing amino acid sequence variants of target
polypeptides, the location of the mutation site and the nature of
the mutation will depend on the target polypeptide
characteristic(s) to be modified. The sites for mutation can be
modified individually or in series, e.g., by (1) substituting first
with conservative amino acid choices and then with more radical
selections depending upon the results achieved, (2) deleting the
target residue, or (3) inserting residues of the same or a
different class adjacent to the located site, or combinations of
options 1-3.
[0079] A useful method for identification of certain residues or
regions of the target polypeptide that are preferred locations for
mutagenesis is called "alanine scanning mutagenesis" as described
by Cunningham and Wells (Science, 244: 1081-1085 [1989]), and
Duncan, A. R. and Winter, G. (Nature, 322: 738-740 [1988]). Here, a
residue or group of target residues are identified (e.g., charged
residues such as arg, asp, his, lys, and glu) and replaced by a
neutral or negatively charged amino acid (most preferably alanine
or polyalanine) to affect the interaction of the amino acids with
the surrounding aqueous environment in or outside the cell. Thus,
while the site for introducing an amino acid sequence variation is
predetermined, the nature of the mutation per se need not be
predetermined. For example, to optimize the performance of a
mutation at a given site, Ala scanning or random mutagenesis may be
conducted at the target codon or region and the expressed target
polypeptide variants are screened for the optimal combination of
desired activity.
[0080] There are two principal variables in the construction of
amino acid sequence variants: the location of the mutation site and
the nature of the mutation. In general, the location and nature of
the mutation chosen will depend upon the target polypeptide
characteristic to be modified.
[0081] Amino acid sequence insertions include amino- and/or
carboxyl-terminal fusions ranging in length from one residue to
polypeptides containing a hundred or more residues, as well as
intrasequence insertions of single or multiple amino acid residues.
Intrasequence insertions (i.e., insertions within the target
polypeptide sequence) may range generally from about 1 to 10
residues, more preferably 1 to 5, most preferably 1 to 3. Examples
of terminal insertions include fusion of a heterologous N-terminal
signal sequence to the N-terminus of the Fv polypeptide to
facilitate the secretion of the mature Fv polypeptide from
recombinant host cells. Such signal sequences generally will be
obtained from, and thus homologous to, the intended host cell
species. Suitable sequences for E. coli include STII or Ipp.
[0082] Other insertional variants of the target polypeptide include
the fusion to the N- or C-terminus of the target polypeptide of
immunogenic polypeptides, e.g., bacterial polypeptides such as
beta-lactamase or an enzyme encoded by the E. coli trp locus, or
yeast protein, and C-terminal fusions with proteins having a long
half-life such as immunoglobulin constant regions (or other
immunoglobulin regions), albumin, or ferritin, as described in WO
89/02922 published 6 Apr. 1989. Additional suitable polypeptide
sequences include but are not limited to sequences derived from
immunoglobulins, carrier proteins, receptors, growth factors,
nutrient substances, cellular adhesive molecules, tissue-specific
adhesion factors, enzymes, toxins and the like. These examples will
be in addition to insertions of cysteine or cysteine-containing
polypeptides such as hinge regions which provide free thiol
cysteinyl.
[0083] Another group of variants are amino acid substitution
variants. These variants have at least one amino acid residue in
the immunoglobulin polypeptide removed and a different residue
inserted in its place. The sites of greatest interest for
substitutional mutagenesis include the CDRs, FR and hinge regions.
They include substitutions of cysteine for other residue and
insertions which are substantially different in terms of side-chain
bulk, charge, and/or hydrophobicity. Other sites for substitution
are described infra, considering the effect of the substitution of
the antigen binding, affinity and other characteristics of a
particular target antibody.
[0084] Fab itself is a deletional variant of intact immunoglobulin
in which the heavy chain constant domain downstream from C.sub.H1
is deleted. Further, in preferred embodiments the C.sub.H1 domain
is followed C-terminally by a cysteine-containing sequence such as
Cys Ala Ala, or sequences having more than one Cys.
[0085] DNA encoding Fab amino acid sequence variants is prepared by
a variety of methods known in the art. These methods include, but
are not limited to, isolation from a natural source (in the case of
naturally occurring amino acid sequence variants) or preparation by
oligonucleotide-mediated (or site-directed) mutagenesis, PCR
mutagenesis, and cassette mutagenesis of an earlier prepared
variant or a non-variant version of the target polypeptide or by
total gene synthesis. These techniques may utilize target
polypeptide nucleic acid (DNA or RNA), or nucleic acid
complementary to the target polypeptide nucleic acid.
Oligonucleotide-mediated mutagenesis is a preferred method for
preparing substitution, deletion, and insertion variants of target
polypeptide DNA.
[0086] PCR mutagenesis is also suitable for making amino acid
variants of the Fab polypeptide. While the following discussion
refers to DNA, it is understood that the technique also finds
application with RNA. The PCR technique generally refers to the
following procedure (see Erlich, supra, the chapter by R. Higuchi,
p. 61-70): When small amounts of template DNA are used as starting
material in PCR, primers that differ slightly in sequence from the
corresponding region in a template DNA can be used to generate
relatively large quantities of a specific DNA fragment that differs
from the template sequence only at the positions where the primers
differ from the template. For introduction of a mutation into a
plasmid DNA, one of the primers is designed to overlap the position
of the mutation and to contain the mutation; the sequence of the
other primer must be identical to a stretch of sequence of the
opposite strand of the plasmid, but this sequence can be located
anywhere along the plasmid DNA. It is preferred, however, that the
sequence of the second primer is located within 200 nucleotides
from that of the first, such that in the end the entire amplified
region of DNA bounded by the primers can be easily sequenced. PCR
amplification using a primer pair like the one just described
results in a population of DNA fragments that differ at the
position of the mutation specified by the primer, and possibly at
other positions, as template copying is somewhat error-prone.
[0087] If the ratio of template to product material is extremely
low, the vast majority of product DNA fragments incorporate the
desired mutation(s). This product material is used to replace the
corresponding region in the plasmid that served as PCR template
using standard DNA technology. Mutations at separate positions can
be introduced simultaneously by either using a mutant second
primer, or performing a second PCR with different mutant primers
and ligating the two resulting PCR fragments simultaneously to the
vector fragment in a three (or more)-part ligation.
[0088] Another method for preparing variants, cassette mutagenesis,
is based on the technique described by Wells et al. (Gene, 34: 315
[1985]). The starting material is the plasmid (or other vector)
comprising the Fab DNA to be mutated. The codon(s) in the Fab
polypeptide DNA to be mutated are identified. There must be a
unique restriction endonuclease site on each side of the identified
mutation site(s). If no such restriction sites exist, they may be
generated using the above-described oligonucleotide-mediated
mutagenesis method to introduce them at appropriate locations in
the Fab polypeptide DNA. After the restriction sites have been
introduced into the plasmid, the plasmid is cut at these sites to
linearize it. A double-stranded oligonucleotide encoding the
sequence of the DNA between the restriction sites but containing
the desired mutation(s) is synthesized using standard procedures.
The two strands are synthesized separately and then hybridized
together using standard techniques. This double-stranded
oligonucleotide is referred to as the cassette. This cassette is
designed to have 3' and 5' ends that are compatible with the ends
of the linearized plasmid, such that it can be directly ligated to
the plasmid. This plasmid now contains the mutated Fab polypeptide
DNA sequence.
[0089] Insertion of DNA into a Vector
[0090] The cDNA or genomic DNA encoding the Fab polypeptide is
inserted into a replicable vector for further cloning
(amplification of the DNA) or for expression. Many vectors are
available, and selection of the appropriate vector will depend on
1) whether it is to be used for DNA amplification or for expression
of the encoded protein, 2) the size of the DNA to be inserted into
the vector, and 3) the host cell to be transformed with the vector.
Each vector contains various components depending on its function
(amplification of DNA or expression of DNA) and the host cell for
which it is compatible. The vector components generally include,
but are not limited to, one or more of the following: a signal
sequence, an origin of replication, one or more marker genes, a
promoter, and a transcription termination sequence.
[0091] (a) Signal Sequence Component
[0092] In general, the signal sequence may be a component of the
vector, or it may be a part of the target polypeptide DNA that is
inserted into the vector. Included within the scope of this
invention are Fab polypeptides with any native signal sequence
deleted and replaced with a heterologous signal sequence. The
heterologous signal sequence selected should be one that is
recognized and processed (i.e. cleaved by a signal peptidase) by
the host cell. For prokaryotic host cells that do not recognize and
process the native Fab polypeptide signal sequence, the signal
sequence is substituted by a prokaryotic signal sequence selected,
for example, from the group of the alkaline phosphatase,
penicillinase, lpp, or heat-stable enterotoxin II leaders.
[0093] (b) Origin of Replication Component
[0094] Expression and cloning vectors may, but need not, contain a
nucleic acid sequence that enables the Fab nucleic acid to
replicate in one or more selected host cells. Generally, in cloning
vectors this sequence is one that enables the vector to replicate
independently of the host chromosomal DNA, and includes origins of
replication or autonomously replicating sequences. Such sequences
are well known for a variety of microbes. The origin of replication
from the plasmid pBR322 is suitable for most Gram-negative
bacteria.
[0095] DNA may also be replicated by insertion into the host
genome. This is readily accomplished using Bacillus species as
hosts, for example, by including in the vector a DNA sequence that
is complementary to a sequence found in Bacillus genomic DNA.
Transfection of Bacillus with this vector results in homologous
recombination with the genome and insertion of the target
polypeptide DNA. However, the recovery of genomic DNA encoding the
target polypeptide is more complex than that of an exogenously
replicated vector because restriction enzyme digestion is required
to excise the target polypeptide DNA. Similarly, DNA also can be
inserted into the genome of vertebrate and mammalian cells by
conventional methods.
[0096] (c) Selection Gene Component
[0097] Expression and cloning vectors should contain a selection
gene, also termed a selectable marker. This gene encodes a protein
necessary for the survival or growth of transformed host cells
grown in a selective culture medium. Host cells not transformed
with the vector containing the selection gene will not survive in
the culture medium. Typical selection genes encode proteins that
(a) confer resistance to antibiotics or other toxins, e.g.
ampicillin, neomycin, methotrexate, or tetracycline, (b) complement
auxotrophic deficiencies, or (c) supply critical nutrients not
available from complex media, e.g. the gene encoding D-alanine
racemase for Bacilli.
[0098] One example of a selection scheme utilizes a drug to arrest
growth of a host cell. Those cells that are successfully
transformed with a heterologous gene express a protein conferring
drug resistance and thus survive the selection regimen. Examples of
such dominant selection use the drugs neomycin (Southern et al., J.
Molec. Appl. Genet., 1: 327 [1982]), mycophenolic acid (Mulligan et
al., Science, 209: 1422 [1980]) or hygromycin (Sugden et al., Mol.
Cell. Biol., 5: 410-413 [1985]). The three examples given above
employ bacterial genes under eukaryotic control to convey
resistance to the appropriate drug G418 or neomycin (geneticin),
xgpt (mycophenolic acid), or hygromycin, respectively.
[0099] (d) Promoter Component
[0100] Expression and cloning vectors will usually contain a
promoter that is recognized by the host organism and is operably
linked to the Fab polypeptide nucleic acid. Promoters are
untranslated sequences located upstream (5') to the start codon of
a the Fab structural gene (generally within about 100 to 1000 bp)
that control its transcription and translation. Such promoters
typically fall into two classes, inducible and constitutive.
Inducible promoters are promoters that initiate increased levels of
transcription from DNA under their control in response to some
change in culture conditions, e.g. the presence or absence of a
nutrient or a change in temperature.
[0101] Inducible promoters under high regulation are preferred for
the microbial expression of Fv-containing polypeptides. At this
time a large number of promoters recognized by a variety of
potential host cells are well known. These promoters are operably
linked to DNA encoding the Fab polypeptide by removing the promoter
from the source DNA by restriction enzyme digestion and inserting
the isolated promoter sequence into the vector. Both the native Fab
polypeptide promoter sequence and many heterologous promoters may
be used to direct amplification and/or expression of the Fab
polypeptide DNA. However, heterologous promoters are preferred, as
they generally permit greater transcription and higher yields of
expressed target polypeptide as compared to the native target
polypeptide promoter.
[0102] Promoters suitable for use with prokaryotic hosts include
the .beta.-lactamase and lactose promoter systems (Chang et al.,
Nature, 275: 615 [1978]; and Goeddel et al., Nature, 281: 544
[1979]), alkaline phosphatase, a tryptophan (trp) promoter system
(Goeddel, Nucleic Acids Res., 8: 4057 [1980] and EP 36,776) and
hybrid promoters such as the tac promoter (deBoer et al., Proc.
Natl. Acad. Sci. USA, 80: 21-25 [1983]). However, other known
bacterial promoters are suitable. Their nucleotide sequences have
been published, thereby enabling a skilled worker operably to
ligate them to DNA encoding the target polypeptide (Siebenlist et
al., Cell, 20: 269 [1980]) using linkers or adaptors to supply any
required restriction sites. Promoters for use in bacterial systems
also generally will contain a Shine-Dalgarno (S.D.) sequence
operably linked to the DNA encoding the target polypeptide.
[0103] Construction of suitable vectors containing one or more of
the above listed components employs standard ligation techniques.
Isolated plasmids or DNA fragments are cleaved, tailored, and
religated in the form desired to generate the plasmids
required.
[0104] Selection and Transformation of Host Cells
[0105] Suitable host cells for expressing Fab are microbial cells
such as yeast, fungi, and prokaryotes. Suitable prokaryotes include
eubacteria, such as Gram-negative or Gram-positive organisms, for
example, E. coli, Bacilli such as B. subtilis, Pseudomonas species
such as P. aeruginosa, Salmonella typhimurium, or Serratia
marcescans. One preferred E. coli cloning host is E. coli 294 (ATCC
31,446), although other strains such as E. coli B, E. coli
.sub.X1776 (ATCC 31,537), E. coli RV308 (ATCC 31,608) and E. coli
W3110 (ATCC 27,325) are suitable. These examples are illustrative
rather than limiting. Preferably the host cell should secrete
minimal amounts of proteolytic enzymes, and additional protease
inhibitors may desirably be incorporated in the cell culture.
[0106] Host cells are transfected and preferably transformed with
the above-described expression or cloning vectors of this invention
and cultured in conventional nutrient media modified as appropriate
for inducing promoters, selecting transformants, or amplifying the
genes encoding the desired sequences.
[0107] Culturing the Host Cells
[0108] Cells used to produce the Fab polypeptide of this invention
are cultured in suitable media as described generally in Sambrook
et al., (Molecular Cloning: A Laboratory Manual New York: Cold
Spring Harbor Laboratory Press, 1989). Any other necessary
supplements may also be included at appropriate concentrations that
would be known to those skilled in the art. The culture conditions,
such as temperature, pH, and the like, are those previously used
with the host cell selected for expression, and will be apparent to
the ordinarily skilled artisan.
[0109] It is currently preferred that the bacterial host cells be
cultured at temperatures from 37.degree. C. to 29.degree. C.,
although temperatures as low as 20.degree. C. may be suitable.
Optimal temperatures will depend on the host cells, the Fab
sequence and other parameters. 37.degree. C. is generally
preferred.
[0110] Purification of Fab Polypeptides
[0111] Soluble polypeptides are recovered from recombinant cell
culture to obtain preparations that are substantially homogeneous
as to Fab. As a first step, the culture medium or periplasmic
preparation is centrifuged to remove particulate cell debris.
Periplasmic preparations are obtained in conventional fashion, e.g.
by freeze-thaw or osmotic shock methods. The membrane and soluble
protein fractions are then separated. The Fab polypeptide is then
purified from the soluble protein fraction. The following
procedures are exemplary of suitable purification procedures:
fractionation on immunoaffinity or ion-exchange columns; ethanol
precipitation; reverse phase HPLC; chromatography on silica or on a
cation exchange resin such as DEAE; chromatofocusing; SDS-PAGE;
ammonium sulfate precipitation; gel filtration using, for example,
Sephadex G-75; protein A or protein G affinity matrix (e.g.
Sepharose) columns; and hydrophobic interaction chromatography.
[0112] Fab polypeptide variants in which residues have been
deleted, inserted or substituted are recovered in the same fashion,
taking account of any substantial changes in properties occasioned
by the variation. For example, preparation of a Fab polypeptide
fusion with another protein or polypeptide, e.g. a bacterial or
viral antigen, facilitates purification since an immunoaffinity
column containing antibody to the antigen is used to adsorb the
fusion. Immunoaffinity columns such as a rabbit polyclonal
anti-target polypeptide column can be employed to absorb the target
polypeptide variant by binding it to at least one remaining immune
epitope. A protease inhibitor also is useful to inhibit proteolytic
degradation during purification, and antibiotics may be included to
prevent the growth of adventitious contaminants.
[0113] Utility of the Fab-Containing Polypeptides
[0114] The antibody fragments of this invention are useful in
diagnostic assays for antigen in specific cells, fluids or tissues,
for immunoaffinity purification of the antigens and for therapies
which are based on antigen antagonism.
[0115] Analytical methods for the antigen bound by the Fab
polypeptide are conventional and may use a label bound to the Fab.
The label used with the Fab polypeptide is any detectable
functionality that does not interfere with its binding to Fab.
Numerous labels are known, including the radioisotopes .sup.32P,
.sup.32S, .sup.14C, .sup.125I, .sup.3H, and .sup.131I, fluorophores
such as rare earth chelates or fluorescein and its derivatives,
rhodamine and its derivatives, dansyl, umbelliferone,
luceriferases, e.g., firefly luciferase and bacterial luciferase
(U.S. Pat. No. 4,737,456), luciferin, 2,3-dihydrophthalazinediones,
horseradish peroxidase (HRP), alkaline phosphatase,
.beta.-galactosidase, glucoamylase, lysozyme, saccharide oxidases,
e.g., glucose oxidase, galactose oxidase, and glucose-6-phosphate
dehydrogenase, heterocyclic oxidases such as uricase and xanthine
oxidase, lactoperoxidase, biotin/avidin, spin labels, bacteriophage
labels, stable free radicals, imaging radionuclides (such as
Technetium) and the like.
[0116] Conventional methods are available to bind these labels
covalently to proteins or polypeptides. For instance, coupling
agents such as dialdehydes, carbodiimides, dimaleimides,
bis-imidates, bis-diazotized benzidine, and the like may be used to
tag the antibodies with the above-described fluorescent,
chemiluminescent, and enzyme labels. See, for example, U.S. Pat.
No. 3,940,475 (fluorimetry) and U.S. Pat. No. 3,645,090 (enzymes);
Hunter et al., Nature, 144: 945 (1962); David et al., Biochemistry,
13: 1014-1021 (1974); Pain et al., J. Immunol. Methods, 40: 219-230
(1981); and Nygren, J. Histochem, and Cytochem., 30: 407-412
(1982). Preferred labels herein are enzymes such as horseradish
peroxidase and alkaline phosphatase. The conjugation of such label,
including the enzymes, to the Fab-containing polypeptide is a
standard manipulative procedure for one of ordinary skill in
immunoassay techniques. See, for example, O'Sullivan et al.,
"Methods for the Preparation of Enzyme-antibody Conjugates for Use
in Enzyme Immunoassay," in Methods in Enzymology, ed. J. J. Langone
and H. Van Vunakis, Vol. 73 (Academic Press, New York, N.Y., 1981),
pp. 147-166. Such bonding methods are suitable for use with the Fab
polypeptides of this invention.
[0117] The Fab-containing polypeptides also may comprise an
immunotoxin. For example, the Fab heavy chain is optionally
conjugated to a cytotoxin such as ricin for use in AIDS therapy.
Alternatively, the toxin may be a cytotoxic drug or an
enzymatically active toxin of bacterial, fungal, plant or animal
origin, or an enzymatically active fragment of such a toxin.
Enzymatically active toxins and fragments thereof include
diphtheria A chain, nonbinding active fragments of diphtheria
toxin, exotoxin A chain (from Pseudomonas aeruginosa), ricin A
chain, abrin A chain, modeccin A chain, alpha-sarcin, Aleurites
fordii proteins, dianthin proteins, Phytolaca americana proteins
(PAPI, PAPII, and PAP-S), momordica charantia inhibitor, curcin,
crotin, sapaonaria officinalis inhibitor, gelonin, mitogellin,
restrictocin, phenomycin, enomycin and the tricothecenes. In
another embodiment, the antibodies are conjugated to nonpeptidyl
drugs such as cis-platin or 5FU. Conjugates of the monoclonal
antibody and such cytotoxic moieties are made using a variety of
bifunctional protein coupling agents. Examples of such reagents are
SPDP, IT, bifunctional derivatives of imidoesters such as dimethyl
adipimidate HCl, active esters such as disuccinimidyl suberate,
aldehydes such as glutaraldehyde, bis-azido compounds such as bis
(p-azidobenzoyl) hexanediamine, bis-diazonium derivatives such as
bis-(p-diazoniumbenzoyl)- - -ethylenediamine, diisocyanates such as
tolylene 2,6-diisocyanate and bis-active fluorine compounds such as
1,5-difluoro-2,4-dinitrobenzene. The lysing portion of a toxin may
be joined to the Fab antibody fragment.
[0118] Immunotoxins can be made in a variety of ways, as discussed
herein. Commonly known crosslinking reagents can be used to yield
stable conjugates.
[0119] When used in vivo for therapy, the Fab-containing fragments
of the subject invention are administered to the patient in
therapeutically effective amounts (i.e. amounts that have desired
therapeutic effect) in the same fashion as intact immunoglobulins.
The products prepared in accord with the methods of this invention
offer the advantage of substantial molecular homogeneity and are
devoid of toxic contaminants heretofore used in preparing
F(ab').sub.2.
[0120] The antibody compositions used in therapy are formulated and
dosages established in a fashion consistent with good medical
practice taking into account the disorder to be treated, the
condition of the individual patient, the site of delivery of the
composition, the method of administration and other factors known
to practitioners. The antibody compositions are prepared for
administration according to the description of preparation of
polypeptides for administration, infra.
EXAMPLE
Expression of Active Fab, Fab', and F(ab').sub.2, Antibody
Fragments
[0121] Overexpression of the HER2 proto-oncogene product
(p185.sup.HER2) has been associated with a variety of aggressive
human malignancies. An Escherichia coli expression system has been
developed that secretes functional Fab and Fab' fragments of a
humanized antibody, huMAb4D5-8, at titers of about 1 to in excess
of about 2 grams per liter as judged by binding to antigen,
p185.sup.HER2. The Fab' fragment was recovered with the single
hinge region cysteine present mainly as the free thiol (up to about
90 mole %) permitting efficient directed disulfide bond formation
in vitro to form the bivalent F(ab').sub.2 antibody fragment. This
molecule is indistinguishable from F(ab').sub.2 derived from
proteolysis of intact antibody in antigen binding affinity and in
anti-proliferative activity against the human breast tumor cell
line, SK-BR-3, which over-expresses p185.sup.HER2, but unlike the
proteolytic product, the F(ab').sub.2 here is C-terminally
homogenous. This invention facilitates the construction of
monospecific and bispecific F(ab').sub.2 antibody fragments,
including naturally derived or humanized antibody fragments for
research and therapeutic purposes. This invention is particularly
applicable for developing antibody fragments capable of directing
diagnostic or therapeutic moieties to target antigens such as tumor
foci characterized by overexpression of the HER2
proto-oncogene.
[0122] The murine monoclonal antibody known as muMAb4D5 (Fendly, B.
M. et al., Cancer Res. 50:1550-1558 (1990)) is directed against the
extracellular domain (ECD) of p185.sup.HER2. The muMAb4D5 and its
uses are described in copending PCT application WO 89/06692
published 27 Jul. 1989. This murine antibody was deposited with the
ATCC and designated ATCC CRL 10463. In this description, the terms
muMAb4D5, chMAb4D5 and huMAb4D5 represent murine, chimerized and
humanized versions of the monoclonal antibody 4D5,
respectively.
[0123] muMAb4D5 has potential for clinical intervention in that it
specifically inhibits the growth of human tumor cell lines
overexpressing p185.sup.HER2 (2, 3) and is rapidly internalized by
target cells (data not shown). Amplification and/or overexpression
of HER2 is associated with multiple human malignancies and appears
to be integrally involved in progression of 25-30% of primary human
breast and ovarian cancers (5). The muMAb4D5 molecule was
previously "humanized" (6) in an attempt to improve its clinical
efficacy by reducing immunogenicity and allowing it to support
human effector functions. The humanized antibody, huMAb4D5-8,
contains essentially only the antigen binding loops from the murine
parent antibody together with human variable region framework
residues and constant domains. This humanized antibody has 3-fold
higher affinity for p185.sup.HER2 ECD than the murine parent,
comparable anti-proliferative activity against p185.sup.HER2
overexpressing tumor cells and supports antibody-dependent cellular
cytotoxicity.
[0124] It was desired to enlarge the repertoire of
anti-p185.sup.HER2 antibody reagents to include the Fab and
F(ab').sub.2 antibody fragments of huMAb4D5-8. The smaller size of
these fragments compared to the intact antibody is likely to
enhance their specific localization to solid tumors by improving
tumor penetration and promoting more rapid removal from serum
(reviewed in ref. 8). huMAb4D5-8 Fab and F(ab').sub.2, fragments
were obtained by direct recombinant expression to facilitate
engineering of these fragments. Furthermore this strategy provided
more homogeneous preparations of antibody fragments than could be
obtained by limited proteolysis and partial reduction of intact
antibodies (reviewed in ref. 9). Functional Fv and Fab fragments
had been secreted from E. coli (10).
[0125] The strategy here for the E. coli secretion of antibody
fragments (FIG. 1) shares two basic similarities with the work of
others (10). Firstly a dicistronic operon is used to direct the
co-expression of corresponding light and heavy chain fragments.
Secondly the antibody chains are preceded by bacterial signal
sequences to direct secretion into the periplasmic space of E. coli
where the redox environment favors disulfide bond formation and the
light and heavy chain fragments may assemble. The system here
differs from earlier strategies in three basic ways. Firstly the
transcription unit utilizes the a highly regulated promoter, the E.
coli PhoA promoter (11) inducible by phosphate starvation, and
heat-stable enterotoxin II signal sequence (12). Secondly, the gene
segment for the light chain precedes that for the heavy chain Fd
fragment (V.sub.H and C.sub.H1 domains). Thirdly, in order to
express the Fab' fragment of huMAb4D5-8 the C.sub.H1 gene segment
was extended to encode part of the cysteine-containing antibody
hinge region. The sequence Cysteine followed by two Prolines and
another Cysteine (CPC terminus) was initially chosen since it is
found in the hinge region of human IgG.sub.1 molecules (17)
including the full length version of huMAb4D5-8 (6). The
construction of additional Fab' variants by cassette mutagenesis
(18) of the pBR322-based expression vector was facilitated by
installing unique Sal I and Sph I restriction sites towards the end
of the C.sub.H1 gene segment and immediately 3' to the stop codon,
respectively.
[0126] The huMAb4D5-8 Fab fragment was expressed in a phage
resistant derivative of E. coli RV308 (19) grown at high cell
density in a fermentor (20). The titer of functional huMAb4D5-8 Fab
in fermentation media is routinely 1 to 2 grams per liter as judged
by p185.sup.HER2 ECD binding ELISA (6). Modest amounts of
huMAb4D5-8 Fab (usually <200 mg/l) are found associated with the
cell paste and may be released by osmotic shock. The consensus
framework region of the huMAb4D5-8 Fab fragment was found to bind
tightly to both staphylococcal protein A and to streptococcal
protein G, allowing their use for affinity purification. Very
similar estimates of titers are obtained for culture media or cell
paste samples after affinity purification on protein A prior to
antigen-binding ELISA. Similar expression titers of p185.sup.HER2
ECD binding activity have been obtained for the huMAb4D5-8 Fab
variant (cysteine, two prolines and another cysteine) and the
additional Fab' variant described below.
[0127] Formation of F(ab').sub.2 molecules requires chance
encounters of Fab' hinge cysteinyl thiols to form disulfide bonds
without the assistance of extensive interactions between C.sub.H3
domains possible in the case of intact antibodies. Thus high level
expression of Fab' in the periplasmic space of E. coli was
anticipated to drive formation of F(ab').sub.2 in vivo. In fact
<10% of the Fab' molecule having the CPC terminus (isolated from
either media or cell paste) were recovered as the bivalent form as
judged by SDS-PAGE analysis after protein A purification. High
resolution mass spectrometry and other studies suggested that
substantial formation of an intramolecular disulfide bond between
the two hinge cysteine residues had occurred. This possibility was
precluded by the construction of an additional Fab' variant with a
single hinge cysteine residue having the C-terminal sequence, Cys
Ala Ala. Negligible quantities of F(ab').sub.2 are formed when this
Fab' variant is secreted from E. coli and DTNB analysis does not
detect free thiol.
[0128] The Fab' molecules are recovered under conditions that
maintain the hinge cysteine present as the free thiol and then
readily and efficiently formed into F(ab').sub.2 by directed
coupling in vitro. For example, Brennan et al. (23) reacted a Fab'
free thiol (Fab'-SH) with DTNB to form the thionitrobenzoate
derivative (Fab'-TNB) which was then coupled to a second
Fab'(Fab'-SH) to form the bispecific F(ab').sub.2. The following
strategy allows the routine purification of intact functional
huMAb4D5-8 Fab' Cys Ala Ala variant with 75-90% of the molecules
containing a free hinge thiol as judged by DTNB analysis (FIG. 2):
firstly, growth conditions were judiciously modified (20) to target
secretion of Fab' to the periplasmic space of E. coli rather than
into the culture media where the Fab' hinge thiol was found to be
quantitatively and covalently blocked. Secondly, the Fab' fragment
was isolated from the cell paste and affinity purified on protein G
sepharose at low pH (pH 5.0) to maintain the cysteinyl thiol in the
less reactive protonated form. Thirdly, EDTA was added to chelate
metal ions capable of catalyzing disulfide bond formation and to
inactivate metallo proteases. Finally a cocktail of additional
protease inhibitors (phenylmethylsulfonyl fluoride (PMSF),
leupeptin, pepstatin and benzamidine) virtually eliminated
proteolysis of Fab' by E. coli proteases during the purification.
Small amounts of contaminating proteolytic fragments in Fab'
preparations were readily removed by hydrophobic interaction
chromatography. The Fab'-TNB derivative was prepared in a similar
manner except that Fab'-SH released from freeze-thawed cells by
osmotic shock was adjusted to neutral pH in the presence of excess
DTNB.
[0129] Equimolar quantities of Fab'-SH and Fab-TNB were coupled
together to efficiently form the F(ab').sub.2 antibody fragment by
a disulfide exchange reaction (FIG. 2). The coupling reaction was
followed by monitoring the increase in absorbance at 412 nm upon
release of the thionitrobenzoate anion and found to reach
completion after 30 to 60 min at 37.degree. C. No free thiol was
detected in the reaction mix after coupling and the amount of
remaining Fab' is consistent with that amount of unreactive
material in the Fab'-SH and Fab'-TNB preparations. F(ab').sub.2 was
separated from Fab' by gel filtration on a S100-HR sizing column.
Only trace quantities of huMAb4D5-8 F(ab').sub.2 are were formed in
mock coupling reactions containing either Fab'-SH or Fab'-TNB
alone, as anticipated (23).
[0130] The physical and chemical integrity of purified huMAb4D5-8
Fab and F(ab').sub.2 was evaluated by SDS-PAGE (FIG. 2), analysis
of the amino terminal sequence, amino acid composition plus free
thiol content, and by circular dichroism. Purified huMAb4D5-8 Fab
(M.sub.r=47.7 kdal) and F(ab').sub.2 (M.sub.r=96.0 kdal) fragments
analyzed by SDS-PAGE under non-reducing conditions each gave a
single major band of the expected mobility. After SDS-PAGE under
reducing conditions both Fab and F(ab').sub.2 antibody fragments
gave a doublet of bands of similar intensities as expected from
release of stoichiometric quantity of free light chain (23.4 kdal)
and heavy chain Fd (24.3 kdal) or Fd' (24.6 kdal) fragments (not
shown). Amino terminal sequence analysis (8 cycles) of Fab and
F(ab').sub.2 antibody fragments gave the expected mixed sequence
from a stoichiometric 1:1 mixture of light and heavy chains
(V.sub.L/V.sub.H: Asp/Glu, Ile/Val, Gln/Gln, Met/Leu, Thr/Val,
Gln/Glu, Ser/Ser, and Pro/Gly with no evidence of additional
sequences. No free thiol was detected in either Fab or F(ab').sub.2
preparations by DTNB analysis as expected. Amino acid analysis (27)
of acid hydrolysed Fab or F(ab').sub.2 was in excellent agreement
with the expected composition (6). The circular dichroism spectrum
of the Fab fragment is characteristic of an immunoglobulin
fold.
[0131] The function of huMAb4D5-8 Fab and F(ab').sub.2 antibody
fragments was investigated by measuring the binding affinity for
the p185.sup.HER2 ECD and by investigating their effect upon the
proliferation of the p185.sup.HER2 overexpressing human breast
carcinoma line, SK-BR-3 (Table 1).
1TABLE 1 Analysis of huMAb4D5-8 fragments by p185.sup.HER2 ECD
binding affinity and anti-proliferative activity with breast
carcinoma, SK-BR-3 cells. huMAb4D5-8 variant Source K.sub.d
.multidot. pM Relative cell proliferation.sup..dagger. Fab E. coli
570 91 F(ab').sub.2.sup..dagger-dbl. E. coli 290 53 F(ab').sub.2
293 cells 300 50 *K.sub.d values for the p185.sup.HER2 ECD were
determined as previously described (5) and the standard error of
the estimates are .ltoreq. .+-. 10%. .sup..dagger.Proliferation of
SK-BR-3 cells incubated for 96 hr with huMAb4D5 variants shown as a
percentage of the untreated control as described (5). Data
represent the maximal anti-proliferative effect for each variant
calculated as the mean of triplicate determinations at a fragment
concentration of 10 .mu.g/ml. Data are all taken from the same
experiment and the estimated standard error .ltoreq. .+-. 15%.
.sup..dagger-dbl.Cys Ala Ala variant.
[0132] The binding affinity of huMAb4D5-8 F(ab').sub.2 antibody
fragment for p185.sup.HER2 ECD is identical to that of the
corresponding fragment derived from limited proteolysis of whole
antibody expressed in mammalian cells. The bivalent F(ab').sub.2
antibody fragment derived from E. coli has identical
anti-proliferative activity with SK-BR-3 cells to both the intact
bivalent huMAb4D5-8 parent antibody derived from 293 cells (6) and
the F(ab').sub.2 antibody fragment derived from limited pepsin
digestion of intact antibody. In contrast, the monovalent Fab
molecule does not significantly affect the growth of SK-BR-3 cells.
This suggests that the crosslinking of p185.sup.HER2 on the surface
of cells may be required for inhibiting their proliferation. It
appears exceedingly unlikely that the 2-fold weaker antigen binding
affinity of the Fab compared to the F(ab').sub.2 could account for
the lack of anti-proliferative activity of the Fab at
concentrations up to several hundred fold above the K.sub.d.
Furthermore, it is possible to block the anti-proliferative
activity of the bivalent parent antibody, muMAb4D5, with a
monovalent huMAb4D4 Fab fragment.
[0133] In this example, the expression titer of functional Fab
fragments was increased compared to the reports in the literature
for E. coli, from about 1000-fold to 1 to 2 grams per liter.
Additionally, Fab' molecules were recovered. This enhanced
expression is very likely due only in part to the higher cell
densities (10 to 20-fold) and more precisely controlled environment
of the fermentor than the simple shake flask, the very tight
control of pre-induction expression and the characteristics of the
humanized variable domain sequence employed. The titer of Fab'
obtained here using the phoA promoter are surprisingly high and may
result from a combination of the use of this powerful promoter in a
low copy number (pBR322-based) vector. Thus the repressor is not
titered to less effective levels. The cells also could be
transformed to make surplus repressor. What is important is that
the promoter be inactive prior to induction. It is possible that
vector design and the high thermal stability of the Fab fragment
(T.sub.m>80.degree. C.) also may be important. This system
should greatly facilitate clinical or biophysical studies requiring
large quantities of antibody fragments.
[0134] Similar high expression levels have been observed for
several additional variants of huMAb4D5-8 Fab which contain one or
more amino acid changes in the antigen binding loops or nearby
framework residues (6). Two alternative antigen-binding
specificities have been recruited into huMAb4D5-8 Fab' by judicious
replacement of antigen binding residues and found to give high
expression titers. This is consistent with the notion that the
framework of a polypeptide comprising substantially huMAb4D5-8 Fab'
might be generically useful for highly expressed humanized Fab'
molecules.
[0135] The huMAb4D5-8 Fab' Cys Ala Ala shows very little tendency
to form F(ab').sub.2 in vivo despite the apparently quantitative
formation of intra-domain disulfides in the variable regions.
However, in vitro F(ab').sub.2 forms readily by air oxidation of
Fab'-SH at pH 7.5 in the absence of EDTA at concentrations that are
at least 10-fold lower than are found in vivo.
[0136] Without committing to any particular mechanism, it is
believed that the redox potential of the periplasmic space of E.
coli is sufficiently oxidizing to allow formation of the
intra-domain disulfide bonds but not the inter-heavy chain
disulfide which is presumably thermodynamically less favorable.
Nevertheless, recovery of functional Fab' fragments secreted into
the periplasmic space of E. coli with the unpaired hinge cysteine
mainly as the free thiol provides the essential starting material
for directed coupling (23, 32, 33). We have additionally exploited
the free thiol for immobilization of the huMAb4D5-8 Fab' fragment
on an activated thiol support as previously described (34),
enabling the p185.sup.HER2 ECD to be affinity purified from
solution. The free hinge thiol has also been used for attachment of
fluorescent probes for fluorescence-activated cell sorting. It is
also within the scope of this invention to use the free cysteinyl
thiol for the site-directed attachment of radionuclides for imaging
or therapy. This would offer the advantage over conventional
labelling strategies of a defined stoichiometry and attachment site
without the risk of compromising antigen binding affinity.
BIBLIOGRAPHY
[0137] 1. B. M. Fendly et al., Cancer Res. 50: 1550 (1990).
[0138] 2. R. M. Hudziak et al., Molec. Cell. Biol. 9: 1165
(1989).
[0139] 3. R. Lupu et al., Science 249: 1552 (1990).
[0140] 5. D. J. Slamon et al., Science 235: 177 (1987); D. J.
Slamon et al., Science 244: 707 (1989).
[0141] 6. P. Carter et al., Proc. Natl. Acad. Sci. U.S.A.
89(10):4285-9 (1992).
[0142] 7. P. T. Jones et al., Nature 321: 522 (1986); L. Riechmann
et al., Nature 332: 323 (1988); M. Verhoeyen et al., Science 239:
1534 (1988).
[0143] 8. R. D. Blumenthal et al., Adv. Drug Del. Rev. 4: 279
(1990).
[0144] 9. P. Parham in Cellular Immunology (E. M. Weir, Ed.,
Blackwell Scientific, CA) 4th edition, vol 1 chapter 14 (1983).
[0145] 10. A. Skerra and A. Pluckthun, Science 240: 1038 (1988); M.
Better et al., ibid p. 1041.
[0146] 11. C. N. Chang et al., Gene 44: 121 (1986).
[0147] 12. R. N. Picken et al., Infect. Immun. 42: 269 (1983).
[0148] 13. W. Palm and N. Hilschmann, Z. Physiol. Chem. 356: 167
(1975).
[0149] 14. J. W. Ellison et al., Nucleic Acids Res. 10: 4071
(1982).
[0150] 15. F. Bolivar et al., Gene 2: 95 (1977).
[0151] 16. S. Scholtissek, et al., Nucleic Acids Res. 15: 3185
(1987)
[0152] 17. E. A. Kabat, et al., Sequences of Proteins of
Immunological Interest 3rd edition (National Institutes of Health,
Bethesda, Md., 1987).
[0153] 18. J. A. Wells et al., Gene 34: 315 (1985).
[0154] 19. R. A. Maurer, Ph.D. thesis, Harvard University
(1978).
[0155] 20. The huMAb4D5-8 Fab' fragment was expressed in E. coli
strain 25F2 derived from the strain RV308 (ref. 22, ATCC#31608) by
inactivating the tonA gene. Cells were grown for 32-40 hours in an
aerated 10 liter fermentor at 37.degree. C. at a low agitation rate
(650 rpm, KIa .about.600 mmol l.sup.-1hr.sup.-1atm.sup.-1) in a
medium that initially contained 12 g l.sup.-1 digested casein, 17
mM glucose, 2.4 mM isoleucine hydrochloride, 47 mM
(NH.sub.4).sub.2SO.sub.4, 10 mM NaH.sub.2PO.sub.4, 18 mM
K.sub.2HPO.sub.4, 4.1 mM trisodium citrate, 12 mM MgSO.sub.4, 125
.mu.M FeCl.sub.3 and 20 .mu.M each of ZnSO.sub.4, MnSO.sub.4,
CuSO.sub.4, CoCl.sub.2, H.sub.3BO.sub.3 and NaMoO.sub.4) plus 12 mg
l.sup.-1 tetracycline and received automated feeds of ammonia to
maintain the pH at 7.0 and also glucose to maintain a slight excess
or avoid anaerobisis depending upon cell density of 80 to 100
OD.sub.550. The cell density at harvest is usually 120 to 150
OD.sub.550.
[0156] 21. G. L. Ellman, Arch. Biochem. Biophys. 82: 70 (1959).
[0157] 23. M. Brennan et al., Science 229: 81 (1985).
[0158] 24. E. Lamoyi, et al., Methods Enzymol. 121: 652 (1986).
[0159] 25. T. E. Creighton, Protein Structure, a Practical Approach
(IRL Press, Oxford, UK, 1990), p. 157.
[0160] 26. P. Matsudaira, J. Biol. Chem. 262: 10035 (1987)
[0161] 27. S. Moore and W. H. Stein, Methods Enzymol. 6: 819
(1963)
[0162] 32. M. J. Glennie et al., J. Immunol. 139: 2367 (1987); M.
J. Glennie et al., J. Immunol. 141: 3662 (1988).
[0163] 33. T. Nitta et al., J. Immunol. 19: 1437 (1989). T. Nitta
et al., Lancet 335: 368 (1990).
[0164] 34. P. Carter and J. A. Wells, Science 237: 394 (1987).
* * * * *