U.S. patent application number 14/469863 was filed with the patent office on 2015-03-05 for therapeutic uses of humanized antibodies against alpha-4 integrin.
This patent application is currently assigned to BIOGEN IDEC MA INC.. The applicant listed for this patent is BIOGEN IDEC MA INC.. Invention is credited to Mary M. BENDIG, Tarran S. JONES, Olivier J. LEGER, Jose SALDANHA, Ted A. Yednock.
Application Number | 20150064177 14/469863 |
Document ID | / |
Family ID | 46205415 |
Filed Date | 2015-03-05 |
United States Patent
Application |
20150064177 |
Kind Code |
A1 |
BENDIG; Mary M. ; et
al. |
March 5, 2015 |
THERAPEUTIC USES OF HUMANIZED ANTIBODIES AGAINST ALPHA-4
INTEGRIN
Abstract
The invention provides methods of treatment using humanized
immunoglobulins that specifically bind to alpha-4 integrin. The
methods are useful for treatment of asthma, atherosclerosis, AIDS
dementia, diabetes, inflammatory bowel disease, rheumatoid
arthritis, transplant rejection, graft versus host disease, tumor
metastasis, nephritis, atopic dermatitis, psoriasis, myocardial
ischemia, and acute leukocyte mediated lung injury.
Inventors: |
BENDIG; Mary M.; (Villanova,
PA) ; LEGER; Olivier J.; (Hertfordshire, GB) ;
SALDANHA; Jose; (Enfield, GB) ; JONES; Tarran S.;
(Radlett, GB) ; Yednock; Ted A.; (Fairfax,
CA) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
BIOGEN IDEC MA INC. |
Cambridge |
MA |
US |
|
|
Assignee: |
BIOGEN IDEC MA INC.
Cambridge
MA
|
Family ID: |
46205415 |
Appl. No.: |
14/469863 |
Filed: |
August 27, 2014 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
13540462 |
Jul 2, 2012 |
|
|
|
14469863 |
|
|
|
|
12175116 |
Jul 17, 2008 |
8246958 |
|
|
13540462 |
|
|
|
|
11006808 |
Dec 8, 2004 |
7435802 |
|
|
12175116 |
|
|
|
|
09155739 |
Sep 11, 1998 |
|
|
|
PCT/US96/18807 |
Nov 21, 1996 |
|
|
|
11006808 |
|
|
|
|
08561521 |
Nov 21, 1995 |
5840299 |
|
|
09155739 |
|
|
|
|
PCT/US95/01219 |
Jan 25, 1995 |
|
|
|
08561521 |
|
|
|
|
08186269 |
Jan 25, 1994 |
|
|
|
PCT/US95/01219 |
|
|
|
|
Current U.S.
Class: |
424/133.1 ;
435/7.24; 530/387.3; 536/23.53 |
Current CPC
Class: |
C07K 16/2839 20130101;
C07K 2317/24 20130101; C07K 2319/00 20130101; A61K 38/00 20130101;
C07K 2317/92 20130101; C07K 2317/55 20130101; C07K 2317/567
20130101; A61P 29/00 20180101; G01N 33/6872 20130101; C07K 2317/565
20130101; A61P 7/00 20180101; C07K 16/2842 20130101; G01N 2333/7055
20130101; A61K 2039/505 20130101; A61P 37/00 20180101 |
Class at
Publication: |
424/133.1 ;
530/387.3; 536/23.53; 435/7.24 |
International
Class: |
C07K 16/28 20060101
C07K016/28; G01N 33/68 20060101 G01N033/68 |
Claims
1-26. (canceled)
27. A humanized immunoglobulin comprising a humanized heavy chain
and a humanized light chain: (1) the humanized light chain
comprising three complementarity determining regions (CDR1, CDR2
and CDR3) having amino acid sequences from the corresponding
complementarity determining regions of the mouse 21-6
immunoglobulin light chain variable domain designated SEQ. ID. NO:
2, and a variable region framework from a human kappa light chain
variable region framework sequence except in at least one position
selected from a first group consisting of L45, L49, L58 and L69,
wherein the amino acid position is occupied by the same amino acid
present in the equivalent position of the mouse 21-6 immunoglobulin
light chain variable region framework; and (2) the humanized heavy
chain comprising three complementarity determining regions (CDR1,
CDR2 and CDR3) having amino acid sequences from the corresponding
complementarity determining regions of the mouse 21-6
immunoglobulin heavy chain variable domain designated SEQ. ID. NO:
4, and a variable region framework from a human heavy chain
variable region framework sequence except in at least one position
selected from a second group consisting of H27, H28, H29, H30, H44,
H71, wherein the amino acid position is occupied by the same amino
acid present in the equivalent position of the mouse 21-6
immunoglobulin heavy chain variable region framework; wherein the
humanized immunoglobulin specifically binds to VLA-4 with a binding
affinity having a lower limit of about 10.sup.7 M.sup.4 and an
upper limit of about five-times the binding affinity of the mouse
21-6 immunoglobulin.
28. The humanized immunoglobulin of claim 27, wherein the humanized
light chain variable region framework is from an RE1 variable
region framework sequence except in at least one position selected
from the first group, and except in at least one position selected
from a third group consisting of positions L104, L105 and L107,
wherein the amino acid position is occupied by the same amino acid
present in the equivalent position of a kappa light chain from a
human immunoglobulin other than RE1.
29. The humanized immunoglobulin of claim 28, wherein the humanized
heavy chain variable region framework is from a 21/28'CL variable
region framework sequence.
30. The humanized immunoglobulin of claim 29, wherein the humanized
light chain variable region framework comprises at least three
amino acids from the mouse 21.6 immunoglobulin at positions in the
first group and three amino acids from the kappa light chain from
the human immunoglobulin other than RE1 at positions in the third
group, and the humanized heavy chain variable region framework
comprises at least five amino acids from the mouse 21.6
immunoglobulin at positions in the second group.
31. The humanized immunoglobulin of claim 30, wherein the humanized
light chain variable region framework is identical to the RE1 light
chain variable region framework sequence except for the at least
three positions from the first group and the three positions from
the third group, and the heavy chain variable region framework is
identical to the 21/28'CL heavy chain variable region framework
sequence except for the at least five positions from the second
group.
32. The humanized immunoglobulin of claim 31, wherein the at least
three positions from the first group are positions L45, L58 and
L69, and at the least five positions from the second group are
positions H27, H28, H29, H30 and H71.
33. The humanized immunoglobulin of claim 32, wherein the humanized
light chain comprises complementarity determining regions that are
identical to the corresponding complementarity determining regions
of the mouse 21-6 heavy chain, and the humanized heavy chain
comprises complementarity determining regions that are identical to
the corresponding complementarity determining regions of the mouse
21-6 heavy chain, except that the CDR3 region of the humanized
heavy chain may or may not comprise a phenylalanine residue at
position H98.
34. The humanized immunoglobulin of claim 33, wherein the CDR3 of
the humanized heavy chain comprises a phenylalanine residue at
position H98.
35. The humanized immunoglobulin of claim 27, wherein the amino
acid sequence of the mature light chain variable region is selected
from the sequence designated La (SEQ. ID. NO:7) in FIG. 6 or the
sequence designated Lb (SEQ. ID. NO:8) in FIG. 6.
36. The humanized immunoglobulin of claim 27, wherein the amino
acid sequence of the mature heavy chain variable region is selected
from the sequence designated Ha (SEQ. ID. NO:11) in FIG. 7, the
sequence designated Hb (SEQ. ID. NO: 12) in FIG. 7, or the sequence
designated Hc (SEQ. ID. NO:13) in FIG. 7.
37. The humanized immunoglobulin of claim 35, wherein the amino
acid sequence of the mature light chain variable region is the
sequence designated La (SEQ. ID. NO: 7) and the mature heavy chain
variable region is selected from Ha (SEQ. ID. NO:11) or Hc (SEQ.
ID. NO: 13).
38. An antigen-specific binding fragment of the humanized
immunoglobulin of claim 37.
39. A humanized immunoglobulin of claim 37, further comprising a
constant region domain.
40. A humanized immunoglobulin of claim 39, wherein the constant
region domain has an effector function.
41. The humanized immunoglobulin of claim 37, wherein the humanized
immunoglobulin is a Fab fragment.
42. A nucleic acid encoding a heavy chain of a humanized antibody
of claim 1 or an antigen-specific binding fragment thereof, or a
light chain of a humanized antibody of claim 27, or an
antigen-specific binding fragment thereof.
43. A pharmaceutical composition comprising a humanized antibody of
claim 37, or a binding fragment thereof, and a pharmaceutically
acceptable carrier therefor.
44. A method for detecting VLA-4 antigen, the method comprising:
administering a humanized immunoglobulin of claim 37, or an
antigen-specific binding fragment thereof, to a tissue sample from
a patient; and detecting complexes formed by specific binding
between the antibody or fragment and VLA-4 present in the target
sample.
45. A method of inhibiting adhesion of a leukocyte to an
endothelial cell, the method comprising administering a
therapeutically effective amount of the pharmaceutical composition
of claim 43.
46. A method of treating a disease in a patient comprising
administering to the patient a therapeutically effective amount of
the pharmaceutical composition of claim 43, wherein the disease is
selected from multiple sclerosis, inflammatory bowel disease,
ulcerative colitis, Crohn's disease, rheumatoid arthritis, asthma,
diabetes, atherosclerosis, AIDS dementia, transplant rejection,
graft versus host disease, tumor metastasis, nephritis, atopic
dermatitis, psoriasis, myocardial ischemia, and acute leukocyte
mediated lung injury.
Description
CROSS REFERENCE TO RELATED APPLICATIONS
[0001] This application is a continuation of U.S. application Ser.
No. 13/540,462, filed Jul. 2, 2012, which is a continuation of U.S.
application Ser. No. 12/175,116, filed Jul. 17, 2008, now U.S. Pat.
No. 8,246,958, which is a continuation of U.S. application Ser. No.
11/006,808, filed Dec. 8, 2004, now U.S. Pat. No. 7,435,802, which
is a continuation of U.S. application Ser. No. 09/155,739, filed
Sep. 11, 1998, now abandoned, which is a national phase entry under
35 U.S.C. .sctn.371 of International App. No. PCT/US96/18807, filed
Nov. 21, 1996, which is a continuation-in-part of U.S. application
Ser. No. 08/561,521, filed Nov. 21, 1995, now U.S. Pat. No.
5,840,299, which is a continuation-in-part of International App.
No. PCT/US95/01219, filed Jan. 25, 1995, which is a
continuation-in-part of U.S. application Ser. No. 08/186,269, filed
Jan. 25, 1994, now abandoned, all of which are incorporated by
reference in their entirety for all purposes.
TECHNICAL FIELD
[0002] This invention relates generally to humanized antibodies
specific for the alpha-4 (.alpha.4) integrin and therapeutic uses
of the same.
BACKGROUND OF THE INVENTION
[0003] Inflammation is a response of vascularized tissues to
infection or injury and is effected by adhesion of leukocytes to
the endothelial cells of blood vessels and their infiltration into
the surrounding tissues. In normal inflammation, the infiltrating
leukocytes release toxic mediators to kill invading organisms,
phagocytize debris and dead cells, and play a role in tissue repair
and the immune response. However, in pathologic inflammation,
infiltrating leukocytes are over-responsive and can cause serious
or fatal damage. See, e.g., Hickey, Psychoneuroimmunology II
(Academic Press 1990).
[0004] The attachment of leukocytes to endothelial cells is
effected via specific interaction of cell-surface ligands and
receptors on endothelial cells and leukocytes. See generally
Springer, Nature 346:425-433 (1990). The identity of the ligands
and receptors varies for different cell subtypes, anatomical
locations and inflammatory stimuli. The VLA-4 leukocyte
cell-surface receptor was first identified by Hemler, EP 330,506
(1989) (incorporated by reference in its entirety for all
purposes). VLA-4 is a member of the .beta.1 integrin family of cell
surface receptors, each of which comprises .alpha. and .beta.
chains. VLA-4 contains an .alpha.4 chain and a .beta.1 chain. VLA-4
specifically binds to an endothelial cell ligand termed VCAM-1. See
Elices et al., Cell 60:577-584 (1990) (incorporated by reference in
its entirety for all purposes). The .alpha.4 chain also associates
with a .beta.7 chain to form an integrin referred to as
.alpha.4.beta.7. Although VCAM-1 was first detected on activated
human umbilical vein cells, this ligand has also been detected on
brain endothelial cells. See commonly owned, co-pending application
U.S. Ser. No. 07/871,223 (incorporated by reference in its entirety
for all purposes).
[0005] Adhesion molecules such as .alpha.4 integrin are potential
targets for therapeutic agents. The VLA-4 receptor of which
.alpha.4 integrin is a subunit is a particularly important target
because of its interaction with a ligand residing on brain
endothelial cells. Diseases and conditions resulting from brain
inflammation have particularly severe consequences. For example,
one such disease, multiple sclerosis (MS), has a chronic course
(with or without exacerbations and remissions) leading to severe
disability and death. The disease affects an estimated 250,000 to
350,000 people in the United States alone.
[0006] Antibodies against .alpha.4 integrin have been tested for
their anti-inflammatory potential both in vitro and in vivo in
animal models. See U.S. Ser. No. 07/871,223 and Yednock et al.,
Nature 356: 63-66 (1992) (incorporated by reference in its entirety
for all purposes). The in vitro experiments demonstrate that
.alpha.4 integrin antibodies block attachment of lymphocytes to
brain endothelial cells. The animal experiments test the effect of
.alpha.4 integrin antibodies on animals having an artificially
induced condition (experimental autoimmune encephalomyelitis),
simulating multiple sclerosis. The experiments show that
administration of anti-.alpha.4 integrin antibodies prevents
inflammation of the brain and subsequent paralysis in the animals.
Collectively, these experiments identify anti-.alpha.4 integrin
antibodies as potentially useful therapeutic agents for treating
multiple sclerosis and other inflammatory diseases and
disorders.
[0007] A significant problem with the anti-.alpha.4 integrin
antibodies available to-date is that they are all of murine origin,
and therefore likely to raise a human anti-mouse response (HAMA) in
clinical use. A HAMA response reduces the efficacy of mouse
antibodies in patients and prevents continued administration. One
approach to this problem is to humanize mouse antibodies. In this
approach, complementarity determining regions (CDRs) and certain
other amino acids from donor mouse variable regions are grafted
into human variable acceptor regions and then joined to human
constant regions. See, e.g., Riechmann et al., Nature 332:323-327
(1988); Winter, U.S. Pat. No. 5,225,539 (1993) (each of which is
incorporated by reference in its entirety for all purposes).
[0008] Although several examples of humanized antibodies have been
produced, the transition from a murine to a humanized antibody
involves a compromise of competing considerations, the solution of
which varies with different antibodies. To minimize immunogenicity,
the immunoglobulin should retain as much of the human acceptor
sequence as possible. However, to retain authentic binding
properties, the immunoglobulin framework should contain sufficient
substitutions of the human acceptor sequence to ensure a
three-dimensional conformation of CDR regions as close as possible
to that in the original mouse donor immunoglobulin. As a result of
these competing considerations, many humanized antibodies produced
to-date show some loss of binding affinity compared with the
corresponding murine antibodies from which they are derived. See,
e.g., Jones et al., Nature 321:522-525 (1986); Shearman et al., J.
Immunol. 147:4366-4373 (1991); Kettleborough et al., Protein
Engineering 4:773-783 (1991); Gorman et al., Proc. Natl. Acad. Sci.
USA 88:4181-4185 (1991); Tempest et al., Biotechnology 9:266-271
(1991).
[0009] Based on the foregoing it is apparent that a need exists for
humanized anti-.alpha.4 integrin antibodies demonstrating a strong
affinity for .alpha.4 integrin, while exhibiting little, if any,
human-antimouse response. The present invention fulfill this and
other needs.
SUMMARY OF THE INVENTION
[0010] The invention provides uses of a humanized antibody to
alpha-4 integrin in the manufacture of a medicament for treating a
disease selected from the group consisting of asthma,
atherosclerosis, AIDS dementia, diabetes, inflammatory bowel
disease, rheumatoid arthritis, transplant rejection, graft versus
host disease, tumor metastasis, nephritis, atopic dermatitis,
psoriasis, myocardial ischemia, and acute leukocyte mediated lung
injury.
[0011] The humanized immunoglobulins used in the above methods
specifically bind to a alpha-4 integrin. The humanized antibodies
comprise a humanized light chain and a humanized heavy chain. A
preferred humanized light chain comprises three complementarity
determining regions (CDR1, CDR2 and CDR3) having amino acid
sequences from the corresponding complementarity determining
regions of a mouse 21-6 immunoglobulin light chain, and a variable
region framework from a human kappa light chain variable region
framework sequence except in at least one position selected from a
first group consisting of positions L45, L49, L58 and L69, wherein
the amino acid position is occupied by the same amino acid present
in the equivalent position of the mouse 21.6 immunoglobulin light
chain variable region framework. A preferred humanized heavy chain
comprises three complementarity determining regions (CDR1, CDR2 and
CDR3) having amino acid sequences from the corresponding
complementarity determining regions of a mouse 21-6 immunoglobulin
heavy chain, and a variable region framework from a human heavy
chain variable region framework sequence except in at least one
position selected from a group consisting of H27, H28, H29, H30,
H44, H71, wherein the amino acid position is occupied by the same
amino acid present in the equivalent position of the mouse 21-6
immunoglobulin heavy chain variable region framework. The
immunoglobulins specifically bind to alpha-4 integrin with an
affinity having a lower limit of about 10.sup.7M.sup.-1 and an
upper limit of about five times the affinity of the mouse 21-6
immunoglobulin.
[0012] Usually, the humanized light and heavy chain variable region
frameworks are from RE1 and 21/28'CL variable region framework
sequences respectively. When the humanized light chain variable
region framework is from RE1, at least two framework amino acids
are replaced. One amino acid is from the first group of positions
described supra. The other amino acids is from a third group
consisting of positions L104, L105 and L107. This position is
occupied by the same amino acid present in the equivalent position
of a kappa light chain from a human immunoglobulin other than
RE1.
[0013] Some humanized immunoglobulins have a mature light chain
variable region sequence designated La or Lb in FIG. 6, or a mature
heavy chain variable region sequence designated Ha, Hb or Hc in
FIG. 7. Preferred humanized immunoglobulins include those having an
La light chain and an Ha, Hb or Hc heavy chain.
[0014] In another aspect the invention provides pharmaceutical
compositions for use in treating the above diseases. The
pharmaceutical compositions comprise a humanized immunoglobulin or
binding fragment as described supra, and a pharmaceutically
acceptable carrier. In some methods of treatment a therapeutically
effective amount of a pharmaceutical composition is administered to
a patient suffering from one of the diseases listed above.
BRIEF DESCRIPTION OF FIGURES
[0015] FIGS. 1A and 1B: DNA (SEQ. ID NO:1) and amino acid (SEQ. ID
NO:2) sequences of the mouse 21.6 light chain variable region.
[0016] FIGS. 2A and 2B: DNA (SEQ. ID NO:3) and amino acid (SEQ. ID
NO:4) sequences of the mouse 21.6 heavy chain variable region.
[0017] FIGS. 3A and 3B: Light (A) and heavy (B) chain expression
vectors used to produce chimeric and reshaped human antibodies with
human kappa light chains and human gamma-1 heavy chains in
mammalian cells.
[0018] FIG. 4: ELISA comparison of chimeric and mouse 21.6 antibody
binding to L cells expressing human .alpha.4.beta.1 integrin on
their surface.
[0019] FIG. 5: Molecular model of the variable regions of mouse
21.6 antibody. Residues of special interest are labelled.
[0020] FIG. 6: Comparisons of the amino acid sequences of mouse and
reshaped human 21.6 (SEQ. ID NO:5) light chain variable regions.
The amino acid residues that are part of the Chothia canonical
sequences for the CDR loop structures are marked with an asterisk.
REI (SEQ. ID NO:6) shows the FRs and CDRs from the V.sub.L region
of human REI light chain. La (SEQ. ID NO:7) and Lb (SEQ. ID NO:8)
are the two versions of reshaped human 21.6 V.sub.L region. The
residues in the FRs of La that differ from those in the REI
sequence are underlined. In Lb, only the residues in the framework
regions that differ from those of REI are shown.
[0021] FIG. 7: Comparisons of the amino acid sequences of the mouse
and reshaped human 21.6 (SEQ. ID NO:9) heavy chain variable
regions. The amino acid residues that are part of the canonical
sequences for the Chothia CDR loop structures are marked with an
asterisk. 2*CL (SEQ. ID NO:10) shows the FRs and CDRs from the
V.sub.H region of human 21/28'CL antibody. Ha (SEQ. ID NO:11), Hb
(SEQ. ID NO:12), and Hc (SEQ. ID NO:13) are the three versions of
reshaped human 21.6 V.sub.H region. The residues in the FRs of Ha
that differ from those in the 21/28'CL sequence are underlined. In
Hb and Hc, only the residues in the framework regions that differ
from those of 21/28'CL are shown.
[0022] FIG. 8: PCR-based construction of version "a" of reshaped
human 21.6 light chain variable region. The dotted lines indicate a
complementary sequence of at least 21 bases between the
primers.
[0023] FIG. 9: PCR-based construction of version "a" of reshaped
human 21.6 heavy chain variable region.
[0024] FIGS. 10A and 10B: cDNA and amino acid sequences (SEQ. ID
NOS: 14 and 15) of the first version ("a") of reshaped human 21.6
light 35 chain variable region.
[0025] FIGS. 11A and 11B: DNA and amino acid sequences (SEQ. ID
NOS: 16 and 17) of the first version ("a") of reshaped human 21.6
heavy chain variable region.
[0026] FIGS. 12A and 12B: ELISA comparison of chimeric and reshaped
human 21.6 antibodies to bind to L cells expressing human
.alpha.4.beta.1 integrin on their surface. (A) represents the data
from the reshaped human 21.6 antibody (La+Ha). (B) represents the
data from the reshaped human 21.6 antibody (La+Hc).
[0027] FIGS. 13A and 13B: Comparison of mouse 21.6 antibody with a
different anti-alpha-4 integrin antibody, L25. Panel A compares the
ability of the antibodies to block binding of U937 monocytic cells
to purified VCA-1 in the presence and absence of Mn.sup.2+. Panel B
compares the ability of the antibodies to block binding of Jurkat
cells to increasing concentrations of VCAM-1.
[0028] FIG. 14: Delay of weight loss in animals treated with mouse
or human 21.6 antibody.
[0029] FIG. 15: Reversal of clinical symptoms in animals treated
with mouse or human 21.6 antibody.
[0030] FIG. 16: Reversal of weight loss in animals treated with
mouse or human 21.6 antibody.
DEFINITIONS
[0031] Abbreviations for the twenty naturally occurring amino acids
follow conventional usage (Immunology--A Synthesis (2nd ed., E. S.
Golub & D. R. Gren, eds., Sinauer Associates, Sunderland,
Mass., 1991)). Stereoisomers (e.g., D-amino acids) of the twenty
conventional amino acids, unnatural amino acids such as
a,a-disubstituted amino acids, N-alkyl amino acids, lactic acid,
and other unconventional amino acids may also be suitable
components for polypeptides of the present invention. Examples of
unconventional amino acids include: 4-hydroxyproline,
.gamma.carboxyglutamate, .epsilon.-N,N,N-trimethyllysine,
.epsilon.-N-acetyllysine, 0-phosphoserine, N-acetylserine,
N-formylmethionine, 3-methylhistidine, 5-hydroxylysine,
co-N-methylarginine, and other similar amino acids and imino acids
(e.g., 4-hydroxyproline). Moreover, amino acids may be modified by
glycosylation, phosphorylation and the like.
[0032] In the polypeptide notation used herein, the lefthand
direction is the amino terminal direction and the righthand
direction is the carboxy-terminal direction, in accordance with
standard usage and convention. Similarly, unless specified
otherwise, the lefthand end of single-stranded polynucleotide
sequences is the 5' end; the lefthand direction of double-stranded
polynucleotide sequences is referred to as the 5' direction. The
direction of 5' to 3' addition of nascent RNA transcripts is
referred to as the transcription direction; sequence regions on the
DNA strand having the same sequence as the RNA and which are 5' to
the 5' end of the RNA transcript are referred to as "upstream
sequences"; sequence regions on the DNA strand having the same
sequence as the RNA and which are 3' to the 3' end of the RNA
transcript are referred to as "downstream sequences."
[0033] The phrase "polynucleotide sequence" refers to a single or
double-stranded polymer of deoxyribonucleotide or ribonucleotide
bases read from the 5' to the 3' end. It includes self-replicating
plasmids, infectious polymers of DNA or RNA and non-functional DNA
or RNA.
[0034] The following terms are used to describe the sequence
relationships between two or more polynucleotides: "reference
sequence", "comparison window", "sequence identity", "percentage of
sequence identity", and "substantial identity". A "reference
sequence" is a defined sequence used as a basis for a sequence
comparison; a reference sequence may be a subset of a larger
sequence, for example, as a segment of a full-length CDNA or gene
sequence given in a sequence listing, such as a polynucleotide
sequence of FIG. 1 or 2, or may comprise a complete DNA or gene
sequence. Generally, a reference sequence is at least 20
nucleotides in length, frequently at least 25 nucleotides in
length, and often at least 50 nucleotides in length. Since two
polynucleotides may each (1.sup.-) comprise a sequence (i.e., a
portion of the complete polynucleotide 30 sequence) that is similar
between the two polynucleotides, and (2) may further comprise a
sequence that is divergent between the two polynucleotides,
sequence comparisons between two (or more) polynucleotides are
typically performed by comparing sequences of the two
polynucleotides over a "comparison window" to identify and compare
local regions of sequence similarity. A "comparison window", as
used herein, refers to a conceptual segment of at least 20
contiguous nucleotide positions wherein a polynucleotide sequence
may be compared to a reference sequence of at least 20 contiguous
nucleotides and wherein the portion of the polynucleotide sequence
in the comparison window may comprise additions or deletions (i.e.,
gaps) of 20 percent or less as compared to the reference sequence
(which does not comprise additions or deletions) for optimal
alignment of the two sequences. Optimal alignment of sequences for
aligning a comparison window may be conducted by the local homology
algorithm of Smith & Waterman, Adv. Appl. Math. 2:482 (1981),
by the homology alignment algorithm of Needleman & Wunsch, J.
Mol. Biol. 48:443 (1970), by the search for similarity method
of Pearson & Lipman, Proc. Natl. Acad. Sci. (USA) 85:2444
(1988) (each of which is incorporated by reference in its entirety
for all purposes), by computerized implementations of these
algorithms (GAP, BESTFIT, FASTA, and TFASTA in the Wisconsin
Genetics Software Package Release 7.0, Genetics Computer Group, 575
Science Dr., Madison, Wis.), or by inspection, and the best
alignment (i.e., resulting in the highest percentage of sequence
similarity over the comparison window) generated by the various
methods is selected. The term "sequence identity" means that two
polynucleotide sequences are identical (i.e., on a
nucleotide-by-nucleotide basis) over the window of comparison. The
term "percentage of sequence identity" is calculated by comparing
two optimally aligned sequences over the window of comparison,
determining the number of positions at which the identical nucleic
acid base (e.g., A, T, C, G, U, or I) occurs in both sequences to
yield the number of matched positions, dividing the number of
matched positions by the total number of positions in the window of
comparison (i.e., the window size), and multiplying the result by
100 to 30 yield the percentage of sequence identity. The terms
"substantial identity" as used herein denotes a characteristic of a
polynucleotide sequence, wherein the polynucleotide comprises a
sequence that has at least 85 percent sequence identity, preferably
at least 90 to 95 percent sequence identity, more usually at least
99 percent sequence identity as compared to a reference sequence
over a comparison window of at least 20 nucleotide positions,
frequently over a window of at least 25-50 nucleotides, wherein the
percentage of sequence identity is calculated by comparing the
reference sequence to the polynucleotide sequence which may include
deletions or additions which total 20 percent or less of the
reference sequence over the window of comparison. The reference
sequence may be a subset of a larger sequence, for example, the
sequence shown in FIG. 1 or 2.
[0035] As applied to polypeptides, the term "sequence identity"
means peptides share identical amino acids at corresponding
positions. The term "sequence similarity" means peptides have
identical or similar amino acids (i.e., conservative substitutions)
at corresponding positions. The term "substantial identity" means
that two peptide sequences, when optimally aligned, such as by the
programs GAP or BESTFIT using default gap weights, share at least
80 percent sequence identity, preferably at least 90 percent
sequence identity, more preferably at least 95 percent sequence
identity or more (e.g., 99 percent sequence identity). Preferably,
residue positions which are not identical differ by conservative
amino acid substitutions. The term "substantial similarity" means
that two peptide sequences share corresponding percentages of
sequence similarity.
[0036] The term "substantially pure" means an object species is the
predominant species present (i.e., on a molar basis it is more
abundant than any other individual species in the composition), and
preferably a substantially purified fraction is a composition
wherein the object species comprises at least about 50 percent (on
a molar basis) of all macromolecular species present. Generally, a
substantially pure composition will comprise more than about 80 to
90 percent of all macromolecular species present in the
composition. Most preferably, the object species is purified to
essential homogeneity (contaminant species cannot be detected in
the composition by conventional detection methods) wherein the
composition consists essentially of a single macromolecular
species.
[0037] For purposes of classifying amino acids substitutions as
conservative or nonconservative, amino acids are grouped as
follows: Group I (hydrophobic sidechains): norleucine, met, ala,
val, leu, ile; Group II (neutral hydrophilic side chains): cys,
ser, thr; Group III (acidic side chains): asp, glu; Group IV (basic
side chains): asn, gln, his, lys, arg; Group V (residues
influencing chain orientation): gly, pro; and Group VI (aromatic
side chains): trp, tyr, phe. Conservative substitutions involve
substitutions between amino acids in the same class.
Non-conservative substitutions constitute exchanging a member of
one of these classes for another. Amino acids from the variable
regions of the mature heavy and light chains of immunoglobulins are
designated Hx and Lxx respectively, where x is a number designating
the position of an amino acids according to the scheme of Kabat et
al., Sequences of Proteins of Immunological Interest (National
Institutes of Health, Bethesda, Md. (1987) and (1991)) (hereinafter
collectively referred to as "Kabat et al.," incorporated by
reference in their entirety for all purposes). Kabat et al. list
many amino acid sequences for antibodies for each subclass, and
list the most commonly occurring amino acid for each residue
position in that subclass. Kabat et al. use a method for assigning
a residue number to each amino acid in a listed sequence, and this
method for assigning residue numbers has become standard in the
field. Kabat et al.'s scheme is extendible to other antibodies not
included in the compendium by aligning the antibody in question
with one of the consensus sequences in Kabat et al. The use of the
Kabat et al. numbering system readily identifies amino acids at
equivalent positions in different antibodies. For example, an amino
acid at the L50 position of a human antibody occupies the
equivalence position to an amino acid position L50 of a mouse
antibody.
DETAILED DESCRIPTION
I. Humanized Antibodies Specific for Alpha-4 Integrin
[0038] In one embodiment of the invention, humanized
immunoglobulins (or antibodies) specific for the alpha-4 integrin,
a subunit of VLA-4 are provided. The humanized immunoglobulins have
variable framework regions substantially from a human
immunoglobulin (termed an acceptor immunoglobulin) and
complementarity determining regions substantially from a mouse
immunoglobulin termed mu MAb 21.6 (referred to as the donor
immunoglobulin). The constant region(s), if present, are also
substantially from a human immunoglobulin. The humanized antibodies
exhibit a specific binding affinity for alpha-4 integrin of at
least 10.sup.7, 10.sup.8, 10.sup.9, or 10.sup.10 M.sup.-1. Usually
the upper limit of binding affinity of the humanized antibodies for
alpha-4 integrin is within a factor of three or five of that of mu
MAb 21.6 (about 10.sup.9 M.sup.-1). Often the lower limit of
binding affinity is also within a factor of three or five of that
of mu MAb 21.6.
[0039] A. General Characteristics of Immunocilobulins
[0040] The basic antibody structural unit is known to comprise a
tetramer. Each tetramer is composed of two identical pairs of
polypeptide chains, each pair having one "light" (about 25 kDa) and
one "heavy" chain (about 50-70 kDa). The amino-terminal portion of
each chain includes a variable region of about 100 to 110 or more
amino acids primarily responsible for antigen recognition. The
carboxy-terminal portion of each chain defines a constant region
primarily responsible for effector function.
[0041] Light chains are classified as either kappa or lambda. Heavy
chains are classified as gamma, mu, alpha, delta, or epsilon, and
define the antibody's isotype as IgG, IgM, IgA, IgD and IgE,
respectively. Within light and heavy chains, the variable and
constant regions are joined by a "J" region of about 12 or more
amino acids, with the heavy chain also including a "D" region of
about 10 more amino acids. (See generally, Fundamental Immunology
(Paul, W., ed., 2nd ed. Raven Press, N.Y., 1989), Ch. 7
(incorporated by reference in its entirety for all purposes).
[0042] The variable regions of each light/heavy chain pair form the
antibody binding site. The chains all exhibit the same general
structure of relatively conserved framework regions (FR) joined by
three hypervariable regions, also called complementarity
determining regions or CDRs. The CDRs from the two chains of each
pair are aligned by the framework regions, enabling binding to a
specific epitope. CDR and FR residues are delineated according to
the standard sequence definition of Kabat et al., supra. An
alternative structural definition has been proposed by Chothia et
al., J. Mol. Biol. 196:901-917 (1987); Nature 342:878-883 (1989);
and J. Mol. Biol. 186:651-663 (1989) (hereinafter collectively
referred to as "Chothia et al." and incorporated by reference in
their entirety for all purposes). When framework positions, as
defined by Kabat et al., supra, that constitute structural loop
positions as defined by Chothia et al., supra, the amino acids
present in the mouse antibody are usually incorporated into the
humanized antibody.
[0043] B. Production of Humanized Antibodies
[0044] (1) Mouse MAb 21.6
[0045] The starting material for production of humanized antibodies
is mu MAb 21.6. The isolation and properties of this antibody are
described in U.S. Ser. No. 07/871,223. Briefly, mu MAb 21.6 is
specific for the alpha-4 integrin and has been shown to inhibit
human lymphocyte binding to tissue cultures of rat brain cells
stimulated with tumor necrosis factor. The cloning and sequencing
of CDNA encoding the mu MAb 21.6 antibody heavy and light chain
variable regions is described in Example 1, and the nucleotide and
predicted amino acids sequences are shown in FIGS. 1 and 2. These
figures also illustrate the subdivision of the amino acid coding
sequencing into framework and complementarity determining domains.
From N-terminal to c-terminal, both light and heavy chains comprise
the domains FR1, CDR1, FR2, CDR2, FR3, CDR3 and FR4. The assignment
of amino acids to each domain is in accordance with the numbering
convention of Kabat et al., supra.
[0046] (2) Selection of Human Antibodies to Supply Framework
Residues
[0047] The substitution of mouse CDRs into a human variable domain
framework is most likely to result in retention of their correct
spatial orientation if the human variable domain framework adopts
the same or similar conformation to the mouse variable framework
from which the CDRs originated. This is achieved by obtaining the
human variable domains from human antibodies whose framework
sequences exhibit a high degree of sequence identity with the
murine variable framework domains from which the CDRs were derived.
The heavy and light chain variable framework regions can be derived
from the same or different human antibody sequences. The human
antibody sequences can be the sequences of naturally occurring
human antibodies or can be consensus sequences of several human
antibodies. See Kettleborough et al., Protein Engineering 4:773
(1991); Kolbinger et al., Protein Engineering 6:971 (1993).
[0048] Suitable human antibody sequences are identified by computer
comparisons of the amino acid sequences of the mouse variable
regions with the sequences of known human antibodies. The
comparison is performed separately for heavy and light chains but
the principles are similar for each. This comparison reveals that
the mu 21.6 light chain shows greatest sequence identity to human
light chains of subtype kappa 1, and that the mu 21.6 heavy chain
shows greatest sequence identity to human heavy chains of subtype
one, as defined by Kabat et al., supra. Thus, light and heavy human
framework regions are usually derived from human antibodies of
these subtypes, or from consensus sequences of such subtypes. The
preferred light and heavy chain human variable regions showing
greatest sequence identity to the corresponding regions from mu MAb
21.6 are from antibodies RE1 and 21/28'CL respectively.
[0049] (3) Computer Modelling
[0050] The unnatural juxtaposition of murine CDR regions with human
variable framework region can result in unnatural conformational
restraints, which, unless corrected by substitution of certain
amino acid residues, lead to loss of binding affinity. The
selection of amino acid residues for substitution is determined, in
part, by computer modelling. Computer hardware and software for
producing three-dimensional images of immunoglobulin molecules are
widely available. In general, molecular models are produced
starting from solved structures for immunoglobulin chains or
domains thereof. The chains to be modelled are compared for amino
acid sequence similarity with chains or domains of solved three
dimensional structures, and the chains or domains showing the
greatest sequence similarity is/are selected as starting points for
construction of the molecular model. For example, for the light
chain of mu MAb 21.6, the starting point for modelling the
framework regions, CDR1 and CDR2 regions, was the human light chain
RE1. For the CDR3 region, the starting point was the CDR3 region
from the light chain of a different human antibody HyHEL-5. The
solved starting structures are modified to allow for differences
between the actual amino acids in the immunoglobulin chains or
domains being modelled, and those in the starting structure. The
modified structures are then assembled into a composite
immunoglobulin. Finally, the model is refined by energy
minimization and by verifying that all atoms are within appropriate
distances from one another and that bond lengths and angles are
within chemically acceptable limits. Example 4 discusses in more
detail the steps taken to produce a three dimensional computer
model for the variable regions of the mu MAb 21.6, and the model is
shown in FIG. 5. This model can in turn serve as a starting point
for predicting the three-dimensional structure of an antibody
containing the mu MAb 21.6 complementarity determining regions
substituted in human framework structures. Additional models can be
constructed representing the structure when further amino acid
substitutions to be discussed infra, are introduced.
[0051] (4) Substitution of Amino Acid Residues
[0052] As noted supra, the humanized antibodies of the invention
comprise variable framework regions substantially from a human
immunoglobulin and complementarity determining regions
substantially from a mouse immunoglobulin termed mu MAb 21.6.
Having identified the complementarity determining regions of mu MAb
21.6 and appropriate human acceptor immunoglobulins, the next step
is to determine which, if any, residues from these components
should be substituted to optimize the properties of the resulting
humanized antibody. In general, substitution of human amino acid
residues with murine should be minimized, because introduction of
murine residues increases the risk of the antibody eliciting a HAMA
response in humans. Amino acids are selected for substitution based
on their possible influence on CDR conformation and/or binding to
antigen.
[0053] Investigation of such possible influences is by modelling,
examination of the characteristics of the amino acids at particular
locations, or empirical observation of the effects of substitution
or mutagenesis of particular amino acids.
[0054] When an amino acid differs between a mu MAb 21.6 variable
framework region and an equivalent human variable framework region,
the human framework amino acid should usually be substituted by the
equivalent mouse amino acid if it is reasonably expected that the
amino acid: [0055] (1) noncovalently binds antigen directly (e.g.,
amino acids at positions L49, L69 of mu MAb 21.6), [0056] (2) is
adjacent to a CDR region, is part of a CDR region under the
alternative definition proposed by Chothia et al., supra, or
otherwise interacts with a CDR region (e.g., is within about 3A of
a CDR region) (e.g., amino acids at positions L45, L58, H27, H28,
H29, H30 and H71 of mu MAb 21.6), or [0057] (3) participates in the
V.sub.L-V.sub.H interface (e.g., amino acids at position H44 of mu
MAb 21.6).
[0058] Other candidates for substitution are acceptor human
framework amino acids that are unusual for a human immunoglobulin
at that position (e.g., amino acids at positions L104, L105 and
L107 of mu MAb 21.6). These amino acids can be substituted with
amino acids from the equivalent position of more typical human
immunoglobulins. Alternatively, amino acids from equivalent
positions in the mouse MAb 21.6 can be introduced into the human
framework regions when such amino acids are typical of human
immunoglobulin at the equivalent positions.
[0059] In general, substitution of all or most of the amino acids
fulfilling the above criteria is desirable. Occasionally, however,
there is some ambiguity about whether a particular amino acid meets
the above criteria, and alternative variant immunoglobulins are
produced, one of which has that particular substitution, the other
of which does not. The humanized antibodies of the present
invention will usually contain a substitution of a human light
chain framework residue with a corresponding mu MAb 21.6 residue in
at least 1, 2 or 3, and more usually 4, of the following positions:
L45, L49, L58 and L69. The humanized antibodies also usually
contain a substitution of a human heavy chain framework residue in
at least 1, 2, 3, 4, or 5, and sometimes 6, of the following
positions: H27, H28, H29, H30, H44 and H71. Optionally, H36 may
also be substituted. In preferred embodiments when the human light
chain acceptor immunoglobulin is RE1, the light chain also contains
substitutions in at least 1 or 2, and more usually 3, of the
following positions: L104, L105 and L107. These positions are
substituted with the amino acid from the equivalent position of a
human immunoglobulin having a more typical amino acid residues.
Appropriate amino acids to substitute are shown in FIGS. 6 and
7.
[0060] Usually the CDR regions in humanized antibodies are
substantially identical, and more usually, identical to the
corresponding CDR regions in the mu MAb 21.6 antibody.
Occasionally, however, it is desirable to change one of the
residues in a CDR region. For example, Example 5 identifies an
amino acid similarity between the mu MAb 21.6 CDR3 and the VCAM-1
ligand. This observation suggests that the binding affinity of
humanized antibodies might be improved by redesigning the heavy
chain CDR3 region to resemble VCAM-1 even more closely.
Accordingly, one or more amino acids from the CDR3 domain can be
substituted with amino acids from the VCAM-1 binding domain.
Although not usually desirable, it is sometimes possible to make
one or more conservative amino acid substitutions of CDR residues
without appreciably affecting the binding affinity of the resulting
humanized immunoglobulin.
[0061] Other than for the specific amino acid substitutions
discussed above, the framework regions of humanized immunoglobulins
are usually substantially identical, and more usually, identical to
the framework regions of the human antibodies from which they were
derived. Of course, many of the amino acids in the framework region
make little or no direct contribution to the specificity or
affinity of an antibody. Thus, many individual conservative
substitutions of framework residues can be tolerated without
appreciable change of the specificity or affinity of the resulting
humanized immunoglobulin. However, in general, such substitutions
are undesirable.
[0062] (5) Production of Variable Regions
[0063] Having conceptually selected the CDR and framework
components of humanized immunoglobulins, a variety of methods are
available for producing such immunoglobulins. Because of the
degeneracy of the code, a variety of nucleic acid sequences will
encode each immunoglobulin amino acid sequence. The desired nucleic
acid sequences can be produced by de novo solid-phase DNA synthesis
or by PCR mutagenesis of an earlier prepared variant of the desired
polynucleotide. Oligonucleotide-mediated mutagenesis is a preferred
method for preparing substitution, deletion and insertion variants
of target polypeptide DNA. See Adelman et al., DNA 2:183 (1983).
Briefly, the target polypeptide DNA is altered by hybridizing an
oligonucleotide encoding the desired mutation to a single-stranded
DNA template. After hybridization, a DNA polymerase is used to
synthesize an entire second complementary strand of the template
that incorporates the oligonucleotide primer, and encodes the
selected alteration in the target polypeptide DNA.
[0064] (6) Selection of Constant Region
[0065] The variable segments of humanized antibodies produced as
described supra are typically linked to at least a portion of an
immunoglobulin constant region (Fc), typically that of a human
immunoglobulin. Human constant region DNA sequences can be isolated
in accordance with well-known procedures from a variety of human
cells, but preferably immortalized B-cells (see Kabat et al.,
supra, and WO87/02671) (each of which is incorporated by reference
in its entirety for all purposes). Ordinarily, the antibody will
contain both light chain and wavy chain constant regions. The heavy
chain constant region usually includes CH1, hinge, CH2, CH3, and
CH4 regions. The humanized antibodies include antibodies having all
types of constant regions, including IgM, IgG, IgD, IgA and IgE,
and any isotype, including IgG1, IgG2, IgG3 and IgG4. When it is
desired that the humanized antibody exhibit cytotoxic activity, the
constant domain is usually a complement-fixing constant domain and
the class is typically IgG.sub.1. When such cytotoxic activity is
not desirable, the constant domain may be of the IgG.sub.2 class.
The humanized antibody may comprise sequences from more than one
class or isotype.
[0066] (7) Expression Systems
[0067] Nucleic acids encoding humanized light and heavy chain
variable regions, optionally linked to constant regions, are
inserted into expression vectors. The light and heavy chains can be
cloned in the same or different expression vectors. The DNA
segments encoding immunoglobulin chains are operably linked to
control sequences in the expression vector(s) that ensure the
expression of immunoglobulin polypeptides. Such control sequences
include a signal sequence, a promoter, an enhancer, and a
transcription termination sequence. Expression vectors are
typically replicable in the host organisms either as episomes or as
an integral part of the host chromosomal DNA. Commonly, expression
vectors will contain selection markers, e.g., tetracycline or
neomycin, to permit detection of those cells transformed with the
desired DNA sequences (see, e.g., U.S. Pat. No. 4,704,362.)
[0068] E. coli is one prokaryotic host useful particularly for
cloning the polynucleotides of the present invention. Other
microbial hosts suitable for use include bacilli, such as Bacillus
subtilus, and other enterobacteriaceae, such as Salmonella,
Serratia, and various Pseudomonas species. In these prokaryotic
hosts, one can also make expression vectors, which will typically
contain expression control sequences compatible with the host cell
(e.g., an origin of replication). In addition, any number of a
variety of well-known promoters will be present, such as the
lactose promoter system, a tryptophan (trp) promoter system, a
beta-lactamase promoter system, or a promoter system from phage
lambda. The promoters will typically control expression, optionally
with an operator sequence, and have ribosome binding site sequences
and the like, for initiating and completing transcription and
translation.
[0069] Other microbes, such as yeast, may also be used for
expression. Saccharomyces is a preferred host, with suitable
vectors having expression control sequences, such as promoters,
including 3-phosphoglycerate kinase or other glycolytic enzymes,
and an origin of replication, termination sequences and the like as
desired.
[0070] In addition to microorganisms, mammalian tissue cell culture
may also be used to express and produce the polypeptides of the
present invention (see Winnacker, From Genes to Clones (VCH
Publishers, N.Y., N.Y., 1987). Eukaryotic cells are actually
preferred, because a number of suitable host cell lines capable of
secreting intact immunoglobulins have been developed in the art,
and include the CHO cell lines, various Cos cell lines, HeLa cells,
preferably myeloma cell lines, or transformed B-cells or
hybridomas. Expression vectors for these cells can include
expression control sequences, such as an origin of replication, a
promoter, and an enhancer (Queen et al., Immunol. Rev. 89:49-68
(1986)), and necessary processing information sites, such as
ribosome binding sites, RNA splice sites, polyadenylation sites,
and transcriptional terminator sequences. Preferred expression
control sequences are promoters derived from immunoglobulin genes,
SV40, adenovirus, bovine papilloma virus, cytomegalovirus and the
like.
[0071] The vectors containing the polynucleotide sequences of
interest (e.g., the heavy and light chain encoding sequences and
expression control sequences) can be transferred into the host cell
by well-known methods, which vary depending on the type of cellular
host. For example, calcium chloride transfection is commonly
utilized for prokaryotic cells, whereas calcium phosphate treatment
or electroporation may be used for other cellular hosts. (See
generally Sambrook et al., Molecular Cloning: A Laboratory Manual
(Cold Spring Harbor Press, 2nd ed., 1989) (incorporated by
reference in its entirety for all purposes). When heavy and light
chains are cloned on separate expression vectors, the vectors are
cotransfected to obtain expression and assembly of intact
immunoglobulins.
[0072] Once expressed, the whole antibodies, their dimers,
individual light and heavy chains, or other immunoglobulin forms of
the present invention can be purified according to standard
procedures of the art, including ammonium sulfate precipitation,
affinity columns, column chromatography, gel electrophoresis and
the like (see generally Scopes, Protein Purification
(Springer-Verlag, N.Y., 1982). Substantially pure immunoglobulins
of at least about 90 to 95% homogeneity are preferred, and 98 to
99% or more homogeneity most preferred, for pharmaceutical
uses.
[0073] C. Fragments of Humanized Antibodies
[0074] In another embodiment of the invention, fragments of
humanized antibodies are provided. Typically, these fragments
exhibit specific binding to alph-4 integrin with an affinity of at
least 10.sup.7 M.sup.-1, and more typically 10.sup.8 or 10.sup.9
M.sup.-1. Humanized antibody fragments include separate heavy
chains, light chains Fab, Fab', F(ab').sub.2, Fabc, and Fv.
Fragments are produced by recombinant DNA techniques, or by enzymic
or chemical separation of intact immunoglobulins.
II. Nucleic Acids
[0075] The humanized antibodies and fragments thereof are usually
produced by expression of nucleic acids. All nucleic acids encoding
a humanized antibody or a fragment thereof described in this
application are expressly included in the invention.
III. Computers
[0076] In another aspect of the invention, computers programmed to
display three dimensional images of antibodies on a monitor are
provided. For example, a Silicon Graphics IRIS 4D workstation
running under the UNIX operating system and using the molecular
modelling package QUANTA (Polygen Corp. USA) is suitable. Computers
are useful for visualizing models of variants of humanized
antibodies. In general, the antibodies of the invention already
provide satisfactory binding affinity. However, it is likely that
antibodies with even stronger binding affinity could be identified
by further variation of certain amino acid residues. The three
dimensional image will also identify many noncritical amino acids,
which could be the subject of conservative substitutions without
appreciable affecting the binding affinity of the antibody.
Collectively even conservative substitutions can have a significant
effect on the properties of an immunoglobulin. However, it is
likely many individual conservative substitutions will not
significantly impair the properties of the immunoglobulins.
IV. Testing Humanized Antibodies
[0077] The humanized antibodies of the invention are tested by a
variety of assays. These include a simple binding assay for
detecting the existence or strength of binding of an antibody to
cells bearing VLA-4 of which one subunit is alpha-4 integrin. The
antibodies are also tested for their capacity to block the
interaction of cells bearing the VLA-4 receptor with endothelial
cells expressing a VCAM-1 ligand. The endothelial cells may be
grown and stimulated in culture or may be a component of naturally
occurring brain tissue sections. See Yednock et al., supra, and
U.S. Ser. No. 07/871,223. The humanized antibodies are also tested
for their capacity to prevent or reduce inflammation and subsequent
paralysis in laboratory animals having experimental autoimmune
encephalomyelitis (EAE). EAE is induced by injection of a
laboratory animal with CD4.sup.+ T-cells specific for myelin basic
protein or by directly immunizing animals with myelin basic
protein. This protein is localized in the central nervous system,
and the reactive T-cells initiate destruction of sheaths containing
this protein in a manner that simulates the autoimmune response in
multiple sclerosis. See Yednock et al., supra, and copending U.S.
Ser. No. 07/871,223.
V. Pharmaceutical Compositions
[0078] The invention provides pharmaceutical compositions to be
used for prophylactic or therapeutic treatment comprising an active
therapeutic agent, i.e., a humanized 21.6 antibody or a binding
fragment thereof, and a variety of other components. The preferred
form depends on the intended mode of administration and therapeutic
application. The compositions can also include, depending on the
formulation desired, pharmaceutically-acceptable, non-toxic
carriers or diluents, which are defined as vehicles commonly used
to formulate pharmaceutical compositions for animal or human
administration. The diluent is selected so as not to affect the
biological activity of the combination. Examples of such diluents
are distilled water, physiological phosphate-buffered saline,
Ringer's solutions, dextrose solution, and Hank's solution.
addition, the pharmaceutical composition or formulation may also
include other carriers, adjuvants, or nontoxic, nontherapeutic,
nonimmunogenic stabilizers and the like.
[0079] For parenteral administration, the antibodies of the
invention can be administered as injectionable dosages of a
solution or suspension of the substance in a physiologically
acceptable diluent with a pharmaceutical carrier which can be a
sterile liquid such as water and oils with or without the addition
of a surfactant and other pharmaceutically preparations are those
of petroleum, animal, vegetable, or synthetic origin, for example,
peanut oil, soybean oil, and mineral oil. In general, glycols such
as propylene glycol or polyethylene glycol are preferred liquid
carriers, particularly for injectable solutions. The antibodies of
this invention can be administered in the form of a depot injection
or implant preparation which can be formulated in such a manner as
to permit a sustained release of the active ingredient. A preferred
composition comprises monoclonal antibody at 5 mg/mL, formulated in
aqueous buffer consisting of 50 mM L-histidine, 150 mM NaCl,
adjusted to pH 6.0 with HCl.
VI. Methods of Diagnosis
[0080] The humanized antibodies and their binding fragments are
useful for detecting the presence of cells bearing alpha-4
integrin. The presence of such cells in the brain is diagnostic of
an inflammatory response and may signal the need for commencement
of a therapeutic method discussed infra. Diagnosis can be
accomplished by removing a cellular sample from a patient. The
amount of expressed alpha-4 integrin in individual cells of the
sample is then determined, e.g., by immunohistochemical staining of
fixed cells or by Western blotting of a cell extract with a
humanized MAb 21.6 antibody or a binding fragment thereof.
[0081] Diagnosis can also be achieved by in vivo administration of
a labelled humanized MAb 21.6 (or binding fragment) and detection
by in vivo imaging. The concentration of humanized MAb 21.6
administered should be sufficient that the binding to cells having
the target antigen is detectable compared to the background signal.
The diagnostic reagent can be labelled with a radioisotope for
camera imaging, or a paramagnetic isotope for magnetic resonance or
electron spin resonance imaging.
[0082] A change (typically an increase) in the level of alpha-4
integrin in a cellular sample or imaged from an individual, which
is outside the range of clinically established normal levels, may
indicate the presence of an undesirable inflammatory response
reaction in the individual from whom the sample was obtained,
and/or indicate a predisposition of the individual for developing
(or progressing through) such a reaction. Alpha-4 integrin can also
be employed as a differentiation marker to identify and type cells
of certain lineages and developmental origins. Such cell-type
specific detection can be used for histopathological diagnosis of
undesired immune responses.
VII. Methods of Treatment
[0083] The invention also provides methods of treatment that
exploit the capacity of humanized MAb 21.6 to block a4-dependent
interactions. The a4-dependent interaction with the VCAM-1 ligand
on endothelial cells is an early event in many inflammatory
responses, including those of the central nervous system. Undesired
diseases and conditions resulting from inflammation and having
acute and/or chronic clinical exacerbations include multiple
sclerosis (Yednock et al., Nature 356, 63 (1992); Baron et al., J.
Exp. Med. 177, 57 (1993)), meningitis, encephalitis, stroke, other
cerebral traumas, inflammatory bowel disease including ulcerative
colitis and Crohn's disease (Hamann et al., J. Immunol. 152, 3238
(1994)), (Podolsky et al., J. Clin. Invest. 92, 372 (1993)),
rheumatoid arthritis (van Dinther-Janssen et al., J. Immunol. 147,
4207 (1991); van Dinther-Janssen et al., Annals Rheumatic Diseases
52, 672 (1993); Elices et al., J. Clin. Invest. 93, 405 (1994);
Postigo et al., J. Clin. Invest. 89, 1445 (1992), asthma (Mulligan
et al., J. Immunol. 150, 2407 (1993)) and acute juvenile onset
diabetes (Type 1) (Yang et al., PNAS 90, 10494 (1993); Burkly et
al., Diabetes 43, 529 (1994); Baron et al., J. Clin. Invest. 93,
1700 (1994)), AIDS dementia (Sasseville et al., Am. J. Path. 144,
27 (1994); atherosclerosis (Cybulsky & Gimbrone, Science 251,
788, Li et al., Arterioscler. Thromb. 13, 197 (1993)), nephritis
(Rabb et al., Springer Semin. Immunopathol. 16, 417-25 (1995)),
retinitis, atopic dermatitis, psoriasis, myocardial ischemia and
acute leukocyte-mediated lung injury such as occurs in adult
respiratory distress syndrome.
[0084] Inflammatory bowel disease is a collective term for two
similar diseases referred to as Crohn's disease and ulcerative
colitis. Crohn's disease is an idiopathic, chronic
ulceroconstrictive inflammatory disease characterized by sharply
delimited and typically transmural involvement of all layers of the
bowel wall by granulomatous inflammatory reaction. Any segment of
the gastrointestinal tract, from the mouth to the anus, may be
involved, although the disease most commonly affects the terminal
ileum and/or colon. Ulcerative colitis is an inflammatory response
limited largely to the colonic mucosa and submucosa. Lymphocytes
and macrophages are numerous in lesions of inflammatory bowel
disease and may contribute to inflammatory injury.
[0085] Asthma is a disease characterized by increased
responsiveness of the tracheobronchial tree to various stimuli
potentiating paroxysmal constriction of the bronchial airways. The
stimuli cause release of various mediators of inflammation from
IgE-coated mast cells including histamine, eosinophilic and
neutrophilic chemotactic factors, leukotrines, prostaglandin and
platelet activating factor. Release of these factors recruits
basophils, eosinophils and neutrophils, which cause inflammatory
injury.
[0086] Atherosclerosis is a disease of arteries (e.g., coronary,
carotid, aorta and iliac). The basic lesion, the atheroma, consists
of a raised focal plaque within the intima, having a core of lipid
and a covering fibrous cap. Atheromas compromise arterial blood
flow and weaken affected arteries. Myocardial and cerebral infarcts
are a major consequence of this disease. Macrophages and leukocytes
are recruited to atheromas and contribute to inflammatory
injury.
[0087] Rheumatoid arthritis is a chronic, relapsing inflammatory
disease that primarily causes impairment and destruction of joints.
Rheumatoid arthritis usually first affects the small joints of the
hands and feet but then may involved the wrists, elbows, ankles and
knees. The arthritis results from interaction of synovial cells
with leukocytes that infiltrate from the circulation into the
synovial lining of joints. See e.g., Paul, Immunology (3d ed.,
Raven Press, 1993).
[0088] Another indication for humanized antibodies against alpha-4
integrin is in treatment of organ or graft rejection. Over recent
years there has been a considerable improvement in the efficiency
of surgical techniques for transplanting tissues and organs such as
skin, kidney, liver, heart, lung, pancreas and bone marrow. Perhaps
the principal outstanding problem is the lack of satisfactory
agents for inducing immunotolerance in the recipient to the
transplanted allograft or organ. When allogeneic cells or organs
are transplanted into a host (i.e., the donor and donee are
different individuals from the same species), the host immune
system is likely to mount an immune response to foreign antigens in
the transplant (host-versus-graft disease) leading to destruction
of the transplanted tissue. CD8 cells, CD4.sup.+ cells and
monocytes are all involved in the rejection of transplant tissues.
Antibodies directed to alpha-4 integrin are useful, inter alia, to
block alloantigeninduced immune responses in the donee thereby
preventing such cells from participating in the destruction of the
transplantedtissue or organ. See, e.g., Paul et al., Transplant
International 9, 420-425 (0.1996); Georczynski et al., Immunology
87, 573-580 (1996); Georcyznski et al., Transplant. Immunol. 3,
55-61 (1995); Yang et al., Transplantation 60, 71-76 (1995);
Anderson et al., APMIS 102, 23-27 (1994).
[0089] A related use for antibodies to alpha-4 integrin is in
modulating the immune response involved in "graft versus host"
disease (GVHD). See e.g., Schlegel et al., J. Immunol. 155,
3856-3865 (1995). GVHD is a potentially fatal disease that occurs
when immunologically competent cells are transferred to an
allogeneic recipient. In this situation, the donor's
immunocompetent cells may attack tissues in the recipient. Tissues
of the skin, gut epithelia and liver are frequent targets and may
be destroyed during the course of GVHD. The disease presents an
especially severe problem when immune tissue is being transplanted,
such as in bone marrow transplantation; but less severe GVHD has
also been reported in other cases as well, including heart and
liver transplants. The therapeutic agents of the present invention
are used, inter alia, to block activation of the donor T-cells
thereby interfering with their ability to lyse target cells in the
host.
[0090] A further use of humanized antibodies of the invention is
inhibiting tumor metastasis. Several tumor cells have been reported
to express alpha-4 integrin and antibodies to alpha-4 integrin have
been reported to block adhesion of such cells to endothelial cells.
Steinback et al., Urol. Res. 23, 175-83 (1995); Orosz et al., Int.
J. Cancer 60, 867-(1995); Freedman et al., Leuk. Lymphoma 13, 47-52
(1994); Okahara et al., Cancer Res. 54, 3233-6 (1994).
[0091] A further use of the claimed antibodies is in treating
multiple sclerosis. Multiple sclerosis is a progressive
neurological autoimmune disease that affects an estimated 250,000
to 350,000 people in the United States. Multiple sclerosis is
thought to be a the result of a specific autoimmune reaction in
which certain leukocytes attack and initiate the destruction of
myelin, the insulating sheath covering nerve fibers. In an animal
model for multiple sclerosis, murine monoclonal antibodies directed
against alpha-4-beta-1 integrin have been shown to block the
adhesion of leukocytes to the endothelium, and thus prevent
inflammation of the central nervous system and subsequent paralysis
in the animals.
[0092] The humanized MAb 21.6 antibodies of the present invention
offer several advantages over the mouse antibodies already shown to
be effective in animals models:
[0093] 1) The human immune system should not recognize the
framework or constant region of the humanized antibody as foreign,
and therefore the antibody response against such an injected
antibody should be less than against a totally foreign mouse
antibody or a partially foreign chimeric antibody.
[0094] 2) Because the effector portion of the humanized antibody 25
is human, it may interact better with other parts of the human
immune system.
[0095] 3) Injected mouse antibodies have been reported to have a
half-life in the human circulation much shorter than the half-life
of normal human antibodies (Shaw et al., J. Immunol. 138:4534-4538
(1987)). Injected humanized antibodies have a half-life essentially
equivalent to naturally occurring human antibodies, allowing
smaller and less frequent doses.
[0096] The pharmaceutical compositions discussed supra can be
administered for prophylactic and/or therapeutic treatments of the
previously listed inflammatory disorders, including multiple
sclerosis, inflammatory bowel disease, asthma, atherosclerosis,
rheumatoid arthritis, organ or graft rejection and graft versus
host disease. In therapeutic applications, compositions are
administered to a patient suspected of, or already suffering from
such a disease in an amount sufficient to cure, or at least
partially arrest, the symptoms of the disease and its
complications. An amount adequate to accomplish this is defined as
a therapeutically- or pharmaceutically-effective dose.
[0097] In prophylactic applications, pharmaceutical compositions
are administered to a patient susceptible to, or otherwise at risk
of, a particular disease in an amount sufficient to eliminate or
reduce the risk or delay the outset of the disease. Such an amount
is defined to be a prophylactically effective dose. In patients
with multiple sclerosis in remission, risk may be assessed by NMR
imaging or, in some cases, by presymptomatic indications observed
by the patient.
[0098] The pharmaceutical compositions will be administered by
parenteral, topical, intravenous, oral, or subcutaneous,
intramuscular local administration, such as by aerosol or
transdermally, for prophylactic and/or therapeutic treatment.
Although the proteinaceous substances of this invention may survive
passage through the gut following oral administration,
subcutaneous, intravenous, intramuscular, intraperitoneal
administration by depot injection; or by implant preparation. are
preferred.
[0099] The pharmaceutical compositions can be administered variety
of unit dosage forms depending upon the method of administration.
For example, unit dosage forms suitable for oral administration
include powder, tablets, pills, capsules, and lozenges.
[0100] Effective doses of the compositions of the present
invention, for the treatment of the above described conditions will
vary depending upon many different factors, including means of
administration, target site, physiological state of the patient,
and other medicants administered. Thus, treatment dosages will need
to be titrated to optimize safety and efficacy. These compositions
may be administered to mammals for veterinary use and for clinical
use in humans in a manner similar to other therapeutic agents,
i.e., in a physiologically acceptable carrier. In general, the
administration dosage will range from about 0.0001 to 100 mg/kg,
and more usually 0.01 to 5 mg/kg of the host body weight.
[0101] In a preferred treatment regime, the antibody is
administered by intravenous infusion or subcutaneous injection at a
dose from 1 to 5 mg antibody per kilo of bodyweight. The dose is
repeated at interval from 2 to 8 weeks. Within this range, the
preferred treatment regimen is 3 mg antibody per kilo of bodyweight
repeated at a 4 week interval.
[0102] The humanized antibodies of the invention can be used with
effective amounts of other therapeutic agents against acute and
chronic inflammation. Such agents include antibodies and other
antagonists of adhesion molecules, including other integrins,
selectins, and immunoglobulin (Ig) superfamily members (see
Springer, Nature 346, 425-433 (1990); Osborn, Cell 62, 3(1990);
Hynes, Cell 69, 11 (1992)). Integrins are heterodimeric
transmembrane glycoproteins consisting of an a chain (120-180 kDa)
and a 0 chain (90-110 kDa), generally having short cytoplasmic
domains. For example, three important integrins, LFA-1, Mac-1 and
P150,95, have different alpha subunits, designated CD11a, CD11b and
CD11c, and a common beta subunit designated CD18. LFA-1
(.alpha..sub.L.beta..sub.2) is expressed on lymphocytes,
granulocyte and monocytes, and binds predominantly to an Ig-family
member counter-receptor termed TCAM-1 and related ligands. SCAM-1
is expressed on many cells, including eukocytes and endothelial
cells, and is up-regulated on vascular endothelium by cytokines
such as TNF and IL-1. Mac-1 (.alpha..sub.M.beta..sub.2) is
distributed on neutrophils and monocytes, and also binds to ICAM-1.
The third .beta.2 integrin, P150,95 (.alpha..sub.x.beta..sub.2) is
also found on neutrophils and monocytes. The selectins consist of
L-selectin, E-selectin and P-selectin.
[0103] Other antiinflammatory agents that can be used in
combination with the antibodies against alpha-4 integrin include
antibodies and other antagonists of cytokines, such as interleukins
IL-1 through IL-13, tumor necrosis factors .alpha. & .beta.,
interferons .alpha. & .beta. and .gamma., tumor growth factor
Beta (TGF-.beta.), colony stimulating factor (CSF) and granulocyte
monocyte colony stimulating factor (GM-CSF). Other antiinflammatory
agents include antibodies and other antagonists of chemokines such
as MCP-1, MIP-1a, MIP-1.beta., rantes, exotaxin and IL-8. Other
antiinflammatory agents include NSAIDS, steroids and other small
molecule inhibitors of inflammation. Formulations, routes of
administration and effective concentrations of agents for combined
therapies are as described above for the humanized antibodies
against alpha-4 integrin.
VIII. Other Uses
[0104] The humanized antibodies are also useful for affinity
purification of alpha-4 integrin. The antibodies are immobilized to
a solid support and a solution of dispersed proteins is passed over
the support. Alpha-4 integrin and associated 13 chain bind to the
support and is thereby separated from other proteins. The purified
alpha-4 integrin or a fragment thereof, made available by this
method, can be used as a vaccine or as an immunogen for producing
further antibodies.
[0105] The humanized antibodies of the invention are also useful
for generating idiotypic antibodies by, for example, immunization
of an animal with a humanized antibody. An anti-idiotype antibody
whose binding to the human antibody is inhibited by alpha-4
integrin or fragments thereof is selected. Because both the
anti-idiotypic antibody and the alpha-4 integrin or fragments
thereof bind to the humanized immunoglobulin, the anti-idiotypic
antibody may represent the "internal image" of an epitope and thus
may substitute a ligand of alpha-4 integrin, i.e., VCAM-1.
EXAMPLES
Example 1
Cloning and Sequencing of the Mouse 21.6 Variable Regions
[0106] The mouse anti-alpha-4 integrin antibody 21.6 has been
described in co-pending application U.S. Ser. No. 07/871,223. Total
RNA was isolated from hybridoma cells producing mouse 21.6
antibody. First-strand cDNA was synthesized using a kit (Pharmacia
Biosystems Limited). Heavy and light chain variable regions were
obtained by using PCR primers designed to hybridize to sequences
flanking and external to the sequences coding for the variable
regions, thereby allowing cloning of the entire coding sequences
for the mouse 21.6 antibody variable regions. Sense PCR primers
hybridizing to the 5'-ends of mouse kappa light-chain leader
sequences and of mouse heavy-chain leader sequences were designed
based on databases of 42 mouse kappa light-chain leader sequences
and of 55 mouse heavy-chain leader sequences (Jones & Bendig,
Bio/Technology 9:88-89 (1991) (incorporated by reference in its
entirety for all purposes)). These primers were used in conjunction
with anti-sense PCR primers hybridizing to the 3'-ends of the mouse
constant regions (kappa or gamma).
[0107] Mouse 21.6 kappa V.sub.L regions were PCR-amplified in a 50
ul reaction typically containing 10 mM Tris-HCl (pH 8.3), 50 mM
KCl, 200 .mu.M dNTPs, 1.5 mM MgCl.sub.2, 1 unit of AmpliTaq (Perkin
Elmer Cetus) DNA polymerase, 1 ul of cDNA template, 0.25 uM of MKV
primer and 0.25 uM of mouse kappa light chain anti-sense PCR primer
(FIGS. 1A and 1B). Mouse 21.6 V.sub.H regions were PCR-amplified as
described above except that MHVH primer and an anti-sense PCR
primer specific for the mouse IgG1 heavy chain constant region were
used (FIGS. 2A and 2B). Each PCR reaction was cycled, after an
initial melt at 94.degree. C. for 5 min, at 94.degree. C. for 1
min, 55.degree. C. for 1 min, and 72.degree. C. for 2 min over 25
cycles. The completion of the last cycle was followed by a final
extension at 72.degree. C. for 10 min. The ramp time between the
primer-annealing and extension steps was 2.5 min. Following PCR
amplification, 10 ul aliquots from each reaction were analyzed on
ethidium-bromide-stained 1.5% agarose gels.
[0108] The PCR products were cloned using the "TA Cloning System"
(Invitrogen Corporation). Vectors containing inserts of the correct
size were sequenced using double-stranded plasmid DNA and Sequenase
(United States Biochemical Corporation). To avoid any errors that
might have been introduced during the PCR amplification steps, at
least two independently PCR-amplified and cloned DNA fragments were
sequenced for each variable region.
[0109] The sequences of PCR products were compared with other mouse
light chain and heavy chain variable regions (see Tables 1 and 2).
This comparison indicated that the PCR products from MKV2 and MKV4
primers represent authentic mouse 21.6 kappa variable regions, and
those from MHV1 and MHV2 primers represent authentic mouse V.sub.H
regions, and it was concluded that the sequences of these product
are those of the mouse 21.6 antibody variable regions. The DNA and
amino acid sequences of the cDNA coding for the mouse 21.6 V.sub.L
and V.sub.H regions are shown in FIGS. 1 and 2.
TABLE-US-00001 TABLE 1 Comparison of the mouse 21.6 light chain
variable region to other light chain variable regions. Percent
Percent Mouse 21.6 V.sub.L versus: Similarity' Identity Consensus
sequence for 84.0 72.6 mouse kappa V,, subgroup 5.sup.2 Consensus
sequence for 84.0 69.8 human kappa V.sub.L subgroup 1.sup.2
Consensus sequence for 65.1 52.8 human kappa V.sub.L subgroup
2.sup.2 Consensus sequence for 72.6 57.5 human kappa V.sub.L
subgroup 3.sup.2 Consensus sequence for 72.6 58.5 human kappa
V.sub.L subgroup 4.sup.2 Sequence of V.sub.L from human REP.sup.3
81.0 72.4 (Member of human kappa V.sub.L subgroup 1) .sup.1Percent
similarity was determined using the "GAP" program of the University
of Wisconsin Genetics Computer Group. .sup.2Consensus sequences
were taken from Kabat et al., supra. .sup.3REI as sequenced by Palm
et al., Hoppe-Seyler's Z. Physiol. Chem 356: 167-191 (1975).
TABLE-US-00002 TABLE 2 Comparison of the mouse 21.6 heavy chain
variable region to other heavy chain variable regions. Percent
Percent Mouse 21.6 V.sub.H versus: Similarity' Identity Consensus
sequence for 94.3 91.1 mouse V.sub.H subgroup 2c.sup.2 Consensus
sequence for 78.0 65.0 human V.sub.H subgroup 1.sup.2 Consensus
sequence for 70.5 53.3 human V.sub.H subgroup 2.sup.2 Consensus
sequence for 67.5 52.8 human V.sub.II subgroup 3.sup.2 Sequence of
V.sub.H from human 21/28'CL.sup.3 76.5 64.7 (Member of human
V.sub.H subgroup 1) 'Percent similarity was determined using the
"GAP" program of the University of Wisconsin Genetics Computer
Group. .sup.2Consensus sequences were taken from Kabat et al.,
supra. .sup.321/28'CL as sequenced by Dersimonian et al., I.
Immunol. 139: 2496-2501 (1987).
Example 2
Construction of Chimeric 21.6 Antibody
[0110] Chimeric light and heavy chains were constructed by linking
the PCR-cloned cDNAs of mouse 21.6 V.sub.L and V.sub.H regions to
human constant regions. The 5'- and 3'-ends of the mouse CDNA
sequences were modified using specially designed PCR primers. The
5'-end PCR-primers (Table 3), which hybridize to the DNA sequences
coding for the beginnings of the leader sequences, were designed to
create the DNA sequences essential for efficient translation
(Kozak, J. Mol. Biol. 196:947-950 (1987)), and to create a HindIII
restriction sites for cloning into an expression vector. The 3'-end
primers (Table 3), which hybridize to the DNA sequences coding for
the ends of J regions, were designed to create the DNA sequences
essential for splicing to the constant regions, and to create a
BamHI site for cloning into an expression vector. The products of
PCR amplification were digested with HindIII and BamHI, cloned into
a pUC19 vector, and sequenced to confirm that no errors had
occurred during PCR amplification. The adapted mouse 21.6 variable
regions were then subcloned into mammalian cells expression vectors
containing either the human kappa or gamma-1 constant regions (FIG.
3).
TABLE-US-00003 TABLE 3 PCR primers for the construction of chimeric
21.6 antibody. A. Light chain variable region 1. Primer for
reconstruction of the 5'-end (37mer) (SEQ. ID NO: 18) 5' C AGA AAG
CTT GCC GCC ACC ATG AGA CCG TCT Hindlil Kozak M R P S Consensus
Sequence ATT CAG 3' I Q 2. Primer for reconstruction of the 3'-end
(35mer) (SEQ. ID NO: 19) 5' CC GAG GAT CCA CTC ACG TTT GAT TTC CAG
CTT BamHI Splice donor site GGT 3' B. Heavy chain variable region
1. Primer for reconstruction of the 5'-end (37mer) (SEQ. ID NO: 20)
5' C AGA AAG CTT GCC GCC ACC ATG AAA TGC AGC HindIII Kozak M K C S
Consensus Sequence TGG GTC 3' W V 2. Primer for reconstruction of
the 3'-end (33mer) (SEQ. ID NO: 21) 5' CC GAG GAT CCA CTC ACC TGA
GGA GAC GGT GAC T 3' BamHI Splice donor site
Example 3
Expression and Analysis of 21.6 Chimeric Antibody
[0111] The two plasmid DNAs coding for the chimeric 21.6 light and
heavy chains were cotransfected into Cos cells. After two or three
days, media from the Cos cells was analyzed by ELISA (1) for the
production of a human IgG-like antibody and (2) for the ability of
this human-like antibody to bind to L cells expressing human
.alpha.4/.beta.1 integrin on their surface. FIGS. 4 and 12 show
analyses of unpurified and protein-A purified samples of chimeric
21.6 antibody for binding to human a401 integrin, in comparison
with purified mouse 21.6 antibody control. These figures show that
the chimeric 21.6 antibody bound well to antigen and confirm that
the correct mouse 21.6 V.sub.L and V.sub.H regions had been
cloned.
Example 4
Modelling the Structure of the Mouse 21.6 Variable Regions
[0112] A molecular model of the V.sub.L and V.sub.H regions of
mouse 21.6 antibody was built. The model was built on a Silicon
Graphics IRIS 4D workstation running under the UNIX operating
system and using the molecular modelling package QUANTA (Polygen
corp., USA). The structure of the FRs of mouse 21.6 V.sub.L region
was based on the solved structure of human Bence-Jones
immunoglobulin REI (Epp et al., Biochemistry 14:4943-4952 (1975)).
The structure of the FRs of mouse 21.6 V.sub.H region was based on
the solved structure of mouse antibody Gloop2. Identical residues
in the FRs were retained; non-identical residues were substituted
using the facilities within QUANTA. CDR1 and CDR2 of mouse
21.6V.sub.L region were identified as belonging to canonical
structure groups 2 and 1, respectively (Chothia et al., supra).
Since CDR1 and CDR2 of REI belong to the same canonical groups,
CDR1 and CDR2 of mouse 21.6, V.sub.L region were modelled on the
structures of CDR1 and CDR2 of REI. CDR3 of mouse 21.6 V.sub.L
region did not appear to correspond to any of the canonical
structure groups for CDR3s of V.sub.L regions. A database search
revealed, however, that CDR3 in mouse 21.6 V.sub.L region was
similar to CDR3 in mouse HyHEL-5 V.sub.L region (Sheriff et al.,
Proc. Natl. Acad. Sci. USA 84:8075-8079 (1987)). Thus, the CDR3 of
mouse 21.6 V.sub.L region was modelled on the structure of CDR3 in
mouse HyHEL-5 V.sub.L region. CDR1 and CDR2 of mouse 21.6 V.sub.H
region were identified as belonging to canonical structure groups 1
and 2, respectively. CDR1 of mouse 21.6 V.sub.H region was modelled
on CDR1 of Gloop2 V.sub.H region which closely resembles members of
canonical group 1 for CDR1s of V.sub.H regions. CDR2 of mouse 21.6
V.sub.H region was modelled on CDR2 of mouse HyHEL-5 (Sheriff et
al., supra), which is also a member of canonical group 2 for CDR2
for V.sub.H regions. For CDR3s of V.sub.H regions, there are no
canonical structures. However, CDR3 in mouse 21.6 V.sub.H region
was similar to CDR3 in mouse R19.9 V.sub.H region (Lascombe et al.,
Proc. Natl. Acad. Sci. USA 86:607-611 (1989)) and was modelled on
this CDR3 by removing an extra serine residue present at the apex
of the CDR3 loop of mouse R19.9 V.sub.H region and annealing and
refining the gap. The model was finally subjected to steepest
descents and conjugate gradients energy minimization using the
CHARMM potential (Brooks et al., J. Comp. Chem. 4:187-217 (1983))
as implemented in QUANTA in order to relieve unfavorable atomic
contacts and to optimize van der Waals and electrostatic
interactions.
[0113] A view of the structural model of the mouse 21.6 variable
regions is presented in FIG. 5. The model was used to assist in
refining the design of the humanized 21.6 antibody variable
regions.
Example 5
Design of Reshaped Human 21.6 Variable Regions
[0114] (1) Selection of Homologous Human Antibodies for Framework
Sequence
[0115] Human variable regions whose FRs showed a high percent
identity to those of mouse 21.6 were identified by comparison of
amino acid sequences. Tables 4 and 5 compare the mouse 21.6
variable regions to all known mouse variable regions and then to
all known human variable regions. The mouse 21.6 V.sub.L region was
identified as belonging to mouse kappa V.sub.L region subgroup 5 as
defined by Kabat et al., supra. Individual mouse kappa V.sub.L
regions were identified that had as much as 93.4% identity to the
mouse 21.6 kappa V.sub.L region (38C13V'CL and PC613'CL). Mouse
21.6 V.sub.L region was most similar to human kappa V.sub.L regions
of subgroup 1 as defined by Kabat et al., supra. Individual human
kappa V.sub.L regions were identified that had as much as 72.4%
identity to the mouse 21.6 kappa V.sub.L region. The framework
regions (FRs) from one of the most similar human variable regions,
REI, were used in the design of reshaped human 21.6 V.sub.L region.
Mouse 21.6 V.sub.H region was identified as belonging to mouse
V.sub.H region subgroup 2c as defined by Kabat et al., supra.
Individual mouse heavy chain variable regions were identified that
have as much as 93.3% identity to the mouse 21.6 V.sub.H region
(17.2.25'CL and 87.92.6'CL). Mouse 21.6 V.sub.H region was most
similar to human V.sub.H regions of subgroup 1 as defined by Kabat
et al., supra. Individual human V.sub.H regions were identified
that had as much as 64.7% identity to the mouse 21.6 V.sub.H
region. The FRs from one of the most similar human variable
regions, 21/28'CL, was used in the design of reshaped human 21.6
V.sub.H region.
[0116] (2) Substitution of Amino Acids in Framework Regions
[0117] (a) Light Chain
[0118] The next step in the design process for the reshaped human
21.6 V.sub.L region was to join the CDRs from mouse 21.6 V.sub.L
region to the FRs from human REI (Palm et al., supra). In the first
version of reshaped human 21.6 V.sub.L region (La), seven changes
were made in the human FRs (Table 4, FIG. 6).
[0119] At positions 104, 105, and 107 in FR4, amino acids from RE1
were substituted with more typical human J region amino acids from
another human kappa light chain (Riechmann et al., Nature
332:323-327 (1988)).
[0120] At position 45 in FR2, the lysine normally present in REI
was changed to an arginine as found at that position in mouse 21.6
V.sub.L region. The amino acid residue at this position was thought
to be important in the supporting the CDR2 loop of the mouse 21.6
V.sub.L region.
[0121] At position 49 in FR2, the tyrosine normally present in REI
was changed to an histidine as found at that position in mouse 21.6
V.sub.L region. The histidine at this position in mouse 21.6
V.sub.L region was observed in the model to be located in the
middle of the binding site and could possibly make direct contact
with antigen during antibody-antigen binding.
[0122] At position 58 in FR3, the valine normally present in REI
was changed to an isoleucine as found at that position in mouse
21.6 V.sub.L region. The amino acid residue at this position was
thought to be important in the supporting the CDR2 loop of the
mouse 21.6 V.sub.L region.
[0123] At position 69 in FR3, the threonine normally present in REI
was changed to an arginine as found at that position in mouse 21.6
V.sub.L region. The arginine at this position in mouse 21.6 V.sub.L
region was observed in the model to be located adjacent to the CDR1
loop of mouse 21.6 V.sub.L region and could possibly make direct
contact with the antigen during antibody-antigen binding.
[0124] A second version of reshaped human 21.6 V.sub.L region
(termed Lb) was designed containing the same substitutions as above
15 except that no change was made at position 49 in FR2 of REI.
(FIG. 6).
[0125] (b) Heavy Chain
[0126] The next step in the design process for the reshaped human
21.6 V.sub.H region was to join the CDRs from mouse 21.6 V.sub.H
region to the FRs from 21/28'CL (Dersimonian et al., J. Immunol.
139:2496-2501 (1987)). In the first version of reshaped human 21.6
V. region (Ha), five changes were made in the human framework
regions (Table 5, FIG. 7). The five changes in the human FRs were
at positions 27, 28, 29, 30, and 71.
[0127] At positions 27, 28, 29, and 30 in FR1, the amino acids
present in human 21/28'CL were changed to the amino acids found at
those positions in mouse 21.6 V.sub.H region. Although these
positions are designated as being within FR1 (Kabat et al., supra),
positions 26 to 30 are part of the structural loop that forms the
CDR1 loop of the V.sub.H region. It is likely, therefore, that the
amino acids at these positions are directly involved in binding to
antigen. Indeed, positions 27 to 30 are part of the canonical
structure for CDR1 of the V.sub.H region as defined by Chothia et
al., supra.
[0128] At position 71 in FR3, the arginine present in human
21/28'CL was changed to a alanine as found at that position in
mouse 21.6 V.sub.H region. Position 71 is part of the canonical
structure for CDR2 of the V.sub.H region as defined by Chothia et
al., supra. From the model of the mouse 21.6 variable regions, it
appears that the alanine at position 71 is important in supporting
the CDR2 loop of the V.sub.H region. A substitution of an arginine
for an alanine at this position would very probably disrupt the
placing of the CDR2 loop.
[0129] A second version (Hb) of reshaped human 21.6 V.sub.H region
contains the five changes described above for version Ha were made
plus one additional change in FR2.
[0130] At position 44 in FR2, the arginine present in human
21/28'CL was changed to a glycine as found at that position in
mouse 21.6 V.sub.H region. Based on published information on the
packing Of V.sub.L-V.sub.H regions and on the model of the mouse
21.6 variable regions, it was thought that the amino acid residue
at position 44 might be important in the packing of the
V.sub.L-V.sub.H regions (Chothia et al., supra) (FIG. 5).
[0131] Reshaped human 21.6 V. region version Hc was designed to
make the CDR3 loop look more similar to human VCAM-1. Both mouse
21.6 antibody and human VCAM-1 bind to the .alpha.4/.beta.1
integrin. The CDR3 loop of the V.sub.H region of antibodies is the
most diverse of the six CDR loops and is generally the most
important single component of the antibody in antibody-antigen
interactions (Chothia et al., supra; Hoogenboom & Winter, J.
Mol. Biol. 227:381-388 (1992); Barbas et al., Proc. Natl. Acad.
Sci. USA 89:4457-4461 (1992)). Some sequence similarity was
identified between the CDR3 of mouse 21.6 V.sub.H region and amino
acids 86 to 94 of human VCAM-1, particularly, between the YGN
(Tyrosine-Glycine-Asparagine) sequence in the CDR3 loop and the FGN
(Phenylalanine-Glycine-Asparagine) sequence in VCAM-1. These
sequences are thought to be related to the RGD
(ArginineGlycine-Aspartic acid) sequences important in various cell
adhesion events (Main et al., Cell 71:671-678 (1992)). Therefore,
at position 98 in CDR3, the tyrosine present in mouse 21.6 V.sub.H
region was changed to a phenylalanine as found in the sequence of
human VCAM-1.
[0132] Possible substitution at position 36 in FR2 was also
considered. The mouse 21.6 VH chain contains an unusual cysteine
residue at position 36 in FR2. This position in FR2 is usually a
tryptophan in related mouse and human sequences (Table 5). Although
cysteine residues are often important for conformation of an
antibody, the model of the mouse 21.6 variable regions did not
indicate that this cysteine residue was involved either directly or
indirectly with antigen binding so the tryptophan present in FR2 of
human 21/28'CL V.sub.H region was left unsubstituted in all three
versions of humanized 21.6 antibody.
Example 6
Construction of Reshaped Human 21.6 Antibodies
[0133] The first version of reshaped human 21.6 V.sub.L region
(resh21.6VLa) was constructed from overlapping PCR fragments
essentially as described by Daugherty et al., Nucleic Acids Res.
19:2471-2476 (1991). (See FIG. 8). The mouse 21.6 V.sub.L region,
adapted as described in Example 2 and inserted into pUC19, was used
as a template. Four pairs of primers, APCR1-vla1, vla2-vla3,
vla4-vla5, and vla6-vla7 were synthesized (Table 6 and FIG. 8).
Adjacent pairs overlapped by at least 21 bases. The APCR1 primer is
complementary to the pUC19 vector. The appropriate primer pairs
(0.2 umoles) were combined with 10 ng of template DNA, and 1 unit
of AmpliTaq DNA polymerase (Perkin Elmer Cetus) in 50 ul of PCR
buffer containing 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 200 uM dNTPs,
and 1.5 mM MgCl.sub.2. Each reaction was carried out for 25 cycles.
After an initial melt at 94.degree. for 5 min, the reactions were
cycled at 94.degree. C. for 1 min, 55.degree. C. for 1 min, and
72.degree. C. for 2 min, and finally incubated at 72.degree. C. for
a further 30 min. The ramp time between the primer-annealing and
extension steps was 2.5 min. The products of the four reactions (A,
B, C, and D) from the first round of PCR reactions were
phenol-extracted and ethanol-precipitated.
TABLE-US-00004 TABLE 6 PCR primers for the construction of reshaped
human 21.6 variable regions. A. Light chain variable region 1.
Primers for the synthesis of version "a" 21.6VLa1 (39mer) (SEQ. ID
NO: 22): 5' GAT GGT GAC TCT ATC TCC TAC AGA TGC AGA CAG TGA GGA 3'
21.6VLa2 (32mer) (SEQ. ID NO: 23): 5' CTG TAG GAG ATA GAG TCA CCA
TCA CTT GCA AG 3' 21.6VLa3 (39mer) (SEQ. ID NO: 24): 5' AGG AGC TTT
TCC AGG TGT CTG TTG GTA CCA AGC CAT ATA 3' 21.6VLa4 (41mer) (SEQ.
ID NO: 25): 5' ACC AAC AGA CAC CTG GAA AAG CTC CTA GGC TGC TCA TAC
AT 3' 21.6VLa5 (40mer) (SEQ. ID NO: 26): 5' GCA GGC TGC TGA TGG TGA
AAG TAT AAT CTC TCC CAG ACC C 3' 21.6VLa6 (42mer) (SEQ. ID NO: 27):
5'ACT TTC ACC ATC AGC AGC CTG CAG CCT GAA GAT ATT GCA ACT3'
21.6VLa7 (59mer) (SEQ. ID NO: 28): 5' CCG AGG ATC CAC TCA CGT TTG
ATT TCC ACC TTG GTG CCT TGA CCG AAC GTC CAC AGA TT 3' 2. Primers
for the synthesis of version "b" 21.6VLb1 (33mer) (SEQ. ID NO: 29):
changes H-49 to Y-49 5' GGA AAA GCT CCT AGG CTG CTC ATA TAT TAC ACA
3' 21.6VLb2 (38mer (SEQ. ID NO: 30)): changes ACC-101 to ACA-101 to
destroy an Styl site 5' CCG AGG ATC CAC TCA CGT TTG ATT TCC ACC TTT
GTG CC 3' B. Heavy chain variable region 1. Primers for the
synthesis of version "a" 21.6VHal (51mer) (SEQ. ID NO: 31): 5' AAC
CCA GTG TAT ATA GGT GTC TTT AAT GTT GAA ACC GCT AGC TTT ACA GCT 3'
21.6VHa2 (67mer) (SEQ. ID NO: 32): 5' AAA GAC ACC TAT ATA CAC TGG
GTT AGA CAG GCC CCT GGC CAA AGG CTG GAG TGG ATG GGA AGG ATT G 3'
21.6VHa3 (26mer) (SEQ. ID NO: 33): 5' GAC CCG GCC CTG GAA CTT CGG
GTC AT 3' 21.6VHa4 (66mer) (SEQ. ID NO: 34): 5' GAC CCG AAG TTC CAG
GGC CGG GTC ACC ATC ACC GCA GAC ACC TCT GCC AGC ACC GCC TAC ATG GAA
3' 21.6VHa5 (64mer) (SEQ. ID NO: 35): 5' CCA TAG CAT AGA CCC CGT
AGT TAC CAT AAT ATC CCT CTC TGG CGC AGT AGT AGA CTG CAG TGT C 3'
21.6VHa6 (63mer) (SEQ. ID NO: 36): 5' GGT AAC TAC GGG GTC TAT GCT
ATG GAC TAC TGG GGT CAA GGA ACC CTT GTC ACC GTC TCC TCA 3' 2.
Primer for the synthesis of version "b" 21.6VHb (37mer) (SEQ. ID
NO: 37): changes R-44 to G-44 5' CCA GGG CCG GGT CAC CAT CAC CAG
AGA CAC CTC TGC C3' 3. Primer for the synthesis of version "c"
21.6VHc (27mer) (SEQ. ID NO: 38): changes Y-98 to F-98 5' CAG GCC
CCT GGC CAA GGG CTG GAG TGG 3' C. Both light and heavy chain
variable regions Primers hybridizing to the flanking pUC19 vector
DNA APCR1 (17mer (SEQ. ID NO: 39), sense primer) 5' TAC GCA AAC CGC
CTC TC 3' APCR4 (18mer (SEQ. ID NO: 40), anti-sense primer) 5' GAG
TGC ACC ATA TGC GGT 3'
[0134] PCR products A and B, and C and D were joined in a second
round of PCR reactions. PCR products A and B, and C and D, (50 ng
of each) were added to 50 ul PCR reactions (as described above) and
amplified through 20 cycles as described above, except that the
annealing temperature was raised to 60.degree. C. The products of
these reactions were termed E and F. The pairs of PCR primers used
were APCR1-vla3 and vla4-vla7, respectively. PCR products E and F
were phenol-extracted and ethanol-precipitated and then assembled
in a third round of PCR reactions by their own complementarity in a
two step-PCR reaction similar to that described above using APCR1
and vla7 as the terminal primers. The fully assembled fragment
representing the entire reshaped human 21.6 V.sub.L region
including a leader sequence was digested with HindIII and BamHI and
cloned into pUC19 for sequencing. A clone having the correct
sequence was designated resh21.6VLa.
[0135] The second version of a reshaped human 21.6 V.sub.L region
(Lb) was constructed using PCR primers to make minor modifications
in the first version of reshaped human 21.6 V.sub.L region (La) by
the method of Kamman et al., Nucl. Acids Res. 17:5404 (1989). Two
sets of primers were synthesized (Table 6). Each PCR reaction was
essentially carried out under the same conditions as described
above. In a first PCR reaction, mutagenic primer 21.6VLb2 was used
to destroy a Styl site (Thr-ACC-97 to Thr-ACA-97) to yield
resh21.6VLa2. Then, in a second PCR reaction, mutagenic primer
21.6VLb1 (His-49 to Tyr-49) was used with pUC-resh21.6VLa2 as
template DNA. The PCR product was cut with Styl and. BamHI and
subcloned into pUC-resh21.6VLa2, cleaved with the same restriction
enzymes. A clone with the correct sequence was designated
pUC-resh21.6VLb.
[0136] Version "a" of a reshaped human 21.6 V.sub.H region was
constructed using the same PCR methods as described for the
construction of version "a" of reshaped human 21.6 V.sub.L region
(Table 6 and FIG. 9). The HindIII-BamHI DNA fragments coding for
version "g" of reshaped human 425 V.sub.H region (Kettleborough et
al., supra) and version "b" of reshaped human AUK12-20 V.sub.H
region were subcloned into pUC19 vectors yielding pUC- and
pUC-reshAUK12-20b, respectively. (Version "b" of AUK12-20, was
derived by PCR mutagenesis of a fragment V.sub.Ha425 described by
Kettleborough et al., supra, and encodes the amino acid sequence
(SEQ. ID N0:41):
TABLE-US-00005 QVQLVQSGAEVKKPGASVKVSCKASGYSFT SYYIH WVRQAPGQGLEWV
GYIDPFNGGTSYNQKFKG KVTMTVDTSTNTAYMELSSLRSEDTAVYYCA R GGN-RFAY
WGQGTLVTVSS (spaces separate FR and CDR regions)).
Plasmid pUC-resh425g and pUC-reshAUK12-20b, as well as the pUC
vector containing the mouse 21.6 V.sub.H region as modified for use
in the construction of the chimeric 21.6 heavy chain
(pUC-chim21.6VH), were used as template DNAs in the subsequent PCR
reactions. PCR primers were designed and synthesized for the
construction of version "a" of reshaped human 21.6 V.sub.H region
(Table 6). PCR product A (FIG. 9) was obtained using
pUCreshAUK12-20b as DNA template and APCR1-vha1 as the PCR primer
pair. PCR products B and D were obtained using pUC-chim21.6VH as
DNA template and vha2-vha3 and vha6-APCR4 as PCR primer pairs,
respectively. Finally, PCR product C was obtained using
pUC-resh425g as DNA template and vla4-vla5 as the PCR primer pair.
The final PCR product was subcloned into pUC19 as an HindIII-BamHI
fragment for DNA sequencing. A clone with the correct DNA sequence
was designated pUC-resh21.6VHa. The DNA and amino acid sequences of
the first version of the reshaped 21.6 variable region are shown in
FIGS. 10A and 10B.
[0137] The remaining versions of reshaped human 21.6 V.sub.H region
were constructed essentially as described above for the
construction of version "b" of reshaped human 21.6 V.sub.L region.
Two sets of primers were synthesized (Table 6). For the second (Hb)
and third (Hc) versions, mutagenic primers 21.6VHb (Arg-44 to
Gly-44) and 21.6VHc (Tyr-98 to Phe-98), respectively, were used in
PCR reactions with pUC-resh21.6VHa as the template DNA. The PCR
products VHb and VHc were cut with restriction enzymes and
subcloned into pUC vector pUC-resh21.6VHa as MscI-BamHI and
PstI-BamHI fragments, respectively, to yield pUC-resh21.6VHb and
pUC-resh21.6VHc.
[0138] The first version of a reshaped human 21.6 V.sub.H region
(Ha) was constructed in a similar manner to that used for the
construction of the first version of reshaped human 21.6 V.sub.L
region (La). In this case, however, PCR primers were used with
three different template DNAs, mouse 21.6 V.sub.H region as already
adapted for expression of chimeric 21.6 heavy chain, humanized 425
V.sub.H region version "g" (Kettleborough et al., supra), and
humanized AUE12-20 version "b" V.sub.H region (Table 6, FIG. 9).
The DNA and amino acid sequences of the first version of the
humanized 21.6 heavy chain variable region are shown in FIGS. 11a
and 11B. The second and third versions of a humanized 21.6 V.sub.H
region (Hb and Hc) were constructed using PCR primers to make minor
modifications in the first version of humanized 21.6 V.sub.H region
(Ha) (Table 6).
Example 7
Expression and Analysis of Humanized Antibodies
[0139] 1. Linkage of Variable Regions to Constant Regions in
Expression Vectors
[0140] The DNA fragments coding for the chimeric and reshaped 21.6
V.sub.L and V.sub.H regions were subcloned into HCMV vectors
designed to express either human kappa light chains or human
gamma-1 heavy chains in mammalian cells (see FIG. 3) and Maeda et
al., Hum. Antibod. Hybridomas 2:124-134 (1991). Both vectors
contain the human cytomegalovirus (HCMV) promoter and enhancer for
high level transcription of the immunoglobulin light and heavy
chains. The light chain expression vector is exactly as described
in Maeda et al., supra, and contains genomic DNA coding for the
human kappa constant region (Rabbitts et al., Curr. Top. Microbiol.
Immunol. 113:166-171 (1984)). The heavy chain expression vector is
essentially as described in Maeda et al., supra, with the exception
that the genomic DNA coding for the human gamma-1 constant region
was replaced with a cDNA. cDNA coding for human gamma-1 constant
region was cloned from a human cell line that secreted a human
gamma-1 antibody by PCR. For convenient subcloning into the
expression vector, BamHI sites were created at each end of the
cDNA. In addition, a splice acceptor site and a 65 by intron
sequence were created at the 5'-end of the cDNA sequence. The BamHI
fragment (1176 bp) containing the human gamma-1 cDNA splice
acceptor site and intron sequence was substituted for the BamHI
fragment (approximately 2.0 kb) in the existing heavy chain vector
(Maeda et al., supra). The BamHI site to the 3'-side of the human
gamma-1 constant region was then removed with Klenow
polymerase.
[0141] 2. Transfection of Expression Vectors
[0142] Expression vectors were introduced into Cos cells by
electroporation using the Gene Pulsar apparatus (BioRad). DNA (10
Ag of each vector) was added to a 0.8 ml aliquot of
1.times.10.sup.7 cells/ml in PBS. A pulse was delivered at 1,900
volts, 25 AP capacitance. After a 10 min recovery period at ambient
temperature, the electroporated cells were added to 8 ml of DMEM
(GIBCO) containing 5% heat-inactivated gamma globulin-free fetal
calf serum. After 72 h incubation, the medium was collected,
centrifuged to remove cellular debris, and stored under sterile
conditions at 4.degree. C. for short periods of time, or at
-20.degree. C. for longer periods.
[0143] 3. Purification of Humanized Antibodies
[0144] Supernatants from Cos cell transfections were pooled and
purified on immobilized Protein A (ImmunoPure IgG Purification Kit,
Pierce). The supernatant was sterilized by filtration through a
0.22 um filter. After mixing with an equal volume of ImmunoPure IgG
binding buffer (pH 8.0), the diluted sample was applied to a 1 ml
protein A column and allowed to flow completely into the gel. After
washing with 15 ml of ImmunoPure IgG binding buffer, the bound
antibody was eluted with 5 ml of ImmunoPure IgG elution buffer (pH
2.8), and 1 ml fractions were collected. The pH of the first and
second fractions was approximately 8.0. The pH of the third
fraction was adjusted to physiological pH by the addition of 100 ul
of ImmunoPure binding buffer. The five 1 ml fractions containing
the Protein A-purified antibody were then assayed by ELISA to
determine the amount of human IgG antibody present in each
fraction. Antibody was detected using goat alkaline
phosphate-conjugated anti-human IgG (whole molecule, Sigma).
[0145] 4. Measurement of Binding Affinity
[0146] The binding of reshaped human 21.6 antibodies to a4.beta.1
integrin was assayed by ELISA in comparison with mouse and chimeric
antibodies. Briefly, L cells transformed to express a4.beta.1
integrin on their cell surface were plated out and grown to
confluence in 96-well tissue culture plates. The samples to be
tested (either crude supernatants or protein-A-purified) were
serially diluted and added to each well. After incubation for 1 h
on ice and very gentle washing, goat anti-mouse or anti-human
(gamma-chain specific) peroxidase conjugates (Sigma) were added.
After a further 1 h incubation on ice and very gentle washing, the
substrate (o-phenylenediamine dihydrochloride, Sigma) was added.
After incubation for 30 min at room temperature, the reaction was
stopped by adding 1 M H.sub.2SO.sub.4, and the A.sub.490 was
measured.
[0147] Results from analyzing crude supernatants of the two
versions of reshaped human 21.6 light chains (La and Lb), in
combination with version Ha of reshaped human 21.6 heavy chain,
indicated that the La version of reshaped human 21.6 V.sub.L region
gave slightly better binding to antigen than version Lb. The La
version was therefore used in subsequent experiments. Results from
analysis of the crude supernatants of humanized 21.6 heavy chains
(Ha and Hb), in combination with version La of humanized 21.6 light
chain, showed no significant difference between the two versions
(Ha and Hb) of reshaped human V.sub.H regions. Version Ha was
selected for use in further experiments because it contained only
five changes in the human FRs compared with six changes in the
human Hb.
[0148] FIG. 12A compares binding of humanized 21.6 antibody (La+Ha)
with chimeric 21.6 antibody. The data indicate that the reshaped
human 21.6 antibody (La+Ha) bound to antigen as well as, and
perhaps slightly better than, the chimeric 21.6 antibody. The
chimeric 21.6 antibody is expected to be equivalent to mouse 21.6
antibody in its antigen binding characteristics because it contains
the intact mouse 21.6 variable regions. The reshaped human 21.6
antibody (La+Ha) has also been shown to block binding to human
.alpha.4.beta.1 integrin with an efficiency comparable to the
original mouse 21.6 antibody and to the chimeric antibody. It is
therefore concluded that reshaped human 21.6 antibody (La+Ha) has a
specific binding affinity essentially equal to that of mouse 21.6
antibody. Moreover, because only minor modifications in the human
FRs were necessary to recreate the antigen binding site of mouse
21.6 antibody within human variable regions, the reshaped human
21.6 antibody is predicted to behave like an authentic human
antibody.
[0149] Reshaped human 21.6 antibody containing version La of the
reshaped human 21.6 VL region and version Hc of the reshaped human
21.6 VH region was also tested for binding to L cells expressing
human (.alpha.431 integrin on their surface in parallel with
chimeric 21.6 antibody. The results indicate that reshaped human
21.6 antibody (La+Hc) binds well to antigen. The alteration in the
CDR3 of the V.sub.H region did not impair binding to antigen.
Indeed, there is some indication that the alteration in the CDR3
may have slightly improved binding to antigen (FIG. 12B).
Conceivably, the improvement may be more pronounced in a functional
blocking assay.
Example 8
Blocking Properties of Mu 21.6 Antibody
[0150] Mu 21.6 was compared with another antibody against a.sub.4
integrin called L25. L25 is commercially available from Becton
Dickinson, and has been reported in the literature to be a good
inhibitor of .alpha.4/.beta.1 integrin adhesive function. As shown
in FIG. 13 (Panel A), both Mu 21.6 and L25 completely inhibited
.alpha.4/.beta.1 integrin-dependent adhesion of human monocytic
cells to purified VCAM-1 in the absence of Mn.sup.4-2. However, in
the presence of Mn.sup.+2 (1 mM) (one of several activators of
a.sub.4.beta..sub.1 integrin) L25 was no longer an effective
inhibitor. Similar results were observed when a.sub.4.beta..sub.1
integrin was activated by other stimuli. The capacity to block
activated .alpha.4/.beta.1 integrin is likely to be of value in
treating inflammatory diseases such as multiple sclerosis.
[0151] As a further comparison between mu 21.6 and L25, the
capacity of antibody to inhibit human T cell adhesion to increasing
amounts of VCAM-1 was determined. In this experiment, increasing
amounts of VCAM-1 were coated onto plastic wells of a 96 well assay
plate, and the ability of the human T cell line, Jurkat (which
expresses high levels of a.sub.40.sub.1 integrin), to bind to the
coated wells was measured. Values on the Y-axis represent the
percentage of Jurkat cells originally added to each well that
remained bound after washing the well four times (FIG. 13 (Panel
B)). This experiment demonstrates that L25 is a good inhibitor of
cell adhesion when low levels of VCAM-1 are encountered, but
becomes completely ineffective at higher levels of VCAM-1. Mu 21.6,
on the other hand, inhibits cell adhesion completely, regardless of
the amount of VCAM-1 present. The capacity to block at high
concentrations of VCAM-1 is desirable for therapeutic applications
because of upregulation of VCAM-1 at sites of inflammation.
Example 9
Efficacy of Humanized 21.6 Antibody in an Animal Model
[0152] This example establishes the efficacy of humanized 21.6
antibody in prophylactic and therapeutic treatment of EAE in an
animal model simulating multiple sclerosis in humans.
[0153] (a) Methods
[0154] (1) Induction of EAE
[0155] The brain and spinal cord were removed from each of five
guinea pigs euthanized by CO.sub.2 narcosis. The tissue was kept in
PBS on wet ice until it was weighed and homogenized at a
concentration of 1 gram of tissue per ml PBS. The tissue was
completely homogenized using an electric hand-held homogenizer and
subsequently mixed with an equal volume of Freund's complete
adjuvant (FCA). FCA was made by adding 100 mg of mycobacterium
tuberculosis H37 RA (DIFCO, 3114-33-8) to 10 ml of Freund's
incomplete adjuvant (Sigma, F-5506). The mixture was emulsified
into the consistency of mayonnaise by passing the solution between
two syringes connected by a two way stopcock. Each guinea pig was
immunized with 600 ul emulsion divided between three sites of
administration.
[0156] (2) Scoring Animals for Disease Symptoms
[0157] The disease symptoms were assessed by prompting each animal
to walk and assigning the animal a score by the following commonly
accepted criteria: [0158] 0 No disease [0159] 1 Hind limb weakness
[0160] 2 Complete hind limb paralysis [0161] 3 Complete hind limb
and some forelimb paralysis [0162] 4 Moribund or dead
[0163] (3) Serum and Tissue Collection
[0164] Samples were collected by cardiac puncture from
methoxyflurane-anesthetized guinea pigs. About 300-400 ul of blood
were collected and placed in microtainer serum separator and
allowed to clot for between 20-30 min at room temperature. The tube
was then spun for 5 min at room temperature. The serum was drawn
off into Eppendorf tubes and stored at -20.degree. C. for
subsequent analysis of antibody titers by fluorescence activated
cell sorting (FACS).
[0165] For hematological analysis, blood was collected into
ethylenediaminetetraacetic acid-coated microtainer tubes. A 100 ul
aliquot was aspirated into an acridine-coated hematocrit tube. The
tube was capped and the blood was mixed with acridine orange by
gently inverting the tube 15 times. A float was put into the
hematocrit tube and the sample was centrifuged for 5 minutes. The
hematocrit tube was placed into a precalibrated Idexx QBC Vet
Autoreader designed for quantitative buffey coat analysis. Values
were read under the 5 horse calibration system and adjusted to
guinea pig equivalents using a predetermined conversion factor.
[0166] At the end of the experiment, the guinea pigs were killed by
CO.sub.2 narcosis and the brains and spinal cords removed. Half of
the brain and spinal cord from every guinea pig was snap frozen in
2-methyl butane on dry ice (-20 to -40.degree. C.). This tissue was
cut and immunostained with a pan macrophage marker (Serotec
MCA-518) and a T-lymphocyte marker (Serotec MCA-751) using the
avidin-biotin linking peroxidase assay (Vector Laboratories, Inc.,
Burlingame, Calif.) and diaminobenzidine as a chromagen. The tissue
was scored for cellular infiltration according to the following
scoring system: [0167] 0 No infiltrating cells. [0168] 0.5 Very
little staining; may be artifactual; usually 20 associated with
vessels. [0169] 1 Staining of a few cells (less than 15) usually
near a vessel. [0170] 2 Staining of many cells (20-50), usually
radiating out from a vessel. [0171] 3 Staining of many cells
(>50) scattered throughout the tissue; many cuffed vessels.
[0172] (b) Prophylactic Treatment
[0173] This experiment was designed to evaluate the efficacy of
humanized 21.6 antibody in delaying the onset of clinical symptoms.
Previous data have demonstrated that leukocyte influx into the
brain and spinal cord of EAE guinea pigs typically starts between
day 7 and day 8. Therefore, antibodies were administered on day 7
and on day 10 post-immunization. To compare mouse and humanized
21.6 antibody, equivalent doses of each of the antibodies (3.0,
0.30 and 0.03 mg/kg) were administered. Preliminary pharmacokinetic
studies revealed that saturating blood levels of mouse 21.6
antibody were attained within 24 hours after subcutaneous
administration, and remained elevated up to 48 hours.
[0174] On day 11, 24 hours after the second dose of antibody, blood
samples were drawn from three randomly selected animals in each
group. For each treatment group a mean for the number of days for
each guinea pig to reach a clinical score of 1 was calculated
(Table 7). The mean value for the PBS-treated group in this
experiment was 11 days post-immunization (which is typical of
previous results). Treatment with the highest dose of humanized and
mouse antibody resulted in a significant delay of disease by 4.6
(p=0.000) and 3 (p=0.007) days, respectively. The lower doses of
antibody had no effect on the course of disease.
TABLE-US-00006 TABLE 7 Effect of mouse or humanized 21.6 antibody
on time post immunization to reach a clinical score of 1. GROUP 1 2
3 4 5 6 7 mg/kg 0.03 M.sup.# 3.0 H.sup.@ 3.0 H 3.0 M 0.03 H PBS 0.3
M 8 9 13 10 8 9 9 9 10 15. 12 10 9 9 9 10 15 14 10 11 11 9 11 16 14
11 11 12 11 11 16 14 12 11 12 12 11 16 15 12 12 13 12 12 17 15 12
12 13 13 17 18 12 13 Mean + 10.0 .+-. 10.9 .+-. **15.6 .+-. *14.0
.+-. 10.9 .+-. 11.0 11.6 .+-. SD 1.6 1.2 1.3 2.3 1.5 1.4 1.4
.sup.@H denotes humanized antibody; *M denotes mouse. **p = 0.000
and *p = 0.007, as compared to PBS.
[0175] Daily body weights of the guinea pig reflected a similar
effect of the high doses of humanized and mouse antibody. (FIG.
14). Animals in these treatment groups steadily gained weight.
Guinea pigs in all other treatment groups lost 35 weight starting
from just before the day of onset of disease.
[0176] Serum titers of antibody were measured in three randomly
selected animals from each group by cardiac puncture on day 11,
roughly 24 hr after the second treatment. Efficacy of the
antibodies to delay disease correlated tightly with serum levels.
About 20 mg/ml serum antibody was present in the circulation of all
animals injected with the highest dose of both humanized and mouse
antibodies. This concentration is of the same order of magnitude as
the concentration of 21.6 antibody required to saturate alph-4
integrin sites in vitro. In contrast, animals from all other groups
had little to no detectable serum antibody.
[0177] (c) Reversal of On-Going Disease
[0178] About 60 guinea pigs were immunized and allowed to develop
clinical symptoms of EAE. On day 13, all guinea pigs that attained
a clinical score of 1 were randomly assigned to a treatment group.
FIG. 15 shows that animals treated with 3 mg/kg humanized antibody
began to recover hind limb function within 48 hr of treatment. On
Days 17 and 18, one and two days after the second dose, all eight
animals were disease free. ANOVA of the area under the curve values
for each treatment group revealed that only the 3 mg/kg humanized
antibody treated group value was statistically lower than the PBS
control group (p=0.042). These animals progressively gained weight
within 24 hrs after the first administration until the experiment
was terminated on Day 19 (FIG. 16).
[0179] Antibody serum titers were measured by FACS analysis on
samples taken 24 hrs after the first injection (Day 14) and at
sacrifice (Day 19). Treatment with mouse 21.6 antibody resulted in
slightly lower serum antibody titers than treatment with humanized
21.6 antibody (9.1 vs. 12.6 .mu.g/ml). This difference became more
profound on Day 19, three days after the second administration,
when there was very little detectable serum mouse antibody, while
the levels of humanized antibody on Day 19 had dropped below
saturating but were still measurable (6.1 ug/ml). These data
demonstrate a correlation between plasma levels of antibody and
physiologic efficacy and suggest that the effective circulating
antibody level is in the range of 10-20 ug/ml in the guinea
pig.
[0180] Leukocyte infiltration onto brain and spinal cord was
evaluated in tissue from animals killed on Day 19. Table 8 shows
significant differences in the degree of infiltration as a function
of antibody treatment. The reduction in T cell infiltration into
brain and spinal cord and macrophage infiltration into spinal cord
was significant after treatment with 3 mg/kg. Lower doses tended to
reduce infiltration, but did not reach significance. There was no
significant difference in cellular infiltrate of macrophages into
the spinal cord at any dose. Since the immunohistochemical
technique used to evaluate macrophages does not distinguish
resident from invading cells, the lack of effect on macrophages
likely represents the sustained presence of resident macrophages
and microglia.
[0181] The reduction in T-cells and monocytes in brain tissue by
administration of the antibody after establishment of the disease
suggests that cell trafficking is not a cumulative process, but a
dynamic movement of cells into and out of CNS tissue. Importantly,
the data suggest that interruption of the entry of leukocytes into
parenchymal tissue allows the CNS to rid itself of the invading
pathological element.
TABLE-US-00007 TABLE 8 Significant differences in T-cell and
macrophage infiltration into brain and spinal cord on Day 129.
BRAIN SPINAL CORD GROUP MACRO- MACRO- PBS T-CELLS PHAGES T-CELLS
PHAGES 3 mg/kg @ H p = 0.001 p = 0.005 p = 0.007 NS 3 mg/kg # M p =
0.001 p = 0.005 p = 0.008 NS lmg/kg H NS NS NS NS 0.3 mg/kg H NS NS
NS NS NS = not significant.
[0182] Hematology data revealed that treatment with mouse or
humanized 21.6 antibody caused no difference in whole white blood
cell counts, mononuclear and granulocyte number or in red blood
cell count. The high dose of mouse or humanized antibody resulted
in a significant increase in platelet counts as compared to PBS
treated animals (Table 9). In normal guinea pig platelet counts are
755.+-.103 cells/ml, about double that of PBS-treated EAE animals.
Thus, treatment with doses of mouse and humanized antibody that
effectively reversed disease, also restored platelet count to
normal.
TABLE-US-00008 TABLE 9 Effect of antibody treatment on platelet
count in EAE animals. TREATMENT PLATELETS .times. 10.sup.-6
CELLS/ML ++Non EAE guinea pigs 755 .+-. 103 (9) PBS 373.7 .+-.
167.5 (7) 3 mg/kg @ H 622.2 .+-. 97.0 (6)** 3 mg/kg # M 587.5 .+-.
57.8 (6) 1 mg/kg H 578.3 .+-. 123.2 (6) 0.3 mg/kg H 492.5 .+-.
168.6 (6) .sup.++Platelet counts in non-EAE guinea pigs were
determined in a separate experiment. *p = 0.05 vs PBS.
[0183] In conclusion, both humanized and mouse 21.6 antibodies are
effective in delaying and reversing clinical symptoms in an animal
model simulating multiple sclerosis in humans. The humanized
antibody is more effective than the same dosage of mouse antibody
in reversing symptoms.
[0184] Although the foregoing invention has been described in
detail for purposes of clarity of understanding, it will be obvious
that certain modifications may be practiced within the scope of the
appended claims. All publications and patent documents cited above
are hereby incorporated by reference in their entirety for all
purposes to the same extent as if each were so individually
denoted.
Sequence CWU 1
1
* * * * *