Characterization of B1(Fv)PE38 and B1(dsFv)PE38: Single

Vol.
1, 1023-1029,
September
Clinical
1995
Characterization
and
of B1(Fv)PE38
Disulfide-stabilized
That
Cause
Carcinoma
Itai
Benhar
Laboratory
and
Diagnosis
Maryland
Remissions
Xenografts
in Nude
and Centers,
20892-4255
with
Cancer
of Cancer
Institute,
NIH,
Bethesda,
which
The mAb B! (mouse IgG1 K) recognizes
a carbohydrate
on human carcinoma
cells (I. Pastan et al., Cancer
Res., 51: 3781-3787,
1991). We have generated
plasmids
encoding
immunotoxins
in which
the Fv domain of B!, either as
a single-chain
Fv or as a disulfide-stabilized
Fv (dsFv), was
fused to PE38, a tnmcated
form of Pseudomonas
exotoxin
A.
To compare
the activities
of the two types of recombinant
the
proteins
produced
ins were
evaluated
toxicity,
pharmacokinetics,
mouse
model.
tively
stable
were
prepared
in Escherichia
for stabifity,
antigen
and
Although
from
coil.
and
the single-chain
spedfic
cyto-
in a nude
activity
immunotoxin
is rela-
when incubated
at 37#{176}C
(t#{189} -4
h), the dsFv
immunotoxin
is much more stable, with no hoss ofactivity
after
8 h at 37#{176}C.
The single-chain
immunotoxin
has a 2-fold better
binding
affinity
and
cytotoxicity
hired
cells
bhood
of mice of B1(Fv)PE3S
toward
antigen-positive
cal-
The half-lives in the
and B1(dsFv)PE38
(disulfide-stabilized)
are 23 and 27 mm, respectively.
Their
therapeutic
potential
was evaluated
in athymic nude mice bearing human
epidermoid
carcinoma
xenografts.
Both immunotoxins caused complete
regressions
of the s.c. (30-40
mm3)
tumors
when given i.v. in three doses of 0.025 mg/kg every
other day. This is one-twentieth
of the mouse LD.
Recombinant immunotoxins
containing
the B1(Fv) are 2-3-fold
more
potent
antitumor
agents
than previously
described
immunotoxins
containing
the B3(Fv) (Brinlunann
et aL, Proc. Nail.
Acad. Sd. USA, 88: 8616-8620,
1991),
which
also target
Le’
and related
carbohydrates
in human
tumors,
but have a similar toxicity in mice. Thus, their therapeutic
window
is 2-3-fold
larger.
In addition,
B1(dsFv)PE38
has only a 50% decrease
in
the apparent
binding
affinity
of B1(Fv)PE38,
whereas
B3(dsFv)PE38
has a much greater
loss in antigen binding.
than
the dsFv
is
fused
moiety.
as well
breast,
stomach,
colon,
tomas
(for
review,
clinical
trials.
The
immunotoxin.
(single-chain)
domains
framework
revised
3/31/95;
accepted
Refs.
and
2,
of these
form,
unstable
vary
with
a high
in their
The
abbreviations
used
are:
HSA, 0.2%
dsFV,
disulfide-stabilized
human
serum
albumin.
Fv;
Oligo,
and
therein).
evaluated
immunotoxins
VH and
the
of many
BR96
with
some
being
(3, 20).
dsFvs2
and some
also
to aggregate
are much
carcinomas
react
because
with
more
stable,
(3, 21).
Le’
but
it has a different
carbohydrate
structures
mimitope.3
The
the targeting
as anticancer
Bi
pattern
(21),
of immunotoxins
both a whole
recombinant
immunotoxin
(1, 21-23).
with BR96
A single-chain
(1 1). We seek
with
improved
drug
will
increase
have
the
with
immunotoxin
that B1(Fv)PE38
and had a higher
was
initially
oh3
I. Benhar
(3).
addressed
significant
et a!.,
unpublished
data.
epitope
used
as
have
has
potent
binding
and
been
also been
immuno-
properties
so
so that
less
a response.
To that
and initial charBi as the single-
In that
study,
we
found
stable at 37#{176}C
than B3(Fv)PE38,
affinity.
The stability
problem
by making
form
B3(dsFv)PE38
(13, 14).
found
that the disulfide-stabilized
has a very
B3 and
to a B3 peptide
window,
to achieve
B1(Fv)PE38
and
sur-
Le”-related
immunotoxin)
and specific
is much more
antigen-binding
and
show
are currently
immunotoxin
to make more
therapeutic
to be administered
mAbs
to bind
very
that are being developed
IgG conjugate
and a single(Fv
stability
like
end, we have previously
reported
the cloning
acterization
of the variable
domain
of mAb
chain
VL
Lewis’’
on the
a different
of reactivity
and Bi fails
made
made
we can
Bi
with
B3 IgG or its Fv fragments
moiety
agents:
that
mAb
reacts
chain
toxins
VL
1). More
in which
VH and
improved
antigen-binding
affmity
(19).
mAb Bi is a murine
antibody
directed
against
related
carbohydrate
antigens,
which
are abundant
face
in preare usu-
the
where
for
of the
glioblas-
references
in which
stability,
tendency
To whom requests for reprints should be addressed,
at Laboratory
of
Molecular
Biology, National Cancer Institute, NIH, Building 37, Room
4E16, 37 Convent Drive, MSC 4255, Bethesda,
MD 20892-4255.
2
drugs
immunotoxins
form,
I
godeoxynucleotide;
prostate,
being
the
and
carcinomas
and
are currently
as
generated
as candidate
lung
in a single-chain
serves
been
models
1 and
Fv domains
Fv immunotoxins
munotoxin
5/4/95.
that
have
are connected
by a disulfide
bond engineered
into the
region
(12-19).
Single-chain
Fvs and single-chain
stabilized
recently
1/31/95;
fragment
in
targeting
are connected
by a flexible
peptide
linker (3-i
Fv and Fv immunotoxins
have been described
domains
recently,
dsFv
as the
and leukemias,
ovary,
see
Fv immunotoxins
present
toxin
as in animal
Several
ally
a
molecules
serves
immunotoxins
of bymphomas
of B3(Fv)PE38
Received
to
Fv
tested in vitro
the eradication
are chimeric
of an antibody
the Fv is in a disulfide-stabilized
The immunotox-
binding,
antitumor
cytoplasmic
Fv immunotoxins
an Fv fragment
moiety
epitope
bodies
Activity
Human
Recombinant
Biology
ABSTRACT
inclusion
Increased
Mice
cytotoxic
immunotoxins,
1023
INTRODUCTION
Division
National
Research
Single-Chain
of Established
Pastan’
Biology,
B1(dsFv)PE38:
Fv Immunotoxins
Complete
Ira
of Molecular
and
Cancer
reduction
the disulfide-
However,
we
B3(dsFv)PE38
in apparent
have
im-
binding
1024 Activity
of B1Fv
affinity.4
analyzed
Immunotoxins
Since no other anti-La”
antibody
fragment
has been
in the dsFv form, we wished
to investigate
whether
the
dsFv
technology
with
better
describe
is applicable
immunotoxin
and
of Bi
the
characterization
and cytotoxic
stability
of
totoxic
stable
activity
B1(dsFv)PE38
we
of both
We find
-
that
-
its correspond-
a 2-4-fold
decrease
for
-
in
experiments
its
diminished
Oligos
AND
cy-
B
which
the single-chain
encodes
been
described
pBl(Fv)PE38
expression
immunotoxin.
plasmids
The
Plasmid
pB1VL
cysteine,
was
A1OSC
pBl(Fv)PE38,
(24)
was
used
STOP,
which
by
a
encodes
reaction
oligos
(Fig.
1, Oligos
1-3)
were
S’-GGTCATCACAACATCCATATGTATAT-
oligo
STOP
4:
Oligo
5:
STOP
delete
VH
VL
R44C
A1O5C
STOP
4-5
R44C
the
of the dsFv
in Fig. 1.
for VL with
mutagenesis
3:
Oligo
A1O5C
uracil-
to construct
for the components
scheme
is described
Oligo
H
delete
PE38
Oligos
C
has
using
VL
Im-
immunotoxin,
mutagenesis
encoding
mutagenesis
of B1Fv
plasmid
as template
generated
mutagenic
The
of
B1(Fv)PE38
(3). Site-specific
containing
pB1VL
for Expression
1:
‘FA1OSC
METHODS
of Plasmids
The
construction
2:
1.3
I
Construction
munotoxins.
Oligo
Oligo
Mutagenesis
cancer
cells, the two types of
suggesting
that the increased
compensated
I
pB 1(Fv)PE3S
activity.
MATERIALS
three
comparison
In antitumor
in vitro.
A431 epidermoid
are equally
active,
Here,
and in vivo.
it has
EcoR
B1(dsFv)PE38
at 37#{176}C
than
but
HindIII
antibody
affinity.
and
in vitro
more
immunotoxin,
in mice using
immunotoxins
anti-Le’’
binding
of the disuffide-stabilized
is much
single-chain
binding
the
Fv immunotoxins
B1(dsFv)PE38
ing
for
the preparation
forms
to a different
consequences
PstI
A
a free
in
PBIVH
R44C-PE3S
Fig. 1 Plasmids
for expression
of Bi Fv immunotoxins.
pBl(Fv)PE38
(A) codes
for the single-chain
immunotoxin
B1(Fv)PE38
(3). In this
plasmid,
the V
and VL domains
of mAb
Bi are connected
by a
(Gly4Ser)3
linker and then fused to a PE38 gene encoding
the translocation and ADP ribosylation
domains
of Pseudomonas
exotoxin
A (PE).
which
combined.
CTCC-3’
sequences
(Fig.
1, Oligo
1) was
used for the deletion
of the
encoding
VH so that VL was juxtaposed
to the ATG
The plasmids encoding
the components
of the double-chain
immunotoxins were prepared
by site-specific
mutagenesis
(24), where uracil-
translational
the oligos
GCGG-3’
initiation
codon.
It was used in combination
S’-CTCCAGCVFGGTACCACAACCGAACGTGA(Fig. 1, Oligo
2), which
was used to replace
containing
pBl(Fv)PE38
single-stranded
DNA was used as a template.
In the first mutagenesis
reaction
Oligos
1-3 were
combined
in one
mutagenesis
reaction,
with Ohigo 1 producing
deletion
of VH and juxtaposing
VL immediately
3’ to the ATG initiation
codon.
Oligo 2 caused
replacement
of VL alanine
codon
105 with a cysteine
codon,
and Ohigo
3 caused
insertion
of two tandem
translation
termination
codons
immediately
3’ to VL. This
resulted
in the generation
of plasmid
pB1VL
A1O5C STOP (B), which encodes
for B1VL with a free cysteine.
In the
second
mutagenesis
reaction,
Ohigos
4 and 5 were
combined
in one
mutagenesis
reaction,
with Oligo
4 causing
deletion
of VL and juxtaposing
V immediately
5’ to PE38.
Ohigo 5 caused
replacement
of VH
arginine
codon 44 with a cysteine
codon.
This resulted
in the generation
of plasmid pB1VH R44C-PE38
(C), which encodes
for the VH with a
free cysteine,
fused to PE38.
The ohigos are shown
below
their respective positions
in the template
to which
they anneal.
In Ohigos
1 and 4,
which
are deletion-mutagenesis
ohigos,
thin lines indicate
absent
Sequences (these oligos anneal to uncontiguous
sequences
in the template).
codon
codon,
AG-3’
TAA
105
and
in framework
region
4 of VL
S’-GGGACCTCCGGAAGCGAAfl’CVFAmC-
(Fig.
1, Oligo
translational
1, underlined)
which
3),
stop
at the
was
codons
3’ end
of
Plasmid
pB1VH
R44C-PE38,
a free cysteine,
fused to PE38
esis
reaction
in which
The
with
two
tandem
by an EcoRI
site
(Fig.
which
encodes
for the VH
was generated
by a mutagen-
mutagenic
oligos
(Fig.
1, Oligos
4
combined.
oligo
CAGT-3’
S’-GGGACOCCGGAAGCVVVFGCAGAGA-
(Fig.
encoding
with the
two
to insert
alanine
a cysteine
VL.
with
and 5) were
used
followed
with
with
1, Oligo
was
4)
the peptide
linker and
PE38 coding
sequence.
used
to delete
the VL
It was
that
used
the
sequences
VH was flushed
in combination
S ‘-GGTFGCGACCCACTCGAGACACTTCTCCGGA-
GT-3’
(Fig.
codon
codon
1, Oligo
5),
which
was
44 in the framework
region
as described
(13, 16).
Expression
Expression
B1VH
and
R44C-PE38
single-chain
refolding
immunotoxin
of
of Recombinant
encoding
were
inclusion
B1(Fv)PE38,
expressed
was
body
to replace
2 of VH with
Purification
plasmids
used
B1VL
in Escherichia
obtained
protein
Proteins.
A1OSC,
coli.
described
(25).
immunotoxin
was
equimolar
amounts
of solubilized
VH R44C-PE38
inclusion
bodies
a cysteine
by solubilization
as
disubfide-stabilized
arginine
or
The
and
The
except
that
was carried
separated
quentiab
Mono
phy
the final oxidation
out at pH 9.5.
step was omitted
Properly
refolded
from contaminating
proteins
ion exchange
chromatography
Q (Pharmacia)
on
a TSK
followed
G3000SW
4
Y. Reiter
and I. Pastan,
unpublished
data.
binding
by
to A431
competition
adenocarcinoma
by
mixing
and refolding
proteins
were
and aggregates
on Q-Sepharose
by size
(TosoHaas)
exclusion
by
seand
chromatogra-
column.
Antigen Binding and Specific
toxins.
Relative binding affinities
determined
obtained
and reduced
VL A1OSC
and
essentially
as described
(15),
against
Cytotoxicity
of Immuno-
of the immunotoxins
‘25I-labeled
Bi
cells
at 4#{176}C
as described
IgG
were
for
(26).
Clinical
The
cytotoxic
were
tested
activities
synthesis
in cultured
cells
Assay.
The
stability
treatment
was
determined
in
PBS
at
chromatography
37#{176}Cfor
on
quantitate
activities
the untreated
mouse
LD50
old,
different
above,
was
-20
p.1 PBS-HSA.
Mice
injection.
after
immunotoxins
were
Blood
at 2, 5, 10, 20,
30,
60,
data
for the various
three
mean
centration
of
ability
by incubating
A431
cells,
each
serum
bated
using
was
Antitumor
mice
were
cells
suspended
mm3.
and
s.c.
Mice
volumes
and
of the
with
were
curve,
RSTRIP
days
4,
6,
a caliper
and
using
i.v.
and
lized
27);
Methods”
ered
Fig.
system,
Fig.
immunotoxins
separate
the dsFv
buffer
raphy
and
1. The
were
cultures
immunotoxin
used
(15,
by refobding
single-chain
body
followed
by ion exchange
as described
in “Materials
2, proteins
12-16
over
g
wet
95%
cell
pure
paste
express
immunotoxins
affinities
used
(XDE3;
components
immunotoxins
were
protein
obtained.
are
of the
were
the
obtained
As
to A431
a higher
Ref.
of
was
protein
analyzed
serial
(3).
pared
after
2 or 20 h of incubation
mAb,
when
3B,
B1(Fv)PE38
active
with
dilutions
The
ofthe
activities
tested
on A431
has an IC50
In this
type
the
which
an IC50 of 3.0 ng/ml
1 liter
the
cell
5
I. Benhar
and I. Pastan,
unpublished
IgG
to
incorporation
lines after
containing
ng/ml
B1(dsFv)PE38
after
com-
As shown
strongly
observations.
IgG
of
treat-
were
the cells.
and 0.4 ng/ml
Bi
B1(Fv)PE38
has
activity
of each
in PBS
in
of
Bi
of ‘25I-labeled
of 0.4 and 0.25
of assay,
as-
immunotoxin
immunotoxins
with
cells
of their
competed
for the
by 50% at 1.2 i.M.
In our expression
from
each
immunotoxin
of
Activity
of
antigen bind-
of ‘25I-labeled
was assessed
by measuring
by various
human
carcinoma
BSA
mg
found that the
immunotoxin
by competition
of
for the binding
with
to 1 liter
of 20-30
determination
in Fig. 3A, B1(Fv)PE38
Bi IgG to A431 cells
0.2%
Fig.
by
cells
for the binding
ment
10were
yields for Fv immunotoxins
with Fv forms of B15 (16).
cells by 50% at about 2.5 riM. Thus,
affinity
for A431 cells. The cytotoxic
immunotoxin
F3H]leucine
of both
was
added
yield
concentrations
competed
mg pro-
yield
immunotoxin
production
on antigen-positive
to compete
20 h of incubation.
recov-
chromatogshown
recombinant
increasing
B1(dsFv)PE38
in a redox-shuffle
and gel filtration
and Methods.”
were
ing
binding
formula:
disulfide-stabi-
in E. coli BL21
to
17). The
of inclusion
and
monomeric
this
solubibi-
200-300
purification
immunotoxins
an overall
from
Upon
Binding
and Specific
Cytotoxic
and B1(dsFv)PE38.
The specific
in which
in “Materials
3 h, and
bodies,
and
active
mg
cells at 4#{176}C.
As shown
binding
of 125I-labeled
is described
produced
were
Antigen
B1(Fv)PE38
says,
X (0.4).
plasmids
mg
for
weight).
from 1 liter of bacterial
culture. We have previously
refobding
and purification
yield of a recombinant
intervals,
the
refolding
correlates
with its stability,
with 2-4%
that are less stable than those prepared
of immunotoxins
in which the targeting
moiety
is the Fv of mAb
Bi, we prepared
the recombinant
single-chain
B1(Fv)PE38
and
the disulfide-stabilized
B1(dsFv)PE38
immunotoxins.
The conof the expression
10-12
with
RESULTS
Plasmid
Construction
and Production
of B1(Fv)PE38
and B1(dsFv)PE38.
To characterize
the therapeutic
potential
struction
12%.
3.0
=
(wet
disulfide-stabilized
100
in PBS-
at 2-day
and
from
about
8 by
single-chain
Thus,
at A
bodies
of the inclusion
The
buffer,
(nu/nu)
diluted
reduction
obtained.
refobding
(mi-
were
and
are
induced
g inclusion
recovered
3 X 106 A431
4, tumors
tein
culture
2-3
in
UT).
0 with
flask
on
calcu-
Athymic
shaker
zation
ob-
toxins
program,
calculated
X (width)2
incu-
were
Mice.
on
con-
by
concentration
City,
Fig. 2 Recombinant
BI Fv immunotoxins.
SDS-PAGE
of purified
immunotoxins.
M, molecular
weight
standards
indicated
on the left in
kDa. Lanes
1 and 3, B1(Fv)PE38;
Lanes
2 and 4, B1(dsFv)PE38.
Samples
in Lanes
3 and 4 were reduced
while samples
in Lanes I and
2 were not.
we obtain
The
parameters
By day
to
repre-
measuring
injected
toxin
of immunotoxins
measured
(length)
mice.
cells
on day
treated
point
A standard
Lake
used
determined
A431
fitting
ml PBS.
time
-
Each
were
three
was
in Nude
doses
were
tumor
=
Each
the
Salt
were
of different
the
injection.
mice
from
injections
in 0.2
Tumors
after
mm
Pharmacokinetic
Activity
given
240
points.
curve
weeks
into
with
Software,
(6-8
vein
dilutions
an exponential
mice
for
of the
the orbital
to determine
sample.
loss or death
by injection
synthesis.
serial
with
in 200
from
toxin
protein
(6 -8
of 10 p.g B1(Fv)PE38
so different
serum
mice
collected
obtained
used
Scientific
volume
29
18
14
single-
diluted
BALB/c
and
active
to inhibit
The
_
43
of
i.v. injection
p.1 PBS-HSA
time
results
of the
tamed
HSA.
activities
BALB!c
a single
i.v. dose
times,
of remaining
dilutions
female
for weight
were
120,
collect
injections
at 37#{176}C
were
to the
the pharmacokinetics
in 200
bled
30-40
to
Cytotoxic
in Mice.
female
samples
was
croMath
97
68
analytical
column,
or B1(dsFv)PE38
a single
mouse
their
by
incubated
monitored
diluted
the tail vein.
34
at 0.1
aggregates.
compared
To evaluate
given
or B1(dsFv)PE38
bating
and
given
in the blood,
g) were
the
followed
(TosoHaas)
using
were
of B1(Fv)PE38
2 weeks
sents
8 h,
and
tested
g), who
doses
-20
to
G3000SW
Pharmacokinetics
and
2
fob-
incubation
immunotoxins.
Toxicity
weeks
Ml
protein
immunotoxins
by
of immunotoxins
as described
1025
B1(dsFv)PE38
to inhibit
of the
of monomers
of aliquots
determined
up
a TSK
the amount
and
ability
Research
(22).
heat
mg/ml
old,
B1(Fv)PE38
of their
Stability
lowing
dose
of the
by determination
Cancer
bind
in
the BI
after
was
2 or
less
2 or 20 h of
1026
Activity
of B1Fv
Immunotoxins
Table
A
-C
C
C
0
.0
1
Cytotoxicity
tM
B1(Fv)PE38
(ngiml)
B1(dsFv)PE38
Epidermoid
carcinoma
+++
MCF7
LnCap
Breast carcinoma
Prostate
+ +
KB
carcinoma
Cervical
carcinoma
-
>1000
>1000
HUT1O2
T-cell
leukemia
-
>1000
>1000
L929
Mouse
fibroblast
-
>1000
3-1
CompetItor
+
+
0.25
0.4
0.6
2.7
1.5
9.0
>1000
a
All of the cell lines except
L929 are of human
S
Expression
of the B1 antigen
level
nofluorescence.
C Cytotoxicity
C
0
toward
IC50’
Cytotoxicity
Antigen
expression
Source
B1(dsFv)PE38
A431
In
B
and
lines
.
Cell
line”
C.,
of B1(Fv)PE38
various
cell
immunotoxin
its incubation
estimation
weak;
+ + +, strong;
+,
data are given
is based
on immu-
not detected.
values,
the concentration
,
-
as IC50
that causes
a 50% inhibition
on the cells for 20 h.
origin.
of protein
synthesis
of
following
2
0
0.
I.
-
0
cJ
C
C
0
.-
II.
0
be incubated
at 37#{176}C
for more
its
activity.6
cytotoxic
Pharmacokinetics
netic
behavior
levels
Antigen
binding
(#{149},
0) and B1(dsFv)PE38
competition
of I-habeled
cytotoxicity
toward
A431
incorporation
of [3H]leucine
(S,
3
A) of incubation
in PBS
0.2%
+
and specific
cytotoxicity
of B1(Fv)PE38
(A, Li). A, antigen
binding
was estimated
by
Bi IgG binding
to A431
cells at 4#{176}C.
B,
cells
was measured
by the inhibition
of
into cell protein
after 2 h (0, L) or 20 h
of the cells with serial dilutions
of immunotoxins
or
B1(Fv)PE38
in the samples
its dilutions
comparison
notoxins
were
to standard
had a similar
amounts
observed
notoxins,
lines
21;
the same
which
see
differ
Table
B1(dsFv)PE38
cancer
cell lines
activity
toward
correlates
cytotoxic
1).
As
shown
the
same
toward
Stabffity
of aggregation
“Materials
aggregated
icity
of
loss
in PBS
of activity
showed
8
h
very
immunotoxins
the fraction
each
point.
We have
following
recently
loss
of
pharmacokiin mice
was
was
aggregation
incubation
found
protein
4B,
synthesis
as well
as to other
of both
was
sisted
that B1(dsFv)PE38
in
(15),
have
toxicity
of single
by i.v injections
doses
of different
into BALB!c
mice. The mice were
postinjection.
As shown
in Table 2, the
immunotoxins
was
a similar
Activity
of three
about
mouse
0.5
LD50
mg/kg.
of 0.5
of Immunotoxins.
i.v. injections
immunotoxin.
month
on days
Control
mice
Similar
mg/kg
Fv
(15).
The therapeutic
was initiated
on day
in volume,
and con-
4, 6, and 8 of various
were
treated
or longer
day
occurred
X 3. At 0.0125
at 0.05
and
and 0.00625
mg/kg
thereafter.
euthanized
Control
on day
animals
18 when
tumor
developed
with
at
6
I Benhar
et al.,
manuscript
mg/kg
every
of the
doses
PBS-HSA
in preparation.
given
other
day
treatment
large
tumors
X 3,
or a few
and were
1.0 cm.
1
every
(more
its completion
size was about
at 37#{176}C
can
0.025
tumor growth was arrested
for the duration
so with B1(dsFv)PE38)
and resumed
after
days
8 h of
still present
cells
in Fig. 5, both immubehavior
when
comFv immunotoxins
The
evaluated
the cytotox-
in PBS
A431
mm.
was
Antitumor
of the serum
on
potential
of the immunotoxins
was evaluated
by assessing
their
ability
to cause
regressions
of established
human
carcinoma
xenografts
in nude mice. Tumors
were induced
by s.c. injection
other
4 h
aggregated.
following
in Fig.
in
immunoafter
by the ability
protein
only. As shown in Fig. 6, both immunotoxins
had a dose-dependent
antitumor
effect. Complete
regressions
of the tumors that lasted
amount
of B1(Fv)PE38
75%
of monomeric
both
4A,
whereas
half
about
little
im-
of B1(Fv)PE38
of their
in Fig.
at 37#{176}C.
As shown
with
stability
the incubation,
at 37#{176}C,
about
after
of each
estimated
of 3 X 106 A431 cells on day 0. Treatment
4 when
the tumors
averaged
30-40
mm3
activity
at 37#{176}C
as described
As shown
before
The
curves.
As shown
pharmacokinetic
of immunotoxin
for 7-14 days
immunotoxins
of
their
Methods.”
of cytotoxic
cytotoxic
affinity
and
of the
site.
The
correlated
time
bevels
The
the binding
binding
cell
(Ref.
B1(Fv)PE38
different
cells.
by determination
in PBS
the
1,
of recognition
tested
and
incubation
expression
Table
having
with
monomeric
B1(dsFv)PE38
on additional
was
and
of incubation
done
of Immunotoxins.
and
were
in
LD50
of the immu-
antigen
spectrum
its cellular
and B1(dsFv)PE38
toxins
albeit
was
of Bi
the antigen-positive
approximately
munotoxin
level
had
tested
the specificity
assay
in their
<10%
B1(dsFv)PE38
of Immunotoxins.
Toxicity
To check
and
to inhibit
pared to each other
with a t#{189}Of 23-27
BSA.
respectively.
with
of Immunotoxins.
of
of the immunotoxins
incubation,
1 week
determined
by measuring
the immunotoxin
bevels in blood samples drawn from the mice at various
time points following
an i.v.
injection
of 10 xg of each immunotoxin.
The immunotoxin
ng/ml
Fig.
than
Clinical
A
ic
1._.
1._
t=0
o.C
0.1
o.t.
0.1
0.1
0.4
0.8
‘r
8h
0.1
0.1
E Lk
0..
:.
Research
I
1._
4h
2h
Cancer
“c
.
I.q
10
15
0.0
20
#{149}#{149}I
5
25
0.1
0..
15
20
us-..-’
0-c
.
10
25
3i 3
10
15
20
25
i 0
1.c
t=0
‘I)
2h
0.1
0.8
0.1
or
0.4
0.4
0.
0..
Sh
-C
u
oc
5
10
20
15
25
30
0.0
0.6
0.8
0.2
0.4
.
5
10
15
20
25
Elutlon
J
10
30
volume
0_I
or
0.4
o..
iS
.s-.u
4h
20
25
0_c
30
5
10
15
25
20
30
(ml)
C
B.!
Ce
I0
0.
I-
0
c.
e
.!
C
0
41
t=O
2hr
S
4hr
_____
0
.;
C
I
A
61
8hr
.01
1
.1
10
.01
.1
nglml
Fig.
4
Stability
0
ng/mI
of immunotoxins.
B1(Fv)PE38
and
B1(dsFv)PE38
were
diluted
in PBS
to 0.1
molecular
forms of the immunotoxins
were then analyzed by size exclusion
chromatography
Monomers
elute at 18-20 ml, while the aggregates
elute at 11-13 ml. B, cytotoxic
activity
was measured
as in Fig. 3.
B1(Fv)PE38
DISCUSSION
Here,
immunotoxin
we
describe
the
B1(dsFv)PE38,
than the corresponding
immunotoxin
had the
single-chain
activity
when
construction
and
counterpart
single-chain
was
B1Fv
are
relatively
or
Fv
(Table
compared
as well
stable
immunotoxins
B1(dsFv)PE38
show
that
mAb
Bi
it is more
dsFv
stable
single-chain
immunotoxin.
The dsFv
same antigen-binding
specificity
as the
1) but only
and binding
affinity
(Fig.
compared
to a 20-fold
B3(dsFv)PE38
of the
is even
3). This
lower
to its
as the
in comparison
(3).
more
of its cytotoxic
loss is modest
found
when
single-chain
form.4
immunotoxin
single-chain
the
disulfide-stabilized
(Fig.
4), and
stability
of B1(dsFv)PE38
appears
to compensate
loss in binding
affinity
in cytotoxicity
assays.
when tumor cells are exposed
to the immunotoxins
The
B1(Fv)PE38
to other
However,
stable
half
affinity
affinity
the
improved
for the small
For example,
for only 2 h,
is
B1(dsFv)PE38
times
were
ng/ml
for
extended
and
munotoxin
ability
IC50s
ng/mh
space
a simibar
with
show
extracellular
were
for
ng/ml)
very
that
the
distributed
antitumor
of 64,000,
rapidly
weight
of the drug
of B1(Fv)PE38
BSA.
Assuming
and B1(dsFv)PE38
that albumin
and
into
The
of
cells
is inversely
molecular
is very similar
the recombinant
5).
the readily
efficacy
hatter
(28).
into
(Fig.
for the tumor
The
to the molecular
to or
6). B1(Fv)PE38
pharmacokinetics
of its affinity
(0.25
Fur-
weight
to distribute
once
into the tumor.
similar
is equal
(Fig.
appear
than
the incubation
B1(dsFv)PE38).
molecular
similar
space,
is a function
0.4
of B1(dsFv)PE38
i.v. into mice
experiments
to penetrate
0.4
and Methods.”
at 37#{176}C
in PBS
when
that of B1(Fv)PE38
have
the extravascular
(IC50,
However,
activity
than
injected
active
to 20 h, the
better
B1(dsFv)PE38
accessible
more
B1(Fv)PE38
somewhat
and when
at 37#{176}C
for 0 to 8 h. A, the
incubated
3.0 ng/ml).
the antitumor
and
and
much
(IC50,
thermore,
In vitro
Fvs
of
mg/mI
at 4#{176}C
as described
in ‘ ‘Materials
immunotoxins
following
incubation
an
im-
and its
related
weight
to that of
immuno-
1027
1028 Activity
of B1Fv
Immunotoxins
Table
Groups
of four or two BALB/c
Days
postinjection
Diluent
Toxicity
i.v.
B1(Fv)PE38
of
B1(Fv)PE38
(mg/kg)
0.25
0.5
0.125
and B1(dsFv)PE38
of 200
injections
in mice
p.1 dil uent or increas ing doses
0.75
o f B1(Fv)PE38
or B1(dsFv)PE38
B1(dsFv)PE38
(mg/kg)
Diluent
0.125
0.25
0.5
0.75
2/2
2/2
2/2
4/4
4/4
4/4
2/2
4/4
4/4
0/4
2
2/2
2/2
4
2/2
2/2
4/4
4/4
3/4
2/4
2/4
0/4
7
2/2
2/2
3/4
2/4
0/4
2/2
2/2
4/4
3/4
0/4
2/2
2/2
3/4
2/4
0/4
2/2
2/2
4/4
3/4
0/4
14
Number
a
2
mice were given
of mice
alive
of mice
given
injections.
0
0
0
.0
Mice
A
C
treated
with
B1(Fv)PE38
C,)
E
C.)
C
‘4
8)
0
0
N
0)
C
0
E
E
E
I-
E
0
30
60
90
120
150
post
mInutes
180
210
0
240
Blood levels of B1(Fv)PE38
and B1(dsFv)PE38
in mice. Female BALB/c mice were given i.v. injections
of 10 p.g immunotoxin.
The immunotoxin
level was measured
at different times using a cytoFig.
4
8
12
16
20
24
20
24
Injection
B
5
toxicity
assay as described
in ‘ ‘ Materials
and Methods.
mean from three mice for each time point ± SE.
‘ ‘
Results
MIce
treated
4
8
with
B1(dsFv)PE38
are the
C.)
E
C.)
8)
N
toxins
behave
LS1477T
similarly
tumors
and that A431
(29),
we calculate
6 h for the recombinant
Since
toxins
the t112 of B1(Fv)PE38
tumors
that
are comparable
it should
to equilibrate
is about
take
within
to
about
5-
the tumors.
4 h at 37#{176}C
(Fig.
4), there
U)
0
E
I-
should
be significant
inactivation
of single-chain
immunotoxin
molecules
during
this period,
whereas
no dsFv toxin molecules
would
be inactivated.
Thus,
a more stable
dsFv immunotoxin
can probably
penetrate
tivated
in comparison
Another
factor
more stable
site barrier”
deeper
into a tumor before
being
to a single
chain Fv immunotoxin.
that
can
contribute
to the superiority
Fv immunotoxin
in tumor
therapy
(30, 31). According
to this model,
themselves,
being antigen
positive,
can form
the cells which are more proximal
to the tumor
a “sink”
for the drug,
preventing
of the tumor
substance.
in the number
of free diffusible
into
the tumor.
A more
the additional
delay
reaches
the interior
with
reduced
affmity
The
antitumor
specific
activity
than
reported
that
This
stable
it from
phenomenon
molecules
dsFv
macof a
is the “binding
the tumor cells
a barrier,
vasculature
reaching
results
the distal
so that
form
parts
should
other
immunotoxins
by this barrier.
that
Fig. 6
Antitumor
effect
of B1(Fv)PE38
and B1(dsFv)PE38
in a nude
mouse
model.
Groups
of five mice were given s.c. injections
of 3 X 106
cells on day 0, and were treated by iv. injections
of B1(Fv)PE38
(A) or
B1(dsFv)PE38
(B) diluted
in PBS-HSA
on days 4, 6, and 8 (arrows)
when the tumors
PBS-HSA.
Error
0.025
mg/kg; U,
were
bars,
0.0125
established.
Control
mice were treated with
SE of the data. #{149},
control; 0, 0.05 mg/kg; U,
mg/kg; A, 0.00625 mg/kg.
deeper
cytotoxicity
on cultured
cells as well as the
of B1(Fv)PE38
and B1(dsFv)PE38
is better
for
16
endure
caused by that barrier, so that more active drug
of the tumors.
Furthermore,
an immunotoxin
will be bess affected
12
Time/days
in a decrease
that can penetrate
immunotoxin
0
target
related
carbohydrate
mas (15, 22). Since
antigen
on the surface
of human
B1(Fv)PE38
and B1(dsFv)PE38
carcinohave the
same toxicity
in mice as those other immunotoxins,
they have
larger therapeutic
window
in mice. This makes them candidates
for development
as anticancer
drugs
for clinical
application.
a Le’#{176}’- However,
since
adult
mice
do
not
express
the
Bi
or the
a
B3
Clinical
antigen,
the
toxicity
destruction
of
is nonspecific,
normal
immunotoxins
cells.
should
monkeys,
which
be
express
not
i.e.,
Therefore,
the
evaluated
antigens
next
due
to targeted
toxicity
in animals
reacting
with
experiments,
yet
the more
stable
more
active.
We
next
to study
blood
supply
may
be more
the immunotoxins
activity
of
plan
need
the
more
molecule
heterogeneous
to move
stable
such
as
Bi
on
15. Reiter,
Antitumor
tumors
treatment
to ascertain
are
more
that
19: 35-47,
by
17.
Reiter,
1. Pastan,
I., Pai, L., Brinkmann,
U., and
toxins:
new therapeutic
agents
for cancer.
345-354,
1995.
2. Brinkmann,
chim.
Biophys.
3. Benhar,
I., and Pastan,
I. Cloning
the Fv fragments
of the anticarbohydrate
B5 as single-chain
4. Bird,
Single-chain
423-426,
against
K. D., Jacobson,
J. W., Johnson,
Lee, T., Pope, S. H., Riordan,
antigen
1988.
binding
19.
cancer.
proteins.
Science
S., Kaufman,
G. S., and Withlow,
(Washington
DC),
6. Kreitman,
R. J., Chaudhary,
V. K., Kozak,
R. W., FitzGerald,
D. J.,
Waldman,
T. A., and Pastan,
I. Recombinant
toxins
containing
the
variable
domains
of the anti-Tac
monoclonal
antibody
to the interleukin-2 receptor
kill malignant
cells from patients
with chronic
lymphocytic leukemia.
Blood,
80: 2344-2352,
1992.
7. Brinkmann,
U., Pai, L., FitzGerald,
protease
sensitive
region
of Pseudomonas
in the circulation
of mice. Proc. Natl.
1992.
8. Chaudhary,
V. K.,
Queen,
C.,
D., and Pastan,
I. Alteration
of a
exotoxin
prolongs
its survival
Acad. Sci. USA, 89: 3065-3069,
(Land.),
9. Batra,
339:
394-397,
J. K., Kasprzyk,
Recombinant
toxin.
Proc.
anti-erbB2
Nath. Acad.
R. P.,
Waldmann,
T. A.,
consisting
Friedman,
P.,
P. G., Bird, R. E., Pastan,
immunotoxins
Sci. USA, 89:
McAndrew,
S.,
Reiter,
containing
5867-5871,
I., and King,
Pseudomonas
1992.
C. R.
exo-
Gawlak,
S.,
Chace,
D.,
Trail,
P.,
Y.,
Webber,
K.,
framework
regions.
Jung,
S-H.,
Lee,
B.,
and
Brinkmann,
U.,
nani,
PCR.
antibodies
and
Bi
269:
Protein
and
B3 that
Lee,
Pseudomonas
B.,
Kasprzyk,
and antitumor
P. G.,
activity
stabilization
18327-18331,
of
of the
1994.
B. S., and Wilson,
B. S. Selection
of an
from a protein
linker library
prepared
by
14: 256-262,
E. T., Gallo,
Wilhingham,
truncated
by disulfide
Biotechniques,
I., Lovelace,
J. L.,
S. H.,
binding
immunotoxin
W. P. C., Morris,
chain Fv antibody
inverse
Jung,
I. Improved
J. Biol. Chem.,
Pastan,
Fv.
U., and Pastan,
I. Cytotoxic
immunotoxin
composed
of
Fv fragment
and
142-149,
1994.
20. Stemmer,
active single
enzymatic
disulfide-stabilized
R. J., Brinkmann,
a recombinant
of
anti-erbB2
M.
react
C.
with
1993.
M. G., Rutherford,
A. V., Mag-
Characterization
mucinous
of
monoclonal
adenocarcinomas.
Can-
1991.
22. Brinkmann,
U., Pai, L. H., FitzGerald,
D. J., Wilhingham,
M., and
Pastan, I. B3(Fv)-PE38KDEL,
a single-chain
immunotoxin
that causes
complete
Sci.
regression
USA,
88:
of a human
8616-8620,
carcinoma
in mice.
Proc.
Nath.
Acad.
1991.
23. Pai, L. H., Gallo, M. G., FitzGerald,
D. J., and Pastan, I. Antitumor
activity
of a transforming
growth factor a-Pseudomonas
exotoxin fusion
protein
(TGF-ca-PE4O).
Cancer
Res., 51: 2808-2812,
1991.
24.
T. A. Rapid
Kunkel,
phenotypic
25.
selection.
Buchner,
and efficient
Proc.
J., Pastan,
Natl.
site-directed
Acad.
Sci.
I., and Brinkmann,
mutagenesis
USA,
82:
without
488-492,
U. A method
1985.
for increasing
the yield of properly
folded
recombinant
fusion
proteins:
immunotoxins
from
renaturation
of bacterial
inclusion
Biochem.,
205:
263-270,
1992.
single-chain
bodies.
Anal.
Benhar, I., Brinkmann,
U., Webber, K. 0., and Pastan, I. Mutations
of two lysine residues in the CDR hoops of a recombinant
immunotoxin
reduce
27.
Studier,
merase
M., Pfitzinger,
I., and
stabilize
immunoglobulin
Pl#{252}ckthun, A.
Fv-fragments.
A
1990.
13. Brinkmann,
U., Reiter,
Y., Jung,
S. H., Lee, B., and
recombinant
immunotoxin
containing
a disulfide-stabihized
ment.
Proc. Nath. Acad.
Sci. USA,
90: 7538-7542,
1993.
Pastan,
Fv
I. A
frag-
sensitivity
to
chemical
F. W.,
and
Moffatt,
B. A. Use
expression
R. Physiological
of cloned
barriers
Yuan, F., Leunig, M., Huang,
D., and Jam, R. K. Microvascular
tion of stericably
nograft.
Cancer
Weinstein,
stabilized
Res., 54:
J. N., Eger,
of bacteriophage
T7
gene. J. Mol. Biol.,
poly189:
Cancer
liposomes
1994.
R. R., Covell,
antibodies.
of monoclonal
Res., 50: 814s-819s,
in a human
D. G., Black,
S. M., and Keenan,
Ann.
in solid
Cancer
tumor
xc-
C. D., Mulshine,
A. M. The pharma-
N. Y. Aced,
Sci.
31. Juweid,
M., Neumann,
R., Paik, C., Perez-Bacete,
van Osdol, W., and Weinstein,
J. N. Micropharmacology
antibodies
site barrier.
antibodies
1990.
S. K., Berk, D. A., Papahadjopoulos,
permeability
and interstitial
penetra-
(stealth)
3352-3356,
J. A., Larson,
of monoclonal
to delivery
in tumors.
29.
cology
1987.
Bioconjugate
1986.
Jam,
and other macromolecules
30.
derivatization.
1994.
to direct selective
113-130,
28.
its
5: 321-326,
J., Carrasquillo,
1993.
12. Glockshuber,
R., Mahia,
comparison
of strategies
to
Biochemistry,
29: 1362-1367,
conserved
containing
C. R., and Pastan,
Chem.,
Brown, J., and Siegall, C. BR96 sFv-PE3O,
a potent single-chain
immunotoxin
that selectively
kills carcinoma
cells. Cancer Res., 53: 334339,
Cancer
1994.
Reiter,
Y., Kreitman,
antitumor
activity
that
exotoxin.
1989.
10. Wels,
W., Harwcrth,
I., Mueller,
M., Groner,
B., and Hynes,
N.
Selective
inhibition
of tumor cell growth
by a recombinant
single-chain
antibody-toxin
specific
for the erbB-2
receptor.
Cancer
Res., 52: 63106317, 1992.
11.
Fv fragment.
I.
26.
Junghans,
FitzGerald,
D. J., and Pastan, I. A recombinant
immunotoxin
of two antibody
variable
domains
fused
to Pseudomonas
Nature
immunotoxins
cer Res., 51: 3781-3787,
5. Huston,
J. S., Levinson,
D., Mudgett-Hunter,
M., Tai, M. S., Novotny, J., Margohies,
M. N., Ridge, R. J., Bruccoleri,
E. H., Crea, R., and
Oppermann,
H. Protein
engineering
of antibody
binding
sites: recovery
of specific
activity
in an anti-dogoxigenin
single-chain
Fv analogue
produced
in Escherichis
coli. Proc. Natl. Acad.
Sci. USA,
85: 58795883,
1988.
into
U.,
Fv fragment.
21.
M.
242:
Wang,
Q. C., and Pastan,
in mice
of a recombinant
1994.
Brinkmann,
a recombinant
Bio-
expression
and characterization
of
monoclonal
antibodies
B1 and
Protein Eng., 7: 1509-1515,
1994.
immunotoxins.
R. E., Hardman,
B. M., Lee, S.-M.,
Y.,
engineered
5451-5459,
7: 697-704,
King,
U., and Pastan,
I. Immunotoxins
Acta,
1/98:
27-45,
1994.
U.,
a disulfide-stabihized
disulfide-stabihized
anti-Tac
exotoxin.
Int. J. Cancer,
58:
FitzGerald,
D. Recombinant
Ann. N. Y. Acad.
Sci., 758:
B3.
I. Engineering
interchain
disulfide bonds into conserved
frameregions of Fv fragments:
improved
biochemical
characteristics
of
Eng.,
18.
and
REFERENCES
disuhfide
antibody
1994.
bonds
33:
recombinant
and
of interchain
of the Fv of the monoclonal
L. H., Brinkmann,
and pharmacokinetics
containing
disulfide
work
assistance,
B. Design
1029
1994.
Y., Pai,
activity
Biochemistry,
Pastan,
ACKNOWLEDGMENTS
Lee,
Research
16. Reiter, Y., Brinkmann,
U., Kreitman,
R. J., Jung, S. H., Lee, B., and
Pastan, I. Stabilization
of the Fv fragments
in recombinant
immunotoxins
the
evident.
We thank F. Lovehace and A. Harris for cell culture
A. Jackson and J. Evans for editorial assistance.
I., and
region
Res., 54: 2714-2718,
the
whether
Pastan,
in the framework
immunotoxin
was
in which
and the distances
larger
molecules
tumors
S-H.,
Proteins,
B1(dsFv)PE38
larger
Jung,
bonds
mAb
similar
tissues
as in humans.
In the current
experiments,
that were
4-5
mm in diameter
were
used in the
14.
Fv
of Bi
Cancer
tumors:
direct
experimental
evidence
Res., 52: 5144-5153,
1992.
507:
199-210,
M. J., Sato, J.,
of monoclonal
for
a binding