Download Polyethylene Glycol-modified Chimeric Toxin

[CANCER RESEARCH53, 4588-4594, October 1,1993]
Polyethylene Glycol-modified Chimeric Toxin Composed of Transforming Growth
Factor a and Pseudomonas Exotoxin
Qing-cheng Wang, Lee H. Pai, Waldemar
D e b i n s k i , 1 D a v i d J. F i t z G e r a l d , a n d Ira P a s t a n 2
Laboratory of Molecular Biology, Division of Cancer Biology, Diagnosis and Centers, National Cancer Institute, NIH, Bethesda, Maryland 20892
ABSTRACT
Modification of proteins with monomethoxy-polyethylene glycol
(mPEG) has been shown to prolong circulation time and to reduce immunogenicity. To make a mPEG-modified recombinant toxin that retained
cytotoxic activity but had a longer residence time in circulation, we have
constructed an altered form of TGFa-PFA0, a recombinant toxin composed of human transforming growth factor a (TGFa) fused to a fragment
ofPseudomonas exotoxin (PE38) devoid of its cell- binding domain. In the
newly designed protein, termed TGFaR29-L2-CH2-PE38QQA (TCP),
there are no lysine residues in the TGFa and PE38 portions. Human IgG4
constant region CH2 and a tetradecapeptide linker, L2, are inserted between TGFa and PE38. Together, 1.,2 and CH2 contain 13 iysine residues
as potential modification sites for mPEG. mPEG conjugates of TCP (PEGTCP) were generated and the products were resolved by ion exchange
chromatography. Two PEG-TCP species termed B4 and B6 retained 15
and 4% of cytotoxicity, respectively, and 26% of their receptor binding
activity compared with the unmodified TCP. Both B4 and B6 had prolonged circulation times in the blood and reduced toxicity in animals. The
mean residence times of B4 and B6 were 37 and 68 min, respectively,
compared to 7 min for TCP. When administered i.v. to tumor bearing
mice, both B4 and B6 produced marked antitumor effects whereas the
unmodified TCP had none. Also, the immunogenicity of PEG-TCP was
5--10 times less than that of TCP. We suggest that the prolonged circulating
time and reduced toxicity of PEG-TCP compensate for a diminished cytotoxic activity and enlarge significantly the therapeutic window of this
chimeric toxin.
INTRODUCTION
Recently, a number of recombinant toxins have been developed
using gene splicing DNA techniques (1-3). In our laboratory, we have
produced recombinant toxins containing growth factors, single chain
antigen binding proteins and CD4 to mutant forms of PE 3 (4-10).
These chimeric toxins, which are very cytotoxic to target cells, can be
produced in large amounts and are of uniform composition. Moreover,
their properties can be modified by mutating the genes that encode
them. One feature of recombinant toxins is that they have a lower
molecular weight than conventional antibody-toxin conjugates (Mr
65,000 versus about Mr 195,000). This leads to a shorter survival in
the circulation but may also facilitate penetration into tumors. In
addition, these chimeric toxins may not be as compactly folded as
natural proteins and hence be more susceptible to proteolysis. Recombinant toxins have been found to produce complete regression of some
tumors and partial regression of others when tested in nude mice
Received 3/19/93; accepted 7/27/93.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance with
18 U.S.C. Section 1734 solely to indicate this fact.
1 W. D. received a postdoctoral fellowship from the Medical Research Council of
Canada and NIH CRADA. Present address: Laboratory of Molecular Targeting, Research
Centre, HDM-UM, 3850 St. Urbain St., Montreal, Quebec H2W 1T8, Canada.
2To whom requests for reprints should be addressed, at Laboratory of Molecular
Biology, National Cancer Institute, NIH, 9000 Rockville Pike, Bldg. 37, Rm. 4E16,
Bethesda, MD 20892.
3 The abbreviations used are: PE, Pseudomonas exotoxin; mPEG, monomethoxypolyethylene glycol; TGFa, transforming growth factor or; TCP, TGFctR29-L2-Crt2PE38QQA; PEG-TCP, mPEG conjugates of TCP; B1-B6, PEG-TCP species; EGF, epidermal growth factor; NPC, p-nitrophenyl carbonate; ELISA, enzyme-linked
immunoabsorbent assay; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; LDso, dose of toxin that kills 50% of animals in 10 days or less.
bearing human xenografts (9, 11, 12). There is some evidence that
efficacy is enhanced if recombinant toxins survive longer in the bloodstream (12).
One way to increase the lifetime of proteins in the blood is to
modify them with mPEG. This agent has been used extensively to
modify various enzymes such as adenosine deaminase, asparaginase,
superoxide dismutase, monoamine oxidase, immunoglobulins, interleukin 2, and also some allergens (13-22). The mPEG-modified proteins usually have longer circulating lives and reduced immunogenicity compared to unmodified proteins. In some cases, they are also
more resistant to proteolytic degradation (23). Katre et al. (19) and
Zimmerman et al. (20) have reported that mPEG modification of
interleukin 2 increased its potency against a Meth A sarcoma. Kitamura et al. (21) have reported improved tumor localization with a
mPEG-modified F(ab')2 derived from the monoclonal antibody A7.
One common feature of the proteins that have been successfully
modified by mPEG is that they have a single functional domain. In
contrast, recombinant toxins have several functional domains. Therefore, our goal was to design a recombinant toxin that could be modified by mPEG yet retain all of its functions.
PE consists of three major domains (24, 25). Domain Ia (amino
acids 1-252) is responsible for cell binding. Domain II (amino acids
253-364) mediates the translocation of the ADP-ribosylating portion
of PE into the cytosol. A subdomain termed Ib (amino acids 365-399)
has no known function and most of domain Ib can be deleted without
loss of activity (26). Domain III (amino acids 400-613) catalyzes
ADP-ribosylation of eukaryotic elongation factor 2. PE38 consists of
amino acids 252-613 of PE with a deletion of amino acids 365-380.
TGFa-PE38 is a chimeric toxin in which TGFo~ is fused to PE38. This
chimeric toxin binds to EGF receptors that are expressed in large
amounts on some types of tumor cells (27-30). Because PE38 retains
the ADP-ribosylation and translocation functions of PE, the chimeric
toxin is extremely cytotoxic but only to the target cells. To prepare a
form of TGFa-PE38 that can be modified by mPEG, we altered all of
the lysine residues in TGFoL and PE38 by mutations to other amino
acids. We then inserted the small linking peptide L2, containing 2
lysine residues, and a human CH2 domain of IgG4, consisting of 111
amino acids and containing 11 lysine residues (31), between TGFo~
and PE38 (Fig. 1). The novel mutant protein TGFo~R29-L2-CH2PE38QQA is abbreviated to TCP. We anticipated that mPEG would
modify primarily the lysine residues in the linker and the CH2 region
because the reactive group p-nitrophenyl carbonate reacts with primary amino groups in proteins. This paper reports on the properties of
mPEG-modified TCP.
MATERIALS AND
METHODS
Materials. Restriction endonucleases and alkaline phosphatase were obtained from Bethesda Research Laboratories (Gaithersburg, MD), United
States Biochemical (Cleveland, OH), or New England BioLabs (Beverly, MA)
and used under conditions recommended by the suppliers. T4 DNA ligase was
purchased from Boehringer Mannheim (Indianapolis, IN). mPEG-p-nitrophenyl carbonate was obtained from Sigma Chemical Co. (St. Louis, MO). [3H]leucine was obtained from Amersham (Arlington Heights, IL). 12SI-EGFwas
4588
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1993 American Association for Cancer
Research.
RECOMBINANT
TOXINS
Linker 2: 5'-CGGACCTCCTGGCTGAAGGTAAATCqTCAGGCTCTGGqTCAGAATCTAAATCAACTCA-3 '
"ie' T'
Ndel
Pstl
d
'~176
pWDI"l
166C -
3 '-GTACGCCTGGAGGACCGACTTCCATITAGAAGTCCGAGACCAAGTCITAGATITAGqTGAG'ITCGA-5 '
b
L1
TG FaR 29
d
-
pWO 1661
--i~
d
bSphl
Sac.
" ~
Hindlll
~ ~ ~ -
~ ~
-
,a
Hindlll
)
Sa~ll
Ecoil
PE38QQA
782 bp 1
-
/
b
pQW 4
s"
Hindlll
Hindlll
i
!
,,,~ ~"
d ''
I
RI
t
,!, ~ J
~D~
~"
b
Fig. 1. Scheme of construction of plasmid pQW 5 fusion protein and TCP, encoding.
The intermediate plasmids pQW 1 and 2 are omitted in this scheme, bp, base pairs.
purchased from NEN Research Products (Boston, MA). Oligonucleotide
linkers were synthesized on an Applied Biosystems DNA synthesizer (Foster,
CA) at the National Cancer Institute.
Cell Line. A431 cell line was obtained from the American Type Culture
Collection (Rockville, MD).
Bacterial Strains and Plasmid Construction. All the plasmid constructions were carried out using standard cloning techniques. Competent Escherichia coli cells of the HB101 or DH5a strains, purchased from Bethesda
Research Laboratories, were used for transformation and isolation of plasmids.
BL21 (ADE3) cells which carry a T7 RNA polymerase gene in a lysogenic and
inducible form were used as hosts for the expression of recombinant proteins
(32).
The DNA encoding TGFa-PE38 was modified in several steps. The goal
was to generate a form of the recombinant toxin that lacked lysine residues
except in a newly mutated linker region between TGFc~ and PE38. We began
with a gene encoding TGFa in which lysine 29 in TGFa had been changed to
arginine (a kind gift from Dr. Alan Oliff, Merck Sharp & Dohme Research
Laboratories, West Point, PA). Plasmid pWD1661 was formed in a three
fragment ligation (Fig. 1). A 140-base pair fragment (encoding most of the
TGFaR 29) from a plasmid containing the mutant TGFc~ digested with PstI and
SphI restriction enzymes was ligated to a vector originating from plasmid
pWD166C (encoding TGFa-PE38) cut with NdeI and SphI and joined to linker
1, which had NdeI and PstI cohesive ends. Plasmid pWD1661 encodes a
protein which has the amino acid Leu-Ala-Ala-Ala preceding wild type
TGFc~. These four amino acids were eliminated in subsequent subcloning
steps. Next, plasmid pQW3 was produced by the ligation of a 4317-base pair
fragment obtained after digestion of plasmid pWD1661 with SphI and HindlII
to linker 2, containing SphI and HindlII cohesive ends and encoding a tetradecapeptide, L2. pQW4 was created by ligation of a 782-base pair fragment
digested with SaclI and EcoRI from pJVlys4Qdel-l,2 encoding pEQ59~176 613
(PEQQA) to a 4367-base pair fragment originating from pQW3 digested with
the same enzymes. The protein encoded by pQW4 possesses two lysine residues exclusively within the sequence of L2.
Plasmid pQW5 encodes TCP. It was constructed by inserting the CH2 domain
of a human IgG 4 constant region between TGFaR29-L2 and PE38QQA. The
fragment encoding CH2 was cleaved by HindlII digestion of the plasmid
pB387CH2 encoding TGFa-CH2-PE384 and ligated to the plasmid pQW4
nicked with HindlII.
Purification ofTCP. TCP was localized to the inclusion bodies, denatured,
renatured, and purified using the same protocol as for the purification of
B3Fv-PE38 described by Buchner et al. (33).
Preparation of mPEG-modified TCP (PEG-TCP). TCP was reacted with
30-fold molar excess of mPEG p-nitrophenyl carbonate in 0.2 M phosphate
buffer (pH 8.0) containing 1 mM EDTA at 4~ overnight. The reaction mixture
was dialyzed twice against 20 mM Tris-HC1 (pH 7.4) containing 1 mM EDTA
(buffer A) at 4~ The crude PEG-TCP was separated from other reactants by
sequential chromatography on Q-Sepharose and Mono Q with a NaCI gradient
(0-0.3 M) in buffer A. Uncoupled mPEG derivatives did not bind to the
Q-Sepharose column. Unmodified TCP was eluted last from both columns and
was mPEG negative as measured by BaC12-I2 reagent (34, 35).
The SDS-PAGE profile of the crude PEG-TCP showed that there were six
major bands in the gel. They were referred to as B1-B6, respectively, according to their apparent molecular weights from low to high. The fractions eluted
from Mono Q containing a given species of PEG-TCP were pooled and
enriched by the rechromatography on Mono Q.
Protein Assay. The protein concentration of TCP and PEG-TCP species
was determined using Pierce Coomassie Plus Protein Assay Reagent. Bovine
serum albumin was used as the standard protein.
Cytotoxicity and Binding Assays. A431 cells expressing more than 2 •
106 receptors/cell were used as target cells. The cytotoxicity of TCP and
PEG-TCP was measured as described previously (25). The binding assay was
based on the competition of 125I-EGE The assay was carried out basically as
the experiment described by Kreitman et al. (36).
LDso in Mice. TCP and PEG-TCP species of different molecular weights
were diluted in 0.2% human serum albumin-phosphate-buffered saline and
BALB/c mice (6 weeks old, 20-22 g) received injections into the tail veins of
0.2-ml aliquots of various doses. Animals were observed for 10 days for signs
of toxicity or death.
Pharmacokineties of TCP and PEG-TCP in Mice. BALB/c mice (6
weeks old, 20 g) were given a single dose i.v. of TCP (250/xg/kg) or PEG-TCP
species (500/zg/kg). Blood samples were collected at 2, 10, 30, 45, 60, 90, 120,
180, and 240 min. Each time point represents the mean results obtained from
three animals. The concentration of the toxin was determined by incubating the
serum with A431 cells and by measuring its ability to inhibit protein synthesis.
A standard curve was used to determine the concentration of toxin in each
sample.
Antitumor Activity of TCP and PEG-TCP in Nude Mice Bearing a
Human Epidermoid Carcinoma. A431 cells (3 • 106 cells) were injected
s.c. on day 0 into female athymic nude mice (4-6 weeks old, 18-20 g). Tumors
measuring about 5 x 5 mm developed in all animals by day 5. Treatment with
TCP and PEG-TCP was started on day 5 after tumor implantation. Each
treatment group consisted of five or six animals. Tumors were measured with
a caliper every second day and the volume of the tumor was calculated as
described (37).
Detection of Anti-PE Antibody. An ELISA was used to determine the
presence of mouse anti-PE antibodies in the serum of treated and control
animals. Microtiter 96-well plates (Dynatech Laboratories, Alexandria, VA)
were coated with PE38QQA (50 ng/well), incubated overnight at 4~ and
thoroughly washed. The ELISA was carried out as described previously (38).
Chimeric Toxin Neutralization by Serum Containing Anti-PE Antibodies. Serum from animals immunized with TCP and PEG-TCP and control
serum were diluted with TCP at a final concentration of 0.1, 1, 10, 100, and
1000 ng/ml and then incubated for 30 min at 37~ prior to the protein synthesis
assay.
Linker 1: 5'-TATGCTGGCAGCTGCA-3'
3'-ACGACCGTCG-5'
4 j. Batra and I. Pastan, unpublished work.
4589
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1993 American Association for Cancer
Research.
RECOMBINANT TOXINS
RESULTS
[ ~
Preparation of TCP and P E G - T C E To prepare a chimeric toxin
that could be derivatized with mPEG in such a manner that cell
binding, translocation, and ADP-ribosylation activities would be retained, we engineered an altered form of TGFc~-PE38 in which potential reactive lysine residues were located only in positions not
involved with any of these three essential processes. To this end, the
single lysine residue in TGFot was converted to an arginine
(TGFo~R29), the two lysine residues within PE38 at positions 590 and
606 were converted to glutamines, and lysine 613 was deleted
(PE38QQA). In addition, a lysine-rich domain was placed between
TGFot and PE38. This lysine-rich domain consists of a CH2 domain
derived from a human IgG4 constant region preceded by a small
peptide containing two lysine residues (L2). The construction of the
plasmid used to express TCP is shown in Fig. 1. The chimeric toxin
accumulated in E. coli within inclusion bodies and was purified as
previously described (33). This molecule is designed so that, despite
derivatization with mPEG, it could be normally processed within
target cells and could release a Mr 37,000 fragment that has no mPEG
attached to it.
The derivatization procedure with polyethylene glycol was carried
out as described in "Materials and Methods." Analysis of the derivatized products by SDS-PAGE showed that six major species with
different molecular weights were present and referred to as B1-B6,
respectively. To separate these species, the mixture of derivatized
molecules was applied on a Mono Q column and eluted with a gradient of NaCI as shown in Fig. 2. It is evident that the higher molecular weight proteins eluted from the column before the lower
molecular weight species. However, this single chromatographic step
did not adequately resolve species of different molecular weights. To
obtain better separation, selected column fractions were pooled and
rechromatographed on Mono Q columns. The composition of the
repurified mPEG conjugates is shown in Fig. 3. The apparent mo-
0.20.
S
SS
i ,
i
SS
B1
PEG-TCP ~ ]
B2
B3
B4
B5
KDa
B6
~,~,.
-97
I
-36
Fig. 3. Reducing SDS-PAGE analysis (8% gel) for PEG-TCP species B1-B6 which
have been repurified, respectively,by Mono Q.
Table 1 Apparent molecular weight, cytotoxicity, and binding abilities of TCP and
PEG-TCP
The molecularweight was estimated on the basis of protein mobility on the reducing
SDS-PAGE shown in Fig. 3.
TCP
B1
Apparent molecular
72
72
weight (thousands)
IC5o" (ng/ml)
1.3
1.3
50% binding (mM)
21
12
a IC5o, 50% of inhibiting concentration.
B2
B3
81
87
3.5
44
3.1
44
B4
B5
B6
103
122
135
20
100
30
80
8.5
80
lecular weights of B1-B6 are 72,000, 81,000, 87,000, 103,000,
122,000, and 135,000, respectively.
Cytotoxicity a n d Binding Activity of P E G - T C P . Each fraction
(B1-B6) was tested for its ability to inhibit protein synthesis using
A431 cells and for its ability to compete for the binding of a25I-EGF
to the EGF receptor on A431 cells. The data are summarized in Table
1 and Fig. 4. Generally, we found that as the molecular weight of
PEG-TCP increased, reflecting the degree of derivatization by PEG,
the binding activity and cytotoxic activity decreased. Fractions B2 and
200
200
9 CYTOTOXICITY
[ ] BINDING
I
ssSS/~
$sSS / I
A
,,..,.,.
L~
SI
SS
SS
S
$
S
SS
TCP
m
SSSS/
L~
Z
O.lO-
S~SSSS
SSSSS$SS
1
/
/
S$SJ
0.00
0
I ~
10
i
20
Fraction
~1
30
4O
50
numbers
Fig. 2. Profile of Mono Q chromatography.About 14 mg of pooled mPEG conjugates
of TCP which were eluted from Q-Sepharosewere loaded on a Mono Q column (8 ml).
The columnwas equilibratedwith 20 mMTris-HCl (pH 7.4) containing 1 mMEDTAand
eluted with a linear NaCI gradient(0-0.3 M)ifi the same buffer.The protein concentration
was assayed by CoomassiePlus Reagent(Pierce).A595nmrepresentsthe protein concentration.
O
~
TCP
B1
I
0
B2
B3
B4
PEG TCP
B5
B6
I
Fig. 4. Comparison of the cytotoxicity and binding ability of TCP and PEG-TCP
species B1-B6. This figure is based on the data listed in Table 1.
4590
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1993 American Association for Cancer
Research.
RECOMBINANT TOXINS
B3 have similar cytotoxicities and similar binding properties, even
though they differ in molecular weight. The higher molecular weight
compounds, B4-B6, retained between 15 and 4% of their cytotoxic
and binding activities.
Pharmacokinetics of TCP and PEG-TCP in Mice. Pharmacokinetics was determined after a single i.v. bolus administration of TCP
and three different species of PEG-TCP. Table 2 shows the pharmacokinetic data of TCP and the PEG-TCP species B2, B4, and B6.
Pharmacokinetic parameters were calculated using the exponential
curve fitting program RSTRIP (MicroMath Scientific Software, Salt
Lake City, UT). Each time point represents the mean from 2-3 animals. The mean residence time increased from 7 min for the unmodified molecule to 67 min for the most highly derivatized molecules
(B6). Clearance rates changed accordingly. The major changes were
evident in the tl/213 of the clearance curves (Table 2). Fig. 5 shows the
relative serum concentrations of chimeric toxins as a percentage of
TCP concentration at time 0. High molecular weight mPEG derivatized TCP molecules (B4 and B6) had a much longer survival in the
plasma than less modified (B2) or unmodified TCP.
Animal Toxicity Studies. To determine the amount of TCP that
could be administered to animals for therapeutic studies, increasing
amounts of the parental molecules and the modified mPEG molecules
were injected into mice and the LD5o was obtained by noting the doses
that caused death. As shown in Table 3, a significant increase in LDso
was observed upon "PEGylation." The unmodified TCP had an LDso
of approximately 50/.Lg/kg, whereas the most derivatized mPEG-TCP
had an LDso of 400 p~g/kg.
Antitumor Activity. The data in Fig. 6 compare the antitumor
activity of PEGylated molecules with that of unmodified TCP. In these
experiments, A431 cells (3 • 106 cells) were injected s.c. into athymic
mice on day 0. Therapy with TCP or PEG-TCP was delivered on days
4, 5, and 6, when the tumor reached 0.5 x 0.5 cm. Animals received
various doses of TCP (1.25, 2.5, 5, or 10/~g/kg/day for 3 doses) or
PEG-TCP species B2, B4, or B6 (50, 100, 150, 200, or 250/~g/kg/day
for 3 doses). The control group received the phosphate buffer solution.
Fig. 6 shows the effect of TCP and B4, B5, and B6 when given at the
maximum tolerated doses (the highest dose which does not produce
weight loss or death). The maximum tolerated dose for TCP is 2.5
/xg/kg/day for 3 doses; for B4, B5, and B6 maximum tolerated doses
are 100, 150, and 200/xg/kg/day for 3 doses, respectively. As shown
in Fig. 6, high molecular weight derivatized chimeric toxin molecules
(B5 and B6) are better tolerated and are capable of causing tumor
regression in nude mice when given by i.v. bolus injection. In contrast,
no antitumor effect was noted with the unmodified molecule, even
near or at the MTD.
Detection of Anti-PE Antibodies. Other researchers have demonstrated that derivatization of proteins, such as bovine serum albumin
and liver catalase, with mPEG makes these molecules less immunogenic to animals (23, 39). To study the immunogenicity of TCP and
PEG-TCP, three experiments were performed. First, we administered
similar doses of TCP and conjugates B4 and B6 to BD2 F1 mice (6-8
weeks old) and compared the antibody response to that of PE. Animals
received daily i.p. injections of 2.5 or 5/zg/kg of TCP or PEG-TCP
100
A
.<
r
10
.1
Protein
TCP
PEG-TCP (B2)
PEG-TCP (B4)
PEG-TCP (B6)
Clearance
(ml/min/kg)
t~a
t~[3
(min)
(min)
6.6
37.7
37.1
67.6
19.1
16.2
4.0
1.3
1.2
3.3
1.5
1.7
17.7
76.1
39.5
80.0
30
60
90
120
Time
150
180
210
i
240
(minutes)
Fig. 5. Pharmacokinetic profile of PEG-TCP species B2, B4, and B6 in mice. The toxin
concentration was determined as described in "Materials and Methods." The relative
serum concentration is exhibited as percentage of the toxin concentration at time 0. ( 9
TCP, (0) B2, (0) B4, (El) B6.
species B4 or B6 for 7 days. Blood samples were collected on day 10
and mouse anti-PE antibody levels were determined by ELISA. At 2.5
/xg/kg/day, no antibody was detected in the animals that received B4
or B6 of PEG-TCP. At 5 ~g/kg/day, anti-PE was present in the TCP
group (Fig. 7) but not in the animals that received either B4 (not
shown) or B6. We next administered higher doses of B6 conjugate (25,
50, or 100/xg/kg/day for 7 doses) and compared its immunogenicity
with TCP, 5 /xg/kg/day for 7 doses. Higher doses of chimeric toxins
Table 3 LDso of TCP and PEG-TCP species B2, B4, and B6 in mice
Protein
LDso (mg/kg)
TCP
PEG-TCP (B2)
PEG-TCP (B4)
PEG-TCP (B6)
50
75
200
400
0.8
S
I--
i
Table 2 Serum pharmacokinetics of TCP and PEG-TCP species B2, B4 and B6 in mice
The calculations of MRT, clearance, t~a, and t~/3 were based on the data shown in Fig.
5. Pharmacokinetic parameters were calculated using the exponential curve fitting program RSTRIP (MicroMath Scientific Software, Salt Lake City, UT).
MRT
(min)
I
0
0
2
4
Time
6
8
10
12
(days)
Fig. 6. Effect of TCP and PEG-TCP on the growth of A431 tumors in nude mice. Mice
received injections of 3 • 106 A431 cells on day 0 and were treated i.v. on days 4, 5, and
6 with TCP (0) and PEG-TCP species B4 (11), B5 (1~), and B6 (&). Control group received
PBS (0).
4591
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1993 American Association for Cancer
Research.
RECOMBINANT TOXINS
0.4
0.3
m
o
0.2
a
o
0.1
0.0
TCP
5 ug
B6
5 ug
B6
25 ug
B6
50 ug
B6
100 ug
Fig. 7. Histogram of mouse IgG + IgM response to PE following 7 daily doses of TCP
(5 ~g,/kg/day) and PEG-TCP (B6) (5, 25, 50, and 100/xg/kg/day). Antibody response was
measured on day 10 by ELISA. Results are reported as absorbance (OD) values of serum
specimens diluted 1:100.Bars, SD.
could not be administered due to toxicity. Fig. 7 shows a histogram of
mouse immunoglobulin response following immunization. Native PE
was used to detect the antibody response. At 5 and 25 ~g/kg, the
mPEG derivatized TCP was weakly immunogenic. At higher doses,
however, the mPEG-conjugated protein was as immunogenic as TCP.
To determine if the antibody produced in response to mPEG-TCP
had neutralizing activity, serum from animals immunized with TCP
and PEG-TCP and control serum were mixed with TCP at a final
concentration of 0.1, 1, 10, 100, and 1000 ng/ml; incubated for 30 min
at 37~ then tested for inhibition of protein synthesis. As shown in
Fig. 8, sera from animals immunized with 5 ~g of TCP or 50 and 100
p,g of B6 were capable of neutralizing up to 10 ng/ml of TCP, while
serum from animals immunized with 5 (not shown) and 25 ~g/kg of
B6 were not.
DISCUSSION
The lifetime of the B2 conjugate in the plasma was only slightly
prolonged, whereas B4 and B6 survived much longer than the parent
protein. At the 90-rain point, 4.5% of B4 and 13% of B6 were left in
the circulation while only 0.13% of the parent protein was present
(Fig. 5). In addition, the toxicity to mice of B4 and B6 was 4-8 times
lower than that of the parent protein. The extended plasma half-life
and the reduced toxicity of B4 and B6 conjugates in mice were enough
to compensate for their reduced cytotoxic activity to target cells and
made these molecules more therapeutically effective. Kitamura et al.
(21, 40) found that PEGylated F(ab')2 of the monoclonal antibody A7
had less uptake in the liver and kidney and had a higher tumor:normal
tissue ratio when compared with the parent F(ab')2, thus achieving a
more specific localization of the F(ab')2. Also, less uptake of PEGylated tissue plasminogen activator by the liver was found by
Bergeret et al. (41). Thus, we speculate that in addition to the reduction in cytotoxicity, the reduction of the animal toxicity of PEG-TCP
may partly be due to less liver uptake of PEG-TCE The improved
antitumor effect of PEG-TCP may be attributed to a higher tumor:normal tissue ratio of PEG-TCP and a higher tolerance of the mice to
it.
The results shown in Figs. 7 and 8 show that the more heavily
PEGylated TCP has lower immunogenicity in comparison with the
less PEGylated material. It is plausible, as hypothesized by Sehon
(16), that down-regulation of the primary antibody response was due
in part to activation of antigen-specific suppressor T-cells (16, 42). It
is also possible that PEGylation of antigens influences their degradation in macrophages or that the PEG chains shield some epitopes of
the peptides derived from the antigens after the degradation. We
believe it is likely that some portion of PE is released by proteolysis
in vivo, thereby eliciting an immune response. Therefore, it will be
necessary in the future to couple PEG chains to appropriate sites on
PE to reduce the immunogenicity of PE-containing chimeric toxins
further.
A number of enzymes and immunoglobulins have been highly
modified by mPEG. For instance, adenosine deaminase coupled with
dozens of flexible mPEG chains still retains about 60% of its original
enzyme activity (43). Because of its relatively large size, mPEG
couples only to residues on the surface of protein molecules and
usually does not block the active site of enzymes which are usually
buried deeper within the protein structure. Because the mPEG-masked
surface remains permeable to low molecular weight substrates, the
enzymes can continue to operate despite extensive modifications.
In this study, we have constructed an altered form of TGFa-PE38
100
that can be derivatized with polyethylene glycol yet still maintains its
i
cytotoxic activity. We found that PEGylated conjugates, with increasing amounts of PEGylation (B2, B4, and B6), were active and had
80
extended lifetimes in the blood. Moreover, B4 and B6 exhibited
markedly enhanced antitumor activities when compared to unmodified chimeric toxin. Thus, we achieved our goal of preparing an active
PEGylated chimeric toxin that had a prolonged lifetime in the blood
g
and enhanced antitumor activity.
We observed that the chimeric toxin with the lowest amount of
PEGylation, such as the B2 or B3 conjugate, had higher cytotoxicity
and cell-binding ability than molecules with more polyethylene glycol
chains added, i.e., B4 and B6 conjugates. However, the decrease in
20
cytotoxicity with increasing derivatization was not parallel to the
observed decrease in cell binding. For example, B4, B5, and B6
PEG-TCPs have quite similar binding activities, yet B6, which has the
0
.01
,1
1
1 0
1 O0
.001
highest molecular mass of 135 kDa, retains only 4% of its original
cytotoxicity. These results suggest that the decrease in binding to the
ng/ml
EGF receptors is not the sole reason for the reduction in PEG-TCP
Fig. 8. Serum from mice immunized with multiple doses of TCP, 5 ~g/kg/day (ll), and
cytotoxicity. The mPEG chains probably interfere with some other PEG-TCP (B6), 25/zg (O), 50/xg (~), and 100/xg/kg/day (&) and control sera (9 were
steps in toxin function.
incubated with TCP and tested for neutralizing activity in A431 cells.
4592
q
. . . . . . .
J
. . . . . . . .
i
. . . . . . .
,!
. . . . . . . .
i
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1993 American Association for Cancer
Research.
. . . . . . .
RECOMBINANT TOXINS
However, the same scenario is not applicable to chimeric toxins. In the
case of TGFce-PE38, its cytotoxicity is dependent on several different
interactions. For example, TGFo~ must bind to its receptor, and domain
III of PE must interact with NAD + and elongation factor 2. In addition, there must be a proteolytic cleavage between arginine 279 and
glycine 280 that generates a Mr 37,000 peptide fragment that must be
translocated across a membrane into the cytosol. Therefore, we had to
prepare a recombinant toxin that would still undergo these interactions
after PEGylation. We used TGFotR 29 because Defeo-Jones et al. (44)
reported that a conservative substitution of arginine for lysine 29 of
human TGFo~ had little effect on cell-binding or mitogenesis.
Chaudhary et aL (45) found that deletion of Lys613 from PE and the
substitution of glutamine for Lys590 and Lys606 had no effect on the
cytotoxicity of PE. Also, Debinski and Pastan5 found that PE40QQA
retained full activities of PE40. Finally, Kasturi et al. 6 found that a
Ca2 domain of IgG4 can be inserted in CD4-PE40 between the CD4
and FE40 domains without loss of cytotoxic activity. On the basis of
these observations, we designed TCP, expecting that it could be PEGylated and would retain its cytotoxicity. The mPEG chains should
couple mainly to the L2 and CIa2 regions of TCP, although the NH2terminal methionine residue was available for PEGylation. Using this
strategy, we produced a series of active PEGylated conjugates of TCP.
We found that PEGylation often resulted in a considerable loss in
cytotoxicity, as discussed above. To explain this, we suggest that the
unfolded, flexible mPEG chain that has a backbone of about 150
atoms may sterically interfere with TGFe~ binding or with functions of
domains II or III of PE38. The NH2 terminus of TGFot is exposed on
the surface of the molecule (46). Therefore, we cannot exclude the
possibility that a mPEG chain may also be coupled to the NH2 terminus and in this way interfere with the binding of TGFo~ to its
receptor.
mPEG has only one hydroxyl group in each polymer chain and this
can be transformed to an activated form. Three major reactive derivatives of mPEG have been studied. These are cyanuric chloride activated mPEG (39), mPEG-succinimidyl succinate (47), and mPEGNPC (48). The cyanuric chloride derivative is toxic and may also react
with amino acid residues other than lysine (47). The ester linkage
between mPEG and the succinic acid residue is labile in aqueous
solution (22, 49). The reaction of mPEG-NPC with proteins yields
urethan-linked proteins that are stable under physiological conditions
(48). Therefore, we chose to use mPEG-NPC to modify our chimeric
toxins. Because the lysine residues of TCP were modified by mPEGNPC, the resulting conjugates were more acidic than the parent protein. Yet contrary to expectations, the less PEGylated toxins eluted
later from the Mono Q column than the heavily PEGylated molecules,
and the unmodified toxin eluted last. To account for this behavior, we
suggest that the mPEG chains may shield the charges of the TCP
molecule or hamper the binding of the negatively charged functional
groups of the molecule to the Mono Q column. Similar phenomena
have been noted by Davis et aL (50) and Somack et al. (51).
In the presence of 2-mercaptoethanol, the ester linkage between
PEG and p-nitrophenyl carbonate is not completely stable at high
temperature. Therefore, some PEG chains may be released when
heating the PEG-TCP samples for SDS-PAGE, generating lower molecular weight PEG-TCP species, and even free TCP. Using nonreducing SDS-PAGE, B5 and B6 were quite pure and not contaminated
with free TCP, and B4 was contaminated with a small amount of free
TCP (data not shown). Since the pharmacokinetic data were measured
by cytotoxicity assays and unPEGylated TCP was cleared very rapidly
in vivo, we cannot attribute the prolonged half-life of PEG-TCP to the
5 W. Debinski and I. Pastan, unpublished work.
6 Kasturi, J. Batra, and 1. Pastan, unpublished work.
presence of free TCP. For a similar reason, it is very unlikely that the
enhanced antitumor effects shown by PEG-TCP are due to the contamination with free TCP because free TCP was much more toxic to
mice and ineffective at its MTD.
REFERENCES
1. Pastan, I., and FitzGerald, D. Recombinant toxins for cancer treatment. Science
(Washington DC), 254: 1173-1177, 1991.
2. Pastan, I., Chaudhary, V., and FitzGerald, D. J. Recombinant toxins as novel therapeutic agents. Annu. Rev. Biochem., 61." 331-354, 1992.
3. Strom, T. B., Anderson, P. L., Rubin-Kelley, V. E., Williams, D. P., Kijokawa, T., and
Murphy, J. R. Immunotoxins and cytokine toxin fusion proteins. Semin. Immunol., 2:
467-479, 1990.
4. Chaudhary, V. K., FitzGerald, D. J., Adhya, S., and Pastan, I. Activity of a recombinant fusion protein between transforming growth factor ~ and Pseudomonas toxin.
Proc. Natl. Acad. Sci. USA, 84: 4538--4542, 1987.
5. Siegall, C. B., Xu, Y. H., Chaudhary, V. K., Adhya, S., FitzGerald, D., and Pastan, I.
Cytotoxic activities of a fusion protein comprised of TGF a and Pseudomonas
exotoxin. FASEB J., 3: 2647-2652, 1989.
6. Batra, J. K., FitzGerald, D. J., Chaudhary, V. K., and Pastan, I. Single-chain immunotoxins directed at the human transferrin receptor containing Pseudomonas exotoxin
A or diphtheria toxin: anti-TFR(Fv)-PE40 and DT388-anti-TFR(Fv). Mol. Cell. Biol.,
11: 2200-2205, 1991.
7. Chaudhary, V. K., Queen, C., Junghans, R. P., Waldmann, T. A., FitzGerald, D. J., and
Pastan, I. A recombinant immunotoxin consisting of two antibody variable domains
fused to Pseudomonas exotoxin. Nature (Lond.), 339: 394-397, 1989.
8. Chaudhary, V. K., Batra, J. K., Gallo, M. G., Willingham, M. C., FitzGerald, D, J., and
Pastan, I. A rapid method of cloning functional variable region antibody genes in E.
coli as single chain immunotoxins. Proc. Natl. Acad. Sci. USA, 87: 1066-1070, 1990.
9. Brinkmann, U., Pal, L. H., FitzGerald, D. J., Willingham, M., and Pastan, I. B3(Fv)-PE38KDEL~ a single chain immunotoxin that causes complete regression of a human
carcinoma in mice. Proc. Natl. Acad. Sci. USA, 88: 8616-8620, 1991.
10. Chaudhary, V. K., Mizukami, T., Fuerst, T. R., FitzGerald, D. J., Moss, B., Pastan, I.,
and Berger, E. A. Selective killing of HIV-infected cells by a recombinant human
CD4-Pseudomonas exotoxin hybrid protein. Nature (Lond.), 335: 369-372, 1988.
11. Siegall, C. B., Kreitman, R. J., FitzGerald, D. J., and Pastan, I. Antitumor effects of
interleukin 6-Pseudomonas exotoxin chimeric molecules against the human hepatocellular carcinoma, PLC/PRF/5 in mice. Cancer Res., 51: 2831-2836, 1991.
12. Pal, L. H., Gallo, M. G., FitzGerald, D. J., and Pastan, I. Antitumor activity of a
transforming growth factor u-Pseudomonas exotoxin fusion protein (TGF a-PE40).
Cancer Res., 51: 2808-2812, 1991.
I3. Veronese, E M. Enzymes for human therapy: surface structure modification. Chim.
Oggi, 1-2: 53-56, 1989.
14. Abuchowsky, A., and Davis, E E Soluble polymer-enzyme adducts. In: J. W. Holcenberg and J. Roberts (eds.), Enzymes as Drugs, pp. 367-383. New York: John Wiley
& Sons, Inc., 1981.
15. Haris, J. M. Laboratory synthesis of polyethylene glycol derivatives. Rev. Macromol.
Chem. Phys., 25: 325-373, 1985.
16. Sehon, A. H. Suppression of IgE antibody responses with tolerogenic conjugates of
allergens and haptens. Prog. Allergy, 32: 161-202, 1982.
17. Hershfield, M. S., Chaffee, S., Koro-Johnson, L., Mary, A., Smith, A. A., and Short,
S. A. Use of site-directed mutagenesis to enhance the epitope-shielding effect of
covalent modification of proteins with polyethylene glycol. Proc. Natl. Acad. Sci.
USA, 88: 7185-7189, 1991.
18. Knauf, M. J., Bell, D. P., Hirtzer, P., Luo, Z. P., Young, J. D., and Katre, N.
Relationship of effective molecular size to systemic clearance in rats of recombinant
interleukin-2 chemically modified with water-soluble polymers. J. Biol. Chem., 263:
15064-15070, 1988.
19. Katre, N. V., Knauf, M. J,, and Laird, W. J. Chemical modification of recombinant
interleukin 2 by polyethylene glycol increases its potency in the murine Meth A
sarcoma model. Proc. Natl. Acad. Sci. USA, 84: 1487-1491, 1987.
20. Zimmerman, R. J., Aukerman, S. L., Katre, N. V., Winkelhake, J. L., and Young, J. D.
Schedule dependency of the antitumor activity and toxicity of polyethylene glycolmodified interleukin 2 in routine tumor models. Cancer Res., 49: 6521-6528, 1989.
21. Kitamura, K., Takahashi, T., Takashina, K., Yamaguchi, T., Noguchi, A., Tsurumi, H.,
Tojokuni, T., and Hakomori, S. Polyethylene glycol modification of the monoclonal
antibody A7 enhances its tumor localization. Biochem. Biophys. Res. Commun., 171:
1387-1394, 1990.
22. Dreborg, S., and Akerbom, E. B. Immunotherapy with monomethoxypoly-ethylene
glycol modified allergens. Crit. Rev. Ther. Drug Carrier Syst., 6: 315-365, 1990.
23. Abuchowski, A., McCoy, J. R., Palczuk, N. C., Van Es, T., and Davis, F. F. Effect of
covalent attachment of polyethylene glycol on immunogenicity and circulating life of
bovine liver catalase. J. Biol. Chem., 252: 3582-3586, 1977.
24. Allured, V. S., Collier, R. J., Carroll, S. E, and McKay, D. B. Structure of exotoxin
A of Pseudomonas aeruginosa at 3.0-Angstrom resolution. Proc. Natl. Acad. Sci.
USA, 83: 1320-1324, 1986.
25. Hwang, J., FitzGerald, D. J., Adhya, S., and Pastan, I. Functional domains of Pseudomonas exotoxin identified by deletion analysis of the gene expressed in E. coll.
Cell, 48: 129--136, 1987.
26. Siegall, C. B., Chaudhary, V. K., FitzGerald, D. J., and Pastan, I. Functional analysis
of domains II, Ib and III of Pseudomonas exotoxin. J. Biol. Chem., 264: 1425614261, 1989.
27. Velu, T. J., Beguinot, L., Vass, W. C., Willingham, M. C., Merlino, G. T., Pastan, I.,
and Lowy, D. R. EGF-dependent transformation by a human EGF receptor proto-
4593
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1993 American Association for Cancer
Research.
RECOMBINANT TOXINS
oncogene. Science (Washington DC), 238: 1408-1410, 1987.
28. Hall, W. A., Marsha, J., Merrill, M. J., Walbridge, S., and Youle, R. J. Epidermal
growth factor receptors on ependymomas and other brain tumors. J. Neurosurg., 72:
641-646, 1990.
29. Hendler, F. J., and Ozanne, B. W. Human squamous cell lung cancers express
increased epidermal growth factor receptors. J. Clin. Invest., 74: 647-651, 1984.
30. Scambia, G., Panici, P. B., Battaglia, E, Ferrandina, G., Almadori, G., Paludetti, G.,
Maurizi, M., and Mancuso, S. Receptors for epidermal growth factor and steroid
hormones in primary laryngeal tumors. Cancer (Phila.), 67: 1347-1351, 1991.
31. Ellison, J., Buxbaum, J., and Hood, J. Nucleotide sequence of a human immunoglobulin C-v4 gene. DNA, 1: 11-18, 1981.
32. Studier, F. W., and Moffat, B. A. Use of bacteriophage T7 RNA polymerase to direct
selective high-level expression of cloned genes. J. Mol. Biol., 189: 113-130, 1986.
33. Buchner, J., Pastan, I., and Brinkmann, U. A method for increasing the yield of
properly folded recombinant fusion proteins: single-chain immunotoxins from renaturation of bacterial inclusion bodies. Anal. Biochem., 205: 263-270, 1992.
34. Child, C. E. The determination of polyethylene glycol in gamma globulin solution.
Microchem. J., 20: 190-192, 1975.
35. Sims, G. E. C., and Snape, T. J. A method for the estimation of polyethylene glycol
in plasma protein fractions. Anal. Biochem., 107: 60--63, 1980.
36. Kreitman, R. J., Chaudhary, V. K., Siegall, C. B., FitzGerald, D. J., and Pastan, I.
Rational design of a chimeric toxin: an intramolecular location for the insertion of
transforming growth factor ~ within Pseudomonas exotoxin as a targeting ligand.
Bioconjugate Chem., 3: 58-62, 1992.
37. Pai, L. H., Batra, J. K., FitzGerald, D. J., Willingham, M. C., and Pastan, I. Antitumor
activities of immunotoxins made of monoclonal antibody B3 and various forms of
Pseudomonas exotoxin. Proc. Natl. Acad. Sci. USA, 88: 3358-3362, 1991.
38. Pai, L. H., FitzGerald, D. J., Tepper, M., Schacter, B., Spitalny, G., and Pastan, 1.
Inhibition of antibody response to Pseudomonas exotoxin and an immunotoxin containing Pseudomonas exotoxin by 15-deoxyspergualin in mice. Cancer Res., 50:
7750-7753, 1990.
39. Abuchowski, A., Van Es, T., Palczuk, N. C., and Davis, F. F. Alteration of immunological properties of bovine serum albumin by covalent attachment of polyethylene
glycol. J. Biol. Chem., 252: 3578-3581, 1977.
40. Kitamura, K., Takahashi, T., Yamaguchi, T., Noguchi, A., Noguchi, A., Takashina,
K-I., Tsurumi, H., Inagake, M., Toyokuni, T., and Hakomori, S-I. Chemical engineer-
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
ing of the monoclonai antibody A7 by polyethylene glycol for targeting cancer
chemotherapy. Cancer Res., 51: 4310-4315, 1991.
Berger, H., Jr., and Pizzo, S. V. Preparation of polyethylene glycol-tissue plasminogen
activator adducts that retain functional activity: characteristics and behavior in three
animal species. Blood, 71: 1641-1647, 1988.
Sehon, A. H. Suppression of antibody responses by chemically modified antigens. Int.
Arch. Allergy Appl. Immunol., 94: 11-20, 1991.
Pool, R. "Hairy enzymes" stay in the blood [news]. Science (Washington DC), 248:
305, 1990.
Defeo-Jones, D., Tai, J. Y., Wegrzyn, R. J., Vuocolo, G. A., Baker, A. E., Payne, L. S.,
Garsky, V. M., Oliff, A., and Riemen, M. W. Structure-function analysis of synthetic
and recombinant derivatives of transforming growth factor or. Mol. Cell. Biol., 8:
2999--3007, 1988.
Chaudhary, V. K., Jinno, Y., FitzGerald, D., and Pastan, I. Pseudomonas exotoxin:
contains a specific sequence at the carboxyl terminus that is required for cytotoxicity.
Proc. Natl. Acad. Sci. USA, 87." 308-312, 1990.
Harvey, I. D., Winkinson, A. J., Tappin, M. J., Cooke, R. M., and Campbell, I. D. The
solution structure of human transforming growth factor a. Eur. J. Biochem., 198:
555-562, 1991.
Abuchowski, A., Kazo, G., Verhoest, C. R., Van, E. S., Kafkewitz, D., Nucci, M. L.,
Viau, A. T., and Davis, E E Cancer therapy with chemically modified enzymes. I.
Antitumor properties of polyethylene glycol-asparaginase conjugates. Cancer Biochem. Biophys., 7: 175-186, 1984.
Veronese, F. M., LargajoUi, R., Boccu, E., Banassi, C. A., and Schivan, O. Surface
modification of proteins. Activation of monomethoxy polyethylene glycols by phenylchloroformates and modification of ribonuclease and superoxide dismutase. Appl.
Biochem. Biotechnol., 11: 141-152, 1985.
Ulbrich, K., Strohalm, J., and Kopecek, J. Polyethylene glycols containing enzymatically degradable bonds. Makromol. Chem., 187: 1131-1144, 1986.
Davis, F. E, Abuchowski, T., Van Es, T., Palczuk, N. C., Chen, H-L., Savoca, K., and
Wieder, K. Enzyme-polyethylene glycol adducts: modified enzymes with unique
properties. In: G. B. Braun, G. Manecke, and L. B. Wingard, Jr. (eds.), Enzyme
Engineering, Vol. 3, pp. 169-173. New York: Plenum Publishing Corp., 1978.
Somack, R., Sailer, M. G. P., and Williams, L. D. Preparation of long-acting superoxide dismutase using high molecular weight polyethylene glycol (41,000-72,000
daltons). Free Radical Res. Commun., 12-13 pt. 2: 553--562, 1991.
4594
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1993 American Association for Cancer
Research.
Polyethylene Glycol-modified Chimeric Toxin Composed of
Transforming Growth Factor α and Pseudomonas Exotoxin
Qing-cheng Wang, Lee H. Pai, Waldemar Debinski, et al.
Cancer Res 1993;53:4588-4594.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/53/19/4588
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1993 American Association for Cancer
Research.