The Anti-Human Tumor Effect and Generation of

ICANCER RESEARCH57. 1335-1343, April 1. l997J
The Anti-Human Tumor Effect and Generation of Human Cytotoxic T Cells in
SCm Mice Given Human Peripheral Blood Lymphocytes by the in Vivo
Transfer of the Interleukin-6 Gene Using Adenovirus Vector'
FUniiaki
Tanaka,
Masako
Abe,
Tsuyoshi
Akiyoshi,
Tatsuji
Nomura,
Keizo
Sugimachi,
Tadamitsu
Kishimoto,
Tomokazu Suzuki, and Masaji Okada2
Departments of Clinical Genetics (F. T., M. A., T. S.. M. 0.1 and Surgical Oncology IT. A.J. Medical institute of Bioregulation. Kyushu University. 4546 Tsurumihara, Beppu,
Oita 874; Central Institutefor Erperimental Animals, Kawasaki, Kanagawa 210 fT. NJ; Department of Surgery ii, Faculty of Medicine, Kyushu University, FukuOka, FukuOka
812 (K. 5.1; and Department of Medicine III. Osaka University Medical School, Suita, Osaka 565 (T. K.J, Japan
ABSTRACT
Interleuldn-6
@
(IL-6) was found to function as a late-acting
killer helper
factor In the differenthtion of CTLs. In the model oftumor-bearing mice,
the systemic administration ofrecombinant IL.6 was found to mediate the
antitumor
effect on the Immunogenic
induction
of murine
CTLs
but
murine
tumors
not on the poorly
via the in vivo
immunogenic
murine
tumors In our previous study. However, an in vivo experimental model
capable ofanalyzing the anti-human tumor effect via the in vivo induction
of human CTLs has not yet been established. Therefore, in the present
study, severe combined immunodeficient mice were given human periph
cml blood lymphocytes
(SCID-PBLIhu),
and thereafter
human tumor cells
were administered Lp. Into these SCID-PBLIhumice as a model of human
patients with cancer. When these SCID-PBIJhu mice bearing allogeneic
human CESS B blastold tumor cells were treated in vivo with recombinant
Menovirus vector expressing IL-6 eDNA, both the Induction of CDS'
human CTLs against CESS cells In the spleen cells and peritoneal exudate
cells and a prolongation in the survival of these mice were observed.
Furthermore, SCID-PBL/bu mice were given peripheral blood lympho..
cytes from patients with cancer (gastric or rectal cancers) and autologous
human tumor cells. The in vivo administration of recombinant adenovirus
vector expressing
IL-ti eDNA Induced
CDS'
human
CTLs specific for
autologous human tumor cells from human precursor T cells. The in vivo
Injection
of the IL-6
gene also inhibited
growth
and metastasis
In autolo
gous human cancers. Based on the above findings, the experimental model
using SCID-PBL4'hu mice and the 11-6 gene delivered in vivo by an
adenovirus
vector
might
therefore
provide
a new strategy
capable
of
analyzing an anti-human tumor effect and the in vivo induction of human
CTLs by cytokine gene therapy without using the human body.
gene-transfected
murine
lung cancer
cells could
exert
an antitumor
effect by the induction of tumor-specific CTLs and the activation of
macrophages. The IL-6 gene-transfected murine fibrosarcoma also
could mediate an antitumor effect by T cells or macrophages (14, 15).
Recently, murine tumor cells transfected with cDNA for several
cytokines (such as IL-i, IL-2, IL-3, IL-4, IL-S. IL-6, IL-7, IL-12,
tumor necrosis factor a, granulocyte CSF, GM-CSF, and IFN--y) were
found to exert an antitumor activity based on different immunological
mechanisms.
These
methods
all strongly
induce
local
antitumor
im
mune responses by the activation of CTLs, CD4@ T cells, macro
phages, eosinophils, granulocytes, or NK cells while causing a mm
imal degree of systemic toxicity (13—25). Fearom et al. (16)
demonstrated that poorly immunogenic murine colon tumor cells
transfected with IL-2 cDNA could induce the generation of CD8@
cTLs and elicit a strong antitumorimmuneresponse.Tepper et a!.
(21) reported that murine plasmacytomas and mammary adenocarci
nomas transfected with IL-4 cDNA could mediate an antitumor effect
by an inflammatory infiltrate composed of eosinophils and macro
phages. In contrast, Golumbek et aL (22) demonstrated that gene
therapy using IL-4 cured renal cancer-bearing mice by the generation
INTRODUCTION
CytotoxicT cells play an importantrole in the immuneresponses
against
survival time and an inhibition of micrometastasis; Refs. 1, 10, and
11). However, the systemic administration of rIL-6 did not mediate an
antitumor effect on such poorly immunogenic murine tumors as
RL 9 8 and
2T lymphoma (12).
In contrast, murine tumor cells transfected with IL-6 cDNA using
a retrovirus vector induced a strong antitumor effect even on these
poorly immunogenic murine tumors via the activation of CD4@ T
cells and CD8@ T cells (12). Porgador et al. ( 13) reported that IL-6
tumor cells and also induce
an antitumor
effect in vivo (1—3).
ofCD8@ CTLs.
Dranoff et a!. (23) compared the ability of different cytokines
(GM-CSF, IL-6, IL-4, IFN, stem cell factor, granulocyte CSF, IL-2,
IL-7, and so forth) and other molecules (such as adhesion molecules)
Cytokines from helper T cells are required for the induction of
cytotoxic T cells in both the murine and human systems (1, 4—9). to enhance the immunogenicity of tumor cells. Irradiated Bl6 mela
noma cells expressing murine GM-CSF produced the most potent,
IL-63 was found to function as a late-acting killer helper factor in the
differentiation
of CTLs (7). Furthermore,
the systemic
long-lasting,
administration
of rIL-6 was found to induce the in vivo generation of CD8@ CTLs
and specific
antitumor
immunity,
requiring
both CD4@
and CD8@ T cells. However, Saito et a!. (24) argued that GM-CSF
against syngeneic murine tumors, FBL-3 erythro-leukemia, or fibro
gene-transfected murine bladder cancer vaccines are consistently less
sarcoma while it induces an antitumor effect (a prolongation
effective than IL-2 gene-transfected vaccines. Therefore, the antitu
mor effect of cytokine gene therapy may vary depending on the assay
systems used and the type of tumors.
In human cancer patients, the antitumor effect of cytokine gene
of
Received 8/19/96; accepted 1/31/97.
Thecostsof publicationof this anicleweredefrayedin partby the paymentof page
charges. This article must therefore be hereby marked advertisement in accordance with
18 U.S.C. Section 1734 solely to indicatethis fact.
therapy
might be different
from mice with murine
tumors
depending
for Scientific Research on Priority Areas
on the tumors and cytokines. Therefore, by using the SCID-PBLIhu
of Cancer 06282244 and 04253226 and Grant-in-Aidfor Scientific Research(type C)
mice, we analyzed the in vivo anti-human tumor effect of cytokines.
1 This work was supported by Grants-in-Aid
3963 from the Ministry of Education, Science, Sports and Culture of Japan.
2 To whom requests for reprints should be addressed. Phone: 81-977-27-1666;
Fax:
81-977-27-1607.
3 The abbreviations
used are: IL,
interleukin;
rIL,
recombinant
IL; SCID,
severe
combined immunodeficient; PBL, peripheral blood lymphocyte; SCID-PBIJhu mice,
SCID mice given human PBL; Adex IL.6, recombinant adenovinu vector expressing
interleukin 6 cDNA (AdexlSRahlL-6);
Adex SW, recombinant adenovints vector cx
SCID-PBIJhu mce were established by the administration of human
PBLs i.p. into CB-17-SCID mice. CD3@ and CD8@ human T cells
were observed in vivo in spleen cells and PECs in these mice. The
administration
pressing no specific gene (AdexlSRaW); PEC, peritoneal exudate cell; m.o.i., multiplic
ity ofinfection; NIL natural killer; GM.CSF, granulocyte macrophage colony-stimulating
factor, poly(A), polyadenylic acid.
of human
rIL-6
augmented
the generation
of human
CTLs in the SCID-PBLJhu mice stimulated with human B cell tumor
CESS cells (12, 26).
Furthermore, cytokine genes introduced into human cancer cells
1335
IL-6 GENEThERAPYFOR HUMANCANCERSIN SCID-PBlJhuMICE
need to be transfected ex vivo when a retrovirus vector is used.
However, some human cancers cannot be removed by surgical treat
ment. Moreover, it is very difficult to establish cell lines of human
cancers obtained from patients. In such circumstances, the ability of
such a vector to transfect cytokine genes directly into an in vivo tumor
may be a very important factor for applying gene therapy to all human
patients with cancer. The direct in vivo transfer of cytokine gene(s)
(MAGE)-l, -2, -3, -4, BAGE, GAGEI-2, or GAGEJ-6genes were obtained
from the Japanese Cancer Research Bank (Tokyo, Japan; Refs. 27 and 28).
TE5, TEl, TE9, TE13, KY3O, KY450, and KY51Ohuman esophageal squa
using an adenovirus vector appears to be promising for human cancer
cer cell line, was established from a patient (a 26-year-old female) with
moderately differentiated adenocarcinoma. 00108 and GC814, both gastric
cancer cell lines, were established from a 61-year-old female with mucinous
adenocarcinoma and an 84-year-old male with poorly differentiated adenocar
treatment.
We
used
nonproliferative
human
adenovirus
deleted
from
the E1A, E1B, and E3 regions. The IL-6 gene was inserted into this
adenovinis
vector
and was named
Adex
IL-6.
In the murime tumor
cell line, was established in our laboratory from a patient (a 68-year-old
female)with moderatelydifferentiatedadenocarcinoma.RC422,a rectalcan
system, the in vivo administration of Adex IL-6 into mice bearing
tumors mediated a stronger in vivo antitumor effect against poorly
cinoma,
immunogenic
formation
tumors
by the activation
of murine
CTLs
than that of
adenovirus vector.4
against
CESS
cells
and thus prolong
their
survival.
of tumor masses
maintained
in vitro. Murakami,
supplemented
in our specific
pathogen-free
into
an EB virus transformed
human
B blastoid
cell
100
@g/mlstreptomycin,
(A)
hlL-6
poly A
Promotor
(SRa)
@:L@EIA,E1BE3deletiondeletion(.2874bp)(-1 AdexiSRu
c@
animal
hlL-6
(Adex IL-6)
879bp)poly
A
Mice. CB-l7 scid/scid (SCID) mice were obtainedfrom the CentralInsti
tute for Experimental Animals (Kawasaki, Kanagawa, Japan) and maintained
cages
from
Further
AND METHODS
isolator
treatment
plate), and these tumor cell lines were
with 10% FCS, 100 units/mi penicillin,
Promotor
(SRa)
AdexiSRaW
A
E1A,E1B
in filter-capped
by surgical
and 50 @.LM
2-mercaptoethanol.
PBLJhu mice were given PBLs from a patient with gastric cancer or
rectal cancer, injected with autologous cancer cells, and then treated
with Adex IL-6. Human CTLs against autologous human cancer cells
were generated in vivo by the treatment of Adex IL-6 from naive
human T cells. An in vivo anti-human tumor effect also was observed
in these SCID-PBLIhu mice after they were treated with Adex IL-6.
Therefore, the experimental model using SCID-PBL/hu mice and
cytokine genes delivered by an adenovirus vector might provide a new
strategy capable of analyzing the anti-human tumor effect by cytokine
gene therapy without using an actual human subject.
MATERIALS
obtained
in these mice, the tumor cells were transferred
in vitro cultures (Linbro multiwell
more, Adex IL-6 was injected into the SCID-PBLIhu mice bearing
either autologous gastric cancer cells or rectal cancer cells. SCID
@
The tumor cells
line, was a gift from Dr. Fujie (Kyushu University). The tumor cell lines were
maintained in vitro in RPMI 1640 (Flow Laboratories, Inc., McLean, VA) and
In the present study, the administration of Adex IL-6 into SCID
PBLIhu mice bearing the human allogeneic B blastoid tumor CESS
was found to induce the in vivo generation of human CD3 , CD8
CTLs
respectively.
these patients with cancer were injected s.c. into SCID mice. After the
JL-2, JL-7, twnor necrosis factor a, IFN-y, or GM-CSF genes using
@
mous cancer cell lines expressing the MAGE-1, -2, -3, -4, -9, or -12 gene were
provided by the cell bank of Tohoku University (Sendai, Japan) and Dr. Y.
Shimada (Kyoto University, Kyoto, Japan; Ref. 29). GC629, a gastric cancer
deletion
E3
deletion
(.2874bp)
(-1879bp)
(Adex SW)
(B)
center
(Kyushu University). The SCID mice were fed autoclaved food and water. All
mice received daily trimethoprim/sulfamethoxazole (Baktar; 40 mg of tn
methopnim and 200 mg of sulfamethoxazole
per 5 ml of suspension; 0.125 ml
of suspension for every 4 ml of drinking water per mouse) supplied by the
Shionogi Co. (Osaka, Japan) to prevent infection with Pneumocystis carinii.
Reagents and Antibodies. FCS (lot 1111958) was purchasedfrom Hy
clone Laboratories, Inc. (Logan, UT). Monoclonal anti-human CD3 (lot T329)
and anti-human CD8 (lot T82l) antibodies were purchased from Ortho Diag
nostic
System,
Inc. (Raritan,
NJ). Baby rabbit complement
(lot 3441)
:@‘600
U
4
-J
400
was
purchased from Cedarlane Laboratories Ltd., (Hornby, Ontario, Canada).
Generation of SCID-PBLFhu Mice. PBLs were obtained from patients
with cancer or healthy volunteers after Ficoll Hypaque density centnifugation
as described previously (5, 8). PBLs (2.5 X l0@)suspended in 300 @l
of HBSS
were mixed with 100 p1 of matnigel (Becton Dickinson Labware, Bedford,
MA). Six- to 8-week-oldfemale SCIDmice receivedpretreatmentwithwhole
200
body X-ray irradiation (2 Gy; 100 kV, and 3.5 mA for 5 mm) by using an
X-raygenerator(model MBR-l505R, HitachiMedicalCo., Tokyo, Japan)1 h
before the i.p. injection of PBLs.
Tumor Cell Lines. CESS, an EBV-tnansformed human B-blastoid cell
line, and K562, a NK-sensitive leukemia cell line, were maintained in complete
adenocancinoma
7
14
days after the treatment of Adex IL.6
gene
R. R. Nyuta. J. Takano, F. Tanaka,
deleted El region of pAdexlw (pAdexlSRahIL-6).
No expressing gene with SRa
promoter and SV4O poly(A) signal sequences was inserted in the deleted El region of
pAdexlw (pAdexlSRaW).
B, the kinetics of human IL-6 activity in sent from SCID
cell lines expressing
4 M. Okada, H. Hamada, M. Abe, H. Tomonaga,
4
Fig. 1. Establishment of AdexlSRahlL-6 (Adex IL-6) and AdexlSRaW (Adex SW)
and the production of IL-6 in SCID-PBLJhu mice. A, the recombinant adenovirus is based
on adenovirus type 5 and lacks the E1A, E1B, and E3 regions. The expression unit of
human IL-6 with SRa promoter and SV4Opoly(A) signal sequences was inserted in the
medium (5). MKN7, MKN45, MKN74, NUG-C3, and Kato-Ill human gastric
adenocancinoma cell lines and Colo2OS, Co1o32ODM,RCMI, and WiDr
colorectal
012
melanoma
antigen
I. Salto, T. Kishimoto,and T. Suzuki. The antitumoreffect againstpoorly immunogenic
tumors by the in vivo injection of IL-6 and cytokine genes using adenovirus vector,
manuscript in preparation.
PBLJhu mice treated with i.p. injection of Adex IL-6 was assessed using the chemilumi
nescence enzyme immunoassay system as described in “Materials
and Methods.―Data
points, mean of II..-6 activity in five mice; bars, SD.
1336
lI@-6GENETHERAPYFOR HUMANCANCERSIN SCID-PBlJbuMICE
Adenovirus Vector. The recombinant adenovirus vector was based on
adenovirus type 5 and lacks the E1A, E1B, and E3 regions (Fig. lA; Ref. 30).
human tumor cells as target cells (5, 8). Specific cytotoxicity (%) = [(exper
imental 51Crrelease —control 5tCr release)/(maximum 5tCr release —control
The recombinant human IL-6 adenovirus vector (AdexlSRaIL-6; Adex IL-6)
5 ‘Crrelease)]
contained SRa promoter, human IL-6 cDNA, and the SV4O poly(A) signal
sequence inserted into the El-deleted region of a cassetted cosmid vector.
Assay for ll@-6ActIvity. Human IL-6 activity was assessed using the
Chemiluminescence Enzyme Immuno Assay system (SRL Co., Tokyo, Japan).
Recombinantadenovirus(AdexlSRaW;AdexSW)that wasusedas a control
Statistics.
X
100.
Survival and tumor growth were observed every day after the
contained the SRa promoter and the SV4O poly(A) signal. Viral stocks were
injection of human tumor cells into SCID-PBLJhu mice for more than 100
generatedfrom confluent monolayersof 293 cells. The supernatantthatcon
days.
mined virus particles was stored at —80°Cuntil use. After the viruses were
tumors treated with Adex IL-6 or Adex SW was statistically evaluated using
concentrated, the virus stock (1.0 X lO@plaque-forming units/ml) was stored
at —
80°C(31, 32). The tumor cells were found to be transduced with the
adenovirus in vivo by using an adenovirus vector containing the LacZ gene
(Adex LacZ) and p3-galstaining.
the Kaplan-Meier test (33).
Human
Tumors
by the Injection
of Adex
IL-6.
CESS
was used as an allogeneic tumor cell line. PBLS (2.5 X l0@) from the healthy
donor were administered i.p. into SCID mice on day 0. These mice (eight
animals in each group) were inoculated i.p. with 2 X i07 human tumor CESS
cells on day 2. The SCID-PBL/hu mice bearing human tumor cells were then
treated
i.p. with Adex LL-6 at a m.o.i. of 100 or with Adex SW on day 3 after
the injection ofPBLs. Twenty-four days after the injection ofhuman PBLs, the
mice were sacrificed. The spleen cells and PECs obtained from these SCID
PBL/hu
mice were assessed
for lyric activity
against
human
tumor
cells.
In Vivo Induction of Human Cytotoxic T Cells in SCID-PBLflau Mice
against Autologous
Human Tumors by the Injection ofAdex
IL-6. GC629
and GC408 gastric cancer cell lines and RC422, a rectal cancer cell line, were
used as autologous tumor cells. PBLs (2.5 X l0@) from patients with gastric
against
day 0. These mice (either eight or five animals in each group) were injected
with autologous human cancer cells on day 2 and were then treated with Adex
IL-6 at a m.o.i. of 100 or Adex SW on day 3 after the administration of human
PBLs. Cytolytic activity in the spleen cells and PECs from these SCID-PBLJhu
mice at 24 days after the injection of human PBLs was assessed against several
kinds of human tumor cells.
Assay for Cytolytic ActIVity. The percentage of specific cytolytic activity
by means of a microcytotoxicity
mice bearing
human
assay using 5tCr-labeled
Allogeneic
Human
Tumor
Cells
in SCID-PBL/hu
Mice.
To investigate the in vivo activation of human T cells, PBLs from
healthy volunteers were administered i.p. into CB-l7 SCID mice and
then these SCID-PBLIhu mice were in vivo stimulated with human
CESS tumor cells. In the spleen cells, lymph node cells, and PECs
obtained from SCID-PBLIhu mice on a day between day 21 and day
28 after the injection of human PBLs, 10—30%of CD34 CD84
human T cells and CD34 Ct)4@ human T cells were observed (data
not shown).
On day 0, human PBLs were administered in amounts ranging from
1.25 x l0@to 2.5 x i0@ into SCUD mice. CESS cells (1 .0 X i0@or
2.0 X l0@) were i.p. administered on day 2, and then Adex IL-6 was
injected i.p. into these SCID mice on day 3. Human serum IL-6
activity in the SCID-PBLJhu mice was 762.5 pg/ml 24 h after the
injection of Adex IL-6 and was still detected until 24 days after
injection (7.8 pg/mi; Fig. 1B).
cancer or a patient with rectal cancer were administered i.p. into SC@ mice on
was measured
in SCID-PBLJhu
In Vivo Generationof CD34 CD8@Human Cytotoxic T Cells
or CD8 monoclonal antibody (1:20 diluted) for 15 mm at 4°Cand then
incubated with baby-rabbit serum complements (1:10 diluted) for 45 mm at
37°Cas described previously (1, 4, 5).
In Vivo Induction of Human Cytotoxic T Cells in SCID-PBLIhu Mice
Allogeneic
data on survival
RESULTS
Treatment
of Cells with Antibodies and Complements.
Spleen cells
(2 X l06/ml SCID-PBLJhu) and PECs were treated with an anti-human CD3
against
Experimental
As shown
in Table
1 (Experiments
I and II), the cytotoxic
cells
were generated in the in vivo spleen cells and PECs from these SCID
mice on day 24 after the injection of PBLs. When SCID-PBLIhu mice
were given 2.5 x l0@ human PBLs and 2.0 X iO@CESS cells and
treated with Adex IL-6, spleen cells and PECs from these SCID
PBL/hu mice exhibited higher cytotoxic activity against CESS cells
but not against NK-sensitive K562 cells. Therefore, in the following
experiment, SCID-PBLIhu mice given 2.5 X l0@ human PBLs and
1L-6Experiment
Table I The in vivo inductionofhuntan CD8 positive cytoroxic T cells in the SCJD-PBUhu mice by administration ofAdex
1b Spleen
SCIDmice
cells Effector cells from
treated with:%
cytotoxicity―Human
specific
K56220:1 CESS (E:T ratio)
4:12.5
PBLS
CESS cells
0.12.5
X l0@
0.11.25
X l0@
0.12.5
X lO@
0.02.5
X l0@
0.1ExperimentllcPEC
X itf
0.52.5
2.5 X l0@
2.0 X l0@
1.0 x l0@
Adex IL-685.5
Adex IL-652.2
1.0X l0@
AdexIL-623.0
2.0 X i0@
2.0 X [email protected]
Adex SW1.0
2.0 X i0@
2.0 X l0@
x l0@
0.2Experimentlll―
4:1
Gene(s)Against
±9.2
24.0 ±2.2
1.0 ±0.3
0.2 ±
±4.4
±3.2
±0.1
20.2 ±2.2
9.1 ±2.1
1.1 ±0.2
6.3 ±1.1
1.8 ±0.2
0.8 ±0.1
0.8 ±
0.2 ±
0.0 ±
±0.3
1.5 ±0.3
0.2 ±0.1
0.3 ±
4:1
16:1
88.9 ±6.5
Adex IL-616:1
Adex SWI
5.2 ±1.1
3.9 ±
6.1 ±0.2
2.7 ±0.4
1.6 ±
Specific
25:1
3.2Anti-human
0.1Anti-human
CD3 ,@Abe0.0
CD8 mAb0.0
represent
the
mean
± SD
of
b Two days after the administration
or Adex SW (m.o.i.
cytotoxic
activity
in
the
triplicate
4:1
27.8 ±3.1
1.1 ±1.1
Effector splenic cells Cytotoxicityantibody(—)50:1
treated with%
a Data
Against
20:1
cultures
of
spleen
cells
21.9 ±2.3
18.8 ±1.1
±0. 1
±0.0
2.6 ±0.2
2.5 ±0.1
from
eight
mice
in
each
12:1
17.7 ±
0.4 ±
1.1 ±0.2
group.
of various numbers of human PBLS i.p. into the SCID mice, various numbers of tumor cells (CESS) were inoculated i.p., and then Adex IL-6.
100) was injected i.p. on day 3. These mice (eight mice in each group) were sacrificed on day 24. The cytotoxic activity in spleen cells was assessed against
CESScellsand K562cells.
Cscm miceconstructed
with2.5X l0@
human
PBLS
wereinjected
with2.0X lO@
CESS
cellsandthentreated
withAdexIL-6orAdexSW.Onday24aftertheinjection
ofPBLS,
the cytotoxic activity in the PECs was assessed against CESS cells and K562 cells.
d Splenic effector cells from SCID-PBIJhu
mice injected with Adex IL-6 were treated with a monoclonal anti-human CD3 antibody or an anti-human CD8 antibody in the presence
of complement as described in “Materials
and Methods.―After the treatment, effector cells were assessed for cytotoxic activity against CESS.
e mAb.
monoclonal
antibody.
1337
lL-6 GENE THERAPY FOR HUMAN CANCERS IN SCID-PBLIhu MICE
higher than that of Adex SW (P < 0.01). The number and mass of the
CESS tumors in the peritoneal cavity of SCID-PBLJhu mice de
(A) (%)
creased
more significantly
in the mice treated
with Adex IL-6 than in
those treated with Adex SW (Fig. 3). These results indicate that IL-6
gene transfer using adenovirus vector could induce an in vivo antitu
mor effect against the human tumor cell line CESS, in SCJD-PBLJhu
mice.
Furthermore, by using the healthy donors and HLA type-matched
autologous EB-transformed B cell tumors, the human CTLs and
protective immunity (prolongation of survival time) against autolo
gous EB-transformed tumors were also induced in the SCID mice
given human PBLs by treatment with Adex IL-6 (data not shown).
in Vivo Generation of Human CTLs against Syngeneic (Autol
ogous) Cancer Cells in SCID-PBLIhu Mice by the Injection of
>
Cl)
Adex IL-6. To investigatethe in vivogenerationof cytotoxicactivity
against
(B)
syngeneic
(5
Cl)
and PECs
30
40
50
60
70
80 (day)
Fig. 2. The antitumor effect of Adex IL-6 on SCID-PBLJhu mice bearing human
tumors.A, 2.0 X l0@viable CESS human tumorcells were administeredi.p. into the
SCID-PBIJhu
(autologous)
human
tumors,
SCID-PBL/hu
mice
given PBLs from cancer patients and autologous cancer cells were
treated with Adex IL-6. Gastric cancer cell lines and rectal cancer cell
lines were established from cancer patients with gastric and rectal
cancer, respectively.
SCID mice were given PBLs from a patient with gastric cancer and
autologous gastric cancer cell line GC629 cells. Those SCID-PBLIhu
mice were then treated with Adex IL-6 in vivo. Cytotoxic activity
against autologous GC629 cancer cells was induced in the spleen cells
mice given 2.5 x l0@human PBLS (16 mice). On day I after the injection
of tumor cells, these mice were treated with Adex IL-6 (8 mice; •)or Adex SW (8 mice;
0). The survival of these mice is shown. B, 5.0 X l0@viable CESS human tumor cells
were administered i.p. into SCID-PBL/hu mice given 2.5 X l0@human PBLS (10 mice).
On day I after the injection of tumor cells, these mice were treatedwith Adex IL-6
(5 mice; •)or Adex SW (5 mice; 0). The survival of thesemice is shown.
2.0 X l0@ CESS cells were used. On the other hand, little cytotoxicity
was observed in those cells from SCID-PBLIhu mice resulting from
the administration of Adex SW (Table 1, Experiments I and II).
Furthermore, the treatment of effector cells in spleen cells from
SCID-PBL/hu mice with a monoclonal anti-human CD3 antibody or
a monoclonal antihuman CD8 antibody in the presence of comple
ments almost abolished the cytotoxic activity against CESS cells
(Table I, Experiment III). These results demonstrated that the cyto
toxic cells generated in the SCID-PBL/hu mice immunized with
CESS cells and treated with Adex IL-6 were CD3@ CD84 human
cytotoxic T cells capable of lysing CESS human allogeneic tumor
cells specifically.
in Vivo Anti-Human Tumor Effect Caused by the Injection of
Adex IL-6. To examine whether or not the in vivo injection of Adex
IL-6 into SCID-PBLIhu mice bearing human tumors mediates an
anti-human tumor effect, SCID-PBLJhu mice were injected i.p. with
of these
mice by treatment
with Adex
IL-6 but not with
Adex SW. Cytotoxic cells against GC629 cells were not generated
when SCID mice were given GC629 cells in the absence of PBLS
from a patient with gastric cancer even after treatment with Adex IL-6
(Table 2, Experiment I). On the other hand, cytotoxic cells specific
against autologous
RC422 cancer cells were generated in the SCID
PBL/hu mice when SCID mice were given PBLs from a patient with
rectal cancer, and the autologous rectal cancer cell line RC422 cells
were them treated with Adex IL-6 in vivo (Fig. 4B and Table 2,
Experiment II). The cytotoxic activity of these effector cells was
almost completely abolished by the treatment of the effector cells in
the spleen cells and PECs from SCID-PBLIhu with an anti-human
CD3 or CD8 antibody and complements. These results demonstrate
that the in vivo administration of Adex IL-6 into the SCID-PBLIhu
mice bearing autologous human cancer cells induced CD3@, CD8@
human cytotoxic T cells against autologous cancer cells.
Specificity of in Vivo CTLs Generated in the SCID-PBIJhu by
the Administration of Adex IL-6. To investigate the specificity of
human CTLs generated in vivo in the SCID-PBLIhu mice, cytotoxic
activity
of these CTLs
was assessed
using a panel of many
kinds of
allogeneic human gastric, colorectal, and esophageal cancer cells. The
cytotoxic activity induced by the administration of Adex IL-6 in
SCID-PBUhu mice given PBLS from a patient with gastric cancer
(GC629) and autologous cancer cells (GC629) was not exerted against
other cancer cell lines such as GC408, GC814, NUG-C3, MKN7,
MKN45, MKN74 (gastric adenocarcinoma cells), RC422, Colo2O5,
Colo32ODM, RCM1, WiDr (colorectal adenocarcinoma cells), KY3O,
KY450, KY51O, TE5, TE7, TE9, TE13 (esophageal squamous cell
carcinoma) cells, CESS, Murakami K562 cells, and autologous PBLS
2 X l0@ CESS cells on day 2 and were treated i.p. with 100 m.o.i. of
Adex IL-6 or Adex SW on day 3. As shown in Fig. 2A, 75% of the (Table 2, Experiment I; Fig. 4A). In contrast, the cytotoxic activity
mice (6 of 8) were alive on day 80, when Adex IL-6 was injected. The
induced by the administration of Adex IL-6 in the SCID-PBIJhu mice
cure rate in SCID-PBLIhu mice injected with Adex IL-6 was signif
given PBLs from a patient with rectal cancer (RC422) and autologous
icantly higher than that of 12.5% (1 of 8) in SCID-PBLIhu mice
cancer cells (RC422) was specifically exerted against autologous
injected with Adex SW (P < 0.05). In another experiment, SCID
RC422 tumor cells (Fig. 4B; Table 2, Experiment H). As shown in
PBLJhu mice were injected i.p. with 5 X 106 CESS cells and were
Fig. 4G. the cytotoxic activity induced by the administration of Adex
treated i.p. with Adex IL-6. As shown in Fig. 2B, 60% (3 of 5 mice)
IL-6 in the SCID-PBL/hu mice given PBLs from a patient with gastric
were alive on day 80. In contrast, all the mice died of tumor devel
cancer GC408 and autologous cancer cells (GC408) was not exhibited
opment by day 36 when treated with Adex SW. The cure rate in the
against RC422 cells.
SCID-PBL'hu mice treated with Adex IL-6 was also significantly
These results suggest that human CTLs, induced in spleen cells and
1338
IL-6 GENE ThERAPY FOR HUMAN CANCERS IN SCID-PBIJbu MICE
(A)
Adx SW
Fig. 3. A, inhibitim
of tumor engraftment
Adx 1L4
and growth in SCID
PBIJhu mice by treatment with Adex IL-6. 2.0 X l0@viable CESS
human tumor cells (lILA types A2, B5l, B27, DR7, MTI, and MT3)
were administered i.p. into SCID-PBL/hu mice given 2.5 X l0@alloge
neic human PBLS (liLA types Al, A3I, B55, B6l, Cl, C3, DPi, DR8,
and DQ1; 16 mice). On day 1 after the injection of tumor cells, these
mice were treated with Adex IL-6 (8 mice) or Adex SW (8 mice). An
autopsy was
@ormedon the mice when they died or were sacrificed on
day 100. Tumor masses in the peritoneal cavity of these SCID-PBIJhu
micewerefixedwith 10%formaldehydeandconfirmedby the staining
with monoclonal anti-human B cell antigens antibodies as human CESS
tumor cells. B, tumor volume in A was measured in mm using a caliper
andcalculatedas widthx lengthx height.
(B)
Adex SWAdex
IL-6tumor
size (mm3)totaltumor
(mm3)totalNo.11318,
330, 527, 224239960,
40160No.21054,105115966,2894No.3374,
size
60,
05479(-)(-)No.4184,
1
20227(-)(-)No.5288288(-)(-)No.6288288(-)(-)No.7168168(-)(-)No.8(—)
23,
@(—)(—)(—)
PECs from SCID-PBL/hu mice bearing autologous human cancer
cells by treatment with Adex IL-6, can recognize the antigen specific
to the autologous human cancer cells.
in Vivo Antitumor Effect on Autologous Human Cancer Cells
by the Administration of the iL-6 Gene Using Adenovirus Vector.
An in vivo anti-human tumor effect was observed in the SCID-PBL/hu
mice bearing autologous human gastric cancer (C1C629cells), as well
as CFL activity against autologous cancer cells in those mice by the
administration of Adex IL-6 (Fig. 5A). Neither a tumor mass nor
metastasis
was
shown
in the peritoneal
cavity
and organs
(liver,
spleen, kidney, pancreas, lung, and brain) in these SCID-PBLJhu mice
treated
with
Adex
IL-6.
On the
other
hand,
tumor
masses
vivo administration
of Adex IL-6 7 days after the injection
of GC629
cancer cells also mediated the in vivo antitumor effect (tumor regres
sion and prolongation of survival) in the SCID-PBLJhu mice (data not
shown), suggesting that Adex IL-6 is effective even against estab
lished human tumors.
These results indicate that the direct injection of the IL-6 gene in
vivo using adenovirus
vector
could mediate
anti-human
tumor effect
on the autologous cancer and that the experimental model of SCID
PBIJhu providesa usefulstrategyfor analyzingthe in vivoimmuno
regulatorymechanismsagainstanti-humantumorsby the activationof
human T cells.
were
observed in the pentoneal cavity of those SCID-PBLIhu mice bearing
GC629 cancer cells treated with Adex SW (Fig. 5B). No antitumor
effect was observed in the SCID mice bearing autologous human
GC629 tumor without the injection of PBLs from a GC629 cancer
patient even when treated with Adex IL-6 (data not shown). The in
DISCUSSION
In the present study, CD8@ human CTLs specific against autolo
gous human cancer cells, as well as CD8@ human CTLs specific
against allogeneic human cancer cells, were generated in the in vivo
1339
II.-6 GENEThERAPYFOR HUMANCANCERSIN 5CID-PBIJhuMICE
IL-6Experiment
Table2 The in vivo inductionof humancytotoxic T cells against autologous humancancer cells in the SCID-PBJJhumice by treatmentwithAdex
ratioPBL
1.2:1+
r Treatedwith:
Spleen Cells―
specific cytotoxicity,bE:T ratio
0C629
Adex IL-6
0.6—
+
+
GC629
0.0±0.1+
+
+
0C629
2.6±0.6
1.8±0.3
0.3±0.1
2.1
±0.1
1.0±0.1
0.1+
0.2+
0.1+
0.6—
+
+
+
+
—
+
+
+
GC629
K562
PBLC
RC422
1.9
0.3
0.0
6.5
1.7
0.0
0.0
5.3
0.4
0.4
0. 1
4.5
1.1
0.3
0. 1
6.8
±0.2
±0.1
±0.0
± 1.1
1.7
0.0
0.0
6.2
—
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
RC422
NUG-C3
MKN7
MKN45
MKN74
GC408
Colo2OS
Co1o32ODM
RCM1
WiDr
KY3O
KY450
TE5
TE7
TE9
TE13
2.1 ±0.2
0.8
0.4
1.3
0.0
0.4
0.7
0.2
2.0
1.4
0.3
0.7
1.8
0.2
0.0
0.0
0.4
±0.1
±
0.7 ±0.2
0.1+ +
0.1+
0.1+
±0.2+
0.1+
0.1+
0.2+
0.1+
0.3+
0.2+
0.1+
0.1+
0.2+
0.0+
0.1+
0.1+
Against
PEC cells―
% specific cytotoxicity, E:T
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
0.2Experiment
+
ratioPBL
50:1
10:1
2.5:1
20:1
5:1
24.2 ±2.1
13.5 ± 1.2
8.8 ±0.9
18.6 ± 1.5
10.7 ± 1.0
RC422
±0.1
±0.1
±0.2
±0.2
1.1±
0.3
0.4Â0.
±1
1.0±
0.2
0.3 ±0.1
1.2 ±0.3
0.7±0.1
0.9 ±0.3
1.3 ±0.4
0.7 ±0.1
0.0 ±0.0
0.4 ±0.1
1.1 ±0.3
0.1 ±0.0
1.0 ±0.3
1.3 ±0.3
0.8 ±0.2
0.1 ±0.0
0.9 ±0.2
0.0 ±0.0
AdexIL-6
+
+
+
+
Against
50:1
RC422
RC422
GC629
+
3:1
35.0 ±4.1
3.8 ±0.5
3.8 ±0.2
1.2
0.7
0.0
4.5
±
±
±
±
0.2±
0.3 ±
±
±
±
±
±
±
±
±
±
±
±
±
±
PECs E:T
12:1
42.4 ±3.1
4.7 ±0.8
0.7%
0.0 ±0.1
±0.3
±0.1
±0. 1
±0.8
0.5Â0.
±1
Spleen cellse E:T ratio
RC422
4.2+
0.5+
±0.1
±0.1
±0. 1
±0.4
0.1±
0.0
0.2 ±0.1
0.0 ±0.1
1.4±0.3
0.0 ±0.0
0.4 ±0.1
1.0 ±0.2
0.3 ±0.1
0.0 ±0.0
0.0 ±0.1
0.0 ±0.0
0.2 ±0.0
0.8 ±0.2
1.2 ±0.1
0.0 ±0.1
0.0 ±0.0
0.0 ±0.1
11dTreated with:
5:1+
±0.2
±0.2
±0. 1
±0.6
5.2 ±
100:1
19.3 ± 1.9
0.9 ±0.1
0.7 ±0.2
20:1
84.5 ±6.5
11.1 ± 1.3
1 1.8 ± 1.7
83.3 ± 10.1
10.5 ± 1.2
9.2 ± 1.5
37.2 ±
6.0 ±
5.1 ±
specific
RC422)Spleen
cytotoxicity after the treatment with antibody and complement (against
ratioExperiment
cells E:T ratio
lIIi
0.5:116.4
40:1
8:1
2:1
10:1
2:1
±1.8
11.5 ±1.5
8.8 ±1.0
42.1 ±2.4
23.6 ±1.5
5.3 ±
0.1 ±0.0
2.3 ±0.3
18.5 ±1.3
0.1 ±0.1
1.6 ±0.3
13.6 ±1.4
0.3 ±0.1
2.3 ±0.4
12.2 ±
0.9 ±0.2
7.9 ±0.9
3.5 ±0.2
4.8 ±0.3
3.0 ±
4.3 ±
1.0Anti-human
0.3Anti-human
CD3 mAb + C'
0.4C'
CD8 mAb + C'
0.6a
PBLs,GC629
2.5
X
l0@
PBLs
from
a patient
with
gastric
cancer
PEC cells E:T
(0C629)
were
administered
i.p.
into
the
SCID
mice.
On
day
2 after
the
injection
of
PBLs
or
without
the
injection
of
cells (obtained from 0C629 cell line) were injected i.p. and then treated with Adex IL-6 or Adex SW. On day 24, the cytotoxic activity in spleen cells and PECs was assessed.
GC629gastric
b Results were expressed as mean % cytotoxicity
±SD in triplicate cultures of spleen cells or PECs in each group (eight mice per group). The cytotoxic activity against
(Colo2OS,Co1o32ODM,
cancer cells, K562, autologous PBL, RC422 rectal cancer cells. gastric cancer cell lines (NUG-C3, MKN7, MKN45, MKN74, or GC408), colorectal cancer cell lines
assessed.C RCM1. or WiDr), or esophageal cancer cell lines (KY3O, KY450, 1'E5, TE7, TE9, or TE13) was
Autologous
PBLS
were
stimulated
with
0.2%
phytohemagglutinin-P
onday
d 2.5 x l0@ PBL from a patient with rectal cancer (RC422)
in the presence
of
1000
units/mI
of rIL-2
for
2 days
in vitro,
and
blastoid
PBLS
againstRC422
2 and then treated with Adex IL-fl or Adex SW on day 3. On day 24, the cytotoxic activity in the spleen cells or in the PECs from these SCID-PBIJhu
cells.eautologous human cancer cells or 0C629 human cancer
Treated
@
2and
p2.5
with
Adex
SW
(all
others
were
treated
with
Adex
were
used
as target
cells.
were administered i.p. into the SC@ mice on day 0. These mice were injected with autologous human cancer cells
mice was assessed
IL-6.
l0@PBLs from a patient with rectal cancer were administered i.p. into the SCID mice on day 0. These mice were injected with autologous human cancer cells on day
thepresence
then treated with Adex IL-6 on day 3. On day 24, spleen cells or PEC from these SCID-PBIJhu mice were treated with a monoclonal anti-human CD8 or CD3 antibody in
of complement. After the treatment, cytotoxic activity was assessed against RC422 autologous human cancer cells.
g ,@Ab,
monoclonal
SCID-PBLIhu
antibody;
C'.
complement.
mice by the administration
of the IL-6 gene using an
adenovirus vector (Adex IL-6). Furthermore, the in vivo antitumor
effect (prolongation of survival, tumor regression, and inhibition of
metastasis)
on these
SCID-PBLJhu
mice
bearing
human
cancer
cells
(autologous gastric cancer, autologous rectal cancer, or B blastoid
tumor CESS) was demonstrated by the treatment of Adex IL-6.
Several lines of evidence support these conclusions. (a) When ade
novirus vector without the insertion of the IL-6 gene (Adex SW)
instead of Adex IL-6 was administered into SCID-PBLIhu mice
bearing human tumor cells, no cytotoxic cells specific against human
immunized tumor cells were generated. Therefore, these SCID
PBL/hu
mice died because
of tumor
development.
(b) No cytotoxic
cells were induced in vivo in the SCID mice bearing these human
tumor cells even when Adex IL-6 and human tumor cells were given
to the SCID mice without the administration of human PBLs. (c) The
cytotoxic activity of effector cells generated in the SCID-PBL/hu
mice by the injection of Adex IL-6 was almost completely abolished
by the treatment of either anti-human CD8 antibody or anti-human
CD3 antibody in the presence of complements. Human T cells en
riched in E rosette-sheep RBC-positive fraction exerted cytotoxic
activity, whereas PBL cells in E-rosette-SRBC-negative
fraction did
not exert such activity (data not shown). (d) Based on the analysis of
surface antigens on the cells using monoclonal antibodies against
human T cells antigens and fluorescence-activated cell sorting, 15—
30% CD3@ human T cells, 5—15%human CD44 T cells, and 10—25%
CD8@ human T cells were observed in the spleen cells from SCID
PBL/hu mice. Thirty to 60% CD3@ or CD8@ human T cells were also
shown in PECs from those mice. (e) CD84 human CTh cell lines
capable of exerting the cytolytic activity against human tumor cells
(CESS) and living for more than 10 weeks were established from the
spleen cells of the SCID-PBLIhu mice stimulated with CESS cells and
Adex IL-6 by in vitro re-stimulation with CESS cells (data not
shown). (I) Walker and Gallagher (34) described the antibodies
against human ovarian cancer cells that play an important role in the
successful
active
immunotherapy
against
those
human
tumors
in
human PBL-reconstituted SCID mice. However, antibodies against
1340
ll@-6 GENE ThERAPY
FOR HUMAN CANCERS
½oof CTL
(A) (%)
activity
administration
eftector
@
IN 5CID-PBLJhu
autologous
MICE
induced
by Adex
IL-6.
However,
the systemic
of rIL-6 did not induce either human CTLs against
human cancer cells in the SCID-PBUhu
mice immunized
—spleen withautologous
human
tumors
ortheinvivoantitumor
effectinthe
= PEC SCID-PBL/hu
miceinjectedwith allogeneic
CESS,autologous
20
GC629,RC422,or GC408tumorcells(datanot shown).Therefore,in
.@
@
the present study, we have established for the first time the new model
that was capable of analyzing the in vivo anti-human tumor effect and
the in vivo induction of human CTLs against autologous human
tumors, using both SCID-PBLIhu mice and Adex IL-6.
10@
I
target
—
GC629
-
—
GC814
RC422
—
K562
—
KY51O
Murakami
McCune
et a!.
(35)
have
reported
the
establishment
of
SCID-hu
mice by the transplantation of human fetal thymus and liver cells into
SCID mice. However,by using this SCID-humodel,the analysisof
the in vivo induction of tumor-specific human CTLs and the human
antitumor effect might be difficult. In such circumstances, the exper
imental model using SCID-PBIJhu mice might be useful for eluci
dating a human antitumor effect (36). The antitumor property of
recombinant human IL-7 on a human tumor was evaluated by grafting
allogeneic human colon carcinoma into SCID mice and then treating
the mice with IL-7 and adoptively transferred human peripheral blood
T cells (37). The in vitro activated human NK-enriched population
target
GC629
RC422
GC408
K562
Katol
from PBLs reduced the lung metastatic nodules count of human
CESS
melanoma cells in vivo following i.v. injection into SCID mice and
(C) (%)
20
prolonged survival when coinjected with IL-2. In addition, signifi
I
5
‘0
target
0C408
RC422
1(562
Fig. 4. The specific cytolytic activity of human cytotoxic cells in vivo generated in
SCID-PBL/hu mice against autologous human cancer cells. 2.5 X l0@PBLS from patients
with gastric cancer or a patient with rectal cancer were administered i.p. into the SCID
mice on day 0. These mice were injected with autologous human cancer cells on day 2 and
thenweretreatedwithAdexIL-bon day3. Thecytolyticactivityin the spleencellsand
PECsfromtheseSCID-PBL/hu
mice(5 miceineachgroup)24daysaftertheinjectionof
humanPBLSwas assessed againstseveralkindsof humancancercells. Columns,meanof
cytotoxicactivityin eachgroup;bars,SD.A,SCID-PBLIhu
miceweregivenPBLSfrom
a patientwithgastriccancer(GC629),andautologous0C629cancercellsobtainedfrom
thiscelllineweretreatedwithAdexIL-6Invivo.l'he cytotoxicactivityinthespleencells
4 or inthePECs(0) wasassessedagainst(1C629,RC422humanrectalcancercells,
cantly increased antitumor activity against human melanoma in vivo
was also observed in mice receiving IL-2 and IL- 12 in comparison
with those receiving IL-2 only (38). The antitumor activity was also
evaluated in the SCID mouse model by the administration of bone
marrow cells treated with cytokine-activated cytotoxic cells from
patients with lymphoma (39), human lymphokine activated killer cells
with antibodies to tumor associate antigens (40), anti-CD3 antibody
activated human PBLs (41), or the adoptive transfer of human CTL
lines generated in vitro against human melanoma into SCID mice
(42).
However, in these models using SCID mice, the in vivo induction
of human CTLs against autologous human tumors from precursor T
cells has not yet been elucidated. To our knowledge, the present study
is the first report
to show that an anti-human
tumor
effect
could
be
exerted by the induction of human tumor-specific CTLs into in vivo
GCSl4 human gastric cancer cells, K562, Murakami B biastoid cells. and KY51O human
esophageal cancer cells. The cytolytic activity in the spleen cells is shown at an E:T ratio
of 40:1andin the PECsat an E:Tratioof 5:1.B, SCID-PBL/humicegivenPBLSfrom
a patientwithrectalcancer(RC422)andautologousRC422cancercells obtainedfromthe
RC422in vitrocelllineweretreatedwithAdexIL-6in vivo.Thecytotoxicactivityin the
spleen cells (W or in the PECs(O)was
assessed against RC422, 0C629, GC408, Kato-Ill
humangastriccancer,CESS,andK562cells.Thecytolyticactivityin the spleencellsis
shown atan E:Tratio of 12:1and in the PECs atan E:Tratio of2:l. C, SCID.PBLIhu mice
given PBLS from a patient with gastric cancer (0C408) and autologous 0C408 cancer
cellsobtainedfromthe0C408 in vitrocelllineweretreatedwithAdexIL-6in vivo.The
cytotoxic activity in the spleen cells (R) at an E:T ratio of 20:1 and in the PECs (0) at
an E:Tratioof 20:1wasassessedagainstGC408,RC422,and K562cells.
CESS, 0C629, RC422, or 0C408 tumors in the sera of SCID-PBIJhu
mice bearing these tumors were not detected in the present study (data
not shown)even when the mice were treatedwith Adex IL-6, by the
assay of the complement-dependent
cytotoxicity, or by the method
describedby Walkerand Gallagher(34). Augmentationof antibody
AlA
dependent cell-mediated cytotoxicity against these human tumors was
@
B
not observed upon the addition of the sera (data not shown).
hun@
@:
The systemicadministrationof humanrIL-6was foundto augment gastric
cancer(0C629)
wereadministered
intotheSCIDmice(10mice).Onday2after
@
the generation of alloantigen-specific
.
.
.
.
CD3
human CTLs in the
.
injectionof PBLS,0C629autologous
humangastriccancercellsobtainedfromthein
vitro
cell
line
were
inoculated
i.p. into
the
SCID-PBL/hu
mice.
These
mice
were
treated
SCID-PBLIhumice lmmunlzedwlth CESS human tumor cell lines withAdexIL-6(5 mice;A)orAdexSW(5mice;B).Onday24,thesemicewere
(12,26),eventhoughtheaugmentation
of CTLsbyrIL-6wasabout sacrificed
andtheengraftinent
ofhuman
gastric
cancer
GC629
wasinvestigated.
1341
IL-6 GENETHERAPYFOR HUMANCANCERSIN 5CID-PBIJhuMICE
SCID-PBLIhuin the present study mice model. CTLs generatedby
treatment with Adex IL-6 in the SCID-PBLIhu mice given PBLS from
a patient with GC629 gastric cancer (expressing the MAGE-3 gene)
showed little cytotoxic activity against many kinds of gastric and
colorectal
adenocarcinoma
cells or several
kinds of esophageal
squa
mous cancer cells, thus suggesting that these CTLs might not recog
nize shared antigens such as the MAGE-3 antigen. To clearly prove
the tumor specificity of CTLs, SCID-PBLIhu mice should be given
PBLs from a patient with double cancers and injected with one of
these cancer cells, and then cytotoxic activity against other cancer
cells should be assessed. We have had no opportunity to study such a
patient with double cancer. Therefore, these SCID-PBLIhu mice
might provide a useful tool for the establishment of cytokine gene
therapy against cancer in humans.
SCID-PBL/humice given syngeneic PBLs from a lung cancer
patient were injected i.p. with lung cancer cells transfected with IL-6
gene.5 As a result, in vivo cytotoxic activity was generated. In addi
tion, SCID-PBL/hu mice treated with PBLs from a cancer patient and
JL-6 gene-transferred cancer cells showed a prolonged survival time.5
Recently, IL-6 transgenic SCID mice have been established by Y.
Osugi.6 IL-6 transgenic SCID-PBLJhu mice treated with PBLs from a
lung cancer patient and syngeneic lung cancer cells both exhibited a
strong cytotoxic activity (12). Therefore, IL-6 transgemc SCID mice
may provide a useful model for the rapid analysis of anti-human
tumor immunity without using gene-transfected tumor cells.5
However, the findings regarding the engraftment of human PBLs
into SCID mice have varied depending on the investigators; some
report more success than others (41). This may be due to the human
PBL graft rejection mediated by NK cells ofthe SCID mouse (41, 43).
In such situations, human growth hormone promotes the engraftment
of murine or human T cells in SCID mice (44). It has been reported
that the activation of human PBLs with anti-CD3 antibody and the
subsequent
administration
of human
rIL-2 after cell transfer
provides
a means for optimizing human T cell engraftment in SCID mice (41).
Barry et al. (43) have shown that the irradiation of SCID mice before
the engraftment or treatment of SCID mice with anti-asialo GM-i
antiserum
throughout
the engraftment
improved
the survival
of human
thymic grafts by the abrogation of NK cell activity and/or macrophage
activity.
In the present SCID-PBLIhu mouse model study, spleen cells and
PECs contained monocytes/macrophages that were detectable by anti
human monocyte monoclonal antibodies, which thus might contribute
to the capability of the in vivo induction of human CTLs. Irradiation
of SCID mice and the use of matrigel may promote the engraftment of
human macrophages from human PBLS into spleen cells and PECs.
Low-dosage irradiation played an important role in the engraitment of
human spleen cells in the hu-Spl-SCID mice (SCID mice prepared
human spleen cells) as reported by Alegre et a!. (45). In their SCID
model, human T cells retained their proliferative responses to mito
gens and to alloantigens when tested 3 weeks after engraitment into
SCID mice. Mice engrafted with unprimed human splenic T cells
rejected human allografts. Sandhu et a!. (46) also described a SCID
PBLIhu
model
in which
recipient
mice were treated
with irradiation
and anti-asialo GM-i antiserum. In the model, high levels of human
leukocyte engraftment were observed and primary human antibody
responses to antigens were obtained. Because the degree of human
CD4@ proliferation in the spleen cells from these SCID-PBLIhu mice
increases along with human KLH-specific antibody titers, it was
suggested that human antigen presenting cells, B cells, and helper T
cells are functional in the SCID mouse environment.
In contrast, it has been reported that human mature T cells in
SCID-PBL/humice were anergic in vitro. Tary-Lehmanet al. (47)
discussed
that freshly
isolated
cells from SCID-PBLIhu
mice lacked
any in vitro proliferative response to anti-CD3 antibody. The differ
ence between these results and our findings might be due to these
components of our study: (a) the use of matrigel, (b) the stimulation
of IL-6, (c) the period of sacrifice after the administration of human
PBLs into SCID mice, (d) the pretreatment with irradiation before the
administration of PBLs, and (e) the in vivo stimulation of human
tumor cells after the administration of PBLs. Tary-Lehman did not
inject antigens in vivo into SCID mice after the injection of human
PBLs.
The production of IL-6 in SCID mice by the injection of Adex IL-6
may also promote
the engraftment
of human
T cells and the in vivo
generation of CTLs, as reported in IL-3 transgenic SCUDmice. In the
IL-3 transgenic SCID mice, the engra.ftment of bone marrow cells was
augmented (48). In fact, using IL-6 transgenic SCID mice, an increase
in the number of CD3@ human T cells and CD8@ human T cells in
spleen
cells and PECs from IL-6 transgenic
SCID-PBLIhu
mice was
observed (data not shown).
Furthermore, the in vivo injection of the 1L-2 gene using an ade
novirus
vector could induce
the regression
of tumors
(49). Therefore,
using the administration of the IL-6 and IL-2 genes, significant anti
tumor immune responses were induced as shown using rIL-2 and
rIL-6 (1). The problem with using adenovirus vector is that repetitive
injections induce the generation of neutralizing antibody against the
protein of the adenovirus vector. A strategy for overcoming this
problem has been reported recently (50). When IFN-'y or rIL-12 was
administered with the recombinant adenovirus vector, the formation
of neutralizing antibody to adenovirus was blocked. The infiltration of
inflammatory cells was observed slightly in the liver from SCID
PBL/hu mice treated with Adex IL-6 by a microscopichistological
examination.Onthe otherhand,few inflammatorycellsweredetected
in the liver of SCID-PBL/hu mice treated with Adex SW (data not
shown). However, no histological changes were observed in other
organs in the peritoneal cavity, such as the kidneys and intestine in the
SCID-PBL/hu mice treated with Adex IL-6.
In the present study, the in vivo administration of the IL-6 gene using
the adenovirus vector into SCID-PBIJhu mice mediated an anti-human
tumor effect via the induction and differentiation of human CTLs from
precursors of human T cells. By using an adenovirus vector, many
cytokine genes can thus be directly delivered in vivo without the need to
manipulate the genes ex vivo. Therefore, the experimental model using
SCJD-PBL1hu mice and cytokine genes delivered by an adenovirus
vector might be a useful tool for elucidating
the anti-human
tumor effect
of gene therapy using various kinds of cytokine genes.
ACKNOWLEDGMENTS
We thank Hiroko Tomonaga,Junko Takano, and Ryoko Nyuta for their
excellent technical assistance. We are grateful to Prof. Izumu Saito and Dr.
Shizuko Harada (Laboratory of Molecular Genetics, University of Tokyo) for
helpful discussions and for the generous gift of adenovirus vectors. We are
grateful to Associate Prof. Yasuji Yoshikawa (Department of Pathology,
Medical Institute of Bioregulation, Kyushu University) for help with the
microscopic pathological examination of SCID mice treated with Adex IL-6
and human gastric and rectal cancer cells in the tumors and metastatic lesions
in the SC11)mice. We thank Prof. Shizuo Akira (Department of Immunology,
S M. Okada. Y. Ohsugi, K. Nakahara, M. Kitahara, S. Akira, T. Nomura, T. Kishimoto,
Hyogo College of Medicine) for helpful discussions. Finally, we thank Yo
and T. Suzuki. The in vivo anti-human tumor effect using lung cancer cells transfected
shikazu Watanabe (Kyushu University) for his technical support in maintain
with IL-6 cDNA and IL-6 transgenic SC@ mice, manuscript in preparation.
6 Personal communication.
ing the animals used in this study.
1342
IL-6 GENE THERAPY FOR HUMAN CANCERS IN SCID.PBIJhu MICE
transfected with IL-6 cDNA and SCID-hu mice via the induction of cytotoxic T cells.
REFERENCES
1. Kitahara, M., Kishimoto, S., Hirano, T., Kishimoto, T., and Okada, M. The in vivo
anti-tumor effect of human recombinant interleukin-6. Jpn. J. Cancer Res., 81:
1032—1038,
1990.
2. Dranoff, G., and Mulligan, R. C. Gene transfer as cancer therapy. Adv. Immunol., 58:
417—454, 1995.
3. Pardoll, D. M. Paracrine cytokine adjuvants in cancer immunotherapy. Amu. Rev.
hnmunol.,13: 399—415,
1995.
4. Okada,M., Klimpei,G. K., Kuppers,R. C., and Henny,C. S. The differentiationof
cytotoxic T cells in vitro. I. Amplifying factor(s) in the primary response is LytI
cell-dependent.J. Immunol.,122: 2527—2533,
1979.
5. Okada, M., Yoshimura, N., Kaieda, T., Yamamura, Y., and Kishimoto, T. Establish
ment and characterization of human T hybrid cells secreting immunoregulatory
molecules. Proc. Natl. Aced. Sci. USA, 78: 7717—7721,
1981.
6. Okada,M., Yoshimura,N., Ichimori,Y., Kishimoto.S., Nakai, S., Nishino, N., and
Kishimoto,T. Immunologicand molecularcharacterization
of T cell derivedI cell
activating factor. J. Immunol., 136: 1288—1294,1986.
7. Okada, M., Kitahara, M.. Kishimoto, S., Matsuda, T., Hirano, T., and Kishimoto, 1.
IL-6IBSF-2functionas a killer helperfactorin the in vitro inductionof cytotoxic T
cells.J. Immunol.,141: 1543—1549,
1988.
M., Tsuji. K., Ueo, H., Barnard,F. G., and Akiyoshi, T. Expressionof MAGEgenes
in human colorectal carcinoma. Ann. Surg., 224: 183—188,1996.
29. lnoue, H., Mon. M., Li, J., Mimori. K., Honda, M.. Nakashima, H., Mafune, K..
Tanaka, Y., and Akiyoshi, T. Human esophageal carcinomas frequently express the
tumor-rejection antigens of MAGE genes. mt. J. Cancer, 63: 523—526,1995.
30. Kanegae, Y., Makimura, M., and Saito. I. Simple and efficient method for purification
of infectious recombinant adenovirus. Jpn. J. Med. Sci. Biol., 47: 157—166,1994.
31. Yoshida, M., Sakuma, J., Hayashi, S., Abe, K., Saito, I., Harada, S., Sakatani, M.,
Yamamoto, S., Matsumoto, N., Kaneda, Y., and Kishimoto. 1. A histologically
distinctive interstitial pneumonia induced by overexpression of interleukin 6. trans
forming growth factor @l,
or platelet-derived growth factor B gene. Proc. NatI. Aced.
Sd. USA, 92: 9570—9574,1995.
32. Miyake, S., Makimura, M.. Kanegae, Y., Harada, S., Sato, Y., Takamori, K., Tokuda,
C., and Saito, I. Efficient generationof recombinantadenovirususing adenovirus
DNA-terminal
8. Kaieda, T., Okada, M., Yoshimura, N., Kishimoto, S., Yamamura, Y., and Kishimoto,
T. A human helper T cell clone secreting both killer helper factor(s) and T cell
replacingfactor(s).J. Immunol.,129: 46—51,1982.
9. Okada,M.,Sakaguch@
N., Yoshimura,N., Hara,H.,Shimizu,K.,Yoshida,N., Yoshizaki,
K., Kishimoto,S., Yamamura,Y., and Kishimoto,T. B cell growthfactorand B cell
differentiationfactor from human T hybtidomas. Two distinct kinds of B cell growth
factorandtheirsynergismin B cell proliferation.J. Exp.Med., 157: 583-590, 1983.
10. Mule, j. j., Mcintosh,J. K., Jablons,D. M., andRosenberg,S. A. Antitumoractivity
of recombinant interleukin 6 in mice. J. Exp. Med., 171: 629—636,1990.
I 1. Mui& J. J., Custer, M. C., Travis, W. D., and Rosenberg, S. A. Cellular mechanisms
of the antitumor activity of recombinant IL-6 in mice. J. Immunol., /48: 2622—2629,
1992.
12. Okada, M., and Kishimoto, T. The potential application and limitation of cytokine/
growth factor manipulation in cancer therapy. In: L Pusztai. C. E. Lewis, and E. Yap
(eds.), Cell Proliferation in Cancer, Regulatory Mechanisms of Neoplastic Cell
Growth, pp. 218-244. Oxford, United Kingdom: Oxford University Press, 1996.
13. Porgador, A., Tzehoval, E., Katz, A., Vadai, E., Revel, M., Feldman, M.. and
Eisenbach,L. Interleukin6 gene transfectioninto Lewis lung carcinomatumorcells
suppresses the malignant phenotype and confers immunotherapeutic competence
againstparentalmetastaticcells.CancerRes.,52: 3679—3686,
1992.
14. Mullen. C. A., Conic, M. M., Levy, A. T., Stetler-Stevenson,W. G., Liotta, L. A.,
Brandt, S., and Blaese, R. M. Fibrosarcoma cells transduced with IL-b gene exhibit
reduced tumorigenicity, increased immunogenicity, and decreased metastatic poten
tial. Cancer Res., 52: 6020—6024,1992.
15. Dougherty,0. J., Thacker,J. D., Lavey. R. S., Belldegrun,A., and McBride,W. H.
Inhibitoryeffectof locallyproducedandexogenousinterleukin-6on tumorgrowthin
vivo. Cancer Immunol. Immunother., 38: 339-345.
Adv. Allergy Immunol., 1: 13—31,1992.
27. Li, J., Yang, Y., Fujie. T., Baba, K., Ueo, H., Mon. M., and Akiyoshi, T. Expression
of BAGE, GAGE, and MAGE genes in human gastric carcinoma. Clin. Cancer Res.,
2: 1619—1625,
1996.
28. Mori, M., Inoue, H., Mimori, K., Shibuta, K., Baba, K., Nakashima, H., Haraguchi.
1994.
16. Fearon,E. R., Pardoll,D. M., hays, T., Golumbek,P., Levitsky,H. I., Simons,J. W.,
Karasuyama, H., Vogelstein, B., and Frost, P. Interleukin-2 production by tumor cells
bypasses T helper function in the generation of an antitumor response. Cell, 60:
397—403, 1990.
17. Hock, H., Drush, M., Diamantstein, T., and Blankenstein, T. Interleukin-7 induces
protein complex and a cosmid bearing the full-length virus genome.
Proc. NatI. Acad. Sci. USA, 93: 1320—1324,1996.
33. Peto, R.. Pike, M.. Armitage. P., Bresiow, N.. Cox, D.. Howard, S., Mantel, N.,
McPherson, K., Peto, J., and Smith. P. Design and analysis of randomized clinical
trials requiring prolonged observation of each patient. II. Analysis and examples.
Br. J. Cancer, 35: 1—39,
1977.
34. Walker, W., and Gallagher, G. The development of a novel immunotherapy model of
human ovarian cancer in human PBL-severe combined immunodeficient (SCID)
mice. Clin. Exp. Immunol., 101: 494—501, 1995.
35. McCune. J. M., Namikawa, R.. Kaneshima, H., Shultz, L. D., Lieberman, M., and
Weissman, I. L. The SCID-hu mice. Murine model for the analysis of human
hematolymphoid differentiation and function. Science (Washington DC). 241: 1632—
1639, 1988.
36. Mosier. D. E., Gulzia, R. J., Baird, S. M., and Wilson, D. B. Transfer of a functional
human immune system to mice with severe combined immunodeficiency. Nature
(Lond.), 335: 256—259, 1988.
37. Murphy, W. J., Back, T. C.. Conlon, K. C., Komschlies, K. L., Ortaldo, J. R., Sayers,
T. J., Wiltrout, R. H., and Longo, D. L. Antitumor effects of interleukin-7 and
adoptive immunotherapy on human colon carcinoma xenografts. J. Clin. Invest., 92:
1918—1924,1993.
38. Hill. L. L., Perussia, B., McCune, P., and Korngold, R. Effect on human natural killer
cells on the metastatic growth of human melanoma xenografts in mice with severe
combined immunodeficiency. Cancer Res., 54: 763—770,1994.
39. Schmidt-Wolf, I. G. H., Negrin, R. S., Kiem, H-P., Blume, K. G., and Weissman, I. L.
Use of SCID mouse/human lymphoma model to evaluate cytokine-induced killer cells
with potent antitumor cell activity. J. Exp. Med., 174: 139—149,1991.
40. Takahashi, H.. Nakada, 1., and Puisieux, I. Inhibition of human colon cancer growth
by antibody-directed human LAK cells in SCID mice. Science (Washington DC),
259: 1460—1463,1993.
41. Murphy, W. J., Conlon, K. C., Sayers, T. J., Wiltrout, R. H., Back, T. C., Ortaldo,
J. R., and Longo, D. L. Engraftment and activity of anti-CD3-activated human
peripheralblood lymphocytes transferredinto mice with severe combined immune
deficiency. J. Immunol., 150: 3634—3642, 1993.
42. Sabzevari, H.. and Reisfeld, R. A. Human cytotoxic T-cells suppress the growth of
CD4@T cell-dependent tumor rejection. J. Exp. Med., 174: 1291—1298,
1991.
18. Asher, A. L., MUle, J. J., Kasid, A., Restifo, N. P., Salo, J. C., Reichert, C. M., Jaffe,
G.. Fendly, B., Kriegler, M.. and Rosenberg, S. A. Murine tumor cells transduced
with gene for tumor necrosis factor-alpha. Evidence for paracrine immune effects of
tumor necrosis factor against tumors. J. Immunol., 146: 3227—3234, 1991.
19. Watanabe, Y., Kuribayashi, K., Miyatake, S., Nishihara, K.. Nakayama, E..
Taniyama,T., and Sakata,T. Exogeneousexpressionof mouseinterferongamma
spontaneous melanoma metastases in SCIDIhu mice. Cancer Res.. 53: 4933—4937.
1993.
43. Barry, T. S., Jones, D. M., Richter, C. B., and Haynes, B. F. Successful engraftment
of human postnatal tymus in severe combined immune deficient (SCID) mice:
differential engraftment of thymic components with irradiation versus anti-asialo
cDNA in mouse neuroblastomaCl 300 cells results in reduced tumorgenicity by
GM-l immunosuppressiveregimens.J. Exp. Med.. /73: 167—180.
1991.
1989.
engraftment of murine or human T cells in severe combined immunodeflcient mice.
Proc. NatI. Acad. Sci. USA, 89: 4481—4485,1992.
augmentedanti-tumorimmunity.Proc. Nail. Acad. Sci. USA, 86: 9456—9460, 44. Murphy, W. J.. Durum, S. K., and Longo, D. Human growth hormone promotes
20. Colombo, M. P., Ferrari,G., Stoppacciam, A., Parenza, M., ROdO1fO,
M.. Mavillo. F., and
45. Alegre, M-L., Peterson, L. J., Jeyarajah, D. R., Weiser, M., Bluestone, J. A., and
Panniani, 0. Granulocytecolony-stimulatingfactorgene transfersuppressestumorgenic
ity ofa murine adenocarcinomain vivo.J. Exp. Med.. 173: 889—897,1991.
Thistlethwaite,J. R. Severe combined immunodeficientmice engraftedwith human
21. Tepper, R. I.. Puttengale, P. K., and Leder, P. Murine interieukin-4 displays potent
anti-tumor activity in vivo. Cell, 57: 503—512,1989.
22. Golumbek, P. T., Lazenby, A. 3., Levitsky, H. J.. Jaffee. L. M., and Karasuyama, H.
Treatment of established renal cancer by tumor cells engineered to secrete interieu
kin-4. Science (Washington DC), 254: 7 13—7
16, 1991.
splenocytes have functional human T cells and reject human allografts. J. Immunol.,
/53: 2738—2749, 1994.
46. Sandhu, J.. Shpitz, B., Gallinger, S.. and Hozumi, N. Human primary immune
response in SCID mice engrafted with human peripheral blood lymphocytes. J. Im
munol., 152: 3806—3813, 1994.
47. Tary-Lehmann, M.. Lehmann, P. V., Schols, D., Ronvarolo. M. G., and Saxon, A.
23. Dranoff, 0., Jaffee, E., Lazenby, A., Golumbek, P., Levitsky, H., Brose, K., Jackson,
V., Hamada,H., Pardoll,D., and Mulligan,R. C. Vaccinationwith irradiatedtumor
cells engineeredto secretemurinegranulocyte-macrophage
colony-stimulatingfactor
Anti-SCID mouse reactivityshapes the humanCD4@T cell repertoirein hu-PBL
SCID chimeras.J. Exp. Med., 180: 1817—1827,
1994.
48. Bock, T. A., Orlic, D., Dunbar, C. E., Broxmeyer. H. E.. and Bodine, D. M. Improved
engraftment of human hematopoietic cells in severe combined immunodeficient
stimulates potent, specific, and long-lasting anti-tumor immunity. Proc. NatI. Acad.
Sci. USA, 90: 3539—3543, 1993.
24. Saito, S., Banneiji, R., Gansbacher, B., Rosenthal, F. M., Romanenko, P., Heston.
(SCID)micecarryinghumancytokinetransgenes.J. Exp.Med..182: 2037—2043,
w. D.W.,Fair,W.R..andOilboa,E.Immunotherapy
ofbladdercancer
withcytokine
1995.
49. Addison, C. L., Braciak, T., Ralston, R., Muller, W. J., Gauldie, J., and Graham. F. L.
gene-modified tumor vaccines. Cancer Res., 54: 3516—3520, 1994.
25. Tahara, H., Zitvogel, L., Storkus, W. J., Zeh, H. J., III, McKinney, T. G., Schreiber,
Intratumoral injection of an adenovirus expressing interleukin 2 induces regression
R. D., Gubler, U., Robbins, P. D., and Lotze, M. T. Effective eradication of
and immunity in a murine breast cancer model. Proc. NatI. Acad. Sci. USA, 92:
8522—8526,1995.
established murine tumors with IL-12 gene therapy using a polycistronic retroviral
vector.J. Immunol.,154:6466—6474,
1995.
26. Okada,M., lUtahara,M., Akira,S., Shimamoto,T., Kuribayashi,K., Watanabe,Y.,
and Kishimoto, T. The in vivoantitumor effect of IL-6 and KHF: by using tumor cells
1343
50. Yang. Y.. Trinctiieri. 0., and Wilson. J. M. Recombinant IL-l2 prevents formation of
blocking IgA antibodies to recombinant adenovirus and allows repeated gene therapy
to mouse lung. Nat. Med., 1: 890—893,1995.