| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
*
Section of Molecular Oncology, First Department of Surgery, Okayama University Medical School, Okayama, Japan;
First Department of Surgery, Shiga University of Medical Science, Shiga, Japan; and
Section of Thoracic Molecular Oncology, Department of Thoracic and Cardiovascular Surgery, The University of Texas M. D. Anderson Cancer Center, Houston, TX 77030
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CD95 ligand (CD95L, also called FasL/APO-1L) is a type II integral membrane protein of the TNF family that transduces an apoptotic death signal by binding to its receptor CD95, which is a type I transmembrane protein of the TNFR family. The CD95-CD95L system has been implicated in the activated T cell- or NK cell-mediated cytotoxicity, some pathologic tissue damage, and the regulation of lymphocyte homeostasis (3). CD95L is expressed in a wide range of normal tissues, such as the spleen, testis, uterus, large and small intestine, and eye as well as some malignant tumor cells, including human melanoma (4), hepatocellular carcinoma (5), colon cancer (6), and lung cancer (7), which has been proposed to contribute to their immune-privileged status by the elimination of infiltrating immune cells (8, 9). However, more recent studies have demonstrated another property of CD95L in different experimental models. Expression of CD95L on myotubes, pancreatic islet cells, or heart grafts induces neutrophilic inflammation, resulting in accelerated graft rejection in a T cell- or B cell-independent manner (10, 11, 12, 13, 14). Furthermore, locally produced CD95L by transfection with CD95L cDNA causes neutrophil-mediated rejection of tumor cells when s.c. injected (15). Membrane-bound CD95L can be proteolytically cleaved by metalloproteinases, thereby producing a soluble and active form of CD95L (16). It has been also reported that human soluble rCD95L is a potent chemoattractant for human and mouse polymorphonuclear neutrophils (PMNs) (17). These observations suggest that CD95L can act directly on neutrophils to induce their recruitment, indicating that CD95L may play a proinflammatory role in a specific microenvironment.
Numerous studies have shown that p53 mediates its effects by modulating the transcription of a number of cellular target genes. We previously reported the p53-mediated up-regulation of CD95 as well as CD95L in human cancer cells, although little is known about the molecular machinery underlying these processes (18). These data raised the hypothesis that increased CD95L expression by p53 might induce neutrophilic infiltration, which could accelerate the destruction of neighboring tumor cells that were not transduced with wt-p53. In the present study, we show that significant CD95L accumulation and neutrophil infiltration take place in established human colorectal tumors after intratumoral injection of the adenovirus expressing the wt-p53 gene. Thus, the p53-mediated up-regulation of CD95L is likely to provide further insights for the molecular basis of the bystander effect in p53 gene therapy.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The human colorectal carcinoma cell lines LoVo that contain the wt-p53 and SW620 that exhibit a homogygous p53 gene mutation were maintained in monolayer cultures in 75-cm2 tissue culture flasks. These cell lines were routinely propagated in RPMI 1640 medium supplemented with 10% FCS, 25 mM HEPES, 100 U/ml penicillin, and 100 mg/ml streptomycin. The transformed embryonic kidney cell line 293 was grown in DMEM (Life Technologies, Grand Island, NY) with high glucose (4.5 g/L) supplemented with 10% FCS, 100 U/ml penicillin, and 100 mg/ml streptomycin.
Recombinant adenoviruses
The recombinant adenovirus vector expressing human wt-p53 cDNA was previously constructed and characterized (19). The resultant virus was named Ad5CMVp53. The E1A-deleted adenovirus vector lacking a cDNA insert (dl312) was used as a control vector. The viral stocks were quantified by a plaque-forming assay using 293 cells and stored at -80°C.
Animal experiments
Animal experiments were conducted in accordance with the institutional animal care and use regulations. Four-week-old female BALB/c nu/nu mice were purchased from Shizuoka Laboratory Animal Center (Shizuoka, Japan). LoVo and SW620 suspensions (5 x 106 cells/100 µl) were s.c. inoculated into the dorsi of BALB/c nu/nu mice. When palpable nodules were established, tumors were injected daily with PBS (100 µl), dl312 (1 x 109 PFU/100 µl), and Ad5CMVp53 (1 x 109 PFU/100 µl) for 3 consecutive days. The tumor growth was assessed by measuring perpendicular diameters with calipers. Tumor volume was calculated from the largest (a) and smallest (b) diameter with the formula 0.5a x b2. For depleting neutrophils, mice were administered 200 µg of anti-Gr-1 mAb (RB6-8C5; PharMingen, San Diego, CA) i.p. 1 day before the first injection of Ad5CMVp53 and on days 2 and 5 after the first injection. Control mice received i.p. administration of isotype-matched rat IgG2b (PharMingen).
Histological examinations
For immunohistochemistry, tumor tissues were fixed in 20% neutral buffered Formalin. Paraffin-embedded sections (4 µm thick) were deparaffinized in xylene and rehydrated through graded alcohols into PBS. Heat-induced epitope retrieval was achieved by immersion of slides in 10 mM citrate buffer (pH 6) and heating for 20 min in the autoclave. Endogenous peroxidase was blocked by 10-min incubation with 3% hydrogen peroxide in methanol. To prevent nonspecific binding, the sections were incubated in 0.25% casein (Dako, Carpenteria, CA) for 5 min at room temperature. Either mouse anti-human CD95L (Transduction Laboratories, Lexington, KY) or mouse anti-human p53 (DO-7; PharMingen) and goat anti-mouse Ig conjugated to peroxidase-labeled dextran polymer (EnVision+; Dako) were mixed together and incubated for 45 min at room temperature to form Ab complexes. Then, inactivated normal mouse serum (Dako) was added to the mixture and incubated for 45 min at room temperature. Sections were then incubated with the mixture for 60 min at room temperature, treated with diaminobenzidine solution, and counterstained with hematoxylin solution. Tumor tissues were also processed for the granulocyte-specific staining that employs naphthol AS-D-chloroacetate as a substrate for esterase. Granulocytes appear to be red with this staining.
Isolation of PMNs
Thioglycolate-elicited PMNs were harvested from peritoneal exudates 4 h after an i.p. injection of 3 ml 3% thioglycolate (Difco, Detroit, MI). The purity of PMN was >85%, as assessed by Diff-Quick staining (International Reagents, Kobe, Japan) of cytospin preparations under light microscopy. PMNs were suspended in PBS containing 1 mM CaCl2 and 1 mM MgCl2.
Chemotaxis assay
PMN chemotaxis was quantified using a modification of the Boyden chamber technique (20). A cell suspension containing 4 x 106 cells/ml (the total cell number loaded per well was adjusted to give equal numbers of PMNs) in PBS supplemented with 1 mM CaCl2, 1 mM MgCl2, and 1 mg/ml BSA (Fraction V; Sigma, St. Louis, MO) was placed in the top wells of a 48-well microchemotaxis chamber (Neuro Probe, Bethesda, MD). A 3-µm pore-size polyvinylpyrolidone-free polycarbonate filter (Neuro Probe) separated the cells from lower wells containing culture supernatants. Recombinant human IL-8 (Genzyme, Cambridge, MA) and human CD95L (Transduction Laboratories) were used as a positive control. After incubation for 60 min at 37°C in a 5% CO2 humidified atmosphere, the filter was removed, gently scraped off the upper face, fixed in 100% methanol, and subsequently stained with Diff-Quick stain solution. Chemotactic activity was estimated by counting the total number of PMNs migrating to the lower face of the filter in five random high power fields (x400) per well. The results were expressed as the mean number of PMNs per high power field. Each experiment was performed at least three times. For the blocking experiment, mouse anti-human CD95L (NOK-1) or control isotype-matched mouse IgG1 (PharMingen) at a concentration of 5 µg/ml was added to both upper and lower chambers.
Statistical analysis
The statistical significance of differences from control was evaluated by Student’s t test. Values of p < 0.05 were considered significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A recombinant, replication-deficient adenovirus vector carrying
human wt-p53 cDNA under control of the CMV immediate early gene
promoter (Ad5CMVp53) was employed to achieve efficient gene transfer
into human cancer cells. When 5 x 106 SW620
human colorectal cancer cells, which are homozygous for a mutation in
p53, were s.c. inoculated into nu/nu mice, palpable tumors
appeared in 100% of mice 2 wk after tumor injection. Twenty-one days
later, SW620 tumors with a diameter of 5–7 mm were treated daily with
direct intratumoral injection of the vehicle or 1 x
109 PFUs of Ad5CMVp53 or control dl312 vector for
3 consecutive days. Ad5CMVp53-injected mice showed a significant tumor
growth inhibition that started on the third day after the first
injection and continued for at least 40 days, at which time
approximately a 68% inhibition was seen relative to the mock-treated
tumors (Fig. 1
). There were no
significant differences in tumor growth among the mock-treated and
dl312-treated groups. We also observed the growth-inhibitory effect of
Ad5CMVp53 in identical experiments using LoVo human colorectal cancer
cells that contain wt-p53 (data not shown).
|
To determine the efficacy of transducing human colorectal cancer
cells in solid tumors, we assessed the presence of p53 immunoreactivity
in LoVo and SW620 tumors 24 h after the first viral injection.
Immunohistochemical evaluation revealed a detectable, widespread wt-p53
protein, which was identified by intense nuclear staining, in
Ad5CMVp53-injected LoVo and SW620 tumors (Fig. 2, A and D),
whereas dl312-injected tumors showed no overexpression of p53 protein
(Fig. 2, C and F). The geographic distribution
pattern of p53-positive cells with peripheral areas of intense staining
and a central portion of cellular infiltrates was noted in tumors
treated with Ad5CMVp53. Seven days after the first viral injection,
there was no p53 immunoreactivity in both tumors (Fig. 2, B
and E). Furthermore, we detected p53 overexpression only in
tumor cells, but not in infiltrating cells, suggeting that the cellular
infiltrates might be a secondary event following p53 gene
transduction.
|
To investigate the mechanisms for the Ad5CMVp53-mediated antitumor
effect, we histopathologically analyzed LoVo and SW620 tumors injected
with Ad5CMVp53 for the distribution of inflammatory and/or immune cells
as well as CD95L immunoreactivity. Histological analysis at 24 h
after the first injection of Ad5CMVp53 revealed massive tumor cell
death and cellular infiltrates at the central portions of the tumors in
which Ad5CMVp53 was injected (Fig. 3, A, B,
E, and F). In contrast, tumors treated with dl312
showed neither tumor cell death nor cellular infiltrates (Fig. 3, C, D,
G, and H). Intratumoral injection of
physiological saline as a control also had no significant effect (data
not shown). These results suggest that destruction of tumor cells was
associated with a massive accumulation of inflammatory cells, and that
the presence of inflammatory infiltrates might be due to transduced p53
expression rather than viral vector injection.
|
, A and D).
At higher magnification, CD95L was shown to be expressed on the
cytoplasm of tumor cells, while infiltrating cells completely lacked
staining (Fig. 4, B and E). Tumors in mice
treated with dl312 vector had no CD95L-overexpressing cells (Fig. 4, C and F). To characterize the infiltrating cells,
we next performed esterase staining, which can specifically detect
neutrophils, on the section of SW620 tumors injected with Ad5CMVp53. As
shown in Fig. 5B, PMNs were
identified morphologically as well as by esterase staining. These
results indicate that p53-mediated CD95L expression may act on
neutrophils, inducing their recruitment into solid tumors and thereby
causing in part antitumor effect of the wt-p53 gene transfer.
|
|
To determine whether CD95L could be attributed to
Ad5CMVp53-induced mouse neutrophil migration, cell migration was
evaluated by a modified Boyden chamber assay with a
nitrocellulose filter. Culture supernatants of LoVo and SW620 cells
infected with Ad5CMVp53 exhibited the chemotactic activity against
thioglycolate-elicited PMNs obtained from BALB/c nu/nu mice
in a dose-dependent manner. Supernatants of dl312-infected cells were
also chemotactic, presumably because of some factors produced by viral
infection; the magnitude of PMN migration induced by AdCMVp53-infected
cells, however, was significantly greater than that generated by
dl312-infected cells as well as by the classic chemoattractant IL-8
(Fig. 6A).
|
B). These results suggest that the wt-p53-induced CD95L is
indeed capable of stimulating neutrophil migration. Contribution of p53-induced neutrophil accumulation to the antitumor effect
To further define the antitumor effect of PMNs accumulated by
Ad5CMVp53, in vivo neutrophil depletion experiments were performed by
using anti-Gr-1 mAb (RB6-8C5) or isotype-matched control mAb
(IgG2b). The growth suppression of SW620 tumors injected with Ad5CMVp53
was partially but significantly reduced in RB6-8C5-treated mice
compared with that in mock-treated or IgG2b-treated mice (Fig. 7). Pretreatment with IgG2b had no effect
on the antitumor effect of Ad5CMVp53. These results suggest that the
wt-p53 gene transfer could not only induce the direct effect in the
individual transduced cell, but could also cause the growth suppression
of bystander, nontransduced cells via neutrophil accumulation.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Intratumoral injection of Ad5CMVp53 led to significant suppression of
the growth of human colorectal tumors s.c. transplanted in
nu/nu mice (Fig. 1). Although histologic examination
demonstrated that rapid apoptotic tumor cell death occurred after the
wt-p53 transduction (data not shown), as we previously reported in a
variety of tumor models (18, 22, 23, 24, 25), our findings that
notable cellular infiltrates were evident at the site of Ad5CMVp53
injection in both LoVo and SW620 tumors (Fig. 3) suggest that these
infiltrating cells might be involved as effector cells in the antitumor
effect. The absence of cellular infiltrates in the peripheral portion
of the tumor, in which most of the cells were still viable, also
supports this hypothesis; these infiltrates, however, were unlikely to
be cytotoxic T lymphocytes, as the antitumor effect was observed in T
cell-deficient nu/nu mice. Consistently, specific esterase
staining indicated that most of inflammatory infiltrates are murine
neutrophils (Fig. 5). Immunohistochemical analysis showed that CD95L
was detected on apototic tumor cells in the area occupied by murine
PMNs (Fig. 4). These observations suggest that Ad5CMVp53-induced CD95L
expression resulted in the accelerated infiltration of neutrophils,
thereby leading to the massive destruction of tumor tissues. This
scenario could be supported by recent studies showing that soluble
CD95L is chemotactic for murine and human PMNs (17) and
showing that tumors expressing CD95L are made to regress by accerelated
PMN infiltration (15, 26, 27).
Host immune interactions with the adenovirus have been known; the
presence of the inflammatory infiltrates, however, is not due
exclusively to the administration of adenovirus vectors, because
dl312-injected tumors showed neither CD95L overexpression nor PMN
infiltration. Other possibilities, such as inflammatory responses
induced by dead tumor cells, seem unlikely in view of the fact that a
massive cellular infiltration could be observed as early as 24 h
after the first Ad5CMVp53 injection (Fig. 3). It has been reported that
caspases released by apoptotic cells could process and activate
IL-1
, thus inducing inflammation (28); our preliminary
experiments, however, demonstrated that IL-1 could be detected
neither in supernatants nor in cell lysates of tumor cells after
Ad5CMV-p53 infection in vitro (data not shown). We reported that levels
of CD95L mRNA can be greatly elevated, reaching their maximum at
24 h after Ad5CMVp53 infection, followed by a rapid decrease
(18). Recently, it was demonstrated that soluble CD95L
acts as a direct chemoattractant against murine PMNs in vitro
(17). Indeed, our modified Boyden chamber assay showed
that supernatants obtained from Ad5CMVp53-infected tumor cells induced
murine PMN chemotaxis, which could be abolished by adding neutralizing
anti-CD95L Ab (Fig. 6). Thus, although the molecular mechanisms
governing soluble CD95L-dependent chemotactic activity are presently
unknown, Ad5CMVp53-induced CD95L secretion exhibits potent chemotactic
properties toward murine neutrophils.
The phagocytosis of apoptotic cells by macrophages has been shown to
produce human proinflammatory cytokine IL-8, which is a common
chemoattractant active on neutrophils (29). Furthermore,
in murine systems, macrophages ingesting apoptotic cells produced
macrophage-inflammatory protein-2, a murine IL-8 homologue, which was
associated with transient infiltration of neutrophils
(30). These observations support the additional hypothesis
that other chemoattractants such as IL-8 and macrophage-inflammatory
protein-2 secreted from surrounding cells including macrophages might
play a role on neutrophil infiltration in vivo. Further studies will be
necessary to confirm these possibilities. Interestingly, supernatants
obtained from dl312-infected tumor cells display some chemoattractant
activity in vitro compared with those from mock-infected tumor cells
(Fig. 6A), while we did not observe detectable PMN migration
in dl312-injected tumors in vivo (Fig. 3). Other chemotactic factors
could be produced by tumor cells treated with dl312; those levels of
expression, however, might be below the in vivo threshold, at which
stimuli for neutrophil infiltration are generated
(10).
Neutrophils are considered to be partially responsible for the
antitumor effect of Ad5CMVp53, because in vivo depletion of neutrophils
by anti-Gr-1 mAb partially inhibited the response (Fig. 7). It has
been reported that the local inflammation elicited by CD95L-expressing
tumor cells induced bystander rejection of parental tumor cells
(15), suggesting that Ad5CMVp53-induced PMN migration via
CD95L up-regulation may be a major mechanism of the bystander effect in
p53 gene therapy. However, how neutrophils exhibit cytotoxic activity
against tumor cells remains to be determined. PMNs have been reported
to interact directly with CD95L-expressing tumor cells to mediate their
destruction (15, 26). Furthermore, Chen et al.
demonstrated that CD95L-induced neutrophil cytotoxicity was dependent
on p38 mitogen-activated protein kinase function (26). In
our system, whether murine neutrophils specifically lyse the
wt-p53-expressing human tumor cells can hardly be examined because
overexpression of the wt-p53 gene itself has a direct proapoptotic
effect against tumor cells. The precise mechanism must be identified in
further studies using different model systems.
Recently, Dewey et al. reported that adenovirus-mediated conditional cytotoxic gene therapy successfully inhibited the syngenic glioma growth, but the inflammatory infiltrate also induced secondary demyelination (31). Neutrophilic inflammation induced by Ad5CMVp53 injection has the potential to affect normal surrounding tissues; no apparent histopathological changes, however, were observed at the neighboring and distant organs in our murine models (data not shown). The widespread presence of immunoreactive transgene expression throughout the brain was noted in their system (31), whereas transduced p53 as well as CD95L expression were localized at the injection sites presumably because of limited distribution of Ad5CMVp53 in our experiments. These observations have important implications for the safety evaluation of intratumoral administration of Ad5CMVp53, although we cannot rule out the possible involvement of activated T cells or NK cells at the site of Ad5CMVp53 injection when immunocompetent mice are used.
Our studies provide the first evidence that overexpression of the wt-p53 gene induces transient CD95L expression with a proinflammatory function, leading to neutrophil-mediated tumor cell destruction. We previously reported that wt-p53 gene transfer can attenuate tumor cell-induced neovascularization in vivo and suppress the growth of bystander tumor cells (32, 33). Cell-to-cell transfer of phosphorylated ganciclovir via gap junctions between the herpes simplex virus-thymidine kinase-transduced tumor cells and neighboring unmodified cells is known to play an important role in the bystander effect (31); other mechanisms such as immune reaction or blood vessel destruction, however, have been proposed to explain the bystander effect in conditional cytotoxic gene therapy. Our data presented in this work thus offer a rational basis for the mechanism of the bystander effct in p53 gene therapy for cancer.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Toshiyoshi Fujiwara, First Department of Surgery, Okayama University Medical School, 2-5-1 Shikata-cho, Okayama 700-8558, Japan.
3 Abbreviations used in this paper: wt-p53, wild-type p53; CD95L, CD95 ligand; PMN, polymorphonuclear neutrophil.
Received for publication May 25, 2000. Accepted for publication August 21, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
cells induces a granulocytic infiltration, but does not confer immune privilege upon islet allografts. Proc. Natl. Acad. Sci. USA 94:3943.
B activity and synergizes with aspirin to induce apoptosis in human colon cancer cells. Oncogene 19:726.[Medline]
release and inflammation induced by the apoptosis inducer Fas ligand. Nat. Med. 11:1287.
This article has been cited by other articles:
|
|
 |
|
  X. Yu, S. L. Harris, and A. J. Levine The Regulation of Exosome Secretion: a Novel Function of the p53 Protein. Cancer Res., May 1, 2006; 66(9): 4795 - 4801. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Fujiwara, N. Tanaka, S. Kanazawa, S. Ohtani, Y. Saijo, T. Nukiwa, K. Yoshimura, T. Sato, Y. Eto, S. Chada, et al. Multicenter Phase I Study of Repeated Intratumoral Delivery of Adenoviral p53 in Patients With Advanced Non-Small-Cell Lung Cancer J. Clin. Oncol., April 10, 2006; 24(11): 1689 - 1699. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Giroux and F. Denis CD1d-unrestricted human NKT cells release chemokines upon Fas engagement Blood, January 15, 2005; 105(2): 703 - 710. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Umeoka, T. Kawashima, S. Kagawa, F. Teraishi, M. Taki, M. Nishizaki, S. Kyo, K. Nagai, Y. Urata, N. Tanaka, et al. Visualization of Intrathoracically Disseminated Solid Tumors in Mice with Optical Imaging by Telomerase-Specific Amplification of a Transferred Green Fluorescent Protein Gene Cancer Res., September 1, 2004; 64(17): 6259 - 6265. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Murakami, N. Tokunaga, T. Waku, S. Gomi, S. Kagawa, N. Tanaka, and T. Fujiwara Antitumor Effect of Intratumoral Administration of Bone Marrow-Derived Dendritic Cells Transduced with Wild-Type p53 Gene Clin. Cancer Res., June 1, 2004; 10(11): 3871 - 3880. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Ohtani, S. Kagawa, Y. Tango, T. Umeoka, N. Tokunaga, Y. Tsunemitsu, J. A. Roth, Y. Taya, N. Tanaka, and T. Fujiwara Quantitative analysis of p53-targeted gene expression and visualization of p53 transcriptional activity following intratumoral administration of adenoviral p53 in vivo Mol. Cancer Ther., January 1, 2004; 3(1): 93 - 100. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
