| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,§,§
*
Departments of Biology, Microbiology-Immunology, and Obstetrics and Gynecology and
Lineberger Comprehensive Cancer Center, University of North Carolina, Chapel Hill, NC 27599; and
Department of Pharmacology and
§
Cancer Research Institute, Kanazawa University, Kanazawa, Japan
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(12), IFN-
(13), and IFN-
(14). The role of
chemokines in tumor etiology is less well understood: these factors
appear to play different roles in tumor growth and incidence
(15, 16, 17, 18, 19, 20, 21). Chemokines are a superfamily of small proteins
secreted primarily by leukocytes and characterized by a conserved
four-cysteine motif. Two subfamilies, the C-X-C and CC chemokines, are
classified by the spacing of the two first cysteines (22).
The CC chemokines, macrophage-inflammatory protein-1
(23) and monocyte chemoattractant protein-1, are potent
chemoattractants for T lymphocytes and monocytes (22),
whereas the C-X-C chemokines, such as IL-8, are strong chemotactic
agents for T cells, neutrophils, and basophils (22). Previously, we have found that the new anticancer drug, paclitaxel (Taxol, Bristol-Myers Squibb, New York, NY), can induce IL-8 gene expression in subsets of human ovarian cancers (24). Paclitaxel, an inhibitor of microtubule dissociation, represents a new class of antineoplastic agents and has shown efficacy in several malignant tumors. The compound has achieved specific prominence in the treatment of ovarian cancer (25, 26) because the majority of ovarian cancer patients have advanced disease (stage III or IV) at the time of diagnosis and the prognosis of these patients is poor in spite of aggressive intervention. Therefore, chemotherapy has attained a fundamental role in the therapy of ovarian cancer (27). Paclitaxel and docetaxel (Taxotere, Aventis Pharmaceuticals, Collegeville, PA) have shown significant success as single agents in treating recurrent ovarian cancer and provide viable options for patients with advanced cancers, including those that are resistant to Vinca alkaloids or anthracycline. Currently, paclitaxel, plus a platinum analogue, is considered a first-line therapy for advanced ovarian cancer and is used in relapsed or cisplatin-refractory ovarian cancer. Although the primary molecular target of paclitaxel is the microtubule, other chemotherapies, which display similar but not identical activity, are less effective in the clinic. These data suggest that paclitaxel may have additional effects on tumor cells. The earlier finding that the drug can induce IL-8 expression (24) suggests that this cytokine is involved in one of the alternative pathways that mediate the action of paclitaxel on ovarian tumor cells.
The current report examines the role of elevated IL-8 expression in a mouse model of human ovarian cell tumorigenesis. Our results indicate a striking correlation between IL-8 production, neutrophil-monocyte infiltration, and the regression of human ovarian cancer. The relevance of this finding to the therapy of human ovarian cancer is discussed.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The human ovarian cancer cell lines OVCA 420, 429, 194, and 494 have been described previously (24, 28). The cells were maintained as monolayer cultures in DMEM/F-12 medium supplemented with 5% FBS (complete media). The expression plasmids pIL-8BCMGSNeo and pBCMGSNeo have been described previously (16). Briefly, IL-8 cDNA was digested with the restriction enzyme SmaI and BamHI from pUC 19 and inserted into the Bluescript KS+ vector. The Bluescript KS+/IL-8 plasmid was digested with XhoI and NotI, and the resultant fragment was cloned into the XhoI-NotI site of BCMGS neo vector.
DNA transfection
Human ovarian cancer cells OVCA 194 and 494 cells were seeded in 100-mm plates (Falcon Plymouth, U.K.) and grown in complete media until they reached 75–80% of confluence. Subconfluent cultures were transfected by electroporation with pIL-8BCMGSNeo or BCMGS neo plasmids. After 48 h, G418 (800 µg/ml) was added to the cells; medium was changed every 3 days. Ninety-nine percent of cells in the control plate died whereas transfected cells produced colonies, which survived 3 wk of G418 treatment. G418-resistant clones were randomly selected, isolated, and expanded individually.
Northern blot
To assess the level of IL-8 expression, total RNA was isolated from stably transfected cells using the guanidinium isothiocyanate/CsCl method (29). From 3 to 5 µg of purified RNA were loaded in each lane of a denaturing agarose gel (30). Northern blotting was performed using a 32P-labeled IL-8 probe on a blot of the separated RNA as described (24).
IL-8 ELISA
To determine the release of IL-8 by transfected cells, conditioned medium was collected and IL-8 was measured as previously described (24). Briefly, a goat anti-human IL-8 Ab (R&D Systems, Minneapolis, MN) was used as the coating Ab, rabbit polyclonal anti-human IL-8 (Endogen, Cambridge, MA) as the primary Ab, and alkaline phosphatase-conjugated goat anti-rabbit Ab (Organon Teknika, Durham, NC) as the secondary reagent (31). The ELISA can accurately detect nanograms per milliliter of cytokine in the complete medium.
Animal studies
Female BALB/c-nu/nu mice, 4–6 wk old, were purchased from Taconic (Germantown, NY). They were maintained under pathogen-free conditions in the facilities of the Division of Laboratory Animal Medicine at the University of North Caroline (Chapel Hill, NC). Tumor cells were suspended in PBS at a concentration of 5 x 106/ml. For most experiments presented in the text, the total number of tumor cells injected per animal was 106. Injections (200 µl) were given s.c. in the right lower back using a 1-ml syringe. Tumor volumes were measured in cubic millimeters with a vernier caliper and calculated by the formula tumor size = a b2/2. The symbol a is the larger and b is the smaller of the two dimensions (32). All animal experiments were in accordance with guidelines provided by the UNC Institutional Animal Care and Use Committee.
Ab injection
Neutralization of endogenous IL-8 in vivo was examined by i.p. administration of 50 µg mouse anti-human IL-8 (R&D Systems) or mouse IgG1 isotype control Ab. The Abs were injected i.p. at day 0 and on every fifth day thereafter.
Histological analysis
Tissue from the site of tumor injection was fixed by immersion in 4% paraformaldehyde/70 mM phosphate buffer and subsequently embedded with paraffin. Five micron sections were stained with hematoxylin and eosin.
Immunohistochemical staining
Inoculation site tissue destined for immunohistochemistry was embedded in OTC compound (Miles, Elkhart, IN) and snap frozen in isopentane chilled to -100°C by liquid nitrogen. Sections (5 µm) were prepared and stained for NK cells (biotin-conjugated mouse anti-mly-49C, PharMingen, San Diego, CA), neutrophils (biotin-conjugated rat anti-mly-6G (Gr-1), PharMingen), macrophage (biotin-conjugated rat anti-Mac-1, PharMingen), or B cells (biotin-conjugated rat anti-B220/CD45R, PharMingen). Biotin-conjugated isotype control Abs were applied for each corresponding Ab. Slides were fixed in 100% acetone at -20°C for 5 min and then air dried. Nonspecific binding sites were blocked by treatment with 1% serum for 30 min followed by exposure to the Ab or isotype control. Sections were subsequently treated for 30 min with fluorescein-conjugated avidin (1:1000, Vector Laboratories, Burlingame, CA) and washed in PBS four times, and coverslips were applied in aqueous media (Vectashield, Vector). Microscopy was performed using an Olympus BX40 microscope.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Four human ovarian cancer cell lines 420, 429, 194, and 494 were
tested for tumorigenicity in nude mice. The IL-8 nonproducers OVCA 194
and 494 (24) formed tumors in vivo when injected s.c. into
nude mice. In contrast, tumor formation by OVCA 429 cells was very
slow, whereas animals injected with the IL-8-producing OVCA 420 cells
were devoid of tumors (Fig. 1
). No tumors
were observed in these latter animals even 3 mo after injection.
|
The dichotomy in the growth pattern of IL-8 producers and
nonproducers is of interest and suggests that IL-8 may suppress the
growth of human ovarian cancers. OVCA 420 cells have a basal level of
IL-8 expression (24). To determine whether IL-8 is a
crucial factor for tumor growth, a neutralization Ab to IL-8 was
injected (i.p.) every fifth day after cell injection for 140 days. The
injection of IL-8 Abs greatly enhanced tumor growth in all mice,
whereas an isotype-matched control Ab had no effect. (Fig. 2). These findings suggest that IL-8
inhibits tumor growth in OVCA 420 cells.
|
To further evaluate the role of IL-8 in tumorigenesis, an
expression vector containing the human IL-8 gene and the neomycin
resistance gene (neoR) was transfected
into the OVCA 194 cell line. Because this line normally lacks IL-8
expression, our hypothesis was that the growth characteristics would be
significantly altered in the presence of IL-8. After selection in G418
for 1 month, preliminary experiments were conducted using a polyclonal
population of human IL-8-transfected 194 cells. These cells were shown
to express high levels of IL-8 mRNA and secrete substantial IL-8
protein (Fig. 3, lanes 1–4,
and Table I). The level of IL-8 was
substantially higher than that produced by bulk cultures of control 194
neo cells, although the control cells did produce a faint basal level
expression of IL-8 (Fig. 3, lane 3). The IL-8-transfected
population of cells was then tested for tumor growth in nude mice. As
shown in Fig. 4, injection of
107 IL-8-transfected 194 cells grew at a
significantly slower rate than vector-transfected control cells.
|
|
|
In light of the observations with the heterogeneous population of
IL-8-transfected 194 cells, clonal populations of transfected tumor
cells, for both the OVCA194 and OVCA494 cell lines, were generated.
Twenty independent clones from each transfection were selected and
expanded. As determined by ELISA assay (data not shown), 95% of the
194 and 80% of the 494 human
(h)4 IL-8-transfected
clones secreted IL-8 protein. Several of these clones expressed
significantly elevated IL-8 transcripts and protein and were selected
for further study (Fig. 3
, lanes 5 and 6, and
Table I). Injection of clones into nude mice produced a
range of consistent results: OVCA 194 IL-8 clone 10 showed significant
retardation of tumor growth and reduction of tumor size (at day 28
postinjection, Fig. 5A).
However, the tumor incidence in mice bearing this clone was not
significantly reduced (Table II). In
contrast, the effect of OVCA 494 cells was dramatic: tumors failed to
grow in all six animals injected with 494 IL-8 clone 2 (Table II and
Fig. 5B). Clones of cells receiving empty vector grew in all
mice injected (Table II, Fig. 5B). In addition, there was no
difference in the in vitro doubling time of control vs IL-8-expressing
clones. These data strongly suggest a role for IL-8 in modulating tumor
growth. The difference between 194 and 494 may be due to the greater
incidence of tumorigenicity of 194 cells (Fig. 1).
|
|
To investigate the host cellular responses activated by hIL-8, the
histology of the tumor injection sites was examined. These tumors were
introduced s.c. at the flank of the mice. The inoculation site of 194
hIL-8 and 494 hIL-8 displayed a marked elevated cellularity (Fig. 6, C, D, G, and H).
In contrast, few infiltrating cells were observed at the 194 neo and
494 neo injection sites (Fig. 6, A, B, E, and F).
This suggests that IL-8 production in human ovarian tumors mediate
cellular infiltration to the tumor injection site.
|
To identify the cells found in the infiltrate, immunocytochemistry
was performed on frozen tissue from the s.c. tumor injection site.
Application of Abs to identify neutrophils (Gr-1) or macrophages
(Mac-1) demonstrated a marked increase of both phenotypes at the
injection site of both IL-8-expressing clonal lines (Figs. 7 and 8).
We also detected the presence of B cells in these tumors, although the
signal was weak (Fig. 8). NK cells were not found. These results
suggest that infiltrating granulocytes, particularly neutrophils, are
attracted by IL-8 and are likely to play a role in initiating immune
responses to human ovarian tumor cells. This response appears to be
correlated with a significant reduction in the rate of tumor
growth.
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
IL-8 is a member of the C-X-C chemokine family and has multiple biological functions (22). In vitro, IL-8 can inhibit tumor cell proliferation (33) but does not cause tumoricidal activity. Based on other reports, IL-8 is a chemotactic factor for T cells, neutrophils, and basophils (34, 35, 36). In our model, the local production of IL-8 and the subsequent neutrophilic infiltration result in markedly decreased tumor growth. If neutrophils are responsible for the inhibition of tumor growth in nude mice, there are direct and indirect mechanisms whereby IL-8-activated neutrophils can exhibit their effect. Direct mechanisms may involve IL-8-activated neutrophils inducing shape change, granular release, and generation of oxygen radicals and proteases, which may lead to tumor cell death (37). Alternatively, an indirect mechanism may involve the production of mediators by neutrophils in response to chemokines, e.g., TNF, IL-1, and IFNs (38).
Although IL-8 was originally shown to attract and activate neutrophils, this chemokine has multiple biological functions. Some of these functions appear to involve pro-oncogenic activity: IL-8 can induce the migration of some tumor cells (20); it can enhance angiogenesis (39, 40); and it enhances metastasis of melanoma (19). Several papers have also demonstrated that the inhibition of IL-8 attenuates angiogenesis in bronchogenic carcinoma (18) and the expression level of IL-8 correlates with the metastatic potential of human melanoma cells in nude mice (17). Moreover, inhibition of IL-8 reduces tumorigenicity of human non-small cell lung cancer in SCID mice (21).
In contrast, we have observed an antitumorigenic effect of IL-8 in human ovarian cancer cells. This is consistent with a recent report by Hirose et al. (32) showing that the introduction of IL-8 into Chines hamster ovary cells, a hamster ovarian tumor cell line, reduces tumorigenicity in nude mice. Our use of human ovarian tumor cells may provide the impetus to further investigate the clinical relevance of IL-8. We have performed an in vitro invasion assay to determine whether human ovarian cancer cells treated with IL-8 exhibit increased invasion. We have also examined whether IL-8-transfected ovarian cell lines exhibit increased metastasis. Both findings were negative (data not shown). A pattern in melanoma and lung cancer cell lines may be explained by differences in the metastatic potential of different tumor cell types. There may be a balance between neutrophils/macrophages chemotaxis and angiogenesis/metastasis. The overall effect of IL-8 appears to require additional factors necessary for metastasis on human ovarian cancer cells (17, 32).
Paclitaxel is highly effective as a clinical therapy for ovarian
cancer, whereas its use against melanoma has not shown promise. The
differential effect of paclitaxel-induced chemokine on the growth of
different tumors may be an underlying factor for these distinct
clinical responses and is worthy of further investigations.
Interestingly, we have found two squamous lung cancer cell lines which
produce IL-8 in response to paclitaxel (T. Collins, L.-F. Lee, and J.
Ting, personal observation), and we are currently analyzing their
outcomes under conditions similar to those of this study. The
comparison is relevant because pulmonary and ovarian tissues are
epithelial tumors which represent primary tumors that are clinically
responsive to paclitaxel to a degree. Previously, we noted that 
50%
of ovarian carcinoma lines synthesize IL-8 after paclitaxel treatment
(24). This observation has been extended to primary
freshly explanted ovarian cancer cells (24). The analysis
of ovarian cancer cell lines that represent each of the two phenotypes
shows that the nonresponsive phenotype exemplified by OVCA 194 and OVCA
494 can be artificially engineered to express IL-8 gene resulting in
reduced tumor growth.
The histological examination clearly shows more Gr-1+ and Mac-1+ cells at the site of 194 and 494 injection when hIL-8 is expressed. In most reports, the inhibition of tumor growth is mediated through an infiltration of T lymphocytes and or macrophages into the tumor site (41). Previously, Tepper et al. (6, 42) devised a very similar methodology to study IL-4 and found that IL-4 production effectively suppressed a wide range of tumor cells in vivo. The antitumor effect is mediated by an inflammatory infiltrate composed of predominantly eosinophils and macrophages. Consistent with their findings, in our model, we found that the inhibition of tumor growth in immunoincompetent mice was associated with macrophages and neutrophils. The possible mechanisms by which neutrophils kill tumors were discussed above. In tumor immunity, macrophages are thought to be important in tumor lysis and killing. Several reports showed that activated monocytes stimulated by M-CSF kill tumor cells both in vivo and in vitro (43, 44). Activation of macrophages has been suggested to have antitumor effects. In addition to chemotaxis of macrophages and neutrophils, IL-8 has been demonstrated in controlling the trafficking of T cells (35, 36). Similar experiments remain to be tested in immunocompetent mice.
In conclusion, artificial expression of IL-8 in human ovarian cancer cells retarded tumor growth. This was associated with an immune cells infiltrate composed of neutrophils and to a lesser degree, macrophages. Our data provide evidence that IL-8 mediates the beneficial effects of paclitaxel toward ovarian cancer. It would be useful to study the cytokine production in response to paclitaxel treatment in the xenograft mouse model and to determine cellularity. Experiments are currently under way to evaluate this postulate.
|
Footnotes |
|---|
2 Current address: Cancer Center, University of California at Davis, Sacramento, CA 95817.
3 Address correspondence and reprint requests to Dr. Jenny P.-Y. Ting, Lineberger Comprehensive Cancer Center, University of North Carolina, Chapel Hill, NC 27599-7295. E-mail address:
4 Abbreviation used in this paper: h, human.
Received for publication July 20, 1999. Accepted for publication December 27, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-interferon gene transfer into tumor cells generates potent and long lasting antitumor immunity. Cancer Res. 50:7820., IFN-, and TNF-. Lab. Invest. 74:146.[Medline]
1 interferon gene transfer into metastatic Friend leukemia cells abrogated tumorigenicity in immunocompetent mice: antitumor therapy by means of interferon-producing cells. Cancer Res. 53:1107.. Proc. Natl. Acad. Sci. USA 90:2774.This article has been cited by other articles:
|
|
 |
|
  S. R. Perri, D. Martineau, M. Francois, L. Lejeune, L. Bisson, Y. Durocher, and J. Galipeau Plasminogen Kringle 5 blocks tumor progression by antiangiogenic and proinflammatory pathways Mol. Cancer Ther., February 1, 2007; 6(2): 441 - 449. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. E. De Larco, B. R. K. Wuertz, and L. T. Furcht The Potential Role of Neutrophils in Promoting the Metastatic Phenotype of Tumors Releasing Interleukin-8 Clin. Cancer Res., August 1, 2004; 10(15): 4895 - 4900. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Mukaida Pathophysiological roles of interleukin-8/CXCL8 in pulmonary diseases Am J Physiol Lung Cell Mol Physiol, April 1, 2003; 284(4): L566 - L577. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Hu, J. Hofmann, C. Zaloudek, N. Ferrara, T. Hamilton, and R. B. Jaffe Vascular Endothelial Growth Factor Immunoneutralization Plus Paclitaxel Markedly Reduces Tumor Burden and Ascites in Athymic Mouse Model of Ovarian Cancer Am. J. Pathol., November 1, 2002; 161(5): 1917 - 1924. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Salcedo, M. Martins-Green, B. Gertz, J. J. Oppenheim, and W. J. Murphy Combined Administration of Antibodies to Human Interleukin 8 and Epidermal Growth Factor Receptor Results in Increased Antimetastatic Effects on Human Breast Carcinoma Xenografts Clin. Cancer Res., August 1, 2002; 8(8): 2655 - 2665. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Kawakami, M. Kawakami, P. J. Snoy, S. R. Husain, and R. K. Puri In Vivo Overexpression of IL-13 Receptor {alpha}2 Chain Inhibits Tumorigenicity of Human Breast and Pancreatic Tumors in Immunodeficient Mice J. Exp. Med., December 17, 2001; 194(12): 1743 - 1754. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. M. Cairns, J. R. Gordon, F. Li, M. E. Baca-Estrada, T. Moyana, and J. Xiang Lymphotactin Expression by Engineered Myeloma Cells Drives Tumor Regression: Mediation by CD4+ and CD8+ T Cells and Neutrophils Expressing XCR1 Receptor J. Immunol., July 1, 2001; 167(1): 57 - 65. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Zhang, S. Wei, J. Sun, D. Coppola, B. Zhong, G. D. Wu, B. Goodwin, S. Sebti, J. Y. Djeu, and G. Blanck Retinoblastoma Protein Activation of Interleukin 8 Expression Inhibits Tumor Cell Survival in Nude Mice Cell Growth Differ., December 1, 2000; 11(12): 635 - 639. [Abstract] [Full Text]
|
|
|
|
 |
|
 S. Huang, J. B. Robinson, A. DeGuzman, C. D. Bucana, and I. J. Fidler Blockade of Nuclear Factor-{{kappa}}B Signaling Inhibits Angiogenesis and Tumorigenicity of Human Ovarian Cancer Cells by Suppressing Expression of Vascular Endothelial Growth Factor and Interleukin 8 Cancer Res., October 1, 2000; 60(19): 5334 - 5339. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
