| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
and IL-1
on Tumorigenicity Patterns and Invasiveness 1
Departments of
*
Microbiology and Immunology and
Pathology, Faculty of Health Sciences, The Cancer Research Center, Ben-Gurion University of the Negev, Beer-Sheva, Israel;
Department of Cell and Molecular Biology, University of Lund, Lund, Sweden; and
University of Colorado Health Sciences Center, Denver, CO 80262
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and IL-1
), in the milieu of tumor cells that produce them, differentially affect the malignant process. Active forms of IL-1, namely precursor IL-1 (pIL-1), mature IL-1 (mIL-1), and mIL-1 fused to a signal sequence (ssIL-1), were transfected into an established fibrosarcoma cell line, and tumorigenicity and antitumor immunity were assessed. Cell lines transfected with pIL-1, which expresses IL-1 on the membrane, fail to develop local tumors and activate antitumor effector mechanisms, such as CTLs, NK cells, and high levels of IFN-
production. Cells transfected with secretable IL-1 (mIL-1 and ssIL-1) were more aggressive than wild-type and mock-transfected tumor cells; ssIL-1 transfectants even exhibited metastatic tumors in the lungs of mice after i.v. inoculation (experimental metastasis). In IL-1 tumors, increased vascularity patterns were observed. No detectable antitumor effector mechanisms were observed in spleens of mice injected with IL-1 transfectants, mock-transfected or wild-type fibrosarcoma cells. Moreover, in spleens of mice injected with IL-1 transfectants, suppression of polyclonal mitogenic responses (proliferation, IFN- and IL-2 production) to Con A was observed, suggesting the development of general anergy. Histologically, infiltrating mononuclear cells penetrating the tumor were seen at pIL-1 tumor sites, whereas in mIL-1 and ssIL-1 tumor sites such infiltrating cells do not penetrate inside the tumor. This is, to our knowledge, the first report on differential, nonredundant, in vivo effects of IL-1 and IL-1 in malignant processes; IL-1 reduces tumorigenicity by inducing antitumor immunity, whereas IL-1 promotes invasiveness, including tumor angiogenesis, and also induces immune suppression in the host. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and IL-1, and one antagonistic cytokine, the IL-1R antagonist (IL-1Ra).
4 IL-1, IL-1, and IL-1Ra are encoded by different genes. Both IL-1 and IL-1 differ from most other cytokines by lacking a signal sequence, thus not trafficking through the endoplasmic reticulum (ER)-Golgi pathway; the precise mechanisms of IL-1 secretion are thus largely unknown.
IL-1 and IL-1 bind to the same receptors, and there are no significant differences in the spectrum of activities of recombinant IL-1 or IL-1 when studied in vitro or in vivo in diverse experimental systems. However, endogenously produced IL-1 and IL-1 differ dramatically in the subcellular compartments in which they are active. IL-1 is active in its secreted form (17.5 kDa), whereas the IL-1 precursor is inactive; IL-1 is mainly active as an intracellular precursor (31 kDa) or as a membrane-associated form (23 kDa), but is only marginally active as a secreted 17.5 kDa molecule. Mononuclear cells manifest the strongest secretory capacity of IL-1 and IL-1, whereas diverse nonphagocytic cells generally secrete low levels of IL-1. IL-1 is only rarely secreted by living cells, except for activated macrophages, and in contrast to IL-1, IL-1 is not commonly detected in blood or in body fluids, except during severe disease, in which case the cytokine may be released from dying cells.
Diverse effects of the IL-1 molecules on tumor development have been described (reviewed in Refs.1, 2). On the one hand, antitumor effects of IL-1 have been described in experimental tumor systems, mainly due to its ability to costimulate T cell activation, to induce cytokine secretion in specific as well as nonadaptive immune cells, and to potentiate the differentiation and function of immune surveillance cells. On the other hand, IL-1 potentiates invasiveness and metastasis of malignant cells, mainly by inducing adhesion molecule expression on the tumor cells as well as on endothelial cells (1, 2). In addition, IL-1 may stimulate the production of invasiveness-promoting factors such as matrix metalloproteinases, growth factors, or angiogenic factors by the malignant cells or by cellular elements in the tumor’s microenvironment. The diverse effects of the IL-1 molecules on malignant processes have hindered the use of IL-1 as an antitumor agent in clinical trials (1).
We have hypothesized that the subcellular compartmentalization of the IL-1 molecules within the producing cell and its local microenvironment dictates its functions in the malignant process. Previously, we have reported on reduced tumorigenicity and increased immunogenicity of oncogene-transfected fibrosarcoma cells that constitutively express IL-1 (3, 4, 5, 6, 7). In this study, we compared the effects of tumor cell-associated IL-1 and IL-1 on invasiveness and antitumor immunity. For this purpose, we transfected wild-type fibrosarcoma cells with constructs bearing the cDNAs of the active forms of the IL-1 molecules, i.e., the precursor form of IL-1 (pIL-1), the mature form of IL-1 (mIL-1), and mIL-1 fused to a signal sequence (ssIL-1), the latter to allow secretion through the ER-Golgi pathway (8). The results demonstrate that tumor cell-associated IL-1 potentiates the development of antitumor immune responses, which limit the growth of the malignant cells, whereas IL-1 expression by the tumor cells potentiates invasiveness patterns and induces anergy in the host.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
NFS/N mice were obtained from Dr. I. Fossar Larson, Animal Section, The Fibiger Institute, Copenhagen, Denmark and subsequently bred at the animal facilities of the Faculty of Health Sciences, Ben-Gurion University, Beer-Sheva, Israel. NFS/N mice, which are syngeneic to NIH 3T3 fibroblasts and transformed cell lines derived from them, were used in this study (9). Either female or male mice, 6–12 wk, were used in the experiments.
Tumor cell lines and culture conditions
Clone 3 (Cl.3), established as a cloned wild-type fibroblastoid cell line, was derived from NIH 3T3 cells transformed with the TPR-met oncogenes (3). Cl.3 induces local tumors in syngeneic NFS/N mice and has been used as a parental clone for IL-1 and IL-1 gene transfections. Cells were maintained in DMEM (Life Technologies, Paisley, Scotland, U.K.) supplemented with 5% heat-inactivated FCS, 2 mM L-glutamine, penicillin G (100 U/ml) and streptomycin (100 µg/ml) (each from Biological Industries, Kibbutz Beit Haemek, Israel). The transfectants were grown in maintenance medium containing 0.5 mg/ml G418 (Life Technologies).
Retroviral expression vector and gene transfer
The parental cell line (Cl.3) was transfected using the pLXSN vector (Clontech Laboratories, Palo Alto, CA), which contains the neomycin resistance gene and SV40 promoter. Various active forms of human IL-1 were inserted into the vector: the full-length pIL-1 (31 kDa) was inserted into the site of Xho/BamHI; mIL-1 (17.5 kDa) and ssIL-1 were inserted at the BamHI site (8). The transfectants were selected and cloned using limiting dilutions in 0.5 mg/ml G418. IL-1 expression was assessed using ELISA and biological assays.
ELISA for cytokines
Murine IFN- and IL-2 levels were measured using ELISA kits (BD PharMingen, San Diego, CA). Human IL-1 and IL-1 were detected using ELISA kits (R&D Systems, Minneapolis, MN). Lysates of transfectants were prepared by three cycles of freeze-thawing.
Assays for tumor development
Tumor growth was routinely analyzed following the intrafootpad injection of wild-type, mock-transfected tumor cells or the various IL-1 transfectants into the left footpads of mice. The tumor cells were adjusted to appropriate numbers in PBS and 50 µl were injected per footpad. Local tumor growth was determined by measurements of the footpad diameters 2–3 times a week using a caliper.
Experimental metastasis
Mice were inoculated i.v. with either wild-type cells or the various IL-1 transfectants at a dose of 5 x 105 cells/mouse. Moribund mice were sacrificed to measure lung metastatic load. The experiments were terminated after 60 days when all the surviving mice were sacrificed to assess metastatic load. The lung metastatic load was assayed by weighing the lungs. Lungs with visible nodules after fixation in Bouin solution were defined as positive for metastasis, and this was confirmed by histological analysis.
Histology analysis
Tumors, removed from footpads of mice 10 days after intrafootpad injection of wild-type, mock-transfected cells or the various IL-1 transfectants at a dose of 2 x 105 cells/mouse, were immediately fixed in 10% phosphate-buffered formalin. The formalin-fixed samples, as well as the Bouin solution-fixed lungs from mice with metastases, were embedded in paraffin, sectioned at 4 µm, and stained with H&E.
Immunohistochemistry analysis and quantification of microvessel density
The DAKO Envision kit (EnVision Plus System; DAKO, Glostrup, Denmark) was used to stain tumor sections for immunohistochemistry, using rat polyclonal Abs against human von Willebrand factor, which also cross-react with murine von Willebrand factor. Briefly, after deparaffinization and rehydration, the sections were incubated 1 h with anti-von Willebrand factor Abs (diluted 1:200), then incubated with labeled polymer and developed with a diaminobenzidine substrate-chromogen solution. The microvessel density (MVD) was determined by counting blood vessels in the areas of highest vascularity, under low magnification (x100), as previously described (10). MVD refers to the mean vessel number of six highly vascularized fields in each section. The blood vessel count was performed by the same pathologist in a blind manner.
Cytotoxicity assays
NK cell activity in fresh spleen cells from mice 15–18 days after tumor cell inoculation (intrafootpad, 2 x 105 cells/mouse), suspended in RPMI 1640 (Life Technologies) supplemented with 10% FCS, 2 mM L-glutamine, penicillin G (100 U/ml), streptomycin (100 µg/ml), and 2-ME (2 x 10-5 M) (Sigma-Aldrich, St. Louis, MO) (complete RPMI 1640), was assayed in a 4-h standard 51Cr release assay using YAC-1 cells, which are immunospecific for murine NK cells, as a target (4). Briefly, effector spleen cells were added to target cells, previously labeled with 100 µCi of 51Cr (NEN, Boston, MA), in different E:T ratios, as indicated, and incubated for 4 h. 51Cr release in culture supernatants was determined in a gamma counter (Minaxi Auto Gamma 5000 Series; Packard Instrument, Downers Grove, IL). Data were expressed as a percentage of specific lysis according to the formula: specific lysis percentage = (experimental cpm - spontaneous cpm)/(total cpm - spontaneous cpm) x 100%. Spontaneous release was 8–12% in all experiments. The results are presented as the mean percentage of specific 51Cr release of triplicates.
For CTL activation, spleen cells were suspended in complete RPMI 1640 (5 x 106 cells/ml) and cultured for 6 days in the presence of wild-type, mock-transfected cells, or the various IL-1 transfectants (2.0 x 105 cells/ml), which were inactivated by mitomycin C (Sigma-Aldrich), and medium was changed every day. Blasts were harvested, and cytotoxic activity against relevant targets was determined in 4-h 51Cr release assay as previously described (5).
Mixed lymphocyte and tumor cell culture (MLTC)
To assess specific antitumor immunity, we performed MLTC, which consists of lymphoid cells from tumor-bearing mice and mitomycin C-inactivated tumor cells (11). Thus, spleen cells, as effector cells, were taken from tumor-bearing mice, 15–18 days after tumor cell inoculation (intrafootpad, 2 x 105 cells/mouse). Spleen cells (2.0 x 106/ml) were mixed with the relevant mitomycin C-inactivated tumor cells (2.0 x 105/ml), which served as stimulator cells. Stimulator cells were inactivated with mitomycin C at the concentration of 100 µg/ml (<1 x 107 cells/ml), incubated 1 h on ice followed by extensive washes. The cell mixtures were cultured in complete RPMI 1640 for 24 h. Supernatants were collected and IFN- and IL-2 were assayed by ELISA.
Proliferation and cytokine production of spleen cells
Freshly prepared single spleen cells from mice 15–18 days after intrafootpad injection of wild-type, mock-transfected cells and the various IL-1 transfectants, at a dose of 2 x 105 cells/mouse, were seeded in 96-well plates (2.0 x 105/0.2 ml/well) for the proliferation assay and in 24-well plates (2 x 106/1.0 ml/well) for cytokine production assays in RPMI 1640 supplemented with 10% FCS, 2 mM L-glutamine, penicillin G (100 U/ml) and streptomycin (100 µg/ml), and 2-ME (2 x 10-5 M). The cells in 96-well plates were cultured in triplicates for 48 h in the presence of 2.5 µg/ml Con A (Sigma-Aldrich) and pulsed with [3H]TdR (1 µCi/well, 38.0 Ci/mmol; Amersham Pharmacia, Buckinghamshire, U.K.) for the final 18 h of incubation. After incubation, the cells were harvested on glass fiber paper and the counts per minute were recorded using a liquid scintillation counter (Wallac Oy, Turku, Finland). The cells in 24-well plates were cultured for 24 h in the presence of Con A (2.5 µg/ml) and supernatants were collected and IFN- and IL-2 were detected by ELISA.
Statistical analysis
The significant differences in results were determined using the two-sided Student t test, and a p < 0.05 was considered significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
IL-1 or IL-1 was measured in the supernatants and lysates of wild-type, mock-transfected cells or the various IL-1 transfectants. As shown in Fig. 1A, in clones transfected with pIL-1, IL-1 was found only in cell lysates that represent the cytosolic and membrane-associated forms of the cytokine; it could not be detected in supernatants of cells (2). In cells transfected with mIL-1, which inefficiently secrete the cytokine, most of the IL-1 (
10 ng/ml generated by 1 x 106 cells in 24 h) was detected in the cytosol; however, significant amounts of the cytokine (3 ng/ml) were also secreted (Fig. 1B). In cells transfected with ssIL-1, in which secretion of the cytokine is through the ER-Golgi exocytic pathway, most of the cytokine was secreted (5 ng/ml) (Fig. 1B). Although each construct was put under the same promoter in the same expression vector, in several transfection and cloning experiments, we have observed that levels of IL-1 were lower than those of IL-1 (100–500 pg vs 5–10 ng generated by 106 cells/ml in 24 h, respectively), as also described in primary cells, such as activated macrophages (reviewed in Ref.2). In contrast, no IL-1 was detected in either supernatants or lysates of wild-type and mock-transfected fibrosarcoma cells. Both IL-1 and IL-1 produced by the various IL-1 transfectants are biologically active, which was demonstrated by an IL-1 bioassay. The bioactivity of membrane-associated IL-1 was also demonstrated in pIL-1 transfectant clones by paraformaldehyde-fixation (data not shown).
|
reduces the tumorigenicity but IL-1 enhances the malignancy of fibrosarcoma cells
To assess the tumorigenicity patterns of fibrosarcoma cells expressing the various forms of IL-1, we injected intrafootpad wild-type, mock-transfected tumor cells or the various IL-1 transfectants (0.2 x 106 cells/mouse) into syngeneic mice and tumor growth was monitored using a caliper. As shown in Fig. 2A, the two clones of pIL-1 transfectants have completely lost their ability to form a tumor mass, whereas clones of mIL-1 and ssIL-1 transfectants developed more virulent tumors than the parental wild-type and mock-transfected cells, manifesting more rapid tumor growth and earlier death of tumor-bearing mice (Fig. 2, A and B).
|
on invasiveness patterns. Thus, we injected the mIL-1 and ssIL-1 transfectants and also the control lines intrafootpad, at lower gradient cell doses that initially produce less IL-1. Similar growth patterns were observed at each cell dose tested; consistently, mIL-1 and ssIL-1 transfectants were more tumorigenic than the control cells, as detected by enhanced tumor growth patterns and mortality rates. Reduced invasiveness patterns were thus observed after injection of mice with small cell doses of both types of IL-1 transfected cells that initially produced small amounts of the cytokine (Fig. 3). For example, 5 x 103 ssIL-1-transfected cells, which secrete 25 pg of IL-1 in 24 h, are highly tumorigenic, whereas 2 x 105 pIL-1-transfected cells, which produce 50 pg of IL-1 under the same conditions, are not tumorigenic. Thus, enhanced tumor invasiveness was manifested by both types of IL-1 transfectants (pIL-1 and ssIL-1), which secrete the cytokine, suggesting a qualitative feature of the secretable cytokine (and not the membrane-associated or cytosolic forms) in promoting the malignant phenotype.
|
-transfected fibrosarcoma cells grew progressively in sublethally irradiated syngeneic mice, indicating that the malignant cells did not lose their invasiveness following transfection with pIL-1 (Ref.4 and data not shown).
Spontaneous metastasis in the lungs was not observed in mice injected with either wild-type, mock-transfected tumor cells or any of the various forms of IL-1 transfectants. In addition, no lung metastases were observed in tumor cell-inoculated mice after amputation of the primary tumor. Induction of lung experimental metastases of tumor cells in the lungs following i.v. inoculation represents a feature of tissue invasiveness from the vasculature. Only cells transfected with ssIL-1, in which most of the IL-1 produced are secreted, developed lung metastases. These were manifested by an increase in the weight of the lungs as well as earlier mortality in comparison with mice injected with wild-type cells or the other transfectants. In contrast, no significant difference in lung weight or survival time was observed in mice inoculated with either pIL-1, mIL-1 transfectants or wild-type tumor cells (Table I). This indicates that transfectants of both isotypes of IL-1 secrete large enough doses of the cytokine for promoting local tumor invasiveness. However, only ssIL-1-transfected cells, which secrete the largest amounts of IL-1, are capable of inducing experimental metastasis in the lungs.
|
-positive fibrosarcoma tumorsTo understand the mechanisms of IL-1-induced fibrosarcoma rejection or progression and to get indications on the development of antitumor immunity, we performed histological analyses of the cellular infiltrate at tumor sites (Fig. 4). Tumor samples were obtained from mice 10 days after intrafootpad injection of the various IL-1 transfectants or wild-type controls.
|
-transfected cells, tumors were very small and limited and a heavy mononuclear cell infiltrate invaded the entire tumor. Single dispersed tumor cells characterized by a low mitotic activity were observed abundantly (Fig. 4B). The pIL-1-positive tumors were completely eradicated on day 21, and at the site of inoculation, fibrotic scar tissue, which replaced the tumor’s mass, was observed (data not shown). Extensive mitosis was observed in sections of both mIL-1- and ssIL-1-transfected tumors (Fig. 4C), compared with the moderate mitosis seen in wild-type and mock-transfected tumors (Fig. 4A). In mIL-1 and ssIL-1-transfected tumors, necrotic areas were observed, possibly due to the extensive multiplication rate of the cells. Few mononuclear cells were detected within mIL-1 and ssIL-1 tumors, although a heavier mononuclear infiltrate was observed in the periphery of such tumors. Tumors induced by ssIL-1-transfected fibrosarcoma cells were infiltrated by neutrophils; the most dense infiltrate was observed in necrotic areas (Fig. 4D).
Recombinant IL-1 induces the production of multiple chemokines that may potentially recruit different types of leukocytes into tumors (1, 2), however, in the case of our IL-1-transfected tumors, the infiltrate of mononuclear cells inside the tumor is poor. As will be described in mice bearing IL-1-transfected tumors, systemic immune suppression is observed and it may also suppress the production of chemokines and the influx of inflammatory cells into the tumors (see also Discussion).
Tumor cell-derived IL-1 enhances angiogenesis of tumors
To further elaborate on the invasiveness patterns of the various control cells and IL-1 transfectants, we have assessed the angiogenic network in tumor sections stained with Abs against von Willebrand factor, a marker of endothelial cells, and counterstained with H&E. Enhanced angiogenesis patterns in tumors are correlative to increased invasiveness. Many blood vessels were observed in sections of tumors taken from mice 10 days after inoculation of mIL-1 or ssIL-1 transfectants (Fig. 5B), whereas fewer blood vessels were seen in the tumors from mice injected with wild-type or pIL-1-transfected fibrosarcoma cells (Fig. 5A). This was confirmed by counting the mean MVD in tumor sections (Fig. 5C). These results indicate that secreted IL-1 by tumor cells potentiates their invasive potential.
|
and IL-1 transfectantsTo assess the role of immune functions in determining the malignant phenotype of IL-1-transfected fibrosarcoma cells, we assayed antitumor effector cells and cytokines in spleen cell cultures from tumor-bearing mice 15–18 days after tumor cell inoculation.
NK cell killing activity was assessed in freshly harvested spleens from tumor-bearing mice, using sensitive YAC-1 cells. As seen in Fig. 6A, moderately elevated NK cell activity, as compared with that in spleens of normal mice, was observed in spleens of mice injected with wild-type cells, mock- and pIL-1-transfected cells. In spleens of mice inoculated with pIL-1 transfectants, the killing activity was the highest. Interestingly, in spleens from mice bearing tumors induced by IL-1 transfectants, inhibition of NK cell activity was observed. In spleens of mice bearing ssIL-1 tumors, which secrete the highest levels of IL-1, no NK killing activity was observed. In mice injected with mIL-1 fibrosarcoma cells, suppressed NK killing activity was observed; NK cell activity was gradually inhibited as the E:T ratio increases, and at a 200:1 ratio no killing activity was observed. This indicates that some suppressive moiety (cell or factor) exists in the spleens of mice injected with both types of IL-1 transfectants.
|
production were assessed in MLTC consisting of spleen cells from tumor-bearing mice challenged in culture with mitomycin C-inactivated tumor cells. Specific CTL responses were observed only in spleens of mice inoculated with pIL-1-transfected cells; specific killing increased in a dose-dependent manner as more effector cells were present in the killing assay (Fig. 6B). IFN- is an important mediator for the development of antitumor immunity, activating both nonadaptive and specific immune effector cells. As shown in Fig. 6C, high levels of IFN- were observed in supernatants from the MLTC consisting of spleen cells from mice injected with pIL-1 transfectants, whereas only low levels of IFN- production were detected in supernatants of the cultures of spleen cells of mice injected with either mIL-1 or ssIL-1 transfectants, wild-type or mock-transfected tumor cells. Induction of CTLs and IFN- secretion are Ag specific and were detected only when the sensitizing tumor cells were added to culture; no response was detected upon challenge in vitro with irrelevant syngeneic fibrosarcoma cells (methycholanthrene-induced tumors from NFS/N mice). Similar patterns of antitumor specific immune responses (CTL activity and IFN- secretion) were observed also in the draining lymph nodes (data not shown).
As previously stated, reduced NK cell activity, no CTL activity, and only low levels of IFN- secretion were observed in spleens of mice injected with IL-1 transfectants. Therefore, we wanted to assess whether this suppression was Ag-specific or represented general anergy. For this goal, spleen cells from mice 15–18 days after inoculation of the various IL-1 transfectants, wild-type, or mock-transfected cells were stimulated by a polyclonal T cell mitogen, Con A, and proliferation and cytokine secretion were assessed. As can be seen in Fig. 7A, an extensive proliferative response was observed in Con A-stimulated spleen cell cultures from naive cells or mice injected with wild-type, mock-, or pIL-1-transfected cells, however, suppressed proliferation was observed in the spleens of mice inoculated with both types of IL-1 transfectants. The same patterns of suppressed responses in spleen cells from IL-1 tumor-bearing mice were observed when IFN- (Fig. 7B) and IL-2 (Fig. 7C) secretion were tested, indicating that IL-1 transfectants induce general anergy, at the levels of NK and T cell functions, as tested here. These patterns of suppression were also observed when the proliferation and IFN- secretion were assessed in Con A-stimulated purified CD3+ T cells from mice bearing ssIL-1 tumors; functions of CD4+ T cells were more suppressed than those of CD8+ T cells (data not shown).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
or IL-1 by tumor cells to their microenvironment, determines its effects on the outcome of the malignant process. Thus, fibrosarcoma cells transfected with pIL-1, in which the cytokine is expressed on the membrane and in the cytosol, but is not secreted, lose their invasiveness in vivo by developing efficient antitumor cell immune responses. In contrast, fibrosarcoma cells, transfected with secretable forms of IL-1, i.e., mIL-1 and ssIL-1, are more malignant than the wild-type or mock-transfected cells. This is, to our knowledge, the first report on in vivo, differential, nonredundant functions of IL-1 and IL-1 under defined physiological conditions, as exemplified in this study in the control of tumor development.
Because recombinant IL-1 and IL-1, added extracellularly to cells or injected into animals, bind to the same cell surface receptors and exert a similar spectrum of biological activities, we hypothesized that the markedly different in vivo effects of the IL-1 molecules were determined by the compartmentalization of the IL-1 molecules within the producing cell and its microenvironment (1). Thus, membrane-associated IL-1, expressed by the malignant cells, preferentially activates in a juxtacrine manner specific as well as nonadaptive immune cells, via ligation to IL-1R. As such, membrane IL-1 functions as a cell-to-cell interaction molecule and potentiates antitumor immunity at low levels of expression, which are below those toxic to the host. In contrast, the secreted form of IL-1 easily diffuses into the tumor’s microenvironment, in which it induces inflammatory responses and thus potentiates tumor invasiveness and metastasis. This IL-1-mediated enhancement of tumorigenicity possibly acts through stimulation of secretion, in the malignant or host-derived stromal cells, of molecules that promote invasiveness and metastasis, such as growth factors, angiogenesis-promoting factors, matrix metalloproteinases and adhesion molecule expression.
Our results have emphasized the role of the "natural" membrane-associated form of IL-1 as a focused immunostimulatory adjuvant that helps to mount antitumor immunity. Regression of IL-1-positive tumors involves the activation of T cell-mediated immunity, CD8+ T cells, Th1-type cytokines, as well as nonadaptive immune effector cells like NK cells and activated macrophages (1, 3, 4, 5, 6, 7). Membrane IL-1 may putatively serve as a cell-associated costimulatory molecule, like B7, for the activation of T cells and nonadaptive effector cells. In accordance, equivalent NF-
B activation and IL-2 secretion were observed in T cell costimulated either by IL-1 or by B7 (12). Other studies have also emphasized the effectiveness of membrane-associated cytokines expressed on engineered tumor cells in increasing the immunogenicity of the malignant cells (i.e., IFN-, GM-CSF, M-CSF, TNF-, and IL-12) (13, 14, 15). In the case of M-CSF, the superiority of the membrane-associated isoform, as compared with secreted M-CSF, was clearly demonstrated.
Our results demonstrate that the effects of secretable IL-1 on tumorigenicity patterns are complex and involve effects on malignant cell invasiveness as well as effects on antitumor immunity. The secretion of IL-1 is an inefficient process, and most of the IL-1 remains intracellular. Following synthesis, the IL-1 precursor, proIL-1, is cleaved by the IL-1 converting enzymes (ICE), but remains primarily cytosolic unless an additional signal, such as exogenous ATP that promotes secretion, is provided (16). Hence, in fibrosarcoma cells transfected with mIL-1, most of the cytokine resides in the cytosol, as there is no additional signal that promotes active secretion, whereas in ssIL-1 transfectants, IL-1 is mainly detected in supernatants, as it is efficiently secreted through the classical ER-Golgi secretory pathway. Active secretion of IL-1 by the fibrosarcoma cells potentiated their invasiveness, but the degree of invasiveness may depend on the concentration of secreted IL-1, because only ssIL-1 transfectants, which secrete the largest amounts of cytokine, induce experimental lung metastasis. The mIL-1 transfectants, wild-type cells, or mock-transfected cells induce only local invasive tumors. Malignancy patterns of IL-1-transfected fibrosarcoma correlated to tumor-mediated angiogenesis, which was tested in this study as a parameter of tumor invasiveness; the most extensive vascularization patterns were observed in tumors induced by ssIL-1-transfected cells. Recently, Saijo et al. (17) reported increased tumorigenicity of 3LL cells, transfected with a construct of mIL-1 ligated to a leader peptide, similar to the ssIL-1 construct used in this study. Increased invasiveness of IL-1 secreting cells was shown to be mediated by enhanced tumor angiogenesis and increased angiogenic factor production (vascular endothelial growth factor, macrophage-inflammatory protein 2, and human growth factor) through communication networks between the malignant cells and stromal cells.
In this study, we have emphasized the role of tumor cell-derived IL-1 on invasiveness patterns. Similarly, we have demonstrated the role of host-derived IL-1 in determining the malignancy of B16 melanoma cells; we have shown that the invasiveness of B16 melanoma cells is inhibited in IL-1 knockout mice, whereas in control mice and in IL-1 knockout mice, invasive tumors developed (18). Thus, both host-derived and malignant cell-derived IL-1 contribute to enhanced tumor invasiveness, possibly because secretable IL-1 activates, by autocrine/paracrine manners, an inflammatory cascade in the tumor’s microenvironment.
When secretable mIL-1 and IL-1 (natural or recombinant) are involved in in vivo tumor systems, increased invasiveness of malignant cells has been described (2, 19, 20, 21, 22, 23, 24). For example, Vidal-Vanaclocha and associates (23, 25, 26) have used the murine B16 model to demonstrate the role of stroma-derived IL-1 in promoting melanoma hepatic experimental metastasis. Enhanced IL-1-mediated metastasis was induced by firm adhesion of the melanoma cells to hepatic sinusoidal endothelial cells, through up-regulation of expression of VCAM-1 and very late antigen-4 on hepatic sinusoidal endothelial cells by IL-1. Also, tumor cells transfected with a mature secretable form of IL-1 were more invasive than the parental tumor cells (1, 20). In this study, we have emphasized the differential in vivo effects of the IL-1 molecules in their native milieu, on tumor invasiveness. Thus, IL-1, which is in vivo mainly active in cell-associated forms (cytosolic precursor or membrane-anchored form) and is only rarely secreted or detected in body fluids, is antitumorigenic, whereas IL-1, which only functions in its secreted mature form and can be detected in body fluids in inflammatory conditions, is protumorigenic. As already indicated, the protumorigenic effects of secreted IL-1 (mainly IL-1) stem from the strong inflammatory cascade that it induces, whereas in the case of cell-associated IL-1 (i.e., IL-1), inflammatory responses are possibly weaker, due to lower levels of expression of the cytokine, less spread of the cytokine in the microenvironment, or due to the domination of noninflammatory functions of cell-associated IL-1, like stimulation of immunity.
Fibrosarcoma cells, secreting IL-1, were shown to induce early general anergy that affects tumor cell specific as well as nonspecific mitogenic responses. Such suppression is not observed in mice inoculated with wild-type, mock-, or pIL-1-transfected cells even at later intervals. Suppression mediated by IL-1 tumors is systemic and is possibly mediated by immunosuppressive factors secreted by the malignant cells that subsequently activate suppressive circuits in the host, as suppression is gradually ablated following resection of the tumor (data not shown). Tumor-mediated suppression may explain in part the inability of mononuclear cells to efficiently invade tumors induced by IL-1 transfectants. Multiple suppression mechanisms, like CD4+CD25+ T cells, aberrant TCR expression and signaling, suppressor macrophages or their secreted products, may account for systemic immune suppression that is induced in mice injected with IL-1 transfectants (27, 28). Suppressed mitogenic responses and IFN- production were also observed in CD3+ enriched T cells; CD4+ enriched T cell suppression was also observed, although it was less pronounced in enriched CD8+ T cells (X. Song and R. N. Apte, unpublished observations). This might indicate that in such mice, inherent T cell functions are impaired, for example, lack of functional TCRs, as it is known that T cells from tumor-bearing mice express less of the 
-chain and as a result are defective in signaling (29). Lower responses may, however, also be due to the existence of suppressor cells in the enriched T cell subpopulations or due to insufficient function of APCs or costimulatory signals provided by them. As IL-1 is a pleiotropic cytokine, it may induce both adjuvant-like effects and immunosuppression under different experimental conditions (2). In in vitro systems, in the absence of immunosuppressive circuits, recombinant IL-1 can be immunostimulatory, as recently shown by its ability to costimulate IFN- production by T cells or NK cells (30, 31). In contrast, repeated exposure of macrophages to IL-1 renders them tolerant to cytokine secretion in response to further inflammatory stimuli (32). This may occur in the vicinity of IL-1-transfected tumors that continuously secrete the cytokine. The mechanisms of anergy induced by IL-1 transfectants are further assessed in our laboratory.
Similar effects of the IL-1 molecules were observed in other murine experimental tumors, i.e., reduced tumorigenicity of IL-1-transfected TS/A and DA/3 breast cancer cells in BALB/c and C57BL/6 mice respectively (E. Vozonov, Y. Krelin and R. N. Apte, unpublished observations) and increased tumorigenicity of IL-1-transfected 3LL cells in DBA/2 mice (17). In contrast to our present study results and those of Saijo et al. (17), Björkdahl et al. (33, 34) described reduced tumorigenicity and increased immunogenicity of B16F10 melanoma cells engineered to over-express IL-1 linked to a leader peptide. This discrepancy might be due to differences in the type of the tumor cells, the ability of IL-1-secreting tumor cells to activate tumor-mediated suppression and the susceptibility of tumor cells to the diverse pleiotropic effects of IL-1. For example, for some malignant cells IL-1 is cytostatic/cytotoxic, whereas for others it potentiates invasiveness and metastasis. The biological effects of IL-1 on tumor cells may thus depend on the type of tumor, the dose of IL-1, and the conditions of IL-1 administration.
Further knowledge on the mechanisms of action of the IL-1 molecules on malignant processes will allow a better understanding of tumor-host interactions and will facilitate the manipulation of malignant processes by intervention with the IL-1 system.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Current address: Pharmexa A/S, Horsholm, Denmark.
3 Address correspondence and reprint requests to Dr. Ron N. Apte, Department of Microbiology and Immunology, Faculty of Health Sciences, Ben-Gurion University of the Negev, Beer-Sheva 84105, Israel. E-mail address: rapte{at}bgumail.bgu.ac.il
4 Abbreviations used in this paper: IL-1Ra, IL-1R antagonist; pIL-1, precursor of IL-1; mIL-1, mature IL-1; ssIL-1, mIL-1 fused to a signal sequence; MVD, microvessel density; ER, endoplasmic reticulum; MLTC, mixed lymphocyte and tumor cell culture.
Received for publication March 24, 2003. Accepted for publication September 19, 2003.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
either spontaneously or following IL-1 gene transfer. Int. J. Cancer 51:822.[Medline]
expressed by fibrosarcoma cells confers reduced tumorigenicity. Immunol. Lett. 39:45.[Medline]
expression. Cancer Res. 59:1029. gene directs the protein from cytoplasmic accumulation to extracellular release. Cell. Immunol. 169:226.[Medline]
B. Mol. Cell. Biol. 18:3140.. Gene Ther. 4:1181.[Medline]
B through the P2Z purinoreceptor by selectively targeting NF-B p65. J. Cell Biol. 139:1635. promotes tumor growth of Lewis lung carcinoma by induction of angiogenic factors: in vivo analysis of tumor-stromal interaction. J. Immunol. 169:469. gene-transfected human melanoma cells increase tumor-cell adhesion to endothelial cells and their retention in the lung of nude mice. Int. J. Cancer 67:856.[Medline]
on metastasis and the location of B16F1 melanoma cell arrest. Cancer Res. 57:4105., interleukin-1 and interferon--inducing factor on the production of interferon- by natural killer. Eur. J. Immunol. 27:2787.[Medline]
induces in vivo tolerance to lipopolysaccharide in mice. Clin. Exp. Immunol. 128:221.[Medline]
gene to B16 mouse melanoma recruits leucocyte subsets and reduces tumour growth in vivo.. Cancer Immunol. Immunother. 44:273.[Medline]
induces tumor growth inhibition and an enhanced immunity against the wild-type B16 tumor. Cancer Gene Ther. 7:1365.[Medline]This article has been cited by other articles:
|
|
 |
|
  C. Guruvayoorappan and G. Kuttan (+)-Catechin inhibits tumour angiogenesis and regulates the production of nitric oxide and TNF-{alpha} in LPS-stimulated macrophages Innate Immunity, June 1, 2008; 14(3): 160 - 174. [Abstract] [PDF]
|
|
|
|
 |
|
 Y. Krelin, E. Voronov, S. Dotan, M. Elkabets, E. Reich, M. Fogel, M. Huszar, Y. Iwakura, S. Segal, C. A. Dinarello, et al. Interleukin-1{beta}-Driven Inflammation Promotes the Development and Invasiveness of Chemical Carcinogen-Induced Tumors Cancer Res., February 1, 2007; 67(3): 1062 - 1071. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. E. M. Dirkx, M. G. A. oude Egbrink, J. Wagstaff, and A. W. Griffioen Monocyte/macrophage infiltration in tumors: modulators of angiogenesis J. Leukoc. Biol., December 1, 2006; 80(6): 1183 - 1196. |
) and lysates (
) were prepared and assessed for IL-1 content by ELISA. Results are presented as the mean ± SD of three experiments.
), mock-transfected cells (MK, 
), pIL-1
, •), mIL-1
