Degradation of the extracellular matrix (ECM) is essential for progression and metastasis of cancer cells. The ECM-degrading enzymes, matrix metalloproteinases (MMPs), are produced mainly by intratumor monocytes/macrophages. MMPs, particularly MMP-9, are reported to be of crucial significance for both growth and tumor invasiveness. Inhibition of the expression of MMP-9 may prevent tumor development. High-dose intravenous gamma globulins (IVIG) effectively inhibit metastatic spread of tumors in mice and humans and a variety of mechanisms have been suggested to explain this effect.
We studied the effect of purified IVIG on MMP-9 secretion and mRNA expression by in vitro differentiated human monocytic cells (cell lines and peripheral blood monocytes). Zymography was employed to measure gelatinase secretion and Northern blot analysis was used to detect mRNA expression. Involvement of F(ab)2 and Fc components in IVIG activity was also evaluated.
IVIG dose dependently and significantly reduced the amount of secreted MMP-9 and its mRNA expression. F(ab)2, but not Fc fragments, led to suppressed MMP-9 activity. However, competitive experiments demonstrated that Fc, but not F(ab)2 fragments, reversed the IVIG-induced inhibitory effects.
Tumor growth, metastasis, and the angiogenesis associated with these processes require proteolytic degradation of the vascular basement membrane (BM) and the extracellular matrix. The proteases that are associated with tumor progression and its invasive potential are members of the multigene matrix metalloproteinase (MMP) family.1, 2 The source of MMPs in human cancers has been attributed mainly to stromal cells in the tumor vicinity, particularly local macrophages.3
Monocytes/macrophages secrete an array of MMPs. The gelatinases MMP-2 and MMP-9 are observed in high-grade tumors and are also associated with angiogenesis.3, 4 They cleave Type IV collagen, a central component of the BM. Regulation of MMP expression and activity is exerted at many levels including both transcriptional and posttranscriptional mechanisms.1, 2 Posttranslational regulation includes the secretion of MMPs as inactive zymogens and their activation by other proteases or their inactivation through binding to endogenous inhibitors, tissue inhibitors of metalloproteinases (TIMPs).1, 2 Generally, MMP-2 is constitutively expressed and MMP-9 is inducible by a variety of factors, including proinflammatory cytokines such as tumor necrosis factor-alpha (TNF-α) and interleukin (IL)-1β.2 The activated form of MMP-9 associates with the cell surface hyaluronan receptor CD44, which is also implicated in tumor growth and metastasis.5 This membrane localization of MMP-9 allows its escape from the intrinsic inhibitory activity of TIMPs. In addition to matrix degradation, MMP-9 is also implicated in the activation of immune-associated factors (e.g., transforming growth factor-β) that enhance tumor angiogenesis.6
High-dose intravenous gamma globulin (IVIG) is used to treat patients with metastatic tumors.7 Antitumor-specific antibodies have been found in the normal immunoglobulin repertoire and in the expanded repertoire in IVIG preparations. IVIG prevents metastatic spreading in mice and indirect evidence points to its effectiveness against human malignancies.8–10 Treatment of patients with Kawasaki disease with IVIG led to the inhibition of MMP-9 expression.11 In addition, in vitro experiments demonstrated that IVIG inhibited the constitutive secretion of MMP-9 in human endothelial cells,12 probably mediated by the binding of specific, natural antiendothelial cell antibodies present in IVIG preparations. The aim of the current study was to examine whether the observed beneficial effects of IVIG on tumor aggressiveness include inhibition of MMP-9 expression in monocytes/macrophages. Using in vitro differentiated human monocytic cell lines or peripheral blood monocytes (PBM), this study demonstrated that IVIG dose dependently inhibited MMP-9 secretion and mRNA expression and that both the F(ab)2 and Fc immunoglobulin fragments were involved in the inhibitory activity.
MATERIALS AND METHODS
The human monocytic cell lines, U937, THP-1, and MonoMac6, were cultured in RPMI-1640 supplemented with 10% decomplemented fetal calf serum, glutamine, and antibiotics (Biological Industries, Bet Haemek, Israel). PBM were separated by attachment to culture dishes for 2 hours. Cells were maintained at 37 °C in a humidified incubator with 5% carbon dioxide. For zymography assays, cells were cultured in microtiter plates (1.86 × 105 cells per well) and in 6-cm culture dishes (8 × 106 cells per dish) for RNA analysis. Cells were incubated with phorbol 12-myristate 13-acetate (PMA; 10 ng/ml) or TNF-α (10 ng/ml) for 24 hours for protein assays or 48 hours PMA for RNA assessment.13 Cells were washed and exposed to IVIG in the absence or presence of human F(ab)2 (prepared from IVIG preparations) or Fc fragments (Jackson Immune Research Laborotories, West Grove, PA) for 24 hours. Evaluation of the noncytotoxicity of reagents used was determined by XTT assay (according to the manufacturer's instructions; Biological Industries) whose results are proportional to cell number.
Purification of IVIG, F(ab)2 Fragments, and Human Serum Albumin (HSA)
Various IVIG batches from two manufacturers (5% Omrigam: Omrix, Rehovot, Israel; Sandoglobulin: Sandoz, Berne, Switzerland) were dialyzed overnight in Tris-buffered saline, pH 8.0 (0.05 M Tris-HCl, 0.5 M NaCl), diluted 1:5 in the same buffer, and filtered through a 0.45 millipore filter (Sartorius, Göttingen, Germany). To remove contaminants such as α-2-macroglobulin (α2M), IVIG preparations were purified three times on protein G-Sepharose columns (Amersham Pharmacia Biotech, Uppsala, Sweden) according to the manufacturer's instructions. Bound IgG was eluted with 0.1 M glycine-HCl, pH 3.5. The pH of elutes was immediately adjusted to 7.4 with 2 M Tris-HCl, pH 9. The IgG fractions were then dialyzed against phosphate-buffered saline (PBS), pH 7.4. The absence of α2M in the final IVIG product was verified by immune precipitation with specific anti-α2M2 antibodies followed by electrophoresis.
F(ab)2 fragments were prepared by IVIG digestion by 25 μg/mL pepsin (Sigma, St. Louis, MO) in 0.2 M sodium acetate buffer, pH 4.5, for 18 hours at 37 °C. The enzymatic reaction was terminated by addition of 2 M Tris-HCl, pH 8.0. The mixture was extensively dialyzed in PBS, pH 8.0, at 4 °C and remaining whole molecules were removed on a protein G-Sepharose CL-4B affinity column. Fragments and whole IgG were concentrated using Centricon filters (Amicon, Beverly, MA). The concentrated IgG was filter sterilized (0.22 μm filters; Sartorius) and dialyzed against sterile Dulbecco's modified eagle medium (DMEM; Biological Industries).
HSA, 25% solution (Kamada, Kibbutz Beit Kama, Israel), was dialyzed against PBS, filter sterilized, and dialyzed against sterile DMEM as described above. The purity of IgG, F(ab)2 fragments and HSA was confirmed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS- PAGE) under reducing and nonreducing conditions.
Northern Blot Analysis
Total cellular RNA was extracted from U937 with the use of Tri-Reagent (Medical Research Center, Cincinnati, OH) according to the manufacturer's instructions. The RNA (20 μg) was separated by electrophoresis on 1% agarose gel containing formaldehyde and its integrity visualized following ethidium bromide staining. The RNA was capillary transferred to Hybond membranes (Amersham, Buckinghamshire, UK) and hybridized to end-labeled MMP-9 oligomer probes (Biognostik, Göttingen, Germany) at 65 °C in 50 mM Tris-HCl (pH 7.5) 10% dextran sulfate, 1N NaCl, 1% SDS, and 100 μg/mL sheared salmon sperm DNA. Following autoradiography, membranes were stripped and exposed to glyceraldehyde 3 phosphate dehydrogenase (G3PDH) probe (Clontech, San Diego, CA) to normalize loading errors. Relative levels of the ratio of MMP-9:G3PDH label were determined following computerized densitometric assessment (BioImaging Gel Documentation System, Jerusalem, Israel) and using TINA software (Raytest, Straubenhardt, Germany).
Gelatinase (MMP-2 and MMP-9) activity was determined by SDS-PAGE. Proteins in the supernatant were separated on 8% polyacrylamide gels containing 0.1% gelatin (Sigma) and renatured by incubation in 2.5% Triton-X-100 for 30 minutes. Gels were incubated overnight in substrate buffer (50 mM Tris-HCl, pH 7.5) containing 10 mM CaCl2 and 0.05% Brij (Sigma) at 37 °C. A clear zone in the blue background of 0.5% Coomassie Brilliant Blue-stained gels demonstrated the presence of proteinase activity. Molecular weight markers (Bio-Rad Laboratories, Hercules, CA), positive controls, and recombinant MMP-2 and MMP-9 (Chemicon, Temecula, CA) were run with each gel. Computerized densitometry was used to evaluate relative enzymatic activity.
Statistical significance of results was determined using the Student t test when comparing two individual parameters and by analysis of variance when comparing parameters in a certain group. P values less than 0.05 were considered significant. Each experiment was repeated three times with good agreement between the results of individual experiments. The data shown are the mean ± standard deviation (SD).
RESULTS AND DISCUSSION
Monocytes and macrophages are archetypal secretors of gelatinases MMP-2 and MMP-9. Increased secretion of MMP-9 by in vitro differentiated monocytes in comparison to naive cells, consistent with the elevated functional activity attributed to differentiated tissue macrophages, has been reported.14, 15 We observed enhanced MMP-9 expression in the monocytic cell lines studied (i.e., U937, THP-1, and MonoMac6 cells) following 24–48 hours of exposure to the common differentiating agent, PMA (Fig. 1A, representative data for U937 cells). Similar results were also observed for PBM. The addition of several purified IVIG preparations from two manufacturers resulted in a dose-dependent inhibition of MMP-9 expression whereas no change was observed in the level of MMP-2 (Fig. 1B–D). The inhibition was dose dependent and highest at 6 mg/mL IVIG, a concentration that was noncytotoxic and noncytostatic as determined by the XTT assay and trypan blue exclusion (data not shown). These results support previous in vitro studies16 and an in vivo therapeutic dosage.10 HSA at the same concentrations did not influence MMP-9 (Fig. 1C). The inhibition may be specific to IVIG because α2M, a major inhibitor of MMPs that may be found in IVIG preparations, had been removed from the preparations (see Materials and Methods). Similar IVIG-induced inhibition was observed for all monocytic cell lines and for PBM.
The tumor microenvironment is rich in inflammatory cytokines such as TNF-α and IL-1. They are produced by stromal cells including monocytes/macrophages, which are potent producers of MMP-9.2 The observed PMA-induced elevation of the gelatinases may be related to its induction of TNF-α. IVIG also inhibited TNF-α–induced MMP-9 (Fig. 2).
IVIG suppresses immune activation, which may lead to attenuated MMP-9 gene expression through antigen-specific mechanisms involving the immunoglobulin F(ab)212, 17 domain and through nonspecific mechanisms attributed to the IgG-Fc domains. In addition, previous studies of autoimmune and inflammatory disorders have demonstrated that IVIG-induced suppression of immune activity may involve IgG binding to Fcγ receptors (FcγR) on monocytes.8 IVIG fragments were used to evaluate their possible involvement in MMP-9 suppression. F(ab)2, but not Fc fragments, led to a suppression of MMP-9 activity (Fig. 3). Figure 3 demonstrates that Fc, but not F(ab)2 fragments, led to abrogation of MMP-9 inhibition, depending on the relative concentrations of the fragments and IVIG concentrations (Fig. 3C).
The IVIG inhibition of MMP-9 was observed at the mRNA and protein levels (Fig. 4A,B). Similar to the results observed for the MMP-9 protein, the Fcγ fragments also reversed the IVIG-induced suppression of MMP-9 mRNA expression (Fig. 4C). These results suggest that the observed IVIG inhibitory effects on MMP-9 were mediated through Fc binding to its receptor and may necessitate the whole IgG molecule, including both the Fc and F(ab)2 components. The added fragments may have prevented possible binding to both antigen-specific and nonspecific receptors crucial for the activation of intracellular transduction pathways. Previous studies have demonstrated the existence of two types of FcγR, which either promote inflammatory processes (FcRγIII) or inhibit them (FcRγIIB). These two receptors are coexpressed on monocytes and their relative levels determine the final outcome of immune activity.18 The IVIG inhibitory signaling through FcRγIIB mediated its antiinflammatory activity.18 Blockade of this receptor by the added Fc fragments may have prevented IVIG binding, thus partially abrogating the inhibitory effects on MMP-9 production.
F(ab)2-dependent and other mechanisms may have influenced the IVIG-mediated inhibition of MMP-9 in the current study. Among these IVIG-mediated mechanisms may be the suppression of TNF-α (also induced by PMA)19 and the binding of F(ab)2 specifically to RGD (Arg-Gly-Asp) sequences,20 which are necessary for the activity of specific integrins (such as β1 and β3) implicated in the expression of MMP-9.16
In summary, preliminary but significant data point to a beneficial role for IVIG in the prevention of tumor spreading in mice and in the therapy of human cancers.10, 11 The mechanisms through which IVIG prevents tumor survival may be multifactorial and analogous to those affecting autoimmune and inflammatory conditions. The current findings are consistent with the beneficial results reported for IVIG therapy. The attenuation of MMP-9 expression in monocytes suggests an additional role for IVIG relevant to both tumor progression and inflammatory autoimmune processes.