Micrometastasis is one of the most serious problems in the treatment of cancer patients who undergo successful resection of the primary tumor. Even if metastasis is not detected at the time of the operation, micrometastases have developed and many patients suffer from metastasis several months or years after the operation. The lung is one of the most frequent target organs of hematogenous metastasis that restricts the prognosis of patients with malignant diseases, including lung, breast, stomach, colorectal and renal cancers. Therefore, treatment of clinically undetectable micrometastases may be important for these patients to control disease progression.
Several chemotherapeutic agents, such as paclitaxel, DOC, gemcitabine and vinorelbine, have been approved for the treatment of solid tumors over the last 5 years. Although these drugs combined with other cytotoxic agents, such as cisplatin, have begun to be used, patient survival has not been prolonged markedly.1, 2, 3, 4
MMPs play important roles in tumor invasion and angiogenesis by mediating degradation of the extracellular matrix and, hence, promote the formation of metastasis.5 Many tumors are characterized by overexpression of MMP activity, in either the tumor cells themselves or the adjacent stroma. In many types of neoplasm, higher levels of activated MMPs have been demonstrated in more invasive and/or metastatic tumors and may give prognostic information independent of stage.6, 7, 8, 9, 10 In particular, gelatinases (MMP-2 and MMP-9) closely correlate with the invasive and metastatic potentials of various cancers.11, 12 Therefore, MMPs may be ideal targets as a new class of inhibitory drugs for cancer patients, and a large number of MMPIs have been developed.13, 14, 15
The first-generation of MMPIs, including batimastat, has broad inhibitory activity against almost all MMPs and is not appropriate for oral administration. The second generation, such as marimastat [MMI270 (CGS27023A) and prinomastat, also has broad spectrum but can be orally administered. The third generation, including BAY12-9566, was designed to have a selective spectrum of MMP inhibition, and many of this generation do not inhibit MMP-1.14
ONO-4817 is a third-generation MMPI.16 It is a novel synthetic hydroxamic acid–based nonpeptide compound designed to be administered orally. ONO-4817 binds reversibly to the zinc-binding region of MMPs and has a selective inhibitory spectrum (MMP-2, MMP-8, MMP-9, MMP-12 and MMP-13 but not MMP-1 or MMP-7). Compared to BAY12-9566, ONO-4817 is more efficient at inhibition of MMP-2 and MMP-9. It has almost the same spectrum of MMP inhibition as prinomastat but less of an inhibitory effect on MMP-1 suppression. ONO-4817 suppressed the release of proteoglycan from the cartilage of the knee joints in an arthritis model of guinea pigs. Pharmacokinetic studies showed that plasma concentrations of ONO-4817 were more than 10 μmol/l at 1 hr and 1 μmol/l at 4 hr after oral administration at a dose of 100 mg/kg. However, the antimetastatic effect of this drug is unknown.16
In the present study, we explored the effect of a third-generation MMPI, ONO-4817, against established lung micrometastasis in human non-small cell lung cancer cell lines expressing MMP-1, MMP-2 and/or MMP-9. We further determined whether combined use of a cytotoxic agent, DOC, augmented the antimetastatic efficiency.
APMA, aminophenylmercuric acetate; DOC, docetaxel; FIZ, film in situ zymography; MMP, matrix metalloproteinase; MMPI, MMP inhibitor; ONO-4817, (2S,4S)-N-hydroxy-5-ethoxymethyloxy-2-methyl-4-(4-phenoxybenzoyl) aminopentanamide; TIMP, tissue inhibitor of metalloproteinase; TNF, tumor necrosis factor.
MATERIAL AND METHODS
ONO-4817 was synthesized by Ono Pharmaceutical (Osaka, Japan), as reported previously.16 The difference in MMP inhibitory activity of ONO-4817, BAY12-956617 and prinomastat18 is shown in Table I. DOC was obtained from Aventis Pharma (Tokyo, Japan).
Table I. Difference of Inhibitory Activities of MMPIS
Values are mean Ki (nM).
Human non-small cell lung cancer cell lines H226 (squamous cell carcinoma) and PC14PE6 (adenocarcinoma), a highly metastatic variant of PC14, were kind gifts from Dr. I.J. Fidler (M.D. Anderson Cancer Center, Houston, TX). Cell cultures were maintained in MEM supplemented with 10% heat-inactived FBS and gentamicin at 37°C in a humidified atmosphere of 5% CO2 in air. Assays were performed when cultured target cells were in the exponential phase of growth.
Measurement of MMP activity
Tumor cells (2 × 105/2 ml) were cultured in 24-well plates for 24 hr in MEM containing 10% FBS. Cultures were washed twice with PBS and incubated in serum-free MEM for another 48 hr, when the conditioned medium was harvested, spun at 1,000g for 10 min and used for the assay. To evaluate the effect of ONO-4817 on the inhibition of MMP activity, culture supernatants were incubated with various concentrations of ONO-4817 for 1 hr at 37°C. The resultant solutions were measured for activity of MMP-1, MMP-2 and MMP-9 by commercially available ELISA kits (Biotrak; Amersham, Aylesbury, UK), as reported previously.19 These ELISA kits use the pro form of detection enzymes, which can be activated by active MMP-1, MMP-2 or MMP-9, through a single proteolytic event. The activated detection enzyme can then be measured using a specific chromogenic peptide substrate. The brief assay protocol is as follows. Standards and samples (100 μl) were incubated overnight at 4°C in microtiter wells precoated with antibody for MMP-1, MMP-2 or MMP-9. Wells were washed 4 times with PBS; detection reagent (100 μl), containing APMA, and urokinase in 50 mM TRIS-HCl buffer (pH 7.6), containing 1.5 mM sodium chloride, 0.5 mM calcium chloride, 1 μM zinc chloride and 0.01% (v/v) BRIJ 35, were added to each well. Absorbance at 405 nm was measured with an MTP-120 microplate reader (Corona Electric, Ibaraki, Japan) before and after incubation at 37°C for 4–8 hr, and activity of MMP-1, MMP-2 and MMP-9 was calculated according to the manufacturer's instructions. Detection limits of MMP-1, MMP-2 and MMP-9 were 0.1, 0.5 and 0.125 ng/ml, respectively.
In vitro proliferation assays
Proliferation of H226 and PC14PE6 cells was measured by the MTT dye reduction method20 and Cell Counting Kit-8 (Wako, Osaka, Japan), respectively. Briefly, 5 × 103 cells were plated into wells of 96-well plates, treated with various concentrations of drugs and incubated for 72 hr at 37°C. For H226 cells, 50 μl of the MTT stock solution (2 mg/ml) were added to each well, and cells were further incubated for 2 hr at 37°C. Then, culture media were removed, and 100 μl of DMSO were added to dissolve the dark blue crystals. For PC14PE6 cells, 10 μl of Cell Counting Kit-8 solution, which produces WST-8 formazan, were added to each well; and the cells were further incubated for 2 hr at 37°C. Absorbance were measured with the MTP-32 Microplate Reader at test and reference wavelengths, respectively, of 550 and 630 nm for H226 cells and 450 and 630 nm for PC14PE6 cells.
Male athymic nude mice (6–8 weeks old) were obtained from Clea (Osaka, Japan) and maintained under specific pathogen-free conditions throughout the study. Experiments were performed according to the guidelines of our university.
Effect of ONO-4817 and DOC on experimental lung metastasis
To produce experimental lung metastasis, viable tumor (H226 or PC14PE6) cells suspended in 300 μl of PBS were injected into the lateral tail vein. As reported previously, H226 and PC14PE6 cells developed micrometastasis in the lung by day 7.22 Mice were i.v.-injected with PBS or DOC (8 mg/kg for mice inoculated with H226 cells and 2 mg/kg for mice inoculated with PC14PE6 cells) on day 7 and given food mixed with ONO-4817 (1%) or not from day 7 until they were killed. Optimal doses of DOC for each cell line were determined in preliminary experiments (data not shown). After the indicated periods, mice were killed and metastasis formation to the lungs and lung weight was examined. Since some of the mice inoculated with PC14PE6 cells developed pleural effusion in addition to lung metastasis, the effect of ONO-4817 treatment on the formation of pleural effusion was also determined. The number of lung metastases was determined macroscopically after lung staining with Bouin's solution.
Fresh lung tumor specimens were dissected from tumor-bearing mice, embedded in crymold O.C.T.4583 compound (Miles, Elkhart, IN) and immediately frozen on dry ice. These frozen blocks were then sliced sequentially using a Tissuetek II cryostat microtome (Miles) to prepare serial frozen thin sections (8 μm thick). Thin sections were placed on FIZ-GN, a polyethylene terephthalate-base film coated with cross-linked gelatin, 7 μm thick (Fuji, Tokyo, Japan).21 As a negative control, FIZ-GI films treated with 1,10-phenanthroline, a chelating agent functioning as an MMPI, were also used to fix one of the serial sections. A set of FIZ-GN and FIZ-GI films with sections was incubated in a moist chamber (Cosmo Bio, Tokyo, Japan) at 37°C for 6 hr. After incubation, specimens were stained in a Coplin jar over 4 min with Biebrich scarlet (Aldrich, Milwaukee, WI), which was diluted to a 1.0% solution with distilled water. Then, these films were decolored for 10 min with distilled water, followed by counterstaining with Mayer's hematoxylin solution (Muto, Tokyo, Japan) for 2 min and rinsed with water for 10 min. Degraded gelatin was not stained with Biebrich scarlet, and an area of gelatinolytic activity was visualized as white to weak red on a red background. In the serial sections, if FIZ-GN revealed unstained areas and FIZ-GI did not, this suggested that gelatinolytic activity originated from MMPs but not from other proteases, such as trypsin. For verification with FIZ films, one of the serial sections was fixed with 4% paraformaldehyde fixative and rinsed thoroughly with PBS, followed by staining with Gill's hematoxylin and eosin solution (Wako). To analyze gelatinolytic activity, the photographic image under the light microscope was scanned with a Nikon (Tokyo, Japan) Scantouch 210 scanner.
The significance of differences of in vivo findings was analyzed by the Mann-Whitney U-test. Survival of mice was analyzed by the log rank test.
Expression of MMPs by human lung cancer cell lines in vitro
We first determined the activity of MMP-1, MMP-2 and MMP-9 constitutively produced by human lung cancer cell lines. PC14PE6 cells expressed high levels of MMP-1 and MMP-2 and a low level of MMP-9 (Table II). H226 cells expressed MMP-2 activity but not discernible levels of MMP-1 or MMP-9.
Table II. Detection of MMP Activitiy in Cultured Supernatants
MMP activity (ng/ml) in cultured supernatants of human lung carcinoma cell lines was measured by a MMP activity assay system. PC14PE6, adenocarcinoma cells; H226, squamous cell carcinoma cells.
Inhibition of MMP activity by ONO-4817 in vitro
We next examined the inhibitory effect of ONO-4817 on MMP activity in vitro. While pretreatment of culture supernatants of PC14PE6 cells with ONO-4817 had no effect on MMP-1 activity, the same treatment inhibited activity of MMP-2 and MMP-9 in a dose-dependent manner (Fig. 1), indicating selectivity of ONO-4817 on inhibition of MMP activity. In a parallel experiment, pretreatment of culture supernatants of H226 cells with ONO-4817 also inhibited MMP-2 activity in a dose-dependent manner.
Effect of ONO-4817 with DOC on proliferation of tumor cell lines in vitro
ONO-4817 up to 10 μM did not affect the proliferation of PC14PE6 or H226 cells, while DOC inhibited the proliferation in a dose-dependent manner (Fig. 2). Addition of ONO-4817 to DOC did not augment the susceptibility of the 2 cell lines to DOC.
Therapeutic effect of ONO-4817 and DOC against established micrometastasis
We previously reported that i.v. injection of PC14PE6 cells and H226 cells developed micrometastasis in the lung by day 7.22 To examine the therapeutic effect of ONO-4817 and DOC on lung micrometastasis, nude mice were inoculated with PC14PE6 or H226 cells on day 0. Mice were i.v.-injected with DOC on day 7 and given food mixed with 1% ONO-4817 or not from day 7 until death. Mice were killed 5–8 weeks later, and metastasis formation to the lung was evaluated. H226 cells developed numerous small lesions (<2 mm) on the surface of the lung in control mice, as reported previously.22 Although more than 150 lesions were detected on the lung even in mice treated with ONO-4817 alone or DOC alone, each lesion was clearly smaller in treated than in control mice (Table III). Consistent with the size of lung lesions, the lung weight of ONO-4817- or DOC-treated mice was significantly lower than that of control mice. Moreover, the lung weight of mice treated with combination therapy (ONO-4817 and DOC) was significantly lower than that of mice treated with ONO-4817 alone or DOC alone (Table III), suggesting that the lowest tumor burden was in the combination therapy group, though the number of lung lesions was still higher than 150.
Table III. Effect of Treatment with ONO-4817 and DOC on Lung Metastasis by H226 Cells in Nude Mice
PC14PE6 cells injected i.v. developed numerous lung metastases and bloody pleural effusion. Treatment with ONO-4817 alone or DOC alone significantly reduced the number of lung metastases, lung weight and volume of pleural effusion compared to controls. Each therapy was so effective that combination therapy did not show superiority over either alone (Table IV). However, the survival of mice treated with combined therapy was significantly longer compared to all other groups (p < 0.05, log rank test) (Fig. 3). During these treatments, body weight and food intake were not different among the 4 groups (Fig. 4), indicating the feasibility of combined therapy of ONO-4817 and DOC in our model.
Table IV. Effect of Treatment With ONO-4817 and DOC on Lung Metastasis by PC14PE6 Cells in Nude Mice
Weight of lungs (g)
PC14PE6 (1 × 106) cells were inoculated i.v. into nude mice on day 0. Mice were given food mixed with ONO-4817 (1%) or not on days 7–36. Mice were treated with PBS or DOC (2 mg/kg) on day 7 and killed on day 36. Treatment with ONO-4817 significantly inhibited the formation of lung metastasis and pleural effusion compared to no treatment.
Since lung lesions of PC14PE6 cells in mice treated with ONO-4817, DOC or both were very rare on sections, histologic analyses were performed using tumors with H226 cells. MMP activity in the in vivo conditions was determined by means of FIZ (Fig. 5). Lung tumors with H226 cells in the control group showed gelatinolytic activity on GN film, while nontumor parenchyma of the lung did not. Gelatinolytic activity in lung colonies by H226 cells was not detected on FIZ-GI film, suggesting that the gelatinolytic activity was due to MMPs. Gelatinolysis of lung lesions in the ONO-4817-treated group was much lower than that of the lung lesions in the control group, suggesting that ONO-4817 successfully inhibited MMP activity of lung lesions in vivo.
It has been suggested that MMPs degrade the extracellular matrix of the basement membrane and regulate the formation of metastasis by stimulation of tumor invasion and angiogenesis.7 In our study, we demonstrated that a third-generation MMPI, ONO-4817, had the potential to inhibit the growth of established lung micrometastasis. Therapeutic efficacy in terms of tumor burden and survival of mice was further augmented when combined with a cytotoxic agent, DOC.
Metastasis formation can be facilitated if MMPs were expressed at sites of tumor growth, irrespective of their source (tumors or nontumor organ parenchyma) (unpublished data). Since nontumor parenchyma in the lung have no detectable levels of MMPs (Fig. 5), MMP activity derived from tumor cells may be essential for development of lung metastasis. We evaluated the activity of 3 (MMP-1, MMP-2, MMP-9) MMP family members using ELISA kits. Since (i) both PC14PE6 (expressing MMP-1, MMP-2 and MMP-9) and H226 (expressing MMP-2) cells developed lung metastasis, (ii) ONO-4817 did not inhibit MMP-1 activity in vitro and (iii) treatment with ONO-4817 inhibited development of lung metastasis by both PC14PE6 and H226 cells, MMP-2, rather than MMP-1, may be important for production of hematogenous lung metastasis. This hypothesis is supported by evidence that overexpression of MT1-MMP, which is essential for activation of pro-MMP-2, resulted in augmentation of experimental lung metastasis by mouse lung carcinoma cells.23
Several reports have shown that the antitumor effect of many antiangiogenic drugs and MMPIs was stronger when treatment was started earlier,14 suggesting that MMPs may play critical roles in early aspects of cancer progression (local invasion and micrometastasis). In our metastasis model, i.v.-inoculated PC14PE6 and H226 cells were distributed to the lung capillary bed by 4 hr (data not shown) and developed lung micrometastasis by 7 days.22 ONO-4817 started on day 7 (after establishment of micrometastasis) inhibited the formation of lung metastasis by these 2 cell lines, indicating that ONO-4817 inhibited the growth of established lung micrometastasis, rather than inhibition of tumor cell distribution or extravasation. Since we could not detect any difference in intratumoral vascularization by immunohistochemistry of CD31 staining in the lung lesions between the control and ONO-4817-treated groups (data not shown), angiogenesis might not be involved in the inhibition of growth of lung micrometastasis by ONO-4817. Further examinations are warranted to clarify the mechanisms of the inhibition by ONO-4817.
To date, MMPIs have shown little clinical benefit when used as monotherapy in patients with advanced diseases.24, 25, 26, 27 Thus, combined use of MMPIs with other modalities is a reasonable strategy for cancer treatment. Haq et al.28 reported that combination therapy of batimastat with gemcitabine significantly reduced the tumorigenicity of orthotopically implanted human pancreatic cancer cells and improved survival of tumor-bearing mice compared to controls or mice treated with either agent alone. In addition, activated MMP-2 levels of tumor cells and mouse serum were most decreased in the group treated with combination therapy. They speculated that inhibition of MMP-2 activity was due to both direct inhibition by batimastat and indirect inhibition by gemcitabine, reducing tumor growth kinetics. In the present study, we also demonstrated that combined therapy of ONO-4817 with DOC was more effective against lung metastasis of tumor cells expressing MMPs. Although ONO-4817 did not affect the susceptibility to DOC of PC14PE6 or H226 cells in vitro, it is possible that a single injection of DOC reduced the number of lung micrometastatic colonies and continuous treatment with ONO-4817 suppressed the growth kinetics of tumor cells, as speculated by Haq et al.28
The major dose-limiting toxicity of broad-spectrum MMPIs involves time- and dose-dependent musculoskeletal adverse events, including joint pain and stiffness, which limit both the dose level administered and the duration of therapy.26, 27, 29 These adverse events are thought to be due to toxicity associated with inhibition of MMP-1.26, 27 Because ONO-4817 is a third-generation MMPI designed to have a selective spectrum of MMP inhibition (without MMP-1),16 it may have fewer adverse effects than the first- or second-generation MMPIs. In addition, MMPIs have fewer hematologic adverse events23, 26, 29 than many cytotoxic chemotherapeutic agents. Therefore, combined use of MMPIs and cytotoxic chemotherapeutic agents may be ideal in terms of not only efficacy but also adverse events. Indeed, the combined therapy used in our study was feasible and did not affect food intake or body weight of tumor-bearing mice.
Wielockx et al.30 reported that batimastat suppressed apoptosis and necrosis of hepatocytes and, hence, prevented the lethal hepatitis induced by TNF. Moreover, batimastat also prevented lethality in TNF/IFN-γ therapy of melanoma cell-bearing mice while maintaining antitumor efficacy. Thus, it is possible that MMPIs reduce undesirable adverse events and augment the therapeutic efficacy of certain modalities, including cytokine therapy.
A randomized, double-blind, placebo-controlled study showed that marimastat, a second-generation MMPI, prolonged overall survival and progression-free survival of patients with inoperable gastric cancer, especially those who had previously received chemotherapy and did not have clinically detectable metastasis at the time of enrollment.31 We reported that ONO-4817 inhibited progression of established lung micrometastasis by tumor cells expressing MMPs and that therapeutic efficacy was further augmented when combined with DOC. Collectively, combined use of ONO-4817 with chemotherapy may be helpful as adjuvant therapy for patients with resected primary tumors who have no detectable metastasis.