Obesity has become one of the most critical problems in human health, because it can lead to more serious disorders such as cardiovascular disease, pulmonary disorders and metabolic syndrome.1 Attention has recently been focused on the influence of obesity on carcinogenesis and the progression of cancer. A cohort study in the United States revealed that an increased body-mass index is significantly associated with higher rates of death because of cancer.2 In addition, obesity increases the risk of cancer developing in the prostate, colorectum, breast, endometrium and elsewhere.3, 4, 5, 6, 7 Similarly, the risk of renal cell cancer is increased in patients with diabetes mellitus, a disease closely associated with obesity.8 These observations are well correlated with the finding that the carcinogen-induced multiplicity of premalignant colonic lesions is increased in db/db obese and diabetic mice.9
Metastasis is the major cause of mortality in cancer patients. Maehle et al. reported that breast cancer patients with obesity had a higher risk of lymph node metastases.10 Although evidence has accumulated of an association of cancer with obesity, the influence of obesity and diabetes on cancer metastasis remains to be investigated. In addition, the mechanisms by which obesity promotes the development and progression of cancer are still unknown.
Material and methods
Male C57BL/6, C57BL/KsJ-db/db (db/db) and C57BL/6J-ob/ob (ob/ob) mice, 5–6 weeks old, were maintained in a specific pathogen-free barrier facility under laminar airflow conditions with a 12-hr light/dark cycle, a controlled humidity of 55% ± 5% and a temperature of 22°C ± 1°C. This study was conducted in accordance with the standards outlined in the Guidelines for the Care and Use of Laboratory Animals of the University of Toyama.
B16-BL6 melanoma and Lewis lung carcinoma (LLC) cell lines were maintained as monolayer cultures in Eagle's minimal essential medium (EMEM; GIBCO BRL, Grand Island, NY) supplemented with 10% fetal calf serum (FCS), 2 mM of L-glutamine, 100 units/ml of penicillin and 100 μg/ml of streptomycin. All cultures were kept at 37°C in a humidified atmosphere of 5% CO2/95% air.
Establishment of B16-BL6 cells stably expressing firefly luciferase
B16-BL6 cells were transfected with pGL3-promoter and pCI-neo (Promega, Madison, WI). Cells were selected with EMEM containing 550 μg/ml of antibiotic G418 sulfate (Promega) and then a cell line (BL6-luc) was established by limiting dilution. The metastatic potential of BL6-luc cells is comparable with that of the parental B16-BL6 cells.
Experimental lung metastasis
B16-BL6 and LLC cells were harvested with trypsin-EDTA and EDTA, respectively, and then resuspended in cold PBS. For the pulmonary metastasis assay, B16-BL6 (0.7–1.0 × 105 cells) or LLC (2 × 105 cells) cells were injected into the tail vain. They were killed on day 14 after the inoculation and the tracheas with attached lungs were removed and fixed in picric acid. Metastatic foci visible on the surface of the lung were enumerated under a microscope. For B16-BL6 cells, the diameter of the colonies was measured.
Tumor proliferation in vivo
The procedure for the intrapulmonary inoculation of LLC cells was reported previously.11 In brief, the left chest of anesthetized mice was incised ∼5 mm, and 20 μl aliquots of an LLC suspension (2 × 103 cells) admixed with 20 μg of Matrigel™ basement membrane matrix (BD Biosciences, San Jose, CA) were injected into the left lung parenchyma through the intercostal space. Mice were killed 15 days after implantation, and the weight of the tumors was measured. Footpad implantation of tumor cells was performed as follows; B16-BL6 cells (1 × 105) were injected subcutaneously into the right footpad. Tumor size was monitored by the anteroposterior diameter with calipers three times a week. And 14 days after implantation, the tumor size was evaluated. Tumor volume at the implanted site was calculated with the following formula: Tumor volume (mm3) = 1/2 × (long diameter) × (short diameter).2
In vivo imaging
Mice were injected via a tail vein with 1 × 105 BL6-luc cells. The hair of the anterior chest was shaved because black hair reduces the luminescence. Mice were anesthetized with isoflurane and injected intravenously (tail vein) with 3 mg of luciferin (Promega)/200 μl of PBS. Five minutes after the luciferin injection, mice were imaged for 60 sec using the IVIS imaging system (Xenogen, Alameda, CA). Photons emitted from specific regions were quantified using Living Image software (Xenogen). In vivo luciferase activity is expressed as photons/second/cm2/steradian (sr). The signal was measured 1, 3, 5, 7, 10 and 14 days after the tumor inoculation.
Mice were injected via a tail vein with 1 × 105 BL6-luc cells and were killed after 5 min or 6 hr. The removed lungs were homogenized with passive lysis buffer for luciferase assays (Promega). After centrifugation at 14,000 rpm for 10 min at 4°C, 100 μl of supernatant was incubated with 300 μl of PicaGene luciferase substrate solution (Toyo-inks, Tokyo, Japan), and the luciferase activity in the mixture was measured using the IVIS system.
Spleen was removed and homogenized in 10 ml of RPMI-1640 medium (GIBCO BRL) with 3% FCS. Splenocytes were centrifuged and resuspended in 1 ml of red blood cell lysis solution. After 1 min, the cell suspension was centrifuged and the pellet was washed twice and fixed with PBS containing 2% FCS and 0.02% sodium azide. For flow cytometry, the splenocyte suspensions were blocked with anti-Fcγ RII/III Ab and stained with fluorescein isothiocyanate-conjugated anti-NK1.1 (PK136) and peridinin chlorophyll-a protein-conjugated anti-CD3ε (145-2C11). The stained cells were analyzed with a FACScalibur, and the data were analyzed using CELLQuest™ software (BD Biosciences).
Treatment with leptin, anti-asialo GM1 antibody and pioglitazone
The ob/ob mice received a daily intraperitoneal injection of murine recombinant leptin (rmLPT; ProSpec-Tany TechnoGene, Rehovot, Israel) at a dose of 2 μg/g body weight for 14 days. The db/db mice were injected intravenously with 40 μg of reconstituted rabbit anti-asialo-GM1 antibody (Wako Chemicals, Osaka, Japan) 4 and 2 days before the tumor inoculation. Pioglitazone, a specific ligand of peroxisome proliferator-activated receptor-γ (PPAR-γ), (kindly provided by Takeda Pharmaceutical Co., Osaka, Japan) was dissolved in 0.5% methylcellulose and administered daily by oral gavage for 14 days at a dosage of 15 mg/kg body weight. After these treatments, the mice were injected via a tail vein with 1 × 105 BL6-luc cells and were killed 5 min or 6 hr later. Lung and spleen were removed to measure luciferase activity and the number of NK cells, respectively.
Plasma insulin and glucose
Plasma insulin was measured by enzyme-linked immunosorbent assay kit (Morinaga Institute of Biological Science, Yokohama, Japan). Plasma glucose was determined using commercially available kit, based on the glucose oxidase method (Wako Chemicals).
The statistical significance of differences between groups was calculated by applying Mann-Whitney U-test or Student's two-tailed t test. All columns in the figure are expressed as the mean ± S.D. Statistical significance was defined as a p value < 0.05.
The experimental pulmonary metastasis in db/db and ob/ob mice
To examine the influence of obesity on cancer metastasis, we employed an experimental model of pulmonary metastasis using typical obese mice. B16-BL6 cells, a high metastatic mouse melanoma cells, were injected into the tail vein of db/db (Fig. 1a) and ob/ob (Fig. 1c) mice, and then tumor colonies in the lung were enumerated on day 14. The number of colonies was increased remarkably in db/db and ob/ob mice when compared with the C57BL/6 control mice. Metastasis is similarly promoted in db/db (Fig. 1b) and ob/ob (Fig. 1d) mice on the intravenous inoculation of LLC cells, a lung cancer cells, indicating that the increase in metastasis is independent of cancer cell type. These results were summarized in Figure 1e that shows the median number and range of metastatic lung nodules.
The tumor growth in db/db and ob/ob mice
To address the mechanisms of the obesity-mediated enhancement of metastasis, we first focused on tumor growth in the lung. Fourteen days after the experimental pulmonary metastasis of B16-BL6 cells, the diameters of tumor colonies in C57BL/6 and db/db mice were measured. Although the number of colonies in the lung was extremely increased, the ratio of colony size (Fig. 2a) and the average diameter (Fig. 2b) were not significantly different between these groups. In addition, tumor growth at the implanted site was comparable when B16-BL6 cells were inoculated orthotopically in the footpad (Fig. 2c). Similarly, direct implantation of LLC cells into the lung revealed that the weight of primary tumor was not significantly different between C57BL/6 and db/db mice (Fig. 2d).
Promotion of the early phase of metastasis in db/db and ob/ob mice
Recently, the optical bioluminescent imaging, using a cooled charged-coupled device camera, can chase the signal of luminescent-labeled cells in living animals.12, 13 Then, metastatic processes in vivo were investigated by using B16-BL6 cells stably expressing luciferase (BL6-luc cells). Pulmonary metastasis in living mice was evaluated with the in vivo imaging system. Luciferase expression was evident on day 5 in db/db mice and day 10 in control mice, and was stronger in db/db mice than C57BL/6 mice during the experiment (Fig. 3a).
We next tried to investigate the metastatic behavior within 6 hr of the injection. The luciferase activity in lung homogenates was measured to compare the cell number more quantitatively. The luciferase activity at 5 min after the injection of BL6-luc cells was slightly stronger in db/db than the control mice (Fig. 3b). The luciferase activity of all groups, especially control C57BL/6 mice, was decreased at 6 hr after the injection (Figs. 3b and 3c). However, the luciferase activity of db/db and ob/ob mice was significantly stronger than that of control mice (Figs. 3b and 3c).
The effect of pioglitazone in the pulmonary metastasis in db/db mice
Pioglitazone is a thiazolidinedione derivative of insulin sensitizer that is related to differentiation of adipocytes.14 To explore the role of insulin resistance on the metastasis, pioglitazone was administered for 2 weeks and the BL6-luc cells were injected on the next day, because PPARγ agonist has direct antiproliferative and antimetastatic activities.15, 16 As reported previously, pioglitazone improved hyperinsulinemia (Fig. 4a) in db/db mice. However, it was not able to suppress the early stages of metastasis (Fig. 4b) and the formation of metastatic colonies in the lung during 14 days (Fig. 4c). These results demonstrate that insulin resistance may not be a major cause of the increased metastasis in db/db mice.
Suppression of metastasis by replacement of ob/ob mice with recombinant leptin
Treatment with recombinant mouse leptin (rmLPT) has been shown to rescue the phenotypes of the leptin deficient ob/ob mice.17, 18 To evaluate the effects of rmLPT on the increased metastasis, 5-week-old ob/ob mice were given a daily intraperitoneal injection of rmLPT for 2 weeks. The increase in body weight during the treatment was almost completely reduced to the level of control C57BL/6 mice (Fig. 5a). One day after the final injection, BL6-luc cells were injected intravenously and luciferase activity in the lung homogenates was measured at 6 hr. The increased luciferase activity was reduced to the level of control mice (Fig. 5b). These results indicate that the 2-week treatment with rmLPT rescued the phenotypes of ob/ob mice as well as the increased metastasis.
Relationship between NK cells and pulmonary metastasis at early stage
Natural killer (NK) cells play a critical role in antitumor immunity and leptin has been shown to be involved in the differentiation and activation of NK cells. In the leptin rescue experiment in ob/ob mice, flow cytometric analysis demonstrated that splenic NK cell numbers were decreased and rmLPT treatment recovered cell numbers to the level in wild-type mice (Fig. 5c). To explore the role of NK cells, db/db mice received injections of an anti-asialo-GM1 antibody to deplete NK cells (Fig. 6). Two days after the final injection, BL6-luc cells were injected intravenously. At 5 min after tumor inoculation, luciferase activities in C57BL/6 and db/db mice were not affected. In contrast, the elimination of cancer cells during 6 hr was blocked by the antibody, but not by normal rabbit serum, in both C57BL/6 and db/db mice. Of note, the difference between db/db and control mice disappeared.
In the present study, we found that experimental pulmonary metastasis is markedly promoted in leptin-deficient (ob/ob) and leptin receptor-deficient (db/db) mice. The metastasis is similarly promoted on the intravenous inoculation of B16-BL6 (Fig. 1a and 1c) and LLC cells (Fig. 1b and 1d), indicating that the increase in metastasis is independent of cancer cell type. We next focused on tumor growth in the lung to address the mechanisms of the obesity-mediated enhancement of metastasis. The results in Figure 2 show that there were no significant differences on the diameter of metastatic tumor colonies in the lung (Fig. 2a and 2b) and the volume of primary tumor mass in the orthotopic implantation models (Fig. 2c and 2d). Considering the processes of lung metastasis, these results suggest that the increased metastasis is not due to accelerated tumor growth, but rather reflects events before the extravasation of tumor cells released from the primary site into target organs. The in vivo imaging using a stable transfectant of lucifearase gene made it apparent that the increased pulmonary metastasis was occurred within 5 days in db/db mice (Fig. 3a). Monitoring the luciferase activity in lung homogenates revealed that obesity-promoted metastasis occurred mainly within 6 hr after the injection of cancer cells. The ratio of luciferase activity in control vs. obese mice was similar to the difference in the number of metastatic colonies, which raises the idea that the antitumor immunity is impaired during the early stage of metastasis in obese mice.
The role of leptin in the modulation of immune system is becoming increasingly evident.19, 20, 21 A deficiency of leptin is responsible for severe impairments of both innate and acquired immunity, including the differentiation and activation of NK cells.22 During the early phase of metastasis, cancer cells are recognized and destroyed by the host's immune system, especially by NK cells.23, 24 Therefore, we have focused on NK cell function in obesity-promoted metastasis. Splenic NK cell numbers were decreased in db/db and ob/ob mice, and 2-week leptin replacement in ob/ob mice recovered NK cell numbers to the level in wild-type mice. It has been shown that leptin also affects the activation as well as differentiation of NK cells. We employed another protocol in which a single leptin treatment was administered just after the inoculation of BL6-luc cells; however, no inhibition of obesity-induced metastasis was detected (data not shown). These results suggested that leptin restored a normal peripheral NK-cell pool but not NK cell activity in the effecter phase of cancer cell clearance. Moreover, depletion of NK cells with the anti-asialo-GM1 antibody demonstrated that NK cells are involved in the increased metastasis at 6 hr. Collectively, these results demonstrate that metastasis is markedly increased in obese mice mainly because of decreased NK cell differentiation during the early phase of metastasis.
In 1983, Margules's group reported contradictory articles that show resistance to spontaneous metastasis of B16 melanoma in ob/ob mice.25, 26 When the mice were inoculated subcutaneously with melanoma cells, the frequency of lung metastasis was greatly reduced in obese than in lean littermates. An enhanced proliferative response of splenic lymphocytes to T-cell mitogens, but not to a B-cell mitogen, was involved in the mechanism of reduction. On the contrary, we found the promotion of lung metastasis in an experimental lung metastasis model by inoculating cancer cells intravenously. The reason for this discrepancy is not solved, but provably derived from the different route of inoculation of cancer cells. The spontaneous metastasis model includes multiple steps to form tumor colonies in the lung. In contrast, our results suggested a role of NK cell-mediated antitumor immunity in the early stage of metastasis. Evidences have recently been accumulating that leptin plays an important role in the regulation of both innate19, 20, 22 and acquired immunity.21 In addition, Hirose et al. reported enhancement of chemical-induced carcinogenesis in db/db mice.9 These results indicate impaired immune surveillance against tumor cells in obese mice, which support our findings that deficiency of the leptin/leptin receptor system causes the severe pulmonary metastasis in the experimental model.
Leptin deficiency could be detected in patients with cancer-related cachexia,27 and low adipocyte mass causes a reduction in serum leptin in undernourished individuals.28 In the leptin-deficient condition, an impaired type I helper T cell (Th1) immune response increases the risk of infection.21 Also, the serum leptin concentration rises in proportion to the amount of body fat in obese human subjects.29, 30 The notion of leptin resistance, a state of reduced responsiveness to leptin in various tissues, has recently come to be accepted,31, 32 and may explain why obesity causes immune dysregulation including tumor immunity. Although Dovio et al. reported that the activity of NK cells did not differ between obese and nonobese humans,33 it is not clear whether leptin resistance reflects the NK cell activity in obese individuals. In addition, obesity has been associated with an increased susceptibility to infection and bacteremia,34 suggesting that it causes immune deficiency in humans.
The functional relation between cancer and leptin has been discussed in recent clinical reports showing that elevated plasma leptin concentration is linked to an increased risk of carcinogenesis including prostate cancer.6, 7, 35 The effect of leptin on cancer cell in vitro has also been reported. Leptin stimulates the proliferation of breast, prostate and colorectal cancer cell lines.35 However, there are equivocal clinical results on the relation between the risk of cancer and plasma leptin concentration.5, 35 In any case, although the loss-of-function animal experiments have clearly demonstrated that the defect in the leptin signaling causes the increased carcinogenesis and tumor metastasis, it is essential to further investigate the role of leptin resistance in tumor progression including metastasis in humans.
In summary, we found severe experimental pulmonary metastasis in animal models of obesity and diabetes. Our findings raise the possibility that life-style-related diseases are associated with an increased risk of cancer metastasis. It is necessary to directly test this hypothesis in humans. In addition, further mechanistic investigations of obesity-promoted metastasis in animal models will aid our understanding of the risk involved.
The authors are grateful to Dr. Shigeaki Watanabe at SC BioSciences Corporation for technical assistance with the IVIS system.