|Home | About | Journals | Submit | Contact Us | Français|
Dietary energy restriction has been shown to repress both mammary tumorigenesis and aggressive mammary tumor growth in animal studies. Metformin, a caloric restriction mimetic, has a long history of safe use as an insulin sensitizer in diabetics and has been shown to reduce cancer incidence and cancer-related mortality in humans. To determine the potential impact of dietary energy availability and metformin therapy on aggressive breast tumor growth and metastasis, an orthotopic syngeneic model using triple negative 66cl4 tumor cells in Balb/c mice was employed. The effect of dietary restriction, a standard maintenance diet or a diet with high levels of free sugar, were tested for their effects on tumor growth and secondary metastases to the lung. Metformin therapy with the various diets indicated that metformin can be highly effective at suppressing systemic metabolic biomarkers such as IGF-1, insulin and glucose, especially in the high energy diet treated animals. Long-term metformin treatment demonstrated moderate yet significant effects on primary tumor growth, most significantly in conjunction with the high energy diet. When compared to the control diet, the high energy diet promoted tumor growth, expression of the inflammatory adipokines leptin and resistin, induced lung priming by bone marrow-derived myeloid cells and promoted metastatic potential. Metformin had no effect on adipokine expression or the development of lung metastases with the standard or the high energy diet. These data indicate that metformin may have tumor suppressing activity where a metabolic phenotype of high fuel intake, metabolic syndrome, and diabetes exist, but may have little or no effect on events controlling the metastatic niche driven by proinflammatory events.
Epidemiological studies have established a clear relationship between breast cancer and obesity. Specifically, obesity and weight gain have been directly correlated to increased breast cancer risk and mortality [1, 2]. A large prospective study has defined an increased risk of mortality from breast cancer up to 2.12-fold for women with increased body mass index (BMI) compared to normal weight individuals .
The relationship between excess dietary fuel intake and cancer has been modeled in a number of transplant and endogenous cancer animal models. In a review by Freedman et al., it was highlighted that higher caloric intake was sufficient to significantly increase the development of mammary tumors in rats and mice . More recently, diet-induced obesity has been shown to promote the incidence of mammary tumor development in the MMTV-neu model of breast cancer . However, few studies have directly evaluated primary mammary tumor growth and metastasis in conjunction with increased caloric intake.
Caloric restriction has been demonstrated to be an effective method to decrease cancer incidence and tumor growth including mammary tumors (reviewed by ). While a number of studies have demonstrated reduced mammary tumorigenesis associated with caloric restriction or dietary energy restriction, few have investigated dietary energy restriction as a direct treatment for primary or metastatic breast cancer. However, models of primary prostate and brain tumor growth have shown to be effectively suppressed with dietary restriction [7, 8].
A number of therapeutic compounds have been investigated to replace direct modulation of caloric or energy intake, collectively termed caloric restriction mimetics (CRM) . One compound suggested to be a useful CRM candidate is metformin. Metformin (1,1-dimethylbiguinide hydrochloride) is a drug used as a first-line treatment for type-2 diabetes , where it acts by increasing insulin sensitivity. Increased insulin signaling results in reduced insulin levels, reduced hepatic gluconeogenesis and increased glucose uptake by muscle . Epidemiological studies have revealed that metformin therapy decreased cancer incidence and the risk of cancer-related mortality in diabetics when compared to those treated with sulfonylureas or other therapies [12, 13].
Metformin has been shown to significantly repress the growth of a number of cancer cell lines including breast cancer cell lines in vitro [14–17]. Systemic treatment with metformin in pre-clinical animal models has also demonstrated some benefit, such as the repression of pancreatic cancer progression in hamsters and the decrease of Her-2/neu and ApcMin/+ driven tumorigenesis in mice [18, 19]. Treatment with metformin effectively repressed xenograft tumor models including, LNCap prostate and p53-deficient colon cancer and was able to attenuate the increases in Lewis Lung Carcinoma growth promoted by a high energy diet [16, 20]. Interestingly, using a triple negative human breast tumor xenograft model with a standard diet, systemic metformin treatment resulted in increased tumor size primarily mediated through increased angiogenesis . Finally, Hirsch et al. recently reported that while metformin alone has a minimal effect in a human model of breast tumor growth in vivo, the combination of metformin and doxorubicin treatment represses tumor growth to near elimination . Interestingly, this report also demonstrates that metformin may be selectively targeting cancer stem cells within these tumors, suggesting that combination therapy with metformin may significantly reduce tumor burden and prolong remission.
Here, a syngeneic model of aggressive triple negative mammary cancer was utilized in immune-competent mice to evaluate the contribution of dietary energy to primary tumor growth and progression to metastatic disease. Additionally, the effect of systemic metformin was analyzed to determine how metformin affects the primary tumor and/or metastatic events. Systemic mediators were evaluated to determine how these were affected by diet and metformin therapy and whether they correlated with tumor progression and/or metastatic events.
The 66cl4 cell line was obtained from Dr. Fred Miller (Karmanos Cancer Institute and Wayne State University, Detroit, MI). The 66cl4 cell line was derived from a single spontaneously arising mammary tumor from a Balb/c mouse . 66cl4 cells were transfected to express fluorescence red protein, using pDsRed2-ER vector (Invitrogen, Carlsbad, CA). The cell line was maintained using Dulbecco’s Modified Eagle Medium (DMEM) (Invitrogen, Carlsbad, CA) supplemented with 10% FBS, penicillin (100 units/ml)-streptomycin (100 µg/ml), and Geneticin sulfate (G418) (1.2 mg/ml). Cells were cultured at 37°C in 5% CO2. Antibodies used for immunoblots for phospho-AMPKα (Thr172), AMPKα, phospho-p44/42 MAPK (Thr202, Tyr204), p44/42 MAPK, phospho-p70 6S kinase (Thr389) and p70 S6 kinase were from Cell Signaling Technology (Beverly, MA), cyclin D1 and cyclin E antibodies were from Santa Cruz Biotechnology (Santa Cruz, CA) and the antibody for β-actin was from Abcam (Cambridge, MA). Metformin was purchased from Sigma–Aldrich (St. Louis, MO).
Cell growth was determined as previously described using methylthiazolyldiphenyl-tetrazolium bromide (MTT) (Sigma) following the manufacturer’s protocol .
Whole cell extracts were obtained using RIPA lysis buffer (1× PBS, 1% NP40, 0.1% SDS, 1.0% Deoxycholate) as described previously . Blots were developed using ECL reagents (Millipore) on a Kodak Multimodel Imager (2000MM).
Mice were kept in standardized housing and fed either the standard chow diet (Teklad 2018) or a high energy diet (Teklad Custom Diet, TD08445). The breakdown of each diet is listed in Table 1. Dietary energy restriction (DER)—Daily food intake was determined for each individually housed animal for one week on the standard diet. At the initiation of DER, 70% of each animal’s ad libitum (ad lib) food intake was measured and given to the mouse daily. Metformin treatments—During the metformin studies, mice were fed the standard or high energy diet as stated in each experiment and half of the mice were given oral metformin (3.3 mg/ml) in drinking water. Water and metformin-water were kept in light-protected, amber glass bottles and changed every three days. It was estimated that mice drank approximately 5 ml of water each day, yielding a dose of 825 mg/kg/day for a 20 g mouse, the maximum tolerated dose as determined in pilot studies (data not shown).
66cl4-DsRed cells were injected subcutaneously (1 × 106 cells/mouse) in the mammary fat pad of 6- to 8- week-old female Balb/c mice and monitored as previously described . Tumor and lung tissue were harvested and processed for frozen and/or paraffin histology using standard methods.
Paraffin sections of tumors were used for immunohistochemical analyses/quantification with H&E staining and with antibodies to platelet endothelial cell adhesion molecule (PECAM-1/CD31) (Santa Cruz, Inc. Santa Cruz, CA) and apoptosis with ApopTag TUNEL kit (Intergen, Purchase, NY) as described previously . To analyze lung metastases, three lung sections from each animal, 100 µm apart, were examined for lung lesions. A lung lesion was counted as >10 fluorescent cells clustered together when examined at 40× magnification.
Frozen sections of lung tissue were stained with CD11b (BD Pharmigen), F4/80, phospho-AMPKα, phospho-ACC (Upstate) or LC3B (Abcam) antibodies as described previously . Images were analyzed in areas free of any metastatic tumor foci or within tumor foci and quantified as previously described .
Serum insulin-like growth factor (IGF-1) levels were determined with the Quantikine Mouse IGF-1 Immunoassay from R&D Systems, Inc. (Minneapolis, MN).
Serum adipokine levels (insulin, leptin, resistin, TNF-α, MCP-1, and IL-6) were determined by the Milliplex Mouse Cytokine/Chemokine Panel (Millipore) using Luminex MAP technology (Luminex Corp, Austin, TX) according to the manufacturer’s protocol.
Results from individual experiments are represented as mean ± standard error unless otherwise stated. Statistical comparison of groups was performed using two-tailed Student t-tests or ANOVA. Statistical significance was defined as *P ≤ 0.05 or **P ≤ 0.01.
Many studies on caloric and dietary energy restriction have demonstrated significant reductions in mammary tumorigenesis [24–26]; however, few studies have examined the effect of these dietary interventions on established primary breast tumor growth and progression to metastatic disease. To directly examine the effect of DER on tumor growth and progression to metastatic lesions, the 66cl4 tumor model was employed in Balb/c mice. Tumor growth was monitored and the average tumor volume calculated twice weekly (Fig. 1a). DER mice displayed significantly reduced tumor growth rates compared to ad lib fed mice. At harvest, the DER treatment tumors were 35% smaller than the ad lib tumors. Since 66cl4 tumor cells are known to preferentially metastasize to the lung , lungs were collected and evaluated for metastatic disease (Fig. 1b). All ad lib mice displayed multiple metastatic lung lesions, while only 38% of the DER mice had lung lesions. DER resulted in an 87% reduction in the average number of metastatic nodules per lung compared to ab lib controls. DER efficacy was evaluated throughout the experiment by monitoring blood glucose levels. Prior to harvest, the ad lib group displayed an average blood glucose of 95 mg/dl, where the DER group averaged only 75 mg/dl which were statistically different (Fig. 1c).
As a reported CRM, the ability of metformin to affect primary tumor growth in this model was evaluated in mice maintained on a standard rodent diet (Table 1). Prior to use in this tumor model, cultured 66cl4 cells were evaluated for their response to metformin treatment in vitro (Supplemental Fig. 1). Metformin effectively activated AMP-dependent kinase and repressed cell proliferation and protein synthesis. Along with reduced proliferation, MAPK activation and cyclin E and D1 expression were significantly reduced by metformin treatments, consistent with observations in other cell lines [15, 17, 28–31]. Metformin showed no significant effect on the growth of the 66cl4 tumors in animals maintained on standard diet when compared to untreated controls as determined by external measurements, Fig. 2a or by weight of harvested tumors (Fig. 2b). Despite similar tumor size at harvest, the amount of viable tumor area in the metformin-treated group was significantly decreased when compared to controls (Fig. 2c). Additionally, intra-tumoral apoptotic cells were increased and the number of mitotic figures was suppressed by metformin treatment (Fig. 2d and e, respectively). The microvascular density (MVD) was evaluated with quantitative PECAM-1/CD31 staining; however, there were no differences observed between the two groups (data not shown). Finally, the effect of metformin on metastatic dissemination to distal lung tissue was evaluated. No difference was found in the number of metastatic lesions in control versus metformin-treated mice, 0.714 ± 0.184 vs. 0.750 ± 0.412, respectively, P = 0.941.
It has been reported that a high energy diet can promote the growth of syngeneic Lewis Lung Carcinoma in C57Bl/6 mice and that metformin treatment can attenuate this increased growth . Given these observations, the effect of a high energy diet on the growth of the 66cl4 mammary tumors and the effect of metformin on these tumors were evaluated. A diet with increased free sugar content (HED) (Table 1) was provided with and without metformin. Mice on HED established larger tumors more quickly, and the final tumor volumes were larger than those on the standard diet (Fig. 3a compared to Fig. 2a). In agreement with the Lewis Lung report , mice receiving HED plus metformin displayed significantly smaller tumors than mice on the same diet without metformin (Fig. 3a). The average tumor weight at harvest was significantly reduced in the metformin-treated group (Fig. 3b). No significant difference was observed in tumor viability between the HED alone and HED with metformin (Fig. 3c). Apoptosis was increased and the number of mitotic figures was decreased in the viable tumor regions of the metformin-treated group, similar to observations of tumors in the standard diet experiments (Fig. 3c and d, respectively).
It has been demonstrated that AMPK activation with metformin can promote the process of autophagy in vitro and in human colon tumor xenografts . Although viable tumor area was unchanged, it appeared that tumor cell density may be higher in the metformin-treated animals as indicated by the nuclear hematoxylin staining in Fig. 3e, HED compared to HED + Met. Therefore, tumors from both groups were analyzed for cell density by quantitative nuclear staining. The HED plus metformin group displayed a significant increase (23%) in tumor cell density compared to the HED group alone (2533 ± 104.3 cells/hpf vs. 1959 ± 124.3 cells/hpf, respectively, P = 0.0028).
To assess possible implications for metastatic potential, lungs were analyzed for metastatic lesions. There was no significant difference in the number of lung lesions between HED and HED plus metformin groups, 2.20 ± 0.611 vs. 4.00 ± 1.130, respectively, P = 0.168. However, mice treated with HED displayed a significant increase in both size and number of lung lesions compared to mice on the control diet (Fig. 4a–c).
To evaluate the differences in the systemic responses to dietary modifications and metformin treatment, several systemic physiological mediators, blood glucose, IGF-1, and insulin levels, were evaluated from blood samples obtained at the end of each experiment (Table 2). Blood glucose and IGF-1 levels were significantly reduced in animals on the DER diet to 79% and 76% of controls. In addition, insulin levels were reduced by 53% in DER animals when compared to the ad lib group.
The use of metformin with the standard diet did not significantly affect blood glucose, IGF-1 or insulin levels when compared to the controls. However, when metformin was used in animals provided the HED, metformin effectively suppressed blood glucose and IGF-1 levels by 17% and 27% of the HED group, respectively. The metformin suppression of glucose and IGF-1 levels in HED was brought down to equivalent levels as observed in the control diet animals. Although not statistically significant (P = 0.1366), insulin levels were decreased in HED with metformin compared to HED alone by 24%.
HED significantly increased the levels of blood glucose by 22% compared to control diet. The effect of HED was also reflected in increased weight gain over the course of the tumor model where final body weights were significantly greater for the HED at 21.81 ± 0.383 g vs. 18.26 ± 0.391 g for animals on standard diet (P ≤ 0.001). Metformin treatment also significantly suppressed the increase in body weight by 10.5% to 19.52 ± 1.77 g vs. HED alone, 21.81 ± 0.383 (P = 0.0041).
While metformin showed significant effects on metabolic growth factors in this model, the number of metastatic lung lesions was unchanged, suggesting that other systemic mediators may be promoting tumor cell inoculation and survival. To investigate systemic mediators that have been suggested to promote disease progression and facilitate lung metastasis , a panel of cytokines and adipokines were assessed in serum samples obtained at harvest. Serum samples from HED animals were compared to those on standard control diet, and two adipokines, leptin and resistin, were found to be significantly elevated (Table 3). Leptin was increased 2.9-fold and resistin 4.1-fold compared to those on standard diet. Resistin was significantly reduced (30%) by metformin treatment in animals on the standard diet while leptin was unaffected. Interestingly, metformin was ineffective at regulating either leptin or resistin in animals on HED, consistent with their increased adiposity. Other proinflammatory mediators such as TNF-alpha, IL-6, and MCP-1 were evaluated; however, most samples were at, or below, the minimal level of detection (data not shown).
While metformin was able to suppress systemic levels of glucose and growth factors that reflected in reduction of primary tumor growth in the high energy diet group, there was no effect on the number or size of metastatic lung lesions by metformin treatment. To evaluate whether metformin treatment was exerting any effect on the metastatic lesions, AMPK signaling events were evaluated in lung tissue and metastatic lung lesions. Immunofluorescence was performed on sections of lung from HED and HED plus metformin-treated animals to evaluate AMPK activation and the AMPK substrate, acetyl-CoA carbosylase (ACC) in the metastatic lung lesions or unaffected lung tissue (Fig. 5). There were no differences observed in the amount of p-AMPK or phosphorylated ACC (p-ACC) in the normal lung tissue. However, metastatic lung lesions displayed increased signal for both p-AMPK and p-ACC, when compared to surrounding normal lung tissue. It was also noted that there was significant variation between individual lesions and subsequently no significant differences were observed between the HED and HED plus metformin groups. Interestingly, while evaluating the metastatic lung lesions, the tumor foci in the HED plus metformin group appeared have an increased cell density compared to the HED alone. Since the primary tumors displayed a significant increase in cellular density with metformin treatment, lung tumor lesions were evaluated for active autophagy (Fig. 6a). Immunofluorescence for the autophagy marker, LC3B, revealed distinct punctuate perinuclear staining consistent with positive staining for autophagosomes. Quantification of the positive cells indicated that the HED plus metformin lung lesions displayed a significant increase (50%) in the number of LC3B positive clusters (Fig. 6b).
An emerging molecular theme that seems to correlate proinflammatory events with the promotion of metastatic foci has indicated that tissues, especially lung parenchyma, can be “primed” to promote the homing and survival of tumor cells. This priming can be promoted in part, by the pre-localization of bone marrow-derived myeloid cells and macrophages in lung tissue . Since HED treatment resulted in significantly increased levels of the proinflammatory adipokines leptin and resistin, lungs from mice-treated standard control diet and HED with and without metformin were evaluated for myeloid infiltrating cells expressing the CD11b and/or F4/80 antigens. Very few CD11b positive cells were present in the lungs of mice on the control diet, while mice on HED had a dramatic 20-fold increase in the number of CD11b positive cells (Fig. 7a). The presence of these cells appeared to be predominantly in the lung parenchyma as evidenced by immunofluorescence images (Fig. 7b). Metformin treatment appeared to increase the number of CD11b positive cells in conjunction with the control diet and decreased the number of CD11b positive cells in conjunction with the HED, however neither of these reached statistical significance. No differences were observed in the F4/80 staining that predominantly recognized resident alveolar macrophages in all of the treatment groups (data not shown).
This study used a syngeneic model of murine primary and metastatic mammary cancer to evaluate and compare the effects of nutrient availability, dietary energy restriction, and metabolic therapeutic intervention with metformin. The 66cl4 cells are a subpopulation of a spontaneously occurring mammary tumor from a Balb/c mouse and are known to metastasize to the lung . The aggressive nature of this tumor model suggests highly activated growth factor signaling as evidence by high proliferation rates in vitro, MAPK activation signals and lack of estrogen dependence. These cells do not express the hormone receptors for estrogen or progesterone or the Her-2 receptor and appear to represent a murine model of triple negative breast cancer. Triple negative breast cancer is more frequent in younger women as well as women who are obese. In addition, women diagnosed with triple negative breast cancer have especially poor prognosis and rapid occurrence of metastases [35, 36]. It is especially important to develop adjuvant interventions for women with this aggressive subtype of breast cancer.
From the perspective of primary tumor expansion, DER was the most effective at suppressing tumor growth and metastatic disease in this model. The application of metformin resulted in moderate, yet significant, effects on the primary tumor phenotype. Most prominently, increased apoptosis and reduced proliferation was observed. This data are in agreement with the in vitro observations that metformin treatment reduced proliferation, protein synthesis, and MAPK signaling in the 66cl4 cells and is consistent with findings reported in the human triple negative breast cancer cells, MDA-MB-231, when treated with metformin . Primary tumor expansion was promoted by increased dietary energy availability and the effect attenuated by the addition of metformin, which is consistent with the report by Algire et al. that utilized a HED and metformin in the Lewis Lung Carcinoma tumor model [16, 20]. These two studies confirm the efficacy of metformin to repress the growth of aggressive primary tumors in the presence of excess dietary energy in two independent tumor models. When metastatic disease was evaluated, only the application of DER resulted in a significant repression of lung lesions. Even though metformin effectively repressed primary tumor growth when used with HED, it was unable to repress the number of metastatic lung lesions in conjunction with either diet. Importantly, the HED group developed an increased number of metastatic lesions as well as larger lesions when compared to the standard diet group. Even though metformin was able to induce autophagy as demonstrated by increased cell density and LC3b staining of the HED group, metformin therapy was not sufficient to reduce the disease burden in the lungs of these animals.
It has been proposed that growth factor pathways can be critical in supporting breast cancer growth, especially in relation to obesity and diabetes. In addition, the efficacy of DER against tumor progression has been attributed to reduced IGF-1 levels and downstream signaling . The use of DER most efficiently repressed IGF-1 levels as well as blood glucose and insulin levels, while the use of metformin repressed blood glucose and IGF-1 levels only when given with the HED, as has been reported in other studies [19, 20, 39]. These data demonstrate that, while metformin can repress glucose and IGF-1 levels in the presence of excess dietary energy, metformin is not as effective under a balanced energy status and falls short of DER in this regard.
The effect of increased dietary energy in the form of free sugar and therapeutic modifiers such as metformin on the metastatic process is a more complex process to model and interpret. A suppression of distal metastasis was observed with DER compared to control diet, but surprisingly, metformin did not suppress metastasis in conjunction with the HED diet despite the fact that metformin was effective at controlling both IGF-1 and insulin levels. The implications of these results may be significant and imply that the contribution of glucose and IGF-1 in secondary metastasis is offset by other pro-metastatic events as suggested by other recent reports [40, 41]. Not surprisingly, the HED regimen dramatically increases both leptin and resistin levels well above the standard diet. The increased levels of both adipokines were not suppressed by metformin treatment. These data suggest that clinical measures other than glucose, IGF-1, and insulin may be useful when considering the effectiveness of metformin or other metabolic therapeutics when used as adjuvant breast cancer therapies.
As recent reports indicate, there is a growing interest in circulating adipokines as proinflammatory and tumorigenic factors for primary and metastatic cancer [42–44]. These data suggest that the effects of metformin on primary tumors are distinct from the events required for secondary metastasis to the lung, especially in this aggressive breast cancer model. It would be expected that HED may be promoting distal tissue priming to promote tumor cell homing, which is strongly supported by the evidence of significant infiltration of myeloid cells within the lung parenchyma with the HED regimen. It should be emphasized that leptin and resistin levels are only two adipokines that may be involved in this process, but any number of proinflammatory agents and growth factors may be involved, including cytokines such as TNF-α, IL-6, IL-8, or SDF-1α. Additional work is needed to investigate the role for each of these adipokines and local mediators to determine their contribution to the metastatic process.
Since epidemiological studies suggest that up to 50% of our population over the age of 50 exhibit metabolic syndrome [45, 46], representing a pre-diabetic condition related to high caloric intake and a sedentary lifestyle, the relationship of metabolic imbalance and the initiation and/or progression of breast cancer has become critically relevant. The data presented here imply that primary tumor growth and secondary tumor cell metastasis is supported by high energy consumption. Second, where potential novel therapies might be employed to suppress metastatic tumor growth, the identification of adipokines and local mediators that promote tissue priming may be critical in targeting tumor cell inoculation and/or survival in metastatic sites.
The authors would like to thank Nancy Ryan and Xiaoxiao Hong for their technical assistance. This work was supported by NIH:NCI CA064436 and the Connecticut Breast Health Initiative, Inc.
Electronic supplementary material The online version of this article (doi:10.1007/s10549-009-0647-z) contains supplementary material, which is available to authorized users.
Kathryn N. Phoenix, Center for Vascular Biology, Department of Cell Biology, University of Connecticut Health Center, 263 Farmington Ave., Farmington, CT 06030-3501, USA.
Frank Vumbaca, Center for Vascular Biology, Department of Cell Biology, University of Connecticut Health Center, 263 Farmington Ave., Farmington, CT 06030-3501, USA.
Melissa M. Fox, Center for Vascular Biology, Department of Cell Biology, University of Connecticut Health Center, 263 Farmington Ave., Farmington, CT 06030-3501, USA.
Rebecca Evans, Center for Vascular Biology, Department of Cell Biology, University of Connecticut Health Center, 263 Farmington Ave., Farmington, CT 06030-3501, USA.
Kevin P. Claffey, Center for Vascular Biology, Department of Cell Biology, University of Connecticut Health Center, 263 Farmington Ave., Farmington, CT 06030-3501, USA.