Metformin and soybean-derived bioactive molecules attenuate the expansion of stem cell-like epithelial subpopulation and confer apoptotic sensitivity in human colon cancer cells
- 764 Downloads
Colorectal cancer (CRC) is a disease whose genesis may include metabolic dysregulation. Cancer stem cells are attractive targets for therapeutic interventions since their aberrant expansion may underlie tumor initiation, progression, and recurrence. To investigate the actions of metabolic regulators on cancer stem cell-like cells (CSC) in CRC, we determined the effects of soybean-derived bioactive molecules and the anti-diabetes drug metformin (MET), alone and together, on the growth, survival, and frequency of CSC in human HCT116 cells. Effects of MET (60 μM) and soybean components genistein (Gen, 2 μM), lunasin (Lun, 2 μM), β-conglycinin (β-con, 3 μM), and glycinin (Gly, 3 μM) on HCT116 cell proliferation, apoptosis, and mRNA/protein expression and on the frequency of the CSC CD133+CD44+ subpopulation by colonosphere assay and fluorescence-activated cell sorting/flow cytometry were evaluated. MET, Gen, and Lun, individually and together, inhibited HCT116 viability and colonosphere formation and, conversely, enhanced HCT116 apoptosis. Reductions in frequency of the CSC CD133+CD44+ subpopulation with MET, Gen, and Lun were found to be associated with increased PTEN and reduced FASN expression. In cells under a hyperinsulinemic state mimicking metabolic dysregulation and without and with added PTEN-specific inhibitor SF1670, colonosphere formation and frequency of the CD133+CD44+ subpopulation were decreased by MET, Lun and Gen, alone and when combined. Moreover, MET + Lun + Gen co-treatment increased the pro-apoptotic and CD133+CD44+-inhibitory efficacy of 5-fluorouracil under hyperinsulinemic conditions. Results identify molecular networks shared by MET and bioavailable soy food components, which potentially may be harnessed to increase drug efficacy in diabetic and non-diabetic patients with CRC.
KeywordsMetformin Soy Lunasin Genistein Colon cancer Stem-like cell
Cancer stem cell
Fatty acid synthase
Phosphatase and tensin homologue deleted on chromosome ten
Real-time quantitative polymerase chain reaction
Colorectal cancer (CRC) remains a major contributor to cancer morbidity and mortality worldwide and is the third leading cause of cancer deaths in the USA (Siegel et al. 2015). While the molecular mechanisms by which CRC begins and then progresses to metastatic disease are better understood relative to some other cancer types (Tomasetti et al. 2014; Linnekamp et al. 2015), current treatments such as with the first-line chemotherapeutic drug 5-fluorouracil (5-FU) are largely ineffective once the tumor becomes invasive, metastatic and/or chemo-resistant (Paldino et al. 2014). Hence, preventing CRC initiation, progression and recurrence and defining bona fide targets for effective therapies are paramount for reducing CRC-related mortality.
Cancer stem cells (CSC) have emerged as attractive therapeutic targets to inhibit tumor initiation, chemo-resistance, and relapse in many cancer types (Paldino et al. 2014; Zeuner et al. 2014; White and Lowry 2015). According to the CSC hypothesis, a small population of normal stem cells requisite for maintaining tissue homeostasis/renewal can initiate a tumor upon acquiring one or more driver mutations (Chandler 2010). Moreover, a recent study (Tomasetti and Vogelstein 2015) has suggested that the lifetime risk of many different types of cancer is strongly correlated with the total number of divisions of these normal self-renewing cells, and that random mutations arising during DNA replication are likely causative for the emergence of CSC. Interestingly, among the different epithelial stem cells analyzed in that study, those derived from the colon were shown to undergo one of the highest numbers of cell divisions. Inhibiting the expansion of normal SC as well as CSC may thus constitute a useful approach for CRC prevention and treatment.
Diet is a well-accepted risk factor for CRC, the latter potentially mediated by bioactive and bioavailable components in foods that can influence colon epithelial cell growth and survival (Yan et al. 2010; Spector et al. 2013). Previous work in our laboratory demonstrated that dietary intake of soy (Xiao et al. 2008) and whey (Xiao et al. 2006) protein-based diets favorably influenced insulin-signaling components and reduced intestinal tumor incidence in rats exposed to the intestinal carcinogen azoxymethane. Further, we showed in the context of a high-fat diet that a soy protein-based dietary regimen suppressed intestinal epithelial cell proliferation and systemic insulin levels in mice (Al-Dwairi et al. 2012, 2014). The association between colonic intestinal proliferation and insulin is noteworthy, given that the anti-diabetic drug metformin (1,1-dimethylbiguanide hydrochloride; MET) has recently attracted much interest as a potential anticancer medication. Several epidemiological studies reported an inverse correlation between MET intake (at standard doses of 1500–2250 mg/day) and CRC incidence and related mortality in diabetic patients (Zhang et al. 2013; Sehdev et al. 2015). Further, MET at a low dose (250 mg/day) was found to reduce colon epithelial cell proliferation and numbers of aberrant crypt foci in non-diabetic patients (Dowling et al. 2011). Moreover, combinations of MET with conventional chemotherapeutic agents 5-fluorouracil and oxaliplatin were reported to decrease recurrence of colon tumors in a mouse xenograft model (Nangia-Makker et al. 2014). Nevertheless, the potential interactions of MET and bioactive dietary components with known anti-tumor activities for CRC prevention and treatment, and specifically in the context of targeting CSC, have not been delineated. The importance of the latter is underscored by recent reports that dietary factors including the soy isoflavone genistein (Gen) (Montales et al. 2012; Rahal et al. 2013), polyphenolic compounds present in blueberries (Montales et al. 2012), curcumin (Kakarala et al. 2010), and resveratrol (Kao et al. 2009), all of which have been reported to possess anti-metabolic/anti-diabetic properties (Gilbert and Liu 2013; Valsecchi 2013; Yeh et al. 2014; Rouse et al. 2014), can inhibit expansion of basal stem-like mammary cells with tumorigenic potential.
Here, we evaluated the effects of MET and several bioavailable and bioactive molecules present in soy protein isolate, namely the isoflavone Gen, peptide lunasin (Lun), and proteins β-conglycinin (β-con) and glycinin (Gly), alone and in combination, on the survival of CSCs present in the human colon cancer cell line HCT116 (Yeung et al. 2010). We examined potential chemo-preventive mechanisms of these components in CSCs and determined whether the combination of MET, Gen, and Lun with 5-FU could increase CRC cell death relative to 5-FU alone. Our results show that MET, Gen, and Lun together confer enhanced inhibitory effects on CSC expansion and suggest that the targeting of CSC with these molecules should be further explored for the clinical management of CRC.
Materials and methods
Cell culture and treatments
The human colon cancer cell lines HT29 and HCT116 (American Type Culture Collection; ATCC, Manassas, VA) were propagated in McCoy’s medium (ATCC) supplemented with 10 % fetal bovine serum (FBS; GIBCO, Carlsbad, CA) and 5 % antibiotic–antimycotic solution (ABAM; GIBCO) in a humidified incubator (5 % CO2:95 % air) at 37 °C. Cells were seeded in six-well plates at an initial density of 2 × 105 per well, and treated (in culture medium) with metformin (MET 60 µM; Sigma-Aldrich, St. Louis, MO), lunasin (Lun 2 µM; American Peptide Co., Sunnyside, CA), β-conglycinin (β-con 3 µM; provided by Dr. Ferreira, Federal University of Bahia, Brazil), glycinin (Gly 3 µM; provided by Dr. Ferreira), and genistein (Gen 2 µM; Sigma-Aldrich), alone or in combination. β-con and Gly were isolated and purified as described below. Metformin, Lun, β-con, and Gly were dissolved in phosphate-buffered saline (PBS; GIBCO), whereas Gen was dissolved in dimethyl sulfoxide (DMSO; Sigma-Aldrich). In other experiments, cells were treated with insulin (2 µM; Sigma), PTEN inhibitor SF1670 (2 µM; Echelon Biosciences, Salt Lake, UT), and 5-fluorouracil (5-FU 50 µM; Teva Parenteral Medicines, Irvine, CA). Treated cells were collected at select time points for subsequent analyses.
Extraction and isolation of β-conglycinin and glycinin
Soybean flour (60 mesh) was defatted with hexane [1:8 (wt/vol) flour-to-solvent ratio] by rocking at room temperature for 12 h and then re-extracted in the same solvent [1:6 (wt/vol)]. The extract was evaporated to dryness at room temperature and used for isolation of β-conglycinin and glycinin as previously described (Ferreira et al. 2010). The β-conglycinin and glycinin proteins were further purified by size exclusion chromatography on a Sepharose CL-6B column (1.0 × 100 cm) equilibrated with 0.05 M potassium phosphate, pH 7.5 containing 0.5 M NaCl and 0.01 % NaN3 at a column flow rate of 5.8 ml/tube. Column eluates were monitored at 280 nm, and protein-containing fractions were dialyzed and lyophilized. Column fractionation produced one major peak for each of the isolated proteins. SDS-PAGE confirmed the specific molecular weight and indicated over 95 % purity for β-con and Gly proteins (data not shown).
Cell viability, cell cycle distribution, and apoptosis
Cells were seeded in six-well plates at an initial density of 2 × 105 cells per well and incubated overnight prior to treatment. Cells were collected 48 h post-treatment, and the quantity of viable cells was determined by the trypan blue exclusion method using the Vi-Cell cell viability analyzer (Beckman Coulter Inc., Atlanta, GA). For cell cycle analysis, cells were serum-starved for 24 h after plating in 0.5 % charcoal-stripped FBS (CS-FBS; GIBCO), and 24 h later were treated with medium supplemented with 2.5 % CS-FBS and the described treatments. The analysis of cell cycle parameters followed previously described protocols (Rahal and Simmen 2010). At least 1 × 105 cells were stained with propidium iodide (PI; Sigma-Aldrich), and the proportion of cells in Sub G1, G1, S, and G2 phases was analyzed with the Becton–Dickinson LSRFortessa Flow Cytometer and associated software (BD Biosciences, San Jose, CA). The percent of apoptotic cells 48 h post-treatment was evaluated by Annexin V staining (Trevigen, Gaithersburg, MD), followed by analysis in the Becton–Dickinson LSRFortessa Flow Cytometer.
Single-cell suspensions were evaluated for relative frequency of CD133+CD44+-stained cells by flow cytometry and the use of PE-conjugated anti-CD133 (Miltenyi Biotec Incorporated, San Diego, CA) and FITC-conjugated anti-CD44 (BD Pharmingen, San Jose, CA), as previously described (Chen et al. 2011). Cells were sorted on a FACS Aria Cell Sorting Flow Cytometer (BD Biosciences); Supplementary Figure S1 shows the gating strategy used for FACS analysis. Dead cells were excluded using 4′,6-diamidino-2-phenylindole (DAPI; 1 µg/ml; Sigma-Aldrich). CD133+CD44+-positive cells were collected for colonosphere formation assay and gene expression analysis (below).
To examine treatment effects on colonosphere formation, unsorted HCT116 cells and FACS-sorted CD133+CD44+ cell fractions were seeded at a density of 2 × 103 cells per well in ultra-low attachment plates (Corning Inc., Corning, NY) in plating medium [phenol red-free MEM supplemented with B27 media (Invitrogen); 20 ng/ml human basic fibroblast growth factor (Invitrogen, Carlsbad, CA), 20 ng/ml epidermal growth factor (Invitrogen), 10 μg/ml heparin (Sigma-Aldrich), 1 % ABAM, and 100 μg/ml gentamicin (Sigma-Aldrich)] (Montales et al. 2012). Treatments were MET (60 µM), soy bioactive factors (3 µM each for β-con and Gly; 2 µM each for Lun and Gen), and insulin (2 µM), singly and with the indicated combinations. To investigate whether MET limited colonosphere-forming ability, HCT116 cells were treated after initial plating with MET or vehicle, collected 24 h later and then grown as single-cell suspensions under low attachment-plating conditions. For PTEN inhibitor studies, unsorted cells (2 × 105 per well) were plated in six-well plates, allowed to attach overnight, and then treated with the PTEN inhibitor SF1670 (2 µM) for 24 h, prior to collection and re-plating in ultra-low attachment plates. Colonosphere number was counted 5 days after the initial (one-time) treatment using an inverted microscope. To evaluate self-renewal, primary colonospheres (P1) were collected after counting, dissociated into single cells, filtered using a 40 μM sieve, and re-plated in ultra-low attachment plates. Colonospheres formed at secondary passage (P2) were counted at 7 days post-plating and expressed as a percentage of the total number of initially plated cells. Treatment effects were determined from at least three independent experiments, with 3–4 replicates per experiment.
RNA isolation and quantification
Total RNA was extracted from unsorted and FACS-sorted cells and from colonospheres using TRIzol reagent (Invitrogen). cDNA synthesis and real-time quantitative PCR (QPCR) were performed following published methods (Montales et al. 2012). Primers (Supplemental Table 1) were designed using Primer Express (Applied Biosystems, Foster City, CA) and synthesized by Integrated DNA Technologies, Inc. (Coralville, IA). Gene expression was determined by QPCR using Bio-Rad iTaq SYBR Green Supermix (Bio-Rad; Hercules, CA). Transcript levels for target genes were calibrated to a standard curve generated using pooled cDNAs and normalized to a factor that was derived from the geometric mean of expression for β-actin (ACTB) and cyclophilin A (CYPA) genes, using GeNorm excel file software as previously described (Vandesompele et al. 2002).
Western blot analysis
Whole extracts from treated cells (60 µM MET or vehicle; 24 h) were harvested in RIPA buffer containing protease inhibitor cocktail (Santa Cruz Biotechnology; Santa Cruz, CA). Protein concentrations were determined using the bicinchoninic acid protein assay kit (Pierce Biotechnology; Rockford, IL). Equal amounts of lysate protein (50 μg) were subjected to immunoblotting as described previously (Al-Dwairi et al. 2012). Primary antibodies used were anti-PTEN (1:1000; Cell Signaling, Danvers, MA), anti-FASN (1:1000; Abcam, Cambridge, MA), and anti-β-actin (1:10,000; Sigma-Aldrich). Blots were stripped with Restore Western blot stripping buffer (Pierce Biotechnology) prior to incubation with another primary antibody. Immunoreactive proteins were visualized with Amersham ECL Plus kit (GE Health Care Life Sciences, Piscataway, NJ). Digital images were captured with a FlourChem HD2 Imager system (Alpha Innotech, San Leandro, CA), and densitometry was performed using Alpha View software (Cell Biosciences; Santa Clara, CA). Levels of immunoreactive PTEN and FASN proteins were normalized to that for β-actin.
All experiments were performed at least twice, with each experiment conducted in triplicate or quadruplicate. Data are presented as the mean ± SEM; except for that in Fig. 5b which are LS mean ± SEM. Statistical significance between diet groups was evaluated by Student’s t test or one-way analysis of variance using Sigma Stat version 3.5 for Windows. Data in Fig. 5b were subjected to a two-way ANOVA with ‘experiment’ being considered as a ‘random’ factor, and pairwise multiple comparison procedures (Holm–Sidak method) were used to ascribe statistically significant differences between treatment groups. A P value <0.05 was considered to be statistically significant, with tendency for significance at 0.05 < P < 0.1.
HT29 and HCT116 colonospheres
MET attenuated cell viability and colonosphere formation, induced cell apoptosis and altered PTEN and FASN expression of HCT116 cells
We next quantified the temporal expression of the tumor suppressor PTEN and the lipogenic enzyme FASN in control and MET-treated cells. MET increased PTEN mRNA levels (Fig. 2e) and decreased FASN mRNA abundance (Fig. 2f) in a time-dependent manner. Interestingly, MET effects on PTEN and FASN gene expression were maximal at 6 h and persisted to 24 h (Fig. 2e, f) and 48 h (data not shown). The increase in PTEN and decrease in FASN mRNA abundance at 24 h with MET were confirmed at the level of their respective proteins (Fig. 2g, h), demonstrating concordance of mRNA and protein expression.
Soy bioactive components alter HCT116 cell viability, apoptosis, and colonosphere formation
We next examined whether the above soy factors affected colonosphere formation in HCT116 cells. Gen, Lun, and Gly each suppressed colonosphere formation by ~50 %, while β-con was ineffective at the concentration tested (Fig. 3e). Co-treatment with MET and Lun reduced colonosphere formation to a slightly lesser extent (by ~60 %) than did co-treatment with MET and Gen (by ~76 %) (Fig. 3f). Interestingly, co-treatment with all three components failed to reduce colonosphere formation further than that elicited with MET in the presence of either Lun or Gen (Fig. 3g).
MET, Lun, and Gen limit the cancer stem cell-like cell subpopulation, in part, through PTEN
Epidemiological studies suggest that hyperinsulinemia is associated with increased CRC risk (Giovannucci 2007). Moreover, insulin signaling was found to inhibit PTEN expression (Liu et al. 2014), and insulin resistance has been linked to increased FASN expression (Menendez et al. 2009), indicating that aberrant insulin signaling may underlie altered PTEN and FASN expression levels and activity in CRC. To evaluate whether Lun, Gen, and/or MET attenuate CSC frequency by opposing insulin action, we exposed HCT116 cells to an elevated level of insulin (2 µM, modeling hyperinsulinemia) and determined effects of added factors on colonosphere formation and on the CD133+CD44+ subpopulation. Insulin increased colonosphere formation (Fig. 4c) and the percentage of CD133+CD44+ cells (Fig. 4d). Lun, Gen, and MET reduced CSC formation under in vitro hyperinsulinemic condition, with Gen demonstrating greater inhibitory activity than either MET and Lun (MET = Lun) (Fig. 4c, d).
HCT116 cell killing by 5-FU plus MET, Gen, and Lun
CRC is increasingly considered to have metabolic underpinnings; hence, dietary factors that maintain metabolic homeostasis or normalize metabolic dysfunction may be beneficial to inhibiting CRC progression, metastasis, and relapse. Clinical and epidemiological studies have implicated the anti-diabetic drug metformin (MET) in reducing the risk of CRC among diabetic patients as well as improving CRC outcome in patients with diabetes (Dowling et al. 2011; Zhang et al. 2013; Nangia-Makker et al. 2014; Sehdev et al. 2015). Additionally, in diabetic patients with CRC, the decrease in mortality with MET was inversely associated with the frequency of CD133+ subpopulation with CSC properties (Zhang et al. 2013). Nevertheless, MET effects on colon cancer remain controversial (Sui et al. 2014). Moreover, MET elicits unfavorable side effects in some individuals, precluding its use at typically prescribed doses for diabetic patients. Thus, factors that could increase the efficacy of MET may be advantageous for its inclusion as a therapy to address CRC prevention and/or treatment.
Previous studies in several laboratories including our own (Xiao et al. 2008; Al-Dwairi et al. 2012; Yeh et al. 2014; Isanga and Zhang 2011) have shown that soybean components when consumed can favorably affect metabolic pathways and associated genes in animal models and in humans. Moreover, in a mouse model of breast cancer and in breast cancer cells in vitro, dietary intake of soy protein isolate and treatment with Gen, respectively, decreased the frequency of the mammary basal stem cell-like subpopulation, which was associated with decreased mammary tumor incidence and oncogenic molecular profile (Montales et al. 2012; Rahal et al. 2013). Based on this previous body of work, we hypothesized that combination(s) of MET and bioactive soy factors would increase the therapeutic efficacy of the CRC drug 5-FU by targeting cancer stem cells. Our findings present a novel and potentially important role for combination MET, Lun, and Gen in not only reducing cancer cell viability and survival, but in limiting the frequency of the cancer cell subpopulation (CD133+CD44+) with self-renewal properties. We also provide evidence, using the PTEN-specific inhibitor SF1760 and insulin, to indicate that MET, Gen, and Lun actions are comparably mediated by their concurrent induction of tumor suppressor PTEN and inhibition of pro-oncogenic FASN expression. The functional significance of these results is supported by the increased efficacy of 5-FU in eliminating CSCs under high-insulin conditions, the latter mimicking aspects of type 2 diabetes and/or obesity, when co-exposed with MET, Gen, and Lun.
In these studies, we primarily utilized the human colon cancer cell line HCT116 since its highly aggressive nature and increased proportion of non-differentiating CSCs are clinically correlated with lymph node metastases and reduced survival (Yeung et al. 2010). We show that relative to the more differentiated, less aggressive colon cancer cell line HT29, HCT116 cells exhibit a greater proportion of the CSC subpopulation (designated CD133+CD44+), increased ability to form colonospheres (a measure of tumor-forming ability and self-renewal in vitro), and enhanced expression of CSC genes. Using these cells to model aggressive CRC in vitro, we observed that MET, Gen, and Lun, individually and together, limited the expansion of the CSC subpopulation, suggesting their potential value in inhibiting disease progression and conceivably, metastasis and recurrence. The higher efficacy of Gen relative to MET, at the doses used here, in reducing colonosphere formation under high-insulin conditions is a novel and an unexpected finding. Previous studies have shown a dose-related inverse association between soy consumption and CRC risk (Yan et al. 2010). Nevertheless, molecular mechanisms to explain the preventative effects of soy on colon cancer remain relatively unknown. The demonstration of the efficacy of Gen, Lun, and Gly in targeting CSCs at relatively low doses (2–3 uM) suggests their potential to reduce currently prescribed doses of CRC conventional therapies with adverse side effects. While our data point to potential combination therapies of soybean components and 5FU, the current study has not attempted the complex task of defining optimal doses for these single agents and combinations (future studies). Our findings also provide an impetus to pursue preclinical studies of these bioactive molecules in animal models of colon cancer metastasis and under conditions of hyperinsulinemia.
The present study utilizing an in vitro model of aggressive colon cancer in a hyperinsulinemic environment, provides insight into molecular networks shared by the well-prescribed, anti-diabetic drug MET and bioactive/bioavailable soy food components, which potentially may be harnessed to increase drug efficacy in patients with CRC. Additionally, given that PTEN loss of function/expression is a hallmark of many primary and metastatic CRCs, our results suggest further exploration of the potential of combination therapy with MET and the soy factors to improve CRC outcome. Induction of tumor suppressor PTEN expression in cancer cells by dietary bioactive factors is an emerging theme (Liu et al. 2013; Pabona et al. 2013) that warrants further confirmation in vivo. In addition, our results for β-conglycinin and glycinin suggest further examination of these molecules for efficacy in CRC, in light of their high abundance and bioavailability either as intact or smaller proteins/peptides in consumed soy products. The mode(s) of action of these two latter proteins in cancer cell growth regulation remain virtually unexplored, with the exception of a previous report on β-conglycinin as being inhibitory to leukemia cell growth in vitro (Wang et al. 2008). Finally, the identification of food constituents that target CSCs underscores the potential importance of diet/nutrition in the genesis (and treatment) of primary and metastatic CRC and of other cancer types with metabolic underpinnings.
We thank other members of our laboratories for helpful discussions during the course of this work. We are also grateful to Lorenzo Fernandes and Eric Siegel for their assistance with statistical analysis and data presentation.
MTM carried out the experiments and performed statistical analysis. EF and VN performed the extraction and purification of β-conglycinin and glycinin. MTM, RCMS, and FAS interpreted the data and drafted the manuscript. RCMS and FAS supervised the entire project. All authors read and approved the final manuscript.
This work was supported by NIH R01CA136493, NIH UL1TR000039, and the Huie Family Trust.
- Al-Dwairi A, Brown AR, Pabona JM, Van TH, Hamdam H, Mercado CP, Quick CM, Wight PA, Simmen RC, Simmen FA (2014) Enhanced gastrointestinal expression of cytosolic malic enzyme (ME1) induces intestinal and liver lipogenic gene expression and intestinal cell proliferation in mice. PLoS ONE 9(11):e113058PubMedCentralCrossRefPubMedGoogle Scholar
- Iwasaki M, Inoue M, Otani T, Sasazuki S, Kurahashi N, Miura T, Yamamoto S et al (2008) Plasma isoflavone levels and subsequent risk of breast cancer among Japanese women: a nested case-control study from the Japan Public Health Center-based prospective study group. J Clin Oncol 26:1677–1683CrossRefPubMedGoogle Scholar
- Montales MT, Rahal O, Kang J, Rogers TJ, Prior R, Wu X, Simmen RC (2012) Repression of mammosphere formation of human breast cancer cells by soy isoflavone genistein and blueberry polyphenolic acids suggests diet-mediated targeting of cancer stem-like/progenitor cells. Carcinogenesis 33:652–660CrossRefPubMedGoogle Scholar
- Pabona JM, Dave B, Su Y, Montales MT, de Lumen BO et al (2013) The soybean peptide lunasin promotes apoptosis of mammary epithelial cells via induction of tumor suppressor PTEN: similarities and distinct actions from soy isoflavone genistein. Genes Nutr 8:79–90PubMedCentralCrossRefPubMedGoogle Scholar
- Rahal O, Machado H, Montales MT, Pabona JM, Heard M, Nagarajan S, Simmen RC (2013) Dietary suppression of the mammary CD29(hi)CD24(+) epithelial subpopulation and its cytokine/chemokine transcriptional signatures modifies mammary tumor risk in MMTV-Wnt1 transgenic mice. Stem Cell Res 11:1149–1162PubMedCentralCrossRefPubMedGoogle Scholar
- Sui X, Xu Y, Yang J, Fang Y, Lou H, Han W et al (2014) Use of metformin alone is not associated with survival outcomes of colorectal cancer cell but AMPK activator AICAR sensitizes anti-cancer effect of 5-fluorouracil through AMPK activation. PLoS ONE 9(5):e97781PubMedCentralCrossRefPubMedGoogle Scholar
- Xiao R, Carter JA, Linz AL, Ferguson M, Badger TM, Simmen FA (2006) Dietary whey protein lowers serum C-peptide concentration and duodenal SREBP-1c mRNA abundance, and reduces occurrence of duodenal tumors and colon aberrant crypt foci in azoxymethane-treated male rats. J Nutr Biochem 17:626–634CrossRefPubMedGoogle Scholar