Metformin Represses Bladder Cancer Progression By Inhibiting Stem Cell Repopulation Via COX2/PGE2/STAT3 Axis

ObjectiveCancer stem cells(CSCs),a sub-population of cancer cells,has been shown to be involved in the initiation,differentiation,recurrence,metastasis,and drug resistance of various cancers,including bladder cancer.Metformin,a widely used antidiabetic agent,has been demonstrated to possess anti-tumor characteristics towards various cancers in different pathways,such as inhibiting CSCs.However,the inhibition of metformin on bladder cancer CSCs remains unrevealed.We hope that through in vitro and in vivo experiments,explore the inhibitory effect of metformin on bladder cancer stem cells.Methods1.Rats were divided into three groups: control group(n=30);N-methyl-N-nitrosoureatreated group(MNU group);and MNU + metformin(MET group).For MNU treatment,2.5 mg of MNU dissolved in sodium citrate buffer(5g/L)was irrigated into the rat bladders via epidural catheter every other 3 weeks(i.e.week 0,3,6,9).The rats in MET group can freely access tap water with metformin(1g/L).Rats were sacrificed at week 3(n=30),week 6(n=26),and week 12(n=44),respectively,after the first MNU treatment.Blood was drawn from each rat and used for glucose measurement.The bladders were fixed in 10% formalin and paraffin embedded for hematoxylin and eosin(H&E)and Ki67,COX2,Bcl-2,CK14andCK20,OCT3/4,p-STAT3(Tyr705)immunostaining.2.MTT method is used to study 20 mM metformin effects on T24,RT4 bladder cancer cell proliferation;Using Western blot method to research 0 mM,1 mM,5 mM,10 m M,20 m M metformin effect on the Bcl-2,Cyclin D1,AMPK,p-AMPK proteins in T24 or RT4 bladder cancer cell;Flow cytometry study 0 m M,1 mM,5 mM,10 mM,20 mM metformin effect on cell cycle and apoptosis in T24 bladder cancer cell.3.Using ELISA method to detect 0 mm,1 mm,5 mm,10 mm,20 mm metformin effect on generation of PGE2 in T24 bladder cancer cell;Western blot method to detect 0 m M,1 mM,5 mM,10 mM,20 mM metformin,celecoxib,exogenous PGE2 and siCOX2 effect on the COX2,p-STAT3,STAT3,CK14,OCT3/4 protein expression in T24 bladder cancer cell.Results1.At week 3,the total number of mild dysplasia lesions between the MNU group and MET group were not significantly different(24 lesions/10 rats in the MNU group vs 28 lesions/10 rats in the MET group,P=0.466).However,the number of moderate/severe dysplasia lesions in the MNU group(36/10 rats)is significantly higher than that of the MET group(16/10 rats,P=0.007).At week 6,metformin treatment increased the number of the mild dysplasia lesion significantly with 29 and 15 mild dysplasia lesions/8 rats in the MET and MNU group,respectively(P=0.026);whilst decreased moderate/severe dysplasia lesion dramatically with 5 and 30 moderate/severe dysplasia lesions/8 rats in the MET and MNU group,respectively(P=0.02).In addition,3 CIS lesions were also seen in the MNU group with none of such lesion was seen in the MET group.At week 12,there were more papillary tumor lesions in MET group than that of the MNU group(53 lesions/20 rats in MET group vs 19 lesions/14 rats in MNU group,P=0.03).However,the invasive lesions in MET group(46 invasive lesions/20 rats)are significantly less than that of MNU group(49 lesions/14 rats)(P=0.034).2.We treated two bladder cancer cell lines T24 and RT4 with or without 20 mM of metformin for 5 days and the numbers of viable cells were estimated by MTT.It shows that metformin is able to inhibit both T24 and RT4 cell growth.AMPK(Adenosine Monophosphate Activated Protein Kinase)is a well-established molecular target of metformin.We estimated the levels of the phosphorylated AMPK in T24 cells treated with different concentrations of metformin.It shows that metformin is capable of upregulating the phosphorylation of AMPK in a dose-dependent manner without affecting the level of total AMPK.In addition,results from western blotting assays showed that metformin is capable of repressing the levels of cyclin D1 and Bcl-2 in both cell lines in a dose-dependent manner.Then,we estimated the effect of metformin on cell cycles distribution by treating T24 cells with different concentrations of metformin.It shows that metformin treatment increased the cell populations in G1/S phases and decrease the population in G2 phase.Furthermore,the apoptotic cell populations were upregulated by metformin in a dose-dependent manner.More importantly,immunohistochemical staining showed the expression of Ki67 and Bcl-2 in bladder specimens were significantly repressed by metformin treatment.Of note,not only the intensity of Ki67 but also the numbers of Ki67-positive cells were significantly reduced in metformin-treated group at all time points.These data suggest that metformin exerts its anticancer effects by slowing down cell growth and enhancing apoptosis.3.In order to determine if metformin exerts distinct effect on different precursor cells,we examined the expression of two stem cell markers,CK14 and OCT3/4 in tumors derived from animals treated with or without metformin.Metformin treatment reduced the number of CK14-and OCT3/4-positive cells.On the other hand,at all time points examined(3,6 and 12 weeks)more CK20-positive cells were seen in the tissues of the animals treated with metformin.It appears that metformin selectively inhibits CIS and invasive lesions without affecting papillary lesions.These data suggest that metformin may inhibit the repopulation of basal stem cell and subsequently repress early CIS and invasive cancer progression.4.To dissect the mechanism of metformin inhibition of CK14-and OCT3/4-positive cell repopulation,we examined the effect of metformin on COX2,a key enzyme catalyzing the biosynthesis of PGE2.It shows that metformin is capable of repressing the expression of COX2 at all time points in both papillary and invasive lesions.These observations were further substantiated by the dose-dependent inhibition of COX2 in vitro when T24 cells were treated with different concentrations of metformin.Consistent with COX2 inhibition,the levels of PGE2 in culture media were significantly lower when the cells were treated with metformin.Based on the facts that in colorectal cancer the JAK2/STAT3 signaling pathway is regulated by PGE2,and STAT3 is not only involved in the expansion of CK14-positive stem cells in bladder cancer,but STAT3 activation in urothelial stem cells also leads to invasive bladder cancer progression,we decided to examine if metformin-inhibited bladder cancer development is through the COX2/PGE2/STAT3 axis.Immunohistochemistry shows the levels of p-STAT3 in the bladder cancer animal model treated with metformin are consistently lower.These observations were further substantiated by the dose-dependent downregulation of phosphorylated STAT3 in cellular model.Of note,metformin treatment has no effect on the levels of total STAT3.Consistent with downregulation of the levels of COX2 and PGE2 as well as inhibition of STAT3 phosphorylation/activation,the levels of both CK14 and OCT3/4 were repressed by metformin in a dose-dependent manner;suggesting metformin may represses bladder cancer development through the COX2/STAT3 pathway.To further substantiate our hypothesis,we treated T24 cells with either a COX2 specific inhibitor celecoxib,or different concentrations of metformin.Metformin repressed all above-mentioned factors as celecoxib did.In addition,we knocked down COX2 in T24 cells by si RNA technique and monitored the changes of the molecules including COX2,STAT3,p-STAT3,CK-14,OCT3/4.Both siCOX2-1 and siCOX2-3 knocked down COX2 efficaciously but for some unknown reason that si COX2-2 failed to knockdown COX2.When COX2 is knocked down,all three markers p-STAT3,CK14 and OCT3/4 were downregulated significantly.However,the total STAT3 levels were not affected at all.To demonstrate that metformin affects CSC repopulation through PGE2,we treated the cells with metformin in the presence or absence of exogenously added PGE2.It shows that exogenously added PGE2 is able to counteract metformin and upregulate all CSC markers tested without affecting the levels of COX2;demonstrating the essentiality of PGE2 in bladder cancer stem cell repopulation.These data collectively suggest that metformin represses CK14+ stem cell repopulation by downregulating COX2 and subsequently inhibiting the COX2/PGE2/STAT3 axis.Conclusion1.Metformin may not be able to prevent bladder cancer initiation but it appeared to be capable of slowing down the progression from mild to moderate/severe dysplasia lesions,and from CIS to invasive lesions;2.Metformin could inhibit bladder cancer CSCs repopulation and promote apoptosis by downregulating cyclinD1 and Bcl-2;3.Metformin could downregulation of the levels of COX2 and PGE2 as well as inhibition of STAT3 phosphorylation/activation.In a word,metformin represses bladder cancer progression by inhibiting stem cell repopulation via COX2/PGE2/STAT3 axis...