Carcinogenesis vol.33no.3pp.661–669,2012doi:10.1093/carcin/bgr320Advance Access publication January4,2012Hypoxia-inducible factor-1a inhibition by a pyrrolopyrazine metabolite of oltipraz as a consequence of microRNAs199a-5p and20a inductionSeul Gi Kang1,2,y,Woo Hyung Lee1,2,y,Young Hun Lee3, Yong Sup Lee3and Sang Geon Kim1,2,Ã1Department of Pharmacy,College of Pharmacy and2Research Institute of Pharmaceutical Sciences,Seoul National University,Seoul151-742,South Korea and3Department of Pharmaceutical Science,College of Pharmacy, Kyung Hee University,Hoegi-Dong,Seoul130-701,South KoreaÃTo whom correspondence should be addressed.College of Pharmacy,Seoul National University,Sillim-dong,Kwanak-gu,Seoul151-742,South Korea. Tel:þ822-880-7840;Fax:þ822-872-1795;Email:sgk@snu.ac.krOltipraz,a cancer chemopreventive agent,has antitumor and anti-angiogenic effects.In animal models and clinical studies,a consider-able amount of oltipraz is metabolized to pyrrolopyrazines, including M2,7-methyl-6,8-bis(methylthio)pyrrolo[1,2-a]pyrazine; M3,7-methyl-8-(methylsulfinyl)-6-(methylthio)pyrrolo[1,2-a]pyra-zine and M4,7-methyl-6,8-bis(methylsulfinyl)pyrrolo[1,2-a]pyra-zine.In view of the role of hypoxia-inducible factor-1(HIF-1)a in tumor growth and angiogenesis,this study investigated whether pyrrolopyrazine metabolites of oltipraz inhibit HIF-1a induction, and if so,what the molecular basis is.M2treatment inhibited the induction of HIF-1a by a variety of stimuli including insulin, hypoxia,CoCl2and hydrogen peroxide in HCT116cells,whereas M3or M4failed to do so.Consistently,M2prevented HIF-1a target gene induction.Moreover,it inhibited cancer cell invasion and migration.M2caused no change in the expression of HIF-1a transcript but increased the levels of precursor forms of micro-RNAs(miRNAs)199a-5p and20a,but not those of primary forms, indicating facilitation of the maturation process of the miRNAs by M2.Increased levels of the miRNAs contributed to HIF-1a re-pression,as shown by the results of experiments using mimics. Consistently,M2treatment inhibited de novo synthesis of HIF-1a, as supported by decreased incorporation of[35S]-methionine into HIF-1a with no changes in its ubiquitination or degradation. 7-Ethyl-6,8-bis(methylthio)pyrrolo[1,2-a]pyrazine,a synthetic analog of M2,had a similar inhibitory effect.In conclusion,M2 with pyrrolopyrazine structure and its7-ethyl congenor have the ability to prevent the induction of HIF-1a,which may result from the inhibition of HIF-1a de novo synthesis,as mediated by the induction of miR-199a-5p and miR-20a.IntroductionAn activated complex of hypoxia-inducible factor-1(HIF-1)a facil-itates to induce the expression of the genes implicated in the adapta-tion of cancer cells to tumor microenvironments where the utilization of oxygen and nutrients are significantly restricted(1–3),which plays a crucial role in cancer cell proliferation,angiogenesis and invasion/ migration(4,5).Moreover,a considerable amount of HIF-1a is de-tected in aggressive and/or malignant cancers(6,7).A series of com-pounds(e.g.rapamycin,YC-1,resveratrol,radicicol and17-AAG) have been unveiled as the inhibitors of HIF-1a(4,5,8,9).Since they may possess toxic and/or adverse effects(5,9),recent pharmacolog-ical interventions in the activity of HIF-1a are limited.Thus,advances are required in the development of tailor-made inhibitors of HIF-1a.Oltipraz[4-methyl-5-(2-pyrazinyl)-1,2-dithiole-3-thione]is a can-cer chemopreventive agent(10–12)and exerts antitumor and anti-neoangiogenic effects in tumor xenograft animal models(13,14).It has an inhibitory effect on the growth of pulmonary adenomas and forestomach cancers(15,16).In mammals,oltipraz is metabolized into oxidized forms via main pathways:first,oxidative desulfuration of the thione to yield4-methyl-5-(pyrazin-2-yl)-3H-1,2-dithiol-3-one (M1)with no further metabolism and second,desulfuration,methyl-ation and molecular rearrangement to produce7-methyl-6,8-bis(me-thylthio)pyrrolo[1,2-a]pyrazine(M2,a pyrrolopyrazine structure), which may be further metabolized to other forms(M3and M4) (17,18).In our clinical trials,oltipraz underwent prompt and substan-tial molecular conversion into mostly M2after oral administration in human(19).At a high dose(i.e.90mg,q.d.),extended residence time of oltipraz in the body may have caused increased formation of M2(19).Ourfindings indicated that oltipraz or its synthetic derivatives have the ability to inhibit the induction of HIF-1a in human colon cancer cells by not only facilitating its degradation but also decreasing its de novo synthesis(14).It is noteworthy that oltipraz and M2share sim-ilar pharmacokinetic profiles,suggesting that the M2metabolite is highly likely to contribute to the beneficial effect of oltipraz and serve as the bioactive metabolite.In addition,both M1and M2induce the expression of GSTA2gene,prevent mitochondrial injury by activating adenosine monophosphate-activated protein kinase and protect hep-atocytes from reactive oxygen species(20,21).M2treatment effec-tively activated NF-E2-related factor2,suggesting that it may be pharmacologically active(20).Some NF-E2-related factor2activa-tors have the ability to inhibit the expression of HIF-1a.For example, sulforaphane and resveratrol inhibit HIF-1a activity by reducing its stability(22,23).Nonetheless,it remains to be established whether M2has an effect on HIF-1a.In view of the possibility that M2inhibits HIF-1a and has antican-cer potential,this study examined the effects of M2and its synthetic congenors on the expression and activity of HIF-1a.We also explored whether M2suppresses HIF-1a-dependent gene transactivation and,if so,the underlying basis.Here,we identified the induction of specific microRNAs(miRNAs)by M2treatment and their inhibitory effect on the translation of HIF-1a.Moreover,new derivatives of M2that have pyrrolopyrazine structure were synthesized with the aim of discover-ing other candidates that similarly inhibit HIF-1a.Now,we report the inhibitory efficacy of M2and its7-ethyl substitute on the induction of HIF-1a.Ourfindings indicate that these agents increase the levels of mature miR-199a-5p and miR-20a for HIF-1a repression,which is a unique and novel mechanism for the inhibition of HIF-1a activity by pharmacological means.Materials and methodsCells and cell culture conditionsHCT116and HT29cells,human colon cancer cell lines,were obtained from ATCC(Rockville,MD).The cells were maintained in growth medium con-taining Dulbecco’s modified Eagle’s medium,10%fetal bovine serum and5% penicillin–streptomycin at37°C in a humidified atmosphere containing5% CO2.For all experiments,cells were grown to80–90%confluency and were deprived of serum for16h before drug treatment.To create hypoxic conditions, they were transferred to a hypoxic chamber(Forma Scientific,Marietta,OH), where the cells were maintained at37°C in an atmosphere containing5%CO2, 1%O2and94%N2.MaterialsInsulin,H2O2,MG132and CoCl2were purchased from Sigma–Aldrich (St Louis,MO).Antibodies specifically directed against HIF-1a and HIF-1b were obtained from BD Biosciences Pharmingen(San Jose,CA).Anti-ubiquitin antibody was supplied from Sigma–Aldrich.Antibodies recognizingAbbreviations:HIF-1,hypoxia-inducible factor-1;miRNA,microRNA;mRNA,messenger RNA;S6K1,p70ribosomal S6kinase1;UTR,untranslatedregion.y These authors contributed equally to this work.ÓThe Author2012.Published by Oxford University Press.All rights reserved.For Permissions,please email:journals.permissions@661 at Shanghai Jiao Tong University on March 27, 2012 / Downloaded fromp70ribosomal S6kinase1(p70S6K1),p-p70S6K1,lamin A/C and HSP70 were obtained from Cell Signaling Technology(Beverly,MA).Chemical synthesis of M2and congeners1)Synthesis of6,8-bis(methylthio)-7-phenylpyrrolo[1,2-a]pyrazine(N3a)at Shanghai Jiao Tong University on March 27, 2012/Downloaded fromChromosomal DNA was precipitated with trichloroacetic acid and extracted with a solution containing0.5%SDS and0.5N NaOH.The radioactivity was measured using a liquid scintillation counter.In vitro cell invasion/migration assaysAn in vitro cell invasion/migration assay was performed using a24-well TranswellÒas described previously.The lower side of thefilter was covered with type I collagen,whereas its upper side was coated with Matrigel(Collaborative Research,Lexington,KY).The lower compartment was occupied with serum-free media with0.1%bovine serum albumin.HCT116cells were located in the upper compartment of the TranswellÒplate,maintained with10%serum for24h with vehicle or M2,fixed with methanol and then were stained with he-matoxylin for10min,briefly followed by eosin staining.The invasive phenotype was determined by quantifying the cells that migrated to the lower side of thefilter with microscopy(magnification,Â200). Thirteen visualfields were counted for eachfilter,and each sample was assayed in triplicate.An in vitro cell migration assay was per-formed using a24-well TranswellÒunit with polycarbonatefilters. Experimental procedures were the same as for the cell invasion assay but that thefilter was not coated with Matrigel.Data analysisOne-way analysis of variance procedures were used to assess significant differences among treatment groups.For each significant treatment effect, the Newman–Keuls test was utilized to compare multiple group means.ResultsInhibition of HIF-1a and its target gene inductionInsulin treatment induces HIF-1a through both de novo synthesis and protein stabilization(26,27).First,we investigated whether the oxi-dized metabolites of oltipraz inhibit HIF-1a induction by insulin in HCT116cells.Previously,oltipraz treatment at10or30l M sup-pressed the induction of HIF-1a(14).Treatment of the cells with M2inhibited HIF-1a induction,whereas M3,M4or M1failed to do so(Figure1B).M2at30l M completely inhibited the induction of HIF-1a,beginning from1h at least up to6h,and exhibited a dose–response effect(Figure1C).The inhibitory effect of M2on HIF-1a was con-firmed in another cell line(HT29).In addition,M2had an inhibitory effect on HIF-1a induction by other stimuli including hypoxia,CoCl2 or H2O2in HCT116cells(Figure1D).HIF-1a exists as a full length with826amino acids(28).The expression of a shorter form of HIF-1a is observed presumably because alternative splicing produces the smaller form of HIF-1a(735aa)with a weak activity(28).HIF-1a forms a heterodimer with its binding subunit HIF-1b, which undergoes nuclear translocation for target gene induction (3,29).Nuclear HIF-1a content was assessed in HCT116cells treated with M2in the presence or absence of insulin.M2treatment almost completely prevented the ability of insulin to elevate nuclear HIF-1a content in either HCT116or HT29cells(Figure2A).Consistently, M2abrogated increases in HIF-1a target gene transcript levels(i.e. vascular endothelial growth factor and glucose transporter1) (Figure2B).Moreover,the results of hypoxia-response element re-porter gene assays confirmed the inhibitory effect of M2on HIF-1a-dependent gene transcription(Figure2C).Fig.1.M2inhibition of HIF-1a induction by insulin.(A)The chemical structures of oltipraz and its pyrrolopyrazine metabolites.(B)The effects of oltipraz’s metabolites on the induction of HIF-1a by insulin.HCT116cells were treated with each metabolite or oltipraz(30l M each)for1h and continuously incubated in a medium containing100nM insulin for6h.(C)The time course and concentration–response effects of M2.HCT116or HT29cells were treated with30l M M2for 1–6h(left).They were also exposed to the indicated concentration of M2for6h(middle).The band intensity of HIF-1a relative to HIF-1b was quantified by scanning densitometry of the immunoblots(right).Value represents mean±SE fromfive independent experiments(treatment mean significantly different from vehicle-treated control,ÃÃP,0.01,or insulin,##P,0.01).(D)The effects of M2on the induction of HIF-1a by a variety of stimuli.HCT116cells were treated with30l M M2for 1h and were continuously incubated under the condition of hypoxia(1%oxygen)or in a medium containing100l M CoCl2or0.5mM H2O2for the indicated times.HIF-1a inhibition by pyrrolopyrazines663 at Shanghai Jiao Tong University on March 27, 2012 / Downloaded fromInduction of miR-199a-5p and miR-20a/bDespite the repression of HIF-1a protein,M2had no effect on the level of HIF-1a transcript,indicating that posttranscriptional mecha-nism might be involved in this process (Figure 3A).Since miRNAs play a role in the posttranscriptional regulation of HIF-1a ,we exam-ined the effect of M2on the expression of miRNAs that potentially interact with the 3#-untranslated region (UTR)of HIF-1A messenger RNA (mRNA).In an effort to identify miRNAs responsible for the inhibition of HIF-1a expression by M2,we searched the TargetScanDatabase and discovered that four sequence motifs of the 3#-UTR of HIF-1A match seed sequences of 10candidate miRNAs (i.e.miR-199a-5p,miR-20a/b,miR-93,miR-138,miR-18a/b,miR-106a/b and miR-519c).Among the candidate miRNAs,M2treatment signifi-cantly elevated the miRNA levels of miR-199a-5p,miR-20a and miR-20b in HCT116cells (Figure 3B).This effect of M2was also observed in HT29cells.Consistently,transfection of HCT116cells with each miRNA mimics or a mixture of the mimics almost com-pletely abrogated HIF-1a induction (Figure 3C).The mixture was also active in HT29cells (data not shown).These results support the contention that M2inhibits HIF-1a induction and which may result from the upregulation of miR-199a-5p and miR-20a/b that pair to the complementary-binding sites within the 3#-UTR of HIF-1A mRNA.To determine how M2elevates the levels of miR-199a-5p and miR-20a,their precursor levels were assessed by real-time PCR assays (Figure 3D).Interestingly,M2treatment increased the levels of pre-miR-199a and pre-miR-20a,but not their primary precursors,suggest-ing that the miRNAs induction may be mediated by the maturation process of the primary miRNAs.Since miR-20a is transcribed within the miR-17-92cluster,other mature miRNAs expressed from the cluster (miR-17,miR-18a,miR-19a,miR-19b-1and miR-92a-1)were also monitored (Figure 3E).M2treatment had no effects on the levels of the miRNAs,being consistent with our contention that M2affects the maturation process of miR-199a and miR-20a but not their general transcription.In addition,M4,another pyrrolopyrazine metabolite,did not change miR-20a/b levels but decreased miR-199a-5p level (Figure 3F).Taken together,our results demonstrate that M2has the ability to facilitate the maturation of primary precursors to miR-199a-5p and miR-20a that block the translation of HIF-1A mRNA.Inhibition of HIF-1a de novo synthesis35S-Methioninepulse labeling experiments were performed to verifythe inhibitory effect of M2on the synthesis of HIF-1a protein.As expected,M2treatment abolished the de novo synthesis of HIF-1a in HCT116cells (Figure 4A).Since HIF-1a is continuously degraded by the 26S proteasome complex after multiple ubiquitination (3),treat-ment with MG132(a proteasome inhibitor)promoted HIF-1a accu-mulation in cells treated with insulin (Figure 4B,upper).This effect was almost entirely abolished by simultaneous M2treatment in HCT116or HT29cells.Moreover,M2diminished the accumulation of ubiquitinated HIF-1a by MG132(Figure 4B,lower).These results are in line with the notion that the inhibition of HIF-1a de novo synthesis by M2restrained the ubiquitination and proteosomal degra-tion of HIF-1a because of a decrease in its supply.In a continuing effort to find the molecular mechanism of HIF-1a inhibition by M2,we next investigated whether M2alters the stability of HIF-1a .Cycloheximide,a general inhibitor of protein synthesis,was used to prevent HIF-1a de novo synthesis.HCT116cells treated with cycloheximide in combination with insulin displayed a gradual decrease in the level of HIF-1a as a function of time (Figure 4C).Simultaneous treatment of the cells with M2did not change the rate of HIF-1a degradation.These results demonstrate that HIF-1a repres-sion by M2may stem from the inhibition of HIF-1a de novo synthesis but not the change in the protein stability.Inhibition of cell proliferation and invasion/migrationHIF-1a activation promotes neovascularization,which features a key pro-cess in blood and nutrient supply (1–3).It is expected that HIF-1a sup-pression by M2inhibits cell proliferation.Having identified the ability of M2to suppress HIF-1a ,we examined its effect on cell proliferation and invasion/migration.M2significantly decreased the rate of DNA synthe-sis,as indicated by the result of [3H]-thymidine incorporation assay (Figure 5A).Moreover,M2treatment attenuated serum-induced inva-sion/migration of HCT116cells (Figure 5B).Similarly,transfection with a mixture of miR-199a-5p and miR-20a mimics significantly decreased the cell invasion and migration (Figure 5C).Our observations show that the inhibition of HIF-1a by M2may contribute to decreasing the pro-liferation and invasion/migration of the cancercell.Fig.2.M2inhibition of HIF-1a target gene induction.(A )Inhibition of HIF-1a nuclear accumulation by M2.The levels of nuclear HIF-1a weremeasured in the lysates of HCT116or HT29cells treated as described in Figure 1B.Immunoblottings for lamin A and HSP70confirmed the purities of nuclear (NF)and cytoplasmic fractions (CF),respectively.(B )Real-time PCR assays of HIF-1a target gene transcripts.After serum deprivation (16h),HCT116cells were treated with 30l M M2for 1h and were continuously incubated with 100nM insulin for 12h.b -Actin was used as a normalizing reference.Value represents mean ±SE from five independent experiments (treatment mean significantly different from vehicle-treated control,ÃÃP ,0.01,or insulin,##P ,0.01).(C )Hypoxia-response element (HRE)reporter gene assay.HRE-A549cells that had been stably transfected with the HRE-luciferase construct were incubated with 100nM insulin for 24h following M2treatment for 1h.Luciferase activity was measured on the lysates.Value represents mean ±SE from five independent experiments (treatment mean significantly different from vehicle-treated control,ÃÃP ,0.01,or insulin,#P ,0.05;##P ,0.01).S.G.Kang et al.664at Shanghai Jiao Tong University on March 27, 2012/Downloaded fromHIF-1a repression by 7-ethyl-6,8-bis(methylthio)pyrrolo[1,2-a ]pyrazine Based on the novel pharmacologic effect of M2,pyrrolopyrazine analogs were synthesized with the aim of identifying other candidates that inhibit HIF-1a (Figure 6A).7-Ethyl-6,8-bis(methylthio)pyrro-lo[1,2-a ]pyrazine (N3b)had an inhibitory effect on HIF-1a induction,whereas 6,8-bis(methylthio)-7-phenylpyrrolo[1,2-a ]pyrazine (N3a)failed to do so (Figure 6B).Thus,7-ethyl derivative (N3b),but not 7-phenyl derivative (N3a),of pyrrolopyrazine was active in inhibiting HIF-1a .N3b treatment also induced the expression of miR-199a-5p and miR-20a,as did M2(Figure 6C).Overall,our results demonstrate that M2with pyrrolopyrazine structure and its 7-ethyl congenor have the ability to prevent the induction of HIF-1a ,and which may result from the inhibition of HIF-1a de novo synthesis,as mediated by the induction of miR-199a-5p and 20a.DiscussionHIF-1a is closely involved in angiogenesis and tumor growth (1–3).In previous studies,oltipraz inhibits the expression of vascular en-dothelial growth factor at the molecular level and has the ability to inhibit tumor growth and angiogenesis in tumor-bearing animal models (e.g.angiosarcoma and colon cancer)(13,14,30).Oltipraz treatment inhibits the activity of HIF-1a and its target gene trans-activation,thereby repressing angiogenesis and tumor growth (14).This effect appears to result from not only an increase in HIF-1a ubiquitination but also its accelerated degradation.In addition,olti-praz decreased the ability of insulin to produce H 2O 2,reactive oxy-gen species that inhibit prolyl hydroxylase domain protein-mediated hydroxylations of two proline residues of HIF-1a and consequent degradation of HIF-1a by the ubiquitin-proteasome system (14).Thus,the inhibition of HIF-1a by oltipraz results from destabiliza-tion of HIF-1a .Our results and others showed that the formation of M2occurs quickly after oltipraz administration in human or animals (19)and that M2exerts similar or possibly more potent beneficial effects than its parent compound in several experimental models.Therefore,it was anticipated that M2inhibits the activity of HIF-1a in cancer cells.Here,we demonstrated for the first time that M2and its 7-ethyl congener,but not M3and M4,have an inhibitory effect on the in-duction of HIF-1a .Moreover,our results demonstrated that M2treat-ment diminished cancer cell invasion and migration.In this study,we found no accumulation of ubiquitinated HIF-1a after M2treatment:Fig.3.The induction of mature miR-199a-5p and miR-20a/b by M2.(A )The effect of M2on HIF-1a mRNA level.HIF-1a mRNA levels were measured in HCT116cells treated as described in Figure 1B.Actinomycin D (an inhibitor of gene transcription)was used as a positive control.Value represents mean ±SE from five independent experiments (treatment mean significantly different from vehicle-treated control,ÃÃP ,0.01).(B )The effect of M2on the expression of miR-199a-5p and miR-20a/b.The levels of miRNAs were determined by real-time PCR assays in cells treated with 30l M M2for 3h after serum deprivation.Each miRNA level was normalized to that of U6snRNA.Value represents mean ±SE from three to five independent experiments (treatment mean significantly different from vehicle-treated control,ÃÃP ,0.01).(C )HIF-1a repression by the mimics of miR-199a-5p or miR-20a.HIF-1a was immunoblotted on the lysates of HCT116cells treated with the mimics of 100nM miR-199a-5p,miR-20a or combination of both (50nM each)for 3h.Equal loading of proteins was verified by immunoblottings for HIF-1b .(D )Real-time PCR assays.The levels of primary precursor or precursor miRNAs were measured as described in the legend to panel B.(E )The effect of M2on the levels of miRNAs in the miR-17-92cluster.(F )The effect of M4on miR-199a-5p and miR-20a/b levels.For the panels D,E and F,HCT116cells were treated with M2(or M4)as described above.Value represents mean ±SE from four to five independent experiments (treatment mean significantly different from vehicle-treated control,ÃP ,0.05,ÃÃP ,0.01).N.S.,Not significant.HIF-1a inhibition by pyrrolopyrazines665at Shanghai Jiao Tong University on March 27, 2012/Downloaded fromM2failed to change the half-life time of HIF-1a .Thus,the basis of HIF-1a inhibition by M2seems to differ from that of oltipraz (14).This is in line with our previous observation that M2,unlike oltipraz,failed to protect hepatocytes from oxidative stress elicited by arach-idonic acid and iron (21).So,it is probably that M2has a unique molecular basis for HIF-1a repression.Pyrrolopyrazine thione,an intermediate metabolite of oltipraz,causes the reduction of the heme iron via the interaction with cytochrome c (31–33).The membrane-bound cytochrome c then scavenges superoxide radicals and dimin-ishes excessive H 2O 2production (31–33).However,M2did not have this effect (21).Thus,the lack of accumulation in ubiquiti-nated HIF-1a by M2might result from decreased de novo synthesis.NF-j B heterodimer complex consisting of p50and p65transactivates the HIF-1A gene (34),which may be antagonized by certain chemicals (e.g.YC-1)(35).Unlike YC-1,M2had no effect on p50/p65expression or NF-j B reporter activity (data not shown),excluding the possibility that M2inhibits HIF-1a by repressing NF-j B activity.The lack of change in HIF-1A mRNA content by M2suggested that it inhibits HIF-1a at the posttranscriptional level.This hypothesis is strengthened by our observation showing a decreased translational rate of HIF-1a in cells treated with insulin that elevates HIF-1a level through a translational mechanism (26,27).Consistently,M2treatment significantly restrained the synthesis rate of HIF-1a .Since 5#-terminal oligo pyrimidine sequences play a role in p70S6K1-dependent translation of HIF-1A mRNA (36),the inhibition of p70S6K1by olti-praz might contribute to inhibiting HIF-1a (14).Unlike oltipraz,M2failed to change p70S6K1phosphorylation (supplementary Figure S1is available at Carcinogenesis Online;the capital S denotes that the figure can be found in the supplemental material),implying that a dif-ferent mechanism exists for the translational regulation of HIF-1a by M2.The miRNAs are small non-coding RNAs that decrease the expres-sion of target proteins via the inhibition of mRNA translation with enzymatic and regulatory functions (37,38).The inhibitory action of miRNA results from imprecise base pairing of miRNAs withtheFig.4.Inhibition of HIF-1a de novo synthesis by M2.(A )[35S]-Methionine incorporation assay.HCT116cells that had been cultured in a medium free of methionine,cysteine and glutamine were pulse labeled with 100mCi/ml [35S]-methionine for 1h and were incubated in a medium containing methionine,cysteine and glutamine for 1.5h.[35S]-Methionine incorporation to HIF-1a was assessed in the lysates.Value represents mean ±SE from five independent experiments (treatment mean significantly different from vehicle-treated control,ÃÃP ,0.01,or 35S labeling alone,##P ,0.01).(B )The effect of M2on the ubiquitination and degradation of HIF-1a .HCT116or HT29cells transfected with the plasmid encoding for His-tagged ubiquitin (His-Ubi)were treated with MG132(10l M,3h)and were treated with insulin (100nM,3h)or insulin plus M2.Immunoblotting for ubiquitin was done on HIF-1a immunoprecipitates.HIF-1b wasimmunoblotted on the lysates.Value represents mean ±SE from five independent experiments (treatment mean significantly different from vehicle-treated control,ÃÃP ,0.01,or insulin alone,##P ,0.01).(C )The effect of M2on HIF-1a stability.HCT116cells were incubated with insulin or insulin þM2for the indicated times after cycloheximide (CHX,20l g/ml)treatment (upper).HIF-1a or HIF-1b was immunoblotted on the lysates (lower).Value represents mean ±SE from five independent experiments (treatment mean significantly different from zero-time control,ÃÃP ,0.01,or insulin þCHX at respective time,##P ,0.01).S.G.Kang et al.666at Shanghai Jiao Tong University on March 27, 2012/Downloaded from3#-UTR of complementary mRNAs (37,38).It is known that a cluster of miR-17-92and miR-519c participate in hypoxia-induced HIF-1a accumulation (39,40).Under a normoxic condition,miR-17-92tran-scribed from the same cluster may target complementary HIF-1A mRNA for the regulation of its translation,playing an inhibitory role in cancer proliferation in an oxygen-independent manner (39).In rat hepatic sinusoidal endothelial cells,downregulation of miR-199by ethanol increased the levels of HIF-1A and endothelin receptor 1tran-scripts (41).Certain phytochemicals (i.e.curcumin,epigallocatechin-3-gallate and isoflavone)also affect the expression of miRNAs,leading to alterations in cellular signaling and biological behavior (42).However,the modulation of miRNAs responsible for HIF-1a translation by chemical means had never been explored.Here,we demonstrate for the first time that M2and its 7-ethyl congenor induce the expression of miR-199a-5p and miR-20a responsible for the pre-vention of HIF-1a induction.Oltipraz and its major metabolite of M1,however,failed to increase the levels of miR-199a-5p andmiR-20aFig.5.Inhibition of cancer cell proliferation and invasion/migration by M2.(A )Inhibition of DNA synthesis by M2.HCT116cells were treated with 30l M M2for 24h and continuously exposed to 1l Ci/ml[methyl-3H]-thymidine for 8h.Cell proliferation was determined by measuring radioactivity.Value represents mean ±SE from fiveindependent experiments (treatment mean significantly different from vehicle-treated control,ÃÃP ,0.01,or serum alone,##P ,0.01).(B )Repression of cell invasion/migration by M2.Invaded/migratedHCT116cells were counted under light microscope (magnification,Â200).Value represents mean ±SE from four independent experiments (treatment mean significantly different from vehicle-treated control,ÃP ,0.05,ÃÃP ,0.01,or serum alone,#P ,0.05).(C )The effect of a mixture of miR-199a-5p and miR-20a mimics on cell invasion/migration.HCT116cells were treated with miR-199a-5p and miR-20a mixture (50nM each).Value represents mean ±SE from four independent experiments (treatment mean significantly different from vehicle-treated control,ÃÃP ,0.01,or serum alone,#P ,0.05,ÃP ,0.01).Fig.6.The effects of synthetic analogs of M2on the induction of HIF-1a .(A )The chemical structures of pyrrolopyrazine derivatives (N3a and N3b).(B )HIF-1a repression by N3b.HCT116cells were treated with each pyrrolopyrazine derivative (30l M each)for 1h and were continuously incubated in a medium containing 100nM insulin for 6h.Value represents mean ±SE from five independent experiments [treatment mean significantly different from insulin treatment alone (100%),ÃÃP ,0.01].(C )The effect of N3b treatment on miR-199a-5p and miR-20a expression.miR-199a-5p and miR-20a levels were measured by real-time PCR assays in HCT116cells treated with 30l M N3b for 3h after serum deprivation.Each miRNA level was normalized to that of U6snRNA.Value represents mean ±SE from five independent experiments (treatment mean significantly different from vehicle-treated control,ÃP ,0.05,ÃÃP ,0.01).HIF-1a inhibition by pyrrolopyrazines667at Shanghai Jiao Tong University on March 27, 2012/Downloaded from。
