Research Article Cancer Prevention Research Transcriptional … · Transcriptional Attenuation in...

Rese

TraRes

Maria

Abst

Intro

Thecolonvels o2 to 5and dcells isourceacidstion (ture, bgenession mcascaddriven

AuthorCenter,

Note:SResear

CorresMontef718-92

doi: 10

©2010

Cance1292

Do

Published OnlineFirst September 14, 2010; DOI: 10.1158/1940-6207.CAPR-10-0083

Cancer
eventionesearch
arch Article

nscriptional Attenuation in Colon Carcinoma Cells in

PrR

ponse to Butyrate

C. Daroqui and Leonard H. Augenlicht

ractThe

arrest,for Nc-mycwe deSW83tion soccupchromlase inbutionwith con aceysis ugenombutyra

e of evenby seque

s' AffiliatioMontefiore

upplementach Online (h

ponding AioreMedical0-2093; Fax

.1158/1940-

American A

r Prev Res

wnloade

short-chain fatty acid sodium butyrate (NaB), produced in the colonic lumen, induces cell cycledifferentiation, and/or apoptosis in colorectal carcinoma cells in vitro, establishing a potential roleaB in colon cancer prevention. We have previously shown that butyrate decreases cyclin D1 andexpression, each essential for intestinal tumor development, by transcriptional attenuation. Here,termined that butyrate-induced transcriptional attenuation of the cyclin D1 and c-myc genes in7 human colorectal adenocarcinoma cells occurs at ∼100 nucleotides downstream of the transcrip-tart site, with a similar positioning in Caco-2 cells. A concomitant decrease in RNA polymerase IIancy at the 5′ end of each gene was observed. Because transcriptional regulation is associated withatin remodeling, we investigated by chromatin immunoprecipitation whether the histone deacety-hibitory activity of butyrate altered chromatin structure at the attenuated loci. Although the distri-s of histone H3 trimethylated on K4 and K36 along the cyclin D1 and c-myc genes were consistenturrent models, butyrate induced only modest decreases in these modifications, with a similar effecttylated H3 and a modest increase in histone H3 trimethylated on K27. Finally, transcriptome anal-sing novel microarrays showed that butyrate-induced attenuation is widespread throughout thee, likely independent of transcriptional initiation. We identified 42 loci potentially paused byte and showed that the transcription patterns are gene specific. The biological functions of these

loci encompass a number of effects of butyrate on the physiology of intestinal epithelial cells. Cancer PrevRes; 3(10); 1292–302. ©2010 AACR.

hallmtissuec-m

rate dcomplatescells m(8). MnalingcolonwhichT-cellgenesD1 re

duction

short-chain fatty acid butyrate is produced in theby fermentation of dietary fiber, which generates le-f ∼20 mmol/L in the colonic lumen, well above themmol/L concentration that induces cell cycle arrestifferentiation in culture (1–3). Colonic epithelialn vivo use these high levels as their principal energy(4), and efficient metabolism of short-chain fatty

is necessary for their induction of colon cell matura-5, 6). Gene expression profiling showed that, in cul-utyrate altered the steady-state levels of ∼7% of theinterrogated, either increasing or decreasing expres-onotonically with time (7). This implied an orderly

ts, with successive stages in cell maturationntial recruitment of genes and pathways, a

showiThus,and cactivitTra

nizedcyclooxdameas concarcinbutyrac-myc,

n: Department of Oncology, Albert Einstein CancerMedical Center, Bronx, New York

ry data for this article are available at Cancer Preventionttp://cancerprevres.aacrjournals.org/).

uthor: Maria C. Daroqui, Department of Oncology,Center, 111East 210thStreet, Bronx, NY10467. Phone:: 718-882-4464; E-mail: [email protected].

6207.CAPR-10-0083

ssociation for Cancer Research.

; 3(10) October 2010

Cancer Research. on Jcancerpreventionresearch.aacrjournals.org d from

ark of an integrated response necessary for intestinalhomeostasis.yc and cyclin D1 are genes downregulated by buty-uring colonic cell maturation. Cyclin D1 is a drivingonent of the cell division cycle, and c-myc stimu-cell cycling and is important in differentiation asigrate from the crypt toward the intestinal lumenoreover, both genes are direct targets of Wnt sig-(9, 10), a fundamental pathway deregulated intumors, most commonly by mutation of Apc,alters the transcription mediated by β-catenin-factor targeting the cyclin D1, c-myc, and other(11). Targeted inactivation of either c-myc or cyclinduces Apc-initiated intestinal tumor formation,ng that both are essential for tumorigenesis (12, 13).butyrate-mediated decreased expression of cyclin D1-myc may be fundamental to its chemopreventivey (14, 15).nscriptional attenuation or pausing, first recog-for c-fos, c-myb, adenosine deaminase, tubulin, andygenase-2 (16–20), is now understood to be fun-ntal for the regulation of many inducible as wellstitutively expressed genes (21, 22). In colorectaloma cells, a transcriptional block in response to
te was shown by Heruth and coworkers (14) forwhich was further explored by us (23). Cyclin D1
uly 21, 2021. © 2010 American Association for

http://cancerpreventionresearch.aacrjournals.org/

transcto bepromohibitogrowicyclindownThus,butyramRNATran

landmviewe

of HDand chas hisscriptdetermscriptbutyrlengthinvestarraysinduc

Fig. 1.transcrto butytreated5 mmosynthestiling mMethod(Cy5-la(Cy3-lapoint ofirst 355(A) andPoints,indepenSD. Onprobe nnucleotratio ofcells fofor eacin the mwhere twas ind

Butyrate Induction of Transcriptional Attenuation

www.a

Do


riptional downregulation, however, was reportedassociated with transcription factors acting at theter in response to histone deacetylase (HDAC) in-rs and other agents (15, 24). We used imaging ofng transcripts at their site of synthesis to show thatD1 was also transcriptionally paused by butyratestream of the transcription start site (TSS; ref. 25).transcriptional pausing of c-myc and cyclin D1 byte can play a major role in determining steady-statelevels (23, 25).scriptional regulation is associated with chromatin
arks, such as specific histone modifications (re-
d in ref. 26). In this regard, butyrate is an inhibitorrate wsugge

uced by NaB. nt, nucleotide.

acrjournals.org

Cancer Research. on Jcancerpreventionresearch.aacrjournals.org wnloaded from

AC activity, which leads to histone hyperacetylationromatin remodeling of induced target genes, where-tone hypoacetylation has been associated with tran-ionally silent chromatin (27, 28). Here, we firstined the position near the 5′ end at which tran-

ional pausing of c-myc and cyclin D1 is initiated byate, as well as chromatin modification along theof the gene-coding regions. We then extended the

igation to the transcriptome level using novel micro-, identifying a subset of sequences for which butyrateed transcriptional pausing. This regulation by buty-
as highly complex and gene specific, and the datast that the attenuation mechanism is independent
Cyclin D1 and c-mycipt expression in responserate. SW837 cells werefor 0, 6, and 8 h withl/L NaB, and cDNA wasized and hybridized to aicroarray (Materials ands). The ratio of NaB-treatedbeled) to untreatedbeled) cells at each timef treatment is shown for thenucleotides of the cyclin D1c-myc (B) sequences.mean from threedent experiments; bars,the abscissa, theumbers indicate the firstide of the probe. Insets,NaB-treated to untreatedr every probe on the arrayh sequence (color-coded asain figure). *, positionranscriptional attenuation

Cancer Prev Res; 3(10) October 2010 1293



of metional

Mate

CellsHum

235)(23, 2(MycoExponted o(Cacotime pCaco-induc

RNA aRNA

for dimicro(Nimbtrogen

CustocDN

facturcleotidDetail

HistoWes

tonesSuppl

ChromTo

modifprecippleme

QuancDN

as DNanalyz(Applcol, as

AnalydistriEac

the 5′tologyNIH),

StatisThe

dent's

comping ar

Resu

TransWe

of cycma SWarray(Mateto Cy(butynucle5 mmceded∼24 hsionquencexperiout thbridizeitherquenctreatm

Fig. 2.and his6 to 2416 h of butyrate treatment. The total level of histone H3 is also shown.NB, no

Daroqui and Augenlicht

Cance1294

Do


chanisms that determine the frequency of transcrip-initiation.

rials and Methods

and butyrate treatmentan colorectal adenocarcinoma SW837 cells (CCL-

and Caco-2 cells (HTB-37) were grown as described9) and monitored for Mycoplasma contaminationplasma Plus PCR Primer Set, Agilent Technologies).entially growing cells at ∼70% confluency were trea-r not with 5 mmol/L (SW837 cells) or 2 mmol/L-2 cells) sodium butyrate (NaB; Sigma) for differenteriods. A lower concentration of NaB was used in2 cells, as these cells are more sensitive to butyratetion of cell death.

nd probe preparationwas isolated from cells treated or not with NaB

fferent time periods. Double-stranded cDNA forarrays was synthesized from RNA as describedleGen) using SuperScript cDNA synthesis kit (Invi-). Details are given in Supplementary Methods.

m-designed oligonucleotide microarraysA gene expression arrays were designed and manu-

ed in collaboration with NimbleGen. Two oligonu-e arrays were used: “5′ 3′ array” and “tiling array.”s are provided in Supplementary Methods.

ne purification and Western blot analysistern blot analysis and isolation of acid soluble his-were done as described (7, 30). Details are given inementary Methods.

atin immunoprecipitationstudy polymerase II (Pol II) occupancy and histoneication at a specific sequence, chromatin immuno-itation assay was done as described in detail in Sup-ntary Methods.

titative real-time PCRA from butyrate-treated and untreated cells, as wellA isolated by chromatin immunoprecipitation, wased by quantitative real-time PCR using SYBR Greenied Biosystems) following the manufacturer's proto-described in Supplementary Methods.

sis of gene ontology and chromosomebutionh sequence potentially paused by butyrate based on3′ array data was classified according to its gene on-by DAVID Bioinformatics Resources 2008 (NIAID/as described in Supplementary Methods.

tical analysis
mean value of replicates was analyzed using Stu-t test. One-way ANOVA and Bonferroni multiple
at nufor c-

r Prev Res; 3(10) October 2010


arison test were used to analyze the data from the til-ray. Results were considered significant when P < 0.05.

lts

cription of the cyclin D1 and c-myc genesfirst analyzed the effects of NaB on the transcriptionlin D1 and c-myc in human colorectal adenocarcino-837 cells using a custom-designed oligonucleotidethat tiled through the length of these sequencesrials and Methods). These arrays were hybridized3-labeled (untreated, cycling cells) or Cy5-labeledrate-treated cells) cDNA transcribed from total,ar, or cytoplasmic RNA. Cells were treated withol/L NaB for 0, 6, or 8 hours, time points that pre-the G1 arrest and apoptosis induced by NaB atours (not shown). Figure 1 focuses on the expres-level across the first 355 nucleotides of each se-e, shown as the average ratio of three independentments, and the insets show the expression through-e entire coding regions. As expected, the ratio of hy-ation of probes from untreated cells labeled withfluorochrome was approximately 1 along each se-e (Fig. 1). For cyclin D1, 6 or 8 hours of butyrateent decreased its expression by ∼50%, beginning

nspecific band.

Effect of butyrate on cyclin D1 and c-myc protein expressiontone modification. A, Western blot for cyclin D1 and c-myc ath of butyrate treatment. B, AcH3 levels in response to 30 min to

cleotide 92 (Fig. 1A). There was a similar patternmyc, with its expression decreased by 80% to 90%

Cancer Prevention Research



downvels olengthstep wcyclinods).izatioBonfenificanno sigCy5. Wand cadenowereexprestary Fther dFig. Sof thelengthcells.

spikesalongtial hprecausets)cyclinby Natreatm

RNA PcyclinIn

matincy tosequelevels(Fig.of thedecrea

Fig. 3.three e


www.a

Do


stream of nucleotide 84 (Fig. 1B). The repressed le-f both genes generally persisted throughout theof each sequence (Fig. 1, insets). A normalizationas done for each of the 628 and 347 probes forD1 and c-myc, respectively (see Materials and Meth-ANOVA for the NaB-treated data sets (after normal-n) versus untreated cells (0 hours), followed byrroni multiple comparison test, showed a highly sig-t difference at both 6 and 8 hours (P < 0.0001), butnificant difference between 0-hour Cy3 and 0-houre further investigated the effects of NaB on cyclin D1

-myc transcription in another human colorectalcarcinoma cell line, Caco-2, and similar resultsobtained. For cyclin D1, the decrease in transcriptsion was localized at nucleotide 107 (Supplemen-ig. S1A). For c-myc, the decrease was somewhat fur-ownstream, at nucleotide 186 (Supplementary

1B), but still well within the 5′ untranslated regionmessage, thus abrogating the generation of a full-
coding mRNA in Caco-2 cells, as it did in SW837 as sho
Chromatin immunoprecipitation of Pol II and modified histones. A, Pol II occupanxons that make up the cyclin D1 and c-myc coding sequences, respectively. B,

acrjournals.org


in the generally consistent pattern of repressionboth sequences, which could be related to differen-ybridization of some probes on the array, despitetions taken in designing the arrays [e.g., Fig. 1 (in-and Fig. 5B]. By Western blot, we confirmed thatD1 and c-myc expression was ultimately reducedB at the protein level between 6 and 24 hours ofent (Fig. 2A).

ol II occupancy and histone modification alongD1 and c-myc

untreated SW837 cells, analysis of Pol II by chro-immunoprecipitation revealed a higher occupan-ward the 5′ end, relative to the rest of thence, in the cyclin D1 gene, whereas more similarwere observed throughout the length of c-myc

3A, black columns). Following butyrate treatmentcells for 6 hours, Pol II occupancy for cyclin D1

sed most substantially in exon 1 (nucleotides 1-407),
wn in Fig. 3A (gray columns). Whereas the butyrate-
In both cell lines, there were highly reproducible induced difference at the 5′ end of cyclin D1 was not

cy at 6 h of butyrate treatment (5 mmol/L) along the five andAcH3 levels throughout each sequence.




observwith tapproBas

genera

of butin the30 m

Fig. 3.normali(Ct) valuData ar


Cance1296

Do


ed downstream, for c-myc the pattern was different,he butyrate-induced decrease being maintained atximately the same level in exons 2 and 3 (Fig. 3A).
ed on the well-characterized activity of butyrate as a and p
es (Materials and Methods), and gray columns indicate butyrate effects, with dae from at least three independent experiments. Similar results were found at 8 h



yrate on chromatin remodeling. An ∼8-fold increaseoverall level of histone H3 acetylation (AcH3) at

inutes following butyrate treatment was observed
ersisted for ∼6 hours (Fig. 2B). Despite this overall
l inhibitor of HDACs (31), we determined the effects induction of AcH3, it has been reported that butyrate

Continued. C to E, H3me3K4, H3me3K36, and H3me3K27 levels. Immunoprecipitated samples were analyzed by quantitative real-time PCR andzed by their respective input. In each case, black columns indicate basal levels in untreated cells, with data expressed as the mean ± SD of linear

ta expressed as the ratio of NaB-treated to untreated cells.of butyrate treatment (not shown).




induclatedtern ounindtreatmcyclinpolymremaidecrea4 andassociB). Thsequegene ocompfold bAcH3colocaWe

impomethyassocitone Hwith e27 (HlencinH3meconsisthe hithe hiof c-mH3meincreation win thebutyrageneraal atteaccouH3meand wwith tsion, aattenuSim

sponstary Fwhichlevelsboth s

TransThe

rate dclear frationbutyraalterin

in outent tocontriused athe 5′codinIn SWfor 2 tidentsequeregul(Fig. 4regulamainiof dowber noEach oattenuBec

we invat thegenomtion osentatTheresequebandwithshowninducrathermodifRec

thougeratesmuchscriptof moing coviduaWe

latedcustomeachc-mycthat s85 seqexpres3′ arrausingweretowarpause42 seqpresenI sequ


www.a

Do


es histone hypoacetylation of target genes downregu-in expression (32, 33). For both genes, a similar pat-f distribution for AcH3 as for Pol II was found inuced cells (Fig. 3B). However, following butyrateent, AcH3 levels were differentially altered alongD1, with modest decreases associated with decreasederase loading at the 5′ end, whereas for c-myc, AcH3ned consistently decreased, similar to the persistentse of Pol II (Fig. 3B). The increases in AcH3 in exons5 of cyclin D1 after butyrate treatment were notated with increases in Pol II occupancy (Fig. 3A andis may be a technical issue because there is a humannce (FLJ42258) located downstream of the cyclin D1n chromosome 11q13.2 and transcribed from thelementary strand, whose transcript is induced ∼8-y butyrate (not shown). Hence, the elevation inmay be due to the overlap of this transcript thatlizes with the 3′ end of cyclin D1.investigated additional histone methylation marksrtant in transcriptional regulation. These are tri-lated histone H3 on lysine 4 (H3me3K4), generallyated with transcription initiation; trimethylated his-3 on lysine 36 (H3me3K36), generally associatedlongation; and trimethylated histone H3 on lysine3me3K27), considered a mark of transcriptional si-g (32, 33). Figure 3C shows that in untreated cells,3K4 was relatively higher at the 5′ end of cyclin D1,tent with its role in transcription initiation, althoughghest level was in exon 2, not in exon 1. Similarly,ghest level of this modification was found in exon 2yc (Fig. 3C). As shown in Fig. 3D, the pattern of3K36 in uninduced cells was different, in each genesing toward the 3′ end, consistent with its associa-ith elongation (33). Whereas there were decreasesse histone modifications along each gene followingte treatment (Fig. 3C and D), these were modest andlly further downstream of the marked transcription-nuation seen in Fig. 1A and B and, thus, may notnt for this butyrate effect. Finally, the level of3K27 was low along each gene (note scale of axis)as increased by butyrate (Fig. 3E), in accordancehe role of this modification in transcriptional repres-lthough the data do not suggest a specific role in theation of cyclin D1 and c-myc by butyrate.ilar patterns and modest histone modification in re-e to butyrate were seen for Caco-2 cells (Supplemen-ig. S2), with the exception of Pol II distribution,increased at exon 1 of c-myc, and H3me3K4, whosein response to butyrate were somewhat elevated inequences (Supplementary Fig. S2).

criptome analysisabove studies and our prior work focused on buty-ownregulation of cyclin D1 and c-myc because of theunctional roles of these genes in colonic cell matu-and colon tumorigenesis. However, the effects of
te on colon carcinoma cells are highly pleiotropic,g the expression of ∼7% of sequences interrogated
eithersion f

acrjournals.org


r prior report (7). We therefore investigated the ex-which the regulation of transcriptional elongation

butes to downregulation of gene expression. We first5′ 3′ microarray, with five probes within 100 bp ofend and 3′ end, for each of the 17,378 canonical

g sequences in GenBank (Materials and Methods).837 cells untreated or treated with 5 mmol/L NaBo 12 hours, data from two independent experimentsified an overlap of 367 sequences (∼2% of thences analyzed), the expression of which was down-ated at the 3′ end following butyrate treatmentA). Approximately 53% of these sequences were up-ted by butyrate at the 5′ end. For many of the re-ng sequences, there were differences in the extentnregulation at the 5′ and 3′ ends, with a large num-t altered in expression at all at the 5′ end (∼43%).f these sequences was a candidate for transcriptionalation downstream of the TSS.ause butyrate can induce chromatin modification,estigated whether the 367 sequences downregulated3′ end were clustered in regions in the humane. Figure 4B shows that there was a linear distribu-f these loci as a function of the frequency of repre-ion of genes per chromosome on the 5′ 3′ array.fore, there was no overall physical clustering of thesences. A further classification by chromosomal cyto-showed that only ∼12% were clustered in groups,a maximum of four sequences per cytoband (not). Therefore, as for cyclin D1 and c-myc, butyrate-ed transcriptional attenuation may be targeted,than involving broad regions affected by chromatinication.ent data show that transcriptional units, previouslyht to have a straight-forward organization that gen-a single canonical pre-mRNA transcript, are in factmore complex. In particular, multiple short tran-

s are generated at the 5′ end as well as the 3′ endst genes (34). Thus, there is likely to be an underly-mplexity in the structure and regulation of the indi-l transcripts identified in Fig. 4A.therefore analyzed the 367 sequences downregu-at the 3′ end in response to NaB using another-designed microarray with probes that tile through

of these sequences, as was done for cyclin D1 and(Fig. 1). Focusing on the 6- to 8-hour time periodhowed the greatest changes in expression, there wereuences whose 3′ end of the canonical transcript wassed at a lower level than the 5′ end, based on the 5′y analysis. Data from four independent experimentsthe tiling array confirmed that 42 of these sequencesexpressed to a significantly lower extent (>1.5-fold)d the 3′ end than at the 5′ end and, thus, potentiallyd by butyrate. Closer examination revealed that theseuences fell into two general classes, which were re-ted equally, as shown in Table 1 and Fig. 5. Categoryences were upregulated at the 5′ end by butyrate, but
downregulated or not significantly altered in expres-urther downstream, as illustrated by the TMOD1 gene



(Fig. 5showealongative pbutyration, shalf oof exosuggesever,time Pform(datateredfor cycneousproduencom5B), oby buresultple sitThe

their gIDs gigets inizatipossibcycle aIn cothe tetransction (was a

TabSW

Cat

I

II

NOpauaccregilatiocateExa


Cance1298

Do


A). Twenty-eight of the 29 sequences in category Id similar patterns of graded decrease in expressionthe coding sequence, although they varied in the rel-osition at which expression was downregulated byte. TMEM49 sequence (NM_030938) was the excep-howing increased expression by butyrate for the firstf the sequence but an abrupt decrease at the junctionns 10 and 11 (Supplementary Fig. S3), which couldt induction of alternative splicing by butyrate. How-by reverse transcription-PCR and quantitative real-CR, we have not detected any alternatively splicedof this gene in SW837 cells in response to butyratenot shown). Category II genes (Table 1) were not al-by butyrate in expression at the 5′ end, as is the caselin D1 and c-myc (Fig. 1), but then showed heteroge-patterns among the sequences, which was highly re-cible across independent experiments. These patternspassed decreased expression, such as for TFAP4 (Fig.r increased expression before being downregulatedtyrate, as delineated in Table 1. These patterns mayfrom complex transcription, characterized by multi-es of initiation and termination as well as pausing.distribution of these two categories with respect toene ontology is shown in Table 2, with their geneven in Supplementary Table SIA. The repressed tar-n category I were enriched in “cytoskeleton orga-on” (∼18%; Table 2A) and “apoptosis” (∼15%),ly contributing to the induction by butyrate of cellrrest and differentiation in the cells analyzed (2, 3).ntrast, category II sequences were associated withrms “zinc ion binding,” a characteristic of manyriptional factors involved also in transcriptional elonga-
35, 36); indeed, the term “regulation of transcription”lso linked to this category (Table 2B). However, for
categoand Msignintial pawhich

Fig. 4.assessdownreData focells atexpressMethodRed anrespectof the 3the seq

le t83

eg r e

d d d

4

Up o

mples of each category are shown in Fig. 5.



ry II, the input number of genes analyzed (Materialsethods) was below the threshold required for as-g a P value. Nevertheless, the data suggest a differen-

67 sequences downregulated by butyrate at the 3′ end relative touence representation on the array.

Transcript expression of sequences downregulated by butyrate,ed by the 5′ 3′ microarray. A, a total of 367 sequences weregulated by NaB at 2 to 12 h of treatment (Materials and Methods).r the 5′ end are expressed as the ratio of NaB-treated to untreatedeach time point of treatment, whereas the 3′ end data areed relative to the 5′ end, as the “ratio of ratios” (Materials ands). A heat map of two independent experiments is shown.d green, upregulation and downregulation in response to butyrate,ively; black, no significant change. B, chromosome distribution

1. Class7 cells in

ification of transcript paresponse to butyrate

terns in

ory

5′ En
Coding
Mid

egion

le 3′ En
Sequ nces (%)
Up
NS 7.6 UpUp Dow n NS 2.4
Up/D
wn Down 2.4 Dow n Down 1.4 NS 2
NS Up NS 16.7NS Down NS 9.5

TE: (A) The 42 sequences found to be potentiallysed by the 5′ 3′ array and the tiling array were classifiedording to NaB effect throughout the canonical codingon. Up and Down indicate upregulation and downregu-n, respectively (>1.5-fold change), whereas NS indi-s no significant change in response to butyrate.

ttern of transcriptional downregulation by butyrate,may be related to different biological functions.




Discu

Buthibitoand wchemoButyra

epige(revieon twfundagenes

Fig. 5.patternby butymicroarcategorTable 1and TFData arNaB-treArrowsanalyzePCR (into untre


www.a

Do


ssion

yrate, a natural dietary product and an HDAC in-r fundamental to normal colonic cell maturationith potential activity in chemoprevention and
therapy, has extensive effects on gene expression.te alters gene expression as we
tome

ated cells).

acrjournals.org


netic mechanisms of transcriptional regulationwed in ref. 37). We first focused on the effectso specific genes, cyclin D1 and c-myc, which aremental in intestinal cell maturation and tumori-is, and then expanded the analysis to the transcrip-
level, which revealed a complexity of response to
ll as the underlying butyrate not previously appreciated.

Transcript expressions of genes downregulatedrate, assessed by the tilingray. Category I (A) andy II (B) sequences fromare exemplified by TMOD1AP4 genes, respectively.e expressed as the ratio ofated to untreated cells.indicate the regionsd by quantitative real-timesets, ratio of NaB-treated




Tranexontumo(38, 3tilingafter 6compmarkedue tobegin<1% arespecthat tto thesteadymay bin c-mwell aintestideletidecreagestinsensitTran

attribunationin Cactreatminduclocus,towarSW83ciatedplexesBut

ty, indtionalon CD(44).to theinducacrosslatedcreaseSimilathe 5′wardficatioelong(46, 4creaseanalycreasecyclintionaldetectsilencgion o

withnot riand thplex tOu

pleiotcade olowinof genfore, wing a42 geof thecreasethe inthe pring trtranscin devduceinitiatOu

genesbased

Tablepatte

A

Geneseque

Cytoskbiog

ApoptEnzymActin bCaspa

B

Geneseque

Zinc ioIntraceRegulaProPeptRN

NOTgorybyMethods).


Cance1300

Do


scriptional pausing was previously reported at the1/intron 1 boundary of c-myc in hematopoieticr cells carrying translocations involving this gene9). Here, by using high-resolution custom-designedarrays, we determined that in colon carcinoma cells,to 8 hours of butyrate treatment, there is almost a

lete decrease in c-myc transcript expression and ad downregulation of cyclin D1 expression (∼50%)a decreased steady-state level of the transcripts thatat ∼100 nucleotides downstream of the TSS (i.e.,nd ∼2% of cyclin D1 and c-myc coding sequences,tively; Fig. 1), in agreement with previous reportsranscriptional attenuation of the genes occurs closeTSS (40). The more complete decrease in the-state level of c-myc than of cyclin D1 transcriptse related to the fundamental role that alterationsyc play in colon tumor development (12, 41), ass the critical role of c-myc in reprogramming normalnal cell maturation (8). It should also be noted thaton of only a single allele of cyclin D1 was sufficient tose intestinal tumor formation in ApcMin/+ mice, sug-g that the effects of this locus may be particularlyive to more modest changes in gene expression (13).scriptional attenuation in the c-myc gene has beented to paused RNA Pol II and/or premature termi-in the promoter-proximal region (21, 39). Whereaso-2 cells, Pol II distribution increased with butyrateent at the very 5′ end of c-myc, in accord with thetion of transcriptional attenuation by butyrate in thisin SW837 cells butyrate inhibited Pol II occupancyd the 5′ end of cyclin D1 and throughout c-myc. In7 cells, butyrate-induced attenuation could be asso-with a decrease in elongating polymerase com-, as described before for other cell types (21, 42, 43).yrate, a well-characterized inhibitor of HDAC activi-uced a large increase in AcH3, a mark of transcrip-activity, which we determined to occur, for example,KN1A (not shown), confirming published data

Butyrate also induces histone hypoacetylation closeTSS in downregulated genes (45). The butyrate-

ed differences in the levels of AcH3 that we detectedthe exons of cyclin D1 and c-myc were generally re-to Pol II occupancy. However, although they de-d toward the TSS, the AcH3 changes were modest.rly, basal levels of H3me3K4 were higher nearerend of each gene, whereas H3me3K36 increased to-the 3′ end, consistent with reports that these modi-ns associate with transcriptional initiation andation, respectively, in actively transcribed genes7). However, butyrate again only modestly de-d the levels of these two modifications in both geneszed. On the other hand, there was a moderate in-in the levels of H3me3K27 by butyrate throughoutD1 and c-myc, consistent with its role in transcrip-silencing (33). Thus, although the alterations weed in these histone marks were consistent with gene
ing by butyrate, direct correspondence with the re-f attenuation was not definitive. This is consistent
alongof tra



the proposal that these histone modifications aregidly associated with the transcriptional state (26)at the functional epigenetic landscape is more com-han previously thought (48).r previous data have shown that butyrate is highlyropic in its effects on gene expression, with a cas-f changes that increases as a function of time fol-g exposure, a hallmark of sequential recruitmentes and pathways with cell maturation (7). There-e pursued these widespread effects of butyrate us-

whole transcriptome analysis, identifying a subset ofnes potentially attenuated by butyrate. Transcriptionse genes, as well as of cyclin D1 and c-myc, was de-d by NaB downstream of the TSS, suggesting thatitiation step of the transcriptional process was notincipal target, consistent with previous work show-anscriptional elongation as a rate-limiting step ofription (49). Moreover, a recent report showed thatelopmental systems, many genes that do not pro-full-length mRNAs are nonetheless transcriptionallyed (47).r transcriptome analysis revealed that whereasresponding to butyrate fell into general categorieson the location and nature of altered expression
thns
ul

2. Functional classification of transcriptrns in SW837 cells in response to butyrate

e coding sequence (Table 1; Fcription was distinct and high

Cancer Pre

y 21, 2021. © 2010 American

ig. 5), thely reproduc

vention Re

Association

Ontology, category Inces

Perce
ntage P
eleton organization and 18
.5 0.00 1 enesis osis 14 .8 0.03 8 e binding 11 .1 0.03 0 inding 11.1 0.03se activation 7.4 0.03
66

Ontology, category IInces

Percent
age
n binding, zinc finger
33.3 llular organelle 33.3 tion of transcription 22.2 transport and binding 22.2 tein
tidase activity 22.2A metabolic process 11.1

E: The biological function and gene ontology of cate-I (A) and category II (B) sequences were determined

DAVID Bioinformatics database (see Materials and

patternible for

search

for


each gscriptilel seunderThe

ity ofrate isand gprimaTherecomptriggercoordhomehibitof prodirectcell gattenuIn c

D1 an100 nmatinto butinhib

specifscriptiof genthat tacid bmech

Discl

No p

Ackn

WeCorner

Grant

AmeInstitutand Ca

Thepaymenadvertiindicate

Refe1. Ba

and2. Ha

PatumpoInt

3. Hesholon

4. Rothe

5. Aumemo

6. Hegec oJ B

7. Mainaciim200

8. MafilinGa

9. Heof

10. Tetin c

11. Saep

12. Sade

13. Hu


www.a

Do


ene, reflecting the high complexity of genome tran-on recently elucidated (50). High-throughput paral-quencing of transcripts should be informative instanding this heterogeneity along these specific loci.re are two important points regarding the complex-transcriptional regulation by butyrate. First, buty-a short-chain fatty acid, consumed as a nutrient

enerated by fermentation of dietary fiber, and thery energy source for colonic epithelial cells (4).fore, the intestine evolved in the presence of thisound, and the complex patterns of gene expressioned by butyrate may be key to generating the highlyinated response that establishes and maintains theostasis of the colon. Second, butyrate may also ex-antitumor properties, and the control of expressionto-oncogenes such as c-myc, as well as of genes thatly promote selective advantage of tumor cells (e.g.,rowth, survival, vascularization), by transcriptionalation is broader than originally thought.onclusion, transcriptional attenuation of the cyclind c-myc genes in response to butyrate begins at aboutucleotides downstream of the TSS. The role of chro-remodeling and histone modification in response
yrate is complex, may be independent of the HDAC Rece
nsom OJ, Meniel VS, Muncan V, et al. Myc deletion rescues Apcficiency in the small intestine. Nature 2007;446:676–9.lit J, Wang C, Li Z, et al. Cyclin D1 genetic heterozygosity regu-

latMi

14. Heratthe19

15. LainhBio

16. Bec-melo

17. Chacad

18. MeSethe28

19. MicaMo

20. Toregelo15

21. Kruscrin

22. Rageno

23. Wiutiven

24. Zh

acrjournals.org


ic pattern. It has been proposed that control of tran-onal elongation could be a widespread mechanisme regulation in eukaryotes (38). Our results showhis is the case for the effects of the short-chain fattyutyrate and identify gene targets regulated by thisanism.

osure of Potential Conflicts of Interest

otential conflicts of interest were disclosed.

owledgments

thank Paolo Norio, John Mariadason, Anna Velcich, and Georgiafor help in conducting these experiments.

Support

rican Institute for Cancer Research (M.C. Daroqui); National Cancere grantsU54CA100926,CA123473, andCA135561 (L.H.Augenlicht);ncer Center Support Grant P30CA13330.costs of publication of this article were defrayed in part by thet of page charges. This article must therefore be hereby markedsement in accordance with 18 U.S.C. Section 1734 solely tothis fact.

ived 04/09/2010; revised 05/28/2010; accepted 06/10/2010;
itory activity of butyrate, and may follow a gene- published OnlineFirst 09/14/2010.
rencesrnard JA, Warwick G. Butyrate rapidly induces growth inhibitiondifferentiation in HT-29 cells. Cell Growth Differ 1993;4:495–501.gue A, Manning AM, Hanlon KA, Huschtscha LI, Hart D,raskeva C. Sodium butyrate induces apoptosis in human colonicour cell lines in a p53-independent pathway: implications for the

ssible role of dietary fibre in the prevention of large-bowel cancer.J Cancer 1993;55:498–505.erdt BG, Houston MA, Augenlicht LH. Potentiation by specificrt-chain fatty acids of differentiation and apoptosis in human co-ic carcinoma cell lines. Cancer Res 1994;54:3288–93.ediger WE. Role of anaerobic bacteria in the metabolic welfare ofcolonic mucosa in man. Gut 1980;21:793–8.

genlicht LH, Anthony GM, Church TL, et al. Short-chain fatty acidtabolism, apoptosis, and Apc-initiated tumorigenesis in theuse gastrointestinal mucosa. Cancer Res 1999;59:6005–9.erdt BG, Augenlicht LH. Effects of fatty acids on expression ofnes encoding subunits of cytochrome c oxidase and cytochromexidase activity in HT29 human colonic adenocarcinoma cells.iol Chem 1991;266:19120–6.riadason JM, Corner GA, Augenlicht LH. Genetic reprogrammingpathways of colonic cell maturation induced by short chain fattyds: comparison with trichostatin A, sulindac, and curcumin andplications for chemoprevention of colon cancer. Cancer Res0;60:4561–72.riadason JM, Nicholas C, L'Italien KE, et al. Gene expression pro-g of intestinal epithelial cell maturation along the crypt-villus axis.stroenterology 2005;128:1081–8.TC, Sparks AB, Rago C, et al. Identification of c-MYC as a targetthe APC pathway. Science 1998;281:1509–12.su O, McCormick F. β-Catenin regulates expression of cyclin D1olon carcinoma cells. Nature 1999;398:422–6.ncho E, Batlle E, Clevers H. Live and let die in the intestinalithelium. Curr Opin Cell Biol 2003;15:763–70.

es colonic epithelial cell differentiation and tumor number in Apc-n mice. Mol Cell Biol 2004;24:7598–611.ruth DP, Zirnstein GW, Bradley JF, Rothberg PG. Sodium buty-e causes an increase in the block to transcriptional elongation inc-myc gene in SW837 rectal carcinoma cells. J Biol Chem

93;268:20466–72.llemand F, Courilleau D, Sabbah M, Redeuilh G, Mester J. Directibition of the expression of cyclin D1 gene by sodium butyrate.chem Biophys Res Commun 1996;229:163–9.nder TP, Thompson CB, Kuehl WM. Differential expression ofyb mRNA in murine B lymphomas by a block to transcriptionngation. Science 1987;237:1473–6.en Z, Harless ML, Wright DA, Kellems RE. Identification and char-terization of transcriptional arrest sites in exon 1 of the humanenosine deaminase gene. Mol Cell Biol 1990;10:4555–64.chti N, Piechaczyk M, Blanchard JM, Jeanteur P, Lebleu B.quence requirements for premature transcription arrest withinfirst intron of the mouse c-fos gene. Mol Cell Biol 1991;11:

32–41.ddleton KM, Morgan GT. Premature termination of transcriptionn be induced on an injected α-tubulin gene in Xenopus oocytes.l Cell Biol 1990;10:727–35.ng X, Yin L, Joshi S, Rosenberg DW, Giardina C. Cyclooxygenase-2ulation in colon cancer cells: modulation of RNA polymerase IIngation by histone deacetylase inhibitors. J Biol Chem 2005;280:503–9.mm A, Meulia T, Brunvand M, Groudine M. The block to tran-iptional elongation within the human c-myc gene is determinedthe promoter-proximal region. Genes Dev 1992;6:2201–13.schke EE, Albert T, Eick D. Transcriptional regulation of the Igκne by promoter-proximal pausing of RNA polymerase II. J Immu-l 1999;163:4375–82.lson AJ, Velcich A, Arango D, et al. Novel detection and differential
lization of a c-myc transcriptional block in colon cancer chemopre-tion. Cancer Res 2002;62:6006–10.ang ZK, Davies KP, Allen J, et al. Cell cycle arrest and repression



of597

25. Maducin

26. Li Btion

27. Bocet253

28. Tu200

29. MaDivCation

30. DaKiegreRe

31. Semo

32. Bahis

33. ValysCe

34. Cama38:

35. Priby

36. Tanandtio256

37. Kim200

38. Beter53

39. KearrN-

40. WrofNa

41. Ignint20

42. BeforNa

43. Netio

44. Hinfecno13

45. Racrean70

46. BaC,his

47. Gumama

48. Betra

49. Saelo


Cance1302

Do


cyclin D1 transcription by INI1/hSNF5. Mol Cell Biol 2002;22:5–88.ier S, Daroqui MC, Scherer S, et al. Butyrate and vitamin D3 in-ce transcriptional attenuation at the cyclin D1 locus in colonic car-oma cells. J Cell Physiol 2009;218:638–42., Carey M, Workman JL. The role of chromatin during transcrip-. Cell 2007;128:707–19.ffa LC, Vidali G, Mann RS, Allfrey VG. Suppression of histone dea-ylation in vivo and in vitro by sodium butyrate. J Biol Chem 1978;:3364–6.

rner BM. Histone acetylation and an epigenetic code. Bioessays0;22:836–45.riadason JM, Rickard KL, Barkla DH, Augenlicht LH, Gibson PR.ergent phenotypic patterns and commitment to apoptosis ofco-2 cells during spontaneous and butyrate-induced differentia-. J Cell Physiol 2000;183:347–54.roqui MC, Puricelli LI, Urtreger AJ, Elizalde PV, Lanuza GM, Bal der Joffe E. Involvement of TGF-β(s)/T(β)Rs system in tumor pro-ssion of murine mammary adenocarcinomas. Breast Cancers Treat 2003;80:287–301.aly L, Chalkley R. The effect of sodium butyrate on histonedification. Cell 1978;14:115–21.rski A, Cuddapah S, Cui K, et al. High-resolution profiling oftone methylations in the human genome. Cell 2007;129:823–37.koc CR, Sachdeva MM, Wang H, Blobel GA. Profile of histoneine methylation across transcribed mammalian chromatin. Molll Biol 2006;26:9185–95.rninci P, Sandelin A, Lenhard B, et al. Genome-wide analysis ofmmalian promoter architecture and evolution. Nat Genet 2006;626–35.ce DH. P-TEFb, a cyclin-dependent kinase controlling elongationRNA polymerase II. Mol Cell Biol 2000;20:2629–34.S, Aso T, Conaway RC, Conaway JW. Roles for both the RAP30RAP74 subunits of transcription factor IIF in transcription initia-

n and elongation by RNA polymerase II. J Biol Chem 1994;269:
84–91.YS, Milner JA. Dietary modulation of colon cancer risk. J Nutr7;137:2576–9S.
50. BiranEN



ntley DL, Groudine M. Sequence requirements for prematuremination of transcription in the human c-myc gene. Cell 1988;:245–56.ene RG, Mueller A, Landick R, London L. Transcriptional pause,est and termination sites for RNA polymerase II in mammalianand c-myc genes. Nucleic Acids Res 1999;27:3173–82.ight S, Mirels LF, Calayag MC, Bishop JM. Premature terminationtranscription from the P1 promoter of the mouse c-myc gene. Proctl Acad Sci U S A 1991;88:11383–7.atenko NA, Holubec H, Besselsen DG, et al. Role of c-Myc inestinal tumorigenesis of the ApcMin/+ mouse. Cancer Biol Ther06;5:1658–64.ntley DL, Groudine M. A block to elongation is largely responsibledecreased transcription of c-myc in differentiated HL60 cells.ture 1986;321:702–6.pveu A, Marcu KB. Intragenic pausing and anti-sense transcrip-n within the murine c-myc locus. EMBO J 1986;5:2859–65.nebusch BF, Meng S, Wu JT, Archer SY, Hodin RA. The ef-ts of short-chain fatty acids on human colon cancer cell phe-type are associated with histone hyperacetylation. J Nutr 2002;2:1012–7.da-Iglesias A, Enroth S, Ameur A, et al. Butyrate mediates de-ase of histone acetylation centered on transcription start sitesd down-regulation of associated genes. Genome Res 2007;17:8–19.nnister AJ, Schneider R, Myers FA, Thorne AW, Crane-RobinsonKouzarides T. Spatial distribution of di- and tri-methyl lysine 36 oftone H3 at active genes. J Biol Chem 2005;280:17732–6.enther MG, Levine SS, Boyer LA, Jaenisch R, Young RA. A chro-tin landmark and transcription initiation at most promoters in hu-n cells. Cell 2007;130:77–88.rger SL. The complex language of chromatin regulation duringnscription. Nature 2007;447:407–12.unders A, Core LJ, Lis JT. Breaking barriers to transcriptionngation. Nat Rev Mol Cell Biol 2006;7:557–67.ney E, Stamatoyannopoulos JA, Dutta A, et al. Identification and
alysis of functional elements in 1% of the human genome by theCODE pilot project. Nature 2007;447:799–816.



2010;3:1292-1302. Published OnlineFirst September 14, 2010.Cancer Prev Res Maria C. Daroqui and Leonard H. Augenlicht Response to ButyrateTranscriptional Attenuation in Colon Carcinoma Cells in

Updated version

10.1158/1940-6207.CAPR-10-0083doi:

Access the most recent version of this article at:

Material

Supplementary

0083.DC1

http://cancerpreventionresearch.aacrjournals.org/content/suppl/2011/04/14/1940-6207.CAPR-10-Access the most recent supplemental material at:

Cited articles

http://cancerpreventionresearch.aacrjournals.org/content/3/10/1292.full#ref-list-1

This article cites 50 articles, 26 of which you can access for free at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerpreventionresearch.aacrjournals.org/content/3/10/1292To request permission to re-use all or part of this article, use this link

Cancer Research. on July 21, 2021. © 2010 American Association forcancerpreventionresearch.aacrjournals.org Downloaded from


http://cancerpreventionresearch.aacrjournals.org/lookup/doi/10.1158/1940-6207.CAPR-10-0083

http://cancerpreventionresearch.aacrjournals.org/content/suppl/2011/04/14/1940-6207.CAPR-10-0083.DC1

http://cancerpreventionresearch.aacrjournals.org/content/suppl/2011/04/14/1940-6207.CAPR-10-0083.DC1

http://cancerpreventionresearch.aacrjournals.org/content/3/10/1292.full#ref-list-1

http://cancerpreventionresearch.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://cancerpreventionresearch.aacrjournals.org/content/3/10/1292


Research Article Cancer Prevention Research Transcriptional … · Transcriptional Attenuation in...

Documents

Transcript of Research Article Cancer Prevention Research Transcriptional … · Transcriptional Attenuation in...