GENES, CHROMOSOMES & CANCER 00:000–000 (2012)

Aberrant Promoter Methylation of Beta-1,4Galactosyltransferase 1 as Potential Cancer-SpecificBiomarker of Colorectal Tumors

Maria Luana Poeta,1*† Emanuela Massi,2† Paola Parrella,3 Pasquale Pellegrini,2,4 Mariangela De Robertis,2

Massimiliano Copetti,3 Carla Rabitti,5 Giuseppe Perrone,5 Andrea Onetti Muda,5 Francesca Molinari,6

Elena Zanellato,6 Stefano Crippa,6 Damiano Caputo,7 Marco Caricato,7 Milo Frattini,6 Roberto Coppola,7

and Vito Michele Fazio2,3

1Departmentof Bioscience,Biotechnologyand Pharmacological Sciences,Universityof Bari,Italy2Laboratory for Molecular Medicine and Biotechnology,CIR,University Campus Bio-Medico of Rome,Italy3Laboratoryof Oncology,IRCCSCasa Sollievo della Sofferenza,San Giovanni Rotondo,FG,Italy4Cancer Epigenetics and Biology Program (PEBC),Bellvitge Institute for Biomedical Research (IDIBELL),Barcelona,Spain5Departmentof Pathology,University Campus Bio-Medico of Rome,Italy6Institute of Pathology,Locarno,Switzerland7Departmentof Surgery,University Campus Bio-Medico of Rome,Italy

Epigenetic alterations, such as CpG islands methylation and histone modifications, are recognized key characteristics of

cancer. Glycogenes are a group of genes which epigenetic status was found to be changed in several tumors. In this study,

we determined promoter methylation status of the glycogene beta-1,4-galactosyltransferase 1 (B4GALT1) in colorectal can-

cer patients. Methylation status of B4GALT1 was assessed in 130 colorectal adenocarcinomas, 13 adenomas, and in paired

normal tissue using quantitative methylation specific PCR (QMSP). B4GALT1 mRNA expression was evaluated in methyl-

ated/unmethylated tumor and normal specimens. We also investigated microsatellite stability and microsatellite instability

status and KRAS/BRAF mutations. Discriminatory power of QMSP was assessed by receiving operating curve (ROC) analysis

on a training set of 24 colorectal cancers and paired mucosa. The area under the ROC curve (AUC) was 0.737 (95%

confidence interval [CI]:0.591–0.881, P ¼ 0.005) with an optimal cutoff value of 2.07 yielding a 54% sensitivity (95%

CI: 35.1%–72.1%) and a specificity of 91.7% (95% CI: 74.1%–97.7%). These results were confirmed in an independent vali-

dation set where B4GALT1 methylation was detected in 52/106 patients. An inverse correlation was observed between

methylation and B4GALT1 mRNA expression levels (r ¼ �0.482, P ¼ 0.037). Significant differences in methylation levels

and frequencies was demonstrated in invasive lesions as compared with normal mucosa (P ¼ 0.0001) and in carcinoma

samples as compared with adenoma (P ¼ 0.009). B4GALT1 methylation is a frequent and specific event in colorectal cancer

and correlates with downregulation of mRNA expression. These results suggest that the glycogene B4GALT1 represent a

valuable candidate biomarker of invasive phenotype of colorectal cancer. VVC 2012 Wiley Periodicals, Inc.

INTRODUCTION

Colorectal cancer (CRC) represents the third

most commonly diagnosed malignancy in males

and the second in females world-wide, with over

1.2 million of new cases and over 600 thousand

deaths estimated to have occurred in 2008 (Jemal

et al., 2010; 2011). An effective screening for pre-

malignant adenomas and early stage cancers

would have substantial clinical benefits, and

would reduce the mortality of patients with CRC

(Levin et al., 2008; Rennert et al., 2009; Smith

et al., 2011). In recent time, the United States is

the only country where a significant decrease in

incidence rates was recorded in both males and

females, reflecting the relevant impact of early

detection and removal of precancerous lesions

through CRC screening programs (Whitlock

et al., 2008; Center et al., 2009; Edwards et al.,

2010). Despite significant improvements in

screening methods for CRC, only 30%–40% of

patients are diagnosed at an early stage (Smith

Additional Supporting Information may be found in the onlineversion of this article.

Supported by: AIRC MFAG 10520; Progetto ITINERIS 2FILAS Regione Lazio ‘‘articolo 180 comma 4 lettera C) L.R.04/06-Azioni Verticali’’. E. Massi is recipient of a post-doctoral fel-lowship from University Campus Bio-Medico of Rome (BandoAteneo 2009), Center for Integrated Research.

yThese authors contributed equally to this work.

*Correspondence to: Maria Luana Poeta, Department of Bio-science, Biotechnology and Pharmacological Sciences, Universityof Bari, Via Amendola 165/A, 70126, Bari, Italy.E-mail: poetaluana@gmail.com

Received 13 January 2012; Accepted 25 July 2012

DOI 10.1002/gcc.21998

Published online inWiley Online Library (wileyonlinelibrary.com).

RESEARCH ARTICLE

VVC 2012 Wiley Periodicals, Inc.

et al., 2006). Several molecular markers have

been proposed for diagnostic, prognostic, and sur-

veillance purpose in CRC, but their clinical util-

ity remain controversial, due to the lack of

sensitivity and specificity. The first step of a bio-

marker discovery pipeline begins with preclinical

studies and requires a comparison between tumor

tissues and non-tumor tissues. A panel of differ-

ent genes can help reach higher sensitivity levels,

but a more critical issue in CRC screening is

specificity. Specificity is a very high priority for

screening markers as recommended by Early

Detection Research Network, given that even a

small false-positive rate translates itself into a

large number of people subjected to unnecessary

costly diagnostic procedures (Pepe et al., 2001).

CRC arises as a multistep process involving pro-

gressive accumulation of genetic (Fearon and

Vogelstein, 1990) and epigenetic alterations

(Grady and Markowitz, 2002; Herceg and Hainaut,

2007), which could be used as markers for CRC

(Verma and Manne, 2006; Corvalan and Maturana,

2010). Aberrant methylation of CpG islands at the

promoter region of a gene is an epigenetic change

that induces the transcriptional silencing of tumor

suppressor genes (TSGs) (Herman and Baylin,

2003; Ballestar and Esteller, 2008). Hypermethyl-

ation of CpG islands in TSGs has been reported

for several human cancers, including CRCs (Jones

and Baylin, 2007; Rodrıguez-Paredes and Esteller,

2011; van Engeland et al., 2011). The epigenetic

progenitor stem cell model of carcinogenesis

(Feinberg et al., 2006; Mathews et al., 2009;

Yamada and Watanabe, 2010; Tsai and Baylin,

2011) and the specificity of cancer-related epige-

netic modifications strongly support these new mo-

lecular markers for early diagnosis (Hoque et al.,

2006a, b; Schmid, 2010), prognosis (Herranz and

Esteller, 2007), follow-up and therapy response

(Parrella et al., 2009; Sharma et al., 2009). Although

hypermethylated TSGs have been proposed as in-

triguing biomarkers for CRC, only few genes har-

boring high level of specificity have been

identified (Lind et al., 2011).

Previous studies included a genome-wide

methylation analysis using a new promoter struc-

ture algorithm and microarray data sets generated

from 20 cancer cell lines of five different major

cancers treated with demethylating agent 5-aza-

2’-deoxycytidine (Hoque et al., 2008). The aim of

the present study was to identify novel cancer

specific methylated genes. Preliminary data

obtained using bisulfite DNA sequencing and

MSP (methylation specific PCR) in colon cancer

cell lines and tissues samples allowed us to

identify novel cancer-specific methylated genes,

thus showing a methylation frequency ranging

from 60% to 100%. Among cancer-specific meth-

ylated genes in CRC we found that the promoter

of the glycogene beta-1,4-galactosyltransferase 1

(B4GALT1) was hypermethylated in colon cancer

cells and in an initial set of tumor primary tissues

(Hoque et al., 2008, Kim et al., 2009). Despite

the fact that DNA methylation, histone modifica-

tion, chromatin remodeling, and noncoding RNA

have been identified as fundamental elements

controlling gene expression, only a few studies

have investigated the epigenetic mechanisms

underlying aberrant expression of glycan genes in

cancer cells. In this study, we validated the pro-

moter methylation status of B4GALT1 in an inde-

pendent and large series of 130 CRCs and paired

histopathologically normal colonic mucosa. We

found that B4GALT1 aberrant promoter methyla-

tion is associated with reduced mRNA expression

in tumor tissues as compared with normal tissues

and unmethylated cancer cases. We further inves-

tigated the relationship between the methylation

status of B4GALT1 and KRAS/BRAF mutation

status in microsatellite stable and microsatellite

instable (MSI) CRC. Synchronous colon adeno-

mas have been analyzed. Although they harbored

some degree of methylation at B4GALT1 pro-

moter, only invasive lesions showed a statistically

significant difference in methylation levels and

frequencies, compared with normal colic mucosa.

MATERIALS AND METHODS

CRC Patient Samples

A total of 130 CRC and paired normal samples

were retrospectively collected from the archive of

the department of pathology of two hospitals

(University of Campus Bio-Medico of Rome,

Italy N ¼ 52 cases; Institute of Pathology,

Locarno Switzerland N ¼ 78 cases) and stored in

accordance with institutional guidelines.

Fifty-two cases were immediately frozen after

surgery and kept at �80�C until nucleic acid

extraction. Seventy-eight cases consisting of tu-

mor and paired normal mucosa were formalin-

fixed paraffin embedded specimens (FFPE). Syn-

chronous adenocarcinoma, adenoma, and normal

colonic mucosa were obtained from 13 cases

(Institute of Pathology, Locarno). Twenty-four

cases have been used as training set and 106

specimens represented a validation set.

2 POETA ETAL.

Genes, Chromosomes & Cancer DOI 10.1002/gcc

Tumor samples with at least 70% cancer cells

were eligible for molecular studies. Tumors with

low neoplastic cellularity (<70%) were microdis-

sected to enrich the malignant cells.

Pathological data of 130 patients included pT,

pN, and pM classification, grading, histologic sub-

type, tumor location, and stage. Clinical data,

including gender, age, and follow-up times were

retrieved for the 52 Italian cohort of patients,

with a median follow-up time of 66 months

(range 1–93 months). All human materials used in

the study were collected according to the guide-

lines of the Local Ethical Committee. Prior writ-

ten and informed consent was obtained from each

patient accordingly with institutional guidelines.

DNA and RNA Extraction

Colon tumor and normal mucosa samples were

rapidly frozen at –80�C (52/130) or paraffin-em-

bedded (78/130). A series of 5-micron sections

were cut from the primary tumor specimens for

hematoxylin-eosin staining to inspect the presence

of neoplastic cells and to ensure absence of neo-

plastic cells in normal mucosa specimens. From

frozen tissues, additional 12 micron sections were

cut and placed in 1% SDS and proteinase K at

48�C overnight followed by DNA extraction with

phenol/chloroform and precipitated with ethanol.

From FFPE specimens, 10 micron sections were

used for DNA extraction using the QIAamp Mini

kit (Qiagen, Chatsworth, CA) according to the

manufacturer’s instructions. Total RNA was also

available for a subset (n ¼ 19) of the Italian CRC

patients. Total RNA was extracted using the Invi-

trogen TRIzol reagent (Carlsbad, CA). DNA and

RNA concentrations were quantified by the ab-

sorbance measurement at 260 and 280 nm using

the Nanodrop spectrophotometer.

Bisulfite Treatment and Quantitative

Methylation Specific PCR

Two lg of DNA extracted from tumor, ade-

noma, and normal colon mucosa was subjected to

bisulfite treatment and DNA purification using

the Epitect Bisulfite kit (Qiagen Sci, MD)

according to manufacturer’s instructions.

Bisulfite-modified DNA was used as template

for fluorescence-based real-time quantitative meth-

ylation specific PCR (QMSP) as previously

described (Hoque et al., 2006b). Amplification

reactions were carried out in triplicate in a volume

of 20 lL that contained 90 ng of bisulfite-modified

DNA, 600 nM forward and reverse primers, 200 nMprobe, 0.6 U of Platinum Taq polymerase

(Invitrogen, Inc., Rockville, MD), 200 lM each of

dATP, dCTP, dGTP, and dTTP, and 5.5 mMMgCl2. Primers and probe of the gene of interest

(B4GALT1 Forward 50-TAGGAAACGGGTTTCG

ACG-30, Reverse 50-CCGTCCACTTTCTTTA

CCG-30, probe 50-CGTTAAACAACGAAATCCA

ACCGAA-30) were designed to amplify specifically

the bisulfite modified region containing the puta-

tive methylated CpGs, whereas primers and probe

for the reference gene (ACTB Forward 50-TGG

TGATGGAGGAGGTTTAGTAAGT-30, Reverse50-AACCAATAAAACCTACTCCTCCCTTAA-30,probe 50-ACCACCACCCAACACACAATAACAA

ACACA-30) were designed to amplify specifically a

bisulfite modified region not containing CpGs. The

genomic region of B4GALT1 where CpG have been

analysed for methylation status is shown in the Fig-

ure 1, Supporting Information. Amplifications were

carried out using the following conditions: one step

at 95�C for 2 min, 40 cycles at 95�C for 15 sec, and

56�C for 1 min. Amplification reactions were carried

out in 384-well plates in a 7900HT Fast Real Time

PCR Systems (Applied Biosystems, Carlsbad, CA)

and were analyzed by sequence detection system

2.3 software (Applied Biosystems, Foster City, CA).

Serial dilutions (90–0.009 ng) of CpGenomeTM

Universal Methylated DNA (Serologicals Corp.,

Norcross, GA) were used to construct a calibration

curve for both target and reference gene. Each

plate included calibration curves for the ACTB and

B4GALT1 genes, patient DNA samples, positive

Figure 1. Box plot for B4GALT1/ACT ratios determined by QMSPin tumor samples (n ¼ 24) and normal paired colonic tissue (n ¼24). The boxes mark the IQR (interval between the 25th and 75thpercentile). The lines inside the boxes denote median values. Thewhiskers represent the interval between the 10th and 90th percen-tiles (Mann-Whitney U test P ¼ 0.001). The asterisks (*) indicate out-liers. [Color figure can be viewed in the online issue, which isavailable at wileyonlinelibrary.com.]

B4GALT1 PROMOTER HYPERMETHYLATION IN COLORECTAL CANCER 3

Genes, Chromosomes & Cancer DOI 10.1002/gcc

control (CpGenomeTM Universal Methylated

DNA, Serologicals Corp., Norcross, GA) and multi-

ple water blanks. For each sample, the analysis

was repeated in two separate plates. The number

of cycles at which the fluorescence signal crossed a

detection threshold was determined automatically

with the ABI PRISM 7900HT Sequence Detec-

tion System Software, and referred to as CT (cycle

threshold). CTs average value of triplicates for all

samples were converted to a relative ‘‘quantity’’

scale using a standard curve of serially diluted

methylated DNA. The level of methylated DNA

in each sample was determined as a ratio of

B4GALT1 quantity to ACTB quantity and then

multiplied by 1,000 for easier tabulation

(B4GALT1 Qty/ACTB Qty � 1,000) (Ostrow

et al., 2010; Durr et al., 2010). The samples were

categorized as unmethylated or methylated based

on optimal cutoff derived from receiving operating

curve (ROC) analysis.

Real Time RT-PCR

One lg of RNA was reverse-transcribed using

High Capacity cDNA Reverse Transcription kit

(Applied Biosystems) and RNasin Plus RNAse In-

hibitor (Promega Corporation). Quantitative fluo-

rescence-based reverse transcription PCR (RT-

qPCR) was set up in 384-well plates in a total vol-

ume of 20 ll containing 2 ll cDNA (100 ng), 1 llB4GALT1 Taqman

VR

Gene Expression Assays

(Assay ID: Hs00155245_m1 Applied Biosystems),

and 10 ll of TaqmanVR

Gene Expression Master

Mix (Applied Biosystems). Reactions were run on

7900HT Fast Real-Time PCR System (Applied

Biosystems) and analyzed by the sequence detec-

tion system 2.3 software (Applied Biosystems).

The target gene transcription levels were normal-

ized using the Gene Expression Assay for Human

Large Ribosomal Protein, RPLPO (Assay ID:

4326314E, Applied Biosystems) as the reference

gene. A relative quantification method with stand-

ard curve was developed, mRNA levels in each

sample were determined as the ratio of the

B4GALT1 expression level to the RPLPO expres-

sion and then multiplied by 1,000 for easier tabu-

lation (B4GALT1/RPLPO � 1,000).

Real Time-PCR was performed using the fol-

lowing thermal conditions: one step at 95�C for

10 min, 40 cycles at 95�C for 15 sec and 60�C for

1 min. Relative expression in tumor samples as

compared with paired normal colon tissue was

calculated according to the following formula:

2�DDCt (see ABI PRISM 7900 Sequence Detec-

tion System User Bulletin #2). Relative expres-

sion ratios (RERs) were determined as the level

of mRNA expression in the tumor sample to the

expression in the paired normal tissue.

Immunohistochemistry

Immunohistochemistry (IHC) experiments

were performed on 10 surgical pathology speci-

mens used for diagnosis. Samples were fixed in

4% neutral buffered formaldehyde. Representa-

tive tumor blocks were sectioned at 3 lm thick-

ness. IHC was performed by the streptavidin–

biotin method. Endogenous peroxidase in the

section was blocked by incubation with 3%

hydrogen peroxide. A rabbit polyclonal antibody

against the human B4GALT1 protein

(HPA010806, Sigma-Aldrich) was used as primary

antibody at a 1/10 dilution. Sections were incu-

bated with LSAB2 (Dakocytomation, Carpinteria,

CA). 3-30-diaminobenzidine was used for color

development and hematoxylin was used for coun-

terstaining. Negative controls were obtained by

omitting primary antibody. Scoring for B4GALT1

was based on relative intensities of staining of

tumor cells with reference to the normally pres-

ent B4GALT1 staining of tissue-associated mac-

rophages. The staining intensity was scored on a

scale from 0 to 3þ (0, negative; 1þ, weak; 2þ,

moderate; 3þ, strong) in the tumor cells in 10

high power fields. Immunostaining was assessed

by two blinded and independent pathologists and

differences in scoring were resolved by a concur-

rent review and consensus score.

KRAS/BRAF Mutation and MSI Analysis

Genomic DNA obtained from tissue specimens

was amplified by PCR for the clinically relevant

KRAS codons 12/13 (exon 2) and BRAF codon

600 (exon 15) mutations as already published

(Molinari et al., 2011).

Primer sequences for KRAS and BRAF were:

KRAS (F 50-TGGTGGAGTATTTGATAGTGT

A-30; R 50-CATGAAAATGGTCAGAGAA-30) andBRAF (F 50-TCATAATGCTTGCTCTGATAG

GA-30; R 50-GGCCAAAAATTTAATCAGTGG

A-30). PCR conditions used were: 95�C 10 min,

95�C 15 sec, 55�C for KRAS and 52�C for BRAF

30 sec, 72�C 30 sec, 72�C 3 min, 40 cycles. Sam-

ples were then subjected to automated sequenc-

ing of single stranded PCR products using the

ABI Prism 3130 Genetic Analyzer (Applied Bio-

systems). All mutated cases were confirmed twice

with independent PCR reactions.

4 POETA ETAL.

Genes, Chromosomes & Cancer DOI 10.1002/gcc

Analysis of MSI status was conducted on multi-

plex amplification of the 5 microsatellite Be-

thesda-panel (BAT25, BAT26, D2S123, D5S346,and D17S250), as previously reported (Frattini

et al., 2004). Primer sequences for MSI analysis

were: BAT25 (F 50-TCGCCTCCAAGAATGTA

AGT-30; R 50-TCTGCATTTTAACTATGGCT

C-30); BAT26 (F 50-TGACTACTTTTGACTTC

AGCC-30; R 50-AACCATTCAACATTTTTAAC

CC-30), D2S123 (F 50-AAACAGGATGCCTGCC

TTTA-30; R 50-GGACTTTCCACCTATGGGA

C-30), D5S346 (F 50-ACTCACTCTAGTGATAA

ATCGGG-30; R 50-AGCAGATAAGACAGTATT

ACTAGTT-30), D17S250 (F 50-GGAAGAATCAA

ATAGACAAT-30; R 50-GCTGGCCATATATAT

ATTTAAACC-30). Forward primers were labeled

with 6-FAM at the 50-end. MSI was confirmed by

the presence of an additional peak in tumor sam-

ple in comparison with normal paired tissue.

Statistical Analysis

Patients’ baseline characteristics were reported

as median and interquartile range (IQR). The

discriminatory power of B4GALT1 was assessed

by estimating the area under the receiver operat-

ing characteristics (ROC) curves, using methyla-

tion levels in normal colon tissues and tumor

samples. The optimal cutoff was assessed maxi-

mizing jointly sensitivity and specificity. Sensitiv-

ity and specificity, computed at the optimal

cutoff, were reported along with their 95% confi-

dence intervals (CI). The area under the ROC

curve (AUC) was also reported along its 95% CI.

Colorectal samples were classified as methylated

(M) or unmethylated based on whether B4GALT1to ACTB ratios were above or below cutoff values,

respectively. Correlation between B4GALT1/ACTB ratios and mRNA transcript levels were

assessed using the Pearson’s correlation coeffi-

cient (r). Time to progression was defined as the

time between surgery and the first progression

event. Overall survival time was defined as the

time between surgery and death.

Associations between B4GALT1 methylation

status and clinicopathological parameters were

assessed using the Pearson Chi-Squared test and

Mann-Whitney U test for categorical and continu-

ous variables, respectively.

Time to event analyses were performed using

the Kaplan–Meier method with a log-rank test

and the proportional hazards Cox regression mod-

els. A P value < 0.05 was considered statistically

significant. All analyses were performed using

SAS v9.1 and SPSS 17.0.

RESULTS

Tumor Specific Hypermethylation of the B4GALT1

Promoter in CRC Patients

Using the QMSP assay the methylation status

of the B4GALT1 promoter was assessed in an ini-

tial training set of 24 CRC patients, obtained

from the Department of Pathology, University

Campus Bio-Medico of Rome, Italy. For each

sample, the analysis was repeated in two separate

plates and the median value of the B4GALT1/ACTB ratio of the two plates was used for statisti-

cal analyses.

The median values and IQR of B4GALT1/ACTB ratios were 0 (IQR 0–0) for normal colon

mucosa and 3.50 (IQR 0–50.75) for tumor speci-

mens (P ¼ 0.001 Mann Whitney test) (Fig. 1).

No differences in B4GALT1/ACTB ratios were

demonstrated among paraffin embedded samples

and snap frozen specimens.

The B4GALT1/ACTB ratios in the normal colon

tissues and in tumor samples were used to draw a

ROC curve (Fig. 2) in order to assess the discrim-

inatory power of the B4GALT1 QMSP assay. The

AUC value was 0.737 (95% CI: 0.591–0.881, P ¼0.005) (Fig. 2). Based on the ROC curve, an opti-

mal cut off value of 2.07 was assessed maximizing

sensitivity and specificity yielding a 54% (95%

CI: 35.1%–72.1%) sensitivity and a 91.7% (95%

CI: 74.1%–97.7%) specificity. Methylation of

B4GALT1 showed a highly discriminative ROC

Figure 2. ROC Curve of B4GALT1: ROC curve for QMSP assaywas designed on the basis of B4GALT1/ACTB ratios in a training set ofcolorectal tumor samples (n ¼ 24) and normal paired colon samples(n ¼ 24). The AUC is 0.736 (95% CI: 0.591–0.881; P ¼ 0.005). Opti-mal cut-off value (2.07) was considered for methylation analysis in val-idation set (n ¼ 106). [Color figure can be viewed in the onlineissue, which is available at wileyonlinelibrary.com.]

B4GALT1 PROMOTER HYPERMETHYLATION IN COLORECTAL CANCER 5

Genes, Chromosomes & Cancer DOI 10.1002/gcc

curve profile, clearly distinguishing CRC from

corresponding normal distant mucosa.

In a second independent validation set of 106

tumors, including 28 Italian cases and a Swiss

CRC cohort of 78 patients obtained from Insti-

tute of Pathology, Locarno (Switzerland),

B4GALT1 promoter methylation was detected in

49% (52/106) of patients.

Baseline and pathological characteristics of the

Italian and Swiss populations (130 cases) are

shown in Table 1. Clinical data were available for

52 Italian cases (24 training set and 28 validation

set). No statistically significant association was

found between B4GALT1 methylation status and

clinicopathological data.

Correlation of B4GALT1 Promoter Methylation

and B4GALT1 mRNA and Protein Expression in

CRC Patients

To elucidate whether the hypermethylation of

the B4GALT1 promoter impacts the expression of

the gene we analyzed the mRNA expression lev-

els in 19 representative clinical samples, including

primary CRC specimens and paired normal colo-

rectal mucosa (9 methylated and 10 unmethylated

samples). A statistically significant inverse correla-

tion was observed between methylation levels and

B4GALT1 mRNA expression levels (r ¼ �0.482,

P ¼ 0.037, Pearson correlation). For each patient,

we calculated the RERs as the ratio between

mRNA in the tumor sample to the expression in

the paired normal tissue. The B4GALT1 RER val-

ues ranged from 0.06 to 4.76 with a median value

of 0.65 (IQR 0.41–1.54). Moreover, as shown in

Figure 3, methylated tumors showed lower RERs

(Median 0.43 IQR 0.29–1.17) as compared with

unmethylated samples (Median 0.92 IQR 0.57–

2.93) (P ¼ 0.015 Mann Whitney test), further

suggesting that mRNA B4GALT1 expression is

epigenetically regulated.

In order to assess the B4GALT1 protein

expression level, we performed immunohisto-

chemistry analysis on 10 tumor cases and 10 nor-

mal matched mucosa specimens. We divided

samples into methylated and unmethylated

groups (5 vs. 5). B4GALT1 immunohistochemis-

try of colorectal tissue revealed a membrane/cyto-

plasmic staining in tumor and normal epithelial

cells. A weak-moderate staining for B4GALT1

was observed in 80% (4/5) of methylated cases

and in 40% (2/5) of unmethylated cases, but

the difference was not statistically significant

(P >0.05 Fisher Exact Test). Representative

images of immunohistochemical staining are

shown in Figure 4.

B4GALT1 Promoter Hypermethylation Marks the

Onset of Invasive Phenotype in CRC

In order to establish the timing of B4GALT1promoter methylation in colorectal carcinogenesis,

we also analyzed 13 CRC for which synchronous

tubulovillous adenoma lesions were available along

with normal colonic mucosa. The median values

and inter quartile ranges (IQR) of B4GALT1/ACTBratios were 37.55 (IQR 0–292.19) for CRC and 0.0

for colorectal adenoma (IQR 0–9.14) and normal

colic mucosa (IQR 0–0) (P ¼ 0.0001 Kruskall

Wallis Test) (Table 2). Between groups analysis

demonstrated statistically significant differences in

B4GALT1 promoter methylation levels and fre-

quencies in invasive carcinomas as compared with

normal mucosa samples (P ¼ 0.0001 Mann Whit-

ney test; P ¼ 0.0001 Chi-Squared test) and ade-

noma (P ¼ 0.009 Mann Whitney test; P ¼ 0.04

Chi-Squared test), whereas no statistically signifi-

cant differences were demonstrated between nor-

mal tissue and adenoma samples (P ¼ 0.07 Mann

Whitney test; P ¼ 0.22 Chi-Squared test).

B4GALT1 Promoter Methylation and KRAS,

BRAF Mutation, and MSI Status

Key molecular events in CRC are mutations

in the KRAS and BRAF genes as well as MSI

(Table 1). We examined KRAS (exon 2, including

codons 12 and 13) mutational status in 128/130

CRC cases. Mutations of KRAS occurred in 50/

128 (39%) of cases. Among the observed muta-

tions, 37/50 (74%) were found in codon 12 and

the remaining 13/50 (26%) in codon 13.

BRAF (exon 15, including codon 600) mutation

and MSI status were investigated in 77 of 130

patients. The BRAFV600E mutation was

observed in 5/77 (6.5%), whereas MSI was

detected in 7/77 (9%) cases. No statistically sig-

nificant relationship was found between muta-

tions in KRAS or BRAF, or MSI status and

methylation of B4GALT1 promoter.

DISCUSSION

Glycoproteins like carcinoembryonic antigen

(CEA) and glycosylated mucins like CA 19-9 are

biomarkers commonly used in the clinical setting

of CRC. However, the clinical usefulness of these

glycomarkers for diagnostic, prognostic, and sur-

veillance purpose remain controversial (Locker

6 POETA ETAL.

Genes, Chromosomes & Cancer DOI 10.1002/gcc

TABLE 1. Demographic and Clinicopathological Characteristics of Colorectal Cancer Cases (130)

B1,4GALTI methylation status

Methylated Unmethylated

Variable Patients no. No. (%) P value

Total patients 130 64/130 (49.2) 66/130 (50.7)

Gender 130 0.0783Male 67 38/67 (56.7) 29/67 (43.3)Female 63 26/63 (41.3) 37/63 (58.7)

Age 130 0.7298Median (Range) 71.0 (61.0–78.0) 73.0 (65.0–78.5)

Tumor site 129 0.8305Rectum 16 6/16 (37.5) 10/16 (62.5)Sigmoid colon 28 14/28 (50.0) 14/28 (50.0)Descending colon 33 16/33 (48.5) 17/33 (51.5)Transverse colon 3 1/3 (33.3) 2/3 (66.7)Ascending colon 49 26/49 (53.0) 23/49 (47.0)

Histologic type and grading 52 0.728Adenocarcinoma G2 43 17/43 (39.5) 26/43 (60.5)Adenocarcinoma G3 9 3/9 (33.3) 6/9 (66.7)

Pathologic tumor stage 130 0.8118T1 9 3/9 (33.3) 6/9 (66.7)T2 8 4/8 (50.0) 4/8 (50.0)T3 87 46/87 (52.9) 41/87 (47.1)T4 26 11/26 (42.3) 15/26 (57.7)

Pathologic nodal stage 129 0.1329Nþ 62 26/62 (41.9) 36/62 (58.1)N- 67 37/67 (55.2) 30/67 (44.8)

UICC stagea 130 0.2405I 12 5/12 (41.7) 7/12 (58.3)II A 38 17/38 (44.7) 21/38 (55.3)II B 11 4/11 (36.4) 7/11 (63.6)III A 5 2/5 (40.0) 3/5 (60.0)III B 37 22/37 (59.5) 15/37 (40.5)III C 22 12/22 (54.5) 10/22 (45.5)IV 5 2/5 (40.0) 3/5 (60.0)

Alcohol 48 0.4888Never-drinkers 25 10/25 (40.0) 15/25 (60.0)Former/Current drinkers 23 7/23 (30.4) 16/23 (69.9)

Smoking 48 0.3111No 18 8/18 (44.4) 10/18 (55.6)Yes 30 9/30 (30.0) 21/30 (70.0)

KRAS mutation 128 0.777WTa 78 37/78 (47.4) 41/78 (52.6)Codon 12/13 exon 2 50 25/50 (50.0) 25/50 (50.0)

Codon 12 37 16/37 (43.2) 21/37 (56.8) 0.2616Codon 13 13 9/13 (69.2) 4/13 (30.8)

BRAF mutation 77 0.8465WTb 72 40/72 (55.5) 32/72 (44.5)V600E exon 15 5 3/5 (60.0) 2/5 (40.0)

MSI status 77 0.1275MSS 70 41/70 (58.6) 29/70 (41.4)MSI 7 2/7 (28.6) 5/7 (71.4)

Recurrence 49 0.6741No 37 15/37 (40.5) 22/37 (59.5)Locoregional 1 0 1/1 (100.0)Distant 11 5/11 (45.5) 6/11 (54.5)

Death (at last follow-up) 52 0.872Alive 41 16/41 (39.0) 25/41 (61.0)Died 11 4/11 (36.4) 7/11 (63.6)

aUICC, Union for International Cancer ControlbWT, wild type

B4GALT1 PROMOTER HYPERMETHYLATION IN COLORECTAL CANCER 7

Genes, Chromosomes & Cancer DOI 10.1002/gcc

et al., 2006). CEA is actually the only marker rou-

tinely used for monitoring metastatic CRC during

systemic therapy. Bearing in mind that aberrant

glycosylation could be considered an hallmark for

cancer, extensive studies have highlighted the role

of glycans as cancer biomarkers (Hakomori, 2002;

Abd Hamid et al., 2008; Hart and Copeland, 2010).

This study provides evidence that the pro-

moter of B4GALT1, a galactosyltransferase

involved in glycan biosynthesis (Furukawa and

Sato, 1999), is hypermethylated in 50% (65/130)

of CRC cases with an higher-than-90% specificity

rate. Highly specific markers can make bio-

markers in clinical setting more robust, thus

enhancing their cost-effectiveness. Promoter

methylation of B4GALT1 was analyzed in primary

tumor and normal matched mucosa by QMSP. In

the last years, a plethora of genes have been

found to be hypermethylated in CRC and two

promising markers Vimentin (ColoSure test) and

SEPT9 (Epi ProColon) are the only biomarkers

currently included in patented noninvasive tests.

They are used for early detection of stool and

blood analyses in CRC patients, respectively

(Chen et al., 2005; de Vos et al., 2009). In partic-

ular Vimentin is methylated in 0–29% of colo-

rectal normal mucosa (Kann et al., 2006; Zou

et al., 2007; Cho et al., 2010). Bearing in mind

that B4GALT1 has shown a methylation fre-

quency of 50% and a specificity overlapping with

the above mentioned markers (see Vimentin),

this gene could be considered a valuable cancer-

specific biomarker for CRC, especially if included

in a panel of hypermethylated genes.

Glycans of cancer cell surfaces undergo rele-

vant changes during malignant transformation

Figure 4. B4GALT1 protein expression by immunohistochemicalanalysis. Representative results of B4GALT1 immunostaining: Image A:normal colon mucosa; Images B, C: weak staining in CRC cases withmethylated B4GALT1; Images D, E: strong staining in CRC cases withno methylation at the B4GALT1 promoter. Tissue-associated macro-phages are positive for B4GALT1 staining and are used as internalpositive controls. Original magnification. X200 A, B, C, D, E images;X400 squared area of A image. [Color figure can be viewed in theonline issue, which is available at wileyonlinelibrary.com.]

Figure 3. Inverse correlation between B4GALT1 RERs and pro-moter methylation in colorectal tumor samples. B4GALT1 RERs of 13CRC patients (with fold change >2) were categorized by relativemethylation status (UNMET; unmethylated group, N ¼ 7, MET; meth-ylated group, N ¼ 6). P-value was calculated using the Mann-WhitneyU test. The asterisks (*) indicate outliers. [Color figure can be viewedin the online issue, which is available at wileyonlinelibrary.com.]

8 POETA ETAL.

Genes, Chromosomes & Cancer DOI 10.1002/gcc

(An et al., 2009; Satomaa et al., 2009), suggesting

that the expression of complex glycan structures

is in some way impaired upon malignant transfor-

mation. Incomplete synthesis and neosynthesis

are still the two major and relevant concepts for

cancer-associated alterations of cell surface carbo-

hydrate determinants, as formulated by Hakomori

almost 25 years ago (Hakomori et al., 1985). In

particular suppression of normal gene transcrip-

tion is proposed to lie behind incomplete glycan

synthesis in cancer cells. Despite the fact that

epigenetic mechanisms, such as DNA methyla-

tion, histone modification, chromatin remodeling,

and noncoding RNA, are key regulatory elements

needed to control gene expression, there have

been few studies addressing the epigenetic mech-

anisms of glycosyltransferase expression and only

a small number of glycogenes are known to be

silenced by DNA methylation (Kawamura et al.,

2008; Yusa et al., 2009). Indeed, cancer cells tend

to accumulate glycan branches with simpler struc-

tures in comparison to normal cells through epi-

genetic silencing of glycogenes involved in the

synthesis of complex carbohydrate determinants

and this is in line with the classical concept of

‘‘incomplete synthesis’’ advanced by Hakomori

(Kannagi et al., 2008). As promoter methylation

plays a critical role in transcription gene silencing,

we investigated the mRNA expression status of

B4GALT1 in primary CRC showing that the

expression level of B4GALT1 decreased in 68%

(13/19) of cases, and that promoter methylation

was correlated with reduced B4GALT1 transcript

levels, compared with normal colonic mucosa and

unmethylated cancer specimens. The concord-

ance between the mRNA level and the DNA

methylation strongly suggests that B4GALT1 is

downregulated through the DNA methylation in

CRC. Thus, we also evaluated the concordance

between methylation levels and protein expres-

sion, as determined by IHC analysis in five

unmethylated and five methylated tumors.

Although methylated tumors more frequently

showed weak/moderate B4GALT1 immunostain-

ing compared with unmethylated cases, this

difference was not statistically significant. We

believe that the reason for this imperfect correla-

tion between B4GALT1 expression and promoter

methylation might be due to the intrinsic tumor

heterogeneity. Indeed, IHC was performed on

the FFPE block used for diagnosis, whereas

methylation and mRNA expression were eval-

uated on the same fresh frozen specimen.

B4GALT1 is a beta-1,4-galactosyltransferase

that catalyzes the transfer of galactose from the

sugar nucleotide donor uridine diphosphate

galactose to glycoside residues with a terminal

N-acetylglucosamine (GlcNAc) moiety (Furukawa

and Sato, 1999). B4GALT1-deficient mice (Asano

et al.,1997) exhibit enhancement of epithelial cell

proliferation of the skin and small intestine and

abnormal cell differentiation in intestinal villi,

thus suggesting that B4GALT1 plays a critical role

in the regulation of proliferation and in the differ-

entiation of epithelial cells after birth. It is rea-

sonable to speculate that epigenetic silencing of

B4GALT1 expression could contribute to aberrant

epithelial differentiation. Previous studies have

suggested that epigenetic silencing of glycogenes

may occur at early stage of malignant transforma-

tion (Kannagi et al., 2008, 2010; Kawamura et al.,

2008). Although we found some levels of

B4GALT1 aberrant methylation in 3 out of 13

adenoma synchronous to adenocarcinoma, meth-

ylation levels were significantly higher in invasive

lesions compared with benign lesions, suggesting

that B4GALT1 methylation may mark the

onset of invasive phenotype in colorectal

cancerogenesis.

Mutations in the KRAS and BRAF genes as

well as MSI represent key genetic events in CRC

(Zlobec et al., 2010; Price et al., 2011). Previous

works have shown that B4GALT1 may play a role

in the RAS/BRAF pathway; SMMC-7721 human

hepatocarcinoma cells transfected with B4GALT1were more susceptible to cycloheximide-induced

apoptosis than the mock transfected cells trough

EGFR-PKB/Akt pathway (Zhu et al., 2003; Li

TABLE 2. Methylation Status of Synchronous Lesions of Colorectal Adenocarcinoma and Adenoma

Normal tissue, n(%) Adenoma, n(%) Carcinoma, n(%) P

UMa 13/13 (100%) 10/13 (77%) 4/13 (31%) 0.001b

Mc 0/13 (0) 3/13 (23%) 9/13 (69%)Median (IQR) 0 (0–0) 0 (0–9.14) 37.55 (0–292.19) 0.0001d

aUM, Unmethylated.bFisher Exact Test.cM, Methylated.dKruskall Wallis Test.

B4GALT1 PROMOTER HYPERMETHYLATION IN COLORECTAL CANCER 9

Genes, Chromosomes & Cancer DOI 10.1002/gcc

et al., 2006). We investigated the correlation

between these alterations and the methylation

status of B4GALT1. Aberrant epigenetic status of

this glycogene seems to be unrelated to gene

mutations in KRAS or BRAF and MSI. We also

evaluated whether aberrant methylation of

B4GALT1 could be related to pathological data

and outcome. Follow-up data were available for

52 out of 130 cases and no statistically significant

correlations were found. However, these results

may be due to the low number of cases and

events (death or progression) in this cohort.

In conclusion, our findings suggest that the gly-

cogene B4GALT1 is frequently methylated in

CRC and that it may be a specific biomarker for

colorectal neoplasia. Furthermore, methylation of

B4GALT1 have also been detected in adenoma

lesions, but methylation levels and frequencies

are significantly different than normal mucosa

only in invasive cancer lesions, thus suggesting

that this glycogene could be a valuable and spe-

cific candidate biomarker for detection of invasive

phenotype, providing preliminary evidence for

future studies aiming at investigating a wider

panel of galactosyltransferases which can elucidate

the lack of synthesis of cancer surface glycans. A

characterization of the epigenetic profile of glyco-

syltrasferase can be adopted as an alternative

approach to infer possible changes that may occur

in the structures of normal and cancer glycans,

thus overcoming the limits and technical difficul-

ties of studying complex glycoproteic markers.

REFERENCES

Abd Hamid UM, Royle L, Saldova R, Radcliffe CM, Harvey DJ,Storr SJ, Pardo M, Antrobus R, Chapman CJ, Zitzmann N,Robertson JF, Dwek RA, Rudd PM. 2008. A strategy to revealpotential glycan markers from serum glycoproteins associatedwith breast cancer progression. Glycobiology 18:1105–1118.

An HJ, Kronewitter SR, de Leoz ML, Lebrilla C. 2009. Glycomicsand disease markers. Curr Opin Chem Biol 13:601–607.

Asano M, Furukawa K, Kido M, Matsumoto S, Umesaki Y,Kochibe N, Iwakura Y. 1997. Growth retardation and earlydeath of b1,4-galactosyltransferase knockout mice with aug-mented proliferation and abnormal differentiation of epithelialcells. EMBO J 16:1850–1857.

Ballestar E, Esteller M. 2008. Epigenetic gene regulation in can-cer. Adv Genet 61:247–267.

Center MM, Jemal A, Smith RA, Ward E. 2009. Worldwide varia-tions in colorectal cancer. CA Cancer J Clin 59:366–378.

Chen WD, Han ZJ, Skoletsky J, Olson J, Sah J, Myeroff L, Plat-zer P, Lu S, Dawson D, Willis J, Pretlow TP, Lutterbaugh J,Kasturi L, Willson JK, Rao JS, Shuber A, Markowitz SD. 2005.Detection in fecal DNA of colon cancer-specific methylation ofthe nonexpressed vimentin gene. J Natl Cancer Inst 97:1124–1132.

Cho MH, Lee YM, Kim JS, Kim HS, Lee KH, Juhng SW, LeeJH. 2010. Aberrant promoter methylation of the vimentin genein colorectal cancer associated with the adenoma-carcinomasequence. Korean J Pathol 2:179–186.

Corvalan AH, Maturana MJ. 2010. Recent patents of DNA meth-ylation biomarkers in gastrointestinal oncology. Recent PatDNA Gene Seq 4:202–209.

de Vos T, Tetzner R, Model F, Weiss G, Schuster M, Distler J,Steiger KV, Grutzmann R, Pilarsky C, Habermann JK, FleshnerPR, Oubre BM, Day R, Sledziewski AZ, Lofton-Day C. 2009.Circulating methylated SEPT9 DNA in plasma is a biomarkerfor colorectal cancer. Clin Chem 55:1337–1346.

Durr ML, Mydlarz WK, Shao C, Zahurak ML, Chuang AY,Hoque MO, Westra WH, Liegeois NJ, Califano JA, SidranskyD, Ha PK. 2010. Quantitative methylation profiles for multipletumor suppressor gene promoters in salivary gland tumors.PLoS One 5:e10828.

Edwards BK, Ward E, Kohler BA, Eheman C, Zauber AG, Ander-son RN, Jemal A, Schymura MJ, Lansdorp-Vogelaar I, SeeffLC, van Ballegooijen M, Goede SL, Ries LA. 2010. Annualreport to the nation on the status of cancer, 1975–2006, featur-ing colorectal cancer trends and impact of interventions (riskfactors, screening, and treatment) to reduce future rates. Cancer116:544–573.

Fearon ER, Vogelstein B. 1990. A genetic model for colorectal tu-morigenesis. Cell 61:759–767.

Feinberg AP, Ohlsson R, Henikoff S. 2006. The epigenetic pro-genitor origin of human cancer. Nat Rev Genet 7:21–33.

Frattini M, Balestra D, Suardi S, Oggionni M, Alberici P, RadiceP, Costa A, Daidone MG, Leo E, Pilotti S, Bertario L, PierottiMA. 2004. Different genetic features associated with colon andrectal carcinogenesis. Clin Cancer Res 10:4015–4021.

Furukawa K, Sato T. 1999. Beta-1,4-galactosylation of N-glycansis a complex process. Biochim Biophys Acta 1473:54–66.

Grady WM, Markowitz SD. 2002. Genetic and epigenetic alterationsin colon cancer. Annu Rev Genomics Hum Genet 3:101–128.

Hakomori S. 1985. Aberrant glycosylation in cancer cell mem-branes as focused on glycolipids: Overview and perspectives.Cancer Res 45:2405–2414.

Hakomori S. 2002. Glycosylation defining cancer malignancy: Newwine in an old bottle. Proc Natl Acad Sci USA 99:10231–10233.

Hart GW, Copeland RJ. 2010. Glycomics hits the big time. Cell143:672–676.

Herceg Z, Hainaut P. 2007. Genetic and epigenetic alterations asbiomarkers for cancer detection, diagnosis and prognosis. MolOncol 1:26–41.

Herman JG, Baylin SB. 2003. Gene silencing in cancer in associationwith promoter hypermethylation. N Engl J Med 349:2042–2054.

Herranz M, Esteller M. 2007. DNA methylation and histone mod-ifications in patients with cancer: Potential prognostic and ther-apeutic targets. Methods Mol Biol 361:25–62.

Hoque MO, Begum S, Topaloglu O, Chatterjee A, Rosenbaum E,Van Criekinge W, Westra WH, Schoenberg M, Zahurak M,Goodman SN, Sidransky D. 2006a. Quantitation of promotermethylation of multiple genes in urine DNA and bladder can-cer detection. J Natl Cancer Inst 98:996–1004.

Hoque MO, Feng Q, Toure P, Dem A, Critchlow CW, HawesSE, Wood T, Jeronimo C, Rosenbaum E, Stern J, Yu M, TrinkB, Kiviat NB, Sidransky D. 2006b. Detection of aberrant meth-ylation of four genes in plasma DNA for the detection of breastcancer. J Clin Oncol 24:4262–4269.

Hoque MO, Kim Ms, Ostrow KL, Liu J, Wisman GB, Park HL,Poeta ML, Jeronimo C, Henrique R, Lendvai A, Schuuring E,Begum S, Rosenbaum E, Ongenaert M, Yamashita K, CalifanoJ, Westra W, van der Zee AG, Van Criekinge W, Sidransky D.2008. Genome-wide promoter analysis uncovers portions of thecancer methylome. Cancer Res 68:2661–2670.

Jemal A, Siegel R, Xu J, Ward E. 2010. Cancer statistics, 2010.CA Cancer J Clin 60:277–300.

Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D. 2011.Global cancer statistics. CA Cancer J Clin 61:69–90.

Jones PA, Baylin SB. 2007. The epigenomics of cancer. Cell128:683–692.

Kann L, Han J, Ahlquist D, Levin T, Rex D, Whitney D, Marko-witz S, Shuber A. 2006. Improved marker combination fordetection of de novo genetic variation and aberrant DNA incolorectal neoplasia. Clin Chem 12:2299–2302.

Kannagi R, Yin J, Miyazaki K, Izawa M. 2008. Current relevanceof incomplete synthesis and neo-synthesis for cancer-associatedalteration of carbohydrate determinants-Hakomori’s conceptsrevisited. Biochim Biophys Acta 1780:525–531.

Kannagi R, Sakuma K, Miyazaki K, Lim KT, Yusa A, Yin J, IzawaM. 2010. Altered expression of glycan genes in cancers inducedby epigenetic silencing and tumor hypoxia: Clues in theongoing search for new tumor markers. Cancer Sci 101:586–593.

10 POETA ETAL.

Genes, Chromosomes & Cancer DOI 10.1002/gcc

Kawamura YI, Toyota M, Kawashima R, Hagiwara T, Suzuki H,Imai K, Shinomura Y, Tokino T, Kannagi R, Dohi T. 2008.DNA hypermethylation contributes to incomplete synthesis ofcarbohydrate determinants in gastrointestinal cancer. Gastroen-terology 135:142–151.

Kim MS, Louwagie J, Carvalho B, Terhaar Sive Droste Js, ParkHL, Chae YK, Yamashita K, Liu J, Ostrow KL, Ling S, Guer-rero-Preston R, Demokan S, Yalniz Z, Dalay N, Meijer GA,Van Criekinge W, Sidransky D. 2009. Promoter DNA methyla-tion of oncostatin m receport–beta as novel diagnostic and ther-apeutic marker in colon cancer. PLoS One 4:e6555.

Levin B, Lieberman DA, McFarland B, Smith RA, Brooks D,Andrews KS, Dash C, Giardiello FM, Glick S, Levin TR, Pick-hardt P, Rex DK, Thorson A, Winawer SJ; American CancerSociety Colorectal Cancer Advisory Group; US Multi-SocietyTask Force; American College of Radiology Colon CancerCommittee. 2008. Screening and surveillance for the earlydetection of colorectal cancer and adenomatous polyps, 2008: Ajoint guideline from the American Cancer Society, the USMulti-Society Task Force on Colorectal Cancer, and the Ameri-can College of Radiology. CA Cancer J Clin 58:130–160.

Li Z, Zong H, Kong X, Zhang S, Wang H, Sun Q, Gu J. 2006.Cell surface beta 1, 4-galactosyltransferase 1 promotes apoptosisby inhibiting epidermal growth factor receptor pathway. MolCell Biochem 291:69–76.

Lind GE, Danielsen SA, Ahlquist T, Merok MA, Andresen K,Skotheim RI, Hektoen M, Rognum TO, Meling GI, Hoff G,Bretthauer M, Thiis-Evensen E, Nesbakken A, Lothe RA.2011. Identification of an epigenetic biomarker panel with highsensitivity and specificity for colorectal cancer and adenomas.Mol Cancer 10:85.

Locker GY, Hamilton S, Harris J, Jessup JM, Kemeny N, Mac-donald JS, Somerfield MR, Hayes DF, Bast RC Jr; ASCO.2006. ASCO 2006 update of recommendations for the use oftumor markers in gastrointestinal cancer. J Clin Oncol 24:5313–5327.

Mathews LA, Crea F, Farrar WL. 2009. Epigenetic gene regula-tion in stem cells and correlation to cancer. Differentiation78:1–17.

Molinari F, Felicioni L, Buscarino M, De Dosso S, Buttitta F,Malatesta S, Movilia A, Luoni M, Boldorini R, Alabiso O, Girl-ando S, Soini B, Spitale A, Di Nicolantonio F, Saletti P, CrippaS, Mazzucchelli L, Marchetti A, Bardelli A, Frattini M. 2011.Increased detection sensitivity for KRAS mutations enhancesthe prediction of anti-EGFR monoclonal antibody resistance inmetastatic colorectal cancer. Clin Cancer Res 17:4901–4914.

Ostrow KL, Hoque MO, Loyo M, Brait M, Greenberg A, Sieg-fried JM, Grandis JR, Gaither Davis A, Bigbee WL, Rom W,Sidransky D. 2010. Molecular analysis of plasma DNA for theearly detection of lung cancer by quantitative methylation-spe-cific PCR. Clin Cancer Res 6:3463–3472.

Parrella P, la Torre A, Copetti M, Valori VM, Barbano R, Nota-rangelo A, Bisceglia M, Gallo AP, Balsamo T, Poeta ML, Care-lla M, Catapano D, Parisi S, Dallapiccola B, Maiello E,D’Angelo V, Fazio VM. 2009. High specificity of quantitativemethylation-specific PCR analysis for MGMT promoter hyper-methylation detection in gliomas. J Biomed Biotechnol2009:531692.

Pepe MS, Etzioni R, Feng Z, Potter JD, Thompson ML, Thorn-quist M, Winget M, Yasui Y. 2001. Phases of biomarker devel-opment for early detection of cancer. J Natl Cancer Inst93:1054–1061.

Price TJ, Hardingham JE, Lee CK, Weickhardt A, Townsend AR,Wrin JW, Chua A, Shivasami A, Cummins MM, Murone C,Tebbutt NC. 2011. Impact of KRAS and BRAF gene mutationstatus on outcomes from the phase III AGITG MAX trialof capecitabine alone or in combination with bevacizumaband mitomycin in advanced colorectal cancer. J Clin Oncol 29:2675–2682.

Rennert G. 2009. Are we getting closer to molecular populationscreening for colorectal cancer? J Natl Cancer Inst 101:902–903.

Rodrıguez-Paredes M, Esteller M. 2011. Cancer epigeneticsreaches mainstream oncology. Nat Med 17:330–339.

Satomaa T, Heiskanen A, Leonardsson I, Angstrom J, Olonen A,Blomqvist M, Salovuori N, Haglund C, Teneberg S, Natunen J,Carpen O, Saarinen J. 2009. Analysis of the human cancer gly-come identifies a novel group of tumor-associated N-acetylglu-cosamine glycan antigens. Cancer Res 69:5811–5819.

Schmid G. 2010. The use of molecular markers in the diagnosis ofcolorectal cancer screening. Dig Dis 28:625–628.

Sharma S, Salehi F, Scheithauer BW, Rotondo F, Syro LV, KovacsK. 2009. Role of MGMT in tumor development, progression,diagnosis, treatment and prognosis. Anticancer Res 29:3759–3768.

Smith RA, Cokkinides V, Eyre HJ. 2006. American Cancer Soci-ety for the guidelines early detection of cancer, 2006. CA Can-cer J Clin 56:11–25.

Smith RA, Cokkinides V, Brooks D, Saslow D, Shah M, BrawleyOW. 2011. Cancer screening in the United States, 2011: Areview of current American Cancer Society guidelines andissues in cancer screening. CA Cancer J Clin 61:8–30.

Tsai HC, Baylin SB. 2011. Cancer epigenetics: linking basic biol-ogy to clinical medicine. Cell Res 21:502–517.

van Engeland M, Derks S, Smits KM, Meijer GA, Herman JG.2011. Colorectal cancer epigenetics: complex simplicity. J ClinOncol 29:1382–1391.

Verma M, Manne U. 2006. Genetic and epigenetic biomarkers incancer diagnosis and identifying high risk populations. Crit RevOncol Hematol 60:9–18.

Whitlock EP, Lin JS, Liles E, Beil TL, Fu R. 2008. Screening forcolorectal cancer: A targeted, updated systematic review for theU.S. Preventive Services Task Force. Ann Intern Med 149:638–658.

Yamada Y, Watanabe A. 2010. Epigenetic codes in stem cells andcancer stem cells. Adv Genet 70:177–199.

Yusa A, Miyazaki K, Kimura N, Izawa M, Kannagi R. 2009. Epi-genetic silencing of sulfate transporter DTDST in colon cancercells leads to decreased sulfated glycan expression. Proceedingsfor 68th Annual Meeting of the Japanese Cancer Association,Tokyo: Japanese Cancer Association 41.

Zhu X, Chen S, Yin X, Shen A, Ji S, Shen Z, Gu J. 2003. Consti-tutively active PKB/Akt inhibited apoptosis and down-regulatedbeta 1,4-galactosyltransferase 1 in hepatocarcinoma cells. Bio-chem Biophys Res Commun 309:279–285.

Zlobec I, Kovac M, Erzberger P, Molinari F, Bihl MP, Rufle A,Foerster A, Frattini M, Terracciano L, Heinimann K, Lugli A.2010. Combined analysis of specific KRAS mutation, BRAF andmicrosatellite instability identifies prognostic subgroups ofsporadic and hereditary colorectal cancer. Int J Cancer 127:2569–2575.

Zou H, Harrington JJ, Shire Am, Rego RL, Wang L, CampbellME, Oberg AL, Ahlquist DA. 2007. Highly methylated genesin colorectal neoplasia: Implications for screening. Cancer Epi-demiol Biomarkers Prev 12:2686–2696.

B4GALT1 PROMOTER HYPERMETHYLATION IN COLORECTAL CANCER 11

Genes, Chromosomes & Cancer DOI 10.1002/gcc