A new strategy for the diagnosis of MAGE-expressing cancers

A new strategy for the diagnosis of MAGE-expressing cancers

Jong-Wook Park a, Taeg Kyu Kwon a, In-Ho Kim a, Soo-Sang Sohn a, You-Sah Kim a,Chun-Il Kim a, Ok Suk Bae a, Kyung Seop Lee b, Kang-Dae Lee c, Cheong-Sam Lee c,

Hee-Kyung Chang d, Byung-Kil Choe d, Su Yul Ahn e, Chang-Ho Jeon f,*

aThe Institute of Medical Science, Keimyung University School of Medicine, iC&G Co., Taegu, South KoreabDepartment of Urology, Dong-Guk University School of Medicine, Kyungju, South Korea

cDepartment of Otolaryngology-Head and Neck Surgery, Kosin University, Pusan, South KoreadDepartment of Pathology, College of Medicine, Kosin University, Pusan, South Korea

eDepartment of Internal Medicine, Yeungnam University School of Medicine, Taegu, South KoreafDepartment of Clinical Pathology, School of Medicine, Catholic University of Taegu, 3056-6 Daemyung-Dong, Nam-Gu, Taegu, South Korea

Received 21 November 2001; received in revised form 3 April 2002; accepted 3 April 2002

Abstract

The expression of melanoma antigen gene (MAGE), coding for tumor antigens recognized by cytotoxic T cell, is highly

specific to cancer cells, but their use in the detection of a few cancer cells by reverse transcription-polymerase chain reaction

(RT-PCR) has been limited by the low frequency of expression of individual MAGE genes. In order to increase MAGE

detection rate in RT-PCR assay, here, we designed multiple MAGEs recognizing primers (MMRPs) that can bind to the

sequences of cDNA of MAGE-1, -2, -3, -4a, -4b, -5a, -5b and-6 (MAGE 1–6) together. The nested RT-PCR assay using

MMRPs, MAGE 1–6 assay, detected MAGE messages of 1 to 5 SNU484 cells in a background of 107 SNU638 cells. MAGE

detection rate of MAGE 1–6 assay in cancers was higher than that of nested RT-PCR that detects single MAGE gene

expression. The expressions of MAGE genes was detected by MAGE 1–6 assay in 70.4% (19/27) of head and neck cancer

tissues, 91.7% (11/12) of breast cancer tissues, 75% (9/12) of lung cancer tissues. However, they were not detected in 18 benign

lesions and 20 normal head and neck tissues and 30 blood samples from healthy donor. In conclusions, MAGE 1–6 assay can

detect any cancer cells that express at least one of eight MAGE subtype genes, and this method may be very useful for the

diagnosis of MAGE-expressing cancers.

D 2002 Elsevier Science B.V. All rights reserved.

Keywords: MAGE; Nested RT-PCR; Cancer diagnosis

1. Introduction

Many human melanomas express antigens that are

specific targets of the cytotoxic T lymphocyte of tumor-

bearing patients. Melanoma antigen gene (MAGE),

which is one of them, has been studied for tumor

immunotherapy and diagnosis (Tureci et al., 1998;

Castelli et al., 2000; Nishiyama et al., 2001). There

0022-1759/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved.

PII: S0022 -1759 (02 )00105 -9

Abbreviations: MAGE, melanoma antigen gene; MMRP, multi-

ple MAGE subtypes recognizing primers; RT-PCR, reverse tran-

scription-polymerase chain reaction; ADC, 5-aza-2V deoxycytidine.* Corresponding author. Tel.: +82-53-650-4144; fax: +82-53-

255-1398.

E-mail address: [email protected] (C.-H. Jeon).

www.elsevier.com/locate/jim

Journal of Immunological Methods 266 (2002) 79–86

are many kinds of MAGE-expressing tumors. MAGE

is expressed in stomach cancer (Li et al., 1997),

esophageal cancer (Inoue et al., 1995), colorectal

cancer (Mori et al., 1996), lung cancer (Weynants et

al., 1994), breast cancer (Russo et al., 1995), hepato-

cellular cancer (Yamashita et al., 1996), ovary cancer

(Russo et al., 1996), lymphocytic leukemia (Shichijo et

al., 1995), and so on. Because MAGE is expressed in

many kinds of cancers and MAGE gene expression is

highly specific to cancer cells, MAGE has been studied

as an important marker for cancer diagnosis.

MAGE A family consists of several subtypes

including MAGE-1 to MAGE-12. During last several

years, many researchers have studied the expression

of individual MAGE A genes for cancer diagnosis,

however, use of MAGE genes in the detection of a

few cancer cells by reverse transcription-polymerase

chain reaction (RT-PCR) has been limited by the low

expression frequency of individual MAGE genes in

various cancer tissues. Many researchers reported that

the theoretical frequency that cancer expresses at least

one of the MAGE genes tested was very high (Eura et

al., 1995; Yoshimatsu et al., 1998; Chen et al., 1999;

Otte et al., 2001). We reasoned that if primers can

bind to the cDNAs of multiple MAGE genes together,

nested RT-PCR based on these primers might detect

cancers that express at least one of MAGE genes

tested. After comparing gene sequences, we designed

primers, multiple MAGE recognizing primers

(MMRPs), that can bind to the cDNA of MAGE-1,

-2, -3, -4a, -4b, -5a, -5b and -6 (MAGE 1–6) together.

Here, we show the results of nested RT-PCR assay

using MMRPs for the diagnosis of head and neck

cancer, breast cancer and lung cancer.

2. Materials and methods

2.1. Cell culture and induction of MAGE gene

expression

Cell lines established from stomach cancer

(SNU484, SNU638) were cultured in RPMI1640 sup-

plemented with 10% FBS and 1� antibiotics and

antimycotics. For the induction of MAGE gene expres-

sion, SNU 484 and SNU 638 were treated with 5-aza-

2V-deoxycytidine (ADC, 1 Amol) for 48 h and the

message of MAGE was detected by nested RT-PCR.

MAGE protein induced by ADC treatment in SNU 638

was detected by immunocytochemistry. After fixing

and blocking SNU 638 cells treated with ADC for 48,

MAGE protein was detected by anti-MAGE-3 mono-

clonal antibody 57B (Rimoldi et al., 2000) and perox-

idase conjugated rabbit anti-mouse IgG antibody.

2.2. Specimens

We collected 12 breast cancer tissues from Dong-

san medical center in Taegu, South Korea; 27 head

and neck cancer tissues, 18 benign lesions and 20

normal tissues of head and neck, 20 blood samples of

head and neck cancer patients and 30 blood samples

of normal healthy volunteers from Kosin medical

center in Pusan, South Korea; 12 lung cancer tissues

from Taegu catholic university, in Taegu, South

Korea. Histological diagnosis of cancer and benign

disease was obtained in all cases. The tissues were

kept at � 70 jC until total RNA isolation.

2.3. Designing primers

MAGE gene sequences were obtained from Gen-

bank data (National Center for Biotechnology Infor-

mation, USA), and DNA homology of each MAGE

genes was studied by DNAsis program (Hitachi,

Japan). The common DNA sequences that are present

in MAGE subtypes were used for designing MMRPs

(MMRP1, 5V-CTGAAGGAGAAGATCTGCC-3V;MMRP2, 5V-CTCCAGGTAGTTTTCCTGCAC-3V;MMRP3, 5V-CTGAAGGAGAAGATCTGCCWGTG-

3V, W is A or T; MMRP4, 5V-CCAGCATTTCT-

GCCTTTGTGA-3V), and the specific DNA sequences

that present in each MAGE subtype were used for

designing MAGE specific primers (M1 primer for

MAGE-1, 5V-CGGAACAAGGACTCCAGGATA-

CAA-3V; M2 primer for MAGE-2, 5V-GAAAGA-AGTCCTGGCAATTTCTGAG-3V; M3 primer for

MAGE-3, 5V-CCAAAGACCAGCTGCAAGGAACT-3V; M4 primer for MAGE-4, 5V-CGTCAATGCCAAA-GATCATCTTCAG-3V; M5 primer for MAGE-5, 5V-CCTTTGTGACCAGCTCCTTGACTTA-3V; M6 pri-

mer for MAGE-6, 5V-CCAGGCAGGTGGCAAA-

GATGTACAC-3V). The binding site of MMRP1 and

MMRP3 onMAGE-3 genomic DNA and DNA homol-

ogy between primers and genomic DNAs were sum-

marized in Fig. 1.

J.-W. Park et al. / Journal of Immunological Methods 266 (2002) 79–8680

2.4. Nested RT-PCR

Total RNA of tissue specimens, blood leukocytes

(5–6 ml) and cell lines was extracted using Trizol

according to the manufacturer’s instructions (Gibco

BRL). Purity and quality of RNA were assessed by

ultraviolet spectrophotometry. Reverse transcription

reactions were carried out in a 20-ul reaction mixture

that contains 50 mM Tris–HCl, 75 mM KCl, 3 mM

MgCl2, 10 mM DTT, 250 AM dATP, 250 uM dCTP,

250 AM dTTP, 250 AM dGTP, RNase inhibitor (0.75

U/Al), MMLV reverse transcriptase (5 U/Al), 2.5 AMOligo dT primer and 4 Ag denatured RNA. The

reaction mixture was incubated at 42 jC for 60 min,

at 95 jC for 5 min and then stored at � 20 jC until

further use.

The first PCR reactions were carried out in 20 Alreaction mixture that contains 10 mM Tris–HCl, 50

mM KCl, 1.5 mM MgCl2, 200 AM dATP, 200 AMdCTP, 200 AM dTTP, 200 AM dGTP, 0.6 U Taq DNA

polymerase, 0.5 AM sense primer (MMRP1), 0.5 AManti-sense primer (MMRP2) and 2 Al of the products

of the reverse transcription reaction. The cycling

parameters were as follows: initiated denaturation at

95 jC for 5 min, followed by 30 cycles of 95 jC for

30 s, 60 jC for 45 s, and 72 jC for 45 s. The final

extension incubation was performed at 72 jC for 10

min. The 1 Al of first PCR products was used as

template for second (nested) PCR. The nested PCR

conditions were same with the first PCR except

primers and template. The combination of MMRP3

(sense primer) and MMRP4 (Antisense primer) was

used for the detection of the gene expression of

MAGE 1–6 together, and the combination of

MMRP3 (sense primers) and one of MAGE isotype

specific primers (M1, M2, M3, M4, M5, M6; Anti-

sense primers) was used for the detection of the gene

expression of individual MAGE isotype. All PCR

products were separated on 1% agarose gels impreg-

nated with ethidium bromide (0.5 Ag/ml).

2.5. Sequence analysis

To verify that PCR products amplified by the nested

RT-PCR using MMRPs were the expected members of

the MAGE 1–6, cDNA was amplified by nested RT-

Fig. 1. The map and DNA sequences of MMRPs. (A) MMRPs

binding site on MAGE-3 genomic DNA. DNA sizes of PCR

products using MMRP1/2 and MMRP3/4 are 831–855 and 469–

493 bp, respectively. (B) DNA homology between MMRPs and

MAGE 1–6. Blank box represents that DNA base of primer and

genes of MAGE members is the same. W is A or T.

Fig. 2. Analysis of MAGE gene expression in SNU 484 and SNU

638. (A) Detection of the gene expression of each MAGE subtype

in SNU484 and SNU638 by RT-PCR using MMRP1/2 and nested

PCR using MMRP3 and MAGE specific primers (M1–M6). (B)

Detection of the gene expression of MAGE 1–6 together in

SNU484. Lane 1, results of first PCR using MMRP-1/2; lane 2,

results of second PCR using MMRP3/4. (C) Detection of the gene

expression of MAGE 1–6 together in SNU484 and SNU 638

treated with/without ADC by MAGE 1–6 assay. (D) Detection of

MAGE-3 protein in SNU 638 treated with/without ADC by anti-

MAGE-3 monoclonal antibody.

J.-W. Park et al. / Journal of Immunological Methods 266 (2002) 79–86 81

PCR from total RNA of SNU 484. The PCR products

from the SNU484 cell line were extracted from 1%

agarose gels using the QIAquick gel extraction method

(Qiagen) according to the manufacturer’s instructions.

After cloning cDNA into pGEM-T vector (Promega),

the sequence of insert was analyzed by Sanger’s

dideoxyneucleotide chain termination method, and it

was compared with that of MAGE 1–6 registered in

Genbank using DNAsis program (Hitachi).

3. Results

3.1. Analysis of MAGE gene expression in SNU 484

and SNU 638

In order to know MAGE isotype expression profile

of SNU 484 and SNU 638, nested RT-PCR using

MMRP3 and MAGE-isotype specific primers was

carried out. SNU 484 expressed all of six MAGE

subtypes, but SNU 638 did not (Fig. 2A). MAGE

cDNAs amplified from SNU 484 were cloned and

sequenced. The DNA sequences of each cloned

cDNAs showed 100% homology with the mRNA

sequences of MAGE 1–6 that were registered in

Genbank. Total RNA and cDNA of SNU 484 were

used as template for PCR using MMRP1 and MMRP2

(MMRP1/2) and nested PCR using MMRP3 and

MMRP4 (MMRP3/4). About 831–855 bp (Fig. 2B,

lane 1) and 469–493 bp (Fig. 2B, lane 2) DNA

fragments were amplified from cDNA by first RT-

PCR and second PCR, respectively. However, without

reverse transcription, these DNA bands were not

detected (Fig. 2A, lanes 3 and 4). We refer to nested

RT-PCR assay using MMRPs (MMRP1/2 for first

PCR, MMRP3/4 for nested PCR) as the MAGE 1–

6 assay in the following studies.

MAGE expression can be induced by DNA deme-

thylation (De Smet et al., 1996). In order to confirm

that MAGE 1–6 cDNA is amplified from mRNA by

MAGE 1–6 assay, SNU 484 and SNU 638 were

treated with or without ADC for 48 h. After isolating

total RNA of each cell lines, MAGE 1–6 assay was

performed. The expression of MAGE 1–6 genes was

detected in SNU 638 treated with ADC as well as

SNU 484 (Fig. 2C). These results suggest that the

expression of MAGE 1–6 can be induced by DNA

demethylating agent and MAGE 1–6 assay can detect

only the transcript of MAGE 1–6 gene. SNU 638

treated with ADC was also stained with anti-MAGE-3

monoclonal antibody (Rimoldi et al., 2000) that binds

to MAGE-1, -2, -4, -6 as well as MAGE-3 protein by

cross-reaction (Fig. 2D).

Fig. 3. MAGE detection sensitivity of MAGE 1–6 assay. SNU484

was mixed with SNU638 with various rates and the total RNA (4

Ag) isolated from mixed cells was used for MAGE 1–6 assay. One

microliter of 1st RT-PCR products diluted with DW (100-fold) was

used for 2nd PCR.

Fig. 4. Detection of the MAGE expressing cells by MAGE 1–6 assay in blood samples. Blood leukocytes of healthy volunteers (A) and head

and neck cancer patients (B) were isolated by ACK lysis method. Total RNA (4 Ag) isolated from cells was used for the detection of the

expression of MAGE 1–6 by MAGE 1–6 assay. (+) Normal blood samples containing five SNU484 cells was used as positive control.


3.2. MAGE detection sensitivity of MAGE 1–6 assay

Various numbers of SNU 484 cells, MAGE-pos-

itive cell line, were mixed with SNU 638 cells, and

the total RNA isolated from mixed cells was used for

MAGE 1–6 assay. The results showed that 20 PCR

cycles of nested PCR were enough to detect MAGE

messages of 1 to 5 SNU484 cells in a background of

107 SNU638 cells (Fig. 3).

3.3. Detection of MAGE gene expression in blood and

tissue

The expression of MAGE genes in blood cells of

healthy volunteers and head and neck cancer patients

was detected by MAGE 1–6 assay. MAGE cDNA

was not detected in normal blood samples (n = 30).

The (+) symbols in Fig. 4A is normal blood sample

that contains five SNU 484 cells. In the blood samples

(n = 20) of head and neck cancer patient, MAGE was

detected in three cases (Fig. 4B).

In order to know MAGE detection efficacy of

MAGE 1–6 assay, we isolated total RNA from 27

head and neck cancer tissues, 12 breast cancer tissues,

and 12 lung cancer tissues. The results of MAGE 1–6

assay were compared with those of first RT-PCR

using MMRP1/2 and nested PCR using MMRP3

and MAGE specific primers. MAGE-1, -2, -3, -4, -5,

-6 genes were expressed in 22.2%, 37.0%, 55.6%,

40.7%, 37.0% and 48.1% of 27 head and neck cancer,

respectively. Out of 12 samples of breast cancer,

58.3%, 41.7%, 66.7%, 41.7%, 25.0% and 33.3%

expressed MAGE-1, -2, -3, -4, -5 and -6 gene, respec-

tively. Out of 12 samples of lung cancer, 25.0%,

16.7%, 41.7%, 33.3%, 8.33% and 25.0% expressed

MAGE-1, -2, -3, -4, -5 and -6 gene, respectively. In

the MAGE 1–6 assay, however, MAGE gene expres-

sion was detected in 70.4% (19/27) of head and neck

cancer, 91.7% (11/12) of breast cancer and 75.0% (9/

12) of lung cancer (Table 1, Fig. 5).

Among 27 head and neck cancer tissues, 12 breast

cancer tissues and 12 lung cancer patient, only four

cases of head and neck cancer tissues expressed all of

six MAGE subtypes together, and many cases of them

(40.7% of head and neck cancer, 58.3% of breast

cancer and 66.3% of lung cancer) expressed only

three or less kinds of MAGE subtypes (Table 2). In

20 normal tissues and 18 benign lesions of head and

Table 1

Expression rate of MAGE 1–6 genes in cancer tissue of head and

neck, breast and lung

MAGE No. of MAGE-expressing patients (%)

H&N cancer

(n= 27)

Breast cancer

(n= 12)

Lung cancer

(n= 12)

MAGE-1 6 (22.2) 7 (58.3) 3 (25.0)

MAGE-2 10 (37.0) 5 (41.7) 2 (16.7)

MAGE-3 15 (55.6) 8 (66.7) 5 (41.7)

MAGE-4 11 (40.7) 5 (41.7) 4 (33.3)

MAGE-5 10 (37.0) 3 (25.0) 1 (8.33)

MAGE-6 13 (48.1) 4 (33.3) 3 (25.0)

MAGE 1–6 19 (70.4) 11 (91.7) 9 (75.0)

Fig. 5. Detection of the messages of MAGE subtypes in head and neck cancer tissues. Expression of each MAGE subtype was detected by RT-

PCR using MMRP1/2 and nested PCR using MMRP3 and M1–M6. S, size marker; (� ) no cDNA as negative control, MAGE 1–6, results of

MAGE 1–6 assay.


neck cancer patients, MAGE messages were not

detected at all by MAGE 1–6 assay (data not shown).

4. Discussion

Although RT-PCR can detect a few cancer cells

mixed in millions of normal cells (Bostick et al., 1999),

there are limitations on the specificity and amount of

mRNA detected by individual molecular marker(s)

(Bostick et al., 1998). MAGE gene expression is highly

specific to cancer cells, but their use in the detection of

a few cancer cells by RT-PCR also has been limited by

the low expression frequency of individual MAGE

genes. For the last several years, many researchers

reported that the theoretical frequency that cancer

expresses at least one of MAGE genes tested was high

(Eura et al., 1995; Yoshimatsu et al., 1998; Chen et al.,

1999; Otte et al., 2001). These suggestions mean that

the detection of multiple MAGE gene expression

together may be better than that of single gene expres-

sion for the diagnosis of MAGE-expressing cancer.

Here, we have developed a new nested RT-PCR

method that uses MMRPs for detection of several

MAGE gene expressions together.

In order to detect several MAGE gene expressions

together by RT-PCR, here, we used high DNA

homology of MAGE genes to design primers. There

were several candidate sequences for designing pri-

mers in one main exon that contains full length of

coding sequences for MAGE, but when these sequen-

ces were used as primers, genomic MAGE DNA was

mainly amplified by PCR without reverse transcrip-

tion (data not shown). Theses results suggest that

these primers may bind to the exons in genomic

DNA of multiple individual MAGE genes together

and nested PCR were enough to amplify DNA con-

taminants in total RNA samples. In order to rule out

the possible genomic DNA amplification, we used

DNase-treated total RNA as template or mRNA

selective PCR Kit (TaKaRa, Japan), but all of these

methods were not enough to remove genomic DNA

amplification completely. Finally, we designed new

two upstream primers (MMRP1 and MMRP3) that

can bind to two exons of MAGE gene together. These

primers can bind to cDNA but not bind to genomic

DNA, because 3V end of primer sequence was not

able to bind to intron between two genomic DNA

exons. First RT-PCR using MMRP1/2 and nested

PCR using MMRP3/4 amplified only MAGE 1–6

cDNAs. We refer to this method as the MAGE 1–6

assay. MAGE 1–6 assay can detect the expression of

MAGE-1, -2, -3, -4a, -4b, -5a, 5b, -6, together. It

detected one to five SNU484 cells mixed in 2� 107

SNU638 cells and five SNU484 cells mixed in 6 ml

whole normal blood. The sum of amplified cDNAs of

several genes by MAGE 1–6 assay increased the

intensity of DNA stained with ethidium bromide in

electrophoresis gel. These properties of MAGE 1–6

assay may be very important for detection of small

numbers of cancer cells mixed in many normal cells.

MAGE can be expressed in head and neck cancer

(Eura et al., 1995), breast cancer (Otte et al., 2001) and

lung cancer (Yoshimatsu et al., 1998), but the expres-

sion frequency of each subtype gene is low. The

MAGE-1, -2, -3, -4, -41 and -6 genes were expressed

at the mRNA level in 27, 34, 36, 22, 16 and 35,

respectively, of 83 fresh head and neck tumor samples

(Eura et al., 1995); and MAGE-1, -2, and -3 was

positive in 9/43, 13/43, and 22/48, respectively, in the

lung cancer samples (Yoshimatsu et al., 1998), and

MAGE-1, -2, -3, -4, -6, -12 was positive in 4, 13, 7, 9,

10, 6, respectively, of 67 breast cancer samples (Otte et

al., 2001). In this report, we also showed that expres-

sion frequency of each subtype gene was not high and

lots of samples expressed less than 3 kinds of MAGE

subtypes. However, most researchers who studied

MAGE gene expression in cancer estimated theoret-

ical frequency that cancer expresses at least one gene.

Fifty nine/83 (71.1%) of head and neck cancer sam-

ples (Eura et al., 1995) and 18/28 (64.3%) of breast

cancer samples (Otte et al., 2001) showed expression

Table 2

Numbers of positively detected MAGE 1–6 genes in cancer tissue

of head and neck, breast and lung

Expressed No. of MAGE-expressing patients (%)

gene no. H&N cancer Breast cancer Lung cancer

0 8 (29.6) 1 (8.3) 3 (25)

1 3 (11.1) 1 (8.3) 3 (25)

2 4 (14.8) 5 (41.7) 4 (33)

3 4 (14.8) 1 (8.3) 1 (8.3)

4 2 (7.4) 2 (16.7) 1 (8.3)

5 2 (7.4) 2 (16.7) 0 (0)

6 4 (14.8) 0 (0) 0 (0)

Total 27 12 12


of at least one gene. With MAGE 1–6 assay, we also

found that 70.4% of head and neck cancer (n = 27),

91.7% of breast cancer (n = 12) and 75.0% of lung

cancer (n = 12) expressed at least one gene. MAGE is

also detected in another cancers. In gastric cancer, 44/

54 (81.5%) cases expressed at least one of 7 MAGE

genes (Li et al., 1997), and in hepatocellular carcino-

mas, 37/50 (74%) cases expressed at least one of four

MAGE genes (Chen et al., 1999). MAGE expression

frequency and subtype expression profile are various

according to cancer origin and types, and there may be

no significant correlation between MAGE expression

and clinical parameters including clinical stages and

metastasis (Lee et al., 1999). All these reports suggest

that expression profile of MAGE subtypes may not be

important for cancer diagnosis, and detection of multi-

ple MAGE gene expression together, like MAGE 1–6

assay or uMAGE-A assay (Miyashiro et al., 2001),

may be more useful than that of single gene expression

for the diagnosis of MAGE-expressing cancer.

RT-PCR-based amplification of transcripts

expressed in cancer but not in normal non-neoplastic

cells is increasingly used for the sensitive detection of

rare disseminated or exfoliated cancer cells to improve

cancer staging and early detection protocols. How-

ever, these assays are frequently hampered by false-

positive test results due to low-level transcription of

the marker genes in normal cells (Bialkowska-Hobr-

zanska et al., 2000; Piva et al., 2000; Lacroix et al.,

2001; Aerts et al., 2001). In this study, in spite of high

sensitivity of MAGE 1–6 assay, the messages of

MAGE 1–6 were not detected at all in normal blood

samples and benign head and neck tissue. These

results suggest MAGE genes are silent in normal head

and neck tissue and peripheral blood leukocyte, and

MAGE 1–6 assay may detect cancer cells without

false-positive reaction.

RT-PCR has been used as a powerful tool to detect

small numbers of cancer cells in blood. Mori et al.

(1997) reported 6 of 18 gastrointestinal cancer patients

had positive expression of MAGE gene in blood

samples. Three of six MAGE-positive patients devel-

oped metastatic disease, whereas none of 12 MAGE-

negative patients in blood developed metastasis. In our

results, MAGE 1–6 assay detected MAGE gene

expression in 6 ml of whole normal blood samples

containing five SNU 484 cells, and 3 of the 20 blood

samples from head and neck cancer patients. Recently,

we also have been appliedMAGE 1–6 assay to sputum

samples for the diagnosis of lung cancer. MAGE 1–6

assay detected MAGE gene expression in normal

sputum samples containing 1–2 SNU 484 cells, 4/

192 (2.1%) of sputum samples of non-lung cancer

patients and 23/31 (74.2%) of sputum samples of lung

cancer patients. These results suggest MAGE 1–6

assay could be used to detect small numbers of cancer

cell in blood or sputum for early cancer diagnosis or

detection of metastasis.

In conclusions, the primers usually have been

designed to detect the specific expression of one gene,

but here, we designed MMRPs to detect several

MAGE gene expressions together. MAGE 1–6 assay

could detect cancer cells that express at least one of

eight MAGE subtypes. We suggest that this method

may be used for the diagnosis of many kinds of

MAGE-expressing cancers.

Acknowledgements

This study was supported by the Dongsan Medical

Center grant (Suggestion Research Fund, 1999), and

iC&G company grant, Taegu 700-712, South Korea.

References

Aerts, J., Wynendaele, W., Paridaens, R., Christiaens, M.R., van den

Bogaert, W., van Oosterom, A.T., Vandekerckhove, F., 2001. A

real-time quantitative reverse transcriptase polymerase chain re-

action (RT-PCR) to detect breast carcinoma cells in peripheral

blood. Ann. Oncol. 12, 39–46.

Bialkowska-Hobrzanska, H., Bowles, L., Bukala, B., Joseph, M.G.,

Fletcher, R., Razvi, H., 2000. Comparison of human telomerase

reverse transcriptase messenger RNA and telomerase activity as

urine markers for diagnosis of bladder carcinoma. Mol. Diagn.

5, 267–277.

Bostick, P.J., Chatterjee, S., Chi, D.D., Huynh, K.T., Giuliano, A.E.,

Cote, R., Hoon, D.S., 1998. Limitations of specific reverse-tran-

scriptase polymerase chain reaction markers in the detection of

metastases in the lymph nodes and blood of breast cancer pa-

tients. J. Clin. Oncol. 16, 2632–2640.

Bostick, P.J., Morton, D.L., Turner, R.R., Huynh, K.T., Wang, H.J.,

Elashoff, R., Essner, R., Hoon, D.S., 1999. Prognostic signifi-

cance of occult metastases detected by sentinel lymphadenec-

tomy and reverse transcriptase-polymerase chain reaction in

early-stage melanoma patients. J. Clin. Oncol. 17, 3238–3244.

Castelli, C., Rivoltini, L., Andreola, G., Carrabba, M., Renkvist, N.,

Parmiani, G., 2000. T-cell recognition of melanoma-associated

antigens. J. Cell. Physiol. 182, 323–331.


Chen, C.H., Huang, G.T., Lee, H.S., Yang, P.M., Yan, M.D., Chen,

D.S., Sheu, J.C., 1999. High frequency of expression of MAGE

genes in human hepatocellular carcinoma. Liver 19, 110–114.

De Smet, C., De Backer, O., Faraoni, I., Lurquin, C., Brasseur, F.,

Boon, T., 1996. The activation of human gene MAGE-1 in

tumor cells is correlated with genome-wide demethylation. Proc.

Natl. Acad. Sci. U. S. A. 93, 7149–7153.

Eura, M., Ogi, K., Chikamatsu, K., Lee, K.D., Nakano, K., Masuya-

ma, K., Itoh, K., Ishikawa, T., 1995. Expression of the MAGE

gene family in human head-and-neck squamous-cell carcino-

mas. Int. J. Cancer 64, 304–308.

Inoue, H., Mori, M., Honda, M., 1995. Human esophageal carcino-

mas frequently express the tumor-rejection antigens of MAGE

genes. Int. J. Cancer 63, 523–526.

Lacroix, J., Becker, H.D., Woerner, S.M., Rittgen, W., Drings, P.,

von Knebel Doeberitz, M., 2001. Sensitive detection of rare

cancer cells in sputum and peripheral blood samples of patients

with lung cancer by preproGRP-specific RT-PCR. Int. J. Cancer

92, 1–8.

Lee, K.D., Chang, H.K., Jo, Y.K., Kim, B.S., Lee, B.H., Lee, Y.W.,

Lee, H.K., Huh, M.H., Min, Y.G., Spagnoli, G.C., Yu, T.H.,

1999. Expression of the MAGE 3 gene product in squamous

cell carcinomas of the head and neck. Anticancer Res. 19,

5037–5042.

Li, J., Yang, Y., Fujie, T., Tanaka, F., Mimori, K., Haraguchi, M.,

Ueo, H., Mori, M., Akiyoshi, T., 1997. Expression of the

MAGE gene family in human gastric carcinoma. Anticancer

Res. 17, 3559–3563.

Miyashiro, I., Kuo, C., Huynh, K., Iida, A., Morton, D., Bilchik, A.,

Giuliano, A., Hoon, D.S., 2001. Molecular strategy for detecting

metastatic cancers with use of multiple tumor-specific MAGE-A

genes. Clin. Chem. 47, 505–512.

Mori, M., Inoue, H., Mimori, K., 1996. Expression of MAGE genes

in human colorectal carcinoma. Ann. Surg. 224, 183–188.

Mori, M., Mimori, K., Tanaka, F., Ueo, H., Sugimachi, K.,

Akiyoshi, T., 1997. Molecular diagnosis of circulating cancer

cells using MAGE gene assays. JAMA 278, 476–477.

Nishiyama, T., Tachibana, M., Horiguchi, Y., Nakamura, K., Ikeda,

Y., Takesako, K., Murai, M., 2001. Immunotherapy of bladder

cancer using autologous dendritic cells pulsed with human lym-

phocyte antigen-A24-specific MAGE-3 peptide. Clin. Cancer

Res. 7, 23–31.

Otte, M., Zafrakas, M., Riethdorf, L., Pichlmeier, U., Loning, T.,

Janicke, F., Pantel, K., 2001. MAGE-A gene expression pattern

in primary breast cancer. Cancer Res. 61, 6682–6687.

Piva, M.G., Navaglia, F., Basso, D., Fogar, P., Roveroni, G., Gallo,

N., Zambon, C.F., Pedrazzoli, S., Plebani, M., 2000. CEA

mRNA identification in peripheral blood is feasible for color-

ectal, but not for gastric or pancreatic cancer staging. Oncology

59, 323–328.

Rimoldi, D., Salvi, S., Schultz-Thater, E., Spagnoli, G.C., Cerottini,

J.C., 2000. Anti-MAGE-3 antibody 57B and anti-MAGE-1 anti-

body 6C1 can be used to study different proteins of the MAGE-

A family. Int. J. Cancer 86, 749–751.

Russo, M., Traversari, C., Verrecchia, A., 1995. Expression of

MAGE gene family in primary and metastatic human breast

cancer: implications for tumor-specific immunotherapy. Int. J.

Cancer 64, 216–221.

Russo, V., Dalerba, P., Ricci, A., Bonazzi, C., Leone, B.E., Man-

gioni, C., Allavena, P., Bordignon, C., Traversari, C., 1996.

MAGE, BAGE and GAGE genes expression in fresh epithelial

ovarian carcinomas. Int. J. Cancer 67, 457–460.

Shichijo, S., Tsunosue, R., Masuoka, K., Natori, H., Tamai, M.,

Miyajima, J., Sagawa, K., Itoh, K., 1995. Expression of the

MAGE gene family in human lymphocytic leukemia. Cancer

Immunol. Immunother. 41, 90–103.

Tureci, O., Sahin, U., Zwick, C., Koslowski, M., Seitz, G., Pfreund-

schuh, M., 1998. Identification of a meiosis-specific protein as a

member of the class of cancer/testis antigens. Proc. Natl. Acad.

Sci. U. S. A. 95, 5211–5216.

Weynants, P., Lethe, B., Brasseur, F., Marchand, M., Boon, T.,

1994. Expression of MAGE genes by non-small cell lung car-

cinomas. Int. J. Cancer 56, 826–829.

Yamashita, N., Ishibashi, H., Hayashida, K., Kudo, J., Takenaka, K.,

Itoh, K., Niho, Y., 1996. High frequency of the MAGE-1 gene

expression in hepatocellular carcinoma. Hepatology 24, 1437–

1440.

Yoshimatsu, T., Yoshino, I., Ohgami, A., Takenoyama, M., Hana-

giri, T., Nomoto, K., Yasumoto, K., 1998. Expression of the

melanoma antigen-encoding gene in human lung cancer. J. Surg.

Oncol. 67, 126–129.


A new strategy for the diagnosis of MAGE-expressing cancers

Documents

Transcript of A new strategy for the diagnosis of MAGE-expressing cancers