Serological tumor markers have proven valuable in the care of individuals with cancer for the early detection of primary cancers,early detection of cancer relapse, monitoring the response of cancers to therapy, and as predictors of cancer prognosis. Recently, the aberrant hypermethylation of the hMLH1 promoter and its consequent transcriptional silencing has been shown to be a common event in the formation of sporadic microsatellite unstable colon cancer. The silencing of hMLH1 expression appears to be controlled by the hypermethylation of a specific region in the hMLH1 promoter. We developed a methylation-specific PCR assay that assesses this region of the hMLH1 promoter. We found that this assay is able to detect methylated hMLH1 promoter DNA in the serum of some patients with microsatellite unstable colon cancers. In a panel of sera from 19 colon cancer cases, 9 with hMLH1 promoter methylation in the tumor primary, the assay proved 33% sensitive and 100% specific. This assay offers a potential means for the serum-based detection and/or monitoring of microsatellite unstable colon cancers.

Colorectal cancer is the second most common cause of cancer mortality in the adult American population (1). Approximately 40% of affected individuals will ultimately die from their cancer. The chance of cure is greatest in those individuals whose tumor primary or tumor relapse is detected at an early stage that permits curative surgery. Early detection is particularly challenging for tumors that arise in the proximal colon, which is farthest from the rectum and requires full colonoscopy for direct inspection. One novel approach to colon cancer detection has been the attempt to detect tumor-specific DNA alterations in DNA extracted from the blood of patients with cancer (2). In particular, circulating colon cancer DNA has been identified by detection in serum of p53and K-RAS mutations, genetic alterations that are common in tumors that arise in the distal colon. However, tumors of the proximal colon commonly arise via the pathway of MSI3in which p53 and K-RAS are most commonly wild type (3, 4). Previous efforts attempting to detect the signature of MSI in DNA extracted from the serum of individuals bearing tumors of this class have been unsuccessful, suggesting a need for alternative approaches (2).

Colon cancers with MSI are initiated by inactivation of genes that encode components of the DNA mismatch repair pathway. Germ-line mutation of any member of this group of genes gives rise to a syndrome of familial MSI colon cancers, known as hereditary nonpolyposis colon cancer (5, 6). However, most commonly, MSI colon cancers arise sporadically in individuals with no familial colon cancer history. In these cases, the MSI pathway most commonly is initiated by silencing of the hMLH1 DNA repair gene via aberrant methylation of the hMLH1 promoter (7, 8, 9). The methylated hMLH1 promoter DNA is thus a tumor-specific modification that potentially could serve as a target for detection by serological assays. Because the reaction of DNA with sodium bisulfite chemically converts cytosines to uracil but does not alter methyl-cytosines, PCR primers can be designed that, following bisulfite treatment, will specifically amplify methylated but not nonmethylated target DNA templates (10). The present study was designed to determine the ability of such a methylation-specific PCR assay to detect methylated hMLH1 promoter DNA as a serum marker of MSI colon cancer in individuals with this disease.

Sample Collection and DNA Preparation.

Primary colon cancer samples and preoperative serum samples were obtained from 19 patients. The tumor samples were immediately snap frozen and prepared as described previously (11). Serum samples were prepared immediately after collection by centrifuging the sample at 2700 × g for 15 min and aspirating the serum. The serum was then stored at −80°C until use. DNA was extracted from the serum by incubating 1 ml of serum in a buffered extraction solution (140 mm Tris, 140 mm EDTA, and 0.57% SDS) with proteinase K (500μg) overnight at 50°C in a shaking water bath. The sample was then subject to phenol:chloroform/chloroform DNA extraction twice,chloroform extraction, and ethanol precipitation. The precipitated DNA was resuspended in 25 μl of distilled, deionized water and stored at−20°C. Approximately half of the sample was used in the subsequent MSP assay.

Cell Lines and Primary Tumors.

The cell lines RKO, SW480, Vaco-5, and Vaco-457 were used as known controls in the studies described below. The establishment and maintenance of the Vaco colon cancer cell lines has been described previously (11, 12). Genomic DNA from the cell lines and tumors was extracted using protocols published previously(11).

MSP of hMLH1.

MSP exploits the effect of sodium bisulfite on DNA, which efficiently converts unmethylated cytosine to uracil but which leaves methylated cytosine unchanged. Consequently, after treatment the methylated and unmethylated alleles have different sequences that can be used to design allele-specific primers.

Genomic DNA was modified with sodium bisulfite as has been described previously (10, 13). Briefly, genomic DNA from the sample(1 μg) was diluted in 50 μl of distilled water and then denatured by 0.2 m sodium hydroxide for 15 min at 37°C. Thirty μl of 10 mm hydroquinone (Sigma) and 520 μl of 3 m sodium bisulfite (Sigma) at pH 5.0 were then added, and the samples were incubated at 53°C for 18–20 h. After treatment, the DNA was purified using the Wizard DNA Clean-Up System (Promega),following the manufacturer’s protocol. The DNA was then desulfonated with 0.3 m sodium hydroxide for 10 min at room temperature,neutralized with 17 μl of 10 mNH4OAc, and then precipitated in 100% ethanol overnight at −80°C. The samples were resuspended in distilled water at a final concentration of ∼50 ng/μl and stored at −80°C for up to 8 weeks. A set of four known methylated and unmethylated control DNA samples were included in each round of bisulfite treatment.

The modified DNA was then subject to MSP using primer pairs engineered to amplify either methylated or unmethylated DNA. The primers anneal to hMLH1 promoter sequence that contains a total of five CpG dinucleotides, with at least one CpG dinucleotide placed at the 3′ end of each primer to maximize discrimination between methylated and umethylated alleles after sodium bisulfite treatment (Fig. 1 A). Furthermore, the primers were designed to recognize the“C” region in the hMLH1 promoter, the methylation status of which has been demonstrated to invariably correlate with hMLH1 expression in colon cancer (14). The methylated and unmethylated primer pair sequences were as follows: methyl specific, 5′-AACGAATTAATAGGAAGAGCGGATAGCG-3′ (sense)and 5′-CGTCCCTCCCTAAAACGACTACTACCC-3′ (antisense); and unmethyl specific, 5′-TAAAAATGAATTAATAGGAAGAGTGGATAGTG-3′ (sense) and 5′-AATCTCTTCATC CCTCCCTAAAACA-3′ (antisense). The primer pairs were purchased from Sigma-Genosys (The Woodlands, TX).

The PCR was performed in 20-μl reaction volumes containing 1× PCR buffer II (Perkin-Elmer), 1.5 mmMgCl2, 200 μm deoxynucleotide triphosphates mixture (Perkin-Elmer), 1 μm of each primer, 100–150 ng of modified DNA, and 1.5 units of Amplitaq gold DNA polymerase (Perkin-Elmer). The thermocycler conditions were as follows:methyl specific, 95°C for 10 min, (92°C for 30 s, 62°C for 30 s, 70°C for 30 s) for 39 cycles, and final extension of 70°C for 7 min; and unmethyl specific, 95°C for 10 min, (92°C for 30 s, 57°C for 30 s, 70°C for 30 s) for 39 cycles,and final extension of 70°C for 7 min. The PCR products were subject to gel electrophoresis through a 2.5% agarose gel, stained with ethidium bromide, and then visualized with UV illumination using a digital imaging system (Alpha Inotech).

To attain maximal sensitivity and specificity, primers for hMLH1 MSP were designed that, following bisulfite conversion, would selectively amplify templates derived from either methylated or unmethylated versions of sequences contained within the “C” region of the hMLH1 promoter (Ref.14; Fig. 1 A). Methylation of the “C”region, which corresponds to bp −248 to −178 relative to the transcription start, has been demonstrated to invariably correlate with hMLH1 expression in colon cancer (14). To further enhance the discrimination between products amplified by the primer sets that amplified methylated versusunmethylated hMLH1-derived templates, the primers were designed to generate PCR products that varied in size by 11 bp (102 bp versus 91 bp). The short size of these PCR products also increases the utility of these primers for determining the methylation status of the hMLH1 promoter in samples of paraffin-embedded, formalin-fixed tumor tissues.

These primers were first tested for their specificity in amplifying methylated or unmethylated hMLH1 promoter DNA by assaying four cell lines that previously had been characterized as,respectively, demonstrating an active or a methylated and silenced hMLH1 promoter. As shown in Fig. 1 B, a PCR product was obtained with the methylation-specific primer only from those cell lines shown previously by other assays to have a silenced and methylated hMLH1 promoter (8, 9).

To determine the sensitivity of this methylation-specific PCR assay for detecting methylated hMLH1, DNA from cell lines with methylated hMLH1 alleles was mixed with DNA from cell lines bearing unmethylated hMLH1 alleles. These mixtures were treated with sodium bisulfite and then subjected to the MSP assay. A minimum of 1.0% of methylated DNA could be detected with this assay. On the basis of the amount of input DNA used in these reactions, the assay successfully detected the methylated hMLH1 promoter in 1.5 ng of total genomic DNA, which is equivalent to the amount of DNA present in 100–200 cells.

This assay was then used to determine the presence or absence of hMLH1 promoter methylation in a collection of 19 colon cancers that we had previously characterized for their MSI status (data not shown) and for which matched serum samples had been obtained prior to the surgical resection of the cancer. Of 10 MSI colon cancers, 90%(n = 9 of 10) demonstrated methylation of the hMLH1 promoter. In contrast, 0 of 9 MSS colon cancer demonstrated hMLH1 promoter methylation (see Table 1 and Fig. 2).

Using the hMLH1 promoter MSP assay, we then tested the DNAs extracted from the preoperative serum samples that matched these same 19 colon cancers. Of nine individuals whose MSI colon cancers tested positive for hMLH1 promoter methylation, 33%(n = three of nine) also demonstrated a positive assay for hMLH1 promoter methylation in DNA from their preoperative serum samples (see Table 1 and Fig. 2). A positive serum result was obtained in two of two individuals found at surgery to have distant metastases (stage D); in one of three individuals found to have nodal metastases (stage C); and in none of four individuals whose tumors proved to be confined to the colon (stage B). None of 10 preoperative serum samples tested positive in the cases of individuals whose tumors lacked hMLH1 promoter methylation (Table 1). Individuals whose tumors lacked hMLH1 promoter methylation were similar to those whose tumors demonstrated hMLH1promoter methylation with regard to stage, location of the primary tumors, and age of the patients at diagnosis. Thus in this sample set,the assay of serum DNA for hMLH1 promoter methylation demonstrated a sensitivity of 33% and a specificity of 100% for the detection of colon cancers bearing this molecular alteration.

The detection of methylated hMLH1 promoter DNA in the serum of a group of patients with MSI colon cancers contrasts with the previous inability to detect microsatellite alterations in serum-DNA samples from a similar group of MSI colon cancer patients(2). This difference likely reflects the fact that PCR amplification of the methylation-specific hMLH1 promoter product from bisulfite-treated DNA eliminates the background signal from normal DNAs. In contrast, sensitivity for detecting in serum the characteristic microsatellite shifts of the DNA from MSI tumors is likely compromised by a considerable background of the normal microsatellite alleles that would also be expected to be amplified from normal DNA also present in serum. Because MSI cancers are in general wild type for p53 and K-RAS, the serological detection of methylated hMLH1 promoter DNA provides a potential unique serological marker for this subset of cancers.

The sensitivity of 33% that we find for the detection of methylated hMLH1 promoter DNA in serum is similar to the sensitivity reported for detecting methylated p16 promoter DNA in the serum of patients with non-small cell lung cancer, head and neck cancer, and breast cancer (15, 16, 17) but is less than the sensitivity reported for the detection of methylated p16 DNA in the serum of patients with hepatocellular carcinomas that carry methylated p16 (80%) and less than the sensitivity reported for detection of p53 mutations in the serum of patients with p53 mutant colon cancers (70%; Refs. 2 and18). These differences suggest that future technical advances may permit additional improvement in the sensitivity of serological detection of methylated hMLH1 promoter DNA. This would potentially allow application of the assay for early detection of hMLH1 methylated colon cancers, which usually occur in the proximal (right-sided) colon and so are occult to early detection by such standard screening modalities as flexible sigmoidoscopy. Detection of hMLH1 promoter methylation in serum could thus play a role in cancer detection as part of a panel of complementary serological markers that included assays for p53 mutation, K-RAS mutation, and methylated p16. Moreover, our apparent current sensitivity for detecting later stage MSI colon cancers suggests that as currently constituted, serological assay of hMLH1 promoter methylation could have potential application for the early detection of postsurgical relapse of hMLH1methylated cancers. Additionally, our observations suggest the hypothesis that serological detection of hMLH1 promoter methylation may also be of potential use as a prognostic marker for discriminating Dukes’ C patients at high risk for distant metastases,who could benefit from postoperative chemotherapy, versusDukes’ C patients at lower risk for disseminated disease. Lastly, we cannot formally exclude that the methylated hMLH1 DNA detected in serum might in some individuals reflect a constitutional susceptibility for aberrant methylation of this locus that extends to both gastrointestinal and lymphoid cells. Such questions will provide fruitful material for future studies.

Fig. 1.

A, location of primers for hMLH1 MSP assay in relation to the hMLH1promoter C region identified by Deng et al.(14). Primers cover 5 CpG dinucleotides in this region and are designed to specifically amplify sodium bisulfite-converted sequences arising from either methylated (methyl) or unmethylated (unmethyl) templates. The numbering scheme is based on the GenBank sequence entry U83845. B,results of hMLH1 MSP assay using primers that amplify methylated (M) or unmethylated (U)alleles specifically. M1 and M2 are DNA samples from cell lines V5 and RKO, respectively, shown previously to have only methylated hMLH1 alleles, whereas U1 and U2 are DNA samples from cell lines SW480 and V457, respectively, that bear only unmethylated hMLH1 alleles (9, 14).

Fig. 1.

A, location of primers for hMLH1 MSP assay in relation to the hMLH1promoter C region identified by Deng et al.(14). Primers cover 5 CpG dinucleotides in this region and are designed to specifically amplify sodium bisulfite-converted sequences arising from either methylated (methyl) or unmethylated (unmethyl) templates. The numbering scheme is based on the GenBank sequence entry U83845. B,results of hMLH1 MSP assay using primers that amplify methylated (M) or unmethylated (U)alleles specifically. M1 and M2 are DNA samples from cell lines V5 and RKO, respectively, shown previously to have only methylated hMLH1 alleles, whereas U1 and U2 are DNA samples from cell lines SW480 and V457, respectively, that bear only unmethylated hMLH1 alleles (9, 14).

Close modal
Fig. 2.

Results of hMLH1 MSP from colon cancers and paired serum samples. M and U, assays specific for amplification of the methylated versusunmethylated, respectively, hMLH1 promoter. The unmethylated alleles detected in tumor samples likely represent contaminating normal tissue elements and in serum samples likely arise from DNA released from circulating leukocytes.

Fig. 2.

Results of hMLH1 MSP from colon cancers and paired serum samples. M and U, assays specific for amplification of the methylated versusunmethylated, respectively, hMLH1 promoter. The unmethylated alleles detected in tumor samples likely represent contaminating normal tissue elements and in serum samples likely arise from DNA released from circulating leukocytes.

Close modal

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

Supported by USPHS Grants KO8 CA77676 (to W. G.), RO1 CA67409, RO1 CA72160 (to S. D. M.), P30 CA43703,P30CA16058, CA8485, and DK20593. S. D. M. is an associate investigator of the Howard Hughes Medical Institute.

3

The abbreviations used are: MSI, microsatellite instability; MSS, microsatellite stable; MSP, methylation-specific PCR.

Table 1

Results of hMLH1 MSP in tumors and serum of colon cancer patients

Description of patient and tumor characteristics compared with results of hMLH1 MS-PCR analysis of matched tumor and serum samples.

Patient no.hMLH1 Tumor DNA/Serum DNAaMSI statusStageGrade
M /M MSI Poor 
M /M MSI Moderate 
M /M MSI Poor 
M /U MSI Moderate 
M /U MSI Moderate 
M /U MSI B2 Moderate 
M /U MSI Moderate 
M /U MSI Moderate 
M /U MSI Moderate 
10 U /U MSS Moderate 
11 U /U MSS Moderate 
12 U /U MSS Poor 
13 U /U MSS Moderate 
14 U /U MSS Moderate 
15 U /U MSS Moderate 
16 U /U MSS Moderate 
17 U /U MSI Moderate 
18 U /U MSS Moderate 
19 U /U MSS HGDb HGD 
Patient no.hMLH1 Tumor DNA/Serum DNAaMSI statusStageGrade
M /M MSI Poor 
M /M MSI Moderate 
M /M MSI Poor 
M /U MSI Moderate 
M /U MSI Moderate 
M /U MSI B2 Moderate 
M /U MSI Moderate 
M /U MSI Moderate 
M /U MSI Moderate 
10 U /U MSS Moderate 
11 U /U MSS Moderate 
12 U /U MSS Poor 
13 U /U MSS Moderate 
14 U /U MSS Moderate 
15 U /U MSS Moderate 
16 U /U MSS Moderate 
17 U /U MSI Moderate 
18 U /U MSS Moderate 
19 U /U MSS HGDb HGD 
a

M, methylated hMLH1 promoter; U, unmethylated hMLH1 promoter.

b

HGD, high-grade dysplasia in adenoma.

1
Landis S., Murray T., Bolden S., Wingo P. Cancer Statistics..
CA Cancer J. Clin.
,
49
:
8
-31,  
1999
.
2
Hibi K., Robinson C., Booker S., Wu L., Hamilton S., Sidransky D., Jen J. Molecular detection of genetic alterations in the serum of colorectal cancer patients..
Cancer Res.
,
58
:
1405
-1407,  
1998
.
3
Iacopetta B., Welch J., Soong R., House A., Zhou X., Hamelin R. Mutation of the transforming growth factor-β type II receptor gene in right-sided colorectal cancer: relationship to clinicopathological features and genetic alterations..
J. Pathol.
,
184
:
390
-395,  
1998
.
4
Laurent-Puig P., Blons H., Cugnenc P. Sequence of molecular genetic events in colorectal tumorigenesis..
Eur. J. Cancer Prev.
,
9(Suppl.)
:
S39
-S47,  
1999
.
5
Marra G., Boland C. Hereditary nonpolyposis colorectal cancer: the syndrome, the genes, and historical perspective..
J. Natl. Cancer Inst.
,
87
:
1114
-1125,  
1995
.
6
Lynch H., Smyrk T. Hereditary nonpolyposis colorectal cancer (Lynch syndrome): an updated review..
Cancer (Phila.)
,
78
:
1149
-1167,  
1996
.
7
Kane M., Loda M., Gaida G., Lipman J., Mishra R., Goldman H., Jessup J., Kolodner R. Methylation of the hMLH1 promoter correlates with lack of expression of hMLH1 in sporadic colon tumors and mismatch repair-defective human tumor cell lines..
Cancer Res.
,
57
:
808
-811,  
1997
.
8
Herman J., Umar A., Polyak K., Graff J., Ahujia N., Issa J-P. J., Markowitz, S., Willson, J., Hamilton, S., Kinzler, K., Kane, M., Kolodner, R., Vogelstein, B., Kunkel, T., and Baylin, S..
Incidence and functional consequences of hMLH1 promoter hypermethylation in colorectal carcinoma. Proc. Natl. Acad. Sci. USA
,
95
:
6870
-6875,  
1998
.
9
Veigl M., Kasturi L., Olechnowicz J., Ma A., Lutterbaugh J., Periyasamy S., Li G-M., Drummond J., Modrich P., Sedwick D., Markowitz S. Biallelic inactivation of hMLH1 by epigenetic gene silencing, a novel mechanism causing human MSI cancers..
Proc. Natl. Acad. Sci. USA
,
95
:
8698
-8702,  
1998
.
10
Herman J., Graff J., Myohanen S., Nelkin B., Baylin S. Methylation-specific PCR..
a novel PCR assay for methylation status of CpG islands. Proc. Natl. Acad. Sci. USA
,
93
:
9821
-9826,  
1996
.
11
Markowitz S., Wang J., Myeroff L., Parsons R., Sun L., Lutterbaugh J., Fan R., Zborowska E., Kinzler K., Vogelstein B., Brattain M., Willson J. Inactivation of the type II TGF-β receptor in colon cancer cells with microsatellite instability..
Science (Washington DC)
,
268
:
1336
-1338,  
1995
.
12
Willson J., Bittner G., Oberley T., Meisner G., Weese J. Cell culture of human colon adenomas and carcinomas..
Cancer Res.
,
47
:
2704
-2713,  
1987
.
13
Herman J., Baylin S. Methylation-specific PCR. Dracopoli N. C. eds. .
Current Protocols in Human Genetics
,
2
:
10.6
John Wiley & Sons 1–10.6.10. New York  
1998
.
14
Deng G., Chen A., Hong J., Chae H., Kim Y. Methylation of CpG in a small region of the hMLH1 promoter invariably correlates with the absence of gene expression..
Cancer Res.
,
59
:
2029
-2033,  
1999
.
15
Esteller M., Sanchez-Cespedes M., Rosell R., Sidransky D., Baylin S., Herman J. Detection of aberrant promoter hypermethylation of tumor suppressor genes in serum DNA from non-small cell lung cancer patients..
Cancer Res.
,
59
:
67
-70,  
1999
.
16
Silva J., Dominguez G., Villanueva M., Gonzalez R., Garcia J., Corbacho C., Provencio M., Espana P., Bonilla F. Aberrant DNA methylation of the p16ink4A gene in plasma of breast cancer patients..
Br. J. Cancer
,
80
:
1262
-1264,  
1999
.
17
Sanchez-Cespedes M., Esterller M., Wu L., Nawroz-Danish H., Yoo G., Koch W., Jen J., Herman J., Sidransky D. Gene promoter hypermethylation in tumors and serum of head and neck cancer patients..
Cancer Res.
,
60
:
892
-895,  
2000
.
18
Wong I., Lo Y., Zhang J., Liew C-T., Ng M., Wong N., Lai P., Lau W., Hjelm N., Johnson P. Detection of aberrant p16 methylation in the plasma and serum of liver cancer patients..
Cancer Res.
,
59
:
71
-73,  
1999
.