Journal of Surgical Oncology 2008;98:377–383

Expression of Mismatch Repair Proteins, hMLH1/hMSH2, in Non-Small Cell

Lung Cancer Tissues and Its Clinical Significance

HIDENORI KOUSO, MD,* ICHIRO YOSHINO, MD, NAOKO MIURA, MD, TOMOYOSHI TAKENAKA, MD,TARO OHBA, MD, TOMOFUMI YOHENA, MD, ATSUSHI OSOEGAWA, MD, FUMIHIRO SHOJI, MD,

AND YOSHIHIKO MAEHARA, MD

Department of Surgery and Science, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan

Background: hMLH1 and hMSH2 have been implicated to be involved in the DNA mismatch repair (MMR) system. The purpose of this study is

to investigate the expression of hMLH1 and hMSH2 DNA MMR proteins in non-small cell lung cancer (NSCLC) tissue and to elucidate their

clinical significance.

Methods: The hMLH1 and hMSH2 protein expression was evaluated by immunohistochemistry for a consecutive series of 113 NSCLC patients.

The expressions of each protein were examined for an association with the clinicopathological variables, including genetic alterations analyzed by

high resolution fluorescent microsatellite analysis.

Results: Regarding the hMLH1 expression, the MSI-positive patients showed significantly lower scores than the MSI-negative patients. For

hMSH2 expression, the patients with a 20 or higher pack-year index (PYI) showed significantly higher scores than the patients with a PYI less

than 20. The expression status of proteins did not affect both the disease free and overall survival of the patients. No significant correlation was

observed among the scores for the proteins.

Conclusions: The expressions of hMLH1 and hMSH2 are independently regulated and play different roles in NSCLC. The genetic instability is

possibly due to the reduced expression of hMLH1 protein, and hMSH2 expression is associated with smoking status.

J. Surg. Oncol. 2008;98:377–383. � 2008 Wiley-Liss, Inc.

KEY WORDS: hMLH1; hMSH2; microsatellite instability; mismatch repair; non-small cell lung cancer

INTRODUCTION

Lung cancer is one of the most prevalent cancers in the world, and

its mortality rate is significantly associated with the consumption of

cigarettes [1]. The main factor in the process of lung carcinogenesis is

the accumulation of genetic alterations caused by the toxic compounds

contained in cigarette smoke [2]. Cigarette smoke contains approxi-

mately 4,000 chemicals including around 40 carcinogens [3], some of

which generate DNA adducts in the respiratory cells. During DNA

replication, these lesions form DNA mismatches, which have to be

repaired by various systems, otherwise resulting in genetic mutations

[4–6]. Of the various DNA repair systems, the mismatch repair

(MMR) system is essential to maintain the stability of the genome

during repeated duplication. Its primary function is to eliminate any

single-base mismatches and insertion-deletion loops that may arise

during DNA replication [7]. In clinical lung cancer, A to T conversion

of p53 mutations and K-ras codon 12 mutation are closely related to

smoking status and microsatellite instability of the genome [8,9].

Therefore, the MMR system may play a crucial role in lung

carcinogenesis.

The MMR system is composed of a few well-conserved proteins.

The essential components of the MMR system, MutS, MutL, MutH and

Uvr, were identified in Escherichia coli [10,11]. In addition, all

eukaryotic organisms, including yeast, mice and humans, have MutS

homologs and MutL homologs [12]. Genetic and biochemical studies

have indicated that MSH2, one of the MutS homolog proteins, is

required for all mismatch corrections in nuclear DNA, whereas MSH3

and MSH6, which are also MutS homolog proteins, are required for the

repair of some distinct and overlapping types of mismatched DNA

during replication [13]. These three MutS homologs make two

heterodimers, MutSa (MSH2/MSH6) and MutSb (MSH2/ MSH3),

which recognize the mismatched DNA in eukaryotic MMR [7,14,15].

Although not as well studied as the MutS homologs, the eukaryotic

MutL homologs also appear to function as heterodimers. The best

characterization of these in humans has been MutLa, a MLH1/PMS2

heterodimer, which is capable of supporting the repair initiated by

MutSa or MutSb [15–18].

The loss of this system leads to carcinogenesis by accelerating the

accumulation of mutations in the oncogenes and tumor suppressor

genes [19,20], and, in general, such a disorder of the MMR system is

detected by gains or losses of short repeat units within the microsatel-

lite sequences, which is also known as microsatellite instability (MSI)

[9]. MSI is in fact detected in various clinical malignancies, including

colorectal, stomach, endometrial, and lung malignancies [21].

Regarding the mechanisms of the dysfunction of the MMR system,

germ line mutations of the human MMR genes, such as hMLH1 and

hMSH2, cause susceptibility to hereditary nonpolyposis colon cancer

(HNPCC) [7,22,23], whereas hypermethylation of the promoters of

these genes is considered to be another cause of the loss of the MMR

system in various sporadic cancers [24–27]. The inactivation of the

MMR genes is generally associated with the loss of immunohisto-

chemical expression of the corresponding protein [28].

The hMLH1 and hMSH2 DNA MMR proteins are known to be

implicated in human cancer, with colon cancer being the most well-

studied model. However, information on the status of these two

*Correspondence to: Hidenori Kouso, MD, Department of Surgery andScience, Graduate School of Medical Sciences, Kyushu University,Maidashi 3-1-1, Higashi-ku, Fukuoka 812-8582, Japan. Fax: 81-92-642-5482. E-mail: k-nori@surg2.med.kyushu-u.ac.jp

Received 15 November 2007; Accepted 20 May 2008

DOI 10.1002/jso.21108

Published online 21 July 2008 in Wiley InterScience(www.interscience.wiley.com).

� 2008 Wiley-Liss, Inc.

proteins in lung cancer is limited. In this study, we investigated the

expression of hMLH1 and hMSH2 DNA MMR proteins in non-small

cell lung cancer (NSCLC) tissues and their clinical significance.

MATERIALS AND METHODS

Patients

A consecutive series of 113 patients with NSCLC who underwent

surgical resections at Kyushu University Hospital between 2000 and

2003 were analyzed in this study. The histological diagnosis of the

tumors was based on criteria of the World Health Organization [29],

and the TNM stage was determined according to criteria revised in

1997 [30]. The age of the patients ranged from 36 to 84 years (mean of

68.8 years) and included 71 men and 42 women; 72 stage I, 15 stage II,

24 stage III and 2 stage IV; 46 with a pack-year index (PYI) of less than

20, and 67 with a PYI of 20 or more; 89 adenocarcinomas, 20

squamous cell carcinomas and 4 others (Table I). Written informed

consent was obtained from each patient for this study. The institutional

review board of our university gave approval for this study.

Follow-Up of Patients

In order to investigate the overall survival and the disease free

survival of the patients who underwent complete resections, a follow-

up examination was prospectively performed every 2 months for the

first 2 years and thereafter every 3–4 months. The mean follow-up

period of the patients was 37.8 months (median, 43.4 months; range,

0.4–75.2). The follow-up included a physical examination, complete

blood count, blood chemistry, and chest radiography. The majority of

the patients underwent CT and a radionuclide bone scan twice a year, at

the time when symptoms related to recurrence appeared. For recurrent

disease, feasible treatments such as systemic chemotherapy, radio-

therapy or surgery were performed.

High Resolution Fluorescent Microsatellite

Analysis (HRFMA)

Five dinucleotide microsatellites, D2S123, D5S107, D10S197,

D11S904 and D13S175, were used as markers for the analysis of MSI

and LOH [31]. Using genomic DNA derived from the tissue

specimens, the five microsatellite sequences were amplified by

polymerase chain reaction (PCR). Oligonucleotide primers corre-

sponding to the microsatellite sequences [31] were then synthesized

and purified using high performance liquid chromatography. 50 primers

were labeled with the fluorescent compound, ROX (6-carboxy-x-

rhodamine) or HEX (6-carboxy-20,40,70,4,7-hexachloro-fluorescein).

PCR reactions were performed using TAKARA Taq Reagent Kits

(TAKARA Co. Ltd., Tokyo, Japan) and analyzed using the Perkin–

Elmer GeneAmp PCR system 9600 or 2400 (Norwalk, USA). Each

50 ml reaction mixture contained 1� reaction buffer, 350 mM of each

dNTP, 10 pmol of each primer, 2.5 U of the polymerase, and 25 mg of

the genomic DNA. The thermal conditions of the system were as

follows: one cycle at 958C for 4 min, 35 cycles at 958C for 0.5 min,

558C for 0.5 min, 728C for 0.5 min, one cycle at 728C for 10 min. Next,

0.5 U of T4 DNA polymerase was added to the mixture, followed by

incubation at 378C for 10 min. Each 1.5 ml product was mixed with

0.5 ml loading buffer (blue dextran, 25 mM EDTA), 2.5 ml of

formamide, and 0.5 ml of dH2O. In order to compare the

electrophoretic profiles of the two samples, 1.2 ml of ROX-labeled

product and 0.3 ml of HEX-labeled product were mixed together. The

samples were denatured and loaded onto the ABI 373A sequencer

(Applied Biosystems, Foster City, CA). In each case, a size marker

labeled with TMRA (N,N,N0,N0-tetramethyl-6-carboxyrhodamine) was

always electrophoresed in each lane in order to standardize the

mobility of the sample. The running conditions were 1500 V, 20 mA,

and 30 W for 5.5 h. The data were processed using the ABI software

package from GeneScan (Applied Biosystems). The method of

Journal of Surgical Oncology

TABLE I. Demographics of the Patients in This Study

Factors Number of cases

Age

Median 66.8 years

Range 36–84

Sex

Male 71

Female 42

Pack-year indexa

Median 33.9

Range 0–156

20 or more 67

Under 20 46

Tumor histology

Adenocarcinoma 89

Squamous cell carcinoma 20

Others 4

Pathological stage

IA 49

IB 23

II 15

III 24

IV 2

MSI

Positive 9

Negative 103

LOH

Positive 61

Negative 52

aCalculated by the number of packs of cigarettes consumed a day� years.

Fig. 1. Analysis of microsatellite instability and loss of hetero-zygosity by high resolution fluorescent system. (A) A case judgedpositive for microsatellite instability. (B) A case judged positive forloss of heterozygosity.

378 Kouso et al.

detecting MSI and LOH used here has been described previously [31].

Briefly, for the detection of MSI, the y-axis corresponds to length of

DNA fragment and fluorescence of y-axis is longer than that of the

normal control when the MSI occurred in the amplified region of the

genomic DNA derived from the tumor tissue. For the detection of

LOH, the fluorescence of a peak is decreased more than 30% of that of

the normal control when the LOH occurred (Fig. 1). MSI in at least one

locus was detected in 9 cases and LOH in at least 1 locus was detected

in 61 cases (Table I).

Immunohistochemistry

An immunohistochemical study was carried out on the formalin-

fixed, paraffin-embedded tissue sections. Four-micrometer sections

were deparaffinized with xylene and then rehydrated in a series of

ethanols. Heat-induced epitope retrieval was performed in 0.1 M of

NaOH-citrate buffer (pH 7.0) for hMLH1 and hMSH2 immunostain-

ing, and the samples were heated in an autoclave at 1218C for 15 min.

Endogenous peroxidase was blocked at room temperature using 3%

hydrogen peroxide in methanol for 30 min. After blocking with normal

goat serum, the slides were incubated with mouse monoclonal

antibody against hMLH1 (BD Bioscience, Franklin Lakes, NJ) and

hMSH2 (CALBIOCHEM, Germany) using a dilution of 1:50 and

1:200 at 48C overnight, respectively. After washing the sections, they

were treated for 60 min at room temperature with goat-antimouse

immunoglobulin. Staining for hMLH1 and hMSH2 was completed

using the streptavidin–biotin–peroxidase complex method (DAKO,

Denmark) with diaminobenzidine as a chromogen, and then the slides

were counterstained with hematoxylin. The normal staining pattern for

both hMLH1 and hMSH2 was localized at the nuclear portion. The

tumor cells that exhibited an absence of nuclear staining in the

presence of non-neoplastic cells with nuclear staining (with normal

bronchial epithelium as an internal control) were considered to have an

abnormal pattern. According to the staining intensity and the number

of positive cells, the immunoreactivity score (IRS) was assessed using

a previously described procedure [32,33]. In short, the IRS was

determined by multiplication of the values for the grade of intensity

(0: no staining; 1: weak staining; 2: moderate staining; 3: strong

staining) and the values for the grade of the number of positive cells

(1: 0–10%; 2:11–50%; 3: 51–80%; 4: >80%) (Fig. 2). The IRS was

determined by five researchers, including two pathologists, who did not

know any clinical information for the patients.

Statistical Analysis

The relationships among the clinicopathological factors and

hMLH1/hMSH2 expression were analyzed by unpaired t-test. The

correlation between hMLH1 and hMSH2 were examined by using a

Journal of Surgical Oncology

Fig. 2. Immunohistochemical staining patterns for hMLH1 and hMSH2 in NSCLC. (A) A case with the grade for staining intensity of hMLH11, the grade for the number of positive cells of hMLH1 1 and IRS of hMLH1 1. (B) A case with the grade for staining intensity of hMLH1 3, thegrade for the number of positive cells of hMLH1 4 and IRS of hMLH1 12. (C) A case with the grade for staining intensity of hMSH2 2, the gradefor the number of positive cells of hMSH2 1 and IRS of hMSH2 2. (D) A case with the grade for staining intensity of hMSH2 3, the grade forthe number of positive cells of hMSH2 4 and IRS of hMSH2 12.

Mismatch Repair Protein in NSCLC 379

single regression analysis. The survival curves were prepared by the

Kaplan–Meier method, and comparisons among the survival curves

were made using the log rank test. The data were considered to be

significant when the P value was less than 0.05. All of the analyses were

performed with using the Stat View software program, version 5.0.

RESULTS

Relationship Between hMLH1/hMSH2 Protein

Expression and Clinicopathological Factors

The histograms of the IRS of hMLH1/hMSH2 are shown in

Figure 3. The mean IRS of hMLH1 was 7.0� 3.5, ranging from 0 to

12 and one of hMSH2 was 7.1� 2.6, ranging from 2 to 12. The

relationships between the each IRS and the clinicopathological factors

were examined (Tables II and III). The MSI-positive patients showed a

significantly lower IRS of hMLH1 (3.8) than the MSI-negative patients

(7.3) (P¼ 0.0034). Although there was no relationship between the

IRS of hMSH2 and MSI, the patients who had a 20 or higher PYI

showed a significantly higher IRS (7.6) than the patients who had a PYI

less than 20 (6.4) (P¼ 0.0189). No relationship was found regarding

the other clinicopathological factors.

The Expression Status of hMLH1/hMSH2 and the

Survival of Patients

In order to investigate whether the expression of hMLH1 and

hMSH2 are associated with the malignant potential of lung cancer, the

expression status of the both proteins and the overall survival of the 95

patients who underwent complete resection were analyzed. The 48

patients with an IRS of hMLH1 greater than 7 showed a similar 5-year

survival rate with the 47 patients having an IRS of hMLH1 less than 7

(80% vs. 90%, P¼ 0.234). Also regarding hMSH2, the 42 patients with

an IRS greater than 7 and the 53 patients with an IRS less than 7

showed comparable 5-year survival rate (80% and 89%, P¼ 0.396).

Though the disease free survival was also examined, no statistical

difference was observed between the expression status of both proteins

and the disease free survival (P¼ 0.314 and 0.663, respectively)

(Fig. 4A–D).

Relationship Among the Expression Status of hMLH1

and hMSH2

In order to examine the relationship between hMLH1 and hMSH2

protein expression, the cases were divided into four groups: both of the

higher expression groups (both IRS for hMLH1/hMSH2 were over 7;

n¼ 32), the only hMLH1 higher expression group (only the IRS of

hMLH1 was over 7; n¼ 23), the only hMSH2 higher expression group

(only the IRS of hMSH2 was over 7; n¼ 22), and both of the lower

expression groups (neither IRS of hMLH1/hMSH2 was less than 7;

n¼ 36). A venn diagram of the population according to hMLH1/

hMSH2 expression is shown in Figure 5A. The analysis of the

clinicopathological factors, such as gene alterations, smoking status,

cell type and pathologic stage, revealed no significant features for

each subgroup. A regression analysis also revealed no relationship

between the two protein expressions (Fig. 5B). These results may

indicate that there is no relationship between both mechanisms of

expression of the two proteins.

DISCUSSION

The human DNA MMR system is essential to reduce the

accumulation of mutations and to maintain genomic stability. In new

DNA replication, base pair geometry and the nature of the DNA

Journal of Surgical Oncology

Fig. 3. Histograms for the IRS of hMLH1 and hMSH2. (A)Histogram of the IRS of hMLH1. (B) Histogram of the IRS of hMSH2.

TABLE II. Relationship Between the IRS of hMLH1 and

Clinicopathological Factors

Factors

Number of

cases

Average of

IRS P-value

MSI [þ]/[�]a 9/104 3.778/7.288 0.0034

LOH [þ]/[�]a 61/52 7.033/6.981 0.9376

PYIb 20 or more/under 20 67/46 7.000/7.022 0.9743

Adenocarcinoma/others 89/24 6.955/7.208 0.7543

pStage I/others 72/41 6.917/7.171 0.7121

aDetermined [þ] when MSI or LOH was detected in at least one locus.bPYI: pack-year index was calculated by the number of packs of cigarettes

consumed per day� years.

TABLE III. Relationship Between the IRS of hMSH2 and

Clinicopathological Factors

Factors

Number of

cases

Average of

IRS P-value

MSI [þ]/[�]a 9/104 7.889/7.067 0.3717

LOH [þ]/[�]a 61/52 7.295/6.942 0.4804

PYI 20 or more/under 20 67/46 7.612/6.435 0.0189

Adenocarcinoma/others 89/24 6.989/7.667 0.2650

pStage I/others 72/41 7.366/7.000 0.4802

aDetermined [þ] when MSI or LOH was detected in at least one locus.

380 Kouso et al.

polymerase involved result in an error rate of 10�4 to 10�5 at the

nucleotide insertion step of DNA synthesis [34], and the proofreading

exonuclease associated with some DNA polymerases edits this

mistake. Mistakes that escape these fidelity devices are corrected by

the MMR system, further elevating the fidelity by 50–1,000-fold [15].

These replication errors are particularly evident in the microsatellite

sequences, consisting of repeats of 1–4 base pairs. If the MMR system

is inactivated, the errors are fixed as a mutation, such as the addition or

deletion of one or more repeat units, after another next round of

replication. These altered microsatellite sequences are called micro-

satellite instability and are considered to reflect a cellular MMR

deficiency [35].

In the ubiquitous process of MMR, the hMLH1 and hMSH2

proteins have different roles but act serially. The hMSH2 protein

works to recognize mismatched DNA, whereas the hMLH1 protein

coordinates the interplay among the mismatch recognition complex

and other involved proteins [7]. However, in clinical cancer, these

protein expressions are not correlated well with each other. Xinarianos

et al. [36] reported that the comparative analysis of hMLH1 and

hMSH2 protein expression in the NSCLC tissue did not reveal any

associations between the expressions of these two genes, and that

the tumor specimens with a combined reduced expression of both

the hMLH1 and hMSH2 proteins did not correlate with any of the

clinicopathological parameters. In this study, we examined the

expression status of both the hMLH1 and hMSH2 in 113 surgical

specimens of NSCLC, and examined them for a relationship with

the clinicopathological parameters and microsatellite alterations of the

DNA. The expressions of the hMLH1 and hMSH2 proteins were

independent of an association with smoking status with hMSH2 but not

with hMLH1, which was determined by a regression analysis among

the IRSs of the two proteins. Moreover, only the hMLH1 expression

showed a significant relationship with the presence of MSI but not of

hMSH2. It is thus likely that the clinical significance of the hMLH1

and hMSH2 expressions is quite different in lung cancers.

In several types of malignancies, the protein expressions of hMLH1

and/or hMSH2 have been demonstrated to be significantly associated

with a DNA mismatch repair deficiency and MSI in the stomach,

endometrial and lung [37–42]. However, which of the proteins is

essential to the presence of MSI remains to be clarified. In this study,

the IRS of hMLH1 was expectedly lower in the MSI-positive cases

(P¼ 0.0034), whereas there was no statistical difference in the

relationship between hMSH2 and MSI. Chang et al. [42] also reported

a relationship to exist between MSI and a reduced hMLH1 protein

expression, but not with hMSH2 protein expression in NSCLC. In

an experiment of cancer cell lines with MSI, the reversal of

methylation by a demethylating agent, 5-azacytidine, not only induced

Journal of Surgical Oncology

Fig. 4. Correlation of the IRS of hMLH1 and hMSH2 with the prognosis of patients. No statistical difference was observed between theexpression status of hMLH1/hMSH2 and both the disease free and overall survival.

Mismatch Repair Protein in NSCLC 381

the reexpression of the hMLH1 protein, but also restored the MMR

capacity in the MMR-deficient cell line [24,25]. These results suggest

that isolated hMLH1 deficiency leads to MSI.

Although germ line mutations in the MMR genes causes MSI in

HNPCC and sporadic colorectal cancer with MSI, there are a subset of

sporadic colorectal cancers with MSI which have been identified as

having no mutations of the MMR genes [43]. The hypermethylation of

the promoter of hMLH1 also causes the inactivation of the hMLH1

protein expression, and consequently, MSI is caused in most sporadic

cancers including colorectal, stomach, and ovarian carcinoma [24–26].

In NSCLC, hypermethylation of the promoter of the hMLH1 gene was

reported to be significantly associated with the reduction in protein

expression [27]. This report also indicated a significant relationship

between MSI and hypermethylation of the promoter of hMLH1.

Hence, in sporadic cancers, hypermethylation of the promoter of

hMLH1 causes alteration of its protein expression and MSI. Therefore,

hypermethylation of the promoter of hMLH1 may be a meaningful

factor for carcinogenesis. In NSCLC, the regulation of hMLH1 protein

expression, possibly through gene silencing, may have a valuable role

in lung carcinogenesis.

It should be noted that the IRS of hMSH2 was higher in the over-20

PYI cases than in the less-than-20 PYI cases (P¼ 0.0189). The

chemical compounds included in tobacco smoke, such as aromatic

hydrocarbons, frequently cause genetic alterations through the

formation of DNA adducts [44]. On the other hand, inactivation of

the MMR genes promotes carcinogenesis by accelerating the

accumulation of mutations in many oncogenes and tumor suppressor

genes, including p53, K-ras, APC, TGFbIIR, and BAX [20,45–48].

In clinical lung cancer, A to T conversion of p53 mutations and K-ras

codon 12 mutation are closely related to smoking status and an

unfavorable prognosis [8,9]. The MMR system may play a significant

role in overcoming the crisis of the respiratory epithelial cells;

otherwise, the reduced expression of the MMR proteins may result in a

susceptibility to lung carcinogenesis. The present results showed the

close relationship between the hMSH2 expression in lung cancer

tissues and the smoking status of the corresponding patients, and

supported the above hypothesis.

Scartozzi et al. [49] demonstrated that a lower expression of

hMLH1 in the tumor cells was associated with a favorable survival in

patients with NSCLC. However, Skarda et al. [50] reported no

relationships to exist between the altered protein expression of

hMLH1/hMSH2 and both the disease free and overall survival in

NSCLC patients. In this study, the expression status of both hMLH1

and hMSH2 showed no significant impact on the disease free survival

or overall survival of patients. In the present cohort, the mean follow-

up period of the patients was 37.8 months and 72 patients (64%) were

stage I; therefore, it may be difficult to detect survival differences.

CONCLUSION

The genetic instability of NSCLC is due to the reduced expression

of the hMLH1 protein. The hMSH2 expression is increased by

smoking stimulation, and it is associated with the repair of gene

damage due to smoking-related chemical substances. Further investi-

gation is required to elucidate the complete pathways of activation and

inactivation of the expression of these two proteins in NSCLC.

REFERENCES

1. World Health Organization. World Health Statistics Annual.Geneva: WHO; 1991.

2. Phillips DH: Smoking-related DNA and protein adducts in humantissues. Carcinogenesis 2002;23:1979–2004.

3. Hoffmann D, Hecht SS: Advances in tobacco carcinogenesis. In:Cooper CS, Grover PL, editors. Chemical Carcinogenesis andMutagenesis I, Springer-Verlag Berlin: 1990; 63–102.

4. Doherty KM, Sharma S, Uzdilla LA, et al.: RECQ1 helicaseinteracts with human mismatch repair factors that regulate geneticrecombination. J Biol Chem 2005;280:28085–28094.

5. Elliott B, Jasin M: Repair of double-strand breaks by homologousrecombination in mismatch repair-defective mammalian cells.Mol Cell Biol 2001;21:2671–2682.

6. Mu D, Tursun M, Duckett DR, et al.: Recognition and repair ofcompound DNA lesions (base damage and mismatch) by humanmismatch repair and excision repair systems. Mol Cell Biol1997;17:760–769.

7. Peltomaki P: Role of DNA mismatch repair defects in thepathogenesis of human cancer. J Clin Oncol 2003;21:1174–1179.

8. Levine AJ, Momand J, Finlay CA: The p53 tumour suppressorgene. Nature 1991;351:453–456.

9. Slebos RJ, Hruban RH, Dalesio O, et al.: Relationship between K-ras oncogene activation and smoking in adenocarcinoma of thehuman lung. J Natl Cancer Inst 1991;83:1024–1027.

10. Modrich P, Lahue R: Mismatch repair in replication fidelity,genetic recombination, and cancer biology. Annu Rev Biochem1996;65:101–133.

11. Wang H, Yang Y, Schofield MJ, et al.: DNA bending andunbending by MutS govern mismatch recognition and specificity.Proc Natl Acad Sci U S A 2003;100:14822–14827.

12. Jun SH, Kim TG, Ban C: DNA mismatch repair system. Classicaland fresh roles. FEBS J 2006;273:1609–1619.

13. Marsischky GT, Filosi N, Kane MF, et al.: Redundancy ofSaccharomyces cerevisiae MSH3 and MSH6 in MSH2-dependentmismatch repair. Genes Dev 1996;10:407–420.

Journal of Surgical Oncology

Fig. 5. Relationship between hMLH1 and hMSH2 protein expression(A) A venn figure of the population according to hMLH1/hMSH2expression. (B) A dot plot analysis for hMLH1/hMSH2. There was norelationship between the expression of two proteins.

382 Kouso et al.

14. Jascur T, Boland CR: Structure and function of the components ofthe human DNA mismatch repair system. Int J Cancer 2006;119:2030–2035.

15. Iyer RR, Pluciennik A, Burdett V, et al.: DNA mismatch repair:functions and mechanisms. Chem Rev 2006;106:302–323.

16. Li GM, Modrich P: Restoration of mismatch repair to nuclearextracts of H6 colorectal tumor cells by a heterodimer of humanMutL homologs. Proc Natl Acad Sci U S A 1995;92:1950–1954.

17. Fishel R, Lescoe MK, Rao MR, et al.: The human mutator genehomolog MSH2 and its association with hereditary nonpolyposiscolon cancer. Cell 1993;75:1027–1038.

18. Bronner CE, Baker SM, Morrison PT, et al.: Mutation in the DNAmismatch repair gene homologue hMLH1 is associated withhereditary non-polyposis colon cancer. Nature 1994;368:258–261.

19. Eshleman JR, Markowitz SD: Mismatch repair defects in humancarcinogenesis. Hum Mol Genet 1996;5:1489–1494.

20. Kinzler KW, Vogelstein B: Lessons from hereditary colorectalcancer. Cell 1996;87:159–170.

21. Lawes DA, SenGupta S, Boulos PB: The clinical importance andprognostic implications of microsatellite instability in sporadiccancer. Eur J Surg Oncol 2003;29:201–212.

22. Han HJ, Maruyama M, Baba S, et al.: Genomic structure ofhuman mismatch repair gene, hMLH1, and its mutation analysisin patients with hereditary non-polyposis colorectal cancer(HNPCC). Hum Mol Genet 1995;4:237–242.

23. Liu B, Parsons RE, Hamilton SR, et al.: hMSH2 mutations inhereditary nonpolyposis colorectal cancer kindreds. Cancer Res1994;54:4590–4594.

24. Herman JG, Umar A, Polyak K, et al.: Incidence and functionalconsequences of hMLH1 promoter hypermethylation in color-ectal carcinoma. Proc Natl Acad Sci U S A 1998;95:6870–6875.

25. Leung SY, Yuen ST, Chung LP, et al.: hMLH1 promotermethylation and lack of hMLH1 expression in sporadic gastriccarcinomas with high-frequency microsatellite instability. CancerRes 1999;59:159–164.

26. Geisler JP, Goodheart MJ, Sood AK, et al.: Mismatch repair geneexpression defects contribute to microsatellite instability inovarian carcinoma. Cancer 2003;98:2199–2206.

27. Wang YC, Lu YP, Tseng RC, et al.: Inactivation of hMLH1 andhMSH2 by promoter methylation in primary non-small cell lungtumors and matched sputum samples. J Clin Invest 2003;111:887–895.

28. Marcus VA, Madlensky L, Gryfe R, et al.: Immunohistochemistryfor hMLH1 and hMH2: A practical test for DNA mismatch repair-deficient tumors. Am J Surg Pathol 1999;23:1248–1255.

29. World Health Organization. Histological Typing of Lung andPleural Tumors. (3 ed). Springer-Verlag Geneva: 1999.

30. Mountain CF: Revisions in the International System for StagingLung Cancer. Chest 1997;111:1710–1717.

31. Oda S, Oki E, Maehara Y, et al.: Precise assessment ofmicrosatellite instability using high resolution fluorescent micro-satellite analysis. Nucleic Acids Res 1997;25:3415–3420.

32. Friedrich M, Villena-Heinsen C, Meyberg R, et al.: Immunohis-tochemical analysis of DNA ’mismatch-repair’ enzyme humanMut-S-Homologon-2 in ovarian carcinomas. Histochem J 1999;31:717–722.

33. Friedrich M, Meyberg R, Villena-Heinsen C, et al.: Immunohis-tochemical analysis of DNA mismatch-repair enzyme hMSH-2

and Ki-67 in breast carcinomas. Anticancer Res 1999;19:3349–3353.

34. Kunkel TA: DNA replication fidelity. J. Biol Chem 2004;279:16895–16898.

35. Boland CR, Thibodeau SN, Hamilton SR, et al.: A NationalCancer Institute Workshop on Microsatellite Instability for cancerdetection and familial predisposition: development of interna-tional criteria for the determination of microsatellite instability incolorectal cancer. Cancer Res 1998;58:5248–5257.

36. Xinarianos G, Liloglou T, Prime W, et al.: hMLH1 and hMSH2expression correlates with allelic imbalance on chromosome 3p innon-small cell lung carcinomas. Cancer Res 2000;60:4216–4221.

37. Lanza G, Gafa R, Maestri I, et al.: Immunohistochemical patternof MLH1/MSH2 expression is related to clinical and pathologicalfeatures in colorectal adenocarcinomas with microsatelliteinstability. Mod Pathol 2002;15:741–749.

38. Staebler A, Lax SF, Ellenson LH: Altered expression of hMLH1and hMSH2 protein in endometrial carcinomas with micro-satellite instability. Hum Pathol 2000;31:354–358.

39. Rosen DG, Cai KQ, Luthra R, et al.: Immunohistochemicalstaining of hMLH1 and hMSH2 reflects microsatellite instabilitystatus in ovarian carcinoma. Mod Pathol 2006;19:1414–1420.

40. Wilentz RE, Goggins M, Redston M, et al.: Genetic, immuno-histochemical, and clinical features of medullary carcinoma of thepancreas: A newly described and characterized entity. Am JPathol 2000;156:1641–1651.

41. Kawaguchi K, Oda Y, Takahira T, et al.: Microsatellite instabilityand hMLH1 and hMSH2 expression analysis in soft tissuesarcomas. Oncol Rep 2005;13:241–246.

42. Chang JW, Chen YC, Chen CY, et al.: Correlation of geneticinstability with mismatch repair protein expression and p53mutations in non-small cell lung cancer. Clin Cancer Res 2000;6:1639–1646.

43. Liu B, Nicolaides NC, Markowitz S, et al.: Mismatch repair genedefects in sporadic colorectal cancers with microsatelliteinstability. Nat Genet 1995;9:48–55.

44. Polynuclear aromatic compounds. General remarks on thesubstances considered. IARC Monogr Eval Carcinog Risk ChemHum 1983;32:33–91.

45. Konishi M, Kikuchi-Yanoshita R, Tanaka K, et al.: Molecularnature of colon tumors in hereditary nonpolyposis colon cancer,familial polyposis, and sporadic colon cancer. Gastroenterology1996;111:307–317.

46. Losi L, Ponz de Leon M, Jiricny J, et al.: K-ras and p53 mutationsin hereditary non-polyposis colorectal cancers. Int J Cancer1997;74:94–96.

47. Furukawa T, Konishi F, Shitoh K, et al.: Evaluation of screeningstrategy for detecting hereditary nonpolyposis colorectal carci-noma. Cancer 2002;94:911–920.

48. Fujiwara T, Stolker JM, Watanabe T, et al.: Accumulated clonalgenetic alterations in familial and sporadic colorectal carcinomaswith widespread instability in microsatellite sequences. Am JPathol 1998;153:1063–1078.

49. Scartozzi M, Franciosi V, Campanini N, et al.: Mismatch repairsystem (MMR) status correlates with response and survival innon-small cell lung cancer (NSCLC) patients. Lung Cancer 2006;53:103–109.

50. Skarda J, Fridman E, Plevova P, et al.: Prognostic value of hMLH1and hMSH2 immunohistochemical expression in non-small celllung cancer. A. tissue microarray study. Biomed Pap Med FacUniv Palacky Olomouc Czech Repub 2006;150:255–259.

Journal of Surgical Oncology

Mismatch Repair Protein in NSCLC 383