Mismatch Repair Gene Expression Defects Contribute
to Microsatellite Instability in Ovarian Carcinoma
John P. Geisler, M.D.
Michael J. Goodheart, M.D.
Anil K. Sood, M.D.
Richard J. Holmes, M.D.
Melanie A. Hatterman-Zogg, M.D.
Richard E. Buller, M.D.
Division of Gynecologic Oncology, Department of
Obstetrics and Gynecology, Holden Comprehensive
Cancer Center, University of Iowa Hospitals and
Clinics, Iowa City, Iowa.
See editorial on pages 2091–4, this issue.
Supported in part by a Department of Health and
Human Services, Public Health Service–National
Institute of Health Grant 1R21 CA 84121-01 to R.E.
Buller and T32 CA 79445-01A1 to John P. Geisler.
Portions of this research were supported by Grant
IN-1227 from the American Cancer Society admin-
istered through the University of Iowa Holden Com-
prehensive Cancer Center to A. K. Sood, and the
through National Institutes of Health Grant
5KIZHD00849 and the American Board of Obstet-
rics and Gynecology to A. K. Sood.
The authors wish to thank Frederick E. Domann,
Ph.D., for his general laboratory assistance in the
development of the methylation-specific–polymer-
ase chain reaction (MS-PCR).
Address for reprints: John P. Geisler, M.D., Indiana
Gynecologic Oncology, 8424 Naab Road, Suite 2M,
Indianapolis, IN; Fax: (317) 872-5347; E-mail:
BACKGROUND. hMLH1, the human MutL homologue, has been linked to microsat-
ellite instability (MSI) in gastrointestinal tumors. However, to the authors’ knowl-
edge, the role of hMLH1, the other mismatch repair genes (MMR), and MSI in
ovarian carcinoma has not been well defined. The purpose of the current study was
to determine the relation between MSI of ovarian carcinoma and MMR gene
expression, hMLH1 and hMSH2 hypermethylation, and hMLH1 and hMSH2 null
METHODS. hMLH1 mRNA was detected by reverse transcriptase-polymerase chain
reaction (RT-PCR) and amplification of cDNA using a housekeeping gene (glycerol
3-phosphate dehydrogenase) as a control for mRNA quality and quantity. Methy-
lation-specific PCR (MS-PCR) was used to correlate methylation of the hMLH1 and
hMSH2 CpG islands with mRNA expression status. Similar techniques were used to
evaluate the concomitant expression of five other MMR: hMSH2, hMSH3, hMSH6,
PMS1, and PMS2. Microsatellite instability was studied using the National Cancer
Institute consensus markers (D2S123, D5S346, D17S250, BAT25, and BAT26) and
NM23 as described previously.
RESULTS. One hundred twenty-five primary tumors were analyzed. High-fre-
quency MSI (MSI-H) was found in 21 tumors (16.8%). hMLH1 mRNA was absent in
10 of these 21 tumors (47.6%). In each case, coordinated hypermethylation of both
regions A and C of the promoter was identified. Microsatellite stable and low-
frequency MSI tumors all were found to express not only hMLH1 but the other
MMR genes as well (P ? 0.001). Absence of expression of hMSH2 and the four other
MMRs occurred in tumors with absent hMLH1 mRNA expression because of CpG
island hypermethylation. No absence of expression of hMSH2, hMSH3, hMSH6,
PMS1, or PMS2 was found to occur in tumors expressing hMLH1. None of the 11
MSI-H tumors without promoter hypermethylation demonstrated a null mutation
in hMLH1 or hMSH2.
CONCLUSIONS. A molecular mechanism to explain ? 50% of the MSI-H phenotype
in ovarian carcinoma cases was demonstrated. MSI-H may occur because of MMR
defects, especially hMLH1 promoter hypermethylation. Additional mechanisms
are required to explain the balance between the cases of MSI-H as well as the
phenomenon of MSI-L tumors. Cancer 2003;98:2199–206.
© 2003 American Cancer Society.
KEYWORDS: mismatch repair, mRNA, CpG island methylation, microsatellite insta-
bility, ovarian carcinoma.
sion or sequence of the mismatch repair genes (MMR) is relatively
common.1–5To our knowledge, the correlation among these occur-
rences is not fully understood, but MSI appears to occur after the
mutation or loss of expression of hMLH1 or hMSH2.5,6Generally,
n gastrointestinal tumors, especially gastric and colon carcinomas,
microsatellite instability (MSI) secondary to alterations in expres-
© 2003 American Cancer Society
hereditary colon and related tumors in hereditary
nonpolyposis colon carcinoma (HNPCC) families are
secondary to a mutation in one of the MMR genes.7In
contrast, only 2–13%of sporadic colon carcinomas are
reported to be associated with MSI.5,8For these tu-
mors, as well as a cohort of endometrial and gastric
carcinomas, MSI is associated with absent MMR gene
expression secondary to hMLH1 promoter methyl-
Ovarian carcinoma is one of the most common
forms of hereditary cancer in adult females12and is
the leading cause of death from gynecologic cancer in
the U.S.13The role of the MMR genes in this malig-
nancy to our knowledge is still largely unknown.14
Germline mutation of hMLH1 or hMSH2 is relatively
rare, with each reported to occur in ? 1% of consec-
utive ovarian carcinoma cases.15The presence of MSI
in ovarian carcinoma has been well documented by
some investigators, but not by others.16–22The fre-
quency of MSI ranges from a low of 3.8% to a high of
37% depending on the markers studied.16–23Together,
these observations have lead us to hypothesize on the
existence of a relation between MSI and the absence
of expression of MMRs because of promoter down-
regulation.20,24Preliminary support for such a corre-
lation has been derived from the work of Strathdee et
MSI and MMR have both been examined inde-
pendently in ovarian carcinoma patients, and the pur-
pose of the current study was to investigate the cor-
relation between MSI and MMR expression (human
MutL homologues: hMLH1, PMS1, and PMS2; and
human MutS homologues: hMSH2, hMSH3, and
hMSH6), hMSH2 and hMLH1 promoter hypermethyl-
ation, and hMSH2 and hMLH1 null mutations in a
series of ovarian tumors.
MATERIALS AND METHODS
The current study was performed in accordance with
the standards of our institutional committee for the
Protection of Human Subjects. Patient selection was
made solely on the basis of the availability of snap-
frozen tissue for reverse transcriptase-polymerase
chain reaction (RT-PCR)/cDNA amplification. The
majority of these tumors were 100% tumor with none
reported to be ? 90% tumor. Of the 125 patients in the
study, 107 had ovarian carcinoma, 11 had peritoneal
carcinoma, and 7 had carcinoma of the fallopian tube.
Because of the similarities among these tumors, they
are analyzed together as “ovarian carcinomas.”
Microsatellite instability was studied using six mark-
ers: D2S123, D5S346, D17S250, BAT25, BAT26, and
NM23. The first five markers were chosen because of a
National Cancer Institute workshop recommendation
to include these markers when examining MSI in colo-
rectal carcinomas.26The NM23 locus was chosen be-
cause of its proximity to the BRCA1 locus and its
demonstrated utility as a marker for MSI in ovarian
carcinoma.18,27,28Methods for this analysis have been
described previously by our group.18,27,29
Normal and tumor DNA amplimers were evalu-
ated side by side on 7% denaturing polyacrylamide
gels. A marker was considered informative if two al-
lelic bands were observed in normal DNA. MSI was
identified when shifts were noted in either of the al-
lelic bands in tumor DNA compared with normal DNA
consistently in repeated experiments. A tumor was
considered to be microsatellite stable (MSS) if there
were no shifts at any of the six markers. Low-fre-
quency MSI (MSI-L) was defined by a shift in one
marker whereas high-frequency MSI (MSI-H) was de-
fined by shifts at two or more loci.26
mRNA Expression by RT-PCR
The techniques used for RNA isolation and cDNA syn-
thesis by RT-PCR were as described previously begin-
ning with snap-frozen tumor samples stored at ?140
°C.30Failure to amplify a product in the cDNA reac-
tions (despite appropriate amplification of a house-
keeping gene sequence such as glycerol 3-phosphate
dehydrogenase [G3PD]) provided candidate tumors in
which epigenetic phenomenon, including promoter
silencing, may be operational. Six MMR genes were
coamplified (hMLH1, PMS1, PMS2, hMSH2, hMSH3,
and hMSH6) from a single multiplex PCR reaction by
modifying the methods of Wei et al.31,32and Chang et
al.33The ?-actin in the analysis by Wei et al. was
replaced with the one MMR not studied by their
group–hMSH6 (Table 1). A buffer with an increased
magnesium concentration (6.7 mM MgCl2)34was used
with 10 pmoles of each primer pair for hMLH1, PMS1,
and PMS2 and 16.7 pmoles of each hMSH2, hMSH3,
and hMSH6 primer pair; 5 U of Taq polymerase; and 3
?L of cDNA. G3PD expression was measured in a
separate, concurrent reaction.30
After EcoR1 restriction, methylation-specific PCR (MS-
PCR) was performed on NaHSO3-converted DNA. The
NaHSO3reaction was described previously by several
authors.34–37Briefly, DNA (0.5–5 ?g) was incubated
first with 0.3 M NaOH at 37 °C. The alkalinized mixture
was exposed to 3.6 M of NaHSO3and 1 mM of hydro-
quinone at 55 °C for 14 hours before the products were
recovered and desalted with Promega? Wizard Prep
(Promega, Madison, WI). Desalting was performed per
2200CANCER November 15, 2003 / Volume 98 / Number 10
the manufacturer’s recommendation except for the
last elution in which 75 °C deionized water was incu-
bated on the column at room temperature for 5 min-
utes before the final centrifuging. The solution then
was incubated with 0.3 M of NaOH at 37 °C again
before the addition of 3 M of ammonium acetate and
95% ethanol. The mixture was next incubated at -20 °C
for 20 minutes and then centrifuged at 18,620 ? g (4
°C) for 30 minutes. The supernatant fluid was re-
moved, and the DNA was lyophilized and finally re-
suspended in 100 ?L double deionized water (ddH20).
MS-PCR for hMLH1 was performed on the con-
verted DNA using the primers and conditions de-
scribed in Table 1. MS-PCR of hMSH2 was performed
utilizing published primers and conditions.8The same
buffer used for the multiplex PCR was used for all
MS-PCR reactions.34CpGenome™- Universal Methyl-
ated DNA (Intergen Company, Purchase, NY) was
used as the methylated control after NaHSO3conver-
sion. Nonneoplastic ovarian epithelium and human
placental tissue after NaHSO3conversion were used as
unmethylated controls. A null control with all reagents
except DNA template also was performed. The prim-
ers covered a portion of the CpG island in the hMLH1
promoter (GenBank #AB017806; GenBank, National
Institutes of Health, Bethesda, MD) that flanks the
transcription start site (region A: methylated product
at 113 base pairs [bp] and unmethylated product at
315 bp) as well as a sequence 5? to this designated as
region C (methylated product at 96 bp and unmeth-
ylated product at 189 bp).38The hMSH2 primers cov-
ered an area in the CpG island of exon 1 (GenBank
#U041206) near the transcription start site (methyl-
ated product at 132 bp and unmethylated product at
Protein Truncation Test Analysis
The open reading frames of both hMLH1 and hMSH2
were divided into two overlapping fragments as pre-
viously described.39,40cDNA templates (generated
from random hexamers) were used to generate the
fragments. The technique for RNA isolation and cDNA
synthesis for RT-PCR starting from snap-frozen tumor
samples stored at ?140 °C was described previously.30
Primers containing both a eukaryotic translation
initiation site (ATG) and a T7 promoter were used to
generate PCR products that were amenable to protein
truncation testing. The individual PCR reactions were
performed with complete translation of the product in
the TNT? Quick Coupled Transcription/Translation
System (Promega) essentially as shown in the manu-
facturer’s protocol. After the addition of sodium do-
decyl sulfate (SDS) buffer, the samples were heated to
85 °C for 2 minutes. An aliquot of the sample was then
subjected SDS-polyacrylamide gel electrophoresis un-
til the leading dye completely crossed the gel (45 watts
for 15 minutes). Gels were then fixed, dried, and ex-
posed to Scientific Imaging Film (Kodak™, Rochester,
Primers for mRNA Expression and MS-PCR
Gene Template type
no. Primer sequence 5?33?
DNA - methylated
DNA - unmethylated
DNA - methylated
DNA - unmethylated
MS-PCR: methylation-specific-polymerase chain reaction; bp: base pair.
aAdapted from Chang DK, Ricciardiello L, Goel A, Chang CL, Boland CR. Steady-state regulation of the human DNA mismatch repair system. J Biol Chem. 2000;275:18424–18431.
bAdapted from Grady WM, Rajput A, Lutterbaugh JD, Markowitz SD. Detection of aberrantly methylated hMLH1 promoter DNA in the serum of patients with microsatellite unstable colon cancer. Cancer Res.
Polymerase chain reaction conditions for mismatched repair gene cDNA multiplex: 95 °C for 5 minutes, 35 cycles at 95 °C for 30 seconds (59 °C for 30 seconds, 72 °C for 45 seconds), and 72 °C for 10 minutes.
Polymerase chain reaction conditions for NaHSO3Region A methylated DNA: 95 °C for 5 minutes, 35 cycles (95 °C for 30 seconds, 62 °C for 45 seconds, and 72 °C for 30 seconds), and 72 °C for 4 minutes.
Polymerase chain reaction conditions for NaHSO3Region A unmethylated DNA: 95 °C for 5 minutes, 35 cycles (95 °C for 30 seconds, 54 °C for 45 seconds, and 72 °C for 30 seconds), and 72 °C for 4 minutes.
PCR conditions for NaHSO3Region C methylated DNA: 95 °C for 5 minutes, 35 cycles (95 °C for 30 seconds, 55 °C for 45 seconds, and 70 °C for 30 seconds), and 70 °C for 4 minutes.
Polymerase chain reaction conditions for NaHSO3Region C unmethylated DNA: 95 °C for 5 minutes, 40 cycles (95°C for 30 seconds, 58 °C for 45 seconds, and 70 °C for 30 seconds), and 70 °C for 4 minutes.
MMR Genes in Ovarian Carcinoma/Geisler et al.2201
Direct PCR-based sequencing was performed on any
sample in which a shift in band pattern was observed
on the protein truncation test. DNA products were
purified with the Wizard PCR DNA Purification System
(Promega) or the QIAquick™ PCR Purification Kit
(Qiagen Inc., Valencia, CA). The PCR sequencing re-
action was completed using the DNA Cyclic Sequenc-
ing System (Promega). Sequencing of the candidate
mutations (using the Licor IR2; LI-COR, Lincoln, NE)
based on protein truncation testing was performed by
selecting an appropriate region to sequence on a 41-
cm, 7% Long Ranger™ polyacrylamide gel (FMC Bio-
products, Rockland, ME). Specific M13-tagged primers
were chosen based on the size of the truncated protein
product. Electrophoresis was performed at 50 °C, 31.5
volts, and 35 milliamps for approximately 6 hours.
Gels were evaluated by using LI-COR ImagIR 4.0 data
collection software and image manipulator software
(LI-COR). Mutations were confirmed with bidirec-
tional sequencing of products from a second indepen-
Statistical analyses, including the chi-square test, were
performed utilizing SPSS for Windows version 10.0
(SPSS Inc., Chicago, IL).
One hundred twenty-five primary ovarian tumors
were analyzed for MSI and MMR expression. MSI-H
was present in 21 tumors (16.8%) and MSI-L was
present in 13 tumors (10.4%) whereas 91 tumors
(72.8%) were MSS. The correlation between MSI and
tumor histology is demonstrated in Table 2. Although
not statistically significant, serous and clear cell car-
cinomas were found to be less likely to be MSI-H
? 0.077). No primary peritoneal carcinomas demon-
strated MSI-H whereas a similar percentage of fallo-
pian tube carcinomas (28.6%) and ovarian carcinomas
(17.8%) were found to have MSI-H (P ? 0.23).
Multiplex PCR was utilized to study the simulta-
neous expression of all six MMRs. Figure 1 shows the
results of a representative gel. Tumor 107 was MSS
and expressed all six MMRs. Tumors 519, 2, and 109 all
were MSI-H and demonstrated the absence of expres-
sion of one or more MMRs.
Expression of all six MMR mRNAs were readily
detectable in the 104 tumors designated as MSI-L or
MSS. In the 21 MSI-H tumors, hMLH1 mRNA expres-
sion was absent in 10 tumors (47.6%). Eight of the 10
tumors lacking hMLH1 expression also failed to ex-
press hMSH2 and various other MMRs (Table 3). Two
tumors (Tumor 109 and Tumor 302), which were lack-
ing hMLH1 expression, had readily detectable MRNA
from the five other MMRs (Table 3). Eleven MSI-H
tumors appeared to express all 6 MMR mRNAs at
levels consistent with the level of expression reported
in the MSI-L and MSS tumors.
MS-PCR was utilized to examine the CpG island of
the promoter region of hMLH1 for hypermethylation.
Percentage of Tumors of Each Site and Histology with Specific Microsatellite Status
(n ? 87)
(n ? 22)
(n ? 9)
(n ? 4)
(n ? 2)
(n ? 1)
Ovary (n ? 107)
Peritoneum (n ? 11)
Fallopian tube (n ? 7)
ANOS: adenocarcinoma, not otherwise specified.
2202CANCER November 15, 2003 / Volume 98 / Number 10
Both the A and C regions of the promoter were stud-
ied.41Figure 2 shows a representative MS-PCR gel of
region C. Tumor 109 was MSI-H, lacked hMLH1
mRNA expression, and demonstrated only methylated
product. Tumor 60 also was found to be MSI-H
whereas Tumor 87 was MSS. hMLH1 mRNA expres-
sion was detectable in both and the promoters were
unmethylated. By convention, the status of the pro-
moter is determined by the ratio of methylated to
unmethylated product.8,36Hypermethylation of the
promoter A region (near the transcription start site)
was noted in all 10 MSI-H tumors lacking hMLH1
mRNA expression but in none of the tumors express-
ing hMLH1 mRNA (P ? 0.001). hMLH1 down-regula-
tion secondary to promoter hypermethylation was
found to be uniform at both the A and C regions (P
? 0.001). No tumors with hMLH1 mRNA expression
demonstrated hypermethylation in either Region A or
C. The eight MSI-H tumors (Table 3) without hMSH2
mRNA all demonstrated hypermethylation. This find-
ing has been shown in other tumors.42The 13 MSI-H
tumors that demonstrated hMSH2 mRNA (including 2
tumors that did not have hMLH1 mRNA) were un-
methylated at that locus.
The 11 remaining MSI-H tumors, all which ex-
pressed both hMLH1 and hMSH2, were studied by
protein truncation testing for the presence of null
mutations of hMLH1 and/or hMSH2. None of the
tumors demonstrated a truncating defect in hMSH2 or
The MMR system is complex. It is at least partially
responsible for preventing DNA damage from propa-
gating through the cell cycle. The six main MMR genes
(hMLH1, hMSH2, PMS1, PMS2, hMSH3, and hMSH6)
all are present throughout the cell cycle, although
some fluctuation in levels exists.43Expression of MMR
mRNA has been correlated closely with protein ex-
pression.33The two keys appear to be hMLH1 and
hMSH2 because the other MutL homologues (PMS1
and PMS2) as well as the other MutS homologues
(hMSH3 and hMSH6) are reported to be less stable in
the absence of the first member of their lineage.33In
humans, analogous to the murine model, the homol-
FIGURE 1. Representative agarose gel showing
interrelations between expression of mismatch
repair genes (MMR) mRNAs (product sizes: PMS1:
153 base pairs (bp); hMLH1: 194 bp; hMSH6: 267
bp; PMS2: 337 bp; hMSH2: 408 bp; and hMSH3:
486 bp). A microsatellite stable tumor, Tumor
107, expressed all six MMR mRNAs. Tumors 519,
2, and 109 were MSI-H. Tumor 519 demonstrated
absence of expression (AOE) of hMLH1. Tumor 2
had AOE of PMS1, hMLH1, hMSH2, and hMSH3
whereas Tumor 109 had AOE of hMLH1 and
Patterns of MMR mRNA Expression in MSI-H Tumors Lacking
hMLH1 mRNA Expression
Tumor no. hMLH1 hMSH2hMSH3 hMSH6PMS1 PMS2
MMR: mismatch repair genes; open circle: no expression of corresponding mRNA; Solid circles:
expression of corresponding mRNA present.
MMR Genes in Ovarian Carcinoma/Geisler et al.2203
ogous proteins to MSH2 and MSH6 form a het-
erodimer, MutS?, and MSH6 is less stable in the ab-
sence of MSH2 than in its presence.44
The data from the current study appear to corre-
late well with the current functional model of the
MMR system. In the current study, hMSH6 was not
expressed if hMSH2 expression was absent. Absence
of expression of PMS1, PMS2, hMSH3, or hMSH6 only
occurred in the absence of both hMLH1 and hMSH2.
However, only 8 of the 10 tumors with absent hMLH1
expression also were found to lack hMSH2 expression.
This observation suggests alternative hypotheses for
the interrelations between MSI, MMR promoter meth-
ylation, and MMR gene expression. Because greater
than half of MSI-H tumors are reportedly not ex-
plained by hMLH1 or hMSH2 CpG island hypermeth-
ylation, perhaps a gene mutation in hMLH1 or hMSH2
may explain the high frequency of MSI in these tu-
mors. This theory is supported by the current under-
standing of MSI and MMR defects present in other
tumor types.24,44,45The apparent rarity of hMLH1 and
hMSH2 mutations in ovarian carcinoma does not pre-
clude this explanation because the majority of inves-
tigators have to our knowledge examined only germ-
line mutations ofthese
mutations do not appear to have been as carefully
evaluated. Support for this hypothesis is generated
from the observation that 6 of 11 tumors with unex-
plained MSI were of endometrioid histology. Fujita et
al. reported a 50% incidence of somatic MMR gene
mutations in endometrioid ovarian carcinoma.46
However, we found no truncating mutations in either
hMLH1 or hMSH2.
Alternatively, one could conclude that MSI occurs
proximal to MMR gene silencing with an unknown
step preceding MSI or that unknown/unstudied MMR
are involved. Next, promoter methylation of hMLH1
and lack of hMLH1 expression occurs in those tumors
that are MSI-H. After this step, hMSH2 expression is
lost in a certain percentage of tumors with hMLH1
promoter hypermethylation. Finally, loss of expres-
sion of other MMRs occurs after down-regulation of
both hMLH1 and hMSH2. To our knowledge, there is
no supporting evidence for this conclusion from other
tumor types. Essentially, another, yet to be deter-
mined, genetic or epigenetic phenomenon would
have to be the proximal cause of MSI.
In addition to mismatch repair defects, tumors
with hMLH1 defects demonstrate disordered apopto-
sis.24Because hMLH1 methylation can be reversed
with subsequent reexpression of the mRNA and pro-
teins in cell culture, it presents itself as an attractive
therapeutic target potentially treatable with 5-azacyti-
dine.1,24,45Although hypermethylation is reported to
occur in only a small percentage of ovarian carcinoma
cases, it is important because those with hMLH1 hy-
permethylation generally are platinum resistant.25
Clearly, hMLH1 promoter methylation with a sub-
sequent lack of expression plays a role in the propor-
tion of ovarian carcinomas in which it is related inti-
mately to MSI. This mechanism appears to be causal
for approximately half of the MSI-H tumors but none
of the MSI-L tumors. Analysis of the mutational spec-
trum of those MSI-H tumors with apparently normal
MMR gene expression currently is ongoing in our lab-
oratory, in addition to further studies of the MSI-L
subgroup of ovarian carcinoma.
FIGURE 2. Representative agarose gel
merase chain reaction (MS-PCR) of region
C of the hMLH1 promoter. MC indicates
methylation control (methylated product
was 96 base pairs (bp) and unmethylated
product was 189 bp). Tumor 109 had a
high frequency of microsatellite instability
(MSI-H) and absence of expression (AOE)
Tumor 60, expressed hMLH1 mRNA and
demonstrated primarily unmethylated (U)
amplimers. Only unmethylated amplimers
were detected in the microsatellite stable
Tumor 87 that expressed hMLH1 mRNA
2204CANCER November 15, 2003 / Volume 98 / Number 10
1.Veigl ML, Kasturi L, Olechnowicz J, et al. Biallelic inactiva-
tion of hMLH1 by epigenetic gene silencing, a novel mech-
anism causing human MSI cancers. Proc Natl Acad Sci USA.
2.Fleisher AS, Esteller M, Wang, SM, et al. Hypermethylation
of the hMLH1 gene promoter in human gastric cancers with
microsatellite instability. Cancer Res. 1999;59:1090–1095.
3. Suzuki H, Itoh F, Toyota M, et al. Distinct methylation
pattern and microsatellite instability in sporadic gastric can-
cer. Int J Cancer. 1999;83:309–313.
4.Leung SY, Yuen ST, Chung LP, Chu KM, Chan ASY, Ho JCI.
hMLH1 promoter methylation and lack of hMLH1 expres-
sion in sporadic gastric carcinomas with high frequency
microsatellite instability. Cancer Res. 1999;59:159–164.
5.Pedroni M, Sala E, Scarselli A, et al. Microsatellite instability
and mismatch-repair protein expression in hereditary and
sporadic colorectal carcinogenesis. Cancer Res. 2001;21:
6.Wheeler JMD, Loukola A, Aaltonen LA, Mortensen NJ, Bod-
mer WF. The role of hypermethylation of the hMLH1 pro-
moter region in HNPCC versus MSI? sporadic colorectal
cancers. J Med Genet. 2000;37:588–592.
7. Huang J, Kuismanen SA, Liu T, et al. MSH6 and MSH3 are
rarely involved in genetic predisposition to nonpolypotic
colon cancer. Cancer Res. 2001;61:1619–1623.
8. Herman JG, Umar A, Polyak K, et al. Incidence and functional
consequences of hMLH1 promoter hypermethylation in colo-
rectal carcinoma. Proc Natl Acad Sci USA. 1998;95:6870–6875.
9. Esteller M, Catasus L, Matias-Guiu X, et al. hMLH1 promoter
hypermethylation is an early event in human endometrial
tumorigenesis. Am J Pathol. 1999;155:1399–1402.
10. Gurin CC, Federici MG, Kang L, Boyd J. Causes and conse-
quences of microsatellite instability in endometrial carci-
noma. Cancer Res. 1999;59:462–466.
11. Salvesen HB, MacDonald N, Ryan A, et al. Methylation of
hMLH1 in a population-based series of endometrial carci-
nomas. Clin Cancer Res. 2000;6:3607–3613.
12. Moslehi R, Chu W, Karlan B, et al. BRCA1 and BRCA2 mu-
tation analysis of 208 Ashkenazi Jewish women with ovarian
cancer. Am J Hum Genet. 2000;66:1259–1272.
13. Greenlee RT, Murray T, Bolden S, Wingo PA. Cancer statis-
tics, 2000. CA Cancer J Clin. 2000;50:7–33.
14. Geisler JP, Geisler HE, Miller GA, Wiemann MC, Zhou Z,
Crabtree W. Immunohistochemical staining of the mis-
match repair gene, hMSH2, and survival in patients with
ovarian carcinoma. Eur J Gynaecol Oncol. 2000;21:237–240.
15. Rubin, SC, Blackwood MA, Bandera C, et al. BRCA1, BRCA2,
and hereditary nonpolyposis colorectal cancer gene muta-
tions in an unselected ovarian cancer population: relation-
ship to family history and implications for genetic testing.
Am J Obstet Gynecol. 1998;178:670–675.
16. Sood AK, Buller RE. Genomic instability in ovarian cancer: a
reassessment using an arbitrarily primed polymerase chain
reaction. Oncogene. 1996;13:2499–2504.
17. Arzimanoglou II, Lallas T, Osborne M, Barber H, Gilbert F.
Microsatellite instability differences between familial and spo-
radic ovarian cancers. Carcinogenesis. 1996;17:1799–1804.
18. Sood AK, Skilling JS, Buller RE. Ovarian cancer genomic
instability correlates with p53 frameshift mutations. Cancer
19. Buller RE, Skilling JS, Sood AK, Plaxe S, Baergen RN, Lager
DJ. Field cancerization: why late “recurrent” ovarian cancer
is not recurrent. Am J Obstet Gynecol. 1998;178:641–649.
20. Colella G, Vikhanskaya F, Codegoni AM, Bonazzi C,
D’Incalci N, Broggini M. hMLH1 and hMSH2 expression and
BAX frameshift mutations in ovarian cancer cell lines and
tumors. Carcinogenesis. 1998;19:691–694.
21. Codegoni AM, Bertoni F, Colella G, et al. Microsatellite in-
stability and frameshift mutations in genes involved in cell
cycle progression or apoptosis in ovarian cancer. Oncol Res.
22. Johannsdottir JT, Jonasson JG, Bergthorsson JT, et al. The
effect of mismatch repair deficiency on tumorigenesis: mi-
crosatellite instability affecting genes containing short re-
peated sequences. Int J Oncol. 2000;16:133–139.
23. Allen HJ, DiCioccio RA, Hohmann P, Piver MS, Tworek H.
Microsatellite instability in ovarian and other pelvic carci-
nomas. Cancer Genet Cytogenet. 2000;117:163–166.
24. Wheeler JMD, Bodmer WF. DNA mismatch repair genes and
colorectal cancer. Gut. 2000;47:148–153.
25. Strathdee G, MacKean MJ, Illand M, Brown R. A role for
methylation of the hMLH1 promoter in loss of hMLH1 ex-
pression and drug resistance in ovarian cancer. Oncogene.
26. Boland CR, Thibodeau SN, Hamilton SR, et al. A National
Cancer Institute Workshop on Microsatellite Instability for
cancer detection and familial predisposition: development
of international criteria for the determination of microsat-
ellite instability in colorectal cancer. Cancer Res. 1998;58:
27. Sood AK, Holmes RW, Hendrix MJC, Buller RE. Application
of the NCI international criteria for determination of mic-
rosatellite instability in ovarian cancer. Cancer Res. 2001;61:
28. Jacobs IJ, Smith SA, Wiseman RW, et al. A deletion unit on
chromosome 17q in epithelial ovarian tumors distal to the
familial breast/ovarian cancer locus. Cancer Res. 1993;53:
29. Buller RE, Shahin MS, Holmes R, Hatterman MA, Kirby PA,
Sood AK. p53 mutations and microsatellite instability in
ovarian cancer: Yin and Yang. Am J Obstet Gynecol. 2001;
30. Skilling JS, Sood AK, Niemann T, Lager DJ, Buller RE. An
abundance of p53 null mutations in ovarian cancer. Onco-
31. Wei Q, Bondy ML, Mao L, et al. Reduced expression of
mismatch repair genes measured by multiplex reverse tran-
scription-polymerase chain reaction in human gliomas.
Cancer Res. 1997;57:1673–1677.
32. Wei Q, Guan Y, Radinsky R, et al. Expression of five selected
human mismatch repair genes simultaneously detected in
normal and cancer cell lines by a nonradioactive multiplex
reverse transcription-polymerase chain reaction. Pathobiol-
33. Chang DK, Ricciardiello L, Goel A, Chang CL, Boland CR.
Steady-state regulation of the human DNA mismatch repair
system. J Biol Chem. 2000;275:18424–18431.
34. Esteller M, Silva JM, Dominguez G, et al Promoter hyper-
methylation and BRCA1 inactivation in sporadic breast and
ovarian tumors. J Natl Cancer Inst. 2000;92:564–569.
35. Clark SJ, Harrison J, Paul CL, Frommer M. High sensitivity
mapping of methylated cytosines. Nucleic Acids Res. 1994;
36. Herman JG, Graff J, Myohanen S, Nelkin BD, Baylin SB.
Methylation-specific PCR: a novel PCR assay for methyl-
ation status of CpG islands. Proc Natl Acad Sci USA. 1996;
MMR Genes in Ovarian Carcinoma/Geisler et al.2205
37. Domann FE, Rice JC, Hendrix MJC, Futscher BW. Epigenetic Download full-text
silencing of maspin gene expression in human breast can-
cers. Int J Cancer. 2000;85:805–810.
38. Grady WM, Rajput A, Lutterbaugh JD, Markowitz SD. De-
tection of aberrantly methylated hMLH1 promoter DNA in
the serum of patients with microsatellite unstable colon
cancer. Cancer Res. 2001;61:900–902.
39. Liu B, Parsons RE, Hamilton SR, et al. hMSH2 mutations in
hereditary nonpolyposis colorectal cancer kindreds. Cancer
40. Liu B, Nicolaides NC, Markowitz S, et al. Mismatch repair
gene defects in sporadic colorectal cancers with microsat-
ellite instability. Nat Genet. 1995;9:48–55.
41. Deng G, Chen A, Hong J, Chae HS, Kim YS. Methylation of
CpG in a small region of the hMLH1 promoter invariably
correlates with the absence of gene expression. Cancer Res.
42. Wang YC, Lu YP, Tseng RC, et al. Inactivation of hMLH1 and
hMSH2 by promoter methylation in primary non-small cell
lung tumors and matched sputum samples. J Clin Invest.
43. Meyers M, Theodosiou M, Acharya S, et al. Cell cycle regu-
lation of the human DNA mismatch repair genes hMSH2,
hMLH1 and hPMS2. Cancer Res. 1997;57:206–208.
44. Schweizer P, Moisio AL, Kuismanen SA, et al. Lack of MSH2
and MSH6 characterizes endometrial but not colon carci-
nomas in hereditary nonpolyposis colorectal cancer. Cancer
45. WheelerJMD,Beck NE,
Mortensen MC, Bodmer WF. Mechanisms of inactivation of
mismatch repair genes in human colorectal cell lines: the
predominant role of hMLH1. Proc Natl Acad Sci USA. 1999;
46. Fujita M, Enomoto T, Yoshino K, et al. Microsatellite insta-
bility and alterations in the hMSH2 gene in human ovarian
cancer. Int J Cancer. 1995;64:361–366.
2206 CANCER November 15, 2003 / Volume 98 / Number 10