Content uploaded by Nada Al Tassan
Author content
All content in this area was uploaded by Nada Al Tassan
Content may be subject to copyright.
Fig. 1 Pedigree of family N. Siblings
II:1 and II:2 were found to have
adenomas at 50 and 46 y, respec-
tively. Both had roughly 50 macro-
scopically visible adenomas at
colectomy at 59 and 55 y, respec-
tively. Sibling II:3 died after a
colonic adenocarcinoma and an
adjacent adenoma were discovered
at 46 y, but without full assessment
of the large bowel. Siblings II:4–7
were normal on colonoscopic
assessment between 36 and 49 y,
and III:1–5 were normal on colono-
scopic assessment between 24 and
33 y. APC haplotypes with the intra-
genic markers Glu1317Gln (1317
E/Q), Ser2497Leu (2497 S/L) and the
closely linked DP1 (CA)
n
repeat are
shown. The wildtype APC haplo-
type shared by each of the three
affected siblings was also present in
five other family members who
were phenotypically normal.
letter
nature genetics •
volume 30 • february 2002
227
Inherited variants of MYH associated with somatic
G:C→T:A mutations in colorectal tumors
Nada Al-Tassan
1
, Nikolas H. Chmiel
2
, Julie Maynard
1
, Nick Fleming
1
, Alison L. Livingston
2
, Geraint T.
Williams
3
, Angela K. Hodges
1
, D. Rhodri Davies
4
, Sheila S. David
2
, Julian R. Sampson
1
& Jeremy P. Cheadle
1
1
Institute of Medical Genetics, University of Wales College of Medicine, Heath Park, Cardiff, CF14 4XN, UK.
2
Department of Chemistry, University of Utah,
Salt Lake City, Utah 84112, USA.
3
Department of Pathology, University of Wales College of Medicine, Heath Park, Cardiff, CF14 4XN, UK.
4
Department of
Gastroenterology, University Hospital of Wales, Heath Park, Cardiff, CF14 4XN, UK. Correspondence should be addressed to J.R.S. (e-mail:
sampson@cardiff.ac.uk).
Inherited defects of base excision repair have not been associ-
ated with any human genetic disorder, although mutations of
the genes mutM and mutY, which function in Escherichia coli
base excision repair, lead to increased transversions of G:C to
T:A
1–4
. We have studied family N, which is affected with multi-
ple colorectal adenomas and carcinoma but lacks an inherited
mutation of the adenomatous polyposis coli gene (APC) that is
associated with familial adenomatous polyposis
5
. Here we
show that 11 tumors from 3 affected siblings contain 18
somatic inactivating mutations of APC and that 15 of these
mutations are G:C→T:A transversions—a significantly greater
proportion than is found in sporadic tumors or in tumors asso-
ciated with familial adenomatous polyposis. Analysis of the
human homolog of mutY, MYH
6
, showed that the siblings were
compound heterozygotes for the nonconservative missense
variants Tyr165Cys and Gly382Asp. These mutations affect
residues that are conserved in mutY of E. coli (Tyr82 and
Gly253). Tyrosine 82 is located in the pseudo-helix-hairpin-helix
(HhH) motif and is predicted to function in mismatch
specificity
7
. Assays of adenine glycosylase activity of the
Tyr82Cys and Gly253Asp mutant proteins with 8-oxoG:A and
G:A substrates show that their activity is reduced significantly.
Our findings link the inherited variants in MYH to the pattern
of somatic APC mutation in family N and implicate defective
base excision repair in predisposition to tumors in humans.
The base excision repair (BER) pathway has a principal role in
the repair of mutations caused by reactive oxygen species that are
generated during aerobic metabolism
8
. Although oxidative DNA
damage has been implicated in the etiology of degenerative dis-
eases, aging and cancer
9
, so far there is no evidence to link inher-
ited deficiencies of BER to these processes.
We studied a British family, family N, in which three siblings
(II:1–3) were affected by multiple colorectal adenomas and carci-
noma (Fig. 1). Sequencing of the 8532-bp ORF of APC
10,11
in
constitutional DNA samples from siblings II:1 and II:3 showed
that this gene contained five silent base substitutions, 1458C→T
(Tyr486), 1635A→G (Ala545), 4479G→A (Thr1493), 5265
G→A (Ala1755) and 5268G→T (Ser1756), and two missense
variants, Glu1317Gln and Ser2497Leu, but no clear pathogenic
change. None of these variants was present in all three affected
siblings (Fig. 1). We sequenced the RT–PCR products from exons
1–14 of APC derived from normal colon tissue from sibling II:1,
which confirmed that there was equal expression of both APC
alleles with alternate splicing of exons 9a and 10a, as reported
previously
10,12
. These data exclude the possibility that inactiva-
tion of APC was the primary inherited defect in family N.
Inherited mutations of the mismatch repair (MMR) genes
cause hereditary nonpolyposis colorectal cancer (HNPCC),
which is characterized by microsatellite instability (MSI) in the
associated tumors
13
. However, assessment of MSI at seven
Published online: 30 January 2002, DOI: 10.1038/ng828
I.
II.
III.
E Q
S S
3 2
E E
S S
3 4
E Q
S S
3 2
E E
S S
3 1
E E
S S
3 1
E E
S S
3 5
E Q
S S
3 2
E E
S L
3 6
E E
S L
8 6
E Q
S S
8 2
E E
S L
8 6
E E
S L
3 6
1317 E/Q
2497 S/L
DP1
key
E E
S S
3 1
© 2002 Nature Publishing Group http://genetics.nature.com
letter
228 nature genetics •
volume 30 • february 2002
marker loci in DNA extracted from 11 tumors from family N
showed that there was instability with only the Mfd15 marker in
a single adenoma. This observation, together with the multiple
adenoma phenotype, provides evidence against the presence of a
defect in an MMR gene in family N.
As biallelic inactivation of APC occurs in most colorectal ade-
nomas and carcinomas
5
, we sequenced the APC ORF to identify
somatic mutations in each of the 11 tumors obtained from family
N. We characterized 18 somatic inactivating mutations, of which
15 were G:C→T:A transversions that included 14 nonsense
changes and 1 splice site mutation (Table 1 and Fig. 2a–c). The
other three mutations were two C:G→T:A transitions at CpG
dinucleotides and an instance of allelic loss (Table 1). Bialleic
inactivation was confirmed in six of the eight tumors with two
APC mutations.
We compared the proportion of G:C→T:A transversion muta-
tions detected in tumors from family N with 503 reported
somatic APC mutations from sporadic colorectal adenomas and
carcinomas, and with 308 somatic mutations from tumors asso-
ciated with familial adenomatous polyposis (FAP). We found
that the excess of G:C→T:A transversions in family N was highly
significant (15/18 versus 49/503 (P=2.77 × 10
–12
) and 15/18 ver-
sus 30/308 (P=7.69 × 10
–12
), respectively).
The most stable product of oxidative DNA damage
9
, 8-oxo-
7,8-dihydro-2′-deoxyguanosine (8-oxoG), is highly mutagenic
because it mispairs readily with adenine residues
14
, which leads
to an increased frequency of spontaneous G:C→T:Atransversion
mutations in repair-deficient bacteria and yeast cells
1–4
. In E. coli,
the enzymes mutM, mutY and mutT function synergistically to
protect cells from the deleterious effects of guanine oxidation
2
.
The DNA glycosylase mutM removes the oxidized base from 8-
oxoG:C base pairs in duplex DNA, the DNA glycosylase mutY
excises adenine misincorporated opposite unrepaired 8-oxoG
during replication, and the 8-oxo-dGTPase mutT prevents incor-
poration of 8-oxo-dGMP into nascent DNA. Human homologs
of mutM (OGG1)
15
, mutY (MYH)
6
and mutT (MTH)
16
have
been identified.
To determine whether an inherited defect in one of these
proteins was responsible for the pattern of somatic G:C→T:A
mutations in family N, we sequenced the coding regions of the
corresponding genes in a blood DNA sample from sibling II:1.
We identified two nonconservative amino-acid variants in
MYH (Fig. 3a,b): Tyr165Cys (an adenine-to-guanine substitu-
tion at nt 494 in exon 7) and Gly382Asp (a guanine-to-adenine
substitution at nt 1145 in exon 13). We did not identify any
missense variants or other probable pathogenic changes in
OGG1 or MTH.
We assayed both MYHvariants in blood DNA samples from all
members of family N and in 100 British control individuals with
no history of colorectal adenoma or carcinoma. In family N, the
three affected siblings were compound heterozygotes for the
mutations Tyr165Cys and Gly382Asp, and the unaffected family
members were either heterozygous for one of these variants or
normal (odds ratio≈50:1 assuming full penetrance in compound
heterozygotes; Fig. 3c). Each missense variant was also identified
once in different normal controls.
Because the guanine to adenine substitution that causes the
mutation Gly382Asp is located at the first base in exon 13, we
examined its potential affect on splicing and expression. Only 31
of 100 clones obtained by RT–PCR of normal colonic mucosa
total RNA from sibling II:1 carried the Gly382Asp allele, but no
aberrant splicing could be detected.
We looked for somatic muta-
tions of MYH in each of the 11
tumors both by denaturing
high-performance liquid chro-
matography (dHPLC) and sin-
gle-strand conformation
polymorphism (SSCP) analysis
of all exons, and by screening
for allelic loss through assaying
the exon 7 and 13 missense
variants. We did not identify
any somatic mutations to sug-
gest that MYH might function
as a tumor suppressor in a
manner analogous to the MMR
genes in HNPCC
13
. Neither
was there clear evidence that
the Tyr165Cys or Gly382Asp
variants were dominant to wild
type, because heterozygotes for
each variant were phenotypi-
cally normal. Instead, the
occurrence of the multiple ade-
noma phenotype in the three
compound heterozygotes sug-
gests transmission as an auto-
somal recessive trait.
We searched for germline
mutations of MYH, OGG1 and
MTH by sequence analysis of
their ORFs in 16 unrelated
individuals with between 3 and
roughly 50 colorectal adeno-
mas, with or without carci-
Table 1 • Somatic APC mutations identified in family N
Sample
a
Nucleotide change Amino-acid change No. of clones (x/y)
b
Sequence context
c
A1 2602G→T Glu868X 2/6 AGAAAAT
4351G→T Glu1451X 2/6 AG
AAGTA
A2 721G→T Glu241X NA AG
AAGCA
4381G→T Glu1461X 2/6 TG
AAAAG
A3 4717G→T Glu1573X 4/5 TG
AAATA
NI NI
A4 423-1G→T
d
NA 2/2 NA
4351G→T Glu1451X 6/6 AG
AAGTA
A5 601G→T Glu201X NA GG
AAGAA
4348C→T Arg1450X 3/6 NA
B2 3331G→T Glu1111X 7/10 AG
AAACA
LOH LOH NA
B4 3586C→A Ser1196X 3/7 TG
AAAAT
3856G→T Glu1286X 4/5 TG
AAATA
B5 604G→T Glu202X 3/6 AG
AACAA
3850G→T Glu1284X 6/6 TG
AAGAT
B6 2863G→T Glu955X 5/7 AG
AATAC
3949G→T Glu1317X 4/6 TG
AAGAT
C2b 1495C→T Arg499X 3/6 NA
NI NI
C1a
e
NI NI
a
For somatic APC mutations, we analyzed five adenomas from sibling II:1 (A1–5), four adenomas from sibling II:2 (B2,
B4, B5, B6), and one adenoma (C2b) and one carcinoma (C1a) from sibling II:3. Mutations are described according to
the established nomenclature system
29
. Biallelic mutations were found to be on opposite alleles in all tumors except
A2 and A5, for which this could not be determined.
b
Number of clones, where x represents the number with the muta-
tion and y represents the total number from that allele. In general, mutations were found in only a proportion of
clones. Nonmutated clones from the same allele most probably represent contaminating normal tissue. All mutations
were confirmed by an independent assay on a fresh PCR product.
c
Sequence context surrounding the coding region
G:C→T:A mutations (underlined); the sequence of the nontranscribed strand is shown except for the Ser1196X variant
in B4.
d
423-1G→T was shown to cause skipping of exon 4 of APC and is predicted to terminate the reading frame at
the seventh codon of exon 5.
e
C1a did not contain any identified APC mutations, despite re-sequencing of the ORF in
DNA from a second micro-dissected tumor sample. Sequence analysis of the coding regions of CTNNB1 and TP53 in
DNA from this carcinoma was also normal, which suggests that genes in an alternative tumorigenic pathway are
mutated. NA, not applicable; NI, not identified.
© 2002 Nature Publishing Group http://genetics.nature.com
letter
nature genetics •
volume 30 • february 2002
229
noma. We also screened all exons of MYH by dHPLC analysis in
42 unrelated individuals affected with colorectal cancer. We iden-
tified several frequent missense polymorphisms, Ser326Cys in
OGG1 (ref. 17), and Val22Met, His324Gln (ref. 6) and Ser501Phe
in MYH. The allele frequencies of the MYH polymorphisms were
not significantly different in the affected individuals from those
in 100 unaffected controls (Val22Met, 9% of normal chromo-
somes; His324Gln, 24%; Ser501Phe, 3%).
One individual, MA12, who had three adenomas and a carci-
noma, was compound heterozygous with respect to the unique
MYH missense variant Arg260Gln (779G→A) and the
Ser501Phe polymorphism. Analysis of the APC ORF in the four
tumors from MA12 showed two G:C→T:A transversions that
produce nonsense changes (Glu477X and Ser1344X), two
frameshift mutations and one instance of allelic loss. Other fam-
ily members were not available for study and the limited number
of tumors prevented us from
establishing a meaningful pat-
tern of mutation in APC.
Inherited factors are thought to
have a principal role in at least
15% of colorectal cancers
18
;
however, the negative findings
in the additional cases studied
here suggest that predisposi-
tion to colorectal tumors is not
associated commonly with
mutationsin MYH.
Comparison of MYH
homologs in diverse organisms
showed that there are identical
or similar amino acids at the
positions of the missense
changes identified in family N
(Fig. 4a,b). So far it has not
been possible to characterize
fully the protein product of
MYH; thus, to gain insight into
the functional consequences of
the missense variants, we
assessed the effects of the
equivalent E. coli mutY muta-
tions, Tyr82Cys and
Gly253Asp, on the intrinsic
rate of adenine removal from
DNA duplexes containing a
centrally located G:A or 8-
oxoG:A mismatch. Compared
with the wildtype protein, the
mutant proteins showed a
roughly 98% (Tyr82Cys) and
86% (Gly253Asp) reduction in
the rate for adenine removal
from the G:A substrate at 37 °C
(the rate constant describing
chemistry, k
2
, was 1.6 ± 0.04
min
–1
for wild type, 0.04 ± 0.01
min
–1
for Tyr82Cys and 0.22 ±
0.04 min
–1
for Gly253Asp).
Because the high affinity of
mutY for 8-oxoG:A sub-
strates
19
results in reaction
rates that are too fast at 37 °C to
be measured using our manual
methods
20
, we analyzed the
reaction rates with this duplex at 2 °C (Fig. 5). The Tyr82Cys
enzyme is so severely compromised in its catalytic activity that
minimal conversion of substrate to product was observed dur-
ing the time period that was monitored, and the Gly253Asp
enzyme shows an 85% decrease in the rate of adenine removal.
The marked effect of the Tyr82Cys mutation is consistent with
structural studies of mutY, which show that Tyr82 is located in
the pseudo-HhH motif (
79
G-X-G-Y-Y-A
84
) and suggest that
this residue has a role in mismatch specificity and flipping of
adenine into the base specificity pocket
7
. The reduction in
activity of the Gly253Asp mutant for the G:A substrate is simi-
lar to that observed with a truncated form of mutY that lacks
the carboxy-terminal third of the protein
21
. In the colonic
mucosa, the activity of the Gly382Asp allele of MYH might be
compromised further by the reduced expression that we noted
during RT-PCR analysis.
alignments
G A T C G A T C A C G T A C G T
A1
2602G→T
E868X
A C G T A C G T A C G T A C G T
A1
4351G→T
E1451X
wt mut
A C G T A C G T
m
B5
3850G→T
E1284X
confirmationclone sequencing
normal mutant
normal mutant
normal mutant normal mutant
normal
mutant
I
II
I
II
I
II
Fig. 2 Identification of somatic G:C→T:A mutations of APC in colorectal tumors from family N. Sequences of LD–PCR
product clones were aligned. Variants in two or more clones from the same allele (I or II) were confirmed by an inde-
pendent assay on a fresh PCR product. a, The G→T mutation at position 2602 (Glu868X) in adenoma A1 confirmed by
the direct sequencing of standard PCR products. b, The G→T mutation at position 4351 (Glu1451X) on the second APC
allele from adenoma A1 was confirmed by the direct sequencing of LD–PCR products. c, The G→T mutation at position
3850 (Glu1284X) in adenoma B5 was confirmed by restriction enzyme analysis. Arrows indicate the position of the
G:C→T:A mutations on the sequencing gels and the mutant allele on BfrI cleavage of a PCR product amplified from ade-
noma B5. m, DNA size marker (φ×174 HaeIII); mut, B5 adenoma DNA; wt, wildtype control DNA.
a
b
c
© 2002 Nature Publishing Group http://genetics.nature.com
The activity of mutY on mismatched
DNA substrates is influenced by the
immediate sequence context. Methyla-
tion interference experiments have
shown that mutY interacts with
purines, including the G:A mis-
matched bases and two bases on each
side
22
. Examination of the sequence
surrounding the 14 G:C→T:A muta-
tions in the coding region of APC in
family N showed that the two bases
immediately 3′ to the mutated G are
always two adenines, even though
other sequences that can undergo
G:C→T:A mutation to stop codons are
equally prevalent in the APC coding
region (216 GAA sites versus 213 non-
GAA sites,
χ
2
=13.28, P=2.7×10
–4
). In
addition, 13 of 14 sites match three or
all bases in a sequence extending one base 5′ (A/T) and three
bases 3′ (G/A,A,A) to the mutated G
AA (Table 1). These simi-
larities at the G:C→T:A mutated sites might reflect sequence
specificity in MYH.
Our findings implicate MYH in inherited predisposition to
colorectal tumors. The contribution of inherited defects of the
BER pathway both in this setting and more widely demands fur-
ther exploration.
Fig. 4 Evolutionary conservation
of the variant residues in MYH.
a,b, Comparison of the variant
residues Tyr165Cys (a) and
Gly382Asp (b) identified in family N
in Homo sapiens MYH (H. sap.,
accession number AAC50618) with
homologs from Mus musculus (M.
mus., AAG16632), Arabidopsis
thaliana (A. tha., CAB40991), Schizo-
saccharomyces pombe (S.pom.,
AAC36207), Hemophilus influenzae
(H. inf., C64091), Vibrio cholerae (V.
cho., AAF93625), Salmonella
typhimurium (S. typ., AAA27165)
and E. coli (AAG58092) using
Clustal W. Arrows indicate the
position of the variant residues.
Identical, conserved and semi-
conserved residues are shaded
black, dark gray and light gray,
respectively. // indicates the posi-
tion of 18 aa in A. thaliana that
are not present in the other
organisms.
Fig. 3 Identification and segregation of
germline MYH variants in family N. a,b, Direct
sequencing of constitutional DNA from sibling
II:1 identified an A to G substitution at nt 494 in
exon 7, which corresponds to Tyr165Cys (arrow,
a), and a G→A substitution at nt 1145 in exon
13, which corresponds to Gly382Asp (arrow, b).
c, Screening for Tyr165Cys by the amplification-
refractory mutation system (ARMS) and
Gly382Asp by a BglII digest showed that the
three affected siblings (filled symbols) were
compound heterozygotes for these MYH mis-
sense variants, whereas normal family members
(open symbols) were either heterozygous with
respect to one of the variants, or normal. N,
normal ARMS reaction; M, mutant ARMS reac-
tion. Arrows indicate the positions of the
mutant alleles.
letter
230 nature genetics •
volume 30 • february 2002
normal sib II:1 normal sib II:1
1145G→ A
G382D
494A→ G
Y165C
N M N M N M N M N M N M N M N M N M N M N MN M
ARMS
Bgl
II
digest
G382D
Y165C
GATC GATC GATC GATC
H.sap.
353 -PREESSATCVLE-QP---GALGAQ-ILLV-QRPNSGLLAGLWEFPSVTW--E--PSE---QLQRKA---L 406
M.mus.
335 -PREEYSATCVVE-QP---GAIGGPLVLLV-QRPDSGLLAGLWEFPSVTL--E--PSE---QHQHKA---L 389
A.tha.
361 -PRHDFCCVCVLEIHNLERNQSGGR-FVLV-KRPEQGLLAGLWEFPSVILN-E--EADS--ATRRNAINVY 423
S.pom.
296 QR-EERALVVIFQ-KT---DPSTKEKFFLIRKRPSAGLLAGLWDFPTIEFGQESWPKDMDAEFQKSIA-QW 360
H.inf.
233 MP-EKTTYFLILS-KN---GK-----VCLE-QRENSGLWGGLFCFP--QF--E--DKS---SLLH-----F 278
V.cho.
226 KP-VKATWFVMLY-HD---NA-----VWLE-QRPQSGIWGGLYCFP--Q---S--EIA---NIQT-----T 270
S.t yp.
228 LP-ERTGYFLLLQ-HN---QE-----IFLA-QRPPSGLWGGLYCFP--QF--A--RED---ELRE-----W 273
E. co li
228 LP-ERTGYFLLLQ-HE---DE-----VLLA-QRPPSGLWGGLYCFP--QF--A--DEE---SLRQ-----W 273
H.sap.
132 INYYTGWMQKWPTLQDLASASLE---EVNQLWAGLGYYSRGRRLQEGARKVVEELGGHMPRTAETLQQLLP 199
M.mus.
117 IDYYTRWMQKWPKLQDLASASLE---EVNQLWSGLGYYSRGRRLQEGARKVVEELGGHMPRTAETLQQLLP 184
A.tha.
180 MKYYKRWMQKWPTIYDLGQASLEN//EVNEMWAGLGYYRRARFLLEGAKMVVAGTEG-FPNQASSLMK-VK 264
S.pom.
81 KRYYTKWMETLPTLKSCAEAEYNT--QVMPLWSGMGFYTRCKRLHQACQHLAKLHPSEIPRTGDEWAKGIP 149
H.inf.
54 IPYFERFIKTFPNITALANASQD---EVLHLWTGLGYYARARNLHKAAQKVRDEFNGNFPTNFEQVWA-LS 120
V.cho.
47 IPYFERFLERFPTVHALAAAPQD---EVLHFWTGLGYYARARNLHKAAQMVVSEYGGEFPTDLEQMNA-LP 113
S.t yp.
49 IPYFERFMARFPTVTDLANAPLD---EVLHLWTGLGYYARARNLHKAAQQVATLHGGEFPQTFAEIAA-LP 115
E. co li
49 IPYFERFMARFPTVTDLANAPLD---EVLHLWTGLGYYARARNLHKAAQQVATLHGGKFPETFEEVAA-LP 115
Methods
Samples. For family N, we prepared DNA and RNA from venous blood
samples and from normal, adenoma and carcinoma tissue from colon that
had been snap frozen at colonoscopy or surgery, or micro-dissected from
tissue embedded in paraffin blocks. The nature of all tissues was verified
histologically. We also extracted DNA from blood samples from individu-
als with multiple colorectal adenomas (all of whom were normal on
sequencing of exon 4 and the alternatively spliced region of exon 9 of APC,
mutations of which are associated with attenuated FAP
5
), and from indi-
a
b
a
b
c
© 2002 Nature Publishing Group http://genetics.nature.com
letter
nature genetics •
volume 30 • february 2002
231
viduals either with colorectal cancer diagnosed at 40 yr or younger, or with
a family history of at least one first degree relative also affected by colorec-
tal cancer. We micro-dissected archived tumor tissues and extracted the
DNA using standard methods. This study was approved by the Bro Taf
Local Research Ethics Committee.
PCR and microsatellite analysis. We amplified exons 1–3 and 5–14 of APC
using published primers
10
, and exon 4 using ex4F and ex4R (PCR primers
are available upon request). For DNA extracted from paraffin-embedded
blocks, we amplified exon 15 of APC as 40 overlapping fragments. We
amplified exons 2–15 of CTNNB1, 2–11 of TP53, 1–16 of MYH, 1–8 of
OGG1 and 2–5 of MTH as 18, 11, 16, 11 and 4 fragments, respectively. For
DNA extracted from fresh tissue, we amplified exon 15 of APC either as a
single 6.67-kb long-distance (LD)–PCR fragment, or as two overlapping
3.59-kb and 5.07-kb LD-PCR fragments. We amplified exons 10–16 of
MYH as a 3.1-kb LD-PCR fragment. We tested DNA extracted from nor-
mal and tumor tissue for MSI using the markers D2S123, BAT 26, BAT 40,
Mfd15, DP1(APC), D18S69and BAT 25 (ref. 23).
RT–PCR and expression analysis. We used between 100 ng and 2 µg of
RNA for first-strand cDNA synthesis using oligo (dT)
15
and Superscript II
RNase H
–
Transcriptase (Invitrogen Life Technologies). We amplified
exons 1–14 of APC as a 1.958-kb fragment, as described
24
. To determine
the expression levels of individual APCalleles, we assayed the exon 11 poly-
morphism at Tyr486 in recombinant RT–PCR product clones by restric-
tion digestion or sequence analysis. To characterize aberrant splicing asso-
ciated with the 423-1G→T somatic mutation, we amplified exons 3–12 of
APC by RT–PCR and cloned and sequenced the products. To quantify the
expression level of the MYHallele with the Gly382Asp mutation, we ampli-
fied normal colonic mucosa cDNA from sibling II:1 and cloned and
assayed the products by a BglII digest.
Sequencing. We sequenced standard PCR products manually using the Ther-
moSequenase cycle sequencing kit (Amersham) and analyzed the reactions
on 6% polyacrylamide gels. For automated plasmid based sequencing, we
purified standard, LD–PCR and RT–PCR products using the PCR purifica-
tion kit (Qiagen) and cloned these into pGEM-T Easy (Promega) propogated
in E. coli strain JM109; we sequenced at least 12 recombinant clones of each
product. We carried out automated sequencing of RT–PCR product clones
spanning exons 1–14 of APC and LD–PCR products and clones spanning
exon 15 of APC, using two and eight overlapping bi-directional sequencing
reactions, respectively. We carried out automated sequencing of LD–PCR
product clones spanning exons 10–16 of MYH, RT–PCR product clones
spanning exons 3–12 of APC and 12–14 of MYH, and standard PCR product
clones, using M13 forward and reverse primers as described
25
. We aligned
sequence data for 12 or more clones (AlignIR v1.2, Li-Cor) and analyzed vari-
ants in 2 or more clones from the same allele by an independent assay on a
fresh PCR product, to confirm that they represented real mutations and were
not PCR or cloning-induced errors.
Assays for sequence variants. We assayed Tyr486 (1458C→T) in exon 11
of APC using an RsaI digest as described
10
. We assayed Glu1317Gln
(3949G→C) in exon 15 using a PvuII digest and Ala545 (1635A→G) in
exon 13 and Thr1493 (4479G→A), Ala1755 (5265G→A), Ser1756
(5268G→T) and Ser2497Leu (7490C→T) in exon 15 by sequencing. We
assayed the somatic APCmutations Glu1284X (3850G→T) and Glu1317X
(3949G→T) in exon 15 using a BfrI digest. We assigned the somatic APC
mutations to an allele by linking them to one of the polymorphic markers
using either standard, RT–PCR or LD–PCR, followed by cloning and
sequencing. We assayed the missense variants in MYH in 100 British nor-
mal control people. We carried out SSCP and dHPLC analysis at the RTm
and RTm+2 °C as described
26
.
Somatic APCmutation database and statistical analysis. We reviewed litera-
ture reports of characterized somatic APC mutations in colorectal tumors.
This included publications cited in the APCmutation database
27
and publica-
tions from the period 1991–2001, which we identified through a PubMed
search. We excluded reports of truncating mutations that were inconsistent
with the published cDNA sequence
11,28
, and putative missense mutations
because the evidence for their pathogenicity was inconclusive. We retrieved
data on 503 somatic mutations observed in sporadic tumors and 308 somatic
mutations observed in tumors associated with FAP and attenuated FAP. Sta-
tistical analyses were done by Fisher’s exact and the
χ
2
-test.
Site-directed mutagenesis and glycosylase activity assays. We carried out
site-directed mutagenesis of mutY with the primers Tyr82Cys_F and
Gly253Asp_F, and cloning, expression and purification of wildtype and
mutant mutY, as described
21
. To determine the effect of the Tyr82Cys and
Gly253Asp mutations on the intrinsic rate of adenine removal as compared
with wildtype enzyme, we carried out glycosylase assays under single-
turnover conditions ([DNA]<[MutY]) as described
19
, using a 30-bp duplex
containing a centrally located 8-oxoG:A or G:A base pair. We determined the
amount of active protein (wild type, 39%; Tyr82Cys, 53%; Gly253Asp, 58%)
using active-site titration methods
19
. We fitted the resulting data to a single
exponential equation, [P]
t
=A
0
[1 – exp(–k
obs
t)]. Under the conditions used
for these experiments, k
obs
approximates k
2
(ref. 19).
URL. Primer sequences and additional information can be found on our web
site, http://www.uwcm.ac.uk/study/medicine/medical_genetics/research/
tmg/projects/hMYH.html).
GenBank accession numbers. CTNNB1, X89579, NT_005980; TP53,
U94788; OGG1, AC066599 and AC011610; MTH, D35691, D35692,
D35693, D35694; published APC cDNA sequence, M74088 and M73548.
Acknowledgments
We thank A. Wilkie, I. Frayling and R. Snell for advice; I. Tomlinson for
advice and primers for exons 1–3 and 5–14 of APC; M.M. Slupska for
genomic sequence spanning the MYH locus; M. Krawczak for statistical
assistance; J. Best, J. Myring, S. Palmer-Smith, M. McDonald, L. Parry, A.
Radcliffe, N. Dallimore and C. Simpson for help with sample collection and
preparation; P. Davies for help with dHPLC analysis; and S. Owen and D.S.
Jackson for clinical assistance. This work was supported by grants from the
King Saud University through the Saudi Cultural Bureau, Tenovus, and from
the National Institutes of Health. S.S.D. is an A.P. Sloan research fellow and
A.L.L. is an NIH predoctoral trainee.
Received 27 August; accepted 20 December 2001.
1. Nghiem, Y., Cabrera, M., Cupples, C.G. & Miller, J.H. The mutY gene: A mutator
locus in Eschericia coli that generates G:C to T:A transversions. Proc. Natl Acad.
Sci. USA 85, 2709–2713 (1998).
2. Michaels, M.L. & Miller, J.H. The GO system protects organisms from the
mutagenic effect of the spontaneous lesion 8-hydroxyguanine (7,8-dihydro-8-
oxoguanine). J. Bacteriol. 174, 6321–6325 (1992).
wild type
time (min)
Y82C
G253D
product ( nM)
Fig. 5 Representative plots of single-turnover adenine glycosylase assays. Wild-
type, Tyr82Cys and Gly253Asp mutY were assayed for glycosylase activity at 2
°C with a 20 nM duplex substrate containing an 8-oxoG:A mismatch and 30 nM
mutY (active-site concentration). The rate constant k
2
was 1.2 ± 0.2 min
–1
for
wild type, <0.0006 min
–1
(estimated) for Tyr82Cys and 0.26 ± 0.05 min
–1
for
Gly253Asp. All values represent the mean ± s.d. of an average of at least four
separate determinations.
© 2002 Nature Publishing Group http://genetics.nature.com
letter
232 nature genetics •
volume 30 • february 2002
3. Moriya, M. & Grollman, A.P. Mutations in the mutY gene of Escherichia coli
enhance the frequency of targeted G:C to T:A transversions induced by a single 8-
oxoguanine residue in single-stranded DNA. Mol. Gen. Genet. 239, 72–76 (1993).
4. Thomas, D., Scot, A.D., Barbey, R., Padula, M. & Boiteux, S. Inactivation of OGG1
increases the incidence of G:C to T:A transversions in Saccharomyces cerevisiae:
Evidence for endogenous oxidative damage to DNA in eukaryotic cells. Mol. Gen.
Genet. 254, 171–178 (1997).
5. Fearnhead, N.S., Britton, M.P. & Bodmer, W.F. The ABC of APC. Hum. Mol. Genet.
10, 721–733 (2001).
6. Slupska, M.M. et al. Cloning and sequencing a human homolog (hMYH) of the
Escherichia coli mutY gene whose function is required for the repair of oxidative
DNA damage. J. Bactiol. 178, 3885–3892 (1996).
7. Guan, Y. et al. MutY catalytic core, mutant and bound adenine structures define
specificity for DNA repair enzyme superfamily. Nature Struct. Biol. 5, 1058–1064
(1998).
8. Lindahl, T. Instability and decay of the primary structure of DNA. Nature 362,
709–715 (1993).
9. Ames, B.N. & Gold, L.S. Endogenous mutagens and the causes of aging and
cancer. Mutat. Res. 250, 3–16 (1991).
10. Groden, J. et al. Identification and characterization of the familial adenomatous
polyposis-coli gene. Cell 66, 589–600 (1991).
11. Kinzler, K.W. et al. Identification of FAP locus genes from chromosome 5q21.
Science 253, 661–665 (1991).
12. Sulekova, Z. & Ballhausen, W.G. A novel coding exon of the human adenomatous
polyposis coli gene. Hum. Genet. 96, 469–471 (1995).
13. Peltomaki, P. Deficient DNA mismatch repair: a common etiologic factor for colon
cancer. Hum. Mol. Genet. 10, 735–740 (2001).
14. Shibutani, S., Takeshita, M. & Grollman, A.P. Insertion of specific bases during
DNA synthesis past the oxidation-damaged base 8-oxodG. Nature 349, 431–434
(1991).
15. Roldan-Arjona, T. et al. Molecular cloning and functional expression of a human
cDNA encoding the antimutator enzyme 8-hydroxyguanine-DNA glycosylase.
Proc. Natl Acad. Sci. USA 94, 8016–8020 (1997).
16. Sakumi, K. et al. Cloning and expression of cDNA for a human enzyme that
hydrolyzes 8-oxo-dGTP, a mutagenic substrate for DNA-synthesis. J. Biol. Chem.
268, 23524–23530 (1993).
17. Kohno et al. Genetic polymorphisms and alternative splicing of the hOGG1 gene
that is involved in the repair of 8-hydroxyguanine in damaged DNA. Oncogene
16, 3219–3225 (1998).
18. Kinzler, K.W. & Vogelstein, B. Lessons from hereditary colorectal cancer. Cell 87,
159–170 (1996).
19. Porello, S.L., Williams, S.D., Kuhn, H., Michaels, M.L. & David, S.S. Specific
recognition of substrate analogs by the DNA mismatch repair enzyme mutY. J.
Am. Chem. Soc. 118, 10684–10692 (1996).
20. Porello, S.L., Leyes, A.E. & David, S.S. Single-turnover and pre-steady-state kinetics
of the reaction of the adenine glycosylase mutY with mismatch-containing DNA
substrates. Biochemistry 37, 14756–14764 (1998).
21. Chmiel, N.H., Golinelli, M.P., Francis, A.W. & David, S.S. Efficient recognition of
substrates and substrate analogs by the adenine glycosylase mutY requires the C-
terminal domain. Nucleic Acids Res. 29, 553–564 (2001).
22. Lu, A.L., Tsai-Wu, J.J. & Cillo, J. DNA determinants and substrate specificities of
Escherichia coli mutY. J. Biol. Chem. 270, 23582–23588 (1995).
23. Dietmaier, W. et al. Diagnostic microsatellite instability: Definition and
correlation with mismatch repair protein expression. Cancer Res. 57, 4749–4756
(1997).
24. Ishioka, C. et al. Detection of heterozygous truncating mutations in the BRCA1
and APC genes by using a rapid screening assay in yeast. Proc. Natl Acad. Sci. USA
94, 2449–2453 (1997).
25. Cheadle, J.P. et al. Long-read sequence analysis of the MECP2 gene in Rett
syndrome patients: Correlation of disease severity with mutation type and
location. Hum. Mol. Genet. 9, 1119–1129 (2000).
26. Jones, A.C., Sampson, J.R., Hoogendoorn, B., Cohen, D. & Cheadle, J.P.
Application and evaluation of denaturing HPLC for molecular genetic analysis in
tuberous sclerosis. Hum. Genet. 106, 663–668 (2000).
27. Beroud, C. & Soussi, T. APC gene: database of germline and somatic mutations in
human tumours and cell lines. Nucleic Acids Res. 24, 121–124 (1996).
28. Joslyn, G. et al. Identification of deletion mutations and three new genes at the
familial polyposis locus. Cell 66, 601–613 (1991).
29. Antonarakis, S.E. & The Nomenclature Working Group. Recommendations for a
nomenclature system for human gene mutations. Hum. Mutat. 11, 1–3 (1998).
© 2002 Nature Publishing Group http://genetics.nature.com
