| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Carcinogenesis |
Laboratory of Molecular Pathology, Department of Pathology, University of Texas Southwestern Medical Center, Dallas, Texas 75235 [A. M. R., D. L.C., L. B. M., D. N., E. C. F.], and Departments of Medicine [M. S. G.] and Microbiology and Molecular Genetics [J. P. B.], University of Vermont College of Medicine, Burlington, Vermont 05405
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
90% of >80 different skin cancers examined. In
contrast to Trp53+/+ mice in which most
mutations in the Trp53 gene were located in exons 5–8, the
majority of the mutations in Trp53+/- mice
were at other exons. We observed a high predilection for C
T
transition mutations at a unique CpG site in codon 122 (exon 4) of the
Trp53 gene in Xpc-/-
Trp53+/- mice. This site is not part of a
pyrimidine dinucleotide. Mutations at this codon, as well as in codons
124 and 210, were observed exclusively in
Xpc-/- or Xpc+/- mice. Mutations at the corresponding codons (127 and 213) in the
human p53 gene have been reported in skin tumors from human
patients with xeroderma pigmentosum. Hence, mutations at codons 122
(125), 124 (127), and 210 (213) may constitute signatures for defective
or deficient nucleotide excision repair in mice (humans). In
Xpc-/- mice, the majority of mutations
were located at C residues in CpG sites, in which the C is presumably
methylated. A similar bias can be deduced from studies in human XP
individuals. |
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In the present study, we examined the mutational spectrum in the
Trp53 coding region in >100 different skin cancers from
Trp53+/+ and
Trp53+/- mice that were additionally
either wild-type, heterozygous mutant, or homozygous mutant for the
Xpc gene. We have confirmed the presence of a number of
previously identified codon mutational hot spots. Additionally, we have
identified several novel hot spots, especially in
Trp53+/- mice, in which recessive
mutations can be selected in tumors. Among these, we encountered a
prominent hot spot in codon 122 of the Trp53 ORF exclusively
in Xpc-/-Trp53+/- mutant mice. The mutations in this hot
spot involved a cytosine residue at a nondipyrimidine CpG site in which
the cytosine is believed to be methylated. We additionally observed
that mutations in codons 122, 124, and 210 were detected exclusively in
Xpc-/- or
Xpc+/- mice. A review of the literature
(7
, 12)
indicates that CT mutations in the
corresponding codons 127 and 213 of the human p53 gene have
been reported in skin cancers in individuals with the NER-defective
disease XP, but only in a single normal (non-XP) individual (who may
conceivably be genetically XP heterozygous). Hence, inactivating
mutations at these codons may be diagnostic of defective or deficient
NER in mice and humans.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
cDNA Synthesis and Sequencing Analysis.
Total RNA was extracted from frozen samples using Trizol Reagent (Life
Technologies, Inc.). RNA samples were treated with DNase I (Life
Technologies) and reverse-transcribed with the Super Script
Preamplification System kit (Life Technologies). To avoid amplification
of constitutive mutant alleles, amplification of Trp53 and
Xpc cDNAs was carried out independently using two pairs of
primers. For each pair, one primer binds to a region that is missing in
the mutant allele. PCR to amplify the Trp53 coding region
(GenBank accession no. K01700) was carried out with the following
primers:
(a) first set of primers: forward, 5'-CCTGGCTAAAGTTCTGTAGC-3' (nucleotides 22–41); reverse, 5'-GCCTGTCTTCCAGATACTCG-3' (nucleotides 776–757).
(b) second set of primers: forward, 5'-CCTGTCATCTTTTGTCCCTTC-3' (nucleotides 424–444); reverse, 5'-GCAGAGACCTGACAACTATC-3' (nucleotides 1520–1501).
PCR reactions were carried out according to the Expand High Fidelity PCR system (Boehringer Mannheim). Sequencing reactions were run on an ABI 377 automated DNA sequencer. Sequencing forward primers were 5'-TAGCATTCAGGCCCTCATCC-3' (nucleotides 97–116) and 5'-GAAGTCACAGCACATGACGGA-3' (nucleotides 640–660). Reverse primers used to confirm mutations on an independent PCR were 5'-AGGTGGAAGCCATAGTTGCC-3' (nucleotides 480–461), 5'-AGAGGCGCTTGTGCAGGTG-3' (nucleotides 1093–1075), 5'-TCTCAGCCCTGAAGTCATAAG-3' (nucleotides 1400–1380), and the one between nucleotides 776–757 (see above).
DNA Extraction and Southern Blot Analysis.
DNA was extracted according to standard procedures (16)
.
DNA from each sample (10 µg) was digested with StuI and
fractionated by electrophoresis in 0.7–1.0% agarose gels. DNA was
transferred to nylon membranes by alkaline blotting (17)
.
Probes were labeled with [
-32P]dCTP by
random primer extension (18)
. Prehybridization and
hybridization were performed according to standard procedures
(19)
. Membranes were autoradiographed at -70°C for
1–10 days. Probes used for hybridization were prepared by PCR using
oligonucleotide primers (exons 2–10) specific for the mouse
Trp53 gene based on published sequences (GenBank accession
no. K01700).
Restriction Digestion-based Assays to Detect Mutations in Codons
122, 124, and 210.
DNA from skin tumors (500 ng) was amplified by PCR (Platinum
Taq PCR system; Life Technologies). Specific primers were
used to amplify a region containing codon 122 (forward,
5'-CATCACCTCACTGCATGGACGATCT-3', nucleotides 264–288; reverse,
5'-AGAATATGAGAGAAAGGAGAAGAGGCT-3', located in intron 4), codon 124
(forward, 5'-CCTTGACACCTGATCGTTAC-3', located in intron 4; reverse,
5'-GAGCAAGAATAAGTCAGAAGC-3', located in intron 5), and codon 210
(forward, 5'CTACAAGAAGTCACAGCACATGACG-3', nucleotides 631–658;
reverse, 5'-GCTAGAAAGTCAACATCAGTCTAGG-3', located in intron 6). PCR
products were purified by the High Pure PCR Product Purification kit
(Boehringer-Mannheim) and divided into two parts. One part (25 µl)
was digested with a specific restriction enzyme (ApaLI for
codon 122, BseRI for codon 124, and BcgI for
codon 210) according to manufacturer’s instructions (New England
Biolabs, MA). The second half was subjected to the same
conditions except that the restriction digestion enzyme was not added.
Digested and undigested pairs were electrophoresed in 2% agarose gels
(Life Technologies). For samples digested with BseRI, 2.5%
agarose MS gels (Boehringer-Mannheim) were used. The presence of uncut
DNA treated with the respective enzyme was interpreted as possible
evidence for mutation. Independent PCR was then carried out. Purified
PCR products or alternatively the undigested band (cut out of the
agarose gels and purified by the QIAquick Protocol; Qiagen), were
sequenced in an ABI 377 automated DNA sequencer to confirm mutations.
Assays to Evaluate the Methylation Status of Codons 122 and 210.
Genomic DNA from unirradiated skin from either wild-type or
Xpc-/- mice was incubated with or
without a specific enzyme (ApaLI for codon 122;
BcgI for codon 210), according to manufacturer’s
instructions (New England Biolabs, Beverly, MA). None of these enzymes
cut methylated DNA. As a positive control, a PCR product containing the
wild-type sequence of codon 122 or codon 210 was treated identically.
After digestion, 100 ng of each sample were PCR amplified using
primers and conditions described above for amplification of these
codons. The same amount of DNA was also used for PCR amplification of a
region outside the Trp53 locus, where the ApaLI
site does not contain a CpG dinucleotide. Robust PCR amplification was
interpreted as indicative of methylation.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A description of the mutations observed in
Trp53+/- animals of all three possible
Xpc genotypes is presented in Tables 1
2
3
, and those
detected in Trp53+/+ mice are shown in
Tables 4
5
6
. Fig. 1
shows the amino acids targeted by these mutations and their
distribution in various genotypes. All codons in which mutations
occurred more than once in our study or which have been documented
previously as mutational sites in UVB radiation-induced skin cancer in
mice or humans are defined as hot spots for the purposes of this study.
|
|
|
|
|
|
|
documents mutations observed in the Trp53 ORF in
Xpc+/+ Trp53+/-
mice. A total of 11 mutations were detected in 11 skin tumors examined.
In three instances, two different mutations were observed in the same
tumor. In the three tumors in which no mutations were detected, we
could not eliminate the possibility that other mechanisms might have
inactivated Trp53 protein. All of the mutations were on the
NTS of the Trp53 gene, reflecting the well-recognized bias
for preferential NER of the TS of transcriptionally active genes in
NER-proficient cells exposed to UVB radiation (21)
. This
bias is, as expected, exaggerated in Xpc mutant mice because
such mice (as well as XP-C humans) manifest defective NER specifically
of the NTS of transcriptionally active genes (22
, 23)
.
Consistent with previous studies (24, 25, 26, 27)
, all mutations
were at dipyrimidine sites, predominantly affecting the C residue of a
dipyrimidine pair and resulting in CT transitions. Three TA
transversions and a single TG transversion were also noted (Table 1)
.
Examination of the nucleotide sequence context at the sites of
mutations revealed that the affected nucleotide was in a stretch of
3–6 consecutive pyrimidines (Table 1)
. These results are consistent
with previous studies showing that polypyrimidine tracts in DNA are
preferred sites for the formation of CPDs and 6-4 photoproducts
(28)
. The C residue in CpG dinucleotides is typically
methylated in the Trp53 gene (29)
. When
mutations involved a C residue, we observed that the C was in a CpG
dinucleotide in 27% of instances in this genotype (Fig. 2)
. This is consistent with the previously reported frequency of
mutations at methylated C residues in wild-type mice (26%; Refs.
24, 25, 26, 27
).
|
. Mutations in these exons have not been documented previously in UV
radiation-induced skin cancer (24, 25, 26, 27)
. Although this may
in part reflect a bias for examining only exons 5–8, a more cogent
explanation is that recessive mutations can readily manifest in mice
with only one functional Trp53 allele.
Two of the mutations resulted in nonsense codons
(L25Stop; Fig. 1
, Table 1
). The remaining nine
mutations resulted in missense amino acid substitutions. Amino acid
substitutions at codons 270 and 275 (both in exon 8) occurred twice
(Fig. 1
; Table 1
). Codons 270 and 275 are the commonest sites of
mutation in skin cancers in wild-type mice, as determined from a
database of 120 mutations that we compiled from the literature
(Refs. 24, 25, 26, 27
; Fig. 1
, top panel). The human
p53 codon corresponding to mouse codon 275 (codon 278) is
the second most frequent site of mutation in human skin cancers
(7)
, and codon 273 (corresponding to mouse codon 270) is
also frequently mutated in human skin cancers (7)
.
Previous studies in mice have not identified codon 245 as a mutational
hot spot (Fig. 1
; Table 1
). However, mutations in the corresponding
human codon 248 have been reported as the most frequent site of
mutation in skin cancers (7)
. The amino acid substitution
R245C was identified in other Xpc
genotypes in the present study (see later). Mutations at codon 238,
another mutational hot spot in the wild-type mouse database (Fig. 1)
,
were also encountered in other genotypes (see later).
Amino acid substitutions at codons 93, 106, and 200 were observed in
tumors from Xpc+/+Trp53+/- mice (Fig. 1
and Table 1
). Mutations
at codons 93 and 200 have not been reported previously in skin cancers
from mice or humans. However, mutations in human codon 109
(corresponding to mouse codon 106) have been reported in skin cancer
(7)
.
Trp53 Mutations in Skin Cancers from
Xpc+/- Trp53+/-
Mice.
Table 2
shows the mutational spectrum in tumors from double heterozygous
(Xpc+/-Trp53+/-) mice. Once again, a strong bias for
mutations in the NTS is evident. However, two mutations were deduced to
be in the TS. With the exception of a single mutation in codon 122, the
remaining 17 mutations were at dipyrimidine sites, and 15 of these
involved a C residue. In one instance [one of the two mutations in
case 2017-T2 (Table 2)
], we observed the dinucleotide substitution
CATT. Conceivably, the A residue in the sequence TCA (Table 2)
was
misreplicated during DNA synthesis across the 5' flanking damaged
dipyrimidine site (Table 2)
. As was the case with skin cancers in
Xpc+/+ Trp53+/-
mice (Table 1)
, a significant number of the mutations (35%) involved C
residues at (presumably methylated) CpG sites (Table 2
; Fig. 2
). In
contrast to the results obtained in
Xpc+/+ Trp53+/-
mice (Table 1)
, where all of the mutations involved mononucleotide
changes, in double heterozygous mutants
(Xpc+/-Trp53+/-) a significant number of mutations
involved dinucleotides (Table 2)
. Additionally, the trinucleotide
mutation (TCCATT) was observed (Table 2)
.
As in Xpc+/+Trp53+/- mice (Table 1)
, a number of the
mutations in Xpc+/-Trp53+/- mice were located outside exons 5–8
(Fig. 1)
. These resulted in the amino acid substitutions
S58C, L61H,
R62Stop, T122M,
T326S, A344D, and
E346Stop (Table 2)
. None of these mutations
have been identified previously in skin cancers in mice or humans. We
suggest that their location in sites other than exons 5–8 once again
reflects their recessive nature and/or a bias against their detection
in previous studies.
Amino acid substitutions at the previously identified hot spot codons
270 and 275 were again noted (Fig. 1)
. New hot spots were identified at
codons 210 (the single frameshift mutation observed in our study), 122,
and 124 (Fig. 1
; Table 2
). The CT transition mutation at codon 210
has been reported previously in a tumor from a mouse defective in the
Xpa gene (6)
. This mutation has also been
reported in the corresponding human codon in 2 skin cancers, one of
which was from an XP individual (7
, 30) . The hot spot at
codon 124 (human codon 127) has been identified in skin cancers from
two human XP-C individuals (12)
but not in non-XP
individuals.
The amino acid substitution H176Y, observed
once in our entire data set, is a prominent hot spot in the wild-type
mouse skin cancer database (Fig. 1)
. The corresponding human codon 179
is also a prominent site of mutation in skin cancers in both XP and
non-XP individuals (7)
. Mutations in codon 191 of the
Trp53 gene have been observed in wild-type mice and in the
corresponding human codon in a solar keratosis (7)
.
Trp53 Mutations in Skin Cancers from
Xpc-/- Trp53+/-
Mice.
Table 3
and Fig. 1
show the mutational spectrum in skin cancers in
Xpc-/- Trp53+/-
mice. In this genotype, we observed a marked predominance of mutations
in codon 122 of the Trp53 ORF. Of 22 tumors
examined, 14 (64%) involved a C residue at a nondipyrimidine site in
codon 122. As already indicated, this mutation has not been reported
previously in skin tumors associated with UV radiation exposure in mice
or humans. In five instances, the affected C residue was the only
altered nucleotide, resulting in CT transitions (Table 3)
. In the
remaining nine cases, the adjacent A was also mutated, resulting in the
tandem mutation ACTT in seven instances and the tandem mutation
ACCT in 2 instances (Table 3)
. The mutated C residue in codon 122 is
in a CpG dinucleotide. CpG dinucleotides were additionally implicated
in two other mutations at dipyrimidine sites in
Xpc-/- Trp53+/-
mice (cases 406 and 653-T1; Table 3
). Hence, in this genotype 16 of 24
(67%) of the mutations involve a C at (presumably methylated) CpG
sites (Fig. 2)
.
We have determined that the tandem mutation ACTT in codon 122 is
strictly UVB radiation
dependent.5
It remains to be determined whether the CT mutations are also
qualitatively or quantitatively dependent on UV radiation exposure,
because such mutations can arise from spontaneous deamination of C at
methylated CpG sites (31)
. No mutations were detected at
codon 122 in genomic DNA from unirradiated normal mouse skin. Hence,
this mutation is not part of the genetic background of the mouse
strains used. In addition, mutations detected in cDNA were confirmed in
tumor genomic DNA in some cases.
The presence of mutations in codon 122 in skin tumors from mice with a single functional Trp53 allele is consistent with the notion addressed above that mutations outside exons 5–8 are often recessive and hence are not usually expressed in animals that are wild type for Trp53. However, the extraordinary frequency of these mutations in mice that are additionally defective in NER (Xpc-/-) suggests the operation of one or more selecting factors (see "Discussion").
With regard to other mutations in this genotype, amino acid
substitutions at codons 192 have been observed in skin cancers in
wild-type mice or (in the corresponding codon) in humans, and mutations
at codon 193 have been reported in the corresponding codon in humans.
Among eight individuals with mutations at codons 195 or 196, six were
XP patients from three different genetic complementation groups
(32)
. The amino acid substitutions
P216L and Y217N have
not been reported previously in skin cancers in mice. However,
mutations in the human codon 220 have been reported in two skin cancers
(7)
.
Another example of the hot spot amino acid substitution
R210C was identified in skin cancers from
Xpc-/- Trp53+/-
mice (Fig. 1
; Table 3
). We also detected the triple transition mutation
CCCTTT, resulting in the amino acid substitution
P188F (Table 3)
. This codon is mutated in skin
cancers from both humans and mice (7
, 24, 25, 26, 27)
. The amino
acid substitution V119G observed in two tumors
represents another mutational hot spot. Like codon 122, codon 119 is
located in exon 4 and has not been identified previously as a site of
mutation in either mice or humans. The same is true of the amino acid
substitution Q19Stop in codon 2 (Table 3)
.
In summary, in UVB radiation-induced skin cancers in mice in which only a single Trp53 allele is present, the following salient observations emerged.
(a) Codons 25, 119, 122, 192, 210, 270, and 275 are
mutational hot spots based on the present studies (Fig. 1)
. The first
three codons lie outside exons 5–8, the most frequently documented
sites for mutations in the mouse/human Trp53/p53
gene (Fig. 1)
. Pooling our results with those from studies published
previously involving either mice or humans indicates that codons 106,
126, 176, 188, 191, 193, 217, 238, and 245 (or their corresponding
human codons) are also mutational hot spots in skin cancer (Fig. 1)
.
(b) In the genotype Xpc-/-Trp53+/-, mutations at codon 122 of the
Trp53 gene are especially prominent (Fig. 1
; see
"Discussion").
(c) In a significant fraction of the CT transition
mutations, which dominate the mutational spectrum in most studies on UV
radiation-induced mutagenesis (24, 25, 26, 27
, 33)
, the C residue
is in a CpG dinucleotide and is hence presumably methylated. In
Xpc-/- Trp53+/-
mice, the majority of mutations are associated with such sites.
Trp53 Mutations in Skin Cancers from
Trp53+/+ Mice of All Xpc
Genotypes.
On the basis of the correlation between the presence of (presumed
recessive) mutations outside exons 5–8 and the presence of only one
Trp53 allele, we did not expect to observe mutations outside
exons 5–8 in Trp53+/+ mice unless a
second mutation inactivated the remaining Trp53 allele. Such
was indeed the case (Tables 4
5
6)
. In view of the extensive information
available in the literature on Trp53 mutations in skin
cancers in wild-type mice (24, 25, 26, 27
, 33)
, we only sequenced
the Trp53 gene in two tumors from this genotype. In both
cases we observed CT transitions at dipyrimidine sites on the NTS,
resulting in further examples of the hot spot amino acid substitution
R270C (Fig. 1
; Table 4
).
Trp53 Mutations in Skin Cancers from
Xpc+/- Trp53+/+ Mice.
Four of the 10 (40%) mutations observed in the
Xpc+/- Trp53+/+
genotype involved a C residue in a CpG dinucleotide (Fig. 2)
. The amino
acid substitution R270C dominated the spectrum
of mutations observed in tumors from these mice (Table 5
and Fig. 1
). Additional examples of the amino acid substitutions
P275S and S238F, as
well as a second example of the amino acid substitution
S124F, were also encountered. Once again the
latter hotspot correlated exclusively with the
Xpc-/- or
Xpc+/- genetic background (Tables 2
and 5
; Fig. 1
). A mutation in codon 249 has been reported previously in
wild-type mice (24, 25, 26, 27)
, and a silent mutation at the
corresponding codon has been observed previously in humans
(7)
. However, the substitution
D256V, resulting from a TA transversion on
the TS (Table 5)
, is novel in both mice and humans (7
, 24, 25, 26, 27
, 33)
. Like other codons rarely mutated in the
Trp53/p53 gene in previous reports, the
D256V mutation is likely recessive, because a
second amino acid substitution (R270C) was
observed in the same tumor (Table 5)
.
Trp53 Mutations in Skin Cancers from
Xpc-/- Trp53+/+ Mice.
Table 6
presents the mutational spectrum in
Xpc-/- Trp53+/+
mice. The majority of the mutations are CT transitions. As with the
mutational spectrum in the Xpc-/-Trp53+/- genotype, C mutations at CpG sites are
prominent, in this case accounting for 13 of 23 (57%) of the mutations
(Fig. 2)
. Hence, the high frequency (67%) of cytosine mutations at CpG
sites observed in Xpc-/-Trp53+/- mice is not the exclusive result of
the predominance of mutations at codon 122 in this genotype (Fig. 2)
.
All mutations in Xpc-/-Trp53+/+ mice were correlated with the NTS, and
all except two (at codon 122) were at dipyrimidine sites. In six
tumors, no mutations were detected in the Trp53 ORF.
The previously noted hot spot codons 238, 245, and 270 are represented
in Xpc-/-Trp53+/+ mice (Fig. 1
; Table 6
). Additionally,
other examples of mutations at codons 122, 124, 210, and 217 were noted
(Fig. 1
; Table 6
). Once again mutations at codons 122, 124, and 210
correlated exclusively with the Xpc-/-
or Xpc+/- genotypes (Fig. 1)
. Mutations
in codons 127, 148, and 174 have been documented in mice and/or in the
corresponding human codons in skin cancers (7
, 24, 25, 26, 27)
.
The amino acid substitutions P30S and
S96F are novel mutations in exons 3 and 4,
respectively (Fig. 1
; Table 6
). Both are one of two mutations in the
tumors in question. The same is true for the two examples of the
T122L amino acid substitution in
Xpc-/- Trp53+/+
mice. This observation supports the contention that the
P30S, S96F, and
T122L/M amino acid substitutions are recessive
and require inactivation of the second Trp53 allele in
Trp53+/+ cells to be selected in tumors.
In an effort to prove that the T122L mutation
and a second (inactivating) mutation are present on different
Trp53 alleles, we cloned the PCR-derived products from case
363-T1 (Table 6)
and sequenced multiple independent clones. In 50% of
the clones examined, the T122 mutation and the
silent L127 mutation cosegregated, whereas in the
remaining 50%, only the P148 mutation was
detected (data not shown). These results suggest that all three
mutations occurred in both Trp53 alleles in the same cell.
Fig. 3
summarizes the spectrum of mutations at dipyrimidine sites observed in
a total of 71 mutations shown in Tables 1
2
3
4
5
6
. The majority of mutations
at dipyrimidine sites were CT transitions (75%), consistent with
previous studies. Transversions and complex mutations involving more
than a single nucleotide comprised 17 and 4%, respectively.
|
; Tables 1
2
3
4
5
6
). Additionally, amino acid
substitutions at the corresponding human codons 125 and 127 have not
been reported in skin tumors from non-XP humans (7)
.
However, consistent with the results in mice, mutations at human codon
127 (mouse codon 124) have been observed in skin cancers from two XP-C
cases (12)
. Additionally, the CT mutation at codon 210
has been reported in an Xpa-/- mouse
(6)
and in the corresponding human codon (213) in both an
XP and a non-XP patient (7
, 30)
. We are investigating the
possibility that the latter individual may be heterozygous for an XP
gene. This overall correlation suggests that mutations at mouse/human
codons 122/125, 124/127, and 210/213 of the
Trp53/p53 gene in skin cancers associated with
exposure to UV radiation may be signatures for defective or deficient
NER.
To further test the correlation between signature mutations at codons
122, 124, and 210 and the Xpc genotype in mice, we developed
restriction digestion-based assays for the detection of the
T122, S124, and
R210 mutations (see "Materials and Methods")
and examined 50 additional tumors. Once again, mutations were not
observed at these codons in tumors from
Xpc+/+ animals. In contrast, additional
examples of all three mutations were detected in
Xpc+/- and
Xpc-/- mice with either Trp53
genotype. Table 7
shows the frequency of mutations at codons 122, 124, and 210 detected
by both sequencing cDNAs or using the restriction enzyme-based data.
|
; Tables 1
2
3
4
5
6
). Additional studies are required to determine whether these also
represent signature mutations associated with deficient or defective
NER. |
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Our studies demonstrate that hemizygosity for Trp53 in mice can unmask recessive mutations in sites other than exons 5–8, particularly exon 4. Despite the fact that most mutational analyses use mice/humans that are genetically wild-type for Trp53/p53, studies that focus exclusively on exons 5–8 can clearly result in a bias with respect to the nature and distribution of mutations in the Trp53 gene and may miss significant information (35) . Indeed, one study on human skin cancer noted that close to 50% of the mutations were located in exon 4 (36) . Direct examination of the dominant or recessive nature of the mutant alleles identified in our studies is in progress.
A significant observation from our studies is the high frequency of mutations in codon 122 of the mouse Trp53 gene in Xpc-/- Trp53+/- mice. Studies with Xpa-/- mice are in progress to determine whether this mutational preference is specific to mice defective in the Xpc gene [which retain the ability to carry out NER of the TS of transcriptionally active genes (6) ] or results from any defect in NER. Preliminary studies have shown that mutations in this codon are not endogenous to the mouse strains used and that at least one of the mutations observed is specifically related to UVB radiation exposure.
Structural analysis of human wild-type p53 protein indicates that the side chain of amino acid residue T125 (equivalent to mouse T122) forms hydrogen bonds with residues G117 and/or R282. These hydrogen bonds would likely be disrupted by mutation to either L or M. In addition, the side chains of L and M are bulkier than that of T, and molecular modeling of these amino acid substitutions in the human protein suggests that unfavorable steric hinderances would result, which are expected to disrupt the structure of the DNA-binding domain.
The presence of the T122 mutation in mice that are heterozygous for Trp53 is readily explained by the suggestion that the mutation is recessive in nature. However, a salient question is why mutations in this particular codon are so prevalent specifically in Trp53+/- mice that are also defective in NER. The following (nonexclusive) scenarios merit consideration. All these scenarios assume that the photoproduct(s) resulting in mutations are on the NTS, concordant with the requirement of the Xpc gene for NER of this strand of the DNA duplex:
(a) Either the A or the C (or both) residues in the
trinucleotide sequence ACG (codon 122) may be hot spots for a
nondipyrimidine photoproduct, the repair of which specifically requires
NER, or the Xpc protein. The precise nature of the base damage at codon
122 is presently unknown. However, several generic possibilities merit
consideration. CT transitions may arise from spontaneous deamination
of 5-methylcytosine, the frequency of which may be enhanced by exposure
to UVB radiation. In this case, a putative photoproduct in the
trinucleotide ACG may involve the adjacent A rather than the suspected
C residue. Alternatively, 5-methylcytosine may be chemically altered by
exposure to UVB radiation to form photoproducts (such as cytosine
hydrates), the repair of which may require NER. Finally, UVB radiation
may generate photoproducts at either the C, the adjacent A, or both
nucleotide residues. The frequency of these putative photoproducts may
be strongly influenced by the DNA sequence context. In this regard, it
is interesting to note that the tetranucleotide sequence CACpG at codon
122 is unique in the entire Trp53 ORF.
(b) Codon 122 may be a preferred site for mutations by some sort of error-prone translesion synthesis mechanism. Such a mechanism may operate preferentially in NER-defective mice.
(c) Trp53 protein carrying the amino acid substitutions
T122L/M may acquire a gain of function that
renders it highly oncogenic. Hence, tumors with this mutation may be
preferentially selected in mouse skin.
The CT transition at codon 122 (but not the double mutations ACCT
or ACTT) has been reported in skin cancers in wild-type mice exposed
to beta radiation (37)
. Codon 122 is also a frequent site
of spontaneous mutations in the Trp53 pseudogene of
wild-type mice (38)
. Interestingly, this region is the
most significant site of nucleotide sequence divergence between the
Trp53 ORF and its pseudogene. These observations have led to
the suggestion that diversification of mouse subspecies may have been
driven during periods of enhanced levels of natural radiation
(38)
. A simple CT transition mutation in codon 125 of
the human p53 gene has been reported in a spontaneous lung
cancer in a Japanese patient (39)
. Once again, the
interesting question arises as to whether this individual was an XP
heterozygote.
Our studies confirm multiple mutational hot spots in the
Trp53 gene in skin cancers in mice, many of which have been
identified previously in either wild-type or Xpa mice
(6)
, and in some cases in normal or XP humans
(7)
. It is remarkable that both in the present study and
in reports published previously concerning UVB radiation or
sunlight-related skin cancer, mutations at codons 122, 124 and 210 (or
at the corresponding human codons) were, with a single exception,
exclusively observed in NER-defective or -deficient mice and humans. In
our study, these signature mutations were observed in 42 of 117 (36%)
of skin cancers from Xpc-/- or
Xpc+/- mice, with not a single example in
29 tumors from wild-type mice, regardless of the Trp53
genotype, or among 140 mutations in the Trp53 gene
associated with UVB radiation-induced skin cancer in wild-type mice
reported in the literature (24, 25, 26, 27
, 33)
. These
observations suggest that certain sites in the Trp53 gene
are especially vulnerable to UVB radiation-induced mutagenesis in the
absence of NER. Additionally, these observations suggest a relatively
simple method for screening mammalian cells for defective or deficient
NER.
We failed to detect mutations in the coding region of the remaining Xpc allele in Xpc+/- mice with skin cancer (2) . We cannot exclude the possibility that Xpc protein was inactivated by some other mechanism(s). However, it would appear that mice (and humans) may be at greater risk for skin cancers associated with mutations in Trp53/p53 attributable to haploinsufficiency of an XP locus. As shown in the accompanying report, Xpc+/- mutant mice are significantly more cancer-prone than Xpc+/+ controls in both Trp53 wild-type and heterozygous genetic backgrounds (2) .
In Xpc-/- mice, the majority of
mutations affected C residues at CpG sites, regardless of the
Trp53 genotype. Similar conclusions derive from a comparison
of mutations in human XP and non-XP individuals (Fig. 2
; Refs. 7
, 12,
and 32
). Assuming that most if not all CpG
sites in the mouse Trp53 coding region are methylated, this
suggests that methylated CpG sites are preferentially repaired in
NER-proficient mice. This conclusion is consistent with studies
indicating that methylation at dipyrimidine sites
(PyrmCpG) in the human p53 gene is
associated with an increased rate of formation of CPDs and with slower
NER (28
, 40)
. Our results additionally indicate that
mutational hot spots at methylated C residues are not confined to
dipyrimidine sites after exposure to UVB radiation.
In conclusion, we note that three of the codon hotspots for mutation in
Trp53 in Xpc-/-Trp53+/+ mice exposed to UVB radiation are
identical to those in XP humans exposed to sunlight (Fig. 4)
. This and other observations noted in this study indicate that
genetically engineered mice that are defective in NER provide
informative models for studying various aspects of the pathogenesis of
skin cancer in humans associated with exposure to solar radiation and
presumably to other carcinogens.
|
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 These studies were supported by Research Grant
CA44247 (to E. C. F.) and by postdoctoral fellowships from the
American Cancer Society (to D. L. C.) and The Friends of the Center
for Human Nutrition, University of Texas Southwestern Medical Center
(to L. B. M.). 
2 Present address: Life Technologies, Inc.,
Rockville, MD 20850.
3 To whom requests for reprints should be
addressed at, Southwestern Medical School, 5323 Harry Hines Boulevard,
Dallas, TX 75235-9072. E-mail: friedberg.errol{at}pathology.swmen.edu
4 The abbreviations used are: NER, nucleotide
excision repair; ORF, open reading frame; CPD, cyclobutane pyrimidine
dimer; XP, xeroderma pigmentosum; Xpc, xeroderma
pigmentosum group C gene; NTS, nontranscribed strand; TS, transcribed
strand.
5 D. Nahari, D. Cheo, and E. C. Friedberg,
unpublished observations.
Received 8/31/99. Revised 12/ 6/99. Accepted 1/19/00.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. W.P. Wijnhoven, E. N. Speksnijder, X. Liu, E. Zwart, C. Th. M. vanOostrom, R. B. Beems, E. M. Hoogervorst, M. M. Schaap, L. D. Attardi, T. Jacks, et al. Dominant-Negative but not Gain-of-Function Effects of a p53.R270H Mutation in Mouse Epithelium Tissue after DNA Damage Cancer Res., May 15, 2007; 67(10): 4648 - 4656. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. R. Denver, S. Feinberg, C. Steding, M. Durbin, and M. Lynch The Relative Roles of Three DNA Repair Pathways in Preventing Caenorhabditis elegans Mutation Accumulation Genetics, September 1, 2006; 174(1): 57 - 65. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. G. Khan, K.-S. Oh, T. Shahlavi, T. Ueda, D. B. Busch, H. Inui, S. Emmert, K. Imoto, V. Muniz-Medina, C. C. Baker, et al. Reduced XPC DNA repair gene mRNA levels in clinically normal parents of xeroderma pigmentosum patients Carcinogenesis, January 1, 2006; 27(1): 84 - 94. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. M. Hoogervorst, W. Bruins, E. Zwart, C. Th.M. van Oostrom, G. J. van den Aardweg, R. B. Beems, J. van den Berg, T. Jacks, H. van Steeg, and A. de Vries Lack of p53 Ser389 Phosphorylation Predisposes Mice to Develop 2-Acetylaminofluorene-Induced Bladder Tumors but not Ionizing Radiation-Induced Lymphomas Cancer Res., May 1, 2005; 65(9): 3610 - 3616. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Resnick and A. Inga Functional mutants of the sequence-specific transcription factor p53 and implications for master genes of diversity PNAS, August 19, 2003; 100(17): 9934 - 9939. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y.-H. You, P. E. Szabo, and G. P. Pfeifer Cyclobutane pyrimidine dimers form preferentially at the major p53 mutational hotspot in UVB-induced mouse skin tumors Carcinogenesis, November 1, 2000; 21(11): 2113 - 2117. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. C. Friedberg Nucleotide Excision Repair and Cancer Predisposition : A Journey from Man to Yeast to Mice Am. J. Pathol., September 1, 2000; 157(3): 693 - 701. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
