Exonucleolytic proofreading and DNA mismatch repair (MMR) act in series to maintain high-fidelity DNA replication and to avoid mutagenesis. MMR defects elevate the overall mutation rate and are associated with increased cancer incidence. Hypermutable colorectal and endometrial tumors with functional MMR were recently reported to carry amino acid substitutions in the exonuclease domain of DNA polymerase ϵ (Polϵ). This created a notion that loss of the proofreading activity of Polϵ is an initiating cause of some sporadic human cancers. In this study, we identified a somatic P286R substitution in the conserved ExoI motif of Polϵ in a collection of 52 sporadic colorectal tumor specimens. This change has been repeatedly observed in colorectal and endometrial tumors in previous studies despite many possible ways to inactivate Polϵ proofreading. To understand the reasons for the recurrent appearance of the P286R variant, we characterized its functional consequences using the yeast model system. An analogous substitution in the yeast Polϵ produced an unusually strong mutator phenotype exceeding that of proofreading-deficient mutants by two orders of magnitude. This argues that the P286R mutation acts at some level other than loss of exonuclease to elevate cancer risk. Heterozygosity for the variant allele caused a strong mutator effect comparable with that of complete MMR deficiency, providing an explanation for why loss of heterozygosity is not required for the development of Polϵ-mutant human tumors. Cancer Res; 74(7); 1895–901. ©2014 AACR.

Cancer cells contain thousands of mutations not present in normal cells. Changes in oncogenes and tumor suppressor genes directly promote carcinogenesis. In addition, the genomic instability leads to a genetic heterogeneity of tumors, contributing to their ability to adapt and develop resistance to therapy (1). For most cancers, the mechanisms underlying the high frequency of genetic changes are unknown. Maintenance of the low mutation rate relies heavily on three DNA replication fidelity mechanisms: (i) accurate nucleotide selection by the replicative DNA polymerases (Pol) α, δ, and ϵ, (ii) proofreading of errors by the 3′-exonuclease activity of Polδ and Polϵ, and (iii) correction of mispairs by the DNA mismatch repair (MMR) system. These processes act in series keeping the mutation rate per cell division at a low level of less than 10−10. Defects in any of the three mechanisms accelerate tumorigenesis in mice (2, 3), and inherited mutations in the human MMR genes cause predisposition to colorectal cancer in Lynch syndrome (4). The MMR gene MLH1 is also somatically inactivated in approximately 12% of sporadic colorectal cancer.

A recent large-scale molecular characterization of sporadic colorectal cancer revealed that although MMR-deficient cancers are hypermutated, tumors with the highest mutational load had alterations in Polϵ rather than MMR defects (5). Similarly, hypermutated sporadic endometrial tumors frequently contained Polϵ alterations (6, 7). In both cases, the changes affected conserved amino acid residues in the exonuclease domain, suggesting that inactivation of exonuclease activity was responsible for the hypermutator phenotype. In addition, germline mutations affecting the exonuclease domains of Polϵ and Polδ were found to cause a high-penetrance hereditary colorectal cancer predisposition (8). These discoveries strongly suggested that loss of proofreading activity of replicative DNA polymerases is the initiating cause of some hereditary and sporadic human cancers.

In this work, we assessed the status of the exonuclease domain of Polϵ in sporadic colon and rectum adenocarcinomas from 52 patients undergoing treatment at the University of Nebraska Medical Center (UNMC; Omaha, NE). We identified a P286R substitution in the conserved ExoI motif essential for the proofreading activity. This change was observed previously in two other colon tumors (9) and 14 endometrial tumors (6, 7), and is by far the most frequent somatic Polϵ mutation in both types of cancer. The same mutation was found in a colon carcinoma cell line HCC2998 (10). A P286H substitution was reported in one additional colon tumor (5), and a germline mutation of the homologous Pro327 in Polδ was present in a patient with a family history of colorectal cancer (8). The repetitive occurrence of changes at this proline was surprising, because studies in model organisms have demonstrated that proofreading can be inactivated by altering a variety of amino acid residues in the exonuclease domain (11–13). To gain insight into the unique properties of the P286R variant, we studied its genome-destabilizing potential in the yeast model system using several mutational assays. The corresponding P301R change in yeast Polϵ conferred an exceptionally strong mutator phenotype greatly exceeding that of any previously characterized Polϵ mutant, including proofreading-deficient mutants. Moreover, in contrast with mutations inactivating the exonuclease of Polϵ, which barely affects the mutation rate in the heterozygous state, heterozygosity for the P301R produced a strong mutator effect comparable with that of MMR deficiency. The results (i) suggest that the extraordinary mutagenic potential of this Polϵ variant results from functional defects distinct from loss of proofreading; (ii) explain earlier observations that loss of heterozygosity for the POLEP286R mutation is not required for tumor development; and (iii) advocate an idea that the frequent occurrence of P286R in sporadic tumors is related to its stronger mutator effect in comparison with other exonuclease domain changes and the resulting greater chance of accumulating cancer-driving mutations.

Snap-frozen colon and rectum adenocarcinoma tissue and matching normal mucosa were obtained from the UNMC Tissue Procurement Shared Resource. The study was approved by the UNMC Institutional Review Board. Exons 7 to 16 of POLE (or all coding exons for the tumor carrying P286R) were amplified by PCR using DNA isolated from the tissue (Supplementary Fig. S1 and Supplementary Table S1) and Sanger sequenced.

Yeast strains used in the study (Supplementary Table S2) are isogenic to E134 (14). Diploids contain the Kluyveromyces lactis LEU2 gene 26 nucleotides downstream of CAN1 in one copy of chromosome V and a deletion of CAN1 in the homologous chromosome, allowing for selection of recessive can1 mutants on medium containing canavanine and lacking leucine. To replace the chromosomal POL2 with the pol2P301R allele, an integrative plasmid YIpDK1 was constructed by cloning a 2.1-kb HpaI–EcoRI POL2 fragment into NruI and EcoRI sites of YIp5 (15). The pol2P301R mutation was created in YIpDK1 by site-directed mutagenesis. The spontaneous mutation rate was measured by fluctuation analysis as described previously (14).

Inspired by the association of Polϵ exonuclease domain mutations with intestinal cancer in mice (2) and humans (5, 8, 9), we began this study by analyzing this region of Polϵ in 52 patients with sporadic colorectal cancer treated at UNMC. The age distribution of the patients and the anatomical site of the tumor are listed in Table 1. Ten of the 52 tumors carried a common variant allele producing an A252V change in Polϵ (nine in heterozygous and one in homozygous state; Table 1). The variant allele frequency did not differ significantly (P = 0.11; χ2) from that reported for the general population (16). One tumor carried an R259H variant, also observed previously in the general population (allele frequency 0.5%–1%). Remarkably, one tumor (T-04-364) had a heterozygous mutation leading to a P286R substitution in the ExoI motif (Table 1; Fig. 1). This change was not present in the publicly available databases of common polymorphisms, but was reported previously as a somatic mutation in 16 other sporadic cancer cases (6, 7, 9) and as a potential cancer-specific variant in a colon carcinoma cell line HCC2998 (10). Other mutations affecting Pro286 in Polϵ or a homologous Pro327 in Polδ were also observed in patients with colorectal cancer (5, 8). The P286R change found in tumor T-04-364 was not present in normal cells of the patient (Fig. 1A), indicating that it was a somatic mutation. The recurrent emergence of Pro286 alterations is in sharp contrast with the apparently random incidence in tumors of other changes in the exonuclease domain of Polϵ, most of which were only observed once (5–7, 9).

Figure 1.

Mutation at codon 286 of POLE in a sporadic colon tumor. A, DNA sequencing analysis of sample T-04-364 and matching normal mucosa. B, conservation of Pro286 in B family DNA polymerases. Hatched boxes on the POLE schematic show conserved exonuclease and polymerase motifs. Hs, human; Sc, S. cerevisiae; Ec, E. coli. Asterisks mark Pro286 and the catalytic carboxylates affected by the pol2-4 mutation.

Figure 1.

Mutation at codon 286 of POLE in a sporadic colon tumor. A, DNA sequencing analysis of sample T-04-364 and matching normal mucosa. B, conservation of Pro286 in B family DNA polymerases. Hatched boxes on the POLE schematic show conserved exonuclease and polymerase motifs. Hs, human; Sc, S. cerevisiae; Ec, E. coli. Asterisks mark Pro286 and the catalytic carboxylates affected by the pol2-4 mutation.

Close modal
Table 1.

Summary of sequence alterations in the POLE gene in colorectal adenocarcinoma samples from 52 patients

Sample IDAge at diagnosisTumor siteNucleotide changeaAmino acid changea
T-07-285 33 Colon   
T-99-082 36 Colon   
T-00-028 42 Colon   
T-01-076 47 Colon   
T-08-756 48 Colon   
T-05-498 53 Colon   
T-04-364b 54 Colon C857C/G P286P/R 
T-00-075 54 Colon C755C/T A252A/V 
T-02-465 56 Colon   
T-01-336 58 Rectum   
T-08-225 60 Colon C755C/T A252A/V 
T-98-285 60 Colon   
T-03-418 61 Colon   
T-00-067 62 Colon C755C/T A252A/V 
T-05-616 63 Colon   
T-99-114 64 Rectum   
T-09-130 65 Colon   
T-01-125 66 Colon C755C/T A252A/V 
T-99-040 66 Rectum   
T-04-019 67 Colon   
T-04-099 67 Colon   
T-09-113 67 Colon C1470C/T synonymous 
T-02-578 67 Colon   
T-03-492 68 Colon   
T-07-332 68 Colon   
T-99-068 69 Colon C755T A252V 
T-02-052 69 Colon   
T-10-274 69 Colon   
T-00-102 69 Colon   
T-09-366 71 Colon   
T-08-387 72 Colon C755C/T A252A/V 
T-08-483 72 Colon   
T-97-183 72 Colon   
T-07-595 72 Colon   
T-97-025 73 Colon   
T-04-683 73 Colon G776G/A R259R/H 
T-08-400 75 Colon   
T-09-675 75 Colon C755C/T A252A/V 
T-09-676 75 Colon C755C/T A252A/V 
T-10-257 75 Colon   
T-05-513 76 Rectum   
T-02-459 77 Colon   
T-01-235 79 Colon   
T-02-499 80 Colon   
T-04-727 82 Colon   
T-02-568 86 Colon C755C/T A252A/V 
T-08-129 86 Colon   
T-08-654 87 Colon   
T-97-010 88 Colon   
T-97-122 88 Colon   
T-08-754 88 Colon   
T-01-149 96 Colon C755C/T A252A/V 
Sample IDAge at diagnosisTumor siteNucleotide changeaAmino acid changea
T-07-285 33 Colon   
T-99-082 36 Colon   
T-00-028 42 Colon   
T-01-076 47 Colon   
T-08-756 48 Colon   
T-05-498 53 Colon   
T-04-364b 54 Colon C857C/G P286P/R 
T-00-075 54 Colon C755C/T A252A/V 
T-02-465 56 Colon   
T-01-336 58 Rectum   
T-08-225 60 Colon C755C/T A252A/V 
T-98-285 60 Colon   
T-03-418 61 Colon   
T-00-067 62 Colon C755C/T A252A/V 
T-05-616 63 Colon   
T-99-114 64 Rectum   
T-09-130 65 Colon   
T-01-125 66 Colon C755C/T A252A/V 
T-99-040 66 Rectum   
T-04-019 67 Colon   
T-04-099 67 Colon   
T-09-113 67 Colon C1470C/T synonymous 
T-02-578 67 Colon   
T-03-492 68 Colon   
T-07-332 68 Colon   
T-99-068 69 Colon C755T A252V 
T-02-052 69 Colon   
T-10-274 69 Colon   
T-00-102 69 Colon   
T-09-366 71 Colon   
T-08-387 72 Colon C755C/T A252A/V 
T-08-483 72 Colon   
T-97-183 72 Colon   
T-07-595 72 Colon   
T-97-025 73 Colon   
T-04-683 73 Colon G776G/A R259R/H 
T-08-400 75 Colon   
T-09-675 75 Colon C755C/T A252A/V 
T-09-676 75 Colon C755C/T A252A/V 
T-10-257 75 Colon   
T-05-513 76 Rectum   
T-02-459 77 Colon   
T-01-235 79 Colon   
T-02-499 80 Colon   
T-04-727 82 Colon   
T-02-568 86 Colon C755C/T A252A/V 
T-08-129 86 Colon   
T-08-654 87 Colon   
T-97-010 88 Colon   
T-97-122 88 Colon   
T-08-754 88 Colon   
T-01-149 96 Colon C755C/T A252A/V 

NOTE: Sample containing the P286R mutation is highlighted in bold.

aNucleotides and amino acids are numbered from the first codon of POLE and the first amino acid, respectively.

bSample T-04-364 contained an additional heterozygous C→T mutation leading to Arg1826Trp substitution, although we could not determine whether it was present in cis or in trans with the codon 286 mutation. Arg1826 is a nonconserved residue away from the known functional domains, so the significance of this change is presently unclear.

Pro286 is located at the end of the ExoI motif several amino acids downstream of the carboxylate residues Asp275 and Glu277 that are essential for catalysis of the exonuclease reaction (Fig. 1B). Mutations affecting these carboxylates abolish the exonuclease activity of Polϵ (15), and so do analogous mutations in other proofreading exonucleases (13). In addition, multiple other amino acid residues in the exonuclease domain, including those in the conserved motifs ExoII and ExoIII, are known to be essential for proofreading (11–13). To understand why the P286R variant is uniquely observed at a high frequency in tumors, we examined its effects on mutagenesis using the yeast S. cerevisiae model system. Haploid and diploid strains with the corresponding pol2P301R allele were constructed, and mutagenesis was measured at three different reporter loci (Table 2). The Canr forward mutation assay scores a variety of base substitutions, frameshifts and larger rearrangements that inactivate the CAN1 gene. The lys2-InsEA14 and his7-2 reporters are frameshift alleles that revert to wild-type predominantly via −1 and +1 events, respectively. In haploid pol2P301R mutants, the rate of Canr, Lys+, and His+ events was increased 150-, 44-, and 370-fold, respectively, in comparison with the wild-type strain. Strikingly, for all three markers, the mutator effect of pol2P301R was approximately 50-fold higher than that of the pol2-4 mutation inactivating Polϵ proofreading due to the replacement of both catalytic carboxylates at the ExoI motif with alanines (15). Thus, despite the location of Pro286/301 in the ExoI motif, its alteration results in fidelity defects substantially more severe than disruption of proofreading. Moreover, the mutator effect of pol2P301R greatly exceeded that of any previously characterized Polϵ mutation, including mutations strongly reducing nucleotide selectivity (Fig. 2).

Figure 2.

Yeast mimic of POLE-P286R is the strongest Polϵ mutator known. A, location of the known mutator mutations on the yeast Pol2 schematic. Hatched boxes, conserved exonuclease and polymerase motifs. B, rate of Canr mutation (relative to wild-type, wt) in haploid pol2 mutants. Data are from Table 2 and Supplementary Table S3.

Figure 2.

Yeast mimic of POLE-P286R is the strongest Polϵ mutator known. A, location of the known mutator mutations on the yeast Pol2 schematic. Hatched boxes, conserved exonuclease and polymerase motifs. B, rate of Canr mutation (relative to wild-type, wt) in haploid pol2 mutants. Data are from Table 2 and Supplementary Table S3.

Close modal
Table 2.

Rate of spontaneous mutation in haploid and diploid pol2P301R mutants

Canr mutationLys+ reversionHis+ reversion
PloidyGenotypeMutation rate (×10−8)aFold increaseMutation rate (×10−8)aFold increaseMutation rate (×10−8)aFold increase
n 15 (10–20) 14 (11–16) 0.57 (0.42–0.75) 
 pol2P301R 2,300 (1,800–2,800) 150 610 (450–710) 44 210 (160–250) 370 
 pol2-4 44 (39–51) 2.9 14 (11–15) 3.3 (2.6–4.2) 5.8 
2n +/+ 19 (16–24) 16 (14–18) 0.87 (0.62–0.92) 
 pol2P301R/+ 510 (430–680) 27 29 (23–53) 1.8 25 (20–31) 29 
 pol2-4/+ 41 (37–46) 2.2 24b (19–29) 0.8b 3.3 (2.5–4.1) 3.8 
 pol2P301R/pol2P301R 14,000 (6,900–27,000) 740 4,100 (2,000–7,500) 280 1,400 (590–3,100) 1,600 
 pol2-4/pol2-4 97 (73–110) 5.1 22 (18–27) 1.4 7.6 (5.7–8.6) 8.7 
Canr mutationLys+ reversionHis+ reversion
PloidyGenotypeMutation rate (×10−8)aFold increaseMutation rate (×10−8)aFold increaseMutation rate (×10−8)aFold increase
n 15 (10–20) 14 (11–16) 0.57 (0.42–0.75) 
 pol2P301R 2,300 (1,800–2,800) 150 610 (450–710) 44 210 (160–250) 370 
 pol2-4 44 (39–51) 2.9 14 (11–15) 3.3 (2.6–4.2) 5.8 
2n +/+ 19 (16–24) 16 (14–18) 0.87 (0.62–0.92) 
 pol2P301R/+ 510 (430–680) 27 29 (23–53) 1.8 25 (20–31) 29 
 pol2-4/+ 41 (37–46) 2.2 24b (19–29) 0.8b 3.3 (2.5–4.1) 3.8 
 pol2P301R/pol2P301R 14,000 (6,900–27,000) 740 4,100 (2,000–7,500) 280 1,400 (590–3,100) 1,600 
 pol2-4/pol2-4 97 (73–110) 5.1 22 (18–27) 1.4 7.6 (5.7–8.6) 8.7 

aMutation rates are given as medians for at least nine cultures, with 95% CIs in parentheses.

bThe pol2-4/POL2 heterozygote contains two copies of the lys2-InsEA14 allele. All other diploids have one copy of lys2-InsEA14 and a nonrevertible mutation at the same position of the LYS2 gene in the homologous chromosome (Supplementary Table S2). The fold increase shown is the fold difference in the Lys+ reversion rate between the wild-type and pol2-4/POL2 diploids divided by two.

The mutant POLE alleles were always found in the heterozygous state in sporadic human tumors (Fig. 1A; refs. 5, 6, 9). Diploid yeast strains heterozygous for the pol2P301R showed a remarkable increase in the rate of Canr mutation and his7-2 reversion (27- and 29-fold, respectively; Table 2). This is drastically different from strains heterozygous for the pol2-4 mutation inactivating proofreading that showed only a 2.2-fold elevated Canr mutation and 3.8-fold elevated his7-2 reversion. The Canr mutation rate in the POL2/pol2P301R heterozygotes was comparable with that in strains with completely inactivated MMR, which is a well-known cancer-driving replication fidelity defect. For example, the haploid mlh1 strain is a 26-fold mutator (14). This suggests that heterozygosity for the POLEP286R seen in tumors can have a potent genome-destabilizing effect and explain the hypermutated phenotype of these tumors. Diploid yeast homozygous for the pol2P301R allele showed extreme hypermutability (a 740-, 280-, and 1,600-fold increase in the rate of Canr, Lys+, and His+ events; Table 2). The dramatic difference in the apparent mutation rate between diploids and haploids is commonly observed when the mutation load approaches a lethal threshold. It is explained by death in haploids of a fraction of cells with the highest level of mutagenesis that survives and produces the majority of mutants in diploids (17–19, and references therein). In contrast, the pol2-4/pol2-4 diploids are only slightly more mutable than pol2-4 haploids (Table 2; ref. 19), consistent with a modest level of replication errors and fewer cells accumulating lethal mutations. This further emphasizes the differences between the mutator effects of pol2P301R and the proofreading deficiency.

Because of the location of cancer-associated Polϵ changes at conserved residues in the exonuclease domain, a concept has formed in the past year that loss of Polϵ proofreading is responsible for the development of hypermutated human tumors. The functional consequences of the Polϵ variants found in cancers, however, have not been tested previously. The studies described here indicate that the P286R variant confers fidelity defects well beyond those expected from proofreading deficiency. Pro286 is conserved in a B-family Polζ lacking proofreading activity (Fig. 1B), consistent with its possible other functions. Thus, the higher frequency of P286R in cancers in comparison with other exonuclease domain changes could be explained by its exceptionally strong mutator effect and the resulting increased chances of accumulating cancer-driving mutations. Interestingly, we found the P286R variant in a tumor from a 54-year-old individual, indicating a relatively early cancer onset (the median age at diagnosis in our randomly selected group of patients was 69 with 95% confidence interval (CI) of 66–71; Table 1). Further studies could determine whether this mutation is more frequent in younger patients.

The uniquely strong mutator effect of the yeast P286R analog resembles our earlier observations with an R696W variant of Polδ, the only extensively characterized colorectal cancer-associated Polδ variant (17). Yeast strains expressing Polδ–R696W had a mutation rate at least two orders of magnitude higher than any known Polδ mutator and were generally the strongest eukaryotic mutators ever seen. Together with the present studies of Polϵ–P286R, this draws an interesting and unexpected picture wherein the cancer endpoint acts as an efficient selective system for the strongest mutators. Characterization of other replicative polymerase variants present in tumors in future studies will help expand and circumstantiate this view. Long-term expression of a strong mutator phenotype is believed to be disadvantageous, because it leads to an accumulation of deleterious mutations that could impede tumor growth (20). The mutagenic potential of the mutator polymerase variants, however, is buffered by the presence of the wild-type polymerases in the heterozygous tumors. Importantly, heterozygosity for the mutator alleles is still quite mutagenic (Table 2), because both wild-type and mutator enzymes contribute to replication. Thus, DNA polymerase genes constitute an uncommon class of tumor suppressors, for which loss of heterozygosity is not required to promote tumorigenesis. Although limiting the genome-damaging activity of supermutator enzymes, the heterozygosity also creates opportunities for driving the genome instability by transient spikes of hypermutation due to imbalanced expression of the wild-type and mutant alleles as we discussed previously (17).

No potential conflicts of interest were disclosed.

Conception and design: D.P. Kane, P.V. Shcherbakova

Development of methodology: D.P. Kane

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): D.P. Kane, P.V. Shcherbakova

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): D.P. Kane, P.V. Shcherbakova

Writing, review, and/or revision of the manuscript: D.P. Kane, P.V. Shcherbakova

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): D.P. Kane

Study supervision: P.V. Shcherbakova

The authors thank Brad Preston for discussions that led to the development of this project, the UNMC Tissue Bank personnel for help in identification of the human tissue samples appropriate for the study, Tony Mertz for yeast strains, and Elizabeth Moore and Krista Brown for technical assistance.

This research was supported by NIH ES015869, CA127297, and Nebraska DHHS LB506.

1.
Salk
JJ
,
Fox
EJ
,
Loeb
LA
. 
Mutational heterogeneity in human cancers: origin and consequences
.
Annu Rev Pathol
2010
;
5
:
51
75
.
2.
Albertson
TM
,
Ogawa
M
,
Bugni
JM
,
Hays
LE
,
Chen
Y
,
Wang
Y
, et al
DNA polymerase ϵ and δ proofreading suppress discrete mutator and cancer phenotypes in mice
.
Proc Natl Acad Sci U S A
2009
;
106
:
17101
4
.
3.
Wei
K
,
Kucherlapati
R
,
Edelmann
W
. 
Mouse models for human DNA mismatch-repair gene defects
.
Trends Mol Med
2002
;
8
:
346
53
.
4.
de la Chapelle
A
. 
Genetic predisposition to colorectal cancer
.
Nat Rev Cancer
2004
;
4
:
769
80
.
5.
Cancer Genome Atlas Network
. 
Comprehensive molecular characterization of human colon and rectal cancer
.
Nature
2012
;
487
:
330
7
.
6.
Church
DN
,
Briggs
SE
,
Palles
C
,
Domingo
E
,
Kearsey
SJ
,
Grimes
JM
, et al
DNA polymerase ϵ and δ exonuclease domain mutations in endometrial cancer
.
Hum Mol Genet
2013
;
22
:
2820
8
.
7.
Cancer Genome Atlas Network
. 
Integrated genomic characterization of endometrial carcinoma
.
Nature
2013
;
497
:
67
73
.
8.
Palles
C
,
Cazier
JB
,
Howarth
KM
,
Domingo
E
,
Jones
AM
,
Broderick
P
, et al
Germline mutations affecting the proofreading domains of POLE and POLD1 predispose to colorectal adenomas and carcinomas
.
Nat Genet
2012
;
45
:
136
44
.
9.
Seshagiri
S
,
Stawiski
EW
,
Durinck
S
,
Modrusan
Z
,
Storm
EE
,
Conboy
CB
, et al
Recurrent R-spondin fusions in colon cancer
.
Nature
2012
;
488
:
660
4
.
10.
Abaan
OD
,
Polley
EC
,
Davis
SR
,
Zhu
YJ
,
Bilke
S
,
Walker
RL
, et al
The exomes of the NCI-60 panel: a genomic resource for cancer biology and systems pharmacology
.
Cancer Res
2013
;
73
:
4372
82
.
11.
Herr
AJ
,
Ogawa
M
,
Lawrence
NA
,
Williams
LN
,
Eggington
JM
,
Singh
M
, et al
Mutator suppression and escape from replication error-induced extinction in yeast
.
PLoS Genet
2011
;
7
:
e1002282
.
12.
Jin
YH
,
Obert
R
,
Burgers
PM
,
Kunkel
TA
,
Resnick
MA
,
Gordenin
DA
. 
The 3′→5′ exonuclease of DNA polymerase δ can substitute for the 5′ flap endonuclease Rad27/Fen1 in processing Okazaki fragments and preventing genome instability
.
Proc Natl Acad Sci U S A
2001
;
98
:
5122
7
.
13.
Reha-Krantz
LJ
. 
DNA polymerase proofreading: multiple roles maintain genome stability
.
Biochim Biophys Acta
2010
;
1804
:
1049
63
.
14.
Shcherbakova
PV
,
Kunkel
TA
. 
Mutator phenotypes conferred by MLH1 overexpression and by heterozygosity for mlh1 mutations
.
Mol Cell Biol
1999
;
19
:
3177
83
.
15.
Morrison
A
,
Bell
JB
,
Kunkel
TA
,
Sugino
A
. 
Eukaryotic DNA polymerase amino acid sequence required for 3′→5′ exonuclease activity
.
Proc Natl Acad Sci U S A
1991
;
88
:
9473
7
.
16.
NCBI SNP database links
.
[Last accessed 10/5/13]. Available from
: http://www.ncbi.nlm.nih.gov/projects/SNP/snp_ref.cgi?geneId=5426.
17.
Daee
DL
,
Mertz
TM
,
Shcherbakova
PV
. 
A cancer-associated DNA polymerase δ variant modeled in yeast causes a catastrophic increase in genomic instability
.
Proc Natl Acad Sci U S A
2010
;
107
:
157
62
.
18.
Lada
AG
,
Stepchenkova
EI
,
Waisertreiger
IS
,
Noskov
VN
,
Dhar
A
,
Eudy
JD
, et al
Genome-wide mutation avalanches induced in diploid yeast cells by a base analog or an APOBEC deaminase
.
PLoS Genet
2013
;
9
:
e1003736
.
19.
Tran
HT
,
Degtyareva
NP
,
Gordenin
DA
,
Resnick
MA
. 
Genetic factors affecting the impact of DNA polymerase δ proofreading activity on mutation avoidance in yeast
.
Genetics
1999
;
152
:
47
59
.
20.
McFarland
CD
,
Korolev
KS
,
Kryukov
GV
,
Sunyaev
SR
,
Mirny
LA
. 
Impact of deleterious passenger mutations on cancer progression
.
Proc Natl Acad Sci U S A
2013
;
110
:
2910
5
.