Cis lethal genetic interactions attenuate and alter ... - Semantic Scholar

1 downloads 0 Views 815KB Size Report
Mar 23, 2010 - aDepartment of Cell Biology, Albert Einstein College of Medicine, Bronx, NY 10461; bDepartment of Medicine and ..... Sézary syndrome.
Cis lethal genetic interactions attenuate and alter p53 tumorigenesis Yuxun Wanga,1,2, Weijia Zhangb, Lisa Edelmannc, Richard D. Kolodnerd,2, Raju Kucherlapatie, and Winfried Edelmanna,2 a Department of Cell Biology, Albert Einstein College of Medicine, Bronx, NY 10461; bDepartment of Medicine and cDepartment of Genetics and Genomic Sciences, Mount Sinai School of Medicine, New York, NY 10029; dLudwig Institute for Cancer Research, Departments of Medicine and Cellular and Molecular Medicine, and Cancer Center, University of California San Diego, La Jolla, CA 92093; and eHarvard Medical School–Partners Healthcare Center for Genetics and Genomics, Harvard Medical School, Boston, MA 02115

Contributed by Richard D. Kolodner, February 2, 2010 (sent for review October 22, 2009)

|

DNA repair genome instability tumor suppressor gene

| loss of heterozygosity | murine model |

R

PA is a heterotrimeric single-stranded DNA binding protein complex consisting of three subunits, RPA1, RPA2, and RPA3 (1, 2). We have previously shown that a homozygous missense mutation in Rpa1, L230P resulted in cell lethality, whereas the heterozygous Rpa1L230P cells displayed defective DNA repair and genetic instability because Rpa1L230P is a semidominant mutation (3). When these genetic instabilities included copy number aberrations at genomic regions containing critical cancerrelated genes, such as increased oncogene copy number, as demonstrated by array comparative genomic hybridization (aCGH) analysis of cancer cell genomes in the Rpa1L230P/+ mutant mice, they resulted in tumor development in vivo. We further showed that the penetrance of the tumor phenotype was incomplete, suggesting that additional genetic events might be required to drive tumorigenesis with complete penetrance. Rpa1 and Trp53 are syntenic and map in close proximity to each other in both mice (chromosome 11) and humans (chromosome 17). In humans, RPA1 is located on chromosome 17 at 17p13.3 and linked to TP53, which is located at 17p13.1. Loss of 17p encompassing both of these loci has been shown to be one of the most frequent genetic defects found in human cancers including colorectal cancer, breast cancer, lymphoma, and leukemia (4–8). However, to what extent the loss of each of these two genes contributes in this context to the tumorigenic process is unknown, nor is it clear whether TP53 and RPA1 play cooperative, synergistic, additive, or antagonistic roles in tumorigenesis. Loss of p53 function is one of the key causal events in tumorigenesis, and approximately half of all human cancers harbor mutations in TP53. Consistent with its essential role in tumor suppression in humans, murine models carrying loss-of-function mutations in Trp53 develop tumors in multiple tissues (9, 10). Although pathogenic mutations in RPA1 have not been reported in humans, genetic polymorphisms in RPA1 were shown to be present in some human colon cancer tissues and cell lines (11) and www.pnas.org/cgi/doi/10.1073/pnas.1001223107

are also associated with head and neck cancers (12). We have shown that a murine model expressing the mutant Rpa1L230P allele was susceptible to tumorigenesis in lymphoid tissues (3) because of defects in multiple DNA repair pathways that caused increased genetic instability. These studies demonstrated that Rpa1 is essential not only for cell survival but also for the prevention of cancer. In this context it is important to note that although p53 function is essential for tumor suppression, it is not required for cell survival. In addition to their genetic linkage in both mice and humans, RPA1 protein also physically interacts with p53 protein (13, 14), suggesting a functional link between these two genes. In this report, we describe mice carrying mutations in both Rpa1 and Trp53 genes in either trans or cis configurations. The analysis of tumorigenesis in these mouse lines provides compelling evidence that mutations or polymorphisms in cell essential genes can modify the genetic events underlying tumorigenesis at linked diseaseassociated loci via allelic phasing. Results Generation of Cis and Trans Rpa1;Trp53 Compound Mutant Mice.

To study the genetic interaction between Rpa1 and Trp53, we intercrossed Rpa1L230P/+ and Trp53+/− knockout mice and generated double mutant mice carrying heterozygous mutations in both Rpa1 and Trp53 (Fig. 1). Because the Rpa1 and Trp53 genes are linked on mouse chromosome 11, we created Rpa1L230P/+; Trp53+/− double heterozygous mice carrying both mutant alleles on the same chromosome (cis configuration, termed cis mice) or on opposite chromosomes (trans configuration, termed trans mice) (Fig. 1). Cis mice were generated via meiotic recombination by crossing trans mice with Rpa1+/+; Trp53+/+ wild-type mice followed by genetic mapping of the resulting offspring (Fig. 1). We also generated Rpa1L230P/+; Trp53−/− mice by breeding Rpa1L230P/+; Trp53+/− cis mice with Trp53+/− mice. Cis and Trans Configurations Alter Survival and Tumorigenesis in Rpa1;Trp53 Mutant Mice. The survival of cohorts of mice with

each of the different Rpa1 and Trp53 genotypes was analyzed. As expected, Trp53+/− mice had a reduced life span compared with wild-type littermate control mice. One half of the Trp53+/− mice (19 of 38) died by 16 months of age, whereas all of the wild-type mice (n = 27) were still alive. Interestingly, trans mice exhibited a further reduced survival compared with Trp53+/− mice, with a 50% (13 of 26) survival of 14 months (P = 0.0098). Surprisingly, cis mice

Author contributions: Y.W., R.D.K., and W.E. designed research; Y.W., W.Z., and L.E. performed research; Y.W., W.Z., L.E., R.D.K., R.K., and W.E. analyzed data; and Y.W., R.D.K., and W.E. wrote the paper. The authors declare no conflict of interest. 1

Present address: Department of Pharmaceutical and Biomedical Sciences, University of South Carolina, Columbia, SC 29208.

2

To whom correspondence may be addressed. E-mail: [email protected], edelmann@ aecom.yu.edu, or [email protected].

This article contains supporting information online at www.pnas.org/cgi/content/full/ 1001223107/DCSupplemental.

PNAS | March 23, 2010 | vol. 107 | no. 12 | 5511–5515

GENETICS

Rpa1, an essential gene involved in DNA replication and genome maintenance, is syntenic and linked to Trp53 in mice and humans. To study the genetic interaction between Rpa1 and Trp53 in tumorigenesis, we generated compound Rpa1L230P/+; Trp53+/− mutant mice with the mutant alleles in either trans or cis configuration. We demonstrate that the Rpa1L230P missense mutation significantly alters the tumor phenotype and spectrum of Trp53 mutant mice by modifying the genetic mechanisms underlying tumorigenesis. Importantly, when the Rpa1L230P and Trp53 mutant alleles are in cis, the tumor phenotype is attenuated and altered and loss of heterozygosity (LOH) at the Trp53 wild-type locus is selected against, whereas in the trans configuration, Rpa1L230P enhances the Trp53+/− tumor phenotype even though Rpa1L230P is ultimately lost by LOH. These studies indicate that polymorphic genetic variants in cell essential genes can genetically affect closely linked tumor suppressor loci via allelic phasing, which can result in profound phenotypic variations in tumorigenesis.

Rpa1 Het + +

trans mutant mice aged up to 18 months uncovered a spectrum of tumors that included lymphomas, soft tissue sarcomas, and osteosarcomas (Fig. 2 B–E). However, the distribution of the various tumor types found in each of the mutant mouse lines varied (Fig. 2E). Although approximately one fourth of the Trp53+/− mice (7 of 30) were tumor-free at the time of analysis, the remainder of Trp53+/− mice developed mainly lymphomas, and only a small number of the mice developed sarcomas. All trans mice (n = 18) had developed tumors (P = 0.0356 trans vs. Trp53+/−), and both lymphomas and sarcomas were found at equal frequencies. In contrast, there was a significant reduction in tumor load in cis mice. Only half of the cis mice (18 of 36) carried tumors (P = 0.0001 cis vs. trans; P = 0.041 cis vs. Trp53+/−), and sarcomas were the most common tumors in cis mice. All of the Rpa1L230/+; Trp53−/− mice (n = 20) developed lymphomas with very short latency (Fig. 2E).

p53 Het p53 (-)

+ +

+

Rpa1L230P

+

F1

Trans Rpa1L230P +

p53 (-)

+ +

+

+ +

Meiotic recombination

p53 Het + +

p53 (-)

Rpa1 Het/p53 Hom

p53 (-) Rpa1L230P

+ +

+

p53 (-)

+

Cis

p53 (-) Rpa1L230P

Fig. 1. Generation of compound Rpa1; Trp53 mice. The breeding scheme shown was used to produce trans, cis ,and Rpa1L230P/+; Trp53−/− compound mutant mice.

displayed a prolonged survival compared with Trp53+/− mice, with 70% of cis mice (42 of 60) surviving beyond 16 months (P < 0.0001). To determine the cause of the different survival rates, we generated new cohorts of mice for each genotype and analyzed the tumor phenotypes in moribund mice. As shown in Fig. 2A, by 16 months, all (n = 42) trans mice had developed tumors, whereas by 17 months, 75% of Trp53+/− mice (n = 30) had developed tumors (P = 0.0042). In contrast, only half (20 of 40) of the cis mice developed tumors at the somewhat older age of 18 months (P = 0.0046, cis vs. Trp53+/−; P = 0.0027, cis vs. trans). Only 1 of 20 wildtype mice (5%) had developed a tumor by 17 months. The difference in tumor prevalence was consistent with the differential rates of survival associated with each of the mouse lines. Systemic histopathological analysis of a subset of Trp53+/−, cis and

A Mice with tumors (%)

E 100

WT +/p53 Trans Cis

80 60

Cis or Trans Configuration Prevents or Augments Loss of Heterozygosity at the Trp53 Locus During Tumorigenesis. To understand the genetic

mechanism generating the distinct tumor phenotypes associated with each of the different Rpa1 and Trp53 genotypes, we carried out a series of studies in the compound mutant mice. First, we tested whether tumorigenesis in Trp53+/−, cis and trans mice was due to inactivating mutations in the remaining wild-type Trp53 allele. The mutation hot spot region (15) of the Trp53 gene encompassing exons 5, 6, 7, and 8 was sequenced in genomic DNA isolated from tumors that developed in these mice. No mutations were found in any of the tumors analyzed (none of six Trp53+/−, none of six cis, and none of six trans; Fig. S1). An alternative genetic mechanism underlying tumorigenesis in the mice could be loss of heterozygosity (LOH) at the wild-type Trp53 or Rpa1 locus resulting in complete loss of p53 or Rpa1 function. To test this possibility, we developed PCR-based assays to detect the presence or absence of the remaining wild-type Rpa1 or Trp53 allele in both normal and tumor tissues of the same mice. We found no LOH at the wild-type Rpa1 locus in any of the cis and trans tumors analyzed (Fig. S2A).

Soft tissue sarcoma 10%

Tumor-free 0%

Osteosarcoma 0% Soft tissue sarcoma 0%

Tumor-free 23%

40 20

Lymphoma 54%

0 0

5

10

15

20

Age (months)

B

Osteosarcoma 13%

C

Rpa1

p53+/-

D

Soft tissue sarcoma 11%

Osteosarcoma 19%

Lymphoma 100% L230P/+

; p53-/-

Soft tissue sarcoma 11% Tumor-free 0% Osteosarcoma 39%

Lymphoma 40X Lymphoma 19%

Osteosarcoma 40X

Tumor-free 51%

Lymphoma 50%

Cis

Trans

Fig. 2. Mutation in Rpa1 modifies tumor phenotype of Trp53 mutant mice. (A) Tumor incidence and age in mice with indicated Rpa1 and Trp53 genotypes. (B) Gross morphology showing a lymphoma in the mesenteric lymph node of a trans mouse (Upper Left) and close-up image of the tumor mass in the dissected gastrointestinal tract (Upper Right). A lymphoma in the spleen of another trans mouse (Lower Left), with the spleen being grossly enlarged compared with the spleen in the wild-type mouse (Lower Right). (C) X-ray imaging of an osteosarcoma on the skull of a cis mouse (Upper) and close-up x-ray image of the bone mass (Lower). (D) Representative histopathological images of lymphoma (Upper) and osteosarcoma (Lower) in cis mice. (E) Rpa1L230P modifies tumor spectrum in Trp53+/− mice. A comparison of tumor incidence and type in mice with the different Rpa1 and Trp53 genotypes is shown.

5512 | www.pnas.org/cgi/doi/10.1073/pnas.1001223107

Wang et al.

B

A

p53+/-

Trans

Cis

1 2 3 4 5 6

1 2 3 4 5 6

1 2 3 4 5 6

D11Mit226 (1.5 cM) D11Mit78 (2.0 cM) D11Mit20 (20 cM)

LOH in Rpa1L230P/+;p53+/-

p53+/-

8/10

Trans

8/8

Cis

0/10

tumors

p53 (39 cM) Rpa1 (44 cM)

)

3 Tumor

Tumor

Tail

Tumor

Tail

it2 24

Trans 2

Tail

(6 6

cM

)

) cM

cM

(2 0

it2 0

11 M

129 D

1

B6

6

1

Retainment Loss Loss of mutant allele

129 D 11 M B6 it2 129 D 2

C

Chr. 11

(1 1M B6 .5 it7 cM 129 D 8 ) 11 (2 .0 B6 M

D11Mit224 (66 cM)

D11Mit78

Fig. 3. Mutation in Rpa1 prevents LOH of wild-type Trp53 in cis mice. (A) Frequency of LOH in tumors isolated from the mice with indicated genotypes. (B) Chromosomal mapping indicates that LOH involves whole chromosome 11. Results from six representative tumors are shown for each genotype. (C) Verification of ability of selected polymorphic markers to differentiate polymorphisms derived from either C57Bl6 or 129 background, and representative results showing LOH at the D11Mit78 locus in tumors from trans mice.

essential role of Rpa1 in maintaining chromosomal stability. We also performed a CGH on lymphomas that developed in Trp53−/−; Rpa1L230P/+ mice and found that these tumors contained severe chromosomal aberrations including aneuploidy of chromosomes 4, 9, 11, 14, and 15. Compared with the lymphomas of Trp53−/− mice in a previously published aCGH dataset (19) (Fig. S6), the abnormalities in Trp53−/−; Rpa1L230P/+ mice were more extensive, although this difference in genomic instability could also be due to differences in the genetic backgrounds of the mice. Discussion Mouse chromosome 11 is gene-rich and, in addition to Trp53, other tumor suppressor genes map to this chromosome (Fig. S7). Allelic phasing mouse models have been generated previously in which each mouse carried mutations in Trp53 and either a second tumor

Increased Genomic Instability Modifies Tumor Spectrum in Cis and Trans Rpa1;Trp53 Mice. To further dissect the molecular mecha-

nism underlying tumorigenesis in each of the murine models, we analyzed genome-wide genomic instability in the tumors by aCGH. Chromosomal instability appeared to be a rare event in tumors of Trp53+/− mice (Fig. 4). Only one of six Trp53+/− tumors displayed large-scale chromosome aberrations involving whole chromosomes or large chromosomal regions (Fig. S3). Compared with Trp53+/− tumors, both trans and cis tumors displayed increased numbers of chromosomal aberrations (P = 0.03, trans or cis vs. Trp53+/−) (Fig. 4), indicating that the Rpa1L230P mutation enhanced genetic instability in Trp53+/− tumors. We also compared the genetic instability in lymphomas and sarcomas that developed in cis and trans mice. As indicated above, trans and cis mutant mice developed both lymphomas and sarcomas; however, the percentage of sarcomas was higher than in Trp53+/− mice (P = 0.002 and P < 0.0005, respectively). Strikingly, sarcomas contained many more chromosomal aberrations than lymphomas in both cis (Fig. S4) and trans mice (Fig. S5). These data provide further strong evidence for the Wang et al.

Fig. 4. Mutation in Rpa1 enhances genetic instability in tumors of Trp53+/− mice. aCGH analysis revealed higher levels of chromosomal abnormalities in trans and cis compared with Trp53+/− tumors.

PNAS | March 23, 2010 | vol. 107 | no. 12 | 5513

GENETICS

However, 80% of the Trp53+/− tumors and all of the trans tumors displayed LOH at the wild-type Trp53 locus (Fig. S2 A, B, and D). Consistent with previous reports, the frequency of LOH at the Trp53 locus in Trp53+/− mice was similar in lymphomas and sarcomas (16). In contrast, none of the cis tumors displayed LOH at the wild-type Trp53 allele (Fig. 3A and Fig. S2 A and C). To further investigate the genetic events underlying the loss of the wild-type Trp53 locus in tumors of trans and Trp53+/− mice, we mapped the chromosome regions involved in the LOH by PCRbased assays analyzing the Trp53 and Rpa1 loci and several polymorphic markers along chromosome 11 (Fig. 3 B and C). Consistent with our previous results, most of the Trp53+/− tumors lost the wild-type Trp53 allele, whereas only a small percentage of the tumors retained the wild-type allele (Fig. 3B). For those tumors exhibiting LOH, all of the polymorphic markers on chromosome 11 showed LOH indicating loss of the whole chromosome. Similar to Trp53+/− tumors, in all of the trans tumors tested, the LOH event involved the entire chromosome 11. The observation of whole chromosome involvement in LOH is not unprecedented. Several recent reports have shown that in contrast to localized deletion/amplification events in radiation-induced Trp53+/− tumors, LOH in spontaneous Trp53+/− tumors involved whole chromosomes (17, 18). We also found loss of the mutant Rpa1 allele in all of the trans tumors, indicating that the mutant Rpa1 allele was lost concomitant with the loss of the wild-type Trp53 allele. In contrast to the Trp53+/− and trans tumors, all of the cis tumors retained their wild-type chromosome 11 (Fig. 3B).

suppressor gene (Nf1 or Hic1) or a deletion of a chromosomal segment on chromosome 11 (20–25). Compared with the Nf1 and Hic1 genes, the Rpa1 gene is more closely linked to the Trp53 gene with a genetic distance of 5 cM between the two genes (Fig. S7). Nf1 and Hic1 are located 7 and 9 cM distal to Trp53, respectively. In addition, in humans, RPA1 also maps in close proximity to TP53 on the short arm of chromosome 17 at 17p13.3 and 17p13.1, respectively (Fig. S7), suggesting that the genetic linkage between RPA1 and TP53 is evolutionarily conserved. In contrast, in humans, the NF1 gene maps further away from the TP53 gene and is located on the long arm of chromosome 17 at 17q11.2, whereas HIC1 is located on the short arm at 17p13.3 (Fig. S7). Interestingly, allelic phasing in the compound mutant Nf1; Trp53 or Hic1; Trp53 mouse models appears to affect the p53-driven tumorigenesis differently compared with the Rpa1; Trp53 cis and trans mouse models. Although tumorigenesis in both cis and trans Nf1; Trp53 and Hic1; Trp53 mice was accelerated, the cis and trans configuration of Rpa1 clearly affected p53-driven tumorigenesis in opposing ways. We previously reported that the Rpa1L230P mutation caused cell lethality in homozygous mutant cells (3), providing an explanation for the differences in LOH in Trp53+/−, trans and cis tumors. If LOH had occurred at the wild-type Trp53 locus during the development of cis tumors, this loss would likely also have resulted in homozygosity at the closely linked mutant Rpa1L230P locus, which would in turn have caused cell lethality in the tumor cells. Therefore, LOH at the wild-type Trp53 locus is prevented in cis mice because of the close linkage of Trp53 to the wild-type Rpa1 allele on the same chromosome. Interestingly, the presence of the mutant Rpa1L230P allele in trans mice appears to precipitate the loss of the wild-type Trp53 allele as well as its own loss, most likely because a heterozygous Rpa1L230P allele results in increased

genome instability. A schematic description of LOH events at the Rpa1 and Trp53 loci during tumorigenesis is given in Fig. 5. Our studies delineate a genetic mechanism in tumorigenesis in which a mutation in a cell essential gene can profoundly modify LOH events at closely linked tumor suppressor genes and significantly impact tumorigenesis. Furthermore, we postulate that this genetic mechanism of allelic phasing may also exist in the context of other closely linked genetic loci involving cell essential genes in both murine and human genomes. Given the density of cell essential genes within the genomes of all organisms, our results suggest that this intrinsic mechanism may also play a crucial role in modifying the cancer phenotypes and possibly other disease phenotypes. Although there is currently very limited information about cell essential genes in the human genome, we were able to identify a significant number of tumor suppressor genes, or genes thought to be important for tumorigenesis, that are linked to the human orthologs of yeast genes that have been previously identified as cell essential in yeast knockout studies (Saccharomyces Genome Project, http://www.yeastgenome.org). Because these linked tumor suppressor/essential gene pairs occur within 6 Mb of each other, similar to the distance between RPA1 and TP53, it is possible that they may also be important determinants in tumorigenesis in humans (Table S1). As more detailed information about the human genome becomes available, including the locations of regions of copy number variation (CNV) and the results of whole genome associations studies in cancer, the discovery of genetic loci that modify the activity of linked disease loci will be facilitated. For example, copy number losses that include cell essential genes that are linked to tumor suppressor loci may have a significant impact on cancer susceptibility phenotypes of individuals who carry them. Notably, a recent report of a CNV in healthy male individuals from northern France identified a CNV on 17p13.3 that included dele-

L230P

Rpa1L230P

Rpa1 L230P

Rpa1

Rpa1

LOH: 0/12

+

A

L230P

Rpa1

p53+/-

+

L230P

Rpa1

+

+

p53 -

p53 -

+

p53 -

p53 +

p53 +

+

p53 +

p53 -

LOH: 8/10

B

Trans

p53 + N

C

LOH: 8/8 ot s e

lec t

ed

L230P

Rpa1

p53 -

p53 L230P Rpa1

p53 -

+ +

L230P

Cis

D

p53 -

L230P

Rpa1

L230P

Rpa1

LOH: 0/10

+ + L230P

Rpa1

Fig. 5. Diagram depicting the LOH events in tumors carrying the different Rpa1 and Trp53 alleles. (A) LOH at the wild-type Rpa1 locus is prevented during tumorigenesis in Rpa1L230P/+ mice due to cell lethality caused by the L230P mutation. (B) In Trp53+/− mice, the wild-type Trp53 allele undergoes frequent LOH during tumorigenesis. (C) In trans mice, LOH at the wild-type Trp53 and mutant Rpa1L230P gene loci occurs in all tumors and is likely due to the increased genetic instability caused by the Rpa1L230P mutation, which also selects against the retention of both alleles and a strong selection pressure for tumor cells to Trp53 homozygosity. (D) In cis mice, the cell lethal phenotype caused by homozygosity for the Rpa1L230P mutation prevents LOH at the wild-type Rpa1 allele and the closely linked wild-type Trp53 allele. As a consequence, tumor cells retain the mutant Rpa1L230P allele and the Trp53 wild-type allele and tumorigenesis is altered.

5514 | www.pnas.org/cgi/doi/10.1073/pnas.1001223107

Wang et al.

Rpa1L230P/+ and Trp53+/− mice. Cis compound Rpa1; Trp53 mice were generated by crossing trans mutant mice with wild-type mice, and the offspring were genotyped for the presence of both the Rpa1L230P and the Trp53 knockout allele indicating a meiotic recombination event between these loci. The cis mice were also bred to Trp53+/− mice to produce Rpa1L230P/+; Trp53−/− offspring. The mice used in this study were on a mixed C57BL/6 and 129/SvJae genetic background. All animal experiments in this study were conducted with the Institutional Animal Care and Use Committee approval. Histopathology and Imaging Analysis of Tumors. The mice were killed when they became moribund and were subjected to systemic histopathological analysis. The tissues were processed following standard procedures, and tumor and normal tissue paraffin sections were examined by staining with H&E. For tumor imaging, mice were anesthetized with ketamine and xylazine and analyzed for the presence of increased bone mass using a cabinet x-ray system. PCR and LOH Analysis. Genomic DNA was isolated using the Qiagen DNAeasy kit and used in PCR reactions. For LOH mapping analyses, polymorphic dinucleotide markers from Jackson Laboratories and previously published markers were used (22). Mutational Analysis of Trp53. The mutation hot spot regions in Trp53 were PCR amplified using primers flanking each of exons 5, 6, 7, and 8. The DNA fragments were sequenced. Genomic DNA Isolation and aCGH Analysis. The aCGH procedures are described in the SI Text and performed as previously reported (3). Statistical analysis was performed using Student’s t test.

Animals. The Rpa1L230P mice were previously generated by us (3), and the Trp53+/− knockout mice were purchased from the Jackson Laboratories. Trans compound Rpa1; Trp53 mutant mice were generated by intercrossing

ACKNOWLEDGMENTS. We thank Drs. Steven Brunnert and Robert Russell for assistance with the histopathological analysis and x-ray imaging of the mutant mice. We also thank Bo Jin for excellent technical assistance and mouse maintenance. This work was supported by National Institutes of Health Grants CA76329 and CA93484 (to W.E.), CA84301 (to W.E. and R.K.), GM26017 (to R.D.K.), and Center Grant CA13330 (to the Albert Einstein College of Medicine).

1. Wold MS (1997) Replication protein A: A heterotrimeric, single-stranded DNA-binding protein required for eukaryotic DNA metabolism. Annu Rev Biochem 66:61–92. 2. Haring SJ, Mason AC, Binz SK, Wold MS (2008) Cellular functions of human RPA1. Multiple roles of domains in replication, repair, and checkpoints. J Biol Chem 283: 19095–19111. 3. Wang Y, et al. (2005) Mutation in Rpa1 results in defective DNA double-strand break repair, chromosomal instability and cancer in mice. Nat Genet 37:750–755. 4. Mackay J, Steel CM, Elder PA, Forrest AP, Evans HJ (1988) Allele loss on short arm of chromosome 17 in breast cancers. Lancet 2:1384–1385. 5. Baker SJ, et al. (1989) Chromosome 17 deletions and p53 gene mutations in colorectal carcinomas. Science 244:217–221. 6. Seitz S, et al. (2001) Detailed deletion mapping in sporadic breast cancer at chromosomal region 17p13 distal to the TP53 gene: Association with clinicopathological parameters. J Pathol 194:318–326. 7. Tharapel SA, Kadandale JS (2002) Primed in situ labeling (PRINS) for evaluation of gene deletions in cancer. Am J Med Genet 107:123–126. 8. Vermeer MH, et al. (2008) Novel and highly recurrent chromosomal alterations in Sézary syndrome. Cancer Res 68:2689–2698. 9. Jacks T, et al. (1994) Tumor spectrum analysis in p53-mutant mice. Curr Biol 4:1–7. 10. Kirsch DG, et al. (2007) A spatially and temporally restricted mouse model of soft tissue sarcoma. Nat Med 13:992–997. 11. Popanda O, et al. (2000) Mutation analysis of replicative genes encoding the large subunits of DNA polymerase alpha and replication factors A and C in human sporadic colorectal cancers. Int J Cancer 86:318–324. 12. Michiels S, et al. (2007) Polymorphism discovery in 62 DNA repair genes and haplotype associations with risks for lung and head and neck cancers. Carcinogenesis 28: 1731–1739. 13. Abramova NA, Russell J, Botchan M, Li R (1997) Interaction between replication protein A and p53 is disrupted after UV damage in a DNA repair-dependent manner. Proc Natl Acad Sci USA 94:7186–7191. 14. Fanning E, Klimovich V, Nager AR (2006) A dynamic model for replication protein A (RPA) function in DNA processing pathways. Nucleic Acids Res 34:4126–4137.

15. Goodrow TL, et al. (1992) Murine p53 intron sequences 5-8 and their use in polymerase chain reaction/direct sequencing analysis of p53 mutations in CD-1 mouse liver and lung tumors. Mol Carcinog 5:9–15. 16. French JE, et al. (2001) Loss of heterozygosity frequency at the Trp53 locus in p53deficient (+/-) mouse tumors is carcinogen- and tissue-dependent. Carcinogenesis 22: 99–106. 17. Mao JH, et al. (2003) Genetic interactions between Pten and p53 in radiation-induced lymphoma development. Oncogene 22:8379–8385. 18. Mao JH, et al. (2005) Genomic instability in radiation-induced mouse lymphoma from p53 heterozygous mice. Oncogene 24:7924–7934. 19. Mao JH, et al. (2007) Crosstalk between Aurora-A and p53: Frequent deletion or downregulation of Aurora-A in tumors from p53 null mice. Cancer Cell 11:161–173. 20. Cichowski K, et al. (1999) Mouse models of tumor development in neurofibromatosis type 1. Science 286:2172–2176. 21. Vogel KS, et al. (1999) Mouse tumor model for neurofibromatosis type 1. Science 286: 2176–2179. 22. Chen W, et al. (2004) Epigenetic and genetic loss of Hic1 function accentuates the role of p53 in tumorigenesis. Cancer Cell 6:387–398. 23. Zhu Y, et al. (2005) Early inactivation of p53 tumor suppressor gene cooperating with NF1 loss induces malignant astrocytoma. Cancer Cell 8:119–130. 24. Alcantara Llaguno S, et al. (2009) Malignant astrocytomas originate from neural stem/ progenitor cells in a somatic tumor suppressor mouse model. Cancer Cell 15:45–56. 25. Biggs PJ, et al. (2003) Allelic phasing of a mouse chromosome 11 deficiency influences p53 tumorigenicity. Oncogene 22:3288–3296. 26. de Smith AJ, et al. (2007) Array CGH analysis of copy number variation identifies 1284 new genes variant in healthy white males: Implications for association studies of complex diseases. Hum Mol Genet 16:2783–2794. 27. Overholtzer M, et al. (2003) The presence of p53 mutations in human osteosarcomas correlates with high levels of genomic instability. Proc Natl Acad Sci USA 100: 11547–11552.

Materials and Methods

Wang et al.

PNAS | March 23, 2010 | vol. 107 | no. 12 | 5515

GENETICS

tion of the RPA1 gene. This copy number loss was present in 14% of individuals tested and may represent a common allele in the Caucasian population. Our data suggest that this allele could provide a protective effect against TP53 LOH events that occur in cis (26). Our observations that cis and trans mice developed more sarcomas than lymphomas compared with Trp53+/− mice, and that the sarcomas contained more chromosomal abnormalities than lymphomas, suggest that the increased chromosomal instability precipitated by the mutation in Rpa1 programmed the genetic events in Trp53+/− tumor cells toward the development of sarcomas even though the Rpa1L230P mutation ultimately was lost by LOH. This assumption is further supported by previous studies showing a high level of genomic instability in osteosarcomas associated with TP53 mutations (27). In summary, the compound Rpa1; Trp53 mutant mice represent an excellent model to demonstrate the complex effects of genetically linked mutations on cancer predisposition. We show that a single Rpa1 missense mutation serves as a genetic modifier of p53dependent tumorigenesis in mice. In trans mice, the mutation enhanced tumorigenesis by increasing the overall genetic instability and augmenting LOH throughout the entire chromosome 11 due to what is essentially a “hit and run” mechanism. In contrast, in the cis configuration the Rpa1L230P mutation prevented LOH of the closely linked Trp53 locus thereby reducing tumor incidence and altering the tumor spectrum. These studies demonstrate that polymorphic genetic variants in cell essential genes can genetically affect closely linked disease-associated loci such as tumor suppressor genes via allelic phasing, which can result in profound phenotypic variations in tumorigenesis.