Pathogenesis and Consequences of Uniparental …...to autozygosity, germline UPD can arise as a result of mistakes in meiosis, in which all cells in the organism contain the change,

Review

Pathogenesis and Consequences of Uniparental Disomy in Cancer

Hideki Makishima1 and Jaroslaw P. Maciejewski1,2

AbstractThe systematic application of new genome-wide single nucleotide polymorphism arrays has demon-

strated that somatically acquired regions of loss of heterozygosity without changes in copy number

frequently occur in many types of cancer. Until recently, the ubiquity of this type of chromosomal defect

had gone unrecognized because it cannot be detected by routine cytogenetic technologies. Random and

recurrent patterns of copy-neutral loss of heterozygosity, also referred to as uniparental disomy, can be

found in specific cancer types and probably contribute to clonal outgrowth owing to various mechanisms.

In this review we explore the types, topography, genesis, pathophysiological consequences, and clinical

implications of uniparental disomy. Clin Cancer Res; 17(12); 3913–23. �2011 AACR.

Introduction

Chromosomal aberrations constitute a hallmark of acancer genome. Recurrent balanced chromosomalabnormalities, such as those found in distinct types ofleukemia and lymphoma, can be diagnostic and oftenexplain the pathogenesis of such conditions. Chromoso-mal defects also serve as excellent clonal markers and areessential for the diagnosis of a malignant clone or thedetection of minimal residual disease or relapse, especiallyfor cancers that arise in the hematopoietic system (1–3).However, the frequent complexity of chromosomal defectsand inability to obtain viable cells make it difficult to applydiagnostic cytogenetic testing in solid tumors. Conse-quently, relatively few chromosomal defects that can func-tion as diagnostic or prognostic markers have beenidentified to date, although this has begun to change inrecent years. Loss of heterozygosity (LOH) owing to seg-mental or numerical chromosomal deletion is particularlyimportant and is one of several paradigms of malignanttransformation, including the concept of tumor suppressorgene inactivation and Knudson’s 2-hit hypothesis (4). Afterthe loss of chromosomal materials containing one allele,the remaining allele can be affected by somatic mutation orharbor a disease-prone polymorphic variant. Similarly, lossof chromosomal material can lead to LOH and the con-version of a heterozygous inherited (potentially function-ally silent) mutation to a hemizygous mutation. However,the discovery of uniparental disomy (UPD), also referred toas copy-neutral LOH, suggests that LOH may not necessa-

rily be due to the loss of chromosomal material. Undernormal circumstances, each autosome has 2 copies (pater-nal andmaternal) that carry discrete differences encoded bysingle nucleotide polymorphisms (SNPs), and these differ-ences can be used to distinguish between them. In regionsof UPD, portions of one of the chromosomes are lost andreplaced by the exact copy of the remaining chromosome(either paternal or maternal), resulting in the retention of 2copies of genetic information but the loss of polymorphicdifferences that existed due to the presence of maternal andpaternal genes in this region of a diploid chromosome set.Because of the lack of change in the copy number, UPDremains undetected by metaphase cytogenetics. However,studies using microsatellite analysis or genotyping ofsequential SNPs combined with copy number determina-tion have shown that various types of UPD frequently occurin the cancer genome (5, 6). In this review, we discuss thegenesis and types of UPD seen in malignant and normalcells. We focus on mechanisms of selective growth advan-tage that can result from this lesion and discuss theirmechanistic role in cancer pathogenesis. We also reviewthe recurrent regions of UPD identified in various forms ofcancer and the clinical implications associated with thesedefects.

Genetic and Genomic Implications of UPD

LOH due to loss of chromosomal material versus UPDUntil recently, LOH has been most consistently linked to

deletions of chromosomal material in somatic cytogeneticdefects encountered in cancer. In contrast, UPD has beenidentified through the study of inherited diseases, becauseUPD can occur as a germline lesion leading to isodisomy.Heterodisomy, another possible outcome of germlineUPD, does not result in LOH (7). Inherited UPD was firstdescribed by Engel (8) in 1980. It can affect whole chro-mosomes or fragments of chromosomes, and can be inter-stitial or telomeric (Fig. 1). Principally, UPD corresponds toa duplication of either paternal (unipaternal disomy) or

Authors' Affiliations: 1Department of Translational Hematology andOncology Research, Taussig Cancer Institute, and 2Department of Hema-tologic Oncology and Blood Disorders, Cleveland Clinic, Cleveland, Ohio

Corresponding Author: Jaroslaw P. Maciejewski, Taussig Cancer Insti-tute/R40, 9500 Euclid Ave., Cleveland, OH 44195. Phone: 216-445-5962;Fax: 216-636-2498; E-mail: [email protected]

doi: 10.1158/1078-0432.CCR-10-2900

�2011 American Association for Cancer Research.

ClinicalCancer

Research

www.aacrjournals.org 3913

Research. on April 1, 2020. © 2011 American Association for Cancerclincancerres.aacrjournals.org Downloaded from

Published OnlineFirst April 25, 2011; DOI: 10.1158/1078-0432.CCR-10-2900

http://clincancerres.aacrjournals.org/

maternal (unimaternal disomy) alleles, and thus to homo-zygosity for germline allelic variants. LOH due to deletionresults in hemizygosity, whereas UPD results in homozyg-osity (Fig. 1). Theoretically, it is also possible that trisomy isassociated with LOH in the form of uniparental trisomy,which is invariably related to numerical aberrations(Fig. 1). Conceptually, any trisomy might represent a formof UPD without LOH, as both parental alleles are retainedwhile one is duplicated.

In contrast to constitutional UPD, the genesis of somaticUPD is not well understood. However, it may be a result ofmitotic recombination or a successful attempt to correct forthe loss of chromosomal material by duplicating theremaining allele (Fig. 2).

Mechanics of UPD. The chromosomal mechanicsbehind UPD have been intensely investigated in embry-ology and inherited conditions. In consanguineous popu-lations, homozygosity is frequent and can accumulatein the form of stretches of autozygosity. Althoughsuch changes do not always result in pathology, auto-zygosity represents a risk for inherited disease, includinggenetic predisposition to various cancers (9). In contrast

Translational Relevance

We have prepared a review on uniparental disomy(UPD), a new genomic defect that is shedding light onthe pathogenesis of cancer, interactions among muta-tions and chromosomal defects, and why certain muta-tions occur in homozygous form. The review explains ina nonspecialist, medical/pathogenetic context theimplications of UPD for embryogenesis and cancerevolution. Previous Clinical Cancer Research articlesreported the identification of several important genemutations, which was greatly facilitated by the detectionof UPD in the affected areas. UPD highlights areascontaining gene mutations with a homozygous config-uration. In this way, we and other investigators haveidentified CBL and TET2 mutations in myelodysplasticsyndrome, and novel EZH2 mutations in associationwith UPD7q. With the development of new genomictools, it is becoming easier to detect UPD. We believethis review will be educational for all oncologists and ofinterest to any reader.

© 2011 American Association for Cancer Research

Duplication of deletion event

Duplication of somatic mutationa b

b b

Acquiredhomozygosityfor minor(disease-prone)allelic variants

SegmentalNumerical

Deletion

Trisomy:UPD withoutLOH

SegmentalNumerical

UPD

UTP

Diploid

Change in epigenetic silencing patterns

Loss of silencedallele: repression

Loss of unmethylatedallele: silencing

Figure 1. Pathogeneticconsequences of UPD. Lightorange background, types ofsomatic UPD: segmental,numerical, uniparental trisomy,and trisomy with UPD withoutLOH as both parental alleles areretained while one is duplicated.Right, consequences of UPD,including duplication ofmonoallelic deletion leading tobiallelic deletion, duplication ofdisease-prone germlinepolymorphism or mutation,duplication of a somaticmutational event, and gain or lossof imprinting.

Makishima and Maciejewski

Clin Cancer Res; 17(12) June 15, 2011 Clinical Cancer Research3914




to autozygosity, germline UPD can arise as a result ofmistakes in meiosis, in which all cells in the organismcontain the change, or in the initial mitoses after fertiliza-tion, resulting in tissue mosaicism. Autozygosity andmeiotic UPD are not distinguishable without a pedigreeanalysis (Fig. 3). Germline UPD can be due to meioticchromosomal missegregation and subsequent mitoticreassortment leading to a balanced genome. Variousscenarios can lead to germline meiotic UPD, and hetero-disomy and isodisomy (resulting in LOH) have to bedistinguished from numerical chromosomal defects. Tri-somic rescue following errors in meiosis I or II can resultin UPD as one of the possible outcomes (Fig. 3) (10).Unlike germline UPD and autozygosity, somatic UPD

results frommitotic homologous recombination events, orit may represent an attempt to correct for the unbalancedloss of chromosomal material by using the remainingalleles as a template (Fig. 2). Numerical somatic UPDcan occur as a result of mitotic errors, including nondis-junction or loss of a chromosome due to anaphase lagfollowed by reduplication of the remaining chromosome(11). Segmental telomeric UPD may be due to mitotic

homologous recombination between highly homologous,low-copy-number repeats (12, 13). However, such amechanism would be more difficult to invoke for segmen-tal interstitial UPD because it would require 2 consecutiveor simultaneous homologous recombination steps. It ispossible that segmental UPD results from initial deletionfollowed by a compensatory reduplication of the remain-ing chromosomal fragment (Figs. 2 and 3). For diagnosticand investigational purposes, the ability to distinguishbetween a somatic, clonal UPD and germline UPD orautozygosity is of the utmost importance.

UPD and Cancer

Specific chromosomal regions affected by UPD incancer

The pivotal description of UPD in a hematologicalmalignancy came from a study of polycythemia vera(PV) (14), in which UPD9p was found in 33% of patients,constituting the most common chromosomal lesion in thisdisease (15). Later this chromosomal defect was linked to aJAK2V617F mutation (16). More-comprehensive studies

Figure 2. Possible mechanismsleading to somatic UPD. Upperportion, occurrence of somaticUPD may lead to clonalprogression; UPD will be presentin clonal cells only. Lower portion,panels 1 and 2: segmental UPD;panel 3: numerical UPD. Panel 1. Asegmental deletion event iscorrected through duplication ofthe deleted region, using theremaining chromosome as atemplate. a1–d1, possibleoutcomes of chromosomesegregation in the progeny thatcould lead to various types ofsegmental UPD. Panel 2. A mitoticrecombination event leads toexchange of chromatids withvarious possible outcomes(a2–d2). Panel 3. A numerical UPDcan be a result of chromosomalmissegregation (a3–d3).


Acquired Clonal UPD

Mitotic recombination

a1

a3 b3 c3 d3

Mitosis b1 c1 d1

a2 b2 c2 d2

Somatic UPD upon Malignantprogression

UPD in Cancer

www.aacrjournals.org Clin Cancer Res; 17(12) June 15, 2011 3915




demonstrated that JAK2V617F mutations with UPD9p canalso be found in other myeloproliferative neoplasms(MPNs) (17, 18). For example, primary myelofibrosis(PMF) reveals a high frequency of UPD9p with JAK2V617Fmutations (44%) (19). However, the homozygous muta-tional burden varies because of differences in the popula-tion size of mutant cells. Even in purified myeloid cellpopulations, heterozygous and homozygous cells can befound. Moreover, patients with essential thrombocytosis(ET) exhibit a lower frequency of UPD9p, and the resultantJAK2V617F mutational burden in ET is low compared withPMF and PV (16, 20).

Systemic application of SNP arrays as a karyotyping tool(Fig. 4) led to further discoveries of recurrent regionsofUPD

in various myeloid and lymphoid malignancies, with sec-ondary acute myeloid leukemia (sAML), myelodysplasticsyndrome (MDS) or MPN, and chronic myelomonocyticleukemia (CMML) showing particularly high frequencies ofthis type of chromosomal lesion. Investigations of solidtumors produced comparable results with the identificationof recurrent areas of acquired UPD in a wide spectrum ofcancers, some of which show a remarkable predilection forthis type of chromosomal defect (Table 1).

Tumor-specific recurrent regions of UPD can be mappedfor AML and MDS (Fig. 5), as well as for a representativecollection of cell lines derived from various malignancies(Supplemental Fig. 1). Many of the commonly affectedareas contain important genes that have been implicated in


Gamete Zygote

Zygote

Zygote

Trisomicrescue

Non-disjunctionin Meiosis II

Tissue mosaicismNot a clonal event

Constitutional non-clonal UPD

Germ line UPD/Autozygosity

Early embryonicmitotic UPD

Meiotic UPD/Autozygosity

Gamete

Zygote

Parent

Autozygosity

Postfertilization(mitotic) error

Gametecomplementation

Parent

Gamete

Parent

Normalmeiosis

Figure 3. Constitutional versus somatic causes of UPD. Upper portion, early embryonic UPD results in nonclonal tissue mosaicism, whereas inautozygosity and meiotic UPD, all cells of the body will be affected. If it occurs during meiosis I, the gamete will contain 2 chromosomes inherited fromsperm (ovum) and one chromosome from ovum (sperm) leading to heterodisomy (2 different chromosomes inherited from one of the parents). If it occursduring meiosis II, the gametes could contain 2 homologous chromosomes inherited from the ovum (sperm), resulting in isodisomy (lower portion).Heterodisomy can also result from fertilization of a nullisomic gamete and disomic gamete, and isodisomy by fertilization between nullisomic andmonozygousgametes followed by duplication of the remaining copy of the chromosome. Segmental disomy is likely to occur via nonhomologous recombination.






malignant pathogenesis. It should be noted that manystudies probably overestimated the frequency of somaticUPD owing to the lack of distinction from germline-encoded UPD. Nevertheless, the somatic nature of thesealterations can be clearly deduced from the size and loca-tion of the reported regions of UPD and their recurrence.The impact that specific regions of UPD have in terms ofprognosis or diagnosis is currently being evaluated. Forexample, UPD7q, UPD11q, and UPD17p have been linkedto poor outcomes in myeloid malignancies (21–23).Predisposition to UPD in cancer. The fact that certain

cancers, such as MUTYH-associated polyposis colon carci-nomas (24), display a higher frequency of somatic UPDand also can accumulate multiple areas of UPD (complexUPD) implies that there is an inherited or acquired pre-disposition to this type of defect owing to the presenceof fragile sites prone to recombination or a specific type ofchromosomal instability. Particularly high frequencies of

somatic UPD have been described for some malignancies,suggesting that this type of defect may be related to patho-logical pathways that are common in some cancers butabsent in others. For instance, in sporadic colon cancer,physical loss of chromosomal material is characteristic andUPD is less common (25).

Preferred sites of mitotic recombination leading to UPD,with a clustering of the centromeric and telomeric break-points, have been identified (12). In a study of mantle celllymphoma (MCL), the breakpoints flanking all of the geno-micUPDswere significantly associatedwithgenomic regionsenriched in copy number variants and segmental duplica-tions, suggesting that recombination in these regions mayplaya role in thegenomic instabilityofMCL(26). Similarly, acareful analysis of the sites of acquiredUPDorigin in low-riskMDS showed that 43% of UPD regions were localized to orformed part of a previously identified fragile site (27). Fragilesites correspond to known locations of frequent genomic


Totalcopy number

Affymetrix 250K array (CNAG)

Affymetrix 6.0 array (Genotyping Console)

UPD 6pterp12.3DiploidChr. 7

UPD7q

DiploidChr. 4

UPD4q

Copy number

Loss ofheterozygosity

Hetero SNPcall

Allele specificcopy number

AAAB

Allelicdifference Germ

line

Germline

Tumor

Tumor

BB

Figure 4. Detection of UPD using SNP arrays. Left, chromosome ideograms showing exemplary UPD of chromosome 6 detected by (upper portion)250K Affymetrix array (CNAG software) and (lower portion) Affymetrix 6.0 arrays (genotyping console). Right, examples of UPD: UPD7q (250K array)and UPD4q (6.0 arrays).

UPD in Cancer





Table 1. UPD and affected genes in various cancers

UPD Disease Frequency of UPD Affected gene Abnormality

1p MDS/MPN, MPN 8.9% (MDS/MPN), 16.7%(RARST), 4.7% (MPN)

MPL mutation

2p Colorectal cancer 11.1% (cell line), 52% (HNPCC) MSH2 mutation deletion2q MCL 10–16.7% (cell line) MAP2 deletion3p Colorectal cancer 22.2% (cell line), 4% (HNPCC) MLH1 mutation

Colorectal cancer,esophageal cancer

73.9% (esophageal cancer),1.1% (CC)

FHIT deletion

4q MDS, MPN, MDS/MPN,AML

8.8% (MDS/MPN), 3.9–8.7%(MDS)

TET2 mutation

5q Colorectal cancer 28.6–44.4% (cell line) APC mutation6p Loss of graft versus

leukemia effect29.4% (leukemia relapse afterhaploidentical transplantation)

HLA-A, -B, -C loss of mismatch

6q B-cell lymphomas 8% (FL), 3.1% (DLBCL), 10.3%(MALT)

A20 mutation deletion

7q MDS, MDS/MPN, AML 6% (MDS/MPN) EZH2 mutation9p AML 2.6–3.1% (AML with normal

karyotype), 5% (AML)JAK2 mutation

MPN, MDS/MPN 11% (MDS/MPD), 25–43%(MPD), 41–80% (PV), 5.9–17% (ET), 43.8–67% (PMF)

mutation

AML 2.6% (AML with normalkaryotype)

CDKN2A deletion

ALL 7.1–29% (pediatric ALL) deletionFollicular lymphoma 33% (cell line) deletionMCL 60% (cell line), 7.1% (primary

sample)deletion

Esophagealcarcinoma

26.1% (primary sample) deletion

Ovarian cancer 7.5% (primary sample) deletionGlioblastoma 3.3% (primary sample) deletionNeuloblastoma 4.3% (primary sample) deletionCNS lymphoma 21.1% (primary sample) CDKN2A methylation deletionColorectal cancer 55.6% (cell line) methylation

9q BCC 35.7% (primary sample) PTCH mutation11p AML 3.2–4.5% (AML primary

sample), 6.4% (APL)WT1 mutation

AML 4.7% (primary sample) H19 methylationHepatoblastoma 23.5% (primary sample) IGF2, H19 methylationRhabdomyosarcoma 33.3% (primary sample) HRAS mutationWilms’ tumor 2.5–5.6% (primary sample) CDKN1C, IGF2, H19 methylationWilms’ tumor 36% (primary sample) WT1 mutationBeckwith-Wiedemann

syndrome7.2–16.8% (primary sample) CDKN1C, IGF2, H19 methylation

11q MDS/MPN 4.9% (primary sample) CBL mutation13q AML 2.3–5.4% (primary sample) FLT3 mutation

CLL 3.6% (primary sample) miR-15a, miR-16–1 deletionMCL 10% (MCL cell line) RB1 deletionBreast cancer 6% (primary sample) deletionOvarian cancer 23.8% (primary sample) deletionRetinoblastoma 59.5% (primary sample) mutationOvarian cancer 15% (primary sample) BRCA2 mutation

17p MDS, CLL 1.8% (MDS), 6.1% (CLL) P53 mutation

(Continued on the following page)






instability and are associated with breakpoints occurring inthegenerationof chromosomalaberrations inhematologicalmalignancies (28). Fragile sites have also been associatedwith regulatory micro RNA amplifications and deletions(29).Risk factors for the acquisition of UPD also include the

presence of BRCA mutations, which have been associatedwith an increased frequency of UPD that is not observed incases of spontaneous breast cancer (30). In ovarian cancer,UPD is frequently observed in tumors with an inheritedBRCA mutation (31). Microsatellite instability (MSI) hasbeen shown to be associated with an increased frequency ofUPD (32). In one study (33), MSI was present in 60% ofpatients who had AML and regions of UPD (Fig. 5),whereas single-locus MSI was absent in patients withAML in whom UPD was not detected.Pathogenic consequences of UPD in cancer. Although it

is likely that chromosomal deletions occur randomly, thosethat result in a proliferative advantage or resistance to, e.g.,physiological apoptosis could initiate clonal outgrowth.Selection for clones with a specific region of LOH could berelated to a somatic or germline loss of a wild-type allele,resulting in hemizygosity for an SNP-encoded, disease-prone allele or a somatic or germline mutated allele(Fig. 1). If the affected area includes promoters of allelesthat are differentially silenced (imprinted), deletion canlead to either a gain of imprinting (GOI) or loss of imprint-ing (LOI). This can result in changes in gene expression.UPD can also lead to the duplication of an imprintedexpressed allele or a silenced (methylated), imprintedallele. When the transcription of both alleles is requiredfor normal cellular physiology, deletions can result inpathological haploinsufficiency, and thus LOH is less likelyto play a pathogenic role (Fig. 1).There are similarities and important differences

between the consequences of LOH due to deletion orUPD. UPD may convey a permissive growth advantagewhen, in accordance with the 2-hit hypothesis, an initialheterozygous mutation is duplicated through UPD. Thismay result in homozygosity of a somatic mutation thatinactivates a tumor suppressor gene, such as occurs withTP53 in UPD17p, RUNX1 in UPD21q, and many others

(Fig. 6). Activating oncogenic mutations can be dupli-cated through UPD, leading to increased proliferativedrive though a double dose of the mutated gene product.Such a scenario has been encountered with JAK2(UPD9p) (14, 16, 34), FLT3 internal tandem duplication(UPD13q) (35, 36), WT1 (UPD11p) (37, 38), and MPL(UPD1p) (19, 39). Recently, we and other groups foundloss-of-function mutations of EZH2, encoding trimethyl-transferase of H3K27, in patients with UPD7q (Fig. 6)(40–42). Because methylation of H3K27 is a histonerepressive mark associated with condensation of chroma-tin, its loss-of-function mutation results in chromatinrelaxation and accelerated gene transcriptions, as inWNT1 and HOXA family genes. EZH2 mutations aremore commonly homozygous (UPD7q) than heterozy-gous (40, 42). In myeloid malignancies, UPD is fre-quently observed on chromosome 11q specifically inthe MDS/MPN phenotype. CBL mutations were observedin 76% of patients with UPD11q, but were relatively rare(<5%) in patients with deletion or without LOH 11q. Inmutations associated with recurrent UPD, homozygositymay provide a malignant clone with a further growthadvantage. Theoretically, a similar effect could be pro-duced by LOH as a result of deletion, but for some genes,such as CBL, most mutations are homozygous, and cor-responding deletions have not been found to harborhemizygous mutations (22, 43, 44). Consequently, theCBL knockout is less leukemogenic than ring fingerdomain mutant knockin CBL null mice (43, 45). Bycontrast, TP53 or TET2 mutations are associated withboth deletions and UPD. UPD can also affect germlineheterozygous mutations. Examples of such mutationsinclude UPD11q and UPD17q leading to duplicationof CBL and NF1 mutations in juvenile myelomonocyticleukemia (JMML) (Fig. 6) (46–48).

Somatic UPD can also lead to duplication and hencehomozygosity of disease-prone SNPs that are silent in aheterozygous configuration. Typically, such SNPs show anexceeding low frequency of homozygosity for the minorallele in the general population. Of note, however,although duplication of initially heterozygous mutationsis a driving force for clonal dominance, areas of UPD

Table 1. UPD and affected genes in various cancers (cont'd )

UPD Disease Frequency of UPD Affected gene Abnormality

Follicular lymphoma 19.2% (transformed case) mutationMCL 3.8–10.7% (MCL), 10%

(cell line)mutation

Colorectal cancer 57.1% (cell line) mutationBreast cancer 6% (primary sample) mutationGlioblastoma 3.3% (primary sample) mutation

17q JMML 25–80% (primary sample) NF1 mutationOvarian cancer 40% (primary sample) BRCA1 mutation

19q AML 0.6–1.6% (primary sample) CEBPA mutation21q AML 2.6% (AML with normal karyotype) RUNX1 mutation

UPD in Cancer





contain a large number of genes that can include germlinepolymorphisms and imprinted sites. For example, CNTLN,KANK1, DMRT1, TOPORS, and MLANA genes have beenreported to be either imprinted or differentially methylatedon chromosome 9p, which is frequently affected by UPD.Therefore, UPD at this site can lead to GOI and LOI forspecific genes (49–52). In addition, UPD9p contains alarge number of nonsynonymous polymorphisms forwhich either minor or major alleles can be duplicatedand result in discrete changes of the phenotype. Thesefindings indicate that association of the high allelic burdenof JAK2V617F mutation due to UPD9p may be influencedby homozygous SNPs and/or loss or gain of expression ofmonoallelically expressed genes in the region affected byUPD (53). More recently, multiple groups described therelationship between JAK2’s genetic predisposition andJAK2V617F, and reported that the 46/1 JAK2 haplotypepredisposes to the development of JAK2V617F-associatedMPN (54–56). Subsequently, Tefferi and colleagues (57)

observed that a JAK2 germline genetic variation(rs12343867 genotype CC) was less frequent in PMF witha high JAK2V617F burden. This suggests that the allelicdistortion from acquired UPD contributes to the appear-ance of a more pronounced effect on disease susceptibilityin JAK2V617F-positive patients.

Examples of changes in LOI and GOI can also be foundin cancer-prone inherited disorders associated with UPD,such as GOI of H19 and LOI of IGF2 (58, 59). A similaralteration of imprinting patterns has been found in hepa-toblastoma, a tumor characterized by frequent UPD11paffecting H19 and IGF2 (60). LOI for IGF2 and H19 due toUPD is evident in colon carcinoma (61) andWilms’ tumor(WT) (62), ARH1 LOI is evident in ovarian and breastcancer (63), and H19 LOI is evident in AML (64). Some ofthese events may be due to a shared mechanism of UPD.Thus, it is possible that although deletion or duplicationcan randomly affect each parental chromosome, clonalselection may favor the expressed or silenced imprinted


1

5

11

17 18 19 20 21 22

12 13 14 15 16

6 7 8 9 10

2 3 4

Figure 5. Mapping of recurrent UPD by chromosome in AML and MDS. Ideograms of chromosomes: (left) blue and red bars indicate somatic UPD inMDS and AML, respectively; (right) blue bars depict regions of UPD detected in DNA from 1003 controls. In our own analysis of healthy controls, nonclonalUPD and autozygosity were found in 12% of the samples; the majority (97%) of these regions of homozygosity are interstitial and <25 Mb in length. Thetopography and size of the somatic areas of UPD differ strikingly from those of the copy-neutral regions of LOH in the germline.






allele and thus may not be random. In MDS, for example,the FZD9 promoter has been found to be consistentlyhypermethylated in patients with LOH7q as a result ofUPD or chromosomal deletion (65). Theoretically, severalof these mechanisms could be operative in UPD, affecting alarge number of genes and contributing to the heteroge-neity of the resulting tumor phenotype and clinical beha-vior (Fig. 6).

Conclusions

UPD is a previously cryptic type of chromosomal aberra-tion that occurs ubiquitously in cancer and often maps toregions that are affected by loss of chromosomal material.During malignant evolution, the clonal selection process

favors duplication of heterozygous somatic or germlinemutations, disease-prone SNP alleles, or imprinting pat-terns that can produce a selective advantage. The develop-ment of whole-genome scanning technologies with SNParrays has greatly facilitated the detection of UPD. Inaddition to the somatic form of UPD, SNP arrays can detectstretches of homozygosity due to inherited UPD or auto-zygosity, which requires truly clonal events to be distin-guished from nonclonal homozygosity. Inherited UPD orautozygosity may constitute an independent predisposi-tion factor for the development of malignancy. Althoughthis theory is supported by the increased prevalence ofcancer found in inbred populations, it needs to be furtherexplored. Similarly, the mechanisms that lead to the acqui-sition of somatic UPD remain to be clarified. However,


Germ line mutationsHomozygous deletion Genomic imprinting

Chr 9

AMLALLFL

CDKN2A

Chr 13

CLL

miR – 15amiR – 16 – 1

Chr 1

BC OC

Chr 11

HCC HB CCWT AML

Chr 2 Chr 3

HNPCC

MSH2 MLH1

Chr 5

Desmoidtumor (FAP)

APC

Chr 9

BCC(Gorlin

syndrome)

PTCH1

Chr 17

JMML(Neurofibromatosis

type 1)

NF1

Chr 12

ALL(Noonan syndrome)

PTPN11

Chr 11

Rhabdomyosarcoma(Costello syndrome)

HRAS

Somatic mutations

Chr 1

MDS/MPN

MPL

NRAS

Chr 9

BCC MPN

PTCH1

JAK2

Chr 13

AML

FLT3Chr 17

FL

TP53 Chr 19

AMLCEBPA

Chr 21

AMLRUNX1

Chr 11

WT CMML

WT1

CBL

Chr 7

EZH2MDS/MPN

Chr 4

CMML

KIT

TET2

ARH1

CDKN1CIGF2H19

Figure 6. Examples of recurrent UPDs and corresponding homozygous mutations. Examples of recurrent areas of UPD in various human cancers associatedwith specific molecular lesions. Pale orange background, duplications of somatic mutations. Blue background, duplication of segmental losses ofchromosomal material leading to biallelic deletions. Green background, changes in genomic imprinting due to UPD. Pink background, duplication of germlinemutations. (MDS, myelodysplastic syndrome; MPN, myeloproliferative neoplasm; AML, acute myeloid leukemia; CMML, chronic myelomonocytic leukemia;FL, follicular lymphoma; BCC, basal cell carcinoma; WT, Wilms’ tumor; ALL, acute lymphoblastic leukemia; CLL, chronic lymphocytic leukemia; BC, breastcancer; OC, ovarian cancer; HCC, hepatocellular carcinoma; HB, hepatoblastoma; CC, colorectal cancer; HNPCC, hereditary nonpolyposis colorectal cancer;FAP, familial adenomatous polyposis; JMML, juvenile myelomonocytic leukemia).

UPD in Cancer





distinct pathways are probably involved in the develop-ment of numerical and segmental (interstitial or telomeric)UPD. In similarity to other chromosomal lesions, includ-ing gains, losses, and balanced translocations, variousacquired or inherited defects of the mitotic machinery orDNA repair pathways may be involved in UPD. By eluci-dating themechanisms behind this process, wemay be ableto identify people at risk for developing cancer as a result ofUPD, and to develop new drug targets for the treatment oftumors associated with UPD. In the diagnostic setting,identification of UPD allows one to distinguish betweentruly homozygous genetic changes resulting from contam-ination with wild-type cells and those caused by cells withheterozygous mutations. This distinction can be used tomore precisely assess the extent of a mutation and to better

understand the effect of specific mutations in a particulargenetic context. For example, mutations in TP53most oftenoccur in homo- or hemizygous form, and the presence of awild-type allele is protective. The prognostic significance ofsome of the recurrent regions of UPD has been evaluated,and such defects could be included in a cytogeneticallybased prognostic scoring system. Should UPD prove to beof diagnostic value, cytogenetic methods could be intro-duced to allow for routine detection of this type of defect.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Received October 29, 2010; revised January 15, 2011; accepted January18, 2011; published OnlineFirst April 25, 2011.

References1. Haase D, Germing U, Schanz J, Pfeilst€ocker M, N€osslinger T,

Hildebrandt B, et al. New insights into the prognostic impact ofthe karyotype in MDS and correlation with subtypes: evidencefrom a core dataset of 2124 patients. Blood 2007;110:4385–95.

2. Grimwade D, Walker H, Oliver F, Wheatley K, Harrison C, Harrison G,et al. The importance of diagnostic cytogenetics on outcome in AML:analysis of 1,612 patients entered into the MRC AML 10 trial. TheMedical Research Council Adult and Children's Leukaemia WorkingParties. Blood 1998;92:2322–33.

3. Byrd JC, Mr�ozek K, Dodge RK, Carroll AJ, Edwards CG, Arthur DC,et al. Pretreatment cytogenetic abnormalities are predictive of induc-tion success, cumulative incidence of relapse, and overall survival inadult patients with de novo acute myeloid leukemia: results fromCancer and Leukemia Group B (CALGB 8461). Blood 2002;100:4325–36.

4. Devilee P, Cleton-Jansen AM, Cornelisse CJ. Ever since Knudson.Trends Genet 2001;17:569–73.

5. Maciejewski JP, Mufti GJ. Whole genome scanning as a cytogenetictool in hematologic malignancies. Blood 2008;112:965–74.

6. Maciejewski JP, Tiu RV, O’Keefe C. Application of array-based wholegenome scanning technologies as a cytogenetic tool in haematolo-gical malignancies. Br J Haematol 2009;146:479–88.

7. Smith AC, Shuman C, Chitayat D, Steele L, Ray PN, Bourgeois J,et al. Severe presentation of Beckwith-Wiedemann syndromeassociated with high levels of constitutional paternal uniparentaldisomy for chromosome 11p15. Am J Med Genet A 2007;143A:3010–5.

8. Engel E. A new genetic concept: uniparental disomy and its potentialeffect, isodisomy. Am J Med Genet 1980;6:137–43.

9. Bacolod MD, Schemmann GS, Giardina SF, Paty P, Notterman DA,Barany F. Emerging paradigms in cancer genetics: some importantfindings from high-density single nucleotide polymorphism arraystudies. Cancer Res 2009;69:723–7.

10. Tuna M, Knuutila S, Mills GB. Uniparental disomy in cancer. TrendsMol Med 2009;15:120–8.

11. Strachan T, Read AP. Human Molecular Genetics. New York: GarlandScience; 2003.

12. Stephens K, Weaver M, Leppig KA, Maruyama K, Emanuel PD, LeBeau MM, et al. Interstitial uniparental isodisomy at clustered break-point intervals is a frequent mechanism of NF1 inactivation in myeloidmalignancies. Blood 2006;108:1684–9.

13. van Dartel M, Hulsebos TJ. Amplification and overexpression of genesin 17p11.2 � p12 in osteosarcoma. Cancer Genet Cytogenet 2004;153:77–80.

14. Kralovics R, Guan Y, Prchal JT. Acquired uniparental disomy ofchromosome 9p is a frequent stem cell defect in polycythemia vera.Exp Hematol 2002;30:229–36.

15. Kralovics R, Buser AS, Teo SS, Coers J, Tichelli A, Van Der Maas AP,et al. Comparison of molecular markers in a cohort of patients withchronic myeloproliferative disorders. Blood 2003;102:1869–71.

16. Kralovics R, Passamonti F, Buser AS, Teo SS, Tiedt R, Passweg JR,et al. A gain-of-function mutation of JAK2 in myeloproliferative dis-orders. N Engl J Med 2005;352:1779–90.

17. Yamamoto G, Nannya Y, Kato M, SanadaM, Levine RL, Kawamata N,et al. Highly sensitive method for genomewide detection of alleliccomposition in nonpaired, primary tumor specimens by use of Affy-metrix single-nucleotide-polymorphism genotyping microarrays. AmJ Hum Genet 2007;81:114–26.

18. Jones AV, Kreil S, Zoi K, Waghorn K, Curtis C, Zhang L, et al. Wide-spread occurrence of the JAK2 V617F mutation in chronic myelopro-liferative disorders. Blood 2005;106:2162–8.

19. Kawamata N, Ogawa S, Yamamoto G, Lehmann S, Levine RL, PikmanY, et al. Genetic profiling of myeloproliferative disorders by single-nucleotide polymorphism oligonucleotide microarray. Exp Hematol2008;36:1471–9.

20. Scott LM, Scott MA, Campbell PJ, Green AR. Progenitors homozy-gous for the V617Fmutation occur in most patients with polycythemiavera, but not essential thrombocythemia. Blood 2006;108:2435–7.

21. Gondek LP, Tiu R, O'Keefe CL, SekeresMA, Theil KS, Maciejewski JP.Chromosomal lesions and uniparental disomy detected by SNParrays in MDS, MDS/MPD, and MDS-derived AML. Blood 2008;111:1534–42.

22. Makishima H, Cazzolli H, Szpurka H, Dunbar A, Tiu R, Huh J, et al.Mutations of e3 ubiquitin ligase cbl family members constitute a novelcommon pathogenic lesion in myeloid malignancies. J Clin Oncol2009;27:6109–16.

23. Jasek M, Gondek LP, Bejanyan N, Tiu R, Huh J, Theil KS, et al. TP53mutations in myeloid malignancies are either homozygous or hemi-zygous due to copy number-neutral loss of heterozygosity or deletionof 17p. Leukemia 2010;24:216–9.

24. Middeldorp A, van Puijenbroek M, Nielsen M, Corver WE, JordanovaES, ter Haar N, et al. High frequency of copy-neutral LOH in MUTYH-associated polyposis carcinomas. J Pathol 2008;216:25–31.

25. Kurashina K, Yamashita Y, Ueno T, Koinuma K, Ohashi J, Horie H,et al. Chromosome copy number analysis in screening for prognosis-related genomic regions in colorectal carcinoma. Cancer Sci 2008;99:1835–40.

26. Be�a S, Salaverria I, Armengol L, Pinyol M, Fern�andez V, Hartmann EM,et al. Uniparental disomies, homozygous deletions, amplifications,and target genes in mantle cell lymphoma revealed by integrativehigh-resolution whole-genome profiling. Blood 2009;113:3059–69.

27. Mohamedali A, G€aken J, Twine NA, Ingram W, Westwood N, Lea NC,et al. Prevalence and prognostic significance of allelic imbalance bysingle-nucleotide polymorphism analysis in low-risk myelodysplasticsyndromes. Blood 2007;110:3365–73.






28. G€um€usG,Sunguro�gluA,T€uk€unA,SayinDB,B€okesoy I.Common fragilesites associated with the breakpoints of chromosomal aberrations inhematologic neoplasms. Cancer Genet Cytogenet 2002;133:168–71.

29. Calin GA, Sevignani C, Dumitru CD, Hyslop T, Noch E, Yendamuri S,et al. Human microRNA genes are frequently located at fragile sitesand genomic regions involved in cancers. Proc Natl Acad Sci U S A2004;101:2999–3004.

30. Johnson N, Speirs V, Curtin NJ, Hall AG. A comparative study ofgenome-wide SNP, CGH microarray and protein expression analysisto explore genotypic and phenotypic mechanisms of acquired anti-estrogen resistance in breast cancer. Breast Cancer Res Treat2008;111:55–63.

31. Walsh CS, Ogawa S, Scoles DR, Miller CW, Kawamata N, Narod SA,et al. Genome-wide loss of heterozygosity and uniparental disomy inBRCA1/2-associated ovarian carcinomas. Clin Cancer Res 2008;14:7645–51.

32. Melcher R, Al-Taie O, Kudlich T, Hartmann E, Maisch S, Steinlein C,et al. SNP-Array genotyping and spectral karyotyping reveal unipar-ental disomy as early mutational event in MSS- and MSI-colorectalcancer cell lines. Cytogenet Genome Res 2007;118:214–21.

33. Serrano E, Carnicer MJ, Orantes V, Estivill C, Lasa A, Brunet S, et al.Uniparental disomymay be associated with microsatellite instability inacute myeloid leukemia (AML) with a normal karyotype. Leuk Lym-phoma 2008;49:1178–83.

34. Levine RL, Wadleigh M, Cools J, Ebert BL, Wernig G, Huntly BJ, et al.Activating mutation in the tyrosine kinase JAK2 in polycythemia vera,essential thrombocythemia, and myeloid metaplasia with myelofibro-sis. Cancer Cell 2005;7:387–97.

35. Griffiths M, Mason J, Rindl M, Akiki S, McMullan D, Stinton V, et al.Acquired isodisomy for chromosome 13 is common in AML, andassociated with FLT3-itd mutations. Leukemia 2005;19:2355–8.

36. Raghavan M, Smith LL, Lillington DM, Chaplin T, Kakkas I, Molloy G,et al. Segmental uniparental disomy is a commonly acquired geneticevent in relapsed acute myeloid leukemia. Blood 2008;112:814–21.

37. Grundy P, Wilson B, Telzerow P, Zhou W, Paterson MC. Uniparentaldisomy occurs infrequently in Wilms tumor patients. Am J Hum Genet1994;54:282–9.

38. Fitzgibbon J, Smith LL, Raghavan M, Smith ML, Debernardi S,Skoulakis S, et al. Association between acquired uniparental disomyand homozygous gene mutation in acute myeloid leukemias. CancerRes 2005;65:9152–4.

39. Szpurka H, Gondek LP, Mohan SR, Hsi ED, Theil KS, Maciejewski JP.UPD1p indicates the presence of MPL W515L mutation in RARS-T, amechanism analogous to UPD9p and JAK2 V617F mutation. Leuke-mia 2009;23:610–4.

40. Ernst T, Chase AJ, Score J, Hidalgo-Curtis CE, Bryant C, Jones AV,et al. Inactivating mutations of the histone methyltransferase geneEZH2 in myeloid disorders. Nat Genet 2010;42:722–6.

41. Nikoloski G, Langemeijer SM, Kuiper RP, Knops R, Massop M,T€onnissen ER, et al. Somatic mutations of the histone methyltransfer-ase gene EZH2 in myelodysplastic syndromes. Nat Genet 2010;42:665–7.

42. Makishima H, Jankowska AM, Tiu RV, Szpurka H, Sugimoto Y, Hu Z,et al. Novel homo- and hemizygous mutations in EZH2 in myeloidmalignancies. Leukemia 2010;24:1799–804.

43. SanadaM, Suzuki T, Shih LY, OtsuM, KatoM, Yamazaki S, et al. Gain-of-function of mutated C-CBL tumour suppressor in myeloid neo-plasms. Nature 2009;460:904–8.

44. Grand FH, Hidalgo-Curtis CE, Ernst T, Zoi K, Zoi C, McGuire C, et al.Frequent CBL mutations associated with 11q acquired uniparentaldisomy in myeloproliferative neoplasms. Blood 2009;113:6182–92.

45. Rathinam C, Thien CB, Flavell RA, Langdon WY. Myeloid leukemiadevelopment in c-Cbl RING finger mutant mice is dependent on FLT3signaling. Cancer Cell 2010;18:341–52.

46. Muramatsu H, Makishima H, Jankowska AM, Cazzolli H, O'Keefe C,Yoshida N, et al. Mutations of an E3 ubiquitin ligase c-Cbl but notTET2 mutations are pathogenic in juvenile myelomonocytic leukemia.Blood 2010;115:1969–75.

47. Steinemann D, Arning L, Praulich I, Stuhrmann M, Hasle H, Stary J,et al. Mitotic recombination and compound-heterozygous mutationsare predominant NF1-inactivating mechanisms in children with juve-nile myelomonocytic leukemia and neurofibromatosis type 1. Hae-matologica 2010;95:320–3.

48. Loh ML, Sakai DS, Flotho C, Kang M, Fliegauf M, Archambeault S,et al. Mutations in CBL occur frequently in juvenile myelomonocyticleukemia. Blood 2009;114:1859–63.

49. Gimelbrant A, Hutchinson JN, Thompson BR, Chess A. Widespreadmonoallelic expression on human autosomes. Science 2007;318:1136–40.

50. Sarkar S, Roy BC, Hatano N, Aoyagi T, Gohji K, Kiyama R. A novelankyrin repeat-containing gene (Kank) located at 9p24 is a growthsuppressor of renal cell carcinoma. J Biol Chem 2002;277:36585–91.

51. Lerer I, SagiM,Meiner V, Cohen T, Zlotogora J, AbeliovichD.Deletion oftheANKRD15geneat9p24.3causesparent-of-origin-dependent inheri-tance of familial cerebral palsy. Hum Mol Genet 2005;14:3911–20.

52. Luedi PP, Dietrich FS, Weidman JR, Bosko JM, Jirtle RL, HarteminkAJ. Computational and experimental identification of novel humanimprinted genes. Genome Res 2007;17:1723–30.

53. Kralovics R. Genetic complexity of myeloproliferative neoplasms.Leukemia 2008;22:1841–8.

54. Jones AV, Chase A, Silver RT, Oscier D, Zoi K, Wang YL, et al. JAK2haplotype is a major risk factor for the development of myeloproli-ferative neoplasms. Nat Genet 2009;41:446–9.

55. Olcaydu D, Harutyunyan A, J€ager R, Berg T, Gisslinger B, Pabinger I,et al. A common JAK2 haplotype confers susceptibility to myelopro-liferative neoplasms. Nat Genet 2009;41:450–4.

56. Kilpivaara O, Mukherjee S, Schram AM, Wadleigh M, Mullally A, EbertBL, et al. A germline JAK2 SNP is associatedwith predisposition to thedevelopment of JAK2(V617F)-positive myeloproliferative neoplasms.Nat Genet 2009;41:455–9.

57. Tefferi A, Lasho TL, Patnaik MM, Finke CM, Hussein K, Hogan WJ,et al. JAK2 germline genetic variation affects disease susceptibility inprimary myelofibrosis regardless of V617F mutational status: nulli-zygosity for the JAK2 46/1 haplotype is associated with inferiorsurvival. Leukemia 2010;24:105–9.

58. Henry I, Bonaiti-Pelli�e C, Chehensse V, Beldjord C, Schwartz C,Utermann G, et al. Uniparental paternal disomy in a genetic can-cer-predisposing syndrome. Nature 1991;351:665–7.

59. Rainier S, Johnson LA, Dobry CJ, Ping AJ, Grundy PE, Feinberg AP.Relaxation of imprinted genes in human cancer. Nature 1993;362:747–9.

60. Suzuki M, Kato M, Yuyan C, Takita J, Sanada M, Nannya Y, et al.Whole-genome profiling of chromosomal aberrations in hepatoblas-toma using high-density single-nucleotide polymorphism genotypingmicroarrays. Cancer Sci 2008;99:564–70.

61. Nakagawa H, Chadwick RB, Peltomaki P, Plass C, Nakamura Y, de LaChapelle A. Loss of imprinting of the insulin-like growth factor II geneoccurs by biallelic methylation in a core region of H19-associatedCTCF-binding sites in colorectal cancer. Proc Natl Acad Sci U S A2001;98:591–6.

62. Scott RH, Douglas J, Baskcomb L, Huxter N, Barker K, Hanks S, et al.Constitutional 11p15 abnormalities, including heritable imprintingcenter mutations, cause nonsyndromic Wilms tumor. Nat Genet2008;40:1329–34.

63. Yu Y, Xu F, Peng H, Fang X, Zhao S, Li Y, et al. NOEY2 (ARHI), animprinted putative tumor suppressor gene in ovarian and breastcarcinomas. Proc Natl Acad Sci U S A 1999;96:214–9.

64. Tessema M, L€anger F, Bock O, Seltsam A, Metzig K, Hasemeier B,et al. Down-regulation of the IGF-2/H19 locus during normal andmalignant hematopoiesis is independent of the imprinting pattern. IntJ Oncol 2005;26:499–507.

65. Jiang Y, Dunbar A, Gondek LP, Mohan S, Rataul M, O'Keefe C, et al.Aberrant DNA methylation is a dominant mechanism in MDS progres-sion to AML. Blood 2009;113:1315–25.

UPD in Cancer





2011;17:3913-3923. Published OnlineFirst April 25, 2011.Clin Cancer Res Hideki Makishima and Jaroslaw P. Maciejewski CancerPathogenesis and Consequences of Uniparental Disomy in

Updated version

10.1158/1078-0432.CCR-10-2900doi:

Access the most recent version of this article at:

Material

Supplementary

http://clincancerres.aacrjournals.org/content/suppl/2011/06/09/1078-0432.CCR-10-2900.DC1Access the most recent supplemental material at:

Cited articles

http://clincancerres.aacrjournals.org/content/17/12/3913.full#ref-list-1

This article cites 64 articles, 27 of which you can access for free at:

Citing articles

http://clincancerres.aacrjournals.org/content/17/12/3913.full#related-urls

This article has been cited by 6 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

SubscriptionsReprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. (CCC)Click on "Request Permissions" which will take you to the Copyright Clearance Center's

.http://clincancerres.aacrjournals.org/content/17/12/3913To request permission to re-use all or part of this article, use this link



http://clincancerres.aacrjournals.org/lookup/doi/10.1158/1078-0432.CCR-10-2900

http://clincancerres.aacrjournals.org/content/suppl/2011/06/09/1078-0432.CCR-10-2900.DC1

http://clincancerres.aacrjournals.org/content/17/12/3913.full#ref-list-1

http://clincancerres.aacrjournals.org/content/17/12/3913.full#related-urls

http://clincancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://clincancerres.aacrjournals.org/content/17/12/3913


Documents

Pathogenesis and Consequences of Uniparental …...to autozygosity, germline UPD can arise as a result of mistakes in meiosis, in which all cells in the organism contain the change,