dicer1 mutations in familial multinodular goiter with and ... · author affiliations are listed at...

ORIGINAL CONTRIBUTION

DICER1 Mutations in FamilialMultinodular Goiter With and WithoutOvarian Sertoli-Leydig Cell TumorsThomas Rio Frio, PhDAmin Bahubeshi, BScChryssa Kanellopoulou, PhDNancy Hamel, MScMarek Niedziela, MD, PhDNelly Sabbaghian, MScCarly Pouchet, MScLucy Gilbert, MD, MScPaul K. O’Brien, MBKim Serfas, MScPeter Broderick, PhDRichard S. Houlston, MD, PhDFabienne Lesueur, PhDElena Bonora, PhDStefan Muljo, PhDR. Neil Schimke, MDDorothee Bouron-Dal Soglio, MD, PhD

Jocelyne Arseneau, MDKris Ann Schultz, MD, MSJohn R. Priest, MD, MBAVan-Hung Nguyen, MDH. Rubén Harach, MD, PhDDavid M. Livingston, MDWilliam D. Foulkes, MBBS, PhDMarc Tischkowitz, MD, PhD

MULTINODULAR GOITER

(MNG) is a common dis-order characterized bynodular overgrowth of the

thyroid gland, usually in the setting ofdiffuse parenchymal hyperplasia.1,2 Thebiochemical and genetic basis of manyof the known subtypes of thyroid goiterhas been elucidated,3 but little is knownabout underlying genetic factors in thenontoxic form. Two loci for familialMNG have been identified—MNG1 on

chromosome 14q (Online Mendelian In-heritance in Man [OMIM] %138800)4

and MNG2 on the X chromosome(OMIM %300273).5 Several publishedreports identify ovarian Sertoli-Leydigcell tumors (SLCTs) occurring with fa-milial MNG (eTable 1, available at http://www.jama.com). SLCTs represent lessthan 0.5% of all ovarian neoplasms6 buttypically occur in younger women andare the most common androgen-producing ovarian tumors.7,8

Pleuropulmonary blastoma (PPB), arare pediatric mesenchymal thoracic tu-mor,hasrecentlybeenlinkedtothesameregionasMNG1on14q.Germlinemuta-

tions in DICER1 (locatedat14q32)wereidentified in11of11PPBfamilies featur-ingchildrenwithPPB, cysticnephroma,or embryonal rhabdomyosarcoma,9

knownasthePPBFamilyTumorandDys-plasia Syndrome (PPB-FTDS) (OMIM#601200). Some PPB families also con-tain both MNG and gonadal tumors (in-cluding SLCTs).10,11

As a member of the ribonuclease III(RNase III) family, DICER1 is involved

Author Affiliations are listed at the end of this article.Corresponding Author: William D. Foulkes, MBBS,PhD, Program in Cancer Genetics, Gerald BronfmanCentre for Clinical Research in Oncology, 546 av-enue des Pins, Montreal, QC H2W 1S6, Canada([email protected]).

Context Nontoxic multinodular goiter (MNG) is frequently observed in the generalpopulation, but little is known about the underlying genetic susceptibility to this dis-ease. Familial cases of MNG have been reported, and published reports describe 5 fami-lies that also contain at least 1 individual with a Sertoli-Leydig cell tumor of the ovary(SLCT). Germline mutations in DICER1, a gene that codes for an RNase III endoribo-nuclease, have been identified in families affected by pleuropulmonary blastoma (PPB),some of whom include cases of MNG and gonadal tumors such as SLCTs.

Objective To determine whether familial MNG with or without SLCT in the ab-sence of PPB was associated with mutations in DICER1.

Design, Setting, and Patients From September 2009 to September 2010, wescreened 53 individuals from 2 MNG and 3 MNG/SLCT families at McGill Universityfor mutations in DICER1. We investigated blood lymphocytes and MNG and SLCTtissue from family members for loss of the wild-type DICER1 allele (loss of heterozy-gosity), DICER1 expression, and microRNA (miRNA) dysregulation.

Main Outcome Measure Detection of germline DICER1 gene mutations in famil-ial MNG with and without SLCT.

Results We identified and characterized germline DICER1 mutations in 37 individualsfrom 5 families. Two mutations were predicted to be protein truncating, 2 resulted in in-frame deletions, and 1 was a missense mutation. Molecular analysis of the 3 SLCTs showedno loss of heterozygosity of DICER1, and immunohistochemical analysis in 2 samples showedstrong expression of DICER1 in Sertoli cells but weak staining of Leydig cells. miRNA pro-filing of RNA from lymphoblastoid cell lines from both affected and unaffected membersof the familial MNG cases revealed miRNA perturbations in DICER1 mutation carriers.

Conclusions DICER1 mutations are associated with both familial MNG and MNGwith SLCT, independent of PPB. These germline DICER1 mutations are associated withdysregulation of miRNA expression patterns.JAMA. 2011;305(1):68-77 www.jama.com

68 JAMA, January 5, 2011—Vol 305, No. 1 ©2011 American Medical Association. All rights reserved.Corrected on January 19, 2011

Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 09/26/2020

in the generation of microRNAs(miRNAs), short, double-stranded, non-coding RNAs that modulate gene ex-pression at the posttranscriptionallevel.12,13 Relevant to this function are 2RNase III domains and a PAZ domain,which is a single-stranded RNA bind-ing domain that binds the 3� single-stranded overhang of the precursormiRNA and whose distance from theRNase III domains determines the lengthof mature miRNAs generated(FIGURE 1).14 We hypothesized thatDICER1 mutations could be present inkindred with multiple cases of MNG(with and without SLCTs). We subse-quently explored the phenotypic andmolecular consequences of the DICER1mutations observed.

METHODSWe used the following criteria to iden-tify cases and families: 3ormorecasesofMNG,2ormorecasesofMNGplus1caseof SLCT, or MNG and SLCT occurringin the same individual. Cases were iden-tified through PubMed searches andthrough patients referred to the McGillUniversity Cancer Genetics program.Members of 2 MNG/SLCT families de-scribedinpreviouspublications15,16 werecontacted through the authors (familiesA and B). Another case of MNG/SLCT(family C) was identified through a lo-cal clinician. Histopathological materialwas reviewed from the MNG cases (V.-H.N., H.R.H., and D.B.-D.S.), the SLCTcases (D.B.-D.S. and J.A.), and the rhab-domyosarcomacase(V.-H.N.).Diagnosesof MNG and SLCT were made by ex-amining tissue slides stained withhematoxylin-eosin.Therhabdomyosar-coma diagnosis was made as previouslydescribed.17 Families D and E were de-scribed previously as MON236 andMON152, respectively4,17,18;updatedde-tailsincludingnewlyaffectedfamilymem-bers are presented in TABLE 1.

ThestudywasapprovedbytheMcGillUniversity institutional review board.Eachfamilymemberprovidedwrittenin-formedconsent.Forminors,therisksandbenefitswereexplainedtotheparentsandthey signed on behalf of their children.A larger series of germline DNA from 80

anonymous cases of MNG and differen-tiated thyroid cancer from centers in theUnitedKingdom,Italy,andMontrealwerescreened forDICER1mutations.Anony-mous control DNA samples were usedfrom individuals with no history of can-cerwhoattendedtheJewishGeneralHos-pital in Montreal between June and Sep-tember 2009 and who provided writtenconsent. For the immunohistochemical(IHC) analysis, 28 anonymized nonhe-reditary MNG tissue samples from Hos-pitalDrA.Onativia,Salta,Argentina,wereused as controls.

Mutation Analysis

The 26 coding exons of DICER1 werescreened in 80 cases of differentiated thy-roidcancerandMNGandprobands from5 families by high-resolution meltinganalysis using the LightScanner instru-ment (Idaho Technologies, Salt LakeCity, Utah) or sequencing. Primer se-quences and the protocol for amplifica-tion were adapted from Hill et al9 withsome primers being changed to gener-ate smaller amplicons necessary for high-resolution melting (eTable 2 and eMeth-ods). Nontruncating mutations werefurther investigated by complementaryDNA (cDNA) analysis (eMethods).

RNA Analysis

Total RNA was extracted from indepen-dentculturesof lymphoblastoidcell lines

(LCLs)from9mutationcarrier individu-als(familyA:II-2; familyB:II-1, II-2; fam-ilyC:III-1; familyD:I-5, II-1, III-1; familyE: II-4, III-3), 5 noncarriers (3 unrelatedspouses from families D and E and 2 un-relatedcontrols)usingRNeasyKit (Qia-gen, Valencia, California). MessengerRNA(mRNA)wasretrotranscribedusingan oligo-dT and Superscript III reversetranscriptase(Invitrogen,Carlsbad,Cali-fornia).ExpressionofDICER1mRNAwasmeasuredbyquantitativereal-timepoly-merasechainreaction.PredesignedTaq-Man assays (Applied Biosystems, FosterCity,California)wereusedtospecificallyamplifycDNAderivedfrombothmutantandwild-typeDICER1andGAPDHmR-NAs. Inhibition of nonsense-mediatedmRNA decay (NMD) was performedusing cycloheximide as previously de-scribed.19 NMD is a regulatory processused by the cell to specifically recognizeanddestroymRNAtranscripts thatcarrypremature termination codons beforetheir translation into truncated andpotentially harmful proteins. This pro-cess is blocked by protein synthesis in-hibitors such as cycloheximide. Full ex-perimental details are provided in theeMethods.

miRNA Profiling

miRNA profiling of LCLs from 5 mu-tation carriers (family D: I-5, II-1, III-1;family E: II-4, III-3) and 5 noncarriers

Figure 1. Schematic Diagram of the Role of DICER1 in MicroRNA Processing

miRISCPre-miRNAMaturemiRNA

Complementarystrand degradation

DICER1

processing

N U C L E U S C Y T O P L A S M

PAZ domain

RNase III domains

55

3

3 overhang Regulationof protein expression

DICER1–pre-miRNA complex

miRNAduplex

DI

CD

IC

DDDE

R1

ER

1RRRR

DI

CD

IC

DDDDE

R1

ER

1R

1RRRRR

The PAZ domain of DICER1 plays an important role in microRNA (miRNA) processing by functioning as a mo-lecular ruler to ensure pre-miRNAs are cut to the correct length. Other proteins, such as PASHA and DROSHA,and cofactors involved in miRNA processing are not shown. miRISC indicates miRNA-induced silencing complex.

DICER1 MUTATIONS IN FAMILIAL MULTINODULAR GOITER

©2011 American Medical Association. All rights reserved. JAMA, January 5, 2011—Vol 305, No. 1 69Corrected on January 19, 2011


(3 unrelated spouses from families Dand E, 1 mutation-negative member offamily D, II-14, and 1 unrelated con-trol) was performed by the VancouverProstate Centre Microarray Facility.Total RNA including the small RNAfraction was isolated and quantified.The quality of the RNA was confirmedprior to fluorescent end-labeling and

hybridization to the Unrestricted Hu-man miRNA Microarrays Release 12.0(design ID 021827) using the AgilentmiRNA Microarray System with miRNAComplete Labeling and Hyb Kit ver-sion 2 (Agilent Technologies, SantaClara, California). Following hybrid-ization, the microarrays were scanned,quantified, and analyzed using Ag-

ilent system software components(eMethods). Data was normalized bysetting values below 0.05 to 0.05 andapplying per-chip normalization to a setof positive control genes that were se-lected with the criteria that raw datawere above 20 for all samples. Lists ofsignificantly differentially expressedgenes were determined by both a t test

Table 1. Details of Tested Individuals in This Study

(Family) Mutation Identifieda Individualb SexDisease Status(Age at Diagnosis, y)

MutationStatus LOH (Tissue)c IHC (Tissue)c

(A) c.871_874delAAAG I-1 M MNG (14) PresentI-2 F MNG (22) Presentd

II-1 F Unaffected AbsentII-2 F MNG (16), SLCT (18) Present No (SLCT)II-3 F Unaffected AbsentII-4 F MNG (20) Present

(B) c.2457C�Gr.2437_2457del21

I-1 F MNG (17) Present

II-1 M MNG (12) Present No stainingII-2 F MNG (9), SLCT (14) Present No (MNG, SLCT) No staining (MNG), intense staining

in Sertoli cells and weakstaining of Leydig cells (SLCT)

(C) c.5018_5021delTCAA I-1 F Unaffected AbsentII-1 F Unaffected PresentIII-1 F MNG (18), SLCT (32) Present No (SLCT) Intense staining in Sertoli cells and

weak staining of Leydig cells(SLCT)

(D) c.2516C�T I-1 M MNG (15) PresentI-2 M Unaffected AbsentI-3 F MNG (U) PresentI-4 F MNG (21) PresentI-5 F MNG (19) PresentII-1 F MNG (15) Present Cytoplasmic staining (MNG)II-2 F MNG (14) Present Cytoplasmic staining (MNG)II-3 M Unaffected AbsentII-4 F MNG (22) Present No (MNG)II-5 F Unaffected AbsentII-6 F Unaffected AbsentII-8 M Unaffected AbsentII-9 F MNG (35) PresentII-10 F Unaffected AbsentII-11 F Unaffected AbsentII-12 F MNG (19) PresentII-13 M MNG (17) PresentII-14 M Unaffected AbsentII-15 M Unaffected AbsentII-16 F MNG (U) PresentII-17 M MNG (19) PresentIII-1 M MNG (10) Present Cytoplasmic staining (MNG)III-2 M Unaffected AbsentIII-3 M MNG (17) PresentIII-4 F MNG (40) PresentIII-5 F MNG (23) Present Cytoplasmic staining (MNG)III-8 M MNG (14) PresentIII-9 M MNG (24) PresentIV-1 F MNG (21) PresentIV-2 F MNG (16) Present

(continued)




between the 2 conditions (mutation car-rier vs not a mutation carrier) with theP value set at .05 and fold-change fil-tering with a fold-change value of 2.0.The heatmap was created by Cluster 3.0and Java gene Treeview software.20,21

Frozen tissue samples (2 nodules of1 goiter from a DICER1 mutation car-rier) were homogenized in lysis bufferand total RNA was prepared. RNA ex-tracted from normal thyroid tissue (Ap-plied Biosystems/Ambion, Austin,Texas) was used as a control. One mi-crogram of total RNA was reverse tran-scribed and the cDNA was applied ona TaqMan low-density array, with dataacquisition via the Applied Biosys-tems 7900HT Fast Real-time polymer-ase chain reaction system. Data analy-sis was performed using the R analysissoftware (Applied Biosystems). Indi-vidual miRNA expression was vali-dated using TaqMan miRNA Assays(Applied Biosystems), and values werenormalized to the endogenous con-trol U6 snRNA (the noncoding smallnuclear RNA part of U6 small nuclearribosomal ribonucleoprotein). Foldchange was calculated using the DDCTmethod.22 Further experimental de-tails are provided in the eMethods.

LOH and IHC Analyses

DNA was extracted from lymphocytesfrom the probands and, where available,

from macro-dissected, formalin-fixed,paraffin-embedded tumorandgoiter tis-sue.Polymerasechainreactionandevalu-ationof lossofheterozygosity(LOH)wascarriedoutaspreviouslydescribed.23 Fur-therdetailsareprovided in theeMethods.

IHC analysis was performed on depa-raffinized5-µmtissuesections incubatedwith anti-DICER antibody ab14601 (1:50forgoitersandtherhabdomyosarcoma,1:100 for SLCTs) (Abcam, Cambridge,Massachusetts). A total of 28 tissuesamples of sporadic MNG, 9 samples ofDICER1-related MNG (family B: II-1, II-2; family D: II-1, II-2, III-1, III-5; familyE: II-1, II-4, III-3), 2 samples of SLCTs(familyB:II-2;familyC:III-1),and1rhab-domyosarcoma sample (family E: II-3)were analyzed. The sections were thenincubated with labeled polymer-HRPantimouse immunoglobulins (DakoEn-vision� system-HRP; Dako, Glostrup,Denmark) for 30 minutes. Staining wascompletedbya10-minuteincubationwith3,3�-diaminobenzidine plus substrate-chromogensolution(DAB�).Counter-stainwasperformedbydippingsectionsin a bath of hematoxylin for 5 seconds.

RESULTSDICER1 Mutation Analysis

WesearchedforDICER1mutations in53membersfromthe5studiedfamilieswithMNGand/orSLCTandidentified,intotal,37mutationcarriers in these families, all

but3ofthembeingaffectedbyMNGand/orSLCT.Representativeelectrophoreto-grams are shown in eFigure 1 and re-sults are summarized in Table 1. Amongthe 3 MNG/SLCT families, family A wasinitially described in 1981 as an MNG/SLCT family where the proband (fam-ily A: II-2) had MNG at age 16 years andSLCT at age 18 years.16 The proband and3 other family members carried ac.871_874delAAAG mutation, which re-sults in an mRNA product that con-tains a premature termination codon atposition p.291. mRNA carrying the mu-tation cannot be detected because of theaction of NMD (eFigure 2A). Family Bwas a recently reported MNG/SLCT fam-ily15 in whom we identified a c.2457C�Gmutation, which creates a de novo splicesite with a predicted strength of 0.37 (ona scale ranging from 0 to 1 indicating theprobability for a given site to be a splicesite, where 1 is the score for a perfectsplice site). The authentic site has a pre-dicted strength of 0.50.24 Sequencing ofamplification products from the exon15-16 junction reveals that 100% of themutant allele produces a mutant tran-script containing an in-frame deletion ofthe first 21 base pairs of exon 16(r.2437_2457del21) and therefore is notsubject to NMD (eFigure 2B). This newtranscript generates a predicted DICER1protein with a p.Ile813_Tyr819del mu-tation resulting in an altered PAZ struc-

Table 1. Details of Tested Individuals in This Study (continued)

(Family) Mutation Identifieda Individualb SexDisease Status(Age at Diagnosis, y)

MutationStatus LOH (Tissue)c IHC (Tissue)c

(E) c.2805-1G�Tr.2805_2987del183

I-1 F MNG (21) Present

I-2 M MNG (15) Present No (MNG)II-1 F MNG (17) Present No staining (MNG)II-2 M Unaffected PresentII-3 F MNG (13), RMS (20)e Present No (RMS) Cytoplasmic staining (RMS)II-4 M MNG (14) Present Cytoplasmic staining (MNG)III-1 F Unaffected AbsentIII-2 F Unaffected PresentIII-3 M MNG (17) Present Cytoplasmic staining (MNG)III-4 F Unaffected AbsentIII-5 F Unaffected Absent

Abbreviations: F, female; IHC, immunohistochemical analysis; LOH, loss of heterozygosity; M, male; MNG, multinodular goiter; RMS, rhabdomyosarcoma; SLCT, Sertoli-Leydig cell tu-mor of the ovary; U, exact age at diagnosis not known.

ac. is a DNA coding sequence–based nomenclature; numbering commences from the start codon. r. is RNA-based, indicating the effect of the mutation on the messenger RNA.bRoman numerals refer to generations; Arabic numerals refer to individuals within a generation (ie, siblings and cousins).cNo indicates LOH was not present. Absence of text indicates that the tissue was not analyzed either for LOH or for IHC.d Individual was untested but is an obligate mutation carrier.eWe were unable to unequivocally distinguish between the alveolar and embryonal subtypes of RMS.




ture (FIGURE 2; interactive feature avail-able at http://www.jama.com). Thismutation was detected in 3 individu-als, including the proband (family B: II-2), who developed an SLCT at age 14years. In family C, the proband (fam-ily C: III-1) had MNG at age 18 yearsand SLCT at age 32 years, with no fam-ily history of cancer or MNG. She andher unaffected mother carried ac.5018_5021delTCAA mutation, whichresults in an mRNA product that is sub-ject to NMD because it contains a pre-mature termination codon at positionp.1672 (eFigure 2A).

DICER1 mutations were also found inboth MNG families previously linked tochromosome 14q (families D and E), butnot in germline DNA from 71 individu-als with differentiated thyroid cancer, 31of whom also had MNG, and a further9 individuals with MNG, all of whomhad a family history of MNG and dif-

ferentiated thyroid cancer (TABLE 2). Infamily D, we identified a missenseDICER1 variant, c.2516C�T, pre-dicted to cause a p.Ser839Phe change in20 family members, all of whom hadMNG; there were no cases of MNG inany of the 10 family members withoutthe variant. Taken together, these ob-servations indicate that p.Ser839Phe isvery highly penetrant for MNG. Thevariant was not present in DNA from 455anonymous controls with no history ofcancer from the Jewish General Hospi-tal in Montreal. On the basis of a preva-lence of MNG in the population of 4%,1

and assuming a mutation frequency of0.005, the probability of segregation ofthis mutation with MNG in family Dpurely by chance is approximately 10−5.

Serine 839 is conserved in DICER1proteins of higher vertebrates (eFig-ure 3), and predictive software was usedto help establish the overall effect of the

p.Ser839Phe mutation present in fam-ily D. According to the probabilisticprogram SIFT (Sorting Intolerant FromTolerant, http://sift.jvci.org/), whichpredicts effects of a substitution basedon sequences of similar peptides,p.Ser839Phe was assigned a score of0.05. This is the limit of normal be-cause the SIFT scale designates scoresof less than 0.05 as deleterious whilescores of 0.05 or greater are predictedto be tolerated.25 Similarly, Poly-Phen-2 (http://genetics.bwh.harvard.edu/pph2/) predicted this mutation was“possibly damaging” (score, 0.442),26

because the score was between 0.16 and0.85. The Ser to Phe change was pre-dicted to disrupt an alpha helix in thePAZ domain (Figure 2).

The second MNG kindred (family E)carried a fully segregating splice-sitemutation, c.2805-1G�T, resulting in anin- f rame de le t ion of exon 18,r.2805_2987del183 (eFigure 2C), al-tering the structure of DICER1 by elimi-nating part of the PAZ domain (eFig-ure 4). This mutation was not seen ingermline DNA from 430 controls.

RNA and Protein Analysis

We analyzed mRNA and protein ex-tracted from LCLs of 7 carriers fromfamilies B, D, and E (with nontruncat-ing mutations), 2 carriers from familiesA and C (with NMD-sensitive muta-

Figure 2. Predicted Tertiary Structures of Human DICER1 Protein

PAZ domain

A Wild-type DICER1 B Family B C Family D

Partial loss ofalpha helix

Absent alpha helixAlpha helix Absent alpha helix

Pre-miRNAbinding site

PAZ domain PAZ domain PAZ domainDICER1

Alpha helix

The predicted structure of wild-type DICER1 (A), mutant protein with a p.Ile813_Tyr819del (B), and mutant protein carrying a p.Ser839Phe change (C). The models weregenerated by using Phyre (Protein homology/analogy recognition engine; http://www.sbg.bio.ic.ac.uk/phyre). The key differences are an absent alpha helix in the versionsfrom families B and D. Interactive feature available at http://www.jama.com.

Table 2. Additional MNG and DTC Cases Screened for DICER1 Mutationsa

Center LocationNo. ofCases Diagnoses

Median Age(Range), y Family History

United Kingdom 44 40 DTC only,4 DTC and MNG

43 (4-60) 17 cases had �1 relative withDTC or MNG

Italy 22 9 MNG only,13 DTC and MNG

45 (23-66) All had family history of MNGand DTC; 1-22 cases ofMNG per family, 1-4cases of DTC per family

Montreal, Canada 14 14 DTC and MNG 50 (23-71) 2 cases had �1 relative withDTC or MNG

Abbreviations: DTC, differentiated thyroid cancer; MNG, multinodular goiter.aAll cases were screened for DICER1 mutations by high-resolution melting.




tions), and 5 mutation-negative con-trols (3 from families D and E and 2 non-familial controls). mRNA quantificationshowed highly variable levels of mRNAbetween and within families; no consis-tent relationship between mRNA levelsand either the presence of goiter orSLCTs was observed (FIGURE 3).

LOH and IHC Studies

StudiesofLOHusingDNAextractedfromgoiters fromfamiliesB,D,andE,togetherwithconstitutionalDNA,revealednoevi-dence of LOH (Table 1, FIGURE 4). IHCanalysis of these goiters using anti-DICER1 antibody ab14601 revealed amixed picture, with no staining of 2 goi-tersfromfamilyB(II-1andII-2),butclearcytoplasmic staining of several goitersfrom families D and E (individual III-1fromfamilyD is shown inFIGURE 5).Tocompare these findings with that seeninnonfamilialgoiter,weimmunostainedtissuesamplesof28sporadicMNGfromHospitalDrA.Onativia,Salta,Argentina.Twenty-five of these MNG samplesstained with the anti-DICER1 antibody(Figure5)whereas3showednostaining.Therefore, the amount of DICER1 pro-tein did not appear to be associated withDICER1mutationstatus, similar towhatwe observed in LCL RNA and protein.

SLCT tissue from the 3 DICER1 mu-tation carrier probands in families Athrough C was analyzed for LOH, and ineach case there was no evidence of lossof thewild-typeallele (Table1,Figure4,and eFigure 5A). IHC analysis of the 2available SLCTs from DICER1 muta-tion carriers showed intense expressionof DICER1 in Sertoli cells, but the stain-ing was much weaker in Leydig cells(Table 1, Figure 5, and eFigure 5B). Bycontrast, neither ovarian carcinomas(n=5, not shown) nor normal ovary(Figure 5) stained with the ab14601 anti-DICER1 antibody.

One DICER1 mutation carrier (fam-ily E: II-3) died at age 20 years from amalignant paravertebral tumor charac-terized at the time as an alveolar rhab-domyosarcoma (molecular reconfirma-tion of this diagnosis was impossible dueto lack of tissue). Alveolar rhabdomyo-sarcoma has not been observed in

DICER1-related PPB families, althoughunusual sarcomas are characteristic. NoLOH was seen in this tumor (eFigure 5A)and IHC analysis revealed diffuse cyto-plasmic staining (eFigure 5B).

miRNA Assays

We used RNA extracted from LCLs es-tablished from 5 carriers and 4 mutation-

negative controls, all from families D andE (plus 1 unrelated control) for themiRNA assays. Global miRNA profil-ing of a panel of 851 human miRNAsidentified 94 miRNAs (11%) that weresignificantly differentially expressed inthe 5 affected DICER1 mutation carri-ers compared with the 5 unaffected non-carriers (fold change �2 and P� .05)

Figure 3. DICER1 Messenger RNA Levels in Mutation Carriers and Noncarriers Relative to GAPDH

1.0

0.5

0.1III-1 III-1 II-2 II-2III-3II-1 II-1II-4Con1 Con2 Con4 Con5Con3I-5

Family

DIC

ER1

Exp

ress

ion

Rel

ativ

e to

GA

PD

H

D E C AB

Control (noncarrier)MNGMNG and SLCT

DICER1 expression was measured in lymphoblastoid cell lines derived from carriers and noncarriers from fami-lies A through E and from 2 unrelated noncarriers by real-time polymerase chain reaction. There are no sig-nificant differences in the level of DICER1 expression when comparing mutation carriers and noncarriers (P=.55)and when comparing carriers of truncating mutations (families A and C) and carriers of nontruncating muta-tions (families B, D, and E) (P=.49). The identity of each individual carrier is indicated on the x-axis; Con1through Con5 are noncarrier controls. Error bars indicate 95% confidence intervals.

Figure 4. Loss-of-Heterozygosity Analysis of Sertoli-Leydig Cell Tumor and MultinodularGoiter in an Affected Proband

A gDNA from lymphocytes

G T A N A C A

G T A N A C A

C gDNA from lymphocytes

B gDNA from goiter

G T A N A C A

G T A N A C A

D gDNA from SLCT

There is no loss of heterozygosity in Sertoli-Leydig cell tumor (SLCT) and multinodular goiter tissue from theaffected proband (individual II-2, family B). A and B were sequenced simultaneously, and C and D were se-quenced simultaneously at a different time. gDNA indicates genomic DNA; arrowhead indicates mutation. Nucleo-tides labeled with N indicate sites where dual peaks are present.




(FIGURE 6). We then compared miRNAprofiles in the RNA extracted from 1fresh-frozen DICER1-related MNG withnormal thyroid tissue to determinemiRNAs differentially expressed in theMNG and the normal tissue. Compar-ing the 2 lists, only 5 miRNAs derivedfrom the same precursors (miR-345, let-7a, miR-99b, miR-133, miR-194) weredecreased in RNA from both LCLs andthe DICER1-related MNG. Of these,only miR-345 is highly expresseduniquely in the normal thyroid.27 Wefocused on let-7a and miR-345, andmarkedly lower levels of both miRNAswere seen in the DICER1-related goi-

ter, when compared with both normalthyroid gland tissue and a follicular thy-roid carcinoma (FIGURE 7).

COMMENTHere we report germline DICER1 muta-tions in familial MNG and MNG/SLCTfamilies. In the 2 larger families withMNG, the DICER1 mutations werehighly penetrant for goiter. MNG isprevalent in the general population andalso occurs within the wide spectrum ofconditions occurring in kindred withPPB.10,28 Germline DICER1 mutations arefound in more than half of all childrenwith PPB and most of the mutations

identified are predicted to result in trun-cated proteins.9,29 The penetrance of thereported DICER1 mutations for PPB (andthe other major clinical characteristics ofthe syndrome) is low, with many genecarriers remaining unaffected into adult-hood. By contrast, the 3 nontruncatingmutations reportedhere (eFigure1)werehighlypenetrant forMNG.Moreover, theabsence of all other known features of thePPB-FTDS in the large family D kin-dred with 4 generations of affected in-dividuals suggests that the functionaleffect of the p.Ser839Phe mutation isqualitatively different from that associ-ated with truncating mutations. Simi-larly, the mutation in family E resultedin a DICER1 protein that lacks most ofthe PAZ domain (eFigure 4) but was oth-erwise normal. In family B, in which 3individuals developed early-onset MNG,the PAZ domain was significantly al-tered by the in-frame deletion of 7 aminoacids.

Two of the 3 DICER1 mutations seenin the MNG/SLCT families were trun-cating and distributed throughout thegene with no clear genotype-pheno-typecorrelationdifferentiating themfromthose PPB-FTDS families with reportedDICER1 mutations9 (eFigure 6). The me-dian diagnosis age of 3 DICER1-relatedSLCTs in this study and 6 reported PPB-associated SLCTs11 was 13 years, con-siderably younger than median age at on-set of 19 years in sporadic SLCTs7

(P=.009, Mann-Whitney test).The disease spectrum associated with

PPB is broad,10 and we have confirmedby molecular means that SLCT belongsin the PPB-FTDS but can occur withoutPPB. It is likely that most of the re-ported MNG/SLCT cases (eTable 1) har-bor DICER1 mutations and that our find-ings explain the observation first madeby Jensen et al.30 Familial SLCT caseshave also been reported, but whether thisresults from DICER1 mutations is not an-swered by our study because the fami-lies described here had only 1 case ofSLCT per kindred. We did screen an af-fected familial SLCT proband31 but didnot identify a DICER1 mutation; it re-mains possible that DICER1 mutationswill be implicated in other familial SLCT,

Figure 5. Immunohistochemistry of Multinodular Goiter and Sertoli-Leydig Cell Tumors

A MNG (sporadic case)

C Normal ovary D SLCT (III-2, family C)

B MNG (III-1, family D)

Representative staining of DICER1 (A) in multinodular goiter (MNG) from the sporadic cases (Hospital Dr A. Ona-tivia, Salta, Argentina) and (B) from a mutation carrier (individual III-1, family D). Images are at 20� magnifica-tion. Tissue sections were incubated with anti-DICER antibody (1:50), labeled with polymer-HRP antimouse, stainedwith DAB�, and counterstained with hematoxylin. DICER1 staining (reddish brown) is detected in cells borderingthe vesicles and is heterogeneous, being partly dependent on the amount of cytoplasm. No difference in stainingis evident when comparing hereditary and sporadic cases of MNG. Representative staining of DICER1 in normalovary (C) compared with the Sertoli-Leydig cell tumors (SLCTs) (D) from individual III-2, family C (see also eFigure5 for the DICER1 immunohistochemical study of the SLCTs from individual II-2, family B). The tumor is composedof prominent cords and tubules of immature Sertoli cells with discrete clusters of Leydig cells. The Leydig cellscontained a round nucleus with prominent nucleolus and an abundant pale and eosinophilic cytoplasm. The im-mature Sertoli cells have small round nuclei and scanty cytoplasm. Cytoplasm of Sertoli cells was strongly stainedand Leydig cells show very weak staining. No staining was detected in normal ovary.




or that there are other genes respon-sible for familial SLCT with genes in themiRNA processing pathway, such asPASHA/DGCR8 or DROSHA, being rea-sonable candidates.

We also analyzed DICER1 in germ-line DNA from probands from large se-ries of differentiated thyroid cancercases with and without MNG, and nopotentially disease-causing variants

were identified. Although in general dif-ferentiated thyroid cancer cases ap-pear not to be associated with DICER1mutations, differentiated thyroid can-cer cases in the context of SLCT or PPBmay harbor DICER1 mutations.

None of the DICER1-related SLCTsshowed evidence of LOH, and eventhough the numbers are small, thesefindings are consistent with the no-

tion from animal models that DICER1does not function as a classic tumorsuppressor gene but that instead tu-mors develop as a result of miRNA dys-regulation through a possible haploin-sufficiency effect.3 2 The lack ofcorrelation, however, between DICER1mutation status and both mRNA andprotein levels of DICER1 we observedhere (Figure 3) suggests that mecha-

Figure 6. miRNA Microarray Study in DICER1 Mutation Carriers and Noncarriers

miR-1224-5pmiR-1236miR-1229miR-196bmiR-634miR-127-3pmiR-345miR-99b*miR-211miR-500miR-1229miR-486-5pmiR-498miR-1228*miR-30c-2*miR-539miR-133amiR-130a*miR-610miR-654-3pmiR-885-3pmiR-520emiR-623miR-1322miR-1323miR-1300_v13.0miR-125b-2*miR-1300_v13.0miR-498miR-877miR-1236miR-1909*miR-373miR-1224-3pmiR-432miR-658miR-139-3pmiR-1179miR-187*miR-936miR-194*miR-1264miR-665miR-584miR-675*miR-1261miR-610miR-662miR-492

let-7a*

DICER1mutation –

DICER1mutation +

miR-101miR-20a*

miR-20a*miR-17

miR-142-5pmiR-142-5p

miR-17*miR-17*miR-32miR-20bmiR-21*miR-21*miR-19amiR-19amiR-629miR-424miR-424miR-155*miR-155*miR-155*miR-22miR-15bmiR-15bmiR-374amiR-374amiR-590-5pmiR-15b*miR-185miR-18amiR-7-1*miR-142-3pmiR-142-3pmiR-7miR-7miR-1260miR-16-2*miR-186miR-186miR-629miR-32miR-590-5pmiR-625miR-625miR-223miR-223miR-27b

miR-101

DICER1mutation –

DICER1mutation +

miR-133amiR-187*miR-1226*miR-1299miR-125b-2*miR-149*miR-422amiR-34c-3pmiR-422amiR-1299miR-610miR-422amiR-877miR-422amiR-513a-3pmiR-662miR-187*miR-760miR-373*miR-576-5pmiR-623miR-623miR-1207-3pmiR-382miR-610miR-584miR-486-5pmiR-1183miR-557miR-516a-5pmiR-99b*miR-1226*miR-877miR-629*miR-514miR-875-5pmiR-520bmiR-520emiR-202miR-564miR-654-5pmiR-760miR-508-5pmiR-629*

DICER1mutation +

miRNA log2 expression -2 -1 0 1 2

DICER1mutation –

Microarray study showing differentially expressed microRNAs (miRNAs) between carriers and noncarriers of mutations in DICER1 from families D and E. From left toright, carriers of mutation are family E, individuals III-3 and II-4, and family D, individuals I-5, II-1, and III-1. Noncarriers are family D, individual II-15, and 3 unrelated(married-in) controls from families D and E. Unsupervised hierarchical clustering showed clustering of the samples according their mutation status. Cluster 3.0 softwarewas used, wherein miRNA expression values were median-centered, normalized, and clustered using Pearson uncentered correlation distance and average linkage. Theheatmap was divided in 3 consecutive panels from left to right. In those cases where a miRNA hairpin precursor gives rise to 2 miRNAs, the less predominant miRNAis indicated by an asterisk. The color scale indicates the magnitude of miRNA expression in log2 with green indicating low expression and red high expression.




nism of tumorigenesis in humanDICER1 mutation carriers may be com-plex and may differ between tissues.

DICER1 has several highly con-served domains, including the PAZ do-main (Figure 1, Figure 2, and eFigure4), which appears to be critical forDICER1 function: a purified C-terminal fragment of DICER1 contain-ing both RNase domains and double-strand RNA binding domain showed nodouble-strand RNA cleavage activ-ity.33 PAZ acts as a molecular ruler, de-termining where the RNase domains ofDICER1 cut the pre-miRNAs to theirfinal size.34 We observed that selectivedisruption of the PAZ domain, in a set-ting of an otherwise normal DICER1protein, is associated with familialMNG. It is possible that these muta-tions do not represent hypomorphicDICER1 alleles, but instead the goiterphenotype is the result of specific PAZdisruption rather than a true haploin-sufficiency. This might indicate thepresence of a close relationship be-tween PAZ domain function and theregulation of thyroid development. Ina similar vein, a single missense muta-tion in the thyroid transcription factorTTF1 results only in thyroid disease,35

whereas loss of an entire allele, in bothhumans and in animal models, resultsin much more serious phenotypes.3

miRNAs have a crucial role in hu-man development12,36,37 and recent datahave shown that germline DICER1 mu-tations are associated with early-onsetmalignancy.9 Perturbations of miRNAsin cancer are common,38 but constitu-tive defects in miRNAs have not previ-ously been reported in humans. In lightof the central role of DICER1 in miRNAprocessing, we looked for downstreamevidence of miRNA dysregulation in tis-sues fromheterozygotes in familiesDandE with nontruncating mutations(Figure 6), and found that 5 miRNAswere consistently decreased. Of these,only let-7a and miR-345 were also de-creased in the goiter tissue of 1 carrierof the c.2805-1G�T mutation in fam-ily E (Figure 7). let-7a is down-regulated in breast, pancreas, and lungcancer and malignant melanoma39 anddisruptions in the lin-28–let-7 pathwayalter glucose metabolism and insulin sen-sitivity in mice,40 but to date let-7a hasnot been implicated in thyroid disease.miR-345 is highly expressed in the thy-roid gland,27 making this an attractivecandidate to further explore MNG patho-genesis in DICER1 mutation carriers.

Our study has the limitation of beingrelatively small and focused on familialMNG with and without SLCT, so the fullrange of the effect of germline DICER1mutations remains uncertain. Studies of

unselected cases of goiter, SLCT, andother hyperplastic and neoplastic con-ditions will be required to resolve thisquestion. In addition, this study is es-sentially descriptive, and as such, we donot present a mechanistic explanationfor how perturbations of DICER1 func-tions lead to thyroid goiter or SLCT.

In summary, we report DICER1 mu-tations to be associated with both au-tosomal-dominant susceptibility toMNG and SLCT occurring with MNG.This latter association was first ob-served more than 30 years ago by Jensenand colleagues.30 Our study confirmsclinical observations10,11 and defini-tively extends the tumor spectrum ofDICER1 mutation beyond PPB, cysticnephroma, embryonal rhabdomyosar-coma, and lung cysts.9,41 Further, mu-tations in other genes in the miRNAprocessing pathway may explain someof these syndromic disease combina-tions. Unlike SLCT, MNG is a very com-mon condition worldwide2 and deter-mining the role of dysregulated miRNAprocessing in the development of spo-radic MNG could be an important av-enue for future research.

Author Affiliations: Program in Cancer Genetics, De-partments of Oncology and Human Genetics (Drs RioFrio, Foulkes, and Tischkowitz; Mr Bahubeshi; and MssHamel, Sabbaghian, and Pouchet), and Departmentsof Obstetrics and Gynecology (Dr Gilbert) and Pathol-ogy (Drs Arseneau and Nguyen), Department of Medi-cal Genetics and Research Institute, McGill UniversityHealth Centre (Drs Rio Frio and Foulkes and Ms Hamel);Segal Cancer Centre, Lady Davis Institute, Jewish Gen-eral Hospital (Mr Bahubeshi, Mss Sabbaghian andPouchet, and Drs Foulkes and Tischkowitz), McGill Uni-versity, Montreal, Quebec, Canada; Department of Can-cer Biology, Dana-Farber Cancer Institute, Boston, Mas-sachusetts (Drs Kanellopoulou and Livingston);Department of Pediatric Endocrinology and Rheuma-tology, Poznan University of Medical Sciences, Poznan,Poland (Dr Niedziela); Department of Pathology, Eto-bicoke General Hospital, Toronto, Ontario, Canada (DrO’Brien); Hereditary Breast Health Clinic, Health Sci-ences Centre Winnipeg, Winnipeg, Manitoba, Canada(Ms Serfas); Section of Cancer Genetics, Institute of Can-cer Research, Sutton, Surrey, United Kingdom (Drs Bro-derick and Houlston); Genetic Cancer SusceptibilityGroup, International Agency for Research in Cancer,Lyon, France (Dr Lesueur); Medical Genetics Unit, De-partment of Gynaecology, S.Orsola-Malpighi Hospi-tal, University of Bologna, Bologna, Italy (Dr Bonora);Laboratory of Immunology, National Institute of Al-lergy and Infectious Diseases, National Institutes ofHealth, Bethesda, Maryland (Dr Muljo); Division of En-docrinology, Metabolism and Genetics, Internal Medi-cine Department, University of Kansas Medical Cen-ter, Kansas City (Dr Schimke); Department of Pathology,CHU Sainte-Justine, Montreal (Dr Bouron-Dal Soglio);Children’s Hospital and Clinics of Minnesota, St Paul(Dr Schultz); The International Pleuropulmonary Blas-

Figure 7. Quantification of Expression Levels of MicroRNAs let-7a and miR-345 byReal-Time Polymerase Chain Reaction in Frozen Tissues

1.5

1.0

0.5

0.1NormalThyroid

FTCGoiter

let-

7a E

xpre

ssio

nR

elat

ive

to U

6 sn

RN

A

1.0

0.5

0.05

0.01

0.1

NormalThyroid

FTCGoiter

miR

-345

Exp

ress

ion

Rel

ativ

e to

U6

snR

NA

let-7a miR-345

miR-345 and let-7a are down-regulated in goiter tissue from individual III-3 from family E compared withnormal thyroid and follicular thyroid cancer (FTC) tissues from noncarriers of DICER1 mutations. Error barsindicate 95% confidence intervals; U6 snRNA indicates the noncoding small nuclear RNA part of U6 smallnuclear ribosomal ribonucleoprotein.




toma Registry, St Paul (Dr Priest); Sector Patologıa, Hos-pital Dr A. Onativia, Salta, Argentina (Dr Harach). DrKanellopoulou is currently with the Laboratory of Im-munology, National Institute of Allergy and InfectiousDiseases, National Institutes of Health, Bethesda.Author Contributions: Dr Foulkes had full access to allof the data in the study and takes responsibility forthe integrity of the data and the accuracy of the dataanalysis.Study concept and design: Rio Frio, Livingston, Foulkes,Tischkowitz.Acquisition of data: Rio Frio, Bahubeshi, Kanellopoulou,Hamel, Niedziela, Sabbaghian, Pouchet, Gilbert,O’Brien, Serfas, Houlston, Lesueur, Bonora, Muljo,Schimke, Arseneau, Nguyen, Harach, Foulkes,Tischkowitz.Analysis and interpretation of data: Rio Frio, Bahubeshi,Kanellopoulou, Hamel, Niedziela, Sabbaghian, O’Brien,Broderick, Bouron-Dal Soglio, Arseneau, Schultz, Priest,Nguyen, Harach, Livingston, Foulkes, Tischkowitz.Drafting of the manuscript: Rio Frio, Bahubeshi, Priest,Foulkes, Tischkowitz.Critical revision of the manuscript for important in-tellectual content: Rio Frio, Bahubeshi, Kanellopoulou,Hamel, Niedziela, Sabbaghian, Pouchet, O’Brien,Serfas, Broderick, Houlston, Lesueur, Muljo, Schimke,Bouron-Dal Soglio, Arseneau, Schultz, Nguyen,Livingston, Foulkes, Tischkowitz.Statistical analysis: Rio Frio.Obtained funding: Foulkes, Tischkowitz.Administrative, technical, or material support: Rio Frio,Bahubeshi, Hamel, Niedziela, Sabbaghian, Pouchet,O’Brien, Serfas, Broderick, Houlston, Lesueur, Bonora,Muljo, Schimke, Bouron-Dal Soglio, Arseneau, Schultz,Priest, Nguyen, Harach, Foulkes, Tischkowitz.Study supervision: Livingston, Foulkes, Tischkowitz.Conflict of Interest Disclosures: All authors have com-pleted and submitted the ICMJE Form for Disclosure ofPotential Conflicts of Interest. Dr Priest reported beingnamed in a US patent US2009/068691 regarding cer-tain technical procedures associated with DICER1 test-ing. Dr Priest has no financial interest and all rights ofDr Priest have been signed over to the Children’s Hos-pitals and Clinics of Minnesota. No other authors re-ported disclosures.Funding/Support: The work was funded by the Jew-ish General Hospital Weekend to End Women’s Can-cers, the Turner Family Cancer Research Fund, and lesFonds de la Recherche en Sante du Quebec (FRSQ).Mr Bahubeshi is funded by the Canadian Institutes ofHealth Research/FRSQ training grant in cancer re-search (FRN53888) of the McGill Integrated CancerResearch Training Program. Dr Rio Frio was fundedby the Research Institute of the McGill University HealthCentre and by the Henry R. Shibata Fellowship of theCedars Cancer Institute. Dr Priest receives funding sup-port from the Pine Tree Apple Tennis Classic and theTheodora H. Lang Charitable trust. Dr Foulkes holdsan FRSQ national scientist award and Dr Tischkowitzholds an FRSQ clinician-scientist award.Role of the Sponsor: The funding agencies had no rolein the design and conduct of the study; in the collec-tion, analysis, and interpretation of the data; or in thepreparation, review, or approval of the manuscript.Disclaimer: The contents of this article are the sole re-sponsibility of the authors.Online-Only Material: The eMethods, eTables 1 and2, eFigures 1 through 6, and an interactive feature areavailable at http://www.jama.com.Additional Contributions: We would like to thank thepatients and families who participated in this study; DebraCollins, MSc, Division of Endocrinology, Metabolism andGenetics, University of Kansas Medical Center, who as-sisted in contacting one of the families; Archana Srivas-tava, MSc, Segal Cancer Center, Jewish General Hos-pital, for her technical support; and Paul Lee, Departmentof Pathology, McGill University, for his help with ac-

crual of pathology material. Gretchen M. Williams, BS,CCRP, International Pleuropulmonary Registry, Chil-dren’s Hospitals and Clinics of Minnesota, assisted withdata collection and analysis. We thank Robert H. Young,MD, Department of Pathology, Massachusetts Gen-eral Hospital, for providing information on his previ-ously published series of sporadic SLCT cases; Anne Hae-gert, BSc, Laboratory for Advanced Genome Analysis,Vancouver Prostate Centre, Vancouver, Canada, for herhelp with the miRNA expression analysis; and Rami Zahr,BS, National Institute of Allergy and Infectious Dis-eases, National Institutes of Health, and Graham Big-nell, PhD, Wellcome Trust Sanger Institute, for techni-cal support. None of the individuals named here receivedcompensation for the contributions besides salary.

REFERENCES

1. Pinchera A, Aghini-Lombardi F, Antonangeli L, VittiP. Multinodular goiter: epidemiology and prevention[in Italian]. Ann Ital Chir. 1996;67(3):317-325.2. Vanderpump MP, Tunbridge WM, French JM, et al.The incidence of thyroid disorders in the community.Clin Endocrinol (Oxf ). 1995;43(1):55-68.3. Knobel M, Medeiros-Neto G. An outline of inher-ited disorders of the thyroid hormone generatingsystem. Thyroid. 2003;13(8):771-801.4. Bignell GR, Canzian F, Shayeghi M, et al. Familialnontoxic multinodular thyroid goiter locus maps tochromosome 14q but does not account for familial non-medullary thyroid cancer. Am J Hum Genet. 1997;61(5):1123-1130.5. Capon F, Tacconelli A, Giardina E, et al. Mapping adominant form of multinodular goiter to chromosomeXp22. Am J Hum Genet. 2000;67(4):1004-1007.6. Koonings PP, Campbell K, Mishell DR Jr, Grimes DA.Relative frequency of primary ovarian neoplasms. Ob-stet Gynecol. 1989;74(6):921-926.7. Young RH, Scully RE. Ovarian Sertoli-Leydig celltumors. Am J Surg Pathol. 1985;9(8):543-569.8. Young RH. Sex cord-stromal tumors of the ovary andtestis. Mod Pathol. 2005;18(suppl 2):S81-S98.9. Hill DA, Ivanovich J, Priest JR, et al. DICER1 muta-tions in familial pleuropulmonary blastoma. Science.2009;325(5943):965.10. Priest JR, Williams GM, Hill DA, Dehner LP, JaffeA. Pulmonary cysts in early childhood and the risk ofmalignancy. Pediatr Pulmonol. 2009;44(1):14-30.11. Schultz K, Dehner L, Williams G, Hill A, Priest J.Ovarian tumors in association with pleuropulmonaryblastoma [poster No. 305]. Poster presented at: Ameri-can Society of Pediatric Oncology/Hematology Meet-ing; April 7-10, 2010; Montreal, Quebec, Canada.12. Bernstein E, Kim SY, Carmell MA, et al. Dicer isessential for mouse development. Nat Genet. 2003;35(3):215-217.13. Ryan BM, Robles AI, Harris CC. Genetic varia-tion in microRNA networks. Nat Rev Cancer. 2010;10(6):389-402.14. Macrae IJ, Zhou K, Li F, et al. Structural basis fordouble-stranded RNA processing by Dicer. Science.2006;311(5758):195-198.15. Niedziela M. Virilizing ovarian tumor in a 14-year-old female with a prior familial multinodular goiter.Pediatr Blood Cancer. 2008;51(4):543-545.16. O’Brien PK, Wilansky DL. Familial thyroid nodu-lation and arrhenoblastoma. Am J Clin Pathol. 1981;75(4):578-581.17. Druker HA, Kasprzak L, Begin LR, Jothy S, NarodSA, Foulkes WD. Family with Graves disease, multi-nodular goiter, nonmedullary thyroid carcinoma, andalveolar rhabdomyosarcoma. Am J Med Genet. 1997;72(1):30-33.18. Couch RM, Hughes IA, DeSa DJ, Schiffrin A, GuydaH, Winter JS. An autosomal dominant form of ado-lescent multinodular goiter. Am J Hum Genet. 1986;39(6):811-816.

19. RioFrioT,WadeNM,RansijnA,BersonEL,BeckmannJS, Rivolta C. Premature termination codons in PRPF31cause retinitis pigmentosa via haploinsufficiency due tononsense-mediated mRNA decay. J Clin Invest. 2008;118(4):1519-1531.20. de Hoon MJ, Imoto S, Nolan J, Miyano S. Opensource clustering software. Bioinformatics. 2004;20(9):1453-1454.21. Saldanha AJ. Java Treeview. Bioinformatics. 2004;20(17):3246-3248.22. Livak KJ, Schmittgen TD. Analysis of relative geneexpression data using real-time quantitative PCR andthe 2(-Delta Delta C(T)) method. Methods. 2001;25(4):402-408.23. Tischkowitz M, Xia B, Sabbaghian N, et al. Analy-sis of PALB2/FANCN-associated breast cancer families.Proc Natl Acad Sci U S A. 2007;104(16):6788-6793.24. Reese MG, Eeckman FH, Kulp D, Haussler D. Im-proved splice site detection in Genie. J Comput Biol.1997;4(3):311-323.25. Kumar P, Henikoff S, Ng PC. Predicting the effectsof coding non-synonymous variants on protein func-tion using the SIFT algorithm. Nat Protoc. 2009;4(7):1073-1081.26. Adzhubei IA, Schmidt S, Peshkin L, et al. A methodand server for predicting damaging missense mutations.Nat Methods. 2010;7(4):248-249.27. Hsu SD, Chu CH, Tsou AP, et al. miRNAMap 2.0.Nucleic Acids Res. 2008;36(Database issue):D165-D169.28. Priest JR, Watterson J, Strong L, et al. Pleuro-pulmonary blastoma. J Pediatr. 1996;128(2):220-224.29. Hill DA, Wang JD, Schoettler P, et al. GermlineDICER1 mutations are common in both hereditary andpresumed sporadic pleuropulmonary blastoma[abstract]. Lab Invest. 2010;90:311.30. Jensen RD, Norris HJ, Fraumeni JF Jr. Familial ar-rhenoblastoma and thyroid adenoma. Cancer. 1974;33(1):218-223.31. Whitcomb RW, Calkins JW, Lukert BP, Kyner JL,Schimke RN. Androblastomas and thyroid disease inpostmenopausal sisters. Obstet Gynecol. 1986;67(3)(suppl):89S-91S.32. Kumar MS, Pester RE, Chen CY, et al. Dicer1 func-tions as a haploinsufficient tumor suppressor. GenesDev. 2009;23(23):2700-2704.33. ZhangH,KolbFA,JaskiewiczL,WesthofE,FilipowiczW.Singleprocessingcentermodels forhumanDicerandbacterial RNase III. Cell. 2004;118(1):57-68.34. Lau P-W, Potter CS, Carragher B, MacRae IJ. Struc-ture of the human Dicer-TRBP complex by electronmicroscopy. Structure. 2009;17(10):1326-1332.35. Ngan ES, Lang BH, Liu T, et al. A germline mu-tation (A339V) in thyroid transcription factor-1 (TITF-1/NKX2.1) in patients with multinodular goiter andpapillary thyroid carcinoma. J Natl Cancer Inst. 2009;101(3):162-175.36. Chen JF, Murchison EP, Tang R, et al. Targeteddeletion of Dicer in the heart leads to dilated cardio-myopathy and heart failure. Proc Natl Acad Sci U SA. 2008;105(6):2111-2116.37. Murchison EP, Stein P, Xuan Z, et al. Critical rolesfor Dicer in the female germline. Genes Dev. 2007;21(6):682-693.38. Wiemer EA. The role of microRNAs in cancer. EurJ Cancer. 2007;43(10):1529-1544.39. Boyerinas B, Park SM, Hau A, Murmann AE, PeterME. The role of let-7 in cell differentiation and cancer.Endocr Relat Cancer. 2010;17(1):F19-F36.40. Zhu H, Shah S, Shyh-Chang N, et al. Lin28a trans-genic mice manifest size and puberty phenotypes iden-tified in human genetic association studies. Nat Genet.2010;42(7):626-630.41. Bahubeshi A, Bal N, Rio Frio T, et al. GermlineDICER1 mutations and familial cystic nephroma. J MedGenet. 2010;47(12):863-866.




dicer1 mutations in familial multinodular goiter with and ... · author affiliations are listed at...

Documents