DICER1 mutations in Familial Multi-Nodular Goiter with and without Ovarian Sertoli-Leydig Cell Tumors

Abstract

Context

Non-toxic multinodular goiter (MNG) is frequently observed in the general population, but little is known about the underlying genetic susceptibility to this disease. Familial cases of MNG have been reported and there are five such published families which also contain individuals with Sertoli-Leydig cell tumors of the ovary (SLCT). Germline mutations in DICER1, a gene that codes for an RNase III endoribonuclease, have recently been identified in families affected pleuropulmonary blastoma (PPB), some of whom include cases of MNG and gonadal tumors such as SLCT.

Objective

To determine whether familial MNG with or without SLCT in the absence of PPB was caused by mutations in DICER1.

Design, Setting and Patients

From September 2009 to September 2010, we studied two MNG families and three MNG/SLCT families. We screened affected probands for mutations in the DICER1 gene. We investigated blood lymphocytes, MNG and SLCT tissue from family members for loss of the wild-type allele (loss of heterozygosity), DICER1 expression and microRNA dysregulation.

Main Outcome Measure(s)

Detection of germline DICER1 gene mutations in familial MNG with and without SLCT.

Results

We identified and characterized germline DICER1 mutations in all five families. Molecular analysis of the three SLCTs showed no loss of heterozygosity at DICER1, and IHC analysis in two available samples showed strong expression of DICER1 in Sertoli cells, but weak staining of Leydig cells. MicroRNA profiling of RNA derived from lymphoblastoid cell lines from both affected and unaffected members of the familial MNG cases revealed miRNA perturbations in DICER1 mutation carriers.

Conclusions

DICER1 mutations predispose to both familial MNG and MNG with SLCT, independent of PPB and germline DICER1 mutations lead to dysregulation of miRNA. This could be investigated further as a possible novel mechanism of tumorigenesis.

Introduction

MNG is a common disorder, characterized by nodular overgrowth of the thyroid gland, usually in the setting of diffuse parenchymal hyperplasia^1–3. At autopsy, up to one-third of individuals have thyroid glands that contain multiple nodules⁴. The biochemical and genetic basis of many of the known sub-types of thyroid goiter has been elucidated⁵, but little is known about underlying genetic factors in the non-toxic form, which predominates in iodine-replete populations. From one twin study it was estimated that genetic factors account for about 40% of non-toxic goiter in females⁶. Two loci for familial MNG have been identified – MNG1 on chromosome 14q (MIM %138800)⁷ and MNG2 on the X-chromosome (MIM %300273)⁸. Interestingly, several published reports identify ovarian Sertoli-Leydig cell tumors (SLCT) occurring with familial MNG (Table 2). SLCTs represent less than 0.5% of all ovarian neoplasms⁹, but typically occur in younger women and are the most common androgen-producing ovarian tumors^10,11.

Table 2.

Published cases of familial SLCT or SLCT+MNG.

Study	Date	Age at diagnosis of SLCT	Thyroid Disease	Familial MNG
Henderson16	1951	sister(1), 16yrs sister(2), 19yrs	not stated	not stated
Javert and Finn17	1951	mother, age not stated daughter 17 yrs	mother, not stated daughter, MNG, 17 yrs	not stated
Javert and Finn17	1951	53 yrs	MNG (at autopsy), 53 yrs	not stated
Accardo and Condorelli18	1966	2 sisters, 23 yrs	not stated	not stated
Goldstein and Lamb19	1970	cousin (1) 17 yrs cousin (2) 15 yrs	(1) not stated (2) MNG, 13 yrs	absent
Jensen et al.14	1974	sister(1), 18 yrs sister(2), 14 yrs mother, ovarian tumor, type unknown, 13 yrs	sister (1), MNG, 17 yrs sister (2), no MNG at 14 yrs mother, MNG, 23 yrs	4 cases in all
Benfield et al.20	1982	16 yrs	MNG, 14 yrs	6 cases in all
O’Brien and Wilansky 9*	1982	18 yrs	MNG, 16 yrs	7 cases in all
Young and Scully 1	1985	age not stated 1/207 cases of SLCT	MNG, age not stated	not stated
Zaloudek and Norris21	1984	2 cases, age not stated 2/64 cases of SLCT	nodule or goiter, age not stated	not stated
Whitcomb et al.13*	1986	sister(1), 55 yrs sister(2), 57 yrs	sister(1) Graves disease, sister(2) PTC, 53 yrs no MNG reported	absent
Gheorghisan-Galateanu et al.22	2003	63 yrs	MNG, 44 yrs	7 cases in all
Niedziela8*	2008	14 yrs	MNG, 13 yrs	3 cases in all
This study (Family C)	2010	32 yrs	MNG 18 yrs	absent

^*= included in this study, PTC = papillary thyroid cancer

Pleuropulmonary blastoma (PPB), a rare pediatric mesenchymal thoracic tumor, has recently been linked to the same region as MNG1 on 14q. Germline mutations in DICER1 (14q31) were identified in 11 of 11 PPB families featuring children with PPB, cystic nephroma or embryonal rhabdomyosarcoma¹², known as the PPB Family Tumor and Dysplasia Syndrome (PPB-FTDS, OMIM #601200). Some PPB families also contain both MNG and gonadal tumors (including SLCT)^13,14.

As a member of the RNase III family, DICER1 is involved in the generation of double-stranded microRNAs (miRNAs) short, non-coding RNAs that modulate gene expression at the post-transcriptional level^15,16. Relevant to this function are two RNase III domains and a PAZ domain whose location within the protein determines the lengths of miRNAs generated¹⁷. Given the tumor spectrum of PPB families and the heterogeneous functions of DICER1, we hypothesized that DICER1 mutations could be present in kindred with multiple cases of MNG (with and without SLCT), We subsequently explored the phenotypic and molecular consequences of the DICER1 mutations observed.

Methods

Cases and controls

We used the following criteria to identify cases and families 1) Three or more cases of MNG, 2) Two or more cases of MNG plus one case of SCLT, or 3) MNG and SCLT occurring in the same individual (see eTable 1). Cases were identified through Pubmed.gov searches and through patients referred to the McGill Cancer Genetics Program. Members of two published MNG/SLCT families ^18,19 were contacted through the authors (families A and B). Another case of MNG/SLCT (family C) was identified through a local clinician. MNG cases were reviewed by V-HN, HRH and DB-DS and SLCT cases were reviewed by DB-DS and JA. Diagnoses of SLCT were made by examining tissue slides stained with haematoxylin and eosin. Families D and E were described previously as MON236 and MON152 respectively ^7,20,21; updated pedigrees including newly affected family members are presented in Figure 1. All families consented to participate in the study which was approved by the appropriate institutional ethics review boards. A larger series of germ-line DNA from anonymized MNG and differentiated thyroid cancer cases from centers in the UK, Bologna and Montreal were screened for DICER1 mutations. This comprised a total of 88 cases of differentiated thyroid cancer, 20 of which also had MNG and a further 22 cases of MNG all of which have a family history of MNG and differentiated thyroid cancer (eTable 2). Anonymous control DNA samples were used from consented individuals with no history of cancer who attended the Jewish General Hospital in Montreal between June and September 2009. For the immunohistochemical (IHC) analysis, 28 anonymized non-hereditary MNGs from Hospital Dr. A. Oñativia, Salta, Argentina were used as controls.

Figure 1. MNG/SCLT and familial MNG pedigrees.

A–E, pedigrees and electrophoretograms of probands with DICER1 mutations and familial MNG/SCLTO (A–C) or familial MNG (D–E). The genomic reference sequence used is refseq NG_016311.1. The vertical line in electrophoretogram E represents the intron/exon boundary. Symbols in black indicate MNG; “+” or “−” denotes mutation status (+ = mutation). SLCT = Sertoli-Leydig Cell Tumor, CRC = colorectal cancer, FM = fibrolamellar meningioma, RMS = rhabdomyosarcoma.

Mutation Analysis

The 26 coding exons of DICER1 were screened by high-resolution melting (HRM) analysis using the LightScanner instrument (Idaho Technologies Inc., Utah, USA) or sequencing. Primer sequences and the protocol for amplification were adapted from Hill et al. ¹² with some primers being changed to generate smaller amplicons necessary for HRM (see eTable 3 and eMethods for further information). Non-truncating mutations were further investigated by cDNA analysis and/or comparative expression analysis (see eMethods).

RNA analysis

Total RNA was extracted from independent cultures of lymphoblastoid cell lines (LCLs). mRNA was retrotranscribed using an oligo-dT and Superscript III reverse transcriptase (Invitrogen). Expression of DICER1 mRNA was measured by quantitative real-time PCR. Pre-designed TaqMan assays were used to specifically amplify cDNA derived from both mutant and wild-type DICER1 and GAPDH mRNAs. Inhibition of nonsense-mediated mRNA decay was performed as previously described²². Complete experimental details are provided in the eMethods.

miRNA profiling

miRNA profiling of LCLs was performed by the Vancouver Prostate Centre Microarray Facility. Total RNA including the small RNA fraction was isolated and quantified. The quality of the RNA was confirmed prior to fluorescent end-labeling and hybridization to the Unrestricted Human miRNA Microarrays Release 12.0 (Design ID 021827) using the Agilent miRNA Microarray System with miRNA Complete Labeling and Hyb Kit V2. Following hybridization, the microarrays were scanned, quantified and analyzed using Agilent system software components (details are provided in the eMethods). Data was normalized by flooring values below 0.05 to 0.05 and applying per chip normalization to a set of positive control genes that were selected on the criteria that raw data was above 20 for all samples. Lists of significantly differentially expressed genes were determined by both a t-test between the two conditions (P = 0.05) and fold-change filtering with a fold change value of 2.0. The heatmap was created by Cluster 3.0 and Java gene Treeview software^23,24.

Frozen tissue samples (normal thyroid tissue from a non-carrier and one goiter from a DICER1 mutation carrier) were homogenized in lysis buffer and total RNA was prepared. Fifty ng was reverse transcribed and the cDNA was applied on a TaqMan low-density array, with data acquisition via the Applied Biosystems 7900HT Fast Real-time PCR system. Data analysis was performed using the R analysis software (ABI). Individual miRNA expression was validated using Taqman miRNA Assays (ABI), and values were normalized to U6 snRNA. Fold change was calculated using the DDCT method²⁵. Further experimental details are provided in the eMethods.

Loss of Heterozygosity in MNG and SLCT

DNA was extracted from lymphocytes from the probands and, where available, from macro-dissected formalin-fixed, paraffin-embedded tumor and goiter tissue. PCR and evaluation of loss of heterozygosity (LOH) was carried out as previously described ²⁶. Further details are provided in the eMethods.

IHC analysis and Western Blotting

IHC analysis was performed on deparaffinized 5-μm tissue sections incubated with anti-DICER antibody (1:50 for goiters, 1:100 for SLCTs, ab14601, Abcam, MA). Staining was completed using Dako Envision+ system-HRP (Dako, Denmark). Western blots were performed using standard procedures using mouse monoclonal ab (13D6) anti-DICER1 (1:1000, ab14601, Abcam) or rabbit polyclonal anti–tubulin (1:1000; Cell Signaling).

Results

DICER1 Mutation Analysis

We identified DICER1 mutations in the five studied families with MNG and/or SLCT. The pedigrees are shown in Figure 1 and results summarized in Table 1. Among the three MNG/SLCT families (Figure 1, A through C), family A was initially described in 1981 as an MNG family where the proband had MNG at the age of 16 and SLCT at the age of 18 years¹⁹. The proband and three other family members carry a c.871_874delAAAG mutation which results in an mRNA product that contains a premature termination codon at position 296. No mutant protein is produced because of the action of nonsense-mediated mRNA decay (NMD) (Figure 2A, left). Family B is a recently reported MNG/SLCT family¹⁸ where we identified a g.49531C>G mutation (refseq NG_016311.1) which creates a de novo splice site with a predicted strength of 0.37 as compared with 0.50 for the authentic site²⁷. Sequencing of amplification products from the exon 15–16 junction reveals that 100% of the mutant allele produces a mutant transcript containing an in-frame deletion of the first 21 base pairs of exon 16 and therefore is not subject to NMD (Figure 2B). This new transcript generates a DICER1 protein with a p.Ile813_Tyr819del mutation resulting in a change to the altered PAZ structure (compare eFigure 1A to eFigure 1B). This mutation was detected in three individuals, including the proband who developed a SLCT at the age of 14. In Family C the proband had MNG at 18 years and SLCT at 32 years, with no family history of cancer or MNG. She and her unaffected mother carry a c.5018_5021delTCAA mutation, which results in an mRNA product that is subject to NMD as it contains a premature termination codon at position 1703 (Figure 2A, right).

Table 1.

Clinical and molecular summary of families reported in this study

Family	Clinical Features	DICER1 Mutation	Effect of Mutation	LOH (tissue studied)	IHC (tissue studied)
A	1 SLCT+MNG, 3 MNG	c.871_874delAAAG	protein truncation	no (SLCT)	not done
B	1 SLCT+MNG, 2 MNG	g.49351C>G (c.2437_2457del21)	creation of de novo splice site leading to an in-frame deletion of 21 base pairs	no (MNG) no (SLCT)	no staining (MNG), increased staining in Sertoli cells (SLCT)
C	1 SLCT+MNG	c.5018_5021delTCAA	protein truncation	no (SLCT)	increased staining in Sertoli cells (SLCT)
D	20 MNG	c.2516C>T	missense	no (MNG)	cytoplasmic staining (MNG)
E	6 MNG 1 A-RMS*	c.2805-1G>T	creation of de novo splice site leading to an in-frame deletion of exon 18	no (MNG)	cytoplasmic staining (MNG)

^*A-RMS = alveolar rhabdomyosarcoma. This DICER1 mutation carrier died at age 20 years from a malignant paravertebral tumor characterized as A-RMS (molecular re-confirmation of this diagnosis was impossible due to lack of tissue). A-RMS has not been observed in DICER1-related PPB families although unusual sarcomas are characteristic.

Figure 2. DICER1 mutations and their effect on mRNA.

A: mutant mRNAs from families A and C are targeted and degraded by the nonsense-mediated mRNA decay. LCLs from carriers of the c.871_874delAAAG (left) or c.5018-5021delTCAA (right) mutations from family A and C, respectively and from a non-carrier were treated or not treated with cycloheximide (CHX), a potent inhibitor of the NMD. Amplification of the 7–8 exon junction (left) or part of exon 23 (right) in treated and non-treated LCLS indicated that in presence of CHX the amount of mutant mRNA was increased in both carriers. No mutant mRNA was detected in the controls. B: amplification of the exon 15–16 junction of mRNA derived from LCLs from one individual from family B indicated that mutant mRNA produced is insensitive to NMD (left). The upper band seen in this individual is a non-specific heteroduplex. Sequencing of the PCR products revealed that the g.49351C>G mutation leads to the creation of a de novo splice site at position 22 of exon 16 and the in-frame deletion of the first 21 bases of exon 16 (right). C: Splice mutation in family E. Primers in DICER1 exons 17 and 19 amplify a heterozygous 183bp deletion in c.2805-1G>T mutation carrier cDNA (left). Sequence analysis of the 2 amplified products in G/T carrier (right) shows perfect excision of exon 18 in the smaller fragment (middle trace) and wild-type sequence in the larger fragment (upper and lower traces).

DICER1 mutations were also found in both familial MNG families previously linked to chromosome 14q (Figures 1D–E), but not in germ-line DNA from 110 MNG and differentiated thyroid cancer cases (see eTable 2 for further information on these cases). In family D (Figure 1D) we identified a missense DICER1 variant, c.2516C>T, leading to p.Ser839Phe, in 23 family members, 20 of whom had MNG. We could not conclusively confirm the presence or absence of MNG in three remaining variant carriers, but there was no MNG in any of the 13 family members without the variant. Taken together, these observations indicate that p.Ser839Phe is very highly penetrant for MNG. The variant was not present in DNA from 455 anonymous controls with no history of cancer from the Jewish General Hospital in Montreal. On the basis of a prevalence of MNG in the population of 4%¹, and assuming a mutation frequency of .005, the probability of segregation of this mutation with MNG in family D purely by chance is approximately 10⁻⁵.

The amino acid serine is conserved in higher vertebrates (eFigure 2) and predictive software was used to help establish the overall effect of the p.Ser839Phe mutation on the DICER1 protein in family D. According to the probabilistic program SIFT (Sorting Intolerant From Tolerant) (http://sift.jvci.org/), p.Ser839Phe was assigned a score of .05, which is at the limit of normal (≥ 0.05)²⁸. Similarly, PolyPhen2 (http://genetics.bwh.harvard.edu/pph2/), predicts this mutation is “possibly damaging” (score .442)²⁹, as the score lies between .16 and .85. The Ser to Phe change is predicted to disrupt an alpha helix in the PAZ domain (eFigure 1C).

The second familial MNG kindred (Family E, Figure 1) carries a fully segregating splice-site mutation, c.2805-1G>T, resulting in an in-frame deletion of exon 18 (Figure 2C), altering the structure of DICER1 by eliminating part of the PAZ domain (eFigure 3). Comparative expression analysis of cDNA synthesized from RNA isolated from lymphocytes from members of families D and E indicated that two alleles were equally expressed in mutation carriers, ruling out the possibility of a promoter mutation or a large genomic deletion of DICER1 that might be in cis with the identified mutations (eFigure 4).

RNA and Protein analysis of DICER1 in LCLs

We analyzed mRNA and protein extracted from LCLs of seven carriers from families B, D and E (with non-truncating mutations), two carriers from families A and C (with NMD-sensitive mutations), and five mutation-negative controls (3 from families D and E and 2 non-familial controls), mRNA quantification showed highly variable levels of mRNA between and within families and no consistent relationship between mRNA levels and either the presence of goiter or SLCT was observed (eFigure 5). DICER1 Western blots supported these findings (data not shown).

LOH and IHC studies in Goiter and SLCT

Studies of LOH using DNA extracted from goiters from families B, D and E, together with constitutional DNA, revealed no evidence of LOH (Table 1, Figure 3A, left). This is not surprising, given that goiters are hyperplastic, rather than neoplastic, lesions. IHC analysis of these goiters using anti-DICER1 antibody ab14601 revealed a mixed picture, with no staining of two goiters from family B (II-1 and II-2), but clear cytoplasmic staining of several goiters from family D and E (an example shown in Figure 3B, bottom left). To compare these findings with that seen in non-familial goiter, we immunostained 28 non hereditary MNGs from Salta, Argentina. Twenty-five of these MNGs stained with the anti-DICER1 antibody (an example is shown in Figure 3B, top left) whereas three showed no staining. Therefore, the amount of DICER1 protein does not appear to be associated with DICER1 mutation status, similar to what we observed in LCL RNA and protein.

Figure 3. Loss of Heterozygosity and Immunohistochemistry studies.

A: LOH analysis of Family B, no loss of heterozygosity in proband affected tissue. I and III = gDNA from lymphocytes; II = gDNA from goiter; IV = gDNA from SLCT. Arrow indicates mutation.

B: Immunohistochemistry of MNG and SLCT. Left: representative staining of DICER1 in MNG from a non-hereditary case (top) and from a mutation carrier from family D (bottom). DICER1 staining is detected in cells bordering the vesicles and is heterogeneous, being partly dependent on the amount of cytoplasm, which is reduced when surrounding large vesicles. No difference in staining is evident when comparing hereditary and non-hereditary cases of MNG. Right: representative staining of DICER1 in normal ovary (top) compared to a Sertoli-Leydig cell tumor. This tumor is composed of prominent cords and tubules of immature Sertoli cells with discrete clusters of Leydig cells. The Leydig cells contained a round nucleus with prominent nucleolus and an abundant pale and eosinophilic cytoplasm. The immature Sertoli cells have small round nuclei and scanty cytoplasm. No staining was detected in normal ovary. Cytoplasm of Sertoli cells (S) was strongly stained and Leydig cells (L) show very weak staining.

SLCT tissue from the three DICER1 mutation carrier probands in families A through C was analysed for LOH, and in each case there was no evidence of loss of the wild-type allele (Figure 3A, right and eFigure 6A). IHC analysis in the two available SLCT in DICER1 mutation carriers showed increased expression of DICER1 in Sertoli cells, but the staining was much weaker in Leydig cells (Figure 3B, bottom right and eFigure 6B). By contrast, neither ovarian carcinomas (n = 5, not shown) nor normal ovary (Figure 3B, top right) stained with the ab14601 anti-DICER1 antibody.

miRNA Assays

We used RNA extracted from LCLs established in five carriers and four mutation-negative controls, all from families D and E (plus one unrelated control) for the miRNA assays. Global miRNA profiling of a panel of 851 human miRNAs identified 94 miRNAs (11% ) that were significantly differentially expressed in the five affected DICER1 mutation carriers compared with the five unaffected non-carriers (fold-change≥2 and P < .05; Figure 4A). We then compared miRNA profiles in the RNA extracted from one fresh frozen DICER1-related MNG with normal thyroid tissue, to determine miRNAs differentially expressed in the MNG and the normal tissue. Comparing the two lists, only five miRNAs (miR-345, let-7a, miR-99b, miR-133, miR-194) were decreased in RNA from both LCLs and the DICER1-related MNG. Of these, only miRNA 345 is highly expressed uniquely in the normal thyroid³⁰. We focused on let-7a and mi345 and markedly lower levels of both miRNAs were seen in the DICER1-related goiter, when compared with both normal thyroid gland tissue and a follicular thyroid carcinoma (Figure 4B).

Figure 4. miRNA microarray studies.

A: Differentially expressed miRNAs between carriers and non-carriers of mutations in DICER1 from family D and E. Unsupervised hierarchical clustering showed clustering of the samples according their mutation status. The Cluster 3.0 software was used, wherein miRNA expression values were median-centered and clustered using Pearson uncentered correlation distance and average linkage.

B: Quantification of expression levels of miRNAs miR-345 and let-7a by real-time PCR in frozen tissues. miRNA-345 and let-7a are downregulated in goiter tissue from individual IV-4 from family E compared to normal thyroid and follicular thyroid cancer tissues from non-carriers of DICER1 mutations.

Comment

Here we report germline DICER1 mutations in familial MNG and MNG/SLCT families. In the two larger families with MNG, the DICER1 mutations are highly penetrant for goiter. MNG is prevalent in the general population and occurs within the wide spectrum of conditions occurring in kindred with pleuropulmonary blastoma^13,31. Germ-line DICER1 mutations are found in more than half of all children with PPB and most of the mutations identified are predicted to result in truncated proteins^12,32. Notably, the penetrance of the reported DICER1 mutations for PPB (and the other major clinical characteristics of the syndrome) is low, with many gene carriers remaining unaffected into adulthood. By contrast, the three non-truncating mutations reported here (Table 1) are highly penetrant for MNG. Moreover, the absence of all other known features of the PPB-FTDS in the large six-generation Family D kindred suggests that the functional effect of the p.Ser839Phe mutation is qualitatively different from that associated with truncating mutations. Similarly, the mutation in family E results in a DICER1 protein that lacks part of the PAZ domain (eFigure 1C), but is otherwise normal. In family B, where three individuals developed early-onset MNG, the PAZ domain is significantly altered by the in-frame deletion of 7 amino acids (eFigure 1B).

Two of the three DICER1 mutations seen in the SLCT/MNG families were truncating and are distributed throughout the gene with no clear genotype-phenotype correlation differentiating them from those PPB-FTDS families with reported DICER1 mutations ¹² (eFigure 7). The median diagnosis age of three DICER1-related SLCTs in this study and six reported PPB-associated SLCT¹⁴ is 13 years, considerably younger than median age of onset of 19 years in sporadic SCLT¹⁰ (P = .009, Mann-Whitney test).

The disease spectrum associated with PPB is broad¹³, and we have confirmed by molecular means that SLCT belongs in the PPB-FTDS, but can occur without PPB. It is likely that some of the reported SLCT/MNG cases (Table 2) harbor DICER1 mutations and that our findings explain the observation first made by Jensen et al.³³. Familial SCLT has also been reported (Table 2) Whether familial SLCT results from genetic predisposition is not answered by our study as the families described here have only one case of SLCT per kindred. We did screen an affected familial SLCT proband³⁴ but did not identify a DICER1 mutation; it remains possible that DICER1 mutations will be implicated in other familial SLCT, or that there are other genes responsible for familial SLCT with genes in the miRNA processing pathway, such as PASHA/DGCR8 or DROSHA, being reasonable candidates.

We also analyzed DICER1 in germ-line DNA from probands from large series of differentiated thyroid cancer cases with and without MNG, and no potentially disease-causing variants were identified. Although in general differentiated thyroid cancer cases appear not to be associated with DICER1 mutations, differentiated thyroid cancer cases in the context of SLCT or PPB may harbor DICER1 mutations.

None of the DICER1-related SCLT showed evidence of LOH, and even though the numbers are small, these findings are consistent with the notion from animal models that DICER1 does not function as a classic tumor suppressor gene but that instead tumors develop as a result of miRNA dysregulation through a possible haploinsuffienciency effect³⁵. The lack of correlation, however, between DICER1 mutation status and both mRNA and protein levels of DICER1 we observed here suggests that mechanism of tumorigenesis in human DICER1 mutation carriers may be complex and may differ between tissues.

DICER1 has several highly conserved domains, including the PAZ domain (eFigures 1 and 3) which appears to be critical for DICER1 function: a purified C-terminal fragment of DICER1 containing both RNase domains and double- strand RNA binding domain (dsRBD) showed no dsRNA cleavage activity³⁶. PAZ acts as a molecular ruler, determining where the RNase domains of DICER1 cut the pre-miRNAs to their final size³⁷. What is particularly notable in our results is that selective disruption of the PAZ domain, in a setting of an otherwise normal DICER1 protein, leads to familial MNG. Perhaps these mutations do not represent truly hypomorphic DICER1 alleles, but instead the goiter phenotype observed is the result of specific PAZ disruption rather than a true haploinsufficiency. It is, moreover, tempting to speculate that there is a close relationship between PAZ domain function and the regulation of thyroid development. In a similar vein, it is interesting that one missense mutation in the thyroid transcription factor TTF1 results only in thyroid disease³⁸, whereas loss of an entire allele, in both humans and in animal models, results in much more serious phenotypes⁵.

Perturbations of miRNAs in cancer are common³⁹, but constitutional defects in miRNAs have not previously been reported in humans. Since miRNAs have a crucial role in human development^15,40,41, it is not surprising that truncating germ-line mutations in DICER1 can lead to early–onset malignancy¹². In light of the central role of DICER1 in miRNA processing, we looked for downstream effects of miRNA dysregulation in tissues from heterozygotes in families D and E with non-truncating mutations (Figure 4A), and found that five miRNAs were consistently decreased, of which only let7a and miR345 were decreased in the goiter tissue of one carrier of the c.2805-1G>T mutation in Family E (Figure 4B). Let-7a is down-regulated in breast, pancreas and lung cancer, and malignant melanoma⁴² and disruptions in the Lin28–Let-7 pathway alter glucose metabolism and insulin sensitivity in mice⁴³, but to date Let-7a has not been implicated in thyroid disease. MiR345 is highly expressed in the thyroid gland³⁰, making this an attractive candidate to further explore MNG pathogenesis in DICER1 mutation carriers.

In summary, we report DICER1 mutations as a cause for autosomal dominant familial MNG and SLCT occurring with MNG observed more than 30 years ago by Fraumeni and colleagues³³. Our study confirms clinical observations^13,14 and definitively extends the tumor spectrum of DICER1 mutation beyond PPB, cystic nephroma, embryonal rhabdomyosarcoma and lung cysts^12,44. Further, mutations in other genes in the miRNA processing pathway may explain some of these syndromic disease combinations. Unlike SLCT, MNG is a very common condition worldwide³ and determining the role of dyrsegulated miRNA processing in the development of sporadic MNG could be an important avenue for future research.