Content uploaded by Dorota Kula
Author content
All content in this area was uploaded by Dorota Kula
Content may be subject to copyright.
486
PRACE POGLĄDOWE/REVIEWS
Endokrynologia Polska/Polish Journal of Endocrinology
Tom/Volume 61; Numer/Number 5/2010
ISSN 0423–104X
Dorota Kula M.D., Department of Nuclear Medicine and Endocrine Oncology, Maria Sklodowska-Curie Memorial Cancer Centreand
Institute of Oncology, Gliwice Branch, 44–100 Gliwice, Wybrzeże Armii Krajowej St. 15, tel.: +48 32278 97 20, fax: + 48 32 278 94 57,
e-mail: dkula@io.gliwice.pl
Genetic predisposition to papillary thyroid cancer
Badanie predyspozycji genetycznej do raka brodawkowatego tarczycy
Dorota Kula, Michał Kalemba, Beata Jurecka-Lubieniecka, Zbigniew Puch,
Małgorzata Kowalska, Tomasz Tyszkiewicz, Monika Kowal, Daria Handkiewicz-Junak
Department of Nuclear Medicine and Endocrine Oncology, Maria Sklodowska Curie Memorial Cancer Centre and Institute
of Oncology, Gliwice Branch
Abstract
Approximately 5% of differentiated thyroid cancers are hereditary. Hereditary non-medullary thyroid cancer may occur as a minor com-
ponent of familial cancer syndromes (e.g. familial adenomatous polyposis) or as a primary feature (familial non-medullary thyroid cancer
[FNMTC]). Among FNMTC, PTC is the most common. Although a hereditary predisposition to non-medullary thyroid cancer is well
established, the susceptibility genes are poorly known. Up to now, by linkage analysis using microsatellite markers, several putative loci
have been described — 1q21, 6q22, 8p23.1-p22, and 8q24; however, validation studies have been unsuccessful. In the present review we
discuss the results of linkage analysis and the most recent results of genome wide association studies (GWAS) with high resolution SNP
(single nucleotide polymorphism) arrays. (Pol J Endocrinol 2010; 61 (5): 486–489)
Key words: papillary thyroid carcinoma, familial non-medullary thyroid cancer, genetic predisposition, SNP
Streszczenie
Około 5% zróżnicowanych raków tarczycy wykazuje predyspozycję dziedziczną. Dziedziczny nierdzeniasty rak tarczycy może występo-
wać jako składowa niektórych dziedzicznych zespołów nowotworowych, na przykład rodzinnej polipowatości jelit oraz jako rodzinny
zróżnicowany rak tarczycy (FNMTC, familial non-medullary thyroid cancer), gdzie najczęściej obserwuje się raka brodawkowatego. Choć
predyspozycja dziedziczna do nierdzeniastych raków tarczycy jest dobrze znana, to jednak geny warunkujące jej występowanie nie
zostały jeszcze poznane. Wykonane jak dotąd badania zidentyfikowały kilka loci — 1q21, 6q22, 8p23.1-p22 oraz 8q24, jednak wyniki te nie
zawsze były jednoznaczne. W niniejszej pracy omówiono rezultaty badań sprzężenia oraz ostatnio uzyskane wyniki badań związku
całego genomu (GWAS, genome wide association study) wykonano przez badania polimorfizmu pojedynczego nukleotydu (SNP, single
nucleotide polymorphism) z wykorzystaniem techniki mikromacierzy wysokiej gęstości. (Endokrynol Pol 2010; 61 (5): 486–489)
Słowa kluczowe: rak brodawkowaty tarczycy, rodzinny zróżnicowany rak tarczycy, predyspozycja genetyczne, SNP
This work was supported by Ministry of Science and Higher Education grant number N N519 579938.
Introduction
Papillary thyroid carcinoma (PTC), follicular thyroid
carcinoma (FTC), poorly differentiated (insular) thyroid
carcinoma (PDTC), and undifferentiated (anaplastic)
thyroid carcinoma are all non-medullary thyroid can-
cer histotypes (NMTC) which originate from thyroid
epithelial cells [1]. Among them, 5% constitute heredi-
tary cases (HNMTC). Very rarely NMTC may occur as
a component of familial cancer syndromes: Cowden’s
disease, familial adenomatous polyposis (FAP), Gard-
ner’s syndrome, Carney’s complex type 1, Werner’s
syndrome, and papillary renal neoplasia, or as a famil-
ial non-medullary thyroid cancer (FNMTC) [2]. FNMTC
is characterized by the presence of differentiated thy-
roid cancer of follicular cell origin in two or more first-
degree relatives [2].
Among FNMTC, PTC is the most common; howev-
er, kindred with follicular and poorly differentiated thy-
roid cancer have also been reported [2]. The pathogen-
esis of papillary thyroid carcinoma (PTC) involves al-
terations in the RET/PTC-RAS-BRAF signalling path-
way, most frequently by BRAF-activating somatic
mutations and RET/PTC rearrangements [3, 4]. The loss
of heterozygosity at the TCO and NMTC1 locus (loci
linked with FNMTC) was demonstrated in some, but not
all, tumour specimens from patients with FNMTC [2].
Genetic predisposition to PTC, which is the highest
of all cancers not displaying Mendelian inheritance [5–7],
is suggested by case-control studies, which show a 3- to
487
Endokrynologia Polska/Polish Journal of Endocrinology 2010; 61 (5)
PRACE POGLĄDOWE
10-fold higher risk in first degree relatives [3, 5, 6]. The
risk is higher for first-degree male relatives of male
probands than for first-degree female relatives of female
probands [7]. Genes responsible for PTC /FNMTC are
poorly known — genetic predisposition is expected to
be multigenetic with low- to moderate-penetrance
genes [4, 6] interacting with each other and with the
environment determining individual susceptibility [6].
Up to the present, by linkage analysis using microsatel-
lite markers, several putative loci have been described:
TCO locus (thyroid tumours with cell oxyphilia) on
19p13.2 (first identified in a family with oxyphilic thy-
roid neoplasms), NMTC1 in 2q21 (identified in a family
with hereditary transmission of the follicular variant of
PTC), MNG1 (multinodular goiter) on 14q32 (a family
with 18 cases of nontoxic multinodular goiter with
2 PTC individuals), and PRN1 locus on 1q21 (identified
in a family with PTC, nodular benign thyroid disease,
and papillary renal neoplasia) [2, 8–10]; however, in part,
the results are contradictory. The reason for this are
multiple: probably the imprecise FNMTC definition
(two or more family members affected) and, as a conse-
quence, dilution of the linkage studies with sporadic
cases plays a role here. The heterogeneity of FNMTC
syndrome also has to be considered [2].
The above-mentioned studies were performed on
individual families, which were not necessarily repre-
sentative of the vast majority of inherited cases, so val-
idation of the results was unsuccessful [10]. Recently,
new techniques — high resolution SNP arrays — have
been used in genetic predisposition studies. Using this
technique, genome-wide linkage analysis performed in
a large family (11 members affected by nodular goiter
and follicular thyroid adenoma, and 5 patients with
thyroid cancer) revealed the linkage of region 8p23.1-
-p22 with familial thyroid epithelial neoplasia [11]. The
same technique, used for linkage analysis performed
on a large, broad sampling of 38 FNMTC families, dis-
covered two distinct loci on chromosome 1q21, the first
one described earlier as a PRN1 (papillary renal neo-
plasia) locus, and the second one previously unknown
[10]. High density SNP arrays were also used to per-
form a genome-wide linkage study (large family with
PTC and melanoma), which revealed the linkage of 8q24
locus with PTC. By using microsatellites markers the
results were confirmed in 25 additional PTC families.
Further analysis of the 8q24 locus implicated a putative
non-coding RNA gene (AK023948) as a candidate gene
for PTC [5].
In the absence of large pedigrees of related individ-
uals, and to discover the chromosomal regions associ-
ated with common disease, SNP array-based techniques
can be used to perform genome-wide association stud-
ies (GWAS) [10]. Association study seems to be an ap-
propriate approach in multigenic disease as linkage
analysis does not have sufficient power to identify low-
penetrance genes [6]. Recently, a GWAS has been per-
formed on 192 Icelandic patients with PTC or FTC and
37,196 healthy controls, in which 304,983 SNPs were
tested for association [7]. The strongest signals were
obtained for two SNP: rs965513 (Aallele) on 9q22.33 and
rs944289 (T allele) on 14q13.3. The results were con-
firmed in case control groups of European descent:
342 cases and 384 controls from the United States and
90 cases and 1343 controls from Spain. Combination of the
results from Iceland, the United States, and Spain result-
ed in OR 1.75 for A allele of rs965513 (p = 1.7 × 10–27) and
1.3 for T allele of rs944289 (p = 2.0 × 10–9). In the gener-
al population, almost 11% of individuals were homozy-
gous for rs965513 A allele, 32% for rs944289 T allele, and
3.7% were homozygous for both variants. Homozygous
carriers of rs965513 A allele had a 3.1-fold higher risk
for thyroid cancer then non-carriers, and a 1.9-fold high-
er for rs944289 T allele, respectively. The risk for dou-
bly homozygous individuals was 5.7-fold greater. For
the combined data, the frequency of rs965513 A allele
carriers was higher among patient diagnosed at
a younger age (rs944289 had no effect) [7]. There were
no differences observed between males and females for
rs965513 or for rs944289. Subsequently, the effects of
two SNPs were analyzed in the combined data in the
main histotypes of thyroid cancer (~ 85% of the Span-
ish and Icelandic cases were PTC, ~ 12% were FTC,
whereas all of the United States cases were PTC). For
rs956613 A allele the OR for PTC was 1.8 (p = 4.7 × 10–23)
and for FTC OR was 1.55 (p = 0.016), and for rs944289 T
allele OR was 1.32 (p = 2.0 × 10–6) and 1.63 (p = 0.007),
respectively. This demonstrated that two SNPs in-
creased the risk of two main histotypes of thyroid can-
cer (the numbers of other histotype cases in the com-
bined data were too small to be considered) [7].
The SNP rs965513 lies on 9q22.33 within the linkage
disequilibrium (LD) region where XPA, FOXE1, C9orf156,
and HEMGN genes are localized, and the FOXE1 (fork-
head box E1, formerly TTF2, Thyroid Transcription
Factor 2) is the closest gene. FOXE1 is a centre of the
regulatory network of transcription factors, which ini-
tiate thyroid differentiation at the embryonic stage. It is
also important for the maintenance of the thyroid dif-
ferentiated state in adults as it is involved in regulating
the transcription of thyroglobulin and thyroperoxidase
(TPO) gene, which are pivotal for thyroid hormone syn-
thesis [6, 7]. The expression of FOXE1 has been shown
to be abnormal in thyroid tumours, and its mutations
among other phenotypes cause human syndromes as-
sociated with thyroid agenesis.
The SNP rs944289 on 14q13.3 is localized in the LD
region where no RefSeq genes are present and the clos-
488
Genetic predisposition to papillary thyroid cancer Dorota Kula et al.
PRACE POGLĄDOWE
est genes are BRMS1L, MBIP, SFTA3, and NKX2-1
(TTF1). The best candidate as a source of the associa-
tion signal obtained for the 14q13.3 locus is NKX2-1,
which has a prominent role in thyroid development.
Its expression is altered in thyroid tumours [7]. As both
FOXE1 and NKX2-1 genes are involved in the biology
of the thyroid gland, in the next step the effects of
rs965513 and rs944289 on the TSH, free T3 and T4 serum
levels were assessed. The measurement was taken over
a period of 11 years on the material from Icelanders not
having thyroid cancer. Both rs966513 A and rs944289 T
alleles were associated with a decreased TSH serum
concentration (p = 2.9 × 10-14 and p = 0.03, respectively).
rs965513 A allele was associated with an increase in T3
levels and with a decrease in T4 levels (p = 0.003 and
p = 6.1 × 10–5, respectively). For rs944289, no effects on
either T3 or T4 were observed [7]. These data showed
the influence of at least the rs945513 in the 9q22.33 lo-
cus on the thyroid function.
An independent, recently performed candidate gene
association study also revealed the association of FOXE1
with PTC [6]. The study comprised “tag SNP” (used to
infer LD blocks according to the HapMap project) and
putative functional SNP in genes involved in thyroid
cell differentiation and proliferation, and in genes found
to be differentially expressed in thyroid carcinoma (as
described in public databases CGAP-SAGE). A total of
615 Spanish cases and 525 controls were genotyped for
768 SNPs localized in 97 genes. The strongest evidence
of association with PTC was observed for SNP in an LD
block spanning the entire FOXE1 gene. The results were
validated in an independent Italian series of 482 cases
and 532 controls. The strongest association was ob-
served for rs1867277 A allele localized in the promoter
sequence, which was a different SNP to that observed
in the Icelandic GWAS study. The combined OR (per
allele) for rs1867277 was 1.49 (p = 5.9 × 10–9). The
FOXE1 in this study was particularly associated with
the classic PTC variant. Functional assays revealed the
recruitment of USF1 and USF2 transcription factors by
rs186277 A allele, while both alleles, G and A, formed
a complex in which DREAM, CREB, and aCREM partici-
pated. Transient transfection study assays revealed that
CREB, and strongly aCREM, activated FOXE1 promot-
er, while DREAM reduced aCREM dependent-tran-
scriptional induction. USF factors also induced signifi-
cant increases in FOXE1 transcription activity when
rs1867277 A allele was present. Transcriptional activation
of the FOXE1 gene by binding transcription factors
aCREM and CREB was regulated by hormonal factors,
particularly by TSH via cAMP. The authors concluded
that in this way transcription factors could regulate
FOXE1 expression in response to TSH in a physiological
situation, but also indicated that FOXE1-specific studies
were needed to understand its role in thyroid tumour
development. FOXE1 gene belongs to the forkhead fam-
ily of transcriptional factors, which has recently been
identified as a molecular signature for epithelial to mes-
enchymal transition in human colon cancer.
The authors hypothesized that increased FOXE1
expression in thyroid carcinomas (which has been ob-
served to parallel the differentiation process of thyroid
carcinomas) could be related to a motile advantage of
malignant thyroid cells, which would be enhanced by
the presence of the rs1867277 A risk predisposing allele
[6]. Until now, no data are available on the expression
of FOXE1 in PTC.
As mentioned previously, PTC may be caused by
the interaction of multiple genes, either protein-encod-
ing genes or regulatory genes. Recently, microRNA
(miRNA) genes have been implicated in contributing
to the pathogenesis of PTC. MicroRNAs are small non-
coding RNA molecules which negatively regulate the
expression of other genes. MicroRNAs are transcribed
from endogenous DNA and they inactivate specific
mRNAs and interfere with the translation of target pro-
teins [4]. Thus, microRNAs regulate many processes like
development, apoptosis, cell proliferation, and hae-
matopoiesis; they also act as tumour suppressor genes
and oncomirs. Expression of microRNAs has been found
to vary between cancers and normal cells and among
different types of cancers [4]. For example, miR-221,
miR-222, and mir-146 are upregulated in PTC compared
with unaffected thyroid tissue [3]. Recently it has been
found that common G/C polymorphism (rs2910164)
within the pre-miR-146a sequence is associated with
PTC [4]. In material coming from 608 sporadic PTC pa-
tients and 901 controls from Finland, Poland, and the
United States the frequency of genotype differed sig-
nificantly (p = 0.000002). The G/C heterozygosity was
associated with an increased risk of PTC (OR = 1.62;
p = 0.0000007) in comparison with homozygosity, while
both homozygous states were protective (OR = 0.42,
p = 0.0027 for CC v. GG + GC; OR = 0.69, p = 0.0006
for GG v. CC + GC; OR = 0.5, p = 0.024 for CC v. GG).
It was shown that this polymorphism reduced the
amount of pre- and mature miR-146a from the C allele
compared with the G allele. By evaluation of a com-
bined group of DNA samples obtained from the blood
of PTC patients and from normal thyroid tissue in pa-
tients with PTC, it has been found that 3.6–6.1% of PTC
cases exhibit mutation from GG or CC in the germ-line
DNA toward GC heterozygosity in the tumour. The
authors concluded that polymorphism in pre-miR-146a
might play a role in tumourigenesis. The effects of miR-
NA polymorphism were expected to be mediated by
target genes whose expression was affected by the SNP
status. miR-146a is known to be involved in NF-kappa
489
Endokrynologia Polska/Polish Journal of Endocrinology 2010; 61 (5)
PRACE POGLĄDOWE
B regulation and the reduction in miR-146a leads to less
efficient inhibition of target genes: RET-PTC1 (frequent
rearranged proto-oncogene in PTC), TRAF6, and IRAK1
(involved in the Toll-like receptor and cytokine signal-
ling pathway). The role of Toll-like receptors and
NF-kappa B in thyroid tumourigenesis is well estab-
lished [4]. It was also shown that miR-146a GC heterozy-
gotes produced three mature microRNAs: one from the
leading strand (miR-146a), and two from the passenger
strand (miR-146a*G and miR-146a*C), each with its dis-
tinct set of target genes, whereas each homozygote pro-
duced two microRNAs (one from leading strand, the
second from passenger strand). As shown by microar-
ray and TaqMan real time PCR data, the difference in
expression of miR-146a and other transcript between
the tumour and unaffected parts of the thyroid from
GC and GG patients can be observed. These data sug-
gested that altered microRNAs might be early factors
playing a role in the tumourigenesis of PTC [12].
To sum up, although many studies have been per-
formed, up to the present the genes predisposing to
FNMTC still seem not to be sufficiently identified. FN-
MTC tends to be more aggressive than sporadic cancer,
is characterized by early age at onset, is often bilateral,
and may have mixed PTC and FTC features [13]. Many
medical centres recommend more aggressive treatment
of FNMTC cases [2, 8, 10]. Identification of the suscepti-
bility genes would enable FNMTC screening, early di-
agnosis, and as a result prophylactic treatment and im-
proved patient outcome [2]. However, because of mul-
tigenetic predisposition to FNMTC due to the existence
of heterogenous groups of histological variants, the is-
sue is still a challenge and multicenter analysis would
be necessary to reach the final goal [2].
Reference
1. Malchoff CD, Malchoff DM. Familial nonmedullary thyroid carcinoma.
Cancer Control 2006; 13: 106–110.
2. Vriens MR, Suh I, Moses W et al. Clinical features and genetic predispo-
sition to hereditary nonmedullary thyroid cancer. Thyroid 2009; 19: 1343–
–1349.
3. He H, Jazdzewski K, Li W et al. The role of microRNA genes in papillary
thyroid carcinoma. Proc Natl Acad Sci USA 2005; 102: 19075–19080.
4. Jazdzewski K, Murray EL, Franssila K et al. Common SNP in pre-miR-
146a decreases mature miR expression and predisposes to papillary thy-
roid carcinoma. Proc Natl Acad Sci USA 2008; 105: 7269–7274.
5. He H, Nagy R, Liyanarachchi S et al. A susceptibility locus for papillary
thyroid carcinoma on chromosome 8q24. Cancer Res 2009; 69: 625–631.
6. Landa I, Ruiz-Llorente S, Montero-Conde C et al. The variant rs1867277
in FOXE1 gene confers thyroid cancer susceptibility through the recruit-
ment of USF1/USF2 transcription factors. PLoS Genet 2009; 5: e1000637.
7. Gudmundsson J, Sulem P, Gudbjartsson DF et al. Common variants on
9q22.33 and 14q13.3 predispose to thyroid cancer in European popula-
tions. Nat Genet 2009; 41: 460–464.
8. Capezzone M, Marchisotta S, Cantara S et al. Familial non-medullary
thyroid carcinoma displays the features of clinical anticipation suggestive
of a distinct biological entity. Endocr Relat Cancer 2008; 15: 1075–1081.
9. Lesueur F, Stark M, Tocco T et al. Genetic heterogeneity in familial non-
medullary thyroid carcinoma: exclusion of linkage to RET, MNG1, and
TCO in 56 families. NMTC Consortium. J Clin Endocrinol Metab 1999;
84: 2157–2162.
10. Suh I, Filetti S, Vriens MR et al. Distinct loci on chromosome 1q21 and
6q22 predispose to familial nonmedullary thyroid cancer: a SNP array-
based linkage analysis of 38 families. Surgery 2009; 146: 1073–1080.
11. Cavaco BM, Batista PF, Sobrinho LG et al. Mapping a new familial thy-
roid epithelial neoplasia susceptibility locus to chromosome 8p23.1-p22
by high-density single-nucleotide polymorphism genome-wide linkage
analysis. J Clin Endocrinol Metab 2008; 93: 4426–4430.
12. Jazdzewski K, Liyanarachchi S, Swierniak M et al. Polymorphic mature
microRNAs from passenger strand of pre-miR-146a contribute to thyroid
cancer. Proc Natl Acad Sci USA 2009; 106: 1502–1505.
13. Morrison PJ, Atkinson AB. Genetic aspects of familial thyroid cancer.
Oncologist 2009; 14: 571–577.