Download Somatic complex I disruptive mitochondrial DNA

Human Molecular Genetics, 2013, Vol. 22, No. 2
doi:10.1093/hmg/dds422
Advance Access published on October 9, 2012
226–238
Somatic complex I disruptive mitochondrial DNA
mutations are modifiers of tumorigenesis that
correlate with low genomic instability in pituitary
adenomas
Ivana Kurelac1, Alan MacKay3, Maryou B.K. Lambros3, Erica Di Cesare1, Giovanna Cenacchi2,
Claudio Ceccarelli2, Isabella Morra4, Antonio Melcarne5, Luca Morandi6, Francesco Maria
Calabrese7, Marcella Attimonelli7, Giovanni Tallini6, Jorge S. Reis-Filho3 and Giuseppe
Gasparre1,∗
1
Dip. di Scienze Mediche e Chirurgiche, U.O. Genetica Medica and 2Unità Operativa di Anatomia e Istologia
Patologica, Policlinico Universitario S. Orsola-Malpighi, Bologna 40138, Italy, 3Molecular Pathology Team, The
Breakthrough Breast Cancer Research Centre, ICR, London SW3 6JB, UK, 4Servizio di Anatomia Patologica, Az.
Ospedaliera OIRM-S.Anna, Torino 10100, Italy, 5Divisione di Neurochirurgia, Az. Ospedaliera CTO-M.Adelaide,
Torino 10100, Italy, 6Dip. di Anatomia Patologica, Ospedale Bellaria, Universita di Bologna, Bologna 40126, Italy and
7
Dip. di Bioscienze, Biotecnologie e Scienze Farmacologiche, Campus ‘E.Quagliariello’, Università di Bari, Bari 70126,
Italy
Received September 19, 2012; Revised and Accepted October 3, 2012
Mitochondrial DNA (mtDNA) mutations leading to the disruption of respiratory complex I (CI) have been
shown to exhibit anti-tumorigenic effects, at variance with those impairing only the function but not the
assembly of the complex, which appear to contribute positively to cancer development. Owing to the
challenges in the analysis of the multi-copy mitochondrial genome, it is yet to be determined whether
tumour-associated mtDNA lesions occur as somatic modifying factors or as germ-line predisposing elements. Here we investigated the whole mitochondrial genome sequence of 20 pituitary adenomas with oncocytic phenotype and identified pathogenic and/or novel mtDNA mutations in 60% of the cases. Using highly
sensitive techniques, namely fluorescent PCR and allele-specific locked nucleic acid quantitative PCR, we
identified the most likely somatic nature of these mutations in our sample set, since none of the mutations
was detected in the corresponding blood tissue of the patients analysed. Furthermore, we have subjected
a series of 48 pituitary adenomas to a high-resolution array comparative genomic hybridization analysis,
which revealed that CI disruptive mutations, and the oncocytic phenotype, significantly correlate with low
number of chromosomal aberrations in the nuclear genome. We conclude that CI disruptive mutations in
pituitary adenomas are somatic modifiers of tumorigenesis most likely contributing not only to the development of oncocytic change, but also to a less aggressive tumour phenotype, as indicated by a stable
karyotype.
INTRODUCTION
Mitochondrial DNA (mtDNA) mutations frequently occur in
many human neoplasms, where they may contribute to
tumorigenic processes by modifying metabolic, apoptotic or
hypoxic mechanisms (1,2). In particular, pathogenic mtDNA
mutations leading to the disruption of respiratory complex I
(CI) assembly have been shown to block tumour growth in
∗
To whom correspondence should be addressed at: Dip. Scienze Mediche e Chirurgiche, U.O. Genetica Medica - Pad.11, Pol.S.Orsola-Malpighi,
via Massarenti 9, 40138 Bologna, Italy. Tel: +39 512088418; Fax: +39 512088416; Email: [email protected]
# The Author 2012. Published by Oxford University Press. All rights reserved.
For Permissions, please email: [email protected]
Human Molecular Genetics, 2013, Vol. 22, No. 2
mice (3,4). These mutations are hallmarks of oncocytic
tumours, which often present with an indolent, senescent-like
phenotype (5,6). Oncocytic change is not uncommon in
neoplasms of epithelial origin and is characterized by the
abnormal accumulation of dysfunctional mitochondria in the
cell cytoplasm, which arises as a compensatory effect due
to the severe mitochondrial damage caused by CI disruptive
mutations (7). Conversely, the same type of mutations is
rarely encountered in non-oncocytic tumours, which essentially harbour missense mtDNA mutations, that may instead positively contribute to cancer development (2,8,9). Therefore,
depending both on the type and mutant load of a mtDNA mutation, mitochondrial genes may act both as tumour suppressors and as oncogenes, and were recently attributed with the
functional definition of oncojanus, based on this two-faceted
role in tumorigenesis (2,4).
Tumours in which CI disruptive mutations occur in the
highest frequency (80%) are renal oncocytomas (10,11),
which are generally associated with a less aggressive behaviour and a relative chromosomal stability (12– 14). Given
that CI disruptive mtDNA mutations have shown to imply
anti-tumorigenic effects through the induction of pro-apoptotic
mechanisms (3) and/or by preventing hypoxic adaptation (4),
it is reasonable to suggest that, as a consequence, the less aggressive behaviour of tumours bearing such mutations might
be reflected in their genomic stability as well. In order to investigate the correlation between CI disruptive mutations
and genomic stability in tumours, we here studied a set of pituitary adenomas, rare tumours that mainly present as sporadic
lesions (15), although they may also occur as a manifestation
of inherited syndromes, including multiple endocrine neoplasia 1 (MEN1) (15– 17). Pituitary adenomas are classified
based on their similarity to normal parenchymal cells, clinical
outcome and the type of hormonal secretion (15). Oncocytic
change can be found focally in any adenoma type, but
diffuse and extensive mitochondrial proliferation is reported
to occur in 6 – 13% of all pituitary adenomas, which are
hence classified as pituitary oncocytomas (15,18). Similar to
other tumours presenting with an oncocytic phenotype, the
only known genetic hallmark of pituitary oncocytomas are
CI disruptive mtDNA mutations, found in 60% of the
cases (19).
An overlooked issue concerning mtDNA mutations is
whether they occur as germ-line predisposing hits or
whether they are truly somatic, randomly arising events,
which may modify tumour progression and result in the development of a low-proliferative oncocytic versus a more aggressive tumour. Since multiple copies of mtDNA are present in
each cell, mutant and wild-type molecules may co-exist in a
condition known as heteroplasmy and mutant loads may therefore vary. This aspect of mitochondrial genetics has rarely
been taken into account when analysing mtDNA mutations
in tumours. For instance, assumptions about the somatic
nature of mtDNA mutations have been made on the basis of
analyses performed with techniques like Sanger sequencing,
which are not sufficiently sensitive to detect low-level heteroplasmy (20– 22). However, recent reports based on highly sensitive next-generation sequencing techniques have revealed
that low-level mutant loads may be present in the germ line
of cancer patients (23), and the occurrence of low loads of
227
germ-line CI disruptive oncocytoma-associated mtDNA mutations has been reported in single cases (24,25).
Aims of this study were (i) to investigate the correlation
between CI disruptive mtDNA mutations and genomic aberrations in pituitary adenomas and (ii) to determine whether
these mutations act as germ-line or somatic players. A thorough genetic characterization of pituitary adenomas (oncocytic and non) was therefore performed, both in terms of
mtDNA and of chromosomal DNA alterations. Our results
corroborated the association between CI disruptive mutations
and oncocytic phenotype, provided evidence that such mutations are unlikely to exist as germ-line predisposing elements
in pituitary adenoma patients and showed that somatic CI
disruptive mtDNA mutations significantly correlate with a
stable complement of chromosomal DNA, which suggested
their modifying role in pituitary tumour progression,
likely contributing through the double-faceted oncojanus
effect (2,4).
RESULTS
Pathogenic mtDNA mutations are somatic events
frequently occurring in pituitary oncocytomas
With the aim to correlate mtDNA data with the degree of
chromosomal number aberrations, we expanded a previously
characterized sample set of 44 pituitary adenomas (19) with
additional 20 pituitary oncocytoma cases, thus increasing statistical power. The cases were carefully selected through histological and ultrastructural examination (Fig. 1). Corresponding
blood samples for the 20 novel cases were also available,
allowing us to investigate whether mtDNA mutations may
exist in the germ line as predisposing elements for oncocytic
transformation of pituitary adenomas.
The mtDNA characterization of the novel cases revealed
that 12 out of 20 (60%) presented with tumour-specific
mtDNA mutations, 6 of which were novel, whereas 6 have
previously been associated with a known disease (Table 1),
such as m.12421delA in MT-ND5 described in cases with isolated CI deficiency (26). The m.11873insC MT-ND4 mutation,
which results in a shift of the open reading frame and in the
generation of a premature stop codon, has already been
reported and shown to cause CI derangement (19,27). Seven
out of the 12 cases (58%) carried mutations in coding
regions, 2 in MT-RNR2 (16SrRNA) and 3 in tRNA genes.
On the other hand, nine mtDNA changes not defining the haplogroup and hence potentially somatic were nonetheless found
both in tumour and blood samples of seven cases (Supplementary Material, Table S1). All of these non-tumour-specific
mutations were found in homoplasmy in both tissues, indicating their probable polymorphic nature.
All novel tumour-specific mutations affecting coding
regions were predicted as probably damaging by PolyPhen1,
having high position-specific independent count (PSIC)
scores .2 (see Materials and Methods). Alignment with ClustalW2 (Fig. 2A) showed that mutations m.3413G.A,
m.3457G.A and m.15092G.A affect amino acids highly
conserved between different metazoan phyla (from Porifera
to Great Apes and Homo sapiens). Among the three somatic
mutations affecting tRNAs, m.1644G.A was previously
228
Human Molecular Genetics, 2013, Vol. 22, No. 2
Figure 1. Morphology of oncocytomas and non-oncocytic pituitary adenomas. Haematoxylin/eosin staining (magnification ×200) showing monomorphous adenomatous cells, with eosinophilic cytoplasms and small nuclei without atypia in pituitary oncocytomas (A). Mitotic figures were rare (less than 1 mitotic figure/
10 high-power fields), and tumour vasculature was delicate. On the other hand, in non-oncocytic pituitary adenoma, nuclei occupied most of the cytoplasm as
presented by predominant haematoxylin staining (B). Electron micrographs of oncocytomas (C) showed cell cytoplasm occupied by innumerable densely packed
mitochondria characterized by partial loss of matrix and irregularly organized cristae lacking intra-mitochondrial dense granules. A few rough endoplasmic reticulum profiles were observed, relatively small Golgi apparatus, sparsely distributed secretory granules and a variable accumulation of mitochondria, which
appeared to occupy up to 50% of the cytoplasmic compartment. These mitochondria were generally swollen, with partial or complete cavitation of the
matrix and a variable loss of cristae. Some mitochondria featured an irregular pattern of residual cristae which appeared short, in a crescent or concentric
array. Non-oncocytic pituitary adenomas selected for our analysis did not show gross mitochondrial alterations and often abundance of secretory granules
was evident (D).
Table 1. Somatic mtDNA mutations in pituitary oncocytomas (H)
Sample
Base change
AA change
Gene
Mutant load (%)
PSIC/DG
Annotation status
H1
m.3094G.A
–
MT-RNR2
50
–
Colonic crypts biochemical defects
(76)
Novel
Novel
Novel
Chronic external ophtalmoplegia (77)
Novel
Pituitary adenoma (19)
Isolated CI deficiency (26)
Adult Leigh syndrome (28),
hypertrophic cardiomyopathy and
MELAS (78)
Novel
Isolated CI deficiency (26)
Novel
H2
H3
H8
H9
H11
H12
H13
H14
m.3413G.A
m.3457G.A
m.2993T.C
m.12276G.A
m.15092G.A
m.11873insC
m.12421delA
m.1644G.A
G36D
D51N
–
–
G116S
–
–
–
MT-ND1
MT-ND1
MT-RNR2
MT-TL2
MT-CYB
MT-ND4
MT-ND5
MT-TV
100
80
100
50
50
50
30
100
2.396
2.075
–
From 218.50 to 211.90 kcal/mol
2.112
–
–
No change in DG
H17
H18
H19
m.3457G.A
m.12421delA
m.4427G.A
D51N
–
–
MT-ND1
MT-ND5
MT-TM
100
45
100
2.075
–
From 215.82 to 212.15 kcal/mol
AA, amino acid; PSIC, position-specific independent counts; CI, complex I; MELAS, mitochondrial myopathy, encephalopathy, lactic acidosis and stroke-like
episodes.
described as associated with adult Leigh syndrome, hypertrophic cardiomyopathy and mitochondrial myopathy, encephalopathy, lactic acidosis and stroke-like episodes (MELAS)
syndrome (28). The remaining two tRNA mutations were
shown to induce a significant change in the minimal free
energy (DG) of the respective secondary structures by RNA
fold (Fig. 2B). Moreover, the application of tRNAscan-SE
confirmed the effect of the two mutations on the functional
conservation of two tRNAs: m.12276G.A found in the H9
sample was predicted to have a less stable D-stem in
Human Molecular Genetics, 2013, Vol. 22, No. 2
229
Figure 2. The effect of pituitary oncocytoma mtDNA mutations on the CI assembly. (A) Phylogenetic alignments showing conservation of the mtDNA mutation
sites. Phylogenetically conserved amino acids are labelled with an asterisk, the numbering corresponds to the Homo sapiens amino acid sequence and the black arrows
indicate the mutation site. The m.3413G.A occurs at highly conserved region in MT-ND1, whereas m.3457G.A and m.15092G.A induce a change in highly
conserved amino acid of MT-ND1 and MT-CYB, respectively. (B) Secondary structure prediction indicates pathogenic effect of tRNA mutations found in
mtDNA of pituitary oncocytomas. RNA fold prediction of wild-type (WT) and mutated mitochondrial transfer RNAs is shown. Black arrows indicate the
mutated bases. The differences in minimal free energy (DG) between the WT and mutated tRNA sequences are reported. (C) IHC for nuclear CI subunits
(NDUFB6 and NDUFB8) (magnification ×100). Complex V subunit ATP5B was used as a positive control of the IHC reaction (magnification ×100). Two representative samples with WT mtDNA (H6 and H16) and seven samples carrying different mtDNA mutations (H1, H3, H19, H33, H37, H38 and H42) are shown.
comparison with the wild-type MT-TL2 (data not shown),
whereas the m.4427G.A mutation found in the H19 sample
was shown to completely prevent the formation of MT-TM.
The nine germ-line mtDNA mutations presented low PSIC
values, high amino acid variability reported in HmtDB, no
influence on minimal free energy in the case of tRNA
mutations and/or were as already known polymorphisms
(Supplementary Material, Table S1).
To determine whether mtDNA mutations may be predisposing or modifying events, DNA from the blood of the 20 pituitary
oncocytoma patients was analysed by fluorescent PCR (F-PCR)
for indels and by allele-specific locked nucleic acid quantitative
230
Human Molecular Genetics, 2013, Vol. 22, No. 2
real-time PCR (ASLNAqRT-PCR) for point mutations, which
allow detection of 2% (29) and 0.1% (30) mutant load, respectively. The assays were set up for all mutations detected by
Sanger sequencing in the tumour but not in corresponding
blood samples. No low-level heteroplasmy was found in the
blood samples upon F-PCR and ASLNAqRT-PCR analyses
(Supplementary Material, Figs S1 and S2). These findings
suggest that mtDNA mutations are not an element predisposing
to oncocytic transformation in pituitary adenomas, but most
likely a secondary event. The entire set of the complete mitochondrial genome sequences from tumour samples described
in the present study was deposited in the public resource
HmtDB (31), and the list of the HmtDB identifiers is presented
in Supplementary Material, Table S2.
Finally, haplogroup comparison with a set of mitochondrial
genomes of healthy Italian individuals available in HmtDB
did not highlight any correlation between a particular haplogroup and the development of oncocytic phenotype. This
finding was further supported by implementing our data with
previously published pituitary oncocytomas and non-oncocytic
pituitary adenomas (Supplementary Material, Fig. S3).
mtDNA mutations found in pituitary oncocytomas prevent
proper CI assembly
In order to stratify our samples for array comparative genomic
hybridization (aCGH) analysis, we next determined the functional effect of the mtDNA mutations on CI assembly.
The status of CI disruptive mutation was assigned to (i)
mtDNA changes which introduced a premature stop codon
in coding regions of CI subunit genes, (ii) previously described
mutations causing CI deficiency, (iii) novel variants found in a
highly conserved site, at which amino acid variability in the
general population as determined by HmtDB was under 1%
for coding mutations, or which substantially influenced DG
for tRNA mutations, and (iv) novel variants found in conserved D-loop functional regions (promoters PH1 and PL).
Furthermore, in order to validate our criteria for mutation
pathogenicity, immunohistochemistry (IHC) for nuclear
encoded NDUFB6 and NDUFB8 subunits of CI was performed on samples for which material was available, since
negative staining for at least one of the two subunits is indicative of CI disassembly (24). Complex V ATP5B subunit was
used as an indicator of mitochondrial abundance. All of the
tested samples showed to be uniformly positive for ATP5B,
indicating homogeneous oncocytic areas (Fig. 2C). On the
other hand, although samples without mtDNA mutations
showed positive NDUFB6 and NDUFB8 staining, the staining
of at least one of these subunits was negative or very weak in
samples carrying mutations in mtDNA genes (Fig. 2C).
Taking together, considering the bioinformatics predictions
and IHC data, 17 cases were defined as bearing CI disruptive
mutations (Supplementary Material, Table S3).
Occurrence of CI disruptive mtDNA mutations correlates
with lower genomic instability in pituitary adenomas
To determine the repertoire of nuclear numerical genomic
aberrations found in pituitary adenomas, 44 previously characterized pituitary adenomas (19) together with the 20 additional
new cases described in this study were subjected to PCR-based
DNA quality control. Eventually, 48 out of 64 pituitary adenomas were shown to have DNA of sufficient quality to be subjected to high-resolution aCGH (Supplementary Material,
Table S3). The aCGH analysis produced a data set which comprised 31 367 bacterial artificial chromosome (BAC) clones
with unambiguous mapping information according to the
August 2009 build (hg19) of the human genome (www.
ensembl.org). The complete aCGH data set has been deposited
and is available at http://rock.icr.ac.uk/collaborations/Macka
y/pituitary_adenoma_aCGH/. Upon exclusion of the data
from sex chromosomes, 2984 copy number aberrations (i.e.
1774 gains, 1149 losses and 61 amplifications) were detected
in the 48 pituitary adenomas. Pituitary adenomas revealed an
average of 9.56% (+SD 15.5) of the genome involved in
copy number aberrations.
Since mtDNA mutations leading to CI disassembly have
been shown to exhibit a suppressive effect on tumour progression (3,4,19), we hypothesized that they might be associated
with the relatively low genomic instability observed in pituitary
adenomas. In order to investigate the potential modifying function of CI disruptive mutations, we correlated mtDNA genotype
with the nuclear genetic features in our data set of pituitary
tumours for which both mtDNA sequence and aCGH data
were available (Supplementary Material, Table S3).
Tumours lacking CI disruptive mtDNA mutations presented
with an average of 12.73% (+SD 18.37%) of the genome
involved in copy number aberrations. On the other hand,
copy number aberrations involved an average of 3.78%
(+SD 4.32%) of the genome in pituitary adenomas harbouring CI disruptive mutations. In general, a greater number
and complexity of chromosomal copy number aberrations
(.5% of the genome involved) were found in tumours
without CI disruptive mutations (16 out of 31), whereas
those harbouring these mutations mostly displayed ‘flat’
genomes (15 out of 17) (two-tailed Fisher’s exact; P ¼ 0.011).
Since the ‘flat’ pattern of genomic profiles as defined by
aCGH is often associated with microsatellite instability
(MSI), we sought to determine whether pituitary adenomas
in our set displayed MSI. Analysis based on BAT26 and
BAT40 markers was performed, revealing that all pituitary adenomas analysed were microsatellite-stable (Supplementary
Material, Fig. S4). Another potential explanation for apparently ‘flat’ aCGH profiles is the multiclonality of the cases analysed, i.e. tumours composed of mosaics of cancers cells
that, in addition to the founder genetic aberrations, harbour
private mutations. Although this question was not directly
addressed in the samples included in this study, it is well
known that pituitary adenomas are clonal (32,33).
Moreover, to exclude that the differences in genomic profiles could be due to mutations in the TP53 tumour suppressor,
a gene whose inactivation is associated with higher levels of
chromosomal aberrations (34), we sequenced the whole
TP53 coding sequence in the 48 pituitary adenomas included
in aCGH analysis. Neither TP53 pathogenic mutations nor
loss of heterozygosity affecting the TP53 locus was found.
Last, since mutations in the oncogene GNAS have been
reported to occur in 40% of pituitary somatotropinomas (35)
but not in non-functioning pituitary tumours (36), we investigated whether its mutations might be associated with the
Human Molecular Genetics, 2013, Vol. 22, No. 2
differences in genomic profiles of pituitary adenomas. GNAS
mutational hotspot analysis revealed the presence of the
previously reported mutation c.2530C.T (p.844R.C) in
two cases (H40 and CNTRH18), none of which harboured
disruptive mtDNA mutations (Supplementary Material,
Fig. S5) (35).
Taken together, our findings demonstrate that, independent
of microsatellites stability, TP53 and GNAS genotype, stable
nuclear genomic architecture is a feature of pituitary adenomas bearing CI disruptive mutations.
231
two main branches may be distinguished as containing ‘flat’
(Branch 1) versus more complex (Branch 2) genomes. Both
oncocytic phenotype and the occurrence of CI disruptive
mtDNA mutations significantly correlated with the ‘flat’
genomes of Branch 1 (two-tailed Fisher’s test; P ¼ 0.039
and P ¼ 0.036, respectively), which additionally corroborated
our previous finding on the association of CI disruptive mutations and oncocytic phenotype with the less complex genomic
architecture of pituitary adenomas.
Low level of copy number aberrations is also a feature
of oncocytic phenotype in pituitary adenomas
Distinct copy number aberrations in pituitary
oncocytomas do not involve nuclear genes with a known
mitochondrial function
It is generally accepted that CI disruptive mutations are hallmarks of oncocytic phenotype. Indeed, this was confirmed in
our ‘expanded’ sample set, where the occurrence of CI disruptive mtDNA mutations significantly correlated with oncocytic
phenotype (two-tailed Fisher’s test; P ¼ 0.0002). In order to
investigate whether the high genomic stability observed in pituitary adenoma cases carrying CI disruptive mutations correlates with the occurrence of oncocytic phenotype, we analysed
the aCGH data by taking into consideration the oncocytic
versus non-oncocytic phenotype. To determine the repertoire
of nuclear numerical genomic aberrations found in pituitary
oncocytomas and the differences in the patterns of genomic
aberrations, we compared aCGH data relative to 31 pituitary
oncocytomas versus 17 non-oncocytic pituitary adenomas
from our sample set (Supplementary Material, Table S3).
Since CI disruptive mutations are known hallmarks of oncocytic phenotype, we hypothesized that the high genomic stability observed in pituitary adenoma cases carrying CI
disruptive mutations could also be translated to the occurrence
of oncocytic phenotype. In fact, the vast majority (77%) of pituitary oncocytomas displayed minimal levels of numerical
chromosomal aberrations (i.e. ‘flat’ genomes). A statistically
significant lower number of BAC clones harbouring gains,
losses and amplifications were found in pituitary oncocytomas
compared with non-oncocytic adenomas (two-tailed t-test;
P ¼ 0.03). In total, 1564 aberrations were detected in oncocytoma cases (50 per sample, +SEM 4.3), and on average corresponding to 5.02% of the genome involved in copy number
aberrations (range 0.74 – 33.92%). On the other hand, nononcocytic pituitary adenomas carried a total of 1420 copy
number aberrations (83 per sample, +SEM 15.5), which on
average corresponded to 17.84% of the genome involved in
chromosomal aberrations (range 0.72 – 81.77%).
In addition, copy number aberrations in non-oncocytic pituitary adenomas often involved large parts of the genome
(Fig. 3A); 11 out of 17 samples (65%) displayed .5% of
their genomes affected by a numerical chromosomal aberration, whereas only 7 out of 31 (23%) oncocytomas presented
with this level of copy number aberrations (Supplementary
Material, Table S3).
Furthermore, unsupervised hierarchical cluster analysis was
performed based on categorical aCGH data derived from 31
367 BACs, which revealed no consistent clustering based on
the type of the lesion (Supplementary Material, Fig. S6),
which indicates that there are no particular copy number aberrations associated with oncocytic phenotype. However, the
Although copy number aberrations have been extensively
studied in other pituitary tumours (37– 40), as well as in oncocytic tumours of other tissue types (5,41,42), limited information is available on chromosomal alterations in pituitary
oncocytomas, partly also due to the generally limited understanding of the molecular pathways leading to tumorigenesis
in the pituitary gland (32,33,43,44). In order to define the
qualitative differences in the repertoire of gene copy number
changes found in pituitary oncocytic and non-oncocytic adenomas, a multi-Fisher’s exact test adjusted for false discovery
was applied to compare genomic profiles of the 31 oncocytomas versus 17 non-oncocytic pituitary adenomas. Among the
regions significantly gained, lost or amplified between the
two groups (Fig. 3B), there was only one apparently
oncocytoma-associated copy number aberration: a loss of
4p16.3. It must be underlined that this region maps to a
locus reported to be uncommonly affected by germ-line
copy number polymorphisms, is peri-telomeric and prone to
aCGH artefacts; hence, this finding ought to be perceived as
hypothesis-generating. The loss within chromosomal band
4p16.3 was found in 10 out of 31 (32%) oncocytomas,
whereas no copy number losses affecting this locus were
present in pituitary adenomas lacking the oncocytic phenotype
(two-tailed Fisher’s test; P ¼ 8 × 1024). This copy number
gain maps to chr4:3394 –3640 kb and encompasses the
RGS12, HGFAC, DOK7 and LRPAP1 gene loci. An additional
genetic aberration of interest was the gain of 11p15.4, which
was significantly more prevalent in oncocytomas (35%) than
in non-oncocytic pituitary adenomas (6%) (two-tailed
Fisher’s exact test; P ¼ 0.03). The chromosomal band
11p15.4 also maps to a known germ-line copy number polymorphisms and is also peri-telomeric. This copy number
gain maps to chr11:10 425 – 10 684 kb and encompasses the
AMPD3, RNF141, LYVE1 and MRVI1 gene loci.
It has been suggested that genetic lesions in genes with mitochondrial function, but encoded by the nuclear genome, may be
called responsible for oncocytic transformation, representing
phenocopies of mtDNA mutations (27,45). It should be noted,
however, that no known mitochondrial function has been
ascribed to the genes found in genomic regions significantly
more frequently targeted by copy number aberrations in pituitary oncocytomas than in non-oncocytic pituitary adenomas.
To investigate whether oncocytoma samples presented losses
of nuclear genes encoding CI subunits and chaperones whose
mutations have been reported to cause CI disassembly (46,47),
the list of genes located in the regions of recurrent losses in
232
Human Molecular Genetics, 2013, Vol. 22, No. 2
Figure 3. Pituitary oncocytomas present with low copy number aberrations compared with their non-oncocytic counterparts. (A) Genome plots are shown representing the differences in chromosomal aberrations between oncocytomas and non-oncocytic pituitary adenomas. Minimal levels of numerical chromosomal
aberrations (i.e. ‘flat’ genomes) are shown in representative oncocytoma cases H25 and H28. In the middle panel, the two cases with the highest copy
number aberrations found in oncocytoma group are shown (H15 and H19). Chromosomes (X-axis) are presented with gains (green), losses (red) and no
change (black) for each BAC clone in genomic order. Log ratios are presented on the Y-axes. (B) Comparative analysis of the prevalence of chromosomal
gains, losses and amplifications in the pituitary oncocytomas (n ¼ 31) and their non-oncocytic counterparts (n ¼ 17). Top frequency plot shows gains and
losses in the oncocytoma group, middle frequency plot displays gains and losses in non-oncocytic pituitary adenomas and the bottom plot shows multi-Fisher’s
comparisons between the oncocytoma and non-oncocytic groups. The proportion of tumours in which each clone is gained (green bars) or lost (red bars) is
plotted (Y-axis) for each BAC clone according to genomic location (X-axis). Multi-Fisher’s exact tests were performed with CBS log2 ratios for each clone,
and those with an adjusted P-value ,0.05 are plotted (inverse log10, Y-axis) according to genomic location (X-axis).
oncocytoma samples was searched for NDUFAF1, NDUFAF2,
NDUFAF3, NDUFAF4, FOXRED1, NUBPL, ACAD9, ECSIT,
NDUFS1, NDUFS2, NDUFS3, NDUFS4, NDUFS6, NDUFS7,
NDUFS8, NDUFV1, NDUFV2, NDUFA1, NDUFA2,
NDUFA11, C2orf56, C6orf66, C8orf38 and C20orf7. None of
these genes was specifically lost in the oncocytoma group. For
instance, FOXRED1 was found lost in 12 out of 31 (39%) pituitary oncocytomas but also in 7 out of 15 (47%) non-oncocytic
pituitary tumours. Moreover, no gains were detected in
genomic loci enclosing known mitochondrial biogenesis
Human Molecular Genetics, 2013, Vol. 22, No. 2
factors such as peroxisome proliferator-activated receptor
gamma, co-activator 1 alpha (PPARGC1A; ENSG00000109819),
which would explain the mitochondrial hyperplasia observed
in oncocytomas.
Genomic profiling of pituitary adenomas revealed
PTTG1IP gain as a possible oncogenic event in pituitary
tumorigenesis
The molecular drivers of pituitary tumorigenesis are still mostly
unclear and their thorough identification is out of the scope of the
present work. Nonetheless, we investigated common copy
number changes in our entire sample set. Regions of recurrent
aberrations are shown in Supplementary Material, Figure S7
and summarized in Supplementary Material, Table S4. Recurrent aberrations found in .50% of cases included gains of
7p22.3 – q31.1, 8q21.2 – q21.3, 10q26.3, 16p13.3, 18q21.1,
19p13.3 – p12, 20q13.33 and 22q11.1, whereas the recurrent
losses present in .30% of cases mapped to 1p36.32– p36.31,
4q35.2, 9p11.1 – q21.11, 11q12.2, 11q23.3 – q25, 15q11.2 –
q13.3, 16q21, 19p13.2 and 22q11.23 – q12.1. No recurrent
amplification, other than regions of prevalent copy number
polymorphisms (e.g. 18q21.1), was identified.
Next, the list of genes located in recurrent regions
(i.e. .30%) was analysed in search of factors suggested
to be involved in pituitary tumorigenesis, namely GNAS,
PRKAR1A,
PIK3CA,
MEN-1,
CDKN1B,
CCND1,
GADD45G, AIP, HMGA2 and PTTG1 (32,43,48). The
results of the analysis are listed in Supplementary Material,
Table S5. Interestingly, PTTG1IP, a pituitary tumourtransforming 1 interacting protein (HGNC: 13524), located
at 21q22.3, was found gained in 17 out of 48 (35%) pituitary
adenomas, 7 of which were non-oncocytic adenomas (41%)
and 10 oncocytomas (32%). A further search revealed gains
of regions including the PTTG1 locus in additional two
cases, one oncocytoma and one non-oncocytic. The PTTG1
protein is involved in cell-cycle regulation, controlling the
segregation of sister chromatids during mitosis (49). It has
been found over-expressed in numerous tumours where it is
considered to induce cellular transformation by contributing
to genomic instability (49,50). Nonetheless, pituitary adenomas revealed to bear relatively few gene copy number aberrations as defined by aCGH, with an average of 9.56%
aberrations per sample.
DISCUSSION
In this work, we elucidated the role of CI disruptive mtDNA
mutations in pituitary tumorigenesis, and showed that
tumours carrying such mutations have significantly fewer
copy number aberrations than those without these mtDNA
lesions.
In addition to mutations in tumour suppressors and oncogenes, neoplastic development involves many additional
genetic events, which may either have negligible functional
significance under a set of selective pressure forces (i.e. passenger mutations) or modify tumour progression in a positive
or a negative manner. These factors may emerge somatically
in a stochastic fashion or exist prior to tumour initiation as
233
predisposing elements, providing a permissive ground for secondary transformation. The role of mtDNA mutations in this
context has yet to be elucidated. There is evidence that
mtDNA mutations may induce either pro- or anti-tumorigenic
effects (2).
It is not clear, however, whether they emerge after the initial
oncogenic hit as somatic events, or if they exist as low mutant
loads in germ-line tissue. In the latter case, they may be subsequently positively selected as it has been described in oncocytic neoplasms (9,24,25), either randomly as passengers in a
rapidly proliferating cell (51), or due to their positive effects
on tumour progression under selective pressure (52,53).
mtDNA mutations causing respiratory CI disassembly are of
particular interest, as they have been shown to induce block
of aggressive osteosarcoma tumour growth in mice and were
functionally implicated in driving the oncocytic transformation in human neoplasia (4,24,54). To investigate whether
such mutations are predisposing elements or modifiers, the
mitochondrial genome was screened in a set of pituitary oncocytomas. Our results confirmed the high occurrence of CI disruptive mutations in these tumours, revealing their somatic
nature as no low-level heteroplasmy was found in the corresponding blood tissue. Therefore, conversely from what has
been observed in single-oncocytoma cases of other tissues
(24,25), CI mutations in pituitary adenomas are not germ-line
predisposing elements (54).
Certain modifying roles of CI disruptive mtDNA mutations
have already been described, such as promotion of apoptosis
(3) and hampering of hypoxic adaptation (19). In line with
the data presented on the anti-tumorigenic effect of CI
mtDNA mutations, here we demonstrated that they are also
associated with tumours bearing higher levels of genomic stability (P ¼ 0.011), at least in pituitary adenomas. CI disruptive
mutations were found to significantly correlate with genomic
profiles where ,5% of the genome was affected by copy
number aberrations, regardless of the tumour size. Interestingly, lower levels of copy number aberrations were found in
tumours carrying CI disruptive mtDNA mutations despite
the frequent gains of the PTTG1IP gene, which has been associated with the deregulation of genomic architecture in pituitary tumours (49,50).
From the functional point of view, the CI bioenergetic
deficit already shown to be a hallmark of oncocytoma
appears to lead cancer cells to arrest in a senescence-like
status. Such deficit is considered to be the distinctive feature
of most oncocytic tumours regardless of the co-occurrence
of CI-specific mutations, although there is evidence that the
latter may be underestimated, especially since nuclear CI
mutations in these tumours have been seldom investigated
(2). In this frame, the significant association between oncocytic phenotype and low genomic instability may be envisioned
as an association between CI dysfunction and a more stable
genomic complement, although this warrants further investigation corroborated with biochemical data on mitochondrial respiratory competence of such tumours. Overall, however, the
absence of mitotic figures in our oncocytic samples, as well
as in other oncocytomas (6), along with low levels of Ki67,
strengthens the observation that oncocytes are stalled into a
low-replicative state. Hence, upon the occurrence of replication arrest, the chromosomal complement would not undergo
234
Human Molecular Genetics, 2013, Vol. 22, No. 2
further re-arrangements and accumulation of aberrations.
Since genomic instability is one of the major driving forces
of tumorigenesis (55), the here observed higher genomic stability associated with the CI disruptive mutations and with
the oncocytic phenotype may therefore be considered a consequence of an early occurrence of mtDNA disruptive mutations
driving the arrest of tumour progression.
We provide direct evidence that low levels of chromosomal
copy number aberrations are preferentially a feature of pituitary oncocytomas, whereas their non-oncocytic counterparts
displayed a significantly higher number of genomic aberrations. From the perspective of genomic instability, therefore,
pituitary oncocytomas may be considered analogous to
renal oncocytomas, which have been shown to harbour
more stable genomes than their non-oncocytic counterparts
(12– 14) and to carry CI disruptive mutations (10,11,27). Likewise, the genomic architecture of mitochondrion-rich breast
cancers has also been demonstrated to be more stable than
that of breast carcinomas lacking oncocytic features, although
no difference in clinical features was observed (41,56). At
variance with oncocytic tumours of other tissues, those of
the thyroid were previously shown to display higher levels
of chromosomal aberrations when compared with their nononcocytic counterparts (5,57). Furthermore, conversely from
oncocytomas of other tissues, oncocytic thyroid lesions not uncommonly have a more aggressive behaviour than their nononcocytic counterparts (58), but also have the lowest prevalence of CI mutations among all other types of oncocytic
tumours (54,59).
Noticeably, no common copy number aberration was
detected between genomic profiles from our sample set of
oncocytomas and the regions previously associated with the
development of oncocytic phenotype (5,14,41,60), suggesting
that either copy number aberrations are not to be held responsible for oncocytic transformation or that those driving oncocytoma tumorigenesis are site-specific. This finding
furthermore establishes CI disruptive mtDNA mutations as
the only common genetic hallmark in these lesions. Further
studies are necessary, such as parallel massive sequencing
analysis of nuclear genes with mitochondrial function to
understand other mechanisms involved in the accumulation
of mitochondria, especially in those oncocytic tumours
which do not harbour mtDNA mutations. Since there are
about 1000 genes in the human nuclear genome predicted to
have a mitochondrial function (61), it has been suggested
that defects in such genes could be called responsible for
oncocytic transformation, representing phenocopies of
mtDNA mutations (2,27). In fact, our group has recently identified distinct copy number aberrations in mitochondrion-rich
breast cancer, which contained diverse nuclear genes with a
known mitochondrial role (41). In the present study,
however, no annotated mitochondria-related functions were
found for the genes located in chromosomal bands 4p16.3
and 11p15.4, the only two regions significantly more frequently altered in oncocytomas than in non-onocytic pituitary adenomas. Nevertheless, loss-of-function mutations in DOK7
have been reported to cause congenital myasthenia, which is
characterized by muscle weakness due to defects in neuromuscular synapse formation (62– 64). It has recently been
pointed out that similar type of genetic lesions may lead to
the oncocytic phenotype in tumour environment, on the one
hand, and to the development of neuromuscular mitochondriopathies, on the other (2). In this context, DOK7 could be a
good candidate to investigate its potential involvement in
oncocytic transformation.
Taken together, we have confirmed the occurrence of
mtDNA mutations in pituitary oncocytomas, revealing their
somatic nature and their potential modifying role in the maintenance of higher genomic stability of pituitary adenomas.
In the context of anti-tumorigenic effect of CI disruptive
mtDNA mutations, together with previously described
effects on apoptotic (3) and hypoxic mechanisms (4,19), we
here demonstrate that the contribution of such mutations in determining a less aggressive tumour phenotype may, at least in
pituitary tumorigenesis, be inversely associated with the presence of chromosomal copy number aberrations, adding up to
the previously underlined features of an oncojanus gene (2,4).
MATERIALS AND METHODS
Sample collection
Samples of 20 sporadic pituitary adenomas (H1 – H20) and
corresponding blood tissues were collected at Regina Margherita Hospital, Turin, Italy. Moreover, for aCGH analysis,
DNA samples from previously described oncocytomas and
non-oncocytic pituitary adenomas were included (samples eligible for aCGH analysis are listed in Supplementary Material,
Table S3) (4,19). The cases were selected consecutively
during the period 2002 – 2009. All procedures were performed
according to internal board review protocols. All cases enrolled in the study were Italians.
Morphological analysis of pituitary adenomas
Oncocytic phenotype was evaluated independently by three
pathologists (I.M., G.C., C.C.). Samples showing oncocytic
phenotype were first selected on the basis of their histological
appearance on haematoxylin – eosin-stained sections and then
subjected to ultrastructural analysis. Specimens from pituitary
adenoma were fixed in 2.5% glutaraldheyde in 0.1% cacodylate buffer, post-fixed in 1% OsO4, dehydrated in ethanol
and embedded in Araldite. Thin sections (1 mm), stained in
uranyl acetate and lead citrate, were studied with a Philips
410 transmission electron microscope (Philips, Eindhoven,
Holland). Only cases which showed diffuse oncocytic
change (.80%) were designated as pituitary oncocytomas,
whereas the absence of mitochondrial defects was the criterion
for the selection of non-oncocytic pituitary adenomas.
IHC for CI subunits
Immunohistochemical analysis with antibodies against
NDUFB6 (MIM: 603322) and NDUFB8 (MIM: 602140)
subunit of CI and ATP5B (MIM: 102910) subunit of
Complex V (Invitrogen, Milan, Italy) was performed on
paraffin-embedded sections as previously described (24).
Human Molecular Genetics, 2013, Vol. 22, No. 2
DNA extraction
Extraction of DNA from formalin-fixed, paraffin-embedded
(FFPE) tissue was performed using the DNeasy Blood &
Tissue Kit (Qiagen, Milan, Italy), following the manufacturer’s instructions. All patient tumour samples were microdissected to ensure .70% of purity of neoplastic cells.
Microdissection was performed with a sterile needle under a
stereomicroscope (Olympus SZ61, Tokyo, Japan) from 6 – 10
consecutive 8 mm thick sections stained with nuclear fast red
as previously described (65). DNA concentration was measured with the PicoGreenw assay (Invitrogen, Monza, Italy).
The DNA quality was assessed by multiplex PCR following
the protocol developed by van Beers et al. (66). Blood
samples were extracted using the GeneElute Mammalian
Genomic DNA Miniprep Kit (Sigma-Aldrich, Milan, Italy)
and quantified by NanoDrop2000 (Thermo Scientific, France).
mtDNA sequencing
Whole mtDNA resequencing was performed with MitoAll
(Applied Biosystems, Foster City, CA, USA) as previously
described by our group (67), using the SeqScape v.2.5 software for electropherogram analysis (Applied Biosystems).
The pathogenic effect of mutations causing amino acid substitutions was predicted with PolyPhen1 server, which is based
on sequence comparison with homologous proteins. Profile
scores (PSIC) are generated for the allelic variants and represent the logarithmic ratio of the likelihood of a given amino
acid occurring at a particular site relative to the likelihood
of this amino acid occurring at any site (background frequency). Generally, PSIC score differences .2 indicate a
damaging effect, scores between 1.5 and 2 suggest that the
variant is possibly damaging, and scores ,1.5 indicate that
the variant is benign. Amino acid conservation analysis was
performed with Clustalw2 (http://www.ebi.ac.uk/Tools/msa/
clustalw2/), retrieving the FASTA sequences from http
://www.ncbi.nlm.nih.gov. The effect of the mutations on
tRNA secondary structure was estimated using the RNA fold
server (http://rna.tbi.univie.ac.at/cgi-bin/RNAfold.cgi) and
the tRNAscan-SE Search Server (http://lowelab.ucsc.edu/
tRNAscan-SE). RNA fold predicts the stability of the RNA,
considering the minimal free energy (DG), whereas tRNAscan
considers the cloverleaf structural constraints which do not
imply that the structure is the most energetically stable. Nucleotide and amino acid variability data for known mutations
were retrieved from the HmtDB database (www.hmtdb.
uniba.it), where the human mtDNA site-specific variability is
estimated and annotated based on 8300 and 1500 deposited
mitochondrial genomes from ‘normal’ individuals and
‘disease’ samples, respectively.
Haplogroup prediction
The haplogroup prediction has been performed by applying
the ‘classify your genome’ option available within HmtDB
(31). The comparison of each complete mtDNA sequence,
based on mutational events along the human mtDNA phylogeny, has been carried out by taking into account the ‘Copernican’ reassessment (68), and using as reference the
235
Reconstructed Sapiens Reference Sequence (RSRS) implemented in Phylotree (http://www.phylotree.org) (69), rather
than the revised Cambridge Reference Sequence (rCRS).
Fluorescent PCR
PCR reactions were performed using AmpliTaq Gold polymerase (Applied Biosystems) in the presence of a primer
with a 5′ fluorescent tag and analysed as previously described
(29). Primers were designed following the instructions for
preventing nuclear mitochondrial sequence (NumtS) coamplification. Primer sequences and amplification conditions
are detailed in Supplementary Material.
Allele-specific locked nucleic acid real-time PCR
PCR reactions were performed using SYBR Green chemistry
(Applied Biosystems) in 15 ml final reaction volume, with
0.4 mM of each primer, one of which was allele specific for
the mutation and labelled with an LNA base on its 3′ end
(Exiqon, Milan, Italy). Germ-line samples were run together
with known standards (stock, 1:10, 1:100, 1:1000 and
1:10 000), following previously described protocol (70).
Primers were designed following the instructions for preventing NumtS co-amplification. Primer sequences and amplification conditions are listed in Supplementary Material.
aCGH analysis
The aCGH platform was constructed at the Breakthrough
Breast Cancer Research Centre and comprises approximately
32 000 BAC clones tiled across the genome. This type of
BAC array platform has been shown to be suitable for the analysis of DNA samples extracted from FFPE tissue; it has a
resolution comparable with that of high-density oligonucleotide arrays (71,72). Labelling, hybridization, washes, image
acquisition and aCGH data analysis were carried out as previously described (65,73). Low-level gain was defined as a
circular binary segmentation (CBS)-smoothed log2 ratio of
between 0.12 and 0.45, corresponding to approximately
three to five copies of the locus, whereas gene amplification
was defined as having a log2 ratio of .0.45, corresponding
to more than five copies. Hierarchical cluster analysis was performed as previously described (41). Briefly, CBS-smoothed
log2 ratio values and categorical data were used for clustering,
considering Euclidean distance and using Ward’s clustering
algorithm. To identify statistically significant differences
between the genomic profiles of oncocytomas versus nononcocytic pituitary tumour samples, categorical aCGH data
were subjected to multi-Fisher’s exact test with adjustment
for multiple testing, using the step-down permutation procedure maxT, providing strong control of the family-wise type I
error rate. Statistical power to address the hypothesis that
oncocytomas versus non-oncocytic pituitary adenomas would
be distinct entities at the genomic level was calculated by applying Walters normal approximation, which showed that with
n ¼ 17 and a ¼ 0.05 we were able to detect differences
between 5 and 50% with the power of 0.783.
236
Human Molecular Genetics, 2013, Vol. 22, No. 2
aCGH data on copy number changes affecting sex chromosomes were excluded from the final analysis given that female
DNA was used as a reference for both male and female cases.
MSI evaluation
The presence of MSI was evaluated by the amplification of the
mononucleotide repeat markers BAT26 and BAT40 as previously described (74). Briefly, samples were amplified using
AmpliTaq polymerase (Applied Biosystems) and primers labelled with a fluorescent tag on their 5′ end. The products
were loaded onto 3730 DNA Analyzer (Applied Biosystems)
together with GeneScan 500 Liz Size Standard (Applied Biosystems) and analysed using GeneMapper v.3.5 (Applied Biosystems). Tumours were considered MSI-high (MSI-H) when
both BAT26 and BAT40 displayed MSI, defined by multiple
extra bands above the background. As a control, a blood
DNA sample either from the corresponding patient or from a
healthy female donor was run in parallel. The microsatellite
marker BAT26 has been shown to be reliable in determining
the MSI-H phenotype without the requirement for matching
normal DNA given that it is ‘quasi-monomorphic’ in DNA
from normal individuals or microsatellite-stable tumours
compared with MSI-H cancers (74,75).
TP53 and GNAS mutational screening
The entire TP53 coding region (transcript ENST00000269305,
exons 2 – 11) was amplified using K2G(2×) polymerase
(Kapabiosystems, Rome, Italy) in 10 ml final reaction
volume with 0.2 mM of each primer and subsequent thermocyler conditions: 3 s hold at 948C; 10 cycles at 948C for 10 s,
62– 578C for 10 s (touch-down protocol with D-0.58C at
each cycle), 728C for 1 s; 25 cycles at 948C for 10 s, 578C
for 10 s, 728C for 1 s; 30 s hold at 728C. Primers are listed
in Supplementary Material. GNAS mutation hotspots (transcript ENST00000371100, exons 7 – 9) were analysed using
K2G(2×) polymerase (Kapabiosystems) in 10 ml final reaction volume, with 0.2 mM of each primer (35) and following
thermocycler conditions: 3 s hold at 948C; 30 cycles at 948C
for 10 s, 628C for 10 s, 728C for 1 s. Direct sequencing of
the PCR amplicons was performed using BigDye v1.1
(Applied Biosystems), following manufacturers’ instructions
on a 3730 DNA Analyser (Applied Biosystems). Data were
analysed using Sequencher v.4.7 (Gene Codes, Ann Arbor,
MI, USA).
SUPPLEMENTARY MATERIAL
Supplementary Material is available at HMG online.
Conflict of Interest statement. None declared.
FUNDING
This work was supported by the Associazione Italiana Ricerca
sul Cancro - AIRC (IG8810); by the Italian Ministry of University MIUR - Futuro in Ricerca 2008 (J31J10000040001)
and Fondazione Umberto Veronesi - DISCO TRIP project to
G.G.; by the Breakthrough Breast Cancer Research Centre
(A.M., M.B.K.L. and J.S.R.-F.) and by ‘Fondazione della
Cassa di Risparmio di Puglia’ to M.A.; I.K. is supported by
a triennial fellowship ‘Borromeo’ from AIRC. A.M.,
M.B.K.L. and J.S.R.-F. acknowledge the NHS support to the
NIHR biomedical research centre.
REFERENCES
1. Shen, L., Fang, H., Chen, T., He, J., Zhang, M., Wei, X., Xin, Y., Jiang,
Y., Ding, Z., Ji, J. et al. (2010) Evaluating mitochondrial DNA in cancer
occurrence and development. Ann. NY Acad. Sci., 1201, 26–33.
2. Iommarini, L., Calvaruso, M.A., Kurelac, I., Gasparre, G. and Porcelli,
A.M. (2012) Complex I impairment in mitochondrial diseases and cancer:
parallel roads leading to different outcomes. Int. J. Biochem. Cell Biol.
[Epub ahead of print, available online 2 March 2012].
3. Park, J.S., Sharma, L.K., Li, H., Xiang, R., Holstein, D., Wu, J.,
Lechleiter, J., Naylor, S.L., Deng, J.J., Lu, J. et al. (2009) A
heteroplasmic, not homoplasmic, mitochondrial DNA mutation promotes
tumorigenesis via alteration in reactive oxygen species generation and
apoptosis. Hum. Mol. Genet., 18, 1578– 1589.
4. Gasparre, G., Kurelac, I., Capristo, M., Iommarini, L., Ghelli, A.,
Ceccarelli, C., Nicoletti, G., Nanni, P., De Giovanni, C., Scotlandi, K.
et al. (2011) A mutation threshold distinguishes the antitumorigenic
effects of the mitochondrial gene MTND1, an oncojanus function. Cancer
Res., 71, 6220–6229.
5. Tallini, G. (1998) Oncocytic tumours. Virchows Arch., 433, 5– 12.
6. Gasparre, G., Romeo, G., Rugolo, M. and Porcelli, A.M. (2011) Learning
from oncocytic tumors: why choose inefficient mitochondria? Biochim.
Biophys. Acta., 1807, 633– 642.
7. Savagner, F., Franc, B., Guyetant, S., Rodien, P., Reynier, P. and
Malthiery, Y. (2001) Defective mitochondrial ATP synthesis in oxyphilic
thyroid tumors. J. Clin. Endocrinol. Metab., 86, 4920–4925.
8. Lu, J., Sharma, L.K. and Bai, Y. (2009) Implications of mitochondrial
DNA mutations and mitochondrial dysfunction in tumorigenesis. Cell
Res., 19, 802–815.
9. Pereira, L., Soares, P., Maximo, V. and Samuels, D.C. (2012) Somatic
mitochondrial DNA mutations in cancer escape purifying selection and
high pathogenicity mutations lead to the oncocytic phenotype:
pathogenicity analysis of reported somatic mtDNA mutations in tumors.
BMC Cancer, 12, 53.
10. Gasparre, G., Hervouet, E., de Laplanche, E., Demont, J., Pennisi, L.F.,
Colombel, M., Mege-Lechevallier, F., Scoazec, J.Y., Bonora, E., Smeets,
R. et al. (2008) Clonal expansion of mutated mitochondrial DNA is
associated with tumor formation and complex I deficiency in the benign
renal oncocytoma. Hum. Mol. Genet., 17, 986– 995.
11. Mayr, J.A., Meierhofer, D., Zimmermann, F., Feichtinger, R., Kogler, C.,
Ratschek, M., Schmeller, N., Sperl, W. and Kofler, B. (2008) Loss of
complex I due to mitochondrial DNA mutations in renal oncocytoma.
Clin. Cancer Res., 14, 2270– 2275.
12. Presti, J.C. Jr, Moch, H., Reuter, V.E., Cordon-Cardo, C. and Waldman,
F.M. (1996) Renal cell carcinoma genetic analysis by comparative
genomic hybridization and restriction fragment length polymorphism
analysis. J. Urol., 156, 281– 285.
13. Presti, J.C. Jr, Moch, H., Reuter, V.E., Huynh, D. and Waldman, F.M.
(1996) Comparative genomic hybridization for genetic analysis of renal
oncocytomas. Genes Chromosomes Cancer, 17, 199–204.
14. Mertz, K.D., Tchinda, J., Kufer, R., Moller, P., Rubin, M.A., Moch, H.
and Perner, S. (2006) Cytogenetic alterations in renal tumors.
Applications for comparative genomic hybridization and fluorescence in
situ hybridization. Urologe A, 45, 316– 318, 320–322.
15. Saeger, W., Ludecke, D.K., Buchfelder, M., Fahlbusch, R., Quabbe, H.J.
and Petersenn, S. (2007) Pathohistological classification of pituitary
tumors: 10 years of experience with the German Pituitary Tumor Registry.
Eur. J. Endocrinol., 156, 203– 216.
16. Karavitaki, N. (2012) Prevalence and incidence of pituitary adenomas.
Ann. Endocrinol. (Paris), 73, 79–80.
17. Beckers, A., Rostomyan, L. and Daly, A.F. (2012) Overview of genetic
testing in patients with pituitary adenomas. Ann. Endocrinol. (Paris), 73,
62– 64.
Human Molecular Genetics, 2013, Vol. 22, No. 2
18. Niveiro, M., Aranda, F.I., Paya, A., Boix, E., Peiro, G. and Pico, A. (2004)
Oncocytic transformation in pituitary adenomas: immunohistochemical
analyses of 65 cases. Arch. Pathol. Lab. Med., 128, 776–780.
19. Porcelli, A.M., Ghelli, A., Ceccarelli, C., Lang, M., Cenacchi, G.,
Capristo, M., Pennisi, L.F., Morra, I., Ciccarelli, E., Melcarne, A. et al.
(2010) The genetic and metabolic signature of oncocytic transformation
implicates HIF1alpha destabilization. Hum. Mol. Genet., 19, 1019–1032.
20. Dasgupta, S., Koch, R., Westra, W.H., Califano, J.A., Ha, P.K., Sidransky,
D. and Koch, W.M. (2010) Mitochondrial DNA mutation in normal
margins and tumors of recurrent head and neck squamous cell carcinoma
patients. Cancer Prev. Res. (Phila.), 3, 1205– 1211.
21. Hung, W.Y., Wu, C.W., Yin, P.H., Chang, C.J., Li, A.F., Chi, C.W., Wei,
Y.H. and Lee, H.C. (2010) Somatic mutations in mitochondrial genome
and their potential roles in the progression of human gastric cancer.
Biochim. Biophys. Acta, 1800, 264–270.
22. Tseng, L.M., Yin, P.H., Yang, C.W., Tsai, Y.F., Hsu, C.Y., Chi, C.W. and
Lee, H.C. (2011) Somatic mutations of the mitochondrial genome in
human breast cancers. Genes Chromosomes Cancer, 50, 800–811.
23. He, Y., Wu, J., Dressman, D.C., Iacobuzio-Donahue, C., Markowitz, S.D.,
Velculescu, V.E., Diaz, L.A. Jr, Kinzler, K.W., Vogelstein, B. and
Papadopoulos, N. (2010) Heteroplasmic mitochondrial DNA mutations in
normal and tumour cells. Nature, 464, 610– 614.
24. Gasparre, G., Iommarini, L., Porcelli, A.M., Lang, M., Ferri, G.G.,
Kurelac, I., Zuntini, R., Mariani, E., Pennisi, L.F., Pasquini, E. et al.
(2009) An inherited mitochondrial DNA disruptive mutation shifts to
homoplasmy in oncocytic tumor cells. Hum. Mutat., 30, 391–396.
25. Pradella, L.M., Zuntini, R., Magini, P., Ceccarelli, C., Neri, I., Cerasoli,
S., Graziano, C., Gasparre, G. and Turchetti, D. (2011) Two distinct
thyroid tumours in a patient with Cowden syndrome carrying both a
10q23 and a mitochondrial DNA germline deletion. J. Med. Genet., 48,
779–782.
26. Alston, C.L., Morak, M., Reid, C., Hargreaves, I.P., Pope, S.A., Land,
J.M., Heales, S.J., Horvath, R., Mundy, H. and Taylor, R.W. (2010) A
novel mitochondrial MTND5 frameshift mutation causing isolated
complex I deficiency, renal failure and myopathy. Neuromuscul. Disord.,
20, 131–135.
27. Zimmermann, F.A., Mayr, J.A., Feichtinger, R., Neureiter, D., Lechner,
R., Koegler, C., Ratschek, M., Rusmir, H., Sargsyan, K., Sperl, W. et al.
(2011) Respiratory chain complex I is a mitochondrial tumor suppressor
of oncocytic tumors. Front. Biosci. (Elite Ed.), 3, 315–325.
28. Chalmers, R.M., Lamont, P.J., Nelson, I., Ellison, D.W., Thomas, N.H.,
Harding, A.E. and Hammans, S.R. (1997) A mitochondrial DNA
tRNA(Val) point mutation associated with adult-onset Leigh syndrome.
Neurology, 49, 589–592.
29. Kurelac, I., Lang, M., Zuntini, R., Calabrese, C., Simone, D., Vicario, S.,
Santamaria, M., Attimonelli, M., Romeo, G. and Gasparre, G. (2012)
Searching for a needle in the haystack: comparing six methods to evaluate
heteroplasmy in difficult sequence context. Biotechnol. Adv., 30,
363–371.
30. Morandi, L., de Biase, D., Visani, M., Cesari, V., De Maglio, G.,
Pizzolitto, S., Pession, A. and Tallini, G. (2012) Allele specific locked
nucleic acid quantitative PCR (ASLNAqPCR): an accurate and
cost-effective assay to diagnose and quantify KRAS and BRAF mutation.
PLoS One, 7, e36084.
31. Rubino, F., Piredda, R., Calabrese, F.M., Simone, D., Lang, M.,
Calabrese, C., Petruzzella, V., Tommaseo-Ponzetta, M., Gasparre, G. and
Attimonelli, M. (2012) HmtDB, a genomic resource for
mitochondrion-based human variability studies. Nucleic Acids Res., 40,
D1150– D1159.
32. Daly, A.F., Tichomirowa, M.A. and Beckers, A. (2009) The epidemiology
and genetics of pituitary adenomas. Best Pract. Res. Clin. Endocrinol.
Metab., 23, 543– 554.
33. Farrell, W.E. (2006) Pituitary tumours: findings from whole genome
analyses. Endocr. Relat. Cancer, 13, 707– 716.
34. Aylon, Y. and Oren, M. (2011) p53: guardian of ploidy. Mol. Oncol., 5,
315–323.
35. Freda, P.U., Chung, W.K., Matsuoka, N., Walsh, J.E., Kanibir, M.N.,
Kleinman, G., Wang, Y., Bruce, J.N. and Post, K.D. (2007) Analysis of
GNAS mutations in 60 growth hormone secreting pituitary tumors:
correlation with clinical and pathological characteristics and surgical
outcome based on highly sensitive GH and IGF-I criteria for remission.
Pituitary, 10, 275– 282.
237
36. Ruggeri, R.M., Santarpia, L., Curto, L., Torre, M.L., Galatioto, M.,
Galatioto, S., Trimarchi, F. and Cannavo, S. (2008) Non-functioning
pituitary adenomas infrequently harbor G-protein gene mutations.
J. Endocrinol. Invest., 31, 946 –949.
37. Harada, K., Nishizaki, T., Ozaki, S., Kubota, H., Harada, K., Okamura, T.,
Ito, H. and Sasaki, K. (1999) Cytogenetic alterations in pituitary
adenomas detected by comparative genomic hybridization. Cancer Genet.
Cytogenet., 112, 38–41.
38. Metzger, A.K., Mohapatra, G., Minn, Y.A., Bollen, A.W., Lamborn, K.,
Waldman, F.M., Wilson, C.B. and Feuerstein, B.G. (1999) Multiple
genetic aberrations including evidence of chromosome 11q13
rearrangement detected in pituitary adenomas by comparative genomic
hybridization. J. Neurosurg., 90, 306–314.
39. Rickert, C.H., Dockhorn-Dworniczak, B., Busch, G., Moskopp, D.,
Albert, F.K., Rama, B. and Paulus, W. (2001) Increased chromosomal
imbalances in recurrent pituitary adenomas. Acta. Neuropathol., 102,
615– 620.
40. Trautmann, K., Thakker, R.V., Ellison, D.W., Ibrahim, A., Lees, P.D.,
Harding, B., Fischer, C., Popp, S., Bartram, C.R. and Jauch, A. (2001)
Chromosomal aberrations in sporadic pituitary tumors. Int. J. Cancer, 91,
809– 814.
41. Geyer, F.C., de Biase, D., Lambros, M.B., Ragazzi, M., Lopez-Garcia,
M.A., Natrajan, R., Mackay, A., Kurelac, I., Gasparre, G., Ashworth, A.
et al. (2012) Genomic profiling of mitochondrion-rich breast carcinoma:
chromosomal changes may be relevant for mitochondria accumulation
and tumour biology. Breast Cancer Res. Treat., 132, 15–28.
42. Neuhaus, C., Dijkhuizen, T., van den Berg, E., Storkel, S., Stockle, M.,
Mensch, B., Huber, C. and Decker, H.J. (1997) Involvement of the
chromosomal region 11q13 in renal oncocytoma: case report and literature
review. Cancer Genet. Cytogenet., 94, 95– 98.
43. Xekouki, P., Azevedo, M. and Stratakis, C.A. (2010) Anterior pituitary
adenomas: inherited syndromes, novel genes and molecular pathways.
Expert Rev. Endocrinol. Metab., 5, 697–709.
44. Lania, A. and Spada, A. (2009) G-protein and signalling in pituitary
tumours. Horm. Res., 71 (Suppl. 2), 95– 100.
45. Maximo, V., Botelho, T., Capela, J., Soares, P., Lima, J., Taveira, A.,
Amaro, T., Barbosa, A.P., Preto, A., Harach, H.R. et al. (2005) Somatic
and germline mutation in GRIM-19, a dual function gene involved in
mitochondrial metabolism and cell death, is linked to mitochondrion-rich
(Hurthle cell) tumours of the thyroid. Br. J. Cancer, 92, 1892–1898.
46. Sugiana, C., Pagliarini, D.J., McKenzie, M., Kirby, D.M., Salemi, R.,
Abu-Amero, K.K., Dahl, H.H., Hutchison, W.M., Vascotto, K.A., Smith,
S.M. et al. (2008) Mutation of C20orf7 disrupts complex I assembly and
causes lethal neonatal mitochondrial disease. Am. J. Hum. Genet., 83,
468– 478.
47. McKenzie, M., Tucker, E.J., Compton, A.G., Lazarou, M., George, C.,
Thorburn, D.R. and Ryan, M.T. (2011) Mutations in the gene encoding
C8orf38 block complex I assembly by inhibiting production of the
mitochondria-encoded subunit ND1. J. Mol. Biol., 414, 413– 426.
48. Qian, Z.R., Asa, S.L., Siomi, H., Siomi, M.C., Yoshimoto, K., Yamada,
S., Wang, E.L., Rahman, M.M., Inoue, H., Itakura, M. et al. (2009)
Overexpression of HMGA2 relates to reduction of the let-7 and its
relationship to clinicopathological features in pituitary adenomas. Mod.
Pathol., 22, 431– 441.
49. Smith, V.E., Franklyn, J.A. and McCabe, C.J. (2010) Pituitary
tumor-transforming gene and its binding factor in endocrine cancer.
Expert Rev. Mol. Med., 12, e38.
50. Quereda, V. and Malumbres, M. (2009) Cell cycle control of pituitary
development and disease. J. Mol. Endocrinol., 42, 75– 86.
51. Coller, H.A., Khrapko, K., Bodyak, N.D., Nekhaeva, E., Herrero-Jimenez,
P. and Thilly, W.G. (2001) High frequency of homoplasmic mitochondrial
DNA mutations in human tumors can be explained without selection. Nat.
Genet., 28, 147–150.
52. Sharma, L.K., Fang, H., Liu, J., Vartak, R., Deng, J. and Bai, Y. (2012)
Mitochondrial respiratory complex I dysfunction promotes tumorigenesis
through ROS alteration and AKT activation. Hum. Mol. Genet., 20, 4605–
4616.
53. Zhou, S., Kachhap, S., Sun, W., Wu, G., Chuang, A., Poeta, L., Grumbine,
L., Mithani, S.K., Chatterjee, A., Koch, W. et al. (2007) Frequency and
phenotypic implications of mitochondrial DNA mutations in human
squamous cell cancers of the head and neck. Proc. Natl Acad. Sci. USA,
104, 7540–7545.
238
Human Molecular Genetics, 2013, Vol. 22, No. 2
54. Gasparre, G., Porcelli, A.M., Bonora, E., Pennisi, L.F., Toller, M.,
Iommarini, L., Ghelli, A., Moretti, M., Betts, C.M., Martinelli, G.N. et al.
(2007) Disruptive mitochondrial DNA mutations in complex I subunits
are markers of oncocytic phenotype in thyroid tumors. Proc. Natl Acad.
Sci. USA, 104, 9001–9006.
55. Shen, Z. (2011) Genomic instability and cancer: an introduction. J. Mol.
Cell. Biol., 3, 1– 3.
56. Ragazzi, M., de Biase, D., Betts, C.M., Farnedi, A., Ramadan, S.S.,
Tallini, G., Reis-Filho, J.S. and Eusebi, V. (2011) Oncocytic carcinoma of
the breast: frequency, morphology and follow-up. Hum. Pathol., 42,
166– 175.
57. Wada, N., Duh, Q.Y., Miura, D., Brunaud, L., Wong, M.G. and Clark,
O.H. (2002) Chromosomal aberrations by comparative genomic
hybridization in Hurthle cell thyroid carcinomas are associated with tumor
recurrence. J. Clin. Endocrinol. Metab., 87, 4595–4601.
58. Mete, O. and Asa, S.L. (2010) Oncocytes, oxyphils, Hurthle, and
Askanazy cells: morphological and molecular features of oncocytic
thyroid nodules. Endocr. Pathol., 21, 16– 24.
59. Gasparre, G., Bonora, E., Tallini, G. and Romeo, G. (2010) Molecular
features of thyroid oncocytic tumors. Mol. Cell. Endocrinol., 321, 67–76.
60. Di Palma, S., Lambros, M.B., Savage, K., Jones, C., Mackay, A., Dexter,
T., Iravani, M., Fenwick, K., Ashworth, A. and Reis-Filho, J.S. (2007)
Oncocytic change in pleomorphic adenoma: molecular evidence in
support of an origin in neoplastic cells. J. Clin. Pathol., 60, 492–499.
61. Calvo, S., Jain, M., Xie, X., Sheth, S.A., Chang, B., Goldberger, O.A.,
Spinazzola, A., Zeviani, M., Carr, S.A. and Mootha, V.K. (2006)
Systematic identification of human mitochondrial disease genes through
integrative genomics. Nat. Genet., 38, 576– 582.
62. Beeson, D., Higuchi, O., Palace, J., Cossins, J., Spearman, H., Maxwell,
S., Newsom-Davis, J., Burke, G., Fawcett, P., Motomura, M. et al. (2006)
Dok-7 mutations underlie a neuromuscular junction synaptopathy.
Science, 313, 1975– 1978.
63. Selcen, D., Milone, M., Shen, X.M., Harper, C.M., Stans, A.A., Wieben,
E.D. and Engel, A.G. (2008) Dok-7 myasthenia: phenotypic and
molecular genetic studies in 16 patients. Ann. Neurol., 64, 71–87.
64. Vogt, J., Morgan, N.V., Marton, T., Maxwell, S., Harrison, B.J., Beeson,
D. and Maher, E.R. (2009) Germline mutation in DOK7 associated with
fetal akinesia deformation sequence. J. Med. Genet., 46, 338–340.
65. Hernandez, L., Wilkerson, P.M., Lambros, M.B., Campion-Flora, A.,
Rodrigues, D.N., Gauthier, A., Cabral, C., Pawar, V., Mackay, A.,
A’Hern, R. et al. (2012) Genomic and mutational profiling of ductal
carcinomas in situ and matched adjacent invasive breast cancers reveals
intra-tumour genetic heterogeneity and clonal selection. J. Pathol., 227,
42–52.
66. van Beers, E.H., Joosse, S.A., Ligtenberg, M.J., Fles, R., Hogervorst, F.B.,
Verhoef, S. and Nederlof, P.M. (2006) A multiplex PCR predictor for
aCGH success of FFPE samples. Br. J. Cancer, 94, 333– 337.
67. Bonora, E., Porcelli, A.M., Gasparre, G., Biondi, A., Ghelli, A., Carelli,
V., Baracca, A., Tallini, G., Martinuzzi, A., Lenaz, G. et al. (2006)
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
Defective oxidative phosphorylation in thyroid oncocytic carcinoma is
associated with pathogenic mitochondrial DNA mutations affecting
complexes I and III. Cancer Res., 66, 6087–6096.
Behar, D.M., van Oven, M., Rosset, S., Metspalu, M., Loogvali, E.L.,
Silva, N.M., Kivisild, T., Torroni, A. and Villems, R. (2012) A
‘Copernican’ reassessment of the human mitochondrial DNA tree from its
root. Am. J. Hum. Genet., 90, 675–684.
van Oven, M. and Kayser, M. (2009) Updated comprehensive
phylogenetic tree of global human mitochondrial DNA variation. Hum.
Mutat., 30, E386– E394.
Strand, H., Ingebretsen, O.C. and Nilssen, O. (2008) Real-time detection
and quantification of mitochondrial mutations with oligonucleotide
primers containing locked nucleic acid. Clin. Chim. Acta, 390, 126–133.
Tan, D.S., Lambros, M.B., Natrajan, R. and Reis-Filho, J.S. (2007)
Getting it right: designing microarray (and not ‘microawry’) comparative
genomic hybridization studies for cancer research. Lab. Invest., 87,
737–754.
Coe, B.P., Ylstra, B., Carvalho, B., Meijer, G.A., Macaulay, C. and Lam,
W.L. (2007) Resolving the resolution of array CGH. Genomics, 89,
647–653.
Duprez, R., Wilkerson, P.M., Lacroix-Triki, M., Lambros, M.B., Mackay,
A., Hern, R.A., Gauthier, A., Pawar, V., Colombo, P.E., Daley, F. et al.
(2012) Immunophenotypic and genomic characterization of papillary
carcinomas of the breast. J. Pathol., 226, 427–441.
Xicola, R.M., Llor, X., Pons, E., Castells, A., Alenda, C., Pinol, V.,
Andreu, M., Castellvi-Bel, S., Paya, A., Jover, R. et al. (2007)
Performance of different microsatellite marker panels for detection of
mismatch repair-deficient colorectal tumors. J. Natl Cancer Inst., 99,
244–252.
Boland, C.R., Thibodeau, S.N., Hamilton, S.R., Sidransky, D., Eshleman,
J.R., Burt, R.W., Meltzer, S.J., Rodriguez-Bigas, M.A., Fodde, R.,
Ranzani, G.N. et al. (1998) A National Cancer Institute Workshop on
Microsatellite Instability for cancer detection and familial predisposition:
development of international criteria for the determination of
microsatellite instability in colorectal cancer. Cancer Res., 58,
5248– 5257.
Taylor, R.W., Barron, M.J., Borthwick, G.M., Gospel, A., Chinnery, P.F.,
Samuels, D.C., Taylor, G.A., Plusa, S.M., Needham, S.J., Greaves, L.C.
et al. (2003) Mitochondrial DNA mutations in human colonic crypt stem
cells. J. Clin. Invest., 112, 1351–1360.
Cardaioli, E., Da Pozzo, P., Radi, E., Dotti, M.T. and Federico, A. (2005)
A novel heteroplasmic tRNA(Leu(CUN)) mtDNA point mutation
associated with chronic progressive external ophthalmoplegia. Biochem.
Biophys. Res. Commun., 327, 675–678.
Menotti, F., Brega, A., Diegoli, M., Grasso, M., Modena, M.G. and
Arbustini, E. (2004) A novel mtDNA point mutation in tRNA(Val) is
associated with hypertrophic cardiomyopathy and MELAS. Ital. Heart J.,
5, 460– 465.