Human Reproduction, Vol.26, No.5 pp. 1241– 1251, 2011
Advanced Access publication on February 18, 2011 doi:10.1093/humrep/der018
ORIGINAL ARTICLE Reproductive genetics
Variable expression of the Fragile X
Mental Retardation 1 (FMR1) gene in
patients with premature ovarian failure
syndrome is not dependent on number
of (CGG)n triplets in exon 1
J. Schuettler 1,3,†, Z. Peng 1,†, J. Zimmer 1, P. Sinn 2, C. von Hagens 1,
T. Strowitzki 1, and P.H. Vogt 1,*
1
Department of Gynecology Endocrinology, University Women Hospital, Voßstrasse 9, D-69115 Heidelberg, Germany 2Section Gynecology
Pathology, University Women Hospital, Heidelberg, Germany
3
Present address: Department of Gynecology & Obstetrics, University Hospital Mannheim, Mannheim, Germany
*Correspondence address. Tel: +49-6221-567918– 567916; Fax: +49-6221-5633710; E-mail: [email protected]
Submitted on September 16, 2010; resubmitted on November 15, 2010; accepted on January 12, 2011
background: Increased expression of the Fragile X Mental Retardation 1 (FMR1) gene in blood cells has been claimed to be associated
with variable (CGG)n triplet numbers in the 5′ untranslated region of this gene. Increased CGG triplet numbers, including that of the so-called
premutation range (n ¼ 55 –200), were shown to have a risk of ,26% to impair ovarian reserve leading to primary ovarian insufficiency and
premature ovarian failure (POF).
methods: DNA and RNA samples were isolated from 74 patients with idiopathic POF to evaluate quantitatively the expression of FMR1
in leukocytes and CGG triplet number on FMR1 gene alleles. mRNA levels were normalized and compared with those of control women.
Expression of the encoded protein (FMRP) was analysed by immunohistochemistry on ovarian biopsy tissue sections.
results: A large variance of the FMR1 transcript level was found in the leukocyte RNA samples, but only in patients with POF, and this
variability did not correlate to variance of CGG triplet numbers found on both FMR1 alleles (19 , n . 90). During normal folliculogenesis,
FMRP is predominantly expressed in granulosa cells.
conclusions: Our data suggest that FMR1 expression during human folliculogenesis is probably a quantitative trait. Proper function of
FMRP in granulosa cells seems to depend on an optimal transcript level. All women with CGG triplet numbers outside the range associated
with normal folliculogenesis (26 , n . 34) are therefore expected to have a relaxed FMR1 transcription control. FMR1 transcript levels in
leukocytes might therefore be diagnostic for altered FMRP levels in granulosa cells, which will affect the process of folliculogenesis.
Key words: Fragile X Mental Retardation 1 gene expression / (CGG) triplet numbers / premature ovarian failure risk / ovarian reserve
Introduction
Premature ovarian failure (POF) is a hypergonadotropic ovarian
deficiency with primary or secondary amenorrhoea affecting 1% of
women before the age of 40 years (Coulam et al., 1986; Goswami
and Conway, 2005; Fassnacht et al., 2006). Between 4 and 31% of
the patients with POF have a familial accumulation of the POF syndrome supporting a dominant genetic aetiology (Coulam et al.,
1983; Vegetti et al., 1998; van Kasteren et al., 1999; Laml et al.,
†
2002; Dixit et al., 2010). Therefore, besides autoimmune diseases
or infections, genetic factors are thought to be mainly responsible
for POF development. With view to its original observation and
description in 1942 (Albright et al., 1942), POF is now also coined
Primary Ovarian Insufficiency (POI) in order to remind the fact that
50% of the cases show a variable and unpredictable function of the
ovarian reserve (Perlhoff and Schneeberg, 1957; Nelson et al., 1994;
Welt, 2008). Furthermore, 5– 10% of women with POF conceive
and deliver a child after they have received the diagnosis (Gleicher
These authors contributed equally to the results of this paper.
& The Author 2011. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved.
For Permissions, please email: [email protected]
1242
et al., 2009; Nelson, 2009; Cooper et al., 2010). Therefore, it has
been proposed that the POF/POI condition might only be the most
severe form of a more common premature ovarian senescence
(POS) syndrome found in about 10% of women in different populations (Gleicher and Barad, 2010).
Although human candidate genes of POF are localized on all
chromosomes, they seem to be concentrated on the X chromosome
(Krauss et al., 1987; Powell et al., 1994; Sala et al., 1997; Zinn et al.,
1998; Marozzi et al., 2000; Prueitt et al., 2000; Schlessinger et al.,
2002; Rao et al., 2005; Rizzolio et al., 2006; OMIM database ID
#300510; #300511; #300604; #311360; www.ncbi.nlm.nih.gov/
omim). Three regions with clusters of POF candidate genes could
be distinguished by investigating distinct X-autosomal chromosome
translocations (Davison et al., 2000; Toniolo and Rizzolio, 2006).
Accordingly, these regions were designated as POF-1 (Xq26 –q28),
POF-2 (Xq13 –q22) and POF-3 (Xp11.2– p22.1).
In POF-1, the Fragile X Mental Retardation 1 (FMR1) gene is the most
prominent candidate gene. Dysfunction of FMR1 in men causes the
Fragile X syndrome (FXS: OMIM database: ID #300624)—a severe
neurological disorder. It is mainly caused by the amplification of an
unstable CGG triplet in the 5′ untranslated region (UTR) of the
gene’s first exon beyond a number of 200, thereby repressing the
gene’s transcriptional and translational activity (Oostra and Willemsen,
2003). Amplification of the CGG triplet number above the normal
range (n ¼ 29 –30; Fu et al., 1991) towards the so-called premutation
status (n ¼ 55–200) was found to be a high risk factor (1:20) for
women to get POF syndrome (Conway et al., 1999; Sullivan et al.,
2005; Allen et al., 2007; Wittenberger et al., 2007). Accordingly,
females from families with FXS diagnosed with a carrier status for
the CGG premutation on one allele have a high risk (16 –26%) for
getting POF (Allingham-Hawkins et al., 1999). In contrast to this,
the CGG premutation of the FMR1 gene was found only rarely
(1:300) in the general female population (Cronister et al., 2005).
Men with a single FMR1 gene and a CGG permutation allele in exon
1 are suffering from a variable reduction of FMRP protein expression
in their lymphocytes which causes moderate and variable cognitive
deficit (Tassone et al., 2000a). Later in life, this phenotype develops
into one of the most common single-gene, late-onset neurodegenerative disorder, occurring in one of 3000 men and designated as Fragile
X-associated tremor/ataxia syndrome (FXTAS: OMIM ID: #300623).
However, CGG triplet numbers below the premutation status and
above the normal value, the so-called intermediate stage (n ¼ 31 –
54), were also found to be associated with POI/POF (Bretherick
et al., 2005; Streuli et al., 2009). Some authors claim that this association
is not general but probably dependent on the pattern of single nucleotide
mutations (AGG) in the CGG repeat and on the X-chromosomal inactivation status (Tarleton et al., 2002; Bodega et al., 2006). Indeed, in two
POF-CGG association studies, it was reported that CGG repeat
numbers between 31 and 54 could be found with similar frequencies
as in POF patients also in the female controls (Chatterjee et al., 2009;
Bennett et al., 2010). However, in both studies, these female controls
were not analysed for the putative occurrence of POS (primary
ovarian senescence). Recently, it was found that also CGG numbers
below n ¼ 26 are found predominantly only in POF patients. This
suggests a general association of POF and POS with a functional drawback of the FMR1 gene encoded protein (FMRP) in human folliculogenesis. It is therefore believed that it probably controls quantitatively the
Schuettler et al.
woman’s ovarian reserve (Gleicher et al., 2009, 2010). If this holds
true, the function of FMRP during human folliculogenesis should be considered genetically as a sensitive quantitative trait (Gilad et al., 2008).
Indeed, FMRP is a functionally important RNA-binding protein
located in the cellular RISC (RNA-Induced Silencing Complex) and
controlling the level of translation of multiple transcripts (Schaeffer
et al., 2003; Jin et al., 2004). That means any relaxation of the
gene’s tight transcriptional and translational control mediated, e.g.
by the CGG triplet numbers in exon 1 might then become a causative
agent for POF or POS (see also Gleicher and Barad, 2010).
In this study, we analysed comparatively the quantitative expression
level of the FMR1 gene in the leukocyte RNA samples of patients with
a clinically diagnosed idiopathic POF syndrome and of control samples,
i.e. women with a normal menstrual history (called NMH group). In
parallel, we analysed the number of CGG triplets on both FMR1
gene alleles of the same POF samples in order to reveal their putative
influence on the variability of FMR1 expression. Moreover, we reveal
the expression of FMRP (FMR1 encoded protein) in granulosa cells of
different ovarian tissue sections.
Materials and Methods
Patient sampling
This study was approved by the local ethical committee of the University
of Heidelberg. Accordingly, all blood samples were taken from the individuals of the POF and NMH group only after medical indication (POF group)
and prior written informed consent. The same procedure was followed
when receiving an aliquot of the ovarian biopsy of some women diagnosed
by histology to have a normal folliculogenesis. Biopsies were performed
because the women had a clinical diagnosis of follicular cyst formation.
A total of 120 patients enrolling consecutively in our endocrinological outpatient clinic were selected for the presence of ‘idiopathic’ POF syndrome
with aid of an extensive clinical questionnaire according to Fassnacht et al.
(2006). All POF patients with either a putative history of autoimmune diseases or an iatrogenic background owing to earlier chemo- or radiation
therapies were excluded. Also, patients with primary amenorrhoea and
those with karyotype abnormalities causing Turner Syndrome (45,X0)
were excluded. The remaining ‘idiopathic’ POF patient subgroup includes
74 individuals. With the same questionnaire, 42 healthy women being
over the age of 40 years with still regular menstruation or a physiologically
conditioned menopause were selected as our control group (NMH).
NMH4 was chosen as a calibrator sample in all quantitative RT – PCR
assays because her FMR1 expression was not only near the median value
of the range of FMR1 expression values found in all control samples, but
also because she had given birth to two children and was still undergoing
regular menstruation cycles at the age of 50 years. For NMH4, we therefore
can exclude not only POF/POI but also any other POS phenotype, as discussed by Gleicher and Barad (2010).
DNA/RNA isolation and quantitative
RT– PCR analyses
DNA samples were prepared as described by Fassnacht et al. (2006) from
20 ml EDTA blood samples; for practical reasons, we could isolate in parallel
the RNA samples from only 74 of the 95 POF blood samples. All cDNA
samples were synthesized from total RNA after oligo dT priming with the
SuperScriptw First-Strand Synthesis System of Invitrogen (Invitrogen
GmbH, Darmstadt, Germany; cat. no: 11904-018) and the M-MLV
Reverse Transcriptase RNase H Minus, Point Mutant of Promega
1243
Fragile X gene expression and premature ovarian failure
(Promega GmbH, Mannheim, Germany; cat. no. M 3683). For quantitative
analysis of FMR1 mRNA expression, we used the Light Cycler 1.5 instrument
of Roche (Roche 04 484 495 001) with the Light Cycler DNA Fast Start Kit
and the SYBR Green I detection kit (Roche Diagnostics GmbH, Mannheim,
Germany; cat. no. 2239264). For utilization, programming and analysis of
the expression data we used Light Cycler Software Version 3, Light Cycler
Front Version 3.5.17, Light Cycler Run Version 5.3.2 and Light Cycler
Data Analysis Version 3.5.28. The calculation of relative expression values
was made by RelQuant (Relative Quantification Version 1.01 Software
from Roche) as follows: (i) for each sample, three expression assays for
FMR1 and the hypoxanthine phosphoribosyltransferase (HPRT) housekeeping gene were run in parallel to calculate the mean values always with respect
to the internal HPRT expression values; (ii) for further normalization, we
related all HPRT normalized FMR1 expression values to that of NMH4
running as expression reference also in each experiment. The following
FMR1 exon sequences were used as primers for these quantitative PCR
assays: FMR1: FMR1forLC: 5′ -gat gat ggt caa gga atg ggt c-3′ located
1380 – 1411 nts from ATG and FMR1revLC: 5′ -tcg gga gtg atc gtc gtt
tcc-3′ located 1554 – 1574 nts from ATG in FMR1 cDNA (Gene ID
ENSG00000102081 extracted from: www.ensembl.org database). For
expression analyses of the HPRT gene, we used the following primer pairs:
h-HPRT-1-FP: 5′ -gac ctg ctg gat tac atc aaa gc-3′ located 229 – 251 nts
from ATG and h-HPRT-1-RP: 5′ -gga tta tac tgc ctg acc aag ga-3′ located
424 – 446 from ATG in HPRT cDNA sequence (Ensembl Gene ID
ENSG00000165704).
For analysis of the CGG repeat length in exon 1 of the FMR1 gene by
PCR and subsequent analysis of this sequence region with the ALF
express sequence automat (Pharmacia Biotech, Freiburg, Germany;
Type Amersham 1050), we performed all PCR assays in a volume of
10 ml. The PCR mixture contained: 2.5 pM of each primer: (forward
0-1147 and reverse primer 0-1148; for primer sequences, see Baechner
et al., 1993), 200 ml of each dNTP, 0.4 U Taq polymerase; 1× PCR
buffer (all from: Qiagen GmbH, Hilden Germany) and 40 ng of genomic
DNA. Conditions for PCR were as follows: 5 min at 948C for the first
denaturation step; 35 cycles of amplification with a time – temperature
profile of 15 s. at 948C, 2.5 min at 688C, 3 min + 4 s/cycle at 688C,
3 min at 688C; with additional 4 min at 688C in the last cycle. Both
primers were labelled with the fluorescent Cy5 dye (from TIB
MOLBIOL Syntheselabor GmbH, Berlin, Germany; cat. no. Alexa 488).
A 1.5-ml aliquot of PCR mix was resuspended in 7 ml of loading solution
(Formamide, bromophenol blue) containing 100 and 300 bp internal
markers. All samples, after denaturation at 958C for 5 min, were analysed
on 6% denaturing polyacrylamide gel with 7 M urea. A 50 – 500 size marker
labelled with Cy5 dye was used for calculation of CGG triplet numbers.
Electrophoresis was carried out in 0.6× Tris-borate-EDTA (TBE) Buffer
at 1500 V/min. The helium-neon laser operated at a wavelength of
700 nm and a power value of 2.5 mW. Allele sizes and peak areas of
fluorescent products were analysed with aid of the Fragment Manager
software (Pharmacia GmbH).
In order to exclude the presence of large CGG triplet numbers not
resolvable by the ALF sequence automat but only by Southern blot analyses (Southern, 1975), we also performed blot hybridization experiments
of all POF DNA samples with the a-32P-dCTP radioactively labelled p2
probe containing FMR1 exon 1 with the CGG repeat as described by
Stoyanova and Oostra (2004). After EcoRI and EagI genomic DNA restriction the methylated and unmethylated FMR1 exon 1 allele located on the
inactive and active X chromosome, respectively, can be distinguished by
two different fragment lengths (2.5 and 5.2 kb). An increase of the
CGG triplet number towards the premutation range (n ¼ 55 – 200) and
longer (n , 200) become then visible by an increase of one or both
EcoRI – EagI fragments after autoradiography. Reaction conditions: 3 – 5 U
of restriction enzyme (EcoRI from, Thermo Fisher Scientific, Ulm,
Germany; EagI from New England Biolabs GmbH, Frankfurt, Germany)
were incubated with the DNA samples at 378C overnight and then
loaded onto 1% 1× TBE gel. Electrophoresis was carried out at 100 V
in the first hour, then at the 60 V overnight. After depurination in
0.25 M HCl, denaturation with 1.5 M NaCl + 0.5 M NaOH and neutralization in 1 M Tris Cl + 2 M NaCl, the gels were blotted on nylon membranes soaked in 20×SSC (3 M NaCl, 0.3 M sodium citrate) overnight.
Before hybridization with Church buffer (0.5 M Na phosphate pH 7.0 +
7% SDS + 5 mM EDTA), salmon sperm DNA (Roche; cat. no.
1022364001) was added to the hybridization cocktail and the membrane
incubated overnight with the 32P labelled probe pP2.
Statistics
To calculate the asymptotic significance of the gene expression values in
the two analysed groups, we used the Mann– Whitney U-test. A P-value
of ,0.05 was considered to indicate a statistically significant difference
(Wilcoxon, 1945).
Immunohistochemical experiments
For immunohistochemical experiments with a monoclonal FMRP antiserum (MAB1C3: clone 1C3-1a; 1:200 v/v; Euromedex, Souffelweyersheim, France), we collected ovarian tissue samples from females who
were diagnosed first clinically with follicular cyst development. After
laparoscopy, however, only normal follicles in their gonads were identified
by detailed histological inspection of their ovarian tissue sections.
Aliquots of ovarian tissue biopsies were fixed in buffered formaldehyde
and embedded in paraffin blocks. For immunohistochemical experiments,
4 mm tissue sections were prepared and mounted on SuperFrost Plus glass
slides (Neolab, Heidelberg, Germany). After overnight storage at 378C,
sections were de-paraffinized with xylene (3 × 10 min) and re-hydrated
with an ethanol – water series. After washing for 3 min in permeabilization
buffer (0.1 M Tris, 0.1 M NaCl, 0.1% Triton X-100; pH 7.4), the slides
were incubated with 0.2 M boric acid, pH 7, at 608C overnight. Nonspecific cross-reactions were blocked with 20% acetic acid at 48C for
15 s. After pre-incubation of all slides in 3% goat serum (60 min at
room temperature), incubation with the mouse monoclonal FMRP antiserum was performed overnight at 48C in 1% goat serum. For staining
the slides after the antiserum reaction, we used the standard APAAP protocol (Dako Cytomation, Glostrup, Denmark). Subsequently, slides
(APAAP treated) were stained with HistoMark red (Dunn KPL.; Gaithersburg, Maryland, USA) and counterstained with Gill’s hematoxylin II (VWR
International, Bruchsal, Germany) before mounting in Immu-Mount
(Shandon, Pittsburgh, PA, USA).
Results
Comparative quantitative analyses of FMR1
gene expression in leukocytes
In our FMR1 expression study, we explored the gene transcript level in
the leukocyte RNA samples of 74 POF patients with an ‘idiopathic’
history and compared it with that of 42 control samples (NMH
samples). Four of them (POF5; POF6, POF10; POF11) had a family
history of POF. FMR1 mRNA levels estimated in the leukocytes of our
74 POF patients displayed a surprisingly large variability ranging
between 54 and 754% with regard to the control NMH4 as internal calibrator, i.e. setting its FMR1 expression value to ‘100%’ (Table I). In the
42 RNA samples of the NMH control group, FMR1 transcript levels also
displayed some variance (Table I) although within a significantly lower
range with an arithmetic mean value of 112% and a median of 95%.
1244
Schuettler et al.
Table I FMR1 expression rates in leukocytes of patients with POF and NMH controls with reference to cellular HPRT
expression and to the NMH4 calibrator (51.00).
POF code
% Exp. rate
POF code
% Exp. rate
POF code
% Exp. rate
POF code
% Exp. rate
.............................................................................................................................................................................................
(A) FMR1 expression rates in leukocytes of POF patients with reference to NMH4
4
0.87
40
0.93
76
0.8
100
1.95
5
0.95
42
2.03
77
0.8
101
1.89
6
1.41
43
0.54
79
0.73
102
0.55
7
1.13
45
2.05
80
1.3
103
0.94
8
1.81
46
1.33
81
0.89
104
0.93
10
1.83
54
7.54
82
1.46
105
0.66
11
1.28
56
1.32
83
1.01
106
0.82
13
0.74
57
1.23
84
0.65
109
1.04
14
3.08
59
1.95
85
1.07
111
2.71
18
0.63
60
0.87
86
1.57
113
0.59
22
0.79
61
4.87
87
0.96
114
1.12
24
1.09
63
3.53
88
2.57
115
2.46
26
0.72
64
1.25
89
1.00
116
1.00
27
0.89
65
1.15
90
1.67
117
2.02
28
1.1
67
0.59
93
2.57
118
1.75
31
0.81
69
3.46
94
1.00
119
0.92
33
1.39
70
1.07
96
1.02
120
0.97
34
1.49
71
3.04
97
1.78
37
0.9
74
0.99
99
1.06
NMH code
% Exp. rate
NMH code
% Exp. rate
NMH code
% Exp. rate
NMH code
% Exp. rate
.............................................................................................................................................................................................
.............................................................................................................................................................................................
(B) FMR1 expression rates in leukocytes of NMH controls with reference to NMH4
1
1.99
13
1.64
24
2.03
35
0.94
2
0.76
14
2.22
25
1.04
36
0.85
3
1.08
15
1.81
26
1.14
37
0.54
4
1.00
16
0.91
27
2.2
38
0.42
5
0.84
17
0.81
28
1.03
39
1.45
6
1.68
18
0.96
29
0.93
40
0.97
7
1.44
19
0.62
30
1.66
41
0.88
8
1.02
20
0.97
31
0.84
42
0.47
10
0.87
21
0.94
32
0.72
11
1.52
22
0.96
33
0.77
12
1.15
23
1.45
34
0.78
This could be best demonstrated by comparative Box and Whisker plots
(Fig. 1). However, the obvious trend towards remarkable high arithmetic mean and median values (Fig. 1), as well as a SD twice as much
as in the control group was too large to reach statistical significance
(Mann –Whitney U-test: P ¼ 0.091).
Variable FMR1 expression in POF patients is
not associated with CGG triplet numbers in
FMR1 exon 1
It has been reported that increased FMR1 expression levels in POF
patients correlate with the number of CGG triplets in the gene
5′ UTR. We, therefore, estimated in the genomic DNA samples of
the same 74 POF patients the number of CGG triplets in FMR1
exon 1 on both gene alleles. More than half of them (45 samples) displayed a number of CGG triplets in the normal range on both alleles
(Table II). However, despite this their level of FMR1 expression ranged
from 0.54x (i.e. 54%; POF43) to 7.54x (754%; POF54) when compared with the expression of the NMH4 calibrator sample (¼100%)
as described above. This was more variable than found for the two
POF samples, POF99 and POF101, with one allele being in the
so-called premutation range (n ¼ 80–90), the second allele being
below the normal value (n ¼ 20).
Since FMR1 expression in POF patients with CGG values below or
above the normal range (26 , n . 34) might also depend on age
and on the distinct allelic heterozygous or homozygous CGG
1245
Fragile X gene expression and premature ovarian failure
Figure 1 Box and Whisker plot of FMR1 expression profiles in the
leukocytes RNA samples of our POF (n ¼ 74) and NMH (n ¼ 42)
population. The values are shown in a log(10) scale of the proportional expression to calibrator NMH4, as described in text.
Their quartiles Q1 and Q3 as well as their maximum and minimum
values and their displayed medians point to a large variation in
FMR1 leukocyte expression only in the POF samples.
number combination, i.e. whether one or both alleles are above or
below the normal range (Gleicher et al., 2010), we divide our FMR1
expression data of our POF patients in subgroups. The main group
consists of samples with ‘normal (n)’ CGG numbers on both gene
alleles (‘n/n’; 45 samples). Additionally, we distinguished subgroups
with heterozygous (het) allelic constitutions, i.e. one ‘n(ormal)’
allele together with one ‘l(ow)’ or ‘h(igh)’ allele (‘het n/l’ and ‘het
n/h’) and subgroups with homozygous (hom) alleles outside the
normal range, i.e. ‘l/l’ ‘l/h’ and ‘h/h’ (Table II). We subdivided
these CGG genotype subgroups in the table according to their
level of FMR1 expression and into two age groups to reveal some
possible age dependency. For this we chose an arbitrary borderline
of 33 years. FMR1 expression level was called ‘normal’ when it was
still in a range of 20% above or below the transcript level of NMH4
our calibrator control; ‘low’ and ‘high’ when it was below and above
this range, respectively (Table III).
Although we found a strong imbalance in patient numbers with
regard to age, with 17 below and 57 above 33 years, we noticed
more patients in the ‘high’ FMR1 expression subgroup in the older
females. With respect to the total number of patients, this was not
found in the ‘normal’ and ‘low’ FMR1 expression groups. If this
outcome should reflect the well-known tendency towards a reduction
in ovarian reserve with age, it would suggest that higher expression of
FMR1 transcripts in leukocytes can indeed be diagnostic for this, as
first suggested by Gleicher et al. (2010). We also observed a slight tendency of reduced FMR1 expression levels in the group of patients with
‘hom l/n’ and ‘hom l/l’ CGG genotypes when compared with those
with n/n CGG genotype. There were no ‘hom h/h’ CGG alleles in
our POF patient collection (Table III). The number of patients in the
hom l/n or hom l/I group is however too low to draw some conclusions from this observation.
We then selected 15 POF patients, some of which had a high
(.1.3; POF6; POF8; POF14; POF33; POF42; POF46; POF71), a
low (,0.8; POF13) or normal range of FMR1 mRNA level (POF4;
POF5; POF7; POF11; POF31; POF40; POF57) for identification of a
putative mosaic status of the active CGG triplet allele. Such variability
only becomes visible by appropriate Southern blots as described in the
Material and Methods section. We could not identify an extraordinary
CpG methylation status within the FMR1 CpG island (Fig. 2). Moreover, we also could not identify any mosaic status of the CGG
triplet number on the active X chromosome as shown in Fig. 2 for
the genomic DNA sample with a known CGG premutation allele
(‘pre-mut’); nor did we reveal any long CGG repeat blocks, which
cannot be resolved by the ALF sequence automat but only by
Southern blots.
We therefore conclude that the observed variability of FMR1
expression is not caused by some individual gross variation of the
CpG methylation status in the gene promoter CpG island domain.
FMRP is expressed after birth in female
germline predominantly in granulosa cells
A general involvement of FMR1 expression in controlling ovarian
ageing would suggest that the encoded protein (FMRP) is expressed
in granulosa cells controlling ovarian maturation after birth (Binelli
and Murphy, 2009; Gougeon, 2010). We therefore analysed the
expression of FMRP in normal human ovarian tissue sections from
three girls, aged 5 months, 10 years and 19 years, by immunohistochemistry. We found no FMRP expression in the oocytes in these
ovarian tissue sections, but a predominant expression of FMRP in
the granulosa cells surrounding each oocyte (Fig. 3). The observed
decrease in the number of follicles between the samples was in
normal range of the gradient of apoptotic follicle degeneration with
age (Krysko et al., 2008; Ozgur and Oktay, 2008).
Discussion
Variability of FMR1 expression in POF
patients is independent of CGG triplet
numbers
The FMR1 gene is transcribed and translated in many tissues, including
leukocytes. FMR1 was therefore a suitable candidate gene to develop
a reliable diagnostic expression assay using leukocyte RNA samples,
possibly helping to identify any dysfunction of this prominent POF candidate gene during ovarian maturation in POF patients. We thereby
rationalized that any visible variation in the FMR1 expression profile
found in the leukocytes of patients with POF syndrome but not in
the leukocytes of women with a normal menstruation history (NMH
group) would point to the presence of some genomic FMR1 sequence
mutations causing this visible change of the FMR1 expression profile.
One class of such genomic FMR1 mutations associated with POF
has already been identified, namely the number of CGG triplets in
FMR1 exon 1 beyond or below the normal range (26 , n . 34; Gleicher et al., 2009, 2010). Indeed, the most frequent genetic cause of
POF is an increase in the number of CGG triplet towards the so-called
premutation range (n ¼ 55–200; Kenneson et al., 1997; AllinghamHawkins et al., 1999; Sherman, 2000; Allen et al., 2004; Tassone
et al., 2007; Wittenberger et al., 2007; Gleicher and Barad, 2010).
However, also CGG triplet numbers below the pre-mutation range
and outside the normal range are probably associated with distinct
1246
Schuettler et al.
Table II Genomic DNA samples from patients with
POF listed with their relative FMR1 expression levels
and age analysed for CGG triplet numbers in FMR1
exon 1 on both X alleles (#/#).
POF code
#/# CGG
Allele n/h/l
% Exp. rate
Age
........................................................................................
Table II Continued
POF code
#/# CGG
Allele n/h/l
% Exp. rate
Age
........................................................................................
84
30/35
het n/h
0.65
33
85
24/31
het l/n
1.07
40
86
30/42
het n/h
1.57
39
4
28/30
n/n
0.87
26
87
30/30
n/n
0.96
35
5
29/30
n/n
0.95
33
88
20/20
n/n
2.57
34
6
29/31
n/n
1.41
33
89
20/34
het l/n
1.00
31
7
30/31
n/n
1.13
35
90
20/20
hom l/l
1.67
36
8
30/31
n/n
1.81
21
93
30/30
n/n
2.57
38
10
20/33
het l/n
1.83
36
94
28/29
n/n
1.00
43
11
30/32
n/n
1.28
38
96
30/31
n/n
1.02
29
13
29/29
n/n
0.74
22
97
30/31
n/n
1.78
33
14
30/35
het n/h
3.08
41
99
20/80–90
hom l/h
1.06
26
18
20/35
hom l/h
0.63
36
100
30/31
n/n
1.95
35
22
32/41
het n/h
0.79
19
101
20/80–90
hom l/h
1.89
33
24
30/32
n/n
1.09
40
102
31/31
n/n
0.55
39
26
30/34
n/n
0.72
36
103
23/31
n/n
0.94
34
27
21/30
het l/n
0.89
24
105
30/32
n/n
0.93
27
28
30/36
het n/h
1.10
37
106
30/30
n/n
0.66
30
31
30/30
n/n
0.81
33
107
29/30
n/n
0.82
36
33
30/30
n/n
1.39
34
109
30/32
n/n
1.04
34
34
24/30
het l/n
1.49
35
111
29/31
n/n
2.71
38
37
30/31
n/n
0.90
40
113
31/31
n/n
0.59
37
40
30/30
n/n
0.93
35
114
31/31
n/n
1.12
34
42
30/30
n/n
2.06
41
115
30/32
n/n
2.46
42
43
30/32
n/n
0.54
35
116
29/30
n/n
1.00
42
45
22/32
het l/n
2.05
39
117
22/24
hom l/l
2.02
34
46
30/31
n/n
1.33
36
118
20/54
hom l/h
1.75
34
54
30/30
n/n
7.54
27
119
29/30
n/n
0.92
23
56
23/29
het l/n
1.32
34
120
30/31
n/n
0.97
47
57
30/31
n/n
1.23
28
59
60
21/32
20/30
het l/n
het l/n
1.95
0.87
34
38
61
30/31
n/n
4.87
29
63
64
23/32
22/30
het l/n
het l/n
3.53
1.25
39
28
65
30/33
n/n
1.15
33
67
23/29
het l/n
0.59
39
69
30/31
n/n
3.46
35
70
33/37
het n/h
1.07
33
71
31/31
n/n
3.04
37
74
23/30
het l/n
0.99
38
76
19/19
hom l/l
0.80
37
77
20/22
hom l/l
0.80
38
79
30/30
n/n
0.73
24
80
23/41
hom l/h
1.3
32
81
28/30
n/n
0.89
30
82
22/24
hom l/l
1.46
36
83
30/31
n/n
1.01
36
Continued
These were subdivided into the group with a homozygous ‘n(ormal/n(ormal)’ CGG
genotype (45 samples), a group with heterozygous (het) allelic constitutions, i.e. one
‘n(ormal)’ allele together with one ‘l(ow)’ or ‘h(igh)’ allele (‘het n/l’ and ‘het n/h’),
and in a group with homozygous (hom) alleles outside of the normal range, i.e. ‘l/l’
‘l/h’ and ‘h/h’ (grey-toned differentially), according to Gleicher et al. (2010).
impairments of normal folliculogenesis. These were recently summarized under the term ‘POI’ (Gleicher et al., 2009, 2010; Nelson, 2009).
In this paper, we now present strong experimental evidence that
not only CGG triplet amplification might cause a change of the cellular
FMR1 expression level and therefore POF, but also other genomic
alterations are obviously regulating the FMR1 transcript level. We
found a high FMR1 expression variance in leukocytes when the
number of both CGG alleles was in the normal range, but only in
the leukocytes of POF patients (Tables I and II). Interestingly, above
the age of 33 years, the number of women with a high FMR1
expression level seems to be increased (Table III).
A non-linear relationship between the number of CGG triplets and
the associated FMR1 transcript levels was found previously in patients
with POF although only for a small sample size (Garcia-Alegrı́a et al.,
1247
Fragile X gene expression and premature ovarian failure
Table III Comparison of FMR1 expression levels in the three CGG genotypes as defined in Table II.
FMR1 exp. level
n/n CGG alleles
het n/l 1 n/h CGG alleles
hom l/n and l/l CGG alleles
Low
Low
................................................................. ........................................ .....................................
Low
Normal
(<20%)
High
Normal
(<20%)
High
Normal
(<20%)
High
.............................................................................................................................................................................................
POF pat. ,33 years
79;106
Total number of patients in each 2
group
4; 57;81;
96;105;119
8, 54; 61
22
27; 89
64
6
3
1
2
1
0
99
80
1
1
.............................................................................................................................................................................................
POF pat. .33 years
13; 36;
5; 7; 24; 31; 37;
6; 11; 33; 42;
67; 84 28; 60; 70; 10; 14; 34;
43;102; 113 40; 65; 83; 87; 94; 46; 69; 71; 93; 97;
74; 85
45; 56; 59;
103; 107; 109;
100; 111; 115
63; 86
114; 116; 120
Total number of patients in each 5
group
16
13
2
5
8
18
76; 77;
117; 118
82; 88;
90;101
1
4
4
They were called ‘normal’ when they were in a range of 20% above or below the transcript level of NMH4 our calibrator control; ‘low’ and ‘high’ when it was below and above this
range, respectively. Additionally the different FMR1 expression groups were divided into two age groups choosing an arbitrary borderline of 33 years.
Figure 2 Southern blot hybridization of the two FMR1 gene alleles in a number of patients with POF, using the P32 labelled p2 probe of the
FMR1 gene. The female patients always contain two X chromosomes. One gene copy is usually completely methylated in the CpG island region of
exon 1 according to its location on the inactive X chromosome. It is marked by a p2 hybridization signal to a FMR1 genomic fragment of 5.2 kb
(marked on the left). The p2 hybridization signal marking the FMR1 genomic fragment of 2.8 kb (marked on the left) represents the FMR1 allele
with an unmethylated CpG island region thus located on the active X chromosome: its fragment length confirms the presence of CGG repeat
lengths in the normal range, as listed in Table II for these patient samples. No further genomic fragments could be identified in this Southern blot,
indicating the additional presence of long CGG repeat tracts that could not be visualized by the sequence analyser. Moreover, no mosaic status of
the CpG methylation pattern was found in the genomic DNA samples of these patients with POF as shown, compared with the control sample of
a man with a known CGG premutation allele (see pre-mut at the left). Neither did we find complete CpG methylation of the second FMR1 allele,
as shown by the absence of the 2.8-kb hybridization fragment in the control sample of a man with a CGG full mutation allele (see full-mut at the
left).
2007; Tejada et al., 2008). We could confirm this non-linear relationship within our study population of 74 POF patients.
Putative agents causing FMR1 transcript
variation other than CGG triplet numbers
In the literature, three further putative agents causing a variable FMR1
expression are discussed: (i) choice of the transcriptional start site
(coined ‘TSS’) for FMR1, (ii) DNA methylation level in the CGG
triplet block or its adjacent upstream CpG island because being part
of the FMR1 promoter, (iii) compaction of the chromatin around
the different FMR1 gene alleles controlled by specific post-translational
histone modifications called ‘chromatin histone code’.
(i) Expansion of the number of CGG triplets beyond normal range
creates more than one TSS within a 50 nucleotide (nt)
1248
Schuettler et al.
promoter domain this was found to be associated with variable
neurological deficiencies (FXTAS; Tassone et al., 1999; 2000a,
2007; Garcia-Arocena and Hagerman, 2010). It means that
these males with a single FMR1 gene sometimes (but not
always suffer) from the associated significant reduction of FMRP
protein expressed in their lymphocytes (Tassone et al., 2000b).
Thus as in females, the FMR1 CGG premutation does not
always cause a clinical phenotype (Allingham-Hawkins et al.,
1999; Tassone et al., 2000a,b,c). Besides CpG methylation patterns, methylation of two epigenetically controlled genomic
sequence domains (FREE1 and FREE2) located upstream of the
FMR1 promoter might be an additional or alternative transcriptional control agent. Both genomic DNA regions control the
methylation level of the adjacent FMR1 promoter and therefore
the FMR1 transcription rate as well (Godler et al., 2010).
(iii) Local changes in or around the chromatin domain of the FMR1
gene, or some variabilities of the general inactivation mechanism
of the second X chromosome in adult females, might also control
the transcription rate in this genomic region (Coffee et al., 2002;
Gheldof et al., 2006). However, local chromatin studies with
FMR1 gene alleles containing different numbers of CGG alleles
were not yet reported and studying a putative correlation
between the level of mRNA expression and X-inactivation in
blood cells failed to display any significant correlations (Sang
et al., 2008; Tejada et al., 2008).
Figure 3 Expression of the encoded protein (FMRP) in human
female germline after birth and after puberty. Immunohistochemical
staining pattern of our monoclonal FMRP antiserum (MAB1C3;
1:200 v/v) on tissue sections of ovarian biopsies from women of
the indicated ages with the clinical diagnosis of putative cyst development. In all samples, we observed a predominant expression of FMRP
in the granulosa cells of the primary follicles. Bar length in right corner
of each picture corresponds to 10 mm in original tissue section. For
further discussion, see text.
region located 130 nt upstream of the CGG repeats (Beilina
et al., 2004). The authors therefore speculated that increased
FMR1 transcript levels are caused by the use of multiple TSSs.
(ii) The degree of CpG methylation in the CGG triplet blocks
upstream of one or both FMR1 gene alleles also influences the
gene transcription rate (Pietrobono et al., 2002). In males
where a variable CpG methylation level in the CGG triplet
block and the adjacent CpG island of the FMR1 functional
In this context it is worthwhile to recall that in 50% of cases the POF
condition was found to be reversible, and that 5– 10% of POF women
can conceive and deliver a child after they have received their diagnosis
of idiopathic impairment of their ovarian reserve (Gleicher and Barad,
2010). One explanation for these reversing phenotypes might be the
involvement of epigenetic signals, such as DNA methylation and
histone modification, controlling the transcriptional efficiency of the
FMR1 gene as discussed earlier. However, we have to acknowledge
that some of these signals might differ in leukocytes and granulosa
cells depending on cell-specific transfactors. Moreover, a number of
genes and signal pathways are known to be quantitatively involved in
the rate of follicle loss. Thus, the large genetic heterogeneity with variable expressivities of distinct mutations in other functional folliculogenesis genes might cause a variable POF phenotype as well because of
the individual genetic background.
FMRP expression in granulosa cells is a
quantitative trait
Before birth, FMRP expression was found only in the oocyte in human
fetal ovaries (Rife et al., 2004). We therefore revealed in our own
experiments a cellular shift of FMRP expression to granulosa cells
when human follicles further mature after birth (Fig. 3). It is well
known that the cellular transcript level of the FMR1 gene directly controls the amount of FMRP in each cell by a negative feed back loop
(Kenneson et al., 2001; Primerano et al., 2002). This is probably
managed by a binding capacity of FMRP specific to its own transcript
via some purine-quartet motifs (Schaeffer et al., 2001). It suggests that
the cellular FMRP quantity located in the RISC (Jin et al., 2004) must
be tightly controlled. Indeed, this protein is a major cellular translational repressor protein, binding to multiple transcripts and controlling
1249
Fragile X gene expression and premature ovarian failure
the level of their translation (Laggerbauer et al., 2001; Schaeffer et al.,
2003; Till, 2010). Any change in the level of the cellular FMR1 transcript and subsequently protein expression will therefore influence
immediately the quantitative cellular level of proteins encoded by
FMRP-interacting transcripts required in the same target tissue
(Tassone and Hagerman, 2003; Gilad et al., 2008).
For the female germline, we therefore assume that expression of
FMRP in granulosa cells, as first described in this paper (Fig. 3), is probably a sensitive quantitative trait controlling the ovarian reserve, as was
recently first suggested by Gleicher et al. (2009, 2010).
During follicular maturation, the granulosa cells express multiple
endocrine signals and functionally important genetically controlled
signal pathways for proper maturation of the ooocyte. Among them
are the gonadotrophins, FSH and LH, and proteins such as BMP15,
inhibin and activin expressed along the transforming growth factor b
signal pathway. If these products are quantitatively reduced or absent,
they also will cause POS (McNatty et al., 2005; Miyoshi et al., 2006; Fassnacht et al., 2006; Binelli and Murphy, 2009; Gougeon, 2010).
Quantitative trait loci influencing variation of menopausal age were
recently also found for the genes of the POF3 region of the short arm
of the X chromosome and expressed in the ovary (Van Asselt et al.,
2004). If this holds true also for the FMR1 gene in the POF1 region,
as suggested by the data of this paper, any slight variation of FMRP
expression in the granulosa cells of maturating follicles below or
above the optimal level (Chen et al., 2003) might then immediately
interfere with the process of folliculogenesis.
Authors’ roles
J.S. performed all quantitative FMR1 expression assays and their statistical analyses described in this paper. She also prepared the first draft of
this paper.Z.P. performed all (CGG) triplet number assays described in
this paper using the ALF sequence automat for triplet counting and
Southern blotting for genomic hybridization experiments. J.Z. performed all immunohistochemical staining experiments on ovarian
tissue sections with the monoclonal FMRP antiserum. P.S. supported
J.Z. in the practical exploration of the optimal immunohistochemical
staining protocol with the FMRP antiserum and prepared the primary
microscopic images of her results.C.H. collected all clinical data and
blood samples from the women control population, i.e. which had a
normal menstrual history (NMH group). T.S. collected all clinical data
and blood samples from the POF patient group consulting his outpatient
clinic for endocrinological support or IVF/ICSI request. P.V. initiated this
FMR1 expression study in blood cells, coordinated clinical and molecular
genetic data evaluation and prepared the final manuscript.
Acknowledgements
We thank the POF patients and the women of the control population
with a normal menstrual history for their willingness to fill in our
detailed clinical questionnaires and for the blood and ovarian tissue
samples for this FMR1 and FMRP expression study. All clinical colleagues in our outpatient clinics not included in our co-author list
and especially Waltraud Eggert-Kruse, Petra Beuter-Winkler, and
Petra Frank-Herrmann are thanked for their continuous support in
collecting this large number of patient samples. Marie Luise Diarra is
thanked for preparation of the ovarian tissue paraffin sections and
Petra Blim for support in the final editing of the manuscript. We are
also indebted to our colleagues for many helpful discussions during
this work. Bettina Toth is thanked for critical reading and commenting
the final manuscript.
Funding
This work was supported by a clinical grant to T.S. as the head of our
Department of Endocrinology and Reproductive Medicine in
Heidelberg.
References
Albright F, Smith PH, Fraser R. A syndrome characterized by primary
ovarian insufficiency and decreased stature: report of 11 cases with a
digression on hormonal control of axillary and pubic hair. Am J Med
Sci 1942;204:625 – 648.
Allen EG, He W, Yadav-Shah M. A Study of the distributional
characteristics of FMR1 transcript levels in 238 individuals. Hum Genet
2004;114:439 – 447.
Allen EG, Sullivan AK, Marcus M, Small C, Dominguez C, Epstein MP,
Charen K, He W, Taylor KC, Sherman SL. Examination of
reproductive aging milestones among women who carry the FMR1
premutation. Hum Reprod 2007;8:2142– 2152.
Allingham-Hawkins DJ, Babul-Hirji R, Chitayat D, Holden JJ, Yang KT,
Lee C, Hudson R, Gorwill H, Nolin SL, Glicksman A et al. Fragile X
premutation is a significant risk factor for premature ovarian failure:
the International Collaborative POF in Fragile X study-preliminary
data. Am J Med Genet 1999;83:322 – 325.
Baechner D, Manca A, Steinbach P, Wöhle D, Just W, Vogel W,
Hameister H, Poustka A. Enhanced expresssion of the murine FMR1
gene during germ cell proliferation suggests a spezial function in both
the male and the female gonad. Hum Mol Genet 1993;12:2043 – 2050.
Beilina A, Tassone F, Schwartz PH, Sahota P, Hagerman PJ. Redistribution
of transcription start sites within the FMR1 promoter region with
expansion of the downstream CGG-repeat element. Hum Mol Genet
2004;13:543 – 549.
Bennett CE, Conway GS, Macpherson JN, Jacobs PA, Murray A.
Intermediate sized CGG repeats are not a common cause of
idiopathic premature ovarian failure. Hum Reprod 2010;5:1335– 1338.
Binelli M, Murphy BD. Coordinated regulation of folllicle development by
germ and somatic cells. Reprod Fertil Dev 2009;22:1 – 12.
Bodega B, Bione S, Dalpra L, Toniolo D, Ornaghi F, Vegetti W, Ginelli E,
Marozzi A. Influence of intermediate and uninterrupted FMR1 CGG
expansions in premature ovarian failure manifestation. Hum Reprod
2006;21:952 – 957.
Bretherick KL, Fluker MR, Robinson WP. FMR1 repeat sizes in the gray
zone and high end of the normal range are associated with premature
ovarian failure. Hum Genet 2005;117:376 – 382.
Chatterjee S, Maitra A, Kadam S, Patel Z, Gokral J, Meherji P. Archived
issues. Reprod Biomed Online 2009;2:281– 286.
Chen LS, Tassone F, Sahota P, Hagerman PJ. The (CGG)n repeat
element within the 5’ untranslated region of the FMR1 message
provides both positive and negative cis effects on in vivo
translation of a downstream reporter. Hum Mol Genet 2003;
23:3067– 3074.
Coffee B, Zhang F, Ceman S, Warren S, Reines D. Histone Modifications
depict an aberrantly heterochromatinized FMR1 gene in fragile X
syndrome. Am J Hum Genet 2002;71:923 – 932.
1250
Conway GS, Payne NN, Webb J, Murray A, Jacobs PA. Fragile X
premutation screening in women with premature ovarian failure. Hum
Reprod 1999;14:573 – 575.
Cooper AR, Baker VL, Sterlin EW, Ryan ME, Woodruff TK, Nelson LM.
The time is now for a new approach to primary ovarian insufficiency.
Fertil Steril 2010. doi:10.1016/j.fertnstert.2010.01.016.
Coulam CB, Stringfellow S, Hoefnagel D. Evidence for a genetic factor
in the etiology of premature ovarian failure. Fertil Steril 1983;
40:693 – 695.
Coulam CB, Adamson SC, Annegers JF. Incidence of premature ovarian
failure. Obstet Gynecol 1986;67:604 – 606.
Cronister A, DiMaio M, Mohoney MJ, Donnenfeld AE, Hallam S. Fragile X
syndrome carrier screening in the prenatal genetic counselling setting.
Genet Med 2005;7:246 – 250.
Davison RM, Fox M, Conway GS. Mapping of the POF1 locus and
identification of putative genes for premature ovarian failure. Mol Hum
Reprod 2000;6:314 – 318.
Dixit H, Rao L, Padmalatha V, Raseswari T, Kapu AK, Panda B, Murthy K,
Tosh D, Nallari P, Deenadayal M et al. Genes governing premature
ovarian failure. Repr Biomed Online 2010;20:724 – 740.
Fassnacht W, Mempel A, Strowitzki T, Vogt PH. Premature
ovarian failure (POF) syndrome: towards the molecular clinical
analysis of its genetic complexity. Curr Med Chem 2006;13:
1397 – 1410.
Fu YH, Kuhl DP, Pizzuti A, Pieretti M, Sutcliffe JS, Richards S, Verkerk AJ,
Holden JJ, Fenwick RG Jr, Warren ST et al. Variation of the CGG repeat
at the fragile X site results in genetic instability: resolution of the
Sherman paradox. Cell 1991;67:1047 –1058.
Garcı́a-Alegrı́a E, Ibáñez B, Mı́nguez M, Poch M, Valiente A, Sanz-Parra A,
Martinez-Bouzas C, Beristain E, Tejada MI. Analysis of FMR1 gene
expression in female premutation carriers using robust segmented
linear regression models. RNA 2007;13:756 – 762.
Garcia-Arocena D, Hagerman PJ. Advances in understanding the
molecular basis of FXTAS. Hum Mol Genet 2010;19:83 –89.
Gheldof N, Tabuchi TM, Dekker J. The active FMR1 promoter is
associated with a large domain of altered chromatin conformation
with embedded local histone modifications. PNAS 2006;
33:12463 – 12468.
Gilad Y, Rifkin SA, Pritchard JK. Revealing the architecture of gene
regulation: the promise of eQTL studies. Cell 2008.
Gleicher N, Barad DH. Can the FMR1 gene predict early ovarian aging?
Womens Health 2010;6:165– 169.
Gleicher N, Weghofer A, Oktay K, Barad DH. Relevance of triple CGG
repeats in the FMR1 gene to ovarian reserve. Reprod Biomed Online
2009;19:385 – 390.
Gleicher N, Weghofer A, Barad DH. Ovarian reserve determinations
suggest new function of FMR1 (fragile X gene) in regulating ovarian
ageing. Reprod Biomed Online 2010;20:768– 775.
Godler DE, Tassone F, Loesch DZ, Kimball Taylor A, Gehling F, Jenssen
Hagerman R, Brugess T, Ganesamoorthy D, Hennerich D, Gordon L
et al. Methylation of novel markers of fragile X alleles is inversely
correlated with FMRP expression and FMR1 activation ratio. Hum Mol
Genet 2010;8:1618 – 1632.
Goswami D, Conway G. Premature ovarian failure. Hum Reprod Update
2005;4:391 – 410.
Gougeon A. Human ovarian follicular development: from activation of
resting follicles to preovulatory maturation. Ann Endorinol 2010;
71:132 – 143.
Jin P, Zarnescu DC, Ceman S, Nakamoto M, Mowrey J, Jongens TA,
Nelson DL, Moses K, Warren ST. Biochemical and genetic interaction
between the fragile X mental retardation protein and the microRNA
pathway. Nat Neurosci 2004;7:113 – 117.
Schuettler et al.
van Kasteren YM, Hundscheid RD, Smits AP, Cremers FP, van
Zonneveld P, Braat DD. Familial idiopathic premature ovarian failure:
an overrated and underestimated genetic disease. Hum Reprod 1999;
14:2455 – 2459.
Kenneson A, Cramer DW, Warren ST. Fragile X premutations are not
a major cause of early menopause. Am J Hum Genet 1997;
61:1362 – 1369.
Kenneson A, Zhang F, Hagedorn CH, Warren ST. Reduced FMRP and
increased FMR1 transcription is proportionally associated with CGG
repeat number in intermediate-length and premutation carriers. Hum
Mol Genet 2001;10:1449 – 1454.
Krauss CM, Turksoy RN, Atkins L, McLaughlin C, Brown LG, Page DC.
Familial premature ovarian failure due to an interstitial deletion of the
long arm of the X chromosome. N Engl J Med 1987;317:125 – 131.
Krysko DV, Diez-Fraile A, Criel G, Svistunov A, Vandenabeele P,
D’Herde K. Life and death of female gametes during oogenesis and
folliculogenesis. Apoptosis 2008;13:1065 – 1087.
Laggerbauer B, Ostareck D, Keidel EM, Ostareck-Lederer A, Fischer U.
Evidence that fragile X mental retardation protein is a negative
regulator of translation. Hum Mol Genet 2001;4:329 – 338.
Laml T, Preyer O, Umek W, Hengstschlager M, Hanzal H. Genetic
disorders in premature ovarian failure. Hum Reprod Update 2002;
8:483 – 491.
Marozzi A, Manfredini E, Tibiletti MG, Furlan D, Villa N, Vegetti W,
Crosignani PG, Ginelli E, Meneveri R, Dalpra L. Molecular definition
of Xq common-deleted region in patients affected by premature
ovarian failure. Hum Genet 2000;107:304 – 311.
McNatty KP, Juengel JL, Reader KL, Lun S, Myllymaa S, Lawrence SB,
Westren A, Meerasahib MF, Mottershead DG, Groome NG et al.
Bone morphogenetic protein 15 and growth differentiation factor 9
co-operate to regulare granulosa cell function. Reproduction 2005;
129:473– 480.
Miyoshi T, Otsuka F, Suzuki J, Takeda M, Inagaki K, Kano Y, Otani H,
Mimura Y, Ogura T, Makino H. Mutual regulation of follicle-stimulating
hormone signalling and bone morphogenetic protein system in human
granulosa cells. Biol Reprod 2006;74:1073 – 1082.
Mueller-Hartmann H, Deissler H, Naumann F, Schmitz B, Schröer J,
Doerfler W. The human 20-kDa 5’-(CGG)n-3’-binding protein is
targeted to the nucleus and affects the activity of the FMR1
promoter. J Biol Chem 2000;9:6447 – 6452.
Nelson LM, Anasti JN, Kimzey LM. Development of luteinized graafian
follicles in patients with karyotypically normal spontaneous premature
ovarian failure. J Clin Endocrinol Metab 1994;79:1470 – 1475.
Nelson LM. Clinical practice. Primary ovarian insufficiency. N Engl J Med
2009;360:606 – 614.
Oostra BA, Willemsen R. A fragile balance: FMR1 expression levels. Hum
Mol Genet 2003;15:R249– R257.
Ozgur O, Oktay K. The Ovary—anatomy and function throughout human
life. Ann N Y Acad Sci 2008;1127:1 – 9.
Perlhoff WH, Schneeberg NG. The premature climacterium. Am Pract
1957;8:1950.
Pietrobono R, Pomponi MG, Tabolacci E, Oostra B, Chiurazzi P, Neri G.
Quantitative analysis of DNA demethylation and transcriptional
reactivation of the FMR1 gene in fragile X cells treated with
5-azadeoxycytidine. Nucl Acids Res 2002;14:3278 – 3285.
Powell CM, Taggart RT, Drumheller TC, Wangsa D, Qian C, Nelson LM,
White BJ. Molecular and cytogenetic studies of an X; autosome
translocation in a patient with premature ovarian failure and review of
the literature. Am J Med Genet 1994;52:19 – 26.
Primerano B, Tassone F, Hagerman RJ, Hagerman P, Amaldi F, Bagni C.
Reduced FMR1 mRNA translation efficiency in fragile X patients with
premutations. RNA 2002;8:1482 – 1488.
Fragile X gene expression and premature ovarian failure
Prueitt RL, Ross JL, Zinn AR. Physical mapping of nine Xq translocation
breakpoints and identification of XPNPEP2 as a premature ovarian
failure gene. Cytogenet Cell Genet 2000;89:44 – 50.
Rao L, Babu A, Padmalatha V, Kanakavalli M, Deenadayal M, Singh L.
Novel X-chromosomal defect associated with abnormal ovarian
function. J Obstet Gyn Res 2005;1:12 – 15.
Rife M, Nadal A, Mila M, Willemsen R. Immunohistochemical FMRP studies
in a full mutated female fetus. Am J Med Genet 2004;124A:129 – 132.
Rizzolio F, Bione S, Sala C, Goegan M, Gentile M, Gregato G, Rossi E,
Pramparo T, Zuffardi O, Toniolo D. Chromosomal rearrangements in
Xq and premature ovarian failure: mapping of 25 new cases and
review of the literature. Hum Reprod 2006;21:1477– 1483.
Sala C, Arrigo G, Torri G, Martinazzi F, Riva P, Larizza L, Philippe C,
Jonveaux P, Sloan F, Labella T et al. Eleven X chromosome
breakpoints assosiated with premature ovarian failure (POF) map to a
15-Mb YAG contig spanning Xq21. Genomics 1997;40:123– 131.
Sang HY, Young MC, Min AH, Byung MK, Jin JK, Eung GM, Jung GK,
Shin YM. X chromosome inactivation patterns in patients with
idiopathic premature ovarian failure. Hum Reprod 2008;3:688– 692.
Schaeffer C, Baroni B, Mandel JL, Ehresmann B, Ehresmann C, Moine J.
The fragile X mental retardation protein binds specifically to its
mRNA via a purine quartet motif. EMBO 2001;17:4803– 4813.
Schaeffer C, Beaulande M, Ehresmann C, Ehresmann B, Moine H. The
RNA binding protein FMRP: new connections and missing links. Biol
Cell 2003;95:221 – 228.
Schlessinger D, Herrera L, Crisponi L, Mumm S, Percesepe A, Pellegrini M,
Pilia G, Forabosco A. Genes and translocations involved in POF. Am J
Med Genet 2002;111:328– 333.
Sherman SL. Premature ovarian failure in the fragile X syndrome. Am J Med
Genet 2000;97:189 – 194.
Southern EM. Detection of specific sequences among DNA fragments
separated by gel electrophoresis. J Mol Biol 1975;98:503– 517.
Stoyanova V, Oostra BA. Trinucleotide repeat protocols. Methods Mol Biol
2004;277:173 – 184.
Streuli I, Fraisse T, Ivecheole V, Moix I, Morris MA, de Ziegler D.
Intermediate and premutation FMR1 alleles in women with occult
primary ovarian insufficiency. Fertil Steril 2009;2:464 –470.
Sullivan AK, Marcus M, Epstein MP, Allen EC, Anido AE, Paquin JJ,
Yadav-Shah M, Sherman SL. Association of FMR1 repeat size with
ovarian dysfunction. Hum Reprod 2005;2:402– 412.
Tarleton J, Kenneson A, Taylor AK, Crandall K, Fletcher R, Casey R,
Hart PS, Fisch G, Warren ST. A single base alteration in the CGG
repeat region of FMR1: possible effects on gene expression and
phenotype. J Med Genet 2002;39:196 – 200.
Tassone F, Hagerman PJ. Expression of the FMR1 gene. Cytogenet Gen Res
2003;100:124 – 128.
1251
Tassone F, Longshore J, Zunich J, Steinbach P, Salat U, Taylor AK.
Tissue-specific methylation differences in a fragile X premutation
carrier. Clin Genet 1999;55:346– 351.
Tassone F, Hagerman RJ, Taylor AK, Gane LW, Godfrey TE, Hagerman PJ.
Elevated levels of FMR1 mRNA in carrier males: a new mechanism
of involvement in the fragile-X syndrome. Am J Hum Genet 2000a;
66:6– 15.
Tassone F, Hagerman RJ, Taylor AK, Mills JB, Harris SW, Gane LW,
Hagerman PJ. Clinical involvement and protein expression in
individuals with the FMR1 premutation. Am J Med Genet 2000b;
91:144– 152.
Tassone F, Hagerman RJ, Chamberlain WD, Hagerman PJ. Transcription
of the FMR1 gene in individuals with fragile X syndrome. Am J Med
Genet 2000c;97:195 – 203.
Tassone F, Beilina A, Carosi C, Albertosi S, Bagni C, Li L, Glover K,
Bentley D, Hagerman PJ. Elevated FMR1 mRNA in
premutation carriers is due to increased transcription. RNA 2007;
13:555– 562.
Tejada MI, Garcı́a-Alegrı́a E, Bilbao A, Martı́nez-Bouzas C, Beristain E,
Poch M, Ramos-Arroyo MA, López B, Fernandez Carvajal I,
Ribate MP et al. Analysis of the molecular parameters that could
predict the risk of manifesting premature ovarian failure in female
premutation carriers of fragile X syndrome. Menopause 2008;
15:945– 949.
Till SM. The developmental roles of FMRP. Biochem Soc Trans 2010;
38:507– 510.
Toniolo D, Rizzolio F. X chromosome and ovarian failure. Curr Opin Genet
Dev 2006;16:293– 300.
Van Asselt KM, Kok HS, Putter H, Wijmenga C, Peeters PH, van der
Schouw YT, Grobbee DE, te Velde ER, Mosselman S, Pearson PL.
Linkage analysis of extremely discordant and concordant sibling pairs
identifies quantitative trait loci influencing variation in human
menopausal age. Am J Hum Genet 2004;74:444 – 453.
Vegetti W, Grazia-Tibletti M, Testa G, de Lauretis Y, Alanga F, Castoldi E,
Tabolerelli M, Motta T, Bolis PF, Dalpra L et al. Inheritance in idiopathic
premature ovarian failure: analysis of 71 cases. Hum Reprod 1998;
13:1796– 1800.
Welt CK. Primary ovarian insufficiency: a more accurate term for
premature ovarian failure. Clin Endocr 2008;68:499 – 509.
Wilcoxon F. Individual comparisons by ranking methods. Biometrics Bull
1945;1:80 – 83.
Wittenberger MD, Hagerman RJ, Sherman S, McConkie-Rosell A, Wlt CK,
Rebar RW, Corrigan EC, Simpson JL, Nelson LM. The FMR1
premutation and reproduction. Fertil Steril 2007;3:456 – 465.
Zinn AR, Tonk VS, Chen Z. Evidence for a Turner Syndrome locus or loci
at Xp11.2-p22.1. Am J Hum Genet 1998;63:1757 – 1766.