Activation of AKT/mammalian target of rapamycin signaling in the peripheral blood of women with premature ovarian insufficiency and its correlation with FMR1 expression.

BACKGROUND: The protein kinase B (AKT)/mammalian target of rapamycin (mTOR) signaling pathway regulates early follicular activation and follicular pool maintenance in female germline cells. Fragile X mental retardation 1 (FMR1) regulates folliculogenesis and it is variably expressed in patients with Premature Ovary Insufficiency. FMR1 expression is supposed to be linked to AKT/mTOR signaling in an ovarian response dependent manner as demonstrated in recent in vitro and in vivo studies in the female germline in vitro and in vivo.
METHODS: We evaluated changes in the expression of AKT/mTOR signaling pathway genes by real time PCR in the peripheral blood of 74 patients with Premature Ovarian Insufficiency and 56 fertile controls and correlated their expression with FMR1 expression.
RESULTS: Expression of the genes AKT1, TSC2, mTOR, and S6K was significantly more abundant in patients with POI than in the controls. For AKT1, TSC2 and mTOR, gene expression was not affected by FMR1-CGG repeat number in the 5´-untranslated region. FMR1 and S6K expression levels, however, were significantly upregulated in patients with POI and an FMR1 premutation. Independent of a premutation, expression of mTOR, S6K, and TSC2 was significantly correlated with that of FMR1 in all patients. Furthermore, when grouped according to ovarian reserve, this effect remained significant only for mTOR and S6K, with higher significance note in patients with Premature Ovarian Insufficiency than in the controls.
CONCLUSIONS: In Premature ovarian insufficiency patients, activation of AKT/mTOR signaling pathway is remarkable and putatively pathognomonic. Additionally, it seems to be triggered by an FMR1/mTOR/S6K linkage mechanism, most relevant in premutation carriers.

Background

Severe or complete ovarian exhaustion before the age of 40 years is referred to as premature ovarian insufficiency (POI) or premature ovarian failure syndrome [1, 2]. It is characterized by hypergonadotropic oligo- to amenorrhea (> four months) [3]. POI development is multifactorial and is caused by various pathognomonic reasons including chromosomal and genetic aberrations, autoimmune and metabolic diseases (such as galactosaemia, and polyglandular autoimmune syndromes), infections (such as mumps), and iatrogenic treatments [2]. Approximately 1% of women are affected by POI, with 4–30% having a familial element, suggesting the involvement of a genetic component [4, 5]. Several genetic factors and signaling pathways are thought to play crucial roles at different times during folliculogenesis [6]. Folliculogenesis is affected by various factors as described before and ovarian reserves can range from high, above normal, to poor and eventually exhausted at different ages. The transition from physiologically normal to pathologically abnormal reserves is hard to detect. Therefore, identification of key functional steps in ovarian aging, folliculogenesis, and its associated disorders and ovarian reserves are highly needed.

The phosphatidylinositol-3-kinase (AKT)/mammalian target of rapamycin (mTOR) signaling pathway is essential to folliculogenesis, as it maintains the primordial follicular pool. This pathway is functionally involved in primordial follicle activation, granulosa cell (GC) proliferation, oocyte-GC, inter-GC communication, and cell cycle proliferation [7, 8]. Moreover, experimental AKT activation induces follicular maturation in patients with POI [9], and AKT1 expression in cumulus cells might serve as a marker for a positive pregnancy outcome [10].

The FMR1 gene (encoding fragile X-mental retardation 1 protein [FMRP], OMIM: *309,550, localized at Xq27.3) is involved in folliculogenesis and associated disorders [11]. Experimental animal studies have found that FMR1 is required for both germ stem cell maintenance and repression of primordial germ cell differentiation [12]. Moreover, this gene contains a variable CGG base triplet in its 5´-untranslated region (UTR) that is subject to premutation (PM) during expansion of 55–200 nt. Premutation alleles and so-called gray zone alleles (45–54 repeats) are associated with the development of POI in up to 13% of patients and can expand from one generation to the next. This condition is also called fragile X POI (FXPOI, OMIM #311,360) and is the most common monogenic cause of POI [13]. Elevated FMR1 mRNA levels in PM carriers are associated with reduced FMRP levels, reflecting a negative feedback mechanism between FMR1 and FMRP [14], which can induce several ovarian damage mechanisms [15, 16]. Additionally, distinct FMR1 gene expression due to different lengths of CGG repeats before premutation, defined as different genotypes, affect the ovarian reserve prior to POI [17, 18]. However, large variations in leukocyte FMR1 expression in patients with POI without PM have also been identified [19]. Furthermore, analysis of transcriptional changes in peripheral blood of FMR1 PM carriers demonstrated CGG repeat length-dependent downregulation of genes involved in inflammation, neuronal development, apoptosis, and proliferation. One of these highly downregulated genes was AKT1 [20].

FMRP can bind target RNAs within the RNA-interference silencing complex (RISC) [21], thereby regulating the storage, degradation, and translation of their own gene transcripts as well as other gene transcripts. Furthermore, FMRP can bind complementary 3′-UTRs in target mRNAs via their secondary structures (G-quadruplex RNA) [22]. Some of these target genes are located within the mTOR signaling pathway, including tuberous sclerosis complex 2 (Tsc2) and mTOR, as detected in ovarian studies of Fmr−/− mice [23].

We previously identified a putative functional linkage of FMR1/FMRP expression with mTOR/AKT signaling under maintenance of the FMR1/FMRP negative feedback loop, which can be altered by specific inhibition of mTOR [24]. In addition, we recently identified significant correlations between the expression of FMR1 and that of AKT1, TSC2, mTOR, and S6K (encoding ribosomal protein S6 kinase) in fresh GCs from women undergoing controlled ovarian stimulation for in vitro fertilization/intracytoplasmic sperm injection with either a normal or a poor response [25]. However, these relationships have not yet been evaluated in the peripheral blood of patients with POI and might differ according to tissue-specific expression and disease-specific alterations. Importantly, GCs in women with POI cannot be evaluated owing to ovarian exhaustion, and the detection of genes involved in AKT/mTOR signaling and putative correlations with FMR1 in peripheral blood might offer new perspectives for women at risk of POI and associated disorders.

Accordingly, the present study aimed to determine the expression of genes involved in the AKT/mTOR signaling pathway and their putative correlations with FMR1 in leukocytes from the peripheral blood of patients with POI and normal fertile controls (FCs), and to assess the value of these targets as predictive markers or tools in the diagnosis and prognosis of POI.

Material and methods

Design and patients

This prospective, observational, clinical study proceeded at the University Women’s Hospital (Heidelberg, Germany) between February 2017 and October 2020. Written, informed consent to participate in the study was obtained from 74 women with POI and 56 who were fertile (FCs). POI was defined based on the European Society for Human Reproduction and Embryology criteria [3]. The local ethics committee at Ruprecht-Karls-University, Heidelberg, Germany approved this study (ID: S-602/2013), which was conducted according to the principles of the Declaration of Helsinki (2013 amendment). Blood samples were collected from all patients and controls.

DNA and RNA extraction

Samples of DNA and RNA were prepared in parallel from 20 mL of blood samples collected into tubes containing ethylenediaminetetraacetic acid as described previously [1]. Total RNA was primed with oligo dT using the SuperScript First-Strand Synthesis System (Invitrogen GmbH, Darmstadt, Germany; cat. no.: 11904–018) and M-MLV Reverse Transcriptase RNase H Minus, Point Mutant of Promega (Promega Corp., Madison, WI, USA; cat. no.: M 3683) to synthesize cDNA.

Analysis of CGG repeat length

We analyzed CGG repeat lengths in the 5′-UTR of FMR1 (NM_002024.5) exon 1 in patients with POI, using polymerase chain reaction (PCR) and an ALFexpress™ DNA sequencer (Amersham 1050; Pharmacia Biotech, Freiburg, Germany) or an ABI 3100/3130xl sequencer (Thermo Fisher Scientific Inc., Waltham, MA, USA) as described previously [17]. Since May 2020, CGG repeat lengths were analyzed using POI Triplet Repeat Primed Polymerase Chain Reaction (PCR, TP-PCR, and AmplideX® PCR/CE FMR1 Kits; Asuragen Inc., Austin, TX, USA) as per the manufacturer’s protocol, and fragments were separated using a SeqStudio Genetic Analyzer (Thermo Fisher Scientific Inc.). Electropherograms were analyzed using GeneMapper™ v. 5 software (Thermo Fisher Scientific Inc.).

Gene expression analysis

TaqMan predesigned gene expression assays for FMR1 (Hs00924544_m1), AKT1 (Hs00178289_m1), mTOR (Hs00234508_m1), S6K (Hs00177357_m1), TSC2 (Hs01020387_m1), two housekeeping genes HPRT and TBP (Hs99999909_m1; Hs00427620_m1, respectively), and TaqMan universal PCR master mix were obtained from Thermo Fisher Scientific Inc. and performed as per the manufacturer’s protocol and as described previously [25]. All samples were analyzed in triplicates under standard qPCR conditions on a Fast Forward 7500 real-time PCR system (Thermo Fisher Scientific Inc.). Relative gene expression was analyzed using the ΔΔCt method [26]. The cDNA obtained from a lymphoblastoid cell line derived from fertile women was used as the calibrator in each run.

Statistical analysis

The data distribution was determined using Shapiro–Wilk tests. Between-group comparisons of POI vs. FCs and PM carriers vs. non-PM carriers were analyzed using Mann–Whitney tests. Not all data were normally distributed; hence, correlations were analyzed using the Spearman correlation coefficient rho (ρ).

Results are presented as medians with interquartile ranges (25th to 75th percentiles). All data were analyzed using the Statistical Package for the Social Sciences v. 27.0 (IBM Corp., Armonk, NY, USA), and values with p < 0.05 were considered statistically significant.

Results

Gene expression of AKT1, TSC2, mTOR, S6K, and FMR1

The expression of FMR1 was slightly more abundant in peripheral blood from patients with POI than in FCs, although the difference was not significant. By contrast, AKT1, TSC2, mTOR, and S6K were significantly upregulated in women with POI compared with controls (Table 1).

Table 1

AKT1, TSC2, mTOR, and S6K expression in patients with premature ovarian insufficiency vs. fertile controls

	POI (n = 74)	FC (n = 56)
Gene	Median (25th to 75th percentiles)	Median (25th to 75th percentiles)	pa
FMR1	3.539 (2.487–4.594)	3.400 (2.374–4.373)	0.382
AKT1	4.797 (3.972–5.553)	3.590 (1.724–5.342)	0.009
TSC2	3.678 (2.942–4.077)	2.917 (1.269–3.691)	0.001
mTOR	2.533 (1.819–3.996)	1.856 (0.517–3.255)	0.002
S6K	1.824 (1.564–2.281)	1.538 (0.574–2.029)	0.001

Expression of AKT1, TSC2, mTOR, and S6K (relative gene expression) in peripheral blood of patients with patients with premature ovarian insufficiency and fertile controls. Abbreviations: POI premature ovarian insufficiency, FCs fertile controls. aMann-Whitney tests

Effects of CGG repeat length on gene expression levels in patients with POI

We found that FMR1-PM, FMR1, and S6K were significantly upregulated in patients with POI (Table 2).

Table 2

Expression of FMR1, AKT1, TSC2, mTOR, and S6K premutation vs non-premutation carriers

POI	POI without PM	POI with PMb	Pa
n	67	5
FMR1	3.499 (2.433–4.484)	5.376 (4.314–6.133)	0.011
AKT1	4.807 (4.081–5.564)	3.407 (3.305–5.401)	0.231
TSC2	3.730 (3.052–4.107)	2.494 (1.212–3.721)	0.141
mTOR	2.550 (1.878–3.917)	2.769 (1.543–4.138)	1.000
S6K	1.745 (1.555–2.218)	2.370 (2.291–2.909)	0.046

Relative Expression of FMR1, AKT1, TSC2, mTOR, and S6K in patients with premature ovarian insufficiency with and without FMR1 premutation. Abbreviations: PM permutation, POI premature ovarian insufficiency. Data are shown as medians (25th–75th percentiles). aMann-Whitney tests. bPremutated alleles (n); range, 70–81 repeats

Age

Patients were older in the FC group than in the POI group (32.5 ± 3.9 vs. 30.2 ± 7.2 y, p = 0.029). Therefore, the activation of mTOR/AKT signaling due to increased age in patients with POI can be excluded.

Correlations between FMR1 and AKT1, TSC2, mTOR, and S6K

The mRNA expression of mTOR and S6K positively and significantly correlated with that of FMR1 in all patients. These results were consistent after subgroup analysis of patients with POI and FCs. By contrast, the correlation between the expression of TSC2 and FMR1 was less significant, and was not observed after subgroup analysis. The expression of AKT1 and FMR1 did not correlate at the gene level in either group or in all patients together (Table 3).

Table 3

Correlations between FMR1 and AKT/mTOR signaling pathway genes

All patients	Spearman correlation coefficient (ρ) for FMR1	p
AKT	0.136	0.122
mTOR	0.533	< 0.001
S6K	0.551	< 0.001
TSC2	0.217	0.013
Patients with POI
AKT	0.059	0.615
mTOR	0.602	< 0.001
S6K	0.746	< 0.001
TSC2	0.191	0.104
FCs
AKT	0.177	0.193
mTOR	0.388	0.003
S6K	0.288	0.031
TSC2	0.213	0.115

Correlation between FMR1 and AKT/mTOR signaling pathway genes in all women, followed by subgroup analysis. Abbreviations: POI premature ovarian insufficiency, FCs  fertile controls

Discussion

This study aimed to elucidate putative changes in the AKT/mTOR signaling pathway in lymphocytes isolated from the peripheral blood of women with POI compared with women with normal fertility and to determine whether these genes correlate with FMR1 gene expression. This is also clinically interesting as it can potentially identify new markers of ovarian reserve that can improve prediction of ovarian responses during artificial reproductive technologies. Anti-mullerian hormone and antral follicle count are currently the most relevant ovarian reserve markers, followed by age and follicle-stimulating hormone -values; however, these markers have less predictive value in cases of deviant ovarian responses, such as high or poor responses [27-29]. The results observed in the present study in patients with POI are in line with our previous data from germline cells of women with different ovarian responses during controlled ovarian stimulation [24]. Thus, identification of such novel markers that can be used in addition or alternative to existing markers in different patient groups are highly needed.

The expression of FMR1/FMRP is controlled through a negative feedback loop [14] and FMRP, as part of the RISC, interacts with other proteins from the Argonaute family as well as with coding and noncoding RNAs to regulate the translation of various proteins [21, 22]. In the female germline, FMRP is a major regulator of folliculogenesis, and changes in gene expression, genotype, and epigenetics (variable methylation status) are associated with disordered folliculogenesis from diminished ovarian reserve until POI [13, 18, 19, 30].

Both TSC2 and mTOR are putative binding partners of FMR1/FMRP in Fmr−/− mice [23]. The AKT/mTOR signaling pathway plays important roles in various cellular functions via AKT activation, TSC1/TSC2 dimerization, and mTOR activation. Finally, mTOR complex 1 phosphorylates S6K, further regulating cell-specific translation. In the female germline, AKT/mTOR signaling participates in early and late follicular maturation and GC differentiation [31-33].

The present study analyzed gene expression in lymphocytes of peripheral blood from women with normal fertility and women with POI. We found significantly more abundant AKT1, TSC2, mTOR, and S6K expression in patients with POI. Premature recruitment of oocytes therefore fit the data from experimental rodent models. Indeed, elevated mTOR and S6K levels in Fmr-knockout mice lead to premature oocyte recruitment [34]. In addition, we observed that mTOR and S6K was correlated with FMR1 expression in both groups, but significantly correlated in patients with POI. These results further supported the mechanism observed in experimental animals [34] and were consistent with our data from GCs of women with poor ovarian response, in whom mTOR signaling was similarly linked to FMR1 and particularly to S6K [25]. As expected from the findings of previous studies, the expression of FMR1 was significantly more abundant in patients with POI and PM than with POI without PM [9]. Notably, S6K was the only gene in the AKT/mTOR pathway with a simultaneous upregulation in PM carriers. These results are consistent with findings in human neuronal cells, in which S6K is supposed to be a major FMRP-phosphorylating enzyme [35] and suggest that S6K is one of the most promising linkage partner of the AKT/mTOR pathway with FMR1. Noteworthy, although it was not yet significant, is the downregulation of AKT1 that was observed in patients with POI with PM compared to those without PM. These results are consistent with previous transcriptome results from affected males carrying an FMR1 PM [20], and highlight the putative role of AKT/mTOR signaling in FMR1 related disorders, such as FXPOI and diminished ovarian reserve (DOR).

If FMR1/FMRP levels are functional traits in orchestrated follicular maturation, that are linked to mTOR signaling transmitted by, or triggered through mTOR and/or S6K binding, our findings might support further studies and the development of therapeutic approaches. The easy and direct detection of these targets in peripheral blood could facilitate their application as prospective diagnostic tools for women with diminished ovarian reserve and could help estimate their individual risk for POI development.

The AKT/mTOR pathway is related to longevity, and its inhibition is thought to provide protection against age-related pathologies [36]. Therefore, our results could help identify pre-existing conditions or pathological mechanisms associated with premature ovarian aging. Alternatively, our findings might reflect a compensatory mechanism targeting the activation of the last remaining follicles in POI, which would be consistent with initial therapeutic trials of activation in vitro, where AKT activators induced further follicular maturation and development in women with POI [9].

Our results provide the first evidence that activation of the AKT/mTOR signaling pathway in POI is putatively linked to FMR1 in patients with POI, diagnosed using peripheral blood probes. However, our patient cohort was small and not aged matched. Therefore, studies in larger cohort with women with distinct ovarian reserves and ages are needed to evaluate this finding in the complex context of folliculogenesis and progression to POI. The impact of CGG repeat length of this postulated relationship also needs further evaluation in longitudinal studies. Moreover, the results of both groups could be affected by age, although the average age in our cohort was below 35 years. Additionally, women in the control group were slightly older than the POI group, which contradicts age related AKT activation in POI observed in our study.

Conclusions

FMR1 premutation is a known risk factor for the development of POI and is recommended for clinical testing during diagnosis [3]. We believe that FMR1/FMRP regulates proper follicular maturation not only via its premutated CGG repeat length, but also through multiple other ways. In this context, we presumed a functional linkage between FMR1 and the AKT/mTOR signaling pathway based on the findings of previous studies [10, 17, 23–25]. In the present study, we demonstrated upregulated AKT/mTOR signaling in patients with POI that might reflect a pathognomonic reason for POI development or represent a compensatory mechanism for POI. A linkage with FMR1/FMRP, presumably via modified S6K and mTOR binding, could be a functional regulator of this pathogenic mechanism and is perhaps most relevant in PM carriers, and could perhaps be more predictive than age as ovarian reserve marker in patients below 40 years of age. However, further studies are needed to elucidate the regulatory mechanisms. The detectability of these conditions in the peripheral blood of patients reflects conditions within the germline, and offers further perspectives in diagnostics and pathognomonic investigation of POI.