Antioxidative enzyme-related gene expression levels in cumulus cells during in vitro fertilization: A comparison between normal and poor ovarian responders

Background: Preeminent interfering factor caused in vitro fertilization (IVF) failure has been broadly emphasized on poor ovarian response (POR) to controlled ovarian hyperstimulation (COH). The specic etiology and precise infertility treatment in POR patients, until now, has not been established and leads to detrimental effects on physical or psychological patient life qualities. To date, one biological stress phenomena, oxidative stress, is broadly implicated in varieties of female reproductive pathological conditions including poor oocyte quantity and quality. Hence, the present study is principally aimed to investigate oxidative stress scavenging system; enzymatic antioxidant-related gene expression level (SOD1, SOD2 and GPx4), in cumulus cells retrieved from poor ovarian responders (POR) and normal ovarian responders (NOR). Methods: A cross-sectional analytic study; fty-six infertile women, aged between 25-49 years-old who underwent IVF procedure, were enrolled into this study with 28 participants in POR group (1-4 retrieved oocytes) and 28 participants in NOR group (5-15 oocytes). The cumulus cells from rst punctured follicle were processed for mRNA gene expression levels of enzymatic antioxidant-related genes (Superoxide dismutase: SOD1, SOD2 and Glutathione peroxidase: GPx4) by quantitative real-time PCR (qPCR) and their association to other reproductive biological factors. Results: The relative mRNA expression of SOD1, SOD2 and GPx4 neither differed between poor and normal responders nor correlated with other relevant reproductive biological cohorts. However, the relative SOD2 mRNA expression level in patients who received human menopausal gonadotropin (hMG) was notably higher than the recombinant follicle stimulating hormone (rFSH) group (P = 0.017). Conclusions: Our major ndings indicated that the enzymatic antioxidant-related gene mRNA expression in the cumulus cells do not associate with the ovarian response and other clinical outcomes (i.e. fertilization and pregnancy rates) after IVF procedure. However, the enzymatic antioxidant-related genes positively relate with type of gonadotropin administration. ART: reproductive technology; COCs: cumulus cell oocyte complex; COH: controlled ovarian hyperstimulation; GPx: glutathione peroxidase; GSH: total glutathione; GR: glutathione reductase; GST: glutathione S-transferase; hCG: human chorionic gonadotropin; hMG: human menopausal gonadotropin; IVF: in vitro fertilization; LH: luteinizing hormone; LHR: LH receptor; mRNA: messenger RNA; NOR: normal ovarian response; OPU: ovum pick-up; PCOS: polycystic ovary syndrome; POR: poor ovarian response qPCR: real-time PCR; rFSH: follicle hormone; ROS: oxygen


Introduction
According to stress from the social environment and modern lifestyles, the incidence of couples who encounter infertility has dramatically increased. Assisted reproductive technology (ART) including in vitro fertilization (IVF) has been introduced and is widely accepted as standard treatment procedure for couples with tubal factor such as bilateral tubal obstruction, severe male factor, unexplained infertility or couples who failed the initial infertility treatment. However, overall success rates for infertility treatment by IVF are described to be within a range between 30% -40% [1]. Currently, one of the major obstacles causing IVF failure has been emphasized on poor ovarian response (POR) to controlled ovarian hyperstimulation (COH). Obviously, low numbers of collectable oocytes, impairment of embryo development and IVF outcomes are a consequence of POR [2]. Furthermore, clinical pregnancy rates in patients with POR is eventually lower than normal ovarian response (NOR) patients (de ned as 10-15 retrieved oocytes) [2]. Although multiple POR etiologies are proposed such as ovarian aging, premature ovarian insu ciency, ovarian tumor or other ovarian pathologies including endometrioma [3], some infertile women who present with normal baseline characteristics and receive standard doses of gonadotropin still encounter a problem of poor ovarian response without any identi able causes. The speci c etiology together with precise infertility treatment in POR patients, until now, has not been established which leads to recurrent IVF failure and has detrimental effects on physical or psychological patient life qualities [4].
Recently, one biological stress phenomena, oxidative stress, was broadly implicated in varieties of pathological conditions. Oxidative stress refers to an imbalance between pro-oxidant and antioxidant (enzymatic and non-enzymatic antioxidants) homeostasis that leads to an excessive amount of reactive oxygen species (ROS), reactive nitrogen species (RNS) or other free radicals [5.] The excessive accumulation of ROS or other free radicals could interfere with normal cells, organ structures and functions through the following mechanisms; opening of ion channel resulting in unstable mitochondrial membrane potential, lipid peroxidation, protein modi cations and DNA oxidation [6]. In terms of female reproductive dysfunction, a number of previous studies demonstrated that polycystic ovary syndrome (PCOS), endometriosis and preeclampsia was negatively affected by oxidative stress [7] while oxidative stress related to POR and poor IVF outcomes are imperceptibly progressed. A previous study in 1995 initially examined the enzymatic antioxidant activity in follicular uid which documented lower Glutathione peroxidase (GPx) activity in the follicular uid of non-fertilized oocytes compared to fertilized group [8]. Another research study in 2006 compared different gene expressions in cumulus cells retrieved during IVF procedures between old and young women. The ndings presented a gradual decreased level of messenger RNA (mRNA) coding for enzymatic antioxidants; Superoxide dismutase (SOD1, SOD2) and catalase in older women [9]. Notable scienti c support was again published in 2009 when a low level of cumulus cell-SOD was signi cantly correlated with IVF failure [10]. To emphasize the role of oxidative stress markers in IVF, follicular uid-enzymatic antioxidant pro le was recently comprehensive studied [11]. The research data indicated that the activities of GPx, Glutathione reductase (GR) and Glutathione Stransferase (GST) from patients with reduced ovarian response was lower than age-matched healthy oocyte donors [11]. Furthermore, collective data in 2018 described that follicular uid-total glutathione (GSH) activity markedly decreased in patients with low fertilization rate after ICSI (intracytoplasmic sperm injection) [12].
From the above-mentioned data, the oxidative stress phenomenon with decreased enzymatic antioxidant capacity may result in senescence of ovary in patients with poor ovarian response resulting in poor IVF outcomes. However, studies on enzymatic antioxidant-related gene expression levels; SOD1 and SOD2 and GPx4, which are considered potent lipid hydroperoxide scavengers in cumulus cells, and how they are related to ovarian response status have only been partially investigated. Thus, this study principally aimed to I) evaluate enzymatic antioxidant-related gene expression levels (SOD1, SOD2 and GPx (isoform 4; GPx4) in cumulus cells which are under similar milieu as it presents in the oocytes between poor ovarian responders (POR) and normal ovarian responders (NOR) to COH and II) determine any correlation between enzymatic antioxidant-related genes to other reproductive biological factors including IVF

Participants
Fifty-six infertile women, aged between 25-49 years old who underwent IVF were enrolled into this study with 28 participants in normal ovarian response group (NOR) and 28 participants in poor ovarian response group (POR). Patients were categorized to the normal and the poor ovarian response groups according to the number of retrieved oocytes at ovum pick-up (OPU) day.
Poor ovarian response in the present study was de ned as 1-4 retrieved oocytes. The control group or normal ovarian response group, age-matched with POR group, consisted of women with 5-15 retrieved oocyte. Exclusion criteria were women with active ovarian pathologies such as ovarian endometrioma or basal follicle stimulating hormone (FSH) level more than 20 IU/L or women with ovarian hyperstimulation syndrome. None of the participants were taking any antioxidant and vitamin supplements other than folic acid.

Sample size calculation
Sample size calculation was analyzed according to the previous study by Donabela et al [13] with mean of SOD1 relative gene expression in healthy women control group; 2.06 2 -Δ ΔCT (SD = 0.53). To compare two different means, the sample size calculation was performed with the following expectation; SOD1 gene expression level in POR was lower than NOR by 20%, resulting in a number of 25 women for each study group. With an addition of 10% probability for data loss, a total of 56 women were nally included in the present study, 28 patients with NOR and 28 women with POR. Cumulus cell retrieval protocol is described brie y as the following: the collected COCs were stored in fertilization media (LifeGlobal, Guilford, CT, USA) containing 10% human serum albumin (LifeGlobal) and incubated at 37°C in 5% CO 2 for 2 to 3 hours. The COCs were then exposed to 70 IU hyaluronidase (type VIII from bovine testes, Sigma). The cumulus cells were mechanically separated from oocyte under stereo-microscopy (2X magni cation). The oocyte from the rst punctured follicle was further cultured separately from other sibling oocytes. To avoid vaginal cell and white blood cell contamination, cumulus cells were washed with phosphate-buffered saline solution and assessed under light microscope (10X magni cation). Each sample was stored in 15 mL-tube and further processed for the mRNA gene expression study.
Evaluation of antioxidative enzyme-related genes (SOD1, SOD2 and GPx4) mRNA expression The cumulus cell total RNA extraction from each individual COCs was performed using ReliaPrep™ RNA Miniprep System (Promega, Madison, WI, USA). RNA concentration was then measured with NanoDrop 1000 Spectrophotometer (NanoDrop Technologies, Inc., Wilmington, DE, USA). The rst-strand complementary DNA (cDNA) was generated from 50 ng total RNA by using Reverse Transcription System (Promega). The cDNA concentration was again measured with Nanodrop spectrophotometer. Genespeci c primers for the subject sequences, housekeeping gene (GAPDH) and target genes (SOD1, SOD2 and GPx4) were predesigned by integrated DNA technologies (PrimeTime Mini qPCR Assay, Integrated DNA Technologies, Singapore) (Table1). Conventional PCR was carried out according to the manufacturer's instruction (Multiplex PCR kit, Qiagen, Valencia, CA, USA) as following: Taq activation (95°C, 15 min) and 38 cycles of denaturation (94 °C, 30 sec), annealing for 30 sec at a different temperature for each primer (GAPDH, 52 °C; SOD1, 52 °C; SOD2, 56 °C; GPx4, 52 °C) and then extension at 72°C for 20 sec. The nal extension was performed at 72 °C for 1 min. Visualization of the PCR product was evaluated using UV light (Gel Documentation System, Bio-Rad, Hercules, CA, USA). Only samples with con rmation of GAPDH expression were proceeded to quantitative real-time polymerase chain reaction (qPCR) with total volume of 20 µl (consisted of cDNA 50 ng, 10 µl qPCR master mix (Luna Universal Probe qPCR master mix, New England Biolabs, Ipswich, MA, USA) and 500 nM of qPCR Primer and probes (PrimeTime Mini qPCR Assay) ( Table 1). All reactions were performed in duplicate and processed with qPCR (StepOnePlus Instrument, Applied Biosystem, Foster City, CA) according to following conditions: 95 ºC for 60 sec, followed by 45 cycles at 95 ºC for 15 sec and 56 ºC for 15 sec. Two replications were calculated for each gene expression. Relative quanti cation of mRNA expression was calculated using 2-Δ ΔCT method [14].
Baseline characteristics, in vitro fertilization (IVF), embryo transfer and clinical pregnancy outcomes Patients were determined to undergo fresh or frozen-thawed embryo transfer on the basis of their clinical indication. Clinical pregnancy was de ned as the presence of intrauterine gestational sac 4 weeks after embryo transfer. Clinical data such as baseline characteristics, infertility history, basal hormonal pro les, ultrasonography result, IVF and embryo pro les were obtained from medical records.

Statistical analysis
All data analyses were performed using SPSS statistic software package version 22 (SPSS Inc, Chicago, IL, USA). The nonparametric Mann-Whitney test was used for data analysis to compare variables between normal and poor responders when data was absent of normal distribution. Independent t-test for continuous variables and Chi-square test for categorical variables were performed when data distributed normally. The level of signi cance was set at P < 0.05.

Results
Baseline characteristics, in vitro fertilization (IVF), embryo transfer and clinical pregnancy outcomes A total of 56 infertile women who met the inclusion criteria were enrolled into the present study. Only cumulus cells from women with con rmation of GAPDH expression were included in the antioxidative enzyme-related genes expression level analysis; 28 participants in the poor ovarian response group and 28 participants in the normal ovarian response group. The mean age of total participants was 38.6 years old. There were no signi cant differences between POR (39.3 ± 3.7 years old) and NOR group (37.9 + 2.9 years old) (P = 0.15) ( Table 2). The other baseline characteristics including body mass index, duration of infertility, number of previous medically assisted reproduction (MAR) attempts were also not signi cantly different (Table 2). Regarding the baseline of hormonal pro les, the data demonstrated that the level of FSH was marginally higher in POR than those in NOR group (P = 0.05) ( Table 3). Table 4, the selected types of gonadotropin administration were signi cantly different between the two groups; physicians preferred to prescribe rFSH in NOR (78.6%) than in POR group (35.7%), whereas hMG were administered more in POR than in NOR group, 64.3% and 21.4%, respectively. However, the total dose of gonadotropin and the duration of ovarian stimulation were comparable between both groups. As expected, estradiol level on the day of hCG trigger and the numbers of retrieved oocytes in the normal responder group were signi cantly higher than the poor responder group (P < 0.01).

As shown in
According to the IVF outcomes, the oocyte maturation rate did not differ between the two groups (P = 0.72). Additionally, there were no signi cant differences of fertilization rate after ICSI between the two groups (P = 0.34). More than 90.0% of participants in POR group underwent fresh embryo transfer, compared with 75.0% in NOR group. Similar to IVF outcomes, clinical pregnancy rates per aspiration were comparable (21.4%) between poor responders and normal responders ( Table 4).

Relative mRNA expression level of antioxidative enzyme-related genes
The present ndings revealed that the relative mRNA expression level of all enzymatic antioxidant-related genes obtained from cumulus cells did not signi cantly differ between POR and NOR groups (Table 5).
When participants were categorized according to the type of gonadotropin administration, the relative SOD2 mRNA expression level in patients who received hMG was notably higher than in those who obtained rFSH (P = 0.017) ( Table 6). However, no signi cant differences were found in the other antioxidative enzyme-related gene mRNA expression between the two gonadotropin administration cohorts ( Table 6).
The evaluation of cumulus cell mRNA gene expression level from each oocyte per patient revealed that the relative mRNA expression levels were not signi cantly different between the fertilization competence groups, i.e., normal feralization and non-fertilization. (Table 7). Similarly, none of the antioxidative enzyme-related genes expression levels were signi cantly different between pregnant and non-pregnant groups (Table 8).

Discussion
Reactive oxygen species (ROS) induced oxidative stress has been well documented to be balanced by antioxidant scavenging systems. The impairment of antioxidant activity could negatively affect female reproductive functions in various ways including alteration of ovarian steroidogenesis, oocyte maturation, ovulation or embryo development. The aim of this study was to determine the mRNA expression levels of enzymatic antioxidant-related genes (SOD1, SOD2 and GPx4) in cumulus cells and their differences between patients with poor and normal ovarian response to the controlled ovarian hyperstimulation in IVF procedure. However, this study notably revealed that the relative mRNA expression levels of all antioxidant coding genes in the cumulus cells were not different between the two participant groups. Considering the IVF outcomes represented here by the fertilization status and the clinical pregnancy rates, there were no differences in the level of enzymatic antioxidant-related gene mRNA expression.
Interestingly, the signi cantly higher level of SOD2 mRNA expression was observed in patients who received hMG for ovarian stimulation. SOD1 and SOD2 gene encoding enzymatic antioxidant, present in cytoplasm (Cu/ZnSOD) and mitochondria (MnSOD), mainly functions as a superoxide anion (O 2˙ ) scavenger [15]. SOD1 and SOD2 were previously evidenced their existence in human cumulus and granulosa cells [16]. In addition, they have been considered as one of the crucial factors that minimizes oocyte damage related to the oxidative stress environment during follicular growth, oocyte maturation and ovulation [16]. A recent 2019 study revealed that the antioxidant enzymes mRNA expression level in cumulus cells was lower in POR patients than young oocyte donors [17]. However, our ndings indicated that the levels of SOD isoform 1 and 2 (SOD1 and SOD2) mRNA expression were not different between the cumulus cells retrieved from POR and NOR participants.
Regarding another antioxidative enzyme, the family of selenium containing-glutathione-dependent peroxidase (GPx) is composed of four major isoforms; GPx1-GPx4 [18]. GPx4, a monomeric enzyme containing a single selenocysteine moiety, performs the potent function as a lipid peroxidation inhibitor [18]. The remarkable research studies on GPx4 activity have been mainly focused on male reproductive physiology such as subfertility in mice caused by mutation of GPx4 active site selenocysteine (GPx4_U46S), whereas the research data in female is con ned [19]. Recently, relevant supporting data on the role of GPx family in the female reproductive physiology; ovarian microenvironment/ follicular uid, was reported [11]. There is some evidence to suggest that GPx activity markedly decreased in poor IVF responders, whereas the lipid peroxidation level increased [11]. In contrast, the GPx4 mRNA expression level did not signi cantly differ between the two participant groups included in the current study. While more investigations are needed, this so far modest group of collective ndings appears to suggest that cumulus cell-antioxidative enzymes (SOD1, SOD2 and GPx4 isoforms) may not be exerting crucial roles and functions in POR patients. The contradictory results could be principally deliberated by the age difference between the participants from our group and Nuñez-Calonge's research group. The average age of patients in our study was 38 years old (POR = 39.3 ± 3.7 years old and NOR = 37.9 ± 2.9 years old), while the average age of POR patients was 30.6 ± 3.1 years old and those of healthy oocyte donors was 29.7 ± 4.1 years old in the study of Nuñez-Calonge et al. [11]. Previous studies have suggested that SOD gene expression and activity signi cantly changed as age-dependence [9,20]. Similar to SOD, GPx activity gradually declined by cell aging [21]. In addition to the direct effect on the lipid peroxides scavenging system, GPx4 was shown to be involved in mitochondrial membrane potential (ΔΨm) regulation under the oxidative stress condition [22]. There is also some evidence demonstrating that the frequency of cumulus-granulosa mitochondrial membrane rupture was higher in advanced-age than young participant groups [23]. This phenomenon might re ect exquisite changes in the oxidative phosphorylation capacity, i.e. GPx4 activity speci cally in the advanced-age patients [21]. Constantly, the global gene expression analysis of aged oocytes, represented by cumulus cell-oocyte complexes (COCs), in animal research models notably addressed the reduction of mitochondrial and cytosolic antioxidant gene-mRNA expression including peroxiredoxin 3 (Prdx3), thioredoxin 2 (Txn2) and superoxide dismutase 1 (Sod1) [24,25]. With relevant supporting scienti c data, it might be possible to explain our contrasting ndings from other research studies which found that enzymatic antioxidant-related genes (SOD1, SOD2 and GPx4) in cumulus cells retrieved from advanced-age patients may not play prominent roles in determining poor or normal ovarian response to COH. The present ndings could only support the existence of SOD and GPx isoforms in human cumulus cells, while their roles in age-related ovarian response are still to be further investigated.
According to the IVF outcomes, none of the genes in our study were correlated with fertilization or pregnancy rates. Although various evidences indicated the correlation between oxidative stress and IVF outcomes [26], the gradual decrease of antioxidant activity in advanced-age related cohorts, both POR and NOR participants, as presented in our study should be considered. Indisputably, oocyte quantity and quality have tremendously declined in advanced-age patients [27]. However, fertilization competence has been also in uenced by various factors such as sperm quality, microenvironment during the IVF procedure, while clinical pregnancy depends on diverse factors such as embryo quality including ploidy status, uterine receptivity as well as maternal co-morbidities. Thus, it is not surprising that the indifferent IVF outcomes between POR and NOR groups are demonstrated in our study.
Regarding patient groups categorized by the type of gonadotropin administration, the mRNA expression level of SOD2 in patients who received hMG was notably higher than rFSH group. The hMG, composed of FSH and luteinizing hormone (LH), is generally known as hormonally active medication for infertility treatment [28]. Due to the crucial role on antioxidant system, LH-like exogenous hormone could stimulate SOD1 activity in human corpus luteum [29]. Besides, the recent study in other mammalian species, bovine luteal cell cultures revealed that mRNA and protein expression levels of SOD1 and SOD 2 were also mediated by LH supplementation via cAMP-dependent pathway [30]. Theoretically, one key of ovarian function regulation through pituitary gonadotropins is the action of LH via LH receptor (LHR) [31]. In female reproductive physiology, the LHR expressed in various cell types including cumulus cells; both compacted and expanded cumulus cells [32]. Relative to our results, the major nding on the markedly high SOD2 mRNA expression in patients receiving hMG might contribute to the dominant effect of LH. However, the speci c mechanism of hMG related to SOD2 mRNA expression in cumulus cells may be of interest to elucidate the underlying pathway.

Conclusions
Our ndings indicated that the enzymatic antioxidant-related gene mRNA expression (SOD1, SOD2 and GPx4) in cumulus cells did not associate with ovarian response and IVF outcomes. However, one of these genes' mRNA expression was signi cantly related to type of gonadotropin administration. Future research should be conducted to elucidate the different gonadotropin preparations related with oxidative stress status to ultimately achieve optimal ovarian stimulation protocol, especially in poor ovarian response patients.     Data are presented as mean ± SD or n (%).