Growth hormone alleviates oxidative stress and improves oocyte quality in Chinese women with polycystic ovary syndrome: a randomized controlled trial

Oxidative stress (OS) is associated with poor oocyte quality and in vitro fertilization and embryo transfer (IVF-ET) outcomes for patients with polycystic ovary syndrome (PCOS). Growth hormone (GH) can function to reduce OS in some types of cells. Therefore, this prospective randomized study investigated whether GH can significantly improve OS and oocyte quality in women with PCOS. This study enrolled 109 and 50 patients with and without PCOS (controls), respectively. The patients with PCOS were randomly assigned to receive treatment with GH (PCOS-T) or not (PCOS-C). The primary outcome included markers of OS in serum and FF, and secondary outcomes were mitochondrial function in granulosa cells (GCs) and IVF-ET outcomes. The PCOS groups showed higher basal serum total oxidant status (TOS) and OS index (OSI) levels. The follicle fluid (FF) TOS and OSI and GC apoptosis rate were significantly higher, whereas the GC mitochondrial membrane potential (MMP) was significantly lower in the PCOS-C group than in the PCOS-T and non-PCOS control groups (P < 0.05). Significantly more oocytes were fertilised and cleavage stage embryos were produced in the PCOS-T group than in the PCOS-C group (P < 0.05). GH also improved the rates of implantation and clinical pregnancy, but not significantly (P > 0.05). This study showed that GH alleviated the TOS and OSI level in FF and improved GC mitochondrial dysfunction and oocyte quality in patients with PCOS. Clinical Trial Registration Number: This project was prospectively registered on the Chinese Clinical Trial Registry on October 20, 2018. (ChiCTR1800019437) (https://www.chictr.org.cn/edit.aspx?pid=28663&htm=4).


Scientific Reports
| (2020) 10:18769 | https://doi.org/10.1038/s41598-020-75107-4 www.nature.com/scientificreports/ and 9.57% vs. 20.07%, respectively) in the PCOS-T group than in the PCOS-C group (P < 0.05 for all). The MMP was significantly lower (0.23 ± 0.19 vs. 0.76 ± 0.27) and the early and late apoptosis rates were significantly higher (26.12% vs. 11.28%, and 20.07% vs. 12.37%) in the PCOS-C group than in the group without PCOS (P < 0.05 for all; Fig. 2). Mitochondrial function was a main indicator in this study. A power calculation based on sample size and mitochondrial results was performed by Power Analysis and Sample Size (PASS) 11. The result showed that a statistical power of 0.8692 was achieved in the study for the early apoptosis rate between PCOS-T group and PCOS-C group (difference = 0.1965, control group proportion = 0.0646, significance level = 0.05).

Discussion
This prospective, randomized, controlled study found increased OS and dysfunctional mitochondria in the GCs of patients with PCOS undergoing IVF. However, we also determined that GH could lower the TOS and OSI in FF and improve the mitochondrial function of GCs. We also found more fertilised oocytes and cleaved embryos in the PCOS-T group than in the PCOS-C group, whereas the rates of implantation, miscarriage, and clinical pregnancy did not significantly differ among all groups. Although numerous studies have identified OS in patients with PCOS, the experimental findings of specific OS markers are contradictory 9,26-28 . We found significantly higher TOS and OSI levels in serum and FF in patients with PCOS than in those without PCOS, whereas the MDA, TAC, and SOD levels did not significantly differ. However, Wang et al. 26 reported higher values for MDA along with lower values for SOD and TAC. Others have found that the SOD and TAC levels increase to compensate for OS 27,28 . These inconsistent results might be caused by relatively small samples and differing methodologies. Analysis from our findings, MDA is a product of lipid peroxidation, but TOS represents different oxidant molecules' additive effects 29 , ROS/RNS molecules other than MDA might have substantially contributed to the higher TOS. However, because of limited funding

PCOS-T (n = 50) PCOS-C (n = 50) Non-PCOS (n = 50)
Clinical outcomes www.nature.com/scientificreports/ and specimens, we did not measure the levels of other ROS/RNS markers. Furthermore, TAC represents the ability to eliminate free radicals 30 , and OSI (the ratio of TOS/TAC) can evaluate the balance of oxidation and anti-oxidation 31 . We found a slightly higher TAC in patients with PCOS than in those without PCOS, but the difference was not significant. Thus, the absolute TOS was higher, but the ability to eliminate free radicals was relatively insufficient in patients with PCOS. This finding agreed with the notion that OS is extant in PCOS because physiological antioxidants are insufficient to outweigh the excess ROS 2 . In addition, we found that the OS markers in FF and serum concurred. FF originates from blood diffusion and from secretion from GCs and oocytes; thus, OS markers in FF reflect the systemic OS status and that of the ovaries 32 . Through correlation analysis, we found that the basal serum TOS and OSI were both positively correlated with HOMA-IR, FINS and TT levels and negatively correlated with SHBG. These results were consistent with those of other studies showing that the pathophysiology of PCOS, such as hyperandrogenaemia and IR, was associated with OS 11-13 . The above evidence indicated the OS status of patients with PCOS, which suggests that antioxidant therapy might be beneficial. We found that GH reduced the FF TOS by 2.32 mmol H 2 O 2 Eq/L and the OSI by 3.25 in patients with PCOS. However, we did not find the same effects on the systemic OS status because serum OS markers on the trigger day did not significantly differ. Some possible explanations for this finding are as follows. Ten days of GH administration might not be sufficient to alleviate systemic OS. Furthermore, systemic OS is closely associated with obesity, IR, and hyperandrogenism 11 . Patients with PCOS in this study exhibited abnormal endocrine and metabolic statuses; therefore, GH alone might not be able to alleviate systemic OS. The antioxidant effects of GH seem to be exerted mainly in follicles 24 .
Meanwhile, compared with baseline values in all groups, the serum TOS and OSI were significantly increased, whereas the TAC was significantly decreased on the trigger day. The accumulation of ROS during COS disrupts the oxidant-antioxidant balance and leads to OS 7 . Other studies have also reported the OS state and GC apoptosis during COS, which might reduce the oocyte development potential 7,33 . Such findings provide another rationale for treatment with antioxidants in IVF.
With respect to the ovarian OS status, we found that the early and late apoptosis rates of GCs approximately doubled, and the MMP decreased by 70% in patients with PCOS. GCs, which provide nutrients and factors that enable oocyte maturation via intimate cross-talk, are steroidogenic 7 . Mitochondria in GCs are critical for oocyte development and competence and for glucose metabolism 34 . OS leads to GC apoptosis in patients with PCOS by reducing mitochondrial DNA (mtDNA) copy numbers and impairing mitochondrial oxidative Table 3. Serum OS markers at baseline and on the trigger day. Data are presented as mean ± SD. MDA malondialdehyde, SOD superoxide dismutase, TAC total antioxidant capacity, TOS total oxidant status, OSI oxidative stress index. a P < 0.05 parameter on the trigger day versus baseline.

PCOS-T (n = 50)
PCOS-C (n = 50) Non-PCOS (n = 50) www.nature.com/scientificreports/ www.nature.com/scientificreports/ metabolism 7 . Moreover, COS induces ROS production in GCs, which aggravates mitochondrial dysfunction and GC apoptosis 28 . GH might improve mitochondrial function in some types of cells, including oocytes, vascular endothelium, and skeletal muscle cells 24,35 . We found here that GH apparently improved the MMP over three-fold and decreased GC apoptosis by > 50%. The mechanism of how GH improves mitochondrial function is not fully understood. Activating the components of mitochondrial fuel delivery and enhancing ATP generation might be one mechanism 35 . FF and GCs constitute an important microenvironment for oocyte development. Along with alleviating OS, GH administration resulted in significantly more fertilised oocytes and cleaved embryos. The number of higher-quality embryos was improved by 0.86 but this improvement was not significant. Others have found that GH significantly elevates the number and quality of oocytes 36 . However, conclusions from these results should be drawn with caution because of the small sample size.
The poorer IVF outcomes in patients with PCOS than in those without PCOS is mainly due to poorer oocyte quality in patients with PCOS 2 . Along with more oocytes with better quality, we found that GH increased the implantation rate and clinical pregnancy rate and lowered the miscarriage rate, but the difference was not significant. However, Homburg et al. 25 reported that adjuvant GH therapy did not significantly alter the rates of clinical pregnancy and miscarriage in patients with PCOS. However, caution is advised when analysing these results due to the relatively small sample size.
This study has some limitations. We measured only several OS markers due to limited funding and specimens. Moreover, the sample size was too small to analyse outcomes according to the PCOS subtype.
In summary, our study revealed that GH combined with gonadotropin significantly improved mitochondrial dysfunction in GCs and oocyte quality in PCOS women. Because the precise mechanism through which GH influences OS status remains obscure, further basic investigations at the cellular level in vitro are needed to define the involved mechanisms.

Materials and methods
Ethics statements. This prospective, randomized, open-label study was registered in the Chinese Clinical Trial Registry Centre (Registration No. ChiCTR1800019437) before the first patient was enrolled, and was approved by the Chinese Ethics Committee of Registering Clinical Trials. Written informed consent to participate in the study was obtained from all participants. All procedures in this study complied with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975 (2013 revision).
The primary outcomes included markers of OS in serum and FF, and the secondary outcomes were mitochondrial function in GCs and IVF-ET outcomes. The sample size was calculated based on differences in serum TOS 37 between patients with and without PCOS. We used 6.80 µmol H 2 O 2 Eq/L as the mean difference (d) and 7.6 µmol H 2 O 2 Eq/L as the SD for TOS as the key variable 37 . Each group contained 43 participants with an α of 0.05 and a β error of 0.1 (power = 90%). Assuming a dropout of seven participants per group with a dropout rate of 14%, the final sample size was 50 participants per group.  38 and undergoing IVF treatment were enrolled in this study. The patients were randomly assigned (using computer-generated random numbers) to undergo treatment with (PCOS-T, n = 54) or without (PCOS-C, n = 55) GH. Patients were excluded from the study if they met any of the following criteria: (1) hydrosalpinx; (2) congenital uterine malformations and/or endometrial disease, tuberculosis, hyperplasia; (3) systemic lupus erythematosus and/or sicca syndrome; (4) uncontrolled endocrinopathy such as diabetes, hyperthyroidism, hypothyroidism, and hyperprolactinemia; (5) cigarette smoking and/or alcohol consumption; (6) supplementation with vitamin E, vitamin C, or CoQ10, which influence OS markers. Infertile women (aged 20-40 years) who underwent IVF-ET due to tubal issues were recruited as non-PCOS controls (n = 50), and the inclusion criterion for control patients required the absence of all the Rotterdam criteria. The exclusion criteria for this group were the same as those for the patients with PCOS. Each participant had read and signed informed consent before being enrolled in the study. In the PCOS-T group, three patients discontinued recombinant human GH (rhGH) during COS because of the cost, and one patient declined further participation. In the PCOS-C group, two patients abandoned treatment because they left Chengdu City, and three patients declined further participation. No patients in the non-PCOS group dropped out. Therefore, 50 participants per group completed the study.

Study population. Between
Medical histories regarding menstrual cycle regularity, duration of infertility, and treatment were collected from all participants. Physical examinations included measurements of height, body weight (without shoes and heavy clothing), waist circumference at the midpoint between the lower rib margin and the top of the iliac crest at end exhalation, and hip circumference at the level of the greater trochanter. Body mass index (BMI) was calculated as weight divided by height squared (kg/m 2 ). The WHR was calculated as waist circumference divided by hip circumference. Blood samples were collected from elbow veins after an overnight fast on the same day that the participants were assessed by transvaginal ultrasound on days 2-3 of menstruation or progestin-induced withdrawal bleeding. The blood samples were separated by centrifugation at 2132 × g for 15 min at 4 °C within 2 h, and serum was stored at -80 °C. www.nature.com/scientificreports/ day 2 of the menstrual cycle; then gonadotropin (Gn) doses were adjusted according to follicular growth. When serum E 2 concentration reached 300 pg/mL or when the lead follicle grew to a diameter of 13-14 mm, cetrorelix (Cetrotide; Merck-Serono KGaA.) was administered. On the same day as the rFSH administration, only patients in the PCOS-T group were subcutaneously injected with 4 IU/day of rhGH for pharmaceutical use (Jintropin, Changchun GeneScience Pharmaceutical Co., Ltd., Changchun, Jilin, China) until the trigger day. Recombinant human chorionic gonadotropin (hCG) (Ovitrelle; Merck-Serono KGaA.) was administered as the trigger when the diameters of at least two follicles reached ≥ 18 mm, and blood samples were collected on the same day prior to hCG administration. After 36 h, each follicle was measured immediately prior to its aspiration; then, oocytes were retrieved under transvaginal ultrasound guidance, and follicle flushing was not performed. FF and GCs from each patient were separately collected and considered as one sample. FF derived from follicles with a diameter ≥ 16 mm in one case were pooled together and immediately centrifuged at 700 × g for 5 min at room temperature, and the supernatant was stored at -80 °C. The precipitates were suspended with 2 mL of remaining FF and then gently layered onto 3 mL of 50% lymphocyte separation medium (Beijing Solarbio Science and Technology Corporation, Beijing, China) and centrifuged at 700 × g for 10 min at room temperature to remove red blood cells and debris. GCs were collected from the layer at the interface of the gradient and washed twice with 5 mL of PBS (Nanjing KeyGen Biotech. Co., Ltd., Nanjing, Jiangsu, China), and mitochondrial function was immediately examined.

Regimen for COS and
After fertilization in vitro, cultured embryos were evaluated on day 3, and according to the number of blastomeres and the degree of fragmentation, the higher-quality embryos were categorised as grades A/B 39 . One or two higher-quality day 3 embryos were transferred. Luteal phase support with intramuscular (i.m.) injections of progesterone (P) at 60 mg/day started on retrieval day. Serum levels of hCG, E 2 , and P were measured at 14 days after ET and hCG values > 5 IU/mL were considered positive. Clinical pregnancy was defined as a gestational sac containing an embryo with normal cardiac activity. Early miscarriage was defined as loss of pregnancy before gestational week 12. Rates of implantation, clinical pregnancy, and miscarriage were calculated. Ovarian hyperstimulation syndrome (OHSS) was diagnosed according to Navot D et al. 40 . Mild OHSS was defined as patients with at least one symptom consistent with OHSS (nausea, vomiting or diarrhoea) and a maximal ovarian diameter of 5-12 cm. Moderate OHSS was designated as patients with ultrasonic evidence of ascites (sub-hepatic ascites or post-cervical ascites ≥ 10 mm). Severe OHSS was defined as patients with clinical evidence of ascites or hydrothorax or breathing difficulties, or any of the following conditions: haematocrit > 45%; creatinine > 90 µmol/L; alanine aminotransferase > 45 U/L; or oliguria, renal failure, thromboembolic phenomena or adult respiratory distress syndrome 40 . The fresh embryo transfer was cancelled when the patient met any of the following criteria: the number of oocytes retrieved was > 18; the patient was diagnosed with moderate or severe OHSS; a P level was > 1.5 ng/mL on the trigger day; or there was abnormal endometrial thickness (< 7 mm or > 15 mm on the trigger day). Only the first embryo transfer cycle (fresh or frozen-thawed embryo transfer) was analysed in this study. The patients underwent frozen-thawed embryo transfer at least three months later. The physicians making treatment decisions for women with PCOS, such as gonadotropin dosing, trigger timing and oocyte retrieval, embryo evaluation and embryo transfer, were blinded as to study arm (rhGH or not) (Supplementary Information).
Serum MDA concentrations (µmol/L) were determined using micro-MDA detection kits (NanJing Jiancheng Bioengineering Institute Co. Ltd., Nanjing, Jiangsu, China) and ultraviolet spectrophotometry (Shanghai Meipuda Instrument Co., Ltd., Shanghai, China) was performed at 532 nm. SOD was determined using SOD kits (NanJing Jiancheng Bioengineering Institute Co. Ltd.) and spectrophotometry at 450 nm 26 . Using Trolox [(±) 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid] as a standard, the TAC (mmol Trolox Equiv./L) was determined based on the decolourisation of 2,2′-azinobis (3-ethylbenzothiazoline-6-sulphonate) (ABTS). Changes in the colour were measured as absorbance at 420 nm 30 . Using hydrogen peroxide (H 2 O 2 ) as a standard, the TOS (µmol H 2 O 2 Eq/L) was measured as the oxidation of ferrous to ferric ions by oxidant species in acidic medium. Ferric ions complexed with xylenol orange were measured at 594 nm 31 . These semi-automatic microplate colorimetric methods are simple, stable, reliable, and sensitive for determination of TAC and TOS. The OSI was calculated as the ratio of the TOS to that of the TAC. All OS parameters were detected in duplicate. Serum samples from healthy volunteers were pooled for quality control. All OS markers in FF and serum were similarly determined. The intra-and inter-assay coefficients of variations of these parameters were all < 5% and < 10%, respectively.
Apoptosis of GCs was detected using Annexin V-FITC Apoptosis Detection Kits (KeyGEN Bio TECH Co., Ltd.) using flow cytometry (MilliporeSigma Co., Ltd., Burlington, MA, USA). The MMP of GCs was examined using JC-1 Apoptosis Detection Kits (KeyGEN Bio TECH Co., Ltd.). GCs were cultured with JC-1 staining solution for 20 min at 37 °C without exposure to light and then washed twice with incubation buffer. JC-1 accumulates in functional mitochondria with high ΔΨm and forms aggregates that emit red fluorescence. When mitochondrial transmembrane potential is depolarized with low ΔΨm, JC-1 releases from the mitochondria and forms monomers that emit green fluorescence. The fluorescence intensity of JC-1 was detected using flow cytometry. The ratios of red/green fluorescence were calculated to characterise the MMP 42 . www.nature.com/scientificreports/