DOI: https://doi.org/10.21203/rs.3.rs-2341888/v1
Polycystic ovary syndrome (PCOS), one of the most prevalent diseases among women of reproductive age, is an endocrine disorder characterized by anovulation, hyperandrogenism, and polycystic ovary (Witchel 2006). Diagnosis of PCOS in a patient with the diagnostic criteria defined by the NIH (National Institute of Health) and revised at the Rotterdam conference in 2003 was based on three criteria: chronic anovulation, hyperandrogenism, and polycystic ovary. At least two of these criteria must be present for a patient to be diagnosed with PCOS (Witchel 2006; Norman et al. 2007).
Chronic anovulation, the first criterion, involves long-term menstruation (oligomenorrhea) or amenorrhea. Oligomenorrhea is observed in 60% of cases, and these patients’ menstrual cycles are completed in 35-day periods. Hypothalamic dysfunction is present in patients with amenorrhea, which is observed in 40% of cases (Teede et al. 2010). Hyperandrogenism, another of the PCOS diagnostic criteria, is characterized by increased androgen levels. Increased androgen release causes both clinical and biochemical symptoms. Hirsutism and acne are the most frequent clinical symptoms, but these are not used alone to define hyperandrogenism. Circulating free androgens are measured as a biochemical marker of hyperandrogenism. Increased luteinizing hormone (LH) and free testosterone levels are other common symptoms in PCOS (Rotterdam 2004). Polycystic ovary is a term used for ovaries containing 10–15 follicles 2–9 mm in diameter and thickened tunica albuginea in the cortex. The volume of the ovarian stroma usually increases in PCOS, and can exceed 10 ml. A manifestation in only one ovary is sufficient for the diagnosis of polycystic ovary (Rotterdam 2004). Although polycystic ovary and PCOS are frequently confused, the two are quite distinct entities. Patients with polycystic ovaries should meet at least two of the diagnostic criteria in order to be considered to have this syndrome (Norman et al. 2007).
The pathophysiology of PCOS has not yet been fully explained. However, hypothalamic-pituitary dysfunction, exaggerated adrenarche, intraovarian factors, insulin resistance, hyperinsulinemia, obesity, genetic factors, abnormal granulosa cells, and enzymatic disorders are among the causes of the syndrome (Ehrmann 2005). Disruption of the hypothalamic-pituitary axis causes high amounts of gonadotropin to be secreted from the hypothalamus, thus altering the levels of LH secreted from the pituitary gland. The cells most affected by changes in LH levels are theca follicle cells in ovarian follicles, and androgen synthesis increases in line with the rise in LH levels. Increasing androgen levels play a role in the development of numerous symptoms seen in patients with PCOS. Increasing estrogen levels in peripheral tissues and rising free androgen levels cause an increase in LH release and inhibit follicle-stimulating hormone (FSH) release with negative feedback. Anovulation develops as a result of these changes. Since FSH cannot be completely inhibited, the stimulation of the follicles continues. Oocyte development does not occur during the development of the stimulating follicles, and ovulation cannot take place. These follicles occur as cystic structures in the ovary for a period of 5–6 months (Ehrmann 2005). Many recent studies have shown that insulin resistance also plays an important role in PCOS. Insulin affects the theca cells and raises androgen levels, thus reducing the release of sex hormone-binding globin in the liver, and the level of free testosterone gradually increases as a result (Dumesic et al. 2008; McNeilly and Duncan 2013).
Depression is one of the most common PCOS syndromes (Cooney and Dokras 2017; Hung et al. 2014). However, whether depression is the cause or symptom of the disease is still the subject of debate. However, the treatment of depression ameliorates the symptoms of PCOS. For this reason, antidepressant therapy is frequently employed in patients with both diseases. The most commonly used antidepressants in the treatment of depression are tricyclic antidepressants. Amitriptyline (AMI) is one such tricyclic antidepressant. Its mechanism of action has not been fully explained, although some drugs such as serotonin and norepinephrine are thought to suppress the reuptake of neurotransmitters from the membrane (Bryson and Wilde 1996).
The purpose of this study was to investigate the changes occurring in the ovary in PCOS and to identify the positive/negative effects of AMI, a tricyclic antidepressant, in the treatment of PCOS symptoms.
2.1. Experimental animals and procedure
This study was supported by the Ondokuz Mayıs University Project Management Office (PYO.TIP.1904.19.002) and was approved by the Experimental Animal Studies Ethics Committee of Ondokuz Mayıs University (HADYEK 2017/54, 02.03.2018). It was used 40 female Wistar albino 12 weeks-old 200-250g rats and divided into five equal groups, control (Cont), sham (SHAM), AMI, polycystic ovary syndrome (PCOS), and PCOS plus AMI treatment (PCOS + AMI). The sham groups received a single injection of 0.2 mL sesame oil. For the induction of PCOS, the PCOS and PCOS + AMI groups received a single injection of 4 mg/kg estradiol valerate dissolved in 0.2 mL sesame oil. The AMI groups received 10 mg/kg AMI injections for 30 days. The rats were housed in standard plastic cages at 50% ± humidity at room temperature in a 12-hour light/12-hour dark cycle.
2.2.Tissue procedures and analysis
At the end of experimental procedures, the rats were anesthetized with intramuscular ketamine (10 mg/100 g body weight; Sigma Chemical Comp. St. Louis, MO, USA) and prilocaine hydrochloride (0.25 mg/100 g body weight; Sigma Chemical Comp., St. Louis, MO, USA). The ovarian tissues were removed, and 1 cc blood was collected from the heart for biochemical analysis. For the light microscopic and stereological analysis, tissue samples were processed with graded alcohol and xylene (Sigma Chemical Comp, St. Louis, MO, USA) and embedded in paraffin (Merck, Darmstadt, Germany). Sections were cut to a thickness of 10 or 20 µm using a rotary microtome (Leica RM 2135, Leica Instruments, Nussloch, Germany). The slides were stained with hematoxylin and eosin for light microscopic and stereological analysis.
For transmission electron microscopic analysis, ovarian tissue samples were passed through graded acetone and propylene oxide (Sigma Chemical Comp., St. Louis, MO, USA), processed with osmium, acetone, propylene oxide, and araldite, and embedded in resin blocks. Semi-thin sections of 0.5 µm and thin sections of 70 nm were taken from these blocks. Semi-thin sections were stained with toluidine blue, while thin sections were made visible using uranyl acetate and lead citrate solutions. For scanning electron microscopic analysis, one ovary from each group was passed through acetone series and then plated with gold-palladium.
2.3.Biochemical analysis
For the biochemical analysis, 1 cc blood samples collected from the animals’ hearts were centrifuged at 2000 rpm for 15 minutes. Levels of LH, FSH, superoxide dismutase (SOD) (Sunred Biological Technology Co., Ltd, Shanghai, China), and catalase (CAT) (Cayman Chemical Company, Germany) in the serum samples were analyzed using the procedures described with the kits.
2.4.Stereological analysis
Volume analysis
The Cavalieri method was used for volume analysis. Sections of ovarian tissue 5 µm in thickness were taken with a microscope with a camera attachment (Leica DM 7000, Leica Microsystems GmbH, Germany) and opened in Image J software. A 9000 µm2 point counting grid was used to estimate cyst, corpus luteum, and follicle volumes.
The grid was randomly dropped onto the image, and those points intersecting with the area of interest were counted. The volume was calculated using the formula V = t x a(p) x ∑p, where V = volume, t = section thickness, a / p = area represented by a point in the area measurement scale, and ∑P = the total number of points falling on the surfaces of the slices.
Estimation of follicle number
Twenty-five micrometer-thick sections were used for follicle number estimation. Follicle numbers were calculated using the optical fractionator method in the Stereoinvestigator system. The optical fractionator involves counting the particles formed by the optical dissector. Analyses of primordial, preantral, and antral follicles were performed using this method, which allowed particles within a given counting frame to be counted. In the analysis, the counting frame (x, y) was 120x120 µm in size, the step interval (x, y) was 200 x 200 µm, the dissector height was 10 µm, and the upper and lower safety intervals were 3 µm.
2.5.Statistical analysis
The numerical data obtained were analyzed on SPSS software (SPSS version 21.0; SPSS Inc., Chicago, IL, USA) and expressed as mean ± standard error. The Shapiro-Wilk test was used for normal distribution assumptions. Normally distributed data were compared between the groups using One-Way analysis of variance (ANOVA) and the Tukey test. The Kruskal Wallis and Tamhane tests were used to compare multiple groups without normal distribution. p values < 0.05 were considered statistically significant and p values < 0.01 values were considered highly significant.
2.6.Ethical Approval
This animal study was approved by the Experimental Animal Studies Ethics Committee of Ondokuz Mayıs University (HADYEK 2017/54, 02.03.2018).
3.1.Weight change results
The animals’ weights were recorded at the beginning and end of the experiment. Weights increased at the end of the experiment in the PCOS group (p < 0.05). No significant differences were determined in the other groups (p > 0.05) (Fig. 1).
3.2.Stereological Results
Volume result
A statistically significant difference was observed between the Cont and PCOS groups, and the PCOS and AMI groups in terms of the volume of the corpus luteum in the ovarian cortex (p < 0.05; One-Way ANOVA), while a highly significant difference was observed between the PCOS and PCOS + AMI groups (p < 0.01; One-Way ANOVA) (Fig. 2). The volume of the corpus luteum was higher in the PCOS group than in the Cont and PCOS + AMI groups.
No statistically significant difference was observed between the groups in terms of primordial and antral follicles volumes (p > 0.05; One-Way ANOVA). The preantral follicle volume results from the PCOS group were significantly higher than those in the Cont group (p < 0.05; One-Way ANOVA) (Fig. 3).
The thin granulosa layer, thickened theca layer, and degenerative nuclear structures were determined for cyst volume analysis (Wu et al., 2014; Wang et al., 2012; Lim et al., 2011). No statistically significant difference was observed among the groups (p > 0.05; One-Way ANOVA) (Fig. 4).
Follicles numbers
The numbers of follicles in the ovarian tissue were calculated using the optical fractionator method. No significant difference was determined between any of the groups in terms of numbers of primordial and preantral follicles (p > 0.05; One-Way ANOVA). A significant difference between the Cont and PCOS, PCOS and Ami, and AMI and PCOS + AMI groups in terms of antral follicle numbers (p < 0.05; One-Way ANOVA), while a highly significant difference between the Sham and PCOS + AMI groups (p < 0.01; One-Way ANOVA) (Fig. 5).
3.3.Biochemical Results
In order to define the levels of FSH and LH, which play a role in the mechanism of PCOS, and also the oxidative stress that can be caused by PCOS, CAT and SOD values were analyzed biochemically from blood sera, and the results are expressed in IU/L in the graphics (Fig. 6). There were no significant differences between any of the groups in terms of LH levels (p > 0.05; One-Way ANOVA) (Fig. 6b). However, serum FSH levels differed significantly between the Cont and PCOS + AMI groups (p < 0.05; One-Way ANOVA) and highly significantly between the Cont and PCOS groups (p < 0.01; One-Way ANOVA) (Fig. 6a). Highly significant differences in CAT levels were observed between the Cont and Sham, PCOS and AMI, PCOS + AMI and PCOS, and Sham and AMI groups (p < 0.01; One-Way ANOVA) (Fig. 6c). The only significant difference in serum SOD levels was observed only between the Cont and PCOS + AMI groups (p < 0.05; One-Way ANOVA) (Fig. 6d).
3.4. Histopathological analysis
Light microscopic ovarian tissue results
While the epithelium surrounding the ovary appeared in the form of a healthy, cuboidal epithelium in the Cont and Sham groups, it appeared as a cubic prismatic structure in the PCOS and PCOS + AMI groups. A simple squamous epithelium was observed in the Ami group. The tunica albuginea layer in the PCOS, PCOS + AMI, and AMI groups was thicker than in the Cont and Sham groups (Fig. 7). When the follicle structures were examined, the oocyte borders were clear, the zona pellucida and the first row granulosum cells surrounding the oocyte, the theca layers, were healthy, and the glassy membrane could be easily distinguished in the Cont and Sham groups (Fig. 7,8). In the PCOS and PCOS + AMI groups, deformed granulosa cells were observed in which the theca layers of the follicles were more prominent and thought to have entered apoptosis within the follicles. The borders of the oocytes in the follicles could not be distinguished, and the zona pellucida layer was indistinct. In the AMI group, the follicles were degenerated, similarly to the PCOS group. The density of hilus cells characterized by spherical nuclei and lipid droplets in their cytoplasm in the PCOS and PCOS + AMI groups was particularly remarkable. In the PCOS and PCOS + AMI groups, the density of cystic structures characterized by a thin layer of granulosa in the cortex and a thick theca follicle was also striking. The presence of cystic structures similar to those in the PCOS group was also noteworthy in the AMI group (Fig. 8).
Ovarian tissue electron microscopic results
Examination of the surface properties of the ovary revealed that the epithelium around the ovary was a simple cuboidal epithelium with a healthy appearance in the Cont and Sham groups. However, in the PCOS group, the outer surface was severely damaged, and the simple cuboidal epithelium had assumed other forms in some places. In the AMI and PCOS + AMI groups, the epithelium was mostly healthy. In addition, the epithelium in the PCOS + AMI group had indistinct borders, and the morphology of the connective tissue elements was severely impaired (Fig. 9).
Examination of the corpus luteum structures in the ovarian cortex it revealed that the luteal cells in the Cont and Sham groups exhibited a healthy structure with typical endoplasmic reticulum, secretory vesicles, and mitochondria. In the PCOS group, the cytoplasm and mitochondria of the luteal cells were darkly stained, and the agranular endoplasmic reticulum cisterns were enlarged. The corpus luteum was healthy in the AMI and PCOS + AMI groups (Fig. 9).
PCOS is an endocrine disorder seen in the majority women of reproductive age and in which cystic ovaries are observed in cases in which follicular development steps are not fully completed due to increased androgen levels. Various symptoms occur in patients with PCOS, but particularly obesity, anxiety, depression, and blood pressure problems. A high prevalence of obesity and depression is observed in patients with PCOS (Kerchner et al. 2009; Hung et al. 2014; Hart et al. 2015; Cooney and Dokras 2017).
Depression, one of the most frequently observed syndromes in patients with PCOS, is a mental illness that manifests with a decreased sensitivity to stimuli and reinforcement hopelessness and pessimism. The mechanism involved is a complex one. Recent studies have shown a very close relationship between PCOS and depression. Individuals with PCOS are eight times more likely to be depressed than healthy individuals (Kerchner et al. 2009; Hung et al. 2014; Hart et al. 2015; Cooney and Dokras 2017). The relationship between PCOS and depression has generally been examined in clinical studies, and analyses have been performed using anxiety and depression diagnostic questionnaires. The high prevalence of depression observed in patients with PCOS can be attributed to three factors - high androgen levels observed in PCOS, insulin resistance, and infertility (Kerchner et al. 2009; Hung et al. 2014; Hart et al. 2015; Cooney and Dokras 2017). The present study investigated the effects of AMI on ovarian tissues in a rat model of PCOS.
The animals were weighed at the beginning and end of the experiment. The weights of the animals in the PCOS group increased at the end of the study compared to the initial values. No significant difference was observed in the other groups. Considering the obesity rate of 40–70% in adolescents with PCOS, this finding in the present study is quite possible (Vatopoulou and Tziomalos 2020). Sam (2007) showed that exposure to high androgen levels in post-menopausal women causes an increase in visceral adipose tissue. The weight gain occurring in PCOS has been ascribed to insulin and glucose metabolism (Melekoglu et al. 2019; Zeng et al. 2019). No increase in weight between the beginning and end of the experiment in the PCOS + AMI group in this study, suggesting that AMI may have a positive effect on weight gain in patients with PCOS. This can be explained by the reduction of depression and the decrease in food intake that comes with depression. It is also possible that this pathway proceeds via neuropeptide Y (NPY), implicated in both PCOS and depression.
Corpus luteum volumes, follicle volumes and numbers, and cystic structure volumes were analyzed for the quantitative evaluation of morphological changes in the ovary in all groups. Stereological analysis showed that PCOS produced no change in primordial and antral follicle volumes, but increased preantral follicle volume. No statistically significant difference was also observed between the AMI and PCOS + AMI groups. Analysis of follicle numbers and volumes revealed no statistically significant difference between any of the study groups. Although an increase in volume was observed in the PCOS group, no statistically significant difference in follicle numbers was observed, although a decrease was observed in the PCOS group compared to the Cont group. A study involving subcutaneous injection of dehydroepiandrosterone in rats to induce a PCOS model observed large primary follicles at ovarian morphological examination (Misugi et al. 2017). Considering all this information and the data from the present study data, it may be concluded that PCOS increases preantral follicle volume without causing any increase in the number of preantral follicles. This situation results from hormonal imbalances and increased androgen levels under the effect of insulin, changing the FSH/LH balance and inducing follicle development. Similarly, although comparable results were observed between the PCOS + AMI and PCOS groups, the difference was not statistically significant. This shows that AMI treatment is insufficient to ameliorate the adverse effects of PCOS at the primordial and preantral follicle levels.
Analysis of antral follicle numbers in this study revealed a decrease in the PCOS group compared to the Cont group, but there was no significant difference between the PCOS + AMI group and the PCOS group. This situation can be regarded as a natural result of PCOS development. Follicles with a diameter of 2–5 mm in the developmental stage form cystic structures. The absence of a significant difference between the PCOS + AMI group and PCOS group shows that AMI treatment does not affect antral follicle formation.
In addition to follicular changes, the most important parameter in PCOS pathology is the cystic structures located in the ovarian cortex that have not yet completed their development. Many studies show that patients with PCOS have cystic follicles with a diameter of 2 mm (Alsamarai et al. 2009; Azziz 2018; Behmanesh et al. 2019). Although the histological structure of the cystic follicle is evaluated differently in many studies, the thin granulosa layer and the very thick and well-developed theca internal layer are characteristic features of cystic structures (Wang et al. 2017; Manneråset al. 2007; Yaba and Demir. 2012). In their study of 40 rats, Behmanesh (2019) induced a PCOS model with estradiol valerate and reported a cystic structure with impaired follicular maturation due to an altered FSH/LH ratio in the animals in the PCOS group (Behmanesh et al. 2019). Those authors observed increases in cystic structures as well as preantral follicles in rats exposed to the PCOS model. Although there was no statistically significant difference between the groups in the analyses performed in that study, small-volume cystic structures with the criteria described were observed in the PCOS and PCOS + AMI groups (Wu et al. 2014). Due to the changes in FSH and LH levels, the follicular transformation into a cystic structure without development explains the small volume of cystic structures observed in the PCOS group compared to other groups in that study. Takahashi et al. reported cystic structures with small diameters in ovaries with PCOS (Takahashi et al. 1994). The small volume cystic structures observed in the PCOS + AMI group in that study also showed that the application of AMI was insufficient. Other studies have reported the presence of a thickened tunica albuginea layer among the histopathological findings, together with a prominent theca follicle and hyperplasia, increase interstitial cells, and increases in the numbers of cells and volume of the corpus luteum are observed in groups with PCOS (Takahashi et al. 1994; Wang et al. 2012zükara 2013). In the present study, the presence of a thick tunica albuginea was particularly noteworthy at histopathological evaluation of the ovaries from the PCOS group. Dark-stained cells with an angular structure that had lost their spherical structure, most of which had unclear nuclear borders, were observed among the granulosa cells in the follicles. In addition, the outer borders of the oocyte structures in the follicles could not be distinguished. Fragmentation of the nuclei, loss of the nuclear membrane, and a difficult-to-select zona pellucida structure were observed in this group. Lipid vacuoles of theca cells increase in number in PCOS pathology (Gözükara 2013). Histopathological results similar to those in PCOS were also observed in the Ami group, which exhibited oocyte and follicular damage, a thick tunica albuginea layer, and thick and irregular granulosa cells. All these findings suggest that the effects of AMI on the ovary snow need to be investigated more extensively. At histopathological analysis of the PKOS + AMI group, although the healthy follicle numbers and thick corpus luteum were higher than in the PCOS group, some damaged follicular and cystic structures were also observed. The presence of degenerated cells and structures at ultrastructural examination showed the negative effect of AMI on PCOS. These results show that the effect of PCOS cannot be reduced by AMI treatment. In addition to all these results, intense hilus cells, which are involved in testosterone secretion, were found in the PCOS group. Although the number of hilus cells in a healthy ovarian tissue varies, they increase in number in the postmenopausal ovary (Gilks and Clement 2012). Considering the increase in testosterone levels, a characteristic of PCOS, a greater hilus cell density is a possible outcome. This result reveals the need for further research into hilus cell structures and testosterone levels in ovarian structures with PCOS.
The majority of studies of PCOS have observed marked increases in serum LH levels (Azziz 2018; Yıldırım and Memişoğulları 2011; Teede et al. 2010). A study of 20 individuals with PCOS showed that LH levels increased compared to the healthy group and that the ovarian morphology changed accordingly. Another study involving three-week-old Wistar albino rats, produced a PCOS model observed that LH levels increased significantly compared to the control group, although the change in FSH was not significant (Sun et al. 2016). Another study of D vitamin treatment in a PCOS model reported increases in both LH and FSH levels (Çelik 2016). In the present study, both LH and FSH levels decreased in the PCOS group compared to the Cont group. However, only the decrease in FSH levels was statistically significant. Recent studies suggest that the principle marker of androgen level increases in PCOS is a rise in free testosterone levels rather than an increase in LH. A previous study showed that LH levels decreased in PCOS (Tessaro et al. 2015). The FSH and LH results in the present study were compatible with the previous literature. However, the fact that testosterone levels were not determined in this study cannot fully explain this situation. No statistical difference in FSH levels was found between the PCOS and PCOS + AMI groups, although there was a significant difference between the Cont and PCOS + AMI groups. Although this was not statistically significant, the LH value in the PCOS + AMI group was closer to that in the Cont group. This suggests that AMI treatment may cause changes in FSH and LH levels.
The association between PCOS and oxidative stress has been demonstrated by numerous previous studies (Gonzalez et al. 2006; Kuşçu and Var 2009; Blair et al. 2013). Gonzalez et al. showed that the hyperglycemia observed in PCOS patients affects androgen levels by stimulating the secretion of ROS from mononuclear cells. The present study revealed an increased CAT level in the PCOS and AMI groups compared to the Cont group, but there was no significant difference in the PCOS + AMI group. No significant differences in SPD levels were found among any of the groups. The increase in CAT levels in the PCOS group is consistent with previous studies in the literature.
Although a statistically insignificant decrease in SOD is not expected, there are studies reporting decreased SOD levels in individuals with PCOS (Liu and Zhang 2012; Seleem et al. 2014). Several studies have reported that AMI exacerbates oxidative stress (Viola et al. 2000; Cordero et al. 2010; Mytych et al. 2019; Sehonova et al. 2019). The results in the PCOS + AMI group showed that the use of AMI together with PCOS causes a decrease in oxidative stress in rats, even though AMI and PCOS individually cause an increase in oxidative stress.
A previous study of ovarian morphology in rats with PCOS involving AMI showed that the numbers of follicles and the corpus luteum in the PCOS group decreased compared to the control group, while the number of atretic follicles and cystic follicles was significantly higher than the control group (Li et al. 2019). Li et al. stated that the total number of preantral and antral follicles, and the size of the corpus luteum increased significantly in the group treated with AMI, while the number of atretic follicles and cystic follicles decreased. However, these results are not consistent with our stereological and histopathological results. Considering all the data obtained from the PCOS + AMI group in this study, AMI treatment appears to have no positive effect on ovarian morphology.
In conclusion, AMI appears to exhibit no protective or curative effect against the deleterious effects of PCOS on ovarian follicles and hormonal levels, although it may exhibit antioxidant activity capable of reducing the oxidative stress that may occur with PCOS. In addition, the use of AMI alone caused morphological changes in the ovary due to oxidative stress.
Conflicts of Interest
The authors report no conflicts of interest.
Funding
This study was funded by the Project Management Office of Ondokuz Mayıs University (S. Kaplan; PYO.TIP.1904.19.002).