DOI: https://doi.org/10.21203/rs.3.rs-1989089/v1
Study investigated the effects of curcumin treatment on diabetic ovaries at different periods of the disease. Fifty-six female Wistar albino rats (250–300 g) aged 12 weeks were divided into seven groups. No treatment was applied to the control group. The sham group was given 5 mL/kg of corn oil, and the curcumin group 30 mg/kg curcumin. In the DM groups, diabetes was induced by a single intraperitoneal dose of 50 mg/kg streptozotocin (STZ). The DM-treated groups received 30 mg/kg curcumin after either seven days (DC1 group), or 21 days (DC2 group), or simultaneously with STZ injection (DC3 group). Numbers of follicles in the ovaries were estimated using stereological method. FSH, LH, and SOD levels and CAT activity were measured in serum specimens. Follicle numbers and volumes of corpus luteum, blood vessel and cortex volumes, gonadosomatic index, and FSH and SOD levels all decreased significantly in diabetic ovaries, while relative weight loss, connective tissue volume, and CAT activity increased (p < 0.01). Curcumin treatment had a protective effect on the number of primordial follicles in the DC2 group and on antral follicle numbers in the DC3 group. Curcumin also exhibited positive effects on CAT activity and SOD levels, blood glucose levels, and corpus luteum, connective tissue and blood vessel volumes in the DC2 and DC3 groups. Curcumin also ameliorated FSH levels in the DC1 and DC3 groups (p < 0.01). Curcumin exhibits protective effects on ovarian structures and folliculogenesis, especially when used concurrently with the development of diabetes or in later stages of the disease.
Diabetes mellitus (DM) is a metabolic disease the development of which involves both hereditary and environmental factors. It is characterized by excessive blood glucose levels and can result in various complications, such as retinopathy, nephropathy, neuropathy, and cardiovascular diseases (Herman 2007). DM has also been reported to adversely affect both male and female reproductive activities (Wu et al. 2017). Delayed oocyte maturation and stages of meiosis, increased follicular and granulosa cell apoptosis, disorders in oocyte-granulose communication and follicle development, estrus cycle disruptions (especially oligomenorrhea), anovulation, decreased mating ability, mild hyperandrogenism, polycystic ovary syndrome, a decreased live birth rate, and early menopause have been observed in diabetic rats (Codner et al. 2012). Insulin stimulates the secretory activity of the gonadotropic axis by directly affecting gonadotropin releasing hormone (GnRH) neurons (increasing Kiss1 / kisspeptin expressions) (Pralong 2010). Diabetes has been reported to suppress hypothalamic expression of the Kiss1 gene in male and female rats treated with streptozotocin (STZ) (Castellano et al. 2009).
Although many mechanisms have been proposed to explain both the development of diabetes and the pathogenesis of its complications, the most important of these involves increased free radicals in tissue (Volpe et al. 2018). Studies have shown that diabetic individuals are susceptible to oxidative stress. Increased free radicals in tissue as a result of high blood glucose levels are known to exacerbate lipid peroxidation. A decrease in intracellular antioxidant enzyme activities has also been reported as a result of diabetes (Tiedge et al. 1997).
The principal aim of treatment is to minimize and delay complications caused by diabetes. Curcumin is a phenolic compound obtained from the roots of the turmeric plant and is known to exhibit positive effects on various chronic diseases associated with inflammation and oxidative stress (Soleimani et al. 2018). It has also been studied as a potential therapeutic agent in experimental diabetes and in the treatment of diabetic complications due to its effectiveness in reducing hyperglycemia and hyperlipidemia (Zhang et al. 2013). Curcumin suppresses blood glucose levels in diabetic patients, increases the function of pancreatic β cells, and delays the production of reactive oxygen species (ROS) (Perez-Torres et al. 2013; Meghana et al. 2007). It is also thought to possess phytoestrogenic effects, and can therefore interact with the endocrine system, affecting the hypothalamic-pituitary-ovarian axis and repairing these disorders (Bachmeier et al. 2010). Although numerous studies have investigated the effects of DM on ovarian histopathology, the problem remains unsolved. Comprehensive studies are still needed in this area. The purpose of the present study was to determine the structural, functional and hormonal changes occurring in the diabetic ovary and to investigate the possible protective effects and mechanisms of curcumin, with its potential anti-hyperglycemic and phytoestrogenic effects, on the diabetic ovary at the structural and biochemical levels.
An experimental animal model was employed to investigate how curcumin application to the diabetic ovaries would affect follicle counts and volumes, and blood vessel and connective tissue volumes in the ovary. These were investigated in a quantitative manner using the unbiased cell count method, the optical fractionator counting technique, and the Cavalieri principle. The stereological techniques used are design-based methods based on statistical principles (Boyce et al. 2010). In addition, the fact that these methods are based on the principle of impartiality further enhances the value and reliability of the data obtained (Gundersen 1986). The physiological effectiveness of curcumin administration on the hypothalamic-pituitary-gonadal axis in diabetic individuals was also investigated in the present study. The fact that changing gonadotropin levels in diabetic individuals can be regulated by curcumin treatment plays an important role in maintaining ovarian functions. Oxidative stress parameters were measured using biochemical techniques. This study also evaluated the biochemical effect of DM on reproductive activities and the protective antioxidant activity of curcumin.
The results of this study may encourage the use of curcumin to compensate for insulin resistance and hyperglycemia observed in the treatment of diabetic individuals. We hope that our findings will contribute to the development of new therapeutic approaches in cases of diabetes-related infertility and to the reduction of ovarian toxicity, or to the preservation of ovarian quality and reproductive activities in diabetic individuals.
All experimental procedures were carried out at the Ondokuz Mayıs University Experimental Animal Application and Research Center (DEHAM), Turkey, following approval by the Ondokuz Mayıs University Experimental Animals Ethical Committee (no. 2017/53 dated 24.11.2017). The welfare of the experimental animals and relevant ethical principles were taken into account during the procedures performed in the entire experiment. The experimental animals were allowed ad libitum access to water and feed in a 12:h light-dark cycle at a room temperature of 22 ± 2°C and relative humidity of 40–60%.
Fifty-six female Wistar albino female rats aged 12 weeks and weighing 250–300 g were used in the experiment. These were randomly divided into seven groups of eight animals each:
Group 1: Control
No procedure was performed on the animals in this group.
Group 2: Sham
Corn oil was administered by intragastric gavage for 14 days.
Group 3: Curcumin
Curcumin (30 mg/kg) was dissolved in corn oil and administered by intragastric gavage for 14 days (Sharma et al. 2006).
Group 4: Diabetes mellitus group
Diabetes was induced with a single intraperitoneal dose of 50 mg/kg STZ (Cayman, 13104)
Group 5: Diabetes mellitus + Curcumin (DC1)
Seven days after induction of experimental diabetes with STZ, 30 mg/kg curcumin was dissolved in corn oil and administered by intragastric gavage for 14 days.
Group 6: Diabetes mellitus + Curcumin (DC2)
Twenty-one days after induction of experimental diabetes with STZ, curcumin (30 mg/kg) was dissolved in corn oil and administered by intragastric gavage for 14 days.
Group 7: Diabetes mellitus + Curcumin (DC3)
Simultaneously with the injection of STZ to induce experimental diabetes, curcumin (30 mg/kg) was dissolved in corn oil and administered by intragastric gavage for 14 days.
In order to induce the experimental diabetes model, a single dose of 50 mg/kg STZ (Cayman, 13104) was dissolved in freshly prepared 0.1 M citrate buffer (pH: 4.0) and administered i.p. following overnight fasting (Kandemir et al. 2017). Access to feed was permitted after STZ injection (Andrade Cetto et al. 2000).
The experimental animals were given 5% glucose with drinking water to prevent hypoglycemia in the first 12–24 h after STZ injection. Seventy-two hours after STZ administration, blood samples collected from the tail veins of rats fasted overnight were dropped onto a glucometer (GI 300 Plus MED) strip to measure fasting blood glucose levels. Three measurements were taken from each animal, and those with a fasting blood glucose level of 250 mg/dL or above were considered diabetic and included in the study (Yuan et al. 2018). The control, sham, and curcumin groups received 0.1 M citrate buffer (pH: 4.0) i.p.
Body weights and fasting blood glucose values of animals from all groups were measured at the beginning and end of the study. At the end of the experiment, blood samples were collected following the administration of 90/10 mg/kg ketamine (Ketasol®, Richter Pharma, Wels, Austria) and xylazine (Citanest®, AstraZeneca, Istanbul, Turkey) anesthesia.
Tissues from the anesthetized rats were fixed by applying a mixture of 4% paraformaldehyde + 2% glutaraldehyde to the circulation via the transcardiac route. After the perfusion process was completed, the bilateral ovary tissues were extracted, and the surrounding fat and connective tissues were trimmed. The dissected right ovaries were placed into jars containing 4% paraformaldehyde and 2% glutaraldehyde for stereological analysis. One of the left ovarian tissues was also set aside for microscopic examinations. The tissues obtained were subjected to a one-week post-fixation process. Fixation solutions in the bottles were renewed every three days in order to enhance the fixation quality.
Relative weight loss percentages were calculated from the body weights measured at the beginning and end of the study.
The gonadosomatic index (GSI) measures the sexual maturity of animals. The GSI value in the present study was calculated using the formula (Walker et al. 2013)
where GW represents the sum of right and left ovary weights, and BW represents body weight
After the post-fixation process, the right ovaries were placed into tissue processing cassettes for histological tissue examination. They were then washed for half a day to remove the fixation solution. In the dehydration phase, the tissues were passed through increasing graded alcohol series (70%, 80%, 96%, and 100%) and then cleared with xylene. Tissues were infiltrated and embedded with paraffin. Another individual enumerated each tissue from all experimental groups in order to perform a blind study during the tissue-labeling phase. The group names were not used during labeling.
Sections 25 µm in thickness were from taken the right ovary tissue blocks using the systematic random sampling (SRS) method with a rotary microtome (Leica RM 2235, Vienna, Austria) for stereological analysis. The accuracy of the sampling interval was determined by means of a pilot study. The sections obtained were stained with hematoxylin and eosin (H&E) and made ready for analysis.
In contrast to other quantitative techniques, stereological methods are design-based and unbiased methods based on statistical principles and stochastic geometric theory rather than assumptive-based approaches. The "optical fractionator method", one of the stereological methods, was therefore used to determine the numbers of follicles in the ovaries, while the "Cavalieri method" was employed for the volume parameter.
In order to calculate the total number of follicles in the ovary, instead of evaluating all the sections obtained from the ovary in accordance with the "efficiency" principle of stereological methods, Systematic random sampling (SRS) was performed at a ratio of 1/2 in the sections. The sampling rate of the sections to be selected for the study was determined by conducting a pilot study. Next, 25 µm-thick sample sections were taken from all blocks with ½ sampling in accordance with the SRS rule. For this process, after the first 25 µm section on the slide had been selected randomly, the other 25 µm sections were taken at 1/2 systematic intervals. Approximately 40-sampled sections were thus obtained from each tissue block.
The numbers of follicles in the sections were calculated using the optical fractionator probe in the computer-aided and special software stereology analysis system (Stereoinvestigator 9.0., MicroBrield Field; Colchester; USA) in the Recep Tayyip Erdoğan University Faculty of Medicine Department of Histology and Embryology. The parameters to be used in the study were determined by a pilot study (Table 1). During analysis, the morphological classification of ovarian follicles was based on the shape and number of layers of granulosa cells surrounding the oocyte (Myers et al. 2004).
After determining the appropriate parameters for counting, using the optical fractionator probe in the Stereo Investigator program (Stereo Investigator 9.0., MicroBrield Field; Colchester; USA), the boundaries of the area to be counted (the outer boundaries of the ovarian section) under the 1X objective lens were marked out. The size of the counting frame, the dissector height, and the numerical parameters were entered onto the relevant parts of the software and counted under a 63X objective lens. Follicle counting with the optical fractionator method was based on the presence or absence of the nucleus of the oocyte cell. Oocytes with a nucleolus were included in the count, while those without not were excluded. The counting frame was systematically randomly placed on the sections along the determined step intervals using the Stereo Investigator software (Fig. 1A). In order to avoid affecting counting artifacts such as shrinkage and swelling that may occur in the cross-sections during histological procedures (sectioning, staining, etc.), counts were made at 10-µm depths after leaving a 3 µm safety margin from the upper surface of the section. The thickness of section was also measured at each probing site. Follicle counting were performed by applying the rules of the unbiased counting frame in the areas selected by means of the SRS. Oocyte nuclei falling into the count frame or intersecting the free edges of the frame were included in the count (Gundersen 1986)(Fig. 1B).
The total number of follicles "N" was calculated using the formula (West and Gundersen 1990):
According to the formula, the total number of follicles per ovary (N) was determined by multiplying the number of disector particles (number of oocyte nuclei) (\({{\Sigma }Q}^{-}\)) counted after each volume during optical fractionator application and the fraction coefficients (sectional sampling fraction, ssf; area sampling fraction, asf; thickness sampling fraction, tsf).
Corpus luteum, vascular, connective tissue, cortex, medulla, preantral and antral follicle volumes in the ovary sections were estimated using the Cavalieri principle, one of the unbiased stereological counting techniques. Calculation of the reference volume with this method is based on multiplying the total surface area (\({\Sigma }{a}_{i}\)) of the region of interest in the cross sections by the section thickness (\(\widehat{t}\)). A "point" area measurement scale was used to find the surface area of the parameter of interest (corpus luteum, vascular, connective tissue, cortex, medulla, preantral and antral follicle volumes) in the section. The point area measuring scale consists of systematic equidistant points in the form of a cross. The midpoint of each cross represents a point on the grid. Each of these points represents a unit scale area between the four points. Since the fixed distance between the points on the sections is known (P(a)), if the point counting volume estimation grid is placed over a section image, it will intersect with as many points as the area represented by the relevant structure (ΣPi). Multiplication of the total number of points superimposed on section (ΣPi) by (P(a)) yields the total area on that section (Gundersen 1986; Gundersen and Jensen 1987).
In this formula, Vref represents the total volume of the structure of interest, ΣPi the total number of points corresponding to the section projections, \(\widehat{t}\) the average section thickness, and P (a) the area represented by a point. Accordingly, sections were selected with a 1/4 re-sampling from the ovary sections selected by SRS, which were previously made at the rate of 1/2, and volume calculations were made in these sections. The images were then captured with 1/4 sampling on a 4X objective lens (Olympus BX43, Center Valley, PA) in the sample sections selected for calculating cortex and medulla volume. Following the pilot study, the area represented by a single point on the “point counting grid” was determined as 10,000 µm2. A point counting volume estimation grid was placed over the cross-sectional images in Image J (Image Processing and Analysis in Java, NIH, USA) (Fig. 2). The cortex and medulla volumes were calculated by multiplying the number of points superimposing on the cortex and medulla by the section sampling rate, the area sampling rate, and the area represented by a point.
For calculating the volumes of the corpus luteum, blood vessels, connective tissue, and preantral and antral follicles, images of the same sections were taken with a 20X objective lens and using a 1/5 sampling rate. Following a pilot study, the area represented by a point on the test grid was determined as 9000 µm2, and a point counting volume estimation grid was placed over the cross-sectional images in Image J (Image Processing and Analysis in Java, NIH, USA) program. The volumes were calculated by multiplying the number of points superimposing on the structures of interest by the section sampling rate, the area-sampling rate, and the area represented by a point.
Biochemical analyses were performed at the Ondokuz Mayıs University Faculty of Medicine Department of Biochemistry. At the end of the experiment, but before perfusion, 1 mL blood samples were collected from the hearts of anesthetized rats and placed into EDTA tubes. The samples were then separated by centrifugation (Hettich, Tuttlingen, Germany) for 2000 rpm at 4 oC for 15 minutes. Serum samples were stored at -80 oC until investigation. CAT activity, and SOD, FSH and LH levels were measured in the serum samples.
A CAT assay kit (Cayman Chemical Company, USA) was used to compare antioxidant enzyme activities between the groups. The absorbent values of the serum samples were measured at 540 nm with a microplate reader (BioTek Synergy 4, BioTek Instruments Inc., Vermont, USA), and enzyme activity was expressed as µM protein.
A SOD ELISA kit (Bioassay Laboratory Technology, Shanghai, China Cat. No. E0168Ra) was used to measure SOD levels. The amount of enzyme was measured at a 450 nm wavelength using a microplate absorbance reader (SunriseTM, Tecan Trading AG, Männedorf, Switzerland) and was calculated as ng/mL protein.
ELISA (Shanghai Sunred Biological Technology Co. Ltd, Shanghai, China) kits were used to measure FSH and LH levels in serum samples. FSH enzyme (IU / L) and LH (mIU / mL) concentrations were measured at 450 nm.
For histopathological examinations, left ovary tissues were fixed in glutaraldehyde solution (2%), and then divided into 1 mm3 sections and washed with 0.1 M phosphate buffer. The second fixation was performed with osmium tetroxide (OsO4) (1%) for 90 minutes. For resin tissue processing, following dehydration with graded alcohol series, tissues treated with propylene oxide were placed into araldite (an embedding medium suitable for thin sectioning), which was expected to infiltrate the tissues. The tissue particles were placed on the tip of the bullet-shaped molds, which were placed into an oven and allowed to harden at 55–60 ° C (Tables 6–7). Semi-thin sections (0.5 µm) were taken from the prepared resin blocks with an ultramicrotome (Leica EM UC7, Vienna, Austria) and stained with toluidine blue. Images were captured using a microscope with a camera attachment (Olympus BX43, Center Valley, PA, USA) for histopathological examinations.
2.11. Statistical analysis
The numerical data obtained were subjected to statistical analysis using SPSS software (SPSS version 21.0; SPSS Inc., Chicago, IL, USA). The data were expressed as mean ± standard error. The Shapiro-Wilk test was used to test normality. For intergroup comparisons, data conforming to normal distribution were evaluated using the One-Way ANOVA and Tukey tests. The Kruskal Wallis and Tamhane tests were used to compare multiple groups not exhibiting normal distribution. p < 0.05 and p < 0.01 were considered statistically significant.
Relative weight loss (%) was calculated based on the body weights measured at the beginning and end of the experiment. The relative weight loss in the DM group was higher than that in the control, DC2 and DC3 groups (p < 0.05), (Table 2).
The GSI in the DM group was significantly lower compared to the control, sham and curcumin groups (p < 0.01). However, GSI values were significantly lower in the DC1, DC2 and DC3 groups compared with the DM group (p < 0.01), (Table 2).
Fasting blood glucose levels were higher in the DM group than in all the other groups (p < 0.01). Levels in the DC1, DC2 and DC3 groups were significantly lower than in the DM group (p < 0.01), (Table 2).
Primordial Follicle Number
The number of primordial follicles in the DM group decreased significantly compared to those in the control and DC2 groups (p < 0.01).
The number of primordial follicles in the DC1 group was significantly lower than in the control and DC2 groups (p < 0.01 and p < 0.05, respectively).
The number of primordial follicles in the DC2 group was significantly higher than in the DC3 group (p < 0.01). The number of primordial follicles in the control group was statistically significantly higher than in all the other groups (p < 0.01) (Table 3).
Preantral Follicle Numbers
A significant decrease was observed in the numbers of preantral follicles in the DM group compared to the control and sham groups (p < 0.01). However, no statistically significant difference was found between the DM, curcumin, DC1, DC2, and DC3 groups (p ≥ 0.05).
The number of preantral follicles in the control group was significantly higher than in all the other groups (p < 0.01). The number of preantral follicles in the sham group was also significantly higher compared to the DM, DC1, and DC2 groups (p < 0.01) (Table 3).
Antral Follicle Numbers
The number of antral follicles in the DM group was significantly lower than those in the control, curcumin, and DC3 groups (p < 0.01).
The number of antral follicles in the DC1 group was significantly lower than in the sham, curcumin, and DC3 groups (p < 0.01). Similarly, the DC2 group exhibited significantly fewer follicles than the DC3 group (p < 0.01). The number of antral follicles in the control group differed significantly from those in the other groups (except for the curcumin group) (p < 0.01) Table 3.
Total Number of Follicles
The total number of follicles in the DM group decreased significantly compared to the control, sham and curcumin groups (p < 0.01). The total numbers of follicles in the DC2 and DC3 groups were also significantly higher than in the DM group (p < 0.01).
The total number of follicles in the DC1 group decreased significantly compared to the control, sham, curcumin, DC2, and DC3 groups (p < 0.01). In the control group total follicle number was significantly higher than in all the other groups (p < 0.01) (Table 3).
Cavalieri Volume Estimation Results
Preantral Follicle Volumes
The preantral follicle volume in the DM group was significantly lower than in the control and curcumin groups (p < 0.01).
The preantral follicle volume in the DC1 group decreased significantly compared with the control group (p < 0.01). A difference was observed between the DC2 group and the control and curcumin groups (p < 0.05), but not with the other groups (p ≥ 0.05). A significant decrease was observed in the preantral follicle volume in the curcumin group compared to the control group (p < 0.05) (Table 4).
Antral Follicle Volumes
No statistically significant difference was observed between the groups in terms of antral follicle volumes (p ≥ 0.05) (Table 4).
Corpus Luteum Volume
Corpus luteum volumes in the DM group decreased significantly compared to the control, sham, and curcumin groups (p < 0.01). In addition, while there was no significant difference between the DM group and the DC1 group (p ≥ 0.05), the corpus luteum volume of the DM group was lower volume than in the DC2 and DC3 groups (p < 0.01). Similarly, the volume of the corpus luteum in the DC1 group was lower in than the control, curcumin, and DC3 groups (p < 0.01). The volume of the corpus luteum in the DC3 group was significantly lower than in the control group (p < 0.01). In addition, the corpus luteum volume in the sham group was significantly lower than that in the control group (p < 0.01), while the corpus luteum volume in the curcumin group was higher than that in the sham group (p < 0.05) (Table 4).
Connective Tissue Volumes
The connective tissue volume in the DM group increased significantly compared to the control and sham groups (p < 0.01). A significant decrease was observed in the DC1, DC2, and DC3 groups compared to the DM group (p < 0.01), Table 4.
Blood Vessel Volumes
A decrease was determined in blood vessel volume in the DM group compared to the control, DC2, and DC3 groups (p < 0.01).
Also, a significant decrease was determined in the sham group compared to the control group (p < 0.05), (Table 4).
Cortex and Medulla Volumes
The cortex volumes in the DM and curcumin groups decreased significantly compared with the control group (p < 0.05). A significant decrease was also observed in the DC1 and DC2 groups compared to the control group (p < 0.01). Stereological analysis revealed no significant difference in medulla volumes in any of the study groups (p ≥ 0.05) (Table 4).
Catalase Enzyme Activity
A significant increase was observed in CAT activity in the DM group compared to the control, sham, curcumin, DC2, and DC3 groups (p < 0.01), but no significant difference was found with the DC1 group (p ≥ 0.05). CAT activity in the DC1 group increased significantly compared with the control and DC3 groups (p < 0.01), (Table 5).
Superoxide Dismutase Levels
SOD enzyme levels in the DM group decreased significantly compared with all other groups (p < 0.01), (Table 5).
Serum FSH Levels
FSH levels in the DM group decreased significantly compared with the control, sham, curcumin, DC1, and DC3 groups (p < 0.01). FSH levels in the DC2 and DC3 groups exhibited no significant difference from those of the control, sham, curcumin, or DC1 groups (p ≥ 0.05).
FSH levels the DC1 group were significantly higher than in the DC2 group (p < 0.01). FSH levels in the DC2 group were significantly lower than in the control, sham, curcumin, and DC3 groups (p < 0.01), FSH levels the DC3 group were no different to those of the control, sham, or curcumin groups (p ≥ 0.05) (Table 5).
Serum LH Levels
LH levels in the DC2 group decreased significantly compared with the control, sham, and curcumin groups (p < 0.01) (Table 5).
Examination of the general histological structure of the ovaries from the control group revealed a prominent single layer of cubic epithelium surrounding the outside of the ovary and the tunica albuginea layer rich in connective tissue fibers, together with multiple irregular cytoplasmic extensions on the apical surface of the epithelium (Fig. 3A). Oocytes surrounded by a single row of follicle cells with distinct borders were observed in the center of the primordial follicle, immediately below the ovarian epithelium (Fig. 3A). The preantral follicles in the cortex exhibited a normal appearance, the oocyte was located in the center, and these were surrounded by the zona pellucida and granulosa cells (Fig. 3B). The well-preserved oocyte in the center of the antral follicles from the control group, the surrounding zona pellucida, and the corona radiata cell layer surrounding it in a single line could all be easily distinguished. Follicular fluid filling the antrum surrounded the cumulus oophorus (Fig. 3C). The structure of the theca follicle surrounding the follicles was also well preserved (Fig. 3D). In the corpus luteum, numerous normal-appearing, large polygonal-shaped granulosa lutein cells with eosinophilic cytoplasm, as well as surrounding theca lutein cells with well-preserved blood vessels were observed (Fig. 3E). The presence of a large number of blood vessels of different diameters with smooth borders and hilus cells could be seen within the loose connective tissue in the ovarian medulla in the different groups. Other structures in the medulla (connective tissue, blood and lymph vessels) were well preserved (Fig. 3F).
Examination of the sections from the sham group revealed that the single-layer cubic cover epithelium covering the outer surface of the ovary, the tunica albuginea below the epithelium, and the follicles belonging to different stages of development in the ovarian cortex were all normal in appearance (Fig. 4A). Oocytes located on one side of the antrum and corona radiata cells lined up in a single row were observed in a Graafian follicle in sections from this group. Although no nucleus and nucleolus were observed in the structure of the oocyte, scattered chromatin filaments was a noteworthy finding (Fig. 4B). Granulosa lutein and theca lutein cells were distinguishable from one another in the corpus luteum, the cells of both layers exhibiting a normal appearance. The surrounding blood vessels were easily visible, and their lumen had a smooth border (Fig. 4C, D). Lymph vessels with smooth lumen internal boundaries, as well as clusters of cells resembling hilus cells were observed in the lose connective tissue in the medulla. The blood vessels were well protected, and their inner lumen boundaries were clear (Fig. 4E, F).
In the curcumin group ovaries, the cuboidal epithelial cells of the outside of the ovary were normal in appearance, with no microvilli-like extensions on their apical surfaces. Hollow structures were observed in the tunica albuginea, immediately below the epithelium (Fig. 5A). Distinct spaces can be seen between luteal cells belonging to the corpus luteum. All elements of the luteal cells exhibited a normal structure (Fig. 5B). Evaluation of the relationship between theca follicle and granulosa cells showed that both theca follicle cells and granulosa cells exhibited a normal morphology. Some cells among the granulosa cells were apparently in the apoptotic process. There were cells around and inside the theca follicle, thought to be endocrine in character (Fig. 5C, D). Normally structured blood vessels and lymph capillaries were found in the loose connective tissue in the medulla. The presence of mast cells around the blood vessels was noted (Fig. 5E). In the atretic follicle, the glassy membrane immediately after the granulosa cells was indistinct, and the theca follicle layer had lost its normal structure (Fig. 5F).
Single-layer epithelial cells of the ovaries from the DM group were cubic, flat, or prismatic, and had no uniform shape. The tunica albuginea beneath the epithelium consists of tightly arranged collagen fibers (Fig. 6A). The zona pellucida and the surrounding corona radiata cell layer were distinguishable in the Graafian follicle. The distance between the corona radiata cells was also greater. An increased density of organelles and shrunken oocyte nucleus were observed in the oocyte cytoplasm (Fig. 6B). The integrity of the luteal cells of the corpus luteum that developed after ovulation was not preserved, and the distance between the cells had increased. This increased extracellular distance may provide increased capillary lumen diameters in the structure (Fig. 6C). In addition, the capillaries between the luteal cells, whose cell integrity was largely lost due to diabetes, were densely populated with erythrocyte clusters due to increased blood viscosity (Fig. 6D). Similarly, the blood vessel lumens in the structure of the ovary medulla, which has loose connective tissue, were filled with an amorphous, reticular substance due to the increase in blood viscosity. Numerous irregular protrusions were seen on the surface of the endothelium facing the lumen. Cells known as hilus cells were observed in the medulla (Fig. 6E, F).
The ovary of DC1 group is covered by single-layered cuboidal cells in histological sections. Spaces were observed between epithelial cells, and their cytoplasms were stained darker. The tunica albuginea layer beneath the epithelium was not easily visible (Fig. 7A). The primary oocyte located in the center of the preantral follicle in the cortex of the ovary was surrounded by a prominent zona pellucida. Although the nucleus membrane boundaries could not be easily distinguished, the nucleolus was very prominent. While granulosa cells surrounded by theca follicles mostly appeared normal, the labyrinth system between lutein cells was enlarged (Fig. 7B). The boundaries of the granulosa cells surrounding the antrum of the antral follicle in the cross section could not be easily distinguished, there were spaces between them, and they were surrounded by a theca follicle. The granulosa cells issued numerous irregular, microvillus-like cytoplasmic extensions towards the follicular fluid (Fig. 7C). A basement membrane was observed between the theca follicle and granulosa cells, with a distinctive structure around the preantral follicles (Fig. 7D). The presence of numerous blood vessels of different diameters in the loose connective tissue in the ovarian medulla and hilus cells that synthesize androgens, regarded as analogues of Leydig cells in the testis, were seen (Fig. 7E). The oocyte in the lumen of the atretic follicle surrounded by loose connective tissue had disappeared, the space left being filled with tissue fluid (Fig. 7F).
In the DC2 group, there were spaces between the single-layered cubic covering epithelium cells surrounding the ovary, and their cytoplasm were stained darker. The tunica albuginea layer under the epithelium was not clearly visible (Fig. 8A). A basement membrane was observed between the theca follicle and granulosa cells, with a distinctive structure around the preantral follicles (Fig. 8B). The internal and external layers of the theca follicle were well preserved. The corpus luteum cells, blood vessels, and the spaces between the cells all exhibited a normal appearance. The cytoplasm contents were well preserved, while most of the capillaries exhibited a normal structure, with an erythrocyte cluster appearing in the lumen of some (Fig. 8C, D). Loose connective tissue and rich blood vessels were seen in the ovarian medulla. Numerous lymph vessels, some with valves, were observed in addition to the blood vessels in the medulla (Fig. 8E). Examination of the structure of the atretic follicles revealed loss of the genus cell in the center. We think that degeneration of the follicle begins with the death of the oocyte. Collagen fibers were concentrated in the theca externa around the follicle, leading to atresia and the presence of numerous apoptotic cells in the granulosa cell layer (Fig. 8F).
Examination of the sections from the DC3 group ovaries revealed that the single-layered cubic covering epithelial cells surrounding the ovary surface were normal in appearance, with no microvillus-like protrusions on their apical faces. The presence of numerous hilus cells in the connective tissue was particularly noteworthy (Fig. 9A). In the antral follicles, the theca follicle surrounding the granulosa cells was clearly distinguished. Densely stained cells undergoing apoptosis were seen among the follicle cells (Fig. 9B). The cells in the structure of the corpus luteum, blood vessels and the intercellular spaces were well preserved. The cytoplasms and nuclei of the lutein cells exhibited a normal appearance. Erythrocyte clustering was observed in the capillary lumen (Fig. 9C, D). The ovary medulla was filled with normal-appearing loose connective tissue and rich blood vessels. The presence of numerous irregular cytoplasmic extensions on the lumen facing surfaces of the endothelial cells lining the inner surface of the blood vessels was also noted (Fig. 9E, F).
When the fine structure of the ovaries of the control group is examined; it was observed that the interna and externa layers of the theca follicle were easily distinguished, and the follicle cells and the intercellular space had normal appearance (Fig. 10A). The presence of many cavities of different sizes between the cells in the granulosa layer of the DM group and the cells in the theca follicle layer attracted attention. Similar spaces were observed, especially in the cytoplasm of granulosa cells. Capillaries were observed between the theca follicle cells, the lumen of which was filled with a dense substance. Condensation of the cytoplasm and nuclei of some granulosa cells was noted, which is thought to have largely lost their function (Fig. 10B). It was observed that the structures forming the theca layer in the ovaries of the Sham group had a normal appearance and the structure of the cells was preserved (Fig. 10C). In DC1 group, it was noticed that theca lutein cells were well preserved and the intercellular space was normal. Numerous lipid droplets were observed, with well-preserved well-defined nucleated granulosa cells (Fig. 10D). When the cytoplasm of granulosa cells in the curcumin group was examined, it was observed that lipid droplets, cell nuclei and mitochondria were well preserved, and the intercellular space was normal (Fig. 10E). In DC2 group, luteal cell cytoplasm, nuclei and intercellular spaces were observed to be normal. The lipid granules in the cell cytoplasm had smooth borders and the membranes surrounding the organelles were well preserved (Fig. 10F). In the DC3 group, the externa and interna layers of the theca follicle were easily distinguished. Theca externa cells and intercellular space were normal. It was observed that the nucleus and cytoplasm of the lutein cells in the theca interna were well preserved, and the cells had many lipid droplets (Fig. 10G).
It was noticed that the cytoplasm of the control group granulosa cells was filled with many lipid droplets. Well-developed capillaries were observed immediately after the basement membrane surrounding the granulosa cells. In addition, multivesicular structures were observed in some cells (Fig. 11A, B). On the contrary, the presence of numerous spaces between the granulosa cells and in the cell cytoplasm in the DM group was remarkable (Fig. 11C, D). It was observed that erythrocytes accumulated in the lumen of the capillaries between the granulosa cells. This may have occurred as a result of condensation of blood plasma. It was noticed that the cytoplasm and nucleus of the luteal cells in the sham group corpus luteum were in normal structure and they contained many lipid droplets (Fig. 11E, F). It was observed that the cytoplasm of DC1 group granulosa cells, especially lipid droplets, nuclei and nucleoli, had smooth borders. It was noticed that the intercellular space was normal (Fig. 11G, H). When the cytoplasm of curcumin group granulosa cells was examined, it was observed that lipid droplets, cell nuclei and mitochondria were well preserved, and the intercellular space was normal (Fig. 11I, J). DC2 group lutein cells were well preserved and the spacing between cells was normal (Fig. 11K, L). Although cavities were observed in the cytoplasm of some cells in the corpus luteum in the sections belonging to the DC3 group, it was noticed that the cell nucleus and nucleolus had a normal structure (Fig. 11M).
It was also observed that phagosomes with different sizes and densities in the cytoplasm of macrophages with different functional status, which are common in the medulla, in thin sections of ovaries belonging to all groups. It was noticed that the sizes of these macrophages, which are frequently encountered in the ovarian structure, also differ from each other (Fig. 12A-P).
A large number of normal blood vessels were noted in the ovarian sections of the control group. It was observed that the border between the endothelial cells lining the inner surface of the arterioles observed in the medulla was clear. The junctional complexes between the endothelium in the arteriole wall and the prominent inner elastic membrane just below it, the smooth muscle cells surrounding it, and the endothelial cells were easily recognized (Fig. 13A, B). In the sections in the DM group, on the other hand, it was observed that the erythrocytes in the lumen of the capillaries in most areas adhered to the vessel wall (Fig. 13C, D). Similarly, in the Sham group, erythrocytes in the capillary lumen were observed to pile up like a string of coins (Fig. 13E, F). It was noticed that extensive erythrocyte clusters were formed in the capillaries in DC1 and DC2 groups depending on the blood viscosity (Fig. 13G, H- K, L). Macrophages with a large number of phagosomes were observed in the ovarian medulla in the curcumin group. The cells in the connective tissue had smooth borders and normal nuclei (Fig. 13I, J). Heterochromatic nuclei and intensely stained cytoplasm, characteristic of granulosa cells, were striking. In addition, the capillary wall was smooth and vascular integrity was preserved (Fig. 13M, N).
This study investigated the potential protective/healing effects of curcumin, an anti-diabetic agent, on the ovary of in the early and late stages of diabetes (after seven days and 21 days, and during the development of metabolic disorder) using stereological, biochemical, and histopathological methods.
Weight loss has frequently been described among the symptoms caused by hyperglycemia (Ballester et al. 2007; Prasad et al. 2009; Garris and Garris 2004; Kim et al. 2006). Similar findings were obtained in the present study. The relative weight loss (%) in the DM group was higher than in the control, DC2, and DC3 groups. Type 1 DM generally results in a state of hypoleptinemia with mid- to long-term decreases in body weight. The adipose tissue hormone leptin participates in the integrated control of energy balance and reproductive activities. Severe hypogonadism has been observed in animals with leptin deficiency (Tena-Sempere 2007). Weight loss has also been reported to lead to loss of menstrual function, resulting in a lower LH and FSH response for GnRH. These side-effects may be reverses with weight gain (Frisch and McArthur 1974; Warren et al. 1975). In this study, the excessive weight loss in DM group rats (excessive relative weight loss) and the pronounced lipid deficiency in ovarian stroma cells along with the progression of the disease suggest that abnormalities such as angiogenesis, secretion of steroid hormones, and impaired metabolism may be caused by adipose tissue deficiency. Curcumin has been reported to increase serum adiponectin and leptin levels in patients with metabolic syndrome (Panahi et al. 2016). One study stated that 15 mg/kg curcumin administration prevented body weight loss (El-Demerdash et al. 2009). Our study findings showed that treatment with curcumin at 30 mg/kg for 14 days 21 days after induction of diabetes (DC2) and concurrently with diabetes (DC3) had a positive effect on body weight maintenance. However, this effect was not observed in the group (DC1) given curcumin seven days after induction of diabetes. This may be due to properties of curcumin such as suppressing preadipocyte differentiation in the early stages of diabetes and reducing both adipocyte numbers and fat content in adipose tissue (Ahn et al. 2010; Zhao et al. 2011).
The gonadosomatic index (GSI) in the DM group in this study decreased significantly compared to the control, sham and curcumin groups. Studies involving diabetic rats have reported that abnormalities in the HPG axis (Valdes et al. 1991), and GnRH, which regulates sexual maturation and reproductive functions in mammals, cause faster shrinkage of the Graafian follicle, resulting in a decrease in ovarian mass and thus a decrease in GSI. In addition, GnRH exerts inhibitory effects on smaller follicles, and its sustained release may directly affect folliculogenesis, as well as causing a reduction in ovarian mass and GSI (Oluwatoyin et al. 2013). The decrease in folliculogenesis rate, follicle numbers and FSH / LH levels in the present study suggest that the decrease in GSI in diabetic rats may have occurred as a result of a disruption of the HPG axis.
The GSI is an important criterion in evaluating the individual’s reproductive potential. It increased significantly in the DC1, DC2, and DC3 groups compared to the DM group. However, no significant difference was observed between the control, sham, and curcumin groups and the DC1, DC2, and DC3 groups. No previous studies have shown the protective effect of curcumin on the GSI in diabetic rats. Our study is unique from that perspective. However, more comprehensive studies are now needed on this subject.
Examination of the fasting blood glucose levels of the experimental groups revealed a significant increase in the DM group compared to the control, sham and curcumin groups. STZ is a widely used substance for experimental induction of DM in rats. A previous study reported that insulin secretion decreased and blood glucose levels increased as a result of the rapid depletion of pancreatic Langerhans islet β cells in a DM group treated with STZ (Shima et al. 2011). Fasting blood glucose decreased significantly in the DC1, DC2, and DC3 groups in the present study compared to the DM group. However, curcumin administration did not reduce blood glucose level to those of the control, sham, and curcumin groups. Curcumin has been reported to reduce blood glucose levels in diabetic rats through regulation of the polyol pathway (Nabavi et al. 2015). The blood glucose-lowering effect of curcumin has also been observed in human clinical studies of diabetic and pre-diabetic patients (Na et al. 2013). However, curcumin had no effect on blood glucose levels in individuals with normal levels at baseline. Further studies are now needed to evaluate the effects of curcumin on insulin resistance in diseases such as hyperglycemia and diabetes.
Stereological examination of the effects of diabetes on the ovary revealed a significant decrease in the numbers of primordial follicles in the DM group compared with the control group. Primordial follicle numbers increased significantly in the DC2 group compared with the DM group. A significant decrease was observed in the DC1, DC2 and DC3 groups compared with the control group, but there was no significant difference between these groups and the sham group.
Preantral follicle numbers decreased significantly in the DM group compared with the control and sham groups, but there was no difference between them and the curcumin group. No significant difference was also determined between the DC1, DC2, and DC3 groups and the DM group. Numbers of preantral follicles decreased in all these groups compared with the control group.
Antral follicle numbers in the DM group decreased significantly compared with the control and curcumin groups, but exhibited no from those of the sham group. When the protective effect of curcumin was analyzed in terms of antral follicle numbers, the number of follicles in the DC3 group increased significantly compared with the DM group. However, no difference was found between the DM group and the DC1 and DC2 groups. The number of antral follicles in the diabetes groups following curcumin administration was not as high as that in the control group, showing that curcumin was unable to normalize antral follicle numbers in the diabetes groups. Antral follicle numbers in the DC1 group decreased significantly compared with the sham group, but no difference was found between the DC2 and DC3 groups.
Total numbers of follicles in the DM group decreased significantly compared to the control, sham and curcumin groups. The numbers of follicles in the DC2 and DC3 groups exhibited a significant increase compared to the DM group. No difference was observed between the DC1 group and the DM group. However, more follicles were observed in the control group than in the diabetes groups (DC1, DC2, and DC3), while no difference was found between the total number of follicles in the sham group and the number of follicles in the DC2 and DC3 groups.
Examination of follicular numbers can provide important information about the function of the ovary, especially the relationship between folliculogenesis and the factors that regulate it (Myers et al. 2004). A reduction in the number of follicles in diabetic individuals may lead to premature depletion of the follicle pool, leading to an increased risk of premature menopause and/or reduced fertility potential. Mouse, rat, and rabbit studies have investigated the effects of diabetes on the ovary (Farhad et al. 2013; Tiwari-Pandey and Ram Sairam 2009).
In a study of Chinese hamsters, Garris (1984) found that different levels of hyperglycemic state (medium-160-350 mg / dL blood glucose and high level 350 mg / dL blood glucose) caused severe adverse effects on ovarian functions. Total numbers of primary and secondary follicles in that study decreased in the diabetic group compared to the control group. As the hyperglycemic state progressed from moderate to high, the numbers of secondary follicles continued to decrease compared with the controls, and the percentage of available viable follicles dropped dramatically (Garris 1984). The results of our study, indicating a decrease in the numbers of preantral and antral (secondary) follicles in the DM group compared with the control group, are consistent with those of Garris. The number of antral follicles is considered one of the most important markers in evaluating functional ovarian reserves, in predicting the response to gonadotropin stimulation, and in predicting infertility and menopause risk (Depmann et al. 2016). Similarly, in Farhad et al.’s (2013) stereological mouse study, the numbers of preantral follicles and antral follicles decreased, while atretic follicle numbers increased, in the diabetic group compared with the control group. The authors also reported higher cortex and medulla volumes in the diabetic group compared with the control group. The authors suggested that this might be due to the formation of cortical vessels and polycystic structures. This finding differs from the present study, in which the cortex volume decreased significantly in the diabetic group compared with the control group. This decrease in cortex volume may be attributable to the decrease in blood vessel volume in the diabetic ovary, inflammation and increases in atretic follicles, and consequent slowing or pausing of folliculogenesis. Khedr (2016) reported that the numbers of primordial, primary, secondary, tertiary and Graafian follicles in diabetic rat ovaries decreased significantly compared with the control group in a diabetes model induced with STZ. Researchers have reported that uncontrolled diabetes causes adverse biological effects by increasing oxidative stress and lipid peroxidation (Awadin et al. 2015). In that study, factors such as increased lipid peroxidation marker MDA levels in the ovaries of diabetic rats, decreased SOD, CAT, and GPx activities, and decreased gonadotropins contributed to the degeneration of follicles (Khedr 2017).
No previous studies have investigated the effects of curcumin on the diabetic ovary, although it has been reported to protect mouse, rat, and rabbit ovary cells from proapoptotic, pronecrotic and antiproliferative effects in the face of various oxidative stress factors (Tiwari-Pandey and Ram Sairam 2009). Curcumin has frequently been shown to exhibit different effects under varying stress conditions (Tiwari-Pandey and Ram Sairam 2009; Sak et al. 2013; Aktas et al. 2012). Despite the known widespread antioxidant effects of curcumin, the data obtained from the group (DC1) in which this substance was administered seven days after diabetes induction in the present study confirmed that curcumin might exhibit different effects on organs and tissues under varying stress conditions. To summarize, curcumin was protective against diabetes in the DC2 and DC3 groups due to its antioxidant properties, while it exhibited no protective properties in the DC1 group.
Although the antioxidant properties of curcumin have been proved by numerous studies, it should be remembered that it cannot completely eliminate ROS, and that its anti-inflammatory and anti-apoptotic activities may be affected by numerous factors (Sandur et al. 2007). In addition, the apoptotic process can be triggered by the Fas/FasL pathway, as well as by various signals and pathways (Cifone et al. 1995; Li and Yuan 1999). The occurrence of follicular imbalances due to hyperglycemia in the early stages of DM suggests that the antioxidant activity of curcumin is insufficient to eliminate or reduce these adverse effects caused by diabetes. The excessively increased oxidative stress and hyperglycemia that occur in the cells in the chronic period with the induction of DM may have masked the protective role of curcumin in the follicles and the general structure of the ovary. This may be due to the production of antioxidant-induced stress by disturbing the balance between antioxidants and increased reactive species.
Primordial follicles are the most resistant follicles to external factors. However, preantral follicles can easily progress to apoptosis by being more affected by environmental conditions. The reason why curcumin does not exert a protective effect on preantral follicle numbers in diabetic groups treated with curcumin is that the developing follicles are more sensitive to external factors due to their structure, at this stage. Diabetes causes a severe deterioration in the follicle, and the dose of curcumin applied may not be sufficiently strong to restore the structure.
The mechanisms by which the primordial follicles are protected and the principal mechanisms responsible for the loss of preantral and antral follicles at different stages of diabetes have recently emerged as a new research topic. The results of the present study show a pro-oxidant effect on primordial, preantral and antral follicle numbers when curcumin was administered days after the development of diabetes. However, when administered simultaneously with the development of diabetes, it affected only the antral follicles. When given 21 days after the development of diabetes, it exhibited an antioxidant effect on and protected primordial follicle numbers. Histopathological examinations at the level of light and electron microscopy showed that curcumin exhibits protective effects on the structural and functional components of the diabetic ovary. In the curcumin-treated groups, the connective tissue and blood vessels were relatively well protected, the theca follicle around the follicles exhibited a clear organization, the structure of the glassy membrane (basement membrane) was preserved, and the internal and external layers of the theca follicle were easily visible. In the diabetic groups treated with curcumin collagen fiber density / organization in the theca externa layer and the capillary network in the interna layer were smooth and rich. The structure of the zona pellucida was also preserved in these groups, the corpus luteum cells, blood vessels and the spaces between the cells exhibited a normal appearance, and the luteal cell contents were also well preserved.
Ovarian follicular volume is the best indicator of follicular growth (Yavas and Selub 2009). In the present study, ovarian follicle volumes were estimated using the Cavalieri volume estimation method. The preantral follicle volume in the DM group exhibited an extreme decrease compared to the control and curcumin groups. No difference was observed in preantral follicle volumes between the sham and the DM groups. No difference was also determined between the DC1, DC2, and DC3 groups treated with curcumin in terms of preantral follicle volume and the DM group. Finally, no difference was observed between the DM group and the other groups.
In a study using acute and chronic insulin-dependent diabetes models, the authors reported that preovulatory oocytes were significantly smaller in size compared to the control group in both models, and that the developing follicles were smaller in size and contained more apoptotic foci (Chang et al. 2005). In addition, similarly to follicle sizes, the numbers of preovulatory oocytes and corpus luteum decreased in hyperglycemic Akita mice compared to control animals (Chang et al. 2005). Researchers have suggested that abnormally low insulin, which acts as a paracrine factor to facilitate the transition from the primordial follicle to the primary follicle, may inhibit or delay follicle transitions (Kezele et al. 2002). In another study, evaluating the protective effect of Aloe vera, a plant with hypoglycemic effects, on the rat ovarian structure, the researchers found a significant reduction in secondary and tertiary follicle diameters in diabetic rats in the group with experimental diabetes induced by STZ (Shima et al. 2011). The fact that the preantral follicle volumes obtained from our stereological analysis were lower in the diabetic group compared to the control group is consistent with those studies.
Another study examining the effects of maternal diabetes reported decreased numbers and diameters of primary and preantral follicles decreased in the ovaries of offspring born to diabetic mothers (Khaksar et al. 2013). Similarly, Tatewaki et al. (1989) observed that the percentage of primary follicles, secondary follicles, and tertiary follicles decreased in diabetic mice compared with the control group, while the percentage of atretic follicles increased. According to our findings, the decrease in the preantral follicle volume in the DM group may be due to the decrease in androgen production in follicle theca cells and changes in estrogen biosynthesis. This is because the gradual increase in estrogen production correlates with the increase in follicular surface area. In other words, three-dimensional follicular growth occurs as a result of the proliferation of theca interna and externa cells on the surface of the follicle (Gore-Langton and Armstrong 1988). In addition, FSH / LH levels impaired due to diabetes may reduce the amount of estrogen released from follicular cells (Ballester et al. 2007). The absence of any difference in antral follicle volume between the groups in the present study may be due the heterogeneity of the data in the groups, due to the fact that the follicle has a large volume and / or that the coefficients of variation of the groups are high.
Dietary curcumin in rabbits has been found to increase the diameters of primary, secondary and tertiary follicles (Sirotkin et al. 2017). The absence of any difference between the diabetic group (DM) and other groups that are named DC1, DC2 and DC3, treated with curcumin in terms of preantral follicle volumes in the present study, may be due to those authors using diameter measurement (the two-dimensional approach) when evaluating follicles, whereas we used volume calculation (the three-dimensional approach) to evaluate follicle size.
Vascularization and blood flow in the ovary are indispensable for adequate growth and development of ovarian follicles, ovulation and corpus luteum formation (Behrman et al. 2001). In the present study, the blood vessel volume in the DM group was significantly lower compared with that in the control group. In our study, we think that the decrease in blood vessel volume in the ovarian tissue of the diabetes group is related to the decrease in the volume and number of developing follicles. It is thought that the decrease in ovarian stromal vascularity may lead to less delivery of gonadotropins to the granulosa cells of the developing follicles. As it is known, during follicular atresia, the interruption of oxygen support through the blood vessels in the structure can initiate atresia. Images obtained from light microscopic and transmission electron microscopic examinations also support the above findings. The presence of capillaries with partially enlarged lumens in the theca follicle, which contains blood vessels of different diameters in the structure of the ovaries of the DM group, the fact that the capillaries are filled with amorphous material, and the intercellular space between the theca cells (disruption of the normal structure) shows that diabetes adversely affects the blood vessels in the ovaries. It has been observed that there are similar structural defects in the lumens of the blood vessels in the ovarian medulla, which has loose connective tissue characteristics. The remarkable changes as follows: the presence of amorphous substance in the reticular structure in the lumen of blood vessels due to the increase in blood viscosity, the observation of many irregular protrusions on the endothelial surface facing the lumen. All these findings can be counted as one of the causes of atresia and apoptosis-related factors in follicles.
Wu et al. (2017) reported that mice in a diabetes group given a 20-week high-fat diet had low-density blood vessels in their ovaries, and that diabetes suppressed ovarian angiogenesis. Diabetes has been found to inhibit the expression of VEGF, a vital regulatory factor for angiogenesis. We attribute the decrease in blood vessel volume in diabetic group ovarian tissue in the present study to the decrease in the volume and number of developing follicles. A reduction in ovarian stromal vascularity may lead to less transmission of gonadotropins to the granulosa cells of the developing follicles.
Both clinical and experimental diabetes are known to impair endothelial-dependent vascular activity by inactivating NO (Powers et al. 1996). NO has a luteolytic effect, mediating the increase of prostaglandins in the corpus luteum (Fridén et al. 2000). Defects in the vessels that provide blood support in the corpus luteum prevent sufficient blood from reaching the luteal cells, which causes changes in progesterone secretion, resulting in insufficient luteal function in the corpus luteum (Greenwald 1989; Reynolds 1986). Corpus luteum volumes were calculated in the present study, and a significant decrease was observed in the DM group compared with the control, sham, and curcumin groups. A considerable increase was observed in the volume of the corpus luteum in the DC2 and DC3 groups compared to the DM group, while no difference was found between the DC1 group and the DM group. The corpus luteum volumes in the DC1 and DC3 groups exhibited a significant decrease compared with the control group corpus luteum. It was observed that there was no difference between the DC2 group and the Control group in terms of corpus luteum volumes.
A decrease in the volume of the corpus luteum is known to lead to a reduction in progesterone concentrations (Elbaum et al. 1975). A study in which experimental diabetes was induced with STZ (50 mg/kg) reported a decrease in the number and diameter of the corpus luteum in the diabetic group compared to the controls (Farhad et al. 2013). Apoptosis may occur in luteal cells due to hyperglycemia in the corpus luteum. In addition, the decrease in the volume of the corpus luteum may be due to reduced lipid reserves in luteal cells. It has also been suggested that curcumin can increase serum adiponectin and leptin levels in patients with metabolic syndrome (Panahi et al. 2016).
Improving the decreased lipid reserves of luteal cells by regulating plasma leptin levels with curcumin treatment may be the main mechanism involved in the preservation of corpus luteum volume. Further studies are now needed to investigate the effect of leptin levels on corpus luteum functions in diabetic individuals treated with curcumin.
The results of the present study suggest that curcumin may have protective effects on the ovary. In the present study, we found that curcumin positively promoted angiogenesis after long-term diabetes (21 days) when used concurrently with diabetes induction, although we did not expect a positive contribution to angiogenesis with curcumin administration seven days after diabetes induction. We think that this bi-directional effect of curcumin on angiogenesis is regulated by the interaction of many factors, including VEGF, MMPs and TGF-1, fibroblast growth factor (Wang & Chen, 2019). The mechanism(s) underlying the opposite effect of curcumin on angiogenesis under different diabetic conditions now need to be investigated.
Examination of the vascular structure of the DC2 and DC3 groups under light microscopy revealed that the endothelium was normal in structure and very well bonded to the basement membrane due to the protective properties of curcumin. The presence of numerous pinocytic vesicles on the surface of the endothelial cells facing the basement membrane in these groups indicates that these cells are also functionally active.
The volume of ovarian connective tissue in the DM group increased significantly compared with the control and sham groups. In addition, no significant difference was found between the DM group and the curcumin group. The increase in the DM group ovarian connective tissue volume can be attributed to the fibrotic effect and inflammation regulated by TGF-1 and NF-kB. This may be one of the reasons for an adverse impact on ovarian functions. Because these factors increase in diabetic tissues in response to AGEs or high glucose levels (Khedr 2017; Erbas et al. 2014; Park et al. 1997; Sharma et al. 1997).
The increase in the volume of connective tissue in the ovary due to excessive proliferation of fibroblasts and the accumulation of extracellular matrix is one of the main causes of ovarian dysfunction (Zhou et al. 2017). In the present study, connective tissue volume decreased significantly after curcumin administration in the DC1, DC2, and DC3 groups compared with the DM group. These results suggest that curcumin normalizes female oocyte/follicle maturation in diabetic individuals by preventing fibrosis in the ovary. In addition, the absence of a significant difference in connective tissue volumes between these three groups (DC1, DC2, and DC3) and the control, sham and curcumin groups suggest that curcumin administration normalizes the amount of connective tissue in the diabetic groups comparison with the control, sham and curcumin groups. In addition, no difference was observed between the connective tissue volumes of the control and sham groups and those of the curcumin group.
While there was no difference in ovarian cortex volumes in the DC1, DC2, and DC3 groups compared to the DM group, a significant decrease was observed in the DC1 and DC2 groups compared to the control group. The cortex volume of the DC3 group exhibited no difference compared to the control group. No statistically significant difference was also observed in medulla volumes in the DC1, DC2, and DC3 groups compared with the control and DM groups. The decrease observed in the cortex in the DC1 and DC2 groups compared to the control group was directly proportional to the decrease in follicle numbers and connective tissue volume.
Functions such as follicular growth, oocyte maturation, and estrus behavior in the ovary are regulated by the action of gonadotropins (FSH AND LH). The HPG axis has been shown to be adversely affected by diabetic conditions (Castellano et al. 2006; Castellano et al. 2009). Studies have emphasized a strong and direct relationship between insulin and FSH / LH hormones, and this was confirmed by the findings of the present study (Katayama et al. 1984; Ballester et al. 2007). Hypothalamic-pituitary dysfunction in diabetic animals is characterized by low basal FSH and LH levels (Kirchick et al. 1982; Howland and Zebrowski 1980). While this was confirmed in the present study in terms of FSH levels, no difference was found in LH levels. This may be because LH has a shorter half-life than FSH (Liu et al. 1972). Curcumin is known to exhibit phytoestrogenic effects (Bachmeier et al. 2010). It is therefore capable of regulating the endocrine function of reproduction by interacting with the endocrine system and affecting the HPG axis, and by supporting follicular development or the protection of follicles (Sirotkin and Harrath 2014; Yan et al. 2018). Curcumin has been reported to protect against proapoptotic, pronecrotic, antiproliferative and anti-oogenic effects induced by suppression of FSH receptors in mouse ovary cells (Tiwari-Pandey and Ram Sairam 2009).
Examination of the effects of curcumin treatment applied at different times of diabetes on the HPG axis revealed that serum FSH levels increased significantly in the DC1 and DC3 groups compared with the DM group, but there was no difference between these two groups when compared with the control group. No difference between the FSH levels of the DC2 group and the DM group, but a significant decrease was observed compared with the control group. Based on our findings, it may be concluded that curcumin brings FSH levels closer to normal when administered in parallel with the development of diabetes (DC3) or seven days later (DC1). Serum FSH levels could not be normalized when curcumin was used after long-term diabetes (application 21 days after the onset of diabetes). It has been suggested that impairments in the reproductive function of diabetic female rats may be partially related to the secretion of gonadotropins (Johnson and Sidman 1979) or to a decrease in the affinity or number of gonadotropin receptors in ovarian target cells (Tesone et al. 1983). The antioxidant effect of curcumin in the DC2 group may have been insufficient to prevent disorders in the affinity and / or numbers of gonadotropins; the secretion is reduced under long-term diabetic conditions. This may be due to the low dose of curcumin dose we used, or to the duration being insufficient. We think that the antioxidant effects of curcumin on serum FSH levels should now be investigated in studies focusing on dosage and duration.
Examination of serum LH levels revealed no difference between the DC1 and DC3 groups and the control, sham, curcumin, DM, or DC2 groups, while serum LH levels in the DC2 group were significantly lower than in the control, sham, and curcumin groups. We concluded that curcumin was insufficient to normalize serum LH levels under prolonged diabetic conditions (the DC2 group).
Recent studies have shown that oxidative stress may cause infertility by affecting numbers of ovarian follicles and oocytes (Tarin 1995; Tarin 1996; Miyamoto et al. 2010). In the present study, CAT activity and SOD were measured in order to determine the underlying mechanism involved in decreased primordial / preantral follicle numbers, blood vessel volumes, and corpus luteum volumes in the diabetic groups. A significant increase in CAT activity was observed in the diabetic groups compared with the control group. In another study, and consistent with our own findings, CAT enzyme activity, which catalyzes the reduction of hydroperoxides and protects mammalian cells against oxidative damage by playing a role in neutralizing ROS, was significantly higher in diabetic rats compared to controls (Qujeq and Rezvani 2007). Researchers have made a comment that increased in CAT activity with membrane protein and lipid damage is caused by an increase in free radicals. It may therefore be concluded that the antioxidant defense system increases CAT enzyme activity as a result of oxidative stress in diabetic groups.
All the findings obtained in the present study confirm that this sequence of events is interconnected. The increase in CAT enzyme activity in connection with decreased FSH values in the diabetic group in the present study shows that the increase in follicular ROS affects hormone levels, oocyte development, and follicle numbers. In that context, it may be concluded that a delicate balance exists between the synthesis and release of ovarian hormones or changes in the sensitivity of cells to hormones and antioxidant enzymes (Halliwell et al. 1992). The decrease in plasma SOD level in the diabetic groups in the present study may have occurred due to the increased oxidative stress in the ovary. A decrease in SOD levels may cause DNA damage in primordial / preantral and antral follicles, resulting in cell loss.
CAT activity and serum SOD levels were evaluated in order to investigate the effectiveness of curcumin, an antioxidant substance, against the decrease in follicle numbers caused by oxidative stress induced by diabetes and changes in ovarian structure. Accordingly, a significant decrease in CAT activity was observed in the DC2 and DC3 groups compared with the DM group, but no difference was found compared with the control group. Serum SOD levels increased significantly in the DM group compared to the DC1, DC2, and DC3 groups. These groups (DC1, DC2, and DC3) exhibited no difference compared to the control group in terms of SOD levels. The antioxidant effects of curcumin can occur through various mechanisms. One of the most important of these may be increasing the levels of intracellular antioxidants and their mRNA levels, which are adversely affected by diabetic conditions. Due to its antioxidant activity, curcumin may reduce oxidative stress-induced toxicity, which adversely affects ovarian function, by normalizing intracellular antioxidant enzymes. In one study, high oxidative stress marker levels and decreased endogenous antioxidants were observed in the ovaries of PCOS animals. The authors reported significant normalization of SOD, CAT, and GSH activities in the PCOS group with curcumin treatment (Reddy et al. 2016).
Conclusion:
The findings of the present study indicate that diabetes causes a decrease in the numbers of follicles in the ovary, follicle and corpus luteum volumes, blood vessel and ovarian cortex volumes, body weight, GSI, and serum SOD and FSH levels. In addition, diabetes causes an increase in CAT activity and connective tissue volume. Curcumin, which possesses powerful antioxidant properties, was used to eliminate this adverse manifestation caused by diabetes. The protective effect of this substance in different stages of diabetes was extensively investigated using stereological, biochemical and histopathological analyses. It may therefore be concluded that curcumin exerts its protective effects on diabetic ovaries using the following mechanisms;
• Regulation of the HPG axis,
• Preservation of the ovarian cycle and follicle numbers,
• Maintenance of the normal structure of the blood vessel, and
• Reduction of fibrosis in the ovary and lowering hyperglycemia
We suggest that curcumin exerts a protective effect by reducing fibrosis in the ovary and lowering hyperglycemia. In the light of the results of this study, it can be concluded that curcumin may be a desirable and effective substance for restoring ovarian follicle damage in individuals exposed to hyperglycemia. However, comprehensive molecular studies are now needed on this subject. In addition, it should not be forgotten that in addition to its antioxidant properties, curcumin might also have a pro-oxidant effect under some conditions.
Conflicts of Interest
The author declares that there is no conflict of interest.
Acknowledgements
This study was supported by a grant from the Scientific Research Foundation of Ondokuz
Mayis University (Suleyman Kaplan; PYO.TIP.1904.18.022).
Author contribution
K. K. Tufekci conceived and designed the study, performed the experiment, and analyzed the data under the supervision of S. Kaplan. S. Kaplan also managed the funding acquisition and administered all stages of the project. All authors contributed towards the interpretation of findings from this study, played a role in drafting the article or revising it critically for its intellectual content, and gave their final approval for this version of the paper to be published.
|
Parameter |
Value |
|---|---|
|
Area of the counting frame (a) |
6400 µm2 |
|
Step size (X,Y) |
40,000 µm2 |
|
Area sampling fraction (asf) |
6400 µm2/40,000 µm2 |
|
Disector height (h) |
10 µm |
|
Guard zone distance (upper) |
3 µm |
|
Table 2 |
Mean ± SEM of Relative weight loss (%), GSI (%) and Blood glucose level (mg / dL) |
||||||
|
Control |
Sham |
Curcumin |
DM |
DC1 |
DC2 |
DC3 |
|
|
Relative weight loss (%) |
2.00 ± 2.10 d |
7.49 ± 2.56 |
4.81 ± 6.03 |
25.40 ± 8.41 a,f,g |
15.00 ± 2.9 |
4.33 ± 5.38 d |
1.55 ± 3.10 d |
|
GSI (%) |
0.08 ± 0.005 k |
0.12 ± 0.007 k |
0.10 ± 0.006 k |
0.06 ± 0.003 a,l,m,n |
0.10 ± 0.001k |
0.10 ± 0.005 k |
0.14 ± 0.005 k |
|
Blood glucose level (mg / dL) |
94 ± 4.61 k,l,m,n |
152 ± 6.88 k,l,m,n |
130 ± 7.46 k,l,m,n |
488 ± 20.09 h,i,j,k,l,m,n |
308 ± 23.59 h,i,j,k |
340 ± 18.88 h,i,j,k |
317 ± 32.20 h,i,j,k |
a Diff. vs. Control (p < 0.05); b Different vs. Sham (p < 0.05); c Different vs. Curcumin (p < 0.05); d Different vs. DM (p < 0.05);e Different vs. DC1 (p < 0.05); f Different vs. DC2 (p < 0.05); gDifferent vs. DC3 (p < 0.05); h Different vs. Control (p < 0.01); i Different vs. Sham (p < 0.01); j Different vs. Curcumin (p <0.01); k Different vs. DM (p < 0.01); l Different vs. DC1 (p < 0.01); m Different vs. DC2 (p < 0.01); n Different vs. DC3 (p < 0.01).
|
Table 3 |
Mean ± SEM of Follicle numbers |
||||||
|
Control |
Sham |
Curcumin |
DM |
DC1 |
DC2 |
DC3 |
|
|
Primordial Follicle Number |
1648± 100.903 i,j,k,l,m,n |
668.17±71.342h |
539±119.532h |
364.83±36.257h |
406.83±70.663h |
891.33± 75.575h,e, k, g |
533.20±59.008h |
|
CE |
0.13 |
0.27 |
0.27 |
0.29 |
0.28 |
0.18 |
0.26 |
|
CV |
0.13 |
0.23 |
0.77 |
0.22 |
0.38 |
0.18 |
0.22 |
|
Preantral Follicle Number |
1285.20±90.683 j,k,l,m,n |
1156.14±155.43 |
765.17± 51.088h |
412.67±23.571h,b |
467.40±82.844 b |
513.67± 67.252b, h |
784.60±82.456h |
|
CE |
0.15 |
0.2 |
0.22 |
0.26 |
0.29 |
0.24 |
0.21 |
|
CV |
0.21 |
0.32 |
0.24 |
0.28 |
0.43 |
0.29 |
0.21 |
|
Antral Follicle Number |
362.50 ± 11.63 b,k,l,m,n |
214.83 ± 24.33 a, l |
254.00 ± 22.97 k, l |
80.60 ± 7.04 h |
60.80 ± 16.63 h, k |
117.20 ± 17.23 h, k |
283.00 ± 74.03 h, k |
|
CE |
0.25 |
0.46 |
0.4 |
0.64 |
0.72 |
0.6 |
0.46 |
|
CV |
0.07 |
0.28 |
0.41 |
0.36 |
0.61 |
0.45 |
0.73 |
|
Toplam Follicle Number |
3295.33 ± 159.617 i,j,k,l,m,n |
2106.33 ± 196.538 h |
1558.67 ± 124.488 h, d, e |
858.33 ± 53.672 h, i, m, n |
935 ± 132.210 h, i, n |
1622.40 ± 128.439 h, l |
1600.80 ± 86.224 h, k, l |
|
CE |
0.09 |
0.14 |
0.15 |
0.18 |
0.19 |
0.14 |
0.15 |
|
CV |
0.11 |
0.2 |
0.18 |
0.09 |
0.37 |
0.19 |
0.16 |
a Diff. vs. Control (p < 0.05); b Different vs. Sham (p < 0.05); c Different vs. Curcumin (p < 0.05); d Different vs. DM (p < 0.05);e Different vs. DC1 (p < 0.05); f Different vs. DC2 (p < 0.05); gDifferent vs. DC3 (p < 0.05); h Different vs. Control (p < 0.01); i Different vs. Sham (p < 0.01); j Different vs. Curcumin (p <0.01); k Different vs. DM (p < 0.01); l Different vs. DC1 (p < 0.01); m Different vs. DC2 (p < 0.01); n Different vs. DC3 (p < 0.01).
|
Table 4: |
Mean ± SEM of volumes |
||||||
|
Groups |
Control |
Sham |
Curcumin |
DM |
DC1 |
DC2 |
DC3 |
|
Preantral follicle volume (mm3) |
0.99 ± 0.11c,f |
0.55± 0.10 |
0.39 ± 0.03 a, f, k |
0.12 ± 0.23 h |
0.22 ± 0.04 h |
0.22 ± 0.01 a |
0.46 ± 0.23 |
|
CE |
0.02 |
0.03 |
0.04 |
0.06 |
0.05 |
0.05 |
0.04 |
|
CV |
0.26 |
0.23 |
0.18 |
0.3 |
0.37 |
0.16 |
0.13 |
|
Antral Follicle Volume (mm3) |
0.79 ± 0.09 |
0.47± 0.07 |
0.39 ± 0.04 |
0.68 ± 0.04 |
0.59 ± 0.04 |
0.55 ± 0.05 |
0.72 ± 0.19 |
|
CE |
0.03 |
0.03 |
0.03 |
0.04 |
0.03 |
0.03 |
0.03 |
|
CV |
0.27 |
0.26 |
0.36 |
0.31 |
0.14 |
0.21 |
0.48 |
|
Corpus Luteum Volume (mm3) |
8.10 ± 0.20 i,k,l,n |
6.49± 0.17 h |
7.61 ± 0.34 b, k, l |
5.07 ± 0.15 m, n |
5.41 ± 0.31 h, m, n |
7.45 ± 0.49 k, l |
6.49 ± 0.21 h |
|
CE |
0.01 |
0.02 |
0.01 |
0.01 |
0.01 |
0.02 |
0.01 |
|
CV |
0.32 |
0.34 |
0.11 |
0.23 |
0.28 |
0.23 |
0.19 |
|
Connective Tissue Volume (mm3) |
2.43 ± 0.34 k |
1.74 ± 0.26 k |
0.39 ± 0.03 m |
3.38 ± 0.22 h, i, l, m, n |
2.26 ± 0.13 k |
1.85 ± 0.03 k |
1.73 ± 0.17 k |
|
CE |
0.02 |
0.03 |
0.02 |
0.01 |
0.02 |
0.02 |
0.02 |
|
CV |
0.3 |
0.4 |
0.2 |
0.15 |
0.09 |
0.05 |
0.16 |
|
Blood Vessels Volume (mm3) |
1.42 ± 0.08 b, k |
0.96 ± 0.04 a |
1.20 ± 0.10 |
0.82 ± 0.04 h, m, g |
1.20 ± 0.26 |
1.28 ± 0.06 k |
1.14 ± 0.04 d |
|
CE |
0.02 |
0.09 |
0.02 |
0.02 |
0.03 |
0.02 |
0.03 |
|
CV |
0.12 |
0.13 |
0.2 |
0.13 |
0.34 |
0.1 |
0.07 |
|
Ovarian Cortex Volume (mm3) |
11.57 ± 0.35 c, d, l, m |
10.55 ± 1.07 |
7.82 ± 0.69 a |
7.87 ± 0.74 a |
8.02 ± 0.42 h |
7.95 ± 0.25 h |
8.77 ± 0.66 |
|
CE |
0.01 |
0.02 |
0.01 |
0.01 |
0.01 |
0.01 |
0.02 |
|
CV |
0.07 |
0.23 |
0.2 |
0.21 |
0.1 |
0.15 |
0.13 |
|
Ovarian Medulla Volume (mm3) |
3.08 ± 0.16 |
3.16 ± 0.39 |
1.97 ± 0.31 |
3.44 ± 0.52 |
2.06 ± 0.44 |
2.61 ± 0.31 |
3.64 ± 0.90 |
|
CE |
0.01 |
0.01 |
0.02 |
0.02 |
0.02 |
0.02 |
0.02 |
|
CV |
0.12 |
0.28 |
0.36 |
0.35 |
0.34 |
0.32 |
0.39 |
a Diff. vs. Control (p < 0.05); b Different vs. Sham (p < 0.05); c Different vs. Curcumin (p < 0.05); d Different vs. DM (p < 0.05);e Different vs. DC1 (p < 0.05); f Different vs. DC2 (p < 0.05); gDifferent vs. DC3 (p < 0.05); h Different vs. Control (p < 0.01); i Different vs. Sham (p < 0.01); j Different vs. Curcumin (p <0.01); k Different vs. DM (p < 0.01); l Different vs. DC1 (p < 0.01); m Different vs. DC2 (p < 0.01); n Different vs. DC3 (p < 0.01).
|
Table 5: |
Mean ± SEM of CAT activity (ng/mL), SOD level (ng/mL), FSH level (IU/L), LH level (mIU/mL) |
||||||
|
Groups |
Control |
Sham |
Curcumin |
DM |
DC1 |
DC2 |
DC3 |
|
CAT activity (ng/mL) |
130.30 ± 5.24 k, l |
107.67 ± 16.50 k |
97.35 ± 18.74 k |
253.41 ± 7.11 h, i, j, m, n |
180.83 ± 16.95 h, n |
143.69 ± 15.94 k |
104.66 ± 4.31 k, l |
|
SOD level (ng/mL) |
4.11± 0.30 k |
3.89 ± 0.25 k |
3.60 ± 0.16 k |
1.84 ± 0.24 h, I, j, l, m ,n |
3.31 ± 0.15 |
3.64 ± 0.14 |
3.57 ± 0.27 |
|
FSH level (IU/L) |
12.01± 0.33 |
10.61 ± 0.72 k, m |
10.38 ± 1.16 k, m |
6.07 ± 0.49 h, l, n |
10.79 ± 0.51 k, m |
6.39 ± 0.48 h, i, j, n |
9.87 ± 0.68 k, m |
|
LH level (mIU/mL) |
33.06 ± 1.28 m |
35.68 ± 0.832 m |
34.40 ± 2.73 m |
27.13 ± 3.27 |
28.08 ± 3.50 |
21.29 ± 1.70 h, i, j |
30.59 ± 3.65 |
a Diff. vs. Control (p < 0.05); b Different vs. Sham (p < 0.05); c Different vs. Curcumin (p < 0.05); d Different vs. DM (p < 0.05);e Different vs. DC1 (p < 0.05); f Different vs. DC2 (p < 0.05); gDifferent vs. DC3 (p < 0.05); h Different vs. Control (p < 0.01); i Different vs. Sham (p < 0.01); j Different vs. Curcumin (p <0.01); k Different vs. DM (p < 0.01); l Different vs. DC1 (p < 0.01); m Different vs. DC2 (p < 0.01); n Different vs. DC3 (p < 0.01).