Methimazole-Induced Hypothyroidism Increases the Content of Glycogen and Changes the Expression of LDH, GLUT4, and Aromatase in the Pregnant Uterus of Rabbits.

Forty-eight European Chinchilla-breed female rabbits ( Oryctolagus cuniculus 9–12 months old) were housed in individual steel cages under artificial lighting conditions (16:8 h light–dark and kept at 20 ± 2 °C). Virgin (n = 12) and pregnant rabbits (n = 36) were fed using pellet food chow and had continuous access to drinking water. For each reproductive condition, six rabbits were used as euthyroid, and the other six were hypothyroid, induced through the intake of 10 mg/kg/day of methimazole (MMI) in drinking water for 30 days. This treatment with MMI is effective for inducing hypothyroidism in virgin rabbits [ 1 , 29 ]. Thus, we had eight groups: control virgins (C0; n = 6), control at day 5 of pregnancy (C5; n = 6), control at day 10 of pregnancy (C10; n = 6), and control at day 20 of pregnancy (C20; n = 6), MMI-treated virgins (H0; n = 6), MMI-treated at day 5 of pregnancy (H5; n = 6), MMI-treated at day 10 of pregnancy (H10; n = 6), and MMI-treated at day 20 of pregnancy (H20; n = 6). All MMI-treated rabbits had 30 days of treatment after adjusting days to before and after they were matched ( Figure 1 ). Females copulated four times with expert males to facilitate pregnancy. The day of copulation was considered as day 0 of gestation. Virgin rabbits were fed 120 g of solid food per day. After copula, 300 g of food was provided to pregnant females. Feed was recorded daily, while female body weight was measured weekly. All rabbits used in experimental procedures were euthanized with an overdose of sodium pentobarbital (90 mg/kg; i.p.). The feed was stopped on the day before sacrifice at the end of the experimental period. Blood was obtained from the hearts of rabbits from different groups. The serum was separated and kept at −80 °C until use. Immediately after death, the right and left uterine horns were excised. For biochemical measures, a portion of the central uterine horns corresponded to the inter-implantation sites (among two implantation sites) and the other corresponded to the implantation sites (including the placenta; only at days 10 and 20 of pregnancy). On day 20 of pregnancy, the number of implantations and reabsorptions in both uterine horns were counted. The fetus’s body weight, the length of the body, and the abdomen diameters were measured in control and MMI-treated females. The protocol was approved by the Ethics Committee of the Universidad Autónoma de Tlaxcala (accepted on 28 August 2015), following the guidelines of the Mexican Law for the Production, Care, and Use of Laboratory Animals. In dams from all groups, serum concentrations of total triiodothyronine (T3), thyroxine (T4), insulin, and progesterone were quantified using chemiluminescence by Diagnóstico Molecular y Servicio de Referencia S.A. de C.V. (Diagmo Laboratory, Mexico City, Mexico) [ 1 , 29 ]. The serum concentration of estradiol in dams at 20 days of pregnancy was evaluated using an EIA kit (Cayman Chemical Company, Ann Arbor, MI, USA). Assays were carried out according to the manufacturer’s instructions. Serum glucose was measured using a kit (ELITech Clinical Systems, ELITechGroup, Paris, France). Portions of left uterine horns (50 mg) from the inter-implantation and implantation sites were disrupted in lysis buffer. Uterine glycogen content was measured with the same kit using a procedure reported elsewhere [ 1 ]. Glycogen content in the implantation and inter-implantation sites was calculated as the difference between glycogen chains and free glucose (µmol glucosyl units g −1 uterus). Inter- and implantation sites from the central portion of the left uterine horns (50 mg) were disrupted in lysis buffer, as previously reported [ 3 ]. Total protein extracts were obtained via centrifugation at 13,400 rpm for 30 min at 4 °C and quantified afterward. Thus, 30–100 µg of protein was used for the Western blots’ assays depending on the antibody. SDS-PAGE was carried out using 10% acrylamide gels. Later, proteins were transferred to nitrocellulose membranes (BIO-RAD, Hercules, CA, USA, 0.45 µM) and stained with Ponceau’s Red to confirm that protein content was equal in all lines. Subsequently, membranes were blocked with 5–17% non-fat dried milk (depending on the antibody) diluted in 0.02% Tween 20-PBS. Then, they were incubated overnight at 4 °C with antibodies anti-GLUT4 (Santa Cruz Biotechnology, Dallas, TX, USA; sc53566; 1:250), anti-P450 aromatase (Novus Biologicals, Centennial, CO, USA; 1:250; NB100-1596), and anti-LDH (Abcam, Cambridge, UK; 1:200; 98A-1F9BB1. Membranes were incubated for two hours with the secondary antibody 1:250 for GLUT4, 1:1000 for LDH and GLUT4, and 1:20,000 for aromatase. For LDH and aromatase, chemiluminescent signals were detected using a chemiluminescence kit (West Pico Signal, Thermo Fisher Scientific, Waltham, MA, USA) and analyzed with a chemiluminescent signal analyzer (MyECL Imager, Thermo Fisher Scientific). For GLUT4, the signals were captured using photographic plates and Kodak development solutions (developer and mixer) via Kodak GBX Carestream Dental, Rochester, NY, USA. The relative band density for the antigen–antibody complex was calculated using the Image J 14.45S software (National Institute of Health, Bethesda, MD, USA) and normalized to the density of bands covering at least 90% of the length of each Ponceau’s Red-stained Lane [ 3 ]. Data are expressed as the mean ± SEM. Two-way ANOVA tests were performed to compare variables between groups considering the time of pregnancy and MMI treatment as factors in the inter-implantation sites. Two-way ANOVA tests were also applied to investigate the impact of hypothyroidism at implantation sites on days 10 and 20 of pregnancy, considering the type of site and MMI treatment as factors. According to normality tests, Student’s t - or U-Mann–Whitney tests were carried out to determine significant differences between the fetuses’ characteristics. Values of p < 0.05 were considered statistically significant and derived using the GraphPad Prism version 5.01 software.

In human and animal models, hypothyroidism is related to infertility, uterine hyperplasia, uterine inflammation, prematurity, spontaneous abortion, preeclampsia, and low birth weight [ 1 , 2 , 3 ]. In humans, maternal hypothyroidism has been associated with large babies [ 4 ] and a high risk of preterm birth and low birth weights [ 5 ]. In rats, maternal hypothyroidism induced by thyroidectomy or propylthiouracil reduces the weight of the uterus, placenta, and embryos [ 6 ]. Methimazole (MMI) treatment for 50 days in female rabbits also reduces abdominal size in fetuses [ 3 ]. Endometrium and fetal tissues contain thyroid hormone transporters, deiodinases, and alpha and beta subtypes of thyroid hormone receptors [ 7 ]. Glucose, pyruvate, and lactate are energy sources during implantation, placentation, and embryo development. During pregnancy, glycogen is stored in the uterine glands and used by trophoblast and embryonic tissues, being higher toward implantation and decidualization times [ 8 , 9 ] and toward the end of pregnancy [ 10 ]. Hexokinase 1, phosphoglycogen synthase, glycogen synthase, glycogen phosphorylase, glucose-6-phosphate-dehydrogenase, and pyruvate dehydrogenase kinase are also present in the uterus, varying among implantation and inter-implantation sites during the decidualization process [ 9 , 11 ]. The glucose uptake in uterine and placental tissues depends on the different glucose transporters, such as GLUT1, GLUT3, GLUT4, and GLUT8 [ 12 , 13 ]. GLUT1 is increased during placentation and at implantation sites [ 11 ]. Lactate regulates immune cell activity in the uterus [ 14 ], promotes decidualization, and upregulates GLUT4, GPR81, and VEGF in Ishikawa cells [ 15 ], as well as inducing uterine muscle contraction during labor [ 16 ]. Lactate and lactate dehydrogenase (LDH) are high in implantation vs. inter-implantation sites during decidualization [ 11 ]. Changes in the uterine glycogen content have been related to pre-eclampsia and abortions. Glycogen accumulation in the placenta has been reported in preeclamptic pregnancies [ 17 ]. In contrast, placenta, spongiotrophoblast, and glycogen cells are diminished in mice with abortions [ 18 ]. In rats, hypothyroidism also reduces the labyrinth zone, spongiotrophoblast, and vascular sinus but increases the area covered by glycogen cells at 18 days of pregnancy [ 6 , 19 ]. The uterine accumulation of glycogen and the expression of GLUT4 are promoted by estradiol but reduced by progesterone [ 20 , 21 ]. Uterine lactate and LDH are also regulated by estradiol and progesterone in mice [ 11 ]. In addition, thyroid hormones modulate the synthesis and actions of estrogen in the uterus [ 22 , 23 ] and the expression of their nuclear receptors [ 1 ]. The regulation of estradiol and progesterone serum levels by hypothyroidism is controversial and depends on the treatment used to induce this thyroid dysfunction (propylthiouracil or MMI). Thus, both reductions [ 24 ] and increases [ 25 ] in estrogen levels have been reported, while an increase [ 25 ] or a null effect [ 24 ] in progesterone levels has been observed. Hypothyroidism has been related to pregnancy pathologies affecting both dam and progeny health [ 2 , 4 , 5 , 7 ]. However, the mechanisms involved in this are unknown. To approach this, we determined the effect of hypothyroidism on serum glucose, insulin, estradiol, and progesterone levels in rabbits. Since hypothyroidism increases the uterine population of glycogen cells during early pregnancy [ 6 , 19 ], we analyzed the amount of glycogen and the expression of LDH, GLUT4, and aromatase (as an indicator of estradiol synthesis) in the uterus of hypothyroid rabbits during implantation (day 5 of pregnancy), placentation (day 10 of pregnancy), and placenta maturation (day 20 of pregnancy) [ 26 , 27 ]. These molecules were quantified and compared in implantation and inter-implantation sites on days 10 and 20 of pregnancy. Although rabbits are polytocous, they have hemochorial-type placentation resembling humans [ 28 ].

The food intake of each group during the last six days of treatments was compared. Dams from the two groups ate more food on days 5 and 10 of pregnancy and decreased their food intake at 20 days compared to virgin rabbits ( Figure 2 a; time of pregnancy, F = 7.5; p = 0.0004; MMI treatment, F = 0.9; p = 0.33; interaction, F = 0.08; p = 0.96). MMI treatment did not modify the significant differences at any time of pregnancy. The body weight gain was significantly higher in control dams on the 20th day of pregnancy compared to the virgin dams and females at day 5 of pregnancy ( Figure 2 b; time of pregnancy, F = 3.0; p = 0.03; MMI treatment, F = 4.7; p = 0.03; interaction, F = 2.5; p = 0.06). No significant changes were measured in MMI-treated dams at any time of pregnancy. The body weight gain of MMI-treated females was lower than that of the controls when measured on day 20 of pregnancy. In control dams, serum T4 levels decreased significantly during pregnancy. In MMI-treated dams, the concentration of T4 was similar between H0, H5, H10, and H20. MMI treatment did not lead to significant differences in T4 levels compared to the control, except on day 0 (C0 vs. H0) ( Figure 2 c; time of pregnancy, F = 7.9; p = 0.0003; MMI treatment, F = 8.8; p = 0.004; interaction, F = 2.7; p = 0.05). The concentration of T3 was stable during pregnancy in both control and MMI-treated dams ( Figure 2 d; time of pregnancy, F = 1.2; p = 0.32; MMI treatment, F = 19.5; p < 0.0001; interaction, F = 0.3; p = 0.82). However, a significant reduction in the MMI-treated vs. control group was found on day 0. Pregnancy did not change the serum concentration of glucose ( Figure 2 e; time of pregnancy, F = 1.0; p = 0.38; MMI treatment, F = 0.27; p = 0.6; interaction, F = 0.4; p = 0.72) and insulin ( Figure 2 f; time of pregnancy, F = 0.2; p = 0.84; MMI treatment, F = 0.3; p = 0.53; interaction, F = 0.1; p = 0.95) in control and MMI-treated rabbits. Glycogen was identified in diverse cells such as epithelium, stroma, smooth muscle, decidua, giant, and glycogen trophoblast ( Figure 3 a). In the inter-implantation sites of the control and MMI-treated rabbits, the uterine glycogen content was significantly low on day 5 of pregnancy. The control and MMI-treated groups showed no differences in uterine glycogen content between days 0, 10, and 20 of pregnancy. Compared to control groups, MMI treatment increased the uterine glycogen content on day 20 of pregnancy ( Figure 3 b; time of pregnancy, F = 33.4; p < 0.0001; MMI treatment, F = 4.6; p = 0.03; interaction, F = 2.6; p = 0.05). On day 10 of pregnancy, the glycogen concentration in the uterus was similar in the inter-implantation and implantation sites in control and MMI-treated dams. In the implantation sites, MMI-treated pregnant rabbits had a higher uterine glycogen content compared to controls ( Figure 3 b; uterine site, F = 1.0; p = 0.30; MMI treatment, F = 4.1; p = 0.05; interaction, F = 3.1; p = 0.08). On day 20 of pregnancy, the glycogen concentration in the uterus was similar in the inter-implantation and implantation sites in control and MMI-treated dams. In both the inter-implantation and implantation sites, MMI-treated rabbits had a higher uterine glycogen content compared to controls ( Figure 3 b; uterine site, F = 3.9; p = 0.06; MMI treatment, F = 12.5; p = 0.002; interaction, F = 1.0; p = 0.32). The LDH expression in the inter-implantation sites did not vary significantly during pregnancy in control rabbits. In MMI-treated groups, a significant increase in the expression of LDH was found on day 20 of pregnancy compared to H0, H5, and H10. On day 20 of pregnancy, the value of LDH in the MMI-treated group was higher than that in the control group ( Figure 3 c; time of pregnancy, F = 8.9; p < 0.0001; MMI treatment, F = 1.5; p = 0.21; interaction, F = 7.2; p = 0.0005). On day 10 of pregnancy, the glycogen concentration in the uterus was similar in the inter-implantation and implantation sites in control and MMI-treated dams. MMI treatment did not modify the expression of LDH in the inter-implantation and implantation sites ( Figure 3 c; uterine site, F = 2.6; p = 0.11; MMI treatment, F = 0.5; p = 0.81; interaction, F = 2.9; p = 0.10). On day 20 of pregnancy, the LDH expression in the uterus was similar in the inter-implantation and implantation sites in control and MMI-treated dams. In both the inter-implantation and implantation sites, MMI-treated rabbits showed increased uterine LDH expression compared to controls ( Figure 3 c; uterine site, F = 1.5; p = 0.22; MMI treatment, F = 23.9; p < 0.0001; interaction, F = 0.95; p = 0.33). The expression of GLUT4 was identified in different cells, such as epithelium, stroma, smooth muscle, decidua, giant, and glycogen trophoblast ( Figure 4 a). In inter-implantation sites, the expression of GLUT4 was higher during pregnancy than in virgin females in the control groups. In MMI-treated groups, the expression of GLUT4 was significantly lower at 5 and 10 days of pregnancy compared to virgin dams and those at 10 days of pregnancy. Compared to control groups, the MMI-treated virgin group showed a higher expression of GLUT4, but this was significantly lower at 5, 10, and 20 days ( Figure 4 b; time of pregnancy, F = 38.0; p < 0.0001; MMI treatment, F = 14.3; p = 0.0005; interaction, F = 44.8; p < 0.0001). On day 10 of pregnancy, the GLUT4 expression in the implantation sites was higher than that in the inter-implantation sites in both control and MMI-treated dams. MMI treatment reduced the expression of GLUT4 in the inter-implantation and implantation sites ( Figure 4 b; uterine site, F = 53.4; p <0.0001; MMI treatment, F = 21.0; p = 0.0002; interaction, F = 0.72; p = 0.40). On day 20 of pregnancy, the GLUT4 expression in the uterus was similar in the inter-implantation and implantation sites in control and MMI-treated dams. MMI-treated rabbits had a lower uterine GLUT4 expression compared to controls in the inter-implantation site but a higher expression in the implantation site ( Figure 4 b; uterine site, F = 18.3; p = 0.0004; MMI treatment, F = 8.2; p = 0.009; interaction, F = 53.9; p <0.0001). In control groups, the serum concentration of progesterone was significantly higher during pregnancy (days 5, 10, and 20) compared to virgin rabbits. A similar pattern was found in MMI-treated female rabbits on days 5, 10, and 20; these rabbits had higher levels of progesterone than virgin rabbits. No significant differences were observed between control and MMI-treated dams at any time of pregnancy ( Figure 5 a, pregnancy time, F = 25.3; p < 0.0001 MMI treatment, F = 0.66; p = 0.42; interaction, F = 1.7; p = 0.17). The serum concentration of estradiol on day 20 of pregnancy was unaffected by MMI treatment (U = 12.0; p = 1.0; Figure 5 b). In inter-implantation sites, the aromatase expression in the uterus was similar between control groups at any time of pregnancy. In contrast, MMI-treated females showed a lower expression of aromatase at day 10 of pregnancy compared to virgin rabbits and those at days 5 and 20 of pregnancy. A significant difference between control and MMI-treated animals was observed at G0 and G20 ( Figure 5 c; time of pregnancy, F = 7.9; p = 0.0003; MMI treatment, F = 23.4; p <0.0001; interaction, F = 2.1; p = 0.10). On day 10 of pregnancy, the aromatase expression was similar in the inter-implantation and implantation sites in control and MMI-treated dams. MMI treatment did not affect the expression of aromatase in the inter-implantation and implantation sites ( Figure 5 c; uterine site, F = 8.5; p = 0.0008; MMI treatment, F = 1.0; p = 0.031; interaction, F = 0.06; p = 0.80). On day 20 of pregnancy, the aromatase expression in the uterus was similar in the inter-implantation and implantation sites in control and MMI-treated dams. MMI-treated rabbits had a higher uterine aromatase expression compared to controls in the inter-implantation and implantation sites ( Figure 5 c; uterine site, F = 4.2; p = 0.05; MMI treatment, F = 27.8; p < 0.0001; interaction, F = 0.77; p = 0.38). Aromatase was localized in the epithelium, stroma cells, smooth muscle cells, decidua cells, spongioblast, giant cells, etc. ( Figure 5 d). Control dams on day 20 of pregnancy had 6–10 fetuses, while MMI-treated dams had 9–12 fetuses. The characteristics of fetuses, such as the number of live fetuses, number of resorptions, body weight, length, and abdominal diameter of fetuses from the control and MMI-treated females, were similar ( Table 1 ). We created a distribution histogram of fetuses according to body weight, which led us to consider three groups: slim (4.6 g). Therefore, we grouped fetuses according to body weight and compared the media for each group. MMI treatment did not affect the mean values for the percentage of slim, medium, or heavy fetuses compared to the control group ( Table 1 ).

In the present study, the efficacy of MMI treatment for inducing hypothyroidism in virgin rabbits was confirmed by the reduction in T4 and T3 levels compared to the control group. This agrees with previous reports on virgin rabbits [ 1 , 29 ]. During pregnancy, hypothyroidism was also found in MMI-treated groups. However, no significant differences between the pregnant control and MMI-treated groups were observed following pregnancy-induced natural hypothyroidism. This was indicated by the reduction in T4 concentration, as reported in humans [ 30 ]. In agreement with these results, rats are found to have a low level of T4 on day 20 of pregnancy [ 31 ]. Also, MMI treatment did not modify food intake and body weight gain in female rabbits, suggesting that the proportion of muscle may be replaced by fat mass, as previously reported for hypothyroidism [ 32 ]. The serum glucose and insulin concentrations were unaffected by MMI treatment in virgin and pregnant rabbits. Indeed, no modifications in these variables have been previously reported in virgin rabbits [ 33 ] or in virgin or pregnant rats [ 31 , 34 ]. In rabbits, GLUT4 expression was detected in the uterine epithelium, stroma, endometrial glands, and myometrium, as in rats [ 35 ]. The expression of GLUT4 in the endometrium changes with the menstrual phase, being higher in the follicular phase and lower in the luteal phase [ 36 ]. In virgin rabbits, hypothyroidism did not affect the uterine glycogen concentration or the expression of LDH, but it increased the expression of GLUT4 and aromatase. Furthermore, no changes were found in the serum concentrations of progesterone and estradiol due to hypothyroidism. In agreement with this, a null effect of hypothyroidism on these hormones in virgin rabbits was previously reported [ 29 ]. Despite this, MMI treatment affects the expression of estrogen and progesterone receptors in the uterus of virgin rabbits [ 1 ], suggesting that hypothyroidism affects actions involving uterine estrogen. Hypothyroidism promotes the infiltration of immune cells and hyperplasia in the virgin uterus [ 1 ]. In agreement, a high expression of uterine aromatase is related to inflammation, as reported in endometriosis [ 37 ]. Thus, the high expression of GLUT4 and aromatase in the uterus of virgin hypothyroid rabbits may be associated with an inflammatory condition. In pregnant control rabbits, uterine glycogen concentration was low on day 5 of pregnancy (implantation) compared to the rest of the pregnancy. No changes were observed in LDH and aromatase, but a low expression of GLUT4 was observed on days 5 and 10 compared to day 20 of pregnancy. In rabbits, GLUT4 expression was detected in the uterine epithelium, stroma, endometrial glands, myometrium, and decidual zone, as in rats [ 35 ]. The aromatase expression was localized in the endometrium and myometrium of rabbits, as in humans [ 38 ]. In rodents, glycogen decreases upon implantation and increases during decidualization [ 9 ]. Lactate regulates uterine muscle contraction during labor [ 16 ]. Uterine glycogen concentration [ 20 ] and GLUT4 expression [ 13 ] are regulated via estradiol and progesterone. This regulation involves the effects of steroid hormones on insulin actions [ 21 ]. During pregnancy, MMI treatment increased the uterine glycogen concentration and the expression of LDH and aromatase but reduced the expression of GLUT4 at the inter- and implantation sites. We also found that serum progesterone levels were increased by pregnancy but unmodified by MMI treatment. On day 20 of pregnancy, the serum estradiol levels were similar between the control and MMI-treated groups. In agreement with our results, an increase in the uterine population of glycogen cells and a high amount of glycogen in the fetal placenta have been reported in thyroidectomized rats [ 6 , 34 ]. The concentration of lactate increases at the end of pregnancy and, according to our results, this may be related to prematurity induced by hypothyroidism [ 5 ]. The reduction in GLUT4 in the pregnant uterus of hypothyroid dams could be related to inflammation since a previous report indicated that the pregnant uterus of hypothyroid rabbits has endometrial hyperplasia [ 3 ]. In addition, an increase in the expression of inflammation markers like TNFα, IL10, IL6, and HIF1α was reported in the endometrium of pregnant hypothyroid rats [ 19 ]. Hypothyroidism also regulates the expression of integrin avß3, leukemia inhibitory factor, and Mucin 1, which are involved in embryo implantation [ 25 ] and associated with the expression of VEGF and vascularization of the placenta [ 6 ]. In agreement, a reduction in the endometrial expression of GLUT4 has also been reported in obese patients with polycystic ovary syndrome (PCOS), associated with high levels of TNF-α [ 39 ]. The possible inflammation induced by hypothyroidism in the uterus of virgin and pregnant rabbits, which modifies the concentration of glycogen and the expression of LDH, GLUT4, and aromatase, could have consequences for the progeny. In the present study (until day 20 of pregnancy), no differences were found in the number of implantations, reabsorptions, or fetal morphometry. We previously reported that prolonged treatment with MMI (50 days) reduces the size of fetuses [ 3 ]. As day 20 of pregnancy, for rabbits, is the end of the second trimester of pregnancy, we do not know whether hypothyroidism could affect embryo development. This is a limitation of this study. Thus, it is necessary to analyze the body weight at birth. In this regard, the consequences of changes in the glycogen reserve and estradiol could be larger babies [ 4 ], a high risk of preterm birth, including low birth weight [ 5 ], or even alterations in organogenesis and neural development [ 40 ]. Another limitation of the present study is that we did not measure the concentration of estradiol and progesterone directly in the uterine tissues. Also, we do not know the sex of each analyzed fetus. Female and male fetuses could have different morphometric characteristics. In contrast, a strength of this investigation is that it shows that hypothyroidism can modify the metabolic and hormonal environment in both implantation and inter-implantation sites in animal models in the first and second trimesters of pregnancy. This study could provide a basis for future studies researching metabolic changes in implantation sites that could affect implantation, placentation, and embryo development. In conclusion, our results suggest that hypothyroidism may affect glucose uptake in the pregnant uterus, which is associated with the local synthesis of estradiol, as can be inferred by the differences in the expression of aromatase depending on the stage of pregnancy.