A novel PCOS rat model and an evaluation of its reproductive, metabolic, and behavioral phenotypes

Abstract Background Although animal models of PCOS have been used in many studies, none of them can reproduce both the reproductive and metabolic phenotypes of PCOS. In addition, behavioral parameters have not been evaluated in PCOS animal models. Purpose We tried to produce an improved rat model of PCOS, and the reproductive, metabolic, and behavioral phenotypes of the model rats were evaluated. Methods Female rats were implanted with silicon tubes containing oil‐dissolved dihydrotestosterone (Oil‐DHT) as a new PCOS model. Their phenotypes were compared with those of conventional PCOS model rats (DHT), into which tubes containing crystalline DHT were implanted, and non‐DHT‐treated rats (control). Results Both the Oil‐DHT and DHT rats showed greater body weight gain, food intake, and fat depot weight than the control rats. Furthermore, these groups showed fewer estrous stages and increased numbers of cystic follicles. The DHT rats exhibited lower ovarian and uterine weights than the control rats, whereas no such changes were observed in the Oil‐DHT rats. The Oil‐DHT and DHT rats showed less locomotor activity in the light phase than the control rats. Conclusions Our proposed PCOS model reproduced both the reproductive and metabolic phenotypes of PCOS and may have potential for PCOS research.

from early in life, and hence, have been used as rodent models of PCOS. 9,10 As PCOS is a systemic disease, these animal models are useful for evaluating the pathophysiology of PCOS in several tissues. On the other hand, no PCOS animal models can completely reproduce the reproductive and metabolic phenotypes of PCOS; that is, some animal models can reproduce the reproductive phenotype well, but poorly reproduce the metabolic phenotype, whereas the opposite is true of other animal models. 9,10 Among the various PCOS animal models, rats treated with dihydrotestosterone (DHT) until the prepubertal period have been used in basic research. [9][10][11][12][13] Although this animal model reproduces some reproductive and metabolic phenotypes of PCOS well, such as an irregular estrous cycle and increases in body weight and fat depot weight, it poorly reproduces the associated changes in ovarian weight and/or ovarian morphology. Most women with PCOS show enlarged ovaries and PCO, 4,14 whereas DHT-induced PCOS model rats show low ovarian weights and do not always exhibit an increased number of cystic follicles. [9][10][11][12][13] In addition, in our preliminary examination it was confirmed that DHT-induced PCOS model rats had low uterine weights. It is possible that these phenotypic discrepancies between human patients and animal models may hamper the evaluation of the pathophysiology of PCOS at the systemic level. Thus, improved PCOS rat models, which reproduce the hallmarks of PCOS more closely, are desired.
Another issue with PCOS animal models is that the safety of chronic hormonal manipulation has not been fully elucidated.
Although it has been reported that the long-term administration of androgens might have some adverse effects, such as liver and kidney toxicity, hepatic and renal function have not been measured in most androgen-induced PCOS animal models. In addition, although it has been suggested that gonadal hormones affect locomotor activity and thermogenesis, 15-17 these physiological parameters have not been measured in any PCOS animal model. Usually, DHT pellets or silicon tubes containing crystalline DHT are used to produce DHT-induced PCOS animal models. [9][10][11]13 However, we suspected that these procedures may result in DHT having supra-physiological effects on tissues and organs, and may induce some of the phenotypic discrepancies observed between PCOS animal models and human PCOS, such as small ovaries and uteri. On the contrary, it has been well established that the implantation of a silicon tube containing oil-dissolved estradiol can reproduce the physiological hormonal milieu in female rodents. 18 Thus, in this study we tried to establish an improved PCOS rat model using a similar method; that is, the implantation of a silicon tube containing oil-dissolved DHT. The reproductive and metabolic phenotypes of this new PCOS rat model were compared with those of a conventional PCOS rat model. In addition, hepatic and renal function were measured to confirm the safety of this new animal model, and locomotor activity and core body temperature were also assessed to evaluate whether changes in these physiological functions might be related to body weight adiposity in DHT-induced PCOS animal models.

| Animals
Twenty-three-day-old female Wistar rats were purchased from Charles River Laboratories Japan, Inc. (Kanagawa, Japan). They were housed in the animal laboratory of Tokushima University under a 12-hour light/dark cycle (lights turned on at 0800 and turned off at 2000) and controlled temperature (24℃) conditions and were given free access to standard chow (type MF; Oriental Yeast Co. Ltd., Tokyo, Japan; 359 kcal/100 g, 12.8% of the provided calories were derived from fat, 25.6% were from protein, and 61.6% were from carbohydrates) and water. In total, 49 rats were used in this study (29 rats were used for experiment 1, and another 20 were used for ex-

| Effects of chronic DHT administration on reproductive and metabolic functions
Twenty-nine rats were divided into conventional PCOS model (DHT, n=10), new PCOS model (Oil-DHT, n=9), and non-DHT-treated (control, n=10) groups. In the DHT group, each rat was implanted with a silicon tube filled with crystalline DHT (As one Co., Ltd., Tokyo, Japan; inner diameter, 3 mm; outer diameter, 5 mm; length of the filled part, 5 mm). In the Oil-DHT group, each rat was implanted with a silicon tube filled with diluted DHT; that is, DHT dissolved in a solution of 80% peanut oil and 20% ethanol (16 mg/mL) (As one Co., Ltd., Tokyo, Japan; inner diameter, 3 mm; outer diameter, 5 mm; length of the filled part, 10 mm). In the control group, each rat was implanted with an empty tube. All surgery was done on postnatal day (PND) 26, and the rats were housed individually after the surgery.
Body weight and food intake were measured every week. Food was placed in the food space of the wire mesh top, and the remaining food weight was measured. Wood chip bedding was changed at measurement, and broken food in the bedding was collected at this time point. In addition, estrous cyclicity was checked for ten days from PND48 to PND57.
Specifically, a glass pipette filled with sterilized water was inserted into the vaginal orifice to a depth of 5 mm, and the vagina was flushed two or three times. Then, a small sample of the collected fluid was dropped onto a slide and dried in air. All of these slides were stained with Giemsa stain.
Then, cytological examinations were performed, and the stages of the estrous cycle (proestrus, estrus, metestrus, and diestrus) were analyzed based on the relative numbers of each cell type in the vaginal smears.
At PND58, the rats were anesthetized and killed by decapitation after 16 h fasting. Then, their blood, visceral fat, subcutaneous fat, ovaries, and uteri were collected. Each sample was around 300-400 mm 3 in size, and the histological tissue samples were fixed in 4% paraformaldehyde, embedded in paraffin, and stained with hematoxylin and eosin. Prior to fat sampling, the visceral and subcutaneous fat were peeled away from the surrounding tissue and weighed immediately. In addition, weights of uterus and bilateral ovaries were measured. Whole blood was centrifuged at 3000 rpm for 20 min at 4℃, and the serum was removed and stored at −40℃, before being used for the subsequent analyses.
The serum levels of blood urea nitrogen, creatinine, aspartate aminotransferase, alanine aminotransferase, glucose, triglycerides, total cholesterol, high-density lipoprotein cholesterol, and low-density lipoprotein cholesterol were measured by a commercial laboratory (Oriental Yeast Co., Ltd). The serum levels of estradiol, progesterone, and dihydrotestosterone were measured using liquid chromatography-tandem mass spectrometry

| Effects of chronic DHT administration on locomotor activity and body temperature
The other 20 female rats were divided into DHT (n=9), Oil-DHT (n=5), and control (n=6) groups. Appropriate tubes were implanted at PND27 as described above. Around PND50, pre-calibrated radio telemetry transmitters (TA11TA-F10; Data Sciences International, New Brighton, MN, USA) were surgically implanted into the rats. The rats were allowed to recover for around one week, and then their core body temperature and locomotor activity were measured for 48 h. The radio signals were recorded every 15 min and directly converted into body temperature and locomotor activity data using the DATAQUEST software (Data Sciences).

| Statistical analyses
All results are expressed as mean ±standard error of the mean (SEM) values. Comparisons among the DHT, Oil-DHT, and control groups were conducted using one-way or two-way repeated-measures analysis of variance (ANOVA) followed by the Tukey-Kramer or Dunnett post hoc test. P-values of <0.05 were considered to be statistically significant in all statistical comparisons.

| Effects of chronic DHT administration on reproductive and metabolic functions
The intervention had significant effects on the observed body weight Oil-DHT groups than in the control group ( Figure 1A). In addition, at PND54 the mean body weight of the DHT group was heavier than that of the Oil-DHT group ( Figure 1A). Similarly, the intervention had  Almost all of the rats in the control group showed four-to-fiveday regular estrous cycles, whereas many of the rats in the DHT and Oil-DHT groups showed acyclic or irregular cycles ( Figure 4A).
During the 10-day assessment period, the number of estrous stages was lower in the DHT and Oil-DHT groups than in the control group ( Figure 4B). Ovarian weight and uterine weight were lower in the DHT group than in the control group, whereas no such differences were seen between the Oil-DHT and control groups ( Figure 4B).
Serum estradiol and dihydrotestosterone levels were not different among three groups, whereas serum progesterone level in Oil-DHT and DHT groups was significantly lower than that in control group ( Figure 4C).
The ovaries of the rats in the control group had normal morphologies, whereas ovaries with a polycystic morphology were seen in the DHT and Oil-DHT groups ( Figure 5A). The number of cystic follicles was significantly larger in the DHT and Oil-DHT groups than in the control group ( Figure 5B). The number of corpora lutea was significantly lower in the DHT and Oil-DHT groups than in the control group ( Figure 5B). In addition, the number of corpora lutea was greater in the Oil-DHT group than in the DHT group. These results are summarized in Table 1.

F I G U R E 4 (A)
The estrous cyclicity of three representative rats from each group and (B) the number of estrous stages during the 10-day assessment period and ovarian and uterine weight on the day of tissue sampling in the control, Oil-DHT, and DHT groups (n=9−10 per group). Data are expressed as the mean ±SEM. * p <0.05, ** p <0.01 vs. control group  Figure 7A). In the light phase, locomotor activity was lower in the DHT and Oil-DHT groups than in the control group, whereas no intergroup differences in locomotor activity were seen in the dark phase ( Figure 7B). These results are summarized in Table 1.

| DISCUSS ION
PCOS is a common endocrine disorder among women of reproductive age, and it is associated with reproductive disorders, such as infertility and anovulation-related menstrual disorders, and metabolic disturbances. 1-3,19,20 As noted above, various kinds of animal  12 On the other hand, these animal models displayed disrupted ovulation and body weight and body composition abnormalities; that is, they showed disturbed estrous cyclicity and increases in body weight and fat depot weight. [11][12][13]21 Thus, a PCOS animal model that reproduces most PCOS phenotypes and overcomes the discrepancies between conventional animal models and human PCOS is desired. 22 In this study, we proposed a new PCOS rat model that reproduces the reproductive and metabolic features of PCOS without any atrophic changes in the ovaries or uterus. In addition, similar to those used in previous studies, this model reproduces most PCOS phenotypes; that is, increases in body weight gain, food intake, and fat depot weight; disturbed estrous cyclicity; and PCO morphology. 9,10 In our previous studies, PCOS model rats were created by implanting a silicon tube containing crystalline DHT into each rat, whereas a tube containing dissolved DHT was implanted into each rat in this study. 11 Other problems with PCOS animal models include the fact that the safety of chronic hormone manipulation has not been fully evaluated and that changes in locomotor activity and thermogenesis have never been examined in such models. It is well known that the longterm administration of androgens often has adverse effects and that gonadal hormones, for example, estrogen and progesterone, affect thermogenesis in females, including experimental animals. 16,17,23 In addition, it has been reported that estrogen increases locomotor activity in female rodents. Female rats show the greatest locomotor activity in the proestrus stage, when they also display concomitant increases in serum estrogen levels, and estrogen supplementation increases locomotor activity in ovariectomized female rats. [24][25][26] These effects of estrogen on locomotor activity are not observed in mice lacking the estrogen receptor. 18 On the other hand, the effects of androgens on female locomotor activity, especially in androgen-induced PCOS model animals, have not been evaluated. If the core body temperature and/ or locomotor activity of PCOS model animals are altered, such changes may be related to the metabolic disturbances seen in such animals.
Thus, in this study we evaluated the safety of chronic DHT administration, as well as locomotor activity and body temperature, in new and conventional DHT-induced PCOS model rats. As a result, we found that hepatic and renal function and core body temperature were not affected by the administration of DHT in these animal models. On the other hand, locomotor activity was decreased in the light phase, but not in the dark phase, in both the new and conventional models. In rodents, food intake and locomotor activity are both increased in the dark phase, whereas they are decreased in the light phase, during which most time is spent sleeping. 27 It is possible that in the present study, both types of PCOS model rats spent more time sleeping or exhibited decreased activity, even when awake, and that these changes might be, at least in part, related to the increases in body weight gain and fat depot weight they exhibited. In addition to metabolic, endocrine, and immunological evaluations, examinations focusing on behavioral phenotypes are needed to clarify the pathophysiology of the metabolic disturbances seen in PCOS.
In conclusion, we proposed a new DHT-induced rat model of PCOS, which reproduces both the reproductive and metabolic phenotypes of PCOS well. In particular, this model does not induce atrophic changes in the ovaries or uterus, which are observed in conventional DHT-induced PCOS model animals. We also clarified that

CO N FLI C T S O F I NTE R E S T
The authors declare that no conflicts of interest exist.

E TH I C A L A PPROVA L
This article does not describe any experiments involving human participants. All of the institutional and national guidelines for the care and use of laboratory animals were followed. The protocol for the research project was approved by a suitably constituted ethics committee.