Introduction
At low latitude (< 25°C), and without the influence of photoperiodic seasonal anestrus, the reproductive performance of caprine native populations depends especially on social and nutritional factors (Walkden-Brown & Restall, Reference Walkden-Brown and Restall1996). Although tropical breeds have a great reproductive potential during the entire year, they can exhibit a long anestrus period or anovulation because of nutritional deficiency, caused by harsh environmental conditions (Walkden-Brown & Restall, Reference Walkden-Brown and Restall1996; Chilliard et al., Reference Chilliard, Bocquier and Doreau1998). Therefore, energy balance is an essential condition for maintenance of reproduction, as well as for estrus synchronization and superovulatory responses. For instance, feeding previously undernourished goats improved their response to estrus synchronization (Paula et al., Reference Paula, Galeati, Teixeira, Lopes Junior and Freitas2005).
Nutritionally induced anestrus accompanies ovulation failure because of insufficient circulating LH, responsible for stimulation of follicular maturation. This alteration occurs due the suppression of gonadotrophin-release hormone (GnRH) secretion, interfering with all other endocrine functions related to reproduction (Garcia-Garcia, Reference Garcia-Garcia2012). Dominant follicles and corpora lutea are smaller and concentrations of circulating LH are lower in animals challenged with dietary restriction (Rhodes et al., Reference Rhodes, Entwistle and Kinder1996). Although goats from tropical areas are exposed to undernutrition conditions, they are still submitted to reproductive techniques such as estrus synchronization and superovulation. There are some alternatives such as combining superovulation with insulin administration, which acts synergistically with FSH and LH to stimulate granulosa cells mitosis (Selvaraju et al., Reference Selvaraju, Agarwal, Karche and Majumbar2003). However, one must bear in mind that this is a palliative solution.
Besides affecting follicular maturation and ovulation, hormonal imbalance caused by undernourishment may affect also early folliculogenesis, i.e. those follicles at early stage of development (preantral follicles) that represent the pool of reserve gametes. Although there is a paucity in studies regarding the effect of nutrition on the development of preantral follicles in ruminants, it is known that folliculogenesis is modulated by energy balance (Garcia-Garcia, Reference Garcia-Garcia2012). For example, insufficient dietary intake negatively affects the onset of follicular growth in rat and goat (Lintern-Moore & Everitt, Reference Lintern-Moore and Everitt1978; Rondina et al., Reference Rondina, Freitas, Spinaci and Galeati2005). When sheep are submitted to nutrient restriction during gestation, the fetus will present delayed follicle development (Borwick et al., Reference Borwick, Rhind, McMillen and Racey1997; Rae et al., Reference Rae, Palassio, Kyle, Brooks and Lea2001). In pregnant cows this situation resulted in offspring with smaller ovarian follicular reserve (Mossa et al., Reference Mossa, Carter, Walsh, Kenny and Sith2013). Hormonal treatment with PMSG increases the mitotic index in preantral and very small antral follicles in cattle, without affecting the population of large antral follicles, thus increasing both size and cell proliferation in these early follicular classes (Webb et al., Reference Webb, Campbell, Garverick, Gong and Gutierrez1999). This situation means that during follicular development, oocyte size will increase together with the mitosis of granulosa cells (Munakata et al., Reference Munakata, Kawahara-Miki, Shiratsuki, Tasaki and Itami2016), which can be assessed by diameter measurements and with the use of mitotic markers such as colchicine (Shinohara et al., Reference Shinohara, Matsumoto and Mori1997).
The present study represents an effort to investigate the influence of nutrition on follicular growth and development, especially regarding granulosa cell proliferation, following an initial assessment focused on responsiveness to progestagen–eCG–Cloprostenol treatment of goats challenged with feed restriction and refeeding.
Materials and Methods
Ethical aspects
All experimental protocols were approved by the Ethical Committee in Animal Research of the Universidade Estadual do Ceará, Brazil (no. 09656828-3/09).
Animals and experimental design
This study was conducted at the State University of Ceará (Brazil) located at 3°43´S and 38°30´W. The region is characterized by a tropical constant photoperiod regimen. The mean ambient temperature recorded during the experimental period was 26.8 ± 0.71°C [mean ± standard error of the mean (SEM)]. Fifteen cycling crossbred Saanen × native goats with similar [mean ± standard deviation (SD)] live weight (23.75 ± 1.28 kg), age (1.8 ± 0.16 years) and body condition score (2.2 ± 0.2) were used. The goats were divided into three groups (n = 5 per group) and kept housed at ground level with free access to water and salt. For 24 weeks, goats received elephant grass plus concentrate to provide 1.5 (n = 5) and 0.72 (n = 10) times of the energetic requirement for maintenance of live weight (Agricultural and Food Research Council, 1998). Finally, five underfed goats were refed (1.5 times maintenance) for 6 weeks with the diet of the nourished group. The diets presented the same concentrate-to-roughage ratio (60:40) and were provided twice a day (07:00 and 15:00h). All concentrate rations were isoenergetic (70% of TDN on DM basis) and isonitrogenous (19% CP on DM basis). A buck was introduced twice daily in each pen to detect the natural estrus (09:00 h, 16:00 h). Body weight of the animals was recorded once a week. At the end of each nutritional treatment the estrus was synchronized using 45 mg FGA (Chrono-gest, Intervet, France) vaginal sponge for 11 days and intramuscular injections of 300 IU eCG (Novormon, Syntex, Argentina) and 50 μg cloprostenol (Ciosin, Coopers, Brazil), 48 h prior to sponge removal.
Ovariectomy, preparation of ovaries and classification of follicles
Prior to surgery, goats were deprived of feed and water for 36 h and 24 h, respectively. One ovary, per animal, was removed by unilateral ovariectomy. Animal suffering was avoided by implementing ovariectomy under anesthesia with 0.5 mg/10 kg of xylazine hydrochloride (Coopazine, Coopers, São Paulo, Brazil) and 25 mg/10 kg of ketamine hydrochloride (Ketamina Agener, União Química, Embu-Guaçu, Brazil) associated with local administration of 2% lidocaine hydrochloride (Anestésico L Pearson, Eurofarma, Rio de Janeiro, Brazil) applied to the incision sites. A first ovary was removed by laparotomy 72 h after sponge removal and fixed in Bouin's solution. Immediately, colchicine (C9754, Sigma-Aldrich, USA), (1 mg/kg body weight i.v.) was administrated. Two hours after treatment with colchicine animals were euthanized by the administration of an intravenous injection of sodium pentobarbitone (80 mg/kg). This procedure was in accordance with the recommendations of Reilly (Reference Reilly2001). The remaining ovaries were also fixed in Bouin's solution. The ovaries were then dehydrated in ethanol and embedded in paraffin wax. Every fifth section sliced (8 µm of thickness) was stained with periodic acid Schiff (PAS) and Harris's hematoxylin. Follicles with two layers of granulosa cells or more were classified according to Pedersen & Peters (Reference Pedersen and Peters1968) with modifications, in which small preantral presented two to four layers of cuboidal granulosa cells surrounding an oocyte; large preantral presented more than four layers of cuboidal granulosa cells surrounding an oocyte; small antral was characterized by follicles with an initial formation of the antrum (200 μm diameter of antrum); large antral follicles were those follicles with more advanced antrum formation measuring up to 1.10 mm diameter.
Histological analysis
Images from normal follicles of each nutritional group were randomly captured with a digital camera (TK-C1381: JVC American Corp., Wayne, New Jersey, USA) mounted on a microscope (Eclipse 400: Nikon Instech Co., Ltd, Japan). Follicle, oocyte and antrum diameter were determined using a computer program (Image J: National Institutes of Health, Bethesda, Maryland, USA). The thickness of granulosa cells layer was determined by the difference between follicle and oocyte diameter. Normal and atretic follicles with visible oocyte nucleus were counted. For the antral follicles, the major diameter of oocyte in the cross-section was used as reference. Number of follicles were estimated by the fractionator method, and the ovarian volume was determined by the Cavalieri principle (Gundersen et al., Reference Gundersen, Bendtsen, Korbo, Marcussen and Moller1988).
Growth rate calculation
The number of granulosa cells was counted in small and large preantral follicle images, while for the small and large antral follicles it was estimated according to Lussier et al. (Reference Lussier, Matton and Dufour1987). Ten follicles of each classes and ovary (with or without colchicine) were chosen to count granulosa cells in prophase and metaphase. The mitotic index and follicular growth rate (µm/day) was determined as described by Lussier et al. (Reference Lussier, Matton and Dufour1987).
Statistical analysis
All data were analyzed using SAS (SAS, Inc, Cary, NC, USA). The effect of nutritional treatment was analyzed by GLM procedure for live weight, morphological and growth characteristics of follicles. Comparison between means of nutritional treatments was performed by the Duncan test. Differences among proportions or number were analyzed by chi-squared test. Before statistical analysis, the number of follicles was log-transformed.
Results
Live weight and onset of anestrus
The effects of different nutritional levels on live weight in goats are shown in Fig. 1. After 24 weeks of feeding treatment live weight of goats was 19.24 ± 1.89 kg and 28.46 ± 2.45 kg (mean ± SEM) for underfed and fed group, respectively. Animals of refed group show a body mass recovery superior of 2.6 kg, and a final live weight of 21.83 ± 1.45 kg. Body condition score at the end of experiment was 2.8 ± 0.1 for fed group, 1.4 ± 0.1 in restricted group and 1.9 ± 0.1 for refed animals. In underfed group, the absence of natural estrus was observed in all animals after 16 weeks of restricted feeding (Fig. 1).

Figure 1 Weekly body weight recorded during the experiment for fed group (black), underfed group (grey), and at refeeding period of underfed group (white). Anestrus was achieved in the animals of underfed group at sixteenth week of underfeeding (black arrow). Values are given as mean ± standard error of the mean (SEM).
Ovary volume, number of normal, atretic follicles and percentage of atretic follicles for class
In underfed and refed animals, ovarian volume was smaller (P < 0.01) than the ovarian volume from the fed group (Table 1). However, number of normal and atretic follicles was not affected by the feeding system (P > 0.05; Table 1). The incidence of atresia for each follicular class is depicted in Fig. 2 Underfed treatment shows a higher level of atresia in preantral follicles (P < 0.05) and small antral follicles (P < 0.01) when compared with fed animals. In large antral follicles atresia rate was higher (P < 0.01) in fed animals when compared with underfed and refed groups.
Table 1 Ovarian volume, and mean (± SEM) number of normal and atretic follicles in goats submitted to different nutritional levels

a,b Different lower-case letters indicate significant differences among nutritional groups within the same evaluated parameter; P < 0.05.

Figure 2 Mean [± standard error of the mean (SEM)] percentage of atretic follicles per class in goat ovarian tissue at different nutritional levels. S: Small; L: Large. a,b,cDifferent lower-case letters indicate statistical difference among nutritional groups within each follicular class; P < 0.05.
Morphological characteristics
Morphological characteristics of follicles and granulosa cells mitotic index are shown in Tables 2 and 3. Refed animals presented the smallest oocytes enclosed in preantral follicles (P < 0.01) when compared with fed and underfed animals. Furthermore, both refed and underfed animals presented the smallest oocytes enclosed in antral follicles. However, the effect of nutritional treatment on the follicular diameter was observed only when comparing large antral follicles, in which undernourished animals presented the smallest (P < 0.01) follicles (Table 2). Nutritional treatments did not affect the follicular antrum (Table 2).
Table 2 Morphological characteristics of normal follicles in goats submitted to different nutritional levels

a,b Different lower-case letters indicate significant differences among nutritional groups within the same evaluated parameter and follicular category; P < 0.05.
Table 3 Morphological characteristics and mitotic index of normal follicles in goats submitted to different nutritional levels

a,b Different lower-case letters indicate significant differences among nutritional groups within the same follicular category; P < 0.05.
GCs, granulosa cells.
The number of granulosa cells was lower in underfed animals (P < 0.01) when compared with fed and refed goats (Table 3).
Mitotic index and growth rate
In large antral follicles, mitotic index was higher (P < 0.01) for fed animals than underfed and refed animals (Table 3). Based on mitotic index data, growth rate (µm/d) was estimated dividing the large antral follicles in subgroups according follicular diameter, i.e. diameters of 350, 600 and 850 µm (Fig. 3). In fed goats, the growth rate was increased according to the increase in the size of large antral follicles. Conversely, underfed and refed group growth rate appeared to be stagnated at 600 µm. In all subclasses considered, growth rate of nourished group was superior (P < 0.01) when compared with the other treatments (Fig. 3).

Figure 3 Mean [± standard error of the mean (SEM)] growth rate (µm/day) of large antral follicular classes in goats at different nutritional levels. a,b,cDifferent lower-case letters indicate statistical difference among nutritional groups within each follicular diameter; P < 0.05. A,B,CDifferent upper-case letters indicate statistical difference among groups based on follicular diameter within each nutritional group; P < 0.05.
Discussion
Nutrition is one of the most important factors influencing animal reproductive performance (Webb et al., Reference Webb, Garnsworthy, Gong and Armstrong2004; Scaramuzzi et al., Reference Scaramuzzi, Campbell, Downing, Kendall and Khalid2006; Zare-Shahneh et al., Reference Zare-Shahneh, Sadeghipanah, Javaheri-Barfourooshi and Emami-mibody2008; Ying et al., Reference Ying, Wang, Wang, Nie and He2011). The relationship between nutrition and reproduction is complex and responses are often variable or inconsistent (Boland et al., Reference Boland, Lonergan and O'Callaghan2001; Somchit-Assavacheep, Reference Somchit-Assavacheep2011; Safari et al., Reference Safari, Kifaro, Mushi, Mtenga and Adnoy2012). Some factors such as timing and duration of flushing, genotypes, and amount and quality of dietary supplements may affect the reproductive performance of small ruminants (Sormunen-Cristian & Jauhiainen, Reference Sormunen-Cristian and Jauhiainen2002; Acero-Camelo et al., Reference Acero-Camelo, Valencia, Rodríguez and Randel2008; Sabra & Hassan, Reference Sabra and Hassan2008). Conversely, O'Callaghan et al. (Reference O'Callaghan, Yaakub, Hyttel, Spicer and Boland2000) demonstrated in ewes that dietary change has an immediate effect on the endocrine system.
This study showed that temporary undernourishment (6 months) affects caprine folliculogenesis, and these effects can be reversible or not by animal refeeding. Macroscopically, underfed animals present an ovarian volume smaller than those receiving an adequate ration, and refeeding these animals after 6 months will not recover the ovarian size. Based on studies with the fetus suffering growth retardation due to different factors including undernutrition, it is stated that, during malnutrition challenges, blood flow will preferentially supply heart and brain, while liver, pancreas, kidneys, and ovaries will be more deprived and develop less (de Bruin et al., Reference De Bruin, Dorland, Bruinse, Spliet and Nikkels1998). These authors suggested that the impaired reproduction would not be reversible. Similar effects were also observed in ewes submitted to undernutrition during gestation (Sen et al., Reference Sen, Sirin and Kuran2013), and in goats submitted to a 50% decrease in energy requirement for 60 days (Kusina et al., Reference Kusina, Chinuwo, Hamudikuwanda, Ndlovu and Muzanenhamo2001) or in goats fasted for 7 weeks with 30% of the required energy (Tanaka et al., Reference Tanaka, Yamaguchi, Kamomae and Kaneda2003).
When evaluating all follicles in general, no nutritional effects are observed on the rates of normal and atretic follicles. However, an evaluation of each follicular class showed that atresia is higher in preantral and small antral follicles from underfed animals and in large antral follicles from fed animals. Importantly, the refeeding program recovered the atresia rates only in preantral follicles. Atresia rates herein are explained by the reduced blood flow and its connection with the vascularization in large antral follicles, which is greater than in small antral follicles or preantral follicles. This phenomenon is similar to ovarian aging, when decreased ovarian stromal blood flow will compromise perifollicular vascularization, contributing to oxidative stress and follicular atresia (Chan et al., Reference Chan, Bernal, Vickers, Gohir and Petrik2015). As the dominant preovulatory follicle contains theca cells with increased activity, other large antral follicles with a lower granulosa and theca cells activity will suffer atresia, which is a physiological phenomenon, and will culminate with the death of the non-dominant antral follicles (Garcia et al., Reference Garcia, Ballesteros, Hernandez-Perez, Rosales and Espinosa1997).
Proliferation of granulosa cells and growth of oocyte indicate if follicular development follow a synchronic pattern. To evaluate this, the marker colchicine was administered. Colchicine is commonly used to study cell division and functionality (Schmidt, Reference Schmidt1942; Hooper, Reference Hooper1961; Bentzen et al., Reference Bentzen, Hansen and Nielsen1999), once it inhibits protein polymerization at the mitotic fuse and allows to stop cell at metaphase stage without interfere with cell division and, consequently, to assess cell mitotic index. Undernourishment did not affect the growth of preantral or small antral follicles, but of large antral ones, which presented the smallest oocytes and lowest granulosa cells proliferation, without affect antrum size. Refeeding goats recovered granulosa cells proliferation, but not their mitotic index and growth rate from those follicles larger than 600 µm in diameter. Furthermore, as observed by the decreased oocyte diameter in follicles from refed animals, the concomitant oocyte growth and granulosa cells proliferation was not recovered after refeeding the undernourished animals. Paula et al. (Reference Paula, Galeati, Teixeira, Lopes Junior and Freitas2005) showed that prolonged underfeeding has a detrimental effect on estrus synchronization response in goats, but this negative effect can be reversible if adequate diet is offered again to these females. However, in the present study, we showed that follicular growth was irreversibly affected, in which mitotic index was negatively affected, even under eCG stimulation. We suggest that the time to recover follicular activity after underfeeding and refeeding process was too short, and ovarian factors other than gonadotrophins will affect the early staged preantral follicles. Moreover, during the underfeed restriction, the release of the metabolic hormone leptin decreases. This hormone plays a role on the growth of ovarian follicles, corpus luteum formation, and production of steroid hormones in the ovary (Zieba et al., Reference Zieba, Amstalden and Williams2005). Therefore, the compromised secretory function of granulosa cells will affect follicular development. Interestingly, Sirotkin (Reference Sirotkin2010) showed that granulosa cells submitted to malnutrition will suffer less atresia, as observed in the large antral follicles, which are surrounded by more blood supply than small antral ones.
In conclusion, goats challenged with prolonged undernutrition, even followed by refeeding with adequate energetic requirement for maintenance, will present a smaller ovarian volume with compromised folliculogenesis. Although atresia will affect more preantral and small antral follicles, the large antral follicles will be smaller, presenting a lower mitotic index, which surely will impair female reproductive efficiency. Besides the physiological effects, undernutrition will also interfere in the success of reproductive technologies, as ovarian follicles from underfed or refed animals will present a lower quality than those from animals fed an adequate diet.
Conflict of interests
The authors declare that there is no conflict of interest that can be perceived as prejudicing the impartiality of the research reported.
Acknowledgements
This research was supported by FUNCAP (n 004/01). V.J.F. Freitas and D. Rondina are senior investigators of CNPq/Brazil.
Author contributions
Conceived and designed the experiments: DR, VJFF. Performed the experiments: DR and VJFF. Analyzed the data: DR, JBB, JJHC and RRS. Wrote the paper: DR, JBB, JJHC and RRS.