Advances in in vitro folliculogenesis in domestic ruminants

Abstract The in vitro follicle culture (IVFC) represents an outstanding tool to enhance our understanding of the control of folliculogenesis and to allow the future use of a large number of immature oocytes enclosed in preantral follicles (PFs) in assisted reproductive techniques in humans as well as in others mammalian species including the ruminants. So far, the best results of IVFC were reported from mice with the production of live offspring from primordial follicles cultured in vitro. Live birth has been obtained after the in vitro culture of bovine early antral follicles. However, in other ruminant species, these results have been limited to the production of a variable number of mature oocytes and low percentages of embryos after in vitro culture of goat, buffalo and sheep isolated secondary preantral follicles. The present review presents and discusses the main findings, limitations, and prospects of in vitro folliculogenesis in ruminants focusing on bovine, caprine, and ovine species.


Introduction
Ruminants are distributed worldwide and have social and economic importance for many countries, contributing to circa 76% of the global livestock biomass (FAO, 2010). To keep animal production within expected levels, nutrition and reproduction are the key factors. Regarding reproduction, it is essential to understand the mechanisms involved in the formation of the gametes, as well as their development to improve techniques and outcomes (Thatcher, 2017).
At birth, ruminant ovaries contain thousands of immature oocytes, being the vast majority of them enclosed in preantral follicles, that represent the main ovarian oocyte reserve. Despite this large follicle population, most of them (approximately 99.9% of the follicles) will undergo atresia during folliculogenesis. The regulation of folliculogenesis in the preantral follicle phase is an extremely complex process and involves the interaction among endocrine, paracrine and autocrine factors, as previously revised .
The Assisted Reproductive Techniques (ARTs) are of great importance for both basic and applied research. Basic, or fundamental, research is crucial to understand the physiology of reproduction (Smith et al., 2014). Applied research helps to overcome severe infertility either in male or female, as well as to increase the genetic selection rate of highly producing animals (Tan et al., 2017). Among the ARTs that aim to optimize the use in the future of the large ovarian oocyte reserve it is important to highlight the in vitro follicle culture (IVFC) (Green and Shikanov, 2016;. This technique represents an outstanding tool to enhance our understanding of the control of folliculogenesis and to allow the future use of a large number of immature oocytes enclosed in PFs in ARTs in humans as well as in others mammalian species. Interestingly, the production of live offspring from primordial follicles cultured in vitro has been successfully achieved in mice, and it was first reported in 1989 (Eppig andSchroeder, 1989). However, in ruminants, the results have been limited to the production a low percentage of embryos after in vitro culture of goat, buffalo, and sheep secondary preantral follicles . Therefore, this review presents and discusses the main findings, limitations, and prospects of in vitro folliculogenesis in ruminants focusing on bovine, caprine, and ovine species.

Overview of follicle structure and populations and folliculogenesis regulation in vivo
It is well known that mammalian ovaries contain from thousands to millions of follicles whereby approximately ninety percent of this population is represented by preantral follicles usually classified as primordial, intermediate, primary and secondary follicles. Despite this large follicle population, the vast majority of them will be eliminated by a physiological process called atresia during folliculogenesis. Folliculogenesis is the physiological process of activation, growth, and maturation of the ovarian follicle. The regulation of folliculogenesis involves a complex interaction among endocrine, paracrine and autocrine factors which in turn affects steroidogenesis, angiogenesis, basement membrane turnover, follicular atresia, oocyte growth, and maturation as well as the proliferation and differentiation of follicular cells . In the ovary, the distribution of the regulating factors (ligands and their corresponding receptors) varies among follicular compartments (oocyte, granulosa, and theca cells) and significant changes in gene expression pattern among follicular categories (Yoon et al., 2006) have been reported. The control of folliculogenesis is extremely complex because the regulating factors act by binding to different types of receptors that activate distinct signaling pathways and, sometimes different ligands share the same receptors . DOI: 10.21451/1984  Also, there are complex interactions among cell signaling pathways which eventually control gene expression that determines cell survival or death, quiescence or proliferation. Therefore, follicular maturation or atresia will depend on a delicate balance between stimulatory and inhibitory stimuli. Folliculogenesis during the preantral follicle phase can be divided into three steps: (i) the activation (recruitment) of primordial follicles, i.e., transition from primordial (quiescent follicle) to growing follicles (intermediate and primary follicle); (ii) development of primary and secondary follicles; (iii) transition from preantral to antral follicle. In general, it has been stated that the growth of primordial follicles up to the early antral stage is pituitary independent, being probably controlled by autocrine/paracrine mechanisms and modulated by gonadotrophins (for review see . With the understanding of the factors involved in the early folliculogenesis, it will be possible to optimize the use of the large oocyte ovarian reserve in ARTs in humans as well as in other mammalian species. Among these technologies, it is important to highlight the IVFC, as discussed in the coming sections.

Purpose, applications, and type of culture systems
Taking into account that the vast majority of follicles will be eliminated by atresia in case they remain in the ovary the ultimate goal of IVFC is to rescue preantral follicles from the ovary before they become atretic, and culture them up to maturational stages for further in vitro fertilization and embryo production, functioning as an artificial ovary. This technology has some current and future applications such as: (i) to study the control of early folliculogenesis; (ii) to complement other reproductive technologies (e.g., in vitro embryo production, nuclear transfer, etc); (iii) to create gamete banks from endangered species and breeds; (iv) to preserve the fertility in individuals subjected to cancer treatment; infertility treatment (human), and (v) to aid in studies on reproductive toxicology .
Basically, there are two ways to culture preantral follicles: in the isolated form or enclosed in ovarian tissue. Isolated follicles can be cultured in a two-dimensional system which means the follicle is placed on the surface, such as plastic or on an extracellular matrix for instance collagen gel or follicles can be cultured in a 3D system enclosed in an extracellular matrix such as alginate. In situ cultures, on the other hand, have been performed using ovarian fragments or the whole ovary .

Main endpoints used to evaluate the efficiency of IVFC
The efficiency of IVFC can be evaluated using the following endpoints that are crucial for understanding folliculogenesis regulation: follicular survival (morphology/viability); follicular activation and progression through folliculogenesis; oocyte and follicular growth; hormone production; gene expression for key factors (Ligands/receptors); antrum formation; production of fully grown (> 110µm) meiotically competent oocytes; finally, it is also possible to evaluate the oocyte developmental competence through the production of embryos and live offspring (Fig. 1).  Anim. Reprod., v.16, n.1, p.52-65, Jan./Mar. 2019

Bovine
Although live birth has been produced after the in vitro culture of cumulus-granulosa cell complexes from early antral follicles (0.3 -0.7 mm in diameter) (Hirao et al., 2004;Yamamoto et al., 1999), the bovine species is the one facing more difficulties when it comes to move forward in the field of IVFC. The best achievement so far from PFs is the antrum formation after the in vitro culture of primordial and intermediate follicles (< 40 µm diameter) enclosed in ovarian tissue into secondary follicles (~110 µm in diameter), followed by the isolation and in vitro culture of those in vitro grown secondary follicles (McLaughlin and Telfer, 2010). Araújo et al. (2014b) made a profound review on this subject, however, significant advances have been published in recent years that are worth mentioning. The main results are summarized in Table 1.
Early PFs, i.e., primordial and primary follicles are usually cultured in vitro enclosed in ovarian tissue (in situ). With this system, Jorssen et al. (2014) showed that neutral red (NR) staining can be used to facilitate follicle evaluation during a short-term culture (6 days) without affecting follicle developmental competence regardless the oxygen tension (high: 20% O2; low: 5% O2). Regarding culture media supplements, the addition of either growth and differentiation factor 9 (GDF-9) or basic fibroblast growth factor (bFGF) to a medium containing follicle stimulating hormone (FSH) enhanced the beneficial effect of FSH alone in terms of follicle morphology, activation and growth (follicle diameter from ~ 25 µm on D0 to ~ 90 µm on D22 of culture) (Tang et al., 2012).
In addition, the base medium itself can affect follicle viability and development. In fact, Jimenez et al. (2016) stated that αMEM is more effective maintaining follicle viability and promoting follicle growth than both TCM199 and McCoy media. Conversely, Castro et al. (2014) recommended the use of TCM199 and McCoy media for the culture of fresh and vitrified bovine ovarian tissue, respectively, meaning that the process to which the ovarian tissue is submitted prior culture must also be considered before selecting a base culture medium.
Besides growth factors and hormones, the role of cytokines during folliculogenesis, such as tumor necrosis factor-alpha (TNF-α) and interleukin-1 (IL-1), as well as their distribution within the bovine ovary has been investigated. Thereby, proteins of the TNF-α system members, i.e., TNF-α and its receptors (TNFR1/TNFR2), have been detected in oocytes from all follicular categories, in granulosa cells from the secondary stage onwards, and in theca cells at the antral stage. Nonetheless, the addition of TNF-α has shown to reduce follicle survival after 6 days of culture (Silva et al., 2017a). Moreover, proteins of the IL-1 system, i.e., IL-1β and its receptors (IL-1RI, IL-1RII, and IL-1RA) have been detected in oocytes and granulosa cells from all follicular categories and in theca cells at the antral stage. But unlike TNF-α, the addition of IL-1β favored follicular activation and development after 6 days of culture (Passos et al., 2016). The in situ system has also complemented other biotechnologies such as ovarian tissue xenotransplantation. As a matter of fact, a 24 h culture of bovine ovarian tissue in the presence of VEGF prior to xenotransplantation into mice enhanced follicle survival for up to 2 weeks (Langbeen et al., 2016).
As seen in the in situ system, bovine follicles cultured in vitro in the isolated form are also affected by the base medium composition. Rossetto et al. (2012) obtained greater follicle growth, viability and antrum formation (60%), when using TCM199 medium compared to both α-MEM and McCoy media with the same supplementations and medium replacement regime i.e., half of the medium (75 µl) was refreshed every 4 days. However, when another medium replacement regime was used (addition of 5 µl of fresh culture medium to an initial volume of 50 µl every other day), α-MEM became equivalent to TCM199 in terms of follicular growth and antrum formation (Araújo et al., 2015).
Less developed follicular categories, i.e., primordial and primary follicles have successfully reached the antral stage in vitro after 21 days of culture: primordial follicles (< 40 µm) cultured in a two-step system consisting on 6 days of in situ culture followed by 15 days of isolated 2D culture (Act A was added to isolated follicles) (McLaughlin and Telfer, 2010); and primary follicles (50-70 µm) in an isolated 3D system (collagen matrix) in the presence of FSH, luteinizing hormone (LH), estradiol (E2), epidermal growth factor (EGF) and bFGF (Sun and Li, 2013).
In summary, bovine PFs have been able to remain viable during in vitro culture and to form antrum from the primordial stage. Nevertheless, no oocyte meiotic maturation has been accomplished yet. Therefore, future research should focus on the specific factors involved on the in vitro obtention of metaphase II oocytes in this species. Furthermore, besides the fact that bovine is indicated as model for reproductive toxicology studies, especially during oocyte maturation , not so much has been done with PFs.

Ovine
Among the three main species of domestic ruminants (bovine, ovine and caprine), the ovine species is the one where the IVFC has improved the most in the last years (Table 2), probably because is the most used animal model for humans. Hence, it has been reported the production of a relatively high rate of metaphase II (MII) and a low number of embryos at the morula stage after in vitro fertilization (IVF) or parthenogenetic activation of in vitro cultured isolated secondary follicles (Arunakumari et al., 2010;Barboni et al., 2011;Luz et al., 2013). In this sense, Arunakumari et al. (2010) obtained 68% of MII oocytes from which 25% developed to the 2-cell embryo stage, and 16% reached the morula stage (out of the cleaved embryos) after the IVF of oocytes derived from the isolated PFs (200-400 µm) cultured in medium TCM 199 containing thyroxin (T4), FSH, IGF-I, and GH for 6 days. Barboni et al. (2011) also produced competent oocytes from smaller PFs (170 µm) cultured for 12 days in α-MEM with fetal calf serum (FCS) and FSH, although after IVF only 10% of the resulting embryos reached >16-cell stage. Interestingly, in vitro grown oocytes from 360 µm early antral follicles (AFs) (final follicle diameter at the end of the culture) presented similar methylation pattern and developmental capability than their in vivo grown counterparts (early AFs with the same diameter). However, oocytes from both in vivo and in vitro grown early AFs (360 µm) presented low competence compared to oocytes from in vivo grown AFs (6 mm) (Barboni et al., 2011).
Aiming to optimize the current IVFC systems and to understand better the process of folliculogenesis in vitro, several substances and/or culture systems have been tested mainly on isolated secondary follicles (> 200 µm). Thus, it has been shown that rutin alone could potentially replace the combination of the three commonly used antioxidants in culture medium (transferrin, selenium and ascorbic acid), and consequently simplify its composition (Lins et al., 2017). Also, leukemia inhibitory factor (LIF) promoted the rupture of the basement membrane without affecting oocyte maturation or embryo development since 8-cell parthenotes were produced (Luz et al., 2012). The association of LIF and kit ligand (KL), on the other hand, stimulated oocyte meiotic resumption and even a morula was produced after IVF. Nevertheless, this association did not improve the results obtained by LIF alone in terms of oocyte maturation and embryo production (Luz et al., 2013). Likewise, the addition of human leptin to the culture medium increased follicular daily growth but did not affect oocyte maturation (Kamalamma et al., 2016).
Despite all these advances on IVFC of secondary follicles, oocyte maturation and embryo production rates are still far below the results obtained from follicles entirely grown in vivo. On this regard, several studies with this follicular category have shown that IVFC (isolated follicles in a 2D system) negatively affect the expression pattern and/or level of genes related to oocyte survival and development such as P450 aromatase (Lakshminarayana et al., 2014); B-cell leukemia/lymphoma -2 (Bcl2) and Bcl2-associated X protein (Bax) (Praveen Chakravarthi et al., 2015); connexins 32 and 43 (CX32 and CX43) (Chakravarthi et al., 2016a); cyclin B1 (CCNB1) and cyclin D1 (CND1) (Chakravarthi et al., 2016b); and GDF-9 and BMP-15 (Kona et al., 2016).
Early PFs (primordial and primary follicles) are usually cultured in situ (Bertoldo et al., 2014), although they have been cultured in the isolated form as well within a 3D matrix (alginate) (Sadeghnia et al., 2016). These follicular categories have provided some knowledge about what factors mediate follicle activation. Nowadays, increasing evidence suggests that some of them pertain to transforming growth factor β (TGF-β) superfamily (Knight and Glister, 2006), which includes the bone morphogenetic proteins (BMPs). Nonetheless, the addition of BMP4 during the culture of ovarian cortex pieces did not affect follicle activation but enhanced follicle survival and growth (Bertoldo et al., 2014). Also, the stiffness of the environment surrounding the follicles seems to affect their activation and further development. The encapsulation of ovarian cortex in 0.5 or 1% alginate was detrimental for follicle development, while the encapsulation of isolated primordial follicles in 2% alginate potentiated their growth (Sadeghnia et al., 2016).
As it was introduced above, one of the multiple potential applications of IVFC is to serve as an in vitro model for toxicology assays. It is noteworthy that in this species, this technique has been used already for this purpose. Thus, it was determined the minimum concentration of some metabolic stressors that impaired preantral follicle function: 300 µM ammonia, 8 mM urea, 210 µM non-esterified fatty acids (NEFA) (30 µM stearic acid + 60 µM palmitic acid + 120 µM oleic acid), and 0.75 µM β-hydroxybutyric acid (BHB) (Nandi et al., 2017). Moreover, it was shown that aqueous extracts of the plant Justicia insularis were able to maintain follicle morphology and to stimulate primordial follicle activation during the in vitro culture of PFs enclosed in ovarian tissue for 7 days (Mbemya et al., 2017). Early PFs (primordial and primary follicles) are usually cultured in situ (Bertoldo et al., 2014), although they have been cultured in the isolated form as well within a 3D matrix (alginate) (Sadeghnia et al., 2016). These follicular categories have provided some knowledge about what factors mediate follicle activation. Nowadays, increasing evidence suggests that some of them pertain to transforming growth factor β (TGF-β) superfamily (Knight and Glister, 2006), which include the bone morphogenetic proteins (BMPs). Nonetheless, the addition of BMP4 during the culture of ovarian cortex pieces did not affect follicle activation but enhanced follicle survival and growth (Bertoldo et al., 2014). Also, the stiffness of the environment surrounding the follicles seems to affect their activation and further development. The encapsulation of ovarian cortex in 0.5 or 1% alginate was detrimental for follicle development, while the encapsulation of isolated primordial follicles in 2% alginate potentiated their growth (Sadeghnia et al., 2016).

Caprine
In comparison with the other two species reviewed in the present article, the caprine has been by far the most studied species regarding folliculogenesis in vitro in the last years (Table 3). A few 2-to 16-cell embryos have been produced from secondary follicles cultured in vitro in the isolated form (Saraiva et al., 2010;Silva et al., 2014), and even one has reached the morula stage after IVF (Magalhães et al., 2011). Despite that, oocyte maturation and embryo production rates are still low when compared to oocytes grown in vivo. In this regard, wishing to understand and improve conditions for successful IVFC in goats, different substances and/or culture systems have been tested on different follicular categories.
Insulin and FSH are present in almost every culture medium for IVFC, although their concentration when combined is still focus of discussion (Dipaz-Berrocal et al., 2017;Paes et al., 2018). It has been suggested that 10 ng/ml insulin, a lower concentration than that which comes in the ITS composition, could be more efficient in promoting meiotic resumption in the presence sequential FSH . However, some authors described that in the presence of either GH or VEGF, a high insulin concentration (10 µg/ml) combined with fixed 100 ng/ml FSH instead of sequential FSH can improve oocyte developmental competence (Ferreira et al., 2016;Silva et al., 2017b); and also stimulate antrum formation in the presence of phytohemagglutinin (PHA) (Cunha et al., 2013). Conversely, Ferreira et al. (2018) showed no positive effect of the association of high insulin and fixed FSH. This fact might be due to the source of FSH since Ferreira et al. (2018) used recombinant human FSH while most studies used recombinant bovine FSH. Conversely, fixed 10 mIU/ml human FSH improved oocyte meiotic resumption when compared to sequential bovine FSH . Even the base medium itself can influence the follicular response to FSH and insulin. Hence, Amburana cearensis (Amb) ethanolic extract (0.2 mg/ml) base medium promoted higher follicle daily growth rate in the presence of sequential FSH and low insulin concentration than α-minimum essential medium (α-MEM) (Gouveia et al., 2016).
It is most likely that the effect of any supplement may also depend on follicular category since it has been shown that PFs and early antral follicles (EAFs) behave differently under the same culture conditions (Cadenas et al., 2017). Thus, human FSH increased follicle and oocyte diameters of EAFs (~350 µm) but did not affect PFs (Ferreira et al., 2018). Likewise, unlike PFs, EAFs have shown the greatest MII rate described so far from in vitro grown oocytes (46.2% calculated out of the total number of cultured follicles) in response to GH added to a medium with low insulin and no FSH . Also, Cadenas et al. (2018) were able to identify some non-invasive signs for the efficiency of IVFC for EAFs: follicle daily growth ≥ 6.1µm, follicle diameter ≥ 600.1µm, and oocyte diameter ≥ 120.1µm.
Besides culture media composition, many other factors have shown to affect the development of isolated early stage follicles such as the reproductive age of the ovary donor (prepubertal vs. adult), culture period, base media, and culture system (2D vs. 3D). Hence, PFs from prepubertal goats have reached the antral stage, but contrary to PFs from adult goats, were not able to produce MII oocytes after 18 days of IVFC (Amin et al., 2013;Silva et al., 2014) regardless the culture system, i.e., 2D vs 0.5% alginate (3D) . However, Brito et al. (2014) related a positive effect of a lower matrix stiffness (0.25% alginate) on follicle growth and oocyte meiotic resumption when compared to 0.5% alginate and 2D system. Furthermore, the coculture of 5 PFs per alginate bead stimulated follicle growth, and also a new matrix composed by 12.5 mg/ml fibrinogen and 0.125% alginate (fibrinalginate) improved oocyte meiotic resumption when compared to 0.25% alginate .
The suitable culture period for IVFC in the isolated form is still a matter of debate. Even though 18 days is the most commonly used for large secondary follicles (Ferreira et al., 2016;Silva et al., 2017b), it seems that this follicular category may benefit from an extended culture period (30 to 36 days) (Pessoa et al., 2014), while 18 days has been described as the most suitable culture period for EAFs .
Primordial and primary follicles have been usually cultured in situ. Within this culture system, several authors have reported the stimulation of follicle activation, and follicle and oocyte growth after a long-term culture (16 days) in the presence of: FSH during the first half (D0-D8) followed by either GH (Magalhães-Padilha et al., 2012a) or fibroblast growth factor-10 (FGF-10) during the second half (D8-D16) of the culture period (Almeida et al., 2015); KL during the first half and FSH during the second half of the culture ; and also, FSH and IGF-I throughout the entire culture period, which turned up in increasing the percentage of secondary follicles (28%) . Other substances added for a shorter culture time (6 to 7 days) have also shown to exert a positive effect on follicle activation, survival and growth such as: the interaction between melatonin and FSH (Rocha et al., 2013); EGF (Lopes et al., 2015); KL (Faustino et al., 2013); and Concavalin A (Con A) (Portela et al., 2014). On the other hand, Keratinocyte growth factor-1 (KGF-1), also known as FGF-7, did not have a positive impact on early folliculogenesis in vitro (Faustino et al., 2013). Table 3. Chronological advances in in vitro culture of caprine preantral follicles*. Secondary -Isolated 2D (18 days) 10 ng/ml or 10 µg/ml insulin alone or associated to either fixed 100 ng/ml FSH or sequential FSH -10 ng/ml insulin extrusion. 10 µg/ml insulin  follicle growth. 100 ng/ml FSH oocyte meiotic resumption. 10 µg/ml insulin +100 ng/ml FSH  mean oocyte diameter Brito et al., 2016 -α-MEM (3 mg/ml BSA, 10 µg/ml INS, 5.5 µg/ml TRAN, 5 ng/ml SEL, 2 mM GLUT, 2 mM HYPO, 50 µg/ml AA, Sequential FSH)
Despite the undeniable progress obtained in goats, oocyte maturation and embryo production rates are still very low compared to oocytes originating from follicles grown in vivo, which must serve for encouraging further research on this topic. Overall, the data generated seem to point out the need to develop future dynamic and customized culture media for IVFC as the differences among follicles regarding their growth rates, and between follicular categories have shown to affect oocyte maturation in vitro.

Future of IVFC: possible new strategies to overcome the current limitations
To further advance in efficient IVFC methods, some limitations and lack of information need to be overcome. For instance, the in vitro production of a fertilizable oocyte from a primordial follicle requires a long-term culture, which may affect oocyte quality and, consequently embryo prodution. Few studies are reported in animal species other than ruminants, and should be consider when improving culture techniques. It was indicated before that IVC of ovine preantral follicles yield oocytes with normal nuclear-epigenetic maturation (Barboni et al., 2011). However, this study was performed with secondary follicles and not starting from primordial ones, where it was shown in mice that deficiencies at transcriptional and epigenetic levels can occur (Wang et al., 2017). Importantly, IVFC of murine follicles is much shorter than that for large mammals, which can increase the urgence of studies at methylation level. It was demonstrated that apoptosis in murine primary oocytes is mediated by retrotransposon activity (Malki et al., 2014), and the suppresion of this activity is determined by DNA epigenetic modification (Findlay et al., 2015). In other words, methods to measure epigenetic risks, as well as to avoid them during IVFC are still needed. Besides this, preantral follicles are commonly cryopreserved for further in vitro culture. It is well known that the stress caused during exposure to cryoprotectants and the cooling process itself affect important organelles like the endoplasmic reticulum (ER) requiring the possible culture medium enrichment with antioxidants . Studying the ultrastructure of caprine preantral follicles, it was observed that atretic preantral follicles usually presented damaged ER (Silva et al., 2000).

Final consideration
The control of the survival, activation and development of ruminant follicles in vitro is hugely complex and involves multiple interactions among extra and intraovarian factors and can be influenced by the type of base culture media, medium replacement regime, type of culture system (2D vs 3D), culture duration, ovarian source (pre-pubertal vs adult), extracellular matrix components, follicular categories (preantral vs early follicles). Unfortunately, these factors do not act on an isolated form, but interact with each other, making the development of IVFC protocols a challenge. Encouraging results have been reported including satisfactory rates of follicle survival, activation, antral formation and the production of fully grown meiotically competent oocytes especially in caprine and ovine species. However, the in vitro embryo production from in vitro grown oocytes is still low. Therefore, improvements in in vitro follicle culture system should be done to improve oocyte quality (oocyte developmental competence) for further production of viable offspring. This fact will allow the future use of a large number of immature oocytes enclosed in PFs in assisted reproductive technologies in humans as well as in others mammalian species.