Morphological and immunohistochemical comparison of the pituitary gland between a tropical Paracheirodon axelrodi and a subtropical Aphyocharax anisitsi characids (Characiformes: Characidae)

Cardinal tetra Paracheirodon axelrodi and bloodfin tetra Aphyocharax anisitsi are two species of characids with high trade value as ornamental fish in South America. Although both species inhabit middle water layers, cardinal neon exhibits a tropical distribution and bloodfin tetra a subtropical one. Generally, these species are difficult to grow, so it becomes essential to know some key components of the neuroendocrine system to achieve their reproduction in captivity. Considering the importance of deepening the knowledge of the reproductive physiology through functional morphology, for the first time in this work we performed an anatomical, morphological and immunohistochemical analysis of the pituitary gland of these two species. In both species, a leptobasic type pituitary is found in the ventral zone of the hypothalamus and it is characterized by a neurohypophysis which has a well-developed pituitary stalk and a globular adenohypophysis. The pituitary components, characterized by histochemistry and immunohistochemistry, shows a distribution pattern of cells types similar to other teleost species, with only slight differences in the distribution of βFSH and βLH for P. axelrodi . Peces ornamentales.


INTRODUCTION
The understanding of the reproductive function, especially focused on the brainpituitary-gonadal axis, is essential to acquire knowledge about the reproductive biology of different species and fundamental for the support of commercial aquaculture. In captivity, reproduction can be modulated by manipulating environmental parameters, such as water temperature, conductivity and pH, among others (Fiszbein et al., 2010;Mylonas et al., 2010). However, the biological aspects of several teleost species are not well-known, and thus it results very difficult to simulate the environmental conditions required for their reproduction in captivity.
The global trade in ornamental fish is dominated (90%) by freshwater fishes, most of which are sourced from breeding facilities located in developing countries, typically in Asia or South America. However, some fish are still obtained from natural (wild) sources and the exact percentage of wild-caught fish is difficult to quantify given a lack of reliable data (Evers et al., 2019). Generally, these species, collected from their natural environment, are difficult to grow, so it is important to know the morphological and physiological key components of neuroendocrine systems to better regulate the reproduction of these species.
In some commercial species, for example, reproduction can be enhanced by several strategies which can modify the synthesis and release of pituitary hormones (Mylonas et al., 2010;Zohar et al., 2010). In this sense, identifying and characterizing cell types that produce pituitary hormones will enhance our knowledge on fish reproductive physiology and will constitute a solid basis for future reproductive induction with hormonal treatments. Authors said that heterologous antisera can be used in these kind of studies, since their specificity and absence of cross reactivity can be demonstrated (Laiz-Carrión et al., 2003;Pandolfi et al., 2005;Honji et al., 2015).
The pituitary gland is an important endocrine organ that presents two distinctive regions: the neurohypophysis, consisting of neurosecretory terminals principally from the hypothalamus, and the adenohypophysis, constituted mainly by secretory cells. In teleost fish, these endocrine cells are arranged in groups that produce different hormones and are located at specific regions of the adenohypophysis (Van Oordt, Peute, 1983;Laiz-Carrión et al., 2003;Pandolfi et al., 2005).
Cardinal tetra Paracheirodon axelrodi (Schultz, 1956) is a characid species endemic from South America, which inhabits the Amazon, Orinoco and Rio Negro basins (tropical habitat). It inhabits middle water layers with temperatures between 23°C and 27°C, pH ranging from 5.5 to 7, hardness oscillating between 5 dH and 12 dH (Anjos, Anjos, 2006) and it is an omnivore species (Walker, 2004;Marshall et al., 2008). Female spawning occurs in shaded territories and each spawning event can range from 154 to 562 oocytes. It is worth mentioning that this species has an asynchronous oogenesis, since more than two oocyte stages can be present simultaneously (Anjos, Anjos, 2006;Brito, Bazzoli, 2009).
Bloodfin tetra Aphyocharax anisitsi Eigenmann, Kennedy, 1903 is a characid species that inhabits the basin of La Plata's river in South America (subtropical habitat). In this habitat, the water temperature range is between 18°C and 28°C, pH ranges from 5.4 to 7.9 and hardness of 30 dH (Burgess, 2004;Casciotta et al., 2005). The reproductive biology and the development of sexual dimorphic structures in this species have been already described (Gonçalves et al., 2005). Other works on its biodiversity and distribution in Argentina report that this is a vulnerable species in the Mesopotamia (López et al., 2005), but it is not a threatened species in La Plata basin (Zayas, Cordiviola, 2007), suggesting that their conservation status may differ among different habitats.
Taking into account the importance of deepening the knowledge of structures related to reproduction, the brain atlas (Obando-Bulla et al., 2013;Rincón et al., 2016aRincón et al., , 2017 and also a histological and anatomical comparison of the pineal complex in P. axelrodi and A. anisitsi (Rincón et al., 2016b) have been recently published. In this context, considering that we intend to continue assessing the brain-pituitary-gonadal axis, the present study assesses a morphological, anatomical and immunohistochemical characterization of the pituitary gland of both species for the first time.

Animals.
Adults of Paracheirodon axelrodi and Aphyocharax anisitsi were obtained from commercial aquaria. In both cases, animals were housed in community aquaria mimicking their natural conditions: 25°C for both species, pH 6.0 to 7.0 for A. anisitsi and pH 5.0 to 6.0 for P. axelrodi. Fish were fed twice a day with commercial fish pellets (Tetra ®) and acclimatized for at least one month before being used. All procedures described in the following sections were conducted in accordance with international standards, Guide for Care and Use of Laboratory Animals (NRC, 2011) on animal welfare as well as being compliant with local regulations (CICUAL, Comision Institucional para el Cuidado y Uso de Animales de Laboratorio).
Eight adult specimens, 5 females and 3 males of P. axelrodi, of standard length (L S ) of 2.86 ± 0.10 cm, total length (L T ) of 3.36 ± 0.38 cm and weight of 0.42 ± 0.17 g were used. Two females and two males were used for classical histological procedures and three females and one male for immunohistochemical assays. Four females and four adult males of A. anisitsi of L S 3.08 ± 0.04 cm, L T of 3.71 ± 0.06 cm and weight of 0.57 ± 0.03 g were processed. Two females and two males were used for histological studies and two females and two males for immunohistochemical procedures. Voucher specimens are deposited at the Ichthyological collection of the Museo Argentino de Ciencias Naturales "Bernardino Rivadavia" (MACN-Ict 12710), Buenos Aires, Argentina.
Histological analysis of the pituitary gland. Animals were anesthetized with benzocaine (0.1 g/l) and killed by decapitation. To preserve both pituitary´s histology and its anatomical relationship with the brain, whole heads were dissected and fixed in Bouin's solution for 24h at 4°C in the dark. Afterwards, heads were dehydrated through an ascending series of ethanol, clarified with xylene and embedded in Paraplast ® (Sigma). Embedded heads were sagitally and transversally sectioned at 7 µm and mounted on gelatine-coated glass slides. Then, these sections were deparaffinized in xylene, rehydrated through a descending ethanol series and stained with haematoxylin eosin (H-E), Masson trichrome (MT) and periodic acid-Schiff (PAS). Finally, slides were mounted in DPX (Sigma), examined using a Nikon Microphot FX microscope and digitally photographed (Coolpix 4500, Nikon; Japan).
Immunohistochemical analysis of the pituitary gland. For this study, the specimens were processed as described for the histological analysis. Briefly, heads (with pituitaries) were sagittal sectioned at 7 µm and mounted on different slides. They were deparaffinized in xylene, and rehydrated through a graded ethanol series to phosphatebuffered saline (PBS, pH 7.4). Sections were treated with 3% H 2 O 2 for ten minutes to saturate endogenous peroxidase activity, after which they were incubated with 5% nonfat dry milk to block unspecific binding sites. Then, slides were incubated overnight at 4°C, in a moist chamber, with different specific primary antisera (antiserum and dilutions used are detailed on Tab. 1). After washing with PBS, sections were incubated for 45 min with a biotinylated anti-rabbit IgG diluted 1:600 (Dako), at room temperature in a moist chamber. Amplification of the signal was achieved by incubating the sections with peroxidase-conjugated streptavidin (STRP-HRP) (Dako) diluted 1:600, at room temperature in a moist chamber, and visualized with 0.1% 3.3'-diaminobenzidine with  0.03% H 2 O 2 . Sections were slightly counterstained with haematoxylin, mounted in DPX and examined. For β-FSH and β-LH antigen immunoreactivity was recovered. Before the block of endogenous peroxidase activity, samples were heated for 10 min at 90°C with citrate buffer (10mM, pH=6) for epitope unmasking, cooled at room temperature (RT) and washed with PBS. Finally, the slides continued with the immunohistochemical method described above. Negative controls (data not shown) were achieved by omission of the primary antisera and preabsorption tests performed on other species showed antisera specificity (see Pandolfi et al., 2006;Honji et al., 2013). Furthermore, the specificity of the antisera could also be inferred from the localization of the immunostained cells, which agree with previously reports in other neotropical fish species (Borella et al., 2009;Honji et al., 2013;Nóbrega et al., 2017).

RESULTS
Anatomical description of the pituitary gland. The pituitary glands of both P. axelrodi and A. anisitsi appears attached to the ventral region of the hypothalamus (H) by a thick infundibular stalk and consists of two components: the neurohypophysis (NH) and the adenohypophysis (ADH). In both species, the NH consists of a welldeveloped infundibular stem, and the ADH is globular with its three areas located in an anterior-posterior position. According to these characteristics, the pituitary gland of both species is of the leptobasic type ( Fig. 1C-E). The ADH is subdivided in three different areas: a rostral part of the gland, rostral pars distalis (RPD), a central part, the proximal pars distalis (PPD), and a posterior part, pars intermedia (PI). Additionally, each of these areas are invaded by the NH, in a greater proportion in the PPD and PI than RPD, and present different cell types with distinct staining characteristics (Figs. 1, 2). Most of the individuals used for this study were mature males and females. Ovaries had vitellogenic oocytes in greater proportion than earlier oogenetic stages, while testis presented mature spermatozoa in their lumen. No differences were observed in the distribution of pituitary cells with respect to the different phases of the reproductive cycle of individuals (data not shown). Ir-ACTH and ir-MSH cells were recognized by the antibody against ACTH, since it recognizes a common region shared by both peptides, which derive from the same precursor molecule, proopiomelanocortin (POMC). These MSH cells show an elongated shape and are close to the neurohypophyseal fibers interspersed with ir-SL cells. In the negative control, no ir-SL or ir-ACTH cells in the negative control were detected.

Histochemistry and immunohistochemistry. Rostral pars distalis (RPD
Finally, Fig. 7 shows a camera lucida drawings of sagittal sections of the pituitary gland, obtained by histological and immunohistochemical analysis, showing the distribution of adenohypophyseal cells of P. axelrodi (A) and A. anisitsi (B).

DISCUSSION
The pituitary of P. axelrodi and A. anisitsi, is in the ventral zone of the hypothalamus, and its structural and cellular characteristics coincide with that reported in other teleost species (Pandolfi et al., 2001;Borella et al., 2009;Kawauchi et al., 2009;Zohar et al., 2010;Honji et al., 2013;Norris, Carr, 2013). According to the scientific literature, the pituitary of both species is the leptobasic type, which is characterized by a neurohypophysis which has a well-developed pituitary stalk, and a globular adenohypophysis (Van Oordt, Peute, 1983).
In both species, ir-ACTH and ir-PRL cells were observed disposed in compact groups in the rostral pars distalis (RPD). In P. axelrodi, particularly, ir-βFSH and ir-βLH cells were also found in the RPD. Ir-PRL cells have previously been demonstrated in the RPD of many other teleosts (Batten, 1986;Parhar et al., 1998;Segura-Noguera et al., 2000;Rodríguez-Gómez et al., 2001;Sánchez Cala et al., 2003), where they were arranged as a compact mass, like in P. axelrodi and A. anisitsi. This is different than described for A. gigas when the ir-PRL cells are scarce, arranged in thin strands, and weakly immunostained with the employed antiserum (Borella et al., 2009). PRL is important for the survival of freshwater teleost fishes, since it causes an increase in the concentration of ions and maintains sodium levels in plasma (Pandolfi et al., 2001;Kawauchi et al., 2009;Zohar et al., 2010;Honji et al., 2013;Norris, Carr, 2013). Several studies have shown that PRL also influences fish reproductive cycling, as PRL may be involved in spermatogenesis, vitellogenesis and ovulation, beyond its important osmoregulatory function in fish (Edery et al., 1984;Sandra et al., 2000;Cavaco et al., 2003;Whittington, Wilson, 2013). In teleosts, ir-ACTH cells are typically distributed in palisade at the RPD, between PRL cells and branches of neurohypophysial tissue (Schreibman et al., 1973;Van Oordt, Peute, 1983;Borella et al., 2009), as seen in P. axelrodi and A. anisitsi. ACTH stimulates the interrenal gland to produce cortisol that is related with several physiological processes: stress response, metabolism, osmoregulation, among others (Pandolfi et al., 2001;Laiz-Carrión et al., 2003;Borella et al., 2009;Kawauchi et al., 2009;Honji et al., 2013).
Within the proximal pars distalis (PPD), ir-GH, ir-βFSH and ir-βLH cells were found dorsoventrally distributed, for both species. Ir-βFSH and ir-βLH cells abundant in an area close to the fibers of the neurohypophysis with a round or oval shape, these characteristics agree with the typical described for several teleosts (Van Oordt, Peute, 1983;Agulleiro et al., 2006). βFSH cells were characterized with both heterologous antibodies, anti-chum salmon and anti-mummichog βFSH, obtaining a weaker immunostaining with the first one. The processes of ovulation and secretion of gonadal steroids, depend mostly βLH, although βFSH in fish has also a steroidogenic function (Levavi-Sivan et al., 2010). Likewise, βFSH stimulates the early follicular development and the preparation of the gonads for the posterior actions of βLH (Pandolfi et al., 2006;Borella et al., 2009;Zohar et al., 2010;Norris, Carr, 2013).
Even if in A. anisitsi no immunostaining was observed in RPD, in P. axelrodi ir-βFSH and ir-βLH cells are observed in the dorsal part of this region, as already reported in A. gigas (Borella et al., 2009). This localization in the RPD constitutes a novel result, since in other teleost species such as C. dimerus and Bagrus bayad (Fabricius, 1775) gonadotropins are found both in the dorsal and ventral portions of PPD, as well as around the PI (Pandolfi et al., 2006;Mousa et al., 2012). Moreover, in P. axelrodi, ir-βFSH and ir-βLH fibers are also observed in the NH. This coincides with the reported distribution in C. dimerus, C. dimerus suggested that, even if the precise role of brain-derived FSH and LH is not clear, the abundance of immunopositive fibers in the NH could imply a neuromodulatory function together with a possible role in the control of pituitary cells activity (Pandolfi et al., 2006). More detailed studies are necessary to confirm the pattern of distribution of βFSH and ir-βLH cells in P. axelrodi.
On the other hand, previous studies demonstrated that antiserum raised against human-β-TSH selectively cross-react with the TSH-producing cells of several teleost species (García-Hernández et al., 1996;Segura-Noguera et al., 2000), although a weak immunoreactivity to this antiserum was observed in some teleosts too (Nozaki et al., 1990;Yan, Thomas, 1991). In this work, we could not detect ir-TSH cells with the antibody used (data not shown). Considering that TSH has been shown in other teleost species before, the lack of immunolabelling in our experiments is probably due to the specific antibody chosen for this study and it will be interesting to describe the localization of TSH in P. axelrodi and A. anisitsi using other antibodies in future studies.
Finally, ir-GH cells were larger than the former with oval shape, some of them with vacuoles in the cytoplasm, this feature agree with other teleosts (Rendon et al., 1997;Vissio et al., 1997;Parhar et al., 1998;Segura-Noguera et al., 2000;Rodríguez-Gómez et al., 2001;Sánchez Cala et al., 2003). These cells were frequently found in teleosts surrounding the neurohypophysial processes of the PPD (Van Oordt, Peute, 1983), as seen in P. axelrodi and A. anisitsi. For example, this arrangement was not observed in A. gigas pituitary, as there are no NH branches penetrating the pars distalis (Borella et al., 2009). GH has the effect of increasing cellular metabolic rates, and it also induces the liver cells to produce somatomedins. These, in turn stimulate the mitotic indexes of chondrocytes of the epiphyseal plate and therefore promotes the lengthening of the bones and, consequently, growth (Pandolfi et al., 2001;Borella et al., 2009;Honji et al., 2015).
In the pars intermedia (PI), ir-MSH and ir-SL cells were found surrounding the neurohypophyseal branches. In the pars intermedia, the identified ADH cells release somatolactin (SL) and melanocyte stimulating hormone (MSH). On both species, the ir-SL cells present in the PI of pituitary gland were distributed mainly in the peripheral areas of this region, forming cords or clusters surrounding the neurohypophysial tissue and blood capillaries, apparently intermingled with the MSH cells (Batten, 1986;Cambré et al., 1986;Quesada et al., 1988;Vissio et al., 1997;Rodríguez-Gómez et al., 2001). However, in P. axelrodi specifically, ir-SL cells are observed in the PPD too. Similar to results showed for S. hilarii (Honji et al., 2013), where is observed a crossreaction of antisera against chum salmon SL with some cells in PPD region. According to Margolis-Kazan et al. (1981) and Batten (1986) the SL and gonadotropins (βFSH and βLH) hormone can contain antigenically similarity, which might be difficult to remove during biochemically purification.
The SL function in teleost fishes is not yet fully established, since mode of action seems to vary between species. For example, in salmonids, SL has been reported to be involved in sexual maturity and smoltification (Rand-Weaver et al., 1992;Bhandari et al., 2003;Borella et al., 2009;Onuma et al., 2010). In other species, however, it has been associated with in the adaptation to background and photoperiod changes, in metabolism, stress response and reproductive physiology (Zhu, Thomas, 1997;Mousa, Mousa, 1999;Rand-Weaver et al., 1992;Vissio et al., 2002;Honji et al., 2013). On the other hand, α-MSH participates in the adaptation to skin color and in the stimulation of ACTH release during stress (Lamers et al., 1992;Laiz-Carrión et al., 2003;Borella et al., 2009).
There are three reproductive dysfunctions in teleost fish are frequently observed in many species when breeding in captivity, according to Zohar, Mylonas (2001) andMylonas et al. (2010). First, female vitellogenesis and male spermatogenesis fail completely when broodstocks are maintained in captivity. Second, cultured females fail to spawn at the end of the reproductive cycle (oocytes undergo normal vitellogenesis, final maturation, and ovulation; however, ovulated eggs are not released to the water). Third, vitellogenesis appears to progress normally in cultured females, but oocytes fail to reach final maturation, resulting in neither ovulation nor spawning. The second case of dysfunction could be the case of P. axelrodi specifically, since gonadally mature females and males are normally found, however, ovulated eggs or sperm are not released to the water.
Nevertheless, before planning the manipulation of ADH hormone secretion for this species, which will enhance fish reproduction in captivity, it is first necessary to establish the morphophysiological parameters of cell types the pituitary gland and specially that involved in the reproductive process. So, this is the first morphological and immunohistochemical study of the pituitary of P. axelrodi and A. anisitsi. Overall, the pituitary components of P. axelrodi and A. anisitsi presented very similar histological and immunohistochemical characteristics. The results showed a pattern of distribution of pituitary cells similar to that of other teleost species, with only slight differences in the distribution of βFSH and SL for P. axelrodi.