INTRODUCTION

Tilapia is the most cultivated group of fishes around the world, with the second highest production, after carp (^{FAO, 2014}). For the commercial production of tilapia it is vital to have a thorough knowledge of the basic aspect of their biological cycle, specifically reproductive aspects. Studies performed over the last decades have provided insight into these aspects and assisted in their integration into cultivation practices (^{Das et al., 2012}). The success of tilapia culture results from the excellent adaptation of this group of fishes to industrial culture conditions but their reproductive precocity requires control measures to prevent the diversion of energy from muscle production to reproduction (^{Das et al., 2012}; ^{El-Greisy & El-Gamal, 2012}). Some of the most frequently used reproductive control measures involve the creation of monosex cultures through sex reversal, interspecific hybridization, and manual sex selection. Although monosex cultures can be either male or female, males are preferred because they reach larger sizes in a shorter amount of time (^{Camacho-Berthely et al., 2000}; ^{Mlalila et al., 2015}). The most commonly used technique for the creation of monosex cultures is known as sex reversal, which consists in the oral administration of the synthetic hormone 17-α-methyltestosterone (17-α-MT) incorporated into feed and supplied during the first days of life after yolk sac assimilation (7 to 10 days after hatching). During this period the gonads are stimulated to develop into testes both in males and genotype females due to the masculinizing action of this synthetic hormone (^{Green & Teichert-Coddington, 2000}; ^{El-Greisy & El-Gamal, 2012}). In the first days of fish life the gonads are in a period of non-differentiation (bipotential) and are susceptible to external influences that can alter the normal process of sexual differentiation (^{Devlin & Nagahama, 2002}). Over the natural process of sexual differentiation, the gonads develop following the chromosomal arrangement and under the action of the hypothalamic-pituitary-gonadal axis, which regulates the process through production of sex steroids, mainly testosterone and estrogen. Testosterone is the hormonal inducer of testicular development and estrogen induces development of the ovaries (^{Bhandari et al., 2006}; ^{Guiguen et al., 2010}). Although testosterone has a masculinizing action and estrogen feminizing, estrogen is synthesized from testosterone through the action of aromatase (^{Devlin & Nagahama, 2002}).

The most recommended treatment for sex reversal is oral administration of powdered fish feed (40 to 50% of total protein) supplemented with 30 to 60 mg of 17- α-MT kg^-1 of feed during the first 25 to 60 days of life, which does not represent a health risk for the organisms during the cultivation period or for consumers. Over the four weeks of treatment, 0.2 mg of 17-α-MT fish^-1 is utilized and the estimated residual 17-α-MT concentration in the body of treated tilapias does not exceed 1.5-10 μg fish^-1 day^-1. Hormone levels in treated tilapia begin to normalize 5 days after the end of treatment, decreasing to less than 10 ρg of 17-α-MT g^-1 of muscle 6 to 50 days after the end of treatment, therefore product safety is assured after the 120 clearance days suggested by the Food and Drug Administration (FDA) (^{Mlalila et al., 2015}). An average adult male has the ability to remove up to 10 mg of androgen per day. This guarantees the elimination of any residual 17-α-MT that could be found in the muscle after the clearance period recommended by the FDA, ensuring the safety of the product offered on the market (^{Green & Teichert-Coddington, 2000}; ^{El-Greisy & El-Gamal, 2012}; ^{Megbowon & Mojekwu, 2014}; ^{Mlalila et al., 2015}). Despite the effectiveness and safety of this technique, in the last decades’ herbal extracts have been proposed for use as active agents to control reproduction in the cultivation of tilapia, in order to find more economically accessible alternatives for aquaculture development (^{Gabriel et al., 2015}). The use of herbal extracts as sex reversal inducers in fish has a variable efficiency percentage, most often no more than 80%, depending on the plant used, concentration of active ingredients, the species studied, and the developmental phase of the organism during application (^{Ghosal & Chakraborty, 2014}; ^{Ghosal et al., 2016}). This places the alternative at a disadvantage in comparison to the 97% sex reversal of the conventional method (^{El-Greisy & El-Gamal, 2012}). In search for alternatives to maintain reproductive control, studies have been carried out such as those by ^{Jegede (2010)} and ^{Kushwaha (2013)} to evaluate the effect of Hibiscus rosa-sinensis (2.0 g kg^-1 feed) and Aloe vera extract (4.0 g kg^-1 feed), respectively, on the development and gonadal maturation of Oreochromis niloticus. Both studies used mixed reproductive populations with the respective diets. Testicular and ovarian structural degeneration was found in the population fed a diet supplemented with extract of H. rosa-sinensis, and degeneration of sperm and follicular cells was seen in the population fed a diet supplemented with A. vera. Therefore, it is possible to conclude that both extracts can be used for reproductive control of O. niloticus.^{Obaroh & Nezh-Achionye (2013)} evaluated the effect of extracts of Azadirachta indica and Mangifera indica in the feed of O. niloticus, using mixed juvenile populations, fed for 56 days. Those authors found that production of offspring was inhibited in populations administered feed supplemented with these extracts (1.0 and 2.0 g kg^-1, respectively), concluding that A. indica and M. indica can be used to control the prolific production of offspring in tilapia cultures.

The Passiflora genus is a widely distributed group of plants, of which a wide variety of bioactive effects have been reported, mainly anxiolytics, sedatives, and antispasmodics, among others (^{Dhawan et al., 2004}; ^{Ingale & Kasture, 2014}). The most studied species in this respect is P. incarnata, although research into its effects on reproduction is limited. ^{Dhawan & Sharma (2003)} evaluated the effect of the methanolic Passiflora extract on the deterioration of some sexual parameters (fecundity and sperm count) in male rats exposed to Δ9- tetrahydrocannabinol (Δ9-THC), finding that by administering feed supplemented with 10 mg kg^-1 of Δ9-THC + 20 mg kg^-1 P. incarnata methanolic extract for 30 days, fecundity and sperm count were not altered, in contrast with the positive control (Δ9-THC 10 mg kg^-1) where these parameters were below normal values. Furthermore, ^{Bacchi et al. (2013)} evaluated the effect of prenatal exposure to P. incarnata extract in rats, analyzing the sexual behavior of males 22, 35 and 75 days after birth, and found that males exposed to 300 mg kg^-1 (body weight) through maternal pathways showed less sexual competition than the controls. The effect of P. incarnata on reproduction is attributed to the possible anti-aromatase action of some of its compounds, which include: BZF (tri-substituted benzoflavone moiety), vitexin, iso-vitexin and orentine, among others (^{Dhawan & Sharma, 2003}; ^{Dhawan et al., 2004}; ^{Masteikova et al., 2008}). There is no research evaluating the effect of P. incarnata on reproductive parameters of tilapia, so the present study assessed the effect of Passiflora ethanolic extract on the gonadal maturation in tilapia, in order to document any possible antiestrogenic action and discuss its potential use as a reproductive control agent in tilapia cultures.

MATERIALS AND METHODS

The experiment was carried out over 90 days between September and December 2013, in the Aquaculture and Ecotoxicology laboratory of the Faculty of Marine Sciences of the Universidad de Colima, Campus Universitario El Naranjo, in the municipality of Manzanillo, Colima, Mexico. The fish were donated by the Centro Acuícola de Occidente S.A. de C.V., located in the same municipality. The P. incarnata raw extract (PE), was donated by Rosa Elena Dueñas S.A. de C.V. (REDSA) laboratory (density 0.9870-1.0442 g cm³, alcohol density 0°-12° LG, relative pH 5.0-6.36, total solids 2.63-19.35%). One hundred and fifty young and sexually undifferentiated tilapias (0.01 ± 0.003 g average body weight), randomly distributed in each of ten 500 L tanks, from which five groups were formed in duplicate after a 24 h acclimatization period.

Water quality conditions remained within the optimal range according to ^{Camacho-Berthely et al. (2000)} through the application of daily water changes constant aeration and total disposal of solid waste.

The organisms were fed during the first 90 days of life after assimilation of the yolk sac in two distinct feeding regimens. During the first 30 days of the experiment the organisms were given powdered feed (53% protein, 15% fats, 0.33 mm size for treatments NAT, P1, P2, P3, and 52% protein, 14% fats, 0.33 mm size for the treatment MET) at 20% of body weight in 8 daily servings at 1 hour intervals. From day thirty one, the treatment consisted of feed pellets (43% protein, 15% fats, 1.5 mm for each treatment) at a rate of 15% body weight (^{El-Greisy & El-Gamal, 2012}) in 4 servings daily at 2 hour intervals over the remaining 60 days of treatment. At the end of the experimental period, 20 fish were randomly caught in each treatment per replicate, anesthetized with clove oil, weighed and then killed by hypothermia. The gonads were removed from the body cavity, weighed for the Gonadosomatic Index (GSI) calculation using the equation GSI = (gonad weight/total weight) × 100, and preserved in a buffered formalin solution for histological analysis to determine sex and degree of sexual maturity. The gonads were processed according to the paraffin embedment technique where 5 μm thick sections were obtained from the paraffin blocks and stained with hematoxylin-eosin (HE). The gonadal maturation indices were determined based on those reported in ^{Peña-Mendoza et al. (2011)}.

A one-way ANOVA of weight and GSI was carried out. Likewise, a non-parametric (Q-Cochran) analysis of the percentage of males and females in each treatment, as well as the percentage distribution in the different stages of maturation (PDGMS) was performed. Subsequently a post-hoc (Tukey HSD) test was applied.

RESULTS

There was no significant difference (P > 0.05) between the treatments mean weights as well as between the percentages of males found in the treatments, except between MET and NAT (Table 1). There was no significant difference in the GSI between males, whereas the GSI in females of treatments P2 and P3 presented statistically significant differences with respect to females of Nat (Table 2).

Gonad weights varied between 0.02 g to 0.60 g among males and 0.03 g to 0.60 g among females (Fig. 1). Histological evaluation of the gonads revealed that all organisms, including both males and females, were distributed within stages II and III of gonadal maturation (Table 2, Figs. 1, 2).

Figure 1 a) Male gonad belonging to the first group with weights between 0.2 to 0.6 g, b) male gonad belonging to the second group with weights between 0.02 to 0.15 g, c) female gonad belonging to the first group with weights between 0.03 a 0.15 g, d) female gonad belonging to the second group with weights between 0.23 to 0.6 g. 

Figure 2 Tilapia gonads in maturation state II and III. Histological sections: a) testis in stage II (beginning of maturation), where spermatogonia (E), primary (EI) and secondary (E-II) spermatocytes are observed in the spermatic ducts, and sperm formation (Es) begins in some ducts, b) stage III (maturation phase) where the same structures of stage II are observed, in addition to most efferent ducts being full of mature spermatozoa (Es), c) ovary in stage II of development, where large follicles (FG), small follicles (PF), follicles in previtellogenesis (P), early vitellogenesis (VT) and advanced vitellogenesis (VA) are observed, d) stage III, where the pelucida zone (ZP), granand teak cells (T) are observed. Hematoxylin-eosin staining (10x) (40x). 

DISCUSSION

The results on the sex ratio indicates that Passiflora extract does not have a significant effect on the percentage of males (Table 1), which explains why there are no previous studies reporting the use of Passiflora flour or extract as a sex reversal inducer in fish (an exhaustive search was made in the CONRICyT; Science Direct; Google Scholar databases).

No lesions were observed on the gonads under the macroscopic and histological assessments (Figs. 1, 2), nor was there a statistically significant difference in the mean weight attained in the experimental treatments with respect to the controls (Table 1). Therefore, it can be inferred that PE does not have toxic effects on the gonadal cells or anti-nutritional consequences for young tilapia. This supports the findings of ^{Boll et al. (2014)}, who reported no negative effects from P. incarnata in pregnant and lactating rats, and ^{Devaki et al. (2012)} who reported that P. incarnata does not produce toxic effects in rats; however, ^{Boeira et al. (2010)} reported that P. alata produces mild genotoxic damage and inhibits weight gain in mice.

The GSI and PDGMS of the females in P2 and P3 treatments showed statistically significant differences with respect to those obtained from the NAT females. They showed lower GSI values for P2 (1.11 ± 0.88) and P3 (1.04 ± 0.99) compared to NAT (1.71 ± 0.72), a PDGMS of 45% in state of maturation for P2, and 30% in III for P3, while NAT presented a PDGMS of 80% in III. The males presented no significant differences between these same parameters (Table 2). The organisms used in the present study were able to reach their first sexual maturity, and the GSI and the gonadal maturity rate were observed to be directly related, with a lower GSI, corresponding to an earlier stage of maturity. It was also found that organisms with a lower GSI presented macroscopic differences. In general, both males and females presented thin and translucent gonads, while males with a higher GSI presented opaque cream-colored gonads with a sinuous appearance, which could be due to the increase of spermatozoa and the corresponding increase of testicular lumen. Likewise, females with a larger GSI presented orange cream-colored gonads with the appearance of granular sacs, corresponding to the greater presence of follicles in advanced and late vitellogenesis (Figs. 1, 2).

The results show that PE has no effect on gonadal maturation in males, although there was a decrease in the percentage of females that reached stage III of gonadal maturation with respect to the natural response control (NAT). This could indicate that the effect of PE is related to the synthesis or bioactivity of estrogen (^{Dhawan et al., 2004}), possibly an inhibitory effect that impedes gonadal maturation in females due to the fact that there was no similar effect in males. ^{Masteikova et al. (2008)} reported that the ethanolic extract of P. incarnata (50.0 g of crude material 100 mL^-1, percolation method) contains: Chlorogenic acid (284 ± 7.0 μg mL^-1), Hyperoside (178 ± 8 μg mL^-1), Isovitexin (4963 ± 142.0 μg mL^-1), Caffeic acid (175 ± 11.0 μg mL^-1), Quercetin (7 ± 0.5 μg mL^-1), Luteolin (21 ± 4.0 μg mL^-1), Orentin (3299 ± 89.0 μg mL^-1), Rutin (164 ± 9.0 μg mL^-1), Scutelarein (42 ± 3.0 μg mL^-1), Vicenin (6102 ± 45.0 μg mL^-1), and Vitexin (1436 ± 21.0 μg mL^-1). Bioactive effects have been reported among them, such as antioxidant, anti-inflammatory, anticancer and anti-aromatase (^{Olthof et al., 2001}; ^{Gülçin, 2006}; ^{Kim et al., 2011}; ^{Marrassini et al., 2011}; ^{An et al., 2012}; ^{Choo et al., 2012}; ^{Lu et al., 2012}; ^{Lee et al., 2013}; ^{Shu-Yao et al., 2013}). However, in vivo effects (especially anticancer and anti-aromatase), of both flavonoid rich extracts and purified extracts are still debated (^{Saarinen et al., 2001}; ^{Moon et al., 2006}; ^{Genoux et al., 2011}; ^{Lephart, 2015}). The works by ^{Dhawan & Sharma (2002}, ²⁰⁰³); ^{Dhawan (2003)} and review by ^{Dhawan et al. (2004)}, suggest that the compound known as BZF, which is described as a tri-substituted α-naphthoflavone isolated from the methanolic extract of P. incarnata, has a strong anti-aromatase activity by restoring libido decline, sperm count and fertility in male rats exposed to substances such as cannabinoids, alcohol or nicotine. It is thought that the BZF activity mechanism acts through the inhibition of enzyme aromatase, which would prevent the transformation of testosterone to estrogen. This appears to correlate with the results obtained in this study, because aromatase is the key enzyme in ovarian development and maturation in fish, and is involved in the biosynthesis of ovarian estradiol (^{Guiguen et al., 2010}). The biosynthesis of ovarian estradiol is a key step in the process of vitellogenesis, in which the oocyte increases in size and develops its structures (^{Mañanós et al., 2008}), and correlates with the results obtained from the GSI and PDGMS in the females of treatments P2 and P3. Those females presented a lower GSI, more females in previtello-genesis (stage II), and fewer females in late vitellogenesis (stage III) with respect to the females of NAT, which suggests that the PE has a possible antiestrogenic and/or anti-aromatase effect. These results may serve as a basis for future research focused on the use of PE on gonadal maturity development, as well as on reproduction control in tilapia culture.