The Institutional Animal Care and Use Committee approved all animal handling procedures (case number CEUA/UFC/255225-718/2018), and the experiment was conducted according to the experimental protocol for the use of live birds from the Brazilian College of Animal Experimentation.

The experiment was conducted in Fortaleza, CE, Brazil (3°43′6″ S, 38°32′36″ O). Overall, 22 male quail (Coturnix coturnix coturnix) were raised until 140 days of age with ad libitum access to feed and water. At 140 days, the birds were weighed and allocated to four treatments and six repetitions in a completely randomized manner: treatment 1 (six animals), treatment 2 (six animals), treatment 3 (five animals), and treatment 4 (five animals). Treatments were formulated by the addition of three different CA levels. In treatment 1 (control group), the diets were not supplemented with CA. In treatments 2, 3, and 4, 0.25%, 0.50%, and 0.75% of calcium anacardate, respectively, were added to the feed over an experimental period of 210 days. Calcium anacardate was extracted from cashew nut liquid through precipitation, according to the methodology adapted from Paramashivappa et al. (2001). Males were raised together with five females in a conventional shed in cages measuring 35 × 25 × 20 cm (length, width, and diameter), equipped with a linear trough feeder, nipple drinker, and egg tray. The experimental diets were formulated to be isonutritive according to the nutritional requirements presented by Silva (2009), and the nutritional and energy composition values of the ingredients were considered according to Rostagno et al. (2017). There was a replacement of the inert according to the inclusion of CA in the proportion of each treatment (Table 1). Calcium from CA was not considered. The animals were subjected to 16 h light per day. After the experimental period, the animals were sacrificed by exsanguination in compliance with the guidelines established by the National Council for Animal Experimentation Control (CONCEA, 2015) and weighed. After weighing, all birds were placed in supine position for posterior incisions in the ventral region for collection of the testes and epididymis. The epididymis was dissected from the testicles before weighing. The left and right testes were weighted and measured (length, diameter, and volume). Testicular volume was calculated using Bissonett's formula, 4/3 πab² (in which a = half of the long axis and b = half of the short axis), and averaged (Chaturvedi and Bhatt, 1990; Chaturvedi et al., 1993). The testicular gonadosomatic index was calculated using the equation (testis weight/final body weight) × 100%. Samples of the testes and epididymis tissue were reserved for proteomic and histological analysis.

The testicles used for proteomics were frozen in liquid nitrogen and lyophilized at −55 °C for 24 h, with a minimum pressure of 60 mTorr. For the extraction of proteins, 10 mg of lyophilized sample was diluted in 500 μL of lysis solution (Triton X-100 0.1%) and stored at 4 °C for 1 h, with homogenization conducted at 15-min intervals. Subsequently, 400 μL of sample buffer was added to the samples, followed by sonication for 1 h. The samples were centrifuged at 12,000 × g for 30 min at 4 °C, and the supernatant was collected for protein quantification using Bradford's method (Bradford, 1976). The proteins extracted from the testicles were separated by the 1D SDS-PAGE method based on the protocol described by Souza et al. (2020), with modifications. After the run, the gels were stained using Coomassie's Brilliant Blue solution for 12 h and then discolored with methanol and acetic acid solution (40%:10%). The gels were subsequently digitized using the ImageScanner III (GE Healthcare, Piscataway, NJ, USA) and saved in TIFF format, followed by analysis using the Quantity One Software^® version 4.6.3 (Bio-Rad, Rockville, MD, USA).

The bands were cut into pieces of approximately 1 mm³, and tryptic digestion was performed according to De Lazari et al. (2019). Tryptic peptides were separated on a BEH300 C18 column (100 μm × 100 mm) using the nanoAcquity^TM system (Waters, USA) and eluted at 600 μL/min with an acetonitrile gradient (5–85%) containing 0.1% formic acid. The liquid chromatography system was connected to a nanoelectrospray mass spectrometry source (SYNAPT HDMS system, Waters, Milford, MA, USA), and mass spectrometry was operated in positive mode at 90 °C and a capillary voltage of 3.5 kV. The instrument was calibrated with fragments of the protonated double ion (phosphoric acid, m/z 686.8461), and the blocking mass used during data-dependent acquisition (DDA), selecting the precursor MS/MS ions double- or triple-charged. The LC-MS/MS procedure was carried out according to the data-dependent acquisition method (DDA), selecting the double- or triple-charged percussion ions MS/MS. Ions were fragmented by collision-induced dissociation, using argon as the collision gas and a ramp collision energy that varied according to the charge state of the selected precursor ion. Data acquisition was carried out in an m/z interval of 300 to 2,100 for the survey MS (1 scan/s) and an m/z interval of 50 to 2,500 for MS/MS. Data were collected using the MassLynx 4.1 software, processed using Protein Lynx Global Server 2.4 (Waters Corp. Milford, MA, USA), and converted to peak list text files for database searching (.pkl). The Mascot platform (Matrix Science, London, United Kingdom, v. 2.6) was then used to search the NCBIprot and SwissProt database, carrying out gene ontology with the aid of the software for researching protein annotations (STRAP v. 1.5.0.0) and using the UniProtKB database to obtain the terms of the genetic ontology (biological process, molecular function, and cellular component).

Protein-protein networks were retrieved from the STRING database (http://string-db.org) version 10.0 (Viana et al., 2018; Snel et al., 2000), which consists of known and predicted protein interactions collected from direct (physical) and indirect (functional) associations. This database quantitatively integrates interaction data from four sources-genomic context, high-throughput experiments, co-expression, and data from previous publications. The network analysis was evaluated only with the proteins expressed differentially in the bands found; those with antioxidant function and those with interactions with statistical significance (PPI enrichment P-value ≤ 0.05) were included in the network analysis. When there were no reports on the European quail or the genus Cortunix, descriptions with the domestic rooster (Gallus gallus) were used.

Fragments of the testicles and epididymis were fixed by immersion in a 10% formaldehyde solution for 24 h. Subsequently, only the testes, being larger, were cut to obtain sizes close to 5 mm. Testicle and epididymis samples were placed in microtubes with fixation solution for technical processing, which consisted of the following steps: washing, dehydration, infiltration, and blocking. Samples were washed three times at 40-min intervals in 50% ethyl alcohol. After washing, the samples were dehydrated in an ethanol series at 50, 70, 90, and 100% ethanol, with 45 min between the different concentrations. The samples were infiltrated with historesin 1:1 ethyl alcohol for 48 h and again with historesin for 24 h. For blocking, the samples were placed in gelatin capsules, historesin and hardener were added and the capsules were exposed to light for 24 h. After historesin polymerization, the blocks were placed on pieces of wood, and cuts were made at 0.6 μm using a microtome (LEICA, Solms, Germany); the obtained slices were placed on slides for microscopy. Staining was performed with 1% toluidine blue solution, and the coverslips were placed with the aid of VERNIZ^® (Acrilex). Images were captured using a Zeiss Primo Star microscope with an Axiocam 150 color capture camera, and the ZEN lite 2.1 software (Pleasanton, CA, USA) was used for image analysis. To collect image measurements, the Image J 1.8.0 software was used.

Statistical analysis was conducted considering a completely randomized design with four diets and six repetitions. Data were initially subjected to the Kolmogorov-Smirnov normality test to confirm normal distribution. When data were not normally distributed, they were log 10-transformed. Data were analyzed via one-way analysis of variance (ANOVA), according to the following general model:

in which Y_ij is the measured dependent variable, μ is the overall mean, α_i is the treatment effect, and ε_ij is the random error.

Comparison among treatment averages was performed using Tukey's test at a significance level of 0.05. All analyses were performed using the R program (version 3.4.4).

There was no significant (P≤0.05) effect of treatment on testicular weight, length, diameter, and volume as well as on the gonadosomatic index and weight of the birds (Table 2).

Using 1-D electrophoresis, a total of 35 bands were observed in the testicular tissue gels of European quail, with molecular weights from 12 to 150 kDa (Figure 1). The intensity of 13 bands differed among treatments (Figure 2). Of the 13 bands, treatment with 0.75% CA was different from the control treatment in all bands. In two bands (29 and 32), with a molecular weight of 23 and 17 kDa, respectively, treatment with 0.75% CA differed from that with 0.25% CA.

The 1-D electrophoresis and mass spectrometry allowed the identification of 152 proteins in the testes of European quail (Table 3), including heat shock cognate 71-kDa protein (HSPA8), stress-70 protein, mitochondrial (HSP9), superoxide dismutase [Cu-Zn] (SOD1), and peroxiredoxin-6 (PRDX6). Gene ontology was employed to collect data related to molecular function, biological processes, and cellular components of proteins found in the testes (Figure 3). In view of these results, the main functions of the proteins identified in the testicular tissue of European quail are binding (49%) and catalytic activity (36%), with only 1% being attributed to antioxidant activity.

Protein interactions for five proteins involved in antioxidative processes and that differed significantly among treatments from each other in this study (P≤0.05) were analyzed. The proteins were superoxide dismutase [Cu-Zn] (SOD1), peroxideroxin-6 (PRDX6), stress-70 protein (HSPA9), 60-kDa heat shock protein (HSPD1), and heat shock cognate 71-kDa protein (HSPA8). Figure 4 illustrates the interaction of these five proteins, and Figure 5 illustrates the interactions of each mentioned protein with other proteins. Peroxiredoxin-6 showed co-expression interaction with proteins that protect the body against oxidative agents, such as catalase and superoxide dismutase [Cu-Zn] SOD1 (Figure 5A). Superoxide dismutase [Cu-Zn] showed a text-mining interaction with other proteins capable of extinguishing radicals produced during oxidative stress, such as catalase and thioredoxin (Figure 5B). The HSPD1 protein was also found in this study. This protein, which is considered an HSP60, has co-expression and text-mining with HSP70, and therefore, both have similar biochemical characteristics (Becker and Craig, 1994) (Figure 5C). The HSPA9 protein showed a text-mining interaction with proteins responsible for protecting other proteins from oxidative stress (60-kDa heat shock protein, mitochondrial, and DnaJ homolog subfamily B member 6; Figure 5D).The HSPA8 protein was also found in this study, with co-expression interproteic interaction with HSP70 and DJ-1 protein (Figure 5E).

The diameter of the seminiferous tubules of quail was not influenced by treatments. However, the height of the seminiferous epithelium and the diameter of the epididymal duct significantly differed among treatments (Table 4). There was no significant effect of treatments on the quantities of testicle cells of quail (Table 5). Through optical microscopy, images of the seminiferous tubule (Figure 6) and the epididymis (Figure 7) of European quail were captured, with germ cells at different stages of development, as well as Sertoli and Leydig cells (Figures 6 and 7) and sperm formed in the epididymis. Hair and non-hair cells in the efferent duct could also be observed.

The reproductive system of birds differs from that of mammals in that it does not have important glands for semen production, such as the bulbourethral gland, the seminal vesicle, and the prostate. In addition, the epididymis of birds does not have a head, body, and tail division such as the epididymis of mammals (Macari and Maiorka, 2017). Animals subjected to a high-temperature environments undergo heat stress, inducing the production of ROS, which in turn induces molecular, physiological, and morphological changes in the reproductive organs, affecting reproduction (Hanafi et al., 2010). Garrigue et al. (2017), when subjecting rats to high temperatures (40 to 42 °C) for 60 days, noticed a reduction in the size of the reproductive organs of these animals. In a study by Meade et al. (2019), heat stress induced damage to the testicular morphology, whereas older animals were more susceptible to this damage than young ones.

In this experiment, testicular morphology was not influenced by the use of different CA levels. Santos et al. (2012) found that, at 60 days of age, testicle weight stabilized at 3.59 g for meat quail and 2.49 g for laying quail. Therefore, it is believed that the animals in this study were not subjected to intense oxidative stress that would result in changes in testicular morphology.

The gonadosomatic index (GSI) was another biometric variable collected in this study. It is essential in the assessment of testicular size and is related to sperm production (Kenagy and Trombulak, 1986). In this study, the GSI did not differ among treatments (Table 2), and the average GSI in all treatments was 2.52%. Orsi et al. (2005) reported similar GSI values for Japanese quail (2.26%) and Italian quail (2.6%), whereas Lanna et al. (2013) found values for Japanese quail close to 3.7%, indicating greater sperm production capacity and storage. Therefore, light and small testicles, with a low GSI, are related to a lower reproductive capacity due to low rates of spermatogenesis and testosterone production (Santos et al., 2012).

The expression of equivalent bands among treatments decreased with increasing CA concentrations in the feed (Figure 2). The animal naturally produces enzymes that eliminate free radicals, which is known as the enzymatic defense system, producing enzymes such as catalase, glutathione peroxidase, and superoxide dismutase to protect the body against oxidative agents such as hydrogen peroxide, superoxide, and nitrite oxide (Ali et al., 2020). In view of this, the decrease in protein expression is believed to be due to the gradual increase in CA level as a result of the exogenous supply of antioxidant enzymes present in the feed, and therefore, the enzyme defense system does not need to produce antioxidant enzymes.

In the bands that were significantly different (Figure 2), important proteins linked to reproduction were found, such as piwi-like protein 1 (PIWL1) and heat shock protein HSP 90-alpha (HS90AA1). Of these, PIWL1 is involved in the development of the embryo and is related to spermatogenesis; its absence can therefore result in damage to germ cells during meiosis (Chen et al., 2013). The HS90AA1 is a heat shock protein (HSP) produced to protect the testis against lipid oxidation when the animal is subjected to oxidative stress, thus protecting spermatogenesis (Ohsako et al., 1995; Biggiogera et al., 1996; Grad et al., 2010).

Peroxiredoxin-6 and superoxide dismutase [Cu-Zn] (SOD1) proteins, which are important for protecting the body against ROS, were also found in this study. Peroxiredoxin-6 belongs to the class of peroxiredoxins (PRDX), which are the main antioxidant proteins of peroxides produced endogenously in eukaryotes (Han et al., 2005). According to this same author, the antioxidant activity of PRDX in chickens is similar to that found in mammals, and they are involved in the protection and repair of oxidative damage. Regarding gene ontology, this is the first mention of this protein in quail tissues, where it is closely related to the antioxidant function, testicular tissue protection, and decrease in the effects of oxidative stress on fertility.

The enzyme superoxide dismutase [Cu-Zn] plays a role in the body's defense system and neutralizes excess ROS, thus preventing damage to the cell structure (Luz et al., 2011). It naturally occurs in the testes of birds and protects them against lipid peroxidation (Mahmoodpour et al., 2017). Froman and Thurston (1981) compared SOD1 activity in the sperm of turkeys and roosters and observed that it is higher in turkey sperm than in rooster sperm. According to Mruk et al. (2002), a low SOD level in the testis would not only make the organ susceptible to oxidative damage but would also alter specific testicular functions and result in the loss of homeostasis.

The oncological results show that only 1% of the total proteins found have antioxidant function. This level is close to that reported by Słowińska et al. (2017), who studied semen of wild turkey (Meleagris gallopavo) and reported that of 137 proteins observed, only 2% had antioxidant function. Antioxidant activity in the testes is fundamentally important because these enzymes protect against oxidative factors, ensuring high fertility (Surai, 2002).

The seminiferous tubules of most animals have a diameter ranging from 150 to 300 μm (Razi et al., 2010). In this experiment, the lowest and highest values obtained for the diameter of the seminiferous tubule were 183.5 and 221.3 μm, respectively. However, the values obtained in the four treatments were not influenced by the CA level used in this experiment. The heights of the germinal epithelium were significantly different according to the CA level; that is, as the CA concentration in the feed increased, the germinal epithelium shortened, resulting in faster spermatogenesis. Spermatogenesis can suffer failures due to several favors, such as nutrition, handling, and ambience. Animals subjected to heat stress, for example, may have a lower sperm concentration and motility, in addition to an affected spermatogenesis (Turk et al., 2015, 2016; Fouad et al., 2016). With the use of CA, the compounds present in it, such as carnadol, can delay the harmful effects that heat stress has on quail, resulting in increased sensitivity to steroids, which improves gamete production (Durmic and Blache, 2012).

In this study, animals in the control group had an average epithelium height of 71.2 μm, and treatment with 0.75% CA resulted in an average height of 45.6 μm. The diameter of the epididymal duct was also influenced by treatments, with a statistically significant difference between the control (29.9 μm) and the treatment with 0.75% CA (101.5 μm).

When analyzing the histological images of the testicles, we observed germ cells at different stages of development, such as spermatogonia, spermatocytes, spermatids, and spermatozoa. We also noticed the presence of Leydig cells, which are responsible for the production of testicular androgens in the interstitial space (Liu et al., 2014). Sertoli cells were also present in the seminiferous tubules of quail; they regulate germ cell development and provide energy substrates, such as lactate, and hormones (Oliveira et al., 2012). The epididymal ducts were uniform, with a high sperm presence, and efferent ducts with ciliated and non-ciliated cells. According to Franzo et al. (2008), most of the epididymis is made up of efferent ducts, which are connected to the testis network and are the sites of sperm transition to the epididymal lumen, making them fundamental for fluid absorption (Bedford, 1978; Aire, 1979).