The AEPG was prepared from whole fruits kindly donated by Nutracitrus SL (Elche, Alicante, Spain). The dose was selected based on a previous study where the extract induced an antidepressant-like effect in ovariectomized rats (Valdés-Sustaita et al., 2017). PNCG and EA was obtained from Sigma Aldrich (St. Louis, MO, United States), malonaldehyde (MDA), 2’,7’-dichlorofluorescein (DCF), 2’,7’-dichlorodihydrofluorescein diacetate (DCF-DA), Ferrous Sulfate (FeSO₄), ascorbic acid, nitroblue tetrazolium (NBT), phenazine methosulfate (PMS), ethylenediamine-tetra acetic acid (EDTA), nicotinamide adenine dinucleotide (NADH), N-2-piperazine hydroxy ’-2-ethane sulphonic acid (HEPES) and hydrogen peroxide (H₂O₂) were all obtained from Sigma Chemical Company (St. Louis, MO, United States). The solutions were prepared using deionized water from a MilliRQ purifier system (Millipore. Billerica, MA, United States).

Ovariectomized female Wistar rats (250–300 g, 3 months old) were housed in groups of five per polycarbonate cage with food and water available ad libitum, maintained in a 12 h light/dark cycle (lights on at 21 h), in a temperature-controlled room at 22°C. Tissue samples were immediately collected after decapitation of the animals and were preserved at a temperature of −70°C. All experimental procedures with animals were carried out following the official Mexican standard for the care and handling of animals (NOM-062-ZOO-1999) and approved by the Ethics Committee of the National Institute of Psychiatry Ramón de la Fuente Muñiz, CEI-200 (December 10, 2015).

For the study, all the rats used were ovariectomized under anesthesia with tribromoethanol (2% dose 0.1 mL/kg, i.p.). For surgery, a single incision was made in the midline of the ventral region, and oviducts were exposed. The ovaries were removed, which was corroborated by visual inspection. After 3 weeks, the animals were randomly assigned to each experimental and behavioral group (Bekku and Yoshimura, 2005; Estrada-Camarena et al., 2011).

In experiment 1, determinations of LPX, ROS, as well as for the methyl thiazol tetrazolium (MTT) reduction assay in brains from rats treated with a daily intragastric administration of AEPG (1.0 mg/kg for 14 days in a volume of 2 mL/kg) were performed (Figure 1). The anti-immobility effect of the AEPG was previously published (Valdés-Sustaita et al., 2017). The AEPG was prepared from the whole fruit and dissolved in water. The dose selection was based on a previous study in which AEPG showed an antidepressant and anxiolytic-like effect in ovariectomized rats (Valdés-Sustaita et al., 2017; Estrada-Camarena et al., 2020). After behavioral tests, all rats were killed by decapitation, and the brains were immediately removed to be placed in vials for storage in a −80°C freezer for further study.

In experiments 2 and 3, a dose-response curve of commercial PNCG and EA (0.0010, 0.010, 0.10, 1.0 mg/kg) was constructed to evaluate their antidepressant-like effect in the forced swim test (FST) with general activity prior to FST using ovariectomized female Wistar rats. In a previous study, we observed similar results with AEPG after an intraperitoneal or intragastric administration (unpublished results) as well as after a subacute (three injections in 24 h) or 14 days of treatment (Valdés-Sustaita et al., 2017, 2021). For this reason, to evaluate whether EA and PNCG induce an anti-immobility effect in the FST, both compounds were intraperitoneally administered during 14 days/once a day (Figure 1). Each compound was dissolved in vehicle (saline solution 0.9% NaCl).

Finally, from AEPG, a lyophilized mother liquor that main contains PNCG and a precipitated solid that mainly contains EA. To corroborate the effect of PNCG and EA obtained from AEPG, in experiment 4, independent groups of ovariectomized rats were assigned to one of the following groups: vehicle (saline treatment), Fluoxetine (10 mg/kg), AEPG (1.0 mg/kg), EA (0.010 mg/kg) and PNCG (0.010 mg/kg). All treatments were administered i.p., for 14 days/once a day. Thirty minutes after of FST session, rats were euthanized, and brains were removed to be placed in vials for storage in a -80°C freezer for oxidative stress determination.

The presence of EA and PNCG in AEPG was determined by gas chromatography-mass spectrometry.

A general activity test session was performed on day 14 to discard if treatment-related alterations in locomotor activity may influence behaviors on FST test. The rats were gently placed in one of the corners of a rectangular acrylic cage (43 × 33 × 20 cm) with a grid drawn on the floor (12 squares of 11 × 11 cm). The number of times the animal crossed any square with its four legs (crossing) in 5 min was recorded. After each test session, the cage was thoroughly cleaned with a cleaning solution (Estrada-Camarena et al., 2003).

The forced swim test (FST) was performed by exposing the rats to a glass cylinder (46 cm deep and 20 cm in diameter) with water at 23–25°C in which the rats could not touch the bottom or hold on with their legs or tail. Two swimming sessions were performed: an initial 15 min pretest session on treatment day 13, 1 h after drug administration. On day 14, the rats received their treatment, and 60 min later, they were again subjected to the forced swim test for 5 min (test session). After each swimming session, the rats were removed from the cylinders, dried with paper towels, placed in warm cages for 30 min, and returned to their cages. The test sessions were held between 12:00 and 15:00 which were video-recorded for later scoring. A single observer, blind to the treatment conditions, performed the score analysis counting the number of episodes of immobility, swimming, or climbing behavior shown by rats in intervals of 5 s during 5 min (Porsolt et al., 1977, 1978).

Three behaviors scores were defined as follow: (Scapagnini et al., 2012) immobility, where the rat floated in the water without fighting and did only the movements necessary to keep the head above the water; (Nestler et al., 2002) swimming, rat shows active swimming movements, more than necessary to simply keep the head above the water, such as moving around the cylinder or diving; and (Yan et al., 2010) climbing, rat shows active front leg movements in and out of the water, usually directed against the walls (Estrada-Camarena et al., 2003; Cruz et al., 2009).

Brain tissue from rats treated with AEPG, EA, PNCG (14 days, i.p.), or vehicle was homogenized (1:10 w/v) in Krebs buffer (pH 7.4). Then, 375 μL of each brain homogenate was incubated alone or with FeSO₄ (10 μM), or ONOO^– (50 μM) in a final volume of 500 μL for 2 h at 37°C in a water bath. After incubation, the sample was used to determine ROS, LP, and MTT reduction simultaneously.

After incubation with the pro-oxidants FeSO₄ (5 μM) and ONOO^– (25 μM), 500 μL of the TBA reagent (containing 0.75 g of TBA + 15 g of trichloroacetic acid + 2.54 mL of HCl) was added, and the final solutions were re-incubated in a boiling water bath (94°C) for 20 min. Samples were then kept on ice and centrifuged at 12,000 g for 5 min. MDA was determined as a colorimetric product using Synergy™ HTX multi-mode microplate reader (Biotek Instruments, Winooski, VT, United States) at a wavelength of 532 nm. MDA concentrations were calculated by interpolation on an MDA standard curve, constructed in parallel. Protein determinations were made by the Lowry method (Lowry et al., 1951). The results were expressed as μmoles of MDA per milligram of tissue.

Reactive oxygen species (ROS) were determined through DCF-DA oxidation (Ali et al., 1992; Herrera-Mundo and Sitges, 2010). Briefly, 125 μL of forebrain and liver homogenates previously incubated with FeSO₄ and ONOO^– were mixed with DCF-DA solution (final concentration: 75 μM); then, incubated in dark conditions at 37°C for 30 min. After incubation, the samples were centrifuged at 9,000 g for 10 min. ROS were detected in supernatants using fluorescence spectrophotometry (Synergy™ HTX multi-mode microplate reader, Biotek Instruments) at an excitation wavelength of 480 nm and an emission wavelength of 521 nm. Results are expressed as a percentage of ROS production considering the control as 100%.

The cellular function was assessed in forebrain and liver homogenates by determining MTT reduction to formazan crystals. Briefly, the brain homogenate (125 μL) was mixed with 4 μL of the MTT (5 mg/mL) and then incubated for 15 min at 37°C. Samples were then centrifuged at 17,000 g for 3 min and the pellets were suspended in 250 μL with acid-isopropanol. Optical density was determined using an Eon microplate reader (Biotek Instruments) at a wavelength of 570 nm. Results are expressed as the percentage of MTT reduction considering the control as 100%.

Protein was determined according to the Lowry method (PMID 14907713) using bovine serum albumin as a protein standard.

The ability of the study compounds (AEPG 0–4 mg/mL, PNCG 0–4 mg/mL, and EA 0–20 mg/mL) to scavenge hydroxyl radical (^⋅OH) was estimated through the Fe³ + -EDTA-H₂O₂-deoxyribose system (Halliwell et al., 1987; Floriano-Sánchez et al., 2006). The system contained different concentrations of AEPG, PNCG, and EA or an equivalent volume of vehicle (distilled water) for the control, 0.2 mM ascorbic acid, 0.2 mM FeCl₃, 0.208 mM EDTA, 1 mM H₂O₂, 0.56 mM deoxyribose, and 20 mM phosphate buffer (pH 7.4). ^⋅OH was generated by incubating the mixture at 37°C for 60 min. The iron salt (FeCl₃) was mixed with EDTA prior to its addition to the reaction mixture. The degree of degradation of deoxyribose by the ^⋅OH formed was measured directly in the aqueous phase by the thiobarbituric acid test (TBA). Briefly, 250 μL of the TBA solution was added to the samples and boiled in a water bath (94°C) for 10 min. Optical density was estimated at a wavelength of 532 nm using a Synergy™ HTX multi-mode microplate reader (Biotek Instruments). The results are shown as percent of OH scavenging capacity.

O₂^⋅⁣– scavenging was evaluated according to the method previously published (Fontana et al., 2001) based on the reduction of nitroblue tetrazolium (NBT) salt. The non-enzymatic PMS/NADH system that generates O2^⋅⁣– was used to reduce NBT into purple-colored formazan. The reaction mixture contained HEPES buffer (20 mM, pH 7.2), 196 μM NADH, 39.2 μM NBT, 3.92 μM PMS, and increasing concentrations of AEPG (0–1 mg/mL), EA (0–2 mg/mL), PNCG (0–1 mg/mL). The final volume of the mixture was 130 μL. The optical density was determined at a wavelength of 560 nm in a Synergy™ HTX multi-mode microplate reader (Biotek Instruments) each minute for 3 min. All tests were carried out in triplicate and independently. Results are expressed as a percentage of O2^⋅⁣– scavenging capacity.

ONOO^– was synthesized as previously described (Beckman et al., 1994). Briefly, 5 mL of an acidic solution (0.6 M HCl) of H₂O₂ (0.7 M) was mixed with 5 mL of 0.6 M KNO₂ in an ice bath for 1 s, and the reaction was quenched with 5 mL of ice-cold 1.2 M NaOH. Residual H₂O₂ was removed using granular MnO₂ pre-washed with 1.2 M NaOH, and the reaction mixture was left overnight at -20°C. The resulting yellow liquid layer on top of the frozen mixture was collected for the experiment. ONOO^– concentration was determined before each experiment at 302 nm using a molar extinction coefficient of 1,670 M^–1 cm^–1. The ONOO^– scavenging activity was measured by monitoring the oxidation of DCF-DA to DCF by the modified method of Crow and Beckman (1996). The reaction mixture (in a final volume of 1.15 μL) contained 14 μM DTPA, 36.2 μM DCFH-DA; samples were incubated with different concentrations of AEPG (0–5 mg/mL), EA (0–20 mg/mL), PNCG (0–1 mg/mL) and exposed to 35 μM ONOO^–. The optical density at 500 nm was determined in a Synergy™ HTX multi-mode microplate reader (Biotek Instruments). A test containing the reaction mixture, but not the sample, was considered as the 0% trapping capacity, or 100% of the induced DCF-DA oxidation by ONOO^– when added to the assay. To calculate the scavenging capacity of ONOO^–, the readings of all the tubes with the various study compounds were expressed as percentage of oxidation of DCF-DA and converted to the percentage of scavenging using the tube with 100% oxidation as a reference with DCF-DA.

Samples were dissolved in water or methanol (HPLC grade) and filtered through 0.22 μm filters (GHP, Acrodisc 13, Waters) before being injected into the chromatograph. The ultra high-performance liquid chromatography (UPLC) technique was carried out in a Waters Acquity UPLC H-Class liquid chromatography equipment fitted with a Waters photodiode array detector (UPLC, Acquity Waters, Singapore) and the Empower chromatographic software version 3 (Waters, Milford, MA, United States). Sample analysis was performed with a Symmetry C-18 column (100Å, 150 mm × 4.6 mm, 5 mm, Waters, Ireland) with the thermostat at 40°C. The mobile phase consisted in a gradient system of Milli-Q water acidified with 0.3% formic acid (solvent A) and methanol (solvent B). The initial gradient mixture was 90% A: 10% B up to 5 min. Then, a linear 70% A: 30% B was applied from 5 min. Whereas methanol (B) was gradually increased to 100% in the time of 10–12 min. Finally, the gradient was returned to the initial concentrations (90% A: 10% B). A constant flow rate of 0.9 mL/min was considered during a total time of 12 min. The detection wavelength used was 254 nm. Commercial standards of PNCG and EA were used as reference for comparison of the retention times and maximum absorption wavelength data.

Reactive oxygen species, LPX, and MTT data are expressed as mean values ± S.E.M. The Sigma plot 12.0 program (version 12.3, Systat software) was used for the statistical analysis, using U-Mann-Whitney for specific comparisons. For behavioral data, a one-way ANOVA test was applied, followed by the Holm-Sidak test as post-hoc analysis, accepting as significant only values of p ≤ 0.05. The trapping capacity was expressed as 50% of the inhibitory concentration (IC₅₀), which denotes the concentration of the compounds (AEPG, PNCG, and EA, in mg/mL) necessary to achieve a reduction of 50% on the respective molecules in oxidation relative to the test without the compounds of interest. The lowest IC₅₀ value suggests the best elimination capacity of the compound.

First, the AEPG treatment effect was evaluated when the brain homogenates were exposed ex vivo to pro-oxidants (FeSO₄ and ONOO^–). FeSO₄ and ONOO^– induced a significant increase (91 ± 12.4 and 50 ± 23%, respectively) in ROS formation in brain homogenates compared to the control group (p < 0.05; 0.0153 ± 0.001 mg of DCF/mg protein) (Figure 2A). However, when rats were treated with AEPG (1.0 mg/kg, 14 days) and the brain homogenates were exposed ex vivo to the pro-oxidants, no effect was observed on ROS production (p < 0.05), indicating that AEPG prevented the oxidative damage induced by the pro-oxidants. One Way ANOVA values yielding the following values F_(3,14) = 7-0.57, p = 0.003.

Lipid peroxidation was determined as a marker of oxidative Damage. After AEPG pretreatment, the Brain homogenates were exposed to pro-oxidants. As shown in Figure 2B, LPX was enhanced in homogenates of the brain (109 ± 11.3% and 78 ± 4.2%, respectively) by FeSO₄ and ONOO^– compared to the control (p < 0.05; 40.56 ± 5.2 μmoles of MDA/mg protein). However, the administration of AEPG significantly decreased the LPX induced by ONOO^– (p < 0.001), and a decreasing trend is observed with FeSO₄ in the brain. One Way ANOVA values yielding the following values F_(3,14) = 23.42, p < 0.0001.

The reduction of MTT was evaluated as an index of cellular function in the rat brain (Figure 2C). For this, homogenate preparations were exposed to toxic concentrations of FeSO₄ and ONOO^–, and a value of 100% represents the total capacity of MTT reduction by the cell (control), whereas in the groups treated with FeSO4 and ONOO^–, a decrease in cellular functionality of 52.84 and 51.25%, respectively, was observed in the brain (p < 0.001). AEPG pretreatment partially recovered cellular dysfunction induced by the ex vivo exposure to FeSO4 and ONOO^– in rats’ brains (22.7 and 28%, respectively; p < 0.05). One Way ANOVA values yielding the following values F_(2,10) = 11.06, p = 0.002.

Table 1 shows the effect of FLX, AEPG, EA, and PNCG on general activity. As it can be seen, none of the treatments significantly modify the locomotor activity test [F_(4,33) = 0.35, ns], discarding that an alteration of a stimulatory action of any treatment could contribute to decreasing the immobility behavior in FST.

Figure 3A shows that commercial PNCG decreased the immobility behavior compared with the control group at a dose of 0.01 mg/kg (p < 0.03) [F_(5,37) = 4.814, p = 0.002], In addition, PNCG produced a significant increase in the climbing behavior at 0.01 mg/kg (p < 0.001) (F_5,37 = 8.625, p < 0.001) without modification of the swimming behavior [F_(5,37) = 1.242, p = 0.309].

Figure 3B shows that commercial EA at doses of 0.001 and 0.01 mg/kg produced a significant reduction in the immobility behavior (p < 0.001) compared to the control group [F_(5,57) = 15.733, p < 0.001]. Regarding the effect of EA on the swimming behavior, a significant increase of the swimming behavior can be observed (p < 0.001) [F_(5,57) = 17.144, p < 0.001], panel B (Figure 3) without promoting significant changes in the climbing behavior [F_(5,57) = 3.541, p = 0.007].

The redox profiles of AEPG, PNCG, and EA were determined using non-tissue synthetic systems, in which ROS were specifically generated through combinatorial chemistry assays. Figure 4 shows the superoxide scavenging effect of these compounds, being the AEPG and PNCG more effective than EA, since the IC₅₀ for ROS was 0.012 and 0.015 μg/mL for AEPG and PNCG, respectively, while for EA, the IC₅₀ was 0.49 μg/mL (Table 2).

The hydroxyl radical (^⋅OH)-scavenging capacity for AEPG, PNCG, and EA was tested (Figure 5), considering that this radical is one of the strongest oxidants that is also produced by FeSO₄, in which AEPG demonstrated antioxidants effects previously. In this case, AEPG and PNCG were able to trap the ^⋅OH radical with an IC₅₀ 0.36 and 0.42 μg/mL, respectively. However, EA was not able to scavenge the hydroxyl radical at the concentration tested (0–20 mg/mL).

Regarding peroxynitrite (ONOO^–)-scavenging capacity (Figure 6), PNCG showed an IC₅₀ of 0.2467 μg/mL, whereas AEPG was able to trap ONOO- but with a higher IC₅₀ (1.33 μg/mL) than PNCG (Table 2). EA had no effect on trapping ONOO^– in the concentration range tested in this study (0–20 μg/mL).

Chromatographic profile of the AEPG (10 mg/mL) (Figure 7A) allowed observing two major peaks at the retention time of 5.459 and 8.962 min corresponding to the PNCG and EA constituents, respectively, which were also supported by their maximum absorption wavelength data (inserted panels) of the corresponding standards of PNCG (258.5/378.7 nm, Figure 7D) and EA (253.7/366.7 nm, Figure 7E). These references were also considered to characterize the presence of both compounds in the lyophilized mother liquor (Figure 7C) and precipitated solids (Figure 7B) as PNCG and EA, respectively.

The effect of AEPG and its lyophilized fractions were evaluated in the FST. As it can be noted in Figure 8, the treatment for 14 days with AEPG (1 mg/kg), the fraction that main contain PNCG (0.01 mg/kg) as well as that in which EA (0.01 mg/kg) was detected reduce the immobility behavior (p < 0.05) when compared with the control group. These effects were similar to that Induced by the antidepressant FLX [F_(4,33) = 7.41, p < 0.001]. The effect on swimming behavior is shown in Figure 8. As it can be seen, FLX increased significantly swimming behavior when compared against the control group (p < 0.05), while PNCG and EA showed a tendency to increase swimming behavior (p = 0.07). One Way ANOVA values yielding the following values [F_(4,33) = 3.32, p = 0.02]. Neither treatment significantly modify the climbing behavior [F_(4,33) = 2.51, p = 0.06].

Figure 9A shows that the pro-oxidants ONOO^– and FeSO₄ produced a significant increase in ROS formation in the brain of rats treated with vehicle (saline treatment) (p < 0.05). In contrast, in the rat brains treated with AEPG (1.0 mg/kg) and its fractions, EA (0.01 mg/kg) and PNCG (0.01 mg/kg), the pro-oxidant compounds ONOO^– and FeSO₄ had no effect on ROS production (p < 0.05 versus respective control).

When determining LPX in brains of rats treated with AEPG, AE, PNCG, and subsequent exposure to pro-oxidant compounds (ONOO^– and FeSO₄), we observed a significant decrease in LPX with AEPG (1.0 mg/kg) and EA (0.01 mg/kg) despite being subjected to FeSO₄ or ONOO^– (p < 0.05 versus control group), Figure 9B.

When determining the percentage of reduction of MTT in brains treated with FeSO₄ and ONOO^– we observed a cellular functionality decrease (54.64%, red column) and (48.08%, blue column) when compared with the control group (p < 0.05), Figure 9C. In contrast, this decrease was not observed in the brains of rats treated with AEPG and EA (p < 0.05). As it can be seen, the cellular functionality of the brains of rats treated with PNCG did not reach statistical significance.