Chronic caffeine use does not influence behavior and brain oxidative status in mice

Widely consumed in foods and as an energy supplement, caffeine has been studied given its pharmacological effects, especially on Central Nervous System (CNS). In this sense, the present study investigated whether and to what extent caffeine chronic use can influence the brain oxidative status and the behavioral activity of the C57BL/6 female mice. For this, fifteen animals were randomized into the following groups: Control (0.9% saline solution), Caf10 (10 mg/kg caffeine), and Caf50 (50 mg/kg caffeine). The animals received one daily caffeine dose by i.p route for 120 days. Twenty-four hours after the last administration, the animals were subjected to behavioral tests and euthanized. The blood was used by biochemical analysis and the brain to evaluate the oxidative status and micromineral levels. The caffeine did not influence the anthropometric parameters, lipid profile, and Creactive protein levels. Further, superoxide dismutase (SOD) and glutathione-stransferase (GST) activities maintained the same response profile. On the other hand, catalase (CAT) activity was decreased in both groups receiving caffeine compared to the control group. Despite this, malondialdehyde and carbonyl protein levels did not change among the groups, as well as the distribution micromineral levels. In the same way, no caffeine dose altered the findings of anxiety-like behaviors in the animals. Considering the time of caffeine administration, we believe that there was a cellular adaptation triggered by its use, tending to a protective effect on the brain.


INTRODUCTION
Caffeine (1,3,7-trimethylxanthine) is present in the composition of a wide range of foods (e.g., coffee, tea, soft drinks, chocolate), energy drinks, and, commonly, it is used as an ergogenic feature for athletes (1). The effects of caffeine in the brain have been historically described as a psychostimulant, elucidated in behavioral investigations in both humans and rodents [1][2][3][4] . This molecule is a type of methylxanthine alkaloid and makes up the purine group, standing out for its antioxidant and stimulating potential of the central nervous system (CNS) 5,6 . Studies of the pharmacology of caffeine have pointed out that after its consumption and absorption by the organism, this natural molecule is quickly transported to the brain (half-life to 2-12 hours), being capable to influence the response of neurons (e.g., calcium release, ryanodine receptors agonist, trigger mitochondrial reactions), besides influencing the ionotropic receptor's activity (e.g., inhibition of 5'nucleotidases and dephosphorylate the GABAA receptor) [7][8][9][10][11] .
Structurally, caffeine is similar to adenosine, an endogenous neuromodulator purine nucleoside released on the CNS, and that acts as a hormone over different receptors 2,12 . Studies have shown that the effects of this molecule on the CNS can reflect its action as a nonselective antagonist of these adenosine receptors (A1 and A2), with consequent blockade of A2A receptor densely located in the striatum nucleus, showing a positive relationship with the neuroprotector 13,14 . These receptors A2A, when active, have been potentially associated with the control of the N-methyl-D-aspartate (NMDA) pathway that results in the production of caspase-3 and reactive oxygen species (ROS) by mitochondrial pathway predisposing this way the tissues to suffer oxidative damage 9,10 . Additionally, the systemic effects of caffeine on A1 receptors could trigger lipolytic response profiles, which are usually related to modulation in serum lipid levels, such as cholesterol [15][16][17] .
Due mainly to the influence of caffeine in the NMDA pathway, potential antioxidant effects of this compound have been reported in the nervous system 10,18 . This way, the imbalance between the reactive oxygen metabolites production and the downregulation of the endogenous antioxidant defense system has been closely associated with the oxidation and consequent functionality of lipids, proteins, and nucleic acids. Aligned with that fact, these mechanisms of oxidative stress besides predispose the loss of important cognitive and brain functions 8,19 , maintaining a close relationship with reactivity to stress and anxiety disorders [20][21][22][23] . In this perspective, Rammal et al. 24 showed that anxious mice exhibited higher ROS accumulation in the neuronal and glial cells of three regions of the CNS (cortex, cerebellum, and hippocampus) when compared to the group of animals not linked to anxiety-related behaviors. Thus, the use of substances with antioxidant potential, as caffeine has been used as a pharmacotherapeutic co-adjuvant and neuroprotector, in the context of the neuronal and pathological changes in the CNS, including anxiety disorders and neurodegenerative conditions 18,25,26 .
An efficient antioxidant mechanism is preponderant for minimizing the brain vulnerability of the brain against ROS attack. These facts have been commonly associated with the brain's high metabolic rates, its large concentration of polyunsaturated fatty acids (i.e., a target of free radical) with primordial regulatory functions, and lower levels of endogenous antioxidant enzymes of the brain tissue, compared to peripheral organs. Indeed, the brain cells undergo targets constant attacks by free radicals thus requiring the most effective defense mechanism 27-29 . In line with this fact, it is substantiated that the bioavailability of chemical elements as zinc (Zn), selenium (Se), and iron (Fe), besides participating as mediators in different cell signaling processes, when acting as enzymatic co-factors has the potential of the modulating the activity of antioxidant enzymes, thereby improving the system of defense brain against ROS attack 10,28,29 . Hereupon, it is notable that studies with rodents have shown that oxidative stress in the brain can reduce the bioavailability of these minerals 30,31 .
In this context, there have been frequent approaches involving the use of exogenous antioxidant substances, dietary and pharmacological ones, as coadjutants in the improvement of the oxidative function of the different tissues and organs 32,33 . Due to the high frequency of caffeine consumption, from stimulant for athletes to coffee intake by the world population, in addition to its notable antioxidant profile associated with neuroprotective properties, the present study was carried out to evaluate if and to what extent its chronic use, in two different doses, has the potential to influence the brain cells oxidative status and the behavioral activity in C57BL/6 mice.

Animals, treatments, and biometrics
Female C57BL/6 mice (8-week-old; n = 15) were kept in controlled environmental conditions (temperature 22 ± 2 °C, air humidity 60-70%, and 12/12 h daily light/darkness cycles). The animals were randomized into 3 groups with 5 animals per group as follows Control group (0,9% saline solution), Caf10 (10 mg/kg caffeine, and Caf50: 50 mg/kg caffeine. Food and water were provided ad libitum. The caffeine (Sigma, St Louis, MO, USA) was dissolved in 0.9% NaCl (sodium chloride) and provided daily by intraperitoneal (i.p.) dose for 120 days, a chronic exposition in rodents 10,34 . The experimental period and the caffeine doses were selected based on Hughes and Hancock 35 . Additionally, the doses administered were daily determined by animal weight measures. Twenty-four hours after the last caffeine dose administration, the behavioral activity of the animals was tested and they were anesthetized with tribromoethanol (250 mg/kg, i.p.) and euthanized by cardiac puncture for exsanguination. The blood was collected for biochemical analysis, while the whole brain was rapidly removed (< 1minute) and submerged on cold (4ºC) phosphate buffer saline (PBS; 50mM, pH 7.0), followed by the cortex dissection and storage (-80ºC). These frozen samples were subsequently used for enzymatic analysis to assess the oxidative brain profile and for the detection of microminerals Food efficiency was evaluated during the experiment by obtaining parameters such as dietary intake (g/g mass/day), energy intake (kcal/g mass/day), and body mass (g). Dietary intake and body mass were recorded weekly. Energy intake was estimated based on the dietary intake using the reference values provided by each

Biochemical analysis
The biochemical analysis was accomplished by quantifying plasma levels of total HDL cholesterol, LDL cholesterol, and C-reactive protein. These biochemical parameters were analyzed on a spectrophotometer. Ultimately, the C-reactive protein was determined by using immunoturbidimetric assay, with the manufacturer's instructions (Bioclin, Belo Horizonte, Minas Gerais, Brazil).

Enzymatic antioxidant activity
The activity of antioxidant enzymes was investigated by using brain samples homogenized in ice-cold phosphate buffer (pH = 7.0) and centrifuged at 6200 rpm

Lipid and protein oxidation
Lipid peroxidation was evaluated from the quantification of thiobarbituric acid reactive substances (TBARS) according to a standardized biochemical colorimetric method. Briefly, the samples of the brain were homogenized in sodium phosphate buffer and centrifuged at 12.000 rpm for 10 min. The homogenate was collected and incubated with a thiobarbituric acid solution (0.375% thiobarbituric acid, 0.25N HCl, and 15% trichloroacetic acid) for 15min. TBARS levels were quantified spectrophotometrically at 535 nm 38,42 . Protein oxidation was estimated as the cerebral content of protein carbonyl (PCN) 43 . After homogenate removal, PCN content was measured in tissue pellets by the addition of 0.5 mL of 10 mM dinitrophenylhydrazine (DNPH). The reaction is based on the derivatization of the carbonyl group with 2,4-DNPH, which leads to the complementary formation of a stable product (2,4-dinitrophenyl hydrazine). PCN content was measured spectrophotometrically at 370 nm 42 .

Open field test
The open field apparatus consisted of a white wood box measuring 30 cm wide by 30 cm deep and 30 high. The box bottom was divided into 9 squares of 10 x 10 cm area each one, painted with black paint. This marking allowed the free movement of the animals by the total area of the box. The image monitoring and record for subsequent analysis were performed using a Sony® camera (Handycam model), positioned above the open field apparatus. The procedure was performed according to described by Teixeixa et al. 46 . The animals were placed in the box center, and their behavior was filmed for 600 seconds. The variables evaluated were: the number of the square that the animals crossed completely (i.e., with the four legs); the lifting of body and maintaining of weight over the hind legs or Rearing; the number and permanence of the entrance in the square center and self-cleaning or Grooming.

The hole-board test
The hole-board apparatus consisted of a white wood plate (40 cm long by 40 cm wide and 5 cm deep). The plate has 16 holes equidistant with 3 cm in diameter and is coupled in a glass box with 40 cm long by 40 cm wide and 30 cm high. Monitoring and recording of images for later analysis were performed by the same digital structure of the last behavior test. The camera was positioned one meter in front of the perforated plate apparatus. The procedure was performed as described by Kuru et al. 47 . The animals were placed in the center of the plate and their behavior was filmed for 300 seconds. The variable evaluated was the number of diving (i.e., times the animal dipped its head in the holes of the plate). It was considered valid when the head lowered at least enough for the eyes to be immersed by the hole.

Statistical analysis
The normality distribution of data was verified by the Shapiro-Wilk test. Data with parametric distribution were analyzed using a one-way analysis of variance (oneway ANOVA), followed by the Student-Newman-Keuls posthoc test. The nonparametric distributed data were transformed by logarithmic transformation before the analysis to normalize the distribution. Proportion data were transformed by angular transformation before the analysis given their nature as percentages. Data were expressed as mean ± standard deviation (mean ± S.D.). All results with P ≤ 0.05

RESULTS
Caffeine, in both concentrations, contributed to maintaining the anthropometric parameters associated with the final body mass and energy intake (kcal/day).
Moreover, this response profile did not influence the Lee index, the body mass index (BMI), and the tail-muzzle length in the experimental animals (p > 0.05). Similarly, the relation between the biomolecules of lipidic profile HDL and LDL and the inflammatory biomarker CPR (C reactive protein) maintained their baseline values compared to controls animals (p > 0.05; Table 1). In the same way, the relation between the biomolecules of lipidic profile HDL and LDL and the inflammatory biomarker CPR (C reactive protein) further maintained their baseline values compared to controls animals (p > 0.05; Table 1). Finally, caffeine did not influence the animals' weight gain compared to the control group (p > 0.05; Figure 1). Both doses of caffeine contributed to a significant decrease of catalase (CAT) levels in the brain tissue when compared to the control group (p < 0.05; Fig. 2B). On the other hand, the caffeine doses administered maintained the same baseline values of the cerebral superoxide dismutase (SOD) and glutathione-s-transferase (GST) of the control animals (p > 0.05; Figs 2A and C). Although we have observed a differential profile of response to the CAT enzyme, this was not sufficient to cause an imbalance between increased reactive oxygen metabolites and the enzymatic antioxidant defense system, once a time that no disorders associated with lipidic or protein oxidation (Fig. 3A and B)     In association with this response profile, the bioavailability of brain chemical elements, such as sodium (Na), potassium (K), phosphor (P), calcium (Ca), copper (Cu), magnesium (Mg), iron (Fe), zinc (Zn) and selenium (Se), did not present modifications after caffeine intake in both caffeine doses (10 and 50 mg/kg) by 120 days (p > 0.05; Table 2). Na(mg/kg), K (mg/g), P (mg/g), Ca (µg/g), Cu (µg/g), Mg (µg/g), Fe (µg/g), Zn (µg/g) and Se (µg/g) represent, respectively, sodium, potassium, calcium, copper, magnesium, iron, zinc and selenium concentrations. Control group, Caf10 (10 mg/kg caffeine) and Caf50 (50 mg/kg caffeine). Different letters in the rows indicate the statistical difference between groups (P < 0.05).
In the elevated plus-maze (EPM) test, the same entry frequency in the arm open or arm closed was observed between the animals independent of the caffeine doses administrated (p > 0.05; Fig 5A and 4B). The animals also remained by a similar    Hereupon, this enzyme is only effective when the peroxide is used as a substrate for the action of CAT, which converts it into water molecules, consolidating the antioxidant effect 57 . Besides, peroxide can interact with the superoxide anion and trigger the production of the highly reactive hydroxyl radical. In this sense, it would be possible that the reduction of catalase activity contributed to increasing the vulnerability of rodents' brains to oxidative stress. However, despite the enzyme This response profile of the antioxidant enzymes can be a relation to the maintenance of chemical element levels of microelements in brain tissue. According to Barbosa et al. 58 microelements like Zn, Cu, Fe, and Se, besides assisting in processes associated with the integrity of brain metabolism (i.e., effective mediators of innate immune activation 59 , modulating the production of TNF-α, TIMP-1, and the blood capillary proliferation marker α-SMA) 60 , many of these minerals also function as enzymatic co-factors. In this sense, Zn and Cu for example, are required by superoxide dismutase for dismutation of superoxide anion to oxygen and hydrogen peroxide, while Fe is essential for enzymatic CAT activity 61,62 . Also, Se is a cofactor for glutathione peroxidase involved in the reduction of hydroperoxides in the cell 63,64 . Despite the functionality of these antioxidant enzymatic groups and the relevance of the minerals' co-factors balance, the caffeine in the experimental conditions evaluated in our investigation was not capable of influence these variables, corroborating with the maintenance of the activities of these respective antioxidant defense enzymes evaluated.
It has been established that the balance between the endogenous antioxidant system and the attack by reactive oxygen and nitrogen metabolites are fundamentals for brain cells' integrity and, consequently, maintenance of the behavioral activities as the mobility and anxiety profile 10,35 . In this perspective, the review of Bishop et al. 65 showed a positive relationship between continuous oxidative damage in brain tissue (i.e., related to different ROS production pathways) and behavior activities functional decline. The modulation of the A2A receptor by caffeine has the potential to mitigate altered behavioral phenotypes in rats, as hyperlocomotion 66  Equally, though the caffeine use had been closely associated to decrease anxiety-like behaviors in rats 4,35 , this substance was not capable of an induced change of these variables in the experimental animals in our study, evaluated by hole-board test. In this test, independent of caffeine dose, the frequency of diving in the apparatus was similar in all animals compared to the control group that received only water. In a study conducted by Pierard et al. 48 , the application of an experimental protocol that Considering the exposure, although the literature is consistent in pointed the effectiveness of the intake of thermogenic drinks with considerable antioxidant effect as the caffeine, besides its psychostimulating potential, these studies, for the most part, are quite heterogeneous, especially related to the dose consumed and to time of caffeine administration. This fact has been reflected in differential response profiles consistent with positive, negative, and/or adaptive effects associated with using this substance. Lined up with these observations, in our investigation the caffeine at 10 and 50 mg/kg dose administered daily for 120 days showed no evidence of significant improvement on brain oxidative status. Despite this fact, we did not observe any significant oxidative damage. This response profile was also consistent with the behavioral analysis of the animals that maintained their baseline values, similar to the control animals. Considering the time of caffeine administration to the animals, we believe that there was a process of cellular adaptation triggered by its chronic use. However, to confirm or refute this inference believed to be a necessary realization of more detailed studies evaluating the consumption of caffeine variable