1. Introduction
We have conducted research on the anti-stress effect of theanine, an amino acid that is mainly found in the leaves of
Camellia sinensis L. (green tea) [
1,
2,
3,
4,
5,
6,
7,
8]. In our previous studies wherein the anti-stress effects of other amino acids in green teas were examined, L-arginine (Arg) was found to exert an excellent anti-stress effect that is similar to or better than that exerted by theanine [
4]. Almost the same anti-stress effect of Arg was observed at 0.032–3.2 mg/kg/day. Arg is the second most abundant amino acid, after theanine, in high-grade green teas [
4]. Chronic psychosocial stress has been demonstrated to shorten lifespan and accelerate age-related alterations such as cerebral atrophy, oxidative damage, cognitive dysfunction, and behavioral depression in stress-loaded senescence-accelerated mouse prone 10 (SAMP10) [
1,
2]. Mice of this strain have been reported to exhibit a short lifespan, aging-related brain atrophy, and cognitive decline even under normal conditions [
8,
9,
10,
11] and are sensitive to stress [
8]. Therefore, we aimed to further elucidate whether Arg has anti-stress potential against chronically-stressed SAMP10 mice in the present study.
Fundamentally, Arg, one of the 20 basic natural amino acids, is functionally classified as an essential amino acid in birds, carnivores, and young mammals and semi-essential for adults, and has been identified to play critical roles in health, including immune response [
12], wound healing [
13], growth hormone release [
14], and cell proliferation [
15]. Dietary sources of Arg include meat, wheat, sea foods, milk, cheese, corn, soy, nuts, and others [
16]. In Japanese green tea, Arg is contained in the range from 0.85 to 3.14 mg/g as a free amino acid [
17]. In previous studies, dietary Arg has been reported to suppress oxidative stress [
18] and inflammatory responses [
19]. Ingested Arg via the gastrointestinal tract is absorbed in the small intestine. Thereafter, approximately 40% of the ingested Arg is circulated systemically [
20]. Dietary Arg is degraded by arginase, which converts Arg into urea, ornithine, proline, polyamines, glutamate, and glutamine [
21]. In addition, nitric oxide, which can be converted from Arg by nitric oxide synthase [
21], acts as a precursor of signaling molecules [
22] and is involved in several functions, including the vasodilation of blood vessels [
21], synaptic plasticity [
23], learning and memory processing [
24], and modulation of neuronal function during stress and anxiety [
25]. Therefore, Arg and its metabolites play many important roles in health.
Chronic psychosocial stress has been associated with various mental disorders such as depression and anxiety [
26]; it elevates the risk of neurodegenerative diseases, including Alzheimer’s disease, dementia [
27], and cardiovascular diseases, accelerates aging, and shortens lifespan [
28]. Numerous animal and human studies have shown the deleterious effects of stress on the brain, behavior, and cognitive function [
1,
8,
29,
30,
31]. The brain is highly susceptible to stress during both early childhood and old age [
32]. Stress activates the hypothalamic–pituitary–adrenal axis which leads to the secretion of glucocorticoids from the adrenal glands [
32,
33]. Increased levels of glucocorticoids have been associated with neuronal loss [
34,
35], cognitive impairment, and Alzheimer’s disease development [
36].
In the present study, to elucidate the anti-stress potential of Arg on stress-loaded SAMP10 mice, the long-term effect of stress was observed by measuring the cognitive function and depressive-like behavior of mice at nine months of age. Furthermore, the lifespan of these mice was measured. Next, to elucidate the mechanism of Arg in the brain, the initial responses of SAMP10 mice loaded with stress for three days were used to observe the changes in lipid peroxidation (LPO) and gene expression levels in the hippocampus and prefrontal cortex.
3. Discussion
In this study, we found that Arg exerts a remarkable anti-stress effect on the brain as a new function. A significant increase in oxidative damage was observed in the brain of stress-loaded SAMP10 mice, and these mice exhibited cognitive decline as well as a shortened lifespan as they aged. No lifespan-shortening effect due to confrontational housing was observed in WT mice such as ddY and C57BL/6 (data unpublished). Oxidative stress plays a crucial role in the aging process, particularly in cognitive dysfunctions [
48]. Previous data suggested that dietary Arg supplementation ameliorated oxidative stress and that oral administration of Arg at a dose of 1.6 g/day for three months substantially reduced LPO levels in patients with senile dementia [
49]. Our Arg dose administered to SAMP10 mice (3 mg/kg) was lower than the dose in the above report, but it supports that Arg has an inhibitory effect on LPO.
Next, we looked at the molecular target of Arg in the brain to elucidate its mechanism in suppressing oxidative damage. The expression of several immediate-early genes (IEGs) such as
Nr4a1,
Arc, and
Cyr61 significantly increased in the hippocampus of stressed mice and was suppressed in mice that ingested Arg (
Figure 5A).
Nr4a1 is associated with adrenal stress response and excessive neuronal excitotoxicity [
37] and is known to be a potent pro-apoptotic molecule that induces nerve cell death [
38,
39]. Most importantly, Nr4a1 activated in response to oxidative stress has been reported to translocate from the nuclei to the mitochondria and induce mitochondrial damage and cell death [
50]. Arc reportedly plays a critical role in the neuronal excitotoxicity mediated by glutamate receptor signaling. Moreover, an elevated mRNA and protein expression of Arc has been detected in rat cortical neurons via neurotoxic stimulation [
40]. Cyr61 induction has been associated with neuronal cell death [
45], and Cyr61 elevation has been reported to be associated with oxidative stress as it was markedly upregulated at both the gene and protein levels against reactive oxygen species induction in human dermal fibroblasts cells [
51].
Arg incorporated into the brain has been shown to completely block glutamate-induced neuronal excitation in the ventromedial hypothalamus of rats [
52]. Our data suggest that dietary Arg incorporated into the brain suppresses the stress-induced elevation of
Nr4a1,
Arc, and
Cyr61 in the hippocampus and that the suppression of these genes may be involved in preventing neuronal cell death through the regulation of excessive neuronal excitotoxicity and mitochondrial damage via the suppression of oxidative damage in the brain.
Contrarily,
Hba-a2 and
Hbb-b2 expression levels were increased in the hippocampus and prefrontal cortex of mice that ingested Arg under both group and confrontational housing. As both Hba and Hbb are co-localized within the mitochondrion of neurons and are closely associated to maintain the neuronal mitochondrial function as well as survival of neurons [
47], we speculate that Arg protects neurons by maintaining the neuronal mitochondrial function. An increase in the expression of Hba and Hbb reportedly has therapeutic effects against neurodegenerative disease, and increasing evidence suggests that a deficiency in these chains in the brain is associated with neurodegenerative disease [
53,
54]. Our results suggest that Arg may also be important in protecting the brain from neurodegenerative diseases such as Alzheimer’s disease.
Arg revealed a similar protective effect to that of theanine on stress-induced shortened lifespan, cognitive decline, and depression in SAMP10 mice. We have shown that theanine significantly altered the gene expression of neuronal PAS domain protein 4
(Npas4) and lipocalin 2 (
Lcn2) in the hippocampus of stressed SAMP10 mice [
8]. Therefore, theanine and Arg suppress stress in different ways.
In the present study, we demonstrated that orally administered Arg modulates psychosocial stress-induced gene expression in the hippocampus and prefrontal cortex of SAMP10 mice. Further study is necessary to elucidate how dietary Arg modulates the gene expression of Hba-a2, Hbb-b2, and IEGs (Nr4a1, Arc, and Cyr61) in the brain. It may be necessary to determine by immunofluorescence study that which cell types in the hippocampus and frontal cortex, namely, glial or neuron, is primarily reduce lipid peroxidation by Arg. In addition, the effect of Arg on mitochondrial morphology and function needs to be elucidated.
4. Materials and Methods
4.1. Animals, Arg Preparation, and Housing Condition
Four-week-old male SAMP10/TaSlc (SAMP10) mice were purchased from Japan SLC Co., Ltd. (Shizuoka, Japan), and bred under conventional conditions in a temperature- and humidity-controlled room with 12/12 h light–dark cycle (light period, 08:00–20:00; temperature, 23 ± 1 °C; relative humidity, 55 ± 5%). Female mice were not used in this experiment because females are not as territorial as males. The mice were fed a normal diet (CE-2; Clea Co., Ltd., Tokyo, Japan). Arg (Wako Pure Chemical Co., Ltd., Osaka, Japan) was dissolved in water at 10 μg/mL. The volume of water containing Arg consumed by the mice was measured. Forty-eight mice were divided into four groups and observed to determine the long-term effects of Arg on their cognitive function and depression at nine months of age. Thereafter, the mice were fed continually to measure their lifespan. To study the effect of Arg on the brain, another group of 24 mice was used: 12 mice ingested Arg by drinking water ad libitum at 10 μg/mL (3 mg/kg) from one month of age. The remaining 12 mice ingested water as a control. Arg solution was freshly prepared twice a week. All experimental protocols were approved by the University of Shizuoka Laboratory Animal Care Advisory Committee (approval No. 166197, 5 April 2016) and were in accordance with the guidelines of the US National Institutes of Health for the care and use of laboratory animals.
4.2. Housing Condition for Confrontation
To induce psychosocial stress in mice, confrontational housing was used, as previously described [
1,
2,
4,
8]. To build territoriality, two mice were housed for one month in a standard polycarbonate cage that is equally divided into two units by a stainless-steel partition. Next, the partition was removed, and the mice were co-housed confrontationally to induce psychosocial stress (
Figure 6). Group housing was used as a model of the low-stressed condition. Two experiments in a linear diagram are shown including each experimental step for long term and short term (
Figure 6).
4.3. Memory Acquisition Test
To measure the learning ability of mice, a step-through passive avoidance task was tested on 9-month-old mice, as previously described [
2]. Briefly, when a mouse enters a dark chamber from a light chamber, the door in the chamber is closed and an electric foot-shock is delivered at 50 μA for 1 s (SGS-003, Muromachi Kikai. Co., Ltd., Tokyo, Japan). The acquisition of the avoidance response was considered successful if the mouse remained in the light chamber for 300 s. The trial was repeated until the mouse satisfied the acquisition criterion within five trials. For each trial, the time spent by the mice in the light chamber was subtracted from 300 s; the results from the successive trials were summed up for each mouse to determine the time required for learning (“learning time”).
4.4. Measurement of Immobility in the Tail Suspension Test
To examine behavioral depression, mice were individually suspended by their tails at a height of 30 cm using a clip for tail suspension (MSC2007, YTS Yamashita Giken, Tokushima, Japan). The duration of immobility was recorded for 15 min, as previously described [
1]. The mice were considered immobile only when they were both passively hanging and completely motionless. The duration during which the mice were immobile was measured.
4.5. Measurement of Oxidative Damage in the Brain
Mice that were confrontationally housed for three days after being singly housed for one month were used for the quantification of LPO. LPO in the brain of SAMP10 mice was measured using a lipid hydroperoxide assay kit (Cayman Chemical Company, Ann Arbor, MI, USA) according to the manufacturer’s instructions. Briefly, approximately 50 mg of the cerebral cortex was homogenized in 500 μL of HPLC-grade water. An equal volume of methanol solution saturated with Extract R? was added following the addition of 1 mL chloroform–methanol (2:1, v/v) solvent. After centrifugation at 1500× g for 5 min, the bottom layer of the chloroform was collected, and the lipid hydroperoxide content was measured via redox reactions with ferrous ions. The resulting ferric ions produced from the reaction of hydroperoxide with the ferrous ions were detected using thiocyanate ion as the chromogen. The absorbance of each sample at 500 nm was obtained (n = 6/group).
4.6. Measurement of DNA Microarray and qRT-PCR
The mice that were confrontationally housed for three days after being singly housed for one month were used for DNA microarray analysis. The mice were provided 10 μg/mL (3 mg/kg) Arg-water ad libitum. RNeasy Mini Kit (NucleoSpin
? RNA, 740955, Takara Bio Inc., Shiga, Japan) was used to extract total RNA from the hippocampus. To synthesize biotinylated cRNA, total RNA was processed using One-Cycle Target Labeling and Control Reagents (Affymetrix, Santa Clara, CA, USA) and hybridized to a Total RNA Mouse Gene 2.0 ST Array (Affymetrix) using three biological repeats per group. The significance of Arg intake was statistically tested using two-way ANOVA at
p < 0.001 [
55,
56].
For the measurement of qRT-PCR, group-housed same-aged mice were used as the reference. Total RNA was isolated from the homogenized hippocampus and prefrontal cortex as described above. cDNA was prepared from the obtained RNA using PrimeScript
? RT Master Mix (RR036A, Takara Bio Inc.). qRT-PCR analysis was performed using PowerUp
? SYBR
? Green Master Mix (A25742, Applied Biosystems Japan Ltd., Tokyo, Japan) and automated sequence detection systems (StepOne, Applied Biosystems Japan Ltd., Tokyo, Japan). Previously validated primers for
Nr4a1,
Arc,
Cyr61,
Hba-a2, and
Hbb-b2 [
37,
57,
58,
59,
60] (
Table 2) were used to quantify their relative gene expression. β-actin was used as the internal control.
4.7. Statistical Analysis
Statistical data are presented as the mean ± standard error of the mean. Statistical analysis was performed using one-way ANOVA followed by Tukey–Kramer’s honest significant difference method for cognition activity and a tail suspension test. Fisher’s least significant differences were used for qRT-PCR and the LPO assay. After calculating survival rates using the Kaplan–Meier method, the difference in survival rate was tested using the log-rank test. p < 0.05 was considered statistically significant.