In the past few years, many new drugs have been developed from natural or marine organisms. Also, there is an increasing trend of using alternative therapies including fermented products. During fermentation of Saccharina japonica (previously known as Laminaria japonica), the degradation products, amino acids and peptides have significant effect on the biochemical characteristics of fermented S. japonica (FSJ). Several studies reported that marine plants produce biological activities. For example, marine organisms including sea oyster might be useful for the prevention and treatment of liver diseases by inhibiting an oxidation process (Watanabe et al. 2012). Many constituents of laminaria have been used for medical treatments. One of the polysaccharide constituent of laminaria, laminarin, has antilipemic and anticoagulant functions. It has been known that the other constituents, histamine and lamine have the hypotensive effects (Walsh et al. 1982). Another constituent of laminaria, algin (sodium alginate) has anti-diarrhea and soothing effects (Gao et al. 2013, Liu and Zhao 2013). Also alginatecontaining kelp has been used for managing radioactive intoxication (Harikrishnan et al. 2011).

Gamma aminobutyric acid (GABA) is biosynthesized by animals, plants, and microorganisms via the α-decarboxylation of glutamic acid by a glutamate decarboxylase (Ueno 2000). A variety of traditional foods produced during process of microbial fermentation contain GABA. Several studies have reported that GABA is produced by most lactic acid bacteria, including Lactobacillus brevis, Lactobacillus paracasei, Lactobacillus buchneri, and Lactobacillus sakei B2-16 (Yokoyama et al. 2002, Choi et al. 2006, Kim et al. 2009, Kook et al. 2010). Glutamate is a major substance for GABA production and is especially abundant in foods such as sea tangle, cheese, soybeans, rice bran, and mushrooms (Kook et al. 2010, Lee et al. 2010a). In our previous study, it was shown that GABA-producing L. brevis BJ20 was isolated from Jotgal, a Korean traditional fermented food and was capable of completing conversion of L. japonica glutamic acid into GABA (Lee et al. 2010a), and GABA-enriched FSJ has been demonstrated to produce a number of biological effects such as hepatotoxicity improvement and anti-oxidant activity (Lee et al. 2010a, 2010b).

Trimethyltin chloride (TMT) is known as a classic neurotoxicant which can cause serious neuronal degeneration diseases. Exposure of the human population to many chemicals has generated concern about the potential neurotoxicity of new and exciting chemicals (Kaur et al. 2013). Our previous studies already proved through experimental observations that TMT-induced neurodegenerative processes support the efficiency and sensitivity of this model for providing insights into the mechanisms of neuro-degeneration (Park et al. 2011, 2012b). TMT-induced neuronal damage particularly in the hippocampal formation resulted in behavioral alterations such as hyperactivity, impairment of learning and memory (Koczyk 1996, Guo et al. 2004), epilepsy and spontaneous seizures (Geloso et al. 2011). TMT also procures cellular oxidative stress which leads to apoptotic death (Gunasekar et al. 2001, Jenkins and Barone 2004, Mundy and Freudenrich 2006). Also it has been shown that TMT-induced deficits with learning and memory may be mediated by the central cholinergic systems or neurotrophic factors such as brain derived neurotrophic factor (BDNF) and cAMP response element binding protein (CREB) (Park et al. 2012b).

The present study was to evaluate the memory improving effect of FSJ on TMT-induced learning and memory deficits and to examine the mechanism underlying these protective effects in rats. Rats were tested on a Morris water maze for spatial learning and memory and then the parameters such as expression of choline acetyltransferase (ChAT), CREB, and BDNF in the hippocampus and the neurite outgrowth of Neuro2a cells were analyzed.

Male Sprague-Dawley rats weighting 250-280 g each were purchased from Samtaco Animal Corp. (Osan, Korea). The animals were acclimatized for at least 7 days prior to the experiment. The animals were housed in individual cages under light-controlled conditions (12 : 12-h light : dark cycle) and at 23℃ room temperature. Food and water were made available ad libitum. All the experiments were approved by the Kyung Hee University institutional animal care and use committee. Also, this experimental protocol was approved by an Institutional Review Committee for the use of Human or Animal Subjects and the procedures were in compliance with the Declaration of Helsinki for human subjects, or the National Institutes of Health Guide for Care and Use of Laboratory Animals (Publication No. 85-23, revised 1985), the UK Animals Scientific Procedures Act 1986 or the European Communities Council Directive of 24 November 1986 (86/609/EEC).

The rats were injected intraperitoneally with TMT (8.0 mg kg^-1, body weight) dissolved in 0.9% saline and then they were returned to their home cages. From the 15th day after the injection of the drug (for 21 days, treatment of FSJ 50, 100, or 200 mg kg^-1, per oral was given), the water maze test was performed for 5 days.

FSJ was prepared as described by Kang et al. (2012). For the production of GABA-enriched preparations, S. japonica were added to water at a ratio of 1 : 15 (w/v) and 3% yeast extract and 1% glucose were then added to the mixture, each based on the amount of yeast extract and L. japonica required to aid in the fermentation process. After autoclaving at 121℃ for 30 min, FSJ was filtered and Lactobacillus brevis BJ20 (accession No. the Korean Collection for Type Culture [KCTC] 11377BP) culture broth was mixed to the fermented sample at a concentration of 5% (v/v), which was well mixed and then incubated at 37℃. The GABA content in the culture broth was measured as follows. A 1-mL sample was taken and diluted 20-fold using 0.02 N HCl. The dilution solution (2 mL) was filtered through a 0.2-μm membrane filter (dismic-25 cs; Toyo Roshi Kaisha, Tokyo, Japan). The GABA and glutamate in the filtered sample (20 μL) were measured using amino acid analyzer (AAA L-8900; Hitachi High-Technologied Co., Tokyo, Japan) with a flow rate of 0.3 mL min^-1. FSJ contained 6.29% (w/w) glutamic acid and GABA was not detected before fermentation of FSJ. However, GABA content (5.56%, w/w dry weight) was dramatically increased during fermentation with L. brevis. Phophoserine, aspartic acid, threonine, serine, and glycine were also increased after fermentation. These results were in agreement with our previous report (Lee et al. 2010a).

The swimming pool of the Morris water maze was a circular water tank 200 cm in diameter and 35 cm deep. It was filled 21 cm height with opaque water at 23 ± 2℃. A platform 15 cm in diameter and 20 cm in height was placed inside the tank with its top surface being 1.5 cm below the surface of the water. Four different visual cues were mounted on the four directions of the walls. A CCD camera was equipped with a computer for the behavior recording. For the acquisition test, each rat was trained three trials a day. The rat had to find the hidden platform for 180 s, and the escape latency was used as an index of performance in the task. For 4 consecutive days, rats were tested with three acquisition trials. On the 5th day, they received a 1 min probe trial in which the platform was removed from the pool. The performance of the test animals in each water maze trial was assessed by a personal computer for the behavioral analysis using a S-mart Program (PanLab, Barcelona, Spain).

For immunohistochemical examination, rats were anesthetized with sodium pentobarbital (100 mg kg^-1, intraperitoneally) and then perfused transcardially with heparinized phosphate-buffered saline (PBS; pH 7.4) for 30 s followed by 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4) for 10-15 min. The brains were post-fixed in the same fixative overnight, cryoprotected in 30% sucrose solution in PBS, embedded and serially sectioned on a cryostat (Leica Microsystem Co., Ltd., Wetzlar, Germany) at 30 μm thickness in the coronal plane and they were collected in PBS. The primary antibodies against the following specific antigen were used: cAMP response element-binding protein (rabbit polyclonal CREB, 1 : 500; Cell Signaling, Danvers, MA, USA) and BDNF responsive element binding protein (rabbit polyclonal BDNF, 1 : 500; Santa Cruz Biotechnology, Santa Cruz, CA, USA) and ChAT (sheep polyclonal ChAT, 1 : 2,000; Chemicon, Temecula, CA, USA). The primary antibodies were prepared and diluted in 0.3% PBST, 2% blocking serum and 0.001% keyhole limpet hemocyanin (Sigma, St. Louis, MO, USA). The sections were incubated in the primary antiserum for 72 h at 4℃. After three more rinses in PBST, the sections were placed in Vectastain Elite ABC reagent (Vector Laboratories, Burlingame, CA, USA) for 2 h at room temperature. Following a further rinsing in PBS, the tissue was developed using diaminobenzadine (Sigma) as the chromogen. The images were captured using a DP2-BSW imaging system (Olympus, San Diego, CA, USA) and they were processed using Adobe Photoshop. For counting the cells that were positive for CREB, BDNF, and ChAT, the 200 × 200 mm rectangle grid was placed on CA1 and CA3 in the hippocampus according to the method of Paxinos et al. (1985).

The Neuro2a cells were kindly provided by Korea Centers for Disease Control and Prevention, Dr. Kim from the Republic of Korea. The cells were grown in Dulbecco’s modified minimum essential medium (DMEM) supplemented with 10% (v/v) heat-inactivated fetal bovine serum under a 5% CO₂/95% humidified air at 37℃, and the cells were sub-cultured every 3-4 days. The medium was changed every 2-3 days.

The Neuro2a cells were plated onto 35-mm plastic culture dishes coated with collagen (type I from calf skin; Sigma) at a density of 5 × 10⁴ cells well^-1 and then they were incubated for 24 h at 37℃ in 5% CO₂. After 24 h, the culture medium was changed by serum-free DMEM medium. The cells were added with different concentrations of FSJ diluted in sterile PBS at pH 7.4 to attain the desired FSJ concentrations of 10 and 50 μg mL^-1. After 72 h, the cells were rinsed with PBS and then 1 mL of 0.1% glutaraldehyde solution was added and then the cells were incubated for 30 min in room temperature. The length of the longest neurite of the individual cells was measured by phase-contrast microscope image analyze system (Model DC-300F; Leica Microsystem Co., Ltd.) that was attached to a phase-contrast microscope (×200 magnification) and by using software (Qwin Plus 271; Leica Microsystem Imaging Solutions Co., Ltd., Cambridge, UK).

The behavioral and immunohistochemical data were statistically analyzed by one-way or two-way ANOVA with repeated measures. Tukey’s post hoc test was carried out to identify the differences among the means of groups (SPSS ver. 15.0; SPSS Inc., Chicago, IL, USA). The results are expressed as means ± standard error of mean (SEM). p-values less than 0.05 were considered statistically significant.

Effects of FSJ (50, 100, and 200 mg kg^-1) on learning and memory were evaluated using the Morris water maze test, in which learning process was tested during acquisition trials and memory aspect was tested in probe trials.

The escape latency of the control group was longer than that of the normal group during all sessions, and Fig. 1A shows the mean time latencies in reaching the hidden platform in the acquisition test of the Morris water maze for all the groups for 4 days. The control group showed a worse performance than did the normal group (p < 0.05 at the day 1 and p < 0.01 at the days 2 and 4). There were significant main effects, and treatment with FSJ (50 mg kg^-1) had a significant interaction effect on the escape latency of control group from the first day (p < 0.01).

[Fig. 1.] (A) The latency of escape onto the hidden platform during the Morris water maze. The task was performed with 3 trials per day during 4 days for the acquisition test. Data were plotted as mean ± standard error of the mean (SEM). FSJ, fermented Saccharina japonica. #p < 0.05, ##p < 0.01 vs. normal group and **p < 0.01 vs. control group, respectively. (B) Probe test was performed on the 5th day for 60 s. Data were plotted as ± SEM. ##p < 0.01 vs. normal group and *p < 0.05, ***p < 0.001 vs. control group, respectively.

In the probe test, FSJ treatment showed an increase in time spent around the platform compared to that of the control group (50 and 100 mg kg^-1, p < 0.05 and p < 0.001, respectively), reflecting enhancing effect of memory function by FSJ.

The expression of CREB immunoreactive cells per section from the different subregions of the hippocampus is shown in Fig. 2A and B. Statistical analysis indicated that the CREB immunoreactivity in the hippocampus of the control group was significantly lower than that of the normal group. In particular, significant differences were observed in the hippocampal CA1 (p < 0.05) and CA3 (p< 0.001). The CREB reactivity in the FSJ 50 mg kg^-1 group was higher than that of the control group (p < 0.05) in the CA1, where, the number of CREB positive neurons in the FSJ 50 mg kg^-1 group was significantly increased by 119.7% compared to that of the control group (p < 0.05) (Fig. 2).

[Fig. 2.] (A) The expression of cAMP responsive element binding protein (CREB) immunostained nuclei in different hippocampal CA1 and CA3 of the experimental groups. Each value represents the mean ± standard error of the mean. #p < 0.05, ###p < 0.001 vs. normal group and *p < 0.05 vs. control group, respectively. (B) Representative photographs showing the distribution of CREB-immunoreactive cells in the hippocampus of normal group, control group, fermented Saccharina japonica (FSJ) 50 mg kg-1 group, and FSJ 100 mg kg-1. Sections were cut coronally at 30 μm. Scale bar represents: 200 μm.

The expression of the evaluations of the BDNF immune-positive neurons per section from the different hippocampal formations is shown in Fig. 3A and B. Statistical analysis indicated that the BDNF immunoreactivity in the hippocampus of the control group was significantly lower than that of the normal group, in the hippocampal CA1 (p< 0.05) and CA3 (p < 0.01). The BDNF reactivity in the FSJ 50 mg kg^-1 group was higher than that of the control group (p < 0.01) in both CA1 and CA3. In the hippocampal CA3, the number of BDNF positive neurons in the FSJ 50 mg kg^-1 group was significantly increased by 404.5% compared to that of the control group (p< 0.01) (Fig. 3).

[Fig. 3.] (A) The expression of brain derived neurotrophic factor (BDNF) immunoreactive neurons in different hippocampal CA1 and CA3 of the experimental groups. Data were plotted as represents the mean ± standard error of the mean. #p < 0.05, ##p < 0.01 vs. normal group and **p <0.01 vs. control group, respectively. (B) Representative photographs showing the distribution of BDNF-immunoreactive cells in the hippocampus of normal group, control group, fermented Saccharina japonica (FSJ) 50 mg kg-1 group, and FSJ 100 mg kg-1. Sections were cut coronally at 30 μm. Scale bar represents: 200 μm.

The expression of the ChAT immunoreactive neurons per section from the different hippocampus is shown in Fig. 4A and B. Statistical analysis indicated that the ChAT activity in the hippocampus of the control group was significantly lower than that of the normal group in the hippocampal CA1 (p < 0.01) where the number of ChAT positive neurons in the FSJ 100 mg kg^-1 group was significantly increased by 280.0% compared to that of the control group (p < 0.01) (Fig. 4).

[Fig. 4.] (A) The expression of choline acetyltransferase (ChAT) immunoreactive neurons in different hippocampal CA1 and CA3 of the experimental groups. Data were plotted as the mean ± standard error of the mean. ##p < 0.01 vs. normal group and **p < 0.01 vs. control group, respectively. (B) Representative photographs showing the distribution of ChAT-immunoreactive cells in the hippocampus of normal group, control group, fermented Saccharina japonica (FSJ) 50 mg kg-1 group, and FSJ 100 mg kg-1. Sections were cut coronally at 30 μm. Scale bar represents: 200 μm.

The neuritogenic effect of the FSJ water extract powder in the Neuro2a cells is shown in Fig. 5. Neurite outgrowth in the Neuro2a cells was significantly increased with treatment of FSJ. The higher doses of FSJ were given to the cells, the greater the increases of neurite outgrowth were observed (p < 0.01 and p < 0.001 in the 10 and 50 μg mL^-1 of FSJ-treated group, respectively). Also, cells treated with nerve growth factor (NGF; 1.0 μg mL^-1) showed significantly increased neurite outgrowth (p < 0.01). Fig. 5B shows the morphology of the Neuro2a cells with neurite outgrowth at 24 h of treatment with NGF (1 μg mL^-1), FSJ extract (10 and 50 μg mL^-1), and vehicle (Fig. 5).

[Fig. 5.] (A) Neurite outgrowth induced by treatment with the fermented Saccharina japonica (FSJ) water extract powder on Neuro2a cells. Data were plotted as the mean ± standard error of the mean of the length of the longest neurite of the individual cells (n = 50) in each group. Separate measure of one-way ANOVA of the length of neurite. *p < 0.05, **p < 0.01, ***p < 0.001 vs. control (vehicle). (B) Phase-contrast photomicrographs of FSJ extract powder and nerve growth factor (NGF)-induced neurites.

The present results demonstrated that TMT injections produced severe deficits in the rats’ performances in a Morris water maze along with signs of neuronal degeneration, including decreased cholinergic neurons and CREB activity in the hippocampus. Repeated treatment with FSJ prevented the TMT-induced learning and memory impairments as shown in the water maze test and it had a neuroprotective effect against the TMT-induced decrease in CREB and BDNF positive neurons. Also, FSJ stimulated transcription and neurite outgrowth of the PC12h cells (Fortemps et al. 1978, Brown et al. 1979, Chang and Dyer 1983a, 1983b, 1985, Chang et al. 1983a, 1983b, Balaban et al. 1988, Koczyk 1996).

TMT induced profound behavioral and cognitive deficits in animals (Koczyk 1996, Ishida et al. 1997). Our previous studies already reported that intoxication with TMT-induced deficits in learning and memory (Park et al. 2012a, 2012b). Other studies proved the loss of learning and memory in TMT-intoxicated rats (Swartzwelder et al. 1982, Alessandri et al. 1994). The deficit in spatial learning produced by TMT in the current study is consistent with the previous studies of spatial learning impairments (Walsh et al. 1982, Zhao et al. 1982, Hagan et al. 1988, Earley et al. 1992, Alessandri et al. 1994, Woodruff et al. 1994, Koczyk et al. 1996). However, this current study proved that spatial memory in the FSJ groups was preserved during the training days compared to that of the control group. Also, the data of the spatial probe trial demonstrated that FSJ protects against the TMT-induced definite cognitive impairment, especially with regards to long-term memory. And, repeated FSJ administration proved beneficial in improving associative learning and memory processes as assessed by the Morris water maze test. Treatment with FSJ extract produced increased behavioral effects, but was without effects on it in the present study. It is not clear why this high dosage did not produce any behavioral effects. These U-shape or bidirectional patterns of dosage have been frequently observed from in vivo studies using natural products. According to analysis of ingredients, S. japonica contains approximately 55 mg of GABA g^-1. Therefore 11 mg of GABA was injected to 200 mg kg^-1 dosage group. GABA is a potent inhibitory neurotransmitter in the central nervous system and this amount of GABA may inhibit behavioral effects. The other reason may be due to a toxic effect from overdose of natural extracts.

The activation of CREB was known to be associated with a neuronal protective effect such as a defense mechanism. CREB is important for activating the transcription of genes controlled by the cAMP-response element, and many of these genes may be involved in neuronal growth and plasticity and they may may play a significant role in neuronal survival (Sala et al. 2000, Lonze and Ginty 2002). Several studies have shown that disruption of the CREB gene leads to neuronal degeneration (Bourtchuladze et al. 1994, Sala et al. 2000, Lonze and Ginty 2002, Mantamadiotis et al. 2002). CREB is also known to be a key molecular marker of long term potentiation and memory consolidation. Genetic and pharmacological studies have provided strong evidence that the CREB signaling pathway is necessary for learning and memory across species (Kogan et al. 1997, Jackson and Ramaswami 2003). The present study proved that the levels of CREB in the hippocampus were significantly different among the groups (Park et al. 2012a, 2012b). The present study demonstrated that the TMT treated group showed a reduction of approximately 40% of the CREB activity. Thus, we may draw a conclusion that the CREB loss after TMT exposure might be at least partially responsible for the TMT-induced cell death and TMT played a role as a stressor on the CREB gene. A previous study reported the protective effect of FSJ in the Caco-2 and Raw 264.7 cell lines (Kuda et al. 2012, Küpper et al. 2013). A clinical study also proved that fermented FSJ enhanced the antioxidant defense system in a healthy population and may be useful as a functional food ingredient (Kang et al. 2012). In this study, we also found the protection of TMT-induced CREB loss in the repeatedly treated FSJ groups.

BDNF, a member of the neurotrophin family, plays a key role in neuronal survival and differentiation and synaptic plasticity (Thoenen 1995, Mizuno et al. 2000, Lipsky and Marini 2007, Lu et al. 2008). The beneficial effect on learning and memory could have been mediated by the increased synaptic plasticity (Cotman and Berchtold 2002, Farmer et al. 2004), neurotransmission and growth factor expression (Leibrock et al. 1989) that were observed in the rats. Many studies have shown that hippocampal BDNF expression was increased during process of spatial learning(Korte et al. 1995, Gooney et al. 2002), and BDNF knockout mice produced deficits in spatial learning and synaptic plasticity (Jones and Clemmons 1995, Heldt et al. 2007). Reduction in the synthesis or availability of central neurotrophic factors such as BDNF might play a critical role in the aging and degenerative processes. Also, in this study, we have focused on the role of BDNF as one of the potential mediators of these effects. In this study, there was a 204-404% increase in the number of BDNF positive cells in the hippocampus of the FSJ 50 mg kg^-1 treated rats, improving their learning and memory.

The rat pheochromocytoma Neuro2a cell line has been largely used as a model for studying neuronal differentiation. Neuro2a cells respond to neutrophic factors, such as NGF and fibroblast growth factor, by differentiating into sympathetic neuron-like phenotypes that are characterized by neurite outgrowth and the expression of many neuron-specific proteins. In the present study, treatment with the FSJ water extract powder significantly increased neurite outgrowth in the Neuro2a cells in a dose dependent manner. These findings support the hypothesis that FSJ improves the functioning of the neurotrophic system.

In conclusion, treatment with FSJ reduced the TMT-induced learning and memory deficits as shown with the Morris water maze test, and FSJ treatment had a neuroprotective effect against a TMT-induced decrease of the BDNF, CREB, and ChAT immunoreactivity. Treatment with FSJ also significantly increased the neurite outgrowth of the PC12h cells. Thus, our findings indicate that FSJ is a good candidate for further investigations that may ultimately result in its clinical use.