Hippocampal memory enhancing activity of pine needle extract against scopolamine-induced amnesia in a mouse model

We evaluated the neuropharmacological effects of 30% ethanolic pine needle extract (PNE) on memory impairment caused by scopolamine injection in mice hippocampus. Mice were orally pretreated with PNE (25, 50, and 100 mg/kg) or tacrine (10 mg/kg) for 7 days, and scopolamine (2 mg/kg) was injected intraperitoneally, 30 min before the Morris water maze task on first day. To evaluate memory function, the Morris water maze task was performed for 5 days consecutively. Scopolamine increased the escape latency and cumulative path-length but decreases the time spent in target quadrant, which were ameliorated by pretreatment with PNE. Oxidant-antioxidant balance, acetylcholinesterase activity, neurogenesis and their connecting pathway were abnormally altered by scopolamine in hippocampus and/or sera, while those alterations were recovered by pretreatment with PNE. As lipid peroxidation, 4HNE-positive stained cells were ameliorated in hippocampus pretreated with PNE. Pretreatment with PNE increased the proliferating cells and immature neurons against hippocampal neurogenesis suppressed by scopolamine, which was confirmed by ki67- and DCX-positive stained cells. The expression of brain-derived neurotrophic factor (BDNF) and phosphorylated cAMP response element-binding protein (pCREB) in both protein and gene were facilitated by PNE pretreatment. These findings suggest that PNE could be a potent neuropharmacological drug against amnesia, and its possible mechanism might be modulating cholinergic activity via CREB-BDNF pathway.

result in damage to proteins, lipids, and nucleic acids in patients with neurodegenerative disorders 11,12 . Brain tissue is extremely sensitive to oxidative stress due to its high oxygen consumption, iron content, polyunsaturated fatty acids, and low antioxidant capacity 13,14 . Moreover, the hippocampus and amygdala are more sensitive to oxidative injury 15 , and excessive oxidative stress can lead to memory deficits by impairing hippocampal synaptic plasticity 16 .
Among numerous herbal plants, the pine needle (Pinus densiflora Sieb & Zucc.) is commonly used as an herbal medicine and health supplement in East-Asian countries, such as Korea and China 17 . The pine needle is utilized as an infusion of tea and supplementary health food, and is reported to be helpful in the treatment of patients with cancer, coronary heart diseases and neurodegenerative disease [17][18][19] . Despite the many uses of pine needle, there have been no studies of its neuropharmacological activities. Thus, in the present study we investigated the anti-amnesic effects of pine needle extract (PNE) on memory deficits in a mouse model of cognitive impairment caused by scopolamine.

Results
Compounds present in the 30% ethanolic pine needle extract. Four different kinds of flavonoids including catechin, quercetin dehydrates, astragalin, and kaempferol, when PNE sample were subjected to the HPLC. Among them, only catechin and astragalin were detected at 19.01 min and 31.34 of retention time (see Supplementary Fig. S1A and B online), and the concentration of catechin and astragalin were 0.35 6 0.01 and 2.68 6 0.04 respectively (see Supplementary Fig. S1C online). GC-MS data confirmed that terpenoids of a-pinene or b-pinene were not present in PNE (data not shown).
Effect on anti-amnesia in Morris water maze task. Scopolamine injection significantly prolonged the escape latency time and cumulative path length compared with the naïve group on fourth day of the acquisition period (approximately 3-and 4-fold respectively, P , 0.001). The time spent in the target quadrant was significantly shortened in the control group compared with naïve group on fifth day (P , 0.001). Pretreatment with PNE significantly ameliorated the escape latency (P , 0.001 for all groups) and the path-length (P , 0.001 for all groups; Fig. 1A and B) prolonged by scopolamine. On day five, PNE pretreatment groups exhibited significantly increased time spent in the target quadrant compared with the control group (P , 0.001 for both 25 and 50 mg/ kg, P , 0.01 for 100 mg/kg; Fig. 1C). Tacrine had effects similar to those of PNE pretreatment.
Effects on ROS levels in serum and hippocampus. Scopolamine injection significantly increased the ROS levels in both serum and hippocampal tissue compared with the naive group (approximately 1.6-fold, P , 0.001 in both samples). Pretreatment with PNE significantly decreased ROS levels in serum and hippocampal tissue compared with the control group (P , 0.001 for all groups; Fig. 2A and B); similar effects were observed in the tacrine group.  Effects on NO and MDA level in hippocampus. NO level in hippocampal tissue was increased significantly by scopolamine injection compared with the naive group (approximately 1.5-fold, P , 0.001), while pretreatment with PNE significantly attenuated the increase of NO level (P , 0.001 for all groups; Fig. 2C).
MDA level was increased significantly by scopolamine injection in hippocampal tissue compared with the naive group (approximately 3.9-fold, P , 0.001), while pretreatment with PNE significantly attenuated the elevation of MDA level compared with the control group (P , 0.01 for 25 mg/kg, P , 0.001 for both 50 and 100 mg/kg; Fig. 2D). Tacrine had effects similar to those of PNE pretreatment.
Effects on antioxidant biomarker profiles in the hippocampus. Scopolamine injection significantly depleted the antioxidant activities in hippocampal tissue, including TAC capacity (P , 0.001), GSH content (P , 0.001), activity of GSH-Rd (P , 0.001), GST (P , 0.01), SOD (P , 0.05), and catalase activity (P , 0.001), compared with the naïve group. These changes were reversed by PNE pretreatment and for all biomarkers, 50 and 100 mg/kg PNE pretreatment were more effective than 25 mg/kg PNE treatment. Interestingly, SOD activity in response to PNE pretreatment was further enhanced compared with the naïve group. In contrast, tacrine had different effects than PNE, with the exception of TAC capacity (Table 2).
Lipid peroxidation, cell proliferation and neurogenesis in the hippocampus. Scopolamine injection significantly induced the lipid peroxidation in hippocampus (P , 0.01 in DG, P , 0.05 in CA3), shown as a deep red color in the dentate gyrus (DG) and cornu ammonis 3 (CA3) regions by 4HNE staining. Pretreatment with PNE significantly attenuated 4HNE-positive staining in the DG including granular cell layer (GCL), subgranular zone (SGZ) (P , 0.05 for 25 mg/kg, P , 0.01 for 100 mg/kg) and CA3 compared with the control group (P , 0.05 for both 25 and 100 mg/kg; Fig. 3A-C).
Scopolamine injection significantly inhibited the production of new granule cells in the hippocampal DG region, particularly in the SGZ (P , 0.05), as evidenced by reduced Ki67 staining. These alterations were significantly attenuated by PNE pretreatment with    notable replacement of proliferating cells with Ki67-positive staining in the SGZ compared with the control group (P , 0.05 for 50 mg/kg; Fig. 4A and B). Scopolamine injection significantly suppressed the adult neurogenesis, shown as a distributed dendrites and neuron bodies in the DG region by DCX staining, particularly in the SGZ (P , 0.05). Pretreatment with PNE completely ameliorated the adult neurogenesis by enhancing immature neurons in the SGZ compared with the control group (P , 0.05 for all groups; Fig. 5A and B). In contrast, tacrine ameliorated these alterations only in DCX staining result.
Effect on AChE activity in hippocampus. AChE activity in hippocampal tissue was significantly increased by scopolamine injection compared with the naïve group (approximately 3.8-fold, P , 0.001), while pretreatment with PNE completely inhibited the hyper-activation of AChE compared with the control group (P , 0.001 for all groups; Fig. 6A). This effect was also seen with tacrine in the hippocampus.
Western blot analysis of CREB/BDNF in the hippocampus. Scopolamine injection slightly reduced the BDNF and phospho-CREB protein levels in the hippocampus. In contrast, pretreatment with PNE increased BDNF and phospho-CREB levels compared with the control group. Tacrine had an effect on the expression of BDNF similar to that of PNE (Fig. 6B).
Changes in mRNA levels in the hippocampus. The CREB1 (P , 0.001), mAChR1 (P , 0.01), BDNF (P , 0.05) and CBP (P , 0.01) mRNA levels were significantly down-regulated by scopolamine injection compared with the naïve group in the hippocampus. This down-regulation was ameliorated significantly by PNE pretreatment compared with the control group (P , 0.01 or P , 0.001 for all groups; Fig. 6C). Moreover, scopolamine injection significantly upregulated the expression of iNOS in the hippocampus compared with the naïve group (approximately 1.6-fold, P , 0.001), while this upregulation of iNOS was significantly ameliorated by pretreatment with PNE (P , 0.001 for all groups; Fig. 6C). Similar effects were observed in the tacrine group.

Discussion
Neurodegenerative disorders are becoming a major healthcare issue as they result in progressive brain dysfunction among the aging population in industrialized countries. The major pathophysiological features involve alterations of the central cholinergic system, including cholinergic neurons, neurotransmitters and their respective receptors 20 . Several inhibitors of AChE, as well as N-methyl-Daspartate (NMDA) receptor antagonists, are available currently for the treatment of Alzheimer's disease at mild and moderate stages, even though the mechanisms mediating these processes remain to be fully elucidated and sometimes theses drugs show adverse effects 21 . Multitarget-directed ligands and natural therapeutic agents are recently drawing attention as complementary and alternative medicine for neurodegenerative disorders 22 .
Herein, we investigated the neurotherapeutic effects of PNE on memory deficits in a mouse model of amnesia induced by scopolamine injection. To evaluate the effects of PNE on spatial learning and memory, a Morris water maze task was performed 23 . Following 4 days of acquisition training, the scopolamine treated group had prolonged escape latency (3-fold) and cumulative path-length (4-fold). This result was consistent with time spent in the target quadrant, which was reduced compared with the naïve group. As expected, pretreatment with PNE ameliorated delayed escape latency ( Fig. 1A-C). These findings suggest that PNE has an anti-amnesic effect in the scopolamine-induced model.
We examined the antioxidant effects of PNE in the hippocampus since oxidative stress contributes to pathogenesis and histological changes in patients with neurodegenerative disorders 24 . The excessive production of ROS triggers neurotoxic activity through thiol-and lipid-dependent mechanisms in cell membrane 25 . The scopolamineinduced memory deficit model exhibits prominent oxidative stress and memory deficits, although the mechanism of oxidative damage in brain tissue remains unclear 26,27 . The hippocampus plays a crucial role in short-and long-term memory, as well as spatial memory 28 , and is highly susceptible to oxidative stress 29 . As expected, scopolamine injection induced oxidative stress in hippocampus, as evidenced by increased levels of ROS, NO and MDA. These alterations were attenuated significantly by PNE pretreatment ( Fig. 2A-D). Previous studies showed that scopolamine treatment elevated the production of MDA in the cerebral cortex and hippocampus 30 . These results were consistent with changes in iNOS gene expression in the hippocampus (Fig. 6C).
To protect tissues against oxidative damage, cellular organisms maintaining an antioxidant system composed of non-enzymatic and enzymatic components. In our study, scopolamine injection significantly depleted antioxidant capacity of TAC, SOD, catalase and the GSH-redox system in the hippocampus, whereas these alterations were ameliorated significantly by pretreatment with PNE (Table 2). These results suggested that the antioxidant potential of PNE contributed to neuronal plasticity and memory function. Furthermore, in the DG and CA3 region of the hippocampus, scopolamine induced lipid peroxidation, shown as positively stained 4HNE cells in the GCL and SGZ. However, pretreatment with PNE completely attenuated the over-production of 4HNE (Fig 3). Increased 4HNE is a key histopathological feature of various diseases, including Alzheimer's disease, diabetes and cancer 31 . In the hippocampus, dentate granule cells are responsible for the separation of each episodic memory, such as object and place, and the CA3 pyramidal neurons receive and store auto-associative memories in their network 32 .
The new neurons are generated as neural progenitor cells in the SGZ of hippocampal DG, in the adult mammalian brain, including humans 33 . As well known, adult hippocampal neurogenesis plays a crucial role in hippocampal memory function 34 . Accordingly, enhancing hippocampal neurogenesis may be an efficient therapeutic target for Alzheimer's disease. We confirmed the significantly decreased proliferating cells and immature neurons by Ki67 and DCX staining of the scopolamine-induced mouse hippocampus, while pretreatment with PNE markedly ameliorated repression of reproducing cell and neuronal precursor cells in the SGZ (Fig. 4 and 5). These findings can predicts the production of new granule cells, maturation of neurons as well as hippocampal neurogenesis. The neurogenesis and gliogenesis are regulated by various molecular factors, such as neurotransmitters, neurotrophins, hormones and their respective signaling pathways 35 .
To investigate the role of PNE in neurogenesis, we examined cholinergic activity, neurotransmitter levels and subsequent signaling pathways. Scopolamine injection augmented AChE activity 4-fold in hippocampal tissue, whereas pretreatment with PNE completely attenuated the excessive activation of AChE (Fig. 6A). Maintenance of ACh levels is necessary for normal memory function, but excessive AChE activity leads to disruption of ACh in hippocampal cholinergic synapses 36 . In the synaptic cleft, ACh binds to post-synaptic mAChRs and the synaptic signal communicates sequentially to cyclic adenosine monophosphate/protein kinase A (cAMP/PKA)-CREB signaling pathway via G-coupled protein receptors 37 . In our result, the downregulation of mAChR-1 gene expression by scopolamine in hippocampal tissues was normalized significantly by PNE pretreatment (Fig. 6C). Previous studies have revealed that patients with Alzheimer's disease have down-regulated expression of mAChR-1 38 , which correlates with memory disruption in a mAChR-1 knockout mouse model 39 .
In contrast, the transcription factor CREB is essential for memory and synaptic plasticity in the CNS. Disturbance of phosphorylated CREB within the hippocampal region leads to the progression of neurodegenerative diseases such as Alzheimer's disease, Parkinson's disease and Huntington's disease 40 . Previous studies indicated that CREB has a neuroprotective effect against ROS-mediated cell toxicity 11 , and activation of CREB ameliorated cognitive impairment via the cholinergic system 8 . In our study, hippocampal phospho-CREB was reduced by scopolamine injection, which was increased by pretreatment with PNE (Fig. 6B). The expression of CREB1 and CBP, a co-factor of CREB, supported these results (Fig. 6C). BDNF is known to improve learning ability and neurogenesis via phospho-CREB signaling, and mice deficient in BDNF and pro-BDNF exhibit an aging process characteristic of Alzheimer's disease 41,42 . In our study, pretreatment with PNE markedly increased BDNF protein levels (Fig. 6B), which was correlated with BDNF mRNA levels in hippocampal tissue (Fig. 6C).
Pinus densiflora, also called Japanese Red Pine, grows naturally in East Asian countries and has been used traditionally as a supplement in various foods and folk medicine. Previous studies well evidenced the pine needle has potent antioxidant effects against oxidative damage and apoptosis, with anti-obesity and anti-cancer effects 17,43,44 . To date, several active compounds have been isolated from organic-solvent extractions, including flavonoids, terpenoids and tannin 45,46 . Among them, the flavonoid-derived compounds are well-known for enhancement of memory function via stimulating neurogenesis, regulating neurotrophins, and protecting neurons against oxidative and metabolic stress 47,48 . In this study we observed the chemical characters of PNE via performing HPLC analysis using four flavonoid-derived compounds including catechin, astragalin, quercetin dehydrate, and kaempferol. As a result of HPLC analysis, catechin and astragalin were detected in PNE, and we further explored the quantitative analysis of those compounds (Fig. S1A-C). It was well corresponded to the previous studies that both catechin and astragalin showed potent neuroprotective effects on the various brain damage model. Catechin showed its protective effects on the severe brain oxidative damage via enhancement of antioxidant capacities as well as especially inactivation of NF-kB in pyramidal cells of hippocampal CA1 region 49 . On the other hand, the astragalin prevented brain ischemic injury through anti-apoptotic effect and decreases of neuronal cell adhesion molecules 50 .
In conclusion, our results demonstrate that PNE has an antiamnesic effect, which could be mediated by hippocampal neurogenesis, cholinergic activity, CREB-BDNF pathway and its antioxidant properties. This suggests that the Pinus densiflora leaf may be a promising treatment for patients with neurodegenerative disorders. Further study should aim to confirm the neuroprotective effects and its corresponded mechanisms using active compounds such as catechin and astragalin.

Material and Methods
Materials. Pine needles (Pinus densiflora Sieb & Zucc.) were collected from Guryong-Mountain in Chungwon-gun, Chungbuk, Korea. Fifty grams of dried pine needles were extracted with 0.5 L of 30% ethanol at 25uC for 72 h, and suspension was filtered using a 300-mesh 50-mm filter paper (Advantec, Toyo Roshi Kaisah, Tokyo, Japan). Filtrate was concentrated in a rotary evaporator and lyophilized. The final yield of the extract was 10.46% w/w, and it was stored at 280uC. Fingerprinting analysis of PNE was measured by high-performance liquid chromatography (HPLC) and gas chromatography/mass spectrometry (GC/MS).
The dose of PNE was determined based on prescreening results. The protocol was approved by the Institutional Animal Care and Use Committee of Daejeon University (DJUARB 2013-031), and was conducted in accordance with the Guide for the Care and Use of Laboratory Animals published by the United States National Institutes of Health (NIH).
Morris water maze task. Morris water maze task was performed in a circular pool (100-cm diameter 3 50-cm height) with a circular acrylic platform (10-cm diameter 3 35-cm height). The location of platform can be discriminated by visual cue. The pool was filled with milk water (22 6 1uC) and divided into equal quadrants. A platform was placed in one of the quadrants ,1 cm below the surface as described previously 51 . Data were recorded using a video camera connected to the corresponding software (Smart Junior, Panlab SL; Barcelona, Spain).
Mice were placed on the platform for 10 s, and removed from the pool. Mice were given acquisition trial for 4 days. The escape latency and cumulative path-length were recorded during each acquisition trial. On day 5 th , all mice were subjected to probe trial without platform, and were recorded for 120 s. The time spent in the target quadrant was measured for spatial learning and memory.
Sample preparation. All mice were sacrificed under ether anesthesia after 1 h following the Morris water maze. Blood was collected by the method followed by serum. Serum was collected by centrifugation at 3000 rpm for 15 min at 4uC, and the hippocampal region from the whole brain was isolated immediately, and then samples were stored at 280uC or in RNAlater (Ambion, TX, USA). In each group, three mouse brains were fixed in 4% paraformaldehyde, and hippocampi of remaining seven mice was used for biochemical analysis, western blot and real-time PCR analysis. The part of hippocampus was homogenized on ice using RIPA buffer, and other part of hippocampus was used for isolation of RNA.
ROS levels. The levels of ROS in serum and hippocampus were determined as described previously 52 . Absorbance was measured at 505 nm using a UV spectrophotometer (Molecular Devices; Sunnyvale, CA, USA). The results were calculated using H 2 O 2 as a standard and expressed as unit/mL.
NO and MDA levels. The level of nitric oxide (NO) in hippocampus was determined using the Griess method 53 . The purple azo dye product was measured at 540 nm using a UV spectrophotometer.
The level of lipid peroxidation in hippocampus was determined by measuring malondialdehyde (MDA) using the thiobarbituric acid reactive substance (TBARS) method as described previously 54 . Absorbance at 535 and 520 nm was measured using a spectrophotometer. The concentration of TBARS was calculated using a standard TEP and the results were expressed as mmol/mg protein.
TAC capacity. Total antioxidant capacity (TAC) in hippocampus was determined as described previously 55 . The level of TAC was expressed as the gallic acid equivalent antioxidant capacity (GEAC). The absorbance at 600 nm was measured using a UV spectrophotometer.
Total GSH content, GSH-Rd and GST activities. Glutathione (GSH) content in hippocampus was determined as described previously 56 . Absorbance at 405 nm was measured using a UV spectrophotometer.
Glutathione S-transferase (GST) activity in hippocampus was determined using a GST assay kit (Sigma, St. Louis, MO, USA) according to the manufacturer's protocol. Absorbance at 340 nm was measured using UV spectrophotometer. Enzyme activity was calculated using the following formula: Enzyme activity: (unit/mL) 5 [(DA340)/min 3 0.2 3 dilution factor]/(5.3 mM -1 cm 21 3 volume of enzyme sample tested).
SOD and catalase activities. Superoxide dismutase (SOD) activity in hippocampus was determined using a SOD assay kit (Dojindo Laboratories; Kumamoto, Japan) according to the manufacturer's protocol. Absorbance was measured at 450 nm using UV spectrophotometer. Dilutions of bovine erythrocyte SOD ranging from 0.01-50 unit/mL were used as standards.
Catalase activity in hippocampus was determined as described previously 58 . Absorbance of the purple formaldehyde product at 550 nm was measured using a UV spectrophotometer.
Immunohistochemical stain analysis. Immunohistochemical analysis was performed by assessing cell proliferation with Ki67, neurogenesis with DCX and lipid peroxidation with 4HNE staining in hippocampus. Three brain tissues of each group were immersed in the fixative solution for 4 h. The brain was cryoprotected in 30% sucrose, embedded in tissue-freezing medium with liquid nitrogen, and cut into coronal frozen sections (35 mm) using a Leica CM3050 cryostat. Sections were stored under anti-freeze buffer.
AChE activity. Acetylcholinesterase (AChE) activity in hippocampus was determined using a AChE activity assay kit (AAT Bioquest; Sunnyvale, CA, USA) according to the manufacturer's protocol. Absorbance at 410 nm was measured using a UV spectrophotometer.
Western blot analysis. The expression of BDNF and CREB/phospho-CREB proteins in hippocampus was evaluated by Western blot. The proteins from homogenates were separated by 10% polyacrylamide gel electrophoresis and transferred to polyvinylidene fluoride (PVDF) membranes. After blocking in 5% skim milk, the membranes were probed overnight at 4uC with primary antibodies (CREB, p-CREB, BDNF and b-actin). The membranes were washed and incubated for 2 h with HRPconjugated anti-rabbit antibody. Western blots were visualized using an enhanced chemiluminescence (ECL) advanced kit.
Quantitative real-time PCR analysis. The mRNA expression of genes encoding CREB1, mAChR1, BDNF, CREB-binding protein (CBP), and inducible nitric oxide synthase (iNOS) in hippocampus was measured by real-time PCR. Total RNA was isolated from hippocampus using an RNeasy Mini Kit (QIAGEN, Valencia, CA, USA) and cDNA synthesized using a High-Capacity cDNA reverse transcription kit (Ambion, Austin, TX, USA). Real-time PCR was performed using SYBRGreen PCR Master Mix (Applied Biosystems; Foster City, CA, USA) and PCR amplification was performed using a standard protocol with the IQ5 PCR Thermal Cycler (Bio-Rad, Hercules, CA, USA). Information of primers was summarized in Table 1.