Effects of NK-4 in a Transgenic Mouse Model of Alzheimer's Disease

Beta-amyloid (Aβ) peptides are considered to play a major role in the pathogenesis of Alzheimer's disease (AD) and molecules that can prevent pathways of Aβ toxicity may be potential therapeutic agents for treatment of AD. We have previously reported that NK-4, a cyanine photosensitizing dye, displays neurotrophic and antioxidant activities. In this study, we report the effects of NK-4 on the toxicity of Aβ and on cognitive function and Aβ concentration in a transgenic mouse model of AD (Tg2576). In vitro, NK-4 effectively protected neuronal cells from toxicity induced by Aβ. In addition, it displayed profound inhibitory activities on Aβ fibril formation. In vivo, Tg2576 mice received an intraperitoneal injection at 100 or 500 µg/kg of NK-4 once a day, five times a week for 9 months. Administration of NK-4 to the mice attenuated impairment of recognition memory, associative memory, and learning ability, as assessed by a novel object recognition test, a passive avoidance test, and a water maze test, respectively. NK-4 decreased the brain Aβ concentration while increasing the plasma amyloid level in a dose-dependent manner. NK-4 also improved memory impairments of ICR mice induced by direct intracerebroventricular administration of Aβ. These lines of evidence suggest that NK-4 may affect multiple pathways of amyloid pathogenesis and could be useful for treatment of AD.


Introduction
Alzheimer's disease (AD) is the most common cause of dementia in elderly people and affects more than 25 million individuals worldwide. The development of therapeutic strategies for AD is a major unmet medical need. Inhibition of acetylcholinesterase (AChE) is currently a primary therapeutic strategy for the mid to late phase of AD because restitution of a close to normal acetylcholine concentration in the synaptic cleft to enhance cholinergic neurotransmission could ameliorate symptoms of AD. However, the therapeutic effect of AChE inhibitors is only symptomatic and short-term, and therefore new treatments are needed as disease-modifying therapy for AD.
Elevation of the levels of b-amyloid (Ab), a ,4 kDa secreted polypeptide, is thought to be involved in the pathogenesis of AD because mutations in the Ab precursor protein (AbPP) or presenilin genes that lead to increased production of Ab are causative for familial AD [1][2][3][4][5][6][7][8]. Elevated levels of Ab are also well correlated with cognitive decline in the early phase of dementia [9]. This indicates that suppression of Ab in the brains of patients in the early phase of dementia is a primary therapeutic target. The majority of new treatments aim to inhibit Ab toxicity. These include Ab immunotherapies that enhance clearance of accumulating Ab in brain and secretase inhibitors that inhibit production of Ab from AbPP. Inhibitors of Ab aggregation are also currently under development, because the secondary structure determines several important properties of Ab that may be relevant to the pathogenesis of AD. The Tg2576 mouse model of AD expresses a high level of the Swedish mutation of AbPP under control of the hamster prion protein promoter, which leads to elevated Ab production and accumulation in brain [10,11]. The Tg2576 strain does not exhibit aggressive neuronal degeneration, but the mice display agedependent cognitive dysfunction [10,[12][13][14] and are therefore an excellent tool for evaluation of Ab-targeted pharmacological agents.
NK-4 is a divalent cationic pentamethine trinuclear cyanine dye that contains three quinolinium rings, short N-alkyl side chains (C2) and two iodine anions. NK-4 has a variety of biological activities, including antimicrobial [15], macrophage-activating [16], and anticancer properties, and is used as an immunomodulator in treatment with antiviral and anticancer agents [17]. NK-4 also has anti-inflammatory properties [18] and has been used to treat allergy [19].
Recently, we found that NK-4 was a potent neurotrophic agent for promotion of growth and differentiation of neuronal PC12 cells. The neuroprotective effects of NK-4 are mediated by activation of PI3K-Akt and inactivation of SAPK/JNK independently of the TrkA receptor at nanomolar concentrations [20]. NK-4 is also a potent scavenger of hydroxyl radicals (IC 50 : 7.6 mM), peroxy radicals (IC 50 : 5.2 mM), and superoxides (IC 50 : 89.3 mM), and these activities are greater than those of ascorbate [21]. In vivo, NK-4 has been shown to be effective for cerebral injury in a rat model of middle cerebral artery occlusion [21] and for motor discoordination and cerebellar atrophy in a genetic animal model of cerebellar ataxia, without causing significant adverse reactions [20].

Thioflavin T (ThT) Fibril Formation Assay
The Ab fibril formation assay was performed as described previously [23]. Stock solutions of Ab 1-40 , Ab 1-42 and Ab [25][26][27][28][29][30][31][32][33][34][35] were prepared in DMSO (1 mg/ml) and stored at 280uC. Just before the experiment, the stock solution was diluted with 10 mM phosphate buffer containing 100 mM KCl (pH 7.4). The diluted Ab solutions were incubated at 37uC in the presence or absence of indicated concentrations of NK-4 for 72 hr. Fifty ml of the incubated samples containing 100 mM Ab peptides were added to 450 ml of 10 mM ThT in 10 mM phosphate and 100 mM KCl (pH 7.4). Fluorescence intensity was measured with excitation at 435 nm and emission at 485 nm using a Hitachi 650-60 fluorescence spectrophotometer. The results are expressed as the mean 6 SD of three samples.

Electron Microscopy
Amyloid fibril formation was verified by electron microscopy (EM) images of negatively stained samples. A 20-ml sample of Ab  or Ab 1-42 solution incubated with or without NK-4 was placed on a polyvinyl butyral-coated copper mesh grid (Oukenshoji, Tokyo, Japan). The sample was allowed to stand for 30-60 s and excess solution was washed away. Subsequently, samples were negatively stained with 2% (w/v) uranyl acetate (Kowayakuhin, Saitama, Japan) and allowed to dry. After staining, the samples were viewed with a JEM-100CX/II electron microscope (Joel Ltd., Tokyo, Japan) at 80 kV.

Animals
All animal protocols were approved by the Institutional Animal Care and Use Committee of Hayashibara Biochemical Laboratories (permit number F-1002) and were conducted in accordance with the guidelines for the Care and Use of Laboratory Animals at Hayashibara Biochemical Laboratories. Female AbPP transgenic mice (Tg2576, 129S6/SvEvTac background) expressing a double mutation of human AbPP [24], and age-and sex-matched nontransgenic mice were obtained from Taconic (Germantown, NY). The mice were housed individually with supplied nesting materials, and food and water were provided ad libitum. NK-4 solution was injected intraperitoneally at a dose of 100 or 500 mg/ kg once a day, five times a week for 9 months, beginning at 3 months of age. Control mice received 200 ml of saline. During the behavioral test periods, NK-4 or saline was administered 2,3 hr before the start of each test. Male ICR mice (5 weeks of age, weighing 25-30 g) were purchased from Charles River Japan (Kanagawa, Japan).

Novel Object Recognition Test
The task was carried out according to a previous report [26] with minor modification. The apparatus used in this study was an opaque plastic box (40650630 cm) with a clear acrylic cover plate on the top, the floor of which was covered with wood tips. The novel object recognition task comprised the following three sessions. First, in a habituation session, mice were individually placed in the box for 10 min of exploration in the absence of objects. Second, a training session was performed 24 hr after the habituation session. Two objects (A and B) were placed in the back corner of the box, 10 cm from the sidewall. Mice were individually placed in the middle front of the box and the time spent in exploring the two objects was recorded. Exploration behavior of the mice was defined as directing the nose toward the object at a distance of less than 2 cm. Third, a retention session was performed 24 hr after the training session. One of the familiar objects (B) used in the training session was replaced by a novel object C, and objects A and C were placed as in the previous session. Mice were then individually placed in the box and allowed to explore freely for 10 min, with recording of the time spent in exploring each of the two objects. The exploratory preference was expressed as a ratio of the time spent exploring the novel object (TC) over that spent on the two objects (TA+TC): exploratory preference (%) = TC/(TA+TC) 6100.

Water Maze Test
A circular pool (diameter = 130 cm, Nishinihonkikou, Okayama, Japan) was filled with water to a height of 15 cm. White ink was used to render the water opaque. The water temperature was maintained at 2361uC throughout the trials. To enhance the positional recognition of mice, four different geometric patterns were placed at inner rim of each quadrant of the pool as well as each wall of the room. On the first day (day0), the platform (diameter = 9 cm) was placed at the midpoint of the pool, 0.5 cm above the water surface. Each mouse was placed on the platform and allowed to move away from the platform and swim in the pool for 15 sec, then gently guided to the original platform. The next day (day1), the platform was placed at the midpoint of one quadrant, submerged 0.5 cm below the water surface. The position of the platform was unchanged throughout the test trials. Two trials each day with an inter-trial interval of 1 min were conducted for 4 consecutive days (day1 to day4). In each trial, the mice were placed in the pool at one starting position, with their back turned to the midpoint of the pool. They were allowed to swim freely or until they found the hidden platform and came up for air while on the platform. The time required to escape onto the hidden platform was recorded as escape latency. If a mouse did not reach the platform within 120 sec, it was gently guided to the platform, where it remained for 10 sec. Starting locations were changed every trial. The results of 2 trials were averaged and used in data analysis.

Passive Avoidance Test
A passive avoidance test was conducted using the apparatus consisted of two separate chambers (30630620 cm height). One of the chambers was illuminated and the other was dark. Each chamber was separated by a small guillotine door (3.565 cm) and grids were attached on the floor in the dark chamber. This test comprised the following 3 sessions. First, in a pre-training session (day1 and day2), mice were individually placed in the illuminated chamber for 1 min and then in the dark chamber for 2 min. Second, in the test session (day3), mice were individually placed in the illuminated chamber. Immediately after the mice entered the dark chamber, the door was locked and an electrical stimulation of 0.36 mA was applied for 2 sec. The time spent in the illuminated chamber was measured. Third, in the retention session (day4), which was performed 24 hr later, mice were placed back in the illuminated chamber and the time spent in this chamber before entering the dark chamber was measured. The latency in the retention session was expressed graphically and used in data analysis.

Measurement of Ab Species from Plasma and Brain Homogenates by ELISA
After mice were deeply anesthetized with pentobarbital (50 mg/ kg), blood samples were withdrawn from the postcaval vein using heparinized syringe and the plasma was prepared and used fresh or was stored at 280uC until use. Cerebrospinal fluid (CSF) was collected from the dura mater over the cistern magna and brains were isolated. Hemi-brains were weighed and homogenized with a teflon-glass homogenizer in 20 mM Tris-buffered saline (TBS; 137 mM NaCl, protease inhibitor cocktail, pH7.6), and centrifuged at 100,0006g for 1 hr at 4uC. Pellets were resuspended in TBS containing 2% SDS. After centrifugation at 100,0006g for 1 hr at 4uC, the supernatant was obtained as the soluble fraction of Ab, and the pellet was extracted with 70% (w/v) formic acid, followed by centrifugation at 100,0006g for 1 hr at 4uC. The supernatant was neutralized by 20-fold dilution with 1 M Tris base [27] and used as the insoluble Ab fraction. Samples were diluted with the diluents provided in the human (h)Ab 1-40 and hAb 1-42 ELISA kits (Wako, Osaka, Japan). hAb  and hAb  in individual samples were quantified in accordance with the manufacturer's instructions and are expressed as picomoles/mL of plasma or picomoles/mg wet weight of brain (mean 6 SEM).

Histology
Hemi-brains were fixed with 10% buffered formalin for 24 hr and embedded in paraffin blocks, and then coronal sections of cerebral cortex with 5 mm thickness were stained with Congo-Red for counting amyloid plaques. For Ab immunohistochemistry, paraffin sections were dewaxed, rehydrated, and then blocked with PBS containing 0.2% Triton X-100 and 5% BSA. Sections were then incubated with rabbit anti-Ab-peptide polyclonal antibody (#71-5800, 1:50) (Invitrogen, Camarillo, CA) at 4uC overnight. After several washes in PBS, sections were incubated with HRPOconjugated secondary antibody (DAKO, anti-rabbit Igs, 1:2000) and then Ab-immunoreactive tangles were visualized by diaminobenzidine.

Statistical Analysis
All values are expressed as means 6 SD or SEM. A Student t test was used for comparison between 2 groups. One-way ANOVA with a Tukey-Kramer test was used to determine the significance of differences in multiple comparisons. In water maze, repeated measures two-way ANOVA followed by a post hoc Sheffe's F test was applied. Pearson correlation analysis was used to examine the correlation between Ab levels in plasma and brain in each mouse. Differences with a probability value of P,0.05 were considered to be significant.

Effect of NK-4 on Ab Protein Fibrillization
To examine the potential mechanisms of NK-4-mediated protection of neuronal cells from amyloid toxicity, we evaluated the effect of NK-4 on aggregation or fibrillization of three forms of Ab peptide. Solution of Ab 1-40 or Ab 1-42 (100 mM each) was incubated at 37uC for 3 days in the absence or presence of NK-4 at the indicated concentrations (Fig. 2). After the incubation, visible sedimentary aggregates were observed in the Ab controls, while the addition of NK-4 had reduced visible aggregates. To examine the effect of NK-4 further, EM was used to monitor Ab fibril formation. In EM, Ab 1-40 alone formed long and unbranched fibrils (Fig. 2a). In contrast, Ab 1-40 incubated with NK-4 (10 mM) had shorter and fewer filaments with soluble assemblies (Fig. 2b). Similarly, Ab 1-42 formed dense fibrils (Fig. 2c) and they were remarkably weakened by the co-incubation with NK-4 (Fig. 2d). ThT fluorescence is enhanced upon binding to Ab fibrils, proportionally to the amount of fibrils in solution [23]. So, we used ThT to evaluate the effect of NK-4 on the Ab fibril formation quantitatively. All three forms of Ab solutions showed enhanced emission at 482 nm, which is characteristic for ThT bound to amyloid fibrils. NK-4 inhibited all these ThT fluorescence dose-dependently (Fig. 2e). The inhibitory effect of NK-4 was most prominent on Ab 1-40 , then followed by Ab  and Ab [25][26][27][28][29][30][31][32][33][34][35] . An equimolar concentration of NK-4 almost totally inhibited the ThT fluorescence in all three forms of Ab. These results suggest that NK-4 is an effective inhibitor of Ab fibril formation in vitro.

Effects of NK-4 on Recognition Memory in the Novel Object Recognition Test
Given the beneficial effects of NK-4 on multiple aspects of amyloid pathology in vitro, we next examined whether NK-4 had an effect on the cognitive deficit induced by Ab in an AbPP transgenic mouse model of AD. Tg2576 mice expressing the Swedish mutation of AbPP were administered with 100 or 500 mg/kg NK-4 once a day, 5 times a week for 9 months, beginning at 3 months of age. Wild type and Tg2576 mice were tested for object recognition memory 24 hr after training to test long-term memory at ages 3, 6 and 12 months. During the training session, there were no significant differences in exploratory preference between the two objects or in the total exploratory time among the groups (data not shown). At 3 months of age, we could not detect significant differences in recognition memory among all groups tested (data not shown). In the retention session, Tg2576 mice at ages 6 and 12 months showed significantly decreased object recognition memory compared with that in the wild type mice at the respective ages (Fig. 3). Tg2576 mice treated with a low dose of NK-4 (100 mg/kg) tended to spend a longer  time exploring the novel object than did the saline-treated controls at both 6 and 12 months of age (treated for 3 and 9 months), but the differences were not significant. In contrast, Tg2576 mice treated with a high dose of NK-4 (500 mg/kg) spent a significantly longer time exploring the novel object than did the saline-treated mice at both 6 and 12 months of age, and the exploratory preference was comparable to that of wild type controls at age 12 months (after 9 months treatment). This suggests that NK-4 administration for a longer period might be more effective for improvement of object recognition memory.

Effects of NK-4 on Learning Ability in the Water Maze Test
We also evaluated learning ability of mice in a water maze test at ages 6 and 12 months (Fig. 4). Group differences in the escape latency of the training trial session were analyzed using two-way ANOVA with repeated measures. There were significant differences in escape latency between genotype (wild type controls vs. saline-treated Tg2576 mice) at 6 months of age [Genotype 6Time F(1,54) = 40.531, P,0.001], and at 12 months of age [Genotype 6 Time F(1,48) = 7.269, P,0.05]. For saline-treated Tg2576 group, the latency to reach the platform did not shorten during the whole test period at 12 months of age. There was a significant Genotype 6 Time interaction [F(3,48) = 5.0732, P,0.01] at this stage, suggesting that the learning ability of saline-treated Tg2576 mice became impaired with age.
We found significant differences in escape latency between treatments (NK-4 vs. saline) at 6 months of age [Group 6 Time   factor and the time factor also proved to be significant at 6 months of age [F(6,81) = 2.672, P,0.05]. Saline-treated Tg2576 group consistently exhibited a longer latency than NK-4-treated Tg2576 groups. Both low (100 mg/kg) and high (500 mg/kg) doses of NK-4 improved learning ability of Tg2576 mice with significant differences compared with saline-treated controls (P,0.01 at 6 months of age, P,0.05 at 12 months of age), but the effects of NK-4 on escape latency did not differ significantly between these two doses. The swimming ability did not differ significantly among Tg2576 groups (data not shown).

Effects of NK-4 on Passive Avoidance Memory
Recognition memory of NK-4-treated mice at ages 3 and 9 months was evaluated in a passive avoidance test. Similar to the novel object recognition test, no significant difference was detected in latencies for entering the dark chamber of a passive avoidance apparatus during the retention session among all groups tested at 3 months of age (data not shown). However at 9 months, there was a significant difference in latencies between wild type control and saline-treated Tg2576 mice (p,0.01). Both low (100 mg/kg) and high (500 mg/kg) doses of NK-4 reversed the decrease in passive avoidance latency of Tg2576 mice, and the differences were significant compared with saline-treated controls (both p,0.01, Fig. 5). There was no difference in latencies for entering the dark chamber between wild type and Tg2576 mice, or among the treatment groups after placement in the light chamber for the first time during the training session (in the absence of shock). All animals took an average of 70 s to move to explore the dark chamber, indicating no baseline differences in anxiety in the treatment groups in this behavioral test. We also found no difference in shock sensitivity among wild type and saline-and NK-4-treated Tg2576 mice.

Effect of NK-4 on Ab Levels in Tg 2576 Mice
We next evaluated whether NK-4 displayed any significant effect on Ab levels in plasma, CSF and brains from Tg2576 mice. Since NK-4 might directly interact with Ab peptides, we examined whether it interferes the Ab ELISA. Although the presence of NK-4 slightly enhanced optical density values in Ab ELISA systems, the range of increase was less than 5% when the NK-4:Ab molar ratio was 0.1,100 (Fig. 6a, b). This indicated that NK-4 hardly affects both Ab 1-40 and Ab 1-42 ELISA systems.
Dose-dependent increases in Ab x- 40 and Ab x-42 were observed in plasma from NK-4 treated mice (Fig. 6c, d). In contrast, CSF Ab x- 40 and Ab x-42 levels were not significantly changed by NK-4 treatment, although the CSF concentrations of Ab x- 40 and Ab x-42 in NK-4 treated mice tended to be lower than those in salinetreated controls (data not shown). Brain levels of Ab were estimated separately in soluble and insoluble fractions. Both soluble and insoluble Ab x-40 levels were significantly decreased by NK-4 treatment (Fig. 7a, c). Similarly, levels of Ab x-42 were lower in NK-4 treated mice, with a significant decrease in insoluble Abx-42 (Fig. 7b, d). The levels of Ab x- 40 and Ab x-42 in brain were inversely correlated with those in plasma (r = 20.69, P,0.01 (n = 27), and r = 20.42, P,0.05 (n = 27), respectively) (Fig. 7e, f).
We next surveyed brain Ab deposition of Tg2576 mice by Congo Red (CR) staining and Ab immunohistochemistry (Fig. 8).
Wild type mice did not show any sign of amyloid deposition in their brain at 12 month of age (Fig. 8a). Tg2576 mice at the same age developed distinguishable, but very few amount of plaques in their cortex region. The number of CR-positive plaques in the cerebral cortex region was not significantly changed among salinetreated group and NK-4-treated groups (Fig. 8a). On the contrary, Ab-immunoreactive small tangles were abundantly seen in the cortex of saline-treated Tg2576 mice (Fig. 8b), and they were obviously reduced by the high dose NK-4 treatment both in size and in quantity (Fig. 8c).

Effects of NK-4 on Ab25-35-Induced Cognitive Impairment in ICR Mice
To confirm whether the effect of NK-4 was directly attributable to mitigation of Ab pathology, we employed an Ab-induced amnesia model, in which ICR mice received intracerebrovantricular (icv) administration of aged Ab [25][26][27][28][29][30][31][32][33][34][35] peptide to provoke memory deficits [25]. Long-term recognition memory was evaluated by the novel object recognition test and the passive avoidance test starting at 6 days and 9 days after Ab 25-35 -icv injection, respectively. In both behavioral assays, daily NK-4treatment dose-dependently, and significantly improved the memory deficits induced by Ab (Fig. 9a, b). These results, in combination with the results from Tg2576 mice, strongly suggested that NK-4 treatment improved Ab-mediated memory impairments.

Discussion
NK-4 is a N-heterocyclic cationic dye derivative that contains three quinolinium rings and two iodine anions. Although NK-4 shows a variety of biochemical and biological activities [15][16][17][18][19][20][21], structure-activity relationships or molecular mechanisms of them remain largely unknown. Extended p-electron conjugated system of NK-4 would considerably contribute to antioxidative effects by trapping free radicals similar to another cyanine dye platonin [29]. The extended conjugated system also confers this molecule a planar structure by which NK-4 seems to bind at the interface of the bsheet domains of the proteins and act as a ''b-sheet breaker'' [30]. On the other hand, it is reported that the side chain length affects water solubility or affinity to the lipid membrane in this kind of dye compounds [31]; therefore, the specific activity of NK-4 might change in accordance with the length of N-alkyl side chains.
Application of neurotrophic small molecules that modulate neuronal survival and synaptic function is a promising therapeutic approach for AD, while modulation of the amyloid pathology is a key strategy for slowing or halting the disease progression. In this study, we showed that NK-4 protected neuronal cells from Ab 25-35 -induced toxicity (Fig. 1). Attenuation of Ab cytotoxicity by NK-4 might be caused by direct action on Ab to prevent its aggregation into more toxic forms, or via activation of intracellular survival signaling pathways and inactivation of death pathways in neuronal cells [20], or both. The effect of NK-4 on Ab fibril formation was significant, but the IC 50 value for Ab aggregation was approximately 10 mM for Ab 1-40 and Ab 1-42 , and 30 mM for Ab [25][26][27][28][29][30][31][32][33][34][35] (Fig. 2e). In contrast, a much lower concentration (nM level) of NK-4 attenuated the neurotoxicity induced by 50 mM Ab 25-35 in PC12 cells (Fig. 1). These results suggest that NK-4 attenuates Ab-induced toxicity via activation of intracellular survival signaling pathways including Akt activation [20], rather than direct inhibition of Ab aggregation, since the concentration of NK-4 required inhibiting Ab aggregation was substantially high. However, the effect of NK-4 on Ab aggregation was greater than those of non steroidal anti-inflammatory drugs (NSAIDs) such as ibuprofen, naproxen, ketoprofen, and indomethacin [32,33], which require at least a threefold higher concentration to inhibit Ab aggregation. Therefore, the anti-fibrillization effect of NK-4 might play some role in attenuating Ab toxicity in vivo.
Chronic administration of NK-4 to Tg2576 mice significantly attenuated cognitive decline as assessed in a set of behavioral tests (Fig. 3,4,5), and also decreased the levels of Ab in brain (Fig. 7) and augmented those in plasma (Fig. 6). These results imply that the NK-4-induced behavioral improvement was attributable to decreased Ab in brain, and that Ab may be cleared across the Figure 6. Effect of NK-4 on the plasma Ab concentrations in Tg2576 mice. NK-4 did not interfere the optical density of both hAb 1-40 (a) and hAb 1-42 (b) ELISA systems, which using the antibody combinations of BAN50/BA27 and BAN50/BC05, respectively. Since these ELISA systems cannot distinguish N-terminal heterogeneity of Ab peptides derived from biological samples, we represented hAb species as hAb x-40 and hAb x-42 , respectively. Plasma from mice aged 12 months was used for measurements of hAb x-40 (c) and hAb x-42 (d). Values are the mean 6 SEM for salinetreated Tg2576 mice (control group, n = 10), a low dose NK-4 (100 mg/kg)-treated group (n = 9), and a high dose NK-4 (500 mg/kg)-treated group (n = 8). **P,0.01 vs. saline-treated Tg2576 group. doi:10.1371/journal.pone.0030007.g006 blood-brain barrier to the blood. A similar large increase in plasma Ab was found after peripheral administration of an anti-Ab antibody in an AD mouse model [34]. Although there is still no consensus as to the mechanism by which anti-Ab antibodies alter amyloid deposition, in vivo binding properties of the antibodies would affect plasma Ab levels by altering the half-life of Ab [34]. In a similar fashion, a direct NK-4 binding to Ab in Tg2576 mice may be one possible explanation of augmented plasma Ab levels by chronic peripheral administration of NK-4. On the other hand, plasma might serve as a sink that enhances clearance of Ab from the brain based on a peripheral sink hypothesis [35,36], because Ab clearance mainly occurred in plasma in Tg2576 mice [34]. It might be also possible that direct NK-4 binding to Ab creates a sink that enhances clearance of Ab deposition from brain. In either case, further study is required to address these issues.
It seems unlikely that the impaired performance of Tg2576 mice in learning and memory tests is due to changes in motivation or sensorimotor function, since the motivation for each behavioral test is different and different skills are required for a good performance in each test. There were no differences in locomotor activity, total time spent exploring objects in the novel object test between wild type and Tg2576 mice. Recognition memory in the novel object test and associative learning in the passive avoidance test are dependent on the hippocampus and/or perirhinal cortex    [37][38][39]. In this context, progressive impairment of the hippocampus and/or perirhinal cortex-dependent memory started as early as 6 months of age and continued to 12 months of age in Tg2576 mice in our experiment. Contrary to this, histological investigation of Ab deposits in Tg2576 mice showed no substantial differences in quantity of CR-positive plaques among the salinetreated group and NK-4-treated groups at 12 month of age (Fig. 8a). On the other hand, Ab-immunoreactive tangles were abundant in the cortex of saline-treated Tg2576 mice and they were clearly decreased by high dose NK-4 treatment in good agreement with the results of Ab ELISAs (Fig. 7, Fig. 8b, c), and it might also reflect the cognitive function (Fig.3, 4). The difference in plaque stainability between CR and anti-Ab would be due to the sensitivity to immature tangles, and therefore, CR staining might not reflect the total amount of Ab deposition in brain.
The Tg2576 mouse is the most thoroughly characterized AD mouse model and is considered to reflect the human amyloid pathology most closely among mouse models [11]. Although many studies using mice on a B6/SJL background (hereafter; Tg2576/ B6), we used here the newly developed 129S6/SvEvTac strain (hereafter; Tg2576/129) established by Taconic, because this strain does not carry a retinal degeneration mutation [24]. The 129 strains are reported not appropriate for water maze tasks according to previous studies [40][41][42][43], however, the wild type mice of 129S6/SvEvTac strain performed well in the acquisition learning of water maze test [44]. Regarding the timing of cognitive decline, our results showed that the Tg2576/129 did not exhibit cognitive decline at the age of 3 months (data not shown), but became apparent at 6 months, which continued to 12 months (Fig. 3,4,5). Reportedly, Tg2576/B6 showed no detectable cognitive impairment at 3-4 months, but became obvious at 5-6 months of age [14,[45][46][47]. Thus, genetic backgrounds appear not to be affected in the timing of cognitive decline.
The water maze protocol we employed in this study is a variant of Morris water maze navigation task. Although the original protocol was designed for evaluating hippocampus-dependent spatial reference learning and memory [48], our modified protocol is not considered as the one for spatial cognition since we used extra intra-maze landmarks and did not run a probe trial. We placed proximal landmarks on the inner rim of the pool in addition to each wall of the room (distal landmarks) aiming to enhance positional recognition of mice, however, it is suggested that the processing of information related to distal landmarks (room cues) or proximal landmarks (intra-maze cues) are mediated by different neural systems [49,50]. The hippocampus circuitry contributes to a spatial strategy, which involves flexible use of spatial/distal cues, whereas a cue strategic processing of proximal landmarks is dependent upon striatal circuitry [50]. Thus, what we evaluated using the extra proximal landmark condition would not be spatial reference learning, but non-spatial or mixture of both. On another note, the lack of probe trial made it difficult to confirm the memory retention. As a result of these modified experimental design, we could not determine whether the learning strategy of the mice was spatial or not. One thing for certain is, however, that NK-4 modulated the learning ability of Tg2576/129 mice in the water maze task (Fig.4). The learning ability of Tg2576/129 mice was impaired with age, and a long-term NK-4 treatment significantly improved it. Results of two other behavioral tests (Fig.3, Fig.5) support this notion.
In relation to the learning ability of the mice on 129 background, it has been known that all 129 mouse substrains have disrupted-inschizophrenia 1 (DISC1) deletion mutation [51]. Because DISC1 is reported to play a critical role in development and migration of hippocampal neurons [52], cognitive deficits observed in 129 strains may linked to lack of DISC1 protein. In this regard, preferable effect of NK-4 on cognitive improvements of Tg2756/129 mice may be attributable to restoring the dysfunctional DISC1-mediated pathway. To address this question, we examined the effect of NK-4 on Ab-dependent amnesic mice model [25]. As a result, NK-4 treatment dose-dependently and significantly attenuated memory dysfunction induced by direct injection of Ab [25][26][27][28][29][30][31][32][33][34][35] peptide into the lateral ventricle of ICR mice (Fig. 9). Therefore, the modification of DISC1-mediated pathway by NK-4 seems not mainly involved in this case. Since NK-4 activated PI3K-Akt pathway independently of the TrkA receptor in vitro [20], activated Akt (p-Akt) may be an important mediator of the beneficial effects on learning and memory. The p-Akt is required for PI3K-mediated synaptic plasticity and memory consolidation by promoting neuronal cell survival and protein synthesis via phosphorylation of CREB, a downstream regulator involved in hippocampus-dependent longterm and spatial memory formation [53]. Therefore, induction of Akt phosphorylation might play critical roles in NK-4-mediated memory improvement in Tg2576/129 mice.
Oxidative mechanisms are also thought to be involved in the cell loss and other neuropathology associated with AD [54,55]. In AD pathogenesis, reactive oxygen species (ROS) impair mitochondrial redox activity and increases further ROS generation [29,56,57]. Ab induces production of ROS and leads to apoptotic neuronal cell death that can be inhibited by antioxidants [57][58][59]. Pathologic and biochemical studies suggest that ROS induced by fibrillar Ab have neurotoxic effects [60,61]. As mentioned before, NK-4 is a potent scavenger of ROS [21] and an inhibitor of Ab aggregation (Fig. 2). Those activities might contribute to attenuation of Ab toxicity in AD.
As to the safety profile, NK-4 was well tolerated in a mouse acute oral toxicity study at up to 2 g/kg [21], based on mortality, clinical observations, body weight, hematology, blood chemistry, organ weights, and histological examinations of a full tissue list. Additionally, NK-4 was not mutagenic in the standard Ames test due to bacterial-specific metabolism (data not shown). Furthermore, there were no specific adverse events in mice those received intraperitoneal injections of NK-4 at the dose of up to 500 mg/kg/ day, 5 days a week, for 9 months (this study). These observations suggest that NK-4 is a safe compound that will not cause serious adverse reactions. Following an intraperitoneal administration at 500 mg/kg, NK-4 penetrated the brain and nM levels were detected using high-performance liquid chromatography [20]. Thus, the nM range used in the in vitro investigations in this study corresponds to the calculated brain concentration in Tg2576/129 mice receiving NK-4.
In summary, the results of this study show that chronic treatment with NK-4 from an early stage of cognitive dysfunction significantly ameliorated learning impairments in Tg2576/129 AbPP mouse. The NK-4-induced behavioral improvement might be attributable to decreased Ab accumulation in brain, and the inhibitory effects on Ab aggregation, neurotrophin-like property, as well as radical-scavenging effect of NK-4. Its multiple functions might work in concert to attenuate AD pathology.