Neuroprotective Effect of SLM, a Novel Carbazole-Based Fluorophore

Dec 29, 2016 - ... if SLM can decrease the cognitive and behavioral deficits observed in ... (SH-SY5Y) cells was found to be reduced through the treat...
1 downloads 0 Views 6MB Size
Research Article pubs.acs.org/chemneuro

Neuroprotective Effect of SLM, a Novel Carbazole-Based Fluorophore, on SH-SY5Y Cell Model and 3xTg-AD Mouse Model of Alzheimer’s Disease Xiaoli Wu, Jayasankar Kosaraju, Wei Zhou, and Kin Yip Tam* Drug Development Core, Faculty of Health Sciences, University of Macau, Taipa, Macau, China ABSTRACT: Amyloid β (Aβ) peptide aggregating to form a neurotoxic plaque, leading to cognitive deficits, is believed to be one of the plausible mechanisms for Alzheimer’s disease (AD). Inhibiting Aβ aggregation is supposed to offer a neuroprotective effect to ameliorate AD. A previous report has shown that SLM, a carbazole-based fluorophore, binds to Aβ to inhibit the aggregation. However, it is not entirely clear whether the inhibition of Aβ aggregation alone would lead to the anticipated neuroprotective effects. In the current study, we intended to examine the protective action of SLM against Aβinduced neurotoxicity in vitro and to evaluate if SLM can decrease the cognitive and behavioral deficits observed in triple transgenic AD mouse model (3xTg-AD). In the in vitro study, neurotoxicity induced by Aβ42 in human neuroblastoma (SH-SY5Y) cells was found to be reduced through the treatment with SLM. In the in vivo study, following one month SLM intraperitoneal injection (1, 2, and 4 mg/kg), 3xTg-AD mice were tested on Morris water maze (MWM) and Y-maze for their cognitive ability and sacrificed for biochemical estimations. Results show that SLM treatment improved the learning and memory ability in 3xTg-AD mice in MWM and Y-maze tasks. SLM also mitigated the amyloid burden by decreasing brain Aβ40 and Aβ42 levels and reduced tau phosphorylation, glycogen synthase kinase-3β activity, and neuroinflammation. From our observations, SLM shows neuroprotection in SH-SY5Y cells against Aβ42 and also in 3xTg-AD mouse model by mitigating the pathological features and behavioral impairments. KEYWORDS: Amyloid β, aggregation, Alzheimer’s disease, SLM, 3xTg-AD, neuroprotection, SH-SY5Y



small molecules, peptides, and antibodies.2 However, none of them cleared clinical trials due to their severe toxicity.4 A study by Jack and colleagues5 showed that deposition of amyloid β aggregates in the neuron is the trigger that lead to AD. It is evident that amyloid feature is the initiation for the neurodegeneration in AD. Several molecular pathways including glycogen synthase kinase-3β (GSK-3β) stimulation are affected by Aβ generation and aggregation in AD.6 GSK-3β over activity leads to Aβ production, tau hyperphosphorylation, and cognitive decline, the cardinal features of AD. It is still uncertain whether Aβ generation stimulates GSK-3β or vice versa in the progression of AD. AD is considered to be a disorder with several interconnected cellular mechanisms. Therefore, drugs that target multiple features of AD could be more effective than a drug with a single target. SLM, a carbazole-based fluorophore, is a non-peptide small molecule that binds to Aβ and inhibits the aggregates in vitro.7 The structure of SLM is given in Figure 1. Due to its blood− brain barrier penetration capacity, SLM was developed as a carbazole-based fluorophore for detecting amyloid β in AD. According to Yang and colleagues,7 SLM is an Aβ binding molecule with moderate binding constant (49 μM). Our previous results demonstrated the pharmacokinetic profile of SLM in ICR mice with a clearance rate of 1.34 l/(h·kg) at 4

INTRODUCTION Alzheimer’s disease (AD) is the most prevalent neurodegenerative disorder affecting 44 million people around the

Figure 1. Chemical structure of SLM.

world.1 The main pathological features present in post-mortem AD brain are plaques formed by the aggregation of amyloid β (Aβ) peptide and tangles formed by the hyperphosphorylation of tau protein. From these, Aβ aggregation is considered as the main neurotoxic feature leading to other complications including neuro-inflammation and oxidative stress.2 The leading cause for the Aβ aggregation is still not known due to the complex nature of the disease. The predictive pathway leading to toxic aggregates is nucleation, low molecular weight oligomers, larger β-sheets-rich oligomers, aggregation of oligomers, and finally formation of fibrils.3 Albeit fibril formation is the final product, Aβ oligomerization or aggregation is believed to be the main culprit in the progression of AD. Full literature is available on targeting Aβ aggregates by © XXXX American Chemical Society

Received: November 11, 2016 Accepted: December 16, 2016

A

DOI: 10.1021/acschemneuro.6b00388 ACS Chem. Neurosci. XXXX, XXX, XXX−XXX

Research Article

ACS Chemical Neuroscience mg/kg dose,8 though it remains unclear whether SLM mitigates the amyloid plaque load by acting on Aβ. In the present study, we sought to undertake a detailed investigation on the neuroprotective effects of SLM. In particular, we evaluated the therapeutic potential of SLM in in vitro against Aβ42 induced neurotoxicity and also on a triple transgenic AD mouse model (3xTg-AD). Treatment with SLM for one month improved the cognitive function of 3xTg-AD mice. After one month treatment, we observed a detectable level of SLM in the brain which might be responsible for the cognitive improvement. Based on the in vitro experiments, Yang and colleagues7 have suggested that the major role of SLM is to bind amyloid β and inhibit the aggregation to show the neuroprotection. In our study, we observed protective activity of SLM against Aβ42 on SH-SY5Y cells and also reduction of amyloid burden in 3xTg-AD mice upon the dosing of SLM. Moreover, SLM also reduces tau phosphorylation, GSK-3β activity, and neuro-inflammation following one-month treatment. We have demonstrated for the first time that SLM, a novel carbazole-based fluorophore, ameliorates pathological features of AD and improves cognitive ability. These results raise the possibility of SLM as a potential therapeutic agent for AD patients.



RESULTS SLM Decelerates SH-SY5Y Cell Death Induced by Aβ42. Preliminary study indicated that SLM does not affect cell viability; however Aβ42 at a concentration of 20 μM significantly reduced cell viability (data not shown). To test the protective effect of SLM on 20 μM Aβ42 induced cell

Figure 3. SLM decreases Aβ42 induced cell death. (A) SH-SY5Y cells were incubated with 2 μM SLM and 20 μM Aβ42, and the apoptotic rate was determined by annexin V/PI staining. (B) Quantitative analysis of apoptotic cells. Significance was analyzed by one-way ANOVA followed by Newman−Keuls post hoc test (mean ± SEM) using Graphpad Prism. ##p < 0.01, compared with control; **p < 0.01, compared with Aβ.

Figure 4. SLM brain concentration following one-month treatment.

death, SH-SY5Y cells were incubated with different concentrations of SLM (0.08, 0.4, and 2 μM) for 24 h, and the cell viability was tested using 3-(4,5-dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide (MTT). As shown in Figure 2A, in the presence of Aβ42, treatment with SLM significantly increased the viability of the cells in a concentration-dependent manner (F(4, 85) = 1071, p < 0.0001, followed by Newman− Keuls post hoc test, p < 0.05 at 0.08 μM and p < 0.01 at 0.4 and

Figure 2. Cell viability of SLM in the presence of Aβ42 by MTT (A) and LDH (B) assay. Significance was analyzed by one-way ANOVA followed by Newman−Keuls post hoc test (mean ± SEM) using Graphpad Prism. #p < 0.05, ##p < 0.01, compared with control; *p < 0.05 and **p < 0.01, compared with Aβ. B

DOI: 10.1021/acschemneuro.6b00388 ACS Chem. Neurosci. XXXX, XXX, XXX−XXX

Research Article

ACS Chemical Neuroscience

Figure 5. SLM improved cognitive ability in the Morris water maze. (A) Latency to reach the escape platform, (B) time spent in the platform quadrant, and (C) number of platform crossings. (D) Tracking details on day 4 trial 2 of corresponding groups. Significance was analyzed by repeated-measures one-way ANOVA (for panel A) and one-way ANOVA (for panels B and C) followed by Newman−Keuls post hoc test (n = 8, mean ± SEM) using Graphpad Prism. *p < 0.05 and **p < 0.01, compared with control mice.

2 μM vs Aβ). We also obtained similar results in the lactate dehydrogenase (LDH) assay. As shown in Figure 2B, the presence of Aβ42 for 24 h significantly increased the level of LDH compared with the control cells. However, treatment of SLM significantly (F(4, 85) = 2961, p < 0.0001 followed by Newman−Keuls post hoc test, p < 0.01 vs Aβ) decreased the level of LDH when compared with the cells treated with Aβ42. Results show that SLM significantly increased the cell viability of SH-SY5Y cells in both MTT and LDH assays (Figure 2). To further determine the effect of SLM on Aβ42 induced cell death, we performed flow cytometric assay using annexin V/PI staining and showed that SLM protects SH-SY5Y cells against Aβ42 induced cell death (Figure 3; F(3, 44) = 4043, p < 0.0001, followed by Newman−Keuls post hoc test, p < 0.01 vs Aβ). These results indicated that SLM ameliorated cellular toxicity caused by Aβ42. SLM Brain Concentration after One-Month Treatment. To determine the brain concentration of SLM in 3xTgAD mice after 1-month treatment, we checked the SLM concentration in the brain homogenate. The mean concentration of SLM at 2 and 4 mg/kg were 78.74 and 101.32 ng/g of tissue, respectively (Figure 4). SLM Improves the Cognitive Ability of 3xTg-AD Mice. To examine the influence of SLM on cognitive performance, we conducted Morris water maze (MWM) and Y-maze tests on 3xTg-AD mice after one-month treatment. Spatial learning and memory in SLM treated mice can be examined by employing MWM. Time to reach the platform (latency time) was measured from day 1 to 4 in control and SLM treated mice. SLM treated mice (2 and 4 mg/kg) showed significant decrease in the latency time on days 3 and 4 (Figure 5, repeatedmeasures one-way ANOVA, F(3, 28) = 4.309, p = 0.01 on day 3

Figure 6. SLM improved cognitive function in the Y-maze. (A) Time spent in the novel arm and (B) percentage of alterations. Significance was analyzed by one-way ANOVA (n = 8, mean ± SEM) followed by Newman−Keuls post hoc test using Graphpad Prism. *p < 0.05 and **p < 0.01, compared with control mice.

C

DOI: 10.1021/acschemneuro.6b00388 ACS Chem. Neurosci. XXXX, XXX, XXX−XXX

Research Article

ACS Chemical Neuroscience

presence of Aβ deposits at the age of 6 months.9 Vehicle treated 3xTg-AD mice displayed amyloid burden as expected by a disease model. ELISA measurement of amyloid burden (Aβ42 and Aβ40) revealed a significant decrease following SLM treatment (Figure 7). SLM at 2 and 4 mg/kg exhibited significant reduction of Aβ42 (F(3, 20) = 4.10, p = 0.02, followed by Newman−Keuls post hoc test, p < 0.01 vs control) and Aβ40 (F(3, 20) = 4.59, p = 0.01, followed by Newman− Keuls post hoc test, p < 0.01 vs control) as compared with control mice (Figure 6). Similar effect was observed in the immunohistochemistry sections, where SLM significantly reduced the plaque count compared to vehicle treated mice (Figure 8). This was further supported by the thioflavin S positive plaque staining, showing reduced plaque count in the SLM treated mice compared to control mice (Figure 9). Overall the amyloid burden was reduced by SLM treatment in 3xTg-AD mice. SLM Reduces GSK-3β Activity and Tau Pathology in 3xTg-AD Mice. Another important pathological feature that is responsible for neuronal death in AD is tau hyperphosphorylation. To determine the possible mechanism of SLM on cognitive improvement, we examined the tau phosphorylation and GSK-3β activity in the 3xTg-AD mice brain by Western blots. We used total tau, p-tau, GSK-3β, and p-GSK-3β antibodies to determine the action of SLM on tau hyperphosphorylation and GSK-3β activity. We observed no change in total tau (data not shown) and total GSK-3β levels in SLM treated mice compared to control mice. However, we observed a significant (F(3, 16) = 1441, p < 0.0001, followed by Newman−Keuls post hoc test, p < 0.01 vs control) reduction in phosphorylated tau levels at Ser202 in the SLM treated mice compared to control mice (Figure 10). SLM increased the pGSK-3β level significantly (F(3, 16) = 319.8, p < 0.0001, followed by Newman−Keuls post hoc test, p < 0.05 at 2 mg/kg and p < 0.01 at 4 mg/kg vs control) compared to control group. Altogether, the effects of SLM showed a decline in tau phosphorylation and GSK-3β activity. We also tested the protein expressions of amyloid precursor protein (APP) and β amyloid cleavage enzyme 1 (BACE1) after SLM treatment. However, no significant change was observed in APP and BACE1 levels with SLM treatment (Figure 11). SLM Reduces Neuro-inflammation in 3xTg-AD Mice. Increase in neuro-inflammation has been observed in the postmortem brain of AD patients and observed in 3xTg-AD brain. We investigated whether the tested doses of SLM were effective to reduce the neuro-inflammatory ionized calcium-binding adapter molecule 1 (Iba1) and glial fibrillary acidic protein (GFAP) expressions. We observed a clear decrease in Iba1 (F(3, 16) = 617.8, p < 0.0001) and GFAP (F(3, 16) = 1267, p < 0.0001) levels in the SLM treated groups at 2 mg/kg (p < 0.05) and 4 mg/kg (p < 0.01) compared to control mice. High dose of SLM (4 mg/kg) reduced the Iba1 and GFAP to approximately 50% compared to control mice (Figure 12). No significance reduction of Iba1 and GFAP was observed at low dose of SLM. These observations revealed that SLM mitigates neuro-inflammation in 3xTg-AD mice.

Figure 7. SLM decreases brain amyloid burden following one month of treatment by decreasing amyloid β 40 and 42 levels. Significance was analyzed by one-way ANOVA (n = 6, mean ± SEM) followed by Newman−Keuls post hoc test using Graphpad Prism. **p < 0.01, compared with control mice. Levels of (A) Aβ42 and (B) Aβ40 in the brain.

and F(3, 28) = 7.065, p = 0.001 on day 4), albeit mice in control group also showed improved performance from day to day. A significant treatment effect (on days 3 and 4 with p < 0.01, except on day 3 with 1 mg/kg (p < 0.05) followed by Newman−Keuls post hoc test vs control) was observed during the training period. Time spent in the target quadrant and platform crossing were calculated during probe trial (day 5). Mice treated with SLM (2 and 4 mg/kg) showed significant increase in both time spent in target quadrant (one-way ANOVA, F(3, 28) = 3.393, p = 0.03, followed by Newman− Keuls post hoc test, p < 0.05 at 2 mg/kg and p < 0.01 at 4 mg/kg vs control) and platform crossings (one-way ANOVA, F(3, 28) = 4.953, p = 0.007, followed by Newman−Keuls post hoc test, p < 0.01 vs control). The search strategies of mice on day 4 are shown in Figure 5D. This result was further supported by the Y-maze assessment in which working memory was tested on SLM treated mice. The time spent in the novel arm was significantly higher (F(3, 28) = 4.324, p = 0.01, followed by Newman−Keuls post hoc test, p < 0.05 at 2 mg/kg and p < 0.01 at 4 mg/kg vs control) in SLM treated mice compared to control mice (Figure 6). Percentage alterations were also increased significantly (F(3, 28) = 6.046, p = 0.002, followed by Newman−Keuls post hoc test, p < 0.01 vs control) in SLM treated mice compared to control mice. Untreated 3xTg-AD mice exhibited cognitive deficits in MWM and Y-maze as anticipated for this strain. SLM Ameliorates Brain Amyloid Burden in 3xTg-AD Mice. The main feature of AD is increased amyloid burden presented with both Aβ42 and Aβ40. 3xTg-AD mice show the



DISCUSSION Amyloid β and phosphorylated tau are the major targets to develop the novel molecule(s) for the prevention or treatment of AD. It is not surprising to state that Aβ peptide exists biologically from birth and is produced throughout life.10 Several attempts to inhibit Aβ aggregation are under D

DOI: 10.1021/acschemneuro.6b00388 ACS Chem. Neurosci. XXXX, XXX, XXX−XXX

Research Article

ACS Chemical Neuroscience

Figure 8. Immunohistochemical analysis of amyloid plaques shows that SLM reduces amyloid burden following one month of treatment. Significance was analyzed by one-way ANOVA (n = 5, mean ± SEM) followed by Newman−Keuls post hoc test using Graphpad Prism. *p < 0.05 compared with control mice. (A) Representative images of immunohistochemistry at 2× and (B) the plaque count per section.

interactions with target, the Kd value is usually well below micromolar (mostly in nanomolar) range. Factoring in the resistance of the blood−brain barrier, it is unlikely that under the highest dose used in this study, the brain concentration of SLM is sufficiently high to elicit the observed neuroprotective effects if the therapeutic action is solely due to the binding to Aβ followed by aggregation inhibition. Nevertheless, the moderate affinity of SLM toward Aβ might facilitate the initial binding of the molecule to the peptides and contribute to the reduction of amyloid burden (Aβ40 and Aβ42) in the 3xTg-AD mice following one-month treatment. To further explore the mechanism, we evaluated the SLM activity on BACE1 and APP (Figure 10) and confirmed that the SLM may not affect APP processing in 3xTg-AD mice. An earlier report proposed that Aβ production was controlled by GSK-3β activity.13 Amyloid burden, neurofibrillary tangles, and cognitive deficits are well related to the activity of GSK-3β.6 Activation of GSK-3β promotes Aβ generation, tau phosphorylation, and also neuronal apoptosis.14,15 We conducted Western blotting of total and p-GSK-3β to determine the effect of SLM on GSK activity in 3xTg-AD mice. SLM significantly decreased the GSK-3β activity by increasing its phosphorylation. This might also be responsible for the decrease in amyloid burden. Ongoing study on another carbazole-based cyanine fluorophore (of the same chemical series as SLM) from our group (data not published) shows a similar kind of effect in the transgenic mouse model of AD. The present outcome indicated that SLM may reduce the formation of Aβ in 3xTg-AD model mice not only by binding Aβ but also by decreasing the GSK-3β activity. The immunohistochemistry and thioflavin S staining slides also showed the reduced amyloid load by SLM treatment. The concentration of SLM present in the brain (Figure 4) appears

investigation, and some of them failed to clear clinical trials. The success rate of the molecules in clinical trials for AD between 2002 and 2012 is 0.4% (99.6% failure).4 Molecules that attract Aβ and bind the toxic assemblies of Aβ may reduce the toxicity and facilitate the elimination of the toxic products. As reported previously,7 SLM targets Aβ and inhibits Aβ aggregation. Potentially, SLM and analogous molecules that could prevent the formation of Aβ toxic oligomers from monomers or are capable of counteracting Aβ association might ameliorate the cardinal pathologies and behavioral deficits of AD. The results of the current study clearly demonstrated that SLM ameliorated the pathological features and behavioral deficits of AD. Neuronal death induced by Aβ in a neuroblastoma cell line is a suitable cellular model to screen novel therapeutics for AD. Neuron death induced by excessive accumulation of Aβ was considered to be a critical hallmark in the pathogenesis of AD.11 It was reported that abnormal Aβ aggregation could stimulate neuronal cell death.12 Thus, in this investigation, we tested the neurotoxicity of Aβ on SH-SY5Y cells and the protective activity of SLM by MTT, LDH, and flow cytometry assays. We found that Aβ could induce cell death; however when cotreated with SLM, the cell death rate was significantly mitigated. These results indicate that SLM could reduce the cell neurotoxicity induced by Aβ. The binding studies between Aβ and carbazole-based cyanine fluorophores including SLM by Yang and colleagues7 showed that SLM has a binding constant (Kd) of 49 μM, which appeared to be the lowest in the same chemical series. In the context of protein−small molecule binding, the binding of SLM to Aβ is in the micromolar range, which cannot be regarded as high affinity. Typically for small molecule drugs showing good E

DOI: 10.1021/acschemneuro.6b00388 ACS Chem. Neurosci. XXXX, XXX, XXX−XXX

Research Article

ACS Chemical Neuroscience

development for AD, it is vital to recognize the role of Aβ on pro-inflammatory properties of microglia. We reported here that increased neuro-inflammation in the 3xTg-AD mice was prevented by SLM treatment. Immunoreactivity against GFAP and Iba1 demonstrated an obvious increase of reactive astrocytes in the brains of control transgenic mice compared to treated groups. In contrast, GFAP and Iba1 immunoreactivity was decreased in the SLM treated mice. The substantial reduction in neuro-inflammation in the SLM treated mice can be attributed to its inhibition of Aβ accumulation. Cognitive deficits present in 3xTg-AD mice make it a reliable model to check cognitive enhancers for the treatment of AD.25 The model used in the present study shows cognitive deficits at the age of 6 months that worsen with age.26,27 MWM and Ymaze were used to determine the spatial learning and memory and short-term and working memory of 3xTg-AD mice. Aβ aggregation causes age-dependent cognitive dysfunction in an APP overexpressing mouse model.28 It was also supported by the immunization studies using Aβ antibodies, which show cognitive enhancement.29−31 The cognitive deficits observed in the present study in 3xTg-AD mice are consistent with the earlier studies.32,33 Our observation showed that SLM, a carbazole-based fluorophore that binds to Aβ and inhibits the aggregation, improved the cognitive ability after one-month treatment. Eventually, the identification of molecules that target several features of AD is crucial for effective treatment. Obviously, SLM possesses some of these features. Our ongoing aim is to find an investigational new drug approval for a novel molecule like SLM or its derivatives for the treatment or prevention of AD. Altogether, SLM, a carbazole-based fluorophore, showed neuroprotection by binding to Aβ in the triple transgenic mouse model of Alzheimer’s disease. A multimechanism is tangled in the neuroprotective activity in triple transgenic mice resulting from SLM treatment. Importantly, SLM also ameliorates tau phosphorylation and neuro-inflammation and improves cognitive skill in the mice following one month of treatment. We can conclude that the neuroprotective property of SLM is due to the compound acting on multiple phenotypes including the inhibition of Aβ aggregation and the amelioration of tau phosphorylation as well as neuro-inflammation. This study provides compelling evidence to support the development of these novel carbazole-based fluorophores against AD. The outcomes promote the prospect of SLM or its derivatives as a beneficial molecule for AD patients. Further studies are warranted to fully explore the mechanisms of neuroprotection.

Figure 9. Thioflavin S staining of amyloid plaques shows that SLM reduces amyloid plaques following one month of treatment. Significance was analyzed by one-way ANOVA (n = 5, mean ± SEM) followed by Newman−Keuls post hoc test using Graphpad Prism. *p < 0.05, compared with control mice. (A) Representative images of thioflavin S staining at 5× and (B) the plaque count per section. Scale bar: 100 μm.

to be sufficient to show the therapeutic activity against AD pathological features To determine the possible mechanism of SLM on GSK-3β, we further examined the tau phosphorylation level that is responsible for neuronal death in AD. Aβ accumulation affects different biological cascades that lead to tau hyperphosphorylation.16 SLM treatment mitigates tau phosphorylation, which might be attributed to reduced amyloid burden and GSK-3β activity. GSK-3β is strongly connected to AD cardinal features and is a potential link between Aβ and tau.17 Phosphorylation of APP is partially regulated by GSK-3β.18 Phosphorylation of tau is enhanced by Aβ by activating GSK-3β.19 From the present investigation, we observed that SLM acts on Aβ and GSK-3β to show neuroprotection. The activity on GSK-3β needs further investigation to elucidate other possible mechanisms of SLM. The reduction of GSK-3β activity might also be due to SLM action of neuro-inflammation, which is a possible trigger for hyperphosphorylation of tau.20 Neuro-inflammation is another important pathological feature of AD. Inflammation initiates to counteract the production of amyloid plaques, which causes the death of neurons due to the production of cytokines and neurotoxins.19 Amyloid β peptides persistently activate microglia in AD and lead to increased release of pro-inflammatory cytokines, which finally causes reduced Aβ clearance.21−23 Microglia affects the Aβ clearance and supports the aggregation process.24 In drug



METHODS

Cell Culture. Human neuroblastoma cell line (SH-SY5Y) supplied from the American Type Culcure Collection was cultured in Dulbecco’s modified Eagle medium (DMEM) supplemented with fetal bovine serum (10%) and penicillin/streptomycin (1%) at 37 °C in a humidified atmosphere with CO 2 (5%). Aβ42 peptide oligomerization was carried out according to a published report.34 Briefly, lyophilized synthetic Aβ42 was dissolved in HFIP (1,1,1,3,3,3hexafluoro-2-propanol). The peptides were reconstituted in DMSO (dimethyl sulfoxide) following HFIP evaporation to form 5 mM suspension. This 5 mM solution was further diluted to 100 μM in DMEM medium and incubated at 37 °C for 5 days. SH-SY5Y cells were seeded onto a 96-well plate at 10,000 cells per well. One day (24 h) later, Aβ42 oligomer and SLM were added to the cells and incubated for another 24 h. Afterward, SH-SY5Y cells were incubated with 0.5 mg/mL MTT (M6494, Life Technologies) for 4 h at 37 °C. Then, 100 μL of DMSO was added to dissolve the formazan crystals, F

DOI: 10.1021/acschemneuro.6b00388 ACS Chem. Neurosci. XXXX, XXX, XXX−XXX

Research Article

ACS Chemical Neuroscience

Figure 10. SLM reduces tau phosphorylation and GSK-3β activity following one month of treatment. Significance was analyzed by one-way ANOVA (n = 5, mean ± SEM) followed by Newman−Keuls post hoc test using Graphpad Prism. *p < 0.05 and **p < 0.01, compared with control mice. (A) Representative immunoreactive species of p-tau, p-GSK-3β, t-GSK-3β, and GAPDH in control and treated mice and (B) their corresponding densities. Lane 1, Control; Lane 2, SLM 1 mg/kg; Lane 3, SLM 2 mg/kg; Lane 4, SLM 4 mg/kg. optimal temperature and humidity. Body weight was recorded weekly once during the treatment period (Figure 13). Morris Water Maze (MWM). Following one month of SLM treatment, learning and memory skills of animals were confirmed on the Morris water maze (MWM) according to Branca and colleagues.38 Briefly, the animals were freely allowed to find a submerged escape platform in opaque water of a circular pool and the latency to reach the platform was measured from day 1 to 4. Throughout the experiment, the temperature of the pool (25 °C) and the location of the platform (14 in. from the southwest (SW) wall) were maintained constant. Mice were trained for 4 days with 4 trials per day and were assigned four pseudorandom starting points (E, SE, NW, and N) each day. A total of 16 trials were performed with each mouse in the acquisition phase. The animals were positioned facing the wall at one of the 4 starting positions and allowed to locate the submerged platform, 1.5 cm beneath the surface of the water. A successful trial was recorded when a mouse found the platform within 60 s. The mouse was allowed to stay on the platform for 5 s prior to being returned to its home cage. If a mouse was unable to locate the platform within 60 s, it was gently guided to the platform and allowed to stay on it for 15 s. The time required (latency) to find the escape platform was measured. A 60 s probe trial was performed without the escape platform on day 5. Time spent in the target quadrant where the escape platform was positioned and number of platform crossings were recorded during this probe trial. Y-Maze. The Y-maze apparatus consists of 3 arms of equal-length prepared with white PVC. Each arm was 35 cm long, 5 cm wide, and 10 cm high. Arms of the Y-maze are labeled A, B, and C and are positioned at an equal angle of 120° from each other and extend from the central platform. The test was performed as given by Sarnyai and colleagues.39 Briefly, one arm (C) was closed during the training phase and kept constant for all the groups. The animal was positioned in one of the open arms (A) and allowed 15 min to explore the accessible arms (A and B) of the Y-maze. The maze was cleaned with 75% alcohol between trials to avoid olfactory clues. All mice were returned to their home cages and given an hour intertrial interval before commencing the next phase. During the test phase mice were allowed to explore the maze for 5 min with all arms open, and the time spent in the novel arm (C) was recorded.

Figure 11. SLM has no activity on APP and BACE1 following one month of treatment. Lane 1, Control; Lane 2, SLM 1 mg/kg; Lane 3, SLM 2 mg/kg; Lane 4, SLM 4 mg/kg. followed by the absorbance measurement at 570 nm in a microplate reader. Cytotoxicity was also assessed using LDH assay kit (11644793001, Roche Life Science). The LDH activity of the cultured medium was determined according to the manufacturer’s instructions. Cell survival rates were expressed as percentages of the control group. Total cell death was tested by a commercially available annexin V− FITC apoptosis detection kit (640914, Biolegend). Following SLM and Aβ42 treatment, SH-SY5Y cells were rinsed with PBS twice, spun for 10 min at 600 × g and resuspended in buffer (0.5 mL) containing annexin V (5 μL) and propidium iodide (5 μL). The final solution was incubated at 37 °C for 15 min in dark. Flow cytometer (BD ACCURI C6) was used to determine the apoptotic rate. Each assay including MTT, LDH, and cell death were carried out using at least six wells of three independent experiments. Animals and Treatment. Female 3xTg-AD mice (10 months old) (n = 32) mutated with Psen1M146V, APPSWE, and tauP301L were procured from The Jackson Laboratory (34830-JAX/3xTg-AD) with prior approval from Institutional Animal Ethical Committee. Reports indicate that female 3xTg-AD mice show greater pathological features than males,35−37 and therefore we only used female mice in our experiments. Mice were randomized and divided into 4 groups (8 mice in each), one control and three experimental groups. Mice from experimental groups received 1, 2, and 4 mg/kg of SLM via intraperitoneal route, while the control mice received DMSO (2% in phosphate buffer saline) once daily for one month. All the mice were housed in individual ventilated cages in a room maintained with G

DOI: 10.1021/acschemneuro.6b00388 ACS Chem. Neurosci. XXXX, XXX, XXX−XXX

Research Article

ACS Chemical Neuroscience

Figure 12. SLM reduces neuro-inflammation following one month of treatment. Significance was analyzed by one-way ANOVA (n = 5, mean ± SEM) followed by Newman−Keuls post hoc test test using Graphpad Prism. *p < 0.05 and **p < 0.01, compared with control mice. (A) Representative immunoreactive species of GFAP, Iba1, and GAPDH in control and treated mice and (B) their corresponding densities. Lane 1, Control; Lane 2, SLM 1 mg/kg; Lane 3, SLM 2 mg/kg; Lane 4, SLM 4 mg/kg. Western Blotting. Brains were homogenized in lysis buffer (10 mM Tris, 150 mM NaCL, 0.5% deoxycholate, 0.5% NP40, 5 mM EDTA, pH 7.4) including protease inhibitor cocktail (05892791001, Roche) and phosphatase inhibitor cocktail (04906845001, Roche). The homogenate was spun for 1 h at 21 000 × g (4 °C), the resultant supernatant was separated and spun again for 30 min at 21 000 × g (4 °C), and the final supernatant was stored at −80 °C until use. All samples were mixed with sample buffer (250 mM Tris-HCl pH 6.8, 8% SDS, 30% glycerol, 10% β-mercaptoethanol, and 0.02% bromphenol blue), boiled for 5 min at 95 °C, loaded on a bis(acrylamide) gel (10%), and electrophoresed. Proteins were transferred onto poly(vinylidene difluoride) (PVDF) membrane and incubated with the following primary antibodies: anti-tau antibody (MAB361, Millipore), phosphor-PHF-tau antibody (MN1020, Thermofisher), anti-GFAP antibody (MAB360, Millipore), anti-APP antibody (2452, Cell Signaling Technology), anti-Iba1 antibody (019-19741, Wako), antiBACE1 antibody (5606, Cell Signaling Technology), anti-GSK-3β antibody (9315, Cell Signaling Technology), anti-p-GSK-3β (9323,Cell Signaling Technology), and anti-GAPDH antibody (MAB374, Millipore). Membranes were blocked with nonfat milk in TBS-T (Tris buffered saline with Tween 20) for 1 h. Primary antibodies were added in nonfat milk and incubated overnight at 4 °C, followed by washing with TBS-T (3 washes, 5 min each), and incubated with HRP-conjugated secondary antibodies for 1 h at room temperature. Membranes were washed with TBS-T (3 washes, 5 min each), and immunoreactivity was detected by ECL (6883S, Cell Signaling Technology). Immunoreactive species were visualized on a Bio-Rad system using Imagelab version 5.1 software and quantified using ImageJ software. Immunohistochemistry. Left hemispheres were postfixed in paraformaldehyde (4%) for 48 h and transferred into a sucrose solution (30%) for 72 h before sectioning. Frozen sections (40 μm thick) were obtained using a cryostat. Sections were washed with TBS (100 mM Tris pH 7.5, 150 mM NaCl) twice for 5 min each. Quenching was performed by incubating the section with 90% formic acid for 7 min followed by 30 min incubation with hydrogen peroxide (3%). Sections were blocked by incubating in TBS-A (100 mM Tris pH 7.5, 150 mM NaCl, 0.1% Triton X-100) for 15 min and TBS-B (100 mM Tris pH 7.5, 150 mM NaCl, 0.1% Triton X-100, 2% bovine serum albumin) for 30 min. Primary antibody, anti-Aβ42 (Biolegend, 805502, 1:200), was diluted with TBS-B, and sections were incubated overnight at 4 °C.38 Sections were washed twice in TBS and incubated with biotinylated secondary antibody (31800, Thermofisher) for 30

Figure 13. Weekly body weight during SLM treatment. Spontaneous Alternation. A 6 min test was performed to measure spatial memory in the Y-maze by placing the mouse in the center of the maze and allowing the animal to freely explore. The maze was cleaned with 75% alcohol between each trial to prevent olfactory clues. The arm entry sequence of each mouse was manually recorded. A successful alteration is represented by the animal visiting all three arms consecutively (ABC, ACB, BCA, BAC, CAB, and CBA). Percentage of alternation was calculated according to the formula: spontaneous alternation/total possible alternations × 100. Concentration of SLM in the Brain. The structure of SLM is given in Figure 1 and was synthesized as described previously.7 The brain samples were collected 1 h after ip injection at the end of treatment (day 30). After weighing, brain sample was homogenized with deionized water (1:10, v/v) using tissue homogenizer and stored at −80 °C until analysis. The analysis was performed according to our previous report.8 Biochemical Estimations. For biochemical characterization, 3xTg-AD mice were administered SLM or vehicle daily for one month. Upon completion of the treatment, animals were anesthetized and perfused with phosphate buffer saline (PBS). Brains were immediately isolated and divided along the sagittal line into two hemispheres; the right hemisphere was stored at −80 °C for ELISA and Western blot analysis, and the left hemisphere was postfixed in 4% paraformaldehyde for histological analysis. Enzyme-Linked Immunosorbent Assay (ELISA). Aβ42 and Aβ40 ELISA kits were obtained from Invitrogen (KMB3441 and KMB3481). Plates were analyzed on a multimode microplate plate reader (Molecular Devices, Berkshire, UK). H

DOI: 10.1021/acschemneuro.6b00388 ACS Chem. Neurosci. XXXX, XXX, XXX−XXX

Research Article

ACS Chemical Neuroscience min. Sections were detected using an ABC kit (32020, Thermofisher) and visualized with DAB quanto (TA-060-QHDX, Thermofisher). Thioflavin S Staining and Microscopy. Immunofluorescence staining was performed according to a previous published report.40 Briefly, sections were incubated in Thioflavin S solution for 30 min followed by washing in water, three times (2 min each) and finally by 6 min incubation in 80% ethanol. Fluorescence images were obtained using fluorescence microscope (Mirax Scan, Carl Zeiss MicroImaging GmbH, Jena, Germany). Statistical Analyses. Statistical analyses were performed using GraphPad Prism (version 6.01, San Diego, CA, USA) ,and data were expressed as mean ± standard error of the mean (SEM). One-way analysis of variance (ANOVA) followed by Newman−Keuls post hoc test was used for all the data analysis. Data equal to or lesser than p < 0.05 was considered significant.



(4) Cummings, J. L., Morstorf, T., and Zhong, K. (2014) Alzheimer’s disease drug-development pipeline: few candidates, frequent failures. Alzheimer's Res. Ther. 6, 37. (5) Jack, C. R., Knopman, D. S., Jagust, W. J., Shaw, L. M., Aisen, P. S., Weiner, M. W., Petersen, R. C., and Trojanowski, J. Q. (2010) Hypothetical model of dynamic biomarkers of the Alzheimer’s pathological cascade. Lancet Neurol. 9, 119−28. (6) Deng, Y., Xiong, Z., Chen, P., Wei, J., Chen, S., and Yan, Z. (2014) β-amyloid impairs the regulation of N-methyl-D-aspartate receptors by glycogen synthase kinase 3. Neurobiol. Aging 35, 449−59. (7) Yang, W., Wong, Y., Ng, O. T. W., Bai, L.-P., Kwong, D. W. J., Ke, Y., Jiang, Z.-H., Li, H.-W., Yung, K. K. L., and Wong, M. S. (2012) Inhibition of beta-amyloid peptide aggregation by multifunctional carbazole-based fluorophores. Angew. Chem., Int. Ed. 51, 1804−10. (8) Zhou, W., Wu, X., Li, J., Hu, X., and Tam, K. Y. (2016) Quantification of permanent positively charged compounds in plasma using one-step dilution to reduce matrix effect in MS. Bioanalysis 8, 497−509. (9) Giménez-Llort, L., Blázquez, G., Cañete, T., Johansson, B., Oddo, S., Tobeña, A., LaFerla, F. M., and Fernández-Teruel, A. (2007) Modeling behavioral and neuronal symptoms of Alzheimer’s disease in mice: a role for intraneuronal amyloid. Neurosci. Biobehav. Rev. 31, 125−47. (10) Vassar, R., and Citron, M. (2000) Abeta-generating enzymes: recent advances in beta- and gamma-secretase research. Neuron 27, 419−22. (11) Cheng, N., Jiao, S., Gumaste, A., Bai, L., and Belluscio, L. (2016) APP Overexpression Causes Aβ-Independent Neuronal Death through Intrinsic Apoptosis Pathway. eNeuro, DOI: 10.1523/ENEURO.015016.2016. (12) Carriba, P., Jimenez, S., Navarro, V., Moreno-Gonzalez, I., Barneda-Zahonero, B., Moubarak, R. S., Lopez-Soriano, J., Gutierrez, A., Vitorica, J., and Comella, J. X. (2015) Amyloid-β reduces the expression of neuronal FAIM-L, thereby shifting the inflammatory response mediated by TNFα from neuronal protection to death. Cell Death Dis. 6, e1639. (13) Ryan, K. A., and Pimplikar, S. W. (2005) Activation of GSK-3 and phosphorylation of CRMP2 in transgenic mice expressing APP intracellular domain. J. Cell Biol. 171, 327−35. (14) Lucas, J. J. (2001) Decreased nuclear beta-catenin, tau hyperphosphorylation and neurodegeneration in GSK-3beta conditional transgenic mice. EMBO J. 20, 27−39. (15) Beurel, E., and Jope, R. S. (2006) The paradoxical pro- and antiapoptotic actions of GSK3 in the intrinsic and extrinsic apoptosis signaling pathways. Prog. Neurobiol. 79, 173−89. (16) Lee, V. M.-Y., and Trojanowski, J. Q. (2006) Progress from Alzheimer’s tangles to pathological tau points towards more effective therapies now. J. Alzheimers. Dis. 9, 257−62. (17) Flaherty, D. B., Soria, J. P., Tomasiewicz, H. G., and Wood, J. G. (2000) Phosphorylation of human tau protein by microtubuleassociated kinases: GSK3beta and cdk5 are key participants. J. Neurosci. Res. 62, 463−72. (18) Aplin, A. E., Gibb, G. M., Jacobsen, J. S., Gallo, J. M., and Anderton, B. H. (1996) In vitro phosphorylation of the cytoplasmic domain of the amyloid precursor protein by glycogen synthase kinase3beta. J. Neurochem. 67, 699−707. (19) Mokhtar, S. H., Bakhuraysah, M. M., Cram, D. S., and Petratos, S. (2013) The Beta-amyloid protein of Alzheimer’s disease: communication breakdown by modifying the neuronal cytoskeleton. Int. J. Alzheimer's Dis. 2013, 910502. (20) Metcalfe, M. J., and Figueiredo-Pereira, M. E. (2010) Relationship between tau pathology and neuroinflammation in Alzheimer’s disease. Mt. Sinai J. Med. 77, 50−8. (21) Yu, Y., and Ye, R. D. (2015) Microglial Aβ receptors in Alzheimer’s disease. Cell. Mol. Neurobiol. 35, 71−83. (22) White, J. A., Manelli, A. M., Holmberg, K. H., Van Eldik, L. J., and Ladu, M. J. (2005) Differential effects of oligomeric and fibrillar amyloid-beta 1−42 on astrocyte-mediated inflammation. Neurobiol. Dis. 18, 459−65.

AUTHOR INFORMATION

Corresponding Author

*Kin Yip Tam. E-mail: [email protected]. Tel: +853 88224988. Fax: +853 88222314. ORCID

Kin Yip Tam: 0000-0001-5507-8524 Author Contributions

X.W. and J.K. are co-first authors. X.W. and J.K. performed experiments, including animal work, biochemistry, and histology, and analyzed the data. J.K. composed and edited the manuscript. W.Z. performed the LC-MS study. K.Y.T. supervised all phases of the study. All authors discussed the results and commented on the final version of the manuscript. Funding

We appreciate the financial support from the Science and Technology Development Fund, Macao S.A.R (FDCT) (118/ 2013/A3). Notes

The authors declare no competing financial interest.



ABBREVIATIONS AD, Alzheimer’s disease; ANOVA, analysis of variance; APP, amyloid precursor protein; Aβ, amyloid β; BACE1, β-amyloid cleavage enzyme 1; DAB, diaminobenzidine; DMEM, Dulbecco’s modified Eagle medium; DMSO, dimethyl sulfoxide; ELISA, enzyme-linked immunosorbent assay; GFAP, glial fibrillary acidic protein; GSK-3β, glycogen synthase kinase-3β; HFIP, 1,1,1,3,3,3-hexafluoro-2-propanol; Iba1, ionized calciumbinding adapter molecule 1; LDH, lactate dehydrogenase; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; MWM, Morris water maze; PBS, phosphate buffer saline; PVDF, poly(vinylidene difluoride); SH-SY5Y, human neuroblastoma cell line; TBS, Tris-buffered saline



REFERENCES

(1) Habchi, J., Arosio, P., Perni, M., Costa, A. R., Yagi-Utsumi, M., Joshi, P., Chia, S., Cohen, S. I. A., Müller, M. B. D., Linse, S., Nollen, E. A. A., Dobson, C. M., Knowles, T. P. J., and Vendruscolo, M. (2016) An anticancer drug suppresses the primary nucleation reaction that initiates the production of the toxic Aβ42 aggregates linked with Alzheimer’s disease. Sci. Adv. 2, e1501244. (2) Klein, W. L., Krafft, G. A., and Finch, C. E. (2001) Targeting small Abeta oligomers: the solution to an Alzheimer’s disease conundrum? Trends Neurosci. 24, 219−24. (3) McKoy, A. F., Chen, J., Schupbach, T., and Hecht, M. H. (2012) A novel inhibitor of amyloid β (Aβ) peptide aggregation: from high throughput screening to efficacy in an animal model of Alzheimer disease. J. Biol. Chem. 287, 38992−9000. I

DOI: 10.1021/acschemneuro.6b00388 ACS Chem. Neurosci. XXXX, XXX, XXX−XXX

Research Article

ACS Chemical Neuroscience

serotonin(1A) receptors. Proc. Natl. Acad. Sci. U. S. A. 97, 14731− 14736. (40) Yousefi, B. H., von Reutern, B., Scherübl, D., Manook, A., Schwaiger, M., Grimmer, T., Henriksen, G., Förster, S., Drzezga, A., and Wester, H.-J. (2015) FIBT versus florbetaben and PiB: a preclinical comparison study with amyloid-PET in transgenic mice. EJNMMI Res. 5, 20.

(23) Craft, J. M., Watterson, D. M., Frautschy, S. A., and Van Eldik, L. J. (2004) Aminopyridazines inhibit beta-amyloid-induced glial activation and neuronal damage in vivo. Neurobiol. Aging 25, 1283−92. (24) Farfara, D., Trudler, D., Segev-Amzaleg, N., Galron, R., Stein, R., and Frenkel, D. (2011) γ-Secretase component presenilin is important for microglia β-amyloid clearance. Ann. Neurol. 69, 170−80. (25) Wu, X., Li, J., Zhou, W., and Tam, K. (2015) Animal models for Alzheimer’s disease: a focused review of transgenic rodent models and behavioral assessment methods. ADMET DMPK 3, 242−253. (26) Stevens, L. M., and Brown, R. E. (2015) Reference and working memory deficits in the 3xTg-AD mouse between 2 and 15-months of age: A cross-sectional study. Behav. Brain Res. 278, 496−505. (27) Billings, L. M., Oddo, S., Green, K. N., McGaugh, J. L., and LaFerla, F. M. (2005) Intraneuronal Aβ causes the onset of early Alzheimer’s disease-related cognitive deficits in transgenic mice. Neuron 45, 675−688. (28) De Felice, F. G., Vieira, M. N. N., Saraiva, L. M., Figueroa-Villar, J. D., Garcia-Abreu, J., Liu, R., Chang, L., Klein, W. L., and Ferreira, S. T. (2004) Targeting the neurotoxic species in Alzheimer’s disease: inhibitors of Abeta oligomerization. FASEB J. 18, 1366−72. (29) Dodart, J.-C., Bales, K. R., Gannon, K. S., Greene, S. J., DeMattos, R. B., Mathis, C., DeLong, C. A., Wu, S., Wu, X., Holtzman, D. M., and Paul, S. M. (2002) Immunization reverses memory deficits without reducing brain Abeta burden in Alzheimer’s disease model. Nat. Neurosci. 5, 452−7. (30) Kotilinek, L. a, Bacskai, B., Westerman, M., Kawarabayashi, T., Younkin, L., Hyman, B. T., Younkin, S., and Ashe, K. H. (2002) Reversible memory loss in a mouse transgenic model of Alzheimer’s disease. J. Neurosci. 22, 6331−5. (31) Birmingham, K., and Frantz, S. (2002) Set back to Alzheimer vaccine studies. Nat. Med. 8, 199−200. (32) Chen, Y., Liang, Z., Tian, Z., Blanchard, J., Dai, C.-L., Chalbot, S., Iqbal, K., Liu, F., and Gong, C.-X. (2014) Intracerebroventricular streptozotocin exacerbates Alzheimer-like changes of 3xTg-AD mice. Mol. Neurobiol. 49, 547−62. (33) Chen, Y., Liang, Z., Blanchard, J., Dai, C.-L., Sun, S., Lee, M. H., Grundke-Iqbal, I., Iqbal, K., Liu, F., and Gong, C.-X. (2013) A nontransgenic mouse model (icv-STZ mouse) of Alzheimer’s disease: similarities to and differences from the transgenic model (3xTg-AD mouse). Mol. Neurobiol. 47, 711−25. (34) Soura, V., Stewart-Parker, M., Williams, T. L., Ratnayaka, A., Atherton, J., Gorringe, K., Tuffin, J., Darwent, E., Rambaran, R., Klein, W., Lacor, P., Staras, K., Thorpe, J., and Serpell, L. C. (2012) Visualization of co-localization in Aβ42-administered neuroblastoma cells reveals lysosome damage and autophagosome accumulation related to cell death. Biochem. J. 441, 579−90. (35) Hirata-Fukae, C., Li, H. F., Hoe, H. S., Gray, A. J., Minami, S. S., Hamada, K., Niikura, T., Hua, F., Tsukagoshi-Nagai, H., HorikoshiSakuraba, Y., Mughal, M., Rebeck, G. W., LaFerla, F. M., Mattson, M. P., Iwata, N., Saido, T. C., Klein, W. L., Duff, K. E., Aisen, P. S., and Matsuoka, Y. (2008) Females exhibit more extensive amyloid, but not tau, pathology in an Alzheimer transgenic model. Brain Res. 1216, 92− 103. (36) Carroll, J. C., Rosario, E. R., Kreimer, S., Villamagna, A., Gentzschein, E., Stanczyk, F. Z., and Pike, C. J. (2010) Sex differences in β-amyloid accumulation in 3xTg-AD mice: Role of neonatal sex steroid hormone exposure. Brain Res. 1366, 233−245. (37) Clinton, L. K., Billings, L. M., Green, K. N., Caccamo, A., Ngo, J., Oddo, S., McGaugh, J. L., and LaFerla, F. M. (2007) Age-dependent sexual dimorphism in cognition and stress response in the 3xTg-AD mice. Neurobiol. Dis. 28, 76−82. (38) Branca, C., Wisely, E. V., Hartman, L. K., Caccamo, A., and Oddo, S. (2014) Administration of a selective β2 adrenergic receptor antagonist exacerbates neuropathology and cognitive deficits in a mouse model of Alzheimer’s disease. Neurobiol. Aging 35, 2726−35. (39) Sarnyai, Z., Sibille, E. L., Pavlides, C., Fenster, R. J., McEwen, B. S., and Toth, M. (2000) Impaired hippocampal-dependent learning and functional abnormalities in the hippocampus in mice lacking J

DOI: 10.1021/acschemneuro.6b00388 ACS Chem. Neurosci. XXXX, XXX, XXX−XXX