NGF‐Dependent Activation of TrkA Pathway: A Mechanism for the Neuroprotective Effect of Troxerutin in D‐Galactose‐Treated Mice

Abstract

D‐galactose‐(D‐gal)‐treated mouse, with cognitive impairment, has been used for neurotoxicity investigation and anti‐neurotoxicity pharmacology research. In this study, we investigated the mechanism underlying the neuroprotective effect of troxerutin. The results showed that troxerutin improved behavioral performance in D‐gal‐treated mice by elevating Cu, Zn‐superoxide dismutases (Cu, Zn‐SOD) activity and decreasing reactive oxygen species levels. Furthermore, our results showed that troxerutin significantly promoted nerve growth factor (NGF) mRNA expression which resulted in TrkA activation. On one hand, NGF/TrkA induced activation of Akt and ERK1/2, which led to neuronal survival; on the other hand, NGF/TrkA mediated CaMKII and CREB phosphorylation and increased PSD95 expression, which improved cognitive performance. However, the neuroprotective effect of troxerutin was blocked by treatment with K252a, an antagonist for TrkA. No neurotoxicity was observed in mice treated with K252a or troxerutin alone. In conclusion, administration of troxerutin to D‐gal‐injected mice attenuated cognitive impairment and brain oxidative stress through the activation of NGF/TrkA signaling pathway.

INTRODUCTION

Reactive oxygen species (ROS) include oxygen ions, free radicals and peroxides. They are small, high reactive and oxygen‐containing molecules that are generated during normal metabolism of oxygen and have important roles in cell signaling (15). However, during the process of environmental stress, ROS levels can increase dramatically, which can damage various biological macromolecules such as DNA, RNA, proteins and lipids 15, 40, 59, 65. This process is known as oxidative stress 49, 62. Recent evidence shows that oxidative stress is involved in the pathogenesis of many diseases, including cancer, neurodegenerative diseases and many psychiatric problems 20, 32, 33, 44, 50, 51, particularly in neurological and psychiatric diseases because of the central nervous system's vulnerability to oxidative stress (58). D‐galactose (D‐gal) is a naturally occurring substance in the body; however, at high levels, it has neurotoxic efficacy and it can be converted to aldose and hydrogen peroxide via the action of galactose oxidase, which causes the accumulation of ROS, or can stimulate free radical production indirectly by the formation of advanced glycation endproducts in vivo, finally resulting in oxidative stress 38, 40, 65, 68. A large number of experiments have further demonstrated that continuous subcutaneous injection of D‐gal in rodent induced neuropathological changes including cholinergic neuronal loss, decreased neurogenesis and alteration of synapse structure in the hippocampus 9, 24, 40, 67, 70. Further investigations show that these pathology lesions result in severely impaired learning and memory function 9, 12, 24, 38, 40, 68, 69. In addition, these changes also include decreased anti‐oxidative enzyme activity and increased lipid peroxidation products 11, 12, 22, 63, which increase cell karyopyknosis, apoptosis and caspase‐3 protein levels in hippocampal neurons 12, 22, 40. Moreover, D‐gal‐induced neurotoxicity has been implicated in calcium homeostasis impairment, mitochondrial dysfunction, low immune response and a shortened lifespan 11, 14, 37, 38, 56. The above findings suggest that oxidative stress induced by D‐gal may contribute to the neurodegeneration in the brain. As a result, D‐gal‐treated mice have been increasingly used as a model for studying the mechanism of oxidative stress‐induced brain aging and screening neuroprotective drugs.

Flavonoids are a group of naturally occurring compounds widely distributed in most plants (10). Evidence has shown that flavonoids have many biological effects, such as anti‐oxidative and anti‐inflammatory effects. Troxerutin is a derivative of the natural bioflavonoid rutin which has been reported to exert neuroprotection through scavenging ROS and consequently attenuating global cerebral ischemia and reperfusion‐induced neuronal injury 19, 48. Our latest findings have further demonstrated that troxerutin has a protective effect against D‐gal‐induced oxidative damage in mouse brain (42). In the present study, we investigated the mechanism underlying the neuroprotective effect of troxerutin.

MATERIALS AND METHODS

Animals and treatment

Sixty‐three 8‐week‐old male Kunming strain mice were purchased from the Branch of National Breeder Center of Rodents (Shanghai). Prior to experiments, the mice had free access to food and water and were kept under constant conditions of temperature (23 ± 1°C) and humidity (60%). Nine mice were housed per cage on a 12‐h light/dark schedule (lights on 08:30–20:30) and divided randomly into seven groups for 1 week of adaptation.

For intracerebroventricular microinjection, mice were equipped with stainless steel cannulas. Mice were housed individually to prevent them from damaging each other's cannula. All surgical procedures were performed under Equithesin anesthesia (3 mL/kg i.p.) (23). Animals were placed in a stereotaxic apparatus (ZH‐LanXing B/S type, Huaibei, China) and a 26‐gauge stainless steel guide cannula (Plastics One, Roanoke, VA, USA) was chronically implanted into the lateral ventricle using the following coordinates relative to bregma: 0.5 mm anterior‐posterior, 1.0 mm mediolateral and 2.0 mm depth obtained from Paxinos and Watson (1997) (46). A 28‐gauge dummy cannula was inserted to prevent clogging of the guide cannula. For all surgeries, a proper and sterile surgical environment was maintained.

Six days after surgery, groups 1, 5 and 6 served as vehicle control with injection of saline (0.9%) only, and the other four groups of mice (groups 2, 3, 4 and 7) received daily subcutaneous injection of D‐gal (Sigma‐Aldrich, MO, USA) at dose of 500 mg/(kg day) for 8 weeks, respectively. At the same time, mice in groups 3, 4 and 6 received troxerutin (troxerutin; Purity >99%; Baoji Fangsheng Biotechnology Co., Ltd, Baoji, China) of 150 mg/(kg day) in distilled water by oral gavage for 8 weeks, and the mice of groups 1, 2, 5 and 7 were given distilled water orally at the same dose. Thirty minutes after troxerutin treatment, 12 µg of K252a (Merck, Darmstadt, Germany) dissolved in 100 µL of the solvent consisting of 75% sterile saline/25% dimethyl sulfoxide was given 10 minutes in group 4, 5 and 7 by means of intracerebroventricular (ICV) infusion. The other four groups (groups 1, 2, 3 and 6) were infused with an equal volume of the solvent, respectively. ICV infusion of K252a was daily performed using a microinjector (KD Scientific Inc., Holliston, MA, USA) at a rate of 3 µL/min for 8 weeks. The dosage of D‐gal and troxerutin in the experiments was based on our previous report, and the dosage of K252a was based on our pilot study data (42). All experiments were performed in compliance with the Chinese legislation on the use and care of laboratory animals and were approved by the respective university committees for animal experiments. After the behavioral testing, mice were sacrificed and brain tissues were immediately collected for experiments or stored at −70°C for later use.

Behavioral tests

Step‐through test

The step‐through passive avoidance apparatus consisted of an illuminated chamber (11.5 cm × 9.5 cm × 11 cm) attached to a darkened chamber (23.5 cm × 9.5 cm × 11 cm) containing a metal floor that could deliver footshocks. The two compartments were separated by a guillotine door. The illuminated chamber was lit with a 25‐W lamp. The step‐through test was performed as described previously 38, 39. Mice were placed in the dimly lit room containing the apparatus 0.5 h before training to acclimatize to the new environment. Each mouse was then placed individually into the illuminated chamber, facing away from the door to the dark chamber, and allowed to acclimatize for 1 minute. When the mouse was observed to turn its body fully away from the dark chamber, the door was raised; when the mouse next turned fully toward the darkened chamber, the timer was started. An initial time measure was from the time that the mouse faced the opened darkened chamber to the time that the mouse fully entered, with all four paws, the dark chamber. As soon as the mouse entered the dark chamber, the door was slid back into place, triggering a mild footshock (0.3 mA, 50 Hz, 5 s). The mouse was then immediately removed from the chamber and returned to its home cage. The retention test was conducted 24 h later with the mouse again being placed in the illuminated chamber and subjected to the same protocol described above in the absence of footshock. The latency to enter the dark compartment with four paws was measured. The upper time limit was set at 300 s.

Morris water maze test

The Morris water maze (MWM) test was performed as described previously 38, 39. The experimental apparatus consisted of a circular water tank (100 cm in diameter, 35 cm in height), containing water (23 ± 1°C) to a depth of 15.5 cm, which was rendered opaque by adding ink. A platform (4.5 cm in diameter, 14.5 cm in height) was submerged 1 cm below the water surface and placed at the midpoint of one quadrant. The pool was located in a test room, which contained various prominent visual cues. Each mouse received four training periods per day for 4 consecutive days. Latency to escape from the water maze (finding the submerged escape platform) was calculated for each trial. On day 5, the probe test was carried out by removing the platform and allowing each mouse to swim freely for 60 s. The time that the mice spent swimming in the target quadrant (where the platform was located during hidden platform training), and in the three nontarget quadrants (right, left and opposite quadrants), was measured, respectively. For the probe trials, the number of times crossing over the platform site of each mouse was also measured and calculated. All data were recorded with a computerized video system.

Preparation of tissue samples

Tissue homogenates

For biochemical studies performed as described previously 38, 39, 40, animals were deeply anesthetized and sacrificed. Two hemispheres were separated and weighed on ice. Left hemispheres were homogenized in 1/5 (w/v) ice‐cold phosphate buffer saline solution (50 mM PBS, pH 7.4) containing a protease inhibitor cocktail (Sigma‐Aldrich) with 10 strokes at 1200 rpm in a Potter homogenizer. Homogenates were divided into two portions and one part was directly centrifuged at 8000 × g for 10 minutes to obtain the supernatant. Supernatant aliquots were used to determine brain ROS levels and protein contents. The second part of homogenates was sonicated four times for 30 s with 20‐s intervals using a VWR Bronson Scientific sonicator, centrifuged at 5000 × g for 10 minute at 4°C, and the supernatants were collected and stored at −70°C for determination of Cu, Zn‐superoxide dismutases (Cu, Zn‐SOD) enzyme activities. For Western blot analysis performed as described previously 40, 65, right hemispheres were homogenized in 1/3 (w/v) ice cold RIPA lysis buffer [1 × Tris Buffered Saline (TBS), 1% NP‐40, 0.5% sodium deoxycholate, 0.1% SDS, 0.004% sodium azide] combining 30 µL of 10 mg/mL PMSF solution, 30 µL of Na₃VO₄, 30 µL of NaF and 30 µL of protease inhibitors cocktail per gram of tissue. The homogenates were sonicated four times for 30 s with 20‐s intervals using a VWR Bronson Scientific sonicator, centrifuged at 15000 × g for 10 minutes at 4°C, and the supernatant was collected and stored at −70°C for Western blot studies. Protein levels in the supernatants were determined using the BCA assay kit (Pierce Biotechnology, Inc., Rockford, IL, USA).

Collection of brain slice

The mice were perfused transcardially with 25 mL of normal saline (0.9%). The brain tissues were fixed in a fresh solution of 4% paraformaldehyde (pH 7.4) at 4°C for 4 h, incubated overnight at 4°C in 100 mM sodium phosphate buffer (pH 7.4) containing 30% sucrose; and embedded in Optimal Cutting Temperature (OCT, Lecia, CA, Germany). Coronal sections (12 µm) from cryofixed tissue were collected on 3‐aminopropyl‐trimethoxysilane‐coated slides (Sigma‐Aldrich) and stored at −70°C.

Biochemical assays

Assay of Cu, Zn‐SOD activity

Chemicals used in the assay, including xanthine, xanthine oxidase, cytochrome c, bovine serum albumin (BSA) and SOD, were purchased from Sigma Chemical Company (St. Louis, MO, USA). Cu,Zn‐SOD activity was measured according to the method described by Lu et al 38, 40. Solution A was prepared by mixing 100 mL of 50 mM PBS (pH 7.4) containing 0.1 mM EDTA and 2 µmol of cytochrome c with 10 mL of 0.001 N NaOH solution containing 5 µmol of xanthine. Solution B contained 0.2 U xanthine oxidase/mL and 0.1 mM EDTA. Fifty microliters of a tissue supernatant was mixed with 2.9 mL of solution A and the reaction was started by adding 50 µL of solution B. Change in absorbance at 550 nm was monitored in a spectrophotometer (Shimadzu UV‐2501PC, Shimadzu Corp., Kyoto, Japan). A blank was run by replacing the supernatant with 50 µL of ultra pure water. Cu, Zn‐SOD level was expressed as units per mg protein with reference to the activity of a standard curve of bovine copper‐, zinc‐, SOD under the same conditions.

Assay of ROS

ROS was measured as described previously, based on the oxidation of 2′7′‐dichlorodihydrofluorescein diacetate (DCFH‐DA) to 2′7′‐dichlorofluorescein (DCF) (42). Briefly, the homogenate was diluted 1:20 times with ice‐cold Locke's buffer (154 mM NaCl, 5.6 mM KCl, 3.6 mM NaHCO₃, 2.0 mM CaCl₂, 10 mM d‐glucose, and 5 mM HEPES, pH 7.4) to obtain a concentration of 5 mg tissue/mL. The reaction mixture (1 mL) containing Locke's buffer (pH 7.4), 0.2 mL homogenate and 10 mL of DCFH‐DA (5 mM) was incubated for 15 minutes at room temperature to allow the DCFH‐DA to be incorporated into any membrane‐bound vesicles and the diacetate group cleaved by esterases. After 30 minutes of further incubation, the conversion of DCFH‐DA to the fluorescent product DCF was measured using a spectrofluorimeter with excitation at 484 nm and emission at 530 nm. Background fluorescence (conversion of DCFH‐DA in the absence of homogenate) was corrected by the inclusion of parallel blanks. ROS formation was quantified from a DCF‐standard curve and data are expressed as pmol DCF formed/min/mg protein.

In situ hybridization

RNA isolation, RT‐PCR, plasmid construction and in vitro synthesis of riboprobes

Whole brains of 18 mice (3 per group) were dissected after euthanasia with an overdose of the anesthetic agent. Total mRNA was extracted from mouse brain with Trizol (Invitrogen, Carlsbad, CA, USA). About 50–100 mg of brain material was used for RNA isolation with 1 mL of Trizol Reagent. AMV reverse transcriptase (Promega, Madison, WI, USA), and oligo‐(dT)₁₅ primers (Promega) were used for generating total cDNA. PCR cloning of nerve growth factor (NGF) (GenBank Accession No. M14805) was performed with primers containing EcoRI and BamHI (New England Biolab, Beverly, MA, USA) restriction enzyme cut site. NGF mRNA levels were also measured by semi‐quantitative RT‐PCR 39, 41. The sequences of PCR primers were as follows:

5′‐ATCGAATTCCTGAAGCCCACTGGACTAAA ‐3′ (forward primer)

5′‐ATCGGATCCCACCTCCTTGCCCTTGATG ‐3′ (reverse primer)

For control purposes, levels of β‐actin mRNA were measured in the same tubes using the following primers:

5′‐TGAACCCTAAGGCCAACCGTGAA‐3′ (forward primer)

5′‐TCTGCTGGAAGGTGGACAGTGAG‐3′ (reverse primer)

The size of the amplified β‐actin (GenBank Accession No. NM_007393) mRNA fragment was 735 base‐pair (bp). PCR was performed in 50 µL containing thermostable High‐Fidelity Taq DNA polymerase (Roche Diagnostics Corporation, Indianapolis, IN, USA), which provides a 3′ to 5′ proofreading activity, 0.1 mM dNTPs and 1 mM MgCl₂, using the Expand High‐Fidelity PCR‐system according to the instructions of the manufacturer. For PCR, the initial melting temperature was 95°C for 5 minutes, followed by 35 cycles at 94°C for 30 s, 55°C for 30 s and 72°C for 1 minute, with a final extension at 72°C for 10 minutes. The PCR products were visualized on a 1% agarose gel and purified with a Wizard SV Gel and PCR Clean‐Up System kit (Promega) and cloned into pGEM3Z plasmids (Promega). The insert was sequenced and verified to be a 374 bp cDNA, corresponding to bases 370–743 of the NGF mRNA. To make NGF riboprobe templates, the recombinant plasmid containing the 374 bp NGF cDNA fragment was linearized with EcoRI and transcribed with SP6 to generate a complementary RNA (cRNA) antisense probe complementary to mouse NGF mRNA. A sense probe was generated by linearizing the plasmids with BamHI, and transcribed with T7 RNA polymerase. Antisense and sense (control) riboprobes were labeled during in vitro transcription in a reaction mixture containing DIG RNA Labeling mix (Roche Diagnostics Corporation), 1 µg of linearized plasmids and 20 units of SP6 or T7 RNA polymerase (Promega) according to the manufacturer's protocol. After transcription, the template DNA was digested with 40 units of DNase I for 15 minute at 37°C. Unincorporated labeling molecules were removed by centrifugation.

RT‐PCR analysis

Semi‐quantification of detected bands on the agarose gel was calculated using the Scion Image analysis software (Scion Corp., Frederick, MD, USA). Each optical density was normalized using each corresponding β‐actin density as an internal control and we standardized the optical density of vehicle control for relative comparison as 1 to compare other groups.

Brain sections in situ hybridization

The in situ hybridization procedure was carried out as previously described 38, 39, 40, all steps being performed in RNAse‐free conditions. Briefly, the frozen mouse brain sections were arranged in metal racks and brought to room temperature for 30 minutes. The sections were fixed for 10 minutes with a 4% paraformaldehyde solution in PBS, rinsed twice with PBS and twice with PBS containing 100 mM glycine, permeabilized for 30 minutes at 37°C with TE buffer (100 mM Tris‐HCl, 50 mM EDTA, pH 8.0) containing 20 µg/mL RNase‐free proteinase K (Merck), postfixed for 5 minutes at 4°C with DEPC‐treated PBS containing 4% paraformaldehyde, acetylated with 0.25% acetic anhydride in 0.1 M triethanolamine–HCl (pH 8.0) for 15 minutes and dehydrated through graded ethanol. The sections were equilibrated for 5 minutes in 4 × SSC, and then prehybridized for 2 h at 55°C in 4 × SSC containing 50% deionized formamide (Ameresco, Solon, OH, USA). Hybridization was performed at 60°C overnight in hybridization buffer solution containing 40% deionized formamide (Ameresco Co.), 400 ng/mL digoxigenin‐labeled probes, 10% dextran sulfate (Sigma‐Aldrich), 1 × denhard's solution [0.02% Ficoll‐400 (Ameresco Co.), 0.02% polyvinylpyrrolidone (Ameresco Co.), 10 mg/mL RNase‐free BSA (Ameresco Co.)], 4 × SSC, 10 mM DTT (Sigma‐Aldrich), 1 mg/mL yeast tRNA (Calbiochem, La Jolla, CA) and 1 mg/mL denatured and sheared salmon sperm DNA (Sigma‐Aldrich). Prior to hybridization, the riboprobes were denatured for 5 minutes at 80°C and then cooled on ice. Following hybridization, the sections were washed three times for 1 h each with 2 × SSC at room temperature, 2 × SSC at 65°C and 0.1 × SSC at 65°C, respectively. Slides were then incubated for 1 h in 0.5% blocking reagent (Roche Diagnostics Corporation) prepared in 150 mM NaCl, 100 mM Tris–HCl pH 7.5 (NT), and the localization of the bound riboprobes was detected by incubating overnight at 4°C with an AP‐coupled anti‐digoxigenin antibody (Roche Diagnostics Corporation) diluted 1:2000 in 0.5% blocking reagent prepared in NT. The slides were washed twice in NT and equilibrated in a solution of 50 mM MgCl₂, 100 mM NaCl and 100 mM Tris–HCl pH 9.5 (MNT). The antibody was visualized using an alkaline phosphatase substrate (338 µg/mL nitroblue tetrazolium, 175 µg/mL 5‐bromo‐4‐chloro‐3‐indolyl phosphate; Roche Diagnostics Corporation) in MNT buffer. Finally, the color reaction was stopped by washing the slides with 10 mM Tris–HCl (pH 8.0) containing 1 mM EDTA.

Immunohistochemistry

For immunohistochemistry, endogenous peroxidase activity in the sectioned tissues was blocked with 3% H₂O₂, and nonspecific binding sites were blocked with 3% normal goat serum (Chemicon International Inc., Temecula, CA, USA). The sections were incubated with rabbit anti‐TrkA (pTyr⁴⁹⁰) (1:10, Sigma‐Aldrich) or rabbit anti‐Akt (pSer⁴⁷³) (1:50, Cell Signaling Technology, Inc., Beverly, MA, USA) in TBS containing 1% goat serum at 4°C overnight, respectively. Subsequently, biotinylated goat anti‐rabbit IgG secondary antibody (diluted as per the recommendations of the supplier; Vector Laboratories, Inc., Burlingame, CA, USA) was applied, followed by incubation for 1 h with an avidin–biotin–horseradish peroxidase complex (ABC Elite Kit, Vector Laboratories, Inc.). Horseradish peroxidase was reacted with diaminobenzidine and H₂O₂ for 5 minutes to yield a permanent deposit. Stained whole mounts were rinsed in distilled water, mounted on 3‐aminopropyl‐trimethoxysilane‐coated slides, air‐dried overnight, dehydrated in ethanol, cleared in xylene and cover‐slipped with cytoseal (Stephens Scientific, Kalamazoo, MI, USA). The specificity of the staining was assessed by omitting the primary antibody.

For quantitative analysis of the in situ hybridization and immunohistochemical staining, specimens from 18 mice (3 mice per group) were captured using a Zeiss Axioskop 40 microscope (10 × objective or 40 × objective) (Carl Zeiss, Oberkochen, Germany). The images were taken with a CCD camera (CoolSNAP Color, Photometrics, Roper Scientific, Inc., Trenton, NJ, USA) and the integral optical density (IOD) was measured by Image‐Pro Plus 6.0 software (Media Cybernetics Inc., Newburyport, MA, USA). For analysis, plaque areas were excluded and IOD in 0.01 mm² area (four areas per slide, three slides for each brain) was estimated at a consistent position per section.

Western blot analysis

Samples (80 µg protein each) were separated by denaturing SDS‐PAGE and transferred to a PVDF membrane (Roche Diagnostics Corporation) by electrophoretic transfer. The membrane was blocked with 5% nonfat milk and 0.1% Tween‐20 in TBS, incubated overnight with monoclonal rabbit anti‐Akt (pThr³⁰⁸) antibody (1:1000, Cell Signaling Technology, Inc.) or monoclonal rabbit anti‐phospho‐cyclic AMP response element‐binding protein [anti‐CREB (pSer¹³³)] antibody (1:1000, Cell Signaling Technology, Inc.) or polyclonal rabbit anti‐phospho‐α subunit of calcium/calmodulin‐dependent protein kinase II [anti‐CaMKII‐α (pThr²⁸⁶)] antibody (1:1000, Santa Cruz Biotechnology, CA, USA) or monoclonal rabbit anti‐phospho‐extracellular signal‐regulated kinase 1 and 2 [anti‐ERK1/2 (pThr²⁰²/Tyr²⁰⁴)] antibody (1:1000, Cell Signaling Technology, Inc.) or monoclonal rabbit anti‐ERK1/2 antibody (1:1000, Cell Signaling Technology, Inc.), or monoclonal mouse anti‐post‐synaptic density‐95 (anti‐PSD95) antibody (1:1000, Abcam, Cambridge, UK), respectively. Quantitation of detected bands was performed with the Scion Image analysis software (Scion Corp.). The data were normalized using t‐Akt, t‐CaMKII, t‐CREB or β‐tubulin as an internal control and standardized with the vehicle control as 1.0.

Statistic analysis

All statistical analyses were performed using the SPSS software, version 11.5. Group differences in the escape latency in the MWM training task were analyzed using two‐way analysis of variance (ANOVA) with repeated measures, the factors being treatment and training day. The other data were analyzed with one‐way ANOVA followed by Newman‐Keuls or Tukey's HSD post hoc test. Data were expressed as means ± SD. Statistical significance was set at P < 0.05.

RESULTS

Troxerutin attenuates cognitive impairment of D‐gal‐treated mice

Step‐through passive avoidance task

Effect of troxertin on the memory retention of D‐gal mice was investigated by step‐through passive avoidance task. None of the mice tested had obvious health problems (eg, weight loss, cataracts or toxicity reaction). In the acquisition trial, the initial latencies did not differ among the seven groups [F(6,56) = 1.606, P > 0.05] (Figure 1). One‐way ANOVA revealed that D‐gal significantly reduced the step‐through latencies in the 24 h‐retention trial [F(6, 56) = 36.677, P < 0.001], suggesting a memory deficit caused by D‐gal. This analysis also showed that the latencies in D‐gal‐treated mice received daily 150 mg/(Kg day) troxerutin for 8 weeks was significantly lengthened as compared with D‐gal‐treated mice (P < 0.001). In contrast, K252a blocked the neuroprotective effect of troxerutin. No significant difference in step‐through latencies between the D‐gal‐treated group and the troxerutin + D‐gal + K252a group was found. And K252a significantly shortened the step‐through latency of D‐gal group (P < 0.05, vs. D‐gal group). There was also no significant difference in step‐through latencies among the control group, the K252a group and the troxerutin group. This result indicated that no obvious neural toxicity in troxerutin‐treated mice or K252a‐treated mice was found.

Figure 1.

Step‐through passive avoidance task (n = 9). All values are expressed as mean ± SD ***P < 0.001 vs. control group; #P < 0.05; ###P < 0.001 vs. D‐galactose (D‐gal) group.

MWMtest

During the training session, all groups of mice improved their performance as indicated by the shortened escape latencies across successive days (Figure 2A). Mice showed significant difference in mean latencies between training days [F(3, 224) = 55.245, P < 0.001] (Figure 2A) and between treatments [F(6, 224) = 53.852, P < 0.001], but no interaction between the factors day and treatment [F(18, 224) = 0.592, P > 0.05].

Figure 2.

Morris water maze test (n = 9). All values are expressed as mean ± SD. A. Mean latency in the hidden platform test. B. The number of crossings over the exact location of the former platform. **P < 0.01, ***P < 0.001 vs. control group; #P < 0.05, ###P < 0.001 vs. D‐galactose (D‐gal group). C. Comparison of the time spent in target quadrant on day 5. *P < 0.05, **P < 0.01, ***P < 0.001 vs. control group; #P < 0.05, ##P < 0.01, ###P < 0.001 vs. D‐gal group.

Two‐way ANOVA showed that the escape latencies were significantly higher in D‐gal‐treated mice than in control mice (P < 0.001). This result indicated that the spatial learning and memory ability in D‐gal‐treated mice was impaired. A comparison between the D‐gal‐treated group and the troxerutin + D‐gal group showed that the troxerutin could shorten escape latencies of D‐gal‐treated mice (P < 0.001). These data showed that troxerutin could improve cognitive performance in the D‐gal‐treated mice, and there was no significant difference in spatial learning ability between the control group and the troxerutin + D‐gal group. However, K252a blocked the effect of troxerutin. No significant difference between the D‐gal‐treated group and the troxerutin + D‐gal + K252a group was found. In addition, K252a could significantly prolong the latency of D‐gal group. There was also no significant difference in escape latencies among the control group, the K252a group and the troxerutin group. Note that mice of every treatment except D‐gal + K252a group started at the same level of performance (no significant individual effect was observed for the first four trials of day 1).

The withdrawal of the platform induced a general tendency to swim, preferentially to other equivalent zones, in the quadrant where the platform was previously located and in the platform zone. The control, the troxerutin + D‐gal, K252a and the troxerutin mice spent more time swimming in the target quadrant (where the platform was located), while the D‐gal and the troxerutin + D‐gal + K252a mice tended to spend their time more averagely in each quadrant. [F(6, 56) = 19.820, P < 0.05, D‐gal and the troxerutin + D‐gal + K252a groups vs. control group] (Figure 2C). Furthermore, similar results were obtained from the former platform crossings experiments. The control, the troxerutin + D‐gal, K252a and the troxerutin mice crossed over the platform more frequently than D‐gal‐treated mice or troxerutin + D‐gal + K252a mice, respectively [F(6, 56) = 24.622, P < 0.001 vs. D‐gal group] (Figure 2B). No significant difference among the control group, the troxerutin + D‐gal group, the K252a group and the troxerutin group was found, as well as no difference between the D‐gal‐treated group and troxerutin + D‐gal + K252a group. The frequency of crossing over the platform position was significantly reduced in D‐gal‐treated mice by K252a treatment (P < 0.05, vs. D‐gal).

Troxerutin attenuates oxidative stress in the brain of D‐gal‐treated mice

The concentration of ROS is determined by the balance between the rate of production and the rate of clearance by various antioxidant compounds and enzymes. Cu, Zn‐SOD, an important antioxidative enzyme, can convert superoxide anions into peroxides which are finally converted into water. Cu, Zn‐SOD activity and ROS levels are usually viewed as oxidative stress markers. Our results (Figure 3) showed that D‐gal treatment resulted in a significant decrease in Cu, Zn‐SOD activity and a significant increase in ROS levels in the brain [Cu, Zn‐SOD: F(6, 14) = 42.278, P < 0.001; ROS: F(6, 14) = 31.697, P < 0.01]. The data indicated that oxidative stress in vivo was elevated in the brain of D‐gal‐treated mice, which was further exacerbated by K252a. Interestingly, the troxerutin could attenuate the decrease of Cu, Zn‐SOD activity and the increase of ROS levels induced by D‐gal (P < 0.01). This neuroprotection was significantly blocked by K252a. There was no significant difference in Cu, Zn‐SOD activity and ROS levels in the brain between D‐gal‐treated mice and troxerutin + D‐gal + K252a mice. No significant difference among the control group, the troxerutin + D‐gal group, the K252a group and the troxerutin group was found.

Figure 3.

Troxerutin attenuates oxidative stress in the brain of D‐galactose (D‐gal)‐treated mice (n = 3). All values are expressed as mean ± SD. A. Comparison of reactive oxygen species (ROS) levels in mouse brain. Values are averages from three independent experiments. **P < 0.01, ***P < 0.001 vs. control group; ##P < 0.01 vs. D‐gal group. B. Comparison of Cu, Zn‐superoxide dismutase (Cu, Zn‐SOD) enzymatic activity in mouse brain. Values are averages from three independent experiments. ***P < 0.001 vs. control group; #P < 0.05, ###P < 0.001 vs. D‐gal group.

Troxerutin upregulates NGF mRNA expression in the brain of D‐gal‐treated mice

NGF belongs to the neurotrophin family and is essential for the development, survival, differentiation of brain peripheral neurons, contextual memory consolidation and spatial learning 18, 35, 64. In the present study, the expression of NGF mRNA in hippocampus and cerebral cortex of mice was analyzed by in situ hybridization and semi‐quantitative RT‐PCR. In situ hybridization (Figure 4) revealed that the IOD of NGF mRNA was significantly decreased in the hippocampus and the cerebral cortex of D‐gal‐treated mice as compared with the control group. [F _Hippocampus(5, 12) = 8.742, P < 0.01; F _{Cerebral cortex}(5, 12) = 108.463, P < 0.001]. Interestingly, the troxerutin significantly increased the IOD of NGF mRNA in the hippocampus and the cerebral cortex of D‐gal‐treated mice as compared with the D‐gal‐treated group (hippocampus: P < 0.01; cerebral cortex: P < 0.001, troxerutin + D‐gal group vs. D‐gal group). The increase of the IOD of NGF mRNA in the hippocampus and the cerebral cortex of D‐gal‐treated mice administered troxerutin via oral gavage was not blocked by K252a. There was no significant difference among the control group, the troxerutin + D‐gal group, the troxerutin + D‐gal + K252a group, the K252a group and the troxerutin group. Semi‐quantitative RT‐PCR also showed that the troxerutin increased the level of NGF mRNA in the D‐gal treated mice brain. [F(5, 12) = 87.51, P < 0.001] (Figure 4). There was no significant difference among the control group, the troxerutin + D‐gal group, the troxerutin + D‐gal + K252a group, the K252a group and the troxerutin group in the NGF mRNA level.

Figure 4.

Troxerutin upregulates nerve growth factor (NGF) mRNA expression in the brain of D‐galactose (D‐gal)‐treated mice (n = 3). A. In situ hybridization of NGF mRNA in the hippocampus and cerebral cortex. Scale bars: 100 µm. B. Analysis of the integral optical density (IOD) value of NGF mRNA in mouse hippocampal and cerebral cortical slices. Values are averages from three independent experiments. Each value is the mean ± SD. **P < 0.01,***P < 0.001 vs. control group; ##P < 0.01, ###P < 0.001, vs. D‐gal group. C. Semiquantitative RT‐PCR analysis of NGF mRNA expression in mouse brain. The relative level of NGF mRNA in mouse brain was calculated and expressed as the ratio of NGF mRNA/β‐actin mRNA and the vehicle is set as 1.0. Values are averages from three independent experiments. Each value is the mean ± SD. ***P < 0.001 vs. control group; ###P < 0.001 vs. D‐gal group.

Troxerutin increases the phosphorylation of TrkA (Tyr⁴⁹⁰) in the brain of D‐gal‐treated mice

TrkA, a receptor tyrosine kinase, is activated by NGF to induce its dimerization and auto‐phosphorylation and results in activation of a number of signalling cascades that have been associated with neuronal survival, differentiation and synaptic plasticity 3, 25, 27, 28, 29, 36, 43, 55. In this study, the activiation of pTrkA (Tyr⁴⁹⁰) in the hippocampus and cerebral cortex was detected by immunohistochemistry (Figure 5A). Analysis of sections of the hippocampus and cerebral cortex revealed that D‐gal treatment significantly decreased the IOD of pTrkA (Tyr⁴⁹⁰) in both brain regions [F _Hippocampus(5, 12) = 37.476, P < 0.001; F _{Cerebral cortex}(5, 12) = 21.282, P < 0.001, D‐gal group vs. control group] (Figure 5B). The troxerutin by oral gavage administration induced a significant increase of the IOD of pTrkA (Tyr⁴⁹⁰) in both brain regions (hippocampus: P < 0.001; cerebral cortex: P < 0.001, troxerutin + D‐gal group vs. D‐gal group). K252a significantly blocked the activation of TrkA. There was no apparent difference in the level of pTrkA in the hippocampus and cerebral cortex between D‐gal‐treated mice and troxerutin + D‐gal + K252a mice (hippocampus: P > 0.05; cerebral cortex: P > 0.05, troxerutin + D‐gal + K252a group vs. D‐gal group). It was very interesting that K252a could inhibit NGF/TrkA pathway in D‐gal/troxerutin mice, but had no effect on control mice. Further studies need to be carried out to clarify this issue.

Figure 5.

Troxerutin increases the phosphorylation of TrkA (Tyr⁴⁹⁰) in the brain of D‐galactose (D‐gal)‐treated mice (n = 3). A. Immunohistochemistry of pTrkA in the hippocampus and cerebral cortex. Scale bars: 100 µm. Arrows indicated positive signals. B. pTrkA immunoactivity was measured by integral optical density (IOD) value. Values are averages from three independent experiments. Each value is the mean ± SD. **P < 0.01, ***P < 0.001 vs. control group; ##P < 0.01, ###P < 0.001 vs. D‐gal group.

Troxerutin protects mouse brain agaisnt D‐gal neurotoxicity via activation of the Akt/PKB pathway and ERK/MAPK pathway

Both Akt/PKB and ERK/MAPK are of tremendous importance for several neuronal key signaling events, including cell differentiation, proliferation and survival (53). In the present study, we investigated whether these pathways were involved in brain protection effects of troxerutin. The phosphorylation of Akt at Ser⁴⁷³ in the hippocampus and cerebral cortex was detected by immunohistochemistry (Figure 6A). Immunohistochemical analysis showed that D‐gal treatment significantly decreased the IOD of pAkt (Ser⁴⁷³) in both brain regions [F _Hippocampus(5, 12) = 6.645, P < 0.05; F _{Cerebral cortex}(5, 12) = 8.410, P < 0.05, D‐gal group vs. control group] (Figure 6B). The troxerutin by oral gavage administration significantly increased the IOD of pAkt (Ser⁴⁷³) in the hippocampus and cerebral cortex of D‐gal‐treated mice (hippocampus: P < 0.05; cerebral cortex: P < 0.05, troxerutin + D‐gal group vs. D‐gal group). K252a blocks this increase (hippocampus: P > 0.05; cerebral cortex: P > 0.05, troxerutin + D‐gal + K252a group vs. D‐gal group). Furthermore, we investigated the effect of troxerutin on the phosphorylation and activation of Akt (Thr³⁰⁸) and ERK (Thr²⁰²/Tyr²⁰⁴) in D‐gal‐treated mice brain by the immunoblot analysis (Figure 7A). Similarly, D‐gal treatment significantly decreased the activation of Akt (Thr³⁰⁸) [F(5, 12) = 37.325, P < 0.001, D‐gal group vs. control group] and ERK (Thr²⁰²/Tyr²⁰⁴) [F(5, 12) = 30.245, P < 0.001, D‐gal group vs. control group] in mouse brain (Figure 7B). The troxerutin treatment significantly increased the activation of Akt (Thr³⁰⁸) (P < 0.001, troxerutin + D‐gal group vs. D‐gal group) and ERK (Thr²⁰²/Tyr²⁰⁴) (P < 0.001, troxerutin + D‐gal group vs. D‐gal group) in D‐gal‐treated mouse brain. K252a blocks this neuroprotective effect of troxerutin (P > 0.05, troxerutin + D‐gal + K252a group vs. D‐gal group).

Figure 6.

Troxerutin increases the phosphorylation of Akt (Ser⁴⁷³) in the brain of D‐galactose (D‐gal)‐treated mice (n = 3). A. Immunohistochemistry of pAkt (Ser⁴⁷³) in the hippocampus and cerebral cortex. Scale bars: 100 µm. Arrows indicated positive signals. B. pAkt (Ser⁴⁷³) immunoreactivity was measured by integral optical density (IOD) value. Values are averages from three independent experiments. Each value is the mean ± SD. *P < 0.05 vs. control group; #P < 0.05 vs. D‐gal group.

Figure 7.

Troxerutin increases the phosphorylation of Akt (Ser³⁰⁸) and ERK (Thr²⁰²/Tyr²⁰⁴) in the brain of D‐galactose (D‐gal)‐treated mice (n = 3). A. Representative immublot for pAkt (Ser³⁰⁸) and pERK (pThr²⁰²/Tyr²⁰⁴) in mouse brain. B. Relative density analysis of pAkt (Ser³⁰⁸) and pERK (Thr²⁰²/Tyr²⁰⁴) protein bands. The relative density is expressed as the ratio (pAkt (Ser³⁰⁸)/t‐Akt, pERK (Thr²⁰²/Tyr²⁰⁴)/Total ERK) and the vehicle control is set as 1.0. Values are averages from three independent experiments. Each value is the mean ± SD. ***P < 0.001 vs. control group; ###P < 0.001 vs. D‐gal group.

Troxerutin increases the levels of pCaMKII, pCREB and PSD95 in the brain of D‐gal‐treated mice

Evidence suggests that pCaMKII, pCREB and PSD95 play key roles in the synaptic plasticity, learning and memory 34, 47. In this study, we examined the changes of memory‐related proteins (Figure 8A). Our data (Figure 8B) clearly indicated that the levels of pCaMKII, pCREB and PSD95 in the brain of the D‐gal‐treated group were markedly reduced [pCaMKII: F(5, 12) = 50.943, P < 0.001; pCREB: F(5, 12) = 145.558, P < 0.001; PSD95: F(5, 12) = 28.073, P < 0.001, D‐gal group vs. control group]. The troxerutin administration significantly increased the levels of pCaMKII, pCREB and PSD95 in the brain of the D‐gal‐treated mice (pCaMKII: P < 0.001; pCREB: P < 0.001; PSD95: P < 0.001, troxerutin + D‐gal group vs. D‐gal group). K252a blocks this effect of troxerutin (P > 0.05, troxerutin + D‐gal + K252a group vs. D‐gal group).

Figure 8.

Troxerutin increases the levels of pCaMKII, pCREB and PSD95 in the brain of D‐galactose (D‐gal)‐treated mice (n = 3). A. Representative immublot for pCaMKII, pCREB and PSD95 in mouse brain. B. Relative density analysis of pCaMKII, pCREB and PSD95 protein bands. The relative density is expressed as the ratio (pCaMKII /t‐CaMKII, pCREB /t‐CREB, PSD95/β‐tubulin) and the vehicle control is set as 1.0. Values are averages from three independent experiments. Each value is the mean ± SD. ***P < 0.001 vs. control group; ###P < 0.001 vs. D‐gal group.

DISCUSSION

Increasing evidence demonstrates that oxidative damage is an essential source of neurodegenerative diseases such as Parkinson's disease (PD) and Alzheimer's disease (AD). Long‐term injection of D‐gal in mouse brain induces overproduction of ROS and leads to neuronal oxidative damage 38, 69. Previous studies from our laboratory have shown that long‐term D‐gal administration induces ROS production, increases malondialdehyde level and decreases Cu, Zn‐SOD activity compared with that in the control 38, 40. Furthermore, long‐term D‐gal administration induces neural cell apoptosis, the decreased expression of synaptic proteins and the impairment of learning and memory function, including locomotive activity, passive avoidant learning and memory performance and ability to spatially orient. In the present study, we confirmed our previous findings of D‐gal‐impaired passive avoidant learning and memory function and spatial learning and memory ability and D‐gal‐induced oxidative stress in the brain of D‐gal‐treated mice. However, troxerutin treatment, a rutoside derivative and known as a radical scavenger drug, effectively improved the impaired learning and memory performance, restored Cu, Zn‐SOD activity and decreased ROS levels in the brain of D‐gal‐treated mice (16). Our actual results might be ascribed at least in part to its ability to indirectly scavenge and prevent free radical generation. The present findings agreed well with our recent report that troxerutin treatment in D‐gal‐treated mice restored learning and memory ability and redox level to nearly normal level (42). In addition, our present results suggest a further explanation for the protective effects of troxerutin on cholinergic function against D‐gal in the basal forebrain, hippocampus and frontal cortex of mice which could be modulated by NGF.

NGF is the first discovered and best characterized member of the neurotrophin family. NGF is secreted by the target cells of cholinergic neurons of the basal forebrain, binds to and activates its high affinity receptor TrkA. The activated NGF/TrkA complex is involved in neuronal survival, differentiation and functioning of cholinergic neurons in the basal forebrain (21). Evidence indicates that alterations in NGF levels have been implicated in neurodegenerative disorders, such as AD, PD and Huntington's disease (8). Recently, efforts to increase levels of NGF by cell therapy in patients with AD have shown some cognitive improvement (60). Our experimental data also showed that D‐gal downregulated NGF mRNA expression and its high‐affinity receptor TrkA activation in mouse brain, which might be the reason that D‐gal could induce behavior impairment. Whereas, troxerutin could reverse D‐gal‐induced neurotoxicity by increasing NGF mRNA expression and activating its high‐affinity receptor TrkA in mouse brain. When K252a, an antagonist for tyrosine kinase receptors, was infused intracerebroventricularly in the D‐gal‐treated mice administered troxerutin via oral gavage, the neuroprotective effect of troxerutin was blocked markedly. These results indicated that the neuroprotective effect of troxerutin was attributed at least in part to the activation of NGF/TrkA pathway against oxidative damage induced by D‐gal. Interestingly, the neuroprotective effect of troxerutin was blocked whereas the mRNA expression of NGF was not downregulated by K252a treatment in D‐gal/troxerutin group which may be caused by the direct effect of troxerutin on the NGF expression. A more detailed investigation needs to be carried out in the future. To further explain how NGF‐activated TrkA exerted the neuroprotective effect, we therefore investigated several TrkA‐dependent signaling cascades including phosphatidylinositol‐3 kinase‐mediated Akt signaling cascade and ERK signaling pathway.

The PI3K‐Akt signaling pathway is a post‐translational event which includes the binding of PI3K‐phosphorylated phosphoinositides to the PKB/Akt‐pleckstrin domain and subsequent translocation to the plasma membrane, where phosphoinositide‐dependent kinase‐1 (PDK1) and phosphoinositide‐dependent kinase‐2 (PDK2) phosphorylate Akt at threonine residue 308 and serine residue 473, respectively 1, 7, 66. Phosphorylated PKB/Akt is considered to be one of the key pro‐survival pathways within the cell (13). ERK signaling pathway is one of the three mitogen‐activated protein kinase (MAPK) cascades that play important roles in the regulation of cell proliferation, differentiation and survival. Evidence shows that phosphorylated ERK may phosphorylate a variety of substrates and promote cell survival 5, 31, 53. The present results suggested that NGF‐activated TrkA mediated Akt and ERK phosphorylation, which could promote neuronal survival against oxidative damage.

The ERK function in the intracellular signaling pathways not only regulates cell proliferation, differentiation and survival but also plays important roles in synaptic plasticity and memory function (57). Evidence suggests that ERK is phosphorylated in the hippocampus after training in hippocampus‐dependent tasks and is necessary for long‐term memory formation 2, 4, 6, 30, 52, 61. ERK activation is associated with the transcriptional factor CREB in cultured hippocampal neurons or brain slices 26, 45. Furthermore, Finkbeiner reported that phosphorylation of CREB can be induced by CaMKIIα when diverse extracellular stimuli are given (17). CaMKIIα is abundant in the central nervous system and a major component of the PSD, where it plays important roles in regulating neuronal plasticity and survival. Based on a previous report, the activation of CaMKIIα is significantly increased by NGF (54), which might help to repair the learning and memory deficits induced by D‐gal. Our results showed that troxerutin enhanced CaMKII, CREB, ERK phosphorylation and PSD95 expression via increased NGF mRNA expression and phosphorylated TrkA, which could ultimately improve learning and memory performance. Following treatment with K252a, a NGF receptor blocker, we observed that the levels of CaMKII, CREB, ERK phosphorylation and PSD95 expression were markedly reduced in the D‐gal‐treated mice administered with troxerutin and finally the neuroprotective effect of troxerutin was blocked.

In conclusion, troxerutin could improve the learning and memory ability and attenuate oxidative stress in the brain of D‐gal‐treated mice by NGF‐dependent activation of TrkA pathway (Figure 9).

Figure 9.

Schematic diagram illustrating the protective effects of troxerutin against D‐galactose (D‐gal)‐induced cognitive impairment and oxidative stress. Abbreviations: Cu, Zn‐SOD = Cu, Zn‐superoxide dismutase; D‐gal = D‐galactose; ERK = extracellular signal‐regulated kinase; NGF = nerve growth factor; ROS = reactive oxygen species.

DISCLOSURE STATEMENT

There are no actual or potential conflicts of interest in this work. All experiments were performed in compliance with the Chinese legislation on the use and care of laboratory animals, and were approved by the respective university committees for animal experiments.