Development of a simple measurement method for GluR2 protein expression as an index of neuronal vulnerability

In vitro estimating strategies for potential neurotoxicity are required to screen multiple substances. In a previous study, we showed that exposure to low-concentrations of some chemicals, such as organotin, decreased the expression of GluR2 protein, which is a subunit of alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA)-type glutamate receptors, and led to neuronal vulnerability. This result suggested that GluR2 decreases as an index of neuronal cell sensitivity and vulnerability to various toxic insults. Accordingly, we developed a versatile method that is a large scale determination of GluR2 protein expression in the presence of environmental chemicals by means of AlphaLISA technology. Various analytical conditions were optimized, and then GluR2 protein amount was measured by the method using AlphaLISA. The GluR2 amounts were strongly correlated with that of measured by western blotting, which is currently used to determine GluR2 expression. An ideal standard curve could be written with the authentic GluR2 protein from 0 ng to 100 ng. Subsequently, twenty environmental chemicals were screened and nitenpyram was identified as a chemical which lead to decrease in GluR2 protein expression. This assay may provide a tool for detecting neurotoxic chemicals according to decreases in GluR2 protein expression.


Introduction
Mammals have been chronically exposed to environmental chemicals at low concentrations, and some environmental chemicals induce toxicity in individuals and in ecological systems. Thus, investigations of toxic environmental chemicals are necessary to prevent exposure. The central nervous system plays key roles in neuropsychiatric functions, such as behavior, learning, and memory, and comprises neuronal cells that recover poorly from damage. Life-long exposures to neurotoxic chemicals reportedly leads to altered behavior, mental retardation, and other neuronal disabilities, as well as diseases [1][2][3]. Moreover, during developmental stages, the immature blood-brain barrier allows the passage of neurotoxic environmental chemicals, even at low concentrations [4,5]. Therefore, an index for neurotoxicity and a screening system for the index are required.
We previously showed that long-term exposure of rat cortical neurons to low concentrations of organotin decreases GluR2 protein expression, leading to increased neuronal susceptibility to glutamate stimulation compared with that in control neurons [6]. Subsequently, we showed that long-term lead exposure induces neuronal cell death in association with decreased GluR2 expression [7]. The GluR2 protein is a subunit of the alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) receptor, which is a glutamate receptor that mediates rapid excitatory synaptic transmissions in the central nervous system. GluR2 is critical for Ca 2+ permeability of AMPA receptors. GluR2-containing AMPA receptors are impermeable to Ca 2+ [8], whereas GluR2-lacking AMPA receptors are highly permeable to Ca 2+ in a steady state, and the majority of functional AMPA receptors contain GluR2 [9]. GluR2 plays important roles in neuronal death, such as that associated with ischemia or Alzheimer's disease [10][11][12].
In addition, recent studies show that GluR2 is an essential regulator of the memory phenomena [13]. These studies suggest that decreases in GluR2 may be utilized as an index for conditions under which neuronal cells are sensitive and vulnerable to other stimulants. To investigate multiple neurotoxic chemicals for their effects on GluR2 expression, we need large-scale determinations of GluR2 protein expression in neuronal cells. However, western blotting for GluR2 expression is unsuitable for high throughput screening.
AlphaLISA ® (developed by PerkinElmer, Inc.) is applied to high throughput screening methods with high reproducibility [14][15][16]. This analysis is a bead-based proximity immunoassay [17] that exploits oxygen channeling technology. AlphaLISA assays are performed with anti-analyte (a target protein) antibodies and two AlphaLISA beads, including streptavidin-coated alpha donor beads and IgG-coated alpha acceptor beads. AlphaLISA is an all-inone-well assay that does not require transfer or wash steps, and allows assessments of analytes on a large scale [18].
In this study, we developed a novel in vitro assay to screen multiple compounds for their effects on GluR2 expression by means of AlphaLISA technology (Fig. 1). Subsequently, twenty potentially neurotoxic chemicals were screened, and then nitenpyram was identified as a novel chemical which lead to decrease in GluR2 protein expression.

Cell culture
The present study was approved by the animal ethics committee of Hiroshima University. The following procedures were performed under sterile conditions. Primary cultures were obtained from the cerebral cortex of fetal rats (at 18 days of gestation) as described previously [19]. Fetuses were taken from pregnant Slc; Wistar/ST rats under pentobarbital anesthesia. Subsequently, parts of the cerebral cortex were dissected using a razor blade, and cells were dissociated by gentle pipetting and plated on culture plates (4 × 10 6 cells/cm 2 ). Cultures were incubated in Eagle's MEM supplemented with 10% fetal calf serum, l-glutamine (2 mM), d-(+)-glucose (11 mM), NaHCO 3 (24 mM), and HEPES (10 mM). Cultures were maintained at 37 • C in a humidified 5% CO 2 atmosphere. Cultures were incubated in MEM containing 10% FCS (1-7 days) or 10% HS (8-11 days) and the medium was exchanged every 2-3 days. Arabinosylcytosine (10 M) was added to inhibit the proliferation of non-neuronal cells at 6 days in vitro (DIV). Cultures were used for experiments at 11 DIV. This protocol has been shown to produce cultures containing approximately 90% neurons, as indicated by immunostaining for the neuron marker, microtubule-associated protein 2 (MAP2).

Sample preparation
Cultures were used at 11 DIV. After chemical treatment, cells were washed with PBS and lysed in TNE buffer containing 50 mM Tris-HCl, 1% nonidet P-40, 20 mM EDTA, Protease Inhibitor Cocktail (1:200), and 1 mM PMSF. Mixtures were mixed by rotation at 4 • C and were centrifuged at 15,000 rpm; supernatants were used as cell lysates for AlphaLISA assays and western blotting.

Biotinylation and desalting of anti-GluR2 antibodies
Two kinds of anti-GluR2 antibodies with differing recognition sites on the GluR2 protein were selected for AlphaLISA, including mouse anti-GluR2 monoclonal antibody, which recognizes 175-430 amino acids of GluR2 protein, and rabbit anti-GluR2 polyclonal antibody, which recognizes 850-880 amino acids of GluR2 protein. Biotinylation of mouse anti-GluR2 monoclonal antibody was performed using ChromaLink TM Biotin (DMF Soluble; Solulink). Briefly, 100 l of 1 mg/ml antibody, 7.6 l of 2 mg/ml NHS-ChromaLink-biotin reagent, and 92.4 l of PBS were incubated for 2 h at 23 • C. The biotinylated antibody was purified using a Zeba Desalt Spin Column of 0.5 mL (Thermo Fisher Scientific K.K) according to the manufacturer's protocol. After biotinylation, absorbances at 280 nm (protein), 354 nm (biotin), and 450 nm (turbidity) were determined using a Multiskan TM GO (Thermo Fisher Scientific K.K). The number of biotins per antibody was calculated using a ChromaLink Biotin E1% MSR Calculator according to absorbance values. Desalting of rabbit anti-GluR2 polyclonal antibody was also performed using a Zeba Desalt Spin Column of 0.5 mL.

AlphaLISA assay
The AlphaLISA assay was performed in a final volume of 50 l containing cell lysates, biotinylated mouse anti-GluR2 monoclonal antibody, rabbit anti-GluR2 polyclonal antibody, streptavidin-coated donor beads, and anti-rabbit IgG AlphaLISA acceptor beads. Cell lysates were prepared in TNE buffer and the other reagents were diluted with AlphaLISA immunoassay buffer. Dilute solutions were used as negative controls.
Each examination was carried out on clamped conditions as follows: • Determinations of antibody concentrations; 1st incubation time, 2 h; 3rd incubation time, 4 h; protein concentration, 3000 ng.
Proteins were separated using SDS-polyacrylamide gel electrophoresis and were transferred to polyvinylidene difluoride membranes. Membranes were blocked with a blocking buffer containing 5% skim milk for 1 h, and were then incubated with anti-GluR2 (1:2000) and anti-␤-actin (1:8000) antibodies overnight at 4 • C. After incubation with secondary antibody for 1 h, proteins were detected using an enhanced chemiluminescence detection system [Chemi-Lumi One L, Nacalai Tesque (Kyoto, Japan)]. Quantitative analyses were performed using digital imaging software [Image J, NIH (Bethesda, MD, USA)], and GluR2 protein levels were normalized to those of ␤-actin. Western blotting was performed to confirm associations of anti-GluR2 antibody with either donor beads or acceptor beads and the subsequent formation of the complex for AlphaLISA assays.

Cell viability assay
Trypan blue dye exclusion assays and WST-1 assay were performed to assess neurotoxicity. Trypan blue dye exclusion assays were performed as described in previous reports [19]. After exposure to nitenpyram, cell cultures were immediately stained with 1.5% trypan blue for 10 min. Subsequently, the cells were fixed with 10% formalin for 2 min and rinsed with physiological saline. Unstained cells were considered viable and stained cells were considered dead. Cell viability of cultures was calculated as the percentage ratio of the number of unstained cells to total cells counted. Over 200 cells per well were randomly counted. In WST-1 assay, after exposure to nitenpyram, a mixed solution of WST-1, 1-methoxy-5methylphenazinium methylsulfate, and DMEM was added to each well, and the plate was incubated for 1 h. Finally, the supernatants were transferred to the wells of a 96well plate, and the absorption was measured at 415 nm, the maximum wavelength of the formazan dye, with a Microplate reader, Multiscan TM GO (Thermo Scientific).

Statistics
All the experiments were replicated and representative data were shown. Data are expressed as means ± standard errors of the mean (SEM). Statistical analyses were performed using ANOVA followed by Tukey's test, and differences were considered significant when p < 0.05.

Complex formation for AlphaLISA assays
Because AlphaLISA is based on the sandwich configuration, associations of each anti-GluR2 antibody with either donor beads or acceptor beads, and formation of corresponding complexes, were confirmed by means of western blotting with AlphaLISA reaction mixtures (Fig. 2). As in the AlphaLISA protocol in Fig. 1B, cell lysates containing GluR2 protein, anti-GluR2 antibodies, and beads for AlphaLISA were added and incubated for 2 h. Subsequently, two types of mixtures were prepared by incubation without donor or acceptor beads and were then centrifugation at 15,000 rpm for 15 min to separate bead and buffer  Fig. 2]. These data demonstrate the formation of essential complexes between the tested anti-GluR2 antibodies and beads, and validate their use in AlphaLISA assays of GluR2 in cell lysates.

Optimization of assay conditions
All antibodies had bead-binding capacity and the peak AlphaLISA signals were observed before saturation (hook point). Optimal concentrations of each antibody were identified according to bell-shaped curves of biotinylated mouse anti-GluR2 monoclonal antibody signals (Fig. 3A). The peak signal was observed at 12.8 nM indicating the hook point. In subsequent optimization experiments, biotinylated anti-GluR2 monoclonal antibody was used at the sub-hooking concentration of 10 nM. A saturated curve was obtained for rabbit anti-GluR2 polyclonal antibody (Fig. 3B), which required coupling to anti-rabbit IgG antibodies on acceptor bead surfaces. Under these conditions, acceptor beads became saturated with anti-GluR2 antibody, which was used at 5 nM to avoid saturation in further experiments. Hook points were also found in AlphaLISA experiments with cell lysates from rat primary cerebral cortical neurons, allowing determination of optimum protein amount in cell lysates (Fig. 3C). An AlphaLISA saturation curve was obtained using lysates from unexposed rat primary cerebral cortical neurons and was linear up to about 5000 ng (R 2 = 0.97; Fig. 3D). Thus, cell lysates with a protein content of 3000 ng were used in further experiments.
Reactions of both the 1st and the 3rd incubations of antibodies were enhanced with the duration of incubation. A time-dependent increase was also observed at both 1st and 3rd incubation times, and counts did not saturate (Fig. 3E). Therefore, for further experiments the 1st and the 3rd incubation times were set at 6 and 15 h, respectively.

Correlation between western blotting analyses and AlphaLISA assays
After optimizing assay conditions, in order to validate the present AlphaLISA assay, experiments were performed with a purified GluR2 standard. GluR2 is a predicted sixtransmembrane protein (Fig. 4A) and we generated GluR2 constructs using inverted PCR (Fig. 4B). In these experiments AlphaLISA signals were linear up to 100 ng (R 2 = 0.98; Fig. 4C) and were specific to the GluR2 protein. In addition, two kinds of the GluR2 determinations, AlphaLISA and western blotting, were conducted to the identical samples which contains 0-100 ng GluR2 standard, and the result showed strong correlation between the methods (R 2 = 0.99,    4D). The correlation indicated that the present AlphaL-ISA assay can be used as an alternative for western blotting. Moreover, we investigated the interference of the other proteins except GluR2 protein by means of the lysate of C6 cells. We conducted to the identical samples which contain the GluR2 standard and the lysates of C6 cells. The result showed strong correlation between the methods (R 2 = 0.97, Fig. 4E) in the presence of the protein from C6 cells and there is slightly difference compared with the samples which only contains GluR2 standard and TNE buffer (Fig. 4C).

Identification of nitenpyram as a novel suppressor of GluR2 expression
Twenty environmental chemicals were screened for the effect on GluR2 expression by means of the present AlphaLISA method. We previously showed that long-term exposure of rat cortical neurons to environmental chemicals, organotin and lead, led to decreased GluR2 protein expression and greater susceptibility of neurons to glutamate stimulation [6,7]. However, We think GluR2 decrease may be caused by other environmental chemicals besides metals, and screened the other type of environmental chemicals, such as agricultural chemicals. We selected twenty environmental chemicals which have various principle uses, such as pesticide, antiparasitic, herbicide and bactericide as test compounds for screening. Neuronal cells were exposed to DMSO (control) or test compounds for 9 days at the concentration of 1 M and 10 M ( Table 1). As the results, the neonicotinoid insecticide nitenpyram decreased the most in GluR2 expression. In subsequent experiments, cortical neurons were exposed to 0.1-100 M nitenpyram for 9 days and GluR2 expression was measured. As we previously reported that showed that long term exposure to 20 nM tributyltin (TBT) leads to decreased GluR2 protein expression in rat cortical neurons [6]. TBT-treated cortical neurons were used as a positive control, and GluR2 expression was reduced by nitenpyram in a dose-dependent manner and the reduction was significant at concentrations above 10 M (Fig. 5A). AlphaLISA and western blotting were conducted to the identical samples which were obtained from neuronal cells which were exposed to 0-100 M nitenpyram, and the result showed strong correlation between the methods (R 2 = 0.86, Fig. 5B), which indicated that AlphaLISA assay is useful to assess the   neurotoxicity of large numbers of hazardous chemicals as an alternative for western blotting.

Vulnerability of nitenpyram-treated cells to glutamate stimulation
To determine whether nitenpyram causes cell vulnerability, we assessed glutamine-induced neuronal death in the presence or absence of nitenpyram. Neurons were exposed to DMSO (control) or 100 M nitenpyram for 9 days, followed by 0, 50, or 100 M glutamate for 24 h. At 50 M concentration, cell viability following glutamate stimulation differed significantly between control neurons and nitenpyram-treated neurons (Fig. 6A). Moreover, this increased susceptibility of nitenpyram-treated neurons was abolished by an antagonist of GluR2-lacking AMPA receptor, 1-naphthylacetylspermine (NAS) (Fig. 6B and C).

Discussion
Some environmental chemicals are believed to have neurotoxicity at low concentrations [3]. However, small numbers of chemicals have been assessed for neurotoxicity as well-developed methods and markers of neurotoxicity are not available. Analyses of behavioral and neuropathological changes in vivo provide powerful tools for assessing neurotoxicity [21]. Since such analyses are time consuming and expensive, they are unsuitable for neurotoxicity screening of multiple chemicals, necessitating a convenient high throughput in vitro method [22,23]. Thus, in this study, we conditioned AlphaLISA assay for the measurement of GluR2 which is known as a key subunit of AMPA receptor and dictate Ca 2+ permeability. The assay was utilized to assess decreases in GluR2 expression following long-term exposure to low concentrations of multiple environmental chemicals.
AlphaLISA assays are performed with anti-analyte (a target protein) antibodies and two AlphaLISA beads, including streptavidin-coated alpha donor beads and IgGcoated alpha acceptor beads. In this method, biotinylated anti-analyte antibodies are conjugated to donor beads and non-biotinylated anti-analyte antibodies are conjugated to acceptor beads. Accordingly, the acceptor bead is not directly coupled to the analyte-specific antibody, but is coupled to another antibody that recognizes the non-biotinylated anti-analyte antibody. Upon excitation, a photosensitizer inside the donor bead converts ambient oxygen to an excited singlet state. In the presence of analytes, antibodies and beads conjugate analytes, and singlet oxygen molecules are provoked by donor beads that are proximal (about 200 nm) to acceptor beads, triggering a cascade of chemical events in nearby acceptor beads and chemiluminescent emissions at 615 nm.
Under optimal conditions, GluR2 protein amount in AlphaLISA assays and western blotting analyses were highly correlated (R 2 = 0.99, Fig. 4D). The regression line is corresponding to the unity because the slope is about 1. Western blotting is used as a standard method for confirming the expression of specific proteins. However, it requires electrophoresis and transfer steps, followed by quantitation of band densities using software, and lacks quantitative accuracy. In contrast, AlphaLISA is an all-inone-well assay that does not require transfer or wash steps and the signals were quantitative with a purified GluR2 standard (R 2 = 0.98, Fig. 4C  AlphaLISA assays and western blotting analyses validates the present AlphaLISA method as a rapid alternative to western blotting for measurements of GluR2 expression. Besides we confirmed that AlphaLISA was not interfered other cell lysate proteins expect GluR2 using the samples which contain the GluR2 standard and the lysates of C6 cells (Fig. 4E). This results also suggest that TNE buffer is enough as a blank, not cell lysates which do not contain GluR2 protein.
After screening twenty environmental chemicals using the present method, nitenpyram, one of the neonicotinoid insecticides, was identified as a candidate that decreases GluR2 expression (Table 1). Nitenpyram treatment significantly increased neuron susceptibility to glutamate and a GluR2-lacking AMPA receptor antagonist rescued this susceptibility (Fig. 6), suggesting that nitenpyram-induced GluR2 decrease leads to susceptibility of neurons. These results validated the present AlphaLISA method for assessments of neurotoxicity.
Neonicotinoid insecticides are currently the newest and largest single insecticide class in the market, representing more than 20% of insecticides used in 2008 globally [24]. Neonicotinoid insecticides are also used to control fleas and ticks in household pets, and act as selective agonists of nicotinic acetylcholine receptors (nAChRs) in insects [25]. However, some reports suggest that neonicotinoid insecticides affect mammalian nAChRs to a greater extent than previously believed [26][27][28] and that further studies of neonicotinoid insecticides are required.
We developed a new in vitro evaluating method utilizing decreased GluR2 protein expression results in detectable neuronal cell vulnerability [6,7], as well as automated image analyses of neurogenesis and apoptosis. GluR2 is a subunit that depresses cell permeability to bivalent cations in a steady state [29,30] and plays an important role in glutamatergic neuron signaling. Previous studies have demonstrated that GluR2 decreases damage active neurons, and that GluR2 knockout mice have impaired associative learning and reduced anxiety [31]. In in vitro studies, neurotoxicity is evoked by kainate-acid stimulation in GluR2-deficient neurons, possibly reflecting other GluR2-associated consequences in addition to Ca 2+ permeability [32]. Moreover, decreases in GluR2 protein expression have been related to neurodegenerative diseases [10][11][12]. Accordingly, about 60-70% of AMPA receptors have intracellular localization [33], but are delivered to the plasma membrane surfaces as functional AMPA receptors. Thus, total decreases in GluR2 reflect those on the plasma membrane surface.
We previously showed that long-term exposure of rat cortical neurons to environmental chemicals led to decreased GluR2 protein expression and greater susceptibility of neurons to glutamate stimulation [6,7]. These data suggest that GluR2 decreases may offer an index of neuronal cell sensitivity and vulnerability to other stimuli. However, further studies are needed to confirm that GluR2 decreases are a suitable endpoint of neurotoxicity. To this end, comparisons of the effects of multiple chemicals on GluR2 protein expression using AlphaLISA, which is a high throughput screening method with high reproducibility, will be valuable. In the present study, we simplified the AlphaLISA procedures following cell solubilization. However, future studies are required to simplify the pre-measurement processing, such as sample preparation, to maximize the practical application of the present method. In addition, in the present method, AlphaLISA needs three steps for measurement. However, we can simply the method to two steps by binding the GluR2 antibody with GluR2 antibodies which are used in the present method to acceptor beads in advance (AlphaLISA direct assay). AlphaLISA direct assay is capable of improvement in both cost and time consumption. In conclusion, we developed a simple high throughput measurement method for GluR2 expression by means of AlphaLISA technology. This AlphaLISA method allows comprehensive investigations of the effects of multiple potentially neurotoxic chemicals, and it may facilitate investigations of the mechanisms behind decreases in GluR2 expression.