Nanoplasmonic immunosensor for the detection of SCG2, a candidate serum biomarker for the early diagnosis of neurodevelopmental disorder

The neural circuits of the infant brain are rapidly established near 6 months of age, but neurodevelopmental disorders can be diagnosed only at the age of 2–3 years using existing diagnostic methods. Early diagnosis is very important to alleviate life-long disability in patients through appropriate early intervention, and it is imperative to develop new diagnostic methods for early detection of neurodevelopmental disorders. We examined the serum level of secretogranin II (SCG2) in pediatric patients to evaluate its potential role as a biomarker for neurodevelopmental disorders. A plasmonic immunosensor performing an enzyme-linked immunosorbent assay (ELISA) on a gold nanodot array was developed to detect SCG2 in small volumes of serum. This nanoplasmonic immunosensor combined with tyramide signal amplification was highly sensitive to detect SCG2 in only 5 μL serum samples. The analysis using the nanoplasmonic immunosensor revealed higher serum SCG2 levels in pediatric patients with developmental delay than in the control group. Overexpression or knockdown of SCG2 in hippocampal neurons significantly attenuated dendritic arborization and synaptic formation. These results suggest that dysregulated SCG2 expression impairs neural development. In conclusion, we developed a highly sensitive nanoplasmonic immunosensor to detect serum SCG2, a candidate biomarker for the early diagnosis of neurodevelopmental disorders.


Results
Conventional ELISA for SCG2 detection. To quantitatively analyze SCG2, a sandwich ELISA method was developed. Anti-SCG2 antibody was raised against human SCG2 peptide in rabbit and purified from the rabbit serum using an antigen peptide. Immunoblotting was performed with membrane proteins from rat brain tissues; the homemade SCG2 antibody detected SCG2 protein to a level comparable to that obtained with a commercial SCG2 antibody (Fig. S1A). In addition, the homemade SCG2 antibody worked well for immunoblotting and immunoprecipitation of recombinant human SCG2 protein expressed in heterologous cells (Fig. S1B). The homemade SCG2 antibody was employed as a capture antibody for the detection of SCG2 in the human serum using sandwich ELISA. Before constructing a standard curve for SCG2, the appropriate concentration of biotinylated SCG2 detection antibody was determined by measuring the absorbance signal at different concentrations of the detection antibody in the presence and absence of recombinant SCG2. Finally, 2.5 μg/mL was chosen as the optimal concentration of the detection antibody (Fig. S2A). The sandwich ELISA was performed under optimized conditions by varying the concentration of SCG2 diluted in 10% human serum to generate a standard curve; the absorbance exhibited linearity (R 2 = 0.9911) corresponding to SCG2 concentration. With this ELISA method, the limit of detection (LOD), defined by the signal at zero analyte concentration plus three standard deviations, was 5.2 ng/mL SCG2 (Fig. S2B).
Nanoplasmonic immunosensor for SCG2 detection. Then, we developed a plasmonic immunosensor combined with enzyme-based precipitation method to quantitatively analyze SCG2 with high sensitivity in small volumes of infant serum samples. Gold nanodot array (GNA) was used as the plasmonic material; the GNA chips were easily fabricated using UV nanoimprint lithography. As shown in Fig. 1A, B cross-sectional transmission electron microscopy (TEM) image of the GNA chip revealed a gold nanodot on a glass wafer with a diameter and height of 150 and 20 nm, respectively. Figure 1B shows that the hexagonal array of GNA was www.nature.com/scientificreports/ uniformly fabricated over a large area, and four nanodot-patterned areas were prepared on a single chip for multiple detection. Following the fabrication process, the GNA surface was functionalized with thiol groups by a self-assembly monolayer (SAM) reaction for antibody immobilization (Fig. S3). Immobilization of SCG2 capture antibody onto the GNA chip resulted in a red shift of the LSPR band from 777.1 to 779.6 nm, indicating that the antibody was immobilized onto the surface of GNA. Following a blocking step, SCG2-spiked human serum samples were incubated above the sensor chip arrayed with the SCG2 capture antibody. After detection with a biotinylated antibody, the streptavidin-alkaline phosphatase (STA-AP) conjugate was attached to the immunocomplex on GNA surface. The enzyme-mediated reaction catalyzed the conversion of multiple substrate molecules (NBT-BCIP, nitro blue tetrazolium & 5-bromo-4-chloro-3-indolyl phosphate p-toluidine) into precipitates (Fig. 1C). Figure 1D displays a sensorgram that demonstrates the changes in the LSPR wavelength monitored in real time corresponding to each binding process. Figure 1E shows the LSPR spectra of the GNA after each stage in the stepwise procedure for Fig. 1D. The LSPR peak shift (∆λ was defined as the shift in LSPR wavelength between before and after NBT-BCIP reaction) increased from 780.9 to 799.2 nm after the enzymemediated precipitation reaction on the GNA chip. The generated insoluble precipitate eventually induced a large shift in the LSPR peak to a longer wavelength region. The titration experiments to detect SCG2 in 10% human serum revealed that the LOD was 0.290 ng/mL and the slope of the LSPR shift was linear (R 2 = 0.9947) as the SCG2 concentration increased in the range of 0-100 ng/mL (Fig. 1F). In comparison with the conventional ELISA, the nanoplasmonic immunosensor showed approximately 18-fold higher sensitivity.
SCG2 detection with nanoplasmonic immunosensor combined with TSA. To enhance the signal generated from the AP-mediated reaction, the nanoplasmonic immunosensor was optimized using the tyramide signal amplification (TSA) strategy. The schematic representation explains the signal enhancement mechanism of the nanoplasmonic immunosensor for SCG2 detection using the TSA system ( Fig. 2A). After the binding of the biotinylated detection antibody to the captured SCG2, streptavidin-horseradish peroxidase (STA-HRP) gets attached to the immunocomplex. HRP then catalyzes the deposition of biotin-tyramide in the presence of hydrogen peroxide. This reaction covers the immunocomplex surface with biotin in large amounts for detection by STA-AP. The NBT-BCIP precipitate formed by AP is stacked on the GNA surface, consequently resulting in a large increase in the local dielectric constant. Before constructing the standard curve for SCG2, we optimized the blocking solution condition to minimize any non-specific binding between biotinylated detection antibodies and STA-HRP to achieve high signal-to-background ratios. The lowest LSPR peak shift in the absence of target molecules illustrated the successful inhibition of nonspecific binding by 5% skim milk to obtain the best sensitivity and LOD (Fig. S4A). Next, the appropriate concentration of biotinylated SCG2 detection antibody was determined by measuring the LSPR peak shift in the presence and absence of SCG2. As TSA is based on the enzyme-driven substrate deposition, the amount of HRP attached to the biotinylated detection antibody is critical for the analytical performance of the assay. High concentration of antibody can lead to relatively high steric hindrance, thereby preventing the biotin-tyramide www.nature.com/scientificreports/ and STA-AP reaction. On the other hand, low HRP concentration may be insufficient to catalyze the deposition of biotin-tyramide molecules and reduce the reaction sensitivity. As the LSPR peak shift increased until 5 μg/mL concentration of biotinylated detection antibody and reached a saturation point thereafter, we chose this value as the optimal concentration of detection antibody for further assays (Fig. S4B). We evaluated the sensitivity and quantitative performance of the proposed method for SCG2 analysis under optimal conditions. The standard curve for SCG2 was established in the concentration range of 0-30 ng/mL (Fig. 2B). Recombinant SCG2 protein was spiked into 10% human serum at final concentrations representative of clinical test samples. The LOD of the enhanced nanoplasmonic immunosensor (E-NPIS) for SCG2 was 0.016 ng/mL, which was 325-fold lower than that of the conventional ELISA. To confirm the specificity of the proposed assay for the detection of SCG2 in human serum, we analyzed specificity using several serum proteins including IL-6 (interluekin-6), PSA (prostate specific antigen), and AFP (alpha fetoprotein). As shown in Fig. S5, a large peak shift was observed only in the presence of SCG2 in 10% human serum. The LSPR peak shifts were comparable to the background signal in the presence of other interfering substances. These results indicated that the proposed assay exhibited good performance for discriminating SCG2 from possible interfering substances. We also confirmed that the results of the enhanced nanoplasmonic immunosensor matched those of the conventional ELISA (Fig. 2C). A good correlation between ELISA (x-axis) and enhanced nanoplasmonic immunosensor (y-axis) was obtained with the linear regression equation y = 1.1471x − 0.7453 (R 2 = 0.9844). These results demonstrate that the proposed method could accurately detect SCG2 in serum samples.

Detection of SCG2 in clinical samples using enhanced nanoplasmonic immunosensor. Finally,
we validated the feasibility of the enhanced nanoplasmonic immunosensor (E-NPIS) for the detection of SCG2 in clinical samples from pediatric patients. Whole serum samples (5 μL) of eight patients with developmental delay and five controls were 1:10 diluted in a buffer for measurement of SCG2 using the E-NPIS (Table S1 and Fig. 3A). SCG2 was successfully detected in the tenfold-diluted serum samples from pediatric patients using E-NPIS in the concentration range of 1.88 to 9.66 ng/mL (average of 4.54 ng/mL, corresponding to 45.4 ng/ mL in whole serum). The concentrations of SCG2 in the tenfold-diluted serum samples from the control group were between 0.53 and 4.24 ng/mL (average of 2.37 ng/mL, corresponding to 23.7 ng/mL in whole serum). The E-NPIS effectively detected SCG2 at low concentrations, which were undetectable using ELISA (LOD, 5.2 ng/ mL). Thus, the sensitivity of the proposed method was sufficient to quantify SCG2 in small volumes of serum samples from pediatric patients. As shown in Fig. 3A, the levels of SCG2 were higher in the developmental www.nature.com/scientificreports/ delay group than in the control group, although the difference was not statistically significant (p = 0.089). When the serum samples were 1:5 diluted and analyzed using the conventional ELISA method, the patient group displayed higher SCG2 concentrations than the control group; this result was similar to that observed with the E-NPIS (Fig. 3B). The conventional ELISA method allowed SCG2 detection in the concentration range of 7.79 to 19.55 ng/mL (average of 10.26 ng/mL, corresponding to 51.3 ng/mL in whole serum) in fivefold-diluted serum samples from the patient group and 5.41 to 8.53 ng/mL (average of 6.97 ng/mL, corresponding to 34.9 ng/mL in whole serum) in fivefold-diluted serum samples from the control group. In conclusion, this highly sensitive nanoplasmonic immunosensor enabled the detection of SCG2 in small volumes of serum, and revealed the higher serum SCG2 levels in patients with developmental delay.

Regulation of neuronal development by SCG2.
Whether SCG2 expression is closely associated with neuronal development or neurological function is questionable; thus, we examined the relationship between SCG2 expression and neuronal development (Fig. 4). We transfected hippocampal neurons with a human SCG2 expression construct. Primary/secondary dendritic analysis and Sholl profile revealed the attenuation of dendritic arborization in SCG2 overexpression (SCG2 OE) group as compared with that in the EGFP expression group (Cont) (Fig. 4A). In addition, the density of dendritic spines and the number of excitatory (puncta of vGLUT-positive PSD-95) and inhibitory (puncta of vGAT-positive gephyrin) synapses decreased after SCG2 overexpression as compared with that in the control group ( Fig. 4B-D). These results show that neuronal development is attenuated by SCG2 overexpression, and that the dysregulation of SCG2 expression could be a cause of neurodevelopmental disorders. We then examined SCG2 expression in a PTPRT −/− null mouse model of developmental delay. Protein tyrosine phosphatase receptor T (PTPRT) is principally expressed in the central nervous system and regulates neuronal development 29,30 . PTPRT gene abnormalities induce neurological diseases such as intellectual disability, autism, and depression [31][32][33][34] . PTPRT −/− null mice were shown to exhibit attenuated synaptic formation and synaptic plasticity as well as reduced pluripotency of neural stem cells in the hippocampus 30 . In the present study, we found that SCG2 expression decreased in the cortical and hippocampal tissues of PTPRT −/− null mice as compared with that in the wild-type mice (Fig. 5A). Therefore, we examined whether neuronal development of hippocampal neurons was attenuated by reduced SCG2 expression. When we knocked down SCG2 expression using SCG2-shRNA (SCG2 KD), dendritic arborization of hippocampal neurons was attenuated as compared with the control (Cont) group, and was rescued by SCG2 construct resistant to SCG2-shRNA (SCG2 res) (Fig. 5B). The density of dendritic spines, and the numbers of excitatory and inhibitory synapses decreased after SCG2 knockdown as compared with control, and were rescued by SCG2 construct resistant to SCG2-shRNA ( Fig. 5C-E). Thus, the appropriate expression level of SCG2 contributes to normal neurodevelopment, and that any dysregulation in SCG2 expression may induce neurodevelopmental disorders. In this respect, the altered SCG2 concentration could serve as a diagnostic indicator for impaired neurodevelopment; thus, we recommend SCG2 as a candidate serum biomarker of neurodevelopmental disorders.

Discussion
The prevalence of neurodevelopmental disorders exceeds ~ 17% in children between 3 and 17 years of age. These disorders are the most prevalent chronic medical conditions demanding pediatric primary care 1,35 . The diagnosis of neurodevelopmental disorders is mainly established based on methods dependent on language, communication, social interaction, facial expression, and intelligence quotient test. These methods pose difficulties to diagnose neurodevelopmental disorders at an early stage, except for some severe developmental disabilities. A child with global developmental delay typically presents with a delay in multiple developmental areas such as  www.nature.com/scientificreports/ speech, language, social, cognition, play, and motor skills. The diagnosis of global developmental delay is limited to children younger than 3-5 years of age, but these children often evolve to meet the diagnostic criteria for intellectual disability. Here, we developed a new diagnostic method using serum biomarkers for infant patients to overcome the limitations of current diagnostic methods. For the detection of SCG2, a candidate serum biomarker of neurodevelopmental disorder, the nanoplasmonic immunosensor employs an enzyme precipitation reaction combined with a tyramide amplification strategy. This enhanced nanoplasmonic immunosensor (E-NPIS) achieved a very low LOD and was approximately 325 times more sensitive than the conventional ELISA method. The performance of the enhanced nanoplasmonic immunosensor is more than comparable to other LSPR based immunosensors (Table S2). With only 5 μL clinical serum samples, the E-NPIS revealed the increase in SCG2 concentration in patients with developmental delay. On the other hand, 5 μL of clinical serum sample was not sufficient for conventional ELISA because SCG2 concentration was below its LOD. As the amount of serum sample was doubled to 10 μL in conventional ELISA, SCG2 concentration was still below the LOD for some clinical samples. Although the number of samples was not sufficient to draw conclusions, the analysis with conventional ELISA showed that serum SCG2 level increased in the patients as compared with that in the control group. These results suggest that SCG2 could be used as a serum biomarker for the early diagnosis of neurodevelopmental disorders. The reason the difference in SCG2 levels between patients and controls was not statistically significant could be that the sample number was too small or that serum SCG2 concentration inherently did not change much. In previous study, there has been not much difference in the serum SCG2 concentration between patients with Parkinson's disease and controls compared to other biomarkers 36 . Various metabolites, amino acids, peptides, and proteins have been recently identified as potential biomarkers for neurodevelopmental disorders [37][38][39] . The blood-brain barrier (BBB) defends the central nervous system by limiting the entry of harmful molecules from the bloodstream into the brain. Therefore, BBB dysfunction is a cause of many neurological diseases, and neuronal proteins secreted into the blood through the damaged BBB can be used as useful biomarkers 4 . SCG2 is expressed in large quantities in the central nervous system and is located mainly in the basal ganglia, hypothalamus, and pituitary gland. Seceretoneurin, a degraded peptide of SCG2, has been found to promote neurite outgrowth in immature cerebellar granule cells 40 . A previous study using mass spectrometry showed that the concentration of SCG2 decreased in the CSF of patients with multiple sclerosis, and on the contrary, increased in those with mild cognitive impairment, a pre-stage Alzheimer's disease 9,10 . In our study, the patients with developmental delay showed an increase in the serum level of SCG2 as www.nature.com/scientificreports/ compared with the control group. Interestingly, SCG2 expression reduced in the brain tissue of an animal model of developmental delay that showed attenuated synaptic plasticity and aberrant neurogenesis 30 . In hippocampal neurons, dendritic arborization and synaptic formation were attenuated by overexpression or knockdown of SCG2. SCG2 is involved in the packing/sorting of peptide hormones/neuropeptides, and its expression is regulated by neurotransmitter inputs; however, the exact role of SCG2 is not well understood. Learning and memory require the formation of new neural networks in the brain, and the key mechanism underlying this process is synaptic plasticity 41 . Synaptic plasticity is a process by which the strength of synaptic transmission is altered in response to the pattern of synaptic activity. Neurotransmitter inputs regulate gene expression, which underlies long-term forms of learning and memory 7 . SCG2 expression is activated by in-phase inputs of neurotransmitters and, thus, SCG2 has been suggested as a signal integrator of synaptic transmission. Therefore, SCG2 appears to play a crucial role in learning and memory, and dysregulated SCG2 expression could induce neurodevelopmental deficits; further studies with SCG2 gene-modified animals are warranted to confirm these hypotheses.
In conclusion, we developed a highly sensitive nanoplasmonic immunosensor for the detection of serum SCG2, which could be used for the early diagnosis of neurodevelopmental disorders such as global developmental delay.

Methods
Materials. 11  www.nature.com/scientificreports/ was raised against the peptide AHKEESSPDYNPYQG (human SCG2, aa 69-83) as an antigen in rabbit and purified from rabbit serum using antigen peptide covalently linked with Sulfolink coupling gel (Thermo Fisher Scientific). Other antibodies were purchased from commercial sources: rabbit anti-SCG2 (Cat# ab126935, Abcam, Cambridge, UK), mouse anti-PSD-95, rabbit anti-vGlut1 (Cat# MA1-045, Thermo Fisher Scientific), mouse anti-gephyrin (Cat# 147 111 The whole blood collected from participants was allowed to clot at room temperature (RT), and then the clot was removed by centrifuging at 2,000 × g for 10 min in a refrigerated centrifuge.
SCG2 detection using conventional ELISA. The conventional ELISA method was used as the standard in this study. Typically, 100 µL of the homemade anti SCG2 antibody (2 μg/mL) was coated in 96-well plates overnight at 4 °C, and then the wells were blocked with skim milk (w/v, 5%) for 2 h at RT. After blocking, the wells were washed with PBST (PBS at pH 7.4 containing 0.1% Tween20) to remove excess skim milk. Next, 50 μL of various concentrations of SCG2 in human serum were added to the wells. The patient serums were diluted fivefold with PBS buffer. The wells were incubated at RT for 2 h and washed thrice. Later, 100 μL of biotinylated anti-SCG2 antibody solution was added to each well and incubated at RT for 1 h. The wells were washed again, and 100 μL of STA-HRP (500 ng/mL) was added to the wells and incubated at RT for 30 min. Then, the wells were washed with PBST four times, and 100 μL of TMB substrates were introduced and incubated at RT for 30 min. Finally, 50 μL of stop solution (2 M H 2 SO 4 ) was added, and the optical density was measured at 450 nm using a microplate reader.
Sensor chip fabrication. The sensor chip fabrication procedure was reported in our previously works 21 .
Briefly, in order to make a sensor chip, the silicon master with the nano-hole pattern used as the master mold had a diameter, pitch, and depth of 150 nm, 300 nm, and 150 nm, respectively. The master surface was covered with a polycarbonate (PC) film after the UV resin (PFPE) to be used for the replica mold was dropped onto the surface. The PC film of the master was pressed with a hand roller to fill the hole pattern with resin, and then the replica mold was cured on the PC film. A thermoplastic resin (PMMA) was spin-coated on a glass substrate for 2 h at 130 °C (5 bar). The replica mold was separated from the substrate after cooling to 90 °C. To protect the top region of the imprinted resin from O 2 plasma etching, a 30 nm layer of titanium was deposited at a tilted angle of 70° during rotation in an e-beam evaporator. The residual layer was removed by O 2 plasma etching (PINK GmbH Plasma-finish, Bestenheid, Germany) at an O 2 flow rate of 300 mL/min and power of 300 W for 6 min. Finally, a Ti/Au layer (1/20 nm) was deposited by e-beam evaporation, and lift-off was carried out using an acetone solution for 5 min. There are four regions with a diameter of 3 mm in one chip; one of the four regions was used as the reference channel, and the remaining three regions were used as the sample channels. After washing the chip 3 times with 50 μL PBST, blocking solution (5% skim milk) was applied for blocking nonspecific binding, followed by incubation at 37 °C for 30 min. 50 μL of recombinant SCG2 at various concentrations (0-40 ng/mL) in human serum was injected, followed by incubation at 37 °C for 1 h. After incubation, the chips were washed with 3 times with PBST. For nanoplasmonic immunosensor, 5 μg/mL of biotinylated anti-SCG2 antibody and 5 μg/mL STA-AP were added to the chip. After the incubation for 1 h, the excess detection antibody and STA-AP were removed by washing the chips 3 times with washing buffer. Next, the enzyme-catalyzed precipitation reaction was triggered by adding 0.1 M 5-bromo-4-chloro-3-indolyl phosphate p-toluidine (BCIP) and 1 mg/mL nitro blue tetrazolium (NBT) in an AP buffer (100 mM Tris-HCl, 100 mM NaCl, 5 mM MgCl 2 ) at pH 9.5 for 30 min. For enhanced nanoplasmonic immunosensor, 5 μg/mL of biotinylated anti-SCG2 antibody and 5 μg/mL of STA-HRP instead of STA-AP were added to the GNA chip. The chip was washed, and then 1 × biotin-tyramine and 0.03% H 2 O 2 in 0.1 M borate buffer (pH 8.5) were added to the chip followed by keeping at room temperature for 15 min. After washing, 5 μg/mL of STA-AP was added to the chip and incubated at 25 °C for 30 min. Unbound STA-AP was removed by washing the chips 3 times with PBST. The enzyme-catalyzed precipitation reaction was triggered by adding 0.1 M BCIP and 1 mg/mL NBT in an AP buffer at pH 9.5 for 30 min. LSPR wavelength change was measured by LSPR observation setup with the back-reflection mode and determined by the centroid fit using a homemade program based on LabVIEW 21 . The scan area was the 1 mm diameter part located in the center of each region, and it was passed through the light source as described previously 21 . Immunoblot analysis. Immunoblotting was performed as previously described 42 . This study was performed in accordance with the regulations outlined by the Korean law. The animal experiment protocols were approved by the Animal Use and Care Committee of Korea Research Institute of Bioscience and Biotechnology (Permit Number: KRIBB-AEC-19115). Animals were sacrificed using CO 2 gas, and all efforts were made to minimize suffering. Briefly, mice were sacrificed and the brain tissue was quickly removed and homogenized in a homogenization buffer (50 mM Tris-HCl, 150 mM NaCl, 1% Nonidet P-40, 0.1% SDS, and 0.1% sodium deoxycholate, pH 8.0) containing protease inhibitor cocktail (Roche, Mannheim, Germany). Protein samples were resolved with SDS-PAGE and then transferred onto a polyvinylidene fluoride membrane (BioRad Laboratories, CA, USA). Blots were incubated with primary and secondary antibodies followed by visualization using an Enhanced Chemiluminescence kit (Atto Corp., Japan). Immunoblot images were quantified using Quantity One 1-D analysis version 4.6.1 software (Bio-Rad Laboratories) or Image J software (NIH) (https:// imagej. nih. gov/ ij/ index. html).
Transfection of neurons and immunocytochemistry. Primary hippocampal neurons were prepared from embryonic day 18 rats as described previously 29 . Briefly, hippocampi were dissected with trypsin and plated on coverslip coated with poly-L-lysine in Neurobasal medium (Thermo Fisher Scientific) supplemented with B27 (Thermo Fisher Scientific). After 2-3 h of incubation, the plating medium was changed with a growth medium (plating medium and glutamate). Cultured hippocampal neurons were transfected by the calcium phosphate method at 5 or 7 days in vitro. For immunofluorescent staining, after 7-9 days of transfection, cultured hippocampal neurons were fixed in 4% (v/v) formaldehyde/ 4% (w/v) sucrose, and permeabilized with 0.2% (v/v) Triton X-100 in phosphate-buffered saline followed by incubation with primary antibodies and fluorophoreconjugated secondary antibodies.
Image analysis and quantification. Images captured by confocal microscopy (LSM 810, Zeiss, Gottingen, Germany) were analyzed blindly using MetaMorph software (Version: 7.10.1.161, Molecular Devices, San Jose, CA, USA). Sholl analysis of dendritic arbors was performed employing the modified method of Nakayama et al. 43 . After images of individual neurons were printed, printouts were placed under a clear sheet featuring concentric circles with diameters increasing in 20 µm increments. The center of the circles was placed at the cell body center and the numbers of dendrites crossing each concentric circle were counted. Quantification of primary and secondary dendrites was also performed by imaging of individual neurons. The density of dendritic spine and synaptic protein clusters were measured from 27 to 32 dendrites of 7 ~ 9 neurons.
Statistical analysis. GraphPad Prism software (GraphPad Software, Inc., La Jolla, CA, USA) was used to perform all statistical analyses. Two-sample comparisons were conducted with Student's t-tests, while multiple comparisons were performed with a one-way analysis of variance (ANOVA) followed by Tukey post hoc tests. All results are presented as the means ± s.d. or means ± s.e.m. Differences with a p value less than 0.05 were considered to be statistically significant.