Noninvasive Brain Stimulation Enhances Memory Acquisition and Is Associated with Synaptoneurosome Modification in the Rat Hippocampus

Behavioral data analysis. A, Analysis of passive avoidance data. Two-way RM ANOVA detected a significant difference in the latency of all groups between the training and testing days. During the testing day, a statistical significance was detected between sham and acquisition groups. #Habituation data were collected from only eight animals per group; * and ** indicate statistical significance (p < 0.05 and p < 0.001, respectively). B, Cox proportional hazard regression analysis for PAT data on training day. No statistical group separation was detected. C, Cox proportional hazard regression analysis for PAT data on testing day. A significant group difference was detected (log-rank Χ ² = 7.0919, df = 2, p = 0.0288). D, PCA with proteomics data show the distribution of each samples across the groups. E, Hierarchical clustering analysis with proteomics data show the distribution of each sample across the groups. Red, green, and blue boxes represent sham, retrieval, and acquisition samples. Protein IDs in the first hierarchical clustering group, from left to right, are KDELC2, EXOC6, UFL1, CDH8, ARHGAP27, BLNK, MPPED2, GFM1, FNTB, WASH, PDE4DIP, TUBA4A, ACVR1B, PRKAR1B, RBMX2, ZFYVE27, CAPN5, PSME1, KCNA6, LLGL1, JAM3, CD151, S1PR5, and GPSM1. The hierarchical clustering group 2 includes, from left to right, LHX1, GUCY1A2, STX8, CSAD, USP19, FTSJD2, DNAH7, LOC362863, RAB9A, GSTZ1, PDIA4, RPH3AL, LRFN1, STK39, CCDC116, LOC100911646, GLS, PYROXD2, LOC100911456, AOC3, MYO1C, and ILK. For the cluster group 3, protein IDs are INHBA, EIF3J, SLC20A1, ZRANB2, CCDC127, FAM195B, GLDN, FHIT, PNPO, PGP, TUBA1C, ACBD6, MCCC1, NOP58, ITGAD, RPL27, ABCF1, WDR81, and COTL1. The cluster group 4 contains proteins named as, from left to right, CACNA1D, GHITM2, SCFD1, PTP4A2, ZW10, GCLM, APMAP, CRP, SRP54, PTPRN2, SRP542, ADCY6, CAR3, COL3A1, IP6K1, GOLPH31, SLC27A1, RILPL1, PFKFB1, PSMA2, and ICOS. Proteins from the cluster group 5 are GALE, EQTN, CLUL1, GABRB2, MRPS25, MRAS, SCYL1, OXCT1, FAM194A, FAM213B, LAMP5, SNX1, CNIH2, GHITM, SLC29A3, and AK6. Proteins in the cluster group 6 are MARK2, HOMER3, GPHN, RAB3GAP2, RBP1, PALMD, MCCC2, ARHGEF11, and LRRC57. For a clear figure, see Extended Data Figure 2-1.

IPA results. A, Behavior-associated functional annotation detected from the comparison between the acquisition and sham groups. B, Network-1 detected from the comparison between the retrieval and sham groups. C, Six networks that were associated with the nervous system were merged in to a network. Nodes that did not pass cutoff in dataset, not in overlaid dataset, and not connected to any other nodes were excluded from the network. Pink-outlined molecules are associated with cognition (p = 1.81E-3). Molecule activity predictor (MAP) was also overlaid to predict the upstream and downstream effects of activation or inhibition on other molecules. For all IPA data, see Extended Data Figures 3-1, 3-2, 3-3, 3-4.

PPI network analysis. A, PPI network for the significant molecules from the acquisition-and-sham comparison. Nodes were colored for functions of glutamatergic synapse (blue), metabolic pathways (yellow), dopaminergic synapse (green), LTP (red), and cholinergic synapse (pink). Disconnected nodes were excluded. B, The first cluster network of MCL algorithm. C, The second cluster network of MCL algorithm.