Fig. 2. Microarray analysis of gene expression profiles changed by GC and the reversion of their gene expression by
Lpc-EV in HT22 cells. (A) A geometric map consisting of hypothetical gene expression signatures of CON
vs. GC on the x-axis and GC
vs. GC+
Lpc-EV on the Y-axis. The x-axis and the y-axis represent the ratios of the logarithms (log10) of the gene expression levels of GC to Con and GC+
Lpc-EV to GC, respectively. The features and locations of up- or down-regulated expression of genes are explained in four quadrants. (B) A geometric map constructed with gene expression signatures of CON
vs. GC on the x-axis and GC
vs. GC+
Lpc-EV on the y-axis. The genes that were up-regulated by GC and subsequently down-regulated by
Lpc-EV are located in Quadrant A. Conversely, the genes that were down-regulated by GC and subsequently up-regulated by
Lpc-EV are located in Quadrant D. Dotted lines: 20% cut-off. Blue-to Red color gradient: the number of genes counted; blue, rare or few; red, many numbers indicated. (C) A summary and experimental flow of the sequential analysis of gene expression profiles of CON
vs. GC on the x-axis and GC
vs. GC+
Lpc-EV on the y-axis. The numbers of the genes located in Quadrants A and D of the RRHO map, the number of the top 20%-ranked genes defined by RRHO in Quadrants A and D, the number of the top 20%-ranked genes that had mouse gene annotations in the STRING database, and the number of the top 20%-ranked genes predicted to form functional interaction with defined gene members with a medium confidence strength of 0.400 are indicated. (D, E) The identified 998 genes of Quadrant A were grouped by K-means clustering into seven clusters (D), of which Cluster 5 contained the stress signature genes
Mkp1 and
Fkbp5. Cluster 1 (162 genes), red; cluster 2 (158 genes), brown; cluster 3 (156 genes), olive; cluster 4 (149 genes), green; cluster 5 (135 genes), blue; cluster 6 (131 genes), light blue; cluster 7 (101 genes), medium blue. See also
Supplemental Table S1 and
S2. The 756 genes of Quadrant D were grouped by K-means clustering into eight clusters (E), of which Cluster 4 carried
Mecp2. Cluster 1 (111 genes), red; cluster 2 (107 genes), brown; cluster 3 (101 genes), olive; cluster 4 (99 genes), green; cluster 5 (94 genes), lime green; cluster 6 (90 genes), cyan; cluster 7 (78 genes), blue; cluster 8 (76 genes), purple. The number 5 in (D) and the numbers 3 and 4 in (E) indicates the cluster names. See also
Supplemental Table S1 and
S3. (F~H) A summary of key features of Cluster 5 of Quadrant A (F) and Clusters 3 and 4 of Quadrant D (G, H).
