Microarray data on altered transcriptional program of Phgdh-deficient mouse embryonic fibroblasts caused by ʟ-serine depletion

Inherent ʟ-Ser deficiency culminates in intrauterine growth retardation, severe malformation of multiple organs particularly the central nervous system, and perinatal or early postnatal death in human and mouse. To uncover the molecular mechanisms underlying the growth-arrested phenotypes of l-Ser deficiency, we compared gene expression profiles of mouse embryonic fibroblasts deficient in 3-phosphoglycerate dehydrogenase (Phgdh), the first enzyme of de novo ʟ-Ser synthetic pathway, between ʟ-Ser-depleted and -supplemented conditions. The datasets (CEL and CHP files) from this study are publicly available on the Gene Expression Omnibus repository (accession number GEO: GSE55687).


a b s t r a c t
Inherent ʟ-Ser deficiency culminates in intrauterine growth retardation, severe malformation of multiple organs particularly the central nervous system, and perinatal or early postnatal death in human and mouse. To uncover the molecular mechanisms underlying the growth-arrested phenotypes of L-Ser deficiency, we compared gene expression profiles of mouse embryonic fibroblasts deficient in 3-phosphoglycerate dehydrogenase (Phgdh), the first enzyme of de novo ʟ-Ser synthetic pathway, between ʟ-Ser-depleted and -supplemented conditions. The datasets (CEL and CHP files) from this study are publicly available on the Gene Expression Omnibus repository (accession number GEO: GSE55687

Value of the data
The gene expression data list the significantly affected genes by reduced ʟ-Ser availability in Phgdhdeficient mouse embryonic fibroblasts.
Enriched GO terms and phenotypically relevant gene networks provide insight into altered cholesterol metabolism and stress responses elicited by L-Ser deficiency in embryonic fibroblasts.
The data suggest that Phgdh-deficient mouse embryonic fibroblasts serve as a valuable mouse cellular model for human inborn ʟ-Ser deficiency including Neu-Laxova syndrome. Table 1 of DAVID analysis shows that cholesterol/sterol biosynthetic/metabolic process was enriched GO terms in the biological process (BP) of the down-regulated 381 genes, whereas apoptosis/cell death, amino acid biosynthetic process, tRNA aminoacylation, cell cycle arrest, and transcription were major significantly enriched GO terms in the up-regulated 560 genes. Ingenuity Pathway Analysis (IPA) determined top-ranked networks in the down-regulated genes ( Fig. 1A) and the up-regulated genes (Fig. 1B). A network containing genes involved in the cholesterol metabolic process including Hmgcs1, Insig, Hmgcr, and Ldlr, was markedly diminished in the down-regulated genes, while the activation of a network containing stress-responsive Atf4-Atf3-Ddit3 (CHOP) axis was most prominently in the up-regulated genes.

Microarray analysis
Total RNA was extracted using the RiboPure kit (Thermo Fisher Scientific, Waltham, MA USA) after a 6 h incubation under ʟ-Ser-depleted or -supplemented conditions as described [1]. cDNA amplification and labeling, and chip hybridization were carried out as described [1]. After washing, the arrays were scanned with a GeneChipScanner (Affymetrix), and the scans data were processed using the GeneSuite software (Affymetrix). Three biological replicates for each treatment were directly compared.

Data processing and statistical analysis
The data in .CEL files were transferred to GeneSpring 8.0 software (Agilent Technologies). After normalization to its median value, filtration was performed based on the following criteria: (i) scaled intensity4 100 under at least one condition; (ii) false discovery rate, qo 0.01; and (iii) absolute value of fold change (ʟ-Ser-depleted condition/ʟ-Ser-supplemented) 4 2.0 or o0.5.

GO term enrichment and pathway analysis
Significantly differentially expressed genes were analyzed using the Database for Annotation, Visualization, and Integrated Discovery (DAVID) to calculate GO term enrichments in the biological process category at all levels [3]. Enriched GO terms (Benjamini-Hochberg correction:Q-value o0.05) were deemed significant. Phenotypically relevant gene networks of significantly differentially expressed genes were analyzed using the web-based expression analysis program Ingenuity Pathways Analysis (http://www.ingenuity.com).