ABSTRACT: To characterize mechanisms responsible for the CNS dopamine deficiency and the resulting neuropathology caused by deficiency of the housekeeping purine salvage function hypoxanthine guanine phospho- ribosyltransferase (HPRT) in the Lesch Nyhan Disease (LND), we have used microarray-based methods of global gene expression together with quantitative PCR and Western blot analysis to identify dysregulation of genes and aberrant cellular processes in human fibroblasts and in SH-SY5Y neuroblastoma cells made HPRT-deficient by transduction with a retrovirus stably expressing an shRNA targeted against HPRT. Analysis of the microarray expression data by Gene ontology (GO) and Gene Set Enrichment Analysis (GSEA) as well as by GeneSpring GX10 and Panther Classification System reveal that HPRT deficiency is accompanied by aberrations in a variety of pathways known to regulate neurogenesis or to be implicated in neurodegenerative disease, including the canonical Wnt/β-catenin and the Alzheimer’s disease/presenilin signaling pathways. Dysregulation of the Wnt/β-catenin pathway is confirmed by Western blot demonstration of cytosolic sequestration of β-catenin during in vitro differentiation of the SH-SY5Y cells toward the neuronal phenotype. We also demonstrate that two key transcription factor genes known to be regulated by Wnt signaling and to be vital for the generation and function of dopaminergic neurons; i.e., Lmx1a and Engrailed 1, are down-regulated in the HPRT knockdown SH-SY5Y cells. In addition to the Wnt signaling aberration, we found that expression of presenilin-1 shows a severely disturbed expression in HPRT-deficient cells, reflected by marked instability of the 23kDa C-terminal fragment of presenilin-1 in knockdown cells. Western blot analysis of primary cultures of two LND patients with 2.5% and 0% residual HPRT activity also shows dysregulated b-catenin and presenilin-1 expression, including elevated levels of cytosolic phospho-catenin and, in one of the two patient cells, failure of nuclear transport. Similarly, the presenilin-1 processing defect was most clearly demonstrated by markedly increased levels of both the N-terminal and C-terminal presenilin-1 fragments in the human cell line with no detectable residual enzyme activity but less marked over-expression in the cell with 2.5% residual enzyme activity. These demonstrations of dysregulated Wnt and presenilin-1 signaling and impaired expression of transcription factors necessary for dopaminergic development reveal broad pleitropic neuro-regulatory defects played by HPRT and suggest new directions for investigating mechanisms of aberrant neurogenesis and neuropathology in LND and potential new targets for restoration of effective signaling in this neuro-developmental defect. For microarray analysis, RNAs from triplicate independent cultures of vector-infected HPRT knockdown and control fibroblasts were prepared separately, pooled and used to prepare cRNA and finally subjected to microarray transcriptional analysis in triplicate. The integrity of total RNA from HPRT knockdown and control cells was confirmed by bioanalyzer (Agilent Technologies, Santa Clara, CA). The quality of total RNA samples from HPRT knockdown and control cells was assessed by 2100 bioanalyzer before application to microarray analysis. We determined that the RNA integrity number (RIN), (maximal degradation = 1; maximal molecular integrity = 10) was 10 for both the normal and knockdown cells (data not shown), indicating that isolated RNA samples were of sufficiently high quality to permit subsequent preparation of cDNA for microarray analysis. Microarray transcriptional analysis was performed in triplicate using the HumanWG-6 v3.0 Expression BeadChip system (Illumina, San Diego, CA). All reagents were obtained from HumanWG-6 v.3 Expression BeadChip Kit (Illumina) and all experimental processes were carried out according to manufacturer’s instruction (Illumina). After scanning of hybridized BeadChip, quantitation of slide images were performed using Illumina’s BeadArray software and the raw data were normalized by Loess normalization method, and then the normalized raw data in BeadStudio was exported to GeneSpring GX 10.0.2 (Agilent, Santa Clara, CA). For identification of genes significantly altered in knockdown cell compared with the control normal gene set, total detected entities were filtered by signal intensity value (upper cut-off 100th and lower cut-off 20th percentile) and error (coefficient of variation: CV < 50.0 percent) to remove very low signal entities and to select reproducible signal values of entities among the replicated experiments, respectively. In statistical analysis, t-test unpaired (p < 0.05) was applied and all significant changes above 2-fold were selected. Signals were selected if they were above microarray background (detection p-value < 0.05) in either all six experiments or in at least three knockdown or control experiments. Analysis of GO, GSEA and signaling pathway was carried out using GeneSpring GX 10.0.2 (Agilent) and the PANTHER Classification System (http://www.pantherdb.org/). In the analysis of signaling pathways using GeneSpring GX 10.0.2 (Agilent), a total of 140 cellular pathways were identified. For GSEA analysis, we used the false discovery rate (FDR) of <0.4 and p ≤ 0.1.
