Computational protocol: Novel cell separation method for molecular analysis of neuron-astrocyte co-cultures

Similar protocols

Protocol publication

[…] Agilent whole rat genome 4x 44K chips were used. cDNA synthesis and labeling for microarray hybridization were performed according to the manufacturers’ instructions. Briefly, oligo dT primers were used and Cy3 and Cy5 labeling was done using the Agilent two-color RNA Spike-In Kit. 0.5 μg of Cy5 and Cy3-labeled cRNA targets were hydrolyzed and mixed for 30 min. Arrays were hybridized with Cy5 and Cy3 solution for 18 h at 60°C in a rotating hybridization chamber, using 1× hybridization solution (Agilent technologies). Subsequently, arrays were washed, spun dry and scanned using an Agilent scanner.Analysis of the microarrays served two purposes: (1) detection of neuron-induced gene expression in astrocytes and (2) determination of the cold jet efficiency on the mRNA level. In total, four arrays were used. For the first purpose, two CC-CJ samples were labeled with Cy3 and two AA-CJ samples were labeled with Cy5. Each CC-CJ sample was hybridized together with an AA-CJ sample to an array. For checking the cold jet procedure on the mRNA level, two co-culture (CC) samples were labeled with Cy5 fluorescence and hybridized on an array together with a CC-CJ sample (Cy3). The microarray dataset can be found at National Center for Biotechnology Information Gene Expression Omnibus (NCBI GEO; accession number GSE52481).Normalization of the raw data was performed, after feature extraction by Agilent software, by optimized local intensity-dependent normalization (OLIN) and optimized scaled local intensity-dependent normalization (OSLIN; ) using the R project for statistical computing (version 6.0; R project). Next, M values were scaled so that upregulated probes were associated with positive values and downregulated probes with negative values. The following criteria were used for selection of regulated transcripts: (1) for each array, the mean and standard deviation (SD) of the M values of all probes was calculated, and only mRNAs with M values equal or higher to the mean + 2× SD, or equal or lower to the mean - 2× SD were selected, (2) M values had to fulfill our criteria on both replicate arrays, (3) for mRNAs that were represented on the microarray with multiple probes, data were checked for consistency, and probes were only included in the analysis when replicate sequences gave comparable results, (4) mRNAs upregulated in the CC-CJ with AA-CJ comparison could represent left-over neuronal RNA when also strongly downregulated in the CC with CC-CJ comparison. Therefore, mRNAs upregulated in the CC-CJ with AA-CJ comparison where only included when downregulation in the CC with CC-CJ comparison was less than the mean - 2× SD. To assess consistency between technical and biological replicate samples, as well as between replicate probes, non-parametric correlations (using Spearman’s rho) within non-normalized red or green fluorescence samples were calculated. For technical/biological replicate samples, correlations were only performed on regulated probes as indicated by the cold jet-astrocyte alone arrays.Thereafter, data were analyzed by TopGo () using the R project for statistical computing. Overrepresented biological processes were identified by scoring statistical significance of predefined annotated groups based on the gene ontology (GO) database. Statistical significance of overrepresented predefined GO biological processes was scored using the Fisher exact test (). […]

Pipeline specifications