A quantitative comparison of cell-type-specific microarray gene expression profiling methods in the mouse brain

Abstract

Expression profiling of restricted neural populations using microarrays can facilitate neuronal classification and provide insight into the molecular bases of cellular phenotypes. Due to the formidable heterogeneity of intermixed cell types that make up the brain, isolating cell types prior to microarray processing poses steep technical challenges that have been met in various ways. These methodological differences have the potential to distort cell-type-specific gene expression profiles insofar as they may insufficiently filter out contaminating mRNAs or induce aberrant cellular responses not normally present in vivo. Thus we have compared the repeatability, susceptibility to contamination from off-target cell-types, and evidence for stress-responsive gene expression of five different purification methods--Laser Capture Microdissection (LCM), Translating Ribosome Affinity Purification (TRAP), Immunopanning (PAN), Fluorescence Activated Cell Sorting (FACS), and manual sorting of fluorescently labeled cells (Manual). We found that all methods obtained comparably high levels of repeatability, however, data from LCM and TRAP showed significantly higher levels of contamination than the other methods. While PAN samples showed higher activation of apoptosis-related, stress-related and immediate early genes, samples from FACS and Manual studies, which also require dissociated cells, did not. Given that TRAP targets actively translated mRNAs, whereas other methods target all transcribed mRNAs, observed differences may also reflect translational regulation.

Figure 1. Repeatability of microarray measurements is high and largely uniform across all cell purification methods.

Each data point represents the mean Pearson product-moment correlation coefficient between biological replicates for each of the analyzed cell types. Cell types are grouped by purification method, corresponding to the labels on the horizontal axis and demarcated by the shaded regions.

Figure 2. Cell purification methods show differential susceptibility to contamination.

(A) GABA contamination indices of non-GABAergic cell types, (B) astrocyte contamination indices of non-astrocyte cell types, and (C) oligodendrocyte contamination indices of non-oligodendrocyte cell types. For an explanation of how the contamination indices were computed, see the Methods . (D) Mean sample contamination indices of each purification method for the three different categories of contaminants (from A, B, and C). Differences in mean contamination indices across methods were significant for each category of contaminant (ANOVA p<0.005). Asterisks indicate which means were significantly different from the lowest means (Tukey's post-hoc test, p<0.05), which were achieved by the Manual sorting method in the case of GABA and astrocyte contamination, and PAN in the case of oligodendrocyte contamination.

Figure 3. PAN samples show moderately heightened expression of (A) immediate early genes, (B) apoptosis-related genes, and (C) stress-responsive genes, however overall global trends are comparable for most samples.

For each heat map, the replicate-averaged log2 microarray signal intensity (normalized) for each cell type (columns) is presented for all genes (rows) in a given category. Horizontal axis labels and vertical lines indicate purification method groups. (D) Average mean signal intensity for each method and each gene category. Asterisks indicate which means were significantly different from the lowest means (Tukey's post-hoc test, p<0.05).

Figure 4. Gene expression profiles of Purkinje cells purified by three different methods show striking differences.

(A) Scatter plot depicting the maximum fold difference of expression level between methods (vertical axis) and the corresponding ANOVA p-value (horizontal axis) for all genes (where an ANOVA was performed for each gene across groups defined by purification method). Upper right quadrant formed by the two intersecting red lines delineates significantly differentially expressed genes (maximum fold difference >2, ANOVA p-value < 1e-3). (B) Heat map depicting the most significantly differentially expressed genes (maximum fold difference between methods >20, ANOVA p-value < 1e-10) where each column represents an individual replicate sample. Gene names given in red font indicate genes that have a strong likelihood of being non-Purkinje gene contaminants (based on Allen Brain Atlas In-situ data and literature searches, see Results ), green font indicates non-translated mRNAs, and the expression specificity of the remaining genes is unknown. (C) Pie graphs depicting the percentage of method enriched and depleted genes that are also enriched in glia, and thus are likely to be the result of contamination. The total number of genes in each category is given beneath each graph. (D) Gene ontology terms associated with Manual and LCM enriched genes and TRAP depleted genes. Number of genes associated with each term is given on the vertical axis.