The Journal of Neuroscience, September 27, 2006, ():
Global Transcriptome Analysis of Genetically Identified Neurons in the Adult Cortex
J. Neurosci. Rossner et al.
26: 9956
Supplemental data
Files in this Data Supplement:
- supplemental material -
Supplemental Methods
- supplemental material -
Supplemental Results
- supplemental material -
Suppl. Table. 1: Description of the samples
All samples were analyzed with Affymetrix MOE430A gene chips; all targets were generated with a T7-RNA polymerase mediated linear amplification method for two rounds. The reference mouse brain and kidney targets were generated from 50 ng of total RNA isolated from gross dissected tissue; all other targets were generated from samples isolated with laser microdissection (LDM).
- supplemental material -
Suppl. Table. 2: Quality Control parameter of all GeneChips
Summary of all genechip quality control parameter for each MOE430A genechip used (SD = standard deviation).
- supplemental material -
Suppl. Table 3: Probe sets found to be differentially expressed between the Y+MCx versus Y+SSCx samples.
- supplemental material -
Suppl. Table. 4: Gene set description and statistical analysis
- supplemental material
-
Suppl. Fig. 1: Perfusion with cryo protective substances does not prevent the loss of the cellular resolution of the EGFP signal in cryo-sections.
(A) Transgenic mice expressing cytoplasmic EGFP under the control of the GFAP promoter were prior perfused with substances as indicated (see below). Cryo-sections (15µm) were analyzed for GFP fluorescence at the level of the cerebellum. Sections were cut from animals that were perfused with 4% buffered paraformaldehyde solution (PFA), phosphate buffered saline (PBS), PBS containing 10% sucrose (sucrose) or with PBS containing 10% glycerol and 1% DMSO. The loss of the cellular restricted GFP signal upon cryo-sectioning could not be prevented by the perfusion with any of the indicated substances except PFA. Other cryo-protective substances, such as different concentrations of PBS buffered trehalose, were also tested. None of the analyzed cryo-protective agents could prevent signal fading and the time dependent loss of GFP signal (data not shown). In addition, we analyzed cryo-sections of other transgenic mice expressing lower levels of EGFP under the control of the CamKIIα or the GAD65 promoter in principal or interneurons of the forebrain. In these mice, we were also not able to observe any cellular localized EGFP signal after cryo-sectioning (data not shown).
(B) The green fluorescent protein retains functionality after freezing. HeLa cells were transfected with EGFP expression plasmids and extracts were analyzed for GFP fluorescence after one and two freeze-thaw cycles and upon the addition of cryo-protective sucrose solutions. The first freeze thaw cycle does not decrease the total fluorescence, after a second freeze-thaw cycle, the addition of 10% sucrose rescued the loss of signal. Denaturing the GFP by boiling the samples in 1% SDS buffer leads to a complete loss of fluorescence.
(C) Summary of GFP derivatives tested for their relative signal stability after transient expression in eukaryotic cell lines. Data were compiled from transfections of COS7-, HeLa- and PC12 cells. Cells were PFA fixed, unfixed, or fixed with 70% ethanol, frozen at -80°C and subsequently dry mounted. The signal of all GFP derivatives attached to membranes or targeted to sub-cellular structures was resistant against dry mounting, if unfixed or even after ethanol fixation. The best signal stability with respect to morphology and intensity was observed for the nuclear targeted EYFP (EYFPnuc).
- supplemental material
-
Suppl. Fig. 2: Cluster analysis of glutamatergic genes
(A) Heatmap of normalized signal intensities for selected glutamatergic genes of the Y+MCx and Y+SSCx replicate samples along with compatible datasets obtained from purified principal neurons (YFP-S1, CSMN, CPN) and interneurons (G30-S1). The dendrogram tree and the increased relative signal intensities of presumable glutamatergic genes show the close relationship of the Y+MCx/Y+SSCx and the YFP-S1, CSMN, CPN samples. Gene symbols are depicted at the right side of the plot.
(B) Heatmap of three previously characterized marker genes for glutamatergic neurons neurogranin (Nrgn), NeuroD2/NDRF and NeuroD6/NEX. All markers are highly expressed in the microarray data sets obtained from presumable principal neurons and show a much lower expression in data obtained from purified interneurons.
- supplemental material
-
Suppl. Fig. 3: Cluster analysis of gabaergic genes
Heatmap of normalized signal intensities for selected gabaergic genes of the Y+MCx and Y+SSCx replicate samples along with compatible datasets obtained from purified principal neurons (YFP-S1, CSMN, CPN) and interneurons (G30-S1). The dendrogram tree and the decreased relative signal intensities of presumable gabaergic genes show the close relationship of the Y+MCx/Y+SSCx and the YFP-S1, CSMN, CPN samples. Among the gabaergic genes are the well known interneuronal marker genes encoding the glutamic acid decarboxylases 1 and -2 (Gad1, Gad2). Gene symbols are depicted at the right side of the plot.
- supplemental material
-
Suppl. Fig. 4: Comparison of Purkinje cell and complete cerebellar cortex expression profiles
(A) Numbers of up or down regulated genes comparing either Purkinje cells (PC400) and cerebellar cortex (Cbx) to other brain regions or cell types. The number of differentially regulated genes was determined with the MAS 5.0 software with a given cut-off with a minimum of 66% change calls and a SLR of at least 1.4. The number of genes found to be up or down regulated in the Purkinje cells isolated from cerebellar cortex (PC400) versus other brain samples is always substantially higher than for the comparisons with the whole cerebellar cortex (Cbx).
(B) Schematic drawing of all genes found to be up regulated in PC400 versus Cbx by a SLR of ≥3.0 (see whole data set in Suppl. Table 3). The genes were further sub-grouped in (I) genes with a PC400 restricted expression (P calls obtained only in the PC400 samples), (II) highly enriched (additionally with a SLR PC400 versus brain ≥3.0) and (III) enriched genes.
(C) Schematic drawing of all genes found to be upregulated in Cbx vs PC400 by a SLR of ≤-3.0 (see whole data set in Suppl. Table 3). The genes were further sub grouped in (I) highly enriched (additionally with a SLR PC400 vs brain ≥2.0), (II) Cbx enriched genes, (III) genes likely to expressed in blood or (IV) in glial cells.
(D) Average signal intensity plots of selected genes found to be enriched in PC400 versus Cbx samples.
(E) Average signal intensity plots of selected genes found to be enriched in Cbx versus PC400 samples.
(F) Northern Blot analysis of selected genes with aRNA samples from single-isolated PC400 samples and cerebellar cortex (Cbx), brain and kidney. Two identical northern blots were double plotted and signals were quantified, one blot is shown.
(G) Comparison of the genechip signal intensities (white bars) to the GAPDH normalized signals obtained with northern blotting (black bars).
All errors are depicted as standard deviation (SD).
- supplemental material
-
Suppl. Fig. 5: Genes enriched in the EYFP positive cortical layer V neurons
(A) Table of genes detected as restricted or enriched in the EYFP positive cells (Y+Cx) when compared to cortical micro-regions (upper block with a SLR≥2.0 compared to layer V cortical samples (Cx V), lower block with a SLR≥2.0 compared to layers I-VI cortical samples (Cx I-VI)). The genes marked in orange were found to be highly enriched in the Y+Cx cells with higher expression levels compared to all other samples of the study. The genes marked in yellow show a higher expression in Purkinje cells and the cerebellar cortex preparations.
(B) Average signal intensity plots of selected Y+Cx enriched genes.
(C) Northern Blot analysis of selected genes with aRNA samples from single isolated Y+Cx samples and cortical subregions (Cx V and Cx I-VI). The lower relative size distributions of the Y+Cx samples is due to a third amplification round that was necessary to obtain sufficient material (see Materials and Methods). Two identical northern blots were double plotted and signals were quantified, one blot is shown.
(D) Comparison of the genechip signal intensities (white bars) to the GAPDH normalized signals obtained by northern blotting (black bars).
All errors are depicted as standard deviation (SD).
