Region-Specific Gene Expression in Adult Mouse CNS Tissues

Preparation of retina sections. Whole eyes of n = 3 C57BL/6 adult (two month old) mice were dissected, flash frozen in tissue freezing medium and stored at –80 °C. Sections (15 µm) were prepared with a Leica CM1850UV cryostat and placed on Leica PEN-membrane slides. Sections were fixed in cold 75 % ethanol (2 min), washed twice in RNAse-free water (30 sec) and stained in Meyer’s hematoxylin (1 min), to visualize the three retinal nuclear layers: ganglion cell layer (GCL), inner nuclear layer (INL) and outer nuclear layer (ONL) (Figure 1A). Sections were then washed in RNAse-free water (2 x 30 sec), dehydrated in graded ethanol series and air-dried (15 min).

LCM procedure. LCM was performed with a Leica LMD6500 microdissector using a 5x objective in brighfield with TL-BF contrast method. The laser parameters used for the dissection were: power 60, aperture 7, speed 7, specimen balance 46 and offset 25. Retinal layers were microdissected using the pencil function to border each single layer in the following order: GCL, INL and ONL. Figure 1A shows the outline used for GCL dissection and Figure 1B that used for ONL dissection.

RNA extraction and qRT-PCR analysis. Microdissected layers derived from the 3 animals were collected and pooled in 500 µL tubes. Total RNA was extracted from each pool using Nucleospin XS RNA columns (Macherey-Nagel) according to manufacturer’s instructions for LCM assays. Amount and quality of RNA were evaluated measuring the OD at 260 nm, the 260/280 and the 260/230 ratios by Nanodrop (Celbio). We obtained the following yields for each pool: GCL 150 ng, INL and ONL 450 ng each.

To test the precision in microdissecting single layers we performed qRT-PCR to analyze the expression of rhodopsin, a protein involved in phototransduction specifically expressed in photoreceptors, whose nuclei are located in the ONL (Liu et al., 2004, Cell Tissue Res 315: 197–201). Reverse transcription was performed on 150 ng of total RNA using the SuperScript VILO cDNA Synthesis Kit (Invitrogen), according to manufacturer's instructions. PCR was performed on 10 ng of cDNA in a Rotor-Gene 6000 (Corbett) using the KAPA SYBR FAST Universal qPCR kit (Resnova), with the following conditions: 95°C 5 min followed by 40 cycles (95 °C 15 sec, 60 °C 20 sec, 72 °C 40 sec). Melting curve ramp was from 55 °C to 95 °C at 0.2 degrees/sec. The relative levels of rhodopsin mRNA were normalized to the ribosomal protein gene L41 (Tripathi et al., 2009, Neuroscience 159: 842–849). All primers pairs showed single peaks in the melting curve analysis.

Primers were as follows:

rhodopsin (NM_145383.1):

• forward 5'GCCTGAGGTCAACAACGAAT3',
• reverse 5'GATAACCATGCGGGTGACTT3'

L41 (NM_018860):

• forward 5'GGTTCTCCCTTTCTCCCTTG3',
• reverse 5'GCACCCCGACTCTTAGTGAA3'   

As expected, rhodopsin expression was detected only in the ONL, thus confirming the specificity of the laser microdissection and the purity of the extracted RNAs (Figure 1C).

Brain slicing. Adult (3–5 month old) mouse brains were dissected, immediately frozen in dry ice and stored at –80 °C until use.  Before cutting, brains were kept overnight at –20 °C, and then allowed to habituate at –16 °C in the Leica CM1850UV cryostat for 1 hour. Coronal sections (25 µm) were cut throughout the full extent of the dorsal hippocampus and mounted onto Leica PEN-membrane slides. Care was taken to avoid rapid changes in temperature to minimize condensation formation on the slide, therefore reducing potential RNase activity. Following cutting, slides were stored in an airtight box and stored at –20 °C.

LCM procedure. Slides were slowly warmed to room temperature and allowed to dry ~30 min. Slides were then fixed with 70 % ethanol (with RNAse-free water) for 1 min, washed in RNAse-free water for 1 min to remove OCT, briefly rinsed in 70 % ethanol (with RNAse-free water) and allowed to dry completely for ~1 hour (incomplete drying results in poor quality laser cutting). CA1 and CA3 subregions from the dorsal hippocampus were dissected using the “Draw and Cut” function of the Leica LMD6500 microdissector and collected in tube caps (Figure 2A–D). The 10x objective with TL-BF contrast was used. The laser parameters used for the dissection were: power 60, aperture 12, speed 10, specimen balance 45 and offset 40. To improve quality and accuracy of cutting the concave side of the area to be dissected was cut first to prevent the remaining tissue from falling out of the focal plane. The total volume of tissue dissected from each subregion was CA1: 3.1 ± 0.3 mm³, CA3: 3.8 ± 0.4 mm³ (approximately corresponding to 75 sections per brain).

RNA extraction and gene microarray analysis. Total RNA was extracted from each pool of CA1 and CA3 tissue using the Qiagen AllPrep RNA/Protein Kit according to the protocol for LCM samples. Mean RNA yields from CA1 and CA3 were 425 ± 51 and 616 ± 96 ng per brain, respectively. Analysis of RNA integrity was performed using the 2100 Bioanalyzer (Agilent) with an RNA integrity number (RIN) of 5.4 ± 0.6 and 5.2 ± 0.2 recorded for CA1 and CA3 respectively. Mouse gene expression arrays (Agilent 4X44K slides) were hybridized with probes obtained from the extracted RNAs. Microarray data were analyzed with the Agilent GeneSpring GX software and gene ontology performed using DAVID functional annotational tools (Huang et al. 2009, Nature Protoc 4: 44–57). Significant regulation of genes involved in a number of biological processes were observed (Figure 2E). Figure 2F displays an expression heatmap of genes associated with potassium ion transport, one of the biological processes significantly regulated in CA1 versus CA3.