Lenses were removed and immediately fixed using 4% paraformaldehide or methanol-based UMFIX reagent (Sakura Finetek USA, Inc.). To map the exact location of both elongating and maturing fibers on the lens slices, lenses were vibratome-sliced (Vibratome 1000, St. Louis, MO) as described previously and the GFP expression pattern was captured by confocal microscopy. For RNA extraction lens tissue was sliced into 5 micron-thick paraffin sections and microdissected using LCM (Leica Microsystems, Bannockburn, IL). The mid-saggital slices were used both for syncytium border measurements and for LCM. Control measurement (data not shown) confirmed that similar rates of shrinkage in paraformaldehyde- and UMFIX-fixed samples did not affect precision of LCM dissection. Fiber cell samples were dissected out of 5 micron-thick slices. In one experiment we typically processed about 40 slices, which was sufficient to collect the minimum of 200 zone-specific cells pooled from P5 littermate lenses. This sample size provided a reliable representation of RNA species in experimental procedure originally designed and tested for just 1-10 cells. Localization of the syncytium border characterized by the abrupt change of GFP labeling pattern was captured by confocal microscopy as described previously. In brief, GFP fluorescence was visualized using LSM510 instrument (Carl Zeiss, NY) equipped with an argon/krypton laser at 488 nm excitation and a 515-565 nm band pass emission filter. Physical parameters of the zones containing young and maturing fibers were measured in fixed lens slices using the morhphometric software provided by Zeiss.
Lenses were fixed in 4% paraformaldehyde/PBS and sectioned with a vibratome. Cells from elongating and maturing fiber regions were dissected out using the Leica DMLA laser capture microscope, (Leica Microsystems, Bannockburn, IL). The cut-out pieces containing captured cells were put directly into tubes containing the lysis buffer supplied in the Absolutely RNA® nanoprep kit (Stratagene, La Jolla, CA). Total RNA from the microdissected tissue sections was extracted and purified using the Absolutely RNA nanoprep kit according to the manufacturer's protocol. Caps briefly placed onto the section without laser activation were used as negative controls. Samples from several age-matched lenses were pooled together to obtain differentiation-specific samples for microarray analysis. Target RNA amplification and labeling with Cy-3 or Cy-5 dyes from CyDye Post Labelling Reactive Dye Pack (Amersham, USA) was carried out in two rounds using the Amino Allyl MessageAmp™ aRNA Kit (Ambion, USA) as specified by the manufacturer. Quality and size distribution of the targets were determined by the Agilent 2100 Bioanalyzer (Agilent Technologies, USA). The Amino Allyl MessageAmp aRNA Kit is configured to incorporate the modified nucleotide, 5-(3-aminoallyl)-UTP (aaUTP) into the aRNA during in vitro transcription. Once purified and fragmented, the dye labeled aRNA was used for microarray hybridization. Labeled amplified RNA from three different biological experiments was hybridized to the 22K Mouse Oligo microarrays (Agilent Technologies)) according to the manufacturer's instructions. For each biological replicate we performed two technical subreplicates for dye-swap. After hybridization microarrays were washed and scanned using GenePix 4000A (Axon Instruments,Inc.).
