To test IFAP specificity, the dorsal raphe, a region rich in serotonergic (5-HT) neurons, was targeted for extraction from surrounding brainstem tissue. ISH histochemistry for TPH2 (Figure 1C) guided IFAP dissection on adjacent sections (Figure 1D). The ISH was used to identify (i) raphe-containing sections and (ii) localize the tissue areas corresponding to the highest TPH2 optical density within a 1.5 mm diameter circular area. For the full ISH protocol, see Reference 6. Template for sense and antisense probes were generated from cDNA [expressed sequence tag (EST): BG084420]. Using ISH as a reference, each slide was briefly brought to room temperature, and a 0.5-μL aliquot of M-1 embedding matrix applied to each section on the slide. After application, slides were quickly placed onto dry ice to snap-freeze the bead (Figure 1D). For punch removal, slides were transferred to the cryostat, and the blunt edge of a razor blade was used to dislodge each bead from the slide, simultaneously lifting tissue (Figure 1E). A total of 10 beads per sample was used. After all beads were collected in a chilled 1.5-mL tube, the tube was briefly centrifuged for 5–10 s at ~4000× g.
To verify IFAP accuracy, RNA was purified (TRIzol; Invitrogen, Carlsbad, CA, USA), resuspended (20 μL), reverse-transcribed (10 μL) using random 8-mers (SuperScript II; Invitrogen), and cDNA amplification from raphe-specific genes, TPH2 and LMX1B, compared against RNA from an intact brainstem section from an adjacent slide. For cDNA amplification, the following exon-spanning primers were used: LMX1B: 5′-CTGCTGTGCAA-GGGTGACTA-3′/5′-AAACCAGA-CCTGGACCACAC-3′, TPH2: 5′-TGTCCTTGGATTCTGCTGTG- 3′/5′-GCCCACCAACTTCATT-CTTC-3′, and HPRT: 5′-CAGTA-CAGCCCCAAAATGGT-3′/5′- GCGCTCATCTTAGGCTTTGT-3′. As expected, because of the higher concentration of raphe tissue extracted using IFAP, IFAP samples successfully amplified TPH2 and LMX1B, whereas RNA from the brainstem section only amplified the ubiquitously expressed HPRT (Figure 1F). This confirms specificity for targeting nuclei and regionally restricted genes, as well as integrity of RNA by this method. Amplicons for TPH2 and LMX1B were TOPO cloned (Invitrogen) and sequence-verified.
For genomic DNA (gDNA) extraction, 10 beads per sample were treated to a standard high-salt TNES DNA extraction protocol (7) and resuspended in 25 μL TE (10 mM Tris, 0.1 mM EDTA). For TE beads, 1 μL 10 mg/mL glycogen was added as a coprecipitant prior to precipitation, whereas M-1 contained an inherent coprecipitant (likely polyethylene glycol) and did not require glycogen. As a result, DNA pellets were also more gelatinous and required additional time (5 min) and heat (50°C) for resuspension. Although in our downstream tests, M-1's unknown proprietary ingredients had no adverse effect on bisulfite conversion or sequencing, Nanodrop (Thermo Fisher Scientific) readings revealed elevated peaks at 220–230 nm, which complicated accurate gDNA quantitation. This necessitated alternative quantitation by SYBR green fluorescence. Each gDNA sample was diluted 1:100 in TE, pH 7.5 (10 mM Tris, 0.1 mM EDTA), and mixed with an equal volume of a 1:1000 dilution of SYBR green I (Invitrogen) in TE, pH 7.5, for a final volume of 200 μL. Fluorescence was compared to a DNA standard curve with M-1 added (0.5 μL per bead in sample) to extrapolate yield. Fluorescence quantitation of gDNA from 10 beads returned ~400 ng gDNA. For comparison, commercial kits typically return ~1–3 μg gDNA/mg brain tissue. Based on 0.5 μL IFAP beads generating ~1.5 mm diameter punches of 12 μm thickness, 10 punches would have an estimated mass of ~212 μg (assuming density of 1), returning 212–646 ng gDNA. Observed yield (400 ng) is well within expectations.
To test usability of IFAP-extracted samples for DNA methylation analysis, we assessed the sequence quality of gDNA obtained using IFAP (M-1 or TE) from slide-mounted thin sections (10 beads) against tissue obtained using a 1.0 mm diameter MP (VWR International, Radnor, PA, USA) on thick sections and replicated across three independent samples. gDNA was converted with sodium bisulfite (EpiTect; Qiagen, Valencia, CA, USA), and DNA methylation was assessed at the TPH2 promoter and 5′ untranslated region (UTR) using a nested PCR approach, with inner nested primers fused to an M13 adaptor to improve sequencing. Outer nested primers were: 5′-TAGAYGTGTAATTTGATTGTGGTTATTAGT-3′ and 5′-TCCCAACAACT-CRCCCAACTAC-3′. Inner nested primers were: 5′-M13fwd-TGATTGTGGTT-ATTAGTAATT-AGAAATGAGTT-3′ and 5′-M13rev-AATCCAAAAA-CAACCCTCTCC-3′. Amplicons were column-purified (QIAquick; Qiagen) and sequenced bidirectionally (Model 3730 XL; Applied Biosystems, Foster City, CA, USA). Percent methylation was estimated by comparing cytosine:thymine peak ratios in CpG dinucleotides [% methylation = (C)/(C + T)]. Sequence chromatograms of MP and IFAP M-1 or TE indicated comparable conversion of unprotected cytosines (Figure 2A). Furthermore, no difference in CpG dinucleotide methylation was observed, indicating minimal IFAP interference with bisulfite conversion (Figure 2B).