Dissection of discrete brain regions for molecular analysis is complicated by trade-offs between accuracy, flexibility, and costs. We developed a flexible and cost-effective method, in situ hybridization (ISH) guided freeze-matrix assisted punches (IFAP), for extracting nanogram quantities of DNA from slide-mounted sections as thin as 12 μm. Using ISH to localize regions of interest, tissue is targeted by applying a small bead of M-1 embedding matrix onto cryosections, snap-freezing, and collecting the beads for nucleic acid purification. The method quantitatively recovers RNA and DNA usable for PCR and DNA methylation analysis.
DNA methylation is a mechanism by which long-term changes in gene expression are effected. Robust changes to the pattern of DNA methylation in homogeneous macroscopic tissues (e.g., carcinomas or cultured cells) are well established (1, 2). It has been posited that experience-dependent DNA methylation also occurs in healthy normal brain tissue, but data has remained limited and controversial (3). Issues pertaining to anatomical specificity, sample recovery, or high cost have limited progress.
The microdissection of tissue sections is tedious, requires thick sections, and is limited by punch availability. Recovery of tissue from fresh-frozen thin sections, suitable for anatomical studies, is all but impossible. We have developed a technique, based on the principle of in situ hybridization (ISH)-guided laser capture microdissection (LCM; 4), for punching portions of tissue from slide-mounted frozen thin sections. By applying a small bead of liquid onto cryosections and snap-freezing, underlying tissue is lifted when beads are pulled away, and DNA is extracted from bead-tissue complexes using conventional high-salt DNA extraction techniques. This method, ISH-guided freeze-matrix assisted punches (IFAP), facilitates rapid and efficient recovery of discrete regions under conditions suitable for DNA methylation analysis. IFAP utilizes commonly available histological resources to maintain low cost and allows for parallel analyses from source tissue.
The IFAP method synergizes well with standard histological workflows. Tissue processing is streamlined, with sections (≥12 μm) mounted on n slide sets (sections 1, 1+n, 1+2n…on set 1; sections 2, 2+n, 2+2n…on set 2; etc., up to sections n, 2n, 3n…on set n). Each set represents a staggered survey through the region of interest. An ISH is performed on one set to localize the target area and is used to guide IFAP dissection on an adjacent set. IFAP in principle works similar to LCM, but instead uses low temperatures to bond tissue with pipet-applied matrix beads. Although not comparable in spatial specificity to LCM, IFAP is low-cost, rapid, and useful in millimeter scale dissections that do not require single-cell resolution.
The default approach to isolating nuclei from surrounding tissue is the micropunch (MP), which utilizes a hollowed sharpened cylinder to punch tissue (5). However, IFAP offers several advantages: (i) flexibility in sample condition and (ii) workflow efficiency for parallel analyses. In our experience, MP is typically more effective on fresh tissues and requires thicker sections (≥ 60 μm) if frozen and slide-mounted. Additional difficulty occurs if sections are mounted on charged or treated slides, which are commonly used for histochemistry to improve sample retention. A theoretical workaround is to alternate thick and thin slices along with slide coatings, but in practice that is cumbersome and complicated. IFAP offers a slice resolution of at least 12 μm and works with previously stored slide-mounted sections, including charged slides. Thus, it introduces nothing new to the workflow and can be readily applied in a range of conditions. In this study, we detail the use of IFAP in excision of brain nuclei and evaluate sample integrity for use in assessing DNA methylation. We use tryptophan hydroxylase 2 (TPH2), a predominantly raphe-restricted gene, as our test gene.
For IFAP, three brains from postnatal day 60 c57/bl6 male mice were sectioned at 12 μm and mounted on Superfrost slides (Thermo Fisher Scientific, Waltham, MA, USA). A fourth mouse brain, for MP comparisons, was sectioned by alternating between 12- and 100-μm slices, with the 12-μm sections used for ISH-guidance and the 100-μm sections reserved for MP. This alternation of section depth is necessary because MP is impractical on thinly sliced, frozen slide-mounted sections. Slides were stored at -80°C, then reequilibrated to -20°C prior to dissection. All supplies were preequilibrated to -20°C to prevent beads from reliquefying. For liquid application, TipOne ultra-low retention pipet tips (USA Scientific, Ocala, FL, USA) were used to minimize volume variance.
Initially, tests were run to characterize the suitability of different liquids—TE (10 mM Tris-HCl, 0.1 mM EDTA, pH 8.0), water, or M-1 embedding matrix (Shandon; Thermo Fisher Scientific)—for binding and lifting tissue. The rationale for testing M-1, a water-soluble embedding matrix conventionally used to support sectioning of frozen tissue, was that its viscosity would minimize variability in bead diameter. Per liquid, beads were tested for (i) dispersal and uniformity and (ii) structural integrity during removal. Diameters were measured (ImageJ) every 30° across each bead. For 0.5 μL, the average bead diameter for M-1 was ~1.5 mm, whereas for TE and water it was ~2 mm (Figure 1A). Moreover, bead diameter variance was less for M-1 compared to TE or water beads (Figure 1B), indicating better uniformity in bead dimensions. Qualitatively, water and TE beads were more prone to fracture during removal from slides, negatively affecting recovery. For these reasons, M-1 was considered optimal for IFAP dissections.