Cellular Distribution of GAP-43 mRNA in Hippocampus and Cerebellum of Adult Rat Brain by In Situ RT-PCR

Abstract

The growth-associated protein GAP-43 is a presynaptic membrane phosphoprotein that plays a key role in guiding the growth of axons and in modulating the formation of new synapses. To identify the cells that synthesize GAP-43 mRNA, we applied direct in situ reverse transcription-polymerase chain reaction (in situ RT-PCR) in cerebellum and hippocampus of adult rat brain. In situ RT-PCR revealed GAP-43 mRNA in cerebellar granule cells, in Purkinje cells and in some interneurons of the molecular layer. Previous in situ hybridization studies had demonstrated a dense label throughout the granular layer of the cerebellar cortex but no labeling of other cerebellar neurons. Hippocampal cells showing distinct GAP-43 mRNA signal after in situ RT-PCR were CA1 and CA3 pyramidal neurons, CA4 hilar cells, and dentate gyrus granule cells, whereas in situ hybridization studies had detected GAP-43 mRNA only in CA3 and CA1 pyramidal neurons. Our data indicate that GAP-43 mRNA is widely distributed, suggesting that many cell types are potentially involved in synaptic plasticity events. (J Histochem Cytochem 49:1195–1196, 2001)

The growth-associated protein GAP-43 is a neuron-specific phosphoprotein whose expression has been used as an indicator of axonal growth (Benowitz and Routtenberg 1997). In situ hybridization (ISH) studies have demonstrated that GAP-43 mRNA is mostly concentrated in mitral cells of olfactory bulb, in cerebral cortex, in CA3 region of hippocampus, and in cerebellar granule cells of adult rat brain (Kruger et al. 1993). Because the threshold level for mRNA detection is 20 copies per cell in ISH, we investigated GAP-43 mRNA distribution in rat brain by in situ RT-PCR, a molecular technique that allows detection of even a single copy of the mRNA molecule.

Animals were sacrificed according to the guidelines of Italian Ministry of Health regarding the use of laboratory animals. Briefly, adult female Wistar rats were anesthetized and perfused intracardially with 4% paraformaldehyde. Brain samples were sectioned at 10 μm on a cryostat and sections were thaw-mounted on silane-coated slides tested for in situ PCR amplification. The hippocampus was permeabilized with 1 μg/ ml and cerebellum with 2 μg/ml proteinase K for 15 min at 37C. After proteinase K inactivation at 95C for 2 min, sections were treated with RNase-free DNase I overnight at 37C and rinsed in DEPC-treated water. The RT reaction was carried out using the first-strand cDNA synthesis kit (Roche; Basel, Switzerland). Direct in situ PCR of cDNA was performed on the GeneAmp In Situ PCR System 1000 (Perkin–Elmer; Norwalk, CT). The following conditions were used: 1 × reaction buffer, 2.5 mM MgCl₂, 200 μM of each deoxynucleotide, 50 μM Bio-11-dUTP, 1 μM of each primer (GAP-43 sense 5′-ATG CTG TGC TGT ATG AGA AGA ACC-3′; GAP-43 antisense 5′-GGC AAC GTG GAA AGC CGT TTC TTA AAG T-3′), 0.2 U/μl Taq DNA polymerase. The hot-start technique was performed. Twenty cycles were run using the following protocol: denaturation at 94C for 45 sec, annealing at 69C for 1 min, and extension at 72C for 1 min. Negative controls were performed by omission of primers. After amplification, sections were washed in 2 × SSC at 50C for 5 min, blocked in 5% BSA, and exposed to avidin DN (10 μg/ml). Sections were then incubated in biotin–alkaline phosphatase (10 μg/ml) and transferred to 0.1 M Tris-HCl, pH 9.5. Visualization of the amplified biotin-labeled cDNA was performed with NBT/BCIP chromogen. Finally, sections were counterstained with methyl green, dehydrated, cleared in xylene, and coverslipped.

Figure 1

In situ RT-PCR of GAP-43 mRNA in rat cerebellum. (A) Positive labeling can be observed in the granular layer (gl), in Purkinje cells (Pc), and in some interneurons (ml). (B) Negative control displays a prevalence of pale staining over the dark specific labeling. Bars = 40 μm.

Figure 2

GAP-43 mRNA in rat hippocampus. (A–C) Positive reaction is seen in granule cells (gc) and in CA1 and CA3. (D–F) Negative controls for each hippocampal subregion. Bars: A, C, D, E = 40 μm; B, F = 15 μm.

Figure 1A shows in situ RT-PCR of GAP-43 mRNA in rat cerebellum. Positive labeling can be observed in the granular layer (gl), in Purkinje cells (Pc), and in some interneurons of the molecular layer (ml). Figure 1B shows the negative control performed by omission of primers, demonstrating a prevalence of pale staining, due to methyl green counterstain, over the dark specific labeling. Figures 2A–2C show the cellular distribution of GAP-43 mRNA in hippocampal granule cells of dentate gyrus and in pyramidal neurons of layers CA1 and CA3. Dentate gyrus granule cells (gc) are stained, as well as some CA4 hilar neurons (h). CA1 and CA3 pyramidal cells are positive, whereas neuropil areas are faintly colored. Figures 2D–2F show the negative controls of dentate gyrus, CA1, and CA3, respectively. Previous ISH studies revealed a dense label throughout the granular layer of the cerebellar cortex, high levels of GAP-43 mRNA in CA3 cells of hippocampus and low levels in CA1 cells, but no labeling of other neurons (Kruger et al. 1993). Our data show that GAP-43 mRNA is present, although at low levels, in many cell types, indicating that these cells might be involved in dynamic processes of synaptic remodeling.