Late memory-related genes in the hippocampus revealed by RNA fingerprinting

Abstract

Although long-term memory is thought to require a cellular program of gene expression and increased protein synthesis, the identity of proteins critical for associative memory is largely unknown. We used RNA fingerprinting to identify candidate memory-related genes (MRGs), which were up-regulated in the hippocampus of water maze-trained rats, a brain area that is critically involved in spatial learning. Two of the original 10 candidate genes implicated by RNA fingerprinting, the rat homolog of the ryanodine receptor type-2 and glutamate dehydrogenase (EC 1.4.1.3), were further investigated by Northern blot analysis, reverse transcription-PCR, and in situ hybridization and confirmed as MRGs with distinct temporal and regional expression. Successive RNA screening as illustrated here may help to reveal a spectrum of MRGs as they appear in distinct domains of memory storage.

Figure 1

MRG cDNA cloning experiments on water maze-trained vs. control rats. (A) Escape latencies of rats swimming to a submerged platform in the water maze. Rats were trained for four consecutive times to locate the submerged platform. The values are mean ± SEM (n = 36). The shorter latency time is statistically significant (ANOVA with repeated measures F_3.102 = 21.95, P < 0.0001; the trend is linear). (B) Example of RNA fingerprinting. Two dilutions (20, 5 ng) of reverse-transcribed hippocampal RNA from control swimming and water maze-trained rats were PCR fingerprinted using a pairwise combination of arbitrary primers. PCR products were electrophoresed on a sequencing gel and visualized by autoradiography. Arrows mark three potential differentially expressed cDNAs. Clones that appeared up-regulated were sequenced and further analyzed by Northern blot analysis, RT-PCR, and in situ hybridization.

Figure 2

Hippocampal changes in RYR2- and GDH-mRNA levels 6 hr after water maze training. (A) Northern blot analysis of RYR2 and GDH mRNAs in control swimming and water maze-trained rats. Hippocampal RNA from control swimming (lane 1) and water maze-trained rats (lane 2) was hybridized with the labeled probes for RYR2, GDH, and PGK1. (B) Relative mRNA levels of RYR2 and GDH. To quantify the amount of RYR2- and GDH-mRNA levels, the densitometry of their hybridization signal was divided by that of PGK1 obtained in the same sample and expressed in relative units.

Figure 3

Time-course changes of RYR2- and GDH-mRNA levels after water maze training. (A) RT-PCR analysis of RYR2 and GDH mRNAs in control swimming and water maze-trained rats at 2, 6, 12, and 24 hr after training. (B) Relative mRNA levels of RYR2 and GDH in control swimming and water maze-trained rats at 2, 6, 12, and 24 hr after training (n = 6; ∗, P < 0.01, ∗∗, P < 0.001).

Figure 4

Localization by in situ hybridization of RYR2 and GDH mRNAs in hippocampal subfields after water maze training. (A) Pseudocolor representation of RYR2- and GDH-mRNA expression in control swimming and water maze-trained rats. Six hours after training, a general increase of GDH mRNA is seen in the pyramidal layer (P) of CA areas and DG; increases also emerge in the lacunosum molecolare (ml) of CA1, cingulate cortex (Cg), and laterodorsal nucleus of thalamus (LD). RYR2-mRNA expression is low in CA1 and increased predominantly in CA3 and DG after water maze training. Color spectrum on the right side of the figure represents the pixel value of gray levels. (Scale bar = 800 μm.) (B) Relative mRNA levels of RYR2 and GDH in different hippocampal subfields of control swimming and water trained rats. Quantification of induction increase is achieved by comparison of pixel values of an area of interest in four sections from each of four pairs of rats. Changes in mRNA levels are expressed as the density ratio of trained-to-control animals. Values are mean ± SEM.