Transcriptome analysis of synaptoneurosomes identifies neuroplasticity genes overexpressed in incipient Alzheimer's disease

Abstract

In Alzheimer's disease (AD), early deficits in learning and memory are a consequence of synaptic modification induced by toxic beta-amyloid oligomers (oAbeta). To identify immediate molecular targets downstream of oAbeta binding, we prepared synaptoneurosomes from prefrontal cortex of control and incipient AD (IAD) patients, and isolated mRNAs for comparison of gene expression. This novel approach concentrates synaptic mRNA, thereby increasing the ratio of synaptic to somal mRNA and allowing discrimination of expression changes in synaptically localized genes. In IAD patients, global measures of cognition declined with increasing levels of dimeric Abeta (dAbeta). These patients also showed increased expression of neuroplasticity related genes, many encoding 3'UTR consensus sequences that regulate translation in the synapse. An increase in mRNA encoding the GluR2 subunit of the alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid receptor (AMPAR) was paralleled by elevated expression of the corresponding protein in IAD. These results imply a functional impact on synaptic transmission as GluR2, if inserted, maintains the receptors in a low conductance state. Some overexpressed genes may induce early deficits in cognition and others compensatory mechanisms, providing targets for intervention to moderate the response to dAbeta.

Figure 1. Enrichment and stability of synaptoneurosomes.

Microscopy: A) Nuclear contamination of the homogenate (2 µl smears) (DAPI fluorescence) decreases after sequential passage through mesh screens with B) the post- 10 µm screen synaptoneurosome pellet (100 µl volume) still retaining a few nuclei (Bar = 20 µm). C) Intact synaptoneurosomes are detected by phase contrast in smears of pelleted post-10 µm screen synaptoneurosome as typical “snowman” pre- and postsynaptic profiles (arrowheads). Larger, empty, membranous structures are also observed. (Bar = 5 µm). D) Electron microscopy reveal “snowman” profiles with apparent postsynaptic densities (arrowheads); (Bar = 300 nm). E) At higher magnification, presynaptic terminals containing synaptic vesicles, 20 nm in size (white arrows) are indicated (Bar = 100 nm). Immunoblots: F) Homogenates (black bars) and synaptoneurosomes (white bars) were compared for stability and enrichment of synaptic proteins and contamination with other cellular debris. Densitometric comparison on immunoblots of postsynaptic proteins PSD-95, NMDAR1 and GluR2 in synaptoneurosomes shows more than a two-fold enrichment compared to homogenates, an increase in presynaptic protein SNAP-25 but a 40% decrease in glial protein GFAP. Cytoplasmic proteins β-tubulin and PKCα show little change. The ratio of each protein to GAPDH is given in arbitrary units based on the integrated density value which is the sum of all pixel values after background correction (IDV). G) In synaptoneurosomes, de novo synthesis of several proteins is observed as newly translated, biotinylated proteins detected with streptavidin (lane 1), and one at 50 kD co-migrates with a band detected with antibody to αCAMKII (lane 4,arrowhead). Actinomycin D, a transcription inhibitor, does not reduce the protein band profile (lane 2). Protein synthesis is inhibited by the translation inhibitor, anisomycin (lane 3), although endogenously biotinylated proteins present in synaptoneurosomes are still seen at ∼140 and 75 kD (asterisks).

Figure 2. Oligomeric Aβ increases with dementia progression.

Oligomeric Aβ dimer (dAβ) is detected with antibody 4G8 on immunoblots of A) whole homogenate or B) synaptoneurosomes prepared from patients with a range of MMSE scores (29-15). A band observed at <10 kD with a migration identical to dAβ, is detectable in IAD patients with MMSE scores of about 26 or less. In C) densitometry reveals that, when normalized to GAPDH, dAβ levels are inversely related to MMSE although patients with an ApoE4 allele (including one control) exhibit higher concentrations of dAβ than patients with comparable MMSE scores but no E4 allele. Homogenate (grey bar) and synaptoneurosome (white bar) samples show increasing dAβ in IAD as the disease progresses. Data are representative of 3 separate experiments. The ratio of dAβ to GAPDH is given in arbitrary units based on the integrated density value which is the sum of all pixel values after background correction (IDV).

Figure 3. Microarray and Cluster Analysis.

In A) Genes were filtered out on the “present” call (P-call) generated by MAS 5.0 leaving 12,536 genes. A 2-way ANOVA test performed on log10-transformed data comparing control with IAD detects the significantly expressed genes. Changes in gene expression in the IAD patients were evaluated at 3 levels of significance (p<0.05, p<0.01, p<0.001) for patients grouped by MMSE values and summarized in the chart below, B). We used the p<0.01 analysis gene list for further study. In C), cluster analysis shows relative mRNA message in controls (black bar) and IAD patients (green bar) with increased expression (green to red) or decreased expression (red to green). Blue line above the heatmap encompasses patients with gene profiles transitional between control and IAD. See Table S2 for genes lists with gene name, symbol and fold change.

Figure 4. Quantitative RT-PCR confirms upregulation of GLUT1 and GluR2.

Total RNA (2 µg) was used to generate cDNA by reverse transcription using oligo-dT primers. Note increased mRNA levels of glutamate transporter (GLUT1) and GluR2 in IAD patients compared to controls. Data are mean values ±SEM comparing 4 control and 4 IAD patients. βCAMKII and GAP43 do not show significant change in mRNA levels. GAP-3 (p<0.1), GLUT1 (* p<0.01), GluR2 (# p<0.03), βCAMKII (p<0.06).

Figure 5. GluR2 protein expression.

To determine if increased GluR2 mRNA generates elevated protein expression, we compared subjects with a range of MMSE values. Representative immunoblots of A) homogenates and B) synaptoneurosomes, (10 µg protein per lane) were detected with antibodies for GluR2 and GAPDH. C) In controls, the increased ratio (2-fold) of synaptoneurosome (white bar) GluR2 to total GluR2 in homogenates (* p<0.0002) is an indication of synaptic protein enrichment and this ratio increases by 50% in IAD patients (p<5.8E-08). In the IAD group there is a significant increase in synaptoneurosome GluR2 protein expression compared to control (** p<6.0E-05). Homogenate GluR2 (black bar) remains constant in controls and IAD. Data are mean values ±SEM for 4 controls and 4 IAD analyzed in 5 separate experiments. The ratio of GluR2 to GAPDH is given in arbitrary units based on the integrated density value which is the sum of all pixel values after background correction (IDV).