Temporal Quantitative Profiling of Newly Synthesized Proteins during Aβ Accumulation

Abstract

Accumulation of aggregated amyloid beta (Aβ) in the brain is believed to impair multiple cellular pathways and play a central role in Alzheimer's disease pathology. However, how this process is regulated remains unclear. In theory, measuring protein synthesis is the most direct way to evaluate a cell's response to stimuli, but to date, there have been few reliable methods to do this. To identify the protein regulatory network during the development of Aβ deposition in AD, we applied a new proteomic technique to quantitate newly synthesized protein (NSP) changes in the cerebral cortex and hippocampus of 2-, 5-, and 9-month-old APP/PS1 AD transgenic mice. This bio-orthogonal noncanonical amino acid tagging analysis combined PALM (pulse azidohomoalanine labeling in mammals) and HILAQ (heavy isotope labeled AHA quantitation) to reveal a comprehensive dataset of NSPs prior to and post Aβ deposition, including the identification of proteins not previously associated with AD, and demonstrated that the pattern of differentially expressed NSPs is age-dependent. We also found dysregulated vesicle transportation networks including endosomal subunits, coat protein complex I (COPI), and mitochondrial respiratory chain throughout all time points and two brain regions. These results point to a pathological dysregulation of vesicle transportation which occurs prior to Aβ accumulation and the onset of AD symptoms, which may progressively impact the entire protein network and thereby drive neurodegeneration. This study illustrates key pathway regulation responses to the development of AD pathogenesis by directly measuring the changes in protein synthesis and provides unique insights into the mechanisms that underlie AD.

Keywords: Alzheimer’s disease; Aβ accumulation; newly synthesized proteins; quantitative profiling.

Figure 1.

Schematic of labeling and AD pathogenesis confirmation. A: Pulse labeling of AHA/hAHA in mice. After feeding mice with AHA or hAHA diet for 4 days at 2, 5, 9 months, the mice were sacrificed and the tissues were extracted. B: ELISA analysis confirmed that Aβ levels are increased in both cerebral cortex and hippocampus along with the age of APP/PS1 mice. The cortex and hippocampus homogenates were assayed at 2, 5 and 9 months for human-specific Aβ42.Bars, means±S.D. (n=3). C: The schematic diagram of the AHA/hAHA labeling in mice cells through 4 days of feeding. D: Incorporation of AHA/hAHA is decreased with age in both cerebral cortex and hippocampus. Equal amount of proteins from hAHA and AHA labeled mice were mixed. Biotin can be added to AHA/hAHA proteins through click reaction and its detection can reflect the Incorporation of AHA/hAHA in cerebral cortex and hippocampus of 2, 5, 9 months mice by dot blot. E: The grayscale of dots in D were measured by image J and was presented by the points. Bars, means±S.D. (n=3).

Figure 2.

Identification and quantitative results A: The number of NSPs identified from an equal amount of AHA /hAHA protein mix by mass spectrometer are decreased with age in both cerebral cortex and hippocampus. Bars, means±S.D. (n=3). One-way ANOVA analysis was performed between time points. *p<0.05. **p<0.01. B: There is no significant difference in the number of NSPs between AD and WT in different age of mice with One-way ANOVA analysis performed between AD and control mice. Bars, means±S.D. (n=3). C&D. Venn diagram analysis for total quantified proteins at 2, 5, 9 months in cerebral cortex and hippocampus.

Figure 3.

Cluster analysis. Heat map showing mean-normalized protein fold changes for 247 overlapped proteins and 210 proteins over all time points in cerebral cortex (A) and hippocampus (B), respectively.

Figure 4.

Differentially expressed changes. A. Graphs of age-dependent increase of the percentage of significantly altered proteins in total quantified datasets. B. The distribution of significantly upregulated and down-regulated proteins at each time point. C&D. Venn diagram analysis for significantly changed proteins at 2, 5, 9 months in cerebral cortex and hippocampus. E. Biological process annotation on the significantly altered proteins at 2, 5, 9 months in cerebral cortex and hippocampus using STRING database. The 6 biological processes identified with the lowest false discovery rate (FDR) are displayed.

Figure 5.

STRING network analysis of significant altered NSPs, which were enriched in vesicle related processes. Only interactions with a STRING score ≥ 0.7 are shown. Node colors are linearly related to fold-change (Rectangular node excluded). Node shapes represent different timepoints and brain region. The shape “rectangular” represents the overlapped NSPs fold-changes indicated in bar graph. The column in red means increased synthesis while that in green means down regulation. The overlapped NSPs were listed with bar graph annotated from left to right. APP: 2mon_CTX; 2mon_Hipp; 5mon_Hipp; 9mon_CTX; Cap1: 2mon_Hipp, 9mon_CTX; Cops5: 2mon_Hipp, 5mon_Hipp; Ddx5: 5mon_CTX; 9mon_Hipp; Ehd3: 5mon_CTX; 9mon_Hipp; Mon2: 2mon_Hipp; 5mon_CTX; Rab1A: 9mon_CTX; 9mon_Hipp

Figure 6.

A: The subcellular distribution of significantly changed NSPs which were enriched in vesicle transportation process. B: More detailed annotation of endosomal proteins from Figure 6A according to previous publications. (Endosome recycling: Vps26b, Arhgap44, Psen1, Tnik, Ehd3, Gpc1; Endosome traffic: Ccdc93, Vps45, Osbpl1a, Sh3gl2; Endosome sorting: Vps37c, Rab1a; Endocytosis/Exocytosis: Lrpap1, Cadps, Rab5c. C: Verification of NSPs changes by western blot. The click reaction was performed on each sample. and Methanol/Chloroform precipitation. Then NSPs were enriched using neutravidin beads and eluted from the resin by boiling in 50ul 4×sample buffer and 2.5ul 20×reducing reagent. Protein were resolved by 4–12% gradient SDS-PAGE and transferred onto PVDF membranes using a semi-dry blotting apparatus. After blocking with 5% milk powder in PBS with 0.05% Tween 20, the membranes were incubated overnight at 4 °C with primary antibodies and then with an HRP-conjugated secondary antibody. Protein bands were visualized by chemiluminescence.

Figure 7.

Comparison of protein synthesis and total abundance changes. A: Scatter plot of individual protein synthesis ratios versus abundance ratios in APP/PS1 AD mouse model versus age matched normal control. P value refers to Pearson correlation coefficient. B: The independence of changes in protein synthesis and abundance is consistent with the notion that both synthesis and degradation contribute to the overall protein pool size.