Proteomic changes in the hippocampus of large mammals after total-body low dose radiation

Abstract

There is a growing interest in low dose radiation (LDR) to counteract neurodegeneration. However, LDR effects on normal brain have not been completely explored yet. Recent analyses showed that LDR exposure to normal brain tissue causes expression level changes of different proteins including neurodegeneration-associated proteins. We assessed the proteomic changes occurring in radiated vs. sham normal swine brains. Due to its involvement in various neurodegenerative processes, including those associated with cognitive changes after high dose radiation exposure, we focused on the hippocampus first. We observed significant proteomic changes in the hippocampus of radiated vs. sham swine after LDR (1.79Gy). Mass spectrometry results showed 190 up-regulated and 120 down-regulated proteins after LDR. Western blotting analyses confirmed increased levels of TPM1, TPM4, PCP4 and NPY (all proteins decreased in various neurodegenerative processes, with NPY and PCP4 known to be neuroprotective) in radiated vs. sham swine. These data support the use of LDR as a potential beneficial tool to interfere with neurodegenerative processes and perhaps other brain-related disorders, including behavioral disorders.

Fig 1. Differential clustering of RAD vs. SH swine.

(A) PCA of proteomic profiles shows distinct differential clustering of RAD (green) vs. SH (blue) hippocampal protein expression. (B) Heat map of expression pattern clustering, where green indicates low and red indicates high differential protein abundance. Hierarchal clustering of proteins reveals group-specific abundance trends (see brackets on heat map). Z-score transformation of normalized protein abundances from a quantitative proteomics analysis using isobaric mass tags was applied before performing the hierarchical clustering based on Euclidean distance and complete (furthest neighbors) linkage. The horizontal dendogram shows the proteins in samples that clustered together. RAD n = 9, SH n = 6.

Fig 2. Hippocampal proteomic profiling of RAD vs. SH swine.

(A) Volcano plot of log2(FC) vs -log10(p-value) of hippocampal proteomic profiles, with up-regulated proteins on the right (red) and down-regulated proteins on the left (green). Significance threshold of p<0.05 is indicated by shading, with all significantly changed proteins with log2(FC)≥0.26 included in our analysis shown in the corresponding color shaded boxes (p-value of per group ratio calculated by t-test; fold changes visualized as log2 of abundance ratio). We identified 190 up-regulated (red square) and 120 down-regulated (green square) proteins within these criteria through the proteomic profiling. Location of target proteins selected for WB verification are indicated by blue dots on the volcano plot (A), and the STRING interaction network of these proteins is illustrated (B).

Fig 3. TPM1, TPM4, PCP4 and NPY are identified as having significantly increased abundance in the hippocampus after LDR exposure.

(A, B) Western blot verifies the significantly increased abundance in the swine hippocampus in RAD vs SH groups (p<0.05; unpaired Student’s t-test, one-tailed; RAD n = 9 (n = 8 for PCP4), SH n = 6) initially identified through proteomic analysis. STRING software analyses of the top 50 (C) protein interactions with the eight WB targets identify potential mechanisms of involvement in the brain’s response to LDR.

Fig 4. Fitting selected WB targets into the broader network of neurodegenerative protein interactions.

(A) STRING analysis of protein interactions following input of eight target proteins and 18 previously investigated proteins involved in neurodegeneration after LDR (MAPT, SNCA, PARK2, LRRK2, APP, TUBB, TH, SYP, CASP3, NEFL, GFAP, GAP43, DLG4, PPP1R9B, TP53, CHEK2, H3F3A and POLB) (22), demonstrating possible interactions of our target proteins within this wider neurodegenerative network. The STRING interaction network with 50 closest interactors is also visualized (B).

Fig 5. DAVID analyses of differentially abundant hippocampal proteins identified through proteomic analysis of RAD vs. SH swine.

Following input of a condensed list of 310 proteins identified through proteomic analysis as having differential abundance in the hippocampus after LDR, DAVID enrichment analysis of GO relationships revealed 41 molecular functions (A), 41 cellular components (B) and 68 biological processes (C) as potentially significantly affected.