Functional 'omics' approaches to assess cell signaling upon endogenous molecule perturbation

Abstract

Functional ‘omics’ approaches especially integrative analyses could elucidate the property of entire interacting gene subset networks or pathways. These approaches rather simply overlapping information, but potentially generate valuable connections from functionally coherent gene modules [2, 3]. Furthermore, models of pairwise interactions between genes, gene products and biomolecules have been developed in recent studies. They were especially useful in systematic assembling of the complex diseases and healthy biological system compendia [4, 5]. Decent gene expressions, co-regulation profiles and organic or inorganic molecules involved signaling pathways determine biological function of a living organism. While destabilization of these co-regulation patterns has been indicated in multiple diseases and stress conditions [6, 7]. We could detect and characterize such destabilization by systems biology approaches to elucidate the initiation and propagation of specific diseases. β-hydroxybutyrate (3OHB) has been demonstrated to be one of endogenous neuroprotective agents. It has been extensively studied to elucidate potential mode of action in many neurodegenerative diseases models including AD, PD and ALS for decades. Combining quantitative proteomics and disease related protein networks approaches, we successfully identified H3K27me3 as a sensitive epigenetic regulator upon 3OHB perturbation. Alterations of both H3K4me3 and H3K27me3 were further detected in 3OHB treated normal neuron and fasted brain samples. We hypothesis that chromatin bivalent status was perturbed by 3OHB and determined epigenetic regulation profiles in normal neuron. The further integration of CHIP-seq and RNA-seq data set has revealed that 3OHB perturbed bivalent genes were majorly responsible for transcriptional regulation of neurodegeneration diseases related genes and neurodevelopment processes. We performed transcriptome analyses on neural stem cells (NSCs) upon 0.02 mM of 3OHB incorporation during neural differentiation process, which revealed that neural differentiation could be greatly affected by 3OHB through impairing the neural precursor cell differentiation and proliferation related biological processes. More interestingly, we have identified relative abundant histone lysine hydroxybutyrylation sites including H2AK118/119bhb and H2BK34bhb and detected both H2AK118bhb and H2AK119ub1 alteration upon 3OHB treatment. Our data reflected a novel scenario in which 3OHB could perturb chromatin bivalency and gene expression pattern through direct occupation of monoubiquitination sites. As far as we know, this is the first study to provide a systematic investigation of 3OHB perturbed neuron proteome. Nevertheless, ketone body induced chromatin bivalency fluctuation implicates possible scenario in which ketone dependent neuroprotective functions may progress with help of histone modification and chromatin structure alteration. Thus far, our results build the connection between ketone body, chromatin bivalent state, neurodegeneration diseases and neural differentiation. Although protein-protein interaction networks were used in our study to decipher possible hub proteins, all of those networks were built based on prediction. In order to decipher the real word of protein interaction network organization, we further conducted alternative proteomic approach to the study of in vivo protein tertiary structure and PPIs in living plant cells. Mass spectrometry analysis in combination with the site-specific chemical cross-linking has emerged as a powerful tool in study of three-dimensional structure of protein complex and in mapping of Protein-Protein Interactions (PPIs). As compared to the typical approaches of studying three-dimensional structure of proteins using X-ray and NMR, mass spectrometry analysis of the chemically cross-linked functional groups of proteins provides direct evidence for protein - protein interaction in protein complex. Even though in vitro cross-linking experiments have been widely applied to investigate the specific interactions of a bait protein and its targets, the measurement of in vivo PPIs has been extremely problematic and difficult due to the dynamic nature of the biological system and the lower number of cross-linking peptides that can be isolated via MudPIT (Multidimensional Protein Identification Technology). Using Arabidopsis thaliana as a model organism, we have attempted to develop an improved chemical cross-linking/ mass spectrometry-based workflow, which aims at optimizing the in vivo cross-linking conditions, establishing of a MudPIT cross-linking peptide (cl-peptide) enrichment procedure, and development of a software to identify cl-peptides. Thus far, we successfully identified several in vivo cross-linked (cl) peptides of high-confidence for functional characterization using an in-house developed software. This work has demarked a beginning of building an in vivo PPI-network mediating cellular events in a eukaryotic system and may be applied into other model organisms for medical research. The data and methodologies described in this thesis demonstrated technological advancement of mass spectrometry based proteomics approach for characterizing direct and indirect protein-protein in plant and mammalian cells, respectively. The integration of proteomics, transcriptomics, epigenomics for systems biology analysis highlighted the possible scenario of 3OHB dependent biological function perturbation. The combination of direct chemical cross-linking and proteomics featured the possibility of direct assessment of PPI events in eukaryotic systems. Therefore, extending traditional proteomics towards systems biology and interactomics provide solid basis for comprehensive exploration of molecular and cellular signaling networks in living organisms.

Date of Award	2017
Original language	English
Awarding Institution	The Hong Kong University of Science and Technology