Quantitative Proteomics And Bioinformatics Combine To Study TLR Signaling Pathways And MyD88 Complexes

The innate immune response is the first line of defense against microbial infections. Since the discovery of the Drosophila protein Toll which induces effective immune responses to bacterial infection, accumulating evidence showed that the innate immune system specifically recognizes invading microorganisms. The targets of innate immune recognition are the conserved molecular patterns (pathogen-associated molecular patterns, PAMPs) of microorganisms. Receptors in innate immunity are therefore called pattern recognition receptors (PRRs). The specific reorganization of PAMPs by PRRs effectively modulates the innate immunity to defense the host from the infection of microorganisms. Toll-like receptors (TLRs) are one of the most important kinds of PRRs which are capable of sensing organisms ranging from bacteria to fungi, protozoa, and viruses. Up till now, about 13 TLRs are found not only play key roles in innate immune system but also an important bridge to link innate immune system and adaptive immune system. Thus, the investigation of TLR signaling pathway has tremendous meaning to provide information for researchers to know better about the immunology. Previous studies about TLR were performed through molecular and cellular biology approaches, but at protein level there is still lack of comprehensive dataset to provide the global view of TLR signaling pathway changes. This thesis systematically investigates the protein changes involved in TLR signaling pathway through the combination of quantitative proteomics and bioinformatics.In Chapter 1, recent research progresses on quantitative proteomics were introduced and summarized. The application of AACT/SILAC based quantitative proteomics approaches was emphasized. The main traditional protein protein interaction detection methods and powerful bioinformatics tools based on proteomics analysis were subsequently reviewed. The advantage and weakness of each method were also included. Based on these useful technologies, the purpose and significance of our study were introduced at last. This dissertation was composed of three main research parts. The first part reveals multiple pathway cross-talk that coordinates specific signaling and transcriptional regulation on macrophages for the early host response to LPS. The second part investigates endogenous MyD88 interacting proteins of TLR2 and TLR4 signaling pathway by AACT quantitative proteomics and agarose beads immunoprecipitation. The third part discuss the combination of magnetic bead immnuoprecipitaion, quantitative proteomics and bioinformatics, revealing LPS-induced interactions with endogenous MyD88 based on protein domain analysis to explore its diverse functions in innate immunity.(1) Subcellular quantitative proteomics and bioinformatics reveals multiple pathway cross-talk that coordinates specific signaling and transcriptional regulation for the early host response to LPSLipopolysaccharide (LPS), a major pro-inflammatory component of the gram-negative bacterial cell wall, is one of the agents very commonly present as contaminant on airborne particles, including air pollution, organic dusts, and cigarette smoke. Acting at the first line of defense of innate immunity, toll-like receptor 4 (TLR4) specifically recognize LPS to initiate and mediate the signal transduction for modulating the host inflammatory responses. Although growing numbers of genes and proteins have been characterized one-gene-at-a-time for their functional involvements in TLR4 signaling, the global details about the LPS-induced TLR4-mediated signaling and regulatory pathways/networks remain largely incomplete. Given the fact that most of signaling pathways are activated at the early phase of LPS-induced cellular response, in the living LPS stimulated macrophages we employed subcellular SILAC-based quantitative proteomics approach to identify those proteins showing cytosol- or nuclei-specific changes in their abundances. Subcellular fractionation not only reduces the spectral complexity for identifying maximum numbers of proteins but also enriches certain low-abundance proteins at their corresponding compartments where they are functional.As a result, following 10 min LPS stimulation,508 proteins were found up-regulated and 103 proteins down-regulated in the cytosol, while 678 proteins were up-regulated and 80 proteins down-regulated in the nuclei, respectively. In coordination with the observations that many key proteins involving the signal relays in the MAPK and NF-κB cascades were found simultaneously regulated in the cytosol, various transcriptional factors (TFs) such as IRFs were found activated in the nuclei. We extended links between these intracellular signaling and apoptotic pathways were also found. For the first time, our combined datasets from quantitative proteomics and bioinformatics analysis further provided a direct and systems insight into how the upstream signaling pathways in cross-talks are modulating the activity of defined TFs for regulating particular sets of pro-inflammatory gene expression which involved in a variety of biological processes and pathways including protein and nucleic acid metabolism, apoptosis, DNA damage recognition and repair, cell cycle control, and host cell defense. The cross talk among multiple signaling cascades represents critical points of convergence for TLR4 signaling for modulating the activity of multiple transcriptional factors where pharmacological targets for therapeutic intervention could be located systematically.(2) Primary investigation on endogenous MyD88 interacting proteins of TLR2 and TLR4 signaling pathway by AACT quantitative proteomics and agarose beads immunoprecipitationUp till now,13 TLRs have been found in mouse. Although different TLRs recognize different PAMPs/ligands, myeloid differentiation primary response gene 88 (MyD88) serves as an key adaptor protein of all TLRs except TLR3 to transduce the signal and activate the downstream signal cascades. MyD88 can recruit the downstream signal molecule to transduce the signal through its Death Domain and trigger a series of pathways which cross talk together to control the immune response. Thus, identification of interacting partners of MyD88 from macrophages will shed much light to our understanding of how the host controls the specificity in the TLRs-mediated signaling pathways.A lot of technologies have been developed to combine with mass spectrometry to study on protein interacting in large scale, such as tandem affinity purification (TAP technology) and in vivo-dual tagging etc. But all these technologies involve constructing target proteins with tag and transfect them to cell, resulting in changes of physiological environment of cells to some extent. Moreover, the transfected target protein may compete with endogenous bait protein when recruit their interacting proteins. Therefore, pull down the endogenous bait protein directly by antibody is the best method to analyze protein protein interacting in native environment. However, one main drawback of the typical immunoprecipitation is the high background when the antibody pulls down the complex from cell, thus will generate a lot of false-positive interacting proteins. Here, we combine the typical immunoprecipitation with AACT/SILAC technology which can distinguish those specific interactors form background according to the quantification information.The TLR4 agonist LPS and TLR2 agonist Zymosan were selected to stimulate the macrophages. Using this combined powerful method, we identified 82 specific MyD88 interacting proteins after LPS stimulation and 56 specific MyD88 complex after Zymosan stimulation. Through the bioinformatics tools, the possible roles of MyD88 complex in regulating TLR2 and TLR4 signaling pathway have been discussed.(3) Combination of magnetic bead immnuoprecipitaion, quantitative proteomics and bioinformatics reveal LPS-induced interactions with endogenous MyD88 based on protein domain analysis to explore its diverse functions in innate immunityAlthough the method of agarose beads immunoprecipitation mentioned in Chaper 2 is powerful to study endogenous protein interaction, there are too many steps involved in this method which lead to generate more errors and the lost of weak interacting proteins. Here, in stead of agarose beads, we use magnetic beads to do immunoprecipitation which simplified the experiment steps and thus reduce the cell number needed for the experiment. Combined this improved immunoprecipitation and AACT/SILAC quantitative proteomics, we distinguished 161 endogenous MyD88 interacting proteins after LPS stimulate macrophages. Bioinformatics play a key role in predicting and validating the protein protein interacting now. One protein interacts with another one through its domain in stead of the whole protein sequence. Some database about domain domain interacting have been established based on the information of large scale known protein protein interacting, thus the domain interacting can validate the newly found protein interacting to some extend.Here, through the domain mapping analysis, we found proteins which include several domains have more chance to be recruited by MyD88. These proteins involved in diverse biological processes, indicating MyD88 would recruit multifunctional proteins to mediate the TLR signaling pathway together. Based on the domain domain interacting relationship of our data, we construct a protein interacting network to show the predicted connection among the proteins we identified. The proteins recruited by MyD88 can be divided into three layers according to their distance to MyD88. The fist layer of indentified interacting proteins means they have chance to interact with MyD88 directly. The proteins in this layer have main function like kinase activity and receptors which can active the upstream signal cascades to transfer the signal form MyD88 to the second layer interacting proteins. The second layer of interacting proteins will then activate and regulate a lot of different biological processes and pass the signal to the third layers which control nucleotide and nucleic acid metabolic process and gene expression. This predicted protein interacting network help us know more clearly about the whole immune response cascades about how the signal transduced form MyD88 to its direct and indirect interacting proteins. The method we used here can also be applied to investigate other bait proteins.