Absolute Quantification of Molecular Representation and Interaction

分子表示和相互作用的绝对定量

• 批准号: 9354862
• 负责人: Aleksandra Nita-Lazar
• 金额: $ 44.87万
• 依托单位: NATIONAL INSTITUTE OF ALLERGY AND INFECTIOUS DISEASES
• 依托单位国家: 美国
• 资助国家: 美国
• 起止时间: 至
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/9354862
• 关键词: ATP Synthesis Pathway; Affinity; Agreement; Amino Acids; Behavior; Binding; Biological; Biological Assay; Catabolism; Cell Culture Techniques; Cell Cycle Regulation; Cell Line; Cells; Chemotaxis; Collaborations; Computer Simulation; Coupled; Data; Development; Experimental Models; Exposure to; Freedom; Gene Expression; Geometry; Immune system; In Vitro; Integrins; Label; Length; Ligands; Liquid Chromatography; Literature; Mammalian Cell; Mass Spectrum Analysis; Measurement; Measures; Mediating; Messenger RNA; Methodology; Methods; Mitochondria; Mitochondrial Proteins; Modeling; Modification; Molecular; Monitor; Mouse Protein; Mus; Mutation; Organelles; Osteoclasts; Osteoporosis; Pathway interactions; Peptides; Phosphorylated Peptide; Post-Translational Protein Processing; Process; Production; Protein Biosynthesis; Proteins; Proteome; Proteomics; PubMed; RNA; Reaction; Reporting; Rheumatoid Arthritis; Sampling; Set protein; Shotguns; Signal Pathway; Signal Transduction; Signaling Molecule; Site; Sphingosine-1-Phosphate Receptor; Stable Isotope Labeling; Stimulus; System; TLR4 gene; TRANCE protein; Testing; Time; Transcript; base; bone; cathepsin K; cell type; chemokine; deamidation; design; differential expression; insight; mRNA Expression; macrophage; mathematical model; monocyte; nano; network models; oxidation; pathogen; predictive modeling; protein expression; receptor; response; simulation; sphingosine 1-phosphate; success; synthetic peptide; text searching; transcriptome sequencing; transcriptomics; vacuolar H+-ATPase

We have adapted the methods for absolute quantification based on the targeted proteomics combined with the use of peptide standards and used the data for robust predictive modeling of the signaling pathways in the immune system. We used osteoclast development from macrophages as the initial experimental model. Understanding the mechanisms of osteoclast formation and action is crucial for progress in studies of rheumatoid arthritis and osteoporosis. We used the well characterized murine monocyte-macrophage RAW 264.7 cell line as the osteoclast precursor model cell line. The cells fuse to form multinucleated osteoclasts when stimulated with receptor activator of nuclear factor kappa B ligand (RANKL), but the differentiation process is inhibited by sphingosine-1 -phosphate (S1P). The mRNA levels of many proteins change and we wanted to see if these changes are reflected in changes of the cell proteome. We have optimized cell culture conditions and methods for osteoclast enrichment. Using SILAC (stable isotope labeling with amino acids in cell culture) we compared the proteomes of untreated RAW 264.7 cells, intermediate osteoclasts and differentiated, multinucleated osteoclasts. The analysis revealed a set of differentially expressed proteins, which we used to design a set of standard peptides for absolute quantification by mass spectrometry. We have also performed mRNA expression analysis using microarrays and identified major differences between all three cell types. Specifically, we found that compared to osteoclast precursors, multinucleated osteoclasts conserve energy by down-regulating pathways involved in cell cycle control, gene expression, and protein synthesis. In agreement with previous reports, multinucleated osteoclasts were found to express relatively high levels of V-ATPase, TRAP, cathepsin K, and integrins. Proteins involved in ATP synthesis and catabolism, localized primarily in the mitochondria, were also upregulated in multinucleated osteoclasts, suggesting that osteoclasts up-regulate ATP production compared with osteoclast precursors and intermediate osteoclasts. We have confirmed that both mitochondrial mass and potential are elevated in mature osteoclasts, and median mitochondrial protein expression was significantly higher than the median protein expression in other organelles (1). S1P regulates the chemoattraction and chemorepulsion of osteoclast precursors to and from bones. The murine macrophage RAW 264.7 cells, used here as a model, express two receptors for S1P: S1PR1 and S1PR2. These receptors have markedly different affinity to S1P and cause the opposite effects upon exposure to low/high concentrations of S1P. To develop a deeper understanding of mammalian cell chemotaxis, we used transcriptomics, shotgun proteomics, targeted proteomics, and pathway simulation to investigate S1P-mediated chemotaxis of osteoclast precursors. Transcriptomics using RNA-seq enabled the identification and quantitation of RNA transcripts and shotgun proteomics enabled the identification of proteotypic peptides. Selection of the target peptides used a wide variety of criteria including peptide proteotypic qualities, sequence uniqueness, and vulnerability to modification (e.g., oxidation and deamidation), eliminating many theoretically possible peptides, which could be non-compatible with mass spectrometric analysis. We used the quantitative data obtained from osteoclast precursors by shotgun proteomics to find the peptides amenable to analysis in our Orbitrap Velos. SPOT synthesis was used to prepare a set of 409 standard, synthetic peptides, which we used to assess the protein expressions in macrophages. Single Reaction Monitoring (SRM) of RAW264.7 cell lysates spiked with the standard peptides resulted in the confident identification and semi-quantitation of 208 of the 409 peptide targets from proteins in the chemokine signaling network. The SRM analysisof a smaller set of 65 heavy-labeled, quantitated internal peptide standards from proteins differentially expressed under different experimental conditions provided absolute numbers of molecules. These data were then used to design targeted proteomics assays of the proteins of the mouse chemotaxis pathway. Targeted proteomics assays using nano-flow liquid chromatography coupled to selected reaction monitoring mass spectrometry (LC-SRM) were performed to produce protein absolute abundance values (in units of copies/cell) for each of the target proteins within RAW 264.7 cells. RAW cells were again used as model osteoclast precusors because they have very similar S1P-directed chemotaxis behavior. Rules-based pathway modeling enabled the simulation of the mouse chemotaxis pathway based on bi-molecular interactions within the geometry of a three-dimensional in silico RAW cell. Measured protein abundance values, used as simulation input parameters, led to in silico pathway behavior matching in vitro measurements. Moreover, once model parameters were established, even simulated responses towards stimuli that were not used for parameterization were consistent with experimental findings. These findings demonstrated the feasibility and value of combining targeted mass spectrometry with pathway modeling for advancing biological insight and defined our experimental approach to modeling other immune system signaling pathways (2, 3). In the TLR signaling network modeling study, we utilize targeted proteomics with transcriptomics to aid in contructing a computational model of the LPS-TLR4 signaling pathway in a mouse monocyte-macrophage cell line RAW264.7. A set of protein targets was identified from a review of current literature and KEGG pathways describing LPS-TLR4 signaling. Corresponding peptides were selected after scoring based on several criteria including length, shotgun proteomics identification, and potential PTM sites as determined by literature mining by motif prediction (Pubmed). Peptides were analyzed in both shotgun-mode and SRM-mode to determine the potential for success in biological samples. RAW cell samples stimulated with LPS for different times were analyzed for the selected peptides. We performed semi-quantitative analysis with the external peptide standards and obtained proteotypic peptides for most of the proteins in the canonical TLR signaling network. Based on these results, we have ordered heavy-labeled internal peptide standards against corresponding protein targets for absolute quantitation measurements. Additionally, we designed and obtained peptides phosphorylated at the crucial regulatory residues of the proteins in the TLR signaling network. In collaboration with Drs. Martin Meier-Schellersheim and Bastian Angermann, we have created the network of essential proteins and their interactions for the Simmune-based model and began modeling the network changes following TLR stimulation with LPS. The model will incorporate also the measurements of PTM changes obtained form project AI001084-07 (Protein Modifications Involved in Cell Signaling). In this projest, we will be able to reach beyond basal level quantification to further develop and test the TLR signaling network model under a variety of biologically relevant perturbations (different TLR ligands, whole pathogens, and cells with mutations in specific signaling molecules). 1. An E, Narayanan M, Manes NP, and Nita-Lazar A. (2014) Mol Cell Proteomics 2014 Oct;13(10):2687-704. doi: 10.1074/mcp.M113.034371 2. Manes NP, Angermann BR, Koppenol-Raab M, An E, Sjoelund VH, Sun J, Ishii M, Germain RN, Meier-Schellersheim M, and Nita-Lazar A. (2015) Mol Cell Proteomics. 2015 Oct;14(10):2661-81. doi: 10.1074/mcp.M115.048918. 3. Manes NP, Mann JM, and Nita-Lazar A. (2015) J Vis Exp 102, doi: 10.3791/52959

我们根据目标蛋白质组学并结合肽标准品的使用，调整了绝对定量方法，并使用这些数据对免疫系统中的信号通路进行稳健的预测建模。 我们使用巨噬细胞的破骨细胞发育作为初始实验模型。了解破骨细胞的形成和作用机制对于类风湿关节炎和骨质疏松症研究的进展至关重要。我们使用已充分表征的鼠单核巨噬细胞 RAW 264.7 细胞系作为破骨细胞前体模型细胞系。当用核因子 kappa B 配体受体激活剂 (RANKL) 刺激时，细胞融合形成多核破骨细胞，但分化过程受到 1-磷酸鞘氨醇 (S1P) 的抑制。许多蛋白质的 mRNA 水平发生变化，我们想看看这些变化是否反映在细胞蛋白质组的变化中。我们优化了破骨细胞富集的细胞培养条件和方法。使用 SILAC（细胞培养物中氨基酸的稳定同位素标记），我们比较了未经处理的 RAW 264.7 细胞、中间破骨细胞和分化的多核破骨细胞的蛋白质组。分析揭示了一组差异表达的蛋白质，我们用它们设计了一组标准肽，用于通过质谱法进行绝对定量。我们还使用微阵列进行 mRNA 表达分析，并确定了所有三种细胞类型之间的主要差异。具体来说，我们发现与破骨细胞前体相比，多核破骨细胞通过下调涉及细胞周期控制、基因表达和蛋白质合成的途径来节省能量。与之前的报道一致，发现多核破骨细胞表达相对较高水平的 V-ATP 酶、TRAP、组织蛋白酶 K 和整合素。参与 ATP 合成和分解代谢的蛋白质主要位于线粒体中，在多核破骨细胞中也上调，表明与破骨细胞前体和中间破骨细胞相比，破骨细胞上调 ATP 产生。我们已经证实，成熟破骨细胞中的线粒体质量和电位均升高，并且中位线粒体蛋白表达显着高于其他细胞器中的中位蛋白表达 (1)。 S1P 调节破骨细胞前体进出骨骼的化学吸引和化学排斥。此处用作模型的鼠巨噬细胞 RAW 264.7 细胞表达两种 S1P 受体：S1PR1 和 S1PR2。这些受体对 S1P 的亲和力明显不同，并且在暴露于低/高浓度的 S1P 时会产生相反的效果。为了更深入地了解哺乳动物细胞趋化性，我们使用转录组学、鸟枪蛋白质组学、靶向蛋白质组学和途径模拟来研究 S1P 介导的破骨细胞前体趋化性。使用 RNA-seq 的转录组学能够识别和定量 RNA 转录本，而鸟枪法蛋白质组学能够识别蛋白型肽。目标肽的选择使用了多种标准，包括肽的蛋白型质量、序列独特性和修饰的脆弱性（例如氧化和脱酰胺），消除了许多理论上可能的肽，这些肽可能与质谱分析不兼容。我们利用鸟枪式蛋白质组学从破骨细胞前体获得的定量数据来寻找适合在 Orbitrap Velos 中分析的肽。 SPOT 合成用于制备一组 409 个标准合成肽，我们用它们来评估巨噬细胞中的蛋白质表达。对掺有标准肽的 RAW264.7 细胞裂解物进行单反应监测 (SRM)，可对趋化因子信号网络中蛋白质的 409 个肽靶点中的 208 个进行可靠的鉴定和半定量。对来自不同实验条件下差异表达的蛋白质的较小组 65 个重标记定量内肽标准品进行 SRM 分析，提供了分子的绝对数量。然后将这些数据用于设计小鼠趋化途径蛋白质的靶向蛋白质组学测定。 使用纳流液相色谱结合选择反应监测质谱 (LC-SRM) 进行靶向蛋白质组学测定，以产生 RAW 264.7 细胞内每种靶蛋白的蛋白质绝对丰度值（以拷贝/细胞为单位）。 RAW 细胞再次被用作模型破骨细胞前体，因为它们具有非常相似的 S1P 定向趋化行为。基于规则的通路建模能够基于三维计算机原始细胞几何结构内的双分子相互作用模拟小鼠趋化通路。测量的蛋白质丰度值用作模拟输入参数，导致计算机模拟途径行为与体外测量结果相匹配。此外，一旦建立了模型参数，即使对未用于参数化的刺激的模拟反应也与实验结果一致。这些发现证明了将靶向质谱与通路建模相结合以推进生物学洞察的可行性和价值，并定义了我们模拟其他免疫系统信号通路的实验方法 (2, 3)。 在 TLR 信号网络建模研究中，我们利用靶向蛋白质组学和转录组学来帮助构建小鼠单核巨噬细胞系 RAW264.7 中 LPS-TLR4 信号通路的计算模型。通过对当前文献和描述 LPS-TLR4 信号传导的 KEGG 通路的回顾，确定了一组蛋白质靶标。根据几个标准评分后选择相应的肽，这些标准包括长度、鸟枪法蛋白质组学鉴定以及通过基序预测（Pubmed）进行文献挖掘确定的潜在 PTM 位点。在鸟枪模式和 SRM 模式下对肽进行分析，以确定在生物样品中成功的潜力。分析用 LPS 刺激不同时间的原始细胞样品中选定的肽。我们使用外部肽标准品进行了半定量分析，并获得了规范 TLR 信号网络中大多数蛋白质的蛋白肽。根据这些结果，我们订购了针对相应蛋白质靶标的重标记内肽标准品，以进行绝对定量测量。此外，我们设计并获得了在 TLR 信号网络中蛋白质的关键调节残基处磷酸化的肽。与博士合作。 Martin Meier-Schellersheim 和 Bastian Angermann，我们为基于 Simmune 的模型创建了必需蛋白质网络及其相互作用，并开始对 LPS 刺激 TLR 后的网络变化进行建模。 该模型还将纳入从项目 AI001084-07（细胞信号转导中涉及的蛋白质修饰）获得的 PTM 变化的测量结果。在这个项目中，我们将能够超越基础水平的定量，进一步开发和测试各种生物学相关扰动（不同的TLR配体、整个病原体和特定信号分子突变的细胞）下的TLR信号网络模型。 1. An E、Narayanan M、Manes NP 和 Nita-Lazar A. (2014) Mol Cell Proteomics 2014 年 10 月；13(10):2687-704。号码：10.1074/mcp.M113.034371 2. Manes NP、Angermann BR、Koppenol-Raab M、An E、Sjoelund VH、Sun J、Ishii M、Germain RN、Meier-Schellersheim M 和 Nita-Lazar A. (2015) Mol 细胞蛋白质组学。 2015 年 10 月；14(10):2661-81。 doi：10.1074/mcp.M115.048918。 3.Manes NP、Mann JM 和 Nita-Lazar A. (2015) J Vis Exp 102，doi：10.3791/52959