開發磷酸化蛋白體質譜法應用於微量組織

Developing mass spectrometry-based phosphoproteomics for microscale tissue

謝宇衡(Yu-Heng Hsieh)

指導教授 ： 陳玉如

國立臺灣大學/生命科學院/基因體與系統生物學學位學程/碩士(2021年)

https://doi.org/10.6342/NTU202103270

摘要

蛋白磷酸化是重要轉譯後修飾之一，能調節體內細胞訊息傳遞及生物功能。先進的質譜與大規模胜肽分離技術已成為一種深具潛力涵括全基因體深度磷酸化蛋白體分析工具。然而，目前要獲得臨床檢體枝深度磷酸化蛋白體分析仍然需要相當大的起始量（>50毫克腫瘤組織及>200微克胜肽用於磷酸化胜肽萃取）以及耗時的樣品製備、質譜數據採集與資料處理。為了增進具全基因體深度之組織磷酸化蛋白體分析，我們開發一種高靈敏且流程精簡的磷酸化蛋白體樣品製備並應用於微量組織檢體（<10毫克組織），如抽吸針組織活檢及石蠟包埋組織切片等。
首先，我們使用新鮮老鼠肺臟組織來優化組織蛋白萃取，並在優化樣品處理過程以省略去除界面活性劑步驟減少樣品損失。我們選擇三種組織裂解液分別為月桂酸鈉、尿素和乙腈，並加入TCEP與CAA達到快速還原/烷基化。相較於尿素(9%)和乙腈(4%)蛋白萃取率，利用月桂酸鈉進行組織裂解可顯著提升蛋白萃取率達11%。此外，與用尿素與乙腈來裂解組織進行磷酸化蛋白體分析相比，使用月桂酸鈉顯著增加了磷酸化蛋白體鑑定深度達9,539條磷酸化胜肽（尿素：6,389條磷酸化胜肽；乙腈：2,248條磷酸化胜肽）。以5毫克到0.5毫克老鼠肺組織評估月桂酸鈉裂解方法的靈敏度，此方法可獲取>10%的蛋白萃取率，並證實微量組織利用月桂酸鈉組織裂解法可獲得較高蛋白萃取率。
第二個部分則是優化控制酸鹼度之固向化金屬親和層析磷酸化胜肽萃取法以精簡實驗流程。我們將6%醋酸(pH 3.0)樣本緩衝液代替為80% ACN, 0.1% TFA，進而省略磷酸化胜肽萃取中的樣本緩衝液置換步驟。我們利用30微克老鼠肺臟組織胜肽進行測試，使用80% ACN, 0.1% TFA能鑑定5,969條磷酸化胜肽，並將萃取專一性提高到99%且具有高再現性（Pearson 相關係數：0.912-0.949）。此外，去除樣本緩衝液置換步驟可以減少>4小時樣本製備時間。在0.5毫克組織中，此優化磷酸化樣品製備流程鑑定6,698條磷酸化胜肽和99%萃取專一性，有高再現性。綜上所述，我們優化後的磷酸化蛋白體樣本製備流程應用於微量組織樣品，具有高再現性，整體製備流程顯著縮短至1天內。
為了在微量組織中進行深度磷酸化蛋白體學分析，我們將single-shot非數據依賴擷取 (Data-Independent Acquisition，DIA)方法納入上述磷酸化蛋白體製備方法。我們以50微克老鼠肺臟組織胜肽評估DIA方法。利用DIA方法鑑定>32,000條磷酸化胜肽和14,000個class 1磷酸化位點。與DDA方法相比，DIA方法提升三倍磷酸化肽鑑定數和兩倍可定量之磷酸化位點。在DIA方法中，所有可定量磷位點的信號動態範圍超過6個數量級，而數據依賴擷取方法(Data-dependent acquisition, DDA)僅為4個級數，此結果證明DIA 方法具有更廣的動態範圍，可以檢測低含量的磷酸化胜肽。與DDA方法具43%資料缺失相比，DIA方法僅有12%資料缺失，能進一步提升多重樣品的定量再現性。我們同時也優化液相層析梯度，並用於微量組織磷酸化蛋白體學分析，加以提高樣品分析效率。使用120分鐘液相層析梯度，可鑑定超過30,000條磷酸化胜肽（超過14,000個class 1磷酸化位點），其中涵蓋可應用於非小細胞肺癌標靶治療之16個藥物靶點。
最後，我們將優化的磷酸化蛋白體學實驗流程和 DIA 方法應用於人類肺癌石蠟組織包埋切片(FFPE section)。在利用libDIA以及 directDIA 的結果中，鑑定出 >15,000條磷酸化胜肽，在非小細胞肺癌訊息傳遞路徑（16個磷酸化蛋白，71個磷酸位點）和人類激酶體（164個激酶）中具有高覆蓋率。與先前研究相比，我們的優化方法將磷酸化蛋白體學深度提升至>10,000條磷酸化胜肽，並鑑定與定量出6個肺癌藥物靶點。我們的結果證明深度磷酸化蛋白體學分析運用在石蠟組織包埋切片的實用性。總體而言，此樣品製備和分析流程具有高度再現性和靈敏度，可應用於微量組織進行深度磷酸化蛋白質體學分析。

關鍵字

微量組織 ； 磷酸化蛋白體學 ； 固定金屬離子親和性層析 ； 數據非依賴性採集質譜

並列摘要

Protein phosphorylation is one of the important post-translational modification (PTM) regulating cellular signaling transduction and biological functions. Advanced mass spectrometry coupled to extensive multiple peptide fractionation have demonstrated as a promising analysis tool towards genome-wide phosphoproteomics depth. However, in-depth phosphoproteomic profiling of clinical specimens still requires a rather large amount (>50mg tumor tissue, >200μg peptide for phosphopeptide enrichment) and days of sample preparation, data acquisition and processing. To facilitate genome-wide tissue phosphoproteomic analysis, we report a highly sensitive and streamlined phosphoproteomic sample preparation for low amount clinical specimens (<10mg tissue) compatible with needle biopsy and FFPE section.
In the first part of the thesis, we used fresh mouse lung tissue for optimization of the protein extraction and bypassed removing detergents during sample processing. Three lysis buffer sodium laurate (SL), urea and acetonitrile (ACN) were selected and incorporated into rapid reduction/alkylation by TCEP/CAA. The SL lysis shows significantly superior protein extraction yield of 11% compared to urea (9%) and ACN (4%). Using mouse lung tissue (5 mg amount) for LC-MSMS analysis, furthermore, the use of SL significantly increased phosphoproteome identification (9,539 phosphopeptides) compared to conventional detergents (urea: 6,389 phosphopeptides; ACN 2,248 phosphopeptides). The sensitivity of SL lysis method was evaluated from 5 mg to 0.5 mg mouse lung tissue and achieved >10% protein extraction yields. Here, we demonstrated the high efficiency of SL-based tissue lysis and protein extraction.
The second part focused on the optimization and development of a streamlined phosphopeptide enrichment protocol based on the pH control immobilized metal affinity chromatography (IMAC). The use of 80% ACN with 0.1% TFA sample loading buffer bypassed buffer reconstitution step in phosphopeptide enrichment, demonstrating identification of 5,969 phosphopeptides from 30μg mouse lung tissue peptide with enhanced enrichment to 99% specificity and high reproducibility (Pearson correlation: 0.912-0.949). Furthermore, bypass buffer reconstitution reduced >4 hours in phosphopeptide enrichment workflow. On the application of sub-microsale tissue, the optimal workflow identified 6,698 phosphopeptides and 99% enrichment efficiency in 0.5mg tissue with high reproducibility. Taken together, our approach significantly reduced the microscale tissue sample preparation to 1 day.
Toward in-depth phosphoproteomic profiling in microscale tissue, we incorporated the single shot data-independent acquisition (DIA) approach into our microscale tissue phosphoproteomics pipeline. The performance of the DIA method was evaluated from 50 μg mouse lung tissue peptide and identified >32,000 phosphopeptides and 14,000 class 1 phosphosites. The DIA method improved three fold phosphopeptide identification and two fold quantified phosphosites compared to the DDA method. In DIA method, the intensity dynamic range of all quantified phosphosites has crossed 6 orders of dynamic, and only 4 orders with DDA method. The DIA method has a wider dynamic range which can detect low abundance phosphopeptides. The DIA further enhanced quantitative reproducibility with 12% missing value compared to 43% missing value from the DDA method. We also optimized LC gradient for microscale tissue phosphoproteomic analysis to improve analytical efficiency. Using 120 min LC gradient identified over 30,000 phosphopeptides (over 14,000 class 1) with 16 druggable targets in lung cancer in non-small lung cancer cells.
Finally, we applied optimal phosphoproteomics workflow and DIA method to human lung cancer FFPE tissue section. We identified >15,000 phosphopeptides in library-based DIA and direct DIA reported with high coverage in non-small lung cancer pathway (16 phosphoproteins, 71 phosphosites) and human kinome (164 kinases). Compared to previous research, our method improved phosphoproteome identification toward 10,000 phosphopeptide numbers and quantified 6 lung cancer druggable targets. Our result demonstrated the utility of the FFPE section and achieved depth of phosphoproteomic profiling in cancer research. Overall, this sample preparation and analysis process is highly reproducible and sensitive and can be applied to microscale tissues for deep phosphoproteomics analysis.

並列關鍵字

Microscale tissue ； Phosphoproteomics ； Immobilized metal affinity chromatography (IMAC) ； Data independent acquisition (DIA) mass spectrometry

參考文獻

1. Sharma, K., et al., Ultradeep human phosphoproteome reveals a distinct regulatory nature of Tyr and Ser/Thr-based signaling. Cell Rep, 2014. 8(5): p. 1583-94.

2. Ardito, F., et al., The crucial role of protein phosphorylation in cell signaling and its use as targeted therapy (Review). Int J Mol Med, 2017. 40(2): p. 271-280.

3. Ubersax, J.A. and J.E. Ferrell Jr, Mechanisms of specificity in protein phosphorylation. Nature Reviews Molecular Cell Biology, 2007. 8(7): p. 530-541.

4. Harsha, H.C. and A. Pandey, Phosphoproteomics in cancer. Mol Oncol, 2010. 4(6): p. 482-95.

5. Lahiry, P., et al., Kinase mutations in human disease: interpreting genotype-phenotype relationships. Nat Rev Genet, 2010. 11(1): p. 60-74.