高效能單管式醣質體末端表位鑑定分析平台之開發與應用

Abstract

細胞的醣質體包含醣脂質或醣蛋白的N型及O型醣鍊，雖其基礎結構有所不同，但在延伸的醣鍊末端，均可發現藉由唾液酸、岩藻糖、或硫酸機團等修飾而形成的末端結構－醣表位(glycotope)。這些末端醣表位藉由與特定蛋白受體的結合，參與並調控細胞與細胞、或細胞與外界的辨識、溝通、與黏附等生理現象，足以代表一個生物樣本(如細胞或組織)在特定的生理或病理狀態下的醣質體特色。傳統的醣質分析法著重於將醣蛋白的N型及O型醣鍊與醣脂質各自純化分離後獨立分析，其繁複的萃取及分離步驟，不僅降低了分析效率，以末端醣表位的角度而言，亦未能達到整體性的全面鑑定。 本研究所-開發的｢單管式醣質分析法｣，藉由省略傳統醣質分析的中間分離步驟，創新地將醣蛋白的N型與O型醣鍊製備成單管分析樣品，達成同時偵測N型及O型糖鍊的目標。質譜分析顯示，單管式製備的N/O型醣鍊混和物展現了與使用傳統法分離的N型及O型醣鍊的高度同質性，並無因單管式製備而遺漏了醣蛋白的特定醣類資訊，且單管式的同時性分析精準地呈現了醣質體樣品中N型及O型糖鍊的相對含量資訊，這是傳統分析法不易達成的。結合本研究團隊所開發的新穎醣質體末端表位液相層析串聯質譜分析平台與數據分析軟體，單管式醣質分析法精確且有效率地鑑定了胃癌及大腸結腸癌細胞中一系列被認為與癌症相關帶有唾液酸及岩藻糖修飾的Lewis血型末端醣表位結構。不僅如此，基於醣質體製備過程中醣脂質會與醣蛋白共沉澱的意外發現，本研究進一步成功地將醣脂質與醣蛋白分析結合，無論是製備同時含有醣脂質醣鍊、N型與O型醣鍊的單管式分析，或是由此法衍生的簡易醣脂質萃取與分析，均拓展了單管式醣質分析法的應用範圍。 應用方面，本研究以所開發的單管式醣質分析法有效解析I.異體移植之腫瘤與II.醣合成抑制劑處理後的細胞醣質體變化。在第一部份，與一般離體培養的癌症細胞相比，由胃癌及大腸結腸癌的異體移植腫瘤分離出的細胞，展現了各具特色的末端醣表位的變化，包含Lewis b/y或聚乳醣胺(poly-N-acetyllactosamine)的上升或下降。在醣合成抑制劑方面，本研究發現兩類N型醣鍊抑制劑swainsonine和2-deoxyglucose並非如預期單純抑制胃癌細胞 N型醣鍊的產生，尚會導誘生成新奇的醣類結構，包括具有聚乳醣胺延伸與多岩藻糖修飾的混和型 (hybrid type) N型醣鍊及攝入了脫氧六碳糖的多甘露醣型 (high mannose type) N型醣鍊。另外，本研究亦發現在細胞株培養中添加廣為人知的O型醣鍊抑制劑benzyl-α-GalNAc後，不單是造成特定O型醣鍊的減少，而是對醣蛋白整體、包含N型醣鏈上唾液酸化與岩藻糖化的系統性抑制，使細胞不正常地堆積未經修飾的N型與O型醣鏈。這些預料之外的醣質體改變多方驗證了我們需要一個全面且精準的醣質分析平台來有效率地鑑定細胞處於不同生物環境下醣質體的動態變化，也突顯了本論文提出的「單管式醣質分析法」之高度適用性與應用價值。 基於其高效化的設計，單管式醣質分析法能簡化醣質體的製備流程，亦縮短分析所需的時間，而提升了醣質分析的效率且維持傳統分析所能達成的分析深度及廣度。本研究僅以分析單管樣品，同時且全面地解析了醣脂質、N型及O型糖鍊，提供一個新穎的整體視角來研究細胞的醣質體。藉由與醣質體末端表位分析平台的整合，單管式醣質分析法具有高度潛力發展為高通量的醣質體分析技術，藉此能夠處理大量的臨床樣品。單管式醣質體末端表位鑑定分析平台藉由有效發掘隱藏在醣質體末端的訊息，預示了｢精準醣質體學｣在個人化醫療上的無限發展。

As a unique type of biomarker, the sugar codes represented by an assorted terminal sialyl-, fucosyl, and/or sulfated glyco-epitopes (glycotopes) collectively define the glycomic features of a cell or tissue type at a specific patho-physiological state and commonly act as essential players of glycan-mediated functions in a myriad of biological processes. Since the same glycotopes can be equally or differentially distributed over glycoproteins and glycosphingolipids (GSLs), a holistic view of their regulated expression necessitates a comprehensive profiling of N-, O-, and GSL-glycans. However, the conventional separate analysis of different glycan types that requires multiple extraction/fractionation steps often compromise the analytical throughputs of MS-based glycomic studies and appears redundant in a pursuit of mapping the glycotopes shared among glycoconjugates. One-fraction glycan analysis was developed in this study as an innovative glycomic approach featuring a sequential release into one single vial and the simultaneous detection of N-glycans and O-glycans by means of eliminating the intermediate fractionation steps that partition a glycan mixture to separate pools in regular MS-based glycomics. Analysis of the non-fractionated N- and O-glycans via MALDI-MS and nanoLC-ESI-MS exhibited a qualitatively unbiased representation and cell-specific characteristic portrayal of protein glycosylation. By interfacing with a glycotope-centric glycomic pipeline based on an Orbitrap MS-based RP C18 nanoLC-MS2-pd-MS3 analysis of permethylated glycans followed by GlyPick-aided data mining for sequence-informative MS2 oxonium ions and the linkage-specific MS3 ions from selected MS2 productive ions, the one-fraction glycan analysis was proven to a reliable semi-quantitative analytical strategy for a precision mapping of the (sialyl-) fucoyslated Lewis antigens comprising (Sialyl) Lewis a/x, Lewis b/y, and H type 1/2 in gastric and colorectal cancer cells. Furthermore, empowered by the unexpected co-extraction of GSLs with proteins upon protein precipitation, the protein-based one-fraction glycan analysis was extended to a GSL-included global glycomic characterization realized through either a direct one-fraction N-, O-, and GSL glycan analysis or a facile recovery of intact GSLs for parallel analysis. Next, the applications of one-fraction glycan analysis to tumor xenografts and glycosylation inhibitor treated cancer cells validated its capability to precisely delineate the glycomic alterations of cancer cells when exposed to characteristic cellular and biochemical factors. Comparing to the typical ex vivo cultured counterparts, the tumor-resident cancer cells isolated from gastric and colorectal tumor xenografts exhibited cell-dependent variable glycomic changes in multiple biologically relevant glycotopes including Lewis b/y and polylactosamine. On the other hand, two N-glycan inhibitors, swainsonine and 2-deoxyglucose, rather than acting as classical blockers of N-glycosylation, actively participated in N-glycan assembly in gastric cancer cells and brought about unusual N-glycans as the complex type-like hybrid type N-glycans with multi-fucosylated ploylactosamines, and the deoxyhexose-incorporated high mannose type N-glycans, respectively. Moreover, the common O-glycan inhibitor benzyl-α-GalNAc was demonstrated to strikingly modulate N-glycans by downregulating the terminal sialylation and fucosylation, leading to aberrant accumulation of unmodified multi-antennary N-glycans and simple core 1/2 O-glycans in multiple cell lines. The serendipitous discovery of glycomic alterations attested to the necessity of a glycome-wide analytical approach like one-fraction glycan analysis to dissect the dynamic cellular glycosylation in different biological contexts. Based on the streamlined design, one-fraction glycan analysis would increase analytical throughput via a simplified sample preparation and a shortened data acquisition time without compromising the attainable glycomic resolution and precision. Moreover, the unprecedented simultaneous one-shot analysis of one-fraction N-, O-, and GSL-glycans could directly unravel the relative abundance of glycans distributed over glycoproteins and glycosphingolipids, conferring a global perspective on cellular glycomes. Through integrating with the semi-quantitative glycotope mapping pipeline, the present one-fraction glycan analysis could be further evolved to an advanced high-throughput glycomic platform suited for efficiently mapping the glycotopes of biological significance and with capability to manage huge cohorts of clinical samples, priming the precision glycomics for a promise of optimized personalized medicine.