利用磷酸化蛋白質體學探勘非小細胞肺癌之抗藥性標的

Abstract

表皮生長因子接受器-酪胺酸激酶抑制劑 (EGFR-TKIs) 已被證實作為EGFR突變的晚期非小細胞肺癌(NSCLC)患者的標靶治療的成功實例之一。舉例來說，吉非替尼(艾瑞莎)作為肺癌第一線用藥，對於EGFR的酪胺酸區域突變(19外顯子的缺失突變及21外顯子的L858R點突變)的患者，具有良好的療效。然而，這些使用艾瑞莎治療的患者終將會產生抗藥性以及腫瘤的復發；目前對發生EGFR-TKI主要的抗藥性機制有兩種說法，其一為EGFR在20外顯子出現了T790M 突變，此突變的產生干擾了TKI與EGFR的kinase domain之間的相互作用。另外一種抗藥機制則是，可能其他受體酪氨酸激酶（RTK）失調或下游分子的異常活化，或是其他逃脫TKI的訊息傳遞路徑的活化，其機制仍有待發掘。找到造成抗藥性的因子，有助我們進一步了解細胞產生抗藥性的調節機制，並且促使合併試劑的開發，希望能克服肺癌對於TKI產生的抗藥性並提供更好的療效。 磷酸化蛋白質體學已成熟成為一有利的工具，能夠系統性地揭示細胞中蛋白質的特定位點磷酸化改變。因此，為了研究在抗藥性細胞中獲得抗藥性的細胞調節機制，我們使用具藥物敏感性的PC9細胞和抗藥的PC9/gef細胞作為研究的模型。在本研究中，利用同重元素相對與絕對定量標定法(iTRAQ8)結合固定化鐵離子親合層析技術的磷酸化蛋白質體學定量分析策略，比較這兩株細胞分別以低劑量(0.03 µM)或高劑量(10µM)艾瑞莎刺激後的蛋白質磷酸化反應，並針對加藥後一天和三天進行分析。推測此方法可以藉著區分此兩株細胞之間不同的磷酸化表現量，鑑別出由其他激酶驅動的下游細胞訊息傳遞，但也可能鑑別出由TKI激活或未被TKI改變的蛋白質或途徑，希望找尋在抗藥性細胞裡，造成細胞對於艾瑞莎產生抗性的異常調節機制。在磷酸化蛋白質體學結果中，2509個蛋白中有3850條磷酸化胜肽被鑑定及定量。接著，將此複雜的蛋白質磷酸化定量結果進行叢集分析，將對艾瑞莎反應的變化分類出兩個變化趨勢。我們假設，在PC9艾瑞莎低劑量處理後蛋白質磷酸化沒有表現或表現被抑制，以及在PC9/gef要同時在低劑量以及高劑量加藥的條件下隨著天數，蛋白質磷酸化的表現量上升或是不變，即是可能造成抗藥性的關鍵蛋白。第一階段我們篩選出104條磷酸化胜肽(95蛋白質)作為抗藥性標靶。進一步將這些蛋白質以Ingenuity Pathway分析（IPA）知識數據庫以及DAVID進行功能分類GO (Gene Ontology)，利用生物資訊解析可能的異常調節網絡，並找出具有抗藥性潛力的蛋白質(ZYX、EIF4BP1、ATG4B、CTNND1、SRF2)。其中，已經有文獻指出EIF4BP1與肺癌抗藥性相關。除此之外，IPA與蛋白質的特定序列特徵資訊 (motif)分析結果預測TGFBR1與CK2可能為參與調控抗藥機制的上游調節因子。這個網絡也顯示ZYX可能受AKT、TGFBR1、CK2所調控，綜合上述推論，ZYX可能是造成PC9/gef抗藥性的關鍵蛋白。因此，我們更進一步驗證ZYX 絲氨酸(142和143) 磷酸化於EGFR-TKI的作用。在後續的生化實驗中證明了ZYX的磷酸化為TGFB1和AKT所調控，而且同時加入艾瑞莎更加抑制PC9/gef細胞生長及細胞遷移能力。將鑑定到的ZYX磷酸化胜肽進行motif分析後，結果預測酪蛋白激酶(CK2)為可能的激酶，PC9/gef細胞處理CK2抑制劑 ( CX-4945 )後，降低了ZYX 絲氨酸(142和143) 磷酸化表現量，與艾瑞莎合併處理亦降低PC9/gef細胞生長。雖然這三種藥物合併艾瑞莎的細胞實驗結果顯示，藥物合併治療可以進一步減少細胞生長，但卻不增加抗藥性細胞對於艾瑞莎的敏感度。因此，藉由轉染小分子干擾核糖核酸 (small interfering RNA, siRNA)，並與艾瑞莎協同使用，此實驗結果顯示降低ZYX 蛋白質的表現量可以增加艾瑞莎對於PC9/gef細胞的生長抑制能力。最後，透過將帶有ZYX點突變 ( S142/3A ) 質體在抗藥性細胞裡大量表現無法磷酸化的ZYX 蛋白，合併處理艾瑞莎後的實驗結果顯示，將此磷酸化位點突變後，可以增加艾瑞莎在抗藥性細胞中的療效。綜上所述，本研究透過系統性的磷酸化蛋白質體學分析，揭示非小細胞肺癌中與EGFR-TKI誘導的抗藥性相關的磷酸化蛋白體變化，且證明了TGFBR1、AKT、CK2調控了ZYX的磷酸化，協同造成抗藥細胞對於艾瑞莎的抗藥性，也證實ZYX為TKI抗藥性標靶。在非小細胞肺癌上的直接應用可提供更好肺癌診斷和治療之新方向。

Epidermal growth factor receptor-tyrosine kinase inhibitors (EGFR-TKIs) have been demonstrated as successful example of targeted therapy for advanced non-small cell lung cancer (NSCLC) patients with EGFR mutation. For example, Gefitinib (Iressa®) is an effective first-line treatment for NSCLC patients with EGFR exon 19 deletions (del19) and exon 21 substitution (L858R) mutations. However, gefitinib-treated patients eventually develop acquired resistance and recurrence with mechanisms yet to be discovered. Two mechanisms have been reported to involve in the resistant process, including acquired secondary mutation (T790M) in the exon 20 of EGFR to interrupt its interaction with TKIs The other has been proposed to be EGFR-independent manner due to dysregulation of other receptor tyrosine kinase or downstream molecules bypassing signaling activations. The discovery of resistance-related factors may advance our understanding for the underlying mechanism of resistance of EGFR-TKIs and to facilitate the design of a companion agents to overcome the resistance for better therapy efficacy. Advancement in phosphoproteomics has offered a powerful tool to reveal the system view of site-specific phosphorylation and their alterations in cellular proteome. To investigate the cellular regulatory mechanism contributing to the gefitinib acquired resistance in NSCLC, we use the TKI-sensitive PC9 cells and TKI-resistant PC9/gef cells as a cell model to quantitatively compare their dynamic phosphoproteomic alteration in response to Gefitinib. We performed an isobaric tags for relative and absolute quantitation (iTRAQ) labeling approach combined with Immobilized Metal Affinity Chromatography (IMAC) for phosphopeptides enrichment to quantitatively compare PC9 and PC9/gef, cells in response to gefitinib stimulation in dosage-dependent (0.03µM in both PC9 and PC9/gef, 10µM in PC9/gef) effects after 1 and 3 days. We hypothesized that this approach would identify the TKI-responsive signaling as well as downstream signaling events driven by other kinase that may potentially contribute to the drug resistance. The phosphoproteomics results identified and quantified a total of 3851 phosphopeptides from 2509 proteins. The time- and dose-dependent phosphoproteomic changes are further clustered into two dynamic trends in low-dose and high-dose treatment. Upon Gefitinib treatment, 104 phosphopeptides (95 phosphoproteins) were down-regulated upon low-dose gefitinib treatment in PC9 cells, whereas they were up-regulated in PC9/gef cells, which may be our hypothetical new drug-resistant targets. Based on functional annotation and pathway enrichment analysis, we systematically reconstructed an integrative network of these 104 phosphoproteins using DAVID and IPA database to identify potential candidate targets of TKI-resistance. The network potentially underlying the gefitinib perturbation phosphoproteome indicated that ZYX, EIF4EBP1, ATG4B, CTNND1, SRF2 may be related to drug resistance. Among them, EIF4EBP1 has been reported to confer EGFR TKI resistance in lung cancer. Moreover, IPA and motif analysis suggested that TGFBR1 and CK2 are the predicted major upstream regulators. The network also illustrated that ZYX may be regulated by AKT, TGFβR1 and CK2, as a critical node contributing to EGFR-TKI resistance in PC9/gef cells. We further validate the roles of ZYX and its phosphorylation site Serine 142 and 143 in EGFR-TKI resistance. Inhibition of TGFBR1 and AKT both reduced the Ser142/143 phosphorylation of ZYX and combined treatment with gefitinib decreased cell viability, and cell migration activity in PC9/gef cells. The motif analysis for the phosphopeptides of ZYX predicted that casein kinase 2 (CK2) is a putative kinase. Treatment of CK2 inhibitor (CX-4945) reduced Ser142/143 phosphorylation of ZYX, and combined treatment with gefitinib decreased cell viability in PC9/gef cells. The potential role of AKT, TGFBR1, and CK2 in recovery from EGFR-TKI resistance are excluded based on the result that their inhibition reduced cell growth but did not enhance the sensitivity of gefitinib. Furthermore, knocking down ZYX expression by transfection with small interfering RNA of ZYX enhanced the cytotoxicity and cell growth inhibition of gefitinib in PC9/gef cells. Finally, site-directed mutagenesis of ZYX Ser142/143 reinforced the efficacy of gefitinib in resistant NSCLC cells, suggesting the promising role of site-specific phosphorylation on ZYX to recover the EGFR-TKI sensitivity. In summary, our result suggests TGFβR1-ZYX, AKT-ZYX, and CK2-ZYX signaling axis that as a collaborative network regulated PC9/gef cells resistance to gefitinib. This study may shed light on the gefitinib-resistant mechanisms through systematic phosphoproteomic analysis and provide a novel landscape for improving therapeutic efficacy by combinational treatment in NSCLC.