探討急性骨髓性白血病細胞對cabozantinib抗藥機制─從生物資訊到實驗室分析

Abstract

急性骨髓性白血病(AML)的病人中，大約有20~30%帶有FLT3基因突變，其中以FLT3之內部串聯重複(FLT3-ITD)最為常見，且具較差的預後，因此近年來成為熱門的AML研究標的。雖然FLT3標靶藥物初期的治療成效可見，但是大多只能使病人得到暫時緩解，病人對治療產生抗性並復發的機率仍高。目前臨床上對於FLT3標靶藥物產生抗藥性的原因仍眾說紛紜，因此有必要持續探討其抗藥機制並發展新穎的治療策略。 在過去實驗室的研究，FLT3標靶藥物cabozantinib (CBZ)可以在低濃度下選擇性的減緩帶有FLT3-ITD之AML細胞株Molm13及MV4-11生長，並抑制FLT3下游的訊息傳遞；另外在小鼠動物實驗中也能有效抑制MV4-11以及Molm13皮下腫瘤的生長。為了探討AML細胞對CBZ抗藥性的相關機轉，實驗室也建立了對於CBZ具有抗藥性的細胞株Molm13-XR以及MV4-11-XR。 現今次世代定序技術發達，RNA-seq為熱門的轉錄體學研究工具，透過RNA-seq的分析，我們可以瞭解生物體內是否有新穎的點突變或融合基因的存在，也可以偵測大量基因的表達量，並結合許多現有的生物資訊工具，進行路徑詮釋或找尋具有治療潛力的小分子藥物。因此本篇研究利用生物資訊工具對兩株抗藥細胞Molm13-XR以及MV4-11-XR在點突變、融合基因以及基因表達量進行抗藥性的探討，並藉此尋找有潛力逆轉抗藥細胞基因表達的小分子藥物，也以細胞實驗以及斑馬魚動物實驗確認其效果、進行驗證。 本篇研究確認Molm13-XR與MV4-11-XR皆增加FLT3 D835Y點突變，且MV4-11-XR為同型合子；另外也初步在兩株抗藥細胞整理出共同出現的60個基因具有SNP以及20個基因具有小片段插入刪去(indel)。在融合基因分析上，我們雖未發現與抗藥性有關的融合基因，但特定融合基因的表達量在母株細胞及抗藥細胞的差異可能暗示著抗藥細胞產生時存在篩選的效應(clone selection)。在基因表達量分析上，利用KEGG及DAVID進行路徑詮釋後，結果顯示兩株細胞皆有JAK-STAT、GPCR等致癌訊息路徑的過度活化，以及在醣類、胺基酸或脂質代謝路徑的基因表達量上升，與實驗室先前研究發現抗藥細胞能量代謝特性的變化相互呼應。 透過分別比對兩株抗藥細胞株中代謝相關基因表達量的變化，我們利用Connectivity map找尋具有潛力改變細胞代謝特性而克服抗藥性的小分子藥物。在Molm13-XR中，我們發現化療藥物lomustine以及HSP90抑制劑radicicol與tanespimycin；以及在MV4-11-XR中有mTOR抑制劑rapamycin以及其他PI3K/mTOR抑制劑，皆具有逆轉細胞代謝特性以及克服抗藥性的潛力。實驗結果指出，lomustine、radicicol可以抑制細胞糖解相關基因的表達以及葡萄糖攝取的能力；而radicicol更能夠進一步抑制乳酸的生成，且可能透過抑制FLT3訊息傳遞路徑來調控HIF-1α、MYC及MCL-1，影響細胞的代謝與存活。另外，我們也使用斑馬魚異種移植模型來驗證前述的實驗，結果顯示合併radicicol與cabozantinib可以有效降低斑馬魚體內的AML細胞量。我們的實驗結果認為，利用生物資訊工具找尋小分子藥物用以逆轉細胞代謝特性，或為可行的治療策略來克服對FLT3-ITD AML細胞對CBZ之抗藥性。

Internal tandem duplication of FLT3 (FLT3-ITD) was found in about 20-30% in AML patients and resulting in poor prognosis. Hence, FLT3 have thought to be an ideal target for AML treatment. Although FLT3 targeted therapy showed modest promising effect in the initial stage, relapse and drug resistance occurred in most patients, which indicated the importance of exploring the resistant mechanisms and discovering novel treatment strategies. Previously, we showed that cabozantinib(CBZ), an oral multi-targeted tyrosine kinase inhibitor, could be selectively cytotoxic in AML cells with FLT3-ITD. However, drug resistance occurred after gradual escalating concentration of CBZ incubation of the FLT3-ITD-harboring Molm13 and MV4-11 cells, with IC50 increased from 1.06 nM of parental Molm13 cells to 473.36 nM of the resistant Molm13-XR cells, and from 9.5 nM of parental MV4-11 cells to 1500 nM of resistant MV4-11-XR. In this research, we used RNA-seq to discover SNVs, fusion genes and signaling pathways changed in cabozantinib resistant cells. Connectivity map was used to predict small molecules which had potentials to reverse gene expression in resistant cells. Also, glucose uptake and lactate production were measured to realize the metabolic alternations after drug treatments. Zebrafish xenograft experiments were performed to evaluate the drug efficacy in vivo. We found FLT3 D835Y mutants and other 60 SNVs and 20 indels occurred in both CBZ-resistant cell lines. Pathway analysis showed that metabolic alternation on glucose, amino acid and lipid, GPCR pathway and JAK-STAT pathway activation were common features of these two resistant cell lines. Connectivity map predicted that lomustine, HSP90 inhibitor radicicol, and tanespimycin had potentials to reverse the metabolic phenotypes in Molm13-XR cells; and also showed that PI3K/mTOR inhibitor rapamycin, gedatolisib, and omipalisib could reverse that on MV4-11-XR. Data showed that lomustine and radicicol could reduce glycolysis-related genes expression and glucose uptake; Radicicol could also inhibit lactate production and may regulate cell metabolism and survival through inhibition of FLT3 downstream signaling to decrease the protein level of HIF-1α, MYC, and MCL-1. Finally, a combination of radicicol and cabozantinib could effectivity reduce tumor burden in zebrafish xenograft model. In this study, we showed reverse metabolic gene expression and phenotype with small molecules predicted by bioinformatic tools could be a powerful way to discover novel compounds to overcome cabozantinib-resistance in FLT3-ITD leukemia cells.