编者按：干细胞基础研究、再生医学、细胞治疗以及临床研究指南的应用，是当前全球血液学与转化医学领域最受瞩目的前沿方向。以此为核心议题，第十届国际干细胞论坛（IFSC）隆重召开，汇集了来自世界各地的顶尖专家学者，共同探讨科学发现与技术创新，集中展示造血干细胞基础、临床及转化研究的最新进展，推动全球干细胞领域协同创新发展。会议期间，《肿瘤瞭望-血液时讯》特邀日本大阪大学Daichi Inoue教授，围绕“预后极差白血病的克隆演化与靶向干预”分享其最新研究成果与思考。现将精粹整理成文，以飨读者。

《肿瘤瞭望-血液时讯》：您的讲题聚焦预后极差白血病的克隆演化。通过您的新型小鼠模型，您发现了哪些驱动高风险急性髓系白血病（AML）演进的关键节点？这对理解其耐药机制有何帮助？

Daichi Inoue教授：在本项研究中，本团队重点关注了SETBP1突变，尤其是其发生在ASXL1突变克隆背景下的情形。该领域存在一个普遍的疑问：为何这两种突变会优先同时出现，且总是ASXL1先发生突变，而SETBP1作为后期事件发生？为阐明这种优先顺序与组合的机制，本团队通过广泛的动物模型与人类患者分析发现，核心机制在于SETBP1蛋白本身的丰度。ASXL1突变引起了表观遗传启动（epigenetic priming），从而增加SETBP1蛋白的丰度；随后SETBP1突变导致其逃避泛素蛋白酶体途径的降解。这两个事件对于维持足够水平的SETBP1蛋白均是必需的。这意味着此类预后不良白血病以SETBP1丰度依赖的方式发生。

最终本团队发现，依赖于SETBP1蛋白的丰度，其会形成核凝聚体（转录凝聚体形成），以维持高水平白血病转录程序，如HOXA9、HOXA10、MYB及MECOM等。因此，本报告阐述了为何这两种突变均为必需，以及为何这一特定顺序对这类预后不良白血病至关重要。

Oncology Frontier-Hematology Frontier: Your presentation focuses on the clonal evolution of leukemia with dismal clinical outcomes. Using your novel mouse models, what key nodes driving the evolution of high-risk acute myeloid leukemia (AML) have you discovered? How does this help understand its resistance mechanisms?

Professor Daichi Inoue: Right, thank you very much for raising the great point. And in my presentation I focused on the SETBP1 mutation, especially occurring in the context of the ASXL1 mutated clones. So basically there is a general question why these two mutations preferentially occur and always ASXL1 occur first and SETBP1 occur in the later event. Why is the preferential order and why is the preferential combination? So to answer this question, through the extensive animal model and human patient analysis we found that the biggest problem is the abundance of the SETBP1 protein itself. So ASXL1 mutation causes the priming, epigenetic priming to enhance the SETBP1 protein abundance. And then SETBP1 mutation causes escape from the degradation of the ubiquitin proteasome pathway. And these two events are required for the enough level of the SETBP1 protein. So that means this type of poor prognostic leukemia occurs in the SETBP1 abundance dependent fashion. And finally we found that depending on the abundance of SETBP1 protein, they form the nuclear condensates, transcriptional condensate formation, to sustain the high level of the leukemic transcriptional program like HOXA9, HOXA10, MYB and MECOM and so on. So that's why in this presentation we introduce the story why these two are required and why this specific order is important for this poor prognostic leukemia.

《肿瘤瞭望-血液时讯》：针对这类难治/复发性极高危白血病，您认为未来的突破口在于单药的新靶点，还是基于克隆演化动态的联合用药策略？在临床转化中，应如何选择最佳干预时机？

Daichi Inoue教授：如前述，此类白血病依赖于致癌基因蛋白的丰度，并形成核凝聚体以增强白血病特异性转录程序。本团队监测了凝聚体的形成，并尝试结合CRISPR全基因组筛选与小分子抑制剂筛选进行平行筛选，以探寻此类形成凝聚体白血病的“阿喀琉斯之踵”（致命弱点）。

最终，本团队发现了一种单分子抑制剂，该抑制剂已被FDA批准用于其他类型的癌症。本团队尝试通过药物重定位来消除核凝聚体的形成。该药物对动物模型乃至PDX模型均表现出高度有效性，且具备根除白血病干细胞、维持长期完全缓解、甚至治愈100%外显率预后不良白血病的潜力。

关于干预时机，本团队在恶性克隆已扩增但尚未完全转化的白血病前期阶段进行了单细胞RNA测序分析，发现在这些阶段已存在基于凝聚体的转录程序。这意味着临床干预应在完全转化之前尽早开展，即在克隆扩增阶段。但这基本上需通过连续的NGS分析来实现（若条件允许）。

Oncology Frontier-Hematology Frontier: For this highly refractory/relapsed high-risk leukemia, do you think the future breakthrough lies in new single-agent targets or combination strategies based on clonal evolution dynamics? In clinical translation, how should the optimal intervention timing be selected?

Professor Daichi Inoue: Well, another fantastic question. So and as I mentioned this type of leukemia is depending on the abundance of the oncogene proteins. And this forms the nuclear condensate formation to enhance the leukemia specific transcriptional program. And we monitored the condensate formation and try to conduct the parallel screening with the CRISPR whole genome CRISPR screening and small molecule inhibitor screening to find out the Achilles heel of these condensate forming type leukemia. And we finally found that the single molecule and this inhibitor is basically approved by FDA in other types of the cancer. And we tried the drug repositioning to cancel the nuclear condensate formation. And this drug was really effective against the animal models and even the PDX model. And and this drug has the potential to eradicate leukemic stem cells. So and sustain the long term complete remission, even cure the 100% penetrance poor prognostic leukemia.

And about the second question about for the timing of the our intervention into the this poor prognostic leukemia. And we found we conducted the single cell RNA seq analysis in the preleukemic stages where the malignant clones expanded is not fully transformed. But at this stages we found that there is the condensate based transcriptional program. So that means we should intervene early before the fully transformation at the stage of the clonal expansion. But this is basically accomplished by the sequential NGS analysis if it is available.

专家简介

井上大地（Daichi Inoue）教授

日本大阪大学

井上大地教授现任大阪大学がん病理学教授。他于2005年毕业于京都大学医学部，2014年获东京大学医学博士学位，2015至2019年赴美国纪念斯隆-凯特琳癌症中心（MSKCC）从事博士后研究，2024年起出任现职。

其研究聚焦骨髓增生异常综合征（MDS）与急性髓系白血病的分子发病机制，在RNA剪接异常与表观遗传调控领域取得系列成果，代表性工作揭示了ASXL1突变诱导MDS的机制及SETBP1突变驱动白血病转化的路径。曾获日本学术振兴会奖（JSPS Prize，2025年）