利用病人腫瘤異體移植小鼠模型探討KMT2A-r亞型B細胞急性淋巴性白血病中樞神經侵犯相關基因

Abstract

急性淋巴性白血病(Acute lymphoblastic leukemia，ALL)為淋巴系前驅白血球發生基因突變，進而不良性增生所造成的惡性血液疾病。其好發於年輕族群，在台灣為兒童癌症發生率之首。根據免疫分型，急性淋巴性白血病可分為T細胞型(T-ALL)及B細胞型(B-ALL)，並再根據基因突變分為不同亞型。在B-ALL中，常見突變有hyperdiploid、hypodiploid、KMT2A rearrangement、BCR::ABL1、ETV6::RUNX1、TCF3::PBX1等，其中KMT2A rearrangement屬於中樞神經侵犯高風險亞型，同時該亞型預後較差，且在一歲以下嬰幼兒新診斷病例占約70%以上。在現行治療下，對於中樞神經侵犯已有良好的成效，然而對於其本身的機制尚未明瞭。過往文獻提出過多種不同假說，指出某些基因與中樞神經侵犯有關，如2018年有文獻提出B-ALL細胞藉由表現integrin α6，利用血管外壁之laminin由顱骨骨髓腔爬行至蜘蛛網膜下腔；2020年一文獻則觀察到ITGA5及ITGA9的表現量上升與B-ALL細胞於中樞神經系統定殖(colonization)成正相關性；2021年又有另一文獻則提出B-ALL細胞利用蝕骨作用以及CXCR4介導的趨化訊號得以從顱骨進入蜘蛛網膜下腔。然而目前尚未有一定論提出某一機制主導中樞神經侵犯發生，不同亞型是否由同一機制主導也並未確定。綜上所述，本研究旨在探討與B-ALL中樞神經侵犯相關之基因，並集中於KMT2A rearrangement亞型；鑒於中樞神經侵犯檢體不易大量取得，本研究亦使用病人腫瘤異體移植模型(Patient-derived xenograft model，PDX model)克服此一困難。本研究先對十五位KMT2A rearrangement B-ALL病人原有檢體進行RNA定序分析，未發現有基因集於有中樞神經侵犯病人之血癌細胞顯著調升，推論其原因為病人原有檢體非中樞神經侵犯檢體，故未能觀察到與中樞神經侵犯能力及發生等相關的基因表現變化。本研究接著使用重度免疫缺陷NSG™小鼠進行PDX模型之建立，並先採用人類血癌細胞株RS4;11進行異體移植前驅實驗。RS4;11移植於小鼠後，可自小鼠頭部切片觀察到其分布於蜘蛛網膜下腔區域，表明該移植模型可呈現中樞神經侵犯表型。鑒於RS4;11移植模型的成功，我們自前述病人中挑選六位病人個別建立PDX模型，其中三位病人診斷時有中樞神經侵犯，三位則無。對自PDX小鼠分離之中樞神經侵犯細胞進行RNA定序分析，發現整合素基因如ITGA6、ITGA7、ITGA9、ITGB5有顯著調升，同時數個整合素基因如ITGA2B、ITGB2、ITGB3等有顯著調降。以過表徵分析(ORA)則自KEGG資料庫辨別出如「ECM receptor interaction」、「PI3K-Akt signaling pathway」等基因集有所調升，前者包含整合素基因ITGA7、ITGA9、ITGB5，後者包含ITGA6。綜上所述，本研究成功建立PDX模型以供辨識主導KMT2A rearrangement B-ALL中樞神經侵犯的基因路徑。

Acute lymphoblastic leukemia (ALL) is a malignant hematologic disease caused by genetic mutations in lymphoid precursor cells, leading to abnormal proliferation. It predominantly affects young individuals and is the most common childhood cancer in Taiwan. ALL can be classified into T-cell type (T-ALL) and B-cell type (B-ALL) by immunophenotyping, and further classified into different subtypes according to the genetic mutation of the cancer cells. In B-ALL, common mutations include hyperdiploid, hypodiploid, KMT2A rearrangement (KMT2A-r), BCR::ABL1, ETV6::RUNX1, TCF3::PBX1, etc.. The KMT2A-r subtype is associated with a high risk of central nervous system (CNS) involvement and has a poorer prognosis, with over 70% of new diagnoses in infants under one year old. Current treatments have shown good efficacy against CNS involvement, but the underlying mechanisms remain unclear. Previous studies have proposed various hypotheses linking certain genes to CNS involvement. For instance, a 2018 study suggested that B-ALL cells express integrin α6, using laminin on the vascular wall to migrate from the cranial bone marrow cavity to the subarachnoid space. A 2020 study observed increased expression of ITGA5 and ITGA9, correlating with B-ALL cell colonization in the CNS. Another study in 2021 proposed that B-ALL cells utilize osteolysis and CXCR4-mediated chemotaxis to enter the subarachnoid space from the cranial bone. However, no definitive mechanism has been suggested, and it is unclear whether different subtypes are governed by the same mechanism. Our study aims to investigate genes related to CNS involvement in B-ALL, especially focusing on the KMT2A-r subtype. Due to the difficulty in obtaining large quantities of CNS involvement samples, we also utilize patient-derived xenograft (PDX) model to overcome this challenge. Initial RNA sequencing analysis of original samples from 15 KMT2A-r B-ALL patients did not reveal significant gene set upregulation in leukemia cells from patients with CNS involvement. We proposed that because the original samples were not from CNS involvement samples, we weren’t able to observe expression changes of genes related to CNS involvement. We then established PDX models using severely immunodeficient NSG™ mice, while first establishing xenograft model with human KMT2A-r B-ALL cell line RS4;11 as a pilot study. After transplantation, we observed RS4;11 in the subarachnoid space in mice head sections. Thus we confirm that the RS4;11 xenograft model can represent CNS involvement phenotypes. Based on the successful results of RS4;11, we select six patients, three with CNS involvement and three without, and we establish PDX models for individual patients. RNA sequencing analysis of CNS involvement cells isolated from PDX mice revealed significant upregulation of integrin genes such as ITGA6, ITGA7, ITGA9, and ITGB5 and downregulation of several integrin genes such as ITGA2B, ITGB2, and ITGB3. Furthermore, Over-Representation Analysis (ORA) identified several gene sets that are enriched in CNS involvement cells, such as “ECM receptor interaction”, in which ITGA7, ITGA9, and ITGB5 are included, and “PI3K-Akt signaling pathway”, in which ITGA6 is included. Taken together, we have established PDX models that allow us to identify novel pathways that govern CNS involvement in KMT2A-r B-ALL