SOX2+細胞與BRAF V600E突變於造釉細胞瘤致病機轉之研究

Abstract

造釉細胞瘤(ameloblastoma)是最常見且具臨床意義的齒源性腫瘤，約占所有頭頸部腫瘤的1%及齒源性腫瘤的40%。目前認為其起源於牙齒發育過程中殘留的齒源性上皮細胞，尤其是牙板殘餘細胞。雖然造釉細胞瘤被歸類為良性腫瘤，但具有局部侵襲性及高度復發傾向等臨床特徵，其預後很大程度上取決於手術切除的完整性。目前廣泛性顎骨切除仍為治療黃金標準，然而此類手術常造成嚴重的顏面缺損、功能障礙及後續重建困難，對病患生活品質造成重大影響。 近年來分子病理學研究已證實，絲裂原活化蛋白激酶（mitogen-activated protein kinase, MAPK）訊息傳遞路徑中的BRAF V600E突變為造釉細胞瘤最重要的驅動基因變異。雖然BRAF抑制劑已在無法接受手術治療且帶有BRAF V600E突變的患者中展現良好的腫瘤縮小效果與臨床應用潛力，但其通常無法完全清除腫瘤，顯示腫瘤細胞可能透過其他替代性訊息傳遞路徑維持生存，並暗示腫瘤幹細胞（tumor stem cells, TSCs）可能參與治療抗藥性之形成。 SOX2為一重要轉錄因子，廣泛表現於胚胎及成體幹細胞中，並影響幹細胞自我更新與多向分化能力之維持。研究顯示，SOX2亦於多種惡性腫瘤中高度表現，且與腫瘤生長、侵襲能力、治療抗性及不良預後密切相關。此外，牙胚中表現SOX2之細胞已被證實具有分化為各類齒源性上皮細胞之潛能，因此SOX2被視為齒源性上皮幹細胞的重要標誌。本研究之前導研究首先利用免疫組織化學染色分析SOX2於齒濾泡齒源性上皮細胞及不同組織型態造釉細胞瘤中的表現情形。合併後續實驗之結果顯示，齒濾泡齒源性上皮細胞與造釉細胞瘤腫瘤細胞皆有SOX2+細胞的存在。並在接續實驗證實初代細胞培養之齒濾泡齒源性上皮細胞與造釉細胞瘤腫瘤細胞皆具有幹細胞相關特徵，且均表現SOX2。進一步功能性研究證實，SOX2對造釉細胞瘤細胞存活具有重要影響，且帶有BRAF突變之病例及對BRAF抑制劑產生抗藥性的腫瘤細胞均呈現較高SOX2表現。 為進一步探討造釉細胞瘤之分子致病機制，本研究利用轉錄體定序分析比較新鮮造釉細胞瘤組織與齒濾泡組織之基因表現差異。為降低組織間質造成之干擾，進一步比較初代培養之造釉細胞瘤上皮細胞與齒濾泡中上皮細胞之轉錄體表現，並取兩種分析情境下差異表現基因（differentially expressed genes, DEGs）之交集進行後續分析。結果顯示，除MAPK訊息傳遞路徑活化外，Wnt/β-catenin及PI3K/AKT/mTOR訊息傳遞路徑亦可能參與造釉細胞瘤之發展。 後續進一步比較帶有BRAF突變與未帶有BRAF突變之造釉細胞瘤組織後發現，BRAF突變型腫瘤中多達七項Wnt/β-catenin訊息傳遞路徑上游抑制因子呈現顯著低表現；相較之下，BRAF無突變病例則表現較高之GLI1與GLI2，顯示Hedgehog訊息傳遞路徑可能於此亞群腫瘤中扮演較重要角色。最後，透過全轉錄體空間轉錄體分析進一步驗證先前發現之差異表現基因，並證實Wnt/β-catenin及Hedgehog訊息傳遞路徑相關活化因子於造釉細胞瘤腫瘤上皮區域具有較高表現。 綜合而言，本研究揭示SOX2於造釉細胞瘤幹細胞特性維持及BRAF抑制劑抗藥性形成中的潛在角色，並透過轉錄體與空間轉錄體分析，建立造釉細胞瘤更完整的致病分子病理機轉。研究結果顯示，除MAPK訊息傳遞路徑外，Wnt/β-catenin、PI3K/AKT/mTOR及Hedgehog等多條致癌訊息傳遞路徑亦共同參與造釉細胞瘤之發生與進展。這些發現不僅拓展了目前以單一路徑為主的造釉細胞瘤致病模型，也為未來發展多重訊息傳遞路徑抑制、合併標靶治療策略提供部分理論基礎。

Ameloblastoma is the most common and clinically significant odontogenic tumor, making up about 1% of all head and neck tumors and 40% of odontogenic tumors. It is thought to originate from residual odontogenic epithelium during tooth development, especially the rests of the dental lamina. Although classified as benign, ameloblastoma is locally aggressive and frequently recurs. Its prognosis depends heavily on how completely it is surgically removed. Currently, extensive jaw resection remains the gold standard treatment. However, this radical surgery often causes severe facial deformities, functional loss, and reconstruction challenges, which seriously damage the patient's quality of life. Recently, molecular pathology has shown that the BRAF V600E mutation in the MAPK signaling pathway is the key driver mutation in ameloblastoma. BRAF inhibitors effectively shrink tumors in patients with this mutation who cannot undergo surgery, but they rarely eliminate the tumor completely. This suggests that tumor cells survive through alternative pathways and implies that tumor stem cells (TSCs) might drive this treatment resistance. SOX2 is a key transcription factor in embryonic and adult stem cells that maintains self-renewal and pluripotency. It is also highly expressed in various malignant tumors, where it links to tumor growth, invasion, drug resistance, and poor prognosis. Additionally, SOX2-expressing cells in the tooth germ can differentiate into different types of odontogenic epithelial cells, making SOX2 an important marker for dental epithelial stem cells. In our pilot study, immunohistochemical (IHC) staining showed that SOX2+ cells are present in both dental follicle odontogenic epithelial cells and ameloblastoma cells. Follow-up experiments confirmed that primary-cultured cells from both tissues show stemness features and express SOX2. Further functional tests proved that SOX2 is vital for ameloblastoma cell survival. Notably, SOX2 expression was higher in BRAF-mutated cases and in cells that developed resistance to BRAF inhibitors. To explore the molecular mechanisms further, this study used RNA sequencing (RNA-seq) to compare gene expression between fresh ameloblastoma and dental follicle tissues. To avoid interference from the stromal microenvironment, we also compared primary-cultured epithelial cells from both tissues. We then focused on the intersection of differentially expressed genes (DEGs) from both setups. The results showed that besides MAPK activation, the Wnt/β-catenin and PI3K/AKT/mTOR pathways also likely drive the development and progression of ameloblastoma. Next, we compared BRAF-mutated and BRAF wild-type ameloblastoma tissues. In BRAF-mutated tumors, seven upstream inhibitors of the Wnt/β-catenin pathway were significantly downregulated. In contrast, BRAF wild-type cases expressed higher levels of GLI1 and GLI2, suggesting the Hedgehog pathway plays a bigger role in this specific subgroup. Finally, spatial transcriptomics validated these DEGs, confirming that activation factors for Wnt/β-catenin and Hedgehog pathways are enriched in the epithelial regions of the tumor. In conclusion, this study highlights the role of SOX2 in maintaining stemness and driving BRAF inhibitor resistance in ameloblastoma. By combining transcriptomic and spatial transcriptomic data, we map out a more complete molecular pathology for the disease. Our findings show that ameloblastoma is driven not just by the MAPK pathway, but by a network of oncogenic pathways including Wnt/β-catenin, PI3K/AKT/mTOR, and Hedgehog. This shifts the current single-pathway model to a multi-pathway view, providing a rationale for future combination targeted therapies.