An Interleukin-25-Mediated Autoregulatory Circuit in Keratinocytes Plays a Pivotal Role in Psoriatic Skin Inflammation.

2020/01/05

與多數上皮癌一樣，口腔癌的發生是漸進性的，口腔癌患者在確診前的一段時間，常先出現扁平苔蘚、紅白斑、口腔粘膜下纖維化、上皮發育不良型增生或過度角化、上皮疣狀增生等種種不正常的病變，統稱為「口腔癌前病症」。台灣自民國88年開始進行口腔癌篩檢後，追蹤癌前病症患者以杜絕其病灶惡性轉換為口腔癌，是臨床上一件重要卻負荷不輕的使命。最新結合一萬六千多位患者的追蹤資料顯示，各型的口腔癌前病症其惡性轉癌風險並不相同 (十年內5–68%)，如何及早將高危險群患者鑑定出來，且於未惡性轉換前施予有效之化學預防療程，是口腔癌篩檢計畫中最後、也是最重要一哩路。此外，儘管國外有一些進行頭頸癌癌前病症的化學預防研究，但其病因多與抽煙與喝酒有關，缺少本土性口腔癌「檳榔」因子，是否適用於國人實屬未知。本計畫乃基於這兩個迫切的議題，深入探討檳榔鹼對人類口腔角質細胞的癌變初始化機制，將其與精準分子診斷技術結合，以產出能轉譯於臨床上口腔癌前病症患者之風險評估及預防措施，以降低國人口腔癌罹患率暨提升患者存活率。

先前成果

免疫基因體學之進展日新月異，實驗室先前以CIBERSORT等免疫生物資訊演算程式，針對NCKU OrCA 40TN之80例組織，進行22種免疫細胞分析，發現40例鄰近正常組織內其 resting CD4 memory T 顯著地較40例腫瘤組織為高 (p = 5.05e-08); 而 activated dendritic cells 則顯著地降低 (p = 4.96e-07)。另外，多項新穎指數計算程式，可有效將四十對檢體加以分類，且與病人整體存活率具顯著關聯性，此份資料庫將成為日後研究口腔癌前病症最佳之腫瘤對照組。

為探討檳榔鹼誘導口腔角質細胞惡性病變之分子機轉，實驗室先前以低於致死劑量之檳榔鹼重複刺激人類口腔上皮細胞五十次以上，建立三對檳榔鹼有無處理之細胞株配對 (CTRL/AC)，並已完成全轉錄體定序分析，這些材料將成為本計畫中檳榔鹼與癌前病症惡性轉化研究之細胞模型。

milestones and contents

	110	111	112	113
milestones	intrinsic factors responsible for immune escape are identified	stromal effects involved in malignant transformation are confirmed	molecular subtype classifier is established	application of classifier and risk association of subtypes in clinical specimens
工作重點	細胞實驗轉錄體定序分析與結果視覺化動物實驗模型建立	動物實驗進行細胞異種移植組織轉錄體定序分析與結果視覺化多種多效價免疫化學組織染色法測試與評估	腫瘤組織亞型分子器程式評估、測試、修改 進行多效價免疫化學組織染色法	臨床檢體分型與風險評估 以細胞與動物實驗驗證惡性轉化因子及相關訊息傳導路徑

計畫第一年將著重於三對CTRL/AC口腔角質細胞株之特性研究以及轉錄體定序分析與結果視覺化。各配對細胞株將於其對AC之耐受性、細胞生長速度、clonogenicity、ROS清除能力等生物功能上逐一做比較。而於轉錄體定序結果分析上，在結合GSEA, DAVID, MOLAS等多種方式分析DEGs後，依照FDR或p-value之評估結果，選擇數項在AC與CTRL組有顯著差異的生物功能，設計實驗逐項予以驗證。本年度亦著手進行動物實驗之pilot study。

免疫癌症學之進展日新月異，實驗室先前以CIBERSORT等腫瘤組織微環境成分估算程式，針對NCKU OrCA 40TN (GSE37991)之八十例組織進行22種免疫細胞成份分析(LM22)，發現四十例鄰近正常組織內其 resting CD4 memory T顯著地較四十例腫瘤組織為高 (p = 5.05e-08); 而activated dendritic cells則顯著地較四十例腫瘤組織為低 (p = 4.96e-07)，顯示NCKU 40TN這組dataset，可最為日後口腔癌前病症之轉錄體分析之腫瘤對照組。

動物實驗模型建立

★       請整體性的說明計畫全程如何透過逐年的工作推展，達成預期目標與最終效益，並加強說明其對本院之增益性。例如：轉譯成果如何突顯本院做為政府智庫之角色；技術移轉或產學合作所帶來的產值與經濟效益；技術上的突破如何提升國內產業發展之競爭力。

免疫與癌症基因體學之進展日新月異，過去單基因、單路徑在特定細胞株中導致癌症發生或惡化的觀察，已多數被認為不敷使用、無法應用於臨床上。腫瘤的形成有來自癌細胞的內在因子與微環境中間質成份提供之外在因子，唯有兩種因子一起考量，方能有效控制腫瘤之生長與蔓延。口腔癌前病症是一類富含免疫反應的慢性發炎癌前病灶，本計畫透過找出與癌前病症惡性轉化有關的內在與外在因子，建立口腔癌前病症分子亞型分類器(classifier)，找出高危險群、予以預防，以完成口腔癌篩檢計畫最重要的目標: 降低口腔癌罹患率並提升存活率。

Aim1	To identify intrinsic factors involved in immune escape in AC trained oral keratinocytes
Aim2	To elucidate extrinsic factors responsible for malignant switch of premalignant cells
Aim3	To develop molecular classifiers applicable in clinical use
Aim4

Based on cancer immunoediting theory, in a chronically inflamed precancerous microenvironment (PME), the genome of a diseased keratinocyte is continuously “edited” by the immune system. Only when cells with low immunogenicity, or are able to counteract with immune surveillance by other means, can escape to initiate a tumor mass. These are the intrinsic factors of tumor cells. On the other hand, the stromal constituents of PME also play a crucial role, including fibroblasts, endothelial cells, and extracellular matrix attached with abundant pro-growth factors. These are the extrinsic factors that foster tumor growth. In this grant, we propose to molecularly stratify the PME of oral premalignant disorders, followed by identifying targetable molecules for preventing malignant transformation in the high-risk group. Specific aims are listed as follows.

Aim 1: To identify intrinsic factors involved in immune escape in AC trained oral keratinocytes

Aim 2: To elucidate extrinsic factors responsible for malignant switch of premalignant cells

Aim 3: To develop molecular classifiers applicable in clinical use

Based on cancer immunoediting theory, diseased keratinocytes attract competent immune system to the We observed that a battery of immune responsive genes and pathways were down regulated in arecoline treated human oral keratinocytes (Fig. 2), suggesting a previously undocumented tumorigenic role of arecoline. In in vivo milieu, after multiple rounds of selection, cells with least immunogenicity will evolve to form a tumor mass, which then be fostered by its own tumor microenvironment (TME). 

By using various omics methodologies, at least 4 molecular subtypes of HNC have been identified: BA (basal), MS (mesenchymal), CL (classical), and AT (atypical) (Table 2). Among them, our NCKU 40T/N cohort comprises BA, MS, and CL subtypes. In addition, OSCC cell line OC3 was predicted to be MS; TW2.6 was predicted to be CL (Fig. 3). The TME of oral cancer was inferred to be rich of immune cells yet is rather immunosuppressive. Stratification of oral cancer into molecular/immune subtypes is crucial to precision therapy because immune-hot and immune-cold tissues will be handled differently. Currently, various immunotherapies are in clinical trails, many preliminary results are very promising (Table 3). 

Before reliable assay for molecular stratification of clinical specimens is available, patient derived xenograft (PDX) maintained in immunodeficient mice (NSG or NOG) provides a valuable source for translational research. Using OC3 and TW2.6 cell pellet as tumor grafts, we established oral cancer cell derived xenograft (CDX) in NOG (Tables 5-6) and huMice (NOG with human immune system, unpublished data). Preliminary results suggested that OC3 is more immunogenic than TW2.6 in NOG mice, of which murine stromal DC and macrophage might be the principal immune elements that restrict human tumor cells grown in these immunodeficient mice. 

This 2-yr research proposal seeks to integrate the most updated information and methodologies to understand mechanisms of immune evasion in oral cancer tumorigenesis. We hypothesized that (1) immunoediting takes place in the OPMD stage in which immune tolerance biomarkers should be elucidated and immunosuppressive PME should be intervened. (2) Immune stratification of each tumor specimen w

Specific Aim 3. Study of tumor infiltrated immune cells in clinical FFPE specimens.

Summary: Although innovative technologies such as RNA-seq or multi-channel cell sorter greatly advance our knowledge in enumerate tumor infiltrated lymphocytes (TILs) present in the tumor microenvironment (TME), however, they require considerable amount of tumor tissues, more time consuming, and are difficult to obtain geographical information of interested cells within the TME. In contrast, results from immunohistochemistry (IHC) and assay alike provide superior spatial evidence for cells of interest. On the other hand, given the complexity and heterogeneity in formalin fixed paraffin embedded (FFPE) tissue, educated decision of target gene/protein/cell to be probed is critical to the success of immune stratification. For example, by meta-data analysis, a recent repot suggests that both CD8 T and NKCD56dim cell infiltration are independent good survival biomarkers in HNC [6]. In addition, the inflamed mesenchymal subtype identified by Keck et al exhibits a strong CD8 T cell infiltration that is associated with good outcome [7]. Results from such studies directly provide information for biomarkers that we should track on. Thus, in this specific aim, efforts will be focused on establishing reliable methods for simultaneous detection of multiple cell types present in clinical FFPE specimens, including those from oral cancer and OPMD patients. Two methods are likely to fulfill this need: TSA IHC and RNA in situ hybridization. Please limit to one page. 。）

綜合上述實驗觀察，我們推測嚼食檳榔而罹癌之患者口腔內，角質細胞在檳榔鹼刺激及組織微環境中免疫系統監控下，重複進行「癌症-免疫週期」(cancer- immunity cycle)。過程中如果細胞本身漸漸關閉免疫反應相關基因路徑，則可應付外在免疫系統的攻擊，長期下來最會躲避免疫監控的腫瘤細胞將被保留下來，例如TW2.6口腔癌細胞。基於這些假設，未來二年將(1)進行驗證實驗以找出檳榔相關口腔癌進展過程中參與的免疫逃脫主要路徑及分子；(2)以口腔癌前病變及口腔癌臨床蠟塊為測試檢體，利用類似多價免疫組織化學染色法(multiplex immunohistochemical staining)之技術，剖繪癌前病變/腫瘤微環境中免疫細胞之時空 (spatiotemporal) 浸潤情形。

英潔早安，可以麻煩妳把昨天報告的檔案分享在雲端嗎？看是要存在AK459相關的地方或者是妳是自己的2018-2019 lab meeting folder (再貼上來link謝謝)