以豬隻模式探討腹膜纖維化分子機制與評估治療效能

Abstract

慢性腎病是現代常見的慢性病之一，泛指長達三個月以上的腎功能問題，並通常按照腎絲球過濾率（Glomerular filtration rate，GFR）的數值被分為五期。其中，第五期又被稱為末期腎病（End-stage kidney disease，ESKD）或是腎衰竭，此時期病患的腎臟已無法負擔正常的代謝與體液平衡，會出現尿毒症的症狀，需要使用腎臟替代療法，如腎臟移植、血液透析或是腹膜透析（Peritoneal dialysis，PD）等來維持生命。腹膜透析的原理是利用患者本身的腹膜作為半透膜，透過高張透析液的灌入、存留與排出，來交換出體內多餘的毒素與水分。研究發現，在多年進行腹膜透析的病患身上，經常觀察到腹膜纖維化（Peritoneal fibrosis）與腹膜功能改變等現象，若嚴重可能導致超過濾失敗（Ultrafiltration failure）的發生，使病患必須終止使用腹膜透析，轉往血液透析或是等待腎臟移植。纖維化的腹膜組織，相較於健康的腹膜，常可見間皮細胞（Mesothelial cells）的脫落、細胞外基質（Extracellular Matrix，ECM）的沉積與血管新生（Angiogenesis）等病變。葡萄糖透析液是傳統上最常使用的腹膜透析液，然而這類的透析液被發現在加熱與儲存的過程中，易產生葡萄糖降解產物（Glucose degradation products, GDPs），引起腹膜發炎反應並導致腹膜產生纖維化病變。目前在臨床上用來治療腹膜纖維化的藥物相當有限，許多開發中的藥物仍停留在動物實驗的階段，並以使用嚙齒類等小型動物的實驗動物為主。然而，囓齒類的實驗動物在解剖、生理等構造上與人類有一定的差異，這些差異經常使得研究結果難以順利轉譯到人類臨床應用。本篇博士論文以建立腹膜纖維化的大動物模式來協助相關藥物篩選與機制探討出發，嘗試建立豬的腹膜纖維化動物模式（Porcine peritoneal fibrosis model），建立後測試有潛力的抗纖維化藥物蕾莎瓦（Sorafenib），也使用細胞實驗探討腹膜纖維化相關機制。 首先，我們在豬身上建立了能夠即時評估腹膜功能的腹膜平衡試驗（Peritoneal equilibration test，PET），也建立以一週內兩劑低濃度的次氯酸鈉腹腔注射來引起腹膜纖維化的豬動物模式。這個動物模式除了可見腹膜纖維化的組織病變，也有發生相似於長期腹膜透析病患腹膜功能上的異常，如Small solute的運輸速率上升、鹽篩濾（Sodium sieving）現象消失等。針對鹽篩濾現象的消失，我們進一步確認了纖維化腹膜組織上水通道蛋白1號（Aquaporin 1，AQP1）的含量與分佈，並使用間皮細胞（Mesothelial cells）與血管內皮細胞（Endothelial cells）的細胞株進行實驗。認為次氯酸鈉在細胞上所引起的氧化壓力（Oxidative stress），很有可能是進一步導致細胞骨架改變與水通道蛋白運輸受損的來源，才會進一步導致鹽篩濾現象的消失。 為了建立更接近臨床病患的豬腹膜纖維化模式，我們接著使用2.5%的葡萄糖腹膜透析液加入葡萄糖降解產物之一的甲基乙二醛（Methylglyoxal，MGO）來建立第二個豬腹膜纖維化模式。成功建立豬模式後，我們嘗試使用此動物模式測試口服蕾莎瓦在腹膜纖維化的療效。實驗結果顯示，口服蕾莎瓦能夠有效減緩腹膜的增厚、肌纖維母細胞（Myofibroblast）的增加與血管新生，然而，對於間皮細胞與腹膜功能的保護則相當有限。我們認為單獨使用口服蕾莎瓦治療腹膜纖維化的療效有限，建議需要搭配其他藥物一起使用。 從嘗試使用蕾莎瓦的結果當中，我們發現藥物可能僅對於腹膜纖維化的部分參與細胞有干涉，因此後續使用間皮細胞與纖維母細胞（Fibroblast）細胞株希望先探討甲基乙二醛可能引起腹膜纖維化的背後機制，以及這兩種細胞在纖維化過程的互動關係。期望在更加瞭解背後機制後，能夠尋找到關鍵的治療點來決定與開發藥物。首先我們將甲基乙二醛刺激下的間皮細胞進行次世代定序（Next generation sequencing，NGS），發現間皮細胞的轉錄體在細胞凋亡（Apoptosis）、促發炎反應（Pro-inflammatory response）與纖維化相關的路徑上有明顯的改變。另外，我們發現若將甲基乙二醛刺激間皮細胞後的上清液濃縮，加入纖維母細胞的培養液，會使得纖維母細胞的細胞骨骼發生改變、增加分泌細胞外基質的能力，被活化成前驅肌纖維母細胞（Proto-myofibroblasts）。然而，因著並未出現平滑肌肌動蛋白（Alpha smooth muscle actin, α-SMA）的表現，可以確認其尚未轉換成造成纖維化的重要細胞——肌纖維母細胞。推測可能還需要其他細胞的參與或因子，才能夠使纖維母細胞完全轉換成肌纖維母細胞。

End-stage kidney disease (ESKD) is the severest state of chronic kidney disease that patients with ESKD cannot survive without the intervention of renal replacement therapy. Belongs to renal replacement therapy, peritoneal dialysis (PD) maintains 11% of ESKD patients’ lives around the world. With high osmotic dialysate in the peritoneal cavity, the peritoneum is used as the semipermeable membrane to remove excess fluids and wastes from the body during PD. However, pathological changes in peritoneal morphology and function, which is termed peritoneal fibrosis, were often observed during long-term PD and can gradually result in the termination of PD. Peritoneal fibrosis, characterized by denudation of mesothelial cells, infiltration of myofibroblasts, and angiogenesis, is considered to result from the repeated inflammation caused by the components in the dialysate. To date, only a limited number of therapeutic options could be provided to patients with peritoneal fibrosis in clinical. Most of the newly developed treatments for peritoneal fibrosis have been tested in rodent-based models; however, no large animal models can be applied to bridge the discrepancies between rodents and humans or validate of which treatment may have better and more effective potential in humans. Therefore, establishing porcine peritoneal fibrosis models was aimed at accelerating the development and screening of therapeutic options for peritoneal fibrosis in this doctoral dissertation. Firstly, a peritoneal equilibration test (PET) for evaluating instant changes of peritoneal function was established on pigs. By intraperitoneal administrations of 0.1% NaClO, histological features of peritoneal fibrosis and altered function on the peritoneum were both observed. Correlated to defects of sodium sieving in PET on NaClO-injured pigs, decreased amounts of aquaporin1 (AQP1) in the fibrotic peritoneum were noted. Based on in vivo and in vitro results, we proposed that disruption of the cytoskeleton induced by excessive reactive oxygen species defected intracellular transport of AQP 1 and resulted in the disappearance of sodium sieving. Next, we established a novel and clinically relevant porcine model of peritoneal fibrosis induced by 40 mM methylglyoxal (MGO) in 2.5% glucose-based dialysate. Upon successful establishment of the MGO-induced peritoneal fibrosis pig model, sorafenib, a tyrosine kinase inhibitor with the potential to treat peritoneal fibrosis, was given orally to evaluate its therapeutic efficacy. We observed that sorafenib effectively alleviated the thickening of the peritoneum, angiogenesis, myofibroblasts infiltration, and decreased endothelial glycocalyx. However, therapeutic efficacy in ameliorating the loss of mesothelial cells, decreased ultrafiltration volume and elevated solutes transport rates was limited. These results demonstrated that applying sorafenib alone was not sufficient to fully rescue MGO-induced peritoneal defects. Since the distinct responses from different cell types upon sorafenib treatment were observed, we further unveiled the roles and interactions between participated cells in MGO-induced peritoneal fibrosis. The transcriptomic data from the MGO-stimulated mesothelial cells showed a significant upregulation of genes involved in pro-inflammatory, apoptotic, and fibrotic pathways. As for fibroblasts, no phenotypic changes were noted after direct stimulation of MGO, whereas supernatant from MGO-stimulated mesothelial cells promoted fibroblasts transforming into proto-myofibroblasts, the first stage of activation toward myofibroblasts. However, additional involvement of other factors or cells (e.g., macrophages) may be needed for a complete transformation of fibroblasts into myofibroblasts.