以新發展之微生物泳動分析演算法揭示 Haloarcula marismortui 之光趨性

Abstract

過去微生物學家認為Haloarcula marismortui 不會泳動，然而近年的研究證明了H. marismortui 鞭毛的存在，顯示其具有泳動能力。本研究觀察H. marismortui 的活菌泳動現象，並以ImageJ 追蹤其泳動軌跡 (swimming trajectory)。由於其軌跡相當雜噪 (noisy)，因此難以適用於前人所建立的泳動分析演算法 (motion analysis algorithm)。在此我提出 “window vector” 的概念，以Microscope Excel-VBA 程式語言建立一個微生物泳動分析演算法 (microbial motion analysis algorithm)，具有以下功能：(1) 不需以實驗結果個別化估算分析參數 (empirical parameter customization)，而用冪次率關係 (power-law relationship) 自動分辨細胞為泳動或非泳動；(2) 減低布朗運動對泳動軌跡造成的雜訊 (noise)，因此提升泳動折返 (swim reversal) 判斷的準確性。藉此演算法，本研究證明了HmSRI 和HmSRII 這兩個近來被確認的感光視紫質 (sensory rhodopsin) 的生理功能，分別是趨光性 (photoattractant) 和避光性 (photorepellent) 反應。H. marismortui 是除了 Halobacterium salinarum 外，唯一被確認具有光趨性 (phototaxis) 的古生菌。

Haloarcula marismortui has been described to be nonmotile prior to the recent identification of flagellar filaments, suggesting the motile nature of H. marismortui. Here we observed the locomotion of freshly cultured H. marismortui cells and tracked the swimming trajectories via ImageJ. Trajectories of H. marismortui are intrinsically noisy, posing difficulties in motion analysis with previously established algorithms. By introducing the concept of ‘‘window vector,’’ a Microsoft Excel-VBA-implemented microbial motion analysis algorithm reported here was able to (1) discriminate nonswimming objects from swimming cells without empirical customization by applying a power-law relationship and (2) reduce the noise caused by Brownian motion, thus enhancing the accuracy of swim reversal identification. Based on this motion analysis algorithm, two recently identified sensory rhodopsins, HmSRI and HmSRII, were shown to mediate photoattractant and photorepellent responses, respectively, revealing the phototactic activity of H. marismortui, the only archaeon showing such phenomenon other than Halobacterium salinarum.