含鹵素結構吡啶類對藻類之毒性影響評估

Abstract

本研究為了探討鹵素結構吡啶類對於藻類之毒性影響評估，選用單細胞藻類-月芽藻(Pseudokirchneriella subcapitata)為試驗物種，毒性物質為含鹵素取代基之吡啶，以Δ DO、Final Yield、Growth Rate等三個反應終點，進行一系列之藻類毒性試驗。針對難溶解藥物以添加丙酮(Acetone)改善其水溶性問題，並以solvent control 試驗做為品管依據。本研究試驗終點敏感度排序分別為FY(81.82%) > ΔDO(18.18%) > GR，代表觀察藻類細胞顆粒數的變化可以很容易觀察到毒化物所造成的毒性變化。在物種敏感度比較的部分本研究所使用的月芽藻具有相當高的敏感度。而透過藻類毒性試驗可發現取代基Br－提供最大毒性且pyridine環上的鹵素取代基數量越多毒性隨之昇高，同分異構物也因為取代基位置不同而有明顯的毒性差異。 透過QSAR迴歸結果可知本研究化合物對於親脂性參數logKow有良好的相關性(r2=0.82-0.86)，對於電性參數Ehomo的迴歸相關性則顯得不盡理想(r2=0.30-0.33)，而將兩種參數合併迴歸後其相關性仍在可接受範圍內(r2=0.80-0.84)，代表未來可利用本研究之吡啶類QSAR方程式簡單迅速的推估同類化合物，大幅度節省試驗成本及人力資源。 為探討傳統機制分類的準確性，本研究藉由主成分分析(Principal component analysis, PCA)及群集統計分析(Hierarchical Cluster Analysis, HCA)重新歸類pyridine類化合物，分類結果在非極性麻醉當中有四個化合物落在範圍之外，分別為2,3-Dichloropyridine、2,5-Dichloropyridine、3,5-Dichloropyridine以及2,3,5-Trichloropyridine，其中2,3,5-Trichloropyridine在此分析結果屬於反應性物質。因此利用統計方法PCA及HCA可以有系統的將化合物的物化特性結合生物反應，進一步歸類化合物的毒性作用機制，不再像傳統分類法只是依照化合物的結構而進行簡單分類。

This study investigates the toxicity of halogen-substituted pyridines using a closed-system algal toxicity test technique. Three test endpoints (i.e., dissolved oxygen production (ΔDO), final yield and algal growth rate) were applied to examine the toxic effects of various halogen-substituted pyridines. The unicellular green microalga (Pseudokirchneriella subcapitata) was the test organism and 11 halogen-substituted pyridines were the test chemicals. For insoluble chemicals, this research added Acetone in the stock solution and conducted quality assurance with solvent control. The ranking of sensitivity of three endpoints were FY(81.82%) > ΔDO(18.18%) > GR, which means when observing alga cells, the obvious variety can be seen. When comparing the sensitive species, Pseudokirchneriella subcapitata is almost sensitive. Through observing the results in this study, one can figure out the bromo substituent provides more toxicity than other halogen substitutes and the toxicity variation is in accordance to number of substituents and analogues. Furthermore, the results of QSAR regression analysis show that chemicals in this study have a good correlation with logKow (r2=0.30-0.33). In contrast, regression correlation with Ehomo is even less than ideal (r2=0.82-0.86). But when combining these two parameters, its relevance is still within the acceptable range (r2=0.80-0.84). This means that in the future the QSAR model developed by this study can easily and quickly estimate similar compounds, to save more time and money. Another addition to the objective of this research, was to investigate the mechanism of traditional classification accuracy with Principal Component Analysis (PCA) and Hierarchical Cluster Analysis (HCA). The results showed that 2,3-Dichloropyridine、2,5-Dichloropyridine、3,5-Dichloropyridine and 2,3,5-Trichloropyridine did not belong to non-polar chemicals, and where 2,3,5-Trichloropyridine was classified as the reactive substance. Therefore, the use of statistical methods PCA and HCA can systematically combine the physical and chemical properties of the compounds with biological reactions, to further classify modes of toxic action of halo-pyridines.