淡水河流域腸道病毒之分子監測與預測

MOLECULAR MONITORING AND PREDICTION OF ENTERIC VIRUSES IN DANSHUEI RIVER, TAIWAN

張博任(Po-Jen Chang)

指導教授 ： 陳建先

大同大學/工程學院/生物工程學系所/碩士(2009年)

摘要

淡水河為北台灣主要河川，上游有石門、翡翠水庫供給大台北民生用水，中游有污水處理廠排放口，下游為渡船碼頭及海水浴場供民眾遊玩。腸道病毒主要是以糞口途徑傳播，當人接觸或食入受污染的水或食物，則會增加感染的風險，因此有必要對水環境進行調查與監測。本研究室自2005年起，監測淡水河下游河水及北台灣海水，皆分析出腸道病毒，因此本研究將擴大監測2006年5月至2007年12月間淡水河流域上、中、下游之腸道病毒 (腸病毒、腺病毒、輪狀病毒及諾羅病毒)，及探討環境因子對於腸道病毒分佈之影響並以生物資訊學的方法探討腸道病毒在淡水河流域中演化的意義。
近年類神經網路 (Artificial Neural Networks, ANNs)被應用在各個不同科學領域，其優點為能處理因子間複雜、有相互關係且非線性的問題。因此本研究將先前調查淡水河流域中腸道病毒的數據，包括物化因子 (溫度、鹽度、導電度及pH)、指標微生物 (E.coli、Total coliform、Fecal coliform、Enterococcus)及病毒，利用類神經網路建立一預測架構，預測水體中腸道病毒的存在與否。
本研究將收集河水樣本以吸附─沖提法(Adsorption-Elution method)將2 L濃縮至2 mL，再以RT-PCR及Seminested-PCR檢測樣本中之腸道病毒，結果腺病毒有20.4% (71/348)，其次為諾羅病毒11.2% (39/348)，輪狀病毒有4.6% (16/348)，最低為腸病毒0.9% (3/348)，顯示淡水河流域確實受到腸道病毒污染，且以腺病毒污染最嚴重。在淡水河流域中以中、下游地區為高度污染區。季節分佈方面，腺病毒在全年皆有檢出情形，諾羅病毒與輪狀病毒有同時檢出情形，在採樣期間有3個檢出高峰，分別在2006年5-8月、2007年1-5月及2007年9-12月。
環境因子方面以水溫對腺病毒及pH對諾羅病毒在淡水河流域有統計上相關意義，而指標微生物與腸道病毒間無統計相關；腸道病毒間只有諾羅病毒與輪狀病毒間有統計上相關意義。本研究經核酸定序後發現腺病毒有ADV-3、ADV-41、ADV-40、CADV-2，諾羅病毒有NV-GII.4，輪狀病毒有RV-G1、RVG3、RVG9分佈在淡水河流域中，又經由分子演化分析後顯示與台灣、日本、韓國及中國大陸等東亞區域共同循環傳播。
在預測方面，經過試誤法測試後，腺病毒以Model 6 (日期、溫度、pH、E.coli) 的預測架構有最好的預測準確，而諾羅病毒以Model 5 (日期、溫度、pH、Total coliform) 的預測架構有最好的預測準確率，其預測準確率分別為79.6%及85.9%。綜合以上結果，淡水河確實受到腸道病毒污染，且以腺病毒最嚴重；類神經網路可適用於預測淡水河中腸道病毒存在，對腺病毒可選取日期、溫度、鹽度、pH及E.coli，而諾羅病毒可選取日期、溫度、鹽度、pH及Total coliform建立預測模式，其預測準確率約為80%。

關鍵字

類神經網路 ； 分子監測 ； 淡水河 ； 腸道病毒

並列摘要

The Damshui River is the main source of drinking water in Taipei and is also used for recreational purposes, especially in summer. The enteric viruses generally spread through the fecal-oral route. When people contact the contaminated food or water, it would increase the public health risk. Therefore, it is necessary to monitor water environment. Since 2005, our research team has monitored enteric viruses at the downstream Damshui River and the coastal water of northern Taiwan, and the results have shown that enteric viruses were detected. The aim of this study is to investigate the distribution of enteric viruses (including enterovirus, adenovirus, rotavirus and norovirus) in Damshui River watershed during May 2006 to December 2007, and the relationship between environmental parameter and enteric viruses.
In recent years, Artificial Neural Networks (ANNs) had been used increasingly in various fields of science and technology, which could handle complex, inter-related and non-linear relationships between multiple parameters. Therefore, we collected the foregoing data, including numerical knowledge (temperature, salinity, conductivity and pH value), Heuristic knowledge (date and sampling site) and biological knowledge (E.coli, Total coliform, fecal coliform and enterococcus). Another aim of this study is to use ANNs to define input parameters that predict the presence or absence of enteric viruses in surface water impacted by multiple parameters related to water quality.
This study adopts adsorption-elution method to concentrate 2 L water to 2 mL, then use RT-PCR and Seminested-PCR detect the enteric viruses in water sample. Our results have indicated that 20.4% (71/348) of samples showed positive for adenovirus, 11.2% (39/348) for norovirus, 4.6% (16/348) for rotavirus, and 0.9% (3/348) for enterovirus. It proves that Damshui River watershed was contaminated by enteric viruses, and the most acute kind of viruses is adenovirus. As to geographical distribution, the most seriously contaminated area were midstream and downstream. As to seasonal distribution, adenovirus were detected all year round, while norovirus and rotavirus were detected during the same periods, including May to August 2006, January to May 2007, and September to December 2007.
As for environmental factors, the temperature was shown significantly correlated with adenovirus, and pH value was shown significantly correlated with norovirus. However, indictor bacteria were not significantly correlated with enteric viruses. As to relationships among enteric viruses, only norovirus and rotavirus had significant correlation. Through sequencing analysis of the Damshui River samples, it revealed the presence of ADV-3, ADV-41, ADV40, CADV-2, NV-GII.4, RV-G1, RV-G3, and RV-G9. We adopted phylogenetic analysis to compare the nucleotide sequence of enteric viruses, and the result showed that the nucleotide sequence was propagated in East Asia. As for ANNs prediction, the results showed adenovirus had the best accuracy in Model 6 (date, temperature, pH value and E.coli) and norovirus had the best accuracy in Model 5 (date, temperature, pH value and Total coliform), with total percent accuracy of 79.6% and 85.9%, respectively.
According to the results of this study, Damshui River watershed was contaminated by enteric viruses, and the ANNs could predict the enteric viruses in water.

並列關鍵字

Molecular monitoring ； Danshuei River ； Enteric virus ； Artificial Neural Networks

參考文獻

88. 江大雄、賴幸湀、陳國東、王躬仁. 1999. 腸病毒的流行病學概論.疫情報導. 第15卷. 第3期.

89. 李君男. 2003. 急性腸胃炎致病病毒性因子之探討. 行政院衛生署疾病管制局九十二年度科技研究發展計劃.

1. Abad, F.X., R. M. Pinto, A. Bosch. 1994a. Survival of enteric viruses on environmental fomites. Appl. Environ. Microbiol. 60:3704–3710.

3. Atmar RL, and Estes MK., 2001. Diagnosis of noncultivatable gastroenteritis viruses, the human caliciviruses. Clin Microbiol Rev. 14:15-37.

4. Auerbach, J., D. Prager, S. Neuhaus, U. Loss, and K. H. Witte, 1994. Grouping of porcine enteroviruses by indirect immunofluorescence and description of two new serotypes. J. Veter. Med. Series B 41:277-282.