【作者】 虞锐鹏；

【导师】 宋启军；

【作者基本信息】 江南大学 ， 化学工程与技术， 2022， 博士

【摘要】 湖泊富营养化是当今世界主要的生态问题之一。过快城镇化、工业化、过度农业养殖等人为因素,使得水体营养物质含量上升,进而导致了蓝藻水华发生频次的增加。蓝藻水华产生多种异味化合物和毒素,对各类水体（特别是饮用水水源）的水质造成直接威胁。然而,现存的水体评价指标不够全面和准确,不能真实反映有毒有害物质具体含量。本研究采用代谢组学研究技术,分析了蓝藻水华（微囊藻为优势藻种）产生的次生代谢产物及其对水质的影响并应用于水质评价。首先,为解决水体评价指标不全面、不准确的问题,使用代谢组学和统计学方法研究了不同时期微囊藻样品中的挥发性和水溶性代谢产物,筛选出统计学差异代谢产物。本研究建立了基于顶空-固相微萃取箭型-气相色谱-质谱联用技术,对不同时期微囊藻样品中挥发性代谢产物进行定性分析,鉴定得到210种挥发性代谢产物。结合多变量统计分析,筛选出10种统计学显著性差异代谢产物:环己醇、二甲基三硫、苯甲醇、樟脑、2-甲氧基苯酚、3-己烯-1-醇、2,4-癸二烯醛、3-甲基吲哚、柠檬醛和正壬醇。优化了固相微萃取箭型萃取温度、萃取时间、解析温度、解析时间,体系盐度等萃取条件,进行定量分析和方法学验证。结果表明,此10种差异代谢产物在0.050 ng/L～1000 ng/L范围内具有良好的线性关系,相关系数>0.998,检出限范围为0.010 ng/L～0.030 ng/L,平均回收率在76.3%～93.0%之间。针对不同生长时期的微囊藻样品,研究建立了UPLC Q-TOF MS非靶向代谢组学分析,运用多种代谢组工具对原始数据发掘分析,对不同生长时期、不同环境的微囊藻之间水溶性差异代谢产物综合分析,鉴定发现从高峰期到衰亡期,细胞内藻毒素（MCs）有着显著降低。使用Progenesis QI对采集的样本数据进行分析,共得到5896个带有详细碎片和峰面积等信息的组分。设定合适的筛选参数与Chemspider天然产物全库中化合物的信息进行匹配,最终获得了24个具有显著差异且定性可靠的潜在标志物。在差异代谢产物间相关性分析中,发现MCs和焦脱镁叶绿素a、金盏花红素表现出高度相关。对筛选得到的水溶性差异代谢物MCs进行定量分析,优化固相萃取条件,建立了微囊藻水华样品中藻毒素（共13种）的检测方法。结果表明,通过固相萃取柱净化处理,减少基质的干扰,该方法可用于实际藻样测定。通过自建藻毒素谱库,发现九种疑似微囊藻毒素化合物。其次,为提高代谢产物测定的灵敏度和稳定性,基于筛选出的统计学差异代谢产物,选取特征化合物3-甲基吲哚（3-MI）作为模板分子,以邻苯二胺为单体,采用电聚合法制备了3-MI分子印迹电化学传感器,大大降低了3-MI的检出限和分析成本。基于密度泛函理论方法计算的分子印迹聚合物计算机辅助设计,评价不同功能单体与模板分子3-MI的相互作用。筛选采用邻苯二胺为单体,3-MI为模板,根据电聚合法制备了3-MI分子印迹电化学传感器。研究了聚合电位、聚合介质、聚合扫描圈数及模板去除方式对印迹电极性能的影响。用K₃[Fe（CN）₆]作为电化学探针,采用间接分析法对3-MI进行分析,线性范围为0.010μM～1.2μM,检测限为4.0 n M。制备的传感器在结构相似分子的干扰下,对模板分子具有良好的识别能力和高选择性。该传感器具有良好的重复性和稳定性,成功用于实际藻水样品中的3-MI测定。第三,为验证差异性代谢物与传统评价因素一致性及优越性,将差异性代谢物分析应用于实际水体样品中的分析,探讨了微囊藻水华代谢产物产生的时空变化趋势。结果表明:通过对不同时期差异性代谢物含量的分析,分析者可以得出该水体藻类生长或污染的情况,初步证明该技术可以用于天然湖水水质评价。2019年3月至11月整个采样监测周期间,研究发现,3-甲基吲哚与叶绿素a、藻类密度、水温等水质参数呈显著正相关;吲哚、β-环柠檬醛与叶绿素a、藻类密度呈显著正相关。游离藻毒素及胞内藻毒素与总磷、叶绿素、藻类密度等水质参数呈显著正相关。为了更全面地评价水质,需要进一步对特征代谢产物进行监测和综合评价。最后,将基于特征代谢产物的分析手段应用于实际样品（鱼肉和蓝藻填埋场渗滤液）检测中,优化了预处理方法,实验结果显示,该方法表现出良好的定量分析能力。文章建立了一种鱼肉中蓝藻异味代谢物的检测方法。通过优化微波蒸馏及固相微萃取等提取富集条件,采用全二维气相色谱分离,有效实现蓝藻中常见的8种异味化合物的完全分离,以高分辨特征离子的精确质量数进行内标法定量分析。8种异味化合物在0.01μg/L～100μg/L质量浓度范围内呈良好线性关系,方法的检出限为0.070μg/kg～3.3μg/kg。此研究中,以二乙烯基苯、多孔活性碳粉和聚二甲基硅氧烷为原料,采用一锅法制备了SPME材料。通过表征自制SPME纤维涂层并优化萃取条件,与进口商用产品相比,自制SPME纤维涂层能兼顾VOCs的萃取,尤其能提高含硫异味物质的萃取效能。结合GC-MS对蓝藻填埋场渗滤液水质进行分析,考察渗滤液处理站对恶臭物质的去除效率。更多还原

【Abstract】 Lake eutrophication is one of the main ecological problems in the world.Human factors,including rapid urbanization,industrialization and excessive agricultural cultivation,lead to nutrient content rise in water body,further increasing the frequency of cyanobacteria bloom.Cyanobacteria produce a variety of toxins and odor compounds,which pose a direct threat to the supply water,especially drinking water.However,the existing water evaluation indicators focus on physical property indicators,yet the chemical indicators are relatively general,which cannot fully and accurately reflect the specific content of toxic and harmful substances.In this study,metabolomics research technology was used to analyze the secondary metabolites produced by cyanobacteria bloom and their impact on water quality,which is applied to water quality evaluation.Firstly,in order to solve the problem of incomplete and inaccurate water evaluation indexes,the volatile and water-soluble metabolites in Microcystis samples at different periods were studied by metabolomics and statistical methods,and the statistically different metabolites were screened out.A novel headspace solid phase microextraction Arrow method combined with gas chromatography-mass spectrometry was developed for extraction and analysis of volatile organic compounds of microcystis under simulated natural conditions.By selecting suitable extraction fibers,210 kinds of volatile metabolites were identified by full scanning qualitative analysis from the samples of Microcystis in producing period.The metabolic changes of volatile molecules during the growth of Microcystis were analyzed by multivariate statistical analysis and clustering Analysis.Chemometric analysis based on principal component analysis,partial least squares-discriminant analysis and heatmap associated with hierarchical cluster analysis provided a suitable tool to differentiate volatile organic compounds in algal material.And 10metabolites in statistical significance were identified,including cyclohexanol,dimethyl trisulfide,benzenemethanol,camphor,2-methoxyphenol,3-hexene-1-ol,2,4-decadienal,3-methylindole,citral and 1-nonanol.Parameters affecting the extraction,such as extraction temperature,time and salt were optimized.In addition,the analysis conditions,including desorption temperature and time as well as gas chromatographic parameters,were optimized.The calibration curves showed a good linearity（correlation coefficient>0.998）in the concentration range of 0.050ng/L～1000 ng/L.The detection limits of this method were 0.010 ng/L～0.030 ng/L,and the recoveries at the concentration level of 100 ng/L were 76.3%～93.0%.The method was satisfactorily precise,with the RSD values of less than 12.7%.The new method is simple,quick,stable,and applicable to complex matrices.It is suitable to the determination of volatile metabolites in natural water at the early stage of blue-green algae bloom.UPLC Q-TOF MS non-target metabolomics analysis was established for four groups of Microcystis samples in different growth periods.The original data was analyzed with various metabolomics tools.The comprehensive evaluation of different metabolisms of Microcystis in different growth periods and different environments showed that MC-RR and MC-LR decreased significantly from the peak to the aging period.Using the latest metabolomics software,the author analyzed the collected sample data and got 5896 components with detailed fragment and peak area information.The screening parameters were set:parent ion matching error,fragment matching error,isotopic similarity,fragment ion fraction,total score of compound identification and ANOVA（p）value.Matched with the information of the compounds in the whole natural product library of Chemsipider,26potential markers with significant difference and qualitative reliability were obtained.In the correlation analysis of the different metabolites,the author found that there was a high degree of consistency between the algal toxin and pyropheophyll A and calendula.The target analysis of MCs was carried out and the solid phase extraction conditions were optimized.The detection methods of 13 microcystins toxins in cyanobacteria were established.Results show that the matrix has little interference to the method,which can be used for the determination of algae samples.Nine suspected microcystin compounds were discovered by self-built microcystin library.Secondly,in order to improve the sensitivity and stability of metabolite determination,based on the selected statistically different metabolites,the 3-MI Molecularly Imprinted Electrochemical sensor was prepared by electropolymerization.The characteristic compound 3-methylindole was selected as the template molecule and o-phenylene was used as the monomer.The electrochemical sensor greatly reduced the detection limit and analysis cost of 3-MI.3-MI is a recalcitrant and toxic substance that can cause damages to human and animal lung cells and may have a pronounced effect on mood and stress levels.Rapid and sensitive detection of 3-MI is therefore of great important for its environmental safety evaluation.A novel electrochemical sensor for the determination of 3-MI was fabricated based on electropolymerization of molecular imprinted polymer film on a glassy carbon electrode.Rational selection of the suitable monomers was conducted by evaluation of the interaction between each monomer and template using the density functional theory method.The results indicated that the stabilization energy of 3-MI-o-Phenylenediamine complex is greater than that obtained from other evaluated monomers,hence o-PD was used to prepare the MIP sensor.The imprinted layer was systematically characterized by cyclic voltammetry,electrochemical impendence spectroscopy scanning electron microscopy and atomic force microscopy.Under optimal experimental conditions,a linear relationship was observed between the peak current change of the electrochemical probe[Fe（CN）₆]^3-and the concentration of 3-MI over the range0.010μM～1.2μM,with a detection limit of 4 n M.This sensor also exhibited superb selectivity to 3-MI as minimum interferences were observed from other structurally similar compounds,thus it was successfully applied for the determination of 3-MI in real samples.Thirdly,in order to verify the consistency and superiority of differential metabolites with traditional evaluation factors,the analysis of differential metabolites was applied to the analysis of actual water samples,and the production and temporal and spatial variation trend of cyanobacteria bloom metabolites were discussed.The results show that through the analysis of differential metabolites,the analyst can obtain information about the algae growth or pollution of the water body,which preliminarily proves that thistechnology can be used to evaluate the water quality of natural lakes.During the sampling and monitoring period from March to November 2019,it was found that there was a significant positive correlation between 3-methylindole and water quality parameters such as chlorophyll a,algal density,water temperature.There was a significant positive correlation between indole,β-cyclocitral and Chl-a,algal density.Free and intracellular microcystins were positively correlated with total phosphorus,chlorophyll and algae density.Finally,the analysis method based on differential metabolites was applied to the actual samples（fish flesh and cyanobacteria landfill leachate）,and the pretreatment method was optimized.The experimental results showed good quantitative analysis ability.A method for the detection of cyanobacteria odor metabolites in fish was established.By optimizing the extraction and enrichment conditions such as microwave distillation and solid-phase microextraction,the complete separation of eight common odor compounds in cyanobacteria was effectively realized by full two-dimensional gas chromatography,and the internal standard method was used for quantitative analysis with the accurate mass number of high-resolution characteristic ions.Eight odor compounds in the range of 0.010μg/L～100μg/L showed good linearity.The detection limit of the method is 0.070μg/kg～3.3μg/kg。Using divinylbenzene,porous activated carbon powder and polydimethylsiloxane as raw materials,SPME materials were prepared by one pot method.The home-made SPME fiber coating was characterized and extraction conditions were optimized.Compared with imported commercial products,the home-made SPME fiber can not only extract VOCs,but also enhance the extraction efficiency of sulfur-containing odor substances.Combined with GC-MS,the water quality of leachate from cyanobacteria landfill was analyzed to investigate the removal efficiency of odor substances in leachate treatment station.更多还原