摘要
以分子电性距离矢量(MEDV-13)和原子类型电拓扑状态指数(ETSI)2种分子结构描述符,分别有效表征了15个酚类雌性激素的分子结构,应用基于预测的变量选择与模型化(VSMP)方法分别建立了该类化合物经辣根过氧化物酶(HRP)催化氧化的速率常数(lnkCAT)与分子结构的定量相关模型.以MEDV-13为分子描述符建立的模型显示:影响去除效果的主要结构因素是由5个MEDV描述子表达的5个子结构碎片,即—CH3、—CH2、—O—、—OH和—cCc.以ETSI为分子描述符所建模型表明:影响该反应速率的主要结构因素是由5个ETSI描述子对应的5个子结构碎片,即—CH3、—CH2、—O—、—OH和>C< .所建模型的估计相关系数分别为0.92和0.93 , LOO检验相关系数分别为0.84和0.85,表明2种模型均具有良好的估计能力与稳健性.以上结果说明,虽然酚类雌激素的种类、结构多样,但决定其HRP酶催化去除效果的关键结构因素是酚环上所有取代基整体给电子能力的强弱.
Molecular electronegative distance vector （MEDV-13 ） and electrotopological state index （ETSI） were employed to describe the molecular structures of 15 estrogenic phenolic chemicals. Quantitative linear relationships were then developed between the two sets of descriptors and the rates of horseradish peroxidase （HRP）-mediated enzymatic reactions with the 15 chemicals using variable selection and modeling based prediction （VSMP）. The model constructed using MEDV-13 descriptors showed that the main structural factors influencing the reaction rates of the estrogenic phenolic chemicals were attributed to such substructural features as -CH3 , -CH2 , -O-, -OH and -cCc （where ＂ c＂ refers to a ehemieal bond in the aromatic ring） and the mode/ tructed by ETSId iptors indicated that-CH3, -CH2, -O-, -OH and 〉C〈 （where ＂ 〈 ＂and ＂ 〉 ＂refer to two single bonds） are important influential features. The two best 5-variable models with a calibrated correlation coefficient of r≥0.92 and a validated correlation coefficient of q≥0. 84, were established by multiple variable linear regression. The models exhibit satisfactory prediction abilities and stabilities. The results indicate that electron-donating substituent groups on phenolic rings are the key factor controlling HRP-mediated enzymatic reactions, although the chemicals under study have diverse structures.
出处
《环境科学学报》
CAS
CSCD
北大核心
2008年第12期2562-2567,共6页
Acta Scientiae Circumstantiae
基金
国家自然科学基金资助项目(No.20677024)~~
关键词
内分泌干扰物
辣根过氧化物酶
酶催化反应
定量结构-活性相关关系
endocrine disrupting chemicals
horseradish peroxidase
peroxidase-mediated reactions
Quantitative Structure-Activity Relationship
