蛋白芯片技术检测肺炎克雷伯菌产ESBLs菌株的研究

Study on the Detection of Klebsiella pneumoniae ESBLs Producing Strains by Protein Chip Technology

目的建立肺炎克雷伯菌（简称肺克）产ESBLs菌株蛋白质组诊断模型，寻找快速鉴定产ESBLs肺克的方法。方法采用PBSⅡ／C型蛋白质指纹图谱仪及Au蛋白芯片筛选肺克产ESBLs菌株差异表达蛋白，通过人工神经网络（ANN）建立并验证肺克产ESBLs菌株的蛋白质谱模型。结果相对分子量为2000--20000范围内共检测到肺克的47个蛋白质峰，用三个质荷比分别为5985．5、8361．6、9169．8的蛋白质峰建立的肺克产ESBLs菌株蛋白质谱模型能够区分肺克产ESBLs菌株和不产ESBLs菌株，模型经盲法验证灵敏度为96.88％、特异度为93．62％。结论本研究建立的蛋白质谱模型，能够准确和快速的鉴定肺克产ESBLs菌株，值得深入研究。

Objective To establish the proteomic diagnostic model for Klebsiella pneumoniae strains producing extended spectrum beta lactamases （ESBLs） and scarch for a rapid method to identify them. Methods Klebsiella pnewumoniae pro reins were detected by the PBS Ⅱ/C protein fingerprint spectrometer and Au protein chip. The differentially expressed prolcins of Klebsiella pneumoniae ESBLs positive and negative strains were analyzed and an artificial neural network （ANN） model of Kileb- siella pneumoniae ESBLs producing strains was established and validated. Results Totally, 47 protein peaks were detected for the moleculars with mass- to charge ratio （m/z） within the range of 2,000 20,000. The peaks with m/z of 5,985.5, 8,361.6 and 9,169.8 were screened by ANN as the protein spectrum model of Klebsiella pneumoniae ESBLs producing slrains. The validation result of the model indicated that the sensitivity was 96.88% and the specificity was 93.62%. Conclusions The protein spectrum model established by this study can identify Klebsiella pneumoniae ESBLs producing strains accurately and rapidly. It is worth further studying.