MALDI-TOF MS技术快速区分携带rmpA2毒力基因的高毒力肺炎克雷伯菌

Rapid detection of rmpA2-carrying hypervirulent Klebsiella pneumoniae by MALDI-TOF MS

目的 评价基质辅助激光解析/电离飞行时间质谱（MALDI-TOF MS）技术快速区分携带rmpA2毒力基因的高毒力肺炎克雷伯菌的能力.方法 对浙江大学医学院附属第二医院和河南省人民医院的57株肺炎克雷伯菌,PCR扩增rmpA2毒力基因和荚膜血清型基因,多位点序列分型方法（MLST）进行分型,拉丝试验进行高黏液表型测定.MALDI-TOF MS对肺炎克雷伯菌进行鉴定并对峰图进行二维分析和算法统计,包括采用支持向量机算法（SVM）,遗传算法（GA）,监督使神经网络算法（SNN）和快速分类算法（QC）进行数据分析并建立相应模型,得到分型区分的特征峰.结果 57株肺炎克雷伯菌中有28株携带rmpA2毒力基因,MLST均为ST11型,其中拉丝试验阳性有5株;29株不携带rmpA2毒力基因,主要ST型为ST11（n=23）,还包括ST15（n=3）,ST1、ST76和ST473各一株.质谱将肺炎克雷伯菌较清晰地划分成两个区域,ClinProTools软件可准确区分84.2%（48/57）的菌株.ClinProTools软件四种算法结果基本相似,SVM特异性和灵敏度最高,分别为93.23%和100%.采用ClinProTools对质谱峰进行统计显示,得到区分rmpA2基因阳性组和阴性组的2个特征峰分别为7168.9和7280.76,在峰强度上存在一定的差异.结论 MALDI-TOF MS快速区分携带rmpA2毒力基因的高毒力肺炎克雷伯菌方法灵敏度和特异性都在90%以上,得到两个特征峰在峰强度上存在一定差异,需进一步验证.

Objective To evaluate the ability of matrix-assisted laser desorption/ionization time-of- flight mass spectrometry （MALDI-TOF MS ） in identifying the rmpA2-carrying hypervirulentKlebsiella pneumoniae. Methods A total of 57nonduplicateK1ebsiella pneumoniae isolates were collected from the Second Affiliated Hospital of Zhejiang University and Henan Provincial People＇s Hospital. Virulence gene rmpA2 and capsule K serotype-specific genes were detected by PCR, muhilocus sequence typing （MLST） was performed for molecular typing, and string test was conducted to identify the hypermucoviscous phenotype. The ClinProTools software was used for peak analysis. Four standard algorithms, including support vector machine （SVM）, genetic algorithm （GA）, supervised neural network （SNN）, and quick classifier （ QC ） , were tested for their power to differentiate between rmpA2-positive and rmpA2-negative strains. Results Among the 57 Klebsiella pneumoniae isolates, 28 isolates belonged to sequence type （ST） 11 and carried the virulence gene rmpA2, of which 5 isolates were positive for string test ; while the other 29 isolates were not detected rmpA2,and MLST divided them into five STs including ST11 （n = 23）, ST15 （ n = 3 ） , ST1 （ n = 1 ） , ST76 （ n = 1 ） , and ST473 （ n = 1 ）. The results of four standard algorithms were similar, while the sensitivity and specificity of SVM was the highest at 93.23% and 100% respectively. Analysis results of ClinProTools software suggested that the specific peaks for differentiating the rmpA2-positive and rmpA2-negative strains were 7 168.9 and 7 280.76, however, the peak intensity of themwere variant. Conclusion The sensitivity and specificity of MALDI-TOF MS for rapid identification of rmpA2- carrying hypervirulentKlebsiella pneumoniae werehigher than 90% and there was a difference on the peak intensity of two specific peaks, which needs further study.