Objective: To establish mouse models of Candidemia, and investigates statistically significant polypeptide peaks to provide auxiliary diagnosis of this disease. Methods: A total of 170 specific pathogen free adult male ICR mice with body mass of 27-30 g were completely randomly divided into Candida albicans infection group (n=80), Candida parapsilosis infection group (n=80) and the normal control group (n=10), and the two kinds of Candidemia mouse models were established via tail vein injection. The serum samples were analyzed by Matrix-assisted laser desorption-ionization time of flight mass spectrometry and relevant software, and the polypeptide peaks with significant differences were screened to establish diagnostic models. Results: A total of 65 differential polypeptide peaks were obtained compared with the Candida albicans infection group and the normal control group. Combined with m/z 1 100.4, 1 581.0, 3 808.0 as differential polypeptide peaks to established the diagnostic model, the sensitivity was 95.24%(40/42), the specificity was 90.63%(29/32), the accuracy rate was 93.24%(69/74), and the AUC value of the ROC curve was 0.972(95%CI: 0.941-1.000). A total of 73 differential polypeptide peaks were obtained compared with Candida parapsilosis infection group and the normal control group. Combined with m/z 1 433.2, 1 148.5, 4 093.5, 4 522.2, 8 140.9, 8 234.6 as differential polypeptide peaks to established the diagnostic model, the sensitivity was 95%(38/40), the specificity was 81.25%(26/32), the accuracy rate was 88.89%(64/72), and the AUC value of the ROC curve was 0.953(95%CI: 0.903-1.000). A total of 78 differential polypeptide peaks were obtained compared with Candida albicans infection group and Candida parapsilosis infection group. Combined with m/z 2 736.9, 8 091.5, 8 153.7 as differential polypeptide peaks to established the diagnostic model, the accuracy of distinguishing C. albicans infection from C. parapsilosis infection was 98.78%(81/82). Conclusions: Successfully screened the differential polypeptides and established the related diagnostic models. Which is helpful to find serum biomarkers for the auxiliary diagnosis of Candidemia, and provides a basis for the early diagnosis and the rational use of drugs.
目的： 建立念珠菌血症小鼠模型，寻找辅助诊断念珠菌血症的差异多肽峰。 方法： SPF级ICR雄性小鼠170只，体质量27~30 g，根据随机数字表法将小鼠完全随机分为白色念珠菌感染组（n=80）、近平滑念珠菌感染组（n=80）和对照组（n=10），通过尾静脉注射的方法建立念珠菌血症小鼠模型。利用基质辅助激光解析电离飞行时间质谱检测模型小鼠血清多肽谱，选择差异较为明显的多肽峰建立诊断模型。 结果： 比较白色念珠菌感染组和对照组共得到65个差异多肽峰，组合质荷比（m/z）为1 100.4、1 581.0、3 808.0这3个差异多肽峰建立诊断模型，敏感度为95.24%（40/42），特异度为90.63%（29/32），准确率为93.24%（69/74），ROC曲线下面积为0.972（95%CI：0.941~1.000）；比较近平滑念珠菌感染组和对照组共得到73个差异多肽峰，组合质荷比（m/z）为1 433.2、1 148.5、4 093.5、4 522.2、8 140.9、8 234.6这6个差异多肽峰建立诊断模型，敏感度为95%（38/40），特异度为81.25%（26/32），准确率为88.89%（64/72），ROC曲线下面积为0.953（95%CI：0.903~1.000）。比较白色念珠菌感染组和近平滑念珠菌感染组，共得到78个差异多肽峰，组合质荷比（m/z）为2 736.9、8 091.5、8 153.7这3个差异多肽峰建立诊断模型，区分白色念珠菌感染和近平滑念珠菌感染的准确率为98.78%（81/82）。 结论： 通过念珠菌血症小鼠模型筛选的差异多肽峰有利于寻找辅助念珠菌血症诊断的血清标志物，为临床早期诊断及合理用药提供了依据。.
Keywords: Candidemia; Diagnostic model; Matrix-assisted laser desorption-ionization; Peptide fingerprinting; Serum biomarkers.