| Objective: To improve an accurate and comprehensive database for rapid identification of Burkholderia pseudomallei(Bp)from different culture medias and culture time by using MALDI-TOF MS(Matrix Assisted Laser Desorption Ioniztion Time of Flight Mass Spectrometer)through the strains of Burkholderia pseudomallei and other related bacterias collected by Hainan Antimicrobial Resistance Surveillance System,in order to provide basic data for the identification of Burkholderia pseudomallei,and to compare the differences of fingerprints of strains in different medias and different culture time,so as to provide laboratory basis for the correct diagnosis.Methods: The genetic background of the strain was clear by whole genome sequencing,50 strains of Burkholderia pseudomallei identified by molecular biological methods were passaged to BA(Blood Agar),MAC(Mac Conkey),CHA(Chocolate Haemophilus Agar)and CBA(China Blue Agar),and purebred colonies were selected on the first,second and third day of culture,the target plate was loaded into the VITEKMS instrument,and the spectra were obtained by Launch Pad software.The quality control strain was ATCC8739 fresh every day.According to the requirements for the construction of mass spectrometry database(the number of peaks is between 80 ～ 250),1092 pieces of atlas information were selected for the construction and verification of Bp primary reference database.Secondly,the reference spectra of 12 Bp related strains were obtained by the same method,and the two were combined to build an exclusion list which is helpful to eliminate the cross peaks,in order to modify the primary reference spectra database through the above steps.Then,according to the requirements of SOP,we further narrowed the number of peaks in the primary database,the number of peaks with the same quality,strain similarity and other requirements of a total of 384 high-quality maps,and built a modified super spectra library containing exclusion list.Finally,30 Bp and 8 related strains were used as strains to be tested,and the strains were identified by MS mass spectrometer,so as to verify the identification performance of the super spectrum database,and the identification results were obtained and recorded.Results: 50 strains of Burkholderia pseudomallei were isolated from different patients.All the strains used in the experiment can be traced to the source,and all the strains were identified as Burkholderia pseudomallei by WGS(Whole Genome Sequencing).The quality control strains on each target were successfully identified as Escherichia coli ATCC8739.It could be seen that on the first day of culture,peaks 9623-9625 could be seen on the four medias,peaks 9625-9626 could be seen on the second day at the same mass-to-charge ratio,but on the third day,it became peaks 9617-9619,and on the first day of BA culture Peak 5197 could be seen at the same time,but it became smaller or even disappeared in the second and third day,and it was difficult to distinguish.On the first day of culture,no matter which medium the colonies were taken from,it could be seen that the peak 4411 can be clearly expressed in the spectrum.On the second and third day,the peak became smaller and even difficult to capture;on the first and second day of culture,peak 3762 had not been captured,but on the third day,the mycoprotein profiles taken from the four media are all acceptable.It was expressed;on the first day of culture,peak 6226 could be seen in the colonies taken from BA,MAC,and CHA except CBA;on the second day of culture,except CHA,peak 7528 could be captured;At three days,peak 7523 could be seen in the protein profiles generated from the other three colonies except CBA.The mass spectrometer with fully expanded database has a satisfactory accuracy for the identification of Burkholderia pseudomallei.Compared with the gold standard WGS,the specificity,accuracy,sensitivity and negative predictive value of the expanded database for Burkholderia pseudomallei identification are 100%,85.75%,82.33% and57.6%,and the positive predictive value is 100%.The difference between two groups was statistically significant(P < 0.05).The mass spectra of Burkholderia pseudomallei and its related bacterias can be clustered and analyzed by SARAMIS software,and the tree map with obvious cluster clustering can be obtained.Conclusion: 1.Even for the same Bp strain,the protein map will change with the change of culture medium and culture time.Although the selection of Bp colonies cultured on CBA for the third or second day is easier to form a high quality map conducive to the construction of a super map library,it does not prevent the early Bp colonies from being accurately identified to the species level.2.SARAMIS software can be used to cluster the mass spectra of Bp and its related strains to get a tree map,and it can be seen from the map that even the strains belonging to different species have obvious cluster clustering.3.After the database has been fully expanded,the mass spectrometer can identify Bp bacteria more accurately than before.The improved database can be applied to the clinical practice of identifying Bp on a large scale. |
