Identification Of Clostridium Botulinum By Confocal Raman Microscopy

Clostridium botulinum will release the Botulinum toxin in the process of reproduction,and the Botulinum toxin is known to be the most toxic biological toxin among the natural and synthetic toxins.It can cause bacterial food poisoning through canned food,vacuum packaged food and sealed pickled food.It is harmful to human health and threatens public safety.The conventional detection method is to use biological detection method after anaerobic separation and cultivation in laboratory.The operation of this method is very complicated,and the detection time is very long(usually more than 10 days),which is not conducive to timely risk control.Therefore,the research on fast and accurate identification method of Clostridium botulinum is an important issue to be solved in the field of disease control.In recent years,Raman spectroscopy can reflect the information of material composition at the molecular level and has characteristic signals.Because of its non-destructive,label-free,high efficiency and simple operation,it has been widely used in qualitative research.Bacterial identification is also a research hotspot these years.The aim of this paper is to construct a confocal Raman spectroscopy system for bacterial identification,and to establish a rapid identification method of Clostridium botulinum by confocal Raman spectroscopy combined with chemometrics.In this study,a Clostridium botulinum B strain was isolated from a stool sample of a suspected infantile botulinum poisoning patient.The strain was identified by bacterial culture and routine identification methods,which provided research samples for Raman experiment.In this paper,the Raman spectrum acquisition experiment method,the spectral information processing method are studied by using the Micro-confocal Raman spectrum system.We used two methods to eliminat the baseline drift and fluorescence background ofRaman spectrum.Finally,we study theclustering and recognition method of bacterial spectrum.The main results are as follows:(1)Raman spectrum acquisition method:Raman spectra of glass,quartz and BaF2films were collected by confocal microscopic Raman spectroscopy system.The results show that glass has strong fluorescence background signal,followed by quartz,which has a serious interference effect on bacterial Raman spectrum signal.BaF2 has no background fluorescence signal except for its sharp characteristic peak at 243 cm-1,so it is chosen as the base material for collecting bacterial spectrum.Through the experiments of different bacterial spectrum acquisition,it is found that the experimental parameters such as integration time and laser power intensity have important influence on the spectral intensity,and the optimal parameters should be based on the type of different bacterial strains.Generally,Gram-positive bacteria need higher laser power and longer integration time,whereas Gram-negative bacteria need the opposite.(2)Spectral pretreatment methods:Bacteria and other biological samples have strong Raman fluorescence background.In addition,due to the random variation of coating thickness,bacterial individuals,laser irradiation time and intensity on the slide,there is a serious baseline drift between Raman spectra collected.Fifty-one bacterial spectra of Shigella flexneri were collected repeatedly,and then be processed by standard normal transformation(SNV),first derivative,fluorescence fading effect difference method(new method)and airPLS method,respectively.The results show that SNV and first derivative can effectively eliminate baseline drift.The difference method of fluorescence fading effect can significantly eliminate the baseline drift and fluorescence background of Raman spectrum,besides,it can effectively enhance the Raman spectrum characteristics of bacteria.The airPLS method can significantly eliminate baseline drift,but can not completely eliminate its fluorescence background.Four pretreatment methods can improve the spectral repeatability.(3)Identification methods:Seven strains including E.coli TOP 10,E.coli EDL933,Shigella flexneri,Shigella sonnei,Luteimonas terricola,Clostridium perfringens and Clostridium difficile were prepared as research objects,and 354 Raman spectra of strains were collected repeatedly.Firstly,all the Raman spectra of the seven bacteria were analyzed by principal component analysis(PC A).The scores of the first three principal components were used to plot,i.e the bacterial spectra were projected into the principal component spectral space,and their Gram type levels were studied.The original spectra and the spectra preprocessed by SNV,first derivative,fluorescence fading effect difference method and airPLS respectively were clustered.And the bacterial spectrum can be completely separated in PCA space at the level of Gram type.Then PCA was used to cluster again at the generic level,but there was a serious overlap between classes.At the same time,it was found that the fluorescence background was beneficial to bacterial classification.Further research shows that the correct recognition rate of the genus and strain models is 100%by using SIMCA pattern recognition method after data pretreatment by Savitzky-Golay smoothing,airPLS and SNV methods.(4)Identification of Clostridium botulinum by Raman spectroscopy:Clostridium botulinum type B was isolated from the stool samples of a suspected infantile botulinum poisoning patient and identified by the routine identification methods.Fifty Raman spectrum of Clostridium botulinum were collected from different batches of cultured Clostridium botulinum bacteria on BaF2 slide at different locations by confocal Raman spectroscopy.The Raman characteristic peak of Clostridium botulinum were found at 726 cm-1,780 cm-1,852 cm-1,1001 cm-1,1328 cm-1,1443 cm-1,1571 cm-1,1667 cm-1 and other places.Comparing with other kinds of bacteria,it is found that the Raman spectra of bacteria are very similar,and there are minor differences between them.The identification model of Clostridium botulinum was established based on SIMCA and confocal Raman spectroscopy,and the recognition accuracy of Clostridium botulinum was 91.67%.Confocal micro Raman spectroscopy system was established in this paper.And Raman spectroscopy could achieve accurate and rapid identification of Clostridium botulinum combined with and SIMCA.This method has the advantages of convenient operation and rapidity compared with the traditional detection method.