| Vibrio parahaemolyticus (Vp) is a halophilic gram-negative bacteriumthat exists in45.7%of seafood products. Food poisoning caused by V.parahaemolyticus is the top one among bacterial food poisoning events in thecoastal areas of China. It is one of the most important pathogens in the fieldof food safety. Heating, freezing, or drying are commonly used to controlbacterial contamination and eliminate pathogens in food industry. After thesetreatment, microorganisms may be killed, survived (non-injured), orsublethally injured. Under suitable conditions, injured microorganisms canrepair themselves and become active and pathogenic, thus posing a seriousthreat to human health. Detection of sublethally injured V. parahaemolyticusis thus of great importance for food safety guarantee.A technique combining Fourier transform infrared (FTIR) spectroscopywith chemometrics has been developed for the detection of sublethally injuredV. parahaemolyticus following thermal treatment at55°C for various periods. Traditional microbiological tests were adopted to verify the results of FTIRspectroscopy. The method was validated repeatedly with different bacterialstrains. The research contents and results are demonstrated as follows.(1) Comparison among different FTIR spectroscopy methods forbacterial detectionA series of FTIR spectroscopy methods were used to obtain bacterialspectra, which included transmission (KBr)-FTIR spectroscopy, ATR(diamond)-FTIR spectroscopy, transmission (BaF2)-FTIR spectroscopy, ATR(BaF2)-FTIR spectroscopy, transmission (BaF2)-micro-FTIR spectroscopy,and specular reflection (Al)-micro-FTIR spectroscopy. Compared to normalFTIR, micro-FTIR greatly improved the sensitivity; for instance, theabsorption intensity of1655cm-1was dramatically enhanced from10%(normal) to60%(micro). Micro-FTIR also greatly shortened detection time(120spectra were collected within one hour), and simplified the samplepreparation process, so that the risk of bacteria state alteration was minimized.Specular reflection (Al)-micro-FTIR spectroscopy absorbed so much that theintensity of1655cm-1almost reached100%, whose spectra was not suitablefor subsequent analysis. The transmission (BaF2)-micro-FTIR spectroscopywas selected as the best approach to detect injured microorganisms.(2) Development of FTIR spectroscopy method for bacterial identificationThe infrared spectra of10V. parahaemolyticus strains representing8serotypes and2Listeria monocytogenes strains representing2serotypes werecollected. In a typical infrared spectrum of V. parahaemolyticus, someimportant spectral peaks were found to be corresponded with such illustratingfeatures, including cell membrane phospholipids, cell wall polysaccharides,cytoplasmic proteins and lipids and nucleic acids. Bacteria producedabundant infrared absorptions in the four spectral regions:3000-2800,1800-1500,1500-1200and1200-800cm-1, and many peaks in these regionsfound their corresponding functional groups.Two L. monocytogenes strains were used as negative control, the infraredspectra of which were similar to those of V. parahaemolyticus. In the secondderivative spectra of L. monocytogenes and V. parahaemolyticus, many subtledifferences exsited in the1200to800cm-1and1500to1200cm-1bands.However, to differentiate V. parahaemolyticus from L. monocytogenes was noteasy by direct infrared and second derivative spectra.To exploit the hidden information, the four regions were analyzed withchemometrics approach. In the four spectral regions, all of thetwo-dimensional principal component analysis (PCA) plots showed distinctsegregations of L. monocytogenes cluster and V. parahaemolyticus cluster, far from each other; and clear segregations of different serotypes of V.parahaemolyticus as well. Using soft independent modeling of class analogy(SIMCA) analysis, all of the12strains were predicted with100%accuracy inthe above four regions. It suggested that FTIR spectroscopy not onlyspecifically differentiated V. parahaemolyticus and L. monocytogenes, buteven discriminated different serotypes.(3) Development of FTIR spectroscopy method for detection ofsublethally injured V. parahaemolyticusThe inactivation and sublethal injury treatment of V. parahaemolyticusATCC17802was carried out at55°C (for0,2,4,6and8min, respectively).The original and second derivative spectra of injured bacteria were similar tothose of intact bacteria. In the four spectral regions, all of the PCA plotsshowed distinct segregations of heat-treated cluster and untreated cluster; andpreliminary discriminations of heat-treated cells with varying time as well.The loadings for PC1, PC2and PC3confirmed that polysaccharide, protein,lipid, and nucleic acid were significantly affected; cell wall, cell membraneand DNA were damaged during heating. Using SIMCA, the injury waspredicted with more than80%accuracy; among20cases, only three ones’accuracies were lower than80%.To verify the results of FTIR spectroscopy, traditional microbiological tests were adopted. Survival curves of V. parahaemolyticus ATCC17802cells following thermal treatment at55°C recovered on unselective TSB agarand selective TCBS agar showed that viable counts decreased on both agarsas heating time increased, indicating increase of dead cells. They alsodisplayed that viable counts on unselective agar were always more than thoseon selective agar, and the gap increased as heating time increased, indicatingthat more and more V. parahaemolyticus bacteria were sublethally injured.FTIR spectroscopy for detection of sublethal Injury in V. parahaemolyticuswas established in this section.(4) Validation of FTIR spectroscopy method for detection of sublethallyinjured V. parahaemolyticusNine V. parahaemolyticus strains representing different serotypes werethermally treated at55°C with different heating time. In the four spectralregions, all of the PCA plots showed distinct segregations of heat-treated cellswith varying time (including intact ones); among the five strains with lethaltreatment, PCA plots displayed relatively farther distance between dead andinjured (including non-injured) cells. In all the four regions, PCA plots couldreveal a clear trend of increasing injury in V. parahaemolyticus withincreasing heating time. The region of1800-1500cm-1worked best in thediscrimination of cell status with an average contribution rate of90%by the first two principle components. SIMCA results showed high averageprediction and rejection rates in all the four spectral regions; the spectralregions of1800-1500cm-1and1500-1200cm-1worked best in the injuryprediction with an average prediction rate of98.89%and an average rejectionrate of99.78%. Combining PCA and SIMCA results, it suggested that thespectral region of1800-1500cm-1worked best in both discrimination of cellstatus and injury prediction for V. parahaemolyticus. After repeatedexperiments with different strains, the technique was confirmed to be able todiscriminate and predict intact, varying-degreed injured, and dead V.parahaemolyticus cells.In conclusion, the detection method by FTIR spectroscopy forsublethally injured V. parahaemolyticus developed in this study was able todifferentiate intact, varying-degreed injured, and dead V. parahaemolyticuscells. The method was confirmed to be specific, sensitive, fast, easy-to-do,and low-cost. As a rapid technology for V. parahaemolyticus detection, themethod could become a supplement of traditional incubation method toimprove the test accuracy. Furthermore, this method may be potentially usefulfor validating the effectiveness of various thermal processing treatments,predicting the degree of cell injury or death, and providing theoretical basisfor improving existing sterilization technologies. |
PDF Full Text Request