Surface-enhanced Raman Scattering Method For Species And Antibiotic Resistance Identification Of Pathogenic Bacteria

BackgroundRapid and accurate identification of pathogenic bacteria is critical to make suitable treatment,control pathogen related diseases and reduce drug resistance.The clinical routine methods for identification of pathogenic bacteria and drug susceptibility tests depend on bacterial culture.Culture-based diagnosis provides highly accurate results.Thus,it remains the gold standard method for identification of bacteria and antibiotic resistance.Altogether the culture-based approach takes a minimum of 48 hours for biochemical and antibiotic susceptibility test.It is not appropriate for rapid screening of pathogens.Faster or accurate methods have been developed.These method include enzyme-linked immunosorbent assay?ELISA?,polymerase chain reaction?PCR?and Mass spectroscopy?MS?.These methodologies are more rapid but suffer from limitations due to background contamination,and high cost.Besides,they often require specially trained staff and specialized technicians,which cannot be operated under field conditions.New methods for general,quick,automatic,and portable detection of bacteria are still urgently needed.Surface-enhanced Raman scattering?SERS?has become a powerful technique for biological applications.SERS provides high sensitivity,great selectivity,rapid response time,and easy sample preparation.In particular,the identification and detection of pathogenic bacteria by SERS has drawn considerable attentions.SERS is able to obtain whole microorganisms fingerprints and detect single cell.SERS provides enhanced Raman signals of molecules by metallic nanoparticles with enhancing factor up to 10⁶-10¹².Current SERS methods for pathogenic bacteria detection are highly sensitive.However,they provided spectra of Raman reporter instead of the bacterial cell.By contrast,some nanoparticles based SERS method does not require any label dye.These methods can directly obtain the intrinsic SERS fingerprint of bacteria cell wall.SERS have potential for identification of pathogens.Thus,there is a continuing need to develop improved SERS techniques.These techniques are ideally simple,rapid,reproducible,and have no interference from the substrate.Therefore,this study aimed at the urgent need for rapid detection and drug resistance analysis of bacteria.In this study,SERS technology,multivariate statistical analysis and machine learning algorithm were used for pathogenic bacteria detection.A label-free SERS detection method was established using positively charged silver nanoparticles?AgNPs⁺?to detect different species bacteria and methicillin-resistant Staphylococcus aureus?MRSA?.Then multivariate statistical analysis containing principal component analysis?PCA?,Hierarchical Clustering analysis?HCA?,partial least squares discriminant analysis?PLS-DA?and latent structure discriminant analysis classification?OPLS-DA?wasused to identify different species bacteria and MRSA.Furthermore,machine learning algorithm was processed to establish and evaluate discriminant analysis models for MRSA and analysis the difference characteristics between MRSA and methicillin-sensitive Staphylococcus aureus?MSSA?.This study included three sections as follows.Methods and materialSection 1SERS effect of different kinds of nanomaterials for pathogenic bacteria detection.1)Synthetic of nanomaterials,including gold-silver nano-core shell composite material?AuNR@Ag?,gold nano-triangle-gold nanoparticle composite?TAuNPs-AuNPs?and silver nanoparticle?AgNPs?colloidal solution.2)Characterization of different nanomaterials and their combination with Staphylococcus aureus?S.aureus?,including ultraviolet-visible?UV-Vis?,scanning electron microscopy?SEM?,transmission electron microscopy?TEM?,zeta potential,particle size analysis,and energy dispersive spectrometer analysis?EDS?.3)Verify the SERS effect of different nanomaterials for detecting S.aureus.Section 2:Analysis performance evaluation of SERS detection method using AgNPs⁺.1)Optimization of experimental conditions.2)Verification the SERS effects of AgNPs⁺for the detection of S.aureus.3)Analysis performance evaluation of this method including sensitivity,specificity,repeatability and the application in clinical samples.Section 3:SERS detection for pathogenic bacteria species and MRSA based on AgNPs⁺.1)Collection of clinical MRSA strains.2)SERS detection of different species of pathogenic bacteria and MRSA.3)Multivariate statistical analysis of different species of pathogenic bacteria and MRSA.4)Machine learning algorithms was processed to establish and evaluate discriminant analysis model of MRSA.The difference characteristics between MRSA and MSSA were analyzed.Main resultsSection 1:SERS effect of AgNPs⁺was higher than other kinds of nanomaterials.1)AgNPs⁺,AgNPs^-and AuNR@Ag were successfully synthesized.2)The SERS effect of AgNPs⁺is more significant than other nanomaterials.AgNPs⁺are tightly covered on the bacterial surface by electrostatic bonding.AgNPs⁺were chosen for SERS detection.3)Q-SERS and AuNR@Ag substrate might be more suitable to detect nanoscale molecule instead of micron sized bacteria.Section 2:SERS detection method using AgNPs⁺showed acceptable analysis performance.1)The experimental conditions were optimized.The incubation time of AgNPs⁺and S.aureus was 30 minutes?min?,the volume ratio of AgNPs⁺to S.aureus was 1:1,and the ultrapure water was solution buffer,and the 65 nm AgNPs⁺were used.2)The SERS detection method based on AgNPs⁺was successfully established.The AgNPs⁺enhancement factor reached 10⁷,and the peak intensity of AgNPs⁺at 730 cm^-1 was3.4 times higher than that of AgNPs^-.3)Single bacteria could be detected by this method.The repeatability of the test method was acceptable.The intra-assay relative standard deviation?RSD?was 12.7%,and the inter-assay RSD was 14.31%.This method was successfully used for the detection of clinical samples.Section 3:Different species of pathogenic bacteria and MRSA could be discriminated by multivariate statistical analysis and machine learning algorithms.1)The Raman spectrum peak position and peak intensity of bacteria and fungi are significantly different,and that of Gram-positive bacteria and Gram-negative bacteria have significant differences.2)Different genus or species of pathogenic bacteria could be distinguished by multivariate statistical analysis.The different characteristics between different S.aureus and other ic bacteriawere located at 721-735 cm^-1,1316-1322 cm^-1,647-656 cm^-1,1618-1626 cm^-1.The different characteristics between S.aureus 29213 and S.aureus 25923were located at 231-256 cm^-1,729-735 cm^-11 and 1319-1335 cm^-1.3)The Raman peak position,intensity and relative intensity ratio between MRSA and MSSA were different.The peak intensity of MRSA at 727-730 cm^-1 and 1322-1325 cm^-1 is relatively low,which is relatively high at 1003-1006 cm^-1,1154-1159 cm^-1,1237 cm^-1 and1518-1521 cm^-1.4)MRSA and MSSA could be distinguished by multivariate statistical analysis.The neural network was selected as the classification model to discriminant MRSA and MSSA.The neural network classifier showed that the area under the curve?AUC?was 0.986,the accuracy was 0.960,and the recall was 0.959.OPLS-DA analysis showed the different characteristics between MRSA and MSSA were at 513-519 cm^-1,727-732 cm^-1,1147-1152cm^-1and 1513-1521 cm^-1,while the machine learning algorithms showed that were at721-727 cm^-1,697.821 cm^-1,960-972 cm^-1,1314.27 cm^-1,750-756 cm^-1 and 510.72 cm^-1.ConclusionsIn summery,a simpled and label-free SERS detection method was established using AgNPs⁺to detect different species of bacteria and MRSA.The enhancement factor reached to 10⁷.This method showed acceptable analysis performance,which could detect single bacteria.Different species of bacteria and MRSA were be successfully distinguished by multivariate statistical analysis.The different characteristics between different S.aureus and other pathogenic bacteria were analyzed.Furthermore,neural network was established as classifier for MRSA with accuracy of 0.96.The difference between MRSA and MSSA were further analyzed.This study provided a new method for rapid detection and drug resistance analysis of bacteria,which was of great importance for the diagnosis of infectious diseases and prevention of antibiotic resistance.