Rapid Molecular Diagnosis Of Invasive Fungal Infection

Background Over the past three decades, the incidence of invasive fungal infection (IFI) has been ever increasing, which has posed a great threat especially for immunocompromised patients such as transplant recipients, chemotherapy recipients, patients on other immunosuppressive therapy, and those who received invasive medical procedures. Early diagnosis of IFI which is the key for timely antifungal therapy, has been extremely challenging, as the imaging feature is often non-typical, histopathology not always available, and antigen/antibody tests have high false positive and/or negative rate due to the various influencing factors. Conventionally, definitive diagnosis of IFI requires discovery of pathogens in culture and phenotypic identification based on the isolates. However, the culture-based method has become increasingly unable to meet the growing demand for early diagnosis of IFI, because it is time-consuming, cumbersome, and sometimes imprecise because of its reliance on professional judgments. Molecular techniques, on the other hand, have provided new approaches for diagnosis of IFI, which is simple, reproducible, with reduced turn-around time and improved sensitivity and specificity, as compared to the traditional culture-based methods. The majority of molecular diagnostic techniques identify fungal pathogens based on culture isolates, while other tests allow for the detection of fungi directly from culture bottles, or even from clinical specimens.Part I Phenotypic and Molecular Identification of Common Clinical Pathogenic FungiObjective To evaluate and compare the performance of phenotypic identification, and molecular identification utilizing ribosomal RNA gene (rDNA) sequencing, in the identification of common clinical pathogenic fungi.Methods Both the phenotypic identification and rDNA sequencing of internal transcribed spacer (ITS)1, ITS2and D1/D2(at the5’ end of28s) regions were performed for8reference strains and47isolates in this study. Microbial identification was validated using combination of phenotypic and rDNA sequencing methods, which was used as standard to calculate the accuracy of phenotypic, single region, and multiple regions sequencing identification methods. Results Phenotypic identification yielded67.7%(21/31) species-level identification accuracy for yeasts and45.5%(10/22) for filamentous fungi. Accuracy of rDNA sequencing was superior to phenotypic method. D1/D2sequencing was more accurate than ITS1(83.9%VS.61.3%) and ITS2(83.9%VS.77.4%) in yeasts identification, but less accurate than ITS1(59.1%VS.68.2%) and ITS2(59.1%VS.63.6%) in filamentous fungi identification. Multiple regions sequencing had better performance than single region sequencing. We also found that some phylogenetically close-related species could not be distinguished even by combination of ITS1, ITS2, and D1/D2sequencing because of the high similarity in nucleotide sequence. Overall,90.9%(50/55) species-level and100%genus-level identification were achieved by combination of phenotypic and rDNA sequencing identification.Conclusions Identification based on phenotypic characteristics is still irreplaceable, especially for species with high genotypic identity. rDNA sequencing, especially multiple regions sequencing, has higher accuracy than phenotypic identification, and the value of different regions varies in different fungi pathogens. However, sequencing methods has the same disadvantage to phenotypic identification in the reliance on culture, and can be a problem when more than one microorganism coexists. The combination of phenotypic and rDNA sequencing identification can accurately identify the vast majority of common fungi pathogens to species, and its results can serve as a standard for evaluating other methods.Part II Microchip-Fingerprinting Identification of Common Clinical Pathogenic FungiObjective To establish a microchip-fingerprinting identification method and to evaluate its role in the identification of common clinical pathogenic fungi.Methods Fungi pathogens represented total of22unique species (10yeasts species and11filamentous fungi species) from10different genera were included in this research, all of which had been validated by combination of phenotypic and rDNA sequencing identification. Fingerprinting PCR amplifications followed by microchip electrophoresis detection of amplicons were performed to establish species-specific microchip-fingerprinting databases with M13and (GACA)4as single primer separately. Intra-and inter-species similarity coefficient (S value) was calculated for each database to analyze the similarity. The established microchip-fingerprinting system was then applied to identify isolates of different species to validate its accuracy.Results Standard microchip-fingerprinting databases with M13and (GACA)4as single primer were established. Similarity analysis showed that the average inter-species similarity coefficient of the microchip-fingerprints obtained with primer M13was26.7%, for (GACA)4it was26.1%. When fingerprints alignment was performed in yeasts and filamentous fungi database independently, high similarity (70%≤S value<100%) occurred between Candida galabrata and Rhodotorula mucilaginosa with primer M13, and the same is true between Penicillium marneffei and Fusarium oxysporum with primer (GACA)4. When different isolates were tested with preselection of yeasts or filamentous fungi,95.5%and96.4%species-level identification was achieved with M13and (GACA)4microchip-fingerprinting database respectively. We also found that isolates with inter-species similarity≥50%might not be differentiated from each other. Quality and quantity of template DNA and amplification condition could influence the resulting fingerprinting pattern and then identification results.Conclusions Microchip electrophoresis shortens the time needed significantly for detection of PCR fingerprints and make it more convenient for establishment of fingerprinting database and to align bands objectively, and thus present a quick, accurate and automated method for identifying clinical common fungi pathogens while further standardization of operation and expension of database are needed. Moreover, this method can also be used in genotyping and molecular epidemiology investigation. But presently it is still culture-based and cannot be used to detect pathogens directly from clinical specimens.PartⅢ PCR coupled with Electrospray Ionization Mass Spectrometry Identification of Common Clinical Pathogenic FungiObjective To evaluate the role of PCR coupled with electrospray ionization mass spectrometry (PCR/ESI-MS) technique in identification of common clinical pathogenic fungi and its prospect in IFI diagnosis.Methods PCR/ESI-MS analysis was performed for8reference strains and47clinical/environmental isolates in this study. All of the fungi pathogens had been validated by combination of phenotypic and rDNA sequencing identification. They were divided into two groups:47species within the coverage of broad fungal assay kit and8out of it. Identification accuracy was assessed separately for the two groups.Results Overall,85.1%(40/47) species or species complex-level identification and87.2%(41/47) genus-level identification were observed in all the strains and isolates contained in the coverage of broad fungal assay kit. While37.5%(3/8) genus-level identification was obtained for strains and isolates out of the coverage. Moreover, PCR/ESI-MS analysis could characterize mixtures of organisms simultaneously.Conclusions PCR/ESI-MS analysis is a new technique for identification and detection of pathogen organisms based on base composition information of multiple loci. It requires shorter turnaround time and lower DNA samples concentration to get accurate and sensitive identification results. PCR/ESI-MS could characterize mixtures of organisms simultaneously which is important for diagnosis of polymicrobial infections. In addition, by adding an internal positive control, the PCR/ESI-MS analysis can provide semi quantitative result (not investigated in this study). All of the advantages above make it a valuable technique for clinical diagnosis, mutation detection and molecular epidemiology research. Further investigations need to be done employing more well-validated reference strains and clinical isolates to evaluate the robustness of PCR/ESI-MS method.