Immunoproteomics Based Identification Of Immunodominant Antigens From Aspergillus Fumigatus For Early Diagnosis Of Invasive Aspergillosis

Background:In recent decades, invasive aspergillosis (IA) has emerged as an important cause of morbidity and mortality in patients with prolonged neutropenia. Moreover, several reports have recently described a rising incidence of IA in critically ill patients, even in the absence of an apparent predisposing immunodeficiency. The diagnosis of IA is still difficult because signs and symptoms are non-specific. The conventional diagnostic methods, such as tissue examination and microbial cultivation, may lack sensitivity in the first stages of infection in critically ill patients. As a result, the diagnosis of IA is often established after a long delay or following autopsy. Currently, the most valuable method used in the diagnosis of IA is GM assay. Galactomannan (GM) is present in the cell walls of most Aspergillus species. However, the sensitivity and specificity of GM testing can not meet clinical needs. Therefore, more prompt and accurate disease markers for early diagnosis are needed, which requires a thorough knowledge of fungal antigens detected in the serum or other body fluids of infected patients. We have recently observed that high levels of antibody against A. fumigatus are often present in the sera of proven IA patients. This finding prompted us to discover the potential novel biomarkers for the diagnosis of IA.Objective:To screen and identify Aspergillus fumigatus immunodominant antigens using immunoproteomics. To obtain the full-length recombinant protein of thioredoxin reductase GliT (TR), the novel immunodominant antigen identified from A. fumigatus. To establish an ELISA-based method for detecting anti-TR antibodies using the recombinant TR as the coating antigen. To investigate the antibody response to TR by a rabbit model of invasive aspergillosis. To evaluate the clinical significane of such anti-TR antibody detection method in diagnosis and therapy monitoring of IA by clinical serum samples.Methods:The secreted proteins and mycelial proteins of A. fumigatus prepared by TCA/acetone method were used for immunoblot assay with the sera of proven IA patients in order to select strongest immunoreactive proteins. The secretory proteins of A. fumigatus, which showed the high immunoreactivity, were separated by 2-DE, and probed with pooled sera of patients with proven IA. The sepecific immunoreactive proteins spots were excised from the 2-DE gels for tryptic in-gel digestion and matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF-MS). Peptide mass fingerprints (PMFs) obtained by the MALDI-TOF-MS analysis were searehed against the NCBI database using Mascot software. One of the proteins, thioredoxin reductase GliT (TR), which showed the best immunoactivity, was further analyzed by bioinformatics. The signal peptide and the probability of TR were predicted using SignalP software. Another subcellular localization prediction tool, WoLF PSORT, was used to analyze the amino acid sequences of proteins for prediction of cellular localization. Homology analysis was performed using the BLAST program. From total RNA of A. fumigatus, cDNA was generated using reverse transcription PCR. The amplified fragment was cloned into pMD18-T vector and sequenced. The recombinant plasmid pMD18-T/TR was digested by the restriction enzymes, and the target fragment was inserted into pET-28a(+) vector. Then, it was used to transform E.coli BL21(DE3) and induced expression by IPTG. The recombinant TR protein was analyzed by SDS-PAGE and western blot, and purifed by TALON Metal Affinity Resins.New Zealand white rabbits were immunosuppressed by intramuscular injection of cortisone acetate. The rabbits were infected intravenously withAspergillus conidia. An indirect ELISA detect the anti-TR antibody levels in the sera was established using the recombinant TR as coating antigen. The rabbit IA model was used to further evaluate the antibody response to TR. Sera from patients with culture- and/or histology-documented IA were obatined. Control sera were obtained from healthy individuals and non-neutropenic patients with culture-documented candidemia and bacteremia. The anti-TR antibody level was detected by ELISA. The sensitivity and specificity of the test were investigated and cut-off value was determined.Results:(1) The immunoreactivity of secretory proteins was higher than that of mycelial proteins. Among 4 kinds of media, proteins secreted by A. fumigatus during growth in YEPG for 14 days showed the highest immunoreactivity and the most immunoreactive bands, which could be further analyzed by immunoproteomic approach to screen specific anitgens of A. fumigatus.(2) The 2-DE and Western blot analyses of the filtrate proteins showed that IA patient sera reacted strongly with many antigens of secretory proteins. A total of 40 distinct immunoreactive spots were identified. The 39 successfully identified spots corresponded to 17 individual proteins. Most of these proteins are metabolic enzymes that are involved in carbohydrate, fatty acid, amino acid, and energy metabolism. Seven of these proteins have been reported as antigens of Aspergillus and other fungi, and others have not been described as antigens before, such as fumarylacetoacetate hydrolase FahA, aldehyde dehydrogenase AldA, aromatic aminotransferase Aro8, G-protein comlpex beta subunit CpcB, actin cytoskeleton protein (VIP1), phytanoyl-CoA dioxygenase family, urate oxydase UaZ,3-hydroxybutyryl-CoA dehydrogenase, proteasome component Pre8, putative and hypothetical protein. Seven proteins occurred in multiple spots. One protein of interest, which showed the best immunoreactivity, was identified as TR.(3) TR was predicted as a secretory protein with the presence of signal sequences with good predictive value (signalP probability,0.808). The protein localization of TR was predicted using WoLF PSORT, and the result also indicated that this protein might be an extracellular protein (Query Protein WoLFPSORT prediction:extr,12.0; cyto,6.5; cyto_nucl,4.0; mito,3.0; pero,2.0). This protein was BLAST-searched for sequence homology with human proteins and other fungi using the BLAST program. The results indicated that TR of A. fumigatus had no matches with human proteins. Furthermore, TR of A. fumigatus had low homology with other fungi, such as Candida albicans (25%), C. tropicalis (25%), C. glabrata (24%), C. guilliermondii (27%), C. dubliniensis (23%), Saccharomyces cerevisiae (24%), Cryptococcus neoformans (28%), and Penicillium marneffei (27%). These results suggested that the TR of A. fumigatus could be developed as a biomarker for the diagnosis of IA.(4) The full-length TR gene were cloned into the pET-28a (+) expression vector. The TR sequence was 100% identical to the gene of A. fumigatus published in GenBank database. After induction by isopropyl-β-D-thiogalactoside (IPTG), the recombinant 6-His-tagged TR was expressed, and a novel protein band corresponding to 36 kDa was detected by SDS-PAGE. Protein identity was unambiguously confirmed by MALDI-TOF MS, whereas following tryptic digestion proteins were identified yielding 37% sequence coverage. Most of the recombinant proteins were soluble. After purification using a TALON metal affinity resin, the protein purity was approximately 91%. Western blot showed that the recombinant proteins could be recognized by the sera from all six patients with proven IA。(5) A rabbit IA model was used to further evaluate the antibody response to TR. An increased amount of antibody was detectable in the serum 7 days after infection. The antibody level showed a constant rise around the time of clinical evidence of IA. This would imply that anti-TR antibodies are probably produced early in the development of IA and indicates the utility of anti-TR antibodies as a reliable marker for IA.(6) An ELISA to detect the anti-TR antibody levels was set up using the recombinant TR as coating antigen. The intra-assay variation and inter-assay variation of the ELISA were 8.57% and 12.17%, respectively. The anti-TR antibody detection had a sensitivity of 80.9% and a specificity of 96%. Overall, the sensitivity of the anti-TR antibody test in diagnosing IA was better than detection of GM (52.3%) in serum (p< 0.01). The anti-TR antibody in the first serum sample was positive in 15 of the 42 cases (35.9%, 15/42), whereas the serum GM remained negative. The first serum sample from 29 of 42 antibody-positive patients (69%) showed a positive result. By combining the GM assay with the anti-TR ELISA, the diagnostic sensitivity among non-neutropenic patients increased to 88.1%.Conclusions:Using immunoproteomics approach, a total of 17 proteins of A. fumigatus were identified as antigens espressed in IA. Ten of the proteins have not been reported as antigens of Aspergillus and/or other fungi. One protein of interest, which showed the best immunoreactivity, was identified as TR. The recombinant TR was cloned and expressed, then a recombinant TR-based ELISA to detect the anti-TR antibody levels was established. The test's performance was assessed by measuring the sensitivity and specificity in discriminating IA patients from non-IA patients. The results indicate that the anti-TR antibody is a useful marker in establishing the diagnosis of IA in non-neutropenic patients.