Diagnostic Value Of Eno And Fba Quantitative Detection In Invasive Cadidiasis

Background:Candida species ranked the fourth most common cause of nosocomial bloodstream infections, and the third cause among intensive care unit (ICU) patients. Candida albicans accounted for the majority of invasive Candida species infections (50%-70%). Diagnosis of invasive candidiasis is extremely challenging, because clinical symptoms are often nonspecific. Poor outcome in patients with invasive candidiasis are related to delayed treatment with an effective antifungal treatment. Significant mortality of invasive candidiasis is as high as40%despite antifungal therapy, partially associated with the lack of rapid and reliable diagnostic assays. Blood culture and histopathologic identification are currently considered as the "gold standard" for diagnosis of invasive candidiasis. However, blood cultures are poorly sensitive, and positive results are often available too late for clinicians to make a decision. Particularly, deep-seated candidiasis patients who are not associated with candidemia are missed by blood cultures. Histopathologic diagnostic techniques are often invasive procedures, which carry significant risks and are often impractical in critical ill patients. Early diagnosis and/or timely treatment have been shown to be critical for successful outcome. Therefore, seeking reliable and effective surrogates in invasive candidiasis and establishing the detection methods are among the most urgent needs. In the past decades, serological methods focused on candida specific antigens and/or antibodies have been widely investigated, due to the sensitivity, simplicity, low costs and easy operation. The presence of candida specific antigens could provide direct evidence for invasive candidiasis diagonosis. Ideal antigen marker could also provide information for therapeutic monitoring and prognosis. Antigen-based tests are major serological methods that have been developed for early detection of invasive candidiasis. Detection of (1,3)-p-D-glucan and mannan are broadly validated antigen-based tests in clinical practice. It has been reported that detection of the above-mentioned antigens could provide opportunity to initiate pre-emptive antifungal therapy even before signs and symptoms of clinical disease appear and before positive blood cultures available. However, the moderate sensitivity and specificity is unsatisfied. There is a need for other surrogates with better sensitivity and specificity that can improve the diagnosis of invasive candidiasis.Enolase (Eno) is an enzyme that catalyzes the reversible conversion of2-phosphoglycerate into phosphoenolpyruvate. It is composed of a440-residue sequence with a molecular mass of46-48kDa. Fructose-1,6-bisphosphate aldolase (Fba) catalyzes the reversible cleavage of fructose-1,6-bisphosphate to dihydroxyacetone phosphate and glyceraldehyde3-phosphate, predicted a359-residue sequence with a molecular mass of39kDa. Eno and Fba are central enzymes in the glycolytic pathway, playing important role in energy metabolism. Eno and Fba locate both in cytoplasma and in cell wall, and are the most abundant cytoplasmic proteins. Eno and Fba have been identified as hyphae-specific proteins because their expressions were upregulated during hyphal transition. Moreover, as shown in serological proteome analysis and immunoblotting data, Eno and Fba elicited high anti-enolase IgG and anti-Fba IgG antibody levels in serum specimens from invasive candidiasis patients. Eno and Fba have been identified as immunodominant antigens in Candida albicans infections. We speculated that Eno and Fba have the potential advantages of diagnositic marker for invasive candidiasis and proposed to evaluate their diagnostic value in invasive candidiasis.Objective:This study is designed to establish sandwich ELISAs for quantitative detection Eno and Fba of Candida albicans based on specific monoclonal and polycolonal antibody to recombinant proteins of Candida albicans Eno and Fba in previous study. To obtain the kinetics of Eno and Fba in different immune status invasive candidiasis mouse, by developing immunocompetent and immunocompromised invasive candidiasis BALB/c mouse model. To determine sera levels of Eno and Fba in invasive candidiasis patients, and to evaluate the diangnostic value of Eno and Fba in invasive candidiasis.Methods:1. Development of sandwich ELISAs for quantitative detection of Eno and FbaPolycolonal antibody was purified by recombinant Eno or Fba coupled NHS-activated Sepharose Fast Flow gel. Monocolonal antibody was purified by Protein G Sepharose media. Cytoplasmic supernatant of Candida albicans, Candida parapsilosis, Candida tropicalis, Candida glabrata, Candida guilliermondii and recombinant Candida albicans Eno and Fba were immunoblotted with purified antibody to detect the specificity of antibody. Monoclonal antibody was employed as coating antibody, while HRP-conjugated goat polyclonal antibody was used as detecting antibody. The performance parameters of the sandwich ELISA including precision, specificity and the limit of detection, linear range, serum recovery were verified by using recombinant Candida albicans Eno and Fba.2. Quantitative detection of Eno and Fba in supernatant of common fungi culturesThe developed assay was applied to determine Eno and Fba levels in supernatant of Candida albicans, Candida parapsilosis, Candida tropicalis, Candida glabrata, Candida guilliermondii, Cryptococcus neoformans and saccharomyces cerevisiae for preliminary evaluation.3. Detection levels of Eno and Fba in sera and kidney homogenates in IC mice4-6weeks old female BALB/c mice were randomly assigned to immunocompromised group (n=25), immunocompetent group (n=25) and control (n=5). Immunosuppression status was induced by administration of100mg/kg of intraperitoneal (i.p.) cyclophosphamide4days prior to infection. This was followed by a second dose of100mg/kg i.p. cyclophosphamide at the day infection and then75mg/kg i.p. cyclophosphamide every other day after intravenous challenge. Immunocompetent and immunocompromised BALB/c mice were injected intravenously (i.v.) in the lateral tail vein with5×105Candida albicans ATCC90028yeast in0.1ml sterile PBS, control group mice were injected with0.1ml sterile PBS. Serum and supernatant of kidney homogenates was collected at8h,1d,2d,3d and5d after infection (5mice per time point). Levels of Eno and Fba in various time point samples were determined to obtain the kinetics of Eno and Fba in invasive candidiasis with different immune status. WBC, blood culture and fungal burden in kidney were also determined.4. Quantitative detection of Eno and Fba in invasive cadidiasis patients200sera from92invasive candidiasis proved by blood cultures and clinical symptom were collected. The Candida species distribution was as follows:C. albicans (n=46,104samples), C. parapsilosis (n=15,27samples), C. tropicalis (n=17,46samples),other Candida species (n=14,23samples)(including Candida glabrata6, C. guilliermondii3, C. pelliculosa2, C. lusitaniae1and C.haemulonii2). Control sera were obtained from healthy individuals (n=100) and Candida species colonized patients (n=30). Levels of Eno and Fba in invasive candidiasis patients and individuals in control group were determined, the diagnostic performance were evaluated.Results:1. The specificity of antibody against Eno or FbaAs shown in Western blot results, anti-Eno polycolonal and monocolonal antibody specially recognized cytoplasmic Eno of C. albicans and recombinant Eno. Cross-reactivity of anti-Eno monocolonal antibody was observed with C. parapsilosis, and that of anti-Eno polycolonal antibody was observed with C. parapsilosis and C. tropicalis. Anti-Fba polycolonal and monocolonal antibodies not only recognized cytoplasmic Fba of C. albicans and recombinant Fba, but also were cross reactive with C. parapsilosis and C. tropicalis.2. Performance of ELISA for quantitative detection of EnoThe intra and inter-coefficient of variation was6.72%,7.18%and10.4%,12.68%, at the Eno concentration of25ng/ml and5ng/ml, respectively. The limit of detection was1.25ng/ml. The linear range is (1.25-50)ng/ml, sera recoveries were97.75%、122.84%and92.13%,corresponding to standard concentration of50ng/ml,25ng/ml and5ng/ml,respectively.3. Performance of ELISA for quantitative detection of FbaThe intra and inter-coefficient of variation was6.88%,7.45%and9.72%,11.92%, at the Fba concentration of50ng/ml and10ng/ml, respectively. The limit of detection was2.5ng/ml. The linear range is (2.5-100)ng/ml, sera recovies were104.02%、102.84%and90.90%,corresponding to standard concentration of100ng/ml,50ng/ml and10ng/ml, respectively.4. Quantitative detection of Eno and Fba in supernatant of fungi culturesEno (3.06ng/ml) was detected in the supernatant of Candida albicans culture after incubated in37℃for24h. The level of Eno in the supernatant of Candida albicans culture gradually increased thereafter and reached33.43ng/ml at120h. Eno in the supernatant of Candida parapsilosis culture was detectable after being incubated in37℃for48h (1.03ng/ml) and kept low level in the incubation period (3.17ng/ml at120h). Incubated in37℃for9h, Fba could be determined in the supernatant of Candida albicans, Candida parapsilosis and Candida tropicalis (2.60ng/ml,1.14ng/ml and2.19ng/ml, respectively). The level of Fba in Candida albicans culture was54.16ng/ml at120h, while the level of Fba in Candida parapsilosis and Candida tropicalis showed same tardy upward trend (13.60ng/ml and7.82ng/ml at120h, respectively). Eno and Fba could not be detected in the supernatant of Candida guilliermondii, Candida glabrata, Cryptococcus neoformans and Saccharomyces cerevisiae throughout the whole incubation period. The growth of Candida albicans hyphae was more efficient in37℃than in25℃.The amount of Eno and Fba in the supernatant incubated in25℃was less than those in37℃. It was suggested that level of Eno and Fba was positive correlated with the growth of hyphae.5. Kinetic patterns of Eno and Fba in IC mice5.1Positive rate of Eno and Fba in seraIn immunocompetent mice, at8h after infection,60%(3/5) mice had candidemia. Candida albicans didn’t be isolated during the period from1d to5d after infection. In immunocompromised mice,60%(3/5) and20%(1/5) mice had candidemia at8h and1d after infection,respectively. Candida albicans could no longer be isolated during the period from2d to5d after infection. Circulating antigenemia of Eno and Fba was present in both group mice during the whole observation period. The positive rate of Eno was33.3%-100%and50%-100%in immunocompetent mice and immunocompromised mice respectively. The positive rate of Fba was75%-100%and100%in immunocompetent mice and immunocompromised mice respectively. The positive rate of Eno and Fba was higher than that of blood cultures at the same time point throughout the test period. There were significant differences between positive rate of Fba at1d after infection in immunocompetent mice, that of Eno at3d after infection, that of Fba at2d,3d and5d after infection in immunocompromised mice and that of controls (P all=0.008, P<0.05/4).5.2Levels of Eno and Fba in sera The levels of Fba at8h,1d after infection were [4.56(3.66,34.03) ng/ml and23.76(20.34,41.06)ng/ml, respectively] in immunocompetent mice and those at all time points in immunocompromised mice were significantly higher than in controls[0(0,0.175) ng/ml](P all=0.008, P<0.05/6). The levels of Eno at1d,3d after infection were [6.34(0.92,12.45) ng/ml and3.09(1.26,6.48) ng/ml, respectively] in immunocompromised mice were significantly higher than in controls (P all=0.005, P<0.05/6). The levels of Eno and Fba in immunocompromised mice were higher than those in immunocompetent mice, but the significant difference was found at3d after infection (P=0.045). There was significant positive correlation between circulating Eno and Fba (r,=0.363, P=0.018, n=42).5.3Levels of Eno and Fba in kidney homogenatesEno and Fba were determined in all supernatants of kidney homogenates during the whole observation period, reflecting the invasive infection in the kidney. Significant difference was found between levels of Eno at8h,1d after infection [(262.99±92.18) ng/ml and (325.24±117.32) ng/ml, respectively] in immunocompromised group and that in immunocompetent group[(139.24±61.48) ng/ml and (133.82±80.15) ng/ml, respectively](P=0.037and P=0.017, respectively). The levels of Eno and Fba in supernatants of kidney homogenates of IC mice were higher than those in control. The levels of Eno and Fba in supernatants of kidney homogenates in immunocompromised group at all time points were significantly higher than those in control (P<0.05or P<0.01). The level of Fba at5d after infection [(999.96±168.78)ng/ml] in immunocompetent group was significantly higher than that in control [(0.38±0.22) ng/ml]. No obvious change was found in levels of Eno and Fba in various time points in both groups. There was significant positive correlation between Eno and Fba in kidney homogenates(rs=0.623, P=0, n=50).6. Quantitative detection of Eno and Fba in IC patients6.1Quantitative detection of Eno and Fba in IC patientsThe level of Eno and Fba in IC patients (n=92) was0(0,1.33) ng/ml and3.58(0,16.32) ng/ml respectively, significantly higher than those in negative controls(n=130)(U value were3026.5and4312.0, P=0). The highest sera levels of Eno and Fba were found in invasive Candida albicans infection patients (n=46), being0.52(0,3.55)ng/ml and5.53(0,31.31)ng/ml, respectively, followed by invasive Candida tropicalis infection patients (n=17), invasive Candida parapsilosis infection patients(n=15) and other Candida species infection(n=14). The sera levels of Eno and Fba in invasive Candida tropicalis, Candida parapsilosis and other Candida species infection patients were (0,0.39)ng/ml,3.67(0,19.81)ng/ml,0(0,0.23)ng/ml,0(0,4.69)ng/ml and0(0,0)ng/ml,0(0,3.72)ng/ml respectively. There was statistical difference between sera levels of Eno and Fba in other Candida species infection patients, controls and that of invasive Candida albicans infection patients (P <0.05/5or P<0.01/5). The significant difference was found in sera level of Fba between the controls and IC patients caused by Candida parapsilosis and Candida tropicalis (P<0.05/5or P<0.01/5). The sensitivity of Eno and Fba in IC was33.70%(31/92) and53.26%(49/92), respectively. The positive rate of Eno in various candida species infections was lower than that of Fba. However, there was no statistic difference. Sera levels of Eno and Fba showed significant positive correlation by Spearman analysis(rs=0.489, P=0, n=200).6.2Quantitative detection of Eno and Fba in IC patients caused by Candida albicansThe sensitivity of Eno in invasive Candida albicans infection was52.17%(24/46) and13.04%(6/46) in invasive non-Candida albicans infection. There was statistic difference between two groups (χ2=16.026, P=0). The sensitivity of Fba in invasive Candida albicans infection was69.57%(32/46) and34.78%(16/46) in invasive non-Candida albicans infection. The former was significantly higher than the latter (χ2=11.152, P=0.001). For invasive Candida albicans infection, the sensitivity, specificity, and area under the receiver operating characteristic curve (AUC-ROC) of detection Eno and Fba were52.17%,99.23%,0.729and69.57%,99.23%,0.839,respectively. The sensitivity increased to80.43%when Eno and Fba were combined for diagnosis. The presence of Eno and Fba in sera sometimes was prior to the positive blood cultures. Multiple sampling improved the detection of Eno and Fba.Conclusions:1. Two specific and sensitive sandwich ELISAs for quantitative detection Eno and Fba of Candida albicans were developed. The performance parameters of sandwich ELISAs including precision, specificity,the limit of detection, linear range and serum recovery were assessed and met the requirement of clinical application, the limit of detection was relatively consistant with previous research.2. Two sandwich ELISAs for quantitative detection Eno and Fba were applied to the supernatant of fungi culture including Candida albicans, Candida parapsilosi, Candida tropicalis, Candida guilliermondii, Candida glabrata, Cryptococcus neoformans and Saccharomyces cerevisia. Natural form of Eno and Fba in the cultures could be specifically detected by the ELISAs. The ELISA for Eno had weak cross-reactivity with Candida parapsilosis, and wheras the ELISA for Fba cross reacted with Candida parapsilosis and Candida tropicalis. The amount of Eno and Fba increased with the growth of hyphae.3. Invasive candidiasis mouse model with immunocompetent or immunocompromised status was developed.The kinetics of Eno and Fba in circulating blood and kidney homogenates for the first time was obtained. Eno and Fba were detected in all kidney homogenates during the whole observation period, reflecting the invasive infection in the kidney. The presence of Eno and Fba antigenmia was determined throughout the whole experiment period in both group mice. The positive rate of Eno and Fba was always higher than that of blood cultures at the same time point. Significant positive correlation was found not only between sera levels of Eno and Fba, but also between levels of Eno and Fba in kidney homogenates. It is suggested that Eno and Fba have the potential of diagnostic marker for invasive candidisis.4. Diagnostic value of quantitative detection of Eno and Fba was evaluated in invasive candidiasis patients. The diagnostic performances of both ELISAs in invasive Candida albicans infections were better than in invasive pooled Candida species infection. The sensitivity, specificity, and area under the receiver operating characteristic curve (AUC-ROC) of detection Eno and Fba were52.17%,99.23%,0.729and69.57%,99.23%,0.839respectively. The sensitivity increased to80.43%when Eno and Fba were combined for diagnosis. The presence of Eno and Fba in sera sometimes was prior to the positive blood cultures. Our findings suggest that Eno and Fba could be promising diagnostic markers especially used in combination for invasive Candida albicans infections. Detection sera level of Eno and Fba may provide the opportunity for early diagnosis and complement the diagnostic value of blood cultures.