DETECTION OF ASPERGILLUS SPP . IN BIOLOGICAL SAMPLES BY REAL-TIME PCR

a Department of Clinical Biochemistry and Diagnostics, Faculty of Medicine and University Hospital, Charles University, 500 05 Hradec Kralove, Czech Republic b Department of Clinical Microbiology, Faculty of Medicine and University Hospital, Charles University, 500 05 Hradec Kralove c Department of Clinical Hematology, Faculty of Medicine and University Hospital, Charles University, 500 05 Hradec Kralove d Department of Biological and Medical Sciences, Faculty of Pharmacy, Charles University, 500 05 Hradec Kralove e Department of Microbiology, Faculty of Medicine and Dentistry, Palacky University, 775 15 Olomouc * Corresponding author fax: +420 495 832 003 e-mail: bolehrad@fnhk.cz

Currently, the proportion of invasive infections caused by the filamentous fungi of the genus Aspergillus is growing in immunocompromised persons, particularly in transplant recipients and neutropenic patients.Invasive aspergillosis (IA) usually develops after inhaling the air-borne conidia.
In such patients, the bronchopulmonary form of IA occurs most frequently, with the infection being capable of further dissemination from the primarily affected lower respiratory tract.This fungal disease is very devastating, mainly due to the strong angioinvasiveness of the Aspergillus species.A. fumigatus is the dominant etiological agent, followed by A. flavus and A. terreus.Therefore, it is not surprising that IA mortality in immunocompromised patients commonly exceeds 50 % despite antifungal therapy 1 .
It is widely accepted that successful therapy of IA closely depends on early detection of the fungus followed by an aggressive specific treatment.This mainly includes treatment by high doses of amphotericin B or its lipid formulations, and most recently by a new triazole voriconazole and echinocandin derivatives [2][3][4][5] .Unfortunately, laboratory diagnostics of IA remains extremely difficult, mainly due to the sensitivity of the methods and to the correct interpretation of the results.
Rapid laboratory diagnostics of IA has recently been based, above all, on immunological detection of the Aspergillus antigen (galactomannan, beta-glucan) and/or nucleic acid in the blood or other body fluid samples, especially from the respiratory tract 6 .
We have optimized previously published quantitative PCR assay 7 for detection of DNA from medically most relevant Aspergillus species.Optimized real-time PCR (rt-PCR) was then used for detection of Aspergillus DNA in various samples.DNA isolation.Plasma or serum, BAL fluid or other fluid samples of 300 µl to 1 ml in volume were centrifuged at 16,060 g.Then, the pellets were used for DNA extraction with the phenol-chloroform technique as described earlier 8 .The isolated DNA was stored at 4°C over 48-72 hrs or at -70 °C in the cases where longer storage was necessary.

Characteristics
All samples were analyzed in duplicate.
The PCR mixture of 10 μl in volume consisted of 1× LightCycler Fast-Start DNA Master HybProbe (Roche Diagnostics, Mannheim, Germany), 3 mmol/l MgCl 2 , 0.2 U UNG (uracil-DNA-glycosylase, Roche Diagnostics, Mannheim, Germany), the respective primer in the concentration of 0.8 μmol/l, the fluorescein isothiocyanatelabeled probe in the concentration of 0.4 μmol/l, and 5 μl of the template.The temperature profile of the PCR reaction was as follows: 50 °C for 5 min for UNG activation, 95 °C for 10 min as the initial denaturation, and 55 cycles with the temperatures of 95 °C for 5 s, 60 °C for 10 s, and 72 °C for 15 s.Sample fluorescence was measured in each cycle at the end of the annealing.
In order to verify the absence of DNA polymerase inhibitors, end-point PCR followed by agarose gel electrophoresis was carried out.Part of the gene for the human beta-globin was used as an internal control of the amplification.

RESULTS
Repeatability (accuracy in a series) of the optimized rt-PCR was determined 10 times using two samples of different concentration of Aspergillus DNA and the coefficient of variation was ≤ 1.28 %.Reproducibility (accuracy between days) was determined by testing two samples of different concentration of Aspergillus DNA measured in duplicate in five independent runs, the coefficient of variation was ≤ 2.43 %.
The real-time PCR reaction was positive with examined strains A. fumigatus AF13, A. flavus ATCC 36607, A. niger ATCC 10535, A. terreus DSM 826 and PCR sensitivity was found to be at 5 copies of Aspergillus DNA per mL.

DISCUSSION
Several studies have demonstrated recently that Aspergillus DNA can be detected successfully in various samples by different PCR-assays 7,[9][10][11][12][13][14] .We have evaluated the specificity of determination, sensitivity, repeability Detection of Aspergillus spp. in biological samples by real-time PCR and reproducibility of an rt-PCR published earlier 7 and optimized by ourselves.The specificity of determination and high sensitivity found was comparable to commonly used nested PCR-assay 10 .Repeatability and reproducibility of the PCR-assay optimized by us was sufficient, and in accordance with the data published by others 7,10,14 .Another general advantage of any PCR-based assay is its capability to detect invasive aspergillosis even if no viable fungal elements are present in the sample.By contrast, the presence of a sufficient amount of viable fungal elements is necessary if a cultivation method is to be successfully employed.Moreover, detection of viable fungal elements in the peripheral blood has been exceptional in the case of Aspergillus.For this reason, Aspergillus DNA detection in the peripheral blood by a PCR-assay should be preferred over cultivation techniques.
On the other hand, a number of practical problems are encountered in using molecular biology methods, e.g.false positivity resulting from contamination or false negativity resulting from presence of PCR-inhibitors, both of which can lead to an incorrect interpretation of the findings 15 .In addition, sensitivity of a PCR-assay can be decreased in patients receiving antifungal therapy 16 .
Material contamination can occur on several levels, and for this reason, the contamination risk should be minimized as much as possible.Firstly, contamination by air-borne conidia may occur during sampling.This risk is low in the case of blood samples, as an airtight sampling system is used.The risk, however, rises significantly in respiratory tract samples, sputum and endotracheal secretion in particular.It also cannot be excluded in the case of BAL, but it is substantially reduced.Hence, a single positive result is not conclusive for the diagnosis of invasive aspergillosis; two consecutive positive findings have much higher validity [17][18][19] .
Another source of contamination may come from laboratory processing, DNA extraction and preparation of reaction mixtures in particular.Negative controls as well as the amplification process may be contaminated, even though the latter occurs relatively rarely 15 .For this reason, we evaluated risk of contamination during DNA extraction using negative samples that were included in the set of the clinical samples.These negative controls were not found positive in any case.This clearly demonstrates high efficiency of the measures adopted to prevent cross-contamination of samples.In addition to strict adherence to general measures these relied also on the use dUTP and UNG (uracil-DNA-glycosylase).
Monitoring of the inhibition of the Taq DNA polymerase, which enables us to exclude falsely negative results, was also included in the assay.However, including an inhibition control into the reaction mixture itself is, in general, rather problematic, often leading to reduced sensitivity of the assay.This problem was resolved by running separate PCR assays targeting the housekeeping beta-globin gene, through which the presence of inhibitors was tested in all samples processed.This internal (inhibition) control must be positive in all samples, but we found 4 cases from 354 samples that were negative (confirmed presence of inhibitors).
A number of reports have shown that positive results for an PCR-assay found either repeatedly in consecutive samples from the same patient or correlating with a positive result of galactomannan (GM) detection from a sample processed in parallel correlated much better with clinical findings compared to single positive findings.Therefore, combined use of a PCR-assay and a GMassay has been demonstrated to significantly increase the chance of early and accurate IA diagnostics 18,19 .If a careful analysis of risk and predisposition factors is included together with the results of imaging methods (HR-CT), microbiological examination and detection of Aspergillus galactomannan and/or DNA, such a complex approach to patient can be of great help in early diagnostics and treatment of IA.Validation of an optimized PCR-assay is a key step to such an approach.