Preliminary evaluation of one conventional nested and two real-time PCR assays for the detection of Toxoplasma gondii in immunocompromised patients

Toxoplasma reactivation represents a rare but potentially life-threatening complication in bone marrow transplant recipients and other immunocompromised patients (Mele et al., 2002). Rapid diagnosis of toxoplasmosis is crucial in patients with impaired immune function, as early treatment may improve the clinical outcome. Classical serological diagnosis of toxoplasma infection relies on the detection of anti-toxoplasma immunoglobulin, but serology may be unreliable in immunodeficient individuals who fail to produce significant titres of specific antibodies. Over the past decade, PCR assays have allowed the sensitive detection of Toxoplasma gondii DNA in clinical specimens. Molecular diagnosis of toxoplasmosis can be accelerated by performing real-time PCR protocols (Bell & Ranford-Cartwright, 2002). These newly developed real-time assays allow amplification and simultaneous detection of DNA in 1 h. However, no comparative data are available on the performance of conventional and real-time PCR assays for the detection of T. gondii DNA under routine conditions. Therefore, the aim of this study was to amplify T. gondii DNA from routine samples of allogenic transplant recipients by applying two real-time PCR protocols on a LightCycler (LC), and a conventional nested PCR. Over a period of 18 months, 1000 routine specimens (83 % EDTA blood, 4 % bronchoalveolar fluid, 6 % cerebrospinal fluid, 2 % pleural effusion, 2 % bone marrow aspirate, 1 % lymph node biopsies and 2 % tissue biopsies) were collected from allogenic stem cell recipients. Extracted DNA was amplified with a conventional nested PCR targeting the T. gondii B1 gene (Roth et al., 1992). DNA preparations that were found to be positive in the conventional nested PCR were amplified with two LC real-time PCR protocols targeting the T. gondii B1 gene (Costa et al., 2000) and the T. gondii 529 bp repeat region (Reischl et al., 2003). In the conventional nested PCR (Roth et al., 1992) the outer primers P3 (5'-CTTCAAGCAGCGTATTGTCG-3') and P7b (5'-TCTTTAAA GCGTTCGTGGTC-3') and the inner primers P7 (5'-TAAAGCGT TCGTGGTCAACT-3') and P8 (5'-GGAACTGCATCCGTTCATGA-3') were used. The amplification mixture (25 µl) contained 200 µM dNTPs (Promega), 3 mM MgCl₂, Taq DNA polymerase (Roche Diagnostics) and 0.2 µM of each primer. The amplification mixture contained 5 µl extracted DNA in the first round of the nested PCR, and 0.5 µl of the PCR product was amplified in the second round for 40 cycles (20 s denaturation at 94 °C, 20 s annealing at 55 °C and 20 s extension at 74 °C). The amplicon (190 bp) was detected by agarose gel electrophoresis and ethidium bromide staining.

For the LC PCR targeting the T. gondii B1 gene (Costa et al., 2000) primers B1 sense (5'-GGAGGACTGGCAACCTGGTGTCG-3') and B1 antisense (5'-TTGTTTCACCCGGACCGTTTAGCAG-3') and probes Tox1 (5'-CGGAAATAGAAAGCCATGAGGCACTCCfluorescein-3') and Tox2 (5'-Red 640ACGGGCGAGTAGCACCTGAGGAGATPh-3') were used. The reaction mixture (20 µl; Master Hybridization Probes kit; Roche Diagnostics) contained 0.5 µM of each primer, 0.2 µM of each probe, 5 mM MgCl₂ and 5 µl extracted DNA. Amplification was performed in a LightCycler (Roche Diagnostics) for 50 cycles: 5 s denaturation at 95 °C, 10 s annealing at 60 °C and 15 s extension at 72 °C, with an overall ramp rate of 20 °C s¹. For the LC PCR targeting the T. gondii 529 bp repeat region (Reischl et al., 2003) primers Tox9 (5'-AGGAGAGATATCAGGACTGTAG-3') and 10as (5'-GCGTCGTCTCGTCTGGATCG-3') and probes Tox3 (5'-GAGTCGGAGAGGGAGAAGATGTTfluorescein-3') and Tox4 (5'-Red 640CCGGCTTGGCTGCTTTTCCTGPh-3') were used. The reaction mixture (20 µl) contained 0.5 µM of each primer, 0.2 µM each probe, 3 mM MgCl₂ and 5 µl extracted DNA. Amplification was performed for 50 cycles: 10 s denaturation at 95 °C, 20 s annealing at 50 °C and 30 s extension at 72 °C, with an overall ramp rate of 20 °C s¹.

To evaluate the sensitivity of the conventional nested PCR assay, the PCR product of a sample containing T. gondii DNA was sequenced and cloned into a plasmid vector (TOPO TA cloning kit; Invitrogen). Toxoplasma-free EDTA blood was spiked with the plasmid. Anti-toxoplasma antibodies were detected in serum by immunosorbent agglutination assay (ISAGA; Toxo Multi Tool; Innogenetics), immunofluorescence test (IFT; Toxo-Spot IF; bioMérieux) and ELISA (IgG/IgM kit; VirionSerion).

All PCR assays were performed according to quality standards for nucleic acid amplification techniques published by the German Society for Hygiene and Microbiology (Roth et al., 2001). This included physical separation of working areas for the preparation of amplification reaction mixtures, for specimen preparation, for preparation of positive controls and for the amplification and detection of nucleic acids. In all rooms the equipment was labelled to indicate clearly the working area to which it belonged. Only plugged pipette tips were used to prevent contamination by aerosols. All areas were equipped with flow benches. Moreover, the risk of contamination was reduced by a unidirectional workflow ensuring that pre- and post-amplification procedures were not performed alternately. To inactivate nucleic acids, working surfaces were cleaned with sodium hypochlorite solutions and flow benches were irradiated with UV light. The following controls were performed in each run. As a negative control for specimen processing, DNA was extracted from a sample that did not contain the target sequence and the DNA extract was subjected to amplification and detection. As a negative control for amplification, reaction mixtures containing reagents only, but not nucleic acids, were amplified (in the case of the nested PCR, in both rounds). As a positive control for extraction and amplification, DNA of a sample containing the target sequence in low copy number was extracted and amplified. Finally, as a control for inhibition, a fragment of the human ß-actin gene was amplified in each sample. Samples with discrepant results in the three PCR assays were analysed in duplicate.

Over the 18-month study, 16 samples obtained from eight allogenic transplant recipients were shown to contain T. gondii DNA by the nested PCR protocol (Table 1). Twelve of these samples yielded a positive result following amplification with each of the three PCR protocols. The remaining four samples were negative in at least one of the real-time PCRs. PCR products from two discordant samples were sequenced, which confirmed the accuracy of the conventional PCR. PCR inhibitors were not detected when a fragment of the human ß-actin gene was amplified (Murray et al., 1990). To investigate the analytical sensitivity of the nested PCR assay, toxoplasma-free EDTA blood was spiked with a plasmid containing the PCR positive control at different concentrations (10⁷ to 10⁰ copies ml¹). DNA was extracted and amplified. Amplicon analysis (190 bp) on an agarose gel indicated a sensitivity of the nested PCR in the range of 110 copies (ml blood)¹. Amplification of DNA from a diluted T. gondii-positive blood sample using each of the three PCR assays indicated similar sensitivities of the three assays (data not shown). Over the entire study, 1000 routine DNA extracts were amplified using the nested PCR, in which no non-specific amplicons were detected. Amplification of DNA extracted from 20 randomly selected routine EDTA-blood samples using either of the two LC assays did not generate a fluorescence signal. Amplification and detection of the amplicon was completed within 1 h using the real-time PCR assays, compared to 3.5 h using the conventional nested PCR plus subsequent gel electrophoretic detection of the amplicon. The clinical data of three patients whose specimens showed discrepant PCR results are summarized in Table 2.

Table 1. Detection of T. gondii DNA in patient samples with real-time and conventional nested PCR assays +, Positive; , negative; BAL, bronchoalveolar lavage; CSF, cerebrospinal fluid.

Table 2. Clinical data of allogenic stem cell recipients with T. gondii-positive samples AML, Acute myeloid leukaemia; ALL, acute lymphoid leukaemia; JMML, juvenile myelomonocytic leukaemia; NMR, nuclear magnetic resonance; CT, computerized tomography.

Our observation that the LC assays failed to detect T. gondii DNA in some specimens confirms the results of a recent study on the performance of LC technology in human samples (Teo et al., 2002). Several parameters were identified to have an adverse effect in the LC assays, including abstraction of PCR reagents on glass capillaries, primer-dimer formation and non-specific product generation (Teo et al., 2002). Also, the amount of human chromosomal DNA in the specimen was shown to influence the amplification efficiency of the target DNA. In our study, the LC assay targeting the T. gondii 529 bp repeat region (200300 copies per parasite) (Homan et al., 2000) revealed false-negative results in only two samples, whereas the LC assay targeting the T. gondii B1 gene (35 copies per parasite) (Burg et al., 1989) revealed false-negative results in four samples. Therefore, it appears likely that the discrepancies in PCR results are partly caused by different copy numbers of the target sequences. This is in agreement with a recent study on T. gondii-positive amniotic fluids, which showed that amplification of a sequence within the T. gondii 529 bp repeat region is at least 10 times more sensitive than targeting the T. gondii B1 gene (Reischl et al., 2003). The three patients for whom the conventional PCR and the LC assays gave discrepant results had non-specific symptoms compatible with toxoplasma reactivation.

In conclusion, real-time PCR allows rapid identification of T. gondii DNA in patient samples. However, discrepancies between real-time PCR and conventional nested PCR results occur in a minority of routine samples. Therefore, real-time and conventional PCR protocols for T. gondii should be optimized and carefully evaluated in patient samples before they are implemented as routine methods.

We thank C. Klotz and S. Richt for excellent technical assistance.