Blog
About

171
views
0
recommends
+1 Recommend
0 collections
    14
    shares
      • Record: found
      • Abstract: found
      • Article: found
      Is Open Access

      Development of an Aptamer-Based Concentration Method for the Detection of Trypanosoma cruzi in Blood

      Read this article at

      Bookmark
          There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.

          Abstract

          Trypanosoma cruzi, a blood-borne parasite, is the etiological agent of Chagas disease. T. cruzi trypomastigotes, the infectious life cycle stage, can be detected in blood of infected individuals using PCR-based methods. However, soon after a natural infection, or during the chronic phase of Chagas disease, the number of parasites in blood may be very low and thus difficult to detect by PCR. To facilitate PCR-based detection methods, a parasite concentration approach was explored. A whole cell SELEX strategy was utilized to develop serum stable RNA aptamers that bind to live T. cruzi trypomastigotes. These aptamers bound to the parasite with high affinities (8–25 nM range). The highest affinity aptamer, Apt68, also demonstrated high specificity as it did not interact with the insect stage epimastigotes of T. cruzi nor with other related trypanosomatid parasites, L. donovani and T. brucei, suggesting that the target of Apt68 was expressed only on T. cruzi trypomastigotes. Biotinylated Apt68, immobilized on a solid phase, was able to capture live parasites. These captured parasites were visible microscopically, as large motile aggregates, formed when the aptamer coated paramagnetic beads bound to the surface of the trypomastigotes. Additionally, Apt68 was also able to capture and aggregate trypomastigotes from several isolates of the two major genotypes of the parasite. Using a magnet, these parasite-bead aggregates could be purified from parasite-spiked whole blood samples, even at concentrations as low as 5 parasites in 15 ml of whole blood, as detected by a real-time PCR assay. Our results show that aptamers can be used as pathogen specific ligands to capture and facilitate PCR-based detection of T. cruzi in blood.

          Related collections

          Most cited references 36

          • Record: found
          • Abstract: found
          • Article: not found

          Development of a real-time PCR assay for Trypanosoma cruzi detection in blood samples.

          The aim of this study was to develop a real-time PCR technique to detect Trypanosoma cruzi DNA in blood of chagasic patients. Analytical sensitivity of the real-time PCR was assessed by two-fold serial dilutions of T. cruzi epimastigotes in seronegative blood (7.8 down to 0.06 epimastigotes/mL). Clinical sensitivity was tested in 38 blood samples from adult chronic chagasic patients and 1 blood sample from a child with an acute congenital infection. Specificity was assessed with 100 seronegative subjects from endemic areas, 24 seronegative subjects from non-endemic area and 20 patients with Leishmania infantum-visceral leishmaniosis. Real-time PCR was designed to amplify a fragment of 166 bp in the satellite DNA of T. cruzi. As internal control of amplification human RNase P gene was coamplified, and uracil-N-glycosylase (UNG) was added to the reaction to avoid false positives due to PCR contamination. Samples were also analysed by a previously described nested PCR (N-PCR) that amplifies the same DNA region as the real-time PCR. Sensitivity of the real-time PCR was 0.8 parasites/mL (50% positive hit rate) and 2 parasites/mL (95% positive hit rate). None of the seronegative samples was positive by real-time PCR, resulting in 100% specificity. Sixteen out of 39 patients were positive by real-time PCR (41%). Concordance of results with the N-PCR was 90%. In conclusion, real-time PCR provides an optimal alternative to N-PCR, with similar sensitivity and higher throughput, and could help determine ongoing parasitaemia in chagasic patients.
            Bookmark
            • Record: found
            • Abstract: found
            • Article: found
            Is Open Access

            International Study to Evaluate PCR Methods for Detection of Trypanosoma cruzi DNA in Blood Samples from Chagas Disease Patients

            Introduction A century after its discovery [1] Chagas disease still represents a health threat to an estimated 28 million people in the Americas, being the second highest illness burden among neglected tropical diseases [2]–[3]. The infection by the protozoan Trypanosoma cruzi can be acquired from blood-sucking triatomine bugs, blood transfusion, transplacental transmission or by the oral contamination foodstuffs by infected triatomine faeces [2]–[3]. Since 1990, a series of international initiatives based on vector control, systematic screening of blood donors in all endemic countries, and detection and treatment of congenital transmission have been launched for control and elimination of Chagas disease. These strategies have led to significant reduction in the number of infected people worldwide. According to information from 21 countries where the disease is endemic, the number of infected persons today is estimated to be 7,694,500, most of them at the chronic stage of disease [2]–[3]. Traditional parasitological procedures, such as xenodiagnosis and haemoculture are laborious and time-consuming and show poor sensitivities in cases of low-level parasitaemias, limiting their usefulness in diagnosis and monitoring of drug efficacy [4]–[6]. Since the past decade, the application of polymerase chain reaction (PCR) to detect T. cruzi directly in blood samples has opened new possibilities for the diagnosis of infection and evaluation of trypanocidal chemotherapy in different clinical and epidemiological settings [7]–[22]. These PCR procedures have revealed highly variable levels of sensitivity and specificity, depending on a number of technical factors such as, the volume of sample collected, the conditions of conservation of the sample, the methods used to isolate DNA, the parasite sequences and primers selected, the reagents used as well as the thermo-cycling conditions. Variability in PCR sensitivity could also be in part explained by the intermittent presence and quantity of circulating parasites at the time of blood collection. In addition, molecular targets from strains belonging to six different T. cruzi discrete typing units (DTUs, [23]) with dissimilar DNA content and gene dosage [22], [24]–[25] have been used for molecular diagnosis by different laboratories. In addition, sequence polymorphisms within amplified fragments among strains from different DTUs may influence the efficiency of amplification [26]–[27]. Moreover, false negative findings due to interference of PCR inhibitory substances co-purified during lysis and DNA extraction of blood samples and false positive results mostly due to carry over DNA contamination [22], [28] may arise. In this context, the assessment of the performances of currently available PCR tests for detection of T. cruzi infection in blood samples and DNA control sets was launched by expert laboratories in PCR detection of T. cruzi infection from different countries of America and Europe. We aimed to compare the performance of currently used PCR strategies for detection of T. cruzi DNA in sets of blind samples, including purified DNA from reference culture stocks from different T. cruzi discrete typing units, human blood samples spiked with cultured parasite cells and clinical samples from seropositive and seronegative patients from different endemic countries, in order to select the best performing tests for validation. Materials and Methods The participating laboratories were selected on the basis of their expertise in current processing of clinical samples for PCR detection of T. cruzi DNA, facilities with pre-PCR and post-PCR working areas and sufficient financial support to enable sustainability of PCR work after conclusion of this study. Twenty six Laboratories from different countries of America and Europe participated in this study, one laboratory from Belgium, Bolivia, France, French Guiana, Mexico, Paraguay, Peru, Spain, United Kingdom, Uruguay, USA, Venezuela, two laboratories from Chile, three from Colombia, four from Argentina and five from Brazil. Aiming to explore the highest extent of currently used PCR tests for detection of T. cruzi DNA, the participating laboratories were encouraged to carry out all available PCR strategies currently in use according to their own protocols and using their own financial resources (LbX1 to LbXN in Table 1, where LbX denotes laboratory and the number denotes a given test). 10.1371/journal.pntd.0000931.t001 Table 1 PCR tests reported by the participating Laboratories. Lb/Test Extraction Method Target Primer Names PCR Master Mix Cycles LbA Solvent extraction kDNA 121-122 Conventional In-House 35 LbB Solvent extraction kDNA S35 - S36 Conventional In-House 30 LbC/1 Solvent extraction kDNA S35 - S36 Conventional In-House 32 LbC/2 Solvent extraction Sat-DNA tcz1 - tcz2 Conventional In-House 40 LbC/3 Solvent extraction 24s D71-D71 Conventional In-House 40 LbC/4 Solvent extraction CO II-DNA Tcmit 31-40 Conventional In-House 48 Lb/C5 Solvent extraction CO II-DNA Nested Tcmit 10-21 Conventional In-House 48 Lb/C6 Solvent extraction SL-DNA Tcc- Tc1-Tc2 Conventional In-House 30 Lb/D1 Solvent extraction kDNA 121-122 Conventional In-House 36 Lb/D2 Solvent extraction Sat-DNA TczF-TczR Real Time Qiuantitect (Kt) 50 Lb/D3 Solvent extraction Sat-DNA TczF-TczR Conventional In-House 41 LbE Chelex Resine kDNA 121-122 Conventional In-House 35 LbF/1 Roche Silica gel col (Kt) Sat-DNA cruzi1-2 Real Time Roche (Kt) 45 LbF/2 Roche Silica gel col (Kt) kDNA 32f-148r Real Time Roche (Kt) 45 LbG/1 Qiagen DNeasy Tissue kit (Kt) kDNA FAM – IPC 32f-148r Real Time Apllied Biosystem (Kt) 55 LbG/2 Qiagen DNeasy Tissue kit (Kt) kDNA FAM 32f-148r Real Time Apllied Biosystem (Kt) 55 LbG/3 Qiagen DNeasy Tissue kit (Kt) kDNA VIC 32f-148r Real Time Apllied Biosystem (Kt) 55 LbG/4 Qiagen DNeasy Tissue kit (Kt) Sat-DNA cruzi1-2 Real Time Apllied Biosystem (Kt) 45 LbH/1 Favorgen Glass fibers col (Kt) kDNA 121-122 Conventional GoTaq (Kt) 33 LbH/2 Favorgen Glass fibers col (Kt) kDNA 121-122 Conventional In-House 33 LbI/1 Favorgen Glass fibers col (Kt) kDNA 121-122 Conventional In-House 40 LbI/2 Favorgen Glass fibers col (Kt) kDNA S35 - S36 Conventional In-House 40 LbJ Solvent extraction Sat-DNA Tcz1-Tcz2 Conventional In-House 40 LbK/1 Silica gel col (Kt) Sat-DNA cruzi1-2 Real Time In-House 40 LbK/2 Silica gel col (Kt) kDNA 121-122 Conventional In-House 40 LbL/1 Blood mini Kit (Kt) Sat-DNA cruzi1-2 Conventional In-House 40 LbL/2 Blood mini Kit (Kt) Sat-DNA Satellite DNA based kit Conventional OligoC-T Coris (Kt) 40 LbM Silica gel col (Kt) kDNA TC1-TC2 Conventional In-House 40 LbN/1 Solvent extraction kDNA 121-122 Conventional In-House 40 LbN/2 Solvent extraction Sat-DNA Tcz1-Tcz2 Conventional In-House 35 LbO Solvent extraction kDNA 121-122 Conventional In-House 40 LbP/1 Solvent extraction kDNA 121-122 Conventional In-House 35 LbP/2 CTAB (IH) kDNA 121-122 Conventional In-House 35 LbQ Solvent extraction kDNA 121-122 Conventional In-House 37 LbR Roche Silica gel col (Kt) kDNA 121-122 Conventional In-House 40 LbS/1 Qiagen Silica gel col (Kt) 18s Tc18s F3-R4 Conventional AmpliTaq Gold (Kt) 40 LbS/2 Qiagen Silica gel col (Kt) Sat-DNA cruzi1-2 Real Time Platinum qPCR (Kt) 40 LbS/3 Qiagen Silica gel col (Kt) 18s Tc18s F1042- R1144 Real Time Platinum qPCR (Kt) 40 LbS/4 Qiagen Silica gel col (Kt) kDNA 121-122 Conventional AmpliTaq Gold (Kt) 40 LbT ATGEN kit (Kt) kDNA 121-122 Real Time Invitrogen (Kt) 40 LbU/1 Solvent extraction kDNA 121-122 Conventional In-House 40 LbU/2 Solvent extraction 24s D71-D72 Conventional In-House 32 LbV/1 Silica gel col (Kt) kDNA 121-122 Conventional In-House 40 LbV/2 Silica gel col (Kt) Sat-DNA Tcz1-Tcz2 Conventional In-House 30 LbW Solvent extraction kDNA 121-122 Conventional In-House 40 LbX Solvent extraction kDNA 121-122 Conventional In-House 35 LbY Solvent extraction kDNA 121-122 Conventional In-House 35 LbZ Silica gel col (Kt) Sat-DNA cruzi1-2 Real Time TaqMan Univ (Kt) 45 LbX/1-6, Laboratory and test identification, kDNA, minicircle DNA; Sat-DNA, satellite DNA; 24s, 24sa rDNA; 18s, 18s rDNA; SL, Spliced Leader; kDNA FAM, kDNA TaqMan probe labeled with FAM; kDNA VIC, kDNA TaqMan probe labeled with VIC; IPC, TaqMan Exogenous Internal Positive Control (Applied Biosystems). The organizing laboratory (LabMECh, INGEBI, Buenos Aires) was in charge of preparing characterised samples in three different sets (A, B and C), as described below. Set A. This set consisted of ten-fold serial dilutions of T. cruzi DNA, plus three negative controls without DNA in bi-distilled sterile water. T. cruzi DNA was purified from epimastigote cells grown in LIT medium from stocks Silvio X10, Cl-Brener and CAN III, which are references for the discrete typing units T. cruzi I (DTU I), T. cruzi VI (DTU IIe) and T. cruzi IV (DTU IIa), respectively [23], [29]. The identity of the DTUs was confirmed using a PCR algorithm targeting several nuclear genes, as detailed in Burgos and coworkers [30]. T. cruzi DNA was extracted from parasite cultures using current chloroform- DNA extraction without vortexing during the procedure [31]. The concentration and quality of DNA was measured at 260/280 nm in triplicate, using a Nanodrop 1000 spectrophotometer (ThermoFisher Scientific, Waltham, MA, USA). Each series of DNA samples was conformed by concentrations ranging from 10 fg/ul to 10−3 fg/ul. Set B. This set contained seronegative human blood samples treated with Guanidine Hidrochloride 6M-EDTA 0.2 M buffer, pH 8.00 [32] and spiked with ten-fold dilutions of cultured CL-Brener epimastigotes. One Guanidine Hidrochloride-EDTA treated blood sample without parasites was included as negative control. The samples were prepared as follows: a pool of human blood samples testing negative for T. cruzi infection in current serological methods and PCR was mixed with an equal volume of Guanidine Hidrochloride-EDTA buffer. An aliquot was withdrawn and spiked with 5 parasite cells/mL, homogenized, let stand overnight at room temperature, boiled the following day during 15 minutes [33], let stand at room temperature overnight and then stored at 4°C. The remaining non-spiked blood was treated in the same way and stored as a negative stock. One day later, the spiked Guanidine Hidrochloride-EDTA blood was used as starter for preparing 5 ten-fold serial dilutions, using as matrix the negative blood stock, to obtain samples ranging from 0.5 to 5.10−5 parasite equivalents/mL of blood. Set C. This was a panel of 42 pre-characterized archived clinical blood samples stored in Guanidine Hidrochloride-EDTA buffer, including 10 from seronegative patients and 32 from seropositive patients from endemic regions of Argentina, Bolivia, Brazil and Paraguay. The seropositive panel was composed by patients at different phases of T. cruzi infection, namely, 2 immunosuppressed patients after heart transplantation, 23 indeterminate Chagas disease and 7 chronic Chagas disease patients with cardiac and/or digestive manifestations. They were selected from archived collections from the Serodiagnostic Laboratory for Chagas Disease, Federal University of Goias, Brazil, and from the organizing Lab. Diagnosis was based on their serological reactivity by at least two out of three routine serological methods (ELISA, IHA and Latex Agglutination tests) and clinical and electrocardiographic findings. Samples from patients were obtained with written informed consent and approval of the Ethics Committee of the Rivadavia Hospital, Government of Buenos Aires city, Argentina and the Serodiagnostic Laboratory for Chagas Disease, Federal University of Goias, Goiania, Brazil. Furthermore, all samples were tested by two PCR tests performed on duplicate at the organizing laboratory, namely a hot-start PCR targeting kDNA according to Burgos et al [30], a Real time PCR targeting satellite DNA sequences and a Real time PCR targeting an internal amplification control, according to Duffy and coworkers [22], allowing confirmation of PCR negativity among seronegative samples and PCR positivity among a subgroup of the 32 seropositive samples (data not shown). Each sample from set A, B and C was aliquoted and distributed into 1 ml Screw Top bar-coded tubes (Matrix Trackmates, UNITEK, USA) to each package. The packages were sent refrigerated to the participating laboratories (World Courier, Arg). Each laboratory received 50 µls of Set A and 500 µls of samples belonging to sets B and C. Best Performing PCR Methods DNA extraction Methods LbD2, LbD3 and LbQ: Solvent DNA extraction was carried out from 100 µl of Guanidine Hidrochloride-EDTA blood aliquots. Briefly, 100 µl aliquots were taken and well mixed with 100 µ l of phenol-chloroform-isoamylic alcohol (25∶24∶1) (phenol Tris–EDTA pH 8, USB Corporation, USA). After centrifugation for 3 min at 13000 rpm 150 µl of distilled water were added. The solution was mixed and centrifuged for 3 min at 13000 rpm. The aqueous phase was transferred to a clean tube, and a final extraction with 200 µ l of chloroform was performed. After centrifugation for 3 min at 13000 rpm the aqueous phase was transferred to a clean tube and mixed with 40 µg of rabbit liver glycogen (Sigma, USA). The DNA was precipitated with 200 µl of isopropyl alcohol during 30 minutes at −20°C. Then the solution was centrifuged at 13000 rpm for 15 min. The pellet was washed with 500 µl of 70% ethanol and centrifuged again 15 min at 13000 rpm. After discard the ethanol the pellet was allowed to dry during 10 min at 37°C. Finally the pellet was suspended in 50 µl 10 mM Tris-HCl, pH 8.5. DNA solution was stored at −20°C. Method LbF1: DNA isolation used a commercial kit (High Pure PCR Template preparation kit, Roche Applied Science) according to the manufacturer's protocol. DNA solution was stored at −20°C. DNA amplification Method LbD3 was carried out in a MJR PTC-100 thermocycler (MJ Research, Watertown, MA, USA). Master mix was composed by 1X Taq platinum amplification buffer, 250 µM deoxynucleotide triphosphate solution (dNTPs), 3 mM MgCl2 solution, 1,5 U Taq Platinum (Invitrogen, Brazil), 0.5 µM sat-DNA specific primers TCZ-F (GCTCTTGCCCACAMGGGTGC) and TCZ-R (CCAAGCAGCGGATAGTTCAGG), 5 µl of template DNA and a quantity of water sufficient to give a final volume of 50 µl. Cycling parameters were one step of 3 min at 94°C; 40 cycles of 45 sec at 94°C, 1 min at 68°C and 1 min at 72°C and one final extension step of 10 min at 72°C, 182 bp Sat-DNA PCR products were analysed in 3% agarose gels (Invitrogen, Life Technologies, USA) stained with ethidium bromide. Method LbQ was carried out in a MJR PTC-100 thermocycler (MJ Research, Watertown, MA, USA). Master mix was composed by 1X Taq platinum amplification buffer, 200 µ M dNTPs, 3 mM MgCl2 solution, 1,5 U Taq Platinum (Invitrogen, Brazil), 10 µM kDNA specific primers 121 (AAATAATGTACGGGKGAGATGCATGA) and 122 (GGTTCGATTGGGGTTGGTGTAATATA), 7.5 µl of template DNA and a quantity of water sufficient to give a final volume of 50 µl. Cycling parameters were one step of 3 min denaturation at 94°C; 2 cycles of 1 min at 97.5°C, 2 min at 64°C; 33 cycles of 1 min at 94°C, 1 min at 62°C and one final extension step of 10 min at 72°C, 330 bp kDNA PCR products were analysed in 2% agarose gels stained with ethidium bromide. Method LbD2 was conducted using a Rotor Gene 3000 (Corbett Research, Sydney, Australia) Real Time thermocycler. Each PCR reaction contained 1X Qiagen QuantiTect Sybr-Green PCR Master Mix (Qiagen), 0.5 µM SatDNA specific primers TCZ-F (GCTCTTGCCCACAMGGGTGC) and TCZ-R (CCAAGCAGCGGATAGTTCAGG), 2 µl of template DNA and PCR-grade H2O (Qiagen) to a final volume of 20 µl. The amplification was conducted under the following cycling conditions after 15 min of denaturation at 95°C, PCR amplification was carried out for 50 cycles (95°C for 10 s, 55°C for 15 s and 72°C for 10 s). Fluorescence data collection was performed at 72°C at the end of each cycle. After quantification, a melt curve was made with 74–85°C raising by 0.5°C each step and waiting for 4 seconds afterwards acquiring on Green channel. Melting temperture (Tm) of the amplicon was 81°C. Finally, data were analyzed with Rotor-Gene 6000 Series Software 1.7 (Corbett Research). Method LbF1 was conducted using a Rotor Gene 3000 (Corbett Research, Sydney, Australia) Real Time thermocycler. Each PCR reaction contained 1X PCR FastStart Universal Probe Master Master Mix (Roche), 0.75 µM SatDNA specific primers cruzi 1 (ASTCGGCTGATCGTTTTCGA) and cruzi 2 (AATTCCTCCAAGCAGCGGATA), 0.25 µM SatDNA specific probe cruzi 3 (CACACACTGGACACCAA), 2 µl of template DNA and PCR-grade H2O to a final volume of 20 µl. The amplification was conducted under the following cycling conditions, after 15 min of denaturation at 95°C, 45 cycles at 95°C for 10 s, 54°C for 60 s. Fluorescence data collection was performed at 54°C at the end of each cycle. Finally, data were analyzed with Rotor-Gene 6000 Series Software 1.7 (Corbett Research). The possibility of contamination of the PCR reagents and of the solutions used to prepare DNA was carefully examined through the use of appropriate controls. Also two dilutions from DNA purified from Cl- Brener strain were analyzed in each round as strong positive and detection limit control, respectively. Data Analysis An access database form was distributed to the participants to standardize reporting of results. Those laboratories performing more than one PCR test per sample sent a separate report for each test. The results were analyzed by using SAS Software and Microsoft Excel. Due to the exploratory nature of the study, a descriptive analysis of results is provided. For set A, the following parameters were evaluated: 1) specificity (Sp): the proportion of negative PCR results in the three negative samples, 2) coherence: (Co) the ability of reporting positive PCR findings in a consecutive way, from the highest to the lowest detected DNA concentration for each series of DNA dilutions of parasite stocks and 3) the detection limits (DL) for each stock. A test was defined as Good Performing Method (GPM) if it was 100% specific and coherent and capable of detecting 10 fg/ul or less DNA for all parasite DTU stocks. For set B the same parameters were evaluated: Sp, Co and DL. A test was defined as GPM if it was 100% specific and coherent and capable of detecting 5 parasite equivalents/mL of Guanidine Hidrochloride-EDTA treated blood or less. For each sample of set C, a consensus PCR result was obtained on the basis of the reports by GPM tests in sets A and B, as done in other PCR interlaboratory studies [34]. A sample was considered PCR positive by consensus if more than 50% of the GPM gave positive results and PCR negative if more than 50% of GPM tests gave negative results. Those samples for which 50% of the GPM methods gave positive reports and 50% gave negative ones were considered indeterminate. The sensitivity, specificity, accuracy and kappa index of the different PCR tests were calculated by using 1) the above mentioned consensus PCR results and 2) the serological diagnosis as the reference methods. Inter-observer kappa coefficients were calculated using GraphPad Software on-line statistical calculators (http://www.graphpad.com/quickcalcs/kappa1. cfm). Kappa values<0.01 indicate no concordance, those between 0.1 and 0.4 indicate weak concordance, those between 0.41 and 0.60 indicate clear concordance, those between 0.61 and 0.80 indicate strong concordance, and those between 0.81 and 1.00 indicate nearly complete concordance. Accuracy was calculated as reported [35]. Results Twenty six laboratories reported PCR results, using one to six different PCR tests (Table 1). The main sources of variability among laboratories and tests included DNA purification procedures using commercial kits (Kt) or in-house methods (IH), T. cruzi target and primer sequences, cycling instrumentation by conventional (C) or Real Time (RT) thermocyclers, cycling conditions, master mix compositions and trade marks of PCR kits (Kt) or reagents for IH master mixes. A total of 48 PCR tests were reported for set A samples and 44 of them for sets B and C. Twenty eight tests targeted minicircle DNA, 24 of them amplified the 330 bp variable region and 4 amplified a 118 bp fragment from the constant region (Lb F2 and Lb G1 to G3, Table 1). Thirteen tests targeted the satellite DNA sequence (Sat-DNA), two targeted the 18s ribosomal RNA genes (18s rDNA), two amplified a fragment from the 24sα ribosomal RNA genes (24sα rDNA), one targeted the intergenic region of spliced-leader genes (SL-DNA) and two the mitochondrial gene for the subunit II of cytochrome oxidase (CO II-DNA) (Table 1). Analysis of Set A Table 2 shows the data obtained by the 48 PCR tests on DNA dilutions from the 3 parasite stocks representing DTUs I, IV and VI. The seven PCR tests targeting sequences other than Sat-DNA or kDNA failed to detect the most concentrated DNA sample (10 fg/ul) of one, two or all three parasite stocks (Tests C3, C4, C6, S1 and S3; Detection Limit  =  ND), or reported false positive findings in the negative controls without DNA (Tests C5 and U2) and thus were not included in the following analysis. 10.1371/journal.pntd.0000931.t002 Table 2 Performances of PCR tests in Sets A and B. Set A Set B GPM T. cruzi I T. cruzi IV T. cruzi VI T. cruzi VI Lb/Test Sp Co DL Co DL Co DL Sp Co DL par/ml LbA Y Y 0.1 N 0.01 N 0.01 Y N 0.005 N LbB Y N 0.001 N 0.001 Y 0.1 Y N 0.005 N LbC/1 N N 0.001 N 0.001 Y 0.1 N Y ND N LbC/2 Y Y ND N 1 Y 10 Y Y 0.00005 N LbC/3 Y Y ND Y ND Y ND NA NA NA N LbC/4 Y Y ND Y ND Y ND NA NA NA N Lb/C5 N Y 1 N 0.1 N 0.001 NA NA NA N Lb/C6 Y Y ND Y ND Y ND NA NA NA N Lb/D1 Y Y 1 Y 10 Y 1 N N 0.005 N Lb/D2 Y Y 1 Y 10 Y 1 Y Y 0.05 Y Lb/D3 Y Y 1 Y 10 Y 1 Y Y 0.05 Y LbE Y Y 1 Y 10 Y 10 Y Y 0.005 Y LbF/1 Y Y 0.1 Y 1 Y 0.01 Y Y 0.05 Y LbF/2 Y Y 1 N 0.01 Y 1 Y Y 0.5 N LbG/1 Y Y 0.1 Y 1 Y ND Y Y 0.05 N LbG/2 Y Y 0.1 Y 1 Y 1 Y Y 0.05 Y LbG/3 Y Y 0.1 Y 1 Y 1 Y Y 0.05 Y LbG/4 Y Y 1 Y 1 Y 1 Y Y 0.5 Y LbH/1 Y Y 1 Y 10 N 0.001 Y N 0.05 N LbH/2 Y Y 1 N 0.1 Y 10 Y N 0.05 N LbI/1 Y Y 1 N 0.001 Y 10 Y Y 0.005 N LbI/2 Y Y 1 Y 10 Y 10 Y Y 0.05 Y LbJ Y Y 0.01 N 0.001 N 0.001 Y N 0.5 N LbK/1 Y Y 10 Y 10 Y 10 Y Y 0.5 Y LbK/2 Y Y 1 Y 10 Y 10 Y Y 5 N LbL/1 Y Y ND Y 10 Y 1 Y Y 0.5 N LbL/2 Y Y ND Y ND Y 1 Y Y 0.5 N LbM N Y 0.001 Y 0.001 N 0.001 Y Y ND N LbN/1 Y Y 0.1 Y ND N 0.1 Y N 0.005 N LbN/2 Y Y 1 Y ND Y 10 Y Y 0.5 N LbO Y Y 10 N 1 Y ND Y Y 0.05 N LbP/1 Y Y 0.1 Y 10 Y 1 Y Y 5 N LbP/2 Y Y 0.1 Y 10 Y 1 Y Y 0.5 N LbQ Y Y 1 Y 10 Y 1 Y Y 0.5 N LbR Y Y 0.1 Y 1 Y 0.1 Y N 0.00005 N LbS/1 Y Y 1 Y ND Y ND Y Y ND N LbS/2 Y Y 1 Y 10 Y 1 Y Y 0.5 N LbS/3 Y Y 10 Y 10 Y ND Y Y ND N LbS/4 Y Y 1 Y 1 Y 10 N N 0.00005 N LbT N N 0.001 Y 1 N 0.01 N N 0.05 N LbU/1 N Y 0.001 Y 0.001 Y 0.001 Y N 0.005 N LbU/2 N Y 0.001 Y 0.001 Y 0.001 N N 0.005 N LbV/1 Y Y 0.1 Y 10 Y 10 Y Y 0.05 N LbV/2 Y Y 0.1 Y ND Y 10 Y Y 0.05 N LbW Y Y 0.01 Y 1 Y 0.1 Y Y 0.005 N LbX N N 0.01 N 0.001 N 0.1 Y N 0.0005 N LbY N Y 1 Y 1 Y 0.1 Y Y ND N LbZ Y Y 1 Y 1 Y 0.01 N N 0.0005 N LbX/1-6, Laboratory and test identification; Bold Type, Good Performing Methods in sets A or B; GPM, Good Performing Methods in sets A and B; Sp, 100% of specificity in all controls without T. cruzi DNA; Co, Coherence in PCR positive reports; DL, Detection limit in fg DNA/ul; Y, Affirmative; N, Negative; NA, Not available; ND, Not detectable. Out of the 41 tests based on kDNA (28 tests) or Sat-DNA sequences (13 tests), 25 (51.2%) provided specific and coherent results for all three parasite stocks (Sp = Y, Co = Y, Table 2). Fourteen of them targeted kDNA, representing 50% of the reported kDNA-PCR tests and 11 targeted Sat-DNA, representing 84.6% of Sat-DNA PCR tests. These data indicated that PCR tests based on Sat-DNA sequences were more specific than those based on kDNA. Figures 1A and 1B show the distribution of the detection limits (DL) of the above mentioned 11 Sat-DNA PCR and 14 kDNA PCR tests, respectively, for each T. cruzi stock. 10.1371/journal.pntd.0000931.g001 Figure 1 Analytical Sensitivity of specific and coherent PCR tests in sets A and B. Distribution of detection limits (DL) of specific and coherent PCR tests targeted to Sat-DNA (A) and kDNA sequences (B) for detecting serial dilutions of purified DNA from 3 parasite stocks (Set A) representative of T. cruzi DTU I (Silvio×10), DTU IV (Can III cl1) and DTU VI (Cl Brener). C. Distribution of detection limits (DL) of specific and coherent PCR tests targeted to Sat-DNA (black bars) and kDNA sequences (white bars) carried out from human blood spiked with serial dilutions of parasite cells (Set B). Analysis of T. cruzi I DNA series: Nine out of 11 Sat-DNA- and all 14 kDNA-PCR tests were capable of detecting at least the most concentrated T. cruzi I DNA sample (Figure 1 A and B, grey bars) and 2 Sat-DNA and 8 kDNA-PCR tests could detect 0.1 fg/µl of T. cruzi I DNA. The lowest detection limit for T. cruzi I DNA was 0.01 fg/ul obtained by laboratory W using conventional kDNA-PCR (Table 2 and Figure 1B). Thus, kDNA- PCR tests were more sensitive than Sat-DNA PCR tests to detect T. cruzi I DNA. Analysis of T. cruzi IV DNA series: 8 out of 11 Sat-DNA- and all 14 kDNA-PCR tests were capable of detecting the most concentrated T. cruzi IV DNA sample (Figure 1 A and B, black bars). The lowest detection limit (1 fg/µl) was reached by three Sat-DNA- and six kDNA-PCR tests, suggesting similar analytical sensitivities of methods based on both molecular targets to detect T. cruzi IV DNA. Analysis of T. cruzi VI DNA series: All 11 Sat-DNA- and 13 out of 14 kDNA-PCR tests were capable of detecting the most concentrated T. cruzi VI DNA sample (Figure 1 A and B, white bars). The only test that did not detect Cl-Brener DNA amplified the constant kDNA region (G1, kDNAc, Table 2). The lowest detection limit (0.01 fg/µl) was obtained by 2 Sat-DNA PCR tests (Z and F1, Table 2) followed by 0.1 fg/µl obtained by 2 conventional kDNA-PCR tests (R and W, Table 2). Overall, the reported PCR tests were less sensitive for detecting DNA from the T. cruzi IV reference stock. Twenty PCR tests showing specific and coherent results and detecting at least the most concentrated DNA samples from each of the parasite stocks were considered Good Performing Methods for Set A (bold fonts, Table 2). They comprised 53.8% of 13 Sat-DNA-PCR and 46.42% of 28 kDNA-PCR tests. Ten GPM tests used in-house (IH) PCR mixtures and 10 used commercial master mixes (Kt), representing 35.7% of the 28 IH and 76.9% of the 13 Kt PCR reagent mixes. In addition, 12 GPM used conventional amplification and eight used real time PCR (C and RT in Table 2), representing 38.7% of 31 C and 72.7% of 11 RT tests. These data showed that commercial master mixes and real time PCR offered better PCR performance in purified DNA samples. Analysis of Set B Out of the 44 PCR tests reported for spiked Guanidine Hidrochloride-EDTA blood samples, the three tests targeting sequences other than Sat-DNA or kDNA were not further analyzed, because they failed to detect the most concentrated sample (S1 and S3 tests) or showed false positive findings in the non-spiked control (U2 test) (Set B, Table 2). Twenty five out of 41 PCR tests based on kDNA and Sat-DNA sequences showed specific, coherent results and detection limits of at least 5 par/ml (GPM, bold fonts, Table 2, Set B). They included 14 kDNA and 11 Sat-DNA PCR tests, representing 50% of 28 kDNA and 84.6% of 13 Sat-DNA based tests. Ten GPM used in-house extraction methods and 15 used DNA extraction kits, representing 41.6% of 24 IH and 62.5% of 24 Kt tests. Thus, methods using commercial DNA extraction and Sat-DNA as amplification target resulted in better performance. Procedures based on kDNA presented more variation in sensitivity than Sat-DNA tests (Figure 1C, white bars). The smallest detected concentration was 5×10−3 par/ml, recorded by three laboratories using conventional kDNA-PCR after DNA extractions with Chelex resine, a blood extraction kit or solvent extraction with Phenol (LbE, LbL1 and LbW, respectively, Table 1). Tests based on Sat-DNA presented sensitivities between 0.05 and 0.5 par/ml (10/11 tests, Figure 1C, black bars) with the only exception of one test based on solvent DNA extraction and IH conventional Sat-DNA PCR (C2, Table 1) that reached a detection limit of 5×10−5 par/mL (Table 2, and Figure 1C). Analysis of Set C Out of the 44 PCR tests performed on clinical samples of Set C, a 18s-rDNA PCR (S3) and a SL-DNA PCR (C6) tests did not detect any positive sample and the 24s α rDNA-PCR test (U2) had only 40% of specificity. Consequently, they were not included for subsequent analysis. The levels of agreement among the 41 remaining PCR tests on the reports for each clinical sample are presented in Table 3. For each sample, 3 series of consensus PCR results were calculated: 1) consensus based on the 28 kDNA-PCR tests, 2) consensus based on the 13 Sat-DNA PCR tests and 3) consensus based on the 16 tests defined as GPM in both sets A and B samples. The sensitivity of consensus kDNA-PCR was 65.62% (21 PCR positive/32 seropositive samples), that of consensus Sat-DNA was 62.5% (20 PCR positive/32 seropositive samples) and that of consensus GPM was 56.25% (B (18 PCR positive/32 seropositive samples) being 4 samples indeterminate (13, 18, 20, 21, Table 3) because the levels of agreement among GPM tests was 50%. 10.1371/journal.pntd.0000931.t003 Table 3 Concordance of PCR results reported for each clinical case of Set C. CLINICAL CASES kDNA PCR n = 28 Sat-DNA PCR n = 13 GPM n = 16 ID G Ag Status Region EN pos/tot % Cons pos/tot % Cons pos/tot % Cons 1 F NA cChHD-HTx Arg- Uk1 Uk 26 92,9 P O S 12 92,3 P O S 15 93,8 P O S 2 M NA cChHD-HTx Arg-Chaco1 Yes 26 92,9 P O S 10 76,9 P O S 15 93,8 P O S 3 F 54 Mega III cChH Br- MG Yes 20 71,4 P O S 12 92,3 P O S 12 75,0 P O S 4 F 42 Pregnant Arg- Salta2 Yes 18 64,3 P O S 12 92,3 P O S 14 87,5 P O S 5 F 25 Pregnant Bo-Uk Yes 18 64,3 P O S 11 84,6 P O S 12 75,0 P O S 6 M 20 Blood donor Br-BA Yes 19 67,9 P O S 10 76,9 P O S 10 62,5 P O S 7 F 41 cChHD Br-BA Yes 18 64,3 P O S 9 69,2 P O S 11 68,8 P O S 8 F 31 Pregnant Bo-Uk Yes 16 57,1 P O S 10 76,9 P O S 11 68,8 P O S 9 F NA Pregnant Par-Uk2 Yes 15 53,6 P O S 11 84,6 P O S 12 75,0 P O S 10 F 22 Ex-pregnant Br-BA Yes 17 60,7 P O S 10 76,9 P O S 12 75,0 P O S 11 F 41 Chronic CD Br-BA Yes 16 57,1 P O S 9 69,2 P O S 11 68,8 P O S 12 F 24 Ex-pregnant Br-Go Yes 15 53,6 P O S 9 69,2 P O S 10 62,5 P O S 13 F 32 Pregnant Arg-Co Yes 16 57,1 P O S 8 61,5 P O S 8 50,0 I N D 14 F 35 Ex-pregnant Br-Ceara Yes 17 60,7 P O S 9 69,2 P O S 12 75,0 P O S 15 F 47 cChHD Br-Go Yes 17 60,7 P O S 8 61,5 P O S 9 56,3 P O S 16 F NA Pregnant Par-Uk Yes 14 50,0 P O S 8 61,5 P O S 9 56,3 P O S 17 M 55 CD Br-MG Yes 15 53,6 P O S 7 53,8 P O S 10 62,5 P O S 18 M 33 cChHD Br-BA Yes 15 53,6 P O S 7 53,8 P O S 8 50,0 I N D 19 F 66 Mega II + CBBB Br-BA Yes 15 53,6 P O S 5 38,5 N E G 9 56,3 P O S 20 F 18 Ex-pregnant Br-Go Yes 16 57,1 P O S 4 30,8 N E G 8 50,0 I N D 21 F 18 Pregnant Arg-Sg Yes 15 53,6 P O S 6 46,2 N E G 8 50,0 I N D 22 F 43 Indeterminate CD Br-BA Yes 13 46,4 N E G 7 53,8 P O S 6 37,5 N E G 23 F 57 Blood donor Br-Piaui Yes 11 39,3 N E G 8 61,5 P O S 9 56,3 P O S 24 F 46 Blood donor Br-BA Yes 7 25,0 N E G 6 46,2 N E G 6 37,5 N E G 25 F 25 Pregnant Par-Uk2 Yes 10 35,7 N E G 5 38,5 N E G 5 31,3 N E G 26 F 32 Pregnant Par-Uk2 Yes 11 39,3 N E G 5 38,5 N E G 6 37,5 N E G 27 F 36 Pregnant Arg-Chaco Yes 11 39,3 N E G 5 38,5 N E G 5 31,3 N E G 28 F 36 Pregnant Arg-Chaco Yes 9 32,1 N E G 4 30,8 N E G 2 12,5 N E G 29 M 59 cChHD Br-Piaui Yes 9 32,1 N E G 2 15,4 N E G 5 31,3 N E G 30 F 29 Mega-II Br-Go Yes 7 25,0 N E G 4 30,8 N E G 5 31,3 N E G 31 F NA Pregnant Par-Uk2 Yes 9 32,1 N E G 1 7,7 N E G 1 6,3 N E G 32 F 28 Pregnant Arg-Sg Yes 2 7,1 N E G 3 23,1 N E G 1 6,3 N E G 33 M 38 Routine Br-Go No 4 14,3 N E G 1 7,7 N E G 1 6,3 N E G 34 M 51 Routine Br-Uk Yes 5 17,9 N E G 1 7,7 N E G 1 6,3 N E G 35 M NA Blood donor Arg-BAs No 6 21,4 N E G 1 7,7 N E G 1 6,3 N E G 36 F 39 Routine Br-Go Yes 8 28,6 N E G 1 7,7 N E G 1 6,3 N E G 37 F 37 Routine Br-Go No 6 21,4 N E G 1 7,7 N E G 2 12,5 N E G 38 M NA Blood donor Arg-BA No 7 25,0 N E G 2 15,4 N E G 3 18,8 N E G 39 F 36 Routine Br-Bh Yes 7 25,0 N E G 4 30,8 N E G 3 18,8 N E G 40 F 40 Routine Br-Go Yes 8 28,6 N E G 4 30,8 N E G 5 31,3 N E G 41 F 40 Routine Br-Go Yes 9 32,1 N E G 6 46,2 N E G 5 31,3 N E G 42 F 58 Routine Br-Go Yes 11 39,3 N E G 6 46,2 N E G 6 37,5 N E G Patients 1 to 32 are seropositive and 33 to 42 seronegative. 28 kDNA tests and 13 Sat DNA tests were performed for each sample. kDNA, minicircle DNA; Sat-DNA, satellite DNA; GPM Good performing Methods in panels A and B; ID, sample identification number; G, Gender; Ag, age in years; EN, Endemic precedence; %: Percentage of positive results; Cons, Consensus PCR result; F, female; M, male; NA, not available; 28 kDNA tests and 13 Sat DNA tests were performed for each sample. 1 T.cruzi DTU I, 2 T.cruzi DTU II/V/VI, NE, not endemic; Uk, Unknown; Pos, positive consensus; Ind, indeterminate consensus; Neg, negative consensus; cChHD, chronic Chagas heart disease, Mega Megacolon, CBBB, Complete Branch Bundle Blockage, HTx, Heart transplantation; Arg: Argentina; Bo: Bolivia; Br: Brazil; Par: Paraguay; BAs, Buenos Aires; Bh, Bahia; Go, Goias; MG: Minas Gerais; Sg: Santiago del Estero. The individual performance of the 41 PCR tests was evaluated in comparison with the consensus PCR results reached by the 16 GPM in sets A and B (18 PCR positive, 20 PCR negative samples) and in comparison with serologic diagnosis (10 seronegative, 32 seropositive samples) (Table 4). There was a high variability among the performances of the different methods (Table 4). The median values of sensitivity, specificity and accuracy of the 41 tests were 72, 77.5 and 68.4%, respectively in comparison to consensus GPM PCR reports, and 59.4, 70 and 59.5%, respectively in comparison to serological diagnosis (Table 4). 10.1371/journal.pntd.0000931.t004 Table 4 Performance of PCR tests in comparison to consensus GPM reports and serodiagnosis. Test PCR performance vs consensus GPM K+S PCR performance versus Serology Lb/Test PCR Target Se Sp Acc kappa Se Sp Acc kappa BPM N = 18 N = 20 N = 38 N = 38 N = 32 N = 10 N = 42 N = 38 LbA C K 33.3 60.0 47.4 −0.1 31 70 40.5 0.0 N LbB C K 72.2 35.0 52.6 0.1 66 30 57.1 0.0 N LbC/1 C K 0.0 100.0 52.6 0.0 0 100 23.8 0.0 N LbC/2 C S 66.7 15.0 39.5 −0.2 69 10 54.8 −0.2 N Lb/D1 C K 94.4 45.0 68.4 0.4 81 40 71.4 0.2 N Lb/D2 RT S 94.4 85.0 89.5 0.8 69 100 76.2 0.5 Y Lb/D3 C S 94.4 85.0 89.5 0.8 63 100 71.4 0.4 Y LbE C K 94.4 65.0 78.9 0.6 81 80 81.0 0.5 N LbF/1 RT S 83.3 95.0 89.5 0.8 63 100 71.4 0.4 Y LbF/2 RT K 72.2 90.0 81.6 0.6 53 90 61.9 0.3 N LbG/1 RT K 100.0 60.0 78.9 0.6 84 60 78.6 0.5 N LbG/2 RT K 100.0 65.0 81.6 0.6 78 40 69.0 0.4 N LbG/3 RT K 100.0 65.0 81.6 0.6 78 40 69.0 0.4 N LbG/4 RT S 94.4 90.0 92.1 0.8 63 60 61.9 0.4 N LbH/1 C K 27.8 80.0 55.3 0.1 22 80 35.7 0.0 N LbH/2 C K 22.2 80.0 52.6 0.0 16 80 31.0 0.0 N LbI/1 C K 83.3 40.0 60.5 0.2 78 50 71.4 0.3 N LbI/2 C K 38.9 40.0 39.5 −0.2 53 40 50.0 −0.1 N LbJ C S 55.6 60.0 57.9 0.2 59 70 61.9 0.2 N LbK/1 RT S 61.1 70.0 65.8 0.3 44 60 47.6 0.3 N LbK/2 C K 0.0 100.0 52.6 0.0 0 100 23.8 0.0 N LbL/1 C S 88.9 45.0 65.8 0.3 84 60 78.6 0.4 N LbL/2 C S 83.3 60.0 71.1 0.4 72 60 69.0 0.3 N LbM C K 66.7 50.0 57.9 0.2 59 50 57.1 0.1 N LbN/1 C K 66.7 80.0 73.7 0.5 47 60 50.0 0.0 N LbN/2 C S 72.2 80.0 76.3 0.5 47 70 52.4 0.1 N LbO C K 66.7 55.0 60.5 0.2 47 30 42.9 −0.2 N LbP/1 C K 88.9 85.0 86.8 0.7 53 80 59.5 0.2 N LbP/2 C K 11.1 100.0 57.9 0.1 6 100 28.6 0.0 N LbQ C K 83.3 90.0 86.8 0.7 63 100 71.4 0.4 Y LbR C K 88.9 55.0 71.1 0.4 81 70 78.6 0.5 N LbS/2 RT S 50.0 90.0 71.1 0.4 38 100 52.4 0.2 N LbS/4 C K 55.6 90.0 73.7 0.5 47 100 59.5 0.3 N LbT RT K 50.0 75.0 63.2 0.3 41 80 50.0 0.1 N LbU/1 C K 16.7 95.0 57.9 0.1 9 90 28.6 0.0 N LbV/1 C K 27.8 100.0 65.8 0.3 16 100 35.7 0.1 N LbV/2 C S 44.4 100.0 73.7 0.5 28 100 45.2 0.2 N LbW C K 100.0 35.0 65.8 0.3 91 40 78.6 0.3 N LbX C K 100.0 50.0 73.7 0.5 88 60 81.0 0.5 N LbY C K 77.8 50.0 63.2 0.3 75 80 76.2 0.5 N LbZ RT S 50.0 90.0 71.1 0.4 38 100 52.4 0.2 N Median 72 77.5 68.4 0.4 59.4 70.0 59.5 0.2 (25-75p) (50–88.9) (55–90) (57.9–781) (0.2–0.6) (37.5–75) (60–100) (47.6–71.4) (0–0.4) LbX/1-6, Laboratory and test identification; BPM, Best Performing Methods; Consensus GPM K + S: consensus findings of GPM by kDNA and Satellite DNA PCRs; C, Conventional PCR, RT, Real Time PCR; K, kDNA; S, Satellite DNA; Se, sensitivity; Sp, specificity; Acc, accuracy; kappa, kappa index; N, negative; Y, affirmative; 25–75p, 25th-75th percentiles; Bold type, Good Performing Methods (GPM) in sets A and B. Best Performing Methods Four GPM showed the best operational parameters in set C (Table 4). Tests LbD2 and LbD3 used solvent DNA extraction followed by conventional hot-start and Real time PCR targeted to Sat-DNA, respectively (primer pairs TCZ-F/TCZ-R). Test LbF1 used a commercial kit for DNA extraction based on glass fiber columns and Real Time PCR targeted to Sat-DNA (primer pairs cruzi 1/cruzi 2 and TaqMan probe cruzi 3) and test LbQ used solvent DNA extraction and conventional hot-start PCR for kDNA (primer pairs 121/122). The performance of these four tests was further evaluated at the coordinating laboratory on a subset of samples from seropositive and seronegative patients, analysed in four independent experiments (Table 5). Examples of the outputs of each method are shown in Figure 2. The degree of concordance among the reported results by the BPM was between 87.5% and 90.62%. This intralaboratory evaluation showed that the selected methods depicted similar operational parameters than when performed by the corresponding laboratories in the international study (Tables 4 and 6). 10.1371/journal.pntd.0000931.g002 Figure 2 Examples of the outputs of the four best performing PCR methods. A. LbD2 ; B.LbD3, C. LbF1 and D. LbQ. The methods are described in Materials and Methods and Table 1. 6, 15: seropositive samples; 35; seronegative sample (Table 3). PC: Positive control: 10 fg/µl of T.cruzi VI. NC. Negative Control: Master Mixes devoid of DNA. 10.1371/journal.pntd.0000931.t005 Table 5 Intra-laboratory evaluation of best performing methods in human samples. N° Positive PCR/N° tested samples ID LbD2 LbD3 LbF1 LbQ % pos Cons Seropositive samples 4 2/4 3/4 2/4 2/4 75 pos 6 4/4 4/4 4/4 4/4 100 pos 11 4/4 3/4 4/4 3/4 90,6 pos 15 2/4 3/4 3/4 3/4 59,4 pos 32 0/4 0/4 0/4 0/4 0 neg Seronegative samples 33 0/4 0/4 0/4 0/4 0 neg 35 0/4 0/4 0/4 0/4 0 neg 38 0/4 0/4 0/4 0/4 0 neg Concordance 28/32 29/32 29/32 28/32 ID, sample identification number; LbX/1-6, Laboratory and test identification; % pos, Percentage of Positivity; Cons, Consensus PCR Result; pos, positive; neg, negative. 10.1371/journal.pntd.0000931.t006 Table 6 Intra-Laboratory Evaluation of the four Best Performing Methods in samples from Table 5. PCR vs Consensus PCR of Table 5 PCR vs Serology Lb/Test Se Sp Acc kappa Se Sp Acc kappa N = 20 N = 12 N = 32 N = 32 N = 20 N = 12 N = 32 N = 32 LbD2 75 100 87,5 0.8 60 100 75 0.5 LbD3 81,25 100 90,6 0.8 65 100 78 0.6 LbF1 81,25 100 90,6 0.8 65 100 78 0.6 LbQ 75 100 87,5 0.8 60 100 75 0.5 LbX/1−6, Laboratory and test identification; Se, sensitivity; Sp, specificity; Acc, accuracy; kappa, kappa index. Discussion PCR technology has been widely used for the diagnosis and monitoring of disease progression and therapy outcome in many infectious diseases [28]. Since 1989, PCR strategies have been developed aiming to analyse clinical samples infected with T. cruzi [7], [36]. However, each laboratory has applied its own protocols and quality controls, making comparison of PCR based findings among different research groups and geographical regions not reliable. This international collaborative study is a crucial first step aiming at the evaluation of currently used PCR procedures for detection of T. cruzi infection, towards the assessment of a standard operative procedure. Out of the 48 PCR tests reported by 26 laboratories, those targeting ribosomal, miniexon or CO II subunit gene sequences were not sensitive enough when challenged against 10 fg/µl or less of purified DNA from the 3 tested parasite stocks, to merit further consideration. Thus, these methods appeared not suitable for sensitive molecular diagnosis of Chagas disease in clinical settings. However, these parasitic targets are been widely used for genotyping parasite discrete typing units [29]–[30], [37]–[41], mitochondrial gene haplotypes [42] or miniexon based T. cruzi I genotypes [43]–[44]. In this regard, multicentric evaluation and standardization of PCR based genotyping methods for identification of T. cruzi DTUs is needed. PCR Performance in Set A Samples In set A, GPM included kDNA and sat-DNA PCR tests in similar proportions. However, Sat-DNA PCR tests were less sensitive than kDNA-PCR tests to detect T. cruzi I DNA. This is most likely due to the fact that T. cruzi DTU I harbors approximately four to ten-fold less number of satellite repeats than DTUs II, V and VI, which has been demonstrated by different molecular approaches [22], [45]. Regarding T. cruzi IV that also harbors a lower dosage for satellite sequences [22], similar analytical sensitivities of kDNA and Sat-DNA PCR tests were observed, being lower than that obtained for the other two tested DTU representative stocks. The genome size and relative DNA contents of Can III cl1 (116.44 Mb, 95% CI 110.4–122.63 and 1.090, respectively) and CL Brener (108.55 Mb, 95% CI 101.41–115.89 and 1.017) are similar [25], although Can III cl1 harbors about 5 fold less satellite repeats than CL-Brener [22]. The relative contribution of the nucleus and kinetoplast has not been measured but normally, kDNA represents 20–25% of the total DNA content [46]. There are no available data regarding the number of minicircles in the kinetoplast of Can III cl1, so it could be speculated that the lower analytical sensitivity of most PCR tests to detect DNA from this clone respect to the other ones, could be due to a lower minicircle copy dosage. PCR Performance in Set B Samples Set B allowed evaluation of the influence of DNA extraction procedures in the PCR performance. A 72.2% of DNA extraction methods based on commercial kits led to GPM in set B, whereas 57.8% of phenol-chlorophorm extracted DNA led to GPM reports. These findings indicated that Guanidine Hidrochloride-EDTA blood was suitable for DNA extraction using kits based on lysis buffers containing Guanidine salts. Out of the 25 GPM in set B, 14 had a sensitivity of 0.05 par/ml, which should be adequate for diagnosis of infection in chronic patients [20]. Indeed, the necessary detection limit in chronic Chagas disease has been stated as one parasite cell in 10 mL of blood [20]. PCR Performance in Set C Samples Analysis of PCR performance in set C clinical samples showed that the four best performing tests presented strong concordance with respect to consensus PCR results obtained by the 16 tests defined as GPM in sets A plus B (kappa index between 0.7 and 0.8). Out of them, three tests targeted sat-DNA sequences and only one targeted kDNA. These data are in agreement with previous works showing that PCRs targeting Sat-DNA performed better than PCRs targeting kDNA sequences [47]–[49], although kDNA based PCR has been more widely used [20]. Moreover, two of the sat-DNA best performing tests used Real Time PCR, one with a Sybr Green fluorescent dye (LbD2) and the other one with a TaqMan probe (LbF1). It must be pointed out that LbD2 and LbD3 tests were performed by the same laboratory. Out of the 16 GPM performed by 11 different laboratories, 3 laboratories performed two methods (LbD, LbK and LbP) and one lab developed 3 tests (LbG). These data point to laboratory dependence concerning PCR performance, which may be due to multiple factors including technical expertise, correct use of quality controls, instrumentation and reagents. For example, tests LbF1, LbS2 (GPM) and LbZ (not GPM) were all based on sat-DNA Real Time PCR using the same primer pair (cruzi 1 – cruzi 2), differing in the trade marks of the DNA extraction and Master Mix kits. Some tests shown as GPM in sets A+B had very low sensitivities in set C (LbK2, LbP2, LbV1, Table 4), suggesting that quality controls might have failed to distinguish false negative clinical samples. A major drawback of most PCR tests is that they do not contain an internal amplification control (IAC). An IAC is a non target DNA sequence present in the same sample reaction tube, which is co-amplified simultaneously with the target sequence [50]. In a PCR without an IAC, a negative result can indicate that the reaction was inhibited, as a result of the presence of inhibitory substances in the sample matrix. The presence of PCR inhibition in Guanidine Hidrochloride-EDTA treated blood samples has been described [22]. The European Standardization Committee (CEN), in collaboration with International Standard Organization (ISO) has proposed a general guideline for PCR testing that requires the presence of IAC in the reaction mixture [51]. Therefore, only IAC-containing PCRs should undergo multicentre collaborative trials, which is a prerequisite for validation. Some other tests shown as GPM in sets A+B had very low specificities (LbI2, LbW, LbG2, LbG3, Table 4). Amplicon carry-over contamination is one of the most probable causes. PCR master mixes with dUTP and Uracil-DNA N-glycosylase (UNG) intended to abolish amplicon carry-over contamination were used in some tests (LbF, LbG, LbL, LbS, Table 1). Nevertheless, some of them did not show good specificity in set C (LbG2, LbG3, LbG4, LbL1, LbL2, Table 4), suggesting that problems during sample processing, such as sample to sample contamination could have arisen. The median values of the sensitivities obtained in testing the Set C samples with the 16 tests determined to be GPMs by testing the Set A and Set B samples varied considerably depending on the clinical characteristics of the persons from whom the Set C samples were drawn. Indeed, sensitivity was 100% (25-75p = 100−100) for immunosuppressed heart transplanted pts, 56.5% (25-75p = 39.1–66.3%) for asymptomatic and 57.1% (25-75p = 14–75%) for symptomatic chronic Chagas disease patients. These data point to the limitations of PCR strategies for diagnosis of patients at the chronic phase of disease. In addition, some of these samples had been stored at 4°C for at least two years before this PCR study; thus higher PCR positivity might be obtained in prospective clinical studies but it is unlikely that the current PCR methods will have a sensitivity comparable to serological assays for diagnosis of chronic Chagas disease. The four BPM methods were transferred to the Coordinating laboratory, where they were evaluated in a subset of clinical samples, each one tested in four independent assays, obtaining good concordance and confirming the performance reported by the participating laboratories in the previous international study (Tables 5 and 6). Further work is still needed to validate them through prospective studies in different settings. In this regard, this collaborative evaluation constitutes a starting point towards technical improvement and development of an international standard operating procedure (SOP) for T.cruzi PCR. In this context, the BPMs could be recommended for alternative diagnostic support, such as in the following settings: a) post-treatment follow-up of patients to look for failure of therapy to achieve parasitologic response [12], [14]–[16], [20], [22]; b) diagnosis of congenital Chagas disease in newborns in whom the presence of maternal anti-T. cruzi antibodies make serological studies useless [11], [15], [48]; c) early diagnosis of reactivation after organ transplantation of T.cruzi infected recipients under immunosuppressive therapy [18], [41], d) differential diagnosis of Chagas reactivation in patients with AIDS [39], and e) suspicion of oral transmission [52]. Moreover it can be useful for post-treatment follow-up of experimental animals to look for failure of therapy to achieve parasitologic cure [53]; in diagnosis in naturally infected triatomines or triatomines used for xenodiagnosis, since it has been shown that PCR tests are much more sensitive than microscopic examination of intestinal contents [37], [54]; and diagnosis of T. cruzi infection in mammalian reservoirs for which serologic tools have not been developed [38].
              Bookmark
              • Record: found
              • Abstract: found
              • Article: not found

              Accurate Real-Time PCR Strategy for Monitoring Bloodstream Parasitic Loads in Chagas Disease Patients

              Introduction Infection with the parasite Trypanosoma cruzi (T. cruzi) remains a major concern in 21 endemic countries of America, with an estimated prevalence of almost 8 million infected people [1]. The infection may be acquired mainly through the triatomid insect vector, blood transfusion or the trans-placental route. Furthermore, in areas under vector control, cases of congenital and transfusional transmission are relatively emerging [2]. This parasitic disease shows a variable clinical course, which ranges from asymptomatic cases, to severe chronic stages characterized by low parasitaemia and cardiac and/or gastrointestinal disorders [1]. Individuals from different endemic regions are infected with distinct parasite populations that may play a role in pathogenesis, clinical forms and severity of the disease [3]. Parasite populations are classified into two main phylogenetic lineages, T. cruzi I (TcI) and T. cruzi II (TcII) [4]; the later composed by five subdivisions designated as TcIIa to TcIIe [5]. Current chemotherapies based on the nitrofuran nifurtimox, and the nitroimidazole benznidazole, are unsatisfactory since these compounds are almost exclusively effective in recent infections and frequently have toxic side effects [6]. In this context, the development of novel drugs is necessary [6]. After etiological treatment (tmt), the criterion of cure relies on serological conversion to negative of the anti- T. cruzi antibody response [2], but in patients initiating therapy at the indeterminate phase, seroconversion usually occurs several years after treatment, requiring long-term follow-up [7]. Moreover, parasitological response to treatment is usually monitored by means of traditional methods such as Strout, hemoculture or xenodiagnoses, which lack sensitivity, and therefore are also inadequate for these purposes [2]. In this context, quantitative real-time PCR (Q-PCR) has the potential to become a novel parasitological tool for prompt evaluation of trypanocidal treatment. As a target for amplification, the nuclear satellite DNA, represented in 104 to 105 copies in the parasite genome is highly conserved [8]–[11] and therefore may provide accurate Q-PCR based measurements. Proper Q-PCR performance also requires high quality DNA extraction procedures from blood samples, which in most cases are collected in guanidine hydrochloride and EDTA buffer (GEB) [12]. The co-purification of trace PCR inhibitors may not impede amplification but may reduce its efficiency resulting in erroneous quantification of the parasitic load. Accordingly, we aimed to develop a satellite-DNA based Q-PCR strategy for accurate quantification of T. cruzi loads in peripheral blood samples along with an adequate DNA extraction protocol. The following features have been regarded: A commercial DNA purification protocol based on silica-membrane technology was adapted for GEB samples, providing DNA lysates without PCR interfering substances. A heterologous internal standard (IS) was incorporated to each GEB sample to follow-up the yield and quality of DNA extraction and PCR amplification. The relative copy number of satellite repeats per genome, according to the parasite lineage, was assessed for a more accurate quantification of the parasitic load, by means of a real-time PCR melting curve analysis (Lg-PCR). Finally, we applied this Q-PCR strategy to: (1) calculate the basal T .cruzi loads in blood specimens collected from Chagas disease pediatric patients, (2) follow-up their parasitological response to treatment with benznidazole, and (3) monitor T. cruzi recrudescence and parasitological response to treatment in chronic Chagas heart disease patients undergoing heart-transplantation and receiving immunosuppressive therapy. Materials and Methods Ethics statement This study was conducted according to the principles expressed in the Declaration of Helsinki. The study was approved by the Institutional Review Boards of the “Ricardo Gutierrez” Children's Hospital and of the Fundacion “Rene Favaloro”. All patients or responsible adults provided written informed consent for the collection of samples and subsequent analysis. Parasite stocks T. cruzi epimastigotes were grown in liver infusion tryptose (LIT) medium containing 10% calf serum at 27–28°C. The parasites were harvested and stored at −70°C. T. cruzi DNA was purified after Phenol-Chloroform extraction and ethanol precipitation. Reference T. cruzi stocks used as controls were: TcI (Silvio X10 cl1, SN3, HA, Pal V2-2, Pav 00, G); TcIIa (CanIII); TcIIb: (Tu18, JG, Gilmar, Y, Basileu, Mas cl1); TcIIc: (M5631, Cu-Tom-229, Cu-Yaya-211), TcIId (Mn cl2, Bug 2148 cl1, SO3 cl5, PAH 265, Tev 41), TcIIe (Cl Brener, Tul 77, Tep 6, Tep 7, MC). They were kindly provided by Patricio Diosque (Universidad Nacional de Salta, Salta, Argentina), Andrea M. Macedo (University Federal of Minas Gerais, Belo Horizonte, Brasil), and Michel Tibayrenc (UR62 “Genetics of Infectious Diseases”, IRD Centre, Montpellier, France). Some T. cruzi I strains were provided by Omar Triana Chavez (University of Antioquia, Medellín, Colombia). Argentinean T. cruzi IIc isolates were provided by Ricardo Gurtler (Universidad de Buenos Aires, Argentina). Chagas disease patients Peripheral blood samples collected from a cohort of 43 T. cruzi-seropositive pediatric patients (mean age: 7.13 years; 15 days–18 years old) admitted between 2005 and 2006 at the Parasitology Unit of the “Ricardo Gutierrez” Children's Hospital, a reference centre for diagnosis and treatment of Chagas disease pediatric patients (Government of Buenos Aires, Argentina). These patients were treated during 60 days with benznidazole (5 to 8 mg/kg/day) [13]. Parasitic loads were determined by Q-PCR at time of diagnosis (t1), at 7 (t2), 30 (t3) and 60 (t4) days of treatment. After t4, patients were followed-up by qualitative kDNA-PCR at 6, 12 and 18 months post-tmt. Blood samples from three chronic Chagas heart disease patients undergoing orthotopic heart transplantation (Tx) in 2003–2005 were provided by the Transplant Unit of the Instituto de Cardiología y Cirugía Cardiovascular, Fundación “René Favaloro”, Buenos Aires, Argentina. [14]. All these protocols were approved by the ethical committees of the corresponding institutions and written informed consents were required from each patient or a responsible adult. DNA extraction from peripheral human blood samples Ten mL or 2 mL blood samples collected from T. cruzi infected adults or infants, respectively, were immediately mixed with one volume of 2× lysis buffer containing 6 M guanidine hydrochloride (Sigma, St Louis, USA) and 200 mM EDTA, pH 8.0 (GE) [12]. QIAmp DNA Mini Kit (Qiagen, Valencia, CA) based extraction was carried out from 400 µl of GEB and eluted with 200 µl of water according to the manufacturer's instructions using the Blood and Body Fluid Spin Protocol, with slight modifications. Briefly, since blood samples were initially mixed with one volume of GE lysis buffer, treatment with proteinase K and “AL” lysis buffer (which contains guanidine hydrocloride) were omitted. The following steps were carried out following the manufacturer's instructions. Phenol-chloroform based DNA extraction was carried out from 100 µl of GEB aliquots and resuspended in 50 µl water as reported [15]. T. cruzi standard calibration curve Cl Brener epimastigotes were added to non-infected human blood to result in a concentration of 105 p/mL of reconstituted blood and immediately mixed with one volume of 2× lysis GE buffer. The resulting GEB was serially diluted 10-fold with non-infected GEB to cover a range between 105 and 0.001 parasite equivalents/mL. Total DNA was purified using the QiAmp DNA Mini Kit based extraction method, as above described. T. cruzi satellite DNA Q-PCR An MJR-Opticon II device (Promega, USA) was used for amplification and detection. The 20 µL reaction tube contained 0.5 µM of novel primers Sat Fw (5′-GCAGTCGGCKGATCGTTTTCG-3′) and Sat Rv (5′-TTCAGRGTTGTTTGGTGTCCAGTG-3′), 3 mM MgCl2, 250 µM of each dNTP, 0.5 U of Platinum Taq polymerase, (Invitrogen, Life Technologies, USA) SYBR Green (Invitrogen, Life Technologies, USA) at a final concentration of 0.5× and 2 µL of sample DNA. After 5 min of pre-incubation at 95°C, PCR amplification was carried out for 40 cycles (94°C for 10 s, 65°C for 10 s and 72°C for 10 s). The plate was read at 72°C at the end of each cycle. Internal standards of Q-PCR A linearized p-ZErO plasmid containing a sequence of Arabidopsis thaliana was used as a heterologous internal standard (IS). All clinical samples were co-extracted with 200 pg of recombinant plasmid, which was assumed as 1 arbitrary unit (AU) of IS. This amount of IS input was chosen because the amplicons are detected at approximately the Cycle threshold (Ct) number 20, the mid point of the dynamic range of the PCR. For each Q-PCR test, the IS was added to 400 µl of GEB lysate, immediately before the DNA extraction procedure. The standard calibration curve for the IS was carried out using the same reconstituted blood samples as for the T. cruzi calibration curve: those samples containing 105, 104 and 103 p/mL were spiked with 2 AU of IS, those containing 100, 10 and 1 p/mL were spiked with 0.2 AU of IS, and those containing 0.1, 0.01 and 0.001 p/mL were spiked with 0.02 AU of IS. Two non-infected blood samples with and without IS DNA were used as negative controls of the extraction procedure. The IS was quantified using 1 µM of primers, IS Fw (5′-AACCGTCATG GAACAGCACGTAC-3′) and IS Rv (5′-CTAGAACATTGGCTCCCGCAACA-3′). All other PCR reagents and cycling conditions were identical to those used for T. cruzi Q-PCR. Qualitative PCR detection of T. cruzi kDNA Post-treatment follow-up of parasitological response to benznidazole in pediatric and heart transplanted patients was conducted by means of kDNA-PCR as previously reported [14],[16]. Determination of the satellite/P2α ratio In order to analyze the variability in the number of satellite sequences detected by Q-PCR, comparative quantification was performed using as a normalizer the single copy ribosomal protein P2α gene (GenBank accession number XM_800089). Assays for quantification of P2α gene were performed using 1 µM of primers P2α Fw (5′-ATGTCCATGAAGTACCTCGCC-3′) and P2α Rv (5′-GCGAATTCTTACGCGCCCTCCGCCACG-3′). All other PCR reagents were used at the same concentrations as for T. cruzi Q-PCR. After 5 min of pre-incubation at 95°C, PCR amplification was carried out for 40 cycles (94°C for 10 s, 60°C for 15 s and 72°C for 10 s). The plate was read at 72°C at the end of each cycle. Identification of parasites according to the number of satellite sequences detected per genome by Lg-PCR Since TcI and TcIIa parasites have a lower number of satellite sequences than TcIIb/c/d/e parasites, we have developed a method to distinguish between both groups according to the melting temperatures (Tm) of their corresponding amplicons to enable more precise parasitic load assessments. The identification of the type of satellite sequence was performed using 0.5 µM of primers TcZ1 (5′-CGAGCTCTTGCCCACACGGGTGCT-3′) and Sat Rv (5′-TTCAGRGTTGTTTGGTGTCC AGTG-3′). All other PCR reagents were used at the same concentrations as for T. cruzi Q-PCR. The PCR conditions consisted of an initial denaturation at 95°C for 5 min, followed by 40 cycles of 94°C for 10 s, 65°C for 10 s and 72°C for 10 s with fluorescence acquisition at 81.5°C and a final step of 2 min at 72°C. Amplification was immediately followed by a melt program with an initial denaturation of 5 s at 95°C and then a stepwise temperature increase of 0.1°C /s from 72–90°C. Since satellite DNA is arranged in tandem repeats, if more than 0.05 parasites (equivalent to approximately 10 p/mL) are loaded in the reaction tube, a satellite sequence dimer is amplified giving place to a melting temperature peak typically above 86°C for both lineage groups. Satellite sequences were obtained by direct sequencing of satellite DNA amplicons obtained with TcZ1 and TcZ2 primers (GenBank Accession numbers EU728662-EU728667). Sequence alignment was conducted using MEGA version 4 [17]. Normalization of parasite loads according to the internal standard and parasite satellite sequence group The parasite load in the clinical sample was normalized respect to the standard curve according to (1) the efficiency of the DNA extraction procedure measured by the amplification of the IS, and (2) the parasite lineage group. The following equation was used: Np/mL  = (Np/well ×LF / AU) / V, where Np/mL is the number of parasites per millilitre of blood, Np/well is the number of parasites per well, LF is the lineage factor, AU are the arbitrary units of IS quantified and V is the volume of extracted DNA sample used per reaction. Results Analytical sensitivity and reproducibility of Q-PCR with purified DNA and reconstituted blood samples The analytical sensitivity of the Q-PCR was tested by using serial dilutions of purified T. cruzi DNAs from TcI (Silvio X10 cl1) and TcIIe (Cl Brener) reference stocks. The detection limits were 2 fg and 0.2 fg DNA per reaction tube with a dynamic range of 107 and 108 for Silvio X10 cl1 and Cl Brener stocks, respectively (data not shown). These detection limits correspond to 0.01 and 0.001 parasite genomic equivalents considering that one parasite cell harbors approximately 200 fg of DNA. Furthermore, we tested the operational parameters of Q-PCR in reconstituted - blood samples spiked with known quantities of Cl Brener and Silvio X10 cl1 cultured epimastigote cells. The dynamic range of Q-PCR performed with samples reconstituted with Silvio X-10 was 0.1–105 p/mL and with those spiked with Cl Brener was 0.01–105 p/mL (Figure 1). 10.1371/journal.pntd.0000419.g001 Figure 1 Dynamic range of the T. cruzi satellite DNA based Q-PCR. Results are expressed as the number of parasites per milliliter of blood and represent the average of 5 independent experiments. Slope = −3.35. Efficiency = 99%. Dynamic range: 0.01–105 p/mL. R square: 0.998. C(t): cycle threshold. The reproducibility of the Q-PCR assay at 100 p/mL and 1 p/mL was estimated by testing each reconstituted sample 15 times in the same PCR run. The coefficients of variation of the Ct values were 1.27% and 2.30%, respectively for Cl Brener and 1.60% and 5.42%, respectively for Silvio X10 cl1. The reproducibility of the DNA extraction was also characterised: aliquots from the same GEB-reconstituted samples containing 10 Cl Brener p/mL were processed in twenty independent DNA purification experiments. For each of the 20 DNA lysates, the p/mL were measured in triplicate PCR runs and a mean value was calculated. The coefficient of variation of the Ct values among the 20 mean values was 1.69%. Estimation of the relative numbers of satellite DNA copies per genome for different Trypanosoma cruzi lineages In order to evaluate if the differences in the detection limits of the Q-PCR obtained from Cl Brener or Silvio X10 cl1 samples were related to a different copy number of satellite repeats in their respective genomes, the amounts of satellite sequences, relative to the single copy gene encoding the ribosomal-P2α protein, were estimated for parasite stocks belonging to the 6 lineages (Table 1). The stocks belonging to TcIIb/d/e lineages showed similar amounts of satellite repeats (Table 1A), the M5631 stock (TcIIc) harbored 2-fold less repeats than the aforementioned stocks (Table 1A), whereas Can III stock (TcIIa) and six stocks belonging to T. cruzi I from Argentina, Colombia and Brazil, harbored a 10-fold lower number of satellite repeats (Table 1A and 1B). This is in agreement with the analytical sensitivities obtained with purified DNA and reconstituted samples of the reference stocks. 10.1371/journal.pntd.0000419.t001 Table 1 Variation in the relative numbers of satellite DNA copies per genome. A. Cultured T. cruzi stocks representing each of the phylogenetic lineages Lineage T.cruzi Reference Stocks Sat/P2α Tc I Silvio X-10 cl1 0.08 Tc IIa Can lll 0.08 Tc IIb Tu18 0.95 Tc IIc M5631 0.47 Tc IId Mn cl2 1.1 Tc IIe Cl Brener 1 B. T. cruzi I stocks from different endemic countries T. cruzi I Stocks Country of Origin Sat/P2a Silvio X-10 cl1 Brasil 0.08 G Brasil 0,13 HA Colombia 0,11 SN3 Colombia 0,13 Pav 00 Argentina 0,09 Pal V2-2 Argentina 0,12 Results are expressed as arbitrary units, obtained calculating the ratio of copy numbers of Cl Brener satellite/ ribosomal-P2α gene DNA targets. Distinction by Lg-PCR of T. cruzi groups according to the relative number of satellite repeats Two groups of lineages, T. cruzi I/IIa (group I) and T. cruzi II (IIb,c,d,e) (group II) were clearly distinguished due to the differential melting temperatures of their corresponding satellite sequence amplicons, above or below 85°C, respectively (Figure 2). 10.1371/journal.pntd.0000419.g002 Figure 2 Melting curve analysis of satellite amplicons from reference stocks. Group I satellite amplicons show melting temperatures above 85°C, whereas Group II render amplification products with melting temperatures below 85°C. The Lg-PCR was validated using a panel of 25 characterised stocks (Table 2): five stocks belonging to TcI, one to TcIIa, six to TcIIb, three to TcIIc, five to TcIId and the remaining five to TcIIe. 10.1371/journal.pntd.0000419.t002 Table 2 Melting temperatures of satellite fragments amplified from T. cruzi stocks belonging to the 6 phylogenetic lineages. Stock Tm (°C) Lineage Tm (°C)±SD Silvio X-10 cl1 85.7 Tc I 85.34±0.15 Pal v2-2 85.4 G 85.5 SN3 85.3 HA 85.4 Can lll 85.7 Tc lla 85.70±0.30 Tu18 84.6 Tc llb 84.60±0.18 JG 84.8 Gilmar 84.8 Y 84.6 Basileu 84.4 Mas cl1 84.4 M5631 83.9 Tc llc 84.03±0.45 Cu-TOM-229 84.5 Cu-Yaya-211 83.6 Mn cl2 84.4 Tc lld 84.44±0.30 Bug 2148 cl1 84.8 SO3 cl5 84.7 PAH265 84.1 Tev 41 84.2 Cl Brener 84.2 Tc lle 84.42±0.33 Tul77 84.7 Tep7 84.3 MC 84.0 Tep6 84.9 The 99% confidence interval of the melting temperatures depicted by the T.cruzi strains from group I (mean: 85.5°C; CI 99% 85.2–85.8) and II (mean: 84.4°C; CI 99% 84.2–84.7) did not overlap. This distinction allows a more accurate parasite load quantification by incorporating a correction factor (Lineage Factor LF), according to the number of target sequences, when the parasitic load is calculated. Comparison of PCR inhibition using two extraction methods in human samples We extracted 55 GEB samples from Chagas disease patients with the QIAmp DNA Mini Kit adapted protocol as well as with a Ph-Chl based method. Using the Ph-Chl based method, we detected traces of PCR inhibitors in 16 (29%) samples, whereas for the commercial kit we did not detect PCR inhibitors in any of the samples. The presence of PCR inhibitors was assessed by (1) the low yield of IS amplification, giving rise to less than 0.1 AU of the IS and (2) the improved yield of IS amplification when the DNA lysate was diluted 1/100 prior to the Q-PCR run. Out of the 16 samples that presented traces of PCR inhibitors when extracted with the Ph-Chl based method, 4 were negative for T. cruzi by both extraction methods, 3 rendered similar parasite loads by both methods, 8 rendered higher parasite loads when extracted with the commercial kit, and one rendered a positive result only when extracted with the commercial kit. For the 39 samples extracted by Ph:Chl that did not contain PCR inhibitors, similar results of T. cruzi quantification were obtained with both methods of extraction. On the basis of these data, the QIAmp DNA Mini Kit extraction protocol was selected for Q-PCR in clinical samples. Application of the Q-PCR assay to clinical specimens A panel of human blood samples was analyzed using the Q-PCR strategy, namely samples from seropositive pediatric patients and chronic Chagas heart disease adults presenting clinical reactivation after heart transplantation. Table 3 describes the necessary steps for calculating the parasitic loads, considering the lineage factor (LF), the AU of the IS and the input volume of DNA sample. For example, in case Tx1c, the Q-PCR alone quantified 0.39 p/mL, but the AU of the IS was 0.43, and the LF was 10 (T. cruzi I), giving a final result of 9.07 p/mL, which is 23 times higher than if the quantification had been made based only on the Q-PCR crude measurement. In case Pd6, a pediatric patient with very low parasitemia, the lineage could not be determined since the low concentration of amplicons gave rise to non-reliable melting temperature peaks and therefore the parasitic load could only be expressed within a range of 10-fold. In cases Tx1a and Tx2c, also with low levels of parasitemia, the lineage was presumed from previous data obtained from genotyped samples of the same patients, because it was demonstrated that the T. cruzi lineages were persistent during recrudescence leading to clinical reactivation in these patients [14]. Cases Pd2a and Tx2b presented high parasitic loads, 512 and 468 p/mL, respectively. In both of them, dimers of the satellite repeats were preferentially amplified when Lg-PCR was performed, giving melting temperatures above 86°C (Table 3). Thus, when the non-corrected Q-PCR results are above 10 p/mL, the DNA lysate should be diluted to approximately 1 p/mL before performing Lg-PCR. 10.1371/journal.pntd.0000419.t003 Table 3 Examples of the calculation of parasitic loads in pediatric and transplanted patients. Case Age/Gender Clinical Diagnosis Sample Q-PCR Np/well / V IS-PCR AU Lg-PCR Tm (°C) Lineage LF Parasitic Load Np/mL  = (Np/well ×LF / AU) / V Pd 1a 7 y / M Congenital ChD Pre-tmt 7.20 0.87 84.6 2 1 8.28 1b 7 days of tmt 0.1 0.97 84.6 2 1 0.1 1c 30 days of tmt 0.76 1.21 84.4 2 1 0.63 1d 60 days of tmt 0.95 0.82 84.6 2 1 1.16 Pd 2a 3 m / M Congenital ChD Pre-tmt 527 1.03 84.5‡ 2 1 512 2b 30 days of tmt 0.29 0.67 84.8 2 1 0.43 2c 60 days of tmt 1.01 1.34 84.4 2 1 0.75 Pd 3 1 y / F Congenital ChD Pre-tmt 14.69 1.27 84.4 2 1 11.57 Pd 4 2 y / M Congenital ChD Pre-tmt 0.34 0.49 84.6 2 1 0.69 Pd 5 3 y / F Congenital ChD Pre-tmt 0.54 1.18 84.6 2 1 0.46 Pd 6 10 y / F Indeterminate ChD Pre-tmt 0.04 0.74 ND ND# ND# 0.05–0.5 Tx 1a 46 y / M Chronic Cardiomyopathy with Heart Transplantation 5 days pre-Tx 0.03 1.26 ND 1* 10* 0.22 1b 29 days post-Tx 0.59 1.53 85.2 1 10 3.85 1c reactivation 78 days post-Tx 0.39 0.43 85.3 1 10 9.07 Tx 2a 61 y / F Chronic Cardiomyopathy with Heart Transplantation 7 days post-Tx 2.23 0.84 84.5 2 1 2.66 2b reactivation 92 days post-Tx 342 0.73 84.6‡ 2 1 468 2c 12 days post-tmt 0.13 1.84 ND 2* 1* 0.07 In the samples from transplanted patients (Tx) whose infecting lineages could not be determined, they were presumed from the data obtained from other samples of the same patient, assuming that the lineage is persistent during reactivation [13]. ND: Not determined. # In case Pd6 no presumption could be made about the parasite lineage. ‡ Samples diluted before Lg-PCR. The reproducibility of the whole Q-PCR assay was evaluated in 5 clinical samples from different Chagas disease patients covering a range of 3 logarithms of parasitic loads (0.06 to 70.29 p/mL, Table 4). For each peripheral blood sample (A to E, Table 4), the entire protocol was carried out in 5 independent replicates. The coefficients of variation of the Ct values and p/mL for the T.cruzi Q-PCRs ranged from 1.30 to 5.49% and from 25.19 to 137,95%, respectively, and those for the IS-PCRs ranged from 0.66 to 2.28% and from 9.56 to 29.51%, respectively (Table 4). The final parasitic load measurements were inversely correlated with their coefficients of variation, which ranged from 32.43 to 114.20% (Table 4). 10.1371/journal.pntd.0000419.t004 Table 4 Reproducibility of the Q-PCR and IS-PCR assays in clinical samples of Chagas disease patients. Patient Replicates Q-PCR Cv % IS-PCR Cv% Parasitic Load Ct p/mL Ct p/mL Ct AU Ct AU p/mL Mean p/mL Cv% A 1 neg 0.00 5.49 137.95 20.83 1.37 1.54 22.66 0.00 0.06 114.2 2 30.29 0.29 20.40 1.83 0.16 3 31.95 0.09 21.23 1.05 0.09 4 34.31 0.02 20.91 1.30 0.01 5 33.56 0.03 21.12 1.13 0.03 B 6 30.83 0.20 2.83 66.52 23.08 0.30 2.28 29.51 0.65 0.33 72.39 7 31,70 0.11 22.48 0.46 0.24 8 30.08 0.33 21.97 0.64 0.51 9 32.13 0.08 22.00 0.63 0.13 10 32.09 0.08 21.86 0.69 0.12 C 11 30.16 0.31 3.20 69.45 21.64 0.80 1.51 21.96 0.39 0.50 88.15 12 30.29 0.29 20.94 1.28 0.22 13 31.19 0.15 21.31 1.00 0.15 14 28.59 0.92 21.76 0.74 1.24 15 29.59 0.46 21.31 1.00 0.46 D 16 26.08 5.16 1.30 25.19 20.97 1.25 0.66 9.56 4.13 4.87 32.43 17 25.93 5.72 21.03 1.20 4.75 18 25.36 8.47 21.11 1.14 7.41 19 26.26 4.56 20.75 1.45 3.15 20 25.91 5.80 21.05 1.18 4,90 E 21 22.89 46.25 2.44 40.68 21.36 0.96 1.42 19.05 47.99 70.29 33.45 22 22,80 49.20 21.95 0.65 76.01 23 23,00 42.88 21.32 0.99 43.32 24 21.72 103.38 21.16 1.10 93.84 25 22.17 75.87 21.56 0.84 90.28 The T.cruzi lineage group was assessed by Lg-PCR as T.cruzi II in all tested patients. Parasitic Loads were calculated as shown in Table 3. Follow-up of pediatric patients under treatment with benznidazole The pre-treatment parasitic loads were assessed in 43 children with Chagas disease. Basal parasitic loads ranged from 640 to 0.01 p/mL, and were correlated to the patients' ages at the time of diagnosis (coefficient of Spearman: −0.5832, P<0.05) (Figure 3A). Thirty eight of these patients were monitored by Q-PCR during 60 days of etiological treatment with benznidazole. Parasitic loads were determined at time of diagnosis (t1) in 38 cases, at 7 (t2), 30 (t3) and 60 (t4) days of treatment in 31 cases. 10.1371/journal.pntd.0000419.g003 Figure 3 Parasitic loads in peripheral blood samples from pediatric patients. (A) Association between basal parasitic loads and patients' ages in 43 pediatric cases. Coefficient of correlation: −0.5832; P<0.05. (B) Monitoring of parasitological response to benznidazole therapy in 38 pediatric patients. The evolution of the parasitic loads for patients with more then one positive sample are depicted. Samples were withdrawn at time of diagnosis (t1), after 7 (t2) and 30 (t3) days of treatment, as well as at the end of treatment (t4, 60 days). Only the PCR positive samples are shown. The horizontal line represents the lower limit of the dynamic range of Q-PCR. Figure 3B and Table 3 show the parasitic response of these patients during treatment monitoring. Q-PCR results at t2 were negative in 24 out of 31 patients (77%), at t3 in 27 out of 31 patients (87%) and at t4 in 29 out of 31 patients (94%). One of the Q-PCR positive patients at t4 was a 7 year-old boy whose parasite load declined from 8.28 p/mL at t1 (Pd1a) to 0.1 p/mL at t2 (Pd1b), relapsing to 0.63 p/mL at t3 (Pd1c) and 1.16 p/mL at t4 (Pd1d) (Figure 3B, red triangle, Table 3). The other Q-PCR positive patient at t4 was a 3 month old infant who displayed detectable parasitic loads in the three analysed samples; 512 p/mL at t1 (Pd2a), 0.43 p/mL at t3 (Pd2b) and 0.75 p/mL at t4 (Pd2c) (Figure 3B, green triangle, Table 3). Both patients were followed-up by kDNA-PCR with persistently positive results at 6, 12 and 18 months post-tmt suggesting treatment failure (data not shown). Q-PCR based monitoring of Chagas disease reactivation in heart-transplanted patients Stored peripheral blood samples from three heart transplanted patients infected with different parasite lineages and presenting different patterns of clinical reactivation were retrospectively analyzed using the Q-PCR strategy (Figure 4). 10.1371/journal.pntd.0000419.g004 Figure 4 Follow-up of Chronic Chagas heart disease patients after heart transplantation. Parasitic loads in peripheral blood samples of Chronic Chagas heart disease patients with clinical reactivation due to immunosupression after heart transplantation. * Time of diagnosis of clinical reactivation and etiological treatment. Case Tx1 presented positive Q-PCR results (0.22 p/mL) prior to heart transplantation. Clinical reactivation was diagnosed at day 78 post-Tx by a skin chagoma biopsy and a positive Strout result. The parasite population was characterized as group I by Lg-PCR. Case Tx2 did not present positive PCR results before heart Tx, but parasitemia became detectable 7 days post-Tx (2.66 p/mL). Clinical reactivation was diagnosed at 92 days post-Tx due to positive Strout findings. Upon etiological tmt, the parasitic load decreased reaching undetectable levels at 21 days post-tmt. The T. cruzi population was identified as group II. Case Tx3 presented a parasite load of 0.02 p/mL 30 days before Tx, showing a 10-fold increase (0.2 p/mL) 7 days after Tx. The parasite load continued to rise until clinical reactivation was diagnosed 38 days after Tx based on positive Strout findings. Accordingly, the patient was treated with benznidazole and the parasite load dropped rapidly, with negative Q-PCR findings at 14 days post-tmt, persisting PCR negative for 53 days. However, 70 days after tmt Q-PCR monitoring revealed detectable loads (0.01 p/mL) (Figure 4). The bloodstream T. cruzi population was identified as group II. Discussion Herein we report a highly sensitive, reproducible, accurate and rapid real-time Q-PCR strategy for quantification of the T. cruzi parasitic load in human blood samples. Indeed, this is a first attempt to develop a real-time Q-PCR strategy for reliable T. cruzi quantification, because it incorporates: (1) a commercial kit for sample processing which minimizes carry-over of PCR inhibitors and standardizes the yield and quality of DNA extraction, (2) a closed-tube single-round PCR reaction which minimizes carry-over contamination, (3) an appropriate internal quality control and (4) a correction of the parasitic load, according to the variation in the number of target sequences between the different lineage groups. This is a key step towards the standardization and validation of in-house Q-PCR tests for application to routine laboratory practice. Although studies showing the advantages of real-time PCR for screening and quantification of T. cruzi have recently appeared, none have implemented procedures to normalize the DNA extraction yield and the representativity of the PCR target according to the lineage group. Some studies [18]–[20] have chosen as an internal control a host DNA sequence, but the human DNA content in blood can be highly variable, especially in immunosupressed patients. In this work, the addition of a standardized amount of a plasmid containing a heterologous sequence, allows normalization of the DNA extraction yields and detection of false negatives due to inhibition under any clinical situation. Regarding the selected molecular target of amplification, Elias et al. [8],[11] and Vargas et al. [10] demonstrated that satellite DNA is 4 to 9 times more abundant in TcIIb/d/e than in TcI stocks. Herein, we have extended this analysis to all 6 T. cruzi lineages, describing for the first time the satellite repeats of T. cruzi IIa, IIc and IId representative stocks (GenBank Accession numbers EU728662-EU728667). Indeed, we detected a 5 to 10-fold variation in the satellite DNA content between group I (TcI/IIa) and group II (TcIIb/c/d/e) parasite stocks (Table 1). Therefore, this variability must be taken into account in order to calculate accurately the parasitic loads. Accordingly, we have also designed a highly sensitive real-time PCR procedure (Lg-PCR) to distinguish T. cruzi group I (with lower satellite sequence copy number and higher melting temperature) from group II lineages (with higher satellite sequences copy number and lower melting temperature). All the analyzed stocks rendered only one temperature melting peak, including hybrid stocks like Cl Brener although harboring both types of satellite sequences. This can be explained by the fact that Cl Brener (Tc IIe) harbors ten times more type II than type I satellite repeats [10] and due to exponential amplification, only the predominant sequence type is detected. The high analytical sensitivity of Lg-PCR makes it useful for direct lineage group characterization in biological samples that may not be typed using other typing methods [16],[21],[22]. Alternatively, a recently reported multiplex PCR strategy might be useful for typing T.cruzi groups in clinical specimens [23] when a Q-PCR test targeted to satellite sequences is carried out. Sample processing and DNA extraction must also be optimized for reliable quantification. In this direction, we adapted a commercial kit, based on silica-membrane technology (QIAmp DNA Mini Kit) for processing GEB samples. Phenol-chloroform DNA extraction is a cost-effective method for qualitative PCR, but traces of PCR inhibitors may be co-purified [24],[25]. These interfering substances may not impede Q-PCR amplification, but could affect its efficiency leading to inaccurate results. In fact, we detected traces of PCR inhibitors in 29% of the samples extracted with Ph-Chl. When present, these inhibitors underestimated the parasite loads in 67% of the positive samples, although they did not seriously affect the positivity of the PCR. These results suggest that Ph-Chl based extraction of GEB samples is not suitable for Q-PCR but can be used for qualitative purposes. In this report, we applied the Q-PCR strategy to blood samples collected from patients under different clinical scenarios. Its wide dynamic range allowed direct measurements in cases with high parasitic loads such as immunosuppressed Chagas disease patients and congenitally infected newborns, as well as in cases with low parasitemias, such as patients at the indeterminate phase or under etiological treatment. In this sense, the coefficients of variation of the Q-PCR measurements obtained from clinical samples were similar to those obtained from reconstituted blood samples (Table 4). When applied to newborns, infants and children with T. cruzi infection, Q-PCR estimated their basal parasitic loads in a vast range, between 0.01 and 640 p/mL of blood (Figure 3A). The highest parasitic loads observed in the younger pediatric population are in agreement with the results obtained by conventional parasitological and kDNA-PCR analysis [15],[26]. Moreover, the lower parasitic loads detected in the older pediatric patients reflect their evolution to the indeterminate phase of congenital infection [15]. Furthermore, we were able to follow-up their parasitological responses to treatment with benznidazole (Figure 3B) with a favorable outcome in 94.7% (36/38) cases. It is worth to note that at t3, under 30 days of tmt, 4 patients still showed detectable parasitic loads. At the end of tmt (t4) 2 of them became Q-PCR negative and remained negative during 18 months of post-tmt follow-up. This observation emphasizes the importance of a treatment regimen of 60 days. Interviews with the mothers of the two patients who persisted Q-PCR positive at t4 revealed the non-adherence in one of them (Pd2, Table 3 and Figure 3B, green triangle), whereas in the other case (Pd1, Table 3 and Figure 3B, red triangle) persistence of parasitemia indicated lack of parasitological response to benznidazole. These cases demonstrate the usefulness of the Q-PCR assay as surrogate marker for early detection of treatment failure. Two pediatric patients (Figure 3B, pink and yellow squares) presented an increase of their parasitic loads from t1 to t2, showing a favorable parasitological response to treatment in the samples collected at t3 and t4, fact that was confirmed by means of kDNA-PCR in the successive post-tmt controls (data not shown). The lower parasitic loads detected at t1, before initiation of tmt, compared with t2, might be due to natural fluctuations of the parasitemia in chronic Chagas disease patients [27]. In this context, it is important to analyze serial blood samples to be able to observe an increasing or decreasing tendency in the parasitic loads in chronic patients under treatment. The Q-PCR test was also used for early detection of T. cruzi reactivation after heart transplantation. This was visualized through the increment of the parasitic loads in all patients who presented clinical manifestations of reactivation [14]. In case Tx1, the patient is infected with T. cruzi I, and the number of parasites increased from 0.22 p/mL (5 days pre-Tx) to 9.07 p/mL (78 days post-Tx) when the patient presented signs and symptoms of skin reactivation and patent parasitemia [14]. In the other two tested cases, Tx2 and Tx3, who were infected with group II populations, the parasitic loads increments were notably higher (Table 3 and Figure 4). In Tx3, treatment with benznidazole after reactivation achieved transitory parasitological response because samples collected at 44 and 70 days after tmt were Q-PCR positive (Figure 4). Parasite relapse was confirmed by means of kDNA-PCR in successive samples until a second episode of clinical reactivation was diagnosed [14]. In Tx cases, the Q-PCR allowed to detect parasitic load increase, previous to diagnosis of reactivation, as well as to follow-up parasitological response during treatment with benznidazole. All together, the high analytical sensitivity of the Q-PCR strategy, the low levels of intra- and inter-assay variation, as well as the accuracy provided by the Lg-PCR correction, promotes this method as a key laboratory tool to follow-up patients under etiological treatment or at risk of clinical reactivation. This will be of particular significance for future drug trials in which an early assessment of efficacy or failure is mandatory [28].
                Bookmark

                Author and article information

                Contributors
                Role: Editor
                Journal
                PLoS One
                PLoS ONE
                plos
                plosone
                PLoS ONE
                Public Library of Science (San Francisco, USA )
                1932-6203
                2012
                22 August 2012
                : 7
                : 8
                Affiliations
                Laboratory of Emerging Pathogens, Division of Emerging and Transfusion Transmitted Diseases, Center for Biologics Evaluation and Research, U. S. Food and Drug Administration, Bethesda, Maryland, United States of America
                Federal University of São Paulo, Brazil
                Author notes

                Competing Interests: The authors have declared that no competing interests exist.

                Conceived and designed the experiments: RN AD. Performed the experiments: RN VB SS FFA. Analyzed the data: RN HLN AD. Contributed reagents/materials/analysis tools: RN AD. Wrote the paper: RN HLN AD.

                Article
                PONE-D-12-01182
                10.1371/journal.pone.0043533
                3425475
                22927983

                This is an open-access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.

                Page count
                Pages: 12
                Funding
                This work was supported by FDA internal funding under the FDA Critical Path Initiative. No external funding sources were used for this study. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
                Categories
                Research Article
                Biology
                Biochemistry
                Nucleic Acids
                RNA
                Medicine
                Diagnostic Medicine
                Clinical Laboratory Sciences
                Transfusion Medicine
                Test Evaluation
                Hematology
                Transfusion Medicine
                Infectious Diseases
                Neglected Tropical Diseases
                Chagas Disease
                Parasitic Diseases
                Chagas Disease

                Uncategorized

                Comments

                Comment on this article