To determine the ability of recombinant antigens to detect cases of infection with Trypanosoma cruzi among cases of infection with Leishmania spp. by serological methods.
To determine the ability of recombinant antigens to detect cases of infection with Trypanosoma cruzi among cases of infection with Leishmania spp. by serological methods.
Sera from 41 patients infected with Leishmania spp. were evaluated with ELISA using single (FRA, CP1 and TSSAVI) or pooled (commercial Rec-ELISA) recombinant proteins or homogenate antigens (commercial H-ELISA). As there is no gold standard antigen to discriminate Chagas disease from leishmaniasis, the correlation of results between defined antigens and the homogenate was made with Kappa Index (KI), the level of correlation considered being used as a criterion of specificity.
Single recombinant antigens and Rec-ELISA showed good correlation (KI > 0.8). A low correlation (KI < 0.66) was observed between the results from single recombinant antigens or the commercial recombinant kit and H-ELISA.
The highly correlated results between T. cruzi single or pooled recombinant proteins are indicative of the usefulness of recombinant antigens for Chagas diagnosis. Our results also indicate that in the city of Oran in Argentina, between 12% and 17% of patients with leishmaniasis are also infected with Chagas disease. The high KI values between TSSAVI and the other recombinant proteins suggest that in these patients, the infection may be caused by T. cruzi II and/or V and/or VI lineages.
Déterminer la capacité des antigènes recombinants pour la détection de cas d'infection par T rypanosoma cruzi dans les cas d'infection par L eishmania spp. au moyen de méthodes sérologiques.
Les sérums de 41 patients infectés par L eishmania spp. ont été évalués par la méthode ELISA en utilisant des protéines recombinantes uniques (FRA, CP1 et TSSAVI) ou poolées (Rec-ELISA commercial), ou des homogénats d'antigènes (H-ELISA commercial). Comme il n'y a pas d'antigène de référence pour distinguer la maladie de Chagas de la leishmaniose, la corrélation entre les résultats des antigènes définis et ceux des homogénats a été faite grâce à l'indice Kappa (KI), le niveau de corrélation considéré étant utilisé comme un critère de spécificité.
Les antigènes recombinants uniques et le Rec-ELISA ont montré une bonne corrélation (KI > 0,8). Une faible corrélation (KI < 0,66) a été observée entre les résultats des antigènes recombinants uniques ou ceux du kit de recombinants commercial et le H-ELISA.
Les résultats fortement corrélés entre les protéines recombinantes uniques ou poolées de T . cruzi sont révélateurs de l'utilité des antigènes recombinants pour le diagnostic de la maladie de Chagas. Nos résultats indiquent également que dans la ville d'Oran en Argentine, entre 12% et 17% des patients atteints de leishmaniose sont également infectées par la maladie de Chagas. Les valeurs élevées de KI entre TSSAVI et les autres protéines recombinantes suggèrent que chez ces patients l'infection, peut être provoquée par les lignées de T . cruzi II, V et/ou VI.
Determinar la habilidad de los antígenos recombinantes para detectar, mediante métodos serológicos, casos de infección por T rypanosoma cruzi entre casos con infección por L eishmania spp.
Se evaluó el suero de 41 pacientes infectados con L eishmania spp. mediante ELISA utilizando proteínas recombinantes únicas (FRA, CP1 and TSSAVI) o en grupo (Rec-ELISA comercial) o antígenos a partir de homogenizado de parásitos (H-ELISA comercial). Como no existe una prueba de referencia para discriminar entre la enfermedad de Chagas y la leishmaniosis, la correlación de los resultados entre los antígenos definidos y los procedentes del homogenizado se realizó mediante el Índice de Kappa (IK), utilizando el nivel de correlación considerado como el criterio de especificidad.
Los antígenos recombinantes únicos y la Rec-ELISA mostraron una buena correlación (IK > 0.8). Se observó una baja correlación (IK < 0.66) entre los resultados de los antígenos recombinantes únicos o el kit comercial recombinante y el H-ELISA.
Los resultados con una alta correlación entre las proteínas recombinantes de T . cruzi, únicas o en grupo, demuestran la utilidad de los antígenos recombinantes para el diagnóstico de la enfermedad de Chagas. Nuestros resultados también demuestran que en la ciudad de Orán en Argentina, entre un 12% y 17% de los pacientes con leishmaniosis presentan también la enfermedad de Chagas. Los altos valores de IK entre TSSAVI y las otras proteínas recombinantes sugiere que en estos pacientes la infección podría estar causada por T . cruzi de linaje II, V y/o VI.
The endemic areas of Trypanosoma cruzi and Leishmania spp. infections in Latin America overlap in many regions. This is especially true for Argentina and specifically for the province of Salta (Gil et al. 2011). Because both species exhibit close phylogenetic relationships, it is difficult to discriminate between Chagas disease and American cutaneous leishmaniasis (ACL) in overlapping areas due to antigenic cross-reactivity (Chiaramonte et al. 1996; El-Sayed et al. 2005).
A pitfall of conventional serological methods to diagnose Chagas disease is the potential cross-reactivity with antibodies from patients infected with Leishmania spp. In an attempt to overcome this limitation, efforts were made to develop ELISA with parasite-specific antigens for Chagas disease. The importance of this non-conventional serological test lies in its feasibility, flexibility and low cost (WHO 2007; Brasil et al. 2010).
Some T. cruzi recombinant antigens are as follows: shed acute-phase antigen (SAPA) (Affranchino et al. 1989), flagellar repetitive antigen (FRA) (Krieger et al. 1992; dos Santos et al. 1992) and trypomastigote small surface antigen (TSSA) (Di Noia et al. 2002; Cimino et al. 2011; De Marchi et al. 2011; Risso et al. 2011). Another alternative is the construction of chimeric proteins, such as CP1 (Camussone et al. 2009). Nowadays, recombinant antigens have been incorporated in commercial kits of important companies, such as ABBOT (PRISM Chagas assay), Omega Diagnostics (Pathozyme Chagas) and Wiener Lab (Chagatest Rec v3.0) (Oelemann et al. 1998; Caballero et al. 2007; Otani et al. 2009). In particular, Chagatest Rec v3.0 is a kit widely used in Argentina and is prepared with a mixture of recombinant antigens (antigens 1, 2, 13, 30, 36 and SAPA) (Sanchez Negrette et al. 2008).
In the province of Salta, Oran and San Martin departments are endemic for ACL and have the highest prevalence values. Oran is the city with the greatest incidence of ACL in Argentina. According to the 1995–2007 database, the prevalence was 6.73% in children <15 years of age (Gil et al. 2010). Serological studies performed with T. cruzi antigens not related to Leishmania spp. determined that more than 35% of patients with symptomatic cutaneous leishmaniasis were infected with T. cruzi (Frank et al. 2003; Gil et al. 2011). Therefore, it is important to differentiate these infections because treatment and clinical follow-up are largely different and sero-epidemiological studies can underestimate or overestimate the prevalence of these parasitic diseases (Gil et al. 2011).
Previously, antigens used to discriminate between Chagas and leishmaniasis were assessed by considering other immunochemical techniques or epidemiological criteria. However, there is still no gold standard to differentiate between infections caused by T. cruzi and those caused by Leishmania spp.; hence, the results have to be confirmed. Taking into account the difficulty to define infection status in the study population, in the present work, we tried a new approach to evaluate the reliability of T. cruzi recombinant antigens to determine Chagas disease in a group of patients with leishmaniasis from a co-endemic area. We analysed the correlation among ELISA assays performed with FRA, CP1, TSSAVI, a commercial kit that contains a pool of six recombinant proteins (Chagatest Rec v3.0) and a commercial kit prepared with homogenate from the parasite [Chagatest ELISA (H-ELISA)]. Using this criterion, we confirmed that recombinant antigen has higher specificity than T. cruzi parasite homogenate. These results also allowed us to estimate the proportion of Chagas disease in patients with leishmaniasis and the specific T. cruzi strains responsible for these infections.
In 2009, 64 ACL cases were recorded at the Instituto de Investigaciones de Enfermedades Tropicales (IIET), a regional reference centre for the diagnosis of leishmaniasis located in Oran (Figure 1). Of those cases, 41 (64%) were analysed in this work. All patients with ACL lived in Oran and San Martin departments and presented classic cutaneous lesions confirmed by the observation of amastigotes in Giemsa-stained smears. In previous studies, among patients with ACL, Leishmania (Viannia) braziliensis was the most prevalent species of the three Leishmania species known to occur in northern Argentina (Marco et al. 2012; Krolewiecki et al. 2013).
We worked with the recombinant purified antigens CP1, TSSAVI and FRA, produced according to the specifications described elsewhere (Camussone et al. 2009; Cimino et al. 2011; Valiente-Gabioud et al. 2011). These antigens show high specificity and reactivity in human sera.
All sera were analysed by conventional serology [Chagatest ELISA (H-ELISA) and Chagatest ELISA recombinante v. 3.0 (Rec-ELISA), Wiener Lab, Rosario, Argentina], following the manufacturer's guidelines.
For the non-conventional serology, assays were standardised by evaluating the following factors: (i) carbonate buffers (pH = 9.6) and PBS (phosphate-buffered saline, pH = 7.2); (ii) optimal antigen concentration (0.25 and 0.5 μg per well/100 μl); (iii) different serum dilutions (1:20,1:40, 1:50 and 1:100) and (iv) conjugate dilutions (1:5000 and 1:10000). The optimal conditions were those that better discriminated between negative and positive control sera. The washing buffer used was 0.1% Tween in PBS. The reaction was developed using tetramethylbenzidine (TMB) (Invitrogen) in H₂O₂. The reaction was stopped with 0.5N H₂SO₄, and the absorbance was measured with an ELISA reader (Biotek ELx800) with 450-nm filter. All samples were measured in duplicate.
For the FRA-ELISA, polystyrene microplates were coated with 0.25 μg per well/100 μl in carbonate buffer. The serum dilution was 1:100, and peroxidase-conjugated anti-human IgG (Sigma, St. Louis) was 1:10000.
For the CP1-ELISA, polystyrene microplates were coated with 0.5 μg per well/100 μl in carbonate buffer. The serum dilution was 1:40, and peroxidase-conjugated anti-human IgG (Sigma, St. Louis) was 1:10000.
TSSAVI-ELISA, protocol described by Cimino et al. (2011).
We standardised ELISA technique with 18 negative control sera (non-chagasic) from Salta city, a non-endemic area for leishmaniasis. Fourteen serum samples were used as positive controls. These samples were confirmed by conventional serology (HAI and Chagatest Rec-ELISA), both from Wiener Lab, Rosario, Argentina.
The analysis was performed with Graph Pad Prism v.5.0 (San Diego, CA). Relative optical density (ROD) was calculated and defined as OD/cut-off, where optical density (OD) is the mean value of each serum sample and cut-off was determined for each assay as the mean OD of all negative control serum samples plus 3 standard deviations. Indeterminate samples are considered those with absorbance within the indetermination zone (cut-off ± 10%). Kappa Index (KI) was calculated using EPIDAT v3.1 software.
We analysed 41 sera from patients with ACL, 9.7% women and 90.3% men, aged between 10 and 70 years. 11 (27%) of these patients were from Hipolito Irigoyen, 16 (39%) from San Ramon de la Nueva Oran, 4 (10%) from Pichanal and 10 (24%) from San Martin department (Figure 1). Of all the samples analysed, 75.6% (31/41) had a ROD lower than the cut-off. A total of 11 samples were positive for at least one assay (Figure 2). Of the sera reactive to Rec-ELISA, 71.4% (5/7) were also positive for the other assays (Table 1).
|No. of sera||Rec-ELISA||H-ELISA||FRA-ELISA||CP1-ELISA||TSSAVI-ELISA|
The highest KI, which indicates the best agreement found in this procedure, was Rec-ELISA vs. FRA-ELISA and FRA-ELISA vs. CP1-ELISA, with a KI = 0.91. The agreement between the recombinant antigen assays was high, with values of KI > 0.8. TSSAVI-ELISA showed an important agreement with FRA-ELISA (KI = 0.89). The greatest disagreement was observed between H-ELISA and the remaining assays, with a KI < 0.7 (Table 2).
Although multicentric studies have determined that the sensitivity of parasite antigen homogenate is highly useful to diagnose T. cruzi infection with ELISA assays, a lack of specificity is observed in patients with leishmaniasis (Caballero et al. 2007; Remesar et al. 2009). When recombinant proteins of T. cruzi are assessed to determine their specificity, a panel of negative sera from patients with Leishmania spp. infection is usually obtained from a Chagas-free area to ensure, using epidemiological criteria, that study participants are not infected with T. cruzi (Umezawa et al. 1999, 2004; Aguirre et al. 2006). This criterion was used because there is no gold standard to discriminate T. cruzi infection from Leishmania spp. infection or from a mixed infection. This methodological restriction leads to biased results because variables (prevalence of different parasites, concomitant infections, nutritional habits) that determine cross-reactivity in individuals from an area are generally different from those present in other areas. To overcome this weakness of traditional assessment, we calculated the concordant performance of different antigens using a population of patients with leishmaniasis from a co-endemic area.
The recombinant antigens that were assessed correspond to antigens that were previously described to be experimentally specific for determination of Chagas disease (Umezawa et al. 1999; Camussone et al. 2009). This specificity can be explained when the identity of the sequence is analysed, that is, the best identity of FRA sequence within Leishmania genome (taxid: 38568) is of 62% but the identity with sequences from the genome from the phylogenetic distant T. cruzi CL Brener strain (discrete typing unit, DTU TcVI) and T. cruzi Silvio strain (DTU TcI) reaches 100 and 91%, respectively. It is therefore expected that serum reactivity against this antigen be mainly due to T. cruzi infection. Recently, a differential evolution of repetitions has been described for this molecule in T. cruzi and Leishmania species, this branching being coincident with the absence of B epitope cross-reactivity between them (Galetovic et al. 2011). However, cross-reactivity against a single antigen cannot be ruled out due to the presence of antibodies developed against other microorganisms. By contrast, the occurrence of this cross-reactivity against more than one protein is very unlikely. Therefore, the concordance of reactivity against specific recombinant proteins seems to be a reliable criterion to confirm a specific infection when a gold standard is not available.
The different assays using the recombinant antigens presented a good agreement compared with Rec-ELISA commercial assay, which uses a mixture of six recombinant antigens. The recombinant antigen FRA has 97% identity with Ag1 from the kit Wiener Lab (Camussone et al. 2009). The greatest disagreement was observed in the comparison of the results obtained with the recombinant antigens to results of H-ELISA commercial assay (protein mixture from epimastigotes). The use of a mixture of unknown proteins from the parasite makes antigenic cross-reaction possible because, as T. cruzi and Leishmania spp. are phylogenetically closely related, they have highly conserved proteins.
FRA recombinant antigen is a cytoskeleton-associated protein with a high performance used for the diagnosis of chronic infections (Krieger et al. 1992; da Silveira et al. 2001). The high reactivity against FRA shows that these patients with ACL might be in the chronic phase of Chagas disease. Moreover, the high agreement (KI ≥ 0.9) of FRA-ELISA with Rec-ELISA, TSSAVI-ELISA and CP1-ELISA is noticeable.
The results obtained with CP1 confirm previous results (Camussone et al. 2009) with respect to antigen specificity in sera of patients with ACL. In this work, when 15 sera of patients with ACL were analysed, CP1 showed a good performance, 93% of sensitivity and a higher reactivity than that obtained by peptide mixture. Our evaluation of sera from an endemic region of northern Argentina (Salta province) provides a novel contribution. The use of this antigen for the diagnosis of T. cruzi infection allowed us to confirm cases of Leishmania spp. – T. cruzi mixed infections, reported by previous studies carried out by our group (Gil et al. 2011). With this protein, however, an odd result was obtained with the HS8 sample, which was positive for this antigen and negative for the others. The difference between CP1 and FRA is the presence of SAPA epitopes, and this sequence might be responsible for the reactivity. However, the negative result of this serum with Chagatest, which has this protein within a mixture of recombinants, is not consistent with CP1 reactivity. This result might be attributed to the structural configuration of the chimeric antigen, which might establish an unspecific conformational epitope.
We used most of the sensitive and specific recombinant proteins that have been assessed to diagnose Chagas disease up to now: single proteins FRA, CP1, TSSAVI and a mixture of antigens 1, 2, 13, 30, 36 and SAPA in Chagatest Rec. Indeed, most of these proteins have been assessed in multicenter studies and their reactivity was found to be slightly different when using sera from regions where predominant DTUs are TcV and TcVI, as in Brazil or Argentina, or TcI, as in Colombia or Ecuador (Marcipar & Lagier 2012). This difference can be attributed to genetic variability in the parasites for these sequences. However, this lower reactivity was not translated into a lower positive detection index of the antigens in different regions. Indeed, these results are consistent with slight variations of amino acids in the sequences of these antigens within T. cruzi species. In the case of FRA, all of the 67 aa of the repetitive sequence blasted with FRA from CL Brener strain, whereas 60 aa blasted with FRA from Sylvio strain. In the case of SAPA antigen, corresponding to the C-terminal fraction from CP1 (the N-terminal fraction is FRA), the identity of the sequence with the one present in CL Brener and Silvio strains corresponds to 12/12 aa and 11/12 aa, respectively.
Contrary to the proteins described above for diagnostic purposes, TSSAVI have been proposed to differentiate patients infected with T cruzi lineages TcII, TcV and TcVI, but not with lineages TcI, TcIII and TcIV, based on the lineage specificity of this peptide (Bhattacharyya et al. 2010). Our result with TSSAVI reactivity (5 of 11 sera positive for at least one of the assays) suggests infections mainly with TcII and/or TcV and/or TcVI. This is the first description of the circulating T. cruzi lineages in the north of Salta province using serological methods. The present results are consistent with previous reports from Argentina, where the more prevalent lineages in humans are TcII, TcV and TcVI, whereas TcI and TcIII are less frequent (Cura et al. 2012). In fact, patients HS1, HS2, HS3, HS4 and HS7 would be infected with parasites belonging to TcII and/or TcV and/or TcVI, and patients HS5 and HS6 are probably infected with TcI and/or TcIII and/or TcIV.
In summary, this work highlights the importance of non-conventional serology for the diagnosis of Chagas disease. Further studies will be necessary to determine specificity of each recombinant antigen and whether these proteins can be used as a confirmatory test for T. cruzi infection in a larger number of cases of diverse geographic origin.
Non-conventional ELISA is a low-cost, simple-to-use technique. T. cruzi recombinant antigens have good correlations, as indicated by Kappa Index. This highlights the importance of non-conventional serology for the diagnosis of Chagas disease. TSSAVI reactivity proposes the presence of T.cruzi TcII, TcV and/or TcVI lineages in the north of Salta province, where no description of the circulating T. cruzi strains has been provided using serological methods.
This work was supported by Consejo de Investigación de la Universidad Nacional de Salta (CIUNSA), Number 2001 and 2056. We are grateful to Wiener Laboratories, Rosario, Argentina, for providing the commercial Chagatest kits.