Corresponding Author Antonio R.L. Teixeira, Chagas Disease Multidisciplinary Research Laboratory, Faculty of Medicine, University of Brasilia, 70.919-970 Federal District, Brazil. Tel.: +5561 349 4987; Fax +5561 273 4645; E-mail: email@example.com
Lack of conservation of the Amazon tropical rainforest has imposed severe threats to its human population living in newly settled villages, resulting in outbreaks of some infectious diseases. We conducted a seroepidemiological survey of 1100 inhabitants of 15 villages of Paço do Lumiar County, Brazil. Thirty-five (3%) individuals had been exposed to Trypanosoma cruzi (Tc), 41 (4%) to Leishmania braziliensis (Lb) and 50 (4.5%) to Leishmania chagasi (Lc) infections. Also, 35 cases had antibodies that were cross-reactive against the heterologous kinetoplastid antigens. Amongst these, the Western blot assays revealed that 11 (1%) had Tc and Lb, that seven (0.6%) had Lc and Tc, and that 17 (1.6%) had Lb and Lc infections. All of these cases of exposures to mixed infections with Leishmania sp, and eight of 11 cases of Tc and Lb were confirmed by specific PCR assays and Southern hybridizations. Two cases had triple infections. We consider these asymptomatic cases showing phenotype and genotype markers consistent with mixed infections by two or more kinetoplastid flagellates a high risk factor for association with Psychodidae and Triatominae vectors blood feeding and transmitting these protozoa infections. This is the first publication showing human exposure to mixed asymptomatic kinetoplastid infections in the Amazon.
L'absence de conservation de la forêt tropicale de l'Amazonie a imposé des menaces graves à sa population humaine vivant dans les villages nouvellement arrangés, avec pour résultat des épidémies de certaines maladies infectieuses. Nous avons mené une surveillance séroépidémiologique sur 1100 habitants de 15 villages dans le comté de Paço de Lumiar au Brésil. 35 individus (3%) avaient été exposés à l'infection àTrypanosoma cruzi (Tc), 41 (4%) àLeishmania braziliensis (Lb) et 50 (4,5%) àLeishmania chagasi (Lc). En outre, 35 cas possédaient des anticorps à réaction croisée avec les antigènes hétérologues de kinétoplastides. Parmi ces derniers, les analyses par western blot ont révélé 11 (1%) cas de coinfections àTc et Lb, 7 (0,6%) cas àLc et Tc et 17 (1,6%) cas àLb et Lc. Tous ces cas d'expositions à des infections mixtes àLeishmania sp et 8 cas sur 11 de Tc et Lb ont été confirmés par des analyses de PCR spécifiques et des hybridations de d'ADN. Deux cas à infection triple ont été observés. Nous considérons l'existence ces cas asymptomatiques avec des marqueurs phénotypiques et de génotypiques consistants pour des infections mixtes par deux kinétoplastides flagellés ou plus comme étant un facteur de risque élevé pour l'association avec des Psychodidae et des Triatominae qui sont des vecteurs se nourrissant de sang et transmettant ces infections à protozoaires. Ceci est la première publication rapportant l'exposition humaine à des infections asymptomatiques mixtes de kinétoplastides en Amazonie.
La falta de conservación del bosque tropical lluvioso Amazónico ha puesto en grave peligro a las poblaciones humanas que viven en poblados de reciente construcción, resultando en brotes de algunas enfermedades infecciosas. Hemos conducido un estudio seroepidemiológico con 1100 habitantes de 15 poblados del condado Paço do Lumiar, Brasil. 35 (3%) individuos habían estado expuestos a infecciones por Trypanosoma cruzi (Tc), 41 (4%) a Leishmania braziliensis (Lb) y 50 (4.5%) a Leishmania chagasi (Lc). Así mismo, 35 presentaban anticuerpos con reacción cruzada frente a antígenos heterólogos de kinetoplastidos. Entre estos individuos, los ensayos por ‘‘Western blot’’ revelaron que 11 (1%) tenían infecciones por Tc y Lb, que 7 (0.6%) tenían infecciones por Lc y Tc, y que 17 (1.6%) tenían infecciones por Lb y Lc. Todos los casos de exposición a infecciones mixtas con Leishmania sp, y 8 de los 11 casos de Tc y Lb fueron confirmados por ensayos específicos de PCR e hibridación de Southern. Dos casos tenían una infección triple. Consideramos estos casos asintomáticos, con marcadores fenotípicos y genotípicos consistentes con infecciones mixtas con dos o más flagelados kintoplástidos, un factor de alto riesgo en asociación con vectores Psychodidae y Triatominae que se alimentan de sangre y transmiten estas infecciones protozoarias. Esta es la primera publicación que muestra exposición humana a infecciones asintomáticas mixtas de protozoos kinetoplástidos en el Amazonas.
The infections by Lb can run asymptomatic in most vertebrate hosts but can cause mucocutaneous leishmaniasis (Follador et al. 2002). On the other hand, the infections by Lc can lead to visceral leishmaniasis (Kumari & Ram 2002). The infections by the Leishmania species causing diseases in man and domesticated animals are transmitted by phlebotomines (Diptera: Psychodidae) of the genus Lutzomyia in the American Continent (Scherlock et al. 1996; Rebelo et al. 1999; Miranda et al. 2002). Thirty-two sand fly species have been implicated in the transmission of leishmanias to humans, with 12 million infected people (Killick-Kendrick 1999).
It has been estimated that Tc infections affect approximately 18 million people, and that two-thirds of these infections are asymptomatic (reviewed in Teixeira et al. 2006). The Tc infections that cause human Chagas disease have been considered enzootic in the Amazon Basin, where the triatomine (Hemiptera: Reduviidae) vectors and mammal hosts dwell (Coura et al. 1999).
Considering the growing population density and the human predation of the periurban villages and the environmental features that favour insect-vector transmission of Lb, Lc and Tc infections, we decided to survey a population cohort using blood samples and search for specific phenotype and genotype markers allowing identification of each of these kinetoplastid infections. Here we show human exposure to mixed infections caused by these kinetoplastid flagellates in the human population of Paço do Lumiar County, State of Maranhão, Brazil.
Materials and methods
Study population and collection of human blood
This study is a vertical seroepidemiological survey of kinetoplastid infections in a population cohort inhabiting 15 villages of the Amazonian Paço do Lumiar County, State of Maranhão, Brazil. The field study was conducted by health personnel with clinical experience in identifying the human kinetoplastid infections. The survey comprised 0.7% of the county population. We collected 1100 samples of 5 ml of venous blood from the inhabitants of 180 households at 15 villages included in the research protocol. As positive controls we used 60 blood samples obtained from patients whose clinical diagnosis of Lb, Lc or Tc infections had been confirmed by parasitological examinations. Negative control serum samples were obtained from 30 healthy individuals inhabiting the Federal District of Brazil, an area considered to be Psychodidae and Triatominae insect-free. These negative controls showed three negative serological examinations for kinetoplastid infections. In addition we used serum samples from three mock control groups of five patients, each patient with proven Plasmodium falciparum, Mycobacterium tuberculosis or Treponema pallidum infection. Positive, negative control and mock samples were obtained from outpatients at the University of Brasília Hospital. The Faculty Committee on the Conduct of Human Research approved the research protocol. Informed consent was obtained from all patients.
The Tc epimastigotes and the Lc promastigote forms were grown in liver-infusion tryptose axenic medium. The promastigote forms of Lb were grown in Dulbecco medium containing 20% foetal calf serum and antibiotics, at 27 °C. The U937 human macrophage cell line was grown in the culture medium at 37 °C (Vexenat et al. 1996; Lauria-Pires et al. 2000).
Parasites. The Tc and Leishmania forms grown in liquid media were used for preparation of antigens. The exponential growth phase parasites were harvested from 100 ml of culture medium. The cells collected by centrifugation at 1500 × g for 10 min were washed three times in PBS pH 7.2. After fixation in 2% formaldehyde (v/v), dead parasites were washed and diluted in double distilled water to a final concentration of 6 × 103 cells/ml. Approximately 5 μl of the cell suspension were placed onto the surface of glass slides, air dried, wrapped in aluminum foil and stored at 4 °C until use.
Whole cell protein. The 10 × 108 culture forms were harvested and washed three times in PBS, pH 7.2. The cells collected by centrifugation were resuspended in 1:10 (w/v) double distilled water containing proteinase inhibitors (20 μm pepstatin, 20 μm Leupeptin, 10 μm TLCK, and 8 μm PMSF) and frozen at −20 °C. After three cycles of freezing and thawing, the ruptured cells were pelleted at 5000 × g for 20 min. The protein content in the supernatant contained the soluble cell antigen.
Enzyme-linked immunosorbent assay (ELISA) was performed with the soluble cell antigen (100 μl of a10 μg/ml protein solution in 0.1 m carbonate buffer, pH 9.6) to coat the microplate well surfaces. After 2-h incubation at 37 °C, the wells were washed three times with PBS containing Tween-20 at 0.05% (v/v) and blocked for 1 h with 0.5% fat-free milk powder in PBS. The human serum was diluted 1:200 in the washing buffer and 100 μl of the dilution were added per well. The tests were run in triplicate wells of micro plates, each coated with a specific parasite antigen. The plates were incubated for 2 h at 37 °C and washed three times as above and each well received 50 μl of peroxidase-conjugated anti human IgG + IgA + IgM (Zymed, Carlsbad, CA, USA) diluted 1:1200 in washing buffer. After further incubation for 2 h the plates were washed twice again and the enzymatic reaction was developed with o-phenylenediamine as substrate. The reaction was stopped with sulphuric acid 2.5 N, 50 μl per well. The absorbance values were determined with an ELISA reader (EIA multi well reader; Sigma, St. Louis, MO, USA). The positive and negative control sera were used in triplicates. A positive reaction was considered the serum dilution yielding absorbance 0.150 and above. This cut-off was determined in 60 serum samples
The indirect immunofluorescence assay (IF) consisted in the incubation of 20 μl of serum dilutions in PBS, pH 7.4, with the glass slides prepared with an aliquot of formalin-killed culture forms of Tc, Lc or Lb. After 1-h incubation in a moist chamber at 37 °C, the slides were thoroughly washed twice with PBS, air-dried and treated with a fluorescein-labelled anti-Igs, as described (Lauria-Pires et al. 2000). A positive apple-green reaction under the epifluorescence microscope (Axiophot Model, Carl Zeiss Mikroscopie, Jena, Germany) indicated a positive reaction in serum dilutions 1:40 and above.
The direct agglutination assay (DA) of formalin-killed Lc culture forms for identification of cryptic infections in serum dilutions 1:40 and above, and the indirect haemagglutination (IH) of the Rh negative blood group O, was used for identification of chronic Tc infections, as described in a previous paper (Vexenat et al. 1996). An agglutinating reaction obtained with 1:40 serum dilution was considered a cut-off, above which a positive Tc or Lc infection was identified.
The SDS-PAGE analyses of the whole cell protein (1 mg/ml) were performed in 10% gels. The nitrocellulose strips were incubated with borate-buffered saline solution (BBS), pH 7.2, containing 2% fat-free milk for 3 h with gentle shaking. The strips were then incubated with a 1:100 dilution of serum samples in BBS-milk for 12 h at room temperature (Vexenat et al. 1996). After three washings in BBS-milk the strips were treated with 4-chloro-naphtol and hydrogen peroxide to reveal the protein bands. High molecular weight markers were used, and the molecular mass of each band was calculated as recommended by the manufacturer (Life Technologies, Gaithersburg, MD, USA).
PCR amplification products, Southern blotting and hybridization
The human DNA was extracted from nucleated cells, which were obtained by Ficoll-Paque (Pharmacia) gradient from 5 ml of heparinized blood. DNA was extracted also from Lb and Lc promastigotes and from Tc epimastigotes, and from the U937 human macrophage cell line. PCR was conducted with primers specific to Lb, Lc and Tc kDNA (Table 1). Appropriate cycling times were chosen for each set of primers (Sturm et al. 1989; Andresen et al. 1997; Alexander et al. 1998). The PCRs were carried out using template DNA that was 20 times above the levels of detection with 100 ng of each pair of primers. The reaction proceeded with 0.5 IU of Taq, 0.2 mm of each dNTP in a final volume of 25 μl. Negative controls (milli-Q water and human DNA) and positive controls (Lb, Lc and Tc DNA) were used. The β-actin primer (Clontech Laboratories, Palo Alto, CA, USA) detected the quality of test DNA employed in the PCR amplification.
Table 1. Oligonucleotide primers used to detect Leishmania braziliensis, Leishmania chagasi, and Trypanosoma cruzi DNA in human blood nucleated cells
Southern hybridization of amplification products was performed to confirm the PCR specificity. The Tc amplification products that hybridized with a kDNA probe (AF399841) representing a constant region of Tc kDNA minicircle, yielded a 350-bp band (Sturm et al. 1989). Using the Lc specific primer set we obtained a PCR amplification product of approximately 800 bp (Barker 1987), which was employed to hybridize with test samples amplified from human DNA. Accordingly, an amplification product of 750 bp (Smyth et al. 1992) was used as a probe specific for the Lb group. The probes were radio labelled using 32P-dATP (sp. ac. 3000 ci/mmol) as recommended by the manufacturer (Life Technologies), and a maximum stringency of 0.1% SSC + 0.1% SDS, at 65 °C for 1 h was used.
We collected serum samples from 1100 inhabitants from 15 villages in Paço do Lumiar County, located on the outskirts of São Luiz City, the Capital of the State of Maranhão, Brazil. Over the last three decades the county population increased from approximately 15 000–100 000. An average of 6 ± 2 inhabitants per house was examined during the regular working day hours. Amongst the village inhabitants, a clinical and epidemiological survey was conducted, which failed to detect either acute symptomatic kinetoplastid infection or a silent sequel of the chronic infection. No past history of blood transfusion was recorded. Considering that an individual with a kinetoplastid infection should yield consistently positive results by two or more immunological assays specific to one type of infection, 35 individuals (3%) of the study population had antibodies against Tc infections, 41 (4%) yielded antibodies against Lb and 50 (4.5%) showed antibodies against Lc infections (Table 2). Additionally, 35 serum samples presented antibodies cross-reactive with the heterologous kinetoplastid antigens.
Table 2. Identification of kinetoplastid infections in inhabitants of Paço do Lumiar County, Maranhão State, Brazil
*Serologic diagnosis of a kinetoplastid infection by enzyme-linked immunosorbent assay, indirect haemagglutination, immunofluorescence assay, and direct agglutination assay performed with patient's serum and homologous antigen.
†Serum samples showing serologic assays positive with heterologous antigens were subjected to western blots; typical clusters of bands that formed with serum antibodies defined a type of infection (see text).
‡PCR amplification of the individual DNA yielded specific band sizes defining a kinetoplastid infection. Two patients showed mixed Tc, Lb and Lc infections, whose were identified by cluster of bands and PCR amplification products.
T. cruzi + L. braziliensis
L. braziliensis + L. chagasi
L. chagasi + T. cruzi
T. cruzi + L. braziliensis
L. braziliensis + L. chagasi
L. chagasi + T. cruzi
Figure 1 shows profiles of antibody titres detected by ELISA in the 1:200 serum dilutions from a total of 35 cases reacting with Tc, Lb and Lc antigens. In all these cases, the optical densities (OD > 0.150) were statistically different (>0.05) from those obtained from 30 negative control sera (OD < 0.150). In view of the serological evidence of antibodies reacting with each of these kinetoplastid antigens, we then subjected these samples to phenotype and genotype characterizations in order to determine whether they harbour the mixed infections.
The Western blot assays to detect specific serum antibodies reacting with protozoan proteins resolved in SDS-PAGE gels showed bands which had their relative molecular mass determined. Standard serum samples (cases) showing at least two serological examinations positive for Tc, for Lb, or for Lc were used to determine phenotypes of homologous protein bands. This procedure defined clusters of bands each appearing in above 25% of samples tested. Cases presenting antibodies exclusively against the Tc antigen revealed several protein bands but only those with molecular masses 230, 180, 92 and 67 kDa were present in excess of 25% of serum samples tested. Thus, the samples with antibodies exclusively against the Lb antigen formed protein bands of 222, 174, 145 and 125 kDa, whereas those reacting exclusively against Lc revealed bands of 25, 21, 18 and 15 kDa. These phenotype markers are illustrated in Figure 2a–c. Furthermore, 30 negative control serum samples formed no band at all, or they formed a paucity of weak bands, none of which had molecular masses similar to those present in the clusters of bands that were formed with antibodies specific for the kinetoplastid parasite antigens. These antibody-differentiated clusters of bands were used to detect the cases with mixed infections.
Using the criterion of cluster of bands, we observed that 11 cases (1%) had profiles of antibody reactivity against Tc and Lb, and seven cases (0.6%) had profiles of Tc and Lc antigens. The remaining 17 cases (1.6%) had antibodies forming clusters of bands typical of Lb and Lc (Table 2 and Figure 2d). Furthermore, the small molecular mass of protein bands in the clusters related to Lc, (25 kDa and below), contrasting with the characteristic high molecular mass cluster bands formed by the antibodies reacting with the Tc and Lb (67 kDa and above) antigens, suggested that the Western blot assay is a convenient method to achieve a ready diagnosis of visceral leishmaniasis.
In these assays, the specificity of the recognition of protein bands by antibodies was confirmed in positive mock controls, in which observed serum antibodies from patients harbouring M. tuberculosis, T. pallidum or P. falciparum infections formed bands with the Lb, Lc and Tc antigens, but none of which had same molecular masses as those present in the specific cluster of bands identifying each kinetoplastid infection (not shown). In this respect, the demonstration of serum antibodies recognizing two or more clusters of specific protein bands with the expected molecular masses suggests that the individual was concurrently infected by two or more kinetoplastid parasites. Interestingly, two cases (cases 7 and 35) had serum antibodies recognizing clusters of proteins specific for the Lb, Lc and Tc, antigens, that is, these individuals had three concurrent infections indicating blood feeding by three different vector types (Figure 2d).
Next, we co-validated previous results using specific primer sets for PCR amplification of template DNA from blood nucleated cells from the individuals showing antibodies that recognized a range of proteins in clusters formed with Lb, Lc and Tc antigens. The primer sets annealing selectively to the DNA of a kinetoplastid protozoan yielded PCR amplification products, whose specificity was demonstrated in each case by Southern hybridization with probes. These findings showing typical amplification products that hybridized with specific Lb, Lc or Tc, radio labelled probes and thus determining the genotype of the kinetoplastid protozoa made the diagnosis of the infection in that individual (Table 2 and Figure 3). Eight of 11 individuals showing mixed infections by Tc and Lb in the Western blot assay were confirmed by PCR amplification with the Tc primer set. Twelve individuals with Tc and Lc and 27 people with Lb and Lc infections were confirmed by PCR amplifications with their respective primer sets. Interestingly, two individuals (cases 7 and 35, Figure 2) showing clusters of protein bands consistent with having concurrent kinetoplastid infections also showed DNA amplification products with primer sets specific for Lb, Lc and Tc. The specificity of these PCR results was further demonstrated by the absence of amplification products in DNA templates from serologically negative test samples, and in negative controls shown in Figure 3.
We present serological and molecular evidence of exposure to asymptomatic mixed infections by kinetoplastid protozoa Tc, Lb and Lc in the inhabitants of Paço do Lumiar County. Although Bastrenta et al. (2003) had shown human mixed infections by kinetoplastid protozoa in patients with active cutaneous leishmaniasis and in one patient with visceral leishmaniasis, this is the first publication showing these asymptomatic kinetoplastids infections in a clinically healthy human population. These findings demonstrate that virulent flagellate protozoa sharing ecological niches with suitable invertebrate vectors and vertebrate hosts are a risk factor for clinical cases of the infections in the population. In the last two decades, outbreaks of clinical cases of visceral or mucocutaneous leishmaniases, and acute Chagas disease have been described in the County (Silva et al. 1985, 1997; Saldanha et al. 1995; Caldas et al. 2002). Our data are clear evidence that new outbreaks of clinically manifest kinetoplastid infections in people sharing households at the villages in the county can be expected.
The reliability of the assays used to determine seroepidemiological prevalence of kinetoplastid infections has been demonstrated. Although antibodies in sera of patients harbouring infections by Plasmodium, Treponema or Mycobacterium sp. tested positive in IF, ELISA and agglutination tests with kinetoplastid protozoa Lb, Lc and Tc, truly false-positive results were discarded by the Western blot assays, which showed that these antibodies formed a paucity of bands, and with molecular masses different from those present in the clusters defining typical infections by the kinetoplastid flagellates. We postulate that the antibody recognition of common antigenic determinants (serological cross reactions) in prokaryotes and in the flagellate protozoa causing human infections might be explained by existing conserved proteins resulting from evolutionarily symbiotic associations (Margulis & Sagan 2002; Simpson et al. 2002; Molinas et al. 2003; Vaughn et al. 2004).
In this study, we suggest that pressure (population growth, deforestation and human predation of the fauna over the past 30 years) for the transmission of parasitic infections by insect vectors in the domicile and peridomicile is leading to an increasing number of cases of indigenous mixed infections with Lb, Lc and Tc affecting the population living in Paço do Lumiar. Since these people live in poor socio-economic conditions sharing the same types of houses, we believe that detailed environmental, entomological and sociological investigations are required to clarify features associated with likelihood of acquiring kinetoplastid infections in the county.
We believe that the epidemiological patterns of infections that were revealed may reflect some variations in excess household risk factors. For example, considering that most participants were probably subjected to similar environmental exposures and that they were living in areas where various insect vectors transmitting the infectious agents occur sympatrically, it becomes difficult to explain why some people acquired mixed infections whereas others acquired none at all. Certainly, genetic or behavioural factors (such as use or not of repellents, protective clothing or mosquito nets) may play an important role in differentiating individuals harbouring kinetoplastid infections from those who do not in Paço do Lumiar County. Published genetic epidemiological and human behavioural studies of the determinants of susceptibility to these kinetoplastid infections are scarce. In one publication, however, a quantitative analysis determined that genetic factors could partially account for the pattern of seropositivity to the Tc infections; the maximum likelihood estimated that the role of heritability of these infections reached 56% of the observed variation in the infection status, and that an additional 23% of the variation could be attributed to the effects of shared environment (Williams-Blangero et al. 1997). Further environmental, behavioural and genetic study of factors influencing susceptibility of a human population to these infections may suggest new interventions for disease prevention and intervention programs.
To the Conselho Nacional de Pesquisa-CNPq and the Fundação Nacional de Saúde-FUNASA, Ministry of Health, Brazil, for financial support.