Abstracts

PUBLIC HEALTH

Entomological characterization and natural infection of anophelines in an area of the Atlantic Forest with autochthonous malaria cases in mountainous region of Espírito Santo State, Brazil

Caracterização entomológica e infecção natural de anofelinos em área de Mata Atlântica, com casos autóctones de malária, em regiões montanhosas do Espírito Santo

Helder R Rezende^I; Renata M Soares^II; Crispim Cerutti Jr^III,IV; Isabel C Alves^II,V; Delsio Natal^VI; Paulo R Urbinatti^VI; Tasciane Yamasaki^II; Aloísio Falqueto^I; Rosely dos S Malafronte^II,IV

^IProgr. Pós-graduação em Biologia Animal/UFES/FUNASA/SESA; Unidade de Medicina Tropical da Univ. Federal do Espírito Santo; heldericas@hotmail.com

^IIInst. Medicina Tropical de São Paulo da Univ. de São Paulo

^IIIDepto. Medicina Social e Unidade de Medicina Tropical da Univ. Federal do Espírito Santo

^IVDepto. Doenças Infecciosas e Parasitárias da Faculdade de Medicina da Univ. de São Paulo

^VHospital das Clinicas de São Paulo/LIM 49

^VIDepto. Epidemiologia da Faculdade de Saúde Pública da Univ. de São Paulo

ABSTRACT

Autochthonous malaria cases in the state of Espírito Santo, Brazil, are distributed in mountainous regions surrounded by the Atlantic Forest. While some aspects of this disease are unclear, detection of possible vector species can help to elucidate epidemiological uncertainties. Entomological and natural infection studies were carried out using anophelines (Diptera: Culicidae) captured in the municipality of Santa Tereza, ES. Monthly captures were made from March 2004 to February 2006. CDC-CO₂ traps were used from dusk (6:00 P.M.) to dawn (6:00 A.M.) to capture anophelines in the following habitats: near the houses, in open areas (at ground level) and inside, and at the margins of the forest (canopy and ground level). Shannon light traps were also used at the same locations of the CDC-CO₂ traps. A total of 2,290 anophelines within 10 species were captured. The relative frequency of Anopheles (Kerteszia) cruzii Dyar & Knab / A.(K.) homunculus Komp was the highest, with the majority captured in CDC-CO₂ traps installed in the forest canopy. The main species captured in Shannon traps was A.(Nyssorhynchus) strodei Root. The largest number of anophelines was captured from July to September and from 6:00 P.M. to 10:00 P.M. Anopheles (K.) cruzii is the probable vector for malaria transmission inside or near the Atlantic Forest fragments, but the role of other species cannot be ignored, as 53% of the sampled anophelines belonged to the subgenus Nyssorhynchus. The natural infection of A. cruzii, A. parvus (Chagas) and A. galvaoi Causey, Deane & Deane by Plasmodium vivax detected by PCR from DNA extracted from their thoraxes supports this view.

Key words: Anopheles cruzii, autochthonous malaria, Plasmodium, PCR, ecology

RESUMO

No Espírito Santo, os casos de malária autóctone estão distribuídos na região serrana próximo aos fragmentos de Mata Atlântica. Uma vez que alguns aspectos da doença são obscuros, a detecção das possíveis espécies de vetores pode auxiliar na elucidação de incertezas epidemiológicas. Estudos entomológicos e de infecção natural foram realizados com anofelinos (Diptera: Culicidae) capturados no município de Santa Tereza, ES. Capturas mensais foram realizadas de março de 2004 a fevereiro de 2006. Armadilhas CDC-CO₂ foram utilizadas do crepúsculo (18:00h) ao amanhecer (6:00h), para capturar anofelinos nos seguintes habitats: próximo ao domicílio e área aberta (solo), margem e interior da mata (solo e copa). Armadilhas Shannon também foram utilizadas nos mesmos locais que as de CDC-CO₂. Capturou-se o total de 2.290 anofelinos distribuídos em 10 espécies. A maior frequência relativa foi de Anopheles (Kerteszia) cruzii Dyar & Knab / A.(K.) homunculus Komp, sendo a maioria capturada em CDC-CO₂ instalada na copa da mata. A principal espécie capturada em armadilha Shannon foi A.(Nyssorhynchus) strodei Root. O maior número de anofelinos foi capturado entre julho e setembro das 18:00h às 22:00h. Provavelmente A.(K.) cruzii é responsável pela transmissão da malária dentro ou próximo aos fragmentos de Mata Atlântica. Entretanto, a participação de outras espécies não pode ser ignorada, visto que 53% da amostragem foi constituída pelo subgênero Nyssorhynchus. A detecção de Plasmodium vivax no tórax de A. cruzii, A. parvus (Chagas) e A. galvaoi Causey, Deane & Deane por meio de PCR reforça esse argumento.

Palavras-chave:Anopheles cruzii, malária autóctone, Plasmodium, PCR, ecologia

In general, the spatial distribution of anophelines (Diptera: Culicidae) follows that of malaria, as their species play an important role in this major health problem. The majority of cases notified in Brazil are reported in the Amazon region (576,963) (Ministry of Health 2005), but residual cases have also been notified in areas outside the Amazon, such as Southern and Southeastern Brazil. These areas have their own particular ecosystems and are covered by the Atlantic Forest (Pinotti 1951). Indigenous malaria in these regions is known as bromeliad malaria (Downs & Pittendrigh 1946).

The possible effect of the coexistence of anophelines, nonhuman primates and human beings on malaria transmission in these regions is not well understood, and little is known on the real dynamics of this infection in such ecosystem (Curado et al 1997, Duarte et al 2006, Cerutti et al 2007).

Extra-Amazonian autochthonous malaria cases are distributed in mountainous regions and are associated with agricultural activities near the forest (Cerutti et al 2007). Cases are restricted to these areas and there is no evidence of outbreaks that could affect either the tourist or agricultural potential.

Studies to identify possible vector species in the state of Espírito Santo are important for the public health system, as they can provide the basis for future preventive measures, and the results obtained can be extrapolated to other regions covered by Atlantic Forest in other states in Brazil.

Anophelines preliminary identification data and the nature of the landscape suggest that Plasmodium transmission in these regions is different from that in the Amazon (Espírito Santo State Secretary of Health, unpublished observations), due to the absence of Anopheles darlingi Root and A. aquasalis Curry, the small number of specimens of A.albitarsis s.l. Linch Arribálzaga, A. cruzii Dyar & Knab and A. bellator Dyar & Knab, and the presence of A. argyritarsis Robineau-Desvoidy, A. evansae (Brèthes), A. strodei Root and A. lutzi Cruz outdoors. The question as to whether these species can participate in the transmission of autochthonous malaria also remains to be answered. In here, we endeavor (a) to determine by means of an entomological survey whether Plasmodium transmission in a mountainous region of the state of Espírito Santo correlates with the existence of a dominant species and (b) to identify by means of molecular techniques which species of anophelines are infected and by which parasite species.

Material and Methods

Area description. Two strategies were used to capture anophelines. The first one, here referred to as "fixed point", was used at a site located at Valsugana Velha (19º57'58.4" S, 40º34'45.2" W and 790 m), in the municipality of Santa Tereza, where autochthonous malaria cases had been reported (Fig 1). This region is characterized by abundant rainfall and low temperatures. The second, referred to as "mobile points", was used at sites spread over an extensive region that includes eleven municipalities (about 5,000 km²) where autochthonous malaria cases were reported during this study. Twenty-four captures were made at the fixed point and 17 at mobile points.

Capture methods and periodicity. Two methods were used: (1) CDC-300gCO₂ light traps installed close to the houses in open areas (at flooor level), at the margin of the forest (canopy and ground) and inside the forest (also canopy and ground); and (2) Shannon traps close to the houses (mobile points) and at the margin of the forest (fixed point).

The six CDC-CO₂ traps were installed simultaneously, with two of them being placed at a height of 10 m in the canopy (at the margin and inside the forest). The source of CO₂ used in the CDC traps was dry ice (approximately 300 g by trap). Three members of the team carried out the captures at the Shannon traps one at a time, at a rotation of 4h per member, giving a total of 12h of captures, which followed a monthly schedule at the fixed point and a demand-based schedule (related to the number of malaria cases that occurred) at the mobile points. The traps were set from 6:00 P.M. to 6:00 A.M., and captures were performed from March 2004 to February 2006. Rainfall and relative-humidity data were obtained from a meteorological station in the municipality of Santa Tereza. Temperature data were estimated according to Feitoza et al (2001).

Specimens storage and identification. Captured specimens were kept in silica gel tubes or isopropanol and identified according to Consoli & Lourenço-de-Oliveira keys (1994). In spite of the researchers' expertise in identifying anophelines, the presence of sibling species led to initial misclassification of some specimens of A. homunculus as A. cruzii. To avoid confusion, in this report we refer to them as A. cruzii / homunculus specimens.

Molecular techniques.Anopheline mosquitoes captured at the various locations were assayed by PCR. DNA from the thoraxes and abdomens was extracted according to the method described by Oskam et al (1996) with modifi cations. The extractions from thoraxes and abdomens were carried out independently to separate mosquitoes that were carrying sporozoites in their salivary glands, and could be considered potential vectors, from those that were carrying oocysts in their abdomen. The supernatant was precipitated by adding two volumes of ice-cold absolute ethanol in the presence of 0.3 M sodium acetate (NaAc). After centrifugation, the DNA was rinsed with 70% ethanol, dried and resuspended in 50 µl of tris-EDTA buffer (TE). The PCR protocol was performed using primers and methodology described by Kimura et al (1997) and modifi ed by Win et al (2002). The products were eletrophoresed in 2% agarose gel and visualized under UV-light.

Statistical analysis. Spearman's rank correlation coefficient was used to test for a correlation between the number of anophelines and the weather variables (temperature, humidity and rainfall). The Kruskal-Wallis non-parametric test and the Berger-Parker dominance index were used to compare the number of anophelines captured at different periods and at different locations in the same period (P < 0,05). Data analysis was conducted using the BioStat 4.0 software.

Results

A total of 2,290 specimens within 10 species were captured, with A. cruzii being the most frequent (Table 1).

Variations in weather conditions (Figs 2, 3). Highest capture frequencies were observed in the colder, drier months of July and September, when the mean temperature and rainfall were 15.5ºC/29.3 mm and 18.1ºC/77.5 mm, respectively. The only month in which all ten species were captured simultaneously was September. A. cruzii, A. strodei, A. evansae and A. lutzi were present all year round, and A. albitarsis s.l. was only absent in January. Of all the anophelines captured, A. cruzii was the predominant species, except in September, and like A. lutzi, the highest capture frequency for this species was observed in July. The lowest frequency for A. cruzii was registered in May, when A. albitarsis s.l. achieved its higher frequency. Anopheles evansae and A. strodei were predominant in September (Fig 3). Although a negative trend was observed between the number of anopheline specimens, temperature (r = - 0,11; P = 0,7), rainfall (r = - 0,08; P = 0,8) and humidity (r = - 0,09; P = 0,7), and a positive trend observed between the number of A. cruzii specimens, temperature (r = 0,3; p = 0,3) and rainfall (r = 0,11; P = 0,7), neither were significant

Spatial distribution. Anopheles evansae, followed by A. strodei and A. albitarsis s.l., were the most frequently captured species in Shannon traps, whereas A. cruzii was the most frequent in CDC CO2 traps (Figs 4, 5). The higher frequency of A. cruzii in CDC-CO2 traps can be explained by their location in the canopy, where most of the specimens prefer to feed. The Parker dominance test revealed indices (d) of 0.9 for A. cruzii inside the forest and 0.8 at the margin, both recorded in the canopy (Fig 4). Anopheles strodei was the most frequent species captured in open areas (d = 0.4) and near the houses (d = 0.2), followed by A. evansae. Anopheles lutzi was frequent in Shannon traps near the houses and in CDC-CO2 traps in the forest canopy (d = 0.2). The same ten species captured in the forest were also captured in the open area and near the houses.

Hourly frequency. The samples obtained from 6:00 P.M. to 10:00 P.M. (64.8% of the specimens) was more representative than those obtained from 10:00 P.M. to 2:00 A.M. or 2:00 A.M. to 6:00 A.M. (P = 0.002; Kruskal-Wallis test). The smallest sample size was that obtained from midnight to 1:00 A.M. (1.9% of the specimens). However, some species, such A. albitarsis s.l., A. lutzi, A. evansae and A. strodei were active throughout the night (Fig 5).

Natural infection (Table 2). Only one pool of A. cruzii / A. homunculus was captured at ground level in the forest. All the anopheline specimens from the subgenus Nyssorhynchus infected by P. vivax were captured at the margins of the forest.

Discussion

Neotropical anophelines are classifi ed into fi ve subgenera: Nyssorhynchus, Kerteszia, Stethomyia, Lophopodomyia and Anopheles (Faran & Linthicum 1981, Sallum et al 1999, Wilkerson & Sallum 1999). In South America, species reported as vectors of human malaria belong to the subgenera Nyssorhynchus, Anopheles and Kerteszia (Deane 1986, Consoli & Lourenço-de-Oliveira 1994). While Nyssorhynchus mosquitoes, and Anopheles (N.) darlingi Root in particular, are the main malaria vectors in the Brazilian Amazon (Forattini 1962, Deane 1988, Lourenço-de-Oliveira et al 1989), a few species of the subgenera Anopheles [A. mattogrossensis Lutz & Neiva, A. peryassui Dyar & Knab and A. mediopunctatus s.l. (Theobald)] are considered less important vectors (Tadei & Thatcher 2000). Kerteszia is a small neotropical subgenus composed of only 12 species (Zavortink 1973) and its geographic distribution extends from Southern Mexico to Southern Brazil (Aragão 1964). Because of the association between Kerteszia mosquitoes and the bromeliads that serve as their breeding sites, Down & Pittendrigh (1946) created the term "bromeliad malaria" to describe malaria transmission vectored by Kerteszia mosquitoes, thus differentiating it from forest malaria, transmitted by anophelines of the subgenera Nyssorhynchus and Anopheles, which breed in other water collections.

Bromeliad malaria was a public health problem in Brazil during the nineteenth and early twentieth centuries (Gadelha 1994). During that period, the close contact between humans and the Atlantic Forest habitat as a result of population growth, the expansion of urban centers and the construction of railways provided conditions for extensive exposure of humans to mosquito bites and bromeliad-malaria transmission. The introduction of control measures to reduce the number of mosquito breeding sites through deforestation and the elimination of the Bromeliaceae population, together with the use of chemical insecticides and anti-malarial drugs over the years, resulted in a significant reduction in the number of malaria cases, from 40,000 in 1940 to only 71 in 1982 (Deane 1988).

Despite these control measures, malaria has never been eradicated from the states of São Paulo or Espírito Santo in Southeastern Brazil (Ministry of Health, 2005) or Santa Catarina (Machado et al 2003) in Southern Brazil.

The results of this entomological survey in Espírito Santo in an area covered by Atlantic Forest with a very low level of transmission of autochthonous malaria revealed a different profile from that of other Brazilian regions covered by the Atlantic Forest and do not support the assumption that A.(Ker.) cruzii / A.(Ker.) homunculus by itself is responsible for the maintenance of human malaria in these regions.

Given the presence of A. strodei, A. parvus, A. evansae and A. galvaoi in mobile capture points, one cannot disregard the co-participation of the subgenus Nyssorhynchus in malaria transmission. These species may be involved in malaria transmission in locations where they can be found in higher frequency and where their presence is probably related to environmental modifications that have reduced the number of sources of domestic and wild blood. However, it can be argued that the fact that these species were found in areas of low endemicity or that malaria cases were not as frequent in Espírito Santo as in the Amazon region constitutes evidence against their role in the transmission of the disease. Our fi nding of P. vivax in the thorax of A.parvus and A.galvaoi, however, reinforces the argument that they are, in fact, acting as vectors in these locations. It is possible that these Nyssorhynchus specimens could have become infected by Plasmodium when biting human beings, but the absence of malaria outbreaks leads us to believe that human-to-human transmission is very low.

Several authors have reported A. cruzii in low frequency in this region of Espírito Santo (Deane et al 1968, Rezende et al 2005). We believe that we were successful in capturing this species because CDC-CO₂ traps were used. A. cruzii was captured with higher frequency in colder, drier months. This finding could be explained by its adaptation to mountain habitats, like other species in the subgenus Kerteszia (Zavortink 1973). The majority, however, were captured inside the forest using CDC-CO₂ traps or during the crepuscular peak using the Shannon trap. Inside the forest, the temperature is milder and there is less variation in weather conditions.

A relationship between rainfall and the maintenance of breeding sites was observed in Valsugana Velha. At this location, one stream close to the margin of the forest dried up in the drier months, limiting the availability of breeding sites for anophelines of the Nyssorhynchus and Anopheles subgenera in the forest environment. Near the houses, on the contrary, the level of water in the breeding sites remained constant, favoring the persistence of these subgenera close to the human population. A different picture was observed for breeding sites for A.cruzii / A.homunculus, as the bromeliads continued to hold water despite the dry season.

The diversity of the species captured in Valsugana Velha (fixed point) and in the mobile points was similar, although there were differences in the number and frequency of each species captured. While A. cruzii / A. homunculus predominated in Valsugana Velha, A. lutzi was the most abundant species in the mobile points. CDC-CO₂ traps showed better results in Valsugana Velha than in mobile points, where Shannon traps were more efficient. The different traps were found to inflouence the species captured, with the best results for A. cruzii /A. homunculus being obtained with CDC-CO₂ traps.

A higher species diversity was expected inside the forest (Guimarães et al 2000), but the diversity was in fact lower than that observed near the houses, where there are more human beings, probably because of the availability of food, shelter and breeding sites. Unlike the findings of previous studies in Southeastern Brazil (Deane et al 1984, Deane 1992, Tubaki et al 1993), our results showed a decrease in frequency of A. cruzii / A. homunculus as the distance from the forest and canopy increased, refloecting their limited domiciliary habits and vertical dispersal. However, other species, such as A. strodei, were found to display different behavior, with their frequency increasing near houses and decreasing inside the forest, indicating possible replacement of A. cruzii by other species.

Anopheles cruzii is frequently considered an acrodendrophilic species (Deane et al 1968, Deane 1992), as it breeds in bromeliads in the canopy and prefers to feed on simian and bird species. It appears that the other species, unlike A. cruzii / A. homunculus, can adapt more easily and live in areas modified and inhabited by human beings (Deane et al 1984, Guimarães et al 2000, 2003, Forattini 2002). Based on this assumption, our findings may demonstrate the replacement of A.cruzii / A. homunculus by A. strodei, A. albitarsis s.l. and A. evansae, as the environment is progressively transformed by humans. Nevertheless, as we were still able to find these wild species (A. cruzii / A.homunculus) in Valsugana Velha and in the mobile points, we conclude that this transformation is not complete. Interestingly, almost 80% of bromeliad-malaria cases in Espírito Santo occur in males (Cerutti et al 2007). In such a scenario, A. cruzii could be maintaining malaria transmission in human males, as women do not normally enter the forest. If A. cruzii is considered the probable vector in this region, malaria is probably being transmitted as a zoonosis, as this species prefers to remain in the canopy feeding on nonhuman primates.

The majority of anophelines were captured from 6:00 P.M. to 10:00 P.M., and the nycthemeral activity of A.albitarsis s.l., A. cruzii, A. evansae, A. lutzi and A. strodei was observed all night long. The risk of contact between vectors and humans increases at the margin of the forest from 6:00 P.M. to 10:00 P.M. during the month of July.

Serological studies revealed a high frequency of antibodies against peptides of the circumsporozoite protein corresponding to

Plasmodium vivax variants (VK210 and VK247),

P.malariae / P. brasilianum and human

P.vivax-like /

P. simiovale in local human populations and in different wild monkey species. Also, PCR analysis revealed that one inhabitant was infected by

P. malariae, suggesting, once more, transmission in the context of a zoonosis, with monkeys acting as malaria reservoirs (Curado

et al 1997, 2006, Duarte

et al 2006).

The nycthemeral activity of A. cruzii captured in Espírito Santo was different from that of A. cruzii reported in other areas. This observation reinforces the hypothesis proposed by several authors that A. cruzii is, in fact, a complex of sibling species instead of a monotypical (Ramirez & Dessen 2000, Carvalho-Pinto & Lourenço-de-Oliveira 2004, Malafronte et al 2007). Such possibility was also supported by the initial misclassifi cation of A. homunculus as A. cruzii in this study (Zavortink 1973, Rosa-Freitas 1998, Forattini 2002).

The few studies that deal with the natural infection of anophelines used different techniques to detect Plasmodium species in Brazilian mosquitoes, including immunoassays (Arruda et al 1986, Branquinho et al 1997) and comparison between PCR and immunoassay techniques (Moreno et al 2004). Anopheles cruzii is still involved in human and simian malaria transmission in valleys in the Atlantic Forest in the states of São Paulo and Santa Catarina (Carvalho et al 1988, Branquinho et al 1997, Curado et al 1997), and when tested by ELISA were found to be infected with P. vivax and one of its variants, P. vivax VK247 (Branquinho et al 1997).

In this study, the PCR technique yielded interesting results. In order to detect potential vectors and those that are carrying sporozoites in their salivary glands (and could consequently be considered vectors), the mosquitoes were split into two parts - the thorax and the abdomen. The positive results obtained with tests performed on the thoraxes indicate that A. cruzii is not the only species involved in malaria transmission, but that the subgenus Nyssorhynchus may also play a role. In fact, it would appear more likely that A.cruzii is involved in the transmission of simian malaria, as the majority of the P. vivax positive specimens were captured in CDC-CO₂ traps in the canopy (minimum infection rate = 0.5%), whereas Nyssohynchus species were captured in Shannon traps at the margins of the forest around human dwellings. Because there is evidence of identity between P. vivax and P. simium (Li et al 2001), such a possibility deserves further study.

A. cruzii / A. homunculus specimens were found infected by P. falciparum (data not shown) in this study and it was surprising, as bromeliad malaria has traditionally been related to P. vivax infections (Carvalho et al 1988, Curado et al 1997). The fact that these specimens were not captured in the CDC-CO₂ traps in the canopy means that the relationship between P. falciparum and simian malaria is less clear-cut. We cannot dismiss the possibility of false positive PCR results, but P. falciparum infections were reported in inhabitants of others Atlantic Forest regions of Brazil (Curado et al 2006). The question also arises as to whether this positive result for P. falciparum could represent evidence of a modified strain, as P.falciparum is always transmitted by anophelines of the Nyssorhynchus subgenus. It is possible that transmission by another subgenus could impose some modification on strain behavior (Li et al 2001). Our fi ndings of P.falciparum in asymptomatic local inhabitants add to such a possibility, which has been associated by others with the antigenic variability present in several isolates of the parasite (Rich et al 2000, Awadalla et al 2001).These preliminary results point to the need for further studies to help on understanding the transmission cycle of extra-Amazonian malaria infection.

Acknowledgments

To Agenor, Edmar, Edson, Gustavo, João, Jorge and Mauro for technical support; to Anselmo and Lúcia, who live in the study area; and to FAPESP (grant number 2003/07631-4).

Received 07/XII/07. Accepted 22/I/09.

Edited by Álvaro Eiras - UFMG

Publication Dates

History