Angiostrongylus minasensis n. sp.: new species found parasitizing coatis ( Nasua nasua ) in an urban protected area in Brazil

https://doi.org/10.1590/S1984-29612019103 Abstract Currently, there are 21 species of Angiostrongylus that parasitize the pulmonary or mesenteric arteries of wild and domestic rodents, felids, canids and human. Species of Angiostrongylus have cosmopolitan distribution covering tropical, subtropical and temperate regions. The procyonid Nasua nasua (coati) is a reservoir host for a wide variety of parasites that may be harmful to its populations or may contain etiological agents with zoonotic potential. In urban areas, coatis are usually found in close association with humans and domestic animals. We morphologically and molecularly characterized a new species of Angiostrongylus found in N. nasua in a protected area within Belo Horizonte, Brazil. The new species of Angiostrongylus differs from other species of the same genus in terms of the length and bifurcation level of the lateral and ventral rays, the length of spicules and female tail morphology. Molecular phylogenetic results based on the mitochondrial cytochrome c oxidase subunit 1 gene suggest that the newly identified species belongs to a genetic lineage that is separate from other species of Angiostrongylus . This new species was collected from the mesenteric arteries of N. nasua . It was named Angiostrongylus minasensis n. Nós caracterizamos morfológica e de Angiostrongylus nasua área protegida de A de Angiostrongylus difere de outras espécies do mesmo gênero pelo comprimento e nível de bifurcação dos raios lateral e ventral, o comprimento dos espículos e a morfologia da cauda da fêmea. Resultados moleculares e filogenéticos baseados no gene mitocondrial citocromo c oxidase subunidade 1 indicam que a espécie recém-identificada pertence a uma linhagem genética separada de outras espécies de Angiostrongylus . O presente relato descreve uma nova espécie de Angistrongylus coletada das artérias mesentéricas de N. nasua , denominada Angiostrongylus minasensis n. sp..

The coati, Nasua nasua (Procyonidae) Linnaeus, 1766, is common in South America (Decker, 1991). Coatis live preferentially in forested environments, are diurnal and omnivorous. In urban environments, and thus in recreational areas, coatis may reach high population densities and are found in close association with humans and domestic animals (Crespo, 1982;Gompper & Decker, 1998;Alves-Costa et al., 2004).
We describe a new species of Angiostrongylus that was found parasitizing the mesenteric arteries of specimens of N. nasua that had been collected in a protected area in the state of Minas Gerais, Brazil. We characterized the new species by means of optical and scanning electron microscopy (SEM) and used molecular phylogenetic analysis to determine its relationships within the family Angiostrongylidae. The species was named Angiostrongylus minasensis n. sp..

Materials and Methods
Origin of the parasites Ten road-killed specimens of Nasua nasua (four adult females, four adult males and two young females) were collected in the Mangabeiras Municipal Park, an environmental conservation area located in Belo Horizonte, the capital of the state of Minas Gerais, Brazil (latitude 19°56'43" S and longitude 43°54'42" W), between August 2011 and August 2016. The carcasses were sent to the Helminthology Laboratory of the Institute of Biological Sciences, Federal University of Minas Gerais (UFMG). The study was authorized through the system for authorizations and for biodiversity information (SISBIO) of the Chico Mendes Institute for Conservation and Biodiversity [ICMBio/39817-1, ICMBio/54546-1] and by UFMG's ethics committee for use of animals [306/2017].
Nematodes were collected from the mesenteric arteries of the coati specimens. They were washed in 0.85% saline solution and fixed in 10% formalin at 80 °C for morphological and morphometric analyses, as described by Lima et al. (1985), or were placed in 100% ethanol and stored at -20 °C for molecular analysis. Feces were recovered from the rectum of all the necropsied animals, and then first stage (L1) larvae were isolated using the Baermann apparatus (Barçante et al., 2003).

Microscopy analysis
The paratypes (10 males and 10 females), included one holotype (male), and one allotype (female) were cleared in Amann's lactophenol. Measurements were made in millimeters (unless otherwise stated). These measurements are presented as holotype or allotype values, followed by the interval, mean and variation coefficient between parentheses.
Images were captured using a digital camera attached to an optical microscope. Drawings were made with the aid of a camera lucida. The new Angiostrongylus species was morphologically identified using the method of Anderson et al. (2009). Nematodes were prepared for conventional SEM as previously described (Lopes et al., 2013) and observed using the Zeiss Auriga Compact equipment at the urogenital research unit of the State University of Rio de Janeiro.

DNA isolation, PCR and sequencing
Genomic DNA was isolated from adult nematode specimens that were recovered from the mesenteric arteries of the coatis using the QIAGEN® QIAamp® DNA mini kit, following the manufacturer's protocol. The polymerase chain reaction (PCR) was carried out using a primer cocktail, as described by Prosser et al. (2013), to produce amplicons of the cytochrome c oxidase subunit 1 gene (MT-CO1), using the Invitrogen™ Platinum™ Taq DNA polymerase.
After checking the PCR products on agarose gel check, they were then purified using the GE Healthcare Illustra™ GFX™ PCR DNA and Gel Band purification kit, following the manufacturer's protocol. Purified amplicons were cycle-sequenced using the Applied Biosystems™ BigDye™ Terminator v3.1 cycle sequencing kit individually using each of the six primers from the above-mentioned cocktail (Prosser et al., 2013). Sequencing was performed using the ABI 3730 DNA analyzer at the DNA sequencing platform of the Oswaldo Cruz Institute (PDTIS/FIOCRUZ). Six overlapping fragments for each sample were assembled into contigs, and ambiguities were edited to generate a consensus sequence for each specimen, using the Geneious 9.1 software package with default parameters (http://www.geneious.com) (Kearse et al., 2012).

Phylogenetic analyses
Consensus sequences were aligned with GenBank sequences of representative members of the family Angiostrongylidae, four species of the metastrongyloid taxa (Protostrongylus rufescens, Metastrongylus pudendotectus, M. salmi and Parafilaroides normani) and two unpublished sequences of Heterostrongylus heterostrongylus. Two strongylid GenBank sequences were used as outgroups: Ancylostoma duodenale and Necator americanus (Table 1). The alignment was done as described by Maddison & Maddison (2015). The incidence of substitution saturation was verified using the test suggested by Xia et al. (2003) and Xia (2009). In addition, plots of Jukes & Cantor (1969) pairwise distances against transitions and transversions were used for each codon position. We carried out both tests using the DAMBE software, version 5.6.14 (Xia, 2013).
To account for different evolutionary processes at each of the three codon positions, ML and BI analyses were performed with distinct models per codon position.
ML-pairwise distances were computed using the same codon-based partitioned models (Schwarz, 1978). The robustness of nodes in ML was assessed by means of nonparametric bootstrap percentages (BP) after 1,000 pseudo-replicates had been produced, and by means of the expected-likelihood weights, which were applied to local rearrangements of tree topology (LR-ELW) after 1,000 replicates.
In the BI analyses, distinct GTR+I+Γ models were used for each codon position, with unlinking of base frequencies and parameters. Markov chain Monte Carlo (MCMC) sampling was performed for 10,000,000 generations with four simultaneous chains, in two runs. The robustness of nodes in BI was assessed by means of Bayesian posterior probabilities (BPP), which were calculated from trees that were sampled every 100 generations. The adequacy of BI analysis sampling was assessed using the Tracer software, version 1.6 (Rambaut et al., 2014). Presence of more than 100 effectively independent samples was considered sufficient. To save computational time, BI analyses were carried out on XSEDE cluster using the CIPRES Science Gateway (Miller et al., 2010).  Prevalence. 100% (n = 10). Type material. Holotype male (CHIOC number 38505a) and allotype female (CHIOC number 38505b). Deposited in the Oswaldo Cruz helminthological collection (CHIOC), Rio de Janeiro, Brazil.

Angiostrongylus minasensis
Etymology. The species name refers to the place where the new species was isolated. Description General. (Figures 2a and 3a) Body of both sexes filiform with transparent transversally striated cuticle. Oral aperture simple. Esophagus short and claviform. Excretory pore posterior to esophago-intestinal junction. Nerve ring anterior to the middle of esophagus. SEM results showed details of the cuticular topography, including important taxonomical structures. Anterior end of male and female is characterized by the presence of a circular oral opening surrounded by six sensory papillae and two amphids. The papillae are distributed in two dorsal and four lateroventral and amphids on the lateral position of the anterior end (Figure 4a and b). Dorsal side of the mouth presents a structure with a cleft, which was named a tooth-like structure (Figure 4b).  3-0.37 (0.34 ± 0.03)] from anterior end, respectively. Caudal extremity ventrally curved. Copulatory bursa symmetrical and small. Ventral rays have a common origin, bifurcating distally; lateroventral ray is longer and more robust than ventro-ventral ray. Lateral rays emerge from a common trunk and are similar in size; antero-lateral ray is robust and separate from the common trunk close to the base; mediolateral and posterolateral rays are separated close to the extremity. Externo dorsal rays are separate from the dorsal ray and are longer and more slender. The dorsal ray is reduced and bifurcates distally into two small branches. Spicules short, equal in size with a striated sheath, 0.25 mm [0.24-0.27 (0.25 ± 0.012)] length. Gubernaculum 0.039 mm [0.034-0.045 (0.04 ± 0.003)] in length. SEM results showed symmetrical copulatory bursa with quadrangular aspect and two spicules (Figure 4d). On the dorsal face, it was possible to identify the external lateral and dorsal rays (Figure 4d). Ventral, lateral and dorsal trunks were observed and, in addition, it was possible to differentiate the lateral trunk divided into three rays: anterolateral, mediolateral and posterolateral (Figure 4d and e). At the end of the anterolateral ray on both sides, near to the bursal edge, we identified one papilla (Figure 4d and f). The spicules were everted, trough-shaped, and the surface presented a striated external appearance (Figure 4f). In live specimens, the uterine tubules can be seen to spiral around the blood-filled intestine, and this can be seen through the transparent cuticle. Anus 0.09 mm [0.08-0.09 (0.09 ± 0.005)] from posterior end. Long tail ventrally curved with a tapered tip. SEM results showed that the vulvar opening was near to the posterior end of the body and the long and sharp tail (Figure 4c).
First stage larvae.
Although eight species of Angiostrongylus also have longer externodorsal rays than their dorsal rays, they can be differentiated from A. minasensis n. sp. by the lateral rays, the ventral rays and the length of the spicules. A. minasensis n. sp. has ventral trunk rays that are more robust, and its spicules are shorter than those of A. vasorum and A. gubernaculatus (Baillet, 1957;Dougherty, 1946;Duarte et al., 2007). Angiostrongylus minasensis n. sp. differs from A. chabaudi and A. costaricensis because its externodorsal ray is separated from the lateral trunk and its spicules are shorter (Biocca, 1957;Morera & Céspedes, 1970). The new species differ from A. taterona because the anterolateral ray bifurcates close to the base of the lateral trunk, while the mediolateral and posterolateral rays bifurcate in the distal third of the trunk. Moreover, the spicules are shorter (spicules of A. minasensis n. sp. measured 240-270 µm versus 500-620 µm in A. taterone) (Baylis, 1928).
The species that was found in N. nasua differs morphologically from A. costaricensis. The rays of the copulatory bursa of A. minasensis n. sp. lack dilations in the distal extremity and have a slender externodorsal ray. The externodorsal ray of A. costaricensis is approximately the same thickness as the rays of the lateral trunk, has the same dilation in the distal extremity and presents longer spicules. The posterior extremity of the female of A. minasensis n. sp. is longer and sharper than that of A. costaricensis and A. vasorum, which is roughly conical and slightly curved in both of these species (Morera & Céspedes, 1970;Duarte et al., 2007). The tails of females of A. siamensis and A. minasensis are of the same length and sharpness, but the copulatory bursa of males of A. siamensis is smaller and presents shorter rays (Ohbayashi et al., 1979).
Finally, A. minasensis n. sp. is morphometrically different from all other species found at the same site of infection (mesenteric arteries) and different from all other species found in the order Carnivora, as described in Table 2.

Phylogenetic analysis
The aligned sequences resulted in a matrix comprising 18 taxa and 702 characteristics, of which 459 were constant and 190 were variable characteristics that were informative for parsimony. The phylogenies described here were based on the MT-CO1 gene and were inferred using two different optimality criteria (ML and BI), which resulted in similar topologies with little variation in nodes and support values ( Figure 5). The ML method resulted in a tree with lnL score = -3642.256. The evolutionary models selected through BIC were as follows: HKY+G with fixed substitution rates (3,1,1,1,1,3); frequency of bases and gamma distribution optimized in the first codon position and independent J1+G with substitution rates; and frequency of bases and proportion of invariable sites optimized in the second and third codon positions.
The BI method resulted in values that were highly significant for all the parameters for the estimated size of the sample (ESS). The genera Angiostrongylus, Heterostrongylus, Aelurostrongylus and Parafilaroides formed a monophyletic group in both topologies with support ranging from low to moderate (LR-ELW = 76%, ML-BP < 50%, BPP = 86%). The genus Angiostrongylus was monophyletic in the BI topology with low support (BPP = 67%).
Angiostrongylus minasensis n. sp. was a sister to the specimen isolated from Nasua narica (Santoro et al., 2016) with strong support in all topologies (LR-ELW = 93%; ML-BP = 99%; BPP = 100%). The genetic lineage formed by Angiostrongylus in coatis was connected via basal polytomy to the lineage formed by A. vasorum and the lineage formed by A. cantonensis, A. malaysiensis and A. chabaudi, with low support in the BI topology (BPP = 61%).
ML-pairwise genetic distances of the MT-CO1 gene between Angiostrongylus species ranged from 6.8% between A. cantonensis and A. malaysiensis to 27.3% between A. chabaudi and A. costaricensis. Distances between genera ranged from 16.1% between Necator and Ancylostoma to 41.4% between Ancylostoma and Angiostrongylus. Distances between Aelurostrongylus, Heterostrongylus and Angiostrongylus species sequences ranged from 20.8% to 31.1%. The distance within Angiostrongylus taxa from coatis was 0.2%, whereas the distance within A. costaricensis was 13.3%. The distances between Angiostrongylus taxa from coatis and A. costaricensis taxa ranged from 16.2% to 18.5% (Table 3).

Discussion
The finding of helminths harbored in the mesenteric arteries of coatis from the Mangabeiras Park in Belo Horizonte (state of Minas Gerais, Brazil) initially suggested the possibility of infection by A. costaricensis, because this species has been reported infecting rodents and N. narica coatis in the Americas (Morera, 1973;Monge et al., 1978;Santoro et al., 2016). An alternative explanation was that the animals were infected by A. siamensis, which has been described infecting rodents in Thailand (Ohbayashi et al., 1979), since the latter species was also found infecting the mesenteric arteries of the caecum.
Given that Angiostrongylus spp. present low specificity to their vertebrate hosts, we examined the possibility that the species described here was A. vasorum, because this species has been described parasitizing domestic and wild canids in the state of Minas Gerais (Lima et al., 1985;Lima et al., 1994). However, morphological evaluation of the posterior extremity of the specimens of the present study that were found parasitizing coatis, in comparison with species found at the same infection sites and all the species found parasitizing carnivores, allowed us to conclude that the Angiostrongylus sp. described here could not be identified as any of the above mentioned species and is, therefore, a new species.
The ML-pairwise genetic distances of the MT-CO1 gene showed that A. minasensis n. sp. is significantly distant from the A. costaricensis taxa isolated from rodents in Brazil and Costa Rica. On the other hand, the genetic distance between A. minasensis n. sp. recovered from Nasua nasua in Brazil and A. costaricensis isolated from N. narica from Costa Rica is small and is only larger than the distance between the European sequences of A. vasorum. Similarly, the distances between the A. costaricensis taxa from N. narica and rodents, both from Costa Rica, are also significantly large. The phylogenetic analyses presented here suggest that A. minasensis n. sp. corresponds to a new taxon. Interestingly, the new species is closely related to the A. costaricensis taxon occurring in N. narica in Costa Rica.
Angiostrongylus costaricensis may also have low host specificity, since it parasitizes primates and the marsupial Didelphis virginiana, and Procyon lotor in Miami (Florida, USA) (Miller et al., 2006). However, L1 larvae isolated from faeces was not reported in these cases (Miller et al., 2006). In the present study, the occurrence of L1 in all the samples of feces from all the coatis that presented the adult parasite suggests that N. nasua is the natural host of A. minasensis n. sp. In this area, coatis occur at high population densities due to the absence of natural predators and the high abundance of food offered and discarded by humans (Porfírio et al., 2006). The Mangabeiras Park is located on the border between the metropolitan area of Belo Horizonte, which is subjected to anthropic pressure, and a conservation area (Serra do Curral), which could facilitate animal dispersion (Porfírio et al., 2006).

Conclusion
Angiostrongylus minasensis n. sp. is a new species of Angiostrongylus found in the mesenteric arteries of Nasua nasua in Brazil.