INTRODUCTION
Cats are known to be susceptible to leishmaniasis, but their participation in the cycle of the disease is still unclear. Since the description of the first case of leishmaniasis in a domestic cat in 1912 (Sergent et al. Reference Sergent, Sergent, Lombaard and Quilichini1912), several cases of leishmaniasis have been reported in cats (Schubach et al. Reference Schubach, Figueiredo, Pereira, Madeira, Santos, Andrade, Cuzzi, Marzochi and Schubach2004; Simões-Mattos et al. Reference Simões-Mattos, Bevilaqua, Mattos and Pompeu2004; Da Silva et al. Reference Da Silva, Cândido, Pereira, Brazil and Carreira2008; Vides et al. Reference Vides, Schwardta, Vicente Sobrinho, Marinho, Laurentic, Biondo, Leuteneggere and Marcondesa2011), occurring mainly in visceral leishmaniasis (VL) endemic areas (Saló et al. Reference Saló, Gonzalez, Altimira and Vilafranca2007; Maia et al. Reference Maia, Nunes and Campino2008; Costa et al. Reference Costa, Rossi, Laurenti, Gomes, Vides, Sobrinho and Marcondes2010).
Although dogs are the main urban reservoir hosts of zoonotic VL (Souza et al. Reference Souza, Luz and Rabello2008), it has been observed by xenodiagnosis that cats infected by VL are able to infect Lutzomyia longipalpis (Maroli et al. Reference Maroli, Pennisi, Di Muccio, Khoury, Gradoni and Gramiccia2007; Silva et al. Reference Silva, Rabelo, Gontijo, Ribeiro, Melo, Ribeiro, Melo, Ribeiro and Michalick2010), and are thus potential reservoirs of this zoonotic disease. Studies about feline LV are limited and the disease in cats is still considered neglected (Pennisi et al. Reference Pennisi, Cardoso, Baneth, Bourdeau, Koutinas, Miró, Oliva and Solano-Gallego2015).
VL is a zoonotic disease that occurs in the Old and New World, and Brazil is one of the countries with the highest incidence of the disease (Alvar et al. Reference Alvar, Velez, Bern, Herrero, Desjeux, Cano, Jannin and Den Boer2012). The first case of natural Leishmania infection in a feline in Brazil was recorded in 1939 in a cat with ulcers on the ears and nose, in the state of Pará, but the species causing the infection was not identified (Mello, Reference Mello1940).
Other studies conducted in Brazil reported Leishmania spp. infection in domestic cats in Aurá, state of Pará (Mello, Reference Mello1940), Leishmania subgenus Viannia in Belo Horizonte, state of Minas Gerais (Passos et al. Reference Passos, Lasmar, Gontijo, Fernandes and Degrave1996), L. infantum in Cotia, state of São Paulo (Savani et al. Reference Savani, Camargo, Carvalho, Zampieri, Santos, D’áuria, Shawc and Floeter-Winter2004), Araçatuba, state of São Paulo (Vides et al. Reference Vides, Schwardta, Vicente Sobrinho, Marinho, Laurentic, Biondo, Leuteneggere and Marcondesa2011) and Andradina, state of São Paulo (Coelho et al. Reference Coelho, Richini-Pereira, Langoni and Bresciani2011), and L. amazonensis in Campo Grande, state of Mato Grosso do Sul (Souza et al. Reference Souza, Barros, Ishikawa, Ilha, Marin and Nunes2005) and Ribas do Rio Pardo, state of Mato Grosso do Sul (Souza et al. Reference Souza, Nunes, Borralho and Ishikawa2009. The presence of anti-Leishmania spp. antibodies in cats has also been reported in other regions in Brazil, such as Barra Mansa, state of Rio de Janeiro (Figueiredo et al. Reference Figueiredo, Bonna, Nascimento, Costa, Baptista, Pacheco, Amendoeira and Madeira2009) and Recife and Petrolina, state of Pernambuco (Silva et al. Reference Silva, Ramos, Pimentel, Oliveira, Carvalho, Santana, Faustino and Alves2014).
Several clinical signs have been observed in infected cats, such as the presence of ulcers in the ears and nose (Mello, Reference Mello1940), interdigital ulcers (Passos et al. Reference Passos, Lasmar, Gontijo, Fernandes and Degrave1996), nodes in the nose, weight loss, dehydration, swollen lymph nodes (Savani et al. Reference Savani, Camargo, Carvalho, Zampieri, Santos, D’áuria, Shawc and Floeter-Winter2004), skin alterations (Bresciani et al. Reference Bresciani, Serrano, Matos, Savani, D'auria, Perri, Bonello, Coelho, Aoki and Costa2010) and facial dermatitis with scabbing and sores and in palmar and plantar areas (Coelho et al. Reference Coelho, Lima, Amarante, Langoni, Pereira, Abdelnour and Bresciani2010). However, the clinical diagnosis is complex and difficult because other skin diseases are common among these animals.
The investigation of the history and the relation of epidemiological aspects are important points to consider for obtaining a correct clinical diagnosis because, in practice, skin diseases, which account for the majority of cases in the clinical routine of these animals, can be confused with some of the signs found in leishmaniasis.
Given that cats can be infected by several species of Leishmania, and that these animals do not always present a detectable humoral immune response by the techniques used in canine serological diagnostics (Poli et al. Reference Poli, Abramo, Barsotti, Leva, Gramiccia, Ludovisi and Mancianti2002; Solano-Gallego et al. Reference Solano-Gallego, Rodriguez-Cortés, Iniesta, Quintana, Pastor, Espada, Pórtus and Alberola2007), the species of parasite involved in the infection should be identified and characterized (Vides et al. Reference Vides, Schwardta, Vicente Sobrinho, Marinho, Laurentic, Biondo, Leuteneggere and Marcondesa2011).
The diagnostic confirmation and correct identification of Leishmania spp. is important for proper species-specific therapy as well as for epidemiologic studies. The method, internal transcribed spacers 1 (ITS1), polymerase chain reaction (PCR), restriction fragment length polymorphism (RFLP), using the restriction enzyme approach was developed as an alternative to existing diagnostic procedures such as direct detection of parasites by microscopic examination of clinical specimens or by cultivation (Schonian et al. Reference Schonian, Nasereddin, Dinse, Schweynoch, Schallig, Presber and Jaffe2003).
Due to the increase in reported cases of Leishmania spp. and especially of L. infantum in domestic cats, the aim of this study was to use a diagnostic method for leishmaniasis, which combines high sensitivity with the differentiation of species, to determine the occurrence of L. infantum infection in domestic cats and to investigate the importance of this species in the epidemiological chain in an area where VL is highly endemic, the municipality of Teresina.
MATERIAL AND METHODS
Animals and samples
The study involved 83 domestic cats of both sexes and a variety of breeds, ages, sizes and weights. The inclusion criteria were animals from the Collects were conducted in the period of February to June of 2013. Inclusion criteria were 10% of attended cats in University Veterinary Hospital from Federal University of Piauí (HVU/UFPI) (n = 22) and 10% of the total population of cats of two neighbourhoods (Planalto Uruguai and Parque Itararé) of the municipality of Teresina, with high LV prevalence, according to information of Zoonosis Management of Teresina, using random houses.
The animals were subjected to a clinical examination, weighed and their nutritional status was assessed. To collect biological material, the animals were given an intramuscular anaesthetic of acepromazine (0·1 mg kg−1) and ketamine (5 mg kg−1). Five millilitre of blood was then drawn into EDTA vacutainer tubes for complete blood count (CBC) and 5 mL in vacutainer tubes without anticoagulant for biochemical tests, serology for retroviruses and Leishmania, as well as popliteal lymph node aspirates to examine for Leishmania.
Parasitological tests
Samples of popliteal lymph node aspirate were placed in test tubes containing NNN (Novy, MacNeal, Nicolle) biphasic culture medium with 1 mL of Schneider's supplemented with penicillin (100 IU mL−1), streptomycin (100 ug mL−1), and 10% fetal bovine serum, which were then incubated in a Biochemical Oxygen Demand (B.O.D.) chamber at 25 °C. The culture was examined every 5 days up to day 20 in an optical microscope under 40 × magnification to observe the promastigote forms of the parasite. Popliteal lymph node sample was also performed smeared on a slide and stained by the Giemsa method to visualize Leishmania amastigotes.
Haematological and biochemical tests
The CBC was performed in a Mindray BC-2800 Vet automatic haematology analyser and the differential leucocyte count in Giemsa stain blood smears.
Serum biochemical quantifications of urea, creatinine, aspartate aminotransferase (AST), albumin and total proteins were performed in a BIOCLIN BA-88 semi-auto chemical analyser using BIOCLIN kits, as recommended by the manufacturer. The globulin concentration was calculated by subtracting the albumin from the total protein.
Detection of feline leukaemia virus (FeLV) antigen and antibodies to feline immunodeficiency virus (FIV)
To evaluate the impact of immunosuppressive retroviruses, 83 of them were tested for FeLV antigen and for FIV antibody. The detection of FeLV antigen (p27) and FIV antibody was performed using a commercial assay kit (SNAP® FIV Antibody/FeLV Antigen Combo Test; IDEXX Laboratories, Westbrook, ME), according to the manufacturer's recommendations.
Leishmaniasis detection enzyme-linked immunosorbent assay (ELISA)
An ELISA protocol was followed as recommended by the manufacturer, using FIOCRUZ Bio-Manguinhos kits (EIE – Canine Visceral Leishmaniasis – Bio-Manguinhos) (Ashford et al. Reference Ashford, Badaro, Eulalio, Freire, Miranda, Zalis and David1993; Alves & Bevilacqua, Reference Alves and Bevilacqua2004), with minor modifications. The sera were diluted 1:200 and anti-mouse IgG conjugated cat. A20-120P at 1:80 000. Testing was performed in duplicate, and the procedure was repeated on different days using different slides to confirm the results. Optical density (OD) was measured at 450 nm in a URIT-660 Microplate Reader (URIT Medical Electronic Co., Ltd, Guangxi, China). Sera of the 10 control cats were placed on the same plate to calculate the limits of the cutoff. These animals became from neighbourhoods from Teresina where LV (canine and human) presents low prevalence and negative bone marrow PCR. The cutoff was calculated multiplying the mean of the negative controls by two (approximately two standard deviations). In addition, the manufacturer of kit suggests adding 20% to the cut-off value to increase the specificity of the test. Therefore, animals that presented optical density higher than the cutoff plus 20% were considered positive.
DNA Extraction
The DNA of the promastigote forms was extracted using the Genomic DNA from tissue kit (NucleoSpin® Tissue, Macherey-Nagel, Durën, Germany), following the protocol described by the manufacturer.
Molecular analysis and identification of isolates
Leishmania-specific ribosomal ITS1 gene was amplified using the primers LITSR (5-CTG GAT CAT TTT CCG ATG-3) and L5·8S (5-TGA TAC CAC TTA TCG CAC TT-3). The PCR reaction mixture contained 1·0 µL of DNA solution, 0·2 mm of each dNTP, 0·1 nm of each primer, 2·5 U of AmpliTaq Gold® (Applied Biosystems), 2·5 µL of 10× buffer [Tris–HCl 50 mm, (pH 8·3), KCl 50 mm] and 2·0 mm MgCl2 in a final volume of 15 µL. Positive controls with L. infantum (MHOM/1973/BH46), L. amazonensis (IFLA/BR/1968/PH8) and L. braziliensis (MHOM/1975/M2903) were used. Genomic DNA was used at 1·0 ng µL−1. All the tests included a negative control without DNA. The amplifying conditions were: initial denaturation at 95 °C for 2 min, 32 cycles at 95 °C for 20 s, 53 °C for 1 min, 72 °C for 1 min, and a final extension at 72 °C for 6 min. The amplicons of about 300–350 base pairs (bp) fragments were analysed at 100 V in buffer (89 mm tris (pH 7·3), 89 mm boric acid and 2 mm EDTA) in 5% acrylamide/bis-acrylamide 39:1 gels stained with silver nitrate.
PCR products (5–10 µL) were digested with endonuclease HaeIII enzyme, according to the manufacturer's instructions. Restriction fragments were analysed at 100 V in buffer [89 mm tris (pH 7·3), 89 mm boric acid and 2 mm EDTA] in 5% acrylamide/bis-acrylamide and compared with the Leishmania species used as control.
Analysis of data
The Kappa agreement index was determined using version 8.2 of the SAS System for Windows (Statistical Analysis System) (SAS Institute Inc., Cary, NC, USA).
RESULTS
The prevalence of VL was 4% (3/83) by ELISA (Fig. 1), and 4% (3/83) animals presented L. infantum in popliteal lymph node aspirate, in both Giemsa-stained smear and culture medium.
Fig. 1. ELISA anti-Leishmania antibody assay of 83 domestic cats in Teresina, PI, Brazil. Dots (optical density of each animal). The horizontal line indicates the cutoff (mean of the negative controls × 2 × 1 2). The three animals were considered positive with optical density value above of the cutoff.
The agreement between tests (parasitological and ELISA) was good (0·31, P = 0·005), based on the classification of the agreement index Kappa described by Landis & Koch (Reference Landis and Koch1997).
Among the 83 animals studied, only one of the cats with positive parasitological tested positive by ELISA, was also positive for FIV, presenting cachexia as a clinical sign suggestive of leishmaniasis. This was the only one of the 83 cats that presented a change in the clinical evaluation. The three cats with positive parasitological were females and adults (>6 months).
In the three animals with positive parasitological examination, a more detailed evaluation of possible biochemical changes was performed. In the haematological tests, animal I presented normocytic normochromic anaemia, with white blood cell count (WBC) within the normal range (Tables 1 and 2). Serum biochemistry showed urea levels above the physiological parameters, hyperproteinaemia, hypergammaglobulinaemia, hypoalbuminaemia and low albumin/globulin (A/G) ratio. The ELISA was positive for Leishmania, FIV test positive and FeLV negative (Table 3).
Table 1. Haematological parameters of the cats under study
Table 2. CBC of cats with positive parasitological examination for leishmaniasis
a (Jain, Reference Jain1993).
b Mean corpuscular volume.
c Mean corpuscular haemoglobin concentration.
Table 3. Biochemical and serological tests for the identification of retrovirus in cats with positive parasitological examination for leishmaniasis
a (Kaneko et al. Reference Kaneko, Harvey and Bruss1997).
Animal II presented normocytic normochromic anaemia, and WBC showed eosinophilia (Tables 1 and 2). Serum biochemistry showed urea levels above the physiological parameters. ELISA (VL), FIV and FeLV tested negative (Table 3).
Animal III presented normocytic normochromic anaemia, with WBC showing neutrophilic leucocytosis (Tables 1 and 2). Serum biochemistry showed urea levels above the physiological parameters, hyperproteinaemia, hypergammaglobulinaemia, hypoalbuminaemia, and therefore, low A/G ratio. ELISA (VL), FIV and FeLV were negative (Table 3).
The molecular biology revealed the presence of amplified fragments of 300–350 pb in the strains isolated from the three infected cats. The causative parasites were identified as L. infantum using ITS1 PCR and subsequent RFLP analysis (Fig. 2).
Fig. 2. ITS1 PCR gel electrophoresis of DNA of Leishmania infantum isolates from three cats (lines 1, 2 and 3 represent cats 1, 2 and 3, respectively). Top panel 300–350 bp, Leishmania ITS1 PCR product. Bottom panel shows the HaeIII restriction fragments of the ITS1 PCR product. Note the band pattern, identical to the pattern of L. infantum control (line 4). Line 5, L. braziliensis; line 6, L. amazonensis; NC, negative control reaction; M, molecular marker.
DISCUSSION
Teresina, the state capital of Piauí, is located in northeastern Brazil, 72 m above sea level, at 05°05′21″ latitude, 42°48′07″ longitude. This region is considered endemic for human and canine VL. Transmission of the disease in the city started in 1981, aided by the precarious living conditions of the population and the environmental changes caused by the intense urbanization process, which presumably facilitated the vector's adaptation to the urban environment (Costa et al. Reference Costa, Pereira and Araujo1990).
The provenance and domicile of cats in regions endemic for VL, living close to infected humans, dogs and rats, may trigger the transmission of Leishmania spp., causing them to participate as accidental hosts. The three infected cats of this study came from the same neighbourhood on the outskirts of Teresina, most of which lack basic sanitation and have tree vegetation in the proximities. The brick houses were unfinished.
The presence of phlebotomine sand flies in the city is widespread, and according to the reports of the Teresina Zoonoses Management, L. longipalpis can be captured throughout the year, with an incidence of 44·4% during the rainy season (December, January and February) (Gezoon, 2010).
The non-selective and remarkably zoophilic feeding behaviour of some sandfly species such as L. longipalpis, which is present in and around human dwellings, reinforces the hypothesis that felines can act as a food source for the vector, favouring Leishmania spp. infection (Dantas-Torres et al. Reference Dantas-Torres, Simões-Matos, Brito, Figueiredo and Faustino2006).
The low prevalence of Leishmania spp. in cats may be attributed to ineffective diagnosis and the natural resistance of the cat to the parasite, due to genetic factors (Maia et al. Reference Maia, Nunes and Campino2008). This possible resistance was observed in an experimental infection performed by Kirkpatrick et al. (Reference Kirkpatrick, Farrell and Goldschmidt1994), who detected the presence of L. infantum in the post-mortem inspection of all the animals in the 4th week of infection, but detected no parasites in the 24th week.
The ELISA-based serological test, the prevalence was 4% (3/83), but only one of these tested positive for Leishmania spp. by parasitological method. Silveira-Neto et al. (Reference Silveira-Neto, Sobrinho, Martins, Machado, Marcondes and Lima2011), who evaluated serum samples from cats in Araçatuba, state of São Paulo, an area endemic for human and canine VL, using ELISA with crude antigen, fucose–mannose ligand-ELISA and ELISA-rK39, reported that 23, 13·3 and 15·9% of the analysed samples, respectively, were positive.
Vides et al. (Reference Vides, Schwardta, Vicente Sobrinho, Marinho, Laurentic, Biondo, Leuteneggere and Marcondesa2011), who studied Leishmania in cats with skin lesions, demonstrated that the congruence between serological and parasitological tests was very low. In fact, when they evaluated the results of the positive animals, they found that most of the cats with positive serology showed negative parasitology, and those with positive parasitology showed negative serology. This incongruence may be related to the natural resistance to infection by L. infantum of cats, which usually do not develop the severe visceral form but a cutaneous form with low production of specific antibodies (Solano-Gallego et al. Reference Solano-Gallego, Rodriguez-Cortés, Iniesta, Quintana, Pastor, Espada, Pórtus and Alberola2007).
Only one animal in this study presented clinical signs suggestive of VL, a fact that is corroborated by other reports which state that a large proportion of these animals are subclinically infected (Vita et al. Reference Vita, Santori, Aguzzi, Petrota and Luciana2005; Solano-Gallego et al. Reference Solano-Gallego, Rodriguez-Cortés, Iniesta, Quintana, Pastor, Espada, Pórtus and Alberola2007; Nasereddin et al. Reference Nasereddin, Salant and Abdeen2008; Sobrinho et al. Reference Sobrinho, Rossi, Vides, Braga, Gomes, De Lima, Perri, Generoso, Langoni, Leutenegger, Biondo, Laurenti and Marcondes2012); hence, the actual number of infected cats may be underestimated (Miró et al. Reference Miró, Rupérez, Checa, Gálvez, Hernández, García, Canorea, Marino and Montoya2014). However, in other studies, clinical signs observed in cats with VL were crusty dermatitis, ulcerative lesions, subcutaneous nodules, weight loss, dehydration and enlarged lymph nodes (Savani et al. Reference Savani, Camargo, Carvalho, Zampieri, Santos, D’áuria, Shawc and Floeter-Winter2004; Coelho et al. Reference Coelho, Lima, Amarante, Langoni, Pereira, Abdelnour and Bresciani2010).
Leishmaniasis in cats coinfected with FIV and FeLV has been reported in the literature (Grevot et al. Reference Grevot, Jaussaud, Marty, Pratlong, Ozon and Haas2005; Duarte et al. Reference Duarte, Arruda, Andrade, Nunes, Souza, Dourado and Costa2010). However, immunosuppression caused by infection by these viruses and the relationship with feline leishmaniasis should be further investigated (Poli et al. Reference Poli, Abramo, Barsotti, Leva, Gramiccia, Ludovisi and Mancianti2002; Grevot et al. Reference Grevot, Jaussaud, Marty, Pratlong, Ozon and Haas2005). In this work, one of the animals with VL tested positive for FIV, and it is suspected that retroviral coinfections increase the susceptibility of cats to leishmaniasis, being the only one to present clinic manifestation (weight loss). The lymph nodes of this animal presented an average of three amastigotes for field, possibly due to immunosuppression caused by viral infection associated with VL.
Unlike dogs affected by VL, which present immune response, particularly humoral, cats seem to have a predominantly cellular immune response to Leishmania infection, a fact that decreases the value of serology for diagnosis (Solano-Gallego et al. Reference Solano-Gallego, Rodriguez-Cortés, Iniesta, Quintana, Pastor, Espada, Pórtus and Alberola2007).
The usefulness of haematological and biochemical laboratory data for the diagnosis of VL in cats is limited, although these parameters are important for the evaluation of the animals’ clinical status and prognosis of the disease (Hagiwara, Reference Hagiwara and Souza2003). According to Solano-Gallego et al. (Reference Solano-Gallego, Koutinas, Miró, Cardoso, Pennisi, Ferrer, Bourdeau, Oliva and Baneth2009), some laboratory findings may be related to canine VL, such as renal azotemia, non-regenerative anaemia, leukocytosis or leukopenia, hyperglobulinaemia, hypoalbuminaemia, low A/G ratio, and changes in liver enzyme activities.
In this study, the cats infected by L. infantum showed normocytic and normochromic anaemia. Anaemia in canine VL may occur by different mechanisms: diminished erythropoiesis, chronicity of the disease, blood loss, lysis of red blood cells and erythrocyte depletion due to the production of autoantibodies, which lead to splenic sequestration (Ikeda et al. Reference Ikeda, Ciarlini, Feitosa, Gonçalves, Luvizotto and Lima2003).
Two animals (I and III) presented hyperproteinaemia with hyperglobulinaemia and hypoalbuminaemia. In canine VL, hyperglobulinaemia may be due to the humoral immune response of polyclonal B lymphocytes, which is evidenced by increased gamma globulin, decreased albumin, and inverted A/G ratio (Abreu-Silva et al. Reference Abreu-Silva, Lima, Macedo, Moraes-Júnior, Dias, Batista, Calabrese, Moraes, Rebêlo and Guerra2008; Trópia de Abreu et al. Reference Trópia De Abreu, Carvalho, Carneiro, Giunchetti, Teixeira-Carvalho, Martins-Filho, Coura-Vital, Corrêa-Oliveira and Reis2011; Nicolato et al. Reference Nicolato, Abreu, Roatt, Aguiar-Soares, Reis, Carvalho, Carneiro, Giunchetti, Bouillet, Lemos, Coura-Vital and Reis2013). Hypergammaglobulinaemia may cause the formation of immune complexes, which are determinants in the pathogenesis of the infection, especially in kidney lesions. In this study, the cats I, II and III presented urea elevation (suggestive of kidney failure). However, the levels of creatinine of these animals were within normal parameters. It is therefore likely that the increase of urea in these animals is due to extra- or pre-renal factors.
It was also observed a change in AST in two cats with positive parasitological examination. Both presented AST below of the normal parameters, but this alteration not shows relation with the LV. One of the causes that favours the reduction of this enzyme in the blood are hormonal problems, however the low concentration of AST presents little clinical importance.
Molecular techniques have proved to be sensitive and powerful tools for detecting Leishmania directly in clinical samples as well as for parasite characterization by PCR. Several scientific papers based on ITS1 analysis have been published on the diagnosis of leishmaniasis and the identification of Leishmania species (Schonian et al. Reference Schonian, Nasereddin, Dinse, Schweynoch, Schallig, Presber and Jaffe2003). Using molecular characterization, this study identified L. infantum in cats in the municipality of Teresina, an area highly endemic for VL (Brasil, 2012).
Finally, the presence of cats infected with L. infantum in the city of Teresina, an area endemic for canine and human VL, indicates a possible participation of these animals in the epidemiological chain of L. infantum in intra- and peridomiciliary environments. Natural infection of this species in endemic areas for VL requires more studies in order to shed light on the possible role of these animals in the transmission of VL cycle.
ACKNOWLEDGEMENTS
The authors thank the clinical veterinarians of the HVU/UFPI, the owners who consented to the collection of biological materials from their animals for this study, and the Federal University of Piauí for its financial support.
FINANCIAL SUPPORT
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.
CONFLICT OF INTEREST
None.
ETHICAL STANDARDS
All procedures contributing to this work comply with the ethical standards on the care and use of animals in experiment. The study was approved by the Ethics Committee for Animal Experimentation – CEEA /UFPI, under Protocol No. 012/12.
Open access
