Frequency of Giardia duodenalis infection and its genetic variability in dogs in Cuiabá, Midwest Brazil

Introduction: Giardia duodenalis, a unicellular, eukaryotic, and flagellated protozoan, presents two evolutionary forms in its life cycle, namely, trophozoites and cysts. It causes diarrhea in humans, dogs, cats, rodents, and ungulates. Despite being morphologically similar, the isolates of G. duodenalis are genetically diverse, affecting the stability and unanimity of taxonomic classification. Since different Giardia assemblages may occur within one isolate, multilocus genotyping is recommended for the genetic identification. Methodology: To determine the frequency of G. duodenalis infections in domiciled dogs in Cuiabá Municipality (State of Mato Grosso, Midwestern Brazil) and characterize its genetic variability, fecal samples were collected from 147 dogs. Results: Overall, 6.8% (10/147) of the samples presented cysts of G. duodenalis, which sequencing and genotypic characterization using tpi and gluD revealed assemblages C and A, genetic grouping of G. duodenalis. Only three samples amplified by tpi and one sample amplified by gluD. Conclusions: The risk factors age, gender, breed, diet and the presence of other dogs in the same house were not correlationated with giardiasis. The host-specific and zoonotic genotype warns of the risk of inter and intraspecies transmission and it provides, for the first time, information about genetic characterization of G. duodenalis isolates in dogs in Cuiabá, Midwest region of Brazil.


Introduction
Giardia duodenalis (syns. Giardia intestinalis and Giardia lamblia) is one of the most important waterborne protozoan pathogens in both developed and developing countries, causing diarrhea in many different species around the world [1]. It is genotypically divided into various assemblages (A-H) of G. duodenalis which include the sub genotypes AI, AII, BIII, BIV [2]. Assemblages C and D are most frequently found in dogs, but also can sporadically infect humans, whereas assemblages A and B are commonly reported in humans [3].
G. duodenalis infection occurs via the ingestion of its cysts in water and contaminated food, as well as via the fecal-oral route. In the stomach of the host, cysts rupture because of gastric acidity, releasing trophozoites that replicate in the intestine and cause watery diarrhea, abdominal pain, nausea, and vomiting. However, the attachment of their ventral disk to the epithelium of the intestine is responsible for pathophysiological mechanisms [4]. Most infections are self-limiting [3], depending on the characteristics of the parasite (strain, number of cysts ingested, ability to adhere and cause tissue damage, and virulence factors) and host (immune and nutritional status, gastric juice pH, and intestinal microbiota) [5].
Although G. duodenalis infection is global, there is a large difference in its prevalence between countries. In Brazil, its prevalence in canines ranges from 2% to 70% [1]. However, in the Midwest region of Brazil, there are only a few studies on the prevalence of giardiasis, especially in the State of Mato Grosso, except for reports of 72,7% of prevalence in children and 0,04% in vegetables [6,7].
In the present study, we evaluated the frequency of G. duodenalis infection in dogs in Cuiabá Municipality, Midwest Brazil. In addition, this study aimed to genotyping cysts of G. duodenalis from dogs, using tpi and gluD genes.

Study design
Fecal samples of dogs were collected at the Veterinary Hospital of Federal University of Mato Grosso, in the Cuiabá Municipality (15°35′56″S, 56°06′01″W) between January 2016 to December 2017. The owners of the 147 dogs from which the fecal samples were obtained answered a questionnaire based on closed-ended questions, with at least two questions to determine the profile of dogs and their households [age (≤ 1 year, puppy; > 1 year, adults), sex, breed, signs of diarrhea, other dogs in the house and diet].

Parasitological analysis
The samples were collected after defecation, stored in a plastic container for a maximum of 24 h under 4ºC refrigeration, and part of each sample was concentrated using zinc sulfate flotation technique with the final sediment being examined by just one trained veterinary medic using an light microscope to view of cysts [12].

Molecular tests
DNA was extracted from feces aliquots using the NucleoSpin Tissue Kit (Macherey-Nagel ® , Düren, Germany). Briefly, the fecal samples with visible G. duodenalis cysts suspended in a sodium chloride solution were washed four times with Tris-EDTA solution (pH = 8.0), then subjected to thermal shock cycles to rupture the cysts [13]. The manufacturer's protocol was then followed to extract DNA.
For G. duodenalis genotyping, nested and seminested PCR amplification was performed as previously described [13,14] on 530 and 659 base pair (bp) fragments of tpi and gluD, respectively. Nucleasefree water and DNA from G. duodenalis obtained from naturally infected domestic dogs were used as the negative and positive controls, respectively. Primers used for PCR are presented in Table 1. The amplified products were analyzed by electrophoresis on a 1.5% agarose gel and visualized using the ChemiDoc ® (Bio Rad, Berkeley, California) transilluminator. After purification using the GFX TM PCR DNA Purification Kit and Gel Band Kit (GE Healthcare, Chicago, Illinois), the samples were sequenced using the ABI-PRISM 3500 Genetic Analyzer (Applied Biosystems, Foster City, California) and subsequently analyzed using the BLAST NCBI program (http://www.ncbi.nlm.nih.gov/blast/Blast.cgi).
The Neighbor-Joining method was used to build the phylogenetic tree with the concatenated tpi and gluD gene sequences after alignment with the Muscle method using MEGA program (Version 10.1.7).

Statistical Analysis
The frequency of infection was determined using the ratio of samples from dogs with cysts from the examined dogs. The association between sex, breed, age, diet, other animals in the house, and G. duodenalis infection was verified using the non-parametric Chisquare or Fischer's Exact test, if the cell values were less than five. These tests were performed in Epi Info ® (CDC, EUA), with the significance level set to 5%.

Ethical statement
The study was approved by the Ethics Committee of the Federal University of Mato Grosso (CEUA-UFMT) with the approval number 23108.170944/2016-16.

Gene tpi
All isolates from dogs (CUIABA20, CUIABA164 and CUIABA165) were characterized as Assemblage C presenting homology with the LC437553.1 and MN270282.1 sequences stored at GenBank® witch were isolated from dogs.

Gene gluD
The isolated CUIABA165 was identified as assemblage A presenting 100% homology with sequences stored at GenBank® with accession numbers LC507405.1 witch was isolated from human.

Phylogenetic analysis
In order to determine the genetic relationship of the isolates, phylogenetic analysis was performed using concatenated data in the respective sequences of the tpi and gluD genes ( Figure 2). The sequences of isolates deposited on GenBank with the specified assemblages were used for the construction of the phylogenetic tree  1 and KF993735.1). The isolate CUIABA165, from this study, remained in the same clade as G. duodenalis assemblage C.

Discussion
The frequency of G. duodenalis infection observed in dogs in this study was 6.8%, which is lower than that reported in Campo Grande, Mato Grosso do Sul (MS) (27.3%) [15] and higher than that reported in Goiânia, Goiás (GO) (1.6%), cities in the Midwest region of Brazil [16].
All dogs evaluated in this study originated from private households, unlike those in MS, which included dogs from a higher proportion of kennels, as well as, zoonoses control centers, which may have contributed to increased prevalence [15]. In GO, the protozoan was found only in domiciled dogs, and this can be attributed to the low prevalence in this group [16]. Therefore, despite the fact that the three cities are located in the Midwest region, other factors contribute to these different rates of prevalence, such as the type of environment in which the animals live and the presence of a population cluster [1].
Age, sex, and breed of infected dogs were not considered risk factors for infection [15,17]. However, some studies claim that puppies are more predisposed [18] due to immunosuppression and their habit of chewing objects [1,11]. In addition, male dogs have been reported to be at a higher risk because of activity in a greater territorial area [19].
The clinical sign of diarrhea was observed in positive and negative dogs, and did not increase the chance of finding G. duodenalis parasitism. This can occur because of the intermittent elimination of cysts [4] or because the variability of clinical manifestations depends on the hosts' immunocompetence [18]. In addition, most dogs show subclinical infection [20].
Regarding the non-association of food as a risk factor, it can be inferred that because they are domiciled dogs, the chance of consuming food contaminated with cysts is negligible when compared with stray dogs that live freely [17].
The presence of other dogs in the same household was not a significant factor for the occurrence of giardiasis in this study, which is consistent with previously reported findings [21], who reported that single dogs and the dogs living in the presence of more than one dog had a similar prevalence. Similarly, other research found a significant statistical difference of infection rates between shelter dogs and household pets [22]. This is probably due to the greater concentration of animals in a shelter and the greater ambient contamination to which they are exposed.
In Brazil, multilocus genotyping to characterize assemblages of G. duodenalis has been conducted only in the southern and southeastern regions with dogs, cats, sheep and cattle [23]. However, this study is the first to characterize G. duodenalis genotypes in dogs in the Midwest.
Concatenated analysis of the sequences showed that the tpi gene showed greater specificity when classifying the assemblage of Giardia species. This fact can be explained by mixed infection or by heterozygosis present in the sequence of gluD gene [24][25][26]. It is estimated that different loci of the same isolate can cause a discrepancy in the identification of assemblages in 15% of the cases [27]. This discrepancy has important implications for molecular studies that use only one gene to classify the assemblages of isolates [25,26].
Assemblage C identified in the sampled dogs has been described in several countries [28] including Brazil. Studies in São Paulo had a prevalence of 25.8% [29] and in Minas Gerais 18.7% [30]. Although this Figure 2. Phylogenetic relationship of G. duodenalis isolate characterized by the sequencing of tpi and gluD genes. The phylogenetic tree was built using concatenated data from tpi and gluD genes by Neighbor-Joining analysis. Bootstrap values above 50% are shown in the figure. Assemblage A is represented by the color yellow, assemblage B by the color green, assemblage C by the color red, assemblage D by the color blue, assemblage E by the color pink and assemblage F by the color purple. genotype is common in dogs, cats, coyotes, and wolves, it has been reported in human samples as well [31].
Another assemblage commonly found in dogs is type D [17,29]. Although assemblages C and D occur more commonly in dogs, assemblages A and B, which are common in humans, have been detected in dog feces in different continents, including South America [3]. Besides that, infection by assemblages C and D occurs as a result of species-specific transmission [32,33].
Moreover, the low rate of prevalence could be explained by the intermittent elimination of cysts and the low rate of DNA amplification can be explained by the small number of cysts in the sample, the loss of the cysts during their recovery, the small amount of DNA, the presence of fecal inhibitors in the samples, the small volume of the sample or the loss from washing the sample [13,30].
Two assemblages of G. duodenalis, assemblage A and C, were identified in the fecal sample of one of the examined dogs, however, when constructing the concatenated tree the sample showed greater proximity with assemblage C, these discrepancies between the loci of the same isolate may occur due to mixed infections or heterozygosis of the species [30].
In this study, G. duodenalis was most prevalent, followed by Ancylostoma spp. Similarly, other authors founded a high prevalence of G. duodenalis in the City of Medellín, Colombia [34] where dogs and cats had rates of Giardia infection higher than Ancylostoma spp.

Conclusions
The frequency of G. duodenalis infection in dogs was 6.8%, and age, gender, breed, diet and the presence of other dogs in the same house were not risk factors for the occurrence of giardiasis from several countries. Furthermore, the observation of assemblages A and C of G. duodenalis warns of the risk of inter and intraspecies transmission in the studied environment, especially for the zoonotic risk. It is suggested that further studies be carried out with other dogs and other mammals so that it is possible to characterize the genotypes that occur in several species of animals in the state of Mato Grosso, Midwest, Brazil.