Introduction

Intestinal helminths and protozoa are common enteroparasites in household and shelter dogs. Many surveys have been conducted worldwide in different canine populations to evaluate the presence of intestinal parasites, because of their veterinary and sanitary importance (Claerebout et al. 2009; Itoh et al. 2009; Joffe et al. 2011; Palmer et al. 2008; Riggio et al. 2013; Rinaldi et al. 2006; Robertson and Thompson 2002; Savilla et al. 2011). Dogs can harbour several helminths, i.e. ascarids (“roundworms”), ancylostomatids (“hookworms”) and tapeworms (e.g. Dipylidium caninum, Echinococcus spp. and Taenia spp.), most of which able to infect also human beings. For instance, Toxocara spp. and Ancylostoma spp. are responsible for different lesions and diseases in humans, being visceral and cutaneous larva migrans the most important (Lee et al. 2010; Traversa 2012). Human cystic and alveolar echinococcoses by Echinococcus granulosus and Echinococcus multilocularis, respectively, are also of great human concern (Deplazes et al. 2011). The zoonotic role of some protozoa affecting dogs, such as Giardia duodenalis and Cryptosporidium spp., is established despite some aspects have to be yet clarified (Bowman and Lucio-Forster 2010). In humans and in dogs, both protozoa may induce different gastrointestinal manifestations, including intermittent/chronic diarrhoea, abdominal pain, nausea, vomiting, anorexia, and weight loss (Ballweber et al. 2010; Lucio-Forster et al. 2010). Of the seven different host assemblages (A–G) in which G. duodenalis is divided (Monis et al. 2003), assemblages A and B are usually isolated from human faeces, and recent studies reported that they can occur frequently in pet dogs, suggesting their potential role as reservoirs in human infections (Monis et al. 2009).

The strictly cohabitation between dogs and humans, sharing the same living areas, increases the risk of infection. For this reason, rescue shelters represent a good observatory for the evaluation of the pathogens that circulate among dog populations. The purpose of the present survey was to determine the occurrence and prevalence of intestinal parasites in dogs living in shelters, with a focus on the characterization of Giardia and Cryptosporidium isolates, together with the assessment of potential risk factors related to canine individual data and anthelmintic treatments.

Materials and methods

Samples and data collection

From November 2008 to June 2012, faecal samples were collected randomly from dogs kept in eight different shelters (S) located in North-eastern Italy, namely seven in Veneto region (S1–S7) and one in Friuli-Venezia Giulia region (S8). Of a total of 318 samples, 268 (84.3 %,) were individual stool samples and 50 (15.7 %) were pooled faeces collected from boxes housing 2–8 dogs. All pooled samples were collected in two shelters, one (S5) located in the province of Padua (38 samples) and one (S7) in the province of Verona (12 samples).

When possible, individual data on breed (crossbred, purebred), gender and age class (<1-year-old, 1–5-years-old, >5-years-old) were collected. Furthermore, information on anthelmintic treatments (yes/no) given within 2 months prior to sampling was recorded in 162 cases. Faecal samples were picked up from the box floor in clean plastic containers, stored at refrigerated conditions (+4 °C) and examined within 2 days. An aliquot of each faecal sample was frozen at −20 °C, pending further molecular analyses, except for 33 samples (i.e. low amount of material, thus not suitable for molecular analysis).

Copromicroscopic analyses

Each faecal sample was macroscopically checked for tapeworm proglottids and roundworms and then analysed (at least 2 g) with a qualitative copromicroscopic technique using a sodium-nitrate solution (specific gravity 1.3) in a double-step sedimentation-floatation procedure (MAFF 1986). Each slide was examined under a light microscope by ×100 and ×400 magnification for helminth eggs and (oo)cysts of protozoa, respectively. Morphometric features were analysed to identify parasitic elements, whose presence was registered for each sample into an Excel 2007 spreadsheet.

Molecular analyses

DNA was extracted from 285 samples using the commercial PSP® Spin Stool DNA Kit (Invitek GmbH, Germany) according to the manufacturer’s instructions. The extracted DNA was screened by a real-time PCR to detect Giardia and Cryptosporidium, separately. Real-time positive samples were submitted to PCR for DNA amplification and sequencing.

Giardia duodenalis

The DNA of G. duodenalis was amplified using a real-time PCR targeting the SSU-rRNA gene with the forward primer Giardia F, the reverse primer Giardia R and the specific double-labeled probe Giardia T (Verweij et al. 2003). The cycling conditions, performed in a LightCycler® Nano (Roche, Germany) and previously described by Verweij et al. (2003), were modified as follows: the final volume of reaction was 10 μl with 5 μl easy-to-use master reagent FastStart Essential DNA Probes Master 2x concentrated (Roche, Germany), 1.3 μl PCR Grade water, 0.5 μl (0.5 μmol) of each specific primer, 0.2 μl (0.2 μmol) double-labeled probe and 2.5 μl of the DNA isolated from the stool samples. PCR amplifications consisted of 10 min at 95 °C followed by 45 cycles at 95 °C for 10 s and 60 °C for 30 s.

Real-time PCR-positive samples were submitted to a two-step nested PCR protocol. The first PCR reaction was performed in T-Personal Thermocycler (Biometra, Germany) in 30 μl volume with the final mix containing 3 μl DNA, 3 μl Buffer 10X, 2 μl dimethyl sulfoxide (DMSO), 1.2 μl (0.4 μM) of each primer RH11 and RH4 (Hopkins et al. 1997), 0.2 μl (1 UI) AmpliTaq Gold® DNA Polymerase (Applied Biosystem®, USA), 0.6 μl deoxynucleotide triphosphates (dNTPs) 10 mM, 0.9 μl MgCl2 25 mM and double-distilled water. Reactions were heated at 94 °C for 11 min and 30 s followed by 35 cycles of 94 °C for 30 s, 65 °C for 30 s and 72 °C for 30 s and 1 cycle of 72 °C for 7 min. The second PCR reaction used the same conditions as previously described using the primers GiarF and GiarR (Read et al. 2002). One microlitre amplicon of the first PCR was used in the reaction mixture of the second step.

Both positive (G. duodenalis DNA) and negative (no template added) controls were included in each PCR reaction. Amplification products were subsequently visualized on 2 % agarose gels with SYBR® Safe DNA gel stain (Invitrogen™, USA).

Cryptosporidium spp.

A touch-down real-time PCR on SYBR® Green I targeting the Cryptosporidium oocyst wall protein (COWP) gene was used with a LightCycler® Nano for the detection of Cryptosporidium spp. with slight variations from the published method (Guy et al. 2003). The amplification reaction was performed in 10 μl of final volume with 5 μl FastStart Essential DNA Green Master 2x concentrated (Roche, Germany), 2.6 μl PCR Grade water, 0.2 μl (0.2 μmol) of each primer (COWP P702-R, COWP P702-F) and 2 μl of DNA extracted from the stool samples. The amplification cycle included an initial step for Taq activation at 95 °C for 10 min, followed by 50 cycles characterized by denaturation (95 °C for 10 s), annealing (65 °C for 10 s) and extension (72 °C for 15 s) steps. The annealing step was conducted under touch-down PCR conditions, i.e. the initial temperature of 65 °C for 10 s was decreased by 0.5 °C s−1 at each cycle during the first ten cycles until the final annealing temperature of 60 °C. The following 40 cycles were carried out under the reached conditions. After the amplification cycle, positive samples were detected using the melting curve analysis. The temperature was increased slowly from 60 to 95 °C for 15 s at a rate of 0.1 °C s−1 with continuous monitoring of fluorescence. The specific melting temperature (Tm) was registered for each amplified sample.

Real-time PCR-positive samples were submitted to a conventional PCR protocol with the following mixture of reaction: 2 μl DNA, 1.2 μl (0.4 μmol) of each primer (COWP P702-R, COWP P702-F), 3 μl Buffer 10X, 1.2 μl (2 mmol) MgCl2, 0.6 μl (0.2 mmol) dNTPs, 1 UI Platinum® Taq DNA Polymerase (Invitrogen™, USA) and double-distilled water in 30 μl of final volume. Cycling conditions were the following: Taq activation step at 95 °C for 5 min, 40 cycles of 95 °C for 15 s, 60 °C for 1 min and 72 °C for 1 min and post-extension step of 72 °C for 7 min. Cryptosporidium amplicons were shown on agarose gel as described for Giardia nested PCR.

In each PCR reaction, positive (Cryptosporidium parvum DNA) and negative (no DNA) samples were added.

Amplicons were sequenced by BMR Genomics (University of Padua, Italy), and then, sequences were aligned using the software ChromasPro version 1.7.5 (Technelysium Pty Ltd, Australia) and compared with those available in the GenBank® database.

Statistical analyses

Differences in prevalence of intestinal parasitism in relation to individual data were evaluated by the Pearson’s chi-squared test (significance level p < 0.05), using SPSS Statistics software, version 22.0.0 (IBM®, New York, USA). Parasites with very low prevalence values were not considered in the statistical analysis.

Results

Data on dogs’ signalment are reported in Table 1. No parasites were detected at the macroscopic examination of stool samples, while more than the half (167/318, 52.5 %) of them were positive for at least one parasite at the microscopic analysis. Among positive samples, 112 (67.1 %) presented one parasite species, 50 (29.9 %) two species and four (2.4 %) three species, and only one was positive for four species. Overall, 106 (63.5 %) samples were positive for helminths, 41 (24.5 %) for protozoa and 20 (12.0 %) for both groups of parasites. Concerning helminths, Trichuris vulpis showed the highest prevalence rate (29.2 %); among protozoa, Giardia and Cystoisospora (oo)cysts were observed in the 15.1 and 5.7 % of the examined samples, respectively (Table 2).

Table 1 Allocation of individual faecal samples according to dogs’ signalment data
Full size table
Table 2 Copromicroscopic prevalence [n (%)] of helminth and protozoan parasites assessed by copromicroscopic technique in single and pooled faecal samples
Full size table

The overall prevalence for protozoa and/or helminth parasites varied significantly (χ 2 = 48.097, p < 0.001) among shelters, from a minimum of 17.6 % to a maximum of 82.7 % (Table 3). A significant relation between prevalence values and provenance was confirmed for the most common helminths, i.e. T. vulpis (χ 2 = 65.757, p < 0.001), Toxocara canis (χ 2 = 33.863, p < 0.001) and ancylostomatids (χ 2 = 77.981, p < 0.001). Such a significant relationship was also observed for Giardia prevalence values obtained by both copromicroscopy (χ 2 = 61.728, p < 0.001) and nested PCR (χ 2 = 22.690, p < 0.01), as reported in Table 3. With regard to the individual faecal samples, no significant correlations (p > 0.05) were found between prevalence of the most common parasites and history data as age, sex and breed. Anyhow, T. canis was most frequently detected in animals ageing less than 1 year, while the prevalence of T. vulpis gradually increased from younger dogs until to almost double in older ones as shown in Fig. 1. The administration of an anthelmintic within 2 months prior to sampling significantly influenced the occurrence of T. vulpis and ancylostomatids (Table 4).

Table 3 Prevalence [n (%)] of helminth (a) and protozoan (b) parasites in each monitored shelter and available data on treated/untreated dogs with anthelmintic drugs
Full size table
Fig. 1
figure 1

Prevalence of T. canis and T. vulpis according to the age classes

Full size image
Table 4 Copromicroscopic results [n (%)] for Trichuris vulpis and ancylostomatids in treated/untreated dogs
Full size table

The presence of Giardia was revealed by real-time PCR in 165/285 (57.9 %) samples. Out of 165 positive samples, 106 were confirmed by nested PCR, and among them, 79 amplicons were successfully sequenced as on the follows: 78 host-specific genotypes, i.e. 49 assemblages C and 29 assemblages D (detected both in all shelters except S4, where only assemblage D was found) and one assemblage B1 (S8).

Real-time PCR on SYBR® Green I revealed the presence of Cryptosporidium spp. in 3/285 (1.1 %) stool samples, two collected in S3 and one in S8. All positive samples were confirmed by PCR, and amplicons were sequenced as C. parvum (GenBank®, accession number JQ349359).

Discussion

Several studies were conducted all over the world to investigate parasite prevalence in dogs kept in shelters, and most of them reported helminth prevalence over 30 % (Blagburn et al. 2008; Capelli et al. 2006; Claerebout et al. 2009; Ortuño and Castellà 2011; Ortuño et al. 2014; Palmer et al. 2008; Turkowicz and Cielecka 2002). The prevalence recorded in the present survey (52.5 %) fits with data previously reported in Italy (Capelli et al. 2006). These high prevalence values in shelter and kenneled dogs are likely related to the great exposure to parasitic infections and, in particular, to infective elements. The density of individuals in a restricted area, especially in poor management conditions, may lead to high environmental contaminations and to an increased risk of infection, especially for giardiosis and geo-helminthoses (Capelli et al. 2003; Leonhard et al. 2007; Ortuño and Castellà 2011). In the present study, the results showed that single infections were more frequent than mixed infections and that helminths were more prevalent than protozoa, in agreement with the results recorded in previous investigations (Ramírez-Barrios et al. 2004; Riggio et al. 2013). Furthermore, T. vulpis was confirmed as the most prevalent helminth in kenneled dogs, followed by roundworms and hookworms. The higher prevalence rates of T. vulpis and T. canis in comparison with other parasites are probably due to the high resistance of their eggs in the environment for long time and to the biological features of ascarid larval stages in transplacental and transmammary transmission (Traversa 2012).

No cestode elements were found, except for D. caninum that was isolated only in one case. In general, low prevalence values are probably underestimated because a single copromicroscopic examination has a very low sensitivity to detect tapeworms, due to the inconstant elimination of proglottids and the undistributed eggs in the faeces.

Among protozoan infections, the overall prevalence of giardiosis by copromicroscopic examinations (15.1 %) was in accordance with other studies (Capelli et al. 2006; Neves et al. 2014; Nikolic et al. 2008; Palmer et al. 2008). After trichurosis, giardiosis was the second most prevalent intestinal infection in the examined dog population, reaching the high value of 57.9 % when investigated with molecular methods, as previously described in Italy (Paoletti et al. 2008; Scaramozzino et al. 2009) and elsewhere (Bowman and Lucio-Forster 2010). The lower copromicroscopic prevalence detected in this survey indicates that a single copromicroscopic analysis may be insufficient to diagnose G. duodenalis infection (Capelli et al. 2006; Epe et al. 2010; Thompson 2004).

According to previous results (Giangaspero et al. 2007; Paoletti et al. 2008; Scaramozzino et al. 2009), all the isolates detected in the present study but one were the dog-specific assemblages C and D, confirming that host-specific genotypes are much more prevalent than other assemblages in kennels, where there is a high frequency of dog-to-dog transmission (Ballweber et al. 2010; Leonhard et al. 2007; Thompson 2004). Zoonotic and host-specific assemblages of G. duodenalis isolated from dog faeces are well documented in literature (Ballweber et al. 2010; Berrilli et al. 2004; Bowman and Lucio-Forster 2010; Hunter and Thompson 2005; Lalle et al. 2005; Monis et al. 2003; Scaramozzino et al. 2009). In fact, the frequency of infections with host-specific assemblages is more common in dogs living in kennels than in household dogs, where zoonotic assemblages are more frequent (Claerebout et al. 2009; Covacin et al. 2011; Leonhard et al. 2007; Scaramozzino et al. 2009; Uehlinger et al. 2013; Upjohn et al. 2010).

The low prevalence rate (1.1 %) of cryptosporidiosis detected in this study is in accordance with values previously found in Central and Southern Italy (Giangaspero et al. 2007; Rinaldi et al. 2008) and worldwide (Claerebout et al. 2009; Uehlinger et al. 2013; Yoshiuchi et al. 2010). The Cryptosporidium isolates in this survey were typed as C. parvum, and none of the three positive samples was identified as a host-specific genotype as reported by other authors worldwide (Giangaspero et al. 2006; Sotiriadou et al. 2013; Uehlinger et al. 2013; Wang et al. 2012; Yoshiuchi et al. 2010). As for Giardia, also, zoonotic Cryptosporidium species are well documented in humans, including C. parvum (Cama et al. 2003; Lucio-Forster et al. 2010; Morgan et al. 2000; Pedraza-Díaz et al. 2001; Pieniazek et al. 1999; Xiao et al. 2001). These results suggest that the risk of transmission to humans for both protozoa is less significant than previously thought, but, on the other hand, these infections are often underestimated and the isolation of dog-specific genotypes from human faeces is not uncommon (Bowman and Lucio-Forster 2010; Soliman et al. 2011).

The statistical analysis showed that the shelter of provenance was one of the main risk factors for helminth infections in the here examined dogs. This result suggests a close relation between occurrence of parasites and shelter management, in particular with the accomplishment of effective control programmes including: (a) hygiene measures, such as environmental cleaning and disinfections; (b) appropriate diagnostic methods in housed animals (including new arrivals) and, as a consequence, (c) appropriate use of parasiticides. In the present study, the adoption of anthelmintic treatments within 2 months before sampling led into a significant reduction (p < 0.001) of helminthoses. Indeed, a higher prevalence was found for T. vulpis (71.2 %) and ancylostomatids (38.5 %) in S3 and for T. canis (27.7 %) in S2, where no or few sampled dogs (0/52 and 21/47, respectively) received an anthelmintic drug within the previous 2 months, suggesting that the management had a low standard for controlling parasites in those sites. The animal density is another key factor (Dubná et al. 2007; Leonhard et al. 2007; Meireles et al. 2008; Ortuño and Castellà 2011), especially in areas for out-of-box time, that may have contributed to the increase of infection risks in these shelters. Also, giardiosis resulted to be significantly influenced by the provenance of dogs. Although these protozoa were detected in all investigated shelters, copromicroscopic and nested PCR prevalence varied significantly from 0 to 33.3 % and from 20.0 to 61.5 %, respectively. With this regard, it is worthy of note that false-negatives may influence faecal examinations and that molecules routinely used in shelters may not have an antiprotozoal activity. Despite some parasiticides (e.g. fenbendazole) are active against both helminths and protozoa, they are frequently administered in a single dose under a parasite control scheme. This approach does not have efficacy for giardiosis, as, for instance, fenbendazole should be given for three consecutive days (ESCAAP 2011). Therefore, incorrect treatments may increase Giardia prevalence and/or exacerbate the occurrence of Giardia superinfections in a shelter.

The prevalence of intestinal parasites in puppies is generally considered higher than in adults, due to their immature immune-system (Gates and Nolan 2009), whereas parasite-specific immunity is thought to be acquired with age, as a consequence of single or repeated exposures (Ramírez-Barrios et al. 2004). Nevertheless, in this study, the age of animals was not statistically related to the presence of parasites, even though T. canis was more prevalent in younger animals, while T. vulpis showed an increasing trend until to be mainly prevalent in older ones. These data are in agreement with those reported in previous studies describing a higher prevalence of roundworms in dogs younger than 12 months of age and whipworms in adult dogs (Fontanarrosa et al. 2006; Gates and Nolan 2009; Riggio et al. 2013). The high occurrence of trichuroid infections in adult dogs is likely due to an absence of a transplacental and/or transmammary transmission and to the inability to elicit a protective immune response (Traversa 2011). Moreover, the high prevalence of whipworm infections detected in this study confirms the resistance of trichuroid eggs in the environment and supports the importance to use anthelmintics which are effective against both larval and/or adult stages of T. vulpis.

In conclusion, this study confirms that intestinal parasites are very common pathogens in dog communities and highlights the need to improve the quality of shelter management in terms of cleaning and disinfection programme, control programmes combining accurate diagnostic methods and therapeutic approaches. Although the close-contact with dogs is not considered as a relevant risk in the transmission of intestinal nematodes, the zoonotic risk due to their infective elements in the environment should always be kept in mind (Traversa 2012). Conversely, Giardia and Cryptosporidium (oo)cysts are infective immediately after they are excreted and this may represent a higher risk of infection by a close-contact between dogs and human beings. For this reason, shelter dogs should be considered a source of infection for shelter workers, towards better health managements, for safe pet-adoption campaigns and a minimization of the environmental faecal pollution with canine intestinal parasites (Traversa et al. 2014).