Detection and molecular characterization of Cryptosporidium spp . in captive canaries ( Serinus canaria ) using different diagnostic methods

This study used several diagnostic methods to examine the occurrence of and molecularly characterize Cryptosporidium spp. in captive canaries (Serinus canaria) in southern and southeastern Brazil. A total of 498 fecal samples were purified by centrifugal-flotation using Sheather’s solution. Cryptosporidium spp. diagnosis was performed using three diagnostic methods: malachite green negative staining, nested PCR targeting the 18S rRNA gene, followed by sequencing the amplified fragments, and duplex real-time PCR targeting the 18S rRNA specific to detect Cryptosporidium galli and Cryptosporidium avian genotype III. The overall positivity for Cryptosporidium spp. (total samples positive in at least one protocol) from the microscopic analysis, nested PCR and duplex real-time PCR protocol results was 13.3% (66/498). The positivity rates were 2.0% (10/498) and 4.6% (23/498) for Cryptosporidium spp. by microscopy and nested PCR, respectively. Sequencing of 20 samples amplified by nested PCR identified C. galli (3.0%; 15/498), Cryptosporidium avian genotype I (0.8%; 4/498) and Cryptosporidium avium (0.2%; 1/498). Duplex real-time PCR revealed a positivity of 7.8% (39/498) for C. galli and 2.4% (12/498) for avian genotype III. Malachite green negative staining differed significantly from nested PCR in detecting Cryptosporidium spp. Duplex real-time PCR was more sensitive than nested PCR/sequencing for detecting gastric Cryptosporidium in canaries.

Parasite-host adaptation and co-evolution among Cryptosporidium spp.and their avian hosts are thought to occur (XIAO et al., 2002) since some Cryptosporidium species/genotypes are found almost exclusively in certain avian orders (NAKAMURA & MEIRELES, 2015).Therefore, epidemiological studies using samples representing avian orders or species would aid in investigating Cryptosporidium species evolution.
The common techniques used to diagnose Cryptosporidium infection are microscopic analysis and nested PCR.Microscopy is an inexpensive and fast technique; however, it does not identify the Cryptosporidium species and is less sensitive and specific.Nested PCR is more expensive than microscopy, despite being highly sensitive and specific and identifying the species after amplicon sequencing (JEX et al., 2008).Another option for specific species/genotype diagnosis of gastric cryptosporidiosis in birds is through duplex real-time PCR (NAKAMURA et al., 2014).
The aim of the present study was to determine the prevalence of Cryptosporidium spp. in fecal samples from captive canaries from southern and southeastern Brazil.Additionally, molecular characterization of positive samples was performed, and three techniques for detecting Cryptosporidium spp.were compared.

Materials and Methods
This study was approved by the Animal Use Ethics Committee (CEUA) of the São Paulo State University (UNESP), School of Veterinary Medicine, Araçatuba, process FOA 01022-2015.

Fecal samples
Fecal samples were obtained from asymptomatic captive canaries housed in 102 aviaries from six southern and southeastern Brazilian states, exhibited at the 64th Ornithological Championship 2015 of the Ornithological Federation of Brazil (FOB), from 09/07/2015 to 19/07/2015, in the city of Itatiba, state of São Paulo, Brazil.
The total population of canaries exhibited in the championship was approximately 40,000 birds.To determine the prevalence of Cryptosporidium spp. in these canaries, the sample number of 385 was calculated using Win Episcope software (THRUSFIELD et al., 2001) considering a diagnostic test with 100% sensitivity and 100% specificity.As lower sensitivity and specificity rates are common when using the diagnostic protocols employed in this study (JEX et al., 2008), and because losses occur in sample storage and processing, 498 samples were collected.
Samples were collected from the bottom of the cage at the time of the bird's reception at the championship to avoid cross-contamination.Each sample was collected using a disposable wooden spatula, transferred to a 2-ml microtube containing 0.9% sodium chloride solution in enough quantity to prevent dehydration, and stored at 4°C.

Purification of oocysts
Samples were fragmented and homogenized using a disposable wooden spatula in a 2 mL microtube containing Sheather's solution (g = 2.05) prepared with phosphate buffered saline and 0.1% Tween 20 (PBS-T).The contents of each microtube were homogenized, and half of the contents were transferred to another microtube so that each sample was purified in two tubes simultaneously.Each microtube was filled with 1.9 mL of Sheather's solution and vortexed and centrifuged at 800 g for 5 minutes.Four hundred μL of supernatant was transferred to a microtube containing 1,500 μL of PBS-T, homogenized by inversion and centrifuged at 10,000 g for 3 minutes.After discarding the supernatant, the microtubes were filled with 1.9 mL of PBS with 0.01% Tween 20 and centrifuged at 10,000 g for 3 minutes.The supernatant was discarded, preserving approximately 100 μL of solution and sediment.One hundred μL of 10% buffered formalin was added to one microtube for Cryptosporidium oocyst microscopy screening using malachite green negative staining (ELLIOT et al., 1999), while the other microtube was frozen at -20°C for DNA extraction and amplification by nested PCR and duplex real-time PCR.

Genomic DNA extraction
Genomic DNA of Cryptosporidium spp. was extracted per the protocol adapted from McLauchlin et al. (2000) and Wang et al. (2011) using Zymo-Spin  IIIC silica columns (Zymo Research, Irvine, USA) to replace activated silica.The DNA was eluted in 100 μL of elution buffer (10 mM Tris, 0.5 mM EDTA, pH 9) and stored at −20°C in two 50 μL aliquots.
The reactions contained a volume of 25 μL, with 2.5 μL of 10x PCR buffer, 2.0 mM MgCl 2 , 0.5 U of JumpStart  Taq DNA Polymerase (Sigma-Aldrich, St. Louis, USA), 200 μM of each deoxyribonucleotide, 200 nM of each primer, 5 μL of target DNA in the PCR and 2.5 μL of DNA in the nested PCR.Samples were subjected to initial DNA denaturation at 94°C for 2 minutes, followed by 40 cycles, each consisting of denaturation at 94°C for 30 seconds, annealing at 58°C and extension at 72°C for 60 seconds (PCR) or 45 seconds (nested PCR), with a final extension at 72°C for 7 minutes.Amplified fragments were analyzed by GelRed  (Biotium, Fremont, USA) stained gel electrophoresis.

Duplex real-time PCR
Duplex real-time PCR was performed to simultaneously detect C. galli and Cryptosporidium avian genotype III (NAKAMURA et al., 2014), amplifying 134-bp and 138-bp amplicons, respectively (Table 1), under the following reaction conditions: 25 μL of solution containing 12.5 μL of JumpStart  TaqReady Mix (Sigma-Aldrich, St. Louis, USA), 4.5 mM MgCl 2 , 250 nM of each probe, 600 nM of each primer, 0.6 μg/μL of non-acetylated bovine serum albumin (Sigma-Aldrich, St. Louis, USA), and 5 μL of target DNA.PCR cycles consisted of 2 minutes of denaturation at 94°C followed by 50 cycles of 30 seconds at 94°C and 1 minute at 61°C in the CFX96  Real-Time PCR Detection System (Bio-Rad, Hercules, USA).

DNA sequencing
Identifying the Cryptosporidium species was performed by sequencing the nested PCR amplicons, after purification using the Illustra ExoProStar  1-Step (GE Healthcare Life Sciences, Champaign, USA) or the QIAquick  Gel Extraction kit (Qiagen, Hilden, Germany), using ABI Prism  Dye Terminator 3.1, in an ABI 3730XL automatic sequencer (Applied Biosystems, Foster City, USA), at the Center for Sequencing and Functional Genomics of UNESP, Jaboticabal Campus, Brazil.Sequencing reactions were performed in both directions using nested PCR primers.
Consensus sequences were determined using the CodonCode Aligner v. 7.1.2(CodonCode Corporation, Dedham, USA) and aligned with homologous sequences published in GenBank using Clustal X software (THOMPSON et al., 1997) and Bioedit Sequence Alignment Editor (HALL, 1999).
Nucleotide sequences generated in this study were submitted to the GenBank database under the accession numbers MG832881-MG832883.

Statistical analysis
The McNemar test was used to compare diagnostic techniques, and the Kappa correlation coefficient test was used to evaluate the agreement between them.Statistical analyses were performed using the BioEstat 5.0 software (Analyst Soft Inc., Walnut, USA) and the results were considered significant when p<0.05.
Duplex real-time PCR was positive for gastric Cryptosporidium in 10.2% (51/498) of the samples, 7.8% (39/498) for C. galli and 2.4% (12/498) for Cryptosporidium avian genotype III.For detecting the gastric species of Cryptosporidium, nested PCR/sequencing and duplex real-time PCR were significantly different (p<0.0001), and the agreement between the two methods was fair (Kappa = 0.40).
C. galli is associated with chronic infection in the passerine proventriculus and is likely responsible for chronic gastric disease and predisposition to concomitant infections (ANTUNES et al., 2008;NAKAMURA & MEIRELES, 2015).The species most frequently detected by nested PCR/sequencing was C. galli, corresponding to 75% (15/20) of the sequenced samples.In addition, Cryptosporidium species identification by nested PCR/sequencing and duplex real-time PCR revealed results similar to those of other authors (NAKAMURA et al., 2014;NARDI, 2015), in which  2015).Herein, C. galli was detected more frequently in canaries than Cryptosporidium avian genotype I.In contrast, Nardi (2015) found a 14.2% (29/204) positivity for Cryptosporidium avian genotype I in fecal samples and cloacal swabs from canaries.This higher positivity rate may have resulted from cloacal swab sampling and by including samples from symptomatic and dead birds.Although the tissue tropism of avian genotype I is undetermined, the close genetic similarity among this genotype, C. avium, C. baileyi and avian genotype II is evidence that avian genotype I exhibits tropism for the cloacal epithelia, the bursa of Fabricius or the respiratory tract (NAKAMURA & MEIRELES, 2015).
In conclusion, nested PCR was more sensitive than microscopic analysis using malachite green negative staining to detect Cryptosporidium spp. in fecal samples from canaries.Duplex real-time PCR was more sensitive than nested PCR/sequencing for diagnosing gastric cryptosporidiosis in canaries.There was a higher prevalence of C. galli, and Cryptosporidium avian genotype III, Cryptosporidium avian genotype I and C. avium were detected at lower positivity rates.

Table 1 .
Primers and probes used to detect C. galli and Cryptosporidium avian genotype III by duplex real-time PCR.Molecular characterization of Cryptosporidium spp. in captive canaries

Table 3 .
Detection and identification of Cryptosporidium species and genotypes in canary fecal samples using microscopy and molecular methods targeting the 18S rRNA gene.

Table 2 .
Results of positivity for Cryptosporidium spp., in at least one protocol, from canary fecal samples from southern and southeastern Brazil.

nº positive/n° sampled) % positive (nº positive/ nº sampled) Species identification % positive (nº positive/nº sampled)
Chelladurai et al. (2016)sitivity than Cryptosporidium avian genotype III in Passeriformes, including canaries.Cryptosporidium avian genotype VI, described byChelladurai et al. (2016), is closely related to C. galli and Cryptosporidium avian genotype III and likely infects the gastric epithelia.Although we did not detect avian genotype VI in canary samples by nested PCR/sequencing, we cannot assure that duplex real-time PCR does not detect this genotype due to the high similarities in its primer and probe annealing regions.Tissue tropism and the clinical importance of Cryptosporidium avian genotype I are undetermined (NAKAMURA & MEIRELES,