Subtype Characterization and Zoonotic Potential of Cryptosporidium felis in Cats in Guangdong and Shanghai, China

Cryptosporidiumfelis is an important cause of feline and human cryptosporidiosis. However, the transmission of this pathogen between humans and cats remains controversial, partially due to a lack of genetic characterization of isolates from cats. The present study was conducted to examine the genetic diversity of C. felis in cats in China and to assess their potential zoonotic transmission. A newly developed subtyping tool based on a sequence analysis of the 60-kDa glycoprotein (gp60) gene was employed to identify the subtypes of 30 cat-derived C. felis isolates from Guangdong and Shanghai. Altogether, 20 C. felis isolates were successfully subtyped. The results of the sequence alignment showed a high genetic diversity, with 13 novel subtypes and 2 known subtypes of the XIXa subtype family being identified. The known subtypes were previously detected in humans, while some of the subtypes formed well-supported subclusters with human-derived subtypes from other countries in a phylogenetic analysis of the gp60 sequences. The results of this study confirmed the high genetic diversity of the XIXa subtype family of C. felis. The common occurrence of this subtype family in both humans and cats suggests that there could be cross-species transmission of C. felis.


Introduction
Cryptosporidium spp. are important apicomplexan parasites inhabiting the gastrointestinal tract of humans and other vertebrates, causing severe diarrhea [1]. Human cryptosporidiosis has been associated with over 20 Cryptosporidium species, but C. hominis, C. parvum, C. meleagridis, C. felis, and C. canis are the most common ones [2]. Among them, C. felis mainly infects cats and is therefore considered a host-adapted species [3]. Human C. felis infections, however, are common in developing countries [4][5][6][7], and at least one possible zoonotic transmission of C. felis between a household cat and the owner has been reported [8]. Nevertheless, the limited number of reports of zoonotic infections with this species has raised questions on the importance of zoonotic transmission in the epidemiology of human C. felis infections [9].
A gp60 subtyping tool has been developed recently for genetic characterizations of C. felis [17]. Thus far, nearly 200 C. felis isolates have been examined, which has led to the

Nucleotide Sequence Variations in the gp60 Gene of C. felis
A comparison of the nucleotide sequences generated led to the identification of 15 sequence types. They differed from each other by both nucleotide substitutions and copy numbers of repetitive sequences. The sequence alignments showed the presence of numerous single nucleotide substitutions (SNPs) over the partial gp60 gene among the 20 C. felis isolates. Three types of simple tandem repeats were detected in the gp60 genes ( Table 2). Among them, 2-5 copies of the 33-bp repeat sequence 5 -CCACCTAGTGGCGGTAGTGGCGTGTCCCCTGCT-3 and 2-4 copies of the 39-bp repeat sequence 5 -CCACCTAGTGGCGGTAGTGGCGTGTCCCCTGCT-3 Pathogens 2021, 10, 89 3 of 8 were observed at nucleotides 460-577 and 778-911 of the sequence alignment, respectively. In both repeat types, the last copy had only half of the usual length. In addition, 5 copies of the trinucleotide repeat GTT were found in the gp60 sequences at nucleotides 1143-1154.
from Guangdong and Shanghai for assessment of their zoonotic potential.

Amplification of the gp60 Gene
Among the 30 DNA preparations that were positive for C. felis based on nested PCR analysis of the small subunit (SSU) rRNA gene, 20 (66.7%) generated the expected products in the gp60 PCR. Among them, 9 were from cats in pet shelters, 5 were from cats visiting animal hospitals, 4 were from cats in pet shops, and 2 were from stray cats ( Table  1). These PCR products differed slightly in size ( Figure 1) and were all sequenced successfully.
IDs labelled with ECUST were from cats in Shanghai, while those with SCAU were from cats in Guangdong; b 33-bp tandem repeat (5 -CCACCTAGTGGCGGTAGTGGCGTGTCCCCTGCT-3 ) with a partial copy at the end; c 39-bp tandem repeat (5 -AGCACAACTGCGGCTACAGCGAGCACTGCGAGTTCGACA-3 ) with a partial copy at the end and 0-2 nucleotide differences.

Cryptosporidium felis Subtypes Identified
Altogether, 15 subtypes were identified, with two subtypes in samples SCAU392 and SCAU754 having nucleotide sequences identical to the XIXa-39 and XIXa-40 subtypes (GenBank reference sequences MH240852 and MH240853) from humans in the United Kingdom, respectively (Table 1). According to the nomenclature system of the XIXa subtypes [18], the 13 new subtypes were named XIXa-81 (3)

Phylogenetic Analysis of C. felis Subtypes
The gp60 sequences from the 20 C. felis isolates were all placed in the XIXa subtype family in the phylogenetic tree ( Figure 2). They formed several subclusters with strong bootstrap support with C. felis subtypes from humans. Among the novel subtypes identified in the study, 4 subtypes (XIXa-90, XIXa-91, XIXa-92, and XIXa-93) formed a subcluster with several subtypes identified from humans and cats in several European countries and another subtype (XIXa-82) formed a subcluster with several subtypes from humans and cats in the United Kingdom and Sweden. In addition, two cat-derived subtypes (XIXa-87 and XIXa-88) in 4 isolates from this study formed their own subcluster. As expected, the known subtypes (XIXa-39 and XIXa-40) identified in the present study and previously in humans from the United Kingdom and Indonesia formed a subcluster (Figure 2).

Discussion
The gp60 subtyping tools have been widely used in assessing the intra-species diversity and zoonotic transmission of human-pathogenic Cryptosporidium spp. Several genes including gp60, other genetic loci with simple tandem repeats, and double-stranded viral RNA have been used to further differentiate the Cryptosporidium species into different

Discussion
The gp60 subtyping tools have been widely used in assessing the intra-species diversity and zoonotic transmission of human-pathogenic Cryptosporidium spp. Several genes including gp60, other genetic loci with simple tandem repeats, and double-stranded viral RNA have been used to further differentiate the Cryptosporidium species into different subtypes [10]. Among them, the gp60 gene is highly polymorphic in all Cryptosporidium spp. examined thus far and therefore can be categorized into multiple subtype families by nucleotide sequence differences [4]. Moreover, the gp60 is an invasion-related protein and the subtype families identified have been linked to differences in host ranges and with virulence in C. parvum and C. hominis [3,23,24]. In this study, the newly developed gp60 subtyping tool was employed to characterize the C. felis and to understand its zoonotic potential in China.
This represents the first subtyping study of C. felis in China. The previous two studies on C. felis subtypes mostly examined C. felis isolates from other countries, with five subtype families (XIXa, XIXb, XIXc, XIXd, and XIXe) being identified [17,18]. In the present study, 20 cat-derived isolates from China were subtyped and the results of the phylogenetic analysis showed that all of them belonged to the subtype family XIXa. The PCR amplification efficiency (66.7%) of this study was similar to that of a previous study (67.0%) [18]. The light infections of C. felis in healthy cats could be one of the reasons for the relatively low PCR amplification efficiency.
An analysis of the gp60 gene confirmed the high genetic diversity of C. felis isolates. Like observations in previous studies [17,18], 15 subtypes were seen in 20 C. felis isolates successfully analyzed from cats. However, all C. felis isolates in this study belonged to the subtype family XIXa. As 13 of the 15 XIXa subtypes identified were novel, there could be geographic isolation among some of the XIXa subtypes, as previously suggested by us [18].
The results of the present study suggest that cross-species transmission of C. felis could be possible. In this study, among the 15 XIXa subtypes, two were identical to subtypes previously found in humans, while others clustered with human-derived subtypes from other countries. Importantly, one of the subclusters formed included two sequences from an owner and the household cat (GenBank reference sequences MH240883 and MH240884) with possible zoonotic transmission [17]. Moreover, we previously subtyped two C. felis isolates from one child in Shanghai (XIXa-14) and one rhesus macaque (XIXa-12) in Guizhou, China [18], while the gp60 sequences were different with the 15 XIXa subtypes identified in the present study. They clustered together in a large clade within the XIXa subtype family in the phylogenetic tree ( Figure 2). The subtype characteristics of C. felis identified in this suggest that the zoonotic XIXa subtype family could be the dominant subtypes in pet cats in China. As the sequences from C. felis in Guangzhou and Shanghai are dispersed throughout the large clade within the XIXa subtype family, they could be good representatives of C. felis in cats in China. Since C. felis infections have been reported in cats and humans in several locations in China [19,[25][26][27][28], an analysis of the gp60 sequences of human-and cat-derived C. felis isolates from the same location is needed to understand the epidemiological importance of zoonotic transmission of C. felis.
Currently, the significance of the divergent C. felis subtypes in cats is not clear. As mentioned above, different C. parvum and C. hominis subtypes have been linked to differences in virulence. Therefore, the hyper-transmissible and virulent C. hominis IbA10G2 and C. parvum IIaA15G2R1 subtypes have caused numerous outbreaks of human cryptosporidiosis worldwide [10] and infections with the former have induced more clinical symptoms than other C. hominis subtypes [23]. In C. felis, although previous studies and results of the present study suggest that the XIXa is the dominant subtype family in humans and cats [18], the relationship between the pathogenicity and its high transmission is not yet clear. More subtyping studies at additional genetic loci are needed to understand the differences in pathogenicity among C. felis subtypes.

Ethics Statement
This research was reviewed and approved by the Ethics Committee of the South China Agricultural University. The fecal specimens were collected with the permission of the owners of the pets. Each sample was collected and placed into a 50-mL plastic centrifuge tube containing 2.5% potassium dichromate and was transferred to the laboratory for storage at 4 • C. The DNA extraction was completed within one week. The DNA was stored at −20 • C for less than three years before being used in the PCR analysis in the present study.

C. felis Isolates
The nested PCR targeting the small subunit (SSU) rRNA gene was used to detect Cryptosporidium spp. [29]. DNA preparations of 30 C. felis-positive fecal samples from China were used in the present study. Among them, 18 samples were obtained from cats in Guangdong province, including 7 from pet shelters, 6 from animal hospitals, 4 from stray cats, and 1 from a pet shop. The remaining 12 samples were obtained from cats in Shanghai, including 8 from pet shelters and 4 from pet shops. They were from two previous studies of molecular epidemiology of cryptosporidiosis in cats [27,28].

PCR Amplification
The newly developed nested PCR targeting the conserved region of the gp60 gene was employed to identify the subtypes of C. felis in this study [17]. Briefly, primers GP60-Felis-F1 (5 -TTT CCG TTA TTG TTG CAG TTG CA-3 ) and GP60-Felis-R1 (5 -ATC GGA ATC CCA CCA TCG AAC-3 ) were used in primary PCR, while GP60-Felis-F2 (5 -GGG CGT TCT GAA GGA TGT AA-3 ) and GP60-Felis-R2 (5 -CGG TGG TCT CCT CAG TCT TC-3 ) were used in secondary PCR. The sizes of primary and secondary PCR products were approximately 1200 and 900 bp, respectively. The PCR reaction and cycling program were described recently [18]. Each DNA preparation was analyzed in duplicate, with the inclusion of both positive (C. felis DNA) and negative (reagent-grade water) controls in each PCR run. The positive products from the secondary PCR were identified by 1.5% agarose electrophoresis.

DNA Sequence Analyses
All positive products of the expected size were sequenced bidirectionally on an ABI3730 autosequencer by the Sangon Biotech (Shanghai, China) using secondary PCR primers. The DNA sequences generated were assembled using ChromasPro 2.1.6 (http://technelysium. com.au/wp/) and edited using BioEdit 7.1.3.0 (http://www.mbio.ncsu.edu/BioEdit/bioedit. html). These and the reference sequences from GenBank were aligned with each other using the MUSCLE program implemented in MEGA 6 (https://www.megasoftware.net/). Tandem Repeats Finder 4.09 (http://tandem.bu.edu/trf/trf.html) was used to identify repetitive sequences within them. A maximum-likelihood tree was constructed using MEGA 6 based on substitution rates calculated using the general time reversible model and gamma distribution. The bootstrap method with 1000 replicates was used to assess the reliability of the phylogenetic clusters formed. Representative nucleotide sequences of the C. felis subtypes identified in the present study were deposited in the GenBank database under accession numbers MW351820-MW351832.

Conclusions
The present study reported the subtype characteristics of C. felis isolates from cats in China for the first time. The results of the phylogenetic analysis suggested the potential zoonotic transmission of this pathogen. More isolates from diverse areas and hosts should be analyzed to confirm this conclusion.