**3. 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 [19,25–28], 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.
