First identification of Cryptosporidium parvum virus 1 (CSpV1) in various subtypes of Cryptosporidium parvum from diarrheic calves, lambs and goat kids from France

Cryptosporidium spp. remain a major cause of waterborne diarrhea and illness in developing countries and represent a significant burden to farmers worldwide. Cryptosporidium parvum virus 1 (CSpV1), of the genus Cryspovirus, was first reported to be present in the cytoplasm of C. parvum in 1997. Full-length genome sequences have been obtained from C. parvum from Iowa (Iowa), Kansas (KSU) and China. We aimed at characterizing the genome of CSpV1 from France and used sequence analysis from Cryptosporidium isolates to explore whether CSpV1 genome diversity varies over time, with geographical sampling location, C. parvum genetic diversity, or ruminant host species. A total of 123 fecal samples of cattle, sheep and goats were collected from 17 different French departments (57 diseased animal fecal samples and 66 healthy animal fecal samples). Subtyping analysis of the C. parvum isolates revealed the presence of two zoonotic subtype families IIa and IId. Sequence analysis of CSpV1 revealed that all CSpV1 from France, regardless of the subtype of C. parvum (IIaA15G2R1, IIaA17G2R1 and IIdA18G1R1) are more closely related to CSpV1 from Turkey, and cluster on a distinct branch from CSpV1 collected from C. parvum subtype IIaA15G2R1 from Asia and North America. We also found that samples collected on a given year or successive years in a given location are more likely to host the same subtype of C. parvum and the same CSpV1 strain. Yet, there is no distinct clustering of CSpV1 per French department or ruminants, probably due to trade, and transmission of C. parvum among host species. Our results point towards (i) a close association between CSpV1 movement and C. parvum movement, (ii) recent migrations of C. parvum among distantly located departments and (iii) incidental transmission of C. parvum between ruminants. All together, these results provide insightful information regarding CSpV1 evolution and suggest the virus might be used as an epidemiological tracer for C. parvum. Future studies need to investigate CSpV1’s role in C. parvum virulence and on subtype ability to infect different species. Supplementary Information The online version contains supplementary material available at 10.1186/s13567-023-01196-4.


Introduction
Species of the genus Cryptosporidium are protozoan parasites specific of vertebrates that are responsible for mild to severe diarrhea in both human and animals [1,2].In ruminants, diarrhea is particularly severe in newborn and young.It causes dehydration, weakness, anorexia, weight-loss and a delayed growth.Adults can remain asymptomatic and serve as reservoir host for the parasite [3].Thus, Cryptosporidiosis represents a significant economic loss for farmers.Cryptosporidium is also a public health issue because contamination of surface water can lead to human infections.Cryptosporidium is among the most important cause of diarrhea in developing countries, and remain a major cause of waterborne outbreaks and illness worldwide in both immunocompetent and immunocompromised people.In healthy immunocompetent people, cryptosporidiosis usually resolves itself within a couple of weeks [4][5][6][7].However, epidemic outbreaks are still associated with fatalities, mostly in young children, and a chronic and life-threatening cryptosporidiosis can develop in immunocompromised individuals, such as those affected by HIV, transplant recipients or infants [7].
The genus Cryptosporidium is composed of 44 recognized species, and 120 genotypes remain of unknown species status [8].They represent different public health significance depending on their host specificity and zoonotic potential.Humans are mainly infected by C. parvum and C. hominis.C. parvum is by far the Cryptosporidium species with the broadest host range and most important zoonotic potential.Nearly 20 C. parvum subtypes have been described using the 60 kDa glycoprotein gene (gp60) as genetic locus.The subtype family IIc appears to have adapted to humans whereas the subtype family IIa is mostly found in cattle and subtype family IId is mostly found in sheep and goats [9,10].The highly transmissible subtype IIaA15G2R1 is the most prevalent in most industrialized nations [10][11][12].In France, this highly transmissible subtype has been reported in feces from cattle, goat and sheep [13][14][15].However, C. parvum has a panmictic/epidemic population structure resulting in a high genetic diversity, with frequent disagreement between gp60 results and other genetic markers due to frequent genetic recombination within this subtype [11].Multilocus typing tools have allowed a better characterization of C. parvum subtype IIaA15G2R1, revealing a geographic segregation with country-specific clusters and a genetic distance that correlates with geographic distance [16,17].Different subpopulations have also been found in different host species due to reduced gene flow [18][19][20].The global expansion of the IIaA15G2R1 zoonotic subtype presents a major challenge.In particular, there is a need to develop novel, easy to use methods to track C. parvum's genetic structure, host-adaptation, virulence and threat to public health.
The C. parvum virus 1, of the genus Cryspovirus, family Partitiviridae, was first reported in C. parvum cytoplasm in 1997 [21,22].The viral genome is bisegmented and dsRNA1 and dsRNA2 each encode a single protein.Closely related strains of the cryspovirus were found in distinct C. parvum strains infecting humans and calves, as well as in C. hominis, C. felis and C. meleagridis, with > 92% amino-acid similarities [23,24].A single study has assessed the cryspovirus prevalence in parasite populations, in Japan [25].It showed that CSpV1 has 100% apparent prevalence and appears to co-diversify with C. parvum so that viral sequences might be used for sensitive detection of C. parvum subtype IIaA15G2R1 and to determine the regional origin of the parasite [25][26][27].
The aims of the present study were the following: (1) to determine by PCR (18S rRNA, PCR-RFLP, gp60) the presence of Cryptosporidium in calves, lambs and kid goats in different regions of France, (2) to characterize the presence and prevalence of cryspovirus in these French Cryptosporidium isolates, (3) to genotype the virus in order to test whether CSpV1 genome diversity varies over time, with geographical sampling location, with C. parvum genetic diversity, or with ruminant host species.

Sample collection
A total of 123 fecal samples were collected voluntarily by veterinarians or departmental veterinarian laboratories, between 2018 and 2022, and originated from 17 different French departments (Table 1).Among those, 57 samples were collected from diseased animals that had been naturally infected by Cryptosporidium and presented a diarrhea at the time of sampling.An additional 66 fecal samples were collected from apparently healthy animals.Samples were collected with sterile gloves, mailed at room temperature, and conserved at 4 °C until use.

Cryptosporidium detection and enrichment
Cryptospridium spp.were initially detected microscopically with direct immunofluorescence assay (DFA) (MeriFluor ® Cryptosporidium/Giardia, Meridian Bioscience Europe, Milano, Italy) as previously described [14].Cryptosporidium oocysts were purified from samples with positive DFA using Dynabeads ™ anti-Cryptosporidium kit per manufacturer protocol.The success of Cryptosporidium purification was controlled using DFA.

DNA extraction, Cryptosporidium species determination and C. parvum subtyping
Disruption of oocyst walls was achieved using ten freeze-thaw cycles as previously described [28].Then, DNA extraction was conducted using the QIAamp DNA Stool Mini Kit (Qiagen, France), according to the manufacturer's instructions.The Cryptosporidium genus was determined using a nested PCR amplifying a 840 bp fragment of 18S rRNA gene [29].PCR products were digested with SspI and MboII endonucleases and restriction fragment length polymorphism (RFLP) analysis was used to determine the Cryptosporidium species [30].C. parvum samples were subtyped by nested PCR-sequence analysis of the partial 60 kDa glycoprotein locus (gp60) [31] and Sanger sequencing of both strands (Genoscreen) (Genbank numbers in Table 2).C. parvum subtypes were named using the recommended nomenclature system [10,32].

Full-length genome sequencing
CSpV1 genome was obtained from a pool of samples B44 and B45, using RNA sequencing.cDNA libraries were prepared using an Ion total RNAseq Kit (Life technologies, Carlsbad, CA, USA) according to the supplier's instructions.The cDNA libraries were sequenced using   Sequence reads were cleaned and trimmed for adapter removal using fastP version 0.20.1 [34] and sequence quality was verified using FastQC version 0.11.8 [35].Reads were assembled with rnaSPADES de novo assembler as implemented in SPAdes assembler version 3.10.0[36].Resulting contigs were aligned on local nt database with Megablast version 2.10.1 to identify viral references.Then, both sequence reads and assembled contigs were aligned using Burroughs-Wheeler Aligner (BWA, version 0.7.8) [37] against the CSpV1 strain Iowa genome fragments (NC_038843 dsRNA1 and NC_038844 dsRNA2) and visualized in Integrative Genome Viewer (IGV) [38,39] to control the quality of the consensus sequence extracted using Samtools pileup version 1.8 [40].

Phylogenetic analyses
The CSpV1 dsRNA1 and dsRNA2 sequences were aligned against all other CSpV1 sequences known to date.Nucleotide and protein percentage of identity were calculated with Clustal Omega [41].Phylogenetic analysis were conducted on the full-length open reading frame.Phylogenetic tree were then inferred using the maximum likelihood method implemented in PhyML (version 3.0) [42] using the best-fit model and best of NNI and Subtree Pruning and Regrafting (SPR) branch swapping.Support for nodes on the trees were assessed using an approximate likelihood ratio test (aLRT) with the Shimodaira-Hasegawa-like procedure.Trees generated using the Neighbor-Joining and Maximum Parsimony methods gave identical results.

Results
None of the healthy goats, sheep and cattle were infected by Cryptosporidium.Among the 57 diseased animals presenting diarrhea, 55 were infected with Cryptosporidium and one individual was infected with Giardia.18S PCR and sequencing revealed the presence of C. parvum only.This was in agreement with previous studies in France showing that C. parvum is responsible for most infections in pre-weaned calves, lambs and goats [45].In France, the two most prevalent C. parvum belong to subtype families IIa and IId [14].For one cattle sample with low parasite load, the 18S gene could not be sequenced.For all other samples, C. parvum subtyping revealed that cattle were infected by subtypes IIaA15G2R1 (18 ind; 90%) and IIaA17G1R1 (2 ind, 10%), sheep were infected by subtypes IIaA15G2R1 (22 ind; 73%), IIaA16G3R1 (2 ind; 7%), IIaA17G2R1 (1 ind, 3%) and IIdA18G1R1 (5 ind; 17%), and all 4 goat kids were infected by IIaA15G2R1 (Figure 1, Table 2).Thus, IIaA15G2R1 C. parvum subtype was the most prevalent, and was found in 11 departments on all four successive years (2018 to 2022).IIaA17G1R1 and IIaA17G2R1 were found in one department in 2020 and 2021, while IIaA16G3R1 and IIdA18G1R1 were found only once, in one department.
The full-length genome of CSpV1 from France was obtained from a pool of C. parvum of subtype IIaA15G2R1 collected from cattle in 2021 (Samples B44 and B45; Genbank acc numbers OQ686777 and OQ686778 for dsRNA1 and dsRNA2 respectively).The representative dsRNA1 and dsRNA2 FR strain genome was obtained with a coverage depth of 50 and 70 reads per position and with 0.6 and 0.8% iupac codes for dsRNA1 and dsRNA2, respectively.The genome sequence was compared to the three available full-length Cryspovirus genomes from Iowa (Iowa), Kansas (KSU) and China.The FR strain dsRNA1 shows a maximum of 97.3% nucleotide (nc) identity to the Iowa strain and a minimum of 96.4% nc identity to the KSU strain from North America.The RdRP protein shows 99% and 97% amino-acid (aa) * Spl: Sample.Letter (B, bovine / cattle; C, Caprine / goats and O, ovine / sheep); Dpt: Department; Cryspovirus PCR was carried out using primers RdRPV and CPV from Jenkins et al. [33] primers VL and VS from Murakoshi et al. [25].GP60 C. parvum subtyping and VL/VS Cryspovirus amplicon were followed with Sanger sequencing.Genbank accession numbers are provided.identity to the Iowa and KSU strains, respectively.The FR strain dsRNA2 shows a maximum of 98.4% nc identity to the Iowa strain and a minimum of 98% nc identity to KSU strain and was more distantly related to the related viruses discovered in Cryptosporidium meleagridis (85.6% nc identity), Cryptosporidium felis (86.7% nc identity), and Cryptosporidium hominis (90.8% nc identity).The capsid protein shows 98.6% and 99.1% aa identity to the Iowa and KSU strains, respectively.Using two different sets of primers per virus genome fragment, we confirmed the presence of CSpV1 within all but two isolates of C. parvum (Table 2), including in the sample for which the C. parvum subtype could not be determined.This high prevalence was in accordance with previous reports [21,25].Good quality partial sequences of CSpV1 virus dsRNA1 encoding the RdRP, and dsRNA2 encoding the Capsid proteins were obtained for 30 and 42 samples, respectively (Genbank accession numbers are provided in Table 2).CSpV1 partial genome sequencing provided evidence of co-infection by two different genotypes of CSpV1 in C. parvum isolate O77, depicted by the presence of two overlapping peaks on chromatograms.Interestingly, point variations observed co-localize with nucleotide positions that show variations when comparing all sequences with each other, further supporting the co-infection hypothesis (Additional file 1).A greater genetic diversity was observed within dsRNA1 that shows an average of 95.6% nc identity to CSpV1 strain Iowa whereas dsRNA2 presents an average of 98% nc identity to CSpV1 strain Iowa.We conducted phylogenetic analyses on the nucleotide sequences of dsRNA1 and dsRNA2 and yielded similar results.As previously described, all sequences from Japan clustered closely together on a single branch.Similarly, sequences from France clustered closely together and are most closely related to the only other CSpV1 partial virus sequence from the European continent (origin: Turkey) (Figures 2  and 3).

Discussion
The parasite population structure and rate of CSpV1 vertical and horizontal transmission is expected to influence virus evolution.Here, phylogenetic analyses conducted on both genome fragments (dsRNA1 and dsRNA2) show that CSpV1 from France clustered together with the sequence of CSpV1 from Turkey, and on a branch distinct from CSpV1 collected from North America or Asia (Figures 2 and 3).This result indicates that the CSpV1 dsRNA1 and dsRNA2 can distinguish sampling area at the continental level.Furthermore, we provide herein the first characterization of CSpV1 from different subtypes of C. parvum.Indeed, studies conducted on CSpV1 from North America and Japan focused on C. parvum subtype IIaA15G2R1 [25,46], while the subtype of C. parvum  between gp60 subtypes of C. parvum in the same animal.Host co-infection by different strains of C. parvum could also explain why we found some evidence of co-infection by different CSpV1 in isolate O77.We did not conduct clonal isolation of C. parvum, which means that if the host was co-infected by different C. parvum strains, they were present in the mixture that was analyzed.Therefore, the working hypothesis that each C. parvum strain carries only a single CSpV1 remains to be tested through more advanced analyses of CSpV1 genotype and C. parvum genotyping.
It has been suggested that CSpV1 evolution can be used to track movement of C. parvum [25].Thus, we compared the phylogenetic position of CSpV1 dsRNA1 and dsRNA2 with the ruminant host, sampling year, and geographic locations.Our results indicate that samples collected on a given year and given location are more likely to host the same subtype of C. parvum and the same CSpV1 strain (Figures 1 and 2).In addition, when sampling occurred on successive years in the same location, we found the same subtype of C. parvum and the same CSpV1 strain (Figure 2).Yet, there is no distinct clustering of viruses per department or ruminant host: closely related CSpV1 were found in distant departments (i.e.viruses of IIaA15G2R1 from cattle of departments 63, 03, 57 and 58, Figures 1 and 2); also, closely related CSpV1 are often found in different host species (Figures 2 and 3 In conclusion, we provide the first genomic data of CSpV1 in France.We also provide the first comparative analysis of the genome of CSpV1 from different C. parvum subtypes, from different ruminant hosts and over successive years.This analysis provides insightful information regarding both C. parvum and CSpV1 transmission and evolution.In agreement with Murakoshi et al. [25], our results suggest that CSpV1 is vertically transmitted in C. parvum, and that transmission between subtypes likely occurs as a result of recombination during the parasite's sexual life-cycle.While gp60 gene coding is the most common marker used for epidemiological surveillance of C. parvum, the limitations associated with this approach are recognized and novel techniques such as whole genome sequencing and multilocus genotyping are being investigated as alternative approaches [49,57].Herein, we identified CSpV1 in all but two C. parvum isolates, which can either suggest the absence of the virus within these two C. parvum-positive samples, or illustrate that our sensitivity was suboptimal.Given that CSpV1 was often identified using a set of primers and not others suggest that neither of our primers provided sufficient sensitivity.Moreover, we failed at obtaining good quality CSpV1 sequences from many samples, which limited the scope of our analyses.In the future, amplicon-based sequencing approaches are likely to provide better sensitivity, while allowing for the sequencing of multiple variants of CSpV1 within the same sample.CSpV1 genome sequencing may represent a low-cost alternative approach to track C. parvum epidemiology and will help identify co-infections, and characterize the rates of outcrossing and inter-species transmission.Coupling amplicon-based sequencing of CSpV1 with multilocus genotyping of C. parvum would most likely provide a more accurate description of the parasite and virus co-diversification.Whether the virus' ability to jump host between C. parvum belonging to different subclades can modulate the protozoan parasite's ability to infect different ruminants, its pathogenicity, and zoonotic potential remains to be investigated through experimental approaches.

Figure 1 C
Figure 1 C. parvum subtyping in France between 2019 and 2021.Gp60 based subtyping of C. parvum was conducted on isolates from 11 different French departments collected from Cattle, Sheep and Goats.A pie chart is provided for each department (dpt) within which C. parvum was detected, and provide the C. parvum subtype and the host.The size of the pie charts is proportional to the number of samples collected.

Figure 3
Figure 3Tracking CSpV1 movements and evolution through phylogenetic analysis of dsRNA2.Phylogenetic tree of a 867 nc long fragment of dsRNA2 of CSpV1 using all available CSpV1 sequences.The tree was inferred in PhyML using the LG substitution model.Branch points indicate that the results of Shimodaira-Hasgawa branch test > 0.8.Scale bar shows the number of nucleotide changes.Genbank virus isolate names are given as follows: accession number_C.parvum subtype_strain name.French virus isolates names are provided as follows: Letter (B: bovine / cattle, C, Caprine / goats and O; ovine / sheep) followed with individual number, department number, C. parvum subtype, and sampling year.A map of France providing the localisation of departments (dpt) is provided next to the tree.
).Our results point towards (i) a close association between CSpV1movement and C. parvum movement, (ii) recent migrations of C. parvum among distantly located departments and (iii) incidental transmission of C. parvum between ruminants.Inter-species local transmission of Cryptosporidium spp. is characteristic of the parasite epidemiology [54-56].Thus, the presence of C. parvum in many animal species, inter-species transmission, movements of animals due to trade among different regions of the country, and the potential role of wild animals in parasite dispersion [56], implies that the sources of C. parvum and CSpV1 movements are multiple.