Cryptosporidium infections in asymptomatic calves up to 4 months in Poland: a cross-sectional population study

Cattle cryptosporidiosis is noted worldwide with varied frequency of infection prevalence depending on geographical, environmental and husbandry factors. In this study, the prevalence of Cryptosporidium infections in cattle was determined on the basis of molecular results obtained by testing 1601 faecal samples collected from calves up to 4 months of age housed in all Polish provinces from 2014 to 2018. Detection and identification of Cryptosporidium species was performed at the 18 small subunit ribosomal RNA (18S rRNA) locus by conducting PCR–RFLP analysis of the amplified DNA fragments. The prevalence of Cryptosporidium infections in the cattle population was 45.3% (CI 95%: 42.8–47.7; 725/1601). The infected animals were housed on 233/267 (87.3%) of monitored farms with regional prevalence ranging from 27.8 to 62%. The restriction pattern of 18S rRNA amplicons for positive samples was characteristic of C. parvum, C. bovis, C. ryanae, C. andersoni, and unexpectedly also of C. baileyi and C. suis. Infections of C. bovis and C. ryanae prevailed in the studied cattle population relegating C. parvum to third in prevalence. Likewise, mixed infections caused by C. bovis and C. ryanae as well as C. parvum and C. bovis were observed. A relationship between the infecting parasite species and animal breed was found. For instance, C. parvum prevailed in Black and White lowland breed, C. ryanae in Limousine cattle and C. andersoni in dairy animals of mixed dairy breeds. Furthermore, differences in prevalence of particular parasite species between cattle breeds were also shown.

Cryptosporidium infections in cattle were recognised for the first time in 1971 in the United States of America in an 8-month-old diarrhoeic female calf.Histopathology of the small intestine revealed atrophy of the villi and the presence of various developmental forms of Cryptosporidium in the epithelium 1 .Since then, Cryptosporidium infections and cryptosporidiosis have been reported in cattle worldwide 2 .The significant role of cattle as a source of Cryptosporidium for humans was recognised during a waterborne epidemic of human cryptosporidiosis in Milwaukee, USA 3 .Likewise in Poland, studies on occurrences of Cryptosporidium in bovine host have been conducted since the 1970s, usually being confined to animals bred on farms from selected Polish regions [4][5][6] and using methods not always conducive to identification of the parasite species 4,7,8 .However, an attempt to assess the prevalence of infections nationwide by individual Cryptosporidium species in Polish cattle of different ages has also been undertaken 9 .As a pilot study, it covered a broad group of animals at the age from 1 day to 6 years of different health status.However, due to the low number of samples obtained from particular Polish provinces, the results did not allow for a regional assessment of the infection prevalence as well as to gather unbiased data on the occurrence of Cryptosporidium species in the population.It has been shown that infections caused by different Cryptosporidium species mainly affect animals up to the age of 4 months with Cryptosporidium bovis (C.bovis) as predominating species.
A wider application of molecular tools in veterinary parasitology has facilitated identification of Cryptosporidium species in livestock.As with other similar studies conducted in Europe, in Poland the infections previously diagnosed in cattle were mostly caused by Cryptosporidium parvum (C.parvum) 8,10,11 .The tested animals were also positive for C. bovis, Cryptosporidium andersoni (C.andersoni) and Cryptosporidium ryanae (C.ryanae) 9,11 .An age-related pattern of Cryptosporidium species infecting cattle was solely observed for C. parvum.C. andersoni was the only species occurring in adult cattle over one year of age 9 .Species which were unusual for the bovine host, such as Cryptosporidium felis (C.felis), were only occasionally detected in cattle populations in Poland 12 .
In an effort to gather data on the prevalence of Cryptosporidium infections in cattle population and in particular Polish provinces, only healthy animals up to the age of 4 months were sampled.Apart from a molecular assessment of the infection prevalence, the study aimed to identify parasite species and their regional distribution pattern in cattle of different breeds and ages.

Ethics approval and consent to participate
This study did not require approval of an Ethics Committee.However, freshly voided faeces were collected during routine veterinary practice in adherence to international guidelines for animal care.Sample collection was not harmful and did not violate animal welfare laws.No clinical interventions were performed.The owners of the cattle included in this study provided informed consent to participate.

Cattle faeces
Over a period of 5 years, 1601 freshly voided faecal samples were collected from cattle from the age of 1 week to 4 months from 2014 to 2018 (Table 1).Each cattle was sampled only once and the sampling plan covered each year different provinces.The animals were housed on 267 farms located across the 16 administrative provinces of Poland.The average number of animals in the herd was 137.The cattle were mainly raised on large farms (175, 65.5%) with > 50 heads.Small farms (< 50 heads) accounted for 92 (34.5%) of the tested establishments.In each year from 306 to 324 faeces were subjected for testing.Animals originated from farms located in districts with the highest cattle populations in 3 or 4 provinces each year 13 .In each province 18 farms were monitored except from Dolnośląskie (DS) (15 farms) and Opolskie (OP), Śląskie (SL) and Podkarpackie (PK) where 12 farms were sampled in each localization.Farms were randomly selected and represented different administrative locations in the province.Generally, from each farm freshly voided faeces of 6 animals (18 per district) of different breeds at the age between 1 week to 4 months were taken.However, there were also farms enrolled on www.nature.com/scientificreports/which all sampled animals were the same age and breed.For instance, on 9 farms, there were only 4-week-old calves and on 15 farms (5 with calves of each age) there was stock at the age of 8, 12, and 16 weeks.The cattle were in good health without symptoms of Cryptosporidium infections.During this study no clinical intervention or animal examination were conducted by vets taking care of these farms who collected samples.The animals were divided into three age groups as shown in Table 1.Faecal samples of 10-15 g were placed individually into plastic containers, labelled and sent to the laboratory.Faeces were collected from cattle representing dairy breeds (Polish Black and White Holstein-Friesian (HO), Jersey (JE), Polish Red and White Holstein-Friesian (RW), Brown Swiss (BS), and dairy animals of mixed dairy breeds (MS)); meat breads (meat cattle of mixed meat breeds (MM), Aberdeen Angus (AN), Charolaise (CH), Salers (SL), Limousine (LM), and Belgian Blue (BB)); and cattle of mixed dairy-meat breeds (mixed dairy-meat breeds (MDM), Black and White lowland (NCB), Polish Red (RP), Montbéliarde (MO), Polish Black and White (ZB) and Simmental (SM)).The sampled animals mainly represented dairy breeds (74.8%), followed by meat (14.5%) and mixed (10.7%)breeds.

Detection and identification of Cryptosporidium species based on the analysis of the 18S rRNA gene fragment
Cryptosporidium DNA was extracted from 0.1 g (100 µl) of the faeces with an alkali wash and a heat lysis method developed by Millar et al. 14 with further modifications 9 .Nucleic acid extracts were purified using a GeneMATRIX PCR/DNA Clean-Up Purification Kit (EURx Ltd., Gdańsk, Poland) according to the manufacturer's instructions.The extracts containing parasite DNA were stored at − 20 °C until use.Detection of Cryptosporidium DNA was performed by a nested-PCR method allowing amplification of the 18S rRNA Cryptosporidium gene fragment using the primers specified by Xiao et al. 15 and conditions described elsewhere 16 .To increase amplification efficiency and to reduce amplification inhibition each reaction mixture was supplemented with 20 µg of bovine serum albumin (Thermo Fisher Scientific, Vilnius, Lithuania) 9 .The correct performance of the nucleic acid extraction and the amplification step were verified by inclusion of an appropriate set of controls as described elsewhere 9 .All reactions were performed in a Biometra TProfessional BASIC thermocycler (Analytik Jena, Jena, Germany).
The identity of Cryptosporidium species in positive faeces was defined by a PCR-RFLP analysis of the amplified 18S rRNA fragments.The analysis used the following enzymes: NdeI for identification of C. parvum 17 , MboII in a parallel digestion for C. bovis and Cryptosporidium ryanae (C.ryanae) detection 18 , XbaI for differentiation of C. andersoni and C. parvum by their lack of a restriction site from C. bovis and C. ryanae 9 , BcuI (SpeI) in digestion for identification of Cryptosporidium suis (C.suis) 17 and SspI for detection of Cryptosporidium baileyi (C.baileyi) 19 .To confirm the correct identification of Cryptosporidium species which were unusual for cattle such as C. suis and C. baileyi, sequencing of the post-PCR 18S rRNA amplicons was conducted.The PCR products excised from the agarose gel and purified were directly sequenced in both directions using the ABI Prism BigDye Terminator v3.1 Cycle Sequencing Kit on an ABI 3730XL DNA sequencer (Life Technologies, Carlsbad, CA, USA) at the Genomed S.A. sequencing service (Warsaw, Poland).Nucleotide sequences were aligned with published sequences from GenBank by using the NCBI-BLAST program (http:// blast.ncbi.nlm.nih.gov/ Blast.

Geographical distribution of Cryptosporidium infections
The infected animals were reared on 233 (87.3%) out of 267 monitored farms located across all 16 Polish provinces.In the cases of the Zachodniopomorskie (ZP) and DS provinces, all monitored farms appeared Cryptosporidium positive.Cryptosporidium infections were detected with varied prevalences ranging from 27.8% in Małopolskie (MP) to 62% in Podlaskie (PL) province (Fig. 1).However, greater than 40% prevalence of infections was revealed in the majority of the monitored regions.

Age related prevalence of Cryptosporidium spp.
Overall Cryptosporidium infection prevalence for the studied cattle population was 45  3).However, statistically significant differences were solely observed between animals at the age of 4-8 and 8-16 weeks (χ 2 = 4.19; p = 0.041).Six different species were detected being C. bovis the most common in the studied cattle population (χ 2 = 189.13,p < 0.001) (Table 3
The provinces at the opposite extreme were SL for C. parvum and LS, ZP and MZ for C. andersoni, where these parasites were not detected.Isolations of C. suis and C. baileyi from cattle were achieved in PL and WP and in LS and SL respectively (Table 4).Mixed infections of C. bovis and C. ryanae with a 2.8% frequency were observed in WP and below 1.8% in LS, ZP, SL, OP and DS.
Cryptosporidium bovis and Cryptosporidium ryanae

Discussion
Molecular methods have become irreplaceable in the detection and identification of Cryptosporidium species.PCR-based methods are gradually replacing microscopic methods in epidemiological studies on Cryptosporidium prevalence in livestock [23][24][25][26][27][28][29] as they offer higher sensitivity and allow precise and direct identification of parasite species 30,31 .In this study, the prevalence of Cryptosporidium infections in cattle was only determined on the basis of molecular testing followed by the identification of parasite species using RFLP analysis.It was fast and reliable alternative to amplicon sequencing which often fails for samples containing a mixture of closely related parasite strains or species.To our knowledge this is the first nationwide epidemiological study on Cryptosporidium prevalence conducted in Europe, although a similar type of research has already been performed in China 31 .In contrast to our work, other Cryptosporidium studies carried out in cattle in Europe were only limited to animal populations kept regionally in each country [32][33][34][35][36][37] .
The prevalence of Cryptosporidium infections in cattle from Poland at the age of 1 to 16 weeks was 45.3%, with only significant difference in parasite prevalence between the older animal groups.Nevertheless, regional differences in prevalence of infections related to animals age were observed.In the north-western, central and eastern regions of Poland, a higher number of infections was found in the youngest animals up to 8 weeks of age.This can to some extend be explained by the presence of a higher cattle population housed in these regions with majority of animals represented by dairy breeds.There is a marked difference from the prevalence found in our previous study aiming at detection and molecular identification of Cryptosporidium species in cattle between the ages of 1 day and 6 years, in which it was on average 17% 9 .This large discrepancy could have resulted from testing of approximately threefold higher number of animals which gave more accurate data on Cryptosporidium prevalence at province level as well as from different sampling plan used in the current study.The narrow age range (1-16 weeks) of the sampled cattle may have also contributed to the results obtained as in older animals which were not tested in this study, Cryptosporidium infections occur less frequently 38,39 .It should be emphasised that sampled animals originated from randomly selected farms of different sizes which were located in the districts of their provinces with the highest cattle populations.Therefore, it is likely that higher density of animal population in the studied regions could have facilitated infection spread between animals and farms.This observation has previously been shown on commercial cattle farms [40][41][42] .Likewise, an association between dairy herd size and an increased risk of Cryptosporidium infection has been demonstrated 43 .Also shown in this study, the intensive cattle rearing system was mainly associated with dairy farming, which might have influence on prevalence of Cryptosporidium infection, as in all provinces with infection rate above 50%, dairy farming was the main rearing system.A direct comparison of the results obtained in this study with those of other European studies is difficult, as most of them focused on small animal populations and a discrete region 25,42 or were limited to selected farms 44 .Certainly, the higher Cryptosporidium prevalence of the present study cannot be attributed to differences in sample processing and detection methodology, because those studies also employed molecular methods.As an outcome comparable to our results, a high 48.6%rate of Cryptosporidium prevalence was observed in cattle in a cohort study conducted in Sweden 45 and 43.8% prevalence in dairy cattle herds in Cyprus 46 .Contrastingly to the European data, a low 14.1% infection rate was recorded in a large population study on pre-weaned cattle under the age of 3 months from northwest China 47 .Surprisingly, a low 9.9% infection prevalence was estimated for diarrhoeic calves in Korea 48 .Differences in infection prevalence can be found despite age similarity of the tested animals and size similarity of the sampled animal population.Various factors such as geographical location, environmental conditions, animal breed, husbandry system and rearing conditions may have an influence on the results obtained.
In the current study the majority of infections in calves up to 4 months were caused by C. bovis and C. ryanae.The species distribution in animals has changed in Poland over time 9 as currently C. parvum infections were reported less frequently.A similar infection pattern of detected species with dominating C. bovis and C. ryanae has also been shown in several studies from different parts of the world 45,49 .As demonstrated here, C. parvum was not the dominant species in healthy calves up to 8 weeks of age, and this finding is in agreement with previous observations 45 .Consequently, it is not surprising that different distribution patterns of C. parvum have been found depending on the health status of the tested animals 33,50,51 .In this study, the overall prevalence of C. parvum in cattle in Poland was estimated at 6.2%.It was also found that animals under the age of 1 month were not the major host for C. parvum , and the number of C. parvum-positive samples decreased with a higher animal age.These results are consistent with our previous observations on C. parvum prevalence in cattle in Poland 9 .Among the Cryptosporidium species identified in this study, C. parvum is considered the most important zoonotic species.
However, in this study unusual Cryptosporidium species for a bovine host, namely C. baileyi and C. suis, were detected in the herds investigated, in which infected animals were carrying those Cryptosporidium species asymptomatically.The calves positive for C. baileyi and C. suis represented different breeds (LM, HO and NCB) as well as rearing types (meat, dairy and dairy-meat varieties).Thus far, C. baileyi 52 , C. suis 53,54 , Cryptosporidium scrofarum 55 , Cryptosporidium xiaoi 34 , Cryptosporidium serpentis, Cryptosporidium tyzzeri 56 , C. felis 57 and Cryptosporidium occultus (previously known as C. suis-like) 34,55 have been detected in grazing ruminants without evidence that they were causing disease.In this study, the presence of C. baileyi and C. suis in cattle might be associated with an accidental oocyst ingestion with contaminated feed.By examining the routes whereby C. baileyi could have been transmitted, it seems likely that feed contamination occurred as a result of environmental parasite abundance linked to extensive poultry farming in areas where the calves were being reared.However, if this finding would be reported as repeatable, a further investigation would be required to reveal any possible parasite-host interactions.In the present study, mixed infections of C. bovis and C. ryanae or C. parvum and C. bovis were detected.They were previously found sporadically in young cattle in Poland 9 , China (C.parvum and C. bovis) 29  Significant differences in Cryptosporidium prevalence (from 27.8% to 62%) in Polish provinces were observed.Those results are also in agreement with reports from Korea describing a wide range of prevalence rates of infections between regions 48 .In this study a higher prevalence of Cryptosporidium infections in particular provinces was not always correlated with a higher population of housed animals, for example, in MP province with the lowest 27.8% Cryptosporidium prevalence, the cattle population was similar to this observed in PM with 50% frequency of infections.Of note is that the prevalence above 50% was mostly observed in provinces in which the tested animals were housed on farms with an intensive cattle rearing system which is mainly associated with dairy farming.Likewise, there were differences observed between species distribution and the region of the country.There were also no differences found in the frequency of infections caused by particular Cryptosporidium species in animals representing the intensive cattle rearing system compared to general cattle population.
Little is known about Cryptosporidium occurrence in cattle of different breeds.When data on Cryptosporidium prevalence was broken down by breed type, a relationship between the infecting specific parasite species and the animal breed was found.For instance, C. parvum prevailed in NCB, C. ryanae in LM and C. andersoni in MS cattle.However, the observed in this study prevalence of breed-associated parasite species should be interpreted with caution because of differences in the numbers of tested animals of each production purpose: the herds were of mostly dairy breeds with HO dominating animals and meat breeds were the next largest proportion.In this light, the current results do not necessarily indicate any greater breed sensitivity to Cryptosporidium infection.It seems to be justified that a higher number of animals of a particular breed type be tested to reveal exact parasite-host interactions.However, breed-related prevalence of Cryptosporidium infections has also been demonstrated in a comparison between European-bred animals, Zebu and animals cross-bred with Zebu 39 .While this relation was not indicated in our previous study related to Cryptosporidium epidemiology in cattle, it was evident for goats which showed breed-related differences in parasite prevalence 59 .In this study the distribution of Cryptosporidium species at the farm level was not analysed.Considering the randomness of sampling and varied ages of calves sampled on each farm, those farm-level variables would not guarantee obtaining reliable species distribution results.As a final observation of a possible limitation, although our sampling scheme covered all Polish provinces, the number of sampled animals may not reflect the actual population size in each region.Nevertheless, the sampling plan used allowed to estimate the prevalence of Cryptosporidium infections with 95% probability for the results with a binomial distribution.Only prevalences assessed for particular Polish provinces can be characterized by lower probability.However, the large, total sample size allowed to draw valid conclusions.Furthermore, testing of young animals < 4 months of age in which Cryptosporidium infections are mostly prevalent can be considered a sentinel study for occurrence of Cryptosporidium species in the population.

Figure 1 .
Figure 1.Province-related prevalence of Cryptosporidium infections in cattle.The numbers within parentheses indicate 95% CI.

Table 1 .
However, greater than 40% prevalence of infections was revealed in the majority of the monitored regions.The number of faecal samples collected from cattle up to the age of four months housed in different Polish provinces.

Table 3 .
Results on detected Cryptosporidium species in cattle in Poland.*Estimated values did not consider mixed C. bovis and C. ryanae or C. parvum and C. bovis infections.
58d Brazil (C.parvum and C. bovis; C. parvum and C. ryanae; C. parvum and C. andersoni) 39 .Majewska et al.58also detected concurrent parasite infections in Polish cattle, but in contrast to our findings, they were caused by C. parvum and C. andersoni.