Cryptosporidium infection in non-human hosts in Malawi

BANDA, Z., NICHOLS, ROSELY A.B., GRIMASON, A.M. & SMITH, H.V. 2009. Cryptosporidium in­ fection in non­human hosts in Malawi. Onderstepoort Journal of Veterinary Research, 76:363–375 Of 1 346 faecal samples from the Chikwawa and Thyolo districts of Malawi, analysed for the presence of Cryptosporidium oocysts between October 2001 and May 2003, 61.3 % were from cattle (29.8 % of these were from calves < 6 months old). Cryptosporidium oocysts were detected during all three seasons studied in Chikwawa and Thyolo. In Chikwawa, 13.6 % of adult cattle and 11.7 % of calves were infected, compared to 28.9 % of adult cattle and 36.7 % of calves in Thyolo. Dependent on sea­ son, between 7.8 % and 37.7 % (Chikwawa) and 16.7 % and 39.3 % (Thyolo) of cattle samples con­ tained oocysts. In Chikwawa, the highest percentage of infections occurred in the cool season, whereas in Thyolo, the highest percentage of infections occurred in the dry season. Faecal samples from goats [n = 225], pigs [n = 92], sheep [n = 6]), rabbits, guinea pigs, chickens, ducks, turkeys, doves and guinea fowls were also analysed. Up to 5.6 % of goat samples contained oocysts in Chikwawa, compared to between 16.7 % and 39.3 % in Thyolo. Again, in Chikwawa, the highest percentage of infections occurred in the cool season and the lowest in the rainy season, whereas, in Thyolo, the highest percentage of infections occurred in the dry season and the lowest in the cool season. In pigs, more infections were detected in the dry season in Chikwawa, but infections in the cool season were similar (17.7 %), whereas in Thyolo, infections occurred in all three seasons (17.9 % in the rainy season, 25 % in the cool season and 60 % in the dry season). Of ten diarrhoeic, oocyst positive cattle faecal samples collected from Chikwawa and subjected to PCR­RFLP, four oocyst positive samples (two from heifers, one from a cow and one unknown) were amplified at an 18S rRNA and Cryptosporidium oocyst wall protein (COWP) loci. RFLP of the 18S rRNA locus indicated that Cryptosporidium parvum, Cryptosporidium hominis, Cryptosporidium bovis and / or Cryptosporidium ryanae DNA, or a mixture of them was present. Cryptosporidium parvum DNA was identified in one sample that amplified at the COWP locus, indicating the presence of the major zoonotic Cryptosporidium species in Malawi.

constitutes part of the complex group of parasitic, bacterial and viral diseases that impair the ability to achieve full potential and impair development and socioeconomic improvements.All diseases includ ed in the World Health Organisation (WHO) Neglected Diseases Initiative have a common link with poverty and, as the current view is to take a comprehensive approach to all these diseases, Cryptosporidium was included in the 'Neglected Diseases Initiative' in 2005 (Savioli, Smith & Thompson 2006).
The epidemiology of Cryptosporidium is complex with many hosts harbouring more than one species that can cross host species barriers, and having at least one hostadapted species, which are less like ly to cross host barriers (Fayer 2008;Smith 2008).Of the 21 recognised Cryptosporidium species, hu mans can be infected with Cryptosporidium parvum, Cryptosporidium hominis, Cryptosporidium melea gridis, Cryptosporidium felis, Cryptosporidium muris, Cryptosporidium canis, Cryptosporidium suis and the Cryptosporidium cervine, monkey, skunk and rab bit genotypes (Caccio et al. 2005;Nichols, Chal m ers, Sopwith, Regan, Hunter, Grenfell, Har rison & Lane 2006), but two species, C. parvum and C. ho minis, cause the majority of human disease (Caccio et al. 2005).In livestock, C. parvum, Cryptosporidium andersoni, Cryptosporidium baileyi and C. melea gridis cause disease (Smith 2008).Cryp to sporidium par vum cryptosporidiosis is a cause of scours in young, unweaned livestock, and is the species rec ognized for having the greatest zoonotic potential.
Transmission of Cryptosporidium occurs following the ingestion of the environmentally robust transmis sive stage, the oocyst.Transmission can be direct or indirect.Direct routes include person to person, an imal to animal, and animal to person (zo onotic) and indirect routes include contaminated faeces and soil, surface and ground water, wastewater, feed, herb age, vectors and equipment (Cacciò et al. 2005;Smith, Cacciò, Tait, McLauchlin & Thompson 2006;Smith et al. 2007;Xiao & Feng 2008).
High mortality due to cryptosporidiosis has been re ported in cattle mainly among meat breeds (Sanford & Josephson 1982;Peeters, Villacorta, Vanop den bosch, Vandergheynst, Naciri, AresMazas & Yvore 1992).Cryptosporidium andersoni infection in cattle is known to affect weight gain (Anderson 1987) and milk production (Esteban & Anderson 1995).Crypto sporidiosis in neonatal small ruminants such as goats and sheep can result in weight loss and re tarded growth during the first few weeks of their life (Naciri, Yvore & Leieux 1984) and mortality has been reported in goat kids (Vieira, Silva, Tolentino, Lima & Silva 1997) and adult goats (Johnson, Muirhead, Windsor, King, AlBusaidy & Cornelius 1999).
At present, there is no information available regard ing the prevalence of Cryptosporidium spp. in non human hosts in Malawi.Owing to the importance of this parasite, and the potential for zoonotic trans mission, a study was carried out in the Chikwawa and Thyolo Districts of Malawi with the main objec tive of determining the prevalence of Cryptosporidium spp.infection, primarily in cattle.

Study areas
The study was carried out in Chikwawa and Thyolo Districts of the southern region of Malawi from Octo ber 2001 to May 2003.Chikwawa is ~50 km from the commercial capital of Malawi, Blantyre, at an al titude of ~200 m above sea level.It has mean annu al maximum temperatures between 30-32 °C and mean annual minimum temperatures between 18-20 °C, and has an annual rainfall of ~800-1 200 mm.
Thyolo is ~39 km from Blantyre at an altitude of ~820 m above sea level.It has mean annual maxi mum temperatures of ~29 °C and mean annual min imum temperatures of ~19 °C and has an annual rainfall of ~1 258 mm.

Study seasons and selection of village households
The Chikwawa study was carried out during the dry, rainy and cool seasons from

Sample collection
Faecal samples were collected once per season from nondiarrhoeic and diarrhoeic young and adult animals into 50 g capped plastic sample containers (Enterprise Containers, Blantyre, Malawi).Farmers were visited between 05:00 and 09:00 before they released the animals from their kraals for foraging, browsing or grazing.The animals were stimulated to defaecate by making them move about in their kraals.A faecal sample was scooped using a clean wooden applicator stick from the centre of the stool and carefully placed into a clean plastic container which was sealed and identified with an unique sample number.The date of collection, village, sex, age, colour and condition of the animal, macroscop ic appearance of the stool (diarrhoeic or formed), and name of the owner of the animal and of the sample collector were entered onto a form.Samples were transported in cool boxes to the laboratory at the University of Malawi, The Polytechnic (UMP), and stored at 4 °C until analysed.

Cryptosporidium oocyst detection
A representative portion of faeces (approx.500 mg) was smeared onto a clean glass slide with a wood en applicator stick, air dried, fixed in methanol for 3 min and subjected to modified Ziehl Neelsen (mZN) staining (Casemore 1991;Smith 1992).

Oocyst abundance
Oocyst abundance was determined by assessing the average number of oocysts seen per ten fields (1+ = < 5 oocysts per field; 2+ = > 5 oocysts per field; 3+ = > 20 oocysts per field).Oocysts were identified using the 40x objective and confirmed under the 100x oil objective.

Epifluorescence microscopy
Oocysts were visualized by epifluorescence micro scopy and their internal morphology assessed us ing Nomarski differential interference contrast (DIC) microscopy.A blue filter block (excitation 490 nm; emission 510 nm) was used to visualize FITCC mAb localization and an ultraviolet (UV) excitation (excitation 355 nm, emission 450 nm) was em ployed to determine the presence of DAPIstained sporozoite nuclei.All assessments of the presence of fluorescent nuclei and internal morphology were undertaken at either 400x or 1 000x magnification.

Oocyst purification and DNA extraction
Cryptosporidium oocyst positive samples were puri fied by water ether concentration (Bukhari & Smith 1995;Nichols, Moore & Smith 2006) or immuno magnetic separation (IMS, Dynal A.S. Oslo, Norway) according to the manufacturer's instructions.Oocysts were suspended in lysis buffer [50 mM TrisHCl pH 8.5, 1 mM EDTA, 0.5 % sodium dodecyl sulphate (SDS)] and DNA was released from oocysts follow ing 15 cycles of freezethawing (freezing in liquid nitrogen for 1 min, followed by thawing at 65 °C for 1 min).Proteinase K (final concentration 200 µg mℓ -1 ) was added, and the samples incubated at 55 °C in a water bath, for 3 h.Proteinase K was heat denatured (90 °C, 20 min), samples chilled on ice for 1 min, and centrifuged for 5 min at 10 000 x g (Nichols & Smith 2004).The supernatant (70 µℓ) was transferred to a clean tube and either used im mediately for PCR amplification or stored at -20 °C until used.

PCR protocol
PCR amplifications were performed in Perkin Elmer thermocycler model 480 in 0.5 mℓ thinwalled tubes.For the direct 18S rRNA PCR reaction volumes of 100 µℓ consisted of premixed reagents consisting of 200 µM of each of the four dNTP's (Amersham Pharmacia Biotech UK Ltd., Amersham Place, Little Chalfont, Bucks., HP7 9NA, UK), 0.2 µM each of primers CPBDIAGF/R (Cruachem, Todd Campus Acre Road, Glasgow, G20 0UA, UK), bovine serum albumin at final concentration of 400 µg mℓ -1 , MgCl 2 at 3.5 mM, 2.5 U of Taq polymerase (Advanced Biotechnologies, Holly Ditch Farm, Mile Elm, Calne, Wilt shire, SN11 0PY, UK) in 1 x PCR buffer IV (Ad vanced Biotechnologies) and Tween 20 at a final concentration of 2 % to inactivate 0.05 % sodium dodecyl sulphate.Approximately 40 µℓ of mineral oil was layered on top of each tube containing the dis pensed PCR mix and 10 µℓ of DNA template was pipetted under the oil.Positive controls consisted of C. parvum oocyst DNA and negative controls were DNAfree reaction tubes.Samples were subjected to 39 cycles of amplification according to Johnson, Pieniazek, Griffin, Misener & Rose 1995) and the 435 bp PCR product was visualized following ethid ium bromide staining of 1.4 % agarose gels (Nichols et al. 2003(Nichols et al. , 2006)).PCR protocol for the STNCOWP locus was according to Homan et al. (1999) with the inclusion of bovine serum albumin and Tween 20 in the PCR mixture as described above.

Cryptosporidium species identification by PCRRflP analysis
Twenty µl of 18S rRNA PCR product were simulta neously digested with 20U of each restriction en zymes Ase I (New England Biolabs, 75-77 Knowl Piece, Wilbury Way, Hitchin, Herts, SG4 0TY, UK and DraI (Invitrogen, 3 Fountain Drive, Inchinnan Bus iness Park, Renfrew, PA4 9RF, UK) in 50 µℓ of 1 x NE Buffer 3 (New England Biolabs) for 2 h at 37 °C.Ten microliters of PCR product from the STN COWP assay (Homan et al. 1999) was digested with 10 U of restriction enzyme Taq I (Invitrogen) accord ing to the supplier's instructions.Undigested con trols were run alongside the digested fragments in a 2 % agarose gel at 100 volts for 1-2 h and stained with ethidium bromide (0.5 µg mℓ -1 ) incorporated in the gel and the running buffer.

Quality assurance
All diagnostic methods performed at SPDL were con ducted in compliance with the Standard Operating Procedures accredited by Clinical Pathology Accred itation (UK) Ltd. and the Drinking Water Inspectorate (Regulatory Cryptosporidium Standard Operating Procedures).All methods performed at UMP are ac credited at SPDL.

Cryptosporidium in Malawian faecal samples
A total of 1 346 faecal samples were analysed.Of these, 905 were from Chikwawa consisting of 559 samples collected in the dry season, 189 in the rainy season and 157 in the cool season, and 441 sam ples were from Thyolo consisting of 219 samples collected in the dry season, 114 in the rainy season and 108 in the cool season.Cryptosporidium oocysts were present in 16.7 % (224/1 346) of samples, with a similar rate of occurrence in Chikwawa (7.7 %, 104/1 346) and in Thyolo (8.9 %, 120/1 346).
The number of animal species sampled and the percentage positivity in the three seasons from both Chikwawa and Thyolo are presented in Table 1.The percentage of samples that was oocyst positive was dependent on season in the two districts sam pled, with the highest percentage positives detected in the cool season (31.9 %) in Chikwawa and in the dry season (34.7 %) in Thyolo (Table 2).

Age
In Chikwawa, between 7.8 % and 37.7 % of cattle samples contained oocysts, dependent on season, with the highest percentage of infections occurring in the cool season and the lowest in the rainy sea son (Table 1).In Thyolo, between 3 % and 32.6 % of cattle samples contained oocysts, dependent on sea son, with the highest percentage of infections occurring in the dry season and the lowest in the cool season (Table 1).In Chikwawa, between 0 % and 5.6 % of goat samples contained oocysts, with the highest percentage of infections again occurring in the cool season and the lowest in the rainy season (Table 1).In Thyolo, between 3 % and 32.6 % of goat samples contained oocysts, with the highest percentage of infections occurring in the dry season and the lowest in the cool season (Table 1).In Chi kwawa, the highest percentage of infections in pigs were detected in the dry season, but infections in the cool season were similar (17.7 %), whereas in Thyolo, infections occurred in all three seasons (17.9 % in the rainy season, 25 % in the cool sea son and 60 % in the dry season) (Table 1).
Cryptosporidium infections in rabbits, guinea pigs, chicken, ducks, turkeys, doves and guinea fowls were also detected, and while the data indicate that infections were detected in these hosts, the data are sparse and do not cover all seasons analysed.Based on mZN positivity, Table 2 identifies the num bers and percentages of stools containing oocysts in our samples, over the three seasons analysed and serves as an indicator of the prevalence of Cryp tosporidium in the Chikwawa and Thyolo Districts of Malawi.Cryptosporidium oocysts were more likely to be detected in the cool season in Chikwawa and in the dry season in Thyolo whereas oocysts were least likely to be detected in the dry season in Chi kwawa and in the cool season in Thyolo.The age based breakdown of Cryptosporidium oocyst posi tive samples from cattle is presented in Tables 3  and 4. We detected Cryptosporidium oocysts during all three seasons in the Chikwawa and Thyolo dis tricts of Malawi.In Chikwawa, 13.6 % of adult cattle and 11.7 % of calves were infected, compared to 28.9 % of adult cattle and 36.7 % of calves in Thyolo (  Anusz, Mason, Riggs & Perryman (1990) reported a detection limit of 10 6 oocysts mℓ -1 of faeces using Kinyoun staining.Oocyst concentration methods can increase the sensitivity of detection.Between 1 x 10 4 and 5 x 10 4 oocysts g -1 of unconcentrated stool are necessary to obtain a 100 % detection efficiency using the Kinyoun staining method (Weber, Bryan, Bishop, Wahlquist, Sullivan & Juranek 1991).Vari a tions in faecal consistency influence the ease of de tection, with oocysts being more easily detected in concentrates made from watery, diarrhoeal speci mens than from formed stool specimens (Weber et al. 1991).Webster, Smith, Giles, Dawson & Robert son (1996) reported that oocysts were not detected in bovine faecal samples seeded with 10 4 C. par vum oocysts g -1 following formol ether concentra tion and examined using AP or IFAT, and that su crose or salt flotation was required to increase the threshold of detection to 4-6 x 10 3 oocysts g -1 using AP or IFAT.
With the exception of the small numbers of sheep samples analysed (n = 6), Cryptosporidium spp.oocysts were detected in all other host species ( In addition to the likelihood of some of these Crypto sporidium species causing disease in their major hosts, a zoonotic potential also exists for those ani mals infected with human infectious species, par ticularly, but not exclusively, those infected with C. parvum. We detected Cryptosporidium oocysts during all three seasons in the Chikwawa and Thyolo Districts of Malawi.In Chikwawa, 13.6 % of adult cattle and 11.7 % of calves were infected, compared to 28.9 % of adult cattle and 36.7 % of calves in Thyolo (Table 4).Season influenced Cryptosporidium occurrence in adult cattle and calves in Chikwawa and Thyolo.Chikwawa experiences higher temperatures and lower annual rainfall than Thyolo, and temperature influences oocyst survival.
Oocysts remain infectious for up to 3 months when stored in water at ambient (20-30 °C) temperatures (Fayer, Trout & Jen kins 1998)  As age influences the species of Cryptosporidium infecting bovines, with calves less than 6 months old being most likely to harbour the zoonotic spe cies, C. parvum (Feng et al. 2007), there should be a greater potential for the zoonotic spread of C. par vum in the cool season in Chikwawa and in the dry season in Thyolo, where infection in calves was highest (~50 %; Table 3).Furthermore, the zoonotic potential might be greater in Thyolo than in Chikwa wa, as 36.7 % of calves were infected in the former compared with 11.7 % in the latter district (Table 4).
Cryptosporidium is a common infection of bovines worldwide.

TABLE 1
Percentage of animals infected withCryptosporidium species oocysts during the cool, dry and rainy seasons in the Chikwawa and Thyolo Districts of Malawi

TABLE 3
Percentage of cattle excreting Cryptosporidium species oocysts by age and season in the Chikwawa and Thyolo Districts of Malawi

assurance at SPDL on UMP mZN putative Cryptosporidium oocyst positive samples obtained from cattle in Chikwawa
Table 4).Season influenced Cryptosporidium oc currence in adult cattle and calves in Chikwawa and Thyolo.The highest percentages of oocyst positivity occurred in both older cattle and calves in the cool season in Chikwawa and in the dry season in Thyolo with nearly 50 % of calves being infected (Table 3).
rRNA locus, following Ase I /DraI digestion, all four samples showed a RFLP pattern that corresponds to C. parvum, C. hominis, C. bovis or C. ryanae (pre viously, Cryptosporidium deerlike genotype), or a mixture of them.Only one of the four 18S rRNA positive samples amplified at the STNCOWP locus and the RFLP pattern indicated the presence of C. parvum DNA.
(Smith et al. 2007)96 1995)and 19 % in adult dairy cows(Bukhari & Smith 1996) have been documented.These findings are in agreement with the results of this study.This study clearly demonstrates the presence of Cryptosporidium infections in a variety of avian and mammalian hosts in Malawi, indicating the potential role of such hosts as reservoirs of infectious para sites, and in particular infections in postweaned and adult cattle.Although we did not use molecular typing tools in our survey, we did use them for qual ity assurance.Four of the ten mZN, AP and IFAT positive samples used for quality assurance at SPDL were amplified by PCR using the direct 18S rRNA gene locus, but only one oocyst positive sam ple amplified at the STNCOWP locus.The direct 18S rRNA gene locus can determine all Crypto spor idium species after sequencing, while the STN COWP locus can identify C. parvum, C. hominis and C. meleagridis by RFLP(Jiang & Xiao 2003).Although each oocyst contains 20 copies of the 18S rRNA gene, the direct 18S rRNA PCR was only marginally sensitive with these ten oocyst positive slides.The expansion of valid Cryptosporidium spe cies in the last 5 years(Smith et al. 2007)identifies that RFLP analysis is becoming less useful current ly, as Ase I and DraI digestion of the 18S rRNA am plicon cannot discriminate between C. parvum, C. hominis, C. bovis and C. ryanae.The C. parvum RFLP pattern was identified from the only oocyst positive sample that amplified at the STNCOWP single copy gene locus.The low PCR positivity rate using these loci may be due to coextraction of in hibitors with DNA, the amount of DNA template ex tracted and/or the absence of nuclear material with in the oocysts.Most of our microscopy positive samples had low oocyst abundances (highest oo cyst abundance = 2+), possibly indicating either the tailing off of cryptosporidiosis or asymptomatic in fection and the low sensitivity of mZN.Crypto spor idium parvum was unambiguously detected in one (10 %) sample (heifer; Table5) establishing that C. parvum infection occurs in Malawian cattle.Since C. parvum can infect susceptible human and non human hosts, this species has both veterinary and public health significance in Malawi.