Abstract
We developed and validated a PCR-based method for identifying Cryptosporidium species and/or genotypes present on oocyst-positive microscope slides. The method involves removing coverslips and oocysts from previously examined slides followed by DNA extraction. We tested four loci, the 18S rRNA gene (N18SDIAG and N18SXIAO), the Cryptosporidium oocyst wall protein (COWP) gene (STN-COWP), and the dihydrofolate reductase (dhfr) gene (by multiplex allele-specific PCR), for amplifying DNA from low densities of Cryptosporidium parvum oocysts experimentally seeded onto microscope slides. The N18SDIAG locus performed consistently better than the other three tested. Purified oocysts from humans infected with C. felis, C. hominis, and C. parvum and commercially purchased C. muris were used to determine the sensitivities of three loci (N18SDIAG, STN-COWP, and N18SXIAO) to detect low oocyst densities. The N18SDIAG primers provided the greatest number of positive results, followed by the N18SXIAO primers and then the STN-COWP primers. Some oocyst-positive slides failed to generate a PCR product at any of the loci tested, but the limit of sensitivity is not entirely based on oocyst number. Sixteen of 33 environmental water monitoring Cryptosporidium slides tested (oocyst numbers ranging from 1 to 130) contained mixed Cryptosporidium species. The species/genotypes most commonly found were C. muris or C. andersoni, C. hominis or C. parvum, and C. meleagridis or Cryptosporidium sp. cervine, ferret, and mouse genotypes. Oocysts on one slide contained Cryptosporidium muskrat genotype II DNA.
The apicomplexan parasite Cryptosporidium is responsible for many outbreaks of human disease affecting both immunocompetent and immunocompromised individuals (3, 12). Cryptosporidiosis is also of veterinary importance, since farm animals can be infected with various Cryptosporidium species that cause morbidity, with subsequent economic consequences. Cryptosporidium has a large range of host species, and infected hosts can shed large numbers of viable, environmentally robust oocysts that contaminate wastewater, watersheds, surface water, and groundwater.
The epidemiology of Cryptosporidium is complex, with many hosts harboring more than one species and having at least one host-adapted species (36). Of the 14 recognized Cryptosporidium species, 2 species, C. parvum and C. hominis, cause the majority of human disease. C. parvum also infects newborn farm animals, mainly calves and kids. Genetic analyses have revealed that more than one species of Cryptosporidium can infect susceptible immunocompromised (2, 6, 9, 15, 19-21), and immunocompetent (13, 18-20, 34) human hosts. Species of Cryptosporidium reported to have crossed host specificity barriers and that have been detected in human stools include C. meleagridis, C. felis, C. muris, C. canis, C. suis, and Cryptosporidium sp. cervine and monkey genotypes (3).
Contamination of the aquatic environment with Cryptosporidium spp. oocysts that are not infectious to susceptible human hosts contributes to the difficulties in assessing the risk to public health from waterborne oocysts. The extent of the occurrence of species other than C. parvum in the environment is only now being addressed. Analysis of U.S. storm water samples revealed the presence of Cryptosporidium spp. in 27 of 29 samples, mainly wildlife Cryptosporidium genotypes (35). The most common genotypes/species found in surface waters were C. parvum, C. hominis, and C. andersoni, with C. andersoni reported to be the most commonly found in wastewater (eight samples). Of the 22 Cryptosporidium species and genotypes identified in 107 of 121 water samples from storm events in three New York area watersheds, only 11 were of known species or genotypes (11). Storm waters are expected to carry greater microbial loads, including larger numbers of Cryptosporidium oocysts (14). Frequently, oocysts occur at low densities in water (23, 24, 26), and methods which can genotype small numbers of organisms reliably and reproducibly from water concentrates are required to determine which species are present, and with what frequency, in water.
In the United Kingdom, government regulations (28) exist to ensure that drinking water is treated adequately to remove Cryptosporidium spp. oocysts. A treatment standard of an average of less than 1 oocyst in 10 liters of water supplied from a treatment works, with at least 40 liters per hour of treated water going into the supply, was set. Also identified (28) is a standard operating protocol (SOP) for monitoring Cryptosporidium oocysts in water supplies. Identification and enumeration of oocysts by using modifications of the methods of Smith et al. (25, 27) and Grimason et al. (8) are performed on air-dried water concentrates that have been methanol fixed onto glass microscope slides and stained with a Drinking Water Inspectorate (DWI)-approved commercially available monoclonal antibody that recognizes exposed epitopes on oocyst walls (DWI-FITC-C-mAb) and the nuclear fluorogen 4′6-diamidino-2-phenyl indole (DAPI). Slides are viewed under the appropriate filters of an epifluorescence microscope, and oocysts are identified, measured, and enumerated. A similar approach for the identification and enumeration of Cryptosporidium oocysts for public health purposes has been adopted by the U.S. Environmental Protection Agency, with methods 1622 and 1623 (30, 31).
This study reports the development and validation of a method for genotyping oocysts from microscope slides that involves retrieving the entire water concentrate sample dried onto a microscope slide. Following identification and enumeration using standardized methods (29-31), DNA is extracted, amplified by PCR, and subjected to restriction fragment length polymorphism (RFLP) analysis for species and genotype identification. The method was developed and validated using (i) oocysts from four Cryptosporidium species seeded onto microscope slides and (ii) Cryptosporidium monitoring slides containing Cryptosporidium spp. oocysts obtained from United Kingdom water companies following monitoring for the presence of oocysts in approximately 1,000 liters of drinking water (29).
MATERIALS AND METHODS
Sources of Cryptosporidium oocysts.
Purified Cryptosporidium oocysts were obtained from the following three sources: (i) C. parvum oocysts of ovine origin (MD, cervine-ovine; Moredun Research Institute, Edinburgh, United Kingdom), (ii) C. muris (Waterborne Inc., New Orleans, La.), and (iii) C. parvum, C. hominis, and C. felis oocysts of human origin, obtained from routine submissions to the Scottish Parasite Diagnostic Laboratory (SPDL), Glasgow, United Kingdom. Oocysts from the first two sources were supplied purified. All human isolates were purified by water-ether concentration and sucrose flotation according to the methods of Bukhari and Smith (1). All oocysts were stored in reverse osmosis water, in the dark at 4°C until used.
Thirty-three Cryptosporidium spp. oocyst-positive slides were received at SPDL from three English water utilities (Severn Trent Cryptosporidium Laboratory [STL], Coventry, United Kingdom; Analytical & Environmental Services Ltd. Cryptosporidium Laboratory [AES]; and Northumberland and Wessex Water Ltd., Bath, United Kingdom), and one Scottish water utility (Scottish Water Cryptosporidium Laboratory [SW]). Slides were reexamined by epifluorescence microscopy and then kept in the dark, at room temperature, until oocyst DNA was extracted. All slides received were prepared according to the SOP for DWI regulatory slides; however, regulatory slides are kept for a considerable length of time, and the slides forwarded to SPDL were nonregulatory, i.e., they had no chain-of-evidence documentation accompanying them.
Evaluation of the sensitivity and efficacy of extraction of Cryptosporidium oocysts from slides during method development.
Either a 10- or 20-μl volume from a suspension of 1 × 103 purified Cryptosporidium spp. oocysts ml−1, enumerated using an improved Neubauer hemocytometer or from suspensions of Cryptosporidium spp. oocysts containing known numbers of oocysts prepared by flow cytometry (modified from the method of Reynolds et al. [22]) was used to prepare control slides. Flow cytometry-sorted oocyst numbers were validated by filtering 10% of suspensions through 1.2-μm Millipore polycarbonate filter membranes (Isopore). Entrapped oocysts were methanol fixed and stained with DWI-FITC-C-mAb (Cellabs Crypto Cel IF antibody; TCS Water Services, Buckingham, United Kingdom) and examined using epifluorescence microscopy (data not shown). Flow cytometry-sorted oocyst suspensions were stored at 2 to 8°C until used.
Oocyst staining procedure.
Each oocyst suspension was pulsified for approximately 5 to 10 seconds before a sample was dispensed onto the well of a single-well microscope slide. The pulsifier (Kalyx Biosciences Inc., Ontario, Canada) is a single-speed pulsifier that generates a near-uniform oocyst suspension following horizontal agitation. Samples were air dried at room temperature, methanol fixed, and stained with DWI-FITC-C-mAb and DAPI, according to the methods of Smith et al. (27).
Microscopy.
An Olympus BH-2 epifluorescence microscope equipped with Nomarski differential interference contrast (DIC) optics was used to view the prepared slides. Epifluorescence microscopy using UV excitation (excitation at 355 nm, emission at 450 nm) was used to determine the presence of the DAPI-stained sporozoite nuclei. A blue filter block (excitation at 490 nm, emission at 510 nm) was used to visualize FITC-C-mAb emissions. Nomarski DIC optics were used to determine internal morphology. All evaluations for the presence of fluorescent nuclei and internal morphology were undertaken using both 40× and 100× objectives.
Microscopic examination of test slides.
Each slide was examined according to the DWI SOP (29). An object, located under the 20× objective, which fit the initial criterion of a Cryptosporidium oocyst was examined further using either the 40× or 100× oil immersion objectives as required. The number of DAPI-stained sporozoite nuclei present in each oocyst was determined and recorded. Nomarski DIC optics were used to assess the internal morphology of oocysts (29).
Protocol for removing oocysts from test slides.
All slides were processed to completion, individually. Slides were placed on absorbent tissue, and a cotton swab moistened in nail varnish remover was applied onto the nail varnish to soften it. The perimeter of the coverslip was swabbed with the impregnated swab to soften the nail varnish, and the opposite end of the swab was used to scrape the softened nail varnish from the coverslip-slide interface onto the absorbent tissue. A clean scalpel blade was used to lever a corner of the coverslip from the slide surface, and then the coverslip was gently lifted off the slide. The opposite end of the coverslip was held gently to avoid sideways movement of the coverslip or slide. Once lifted, the coverslip was inverted and placed onto the absorbent tissue. The Teflon-coated area of the slide surrounding the well was dried with a small piece of folded absorbent tissue, and then 10 μl of lysis buffer (LB; 50 mM Tris-HCl pH 8.0, 1 mM EDTA pH 8.0, 0.5% sodium dodecyl sulfate) was pipetted onto the well of the slide. The entire surface of the well was scraped with a sterile 1-μl bacteriological inoculation loop (Nunc, United Kingdom), and once scraped, the loop was placed on a support so that it did not rest on a contaminated surface.
Residual LB was aspirated by tilting the slide to an angle of about 45° from the horizontal towards the operator and aspirating the fluid which collected at the bottom of the well by placing the tip of a P20 Gilson pipette fitted with a filter-tipped pipette tip close to the fluid.
The scraped sample in LB was pipetted into an appropriately labeled 0.5-ml screw-cap microcentrifuge tube. A further 10-μl volume of fresh LB was deposited onto the sample well using a clean pipette tip, and the sample was scraped using the same inoculation loop. Once scraped, the loop was placed on a support so that it did not rest on a contaminated surface. All liquid was removed from the well as described above, and then the slide was rotated through 180° and the slide-scraping steps were repeated, twice again. The final volume of the sample amounted to ∼40 μl. The loop was snapped, carefully, by pressing it against the inner wall and the rim of the microcentrifuge tube and left inside the tube, which was capped. Swabs, gloves, and absorbent tissues were disposed of immediately after each sample was removed from a slide. Once scraped, the slide was retained and reexamined by epifluorescence microscopy to determine the efficiency of the removal procedure.
During method development, the coverslips removed from the slides were also examined for the presence of oocysts. This was to ensure that no oocysts were lost from the slides onto the coverslips.
Preparation of oocyst lysate by freezing and thawing.
DNA extraction was performed as previously described (16, 17). Briefly, tubes containing the scraped oocyst suspensions and the end of the bacteriological loop were immersed in liquid nitrogen for 1 min and thawed in a 65°C water bath for 1 min. This freeze-thawing cycle was repeated 15 times, and every five cycles, tube contents were mixed by gentle rocking. Each sample was centrifuged at 14,000 × g for 10 s to ensure that all the sample lysate was deposited at the base of the tube. Lysate was transferred into a clean tube containing 1.6 μl of proteinase K at 5 mg ml−1 and incubated at 55°C for 3 h in a water bath. Following incubation, capped tubes were centrifuged at 14,000 × g for 10 s to ensure that all the sample lysate was deposited at the base of the tube. Samples were incubated at 90°C for 20 min in a water bath to denature proteinase K, chilled on ice for ∼1 min, and then centrifuged at 14,000 × g for 5 min at room temperature. All supernatant (∼30 μl) was transferred to a clean, labeled tube and stored at −20°C until used.
PCRs.
Determination of PCR sensitivity was performed with C. parvum oocyst-seeded slides using a nested CPB-DIAG 18S rRNA gene PCR assay (17), a nested 18S rRNA gene PCR-RFLP (32), a single-tube nested Cryptosporidium oocyst wall protein (COWP) PCR-RFLP assay for genotyping of C. hominis and C. parvum (10), and a multiplex allele-specific PCR assay for the dhfr gene, which genotypes C. hominis and C. parvum by direct PCR product size polymorphism (7).
PCRs were set up in a designated laboratory in a UV-presterilized hood. Each reaction was performed in either 50 or 100 μl containing premixed reagents at final concentrations of 200 μM of each of the four deoxynucleoside triphosphates, bovine serum albumin at 400 μg ml−1, MgCl2 at concentrations varying from 2.0 to 6 mM, depending on the PCR assay, 2.5 U of Taq polymerase (ABgene, United Kingdom), Tween 20 at 2%, and primers purchased from MWG Biotechnology UK Ltd. (Milton Keynes, United Kingdom) at the concentration specified for each assay in 1× PCR buffer IV (ABgene, United Kingdom). Five microliters (for direct PCR assays or STN-COWP PCRs) or 2 μl (for primary PCR of two-step nested PCR assays) of oocyst lysate (defrosted at room temperature, mixed by vortexing for 10 s, and centrifuged at 14,000 × g for 10 s in a microcentrifuge) was used for the amplification. Three negative controls were set up for each PCR run: one using the water designated for preparing the megamix (performed in the laboratory designated for pre-PCR manipulations), one using LB set up before dispensing the test samples, and one set up after all the test samples for an individual PCR run had been dispensed. One positive control of a known DNA concentration, which was appropriate to each PCR run, was set up as the last sample. Secondary PCRs were set up by transferring 2 μl of primary PCR mixture to a 100-μl total reaction volume following published protocols. PCR amplifications were performed in a Perkin-Elmer thermocycler, model 9600 or 480, following published amplification protocols.
The PCR product was visualized by gel electrophoresis on either 2% or 1.4% agarose gels and stained with ethidium bromide on a UV transilluminator (UVT-20 M/V; UV emission at 302 nm; Herolab). Band intensities were classified according to the concentration of amplicons as follows: negative PCR, faint band denoting positive PCR but insufficient amplicon concentration for RFLP analysis; 1+, low-intensity bands with sufficient amplicon concentration for RFLP analysis; 2+, medium-intensity bands with sufficient amplicon concentration for RFLP analysis; 3+, high-intensity bands with sufficient amplicon concentration for RFLP analysis; and 4+, very-high-intensity bands with sufficient amplicon concentration for RFLP analysis. Gels were photographed using the Gel Doc 2000 system (Bio-Rad, United Kingdom) equipped with the QuantityOne software for gel documentation and quantitation of PCRs.
RFLP analysis by enzymatic digestion of PCR products.
Restriction enzymes DraI, SspI, DdeI, TaqI (Invitrogen, United Kingdom), and AseI (New England Biolabs, United Kingdom) were used according to the manufacturers' instructions. Twenty microliters of PCR product was digested with 20 U of each enzyme in a total volume of 50 μl in the appropriate buffer provided by the manufacturer. Digestions were completed at 37°C (DraI, AseI, SspI, and DdeI) or 65°C (TaqI) for 1 to 2 h, and the digested products were resolved in 2% agarose gels. When simultaneous digestion with DraI and AseI was performed, NE buffer 3 (New England Biolabs, United Kingdom) was used. This provides 100% efficiency of digestion with AseI and approximately 85% with DraI.
RESULTS
Development and validation of a method for oocyst species detection using control slides.
Oocyst (C. parvum MD isolate) numbers ranged from 2 to 18 (Table 1). Empty oocysts (one to three oocysts per sample) were detected in nearly 50% (22/48) of the samples and, at worst, constituted 50% of oocysts. We analyzed a further 24 samples to determine whether oocysts were retained on coverslips, but we did not detect any (data not shown).
TABLE 1.
Microscopic and molecular assessment of the effectiveness of scraping C. parvum (Moredun isolate C1/03) oocysts from seeded samples deposited onto welled slides
| No. of DAPI+ oocystsa (no. of empty oocysts) | N18SDIAG PCR resultb |
|---|---|
| 2; 2 (2) | +; + |
| 3; 3 (1) | +; + |
| 4; 4 | −; + |
| 5; 5; 5; 5 (1) | +; +; +; + |
| 6; 6; 6; 6; 6 (1); 6; 6; 6 | +; +; +; +; +; +; −; + |
| 7 (2) | + |
| 8 (3); 8 (1) | −; + |
| 9 (1); 9 (2); 9 (1); 9 (2) | +; +; +; + |
| 10; 10 (1); 10 (2); 10 | +; +; +; + |
| 11; 11 (1); 11 (1); 11 (2); 11 (1) | +; +; +; +; + |
| 12 (1); 12; 12 (2) | +; −; + |
| 13; 13; 13; 13 (2) | +; +; +; +; + |
| 15 (1); 15; 15 | +; +; + |
| 16 | + |
| 18; 18 (2); 18 | −; +; + |
Following reexamination of slides after scraping, only two slides (containing 18 oocysts each) retained oocysts (one each). Coverslips from four slides (containing 18, 11, 11, and 12 DAPI-positive oocysts) were examined microscopically after their removal and found to contain no oocysts.
Slides containing the same number of oocysts are presented in the same row and are separated by semicolons. Each PCR result separated by a semicolon represents the analysis of the slide shown in the first column.
DNA released from lysed oocyst scrapings produced amplicons with 87.5% (42/48) of samples (Table 1). Two samples containing two intact oocysts each generated positive amplicons.
Our initial scraping method did not remove all oocysts from the slide well surface (Table 2). Small numbers of oocysts (one to three) were detected after scraping, typically in an unscraped area located on the upper right hand side of the test slide well, which was missed probably because the person processing the slide was right handed. To address this, the slide was rotated through 180° following the first set of two scrapings so that the undisturbed area was scraped, and then two further sets of scrapings were performed. Reexamination of 10 slides containing four (n = 3), five (n = 2), and six (n = 2) C. parvum (MD isolate) oocysts, by epifluorescence microscopy, revealedthat all oocysts were removed, and this procedure was adopted for all further analyses.
TABLE 2.
Examination of C. parvum oocystsa in seeded samples before and after sample removal and assessment of the effectiveness of the four PCR assays tested (N18SDIAG, N18SXIAO, STNCOWP and MAS-PCR) in detecting the presence of C. parvum (≤75 oocysts per sample) oocyst DNA
| No. of DAPI+ oocystsb (no. of empty oocysts) before scraping | Oocysts after scraping (score)c | PCR resultd
|
||
|---|---|---|---|---|
| N18SDIAG | STN-COWP | N18SXIAO | ||
| 50 (4) | 3 | 4+ | − | 3+ |
| 52e (4) | 2 | 3+ | 1+ | 2+ |
| 54e (5) | 0 | 4+ | FB | 3+ |
| 55e (5) | 1 | 4+ | 1+ | 3+ |
| 57e (11) | 2 | 4+ | FB | 3+ |
| 61 (5) | 0 | 2+ | 1+ | 2+ |
| 63 (8) | 3 | 4+ | − | 3+ |
| 63e (6) | 1 | 3+ | 2+ | 3+ |
| 74 (5) | 1 | 4+ | − | 3+ |
| 75 (5) | 1 | 2+ | 1+ | 2+ |
Moredun isolate (batch C2/03).
The number of sporozoite nuclei per oocyst before scraping was determined, and their distribution within this slide set was scored as follows: score of 1, 1 oocyst; score of 2, 10 oocysts; 3, 59 oocysts; 4, 475 oocysts.
No nuclei were seen in any of the oocysts remaining on wells after scraping.
All samples were analyzed by MAS-PCR, and the results were negative. −, negative PCR; FB, faint band, positive PCR but insufficient amplicon concentration for RFLP analysis; 1+, low-intensity bands with sufficient amplicon concentration for RFLP analysis, 2+, medium-intensity bands with sufficient amplicon concentration for RFLP analysis; 3+, high-intensity bands with sufficient amplicon concentration for RFLP analysis, 4+, very-high-intensity bands with sufficient amplicon concentration for RFLP analysis.
Oocysts whose DAPI-positive sporozoite nuclei (either one or two in each sample) were externalized.
Sensitivity of assays for amplification of C. parvum DNA from slides.
We used four loci (N18SDIAG, N18SXIAO, STN-COWP, and MAS-PCR [i.e., identified by multiplex allele-specific PCR]) to amplify DNA from 4 to 23 C. parvum oocysts whose DAPI-stained sporozoite nuclei had also been enumerated (Table 3). The numbers of sporozoite nuclei per oocyst were determined, and their distribution within this slide set was as follows: one for 1 oocyst, two for 16 oocysts, three for 108 oocysts, and four for 186 oocysts. Of these, N18SDIAG consistently generated the greatest number of positive results (Table 3), failing to generate an amplicon with 3 of the 30 samples tested. Amplicon intensities were strong (≥3+) in >96% of positive samples, and only with one sample (containing nine oocysts, one of which contained no DAPI-positive nuclei) was a faint band produced.
TABLE 3.
Assessment of the effectiveness of the four PCR assays tested (N18SDIAG, N18SXIAO, STN-COWP, and MAS-PCR) in detecting the presence of C. parvum (Moredun isolate C1/03) oocyst DNA (<25 oocysts per sample)
| No. of DAPI+ oocysts per slidea (no. of empty oocysts) | PCR resultb
|
|
|---|---|---|
| N18SDIAG | STN-COWP | |
| 4 | 4+ | − |
| 5; 5 (1); 5; 5 | −; 4+; 4+; 4+ | −; −; −; − |
| 6 (1); 6; 6 | 4+; 4+; 4+ | −; −; − |
| 7 (1); 7; 7 | 4+; 4+; 4+ | −; −; − |
| 8; 8 (1) | 4+; − | −; − |
| 9 (1); 9; 9; 9 (1) | 4+; 4+; 3+; FB | −; −; −; − |
| 10; 10 (2); 10 (1) | 4+; −; 4+ | −; −; 1+ |
| 13 | 4+ | −; − |
| 14 | 4+ | −; − |
| 17 (2) | 4+ | −; − |
| 18 | 4+ | −; − |
| 19 (4) | 3+ | −; − |
| 20 (1); 20 (1) | 4+; 3+ | −; 1+ |
| 21 | 3+ | −; − |
| 23 (2); 23 (1) | 3+; 4+ | −; − |
Slides containing the same number of oocysts are presented in the same row and are separated by semicolons.
All samples were tested with the N18SXIAO PCR and were negative. MAS-PCR was used to test samples from 10 slides (containing from 14 to 23 oocysts). All were negative. Slides containing the same number of oocysts are presented in the same row and are separated by semicolons. Each PCR result separated by a semicolon represents the analysis of the slide shown in the first column. −, negative PCR; FB, positive PCR but insufficient amplicon concentration for RFLP analysis; 1+, low-intensity bands with sufficient amplicon concentration for RFLP analysis; 2+, medium-intensity bands with sufficient amplicon concentration for RFLP analysis; 3+, high-intensity bands with sufficient amplicon concentration for RFLP analysis; 4+, very-high-intensity bands with sufficient amplicon concentration for RFLP analysis.
STN-COWP generated predominantly negative results. Previously, our detection limit with STN-COWP was ≥10 intact oocysts in suspension (data not shown). On retesting 10 samples (containing 13 to 23 oocysts), only 1 generated a visible amplicon (containing 20 oocysts), indicating that the outcome of the first testing was probably correct. Neither N18SXIAO nor MAS-PCR generated visible amplicons with these samples (Table 3). Only 10 samples were tested with MAS-PCR, because parallel testing with higher numbers of seeded C. parvum oocysts indicated that, even at a level of ∼50 oocysts per sample, MAS-PCR had difficulty in generating visible amplicons (Table 2).
Sensitivity of the assays in amplifying oocyst DNA from four Cryptosporidium species dried onto slides.
We prepared nine slides containing 2 to 24 oocysts of each of the four species tested (Table 4). Occasionally, individual oocysts contained <4 nuclei (Table 4). The numbers of sporozoite nuclei per oocyst were determined, and their distribution for C. felis was as follows: one for 2 oocysts, two for 5 oocysts, three for 10 oocysts, and four for 86 oocysts. For C. hominis, the numbers of nuclei were one for 2 oocysts, two for 9 oocysts, three for 22 oocysts, and four for 83 oocysts. For C. muris, this information is not available. For C. parvum, the numbers of nuclei were two for 1 oocyst, three for 7 oocysts, and four for 97 oocysts. C. muris oocysts exhibited poor rim fluorescence with variable intensities that were irregular in shape, with some collapsed walls and diffuse DAPI staining, making enumeration of DAPI-positive oocysts impossible. We extracted and amplified DNA from these oocysts, as no alternative source of C. muris oocysts was available.
TABLE 4.
Assessment of the effectiveness of three of the PCR assays tested (N18SDIAG, N18SXIAO, and STN-COWP) in detecting the presence of C. felis, C. hominis, C. muris, and C. parvum oocyst DNA in the samples testeda
| Species | No. of DAPI+ oocystsb (no. of empty oocysts) | PCR resultc
|
||
|---|---|---|---|---|
| N18SDIAG | STN-COWP | N18SXIAO | ||
| C. felis | 5; 5; 5 | 4+; 4+; 4+ | NA; NA; NA | −; −; − |
| 9; 9 | 4+; − | NA; NA | −; FB | |
| 11 | 1+ | NA | − | |
| 21 (1) | 4+ | NA | − | |
| 22 (3) | 4+ | NA | − | |
| 24 (4) | 4+ | NA | − | |
| C. hominis | 5; 5; 5 | 4+; 4+; 4+ | −; −; − | FB; −; FB |
| 6 (1); 6 (1) | 4+; 4+ | −; − | −; − | |
| 10; 10; 10 | 4+; 4+; 4+ | −; −; − | 1+; 1+; − | |
| 20; 20 | 4+; 4+ | 1+; FB | 4+; FB | |
| 21 | 4+ | 1+ | 2+ | |
| C. muris | 2; 2 | 1+; 2+ | NA; NA | −; − |
| 3; 3 | 1+; 1+ | NA; NA | −; − | |
| 5 | 3+ | NA | − | |
| 6 | − | NA | − | |
| 14 | 4+ | NA | 4+ | |
| C. parvum | 5; 5; 5 | −; −; − | −; −; − | −; −; − |
| 10; 10 | −; 4+ | −; − | −; − | |
| 11 | − | − | − | |
| 19 | − | − | − | |
| 20 (1) | − | − | − | |
| 21 | 4+ | − | 1+ | |
Oocyst numbers were ≤25 per sample.
Slides containing the same number of oocysts are presented in the same row and separated by semicolons.
Each PCR result separated by a semicolon represents the analysis of the slide shown in the first column. PCR results: −, negative PCR; FB, faint band, positive PCR but insufficient amplicon concentration for RFLP analysis; 1+, low-intensity bands with sufficient amplicon concentration for RFLP analysis; 2+, medium-intensity bands with sufficient amplicon concentration for RFLP analysis; 3+, high-intensity bands with sufficient amplicon concentration for RFLP analysis; 4+, very-high-intensity bands with sufficient amplicon concentration for RFLP analysis; NA, not available.
Most positives were generated with the N18SDIAG primers; the N18SXIAO primers performed less well, and the STN-COWP primers produced a few low-intensity amplicons (Table 4). N18SDIAG primers exhibited variability in maximizing amplicon intensity for different Cryptosporidium species. For example, C. muris amplicons were less intense than C. felis or C. hominis amplicons, and few (2/9) amplicons were generated with C. parvum samples.
Validation of PCR-RFLP for species identification and genotyping of oocysts from Cryptosporidium monitoring slides from United Kingdom water utilities.
Thirty-three slides were provided from four water companies/water utilities. AES slides were in an acceptable condition, and microscopic reevaluation was performed (Table 5). STL provided 10 slides in poor condition, making microscopic reevaluation impossible, and without accompanying information. Of the 10 samples tested, 9 produced amplicons with N18SDIAG while 6 produced amplicons with N18SXIAO (Table 5). SW slides (1 to 53 oocysts) produced visible amplicons with N18SDIAG with a satisfactory (3+ or 4+) intensity (Table 5), whereas seven produced visible amplicons with the N18SXIAO assay with sufficient amplicon intensity (3+ or 2+) for enzymatic digestion. Mixed species were detected in four slides with the N18SDIAG assay and in one slide with the N18SXIAO assay. Wessex Water Ltd. slides (5, 14, and 99 oocysts) produced amplicons with both assays that were sufficiently intense for digestion. Two of three slides contained larger (∼8 × 6 μm) oocysts. Both slides contained C. andersoni DNA by PCR-RFLP analysis (Table 5).
TABLE 5.
Identification of Cryptosporidium oocyst species directly from slides by nested PCR-RFLP analysis of two 18S rRNA gene loci
| Slide source | No. of DAPI+ oocysts (no. of empty oocysts) | PCR resultb
|
Cryptosporidium species identification by PCR-RFLPc
|
||
|---|---|---|---|---|---|
| N18SDIAG | N18SXIAO | N18SDIAG | N18SXIAO | ||
| AES | 1 | 3+ | 3+ | Mixed (C. muris or C. andersoni and faint band for C. parvum or C. hominis) | C. muris or C. andersoni |
| 2 (1) | 1+ | − | Mixed (C. meleagridis or Cryptosporidium sp. cervine, ferret, or mouse genotype and wild-type band at ∼350 bp) | NA | |
| 2 | 3+ | − | C. meleagridis or Cryptosporidium sp. cervine, ferret, or mouse genotype | NA | |
| 3 | 3+ | − | Wild-type pattern | NA | |
| 15 | 3+ | 3+ | Mixed (C. muris or C. andersoni and C. parvum or C. hominis) | Mixed (C. muris or C. andersoni and C. hominis) | |
| 49 (4) | 3+ | − | C. parvum or C. hominis | NA | |
| 80 (10) | 3+ | 3+ | Mixed (C. muris or C. andersoni and C. parvum or C. hominis) | Mixed (C. muris or C. andersoni and C. hominis) | |
| 130 (15) | 3+ | 3+ | Mixed (C. muris or C. andersoni and C. parvum or C. hominis) | Mixed (C. muris or C. andersoni and C. hominis) | |
| STL | NAd | 3+ | − | Mixed (C. muris or C. andersoni and C. parvum or C. hominis) | NA |
| NA | 3+ | 3+ | Mixed (C. muris or C . andersoni and major bands of ∼220 and 260 bp) | C. muris or C. andersoni | |
| NA | − | − | NA | NA/NA | |
| NA | 3+ | 3+ | Mixed (C. muris or C. andersoni and C. parvum or C. hominis) | Mixed (C. muris or C. andersoni and C. hominis) | |
| NA | 3+ | 3+ | Mixed (C. muris or C. andersoni and C. parvum or C. hominis) | Cryptosporidium sp. muskrat genotype II | |
| NA | 3+ | 3+ | C. parvum or C. hominis | C. hominis | |
| NA | 3+ | − | C. meleagridis or Cryptosporidium sp. cervine, ferret, or mouse genotype | NA | |
| NA | 3+ | 3+ | Mixed (C. muris or C. andersoni and C. parvum or C. hominis) | Mixed (C. muris or C. andersoni and C. parvum) | |
| NA | 1+ | − | Mixed (C. muris or C. andersoni and C. parvum or C. hominis) | NA | |
| NA | 3+ | 3+ | C. parvum or C. hominis | C. hominis | |
| SW | 1 | 3+ | 2+ | C. andersoni | C. andersoni |
| 2 | 4+ | 3+ | C. muris or C. andersoni | C. muris or C. andersoni | |
| 2 | 4+ | 3+ | NA | C. muris or C. andersoni | |
| 2 | 4+ | 2+ | NA | Mixed (C. muris or C. andersoni and C. parvum pattern) | |
| 3 | 3+ | − | C. parvum or C. hominis | NA | |
| 5 | 3+ | 2+ | C. andersoni | C. andersoni | |
| 5 | 4+ | 3+ | NA | C. muris or C. andersoni | |
| 5 | 4+ | 3+ | NA | C. muris or C. andersoni | |
| 22 (8) | 3+ | − | Mixed (wild-type band at ∼250 bp and C. meleagridis or Cryptosporidium sp. cervine, ferret, or mouse genotype) | NA | |
| 34 (6) | 3+ | − | Mixed (C. parvum and C. meleagridis or Cryptosporidium sp. cervine, ferret, or mouse genotype) | NA | |
| 50 (15) | 3+ | − | Mixed (wild-type band at ∼250 bp and C. meleagridis or Cryptosporidium sp. cervine, ferret, or mouse genotype) | NA | |
| 53 (12) | 3+ | − | Mixed (C. parvum and C. meleagridis or Cryptosporidium sp. cervine, ferret, or mouse genotype) | NA | |
| WWa | 5 (size, 8 × 6 mm) | 3+ | 2+ | C. andersoni | C. andersoni |
| 14 (size, 8 × 6 mm) | 3+ | 2+ | C. andersoni | C. muris or C. andersoni | |
| 99 (33) | 3+ | 2+ | Mixed (C. parvum pattern and C. meleagridis or Cryptosporidium sp. cervine, ferret, or mouse genotype) | Mixed (Cryptosporidium sp. cervine genotype and nonidentifiable wild-type genotype pattern) | |
WW, Northumberland and Wessex Water Ltd.
PCR results report the amplicon concentration determined by visual observation of band intensity on agarose gels and were graded as follows: −, negative; 1+, low-intensity bands with sufficient amplicon concentration for RFLP analysis; 2+, medium-intensity bands with sufficient amplicon concentration for RFLP analysis; 3+, high-intensity bands with sufficient amplicon concentration for RFLP analysis; 4+, very-high-intensity bands with sufficient amplicon concentration for RFLP analysis.
Species identification of C. muris or C. andersoni indicates further digestion with DdeI is required before final species identification; identification of C. meleagridis or Cryptosporidium sp. cervine, ferret, or mouse genotypes indicates further digestion with SpeI is required before final species identification; identification of C. parvum or C. hominis indicates sequencing of the amplicon is required before final species identification.
NA, not available.
Overall, the most common identifiable patterns observed with the N18SDIAG PCR-RFLP profile were of C. muris or C. andersoni (15 slides), C. parvum or C. hominis (or C. bovis, since RFLP patterns overlap) (16 slides), or C. meleagridis or the Cryptosporidium sp. cervine, ferret, or mouse genotype pattern (7 slides). With N18SDIAG assays, C. muris can be distinguished readily from C. andersoni by further digestion with DdeI. However, DNA sequencing is needed to distinguish between C. parvum, C. hominis, and C. bovis. The Cryptosporidium sp. cervine genotype amplicon can be further digested with SpeI so that it can be distinguished from C. meleagridis and ferret and mouse genotypes, which do not possess a restriction site for this enzyme. The N18SDIAG assay has less discriminatory power than the N18SXIAO assay with RFLP. All positive slides with the N18SXIAO assay generated PCR-RFLP patterns consistent with published data. Patterns conforming to C. muris or C. andersoni, to C. hominis, and to Cryptosporidium sp. muskrat genotype II previously described (32, 35, 37) were identified in these samples (Table 5). In the N18SXIAO assay, the C. muris amplicon can be distinguished readily from C. andersoni by further digestion with DdeI. C. hominis can be distinguished from C. parvum by digestion with restriction enzymes AseI and SspI; however, the recently described C. bovis RFLP pattern overlaps that of C. parvum.
In most cases the two PCR assays were in agreement concerning the RFLP patterns, with the exception of sample ST5, which appeared to have mixed patterns with the N18SDIAG RFLP profile and the Cryptosporidium sp. muskrat genotype II pattern with the N18SXIAO RFLP profile (37).
Five slides showed PCR-RFLP patterns with extra bands that might be derived from the mixture of species/genotypes present in the water concentrate (STL-2, AES-1, SW-1, and SW-2) or of new profiles from wild-type Cryptosporidium sp. (AES-6) that may correspond to novel species or genotypes. Unfortunately, these slides were positive only with the N18SDIAG assay, and comparison with the published profiles of the N18SXIAO RFLP assay (11, 32, 33, 35, 37) was not possible. Sequencing the N18SDIAG PCR product would be necessary for species identification in cases where a species/genotype RFLP profile overlaps that of another species/genotype; however, sequencing of amplicons generated from mixtures of species/genotypes is more technically challenging.
DISCUSSION
Currently, for many PCR assays, there is a distinct difference between laboratory and field data. An effort has been directed towards efficient extraction of Cryptosporidium DNA, selection of Cryptosporidium-specific primers, and overcoming the effects of inhibitors such as clays, humic and fulvic acids, polysaccharides and other organic compounds, salts, heavy metals, etc. (16, 17). Previously, we showed that oocyst staining with both FITC-C-mAb and DAPI is amenable to PCR amplification (17).
The current method combines an effective freeze-thawing treatment that optimizes disruption of the oocyst wall and the release of sporozoite DNA (16) with precise steps to remove coverslips and scrape the entire sample from a slide previously sealed and analyzed by microscopy for the presence of Cryptosporidium oocysts. We also validated oocyst removal from microscope slides, and we believe that the success of this scraping method is due to methanol fixing of oocysts on the wells of slides, which causes them to adhere to the glass well and places a layer of mounting medium between the sample and the sealed coverslip.
The predominantly negative results generated with STN-COWP, even after retesting, together with some negative results obtained with N18SXIAO, were unexpected, as previous experiences with STN-COWP and N18SXIAO indicated that the detection limit could be ≥10 intact oocysts. The results obtained with MAS-PCR also were disappointing, and parallel testing with higher numbers of seeded C. parvum oocysts indicated that, even at a level of ∼50 oocysts per sample, MAS-PCR had difficulty in generating visible amplicons (data not shown).
When STN-COWP and N18SXIAO were tested using higher oocyst numbers per slide, visible amplicons were produced, but not with all samples. While these results were in far closer agreement with our preliminary data, the STN-COWP was also considered to be of insufficient sensitivity for the detection of low numbers of oocysts on slides.
One reason why the N18SXIAO PCR was less sensitive than the N18SDIAG PCR may be that the amplicon generated with N18SXIAO (primary product, ∼1,325 bp; secondary product, 826 to 864 bp, depending on the Cryptosporidium isolate) is larger than that generated with N18SDIAG (primary product, 655 to 667 bp; secondary product, 425 to 455 bp), resulting in less efficient amplification in a primary PCR.
However, the N18SXIAO assay offers a higher discriminatory level from RFLP profiles than with the N18SDIAG assay, but the recent identification of new species and genotypes of epidemiological importance (4, 5), whose RFLP profiles overlap with other species, will limit the future use of PCR-RFLP assays for Cryptosporidium species and genotype identification. Here, sequencing will be the only option for identifying the Cryptosporidium species present on slides, and the amplicon defined by the CPB-DIAG primers is advantageous since it spans the hypervariable region of the gene.
In general, the results of positive PCR-RFLP patterns obtained with the N18SDIAG and N18SXIAO assays are comparable, but the N18SDIAG assay is more sensitive, as it identifies more positives. C. muris or C. andersoni, C. parvum or C. hominis, and C. meleagridis or Cryptosporidium sp. cervine, ferret, or mouse DNA were common findings in the Cryptosporidium monitoring slides that we analyzed (Table 5).
Microscopically, C. muris or C. andersoni can be differentiated from C. parvum by oocyst morphometry; however, some slides examined in this study did not appear to contain oocysts larger than 4 to 5 μm that could account for the RFLP result, indicating the presence of C. muris/C. andersoni with both nested 18S PCR assays. These results suggest that (i) amplifiable, naked Cryptosporidium DNA is present on the slides and (ii) oocysts may have been present but either were not recognized by DWI-FITC-C-mAb or were not identified. The latter explanation is supported by our observations that not all human-derived isolates of C. hominis and C. parvum are stained by commercially available monoclonal antibodies (R. Nichols and H. Smith, unpublished data). Furthermore, our experience of staining C. muris oocysts with DWI-FITC-C-mAb is that their fluorescence intensity is lower than that of C. parvum oocysts.
Incomplete information on the type of water sampled for this study hindered further conclusions regarding the species present in these waters, but we have embarked on a much larger study that should provide a better understanding of those Cryptosporidium species/genotypes present in Scottish waters.
These preliminary findings highlight the fact that many Cryptosporidium species/genotypes occur in the United Kingdom aquatic environment and that a sensitive molecular approach is required to determine the extent of environmental contamination with accepted and “novel” species/genotypes/subtypes of Cryptosporidium. Our data concur with those of others (35) in that C. hominis, C. parvum, C. canis, C. felis, C. andersoni, C. muris, and C. baileyi can be found in the aquatic environment.
A proportion (18%) of Cryptosporidium oocyst-positive slides tested failed to generate a PCR product at any of the genetic loci tested. These PCR-RFLP negatives occurred with both seeded oocyst samples and Cryptosporidium monitoring slides. Oocyst numbers on these slides were low (mean, 8.6; range, 1 to 20 oocysts) and variable, with varying numbers of DAPI-positive nuclei. The limit of sensitivity appears not to be based on oocyst number, as N18SDIAG was capable of generating amplicons from seeded sample slides containing only two oocysts. Such variability is of importance when developing standardized methods for small numbers of oocysts on Cryptosporidium monitoring slides. Currently, we are unable to identify specific reasons why this inconsistency arose; however, it does support our belief that it is imperative to use more than one 18S locus to determine the species/genotype of small numbers of Cryptosporidium oocysts present in water. Our PCR-RFLP and sequencing approaches for Cryptosporidium monitoring slides should permit the identification of species/genotypes in a significant proportion of samples; however, a minority will not be amenable to this approach.
Acknowledgments
This work was supported by the Drinking Water Inspectorate, United Kingdom (grant DWI-70/2/154).
We thank Amanda Walton and Denise Whitehead for excellent technical assistance. We gratefully acknowledge the following water companies/water utilities for contributing Cryptosporidium-positive Cryptosporidium monitoring slides for this study: Severn Trent Laboratories, Business Centre, Torrington Avenue, Coventry CV4 9GU, United Kingdom; Analytical Environmental Services, STL Business Centre, Torrington Avenue, Coventry CV4 9GU, United Kingdom; Scottish Water, Juniper House, Edinburgh EH14 4AP, Scotland, United Kingdom; and Wessex Water, Scientific Centre, Mead Lane, Saltford, Bristol BS31 3ER, United Kingdom.
REFERENCES
- 1.Bukhari, Z., and H. V. Smith. 1995. Effect of three concentration methods on viability of Cryptosporidium parvum oocysts recovered from bovine feces. J. Clin. Microbiol. 33:2592-2595. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Cacciò, S., E. Pinter, R. Fantini, I. Mezzaroma, and E. Pozio. 2002. Human infection with Cryptosporidium felis: case report and literature review. Emerg. Infect. Dis. 8:85-86. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Cacciò, S. M., R. C. A. Thompson, J. McLauchlin, and H. V. Smith. 2005. Unravelling Cryptosporidium and Giardia epidemiology. Trends Parasitol. 21:430-437. [DOI] [PubMed] [Google Scholar]
- 4.Fayer, R., M. Santin, M. Trout, and E. Greiner. 2006. Prevalence of species and genotypes of Cryptosporidium found in 1-2-year-old dairy cattle in the eastern United States. Vet. Parasitol. 135:105-112. [DOI] [PubMed] [Google Scholar]
- 5.Fayer, R., M. Santin, and L. Xiao. 2005. Cryptosporidium bovis n. sp. (Apicomplexa: Cryptosporidiidae) in cattle (Bos taurus). J. Parasitol. 91:624-629. [DOI] [PubMed] [Google Scholar]
- 6.Gatei, W., R. W. Ashford, N. J. Beeching, S. K. Kamwati, J. Greensill, and C. A. Hart. 2002. Cryptosporidium muris infection in an HIV-infected adult, Kenya. Emerg. Infect. Dis. 8:204-206. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Giles, M., D. C. Warhurst, K. A. Webster, D. M. West, and J. A. Marshall. 2002. A multiplex allele specific polymerase chain reaction (MAS-PCR) on the dihydrofolate reductase gene for the detection of Cryptosporidium parvum genotypes 1 and 2. Parasitology 125:35-44. [DOI] [PubMed] [Google Scholar]
- 8.Grimason, A. M., H. V. Smith, J. F. W. Parker, Z. Bukhari, A. T. Campbell, and L. J. Robertson. 1994. Application of DAPI and immunofluorescence for enhanced identification of Cryptosporidium spp. oocysts in water samples. Water Res. 28:733-736. [Google Scholar]
- 9.Guyot, K., A. Follet-Dumoulin, E. Lelievre, C. Sarfati, M. Rabodonirina, G. Nevez, J. C. Cailliez, D. Camus, and E. Dei-Cas. 2001. Molecular characterization of Cryptosporidium isolates obtained from humans in France. J. Clin. Microbiol. 39:3472-3480. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Homan, W., T. van Gorkom, Y. Y. Kan, and J. Hepener. 1999. Characterization of Cryptosporidium parvum in human and animal feces by single-tube nested polymerase chain reaction and restriction analysis. Parasitol. Res. 85:707-712. [DOI] [PubMed] [Google Scholar]
- 11.Jiang, J., K. A. Alderisio, and L. Xiao. 2005. Distribution of Cryptosporidium genotypes in storm event water samples from three watersheds in New York. Appl. Environ. Microbiol. 71:4446-4454. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Karanis, P., C. Kourenti, and H. V. Smith. Waterborne transmission of protozoan parasites: a worldwide review of outbreaks and lessons learnt. J. Water Health, in press. [DOI] [PubMed]
- 13.Katsumata, T., D. Hosea, I. G. Ranuh, S. Uga, T. Yanagi, and S. Kohono. 2000. Short report: possible Cryptosporidium muris infections in humans. Am. J. Trop. Med. Hyg. 62:70-72. [DOI] [PubMed] [Google Scholar]
- 14.Kistemann, T., T. Classen, C. Koch, F. Dangerdorf, R. Fischeder, J. Gebel, V. Vacata, and M. Exner. 2002. Microbial load of drinking water reservoir tributaries during extreme rainfall and runoff. Appl. Environ. Microbiol. 68:2188-2197. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Morgan, U. M., R. Weber, L. Xiao, I. Sulaiman, R. C. A. Thompson, W. Ndiritu, A. Lal, A. Moore, and P. Deplazes. 2000. Molecular characterization of Cryptosporidium isolates obtained from human immunodeficiency virus-infected individuals living in Switzerland, Kenya, and the United States. J. Clin. Microbiol. 38:1180-1183. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Nichols, R. A. B., and H. V. Smith. 2004. Optimisation of DNA extraction and molecular detection of Cryptosporidium oocysts in natural mineral water sources. J. Food Prot. 67:524-532. [DOI] [PubMed] [Google Scholar]
- 17.Nichols, R. A. B., B. M. Campbell, and H. V. Smith. 2003. Identification of Cryptosporidium spp. oocysts in UK noncarbonated natural mineral waters and drinking waters using a modified nested PCR-RFLP assay. Appl. Environ. Microbiol. 69:4183-4189. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Ong, C. S. L., D. L. Eisler, A. Alikhani, V. W. K. Fung, J. Tomblin, W. R. Bowie, and J. Isaac-Renton. 2002. Novel Cryptosporidium genotypes in sporadic cryptosporidiosis cases: first report of human infections with a cervine genotype. Emerg. Infect. Dis. 8:263-268. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Pedraza-Diaz, S., C. Amar, A. M. Iversen, P. J. Stanley, and J. McLauchlin. 2001. Unusual Cryptosporidium species recovered from human faeces: first description of Cryptosporidium felis and Cryptosporidium dog type from patients in England. J. Med. Microbiol. 50:293-296. [DOI] [PubMed] [Google Scholar]
- 20.Pedraza-Diaz, S., C. Amar, J. McLauchlin, G. L. Nichols, K. M. Cotton, P. Godwin, A. M. Iversen, L. Milne, J. R. Mulla, K. Nye, H. Panigrahl, S. R. Venn, R. Wiggins, M. Williams, and E. R. Youngs. 2001. Cryptosporidium meleagridis from humans: molecular analysis and description of affected patients. J. Infect. 42:243-250. [DOI] [PubMed] [Google Scholar]
- 21.Pieniazek, N. J., F. J. Bornay-Llinares, S. B. Slemenda, A. J. daSilva, I. N. S. Moura, M. J. Arrowood, O. Ditrich, and D. G. Addiss. 1999. New Cryptosporidium genotypes in HIV-infected persons. Emerg. Infect. Dis. 5:444-449. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Reynolds, D. T., R. B. Slade, N. J. Sykes, A. Jonas, and C. R. Fricker. 1999. Detection of Cryptosporidium oocysts in water: techniques for generating precise recovery data. J. Appl. Microbiol. 87:804-813. [DOI] [PubMed] [Google Scholar]
- 23.Smith, H. V., and J. B. Rose. 1990. Waterborne cryptosporidiosis. Parasitol. Today 6:8-12. [DOI] [PubMed] [Google Scholar]
- 24.Smith, H. V., and J. B. Rose. 1998. Waterborne cryptosporidiosis: current status. Parasitol. Today 14:14-22. [DOI] [PubMed] [Google Scholar]
- 25.Smith, H. V., R. A. Gilmour, P. G. Smith, G. M. Morris, A. M. Grimason, J. F. W. Parker, M. H. Jackson, and R. W. A. Girdwood. 1989. A modified method for the detection of Cryptosporidium spp. oocysts in water related samples. Commun. Dis. Scotland 15:7-13. [Google Scholar]
- 26.Smith, H. V., L. J. Robertson, and J. E. Ongerth. 1995. Cryptosporidiosis and giardiasis, the impact of waterborne transmission. J. Water Supply Res. Technol. AQUA 44:258-274. [Google Scholar]
- 27.Smith, H. V., B. M. Campbell, C. A. Paton, and R. A. B. Nichols. 2002. Significance of enhanced morphological detection of Cryptosporidium sp. oocysts in water concentrates determined by using 4′,6′-diamidino-2-phenylindole and immunofluorescence microscopy. Appl. Environ. Microbiol. 68:5198-5201. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.United Kingdom Drinking Water Inspectorate. 1999. Water supply (water quality) (amendment) regulations 1999, SI no. 1524. U.K. Drinking Water Inspectorate, London, England.
- 29.United Kingdom Drinking Water Inspectorate. 2000. The water supply (water quality) (amendment) regulations 1999. Cryptosporidium in water supplies: laboratory analysis, part 2, revision 3, Nov. 2000. Protocol containing standard operating protocols (SOPs) for monitoring Cryptosporidium oocysts in water supplies. DWI information letter 26/2000. U.K. Drinking Water Inspectorate, London, England. [Online.] www.dwi.gov.uk.
- 30.United States Environmental Protection Agency. 2001. Method 1622: Cryptosporidium in water by filtration/IMS/FA. EPA 821-R01-026. U.S. EPA Office of Water, Washington, D.C.
- 31.United States Environmental Protection Agency. 2001. Method 1623: Giardia and Cryptosporidium in water by filtration/IMS/FA. EPA 821-R01-025. U.S. EPA Office of Water, Washington, D.C.
- 32.Xiao, L., U. M. Morgan, J. Limor, L. Escalante, M. Arrowwood, W. Shulaw, R. C. A. Thompson, R. Fayer, and A. A. Lal. 1999. Genetic diversity within Cryptosporidium parvum and related Cryptosporidium species. Appl. Environ. Microbiol. 65:3386-3391. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Xiao, L., K. Alderisio, J. Limor, M. Royer, and A. A. Lal. 2000. Identification of species and sources of Cryptosporidium oocysts in storm waters with a small-subunit rRNA-based diagnostic and genotyping tool. Appl. Environ. Microbiol. 66:5492-5498. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Xiao, L., C. Bern, J. Limor, I. Sulaiman, J. Roberts, W. Checkley, L. Cabrera, R. H. Gilman, and A. A. Lal. 2001. Identification of 5 types of Cryptosporidium parasites in children in Lima, Peru. J. Infect. Dis. 183:492-497. [DOI] [PubMed] [Google Scholar]
- 35.Xiao, L., A. Sing, J. Limor, T. Grazyck, S. Gradus, and A. A. Lal. 2001. Molecular characterization of Cryptosporidium oocysts in samples of raw surface water and wastewater. Appl. Environ. Microbiol. 67:1097-1101. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Xiao, L., R. Fayer, U. Ryan, and S. J. Upton. 2004. Cryptosporidium taxonomy: recent advances and implications for public health. Clin. Microbiol. Rev. 17:72-97. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Zhou, L., R. Fayer, J. M. Trout, U. M. Ryan, F. W. Schaefer III, and L. Xiao. 2004. Genotypes of Cryptosporidium species infecting fur-bearing mammals differ from those of species infecting humans. Appl. Environ. Microbiol. 70:7574-7577. [DOI] [PMC free article] [PubMed] [Google Scholar]