OUP user menu

An immunomagnetic separation polymerase chain reaction assay for rapid and ultra-sensitive detection of Cryptosporidium parvum in drinking water

Sylvie Hallier-Soulier , Emmanuelle Guillot
DOI: http://dx.doi.org/10.1111/j.1574-6968.1999.tb13674.x 285-289 First published online: 1 July 1999


A sensitive and rapid method was developed to detect Cryptosporidium parvum oocysts in drinking water. This molecular assay combined immunomagnetic separation with polymerase chain reaction amplification to detect very low levels of C. parvum oocysts. Magnetic beads coated with anti-cryptosporidium were used to capture oocysts directly from drinking water membrane filter concentrates, at the same time removing polymerase chain reaction inhibitory substances. The DNA was then extracted by the freeze-boil Chelex-100 treatment, followed by polymerase chain reaction. The immunomagnetic separation-polymerase chain reaction product was identified by non-radioactive hybridization using an internal oligonucleotide probe labelled with digoxigenin. This immunomagnetic separation-polymerase chain reaction assay can detect the presence of a single seeded oocyst in 5–100-l samples of drinking water, thereby assuring the absence of C. parvum contamination in the sample under analysis.

  • Immunomagnetic separation-polymerase chain reaction
  • Drinking water
  • Cryptosporidium parvum

1 Introduction

Cryptosporidium parvum is a coccidian protozoan identified as a significant water-borne pathogen responsible for several serious outbreaks of illness worldwide in the past 10 years. The parasite can cause gastrointestinal illness in humans, particularly in immuno-compromised individuals [1]. It can be transmitted through water in the oocyst form, which is resistant to many environmental stresses and water treatments [2]. Since a small number of C. parvum oocysts has often been found in the environment and the number of oocysts required to cause infection is relatively low [3], a highly sensitive method is required for the detection of C. parvum oocysts in water samples. Furthermore, a variety of host species (mammals, birds, aquatic vertebrates, humans) may be the source of Cryptosporidium oocysts found in water. Since only C. parvum is recognized as a human pathogen, the species-specific identification of oocysts is critical.

The standard methods [4,5] for the detection of C. parvum in environmental samples are time-consuming, inefficient and characterized by a low sensitivity, so they can never guarantee the absence of Cryptosporidium oocysts in drinking water.

Genetic methods based on the detection of Cryptosporidium nucleic acid by the polymerase chain reaction (PCR) have been developed [69]. However, PCR-based diagnostic techniques applied directly to concentrated water samples may be inhibited by components present in the concentrates. Immunomagnetic separation (IMS) has been reported as an efficient pre-PCR step for separation and isolation of specific cells from crude samples with simultaneous elimination of specific Taq DNA polymerase inhibitors [10].

In this study, we developed a molecular method for the sensitive and rapid detection of C. parvum oocysts from drinking water samples. The method combines an IMS of the oocysts from water concentrates with a PCR amplification of a C. parvum-specific DNA sequence.

2 Materials and methods

2.1 Oocyst stock

Cryptosporidium oocysts were obtained after purification from the Unité des Maladies à Protozoaires (INRA, Tours, France) from the faeces of naturally infected calves. Oocysts were supplied as a purified suspension in 2.5% potassium dichromate and stored at 4°C. The oocyst concentration was determined by a 10-fold dilution assay and immunofluorescence with the Cryptosporidium/Giardia cell test IF (Cellabs, Sydney, Australia).

2.2 DNA extraction from purified oocysts

Approximately 107 oocysts were suspended in lysis buffer (1% sodium dodecyl sulfate in 10 mM Tris, 1 mM EDTA buffer (TE), pH 8) and then disrupted by thermal shocks (95°C for 4 min, −15°C for 15 min). The mixture was incubated for 1 h at 37°C with 0.3 mg ml−1 proteinase K and then for 30 min at 65°C with 1% cetyl trimethyl ammonium bromide. DNA was extracted twice by phenol:chloroform:isoamyl alcohol (24:24:1) and isopropanol-precipitated and then resuspended in 50 µl TE buffer (pH 8). The concentration and quality of the DNA were estimated by measuring the absorbance ratio at 260 and 280 nm and by agarose gel electrophoresis.

2.3 IMS of oocysts from drinking water samples

Tap water samples (5, 20 or 100 l), previously dechlorinated by adding 20 mg l−1 of sodium thiosulfate, were spiked with a range of oocyst dilutions from 1 to 104. Dilution series of oocysts used in spiking experiments were microscopically enumerated in triplicate by an immunofluorescence assay. 20- and 100-l seeded samples were concentrated by filtration through yarn-wound Envirocheck capsules (Gelman, Ann Arbor, MI, USA) at a flow rate of 2 l min−1. Oocysts were eluted from the capsule with 240 ml of a detergent elution buffer (8 g NaCl, 0.2 g KH2PO4, 2.9 g Na2HPO4·12H2O, 0.2 g KCl, 0.1 g sodium dodecyl sulfate, 3 µl Tween 80, 150 µl Sigma Antifoam B, 1 l deionized H2O). Oocysts contained in the eluate were concentrated by filtration through a polycarbonate membrane filter (25-mm diameter, 1.2-µm pore size). 5-l samples were filtered directly through a polycarbonate membrane filter (25-mm diameter, 0.8-µm pore size). In both cases, the filter was placed into a 50-ml polycarbonate tube containing 5 ml distilled water and vigorously mixed for 5 min at high speed. The filter was then discarded and 50 µl anti-Cryptosporidium Dynabeads (DYNAL, Compiègne, France) was added. The samples were then treated as recommended by the manufacturer. At the end of the immunomagnetic capture, the Dynabeads were resuspended in 10 µl distilled water.

Genomic DNA from the separated oocyst suspension (10 µl) was released in the presence of 25% (w/v) Chelex 100 (Bio-Rad, Hercules, CA, USA) by five cycles of freezing and thawing (−80°C for 2 min, 95°C for 2 min). Following centrifugation for 1 min at 13 000×g, the DNA in the supernatant was used directly as the template in PCR experiments.

2.4 PCR assay

A set of primers previously described by Balatbat et al. [11], based on the sequence of unknown genomic regions reported by Laxer et al. [6], was used in this study. The forward primer LaxA (5′-GCG AAG ATG ACC TTT TGA TTT G-3′) and the reverse primer LaxB (5′-AGG ATT TCT TCT TCT GAG GTT CC-3′), synthesized by Genset (Paris, France), amplified a 210-bp fragment specific to C. parvum. Amplification reaction mixtures contained 2.5 U AmpliTaq (Life Technologies, Gaithersburg, MD, USA), 1×PCR buffer (20 mM Tris-HCl pH 8.4, 50 mM KCl), 1 µM of each forward and reverse primer, 1.5–7 mM MgCl2, 200 µM of each dNTPs and 10 µg ml−1 bovine serum albumin, in a 50-µl volume. PCR was performed in a model 9600 DNA thermal cycler (Perkin-Elmer) with the following temperature cycle: 94°C for 5 min followed by 40 cycles of denaturation at 94°C for 30 s, annealing at 60°C for 45 s and elongation at 72°C for 1 min. Negative controls in which oocyst or DNA was replaced by sterile distilled water were included during the IMS step and PCR amplification. Positive controls (50 ng of C. parvum purified DNA) were run in parallel to avoid false negative results from PCR inhibitors.

2.5 Detection of amplified products

The amplified products (10 µl) were visualized and photographed under UV light after electrophoresis in a 2% agarose gel (Life Technologies, Gaithersburg, MD, USA) containing 0.5 mg ml−1 ethidium bromide. A 1-kbp size marker was included (Life Technologies, Gaithersburg, MD, USA).

To confirm the specificity of the PCR assay, dot-blot hybridization with an internal probe was performed. The sequence of the internal probe selected by using OLIGO version 5 (National Biosciences, Plymouth, MN, USA) was 5′-TGG CCA ATG ATG AAT TAA CC-3′. The oligonucleotide probe was labelled with digoxigenin-11-dUTP at the 3′-end with the 3′-End labelling kit (Boehringer, Mannheim, Germany). Dot-blot hybridization and chemiluminescent detection were performed as specified by the manufacturer (Boehringer, Mannheim, Germany). Briefly, an aliquot (4 µl) of the amplified DNA was denatured by boiling for 10 min, then immobilized onto a nylon membrane (Nytran-Plus, Schleicher and Schuell, Dassel, Germany). Hybridization incubation was carried out overnight at 42°C. An X-ray film (Du Pont NEN Research Products, Boston, MA, USA) was exposed to the membrane for 30 min at room temperature before developing.

3 Results and discussion

For the PCR step of the IMS-PCR assay developed in this study, a set of primers previously designed by Balatbat et al. [11] in a nested PCR test was used in 40 cycles of a single PCR to avoid the added risk of carry-over contamination posed by nested PCR. The optimum MgCl2 concentration and annealing temperature were empirically determined. The yield of amplification product obtained was not increased when the MgCl2 concentrations were raised to 7 mM. The MgCl2 concentration finally used was 2 mM. The optimal PCR specificity was obtained when the annealing temperature was 60°C, whereas the annealing temperature of 45°C reported by Balatbat [11] was less specific. The specificity of the product was confirmed by dot-blot hybridization with an internal digoxigenin-labelled probe. The sensitivity of the PCR assay was evaluated by performing an amplification of a serial 10-fold dilution of purified DNA from C. parvum from 1 ng (i.e. ~105 genomes) to 10 fg (i.e. ~one genome). In our experiment, as little as 10 fg of C. parvum DNA (equivalent to one parasite) could be detected by agarose gel electrophoresis and dot-blot hybridization (data not shown).

To determine the number of C. parvum oocysts detectable by our IMS-PCR assay, the sensitivity of the method was investigated using 1–104 oocysts. For PCR amplification from purified oocysts, addition of Chelex-100 was required prior to the lysis step by thermal shocks. By protecting the extracted DNA from degradation at a high temperature, Chelex-100 allowed a higher yield of amplification products. However, raising the Chelex-100 concentration from 15 to 50% (w/v) did not increase the PCR yield. Thus, an intermediar Chelex-100 concentration of 25% (w/v) was used. Under these experimental conditions, the sensitivity of the IMS-PCR assay was very high with as little as one oocyst giving a clear signal detectable by agarose gel electrophoresis and dot-blot hybridization (data not shown).

To evaluate the sensitivity of the IMS-PCR test on drinking water samples, dechlorinated tap water samples of 5, 20 or 100 l were spiked in triplicate with 10-fold dilutions of oocysts ranging from 1 to 104 cells. Spiked oocysts were concentrated by filtration of the tap water sample through yarn-wound filters and/or polycarbonate membrane filters, depending on the water sample volume. The lower limit of detection was 1 oocyst on agarose gel electrophoresis and dot-blot hybridization in the three different volume samples of 5, 20 and 100 l (Fig. 1A and B). These results demonstrated the efficiency of the oocyst concentration and capture steps from tap water samples with no loss of oocysts occurring during filtration or IMS and no inhibitory substances present in the PCR mixture.

Figure 1

Sensitivity of the IMS-PCR assay specific for C. parvum from 100-l tap water samples. (A) Amplification of a 210-bp fragment visualized by agarose gel electrophoresis. (B) Dot-blot of the PCR products with the internal digoxigenin-labelled probe. Lane M: DNA size marker. Lane 1: PCR positive control. Lanes 2–6: IMS-PCR assay from 100-l tap water sample spiked with 104, 103, 102, 10 and one C. parvum oocysts. Lane 7: Negative control without C. parvum oocysts. Lane 8: PCR negative control without template DNA.

In addition to its high sensitivity, this IMS-PCR assay is rapid, providing results in less than 24 h. By comparison, the recommended methods for Cryptosporidium analysis [4] include concentration of large volumes of water (100–500 l) using cartridge filters, separation from background debris by flotation on a Percoll-sucrose gradient and then analyzing a proportion of the final concentrate by immunofluorescence detection [12]. This is tedious, time-consuming (at least 48 h) and shows poor recovery (30–40%) [13]. Clarification of samples by Percoll-sucrose density gradient centrifugation contributes significantly to losses of Cryptosporidium oocysts, compromising the sensitivity of any subsequent detection method. IMS overcomes these limitations, as sample clarification is not required. In addition, magnetic beads are simple to use, do not require expensive equipment and can effectively separate and concentrate target microorganisms from inhibitory substances. The expert opinion is that the absence of Cryptosporidium oocysts in drinking water can never be guaranteed using the recommended methods for oocyst isolation and enumeration [14]. In this study, the IMS-PCR assay specific for C. parvum allowed detection down to one oocyst in 5–100-l samples of drinking water in 7 h. This represents a significant improvement in the assay speed and sensitivity and could therefore be applied to the routine monitoring of C. parvum contamination in drinking water.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
View Abstract