| Front page | | Contents | | Previous | | Next |

Optimisation and validation of a method for identification of Cyptosporidium and Giardia in drinking water

Summary

The purpose of this project was to optimise and validate a method for concentration, purification, and subsequent identification of the protozoans Cryptosporidium and Giardia in raw water and clean water, respectively.

Interest in knowing the status of the content in water of the protozoans is partly based on the description of extensive waterborne epidemics caused by Cryptosporidium and Giardia in drinking water. Hence, increasing interest has arisen in the development and optimisation of methods to demonstrate these zoonotic parasites in different materials, including raw water and drinking water.

Cryptosporidium parvum and Giardia lamblia (syn. G. intestinalis and G. duodenalis) are protozoans, which can infect the gastro-intestinal tract of both man and animals. The parasites are ubiquitous in nature and among the most common non-bacterial causes of diarrhoea.

Cryptosporidia are excreted with the faeces as oocysts in great numbers. The oocysts are immediately infective and furthermore resilient towards all commonly employed disinfectants, including chlorine in the concentrations used for treatment of bathing and drinking water. In a cool, damp environment the oocysts may survive for several months. Infection occurs by direct contact (person-person, animal-human) or indirectly via ingestion of oocysts from e.g. drinking water, bathing water and foods contaminated with faecal material.

Giardiasis in humans is caused by the species G. lamblia, which, similar to C. parvum, has a very wide host spectrum. The lifecycle encompasses two stages, the trophozoite, which is the infective form, and the cyst stage, which is the resistant form, or resting stage. The cyst stage is excreted with faeces and can survive for up to 2 months outside the host.

For both protozoans, only few (oo) cysts are enough to cause illness.

In this study a method has been optimised and validated to demonstrate Cryptosporidium oocysts and Giardia cysts. The method has been validated through determination of the recovery rate by adding (spiking) a known number of oocysts and cysts to water samples. Six water samples of different quality were employed for the validation; three clean water samples from the waterworks at Kingosvej, Mørkskov, and Regnemark, plus three raw water samples from Kingosvej and Mørkskov waterworks as well as Ejby raw water drilling.

Ten filtrations, each of 10 litres, have been performed with each of the following cyst/oocyst numbers added: 100, 1,000, and 10,000. Following every tenth filtration, a positive control (5 l pure water with addition of the same number of cysts/oocyst as the previous water sample, 100, 1,000 or 10,000) and a negative control (5 l pure water) were carried out.

The water samples were filtered through a 2µm membrane filter, concentrated by centrifugation and immunomagnetic separation (IMS), and finally stained and read by immunofluorescence.

In the course of the project, the methods for concentration and detection of Cryptosporidium and Giardia in water have been optimised at various stages. The rubber tubes were changed in the filtration process to ensure a protozoan-free status. When liberating the (oo)cysts from the filter in the 'pulsifier' employed, the volume of water as well as the general procedure were adjusted to minimize loss of (oo)cysts. The concentration of the filtrate was performed at an adjusted centrifugation configuration, and the process of microscopy was adjusted according to the sample content of (oo)cysts.

As part of the validation it was examined whether there would be an interaction between Giardia and Cryptosporidium, such that, e.g., a high concentration of one might influence the recovery of the other. These experiments were carried out using tap water with addition of 10,000 of one protozoan and 100 of the other, and vice versa. For each combination 5 filtrations were performed. Additionally, the results of the positive controls from experiments with 10,000 (oo)cysts of each protozoan added were compared with the interaction experiments.

Studies using raw water showed less than expected recovery of Cryptosporidium oocysts. To verify this, the oocysts were captured by the filter, and for some samples a 50µl sub-sample was extracted after filtration and before IMS. These sub-samples were stained for immunofluorescence microscopy according to the manufacturer's directions. In all cases, oocysts could be demonstrated in the sample.

The recovery rate of (oo)cysts was not influenced negatively by a high concentration of one or both protozoans and was not significantly different at the different spiking levels of Giardia and Cryptosporidium, respectively. This observation enabled an evaluation of the recovery rate for each protozoan in the water types examined, without using a correcting factor for any interaction.

In order to calculate the recovery rate for the two protozoans in the various water types, the spiking dose was checked by counting three sub-samples. The results of the dose counts varied considerably, but were not significantly different between the two protozoans.

The recovery of Giardia in raw water varied between 39.3-77.3%, with an average of 54.6-62.3%, and with greatest variation in samples spiked with low numbers. The staining of the Giardia cysts was not satisfactory in any of the raw water samples.

The recovery rate for Giardia cysts in clean water varied between 26.8-63.4% with an average of 41.5-52.8%. The variation in recovery rate was greatest in samples with low spiking dose. Only for samples from Kingosvej with 1,000 and 10,000 cysts added, respectively, was recovery significantly lower (P < 0.05) than that of the other clean water types, which did not differ. If the much lower Kingosvej clean water results were omitted, the recovery rate would be 48.9-59.4 %.

The recovery of Cryptosporidium in raw water varied between 0.1-6.0%, with an average of 0.2-3.2%. The variation of the recovery rate was greatest in samples with 100 oocysts added.

The recovery of Cryptosporidium oocysts from clean water samples was within an interval of 25-51.1%, with an average of 37.0-38.7%. The greatest variation was seen in samples with 00 oocysts added, whereas the standard deviation of samples spiked with 10,000 oocysts was less.

As is described above, we found that for both raw water and clean water the recovery of Giardia was greater than that of Cryptosporidium. This may be due to the greater size of the Giardia cysts improving their capture on the membrane filter. Furthermore, the producer of the IMS kit, Dynal Biotech, states that the binding between Giardia and Dynabeads anti-Giardia is stronger than the binding between Cryptosporidium and Dynabeads anti-Cryptosporidium. This could be the reason that many Giardia cysts are recovered from raw water but no, or only few, Cryptosporidium oocysts.

It is a well-known problem that Cryptosporidium oocysts as well as Giardia cysts have a tendency to clump together and sediment, which complicates a precise spiking of the samples. Generally it must be assumed that the large standard deviation of the recovery rates of the protozoans, not least in samples with 100 (oo)cysts added, is caused by both a variation in the actual spiking dose and a methodology-related (filtration, IMS, fluorescence labelling) variation.

The chemical and biological constitution of the raw water types is generally not very different from that of the clean water. However, there is a considerable amount of iron, especially in Ejby raw water (3.7mg/l), and even in Kingosvej raw water (0.05-0.09mg/l), which contains the least iron of the three raw water types, the iron content is higher than in any of the clean water types. In all the raw water types, the content of ammonium is markedly higher than in the clean water samples. Additionally it appears that the raw water contains more bicarbonate and sulphur oxide than the clean water, although this data is not available for all the water types. In contrast to Giardia, the recovery of Cryptosporidium in raw water was practically non-existent. The cryptosporidia were demonstrated in sub-samples of filtration and hence we concluded that the IMS process did not function adequately in raw water. This could be related to the amount of iron in the raw water, or the significant amount of sediment which was characteristic of the raw water, even though the sediment amount did not exceed the IMS process directions. The binding between the Cryptosporidium oocysts and Dynabeads anti-Cryptosporidium is not strong and it is conceivable that compounds in the sediment further weaken the binding.

No lower detection limit has been determined for the method, however, based on the recovery of Cryptosporidium oocysts of at least 25%, it is expected that the method may detect 2-3 oocysts in samples with only 10 oocysts. For Giardia the method can detect 4-6 cysts in samples containing only 10 cysts. This validation process has determined no upper detection limit, but it is likely to depend on the IMS process.

| Front page | | Contents | | Previous | | Next | | Top |