Impact de trois types de cartouches de filtration sur le rendement de la méthode de détection quantitative de Cryptosporidium et Giardia

Impact of three types of cartridges on the recovery of the detection methode of Cryptosporidium and Giardia

Anjou-Recherche. Graal - Département de Recherche en Microbiologie 109-111, rue des Côtes - 78600 Maisons-Lafitte

Résumé

La prévention des épidémies parasitaires de type hydrique nécessite le développement d'outils analytiques appropriés à la quantification précise des parasites viables véhiculés dans les eaux destinées à la potabilisation. La méthode de concentration et de clarification de Cryptosporidium et Giardia recommandée par l'EPA (USA) nous a donné des rendements de récupération de Cryptosporidium faibles et variables (0,17 ± 0,13 %, C. V = 76 %) lorsque des cartouches en polypropylène bobiné et des échantillons d'eau de rivière artificiellement contaminés sont utilisés. Le protocole a été modifié, (centrifugation à 2000 g au lieu de 1050 g) et les rendements de récupération de Cryptosporidium ont été établis en utilisant de cartouches de filtration en polypropylène bobiné, enroulé et en polyéthersulfonate ainsi que des eaux artificiellement contaminées. Lorsque des cartouches en polypropylène bobiné et des cartouches constituées de membranes en polypropylène enroulées sont utilisées, les rendements du protocole modifié restent faibles et variables. Pour les cartouches bobinées ils ont été de 2,0 à 30 % avec un coefficient de variation de 153% et pour les cartouches à membranes enroulées de 0,4 à 39 % avec un coefficient de variation de 101 %. L'utilisation des cartouches en polyéthersulfonate, a permis d'obtenir des rendements plus reproductibles et généralement plus élevés: 12 à 37 % de rendement avec un meilleur coefficient de variation de 36,4 %.

Abstract

Preventing the outbreak of waterborne epidemies requires the development of analytical methods for the precise quantification of viable Cryptosporidium and Giardia carriers in water used for domestic consumption.

The method used most often for the detection of waterborne parasites is that of LeChevallier et al. (1991) which is that recommended by USA-EPA. This method allows the enumeration, by microscopy, of these parasites. However, this technique does not provide information on the viability, infectivity, pathogenicity nor the proportion of human pathogenic species, C. parvum and G. intestinalis. Moreover, the variability observed on the levels of contamination in river water described by several authors (0.001 to 500 Cryptosporidium/l and 0.01 to 70 Giardia/l) and on the percentages of positive samples can be attributed to the varying water quality and to the unreliability of the method used. In fact, with LeChevalier method's, some authors have found high Cryptosporidium recoveries 42 to 80% [Rose. 1988; LeChevallier et al. 1991], and yet others have found low recoveries < 1 to 10 % [Nieminski et al., 1995; Fricker et al., 1994; Watkins et al. 1994, Vesey et Slade.1991].

This method involves the filtration of large quantities of water through a wound polypropylene cartridge, elution of parasites and clarification using a 1.1 specific-gravity Percoll-sucrose density. The parasites are then labeled using the immunofluorescence approach [Rose et al., 1988; LeChevallier et al.1991, Drozd et al., 1993].

In our laboratory, slight (0.02 to 0.4%) and variable (coefficient of variation = 76%) Cryptosporidium recoveries have been obtained for 10 spiked river water samples using this method. So, we have been evaluating the recoveries of the principal steps for this protocol the results of which show that the suggested centrifugation speeds are inappropriate to sufficiently compact the pellet, thus reducing the loss of Cryptosporidium. After evaluating the results of the modified protocol (centrifugation at 2000g. instead of at 1050g) the recoveries show an overall increase of 8 ± 12.3%. Unfortunately, the variability of the results has also increased (coefficient of variation = 153%).

Consequently, the recoveries obtained using two others types of cartridge, rolled polypropylene and polyethersulfone cartridge, were also evaluated. When rolled polypropylene cartridges were used, the recoveries from the modified protocol remained low and variable, the range was from 0.4 to 39% and coefficient of variation was 101%. The use of polyethersulfone cartridges, which have been recently marked, has improved the recoveries by 12 to 37% with a coefficient of variation of 36.4%. In conclusion, the polyethersulfone cartridges seem to be more appropriate to monitor the contamination levels of river waters than the polypropylene cartridges.