In Vitro Cryptosporidium Parvum: Pilot-Scale Evaluation of UV Reactors'e Cacy Against Infectivity of Oocysts
In Vitro Cryptosporidium Parvum: Pilot-Scale Evaluation of UV Reactors'e Cacy Against Infectivity of Oocysts
In Vitro Cryptosporidium Parvum: Pilot-Scale Evaluation of UV Reactors'e Cacy Against Infectivity of Oocysts
pital Saint Louis, Centre for Environmental Analysis of Veolia Environment, Saint Maurice, France; 2Laboratoire de Parasitologie-Mycologie, Ho pitaux de Paris, Paris, France; 3Laboratoire de Parasitologie-Mycologie, Faculte de Medecine Assistance Publique Ho Denis Diderot, Paris, France; and 4 Anjou Recherche-Research Center on Water of Veolia Environment, Maisons-Laftte, France
Correspondence: Yves J.F. Garin, Laboratoire de Parasitologie-Mycologie, de Medecine, Paris VII Denis Faculte Universite Diderot, 16 rue Henri Huchard, 75018 Paris, France. Tel.: 133 1 42 49 95 03; fax: 133 1 42 49 48 03; e-mail: [email protected] Present address: Emilio Entrala, Industrias Espadafor, Avda. Andalucia s/n, Granada 18015, Spain. e-mail: [email protected] Received 29 March 2007; revised 16 August 2007; accepted 18 August 2007. First published online November 2007. DOI:10.1111/j.1574-695X.2007.00335.x Editor: Kai Man Kam Keywords Cryptosporidiosis; Cryptosporidium ; water purication; UV reactors; evaluation.
Abstract An experimental protocol was developed to assess the efcacy of two UV reactors (medium-pressure UVasters, and a low-pressure reactor) on the infectivity of Cryptosporidium parvum oocysts under conditions mimicking small- or mediumsize water distribution units. The protocol included purication of large amounts of viable oocysts from experimentally infected calf feces, pilot spiking, sample concentration and purication after UV radiation, oocyst quantication and in vitro evaluation of oocyst infectivity on HCT-8 cells. Water samples were collected at intervals upstream and downstream from the UV reactor after spiking. Oocysts were concentrated by centrifugation, puried by immunomagnetic capture and quantied using laser-scanning cytometry. An enhanced in vitro infectivity test on HCT-8 cells was developed, where oocysts were pretreated in order to obtain maximized in vitro infectivity, and infectious foci were enumerated after immunouorescence staining after 3 days of culture. This method was superior to viability measured by excystation for assessing oocyst infectivity. The infectivity rate of untreated oocysts ranged between 9% and 30% in replicate experiments. The method allowed us to determine inactivation rates 4 4.92 (log) with UVasters and 4 4.82 with the LP reactor after exposition of oocysts to an effective dose of 400 J m2 at ow rates of 15 and 42 m3 h1, respectively. This dose is required to achieve a 4-log inactivation of a wide range of waterborne pathogens, some of which, like viruses, are particularly resistant to UV light. Computational uid dynamics provide a reliable assessment of ow and dose delivery through the UV reactor, but only assessment of the efcacy on microorganisms can provide the denitive result. Validation tests for Cryptosporidium disinfection are based on spiking experiments performed with puried oocysts obtained from experimentally infected animals. Mouse or in vitro cell culture inoculation, based on most probable number (MPN) tests, can be used to evaluate the infectivity of oocysts after disinfection (Slifko et al., 1999, 2002; Rochelle et al., 2002). The present work describes the protocol used for the validation of two UV reactors (one medium-pressure and one low-pressure reactor) for tap water disinfection in France, including: (1) a method for the preparation of the
2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved
Introduction
The emergence of waterborne outbreaks in the last 30 years due to chlorine-resistant protozoa, and the increasing information on chlorine-derived oxidation byproducts, has increased interest from the water industry in UV-disinfection technologies (Erickson & Ortega, 2006). The sensitivity of Cryptosporidium oocysts to UV light has been demonstrated using a collimated beam apparatus, batch studies and laboratory-scale pilots. Inactivation rates ranging from 90% to 99.99% were demonstrated for UV doses ranging from 30 to 250 J m2 (Lorenzo-Lorenzo et al., 1993; Linden et al., 2001; Rochelle et al., 2005; Hijnen et al., 2006). Medium- and low-pressure UV radiations have also been found to be efcient against Cryptosporidium hominis oocysts (Johnson et al., 2005). A UV dose of 400 J m2 is recommended for industrial NORM, 2001). reactors for optimal tap water disinfection (O
FEMS Immunol Med Microbiol 51 (2007) 555561
556
E. Entrala et al.
spiking suspensions, (2) a method for determining oocyst concentration after reactor inoculation based on concentration by centrifugation and immunomagnetic separation and (3) an enhanced in vitro evaluation of Cryptosporidium parvum oocyst infectivity. The present results show that this protocol allows us to quantify accurately oocyst inactivation in nished water up to nearly 5-log units under conditions that fully resemble UV reactors working conditions in water production units.
1000 g for 10 min at 4 1C, and then the supernatant was removed by aspiration. After two washes with Tris-EDTA buffer (Tris base 50 mM, EDTA 10 mM, no pH adjustment needed) (Sigma-Aldrich, Saint Quentin Fallavier, France), the pellet was poured onto a diphasic KBr gradient [7 and 3 mL, 16% and 6% (w/v) KBr in Tris-EDTA, respectively] and centrifuged at 3000 g for 1 h at 4 1C. The puried oocysts were recovered from the gradient interphase by aspiration, diluted 1/3 in distilled water, centrifuged at 3000 g, 10 min at 4 1C, and then washed twice with Hanks balanced salt solution (HBSS). Oocysts were quantied at this stage using a hematocytometer chamber (Kovaslide, VWR, Fontenay sous Bois, France). Oocyst viability was estimated by measuring the percentage of excystation (Robertson et al., 1993). One hundred microliters of the undiluted suspension of oocysts was added to 900 mL of HBSS without Ca21 and Mg21 salts (Sigma-Aldrich), acidied to pH 2.7 (AHBSS) and incubated for 30 min at 37 1C. The suspension was then centrifuged for 5 min at 9000 g, 950 mL of the supernatant was replaced with an equal volume of HBSS supplemented with 0.75% (w/v) sodium taurocholate and 0.1% sodium bicarbonate (Sigma-Aldrich). After a 2-h incubation at 37 1C, oocysts were examined for excystation under a phasecontrast microscope at a 100 magnication, and the percentage of excysted oocysts was recorded.
Experimental design
Two sets of experiments were conducted on a similar design. The rst set consisted of two repeated experiments with the medium-pressure UV reactor (UVasters) and the second consisted in one experiment with the LP reactor. Experiments were realized in a pilot water circuit of 80200-mm diameter water pipes between a 3-m3 water tank lled with tap water, a water pump providing a 1545 m3 h1 ow rate, one injection pump (0.189 m3 h1 maximum ow rate), a UV reactor and a 5-m3 tank for collecting waste water. For water sampling, two water taps located upstream and downstream from the reactor were available (Fig. 1). The water used for the spiking experiments was tap water (surface water treated conventionally to reach the criteria of the French health authorities concerning drinking water). The municipal treatment included occulation and ltration. Before distribution, chlorine was included at a concentration of 0.3 mg L1. The tap water was collected several days before the experiment with the UV lamp of the reactor in closed loop and stored for at least 24 h in order to destroy residual chlorine and ensure that the water was free from any other disinfectant. Each experiment comprised the following sequential steps: (1) purication of oocysts and assessment of viability through an excystation test,
FEMS Immunol Med Microbiol 51 (2007) 555561
Parasites
Two batches of C. parvum oocysts were used. Both were prepared from experimentally infected calf feces provided by Prof. M. Naciri (Institut National de Recherche Agronomique, Tours, France). Feces had been collected 4 months (batch 1) or 1 month (batch 2) before the experiments, and had been kept at 4 1C until use. Oocysts were puried from the diluted feces by centrifugation in a potassium bromide (KBr) gradient as described previously (Entrala et al., 2000). Ten milliliters of the feces suspension was added to 15 mL of diethyl ether. After a vigorous agitation, the suspension was centrifuged (GR412 Jouan, Saint-Herblain, France) at
c
2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved
557
the automatic ChemScans analyzer coupled to a uorescence microscope as described previously (de Roubin et al., 2002). In experiment 2, this procedure was performed on 2 L of each UV-exposed sample.
Fig. 1. Diagram of the equipment used for experiments. DS, doping suspension; IN, OUT, sample taps; IP, injection pump; R, reactor; SM, static mixer; WP, water pump; WT, water tank.
(2) preparation of a 10-L (Experiments 1, 2) or 20-L (Experiment 3) oocyst spike solution in distilled water (Milli-Q, Millipore); the oocyst concentration was adjusted according to the water pump and injection pump ow rates in order to provide a concentration of 104105 oocysts L1 into the reactor during 5 min, (3) checking the water ow rates of the water pump and the injection pump, (4) collect one water sample from the water tank, (5) checking the UVlamp intensity in the reactor, (6) starting of the injection of the parasite suspension, (7) stabilization of the water ow rate inside the reactor for 30 s, (8) collection of water samples at 30-s intervals upstream and downstream from the reactor. Each water sample (2 L) was collected in a sterile plastic can, and immediately processed for oocyst concentration, enumeration and cell culture.
2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved
558
E. Entrala et al.
(2 g40 mL60 mL). Cultures were examined using a uorescence microscope (Leitz DMBR, Leika, Rueil Malmaison, France) at 200 or 400 magnication. The total number of uorescent parasitic foci in each culture was recorded.
Results
The starting conditions of the three experiments are summarized in Table 1. The ow rates and irradiance values were checked continuously during the trials using internal meters. Under the conditions of the tests (UV transmission at 254 nm over a 1-cm path length greater than 97%), the ow rates of the reactors were 42 m3 h1 for the low-pressure reactor and 15 m3 h1 for the medium-pressure reactor. During the experiments with LP and UVasters reactors, variations of ow rates ranged between 1% and 6%, respectively. For both sets of tests, irradiance was stable during the trials. Residual chlorine was 0.06, 0.01 and 0.05 mg L1 in experiments 1, 2 and 3, respectively. For all experiments, the excystation rate of puried oocysts was 4 90%. For each experiment, spiking batches of oocysts were prepared and their concentration was checked just before inoculation to the water circuit. According to the inoculation pump and water pump ow rates, a dilution factor of the oocysts was calculated.
All scheduled water samples were collected between 1 and 4 min after the inoculation pump was started, i.e. a time allowing the oocyst ow to stabilize in the circuit and largely before the end of oocysts inoculation. Two-liter samples were collected. In experiments 1 and 3, a 1-L aliquot of each sample was processed for oocyst counts and culture. In experiment 2, 2 L of UV-exposed samples were processed in two separate 1-L aliquots. The mean oocyst recovery rates for all water samples in experiments 1, 2 and 3 were 82%, 89% and 95% after centrifugation and 76%, 66% and 74% after IMS concentration, respectively. These recovery rates account for the lower oocyst counts in samples compared with the theoretical value calculated from the water pump ow rates (Table 1). Taking into account the parasite counts after centrifugation and concentration, oocyst suspensions were diluted before inoculation to cultures. Samples that have been exposed to UV (EG) or had not been challenged by oocysts were inoculated undiluted as no infection, or a very low rate of infection, was expected. According to the experiment, four to 16 replicate wells were inoculated with 50 mL of the parasite suspension containing 920026 167 oocysts. Samples that were collected upstream from the reactors (AC) were diluted before inoculation to culture. A dilution of 1/1000 was selected to obtain nonconuent parasitic foci in cultures, i.e. o 40 foci well1. Ten microliters of each oocyst dilution were inoculated into four to eight replicate wells. The results of cell cultures from different water samples are presented in Table 2. No parasite foci and no cellular toxicity were noted in cultures inoculated with unspiked water. In cultures inoculated with oocysts that had not been exposed to UV (i.e. collected upstream from the reactor), 1032 nonconuent parasitic foci per culture well were observed and could be easily enumerated in Labtek II chambers. The coefcients of variation of the number of parasitic focus in replicate cultures were 27%, 20% and 30% in experiments 1, 2 and 3, respectively. From these counts, the mean infection rates were estimated at 8.45% and 7.64% for the rst batch of oocysts used in experiments 1 and 2 (Table 2), and 37.09% for the batch used in experiment 3
Table 1. Experimental conditions for the assessment of UV reactors activities against Cryptosporidium parvum oocysts Stock spike solution Water pump Injection pump Dilution No. of ow rate (m3 h1) ow rate (m3 h1) factor Volume (L) oocysts 0.189 0.189 0.120 83.5 80.9 350 20 20 10 5.8 108 4.3 108 Oocysts concentration (L1) in the circuit Excystation Calculated (%) ( 105) 92 90 3.45 2.68 0.76 Measured ( 105) 2.32 0.14 1.52 0.16 0.53 0.03 Ratio (%) 67 57 69
UV reactor
2.67 108 4 95
Mean values after centrifugation and IMS determined by Chemscan analysis of three samples collected upstream from UV reactors.
2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved
559
Table 2. Cell culture assessment of the inactivation rates of UVasters on Cryptosporidium oocysts Upstream to UV reactors (samples AC) Oocysts L1 ( 105)z Experiment 1 A B C A B C 2.33 2.45 2.17 2.18 2.16 1.75 Inoculated oocyst/well (no. of wells) 233 (6) 245 (6) 217 (6) 187 (8) 212 (8) 172 (8) Downstream to UV reactors (samples EG)w Oocysts L1 ( 105)z E F G E1 E2 F1 F2 G1 G2 3.42 3.24 3.17 Total: 1.84 2.61 2.24 2.21 2.29 2.06 Total: Inoculated oocyst/well (no. of wells) 13 700 (16) 12 970 (16) 12 700 (16) 629 920 9200 (16) 26 167 (8) 11 217 (16) 22 133 (8) 11 483 (16) 20 600 (8) 1 077 600
Foci/well (mean SD) 19.7 3.9 19.0 6.8 19.8 5.8 Mean: 12.7 3.4 16.2 1.9 14.5 2.6
Infection rate (%) 8.44 7.76 9.14 8.45 6.82 7.67 8.43
Log decrease Z4.27 Z4.24 Z4.23 Z4.73 Z4.05 Z4.20 Z4.17 Z4.13 Z4.15 Z4.10 Z4.92
Experiment 2
Mean:
One liter part of the collected 2-L sample.
w z
7.64
Two separate 1-L parts of the sample (named E1, E2. . .). Determined by Chemscan analysis of three aliquots from each water sample.
(Table 3). These values were used to calculate the UV inactivation rates in each experiment. In cultures inoculated with oocysts that had been exposed to UV through the UVasters or the LP reactor, parasite foci were not seen. In most cultures, uorescent oocysts appeared to be lying on the monolayers but with no possible confusion with growing parasites. From these results, the inactivation rates of the reactors were calculated as follows:
public distributed water. Approximately 60% of public water systems are of a very small size, serving o 500 people (Miquel, 2003). Because of this considerable diversity in water resources and distribution, water industries and public communities need to propose adapted disinfecting processes. These processes have to be in agreement with the French regulation, which requires that the drinking water provided should
Inactivation rate Number of inoculated UV exposed oocysts mean infection rate of oocysts Log Number of parasitic foci in culture 1 Because replicate cultures showed a limited variability between the number of parasitic foci in water samples taken upstream or downstream from UV reactors, data from all the water samples taken downstream from the reactors were merged to calculate the inactivation rate of the reactors. Because no parasitic foci were seen after UV exposure, a 11 correction had to be included in the equation. Thus, the calculated value was the minimum estimated rate of inactivation. This rate was Z4.73 and Z4.92 in experiments 1 and 2 with UVasters, and Z4.82 with the LP reactor, respectively. not contain microorganisms at a concentration that might be a risk for public health (Anon, 2001). Although this regulation does not dene a specic requirement for monitoring Cryptosporidium in water resources or in nished water (except when a fecal contamination is suspected), water producers must take into account the potential risk of contamination of water resources by Cryptosporidium to dene the water treatment process. Several water processes can be used and combined according to the resource but only a few of them are efcient in removing or inactivating Cryptosporidium oocysts (Betancourt & Rose, 2004). Two treatments are very efcient towards Cryptosporidium: membrane ltration (ultra-ltration, nanoltration, reverse osmosis) and UV. Depending on the objectives of the nal treatment (reduction of microbiology, of organic matter or softening of hard waters. . .), technico-economical studies can identify the most advantageous treatment.
c
Discussion
There are c. 27 000 public water systems and more than 36 000 water catchments in France. Among these, 95% collect water from underground resources but the 5% collecting surface water resources nally provide 37% of
FEMS Immunol Med Microbiol 51 (2007) 555561
2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved
560
E. Entrala et al.
Table 3. Cell culture assessment of the inactivation rates of LP reactor on Cryptosporidium oocysts Upstream from UV reactors (samples AC) Inoculated oocyst/well (no. of wells) 54 (5) 55 (6) 49 (6) Downstream from UV reactors (samples EG)w
Foci/well (mean SD) 22.8 5.9 16.0 3.3 20.0 6.8 Mean:
Inoculated oocyst/well (no. of wells) 12 650 (4) 12 650 (4) 18 950 (4) 177 000
Determined by Chemscan analysis of three samples. Determined by Chemscan analysis of three aliquots from each water sample.
UV light possesses several advantages compared with other disinfectants, such as a good efcacy over a wide range of microorganisms in a short time of exposure and the absence of by-products. Several bench-scale studies showed that a relatively low UV dose of medium-pressure or lowpressure UV irradiation is able to inactivate C. parvum oocysts, and that despite a capacity of DNA repair (Shin et al., 2001), UV-exposed oocysts are not capable of recovering their infectivity (Rochelle et al., 2005). In the present study, assessment of infectivity was performed by cell culture. A cell culture method was used combined with immunouorescence using a rat polyclonal antibody and quantication of secondary parasite clusters as representative of infective oocysts rather than evaluation by the MPN assay (Slifko et al., 1997). In a way, this method is comparable to counting of CFU, as used for bacteria. The results show that this method, previously used in benchscale experiments (Bukhari & LeChevallier, 2003; Bukhari et al., 2004), was appropriate and reproducible for testing a large number of samples and replicates in a large-scale study. As already noted by other authors, a marked difference in oocyst infectivity to cell cultures according to the parasite batch was observed (Rochelle et al., 2002; Slikfo et al., 2002; Schets et al., 2005). The two batches were from the same source and the same isolate but the rst one was prepared 4 months after stool collection, whereas the second one was prepared o 1 month after collection. For both batches, attention was focused on purifying viable oocysts using the KBr gradient methods, and the excystation rate of puried oocysts was always 4 90%. Despite this, the infectivity in cell culture was low when older oocysts were used, pointing out the importance of using fresh oocysts for cell culture experiments. This also conrms that oocyst viability, assessed by excystation, is not a good indicator of infectivity (Bukhari et al., 2000). Because of this variability in infectivity, and because of the high ow rate through the reactors, a large inoculum size
c
had to be used and samples and cultures had to be repeated in order to be able to estimate the highest inactivation rate. The reproducibility of culture results allowed pooling of data from repeated water samples. It was nally demonstrated that UV lamps at a uence of 400 J m2 resulted in a minimum inactivation rate of 4.92 log with the UVasters reactor provided at a 15 m3 h1 water ow rate and at least 4.82 log with the LP reactor provided at a 42 m3 h1 water ow rate. These results conrm the remarkable efcacy of both polychromatic medium-pressure and monochromatic low-pressure UV lamps. It should be stressed that the uence was set at 400 J m2 because other infectious agents such as viruses require this uence for inactivation and is usually applied for tap water disinfection. The results show that this condition is more than sufcient for the inactivation of Cryptosporidium. Such experimental conditions are very close to those of many small- or medium-size water production units, and the results obtained fulll the requirement of water safety, including the recently effective US Long Term 2 Enhanced Surface Water Treatment Rule (LT2ESWTR) (US Environmental Protection Agency, 2006). The different ranges of power for either low-pressure lamps or medium-pressure lamps allow the use of the UV technologies for a wide range of ow rates with a minimum occupied volume. Because UV systems are compact and easy to install and require very little maintenance, they are well adapted for various water distribution networks. As both reactors proved to be equally efcient in the present study, the choice between low- and medium-pressure lamps relies only on economical considerations on the balance between investment/operational costs, medium-pressure reactors being more powerful (lower investment) while low-pressure reactors are much more efcient (lower electrical consumption). Moreover, in case of accidental parasite contamination of a resource, an additional UV treatment can be installed on
FEMS Immunol Med Microbiol 51 (2007) 555561
2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved
561
the decient water supply system, as it was the case with UVasters during the Cryptosporidium waterborne outbreak in Divonne les Bains (Cellule Interr egionale dEpidemiolo ne-Alpes-Auvergne, 2003). gie Rho The present results show that in the present experimental protocol, UV reactors are remarkably efcient in inactivating Cryptosporidium oocysts, one of the most resistant contaminating microorganisms in drinking water, offering a cost-effective and environmentally friendly means of controlling this waterborne disease.
References
` la consommation Anon (2001) D ecret relatif aux eaux destin ees a ` lexclusion des eaux min humaine, a erales naturelles. Journal Ofciel de la R epublique Franc aise, Lois et D ecrets. 20 D ecembre 2001, No. 1220, pp. 2038120399. Betancourt WQ & Rose JB (2004) Drinking water treatment processes for removal of Cryptosporidium and Giardia. Vet Parasitol 126: 219234. Bukhari Z & LeChevallier M (2003) Assessing UV reactor performance for treatment of nished water. Water Sci Technol 47: 179184. Bukhari Z, Marshall MM, Korich DG, Fricker CR, Smith HV, Rosen J & Clancy JL (2000) Comparison of Cryptosporidium parvum viability and infectivity assays following ozone treatment of oocysts. Appl Environ Microbiol 66: 29722980. Bukhari Z, Abrams F & LeChevallier M (2004) Using ultraviolet light for disinfection of nished water. Water Sci Technol 50: 173178. ne-Alpes-Auvergne Cellule Interr egionale dEpidemiologie Rho ` la pollution du r (2003) Epid emie de gastro-ent erite li ee a eseau de distribution deau potable de la commune de Divonneles-Bains, Ain (01). Aout Septembre 2003. http://rhone-alpes. sante.gouv.fr/sante/cire/acrobat/divonne.pdf. de Roubin MR, Pharamond JS, Zanelli F, Poty F, Houdart S, Laurent F, Drocourt JL & Van Poucke S (2002) Application of laser scanning cytometry followed by epiuorescent and differential interference contrast microscopy for the detection and enumeration of Cryptosporidium and Giardia in raw and potable waters. J Appl Microbiol 93: 599607. Entrala E, Molina-Molina J, Rosales-Lombardo M, SanchezMoreno M & Mascaro-Lazcano C (2000) Cryptosporidium parvum: oocysts purication using potassium bromide discontinuous gradient. Vet Parasitol 92: 223226. Erickson MC & Ortega YR (2006) Inactivation of protozoan parasites in food, water, and environmental systems. J Food Prot 69: 27862808. Hijnen WA, Beerendonk EF & Medema GJ (2006) Inactivation credit of UV radiation for viruses, bacteria and protozoan (oo)cysts in water: a review. Water Res 40: 322. Johnson AM, Linden K, Ciociola KM, de Leon R, Widmer G & Rochelle PA (2005) UV inactivation of Cryptosporidium
hominis as measured in cell culture. Appl Environ Microbiol 71: 28002802. Linden KG, Shin G & Sobsey MD (2001) Comparative effectiveness of UV wavelengths for the inactivation of Cryptosporidium parvum oocysts in water. Water Sci Technol 43: 171174. Lorenzo-Lorenzo MJ, Ares-Mazas ME, Villacorta-Martinez de Maturana I & Duran-Oreiro D (1993) Effect of ultra-violet disinfection of drinking water on the viability of Cryptosporidium parvum oocysts. J Parasitol 79: 6770. Miquel G (2003) Rapport sur la qualit e de leau et de lassainissement en France. Ofce parlementaire d evaluation des choix scientiques et technologiques. No. 215, p.293. http://www.senat.fr/rap/l02-215-2/l02-215-21.pdf. NORM (2001) O NORM M 5873-1. Plants for the disinfection O of water using ultraviolet radiation Requirements and testing sterreichisches Part 1: low pressure mercury lamp plants. O Normungsinstitut, Vienna. NORM (2003) O NORM M 5873-2. Plants for the disinfection O of water using ultraviolet radiation Requirements and testing Part 2: medium pressure mercury lamp plants. sterreichisches Normungsinstitut, Vienna. O Robertson LJ, Campbell AT & Smith HV (1993) In vitro excystation of Cryptosporidium parvum. Parasitol 106: 1319. Rochelle PA, Marshall MM, Mead JR, Johnson AM, Korich DG, Rosen JS & de Leon R (2002) Comparison of in vitro cell culture and a mouse assay for measuring infectivity of Cryptosporidium parvum. Appl Environ Microbiol 68: 38093817. Rochelle PA, Upton SJ, Montelone BA & Woods K (2005) The response of Cryptosporidium parvum to UV light. Trends Parasitol 21: 8187. Schets FM, Engels GB, During M & de Roda Husman AM (2005) Detection of infectious Cryptosporidium oocysts by cell culture immunouorescence assay: applicability to environmental samples. Appl Environ Microbiol 71: 67936798. Shin GA, Linden KG, Arrowood MJ & Sobsey MD (2001) Lowpressure UV inactivation and DNA repair potential of Cryptosporidium parvum oocysts. Appl Environ Microbiol 67: 30293032. Slifko TR, Friedman D, Rose JB & Jakubowski W (1997) An in vitro method for detecting infectious Cryptosporidium oocysts with cell culture. Appl Environ Microbiol 63: 36693675. Slifko TR, Huffman DE & Rose JB (1999) A most-probable-assay for enumeration of infectious Cryptosporidium parvum oocysts. Appl Environ Microbiol 65: 39363941. Slikfo TR, Huffman DE, Dussert D, Owens JH, Jakubowski W, Haas CN & Rose J (2002) Comparison of tissue culture and animal models for assessment of Cryptosporidium parvum infection. Exp Parasitol 101: 97106. US Environmental Protection Agency (2006) National primary drinking water regulations: long term 2 enhanced surface water treatment rule. Fed Regist 71: 654786.
2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved