Detection of Viable Oocysts of Cryptosporidium parvum Following

Anal. Chem. 2001, 73, 1176-1180

Detection of Viable Oocysts of Cryptosporidium parvum Following Nucleic Acid Sequence Based Amplification Antje J. Baeumner,* Michele C. Humiston, Richard A. Montagna,† and Richard A. Durst

Bioanalytical Research Group, Department of Food Science and Technology, Cornell University, Geneva, New York 14456-0462

A reliable method using nucleic acid sequence based amplification (NASBA) with subsequent electrochemiluminescent detection for the specific and sensitive detection of viable oocysts of Cryptosporidium parvum in environmental samples was developed. The target molecule was a 121-nt sequence from the C. parvum heat shock protein hsp70 mRNA. Oocysts of C. parvum were isolated from environmental water via vortex flow filtration and immunomagnetic separation. A brief heat shock was applied to the oocysts and the nucleic acid purified using an optimized very simple but efficient nucleic acid extraction method. The nucleic acid was amplified in a water bath for 60-90 min with NASBA, an isothermal technique that specifically amplifies RNA molecules. Amplified RNA was hybridized with specific DNA probes and quantified with an electrochemiluminescence (ECL) detection system. We optimized the nucleic acid extraction and purification, the NASBA reaction, amplification, and detection probes. We were able to amplify and detect as few as 10 mRNA molecules. The NASBA primers as well as the ECL probes were highly specific for C. parvum in buffer and in environmental samples. Our detection limit was ∼5 viable oocysts/sample for the assay procedure, including nucleic acid extraction, NASBA, and ECL detection. Nonviable oocysts were not detected. Cryptosporidium parvum is a significant waterborne pathogen that has been responsible for multiple outbreaks of cryptosporidiosis throughout the world, with the largest outbreak recorded to date occurring in Milwaukee in 1993 with over 400 000 people infected and over 100 deaths.1 Point prevalence surveys indicate that over 50% of surface waters in the United Kingdom can be contaminated with Cryptosporidium. In continental United States, more than 80% of the surface water supplies and 26% of the treated drinking waters have been shown to be contaminated with Cryptosporidium. From the at least six species of the genus Cryptosporidium, C. parvum is the major species responsible for * To whom correspondence should be addressed. Department of Agricultural and Biological Engineering, Cornell University, 318 Riley-Robb Hall, Ithaca, NY, 14853. E-mail: [email protected]. Fax: 607-255-4080. † Present address: Innovative Biotechnologies International, Inc., 335 Lang Boulevard, Grand Island, NY, 14072. (1) Wilkinson, S. L. Chem. Eng. News 1997, 75 (Nov 10), 24-33.

1176 Analytical Chemistry, Vol. 73, No. 6, March 15, 2001

clinical disease in humans and domestic mammals.2 The parasite is transmitted through water, and one of its life cycle stages includes an oocyst form that is resistant to the commonly used water treatment methods (i.e., physical water treatment processes and disinfection). Although the number of oocysts found in environmental waters is generally very low, the minimum infectious capture material is suggested to be between 1 and 132 oocysts. Thus, sensitive and effective diagnostic tools are required. Current methods used to detect the presence of C. parvum in public water treatment systems rely upon EPA-approved methods (methods 1622 and 1623) that are labor-intensive, time-consuming, and very often unable to effectively distinguish viable from nonviable oocysts. Methods 1622 and 1623 require filtration, immunomagnetic separation of the oocysts from the material capture, and immunofluorescence assay for determination of oocyst concentrations, with confirmation through vital dye staining and differential interference contrast microscopy. In addition, improper performance of current immunofluorescence assays by water facility analysts have led to false positive results with subsequent unnecessary “Boil Water Alerts” in at least one instance.3 Molecular probes have recently come to the scene using a fluorescent in situ hybridization (FISH) detecting ribosomal RNA (rRNA). However, it is questionable that rRNA is a good indicator of viability of an organism. It could indicate viability of oocysts from the environment, but it is not useful for measuring viability after disinfection.4 Another method for assessing viability is in vitro cell culture. However, this is not a method that could be routinely or easily performed in a water treatment facility’s laboratory. Messenger RNA (mRNA) is present only in viable organisms since its lifetime is short, often only a few seconds to minutes. Therefore, mRNA is an optimal target molecule if viable organisms need to be distinguished from nonviable organisms. However, not all mRNAs are produced at any time in the life of an organism. Thus, we have chosen a heat shock protein mRNA (hsp70) from C. parvum as our target sequence.5 On applying a heat shock to the oocysts, hsp70 mRNA is produced only by viable organisms. (2) Smith, H. V.; Rose, J. B. Parasitol. Today 1998, 14, 14-22. (3) Clancy, J. L. J.-Am. Water Works Assoc. 2000, 92, 55-66. (4) Erickson, B. Anal. Chem. 1998, 70, 49A-52A. (5) Khramtsov, N. V.; Tilley, M.; Blunt, D. S.; Montelone, B. A.; Upton, S. J. J. Eur. Microbiol. 1995, 42, 416-422. 10.1021/ac001293h CCC: $20.00

© 2001 American Chemical Society Published on Web 02/17/2001

Figure 1. The NASBA amplification pathway.

The stimulation of hsp70 mRNA production is necessary to ensure that target molecules are present at any of the oocyst’s life stages. Additionally, it leads to an amplification of the target molecule at the molecular level and will allow the detection of fewer organisms. In this publication, we describe the development of a highly sensitive and specific detection method for viable oocysts of C. parvum in environmental samples that uses a simple oocyst of C. parvum disruption method, nucleic acid sequence based amplification (NASBA) amplification, and ECL detection techniques. NASBA has several advantages over PCR and RT-PCR: it exclusively amplifies RNA, it is rapid, and it is isothermal.6-8 Thus, no specialized equipment other than a heating block is required which makes NASBA the better amplification tool. A flow diagram describing NASBA is shown in Figure 1. Previously, excystation assays, prior to nucleic acid extraction and PCR amplification, have been used in order to prove viability of the organism.9 Also, it has been proven difficult to differentiate viable from nonviable oocysts of C. parvum using RT-PCR since genomic DNA interferes with the amplification10 and several precautions have to be taken in an RT-PCR assay in order to detect only viable organisms. However, applying NASBA as the molecular amplification system, only RNA will be detected due to NASBA’s exclusive amplification of RNA. Thus, NASBA is the better-suited amplification method for the detection of viable organisms. Electrochemiluminescence (ECL) technology has been shown earlier to be a reliable and sensitive hybridization technique.11 After only 30 min of hybridization at elevated temperatures, up to 50 samples can be analyzed with ∼1 min analysis time per sample Thus, we developed a reliable and simple oocyst disruption method, optimized the nucleic acid purification, and optimized amplification and detection probes. We tested the specificity of our NASBA and ECL procedures for C. parvum. Finally, we applied the assay system for the detection of viable oocysts of C. parvum in environmental samples of clean and dirty water. (6) van der Vliet, G. M. E.; Schukkink, R. A. F.; van Gemen, B.; Scepters, P.; Cloister, P. R. J. Gen. Microbiol. 1993, 139, 2423-2429. (7) Compton, J. Nature 1991, 350, 91-92. (8) van Gemen, B.; van Beuningen, R.; Nabbe, A.; van Strijp, D.; Jriaans, S.; Lens, P.; Kievits, T. J. Virol. Methods 1994, 49, 157-168. (9) Deng, M. Q.; Cliver, D. O.; Mariam, T. W. Appl. Environ. Microbiol. 1997, 63, 3134-3438. (10) Stinear, T.; Matusan, A.; Hines, K.; Sandery, M. Appl. Environ. Microbiol. 1996, 62, 3385-3390. (11) van Gemen, B.; van Beuningen, R.; Nabbe, A.; van Strijp, D.; Jriaans, S.; Lens, P.; Kievits, T. J. Virol. Methods 1994, 49, 157-168.

MATERIALS AND METHODS Reagents. NASBA reagents and ECL buffer and cleaning solutions were provided by Organon Teknika (OT), Boxtel, The Netherlands. OT labeled DNA probes for the ECL reaction with ruthenium and magnetic beads, respectively. The NucliSense Reader for the analysis of NASBA reactions hybridized with ECL probes was kindly provided by OT. Oocysts of C. parvum were obtained from J. B. Rose’s laboratory, St. Petersburg, FL, and Dave Kelsey, ImmuCell, Portland, ME. ImmuCell Co. provided realwater samples spiked with low oocyst of C. parvum numbers. Certified oocyst of C. parvum samples were purchased from Becky Hoffman, Flow Cytometry Department, Madison, WI. Microorganisms for the specificity analysis were kindly provided by Randy Worobo and John Churey, Cornell University, Geneva, NY, and Joan Rose, St. Petersburg, FL. Primer and probes were synthesized at the Biotechnology Center at Cornell University, Ithaca, NY. For the nucleic acid extraction, Qiagen RNeasy kits were purchased, and materials required for the Boom technique were kindly provided by Organon Teknika. Quantified hsp70 mRNA derived from the C. parvum gene was provided by John Paul and Matt Ewert, USF in St. Petersburg, FL. Ethanol, 200 proof molecular biological grade, water, and buffer reagents were obtained from Sigma Co. The NucliSense Reader for the ECL analysis was made available to our laboratory by Organon Teknika. DNA Primer and Probe Sequences. From the sequence C. parvum hsp70 mRNA, complete cds12 published in the GeneBank from the National Center for Biotechnology Information,13 a C. parvum specific sequence was selected using the Basic Local Alignment Search Tool (BLAST). Using DNAStar’s Primer Select, software sequences were defined as NASBA primers and ECL probes. We followed the general NASBA protocol to design NASBA primer 1.14 In general, it consists of a target-specific region (bold), the T7 RNA polymerase promotor (underlined), and a short adenine- and guanine-rich region. Primer 2, detection probes 3 and 4 each consist of a sequence-specific region for hybridization with their respective target sequences. The sequences of both primers and probes are given in Table 1. Probes 1 and 2 were labeled with biotin at the 5′ end. Subsequently, they were provided to OT where they were coupled to magnetic beads. Probes 3 and 4 were labeled with a C3 amino linker at the 5′ end and were attached to the ECL Tris(2,2′bipyridine)ruthenium label by OT. The probe solutions were prepared in a buffer containing 0.5 mL of BSA (10 g/L) plus 2-chloracetamide (50 g/L), 3.125 mL of 20× SSC, 0.05 mL of probe (33 µM), and 1.325 mL of water. Each probe was prepared at a concentration of 2 × 1014 molecules/mL. Heat Shock Procedure, Nucleic Acid Extraction, and Purification Protocol. For the stimulation of heat shock mRNA production in oocysts of C. parvum, the oocysts were heated for 20 min at 42 °C in a heating block. Subsequently, the oocysts were disrupted by incubation in lysis buffer (provided in the Qiagen RNeasy kit and Organon Teknika Boom extraction kit) at 60 °C for 30 min with 5 s of vortexing every 10 min. Several (12) U71181 Cryptosporidium parvum heat shock protein (Hsp70) mRNA, complete cds gi|1616782|gb|U71181.1|CPU71181 [1616782]. (13) NCBI. http://www.ncbi.nlm.nih.gov/Web/Search/index.html. (14) Kievits, T.; van Gemen, B.; van Strijp, D.; Schukkink, R.; Dircks, M.; Adriaanse, H.; Malek, L.; Sooknanan, R.; Lens, P. J. Virol. Methods 1991, 35, 273-286.

Analytical Chemistry, Vol. 73, No. 6, March 15, 2001

1177

Table 1. Sequences of NASBA Primers and ECL Probes DNA probe/primer (position in target sequence)

sequence

primer 1 (1063-1082) primer 1B (1110-1130) primer 2 (962-984) probe 1 (962-981) probe 2 (985-1004) probe 3 (993-1012) probe 4 (1091-1110)

5′ aat tct aat acg act cac tat agg gag aag gta gaa cca cca acc aat aca 3′ 5′ aat tct aat acg act cac tat agg gag aag gta ccg tta aag aat tcc tga a 3′ 5′ aga ttc gaa gaa ctc tgc gct ga 3′ 5′ aga ttc gaa gaa ctc tgc gc 3′ 5′ tta ctt ccg tgc aac ttt ag 3′ 5′ gtg caa ctt tag ctc cag tt 3′ 5′ cca aag gtt cag gcc ttg at 3′

other techniques were explored including, freeze/thaw cycles in liquid N2/65,15 bead beating, and incubation at various elevated temperatures. Nucleic acid was subsequently purified using Boom technology16 or the Qiagen RNeasy kit.17 The Qiagen procedure was modified, by incorporating an additional ethanol-washing step, and the spin times were increased from 15 to 20 s. Additionally, the drying time was increased from 2 to 4 min. NASBA Reaction and ECL Detection Protocols. The NASBA standard procedure was followed using the Organon Teknika NASBA kit18 containing 70 mM KCl. In brief, primers 1 and 2 were mixed with NASBA buffers and solutions. Nucleic acid extracts were added, and the mixture was incubated for 5 min at 65 °C. Subsequently, the samples were incubated for 5 min at 41 °C, the enzyme cocktail was added, and the reaction incubated for another 60-90 min at 41 °C. The reactions were stored at -20 °C or directly analyzed with the ECL technique: 10 µL of ECL labeled probe and 10 µL of magnetic bead labeled probe were mixed with 5 µL of 1:10 diluted NASBA reaction. The mixture was incubated for 30 min at 41 °C. Finally, 300 µL of ECL assay buffer was added to the mixture and the samples were analyzed in the NucliSense reader. RESULTS AND DISCUSSION NASBA Reaction and ECL Detection Design. We optimized the NASBA reaction concerning KCl concentration and primer combinations. A concentration of 70 mM KCl was found to be optimal for the amplification of C. parvum mRNA. Several primer sequences were investigated (not all sequences and data are shown). However, primers 1 and 2 were shown to be optimal regarding sensitivity and signal amplification. The amplification was kinetically independent from initial mRNA concentrations (Table 2). A time course experiment reveals that the reaction achieves completion after 75 min (Figure 2). High signals were obtained for the optimal primer pair, which allows detection of NASBA amplicons after 45 min. However, we usually incubated reactions for 60-90 min to ensure optimal performance. Detection probes 1 and 3 hybridized specifically and sensitively with the amplicon derived from primers 1 and 2. Thus, it was possible to amplify less than 100 mRNA molecules effectively using the NASBA protocol. (15) Johnson, D. W.; Pieniazek, N. J.; Griffin, D. W.; Misener, L.; Rose, J. B. Appl. Environ. Microbiol. 1995, 61, 3849-3855. (16) Boom, R.; Sol, C. J. A.; Salimans, M. M. M.; Jansen, C. L.; Wertheim-vanDillen, P. M. E.; van der Noordaa, J. J. Clin. Microbiol. 1990, 28, 495-503. (17) Qiagen RNeasy Mini Handbook, 2nd ed.; Qiagen: Valencia, CA, 1997; pp 33-35. (18) Kievits, T.; van Gemen, B.; van Strijp, D.; Schukkink, R.; Dircks, M.; Adriaanse, H.; Malek, L.; Sooknanan, R.; Lens, P. J. Virol. Methods 1991, 35, 273-286.

1178 Analytical Chemistry, Vol. 73, No. 6, March 15, 2001

Table 2. Determination of the NASBA Sensitivity Using Primers 1 and 2 no. of mRNA molecules per NASBA reaction

no. of positives (total no. of rxns, 6)

av ECL value

5× 5 × 105 5000 500 50 25 10 0

6 6 6 6 6 5 6 0

>2 000 000 >2 000 000 >2 000 000 >2 000 000 >2 000 000 1 900 000 1 600 000 400

107

Figure 2. NASBA time course experiment using two different pairs of primers.

Oocyst Disruption and Nucleic Acid Extraction Optimization. Various oocyst disruption methods were investigated in order to optimize the nucleic acid release from oocysts of C. parvum. Standard procedures from the literature were investigated, such as freeze/thaw cycles19 and incubation at 95 °C for 10 min.20 Additionally, new methods were also evaluated. The following different procedures were evaluated: (a) incubation at 95 °C for 5 min in lysis buffer; (b) incubation at 56 °C in lysis buffer for 20-60 min; (c) incubation at 60-95 °C for 10-40 min in lysis buffer; (d) incubation at 60 °C for 30 min in lysis buffer with vortexing every 10 min for 5 s; (e) incubation at room temperature in lysis buffer for 30-60 min; (f) 3-6 times freeze/ thaw cycling in lysis buffer using liquid nitrogen (or a dry ice/ ethanol bath) and a 65 °C water bath, with incubation at each condition for 30 s-1.5 min; (g) 2-3 times freeze/thaw cycling (19) Johnson, D. W.; Pieniazek, N. J.; Griffin, D. W.; Misener, L.; Rose, J. B. Appl. Environ. Microbiol. 1995, 61, 3849-3855. (20) Deng, M. Q.; Cliver, D. O.; Mariam, T. W. Appl. Environ. Microbiol. 1997, 63, 3134-3438.

with subsequent bead beating for 2 min using 0.58 g of 100-µm glass beads in lysis buffer. The extractions were carried out using oocyst concentrations between 10 and 1000 per sample. Nucleic acid extracts were purified using Boom technology or the Qiagen RNeasy kit. We found that method f, 6× freeze/thaw cycling, and method d, 30min incubation at 60 °C, gave the most reproducibly optimal results. Both yielded similar results. However, adding three 5-s vortexing steps to the 60 °C incubation method (i.e., every 10 min) increased its sensitivity and reliability. Thus, this optimized oocyst of C. parvum disruption protocol was selected as our standard method due to its simplicity and effectiveness. The only equipment needed were a heating block and a vortex, while the only chemical required is a lysis solution provided in the nucleic acid purification kits. Nucleic acid was purified using Boom technology and the Qiagen RNeasy kit. The original Qiagen protocol was optimized by adding an additional ethanol-washing step and by prolonging spin times from 15 to 20 s. as well as the drying time from 2 to 4 min. Up to 12 samples could easily be extracted simultaneously within about 30-40 min. To prevent the inhibition of subsequent NASBA reactions by the nucleic acid extracts, additional drying time and washing steps were added. The Boom technology was carried out following the manufacturer’s protocol. Although based on similar purification principles, it was found to be more complicated and time-consuming, taking an average of 60-70 min to complete. Therefore, the Qiagen procedure was chosen as the standard nucleic acid purification method. Heat Shock Optimization. The heat shock protein production in oocysts of C. parvum was stimulated by a brief heat shock at elevated temperatures. Initially, 20 oocysts/sample were incubated for 20 min at 45 °C following a protocol recommended by Stinear et al.21 Then, different temperatures and incubation times were investigated in order to optimize mRNA production. The heat shock temperature was varied from 41 to 47 °C and the incubation time from 5 to 30 min. It was found that incubation for 20 min at 42-43 °C was optimal for stimulation of hsp70 mRNA production. Ten oocysts were detected reliably when these heat shock conditions were applied. We assume that lower temperatures were not high enough to stimulate the heat shock protein production. At higher temperatures, the expression of proteins and their respective mRNAs was probably inhibited as suggested by Lindley et al.22 Oocyst of C. parvum Detection. Optimal conditions for heat shock stimulation, nucleic acid extraction, NASBA reaction, and ECL detection were utilized in order to detect oocysts of C. parvum in buffer solution. Serial dilutions of an oocyst stock solution were made and analyses performed in duplicate. The results are summarized in Table 3. Positive signals were obtained for all oocyst-containing solutions. Since oocyst concentrations were obtained by serial dilution only, a second experiment was performed in which 10 and 1 oocysts/sample were investigated in 10 replicate analyses. As Table 4 shows, 80% of the 10-oocyst sample and 20% of the 1-oocyst samples were positive. Although these sample concentrations were obtained by serial dilution, their (21) Stinear, T.; Matusan, A.; Hines, K.; Sandery, M. Appl. Environ. Microbiol. 1996, 62, 3385-3390. (22) Lindley, T. A.; Chakraborty, P. R.; Edlind, T. D. Mol. Biochem. Parasitol. 1988, 28, 135-144.

Table 3. Detection of Viable Oocysts of C. parvum. Serial Dilution of Oocysts in Buffer Solution, Duplicate Analysis no. of oocysts/ sample

ECL detection signal

1 000 000 100 000 10 000 1 000 100 10 0

positive positive positive positive positive positive negative

Table 4. Detection of Viable Oocysts of C. parvum. Serial Dilution of Oocysts in Buffer Solution, 10 Replicates Per Oocyst Concentration no. of oocysts/sample

no. of replicates

no. of positive samples

10 1

10 10

8 2

Table 5. Detection of Viable Oocysts of C. parvum. Certified Samples Obtained by Flow Cytometry no. of oocysts/sample

no. of replicates

no. of positive samples

50 10 5

7 7 7

7 6 6

Table 6. Detection of Killed Oocysts of C. parvuma

positive negative

viable

killed, 30-min incubn

killed, 3-h incubn

killed, 24-h incubn

5 1

3 3

1 5

0 6

aOocysts were killed and detected at different times after the killing procedure. Analyses performed in six replicates.

analysis provides initial indications of the detection limit (i.e., sensitivity) of the assay. Therefore, certified oocyst samples (containing specific numbers of oocysts enumerated by flow cytometry) were purchased. Table 5 summarizes the results obtained with certified samples. The results show that the detection system can detect as few as 5 oocysts/sample in buffer solution. In a subsequent analysis, the detection of dead oocysts of C. parvum was investigated. Forty oocysts/sample were killed by heating for 15 min at 95 °C in buffer solution. Killed oocysts were analyzed directly after the killing procedure, after 3 h, and after 24 h stored at room temperature (Table 6). During the heating procedure at 95 °C, some hsp70 mRNA was produced which was readily degraded and after 24 h the oocysts were no longer detectable. Thus, the detection system specifically detects only viable oocysts of C. parvum. Analysis of Assay Specificity for C. parvum. The specificity of a nucleic acid amplification and hybridization system is determined by the specificity of the target sequence (i.e., the amplification primers and the detection probes). Therefore, a variety of different microorganisms was investigated (Table 7). Fresh cultures of all listed organisms were diluted to 100-100 000 Analytical Chemistry, Vol. 73, No. 6, March 15, 2001

1179

Table 7. Determination of the Assay Specificity. Investigation of Microorganisms with 100-100 000 Cells Per Sample Bacillus cereus 4552 Bacillus subtilis 6633 Clostridium butyricum Cyclospora DPF (Alicyclobacillus) E. coli a33 E. coli 43895 E. coli 25922 Giardia lamblia

Fecal streptococcus Klebsiella pneumonia Lactobacillus casei Lactobacillus delbrueckii Lactobacillus viridescens Listeria innocua Listeria iranovii Listeria monocytogenes Saccharomyces rouxii

Table 8. Detection of Different Amounts of Oocysts Spiked into Environmental Water Samples Prior to the IMS Procedure % of positive samples type of water

500 oocysts

50 oocysts

0 oocysts

clean dirty

100 100

100 83

0 0

cells/sample, heat shock was applied, and the nucleic acid was extracted, amplified in NASBA, and detected with the ECL system. None of the microorganisms studied produced a false positive signal. Hence, the assay is highly specific for C. parvum. Additionally, samples of Giardia, Cyclospora, Escherichia coli, Streptococcus faecalis, and Salmonella typhimurium were spiked with 40 oocysts of C. parvum. In all cases, oocysts of C. parvum could be detected without any interference by overwhelming concentrations of nonspecific RNA found in the added organisms. Environmental Samples. In a standard Vortex flow filtration (VFF) and immunomagnetic separation (IMS) procedure, 100 L of water is filtered to obtain a final volume of 40 mL. Subsequently, a 10-mL sample is taken and used in the IMS procedure. Captured oocysts are released from the magnetic beads for analysis. To simulate this procedure, we received two different types of environmental samples from Dave Kelsey at ImmuCell Co.: clean water (DI water) and dirty water (0.5-mL packed pellet/10.0-mL sample). Ten milliliters of these water samples was spiked by Dave Kelsey with various amounts of oocysts of C. parvum, and oocysts were purified from the water samples with IMS. We prepared IMS samples in two variations: oocysts released and oocysts nonreleased from the magnetic beads. As shown in Table 8 we were able to detect 50 and 500 oocysts from dirty and clean water samples. There was no difference between nonreleased and released oocysts. Spiking dirty and clean water samples with lower oocyst numbers (i.e., 1, 5, and 7), we were only able to detect samples containing 7 oocysts (7.25 in average). Two out of six clean water samples and three out of six dirty water samples were positive. We assume that not all the

1180 Analytical Chemistry, Vol. 73, No. 6, March 15, 2001

Saccharomyces bisporous Salmonella typhimurium Shigella paradysenteriae 60R Shigella sonnei Staphylococcus aureus Vibrio parahaemolyticus yeast y307 Yersinia enterocolitica Zygosaccharomyces bailii

oocysts of C. parvum were viable at point of detection. Also, a loss of oocysts during the IMS procedure could cause a total loss of oocysts at these low spiking concentrations; thus, it was not possible to determine sample containing less than 7 oocysts. CONCLUSION We have presented in this paper the development of a highly specific and very sensitive assay for the detection of viable oocysts of C. parvum in environmental samples. Our test system is capable of detecting 5 viable oocysts in buffer solution as well as less than 10 oocysts (i.e., 7.25) in a 10-mL environmental sample. To obtain definite detection limits of our test system, we worked with certified samples obtained by flow cytometry. In the future, we will investigate several environmental samples from different sources in order to determine matrix effects on the nucleic acid extraction and NASBA amplification reactions. Also, we will extensively study low oocyst concentrations in environmental samples. ACKNOWLEDGMENT The authors acknowledge Organon Teknika, Boxtel, The Netherlands, for providing NASBA and ECL reagents, the ECL NucliSense Reader, Joan Rose, University of South Florida, St. Petersburg, FL, for initial help with oocyst of C. parvum samples, John Paul and Matt Ewert, University of South Florida, St. Petersburg, FL, for support in the nucleic acid extraction development, and Dave Kelsey, ImmuCell Corp. for providing environmental samples spiked with oocysts of C. parvum. The authors thank for financial support of this work the German Academic Exchange Institution (DAAD), the New York State Energy Research and Development Authority (NYSERDA), Innovative Biotechnologies International, Inc., and the Cornell Center of Advancement in Technology in Biotechnology.

Received for review November 2, 2000. Accepted January 17, 2001. AC001293H