Detection of Viable Cryptosporidium parvum in Soil by Reverse Transcription–Real-Time PCR Targeting hsp70 mRNA

Abstract

Extraction of high-quality mRNA from Cryptosporidium parvum is a key step in PCR detection of viable oocysts in environmental samples. Current methods for monitoring oocysts are limited to water samples; therefore, the goal of this study was to develop a rapid and sensitive procedure for Cryptosporidium detection in soil samples. The efficiencies of five RNA extraction methods were compared (mRNA extraction with the Dynabeads mRNA Direct kit after chemical and physical sample treatments, and total RNA extraction methods using the FastRNA Pro Soil-Direct, PowerSoil Total RNA, E.Z.N.A. soil RNA, and Norgen soil RNA purification kits) for the direct detection of Cryptosporidium with oocyst-spiked sandy, loamy, and clay soils by using TaqMan reverse transcription-PCR. The study also evaluated the presence of inhibitors by synthesis and incorporation of an internal positive control (IPC) RNA into reverse transcription amplifications, used different facilitators (bovine serum albumin, yeast RNA, salmon DNA, skim milk powder, casein, polyvinylpyrrolidone, sodium hexametaphosphate, and Salmonella enterica serovar Typhi) to mitigate RNA binding on soil components, and applied various treatments (β-mercaptoethanol and bead beating) to inactivate RNase and ensure the complete lysis of oocysts. The results of spiking studies showed that Salmonella cells most efficiently relieved binding of RNA. With the inclusion of Salmonella during extraction, the most efficient mRNA method was Dynabeads, with a detection limit of 6 × 10² oocysts g⁻¹ of sandy soil. The most efficient total RNA method was PowerSoil, with detection limits of 1.5 × 10², 1.5 × 10³, and 1.5 × 10⁴ C. parvum oocysts g⁻¹ soil for sandy, loamy, and clay samples, respectively.

INTRODUCTION

As an important enteric protozoon of worldwide health concern, Cryptosporidium parvum is one of the major waterborne causative agents of gastrointestinal illness in humans (17). The infectious form of C. parvum, the oocyst, is highly resistant to environmental stress and hence can remain viable in the environment for months (56). Viability tests are important in the detection of C. parvum oocysts in environmental samples, since only viable oocysts pose a threat to public health (37). The massive devastating consequence of the waterborne cryptosporidiosis outbreak in Milwaukee, WI, in 1993 also illustrates the potential destructive effect of C. parvum during accidental or intentional contamination scenarios (39, 46). Our previous study showed that surface soil can serve as a major source of C. parvum introduced to the environment by cattle ranching (30). Once contaminated, soil serves as a constant source of oocysts to surface water bodies by runoff and potentially to groundwater by infiltration. The current detection methods, however, are mostly confined to water. For fast responses to land management practices or intentional contamination events, rapid and sensitive methods that simultaneously recover and detect viable oocysts in soil are needed. Traditionally, animal infectivity (16), in vitro cell culture (13, 60), in vitro excystation (7), and staining by fluorogenic vital dyes (10, 56) were used to determine the viability of C. parvum oocysts. Prior to those assays, oocysts should be separated from environmental matrix through various purification procedures, which resulted in the loss of oocysts, variable recovery rates, and frequent interference with the oocysts' integrity (1, 8, 35). The detection assays are also expensive, time-consuming, and labor-intensive, which could further limit their applicability in making quick and accurate identification of viable C. parvum from soil.

In addition to vital dye staining, another microscopic method for the detection of Cryptosporidium oocysts in soil includes fluorescence in situ hybridization (FISH) that targets a variable region of the small-subunit (SSU) rRNA. Results from several investigations have demonstrated that both microscopic assays correlate with results from the standard mouse infectivity and in vitro excystation assays (10, 26). However, the interpretation of viability from both assays must be undertaken cautiously, since the tests are known to overestimate viability (7, 10, 28, 50). Therefore, the extent to which rRNA probes are useful for oocyst viability studies depends on the decay rate of SSU rRNA in the environment, which likely varies depending on different environmental conditions (28). Some of the bias that also may interfere with the quantification of viable oocysts is due to the high autofluorescence properties of soil particles.

Previous investigations have shown that the production of mRNA was correlated with the viability of C. parvum oocysts (42, 63). Methods combined reverse transcription (RT) with PCR targeted mRNA coding for various proteins, including heat shock protein 70 (hsp70) (5, 24, 29, 57, 63), β-tubulin (68), amyloglucosidase (27), and CP2 (38) from purified suspension or water samples. Assays targeting hsp70 mRNA are advantageous for enhancing the sensitivity of RT-PCR since hsp70 mRNA is produced in abundance in response to heat or other stresses. Compared with conventional PCR, real-time PCR is more rapid, sensitive, and specific. Quantitative reverse transcription–real-time PCR (qRT-PCR) has also been reported to quantify hsp70 mRNA from viable oocysts in manure (21). However, until now, no qRT-PCR method targeting C. parvum hsp70 mRNA has been reported to assess oocyst viability in soil.

The extraction of high-purity mRNA from soil matrices is the key for a successful RT-PCR. Although many soil RNA extraction procedures have been described (4, 18, 19, 23, 36, 44, 47, 52), in practice, however, the efficient recovery of mRNA from soil remains a challenge. Three interfering factors may include the following: (i) mRNA is prone to degradation by RNase, which is ubiquitous in soil environments; (ii) soils are rich in humic substances, which could be coextracted with mRNA and inhibit the subsequent RT-PCR even at minute concentrations; and (iii) mRNA binds to naturally present soil components such as clay. This is particularly problematic for soil with very low biomass (6), such as subsurface soil.

Commercially available isolation kits represent a useful alternative to conventional soil RNA extraction methods. Kits are time efficient and have the potential to expedite and standardize sample processing and thus increase the throughput and reliability of extraction procedures. As could be expected, the selection of appropriate methods can influence the outcome of the C. parvum oocyst viability determination in terms of quality and quantity of RNA recovered (14, 52, 59) and the sensitivity and specificity of RT-PCR (22). The significant differences in the physical-chemical properties of various soils (i.e., organic matter, clay content, and pH) may lead to variable efficiencies of nucleic acid extraction methods (9, 58). Therefore, it became necessary to modify the existing methods to overcome the matrix impact of the soil texture and type.

In this study, the efficiencies of five commercial RNA extraction methods, with or without modifications, were evaluated for the extraction of RNA from C. parvum oocysts seeded into different soils (i.e., sand, clay, and loam) by using TaqMan qRT-PCR targeting hsp70 mRNA. The extraction methods used represent major methodologies commercially available for RNA extraction from soil, including spin column chromatography using resin or filter membrane, high binding matrix, and magnetic bead purification. The effectiveness of several facilitators to relieve RNA binding on soil components was also assessed. A foreign internal positive control (IPC) RNA was incorporated into the qRT-PCR assay to detect potential inhibitors present in RNA samples. Based on the study results, we developed a procedure for the rapid and accurate detection of viable C. parvum oocysts in soil by using a combination of physical and chemical treatments as well as Salmonella cells followed by the Dynabeads mRNA Direct kit for mRNA extraction and the addition of IPC RNA to qRT-PCR mixtures. This study is the first to demonstrate the extraction of C. parvum oocysts from soil.

MATERIALS AND METHODS

Cryptosporidium oocysts.

Suspensions of flow cytometry-sorted C. parvum (Iowa isolate) shipped in antibiotic solution from Waterborne, Inc., were used for oocyst studies. A preparation of 3 × 10⁵ oocysts in 50 μl of phosphate-buffered saline was used for each spiking experiment: mRNA soil extraction efficiency and/or the effect of RNA extraction facilitators for the relief of RNA binding on soil components. Preparations of 3 × 10¹ to 3 × 10⁵ oocysts in 50 μl of phosphate-buffered saline were used for the evaluation of detection sensitivity. Final concentrations of oocyst preparations were made by serial dilution of concentrated stock solutions in phosphate-buffered saline each day and were counted by hemocytometer using a Zeiss Axiophot2 epifluorescence microscope (Carl Zeiss, Thornwood, NY) as described by EPA Method 1623 (65). Oocyst stock solutions were stored at 4°C and used within 60 days of isolation from infected calves.

Soil samples.

Environmental soil samples were collected at various locations to represent sand, loam, and clay from the vicinity of Byrd's Mill Spring, the water supply of the city of Ada, OK. After stones and plant residues were removed, each soil batch was homogenized separately (2-mm sieve) and stored at 4°C. Soils were analyzed by the Soil Analytical Laboratory in Oklahoma State University (Stillwater, OK) for particle size and organic matter content (Table 1).

Table 1.

Environmental soil characteristics used for spiking and RNA extraction trials

Soil classification	pHa	Organic matterb (%)	Texture (%)
			Sand	Silt	Clay
Sand	6.8	0.48	97.5	2.5	0
Sandy loam	7.2	0.97	77.5	12.5	10
Clay	7.1	2.64	25	30	45

^aMeasured after equilibration with distilled water.

^bMass loss of dried material during ignition.

In addition to the environmental soil samples, the following soil components were purchased for oocyst spiking experiments: reference Ca-rich montmorillonite clay (STX-1b; Gonzales County, TX) from Source Clays Repository of Clay Minerals Society and ground silica sand (SIL-CO-SIL 90; top size, 90 μm) from U.S. Silica (Berkeley Springs, WV), which was acid washed as described previously (41, 55) to remove residual iron and organic matter.

On each extraction day, soil aliquots ranging from 0.5 to 2.0 g as per the manufacturer's recommendations, were spiked with oocysts and thoroughly mixed by vortexing followed by incubation (5 min, 25°C). The spiked samples were then subjected to heat induction for hsp70 mRNA production (if needed) by incubation (20 min, 45°C) followed by immersion in ice water to terminate heat induction. Oocyst isolates inactivated by incubating at 95°C for 20 min were subjected to the same extraction as a negative control.

(m)RNA extraction.

A total of 446 soil samples were extracted. Prior to all extraction procedures, solutions, and plasticware were made RNase free by diethylpyrocarbonate (DEPC) treatment, glassware was baked overnight (200°C), and pipettors and benchtops of the work areas were wiped with UltraClean lab cleaner (Mo Bio, Carlsbad, CA) to inactivate RNase.

With the exception of mRNA extraction from isolates of oocysts (3 × 10⁵) as positive controls, which were first mixed with lysis/binding buffer (200 μl) of the Dynabeads mRNA Direct kit (Invitrogen, Carlsbad, CA) and then were subjected to 6 cycles of freeze (liquid N₂, 1 min)-thaw (65°C, 1 min) to disrupt oocysts, which was followed by the precipitation of residual cell wall by centrifugation (17,000 × g, 3 min) and mixing of the supernatant with 1.5 mg of prewashed Dynabeads oligo(dT)₂₅ in accordance with the manufacturer's protocol. One of the following five methods was used to extract total RNA (methods 1 to 4) or mRNA (method 5) from spiked soil or sand/clay mixtures following the manufacturer-recommended protocols. The following modifications, however, apply to all methods: (i) the addition of β-mercaptoethanol (BME) to lysis buffer at the final concentration of 10 μl ml⁻¹ at the onset of the extraction process to inactivate RNase, and (ii) replacing vortexing with three cycles of bead beating (40 s each at 4,400 rpm) using FastPrep-24 homogenizer (MP Biomedical, Solon, OH) to ensure the complete lysis of oocysts.

(i) Method 1: total RNA extraction with the FastRNA Pro Soil-Direct kit.

Briefly, 0.5 g of spiked environmental soil (sand, loam, or clay) mixed with lysing matrix E and RNApro soil lysis solution of the FastRNA Pro Soil-Direct kit (MP Biomedicals, Solon, OH) was subjected to bead beating. Released RNA was recovered using phenol-chloroform extraction, precipitated using isopropanol, washed with ethanol, and resuspended into DEPC water. RNA was then bound to the RNAMATRIX, cleaned by a washing buffer, and eluted in DEPC water as described in the manufacturer-recommended protocol.

(ii) Method 2: total RNA extraction with the PowerSoil total RNA isolation kit.

Briefly, 2 g of spiked environmental soil (sand, loam, or clay) was treated with a series of solutions from the PowerSoil total RNA isolation kit (Mo Bio, Carlsbad, CA) and then processed with bead beating. Total RNA was isolated by phenol-chloroform extraction, precipitated by isopropanol, captured on RNA capture column, washed and eluted with the elution buffer, and recovered by a final isopropanol precipitation as described in the manufacturer-recommended protocol.

(iii) Method 3: total RNA extraction with E.Z.N.A. soil RNA kit.

Briefly, a 15-ml tube containing 2 g of spiked environmental soil (sand, loam, or clay), glass beads, and a series of lysis solutions from the E.Z.N.A. soil RNA kit (Omega Bio-Tek, Norcross, GA) was subjected to bead beating. RNA was isolated by using phenol-chloroform extraction twice, bound to HiBind RNA column packed with proprietary matrix, cleaned with a wash buffer, and eluted from packed column by DEPC water as described in the manufacturer-recommended protocol.

(iv) Method 4: total RNA extraction with Norgen soil total RNA purification kit.

Briefly, 0.5 g of spiked environmental soil (sand, loam, or clay) was mixed with lysis solution from the Norgen soil total RNA purification kit (Norgen Biotek, Thorold, ON, Canada) and subjected to bead beating. RNA in the supernatant mixed with binding solution I and II was then bound on the proprietary resin in the spin column. Impurity was washed away using washing solution and RNA eluted with the elution buffer as described in the manufacturer-recommended protocol.

(v) Method 5: mRNA extraction using physical and chemical treatments prior to Dynabeads mRNA Direct kit.

The method was developed using the following oocyst-spiked soils: (i) acid-washed quartz sand, (ii) mixture of reference Ca-rich montmorillonite clay STX-1b and acid-washed quartz sand, and (iii) environmental sand, loam, or clay. Briefly, 0.5 g of spiked soil sample was transferred into a 2-ml tube containing a mixture of glass beads with various diameters (2 mm, 0.2 g; 1 mm, 0.2 g; 0.5 mm, 0.2 g) and 1 ml of mRNA extraction buffer (4 M guanidinium thiocyanate, 0.1 M Trizma base, 1% dithiothreitol, 0.5% N-lauroylsarcosine sodium [pH 8.0]) and was subjected to three bead-beating cycles (40 s each at 4,400 rpm, set 6.0) by FastPrep-24 homogenizer (MP Biomedical, Solon, OH) (physical treatment step). The sample was then centrifuged (14,000 × g, 5 min), the supernatant was mixed thoroughly with 800 μl phenol-chloroform-isoamyl alcohol (25:24:1 [vol/vol]) (Amresco Inc., Solon, OH) (chemical treatment step) and centrifuged again (14,000 × g, 5 min at 4°C). RNA in the supernatant was precipitated using 600 μl of ice-cold isopropanol. The Dynabeads mRNA Direct kit (Invitrogen, Carlsbad, CA) was then used to process the sample by following the manufacturer-recommended protocol. Briefly, the RNA precipitate was resuspended into 400 μl Dynabeads lysis/binding buffer and mixed with 1.5 mg Dynabeads oligo(dT)₂₅. Following hybridization (15 min, 25°C) with continuous mild rolling, Dynabeads oligo(dT)₂₅ was isolated using a magnetic concentrator and washed three times each with buffer A and B. The resulting mRNA was then eluted from the Dynabeads oligo(dT)₂₅ by incubating the beads (80°C, 2 min) in 15 μl Tris-HCl.

Total RNA or mRNA was kept on ice and subjected to purification (if necessary) immediately. To remove residual coextracted DNA, 10 μl eluted (m)RNA was then subjected to DNase treatment using a Turbo DNA-free kit (Ambion/Applied Biosystems, Foster City, CA) by following the manufacturer's recommended protocol. After the final elution, RNA samples were subjected to an additional centrifugation step (10,000 × g, 90 s) to completely remove the inactivation reagent that might interfere with the downstream enzymatic reaction. RNA extraction efficiency was defined by the threshold cycle (C_T) values obtained with TaqMan RT-PCR, which indicated the suitability of the extracted RNA for RT-PCR. RNA extraction/purification, DNase treatment, and RT-PCR assays were conducted on the same day.

Quantitative reverse transcription–real-time PCR.

The evaluations of different RNA extraction methods, RNA extraction with facilitators, and different treatments for removing inhibitory material were done by quantitative reverse transcription–real-time PCR (qRT-PCR) targeting the hsp70 gene of C. parvum. PCR primers, probes, and oligonucleotides used in this study are listed in Table 2 along with their sequences.

Table 2.

PCR primers, probes, and oligonucleotides used for detection of C. parvum and RNA IPC construction

Oligonucleotide		Target	Sequence (5′–3′)	Reference or source
Type	Name
Primer pair 1	CP-hsp70f	C. parvumhsp70 gene	AACTTTAGCTCCAGTTGAGAAAGTACTC	21
	CP-hsp70r		CATGGCTCTTTACCGTTAAAGAATTCC
Primer pair 2	Spud-f	IPC	AACTTGGCTTTAATGGACCTCCA	51
	Spud-r		ACATTCATCCTTACATGGCACCA
Primer pair 3	Cp-spudf	IPC	AACTTTAGCTCCAGTTGAGAAAGTACTCAACTTGGCTTTAATGGACCTCCA	This study
	Cp-spudr		CATGGCTCTTTACCGTTAAAGAATTCCACATTCATCCTTACATGGCACCA
Primer pair 4	T7-hsp70f	IPC	TAATACGACTCACTATAGGGAGAAACTTTAGCTCCAGTTGAGAAAGTACTC	This study
	T7-hsp70r		TAATACGACTCACTATAGGGAGACATGGCTCTTTACCGTTAAAGAATTCC
Probe	Cp-hsp70P	C. parvumhsp70 gene	6FAM-AATACGTGTAGAACCACCAACCAATACAACATC-BHQ1a	21
Probe	Spud-p	IPC	CY5-TGCACAAGCTATGGAACACCACGT-BHQ2	51
IPC	IPC		AACTTGGCTTTAATGGACCTCCAATTTTGAGTGTGCACAAGCTATGGAACACCACGTAAGACATAAAACGGCCACATATGGTGCCATGTAAGGATGAATGT	51

^a6FAM, 6-carboxyfluorescein; BHQ1, Black Hole Quencher 1.

All reverse transcription reactions were conducted with a GeneAMP PCR system 9700 (Applied Biosystems, Foster City, CA) using either the TaqMan reverse transcription (TR) system or high-capacity RNA-to-cDNA master mix (HC) (Applied Biosystems, Foster City, CA). TR was used to evaluate the effect of Salmonella during RNA extraction in preventing mRNA degradation. HR was used for all other reverse transcriptions unless otherwise stated. For TR, each reaction (20-μl volume) contained 2 μl (m)RNA, 1× TaqMan RT buffer, 5.5 mM MgCl₂, 500 μM each deoxynucleoside triphosphate, 2.5 μM either oligo(dT)₁₆ or random primer, 8 U RNase inhibitor, and 25 U of reverse transcriptase. The reaction mixtures were incubated for 10 min at 25°C and then at 48°C for 30 min, followed by 95°C for 5 min to inactivate the reverse transcriptase. To test for DNA contamination of the RNA extracts after the DNase treatment, control reaction mixtures were prepared as described above, but reverse transcriptase was not added. For HC, 1 μl of (m)RNA was mixed with 1× RNA-to-cDNA master mix in a 10-μl reaction volume. The reactions were performed by incubating the sample at 25°C for 5 min, 42°C for 30 min, and 85°C for 5 min. Contamination of gDNA was tested for using master mix without reverse transcriptase [(−)RT] included in the kit.

Immediately following the RT reaction, quantitative PCR was performed using an ABI 7500 real-time PCR system (Applied Biosystems, Foster City, CA). The 20-μl qPCR mixture contained 2 μl of RT product, 600 nM (each) forward (CP-hsp70f) and reverse (CP-hsp70r) primer, 300 nM TaqMan probe (CP-hsp70p), DEPC-treated DNase-free water and 1× TaqMan universal PCR master mix (including optimized concentration of AmpliTaq gold DNA polymerase, dinucleoside triphosphates [dNTPs] with dUTP, passive reference ROX dye [Applied Biosystems, Foster City, CA]). The thermal protocol consisted of 5 min of polymerase activation at 95°C, followed by 45 cycles of 15 s of denaturation at 95°C and 1 min of annealing/extension at 60°C.

Construction of RNA internal positive control fragment.

An RNA fragment was added to each sample as the internal positive control (IPC) during the reverse transcription to monitor the efficiency of extraction procedure and detect the presence of inhibitors. The assay we selected targeted a 101-bp fragment of the Solanum tuberosum phyB gene (GenBank accession no. Y14572), which encodes a species-specific regulatory photoreceptor involved in the pathway that results in the purple coloration of potato root (51). This DNA sequence was selected since it did not share homology with any of the other PCR targets used in the lab and was not present in the samples analyzed. Oligonucleotide of this fragment was commercially synthesized and high-pressure liquid chromatography (HPLC) purified (Integrated DNA Technologies, Inc., Coralville, IA). Double-strand DNA (D1) was obtained from oligonucleotide by amplifying with primer pair Spud-f and Spud-r. To introduce 5′ and 3′ terminal primer homology for Cp-hsp70f and Cp-hsp70r, D1 was 1/1,000 diluted and used as the template for amplification with primer pair Cp-spudf and Cp-spudr. The resultant DNA fragment (D2) was purified with a Qiagen PCR purification kit (Qiagen, Valencia, CA), diluted to 1/1,000, and amplified with primer T7-hsp70f. The T7-hsp70f was designed to introduce the T7 RNA polymerase promoter site to the 5′ terminus. To obtain RNA copy, D2 flanked with T7 promoter site at its 5′ terminus was purified with a Qiagen PCR purification kit and subjected to in vitro transcription using the T7 MEGAscript kit (Ambion/Applied Biosystems, Foster City, CA) following the manufacturer's instructions. Briefly, 2.5 μl template DNA (80 ng/μl) was mixed with NTP, reaction buffer, and T7 RNA polymerase in a final reaction volume of 20 μl and incubated (37°C, 6 h). Template DNA was removed by adding 1 μl Turbo DNase to the reaction product and incubating (15 min, 37°C). RNA was purified using a MEGAclear RNA purification kit (Ambion/Applied Biosystems, Foster City, CA) according to the manufacturer's instructions. The absence of DNA was confirmed by direct PCR using Cp-hsp70f and Cp-hsp70r. The identity of the RNA IPC was confirmed by reverse transcription and TaqMan RT-PCR. The RNA IPC concentration was estimated by A₂₆₀ using a NanoDrop 2000 spectrophotometer (Thermo Scientific, Wilmington, DE), and the copy number was calculated based on concentration. RNA IPC was serially diluted to a final concentration that gave a C_T value of between 33.5 and 34 when tested by RT-PCR and stored (−80°C) in 50-μl aliquots.

RNA binding trials.

Method 5 extraction procedures were used to confirm the binding of RNA on soil particles when spiked with 3 × 10⁵ C. parvum oocysts. RNA extraction was performed using the following soil combinations: (i) two aliquots of acid-washed quartz sand of different quantities (0 g, 0.25 g, 0.5 g, and 1 g) (EDTA was also added to only one aliquot to a final concentration of 50 mM before the lysis); (ii) 0.5 g mixture of reference montmorillonite clay STX-1b and acid-washed quartz sand with the following ratios (0, 2%, 10%, 20%, 30%, 40%, 50% of clay). Resultant RNA in each of the elution was quantified by RT-PCR.

RNA extraction facilitators.

Method 5 extraction procedures were used to evaluate the effectiveness of different facilitators to relieve RNA binding on soil components using 15% clay-sand mixture (0.5 g) spiked with 3 × 10⁵ C. parvum oocysts. Extraction facilitators tested were bovine serum albumin (BSA) (Sigma-Aldrich, St. Louis, MO), Saccharomyces cerevisiae RNA (RNA_Y) (Sigma-Aldrich, St. Louis, MO), salmon sperm DNA (Sigma-Aldrich, St. Louis, MO), skim milk powder (SM) (Difco, Detroit, MI), casein from bovine milk (Sigma-Aldrich, St. Louis, MO), polyvinylpyrrolidone (PVP) (molecular weight ∼360,000) (Sigma-Aldrich, St. Louis, MO), sodium hexametaphosphate (SMP) (Fisher Scientific, Wilmington, DE), and Salmonella Typhi (ATCC no. 700931, American Type Culture Collection, Manassas, VA). Final concentrations per gram of soil were as follows: salmon DNA (10 mg), BSA (20 mg), RNA_Y (20 mg), skim milk (40 mg), casein (40 mg), PVP (40 mg), SMP (100 mg), and Salmonella Typhi (10⁸ cells). Using method 5, each facilitator was added prior to the bead-beating step to the lysis tube and then processed as instructed. RNA in the elution was quantified by RT-PCR.

Facilitators proven to be effective to enhance RNA recovery (RNA_Y and Salmonella) were further tested on the three environmental soils (clay, loam, sand) to determine the optimal concentration by changing the amount of facilitators in the extraction. RNA extractions from the three soils were then evaluated by other methods (methods 1 to 4) with the presence of most effective facilitator in its optimal concentration.

The effect of Salmonella cells in preventing RNA degradation during extraction was tested by using various concentrations (0, 2 × 10⁷, 1 × 10⁸ cells/g) in 0.5 g of pure sand spiked with 3 × 10⁵ oocysts. RNA was extracted using method 1. Reverse transcription was performed with TaqMan reverse transcription system (TR) using either oligo(dT)₁₆ or random hexamers as the primer.

Removal of humic substances.

Two factors were considered in determining if additional cleanup steps were required to remove PCR inhibitors: (i) absorbance measurements, and (ii) the level of PCR inhibition. Nanodrop 2000 spectrophotometer (Thermo Scientific, Wilmington, DE) was used to evaluate the ability of each kit to remove humic substances, which absorb light around 320 nm (48). Soil samples, i.e., reference clay, acid-washed sand, and environmental clay, loam, and sand, were extracted without C. parvum oocysts. Light absorbance of nonspiked extracts was then measured at 320 nm. The level of PCR inhibition was determined when the extracts from nonspiked soil samples were mixed with RNA IPC and then evaluated with RT-PCR.

RNA purity was enhanced by two methods: (i) soil pretreatment using Al₂ (SO₄)₃, and (ii) RNA purification after extraction. The pretreatment procedure that was used to precipitate soil humic substances is as follows. Briefly, soil (0.5 g) spiked with C. parvum oocysts was mixed with 500 μl of phosphate buffer (0.1 M NaH₂PO₄-Na₂HPO₄, pH 6.6) and 500 μl of 0.1 M Al₂ (SO₄)₃ and subjected to mild vortexing (2 min). The superfluous Al³⁺ was precipitated by adjusting the pH to 8.0 by adding 250 μl 1 M NaOH. The mixture was then centrifuged (3,500 × g, 2 min), and the supernatant was removed. RNA was then extracted using method 5 with RNA_Y as the facilitator.

To purify RNA after the extraction, two different cleanup kits were used following the manufacturer's protocols: (i) the RNeasy MinElute cleanup kit (Qiagen, Valencia, CA), and (ii) the MicroSpin S-400 HR spin column (GE Healthcare, Little Chalfont, UK). Briefly, for the RNeasy MinElute cleanup kit, an eluted sample was mixed with RNase-free water and 100% ethanol and then transferred to silica membrane spin column to accomplish RNA binding. Impurity was washed away and RNA eluted in 14 μl of RNase-free water. The MicroSpin S-400 HR spin column was based on the gel filtration. The sample was applied slowly to the top center of the premade resin bed in the spin column. The spin column was then spun (735 × g, 2 min) to elute the RNA.

Statistical analysis.

The extraction efficiency as represented by C_T values (the number of cycles when a fluorescence signal above the threshold was detected, defined by the mean plus four times the SD of the fluorescence signal of the reference dye) obtained from TaqMan RT-PCR was used to determine the most efficient method for RNA extraction for each soil type. The methods were compared by one-way analysis of variance (ANOVA) on C_T values. Post hoc Tukey's test with a 95% mean confidence interval was performed. The efficiency of each facilitator to enhance RNA extraction was compared by one-way ANOVA on C_T values obtained and post hoc Tukey's test. The most efficient facilitator(s) was determined as the one with the lowest C_T value(s).

RESULTS

A total of 446 soil samples were extracted during this study. Each soil sample size was defined by the commercially available kits and ranged from 0.5 to 2 g. Without reverse transcription, no qPCR amplification was observed for samples treated with DNase. No RT-PCR amplification was detected for heat-inactivated oocysts when isolated from or spiked onto soil, indicating that the amplifications observed were caused by target hsp70 mRNA extracted from viable C. parvum oocysts.

Efficiency of extraction kits with oocyst-spiked soils following the manufacturers' protocols.

Five methods were tested to extract RNA from 0.5 to 2 g of soil spiked with 3 × 10⁵ C. parvum oocysts. Positive controls using mRNA extracted from 3 × 10⁵ C. parvum oocysts by Dynabeads mRNA Direct kit yielded mean C_T values of 26.5 ± 0.87 (means ± standard deviation [SD]) in RT-PCR. No RNA extract from any type of the soil with any method using the manufacturer's protocol yielded detectable hsp70 cDNA in RT-PCR, even after additional purification or dilution to alleviate the RT-PCR inhibition. However, as shown in Fig. 1, IPC RNA added before reverse transcription showed amplification for sandy soil by using methods 1, 2, 4, and 5; for loamy soil by using methods 2, 4, and 5; and for clay using methods 2 and 4. Factors other than inhibition were believed to be responsible for the failure of amplification. Therefore, modifications to the procedures were required to successfully extract RNA from soil in our experimental setting.

Fig. 1.

TaqMan reverse transcription–real-time PCR detection of IPC RNA mixed before reverse transcription with extract from sand (A), loam (B), and clay (C) samples by using extraction method 1 (solid squares), method 2 (open triangles), method 3 (solid circles), method 4 (solid triangles), and method 5 (open squares). Open circles indicate amplification of IPC RNA only. The fluorescence increase is plotted versus the cycle number of RT-PCR (plot start from cycle 26).

Effect of facilitators on RNA extraction.

Two tests verified the binding of RNA on soil particles. In the first test, C_T values increased from 33.8 with no sand to 37.0 with 1 g of sand. The addition of 50 mM EDTA reduced the C_T values by 2 to 3 units, as shown in Table 3. In the second test, mean C_T values increased from 32.4 with no clay to 37.7 with 10% clay, while no amplification was observed in samples with clay content of more than 10%, as shown in Table 4.

Table 3.

Detection of hsp70 mRNA in C. parvum oocysts spiked in different amounts of quartz sand with method 5^a

Sample (g)	Mean CT value by method 5 (CT ± SD)
	Without EDTA	With 50 mM EDTA
0	33.80 ± 0.27	30.82 ± 0.17
0.25	33.99 ± 0.22	31.42 ± 0.23
0.5	34.76 ± 0.23	32.41 ± 0.18
1	37.29 ± 0.46	33.66 ± 014

^aA minimum of 5 replicates were used.

Table 4.

Detection of hsp70 mRNA in C. parvum oocysts spiked in 0.5 g of quartz sand/reference montmorillonite clay mixture and extracted with method 5^a

Clay ratio (%)	Mean CT value by method 5 (CT ± SD)
0	32.41 ± 0.18
2	34.57 ± 0.34
10	37.73 ± 1.01
≥20	Undetected

^aA minimum of 5 replicates were used.

The effects of facilitators to relieve (m)RNA binding on soil particles during extraction with method 5 were examined with a reference clay/quartz sand mixture of 15:85 by weight, as shown in Table 5. Pure sand and reference clay STX-1 were used to exclude the influence of enzyme inhibitors, and no amplification was observed from samples without facilitator. The addition of RNA_Y at 20 mg/g and Salmonella Typhi at a concentration of 10⁸ cells/g soil significantly enhanced the extraction of mRNA, with C_T values of 32.4 and 32.3, respectively. SM and casein at a concentration of 40 mg/g greatly increased the yield of C. parvum RNA, with C_T values of 33.7 and 34.0, respectively. Salmon DNA at a concentration of 10 mg/g improved the extraction of RNA, with a C_T value of 38.6. BSA slightly enhanced the extraction of RNA from soil with a C_T value of 39.2. In contrast, PVP and SMP did not relieve the binding of RNA on soil particles, as no signals were observed for samples with these two facilitators.

Table 5.

Detection of hsp70 mRNA in C. parvum oocysts spiked in 0.5 g montmorillonite clay/quartz sand mixture (15% clay) extracted with method 5 in the presence of different facilitators and the C_T value of IPC RNA^a

Facilitator added	Mean CT value (CT ± SD)
	hsp70 mRNA	IPC RNA
BSA, 10 mg	39.26 ± 0.26	33.26 ± 0.08
RNAYb, 10 mg	32.44 ± 0.16	33.37 ± 0.32
DNA, 5 mg	38.66 ± 0.57	33.45 ± 0.31
Casein, 20 mg	34.10 ± 0.19	33.33 ± 0.09
Skim milk, 20 mg	33.73 ± 0.40	33.22 ± 0.16
Salmonella (5 × 107 cells)	32.33 ± 0.05	33.29 ± 0.10
PVP, 20 mg	Undetected	33.37 ± 0.14
SMP, 50 mg	Undetected	33.26 ± 1.51
No addition	Undetected	33.38 ± 0.02

^aRNA_Y, S. cerevisiae RNA.

^bA minimum of 5 replicates were used.

One-way ANOVA revealed that the C_T values for RNA obtained by using RNA_Y, salmon DNA, Salmonella, casein, SM, and BSA as facilitators were significantly different (P < 0.05, n = 18). Tukey's posttest analyses found no significant differences between C_T values by using RNA_Y and Salmonella. It was noted that during the phenol-chloroform extraction step, a white gel-like structure was formed between the two phases when using SM or casein as the facilitator and thus made the subsequent separation process very difficult. Since most of the soil RNA extraction kits included a step of phenol-chloroform extraction, SM and casein were excluded in further tests.

Next, the efficiency of RNA_Y and Salmonella on RNA extraction was tested on oocyst-spiked sandy and loamy soil with method 5. For sandy soil (Table 6), the addition of RNA_Y at a concentration as low as 4 mg/g of soil significantly relieved binding of RNA and thus enhanced its extraction with 3 replicates yielding consistent amplifications with a C_T value of 36.9. In the range from 4 to 20 mg/g of soil, the C_T value decreased with the increase in the amount of RNA_Y added. The lowest mean C_T value (30.58 ± 0.52) was obtained from samples with RNA at a concentration of 20 mg/g of soil. The C_T value remained unchanged when RNA_Y was between 20 and 80 mg/g of soil. The inclusion of Salmonella, at the onset of the extraction process, at a level as low as 2 × 10⁷ cells per gram of soil greatly improved the RNA extraction efficiency with a mean C_T value of 36.31 ± 0.60. However, increasing the concentration of Salmonella from 2 × 10⁷ to 2 × 10⁸ cells per gram of soil led to the decrease of C_T values from 36.3 to 31.2. The addition of Salmonella at concentrations higher than 2 × 10⁸ cells per gram of soil reduced the yield of mRNA as evidenced by the increase of C_T values compared with that of 2 × 10⁸ cells per gram of soil. One-way ANOVA on the C_T value was performed to compare the efficiencies of RNA_Y and Salmonella at its optimum concentration on RNA extraction from sandy soil. No significant difference was observed between mean C_T values (P = 0.35). No amplification was observed for loamy soil irrespective of which facilitator was included.

Table 6.

Detection of hsp70 mRNA in C. parvum oocysts spiked in 0.5 g of sandy soil extracted with method 5 in the presence of RNA_Y or Salmonella and the C_T value of IPC RNA^a

Facilitator added	Mean CT value (CT ± SD)
	hsp70 mRNA	IPC RNA
RNAY, 2 mg	36.94 ± 0.42	35.03 ± 0.59
RNAY, 5 mg	33.19 ± 0.06	34.31 ± 0.15
RNAY, 10 mg	30.58 ± 0.52	34.57 ± 0.11
RNAY, 20 mg	30.62 ± 0.34	34.69 ± 0.22
RNAY, 40 mg	30.63 ± 0.17	34.73 ± 0.12
Salmonella (1 × 107 cells)	36.31 ± 0.60	34.14 ± 0.37
Salmonella (2 × 107 cells)	35.16 ± 0.13	34.58 ± 0.09
Salmonella (5 × 107 cells)	33.02 ± 0.16	35.16 ± 0.34
Salmonella (1 × 108 cells)	31.20 ± 0.07	35.14 ± 0.22
Salmonella (2 × 108 cells)	32.38 ± 0.43	34.64 ± 0.19

^aA minimum of 5 replicates were used.

The effect of Salmonella in preventing RNA degradation during extraction is shown in Table 7. Without the addition of Salmonella, the difference between C_T values obtained using oligo(dT)₁₆ and random primer was 3.16; with the addition of 2 × 10⁷ Salmonella cells/g of soil, this difference lowered to 0.39; and in the presence of 1 × 10⁸ Salmonella cells/g of soil, the difference lowered even more to −0.12.

Table 7.

Detection of hsp70 mRNA in C. parvum oocysts spiked in 0.5 g of quartz sand extracted with method 1 in the presence of different amounts of Salmonella by qRT-PCR^a

No. of Salmonella cells	Mean CT value
	Random primer	Oligo(dT)16	Difference
0	33.46	36.62	3.16
1 × 107	31.84	32.23	0.39
5 × 107	30.99	30.87	−0.12
1 × 108	31.36	31.25	−0.11

^aA minimum of 5 replicates were used. Two types of primers [random or oligo(dT)₁₆] were used in the reverse transcription reaction.

Inhibitor removal performance assessment.

RNA extracts by method 5 from loamy soil spiked with 3 × 10⁵ C. parvum oocysts and different concentrations of RNA_Y or Salmonella were further purified with either MicroSpin S-400 HR spin column or RNeasy MinElute cleanup kit to remove RT-PCR inhibitor. hsp70 mRNA in approximately 43% (13/30) of the extract was detected after purification with the MicroSpin S-400 HR spin column. Purification using the RNeasy MinElute method resulted in hsp70 mRNA in 87% (26/30) of the extracts being detected.

Soil spiked with oocysts following pretreatment with Al₂(SO₄)₃ was subjected to RNA extraction using method 5 in the presence of RNA_Y at an optimal concentration. No amplification was observed in RT-PCR for any type of the soil. IPC RNA, however, showed amplification for all of the three environmental soils.

Comparison of (m)RNA extraction efficiencies of 5 methods using modified protocols.

The following modifications were thus made to the manufacturers' recommended protocols to optimize the extraction of RNA from soils. BME (20-μl/ml solution for sandy soil, and 40-μl/ml solution for loamy and clay soil) and EDTA with a final concentration of 50 mM were included in the lysis solution for all methods. Salmonella was used as the RNA extraction facilitator at concentrations of 2 × 10⁸, 4 × 10⁸, and 8 × 10⁸ cells per gram for sandy, loamy, and clay soil, respectively. After DNase treatment to remove the residual DNA, the RNA extract was used directly for TaqMan reverse transcription-PCR. For samples that did not show successful amplification, an additional purification using RNeasy MinElute cleanup kit was performed before RT-PCR. For method 2, the garnet bead in the lysis tube was replaced by a bead mixture of various diameters (2 mm, 0.4 g; 1 mm, 0.4 g; 0.5 mm, 0.4 g) before extraction to ensure higher (m)RNA yield. The comparative efficiencies of the five RNA extraction methods were evaluated on three environmental soils spiked with 3 × 10⁵ oocysts of C. parvum.

For sandy soil, successful amplifications were achieved using methods 1, 2, 4, and 5 without further purification. Mean C_T values were 31.7, 30.5, 34.5, and 30.3 for methods 1, 2, 4, and 5, respectively. RNA extracted by method 3 required further purifications to be detected by RT-PCR (C_T value 38.9). One-way ANOVA on the C_T value indicated that the difference in extraction efficiencies of the five RNA extraction methods was significant (P < 0.05). Post hoc Turkey's test revealed no significant differences between C_T values for method 2 and method 5 (P = 0.097). The differences, however, were significant when C_T values for the two methods were compared to C_T values for the other three methods (P < 0.05).

RNA extracts from loamy soil by methods 2 and 4 were RT-PCR amplifiable without further purification, with C_T values of 32.7 and 37.44, respectively. Purification was required for RNA extracted from loamy soil by methods 1, 3, and 5 to be successfully amplified by RT-PCR. C_T values for methods 1, 3, and 5 after purification were 39.1, 40.2, and 36.6, respectively. Significant differences in the extraction efficiency were revealed by one-way ANOVA on C_T values (P < 0.05). Pairwise comparisons of C_T values by post hoc tests indicated significant differences among all comparisons.

RNA extractions from clay appeared to be the most difficult. With the exception of method 2, no other method produced RT-PCR suitable RNA extract, even after the additional purification by RNeasy MinElute cleanup kit. The mean C_T value for extracts by method 2 was 34.0.

Detection limits are defined as the lowest oocyst level for which all three replicates can be detected as indicated by the C_T value. Table 8 shows the detection limit of RT-PCR using mRNA extracted from three soils by all five methods. Methods 2 and 5 yielded mRNA detections at the lowest concentration by RT-PCR from sandy soil. Detection limits using mRNA extracted from sandy soil with methods 2 and 5 were 1.5 × 10² and 6 × 10² oocysts per gram of soil, respectively. Method 2 yielded mRNA with a detection limit of 1.5 × 10³ oocysts per gram of loamy soil, which was the lowest concentration among all methods. The detection limit for clay was 1.5 × 10⁴ oocysts per gram of soil by using method 2.

Table 8.

Limit of detection of qRT-PCR using C. parvum hsp70 mRNA extracted from three soils by various extraction methods^a

Extraction method	Soil type
	Sandy		Loamy		Clay
	LODb	Avg CTc	LODb	Avg CTc	LODb	Avg CTc
Method 1	6 × 103	38.65 ± 0.66	6 × 105	39.16 ± 0.80	n/a	n/a
Method 2	1.5 × 102	40.33 ± 0.09	1.5 × 103	38.00 ± 0.90	1.5 × 104	37.30 ± 0.61
Method 3	1.5 × 105	38.92 ± 0.27	1.5 × 105	40.2d	n/a	n/a
Method 4	6 × 104	37.76 ± 0.25	6 × 105	37.44 ± 0.29	n/a	n/a
Method 5	6 × 102	40.84 ± 0.45	6 × 104	39.15 ± 0.16	n/a	n/a

^aA minimum of 3 replicates was used. n/a, not applicable.

^bThe limit of detection (LOD) was defined as the lowest concentration for which 3 out of 3 replicates produced a positive result (unit: oocysts g⁻¹).

^cAverage C_T values are expressed as mean ± standard deviation.

^dOnly 1 out of 3 replicates is RT-PCR detectable.

DISCUSSION

Information concerning the viability of C. parvum oocysts is required when developing a rapid estimate of a risk to human health. Heat shock protein (hsp) mRNA decays shortly after cell death (40); therefore, it can be correlated with the viability of C. parvum oocysts (63) and used as a viability marker. The advantage of targeting hsp70 mRNA is that the transcripts are present in higher copy numbers than with other mRNAs, which improves the overall sensitivity of detection. While reverse transcription-PCR (RT-PCR) targeting hsp70 mRNA has been proposed in determining the viability of C. parvum oocysts in water (24, 29, 63) or manure samples (21), this study is the first for the detection of viable oocysts in soil. The method relies largely on the efficiency of RNA extraction from a soil matrix as well as the reduction of inhibitors.

The literature describes only one other investigation where RNA extraction was evaluated with respect to soil samples (14); however, in that study RNA extraction kits were compared utilizing soil spiked with high concentrations of virus MS2 RNA (more than 10⁷ PFU g⁻¹ soil). It is not typical for human pathogens, such as C. parvum oocysts, to be present in soil in such high concentrations. In this study, five RNA isolation methods were evaluated for extracting (m)RNA from C. parvum oocysts in sandy, loamy, or clay soil by using real-time RT-PCR targeting hsp70 mRNA.

The failure to detect hsp70 mRNA by RT-PCR from all samples in the initial extraction attempt might have been caused by the inhibitory effect of humic substances coextracted from soil or the low recovery of m(RNA) in the extracts. Amplification of IPC RNA for some of the extracts (Fig. 1) indicated that the inhibitory effect was not the main reason for the failure of amplification, at least for samples that showed positive IPC RNA signals. This was further supported by the observation that, even after purification or dilution to relieve the inhibition, hsp70 mRNA still could not be detected. The amount of mRNA in the 3 × 10⁵ oocysts of C. parvum spiked was sufficient to be detected in our experimental setting when the recovery rate was high, since a positive control using mRNA extract from an isolate of C. parvum oocysts of the same number yielded a C_T value of 26.5 ± 0.87.

Factors that may have contributed to the low (m)RNA recovery rate include insufficient lysis of oocysts, degradation of the released (m)RNA by RNase in the soil matrix, or the adsorption of released RNA on soil components (clay, sand mineral, and organic matter). It is unlikely that the oocysts were incompletely disrupted since bead beating was used for all five extraction methods. This technique has been recommended to disrupt cells in soil aggregates during nucleic acid extraction (49) and proven to be effective at disrupting oocysts (21). The chance of mRNA being degraded by RNase in a soil matrix is small since it was inactivated by BME before the lysis began. Therefore, the failure to detect (m)RNA by RT-PCR was probably caused by an unsuccessful extraction due to adsorption of (m)RNA on soil components after they were released from oocysts. The results of previous studies suggest that the adsorption of DNA on soil colloids poses a serious challenge to its extraction from soil (20, 53, 67). Compared to DNA, reports of RNA adsorption by soil components are relatively rare (64); however, in our study, two tests using pure sand and reference clay confirmed the binding of (m)RNA on soil components. Since the process of DNA/RNA adsorption on soil particles is prompted primarily by bivalent cations (54), the inclusion of 50 mM EDTA in the lysis solution partially alleviated the adsorption of RNA (lower C_T value) due to its chelating effect (25). EDTA alone proved inadequate to mitigate the binding of RNA on soil with clay contents higher than 10% (Table 4), since clay has a binding capacity much higher than that of sand.

Various additives have been shown to increase the efficiency of DNA extraction from soil by reducing its binding site availability (20, 45, 67, 69). In this study, the addition of additives (SM, casein, salmon DNA, BSA, RNA_Y) did improve the extraction of RNA to various extents. Although improving RNA extraction was the main focus in the selection of the appropriate facilitators, other factors, such as simplifying the method and lowering costs were also considered. RNA_Y was initially found to be the most efficient approach to deal with the adsorption of (m)RNA on soil colloids during RNA extraction using method 5. When total RNA extraction methods 1, 2, 3, and 4 were used, after the lysis step, the presence of excessive extraneous RNA far exceeded the capacity of binding matrix; however, the binding of target RNA was significantly lowered and thus influenced its detection (data not shown). Moreover, the high expense of RNA_Y increased the cost of detection by $24 per sample (three replicates). Therefore, RNA_Y was not used for further experiments.

The effectiveness of using Salmonella to increase the yield of (m)RNA during RNA extraction indicated that its debris during cell disruption, including cell wall components, protein, and nucleic acid, competed for binding sites on soil colloids and thus reduced the binding of C. parvum oocyst (m)RNA. The inclusion of 2 × 10⁸ Salmonella cells per gram of sandy soil has the same effect as 20 mg of RNA_Y in alleviating mRNA adsorption on soil components in the extraction of C. parvum mRNA from sandy soil by using method 5. In contrast to RNA_Y, the debris of Salmonella were removed during the isolation steps and thus did not interfere with RNA binding on the matrix during the purification steps with the total RNA extraction methods. So Salmonella provided an economical alternative to mitigate C. parvum (m)RNA adsorption during RNA extraction from soil. Furthermore, under natural scenarios, C. parvum is released into the environment in animal feces together with high concentrations of bacterial indicators (11). It is estimated that in 1 gram of feces there are approximately 10¹¹ bacteria (66). Therefore, this simple rationale of using prokaryotic biomass to assist eukaryotic nucleic acid extraction from a complex environmental matrix can be generalized and applied in similar studies.

The utility of Salmonella during the lysis step of the extraction not only improved the yield of (m)RNA, as evidenced by the decreased mean C_T values, but also reduced the degradation of mRNA, as demonstrated by declines in the differences of mean C_T values between oligo(dT)₁₆ and random primers (Table 7). The variation in the mRNA-primer hybridization as well as mRNA structural changes during the extraction procedures might also have accounted for the observed differences in mean C_T values by using different primer sets with or without Salmonella. Eukaryotic mRNA degradation is initiated by the shortening of the poly(A) tail (33), which results in the lower efficiency of reverse transcription and ultimately the higher C_T values in RT-PCR using oligo(dT)₁₆. However, the integrity of poly(A) tail rarely influences the efficiency of reverse transcription using random primers and hence the C_T value in RT-PCR. The observed difference of a mean C_T value of 3.16 in the absence of Salmonella was significantly reduced to 0.39 when 1 × 10⁷ Salmonella cells were used per gram of soil, suggesting less degradation or shedding of mRNA.

The inclusion of IPC RNA in reverse transcription enabled the identification of false-negative results due to the presence of inhibitory substances. Unlike previous studies that used a housekeeping gene RNA to monitor RNA extraction and amplification efficiency, our developed IPC RNA was not present in the environmental samples tested. Therefore, being independent of the substrate tested, the detection of IPC RNA by TaqMan reverse transcription-PCR suggested that the efficiency of extraction methods to remove inhibitors was sufficient. Nonamplification of IPC RNA indicated the interference of a PCR inhibitor to RT-PCR. When using housekeeping genes as an internal control, their exact initial amount is unknown in most cases due to different transcription levels. In some cases they might not be present in environmental samples, while in others they might far exceed the amount of target mRNA, resulting in a competitive reverse transcription and PCR amplification which significantly affects the sensitivity, precision, and accuracy of the assay. In the case of IPC RNA, the amount of IPC RNA is set at a low level to limit its influence.

All five RNA methods tested utilized a direct extraction procedure without isolating C. parvum oocysts from the soil matrix. Indirect procedures that extract RNA from microorganisms after they were separated from soil matrix have been suggested to improve purity by reducing humic contents (62), which is more appropriate for soil with high levels of humic substances. A commercial soil RNA indirect extraction kit is available for the effective isolation of bacterial RNA from soil (FastRNA Pro Soil-Indirect kit, MP Biomedicals, Solon, OH). Due to the size of C. parvum oocysts (∼5 μm), further amendments to this kit may make it more suitable for the extraction of C. parvum RNA from soil as both the centrifugation step and the cheesecloth filtration during the separation procedure could eliminate the oocysts from subsequent RNA extraction.

Method 2 (2-g sample) and method 5 (0.5-g sample) were found to be the most efficient with respect to extracting RNA from sandy soil. Differences in the amounts of soil processed are evidenced by using qRT-PCR to show that sensitivities of approximately 150 and 600 C. parvum oocysts per gram of soil were obtained (methods 2 and 5, respectively) using extracted RNA. For loam and clay, method 2 was the most efficient, with detection limits of 1.5 × 10³ and 1.5 × 10⁴ C. parvum oocysts per gram, respectively. Previous studies reported that single viable C. parvum oocysts had been detected by RT-PCR targeting hsp70 mRNA from environmental water samples (29, 63). Since water samples may not contain a complex environmental matrix which efficiently binds RNA during extraction and contain a lower content of PCR inhibitors, making comparisons between water samples and soil samples is untenable. The only comparable study (21) reported a detection limit of around 5,000 C. parvum oocysts per milliliter of manure by using quantitative reverse transcription-PCR targeting hsp70 mRNA. It should be noted, however, that the method used in that study is not suitable to be adopted for soil: (i) a sample size of 60 to 90 μl of manure is insignificant to be used for soil processing, (ii) the components and texture of soil matrix are more heterogonous and complex than are those of manure, and (iii) absorption of mRNA on soil is strong as compared to manure. Furthermore, in that study, before RNA extraction manure samples were washed three times to reduce dissolved inhibitory compounds. In our study, using the modified procedures, a comparable detection limit was observed without any pretreatment of loamy soil; a much lower detection limit was obtained for sandy soil. Since 30 to 100 C. parvum oocysts can cause human infections (15), further modifications can be made to our methods to detect lower numbers of oocysts in soil samples by reducing the influence of humic substance. For instance, mRNA extraction under low pH (pH 5.0) has been suggested to be essential to minimize the coextraction of humic acids from soil samples and to increase the stability of mRNA (47). The pH of the lysis buffer and phenol-chloroform mixture could be adjusted to 5.0 for soil samples with high humic acid content. Additionally, novel mutant Taq and Klen-taq enzymes that are resistant to PCR inhibitors have recently been identified (31). Combining our modified RNA extraction methods with RT-PCR using an inhibitor-resistant enzyme might also significantly facilitate the detection of viable C. parvum oocysts in soil samples with high inhibitor contents.

In conclusion, the development of C. parvum-specific TaqMan RT-PCR assays, as described in this study, provides a valuable new approach for the assessment of viable C. parvum in environmental soil samples. Using an optimized (m)RNA extraction method in combination with RT-PCR, viable C. parvum oocysts were detected from soil samples without the need for previous oocyst purification or any pretreatment to reduce inhibitory substances. The procedure from RNA extraction to RT-PCR can be completed within 8 h. In contrast to other viability-testing methods such as animal infectivity assays, in vitro cell culture, and in vitro excystation, our optimized detection methods provide much faster, less expensive, and more accurate tools to reveal the presence and viability of C. parvum oocysts in soil.

Soil is an important vehicle through which C. parvum oocysts can reach water sources (61) indirectly through runoff from land grazed by livestock (2, 30, 32), directly by runoff from soil containing manure, i.e., fertilizer, or vertical transport through preferential flow paths to groundwater (3, 12, 34, 35, 43). Information on the viability of C. parvum oocysts in soil will be important for rapid responses in making risk analysis and management decisions in cases of contamination due to unintentional release of C. parvum oocysts such as spreading manure on farms. The modified method combined with TaqMan RT-PCR might be an important step in the standardization of identification and detection of viable C. parvum oocysts in soil samples.