Development of a PCR-Enzyme Immunoassay Oligoprobe Detection Method for Toxoplasma gondii Oocysts, Incorporating PCR Controls

Abstract

Infections caused by Toxoplasma gondii are widely prevalent in animals and humans throughout the world. In the United States, an estimated 23% of adolescents and adults have laboratory evidence of T. gondii infection. T. gondii has been identified as a major opportunistic pathogen in immunocompromised individuals, in whom it can cause life-threatening disease. Water contaminated with feces from domestic cats or other felids may be an important source of human exposure to T. gondii oocysts. Because of the lack of information regarding the prevalence of T. gondii in surface waters, there is a clear need for a rapid, sensitive method to detect T. gondii from water. Currently available animal models and cell culture methods are time-consuming, expensive, and labor-intensive, requiring days or weeks for results to be obtained. Detection of T. gondii nucleic acid by PCR has become the preferred method. We have developed a PCR amplification and detection method for T. gondii oocyst nucleic acid that incorporates the use of hot-start amplification to reduce nonspecific primer annealing, uracil-N-glycosylase to prevent false-positive results due to carryover contamination, an internal standard control to identify false-negative results due to inadequate removal of sample inhibition, and PCR product oligoprobe confirmation using a nonradioactive DNA hybridization immunoassay. This method can provide positive, confirmed results in less than 1 day. Fewer than 50 oocysts can be detected following recovery of oocyst DNA. Development of a T. gondii oocyst PCR detection method will provide a useful technique to estimate the levels of T. gondii oocysts present in surface waters.

Infections by Toxoplasma gondii are widely prevalent in animals and humans throughout the world. In the United States, an estimated 23% of adolescents and adults have laboratory evidence of T. gondii infection (13). T. gondii has been identified as a major opportunistic pathogen in immunocompromised individuals, in whom it can cause life-threatening disease. Of the 750 deaths attributed to toxoplasmosis each year, 50% are believed to be foodborne, making toxoplasmosis the third leading cause of foodborne deaths in this country (20). In addition, infections in pregnant women can cause serious health problems to the fetus if the parasites are transmitted (i.e., congenital toxoplasmosis) and cause severe sequelae in the infant (e.g., mental retardation, blindness, and epilepsy). It is estimated that there are 400 to 4,000 cases of congenital toxoplasmosis occurring in the United States each year (13).

T. gondii is an obligate intracellular protozoan pathogen belonging to the phylum Apicomplexa. Other members of this phylum include the human pathogens Plasmodium and Cryptosporidium as well as the animal pathogens Eimeria and Sarcocystis. Members of the cat family (Felidae) are the definitive hosts of T. gondii, with many mammals (including humans) and birds serving as intermediate hosts. Animals acquire Toxoplasma by ingesting any of three infectious stages of the organism: rapidly multiplying forms called tachyzoites, quiescent bradyzoites that occupy cysts in infected tissue, or oocysts shed in feces (8). For humans, ingestion of tissue cysts in undercooked meat and ingestion of food or water contaminated with oocysts are the two major modes of transmission. However, in most instances the source and route of transmission have not been determined. Oocyst-induced infections are considered to be more severe in humans and other animals than are tissue cyst-induced infections (8). Widespread natural infection is possible because cats may excrete hundreds of millions of oocysts after ingesting only a few tissue cysts (7). Oocysts are resistant to environmental and chemical inactivation and can survive in moist conditions for months and even years (35). Water contaminated with feces from domestic cats or other felids may be an important source of human exposure (1-3, 12, 14, 34; S. M. Hall, A. Pandit, A. Golwilkar, and T. S. Williams, Letter, Lancet 354:486-487, 1999).

Because of the lack of information regarding the prevalence of T. gondii in surface waters, there is a clear need for a rapid, sensitive method to detect T. gondii from water. Currently available animal models and cell culture methods are time consuming, expensive, and labor-intensive, requiring days or weeks for results to be obtained (14, 25). Detection of T. gondii nucleic acid isolated from clinical and animal samples by PCR has become the preferred method (4, 10, 16, 18, 21, 22, 24, 25). All of these studies have focused on detection of DNA isolated from tachyzoites and bradyzoites. There has been no development of methods to recover T. gondii oocyst nucleic acid (the stage at which T. gondii is most likely to be present in environmental samples).

Recovery of a human pathogen such as T. gondii from environmental samples presents several unique problems. Although high levels of oocysts can be shed in the stool of a feline, there is a considerable dilution factor as these oocysts are dispersed into an aquatic environment such as a reservoir or stream. This dilution factor requires that large volumes of water be concentrated (usually by filtration) to recover an adequate number of oocysts for subsequent detection. While the pathogen of interest (viral or parasitic) is being concentrated, other substances (such as humic and fulvic acid, complex carbohydrates, metals, salts, and many uncharacterized compounds) inhibitory to subsequent pathogen detection methods may also be concentrated (16, 18, 28).

There are many methods that have been developed to remove or inactivate these inhibitors. These methods include immunomagnetic separation (e.g., antibody capture) (6, 15, 17, 28, 31), guanidine isothiocyanate digestion (32), proteinase K digestion (4, 11, 15, 24), Sephadex gel chromatography (27), and buffer washes (15). In many cases, it is impossible to remove all inhibitors from every sample. Because of incomplete inhibitor removal, it is important to incorporate an internal standard control into each PCR to identify false-negative results due to sample inhibition (17, 18, 24, 29, 31). When processing environmental samples, adequate detection sensitivity following pathogen concentration and purification must be obtained. Many researchers have used nested or seminested PCR to improve detection sensitivity. However, nested PCR is inherently problematic, giving rise to many reports describing laboratory contamination following the procedure (4, 9, 24). Combining PCR with oligoprobe hybridization can eliminate the need for nested PCR. Oligoprobe hybridization increases both the sensitivity of single-round PCR amplification (10, 22, 28) and the specificity of the assay by providing PCR-independent conformation that the amplified product originated from the pathogen of interest and was not a nonspecific, spurious product (10, 21, 30). A recent development in hybridization technology is the use of liquid-based hybridization such as a nonradioactive enzyme immunoassay (EIA) (11, 21, 22, 30). An additional technique, used to prevent PCR carryover contamination, incorporates the use of uracil-N-glycosylase (UNG) (11, 19, 24, 30), an enzyme which degrades uracil-containing PCR products prior to amplification. Contaminating PCR products from previous laboratory experiments (in which dTTP was replaced with dUTP) are degraded by UNG, thus eliminating this potential contamination route.

The purpose of this research was to develop a PCR amplification and detection method for T. gondii oocyst nucleic acid that incorporates the use of hot-start amplification to reduce nonspecific primer annealing, UNG to prevent false-positive results due to carryover contamination, an internal standard control to identify false-negative results due to inadequate removal of sample inhibition, and PCR product oligoprobe confirmation using EIA.

MATERIALS AND METHODS

T. gondii oocysts and plasmid DNA.

For optimization of T. gondii DNA recovery, amplification, and detection conditions, a plasmid containing full-length T. gondii strain RH genomic DNA (kindly provided by Vern Carruthers, Johns Hopkins University Bloomberg School of Public Health) was used. Live and heat-inactivated oocysts from T. gondii strain VEG (kindly provided by Alan Lindquist, U.S. Environmental Protection Agency) were used to confirm optimized PCR conditions and evaluate oocyst DNA recovery and detection efficiency. T. gondii oocysts were purified from feces obtained from a laboratory-reared cat colony. The oocysts were purified by sedimentation and flotation and stored in 2% H₂SO₄. Before being shipped, they were killed by heating to 55°C for at least 15 min. The suspension density was determined by hemocytometer after heat killing and serially diluted in water to the desired concentrations.

Oocyst enzyme digestion and phenol-chloroform DNA extraction.

Known numbers of oocysts in aqueous samples were mixed 1 to 5 with proteinase K lysis buffer (10 mg of proteinase K/ml, 120 mM NaCl, 10 mM Tris [pH 7.5], and 0.1% sodium dodecyl sulfate [SDS]). After mixing, the samples were incubated at 55°C for 1 h. After cooling to room temperature, an equal volume of phenol-chloroform-isoamyl alcohol (25:24:1) was added, and the samples were mixed and centrifuged (16,000 × g) for 2 min at room temperature. The aqueous phase mixture was then transferred to a clean tube, an equal volume of chloroform was added, and the samples were mixed and centrifuged as above. The aqueous phase mixture was then transferred to a clean tube, and 10 μg of glycogen (as a low-titer DNA carrier), sodium acetate (final concentration, 0.3 M), and 2 volumes of ice-cold 100% ethanol were added. The mixture was inverted 10 times and stored at −20°C for 60 min. The sample was then centrifuged 16,000 × g for 30 min at 4°C to pellet precipitated DNA. The supernatant was removed and discarded, 750 μl of 70% ethanol was added, and the sample was centrifuged at 16,000 × g for 15 min at 4°C. The supernatant was removed, and the pellet was air dried for 15 min. The DNA was then suspended in 50 μl of water. During methods optimization, 10- to 25-μl volumes of each sample were analyzed for oocyst DNA by PCR. After optimization, a volume of 10 μl was selected for all subsequent experiments.

Oocyst freeze-thaw DNA extraction.

Known numbers of oocysts in aqueous samples were subjected to three cycles of freezing for 3 min in a dry-ice and ethanol slurry and thawing for 5 min at 55°C. Released DNA was then analyzed directly by PCR.

Ohio River filtrates.

For evaluation of potential sample inhibition, mock filter eluates were obtained by following EPA method 1623 (a method developed for isolation of Giardia and Cryptosporidium from water). Briefly, 10-liter samples of Ohio River water (Cincinnati, Ohio), with a turbidity of 27 nephelometric turbidity units, were filtered through Envirochek capsule filters (Pall Gelman, Ann Arbor, Mich.) and eluted with an aqueous buffered salt and detergent solution according to EPA method 1623 (33). The eluate was centrifuged, and the supernatant was removed to a final packed pellet volume of 0.3 ml and stored at 4°C until use for evaluation of potential PCR inhibition of mock sample concentrates. The packed pellets were diluted 10-, 100-, and 1,000-fold. Portions of each dilution series (n = 4) were seeded with approximately 100 T. gondii oocysts. Oocyst-seeded portions were then processed by either enzymatic digestion-precipitation or by freeze-thaw.

T. gondii primers and probes.

The primers and a probe to the T. gondii B1 gene first described by Burg et al. (5) were used in this study. The sense primer (Tox1, 5′-GGAACTGCATCCGTTCATGAG-3′) and antisense primer (Tox2, 5′-TCTTTAAAGCGTTCGTGGTC-3′) generate a 193-bp product. To facilitate the use of EIA PCR amplicon detection (see below), the Tox1 primer was biotinylated (Tox1bio) at the 5′ end. Oligoprobes (Tox3, 5′-GGCGACCAATCTGCGAATACACC-3′ and Tox4, 5′-TCGTCAGTGACTGCAACCTATGC-3′) internal to Tox1bio and Tox2 were 5′ end labeled with digoxigenin (Tox3dig and Tox4dig, respectively) for Southern hybridization detection using the Boehringer Mannheim (Indianapolis, Ind.) End Labeling kit. In addition, Tox3 and Tox4 oligoprobes were 5′ end labeled with fluorescein isothiocyanate (FITC) (Tox3FITC and Tox4FITC) for EIA detection. Probes and primers were purchased from Invitrogen Custom Primers, Frederick, Md.

Internal standard control.

A 156-bp internal standard control was created by using primer-directed mutagenesis on full-length T. gondii region B1 DNA. Briefly, a 40-bp antisense primer (ToxIS, 3′-TCTTTAAAGCGTTCGTGGTCAAAGTTGCACAGATACTCAT-5′) containing the Tox2 primer sequence and 20 additional T. gondii specific nucleotides located 37 bp upstream from the Tox2 primer sequence was designed. The ToxIS and Tox1 primers were used to amplify T. gondii genomic DNA, generating a 156-bp internal control amplicon with a 37-bp deletion. Subsequently, T. gondii internal standard amplicons were cloned into the pCR 2.1-TOPO (Invitrogen, Carlsbad, Calif.) vector. T. gondii internal standard-containing vectors were transformed into chemically competent Top 10 Escherichia coli (Invitrogen, Carlsbad, Calif.). After overnight incubation at 37°C on selective plates, positive colonies were picked and transferred to a master plate. The presence of internal standard within the clone was confirmed by PCR, restriction enzyme analysis, and sequencing. After confirmation, a stock of internal standard DNA was produced by culturing the clones and extracting plasmid DNA by use of the QIAprep Spin Maxiprep kit (Qiagen, Valencia, Calif.). One microliter of a serial dilution of internal standard DNA, equivalent to 10 PCR units, was seeded into the PCR master mix.

PCR amplification.

Amplification was performed with a 50-μl reaction mixture containing: 2.5 U of AmpliTaq Gold polymerase (Applied Biosystems, Foster City, Calif.), 1× Gene Amp PCR Buffer II (Applied Biosystems), 2.5 mM MgCl₂ (Applied Biosystems), 0.5 U of Amperase UNG (Applied Biosystems), 300 μM dATP, dCTP, and dGTP, 600 μM dUTP (Promega Corporation, Madison, WI), 0.25 μM Tox1bio and Tox2 primers, and 10 μl of template. PCR was performed in a PTC-200 thermal cycler (MJ Research, Waltham, Mass.) with an initial UNG digestion period (10 min at 20°C), an AmpliTaq Gold activation interval (10 min at 95°C), followed by 40 cycles of repeated denaturation (30 s at 92°C), annealing (50 s at 55°C), and extension (30 s at 72°C), concluding with a final extension step for 7 min at 72°C. All PCR experiments contained appropriate positive and negative controls.

Southern hybridization and detection.

Southern hybridization and detection were performed according to the methods outlined by Sambrook and Russell (26). Briefly, after gel electrophoresis (1% agarose plus 1% NuSieve GTG agarose [2% total concentration], 110 V for 1.5 h, ethidium bromide [0.5 μg per ml of agarose] staining of amplicons), double-stranded PCR amplicons were denatured to single-stranded form, transferred to a positively charged nylon membrane, and fixed to the membrane by UV cross-linking using a total of 1,200 μJ. The membrane was prehybridized for 30 min at 54°C in prehybridization buffer (PerfectHyb Plus; Sigma, St. Louis, Mo.) and then placed into a hybridization solution (PerfectHyb Plus containing either 25 nM digoxigenin-labeled Tox3dig or Tox4dig). After hybridization for 1 h at 54°C, the membrane was washed twice at room temperature for 5 min in 2× SSC-0.1% SDS (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate) and twice for 15 min in 0.5× SSC-0.1% SDS. The hybridized probe was detected by use of the Genius 3 nucleic acid detection kit (Boehringer-Mannheim) according to the manufacturer's protocol. Digoxigenin-labeled marker VIII (Boehringer-Mannheim) was used for molecular-weight size determination following gel electrophoresis and Southern hybridization.

EIA amplicon detection.

A sandwich hybridization capture assay was used to detect PCR products with the Light Diagnostic Universal Amplicon detection assay (Chemicon International, Temecula, Calif.). A 10 nM concentration of Tox3FITC or Tox4FITC probe was used according to the manufacturer's instructions to detect T. gondii oocyst DNA or internal standard DNA, respectively. Briefly, biotinylated amplicons were generated by using the Tox1bio primer. Following PCR amplification, amplicons were heat denatured and a 10 nM concentration of Tox3FITC or Tox4FITC probe was hybridized with 11 μl of product at 37°C for 30 min. The solution was then transferred to a 96-well modified adavin-coated plate and incubated for an additional 30 min at 37°C. During incubation, biotinylated amplicons (containing hybridized probe) bind to the coated plate. Anti-FITC monoclonal antibody, conjugated to horseradish peroxidase, was subsequently allowed to react with the bound product for 30 min at 37°C. Any unbound conjugate was removed by washing, and horseradish peroxidase substrate was added and allowed to react for 10 min. The reaction was quenched by the addition of acid with subsequent detection of hybridized product at 450 nm using a spectrophotometer. Results for optical density at 450 nm (OD₄₅₀) that were greater than 0.1 were considered positive.

RESULTS

In initial experiments, low copy numbers of T. gondii plasmid DNA were used to optimize PCR conditions prior to evaluating oocyst DNA detection, inhibitor influence, and DNA losses associated with release and recovery from oocysts. PCR master mix concentrations were optimized on half-log₁₀ dilutions of T. gondii plasmid DNA to provide the best endpoint detection when performing PCR. Endpoints between 50 and 150 fg of T. gondii genomic DNA were consistently achieved. Southern transfer and probing assays confirmed the results determined by gel electrophoresis. Figure 1 is representative of the agreement consistently observed among PCR results evaluated by gel electrophoresis, Southern blot hybridization, and EIA. Lanes with visible bands at 193 bp on the gel were also positive for Southern blot hybridization and had positive EIA results (OD₄₅₀ > 0.10). All negative lanes on the gel were also negative for Southern blot hybridization and EIA analysis. In addition, reverse transcription-PCR products from hepatitis A virus (HAV) and poliovirus (PV) amplified with virus-specific biotinylated primers and confirmed with virus-specific probes were used to test the specificity of Tox3dig and Tox3FITC probes. Neither HAV nor PV amplicons hybridized to the T. gondii-specific probes (Fig. 1).

FIG. 1.

T. gondii half-log₁₀ end point dilution PCR using Tox1bio-Tox2 primers followed by Southern oligoprobing with Tox3dig and EIA detection with Tox3FITC with the inclusion of HAV and PV amplicons as specificity controls. M, digoxigenin-labeled molecular-weight marker VIII; lanes 1 to 3, 10 μl of 10^−7.5 to 10^−8.5 half-log ₁₀ dilution of T. gondii genomic DNA (1.5-μg/μl stock), respectively, amplified under standard PCR conditions; lane 4, HAV amplicons; lane 5, PV amplicons; lane 6, negative reagent control. (A) Two percent agarose gel electrophoresis of 10 μl of PCR product. (B) Southern transfer and Tox3dig probing of panel A. (C) Analysis by EIA using the Tox3FITC probe corresponding to amplicons for each lane.

The cutoff value between positive and negative EIA results was determined empirically for each EIA probe on a set of negative control amplicons. For Tox3FITC the average negative control OD₄₅₀ was 0.001 (standard deviation [SD] = 0.003; n = 20). For Tox4FITC the average negative control OD₄₅₀ was 0.002 (SD = 0.002; n = 7). The cutoff OD₄₅₀ value for positive samples for each experiment was calculated by averaging the EIA negative control and PCR negative control from that experiment and adding 0.03 (greater than 10 SD for any probe condition tested). If this cutoff value was less than 0.1, a more conservative value of 0.1 was used as the cutoff OD.

In an effort to determine the effectiveness of PCR contamination prevention by UNG digestion, increasing concentrations of T. gondii PCR product generated by using dUTP (replacing dTTP) were added to tubes containing UNG in the master mix. The UNG system was able to digest the PCR product over a 4-log₁₀ range (data not shown).

Another concern related to the use of UNG is the degradation of PCR products due to residual UNG enzymatic activity prior to product analysis by gel electrophoresis or EIA. To evaluate the stability of PCR products, equal portions of PCR amplicons generated by PCR amplification of half-log₁₀ T. gondii nucleic acid dilution series were compared by using EIA results obtained immediately following PCR and comparing them to EIA results for PCR product amplicons stored at −20, 4, or 25°C for 24 to 72 h. Storage at 4 or −20°C for up to 72 h did not affect the stability of PCR products. However, storage at room temperature caused the concentration of PCR products to markedly differ from baseline values (data not shown). PCR products generated by using dUTP but with no UNG in the master mix showed no changes in the concentration of the PCR product for any of the temperatures or times evaluated (data not shown).

The accuracy of T. gondii oocyst hemocytometer counts was evaluated in a series of dilution experiments (n = 5) resulting in a high degree of precision with a coefficient of variation of 0.06. The optimum recovery methods for T. gondii oocyst DNA in a clean system (laboratory reagent grade water) were enzymatic digestion with proteinase K followed by phenol-chloroform extraction and ethanol precipitation or by liberation of oocyst DNA by freeze-thaw extraction. Using either of these methods, nucleic acid from fewer than 50 oocysts was consistently detected following PCR amplification (Fig. 2).

FIG. 2.

T. gondii oocyst dilutions processed by enzymatic digestion and phenol-chloroform extraction-ethanol precipitation or freeze-thaw followed by PCR amplification. M, 100-bp molecular-weight marker; lanes 1, 3, and 5, T. gondii-specific amplicons from 408, 41, and 4 T. gondii oocysts, respectively, amplified after enzymatic digestion and phenol-chloroform extraction with ethanol precipitation; lanes 2, 4, and 6, T. gondii-specific amplicons from 408, 41, and 4 T. gondii oocysts, respectively, amplified after freeze-thaw; lane 7, negative reagent control. (A) Two percent agarose gel electrophoresis of 10 μl of PCR product. (B) Analysis by EIA using the Tox3FITC probe corresponding to T. gondii amplicons for each lane.

Internal standard template was added to the PCR master mix to provide a means to evaluate sample inhibition. When low to moderate levels (<3,160 PCR U) of internal standard were added to the PCR master mix, both wild-type (i.e., full-length, unmodified) T. gondii DNA and internal standard DNA could be detected concurrently without any loss of sensitivity (Fig. 3, lanes 1 to 9). The wild-type and internal standard-derived amplicons were differentiated by using selective probing with Tox3 and Tox4 following Southern transfer or EIA. Only low levels of internal standard are added to the PCR master mix to prevent competition of target T. gondii oocyst DNA by internal standard DNA. When ≥3,160 PCR U of internal standard were seeded into the PCR master mix, loss of T. gondii oocyst DNA detection sensitivity was observed, as noted by the absence of oocyst detection in lane 12 of Fig. 3.

FIG. 3.

Interaction between log₁₀ dilutions of T. gondii oocyst DNA and half-log₁₀ dilutions of internal standard DNA during a PCR using Tox1bio and Tox2 primers followed by gel electrophoresis, ethidium bromide staining, and detection by EIA with Tox3FITC and Tox4FITC. M, 100-bp molecular-weight marker; lanes 1, 4, 7, and 10, 5,125 oocysts per 50 μl of PCR mixture; lanes 2, 5, 8, and 11, 513 oocysts per 50 μl of PCR product; lanes 3, 6, 9, and 12, 52 oocysts per PCR; lanes 1 to 3, 100 PCR U of internal standard DNA; lanes 4 to 6, 316 PCR U of internal standard DNA; lanes 7 to 9, 1,000 PCR U of internal standard DNA; lanes 10 to 12, 3,160 PCR U of internal standard DNA; lane 13, negative reagent control; lane 14, internal standard DNA positive control. (A) Two percent agarose gel electrophoresis of 10 μl of PCR product. (B) Analysis by EIA using Tox3FITC probe corresponding to T. gondii amplicons for each lane. (C) Analysis by EIA using the Tox4FITC probe corresponding to internal standard and T. gondii amplicons.

Ohio River water mock filtrate dilutions were used to investigate how the presence of inhibitors in concentrated environmental samples would affect T. gondii oocyst recovery and detection. Packed pellets were diluted 10-, 100-, and 1,000-fold, seeded with oocysts, and then processed by either enzymatic digestion-precipitation or by the freeze-thaw method. Recovered DNA was subsequently analyzed by PCR amplification and EIA detection. Oocyst extraction methods varied in their ability to remove inhibitors and detect T. gondii oocysts. The freeze-thaw method was not as effective in removing inhibitors, with sample inhibition consistently present in 10- and 100-fold dilutions of packed-pellet suspensions (Fig. 4). The enzymatic digestion-precipitation method showed a 10-fold improvement in sensitivity, with inhibition present only at 10-fold dilutions of packed pellets (Fig. 4). Incorporation of the internal standard control enabled easy identification of sample inhibition (Fig. 4, lanes 1, 2, and 4).

FIG. 4.

T. gondii oocyst DNA amplification by PCR using Tox1bio and Tox2 primers followed by gel electrophoresis, ethidium bromide staining, and detection by EIA with Tox3FITC and Tox4FITC from log₁₀ dilutions of Ohio River water filtrates seeded with oocysts. Oocyst DNA was recovered by either proteinase K digestion with phenol-chloroform extraction or freeze-thaw extraction. M, 100-bp molecular-weight marker; lanes 1 to 8, 96 oocysts per 50 μl of PCR product; lanes 1, 3, 5, and 7, phenol-chloroform extraction and ethanol precipitation; lanes 2, 4, 6, and 8, freeze-thaw extraction; lanes 1 and 2, 10⁻¹ dilution of Ohio River water filtrate; lanes 3 and 4, 10⁻² dilution of Ohio River water filtrate; lanes 5 and 6, 10⁻³ dilution of Ohio River water filtrate; lanes 7 and 8, reagent-grade water; lane 9, internal standard positive control; lane 10, negative reagent control. (A) Two percent agarose gel electrophoresis of 10 μl of PCR product. (B) Analysis by EIA using the Tox3FITC probe corresponding to T. gondii amplicons for each lane. (C) Analysis by EIA using the Tox4FITC probe corresponding to internal standard and T. gondii amplicons.

DISCUSSION

This study is one of the first investigations describing the development and optimization of a rapid molecular detection method for T. gondii oocyst DNA. Little information has been gathered on the presence of oocysts in environmental waters, and evaluation of T. gondii oocyst contamination of water has been limited by a lack of reliable detection methodologies. Isaac-Renton et al. (14) attempted to isolate oocysts without success from a drinking water supply implicated in an outbreak of toxoplasmosis. Their method was based on modifications of a method developed for isolation of Cryptosporidium by large volume filtration with subsequent elution and concentration of oocysts from the filter. Processed samples were inoculated into female Swiss-Webster mice. After 60 days of observation, sera from the mice were tested by using a modified agglutination test for toxoplasmosis (23). However, their study made no effort to assess oocyst recovery efficiency during the development of oocyst isolation procedures prior to testing large volume water samples. Molecular techniques such as PCR show promise for the rapid detection of oocysts from the environment. However, prior to environmental isolation and detection of T. gondii oocysts, a robust and sensitive detection method must first be developed.

The methods used in this study incorporate many features that promote specificity and sensitivity in the detection of DNA from T. gondii oocysts. The use of hot-start PCR using Taq Gold polymerase is one such feature. Taq Gold requires heating to activate the enzyme prior to the first round of extension during PCR. Thus, nonspecific amplification that can occur due to polymerase activity at suboptimal temperatures during cycle initiation is eliminated, resulting in increased product yield and decreased nonspecific background.

Southern blot hybridization is a labor-intensive detection system that requires 1 to 2 days to complete. Newer detection assays have been developed that allow a more rapid evaluation of PCR results (19, 21, 22, 30). We developed a DNA EIA to be used in conjunction with the T. gondii PCR assay. This enzyme-linked immunosorbent assay-based format enables very rapid analysis of large numbers of PCR samples and is amenable to automation. As with other enzyme-linked immunosorbent assay-based procedures, it is important to determine and optimize the background absorbance for each probe. Probe concentration is a key factor during optimization. For the probes used in these experiments, parameters that provided good sensitivity and low background were selected. The EIA detected T. gondii DNA with equal sensitivity to Southern blot hybridization (Fig. 1). Nonspecific amplicons of similar size generated by using biotinylated primers did not cross-react with T. gondii-specific EIA probing (Fig. 1, lanes 4 and 5). A distinct advantage of the EIA is the elimination of the subjective interpretation of the presence or absence of bands on a gel or Southern blot. EIA provides a numerical result, in contrast to the band with various levels of intensity present on a gel, Southern, or slot blot. In an EIA, a numerical cutoff value is generated by the negative-control samples. Samples are either above or below this cutoff value. For our assay, an OD₄₅₀ of 0.1 was selected as a conservative cutoff value. In our experiments, the OD₄₅₀s for all negative samples were well below 0.1 and those for all positive samples (even those near endpoint detection) were well above 0.1. Results obtained by following more conventional detection (e.g., Southern blot hybridization) are more subjective, and interpretations of results can vary, especially when signal intensity is weak. The EIA detection achieves a significant savings in time, with results now available in less than 4 h instead of 1 or 2 days. It is important to note that the EIA method is not quantitative because it is used following the final round of PCR amplification (well after the plateau effect present in conventional PCR is observed). The absolute numerical EIA result is irrelevant; for example, in Fig. 2A, the band observed in lane 1 is brighter than that observed in lane 2; however, EIA analysis produced a higher number for lane 2 than for lane 1 (Fig. 2B).

An additional problem encountered by many laboratories is the potential of unamplified template becoming contaminated with PCR product that had been previously amplified in the laboratory. One strategy to prevent PCR product carryover consists of utilizing dUTP in place of dTTP in the PCR mix (19, 24). All samples are treated with UNG before PCR to destroy any dUTP-containing amplicons, which may have contaminated the sample during preparation. This prevents contaminating amplicons from being amplified. We have incorporated a carryover prevention method that uses a heat-labile UNG. Following the elimination of dUTP-containing amplicons by UNG, the UNG enzyme was inactivated by heat during the Taq Gold activation incubation at 95°C prior to amplification. Storage of PCR amplicons for up to 72 h at −20 or 4°C did not affect subsequent analysis. This feature is important because it permits PCR amplification during the evening with amplicon analysis done within the next 3 days. Storage of the PCR product at room temperature may allow any nondenatured residual UNG enzyme to degrade PCR products prior to analysis.

The problems of false-negative results associated with inhibition of PCR are well documented (17, 18, 24, 29, 31). For this reason, an internal standard specific for T. gondii DNA was developed. To simplify the PCR master mix and to provide similar primer annealing temperatures, the internal standard was designed to be amplified by the same primers used to amplify wild-type T. gondii DNA. The 37-bp deletion in the internal standard control contains the sequence corresponding to the Tox3 oligoprobe binding site. As a result, the Tox3 oligoprobe cannot bind to the internal standard control. The oligoprobe Tox4, which corresponds to the sequence internal to the Tox1bio and Tox2 primers for both native T. gondii region B1 amplicons and internal standard, is used as a positive control. Duplicate portions of amplicons must be analyzed by EIA to differentiate between internal standard seeded into the master mix and T. gondii oocyst DNA amplicons. The concentration of internal standard in the PCR master mix must be carefully controlled. When <3,160 PCR U of internal standard were seeded into the PCR master mix, there was no loss of sensitivity in detecting wild-type DNA. However, the presence of excessive internal standard template can lead to competition with the T. gondii oocyst DNA template for essential components in the PCR master mix (Fig. 3). If the T. gondii oocyst template is outcompeted by the internal standard template, a false-negative result could occur. An additional technique used for inhibitor evaluation is to seed internal standard into a separate PCR master mix and run two PCRs for each sample. Although this increases reagent consumption, it eliminates any potential competition between target nucleic acid and internal standard DNA. Care must be taken however, to maintain consistency between master mixes to avoid tube-to-tube variations.

Enzymatic digestion of T. gondii oocysts using proteinase K followed by phenol-chloroform extraction and ethanol precipitation of exposed DNA and the simpler method of freezing and thawing to disrupt oocyst cell wall to liberate the DNA were evaluated. As shown in Fig. 2, both of these methods were able to detect fewer than 50 oocysts in a clean system. However, when low levels of oocysts were seeded into dilutions of packed pellets obtained by filtering Ohio River water by EPA method 1623, the sensitivity of the freeze-thaw method was lower than that obtained with the extraction-elution method (Fig. 4). By incorporating an internal standard control into the assay, we were able to differentiate between the absence of oocysts (true-negative sample) and the presence of sample inhibition (false-negative sample). It was evident that when examining complex sample matrixes such as filtered river eluates, additional sample processing was required beyond a simple disruption of oocyst cell walls. Method 1623 uses Cryptosporidium- and Giardia-specific antibodies to further purify (oo)cyst-containing samples. To date, there are no commercially available antibodies that have been developed to isolate the oocyst form of T. gondii. Our laboratory is continuing to evaluate antibodies in order to develop a T. gondii oocyst antibody capture protocol. Development of such a protocol will require the screening of several closely related organisms (Neospora caninum, Sarcocystis spp., etc.).

In conclusion, we have developed a liquid-based PCR amplification and detection method that rapidly and reliably detects T. gondii oocyst nucleic acid. The method incorporates hot-start amplification to reduce nonspecific primer annealing, UNG to prevent false-positive results due to carry-over contamination, an internal standard to identify false-negative results due to inadequate removal of sample inhibition, inhibition removal via phase separation and extraction, and PCR product oligoprobe confirmation using EIA. This method can provide positive confirmed results in less than 1 day and detect fewer than 50 oocysts. The described T. gondii oocyst DNA recovery, amplification, and detection system will facilitate the development and optimization of a method for the isolation of low numbers of T. gondii oocysts from large-volume water samples.