Discrimination of gastrointestinal nematode eggs from crude fecal egg preparations by inhibitor-resistant conventional and real-time PCR

Abstract

Diagnosis of gastrointestinal nematodes relies predominantly on coproscopic methods such as flotation, Kato-Katz, McMaster or FLOTAC. Although FLOTAC allows accurate quantification, many nematode eggs can only be differentiated to genus or family level. Several molecular diagnostic tools discriminating closely related species suffer from high costs for DNA isolation from feces and limited sensitivity since most kits use only small amounts of feces (<1 g). A direct PCR from crude egg preparations was designed for full compatibility with FLOTAC to accurately quantify eggs per gram feces (epg) and determine species composition. Eggs were recovered from the flotation solution and concentrated by sieving. Lysis was achieved by repeated boiling and freezing cycles - only Trichuris eggs required additional mechanic disruption. Egg lysates were directly used as template for PCR with Phusion DNA polymerase which is particularly resistant to PCR inhibitors. Qualitative results were obtained with feces of goats, cattle, horses, swine, cats, dogs and mice. The finally established protocol was also compatible with quantitative real-time PCR in the presence of EvaGreen and no PCR inhibition was detectable when extracts were diluted at least fourfold. Sensitivity was comparable to DNA isolation protocols and spiked samples with five epg were reliably detected. For Toxocara cati a detection limit below one epg was demonstrated. It was possible to distinguish T. cati and Toxocara canis using high resolution melt (HRM) analysis, a rapid tool for species identification. In human samples, restriction fragment length polymorphism (RFLP) and HRM analysis were used to discriminate Necator americanus and Ancylostoma duodenale. The method is able to significantly improve molecular diagnosis of gastrointestinal nematodes by increasing speed and sensitivity while decreasing overall costs. For identification of species or resistance alleles, analysis of PCR products with many different post PCR methods can be used such as RFLP, reverse-line-blot, Sanger sequencing and HRM.

Figure 1. Flow chart comparing different egg purification protocols.

Steps that were used in all protocols are shown in yellow, variant purification protocol steps in either orange or red. The latter color indicates mandatory steps for successful amplification in the final protocol in method C. The fourth protocol was modified from by increasing the amount of feces to 10 g to allow comparison with the other approaches.

Figure 2. Qualitative identification of trichostrongylid nematodes of goats.

Eggs were purified from five different animals (goats numbered 1 to 5 with epgs of 1241, 178, 307, 210, and 65, respectively) using the final protocol with sieving but without sucrose gradient. Lanes 1–5 present results for the individual goats. Primer pairs used are indicated above each gel. Positive controls (+) contained 1 ng plasmid DNA with the ITS-2 of the target species cloned in pCR4TOPO. Negative controls (−) contained only water. M, marker (100 bp ladder, Fermentas). PCRs were performed at least three times producing identical results and PCRs from extracted DNA (directly from eggs) also identified the same species. All PCR fragments were verified by sequencing.

Figure 3. Compatibility of d-PCR with fecal samples from different animal species.

Cattle samples were analyzed with primers for C. oncophora (A) and O. ostertagi (B). The 28S rDNA (C, E), the ITS-1 (D) and the ITS-2 (F) primer pair were used for feces of horses, carnivores and swine. A Trichuris-specific ITS-2 primer pair was used to detect T. muris and T. vulpis in murine and canine fecal samples (G). Lanes with different numbers represent individual animals. Positive controls (+) contained 1 ng plasmid DNA with an insert containing the corresponding target sequence. Only for Trichuris, genomic DNA isolated from T. vulpis eggs was used as positive control., negative controls were set up with water instead of template. M, marker (100 bp ladder, Fermentas); T can, T. canis, T cat, T. cati; A t, A. tubaeformae, A c, A. caninum, U s, U. stenocephala.

Figure 4. Determination of amplification efficacy by real-time PCR.

Fourfold serial dilutions of fecal sample extract from a T. cati infected cat (1∶4, dilution factor 0.25 to 1∶1024, dilution factor 0.00098) and tenfold serial dilutions of plasmid DNA (10⁶ to 10¹ copies) were used as template for real-time PCR in the presence of EvaGreen fluorescence dye. (A) Amplification plots showing signal accumulation measured in relative fluorescence units (RFU) with increasing cycle number. (b) Regression curves for fecal samples (red) and plasmid DNA (blue) were calculated with GraphPad Prism 5. Goodness of fit in terms of R² and slopes (with 95% confidence intervals) are given and slopes were used to calculate PCR efficiencies (Eff). Both regressions curves are virtually parallel and no significant difference between slopes could be found. (c) Amplification efficacy was also calculated from the slopes of individual amplification plots with LinRegPCR. Individual efficacies for all fecal (dots) and all plasmid samples (squares) as well as means ± SD are presented. A Student’s t test was used to compare PCR efficiencies between both groups but no significant differences could be detected.

Figure 5. Parasite species identification by melting curve analysis.

Fecal extracts from samples containing T. canis, T. cati or a mixture of both species were used as template for real-time PCR in the presence of EvaGreen followed by a high resolution melting curve analysis. Melting peaks obtained from maxima in the plot of the first deviation of the fluorescence intensity d(RFU)/dT were 88.4–88.5°C for T. cati and 90.3–90.4°C for T. canis (a). Amplicons from the mixed samples showed both peaks. Melting curves were normalized (b) and a difference plot (c) was calculated by subtracting the mean of both T. canis samples from all individual curves. Results for one of three representative experiments are shown.

Figure 6. Comparison of different template preparation methods.

Fecal samples negative for nematode eggs were spiked with C. oncophora eggs aiming to obtain epgs of 250, 150, 100, 50, 25 and 5. Every sample was split in three parts and analyzed either by direct DNA isolation from 500 mg feces using a commercial kit or by concentration of eggs from 10 g feces including determination of actual epgs by FLOTAC. Eggs were either obtained by flotation followed by DNA extraction or by flotation and sieving followed by freeze-boiling (d-PCR). C_q values were plotted vs. actual epgs as determined by FLOTAC and semi-logarithmic regression curves were fitted in GraphPad Prism software.

Figure 7. Species determination for human hookworm samples.

(A) Eight human and one canine (A. caninum) sample (+) together with hookworm negative fecal extract (−) were analyzed in duplicate using the ITS-2 primer pair. N. am., Necator americanus; A. du., A. duodenale; M, marker (100 bp ladder, Fermentas). (B, C) Representative PCR products were purified and 150 ng were digested with RsaI and separated using the DNA1000 LabChip® which has a sizing range from 25–1000 bp. Theoretical fragment sizes for N. americanus are 313, 53, 40 and 11 bp and 159, 139, 35, 28, 28, 16 and 12 bp. In silico digestion of A. duodenale or A. caninum fragments leads to identical fragment sizes of 198 and 113 bp. The gel view is shown in (B) for five samples and the electropherogram in (C) for two samples.