Introduction

Diagnosis for parasitic diseases is primarily based on morphological identification of worms or eggs. However, practical diagnosis for larva migrans depends largely on serologic tests because of the difficulty of getting tissue-embedded larvae in biopsy materials. Ascarid nematode larvae are particularly minute and similar in morphology (Nichols 1956a, 1956b). Even when a section of larva is found in a histopathological specimen, the species identification is hard unless the typical morphological features can be observed in the section. Since identification of parasite species is important not only for diagnosis but also for epidemiological surveys for public health assessment, a simple and reliable method for the identification of tissue-embedded larvae is desirable. For this purpose we examined the applicability of a nucleotide sequence analysis using polymerase chain reaction (PCR) to identify ascarid larvae in tissues. Each parasite species has unique ribosomal DNA (rDNA) sequences which can be used as markers to distinguish them from closely related and/or morphologically similar species (Hillis and Dixon 1991; Chilton et al. 1995; Gasser et al. 1998). Two sequences of the internal transcribed spacers (ITS) regions of rDNA, ITS1 and ITS2, have been used as species-specific markers for genetic identification (Hoste et al. 1993; Campbell et al. 1995; Zhu et al. 2000). The ITS region sequences of various ascaridoid parasites are now available from DNA databases (DDBJ, EMBL, GenBank).

Materials and methods

Adult ascarid nematodes examined in this study were as follows: Ascaris lumbricoides, Alj-1 and -2 from humans in Miyazaki, Japan; A. lumbricoides, Alb-3 from humans in Bangladesh; A. suum, Asj-1 and -2 from pigs in Miyazaki, Japan; Toxocara canis, Tnj-1 from dogs in Miyazaki, Japan; T. cati, Tcj-1 from cats in Hokkaido, Japan. Because A. lumbricoides and A. suum can hardly be distinguished between (Crompton 2001), we designated the round worm found in humans as A. lumbricoides and that in pigs as A. suum in this study. Worms were obtained from hosts and examined morphologically, then frozen at −70°C or fixed in 70% ethanol. To prepare infective larvae of A. suum, the eggs were obtained from adult females and were incubated in 0.05% formalin at 27°C with aeration for over a month. Mechanical hatching of embryonated eggs was performed as described previously (Urban et al. 1981).

Total genomic DNA was extracted using proteinase K digestion and phenol-chloroform-isoamyl alcohol extraction. ITS1 and ITS2 were amplified by PCR using the following oligonucleotide primers: F2662, 5′-GGCAAAAGTCGTAACAAGGT-3′ (ITS1, forward); R3214, 5′-CTGCAATTCGCACTATTTATCG-3′ (ITS1, reverse); F3207, 5′-CGAGTATCGATGAAGAACGCAGC-3′ (ITS2, forward); R3720, 5′-ATATGCTTAAGTTCAGCGGG-3′ (ITS2, reverse). ITS2 primers used in this study were identical to LC1 and HC2 reported by Navajes et al. (1992). The numbers in the primer names designate the position of the 3′ end of the primer in the published Caenorhabditis elegans sequence (Ellis et al. 1986). PCR reactions were carried out for 45 cycles of 1 min at 92°C, 1 min at 52°C and 1 min at 72°C with PCR buffer including Ampli Taq DNA polymerase (Perkin Elmer, USA). PCR products were detected in 1% agarose gel stained with ethidium bromide and photographed with FAS-III (TOYOBO, Japan). The products were subjected to cycle sequencing using the Big-Dye Terminator cycle sequencing reaction kit (ABI, USA) and sequenced by an automated sequencer (model 310, ABI).

Results and discussion

ITS region sequences of the samples were compared with those registered in the GenBank, which confirmed that there were no differences within the species, then genetic distances between species were calculated by the Kimura-2-parameter method (Kimura 1980) (Table 1). Both ITS regions could be used to distinguish Ascaris parasites from Toxocara spp. An interspecific difference between T. canis and T. cati was observed in the ITS2 region. In the case of Ascaris parasites, no clear genetic distance was observed between A. lumbricoides and A. suum. The extreme similarity of the ITS regions of these two Ascaris has already been reported (Anderson 1995; Zhu et al. 1999). Nevertheless, the sequence difference in the ITS region of Ascaris from humans and pigs was reproducible, and the ITS analysis may be useful for the determination of the original host, i.e. humans or pigs.

Table 1 Genetic distances of ascarid nematodes estimated by internal transcribed spacer (ITS) a regions
Full size table

To determine the sensitivity of this method, a single larva of A. suum was identified under a dissecting microscope, fixed in 70% ethanol and mixed with approximately 0.005 or 0.01 g porcine liver. Both ITS regions from a single larva of A. suum were amplified (Fig. 1; PCR products for ITS1 and that of ITS2 are not shown) although the DNA extracts required dilution (1:50–100) before amplification. Since this method is sensitive and specific, it can be applied for the genetic identification of ascarid larvae embedded in host tissues.

Fig. 1
figure 1

Results of internal transcribed spacer region amplification by polymerase chain reaction. Lane 1 A single larva of Ascaris suum, lane 2 a single larva with approximately 0.005 g of porcine liver, lane 3 a single larva with approximately 0.01 g of porcine liver, lane 4 porcine liver alone. The molecular weight (bp) is indicated by ➙

Full size image

As a practical trial of the method, a morphologically unidentified nematode larva found in turkey liver was examined. The turkeys had been kept in the backyard of a patient with visceral larva migrans, probably of A. suum as determined by immunological methods. Because the patient sometimes consumed raw liver of the turkeys, they were suspected to be a source of the infection. Birds, including chickens, could be a paratenic host for A. suum (Permin et al. 2000) as well as T. canis (Galvin 1964; Nagakura et al. 1989). When the livers of the five turkeys were examined, a few minute nodules were found on the surface of all five livers, and the nematode larvae were microscopically detected in the lesions of four out of five turkey livers. After extraction, amplification and sequence analysis of the ITS regions, the larva embedded in a turkey liver was identified as T. canis (Table 2). Indeed, the turkeys has been kept right next to a number of adult dogs and puppies at the patient’s home.

Table 2 Genetic distances of the unidentified larva vs. ascarid nematodes estimated by ITS regionsa
Full size table

Isolation of larvae from biopsy materials by pepsin or trypsin digestion is possible, but this is bothersome and time-consuming work. Further the digestion is not appropriate for part of a larva in tissue. Processing the whole host tissue containing a larva is much more convenient and there is less risk of losing the parasites. The method described here is sensitive and parasite specific, and can be performed even for a part of a parasite, and is useful for the species identification of worms embedded in host tissue.

The nucleotide sequences determined in this study were deposited in DNA databases (GenBank, EMBL, DDBJ) with accession nos.  AB110019–AB110034.