- Email: [email protected]
Trichinella papuae n.sp. (Nematoda), a new non-encapsulated species from domestic and sylvatic swine of Papua New Guinea
International Journal for Parasitology 29 (1999) 1825±1839
Trichinella papuae n.sp. (Nematoda), a new nonencapsulated species from domestic and sylvatic swine of Papua New Guinea E. Pozio a, *, I.L. Owen b, G. La Rosa a, L. Sacchi c, P. Rossi a, S. Corona c a Laboratory of Parasitology, Istituto Superiore di SanitaÁ, viale Regina Elena 299, 00161 Rome, Italy National Veterinary Laboratory, Department of Agriculture and Livestock, PO Box 6372, Boroko, Papua New Guinea c Department of Animal Biology, University of Pavia, Piazza Botta 9, 27100 Pavia, Italy
b
Received 8 June 1999; received in revised form 27 July 1999; accepted 28 July 1999
Abstract Encapsulated and non-encapsulated species of the genus Trichinella are widespread in sylvatic animals in almost all zoogeographical regions. In sylvatic animals from Tasmania (Australian region), only the non-encapsulated species Trichinella pseudospiralis has been reported. Between 1988 and 1998, non-encapsulated larvae of Trichinella were detected in ®ve domestic pigs and six wild boars from a remote area of Papua New Guinea. Morphological, biological, and molecular studies carried out on one strain isolated from a wild boar in 1997 suggest that these parasites belong to a new species, which has been named Trichinella papuae n.sp. This species can be identi®ed by the morphology of muscle larvae, which lack a nurse cell in host muscles, and whose total length is one-third greater than that of the other non-encapsulated species, T. pseudospiralis. Adults of T. papuae do not cross with adults of the other species and genotypes. Muscle larvae of T. papuae are unable to infect birds, whereas those of T. pseudospiralis do. The expansion segment V of the large subunit of the ribosomal DNA diers from that of the other species and genotypes. All of these features allow for the easy identi®cation of T. papuae, even in poorly equipped laboratories. The discovery and identi®cation of a second non-encapsulated species in the Australian region strongly supports the existence of two evolutionary lines in the genus Trichinella, which dier in terms of the capacity of larvae to induce a modi®cation of the muscle cell into a nurse cell. # 1999 Australian Society for Parasitology Inc. Published by Elsevier Science Ltd. All rights reserved. Keywords: Molecular systematics; Morphology; Nematode taxonomy; Non-encapsulated larvae; Papua New Guinea; Ribosomal DNA; Sus scrofa; Trichinella papuae
1. Introduction
* Corresponding author: Tel.: +39-06-4990-2304; fax: +3906-4938-7065. E-mail address: [email protected] (E. Pozio)
Non-encapsulated Trichinella larvae were ®rst discovered in a raccoon dog from the Caucasus, Russia, in 1972, and they were described as a new species, Trichinella pseudospiralis Garkavi,
0020-7519/99/$20.00 # 1999 Australian Society for Parasitology Inc. Published by Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 0 - 7 5 1 9 ( 9 9 ) 0 0 1 3 5 - 6
1826
E. Pozio et al. / International Journal for Parasitology 29 (1999) 1825±1839
1972 [1]. For more than 10 years, these parasites were considered an enigma [2] because they do not induce the muscle cell to transform into a nurse cell and because they are able to infect birds, which are not susceptible to encapsulated species of the genus Trichinella. Further studies have shown that T. pseudospiralis is a cosmopolitan species infecting a wide range of animal hosts, including birds, marsupials, rodents, carnivores, swine, and humans [3±6]. Recently, molecular studies have revealed dierences in the ribosomal DNA among T. pseudospiralis isolates originating from the Palaearctic, Nearctic, and Australian regions [7]. There are few reports of the presence of Trichinella in sylvatic animals from the Australian region, and these reports refer exclusively to non-encapsulated Trichinella, identi®ed as T. pseudospiralis. The parasites were found in Tasmania in two species of birds, in four species of marsupials [3, 8, 9], and in a human being [10]. Encapsulated larvae, which were identi®ed as Trichinella spiralis [11], have been found only in domestic pigs and synanthropic rats from New Zealand [12], although this parasite was probably imported by humans. In the present paper, we describe the discovery and the characterisation of a new non-encapsulated Trichinella species isolated from a wild pig in Papua New Guinea (PNG). The discovery and identi®cation of a second non-encapsulated species in the Australian region supports the existence of two evolutionary lineages in the genus Trichinella, which dier in terms of the capacity of larvae to induce a modi®cation of the muscle cell into a nurse cell.
pigs (Sus scrofa, considered to be hybrids between Sus scrofa vittatus and Sus celebensis) [13] and 83 wild animals (seven reptiles; three birds; and 73 mammals, speci®cally, 31 marsupials, 18 rodents, six chiropterans, 17 herbivores, and one carnivore) originating from the Bula Plain, a ¯at savannah grassland (approximately 398 km2 in area) located near the village of Balamuk (Fig. 1). In the wet season (from December to May), the plain is largely submerged under water. The soil is not suitable for agricultural purposes, but the land is rich in wildlife and is part of a wildlife management area. Wild pigs are plentiful; they feed on the plain at night and spend the daytime in the forest. Non-encapsulated Trichinella larvae without in®ltration of in¯ammatory cells around the infected muscle ®bres (Fig. 2) were detected in six (8.8%) wild pigs from the adjacent Bula Plain between 1988 and 1998. No Trichinella larvae were detected in any of the domestic pigs from other regions of PNG or in the other 83 wild animals examined. 2.2. Parasite isolation The muscle larvae from one of the six wild pigs from the Bula Plain, killed in August 1997, were administered per os to three Swiss mice and found to be infective for two of these mice. Two
2. Materials and methods 2.1. Survey in Papua New Guinea After the discovery in 1988 of non-encapsulated Trichinella larvae in the muscles of ®ve domestic sows from the village of Balamuk (latitude 88 56 0 S; longitude 1418 16 0 E), these parasites were searched for in hundreds of domestic pigs from many regions of PNG and in 67 wild
Fig. 1. Map of the island of New Guinea showing the area of origin (Q) of domestic and wild pigs infected with Trichinella papuae in Papua New Guinea.
E. Pozio et al. / International Journal for Parasitology 29 (1999) 1825±1839
1827
Fig. 2. Sagittal section (A) and transverse section (B), stained with H&E, of two non-encapsulated larvae of Trichinella papuae in the diaphragm of an infected sow detected in the village of Balamuk (Papua New Guinea) in 1988. Note the absence of the collagen capsule. There is no in®ltration of in¯ammatory cells around the infected muscle ®bres. Scale bar = 21 mm.
months later, one of the two mice was found to harbour 435 non-encapsulated larvae (isolate code ISS572), which were then administered per os to ®ve CD1 immunosuppressed female mice (85 larvae/mouse). One month later, 1257 larvae were collected from these mice [reproductive capacity index (RCI) = 2.9] and were used to infect 12 CD1 female mice. After the 6th passage, the RCI stabilised at 5.3. To increase the number of larvae in Swiss mice from the 2nd to the 6th passage, each animal was immunosuppressed with cyclophosphamide (4 mg/mouse) at ÿ4, 0, 4, and 8 days p.i. Immunosuppressed animals were bred in a box with ®ltered air to avoid respiratory infections. Muscle larvae were collected at each passage by arti®cial digestion of skinned and
eviscerated mouse carcasses. The digestive ¯uid consisted of 0.5% hydrochloric acid, 0.5% pepsin 1:10 000, and PBS at 408C. Carcasses were minced in a blender with the digestive ¯uid at a concentration of 1:40 (w/vol) for 30 s. The digestive ¯uid containing the minced mouse carcasses was then incubated with magnetic stirring at 39± 408C for 10 min. Muscle larvae were then allowed to sediment at 39±408C for 15 min. The supernatant was discarded gently and the sediment was ®ltered with a sieve of 300 mm. Finally, muscle larvae were washed three times with warm PBS and counted. Some of them were used to infect other mice for the in vivo maintenance of the isolate, and the remainder was stored at ÿ308C.
1828
E. Pozio et al. / International Journal for Parasitology 29 (1999) 1825±1839
2.3. Morphology The morphology of larvae in muscles of the tongue of the ®rst infected domestic sow was studied in histological sections stained with H&E. To study the morphological features of adult and larval stages of the PNG strain, muscle larvae of the 8th passage in laboratory mice were collected after arti®cial digestion, ®xed in 70% ethyl alcohol, and stored in 70% ethanol. Male and female adult worms (syntypes, accession number ISS572) were collected from mouse intestines (8th passage) at 6 days p.i., and then ®xed in 70% ethyl alcohol and stored in 70% ethanol. Measurements were made of 30 adult males, 30 adult females, 30 larval-stage males, and 30 larval-stage females; these measurements included: total length (TL), width at the midbody (W), distance of the nerve ring from the anterior extremity of the nematode (NR), length of the oesophagus (OE), length of the stichosome (ST), distance of the vulva from the anterior extremity of the nematode (V), length of the uterus (UT), length of the ovary (OV), length of the testis (TS), length of the papillae (PAPL), width of the papillae (PAPW), length of the genital primordium (GP), length of the rectum (RE), and distance of the posterior margin of the genital primordium from the posterior extremity of the nematode (GPE). Measurements of worms of the PNG strain were compared to those of T. pseudospiralis and T. spiralis s.l. reported in the literature [1, 2]. 2.4. Ultrastructural study of muscle larvae in host tissue The ultrastructure of PNG larvae in the mouse diaphragm was studied and compared with muscle larvae of the reference strain of T. pseudospiralis (code ISS13). Diaphragm muscles were removed, cut into small pieces (1±21±2 mm), and ®xed in 5% glutaraldehyde in 0.1 M cacodylate buer (pH 7.2) for 4 h at 48C. The samples were then washed in cacodylate buer and post®xed with 1% OsO4 in the same buer for 1.5 h at 48C, dehydrated in ethanol, and embedded in Epon 812. For light microscope analysis, semi-
thin sections (0.5 mm) were stained with 0.5% toluidine blue and observed with a Zeiss photomicroscope III. Thin sections, stained with uranyl acetate and lead citrate, were examined with a Zeiss EM 900 TEM. 2.5. Cross-breeding Single male and female muscle larvae of the PNG strain were crossed with single male and female muscle larvae of the eight reference strains (see below). The muscle larvae were collected after arti®cial digestion according to standard procedures (see above), and were identi®ed under an inverted microscope on the basis of the following morphological characters, proposed by Belosevic and Dick [14] and by Britov [15]. Male: (1) intestinal bulb generally close to the dorsal surface; in some larvae, close to the ventral surface; (2) intestine crossing the gonad from the dorsal to the ventral surface; in some larvae, crossing the gonad from the ventral to the dorsal surface and then recrossing to the ventral surface; and (3) length of rectum approximately 40± 50 mm. Female: (1) intestinal bulb generally close to the ventral surface; (2) intestine on the ventral surface; in some larvae, intestine crossing the gonad from the ventral to the dorsal surface and then recrossing to the ventral surface; (3) length of rectum approximately 20±30 mm; and (4) presence of a thickened subcuticular layer in the region of the vulval primordium (i.e. on the dorsal surface at about two-thirds of the way along the stichosome). Ten mice (CD1 female of 25 g) were inoculated per os with a single male and female larva belonging to dierent genotypes by a stomach tube connected to a 1 ml syringe, whereas ®ve mice were infected with a single male and female larva belonging to the same genotype and used as control. To ensure that no larvae remained in the syringe or in the tube, both were rinsed twice with water, which was then examined under the inverted microscope. All mice that were infected with larvae belonging to two dierent strains were immunosuppressed with 4 mg of cyclophosphamide at ÿ4, 0, 4, and 8 days p.i., since the RCI (number of larvae collected after arti®cial digestion/number of larvae
E. Pozio et al. / International Journal for Parasitology 29 (1999) 1825±1839
given per os) of the PNG strain was found to be low in CD1 mice. Mice were killed on day 35 p.i., and the entire skinned and eviscerated carcasses were digested individually. 2.6. Reference strains Ten reference strains were used for crossbreeding experiments and molecular characterisation: one strain of T. spiralis (code ISS3), one strain of Trichinella nativa (code ISS10), one strain of Trichinella britovi (code ISS2), three strains of T. pseudospiralis (codes ISS13, ISS141, ISS470), one strain of Trichinella nelsoni (code ISS29), one strain of Trichinella murrelli (code ISS35), one strain of Trichinella T6 (code ISS34), and one strain of Trichinella T8 (code ISS124) [16±18]. 2.7. Infectivity to birds A total of 500 muscle larvae collected after arti®cial digestion of mouse carcasses infected with the PNG strain and with the T. pseudospiralis reference strain (code ISS13) were given per os to 20 chickens (10 for each strain) of 2 weeks of age. As a control, the same number of larvae was given per os to 10 Swiss CD1 female mice (®ve for each strain). Chickens and mice were killed at 35 days p.i., skinned, and eviscerated, and the entire carcasses were subjected to arti®cial digestion to evaluate the RCI. 2.8. Puri®cation of genomic DNA DNA was obtained from frozen puri®ed larvae by overnight incubation at 558C in extraction buer (0.2 M Tris, 0.3 M NaCl, 0.025 M EDTA) containing 3% SDS and 200 mg/ml proteinase K. Puri®cation of DNA was done by phenol and phenol±chloroform±isoamyl alcohol extraction, followed by ethanol precipitation. DNA was resuspended in TE buer (10 mM Tris and 1 mM EDTA) and frozen at ÿ208C for use in PCR ampli®cations.
1829
2.9. Polymerase chain reaction protocol Polymerase chain reaction was performed using Taq DNA polymerase, 10 PCR buer, and dNTPs, from Takara (Japan), preparing 50 ml containing a ®nal concentration of 1.5 mM MgCl2 (included in the Takara 10 PCR buer), 200 mM dNTPs, 50 pmol of each primer and 0.5 unit of Taq DNA polymerase. For ampli®cation, 50 ng of DNA were used. Ampli®cations consisted of 35 cycles, as follows: denaturation at 948C for 20 s, annealing at 568C for 30 s, and elongation at 728C for 60 s. The primer set oTsr1 (5 0 -CGA AAA CAT ACG ACA ACT GC-3 0 ) and oTsr4 (5 0 -GTT CCA TGT GAA CAG CAG T-3 0 ) was used; this set ampli®es a region within the lsrDNA, known also as the expansion segment V (ESV) [7]. The primer set SB5 (forward and 5 0 -GGTATAAGGGAAAGCCGGAA-3 0 reverse 5 0 -CACTCACATTGTAT GCCAAG-3 0 ), designed by Wu et al. [19] from a sequenced fragment obtained by random ampli®cation of polymorphic DNA, which ampli®es a fragment of 680 bp exclusively in T. pseudospiralis, was also used to amplify DNA from PNG larvae. 2.10. Sequencing and analysis The ampli®cation products with the primer set oTsr1/oTsr4 were directly sequenced by Perkin Elmer ABI Prism 310 Genetic Analyser after labelling with BigDye Terminator Cycle Sequencing Ready Reaction kit (Perkin Elmer), following the manufacturer's instructions. The primers used for sequencing were the same as those used for PCR ampli®cation (oTsr1 and oTsr4). The products of three independent ampli®cation tubes were completely sequenced in both directions to con®rm the nucleotide sequences. The OMIGA 1.1 sequence-analysis software program was used for the analysis of the nucleotide sequence and for the alignment by Clustal W. Where possible, the alignment was determined by eye. The sequence of the PNG ESV was compared with the sequences present in the GenBank database of BLAST service of the National Centre for Biotechnology Information (NCBI) [20].
1830
E. Pozio et al. / International Journal for Parasitology 29 (1999) 1825±1839
3. Results The morphological features of adults and larvae of both sexes are reported in Table 1. The average TL of the isolated larvae was 966 mm for males (range 816±1066) and 1002 mm for females (range 888±1138). The TL of adult males was found to be equal in size or smaller, compared with the male larval stage. Muscles infected with non-encapsulated PNG larvae showed dramatic changes in the ®bre architecture. The infective larvae appeared to be almost always located in the intracellular space, and no in®ltration of in¯ammatory cells was observed around infected muscle ®bres (Fig. 3A). In some cases, myo®laments were totally disarranged, with the consequent dislocation of mitochondria (Fig. 3B). In fact, a large multilayered cluster of vacuolated mitochondria was observed to have formed around the larva (Fig. 4A). A double-layered envelope covered the larval cuticle, and a host-derived dense plasmalemmalike layer appeared to be ®rmly attached to the
host cell (Fig. 4B). In some cases, a nurse-celllike structure was observed where the sarcomeres appeared to be replaced by a sarcoplasmic reticulum ®lled with cisternae, lamellar bodies, and ribosomes (Fig. 5). An intense in¯ammatory reaction was observed around a larva that was apparently located outside the muscle cell (Fig. 6A±C). Muscles infected with larvae of the T. pseudospiralis reference strain revealed similar modi®cations in the muscle cell±parasite complex. The ultrastructural changes in the muscle ®bre architecture included the disaggregation of myo®laments, structural modi®cations, clustering and vacuolation of mitochondria and formation of the nurse cell-like structure (Fig. 7A, B). Single PNG male and female adult worms did not cross in either direction with females or males of the six species or two genotypes tested, whereas mice in control experiments became infected (data not shown). Papua New Guinea larvae were unable to infect chickens, unlike T. pseudospiralis larvae, which had an RCI of 15.
Table 1 Morphological features of adults and larvae of both sexes of Trichinella papuae n.sp.a Featuresb
Adult male2S.D. (range)
Adult female2S.D. (range)
Larva male2S.D. (range)
Larva female2S.D. (range)
TL MW NR OE ST V UT OV TS PAPL PAPW GP RE GPE
9552 31 (802±1056) 312 4 (24±38) 552 11 (24±72) 1242 17 (82±149) 3842 41 (307±480) Ð Ð Ð 3542 49 (254±456) 192 4 (10±32) 122 2 (8±16) Ð Ð Ð
17622199 (1320±2165) 362 5 (29±48) 572 11 (34±82) 126 213 (106±154) 374 249 (307±494) 396 232 (331±456) 10022171 (571±1392) 166 252 (101±360) Ð Ð Ð Ð Ð Ð
966265 (816±1066) 3324 (24±43) 6229 (38±77) 156241 (96±278) 456251 (360±569) Ð Ð Ð Ð Ð Ð 257240 (168±360) 5329 (42±80) 76219 (40±132)
1002259 (888±1138) 3425 (24±43) 6429 (48±82) 159223 (120±260) 442259 (269±566) Ð Ð Ð Ð Ð Ð 345228 (288±394) 2424 (18±34) 43210 (30±76)
a Measurements (mm) were made on 30 adult males, 30 adult females, 30 larval-stage males, and 30 larval-stage females, collected at the 8th passage in laboratory mice. b TL, total length; MW, width at the midbody; NR, distance of the nerve ring from the anterior extremity of the nematode; OE, length of the oesophagus; ST, length of stichosome; V, distance of the vulva from the anterior extremity of the nematode; UT, length of the uterus; OV, length of the ovary; TS, length of the testis; PAPL, length of the papillae; PAPW, width of the papillae; GP, length of the genital primordium; RE, length of the rectum; and GPE, distance of the posterior margin of the genital primordium from the posterior extremity of the nematode.
E. Pozio et al. / International Journal for Parasitology 29 (1999) 1825±1839
1831
Fig. 3. (A) Oblique longitudinal section of an intracellular larva of Trichinella papuae in a mouse diaphragm 3 months p.i. Note the absence of the collagen capsule. There is no in®ltration of in¯ammatory cells around the infected muscle ®bres. Scale bar = 50 mm. (B) Transmission electron micrograph of an intracellular larva of T. papuae. Myo®laments are almost completely disaggregated and mitochondria appear to be scattered throughout the sarcolemma. M, mitochondria; P, parasite. Scale bar = 1.1 mm.
Nevertheless, larvae of both the PNG and T. pseudospiralis strains from the same batches were able to infect laboratory mice. The PCR ampli®cation of the ESV of the PNG strain using the primer set oTsr1/oTsr4 produced a single fragment of 240 bp. This fragment diers from those of T. pseudospiralis strains, in which polymorphism was observed, with three fragments ranging from 290 to 310 bp. It also diers from fragments of the other species and genotypes (Fig. 8). Unlike T. pseudospiralis isolates, no evidence of polymorphism was found in the ESV of the PNG strain. The primer set
SB5, speci®c for T. pseudospiralis, did not amplify any fragment from the DNA of the PNG strain (Fig. 9). The alignment of ESV sequences of the PNG strain, T. pseudospiralis strain from North America (code ISS470), and T. pseudospiralis strain from Russia (Garkavi's strain, code ISS13) showed that the three strains had some microsatellites, or portions of microsatellites, in common (Fig. 10). In the PNG strain, the TGG microsatellite and its 5 0 ¯anking region were found to be completely conserved; the GT1 and TTG microsatellites and their 5 0 ¯anking regions were par-
1832
E. Pozio et al. / International Journal for Parasitology 29 (1999) 1825±1839
Fig. 4. (A) Transmission electron micrograph of Trichinella papuae in a mouse diaphragm 3 months p.i. Sarcomeres are totally lacking. A thick cluster of mitochondria surrounds the larva. In most mitochondria, the matrix is beginning to become vacuolated, yet the cristae appear normal. Arrows, ribosomes; asterisk, Golgi complex; M, mitochondria; P, parasite. Scale bar = 1.1 mm. (B) Detail of (A) showing the muscle±parasite interface. A double-layered structure covers the cuticle (arrow) and a host-derived plasmalemma-like layer (arrowhead) appears to be ®rmly attached to the underlying mitochondrial cluster. The epi-, exo-, and mesocuticular regions are evident. E, exocuticle; Me, mesocuticle; H, hypodermis; M, muscle. Scale bar = 0.4 mm.
tially conserved; whereas the TGC and CA microsatellites were absent. A BLAST search in the GenBank database revealed no signi®cant homology with the PNG sequence. 4. Discussion Trichinella nematodes were ®rst discovered in PNG in August 1988, when non-encapsulated Trichinella larvae were observed in muscle tissue sections of ®ve domestic pigs from a small breed-
ing facility in the village of Balamuk. At the time, Trichinella parasites were isolated in laboratory rats and pigs, but these isolates were not maintained in laboratory animals. Adults of Trichinella collected from the gut of infected rats in 1988 are preserved under formalin at the Natural History Museum, London (ref. Nos. 1989, 1200 and 1220). The village of Balamuk and the Bula Plain are located in a remote region of western PNG; they have few facilities, a very small population of hunters and subsistence gardeners, and virtually no movement of animals
E. Pozio et al. / International Journal for Parasitology 29 (1999) 1825±1839
1833
Fig. 5. Transmission electron micrograph of an intracellular larva of Trichinella papuae in a mouse diaphragm 3 months p.i. Nurse cell-like structure: contractile elements are replaced by sarcoplasmic reticulum. C, cisternae; L, lamellar bodies; P, parasite; arrows, ribosome; V, vesicles. Scale bar = 1.2 mm.
outside of the area, suggesting the autochthonous origin of the parasite. We cannot rule out that the PNG strain of Trichinella originated from the adjacent area of Irian Jaya (Indonesia), given that there is no physical barrier between these two countries. Muscle larvae of both the PNG strain and T. pseudospiralis can be distinguished from those of the other species and genotypes of the genus Trichinella by the absence of a capsule around larvae in host muscles. The two strains can be distinguished from each other based on the TL of larvae; in fact, the TL of PNG muscle larvae falls within the TL of encapsulated species and genotypes and is about one-third greater than that of T. pseudospiralis. The presence of these morphological hallmarks is of great importance for taxonomic and epidemiological studies. The present investigation reveals the existence of ultrastructural similarities between the PNG strain and the reference strain for
T. pseudospiralis. The muscle-stage larva of the PNG strain lacks a collagen capsule, and the infected muscle ®bres undergo dramatic alterations in their micro-architecture, including: (1) loss of contractile elements; (2) clustering of mitochondria around the muscle larva; (3) enlargement of the sarcoplasmic reticulum; and (4) formation of a nurse-cell-like structure. The surface layers of the cuticle of PNG larvae appear to be ultrastructurally similar to the cuticular envelopes already described in T. pseudospiralis [21, 22]. The presence of a double membrane of host origin around PNG larvae may provide some protection for the worm from the host response; moreover, the absence of a collagen capsule does not prevent the larvae from migrating through the muscle ®bres. This situation may be harmful to the host because it does not limit the invasion of new muscle ®bres. The cellular response of the host observed around the larvae located outside the muscle cell may per-
1834
E. Pozio et al. / International Journal for Parasitology 29 (1999) 1825±1839
Fig. 6. In¯ammatory reaction around an extracellular larva of Trichinella papuae in a mouse diaphragm 3 months p.i. (A) Section of muscle infected with T. papuae: around the parasite (P) there is a cluster of in¯ammatory cells (Ic). Scale bar = 0.5 mm. (B) Transmission electron micrograph of the parasite (P) apparently located outside the muscle cell. Scale bar = 5 mm. (C) Transmission electron micrograph of an in¯ammatory zone. Arrows, collagen ®brils; G, eosinophil granulocyte. Scale bar = 5 mm.
E. Pozio et al. / International Journal for Parasitology 29 (1999) 1825±1839
1835
Fig. 7. Transmission electron micrographs of a larva of Trichinella pseudospiralis in a mouse diaphragm 3 months p.i. (A) The micrograph shows the disaggregation of myo®laments and concentration of mitochondria close to the larval surface. The inner matrix of mitochondria is beginning to become vacuolated. Arrows, collagen ®brils; F, ®broblast; M, mitochondria; P, parasite; arrowhead, sarcolemma. Scale bar = 1.2 mm. (B) The micrograph shows a nurse cell-like structure characterised by an elaborate system of smooth endoplasmic reticulum cisternae. C, cisternae; asterisk, larval envelope remnants; P, parasite; arrows, ribosomes; S, smooth endoplasmic reticulum. Scale bar = 5 mm.
haps reduce tissue damage. These observations are consistent with the severity of the clinical picture observed in persons infected with T. pseudospiralis in Thailand, some of whom died [6]. Adults of the PNG strain do not cross with adults of the other species or genotypes in Trichinella in either sense, suggesting a reproductive isolation not only with the encapsulated genotypes, but also with T. pseudospiralis, the other non-encapsulated species. Although the study of cross-breeding is of great importance in evaluating the relationship between two genotypes, data from the literature show con¯icting results as to whether
cross-breeding of Trichinella genotypes obtained in laboratory conditions can be translated to reallife settings [2, 23±25]; however, since our results were negative, this problem did not arise. Infectivity to birds, which is considered to be one of the most important biological features of T. pseudospiralis, is absent in the PNG strain, suggesting that the two strains have a dierent evolution. In fact, the capability of T. pseudospiralis to infect birds has been considered an important factor in the dispersion of this parasite in several continents, and at the same time an ancestral feature of this species [3]. When a non-
1836
E. Pozio et al. / International Journal for Parasitology 29 (1999) 1825±1839
Fig. 8. Polymerase chain reaction ampli®cation of the expansion segment V of DNA from Trichinella strains by the primer set oTsr1 and oTsr4. Lane 1, Trichinella pseudospiralis from Russia; lane 2, T. pseudospiralis from Alabama (USA); lane 3, T. pseudospiralis from Tasmania; lane 4, Papua New Guinea strain; lane 5, T. britovi; lane 6, T. spiralis; lane 7, T. nelsoni; lane L, 100 bp ladder (Amersham-Pharmacia Biotech).
encapsulated Trichinella was ®rst discovered in PNG, we believed that its presence on the island of New Guinea may have been related to birds. However, the results of experimental infection in chickens suggest that another means of introduction may exist. Considering that T. papuae has been detected only in wild and domestic swine and that this host species was introduced in New Guinea by human migrants from Indonesia and South East Asia approximately 6000±10 000 years ago [26], we speculate that this parasite evolved outside of New Guinea and was introduced with imported swine. The detection of Trichinella in wild animals from nearby regions of New Guinea will be of great importance for explaining the origin and the evolution of this parasite. Zarlenga and Dame [27] demonstrated the presence of a gap in the ribosomal RNA with respect to the ribosomal DNA in T. spiralis. This gap was due to excision of ESV, which seems to be of no use for the functionality of the ribo-
Fig. 9. Polymerase chain rection analysis of dierent species of Trichinella ampli®ed by the primer set SB5F and SB5R. Lane 1, Trichinella pseudospiralis from Russia; lane 2, T. pseudospiralis from Alabama (USA); lane 3, T. pseudospiralis from Tasmania; lane 4, T. papuae; lane 5, T. britovi; lane 6, T. spiralis; lane L, 100 bp ladder (Amersham-Pharmacia Biotech).
some. The ESV is then free to ®x mutations in Trichinella, as demonstrated by the analysis of ribosomal DNA sequences of dierent species in the genus, which show that the dierences are exclusively located inside the ESV [7, 27]. Molecular analysis of ESV shows that the PNG strain greatly diers from the non-encapsulated isolates of T. pseudospiralis, as well as from the other species and genotypes in Trichinella (data not shown). Based on morphological, biological, and molecular features, it can be concluded that parasites of the PNG strain belong to a new species which we propose naming Trichinella papuae n.sp. This is the tenth genotype (Trichinella T10) to be identi®ed in the genus Trichinella. In fact, in addition to the eight genotypes described by La Rosa et al. [11], the Japanese strains from sylvatic animals have recently been recognised as belonging to a new genotype (Trichinella T9) [28]. Trichinella pseudospiralis isolates have been detected several times in Asia (India, Kamchatka, Kazakhstan, and Thailand) [5, 6, 29] and in Europe (Caucasus and Tula regions of Russia, Italy and France) [1, 5, 30] (G La Rosa and E Pozio, unpublished data), once in North America (Alabama) [4], and in many animals from the Australian region (Tasmania) [8, 9]. In
Fig. 10. Multi-alignment of homologous expansion segment V (ESV) fragments from Trichinella pseudospiralis from North America (TpUSA, code ISS470), T. pseudospiralis from Russia (TpRussia, code ISS13), and Trichinella papuae from Papua New Guinea (PNG, code ISS572). U = upper band; L = lower band (see [7]); asterisk, the conservation of the base among all the examined strains; dot, the presence of a gap; horizontal bar, the PCR primer; boxed area, a microsatellite; arrows, boundaries of the ESV.
1838
E. Pozio et al. / International Journal for Parasitology 29 (1999) 1825±1839
South-East Asia and Australian regions, apart from the cosmopolitan species T. spiralis, which has been passively introduced by human activity, only T. pseudospiralis and T. papuae have been detected in sylvatic animals, suggesting that the evolutionary line of non-encapsulated larvae in the genus Trichinella could originate in this area of the world. These data suggest the presence of at least two well-dierentiated non-encapsulated species whose geographical distribution partially overlaps that of encapsulated species and genotypes. The capability of larvae of encapsulated species to induce the modi®cation of the muscle cell into a nurse cell is a very complex molecular mechanism, in part still unknown [31]. This mechanism is absent in T. pseudospiralis and in T. papuae, strongly suggesting the presence of two evolutionary lines in the Trichinella genus, indicating that their classi®cation should be reevaluated. Further studies are needed to evaluate the natural life-cycle and epidemiology of T. papuae as well as the distribution area, the natural reservoirs, and its role as a human pathogen. 5. Taxonomic summary 5.1. Trichinella papuae n.sp. Biological features: RCI in Swiss mice 5.3 at the 9th passage. No resistance to freezing. Type host: Sus scrofa, considered to be a hybrid between Sus scrofa vittatus and Sus celebensis. Other hosts: (1) natural, domestic pig (Sus scrofa); (2) laboratory, Swiss mouse. Site: small intestine (adults) and skeletal muscle (larvae). Morphological features: non-encapsulated male and female larvae in muscle tissues have an average total length of 966 mm (range 816±1066) and 1002 mm (range 888±1138), respectively. Males appear to be equal in size or smaller at the adult stage, compared with the larval stage. Molecular features: expansion segment V of the ribosomal DNA of 240 bp. Type locality: Bula Plain, Papua New Guinea, latitude 88 56 0 S, longitude 1418 16 0 E.
Distribution: unknown. Specimens deposited: syntypes, adult males and females, International Trichinella Reference Centre at the Istituto Superiore di SanitaÁ, Italy (ref. No. ISS572). Adults and ML from this strain are also maintained in CD1 female mice. Etymology: the species is named for the country of origin.
Acknowledgements We wish to thank F. Mancini Barbieri and M. Amati for their technical support. The collection of biological material in Papua New Guinea was made possible by the support received from the Northern Australia Quarantine Strategy (NAQS).
References [1] Garkavi BL. Species of Trichinella isolates from wild animals. Veterinariya 1972;10:90±1. [2] Dick TA. Species, and infraspeci®c variation. In: Campbell WC, editor. Trichinella and trichinellosis. New York: Plenum Press, 1983;31±73. [3] Pozio E, Shaikenov B, La Rosa G, Obendorf DL. Allozymic and biological characters of Trichinella pseudospiralis isolates from free-ranging animals. J Parasitol 1992;78:1087±90. [4] Lindsay DS, Zarlenga DS, Gamble HR, Al-Yaman F, Smith PC, Blagburn BL. Isolation and characterization of Trichinella pseudospiralis Garkavi, 1972 from a black vulture (Coragypus atratus). J Parasitol 1995;81:920±3. [5] Britov VA. Trichinellosis in Kamchatka. Wiad Parazyt 1997;43:287±8. [6] Jongwutiwes S, Chantachum N, Kraivichian P et al. First outbreak of human trichinellosis caused by Trichinella pseudospiralis. Clin Infect Dis 1998;26:111±5. [7] Zarlenga DS, Aschenbrenner RA, Lichtenfels JR. Variations in microsatellite sequences provide evidence for population dierences and multiple ribosomal gene repeats within Trichinella pseudospiralis. J Parasitol 1996;82:534±8. [8] Obendorf DL, Clarke KP. Trichinella pseudospiralis infections in free-living Tasmanian birds. J Helminthol Soc Wash 1992;59:144±7. [9] Obendorf DL, Handlinger JH, Mason RW et al. Trichinella pseudospiralis infection in Tasmanian wildlife. Aust Vet J 1990;67:108±10. [10] Andrews JRH, Bandi C, Pozio E, Gomez Morales MG, Ainsworth R, Abernethy D. Identi®cation of Trichinella
E. Pozio et al. / International Journal for Parasitology 29 (1999) 1825±1839
[11] [12] [13]
[14] [15] [16] [17] [18] [19] [20]
[21]
pseudospiralis from a human case using random ampli®ed polymorphic DNA. Am J Trop Med Hyg 1995;53:185±8. La Rosa G, Pozio E, Rossi P, Murrell KD. Allozyme analysis of Trichinella isolates from various host species and geographic regions. J Parasitol 1992;78:641±6. Buncic S. A case of a pig infested with Trichinella spiralis. Surveillance 1997;24:8. Groves C. Ancestors for the pigs: taxonomy and phylogeny of the genus Sus. Technical Bulletin No. 3, Department of Prehistory, Research School of Paci®c Studies, Australian National University, 1981;1±96. Belosevic M, Dick TA. Trichinella spiralis: comparison with an arctic isolate. Exp Parasitol 1980;49:266±76. Britov VA. Etiological agents of trichinellosis. Moscow: Nauka, 1982 [in Russian]. Pozio E, La Rosa G, Murrell KD, Lichtenfels JR. Taxonomic revision of the genus Trichinella. J Parasitol 1992;78:654±9. Pozio E, La Rosa G. Trichinella murrelli n.sp: etiological agent of sylvatic trichinellosis in temperate areas of North America. J Parasitol, in press. Pozio E, La Rosa G, Rossi P. Trichinella Reference Centre. Parasitol Today 1989;5:169±70. Wu Z, Nagano I, Fukumoto S, Saito S et al. Polymerase chain reaction primers to identify Trichinella spiralis or T. pseudospiralis. Parasitol Int 1997;46:149±54. Altschul SF, Madden TL, Schaer AA et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 1997;25:3389±402. Lee DL, Wright KA, Shivers RR. A freeze±fracture study of the surface of the infective-stage larva of the Nematode Trichinella. Tissue Cell 1984;16:819±28.
1839
[22] Ubelaker JE, Stewart GL, Martin JH. Modi®cation of the ultrastructure of the muscle larva of Trichinella pseudospiralis following exposure to acidi®ed pepsin solution. J Parasitol 1993;79:133±7. [23] Bessonov AS, Penkova RA, Uspensky AV. On the independence of Trichinella species. Wiad Parazyt 1975;21:561±75. [24] Boev SN, Britov VA, Orlov IV. Species composition of Trichinellae. Wiad Parazyt 1979;25:495±503. [25] Ooi HK, Oku Y, Kamiya M, Ohbayashi M. Interbreeding and fecundity of a single pair of two strains of Trichinella spiralis in mice. Jpn J Vet Res 1984;32:177±82. [26] Flannery TF. Mammals of New Guinea. The Australian Museum, Carina, Qld: R Brown and Associates, 1990. [27] Zarlenga DS, Dame JB. The identi®cation and characterization of a break within the large subunit ribosomal RNA of Trichinella spiralis: comparison of gap sequences within the genus. Mol Biochem Parasitol 1992;51:281±90. [28] Nagano I, Wu Z, Matsuo A, Pozio E, Takahashi Y. Identi®cation of Trichinella isolates by polymerase chain reactionÐrestriction fragment length polymorphism of the mitochondrial cytochrome c oxidase. Int J Parasitol 1999;29:1113±20. [29] Shaikenov B, Boev SN. Distribution of Trichinella species in the Old World. Wiad Parazyt 1983;29:595±608. [30] Pozio E, Goredo M, Fico R, La Rosa G. Trichinella pseudospiralis in sedentary night-birds of prey from Central Italy. J Parasitol 1999;85:759±61. [31] Despommier DD. How does Trichinella spiralis make itself at home?. Parasitol Today 1998;14:318±23.
Trichinella papuae n.sp. (Nematoda), a new nonencapsulated species from domestic and sylvatic swine of Papua New Guinea E. Pozio a, *, I.L. Owen b, G. La Rosa a, L. Sacchi c, P. Rossi a, S. Corona c a Laboratory of Parasitology, Istituto Superiore di SanitaÁ, viale Regina Elena 299, 00161 Rome, Italy National Veterinary Laboratory, Department of Agriculture and Livestock, PO Box 6372, Boroko, Papua New Guinea c Department of Animal Biology, University of Pavia, Piazza Botta 9, 27100 Pavia, Italy
b
Received 8 June 1999; received in revised form 27 July 1999; accepted 28 July 1999
Abstract Encapsulated and non-encapsulated species of the genus Trichinella are widespread in sylvatic animals in almost all zoogeographical regions. In sylvatic animals from Tasmania (Australian region), only the non-encapsulated species Trichinella pseudospiralis has been reported. Between 1988 and 1998, non-encapsulated larvae of Trichinella were detected in ®ve domestic pigs and six wild boars from a remote area of Papua New Guinea. Morphological, biological, and molecular studies carried out on one strain isolated from a wild boar in 1997 suggest that these parasites belong to a new species, which has been named Trichinella papuae n.sp. This species can be identi®ed by the morphology of muscle larvae, which lack a nurse cell in host muscles, and whose total length is one-third greater than that of the other non-encapsulated species, T. pseudospiralis. Adults of T. papuae do not cross with adults of the other species and genotypes. Muscle larvae of T. papuae are unable to infect birds, whereas those of T. pseudospiralis do. The expansion segment V of the large subunit of the ribosomal DNA diers from that of the other species and genotypes. All of these features allow for the easy identi®cation of T. papuae, even in poorly equipped laboratories. The discovery and identi®cation of a second non-encapsulated species in the Australian region strongly supports the existence of two evolutionary lines in the genus Trichinella, which dier in terms of the capacity of larvae to induce a modi®cation of the muscle cell into a nurse cell. # 1999 Australian Society for Parasitology Inc. Published by Elsevier Science Ltd. All rights reserved. Keywords: Molecular systematics; Morphology; Nematode taxonomy; Non-encapsulated larvae; Papua New Guinea; Ribosomal DNA; Sus scrofa; Trichinella papuae
1. Introduction
* Corresponding author: Tel.: +39-06-4990-2304; fax: +3906-4938-7065. E-mail address: [email protected] (E. Pozio)
Non-encapsulated Trichinella larvae were ®rst discovered in a raccoon dog from the Caucasus, Russia, in 1972, and they were described as a new species, Trichinella pseudospiralis Garkavi,
0020-7519/99/$20.00 # 1999 Australian Society for Parasitology Inc. Published by Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 0 - 7 5 1 9 ( 9 9 ) 0 0 1 3 5 - 6
1826
E. Pozio et al. / International Journal for Parasitology 29 (1999) 1825±1839
1972 [1]. For more than 10 years, these parasites were considered an enigma [2] because they do not induce the muscle cell to transform into a nurse cell and because they are able to infect birds, which are not susceptible to encapsulated species of the genus Trichinella. Further studies have shown that T. pseudospiralis is a cosmopolitan species infecting a wide range of animal hosts, including birds, marsupials, rodents, carnivores, swine, and humans [3±6]. Recently, molecular studies have revealed dierences in the ribosomal DNA among T. pseudospiralis isolates originating from the Palaearctic, Nearctic, and Australian regions [7]. There are few reports of the presence of Trichinella in sylvatic animals from the Australian region, and these reports refer exclusively to non-encapsulated Trichinella, identi®ed as T. pseudospiralis. The parasites were found in Tasmania in two species of birds, in four species of marsupials [3, 8, 9], and in a human being [10]. Encapsulated larvae, which were identi®ed as Trichinella spiralis [11], have been found only in domestic pigs and synanthropic rats from New Zealand [12], although this parasite was probably imported by humans. In the present paper, we describe the discovery and the characterisation of a new non-encapsulated Trichinella species isolated from a wild pig in Papua New Guinea (PNG). The discovery and identi®cation of a second non-encapsulated species in the Australian region supports the existence of two evolutionary lineages in the genus Trichinella, which dier in terms of the capacity of larvae to induce a modi®cation of the muscle cell into a nurse cell.
pigs (Sus scrofa, considered to be hybrids between Sus scrofa vittatus and Sus celebensis) [13] and 83 wild animals (seven reptiles; three birds; and 73 mammals, speci®cally, 31 marsupials, 18 rodents, six chiropterans, 17 herbivores, and one carnivore) originating from the Bula Plain, a ¯at savannah grassland (approximately 398 km2 in area) located near the village of Balamuk (Fig. 1). In the wet season (from December to May), the plain is largely submerged under water. The soil is not suitable for agricultural purposes, but the land is rich in wildlife and is part of a wildlife management area. Wild pigs are plentiful; they feed on the plain at night and spend the daytime in the forest. Non-encapsulated Trichinella larvae without in®ltration of in¯ammatory cells around the infected muscle ®bres (Fig. 2) were detected in six (8.8%) wild pigs from the adjacent Bula Plain between 1988 and 1998. No Trichinella larvae were detected in any of the domestic pigs from other regions of PNG or in the other 83 wild animals examined. 2.2. Parasite isolation The muscle larvae from one of the six wild pigs from the Bula Plain, killed in August 1997, were administered per os to three Swiss mice and found to be infective for two of these mice. Two
2. Materials and methods 2.1. Survey in Papua New Guinea After the discovery in 1988 of non-encapsulated Trichinella larvae in the muscles of ®ve domestic sows from the village of Balamuk (latitude 88 56 0 S; longitude 1418 16 0 E), these parasites were searched for in hundreds of domestic pigs from many regions of PNG and in 67 wild
Fig. 1. Map of the island of New Guinea showing the area of origin (Q) of domestic and wild pigs infected with Trichinella papuae in Papua New Guinea.
E. Pozio et al. / International Journal for Parasitology 29 (1999) 1825±1839
1827
Fig. 2. Sagittal section (A) and transverse section (B), stained with H&E, of two non-encapsulated larvae of Trichinella papuae in the diaphragm of an infected sow detected in the village of Balamuk (Papua New Guinea) in 1988. Note the absence of the collagen capsule. There is no in®ltration of in¯ammatory cells around the infected muscle ®bres. Scale bar = 21 mm.
months later, one of the two mice was found to harbour 435 non-encapsulated larvae (isolate code ISS572), which were then administered per os to ®ve CD1 immunosuppressed female mice (85 larvae/mouse). One month later, 1257 larvae were collected from these mice [reproductive capacity index (RCI) = 2.9] and were used to infect 12 CD1 female mice. After the 6th passage, the RCI stabilised at 5.3. To increase the number of larvae in Swiss mice from the 2nd to the 6th passage, each animal was immunosuppressed with cyclophosphamide (4 mg/mouse) at ÿ4, 0, 4, and 8 days p.i. Immunosuppressed animals were bred in a box with ®ltered air to avoid respiratory infections. Muscle larvae were collected at each passage by arti®cial digestion of skinned and
eviscerated mouse carcasses. The digestive ¯uid consisted of 0.5% hydrochloric acid, 0.5% pepsin 1:10 000, and PBS at 408C. Carcasses were minced in a blender with the digestive ¯uid at a concentration of 1:40 (w/vol) for 30 s. The digestive ¯uid containing the minced mouse carcasses was then incubated with magnetic stirring at 39± 408C for 10 min. Muscle larvae were then allowed to sediment at 39±408C for 15 min. The supernatant was discarded gently and the sediment was ®ltered with a sieve of 300 mm. Finally, muscle larvae were washed three times with warm PBS and counted. Some of them were used to infect other mice for the in vivo maintenance of the isolate, and the remainder was stored at ÿ308C.
1828
E. Pozio et al. / International Journal for Parasitology 29 (1999) 1825±1839
2.3. Morphology The morphology of larvae in muscles of the tongue of the ®rst infected domestic sow was studied in histological sections stained with H&E. To study the morphological features of adult and larval stages of the PNG strain, muscle larvae of the 8th passage in laboratory mice were collected after arti®cial digestion, ®xed in 70% ethyl alcohol, and stored in 70% ethanol. Male and female adult worms (syntypes, accession number ISS572) were collected from mouse intestines (8th passage) at 6 days p.i., and then ®xed in 70% ethyl alcohol and stored in 70% ethanol. Measurements were made of 30 adult males, 30 adult females, 30 larval-stage males, and 30 larval-stage females; these measurements included: total length (TL), width at the midbody (W), distance of the nerve ring from the anterior extremity of the nematode (NR), length of the oesophagus (OE), length of the stichosome (ST), distance of the vulva from the anterior extremity of the nematode (V), length of the uterus (UT), length of the ovary (OV), length of the testis (TS), length of the papillae (PAPL), width of the papillae (PAPW), length of the genital primordium (GP), length of the rectum (RE), and distance of the posterior margin of the genital primordium from the posterior extremity of the nematode (GPE). Measurements of worms of the PNG strain were compared to those of T. pseudospiralis and T. spiralis s.l. reported in the literature [1, 2]. 2.4. Ultrastructural study of muscle larvae in host tissue The ultrastructure of PNG larvae in the mouse diaphragm was studied and compared with muscle larvae of the reference strain of T. pseudospiralis (code ISS13). Diaphragm muscles were removed, cut into small pieces (1±21±2 mm), and ®xed in 5% glutaraldehyde in 0.1 M cacodylate buer (pH 7.2) for 4 h at 48C. The samples were then washed in cacodylate buer and post®xed with 1% OsO4 in the same buer for 1.5 h at 48C, dehydrated in ethanol, and embedded in Epon 812. For light microscope analysis, semi-
thin sections (0.5 mm) were stained with 0.5% toluidine blue and observed with a Zeiss photomicroscope III. Thin sections, stained with uranyl acetate and lead citrate, were examined with a Zeiss EM 900 TEM. 2.5. Cross-breeding Single male and female muscle larvae of the PNG strain were crossed with single male and female muscle larvae of the eight reference strains (see below). The muscle larvae were collected after arti®cial digestion according to standard procedures (see above), and were identi®ed under an inverted microscope on the basis of the following morphological characters, proposed by Belosevic and Dick [14] and by Britov [15]. Male: (1) intestinal bulb generally close to the dorsal surface; in some larvae, close to the ventral surface; (2) intestine crossing the gonad from the dorsal to the ventral surface; in some larvae, crossing the gonad from the ventral to the dorsal surface and then recrossing to the ventral surface; and (3) length of rectum approximately 40± 50 mm. Female: (1) intestinal bulb generally close to the ventral surface; (2) intestine on the ventral surface; in some larvae, intestine crossing the gonad from the ventral to the dorsal surface and then recrossing to the ventral surface; (3) length of rectum approximately 20±30 mm; and (4) presence of a thickened subcuticular layer in the region of the vulval primordium (i.e. on the dorsal surface at about two-thirds of the way along the stichosome). Ten mice (CD1 female of 25 g) were inoculated per os with a single male and female larva belonging to dierent genotypes by a stomach tube connected to a 1 ml syringe, whereas ®ve mice were infected with a single male and female larva belonging to the same genotype and used as control. To ensure that no larvae remained in the syringe or in the tube, both were rinsed twice with water, which was then examined under the inverted microscope. All mice that were infected with larvae belonging to two dierent strains were immunosuppressed with 4 mg of cyclophosphamide at ÿ4, 0, 4, and 8 days p.i., since the RCI (number of larvae collected after arti®cial digestion/number of larvae
E. Pozio et al. / International Journal for Parasitology 29 (1999) 1825±1839
given per os) of the PNG strain was found to be low in CD1 mice. Mice were killed on day 35 p.i., and the entire skinned and eviscerated carcasses were digested individually. 2.6. Reference strains Ten reference strains were used for crossbreeding experiments and molecular characterisation: one strain of T. spiralis (code ISS3), one strain of Trichinella nativa (code ISS10), one strain of Trichinella britovi (code ISS2), three strains of T. pseudospiralis (codes ISS13, ISS141, ISS470), one strain of Trichinella nelsoni (code ISS29), one strain of Trichinella murrelli (code ISS35), one strain of Trichinella T6 (code ISS34), and one strain of Trichinella T8 (code ISS124) [16±18]. 2.7. Infectivity to birds A total of 500 muscle larvae collected after arti®cial digestion of mouse carcasses infected with the PNG strain and with the T. pseudospiralis reference strain (code ISS13) were given per os to 20 chickens (10 for each strain) of 2 weeks of age. As a control, the same number of larvae was given per os to 10 Swiss CD1 female mice (®ve for each strain). Chickens and mice were killed at 35 days p.i., skinned, and eviscerated, and the entire carcasses were subjected to arti®cial digestion to evaluate the RCI. 2.8. Puri®cation of genomic DNA DNA was obtained from frozen puri®ed larvae by overnight incubation at 558C in extraction buer (0.2 M Tris, 0.3 M NaCl, 0.025 M EDTA) containing 3% SDS and 200 mg/ml proteinase K. Puri®cation of DNA was done by phenol and phenol±chloroform±isoamyl alcohol extraction, followed by ethanol precipitation. DNA was resuspended in TE buer (10 mM Tris and 1 mM EDTA) and frozen at ÿ208C for use in PCR ampli®cations.
1829
2.9. Polymerase chain reaction protocol Polymerase chain reaction was performed using Taq DNA polymerase, 10 PCR buer, and dNTPs, from Takara (Japan), preparing 50 ml containing a ®nal concentration of 1.5 mM MgCl2 (included in the Takara 10 PCR buer), 200 mM dNTPs, 50 pmol of each primer and 0.5 unit of Taq DNA polymerase. For ampli®cation, 50 ng of DNA were used. Ampli®cations consisted of 35 cycles, as follows: denaturation at 948C for 20 s, annealing at 568C for 30 s, and elongation at 728C for 60 s. The primer set oTsr1 (5 0 -CGA AAA CAT ACG ACA ACT GC-3 0 ) and oTsr4 (5 0 -GTT CCA TGT GAA CAG CAG T-3 0 ) was used; this set ampli®es a region within the lsrDNA, known also as the expansion segment V (ESV) [7]. The primer set SB5 (forward and 5 0 -GGTATAAGGGAAAGCCGGAA-3 0 reverse 5 0 -CACTCACATTGTAT GCCAAG-3 0 ), designed by Wu et al. [19] from a sequenced fragment obtained by random ampli®cation of polymorphic DNA, which ampli®es a fragment of 680 bp exclusively in T. pseudospiralis, was also used to amplify DNA from PNG larvae. 2.10. Sequencing and analysis The ampli®cation products with the primer set oTsr1/oTsr4 were directly sequenced by Perkin Elmer ABI Prism 310 Genetic Analyser after labelling with BigDye Terminator Cycle Sequencing Ready Reaction kit (Perkin Elmer), following the manufacturer's instructions. The primers used for sequencing were the same as those used for PCR ampli®cation (oTsr1 and oTsr4). The products of three independent ampli®cation tubes were completely sequenced in both directions to con®rm the nucleotide sequences. The OMIGA 1.1 sequence-analysis software program was used for the analysis of the nucleotide sequence and for the alignment by Clustal W. Where possible, the alignment was determined by eye. The sequence of the PNG ESV was compared with the sequences present in the GenBank database of BLAST service of the National Centre for Biotechnology Information (NCBI) [20].
1830
E. Pozio et al. / International Journal for Parasitology 29 (1999) 1825±1839
3. Results The morphological features of adults and larvae of both sexes are reported in Table 1. The average TL of the isolated larvae was 966 mm for males (range 816±1066) and 1002 mm for females (range 888±1138). The TL of adult males was found to be equal in size or smaller, compared with the male larval stage. Muscles infected with non-encapsulated PNG larvae showed dramatic changes in the ®bre architecture. The infective larvae appeared to be almost always located in the intracellular space, and no in®ltration of in¯ammatory cells was observed around infected muscle ®bres (Fig. 3A). In some cases, myo®laments were totally disarranged, with the consequent dislocation of mitochondria (Fig. 3B). In fact, a large multilayered cluster of vacuolated mitochondria was observed to have formed around the larva (Fig. 4A). A double-layered envelope covered the larval cuticle, and a host-derived dense plasmalemmalike layer appeared to be ®rmly attached to the
host cell (Fig. 4B). In some cases, a nurse-celllike structure was observed where the sarcomeres appeared to be replaced by a sarcoplasmic reticulum ®lled with cisternae, lamellar bodies, and ribosomes (Fig. 5). An intense in¯ammatory reaction was observed around a larva that was apparently located outside the muscle cell (Fig. 6A±C). Muscles infected with larvae of the T. pseudospiralis reference strain revealed similar modi®cations in the muscle cell±parasite complex. The ultrastructural changes in the muscle ®bre architecture included the disaggregation of myo®laments, structural modi®cations, clustering and vacuolation of mitochondria and formation of the nurse cell-like structure (Fig. 7A, B). Single PNG male and female adult worms did not cross in either direction with females or males of the six species or two genotypes tested, whereas mice in control experiments became infected (data not shown). Papua New Guinea larvae were unable to infect chickens, unlike T. pseudospiralis larvae, which had an RCI of 15.
Table 1 Morphological features of adults and larvae of both sexes of Trichinella papuae n.sp.a Featuresb
Adult male2S.D. (range)
Adult female2S.D. (range)
Larva male2S.D. (range)
Larva female2S.D. (range)
TL MW NR OE ST V UT OV TS PAPL PAPW GP RE GPE
9552 31 (802±1056) 312 4 (24±38) 552 11 (24±72) 1242 17 (82±149) 3842 41 (307±480) Ð Ð Ð 3542 49 (254±456) 192 4 (10±32) 122 2 (8±16) Ð Ð Ð
17622199 (1320±2165) 362 5 (29±48) 572 11 (34±82) 126 213 (106±154) 374 249 (307±494) 396 232 (331±456) 10022171 (571±1392) 166 252 (101±360) Ð Ð Ð Ð Ð Ð
966265 (816±1066) 3324 (24±43) 6229 (38±77) 156241 (96±278) 456251 (360±569) Ð Ð Ð Ð Ð Ð 257240 (168±360) 5329 (42±80) 76219 (40±132)
1002259 (888±1138) 3425 (24±43) 6429 (48±82) 159223 (120±260) 442259 (269±566) Ð Ð Ð Ð Ð Ð 345228 (288±394) 2424 (18±34) 43210 (30±76)
a Measurements (mm) were made on 30 adult males, 30 adult females, 30 larval-stage males, and 30 larval-stage females, collected at the 8th passage in laboratory mice. b TL, total length; MW, width at the midbody; NR, distance of the nerve ring from the anterior extremity of the nematode; OE, length of the oesophagus; ST, length of stichosome; V, distance of the vulva from the anterior extremity of the nematode; UT, length of the uterus; OV, length of the ovary; TS, length of the testis; PAPL, length of the papillae; PAPW, width of the papillae; GP, length of the genital primordium; RE, length of the rectum; and GPE, distance of the posterior margin of the genital primordium from the posterior extremity of the nematode.
E. Pozio et al. / International Journal for Parasitology 29 (1999) 1825±1839
1831
Fig. 3. (A) Oblique longitudinal section of an intracellular larva of Trichinella papuae in a mouse diaphragm 3 months p.i. Note the absence of the collagen capsule. There is no in®ltration of in¯ammatory cells around the infected muscle ®bres. Scale bar = 50 mm. (B) Transmission electron micrograph of an intracellular larva of T. papuae. Myo®laments are almost completely disaggregated and mitochondria appear to be scattered throughout the sarcolemma. M, mitochondria; P, parasite. Scale bar = 1.1 mm.
Nevertheless, larvae of both the PNG and T. pseudospiralis strains from the same batches were able to infect laboratory mice. The PCR ampli®cation of the ESV of the PNG strain using the primer set oTsr1/oTsr4 produced a single fragment of 240 bp. This fragment diers from those of T. pseudospiralis strains, in which polymorphism was observed, with three fragments ranging from 290 to 310 bp. It also diers from fragments of the other species and genotypes (Fig. 8). Unlike T. pseudospiralis isolates, no evidence of polymorphism was found in the ESV of the PNG strain. The primer set
SB5, speci®c for T. pseudospiralis, did not amplify any fragment from the DNA of the PNG strain (Fig. 9). The alignment of ESV sequences of the PNG strain, T. pseudospiralis strain from North America (code ISS470), and T. pseudospiralis strain from Russia (Garkavi's strain, code ISS13) showed that the three strains had some microsatellites, or portions of microsatellites, in common (Fig. 10). In the PNG strain, the TGG microsatellite and its 5 0 ¯anking region were found to be completely conserved; the GT1 and TTG microsatellites and their 5 0 ¯anking regions were par-
1832
E. Pozio et al. / International Journal for Parasitology 29 (1999) 1825±1839
Fig. 4. (A) Transmission electron micrograph of Trichinella papuae in a mouse diaphragm 3 months p.i. Sarcomeres are totally lacking. A thick cluster of mitochondria surrounds the larva. In most mitochondria, the matrix is beginning to become vacuolated, yet the cristae appear normal. Arrows, ribosomes; asterisk, Golgi complex; M, mitochondria; P, parasite. Scale bar = 1.1 mm. (B) Detail of (A) showing the muscle±parasite interface. A double-layered structure covers the cuticle (arrow) and a host-derived plasmalemma-like layer (arrowhead) appears to be ®rmly attached to the underlying mitochondrial cluster. The epi-, exo-, and mesocuticular regions are evident. E, exocuticle; Me, mesocuticle; H, hypodermis; M, muscle. Scale bar = 0.4 mm.
tially conserved; whereas the TGC and CA microsatellites were absent. A BLAST search in the GenBank database revealed no signi®cant homology with the PNG sequence. 4. Discussion Trichinella nematodes were ®rst discovered in PNG in August 1988, when non-encapsulated Trichinella larvae were observed in muscle tissue sections of ®ve domestic pigs from a small breed-
ing facility in the village of Balamuk. At the time, Trichinella parasites were isolated in laboratory rats and pigs, but these isolates were not maintained in laboratory animals. Adults of Trichinella collected from the gut of infected rats in 1988 are preserved under formalin at the Natural History Museum, London (ref. Nos. 1989, 1200 and 1220). The village of Balamuk and the Bula Plain are located in a remote region of western PNG; they have few facilities, a very small population of hunters and subsistence gardeners, and virtually no movement of animals
E. Pozio et al. / International Journal for Parasitology 29 (1999) 1825±1839
1833
Fig. 5. Transmission electron micrograph of an intracellular larva of Trichinella papuae in a mouse diaphragm 3 months p.i. Nurse cell-like structure: contractile elements are replaced by sarcoplasmic reticulum. C, cisternae; L, lamellar bodies; P, parasite; arrows, ribosome; V, vesicles. Scale bar = 1.2 mm.
outside of the area, suggesting the autochthonous origin of the parasite. We cannot rule out that the PNG strain of Trichinella originated from the adjacent area of Irian Jaya (Indonesia), given that there is no physical barrier between these two countries. Muscle larvae of both the PNG strain and T. pseudospiralis can be distinguished from those of the other species and genotypes of the genus Trichinella by the absence of a capsule around larvae in host muscles. The two strains can be distinguished from each other based on the TL of larvae; in fact, the TL of PNG muscle larvae falls within the TL of encapsulated species and genotypes and is about one-third greater than that of T. pseudospiralis. The presence of these morphological hallmarks is of great importance for taxonomic and epidemiological studies. The present investigation reveals the existence of ultrastructural similarities between the PNG strain and the reference strain for
T. pseudospiralis. The muscle-stage larva of the PNG strain lacks a collagen capsule, and the infected muscle ®bres undergo dramatic alterations in their micro-architecture, including: (1) loss of contractile elements; (2) clustering of mitochondria around the muscle larva; (3) enlargement of the sarcoplasmic reticulum; and (4) formation of a nurse-cell-like structure. The surface layers of the cuticle of PNG larvae appear to be ultrastructurally similar to the cuticular envelopes already described in T. pseudospiralis [21, 22]. The presence of a double membrane of host origin around PNG larvae may provide some protection for the worm from the host response; moreover, the absence of a collagen capsule does not prevent the larvae from migrating through the muscle ®bres. This situation may be harmful to the host because it does not limit the invasion of new muscle ®bres. The cellular response of the host observed around the larvae located outside the muscle cell may per-
1834
E. Pozio et al. / International Journal for Parasitology 29 (1999) 1825±1839
Fig. 6. In¯ammatory reaction around an extracellular larva of Trichinella papuae in a mouse diaphragm 3 months p.i. (A) Section of muscle infected with T. papuae: around the parasite (P) there is a cluster of in¯ammatory cells (Ic). Scale bar = 0.5 mm. (B) Transmission electron micrograph of the parasite (P) apparently located outside the muscle cell. Scale bar = 5 mm. (C) Transmission electron micrograph of an in¯ammatory zone. Arrows, collagen ®brils; G, eosinophil granulocyte. Scale bar = 5 mm.
E. Pozio et al. / International Journal for Parasitology 29 (1999) 1825±1839
1835
Fig. 7. Transmission electron micrographs of a larva of Trichinella pseudospiralis in a mouse diaphragm 3 months p.i. (A) The micrograph shows the disaggregation of myo®laments and concentration of mitochondria close to the larval surface. The inner matrix of mitochondria is beginning to become vacuolated. Arrows, collagen ®brils; F, ®broblast; M, mitochondria; P, parasite; arrowhead, sarcolemma. Scale bar = 1.2 mm. (B) The micrograph shows a nurse cell-like structure characterised by an elaborate system of smooth endoplasmic reticulum cisternae. C, cisternae; asterisk, larval envelope remnants; P, parasite; arrows, ribosomes; S, smooth endoplasmic reticulum. Scale bar = 5 mm.
haps reduce tissue damage. These observations are consistent with the severity of the clinical picture observed in persons infected with T. pseudospiralis in Thailand, some of whom died [6]. Adults of the PNG strain do not cross with adults of the other species or genotypes in Trichinella in either sense, suggesting a reproductive isolation not only with the encapsulated genotypes, but also with T. pseudospiralis, the other non-encapsulated species. Although the study of cross-breeding is of great importance in evaluating the relationship between two genotypes, data from the literature show con¯icting results as to whether
cross-breeding of Trichinella genotypes obtained in laboratory conditions can be translated to reallife settings [2, 23±25]; however, since our results were negative, this problem did not arise. Infectivity to birds, which is considered to be one of the most important biological features of T. pseudospiralis, is absent in the PNG strain, suggesting that the two strains have a dierent evolution. In fact, the capability of T. pseudospiralis to infect birds has been considered an important factor in the dispersion of this parasite in several continents, and at the same time an ancestral feature of this species [3]. When a non-
1836
E. Pozio et al. / International Journal for Parasitology 29 (1999) 1825±1839
Fig. 8. Polymerase chain reaction ampli®cation of the expansion segment V of DNA from Trichinella strains by the primer set oTsr1 and oTsr4. Lane 1, Trichinella pseudospiralis from Russia; lane 2, T. pseudospiralis from Alabama (USA); lane 3, T. pseudospiralis from Tasmania; lane 4, Papua New Guinea strain; lane 5, T. britovi; lane 6, T. spiralis; lane 7, T. nelsoni; lane L, 100 bp ladder (Amersham-Pharmacia Biotech).
encapsulated Trichinella was ®rst discovered in PNG, we believed that its presence on the island of New Guinea may have been related to birds. However, the results of experimental infection in chickens suggest that another means of introduction may exist. Considering that T. papuae has been detected only in wild and domestic swine and that this host species was introduced in New Guinea by human migrants from Indonesia and South East Asia approximately 6000±10 000 years ago [26], we speculate that this parasite evolved outside of New Guinea and was introduced with imported swine. The detection of Trichinella in wild animals from nearby regions of New Guinea will be of great importance for explaining the origin and the evolution of this parasite. Zarlenga and Dame [27] demonstrated the presence of a gap in the ribosomal RNA with respect to the ribosomal DNA in T. spiralis. This gap was due to excision of ESV, which seems to be of no use for the functionality of the ribo-
Fig. 9. Polymerase chain rection analysis of dierent species of Trichinella ampli®ed by the primer set SB5F and SB5R. Lane 1, Trichinella pseudospiralis from Russia; lane 2, T. pseudospiralis from Alabama (USA); lane 3, T. pseudospiralis from Tasmania; lane 4, T. papuae; lane 5, T. britovi; lane 6, T. spiralis; lane L, 100 bp ladder (Amersham-Pharmacia Biotech).
some. The ESV is then free to ®x mutations in Trichinella, as demonstrated by the analysis of ribosomal DNA sequences of dierent species in the genus, which show that the dierences are exclusively located inside the ESV [7, 27]. Molecular analysis of ESV shows that the PNG strain greatly diers from the non-encapsulated isolates of T. pseudospiralis, as well as from the other species and genotypes in Trichinella (data not shown). Based on morphological, biological, and molecular features, it can be concluded that parasites of the PNG strain belong to a new species which we propose naming Trichinella papuae n.sp. This is the tenth genotype (Trichinella T10) to be identi®ed in the genus Trichinella. In fact, in addition to the eight genotypes described by La Rosa et al. [11], the Japanese strains from sylvatic animals have recently been recognised as belonging to a new genotype (Trichinella T9) [28]. Trichinella pseudospiralis isolates have been detected several times in Asia (India, Kamchatka, Kazakhstan, and Thailand) [5, 6, 29] and in Europe (Caucasus and Tula regions of Russia, Italy and France) [1, 5, 30] (G La Rosa and E Pozio, unpublished data), once in North America (Alabama) [4], and in many animals from the Australian region (Tasmania) [8, 9]. In
Fig. 10. Multi-alignment of homologous expansion segment V (ESV) fragments from Trichinella pseudospiralis from North America (TpUSA, code ISS470), T. pseudospiralis from Russia (TpRussia, code ISS13), and Trichinella papuae from Papua New Guinea (PNG, code ISS572). U = upper band; L = lower band (see [7]); asterisk, the conservation of the base among all the examined strains; dot, the presence of a gap; horizontal bar, the PCR primer; boxed area, a microsatellite; arrows, boundaries of the ESV.
1838
E. Pozio et al. / International Journal for Parasitology 29 (1999) 1825±1839
South-East Asia and Australian regions, apart from the cosmopolitan species T. spiralis, which has been passively introduced by human activity, only T. pseudospiralis and T. papuae have been detected in sylvatic animals, suggesting that the evolutionary line of non-encapsulated larvae in the genus Trichinella could originate in this area of the world. These data suggest the presence of at least two well-dierentiated non-encapsulated species whose geographical distribution partially overlaps that of encapsulated species and genotypes. The capability of larvae of encapsulated species to induce the modi®cation of the muscle cell into a nurse cell is a very complex molecular mechanism, in part still unknown [31]. This mechanism is absent in T. pseudospiralis and in T. papuae, strongly suggesting the presence of two evolutionary lines in the Trichinella genus, indicating that their classi®cation should be reevaluated. Further studies are needed to evaluate the natural life-cycle and epidemiology of T. papuae as well as the distribution area, the natural reservoirs, and its role as a human pathogen. 5. Taxonomic summary 5.1. Trichinella papuae n.sp. Biological features: RCI in Swiss mice 5.3 at the 9th passage. No resistance to freezing. Type host: Sus scrofa, considered to be a hybrid between Sus scrofa vittatus and Sus celebensis. Other hosts: (1) natural, domestic pig (Sus scrofa); (2) laboratory, Swiss mouse. Site: small intestine (adults) and skeletal muscle (larvae). Morphological features: non-encapsulated male and female larvae in muscle tissues have an average total length of 966 mm (range 816±1066) and 1002 mm (range 888±1138), respectively. Males appear to be equal in size or smaller at the adult stage, compared with the larval stage. Molecular features: expansion segment V of the ribosomal DNA of 240 bp. Type locality: Bula Plain, Papua New Guinea, latitude 88 56 0 S, longitude 1418 16 0 E.
Distribution: unknown. Specimens deposited: syntypes, adult males and females, International Trichinella Reference Centre at the Istituto Superiore di SanitaÁ, Italy (ref. No. ISS572). Adults and ML from this strain are also maintained in CD1 female mice. Etymology: the species is named for the country of origin.
Acknowledgements We wish to thank F. Mancini Barbieri and M. Amati for their technical support. The collection of biological material in Papua New Guinea was made possible by the support received from the Northern Australia Quarantine Strategy (NAQS).
References [1] Garkavi BL. Species of Trichinella isolates from wild animals. Veterinariya 1972;10:90±1. [2] Dick TA. Species, and infraspeci®c variation. In: Campbell WC, editor. Trichinella and trichinellosis. New York: Plenum Press, 1983;31±73. [3] Pozio E, Shaikenov B, La Rosa G, Obendorf DL. Allozymic and biological characters of Trichinella pseudospiralis isolates from free-ranging animals. J Parasitol 1992;78:1087±90. [4] Lindsay DS, Zarlenga DS, Gamble HR, Al-Yaman F, Smith PC, Blagburn BL. Isolation and characterization of Trichinella pseudospiralis Garkavi, 1972 from a black vulture (Coragypus atratus). J Parasitol 1995;81:920±3. [5] Britov VA. Trichinellosis in Kamchatka. Wiad Parazyt 1997;43:287±8. [6] Jongwutiwes S, Chantachum N, Kraivichian P et al. First outbreak of human trichinellosis caused by Trichinella pseudospiralis. Clin Infect Dis 1998;26:111±5. [7] Zarlenga DS, Aschenbrenner RA, Lichtenfels JR. Variations in microsatellite sequences provide evidence for population dierences and multiple ribosomal gene repeats within Trichinella pseudospiralis. J Parasitol 1996;82:534±8. [8] Obendorf DL, Clarke KP. Trichinella pseudospiralis infections in free-living Tasmanian birds. J Helminthol Soc Wash 1992;59:144±7. [9] Obendorf DL, Handlinger JH, Mason RW et al. Trichinella pseudospiralis infection in Tasmanian wildlife. Aust Vet J 1990;67:108±10. [10] Andrews JRH, Bandi C, Pozio E, Gomez Morales MG, Ainsworth R, Abernethy D. Identi®cation of Trichinella
E. Pozio et al. / International Journal for Parasitology 29 (1999) 1825±1839
[11] [12] [13]
[14] [15] [16] [17] [18] [19] [20]
[21]
pseudospiralis from a human case using random ampli®ed polymorphic DNA. Am J Trop Med Hyg 1995;53:185±8. La Rosa G, Pozio E, Rossi P, Murrell KD. Allozyme analysis of Trichinella isolates from various host species and geographic regions. J Parasitol 1992;78:641±6. Buncic S. A case of a pig infested with Trichinella spiralis. Surveillance 1997;24:8. Groves C. Ancestors for the pigs: taxonomy and phylogeny of the genus Sus. Technical Bulletin No. 3, Department of Prehistory, Research School of Paci®c Studies, Australian National University, 1981;1±96. Belosevic M, Dick TA. Trichinella spiralis: comparison with an arctic isolate. Exp Parasitol 1980;49:266±76. Britov VA. Etiological agents of trichinellosis. Moscow: Nauka, 1982 [in Russian]. Pozio E, La Rosa G, Murrell KD, Lichtenfels JR. Taxonomic revision of the genus Trichinella. J Parasitol 1992;78:654±9. Pozio E, La Rosa G. Trichinella murrelli n.sp: etiological agent of sylvatic trichinellosis in temperate areas of North America. J Parasitol, in press. Pozio E, La Rosa G, Rossi P. Trichinella Reference Centre. Parasitol Today 1989;5:169±70. Wu Z, Nagano I, Fukumoto S, Saito S et al. Polymerase chain reaction primers to identify Trichinella spiralis or T. pseudospiralis. Parasitol Int 1997;46:149±54. Altschul SF, Madden TL, Schaer AA et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 1997;25:3389±402. Lee DL, Wright KA, Shivers RR. A freeze±fracture study of the surface of the infective-stage larva of the Nematode Trichinella. Tissue Cell 1984;16:819±28.
1839
[22] Ubelaker JE, Stewart GL, Martin JH. Modi®cation of the ultrastructure of the muscle larva of Trichinella pseudospiralis following exposure to acidi®ed pepsin solution. J Parasitol 1993;79:133±7. [23] Bessonov AS, Penkova RA, Uspensky AV. On the independence of Trichinella species. Wiad Parazyt 1975;21:561±75. [24] Boev SN, Britov VA, Orlov IV. Species composition of Trichinellae. Wiad Parazyt 1979;25:495±503. [25] Ooi HK, Oku Y, Kamiya M, Ohbayashi M. Interbreeding and fecundity of a single pair of two strains of Trichinella spiralis in mice. Jpn J Vet Res 1984;32:177±82. [26] Flannery TF. Mammals of New Guinea. The Australian Museum, Carina, Qld: R Brown and Associates, 1990. [27] Zarlenga DS, Dame JB. The identi®cation and characterization of a break within the large subunit ribosomal RNA of Trichinella spiralis: comparison of gap sequences within the genus. Mol Biochem Parasitol 1992;51:281±90. [28] Nagano I, Wu Z, Matsuo A, Pozio E, Takahashi Y. Identi®cation of Trichinella isolates by polymerase chain reactionÐrestriction fragment length polymorphism of the mitochondrial cytochrome c oxidase. Int J Parasitol 1999;29:1113±20. [29] Shaikenov B, Boev SN. Distribution of Trichinella species in the Old World. Wiad Parazyt 1983;29:595±608. [30] Pozio E, Goredo M, Fico R, La Rosa G. Trichinella pseudospiralis in sedentary night-birds of prey from Central Italy. J Parasitol 1999;85:759±61. [31] Despommier DD. How does Trichinella spiralis make itself at home?. Parasitol Today 1998;14:318±23.