Background

Hookworms (Nematoda: Ancylostomatidae) are one of the most common soil-transmitted helminths, causing serious iron-deficiency anemia and protein malnutrition in humans and domestic and wild mammals [1,2,3]. Both major genera Ancylostoma (Dubini, 1843) and Necator Stiles, 1903, relegated to two distinct subfamilies, are responsible for morbidity and socioeconomic burdens [4]. Unlike species in the genus Necator, most Ancylostoma hookworms are considered to be of greater medical and veterinary importance because of distribution, prevalence, and multiple zoonotic species [2]. Currently there are fourteen valid species identified in the genus Ancylostoma that are often considered in the context of the range of hosts that are typically infected. For example, the ‘anthrophilic’ form is limited to Ancylostoma duodenale (Dubini, 1843) which principally infects humans. ‘Anthropozoonotic’ forms, capable of circulating among free-ranging wild hosts, some domestic hosts and humans include Ancylostoma caninum (Ercolani, 1859), Ancylostoma braziliense Gomes de Faria, 1910 and Ancylostoma ceylanicum Looss, 1911. Other species, including most of the recognized diversity in the genus are considered to be primarily of veterinary importance, including Ancylostoma tubaeforme (Zeder, 1800), Ancylostoma malayanum (Alessandrini, 1905), Ancylostoma pluridentatum (Alessandrini, 1905), Ancylostoma paraduodenale Biocca, 1951, Ancylostoma kusimaense Nagayosi, 1955, Ancylostoma buckleyi Le Roux & Biocca, 1957, Ancylostoma taxideae Kalkan & Hansen, 1966, Ancylostoma genettae Macchioni, 1995, Ancylostoma protelesis Macchioni, 1995, and Ancylostoma somaliense Macchioni, 1995 [5, 6]. It is noteworthy that nearly all of these species can also be found in wildlife, such as A. duodenale in Crocuta crocuta (Erxleben); A. caninum and A. braziliense in Acinonyx jubatus (Schreber) and Canis mesomelas Schreber; A. ceylanicum in Canis lupus dingo Meyer; A. paraduodenale in Leptailurus serval (Schreber); A. malayanum in Ursus thibetanus G. Cuvier; A. pluridentatum in Puma concolor coryi (Bangs); A. kusimaense in Nyctereutes procyonoides viverrinus Temminck; A. taxideae in Taxidea taxus taxus (Schreber); A. genettae in Genetta genetta (Linnaeus); A. protelesis in Proteles cristata (Sparrman); and A. somaliense in C. mesomelas [5,6,7,8,9,10,11,12]. Although a diverse assemblage of carnivorans is recognized as hosts for Ancylostoma, only one species had been documented or described previously among the Ursidae [7]; species of the distantly related Uncinaria Frölich, 1789, are considered typical in ursine hosts [13].

Recent molecular-based genetic and epidemiological investigations have shown that among certain wild or domestic animal-derived species of Ancylostoma, A. ceylanicum is becoming the second most common hookworm found to infect and complete its life-cycle in humans [12, 14,15,16,17,18]. Similar transmission and cross-infection cases have been reported for other congeneric species, notably A. caninum [12, 19, 20] and A. braziliense [12]. Such situations highlight the public health significance of hookworm infection and the necessity to assess their prevalence and distribution, and to identify their wildlife hosts. This has become especially important for wildlife hosts that may have recently adapted to the human environment due to rapid urbanization [14, 21] leading to increased interactions with people in conservation centers and zoological gardens constructed for endangered and valuable animals [22]. Regrettably, little attention has been broadly paid to the species of Ancylostoma because of a limited understanding of their diversity, abundance and distribution and the difficulty in etiological and epidemiological sampling in the wild [12, 14].

The giant panda, Ailuropoda melanoleuca (David), one of the most endangered and rare species of China, is regarded as one of the preeminent species for wildlife conservation in the world. Higher taxonomic status for these enigmatic carnivorans had been unresolved, until relatively recent decisions that unequivocally placed giant pandas among the Ursidae (e.g. [23,24,25,26]). Wild giant pandas currently inhabit six small mountain ranges of China i.e. Qinling, Minshan, Qionglai, Daxiangling, Xiaoxiangling and Liangshan (Fig. 1), with an estimated population size of ~1,864 [27,28,29,30]. Since the 1950s, numerous natural reserves, conservation centers, research bases and zoological gardens were specifically established by the Chinese government to protect this threatened species [31]. Some of these wild giant pandas have become closely associated with humans as they are housed for artificial breeding and conservation and biological investigations. Also, some pandas have been displayed publically as the ‘messenger of peace and friendship’ around the world [32]. Although ecological, genetic and etiological studies have shown that the panda faces the threat of extinction due to habitat loss, poor reproduction and low resistance to infectious diseases [33, 34], recent surveys strongly indicate that parasitic infections represent the leading health threat to giant pandas of China [35,36,37,38,39,40].

Fig. 1
figure 1

Sampling site in China (red circle) for Ancylostoma ailuropodae n. sp. in the giant panda. The distribution of the giant panda populations in Shaanxi, Gansu and Sichuan provinces of China is indicated in black with the names of mountain ranges

Full size image

Hookworm parasites have been frequently observed in the intestines of wild dead giant pandas since 1995 [28] and the first record, attributed to a species of Ancylostoma, was reported by Zhang et al. in 2005 [41]. However, detailed morphological descriptions, determination of taxonomic status and indicators of pathogenicity of the Ancylostoma sp. derived from giant panda are lacking. The recent collection of parasites from a wild giant panda that died in the Fengtongzai Natural Reserve in Sichuan Province of China resulted in the recovery of fresh Ancylostoma specimens and provided an opportunity to fill some of these gaps in our knowledge. We have used DNA sequence and morphological analysis, applying clear species criteria established in a phylogenetic context [42], to recognize and describe a previously unknown hookworm species from the giant panda. A putative sister-species relationship with the ‘anthropozoonotic’ A. ceylanicum suggests a possible zoonotic risk for transmission and infection to humans.

Methods

Parasite collection and microscopic examination

In November 2013, a wild female giant panda was found dead in the Fengtongzai Natural Nature Reserve, Sichuan Provence of China (Fig. 1). After a routine necropsy, seventeen hookworm specimens (seven males and ten females) were collected from the small intestine under the Scientific Procedures Premises License for the College of Veterinary Medicine, Sichuan Agricultural University (Sichuan, China). In addition, parasite eggs were isolated from the intestinal content by the centrifuge-flotation method using saturated MgSO4 [43]. After washing in physiological saline, the hookworm specimens were either directly fixed in Berland’s fluid (95% glacial acetic acid and 5% formaldehyde) for morphological analysis or stored in 70% ethanol for subsequent molecular profiling. For morphology, the hookworms were identified to the genus level on the basis of the existing taxonomic keys and descriptions of Ancylostoma spp. (e.g. [44]). In brief, the worms (n = 15; 6 males and 9 females) were prepared as temporary whole mounts in glycerin after clearing in lactophenol and examined under both dissecting and light microscopy at magnifications of 10–40× and 40–200×, respectively; male and female specimens were characterized morphologically including photo-micrographic imaging and morphometrics. Measurements are given in micrometres (μm) unless specified otherwise and presented with the range followed by the mean within parentheses. In addition, some key characteristics of the adults were drawn with the aid of serial photographs for morphological comparison and differentiation from other related species. These specimens including the type-series and vouchers for molecular analyses have been deposited in the Department of Parasitology, Sichuan Agricultural University (accession numbers code GYY-XY).

Molecular profiles and phylogeny

For molecular analysis, two adult specimens of Ancylostoma sp. (one female and one male; sample codes FTZ1 and FTZ2, respectively) preserved in 70% ethanol were air-dried and their mid-body regions (~1 cm) were excised individually for extraction of genomic DNA using the Universal Genomic DNA Extraction Kit (TaKaRa, Dalian, China) according to the manufacturer’s protocol. The cephalic and caudal extremities of each specimen were retained as archived vouchers. The DNA extract was used as template for PCR amplifications at the nuclear internal transcribed spacer ITS1-5.8S-ITS2 region (734 bp) and mitochondrial cytochrome c oxidase subunit 1 (cox1) locus (393 bp) using primer pairs designed based on the alignments of the relatively conserved regions of the congeneric species A. ceylanicum, A. caninum, and A. duodenale in GenBank. Two PCR primer sets were as follows: ITS1-5.8S-ITS2, forward: 5′-GTC GAA GCC TTA TGG TTC CT-3′ and reverse: 5′-TAA CAG AAA CAC CGT TGT CAT ACT A-3′; cox1, forward: 5′-ATT TTA ATT TTG CCT GCT TTT G-3′ and reverse: 5′-ACT AAC AAC ATA ATA GGT ATC ATG TAA-3′. The PCR reactions contained ~20 ng of genomic DNA were performed in 50-μl reaction volumes containing 25 μl 2× Phusion High-Fidelity PCR Master Mix (Finnzymes OY, Espoo, Finland), 3 μl gDNA, 3 μL of each primer and 16 μl of ddH2O. PCR cycling conditions carried out in a Mastercycler Gradient 5331 thermocycler (Eppendorf, Germany) were an initial denaturation at 95 °C for 5 min; then for ITS1-5.8S-ITS2, 35 cycles of 95 °C for 30 s, 39.8 °C for 30 s, and 72 °C for 45 s; but for cox1, 35 cycles at 95 °C for 30 s, 44.1 °C for 30 s, and 72 °C for 30 s; followed by a final step at 72 °C for 10 min. For each amplification, samples without parasite gDNA and host DNA as negative controls were also included. All PCR products were examined on agarose (1%) gels to verify that they represented the target bands. The corrected gel-isolated amplicons were column-purified and sequenced in both directions using terminator-based cycle sequencing with BigDye chemistry (Applied Biosystems, Foster City, CA, USA) on an ABI 3730 DNA sequencer (Applied Biosystems) in TaKaRa Biotechnology Co. Ltd. (Dalian, China). To ensure maximum accuracy, each amplicon was sequenced three times independently. The consensus sequences were utilized for the following bioinformatic analyses and added to GenBank under the accession numbers KP842923 (FTZ1) and KP842924 (FTZ2) for ITS1-5.8S-ITS2 and KP842921 (FTZ1) and KP842922 (FTZ2) for cox1.

Sequences of ITS1-5.8S-ITS2 and cox1 of Ancylostoma sp. in the present study were separately aligned with reference sequences from closely related species (Table 1), including the congeneric species A. ceylanicum, A. caninum, A. duodenale, A. braziliense and A. tubaeforme as well as other hookworm species Necator americanus (Stiles, 1902), Uncinaria hamiltoni Baylis, 1933 [45], U. lucasi Stiles & Hassall, 1901, U. stenocephala (Railliet, 1884), U. sanguinis Marcus, Higgins, Slapeta & Gray, 2014 [46], Uncinaria sp., and Bunostomum phlebotomum (Railliet, 1900), using the Clustal X 1.83 program [47]. During the procedure, the nucleotide alignment of cox1 was further adjusted by a codon-guided protein alignment. Given the presence of the ambiguous regions within these alignments, an online version of GBlocks (http://molevol.cmima.csic.es/castresana/Gblocks_server.html) was also introduced here. After refining the alignments using Gblocks, the sequence datasets were used for phylogenetic analyses using both maximum parsimony (MP) (PAUP* 4.10b [48]) and Bayesian inference (BI) methods (MrBayes 3.2 [49]). In the MP analysis, heuristic searches were executed by branch-swapping utilizing tree-bisection-reconnection (TBR) algorithm and 1,000 random-addition sequence replicates with 10 trees held at each step, and finally the optimal topology with bootstrapping frequencies (BF) was obtained using Kishino-Hasegawa, as described previously [50]. For the BI analysis, the nucleotide substitution model GTR + I + G was determined using the Bayesian Information Criteria (BIC) test in jModeltest v. 2.1.6 [51], and the trees were constructed employing the Markov chain Monte Carlo (MCMC) method (chains = 4) over 100,000 (cox1) or 1,000,000 (ITS1-5.8S-ITS2) generations with every 100th (cox1) or 1000th (ITS1-5.8S-ITS2) tree being saved; when the average standard deviation of the split frequencies reduced to less than 0.01, 25% of the first saved trees were discarded as “burn-in” and the consensus (50% majority rule) trees were inferred from all remaining trees and further plotted in TreeviewX (http://taxonomy.zoology.gla.ac.uk/rod/treeview.html), with nodal supports expressed as posterior probabilities (PP). The livestock hookworm B. phlebotomum was used as outgroup reference and included in each phylogenetic analysis. Paralleled to the phylogenies, among the genus Ancylostoma the new hookworm species coupled with A. ceylanicum, A. caninum, A. duodenale and A. tubaeforme was also subjected to detection of synonymous and non-synonymous mutations in the mitochondrial cox1 gene using their corresponding protein sequences, followed by determination of genetic distances between them using a distance matrix based on the maximum composite likelihood model in MEGA [52].

Table 1 Information of Ancylostoma species used for molecular identification in the present study
Full size table

Results

Family Ancylostomatidae Looss, 1905

Genus Ancylostoma (Dubini, 1843)

Ancylostoma ailuropodae Yang, Hoberg & Xie n. sp.

Type-host: Giant panda Ailuropoda melanoleuca (David) (Mammalia: Carnivora: Ursidae).

Type-locality: Fengtongzai Natural Reserve (30°42′12″N, 102°56′14″E), Baoxing, Sichuan Province, China.

Type-material: Holotype, adult male (GYY-XY 1301); allotype, adult female (GYY-XY 1308); paratypes, three adult males (GYY-XY 1302-4) and three females (GYY-XY 1309-11). All materials, together with nine vouchers (three males, GYY-XY1305-7; six females, GYY-XY13012-17) containing one male and one female represented by cephalic and caudal extremities, with the mid-body sub-sampled for DNA sequence analysis, are deposited at the Department of Parasitology in Sichuan Agricultural University, Sichuan, China. Collectors: GY Yang, TF Zhang and Y Xie.

Site in host: Small intestine (most in the duodenum).

Representative DNA sequences: Representative nuclear ribosomal and mitochondrial DNA sequences were deposited in the GenBank database under the accession numbers KP842923–KP842924 (ITS1-5.8S-ITS2) and KP842921–KP842922 (cox1).

ZooBank registration: To comply with the regulations set out in article 8.5 of the amended 2012 version of the International Code of Zoological Nomenclature (ICZN) [53], details of the new species have been submitted to ZooBank. The Life Science Identifier (LSID) of the article is urn:lsid:zoobank.org:pub:A2492E99-AA70-4A58-AB70-7FED78E726A3. The LSID for the new name Ancylostoma ailuropodae n. sp. is urn:lsid:zoobank.org:act:2C6B6C1E-5F70-49B7-A303-B4D5AE9C7847.

Etymology: The new species is named for the type-host.

Description

General. Slender, relatively small nematodes of white coloration in life (Fig. 2a). Body cylindrical, tapering toward cephalic and caudal extremities with fine transversely striated cuticle; head oriented dorsally in males and females. Buccal capsule widening posteriorly to prominent oral aperture, possessing two pairs of ventrolateral teeth and two pairs of triangular dorsolateral teeth (Fig. 2b-f). Ventrolateral teeth vary in size and shape, with small, sub-aduncate inner and large triangular outer teeth extending dorsally. Dorsal gland well developed, associated with rod-like oesophagus, slightly swollen posteriorly, terminating in a lobed valve at junction with intestine (Fig. 4a, b). Nerve-ring at midlevel of oesophagus. Cervical papillae well developed, conical, situated posterior to level of nerve-ring. Excretory pore opens at level between cervical papillae and nerve-ring (Fig. 5a1, 2).

Fig. 2
figure 2

Photomicrographs of adults of Ancylostoma ailuropodae n. sp. a Total view of males (top) and females (down); b-f Cephalic extremity: lateral view of mouth (b and c), showing dorsolateral and ventrolateral teeth; dorsoventral view of mouth (d-f), showing dorsolateral (d) and ventrolateral (e and f) teeth with their positions, shapes and sizes. The arrangements of dorsolateral (2 pairs; b-d) and ventrolateral (2 pairs; b, c, e and f) teeth are indicated by red arrows

Full size image

Male. [Based on the holotype and three males.] Body length 8.60–12.00 (10.30) mm, maximum width at mid-body 500–520 (510). Buccal capsule 180–220 (200) long, 120–160 (140) wide in dorsoventral view; oesophagus 960–1,500 (1,230) long, 150–190 (170) wide; oesophageal length 12% of total body. Cervical papillae 600–750 (680), excretory pore 500–580 (530), nerve-ring 390–520 (425) posterior to cephalic extremity. Copulatory bursa well developed, broader than long; dorsal lobe small with lateral lobes projecting in direction of lateral trunks (Figs. 3a-f, 5a5). Dorsal ray thick, 280–390 (350) in length, 40–60 (52) in maximum width; bifurcating at 270–295 (280) from anterior into 2 branches; each branch further dividing into 2 sub-branches; externodorsal rays arcuate, arising from dorsal ray at same level (Figs. 3e-f, 5a5). Lateral rays slender, tapering, and arcuate with a common stem. Anterolateral ray bending anteriad, with medio- and posterolateral rays projecting in parallel, extending to edge of bursa (Figs. 3a-b, 5a5). Antero- and posteroventral rays merge at base and then divide, continuing parallel deep into cleft (Figs. 3c-d, 5a5). Spicules tawny colored, paired, equal, filiform, 2,000–2,900 (2,450) long (Figs. 4e-f, 5a5). Gubernaculum fusiform, 80–120 (90) long, 12–20 (16) wide (Figs. 4c, 5a5). Cloacal papillae (n = 7) (Figs. 4d, 5a5): 1 pair disposed dorsally, 1 pair laterally, 3 single papillae ventrally.

Fig. 3
figure 3

Photomicrographs of Ancylostoma ailuropodae n. sp. male, caudal extremity. a, b Lateral view of bursa showing position of lateral rays and genital cone. c, d Ventral view of bursa showing configuration of antero- and postero-ventral rays. e, f Dorsal view of bursa showing relationships of the dorsal and externodorsal rays; note configuration of the bifurcations of the dorsal ray. Arrows in a, c and e denote the rays which are magnified in panels b, d and f, respectively

Full size image
Fig. 4
figure 4

Photomicrographs of adults of Ancylostoma ailuropodae n. sp. a Dorsoventral view of anterior region of female, showing buccal capsule and entire oesophagus. b Lobed valves between oesophagus and intestine. c Ventrolateral view of male tail, showing gubernaculum. d Ventral view of male tail, showing cloacal papillae. e, f Ventral and ventrolateral views of male tail, showing spicules from both proximal (e) and distal (f) extremities. g Lateral view of female tail with spine-like point. Arrows indicate some small structures, including cervical papillae (a), lobed valve (b), gubernaculum (c), cloacal papillae (d), spicules (e, f) and spine-like point of female tail (g)

Full size image
Fig. 5
figure 5

Line drawings of Ancylostoma ailuropodae n. sp. and comparison of dorsal rays among Ancylostoma spp. a Morphological structures of A. ailuropodae n. sp.: 1, dorsoventral view of anterior region; 2, lateral view of anterior region; 3, lateral view of female vulval region; 4, lateral view of female caudal region; 5, dorsal view of male caudal region; 6, egg. b Ten Ancylostoma species for comparison of dorsal rays: 1, A. taxideae [10]; 2, A. duodenale [9]; 3, A. paraduodenale [6]; 4, A. caninum [58]; 5, A. malayanum [7]; 6, A. kusimaense [9]; 7, A. ceylanicum [9, 11]; 8, A. braziliense [9, 11]; 9, A. tubaeforme [58]; 10, A. ailuropodae n. sp

Full size image

Female. [Based on the allotype and three females.] Body 9.80–16.00 (12.90) mm long, with maximum width at mid-body 560–740 (650); width at anus 270–340 (285). Buccal capsule 170–250 (210) long, 130–190 (160) wide in dorsoventral view; oesophagus 1,280–1,320 (1,300) long, 170–250 (200) in maximum width near base. Cervical papillae 800–1,230 (900), excretory pore 760–950 (820), nerve-ring 600–650 (620) posterior to cephalic extremity. Vulva opens ventrally in posterior third of body, at 2,450–4,686 (3,480) from caudal extremity; vagina relatively short. Female reproductive system amphidelphic, with poorly differentiated vestibule, paired sphincters and infundibula confluent with uterine and ovarian stems (Fig. 5a3). Tail 90–370 (230) long, terminating in acute, spine-like point 9–25 (17) in length (Figs. 4g, 5a4). Eggs oval, 54–71 × 28–38 (62 × 33) (n = 20) (Fig. 5a6).

Remarks

Ancylostoma ailuropodae n. sp. is established based on comparisons to available descriptions among congeners in the global fauna [6, 7, 9, 10, 54,55,56,57,58,59,60,61,62]. Ancylostoma ailuropodae is unequivocally differentiated from congeners by structural characteristics of male and female specimens including body size, arrangement, number and dimensions of buccal teeth and shape of the buccal capsule, and in males by the configuration of the dorsal ray and bursa and lengths of spicules and gubernaculum, respectively (see Table 2 and Fig. 5b). Of note, tooth-number appears to represent one of the key morphological characters separating A. ailuropodae from other species of Ancylostoma. Specifically, (i) A. ailuropodae differs from A. caninum, A. tubaeforme and A. taxideae by the number (2 vs 3 pairs) of ventrolateral teeth; and (ii) from A. ceylanicum, A. braziliense, A. duodenale, A. kusimaense, A. paraduodenale and A. malayanum by the number (2 vs 0/1 pairs) of triangular dorsolateral teeth. Furthermore, the shape of the dorsal rays appears to be another potential species-specific morphological indicator (Fig. 5b). Specimens of A. ailuropodae n. sp. vary from A. tubaeforme by differences in cleft length of two digitations in each branch (Fig. 5b9 and 10) and further from A. taxideae, A. duodenale, A. paraduodenale, A. caninum, A. malayanum, A. kusimaense, A. ceylanicum and A. braziliense by the absence of a third digitation in each branch (Fig. 5b1–8 and 10). Verified specimens of A. genettae, A. protelesis and A. somaliense have not yet been described and these three species were not included in the comparison above. Notably, the adults of both A. pluridentatum and A. buckleyi can be distinguished from the new species by the number of ventrolateral teeth, given that A. pluridentatum has only one pair while A. buckleyi has three pairs according to the original descriptions (e.g. [60, 63]). Based on these morphological attributes, A. ailuropodae is considered to be a previously unrecognized species within the genus Ancylostoma.

Table 2 Key comparisons between A. ailuropodae n. sp. and other congeneric Ancylostoma spp
Full size table

Molecular characterization

To further probe the taxonomic position of A. ailuropodae, both nuclear ITS1-5.8S-ITS2 and mitochondrial cox1 sequences from two representative specimens (codes FTZ1 and FTZ2, respectively) were obtained and subjected to sequence characterization and phylogenetic analyses.

Sequence characterization

For ITS1-5.8S-ITS2, the 734 bp sequences from FTZ1 and FTZ2 were identical and had 52.2% A + T content. BLAST analysis revealed that A. ailuropodae shared the highest identity with A. ceylanicum (99.6%), followed by 98.8% identity with A. duodenale, 97.2% with A. tubaeforme, 95.8% with A. caninum, and 92.6% with A. braziliense. Based on the identities, there were a total of 59 variable positions found in the pairwise alignment of ITS1-5.8S-ITS2, including 17 parsimony-informative and 42 singleton sites (data not shown). Within cox1 sequences, same base composition (A = 23.4%; C = 9.7%; G = 22.6%; T = 44.3%) and sequence length (393 bp) were also observed in these two representative individuals of A. ailuropodae, with an A + T content of 67.7%, a typical mitochondrial nucleotide feature in nematodes (towards AT). BLAST search against GenBank/DDBJ/EMBL databases once again showed the highest nucleotide identity existing between the new species and A. ceylanicum (92.6%), followed by 89.2% identity between A. ailuropodae and A. tubaeforme, 88.6% between A. ailuropodae and A. duodenale, and 86.0% between A. ailuropodae and A. caninum, together corresponding to 99.2–100% identities at the amino-acid level. In terms of identity comparisons, there were a total of 78 variable positions in the 378 bp pairwise alignment, including 28 parsimony-informative and 50 singleton sites.

Further, we located these sites and determined if there were non-synonymous substitutions apparent via comparison of their protein sequences, and the results are shown in Fig. 6. Out of 78 variable base sites, 13 were unique for A. ailuropodae (in red); 16 were identical between A. ailuropodae and one of A. ceylanicum, A. duodenale, A. caninum and A. tubaeforme (in orange); and 49 were shared between A. ailuropodae and any two or three of these four congeneric species (in yellow). Among the 49 variable sites, however, the non-synonymous substitutions A/G250 in A. ceylanicum and T/A251 in A. caninum led to their amino acid changes: I (Ilu) → V (Val) in the former and I (Ilu) → N (Asn) in the latter (see Fig. 6). In addition, analysis of genetic distances using maximum composite likelihood estimates placed A. ailuropodae close to A. ceylanicum with the minimum interspecific evolutionary divergence (0.084), compared with 0.121 evolutionary divergence to A. tubaeforme, 0.127 to A. duodenale, and 0.151 to A. caninum (not shown).

Fig. 6
figure 6

Simultaneous alignments of nucleotide and amino-acid sequences of mitochondrial cox1 genes from Ancylostoma ailuropodae n. sp. and its congeneric species. For the alignments, the nucleotide sequences of cox1 genes were retrieved from the GenBank database (species and accession numbers are indicated in parentheses): Aai (A. ailuropodae n. sp.; KP842921), Ace (A. ceylanicum; KF896601), Aca (A. caninum; AB751617), Adu (A. duodenale; NC_003415), and Atu (A. tubaeforme; AJ407940). The corresponding protein sequences were deduced based on the Invertebrate Mitochondrial Code. Both nucleotide and amino-acid sequences were aligned with Clustal X 1.83 program. Regions of identity in either nucleotide (*) or amino-acid (#) are indicated. Variable base loci in Aai unique for A. ailuropodae n. sp. are highlighted in red; those shared between A. ailuropodae n. sp. and one of A. ceylanicum, A. duodenale, A. caninum and A. tubaeforme are highlighted in orange; and those shared between A. ailuropodae n. sp. and any two or three of these four congeneric species are highlighted in yellow. The non-synonymous substitutions A/G250 in A. ceylanicum and T/A251 in A. caninum as well as their amino-acid changes: I (Ilu)/V (Val) and I (Ilu)/N (Asn) are noted in red with a red star. Percentages of nucleotide and amino-acid identities with respect to Aai are shown at the end of each sequence

Full size image

Phylogenetic characterization

Phylogenetic relationships between A. ailuropodae and other species were inferred from the respective sequences of ITS1-5.8S-ITS2 and cox1 using both MP and BI algorithms and their corresponding tree topologies are shown in Fig. 7. Although the two consistent structures (MP/BI) topologically varied from each other due to the different reference species included, both trees provided an identical, robust phylogenetic resolution for A. ailuropodae within the genus Ancylostoma and for the genus Ancylostoma within the family Ancylostomatidae. Specifically, (i) the two A. ailuropodae specimens clustered together as a monophyletic group that was separated from the other Ancylostoma species. (ii) When the congeneric species A. ceylanicum, A. caninum A. duodenale and A. tubaeforme were considered in our cox1-based analysis (Fig. 7a), A. ailuropodae and A. ceylanicum were more closely related to each other than to A. caninum, A. tubaeforme and A. duodenale, with robust support for tree topology (BP = 95 and PP = 0.99). (iii) When another species, A. braziliense, was added to re-construct this phylogenetic relationship using the ITS1-5.8S-ITS2 data (Fig. 7b), A. ailuropodae remained as the putative sister of A. ceylanicum, regardless of isolate origins (one from the UK and another from India; see Table 1), with high statistical support (BP = 89 and PP = 0.91), which was in agreement with the inferences from the cox1 gene analysis (see Fig. 7a). (iv) The inter-relationships of A. ailuropodae, A. ceylanicum, A. caninum, A. duodenale, A. braziliense and A. tubaeforme in the genus Ancylostoma; U. sanguinis, U. hamiltoni, U. lucasi, U. stenocephala and Uncinaria sp. in the genus Uncinaria; and N. americanus in the genus Necator, demonstrated phylogenetic stability of these monophyletic groups, with the current analyses being consistent with previously proposed molecular phylogenies of the hookworms based on the nuclear ribosomal and mitochondrial DNA data [64,65,66,67,68,69,70,71,72,73,74,75].

Fig. 7
figure 7

Phylogenetic relationships of hookworms isolated from the giant panda with the related hookworms in the family Ancylostomatidae. Phylogeny was inferred on the basis of mitochondrial cox1 (a) and nuclear ITS1-5.8S-ITS2 (b) sequences using both maximum parsimony (MP) and Bayesian inference (BI) methods. The livestock hookworm Bunostomum phlebotomum represented the outgroup species. Taxa belonging to the three major genera including Ancylostoma, Uncinaria and Necator in the family Ancylostomatidae are indicated by differently colored rectangles and shown in both phylogenetic topologies. The numbers along the branches indicate bootstrap values resulting from different analyses in the order MP/BI; values less than 50% are shown as “-”

Full size image

Discussion

Hookworms in the genus Ancylostoma cause significant medical and veterinary disease (ancylostomiasis) in various hosts including humans and domestic and wild mammals [2, 71]. Recent epidemiological surveys revealed that some wild animal-derived species of Ancylostoma are emerging as important helminthic zoonotic agents because of rapid urbanization and increased human-wildlife interactions [11, 13,14,15,16,17,18,19,20,21]. The giant panda, for example, is an endangered and rare wild species in China that has been artificially protected and even partially housed for decades due to habitat loss [33]. Clinically unidentified specimens of Ancylostoma in giant pandas had been confirmed by veterinarians and wildlife biologists since the last century, but their potential zoonotic importance remains to be defined [41]. In the present study, A. ailuropodae n. sp. was isolated from the giant panda, morphologically characterized and demonstrated to be closely related to the anthropozoonotic A. ceylanicum by molecular analysis.

In general, morphological identification is a conventional and authoritative approach to define a new nematode parasite species. Concerning the genus Ancylostoma, several common species can be morphologically differentiated by key characters such as body size, teeth of the buccal capsule and shape of bursal rays (see Table 2 and Fig. 5b; cf. [9]). Similarly, specimens of A. ailuropodae from giant pandas are separated from other hookworms on the basis of either ventrolateral and dorsolateral teeth or dorsal rays, supporting the previous conclusions that teeth and rays were reliable morphological indicators in the differential diagnosis of Ancylostoma spp. [55, 56]. Among this assemblage, it is important to note that A. ailuropodae is clearly structurally distinct from A. malayanum, the only other species of Ancylostoma known in ursid hosts (e.g. Ursus thibetanus) (Table 2), with the implication that each of these species endemic to China may be more closely related to other congeners within the genus. Specimens upon which the description and differentiation of A. ailuropodae n. sp. was based were restricted to fully developed adults and eggs. Further work, using a combined laboratory-egg cultivation and Baermann technique, to describe the morphology of developmentally advanced larval stages is needed to complement morphological characteristics of the new species, and to provide valuable information assisting in species identification and differentiation in this genus [55, 76].

Following our morphological evidence, A. ailuropodae from giant pandas was further confirmed as an independent species by molecular analysis. For example, the internal transcribed spacer region (ITS1-5.8S-ITS2) of the nuclear ribosomal DNA is regarded as an appropriate genetic marker to resolve nematode relationships at the species level [77]. Pairwise comparisons of ITS1-5.8S-ITS2 in A. ailuropodae with congeneric species available in the GenBank database revealed a species-specific sequence feature (containing 59 variable informative sites) and overall identity of 92.6–99.6% among A. ceylanicum, A. tubaeforme, A. caninum and A. braziliense. Furthermore, high bootstrap support was evident, based on phylogenetic analysis of ITS1-5.8S-ITS2 that demonstrated monophyly of A. ailuropodae as the putative sister of A. ceylanicum (see Fig. 7b).

Critically, similar conclusions were reinforced by analysis of the mitochondrial cox1 gene. It should also be noted that cox1 analysis was included because recent studies of the substitution patterns for nematode mitochondrial genes (e.g. cox1 and nad4) revealed that they have utility in identifying and differentiating novel or cryptic species among closely related taxa due to assumed faster evolutionary rates than nuclear genes, features of maternal inheritance and absence of recombination [78,79,80]. Compared to the nuclear ITS, the cox1 of A. ailuropodae appeared to have more variable informative sites (n = 78, including 13 unique loci). Nevertheless, results based on cox1 were consistent with inference from ITS, in revealing a sister-species relationship with A. ceylanicum among a broader assemblage of congeners in the genus. Phylogenetic analysis of cox1 data (Fig. 7a) also supported the contention that A. ailuropodae n. sp. is an independent species which is clearly differentiated from A. ceylanicum, A. caninum, A. tubaeforme and A. duodenale.

Based on the results from integrated molecular and morphological comparisons, we propose that A. ailuropodae of giant pandas is a previously unrecognized and separate species that is closely related to the anthropozoonotic A. ceylanicum within the genus Ancylostoma. Additional information regarding the ultrastructure and genomics of this species and other related hookworms is still required. Broader taxonomic comparisons can provide an increasingly precise morphological and molecular basis for species recognition among hookworms. In addition, there were two non-synonymous base substitutions detected in cox1 genes of A. ceylanicum (A/G250) and A. caninum (T/A251) (Fig. 6) that were confirmed to be fixed and species-specific after homologous comparisons with other A. ceylanicum or A. caninum isolates from two sites in the same geographic area.

Ancylostoma ailuropodae identified here is the fourth hookworm to be described from the Ursidae. Previously, the hookworm Uncinaria yukonensis (Wolfgang, 1956) was characterized in black bears and Uncinaria rauschi (Olsen, 1968) in grizzly and black bears [81, 82]. On the basis of comparisons of morphometric and distribution data of ursine hookworms as well as the historical biogeography of bears, Catalano et al. [13] proposed that there was a relatively recent host-switching event of U. rauschi from black bears to grizzly bears.

The occurrence of A. ailuropodae appears consistent with speciation following a host colonization event to giant pandas apparently from a carnivoran source in sympatry, and further indicates a history of independent association with ursine hosts for the broader ancylostomatid hookworm assemblage. The timing and geographic source for these hookworms cannot be elucidated based on the currently available data and the reduced and relictual distribution for pandas, but a history of host colonization is compatible with the current tree topology (for parasites and hosts) and distribution of carnivore hosts for other species of Ancylostoma (e.g. [24]). We suggest that acquisition of Ancylostoma by giant pandas likely occurred prior to 7 million years ago (MYA) when a shift from an omnivorous diet to one dominated strictly by bamboo (by 2.4 MYA) was underway [25].

Divergence of A. ailuropodae appears to have occurred prior to acquisition of A. ceylanicum by humans in Southeast Asia, and prior to the intense bottlenecking of giant panda populations that has characterized the past century (cf. [33, 83] for details about the history of giant pandas). This interpretation is significant, as it would relate to the historical independence of A. ailuropodae and A. ceylanicum before the current intensified conservation campaign for maintaining giant pandas, and the potential for cross-transmission of both hookworm species when infected humans are in contact. The unique niche and specialized bamboo-feeding habits of giant pandas suggest that colonization in ecological time, related to the source or origin of A. ailuropodae, was unlikely given relative isolation with respect to a sympatric assemblage of carnivorans or other mammals that may serve as hosts for species of hookworms [25, 26, 33]. Parasitological inventory among potential carnivoran hosts in Sichuan and nearby regions remains necessary to demonstrate that A. ailuropodae has a narrow host range and may now be limited to the giant panda [84]; apparent narrow host range, however, does not preclude the potential or capacity for contemporary host switches to humans as a zoonotic parasite given opportunity due to permissive ecological circumstances [85,86,87,88].

Phylogenetic and historical isolation of giant pandas from the broader assemblage of ursids and ursine bears (e.g. [23, 24]) in conjunction with apparent structural divergence (e.g. teeth and configuration of the dorsal ray; Table 2 and Fig. 5b5 and 10) of A. ailuropodae and A. malayanum suggests that independent events of host colonization, separated in space and time, were essential in the process of speciation for these hookworms; molecular data, particularly from A. malayanum, is still needed to explore this hypothesis. Moreover, phylogenetic hypotheses for the Ursidae have placed giant pandas distantly from species of Ursus (and other ursines) near the base of an extensive radiation for bears that unfolded across the late Miocene and Pliocene [24]. Among ursine hosts for Ancylostoma, U. thibetanus (Asiatic black bear) is regarded as the sister of U. americanus (American black bear) and placed among crown species in ursid phylogenies [23, 24]. These relationships alone would serve to refute a coevolutionary hypothesis for Ancylostoma hookworms among bears, conversely supporting a history of independent events of host colonization that have structured this fauna.

Unlike U. yukonensis and U. rauschi in bears, the hookworm from giant panda is genetically similar to other Ancylostoma species (Fig. 7). These respective genera are referred to two independent subfamilies within the Ancylostomatide, namely Ancylostomatinae Looss, 1905 for Ancylostoma and Bunostominae Looss, 1911 for Uncinaria, consistent with extended evolutionary trajectories for these taxa among the hookworms. This suggests the independent origin of A. ailuropodae, supporting monophyly of A. ailuropodae and congeneric species A. ceylanicum, A. duodenale, A. tubaeforme, A. caninum and A. braziliense, and strengthens the close relationship between the giant panda hookworm and A. ceylanicum within the clade. Concurrently it suggests that Uncinaria spp. from pinnipeds and ursids are a distinct monophyletic group in the family Ancylostomatidae [70]. The apparent genetic differences of A. ailuropodae n. sp. in pandas and U. rauschi and U. yukonensis in bears, coupled with their divergent biogeographic and ecological histories suggest this system as a good model for exploring the complexities of diversification and faunal assembly in the evolution of host range and associations among hookworms (e.g. [85,86,87,88]).

The potential for genetic partitioning among possible disjunct populations of hookworms in giant pandas should be considered, as it will reflect information about the timing of colonization to giant pandas and the duration of the history of association. Further, the history of fragmentation and isolation for giant pandas across now isolated mountain systems in southwestern China suggests a complex relationship among hosts and hookworms in this region. Such history could be explored through fecal-based approaches in conjunction with molecular diagnostics to examine occurrence and the extent of genetic diversity and distribution for hookworm parasites among populations and subspecies of giant pandas.

Conclusions

This study is the first to describe and define a new member of the genus Ancylostoma, A. ailuropodae, in the wild giant panda using morphological and molecular criteria. Morphological characters (e.g. ventrolateral (two pairs) and dorsolateral (two pairs) teeth and dorsal rays) distinctly separate A. ailuropodae n. sp. from other congeneric species in the genus Ancylostoma. Further, nuclear ITS1-5.8S-ITS2 and mitochondrial cox1-based genetic distance analysis and phylogenies supported the assertion that A. ailuropodae is independent and shares a sister-species relationship with the anthropozoonotic A. ceylanicum. Although additional molecular evidence is warranted, this finding should enhance public awareness of parasitic hookworms in giant pandas, especially in captive populations that have frequent contact with breeders, veterinarians and even tourists. Moreover, the morphological and molecular data presented here enhances the information on species within the genera Ancylostoma, Uncinaria, and Necator and contributes to a more complete understanding of the taxonomy, diagnostics and evolutionary biology of hookworms.