Molecular Diversity of Nematode Parasites in Afrotropical Reed Frogs (Hyperolius spp.)

The diversity of nematodes infecting amphibians is understudied in tropical Africa and unknown in Rwanda. Diversity assessment is hampered by the fact that species descriptions refer mostly to morphological features that are unlinked to DNA sequences of marker genes available in public databases. In this paper, we explore the abundance and diversity of parasitic nematodes in reed frogs Hyperolius kivuensis (n = 115), H. parallelus (n = 45) and H. viridiflavus (n = 100) collected in Rwanda. Five nematode species were identified morphologically as Orneoascaris chrysanthemoides, O. schoutedeni, Gendria leberrei, Aplectana chamaeleonis and Rhabdias collaris. Corresponding DNA sequences of 18S and COI genes were determined and subsequently deposited in GenBank. Aplectana chamaeleonis showed the highest prevalence (8.7%), but O. chrysanthemoides the highest mean intensity of infection (6.0) and largest number (24) of individuals in H. kivuensis. To the best of our knowledge, all amphibian hosts are new records for these nematode species, which are known to infect a wide range of amphibian and reptile species. Our findings suggest that nematode diversity is probably lower than previously assumed due to low host specificity. As morphological species identification is often challenging, our data facilitate molecular identification of adult and specifically larval nematodes found in amphibians of Sub-Saharan Africa.


Introduction
Nematodes are a hyper-diverse worm taxon including free-living and parasitic species [1]. Currently, about 27,000 nematode species have been formally described based on morphological and ecological traits, but conservative estimates yield at least 100,000 species, while some authors expect more than 1 million species [2,3]. The deceptively uniform basic anatomy of nematodes complicates full appreciation of real diversity, rendering morphology-based taxonomy a time-consuming task for well-trained experts [1,4]. Recently, the use of molecular data based on COI and 18S gene sequences represents an independent approach to nematode diversity, available in public sequence databases as GenBank [1,5]. The crucial link between morphological and molecular diversity is still missing for most described nematode species, a serious drawback for the use of sequences to reliably identify voucher specimen [1,5,6].
Parasitic nematodes of vertebrates represent an excellent example for a group in which most species are exclusively identified based on their morphology [7,8]. Amphibians and reptiles have a rich diversity of nematode parasites [9,10], e.g. as many as 173 species have been recorded in the herpetofauna of Mexico [11]. They infect predominately the intestines of their hosts, but also the respiratory tract, body cavity and blood [12]. Nematode parasites may cause a range of damage or alterations, leading to modification of host behavior or metabolism, as demonstrated for lungworms Rhabdias spp. infecting toads [13][14][15]. Thus, the prevalence of nematodes may interfere with the fitness of their amphibian hosts and ultimately affect their survival [16]. A recent enigmatic amphibian decline has been attributed to the combined adverse effects of nematode parasite load and climate change [17].
In this study, we focused on the morphological and molecular diversity of the poorly studied Afrotropical nematode parasites of amphibian hosts. With very few regional exceptions such as Cameroon, Nigeria and South Africa, very little is known about species richness, geographical distribution and host specificity of Afrotropical nematodes [12,[18][19][20]. Presently, identification and determination to the species level is almost exclusively based on morphological features, often relying on very old original publications and descriptions of species found only once or twice [21]. Notable exceptions are recent studies on the Afrotropical Rhabdias lungworms combining morphological and molecular evidence [22,23]. In most genera with widely distributed Afrotropical species such as Amphibiophilus, Aplectana, Falcaustra and Orneoascaris, molecular tools for species identification are still not available. Consequently, most morphological datasets are unlinked to corresponding molecular sets resulting in a very low number of reliably identified Afrotropical nematode parasites with sequence data deposited in GenBank. We present molecular and ecological data on five morphologically identified parasitic nematode species from diverse Afrotropical frog hosts. Molecular markers are used to identify them. We provide a detailed morphological diagnosis including microphotographs of the rarely found species; partial sequences of the nuclear ribosomal 18S-ITS1 locus for all species and the mitochondrial cytochrome oxidase subunit 1 (COI) for one species; and a sequences query with available GenBank entries.

Sampling
In total, 89 adult Kivu reed frogs (Hyperolius kivuensis, HK) and 100 common reed frogs (H. viridiflavus, HV) were sampled from a population inhabiting the cultivated wetlands ("marais", 2.60° S 29.76° E, 1645m a.s.l.) near Huye (Butare), Rwanda [24]. Sampling dates were October 2, 2015 (n = 18 HK; n = 26 HV), October 1, 2016 (n = 4 HV), October 14, 2018 (n = 27 HV), January 10, 2017 (n = 35 HK; n = 25 HV) and March 19, 2017  The detected parasites included 12 species of annelids, nematodes, monogeneans, digeneans and cestodes, but, in this study, we focused exclusively on the nematode species detected in the intestines and lungs of the frogs. We extracted six large ascaridid nematodes from the intestines of one specimen each of H. kivuensis, H. viridiflavus and H. parallelus and transferred them to Eppendorf tubes with 1.5 mL ethanol for molecular identification. One sample each of morphologically different small nematodes found in the small intestines of H. kivuensis and in the lung of H. viridiflavus were treated the same way. All remaining nematodes were fixed in 10% buffered formalin for subsequent morphological examination.
All applicable international, national and/or institutional guidelines for the care and use of animals were followed, and all procedures performed were in accordance with the ethical standards of the University of Koblenz-Landau (approval # Si 2015/01, ethic committee of FB3).

Morphological Identification of Nematode Vouchers
The macroscopically detected nematodes were first observed live using a digital Keyence microscope VHX-6000 and digitally documented using Keyence VHX-H1M1 software. Each formalin-fixed nematode was placed in a drop of lactophenol on a microscope slide, cover slipped and examined under an Olympus BX 50 microscope equipped with a high-resolution camera Olympus DP20 for documentation with the Cell Imaging software. The nematodes were identified to genus level based on the corresponding keys [7,8] and by subsequent comparison with the original descriptions. Parasite terminology is used in accordance with [26]. Vouchers specimens were deposited at the Centrum für Naturkunde (CeNak)-Center of Natural History, Universität Hamburg-Zoologisches Museum, Germany.

Nucleic Acid Extraction and Polymerase Chain Reactions
The ethanol-fixed nematodes were incubated in 1.5 mL reaction tubes, with open cap, at 40 °C in a Thermomixer (Eppendorf, Germany) until the ethanol was completely evaporated. The sample was mixed with 180 µL ATL buffer and 20 µL proteinase K, incubated at 56 °C until the specimen was completely lysed, and then further processed following the protocol for DNA purification from tissues of the QIAamp DNA mini kit (Qiagen, Germany). The extracted DNA was used to amplify parts of nuclear ribosomal locus (18S-ITS1-5.8S) and the mitochondrial cytochrome oxidase subunit 1 (COI) locus (Table 1). Polymerase chain reaction (PCR) was conducted using Taq PCR core kit as recommended by the manufacturer (Qiagen). The following programs were used: 40 cycles of 1 min at 94 °C, 1 min at 53 °C and 1 min at 72 °C for 18S-ITS1-5.8S; 40 cycles of 1 min at 94 °C, 1 min at 56 °C and 1 min at 72 °C for COI. All PCR cycles were initiated with a denaturation step for 3 min at 94 °C and terminated with an extension step of 72 °C for 7 min. Amplicons were separated on a 1% agarose gel, stained with ethidium bromide and visualized on a UV transilluminator. Prior to sequencing, PCR products were purified using QIAquick Purification Kit (Qiagen).

Sequencing and Phylogenetic Analysis
For bidirectional sequencing, the same primers as for the PCR were applied ( Table 1). Assembly of DNA sequence files was conducted with DNA Baser (Heracle BioSoft, Romania) and primer sequences were clipped. Due to poor performance of reverse primer IR8 as a sequencing primer, most sequences only contain partial 18S and ITS1 regions. Sequences were deposited in GenBank (Table  2). Closest matches of sequences were identified by a BLAST search against GenBank entries [27].
Phylogenetic trees were computed from the alignments 1 using Bayesian interference in MrBayes [28]. Posterior probabilities were approximated over 3,000,000 generations. Positions containing gaps and missing data were eliminated. Trees were rooted by the sequences of outgroup species Cylicospirura petrowi (Orneoascaris tree), Ichthyobronema hamulatum (Gendria and Aplectana tree) and Strongylus vulgaris (Rhabdias tree), retrieved from GenBank. The resulting trees were manually refined using Fig Tree v1.4.2. Pairwise nucleotide comparison for COI sequences was generated using GeneDoc [29]. Table 1. Primers used for amplification and sequencing.

Gene
Primer Sequence Reference

Results
The diversity of nematodes in the three representatives of the genus Hyperolius included five species with 60 specimens. Most nematodes were found within the small and large intestines (N = 59) of the frogs examined; only a single specimen was detected in the lungs. The body cavity was nematode-free in all frogs. The nematodes were morphologically identified as Orneoascaris chrysanthemoides Skrjabin, 1916 [21] (hosts: HK, HV), Orneoascaris schoutedeni (Baylis, 1940) Table 3. Blast search in GenBank for corresponding gene sequences to aid molecular identification were in vain for all species.

Orneoascaris chrysanthemoides (Ascaridoidea): Morphological, Ecological and Molecular Features
The most frequent nematode species at the sampling site in Huye (Southern Province) was O. chrysanthemoides, infecting the small and large intestines of a few individuals of H. kivuensis and H. viridiflavus ( Figure 1A-C). The observed features of external morphology agreed with those of the original description [37,38], specifically the presence of relatively large denticles extending over the whole border of lips and of very slender, cylindrical, long spicules in males. The only morphologically studied female O. chrysanthemoides was considerably larger than males (Body length 18.0 mm vs. 11.6-14.6mm). The prevalence rates (2.0-3.5%) was similar to that reported for Leptopelis christyi (6% [39]), but varied during the rainy season. At the beginning of the rainy period in October, the prevalence rate was 0% in H. kivuensis and 1.1% in H. viridiflavus, increasing to 8.8% in H. kivuensis and 4.0% in H. viridiflavus in January, and decreased to zero again at the end of the rainy period. The average intensity of infection and the maximum number of nematodes per host in H. kivuensis exceeded considerably those of H. viridiflavus (Table 3). A captive-held H. viridiflavus male with two intestinal O. chrysanthemoides, but otherwise parasite-free, died 31 days after collection in the wild. Orneoascaris chrysanthemoides never co-occurred with other nematode species in any Hyperolius specimen examined, but with up to two other trematode species (metacercariae) in H. kivuensis. Voucher specimens were deposited at the Zoological Museum of the University of Hamburg (ZMH N 14025; N = 1 originating from HV, and ZMH N 14026; N = 19 from HK).
We succeeded in amplifying a partial 18S and partial ITS1 sequence of two O. chrysanthemoides nematodes collected from H. viridiflavus. These two sequences were identical, but the sequence derived from the morphologically very similar O. schoutedeni collected from H. kivuensis in Musanze (Northern Province) showed a considerable difference (see below).

Orneoascaris schoutedeni (Ascaridoidea): Morphological, Ecological and Molecular Features
At the sampling site in Musanze (Northern Province), we found three adult O. schoutedeni in the small and large intestines of a single H. kivuensis female ( Figure 1D-F). The observed features of external morphology were in full agreement with those described previously [33,38], distinguishing specimens from O. chrysanthemoides by inconspicuous dentigerous ridges at the border of lips and by short spicules and prominent lips of the cloacal aperture in males. In all other morphological aspects, O. schoutedeni and O. chrysanthemoides were very similar. A female O. schoutedeni was slightly larger than two males (Body length 21.1mm vs. 11.3-19.5mm). Prevalence (0.9%) and the maximum number of nematodes per host (n = 3) were lower than those of O. chrysanthemoides (3.5%, n =16; this study) in H. kivuensis at the Huye locality (Table 3). Orneoascaris schoutedeni did not co-occur with other nematode species, but with two trematode species (metacercariae). Voucher specimens were deposited at the Zoological Museum of the University of Hamburg (ZMH N 14024; N = 2 originating from HK).
The partial 18S and partial ITS1 sequence of one O. schoutedeni specimen showed a 3% sequence difference to those of O. chrysanthemoides. A Bayesian tree build with closest sequence matches from GenBank was computed based on a cured sequence dataset of 642 nucleotides ( Figure 1G). The topology of the resulting tree is significantly supported by posterior probability values. The Orneoascaris spp. form a 100% supported sub-clade.

Gendria [Chabaudus] leberrei (Seuratoidea): Morphological, Ecological and Molecular Features
Six nematodes assignable to the genus Chabaudus [40] were collected from the small and large intestines of three H. parallelus males (Figure 2A-C). Recently, the genus Chabaudus has been placed into the synonymy of Gendria because of the absence of reliable morphological features for diagnosis [41]. As independent molecular support is not available for any view on the generic taxonomy, we refer to the nematodes identified based on their morphological features as Gendria [Chabaudus] leberrei. A female was considerably larger than two males (Body length 16.9mm vs. 11.3mm, 12.7mm). The size range observed falls within the range reported for specimens collected from other anuran hosts: females 8.4-19.3mm and males 5.6-14.6mm [34,[42][43][44]. Prevalence (6.7%) and intensity of infection (2.0; Table 3) were much lower than observed in Xenopus spp. (83.3-100% and 5.8-7.5% [42]). Gendria [Chabaudus] leberrei did not co-occur with other nematode or trematode species. Voucher specimens were deposited at the Zoological Museum of the University of Hamburg (ZMH N 14023; N = 2 originating from HP).
We succeeded in amplification of partial 18S-ITS1-5.8S sequence of two G.
[Chabaudus] leberrei specimens showing 100% sequence identity. A Bayesian tree build with closest sequence matches from GenBank was computed based on a cured sequence dataset of 415 nucleotides of partial 18S sequence ( Figure 2D). The topology of the resulting tree is significantly supported by posterior probability values. The Gendria specimens form a 91% supported sub-clade with the most closely related species retrieved from GenBank, the quimperiid Paraseuratum sp.

Aplectana chamaeleonis (Cosmocercoidea): Morphological, Ecological and Molecular Features
A small nematode parasite found exclusively in the large intestine of H. kivuensis was A. chamaeleonis ( Figure 3A-C). The observed features of external morphology were in agreement with those described previously [45,46], specifically the position of postanal papillae in males. A female was considerably larger than five males examined (Body length 6.1mm vs. 2.5-4.1mm). The size range observed falls within the range reported for specimens collected from other hosts: females 3.7-6.7mm and males 2.6-3.9mm [35,45,46]. Prevalence (8.8%; Table 3) was much lower than reported for Leptopelis spp. (25-38% [39]), but decreased continuously from the beginning of the rainy period (11.2% in October) to the end (8.6% in January and 3.8% in March). In one host specimen each, A. chamaeleonis co-occurred either with the parasitic annelid Dero rwandae (Naididae) or with undetermined Echinostomatidae metacercariae. Long-term coexistence of nematodes and host was evident in 6 out of 10 captive-held H. kivuensis specimens in which we detected A. chamaeleonis 20 months following capture in the wild. Voucher specimens were deposited at the Zoological Museum of the University of Hamburg (ZMH N 14027; N = 19 originating from HK).
Fragments of the 18S-ITS1 sequence for four isolates of A. chamaeleonis showed 100% identical sequences. Comparison of a 475 bp-long subfragment of these sequences with the only GenBank entry of another Aplectana species (MH836325) showed 7% sequence divergence. We used a 18S rDNA fragment for computing a Bayesian tree based on closest GenBank entry matches (Figure 2). The topology of the resulting tree is supported by high probability values. The sequences of A. chamaeleonis built a subclade with a Cosmocercidae species previously collected from the smooth newt Lissotriton vulgaris [6]. Note that we failed in amplification of a fragment of the COI gene. Therefore, we cannot compare our specimens with other COI-GenBank entries of Aplectana species.

Rhabdias collaris (Rhabditoidea): Morphological, Ecological and Molecular Features
A single female R. collaris was found in the left lung of a H. viridiflavus male from the population in Huye (Table 3, Figure 4A). The lungs of all other Hyperolius specimens examined in this study were nematode-free. The observed features of external morphology were in agreement with those described previously [36], specifically the inflated and muscular cephalic end and the large uteri, filled with numerous, about 100µm long eggs. Body length was 6.1mm, i.e. smaller than the thus far known seven specimens (8.4-10.2mm) collected from the lungs of Leptopelis vermiculatus and Hyperolius spec. in Amani, Tansania [36].
The topology of the Bayesian tree based on the partial ITS1 sequences (302 nucleotides) is supported by significant posterior probability values ( Figure 4B). Note that Afrotropical Rhabdias spp. do not form a single clade ( Figure 4B). The compared fragment of the COI gene of the six Afrotropical species of Rhabdias showed different levels of pairwise sequence variability (Table 4). Rhabdias engelbrechti and R. picardiae showed the highest level (89%) of sequence similarity to R. collaris.

Discussion
Our study is the first to provide molecular data on five morphologically described Afrotropical nematode species infecting reed frogs Hyperolius spp. DNA sequences facilitate identification of specimens in any developmental stage and enabling the unambiguous distinction between morphologically similar species. The reed frog species are all new host records for theses helminths. The detection of Orneoascaris schoutedeni in H. kivuensis is the first evidence for an amphibian host, while it was previously known only from reptile species [33,38]. This and previously published studies demonstrate that the nematode species reported here are capable of infecting a wide range of African amphibian and reptile host species, as summarized in Table 5. The absence of a strong host specificity facilitates the wide geographical distribution in Sub-Saharan Africa, as demonstrated by Orneoascaris chrysanthemoides currently known from Cameroon, DR Congo, Gabon, Gambia, Rwanda, Uganda and Zambia [21,[37][38][39]47,48]. Considering the fragmentary nature of actual knowledge on the geographical distribution of these nematodes and the usually low prevalence in hosts, wide Sub-Saharan ranges may be characteristic for many of them.

Host-Parasite Ecology
All but one nematode species were detected in the small and large intestines of the three Hyperolius species collected in Rwanda. However, their prevalence was lower than that in other amphibian hosts [39,42]. The low prevalence and intensity of nematode infections in Hyperolius spp. may indicate either some kind of resistance against infections or alternatively short survival of infected hosts. Little is known about the impact of intestinal nematodes on the survival of anuran hosts, but the effect seems to be small [54,55]. Captive-held H. kivuensis survived infections with A. chamaeleonis more than two years, suggesting indeed a low effect on survival (this study). In contrast, a H. viridiflavus infected by two O. chrysanthemoides died a month after capture, suggesting an impact of large intestinal nematodes on survival. Experimental infections with Orneoascaris are needed to quantify the potential impact on Hyperolius survival and to access the adaption level of parasite and host.
Lung infections were limited to a single H. viridiflavus specimen by a single R. collaris, unlike the high prevalence and infection intensities by Rhabdias spp. in other African and Neotropical anuran hosts [12,15,18,19,22,56]. Lung-dwelling nematodes are known to decrease significantly the host´s fitness and cause reduced survival [13,56,57]. The proximate cause for the almost complete absence of lung infections in the Rwandan frog hosts remains enigmatic because Leptopelis and Hyperolius hosts from the neighboring Tanzania were infected by R. collaris [36].

Phylogenetic Implications
The phylogenetic relationships of Orneoascaris spp. and Gendria [Chabaudus] leberrei are currently under debate. In lip morphology, the ascaridoid genus Orneoascaris resembles Ophidascaris [38], but morphological evidence is too weak for a reliable phylogenetic hypothesis [7,8]. Our molecular data support the specific status of the morphologically very similar O. chrysanthemoides and O. schoutedeni, leaving open their relationship to O. sandoshami and an undescribed schoutedeni-like species from Bornean lizards [38]. Due to the fragmentary nature of molecular data, the enigmatic relationship of the genus Orneoascaris to other genera remains unresolved, but Contracaecum, Anisakis and Raphidascaris do not seem to be closely related ( Figure 1G).
The phylogenetic relationships within the seuratioid family Quimperiidae are not wellunderstood [7,58,59]. In our study, Gendria [Chabaudus] leberrei and Paraseuratum sp. formed the sister group of a clade including several cosmocercoid species, whereas the quimperiid Ichthyobronema hamulatum resolved as an outgroup to both clades ( Figure 2D). The association of Paraseuratum sp. with Spectatus spectatus (Cosmocercoidea: Kathlaniidae) in earlier studies [58,59], and more distantly to Megalobatrachonema terdentatum (Cosmocercoidea: Kathlaniidae [6]; this study) also indicates that the genera Gendria and Paraseuratum may be misplaced in the Quimperiidae, but actually belong to the Cosmocercidae. More molecular data on Quimperiidae genera are needed to resolve reliably taxonomic assignment of Gendria and Paraseuratum.
The cosmocercid genus Aplectana includes currently more than 50 species with increasing species descriptions, further expanding its diversity [45,[60][61][62]. Unfortunately, all descriptions are based exclusively on morphological features. To the best of our knowledge, we provide the first DNA sequence corresponding to a morphologically identified Aplectana species. Phylogenetic relationships with other genus members based on molecular features remain unknown.
One of the few nematode genera with ample information on morphological and molecular features are the lung-dwelling Rhabdias parasites [23]. We contribute the first molecular data for the morphologically described R. collaris. Despite the usage of shorter sequence length, the phylogenetic tree based on 18S-ITS1 sequences obtained here for Rhabdias species is very similar to a recently published phylogenetic tree [23]. The lineage including R. collaris consists of species parasitizing different amphibian species occurring in the Afrotropical, Palearctic or Oriental regions ( Figure 4B). Interestingly, our study suggests that Afrotropical Rhabdias spp. form three evolutionary lineages (R. collaris, R. africanus and R. sylvestris lineages).

Conclusions
The diversity of nematodes in frogs from Rwanda, a previously unstudied Afrotropical region, exclusively includes species which have already been described from other amphibian and reptile hosts. Afrotropical diversity of parasitic nematodes may be lower than previously assumed because host specificity is not strongly developed. Phylogenetic insight into the supra-and infra-generic relationships of the four Afrotropical nematode genera detected in this study requires considerably more molecular data for morphologically identified species than currently available. Here, we contribute a dataset linking morphology and molecular features of five nematode species. We recommend that experts on nematode morphology and molecular biologists should join their efforts to establish a sustainable molecular library of DNA sequence data linked to morphological identifications, which in future may serve equally for taxonomic purposes and for unravelling phylogenetic relationships.