Stomatal Dimorphism of Neodiplogaster acaloleptae (Diplogastromorpha: Diplogastridae)

Several genera belonging to the nematode family Diplogastridae show characteristic dimorphism in their feeding structures; specifically, they have microbial feeding stenostomatous and predatory eurystomatous morphs. A diplogastrid satellite model species, Pristionchus pacificus, and its close relatives have become a model system for studying this phenotypic plasticity, with intensive physiological and structural studies having been undertaken. However, the many other species that are morphologically and phylogenetically divergent from P. pacificus have not been examined to date. In the present study, the detailed stomatal structure and induction of dimorphism in Neodiplogaster acaloleptae were examined. N. acaloleptae has a fungal feeding stenostomatous morph and a predatory eurystomatous morph. The predatory morph was induced by starvation, high population density, and co-culturing with its potential prey, Caenorhabditis elegans. The feeding behavior of the stenostomatous and eurystomatous morphs of N. acaloleptae was confirmed, demonstrating that 1) the stomatal and pharyngeal movements of the two morphs were basically identical, and 2) the stomatal elements were protracted to cut open the hyphae and/or prey to feed when a N. acaloleptae flips its dorsal movable tooth dorsally and tilts its subventral stegostomatal cylinder ventrally, forming a pair of scissors to cut the food source. The stoma morphology of N. acaloleptae with a single movable tooth and a long stoma is markedly different from that of Pristionchus, which has two movable teeth and a short stoma. It is, however, similar to that of Mononchoides, tentatively a sister to Neodiplogaster.


Introduction
Feeding is one of the most important behaviors of all animals. To ingest nutrients effectively, every animal has evolved feeding structures according to their ecological needs and feeding resources. Therefore, feeding structures can be highly divergent within each animal group. In addition, in some cases, feeding structures differ within a species. For example, the larvae of some lepidopteran and coleopteran insects are grazers that feed on plant material, whereas the adults have a tube-like proboscis to feed on nectar (Lepidoptera) or a brush-like structure to feed on sap flow (Coleoptera) [1,2]. However, feeding structures do not typically differ within each developmental stage, and relatively few cases of structural polymorphism have been reported [3][4][5].
The nematode family Diplogastridae is one of those animal groups with polymorphic adult stomatal structures [6][7][8]. This ancestral characteristic is widespread within the family, and it is often found in genera with complex stomatal morphology to expand the range of their feeding habitats [7]. Therefore, it would be interesting to establish how this differs across the clade, as each difference must be the result of evolutionary tweaking of an existing mechanism. For example, Pristionchus spp. in the stenostomatous (narrow stoma) morph usually feed on bacteria, but animals with the eurystomatous (wide stoma) morph occur in starved cultures, and they are effective predators of other nematodes [9,10].
The stomatal dimorphism of P. pacificus Sommer, Carta, Kim & Sternberg has been examined as a model of developmental plasticity. Intensive physiological and structural studies on P. pacificus and its close relatives have revealed that the nervous system, chemical compounds, and chemical receptor genes are involved in the developmental switches that result in functional differences between the two morphs [6,[11][12][13][14]. However, the stomatal dimorphism and functional morphology of other diplogastrid nematodes have not been sufficiently examined [14]. Thus, a detailed examination of feeding behavior and ecological characteristics of diplogastrids with stomatal morphology clearly different from Pristionchus will help to elucidate further the structural and behavioral aspects of the satellite model system.
The genus Neodiplogaster Cobb contains several dimorphic species, which have a microbial feeding stenostomatous morph and a predatory eurystomatous morph [15]. Susoy et al. [7] demonstrated that, in several diplogastrid nematode species, including an undescribed Neodiplogaster sp., the eurystomatous morph is induced by co-culture with potential prey [e.g., Caenorhabditis elegans (Maupas)]. However, the morphological characteristics of the eurystomatous morphs, with the exception of N. pinicola Steiner and an undescribed Neodiplogaster sp. [7,15], have not been examined in detail.
Neodiplogaster acaloleptae Kanzaki was originally found as a phoretic associate of a longhorn beetle, Acalolepta luxuriosa (Bates) (Coleoptera: Cerambycidae) from Shiga Prefecture, Japan [16]. The species was not successfully cultured at the time. It was thus described based on fixed-type material of stenostomatous individuals only, and no molecular profile was obtained. Thereafter, the species was reisolated from the type carrier beetle, A. luxuriosa, collected from the experimental nursery of the Forestry and Forest Products Research Institute (FFPRI), Ibaraki Japan, and the nematode was successfully cultured on a fungal lawn of Botrytis cinerea Pers. During the culturing procedure, the eurystomatous morph appeared in the old (starved) culture and was artificially induced using the methodology of Susoy et al. [7].
The present study describes the eurystomatous morph of N. acaloleptae and presents a molecular barcode for the species. Environmental factors involved in induction of the eurystomatous morph were examined experimentally, and the feeding behavior was observed using live specimens, focusing on the functional morphology.

Nematode isolation and culturing
Several adults of A. luxuriosa that fed on the leaves and twigs of Aralia elata (Miq.) were collected at the experimental nursery of the FFPRI, Tsukuba, Ibaraki, Japan (GPS: 36°00' 22.36", N, 140°07' 36.79", 23 m a.s.l.) on June 24, 2014. The collected insects were dissected and examined for associated nematodes. During the dissection, dauer juveniles of diplogastrid nematodes were found in the tracheal system, genitalia, and ovipositor. The dauer juveniles were transferred to a nematode growth medium (NGM) plate seeded with Escherichia coli strain OP50 or malt extract agar (MEA) seeded with B. cinerea. The successfully propagated cultures were then examined morphologically under a light microscope (Eclipse 80i, Nikon, Tokyo) for species identification. The species associated with the tracheal system propagated on the fungal lawn and was identified as N. acaloleptae, whereas the other species propagated on bacteria and was identified as a Diplogasteroides species that has subsequently been described as D. luxuriosae Kanzaki & Ide [17]. These two species were subcultured to MEA-B. cinerea (N. acaloleptae) or NGM-E. coli (D. luxuriosae) medium and maintained as laboratory cultures with the codes NKZ339 and NKZ304, respectively.
None of these nematodes or beetles were endangered or protected species, and all materials were collected on the FFPRI campus. Therefore, no special permits or permissions were required for the present study.

Molecular profiles and phylogenetic analysis
Nematode lysates were prepared for polymerase chain reaction (PCR) according to Kikuchi et al. [18] and Tanaka et al. [19]. The molecular sequences of the near-full-length small subunit (SSU; 18S) and D2-D3 expansion segments of the large subunit (LSU; 28S) ribosomal RNA genes were sequenced according to Ye et al. [20] and Kanzaki and Futai [21], respectively.

Induction of the eurystomatous morph
The eurystomatous morph of N. acaloleptae was induced by co-culturing with C. elegans, using a modification of the methodology suggested by Susoy et al. [7]. A mixed-stage population of N. acaloleptae propagated on B. cinerea in which no eurystomatous morph was found was identified by examination of 200 randomly selected males and females. About 200 mixedstaged individuals, which had been gently washed with sterile water, were inoculated onto a B. cinerea fungal lawn on PDA in 60-mm-diameter Petri dishes or an E. coli bacterial lawn on NGM in 60-mm-diameter Petri dishes and incubated at 25°C. In half of these plates, N. acaloleptae were co-cultured with C. elegans; that is, ca. 100 individuals of mixed-staged C. elegans were inoculated onto the plate. Then, 1 and 2 weeks after inoculation, nematodes were isolated from the culture plates using the Baerrmann funnel technique [22], and the number, sex ratio, and ratios of the eurystomatous morph in females and males were evaluated.
The sex ratio was calculated as the number of males/females in the first 200 adult individuals, and the eurystomatous ratio was calculated as the percentage of eurystomatous morph individuals among the first 100 females or males. Five plates were examined for each treatment.
All data, average number, sex ratio, and ratios of the eurystomatous morph in females and males were analyzed using StatView 4.0 for Macintosh (SAS Institute, Cary, NC). The normality of the data was examined using a histogram. One-way analysis of variance (ANOVA) followed by Tukey-Kramer's post hoc test was conducted for comparisons among the treatments. Sex ratio, namely, whether or not there was a bias towards one sex, was examined by the chisquare test.

Morphological observations and micrographs
Live adults of N. acaloleptae obtained from the culturing experiments were observed using the methodology of Kanzaki [23]. Several morphological characteristics were drawn, and micrographs were obtained using a drawing tube and digital camera system (DS-Ri1; Nikon, Tokyo) connected to a microscope. The digital micrographs were edited using Photoshop Elements v. 3 computer software (Adobe, San Jose, CA).
Following the morphological observations, the remaining nematodes were heat-killed at 60°C for 1 minute and fixed in TAF (triethanolamine formalin: 7.0% formalin, 2.0% triethanolamine, 91% distilled water) for 48 hours. Nematodes were then dehydrated through a glycerin-ethanol series using a modification of Seinhorst's method [24], and mounted in glycerin according to the method of Maeseneer and d'Herde [25]. Mounted voucher specimens were used for morphometrics with the aid of a drawing tube.
Observation of the feeding behavior of the eurystomatous and stenostomatous morphs of N. acaloleptae To observe the feeding behavior of the stenostomatous morph, a small number of fungal hyphae and several stenostomatous adult females were mounted alive on an agar pad and observed under a light microscope. Predation by the eurystomatous morph was observed using a modification of the methodology suggested by Fürst von Lieven [26]; that is, dauer juveniles of C. elegans and eurystomatous females of N. acaloleptae were collected on a 2.0% agarose layer of 3-mm thickness, a cover slip was placed over the nematodes, the agar was cut along the edge of the cover slip, and the cut piece of agar and cover slip were transferred to a glass slide for microscopic observation. Videos of the feeding behavior were captured using a digital microscope camera system, MC170 HD (Leica Microsystems, Wetzlar, Germany).

Induction of the eurystomatous morph
The average number, sex ratio, and eurystomatous ratios of females and males of N. acaloleptae are summarized in Table 1. The sex ratio was slightly skewed towards females; that is, the male/female ratio was usually <0.9, irrespective of the culture conditions (P < 0.05). The average number of nematodes was markedly higher on B. cinerea plates than on E. coli plates, but they were similar in corresponding monoculture and co-culture with C. elegans. However, the percentages of eurystomatous females and males were higher in co-cultures compared with in the corresponding monocultures (Table 1). All values are in the form: mean ± standard deviation (sd). Means followed by the same letter in a column are not significantly different (P < 0.05). doi:10.1371/journal.pone.0155715.t001

Morphology of the eurystomatous morph
The body size of the eurystomatous morph was slightly smaller than that of the stenostomatous morph; however, its general morphology, with the exception of the stomatal region, was not clearly different from that of the stenostomatous morph. Therefore, the stomatal morphology of the eurystomatous morph is described below (Figs 1 and 2). Morphometric data for the eurystomatous and stenostomatous males and females from this study and from the original description (stenostomatous morph) are provided in S1 Table. The stomatal structure of the eurystomatous morph. From the anterior, the stoma was separated into three sections: the cheilostom, gymnostom, and stegostom. The stegostom was separated into three subsections: the pro-meso stegostom, metastegostom, and telostegostom. The cheilostom was short and tube-like, with a width about twice its depth, likely separated into 18 narrow plates (= rugae). The gymnostom was short, ring-like, anteriorly narrowing, and slightly shorter in length than the cheilostom. The interior part of the gymnostom internally overlapped with the posterior part of the cheilostom. The stegostom was deep, appearing as a cuticular cylinder. The pro-mesostegostom was difficult to observe, namely, it was thin, forming a dome-shaped cuticular tube, and a little shorter in length than the cheilostom. The metastegostom formed a square-shaped movable dorsal tooth, a square-shaped right subventral tooth with several spinous processes, and a rounded left subventral ridge; both teeth and a ridge were connected to the stegostomatal cylinder (= telostegostom). The dorsal telostegostom had a triangular dorsal apodeme. The right and left subventral telostegostoms bore wing-or parallelogram-shaped subdorsally oriented apodemes. Connections of the myofilaments to the apodemes and tooth were not clearly observed. The dorsal pharyngeal gland extended to the dorsal tooth, and the gland duct penetrated the dorsal tooth to the buccal cavity. Although the basic composition of stomatal morphology was identical in males and females, the stomatal diameter was smaller in males, and the eurystomatous and stenostomatous morphs in males were sometimes difficult to distinguish.
Some additional morphological characteristics of N. acaloleptae found during microscopic observation of the live materials. Small triangular right and left subventral projections homologous to the eurystomatous right subventral tooth and left subventral ridge were found at the metastegostom of the stenostomatous morph. The hemizonid was faintly visible slightly anterior to the excretory pore. The presence of a deirid at the level of the cardia was confirmed. The postdeirid was found at the level of the middle of the vas deferens of males and the posterior end of the posterior ovary of females. It was confirmed that the vas deferens structure of males possessed a glandular part at the junction between the testis and the vas deferens; the anterior part was composed of large and rounded cells, and the posterior part had a thin wall. The male phasmid was located near the v5-7 genital papillae. These characteristics are shown in S1 Fig.

Voucher materials
The mounted materials examined for their morphometric values were deposited in the laboratory collection of the FFPRI with the slide numbers Neodiplogaster acaloleptae Mv01-Mv40 and Fv01-Fv40.
In addition to the mounted voucher materials, the TAF-fixed materials and cultured material of N. acaloleptae (NKZ339) are available at the Forest Pathology Laboratory (N. Kanzaki) and can be provided to other researchers upon request.

Molecular profiles
Molecular sequences for N. acaloleptae were deposited in the GenBank database under the accession numbers LC107877 (near-full-length of SSU) and LC107878 (D2-D3 expansion segments of LSU).

Feeding behavior of N. acaloleptae
Videos of the stenostomatous individuals on fungi and dead nematode bodies are available as supplemental films (S1 and S2 videos), as is a video of the eurystomatous individuals feeding on C. elegans (S3 video).
The feeding of both morphs comprised three movements: 1) protraction of the stomatal elements (cheilostom, gymnostom, and stegostom) during which the dorsal tooth was flipped (or pushed) out to widen the stomatal opening, the subventral metastegostomatal and telostegostomatal elements were tilted slightly ventrally, the anterior pharynx (procorpus + metacorpus) was opened simultaneously, and the posterior pharynx (isthmus + basal bulb) and cardia were closed; 2) retraction of the stomatal elements, flipping back of the dorsal tooth and closing of the anterior pharynx, and then opening of the posterior pharynx; and 3) closure of the posterior pharynx and opening of the cardia to transfer the food to the intestine (Fig 3). If the contents of a prey nematode leaked out through a cut in the cuticle made by the predator, N. acaloleptae kept the stoma open (Fig 3B and 3C) and moved only the pharynx to suck up the food. The movements of stoma and pharynx were more active when N. acaloleptae individuals fed on dead or alive nematodes than when they fed on fungal hyphae. The movements of the stoma and anterior pharynx appeared to be well coordinated, that is, they moved simultaneously. Additionally, the movements of the posterior pharynx and cardia seemed coordinated with each other. However, the coordination of the anterior part (stoma + anterior pharynx) and posterior part (posterior pharynx + cardia) was a little less stringent; that is, the posterior part sometimes did not open immediately after the anterior part closed.

Food sources and induction of the eurystomatous morph
Previous studies have considered the genus Neodiplogaster to be an omnivorous group [15,27]. The preferred food can vary among species. For example, N. acaloleptae and N. crenatae Kanzaki, Masuya, & Kubono were successfully cultured on a fungal lawn but not on a bacterial one [13], whereas an undescribed Neodiplogaster sp. has long been cultured on bacteria in the laboratory [7]. Similarly, some Mononchoides species, most of which are presumed to be predators [27], could be cultured on bacterial lawns [7,28].
The eurystomatous morph is known to be induced by starvation, high population density, and/or co-culturing with its potential prey [7]. In N. acaloleptae, the proportion of the eurystomatous morph usually increased after 1 week of culture, and it increased further after 2 weeks of culture. Additionally, co-culture with a prey nematode increased the proportion of eurystomatous worms compared with the corresponding monoculture treatment. The nematode did not propagate well in monocultures on bacterial lawns, and the effect of population density can likely be neglected; that is, the high ratio of the eurystomatous morph was likely induced purely by starvation (Table 1).
Thus, three factors-starvation, high population density, and co-culturing-seem to induce the eurystomatous morph in N. acaloleptae. Susoy et al. [7] demonstrated that co-culturing strongly induces the eurystomatous morph in Allodiplogaster, Fictor, Micoletzkya, Mononchoides, and Neodiplogaster, but not in Diplogasteriana, Koerneria, Parapristionchus, or Pristionchus. Interestingly, sensitivity to co-culturing is not clearly correlated with the phylogenetic relationship (Fig 4). In the present and previous studies [7], the eurystomatous morph of these nematodes was induced by the presence of potential prey, C. elegans [29]. Although the precise chemical cues have not been identified in most diplogastrids, except for P. pacificus [12], the conditions for the induction of stomatal polymorphism seem to vary among these genera. Detailed examination of the natural prey and chemical analysis based on the chemical profiles, for example, the pheromone profile [30], of potential prey nematodes should be performed to obtain a deeper understanding of the ecological aspects of the eurystomatous morph.
The proportion of eurystomatous morphs was apparently higher in females than in males, and a similar phenomenon has been observed in other diplogastrid nematodes: for example, the eurystomatous morph has been observed only in females in Allodiplogaster spp. [31] and Acrostichus spp. [32,33]. These genera are not phylogenetically close to Neodiplogaster; thus, the more frequent occurrence of the eurystomatous morph in females may be a general tendency in the family.

Functional morphology of stoma and pharynx
The functional morphology of feeding structures has been examined in several diplogastrid species. Fürst von Lieven [26] examined the movement of stomatal elements of Tylopharynx foetida (Bütschli) and Mononchoides spp., and Bumbarger [34], Serobyan et al. [9,10], Ragsdale et al. [6] and Wilecki et al. [14] examined the detailed feeding pattern of both the stenostomatous and the eurystomatous morphs of Pristionchus pacificus and some other diplogastrid nematodes. The feeding behavior of the eurystomatous morph of P. pacificus [34] can be viewed online (https://www.youtube.com/watch?v=vT8bms2RHe8). The feeding pattern seems to differ among these species; that is, the functional morphologies of the feeding structure of the species (genera) with two movable teeth seem to be similar, but those having a single movable tooth (Tylopharynx, Mononchoides, and Neodiplogaster) seem to differ from each other, and from the species with two movable teeth.
The eurystomatous morph of P. pacificus (and other Pristionchus spp.) has wide stoma and large dorsal and right subventral movable teeth. In a video and observations made available previously [14,34], the movements of the stoma and pharynx of eurystomatous Pristionchus and some other species (Parapristionchus, Allodiplogaster, and Micoletzkya) were as follows. First, the nematode catches prey by sucking its surface into the stoma. This occurs by contacting the prey, opening the anterior pharynx to generate negative pressure, and then flipping the two teeth like a pair of scissors to cut open the cuticle. Thereafter, the body contents of the prey are sucked out by an opening/closing movement of the pharynx, during which the movement of teeth seems to be independent of the movement of the anterior pharynx; that is, the teeth move occasionally for mastication, but they open when the pharynx opens. During the sequence of movements, eurystomatous Pristionchus does not show obvious protraction of stomatal elements, probably because its wide stoma is adapted to sucking out the contents of the prey. Interestingly, these four genera show clear differences in feeding behavior between the stenostomatous and eurystomatous morphs [9,10,14]. In these genera, as typified by P. pacificus, the stenostomatous morph has only one movable tooth (small dorsal tooth), and the tooth is located at the bottom of the stoma. Therefore, the tooth basically moves like a flap to help take in bacteria, while the other stomatal elements do not appear to move [9,10,14]. This clear functional difference between the two stomatal morphs was not observed in Neodiplogaster, which basically has very similar but slightly differently scaled stomatal composition in the two morphs.
On the other hand, Mononchoides spp., which also have a wide stoma, show slight protraction of the gymnostom and stegostom [26] (Fig 5). Because the right subventral tooth of Mononchoides spp. is connected to the subventral stegostomatous cylinder, the cylinder must be tilted to move the tooth [26]. Therefore, Mononchoides spp. must suck the prey into their stoma and move the stomatal elements slightly forward, tilting the subventral cylinder and flipping the dorsal tooth to cut open the prey.
The feeding of Tylopharynx de Man, which has a narrow stoma, clearly differs from that of the above two genera (Fig 5). Because of its narrow stoma, inflexible cheilostomatal ring, and procorpus forming a narrow procorpal tube, Tylopharynx cannot effectively suck its food (hyphae) into the stoma, and the teeth are too small to reach the stomatal opening even when they are flipped. Thus, these nematodes protract the gymnostomatal and stegostomatal elements to outside of the stoma. The feeding of Tylopharynx comprises three movements: 1) protraction of the gymnostom to push it onto the hyphae; 2) protraction of the stegostomatal elements extending the procorpal tube to insert the teeth into the hyphae; and 3) tilting of the gymnostom dorsally, tilting of the stegostomatal cylinder ventrally, and flipping of the dorsal tooth to cut open the hyphae to suck out the contents [26]. Thus, the typical movements in Tylopharynx, namely, the extension of the procorpal tube and tilting of the cheilostom, were not observed in Neodiplogaster and Mononchoides [26].
The feeding behavior of N. acaloleptae seems somewhat intermediate between those of Mononchoides and Tylopharynx, but it is more similar to that of Mononchoides spp. Comparing Neodiplogaster with Monochoides, the protraction of the stoma is much clearer in Neodiplogaster, and the dorsal tilt of the anterior end seen in Monochoides (Fig 5F) is somewhat unclear in Neodiplogaster. Specifically, Neodiplogaster expands the cheilostom and the teeth protrude from the stoma, whereas this does not occur in Monochoides [26]. This extensive protraction may be needed to access the prey since the stoma of Neodiplogaster is considerably narrower than that of Mononchoides, even in the eurystomatous morph. Feeding in Neodiplogaster and Monochoides differs from feeding in Tylopharynx in that protraction and ventral tilt of the stoma happens almost at the same time in these species (Fig 5E). In contrast, in Tylopharynx, protraction is followed by the tilt in a second step (Fig 5B and 5C). These differences could be linked to the difference in the width and flexibility of the cheilostom, which is wider and separated into rugae in Mononchoides and Neodiplogaster but is a narrow inflexible ring in Tylopharynx (Figs 3C, 5B, 5C, 5E and 5F).
Susoy et al. [7] suggested that Mononchoides and Neodiplogaster are phylogenetically close and that Tylopharynx and several other diplogastrids form a sister clade of the Mononchoides/ Neodiplogaster clade (Fig 4); that is, Neodiplogaster shares a narrow stylet-like stoma with Tylopharynx, but is phylogenetically closer to Mononchoides. The similarity in the feeding structure and behavior of Mononchoides and Neodiplogaster seems in accordance with the phylogenetic relationship of these genera. At present, the origin of the narrow stoma shared by the above two genera is unclear, namely, whether this occurred independently in the two genera or was an apomorphic characteristic of their common ancestor. More detailed molecular phylogenetic and morphological characteristic studies may enhance our understanding of the origin and evolutionary history of the stomatal structure of N. acaloleptae.