Host orientation using volatiles in the phoretic nematode Caenorhabditis japonica

Finding a host is the most critical step for parasitic and phoretic nematodes. Host-searching nematodes respond to different types of host-related stimuli such as vibrations (Torr et al., 2004), temperature (Byers and Poinar, 1982), electric fields (Shapiro-Ilan et al., 2009) and chemical compounds (Pye and Burman, 1981; Grewal et al., 1993; Shapiro et al., 2000). However, host-searching strategies vary among species or ecological niches, and mechanisms of host finding, including host recognition and orientation, are poorly understood. Volatiles could be a useful cue for nematodes to orient, particularly those from distant locations. The recognition and importance of volatiles in host searching have been reported in entomopathogenic nematodes (EPNs). The infective juveniles (IJs) of steinernematid and heterorhabditid EPNs respond to volatiles from insects and CO₂ (Lewis et al., 1993; Campbell and Kaya, 1999; Hallem et al., 2011; Dillman et al., 2012). IJs are also attracted to herbivore-induced plant volatiles (HIPVs), which facilitates host detection, resulting in higher infection rates (Rasmann et al., 2005; Hiltpold et al., 2010; Ali et al., 2010; Ali et al., 2011).

Caenorhabditis japonica Kiontke, Hironaka and Sudhaus is a bacterial-feeding nematode that forms a species-specific and female host-biased phoresy with the burrower bug Parastrachia japonensis Scott (Kiontke et al., 2002; Yoshiga et al., 2013). Dauer larvae (DL), a developmentally arrested and phoretic stage of the nematode and a developmental analogue of IJs in EPNs, are predominantly found on P. japonensis adult females throughout the year in fields. Only the nematode benefits from this phoresy. The nematode uses the host insect to prolong its own survival (Tanaka et al., 2012) and for transport to food resources, such as the nest of the host insect where the mother insect stores fruits of Schoepfia jasminodora Siebold & Zuccarini for her nymphs, as well as eggs and nymphal carcasses (Okumura et al., 2013b; Yoshiga et al., 2013). Thus, the phoresy is essential only for the nematode, but the mechanisms of forming the species-specific phoresy are not well understood.

In our previous study, we demonstrated that C. japonica DL specifically embark on P. japonensis and are attracted to the hexane extracts containing body surface components of P. japonensis (Okumura et al., 2013a). These studies indicate the presence of a species-specific kairomone that is directly released by the host insect and attracts DL. However, hexane extracts contain not only volatiles but also non-volatiles, and it is difficult to distinguish whether C. japonica DL are attracted to volatiles or non-volatiles in the agar plate assay, which is commonly used in nematode chemoattraction studies (Bargmann, 2006). Thus, a new method to evaluate volatile attraction is necessary.

Because C. japonica DL embark on a host insect that wanders on the soil surface, DL have to detect the presence of a host on the soil surface and quickly orient themselves for embarkation on the host. Thus, it is possible that C. japonica DL recognize volatiles from a host and orient themselves to the host. To examine this possibility, we developed an assay system for C. japonica DL using a Y-tube olfactometer and demonstrated their recognition of and orientation to a host insect.

Y-tube olfactometers are commonly used in experiments involving insect olfactory responses (Smith et al., 1994). Volatiles pass through one arm of the Y-tube and clean air passes through the other; this gives the possibility of choosing between two different stimuli. However, Y-tube olfactometers are not commonly used in nematode research because of some technical difficulties such as conditioning of humidity and locomotion activity. Thus, we modified a Y-tube olfactometer so that C. japonica DL could freely move in it (Fig. 1A,B). Unlike arthropods, nematodes require high humidity for their movements; thus, using water agar in the assay arena improved nematode movement. In addition, the assay arena was positioned vertically because C. japonica DL have a tendency to move upward during host searching (Okumura et al., 2013c); this improved nematode locomotion and response to volatiles (data not shown).

In the olfactory assay using P. japonensis as the test insect, C. japonica DL moved towards the direction of air flow from P. japonensis soon after the initiation of the experiments, and 43%

(range 13–89%) of inoculated nematodes were found in the test area while only 16% (range 0–49%) were in the control area. In contrast, more than half of the inoculated DL were recovered from the inoculation area in each assay when Acanthocoris sordidus Thunberg, Acrida cinerea Thunberg or Oxya yezoensis Shiraki were used as the test insect, and no attraction toward these insects was observed. The percentage of nematodes in the test area was significantly higher when P. japonensis was used than when other insects or CO₂ was used (Fig. 1C; P<0.001). CO₂ emitted by animals as a by-product of respiration attracts DL of C. elegans and IJs of EPNs (Hallem et al., 2011). In our study, however, CO₂ failed to attract C. japonica DL (Fig. 1C). The three insects used in this study, A. sordidus, A. cinerea and O. yezoensis, also produced CO₂ yet did not attract C. japonica, which suggests that CO₂ may not be a cue for C. japonica. These results also indicate that the movement of C. japonica DL towards the air from P. japonensis appears to be evoked by specific volatiles produced by this host insect. Our results demonstrate that C. japonica DL recognize volatiles from their specific host and orient toward these. To the best of our knowledge, this is the first report that demonstrates the presence of a species-specific volatile kairomone that attracts host-finding stage nematodes.

Many nematodes utilize volatiles for host searching but the utilization of volatiles could vary among species with their ecological background. Although the ecology of EPNs in the wild is not well understood (Strong, 2002), EPNs have a broad host range and can infect most insect orders (Poinar, 1979), and IJs are attracted by different types of volatiles such as HIPVs, insect volatiles and CO₂ (Lewis et al., 1993; Campbell and Kaya, 1999; Hallem et al., 2011; Dillman et al., 2012). In contrast, C. japonica forms a specific phoresy with P. japonensis, and DL are attracted only by volatiles from P. japonensis. The differences in response to volatiles between IJ EPNs and C. japonica DL may reflect the variations in host range and host-searching strategies. To effectively approach a variety of potential hosts, IJs of EPNs may utilize simple indicators of biological activities such as CO₂ as basic host stimuli along with various volatiles as additional information. In contrast, C. japonica DL may not utilize these simple indicators as host stimuli; they may instead exclusively depend on a specific chemical cue from the host. The decrease in response to simple indicators from biological activities and increase in the sensitivity and response to a specific kairomone may be important factors for host recognition and orientation in C. japonica. Analysis of the volatile components of P. japonensis that attract C. japonica DL and studies on sensory responses to known chemicals are thus necessary to better understand the mechanism and evolution of host recognition and orientation in C. japonica.

Caenorhabditis japonica strain H1 isolated from an adult female of P. japonensis collected from Hinokuma Mountain Prefectural Park, Kanzaki City, Saga Prefecture, Japan, was used for the experiments. Dog food agar medium (Hara et al., 1981) seeded with Escherichia coli (Migula) Castellani and Chalmers OP50 was used to obtain a large number of nematodes. Nictating C. japonica DL were collected as described previously (Tanaka et al., 2010). In brief, a sterile yellow 200 μl pipet tip (Watson, Fukaekasei Co., Ltd, Tokyo, Japan) was vertically placed such that the tapered side was up at the centre of a 100 ml culture bottle of dog food agar medium in which E. coli OP50 and the nematodes were inoculated. Approximately 5 days after inoculation, masses of nictating DL at the point of the yellow tip were picked up using a fine needle and used for subsequent experiments.

To determine whether volatiles from insects attract nematodes, we constructed an olfactometer that was a modified version of that used in the study of the Japanese horntail (Matsumoto and Sato, 2007). The olfactometer used in the current study is illustrated in Fig. 1A. The aeration unit (MAU-2, Tokyo Rikakikai Co., Ltd, Tokyo, Japan) was set to obtain an air flow rate of 0.8–1.0 l min⁻¹. Air was passed through ~150 ml of distilled water in a 300 ml glass bottle to remove odours and to maintain a high humidity, and then divided into two lines by the first glass Y-shaped tube joint (Y-tube: inner diameter, 12 mm; Sansyo, Tokyo, Japan). Air flow was adjusted to a rate of 0.4 l min⁻¹ using a flow meter (RK200V, Kofloc, Kyoto, Japan), which was connected to a 100 ml Pyrex test or control bottle. The test bottle contained a test insect, whereas the control bottle had no insects. The air passed through the test and control bottles connected to the assay arena (the second Y-tube), whose longitudinal half volume was filled with 1.5% water agar. Silicone tubes were used for all connections between the aeration unit, water bottles, flow meters and Y-tubes. Approximately 20 μl of the nematode suspension containing 30–100 DL was placed onto the water agar at the centre of inoculation area (Fig. 1B), the arena was set vertically, and aeration was started. Approximately 10 min after starting aeration, we disconnected the arena, and the nematodes in each area (test, control, inoculation and downward areas; Fig. 1) were counted. To test the effect of CO₂, a bottle of compressed CO₂ was connected to one of the flow meters, and the other flow meter was connected to the aeration unit. Air flow was adjusted to a rate of 0.4 l min⁻¹ using a flow meter. Insects used for the olfactometry assay are listed in Table 1. The number of insect individuals of P. japonensis, A. cinerea, O. yezoensis and A. sordidus used for assay was 16, 2, 1 and 1, respectively. The number of replicates of P. japonensis, A. sordidus, A. cinerea, O. yezoensis and CO₂ experiments was 121, 8, 20, 10 and 21, respectively.

Percentages of nematodes in the test area were compared among the insects using a generalized linear model with a Gaussian distribution (StatView Ver. 4.54; Abacus Concepts, Inc., NJ, USA).

We thank Dr S. Tojo for his valuable comments on this study and Dr A. Yamawo for his advice on data analysis.