Effect of Photorhabdus luminescens phase variants on the in vivo and in vitro development and reproduction of the entomopathogenic nematodes Heterorhabditis bacteriophora and Steinernema carpocapsae

Authors

  • Richou Han,

    Corresponding author
    1. Institute for Phytopathology, Department for Biotechnology and Biological Control, Christian-Albrechts-University Kiel, Klausdorfer Str. 28-36, 24223 Raisdorf, Germany
    Search for more papers by this author
  • Ralf-Udo Ehlers

    1. Institute for Phytopathology, Department for Biotechnology and Biological Control, Christian-Albrechts-University Kiel, Klausdorfer Str. 28-36, 24223 Raisdorf, Germany
    Search for more papers by this author

*Corresponding author. Present address: Guangdong Entomological Institute, 105 Xingang Road W., Guangzhou 510260, PR China. Tel.: +86 (20) 84191089; Fax: +86 (20) 84183704; E-mail: en@gdei.gis.sti.gd.cn

Abstract

Photorhabdus luminescens (Enterobacteriaceae) is a symbiont of entomopathogenic nematodes Heterorhabditis spp. (Nematoda: Rhabditida) used for biological control of insect pests. For industrial mass production, the nematodes are produced in liquid media, pre-incubated with their bacterial symbiont, which provides nutrients essential for the nematode's development and reproduction. Particularly under in vitro conditions, P. luminescens produces phase variants, which do not allow normal nematode development. The phase variants were distinguished based on dye absorption, pigmentation, production of antibiotic substances, occurrence of crystalline inclusion proteins and bioluminescence. To understand the significance of the phase shift for the symbiotic interaction between the bacterium and the nematode, feeding experiments tested the effect of homologous and heterologous P. luminescens phase variants isolated from a Chinese Heterorhabditis bacteriophora (HO6), the Heterorhabditis megidis type strain from Ohio (HNA) and the type strain of Heterorhabditis indica (LN2) on the in vivo and in vitro development and reproduction of the nematode species H. bacteriophora (strain HO6) and another rhabditid and entomopathogenic nematode, Steinernema carpocapsae (A24). In axenically cultured insect larvae (Galleria mellonella) and in vitro in liquid media, H. bacteriophora produced offspring on phase I of its homologous symbiont and on the heterologous symbiont of H. megidis, but not on the two corresponding phase II variants. In solid media, nematode yields were much lower on phase II than on phase I variants. On the heterologous phase I symbiont isolated from H. indica the development of H. bacteriophora was not beyond the fourth juvenile stage of the nematode in any of the media tested, but further progressed on phase II with even a small amount of offspring recorded in solid media. Infective juveniles of S. carpocapsae did not develop beyond the J3 stage on all phase I P. luminescens. They died in phase I P. luminescens isolated from H. bacteriophora. Development to adults was recorded for S. carpocapsae on all phase II symbionts and offspring were produced in all media except in liquid. It is concluded that a lack of essential nutrients or the production of toxins is not responsible for the negative impact of homologous phase II symbiont cells on the development and reproduction of H. bacteriophora. The infective juveniles of H. bacteriophora retained cells of the homologous phase I symbiont, but not phase II cells and cells from heterologous symbionts, indicating that the transmission of the symbiont by the infective juvenile is selective for phase I cells and the homologous bacterial associate.

1. Introduction

Photorhabdus and Xenorhabdus (Enterobacteriaceae) are symbionts of entomopathogenic nematodes of the genera Heterorhabditis and Steinernema (Nematoda: Rhabditida), respectively. These bacto–helminthic complexes are safe biocontrol agents [1], which are used to manage soil-borne insect pests in ornamental plants, turf, mushrooms and strawberries [2,3]. For use in plant protection, the nematodes are mass-produced on solid or in liquid media [4–8]. Nematode cultures are inoculated with dauer juveniles (DJs). The scientific term ‘dauer’ (German ‘enduring’) is used for developmentally arrested, third stage juveniles of rhabditid nematodes, which differ structurally from all other stages of the nematode [9]. The free-living DJ represents the infective stages of the nematodes. They carry cells of their specific (homologous) bacterial symbiont in the anterior part of the intestine. When entering an insect host, the DJs exit from the dauer stage (recovery) in response to food signals present in the haemeolymph [10,11], start feeding and develop through another juvenile stage (J4) to adults. During recovery they release the symbiotic bacteria, which multiply, kill the insect within 2–3 days and establish suitable conditions for nematode reproduction. The offspring population consists of DJs and adults, which can produce another adult generation before the DJs emigrate from the insect cadavers searching for new hosts. During the development of DJs the symbiotic bacteria are no longer digested but retained within the anterior region of the DJ intestine [12,13]. A schematic nematode life cycle is shown in Fig. 1.

Figure 1.

Schematic life cycle of entomopathogenic nematodes. Free-living DJ develops through the third and fourth juvenile (J3, J4) to the adult stage (only female phenotype shown), which produce eggs from which DJs develop through the first and second juvenile stage (J1, J2). Dark gray areas mark the intestine, lighter areas represent those parts of the primordial and adult gonads which are hidden by the intestine.

For a high and continuous production of offspring, nematodes of both genera depend on the presence of their symbiotic bacteria. To a limited extent Steinernema spp. can reproduce and develop DJs under axenic in vivo and in vitro conditions [11,14] or on non-symbiotic bacteria [15–19], but they are unable to grow on Photorhabdus luminescens[16]. Offspring of Heterorhabditis bacteriophora cannot develop beyond the first juvenile stage in bacteria-free Galleria mellonella[11]. Significant differences exist between heterologous isolates of P. luminescens in their ability to support reproduction of different Heterorhabditis spp. and their transmission by the DJs [20–25]. For example, when H. bacteriophora is fed with symbionts isolated from Heterorhabditis megidis, reproduction is successful, but the symbiont cells are not retained by the resulting DJs. On the other hand, on symbionts of Heterorhabditis indica, nematodes of H. bacteriophora cannot reproduce [25].

Two forms of Xenorhabdus and Photorhabdus spp. with different phenotypes have been characterized as wild-type primary (phase I) and secondary forms (phase II) [26,27]. Several colony variants with intermediate properties have been described [28,29]. Phase I symbionts are isolated from nematode-infected insects. Phase variants are detected earliest at stationary growth phase in cultures which have been inoculated with phase I. Phase I colonies absorb neutral red and bromothymol-blue from agar media, cells carry inclusion bodies, produce antibiotic substances, lipase, phospholipase, protease, pigments and the outer membrane protein OpnB, and, in case of P. luminescens, produce bioluminescence [26,30–35]. Phase II and other variants lose some or all of these characters. Phase variation also influences the success of nematode production. Phase I Xenorhabdus nematophilus cultures produced higher nematode yields in vivo in G. mellonella than phase II [18,26]. However, phase II variants did not negatively affect the in vitro yields of Steinernema spp. on lipid agar plates [18,36]. Yields of Heterorhabditis spp. from in vitro cultures with phase I were significantly higher than from cultures with phase II [18]. Other studies [24,35] reported no yields on phase II P. luminescens. Investigations on the influence of phase variants on nematode reproduction have always used homologous bacterial symbionts to feed the nematodes. This study tests the effect of homo- and heterologous bacterial phase variants on nematode reproduction.

Phase variation is one of the major reasons for process failure in industrial mass production. The reasons why the phase variants negatively impact nematode reproduction have not been identified yet [37]. The involvement of toxins or inhibitors produced by phase II is one possibility, the lack of essential nutrients another. It has been hypothesized that the phase shift is a result of a prolonged culture under in vitro conditions. As such, phase II would be an artefact without any significance for the symbiosis [26,31]. To understand the significance of phase variants in the nematode–bacterium symbiosis, the present study determined the effect of three P. luminescens isolated from different Heterorhabditis spp. on the development and reproduction of H. bacteriophora and Steinernema carpocapsae in bacteria-free instars of the insect G. mellonella and in artificial production systems using either liquid or solid media.

2. Materials and methods

2.1 Bacterial strains and culture conditions

P. luminescens and X. nematophilus strains were obtained from the hemocoel of G. mellonella larvae 48 h after infection with nematode DJs. Their origin and nematode host are listed in Table 1. They were cultured at 25°C and 200 rpm on a rotary shaker in 100-ml flasks filled with 40 ml sterile liquid culture medium (LCM) containing 10 g trypcase soy broth (bioMerieux), 10 g nutrient broth (Difco), 5 g yeast extract, 5 g casein peptone, 0.35 g KCl, 0.21 g CaCl2, 5.0 g NaCl (Merck) and 30 ml vegetable oil (Azco) in 1000 ml tap water as described by Ehlers et al. [7]. Bacterial growth was assessed by measurement of the optical density of 50-fold dilutions at 725 nm in a spectrophotometer. Phase I bacteria were obtained by selecting green or blue-green colonies from NBTA, nutrient agar supplemented with 25 mg l−1 bromothymol blue and 40 mg l−1 2,3,5-triphenyltetrazolium chloride [26] or red colonies from McConkey agar. Stock cultures were maintained in 15% glycerol (v/v) in oil-free LCM at −80°C. To produce phase II variants, phase I cultures of each P. luminescens strain were subcultured at low osmotic conditions (without NaCl in LCM) until phase II characters were obtained [37]. Phase II variants were used only when they expressed phase II characters after subculturing at standard culture conditions. Several characters were used to distinguish phase variants. Dye absorption was determined on NBTA and McConkey agar, colony pigmentation on NA (Standard I Agar, Merck), NA supplemented with 1% (v/v) Tween 20, 60 or 80 (Sigma) and egg yolk agar. Egg yolk was collected aseptically and poured into an equal volume of 0.9% sterile NaCl solution. After homogenization, 10% (v/v) of this solution was added to NA at 45°C. Pigmentation in oil-free LCM and YS broth [38] (containing in g l−1: 5 yeast extract, 5 NaCl, 0.5 NH4H2PO4, 0.5 K2HPO4, 0.2 MgSO4.7H2O) was checked at alkaline and acidic pH by addition of 1 M NaCl or 1 M HCl. Qualitative assays for lipase activity were performed by plating on Tween 20, 60, 80 agar and lecithinase in egg yolk agar. Lipase and lecithinase activity were recognized by opaque halos around the colonies. Agar plate assays were performed in three replicates, which were cultured for 4 days at 25°C before evaluation. The presence of crystalline inclusion proteins in the bacterial cells was checked at 1000-fold magnification with an interference contrast microscope. Tests for the production of antibiotic substances were conducted as described by Akhurst [39] using Bacillus cereus as test organism and scored positive when a growth inhibition zone of >3 mm was measured around the P. luminescens colonies at 96 h after inoculation of the overlay culture. Bioluminescence was observed in the dark after 20 min adaptation of the eyes. Quantitative measurements were performed with 100 μl bacterial culture suspension from oil-free LCM or YS broth in a luminometer (Lumat 9501, Berthold, Germany). The samples were mixed vigorously for 5 s and then measured immediately for 5 s. Bioluminescence intensity is expressed as in vivo relative light units (RLU s−1). Aldehyde-stimulated bioluminescence was measured after the addition of 10 μl of a 0.1% dodecanol solution in EtOH. Catalase activity was tested by introducing 100 μl of a 30% hydrogen peroxide solution into a bacterial colony and observation of the oxygen release.

Table 1.  Strain designation and origin of the bacteria P. luminescens, X. nematophilus and X. poinariia
SpeciesStrainNematodeGeographical originReceived from
  1. aX. poinarii was only used to produce bacteria-free S. carpocapsae.

  2. b W. Wouts, Landcare Research, Mt. Albert, Auckland, New Zealand.

  3. c S. Easwaramoorthy, Sugarcane Breeding Institute, Coimbatore, India.

  4. d Robin Bedding, CSIRO, Canberra, Australia.

  5. e Randy Gaugler, Rutgers University, New Brunswick, USA.

P. luminescensHO6H. bacteriophoraChina
P. luminescensHNAH. megidisOH, USAWoutsb
P. luminescensLN2H. indicaTamil Nadu, IndiaEaswaramoorthyc
X. nematophilusA24S. carpocapsaeRussiaBeddingd
X. poinariiRS92S. glaseriUSAGauglere

2.2 Production of bacteria-free insects and nematode DJs

Bacteria-free DJs of H. bacteriophora (strain HO6) and S. carpocapsae (strain A24) were obtained according to the method described by Han and Ehlers [25]. Briefly, these nematode strains were monoxenically grown on bacteria from H. megidis HNA and Steinernema glaseri (strain RS92) in LCM, respectively. DJs were collected by centrifugation and cloth migration through a 30-μm sieve under sterile conditions and surface-sterilized in 0.5% streptomycin-sulfate (Merck) for 6 h. The sterilization solution was removed by centrifugation, and the DJs rinsed three times in sterile Ringer's solution. The heterologous bacteria provided are able to support nematode reproduction, however, the cells are not retained by the resulting DJs. The success of the surface-sterilization of the DJs was checked according to Han and Ehlers [25] by transfer of the DJs into sterile media or into culture supernatants of P. luminescens, which induce recovery of the DJs and the release of the symbiotic bacterium from inside the DJ [10]. When no bacterial growth was detected in sterile medium or sterile culture supernatant of P. luminescens the DJs were considered to be bacteria-free.

Axenic insect cultures were prepared according to Han and Ehlers [10]. Surface-sterilized eggs of G. mellonella were transferred into a sterile medium containing 22% ground maize, 22% ground wheat, 11% milk powder, 5.5% yeast extract, 17.5% bee wax, 11% honey and 11% glycerol. Insect eggs were surface-sterilized in 1% sodium hypochlorite for 30 min and then rinsed in sterile distilled water before transfer into the medium. The insects were grown axenically at 35°C in closed containers.

2.3 Effects of phase variants on nematode reproduction under in vivo conditions

Under natural conditions entomopathogenic nematode propagate inside of insect larvae. The effect of phase variants was tested in last instars of the Great Waxmoth G. mellonella. Larvae from axenic cultures were injected with 10 μl bacterial samples from YS broth cultures in stationary growth phase (OD725=2.4–3.0) with a micro-syringe. The insects were then transferred to sterile petri dishes (6 cm diameter) filled with sterile filter paper and exposed to approximately 150 bacteria-free and surface-sterilized DJs of H. bacteriophora or S. carpocapsae at 25°C. For each of the P. luminescens phase variants and phase I X. nematophilus and each of the two nematode species, 10 G. mellonella were treated. Controls of 10 larvae each were injected either with 10 μl sterile YS broth, or with 10 μl filter-sterilized supernatants from each of the tested bacterial cultures or with DJs from monoxenic cultures with their homologous symbiont strain. After 3 days 3 and after 10 days another 7 larvae were dissected and the developmental stages of the nematodes were assessed under a microscope according to Johnigk and Ehlers [13].

2.4 Effect of phase variants in liquid media

Industrial scale mass production of entomopathogenic nematodes is done in liquid media in bioreactors [40]. The effect of bacterial phase variants was tested in liquid media (LCM and YS) in cell wells. Tissue culture wells of 16 mm diameter were filled with 0.5 ml 1-day-old bacterial cultures from oil-free LCM or YS. Each well then received approximately 50 axenic DJs in 0.1 ml sterile Ringer's solution. The plates were sealed with Parafilm and incubated at 100 rpm on a rotary shaker at 25°C. Four replicates were established for each bacterial variant and medium and the experiment was repeated twice. The DJ recovery, nematode development and reproduction and mortality were observed over a period of 10 days. At the same time, samples of the cultures were checked for bioluminescence with the luminometer and were plated onto NBTA and McConkey to check for phase characters. S. carpocapsae was combined with phase I X. nematophilus for a positive control.

2.5 Effect of phase variants in solid agar medium

For maintenance of monoxenic nematode inocula, they are cultured on solid media (lipid agar). Bacteria-free DJs of H. bacteriophora and S. carpocapsae were inoculated together with 0.1 ml of a 2-day-old phase variant culture into lipid agar plates (6 cm diameter) containing 1.6% nutrient broth, 0.5% corn oil, and 1.2% agar [41]. Nematode growth and reproduction were checked after 10 days. At the same time samples were streaked on NBTA and McConkey agar to check for phase characters. For each phase variant and nematode species 10 agar plates were established. In an additional experiment lipid agar plates were inoculated with bacteria-free first stage juveniles (J1) instead of DJs. The J1 were produced from sterile nematode eggs following a protocol described by Lunau et al. [14]. Surface-sterilization was done in a solution of 2.5 ml 4 M NaOH, 0.5 ml 12% NaOCl and 21.5 ml distilled water for 10 min and eggs were transferred to YS broth for hatching. Each plate was inoculated with approximately 50 J1. The cultures were treated as described for DJ-inoculated cultures.

2.6 Effect of phase variants in solid sponge medium

When low capital culture systems are superior to the use of bioreactors, for example, in developing countries [8,42], nematode mass production is done in solid state cultures according to Bedding [4]. The effect of phase variants was also tested in the culture system according to Bedding [43]. About 4500 bacteria-free DJs were inoculated into 200-ml flasks each containing 10 g medium consisting of 1% yeast extract, 5% egg yolk, 15% soya flour, 5% corn oil, 8% polyether polyurethane sponge and 50% distilled water [24]. One day prior to nematode inoculation the flasks had received 2 ml of different phase variant cultures in oil-free LCM. Nematode development was checked, and the final yield was determined by gravity-washing nematodes from the sponge cultures after 15 days. Three flasks were established for each treatment. Samples were streaked on NBTA and McConkey agar and checked for phase characters.

2.7 Retention of phase variants by DJs

From Bedding cultures [43] with successful nematode reproduction DJs were checked for retention of bacterial cells in their intestine. Surface-sterilized DJs were individually transferred to a drop of sterile Ringer's solution on a microscopic slide and cut near the anterior end with an oculist's scalpel. After extrusion of the intestine content and treatment with 0.5% crystal violet, the presence of bacteria inside the nematodes was determined [20,25]. Bacterial retention was also examined in a recovery test according to Strauch and Ehlers [10]. Briefly, surface-sterilized DJs were introduced into tissue culture wells (24-well plates) containing 0.5 ml filter-sterilized (0.2 μm filters, Gelman Sciences) supernatant of a P. luminescens culture. This supernatant had been prepared by centrifugation (14 000×g, 10 min, 4°C) from a 48-h culture in YS. The culture plates were sealed with Parafilm and incubated at 100 rpm on a rotary shaker at 25°C. The presence of bacteria released by the DJs during recovery was checked 5 days after inoculation of the DJs by observation of 50-μl samples under the microscope. In addition, approximately 150 surface-sterilized DJs in 0.1 ml sterile Ringer's solution were added to five G. mellonella larvae on sterile filter paper in sterile Petri dishes of 6 cm diameter. The insects were incubated at 25°C and insect mortality was recorded every 3 days. To check for the presence of bacteria in the hemolymph, insects were dissected and drops of hemolymph observed under the microscope.

3. Results

3.1 Identification of phase variants

Phenotypic characteristics of the P. luminescens phase variants used in this study are presented in Table 2. Phase I of the bacterial strains isolated from H. bacteriophora (strain HO6), H. megidis (strain HNA) and H. indica (strain LN2) possess all characteristics which have been described for other P. luminescens isolates [26,31]. They absorb dyes from NBTA and McConkey agar, produce pigments, catalase, antibiotic substances, crystalline inclusion proteins, lecithinase and lipase and bioluminescence. The phase II of all isolates lost these characteristics. The bioluminescence was not visible, even after adaptation of the eyes to the dark for 20 min. The measurable bioluminescence was significantly less than in phase I. Through aldehyde-stimulation the bioluminescence of a phase I culture of P. luminescens (from strain HO6) was increased by a factor of 4.4, whereas a 77.7-fold increase was recorded for phase II.

Table 2.  Characters used to distinguish phase I and II P. luminescens
CharacterPhase IPhase II
  1. a In vivo bioluminescence of P. luminescens isolated from Heterorhabditis spp. (strains HO6, HNA and LN2) is given as increase in RLU s−1 compared to the measurements obtained with phase II.

  2. b The aldehyde-stimulated bioluminescence is the increase in RLU s−1 after addition of 10 μl dodecanol solution.

  3. c Weak response.

Dye absorption
NBTA+
McConkey+
Pigmentation
Liquid media:
LCMred
YSred
Agar media:
Tween 20red
Tween 60orange
Tween 80red/orange
Egg yolkred/orange
Crystalline inclusion protein in cell+
Catalase+wc
Antibiotic activity+
Lecithinase
Egg yolk+
Lipolysis
Tween 20+
Tween 60+
Tween 80+
Bioluminescence
Eye+
In vivoa
HO6+ (800-fold)w
HNA+ (13 000-fold)w
LN2+ (3800-fold)w
Aldehyde-stimulatedb4.4-fold increase77.7-fold increase

3.2 Effect of phase variants on H. bacteriophora development and reproduction

Samples from all cultures streaked on NBTA and McConkey indicated no change of the phase variants during any of the experiments in the different culture systems. Results from controls without bacteria revealed that H. bacteriophora is unable to reproduce in the absence of P. luminescens. When culture supernatants of P. luminescens were injected into bacteria-free G. mellonella, the insects died and H. bacteriophora DJs recovered, but did not develop any further. All insects survived the injection of YS broth and no nematode reproduction was recorded. In bacteria-free G. mellonella, development to adults with eggs was observed in the living insect. After the hatching of the J1 the development ceased.

The effect of different phase variants on the development and reproduction of H. bacteriophora in different culture systems is presented in Table 3. In vivo, phase I P. luminescens isolated from H. bacteriophora and H. megidis supported nematode reproduction of H. bacteriophora, whereas DJs recovered in phase I P. luminescens from H. indica, but did not develop any further. Phase II variants from H. bacteriophora and H. megidis were not able to support reproduction, but DJs recovered and developed to adults with eggs. Phase I X. nematophilus failed to induce recovery of the H. bacteriophora DJs.

Table 3.  Development and offspring production of H. bacteriophora (strain HO6) and S. carpocapsae (strain A24) on phase I and II variants of P. luminescens isolated from H. bacteriophora (strain HO6), H. megidis (strain HNA), H. indica (strain LN2) and phase I X. nematophilus isolated from S. carpocapsae (strain A24) tested in different culture systems
P. luminescens isolate HO6HNALN2A24
Phase variant IIIIIIIIII
Nematode speciesCulture systemaMost advanced developmental nematode stage recordedb
  1. a For media components see Section 2.

  2. b DJ (no development), J3=recovered third juvenile stage (post DJ), J4=fourth juvenile stage, A=adult, A+E=adult with eggs, +=DJ offspring production. For details on the nematode life cycle see Fig. 1.

  3. c Not tested.

H. bacteriophoraIn vivo+A+E+A+EJ3n.t.cDJ
 Liquid (LCM)+A+AJ4ADJ
 Liquid (YS)+A+E+A+EJ4A+EDJ
 Solid (lipid agar)++++J4+DJ
 Solid (Bedding)++++J4+DJ
S. carpocapsaeIn vivoJ3+J3+J3++
 Liquid (LCM)J3AJ3AJ3AA
 Liquid (YS)J3AJ3AJ3AA
 Solid (lipid agar)J3+J3+J3++
 Solid (Bedding)J3+J3+J3++

In liquid media, H. bacteriophora propagated on phase I bacterial cells isolated from H. megidis and H. bacteriophora, and next generation DJs were first recorded after 5 days. Phase I P. luminescens isolated from H. indica could not support reproduction. DJs recovered but did not develop further than to pre-adult J4. On phase II of all P. luminescens strains only few young adults were recorded in LCM and some adults with one or two eggs in YS. In LCM without oil most of the nematodes died in phase II after 5 days, but survived until 10 days after inoculation in phase II cultures with YS.

On lipid agar and also in Bedding cultures, H. bacteriophora produced offspring DJs in phase I and II, except in phase I P. luminescens isolated from H. indica, in which the development ceased when the J4 stage was reached and the nematodes died after 5 days. On lipid agar no differences were recorded whether the plates had been inoculated with bacteria-free DJs or J1. In Bedding cultures H. bacteriophora DJ yields were highest on the phase I of the homologous symbiont followed by phase I of P. luminescens isolated from H. megidis. On phase II symbionts the nematode yields were significantly reduced (Table 4). On phase I P. luminescens from H. indica the nematode development was not beyond the J4 stage whereas on phase II some offspring were recorded.

Table 4.  Influence of phase variants on H. bacteriophora and S. carpocapsae DJ yields in solid state cultures
Bacterial isolateNematodeNematode Yield
  Phase IPhase II
  Yielda±S.D.bDJ (%)cYield±S.D.DJ (%)
  1. a Mean (n=3) DJ yield (in 1000 DJs/flask of 200 ml).

  2. b Standard deviation.

  3. c Percentage DJs of total nematode population.

  4. d Not tested.

P. luminescens HO6H. bacteriophora31314961081392
P. luminescens HNAH. bacteriophora200119123383
P. luminescens LN2H. bacteriophora003252
X. nematophilus A24H. bacteriophora00n.t.d
P. luminescens HO6S. carpocapsae0057938
P. luminescens HNAS. carpocapsae00921637
P. luminescens LN2S. carpocapsae00861734
X. nematophilus A24S. carpocapsae4601694n.t.

3.3 Effect of phase variants on S. carpocapsae development and reproduction

Samples from all cultures streaked on NBTA and McConkey agar indicated no change of the phase variants during any of the experiments in the different culture systems. Results from controls without bacteria revealed that S. carpocapsae is unable to reproduce in vitro in the absence of X. nematophilus, but few DJ offspring were recorded when bacteria-free DJs had been injected into G. mellonella from axenic cultures.

The effects of different phase variants on the development and reproduction of S. carpocapsae in different culture systems are presented in Table 3. S. carpocapsae could not reproduce on any of the three phase I P. luminescens strains, independent on the culture system used. DJs recovered and never developed any further. In P. luminescens isolated from H. bacteriophora the nematodes died after 3 days, in the other bacterial cultures they survived as long as in the liquid media were supplemented with X. nematophilus.

Phase II P. luminescens supported production of offspring DJs, except in liquid media, where the development ceased in the adult stages. The same results were obtained with the homologous symbiont X. nematophilus, which indicates that the liquid media lack an essential factor, necessary for S. carpocapsae to produce offspring. In vivo, next generation DJs were observed 10 days after infection of G. mellonella with S. carpocapsae and phase II P. luminescens and X. nematophilus. The yields obtained from Bedding flasks are summarized in Table 4. From flasks inoculated with X. nematophilus mean DJ yields of 460 000 DJs were counted. On phase II P. luminescens the DJ yields were significantly lower, whereas no offspring were obtained in phase I P. luminescens. Compared to the cultures on the homologous symbiont X. nematophilus with 94% DJs of the overall nematode population, the percentage never surpassed 40% on phase II P. luminescens.

3.4 Bacterial retention

Bacteria were never detected in the surface-sterilized H. bacteriophora DJs from cultures on phase II symbionts. Even when H. bacteriophora had been cultured on phase II of its homologous bacterial symbiont retention of bacterial cells was not recorded. Phase I P. luminescens isolated from H. bacteriophora (strain HO6) was detected, whereas heterologous phase I P. luminescens isolated from H. megidis was not found in DJs of H. bacteriophora. Surface-sterilized S. carpocapsae DJs which had been cultured on phase II P. luminescens did not retain cells of this bacterium.

4. Discussion

One significant character of the symbiosis between entomopathogenic nematodes and their bacteria P. luminescens or Xenorhabdus spp. is the nutritional dependence of the nematodes on their bacterial associates. Although bacteria-free DJs are able to recover and develop to adults inside a bacteria-free insect, S. carpocapsae can only produce very few DJ offspring. In contrast, the development of H. bacteriophora is not beyond the J1 offspring stage in the absence of any bacteria. Only the addition of the symbiotic bacteria supports the completion of the life cycle and a high offspring production of the nematodes [11]. Reproduction of H. bacteriophora on heterologous P. luminescens isolated from H. megidis results in lower nematode yields when compared to yields obtained with its homologous symbiont and the resulting DJs are free of bacteria, the heterologous symbiont cells are not retained. Homologous phase I symbionts are always superior over heterologous phase I bacterial symbionts in their ability to support nematode reproduction. These observations were made with phase I symbiotic bacteria [11].

The present study has included the effect of phase II variants. In vivo, in the presence of phase II homologous symbiont cells, H. bacteriophora did not develop beyond the adult stage carrying eggs, which were unable to develop to J1. This is one developmental step beyond the development observed under axenic conditions, which resulted in J1 hatching from eggs. The negative effect of phase II within the homologous association can thus not be assigned to the lack of essential nutrients, as all necessary nutrients to complete the development to J1 were available in the bacteria-free insect. As toxins are also not involved (Ehlers, unpublished results), it must be speculated that phase II produce metabolites inhibiting the development of H. bacteriophora. As phase II occurs after resources in the insect cadaver have been consumed (stationary phase), the phase II metabolites might function as signal for the nematode to terminate the propagative phase of the life cycle and develop to DJs, which no longer digest their symbiotic bacteria, but retain the cells in the intestine.

The negative effect of phase II P. luminescens on H. bacteriophora development and reproduction was more pronounced in liquid than in solid media. In the latter, nematodes produced at least some DJ offspring on phase II. These results indicate that the composition of the culture media can have an impact on the influence of the phase II on nematode reproduction. The liquid media contained less nutrients compared to the solid media, which is probably the reason for the cessation of the development in the adult stage on phase II.

Earlier studies have shown [25,44] that H. bacteriophora cannot grow on P. luminescens isolated from H. indica. These results now have to be restricted to phase I of the heterologous symbiont isolated from H. indica. On phase II, H. bacteriophora was able to produce a small amount of DJ offspring in solid media, whereas on phase I the development hardly reached the adult stage. Although the mechanism responsible is still unknown, it can be speculated that phase I P. luminescens isolated from H. indica produce substances which inhibit nematode development of H. bacteriophora beyond the recovered J3. Toxins which kill the recovered nematodes might be involved. It is less probable, that P. luminescens isolated from H. indica cannot provide nutrients for nematode reproduction as on phase II reproduction is possible. Likewise is the situation recorded when P. luminescens is fed to S. carpocapsae. The nematodes were unable to develop beyond the recovered J3 stage on phase I, but reproduced in vivo and in solid media on phase II. Again, the nematodes failed to reproduce on phase II in liquid media, but the same was observed on the phase I homologous bacteria X. nematophilus, which indicates that the addition of oil is essential for S. carpocapsae, whereas P. luminescens can substitute for the oil in liquid culture of H. bacteriophora. Culturing S. carpocapsae in LCM supplemented with oil resulted in successful production of DJs (data not shown).

S. carpocapsae has been reported to develop and produce a limited amount of offspring on some non-symbiotic bacteria [15–18,45], but it is unable to develop on phase I P. luminescens. The results indicate that phase II bacteria of P. luminescens provide nutrients which can be used by bacteria-free S. carpocapsae DJs and J1 to develop, as in control cultures on solid media the bacteria-free nematodes did not even recover without the presence of bacteria. This is another indication that substances are produced by phase I P. luminescens, which impair the development of S. carpocapsae and that phase II lose this character during the phase shift.

This study, for the first time, reports that H. bacteriophora DJs, which had been produced on their homologous P. luminescens phase II (strain HO6), cannot retain phase II bacteria. Earlier studies found that phase II bacteria X. nematophilus were retained in the DJs of S. carpocapsae when the axenic nematodes had been injected into G. mellonella containing phase II cells only [26]. We applied several different methods to verify our results and never could phase II bacteria be recovered from the DJs of H. bacteriophora. The observation should be checked also for other Heterorhabditis sp. and for Steinernema spp. using axenically cultured DJs, in order to prove whether phase II cells are not adapted to be retained by the nematode DJ. If so, this would mean, that the retention of bacterial symbiont cells by the DJ is selective for phase I cells, the variant of P. luminescens which provides best conditions for the establishment of the bacto–helminthic complex inside the insect host and for the consecutive nematode reproduction.

Acknowledgements

The fellowship to R.H., Guangdong Entomological Institute, Guangzhou, China, by the Alexander von Humboldt Stiftung and the financial support by the EU FAIR program to the PRONEMA project CT 97-3116 ‘Improvement of a process technology to scale-up liquid cultures of biocontrol nematodes (Heterorhabditis sp.)’ are gratefully acknowledged. Thanks are due to Stefan Johnigk and Olaf Strauch for their help with the figures.

Ancillary