Plasmid-located pathogenicity determinants of Serratia entomophila, the causal agent of amber disease of grass grub, show similarity to the insecticidal toxins of Photorhabdus luminescens

Serratia entomophila and Serratia proteamaculans cause amber disease in the grass grub Costelytra zealandica (Coleoptera: Scarabaeidae), an important pasture pest in New Zealand. Larval disease symptoms include cessation of feeding, clearance of the gut, amber coloration, and eventual death. A 115-kb plasmid, pADAP, identified in S. entomophila is required for disease causation and, when introduced into Escherichia coli, enables that organism to cause amber disease. A 23-kb fragment of pADAP that conferred disease-causing ability on E. coli and a pADAP-cured strain of S. entomophila was isolated. Using insertion mutagenesis, the pathogenicity determinants were mapped to a 17-kb region of the clone. Sequence analysis of the 17-kb region showed that the predicted products of three of the open reading frames (sepA, sepB, and sepC) showed significant sequence similarity to components of the insecticidal toxin produced by the bacterium Photorhabdus luminescens. Transposon insertions in sepA, sepB, or sepC completely abolished both gut clearance and cessation of feeding on the 23-kb clone; when recombined back into pADAP, they abolished gut clearance but not cessation of feeding. These results suggest that SepA, SepB, and SepC together are sufficient for amber disease causation by S. entomophila and that another locus also able to exert a cessation-of-feeding effect is encoded elsewhere on pADAP.     

Quantification and kinetics of the decline in grass grub endopeptidase activity during initiation of amber disease

Amber disease in the grass grub (Costelytra zealandica White) (Coleoptera: Scarabaeidae), caused by strains of the bacteria Serratia entomophila or S. proteamaculans, is characterised by cessation of feeding and clearance of the midgut. Analysis of the midgut enzyme activity in diseased grass grub larvae showed that proteolytic activity was reduced to low levels. The endopeptidases, trypsin, elastase, and chymotrypsin, were all markedly reduced in activity whereas the exopeptidases (leucine-aminopeptidase and carboxypeptidase A and B) were much less affected. There was no effect on the non-proteolytic enzymes, esterase and alpha-amylase. Sequential analysis of enzyme levels in the gut during onset of disease showed that proteolytic activity dropped after cessation of feeding and preceded gut clearance. In starved, uninfected larvae enzyme activity levels remained high, indicating that decline in enzyme activity is not associated with absence of food and cessation of feeding, but with the onset of disease.     

Plasmid transfer among several members of the family Enterobacteriaceae increases the number of species capable of causing experimental amber disease in grass grub

Abstract Ability to cause amber disease in the New Zealand grass grub, Costleytra zealandica (Coleoptera: Scarabaeidae), by Serratia entomophila and S. proteamaculans (Enterobacteriaceae), is dependent on the presence of a large plasmid in bacterial strains. Transfer of the plasmid alone to several other Enterobacteriaceae resulted in the ability to cause the disease in grass grub larvae. No species other than S . entomophila or S . proteamaculans has previously been recorded causing amber disease.     

Genes Essential for Amber Disease in Grass Grubs Are Located on the Large Plasmid Found in Serratia entomophila and Serratia proteamaculans

The bacteria Serratia entomophila and S. proteamaculans cause amber disease in the grass grub, Costelytra zealandica (Coleoptera: Scarabaeidae), an important pasture pest in New Zealand. Disease symptoms include rapid cessation of feeding and amber coloration of larvae. A 105-kb plasmid (designated pADAP) has consistently been found only in pathogenic isolates of both species. Investigations into the involvement of pADAP in amber disease have been hindered by the lack of both a selectable marker on the plasmid and a reliable transposon delivery system. Kanamycin-resistant transposon insertions into three cloned HindIII fragments (9.5, 9.6, and 10.6 kb) were isolated and introduced into pADAP by shuttle mutagenesis. Inserts into the 9.5-and 9.6-kb HindIII fragments on pADAP did not alter disease-causing ability. When plasmids with inserts into the 9.6-kb region were conjugated into plasmid-minus, nonpathogenic isolates of S. entomophila and S. proteamaculans, all of them became pathogenic. Transposon insertions into two regions of the 10.6-kb HindIII fragment continued to cause cessation of feeding but failed to produce amber coloration. Further analysis of a mutant from each amber-minus region (pADK-10 and pADK-13) demonstrated that the antifeeding effect was produced only at dosages higher than that of the wild-type strain. Complementation with the wild-type HindIII fragment restored full-blown disease properties for pADK-13, but not for pADK-10.     

Pathobiology of amber disease,caused by Serratia Spp., in the New Zealand grass grub,Costelytra zealandica

Amber disease in the New Zealand grass grub (Costelytra zealandica) is caused by some strains of Serratia entomophila or Serratia proteamaculans (Enterobacteriaceae). When treated with pathogenic isolates, larvae ceased feeding within 48 h, developed an amber coloration after 72 h, and entered a long chronic phase without feeding. An acute dose of 2-4 x 10(4) pathogenic bacteria was sufficient to produce disease in 50% of treated larvae. Time to death was directly related to temperature. At 15 degrees C, infected larvae remained in a chronic, nonfeeding state for more than 4 months prior to death. Nonpathogenic isolates, lacking the disease-causing plasmid (pADAP), had no effect on either feeding or disease. Twenty-four hours after ingestion, bacteria were found predominantly in the hindgut and growth occurred primarily within the fermentation chamber and in the head section of the larvae. Nonpathogenic strains did not multiply in treated larvae. Treatment of diseased larvae with antibiotic eliminated Serratia cells from the insects but did not result in restoration of feeding or the dark gut characteristic of the healthy larva.     

Nucleotide sequence of the Serratia entomophila plasmid pADAP and the Serratia proteamaculans pU143 plasmid virulence associated region

Some strains of Serratia entomophila and S. proteamaculans cause amber disease of the New Zealand grass grub Costelytra zealandica (Coleoptera: Scarabaeidae), an important pasture pest in New Zealand. The disease determinants of S. entomophila, are encoded on a 153,404-bp plasmid, termed pADAP for amber disease associated plasmid. The S. proteamaculans strain 143 (Sp143) exhibits an unusual pathotype, where only 60-70% of C. zealandica larvae infected with the bacterium succumb to disease. DNA sequence analysis of the Sp143 pU143 virulence associated region identified high DNA similarity to the pADAP sep virulence associated region, with DNA sequence variation in the sepA gene and the variable region of the sepC component. No pADAP anti-feeding prophage orthologue was detected in the Sp143 genome. The region of pADAP replication was cloned and found to replicate in S. entomophila but not in Escherichia coli. DNA sequence analysis of the plasmid pSG348 repA gene from the French isolate of Serratia grimesii, identified 93% DNA identity to the pADAP repA gene. A comparison of the pU143 virulence associated region with the completed pADAP nucleotide sequence is given.     

Autochthonous bacterial flora indicated by PCR-DGGE of 16S rRNA gene fragments from the alimentary tract of Costelytra zealandica (Coleoptera: Scarabaeidae)

Aims: To locate and identify putative autochthonous bacteria within the grass grub gut that may have a role in symbiosis. Methods and Results: Polymerase chain reaction-denaturing gradient gel electrophoresis (DGGE) fingerprinting was used to investigate bacterial diversity in the grass grub larval gut. The microbial community profiles from five geographically distinct populations were compared and the influence of feeding was analysed. Bacterial community in the midgut was highly variable between locations and was affected by feeding. The hindgut contained a more diverse but stable bacterial community that was less affected by external conditions. Forty-seven distinct DGGE bands, representing different bacterial genotypes, could be distinguished from all samples, with 34 different bands occurring in the hindgut. The 22 most common bands were isolated and DNA was sequenced. Sequence analysis revealed that most bacteria (16/22) were affiliated to the Clostridiales with the predominant bacteria affiliated to the genus Clostridium. The remaining bacteria were aligned to the Betaproteobacteria, Deltaproteobacteria and Bacteroidetes. Conclusions: The grass grub larva has an autochthonous microflora with predominance of Clostridium spp. in the hindgut. Significance and Impact of the Study: Occurrence of an autocthonous microflora in the grass grub hindgut suggests a symbiotic relationship which could help explain the ability of larval scarabs to feed on recalcitrant organic matter.     

Identification of a Serratia entomophila genetic locus encoding amber disease in New Zealand grass grub (Costelytra zealandica).

Serratia entomophila UC9 (A1MO2), which causes amber disease in the New Zealand grass grub Costelytra zealandica, was subjected to transposon (TnphoA)-induced mutagenesis. A mutant (UC21) was found to be nonpathogenic (Path-) to grass grub larvae in bioassays and was shown, by Southern hybridization, to contain a single TnphoA insertion. This mutant failed to adhere to the gut wall (Adn-) of the larvae and also failed to produce pili (Pil-). A comparative study of the total protein profiles of wild-type S. entomophila UC9 and mutant UC21 revealed that the mutant lacked an approximately 44-kDa protein and overexpressed an approximately 20-kDa protein. Transfer of cosmids containing homologous wild-type sequences into mutant strain UC21 restored wild-type phenotypes (Path+, Pil+, and Adn+). One of the complementing cosmids (pSER107) conferred piliation on Pil- Escherichia coli HB101. The TnphoA insertion in UC21 was mapped within an 8.6-kb BamHI fragment common to the complementing cosmids, and we designated this gene locus amb-1. Six gene products with molecular masses of 44, 36, 34, 33, 20, and 18 kDa were detected in E. coli minicells exclusive to the cloned 8.6-kb fragment (pSER201A). The 44-kDa gene product was not detected in E. coli minicells containing the cloned mutant fragment. Saturation mutagenesis of this fragment produced four unlinked insertional mutations with active fusions to TnphoA. These active fusions disrupted the expression of one or more gene products encoded by amb-1. The 8.6-kb fragment cloned in the opposite orientation (pSER201B) expressed only a 20-kDa protein. We propose that these are the products of structural and/or regulatory genes involved in adhesion and/or piliation which are prerequisites in the S. entomophila-grass grub interaction leading to amber disease.     

Native insects as targets and non-target species of biological control

The use of population models for predicting desirable and undesirable outcomes of biological control are described using three case studies from New Zealand. The first reviews the models of Barlow and colleagues used to understand and improve the control of native grass grub Costelytra zealandica populations by augmentative application of pathogenic Serratia entomophila bacteria. A variety of modelling approaches have been used to predict grass grub outbreaks and thus the cost-effectiveness of applying bacteria. Models have also been developed to understand the competitive interactions between pathogenic and non-pathogenic forms of the bacteria. The other two case studies show how retrospective modelling has been used to quantify the non-target impact of introduced biological control agents. The parasitoid Microctonus aethiopoides was introduced to control the lucerne pest Sitona discoideus, but was found to disperse outside of the target habitat and attack several native weevil species in the Entiminae family. Retrospective modelling suggests that, given average parasitism levels of 15%, native Irenimus spp. and Nicaeana spp. weevil populations are likely to have been reduced by 8% due to non-target parasitism. Similarly, population models have shown that native red admiral butterfly (Bassaris gonerilla) populations are likely to have been reduced by 5% due to non- target parasitism by Pteromalus puparum, which was introduced to New Zealand for the control of the cabbage white butterfly (Pieris rapae).     

Isolation and enumeration of Serratia entomophila—a bacterial pathogen of the New Zealand grass grub, Costelytra zealandica

Several agar media were tested for their use in a selective isolation and identification scheme for Serratia entomophila , a bacterium causing amber disease of the New Zealand grass grub, Costelytra zealandica (White). Soil dilutions were plated on caprylate thallous agar (CTA), selective for Serratia spp. Most strains of Ser. entomophila grew well on CTA; the mean efficiency of colony formation on CTA was 94 ± 3% of that on a non-selective medium. The identity of colonies growing on CTA was determined on the basis of their growth reactions on DNase-toluidine blue agar, adonitol agar and itaconate agar. Serratia entomophila could be distinguished from other Serratia spp. found in New Zealand soils, in particular Ser. proteamaculans , another causal agent of amber disease of grass grub. The identification scheme allowed the selective recovery of Ser. entomophila from field soils containing a diverse microflora.     

Cloning Serratia entomophila antifeeding genes--a putative defective prophage active against the grass grub Costelytra zealandica

Serratia entomophila and Serratia proteamaculans (Enterobacteriaceae) cause amber disease in the grass grub Costelytra zealandica (Coleoptera: Scarabaeidae), an important pasture pest in New Zealand. Larval disease symptoms include cessation of feeding, clearance of the gut, amber coloration, and eventual death. A 155-kb plasmid, pADAP, carries the genes sepA, sepB, and sepC, which are essential for production of amber disease symptoms. Transposon insertions in any of the sep genes in pADAP abolish gut clearance but not cessation of feeding, indicating the presence of an antifeeding gene(s) elsewhere on pADAP. Based on deletion analysis of pADAP and subsequent sequence data, a 47-kb clone was constructed, which when placed in either an Escherichia coli or a Serratia background exerted strong antifeeding activity and often led to rapid death of the infected grass grub larvae. Sequence data show that the antifeeding component is part of a large gene cluster that may form a defective prophage and that six potential members of this prophage are present in Photorhabdus luminescens subsp. laumondii TTO1, a species which also has sep gene homologues.     

The lipopolysaccharide biosynthesis core of the Mexican pathogenic strain Serratia entomophila is associated with toxicity to larvae of Phyllophaga blanchardi

The Mexican bacteria Serratia entomophila strain Mor4.1 (Enterobacteriaceae) is pathogenic to coleopteran species of the Phyllophaga genus, which are considered important soil-dwelling pests. Mor4.1 causes anti-feeding activity and mortality to larvae after oral and injection bio-assay either, by bacteria or by cell free culture broth inoculation. The pathogenic determinants of Mor4.1 have not been elucidated. We hypothesize that Mor4.1 produces several toxins and other virulence factors, some acting at the level of the insect gut and others at the hemocoel. To identify and characterize virulence factors, a fosmid library of S. entomophila Mor4.1 was made in Escherichia coli. Five different insecticidal clones were isolated by injecting individual clones into Phyllophaga blanchardi larvae. The complete 40 kb DNA sequence and gene organization of clone G8 was determined. By comparative genomics, 21 genes were associated with virulence. By transposon (Tn5) insertion mutagenesis of G8 and further bio-assays we show that a dUTPase, a flavoprotein and a heptosyltransferase III, are key factors for G8 toxic activity. The heptosyltransferase III, is part of the lipopolysaccharide (LPS) biosynthesis core. We demonstrated that purified LPS from G8 and Mor4.1 are toxic to P. blanchardi larvae by injection bio-assay.     

Preliminary characterization of bacteriophages of Serratia entomophila

C.R. WILSON, T.A. JACKSON AND H.K. MAHANTY. 1993. Bacteriophage was found for the first time associated with the bacterium Serratia entomophila, a pathogen of the New Zealand grass grub ( Costelytra zealandica ). Phage was isolated from the homogenized guts of field-collected grass grubs from sites throughout New Zealand. The main phage type, φCW1, produced opaque plaques in sensitive bacterial lawn and had a lamboid structure consisting of an icosahedral head (55 nm) and a long non-contractile tail (175 times 17 nm) with a bar across the base of the tail. Nucleic acid from φCW1 was digested to nucleotides by DNAse, suggesting double stranded DNA. On further examination of the homogenates, five phage types, φCW1-5, could be distinguished on the basis of plaque morphology. Bacterial host range was determined by testing against a selection of Serratia spp. and other bacteria. All five phage types lysed Ser. entomophila only. Differences in susceptibility to the phage types were found within this species. Lysogeny was demonstrated in φCW1 by immunity to superinfection and induction of free bacteriophage from suspected lysogens. A restriction map for φCW1 was determined with Bam HI, Eco RI and Hind III and a postulated origin of replication (ori) and cohesive site (cos) was suggested. The possible implications of bacteriophage on the use of Ser. entomophila as a biological control agent are discussed.     

Expression of the antifeeding gene anfA1 in Serratia entomophila requires rpoS

The rpoS gene of Serratia entomophila BC4B was cloned and used to create rpoS-mutant strain BC4BRS. Larvae of the New Zealand grass grub Costelytra zealandica infected with BC4BRS became amber colored but continued to feed, albeit to a lesser extent than infected larvae. Subsequently, we found that expression of the antifeeding gene anfA1 in trans was substantially reduced in BC4BRS relative to that in the parental strain BC4B. Our data show that a functional rpoS gene is vital for full expression of anfA1 and for development of the antifeeding component of amber disease.     

Selection,development and testing of phage‐resistant strains of Serratia entomophila for grass grub control

Use of the bacterium Serratia entomophila as an inundative biological control agent for the New Zealand grass grub (Costelytra zealandica) depends on the consistent production of high yields of the bacterium in liquid fermentation. Following the phage related failure of several S. entomophila fermentations, a programme was initiated to isolate phage‐resistant strains. No naturally occurring strains were found to be resistant to eight stocks of phage isolated from either grass grub larvae or the fermenter. Therefore, ethylmethane sulphonate was used to generate phage‐resistant mutants. Strains which showed cross resistance to all eight phage test stocks were tested for pathogenicity towards grass grub. Several strains showing unimpaired pathogenicity were selected for further tests. Four of these phage‐resistant strains produced high cell yields, even when grown in the presence of high numbers of fermenter‐derived phage. Phage‐resistant strains have subsequently been produced in bulk in industrial fermenters, for use in large‐scale field trials throughout New Zealand.     

Stage‐dependent synergism using Metarhizium anisopliae and Serratia entomophila against Costelytra zealandica

Larvae of the New Zealand grass grub (Costelytra zealandica) were treated with the entomopathogenic fungus, Metarhizium anisopliae, and the bacterium, Serratia entomophila, singly and in combination. The mortality of second instar larvae up to 41 days in treatments with both pathogens together was significantly greater than the additive mortalities of single pathogen treatments, and therefore synergistic. Treatment of third instar larvae with both pathogens did not increase mortality compared with the fungus alone. Second instar larvae were more resistant to M. anisopliae than third instar larvae. S. entomophila causes a chronic disease and bacterial treatments alone resulted in disease, but little mortality for either instar within 5 weeks. In both fungus alone and fungus/bacteria treatments, less than half the cadavers supported fungal sporulation. The use of a dual pathogen system for control of grass grub larvae is discussed.     

Investigation of a phage resistantSerratia entomophila strain (BC4B), establishment of generalised transduction and construction ofS. entomophila RecA mutants

ArecA clone was isolated from a cosmid library ofSerratia entomophila constructed in theEscherichia coli strain HB101. Subcloning and transposon mutagenesis were used to identify a 1.36 kb fragment containing therecA gene. A clonedrecA mutation, generated by transposon mutagenesis and the replacement of a portion of therecA gene with an antibiotic resistance cassette, was introduced into the chromosome via a marker exchange technique. TherecA strains created were deficient in DNA repair, homologous recombination and both the spontaneous and UV induction of prophages.S. entomophila recA strains showed continued pathogenicity towards the New Zealand grass grub,Costelytra zealandica. Simple procedures for further construction ofS. entomophila recA strains have been demonstrated.     

Investigation of a phage resistantSerratia entomophila strain (BC4B), establishment of generalised transduction and construction ofS. entomophila RecA mutants

ArecA clone was isolated from a cosmid library ofSerratia entomophila constructed in theEscherichia coli strain HB101. Subcloning and transposon mutagenesis were used to identify a 1.36 kb fragment containing therecA gene. A clonedrecA mutation, generated by transposon mutagenesis and the replacement of a portion of therecA gene with an antibiotic resistance cassette, was introduced into the chromosome via a marker exchange technique. TherecA strains created were deficient in DNA repair, homologous recombination and both the spontaneous and UV induction of prophages.S. entomophila recA strains showed continued pathogenicity towards the New Zealand grass grub,Costelytra zealandica. Simple procedures for further construction ofS. entomophila recA strains have been demonstrated.     

Differences in the virulence of Heterorhabditis bacteriophora and Steinernema scarabaei to three white grub species: The relative contribution of the nematodes and their symbiotic bacteria

Because susceptibility of white grub species to entomopathogenic nematodes differs, we compared the virulence of Photorhabdus temperata and Xenorhabdus koppenhoeferi, the symbiotic bacteria of the nematodes Heterorhabditis bacteriophora and Steinernema scarabaei, respectively, to the three white grub species, Popillia japonica, Rhizotrogus majalis, and Cyclocephala borealis. Both bacteria were pathogenic to all three grub species even at 2 cells/grub. However, the median lethal dose at 48 h post injection and median lethal time at 20 cells/grub showed that P. temperata was more virulent than X. koppenhoeferi to C. borealis. Although H. bacteriophora is less pathogenic than S. scarabaei to R. majalis and P. japonica, their symbiotic bacteria did not differ in virulence against these two grub species, and they also showed similar growth patterns both in vitro and inside R. majalis larvae at 20 °C. We then tested the pathogenicity of oral- and intrahemocoel-introduced H. bacteriophora to R. majalis to determine whether nematodes are able to successfully vector the bacteria into the hemolymph. Hemocoel injected H. bacteriophora was pathogenic to R. majalis indicating successful bacterial release, but orally introduced H. bacteriophora were not. Dissection of grubs confirmed that the orally introduced H. bacteriophora were unable to penetrate into the hemolymph through the gut wall. We conclude that the low susceptibility of R. majalis to H. bacteriophora is not due to the symbiotic bacteria but rather to the nematode’s poor ability to penetrate through the gut wall and the cuticle to vector the bacteria into the hemolymph.     

Differences in penetration routes and establishment rates of four entomopathogenic nematode species into four white grub species

We compared the penetration of the entomopathogenic nematodes Steinernema scarabaei (AMK001 strain), S. glaseri (NC1 strain), Heterorhabditis zealandica (X1 strain), and H. bacteriophora (GPS11 strain) into third-instars of the scarabs Popillia japonica, Anomala orientalis, Cyclocephala borealis, and Rhizotrogus majalis. When larvae were exposed to nematodes for 6-72 h larval mortality and nematode establishment rate and occasionally speed of kill often showed the same pattern within nematode-white grub combinations. But no two nematodes or white grub species had the same pattern for these observations for all white grub or nematode species, respectively. Mortality, establishment, and speed of kill followed a similar pattern for H. zealandica, S. glaseri, and S. scarabaei, but there was no clear relationship for H. bacteriophora. Significant nematode establishment was only observed after at least 48 h exposure in most nematode-white grub combinations. Faster establishment was observed only for H. zealandica in A. orientalis and R. majalis (after 24 h) and for S. scarabaei in P. japonica and R. majalis (after 12 h). Nematode establishment after 72 h in the different scarab species was generally low for S. glaseri (<1.5%) and H. bacteriophora (<3%), higher for H. zealandica (2-5%), and the highest for S. scarabaei (1-14%). However, in another experiment establishment was generally higher after 96h exposure. Nematode penetration sites were determined by comparing nematode establishment in larvae with mouth, anus, mouth+anus, or none sealed with glue. The trends for each nematode species were very similar in the different white grub species. H. zealandica and H. bacteriophora showed excellent cuticular penetration ability but may also penetrate through mouth and/or anus. S. glaseri also penetrated through the cuticle but lower establishment in larvae with mouth or mouth+anus sealed suggested that the mouth is an important penetration site. S. scarabaei showed a preference for the mouth as a penetration site, but it showed some cuticular penetration ability and may also use the anus as a penetration site. The methodology used cannot exclude that cuticular penetration also included penetration through the spiracles. To fully understand the effect of nematode and white grub species on nematode virulence, future studies will have to compare host immune response to the penetrating IJs and the role of the symbiotic bacteria in these interactions.