Search

Larval Galleria mellonella (L.) hemocytes form microaggregates in response to stimulation by Gram-positive bacteria. Hemocyte adhesion to foreign materials is mediated by the cAMP/ protein kinase A pathway and the β-subunit of cholera toxin using a cAMP-independent mechanism. Cholera toxin-induced microaggregation was inhibited by the integrin inhibitory RGDS peptide, implying integrins may be part of the mechanism. Based on the types of mammalian integrin-antibody reactive proteins affecting hemocyte adhesion and bacterial-induced responses α 5 , α v , β 1 , and β 3 subunits occurred on both granular cell and plasmatocyte hemocyte subtypes. A fluorescent band representing the binding of rabbit α 5 -integrin subunit antibodies occurred between adhering heterotypic hemocytes. The frequency of the bands was increased by cholera toxin. The α 5 and β 1 rabbit integrin subunit antibodies inhibited removal of Bacillus subtilis (Cohn) from the hemolymph in vivo. A α 5 β 1 -specific synthetic peptide blocker similarly diminished hemocyte function whereas the α v β 3 -specific inhibitory peptide and the corresponding integrin subunit antibodies did not influence nonself hemocyte activities. Western blots revealed several proteins reacting with a given integrin-antibody subtype. Thus integrin-antibody reactive proteins (which may include integrins) with possible α 5 and β 1 epitopes modulate immediate hemocyte function. Confocal microscopy established plasmatocyte adhesion to and rosetting over substrata followed by granular cell microaggregate adhesion to plasmatocytes during early stage nodulation., (© 2018 Institute of Zoology, Chinese Academy of Sciences.)

Xenorhabdus nematophila and Photorhabdus luminescens are entomopathogenic bacterial symbionts that produce toxic proteins that can interfere with the immune system of insects. Herein, we show that outer membrane proteins (OMPs) could be involved as bacterial virulence factors. Purified totals OMPs of both bacterial species were injected into fifth instar larvae of Spodoptera exigua Hübner. Larvae were surveyed for cellular defenses fluctuations in total haemocyte counts (THC) and granulocyte percentage and for the humoral defenses protease, phospholipase A2 (PLA 2 ), and phenoloxidase (PO) activities at specific time intervals. Changes in the expression of the three inducible antimicrobial peptides (AMPs), cecropin, attacin, and spodoptericin, were also measured. Larvae treated with OMPs of both bacterial species had more haemocytes than did the negative controls. OMPs of X. nematophila caused more haemocyte destruction than did the OMPs of P. luminescens . The OMPs of both bacterial species initially activated insect defensive enzymes post-injection, the degree of activation varying with enzyme type. The AMPs, attacin, cecropin, and spodoptericin were up-regulated by OMP injections compared with the normal larvae. The expression of these three AMPs was maximal at four hours post injection (hpi) with P. luminescens OMPs treatment. Expression of the three AMPs in X. nematophila treated insects was irregular and lower than in the P. luminescens OMPs treatment. These findings provide insights into the role of OMPs of entomopathogenic nematode bacterial symbionts in countering the physiological defenses of insects.

Alternative infection models of bacterial pathogenesis are useful because they reproduce some of the disease characteristics observed in higher animals. Insect models are especially useful for modeling bacterial infections, as they are inexpensive, generally less labor-intensive, and more ethically acceptable than experimentation on higher organisms. Similar to animals, insects have been shown to possess innate immune systems that respond to pathogenic bacteria.

Xenorhabdus nematophila is a symbiotic bacterium of the entomopathogenic nematode Steinernema carpocapsae (Weiser). It produces several toxic proteins which interfere with the immune system of insects. The current study shows that purified pilin protein could be a virulence trait of X. nematophila. The fifth instar larvae of Spodoptera exigua (Hübner) was injected with purified pilin. Changes in the cellular defenses in terms of total haemocyte counts and granulocyte percentage and humoral factors including total protease, phospholipase A 2 , and phenoloxidase activities (humoral defense) as well as the expression of the three main antimicrobial peptides attacin, cecropin, and spodoptericin were measured at specific times. The level of THC and granulocytes in larvae with different concentrations of pilin protein were less than the negative control. Also agglutination of haemocytes was observed 8-16h post-injection. The pilin protein activated phenoloxidase in the initial hour post-injection, by 2hpi, activity was stable. The activities of phospholipase A2 and protease activities reached maximum levels at 12 and 4hpi, respectively, and then decreased. The expressions of attacin, cecropin, and spodoptericin in larvae treated with pilin protein were up-regulated above that of the normal sample. The overexpression of cecropin was greater than the other antimicrobial protein mRNA transcripts. The spodoptericin expression had an irregular trend while expressions of attacin and cecropin reached maximum levels at 4hpi and then decreased. Generally, after the injection of pilin protein, the cellular and humoral immune system of S. exigua is activated but this toxin was able to inhibit them. This is the first report of the role of pilin protein when the bacterial symbiont of S. carpocapsae encounters the humoral defense of an insect., (Crown Copyright © 2017. Published by Elsevier Ltd. All rights reserved.)

Nodulation, the lepidopteran insect immune response to large numbers of microbes in the blood (hemolymph) consists of the coordination of the blood cell (hemocyte) types the granular cells and plasmatocytes in terms of granular cell-bacteria adhesion and hemocyte-hemocyte adhesion (microaggregation). Hemocyte-microbe adhesion is influenced by the secondary messenger, cAMP, and cAMP-dependent protein kinase A. In the present study, cholera toxin, an AB5 protein known to indirectly stimulate adenylate cyclase, is used to examine the hemocyte responses to glass, bacteria and hemocyte-hemocyte microaggregates. In vitro, this toxin induces a bimodal hemocyte adhesion response that varies with the holotoxin concentration in terms of the individual and aggregated hemocyte adhesion responses: the lower CTX concentration (1.2 nM) increases microaggregate adhesion and decreases individual hemocyte binding to glass, as does higher concentrations (6-120 nM), however microaggregates induced by lower concentrations do not adhere to glass. Cholera toxin-induced microaggregation is inhibited by RGDS, suggestive of integrin involvement. In vivo, cholera toxin (1.2-120 nM) injected into larvae induces also a bimodal hemocytic response: low levels (1.2-6 nM) cause reduced hemocyte adhesion, while high levels (12-120 nM) increase hemocyte release or mobilization of adhesive hemocyte counts in the hemolymph. Increasing levels of cholera toxin concomitantly injected with the non-pathogenic bacterium, Bacillus subtilis produces a bimodal pattern in bacterial removal from the hemolymph which correlates with nodule frequency in larvae injected with cholera toxin only. The effects of higher concentrations of cholera toxin in vitro (6-120 nM) and in vivo (12-120 nM) are mediated by the B-subunit, whereas the isolated A-subunit has no effect on hemocyte activity. Cholera toxin and its individual subunits did not detectably alter levels of intracellular cAMP in the hemocytes, suggesting a cAMP-independent mechanism stimulating the nodulation response.

The innate non-self response systems of the deciduous tree pest, the forest tent caterpillar, Malacosoma disstria has been documented by us in terms of in vitro and in vivo reactions towards the Gram-positive nonpathogenic bacterium, Bacillus subtilis and Gram-negative pathogenic microbe, Xenorhabdus nematophila and their respective surface antigens, lipopoteichoic acids (LTA) and lipopolysaccharides (LPS). These studies, often conducted in whole and diluted hemolymph, preclude examination of plasma-free cellular (hemocyte) responses. Plasma-free hemocytes as primary cultures are difficult to obtain. The floating cell line Md66 and attached cell line Md108 from M. disstria hemocytes were examined as a model for plasma-free M. disstria hemocyte non-self responses. Herein, it was established that although both lines differed from each other and from the primary hemocyte cultures of M. disstria in growth parameters, cell composition and sizes both cell lines displayed granular cell-like (GL) cells and plasmatocyte-like (PL) cells according to morphological criteria and to some extent antigenic similarities based on labeling with anti-Chrysodeixis includens hemocyte monoclonal antibodies. Hemocyte-specific neuroglian-like protein was detected on cells of both cell lines and in the primary hemocyte cultures albeit with staining patterns differing according to culture and cell types, confluency levels and cell-cell adhesion. Both cell lines bound B. subtilis and X. nematophila, the reaction extent varying with the cell line and its cell types. LPS damaged both cell types in the two cell lines whereas LTA enhanced the adhesion of Md66 GL cells to flask surfaces followed by PL cell adhesion. PL cells of both lines, like the primary cultures, phagocytosed FITC-labeled B. subtilis; only Md108 GL cells phagocytosed B. subtilis. In either case phagocytosis was always less in frequency and intensity than the primary cultures. Proteins released from the cell lines differed in pattern and magnitude but contained bacterial binding proteins that enhanced differential bacterial adhesion to both cell types in both cell lines: the GL cells both cultures, and those of granular cells in primary cultures, were more involved than the primary plasmatocytes and PL cells. Only Md66 cells possessed lysozyme and both cell types of both lines contained phenoloxidase. Neither enzyme type was released during early phase reaction with the bacteria. LPS inhibited phenoloxidase activity. The similarities and differences between the lines and primary cultures make Md66 and Md108 useful for the systematic examination of plasma-free cellular non-self reactions., (Copyright © 2011 Elsevier Inc. All rights reserved.)

Hemocyte encapsulation reactions of infective juveniles of two Iranian isolates of the entomopathogenic nematodes, Heterorhabditis bacteriophora Poinar (Rhabditina: Heterorhabditidae) and Steinernema feltiae Filipjev (Tylenchina: Steinernematidae), were compared in the economic pest Colorado potato beetle, Leptinotarsa decemlineata Say (Coleoptera: Chrysomelidae), and the greater wax moth, Galleria mellonella L. (Lepidoptera: Pyralidae). The former was a more responsive host than the latter and the hemocyte responses occurred sooner and more extensively. Complete encapsulation of some of the nematodes occurred by 4 h post injection for H. bacteriophora in both L. decemlineata and G. mellonella, and by 2 h pi for S. feltiae in L. decemlineata. The percentage of encapsulation from 24 h to 72 h pi in L. decemlineata was 86.2% for S. feltiae and 39% for H. bacteriophora. In G. mellonella there were no encapsulation or melanization responses against S. feltiae, whereas when H. bacteriophora was encapsulated and melanized (16.7%) the encapsulation level was lower than in L. decemlineata. This study may contribute to effectively selecting entomopathogenic nematode species active against significant economic pests based on the latter's cellular immune response.

Invertebrate intracellular hemocyte signaling pathways affecting cellular-antigen responses, although defined for molluscs and some arthropods including dipteran insects, is less known for lepidopterans. Hemocytic-antigen responses of the arboreal pest lepidopteran Malacosoma disstria are linked to cAMP-dependent protein kinase A implicating cAMP in cellular hemocyte immune responses. The purpose in the present study was to determine intracellular cAMP effects on larval M. disstria hemocytes adhering to slides and bacteria. Altering adenylate cyclase and phosphodiesterase activities as well as cAMP levels in vitro and in vivo changed hemocyte responses to antigens. Quiescent hemocytes had high cAMP levels due to adenylate cyclase activity and possibly low phosphodiesterase (type 4) activity. Antigen contact diminished hemocytic cAMP levels. Inhibiting adenylate cyclase increased hemocyte-antigen and hemocyte-hemocyte adhesion, the latter producing nodules in vivo without bacterial antigens. Inhibiting phosphodiesterase type 4 produced the reverse effects. Pharmacologically increasing intracellular cAMP in attached hemocytes caused many of the cells to detach. Diminished intracellular cAMP changed hemograms in vivo in bacteria-free larvae comparable to changes induced by the bacterium, Bacillus subtilis, by producing nodules. Lowering cAMP enhanced also the removal of Xenorhabdus nematophila and B. subtilisin vivo.

Live adult and juvenile entomopathogenic Steinernema carpocapsae DD136 (P. Nematoda) were not subjected to adhesion by haemocytes of lepidopteran insect larvae of Galleria mellonella or Malacosoma disstriain vitro or in vivo. In vitro freeze-killed nematodes exhibited haemocyte attachment, the intensity increasing with time. Accumulation of haemocytes on the dead nematodes was associated with host phenoloxidase activity; live nematodes and their exudates did not activate the enzyme whereas dead nematodes but not their exudate did activate phenoloxidase. Live-nematode exudate inhibited granular cell and some plasmatocyte adhesion to slides, increased granular cell but not plasmatocyte dissociation from preformed haemocyte monolayers and in vivo elevated total haemocyte counts and changed the floating haemocyte types while impairing bacterial removal from the haemolymph. Dead-nematode exudate did not affect these parameters thus immunosuppressant activity by live nematodes may represent the release of inhibitors not associated with their cuticle. The third stage juveniles released the inhibitors.

Previous research established different interactions of the insect pathogen, Xenorhabdus nematophila and nonpathogen, Bacillus subtilis, with antimicrobial hemocytes and humoral factors of larval Malacosoma disstria [Giannoulis, P., Brooks, C.L., Dunphy, G.B., Mandato, C.A., Niven, D.F., Zakarian, R.J., 2007. Interaction of the bacteria Xenorhabdus nematophila (Enterobacteriaceae) and Bacillus subtilis (Bacillaceae) with the hemocytes of larval Malacosoma disstria (Insecta: Lepidoptera: Lasicocampidae). J. Invertebr. Pathol. 94, 20-30]. The antimicrobial systems were inhibited by X. nematophila and stimulated by B. subtilis. The bacterial surface antigens participating in these reactions were unknown. Thus, herein the effects of lipopolysaccharide (endotoxin) from X. nematophila and lipoteichoic acid from B. subtilis on the larval M. disstria immune factors, the hemocytes and phenoloxidase, were determined. Endotoxin elevated the level of damaged hemocytes limiting the removal of X. nematophila from the hemolymph and enhancing the rapid release of bacteria trapped by nodulation. Similar effects were observed with the lipid A moiety of the endotoxin. The effects of lipopolysaccharide and lipid A on the hemocyte activities were abrogated by polymyxin B (an antibiotic that binds to lipid A) confirming lipopolysaccharide as the hemocytotoxin by virtue of the lipid A moiety. Lipoteichoic acid elicited nodulation and enhanced phenoloxidase activation and/or activity. Although lipoidal endotoxin and lipid A inhibited phenoloxidase activation they enhanced the activity of the enzyme. Apolipophorin-III precluded the effects of lipopolysaccharide, lipid A, and lipoteichoic acid on the hemocytes and prophenoloxidase until the antigens exceeded a critical threshold.

Antioxidants, dimethylsulfoxide (DMSO) and dimethylthiourea (DMTU), at concentrations not affecting the viability of blood cells (haemocytes) from the larval stage of 3 lepidopteran insects - Galleria mellonella, Lymantria dispar, and Malacosoma disstria - differed in their influence on the innate binding of haemocytes to glass, bacteria to haemocytes, and on humoral responses to alien materials. In vitro DMSO had little effect, whereas DMTU substantially impaired the adhesion of the haemocyte types, the plasmatocytes and granular cells, to slides as well as the attachment of Bacillus subtilis to these haemocytes. Although both antioxidants increased lysozyme and phenoloxidase activities, there was no correlation of enzyme activity and haemocyte adhesion responses, possibly reflecting sequestered radicals. Nitric oxide and hydroxyl radicals offset the DMTU effect. In the absence of antioxidants, inactivate protein kinases A (PKA) and C (PKC) enhanced haemocyte aggregation. In general, DMSO, as opposed to DMTU, did not alter the effects of PKA and PKC activators and inhibitors on haemocyte aggregation or of PKC and PKA activities. High concentrations of DMSO and all levels of DMTU, although inhibiting PKA and PKC, inhibited haemocyte adhesion to slides. Comparable results occurred for DMTU-treated haemocytes incubated with B. subtilis. In vivo DMSO, unlike DMTU, did not impair plasmatocyte or granular cell responses to foreign materials, including bacterial removal from the haemolymph and nodulation.

Malacosoma disstria larvae are a pest of deciduous trees. Little is known on the interaction of bacteria with the immediate hemocytic antimicrobial responses of these insects. Incubating dead Xenorhabdus nematophila and Bacillus subtilis with a mixture of serum-free granular cells and plasmatocytes in vitro revealed differential bacterial-hemocyte adhesion and differential discharge of lysozyme and phenoloxidase but not total protein. Although active phenoloxidase adhered equally to both bacterial species, X. nematophila limited enzyme activation whereas B. subtilis enhanced activation. Serum with active phenoloxidase (as opposed to tropolone-inhibited phenoloxidase) and purified insect lysozyme increased bacterial-hemocyte adhesion of both bacterial species. An apolipophorin-III-like protein when incubated with hemocytes, limited their responses to glass slides and bacterial adhesion. However, initial binding of the protein to both bacteria increased granular cell levels with bacteria while lowering the plasmatocyte levels with adhering procaryotes. The protein also increased lysozyme and phenoloxidase activities. Although B. subtilis in vivo elicited a nodulation-based decline in total hemocyte counts and did not affect hemocyte viability, dead X. nematophila elevated hemocyte counts and damaged the hemocytes as lipopolysaccharide levels increased and X. nematophila emerged into the hemolymph. Apolipophorin-III-like protein once bound to the bacteria slowed their removal from the hemolymph.

Signal transduction of the innate immediate responses of insect haemocytes to foreign matter is rarely considered. Herein using a combination of adenylate cyclase inhibitors and activators and phosphodiesterase inhibitors we determined that cyclic adenosine monophosphate (cAMP) at high levels normally impairs non-self response. Haemocyte contact with glass and bacteria lowered cAMP in vitro. Inactive phosphodiesterases, including type 4, impaired haemocyte reactions in vitro. Using the drugs in vivo to modulate adenylate cyclase and phosphodiesterases altered the total and types of haemocytes. Adenylate cyclase inhibitors and etazolate (a type 4 phosphodiesterase inhibitor) alone produced changes in the haemograms similar to those caused by Bacillus subtilis. Sequential injections of an enzyme modulator followed by B. subtilis impaired bacterial removal due (1) in the case of enzyme inhibitors, to the removal of haemocytes prior to bacterial challenge and (2) in the case of forskolin and IBMX to the shut-down of the haemocytes. Activating adenylate cyclase or inhibiting phosphodiesterase impaired bacterial removal when co-injecting the compounds and bacteria.

We used the protein kinase A (PKA) specific activator Sp-8-Br-cAMPS and type I inhibitor Rp-8-Br-cAMPS alone and in combination to define the role of PKA in the non-self responses of larval Galleria mellonella haemocytes in vitro and in vivo. Active PKA depressed haemocyte responses whereas PKA inhibition enhanced activities, including bacterial phagocytosis, the number of haemocytes with adherent bacteria, bacterial-induced haemocytic protein release and haemocyte adhesion to slides in vitro, as well as in vivo bacterial removal from the haemolymph. Non-attached haemocytes had more PKA activity than attached haemocytes; therefore, active PKA limited haemocyte response to foreign materials. We found that (i) PKA inhibitor alone induced non-self responses, including haemocyte protein discharge and lowered haemocyte counts in vivo, and induced nodulation; (ii) the enzyme activator produced effects opposite to those of the inhibitor; and (iii) together, the modulators offset each others' effects and influenced haemocyte lysate PKA activity. These findings establish PKA as a mediator of haemocytic non-self responses.

The non-self cellular (hemocytic) responses of Galleria mellonella larvae, including the attachment to slides and the removal of the bacteria Xenorhabdus nematophila and Bacillus subtilis from the hemolymph, were affected by N-formyl peptides. Both N-formyl methionyl-leucyl-phenylalanine (fMLF) and the ester derivative decreased hemocyte adhesion in vitro, and both elevated hemocyte counts and suppressed the removal of both X. nematophila and B. subtilis from the hemolymph in vivo. The amide derivative and the antagonist tertiary-butoxy-carbonyl-methionyl-leucyl-phenylalanine (tBOC) increased hemocyte attachment to glass. The fMLF suppressed protein discharge from monolayers of granular cells with and without bacterial stimulation, while tBOC stimulated protein discharge. The peptide tBOC offset the effects of fMLF in vitro and in vivo. This is the first report implying the existence of formyl peptide receptors on insect hemocytes in which the compounds fMLF and tBOC inhibited and activated hemocyte activity, respectively.

Culture medium affected the virulence of a strain of Candida albicans toward Galleria mellonella larvae, but the yeast growth rates in yeast extract - peptone - dextrose broth and synthetic Galleria serum were not correlated with yeast virulence. Virulent C. albicans grew rapidly in larval serum, whereas, it limited nodulation and continued development in vivo, producing toxins that damaged the hemocytes and fat body. Nonpathogenic yeast-phase cells grew slowly in larval serum but induced extensively melanized nodules in vivo and developed no further. There was no discernible relationship in 14 exo-enzymes between the virulent and avirulent yeast strains and virulence. The avirulent myosin-I-defective yeast cells were rapidly removed from the hemolymph in vivo because of lysozyme-mediated yeast agglutination and the possible binding of the yeast cells by lysozyme and apolipophorin-III. Both lysozyme and apolipophorin-III are proteins that bind beta-1,3-glucan. Finally, insects with nonpathogenic C. albicans exhibited induced immunity and were more resistant to candidiasis from the wild-type yeast cells than were noninduced insects.

Based on the results from the use of selective inhibitors and activators, active protein kinase A, protein tyrosine kinase, and protein kinase C (PKC) isoforms decreased the adhesion of larval Galleria mellonella hemocytes to glass slides. The protein kinase A inhibitor at all concentrations increased granular cell adhesion only whereas protein tyrosine kinase elevated both granular and plasmatocyte attachment at the lowest concentration. Active, Ca(2+)- and lipid-dependent PKC isoforms limited plasmatocyte and granular cell adhesion whereas PKC that was inhibited by selected compounds (with differed modes of PKC inhibition) enhanced hemocyte attachment. The granular cells were more sensitive to the PKC inhibitors than were plasmatocytes. Phospholipase C and its diacylglyceride product were necessary to reduce hemocyte adhesion and maintain PKC activity. Extracellular Ca(2+), possibly transported through L-channels, was required for plasmatocyte attachment. In contrast, lowering the levels of cytosolic Ca(2+) was associated with decreased PKC activity and was required for hemocyte adhesion., (Copyright 2003 Wiley-Liss, Inc.)

Antibacterial proteins are produced in the reproductive tracts of some insect species. The advent of a pupal ovarian cell line of the lepidopteran Galleria mellonella offered an opportunity for exploring the use of ovarian tissue culture to induce antimicrobial proteins in lieu of the larvae. The ovarian cell growth rates and cell yields were maximized by adjusting Grace's medium to pH 6.5, adding 15% (v/v) qualified heat-inactivated fetal calf serum, and lowering the sucrose concentration to 9.3 g/L. Five cell forms and biochemical profiles of the collective cell types were analyzed throughout the culture growth cycle. The final modified culture medium did not affect morphogenesis, whereas it increased the culture growth rate by 50% and the final cell yield threefold. The molting and immunoprotein-inducing hormone, 20-hydroxyecdysone, increased culture growth rate and altered the levels of cell types A and D. Neither 20-hydroxyecdysone nor the larval immunizing agents, apolipophorin-III or Bacillus subtilis, in combination or alone, induced antibacterial activity. The bacterium did induce immunity in both larval and adult stages.

Studies with Galleria mellonella larvae and the iron chelating agent EDDA showed that iron was essential for the removal of dead Xenorhabdas nematophila and Bacillus subtilis from the haemolymph. The delay in removal of the bacteria from the iron-restricted haemolymph was attributed to reduced adhesiveness of the haemocytes and prophenoloxidase activity. Iron augmentation returned these activities to control levels. Whereas dead B. subtilis had no effect on the concentration of ferrozine-detectable iron (henceforth iron) in the haemolymph, dead X. nematophila was associated with substantially lower levels of iron as the number of damaged haemocytes increased. Haemocyte lysate lowered the concentrations of iron in both FeCl(3) solutions and deproteinized larval serum independent of serum lipids. Haemocyte lysate added to tryptic soybroth lowered the level of iron and limited the growth of X. nematophila. X. nematophila limited iron availability in the plasma by releasing lipopolysaccharides; such a mechanism may be a means of impairing the antimicrobial defences of the insects.

Apolipophorin-III (apoLp-III) impaired the adhesion of plasmatocytes and a granular cell-subpopulation of larval Galleria mellonella to glass slides. The protein bound to haemocytes, limited the responses of the plasmatocytes to Bacillus subtilis and increased the percentage of a subgroup of granular cells with adhering bacteria. The total number of bacteria adhering to all the haemocytes on the slides declined. Injections of apoLp-III slowed bacterial removal from the haemolymph without affecting total haemocyte counts and impaired haemocyte attachment to glass slides. Purified apoLp-III bound to B. subtilis. ApoLp-III in serum bound to bacteria within 5 min, peaked at 15 min and was either shed or dissociated by 60 min. ApoLp-III bound to B. subtilis lowered the adhesion of the bacteria to the haemocytes and slowed the removal of the bacteria from the haemolymph.

The level of apolipophorin-III reached a maximum in the haemolymph of Galleria mellonella at the end of the feeding phase of the seventh larval instar and declined to a plateau value in the pupal and the adult stages. Apolipophorin-III was detected immunologically in fat body tissue, haemocyte lysates, and plasma. In its native state, apolipophorin-III may be associated with another protein with an apparent molecular mass of 77 kDa, possibly apolipophorin-II. Injections of octopamine did not cause lipid loading of high density lipophorin., (Copyright 2001 Wiley-Liss, Inc.)

We investigated the effects of lipoteichoic acids, surface components of Gram-positive bacteria, on the hemocytes and phenoloxidase activity in last instar Galleria mellonella larvae, as well as the binding of apolipophorin-III, an insect lipid-binding protein, to lipoteichoic acids. Binding of apolipophorin-III to lipoteichoic acid was studied using an assay based on 1,9-dimethylmethylene blue. Apolipophorin-III bound the lipoteichoic acids from Bacillus subtilis, Enterococcus hirae, and Streptococcus pyogenes and to intact cells of E. hirae. E. hirae lipoteichoic acid promoted the binding of apolipophorin-III to the cells of this species. All lipoteichoic acids tested caused a dose- and time-dependent drop in the total counts of hemocytes and, depending on the species of lipoteichoic acid, partial or complete depletion of plasmatocytes. Granulocyte counts were not affected. Apolipophorin-III prevented partially the loss of plasmatocytes due to B. subtilis lipoteichoic acid. All three lipoteichoic acids studied activated phenoloxidase in vitro; injections of B. subtilis lipoteichoic acid into the larvae elevated the phenoloxidase activity, whereas injections of E. hirae or S. pyogenes lipoteichoic acid, or apolipophorin-III alone, suppressed it. Apolipophorin-III decreased the activation of phenoloxidase by B. subtilis lipoteichoic acid.

Fifth instar Acheta domesticus nymphs exhibited a decline in total hemocyte counts during the first hour of exposure to dead Xenorhabdus nematophilus; the bacterial level in the hemolymph also declined during this time. Thereafter bacterial numbers in the hemolymph increased as the level of damaged hemocytes increased. The bacteria lowered phenoloxidase activity in vivo by initially reducing the number of hemocytes containing prophenoloxidase and later by inhibiting enzyme activation. Preincubating X. nematophilus in hemolymph with active phenoloxidase in vitro accelerated the removal of the bacteria from the hemolymph in vivo which may be due to modification of the bacterial surface by serine proteases. Lysozyme activity increased in bacteria-injected insects in parallel with an increase in counts of damaged hemocytes; most of the enzyme was located in hemocytes. Lipopolysaccharides of X. nematophilus caused changes in hemocyte counts and phenoloxidase and lysozyme levels comparable to whole bacteria. Lipopolysaccharides also slowed the removal rate of the bacteria from, and accelerated bacterial emergence into, the hemolymph.

The hemocytes to which bacteria adhere were defined and the contribution of the prophenoloxidase system of fifth instar nymphs of Acheta domesticus to adhesion were examined. The physicochemical parameters affecting hemocyte and phenoloxidase activity were determined. Both plasmatocytes and granular cells responded to bacteria, the latter cells entrapping the microorganisms on filopodial extensions. The optimum pH for hemocyte adhesion to glass slides was 6.5, the granular cells being the most sensitive hemocyte type. Although hydrophobic resin beads and positively-charged beads favoured hemocyte attachment, these parameters did not contribute to differential bacterial adhesion to hemocytes. Activation of phenoloxidase was neither enhanced nor inhibited by 0.1 and 1 mg/ml of laminarin or zymosan nor by dead Bacillus subtilis. However, live B. subtilis activated the enzyme and dead Xenorhabdus nematophilus inhibited enzyme activation. Serine protease components of the prophenoloxidase system had opsonic properties for B. subtilis but not for X. nematophilus. Phenoloxidase activity was enhanced by Ca(2+) and Mg(2+) and inhibited by SO(2-)(4).

Heat-inactivated serum of non-immune Galleria mellonella larvae enhanced the lytic activity of larval cell-free hemolymph against Micrococcus lysodeikticus. The increase in bacterial lysis was due to a 17.2 kDa protein known previously to bind to bacterial lipopolysaccharides. The protein enhanced the lytic activity of insect cell-free hemolymph and hen lysozyme in vitro and insect hemolymph in vivo. The hydrophobic protein, which adhered to M. lysodeikticus, was identified by its amino acid sequence homology as apolipophorin-III. The titer of apolipophorin-III in 200-250 mg last instar larvae was 8.7 mg/ml of hemolymph. Apolipophorin-III did not bind to lysozyme. A possible mode of action of apolipophorin-III with lysozyme in the insect is proposed.

Studies on the interaction of the insect pathogenic bacterium, Xenorhabdus nematophilus (Enterobacteriaceae), with its nematode and insect hosts would be greatly assisted if a luminescent phenotype were generated that would allow the detection of viable bacteria in vivo without the necessity for disruption of the cellular interactions. The plasmid, pMGM221, containing the luminescence gene (luxCDABE) of Vibrio harveyi was introduced into different strains (DD136 and 19061) and phases (one and two) of X. nematophilus by triparental mating. For reproducible and efficient conjugation, it was necessary to use older cultures (96-160 h) in the stationary phase of X. nematophilus for mating with relatively small differences (<2-fold) in transconjugant yield for the different strains and phases of X. nematophilus. All transconjugants emitted high levels of light with optimum bioluminescence at 27 degrees C in Luria broth at pH 8.0 containing 20 g/L NaCl; pH, osmolarity, and temperature conditions were similar to those encountered by the bacteria in the hemolymph of the larvae of Galleria mellonella. Plasmids were detected in the transconjugants after 6 months of subculturing the bacteria without antibiotic selection. Aside from light emission, luminescent transconjugants had the same physiological properties as the nonluminescent parental strains, including identical rates of growth, production of exoenzymes, removal from and subsequent emergence into the insect's hemolymph, bacterial-induced hemocyte damage, suppression of prophenoloxidase activation, and the ability to kill G. mellonella larvae. Light-emitting larvae could readily be detected by eye in a dark room, and all bacteria reisolated from dead larvae were luminescent. These properties validate the use of luminescent X. nematophilus not only as a means of following bacterial host interactions, but also as a potential agent to follow the infection and death of the insect population.

Subjecting Aedes aegypti (L.) larvae to conditions that induced the production of oviposition repellency also reduced their wet and dry weights and the concentration of total serum carbohydrates, amino acids, and proteins. Thus, infection with metacercariae of the entomopathogenic digenean Plagiorchis elegans (Rudolphi), starvation for 7 d, or crowding for 2 d reduced larval dry weights by as much as 32, 20, and 23%, respectively, and wet weights by 20, 14, and 11%, respectively. Total serum carbohydrates declined by as much as 36, 21, and 29% for infected, starved, and crowded larvae, respectively, amino acids by 39, 48, and 44%, and protein concentrations by 72, 63, and 62%, respectively. Repellency dilution titers were correlated inversely with movement of the mouth parts and gut. Incubation of infected, starved, and crowded larvae in 0.01 g/liter glucose greatly reduced the level of repellency of their waters, whereas adding glucose to repellent waters had only minor effects. Results indicate that the induction of repellency is associated strongly with nutritional depletion effects.

Xenorhabdus nematophilus (three strains), Xenorhabdus bovienii (one strain), and Photorhabdus luminescens (one strain) showed both exo- and endochitinase activity using p-nitrophenyl-N-acetyl-beta-D-glucosaminide and p-nitrophenyl-beta-D-N,N',N'-triacetylchitotriose, respectively, as substrates. One to three bands were detected on PAGE gel with glycol chitin after electrophoresis. Variation in exo- and endochitinase activity among different species and strains was detected with the strongest activity in X. nematophilus and the weakest in P. luminescens. The primary form of X. bovienii had significantly greater chitinase activity than the secondary form, whereas their growth rate and total protein released into culture medium were similar. The partially purified chitinase of X. bovienii showed significant antimycotic activity against conidial germination and germ tube elongation of Botrytis cinerea.

Aerial conidia of Erynia conica were showered onto the wings of susceptible and nonsusceptible black fly, Simulium rostratum and Simulium decorum, respectively. Secondary type 2 conidia as opposed to type 1 conidia and primary conidia invaded the wings of S. rostratum. The invasion process on both sexes of the host species was rapid, with 40% of the conidia germinating during the 2 hr that the wings were showered with conidia. Approximately 24% of total spores developed appressoria close to the parental spore and 79% of the appressoria developed penetration pegs in the cuticle by 9 hr postinoculation. These conidia also germinated (71%) on the abdominal cuticle of S. rostratum by 6 hr, but no appressoria or penetration tubes were noted throughout the duration of the study. On S. decorum, the invasive conidia exhibited delayed germination and neither formed appressoria nor did they invade the wings, but rather produced high levels of tertiary type 1 spores (26%). Cuticular lipids played a role in the infection process by triggering appressorium formation on the host, but did not seem to act as an inhibitory factor(s) on the nonhost.

Infections by insect-pathogenic fungi of the order Entomophthorales in two adult black fly populations were studied in the Réserve Faunique du Saint-Maurice (Quebec). Entomophaga near limoniae infected Simulium verecundum/rostratum in June whereas Erynia curvispora predominantly parasitized Simulium decorum from the 2nd week of July to September and Erynia conica attacked the Simulium venustum complex, Simulium verecundum/rostratum, and the Simulium vittatum complex from May to September. The fungi did not infect the adult stage of Prosimulium species. Possible evidence of host specificity for Erynia conica is discussed.

Xenorhabdus nematophilus mutants that took longer to kill insects than did the wild type were used to determine the relationship of the physicochemical properties and outer membrane composition to bacterial interaction with the antibacterial systems of Galleria mellonella larvae and to bacterial virulence. Insect serum slowed the growth of the wild-type and mutant bacteria. This was attributed to increased spheroplast formation for the mutants. Spheroplast formation was associated with an increased sensitivity to insect lysozyme and a reduction in overall bacterial cationic charge. Increasing bacterial hydrophobicity was correlated with both increased bacterial attachment to the insect's haemocytes and the accelerated removal of the bacteria from the haemolymph. Attachment of the mutants to the insect haemocytes also increased as the bacterial lipopolysaccharide content increased, the level of prophenoloxidase activation increased, and cationic charge declined. Bacterial emergence into the haemolymph occurred in parallel with haemocyte damage but neither the total lipopolysaccharide levels in the bacteria nor the rate of bacterial emergence were associated with virulence. The rate of lipopolysaccharide release into the haemolymph influenced the rate of haemocyte damage. The contribution of outer membrane proteins to lipopolysaccharide release, bacterial adhesion to haemocytes, and virulence is discussed. Virulence reflects bacterial tolerance to the host's antibacterial defences, favouring an increase in bacteria and toxic lipopolysaccharides.

Xenorhabdus nematophilus subsp. dutki, an entomopathogenic bacterium, is vectored by steinernematid nematodes into insects, where it produces broad-spectrum antibiotics. The use of the nematode-bacterium complex against soil-dwelling pest insects could introduce antibiotics into the soil via the dead insect fragments during the emergence phase of the nematodes. Studies on the stability and activities of these antibiotics produced in the insect Galleria mellonella may contribute to assessing the possible impact of antibiotics on soil bacteria. Two isolates of X. nematophilus subsp. dutki (isolates GI and SFU) produced xenocoumacins 1 and 2 in cadavers of G. mellonella larvae in a 1:1 ratio. Total xenocoumacin 1 and 2 production was 800 ng/200 mg (wet weight) of insect tissue for the GI isolate. Antibiotic activity of water extracts from insects that had been infected with X. nematophilus was stable at 60 degrees C for 1 h and after repeated freeze-thaw cycles. The antibiotic titer of extracts held at 27 degrees C declined by day 10. The spectrum of bacterial species killed by antibiotics produced in insect cadavers varied with the isolate of X. nematophilus. Levels of antibiotic activity were greater in vivo than in tryptic soy broth, which may represent a nutrient effect. The bacterial isolate, culture condition, and presence of nematodes influenced the total antibiotic production in vivo. However, the levels of activity were not correlated with bacterial levels in the different growth environments. Insect cadavers with antibiotic activity transiently lowered the numbers of the bacteria in the soil, the extent of decline varying with the strain of X. nematophilus and the time of sampling.

A state of immunity in Galleria mellonella against the pathogen Pseudomonas aeruginosa is known to be induced by the injection of lipopolysaccharide (LPS), isolated from the homologous organism. An in vitro mixture of the LPS and whole or cell-free hemolymph from non-immunized larvae is not antibacterial. In vitro mixtures of fat body and cell-free hemolymph from non-immunized larvae, incubated at 25 degrees C for 20 hours generated a proteinaceous antibacterial activity. The generation of this activity was enhanced by the presence in the incubation mixture of LPS and/or hemocytes from non-immunized larvae. It is suggested that LPS causes the release of a hemocyte factor(s) which acts in conjunction with or directly on the fat body resulting in an enhanced production of antibacterial factors.

The adhesion of strains of Pseudomonas aeruginosa and Proteus mirabilis to the plasmatocytes and granular cells of nonimmune larval Galleria mellonella was influenced by and varied with the type of carbohydrate. Laminarin enhanced prophenoloxidase activation and bacterial adhesion to the hemocytes whereas sucrose suppressed both activities. For all other sugars there was no correlation between bacterial adhesion to the hemocytes and phenoloxidase activity. It is proposed that bacterial adhesion to the hemocytes may be mediated by both lectinlike binding and components of the prophenoloxidase activating system acting like opsonins.

Three Xenorhabdus nematophilus subspecies influenced Steinernema glaseri growth profiles and growth rates, but this was not necessarily because of different bacterial growth rates. Virulence of dauer nematodes in larval Galleria mellonella varied with the number of dauers retaining bacteria and the bacterial subspecies. Virulence was least for dauers grown on X. nematophilus subsp. bovienii because of the lack of retained bacteria. Virulence was subsequently restored by culturing these nematodes on X. nematophilus subsp. poinari.

The major fatty acids (greater than 10%) of Entomophthora egressa were C16:0 and C18:1 . Minor fatty acids, which varied with the stage of fungal development, included C11:0, C12:0, C13:0, C14:0, C15:0, C16;1, C17:0, C18:0, C18:2, C18:3 C:201, C20:2, C20:3, C20:4, C20:5 and two unidentified unsaturated fatty acids. Differences were observed between the total fatty acid levels of C12:0, C14:0, C17:0, C18:0, and C20:5 and the degree of unsaturation of the fatty acids of 37-h protoplasts grown in modified Grace's medium and a simplified growth medium (SGM). The levels of C12:0, C14:0, C18:1, C20:4, and C20:5 decreased and the levels of C18:0 and C20:2 increased with the formation of spherical hyphal body (shb)initials. With the production of mature shb increased levels of C12:0, C14:0, C15:0, C18:1, C20:4, and C20:5 were detected. During the germination of the shb the levels of C14:0, C16:1, C18;1, and C20:4 increased, whereas C15:0 and C20:5 levels declined. The fatty acid levels, except for C12:0, C13:0, and C20:2, remained constant during the mycelial stage. The degree of fatty acid unsaturation decreased during early stages of development (protoplasts through shb initials). In SGM the degree of fatty acid unsaturation was lowest during the shb initial stage and highest during the shb stage. The total lipid level increased during shb maturation and declined during shb germination.

Immune larval Galleria mellonella removed live Proteus mirabilis from the hemolymph less effectively than did non-immune larvae. This was attributed to a decline in total hemocyte counts, levels of plasmatocytes and granulocytes and hemocyte adhesion capacity. Immune serum possessed factors which reduced bacterial adhesion to the hemocytes. This was not due to altering bacterial surfaces but rather to irreversible suppression of hemocyte activity.