SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Invertebrates, including insects, are being developed as model systems for the study of bacterial virulence. However, we understand little of the interaction between bacteria and specific invertebrate tissues or the immune system. To establish an infection model for Photorhabdus, which is released directly into the insect blood system by its nematode symbiont, we document the number and location of recoverable bacteria found during infection of Manduca sexta. After injection into the insect larva, P. luminescens multiplies in both the midgut and haemolymph, only later colonizing the fat body and the remaining tissues of the cadaver. Bacteria persist by suppressing haemocyte-mediated phagocytosis and culture supernatants grown in vitro, as well as plasma from infected insects, suppress phagocytosis of P. luminescens. Using GFP-labelled bacteria, we show that colonization of the gut begins at the anterior of the midgut and proceeds posteriorly. Within the midgut, P. luminescens occupies a specific niche between the extracellular matrix and basal membrane (lamina) of the folded midgut epithelium. Here, the bacteria express the gut-active Toxin complex A (Tca) and an RTX-like metalloprotease PrtA. This close association of the bacteria with the gut, and the production of toxins and protease, triggers a massive programmed cell death of the midgut epithelium.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Several invertebrate groups, including nematodes (Aballay and Ausubel, 2001; Gallagher and Manoil, 2001; Garsin et al., 2001; O´Quinn et al., 2001) and insects (Daborn et al., 2001; De Gregorio et al., 2001; Wu et al., 2001), are being developed as model systems for the study of bacterial virulence. Invertebrate models have numerous advantages for the study of infection including, low cost, genetic and physiological malleability and fewer ethical considerations than the use of mammalian models. Further, recent genome-wide analysis of the Drosophila immune response to bacteria (De Gregorio et al., 2001) will facilitate analysis of genomic immune responses both to a range of bacterial species and also to their specific virulence factors. Although the phenotypes assayed within these model invertebrate systems, such as gut colonization (Garsin et  al., 2001) or death (Gallagher and Manoil, 2001; O´Quinn et al., 2001), are visually clear, the physiological or molecular basis of these interactions remains obscure. For example, we have little idea of the specific mechanisms relating to colonization of invertebrate tissues or of the toxicological mechanisms whereby the invertebrate host is affected (O´Quinn et al., 2001). Also poorly understood is how infecting bacteria either overcome or avoid the peptide-mediated and cellular components of the invertebrate immune response (De Gregorio et  al., 2001).

We are interested in using a model insect, the tobacco hornworm Manduca sexta, to describe in detail the pattern of infection of an insect by an insect pathogenic bacterium, Photorhabdus luminescens. Manduca sexta has been used extensively as a model for insect physiology and, although it lacks the genetics of Drosophila, the large size of Manduca larvae facilitates a ready dissection of insect bacterial infection. Photorhabdus luminescens has both symbiotic and pathogenic stages of its lifecycle (Forst and Nealson, 1996; Forst et al., 1997; Forst and Clarke, 2001) and therefore itself forms a useful model in which to study genes controlling pathogenicity, symbiosis and the switch between these two states. Photorhabdus luminescens is also a member of the Enterobacteriaceae, facilitating genomic comparisons of putative virulence or symbiosis factors with well-studied bacteria such as Esherichia coli (ffrench-Constant et al., 2000). Photorhabdus luminescens lives in a symbiosis of pathogens with nematodes that invade insects, termed entomopathogenic nematodes (Forst and Nealson, 1996; Forst et al., 1997; Forst and Clarke, 2001). In the ‘symbiotic’ phase of its lifecycle, P. luminescens persists in the gut of entomopathogenic nematodes of the family Heterorhabditidae. After invasion of an insect host by an infective juvenile nematode, 30–200 bacteria (T. Ciche and J. Ensign, unpublished communication) are released directly into the open blood system (haemocoel) of the insect. Within the insect host the bacteria exhibit a ‘pathogenic’ phenotype and are postulated to produce a variety of virulence factors including, the high molecular weight Toxin complexes (Tc), proteases, lipases and lipopolysaccharide (Forst and Nealson, 1996; Forst et al., 1997; ffrench-Constant et al., 2000; Daborn et al., 2001).

Strikingly, P. luminescens bacteria reproduce rapidly within the insect (Daborn et al., 2001) with apparent disregard for both the peptide (antibacterial peptides) and cellular (haemocyte) mediated aspects of the insect immune response. After death of the insect host, the bacteria and nematodes then undergo sequential rounds of growth and reproduction, until new infective juvenile larvae reacquire the bacteria and leave the insect cadaver (Forst and Nealson, 1996; Forst et al., 1997). Although the relative growth rate of P. luminescens subsp. akhurstii strain W14 within the model insect M. sexta has previously been documented (Daborn et al., 2001), we know little about which tissues the bacteria colonize within the insect and how they evade the immune system. Specifically, it is not known if P. luminescens occupies certain tissues preferentially or if the bacteria multiply at different rates within these different tissues. Further, we do not understand how P. luminescens avoids either phagocytosis or encapsulation by insect haemocytes. Finally, we do not know within which tissues of the insect host specific virulence factors are expressed.

In this study, we document the rate of growth of P. luminescens subsp. akhurstii strain W14 within different insect tissues after artificial infection of larval M. sexta. We show that P. luminescens grows most rapidly in the haemocoel and midgut, and then subsequently colonizes the fat body and remainder of the cadaver. During infection P. luminescens W14 secretes an anti-phagocytic agent that inhibits its own phagocytosis and can be recovered from the plasma of infected insects. Midgut colonization is associated with occupation of a specific niche between the basal lamina and the extracellular matrix that surrounds the midgut epithelium itself. Within this niche the bacteria express both the gut active Toxin complex A (Tca) and also an RTX-like metalloprotease, PrtA, while the gut epithelium undergoes massive programmed cell death. This study not only describes the normal infection process for an insect pathogen, but also defines an infection model within which the infection of heterologous or recombinant bacteria can be documented.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Growth of P. luminescens in different insect tissues

We examined the relative growth rates of P. luminescens in different insect tissues by measuring the number of recoverable colony-forming units (CFU) per unit weight of tissues dissected from injected hosts over time. After injection of 70–100 bacterial cells, the number of bacteria recoverable from either the haemolymph or the midgut increases steadily over time (Fig. 1A and B), with 108 CFU being recoverable per gram of infected tissue 72 h post injection. Within these tissues, the number of recoverable bacteria increased by six orders of magnitude within 48 h (24–72 h post infection). In contrast, although the final numbers of recoverable bacteria per gram are similar at 72 h (Fig. 1B), there is a delay in colonization of, and multiplication within, the fat body. Thus it is not possible to recover more than 103 CFU per gram from the fat body until 42 h post infection (Fig. 1C). Similarly P. luminescens W14 multiply slower in the remaining tissues of the insect cadaver and also yield a somewhat lower number (107 CFU per gram) of recoverable bacteria (Fig. 1D).

image

Figure 1.  Numbers of bacteria recoverable from different Manduca tissues over time (h) after injection of Photorhabdus. A, midgut; B, haemolymph; C, fat body and D, remaining carcass. Bacterial cells (50–70 per larva) were injected into a cohort of M. sexta. Larvae were then dissected over time at fixed times post infection. Results are expressed as the mean number (±s.e. of at least three individual larvae) of colony-forming units (CFU) recoverable per gram of wet weight of insect tissue or haemolymph. Note that before 42 h most of the recoverable bacteria are found within the midgut and haemolymph. The line at 1.E + 08 is drawn to facilitate comparisons between the number of bacteria found in the different tissues over time.

Download figure to PowerPoint

These numerical data are supported by visual studies of infection with GFP-labelled P. luminescens W14. Here, the same number of GFP-labelled bacteria were injected into the insect host and then specific tissues were examined over time with a binocular microscope capable of detecting GFP. Using this approach, detectable colonization of the anterior midgut appears at 30 h (Fig. 2A), at which time 106 CFU per gram are recoverable from this tissue (Fig. 1A). Subsequently infection spreads posteriorly along the midgut, colonizing its entire length by 52 h (Fig. 2A). After occupation of the entire gut (after 52 h), GFP expressing bacteria subsequently colonize the fat body. Finally, by 72 h post infection, most tissues within the cadaver appear infected (Fig. 2A). Detailed examination of the pattern of infection of anterior midgut shows that the bacteria are arranged linearly along ‘grooves’ across the midgut surface and along the longitudinal muscles which run down the midline of the gut (Fig. 2B). Higher magnification of these clusters of GFP signal confirms that they correspond to accumulations of individual GFP-expressing bacteria (Fig. 2C).

image

Figure 2.  Colonization of the infection model Manduca by GFP-expressing Photorhabdus.

A. Larvae dissected at different times (Con for control, then 30, 48, 52, and 72 h) post infection to show colonization of the midgut. Note that initial colonization of the midgut (indicated by the length of the white bar) occurs at the anterior midgut (AM) and then proceeds to the posterior midgut (PM) until the gut is completely colonized by 52 h post infection. Later, at 72 h, colonization of other tissues in the cadaver, such as the fatbody (Fb) is also visible.

B. Detail of the anterior midgut showing the pattern of colonizing bacteria within the anterior midgut. Note the longitudinal grooves of bacteria and the highlighting of the longitudinal muscle (LM) which runs down the length of the midgut (arrows).

C. Detail of a colonized groove at higher magnification showing the individual GFP-positive bacteria (arrow).

Download figure to PowerPoint

Viability and phagocytic competence of insect haemocytes

To understand how P. luminescens multiply rapidly within the haemocoel in the presence of insect phagocytes, we examined the effect of bacterial cultures on the ability of haemocytes to phagocytose either FITC-labelled E. coli or P. luminescens W14 itself. Phagocytosis assays using FITC-labelled P. luminescens showed that living W14 bacteria can inhibit their own phagocytosis by insect haemocytes (Fig. 3A). In similar assays, both living and heat-killed FITC-labelled E. coli were phagocytosed. Because heat-killed P. luminescens were phagocytosed (Fig. 3A), it seems likely that P. luminescens W14 actively produces a phagocytosis-inhibiting factor, rather than not being recognized by the insect’s cellular immune system. To test this hypothesis, we used FITC-labelled E. coli to measure phagocytosis in the presence of W14 supernatant or W14 infected insect plasma. The antiphagocytic factor can be found in cell-free filtered W14 supernatants and shows a dose-dependent decline in inhibition upon filtrate dilution (Fig. 3B). During P. luminescens growth in vitro, the time course of production of this antiphagocytic factor mimics bacterial growth itself (Fig. 3C). Thus, significant inhibition is observed from cultures taken as the growing P. luminescens enter exponential growth and percentage phagocytosis then declines in a reciprocal pattern to bacterial growth (Fig. 3C). Finally, cell-free plasma taken from W14 infected insects also has the ability to inhibit haemocyte-mediated phagocytosis (Fig. 3D) showing that the antiphagocytic factor is produced during an actual infection.

image

Figure 3. Photorhabdus luminescens W14 inhibits its own phagocytosis by Manduca haemocytes.

A. Percentage phagocytosis of E. coli or P. luminescens W14, either alive or heat killed, by haemocyte monolayers. Note that living W14 bacteria can inhibit their own phagocytosis, an effect that is reduced by heat treatment of the bacteria.

B. The filtrate (cell-free supernatant) of P. luminescens strain W14 cultures inhibits phagocytosis in a dose (concentration)-dependent manner.

C. The antiphagocytic factor is produced by W14 in vitro as the bacterial culture enters exponential growth phase.

D. The antiphagocytic factor is also detectable in vivo, as cell-free plasma taken from Manduca infected with W14 also inhibits phagocytosis in the haemocyte monolayer assay (see Experimental procedures).

Download figure to PowerPoint

Scanning and transmission electron microscopy of the infected midgut

To localize the gut niche colonized by the invading bacteria, we studied the morphology of uninfected and infected midguts via scanning and transmission electron microscopy (SEM and TEM). The midgut consists of a highly folded epithelium held together by a series of both circular and longitudinal muscles (Fig. 4A and B ). The midgut epithelium itself contains both columnar and goblet cells and is itself encased on the haemocoel side by a basement membrane (Cioffi, 1979) and a series of thick layers of connective tissue, termed the extracellular matrix (Lane et al., 1996). The Malpighian tubules, which perform an equivalent physiological function to the vertebrate kidneys, lie alongside the midgut (Fig. 4A), whereas the midgut itself is innervated by air carrying trachea which carry oxygen to the gut tissues (Fig. 4B).

image

Figure 4.  Examination of the infected Manduca midgut by electron microscopy.

A. Uninfected midgut viewed by scanning electron microscopy (SEM) to show how the folded midgut epithelium is bundled together by muscles, including the dorsal muscle (DM). Note the presence of the Malpighian tubules (Mt) which lie alongside the midgut and the tracheae (T) which invade the midgut tissues to bring oxygen.

B. Detail of uninfected midgut to show how the longitudinal muscles (LM) hold the folds of gut epithelium together.

C. View of an infected midgut 48 h post infection by SEM, showing the presence of numerous bacteria underneath the surrounding extracellular matrix (EM). Individual bacteria can be seen lying alongside each other at high density where the EM has ruptured (black arrows) and the outlines of numerous bacteria can also be seen beneath the EM itself (white arrow).

D. Detail of invading bacteria underneath the EM and alongside a longitudinal muscle (LM).

E. Transmission electron microscopy (TEM) of an infected section of insect midgut. Note the presence of bacteria underneath the EM in close proximity to the cells of the midgut epithelium itself.

F. Detail of the association between the invading bacteria and the cells of the midgut epithelium. Despite the close proximity of the bacteria to the insect cells, no obvious cellular rearrangement of the gut cells is observed, although the cells themselves are often detached from the basal lamina.

Scale bars: A, 500 μm; B, 50 μm; C, 5 μm and D-F, 2 μm.

Download figure to PowerPoint

After invasion of the midgut by P. luminescens at 48 h post infection bacteria are clearly visible beneath the extracellular matrix (Fig. 4C). These colonizing bacteria appear to aggregate around the muscles surrounding the midgut and can be seen underneath the fragmented matrix where it adjoins longitudinal muscle (Fig. 4D). In order to examine the closeness of the association of the bacteria with the midgut epithelium, we used transmission electron microscopy (TEM) to look at the gut in transverse section. In these sections (Fig. 4E and F), the bacteria are in close association with the cells of the epithelium and sandwiched between the epithelium and the sheathing extracellular matrix.

Histopathology of the infected midgut, toxin expression and apoptosis

To relate this localization of gut associated bacteria with gut histopathology, we examined stained sections of infected midgut via light microscopy (Fig. 5A and B ). After colonization of the midgut at 48 h post infection, the cells of the midgut epithelium start to produce rounded blebs, which sometimes also include the cell’s nucleus (Fig. 5B). These nuclei show chromatin condensation, are often surrounded by a perinuclear vacuole (Fig. 5B), and stain TUNEL positive (Fig. 5C), suggesting that they are undergoing apoptosis.

image

Figure 5. Photorhabdus virulence factor expression and programmed cell death in the infected Manduca midgut.

A. Normal morphology of a fold in the Manduca midgut viewed by light microscopy. Note the two types of cells in the epithelium, columnar cells (Cc) and goblet cells (Gc) and the fringing brush-border (Bb). The lumen of the gut (L) and the groove (G) formed by a fold in the midgut epithelium is also indicated.

B. Morphology of an infected fold of the Manduca midgut, 48 h post infection. Note the accumulation of unlabelled bacteria visible within the epithelial groove (G and solid arrow) and that the cells of the midgut epithelium are beginning to detach from the basal lamina and that their nuclei are darkly stained and enlarged (open arrows).

C. Nuclei with the cells of the infected epithelium stain TUNEL positive (open arrows) 48 h post infection, suggesting that they are undergoing programmed cell death.

D. Anti-Tca immunoreactivity detected by a fluorescent conjugate. Immunoreactivity (white arrow) is localized to the same area within the midgut fold as the bacteria themselves. Uninfected midguts gave no signal with the same antibody (data not shown).

E. Anti-PrtA immunoreactivity detected by a fluorescent conjugate 72 h post infection. Immunoreactivity (white arrow) is localized to the basal lamina of tissues such as muscles (M) and fatbody (Fb), which are surrounded by haemolymph (h). Note that both uninfected midguts, infected midguts treated with preimmune sera and infected midguts prior to 48 h post infection (before PrtA was detected at 72 h) gave no signal (data not shown).

F. Overview of anti-PrtA immunoreactivity, at 72 h, within the epithelial fold showing how it highlights the basal lamina of the midgut epithelium.

Scale bar in all panels: 100 μm.

Download figure to PowerPoint

To document if the bacteria within the epithelium-extracellular matrix niche are producing virulence factors, we looked for anti-Toxin complex A (anti-Tca) and anti-RTX-like metalloprotease (anti-PrtA) immunoreactivity. Bacteria within this niche are labelled by the anti-Tca antibody (Fig. 5D) and produce a pattern of immuno-reactivity similar to the distribution of the bacteria themselves. This suggests that the bacteria are producing the gut-active Tca within this epithelium-extracellular matrix niche. Photorhabdus luminescens also produces PrtA within this niche but later in the infection process. Thus anti-PrtA immunoreactivity is not detectable in infected insects until 72 h post infection. In contrast to anti-Tca immunoreactivity, which correlates with the location of the bacteria themselves, anti-PrtA immunoreactivity highlights the membrane (the basal lamina) that forms a barrier between specific tissues such as muscle and fat body, and the haemocoel (Fig. 5E and F ).

Finally, to examine the distribution of these two putative virulence factor proteins at the subcellular level, we used antibody conjugated gold particles under electron microscopy. In sections of infected insects the local distribution of anti-Tca associated gold particles confirms that Tca remains in close association with the bacterial cells themselves (Fig. 6 and B) Thus, Tca complexes appear to be on the extracellular surface of the bacteria (Fig. 6A) and also within extracellular material that either accumulates away from bacterial cells or represents the remnants of lysed bacteria (Fig. 6B). All other tissues of the infected insect appear free of anti-Tca associated labelling. In contrast, anti-PrtA linked gold particles are not associated with the bacteria themselves (Fig. 6C) but are found along strands of the extracellular matrix of the gut and along the basal laminae of other tissues (Fig. 6D).

image

Figure 6.  Subcellular localization of anti-Tca and anti-PrtA associated gold particles within the infected gut by transmission electron microscopy.

A and B. Anti-Tca associated gold particles (arrow) are associated with the extracellular surface of individual bacteria (marked as B within panels) within the gut epithelial niche. Note that other structures such as muscle and other cells show no labelling (data not shown).

B. Anti-Tca associated gold particles (arrows) are sometimes found within extracellular material apparently at some distance from bacterial cells themselves.

C. In contrast to Tca, anti-PrtA associated gold particles are not directly associated with bacterial cells (marked as B within panel) but line the extracellular matrix (Em) of the gut and the basal lamina.

D. This anti-PrtA signal on the extracellular matrix can be clearly seen even in the immediate absence of bacterial cells, suggesting that PrtA is secreted in this niche and then adheres to the insect gut matrix and lamina. Note that other structures such as muscle cells (Mc) are not labelled. For anti-PrtA immunoreactivity, preimmune sera, uninfected sections and infected sections before 48 h (before PrtA expression) show no signal (data not shown).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Photorhabdus and Manduca as a model

The aim of this study was to establish Manduca as a model system for infection by the insect pathogenic bacterium Photorhabdus. Although the Photorhabdus bacteria were introduced into the insect artificially (via injection), and thus without their nematode symbiont, the number of bacterial cells used here (50–70) mimics the number of bacteria (30–200) released by individual nematodes (T. Ciche and J. Ensign, personal communication). Fluctuations in the numbers of recoverable P. luminescens bacteria from infected insects have been previously documented (Forst et al., 1997), however, this is the first detailed numerical and spatial documentation of insect infection by this bacterium. Although the specific colonization events described may be P. luminescens or even strain specific, this dissection of insect infection demonstrates how bacterial infections can be readily detailed in insect models of sufficient size, such as Manduca.

Photorhabdus colonizes different tissues at different times

The ease of dissecting tissues from late instar Manduca larvae, and the well-characterized physiology of this insect model, allowed us to document the relative growth rate of bacteria in different insect tissues (Fig. 1). The colonization of the midgut (Fig. 2) proceeds in a specific anterior to posterior fashion with the bacteria occupying grooves between the highly folded gut epithelium, often next to muscles. These grooves correspond to folds in the midgut epithelium which are held together by both longitudinal and circular muscles. As the extracellular matrix of the gut forms a continuous layer around the midgut epithelium, the bacteria must penetrate this matrix to occupy this niche. The precise mechanism used by the bacteria to penetrate the matrix is not clear but we note that tracheae invade the extracellular matrix and that tracheae may therefore be one potential avenue of entry. Once beneath the extracellular matrix, the bacteria are in very close association with the gut epithelium (Fig. 4). Although there is no obvious reorganization of the gut cells close to the sequestered bacteria, the cells of the epithelium are often detached from their basal lamina suggesting that apoptosis may already have been initiated. Later in infection, large numbers of bacteria (>105 CFU per gram) are recoverable from the fat body and carcass, as well as the midgut and the haemolymph itself. This suggests that colonization of the midgut and haemolymph is initially more important that growth in other tissues and that early in infection Photorhabdus may be actively destroying the gut and the haemolymph based immune system.

Photorhabdus evades the cellular immune response

In assays of the phagocytic competence of haemocyte monolayers (Fig. 3), we found that living W14 bacteria were capable of avoiding phagocytosis, whereas heat-killed P. luminescens were phagocytosed normally. The antiphagocytic factor accumulates in the bacterial supernatant and acts in a dose-dependent manner. These data are consistent with P. luminescens W14 secreting an antiphagocytic factor. In vitro this factor is produced as the bacteria enter exponential growth phase and its presence in vivo is suggested by the observation that cell free plasma taken from W14 infected insects also decreases the ability of haemocyte monolayers to phagocytose. Taken together, this suggests that P. luminescens W14 secretes an antiphagocytic factor both in vitro and in vivo. Because this factor is inactivated by boiling (data not shown), it may be a protein. We note that the antiphagocytic activity of enteropathogenic E. coli (EPEC) is dependent on the presence of the type-III secretion apparatus (Goosney et  al., 1999) whose presence in P. luminescens W14 has been shown by genomic sample sequencing (ffrench-Constant et al., 2000). This raises the formal possibility that the type-III-like secretion machinery of W14 is involved in export of the antiphagocytic factor identified here. Although, we note that antiphagocytic activity is clearly separable in the bacterial supernatant and that contact of the P. luminescens bacterial cells with the haemocytes does not appear to be required.

Localization of putative virulence factor proteins in vivo

Both the toxin Tca and the protease PrtA are produced when the bacterium is within the novel gut-epithelial niche, however, their pattern of expression differs dramatically (Fig. 5). Tca is a gut-active toxin produced by P. luminescens strain W14 and knock-out of the gene that encodes Tca reduces the oral toxicity of W14 supernatant to M. sexta larvae (Bowen et al., 1998). Using fluorescence, anti-Tca immunoreactivity within the infected midgut correlates with the location of the bacteria themselves. Closer examination using electron microscopy-based detection of anti-Tca linked gold particles shows that Tca complexes are found on the extracellular surface of the bacteria while they are within their gut niche (Fig. 6). This shows not only that Tca is produced within this gut niche but also suggests that the toxin complexes themselves remain in close association with the bacteria while they destroy the midgut epithelium. The ‘display’ of the toxin complexes on the surface of the bacterial cell, and the large accumulation of the bacteria within the gut niche, may generate a high effective concentration of Tca and help explain why the gut is destroyed so rapidly during infection.

A second putative virulence factor, PrtA, is an RTX-like zinc-metalloprotease which is encoded next to its own inhibitor and upstream of its own ABC transporter in the W14 genome (Bowen et al., submitted). In the insect, PrtA is expressed late (>72 h) in infection (Daborn et al., 2001) when septicaemia is well established and the insect already dead (death occurs at ~48 h). Although the gene that encodes PrtA and the associated biochemistry of the protease is well documented (Bowen et al., 2000; Bowen et al., submitted), the precise role of this putative virulence factor in infection remains unclear. In contrast to Tca, anti-PrtA immunoreactivity labels the basal lamina of structures within the haemocoel, such as the gut, fat body, muscle and their associated tracheae (Fig. 5). Detailed examination of anti-Tca associated gold particles confirms that secreted PrtA adheres to the extracellular matrix of the gut and basal laminae of other structures in the haemocoel (Fig. 6). This association with basal laminae suggests not only that PrtA is actively exported in vivo but also that the enzyme itself is able to locate and adhere to basal laminae. These observations are consistent with PrtA playing an active role in degrading basal laminae and allowing access to the underlying tissues, such as fat body and muscle. We note that this may be similar to the inferred role of matrix-metalloproteases in facilitating tissue invasion (Belien et al., 1999) and postulate that this lamina specific adherence of PrtA may be involved in breaking down this membrane to enable P. luminescens to bioconvert the underlying tissues.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Bacterial strains, GFP labelling, and insect infection

The bacterial strain used in this study was P. luminescens subsp. akhurstii strain W14 (Bowen et al., 1998). This strain was labelled with Green Fluorescent Protein (GFP) using a GFP-encoding plasmid, which is also arabinose-inducible. Primary-form colonies were inoculated into sterile tubes containing 2 ml of 2% Proteose Peptone no. 3 (PP3) broth, 0.2% arabinose, 2 μg ml–1 ampicillin. The cultures were incubated for 24 h at 30°C on a rotating shaker at 340 r.p.m. For insect injection, bacterial cells were washed with sterile phosphate-buffered saline (PBS; 0.15 M sodium chloride, 10 mM sodium phosphate buffer, pH 7.4) and resuspended in this same buffer. Fifth-instar M. sexta larvae were injected with 10 μl of a suspension containing 50–70 washed bacterial cells. Injections were performed directly into the insect haemocoel using a 50 μl Hamilton syringe with a 30-gauge needle. The numbers of injected bacteria were confirmed by plating a known volume of injected suspension on 2% PP3 plates (1.5% agar plates supplemented with 0.2% arabinose and 2 μg ml–1 ampicillin). Post injection, larvae were held individually with diet, and symptoms of toxicity were noted over time. For arabinose-mediated expression of plasmid-encoded GFP, 10 μl of 20% arabinose in PBS was co-injected into the haemocoel of infected larvae.

To document the numbers of recoverable bacteria within different infected tissues over time, replicates of five larvae were dissected at different time points after injection.

Each larva was surface sterilized with 70% ethanol and then bled to collect its total haemolymph fluid. The total haemolymph collected per larva was weighed. After bleeding, each larva was then also dissected to collect the internal organs. The midgut and fat body were weighed individually and homogenized in PBS using a hand held Potter–Ehlvejen homogenizer. The rest of the insect body was defined as ‘carcass’, weighed and homogenized. To determine the number of recoverable bacteria (CFU per gram of wet weight of tissue or haemolymph), dilution series of tissue homogenates were prepared with PBS and serial dilutions were then plated in duplicate onto PP3-arabinose-ampicilin agar plates. Plates were incubated for 48 h at 30°C and individual colonies counted.

Phagocytic competence of haemocytes

The phagocytic competence of Manduca haemocytes, in the presence or absence of P. luminescens filtrates, was determined in reference to their ability to ingest E. coli. In this context, we measured the phagocytic competence of haemocytes in the presence of W14 supernatant or W14 infected plasma. We also examined the ability of haemocytes to phagocytose W14 itself. P. luminescens W14 culture filtrates were prepared from 72 h cultures grown in PP3. Photorhabdus luminescens cells were removed by centrifugation (3000 r.p.m., 20 min) and passing the resulting supernatant through a 0.22 μm filter. To label bacterial cells with fluorescein-isothiocyanate (FITC), cells of P. luminescens W14 and E. coli DH5α were grown overnight in PP3, and washed three times in peptone water. Cells were then resuspended in a carbonate buffer (pH 9.4) containing 0.1 mg ml–1 FITC, and shaken for 30 min at 28°C in the dark. The cells were then washed six times with peptone water, to remove all traces of free FITC, adjusted to 108 ml–1 in PBS and stored at – 20°C until use. Some cells were heat killed (100°C, 20 min) before being FITC-labelled.

Haemocyte monolayers were prepared as follows. M. sexta larvae were selected 2 days after ecdysis to the fifth larval stage (instar). Insects were chilled on ice for 30 min, and swabbed with 70% ethanol before bleeding from the cut dorsal horn. Approximately 200 μl of haemolymph was allowed to drip into 800 μl of ice-cold Manduca-buffered saline (MBS) solution (Platt and Reynolds, 1985) in a microcentrifuge tube. The tube was immediately inverted to mix, and then centrifuged at 200 g for 5 min at 4°C. The supernatant was replaced with 200 μl of cold Graces Insect Medium (GIM, Sigma) and the haemocytes gently resuspended using a cut pipette tip. Haemocyte density was adjusted to ~5 × 106 ml–1 in GIM using a haemocytometer. Then, 10 mm coverslips were washed in 70% ethanol and placed centrally into each well of a 24-well plate. Cell suspension (100 μl) was pipetted onto each coverslip and then left undisturbed for 30 min (at room temperature) to allow the haemocytes to settle and form a monolayer. The monolayer was washed gently with a few drops of GIM and removed to a new well containing 400 μl of fresh GIM. This procedure leaves >90% of the haemocytes viable, as assessed by Trypan blue exclusion.

Phagocytosis of FITC-labelled E. coli cells in vitro was assessed using a previously described protocol (Hed, 1986; Rohloff et al., 1994) with some modifications. Briefly, haemocyte monolayers were incubated with FITC-labelled E. coli (with or without bacterial culture filtrates) for 2 h in the dark at 28°C. The monolayers were then washed gently in GIM and placed in fresh wells with 300 μl of Trypan blue stain (2 mg ml–1 in GIM) for 20 min. Coverslips were washed in GIM and viewed under fluorescence at × 400 magnification. In this assay, only phagocytosed cells retain their fluorescence after quenching with Trypan blue. The number of haemocytes containing fluorescent bacteria and the total number of haemocytes were determined for each of six fields of view for each treatment.

Insect dissection, histopathology and TUNEL staining

To examine the internal distribution of GFP-expressing Photorhabdus, the larval cuticle was dissected dorsally and laid open to expose the internal organs. GFP-expressing bacteria were detected using a Leica MZFL III binocular microscope fitted with a GFP2 lamp or an Olympus BH-2 microscope fitted with an UV lamp. For insect histopathology, sections of whole larvae or specific tissues from both infected and uninfected larvae were examined. For examination via light microscopy, whole larvae were fixed overnight in Bouin’s solution after several incisions had been made in the cuticle to allow the fixative to permeate the cadaver. Larvae were then embedded in paraffin by using a Leica TP1020 Automatic Tissue Processor. Sections were cut at 3–5 μm and then stained with haematoxylin/eosin and mounted on glass slides with DePeX mounting medium.

For scanning electron microscopy (SEM), larvae were again dissected in PBS to expose internal organs. Midguts were then fixed in 2.5% glutaraldehyde in PBS, post-fixed in 1% osmium tetroxide, dehydrated through an acetone series, and critical point dried. Samples were then coated with gold and examined using a JEOL JSM6310 scanning electron microscope (JEOL Tokyo, Japan).

For transmission electron microscopy (TEM), portions of the anterior midgut epithelium were fixed in 2.5% glutaraldehyde in PBS for 2 h, washed in PBS and then post-fixed in 1% osmium tetroxide for 1 h. Samples were then again dehydrated in an acetone series and embedded in Spurr’s resin (TAAB, Premix). Ultra-thin sections were cut using an ultra-microtome (Leica, Reichert Ultracut E) and stained with uranyl acetate and lead citrate before examination with a JEOL JEM 1200 transmission electron microscope (JEOL Tokyo, Japan).

For immunofluorescence, confocal microscopy was used to examine the distribution of anti-Tca and anti-PrtA immunoreactivity. The anti-Tca antibody used was a rabbit polyclonal antibody raised against the predicted N-terminus of TcaC. The anti-PrtA antibody was also a rabbit polyclonal but was raised against purified PrtA protein. Whole larvae were fixed in 4% paraformaldehyde in PBS for 16 h at room temperature, washed in PBS and then embedded in paraffin and sectioned at 5–7 μm. After being de-waxed and re-hydrated, sections were washed three times (10 min each time) in PBS containing 1% Triton X-100 (TPBS). Sections were then blocked with 5% skimmed milk in TPBS for 1 h and incubated with the primary antibody for 16 h at 4°C. Samples were rinsed in TPBS and incubated with the secondary antibody (CyTM5-conjugated anti-rabbit raised in donkey, Jackson ImmunoResearch) diluted 1 : 500 in TPBS for 3 h at room temperature. After rinsing in PBS, the sections were mounted in MOWIOL and examined with a Zeiss LSM 410 confocal microscope.

For labelling anti-Tca and anti-PrtA immunoreactivity at the ultrastructural level, portions of the anterior midgut epithelium were fixed in 4% paraformaldehyde and 0.5% glutaraldehyde in PBS for 2 h. Samples were then dehydrated in an ethanol series and embedded in L.R. White acrylic resin, incubated with the primary and secondary antibodies (goat anti-rabbit IgG conjugated with 10 nm gold particles) and examined in a JEOL JEM 1200 transmission electron microscope (JEOL Tokyo, Japan).

Apoptosis was detected in whole larvae sections, which were fixed in 4% paraformaldehyde in PBS and embedded in paraffin as described previously, using the TUNEL method according to the manufacturer instructions (In Situ Cell Death Detection Kit; Roche Molecular Biochemical, Indianapolis, IN).

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

This work is supported by a grant from the BBSRC Exploiting Genomics Initiative to R.ff.-C and S.E.R, by a Royal Society Merit Award to R. ff-C, and by a collaborative exchange programme between the University of Bath and the Brazilian Government (CAPES). P. Dean is supported by a BBSRC CASE studentship with CSL York. We thank M. Blight for provision of the purified protease for antibody production and members of the D. Clarke laboratory at Bath for suggestions and discussion.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  • Aballay, A. and Ausubel, F.M. (2001) Programmed cell death mediated by ced-3 and ced-4 protects Caenorhabditis elegans from Salmonella typhimurium-mediated killing. Proc Natl Acad Sci USA 98: 27352739.
  • Belien, A.T., Paganetti, P.A. and Schwab, M.E. (1999) Membrane-type 1 matrix metalloprotease (MT1-MMP) enables invasive migration of glioma cells in central nervous system white matter. J Cell Biol 144: 373384.
  • Bowen, D., Blackburn, M., Rocheleau, T., Grutzmacher, C. and Ffrench-Constant, R.H. (2000) Secreted proteases from Photorhabdus luminescens: separation of the extracellular proteases from the insecticidal Tc toxin complexes. Ins Biochem Mol Biol 30: 6974.
  • Bowen, D., Rocheleau, T.A., Blackburn, M., Andreev, O., Golubeva, E., Bhartia, R. and Ffrench-Constant, R.H. (1998) Insecticidal toxins from the bacterium Photorhabdus luminescens. Science 280: 21292132.
  • Cioffi, M. (1979) The morphology and fine structure of the larval midgut of a moth (Manduca sexta) in relation to active ion transport. Tissue Cell Res 11: 467479.
  • Ffrench-Constant, R.H., Waterfield, N., Burland, V., Perna, N.T., Daborn, P.J., Bowen, D. and Blattner, F.R. (2000) A genomic sample sequence of the entomopathogenic bacterium Photorhabdus luminescens W14: potential implications for virulence. Appl Environ Microbiol 66: 33103329.
  • Daborn, P.J., Waterfield, N., Blight, M.A. and Ffrench-Constant, R.H. (2001) Measuring virulence factor expression by the pathogenic bacterium Photorhabdus luminescens in culture and during insect infection. J Bacteriol 183: 5834 5839.
  • De Gregorio, E., Spellman, P.T., Rubin, G.M. and Lemaitre, B. (2001) Genome-wide analysis of the Drosophila immune response by using oligonucleotide microarrays. Proc Natl Acad Sci USA 98: 1259012595.
  • Forst, S. and Nealson, K. (1996) Molecular biology of the symbiotic-pathogenic bacteria Xenorhabdus spp. & Photor-habdus spp. Microbiol Rev 60: 2143.
  • Forst, S., Dowds, B., Boemare, N. and Stackebrandt, E. (1997) Xenorhabdus and Photorhabdus spp.: bugs that kill bugs. Ann Rev Microbiol 51: 4772.
  • Forst, S. and Clarke, D. (2001) Bacteria-nematode symbiosis. In: Entomopathogenic Nematology.Gaugler, R. (ed)., London: CAB International, pp. 5777.
  • Gallagher, L.A. and Manoil, C. (2001) Pseudomonas aeruginosa PAO1 kills Caenorhabditis elegans by cyanide poisoning. J Bacteriol 183: 62076214.
  • Garsin, D.A., Sifri, C.D., Mylonakis, E., Qin, X., Singh, K.V., Murray, B.E., Calderwood, S.B. and Ausubel, F.M. (2001) A simple model host for identifying Gram-positive virulence factors. Proc Natl Acad Sci USA 98: 1089210897.
  • Goosney, D.L., Celli, J., Kenny, B. and Finlay, B.B. (1999) Enteropathogenic Escherichia coli inhibits phagocytosis. Infect Immun 67: 490495.
  • Hed, J. (1986) Methods for distinguishing ingested from adhering particles. Methods Enzymol 132: 198204.
  • Lane, N.J., Dallai, R. and Ashhurst, D.E. (1996) Structural macromolecules of the cell membranes and the extracellular matrices of the insect midgut. Lehane, M.J. and Billingsley, P.F. (eds). In: >Biology of the Insect Midgut. London: Chapman & Hall, pp. 115150.
  • O’Quinn, A.L., Wiegand, E.M. and Jeddeloh, J.A. (2001) Burkholderia pseudomallei kills the nematode Caenorhabditis elegans using an endotoxin-mediated paralysis. Cell Microbiol 3: 381393.
  • Platt, N. and Reynolds, S.E. (1985) Cardioactive peptides from the CNS of a caterpillar, the tobacco hornworm, Manduca sexta. J Exp Biol 114: 397414.
  • Rohloff, L.H., Weisner, A. and Gotz, P. (1994) A fluorescence assay demonstrating stimulation of phagocytosis by haemolymph molecules of Galleria mellonella. J Ins Physiol 40: 10451049.
  • Wu, L.P., Choe, K.M., Lu, Y. and Anderson, K.V. (2001) Drosophila immunity: genes on the third chromosome required for the response to bacterial infection. Genetics 159: 189199.