Abstract

This study tested the hypothesis that venom from the ectoparasitic wasp Nasonia vitripennis targets brain tissue and fat body from its flesh fly host, Sarcophaga bullata. By 1 h postenvenomation, some brain neurons began to show irregularities in nuclear shape, and though they were predominately euchromatic, there was evidence of heterochromatin formation. Irregularity in the nuclear envelope became more prominent by 3 h after envenomation, as did the condensation of heterochromatin. The severity of ultrastructural changes continued to increase until at least 24 h after parasitoid attack. At this point, cellular swelling and extensive heterochromatic inclusions were evident, multivesicular bodies occurred in the cytoplasm of some cells, and the rough endoplasmic reticulum was dilated in many of the cells. Immunohistochemical staining revealed significant apoptosis in neurons located in brain tissues. By contrast, there was no evidence of any morphological or ultrastructural disturbances in fat body tissues up to 24 h after envenomation, nor did any of the cells display signs of cell death.

1. Introduction

Ectoparasitic wasps typically subdue their insect hosts by induction of some type of halt or delay in development [1, 2]. Host arrest often is the result of venom-induced paralysis [3, 4]. In most cases, paralysis is sustained until host death and the venom constituents operate at neuromuscular junctions and/or block synaptic transmission [57]. The paralyzed host essentially becomes nothing more than a fixed or finite resource for feeding parasitoid progeny. In fact, there is little evidence available to suggest that ectoparasitic wasps relying on paralytic venoms physiologically manipulate their hosts, or even have a need to, beyond triggering permanent paralysis.

The situation can be quite different with ectoparasitic wasps that use nonparalytic venoms. Host developmental suppression is associated with more than just immobilization of the host: a series of physiological and biochemical alterations occur in the host that resemble those changes evoked by koinobiotic endoparasitoids [8, 9]. One example is the pupal parasitoid Nasonia vitripennis (Walker) (Hymenoptera: Pteromalidae). This wasp injects a complex proteinacious venom [10] into fly hosts during oviposition that elicits a developmental arrest sustained until death [11]. The arrestment is characterized by a reduction in respiratory metabolism [12], followed by tissue-specific increases in lipid content [13] that are essential to the wasp’s offspring successfully completing development [14]. Most features of pharate adult development fail to occur in envenomated pupae and pharate adults of the flesh fly, Sarcophaga bullata Parker (Diptera: Sarcophagidae). However, depending on the stage of host development at the onset of parasitism, eye pigment deposition and body bristle formation in fly hosts still occur, albeit as intermediates of wild-type phenotypes and greatly delayed in terms of onset by comparison to normal pharate adult development [2, 11]. The fact that envenomation triggers developmental arrest in the host and that fly development is not completely retarded suggests that wasp venom targets brain tissue [15].

Venom-induced manipulations of host physiology and development appear to depend on signaling pathways involving G-protein sensitive receptors [15, 16]. In vitro assays suggest that disruption of these signaling pathways leads to an imbalance in calcium homeostasis that culminates in cell death. Ultrastructural and morphological evidence using cultured cells (BTI-TN-5B1-4 cells) indicates that venom-induced death shares features consistent with apoptosis, nonapoptotic programmed cell death, and oncosis [17]. Genome mining and proteomic analyses of venom glands from N. vitripennis have led to the identification of multiple venom proteins with defined roles in programmed cell death [10, 18]. However, no functional studies have been performed with isolated venom proteins to determine their roles in the host-parasitoid system, nor have any investigations examined the type of cell death evoked in host tissues in vivo.

In this study, we attempted to address the hypothesis that venom from N. vitripennis targets brain tissue and fat body from the fly host, S. bullata. We specifically examined the morphological and pathological changes that occur in select fly tissues (brain and fat body) following envenomation by N. vitripennis.

2. Materials and Methods

2.1. Insect Rearing

N. vitripennis was maintained as a laboratory colony on pupae and pharate adults of S. bullata as described previously [11]. Adults and larvae were reared under a light-dark cycle of light 15 h: dark 9 hours at 25°C. Twenty to thirty females (3–7 days after emergence from host puparia) were placed in a Petri dish (  mm) with 40–60 nondiapausing pupae (4 days after pupariation at 25°C) of S. bullata and a 50% (v/v) honey-water solution. After 24 h, the adult wasps were removed and parasitized pupae maintained at 25°C, LD 15 : 9 h. Under these conditions, N. vitripennis completes development from egg to adult (emergence) in 12 days.

A colony of S. bullata was maintained as described by Denlinger [19]. Larvae were fed beef liver throughout development at 25°C with a photoperiodic cycle of LD 15 : 9 h. Adults were allowed to feed ad libitum on beef liver, sugar cubes, and water at 25°C with a photoperiodic cycle of LD 15 : 9 h. To synchronize fly development for assessing host age, third stage larvae that had begun to wander from food but prior to crop emptying were placed in a vented glass jar with 1-2 mL tap water. Larvae were held under these conditions for 3 days at 25°C with frequent (3–5 times/d) water changes. This “wet” treatment temporarily inhibits the release of ecdysteroids until the larvae are placed in dry conditions, thereby synchronizing initiation of pupariation [20].

2.2. Exposure of Parasitoids to Hosts

Nondiapausing pharate adults (5 days after pupariation at 25°C) of S. bullata were exposed singly to individual females of N. vitripennis as described previously [11]. Oviposition was restricted to the posterior 1/3 of the fly puparia (individual puparia were wrapped in aluminum foil) [11] to facilitate parasitoid removal. After host exposure, the adult wasps were discarded, the posterior cap of each puparium was opened, and the parasite's eggs were removed. Each pharate adult was then kept separately in glass culture tubes (  mm) capped with cotton plugs and maintained at 25°C, LD 15 : 9 h until brains or fat body were excised.

2.3. Preparation of Brain Tissue

Brains were removed from envenomated and nonenvenomated pharate adults of S. bullata by dissection under a stereo dissecting microscope (Zeiss Stemi 2000, Germany). The head of each fly was severed from the thorax using iris scissors and placed in phosphate buffered saline (25 mM, pH 7.4). The head integument was then cut longitudinally along the dorsal surface to bisect between the optic lobes and expose the brain. Iris scissors were used to gently cut all tracheal and neural connections to the brain so that the brain could be transferred to cold Karnovsky’s fixative (4% paraformaldehyde and 5% glutaraldehyde in 0.1 M phosphate buffer (PB) (pH 7.2) for 8–18 h [21]. After rinses in cold PB, tissues were postfixed in 1% osmium tetroxide for 2 h, rinsed again in PB, and stained with 2% uranyl acetate for 3 h. Tissues were rapidly dehydrated in an ethanol series (50%, 70%, 90%, and 100%) for 10 min each, infiltrated with propylene oxide: plastic, and stored in desiccated plastic overnight. Tissue pieces were placed in molds and polymerized overnight in an oven pre-set at 60°C. Brains were excised from pharate adults at 0, 0.5, 1, 3, 6, and 24 h postenvenomation, with tissues from 2–4 hosts examined by transmission electron microscopy (TEM) at each time point.

Thick sections (1 μm) were cut on an LKB Astrodome 8800 ultramicrotome and collected on glass microscope slides or glass cover slips for light microscopic observations and for immunocytochemical staining. Thin sections were then cut and collected on 200-mesh copper grids, stained with uranyl acetate and lead citrate. All electron microscopic observations were made on a JEOL 100S electron microscope at 80 kV. Section orientations were determined by using optic lobes as reference points. The percentage of cells that displayed formation of plasma membrane blebs, irregularities (i.e., convolutions) of the nuclear envelope, or heterochromatin formation in nuclei were determined from captured images. Image analyses were performed using brain thin sections from three hosts, counting a minimum of 500 cells/treatment/time point. To ensure the same cells were not counted multiple times, sequential thin sections from the same host were not used for cell counts.

2.4. Preparation of Fat Body Tissue

Fat body tissues were removed from envenomated and nonenvenomated pharate adults of S. bullata by dissection under a stereo dissecting microscope essentially as described by Rivers and Denlinger [13]. Lobes of fat body were collected from the head and anterior thoracic regions using fine forceps. Iris scissors were used to gently cut all tracheal and neural connections so that isolated fat body were transferred to cold Karnovsky’s fixative and then subjected to the same fixation, staining, and sectioning procedures described for brain tissue. Fat bodies were excised from pharate adults at 0, 0.5, 1, 3, 6, and 24 h postenvenomation, with tissues from 2–4 hosts examined by TEM at each time point.

2.5. Toluidine Blue Staining

In order to analyze the basic architecture of fat body tissues following envenomation, thick sections were stained with toluidine blue. Toluidine blue stains cytoplasm and, when present, cytoplasmic inclusions or granules [22]. Tissue sections (1 μm) were transferred to glass slides and then stained with 1% toluidine blue in PB, followed by mounting in Permount mounting media (Fisher Scientific Supply, Hanover, Ill, USA) and with a glass cover slip placed over the media. Sections were examined by light microscopy and images captured using an insight 4 SPOT RT fire wire digital camera (14.1 Monochrome with IR filter, Diagnostic Instruments. Inc, Sterling Heights, Mich), mounted on a compound microscope (Nikon), and connected to a Macintosh Power Mac G5 computer (Apple). Images were analyzed using SPOT (v. 4.5) and Adobe Photoshop software (Creative Suite 2, Photoshop v. 9.0).

2.6. Immunocytochemical Staining

Detection of cell death in thick sections of brain tissue isolated from envenomated flies was performed using in situ labeling based on the TUNEL assay (Apoptag Peroxidase in situ Apoptosis Detection Kit, Chemicon International). Tissue sections were washed in three changes of xylene (5 min/wash), followed by two washes (5 min each) in absolute ethanol and a single wash (3 min) each in 95% and 70% ethanol, respectively, prior to following the manufacturer’s instructions with the kit. Apoptotic cell detection relied on a peroxidase reporter molecule with diaminobenzidine serving as the enzymatic substrate. Tissue sections were counterstained in 0.5% (w/v) methyl green in 0.1 M sodium acetate (pH 4.0) to visualize nuclei. Stained sections were mounted in Permount, and then a glass cover slip was placed over the media. Dried specimens were examined by light microscopy and images captured and analyzed as described for thin sections. Apoptosis was determined from captured images of brain thick sections from three hosts, counting a minimum of 500 cells/treatment/time point. To ensure the same cells were not counted multiple times, sequential thick sections from the same host were not used for cell counts.

2.7. Statistical Analyses

Means were compared using one- and two-way analyses of variance (ANOVA) and Student Newman-Keuls multiple comparisons tests using GraphPad statistical software (InStat v. 3.00, ). Percentage data was arcsine transformed prior to analysis.

3. Results

3.1. Brain Organization

Brains from young pharate adults of S. bullata resemble the structural organization of other cyclorrhaphous flies: the brain represents the fusion of supra and subesophageal ganglia, with two distinct optic lobes extending from the protocerebrum [23, 24]. Like in Drosophila, the protocerebrum appeared to be composed of closely arranged neurons and neuroglia. Differentiation during the cryptocephalic to phanerocephalic stage metamorphosis leads to the apoptotically controlled degeneration of the ring gland [25] and the complete formation of two optic lobes in S. bullata [26]. In this study, the cellular organization of phanerocephalic brain tissue observed from unparasitized S. bullata closely resembled that described for Drosophila [27].

3.2. Venom-Induced Ultrastructural Changes in Fly Brains

Following parasitism by N. vitripennis, wasp eggs were removed from young pharate adults of S. bullata so that the impact of venom on brain ultrastructure could be examined. Brain sections from unparasitized flies revealed no irregularities in neuronal cell structure: cell bodies and their nuclei mostly appeared oval in shape, nuclei were euchromatic, and the cells were devoid of obvious vacuoles and blebs (Figure 1(a)). Similarly, brain tissue removed from flies 30 min after envenomation appeared essentially identical to controls, with no detectable changes in cell ultrastructure (Figure 1(a); Table 1). By 1 h postenvenomation, some irregularities in nuclear shape were observed, although the nuclei did not appear convoluted. In these cells, the nuclei were predominantly euchromatic (>70%, Table 1), but there was some evidence of heterochromatin formation (Figure 1(a)). Convolutions in the nuclear envelope of neuronal cells became more prominent by 3 h after envenomation, as did heterochromatin in nuclei and increased electron density of the cytoplasm (Figure 1(b)). Significant ultrastructural changes were evident within 6 h following parasite attack as evidenced by convolutions of the nuclear envelope (Table 1), enlarged and darkened nucleoli, and extensive heterochromatic inclusions (Figure 1(b)). Multilamellar bodies were also evident at this time point (Figure 2(a)). These changes were even more pronounced by 24 h postenvenomation (Table 1). In addition to extensive convolutions of the nuclear envelope and heterochromatic inclusions (Figure 1(b)), multivesicular bodies occurred in the cytoplasm of some cells (Figure 2(c)), and the rough endoplasmic reticulum was dilated in many of the brain neurons (Figure 2(c)). By contrast, mitochondria in brain tissue did not appear to be altered by venom for at least 24 h after envenomation (Figures 2(b) and 2(d)).

3.3. Venom-Induced Death in Fly Brains

The cellular distortions detected in brain tissue following envenomation by N. vitripennis resulted in widespread, but not indiscriminate, cell death. Induction of apoptotic cell death was monitored using an in situ labeling kit based on the TUNEL assay and that relied on a peroxidase reporter molecule. Apoptotic cells were readily distinguished as cells developed blue-purple color intermediates, while living (prior to being embedded), and oncotic cells were unstained. Consistent with venom-induced ultrastructural changes, the number of apoptotic cells detected in brain tissue increased with the length of time after envenomation (Figure 3, Table 2). Few neuronal cells appeared dead due to apoptosis in brain tissue from unparasitized flies or from tissue extracted from envenomated flies up until 3 h postvenom injection (<14%, , Table 2). However, the vast majority of dead cells observed in brain tissue 6 h after parasitism stained positively for apoptosis (Table 2). This pattern of staining was not as evident in 24 h brains, presumably because the cells were irreversibly injured and the nuclear DNA severely degraded due to endonuclease activity. The latter would prevent detection of apoptotic cells since the TUNEL assay relies on in situ labeling of nucleotides.

Apoptosis did not appear to be the only form of cell death induced by venom in brain tissue. Some of the cells were observed at 24 h postenvenomation (Figure 3), and to a lesser extent at 6 h, to be swollen. Cellular swelling is consistent with oncosis, which typically results in lysis. However, there was little evidence that cytolysis occurred in brain tissue at any time point examined.

3.4. Venom-Induced Morphological and Ultrastructural Changes in Fat Body

Light microscopic examination of fat body sections stained with toluidine blue revealed no obvious morphological changes in these tissues at any time point up to 24 h postenvenomation (Figure 4). All fat body cells displayed a large, centrally located nucleus, several prominent vacuoles, and the presence of lipid droplets distributed throughout the cytosol (Figure 4). Similarly, transmission electron micrographs revealed nearly indistinguishable ultrastructure of nuclei, vacuoles, and lipid droplets in fat body excised from unparasitized and envenomated pharate adult of S. bullata at all time intervals examined up to 24 h (Figure 5 only shows 0 and 24 h). There was also no morphological or ultrastructural evidence for induction of apoptosis or any other form of cell death in fat body cells following envenomation by N. vitripennis (Figure 5).

4. Discussion

Parasitic wasp venoms contain a wealth of regulatory agents that are capable of modifying growth and development of host insects to suit the needs of the parasitoid’s progeny [8, 10]. Despite increasing efforts to characterize wasp venoms, including cloning and sequencing of some venom proteins [18, 2830], very little has been revealed regarding the mechanism of action of any of these venoms [5]. Venom from N. vitripennis has been the subject of several recent modes of action studies [16, 31, 32], yet insufficient information is available to determine precisely which tissues are targeted in the fly host and how those tissues are injured to alter normal functions. Venom assays exploiting pupariation, extrication, and posteclosion behaviors of a preferred host S. bullata suggest that N. vitripennis venom alters neurons of the central nervous system (CNS) presumed to reside within the brain [33, 34]. However, the stages of host development (larval and imago) used in those studies are not attacked in nature by adult females of N. vitripennis, so implication of the brain as a target of venom based on these behavioral assays is circumstantial at best. This study has provided the first evidence that venom from N. vitripennis directly targets the brain in natural hosts. The fact that neurons in brain tissue displayed susceptibility to wasp venom, that the onset of significant cell death in neuronal tissue did not occur until several hours after envenomation, and that the predominant form of cell death induced by venom was apoptosis argues that N. vitripennis venom targets specific regions of the brain to manipulate, rather than kill, the host. These observations are also consistent with our recent findings that show in vitro this venom induces multiple forms of cell death and that the dominant mechanism of death triggered by N. vitripennis is apoptosis [17].

The dominant host response to envenomation by N. vitripennis is in fact not death, but instead, the induction of a developmental arrest [11, 14]. The halt in fly development begins a dynamic set of changes in fly physiology characterized by a suppression of respiratory metabolism [12], elevations in lipid synthesis [12, 13], depression of immune responses [35], and altered protein expression [15]. Disrupted protein expression has been detected in several host tissues, but changes in brain heat shock protein (hsp) synthesis appear to be some of the most dramatic deviations from “normal” protein expression. The function of hsps during parasitism has yet to be deciphered, but these proteins may be required to arrest host development via apoptotic pathways [4, 36]. Evidence from Drosophila melanogaster indicates that expression of hsp70 during nonstress conditions leads to slowed development [37] and the cell cycle in cultured cells from D. melanogaster can be arrested in the presence of small hsps [38]. Further investigation is needed to elucidate whether venom-mediated apoptosis in brain tissues and/or altered hsp expression are keys to induction and/or maintenance of developmental arrest following venom injection by N. vitripennis.

The injured neurons in flies envenomated by N. vitripennis appeared morphologically identical to brain neurons in adult D. melanogaster exposed to either high-LET krypton or argon ions [39, 40]. In the case of irradiated flies, the swollen cells eventually lysed and fragmented, whereas there was little evidence of lysis in brain tissue of S. bullata. Similarly, both radiation treatment and envenomation initially displayed no little impact on neuroglia [40]. However, in contrast to D. melanogaster [40], neurons in envenomated S. bullata displayed convoluted nuclear membranes with condensed heterochromatin by 6 h postenvenomation, and, by 24 h, these cells were enlarged due to swelling. The differences evoked by these two types of toxic insults, high energy radiation, and envenomation may be consistent with the earlier prediction of Rinehart et al. [15] that though venom from N. vitripennis injures cells of the host, it does not appear to turn on a typical general stress response in the host [41]. Instead, venom appears to be targeting specific cells in brain tissue of the host to induce developmental arrest and redirect the physiology of the fly for the benefit of its progeny.

The morphology and ultrastructure of fat body tissue did not appear to be affected by wasp envenomation, at least not during the first 24 h following parasitoid attack. This was unexpected since earlier studies have shown sharp elevations in hemolymph and fat body lipid levels in S. bullata following parasitism, envenomation, and artificial venom injections [12, 13]. If venom elicits de novo synthesis of lipid in host fat body, then increased lipid droplet content would be expected in envenomated hosts. However, there were no differences between unparasitized and envenomated fat body detected by light microscopy or transmission electron microscopy. Alternatively, venom may function to liberate accumulated lipids from fat body by inducing cell death, such as occurs with Meteorus pulchricornis and Cotesia kariyai [42, 43]. Indeed, lipases and hydrolases have been identified in venom by genomic mining and proteomic analyses [10, 18], and these enzymes have been predicted to function in fat body digestion [8, 10]. However, there was no evidence in this study that venom from N. vitripennis induced cell death in fat body cells. This suggests that either the cellular events associated with venom-induced elevations in host lipids occur later than 24 h postenvenomation or that the primary changes in host lipids are independent of fat body tissues.

Conflict of Interests

The authors declare that they have no conflict of interests.

Acknowledgment

This work was supported in part by a faculty development grant from Loyola College (University) in Maryland to D. B. Rivers.