Virulence factors and strategies of Leptopilina spp.: selective responses in Drosophila hosts

Abstract

To ensure survival, parasitic wasps of Drosophila have evolved strategies to optimize host development to their advantage. They also produce virulence factors that allow them to overcome or evade host defense. Wasp infection provokes cellular and humoral defense reactions, resulting in alteration in gene expression of the host. The activation of these reactions is controlled by conserved mechanisms shared by other invertebrate and vertebrate animals. Application of genomics and bioinformatics approaches is beginning to reveal comparative host gene expression changes after infection by different parasitic wasps. We analyze this comparison in the context of host physiology and immune cells, as well as the biology of the venom factors that wasps introduce into their hosts during oviposition. We compare virulence strategies of Leptopilina boulardi and L. heterotoma, in relation to genome-wide changes in gene expression in the fly hosts after infection. This analysis highlights fundamental differences in the changes that the host undergoes in its immune and general physiology in response to the two parasitic wasps. Such a comparative approach has the potential of revealing mechanisms governing the evolution of pathogenicity and how it impacts host range.

FIGURE 5.1

Host range and susceptibility. (A) A confluence of ecological, behavioral and host factors contributes to host range. (B) In the event of infection, success of a parasite will depend on the combination of specific individual host and parasite factors, as shown. Host resistance (immune mechanisms), host resources (nonimmune physiological factors that enable host defense) versus mechanisms of evasion and/or virulence on the part of the parasitoid. Clear outcomes of either the parasite [quadrant 1; “high virulence” parasite (e.g., L. heterotoma) infecting “low resistance” host] or the host [(quadrant 4; “low virulence” parasites (e.g., A. tabida) infecting a “high resistance” host)] can be predicted. However, in case of high virulence/high resistance or low virulence/low resistance (quadrants 2 and 3), specific outcomes may be more difficult to predict.

FIGURE 5.2

Hemocyte-mediated encapsulation depends on both hemocyte concentrations and hemocyte differentiation. (A) and (B) L. boulardi-17 infection induces massive encapsulation in D. yakuba. (A) Uninfected and (B) infected third-instar larvae. (C) Hemocyte concentration in D. melanogaster larvae (log normal transformation of hemocyte concentrations) shows a normal distribution. Empirically determined range of normal hemocyte concentration in staged third-instar larvae is shown (vertical lines). Values of hemocyte concentration from heterozygotes and mutants were derived from outcrossed animals. Genotype of mutants is italicized (allele name, where used, is superscripted), heterozygotic genotypes are not. Genotypes shown in bold pertain to JAK-STAT signaling components (os, outstretched; hop, hopscotch; STAT, STAT92E), whereas those that are not bold relate to Toll signaling (snk, snake; ea, easter; spz, spatzle; Tl, Toll; tub, tube; pll, pelle; cact, cactus. CS, Canton S is a wild-type strain). (D,E) Direct correlation of encapsulation capacity with hemocyte concentration in wild-type and control values (D). This correlation is violated in mutant animals. Genotypes are referred to as in legend above. Note: Figures modified from Sorrentino et al. (2004) with permission from the publisher.

FIGURE 5.3

Virus-like particle (VLP) structure and biogenesis. (A,B) Sections of gradient-purified VLPs from L. victoriae. Notice the somewhat irregular pentagonal/hexagonal organization of the VLP core that extends into projections. (C–E) Section of a region from the lumen of L. victoriae long gland with immature VLPs at different magnifications. (F,G) Association of p40 antigen within and around immature VLPs of L. heterotoma (arrows point to immuno-gold particles linked to secondary antibody (G)). Sample in panel (F) was not treated with the primary antibody. (H and I) L. victoriae infection induces lamellocytes to alter morphology. Hemocytes from hemolymph of normal (H) or infected hosts (I) stained with rhodamine-phalloidin (F-actin) and Hoechst (blue). Normal hemolymph has 10-μm wide spherical plasmatocytes (arrowheads) but very few, if any, lamellocytes. None are present in this panel. Infected lamellocytes exhibit altered, spindle-shaped morphology.

FIGURE 5.4

Components and targets of pro-phenol oxidase, Toll-Dorsal/Dif and JAK/STAT pathways. Genes whose expression is modulated by L. boulardi-17 infection (up- or downregulated) are shown. The Toll-Dorsal/Dif and JAK/STAT pathway components that were tested in genetic experiments (Sorrentino et al., 2004) for their requirement for a robust encapsulation response are also shown (genetic evidence). Note: Figure modified from Schlenke et al. (2007).

FIGURE 5.5

Venom apparatus, canals and contents. (A) and (B) A schematic (A) that corresponds to the whole mount (phase, B) of a dissected sample of a venom apparatus from L. heterotoma. Different parts of the organ are labeled. (C) Phalloidin staining of the venom (long) gland from L. heterotoma reveals the scope and organization of a supra-cellular canal system. Made of individual secretory units (one per secretory cell), VLP precursors and other constituent proteins make their way, initially through the rough portion of the canal system, present in the cytoplasm of the large secretory cell itself. This portion (panel inset) is composed of membranous folds of actin-rich microvilli that give the structure a rough appearance. Rough canal (RC) loses the folds and narrows into the smooth canal, which opens into the gland lumen. (D–F) Transmission electron micrographs of the venom gland. (D) Rough canal in cross section, a cell of the intimal layer and immature VLPs are seen in this low magnification view of the gland. (E and F) Cross-sections of the rough (E) and smooth (F) canals of L. victoriae venom glands. Both these structures contain VLP precursors (arrows).

FIGURE 5.6

Host gene expression changes after L. boulardi-17 and L. heterotoma-14 infections. (A) Outline of the design and analysis of the microarray experiment as described in Schlenke et al. (2007). (B) Venn diagram showing number of genes whose expression is modulated at any of the three points after L. heterotoma-14 or L. boulardi-17 infections. Fold change in gene expression was calculated from the published data and analysis threshold was arbitrarily set (fold change ≤0.75 and ≥1.5). (C) Differentially expressed genes identified in panel (B) can be classified into five functional classes as shown. L. boulardi-17-infected larvae differentially express 265 genes compared to 91 genes in L. heterotoma-14-infected larvae. Eighty genes are expressed in the host after either infection. (D) The five functional categories of differentially expressed genes organized by timepoints. Total genes that are differentially expressed at each timepoint after L. boulardi-17 infection are 125, 136 and 163, and after L. heterotoma-14 infection are 42, 40 and 43, respectively. Not all genes are differentially regulated at all three timepoints for the same wasp infection, and therefore, the list of differentially regulated genes at one timepoint is different than the list of genes at another timepoint.

FIGURE 5.7

Wasp-induced fat body expression of Drosomycin-GFP reporter in vivo. (A) Reporter construct of Drosomycin-GFP designed to assay in vivo activation of the promoter (Ferrandon et al., 1998; Tzou et al., 2000). (B–E) Upon L. boulardi infection, Drosomycin-GFP expression is activated through the 24-h period of third-instar larval stages. Hosts were exposed to wasps for 24 h. Fat body samples were dissected 2 or 18 h after infection. Drosomycin-GFP was not expressed in the absence of infection (B, C), but was clearly expressed after infection (D, E).

FIGURE 5.8

Effects of L. boulardi and L. heterotoma on D. melanogaster. Differences in the activities of virulence strategies (active vs. passive) and factors (proteins affecting hemocyte morphology vs. viability) may account for differences in the activation of immune pathways after infection by L. boulardi-17 and L. heterotoma-14. In the former case, only encapsulation is abolished in D. melanogaster hosts. In the latter, all three aspects of immune responses are compromised.