Spodoptera frugiperda X-tox protein, an immune related defensin rosary, has lost the function of ancestral defensins

Abstract

Background: X-tox proteins are a family of immune-related proteins only found in Lepidoptera and characterized by imperfectly conserved tandem repeats of several defensin-like motifs. Previous phylogenetic analysis of X-tox genes supported the hypothesis that X-tox have evolved from defensins in a lineage-specific gene evolution restricted to Lepidoptera. In this paper, we performed a protein study in which we asked whether X-tox proteins have conserved the antimicrobial functions of their ancestral defensins and have evolved as defensin reservoirs.

Methodology/principal findings: We followed the outcome of Spod-11-tox, an X-tox protein characterized in Spodoptera frugiperda, in bacteria-challenged larvae using both immunochemistry and antimicrobial assays. Three hours post infection, the Spod-11-tox protein was expressed in 80% of the two main classes of circulating hemocytes (granulocytes and plasmatocytes). Located in secretory granules of hemocytes, Spod-11-tox was never observed in contact with microorganisms entrapped within phagolyzosomes showing that Spod-11-tox is not involved in intracellular pathogen killing. In fact, the Spod-11-tox protein was found to be secreted into the hemolymph of experimentally challenged larvae. In order to determine antimicrobial properties of the Spod-11-tox protein, it was consequently fractionated according to a protocol frequently used for antimicrobial peptide purification. Over the course of purification, the anti-Spod-11-tox immunoreactivity was found to be dissociated from the antimicrobial activity. This indicates that Spod-11-tox is not processed into bioactive defensins in response to a microbial challenge.

Conclusions/significance: Altogether, our results show that X-tox proteins have not evolved as defensin reservoirs and have lost the antimicrobial properties of the ancestral insect defensins. The lepidopteran X-tox protein family will provide a valuable and tractable model to improve our knowledge on the molecular evolution of defensins, a class of innate immune effectors largely distributed over the three eukaryotic kingdoms.

Figure 1. Expression pattern of Spod-11-tox during the immune response.

A) Spod-11-tox is mainly expressed in hemocytes upon immune challenge. Western blot analyses were performed with protein extracts from plasma (Pl), fat body (FB) and hemocytes (He) of S. frugiperda larvae 24 h post injection of PBS (P) or E. coli (Ec, 10⁵ bacteria/larvae). Proteins were loaded (20 µg/lane) on a 10% SDS-PAGE. The blot was successively probed with both a polyclonal anti-Spod-11-tox antibody (Spod-11-tox, upper panel) and an anti-α-tubulin antibody (α-Tub, lower panel). The absence of a signal with α-tubulin antibody indicates that the plasma preparations are not contaminated by cells. B) Sub cellular location of Spod-11-tox. Membrane (memb.) and cytosolic (cyto.) fractions were prepared with hemocytes withdrawn from S. frugiperda larvae (challenged as described in A) and separated on a 10% SDS-PAGE. Protein separation and western blot were performed as in A.

Figure 2. Hemocyte cell types involved in Spod-11-tox production.

A) Immunofluorescence staining of Spod-11-tox in hemocytes withdrawn from S. frugiperda larvae 6 h post injection with PBS (a, b, c) or E. coli (10⁵ bacteria/larvae) (d, e, f). Cells were DAPI stained (a, d) and immunolabelled with rabbit anti-Spod-11-tox revealed by TRITC-conjugate anti-rabbit (b, e). Overlays of DAPI and TRITC were shown in c and f. Yellow arrows and white arrowheads indicate plasmatocytes and granulocytes respectively. Bars = 20 µm. B) Spod-11-tox appeared mainly aggregated within the cytoplasm in a manner evocative of granule-like vesicles. Higher magnification of a plasmatocyte withdrawn from S. frugiperda larvae 6 h post injection of E. coli (10⁵ bacteria/larvae). Bars = 10 µm. C) Relative percentage of Spod-11-tox immunolabelled circulating hemocytes. Same experiment as in A except that hemocytes were withdrawn at the indicated times. Data are presented as mean±SD of at least three independent experiments (a minimum of 2,000 hemocytes/condition were counted). D) Intracellular location of Spod-11-tox. Same experiment as in B except that hemocytes were withdrawn 3 h post challenge and were (Per.) or were not (N Per.) permeabilized before immunolabelling.

Figure 3. Subcellular localization of Spod-11-tox.

A) Confocal microscopy of phagocytosed E. coli. Immunofluorescence staining of Spod-11-tox containing hemocytes withdrawn from S. frugiperda larvae 3 h post injection of GFP-expressing E. coli (10⁷ bacteria/larvae). Cells were immunolabelled with rabbit anti-Spod-11-tox revealed by TRITC-conjugate anti-rabbit. Bars = 5 µm. B) Immunoelectron microscopy of granulocytes withdrawn from S. frugiperda larvae 3 h post injection of Pichia pastoris (10⁶ yeast/larvae). Note the presence of heterogeneous bodies looking like secondary lysosomes (b), phagocytosed P. pastoris (c) and structured granules (e). Immunogold labelling using 10 nm colloidal gold-conjugated Protein A. Bars = 0.5 µm.

Figure 4. Time course analysis of Spod-11-tox expression in challenged S. frugiperda larvae (E. coli 10⁵ bacteria/larvae).

A) Western blot analyses performed with protein extracts from hemocytes (from 6 to 72 hours post infection). B) Comparison of Spod-11-tox content in hemocytes and plasma early post infection. Proteins were loaded (10 µg/lane) on a Tris-Tricine 10–20% gradient gel. The western blot was realized as in figure 1. Numbers on the right sides of the blots indicate the molecular weight (kDa) estimated using Precision Plus Protein^TM Standards from Bio-Rad. Un; unchallenged.

Figure 5. Purification procedure and properties of S. frugiperda hemolymph.

A) Schematic representation of the protocol used for the purification of the insect hemolymph. Proteins were eluted from the Sep-Pack C₁₈ cartridge with 10, 40, and 80% acétonitrile prepared in acidified water (TFA 0.05%), lyophilizated and reconstituted with MilliQ water. B) Western blot analysis of Spod-11-tox content in the elution fractions of Sep-Pack C₁₈ cartridge. Protein extracts were loaded (15 µg/lane) on a Tris-Tricine 16.5% gel and blotted on a PVDF membrane. The blot was Coomassie stained (upper panel) and probed with polyclonal anti-Spod-11-tox antibody (lower panel). Numbers on the right sides of the blots indicate the molecular weight (kDa) estimated using Precision Plus Protein^TM Standards from Bio-Rad. C) The eluted Sep-Pack fractions were tested for their antimicrobial activities by liquid growth inhibition assay against E. coli (CIP7624). Results are representative of three independents experiments.

Figure 6. Discrimination between antimicrobial properties and anti-Spod-11-tox immunoreactivity of the fractionated S. frugiperda hemolymph.

The 40% Sep-Pack fraction from the extraction of immunized S. frugiperda plasma was subjected to reversed-phase HPLC on an Interchim UPNEC25QS column using a 0–60% linear gradient (dashed line) of acidified acetonitrile over 80 min. Antimicrobial activity against E. coli SBS363 (black rectangles) M. luteus (grey rectangles) and F. oxysporum (white rectangles) was measured by liquid growth inhibition assays. Anti-Spod-11-tox immunoreactivity (hatched rectangles) was measured by dot-blot ELISA. Insets show the dissociation between immunoreactive and antimicrobial fractions over the course of the purification. a: at the first purification step; b: at the second purification step (same column with a 0-27-42% biphasic linear gradient of acidified acetonitrile over 5 and 40 min); c: at the third purification step (Xbridge BEH300 narrowbore column with a 0-23-43% biphasic gradient of acidified acetonitrile over 5 and 80 min). Absorbance (Abs) was monitored at 225 nm.