Targeting an antimicrobial effector function in insect immunity as a pest control strategy

Abstract

Insect pests such as termites cause damages to crops and man-made structures estimated at over $30 billion per year, imposing a global challenge for the human economy. Here, we report a strategy for compromising insect immunity that might lead to the development of nontoxic, sustainable pest control methods. Gram-negative bacteria binding proteins (GNBPs) are critical for sensing pathogenic infection and triggering effector responses. We report that termite GNBP-2 (tGNBP-2) shows beta(1,3)-glucanase effector activity previously unknown in animal immunity and is a pleiotropic pattern recognition receptor and an antimicrobial effector protein. Termites incorporate this protein into the nest building material, where it functions as a nest-embedded sensor that cleaves and releases pathogenic components, priming termites for improved antimicrobial defense. By means of rational design, we present an inexpensive, nontoxic small molecule glycomimetic that blocks tGNBP-2, thus exposing termites in vivo to accelerated infection and death from specific and opportunistic pathogens. Such a molecule, introduced into building materials and agricultural methods, could protect valuable assets from insect pests.

Fig. 1.

Termites exhibit an unusual β(1,3)-glucanase activity. (A) Distance tree of insect GNBPs rooted with B. circulans β(1,3)-glucanase. Numbers depict bootstrap values for basal lineages. Glucanase motif sequences are adjacent to species name. Termite species appear in blue. Numbers following species name indicate GNBP-1, -2, etc. (B) β(1,3)-Glucanase activity in termite tissues and cuticular washes, assayed on Carboxymethyl Curdlan Remazol Brilliant Blue gels (lanes: 1, soldier; 2, large worker; 3, large worker; 4, 10 workers; 5–7, 1, 5, and 30 large workers, respectively) (Left). Comparison of activity in extracts from termites (lanes: 1, Nasutitermes corniger collected from Florida; 2–4, N. corniger collected from different colonies in Panama), Drosophila melanogaster and Galleria mellonella (Right). (C) Activity profile of fractionated termites separated by HPLC. (Inset) Cytotoxic effect of the peak active fraction (marked with an arrowhead) from termites on Metarhizium anisopliae conidia measured by flow cytometry as the effect on cell volume in femtoliters (*, P < 0.05 vs. buffer).

Fig. 2.

tGNBP-2 is an effector β(1,3)-glucanase induced by pathogenic patterns. (A) Immunoprecipitation of tGNBP-2 from termite extracts, nests, unprocessed wood, and soils A and B sampled from different locations. (B) β(1,3)-Glucanase activity of termite immunoprecipitate isolated by rabbit IgG or anti-tGNBP-2 (*, P < 0.05 vs. no protein and rabbit IgG). (C) Expression of tGNBP-2 on hemocyte surface, measured by flow cytometry. (D) Survival of termites infected with Metarhizium anisopliae (M. aniso) following immunization with LPS or laminarin at either 0.5 or 5 mg/mL along a course of 10 days (mean of 2 groups, n = 12/group; all immunized termite groups were P < 0.05 vs. M. aniso only, and P > 0.05 vs. naïve termites with the exception of 0.5 mg/mL LPS). (E) tGNBP-2 expression after exposure of termites to either LPS or laminarin at either 0.5 or 5 mg/mL (left and right lanes in each sample represent two independent repeats). (F) Expression of tGNBP-2 after termite infection with M. anisopliae (M. aniso) or Serratia marcescens (S. marc). Cooled only (cooled) or cooled and pricked (pricked) termites are shown as controls. (G) Effect of crude termite extract on conidial growth (in CFUs) on potato-dextrose agar plates (*, P < 0.05 vs. untreated conidia) (Left). Cytotoxicity map of fractionated termites on M. anisopliae conidia measured by cell volume in femtoliters (Right). F(x),F(y) represent fractions 1–13 on either axis. The tGNBP-2⁺ fraction is marked by an arrowhead; asterisks mark putative antimicrobial peptides. (H) Loss of cytotoxicity after depletion of tGNBP-2 by antibody precipitation. F(x),F(y) represent fractions 1–20 on either axis.

Fig. 3.

GDL blocks pattern recognition and signaling by tGNBP-2. (A) Binding of tGNBP-2 to pathogenic bacteria and fungi (Serratia marcescens, Candida albicans, Aspergillus fumigatus, Metarhizium anisopliae variants 1 and 2, putative M. anisopliae isolated from a dead insect, and Escherichia coli) compared with recombinant human CD14 (rhCD14). IgG, isotype controls; tGNBP-2, tGNBP-2 signal; rhCD14, recombinant human CD14 signal. (B) A computational model of tGNBP-2 (gray) with side chains interacting with β(1,3)-d-glucan and eritoran (an LPS analog) colored as follows: red, acidic; brown, hydropohobic; blue, basic; light blue, Asn/Gln; magenta, Tyr. β(1,3)-d-glucan C atoms are green; eritoran C atoms are cyan. In both, O atoms are red, P atoms orange, and N atoms blue. (C) Dose–response inhibition of β(1,3)-glucanase activity of tGNBP-2, termite lysate, or tGNBP-2-depleted lysate by GDL. (Inset) Zymogram of termite extract with GDL or a d-glucose control. (D) Inhibition of SAPK/JNK signaling in hemocytes by suppression of tGNBP-2 measured by flow cytometry (G, granulocytes; M, monocytes; GDL, d-δ-gluconolactone; n = 2,000) (Left). GDL (20 mM) blocks induction of tGNBP-2 by laminarin challenge in termite cells, measured by Western blot (Right).

Fig. 4.

GDL suppresses termite antimicrobial defense. (A) Effect of gain or loss of tGNBP-2 function on termites infected with Metarhizium anisopliae along the course of 12 days (GDL, d-δ-gluconolactone; mean of 2 groups, n = 12/group; *, P < 0.05 tGNBP-2 treated conidia vs. other groups, P < 0.05 GDL treated termites with untreated conidia vs. other groups, n.s. no significant difference vs. no infection). (B) Postmortem analysis of dead termites from GDL treatment groups representing percentage of termites confirmed to have been infected by microbial pathogens. Numbers above bars represent daily death count.