An odorant-binding protein functions in fire ant social immunity interfacing with innate immunity

Abstract

Social immunity-mediated sanitation behaviours occur in insects when microbially killed corpses are removed and/or dismembered by healthy nestmates. However, little is known concerning the chemical signals or receptor proteins that mediate these responses. Here, we identify cuticular components in the eusocial red important fire ant, Solenopsis invicta: behenic acid, which induces dismemberment behaviour, and oleic and cis,cis-9,12-linoleic acids, which inhibit dismemberment in a process mediated by S. invicta odorant-binding protein-15 (SiOBP15). Yeast two-hybrid screening and protein-protein interaction analyses identified the ant immunity-related proteins apolipophorin-III (SiApoLp-III) and fatty acid binding protein-5 (SiFABP5) as SiOBP15 interacting partners. SiOBP15 and SiFABP5 bound all three dismemberment-related compounds, whereas interactions between SiOBP15 and SiApoLp-III narrowed binding to behenic acid. RNAi-mediated gene expression knockdown of SiOBP15, SiApoLp-III or SiFABP5 revealed that behenic acid chemoreception determines dismemberment behaviour via SiApoLp-III/SiOBP15, whereas SiOBP15 or SiOBP15/SiFABP5 recognition of linoleic acid inhibits dismemberment behaviour. These data identify a host circuit linking olfactory proteins and proteins involved in innate immunity to control the degree of sanitation behaviour elicited in response to microbial infection. We identify specific chemical cues transduced by these proteins, providing a mechanism connecting olfaction-related processes to innate immunity, host-pathogen interactions and social immunity.

Keywords: Solenopsis invicta; apolipophorin-III; fatty acid binding protein; immunity and olfaction; odorant-binding protein; sanitation behaviour.

Figure 1.

Solenopsis invicta social immunity sanitation behaviour: corpse-dismemberment. (A) Quantification of corpse-dismemberment by healthy S. invicta workers towards freeze-killed and B. bassiana-killed conspecifics. (B) Heatmap of levels of indicated fatty acids in freeze-killed (CK) and B. bassiana-killed (IF) workers quantified by GC–MS. Three independent analyses are shown for each condition. (C) Effects of various identified compounds on corpse-dismemberment of freeze-killed conspecifics. All experiments were performed with three independent biological replicates, each consisting of three technical replicates. Error bar = SE. Means followed by different letters are significantly different (Student’s t‐test or one-way ANOVA analysis of variance followed by Tukey’s post hoc test, p < 0.05).

Figure 2.

Ligand-binding specificity and contribution of SiOBP15 to corpse-dismemberment. (A) Dismemberment (%) of freeze-killed workers by untreated, GFP-RNAi and SiOBP15 RNAi treated workers. (B) Dismemberment (%) of B. bassiana-killed workers by untreated, GFP-RNAi and SiOBP15 RNAi treated workers. (C) Kinetics of binding of purified SiOBP15 to the fluorescent reporter 1-NPN. (D) Competition/displacement assays using 1-NPN as the fluorescent reporter ligand bound to purified SiOBP15 using eight identified S. invicta VOCs as competing ligands. All experiments were performed with three independent biological replicates, each consisting of three technical replicates. Error bar = SE. Different letter designations indicate significant difference (p < 0.05).

Figure 3.

Interactions between SiOBP15 and SiApoLp-III and SiFABP5. Pull-down assays using (A) SiOBP15-His and SiApoLp-III-GST and (B) SiOBP15-His and SiFABP5-GST. (C,D) Purified SiOBP15 was immobilized on chips and used for SPR analyses to determine binding kinetics to SiApoLp-III and SiFABP5, respectively. (E) Co-localization of gene expression in S. invicta tissues using three-colour FISH. Worker antennae were probed and visualization using anti-sense RNA corresponding to SiOBP15, SiApoLp-III and SiFABP5 labelled with green, red and pink fluorescence markers, respectively. Labelling of the same cells by anti-sense SiOBP15, SiApoLp-III and SiFABP5 probes are indicated by arrowheads. Co-localization hybridization signals appear as a yellow colour in the overlay (right) of the green, red and pink fluorescence channels. Scale bars = 20 µm.

Figure 4.

Contributions of SiOBP15, SiApoLp-III and SiFABP5 to dismemberment behaviour. Corpse-dismemberment (%) for untreated controls, and workers treated with RNAi targeting GFP (control), SiApoLp-III and SiFABP5 when presented with corpses of workers killed by freezing (A) or by B. bassiana infection (B). Corpse-dismemberment (%) for non-RNAi treated control (WT), and workers treated with RNAi targeting GFP, SiApoLp-III, SiFABP5 and both SiOBP15 + SiApoLp III (dual RNAi treatment) when presented with freeze-killed workers treated with behenic acid (C), linoleic acid (D), trans-11-eicosenoic acid (E) or oleic acid (F). All experiments were performed with three independent biological replicates, each consisting of three technical replicates. Error bar = SE. Different letter designations indicate significant difference (p < 0.05).

Figure 5.

Model for interactions and contributions of S. invicta OBP15, ApoLp-III and FABP5 + ligands to corpse-dismemberment sanitation behaviour.