Some animals express a form of eusociality known as 'fortress defense', in which defense rather than brood care is the primary social act. Aphids are small plant-feeding insects, but like termites, some species express division of labor and castes of aggressive juvenile 'soldiers'. What is the functional basis of fortress defense eusociality in aphids? Previous work showed that the acquisition of venoms might be a key innovation in aphid social evolution. We show that the lethality of aphid soldiers derives in part from the induction of exaggerated immune responses in insects they attack. Comparisons between closely-related social and non-social species identified a number of secreted effector molecules that are candidates for immune modulation, including a convergently-recruited protease described in unrelated aphid species with venom-like functions. These results suggest that aphids are capable of antagonizing conserved features of the insect immune response, and provide new insights into the mechanisms underlying the evolution of fortress defense eusociality in aphids.
Degree of melanziation raw data
To visually evaluate and quantify melanization in larvae attacked by aphid soldiers, we used double-blind observational assays. We quantified the melanization phenotype before and after attack by aphid soldiers by using a subjective visual assay of melanization. We collected P. obesinymphae galls in the vicinity of Nashville, TN in July 2013. To facilitate observation, galls were split in half, and we placed individual Drosophila larvae, a surrogate for common Dipertan predators of galling aphids, in galls with aphids or in galls in which the aphids had been removed. Larvae remained in galls for 1 hr. Each D. melanogaster larvae was photographed before attack, immediately following attack, after removing the attacking soldiers, and 1 hr following attack. Only Drosophila larvae being attacked by at least 10 aphid soldiers were included in the study. Additionally, as controls, Drosophila larvae were stabbed either once or multiple times with a sterile needle. Larvae were then placed in an empty gall for 1 hr. Using a double-blind assay, all Drosophila larvae were assigned a score based on the proportion of the body melanized: 1 (no observable melanization) to 4 (surface completely melaninized). The sample size for each group was greater than or equal to 10.
Surviorship assay raw data
To examine if the over activation of the melanization pathway affected the survivorship of D. melanogaster following attack by aphid soldiers, we compared the survivorship of Drosophila mutants deficient in the melanization response to wild-type Drosophila. To measure survivorship, we introduced larvae to a gall, monitored the larvae every 20 minutes and noted the time to death, defined as when the larvae stopped moving.
qPCR raw data
To examine which genes were differential regulated following attack by aphid soldiers, we used quantitative PCR to assess expression levels of multiple immune genes in two experimental groups of D. melanogaster larvae: 1) Larvae attacked by P. obesinymphae for 1 hr; or 2) Larvae placed in an empty gall for 1 hr. One hour was chosen because it was prior to the average observed time of death caused by aphid soldiers (Lawson et al. 2014), but enough time to observe a melanization response. In order to provide a standardized reference and to establish whether our assays with aphids were sensitive enough to detect immunological activation, we also measured the expression of each these genes in Drosophila that were stabbed with sterile needles.
Phenoloxidase assay raw data
Phenoloxidase levels in Drosophila tissues were measured using a simple procedure based on the transformation of L-dopa to dopachrome in the presence of phenoloxidase, measured as function of optical density, as described by Lario et al. (1993). Larvae were singly placed in freshly collected P. obesinymphae galls for 45 minutes, and subsequently removed with a small paintbrush. A control group of larvae of similar size was not exposed to the aphids and was labeled the “unattacked” group. In order to differentiate between phenoloxidase activation during mechanical wounding of the epithelia by aphid stylets, and circulating phenoloxidase activation in the hemolymph, Drosophila hemolymph was extracted from both the attacked larvae and the control group using fine needles. After hemolymph extraction, the larval midguts were removed using a pair of dissecting tweezers. The remaining cuticle was then placed in a 1.5 mL Eppendorf tube and ground by a pestle. The tubes were centrifuged briefly, and the liquid portion was removed by a pipette for use in the phenoloxidase assay, in addition to the extracted hemolymph. Immediately after extraction, the hemolymph was transferred to ice-chilled ELISA wells containing 50 μL of 20 mM TRIS buffer, pH 7.0, with 1% sodium citrate. The cuticle fluids were transferred by pipette to ELISA wells containing the same mixture. After the hemolymph and cuticle fluids were collected, 50 μL of 3,4-dihydroxyphenyl-alanine (L-dopa), 5 mg/mL in TRIS buffer, were added to each well. Mixtures were incubated at 32°C for 45 min, and the wells were subsequently read in an ELISA spectrophotometer at a wavelength of 490 nm. Blank wells containing 50 μL TRIS buffer and 50 μL L-dopa in TRIS but no hemolymph or cuticle fluids were used as controls.
