Study Organism and Experimental Design

The green veined white butterfly (Pieris napi) is a widespread generalist butterfly. It occurs throughout Europe and in the temperate zone of Asia (GBIF, 2017). For this study, female butterflies were collected in northern Sweden (Abisko township) and Southern Sweden (Kullaberg park, Skåne) in August 2014. These females were transferred to Stockholm University where they were allowed to lay eggs on Garlic mustard (Alliaria petiolata). Larvae from wild-caught females were fed on A. petiolata leaves until pupation in a climate-controlled room (Light:Dark 12:12 h, 17°C). Pupated offspring were placed in cold conditions (4°C) 21 days after pupation.

In order to test the effects of infection with gram-positive or gram-negative bacteria on survival, pupae from Abisko were taken out of diapause in April 2014 and placed in a climate-controlled room (L:D 23:1, 23°C). Unrelated males and females each received a unique identifier before release into the mating cage, and were fed ad lib on 20% sugar solution. Adults were observed every hour to ensure parentage. Once mated, females were placed in individual cups with A. petiolata for oviposition. The leaves were exchanged twice a day, until females stopped laying eggs. The offspring from four females that produced the highest number of eggs were chosen for the experiment. The eggs were kept in containers and placed in climate chambers to develop, and grown under diapausing conditions (L:D 8:16, 17°C). After reaching third instar, larvae were moved to individual cups containing A. petiolata and checked daily to monitor development.

Once larvae reached the second day of 5th instar they were sexed and randomly divided among 8 treatment groups (Supplementary Figure 1). One treatment was injected with 10 μl of sterilized phosphate-buffered saline (PBS), to act as a trauma control. To investigate the effect of different doses of bacteria; three treatment-groups were injected with the live gram-negative E. coli (10⁴, 10⁵, and 10⁶), another three treatment-groups were injected with the gram-positive M. luteus (10⁴, 10⁵, and 10⁶), and the final treatment-group was left as uninjected controls. Details described below. Larvae were weighed to the nearest 0.1 mg and afterward anesthetized by chilling them in containers on ice for 5 min prior to injection. The syringe needle (Hamilton SYR 10 μL 701 ASN) was sterilized by rinsing three times each in two tubes of 95% ethanol, followed by one tube of sterile H₂O. The injection was done at an angle less then 45° behind the hind abdominal proleg, which was sterilized with a 95% ethanol swab beforehand (Hussa and Goodrich-Blair, 2012). Control individuals were weighed and anesthetized without injection. Larval survival was monitored twice daily, until all surviving individuals reached pupation.

For the RNA-seq experiment, larvae from Skåne, southern Sweden (Kullaberg; 56°18′N, 12°27′E 10⁹), from the same stock as Lehmann et al., 2017, were taken out of diapause and reared in the identical conditions as above. The injection treatment was identical as above, the only deviation being the treatments, instead of eight, there were only three treatment groups: PBS, E. coli 10⁶, M. luteus 10⁶. Larva were sampled at 3, 6, 12, and, 24 h after injection. Individuals were sampled by placing them in a 1.5 mL tube, and submerged into liquid nitrogen after which they were stored in −80°C.