Drosophila were cohabitated with female wasps for four days, then separated. Flies were then placed into embryo collection chambers for 24 hr. Embryos were divided into two cohorts and each developed in the absence of adult flies or wasps. One cohort was used to propagate the next generation and never exposed to ethanol food; the second cohort was used in the ethanol preference assay and then discarded (Figure 1a).

Figure 1 with 4 supplements see all

Download asset Open asset

Wasp-exposed Canton-S flies lay approximately 94% of their eggs on ethanol food (Figure 1b). This behavior persists in their offspring despite the F₁ generation never having direct interaction with wasps (Figure 1b). Ethanol preference in F₁ was less potent, with 73% of the eggs laid on ethanol food (p=8.6e⁻⁷, Supplementary file 1). Remarkably, this inherited ethanol preference persisted for five generations, gradually reverting back to the mock exposed baseline (Figure 1b). These observations were replicated in an additional wild type Oregon R (OR) strain (Supplementary file 2), suggesting that the phenomenon is not specific to a particular genetic aberration or background. This indicates that inheritance of ethanol preference is not a permanent germline change, but rather it is a reversible trait. Ethanol preference was measured for two days for initial experiments (Figures 1–3 and Figure 1—figure supplement 1); day one and day two showed similar trends, suggesting that flies do not habituate to ethanol, nor does the preference fade over the course of the experiment (Supplementary file 3).

Figure 2

Download asset Open asset

Figure 3 with 1 supplement see all

Download asset Open asset

To explore the required neural signaling to the germline, mutants defective for long-term memory were assayed. Previous studies have shown that flies defective in long-term memory exhibit an ethanol preference only in the presence of wasps, but not after wasp removal. The long-term memory mutant Orb2^ΔQ produced offspring with an ethanol preference when the embryo collection for the F₁ generation was conducted in the presence of wasps, but this ethanol preference was greatly reduced in offspring collected post-wasp exposure (Figure 1c). Similar results were found for the memory mutant amn¹ (Figure 1—figure supplement 2a). To further probe the role of long-term memory, a conditional knockdown of Orb2 was employed: In this system, upon ingestion of the drug RU486, a RNA hairpin to Orb2 is expressed in the mushroom body, a region of the brain essential for long-term memory (Bozler et al., 2017; Roman et al., 2001). These experiments provide the advantage of generating offspring with wild type long-term memory and avoid the possible developmental complications of mutant lines. Legacy flies collected during the wasp exposure and subsequently tested did exhibit an ethanol preference (Figure 1—figure supplement 1b). However, legacy flies from the parental RU486 treatment collected post-wasp exposure did not display an ethanol preference. These data provide insight in two ways: First, functional long-term memory is not a compulsory requirement to generate ethanol-preferring offspring. Second, the mutant data suggests that intact long-term memory is not required to inherit ethanol preference. Given that ethanol preference in the absence of wasps is long-term memory dependent, this experiment reveals that the neuronal signaling is different for maintained ethanol preference in the F₀ and F₁ flies (Bozler et al., 2017).

We confirmed previous findings, where, following a wasp exposure, F₀ flies have an ethanol preference that persists for more than a week, returning to baseline after ten days (Figure 1—figure supplement 1a). Sister cohorts of F₁ flies were collected at two time points along this F₀ ethanol preference decay; one immediately following wasp exposure (brood 1), and a second, ten days post wasp exposure (brood 2). Brood two did not display an inherited ethanol preference, suggesting that wasp exposure does not inflict a permanent change in the F₀ germline (Figure 1—figure supplement 1d). Again, these findings replicated in OR flies (Supplementary file 2), indicating that these observations are robust and not dependent on the context of a particular genetic background.

To explore further the role of time and dynamics of wasp exposure, multiple generations of flies were exposed to wasps. We found that inherited ethanol preference can be enhanced with successive generations of wasp exposure (Figure 1—figure supplement 1e). This trend did not repeat when nonconsecutive generations were repeatedly exposed to wasps (Figure 1—figure supplement 1f). This suggests that the enhancing effect observed in the successive exposures is time sensitive and may be linked to the ethanol preference of the parental flies.

Several other factors point to distinctions between the F₀ and F₁ ethanol preference behavior. Male F₁ legacy flies, mated to naïve females produced offspring (F₂) with an ethanol preference (Figure 1—figure supplement 3a). Additionally, 14–16 day old F₁ flies displayed an ethanol preference, demonstrating that F₁ flies do not have an ethanol preference decay curve similar to that of the F₀ (Figure 1—figure supplement 3b).

Global transcriptional changes in the female head across generations were examined with RNA sequencing. Heads from the exposed legacy F₁ and F₂ generation were collected and normalized to their respective unexposed legacy group. These results were then compared with the F₀ generation, which was previously reported (Bozler et al., 2017). Analysis of the F₀ data detected 98 differentially expressed transcripts (15 down and 83 up) (Figure 1—figure supplement 4, Supplementary file 4). F₁ and F₂ heads showed very few differentially expressed transcripts, 4 and 5 transcripts respectively. Of the differentially expressed transcripts, no transcript was shared between groups. These data indicate that although wasp exposure itself results in global transcriptional changes in the female head, this observation does not hold true for the subsequent generations.

Mid-oogenesis germline apoptosis (stage 7–8 oocytes) is triggered upon wasp exposure (Figure 2a) (Kacsoh et al., 2018b; Kacsoh et al., 2018a; Kacsoh et al., 2015). However, this wasp response is not heritable like the ethanol preference behavior, and F₁ females do not exhibit germline apoptosis (Figure 2a). Nevertheless, maternal germline knockdown of the effector caspases Dcp-1 and drice (maternal-αtubulin-Gal4 > UAS-Dcp-1[RNAi], and maternal-αtubulin-Gal4 > UAS-Drice[RNAi] respectively) produce offspring without an ethanol preference, regardless of parental treatment (Figure 2c). Although protein-starvation triggers germline apoptosis similar to wasp exposure (Figure 2b), offspring from mothers with starvation-induced apoptosis do not inherit an ethanol preference (Figure 2d). This indicates that germline apoptosis in and of itself is not sufficient for inheritance of ethanol preference.

NPF is known to play a role in food seeking, ethanol consumption, and numerous other reward pathways, and NPF levels decrease in the fan shaped body of female fly brains following wasp exposure (Kacsoh et al., 2013). Even in the presence of wasps, overexpression of NPF (NPF-Gal4 >UAS NPF) inhibits ethanol preference, while in the absence of wasps, knockdown of NPF (NPF-Gal4 >UAS-NPF[RNAi]) is sufficient to induce the ethanol preference behavior (Figure 3a). Given this NPF modulation of ethanol preference in females, we asked whether NPF also signaled to germline cells, triggering caspases and apoptosis. Strikingly, NPF knockdown induces mid-oogenesis apoptosis in the absence of wasps (Figure 3b), while overexpression of NPF results in no elevation in germline apoptosis even in the presence of wasps (Figure 3b). Similarly, NPF-receptor (NPFR) knockdown in neurons (Elav-Gal4 >UAS-NPFR[RNAi]) alone leads to significantly elevated levels of apoptosis (28%, when compared to parental line controls p=6.2e⁻⁴ and 1.5e⁻⁴), and this effect is enhanced with wasp exposure (61%, p=7.0e⁻⁴) (Figure 3c). Taken together these observations link ethanol preference behavior and mid-oogenesis apoptosis in the F₀ females, where both processes are likely caused by changes in NPF and NPFR signaling.

We speculated that the NPF-triggered changes in F₀ behavior and germline might also correlate with observed changes in offspring. For these experiments, it was critical to ensure that F₁ flies did not share the maternal genotype, and a crossing scheme was devised to avoid progeny with transgene expression (see Materials and method section). Legacy flies from mothers with NPF knockdown exhibit ethanol preference, even in the absence of wasp exposure (Figure 3d and Figure 3—figure supplement 1). Inherited ethanol preference is enhanced when the parental NPF knockdown flies are exposed to wasps (Figure 3d). By contrast, NPF overexpression in F₀ mothers exposed to wasps produced offspring lacking the ethanol preference (Figure 3d). NPFR knockdown experiments mirror these findings: Maternal NPFR knockdown produces offspring with an ethanol preference compared to unexposed control lines; again this effect is enhanced when NPFR knockdown is paired with wasp exposure (Figure 3e). Interestingly, overexpression of NPF in F₁ flies blocks ethanol preference in the exposed F₁ legacy group (Figure 3f), raising the possibility that F₁ legacy flies inherit NPF in a repressed or low expression state.

We therefore hypothesized that regulation or depression of NPF might be a means of this behavioral inheritance. Global changes in NPF RNA were not detected in either the F₀ or F₁ female heads (Figure 4—figure supplement 1). However, antibody staining allowed for a region specific examination of NPF protein levels (Figure 4a and (Figure 4—figure supplement 2). Anti-NPF signal has clear overlap with the NPF-Gal4 driving the cd8-GFP reporter (Figure 4a). The fan shaped body has previously been implicated in ethanol preference, and therefore, was a focus in this experiment (Kacsoh et al., 2013). NPF protein levels measured through immunofluorescence were significantly reduced in the fan shaped body of F₀, F₁, and F₂ (two-generations exposed) flies (Figure 4b). We note that NPF was not reduced in all regions of the F₁ and F₂ brains, as intensity of P1 neurons was not reduced in either the F₁ or F₂ flies, although significant reduction was observed in P1 neurons of F₀ flies (Figure 4c).

Figure 4 with 2 supplements see all

Download asset Open asset

Given the observed link between depressed NPF and oocyte apoptosis, it is notable that F₁ flies do not have germline apoptosis. It is possible that apoptosis is due to a localized decrease in NPF not shared between the two generations; perhaps the apoptosis is triggered by other NPF neurons or synapses. It is also conceivable that other neural processes are altered in the flies that we did not detect, decoupling the apoptosis and ethanol preference behaviors in the later generations.

To determine whether maternal or paternal exposure were equally important for transgenerational inheritance of ethanol preference, wasp-exposure and mating were controlled in two separate experiments. First, mated females were exposed to wasps in the absence of male flies. Second, wasp-exposed males were mated to naïve virgin females, removing the maternal exposure as a factor. Interestingly, F₁ offspring from an exclusively maternal wasp exposure inherit ethanol preference, while F₁ offspring from an exclusively paternal exposure did not (Figure 5a). Importantly, this demonstrates a key difference between the F₀ males and the F₁ males, and it may point to an ‘activation phase’ and ‘maintenance phase’ of the epigenetic program. We have previously reported that female flies require sight to induce a behavioral response to wasp exposure (Kacsoh et al., 2015). In further support of the maternal contribution to the inheritance of ethanol preference, blind female flies (ninaB¹; UAS-ninaB) mated to CS males did not produce offspring with an ethanol preference (Figure 5b). In the reciprocal experiment, blind fathers mated to CS females did generate ethanol-preferring offspring following a wasp exposure (Figure 5b).

Figure 5

Download asset Open asset

The maternal epigenetic inheritance of ethanol preference could be conferred by chromosomal elements and/or cytoplasmic factors. If ethanol preference is inherited through a chromatin mark, then chromosome parental-origin tests should reveal a requirement for maternal chromosomal inheritance. However, if inheritance is conferred through a maternally contributed cytoplasmic factor, then passage of all chromosomes through paternal gametes should have no effect, as wasp-exposed females can still maternally deposit molecules and organelles into the oocyte. To test what maternal components may be conferring inheritance of ethanol preference we first focused on chromosomal elements using attached, or compound, chromosomes. Flies where each of the two homologs are fused cannot make haplo-chromosome gametes. Instead, they can only make gametes with one or zero copies of the fused chromosome, and therefore F₁ flies inherit ‘pairs’ of homologs that are entirely maternally or paternally derived (Figure 5c). In this manner, we tested each of the two major autosomes for parent-of-origin effects. Using phenotypic markers, flies were sorted as having either a maternal or paternal exclusive homolog pair and assayed for ethanol preference. Chromosome-II fusion flies had similar results when inheriting exclusively maternal or paternal Chromosome-II elements (Figure 5c). Chromosome-III fusion flies also had inheritance of ethanol preference when receiving both copies of Chromosome-III maternally. However, flies with both copies of Chromosome-III from their fathers failed to inherit an ethanol preference (Figure 5d). This observation has at least three implications: Most importantly, this indicates that some element on Chromosome-III must be inherited from wasp-exposed mothers in order for ethanol preference behavior to be passed on to F₁ legacy flies. This also suggests that maternal copies of the Chromosome-X, Chromosome-II or cytoplasmic factors, if important, are not sufficient for inheritance of ethanol preference. Lastly, that oocytes giving rise to eggs with zero copies of maternal Chromosome-II still confer ethanol preference, indicates that exclusion of maternal chromosomes itself does not generally interfere with transgenerational inheritance.

To further delineate what parts of maternally derived Chromosome-III were required for transgenerational inheritance of ethanol preference we tested chromosomes with well defined deletions. As NPF has previously been shown to control ethanol preference behavior, we speculated that the NPF locus on Chromosome-III may be a target of maternal epigenetic reprogramming (Shohat-Ophir et al., 2012). We also observed that F₁ legacy flies inherit low levels of NPF expression specifically in the fan shaped body of the brain (Figure 4a–b), consistent with the possibility that F₁ flies inherit repressed NPF expression. If the critical maternal Chromosome-III element is the NPF gene locus, then F₁ offspring with maternal deletions of this chromosomal region may prevent inheritance of ethanol preference, much like not having inherited any maternal copies of Chromosome-III (Figure 5d). Using females with one Chromosome-III carrying a large deletion of the NPF gene region and one copy of wild-type NPF on a balancer Chromosome-III allowed us to ask whether an intact maternal NPF gene region was necessary for F₁ inheritance of ethanol preference. We found that legacy F₁ flies from unexposed mothers had no preference for ethanol, regardless of whether they inherited an intact NPF gene on a balancer chromosome or a chromosomal deletion of the NPF region (Figure 5e–f). Legacy F₁ flies from exposed mothers inheriting a wild-type NPF on a balancer chromosome exhibited a strong preference for ethanol, suggesting that multiple rearrangements, deletions and mutations of a balancer Chromosome-III are not sufficient to prevent ethanol preference in F₁ flies. By contrast, legacy F₁ flies from exposed mothers inheriting a Chromosome-III deletion of the NPF gene region do not inherit any preference for ethanol (Figure 5f). This was true for two different Chromosome-III deletions at the NPF locus, whereas a Chromosome-III deletion that does not disrupt the NPF gene had no effect (Figure 5f). Paternally inherited Chromosome-III deletions were not sufficient to prevent ethanol preference in F₁ flies (Figure 5f).

Flies were maintained at room temperature on standard cornmeal-molasses media. A list of fly lines and genotypes used is reported in Supplementary file 5. Female flies were considered mature adults at three to five days post eclosion. Flies outside of this age range were not used for experimentation unless specifically noted, as for example in Figure 1—figure supplement 1d. Experiments involving manipulation of the maternal genotype, such as the maternal NPF knockdown, had a crossing scheme to avoid transgene expression in the F₁ generation. Virgin females with the genotype of interest were crossed to y,w males and offspring were scored by eye color to ensure that flies assayed were not carrying both the Gal4 and UAS constructs.

The Figitid larval endoparasitoid Leptopilina heterotoma (strain Lh14) was used. In order to propagate wasp stocks, we used adult D. virilis in batches of 40 females and 15 males per each vial (Genesse catalog number 32–116). Adult D. virilis were allowed to lay eggs in standard Drosophila vials containing 5 mL standard Drosophila media supplemented with live yeast (approximately 25 granules) for one week before being replaced by adult wasps, using 15 female and six male wasps, for infections. Prior to wasp addition, vials were supplemented with approximately 500 µL of a 50% honey/water solution applied to the inside of the cotton vial plugs. Organic honey was used as a supplement. Wasps aged 3–7 days post eclosion were used for all infections and experiments, and were never reused for experiments.

Mature adult flies were used for wasp exposures: 40 female flies, 10 male flies, and 20 female Lh14 (Leptopilina heterotoma) wasps were placed in a vial with cornmeal-molasses media. This cohabitation (wasp exposure period) lasted for four days. The unexposed control consisted of the 40 female flies and 10 male flies with no wasp cohabitation. Both treatment groups were maintained at room temperature (approximately 22° C) with a 12 hr light-dark cycle for the duration of the exposure period.

At the conclusion of the exposure period, flies were separated into two cohorts. Following the removal of all wasps, one group of flies was used to propagate the next generation, while the second group was assayed for ethanol preference. Group one was placed on molasses-based embryo collection plates, supplemented with yeast paste, for egg collection. The collection period lasted for 24 hr, at which point the adult flies were removed. First instar larvae were transferred from these embryo plates to standard media vials. Larvae were density controlled to approximately 40 larvae per vial.

The second group was assayed for ethanol preference using a food-choice assay (Kacsoh et al., 2015). Briefly, five female flies and one male fly were placed into a modified petri dish with mesh top, termed the ‘fly corral’. Two food sources were placed at opposite ends of the ‘fly corral’. Each food source consisted of 0.45 g of instant drosophila media, hydrated with 2 mL liquid. Control food was hydrated entirely with distilled water, where as ethanol food was prepared with distilled water and a final addition of 95% ethanol to the top of the prepared food, creating a food with 6% ethanol by volume. Food sources were removed and replaced after 24 hr. Figures report the egg laying behavior of the first 24 hr interval unless otherwise noted. Total number of eggs laid on each food source was counted in a blinded fashion with treatment unknown to the counter. These egg counts are reported as a proportion of eggs laid on ethanol food. Flies that encountered ethanol-containing food were excluded from additional experimentation or lineage propagation. Fly corral experiments had ten replicates (cages) per condition.

Legacy flies, those descending from either the unexposed or exposed treatment, were divided into cohorts as described above for behavioral assay or embryo collection. These flies were not re-exposed to wasps except in the instance of multigenerational exposure experiments. Two experiments were conducted that involved multiple generations of treatment. For the successive exposures, three groups of flies were assayed; exposed legacy (two generations), exposed legacy (one generation), and unexposed legacy. In this instance, the exposed legacy (two generations) group was generated by subjecting F₁ exposed legacy flies to an additional round of wasp exposures. These flies therefore had grandparental and parental wasp exposure. Exposed legacy (one generation) had parental wasp exposure only (Figure 1—figure supplement 1B). It is important to note that the parents of the ‘exposed legacy (one generation)’ flies were F₁ unexposed legacy flies, and therefore had the same density control and egg collection as the other groups for the multigenerational duration of the experiment.

It is critical to note that baseline ethanol preference is highly variable depending on environmental conditions. Key factors are temperature and humidity, all ethanol oviposition assays were conducted in an environmentally controlled room at 25°C, approximately 30% humidity (±10%) with overhead lighting and a 12 hr light/dark cycle. Despite these controls, baseline ethanol preference varies day-to-day. For this reason, all groups for direct comparison (used in statistical tests) were tested at the same time.

Pertaining to the nonconsecutive exposure experiments; again three groups were assayed, the exposed legacy F₈ (two generations), exposed legacy (one generation), and the unexposed legacy. For these experiments, the exposed legacy F₈ (two generations) group was created by subjecting F₇-exposed legacy flies to an additional round of wasp exposures. These flies had a six-generation gap between ancestral wasp exposures. Flies in the exposed legacy (one generation) group were produced by exposing F₇ unexposed legacy flies to wasps, and collecting the subsequent offspring.

Several experiment specific modifications were made to the methods described above. To parse the maternal and paternal contributions to the inheritance of ethanol preference, two experiments were conducted. First, 40 mated female flies were used for wasp exposure, in the absence of males. Ten males were added to the population for the embryo collection period. For paternal contribution, male flies were removed from the exposure chamber and mated to unexposed virgin females. To test the role of vision in maternal inheritance, blind female flies mutant in ninaB[1], were crossed to wild type (CS) males. The reciprocal experiment crossed ninaB[1] males to CS female. These experiments were run in parallel and wasp exposures were preformed as previously described.

Compound chromosome experiments crossed two fusion stocks together (either chromosome-II or chromosome-III). The fusion lines retained phenotypic markers, and offspring with maternal or paternal chromosomes were sorted accordingly. Deficiency lines were crossed to CS flies and the genotype of the offspring (balancer or deficiency) was inferred from phenotypic markers.

Particular modifications for the Orb2^ΔQ memory-mutant experiments included an extra day of embryo collection. Following three-days of wasp exposure, flies and wasps were moved to the embryo collection chamber for the final treatment day. Eggs were collected for 24 hr in the presence of the 20 female Lh14 wasps. At the end of this period, wasps were removed and a new embryo collection plate was introduced for the second day of embryo collections. This second day of collection corresponds to the standard embryo collection timeframe in the above-described experiments. F₁ flies had the same genotype as the parental line.

Conditional knockdown of Orb2 was performed as previously described (Bozler et al., 2017). Briefly, flies were bred to have the following transgenes: mushroom-body-Gal4(switch) >UAS-Orb2[RNAi]. The Gal4(switch) driver line contains a Gal4 transcription factor that is active only in the presence of the drug RU486. Vials were prepared with instant food containing two grams of instant drosophila food, hydrated with 8 mL of either RU486 (0.22 mg/mL) in 5% methanol, or vehicle only (5% methanol). Flies and wasps were cohabitated in the prepared vials and were transferred to new food each day regardless of treatment for three days. On the fourth day, an embryo collection was performed on instant food supplemented with yeast (F₁ collection during wasp exposure). The fifth day the wasps were removed and a second egg lay was conducted (F₁ collection post wasp exposure). F1 legacy flies were transferred and raised on cornmeal media, without the addition of vehicle or RU486 and therefore should have fully intact mushroom body and long-term memory capacity.

Sibling cohorts were collected to assess the longevity of the germline change. ‘Brood 1’ flies were collected in the 24 hr immediately following the removal of the wasps. ‘Brood 2’ flies were collected from the same parents, 10 days after the termination of the wasp exposure.

Finally, diet restriction experiments had two groups: one with high protein and the other low protein diets. Low protein flies were maintained on molasses based embryo plates. The high protein group was maintained in similar fashion, but with the addition of yeast paste. High/low diet was maintained for four days prior to embryo collection.

Following the treatment period, ovaries were dissected and fixed in 4% formaldehyde for 30 min. Samples were stained with DAPI and apoptosis was scored based on the morphology of the nurse cell DNA. A researcher blinded to the genotype and treatment group of the samples performed the scoring. At a minimum, 15 ovaries were scored across three replicates (independent wasp exposures) for each group.

Antibody to neuropeptide F was generated in a rabbit to the full length NPF peptide: C-Ahx-SNSRPPRKNDVNTMADAYKFLQDLDTYYGDRARVRFamide. The antibody was subsequently purified using a truncated peptide containing the first 28 amino acids of NPF. Following purification, the antibody was depleted using a peptide of the eight amino acid C-terminal tail, shared by many neuropeptides. All peptide synthesis, antigen injection, serum preparation, peptide purification, and depletions were performed by 21^st Century Biochemicals.

Whole flies were fixed in 4% formaldehyde overnight at 4° C. Female brains were dissected, blocked, and incubated with anti-NPF (1:1000) overnight at 4° C. Antibody solution was removed and samples were blocked before the addition of the secondary antibody, anti-rabbit 488 (1:200), at room temperature for two-hours. Samples were counter stained with DAPI.

For NPF quantification, flies expressing a RFP tagged histone were dissected along with treatment groups and stained in the same solution. Pixel intensity of the fan shaped body (FSB) was measured in Image J. The FSB was outlined by hand and intensity measured. A background measure was made of the region immediately ventral to the FSB, with the same total area as the outlined FSB. The background value was subtracted from FSB measurement. Finally, the background-adjusted intensity value for each brain was divided by the arc length of its’ FSB. This process was repeated for each treatment group and the corresponding histone-RFP flies. These values were normalized to the histone-RFP flies to serve as a control for batch specific variation in staining. Each treatment group was normalized to the unexposed average of that replicate using the formula(s):

Standard fluorescent images were visualized with the Nikon Eclipse E800 microscope and a Olympus DP71 camera. For each experiment, wasp exposure and staining were performed on two separate occasions and final data was pooled after checking for the absence of a batch effect. A minimum of 10 brains were dissected for each treatment replicate as well as RFP-histone co-staining brains. Final quantified sample size range from 15 to 20 (normalized brains), due to sample loss or damage. Imaged samples were only excluded if clear damage or trauma (from dissection or staining process) was evident in the region of interest (FSB or P1 nuerons).

Mature female flies were anesthetized with CO₂ and collected in 15 mL conical tubes, either immediately following the treatment period (F₀), or 3–5 days post eclosion (F₁-F₂). Flies were frozen in liquid nitrogen and briefly vortexed to separate whole heads. Approximately 100 heads were collected for each replicate. A miRNeasy Kit (Qiagen) with on-column DNase treatment was used for RNA isolation. Four samples of each treatment group were prepared.

RNA samples were depleted of rRNA followed by random priming. Minimum sequencing depth per sample was 40 million paired-end reads on the Illumina platform. Sequencing reads were indexed to transcripts using Kallisto and the Ensembl genome (BDGP6) with 100 bootstraps (Aken et al., 2017; Bray et al., 2016). Downstream processing and statistical analyses used Sleuth (Pimentel et al., 2017). Heat maps were generated using hierarchical clustering and the R package pheatmap.

NPF transcript was measured by qPCR (SYBR Green, Thermo-Fisher 4309155). NPF primer targeted mRNA (TCCTGGTTGCCTGTGTGG, TCAGCCATAGTGTTGACATCG). Actin served as the control gene (CGCAAGGATCTGTATGCCAA, ACGGAGTACTTGCGCTCTGG). Fold change was calculated using the delta-delta Ct method.

Statistical tests were run in R (3.0.2 version, ‘Frisbee Sailing’). P-values for egg count data, NPF staining, and apoptosis quantification, were produced by applying a Mann-Whitney Rank Sum test. Error bars presented in the egg count ethanol preference graphs are bootstrap confidence intervals, generated using the boot package.

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

1. Received: January 21, 2019
2. Accepted: May 22, 2019
3. Version of Record published: July 9, 2019 (version 1)
4. Version of Record updated: July 9, 2019 (version 2)

© 2019, Bozler et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.