Bradysia (Sciara) coprophila larvae up-regulate DNA repair pathways and down-regulate developmental regulators in response to ionizing radiation

Abstract

The level of resistance to radiation and the developmental and molecular responses can vary between species, and even between developmental stages of one species. For flies (order: Diptera), prior studies concluded that the fungus gnat Bradysia (Sciara) coprophila (sub-order: Nematocera) is more resistant to irradiation-induced mutations that cause visible phenotypes than the fruit fly Drosophila melanogaster (sub-order: Brachycera). Therefore, we characterized the effects of and level of resistance to ionizing radiation on B. coprophila throughout its life cycle. Our data show that B. coprophila embryos are highly sensitive to even low doses of gamma-irradiation, whereas late-stage larvae can tolerate up to 80 Gy (compared to 40 Gy for D. melanogaster) and still retain their ability to develop to adulthood, though with a developmental delay. To survey the genes involved in the early transcriptional response to irradiation of B. coprophila larvae, we compared larval RNA-seq profiles with and without radiation treatment. The up-regulated genes were enriched for DNA damage response genes, including those involved in DNA repair, cell cycle arrest, and apoptosis, whereas the down-regulated genes were enriched for developmental regulators, consistent with the developmental delay of irradiated larvae. Interestingly, members of the PARP and AGO families were highly up-regulated in the B. coprophila radiation response. We compared the transcriptome responses in B. coprophila to the transcriptome responses in D. melanogaster from 3 previous studies: whereas pathway responses are highly conserved, specific gene responses are less so. Our study lays the groundwork for future work on the radiation responses in Diptera.

Introduction

Maintenance of genome integrity is central to an organism's survival and its ability to faithfully pass genetic information to its offspring. DNA mutation and loss of genome stability can result from errors in endogenous cellular events such as DNA replication and chromosome segregation, and from exposure to exogenous environmental agents that can alter or damage DNA. Living systems have therefore evolved various biochemical and developmental pathways that recognize and respond to errors and/or damage in their genetic code (Clay and Fox 2021). In humans, altered function in these pathways is a common feature in the development and progression of cancer, allowing the accumulation of further mutations in a cancer lineage (Hanscom and McVey 2020). A better understanding of diverse responses to DNA damage may therefore be helpful in developing novel approaches to disease.

Ionizing radiation, including X- and gamma-rays, is an abundant and well-studied DNA damaging agent. Ionizing radiation can directly interact with DNA, causing lesions to individual bases, single-strand breaks (SSBs), and double-strand breaks (DSBs) (Han and Yu 2010). Alternatively, radiation can interact with other molecules in the cell to generate reactive oxygen species (ROS), which can themselves cause DNA damage including single base lesions and SSBs. In metazoans, DNA damage can be repaired by several well-characterized pathways, including base excision repair (BER), nucleotide-excision repair (NER), mismatch repair (MMR), homology directed repair (HDR), and nonhomologous end-joining (NHEJ) (Chatterjee and Walker 2017). In addition, extensive DNA damage can trigger apoptotic pathways to remove cells with potentially unchecked genome instability. Thus, radiation survival involves DNA repair, cell cycle delay, and apoptosis (Su and Jaklevic 2001; Jaklevic and Su 2004; Jaklevic et al. 2006).

While all living systems appear to have some ability to respond to damaged DNA, several species have evolved highly robust responses to ionizing radiation that stem from expanded DNA repair or DNA protection mechanisms. For example, the extremophile bacterium Deinococcus radiodurans can endure doses of gamma-rays several orders of magnitude higher than a typical mammal, in part due to their ability to reassemble highly fragmented genomic DNA into an intact chromosome via specialized synthesis and recombination pathways (Bentchikou et al. 2010). Similarly, tardigrades are highly resistant to extreme conditions such as desiccation and exposure to radiation, and several species encode a tardigrade-specific nuclear protein called Damage Suppressor (Dsup) that binds to nucleosomes and protects them from DNA damage induced by free radicals (Chavez et al. 2019). Notably, Dsup expression in cultured human cells can increase their resistance to X-irradiation (Hashimoto et al. 2016), demonstrating that radiation-protective mechanisms that evolve in one species have the potential to function more broadly.

Historically, ionizing radiation has been an important tool for introducing mutations in model organisms for genetic studies, and was central to our understanding of the mutability of genes through Muller's experiments in Drosophila melanogaster and to the development of the “one gene, one enzyme” hypothesis in Neurospora crassa (Carlson 2013; Strauss 2016). Treatment of model organisms such as Drosophila and Neurospora with moderate doses of radiation creates diverse mutant phenotypes by disrupting the functions of discrete genes, likely through the creation of indels and chromosomal rearrangements that result from non-homology directed repair mechanisms such as NHEJ (Sekelsky 2017). However, for some other organisms that were brought into the lab for potential genetic study in the early part of the 20th century, visible mutants were extremely difficult to obtain via mutagenesis by ionizing radiation (Fabergé 1983). As such, these organisms did not rise to the level of study seen for the more easily mutable Drosophila.

Here, we consider the genetic response to ionizing radiation in the dark-winged fungus gnat Bradysia coprophila (syn. Sciara coprophila; Bradysia tilicola). B. coprophila has long been of interest among many geneticists due to its unique chromosome biology, which features multiple rounds of programmed chromosome elimination in the germline and in the early embryo, among other events (Gerbi 1986; Gerbi 2022). B. coprophila was first cultured as a lab organism by C.W. Metz in the early 1920s (Metz 1925). In the following decades when the mutagenic effects of ionizing radiation were of intense interest, several reports noted that phenotypic changes were exceedingly difficult to induce in many species of Bradysia (including B. coprophila) via ionizing radiation of germline cells, particularly in comparison to Drosophila (Smith-Stocking 1936; Metz and Boche 1939; Metz and Bozeman 1940; Reynolds 1941; Bozeman and Metz 1949; Crouse 1949, 1961; Fabergé 1983). This led to speculation that Bradysia (Sciara) may possess some mechanism of radiation resistance, at least in the germline. However, alternate explanations have been proposed instead of enhanced radiation resistance. For example, the unique sex-determination system of B. coprophila prevents novel recessive autosomal mutations from being isolated as homozygotes until 4 generations after irradiation, and it adds complication to genetic screens (Crouse 1949). In addition, since females have only sons or only daughters, 50% of sex-linked visible mutations will be missed in any one brood (Crouse 1949). It was also speculated that Bradysia (Sciara) may have a high dominant lethal rate (Crouse 1949). Furthermore, visible mutations may be hard to see because of the dark pigmentation of the adult fly (Crouse 1949). In addition, the paucity of radiation-induced visible mutations in Bradysia (Sciara) is not general to the sub-order Nematocera since many mutations have been produced by radiation in mosquitoes (Fabergé 1983).

We have now re-investigated whether B. coprophila has greater radiation resistance than D. melanogaster. Instead of scoring for radiation-induced phenotypic mutations, we have assayed the lethal doses of gamma-irradiation at various developmental stages of B. coprophila. Our analysis showed the highest resistance for pupae and the least for embryos, consistent with analyses of other organisms. Furthermore, we found that irradiated B. coprophila larvae are able to continue their development after exposures of 80 Gy, which triggers a developmental delay of about a week prior to pupation, suggesting an inherent plasticity in the developmental program. Side-by-side comparisons revealed that D. melanogaster larvae can survive 20 Gy, but not 40 Gy, leading to the conclusion that B. coprophila is somewhat (between 2- and 4-fold) more resistant than D. melanogaster to gamma-irradiation.

Recently, the B. coprophila genome was sequenced and its genes annotated (Urban et al. 2021, 2022; Hodson et al. 2022; Baird et al. 2023), enabling an opportunity to study the radiation response at the level of molecules and gene expression. We performed differential gene expression analysis on transcriptomes derived from irradiated and unirradiated larvae, providing a rich dataset for exploration of potential mechanisms of the B. coprophila radiation response. Interestingly, the up-regulated genes include members of the PARP and Argonaute families. Finally, to explore genome-wide transcriptional responses to irradiation across the Dipteran tree, we compared the B. coprophila results to 3 radiation-response studies from D. melanogaster (Brodsky et al. 2004; Akdemir et al. 2007; van Bergeijk et al. 2012). We found that very few specific genes had commonly shared transcriptional responses, even amongst only the Drosophila studies, despite specific pathways (such as DNA repair and apoptosis) being enriched across the transcriptome-wide response in both species. The PARP and Argonaute family responses appear to be unique to the B. coprophila dataset based on these comparisons.

Materials and methods

Fly husbandry

Although the dark-winged fungus gnat is best known as S. coprophila from a century of publications, it goes by several synonymous names, including S. coprophila (Lintner 1895), B. coprophila (Steffan 1966), Sciara tilicola (Loew 1850), Sciara amoena (Winnertz 1867), and others. Here, we refer to it as B. coprophila to be consistent with the current taxonomic name of its genus (Steffan 1966), its genome annotation (Urban et al. 2021, 2022), and the consensus of the community of scientists working with this fly. All fungus gnats analyzed in this study were female B. coprophila strain Holo2 that descends from the parental 7298 stock that carries the dominant marker Wavy on the X′ chromosome (Metz and Smith 1931). The Holo2 stock was recently used for genome sequencing and gene annotation (Urban et al. 2021, 2022). Stocks were obtained from the International Sciara Stock Center at Brown University (https://sites.brown.edu/sciara/).

B. coprophila adult females are monogenic and produce either all female offspring (gynogenic mothers; X′X genotype) or all male offspring (androgenic mothers; XX genotype). The dominant marker Wavy on the X' chromosome permits simple selection of the sex of the offspring by visualizing Wavy wings in X′X mothers vs straight wings in XX mothers. Cultures were maintained in vials containing a 2.2% agar substrate at 21°C in a humid incubator. Larvae were fed every other day with a dry mix of 2 parts ground shiitake mushroom, 1 part spinach powder, 1 part nettle powder, 4 parts ground oat straw, and 2 parts dry Brewer's yeast.

Irradiation and scoring

To collect B. coprophila female embryos for irradiation, gynogenic X′X female flies were mass mated with males (10–15 females and 15–20 males), and a day later the female flies were immobilized on a 2.2% agar petri plate, and the thorax was squeezed with forceps to induce egg laying. This method allows for nearly synchronous embryo age upon collection. Embryos that were 12–14 hours post-egg-laying were used for irradiation. To collect B. coprophila 1st instar larvae, gynogenic female and male adults were mass mated as above to produce vials with a large number of newly hatched female larvae, and agar plugs were moved from the vials to 2.2% agar petri plates for irradiation 1–2 days post-hatching. All B. coprophila female 4th instar larvae and pupae were collected from gynogenic females according to the days since mating and their morphology and placed on 2.2% agar petri plates. Further details on the life cycle and developmental stages of female B. coprophila are given in Fig. 1.

Fig. 1.

Developmental life cycle of B. coprophila. Timing for each stage is for 21°C and corresponds to our own observations. The timing is indicated as days post-mating (dpm) starting when male and female adults were placed together in a vial. The length of time for development varies according to temperature and culture conditions (e.g. underfed larvae develop more slowly than what is shown in this figure). These variables may account for any differences in developmental timing from that reported by Rieffel and Crouse (1966); moreover, (1) their “prepupal” stage is equivalent to the end of our larval eyespot stage; (2) they use the minimum times they found for each stage whereas we focus on average times in our normal culturing conditions. Adult female B. coprophila survive for 2–3 weeks after eclosion when stored at 16°C, unless they are placed in a vial with males in which case they die 1 to 2 days post-mating.

Animals were irradiated with ¹³⁷Cs γ-rays from a JL Shepherd irradiator. Doses are given in Gray (Gy), which corresponds to the absorption of 1 J/kg, where 1 Gy = 100 rad. A continuous dose rate of 1.7 Gy/minute was administered to animals supported on agar petri dishes enclosed in the chamber. Following irradiation, B. coprophila animals were kept on the 2.2% agar petri plates and fed every other day (3×/week). The viability and developmental progression for each stage were scored under a dissecting microscope. Larvae were scored as viable if they showed independent movement during observation. Pupae were scored as viable if they showed movement when prodded with a brush and/or had healthy amber coloration. Pupae that were stiff and dark brown and/or visibly covered with mold were scored as nonviable.

Similar to other reports (Jaklevic and Su 2004; Jaklevic et al. 2006), 4 day old larvae of D. melanogaster Canton-S at 3rd instar (feeding stage) were collected; they were transferred from vials to 2.2% agar petri plates for irradiation at the same time as B. coprophila (see above). After irradiation, the D. melanogaster larvae were transferred to fresh vials with standard Drosophila food.

Library preparation and sequencing

Roughly 590 female 4th instar pre-eyespot larvae (Fig. 1) originating from different mating vials were randomly and evenly distributed among 3 replicate groups to reduce potential batch effects from different cultures. For each replicate group, the larvae were subsequently distributed across 4 petri plates with ∼50 larvae/plate to prevent overcrowding, and divided such that half of the larvae were in the irradiation subgroup and half in the control subgroup for the given replicate. A total of 6 petri plates were used for irradiation and 6 petri plates used for the nonirradiated controls across the 3 replicates. Irradiated larvae were treated with 80 Gy (8000 rad) for 50minutes while the controls remained unirradiated. Approximately 45 minutes following radiation treatment, both irradiated and control larvae were snap-frozen in liquid nitrogen, total RNA was extracted with TRIzol, and cDNA libraries were prepared as previously described using RNeasy columns, AMPure bead cleanup, polyA selection, and NEB adaptor ligation (Urban et al. 2021). Libraries were sequenced to yield 100 bp paired-end reads using the Illumina HiSeq 2000 platform.

Differential expression and Gene Ontology (GO) enrichment analyses

After inspecting Illumina data using FastQC (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/), quantification of transcripts for each control and irradiated RNA-seq replicate was carried out with Salmon v. 1.2.1 (Patro et al. 2017) using an index built on the transcript database generated by the B. coprophila genome project (Urban et al. 2021; also see https://doi.org/10.15482/USDA.ADC/1522618) with no decoys. Differential expression (DE) analysis of irradiated compared to nonirradiated larvae was carried out using DESeq2 v. 1.26.0 (Love et al. 2014).

To analyze the enrichment of Gene Ontology (GO) terms pertaining to biological processes, molecular functions, and cellular components, we used 2 approaches. In our primary approach, we took advantage of the rich functional information for D. melanogaster gene annotations. For each B. coprophila gene that exhibited a significant change in gene expression between irradiated and control samples (Benjamini–Hochberg adjusted P < 0.05; i.e. <5% False Discovery Rate (FDR)), the corresponding predicted protein sequence (Urban et al. 2021) was compared to the D. melanogaster proteome using BLAST hosted at flybase.org (FB2020_03 Dmel Release 6.34; where relevant, gene symbols were updated to those corresponding to FB2022_06 Dmel Release 6.49) (Altschul et al. 1990; Larkin et al. 2021). A D. melanogaster protein was considered a potential homolog if it had an E value less than 1 × 10⁻⁴. A reciprocal BLAST was performed with each “hit” D. melanogaster protein sequence against the B. coprophila proteome; a D. melanogaster protein was considered a bona fide homolog if the same B. coprophila protein that identified the Drosophila protein was subsequently identified as the reciprocal best BLAST (RBB) hit. In addition to RBB hits to find homologs, we used OrthoFinder (Emms and Kelly 2019) on the longest protein isoforms of each gene across the Drosophila (FlyBase r6.49) and Bradysia (Urban et al. 2021; https://doi.org/10.15482/USDA.ADC/1522618) proteomes. The approaches agreed with each other on all shared ortholog assignments, and using both offered a bit more sensitivity in a minority of cases where 1 approach found an ortholog but the other did not. As best BLAST hits were good proxies of ortholog assignments, GO enrichment analysis (Ashburner et al. 2000) was performed with all D. melanogaster genes that were best BLAST hits (261 total) (using Bradysia queries) against the complete D. melanogaster gene set using PANTHER v. 17.0 (The Gene Ontology Consortium 2017; Mi et al. 2019) via geneontology.org.

In our secondary GO enrichment approach, we took advantage of the GO terms that were assigned to B. coprophila proteins with InterProScan as part of the original genome annotation project (Urban et al. 2021; https://doi.org/10.15482/USDA.ADC/1522618). For each GO term, enrichment P values and odds ratios were computed on 2 × 2 contingency tables using Fisher's exact test for the hypergeometric distribution with the SciPy Stats module in Python3. P values were then corrected with the Benjamini–Hochberg false discovery rate procedure with the statsmodels.stats.multitest module in Python3.

Phylogenetic analysis of PARP and Argonaute families

To determine relationships between PARP and Argonaute proteins from B. coprophila and their counterparts from other species, representative proteins were selected from previous phylogenetic analyses of PARP (Citarelli et al. 2010) and Argonaute (Lewis et al. 2016) families, and their sequences were obtained from public databases (NCBI Resource Coordinators 2018; UniProt Consortium 2021). PARP and PIWI domain sequences from each protein, including candidate B. coprophila homologs, were identified and extracted using Pfam (Finn et al. 2014), InterPro (Blum et al. 2021), and/or NCBI's Conserved Domain Database (Lu et al. 2020; Blum et al. 2021). Phylogenetic analysis was performed on extracted domain sequences using the Phylogeny.fr suite of tools (Dereeper et al. 2008), including alignment via MUSCLE (Edgar 2004), removal of poorly aligned positions and divergent regions via Gblocks (Castresana 2000), construction of phylogenies using maximum likelihood via PhyML (Guindon et al. 2010), and rendering of phylogenetic trees via TreeDyn (Chevenet et al. 2006).

Identifying conserved radiation-induced transcriptional responses in Dipterans

To identify conserved transcriptional responses after irradiation between B. coprophila and D. melanogaster, we first compiled a table detailing the transcriptome-wide changes after irradiation of D. melanogaster reported in 3 previous studies (Brodsky et al. 2004; Akdemir et al. 2007; van Bergeijk et al. 2012). Before making comparisons to B. coprophila, to establish which radiation-induced transcriptional responses were found across these 3 D. melanogaster studies, we classified each radiation response gene found in any of the 3 studies as either “low-concordance”, “moderate-concordance”, or “high-concordance” depending on if it was found in 1, 2, or all 3 of the studies, respectively. We then identified B. coprophila orthologs to all of the D. melanogaster radiation response genes with 2 methods, similar to above, using the longest protein isoforms of each gene across the D. melanogaster (FlyBase r6.49) and Bradysia (Urban et al. 2021; https://doi.org/10.15482/USDA.ADC/1522618) proteomes: (1) RBB hits and (2) OrthoFinder (Emms and Kelly 2019). Best B. coprophila BLAST hits were assigned as orthologs to the D. melanogaster queries if they were RBBs with the query, in the same OrthoGroup as the query, or both. Otherwise, the D. melanogaster gene was not assigned a B. coprophila ortholog. In all cases where both RBB and OrthoFinder found an ortholog, the 2 methods agreed. After assigning B. coprophila orthologs to the D. melanogaster genes reported in those 3 studies, we were able to determine which B. coprophila orthologs were also differentially expressed after irradiation in the same direction as found in D. melanogaster.

Results

B. coprophila larvae undergo developmental delay in response to ionizing radiation

To explore how B. coprophila survival and development are impacted by ionizing radiation, we exposed animals of various developmental stages (Fig. 1) to a range of gamma-irradiation dosages. Embryos aged 12–14 hours after egg-laying were highly sensitive to all doses of gamma-radiation tested (40–650 Gy), and no living larvae hatched from embryos irradiated at any dose (Table 1). In contrast, 1st instar larvae survived and grew up to a dosage of 40 Gy, survived but did not grow after 80 Gy irradiation, and did not survive after 160 Gy or 300 Gy irradiation. However, no irradiated 1st instar larvae pupated after any dose. Irradiated pupae were the most radiation-resistant, with nearly the same percent finishing pupal development and eclosing as adults after 40 Gy as after no irradiation treatment (80% vs 85%), with approximately 70% eclosing after 80–300 Gy, and with 27% eclosing after the highest tested dose of 650 Gy (Table 1).

Table 1.

Survival of B. coprophila at different developmental stages in response to ionizing radiation.

Dose (Gy)	Embryos irradiated	Embryos hatcheda	First instar larvae irradiated	First instar larvae surviving to adulthoodb	Pupae irradiated	Pupae eclosed
0	∼200	∼50–80%	290	74%	40	85%
40	∼200	0	106	0	35	80%
80	∼200	0	190	0	35	71%
160	∼200	0	105	0	35	69%
300	∼200	0	90	0	15	73%
650	∼200	0	ND	ND	15	27%

^a Estimated proportion of surviving embryos based on widefield views of egg-laying surface.

^b No evidence of pupation was observed for irradiated larvae up to 35 days post-irradiation. By 24 days post-irradiation, the 40 Gy irradiated larvae had grown larger, whereas the 80 Gy irradiated larvae were still very small but moved when poked with a needle. No first instar larvae survived irradiation at 160 Gy or 300 Gy.

We noticed that the larvae that survived high levels of radiation remained atypically small rather than undergoing the rapid growth observed for unirradiated larvae, suggesting a developmental delay or arrest in response to radiation. To further explore this, we exposed larvae of later developmental stages, primarily 4th instar pre-eyespot, to varying doses of irradiation and tracked their developmental progress through pupation and adulthood (Table 2). At 80 Gy, the lowest dose tested, 100% of larvae successfully pupated, whereas the rate of adult eclosion was reduced to half of that in the unirradiated set. After 100–120 Gy irradiation, 87–97% of the larvae pupated but almost none eclosed as adults. After 160–1100 Gy irradiation, only 20–30% of the larvae pupated and none eclosed as adults. Thus, larval irradiation beyond 80 Gy caused complete pupal arrest and/or lethality for those animals that successfully formed pupae, with no adult flies emerging (Table 2).

Table 2.

Arrested development of 4th instar larvae in response to high dosage of ionizing radiation.

Dose (Gy)	Larvae irradiated	% pupationa	% adult eclosiona
0	20	100%	45%
80	60	100%	20%
100	30	97%	0%
120	30	87%	3%
140	30	60%	0%
160	30	23%	0%
300	30	27%	0%
650	30	17%	0%
900	30	27%	0%
1000	30	37%	0%
1100	30	30%	0%

^a Percentages based on total irradiated larvae for each dose. Larvae were roughly 90% pre-eyespot and 10% eyespot stages.

Although 100% of 4th instar larvae exposed to 80 Gy of gamma-irradiation successfully pupated, the timing of the transition from larva to pupa appeared to be delayed relative to unirradiated controls. To quantify this developmental delay more precisely, we exposed pre-eyespot 4th instar larvae to varying levels of gamma-irradiation and scored the proportion of animals in each of the larval, pupal, and adult stages over time (Fig. 2, Supplementary Fig. 1). By day 6 post-treatment, half of unirradiated control larvae had transitioned to the pupal stage, while 95% of larvae irradiated with 80 Gy remained in the larval stage. By day 14, the majority (85%) of animals in the unirradiated control had emerged as adults, whereas only 10% of the animals irradiated with 80 Gy had eclosed as adults. Consistent with our earlier observations, a higher dose of radiation (160 Gy) resulted in complete developmental arrest in the larval stage for most animals (85%) through day 14, while an intermediate dose of radiation (40 Gy) showed a less severe developmental delay relative to the 80 Gy treatment (Fig. 2, Supplementary Fig. 1). In sum, our data support that B. coprophila larvae (pre-eyespot, 4th instar) are resistant to doses of ionizing radiation up to 80 Gy, which triggers a developmental delay of approximately 7–9 days prior to pupation. In contrast, irradiation of pupae did not significantly delay eclosion to adults (Supplementary Fig. 2), suggesting that the timing for developmental progression had been set by then.

Fig. 2.

Developmental delay of 4th instar larvae in response to moderate dosage of ionizing radiation. Groups of 20–30 4th instar pre-eyespot larvae were exposed to the indicated doses of ionizing radiation and developmental progress was followed over the next 14 days. The percentages of larvae, pupae, and adults for each treatment at each timepoint are shown. See Supplementary Fig. 1 for a graphical representation of these data.

B. coprophila larvae are somewhat more resistant to ionizing radiation than D. melanogaster

We have re-investigated the conclusion that B. coprophila larvae are more resistant to ionizing radiation than D. melanogaster; this was based primarily on the paucity of phenotypic mutations after irradiation of B. coprophila (Smith-Stocking 1936; Metz and Boche 1939; Metz and Bozeman 1940; Reynolds 1941; Bozeman and Metz 1949; Crouse 1949, 1961; Fabergé 1983). Instead, we have now done a side-by-side comparison of eclosion of B. coprophila and D. melanogaster after irradiation of petri plates with larvae from both species in the radiation chamber at the same time. As seen in Table 3, after 20 Gy radiation 75% of B. coprophila early 4th instar larvae had eclosed into adults as compared to only 35% for D. melanogaster (2.1-fold less). Even more strikingly, 87% of B. coprophila early 4th instar larvae had eclosed into adults after 40 Gy irradiation as compared to none for D. melanogaster. After 80 and 160 Gy radiation, neither species eclosed. This indicates that the irradiation level at which 50% of flies eclose must be less than 20 Gy for D. melanogaster in our hands, but between 40 and 80 Gy for B. coprophila. Thus, these data reveal that, in terms of its effects on development and eclosion, B. coprophila larvae are capable of withstanding at least 2-fold higher levels of ionizing radiation than D. melanogaster larvae.

Table 3.

Comparison of irradiation sensitivity in Bradysia coprophila (pre-eyespot 4th instar larvae) and Drosophila melanogaster (early 3rd instar larvae).

Dose (Gy)	Bradysia larvae irradiateda	% adult eclosion	Drosophila larvae irradiated	% adult eclosion
0	55	96%	51	59%
20	57	75%	80	35%
40	53	87%	51	0
80	50	0	41	0
160	50	0	24	0

^a The Bradysia larvae used for this experiment were slightly younger than those used for Table 2, thus explaining their somewhat greater sensitivity here at 80 Gy with 0% eclosion compared to 20% eclosion in Table 2.

B. coprophila larvae up-regulate DNA repair pathways and down-regulate developmental regulators in response to ionizing radiation

To determine the genetic response to ionizing radiation in B. coprophila, we divided female pre-eyespot 4th instar larvae into 3 replicate groups and exposed half of each group to 80 Gy of ionizing radiation, with the remaining half of each group serving as nonirradiated controls. We then generated Illumina RNA-seq libraries to produce roughly 7 to 10 million paired-end reads for each sample (51,316,924 total reads), and subsequently used Salmon (Patro et al. 2017) and DESeq2 (Love et al. 2014) to quantify transcript levels and perform DE analysis between irradiated and nonirradiated groups. Clustering of Salmon-quantified log10 TPMs of expressed genes demonstrated strong reproducibility among control and irradiated samples (Fig. 3a).

Fig. 3.

Differential expression analysis of irradiated B. coprophila larvae. a) Clustering of irradiated and control sample Salmon-quantified log10 TPMs of expressed genes shows reproducibility of the control and radiation procedures that started from randomly selected individual larvae from the same population (see Materials and methods). Radiation samples form a separate cluster from control samples. b) Volcano plot showing log2-fold change in expression vs −log10 of the Benjamini–Hochberg adjusted P value for each gene in the study. Purple, genes with significant (adjusted P < 0.05) up-regulation; green, genes with significant down-regulation; gray, no significant change in expression. Representative gene names reflect the closest D. melanogaster homolog; genes with reciprocal best BLAST hits are bolded. c) Magnified view of the inset shown in panel B.

In our DE analysis, 327 of the 23,117 B. coprophila genes in the current genome annotation (Urban et al. 2021; https://doi.org/10.15482/USDA.ADC/1522618) showed significant changes in their expression in response to irradiation (differentially expressed genes, DEGs), with 232 genes up-regulated and 95 genes down-regulated (Fig. 3b and c, Supplementary Tables 1 and 2). To better understand the functions of these DEGs, we used BLAST searches to identify candidate homologs of each B. coprophila protein in the well-characterized D. melanogaster proteome. Of the 327 B. coprophila genes identified in our experiment, 275 (84.1%; 195 up, 80 down) of their protein sequences identified a candidate homolog in Drosophila (Supplementary Tables 1–3), whereas 52 had no hits in the Drosophila proteome. Of the 275 Drosophila hits, 259 were unique hits (Supplementary Table 3), and 14 Drosophila genes were best hits for more than 1 differentially expressed Bradysia gene (Supplementary Table 4). We then used the best hit Drosophila protein sequences for each Bradysia DEG to perform reciprocal BLAST searches against the B. coprophila proteome, and found that 193 Drosophila proteins (130 up-regulated, 63 down-regulated) returned the RBB hit from B. coprophila (70.2% of DEGs with hits had RBBs; 59.0% of all DEGs), strongly suggesting an orthologous relationship for each of those proteins (Supplementary Tables 1–3). OrthoFinder found the same Drosophila ortholog for 179 of the 193 RBBs, and no ortholog for 14 of them. Thus, 14 DEGs had RBB-only support. OrthoFinder also found orthologs for 11 additional DEGs that did not have RBB support, and this brought the total number to 204.

The remaining 71 DEGs that had significant Drosophila BLAST hits but were not assigned orthologs still tended to share similar functional annotations, such as Pfam domains or GO terms, with their best Drosophila hits (Supplementary Text 1). Moreover, they are likely members of more complex families with multiple paralogs in at least 1 of the 2 species. One example of complex families with different numbers of homologs between species is the expansion of proteins with PARP domains in Bradysia, and another is the expansion of proteins involved in small RNA biology in Bradysia, both of which are discussed in detail later. Further consistent with this, 14 Drosophila proteins were identified as the top BLAST hit for multiple B. coprophila proteins encoded by DEGs in this study (Supplementary Table 4, Supplementary Text 2). Overall, the best Drosophila BLAST hits were largely determined to be orthologs, and those that were not had consistent functional and domain information, suggesting all best hits were good homolog proxies.

For GO analyses, our primary approach was to use the best Drosophila BLAST hit for each Bradysia DEG as a homolog proxy, and use the lists of all such Drosophila homologs to perform Drosophila-powered GO enrichment analyses for the biological pathways that exhibited changes in transcript levels in response to irradiation of B. coprophila (Tables 4 and 5, Supplementary Tables 5–14). As a secondary approach, we analyzed the GO terms assigned directly to Bradysia genes across the entire proteome as part of the original genome assembly and gene annotation (Supplementary Text 3; Urban et al. 2021; https://doi.org/10.15482/USDA.ADC/1522618). Of the 232 up-regulated DEGs, 125 had 313 GO terms (118 unique) and 107 had none. Of the 118 unique GO terms, 52 were deemed significant (Fisher’s exact test odds ratio ≥ 2 and FDR ≤ 10%), although just 19 were associated with more than 1 gene (Supplementary Table 15). Of the 95 down-regulated DEGs, there were 157 GO terms (56 unique) distributed across 54 genes (41 had none). Only 22 GO terms for the down-regulated genes were deemed significant, just 9 of which had more than 1 gene (Supplementary Table 16).

Table 4.

GO analysis of Drosophila homologs of B. coprophila genes that were down-regulated in response to radiation.

Analysis	GO set	Associated genes
Biological processes	Negative regulation of RNA metabolic process	tio, srp, pan, Samuel, en, gro, TfAP-2, salr, ct, bi, noc, bowl, al
	Regulation of cellular biosynthetic process	zfh2, tio, srp, pan, Samuel, en, e, gro, TfAP-2, salr, SMSr, ct, bi, noc, Gbs-70E, bowl, N, fd59A, tna, svp, grh, al, dnr1
	Development of anatomical structures, multicellular organisms, animal organs, larvae, pupae, wing discs, fat bodies, reproductive structures, digestive system, eye-antennal discs, Malpighian tubules	zfh2, boi, Cyp9f2, Mmp2, tio, tkv, srp, ple, pan, hdc, CG10339, en, Ten-a, e, fend, beat-VI, gro, ec, zye, TfAP-2, mab-21, salr, stg, nkd, Toll-7, ct, ex, bi, noc, bowl, shep, N, Lar, Efa6, sev, fd59A, Cpr67B, Mical, svp, ptc, grh, al, fz2, nw, tnc
	Cell fate commitment	Mmp2, tkv, srp, pan, en, nkd, bi, noc, N, sev, svp, ptc, grh, fz2
Molecular functions	Signaling receptor activity	boi, CG12290, tkv, CG10339, Toll-7, N, Lar, sev, svp, ptc, fz2
	Transcription regulator activity	zfh2, tio, srp, pan, en, gro, TfAP-2, salr, ct, bi, bowl, N, fd59A, tna, svp, grh, al
Cellular components	Cell surface	boi, Mmp2, CG10339, Toll-7, N, Lar, nw

Table 5.

GO analysis of Drosophila homologs of B. coprophila genes that were up-regulated in response to radiation.

Analysis	GO set	Associated genes
Biological processes	Nucleotide-excision repair	Xpc, RpII15, RPA2, hay, RpA-70, Ercc1, RPA3, mus201
	Double-strand break repair	nbs, Parp, Ku80, mei-41, DNApol-eta, RPA2, DNAlig3, RpA-70, Snm1, RPA3, CG6171, Irbp18, CycG, SMC5, rad50
	Response to radiation	Inx2, Ku80, mei-41, DNApol-eta, p53, hay, Dark, fbl, Ercc1, hid, Stat92E, babo, Dyrk2, mus201
	Positive regulation of cysteine-type endopeptidase activity involved in apoptotic process	p53, RnrL, Dronc, Dark, hid
	Transcription elongation by RNA polymerase II promoter	Cdk12, Spt5, RpII15, Cdk9
	Telomere binding and/or maintenance	nbs, Ku80, mei-41, RpA-70, Snm1, rad50, dmg5
Molecular functions	Damaged DNA binding	Xpc, nbs, Ku80, DNApol-eta, DNApol-iota, RpA-70, Ercc1, Snm1, RPA3, SMC5
	Regulatory RNA binding	AGO1, r2d2, aub, AGO3
	Dingle-stranded DNA binding	Xpc, RPA2, RpA-70, Ercc1, RPA3, SMC5, mus201, rad50
	ABC-type transporter activity	anne, CG4562, MRP, Mdr50, Hmt-1, CG7627
Cellular components	Apoptosome	Dronc, Dark
	RNA polymerase II, core complex	Polr2E, RpII15, RpII215
	polytene chromosome puff	Parp, Cdk12, Spt5, Cdk9, RpII215
	cytoplasmic ribonucleoprotein granule	shu, AGO1, Dis3l2, roq, aub, AGO3, Rcd-1

Consistent with the developmental delay observed among larvae irradiated with 80 Gy, we found that 45 of 80 (56.2%) Drosophila homologs of down-regulated B. coprophila genes were associated with GO terms related to development (Table 4; also see Supplementary Tables 5, 7, 8, and 12–14), including 11 genes with signaling receptor activity [e.g. sevenless (sev)] and 17 with transcription regulator activity [e.g. aristaless (al) and grainy head (grh)]. Furthermore, several genes that are normally up-regulated during the late larval and early pupal stages of Drosophila development (Graveley et al. 2011) were among the B. coprophila down-regulated gene set, including homologs of ebony, which is required for pigmentation of the Drosophila pupa case (Brehme 1941; Larkin et al. 2021), Samuel (also known as Moses), encoding a co-receptor for “nuclear hormone receptor 78” that is up-regulated in response to the hormone ecdysone (Baker et al. 2007), and narrow, which is required for growth of the Drosophila wing disc (Ray et al. 2015). More generally, the 78 unique Drosophila homologs for down-regulated Bradysia DEGs were enriched for 57 GO terms with the word “development”, 35 with “morphogenesis”, 15 with “differentiation”, 12 with “cell fate”, and 4 with “growth” (Supplementary Tables 8 and 12). GO terms corresponding to signaling, transcription factors, transcription regulation, and cell surface were also enriched (Supplementary Tables 13 and 14). We found similar GO terms directly assigned to down-regulated B. coprophila proteins (not Drosophila homologs) as part of the original automated functional annotation process (Urban et al. 2021; https://doi.org/10.15482/USDA.ADC/1522618) (Supplementary Table 16). For example, in this secondary approach, down-regulated Bradysia genes were enriched for GO terms such as “cell differentiation”, “multicellular organism development”, “nucleoside metabolic process”, “biosynthetic process”, “DNA-binding transcription factor activity”, “regulation of DNA-templated transcription”, “Notch signaling”, and “transmembrane receptor protein serine/threonine kinase activity”. Overall, it is clear that the down-regulated genes are those involved in promoting developmental progression, and the turning down or off of such genes may in part explain the developmental delays induced by irradiation.

In contrast to the down-regulated gene set, the 183 unique Drosophila homolog proxies corresponding to Bradysia genes that were up-regulated in response to irradiation were enriched for GO terms related to known cellular responses to DNA damage (Table 5; also see Supplementary Tables 5, 6, and 9–11), including DNA damage repair via nucleotide-excision and double-stranded break repair pathways (e.g. mutagen-sensitive 201 (mus201) and Ku80) (Sekelsky 2017), cell cycle regulation (dacapo, p53, mei-41) (Baonza et al. 2022), and promotion of apoptosis [e.g. p53, Death regulator Nedd2-like caspase (Dronc), and head involution defective (hid)] (Mollereau and Ma 2014; Baonza et al. 2022). Although the up-regulated genes were enriched with GO terms almost entirely to do with known radiation response genes, DNA repair and maintenance genes, and, to a lesser extent, apoptosis genes, other enriched GO terms included those dealing with transcription elongation, ABC transporter activity, regulatory RNA binding, and cytoplasmic RNA granules. In the latter 2 categories, many of these genes (e.g. AGO1, AGO3, aub, r2d2, Dis3l2, Rcd-1, roq, and shu) have known roles in small RNA biology, RNA catabolism, and translation regulation, suggesting their importance for the radiation response in B. coprophila. Indeed, there are other up-regulated genes with RNA binding and translation regulatory roles, including homologs for CG31957 and eIF6. When analyzing GO terms directly assigned to up-regulated B. coprophila proteins (Urban et al. 2021; https://doi.org/10.15482/USDA.ADC/1522618) (Supplementary Table 15), DNA repair and apoptosis terms were also enriched, but this secondary analysis also featured complementary enriched GO terms, such as “NAD+ ADP-ribosyltransferase activity”, featuring 2 PARP domain proteins, as well as “negative regulation of signal transduction”, “ATPase-coupled transmembrane transporter activity”, “ATP hydrolysis activity”, and “oxidoreductase activity”, the latter of which contained 4 P450 genes.

We also sought to understand more about the 52 B. coprophila genes that were differentially expressed in response to irradiation but that had no Drosophila BLAST hits (Supplementary Text 4; Supplementary Tables 17 and 18). The majority lacked functional or domain-level information, and were annotated as proteins of unknown function. However, of those with functional hits, a few up-regulated genes, for example, seemed to encode DNA repair and DNA modification proteins that may play a role in the DNA damage response (Supplementary Text 4).

Finally, we focused on several of the highest responding B. coprophila genes, which included homologs of human genes that are important in many cancers, such as poly-ADP-ribose polymerase (PARP) (Slade 2020), Guanine nucleotide binding protein like 1 (GNL1, known as Ns4 in Drosophila) (Krishnan et al. 2020), Growth Arrest and DNA Damage-inducible 45 (Gadd45) (Tamura et al. 2012), and Interferon-related developmental regulator 1 (Ifrd1) (Tummers et al. 2015; Lewis et al. 2017). Additionally, several genes involved in small RNA biology were among the up-regulated gene set, including r2d2, required for loading small interfering RNAs (siRNAs) into the RNA-induced Silencing Complex (RISC) in Drosophila (Liu et al. 2003), and 3 homologs that encode Argonaute family proteins, which are central players in small RNA pathways (Wu et al. 2020). Below, we further characterize B. coprophila homologs of 2 notable gene families, PARPs and Argonautes, that are present among the up-regulated genes list and represented by large families in the B. coprophila genome.

The B. coprophila radiation response up-regulates representatives of the PARP gene family

The B. coprophila gene Bcop_v1_g007065, related to Drosophila PARP, showed the largest degree of up-regulation (approximately 68-fold) in response to irradiation. Searches for protein domain signatures in the coding region of Bcop_v1_g007065 (Urban et al. 2021; https://doi.org/10.15482/USDA.ADC/1522618) indicated that it has a PARP catalytic domain as well as a series of Ankyrin repeats and a WGR domain. PARP family proteins are known to act in diverse cellular pathways, including DNA damage repair, signal transduction, apoptosis, and chromatin remodeling (Tulin et al. 2002; Hassa and Hottiger 2008; Jubin et al. 2016). Prior phylogenetic analysis has divided the PARP gene family into 6 clades with varying catalytic activities and functions, and 2 of these clades (clade 1 and clade 4) contain members that encode Ankyrin repeats, whereas Ankyrin repeats are not found in known members of the other 4 clades (Citarelli et al. 2010).

To place Bcop_v1_g007065 among the known PARP family clades, we first carried out a phylogenetic analysis using extracted PARP catalytic domain protein sequences from several representatives of clade 1 and clade 4 from different species. Our analysis shows clear placement of Bcop_v1_g007065 in clade 1, which includes the canonical human PARP1 that is an important regulator of DNA damage repair, and not in clade 4, which is comprised entirely of the Ankyrin repeat-rich PARP proteins known as Tankyrases (Fig. 4a). Furthermore, 2 other Ankyrin repeat-rich PARP genes that group with clade 1, pme-5 from Caenorhabditis elegans and Adprt3 from Dictyostelium discoideum, share the overall domain layout of Bcop_v1_g007065 with a WGR domain encoded between the Ankyrin and PARP domains, whereas the clade 4 Tankyrases instead encode a SAM domain in this position (Fig. 4a and b) (Gravel et al. 2004; White et al. 2009; Citarelli et al. 2010; Perina et al. 2014). We therefore conclude that Bcop_v1_g007065 is likely a functional ortholog of Pme-5 and Adprt3, and propose the name PME-5/Adprt3-Like PARP 1 (BcPALP1) for this highly radiation responsive gene in B. coprophila.

Fig. 4.

Bcop_v1_007065 (BcPALP1) is a clade 1 PARP homolog related to C. elegans Pme-5 and D. discoideum Adprt3. a) Phylogenetic analysis of Bcop_v1_007065 PARP domain with representative sequences from clade 1 and clade 4 PARPs. Pp, Physcomitrella patens; Hs, Homo sapiens; Dd, Dictyostelium discoideum; Dm, Drosophila melanogaster; Sj, Schistosome japonicum; Gg, Gallus gallus; Dr, Danio rerio; Xl, Xenopus laevis. b) Protein domains of H. sapiens Tankyrase-2 (clade 4), Bcop_v1_007065/BcPALP1, D discoideum Adprt3, C. elegans pme-5, and H. sapiens PARP-1 (clade 1). SAM, sterile alpha domain; WGR, Trp/Gly/Arg conserved domain; PRD, PARP regulatory domain; PADR1, domain of unknown function conserved in PARP proteins; BRCT, BRCA1 C-terminal domain. c) Phylogenetic analysis of 17 PARP homologs in B. coprophila. Annotated B. coprophila genes with PARP catalytic domains (Urban et al. 2021) were analyzed along with homologs from other species representing all 6 clades (Citarelli et al. 2010). Catalytic domain sequences for representative PARPs were taken from Physcomitrella patens (Pp), Vitis vinifera (Vv), Arabidopsis thaliana (At), Populus trichocarpa (Pt), Homo sapiens (Hs), Trichoplax adhaerens (Ta), Branchiostoma floridae (Bf), Dictyostelium discoideum (Dd), Drosophila melanogaster (Dm), Schistosome japonicum (Sj), Gallus gallus (Gg), Danio rerio (Dr), Xenopus laevis (Xl), Xenopus tropicalis (Xt), Nematostella vectensis (Nv), and Pyrenophora tritici-repentis (Pt-r). Each gene is labeled with either its gene name or its accession number. All known PARPs were correctly sorted into their published clades (Citarelli et al. 2010). Six B. coprophila PARPs group with clade 1, the DNA repair clade, including Bcop_v1_007065/Bcop-PALP1; 1 groups with the clade 4 tankyrases; 4 group with clade 6, which are likely ancient mono-ADP-ribosyltransferases with roles in membrane biology (Vyas et al. 2013), and the 6 remaining B. coprophila PARP homologs were not placed in known clades (“Novel”). Genes that are significantly up-regulated in irradiated B. coprophila larvae are highlighted in yellow. All B. coprophila homologs except Bcop_v1_008857, Bcop_v1_008859, and Bcop_v1_016259 were expressed in our dataset.

Our further analysis of the genome annotation of B. coprophila (Urban et al. 2021; https://doi.org/10.15482/USDA.ADC/1522618) revealed that it also contains 16 other candidate genes with homology to the PARP catalytic domain (17 total). To further support our characterization of BcPALP1, we carried out a larger scale phylogenetic analysis on all 17 potential B. coprophila PARP homologs using representative sequences of all 6 clades from other species. Of the 17 total B. coprophila genes with PARP catalytic domains, 6 mapped to clade 1, 1 mapped to clade 4, 4 mapped to clade 6, and 6 clustered as a novel outgroup to all known clades (Fig. 4c, Supplementary Tables 19–22). The B. coprophila homolog that grouped with clade 4, Bcop_v1_g000291, was also rich in N-terminal Ankyrin repeats, and likely represents a true Tankyrase ortholog. Conversely, BcPALP1 once again grouped with clade 1, further supporting an orthologous relationship for BcPALP1 with clade 1 PARP family members that participate in DNA repair (Citarelli et al. 2010). Furthermore, one of the PARP domain genes that was placed in the “Novel” clade, Bcop_v1_g016583, was among the genes that were up-regulated in response to radiation according to our RNA-seq data. However, this protein identified Drosophila deltex as its closest homolog in our BLAST searches (Supplementary Table 1). Analysis of the 6 PARP homologs in the Novel clade showed that 4 of them, including Bcop_v1_g016583, encode a C-terminal Deltex domain in addition to a PARP catalytic domain (Supplementary Table 22). Although Deltex domains are not found in human or Drosophila PARP homologs, BLAST searches show that a similar protein structure can be found encoded in the genomes of the closely related Bradysia odoriphaga (Chen et al. 2015) and the springtail Folsomia candida (Faddeeva-Vakhrusheva et al. 2017). Thus, this clade of PARP homologs may represent a specialized functional subgroup found in a subset of organisms, with a potential role in the response to ionizing radiation.

The B. coprophila radiation response up-regulates representatives of the Argonaute gene family

In addition to the strong up-regulation of BcPALP1 in our RNA-seq data, we noted that 3 of the up-regulated B. coprophila genes showed homology to Argonaute family genes. Argonaute proteins are small RNA binding molecules that serve as key effectors in RNAi silencing pathways, including gene silencing by miRNAs, siRNAs, and piRNAs (Wu et al. 2020). All known Dipteran Argonautes encode PIWI and PAZ domains in their protein sequences, and phylogenetic analysis based on their PIWI domains can divide family members into 4 clades: Ago1, Ago2, Ago3, and Piwi/Aubergine (Lewis et al. 2016). Notably, members of each clade carry out different functions, with Ago1 and Ago2 proteins participating in miRNA and/or siRNA silencing, and Ago3, Piwi, and Aubergine participating in biosynthesis and effector steps of piRNA silencing (Meister 2013).

To understand which RNAi pathways may be involved in the radiation response of B. coprophila, we carried out a phylogenetic analysis of B. coprophila Argonaute homologs, including the 3 genes up-regulated in the radiation response (Bcop_v1_g003309, Bcop_v1_g013021, and Bcop_v1_g004567) along with 10 other potential Argonaute homologs identified by the B. coprophila genome project (Urban et al. 2021; https://doi.org/10.15482/USDA.ADC/1522618) that did not significantly change expression in response to radiation. Using extracted PIWI domains from these proteins and from Dipteran Argonaute homologs that had previously been characterized through phylogenetic analysis (Lewis et al. 2016), we found that each of the 3 B. coprophila Argonautes that were up-regulated in response to radiation grouped with a different clade (Fig. 5). Specifically, Bcop_v1_g004569 grouped with the Ago3 clade and Bcop_v1_g013021 fell within the Piwi/Aubergine clade, implicating involvement of the piRNA pathway in the radiation response of B. coprophila, whereas Bcop_v1_003309 grouped near the Ago1 and Ago2 clades but did not fall clearly into either one of them. Subsequent phylogenetic analysis of Bcop_v1_003309 and a larger set of known Ago1 and Ago2 homologs showed a closer relationship with the Ago1 clade (Supplementary Fig. 3), which may indicate a functional divergence from an Ago1-like ancestral gene. Beyond the 3 Argonaute homologs that were up-regulated in response to irradiation, the remaining 10 B. coprophila Argonautes consist of 2 homologs in the Ago2 clade, 2 in the Ago1 clade, 1 in the Ago3 clade, and 5 in the Piwi/Aubergine clade (Fig. 5, Supplementary Tables 23–27), demonstrating the potential for diverse small RNA pathways in the unique biology of B. coprophila.

Fig. 5.

Phylogenetic analysis of 13 Argonaute homologs in B. coprophila. Genes encoding Argonaute homologs that are significantly up-regulated in irradiated larvae are highlighted in yellow. Note that Bcop_v1_003309 groups outside of the Ago1 clade in this tree, but is placed within the Ago1 clade in an expanded analysis of Ago1 and Ago2 sequences (Supplementary Fig. 3). Do, Drosophila obscura; Tb, Tabanus bromius; Db, Drosophila busckii; Eb, Episyrphus balteatus; Dm, Drosophila melanogaster; Ca, Corethrella appendiculata; Bd, Bactrocera dorsalis; aa, Anopheles albimanus; De, Drosophila erectus; Dy, Drosophila yakuba; Ad, Anopheles darlingi; Aa, Aedes aegypti. All B. coprophila homologs except Bcop_v1_018650 were expressed in our dataset.

Establishing a set of universal conserved irradiation response genes across Diptera

B. coprophila is a member species of the so-called “lower Diptera” (Nematocera) that diverged over 200 million years ago from the “higher Diptera” (Brachycera) (Wiegmann et al. 2011), of which D. melanogaster is a member. To establish a putative set of conserved inter-species irradiation response genes across the Dipteran tree, we compared our results to 3 radiation-response studies from D. melanogaster (Brodsky et al. 2004; Akdemir et al. 2007; van Bergeijk et al. 2012). First, we reconstructed a table from Brodsky et al. (2004) that listed well-established DNA damage response genes from mammals, flies, and yeast (Table 5; Supplementary Table 28). We found that B. coprophila has single-copy orthologs with 5 of the 6 DNA damage response genes from D. melanogaster: ATR/mei-41, ATRIP/mus304, ATM/tefu, Chk1/grp, and Chk2/lok. Each of these genes were annotated in B. coprophila as having a single isoform, except for Chk2/lok, which had 2 isoforms. In addition, whereas D. melanogaster has only 1 member of the p53/p63/p73 family, B. coprophila has 2 members, one of which is more similar to the D. melanogaster p53 than the other, but both of which appear to be more similar to mammalian p63 and p73 than p53. Interestingly, D. melanogaster was reported to have no ortholog to mammalian DNA-PKcs, but there is a single-copy ortholog of this gene in B. coprophila, supporting the notion that it was lost in Drosophila, but not all Dipterans. In both species, a p53 member is found up-regulated in response to radiation (Table 6; Supplementary Table 28). In D. melanogaster, chk1/grp and chk2/lok are found in 1 study (van Bergeijk et al. 2012) to be down-regulated in response to irradiation, and although neither were found to be significantly differentially expressed in B. coprophila in this study, both changed in the same downward direction (Table 6; Supplementary Table 28). ATR/mei41 was not reported as differentially expressed in the Drosophila studies and D. melanogaster lacks DNA-PKcs, but both genes were found up-regulated in response to radiation in B. coprophila larvae in this study (Table 6; Supplementary Table 28). Overall, we conclude that B. coprophila has representatives from each member of this set of well-established DNA damage response genes, with an additional p53 family member and an additional gene (DNA-PKcs) compared to Drosophila, and report the protein sequences of these orthologs here (Supplementary Table 28).

Table 6.

DNA damage response homologs in mammals, yeasts, fruit flies, and fungus gnats.

Mammal	Budding yeast	Fission yeast	Drosophila	Differentially expressed in Drosophilaa	Bradysia coprophila	Differentially expressed in Bradysia	Bradysia coprophila ID (Urban et al. 2021)	Bradysia coprophila NCBI annotation protein ID
ATR (upstream kinase)	Mec1	Rad3	mei-41	.	Bcop_mei41/ATR-like	Up	Bcop_v1_g008402	XP_037051938.1
ATRIP (ATR binding protein)	Ddc2	Rad26	mus304	.	Bcop_mus304_ATR-interacting-protein-like	.	Bcop_v1_g019717	XP_037047670.1
ATM (upstream kinase)	Tel1	Tel1	tefu	.	Bcop_tefu/ATM-like	.	Bcop_v1_g000127	XP_037038317.1
DNA-PKcs (upstream kinase)	*	*	*	-	Bcop_DNA-PKcs	Up	Bcop_v1_g001375	XP_037038022.1, XP_037038023.1
Chk1 (downstream kinase)	Chk1	Chk1	grp	Up in 1	Bcop_grp/Chk1-like	.	Bcop_v1_g001888	XP_037052193.1
Chk2 (downstream kinase)	Rad53	Cds1	lok	Down in 1	Bcop_Lok/Chk2-like	.	Bcop_v1_g002778	XP_037052409.1
p53/p63/p73 family of transcription factors	*	*	p53	Up in 2	Bcop_p53/p63-like	.	Bcop_v1_g013595	XP_037028247.1
					Bcop_p53/p73-like	Up	Bcop_v1_g018675	XP_037037422.1, XP_037037423.1, XP_037037424.1

^a The number of studies (of 3 total) where the gene was found to be differentially expressed.

^*This gene/protein is not found in this system.

Next, to compare radiation response genes between D. melanogaster and B. coprophila, we first compiled the reported transcriptome-wide irradiation response genes from all 3 D. melanogaster studies into a single table (Table 7, Supplementary Table 29), noting which genes showed up in all 3 studies (“high-concordance”), 2 of the 3 (“moderate-concordance”), or just 1 (“low-concordance”). We found that very few of the D. melanogaster radiation response genes showed up in more than 1 study, with just 30 of 253 (∼11.9%) in more than 1 study (moderate- and high-concordance), and just 8 (∼3.2%) in all 3 studies (high-concordance only) (Table 7, Supplementary Table 29). The 8 genes found in all 3 studies (or even the 30 found in 2 or more) may represent genes that are part of the core universal radiation response regardless of cell type, tissue, or life stage for mitotically dividing cells. Indeed, the 8 genes that showed up in all 3 studies (Corp, Xrp1, egr, escl, hid, mre11, rpr, skl) were all part of the “high stringency” radiation-induced p53-dependent (“RIPD”) genes from Akdemir et al. (2007) (Supplementary Table 29). Moreover, of the 22 genes that came up in 2 of the 3 studies, 14 were high stringency RIPD genes (Irbp18, IscU, Mabi, PolZ1, RnrL, Xpc, eIF6, pyd, rdog, spo, CG12171, CG18596, CG5664, CG6171), another 4 were in the low stringency RIPD set (Btk, p53, CG15480, CG16815), and just 4 were not in the RIPD sets (Irbp, Ku80, drl, rad50) (Supplementary Table 29). Thus, 26 of the 30 genes in 2 or more studies were supportive of the RIPD gene sets (Akdemir et al. 2007), particularly the high stringency set, which attempted to outline the core universal set of genes that respond to radiation (particularly those dependent on p53). Interestingly, the 8 genes found in all 3 studies were all up-regulated, and 29 of the 30 (96.7%) genes in 2 or more studies were up-regulated, leaving just 1 down-regulated gene found in more than 1 study: “derailed” (drl; a receptor tyrosine kinase in the Wnt5 signaling pathway). Thus, the most universal radiation response genes in Drosophila tend to be genes that are up- and not down-regulated in response to radiation.

Table 7.

Summary of irradiation response genes from 3 Drosophila studies and their Bradysia orthologs in this study.

	# In Drosophila studies	# with Bradysia orthologs	# up- regulated in both species	# down- regulated in both species	DNA repair genes up-regulated in both species	Proapoptotic genes up-regulated in both species	Other genes up-regulated in both species	Genes down-regulated in both species	Drosophila response genes with Bradysia orthologs that had no evidence of DE	Drosophila response genes with no orthologs in Bradysia
High-concordance genes	8	4	2	0	mre11*	hid	.	.	escl, egr	rpr, Corp, skl, Xrp1
Moderate-concordance genes	22	19	12	1	p53, Xpc, RnrL, CG6171, Irbp18, rad50, Ku80, Irbp*	p53	eIF6, pyd, IscU*, spo*	drl	Btk, rdog, PolZ1, CG5664, CG12171, CG18596	Mabi, CG15480, CG16815
Low-concordance genes	223	196	31	22	Ercc1, DNAlig3, RPA1, RPA3, Pol Iota, agt, mu2, DNAlig4*, PolH*, SMC6*, mei-9*, spn-A*, RfC4*, Ssrp*	Dark, Dronc, Clbn, Traf4, Scyl	CIAPIN1, Gadd45, kay, obe, Nfs1, CG1582, CG2017, CG17734, BI-1*, Rdh1*, Ttc19*, CG9411*	Eip75B, en, tkv, Atx-1*, Dref*, Egfr*, Mcm3*, Oli*, ac*, chif*, dnt*, klu*, l(1)sc*, mip130*, nerfin-1*, sc*, unc-5*, vg*, wntD*, woc*, Klp3A*, Ndf*	See Supplementary Table 29	See Supplementary Table 29
Totals	253	219	45	23	.	.	.	.	.

Genes that had weak statistical evidence in this study are marked with asterisks. Strong statistical evidence was any gene with an expected FDR of <5%. Genes with weak statistical evidence had FDRs > 5%, but had either P < 0.05, a nonsignificant sizeable fold-change (>1.49-fold), or both. In all cases, gene expression levels moved in the same direction in both species after irradiation if reported here as up- or down-regulated.

DE, differentially expressed.

Finally, we investigated which of the D. melanogaster radiation response genes were also found in our study of irradiated B. coprophila. These genes might represent a set of the most fundamental radiation response genes in Dipterans and beyond. Of the 8 proteins found in all 3 D. melanogaster studies (“high-concordance D. melanogaster genes”; Table 7, Supplementary Table 29), only 2 are also found differentially expressed in the same direction (up-regulated) in B. coprophila larvae shortly after irradiation. Interestingly, both of the major DNA damage response pathways are represented by 1 gene each: “hid” for apoptosis (Supplementary Text 5), and “mre11” for DNA repair, which had strong (<5% FDR) and weak (P ∼ 0.025, but ∼38% FDR) statistical evidence in our study, respectively. Of the 6 remaining high-concordance D. melanogaster genes, 2 (“escl” and eiger) had no evidence of significant differential expression in B. coprophila larvae and 4 (rpr, corp, skl, Xrp) had no orthologs found among the B. coprophila gene annotations (Supplementary Text 5). As it stands from this analysis, “hid” and “mre11” may be the most universally conserved genes that are up-regulated early in response to irradiation in Diptera, both dependent on p53 (Akdemir et al. 2007). However, other genes that came up in 2 of the 3 D. melanogaster studies, including p53, may also be part of this Dipteran-wide universal core set of genes.

Of the 22 proteins found in 2 of the 3 D. melanogaster studies (“moderate-concordance D. melanogaster genes”; Table 7, Supplementary Table 29), 13 (59%) had evidence of being differentially expressed in the same direction in B. coprophila shortly after irradiation. The only down-regulated D. melanogaster gene in this cohort, drl, was also down-regulated in B. coprophila. The remaining 12 were up-regulated in both species, 9 of which had strong evidence in B. coprophila, including 1 of the 2 p53 orthologs, Xpc, RnrL, CG6171, eIF6, Irbp18, pyd, rad50, and Ku80; and 3 of which had weak evidence (spo, IscU, Irbp). The majority of the conserved up-regulated genes (8) were involved in DNA repair (p53, Xpc, RnrL, CG6171, Irbp18, rad50, Ku80, and Irbp). From that list, 1 (p53) is also proapoptotic. Otherwise, proapoptotic genes were not enriched among the conserved up-regulated moderate-concordance genes. Of the 4 remaining moderate-concordance response genes found up-regulated in both species, 2 may indirectly support DNA repair: (1) IscU is involved in iron-binding and iron-homeostasis, which may be indirectly related to DNA repair as many DNA repair enzymes require iron as a co-factor (Puig et al. 2017), and (2) eIF6 is involved in translation initiation regulation and may facilitate the need for the quick production of the many transcriptionally up-regulated DNA repair proteins. The remaining 2 conserved up-regulated moderate-concordance response genes were Polychaetoid (pyd), potentially involved with the cell surface and cytoskeleton, and a cytochrome P450 enzyme (spook/spo), possibly involved in ecdysteroid biosynthesis and development. The set of 22 moderate-concordance D. melanogaster genes also included 9 (41%) that were either not found or not differentially expressed in the same direction in B. coprophila (Table 7, Supplementary Text 6, Supplementary Table 29).

Of the 253 genes reported in the 3 D. melanogaster studies, 223 (88.1%) were found in only 1 study (“low-concordance D. melanogaster genes”; Table 6, Supplementary Table 29). Of the 223 low-concordance D. melanogaster genes, 196 (87.9%) had orthologs in B. coprophila, but at most, just 53 (23.8%) were differentially expressed in the same direction in both species, of which only 23 (10.3%) had strong evidence (FDR < 5%) in B. coprophila. Of these 53 genes, 31 were up-regulated in B. coprophila, of which 20 had strong evidence. The plurality of low-concordance genes up-regulated in both species (14) were clearly annotated as being involved with DNA repair (Ercc1, DNAlig3, RPA1, RPA3, Pol Iota, agt, mu2, DNAlig4, PolH, SMC6, mei-9, spn-A, RfC4, Ssrp). At least 5 were proapoptotic (Dark, Dronc, Clbn, Traf4, and Scyl). Interestingly, there were 2 up-regulated genes annotated as negative regulators of apoptosis (CIAPIN1, BI-1). Two up-regulated genes, Gadd45 and kay, were annotated as involved in wound healing. Note that Gadd45 and Scyl are also negative regulators of growth, and that Traf4 and Gadd45 are both part of JNK signaling. The remaining up-regulated genes were an assorted mix, including functions involved in mRNA splicing (obe), RNA helicase activity (CG1582), translation elongation (CG2017), cysteine desulfhydrase activity (Nfs1), and behavioral responses to alcohol (CG17734). Overall, low-concordance D. melanogaster genes that are up-regulated in both species were enriched for the DNA repair response, and to a lesser extent, the apoptotic response. There were also up to 22 low-concordance D. melanogaster genes found down-regulated in both species (Table 6, Supplementary Text 7, Supplementary Table 29), although just 3 had strong evidence (Eip75B, en, tkv). All 22 down-regulated genes were associated with development, reproduction, or the cell cycle. Finally, there were 27 low-concordance Drosophila genes without B. coprophila orthologs (Supplementary Table 29), which included cytochrome P450 (Supplementary Text 8) and Glutathione S Transferase (GST; Supplementary Text 9) genes. In both cases, D. melanogaster and B. coprophila up-regulated different members of those families, suggesting those responses are more likely conserved at the level of the gene family than the level of specific genes and orthologs.

Discussion

Level of radiation sensitivity of B. coprophila compared to D. melanogaster

Bradysia (Sciara) coprophila was established as a model genetic organism roughly a century ago, and yet very few visible mutants have been recovered since then, suggesting the hypothesis of its inherent resistance to DNA damage induced by ionizing radiation. This characteristic led to Drosophila being chosen by geneticists a century ago as a popular model organism rather than Bradysia (Sciara). Historically, studies of radiation resistance in B. coprophila have focused on transmission of visible mutations or chromosomal rearrangements to the progeny of irradiated organisms (Metz and Boche 1939; Metz and Bozeman 1940; Reynolds 1941; Bozeman and Metz 1949; Crouse 1949, 1950, 1961). Here, we have performed a side-by-side comparison of the radiation resistance levels of B. coprophila compared to D. melanogaster, using viability and developmental progression rather than adult phenotypic mutations as yardsticks.

In our studies, only 35% of D. melanogaster feeding-stage larvae eclose after 20 Gy radiation and none after 40 Gy radiation. In contrast, 75 and 87% irradiated B. coprophila pre-eyespot larvae eclose after 20 and 40 Gy, respectively, but none after 80 Gy or higher (Table 3). Notably, our eclosion rate of Drosophila after radiation is less than that reported by others (e.g. 84% eclosion for Canton-S and ∼55% eclosion for Oregon-R after 40 Gy radiation; Sudmeier et al. 2015). However, there was great variation in eclosion among 52 DGRP Drosophila lines studied in a previous analysis (Sudmeier et al. 2015). Although we used the Canton-S line for our studies, its genetic background may have changed over time. Also, to treat them identically to B. coprophila, we transferred the third instar Drosophila larvae from their vials with food to agar petri plates for radiation, in contrast to leaving them in the food vials as done by others. Finally, in order to compensate for differences in machine calibration for radiation dose, we irradiated petri plates with an agar substrate with Drosophila or B. coprophila larvae side-by-side at the same time. While our data support that B. coprophila is indeed more resistant to radiation than D. melanogaster, it is not by orders of magnitude. When compared to D. melanogaster for radiation-induced phenotypic mutation, the dramatic paucity of these in B. coprophila most probably reflects the alternate explanations that have been hypothesized, such as difficulty to screen for phenotypic mutations (Crouse 1949) rather than an overwhelming resistance to radiation. Moreover, chromosomal rearrangements can be fairly readily induced by irradiation of B. coprophila (Crouse 1943, 1950, 1960a, 1960b, 1977).

Irradiation of B. coprophila larvae results in a developmental delay and down-regulation of genes that promote development

Using viability and eclosion as parameters, we have assayed the radiation sensitivity of B. coprophila at various developmental stages. As also noted by others for Drosophila (Sudmeier et al. 2015; Paithankar et al. 2017), radiation sensitivity decreases as development progresses. B. coprophila embryos were not viable even at the lowest radiation dose tested here, whereas first instar larvae were viable up to 80 Gy though none eclosed into adults (Table 1). In contrast, about half of B. coprophila fourth instar larvae eclosed into adults after 80 Gy irradiation as compared to unirradiated controls, and irradiated B. coprophila pupae eclosed even after very high doses (73% after 300 Gy and 27% after 650 Gy). It is plausible that the enhanced radiation resistance of pupae might reflect in part the cessation of cell division after tissue remodeling and growth.

Our data demonstrate that irradiated B. coprophila fourth instar larvae delay their transition to pupation by approximately 7–9 days in response to 80 Gy of gamma-irradiation, whereas lower doses have little effect on development. Conversely, higher doses result in a “suspended” motile larval state for surviving larvae with no evidence of pupation several weeks after the expected transition would take place. Thus, there appears to be a close coordination between developmental progression and the response to radiation, perhaps analogous to the molecular “checkpoint” mechanisms that prevent progression through the cell cycle in response to DNA damage. It is known that various kinds of injury induce developmental delays in insects (Hackney and Cherbas 2014), and, similar to our findings reported here for B. coprophila, radiation of Drosophila larvae delays their time to pupation (Villee 1946; Bourgin et al. 1956). The radiation-induced developmental delay might be similar to facultative diapause or quiescence in response to unfavorable environmental conditions (Gill et al. 2017; Hutfilz 2022).

Consistent with the delayed developmental progression of irradiated larvae, RNA-seq analysis showed reduced expression of B. coprophila genes that were homologous to known Drosophila developmental regulators, including several components of signal transduction pathways and homeodomain- and zinc finger-family transcription factors, marking these genes as candidate regulators of late larval and early pupal development in B. coprophila. In holometabolous insects, the transition from larval to pupal stages is largely controlled by ecdysteroid hormones, which increase during late larval stages and impact development via signaling through nuclear hormone receptors such as the Ecdysone Receptor (EcR) (Jindra 2019). Endocrine hormones are generally implicated to perform regulatory roles for insect diapause (Hutfilz 2022). Developmental delay in Drosophila due to injury produces a signal that inhibits ecdysteroid production or secretion by the larval ring gland; this injury-response pathway involves regulation of the neuropeptide PTTH that stimulates the ring gland to produce ecdysone (Hackney and Cherbas 2014). Furthermore, ecdysteroid levels control developmental progression in irradiated Drosophila, and ectopic feeding of 20-hydroxyecdysone can overcome radiation-induced developmental delay (Halme et al. 2010). It is as yet unclear whether the delayed development in irradiated B. coprophila larvae reflects changes upstream or downstream of ecdysteroid biosynthesis and signaling, although we observed a roughly 2-fold down-regulation of a B. coprophila ortholog of Samuel, a gene that plays a role in regulating signaling downstream of EcR in Drosophila (Baker et al. 2007). Also, down-regulation of the ecdysone pathway genes Eip74EF (Ecdysone-inducible protein 74EF) and Eip75B (Ecdysone-inducible protein 75B) were found in irradiated D. melanogaster (van Bergeijk et al. 2012) and are also found down-regulated in B. coprophila. Our data may help to elucidate the molecular connections between B. coprophila hormone signaling and developmental effectors.

Irradiation of B. coprophila larvae results in up-regulation of DNA repair and apoptotic genes, and features members of the PARP and AGO families

Differential gene expression analysis of B. coprophila transcriptomes derived from larvae treated with 80 Gy of ionizing radiation relative to unirradiated controls showed up-regulation of genes related to DNA repair, cell cycle arrest, and apoptosis, and down-regulation of developmental regulators, providing a key dataset to better understand their radiation response. As expected, ionizing radiation led to the up-regulation of genes involved in several DNA repair pathways, including NER and NHEJ, representing a common response to DNA damage among metazoans (Sekelsky 2017). The highest change in expression was observed for a PARP family homolog that we named BcPALP1, which is orthologous to C. elegans pme-5 and D. discoideum Adprt3 according to overall structure and catalytic domain sequence homology. Although functional characterization of Adprt3 is lacking, pme-5 is among the most transcriptionally up-regulated genes in response to ionizing radiation in worms, and DNA damage-induced germ cell apoptosis is increased in worms where pme-5 expression is knocked down via RNAi, providing functional evidence of a role in the DNA damage response (Gravel et al. 2004). PARP catalyzes ADP-ribosylation, a modification of proteins and nucleic acids often seen in a stress response such as to DNA damage (Lϋscher et al. 2022). The PARP complex marks sites of DNA damage via serine-linked ADP-ribosylation in both mammals and Drosophila (Fontana et al. 2023).

BcPALP1 is 1 of 17 PARP genes annotated by the B. coprophila genome project (Urban et al. 2021), representing a large expansion of this gene family, particularly in comparison to the Drosophila genome, which encodes only a single clade 1 PARP1 homolog and a single clade 4 tankyrase (Larkin et al. 2021). The B. coprophila PARP catalytic domain is found in 17 family members in 4 known clades plus a novel clade. The expansion of the PARP gene family may therefore be related to the potential radiation resistance of B. coprophila, with BcPALP1 likely playing a central role. BcPALP1 in clade 1 shows the greatest increase (68.5-fold up-regulated) in differential gene expression after irradiation of B. coprophila. Both Drosophila PARP1 and B. coprophila BcPALP1 (Bcop_v1_g007065) are in clade 1, and share WGR, PRD, and PARP domains in the C-terminal half of the protein. However, BcPALP1 has Ankyrin repeats in the N-terminal half instead of the Zinc finger, PADR, and BRCT domains found in D. melanogaster PARP1. At least 6 of the B. coprophila PARP homologs group within clade 1, where many members from other species have been shown to have roles in DNA repair (Citarelli et al. 2010).

The other up-regulated (2.6-fold) PARP family member in our study (Bcop_v1_g016583) was in a novel clade rather than clade 1, and it has a Deltex domain, which Drosophila PARP1 lacks. B. coprophila has 4 PARP homologs in this novel clade that appear to be a fusion of the PARP catalytic domain with the C-terminal domain of Deltex. The best-known function of Deltex in Drosophila is its regulation of Notch signaling through direct interactions (Diederich et al. 1994; Dutta et al. 2017), but it has other functions, including modulating the expression of Decapentaplegic (dpp) in Drosophila (Sharma et al. 2022). Deltex is an E3-ubiquitin ligase and it ubiquitinates the intracellular domain of Notch (Cornell et al. 1999; Yamada et al. 2011; Zhu et al. 2022). Notably, Deltex (DTX3L) forms a heterodimer with PARP9 in mammals (Chatrin et al. 2020; Ashok et al. 2022). Both PARP and Deltex have a WWE domain that is often identified in proteins associated with ubiquitination or poly-ADP-ribosylation (He et al. 2012); both of these post-translational modifications play roles in pathways including the DNA damage response (Ashok et al. 2022). The fusion of the PARP catalytic domain with the C-terminal domain of Deltex in 4 members of the novel PARP clade in B. coprophila suggests that the interaction of PARP with Deltex seen in trans in mammalian heterodimers might occur in cis in B. coprophila and could shed light on the evolution of their functional interactions.

We also noted 3 Argonaute homologs among the genes that were significantly up-regulated in response to radiation, along with the siRNA loader r2d2, strongly implicating small RNA pathways in the B. coprophila response to radiation. Previous studies from both plants and animals have shown that DSBs in DNA generate small RNAs that can be bound by Argonaute homologs, which facilitates recruitment of repair factors such as Rad51 (d’Adda di Fagagna 2014; Gao et al. 2014; Oliver et al. 2014; Rzeszutek and Betlej 2020; Hu et al. 2021). B. coprophila Argonaute homologs may therefore directly participate in DNA repair pathways. Furthermore, DNA damage is known to activate and mobilize transposable elements (TEs) in both prokaryotes and eukaryotes (McClintock 1984; Bradshaw and McEntee 1989; Walbot 1992; Eichenbaum and Livneh 1998; Rudin and Thompson 2001; Hagan et al. 2003; Farkash and Luning Prak 2006). Phylogenetic analysis showed that 2 of the up-regulated B. coprophila Argonaute genes are orthologous to Ago3 and Piwi/Aubergine proteins that are key components of the piRNA pathway, which plays an important role in silencing TEs in the germline of metazoan organisms (Wu et al. 2020). The up-regulation of B. coprophila Ago3 and Piwi orthologs may therefore be related to maintaining genome integrity in the germline by preventing mobilization of TEs. It is also possible that up-regulation of Argonaute homologs aids in the post-transcriptional regulation of other B. coprophila genes through the activation of RISC (Pratt and MacRae 2009), perhaps suppressing the translation of other developmental regulators until DNA damage has been repaired. A detailed analysis of small RNAs in irradiated larvae may reveal an additional layer of gene regulation that is not captured by DE analysis of mRNAs. For example, irradiation activates bantam microRNA in D. melanogaster, which represses hid and limits radiation-induced apoptosis (Jaklevic et al. 2008).

Comparison of differential gene expression in B. coprophila and D. melanogaster after radiation

In comparing our RNA-seq results to published studies of irradiation responses in Drosophila, we found very few DEGs that were common among 3 analyses of Drosophila, and even fewer that were shared with Bradysia. Differences in the data could reflect different responses to radiation of polytene cells (B. coprophila) to diploid cells (D. melanogaster) in the tissues used for these studies. However, there is a lack of concordance in the transcriptome-wide data even among just the 3 D. melanogaster studies. One interpretation of the prevalence of single-study DEGs is that pathways are more often represented than specific genes. All 3 Drosophila studies have up-regulated genes enriched in pathways for DNA repair and apoptosis, for example, despite little overlap at the gene level. Moreover, genes that need to be differentially regulated as part of the radiation response may not be universal across studies because they are cell-, tissue-, and/or developmental-time-specific genes that are present or absent prior to irradiation by happenstance in the samples of some studies, but not others. For example, in 2 studies, 3–5 (Brodsky et al. 2004) or 2.5–5 (Akdemir et al. 2007) hour embryos were used, whereas van Bergeijk et al. (2012) used wing discs from third instar larvae. Bradysia and Drosophila share many common pathways in response to ionizing radiation even though the specific genes that are differentially expressed in the pathways can vary.

In our study, the PARP family member BcPALP1 has the highest upregulation in response to irradiation, but PARP or PARP domains are not mentioned in the 3 Drosophila studies. Similarly, we identified the radiation-induced up-regulation of genes involved in small RNA pathways: r2d2 and 3 Argonaute homologs (1 each that was similar to Ago3, PIWI, and Ago1/Ago2). In contrast, we only found AGO2 up-regulated in a table of 1 of the 3 Drosophila studies (van Bergeijk et al. 2012), which did not focus on this gene in the text. The specific B. coprophila ortholog corresponding to the D. melanogaster AGO2 gene in that study had no evidence of up-regulation in Bradysia. More generally, none of the 3 Drosophila studies focused on the role of small RNAs and small RNA processing in response to irradiation.

The 3 Drosophila studies were done on diploid cells where radiation can induce apoptosis, and this is also the case for young Drosophila embryos that are primarily diploid (Brodsky et al. 2004). In contrast, the vast majority of tissues in Dipteran larvae are polytene (Cooper 1938), and radiation does not induce apoptosis in the endocycling larval polyploid cells (Mehrotra et al. 2008; Herriage et al. 2023), due in part to low levels of p53 (Zhang et al. 2014). B. coprophila appears to differ in some of these respects from Drosophila, and this may contribute to the somewhat greater radiation resistance of B. coprophila. We also found that B. coprophila has 2 p53 orthologs, compared to 1 in Drosophila, and has an ortholog for mammalian DNA-PKcs, which was lost in the Drosophila lineage, and which was up-regulated in response to radiation in B. coprophila larvae. Of the 4 major proapoptotic genes that were up-regulated in response to radiation in all 3 Drosophila studies (hid, rpr, corp, skl), only 1 (hid) was up-regulated in Bradysia, and it was the only one with an ortholog in this species. However, Bradysia also may have up to 4 orthologs of “hid” that are up-regulated in response to radiation. While specific radiation-induced DNA repair genes seem more broadly conserved as up-regulated targets of radiation than specific proapoptotic genes, apoptotic genes are nevertheless enriched in the sets of radiation response genes from both species. Overall, despite many conserved genes, corresponding response pathways, and overlapping transcriptional responses spanning the Dipteran tree found in this study, there also appears to be enough differences between these 2 lineages to allow the possibility that they may also differ in their robustness to radiation, which reflects our observation, for example, that 87% irradiated B. coprophila larvae but no D. melanogaster larvae eclose after 40 Gy radiation (Table 3).

Data availability

Bradysia (Sciara) coprophila stocks are available upon request. The authors affirm that all data necessary for confirming the conclusions of the article are present within the article, figures, and table. Illumina RNA-seq reads have been submitted to the NCBI BioProject database under accession number PRJNA928089 (https://www.ncbi.nlm.nih.gov/bioproject/PRJNA928089; BioSamples SAMN32919381 and SAMN32919590), and the unprocessed fastq files can be found from the BioProject link above, or directly in the Sequence Read Archive (SRA) under the BioProject accession PRJNA928089 (https://www.ncbi.nlm.nih.gov/sra/PRJNA928089) or under the SRA study accession SRP419156. The datasets have the following individual accessions: SRR23237574, SRR23237573, SRR23237572, SRR23237571, SRR23237570, and SRR23237569.

Supplemental material available at GENETICS online.

Funding

JMU was supported by predoctoral traineeships from the National Institutes of Health (NIH) (T32 GM007601), National Science Foundation (NSF)/EPSCoR (#1004057) and an NSF predoctoral fellowship (GRFP-DGE-1058262); JRB, KRG, AB, and BJT were supported by grants from the NIH (P20 GM0103423 and R15 GM132896-01); SAG was funded by the NIH (R01 GM121455).