Multi-level analysis of reproduction in an Antarctic midge identifies female and male accessory gland products that are altered by larval stress and impact progeny viability

Abstract

The Antarctic midge, Belgica antarctica, is a wingless, non-biting midge endemic to Antarctica. Larval development requires at least 2 years, but adults live only 2 weeks. The nonfeeding adults mate in swarms and females die shortly after oviposition. Eggs are suspended in a gel of unknown composition that is expressed from the female accessory gland. This project characterizes molecular mechanisms underlying reproduction in this midge by examining differential gene expression in whole males, females, and larvae, as well as in male and female accessory glands. Functional studies were used to assess the role of the gel encasing the eggs, as well as the impact of stress on reproductive biology. RNA-seq analyses revealed sex- and development-specific gene sets along with those associated with the accessory glands. Proteomic analyses were used to define the composition of the egg-containing gel, which is generated during multiple developmental stages and derived from both the accessory gland and other female organs. Functional studies indicate the gel provides a larval food source as well as a buffer for thermal and dehydration stress. All of these function are critical to juvenile survival. Larval dehydration stress directly reduces production of storage proteins and key accessory gland components, a feature that impacts adult reproductive success. Modeling reveals that bouts of dehydration may have a significant impact on population growth. This work lays a foundation for further examination of reproduction in midges and provides new information related to general reproduction in dipterans. A key aspect of this work is that reproduction and stress dynamics, currently understudied in polar organisms, are likely to prove critical in determining how climate change will alter their survivability.

Figure 1

Antarctic midge, Belgica antarctica, during reproduction. (A) Mating pair, male on left. Inset, spermatophore transferred to females immediately after copulation. Image is posterior end of female and white material is the spermatophore. (B) Female depositing eggs and accessory gland-derived gel. (C) Accessory gland (left middle, b) and ovaries (top left and right, a) of gravid females 3 days after adult eclosion. (D) Female accessory glands (left, a) and ovaries (top and right, b) following egg and gel deposition. (E) Egg mass following the completion of deposition. (F) Male reproductive tract, a. testes, b, accessory gland, and c, common duct.

Figure 2

Gene expression heat map of Antarctic midge, Belgica antarctica, during development, between sexes, and for accessory glands. Hierarchical clustering of RNA-seq gene expression patterns for males, females, larvae, and accessory glands based on sample distance (Euclidean distance matrix) of differentially expressed contigs.

Figure 3

Genes uniquely enriched for the Antarctic midge, Belgica antarctica in males, females, and larvae and associated gene ontology enrichment. (A) Gene enriched in males (left) and gene ontology (right), (B) gene enriched in females (left) and gene ontology (right), (C) gene enriched in larvae (left) and gene ontology (right). Each box represents a specific category and color represent major GO groups. F, female; M, male; L, larvae.

Figure 4

Genes uniquely enriched in the Antarctic midge, Belgica antarctica, female and male accessory glands and associated gene ontology enrichment. (A) Genes enriched in female accessory glands (left) and gene ontology (right), (B) genes enriched in male accessory gland (left) and gene ontology (right). GO conducted with g:Prolifer. F, female; M, male; L, larvae; MAG, Male accessory glands; FeAG, female accesspry glands.

Figure 5

Weighted gene co-expression network analysis (WGCNA) for larvae, adults, or for male and female accessory glands of the Antarctic midge, Belgica antarctica. (A) Average linkage hierarchical clustering dendrogram of the genes. Modules, designated by color code, are branches of the clustering tree. (B) Correlation of module eigengenes to larvae, adults, and accessory gland traits. Each row corresponds to a module eigengene and columns are traits. *Represents values with a significant positive correlation for Pearson r (P < 0.05). (C) Unsupervised hierarchical clustering heatmap (bottom) and dendrogram (top) of module eigengenes and traits. Gene ontology (GO) analysis of eigengenes associated with larvae (D), males (E), and females (F). GO conducted with g:Prolifer and visualized with REVIGO. Colors for modules between (A–C) are the same. Color patterns in (D–F) is random and size of the boxes represents the relative number of each specific GO categories.

Figure 6

Proteomic analysis of female accessory gland derived gel material from the Antarctic midge, Belgica antarctica. (A) Female depositing eggs with gel and protein components of two gel samples without eggs. Original gel images in Fig. S1. (B) Identification of proteins that represent at least 3% of total protein composition of the accessory gland gel. (C) Congruence of protein abundance and content between two gel samples. (D) Heatmap for transcript levels of gel-specific genes among larvae, females, males, and accessory glands. (E) qPCR validation of RNA-seq data. All tested genes have Pearson correlation coefficients over 0.85. Gel specific genes have a Pearson correlation of 0.87.

Figure 7

Transcription factors (TFs) and TF binding sites associated with reproduction in the Antarctic midge, Belgica antarctica. (A) Relative abundance of transcription factors encoded by genomes of midges and mosquitoes. Assignment of specific TFs to families is present in Table S14. TF families towards the top contain more TFs in B. antarctica. (B) Enrichment for binding site motifs for specific TFs in regulatory regions (2000 bp; 500 bp data not shown due to overlap with 2000 bp) of genes expressed highly in specific stages and accessory glands. Groups of TFs are separated by their motif enrichment profiles across samples. Those highlighted in orange are significantly enriched within the specific stage or tissue. Scale for heatmap is set at relative abundance on a Z scale of − 2 to 2 across each row. (C) Transcript levels of select TFs with significant motif enrichment in the promoters of genes expressed in specific tissues (orange font color in (B)). Scale for heatmap is set at relative abundance on a Z scale of − 2 to 2 across each row. MAG male accessory gland, FeAG female accessory glands.

Figure 8

Comparative analysis of female, male, and larvae-specific gene sets with mosquitoes and midges to Antarctic midge, Belgica antarctica. (A) Female-specific genes compared to midges (left), mosquitoes (middle), and genes with enriched expression in the female reproductive tract (FRT) of mosquitoes (right). (B) Male-specific genes compared to midges (left), mosquitoes (middle), and genes with enriched expression in the male reproductive tract (MRT) of mosquitoes (right). (C) Larvae-specific genes compared to midges (left) and mosquitoes (right). Protein sequences were defined as orthologs if they had reciprocal-best BLASTp hits with an e-value < 10⁻¹⁰.

Figure 9

Comparative analysis of accessory gland gene sets with mosquitoes and midges to the Antarctic midge. (A) Female accessory gland genes compared to midges (left), mosquitoes (middle), and genes with enriched expression in the female reproductive tract of mosquitoes (right). (B) Male accessory gland genes compared to midges (left), mosquitoes (middle), and genes with enriched expression in the male reproductive tract of mosquitoes (right). (C) Overlap between genes expressed in male accessory glands between mosquitoes and B. antarctica. Left, highly enriched in Anopheles male accessory gland. Right, enriched in Anopheles male accessory gland. Enrichment for Anopheles male accessory gland genes is based on values from Izquierdo et al.. Protein sequences were defined as orthologs if they had reciprocal-best BLASTp hits with an e-value < 10⁻¹⁰.

Figure 10

Expression changes in female accessory gland gel-associated proteins in larvae following dehydration stress. (A) Transcript level changes in larvae for gel proteins following dehydration stress. RNA-seq for dehydrated larvae were acquired from Teets et al.. Orange denotes significance between control and dehydrated larvae based on RNA-seq analyses (FDR < 0.05). (B) Heat map of transcript levels for gel-associated proteins during dehydration (D) and cryoprotective dehydration (CD) compared to control (C) that are components of the gel proteome and significantly altered by dehydration. (C) Total mass before birth, (D) total mass after, (E) total mass difference between before and after borth, and (F) total egg production in females when control (non-dehydrated) and dehydrated larvae were allowed to complete development. T-test was utilized to examine statistical differences with the use of R statistic packages. Bars above indicate significance at P < 0.05.

Figure 11

Impact of larval dehydration stress on male fertility. (A) Expression profiles of male-associated genes with expression differences after larval dehydration (D, dehydration, C, control, and CD, cryoprotective dehydration, MAG, male accessory gland). (B) Mass of males used in mating experiments from dehydrated or control larvae. (C) Viable eggs produced (control or dehydrated) following copulation with dehydrated or control males. Analysis of variance (ANOVA) was utilized to examine statistical differences with the use of R statistic packages. Tukey’s test was used as a post-hoc test to examine significance between each treatment. Bars above indicate significance at P < 0.05 unless otherwise noted.

Figure 12

Accessory gland gel is critical for larval development. (A) Amino acid composition and putative phosphate and glycosylation sites of gel proteins determined based on Fig. 6 based on relative number associated with each predicted protein sequence. Gene identification is based on those used in the B. antarctica genome. Relative amounts are based on comparison levels between columns. (B) Survival of developing larvae with (black) and without (gray) gel presence at larval ecdysis. Open circles are the average and filled circles are each replicate. (C) Larvae length after 20 days with and without gel at larval ecdysis. Bar indicates significance at P < 0.05. T-test was utilized to examine statistical differences with the use of R statistic packages. Bars above or beside indicate significance at P < 0.05.

Figure 13

Role of accessory gland gel in relation to thermal buffering of eggs. (A) Representative thermal profile within (orange) and outside (blue) gel-egg mixture under field conditions from Cormorant Island. (B) Egg viability following exposure to 20 °C for 3 h. Eggs without gel (constant) were held at 4 °C for the duration of the trial. Maximum temperature change during 24 h period. N = 6 for each treatment. (C) Maximum temperature change during the course of a single day. (D) Rate of temperature change (minimum to maximum). T-test or Analysis of variance (ANOVA) was utilized to examine statistical differences with the use of R statistic packages. Tukey’s test was used as a post-hoc test to examine significant between each treatment. *or letters above indicates significance at P < 0.05 unless otherwise noted.

Figure 14

Population growth is impacted by dehydration and thermal stress in developing larvae. (A) Population growth following altered egg production due to dehydration exposure as larvae in males, females, and both sexes combined compared to control (no dehydration of larvae). (B) Growth based on the presence or absence of the gel under favorable conditions. (C) Impact of thermal stress on egg viability with and without accessory gland gel.

Figure 15

Summary of Antarctic midge reproduction. Larval development (four stages) is condensed into a single representation for all stages. Adults live approximately 2–3 weeks. Egg development occurs over 30 days. Impact of specific conditions are highlighted based on experimental evidence from this study.