Dynamic Evolution of Repetitive Elements and Chromatin States in Apis mellifera Subspecies

Abstract

In this study, we elucidate the contribution of repetitive DNA sequences to the establishment of social structures in honeybees (Apis mellifera). Despite recent advancements in understanding the molecular mechanisms underlying the formation of honeybee castes, primarily associated with Notch signaling, the comprehensive identification of specific genomic cis-regulatory sequences remains elusive. Our objective is to characterize the repetitive landscape within the genomes of two honeybee subspecies, namely A. m. mellifera and A. m. ligustica. An observed recent burst of repeats in A. m. mellifera highlights a notable distinction between the two subspecies. After that, we transitioned to identifying differentially expressed DNA elements that may function as cis-regulatory elements. Nevertheless, the expression of these sequences showed minimal disparity in the transcriptome during caste differentiation, a pivotal process in honeybee eusocial organization. Despite this, chromatin segmentation, facilitated by ATAC-seq, ChIP-seq, and RNA-seq data, revealed a distinct chromatin state associated with repeats. Lastly, an analysis of sequence divergence among elements indicates successive changes in repeat states, correlating with their respective time of origin. Collectively, these findings propose a potential role of repeats in acquiring novel regulatory functions.

Keywords: Hymenoptera; TE); chromatin landscape; honeybee Apis mellifera; noncoding RNA (ncRNA); repetitive DNA; transposable elements (transposons.

Figure 1

Pattern of repetitive elements in the A. m. mellifera and A. mellifera ligustica genomes. (A) Interspersed repeat landscapes. TE classes are marked with colors: DNA transposons, green; LTR TE, red; rolling-circle TEs (helitrons), yellow; unknown repeats, blue and purple; (B) the classification of repetitive sequences in the genomes; and (C) the distribution of rnd-4_family-321 elements on chromosome maps.

Figure 2

Features of chromatin states and genome occupancy in A. m. mellifera brains. (A) Features of the different chromatin states in queens/drones. Columns from left to right—occupied genome fraction (purple), features comprising states (in blue), and relative enrichment of respective genomic regions (in green); (B) the state transition graph; transition probabilities < 0.05 are not shown. The edge thickness indicates the probability of transition.

Figure 3

Localization features of E13 chromatin state regions and GO analysis of their closest genes. (A) Venn diagrams indicating intersection between closest genes to E13 regions in different castes. (B) Lollipop diagram indicating the Gene Ontology molecular function enrichment of closest genes to E13.

Figure 4

Bar plots depicting chromatin states in young (Kimura 0–3) and middle-aged (Kimura 20–40) repeats. (Top) Young repeats; (bottom) middle-aged repeats. Stacked bar graphs illustrate the proportional genome occupancy of each chromatin state across various Kimura distance intervals.

Figure 5

The MA plot visualizes expression differences of the genes and repeats in queen and worker larvae. The x-axis represents the average expression level, while the y-axis depicts the log-fold change, highlighted with orange—transcripts with differential expression.

Figure 6

Clustering of single-cell transcriptome of the honey bee brain. (A) The UMAP of single cells from the brain of a honey bee, divided into 5 main clusters: hemocytes; glial cells; olfactory projection neurons (OPNs); optic lobe cells (OLCs); and Kenyon cells (KCs). OLCs were subdivided into Tm5c, Lawf2, and PM neurons. Kenyon cells were subdivided into class I small KCs and class I large KCs. (B) Dot plot of predictive gene markers used for cluster annotation. The black frame shows the marker/combination of markers used to identify the cell population.

Figure 7

Dot plot of eight REs predicted as markers. Among the identified markers, the highest expression level (average expression 2.5) was observed for the Copia-5/1071 in cluster 11, which was identified as olfactory projection neurons. Additionally, the EnSpm-5/1919 showed high expression in cluster 17, characterized as hemocytes.