Beetles are the most species-rich group of animals and harbor diverse karyotypes. Most species have XY sex chromosomes, but X0 sex determination mechanisms are also common in some groups. We generated a whole-chromosome assembly of a beetle species with a neo-sex chromosome and utilize eleven additional beetle genomes, to reconstruct karyotype evolution across Coleoptera. We identify ancestral linkage groups, termed Stevens elements, that share a conserved set of genes across beetles. While the ancestral X chromosome is maintained across beetles, we find independent additions of autosomes to the ancestral sex chromosomes. These neo-sex chromosomes evolve the stereotypical properties of sex chromosomes, including the evolution of dosage compensation, and a non-random distribution of genes with sex-biased expression. Beetles thus provide a novel model to gain a better understanding of the diverse forces driving sex chromosome evolution.

The Drosophila obscura species group is one of the most studied clades of Drosophila and harbors multiple distinct karyotypes. Here we present a de novo genome assembly and annotation of D. bifasciata, a species which represents an important subgroup for which no high-quality chromosome-level genome assembly currently exists. We combined long-read sequencing (Nanopore) and Hi-C scaffolding to achieve a highly contiguous genome assembly approximately 193 Mb in size, with repetitive elements constituting 30.1% of the total length. Drosophila bifasciata harbors four large metacentric chromosomes and the small dot, and our assembly contains each chromosome in a single scaffold, including the highly repetitive pericentromeres, which were largely composed of Jockey and Gypsy transposable elements. We annotated a total of 12,821 protein-coding genes and comparisons of synteny with D. athabasca orthologs show that the large metacentric pericentromeric regions of multiple chromosomes are conserved between these species. Importantly, Muller A (X chromosome) was found to be metacentric in D. bifasciata and the pericentromeric region appears homologous to the pericentromeric region of the fused Muller A-AD (XL and XR) of pseudoobscura/affinis subgroup species. Our finding suggests a metacentric ancestral X fused to a telocentric Muller D and created the large neo-X (Muller A-AD) chromosome ∼15 MYA. We also confirm the fusion of Muller C and D in D. bifasciata and show that it likely involved a centromere-centromere fusion.

7The Drosophila obscura species group is one of the most studied clades of Drosophila and 8 harbors multiple distinct karyotypes. Here we present a de novo genome assembly and 9 annotation of D. bifasciata, a species which represents an important subgroup for which no high-10 quality chromosome-level genome assembly currently exists. We combined long-read 11 sequencing (Nanopore) and Hi-C scaffolding to achieve a highly contiguous genome assembly 12 approximately 193Mb in size, with repetitive elements constituting 30.1% of the total length. 13Drosophila bifasciata harbors four large metacentric chromosomes and the small dot, and our 14 assembly contains each chromosome in a single scaffold, including the highly repetitive 15 pericentromere, which were largely composed of Jockey and Gypsy transposable elements. We 16 annotated a total of 12,821 protein-coding genes and comparisons of synteny with D. athabasca 17 orthologs show that the large metacentric pericentromeric regions of multiple chromosomes are 18 conserved between these species. Importantly, Muller A (X chromosome) was found to be 19 metacentric in D. bifasciata and the pericentromeric region appears homologous to the 20 pericentromeric region of the fused Muller A-AD (XL and XR) of pseudoobscura/affinis 21 subgroup species. Our finding suggests a metacentric ancestral X fused to a telocentric Muller D 22 and created the large neo-X (Muller A-AD) chromosome ~15 MYA. We also confirm the fusion 23 2 of Muller C and D in D. bifasciata and show that it likely involved a centromere-centromere 24 fusion. 25 26 INTRODUCTION 27 Recent advances in DNA sequencing technology have dramatically improved the quality and 28 quantity of genome assemblies in both model and non-model species. Long-read sequencing 29 technologies (e.g., PacBio and Nanopore) combined with long-range scaffolding information 30 generated through chromatin conformation capture methods such as Hi-C (Lieberman-Aiden et 31 al. 2009) or Chicago (Putnam et al. 2016) can produce assemblies of unprecedented length and 32 accuracy. However, there are still relatively few assemblies that traverse through megabase-long 33 stretches of highly repetitive sequence, thereby limiting our understanding of the evolution of 34 pericentromere/heterochromatic regions of the genome and the genes, satellites, and transposable 35 elements that inhabit them (Chang et al. 2019, Miga 2019). 36 Drosophila has been at the forefront of genetics and genomics research for over a century 37 and new chromosome-level assemblies are now becoming available for several non-model 38 species (Mahajan et al. 2018, Miller et al. 2018, Bracewell et al. 2019, Karageorgiou et al. 2019, 39 Mai et al. 2019). Recent comparative genomic analysis in the Drosophila obscura group has 40 revealed extensive karyotype evolution and turnover of centromeric satellites that alters 41 chromosome morphology (Bracewell et al. 2019) (Figure 1). Unfortunately, our understanding of 42 karyotype and genome evolution is currently limited because no high-q...