Understanding the Early Evolutionary Stages of a Tandem Drosophilamelanogaster-Specific Gene Family: A Structural and Functional Population Study

Gene families underlie genetic innovation and phenotypic diversification. However, our understanding of the early genomic and functional evolution of tandemly arranged gene families remains incomplete as paralog sequence similarity hinders their accurate characterization. The Drosophila melanogaster-specific gene family Sdic is tandemly repeated and impacts sperm competition. We scrutinized Sdic in 20 geographically diverse populations using reference-quality genome assemblies, read-depth methodologies, and qPCR, finding that ∼90% of the individuals harbor 3-7 copies as well as evidence of population differentiation. In strains with reliable gene annotations, copy number variation (CNV) and differential transposable element insertions distinguish one structurally distinct version of the Sdic region per strain. All 31 annotated copies featured protein-coding potential and, based on the protein variant encoded, were categorized into 13 paratypes differing in their 3' ends, with 3-5 paratypes coexisting in any strain examined. Despite widespread gene conversion, the only copy present in all strains has functionally diverged at both coding and regulatory levels under positive selection. Contrary to artificial tandem duplications of the Sdic region that resulted in increased male expression, CNV in cosmopolitan strains did not correlate with expression levels, likely as a result of differential genome modifier composition. Duplicating the region did not enhance sperm competitiveness, suggesting a fitness cost at high expression levels or a plateau effect. Beyond facilitating a minimally optimal expression level, Sdic CNV acts as a catalyst of protein and regulatory diversity, showcasing a possible evolutionary path recently formed tandem multigene families can follow toward long-term consolidation in eukaryotic genomes.

Characterization and evolutionary dynamics of complex regions in eukaryotic genomes

Complex regions in eukaryotic genomes are typically characterized by duplications of chromosomal stretches that often include one or more genes repeated in a tandem array or in relatively close proximity. Nevertheless, the repetitive nature of these regions, together with the often high sequence identity among repeats, have made complex regions particularly recalcitrant to proper molecular characterization, often being misassembled or completely absent in genome assemblies. This limitation has prevented accurate functional and evolutionary analyses of these regions. This is becoming increasingly relevant as evidence continues to support a central role for complex genomic regions in explaining human disease, developmental innovations, and ecological adaptations across phyla. With the advent of long-read sequencing technologies and suitable assemblers, the development of algorithms that can accommodate sample heterozygosity, and the adoption of a pangenomic-like view of these regions, accurate reconstructions of complex regions are now within reach. These reconstructions will finally allow for accurate functional and evolutionary studies of complex genomic regions, underlying the generation of genotype-phenotype maps of unprecedented resolution.

Detecting PKD1 variants in polycystic kidney disease patients by single-molecule long-read sequencing

A genetic diagnosis of autosomal-dominant polycystic kidney disease (ADPKD) is challenging due to allelic heterogeneity, high GC content, and homology of the PKD1 gene with six pseudogenes. Short-read next-generation sequencing approaches, such as whole-genome sequencing and whole-exome sequencing, often fail at reliably characterizing complex regions such as PKD1. However, long-read single-molecule sequencing has been shown to be an alternative strategy that could overcome PKD1 complexities and discriminate between homologous regions of PKD1 and its pseudogenes. In this study, we present the increased power of resolution for complex regions using long-read sequencing to characterize a cohort of 19 patients with ADPKD. Our approach provided high sensitivity in identifying PKD1 pathogenic variants, diagnosing 94.7% of the patients. We show that reliable screening of ADPKD patients in a single test without interference of PKD1 homologous sequences, commonly introduced by residual amplification of PKD1 pseudogenes, by direct long-read sequencing is now possible. This strategy can be implemented in diagnostics and is highly suitable to sequence and resolve complex genomic regions that are of clinical relevance.