Genomic structural variation: A complex but important driver of human evolution

Abstract

Structural variants (SVs)-including duplications, deletions, and inversions of DNA-can have significant genomic and functional impacts but are technically difficult to identify and assay compared with single-nucleotide variants. With the aid of new genomic technologies, it has become clear that SVs account for significant differences across and within species. This phenomenon is particularly well-documented for humans and other primates due to the wealth of sequence data available. In great apes, SVs affect a larger number of nucleotides than single-nucleotide variants, with many identified SVs exhibiting population and species specificity. In this review, we highlight the importance of SVs in human evolution by (1) how they have shaped great ape genomes resulting in sensitized regions associated with traits and diseases, (2) their impact on gene functions and regulation, which subsequently has played a role in natural selection, and (3) the role of gene duplications in human brain evolution. We further discuss how to incorporate SVs in research, including the strengths and limitations of various genomic approaches. Finally, we propose future considerations in integrating existing data and biospecimens with the ever-expanding SV compendium propelled by biotechnology advancements.

Keywords: brain; gene duplications; genomes; human evolution; structural variation.

Figure 1.. Examples of genomic structural variation.

SVs exist as deletions and duplications (with the largest, most similar duplications termed segmental duplications, or SDs) that change the copy of a genomic segment (i.e., CNVs). Other types of SVs include insertions, translocations, inversions, as well as more complex events not pictured. Figure is adapted from (Alkan, Coe, et al., 2011) via “Genome Structural Variations” by BioRender.com (2022). Retrieved from https://app.biorender.com/biorender-templates.

Figure 2.. Cladogram of Hominidae family.

Divergence time estimates were obtained from Sudmant et al. (2013).

Figure 3.. Difficulties appraising haplotypes between SVs and neighboring SNVs.

(A) Neighboring SNVs (orange circles) are difficult to detect when an SV (colored rectangle) is embedded in repeat-rich regions. (B, C) Haplotypes can be disrupted by (B) IGC and (C) recurrent deletions (H) and duplication (H’). (D) mCNVs can be in the same locus (H) or several kilobases apart (H’).

Figure 4.. Summary of human duplicated genes implicated in neurodevelopmental functions.

Partially duplicated SRGAP2C antagonizes ancestral SRGAP2A through dimerization of their F-BAR domains, causing its degradation and resulting in increases in the rate of neuronal migration, density of dendritic spines, long-range neuronal connections, and synapse density, as well as delayed neuronal maturation. Truncated ARHGAP11B carries 55 distinct terminal amino acids that result in a loss of its ancestral RhoGAP activity, increasing calcium levels in the mitochondria that result in increased glutaminolysis and a higher abundance of basal progenitors that lead to presence of gyrification. HSE of TBC1D3 located in a core duplicon has been expanded multiple times. Studies have revealed an increase in cortical progenitors and subsequently neurons in the presence of this gene, ultimately resulting in mice with gyrencephalic brains. Incomplete duplication of the N-terminal portion of NOTCH2 and subsequent expansion gave rise to several NOTCH2NL paralogs that remained functional likely due to IGC events and that have been found to directly increase the abundance of cortical progenitors and neurons. Figure created with BioRender.com.

Figure 5.. Transcriptional profiles of a subset HSD genes across corticogenesis using data from differentiated hESCs.

RNA-seq quantification across a time-course of 77 days to mimic developmental stages including: pluripotency (plur.), differentiation (diff.), cortical specification (cort. spec.), deep layer formation (form.) and upper layer formation (upp. lay. form.). Expression levels displayed as z-scores of HSD genes in relation to the complete transcriptome indicated as colors (red = high; blue = low expression).

Figure 6.. Short- and long-read SV discovery signals.

R: Reference. S: Sample. SRS: short-read sequencing. LRS: long-read sequencing. Dashed line connects two pairs from the same short-read sequencing DNA fragment. Pink shapes represent long reads.

Figure 7.. Genomic artifacts arising from errors in the human reference genome assembly.

False positive heterozygous SNV calls originating from missing copies in the reference due to identification of paralog-sequence variants due to reads mapping from multiple paralogs.

Figure 8.. Differences in mappability between short and long reads in duplicated genes.

Paralog-specific variants (PSVs) (vertical lines) can distinguish paralogs enabling detection of polymorphic variation (yellow dots). Reads that do not carry PSVs (dashed lines) are unmappable in duplicated regions. SRS: short-read sequencing. LRS: long-read sequencing.