The birth of a human-specific neural gene by incomplete duplication and gene fusion

Abstract

BackgroundGene innovation by duplication is a fundamental evolutionary process but is difficult to study in humans due to the large size, high sequence identity, and mosaic nature of segmental duplication blocks. The human-specific gene hydrocephalus-inducing 2, HYDIN2, was generated by a 364 kbp duplication of 79 internal exons of the large ciliary gene HYDIN from chromosome 16q22.2 to chromosome 1q21.1. Because the HYDIN2 locus lacks the ancestral promoter and seven terminal exons of the progenitor gene, we sought to characterize transcription at this locus by coupling reverse transcription polymerase chain reaction and long-read sequencing.Results5' RACE indicates a transcription start site for HYDIN2 outside of the duplication and we observe fusion transcripts spanning both the 5' and 3' breakpoints. We observe extensive splicing diversity leading to the formation of altered open reading frames (ORFs) that appear to be under relaxed selection. We show that HYDIN2 adopted a new promoter that drives an altered pattern of expression, with highest levels in neural tissues. We estimate that the HYDIN duplication occurred ~3.2 million years ago and find that it is nearly fixed (99.9%) for diploid copy number in contemporary humans. Examination of 73 chromosome 1q21 rearrangement patients reveals that HYDIN2 is deleted or duplicated in most cases.ConclusionsTogether, these data support a model of rapid gene innovation by fusion of incomplete segmental duplications, altered tissue expression, and potential subfunctionalization or neofunctionalization of HYDIN2 early in the evolution of the Homo lineage.Electronic supplementary materialThe online version of this article (doi:10.1186/s13059-017-1163-9) contains supplementary material, which is available to authorized users.

The human-specific paralogs SRGAP2B and SRGAP2C differentially modulate SRGAP2A-dependent synaptic development

“…Restricting the available amount of SRGAP2A by targeting it to the proteasome represents an effective mechanism through which SRGAP2C increases synaptic density and delays maturation. Interestingly, HSGDs have often resulted in the formation of truncated paralogs and several of these truncated HSGDs have recently been implicated in regulating brain development [2][3][4]8,22,23]. This suggests that evolutionary changes affecting levels and patterns of expression not only occurred through modifications in regulators of gene expression at the genomic level [24], but also at the level of protein interactions through the emergence of insoluble proteins.…”

Human-specific NOTCH-like genes in a region linked to neurodevelopmental disorders affect cortical neurogenesis

“…The gene annotated as NOTCH2NL (Duan et al, 2004) on human genome assembly GRCh37 resides on the q arm of human chromosome 1 in the 1q21.1 locus, which long remained one of the most difficult regions in our genome to assemble due to its highly repetitive nature (Szamalek et al, 2006), (Dennis et al, 2012), (Doggett et al, 2006). The 1q21.1 locus has undergone a number of human lineage-specific rearrangements, including a pericentric inversion (Szamalek et al, 2006) and the creation of new humanspecific gene paralogs, including HYDIN2 (Dougherty et al, 2017) and SRGAP2B (Dennis et al, 2012). Resequencing of the pericentric region of chromosome 1 in a haploid human cell line finally resolved previously unmapped regions and led to a revised assembly of the 1q21.1 locus, which is incorporated in the human genome assembly GRCh38 (Steinberg et al, 2014).…”

“…Of particular interest are loci where segmental duplications have created entirely new human-specific gene paralogs that are associated with cortical development, such as SRGAP2C (Dennis et al, 2012), (Charrier et al, 2012), ARHGAP11B (Florio et al, 2015) and TBC1D3 (Ju et al, 2016). Interestingly, human-specific duplicated genes are often located within segmental duplications that mediate recurrent rearrangements associated with neurodevelopmental disorders (Stankiewicz and Lupski, 2010) (Florio et al, 2015), (Nuttle et al, 2016), (Popesco et al, 2006), (Dumas et al, 2012), (Dougherty et al, 2017). One region susceptible to these rearrangements lies on human chromosome band 1q21, which was involved in a large pericentric inversion involving considerable gene loss and duplication during human evolution (Szamalek et al, 2006), contains a disproportionate number of human-specific genes , and also contains the 1q21.1 distal deletion/duplication syndrome interval (Mefford et al, 2008), (Brunetti-Pierri et al, 2008).…”

“…This is due to several tracts of almost identical paralogous DNA. The most recent work (Dougherty et al, 2017) hypothesized that the typical breakpoints for the distal 1q21.1 deletion/duplication syndrome lie within two of these tracts, at ~chr1:146,200,000 and chr1:148,600,000 in GRCh38, respectively, in regions of ~250 kb at 99.7% identity. We confirm this, and show that these regions contain little-studied, paralogous genes we call NOTCH2NLA and NOTCH2NLB that affect Notch signaling during human neurodevelopment.…”

“…We confirm this, and show that these regions contain little-studied, paralogous genes we call NOTCH2NLA and NOTCH2NLB that affect Notch signaling during human neurodevelopment. There are also atypical breakpoints for the distal 1q21.1 syndrome (Dougherty et al, 2017) that reach further almost to a third NOTCH2NL paralog we call NOTCH2NLC. We show both the typical and atypical breakpoints lead to change in NOTCH2NL copy number.…”