Article Figures & Tables
Figures
- Fig. 1. Summary of the analyses performed.
This figure summarizes the analyses performed in this work, indicating the specific questions addressed and the datasets used. For each human protein-coding duplicated gene (PGD) we determined: (1) its duplication age, (2) whether it is within a CNV region in current human populations, and (3) its replication timing (RT) during S phase. We use this gene-centered information to investigate the involvement of CNVs in gene birth through duplication during human evolution and the possible influence of replication timing in these gene duplication events.
- Fig. 2. Phylostratification of human PDGs.
(A) The age of a duplicated gene represents the ancestral species in which the duplication event that led to the generation of the extant gene was detected. A total of 13,909 gene duplicates were assigned to one of the 14 different evolutionary age groups (or phylostrata). Representative extant species that define the gene age classes are indicated (see Table 1 for the complete list). (B) The proportion of CNV genes in each phylostratum is higher in the genes recently duplicated in evolution (P-value <10−150, chi-squared test). A similar result was observed when only CNV gains are considered (supplementary material Fig. S1).
- Fig. 3. Gene duplications, CNVs and RT.
(A) The box plots represent the RT of all human protein-coding genes. The RT was obtained from publicly available microarray-based RT maps. A total of 19,197 human genes were ranked from early to late according to their order of replication. Genes located in CNV regions (CNV genes) replicate later (P-value = 3.4×10−15, Wilcoxon's test). (B) PDGs in CNV regions replicate later than non-CNV PDGs (P-value = 1.3×10−15), a difference that was not observed for singleton genes (P-value = 0.40). (C) Young PDGs (genes duplicated in the primate phylostrata) are preferentially located in CNV regions that replicate late (P-value = 3.8×10−4, Wilcoxon's test), whereas the difference between CNV and non-CNV PDGs is not significant in older duplicates (P-value = 0.41). Note that PDGs duplicated during Primate evolution tend to replicate later than older genes (P-value = 3.9×10−112). The box width is proportional to the number of genes within each figure panel.
- Fig. 4. RT mirrors gene duplication phylogeny.
(A) RT distribution of human PDGs is correlated with duplication age (rho = 0.21, P-value = 5.1×10−150, Spearman's correlation). (B) RT distribution of mouse PDGs is also correlated with duplication age (rho = 0.28, P-value = 5.8×10−278). The box width is proportional to the number of PDGs within each figure panel, and the specific human and mouse lineage age classes are indicated in bold. See also supplementary material Figs S2–S4.
- Fig. 5. The association of PDG age and RT is observed in different human and mouse chromosomal regions.
(A) Human pericentromeric regions (rho = 0.44, P-value = 1.1×10−47, Spearman's rank correlation). (B) Human interstitial regions (rho = 0.18, P-value = 2.7×10−84). (C) Human subtelomeric regions (rho = 0.23, P-value = 5.2×10−24). (D) Mouse pericentromeric regions (rho = 0.17, P-value = 2.0×10−4). (E) Mouse interstitial regions (rho = 0.29, P-value = 5.6×10−255). (F) Mouse subtelomeric regions (rho = 0.32, P-value = 3.6×10−23). Subtelomeric and pericentromeric PDGs were defined as those within 5 Mb of the telomere or centromere, respectively. The rest of the PDGs are considered to be in interstitial regions. The box width is proportional to the number of PDGs within each figure panel.
- Fig. 6. Proposed model based on our observations and previous knowledge.
According to our results, a bias in CNV formation (probably associated with replicative stress) leads to the accumulation of CNV-genes in heterochromatin-rich, late-replicating regions. This scenario increases the intrinsic probability that new gene copies are located in these regions. In the long term, a recurrence of this situation combined with successive selection events would lead to the progressive accumulation of younger genes in late-replicating regions. The location of new genes in heterochromatin would favor the development of cell type-specific patterns of gene expression. This restriction on gene expression will reduce the selection pressure on new genes, resulting in a weaker impact on the whole organism. In this scenario the rapid development of new traits would contribute to the differential evolution of distinct cell types. Obviously, the influence of other unexplored factors would be expected and should not be ruled out.
Tables
- Table 1. List of phylostrata used in the phylogenetic reconstructions.
Article Navigation
Related articles
Cited by...
More in this TOC section
Similar articles
Other journals from The Company of Biologists
Biology Open and COVID-19
We are aware that the COVID-19 pandemic is having an unprecedented impact on researchers worldwide. The Editors of all The Company of Biologists’ journals have been considering ways in which we can alleviate concerns that members of our community may have around publishing activities during this time. Read about the actions we are taking at this time.
Please don’t hesitate to contact the Editorial Office if you have any questions or concerns.
Interview – Savant Thakur
Savant Thakur was a Ph.D. student in Professor Gordon S. Lynch’s lab (The University of Melbourne) working towards understanding the mechanisms of defective muscle repair in muscle diseases and wasting disorders. Sadly, Savant passed away on June 16, 2019 and so one of his supervisors, Professor Lynch, spoke to BiO about Savant's work and character.
A day in the life of a returning spider lab
Back in 2016, two PhD students in Alistair McGregor’s lab at Oxford Brookes shared a day in the life of a spider lab on the Node. Four years and one pandemic later, we caught up with Alistair to find out how his lab has responded to the ongoing Covid-19 crisis.