A genomics approach reveals insights into the importance of gene losses for mammalian adaptations

Abstract

Identifying the genomic changes that underlie phenotypic adaptations is a key challenge in evolutionary biology and genomics. Loss of protein-coding genes is one type of genomic change with the potential to affect phenotypic evolution. Here, we develop a genomics approach to accurately detect gene losses and investigate their importance for adaptive evolution in mammals. We discover a number of gene losses that likely contributed to morphological, physiological, and metabolic adaptations in aquatic and flying mammals. These gene losses shed light on possible molecular and cellular mechanisms that underlie these adaptive phenotypes. In addition, we show that gene loss events that occur as a consequence of relaxed selection following adaptation provide novel insights into species' biology. Our results suggest that gene loss is an evolutionary mechanism for adaptation that may be more widespread than previously anticipated. Hence, investigating gene losses has great potential to reveal the genomic basis underlying macroevolutionary changes.

Fig. 1

Genomics approach to detect gene loss events. a The different steps address a number of challenges related to assembly and alignment issues, and evolutionary changes in gene structures. b Applied to 13,486 human genes that have annotated 1:1 orthologs in mouse, rat, cow, and dog, these steps systematically reduce the number of conserved genes that have inactivating mutations. A total of 85–91% of the genes remaining after step 7 had inactivating mutations only in a single exon that is not entirely conserved (Supplementary Figs. 10 and 11). This shows that mutations in an individual exon of an otherwise-conserved gene is not sufficient to infer gene loss. By requiring that inactivating mutations occur in multiple exons and that less than 60% of the reading frame remains intact, our approach misclassifies ≤0.3% of 13,486 conserved genes as lost

Fig. 2

Adaptations of the cetacean epidermis to the aquatic environment. The figure shows genes with hair- and epidermis-related functions that are specifically lost in cetaceans. The expression pattern of these genes in the skin is shown as gray lines and boxes (expression gradients are indicated). Mice in which these genes are knocked out show epidermal phenotypes that strongly resemble morphological adaptations of the cetacean skin. Since the loss of DSG4, DSC1, TGM5, and GSDMA coincided with a period during which epidermal adaptations evolved in cetaceans (Supplementary Note 1, Supplementary Figs. 14–17, and Supplementary Table 5), these gene losses could have played a causal role in the remodeling of cetacean epidermis. The cetacean-specific loss of ALOXE3, an atypical lipoxygenase that is important for skin barrier function, happened after the split of the baleen and toothed whale lineage (Supplementary Fig. 18 and Supplementary Table 5) and is thus presumably a consequence of epidermal adaptations in these lineages

Fig. 3

Diving and dietary adaptations in sperm whales. The loss of AMPD3 (red, Supplementary Fig. 19) is likely an adaptation to the extreme diving ability of sperm whales. AMPD3 deaminates adenosine monophosphate (AMP) to inosine monophosphate (IMP) in erythrocytes. AMPD3 loss increases the level of ATP (an allosteric hemoglobin effector), which facilitates O₂ release, as illustrated by the O₂-hemoglobin dissociation curve (wildtype, black; AMPD3 knockout, red). In contrast, the loss of the vitamin A synthesizing enzyme BCO1 (blue, Supplementary Fig. 20) in sperm whales is likely a consequence of relaxed selection after sperm whales adapted to their specialized diet that mainly consists of vitamin A-rich but beta-carotene poor squid. The absence of its substrate (beta-carotene) likely made this enzyme obsolete, leading to the loss of BCO1

Fig. 4

Renal and metabolic adaptations in frugivorous bats. A number of renal transporter genes (red, left side) that are specifically lost in fruit bats (large and black flying foxes) reduce urine osmolality in a mouse knockout. Thus, these gene losses likely contribute to the ability of fruit bats to efficiently excrete excess dietary water. Losses of metabolic genes (red, right side) are likely adaptive by improving the processing of the sugar-rich fruit juice. In contrast, gene losses shown in blue are probably a consequence of adapting to the frugivorous diet. These genes provide new insights into the metabolism of bats and corroborate the strong dependence of internal organs on using sugar as the main energy source. Please see Supplementary Figs. 21–31 for the inactivating mutations in these 11 genes

Fig. 5

Convergent gene losses and repeated phenotypic adaptations. a Phylogenetic tree showing the independent lineages that share a derived phenotype (loss of teeth or enamel, body armor in the form of scales, fully aquatic lifestyle). b A new Forward Genomics method discovered genes that are preferentially lost in mammals with the derived phenotype (red dots) compared to other mammals (blue dots). The panel shows the maximum percentage of the intact reading frame of these genes in each species and summarizes gene function (see also main text, Supplementary Notes 6–8 and Supplementary Figs. 33–35). ACP4 has a %intact reading frame value of 86% in the Tibetan antelope (blue dot); however, no unassembled sequencing reads are available to validate the mutation. For some species, we could not compute the maximum intact reading frame due to missing data, therefore the total number of evaluated species slightly differs per gene (Supplementary Tables 6–8)