Reduction of Paraoxonase Expression Followed by Inactivation across Independent Semiaquatic Mammals Suggests Stepwise Path to Pseudogenization

Abstract

Convergent adaptation to the same environment by multiple lineages frequently involves rapid evolutionary change at the same genes, implicating these genes as important for environmental adaptation. Such adaptive molecular changes may yield either change or loss of protein function; loss of function can eliminate newly deleterious proteins or reduce energy necessary for protein production. We previously found a striking case of recurrent pseudogenization of the Paraoxonase 1 (Pon1) gene among aquatic mammal lineages-Pon1 became a pseudogene with genetic lesions, such as stop codons and frameshifts, at least four times independently in aquatic and semiaquatic mammals. Here, we assess the landscape and pace of pseudogenization by studying Pon1 sequences, expression levels, and enzymatic activity across four aquatic and semiaquatic mammal lineages: pinnipeds, cetaceans, otters, and beavers. We observe in beavers and pinnipeds an unexpected reduction in expression of Pon3, a paralog with similar expression patterns but different substrate preferences. Ultimately, in all lineages with aquatic/semiaquatic members, we find that preceding any coding-level pseudogenization events in Pon1, there is a drastic decrease in expression, followed by relaxed selection, thus allowing accumulation of disrupting mutations. The recurrent loss of Pon1 function in aquatic/semiaquatic lineages is consistent with a benefit to Pon1 functional loss in aquatic environments. Accordingly, we examine diving and dietary traits across pinniped species as potential driving forces of Pon1 functional loss. We find that loss is best associated with diving activity and likely results from changes in selective pressures associated with hypoxia and hypoxia-induced inflammation.

Keywords: convergent evolution‌; genomics; hypoxia; marine; pseudogene.

Fig. 1.

Summary of Pon1 pseudogenization across multiple independent aquatic mammal lineages—species whose labels are highlighted in light blue are aquatic or semiaquatic. The colored circles represent presence of results from DNA, RNA, or enzymatic tests; no circle means there are no data to report. Phylogenetic relationships shown are based on data from previously published work (Esselstyn et al. 2017; Swanson et al. 2019; McGowen et al. 2020; de Ferran et al. 2022). Silhouette images are from Phylopic (CC BY‐SA 3.0).

Fig. 2.

Blood plasma from aquatic and semiaquatic species varies in enzymatic activity against Pon1 substrates. Plots show enzymatic activity in units/mL against (from L to R) chlorpyrifos oxon, diazoxon, and phenyl acetate for aquatic/semiaquatic and closely related terrestrial species within (from top to bottom) Cetartiodactyla, Carnivora, and Rodentia. Carnivora species included are M. angustirostris (Monachinae), P. vitulina (Phocinae), O. rosmarus (Odobenidae), E. jubatus (9) and Z. californianus (1) (Otariidae), M. furo (Mustelinae), and A. cinereus (2) and L. canadensis (1) (Lutrinae). Data for Bovids (sheep, goat) and one Muroid are from Furlong et al. (2000). See also Figure S1 and Table S12.

Fig. 3.

Transcriptional activity levels for the three Pon genes in three lineages with independent aquatic transitions (TPM; Transcripts per Million). Each dot represents a value for a singular species—some species had multiple individuals, and therefore their dot is an average value (Supplemental Table S1).

Fig. 4.

Fitted phylogenetic logistic regression of Pon1 status against (A) predicted myoglobin concentration in mg/g (P = 0.051) and (B) calculated aerobic dive limit in minutes (P = 0.1223), estimated using data for nine pinniped species. In (A), both California and Steller Sea Lion have no functional Pon1 and the same predicted myoglobin concentration.

Fig. 5.

Schematic showing the inferred process by which Pon1 loses function during adaptation to aquatic environments. In Step 1, Pon1 is fully functional. In Step 2, a change to a regulatory element reduces transcription and hence Pon1 protein levels. In Step 3, the coding sequence of Pon1 acquires further changes influencing its structure and/or function, leading to low and potentially dysfunctional protein. In Step 4, Pon1 function is fully lost, through a combination of changes preventing transcription and nonsense changes to the coding sequence (lesions), leading to no detectable RNA or protein. Silhouette images are from Phylopic (CC BY‐SA 3.0).