Reconstructed ancestral enzymes suggest long-term cooling of Earth's photic zone since the Archean

Abstract

Paleotemperatures inferred from the isotopic compositions (δ¹⁸O and δ³⁰Si) of marine cherts suggest that Earth's oceans cooled from 70 ± 15 °C in the Archean to the present ∼15 °C. This interpretation, however, has been subject to question due to uncertainties regarding oceanic isotopic compositions, diagenetic or metamorphic resetting of the isotopic record, and depositional environments. Analyses of the thermostability of reconstructed ancestral enzymes provide an independent method by which to assess the temperature history inferred from the isotopic evidence. Although previous studies have demonstrated extreme thermostability in reconstructed archaeal and bacterial proteins compatible with a hot early Earth, taxa investigated may have inhabited local thermal environments that differed significantly from average surface conditions. We here present thermostability measurements of reconstructed ancestral enzymatically active nucleoside diphosphate kinases (NDKs) derived from light-requiring prokaryotic and eukaryotic phototrophs having widely separated fossil-based divergence ages. The ancestral environmental temperatures thereby determined for these photic-zone organisms--shown in modern taxa to correlate strongly with NDK thermostability--are inferred to reflect ancient surface-environment paleotemperatures. Our results suggest that Earth's surface temperature decreased over geological time from ∼65-80 °C in the Archean, a finding consistent both with previous isotope-based and protein reconstruction-based interpretations. Interdisciplinary studies such as those reported here integrating genomic, geologic, and paleontologic data hold promise for providing new insight into the coevolution of life and environment over Earth history.

Keywords: Precambrian; ancestral sequence reconstruction; enzyme thermostability; nucleoside diphosphate kinase; phototroph.

Fig. 1.

Phylogenetic trees used to evaluate ancestral NDK sequences. Nodes from which ancestral NDK sequences were calculated are indicated by red arrows. (Scale bars indicate the number of substitutions per site.) (A) An ML 16S rRNA tree of cyanobacteria based on the NDK sequences of outgroups (archaea and bacteria, black-filled), nonnostocalean cyanobacteria (white-filled), and nostocaleans (gray-filled). (B) An ML 18S rRNA tree of Viridiplantae, based on the NDK sequences showing non-Viridiplantae outgroups (black-filled), nonembryophyte Viridiplantae (white-filled), and Embryophyta (gray-filled). (C) An ML NDK tree constructed using NDK sequences of the Viridiplantae taxa used to construct B, the various parts colored like those in B.

Fig. 2.

Temperature dependence of ancestral NDK specific activity measured by the ATP produced, each value being the average of three or more measurements and the dotted lines spanning activity values measured at temperature intervals greater than the T_ms analyzed by circular dichroism spectroscopy. Data are not available for the specific activity of Cyano16S at temperatures greater than its T_m (∼100 °C).

Fig. 3.

Calibration curves showing the correlation between NDK T_ms and the organismal environmental growth temperatures of extant taxa of archaea and bacteria (22), cyanobacteria, and Viridiplantae. The best-fit linear regression lines for each group were constructed from the indicated data, the black-dashed line being the best fit through all available data and “r” specifying the calculated correlation coefficients. Environmental temperature ranges inferred from ancestral NDK T_ms are indicated by red bars for Cyano16S and Nosto16S and by pink bars for eukaryotic Viridi18S, ViridiNDK, Embryo18S, and EmbryoNDK.

Fig. 4.

Environmental temperature ranges inferred from reconstructed ancestral NDK T_ms plotted against fossil-record-indicated first appearance of the various groups. Paleotemperatures inferred from δ¹⁸O (5) and δ³⁰Si (7) in marine cherts are included for comparison. Blue boxes show the inferred NDK-based temperature ranges (Fig. 3) and fossil-based age uncertainties, the red diamonds denoting temperature and age midpoint values for which ViridiNDK and Viridi18S have been combined due to the similarity of their T_ms.