The Nature of LUCA and Its Impact on the Early Earth System

What kind of paper is this?

This is a Discovery ($\Psi_{\text{Discovery}}$) paper. While it introduces a refined implementation of molecular clock calibration (“cross-bracing”), the primary contribution is the biological inference regarding LUCA’s age, genome size, and metabolic nature. The computational methods serve to characterize a specific biological entity.

What is the motivation?

Understanding the Last Universal Common Ancestor (LUCA) is critical for reconstructing the early evolution of life, yet consensus has been elusive due to disparate data and methods.

• Age Conflicts: Estimates vary widely depending on fossil interpretation and molecular clock calibrations, particularly regarding the “Late Heavy Bombardment” (LHB) constraints.
• Physiological Uncertainty: Debates persist over whether LUCA was a simple “progenote” dependent on geochemistry or a complex prokaryote-grade organism.
• Environmental Context: LUCA is often modeled in isolation, ignoring the ecological interactions that would have shaped its survival and impact on the early Earth system.

What is the novelty here?

The study integrates three advanced computational approaches to provide a holistic reconstruction of LUCA:

• Cross-Braced Dating: It employs a “cross-bracing” strategy in Bayesian molecular clocks, using pre-LUCA gene duplications (paralogues) to constrain the root. This allows the same fossil calibrations to be applied to mirrored nodes, significantly reducing uncertainty.
• Probabilistic Reconciliation: It uses the ALE (Amalgamated Likelihood Estimation) algorithm to reconcile 9,365 gene family trees against the species tree. This explicitly models gene transfer, duplication, and loss, allowing for a much broader reconstruction of the proteome.
• Ecosystem Modeling: The physiological reconstruction is coupled with geochemical modeling to propose that LUCA was a member of a productive, hydrogen-recycling early ecosystem.

What experiments were performed?

• Phylogenomics: Inferred a species tree from 57 single-copy marker genes across 700 diverse prokaryotic genomes (350 Archaea, 350 Bacteria) using maximum likelihood (IQ-TREE 2).
• Molecular Dating: Estimated divergence times using MCMCtree with a partitioned dataset of 5 pre-LUCA paralogue pairs (e.g., ATP synthase, EF-Tu/G). Calibrations included 13 fossil constraints and a “soft” maximum bound based on the Moon-forming impact (4.51 Ga).
• Metabolic Reconstruction: Reconciled 9,365 KEGG ortholog families against the species tree to calculate the posterior probability (PP) of each gene’s presence in LUCA. Metabolic potential was inferred from genes with high PP (typically >0.75).
• Genome Size Prediction: Trained a LOESS regression model on modern prokaryotes to predict LUCA’s genome size based on the inferred number of KEGG families.

What outcomes/conclusions?

• Age: LUCA lived approximately 4.2 Ga (95% CI: 4.09-4.33 Ga), surprisingly soon after the Moon-forming impact (~4.5 Ga).
• Complexity: LUCA was a complex, prokaryote-grade organism with a genome size of ~2.75 Mb (encoding ~2,600 proteins), comparable to modern prokaryotes.
• Physiology:
  • Metabolism: Anaerobic acetogen using a complete Wood-Ljungdahl pathway (WLP) for $CO_2$ fixation and an almost complete TCA cycle. Likely thermophilic (reverse gyrase present, PP = 0.97). The paper found no strong evidence for nitrogenase or nitrogen fixation.
  • Immunity: Possessed 19 Class 1 (Type I and Type III) CRISPR-Cas effector protein families. Cas1 and Cas2 were absent, suggesting an early immune system capable of RNA cleavage and binding but lacking the full CRISPR adaptation machinery.
• Ecology: LUCA likely inhabited one of two major habitats: (1) the deep ocean, where hydrothermal vents and serpentinization provided $H_2$ (supported by reverse gyrase presence, PP = 0.97, consistent with hyperthermophily), or (2) the ocean surface, where atmospheric $H_2$ from volcanism and metamorphism could fuel growth. A shallow hydrothermal vent or hot spring is also considered a possibility. LUCA was part of an established ecosystem whose metabolic by-products would have created niches for other metabolisms, including methanogenesis. If methanogens were also present, the $CH_4$ they produced would have been photochemically recycled to $H_2$ in the atmosphere, boosting biosphere productivity by at least an order of magnitude over abiotic $H_2$ input rates.
• Limitation: The placement of two small-genome lineages (CPR, Candidate Phyla Radiation, and DPANN) remained uncertain. The AU (approximately unbiased) test could not reject either topology (p = 0.517), meaning the data cannot discriminate between the two placements. This phylogenetic uncertainty affects inferences about the early bacterial and archaeal stem lineages.

Reproducibility Details

Data

The study relied on publicly available genomic data and specific subsets of marker genes.

Algorithms

Key computational steps involved sequence processing, tree inference, and probabilistic reconciliation.

• Alignment & Trimming: sequences aligned with MAFFT L-INS-i (v7.407) and trimmed with BMGE (v1.12, BLOSUM30 matrix, entropy 0.5).
• Tree Inference: IQ-TREE 2 (v2.1.2). Species tree: LG+C60+F+G (best-fit by BIC from concatenated 57-marker alignment). Gene family trees for ALE reconciliation (9,365 KEGG families): LG+F+G with 1,000 ultrafast bootstraps.
• Reconciliation: ALE (Amalgamated Likelihood Estimation) program ALEml_undated used to calculate gene presence probabilities, accounting for HGT, duplication, and loss.
• Genome Prediction: LOESS regression (Locally Estimated Scatterplot Smoothing) used to map KEGG family counts to total protein counts/genome size.

Models

The analysis employed sophisticated evolutionary models to handle deep time scales and heterogeneity.

• Substitution Models:
  • Species Tree: LG+C60+F+G (mixture model with 60 profiles, best-fit by BIC).
  • Gene Family Trees (for ALE reconciliation): LG+F+G with 1,000 ultrafast bootstraps.
  • Timetree inference: LG+F+G4 for approximate likelihood calculation (CODEML), as CODEML does not implement the CAT mixture model.
• Molecular Clock:
  • MCMCtree (PAML v4.10.7).
  • Relaxed clock models: GBM (Geometric Brownian Motion) and ILN (Independent Lognormal).
  • Cross-Bracing: Specifically models shared divergence times for duplicated nodes (driver and mirror nodes).

Evaluation

Validation focused on robustness across different topologies and clock models.

Hardware

• Software: PAML v4.10.7 (MCMCtree), IQ-TREE 2, ALE v0.4, HMMER v3.3.2.
• Compute: IQ-TREE runs specified usage of 4 CPUs; MCMCtree approximated likelihood calculation (approx method) to reduce computational cost.

Paper Information

Citation: Moody, E. R. R., Álvarez-Carretero, S., Mahendrarajah, T. A., et al. (2024). The nature of the last universal common ancestor and its impact on the early Earth system. Nature Ecology & Evolution, 8, 1654-1666. https://doi.org/10.1038/s41559-024-02461-1

Publication: Nature Ecology & Evolution 2024

@article{moodyTheNatureLast2024,
  title={The nature of the last universal common ancestor and its impact on the early Earth system},
  author={Moody, Edmund R. R. and Álvarez-Carretero, Sandra and Mahendrarajah, Tara A. and Clark, James W. and Betts, Holly C. and Dombrowski, Nina and Szánthó, Lénárd L. and Boyle, Richard A. and Daines, Stuart and Chen, Xi and Lane, Nick and Yang, Ziheng and Shields, Graham A. and Szöllősi, Gergely J. and Spang, Anja and Pisani, Davide and Williams, Tom A. and Lenton, Timothy M. and Donoghue, Philip C. J.},
  journal={Nature Ecology & Evolution},
  volume={8},
  number={9},
  pages={1654--1666},
  year={2024},
  publisher={Nature Publishing Group},
  doi={10.1038/s41559-024-02461-1}
}

Open Access: This article is published under CC BY 4.0 and is freely available at the paper URL above.

Artifacts: