🧬 biology

Codon bias coevolves with longevity

This study demonstrates that long-lived mammals have evolved specific codon usage biases—such as depleting hypermutable stopogenic codons and favoring non-optimal codons to slow translation and improve protein folding—to minimize somatic mutations and proteostasis loss, thereby extending lifespan.

Original authors: Krisztina Kerekes, Mária Trexler, László Bányai, László Patthy

Published 2026-07-01

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Krisztina Kerekes, Mária Trexler, László Bányai, László Patthy

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

The Big Picture: Why Do Some Animals Live Longer?

Imagine your body is a massive factory that builds millions of tiny machines (proteins) every day. These machines keep you alive. Over time, these machines can get damaged, built incorrectly, or start to rust. When too many of them break, the factory slows down, and the animal ages.

Scientists have long wondered: Why do some animals (like whales or humans) live much longer than others (like mice)?

This paper suggests that the secret isn't just in the instructions for building the machines, but in the rhythm and pacing of how those instructions are read. The researchers found that long-lived animals have evolved a specific way of "speaking" their genetic language that helps build better, longer-lasting machines.

Analogy 1: The "Trap Door" in the Instructions

The Problem:
Your DNA is written in a language of three-letter words called "codons." Some of these words are dangerous because they are like "trap doors" that can easily turn into a "Stop" sign by accident. If a "Stop" sign appears too early, the machine stops building halfway, creating a broken, useless part.

One specific word, CGA, is a notorious trap door. It sits on a slippery patch (a CpG site) where it very easily slips and turns into a "Stop" sign.

The Discovery:
The researchers found that in long-lived animals, this dangerous CGA word is almost completely missing from their instructions. Short-lived animals still use it a lot.

The Takeaway: Long-lived animals have evolved to avoid using the "trap door" words. By removing these risky instructions, they prevent their machines from breaking prematurely, keeping the factory running smoothly for decades instead of years.

Analogy 2: The "Speed Bump" Strategy (The Surprising Part)

The Old Idea:
For a long time, scientists thought that to build things perfectly, you should use the "fastest" and "easiest" words in the genetic language. These are called optimal codons. Think of them as the "green lights" on a highway that let the construction crew (the ribosome) zoom through without stopping. The logic was: Faster and more accurate = Better.

The New Discovery:
This paper found something surprising. While long-lived animals do use some "green lights" to ensure accuracy, they actually use more "red lights" and "speed bumps" for many other parts of the machine.

These "speed bumps" are non-optimal codons. They are harder for the construction crew to read, so the crew has to slow down.

Why slow down?
Imagine you are folding a very complex origami crane. If you fold it too fast, you might mess up a crease, and the whole thing collapses. But if you pause at specific difficult points to make sure the paper is aligned perfectly before moving on, the final result is perfect.

The Paper's Claim: Long-lived animals use these "slow-down" words at specific spots (like the joints between different parts of a protein). This forces the construction crew to pause, giving the new protein enough time to fold into its perfect 3D shape while it is being built.
The Result: Fewer misfolded, tangled, or broken proteins. Since aging is largely caused by a buildup of these "tangled" proteins, slowing down the process actually extends the life of the animal.

Analogy 3: The "Stop Sign" Quality Control

The paper also looked at the final "Stop" signs in the genetic instructions. There are three different words that can mean "Stop": TAA, TAG, and TGA.

TAA is a very clear, loud stop sign.
TAG and TGA are a bit fuzzy; sometimes the construction crew misses them and keeps building, creating a machine with a weird, extra tail attached to the end.

The Discovery:
Long-lived animals almost exclusively use the clear TAA stop sign. Short-lived animals use the fuzzy ones more often.

The Takeaway: By using the clearest possible "Stop" sign, long-lived animals ensure their machines are cut off exactly where they should be, preventing the creation of weird, broken versions of proteins that could clog up the factory.

Summary: The "Proteostasis" Balance

The paper concludes that aging is essentially a loss of proteostasis—the ability of the cell to keep its protein machines clean, folded, and working.

Long-lived animals have evolved a genetic strategy that balances two things:

Accuracy: Avoiding dangerous "trap door" words (like CGA) and fuzzy "Stop" signs.
Folding: Intentionally using "slow-down" words to give proteins time to fold correctly, rather than just rushing to build them as fast as possible.

In short, long-lived animals don't just build their bodies faster; they build them more carefully, taking the time to ensure every part is folded perfectly, which keeps the whole system running for much longer.

Technical Summary: Codon Bias Coevolves with Longevity

Problem Statement
Somatic mutations are a primary driver of carcinogenesis and aging, with gene vulnerability heavily dependent on the abundance of mutation hotspots, particularly CpG dinucleotides. While it is established that long-lived species exhibit increased fidelity of protein synthesis, the specific evolutionary mechanisms linking codon usage to longevity remain unclear. A prevailing hypothesis suggests that longevity evolves through the selection of translationally optimal codons to maximize accuracy and minimize mistranslation. However, this study investigates whether negative selection against hypermutable codons and specific shifts in codon bias—potentially favoring non-optimal codons to regulate translation speed—contribute to the evolution of mammalian longevity.

Methodology
The authors analyzed a dataset of 96 placental mammalian species, integrating lifespan data from the AnAge Database with codon usage statistics (Relative Synonymous Codon Usage, RSCU) and genomic GC content from the Codon Statistics Database.

Normalization: To isolate the effects of selection from the background influence of genomic GC content, the authors calculated the deviation between observed RSCU values ( $RSCU_{obs}$ ) and expected RSCU values ( $RSCU_{exp}$ ) based on the linear relationship between codon usage and GC content.
Correlation Analysis: These deviations were plotted against species lifespan to identify lifespan-dependent codon usage patterns.
Comparative Scope: The analysis was conducted first on a restricted set of 19 rodent species to minimize taxon-specific biological noise, and subsequently expanded to the full spectrum of 96 mammals to test for generalizability.
tRNA Context: The study incorporated tRNA gene copy number data (as a proxy for tRNA abundance) to classify codons as "translationally optimal" (matching abundant isodecoder tRNAs) or "nonoptimal" (lacking matching isodecoder tRNAs).

Key Results

Negative Selection of Hypermutable Codons:
- The study confirmed that the RSCU values of synonymous codons are primarily dictated by genomic GC content.
- However, codons containing hypermutable CpG dinucleotides showed significant depletion relative to their GC-content equivalents.
- Crucially, the CGA codon (arginine), which contains a CpG site and can mutate into a stop codon (TGA) at a high rate, was significantly more depleted in long-lived mammals compared to short-lived ones. This depletion correlates with increased lifespan across mammals, suggesting negative selection against this "stopogenic" hypermutable codon to reduce gene vulnerability.
- Similarly, the use of the TAA stop codon (least prone to readthrough) increased with lifespan, while TGA and TAG decreased, indicating a selection against aberrant protein C-terminal extensions.
Divergent Codon Usage Strategies in Long-Lived Species:
The analysis revealed two distinct patterns of codon usage changes associated with longevity:
- Optimization for Accuracy: For a subset of amino acids (e.g., Ile, Leu, Val, Gly, Thr), long-lived species favored translationally optimal codons (those matching abundant tRNAs). This aligns with the hypothesis that increased translation accuracy reduces the formation of abnormal proteins.
- Optimization for Folding (The Surprising Finding): For a larger group of amino acids (including Asn, Asp, Cys, Phe, Tyr, Ala, Arg, Pro, Ser), long-lived species favored translationally nonoptimal codons (those lacking matching isodecoder tRNAs).
  - In rodents, this trend was strong for Asn, Asp, Cys, Phe, Tyr, Ala, Arg, Pro, and Ser.
  - In the broader mammalian dataset, this correlation remained significant for Asn, Asp, Phe, Tyr, Ala, and Pro.
  - The authors interpret this not as a defect, but as a regulatory mechanism: the use of nonoptimal codons slows down translation at specific sites, providing the nascent polypeptide chain more time to undergo correct co-translational folding. This is particularly critical for multidomain proteins, where linker regions enriched in nonoptimal codons may facilitate a "beads-on-a-string" folding mechanism.
Saturation of Longevity Mechanisms:
The relationship between codon usage shifts and lifespan was found to be non-linear. The slope of codon usage changes was steepest in short-lived mammals (1–40 years) and flattened significantly in extremely long-lived species (>60 years, e.g., humans, whales). This suggests that while codon bias optimization contributes to extended lifespan, extreme longevity may be limited by factors beyond protein misfolding once critical sites for folding and translation efficiency are saturated.

Significance
The paper challenges the simplistic view that longevity is driven solely by the maximization of translation speed and accuracy via optimal codons. Instead, it proposes a more nuanced model where the evolution of longevity involves a balanced proteostasis strategy:

Minimizing Mutational Risk: Through the negative selection of hypermutable, stopogenic codons (specifically CGA).
Enhancing Folding Fidelity: By utilizing translationally nonoptimal codons to decelerate translation, thereby facilitating the correct co-translational folding of complex multidomain proteins.

The authors argue that this reduction in protein misfolding and aggregation addresses a central hallmark of aging—the loss of proteostasis. The findings suggest that the "slowing down" of translation at specific nonoptimal codons is a beneficial evolutionary adaptation that reduces the rate of cellular aging, complementing the benefits of high-fidelity translation in other regions of the proteome.