Expanding the genetic code with diverse backbone structures across diverse sequence contexts

This study reports the discovery of orthogonal aminoacyl-tRNA synthetases capable of incorporating eleven diverse non-canonical monomers into proteins and macrocycles, and demonstrates that evolving orthogonal tRNAs can overcome severe sequence-context dependencies to achieve high-efficiency incorporation across nearly all genetic contexts.

The Gist

Imagine the genetic code as the ultimate instruction manual for building life. For billions of years, this manual has only had 20 standard letters (the 20 natural amino acids) to write the stories of proteins. These proteins are the machines, muscles, and messengers that keep our cells running.

Scientists have long wanted to expand this alphabet. They want to insert "non-canonical monomers" (ncMs)—special, custom-made building blocks that don't exist in nature. Think of these as adding new, exotic letters to the alphabet that give the resulting proteins superpowers, like being able to resist digestion, glow in the dark, or snap together into complex shapes.

However, there's a huge problem: The factory is picky.

The Problem: The Factory Rejects New Parts

In this paper, the researchers tried to sneak these new, exotic building blocks into the cell's protein factory. But the factory (the ribosome) is very strict. It only accepts the standard parts.

Even worse, the factory's acceptance of a new part depends entirely on who is standing next to it.

• The Analogy: Imagine you are trying to insert a new, weirdly shaped Lego brick into a wall of standard Legos. If you try to put it next to a red brick, the wall might hold. But if you try to put it next to a blue brick, the whole structure collapses.
• The Reality: The researchers found that for many of these new building blocks, the cell's machinery would only accept them in less than 1% of the possible positions in a protein. If the "neighbors" (the amino acids before and after the new one) weren't perfect, the factory would reject the new part, or worse, stop building the protein entirely.

The Solution: Evolving the "Delivery Truck"

To fix this, the scientists didn't just try to force the new parts in; they decided to upgrade the delivery system.

In the cell, a molecule called tRNA acts like a delivery truck. It picks up a building block and drives it to the factory floor (the ribosome) to be attached to the growing protein. The researchers realized that the standard delivery trucks were too rigid to handle these weird, new shapes, especially when the "neighbors" were difficult.

Here is what they did:

1. Found the Right Drivers: First, they found special enzymes (synthetases) that could load these exotic blocks onto the delivery trucks. This was the first step.
2. Evolved the Trucks: Then, they ran a "survival of the fittest" experiment. They created millions of slightly different versions of the delivery trucks (mutated tRNAs). They asked the cell: "Which truck can successfully deliver this weird block to the factory, no matter who the neighbors are?"
3. The Result: They found "super-trucks" (evolved tRNAs) that were incredibly adaptable.

The Breakthrough

With these new, evolved delivery trucks, the results were dramatic:

• From 1% to 95%: Previously, a new building block might only work in 1 out of 100 protein sequences. With the new trucks, they could insert the same block into 95% of different sequences.
• New Capabilities: They successfully built proteins with completely new backbones (the "spine" of the molecule) that had never been made inside a living cell before.
• Macrocycles: They even used this system to build tiny, circular protein rings (macrocycles) containing these new blocks. These are like molecular handcuffs or rings that could be used as new drugs or materials.

Why This Matters

Think of this as moving from a world where you can only build with standard, square Lego bricks to a world where you can build with any shape you can imagine, and the factory will build it for you, no matter where you put it.

This technology opens the door to:

• New Medicines: Drugs that are stronger, last longer in the body, and can target diseases more precisely.
• New Materials: Self-assembling nanomaterials that could revolutionize electronics or energy storage.
• Understanding Life: Giving us a deeper understanding of how the machinery of life works by testing its limits.

In short, the researchers didn't just find a way to add a new letter to the genetic alphabet; they rewrote the rules of the grammar so that the new letters can be used anywhere in the sentence, unlocking a whole new universe of biological possibilities.

Technical Summary

1. The Problem

The genetic code has been successfully expanded to include non-canonical amino acids (ncAAs) with variant side chains, but incorporating non-canonical monomers (ncMs) with diverse backbone structures (e.g., $\beta$-amino acids, malonic acids, N-cyclic amino acids) remains a significant challenge.

• Limitations of Current Methods: While in vitro incorporation of these monomers has been achieved, in vivo (cellular) incorporation is inefficient.
• Sequence Context Dependence: Previous attempts showed that incorporating ncMs is highly sensitive to the surrounding amino acid sequence (codon context). Many ncMs could only be incorporated at specific sites (<1% of contexts) or not at all, severely limiting the ability to design and synthesize proteins or macrocycles with these novel backbones.
• Missing Components: Although orthogonal aminoacyl-tRNA synthetases (aaRS) capable of charging tRNAs with these ncMs have been discovered, the downstream translational machinery (specifically the tRNA) often fails to efficiently deliver these monomers to the ribosome for co-translational incorporation.

2. Methodology

The authors employed a multi-stage pipeline combining directed evolution, high-throughput screening, and mass spectrometry:

• Discovery of Orthogonal Synthetases:

  • Used tRNA display to select variants of Methanosarcina mazei pyrrolysyl-tRNA synthetase (MmPylRS) capable of acylating their cognate orthogonal tRNA (tRNAPyl) with eleven new ncMs across five chemical classes: $\beta^2$-amino acids, N-cyclic amino acids, malonic acids, carboxylic acids, and $\alpha,\alpha$-disubstituted amino acids.
  • Validated acylation specificity using fluoro-tREX (ftREX) and LC-MS pulldown assays to ensure minimal cross-reactivity with canonical amino acids.

• Mapping Sequence Context Dependence:

  • Systematically tested the incorporation of four representative ncMs at nine different positions in sfGFP.
  • Generated a 4,096-member library of codon context reporters (varying the two codons flanking the amber TAG codon).
  • Used FACS (Fluorescence-Activated Cell Sorting) coupled with Next-Generation Sequencing (NGS) to map incorporation efficiency scores across all sequence contexts, revealing that wildtype tRNAs failed to incorporate most ncMs in the vast majority of contexts.

• Directed Evolution of Orthogonal tRNAs:

  • Constructed saturation libraries of the orthogonal tRNAPyl, targeting regions interacting with EF-Tu, the ribosome, and the anticodon.
  • Performed FACS-based selection in E. coli using sfGFP reporters containing amber codons and the specific ncM.
  • Iteratively evolved tRNA variants (e.g., $\beta^3\_tRNAPyl$, $\beta^2\_tRNAPyl$, $CA\_tRNAPyl$, $aa\_tRNAPyl$) to improve incorporation efficiency and fidelity.
  • Conducted secondary evolution on specific "recalcitrant" sequence contexts to generate second-generation variants (e.g., $\beta^3\_v2\_tRNAPyl$).

• Validation and Application:

  • Quantified incorporation using tREX (gel-based quantification) and ESI-MS/MS of purified proteins.
  • Demonstrated the synthesis of genetically encoded macrocycles containing these diverse ncMs to prove the utility of the system for library generation.

3. Key Contributions

• Expansion of Monomer Scope: Successfully identified orthogonal synthetases that charge tRNAs with 11 new ncMs, including the first in vivo incorporation of $\beta^2$-amino acids and non-canonical N-cyclic amino acids into proteins.
• tRNA Engineering Pipeline: Established a robust pipeline for evolving orthogonal tRNAs to overcome the kinetic and structural barriers of incorporating non-canonical backbones.
• Context Independence: Demonstrated that evolved tRNAs can expand the range of permissive sequence contexts from <1% (wildtype) to >95% for specific ncMs.
• Macrocyclic Synthesis: Enabled the cellular synthesis of macrocyclic peptides containing diverse backbone modifications, which were previously inaccessible.

4. Key Results

• Synthetase Discovery: Identified PylRS variants (e.g., PylRS($\beta^2$-1), PylRS(CA-1)) that specifically acylate tRNAPyl with $\beta^2$-amino acids, N-cyclic amino acids, and malonic acids with high fidelity.
• Sequence Context Sensitivity: Wildtype tRNAs showed strong context dependence. For example, $\beta^3$-amino acid incorporation was efficient in only 0.6% of sequence contexts, while $\alpha,\alpha$-disubstituted amino acids showed dominant-negative effects in many contexts.
• Evolution Success:
  • $\beta^3\_tRNAPyl$: Increased incorporation of $\beta^3$-amino acids from 0.6% to 55.6% of sequence contexts.
  • $\beta^2\_tRNAPyl$: Increased incorporation of $\beta^2$-amino acids from 26.3% to 54.6% of contexts.
  • $CA\_tRNAPyl$: Increased N-cyclic amino acid incorporation from 52.0% to 78.8% of contexts.
  • $aa\_tRNAPyl$: Increased $\alpha,\alpha$-disubstituted amino acid incorporation from 14.5% to 48.7% of contexts.
  • $\beta^3\_v2\_tRNAPyl$ (Second Gen): Further optimized for difficult contexts, achieving incorporation in 95.5% of all tested sequence contexts (a >100-fold improvement over wildtype for some contexts).
• Efficiency Gains: Evolved tRNAs provided up to 40-fold increases in incorporation efficiency for challenging sequence contexts.
• Macrocycles: Successfully purified and characterized macrocyclic peptides containing $\beta^2$, $\beta^3$, N-cyclic, and $\alpha,\alpha$-disubstituted amino acids. The evolved tRNAs allowed the purification of macrocycle variants that were undetectable with wildtype tRNAs.

5. Significance

This work fundamentally shifts the paradigm of genetic code expansion from merely finding synthetases to co-evolving the entire translational machinery (specifically the tRNA) to accommodate non-canonical backbones.

• Generalizability: The approach proves that direct selection of translational components (tRNAs) can unlock the incorporation of monomers that were previously considered "untranslatable" in living cells.
• Chemical Space: It dramatically expands the accessible chemical space for protein engineering, allowing for the creation of proteins and macrocycles with novel structural motifs, protease resistance, and catalytic activities derived from diverse backbone chemistries.
• Foundation for Discovery: The ability to synthesize these molecules genetically in cells enables high-throughput discovery and manufacturing of novel therapeutics and materials that were previously limited to in vitro synthesis or chemical synthesis.