Rewriting Endogenous tRNA Genes: PERT Emerges as Universal Therapy Platform Targeting Thousands of Genetic Diseases

From CRISPR-Cas9 to prime editing, gene therapy has become increasingly precise. However, this precision has also introduced a paradox of "excessive specificity": to treat over 200,000 different pathogenic mutations, we would need to develop thousands of distinct drugs, a nearly insurmountable cost. Is there a universal strategy that could treat thousands of different genetic diseases with just one "drug component"?

A recent study published in Nature, titled "Prime editing-installed suppressor tRNAs for disease-agnostic genome editing", reported a breakthrough. Researchers used prime editing to convert redundant endogenous tRNA genes into "suppressor tRNAs", proposing a universal therapeutic strategy called PERT. This represents not only a technological innovation but also a deep biological exploration of balancing "efficacy" and "safety".

Among pathogenic alleles recorded in the ClinVar database, approximately 24% are caused by nonsense mutations. These mutations introduce premature termination codons (PTCs) into gene sequences, causing ribosomes to "emergency brake" before protein synthesis is complete. This results in truncated, often nonfunctional proteins and may trigger nonsense-mediated mRNA decay (NMD).

In theory, suppressor tRNAs (sup-tRNAs) are the perfect counter to such mutations. Sup-tRNAs possess anticodons complementary to termination codons, enabling ribosomes to read through PTCs. However, the core challenge hindering the translation of this therapy is the "difficulty in balancing efficacy and toxicity": exogenous overexpression of sup-tRNAs often leads to severe toxicity, potentially causing genome-wide misreading of natural termination codons (NTCs).

The brilliance of the PERT (Prime Editing-mediated Read-Through of PTCs) strategy proposed by David Liu's team lies in its "endogenous replacement" approach rather than "exogenous addition". Leveraging the redundancy of 418 high-confidence tRNA genes in the human genome, the team sacrificed one copy to convert it into a sup-tRNA. This ensures that its expression is regulated by endogenous control elements, maintaining physiological levels and significantly reducing toxicity risks.

Screening the Best Chassis from 418 Candidates

To identify the most suitable candidate for modification, the researchers designed a high-throughput screening system. Using a lentiviral library, they designed corresponding prime editing guide RNAs (epegRNAs) for all 418 high-confidence human tRNA genes, attempting to mutate their anticodons into sequences recognizing TAG, TGA, or TAA termination codons. The study revealed that tRNA backbones for arginine (Arg), leucine (Leu), tyrosine (Tyr), and serine (Ser) showed potential. However, the efficiency was too low.

Although an allele conversion rate of ~29% was achieved at endogenous sites, tRNA with only an altered anticodon (ac-only sup-tRNA) did not induce significant PTC read-through in a single-copy reporter gene model. Even with reporter gene overexpression, protein production recovery was only ~10% compared to wild-type controls. This revealed a harsh reality: at physiological expression levels, merely altering the anticodon is insufficient-deep engineering of the tRNA is necessary.

Engineering a "Super" Suppressor tRNA

To address the lack of potency, the researchers comprehensively optimized the tRNA. First was the interplay between the leader sequence and the termination sequence. They tested 11,543 combinations of "leader sequence + tRNA variant + termination sequence". The results showed that a synthetic termination sequence composed of five thymines (5T) directly adjacent to the mature tRNA sequence outperformed most endogenous termination sequences, likely because efficient transcriptional termination promoted RNA polymerase III recycling.

Through massive data analysis, they identified a series of surprising "potency-enhancing mutations", particularly in the anticodon stem (e.g., A>T at position 38). Intriguingly, these mutations were additive. When the most beneficial mutations were combined in tRNA-Leu-TAA-1-1, a miracle occurred: compared to the ac-only version, the engineered sup-tRNA increased full-length GFP protein production by 5-fold, reaching 35% of wild-type control levels. This far exceeds the therapeutic threshold required for many recessive genetic diseases-all achieved with a single genomic copy under endogenous promoter control.

Pushing Prime Editing to Its Limits

With the perfect sequence in hand, the next step was precise integration. The researchers designed a library of 17,280 engineered pegRNAs (epegRNAs), systematically optimizing the length of the primer binding site (PBS) and reverse transcription template (RTT).

To counteract interference from DNA mismatch repair (MMR) pathways, they co-transfected dominant-negative MMR proteins (MLH1dn) or introduced silent mutations to evade surveillance. After extensive optimization, they achieved editing efficiencies of 60%-80% at endogenous tRNA sites in HEK293T cells. This meant that the redundant leucine tRNA in the vast majority of cells in the dish had been successfully replaced with a custom "super" sup-tRNA.

Figure 1. Optimization of a prime editing strategy for engineered sup-tRNA installation. (Pierce S E, et al., 2025)

From Cell Models to Hurler Syndrome Mice

To validate broad applicability, the researchers constructed a massive library encompassing 14,746 pathogenic TAG mutations. The results showed that a single ac-only sup-tRNA could enable effective read-through in over 70% of sequence contexts. In cell models of various genetic diseases-including Batten disease and Tay-Sachs disease-the PERT strategy successfully restored 17%-70% of enzymatic activity.

The ultimate test came in a Hurler syndrome (mucopolysaccharidosis type I) mouse model, a severe lysosomal storage disorder. By intracerebroventricular injection of AAV9 viral vectors in newborn mice, the team observed 6%-8% editing efficiency in the heart, cerebral cortex, and liver after seven weeks.

For enzyme deficiency diseases, this was enough to reverse the disease. Biochemical analysis revealed that IDUA enzyme activity in treated mice recovered to 7.6%, 6.3%, and 1.7% of wild-type levels in the heart, cerebral cortex, and liver, respectively. Histopathological scoring confirmed that vacuoles in Purkinje cells, liver, and spleen were almost entirely eliminated, with organ states indistinguishable from healthy mice.

Figure 2. Prime editing generates functional sup-tRNAs to rescue animal models of disease. (Pierce S E, et al., 2025)

The Next Era of Gene Editing: "Universalization"

This study marks the transition of gene editing therapies from "customization" to "universalization." The core appeal of the PERT strategy lies in its potential for "one composition of matter to treat diverse genetic diseases." This means resources can be focused on optimizing a single reagent, which, once successful, could benefit thousands of rare diseases, breaking the cost curse of orphan drug development.

Of course, PERT is not without limitations. Current optimizations primarily target TAG and TGA stop codons, with the more stubborn TAA remaining a challenge. Additionally, delivery efficiency and amino acid substitution compatibility require further exploration. Yet, through meticulous evolutionary screening and prime editing technology, David Liu's team has transformed an unassuming cellular tRNA into a "master key" for saving lives. For patients suffering from rare genetic mutations, this research may be the long-awaited blueprint of hope.

Reference

Pierce S E, et al. Prime editing-installed suppressor tRNAs for disease-agnostic genome editing. Nature, 2025: 1-12.