High Throughput Variant Libraries and Machine Learning Yield Design Rules for Retron Gene Editors.

Area of Science:

• Biotechnology
• Molecular Biology
• Genetics

Background:

• Bacterial retron reverse transcriptase systems are used for single-stranded DNA production in biotechnology.
• Modifying natural retron non-coding RNA (ncRNA) enables custom DNA synthesis via reverse transcription.
• Improving reverse transcription efficiency is crucial for retron technology but lacks systematic understanding.

Purpose of the Study:

• To systematically identify retron ncRNA regions amenable to modification for efficient DNA production.
• To establish design rules for 'editrons'—retron-based DNA donors for CRISPR-Cas9 genome editing.
• To enhance retron-mediated DNA production and editron efficiency in biotechnological applications.

Main Methods:

• Thousands of modifications to retron-Eco1 ncRNA were tested using pooled variant library experiments.
• DNA production efficiency was measured to identify tolerant and intolerant regions for modification.
• High-throughput libraries in *S. cerevisiae* were used to define editron design rules.

Main Results:

• Specific regions of the retron ncRNA were identified as critical for maintaining or improving DNA production efficiency.
• Design rules for efficient editron construction were established based on high-throughput screening.
• The optimized retron-Eco1 system achieved unprecedented efficiency in human genome editing.

Conclusions:

• Systematic analysis of retron ncRNA modifications significantly enhances DNA production efficiency.
• Defined design rules enable the development of highly efficient retron-based genome editing tools (editrons).
• This work advances retron technology for custom DNA synthesis and precision genome editing applications.