Designing Cell-Type-Specific Promoter Sequences Using Conservative Model-Based Optimization.

Area of Science:

• Biotechnology
• Molecular Biology
• Bioinformatics

Background:

• Gene therapies require cell-type-specific promoters to target therapeutic genetic cargo effectively and minimize side effects.
• Discovering such promoters is challenging, often needing extensive datasets or manual curation, especially for closely related cell types.

Purpose of the Study:

• To develop a data-efficient framework using model-based optimization (MBO) for designing novel cell-type-specific promoters.
• To address the limitations of existing methods, particularly for distinguishing expression in similar cell types.

Main Methods:

• Utilized a comprehensive framework incorporating conservative objective models (COMs) for MBO.
• Focused on data efficiency and practical considerations like sequence diversity, model uncertainty, and experimental validation.
• Applied the framework to three similar blood cancer cell lines: Jurkat, K562, and THP1.

Main Results:

• Successfully discovered numerous novel cell-type-specific promoters across the tested cell lines.
• Identified a promoter for K562 cells with 75.85% greater cell-type-specificity than the best existing promoter.
• Demonstrated the framework's efficacy in designing promoters for closely related cell types.

Conclusions:

• The developed MBO framework enables efficient discovery of cell-type-specific promoters, even for similar cell types.
• This approach advances the potential of gene therapies by improving the specificity of genetic cargo delivery.
• The findings offer a practical solution for designing regulatory sequences crucial for targeted therapeutic applications.