Lipid discovery enabled by sequence statistics and machine learning.

Area of Science:

• Microbiology
• Biochemistry
• Molecular Biology

Background:

• Bacterial membranes are crucial for cell survival and are shaped by evolutionary pressures.
• The MprF enzyme covalently modifies membrane lipids, influencing membrane composition and properties.
• Previous work identified Streptococcus agalactiae MprF synthesizing lysyl-phosphatidylglycerol (Lys-PG) and a novel cationic lipid, lysyl-glucosyl-diacylglycerol (Lys-Glc-DAG).

Purpose of the Study:

• To investigate the occurrence of Lys-Glc-DAG in other MprF-containing bacteria.
• To identify novel MprF products and understand MprF substrate specificity.
• To model evolutionary constraints on MprF sequences using machine learning.

Main Methods:

• Comparative sequence analysis of MprF proteins across bacterial species.
• Machine learning, specifically restricted Boltzmann machines, to predict MprF substrate specificity.
• Investigating MprF activity in various bacterial genera, including Streptococcus and Enterococcus.

Main Results:

• Several streptococcal MprF enzymes were confirmed to synthesize Lys-Glc-DAG.
• A novel MprF substrate, diglucosyl-diacylglycerol (Glc2-DAG), was identified in Enterococcus species.
• An expanded range of bacteria utilizing MprF for glycolipid modification was discovered.

Conclusions:

• MprF exhibits broader substrate specificity than previously known, including glycolipids.
• Machine learning approaches are effective in predicting enzyme function and discovering novel substrates.
• This study reveals new cationic lipids and expands the known functions of MprF in bacterial membrane biosynthesis.