For this project, I focused on designing L-protein mutants to optimize lysis function in MS2 phage. The L-protein is a 70-residue lysis protein with two distinct domains: a soluble N-terminal domain (residues 1-35) that interacts with the bacterial chaperone DnaJ, and a transmembrane C-terminal domain (residues 36-70) responsible for membrane disruption and lysis activity.
To generate promising mutants, I used ESM-IF1 to score potential mutations across the L-protein sequence, focusing on positions that showed high likelihood ratios for beneficial changes. I also incorporated experimental data from published mutagenesis studies, particularly targeting the critical LS motif region (residues 44-56) which previous research identified as essential for lysis function.
My evaluation process involved several computational analyses. I used the provided experimental dataset to identify positions where mutations maintained protein expression but lost lysis activity, reasoning that these sites are functionally important but structurally stable. I then applied AlphaFold-Multimer to model L-protein:DnaJ interactions for soluble domain mutants, examining how mutations might strengthen or weaken this critical binding interface. For transmembrane domain mutations, I used AlphaFold2 to assess structural stability and considered the oligomerization potential of L-protein, as recent studies suggest it forms membrane-spanning complexes.
After applying these criteria, I selected the following five mutants for further consideration:
| Mutation | Domain | Rationale | Evaluation |
|---|---|---|---|
| R9K | Soluble | Conservative change maintaining positive charge for DnaJ binding | AF2-Multimer shows preserved interface; experimental data confirms maintained expression |
| Y39L | Soluble | ESM high score; removes polar residue to potentially strengthen membrane association | High LLR score (2.24); may reduce DnaJ affinity based on AF2 analysis |
| L44Y | Transmembrane | Introduces aromatic residue near LS motif to enhance oligomerization | Located in critical lysis region; π-stacking potential for complex formation |
| K50L | Transmembrane | Highest ESM score; replaces charged residue with hydrophobic for better membrane insertion | LLR score 2.56; improved pLDDT scores in AF2 predictions |
| T52L | Transmembrane | Removes polar residue from transmembrane helix to improve stability | Located in functionally important region; conservative hydrophobic change |
While these computational predictions provide valuable insights, I recognize several limitations in my approach. AlphaFold-Multimer predictions for membrane proteins have inherent uncertainties, particularly for transmembrane domains where lipid environment effects are not modeled. The experimental dataset, while informative, represents single mutations, and the combinatorial effects of multiple mutations remain unclear. Additionally, ESM scores reflect evolutionary likelihood but may not capture the specific functional requirements of the MS2 lysis mechanism.
As next steps, I would prioritize experimental validation through complementation assays measuring lysis efficiency in E. coli. For DnaJ interaction mutants, I would employ co-immunoprecipitation or surface plasmon resonance to quantify binding affinity changes. Membrane insertion and oligomerization properties could be assessed using fluorescence microscopy and cross-linking experiments. If initial results prove promising, I would explore combinatorial mutations to potentially achieve synergistic improvements in both DnaJ binding and membrane disruption efficiency.