PhaC Enzyme Engineering — LLM Context Document

Version: v1.0
Date: [DATE]
Engineer: [YOUR NAME]
Project goal: [ONE SENTENCE SUMMARY, e.g. “Engineer Class I PhaC to incorporate 3HHx at >15 mol%”]

1. Enzyme Family Background

1.1 Classification

• Class I and II share ~50% sequence identity; Class III/IV are more distantly related
• Class I/II are the primary engineering targets for substrate specificity work

1.2 Reaction chemistry

• Catalyzes polymerization of (R)-3-hydroxyacyl-CoA thioesters into PHA
• Ping-pong (double displacement) mechanism:
  1. Acylation: acyl group transferred to catalytic Cys, CoA released
  2. Transacylation: acyl group transferred to growing polymer chain
• Lipase-like α/β hydrolase fold
• Catalytic triad: Cys – His – Asp
  • C. necator PhaC1 (Cn) reference numbering: C319, D480, H508

1.3 Substrate scope terminology

1.4 Why substrate specificity is structurally interesting

• Substrate-binding tunnel geometry determines acyl chain length tolerance
• Residues within ~5–10 Å of catalytic Cys are primary selectivity determinants
• mcl selectivity often results from removal of steric clash (smaller residues), not addition of new contacts — counterintuitive but well-supported
• Electrostatic environment affects CoA-thioester positioning
• Dimerization interface indirectly influences active site geometry (Class I/II)

2. Structural Information

2.1 Available experimental structures

2.2 AlphaFold models

2.3 Key structural regions

(Using C. necator PhaC1 residue numbering as reference)

2.4 Known substrate-contacting / selectivity residues

(From mutagenesis studies and structural analyses — update as you find more)

2.5 Tunnel geometry notes

• scl enzymes: narrower tunnel, estimated constriction ~4–6 Å
• mcl enzymes: wider tunnel — bulky residues at key positions replaced by smaller ones (Ala, Gly) to accommodate longer acyl chains
• [Add any MD simulation or docking notes here as available]

3. Sequence Dataset Summary

3.1 Dataset composition

• Total sequences collected: [N]
• After 95% identity dereplication (cd-hit): [N]
• Labeled with substrate preference data: [N]
  • scl only: [N]
  • mcl only: [N]
  • broad/mixed: [N]
  • specialty monomer: [N]
• Unlabeled (phylogenetic diversity only): [N]
• Data sources: UniProt/SwissProt, NCBI RefSeq, literature

3.2 Taxonomic distribution

3.3 Alignment properties

• Alignment method: [MUSCLE / Clustal Omega / MAFFT]
• Raw aligned length: [N] columns
• After gap trimming (>80% gap threshold): [N] columns
• Mean pairwise identity, full dataset: [X]%
• Mean pairwise identity, scl group: [X]%
• Mean pairwise identity, mcl group: [X]%

3.4 Top mutual information positions

(Fill in after running Option 2 MI analysis)

4. Experimental Mutation Database

(This section should grow over time as you mine the literature and generate your own data)

4.1 Key literature to mine

• Tsuge et al. (2003) Macromolecules — F420 region, Class I
• Amara et al. (2002) — systematic Class I mutagenesis
• Rehm lab series — Class II specificity determinants
• Nomura et al. — broad-specificity engineered variants
• Insomphun et al. — 3HHx incorporation engineering
• [Add others as you find them]

4.2 Gain-of-function mutations (toward mcl/broader specificity)

4.3 Loss-of-function / specificity-narrowing mutations

4.4 Combinatorial / double mutants

4.5 Thermostability mutations

(Relevant if stacking specificity mutations — need to preserve stability)

4.6 Notes on data quality and comparability

• Monomer incorporation % varies heavily with fermentation conditions (carbon source, growth phase, host strain) — cross-lab comparisons are unreliable
• In vitro assays (purified enzyme + CoA thioesters) are more reliable for intrinsic specificity than in vivo PHA production titers
• [Add any other caveats specific to your dataset]

5. Your Starting Enzyme (Wild-Type)

5.1 Identity

• UniProt accession: [ID]
• Organism: [NAME]
• PhaC class: [I / II / III / IV]
• Gene name: [phaC / phaC1 / phaC2]
• Full sequence length: [N] aa

5.2 Known properties

• Native substrate preference: [e.g. scl — 3HB/3HV, negligible 3HHx]
• Specific activity: [X nmol/min/mg if known]
• Thermostability: [Tm or optimal temperature]
• Expression: [e.g. soluble in E. coli BL21 at 25°C, typical yield X mg/L]
• Any known issues: [e.g. prone to aggregation, requires CoA for stability]

5.3 Sequence — full

[PASTE FULL AMINO ACID SEQUENCE HERE]

5.4 Sequence — substrate-binding pocket region

(~30 residues centered on catalytic Cys; easier to include in prompts)

[PASTE POCKET REGION SEQUENCE HERE — label residue numbers]
e.g. residues 305–335: XXXXCXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
                            ^C319

5.5 Alignment position mapping

(Map your WT residue numbers to the C. necator reference numbering and to your alignment column numbers — critical for interpreting suggestions)

6. Engineering Target

6.1 Primary goal

[State precisely, e.g.:]

Incorporate 3HHx (C6) at >15 mol% in scl-mcl copolymer produced in E. coli BL21 on mixed carbon source (sodium butyrate + sodium hexanoate)

6.2 Secondary goals

• [e.g. Retain 3HB incorporation >50 mol%]
• [e.g. Maintain thermostability — Tm drop <5°C acceptable]

6.3 Acceptable tradeoffs

• [e.g. Up to 30% reduction in overall polymerization activity]
• [e.g. Reduced expression yield acceptable if specificity goal is met]

6.4 Hard constraints — DO NOT VIOLATE

• Maximum simultaneous mutations: [N] (practical screening limit)
• Must retain soluble expression in E. coli
• Do not mutate catalytic triad residues (C319, D480, H508)
• Avoid dimer interface mutations (stability risk)
• [Add any others]

6.5 What has already been tested

(Critical — prevents the LLM from repeatedly suggesting things you’ve tried)

7. Production and Assay Context

7.1 Expression system

• Host: [e.g. E. coli BL21(DE3)]
• Vector: [e.g. pET-28a, His-tag]
• Expression conditions: [e.g. 25°C, 16h, 0.5 mM IPTG]
• Typical yield: [X mg/L culture]

7.2 PHA production conditions

• Carbon source(s): [e.g. 10 mM sodium butyrate + 5 mM sodium hexanoate]
• Co-pathway: [e.g. PhaA/PhaB co-expressed for 3HB-CoA supply; PhaJ for 3HHx-CoA]
• Growth phase at harvest: [e.g. 48h, stationary]
• PHA content typically: [X wt%]

7.3 Analytical method

• PHA extraction: [e.g. chloroform extraction, sodium hypochlorite method]
• Monomer analysis: [e.g. GC-FID after methanolysis, GC-MS for identification]
• Activity assay (if used): [e.g. DTNB assay monitoring CoA release]
• Throughput: [e.g. 24 variants per experiment]

8. Reasoning Guidelines for LLM

8.1 Prioritization criteria (in order)

1. Mechanistic/structural plausibility — a rationale is required
2. Consistency with experimental mutation database (Section 4)
3. Conservation pattern in target-substrate homologs (Section 3.4)
4. Novelty relative to literature

8.2 Required output format for mutation suggestions

For every suggested mutation, provide:

• (a) Mutation in standard notation (e.g. A149F)
• (b) Mechanistic rationale — why this residue, why this substitution
• (c) Supporting evidence — literature, alignment, structural
• (d) Confidence level: High / Medium / Low
• (e) Potential risks — stability, expression, off-target effects
• (f) Tag [SPECULATIVE] if based on analogy with no direct evidence

8.3 Reasoning I do NOT want

• Suggestions based solely on “this residue differs between scl and mcl sequences” without structural or mechanistic reasoning
• Overconfident quantitative predictions (e.g. “this will give 20% 3HHx”)
• Suggestions that violate hard constraints in Section 6.4
• Ignoring the “already tested” table in Section 6.5

8.4 When hypotheses conflict

• Explicitly state the conflict and explain both sides
• Do not silently choose one; flag for experimental resolution

8.5 My background

[Describe your expertise so the LLM calibrates explanation depth, e.g.:]

PhD in microbiology/biochemistry. Comfortable with protein biochemistry and microbial fermentation. Less experienced with structural biology and computational methods — please explain structural reasoning in accessible terms but do not oversimplify the biochemistry.

9. Session Log

(Append after each LLM session — builds institutional memory)

Session [DATE]

Question asked: [Paste your prompt]

Key LLM output / hypotheses generated: [Summarize or paste]

Your assessment: [Which suggestions seem worth pursuing, which to discard and why]

Action items:

• [e.g. Test A149F single mutant]
• [e.g. Check position 171 in AlphaFold model]

Session [DATE]

(repeat block)

10. Experimental Results Log

(Append as data comes in — feeds back into Section 4 and future sessions)

Experiment [DATE / ID]

Variants tested:

Interpretation: [What do these results mean for your hypotheses?]

Updated hypotheses: [How do results change your model of specificity determinants?]

End of context document — keep this file updated and prepend it to every new LLM session