Complestatin-Family BGC LBP-Chassis Feasibility (comp-024)¶
Question. Is the complestatin NRPS BGC tractable in an LBP chassis (E. coli Nissle 1917, Bacteroides thetaiotaomicron) as the next CP0 engineering payload after H05 DAF SCR1-4?
Verdict. RED for the LBP-track framing. Best host (EcN) is YELLOW 0.544; Bacteroides RED 0.225. No host clears GREEN (≥0.60). Complestatin BGC is not the right next CP0 LBP payload. The C1-INH (LBP-luminal) parallel thread scores GREEN-provisional 0.774 and is recommended as the next computational gate instead. Complestatin BGC stays in scope as an aerobic-fermentation production candidate (Streptomyces-class manufacturing), not an LBP-track payload.
BGC summary (grep-verified)¶
48.7 kb (Chiu 2001) / 54.5 kb reconstituted (Park 2016); 16 ORFs; 7 NRPS modules (1 loading + 6 extension + TE) on ComA-D; tailoring = ComI/ComJ (P450 phenolic coupling), ComH (nonheme halogenase), Hmo (FMN oxidase), inferred OxyD β-hydroxylase; in-cluster Hpg pathway (HmaS/Hmo/HpgT/PD); heterologous-expression precedent = S. lividans TK24 only (phylum-internal); none in E. coli, none in Bacteroides. Full architecture + provenance in etc/experiments/comp-024-.../inputs/.
Method¶
Nine feasibility factors per host scored 0.0 → 1.0; geometric mean (right composition when every factor must succeed). Factors: cluster size, GC content (Strep ~72% vs EcN ~50.7% vs B. theta ~42.8%), non-canonical AA precursor supply, PPTase, P450 redox, O2 dependence vs. host lifestyle, toolkit maturity, expression precedent, host toxicity / self-resistance. Comparator: C1-INH on EcN as LBP-luminal payload.
Result¶
| Host | Mean | Verdict | Limiting factor |
|---|---|---|---|
| E. coli Nissle 1917 | 0.544 | YELLOW | O2-dependent tailoring (0.30) |
| B. thetaiotaomicron | 0.225 | RED | P450 redox + O2 (both 0.05) |
| C1-INH on EcN (comparator) | 0.774 | GREEN (provisional) | luminal protease stability (0.35) |
Two independent blocking mechanisms¶
(1) O2-dependent tailoring chemistry. ComI/ComJ P450 + ComH halogenase + Hmo oxidase all need molecular O2. Without P450-mediated phenolic coupling the linear peptide lacks the rigid crosslinked architecture giving complestatin C1q/C4b affinity (Park 2016 M55/S56 deletion derivatives are inactive). EcN runs anaerobic metabolism in colon lumen; Bacteroides is strict anaerobe. (2) No precedent outside Actinomycetota. Negative evidence, not proof of infeasibility, but caps engineering-confidence factor for a multi-year campaign at edge of state-of-art.
Comparator: C1-INH (LBP-luminal) is more tractable¶
C1-INH wins because its dominant question (luminal-protease stability + glycosylation) is a single-axis problem testable with a comp-006-style protease-stability analysis on SERPING1. Complestatin's dominant question (O2-dependent tailoring in anaerobic-resident host) has no known engineering workaround. C1-INH's 0.774 is provisional — depends on two factors at 0.35–0.40 not yet validated. The 11-vs-9-factor apples-to-oranges caveat: restricted 7-factor C1-INH mean ~0.70, still meaningfully above complestatin's 0.544.
Multilingual source check¶
No additional canonical primary literature surfaced beyond English-indexed Chiu 2001 / Park 2016 axis. Park 2016 group includes Korean authors; published in English. TCM materia medica has no entry for complestatin-class actinomycete antibiotics. No translation cross-check needed.
Verification-agent pass¶
Self-reviewed as fresh-eyes reader (subagent spawning blocked per task constraints). Catches: (a) original headline framing was internally inconsistent (YELLOW headline + "do not invest" recommendation) — corrected to RED-for-LBP-track since the original question is binary; (b) C1-INH 0.774 reframed as GREEN-provisional; © 11-vs-9-factor caveat now documented. All reflected in outputs/summary.md and experiment README.
Limitations¶
Factor scores are expert estimates (±0.1 robust). Module substrate specificities partly inferred from chloroeremomycin homology. β-OHTyr biosynthesis gene inferred. No Bacteroides NRPS megacluster precedent (absence of precedent ≠ infeasibility). CAI values genus-typical ±0.1. Aerobic-fermentation route out of scope (Park 2016 already showed S. lividans works for production).
Impact on experimental priorities¶
- Complestatin BGC as next LBP-chassis CP0 payload → dropped. Not the right payload for this chassis class.
- C1-INH (LBP-luminal) → promoted to next computational gate (sister to H05; comp-006-pattern protease-stability analysis on SERPING1).
- Complestatin BGC as aerobic-fermentation candidate → parked open per
chassis-pending-interventions.mdpattern. Re-open as separate comp-NNN if an actinomycete-aerobic chassis track surfaces. - Updated CP0 stack: primary = engineered DAF SCR1-4 (koji-secreted, §1.25); next gate = engineered C1-INH (LBP-luminal); non-engineering axis = dietary rosmarinic acid / luteolin; parked = complestatin BGC.