Skip to content

Phase 6 — Triage of Phase 5-Surviving Leads

Date: 2026-05-06 Method: comp-013-style production-route + dose-feasibility + chokepoint-occupancy assessment, adapted for non-small-molecule compound classes.

Summary

  • PURSUE (koji-track / chemical-synthesis): DAE, cordycepin
  • PURSUE in Phase 7 (medicinal-mushroom cultivation track): GLPP
  • DEFER (Phase 5b prerequisites): AMC-BFE, FZ-Poria
  • DROP: lovastatin

DAE — methyl 2,4-dihydroxybenzoate (DAE)

Class: small molecule (phenolic ester) MW: 168.15 g/mol Source: Ganoderma applanatum (computational attribution; commercial reagent in PMID 35750011) Chokepoints: XO, URAT1 (expression-level)

Verdict: PURSUE

Rationale: Defined small molecule with verified mouse dose-response. Human equivalent dose (~455 mg/day at top mouse dose) is supplement-grade and feasible. Chemical synthesis is the right production route — heterologous expression in koji is not warranted for a simple methyl ester. Wet-lab gate is straightforward.

Dosing math: Mouse top dose → human equivalent ≈ 6.5 mg/kg → ~455 mg/day for a 70 kg person.

Production routes:

  • chemical synthesis ⭐ — trivial (Fischer esterification of 2,4-dihydroxybenzoic acid + methanol); cost: <$1/g at 100g scale
  • commercial purchase — available (Sigma, Combi-Blocks); cost: ~$50/g pharma grade
  • extraction from G. applanatum — low — natural abundance not validated; Model B flagged this gap; cost: high
  • heterologous expression in koji — not warranted — small phenolic, synthesis is much easier than engineering BGC; cost: n/a

Wet-lab gate:

  • Synthesize or purchase 100g methyl 2,4-dihydroxybenzoate (≥99% purity, HPLC)
  • PO+HX mouse HUA model, dose-response 20/40/80 mg/kg PO daily x 14d, n=8/group
  • Primary endpoint: SUA reduction at d14
  • Mechanism: liver XO activity + renal URAT1 mRNA/protein at endpoint
  • Safety: ALT/AST, BUN/Cr, body weight, liver/kidney histology
  • Critical: confirm SUA does not over-correct below normal mouse SUA (~88 µmol/L) — Phase 5 flagged 134 is sub-normal-control

Phase 5b verifications outstanding:

  • Full PDF of PMID 35750011 (~$30) — verify XOD IC50, kinetic constants, inhibition type
  • CNKI dive on Yong/Liang group's TCM diuresis literature (per Model A recommendation)
  • Confirm DAE natural abundance in G. applanatum (Model B flag)

cordycepin — cordycepin (3'-deoxyadenosine)

Class: small molecule (nucleoside analog) MW: 251.24 g/mol Source: Cordyceps militaris (cultivated) — abundant; sparse in wild O. sinensis Chokepoints: URAT1

Verdict: PURSUE

Rationale: Production route is commercial fermentation (mature) or home C. militaris cultivation (substrate-limited but feasible). Mechanism (URAT1 modulation) is in-vivo-validated. The BA caveat is the binding constraint — combination with ADA inhibitor (e.g., GLPP from G. lucidum?) could dramatically extend half-life. Adjacent to GLPP — the comp-014 GLPP+cordycepin combination is a real Phase 6 question.

Dosing math: Mouse top dose → human equivalent ≈ 2.0 mg/kg → ~142 mg/day for a 70 kg person.

BA caveat: Cordycepin has notoriously poor oral BA — rapidly deaminated to 3'-deoxyinosine by serum/intestinal ADA. Effective BA may be <5% without prodrug or co-administration with ADA inhibitor.

Production routes:

  • C. militaris liquid fermentation (commercial) ⭐ — well-established; commercial cordycepin available at $200-500/g pharma-grade; cost: ~$200-500/g
  • Shih et al. published optimized fermentation conditions; this is the production-mature route
  • C. militaris solid-state fermentation (home) — moderate — slower than liquid culture; brown rice substrate works; cost: low (substrate + spores)
  • Home-grown Cordyceps fruiting body cordycepin content varies 0.5-5 mg/g dry weight
  • chemical synthesis — possible but more complex than DAE (nucleoside chemistry); not cost-competitive with fermentation; cost: high
  • heterologous expression in koji — not warranted — C. militaris fermentation is mature; cost: n/a

Wet-lab gate:

  • Source: commercial cordycepin (≥98% HPLC purity) for first wet-lab pass; Phase 7 home-cultivation comparison
  • PO HUA mouse model, dose-response 6.25/12.5/25 mg/kg PO daily x 14d
  • Primary endpoint: SUA reduction
  • Mechanism arm: cordycepin alone vs cordycepin + ADA-inhibitor (pentostatin or GLPP)
  • Pharmacokinetics: cordycepin + 3'-deoxyinosine plasma levels at multiple timepoints
  • Critical question: does ADA inhibition co-treatment extend cordycepin half-life enough to lower the daily dose?

Phase 5b verifications outstanding:

  • PK studies with/without ADA inhibitor co-administration
  • Verify cordycepin content in commercially available C. militaris fruiting bodies (literature varies 0.5-5 mg/g dry weight)

Synergy candidates: - GLPP (G. lucidum) — ADA inhibition could extend cordycepin half-life. Both are accessible via medicinal mushroom cultivation. This is the Phase 7 medicinal-mushroom-complement integration point.

GLPP — GLPP (Ganoderma lucidum polysaccharide-peptide)

Class: polysaccharide-peptide hybrid (~520 kDa; glycopeptide vs proteoglycan distinction unresolved per Model B) MW: 520000 Da Source: Ganoderma lucidum mycelium (Juncao substrate, per Yang lab) Chokepoints: ADA, GLUT9, OAT1, (Keap1/Nrf2 in sister paper PMC11351902)

Verdict: PURSUE in Phase 7 (cultivation track, not koji-engineering)

Rationale: Polysaccharide-peptide is fundamentally not a koji-engineering target — it's a fermentation product of G. lucidum mycelium. This is exactly the Phase 7 medicinal-mushroom-complement strategy use case. Native cultivation route is mature; the binding question is strain selection + extraction protocol standardization rather than genetic engineering. Polysaccharides have ~0% systemic bioavailability — they work via gut signaling (immune modulation, microbiome), not direct ADA binding. The 'ADA inhibition' mechanism is likely indirect via gut-associated lymphoid tissue.

Dosing math: Mouse top dose → human equivalent ≈ 32.5 mg/kg → ~2276 mg/day for a 70 kg person.

Production routes:

  • G. lucidum mycelium liquid fermentation ⭐ — well-established commercial process; multiple manufacturers; cost: ~$1-5/g extract grade
  • Phase 7 territory — native cultivation track, not koji-engineering track
  • G. lucidum solid-state fermentation (home) — moderate — Reishi mushroom kits available consumer-grade; fruiting body extraction yields lower polysaccharide than mycelium; cost: low ($10-30 grow kit)
  • G. lucidum fruiting body extraction (commercial supplement) — ubiquitous; quality varies wildly; 'GLPP' specifically requires defined extraction protocol; cost: consumer supplements $0.50-2/g but uncharacterized
  • heterologous expression in koji — infeasible — polysaccharide-peptide is biosynthesis product, not single gene; mycelium-specific glycosylation machinery; cost: n/a

Wet-lab gate (Phase 7 cultivation track):

  • Strain selection: compare 3-5 commercial G. lucidum strains for GLPP yield + composition
  • Cultivation method comparison: liquid fermentation (defined media) vs Juncao substrate vs other solid-state
  • Extraction protocol: hot water > ethanol precipitation > Sephacryl S-500 chromatography (per Yang lab method)
  • Characterization: MW (SEC-MALS), peptide composition (amino acid analysis), glycan linkage (NMR)
  • Mechanism question: is ADA inhibition direct or indirect? Test purified GLPP fraction vs intestinal lysate ADA in vitro
  • Mouse HUA model verification: 200/400 mg/kg PO daily, replicate 40.6% UA reduction claim

Phase 5b verifications outstanding:

  • Full PDF of PMID 36385640 (Food & Function paywalled)
  • Sister paper PMC11351902 review for Keap1/Nrf2 mechanism evidence
  • Resolve glycopeptide vs proteoglycan distinction (Model B flag) — affects pharmacology framing

Synergy candidates: - cordycepin — GLPP's ADA inhibition could extend cordycepin's notoriously short half-life. Both are home-cultivable medicinal mushrooms. Phase 7 medicinal-mushroom-complement integration.

AMC-BFE — AMC-BFE (Cordyceps militaris × Astragalus membranaceus solid-state fermentation extract)

Class: whole extract — active components NOT identified Source: Cordyceps militaris co-fermented with Astragalus Chokepoints: URAT1, GLUT9, (hepatic) ABCG2, PPARα downstream

Verdict: DEFER — full PDF retrieval required before triage

Rationale: Phase 5 deep-read flagged that authors explicitly hedge 'contribution of individual metabolites requires further investigation.' Whole-extract evidence cannot be triaged with comp-013 IC50-occupancy methodology — there's no defined active compound to model. Phase 6 cannot complete this triage without (a) full PDF + (b) compound-fractionation work.

Production routes:

  • solid-state co-fermentation (academic protocol) — documented in analog studies (Zhao 2025 PMC12805540); reproducibility uncertain without full PDF; cost: moderate — substrate + dual-strain inoculum

Wet-lab gate:

BLOCKED on Phase 5b: retrieve full PDF, attempt compound fractionation, then re-triage. Estimated $30 + 4-6 weeks fractionation → re-enter Phase 6.

lovastatin — lovastatin (mevinolin)

Class: small molecule (HMG-CoA reductase inhibitor, statin class) MW: 404.54 g/mol Source: Aspergillus terreus (industrial), Pleurotus ostreatus (native producer) Chokepoints: HDAC6 (in vitro 16.3 µM), PPARγ (functional 1 µM, Kd ~110 µM)

Verdict: DROP

Rationale: In vitro HDAC6 + PPARγ IC50s are ~100-1000× higher than clinical lovastatin Cmax. The chokepoint hits are pharmacologically irrelevant at clinical doses. Lovastatin's gout-axis relevance, IF any, is via hepatic HMG-CoA reductase inhibition reducing isoprenoid synthesis (NLRP3-priming-adjacent), NOT via the ChEMBL HDAC6/PPARγ off-targets. Statins are already a commercial drug class — Open Enzyme has nothing to add by re-engineering them.

Clinical exposure at standard dose: Lovastatin 20-80 mg/day → systemic Cmax ~50-200 nM In vitro chokepoint IC50: HDAC6 16.3 µM; PPARγ Kd ~110 µM Occupancy at clinical exposure: <0.5% at HDAC6; <0.1% at PPARγ

Wet-lab gate:

Not applicable — DROPPED for chokepoint occupancy reasons. The Pleurotus-as-native-lovastatin-producer thesis remains valid but for HMG-CoA-reductase pathway, not the comp-014 chokepoint set.

FZ-Poria — Wolfiporia cocos (Poria cocos) in FZ multi-herb formula

Class: multi-herb formula component (Poria + Pinellia + Cinnamomum + Achyranthes + Atractylodes) Source: Wolfiporia cocos sclerotium (TCM 茯苓) Chokepoints: ABCG2, GLUT9, OAT1, NLRP3/ASC

Verdict: DEFER — needs single-component Poria study before advancing

Rationale: FZ formula contains 5+ herbs. The chokepoint-modulation is real for the formula but unattributable to Poria specifically. A standalone Poria study would be needed to triage. CNKI likely has single-component Poria pharmacology papers (Wolfiporia is heavily studied in TCM); a Phase 5b CNKI dive could resolve this.

Synergy pairs

GLPP + cordycepin

GLPP's ADA inhibition could extend cordycepin's notoriously short half-life (cordycepin is rapidly deaminated to 3'-deoxyinosine by serum/intestinal ADA). Both are accessible via medicinal mushroom cultivation (G. lucidum + C. militaris). This is the cleanest Phase 7 medicinal-mushroom-complement integration point — two home-cultivable organisms producing complementary compounds.

Wet-lab question: Does combined GLPP + cordycepin co-administration achieve URAT1 modulation at lower individual doses than either alone, with acceptable PK/PD?

Phase 7 handoff

Compounds routed to cultivation track: GLPP (G. lucidum), cordycepin (C. militaris), ergothioneine (Pleurotus ostreatus, koji A. oryzae) — comp-014 anchor species

Compounds staying with koji engineering track: DAE (chemical synthesis preferred; not engineering candidate), DAF SCR1-4 / uricase / lactoferrin (existing koji engineering thesis, not comp-014 derived)

Platform thesis expansion: Engineering chassis (koji) + native-compound complement (medicinal mushrooms) = expanded Open Enzyme platform. Phase 7 scope page formalizes the parallel track.