Koji Home Fermentation — Wild-Type Protocol for Digestive Enzyme Delivery¶

Practical small-batch protocol for producing food-integrated digestive enzymes at home using commercial wild-type koji starter, as an alternative or adjunct to prescription pancreatic enzyme replacement therapy (PERT) for mild-to-moderate exocrine pancreatic insufficiency (EPI).

This is distinct from the project's engineered-strain work in engineered-koji-protocol.md — this page is the wild-type / pre-engineering anchor: what's possible with off-the-shelf koji-kin and home equipment, and what a project-scale engineered version needs to outperform to justify itself.

The fundamental confusion: koji-kin vs. koji rice¶

The two are different things, and most online recipes assume you know which one you have.

Term	What it is	What it looks like	Use
Koji-kin (種麹, tane-koji)	Dried Aspergillus spore inoculum on a small carrier (typically rice flour)	Pink/purple foil packet, fine pinkish-tan powder	Starting material. A pinch inoculates ~1 kg of cooked rice.
Koji rice (麹, kōji)	Cooked rice that has been inoculated with koji-kin and fully colonized by mycelium over 42–48 h	Whole rice grains covered in white fuzzy mycelium, sweet/floral smell	Working enzyme substrate. This is what's called for in shio-koji / amazake / miso recipes.

Common error: eating raw koji rice as a finished food. Koji rice on its own is mildly sweet but unpalatable — it's not meant to be a finished food, it's a fermentation tool that transforms other foods.

The two-stage process¶

┌─────────────────────┐    ┌──────────────────┐    ┌──────────────────────┐
│  Stage 1            │    │  Stage 2         │    │  Use                 │
│  Koji-kin           │ →  │  Koji rice       │ →  │  Shio-koji marinade  │
│  + steamed rice     │    │  + salt + water  │    │  Amazake drink       │
│  (42–48 h, ~37°C)   │    │  (7–14 d, RT)    │    │  Miso (long-term)    │
└─────────────────────┘    └──────────────────┘    └──────────────────────┘

You only buy koji-kin once per ~15 kg of rice you plan to inoculate. Stage 1 produces the koji rice; Stage 2 (and beyond) is where the digestive-enzyme value is delivered to food.

Stage 1: Koji-kin → koji rice¶

Equipment¶

Steamer or rice cooker (steamer preferred — koji rice needs steamed, not boiled, rice)
Incubation environment that holds 32–40°C for ~48 hours: yogurt maker, dehydrator, oven with light bulb, sous vide bath with rack, or a styrofoam cooler with a heating pad
Thermometer (probe-style, food-safe)
Damp clean cotton cloth or muslin
Shallow tray or wooden box (cedar traditional; ceramic, glass, or food-grade plastic also work)

Substrate options¶

Short-grain white rice (standard; sweetest amazake, mildest miso) — start here
Brown rice (more fiber, harder for mycelium to penetrate; possible but advanced)
Pearled barley (麦麹, mugi-koji; traditional in barley miso)
Soybeans (豆麹, mame-koji; for some miso styles; not beginner)

For digestive-enzyme purposes, short-grain white rice is the highest-yield substrate — most starch surface area for amylase, fewest physical barriers to mycelium colonization, sweetest result.

Substrate engineering principles cross-apply (added 2026-05-19, source: substrate-engineering-mushroom-cultivation-lit-scan). The substrate-engineering lit scan (logs/substrate-engineering-mushroom-cultivation-lit-scan-2026-05-19.md) identified four mechanisms — passive accumulation, biotransformation, BGC induction, precursor feeding — all operative in A. oryzae / koji. Carbon-source choice (rice vs. soy vs. mixed grain) and nitrogen-source choice modulate secondary-metabolite expression in koji the same way they do in basidiomycete medicinal mushrooms. The 2 mM methionine supplementation that delivers 1.7–3.1× ergothioneine in Ganoderma / Pleurotus is also documented to enhance ergothioneine in A. oryzae per Lee 2009 (PMC3749454). See Platform Principle 9 (etc/open-source-platform.md) for the platform-level discipline and medicinal-mushroom-extract-sops.md §SOP-7 for the protocol matrix.

Procedure¶

Soak rice 6–12 h in cold water. Drain.
Steam rice 45–60 min until al dente — grains should be cooked through but firm, not mushy. (Boiled rice is too wet and will rot instead of growing koji.)
Cool to ~32–33°C by spreading on a clean tray and turning. Aim for the rice to feel just barely warm.
Inoculate. Sprinkle koji-kin spores evenly across the rice (a small pinch per kg of rice — Hishiroku's instructions say 20 g inoculum / 15 kg rice, so for 1 kg rice you need about 1.3 g — start with ½ tsp). Mix thoroughly with clean hands.
Wrap loosely in damp cloth and place in incubation chamber at 32–33°C.
Turning schedule (this is where the Hishiroku bag's 第1〜3回手入れ instructions come in):
15–20 h after inoculation (1^st turning, 第1回手入れ): rice should be warm to the touch, faint sweet smell. Break up clumps, mix gently, re-wrap. Target temp: 37–39°C.
~24 h (2^nd turning, 第2回手入れ): white mycelium visible. Mix again. Spread to a slightly thicker layer to retain heat. Target: 38–40°C.
~36 h (3^rd turning, 第3回手入れ): mycelium dense and white, sweet/floral smell ("chestnut" / "popcorn" notes). Target: 37–40°C.
Done at 42–48 h total. Rice grains should be visibly fuzzy white, not separated, distinctly sweet-smelling. Should NOT smell of ammonia, alcohol, or be greenish (= contamination, discard).
Refrigerate immediately to halt activity. Use within 1–2 weeks, or freeze for up to 6 months.

Troubleshooting¶

Problem	Likely cause	Fix
No mycelium growth	Temperature too low; rice too wet/dry; expired koji-kin	Check thermometer; rice should be al dente; replace starter
Green/black/pink patches (not white)	Contamination (Penicillium, bacteria)	Discard. Improve sanitation, lower temp
Ammonia smell	Over-fermented (>50 h, too warm)	Discard or use immediately for shio-koji (the salt will mask it)
Yeasty/alcohol smell	Wild yeast competition; rice too wet	Discard; reduce moisture next batch

Stage 2A: Koji rice → shio-koji (salt-koji marinade)¶

This is the highest-leverage application for EPI / digestive-enzyme purposes. Shio-koji is a salt-stable enzyme paste used as a marinade or seasoning — the proteases and amylases pre-digest food in the marinade phase, before it reaches the eater.

Recipe¶

200 g koji rice (from Stage 1)
60 g sea salt or kosher salt (~30% by weight of koji)
250 mL filtered water (~125% by weight of koji)

Combine in a glass jar (1 L size, leave headspace). Stir thoroughly. Cover loosely (cloth + rubber band, or partly-screwed lid — gas needs to escape).

Ferment at room temperature 7–14 days. Stir daily. The mixture will smell sweet, slightly cheesy, and umami as it matures. When the rice grains have softened and mostly dissolved into a porridge-like consistency, blend smooth with an immersion blender. Refrigerate.

Shelf life: 6+ months refrigerated.

Use¶

Marinade meat/fish: apply 5–10% by weight (10 g shio-koji per 100 g protein). Marinate 30 min (light) to 24 h (deep). Brush off excess before cooking — the sugars from amylase activity will burn quickly.
Vegetables: lighter application (3–5%), shorter time (10 min – 2 h)
Salad dressing / sauce base: dilute with vinegar, oil, citrus

Why this helps with EPI (mechanistic extrapolation)¶

Shio-koji's protease pre-digests proteins in the marinade phase. The protein arriving in the eater's small intestine is already partially hydrolyzed (peptides, free amino acids), reducing pancreatic protease demand at the table. Amylase activity does the same for starch, though at lower magnitude in a marinade application (most amylase action happens during cooking with koji-marinated starches, e.g., koji-cured rice dishes).

Evidence level: Mechanistic extrapolation. The enzyme activities are well-characterized in vitro; no RCT-grade data on shio-koji as PERT-equivalent. Anecdotal reports from EPI patients exist but are not formal evidence.

Stage 2B: Koji rice → amazake (sweet rice drink)¶

Traditional Japanese pre-meal digestive aid. Sweet, non-alcoholic, drunk warm or chilled.

Recipe¶

200 g koji rice (from Stage 1)
200 g cooked rice (any short-grain)
400 mL water at 60°C

Combine. Hold at 55–60°C for 8–10 hours (yogurt maker, slow cooker on warm with lid ajar, sous vide, or thermos). The mixture will become progressively sweeter as koji's α-amylase converts rice starch to maltose and glucose.

When the desired sweetness is reached, heat briefly to 80°C to deactivate enzymes (otherwise it will continue to break down and eventually go sour as wild yeast takes over). Refrigerate.

Use: sip 100–200 mL warm before meals as digestive aid. Or use as a sweetener in tea, smoothies. Refrigerated shelf life: ~10 days.

Evidence level: Traditional / folk practice with Japanese clinical literature on amazake's effects on gut microbiota and metabolic markers (multiple papers; modest effect sizes). Not a substitute for prescribed PERT in clinically diagnosed EPI.

Drug-interaction warning: disulfiram + residual ethanol¶

A "failed" amazake or shio-koji batch — wild-yeast contamination, missed 80°C inactivation, extended hold at sweet-stage temperature — can ferment past glucose into ethanol (1–3% in worst cases). For anyone taking disulfiram (Antabuse), even sub-percent ethanol is enough to trigger the acetaldehyde reaction (flushing, tachycardia, hypotension, severe nausea).

Routine precaution if disulfiram is in the stack:

Pre-screen every batch with consumer ethanol test strips (~$15 for a strip pack; detection ±0.1% v/v). Test the finished amazake or shio-koji liquid before consuming.
Discard any batch reading >0.1%. Do not "cook off" — disulfiram users should not rely on volatilization for safety.
Validation/edge cases: send suspect batches to a lab for gas chromatography ($30–80 per sample) only when test strips are ambiguous or the reaction profile (flushing on what should be a clean batch) suggests sub-strip ethanol.

This is the same risk surface as kombucha for disulfiram users — see disulfiram.md for the full interaction table.

Stage 2C: Koji rice → miso (long-term aging)¶

Beyond scope here. Standard ratio: 1 part koji + 1 part cooked soybeans + 0.4 part salt, age 6–12 months minimum. Miso protocols are widely available; reference Christian Weij's Miso, Tempeh, Natto & Other Tasty Ferments or the Cold Mountain miso instructions if pursuing this.

Yellow vs. white vs. black koji for home use¶

Strain	Species	Strengths	Best for
Yellow koji	Aspergillus oryzae	Highest amylase + standard protease; sweetest amazake; most documentation	Default choice for digestive-enzyme home use. Sake, sweet miso, amazake, shio-koji.
White koji	A. luchuensis var. kawachii	High citric acid (anti-contamination in warm climates); good amylase	Shochu, awamori, sour-leaning miso. Tangier shio-koji.
Black koji	A. luchuensis var. awamori	Highest citric acid; robust in hot/humid conditions	Awamori (Okinawan distillate); rare in DIY use

Recommendation for the EPI / home-PERT-alternative use case: yellow koji (A. oryzae) is the better-fit default. White koji is functional but tangier and has slightly lower diastatic power.

If you have white koji on hand: use it first to learn the process (Stage 1 + a basic shio-koji or amazake batch). Switch to yellow when restocking — Cold Mountain (Marukawa Miso, Japan; widely sold in US Asian groceries) and GEM Cultures (Washington State, US) are accessible yellow-koji starter sources.

Format Constraints for Engineered Payloads — Shio-Koji Is Not Universal¶

A critical constraint from engineered-koji-protocol.md §15: shio-koji cannot be used as a delivery format for peptide payloads. The 7–14 day salt ferment maintains active native proteases at room temperature. Any small peptide — carnosine (β-alanyl-L-histidine), KPV (Lys-Pro-Val), BPC-157 (gastric pentadecapeptide), or any future therapeutic peptide — will be hydrolyzed back to its constituent amino acids during the shio-koji fermentation window, destroying the therapeutic molecule. (Mechanistic Extrapolation; source: engineered-koji-protocol.md §15)

Format ranking for peptide payloads, by survival expectation:

Format	Peptide survival	Reason
Dried koji powder (heat-inactivated)	Highest	Heat denatures native proteases; peptides thermally stable across typical food-processing temperatures
Amazake (cooked, <24 h, finished at 80°C)	High	Brief enzyme exposure, then heat inactivation
Fresh koji (refrigerated, days, no salt)	Medium	Proteases active but short exposure window; cold storage slows hydrolysis
Shio-koji (7–14 day salt ferment)	Effectively zero	Sustained protease exposure; salt does not protect peptide bonds

Shio-koji remains the optimal format for robustly folded enzyme payloads (uricase tetramer, lactoferrin glycoprotein) where conformational stability and disulfide bonding provide protease resistance — not exposed peptide bonds. For the multi-format endgame strain, this implies a split delivery model: shio-koji for live-enzyme payloads, dried powder or amazake for peptide payloads. (source: engineered-koji-protocol.md §15)

Project relevance¶

This protocol is the wild-type baseline that the engineered-koji platform (see engineered-koji-protocol.md, koji-construct-design.md, koji-endgame-strain.md) must outperform to justify itself for EPI applications.

Key questions that home wild-type fermentation can illuminate:

Is shio-koji-marinated protein a meaningful PERT-reducer in mild-to-moderate EPI? N=1 / household trials with PERT-dose-per-meal tracking would generate informative observational data.
What's the practical enzyme-activity ceiling of wild-type koji-fermented foods? Measurable in lab (amylase / protease assays of finished shio-koji) vs. commercial PERT activity per dose.
What's the palatability + adherence ceiling? A homemade enzyme strategy only works if the user actually eats the resulting food regularly.

The engineered-strain platform's value proposition gets sharpened by understanding wild-type performance: how much enzyme activity per gram of food at the eater's gastric pH can wild-type already deliver, and where does engineering need to add value (specific activity, pH stability, lipase boost, gastric survival)?

A pedagogical note for the engineered-strain rollout: the engineered-strain platform's distribution model — a frozen / lyophilized master spore stock that users return to for each new starter, with no backslopping past a small generation count — maps directly onto the tane-koji → koji rice pattern documented above. The "buy tane-koji once per ~15 kg of rice" rule already is the limited-generation propagation rule that industrial fermentation enforces with frozen master cell banks. See open-source-platform.md §"Cultural mapping: tane-koji as master, koji rice as working batch" for how the platform's Strain Stability Kit lines up against this convention.

Open questions¶

Quantitative comparison of shio-koji enzyme activity (units/g) vs. commercial PERT (Creon, Zenpep) units per pill. Lab-measurable; methodology now specified in enzyme-quantification-protocol.md (tiered home → bench → outsource path). Tier 3 first-run is the load-bearing experiment that converts ongoing batch QC into clinically interpretable percent-of-Creon-cap-equivalent figures.
Gastric survival of shio-koji-derived enzymes vs. enteric-coated PERT. Hypothesis: poor without coating; useful only for pre-digestion in marinade phase, not in-gut activity post-ingestion.
Lipase activity of A. oryzae shio-koji (low compared to A. niger or engineered strains) — likely the limiting digestive-enzyme axis for fat malabsorption phenotype EPI.
Comparative evidence: any human studies of koji-fermented diets in EPI specifically? (Not aware of any; would be valuable.)