Few foods in the human diet occupy the peculiar intersection of ancient culinary tradition and cutting-edge cardiovascular biochemistry that natto does. Sticky, pungent, threaded with fibrous strands of bacterial biofilm, natto is either immediately repellent or addictively complex depending on your background with fermented foods. But chemistry does not care about palatability. The compound profile produced when Bacillus subtilis var. natto ferments cooked soybeans over 22–24 hours is, by any objective measure, extraordinary.
This guide covers the complete science: the microbiology of B. subtilis fermentation, the mechanism by which nattokinase dissolves fibrin, why the MK-7 form of vitamin K2 found almost exclusively in natto behaves differently from every other dietary K2 source, the clinical evidence for cardiovascular benefit, and a complete home-fermentation protocol you can use tonight.
Important note: Nattokinase has genuine pharmacological activity. If you are on anticoagulant medications (warfarin, heparin, apixaban, rivaroxaban), antiplatelet drugs (aspirin, clopidogrel), or have a bleeding disorder, consult your physician before supplementing or consuming large quantities of natto. This article is educational and not medical advice.
The Microbiology of Bacillus subtilis var. natto Fermentation
The organism at the center of natto's biochemistry is Bacillus subtilis — a gram-positive, rod-shaped, endospore-forming bacterium that exists in soil, plant matter, and the gastrointestinal tracts of many animals. The subspecies or cultivar used for natto production, commonly designated B. subtilis var. natto or B. subtilis natto, is a selected strain optimized over centuries of Japanese culinary practice for aggressive protease and amylase secretion, the production of a characteristic mucilaginous poly-gamma-glutamic acid (γ-PGA) biofilm, and tolerance of the relatively high-temperature fermentation environment required.
Why Soybeans?
The substrate matters enormously. Cooked soybeans provide a near-ideal growth medium for B. subtilis var. natto: roughly 36–40% protein by dry weight (offering abundant amino acid nitrogen), 18–20% lipid (energy substrate), and a complex carbohydrate matrix that B. subtilis amylases can partially digest for fermentation energy. The protein content is particularly critical because nattokinase is itself a protein-digesting enzyme — the raw material for its own biosynthesis is effectively embedded in the substrate.
The Fermentation Sequence
When inoculated soybeans reach the incubation chamber at 38–43°C, B. subtilis var. natto progresses through a predictable growth cycle:
- Lag phase (0–4 hours): Spore germination and early vegetative cell formation. Minimal enzyme secretion. The beans appear unchanged but bacterial density is doubling approximately every 30–45 minutes under ideal conditions.
- Exponential phase (4–14 hours): Explosive bacterial growth. Protease secretion begins in earnest, including the serine protease subtilisin family enzymes. The surface of each bean begins to show early biofilm deposition. Temperature management is critical here — above 45°C, enzyme denaturation begins; below 35°C, competing organisms may gain a foothold and growth rate drops dramatically.
- Stationary/biofilm phase (14–22 hours): Peak nattokinase production. The γ-PGA polymer forms the characteristic sticky white threads that cause natto's distinctive stringiness. Volatile compounds (pyrazines, ammonia from amino acid catabolism) develop the characteristic aroma. Nattokinase concentration in the beans peaks during this window.
- Post-fermentation rest (24–48 hours at 4°C): Refrigeration arrests bacterial growth but allows enzymatic and flavor development to continue slowly. The ammonia sharpness mellows; complex umami compounds deepen. Most traditional producers consider this rest non-optional for flavor quality.
The Poly-Gamma-Glutamic Acid Thread
The white fibrous threads that make natto so visually distinctive — and that confound so many first-time eaters — are strands of poly-gamma-glutamic acid (γ-PGA), a biopolymer secreted by B. subtilis var. natto as a component of its extracellular matrix. γ-PGA is composed entirely of glutamic acid residues linked through their gamma-carboxyl groups rather than the alpha-carboxyl groups used in conventional peptide bonds. This unusual linkage makes it resistant to most peptidases, which is why the threads survive passage through the acidic stomach environment.
Beyond texture, γ-PGA appears to have independent biological activities: it acts as a prebiotic substrate for certain colonic bacteria, may facilitate calcium absorption, and contributes the characteristic tackiness that affects the organoleptic experience of natto. The strand-pulling behavior — stretching the threads between chopsticks — is actually a useful quality indicator: a good ferment produces long, elastic, consistently white threads; poor fermentation produces thin, sparse, or discolored threads.
Nattokinase: Structure, Mechanism, and Fibrinolytic Activity
Nattokinase (officially designated subtilisin NAT, EC 3.4.21.62) was first isolated and characterized by Japanese researcher Hiroyuki Sumi in 1987 during a systematic screen of fermented foods for thrombolytic activity. The story is worth noting: Sumi reportedly dropped a small amount of natto onto a fibrin plate — the standard assay medium for clot-dissolving activity — and observed a dissolution halo several centimeters in diameter after a few hours at body temperature. No other fermented food in the screen came close.
Molecular Structure
Nattokinase is a 275-amino-acid serine protease with a molecular weight of approximately 27.7 kDa. Its active site contains the classic serine protease catalytic triad: Ser221, His64, and Asp32. The enzyme's overall three-dimensional structure is closely homologous to subtilisin Carlsberg, a well-characterized bacterial protease, though nattokinase shows distinct substrate specificity optimized for fibrin cleavage.
The enzyme is stable across a relatively broad pH range (6–12), with optimal activity around pH 10.5, though it retains substantial activity at physiological pH 7.4. Thermal stability is notable: nattokinase maintains its structure and function up to approximately 60°C, which explains why it survives the relatively cool environment of natto fermentation and is not fully destroyed by modest heat exposure, though temperatures above 70°C begin to denature it. This is a critical point for cooking with natto: adding it to hot dishes just before serving preserves enzymatic activity; stir-frying destroys it.
The Fibrinolytic Mechanism
Nattokinase operates through two distinct but complementary pathways to clear fibrin from blood:
Direct fibrinolysis: Nattokinase cleaves fibrin directly — the same substrate that forms the structural backbone of blood clots. It preferentially cleaves fibrin at specific sites, breaking the cross-linked fibrin polymer into soluble degradation products that are cleared by the renal and hepatic systems. In vitro studies have demonstrated that nattokinase cleaves fibrin at a rate approximately 4 times faster than plasmin, the body's own primary fibrinolytic enzyme, under comparable concentrations.
Plasminogen activation: Independently, nattokinase converts plasminogen (the inactive precursor) into plasmin by cleaving the Arg560-Val561 bond, the same site targeted by pharmaceutical thrombolytics like tissue plasminogen activator (tPA). This dual mechanism — direct fibrin dissolution plus endogenous plasmin activation — gives nattokinase a self-amplifying quality: it both dissolves existing clot material and primes the body's own clot-clearing system.
A third effect documented in several studies: nattokinase degrades plasminogen activator inhibitor-1 (PAI-1), the endogenous brake on the fibrinolytic system. PAI-1 levels are elevated in cardiovascular disease, metabolic syndrome, and aging; nattokinase's capacity to reduce PAI-1 provides an additional mechanism independent of direct fibrin or plasmin effects.
Oral Bioavailability: The Critical Question
The first skeptical question about nattokinase is always: does it survive the digestive system? A protein enzyme passing through gastric acid (pH 1.5–3.5) and pepsin exposure would typically be expected to denature and fragment. Early researchers assumed oral bioavailability was negligible. However, multiple subsequent studies have complicated this picture.
Animal studies in rats and dogs demonstrated measurable plasma fibrinolytic activity after oral administration of nattokinase, peaking at 5–8 hours. A 2008 human study (Fujita et al.) administered nattokinase capsules to healthy volunteers and detected significant fibrinolytic activity in plasma with a time-activity profile consistent with intestinal absorption. The mechanism appears to involve partial protection of the enzyme by the mucous layer of the small intestine and possibly by binding to dietary components that reduce pepsin access. Some researchers propose that the enzyme is partially absorbed as an intact or near-intact protein through enterocyte transcytosis.
The practical implication: oral nattokinase from both food and supplements appears to produce measurable systemic fibrinolytic effects in humans, though the bioavailability is lower than parenteral administration and varies significantly by individual digestive function.
Nattokinase Supplement (2,000–4,000 FU)
For those who cannot or will not eat natto regularly, high-potency nattokinase capsules deliver standardized fibrinolytic units with no fermented flavor. Look for products specifying FU (fibrinolytic units) on the label, not just milligrams.
Shop Nattokinase on Amazon →Affiliate link — Borderless Kitchen earns a small commission at no extra cost to you. Always consult your physician before starting any supplement regimen.
Vitamin K2 MK-7: Why Natto's K2 Is Different From Every Other Source
Nattokinase receives most of the research attention in the West, but natto's second major cardiovascular compound may be equally important for long-term arterial health: menaquinone-7 (MK-7), the long-chain form of vitamin K2 that natto produces in extraordinary quantities.
The Vitamin K Family
Vitamin K exists in two primary dietary forms. Vitamin K1 (phylloquinone), found in leafy green vegetables, functions primarily in hepatic clotting factor synthesis. Vitamin K2 (menaquinones, MK-n) is a family of compounds distinguished by the length of their isoprenoid side chain: MK-4 through MK-13, where the number indicates the number of isoprenoid units. The two most biologically significant dietary menaquinones are:
- MK-4: Short-chain menaquinone. Found in animal products (butter, egg yolks, liver, chicken). Half-life in circulation: approximately 1 hour. Activates carboxylation reactions quickly but does not remain in the bloodstream long enough to reach peripheral tissues in significant concentrations.
- MK-7: Long-chain menaquinone. Found in meaningful quantities almost exclusively in natto. Half-life in circulation: approximately 72 hours. Reaches peripheral tissues including arteries, bone, and kidneys at far higher concentrations than MK-4.
The MGP Connection
The critical protein that links vitamin K2 to cardiovascular outcomes is Matrix Gla Protein (MGP), a potent inhibitor of vascular calcification. MGP is produced by vascular smooth muscle cells and must be carboxylated (activated) by vitamin K2 to function. In its uncarboxylated state, MGP is inactive — it cannot bind calcium phosphate crystals and cannot prevent their deposition in arterial walls.
Vascular calcification — hardening of the arteries through calcium phosphate crystal deposition — is one of the strongest independent predictors of cardiovascular mortality. It stiffens arterial walls, increases cardiac afterload, reduces coronary perfusion, and is the central pathological process in many forms of atherosclerosis. Activated (carboxylated) MGP directly binds calcium phosphate crystals and either prevents their nucleation or facilitates their removal from the arterial wall.
The problem: the average Western diet provides roughly 60–70 mcg of vitamin K1 per day through vegetables, but provides minimal MK-7. Tissue studies consistently show high levels of uncarboxylated MGP (ucMGP) in populations with low natto or dairy fat intake — a marker of functional vitamin K2 deficiency in arterial tissue even when clotting function remains normal. This is possible because the liver preferentially uses available vitamin K for clotting factor synthesis; peripheral tissue proteins like MGP get whatever remains.
Natto's MK-7 Concentration
A 100g serving of traditionally fermented natto contains approximately 800–1,100 mcg of MK-7 — a figure that dwarfs every other dietary source. Aged cheeses (Gouda, Brie) contain 50–80 mcg per 100g of mixed menaquinones. Butter, egg yolks, and liver provide primarily MK-4 in concentrations of 10–30 mcg per 100g. No plant food other than natto contains meaningful MK-7.
The recommended supplemental dose for cardiovascular purposes in human trials has typically been 45–360 mcg/day of MK-7. A single 30g serving of natto — the standard Japanese portion size — delivers approximately 240–330 mcg MK-7, placing it solidly in the therapeutic range studied in clinical trials. Daily natto consumption, the norm in many regions of Japan, provides cumulative MK-7 exposure with the long half-life ensuring near-steady-state tissue concentrations.
Population data from Japan support this mechanistically: the Tsurugaya Project and JACC study have documented inverse associations between natto consumption frequency and cardiovascular mortality, bone fracture risk, and all-cause mortality in older adults. These are observational data with confounders, but they align with the mechanistic predictions from MGP and K2 physiology.
Cardiovascular Evidence: What the Research Actually Shows
Interpreting the nattokinase and natto literature requires calibrating your expectations. This is not the database of large, long-term randomized controlled trials that exists for statins or antihypertensives. It is a growing body of mechanistic work, animal studies, small human trials, and population observations that consistently point in the same direction but have not yet produced definitive phase III clinical trial evidence for hard cardiovascular outcomes like myocardial infarction or stroke mortality.
Blood Pressure Reduction
The most consistent human evidence is for modest blood pressure reduction. A 2008 randomized controlled trial by Kim et al. (Hypertension Research) administered nattokinase (2,000 FU/day) to 86 hypertensive subjects over 8 weeks. The nattokinase group showed reductions of 5.5 mmHg systolic and 2.8 mmHg diastolic compared to placebo. A 2012 meta-analysis that pooled similar trial data found comparable effect sizes. The mechanism likely involves nattokinase degrading angiotensin I to produce angiotensin 1-7, which is vasodilatory, in addition to the fibrinolytic effects on blood viscosity.
Atherosclerotic Plaque Studies
A more recent and mechanistically interesting dataset comes from carotid intima-media thickness (CIMT) trials. CIMT is an ultrasound measurement of arterial wall thickness that correlates with systemic atherosclerotic burden. A 2017 randomized trial (Chen et al., JACC Asia) randomized 1,062 participants with existing atherosclerotic plaques to nattokinase supplementation, statin therapy, combination, or control. After 26 weeks, the nattokinase group showed statistically significant reductions in plaque size, with the combination group showing additive effects. This remains one of the largest nattokinase trials in humans and suggests benefit beyond simple fibrinolytic activity.
Vascular Calcification (MK-7 Pathway)
Human intervention studies on MK-7 supplementation have documented dose-dependent reductions in circulating ucMGP — the inactive, uncarboxylated form that reflects K2 insufficiency in vascular tissue. The ECKO trial and subsequent Maastricht group studies demonstrated that 180 mcg/day of MK-7 supplementation for 3 years improved arterial stiffness (measured by carotid-femoral pulse wave velocity) in healthy postmenopausal women compared to placebo, with the largest effects in those with highest ucMGP at baseline. This represents some of the strongest mechanistic-to-clinical linking data in the literature.
Venous Thromboembolism Prevention
Animal and in vitro data are robust for clot dissolution; human data on venous thromboembolism (VTE) prevention remain limited to observational work. A retrospective study of COVID-19 patients with hypercoagulable states found that nattokinase supplementation was associated with reduced D-dimer levels (a marker of clotting activity), but controlled trial data for VTE endpoints are not yet available. This is an active area of research interest.
Evidence Summary Table
| Compound | Target Function | Studied Dose | Best Evidence Level | Verdict |
|---|---|---|---|---|
| Nattokinase | Direct fibrinolysis, plasminogen activation | 2,000–4,000 FU/day | Multiple small RCTs, 1 large RCT (n=1,062) | Strong mechanistic; moderate clinical |
| Nattokinase | Blood pressure reduction (ACE-inhibitor-like) | 2,000 FU/day | RCT (n=86), meta-analysis | Moderate — consistent 3–6 mmHg SBP effect |
| MK-7 (natto) | MGP carboxylation, vascular calcification inhibition | 45–360 mcg/day | RCTs, mechanistic trials, 3-year outcome data | Strong for biomarker; promising for outcomes |
| γ-PGA | Prebiotic activity, mineral absorption | Food quantities | Animal and in vitro primarily | Emerging — insufficient human data |
| Isoflavones (soy) | Endothelial function, LDL oxidation | 40–80 mg/day | Multiple RCTs, meta-analyses | Moderate — modest LDL and vascular effects |
Making Natto at Home: Complete Fermentation Protocol
Home fermentation of natto is genuinely achievable without specialized equipment. The requirements are minimal: soybeans, starter culture (either dried spore powder or a small amount of commercial natto), a temperature-controlled environment in the 38–43°C range (a dehydrator, Instant Pot yogurt setting, or oven with the light on work well), and patience across a 36-hour window from soak to ready-to-eat.
The most common failure modes are contamination (usually from competing bacteria that produce off-odors rather than the clean, sharp ammonia-funk of good natto), insufficient temperature (resulting in poor fermentation and minimal enzyme production), and excessive moisture (causing the beans to become slimy rather than covered in clean white threads).
Natto Starter Spores — Bacillus subtilis var. natto
High-viability dried spore powder specifically selected for natto production. A single packet inoculates multiple kilograms of soybeans. Store in the freezer for extended shelf life between batches.
Shop Natto Starter on Amazon →Affiliate link — Borderless Kitchen earns a small commission at no extra cost to you.
- Soak: 12–18 hours Cover 500g dried soybeans with 3× their volume of cold water. Soak 12 hours minimum; 18 hours is ideal in winter or for larger beans. Beans should double in volume and show no dry center when split.
- Cook: Steam or pressure-cook until very soft Steam for 3–4 hours (traditional) or pressure-cook at 15 psi for 45–60 minutes. Test: a bean should mash between two fingers with minimal force. Firmer beans ferment unevenly; overcooked beans can become too wet. Drain thoroughly — surface moisture invites mold.
- Prepare inoculant solution Dissolve 0.1–0.5g of dried natto starter spores (approximately a match-head volume) in 2 tablespoons of sterile water. Alternatively, thin 2 tablespoons of commercial natto with 2 tablespoons water to use as starter. Prepare this while beans are still hot (not yet touchable).
- Inoculate at the right temperature Allow beans to cool to 43–45°C (a temperature you can just hold your hand near but not touch comfortably). Pour inoculant over beans and mix thoroughly with a sterilized spoon or spatula. This temperature window is critical: too hot kills the spores; too cool allows competing bacteria to establish before B. subtilis dominates.
- Pack into shallow containers Distribute inoculated beans in a shallow layer (2–3cm deep maximum) in food-safe containers. The shallower the layer, the better aeration and the more uniform the fermentation. Cover containers with parchment paper rather than lids to allow CO₂ off-gassing while maintaining humidity. A few small holes in the parchment improve results.
- Ferment at 38–43°C for 22–24 hours Maintain temperature as consistently as possible — fluctuations are the primary cause of uneven batches. Check at 16 hours: the surface should show a thin white haze of bacterial growth. At 20–22 hours, white threads should be clearly visible. Ferment to 24 hours for full nattokinase development. Longer than 28 hours risks ammonia over-production and bitterness.
- Refrigerate for 12–48 hours (the "ripening" stage) Transfer containers to 4°C refrigerator. This step is not optional for flavor — it significantly mellows the ammonia sharpness and develops complex secondary notes (earthy, savory, faintly nutty). The nattokinase content does not decrease appreciably during this cold rest. Most producers consider 24-hour cold rest the minimum for acceptable flavor.
- Serve and store correctly Consume within 7–10 days refrigerated for optimal enzymatic activity, or freeze in small portions for up to 3 months. For maximum nattokinase activity: consume raw or add to dishes after they have cooled below 60°C. Stir vigorously before eating — the mixing action maximizes the characteristic string development and is considered part of the traditional preparation. Traditional accompaniments: soy sauce, karashi mustard, green onion, raw quail egg yolk.
Quality Indicators and Troubleshooting
Good fermentation signs: Dense white threading between beans, clean sharp ammonia-fermented smell (not rancid or sulfurous), beans that are firm but tender, no visible mold growth (green, black, or fuzzy patches indicate contamination from ambient mold spores and the batch should be discarded).
Thin or absent threads: Usually temperature too low, starter too old or insufficiently active, or beans too dry. Ensure beans are fully cooked and that incubation temperature stays above 38°C continuously.
Off-smell (rotten vs. fermented): The smell of good natto is aggressive and ammonia-sharp but clean. Rotten eggs or fecal odor indicates contamination by competing organisms — most commonly coliform bacteria from inadequately sterilized equipment or beans that were not hot enough when inoculated. All equipment should be sterilized with boiling water or food-grade alcohol before use.
Maximizing nattokinase content: Peak enzyme production correlates with fermentation temperature consistency and adequate oxygenation. Thin layers in wide containers outperform deep containers. Commercial producers use perforated trays with forced airflow; at home, parchment covering with small holes in a dehydrator tray approximates this well.
Incorporating Natto Into Your Diet
If you are not acclimated to natto, cold-turkey consumption of the traditional 50g morning portion over rice can be a significant palatability challenge. Several strategies help:
- The miso soup introduction: Stir a small amount of natto (10–15g) into hot miso soup just before eating. The miso's strong umami masks some of the natto flavor while familiarizing the palate with the texture.
- Natto toast: Spread natto on sourdough toast with avocado, a soft-poached egg, and sriracha. The fat from avocado and the acid from sriracha balance the ammonia notes.
- Natto pasta: Toss cold natto with hot pasta (the heat partially tames the flavor while staying below the 60°C denaturation threshold), mentsuyu sauce, and a raw egg yolk.
- Natto kimchi bowl: Combine with kimchi over rice — two fermented cultures that produce complementary flavor profiles. The lactic acid of kimchi provides a bright counterpoint to natto's alkaline sharpness.
Regular consumers report palate adaptation within 2–4 weeks of daily consumption. The sticky texture is typically the last obstacle for Western consumers; the flavor, surprisingly, often becomes enjoyable before the tactile experience does.