Larazotide

AT-1001

Larazotide (AT-1001) is a healing & recovery research compound. Zonulin antagonist — tightens intestinal tight junctions to reduce 'leaky gut' permeability; furthest-along gut-barrier drug (celiac trials).

How Larazotide works

Zonulin antagonist — tightens intestinal tight junctions to reduce 'leaky gut' permeability; furthest-along gut-barrier drug (celiac trials).

Proposed benefits

Researched for soft-tissue and gut repair, reduced inflammation, angiogenesis and faster recovery from injury.

Where to get Larazotide

The evidence for Larazotide

Graded by what exists behind each claim.

✅ Clinically validated

• Phase 3 in celiac disease — and it is an instructive failure. Larazotide reached the largest celiac trial program ever run. The phase 3 was stopped early for futility against its primary endpoint.
• Its earlier phase 2b had been positive on symptom scores, which is exactly the pattern that makes phase 3 necessary. Worth noting what the trial was asking: symptom relief in people already on a gluten-free diet, which is a narrow and demanding endpoint, not 'does it protect the gut barrier'.

📊 Correlative data

• Community use for intestinal permeability, largely on the strength of the mechanism rather than the phase 3 result. Reported experience is subtle.

🧪 Theoretical / extrapolated

• An octapeptide zonulin antagonist — zonulin regulates the tight junctions between intestinal epithelial cells, and larazotide is designed to keep them closed.
• The mechanism is well characterized and the tight-junction effect is demonstrable. What the phase 3 questioned is whether closing tight junctions translates into how someone feels — a gap between mechanism and outcome that recurs constantly in gut-barrier research.

How to read these tiers: they say how much human evidence exists, not how well something works — and ✗ flags harm, never a disappointing trial. How the evidence tiers work →

What Larazotide actually does

Larazotide is an eight-residue peptide, Gly-Gly-Val-Leu-Val-Gln-Pro-Gly, and its ancestry is worth knowing: it is modeled on a fragment of the Vibrio cholerae zonula occludens toxin, the bacterial protein that opens intestinal tight junctions during cholera. Add the residue masses and add water and you get about 726 Da — small, uncharged apart from its termini, mostly glycine and branched aliphatics. Nothing about that molecule is built to be absorbed, and that is deliberate.

What it is antagonizing. Zonulin, the endogenous protein that reversibly disassembles tight junctions, was identified as pre-haptoglobin-2 — the uncleaved precursor of a common haptoglobin variant — and shown to act through protease-activated receptor 2 with transactivation of the EGF receptor Tripathi 2009. That is the pathway larazotide sits in front of, and it explains the whole shape of the drug's design: it is a luminal, competitive, local agent whose entire job is to occupy a site on the apical surface of the enterocyte before zonulin does.

The word “tightens” is wrong and the distinction matters. Larazotide does not close junctions that are open at baseline; the published claim is that it prevents the opening of intestinal epithelial tight junctions in vitro Leffler 2012. A drug that blocks an induced event does nothing when the event is not being induced. That single sentence predicts most of what the human trials found.

Then there is the finding nobody has explained. In the 342-patient phase 2b, 0.5 mg three times daily worked and 1 mg and 2 mg did not Leffler 2015. In the earlier gluten-challenge study the same shape appeared — symptom benefit at some lower doses but not at the higher dose Leffler 2012. A competitive antagonist is supposed to get better with concentration until it plateaus. An inverted dose-response in two independent trials is either an artifact repeated twice, or it is telling you something real about the peptide — self-association at higher luminal concentrations is the usual suspicion for a glycine-rich octapeptide — and no published experiment has distinguished those.

Cell, rodent, human — and where it stops

In cells. Larazotide prevents the opening of intestinal epithelial tight junctions in vitro Leffler 2012; the target it opposes, zonulin, was identified biochemically as pre-haptoglobin-2 and characterized in human and rodent intestinal tissue Tripathi 2009. So far, so clean.

In humans, with the permeability endpoint — and this is the trial people skip. Leffler 2012 randomized 86 diet-controlled celiac patients to larazotide 0.25, 1, 4 or 8 mg three times daily or placebo, with or without a 2.4 g/day gluten challenge, for 14 days. The primary outcome was the urinary lactulose/mannitol fractional excretion ratio — the direct measure of intestinal permeability, and the one the whole mechanism predicts. The result: the ratio was so variable in outpatients that the gluten challenge itself did not raise it significantly above the gluten-free control, and larazotide versus placebo was not significant either. Symptoms did separate — gluten reliably worsened them, and lower doses of larazotide limited that worsening while the higher dose did not.

In humans, with a symptom endpoint, at scale. Leffler 2015 enrolled 342 adults with celiac disease who had been on a gluten-free diet for 12 months or more and stayed on it: a 4-week placebo run-in, 12 weeks of treatment at 0.5, 1 or 2 mg three times daily, then a 4-week placebo run-out. The primary endpoint, the average on-treatment Celiac Disease Gastrointestinal Symptom Rating Scale, was met at 0.5 mg (ANCOVA p = 0.022; mixed model for repeated measures p = 0.005), with a 26% decrease in symptomatic days (p = 0.017), a 31% increase in improved symptom days (p = 0.034), and reductions in headache and tiredness (p = 0.010). The 1 and 2 mg doses were no different from placebo on any endpoint.

In humans, phase 3: stopped. The sponsor announced on 21 June 2022 that a pre-planned interim analysis did not support continuing the phase 3 celiac study (company announcement). That is the largest celiac drug program ever run, ending on futility.

The obstacle, and it is an endpoint problem rather than a mechanism problem. Every human trial of this drug measured symptoms in people already eating gluten-free. The one trial that tried to measure the mechanism found the assay too noisy to detect even the gluten effect it was validating against Leffler 2012. So there is no human study in which larazotide was shown to change intestinal permeability and to change how someone felt. And the community use — “leaky gut” in people without celiac disease — is a step further out again: a population with no diagnostic criterion, taking a drug whose only controlled evidence is in an antigen-specific autoimmune enteropathy that failed its confirmatory trial.

Larazotide pharmacokinetics — how much of it actually gets in

For once the oral barrier is the point of the drug, not the obstacle to it. Larazotide is meant to stay in the lumen. Its target sits on the apical face of the enterocyte, so systemic exposure would be a pure liability, and the design brief was an octapeptide that is not absorbed. Half-life in plasma is close to meaningless here; what matters is how long it survives in the gut and where it is when it does.

What degrades it. Everything the small intestine has. Pancreatic endopeptidases attack internal bonds; brush-border aminopeptidases chew from the N-terminus. Larazotide's own sequence is its only defense: glycine at both ends and a proline at position 7. Proline is the residue most peptidases cannot cut next to, which is exactly why the same trick shows up in GLP-1 analogs, and it is the structural reason an unprotected octapeptide survives the duodenum at all.

The arithmetic that makes the dosing look strange until you do it. The effective dose is 0.5 mg, three times daily, immediately before meals Leffler 2015. At 726 Da that is about 0.7 micromoles. Spread through roughly a liter of small-bowel fluid it gives a starting luminal concentration on the order of 0.7 micromolar, falling with every centimeter of transit as peptidases work on it. That is the number the whole drug rests on, and it is why timing relative to the meal is not a nicety — the peptide has to be at the epithelium at the same moment the gluten peptides are.

And it is why the dose-response is upside down rather than absent. Tripling the dose to 1.5 mg tripled the luminal concentration and produced less benefit, twice Leffler 2012 Leffler 2015. Whatever explains that — aggregation, a bell-shaped occupancy curve, faster proteolysis of a self-associated form — it is a pharmacokinetic question that has never been asked with the peptide measured in an intestinal aspirate.

What would have to be true, and how you would know it was not

1. Permeability should fall on drug, only under challenge, and only at the low dose. The mechanism says larazotide blocks an induced opening. So run a lactulose/mannitol urinary recovery ratio at baseline and at 4 weeks — but run it under a standardized gluten or stress challenge, because Leffler 2012 showed the unchallenged measurement is too noisy in outpatients to detect anything. Prediction: a fall at 0.5 mg three times daily and no fall at 2 mg. If it falls at both, the inverted dose-response is an artifact and two trials were wrong.

2. The prediction that cuts against it: nothing on your blood panel should change. If someone claims larazotide is healing a leaky gut, the cheap test is absorption. In a person with documented deficiency, draw ferritin, vitamin B12 and RBC folate plus a CBC before and after 12 weeks. Prediction: no change. No trial of this drug has ever shown mucosal healing, and the phase 3 stopped for futility (company announcement). A page that predicted improvement here would be predicting something the largest program in the field could not find.

3. Inflammation should not move either, and that is informative. hs-CRP at baseline and 12 weeks. Larazotide is a barrier agent, not an immunosuppressant, and in celiac patients already on a gluten-free diet the inflammatory driver has been removed by the diet. If hs-CRP does fall, the effect is not the advertised mechanism, and the obvious confounder — that people starting a drug also tighten up their diet — has to be excluded first.

4. The symptom read-out has a specific shape. Leffler 2015 measured symptomatic days, not symptom severity at a clinic visit. So count days, not intensity: a 26% fall in symptomatic days over 12 weeks is the published effect size, and it is small enough that a two-week trial of the compound cannot detect it in one person.

What nobody has tested yet

Nobody has run a trial with a working permeability endpoint and a symptom endpoint together. Leffler 2012 tried and the lactulose/mannitol assay defeated it in the outpatient setting; Leffler 2015 dropped the mechanism measurement and measured only symptoms. An inpatient or tightly supervised design — timed urine collections, controlled fluid intake, a standardized gluten challenge — would remove the variance that killed the first attempt. This is the single missing experiment that would tell you whether the mechanism this drug is named for operates in a person.

Nobody has explained the inverted dose-response, and it is measurable. Aspirate small-bowel fluid at 30, 60 and 120 minutes after 0.5 mg and after 2 mg, quantify intact peptide and its fragments by mass spectrometry, and look for concentration-dependent self-association. If the high dose produces less intact peptide at the epithelium, an odd clinical finding becomes an ordinary formulation problem — and a fixable one.

Nobody has tested it in the population that buys it. Every controlled trial was in diagnosed celiac disease. There is no trial in non-celiac intestinal permeability, and the honest reason is that the field cannot yet define who has it: the commercial serum “zonulin” assays do not clearly measure the pre-haptoglobin-2 that Tripathi 2009 identified. Fixing the assay comes before running the trial, and both are undone.

Larazotide — its own safety story, not its class's

The tolerability is genuinely unremarkable, and saying so is part of the honesty. In 342 patients over 12 weeks, safety was comparable with placebo Leffler 2015; in the gluten-challenge study there were no serious adverse events, and the commonest complaints were headache and urinary tract infection Leffler 2012. This is what you expect from a peptide that is not absorbed. The risks worth writing about here are not toxicological.

Risk one: substitution. Every trial required patients to stay on a strict gluten-free diet. Larazotide was tested as an add-on to the diet, never as a replacement for it, and no evidence exists that it protects against a deliberate gluten exposure. Using it to eat bread is not an off-label use of the evidence — it is a use the evidence says nothing about, in a disease whose untreated course includes villous atrophy and, over decades, lymphoma risk.

Risk two: the phase 3 result is the headline, not a footnote. An interim analysis in the largest celiac program ever run concluded that continuing was not supported (company announcement). Anyone selling this today is selling a compound that failed confirmatory testing in the one condition where its mechanism is best established. That is not a reason it cannot work; it is a reason no one should claim it does.

Risk three: you cannot check what you bought. An eight-residue peptide with three glycines and no aromatic residue at all is cheap to synthesize and hard to verify — there is no color, no smell, and no assay a buyer can run. A certificate of analysis reporting a mass near 726 Da is consistent with the right sequence and with several wrong ones, because mass does not distinguish an order of residues. For a compound taken three times a day before meals, indefinitely, that is the practical risk — not the pharmacology.

Sources read for this page

• Leffler DA. Larazotide acetate for persistent symptoms of celiac disease despite a gluten-free diet: a randomized controlled trial.. Gastroenterology 2015 · PMID 25683116
• Leffler DA. A randomized, double-blind study of larazotide acetate to prevent the activation of celiac disease during gluten challenge.. Am J Gastroenterol 2012 · PMID 22825365
• Tripathi A. Identification of human zonulin, a physiological modulator of tight junctions, as prehaptoglobin-2.. Proc Natl Acad Sci U S A 2009 · PMID 19805376
• 9 Meters Biopharma Inc. Interim Analysis of Phase 3 Study of Larazotide for Celiac Disease Does Not Support Trial Continuation.. Company announcement, 21 June 2022

Larazotide — safety, predicted from mechanism

Larazotide — interference & stacking

Bloodwork to run alongside Larazotide

Baseline first, then again at 8–12 weeks.

The Gut Health & Absorption panel covers these in one order — 11 markers, $208.80 with the discount applied.

Check results you already have → · All 103 markers A–Z

Larazotide — frequently asked questions

Larazotide inside a finished plan

One arm of 1 Protocol Blueprint, free to read in full.

What Larazotide is used for

Larazotide appears under 2 goals in the goal router.

Related Healing & Recovery compounds

← Explore the full Protocol Vault

↑ Back to on this page