Is BPC-157 Found in Human Gastric Juice? Nobody Has Shown It

No published evidence shows that BPC-157 occurs in human gastric juice. Every description of the peptide says it does, calling it a fragment of a protein found in the stomach, and that reported origin is where the name body protection compound comes from and why people reason that putting it back can only help.

The fragment is real and synthetic BPC-157 exists. What has never been published is the evidence that its parent protein was ever isolated, sequenced or found in a person.

What does the standard origin story say?

The standard account traces BPC-157 to a protein that a Zagreb laboratory found in human gastric juice in 1989, from which a smaller active piece was identified.¹ That piece is "just 15 amino acids long," and the peer-reviewed literature gives its sequence as GEPPPGKPADDAGLV, with a mass of about 1419 Da.¹²

A 2026 review in Pharmaceutics puts the claim the way almost everyone does, calling BPC-157 "a synthetic pentadecapeptide derived from a gastric protein fragment" and, in its introduction, "derived from a naturally occurring protein fragment present in human gastric juice."² The same review notes that the BPC fractions were "isolated and screened by Sikiric and colleagues at the University of Zagreb" in the early 1990s, and that BPC-157 "represents the synthetic counterpart of one such sequence."²

Read carefully, that is a claim about where the molecule came from and not a finding. The review does not say the parent protein was sequenced, because nothing published says so.

The same review is candid about the limits of its own method elsewhere, describing its recommendations as "derived primarily by analogy with approved peptide therapeutics" and its output as "intended as a developmental roadmap rather than a definitive pharmaceutical specification."² Candor about formulation sitting alongside an unexamined assertion about origin, in one document, is a fair picture of how the field holds the two questions.

What did the 1992 paper actually report?

The earliest BPC-157 entry in the literature is a 1992 study by Predrag Sikiric and 17 others, eighteen authors in all, reporting on "a new potent peptide" the team said it had isolated from human gastric juice and named Body Protection Compound.³

Its experiments were on rats, with a different organ removed from each animal to see which organ systems the peptide interacted with beyond the stomach.³ So the paper is an effects paper. It reports what the peptide did, resting on an isolation described as having happened previously.

That structure matters because every later paper inherits it. The origin is cited, not demonstrated, and the citation chain runs back to work whose own evidence was never put on the record.

Eighteen authors on a 1992 paper also deserves a moment. Eighteen names on a paper and no published isolation record for the step it rests on is an odd combination, whatever the reason for it. The paper reports downstream effects in rats competently enough to have launched three decades of research, and the upstream step it rests on stayed in-house.

Was the parent protein ever sequenced?

No published sequence exists for the parent protein. That full record was never released, and Anna Mapp, who reviewed the team's isolation patent at a reporter's request, called that unacceptable by today's standards.¹

Her specific objection goes further than the missing record. Of the patent describing the isolation, Mapp said: "There was no demonstration that I could tell that it was a single pure isolated substance."¹ The protein's molecular weight is also not pinned down in the record.¹

Without a sequence, the original work cannot be reproduced, and Mapp makes exactly that point about it.⁴ Reproducibility is not a technicality here. A protein nobody else can look for is a protein nobody else can confirm.

A patent is also a different kind of document from a paper. It exists to establish a claim with enough disclosure to support the claim, and not to let a stranger repeat the experiment. Resting an origin claim on patent material is therefore weak even when the patent is sound, and Mapp's objection is that this one does not demonstrate a pure isolated substance at all.

Who is making that assessment?

Anna Mapp is the President of the American Peptide Society, the main professional body for peptide science, and she reviewed the documents at the reporter's request.⁴ Her reaction to what she found was two words: "That is troubling," said Mapp.¹

The society then posted the investigation on its own news page under the title "The BPC-157 Question," crediting the original reporting.⁴ That page is a republication and not a society position statement, and it does not state an APS view on the evidence.⁴

So the assessment carries real weight and it is still one scientist's reading of a patent, offered to a journalist, not a peer-reviewed analysis or an institutional finding. Both of those things are true at once.

Treating it as a formal finding would overstate it, and dismissing it as a journalist's quote would understate it badly. The person who holds the presidency of the field's professional society looked at the founding document of a compound sold worldwide and could not find a demonstration that the starting material was a single pure substance. Few assessments arrive with that combination of standing and brevity.

Does the sequence appear in the human genome?

The sequence does not appear there, according to the two researchers who have commented on it. Mapp says the genetic material coding for the peptide's unique sequence "hasn't been found in the human genome or gut microbiome."¹

Jong-Hwei Su Pang, a professor at Chang Gung University whose group has run several BPC-157 studies, said by email that the peptide might be produced by stomach microbes, because its amino acid sequence does not appear in the human genome.³

I did not run a genome or proteome search myself, and nothing in the sources reports one being run and published. These are two named researchers' statements about an absence. Statements are weaker evidence than a published search, and these two point the same way.

Precision matters about what a published search would settle. The sequence GEPPPGKPADDAGLV is fifteen residues, short enough that a database query against the human proteome is a routine exercise taking minutes.² Nobody involved has published the result of that query, in either direction. For a claim this load-bearing and this cheap to test, the absence of a published test is itself informative.

Could gut bacteria be making it?

Pang has proposed that bacteria might, and Mapp's statement cuts against it. Those two sources conflict and both belong on the record.

Pang proposes stomach microbes as the source, reasoning from the sequence's absence in the human genome.³ Mapp reports that the coding material has not been found in the gut microbiome either.¹

Pang's is the weaker of the two claims. He offers microbial production as an explanation for an absence without reporting that anyone looked in the microbiome, while Mapp reports that the search in exactly that place came up empty. Pang's concession is still notable, because it comes from a researcher inside the field instead of a critic outside it: his own group has published on this peptide.³

The microbial hypothesis also changes the case it was meant to rescue. A peptide produced by gut bacteria is not something the human body makes, and the marketing claim depends on the second thing and not the first. Bacterial origin would make BPC-157 a microbial product that happens to appear in gastric contents. That is a different and far less reassuring story than an endogenous human protective peptide.

What does a co-author say about the sequence?

Sandor Szabo, a co-author on the work, has suggested the original team "misinterpreted an amino acid sequence" decades ago, when analytical techniques were less refined.¹ He offers it as speculation and not a finding.

A 2017 paper Szabo co-authored sets out what a naturally produced substance would normally have: an identified gene, cells that produce it, receptors, and known levels in the body.¹ None of those four are established for BPC-157.¹

Weigh that carefully. A co-author proposing that the founding observation was a misreading is the most serious thing in this story, and it is also an informal suggestion from someone revisiting work from decades earlier, not a retraction or a published reanalysis.

What does the current literature say?

Current peer-reviewed reviews repeat the origin claim and do not examine it. The 2026 Pharmaceutics review states the gastric-juice derivation twice, in the abstract and the introduction, and nowhere addresses whether the parent protein was sequenced or whether the peptide occurs in the human genome.²

That is not a criticism of those particular authors. A review summarizes a field, and this field has treated the origin as settled background since the early 1990s.

The practical effect is a claim that strengthens through repetition. Each review cites the last, the citation count grows, and the original evidence stays unpublished. Our guide to why most peptides have no human evidence covers the general version of a claim that outruns its sourcing.

That pattern repeats in a second place. On delivery, it reports that the BPC-157 literature is "restricted to patent disclosures," notes that "no peer-reviewed pharmaceutical development paper describing a characterised oral formulation of BPC-157 has been published," and cautions that transdermal-patch claims "should be treated with caution" because they "do not appear in peer-reviewed literature."²

So patents carry the origin and patents carry the delivery. For a compound with roughly 200 papers behind it,³ the two questions a buyer would most want answered both sit in documents written to establish ownership instead of to let anyone check the work.

How much of the research comes from one group?

More than 80% of the research comes from one group, by author affiliation. The Pharmaceutics review states that "the overwhelming majority of published BPC-157 studies" come from a single group and that they "originate from a single research group at the University of Zagreb," putting the share above 80%.²

Journalism on the same question reached a similar place by a different route. Of roughly 200 BPC-157 studies in the literature, most list Sikiric or his colleague Sven Seiwerth as a main author, according to a Polish review team's critique.³

Two independent counts landing in the same range is the strongest quantitative claim on this page. It does not say the work is wrong. It says the evidence base has one source, so an error at the origin does not get caught downstream.

What does Sikiric say?

Sikiric rejects the idea that the peptide is not natural, and he has declined to release the material that would settle it. He did not respond to repeated messages or requests for trial data from the reporter who pursued the February story, and no interview with him or a team member was obtained for it.³

For the June story he engaged, and declined to share the screening details, calling them proprietary.¹ Asked about the unpublished parent protein sequence, he compared the protein to a child he wants to keep as his own. That sentence is the article's paraphrase of his answer and not a direct quotation.¹

He has also published a dissenting comment with colleagues arguing that BPC-157 had been "unjustly accused of potentially harmful effects."³ His cited support for the natural-origin position includes Taiwanese work on indirect effects at two receptors and laboratory images of the peptide in human fetal tissue cells; a peptide scientist, Brubaker, said she did not find those images convincing.¹

Does any of this mean BPC-157 does not work?

No. Nothing about the origin problem says BPC-157 does not work, and conflating those two questions is the most common error in this argument. Where a molecule came from and what it does in a body are separate empirical questions, answered by different experiments.

The strongest version of the pro-BPC-157 position deserves stating plainly. A synthetic peptide can have real effects whether or not anything like it exists in a stomach, and whether this one does is answered by the efficacy experiments instead of by its provenance. If BPC-157 heals tissue in animals, where the molecule came from does not undo those results. Our guide to whether BPC-157 actually works covers the efficacy evidence on its own terms, and our page on BPC-157 for anxiety and depression does the same for that claim.

Damage from the origin problem lands on one specific claim and not on the compound. That claim runs: this is something your body already makes, so replacing it is safe and physiological. It needs the origin claim to be true. That claim is unsupported.

Why does the origin claim do commercial work?

The origin claim does commercial work because "your body already makes this" is the safety argument, and it costs nothing to assert. A naturally occurring molecule reads as physiological, and physiological reads as low risk.

Regulators have not accepted that framing. In 2023 the FDA placed BPC-157 on its list of substances that should not be compounded, citing potential significant safety risks, warning that the peptide could trigger an immune response, that products might contain impurities, and that safety information was incomplete.³ Our coverage of the HHS compounding policy and the stopgap before it tracks where that stands now.

Note the tension between the two claims. An immune response is exactly what a foreign molecule provokes and exactly what an endogenous peptide would not. Regulators flagging immunogenicity and marketers claiming endogeneity cannot both be describing the same molecule accurately.

Enforcement has not matched the listing, and that gap is part of why the claim keeps working. Few steps have been taken to discipline pharmacies still selling the peptide, an HHS spokesperson declined to comment on enforcement policy for it, and FDA commissioner Makary has said the agency will need to review peptides to determine which are harmless and which need more study.³ Two lawmakers have asked the agency to limit enforcement against certain peptides.³

That gap between a listing and an enforcement posture is where the origin claim earns its keep. A seller does not need regulators to agree that the compound is endogenous. A seller needs buyers to believe it. A quiet enforcement environment leaves the claim unchallenged at the point of sale.

What has not been measured?

Several things have not been measured, and the list is specific enough to be useful.

No published sequence exists for the parent protein, so no independent laboratory can look for it.¹ No published genome or microbiome search appears in these sources, only two researchers' statements that the coding material has not been found.³¹

No gene, producing cells, receptors or measured body levels have been established. Those four are what a 2017 paper by a co-author names as normal for a naturally produced substance.¹ No independent laboratory has confirmed that BPC-157 occurs in human gastric juice.¹

The screening work that selected this fragment out of the parent protein remains unreleased on proprietary grounds, so the step connecting the protein to the peptide is not on the record either.¹

Gaps extend past provenance into the pharmacy. No peer-reviewed paper describes a characterized oral formulation, aggregation under manufacturing stresses is uncharacterized, and forced-degradation studies have not been reported.² No dose-escalation factor for reaching equivalent systemic exposure by mouth has been reported either.²

Our comparison of BPC-157 and TB-500 and our guide to oral BPC-157 salt forms both treat the molecule as what it demonstrably is, a synthetic 15-residue peptide.

Common questions

Is BPC-157 found in the human body?

What is BPC-157's amino acid sequence?

Does BPC-157 appear in the human genome?

Who discovered BPC-157?

Was the parent protein ever sequenced?

What does "body protection compound" mean?

Has a co-author questioned the origin?

Does the origin problem mean BPC-157 is ineffective?

How much BPC-157 research comes from one group?

Why does the natural-origin claim matter for safety?

What has Sikiric said in response?

Is BPC-157 approved or legal to compound?

Could bacteria in the gut be producing BPC-157?

Is there any independent confirmation of the isolation?

What evidence does Sikiric cite for natural origin?

Does the peer-reviewed literature still call it natural?

Sources

1. Talpos S. The hidden history of BPC-157, a favorite MAHA peptide. Undark, co-published with STAT, 1 June 2026. Read at source, full text, no paywall encountered. Quoted for the 1989 protein and the identification of the active fragment, the 15-amino-acid length, Anna Mapp's review of the isolation patent and her quotations, the unpublished parent protein sequence, the unpinned molecular weight, Sandor Szabo's misinterpretation suggestion and his 2017 paper's four criteria, Sikiric's proprietary-screening position and his paraphrased answer about the sequence, the absence of independent confirmation, and Brubaker's assessment of the fetal tissue images. Szabo's 2017 paper is cited here through this article and was not read at source. ↩1 ↩2 ↩3 ↩4 ↩5 ↩6 ↩7 ↩8 ↩9 ↩10 ↩11 ↩12 ↩13 ↩14 ↩15 ↩16 ↩17 ↩18 ↩19
2. Mateescu DM, Gavrilescu DM, Constantinescu FE, Oancea C, Ilie AC, Folescu R, Popa MD, Iurciuc S, Muresan CO, Enache A. BPC-157 as an Investigational Peptide Therapeutic: Biopharmaceutical Challenges, Formulation Strategies, and Translational Development Barriers. Pharmaceutics 2026;18(5):625. DOI 10.3390/pharmaceutics18050625. Published 20 May 2026. Read at source. The page runs to 108,221 characters and the first 100,000 were read; the final 8,221 were not. Quoted for the sequence GEPPPGKPADDAGLV and its mass, the abstract and introduction wording on gastric-juice derivation, the attribution of the isolation and screening to Zagreb, and the single-group concentration above 80% by author affiliation. This review asserts the origin and does not address the parent protein's sequence, and that silence is cited above as a silence. ↩1 ↩2 ↩3 ↩4 ↩5 ↩6 ↩7 ↩8 ↩9 ↩10
3. Talpos S. From Croatia to MAHA: How an unapproved drug became the next hot peptide. Undark, co-published with STAT, 3 February 2026. Read at source, full text, no paywall encountered. Quoted for the 1992 study's authorship and content, Jong-Hwei Su Pang's emailed comment and institutional affiliation, the Polish review team's authorship count across roughly 200 studies, the FDA's 2023 compounding action and its stated reasons, Sikiric's non-response to the reporter and his published dissenting comment. This is journalism, not peer-reviewed literature, and the FDA action is reported here at second hand and was not read from the agency's own document. ↩1 ↩2 ↩3 ↩4 ↩5 ↩6 ↩7 ↩8 ↩9 ↩10 ↩11 ↩12 ↩13
4. American Peptide Society. The BPC-157 Question. APS News, 1 June 2026. Read at source. This page republishes and summarizes source 2 instead of reporting independently, so it is not corroboration. It is cited only for Anna Mapp's identification as President of the American Peptide Society, for Mapp's statement that the absence of a published parent protein sequence makes the original work impossible to reproduce, and for the fact that the society circulated the investigation on its own news page. The page states no society position on the evidence. ↩1 ↩2 ↩3 ↩4