Biochemistry-first molecular monograph · Version 5.0 · Updated 23 August 2026
BPC-157
Repair-associated signalling with incomplete human translation
Across several experimental systems the peptide changes vascular, nitric-oxide, inflammatory and repair-associated responses rather than acting through one proven universal receptor.
This monograph examines the exact 15-amino-acid peptide BPC-157. It separates cell and animal mechanisms, animal pharmacokinetics, limited human administration and the large gap between experimental tissue effects and controlled patient outcomes.
This monograph begins with the molecule and follows the causal chain through biochemistry, cellular mechanism, exposure and integrated physiology. Population studies enter only at the terminal stress-test: useful for the proposition actually tested, incapable of defining the molecule or replacing the causal model.
Epistemological position
The causal chain carries the science.
Identity establishes what exists. Biochemistry establishes what can react. Pharmacology establishes whether the reaction is reachable. Physiology establishes what the system does with it. Statistical human observation comes last and adjudicates only the route, exposure, population and outcome actually studied.
A population average cannot identify a molecule, create a mechanism or decide an individual response. A demonstrated mechanism remains bounded by the causal step shown; a terminal study can confirm or refute its translation under specified conditions, never rewrite the underlying chemistry.
Molecular dossier
The entity before the outcome.
- ENTITY
- Synthetic 15-residue peptide with a defined primary sequence
- BIOCHEMICAL CLASS
- Gastric-derived repair-associated research peptide
- PRIMARY SYSTEM
- NO signalling · endothelial response · cytoprotection · tissue remodelling
- IDENTITY BOUNDARY
- BPC-157 is not gastric juice, a generic collagen peptide or a universal healing factor. Sequence, route, formulation and achievable tissue exposure must remain explicit.
Action ledger
Describe the action before judging its value.
Benefit is a downstream clinical judgement. The scientific object is the action itself: what moves, under which exposure, in which biological state, and where the inference stops.
Attributed biochemical action
Across several experimental systems the peptide changes vascular, nitric-oxide, inflammatory and repair-associated responses rather than acting through one proven universal receptor.
Exposure and reachability
Rat and dog studies describe short plasma half-life and species-dependent bioavailability. Human pharmacokinetic characterisation remains insufficient.
Physiological direction is contextual
Repair is a coordinated tissue process involving perfusion, inflammatory timing, matrix deposition, remodelling, mechanical loading and resolution. A change in nitric-oxide or angiogenic signalling may be useful in one phase and maladaptive in another. Structure, pain and function must therefore be separated rather than compressed into a generic healing claim.
Individual adjudication boundary
A validated proximal molecular target and human dose–exposure–target-engagement chain. Breadth across animal injury models does not create a pan-tissue human repair claim or establish the direction of angiogenesis in every biological context.
Causal map with terminal adjudication
The scientific hierarchy is explicit.
I established; II–III substantial but distributed across models and incompletely unified
Across several experimental systems the peptide changes vascular, nitric-oxide, inflammatory and repair-associated responses rather than acting through one proven universal receptor.
Class D/E: small uncontrolled human signals; no mature controlled efficacy programme
This descriptor applies only to the named route, population and outcome. It neither ranks nor creates the molecule’s biochemical science.
Translation boundary
A validated proximal molecular target and human dose–exposure–target-engagement chain.
What must not be inferred
Breadth across animal injury models does not create a pan-tissue human repair claim or establish the direction of angiogenesis in every biological context.
Historical and international corpus
Old research is not absent research.
BPC-157 belongs to a Croatian gastric-cytoprotection research programme developed over several decades, not to a century of human therapeutic use. Its historical depth is predominantly experimental and concentrated within a recurring research network.
The breadth of models is real information. It must be discounted neither to zero nor converted into human efficacy without authenticated material, human PK and target engagement.
Long experimental cytoprotection lineage
Early gastric-lesion work was followed by vascular, fistula, tendon and drug-injury models. Repetition over decades strengthens the existence of a context-sensitive biological programme, while concentration of authorship and models limits independent generalisation.[4][6][7][8]
Sparse human observations do not represent the whole corpus
Recent human reports add administration and safety signals but remain too small or uncontrolled to calculate efficacy. They are one layer appended to a much larger preclinical history, not the sole definition of the molecule.[1][3]
Mechanistic and biochemical calculus
Constraint, measurement and inference are not the same claim.
A biochemical proposition can be powerful without a clinical trial, but its premises, exposure and domain must remain explicit. Each statement below is labelled by the operation that supports it.
The exposure bridge is quantitatively weak
Prototype studies reported a parent-peptide half-life below 30 minutes and incomplete intramuscular bioavailability in rats and dogs. Fast disappearance does not prove inactivity, but it makes metabolites, local action and dosing interval decisive unknowns.[2]
A distributed network replaces a single receptor model
NO signalling, endothelial migration, FAK-paxillin, GH-receptor expression and inflammatory responses form a convergent network. Without a validated proximal target, causal direction must be tested node by node.[4][5][7]
Pleiotropy has no automatically positive sign
Angiogenesis, haemostasis and remodelling can support repair or worsen a different tissue state. The same measured action cannot be assigned one universal clinical value.[6][7][8]
Complete scientific web editionOpen the full causal analysis and referencesThe decision-grade summary remains above. The complete publication stays available here and as a downloadable PDF.
Causal architecture
From molecular event to organismal hypothesis.
The mechanism is a network, not a slogan
Published experiments connect BPC-157 with nitric-oxide biology, endothelial behaviour, angiogenic signalling, cytoprotection and inflammatory modulation. These observations form a mechanistic network, but no single universally validated receptor currently explains every reported phenotype.[2][4]
Vascular response can be causal and conditional
Endothelial migration and collateralisation can plausibly support repair in an ischaemic or disrupted tissue. The same remodelling language cannot be assumed favourable in proliferative disease, abnormal vasculature or active bleeding; biological context determines direction.[2][4]
Animal repair models resolve tissue questions
Tendon, ligament, gastrointestinal and drug-injury models test different causal environments. Repetition across tissues supports pleiotropic biology, yet it does not erase differences in mechanical loading, route, local concentration or species metabolism.[2][4]
Human translation begins with exposure
Without adequate human PK and a validated target-engagement marker, a nominal dose cannot be connected rigorously to the preclinical concentration-response programme. A subjective clinical signal can generate a question, but it cannot close this biochemical gap.[1][3]
What is established
✓ Defined 15-amino-acid identity
✓ Substantial preclinical programme
✓ Human administration reported
What remains unresolved
? Human PK
? Controlled efficacy
? Repeated-use and rare-event safety
Causal hierarchy
Seven evidentiary layers. No category substitution.
Levels I-V build the causal and translational model. Human observations and controlled clinical trials then test the complete proposition under the route, exposure, population and outcome actually studied. A trial is terminal adjudication, not the source of molecular truth.
IMolecular identity and chemistryWhat exact sequence, terminal chemistry, formulation and route exist?
IIBiochemistry and target architectureWhat reactions, binding events and signalling steps are physically possible?
IIICells and ex-vivo tissuesWhat changes in the relevant cell or isolated tissue system?
IVAnimal and in-vivo pharmacologyDoes exposure reach the mechanism in an integrated organism?
VIntegrated physiology and translationHow do tissue state, feedback, dose and individual terrain alter the response?
VIHuman PK and observational signalIs the entity present, tolerated and associated with a signal in humans?
VIIControlled clinical trialsDoes the specified protocol change a prespecified patient-relevant outcome?
Method applied to BPC-157
Where this molecule’s evidence can, and cannot, travel.
The complete corpus-wide attribution rules are published in the scientific methodology. This is their molecule-specific application.
Entity rule applied
BPC-157 is not gastric juice, a generic collagen peptide or a universal healing factor. Sequence, route, formulation and achievable tissue exposure must remain explicit.
Reachability rule applied
Rat and dog studies describe short plasma half-life and species-dependent bioavailability. Human pharmacokinetic characterisation remains insufficient. The decisive exposure boundary for BPC-157 therefore remains molecule-, formulation- and route-specific.
Causal-strength rule applied
Across several experimental systems the peptide changes vascular, nitric-oxide, inflammatory and repair-associated responses rather than acting through one proven universal receptor. This places the present causal depth at i established; ii–iii substantial but distributed across models and incompletely unified. That grade describes which links in the mechanism have been demonstrated; it is not a consumer rating and cannot be converted into a general promise of benefit.
Human-adjudication rule applied
Class D/E: small uncontrolled human signals; no mature controlled efficacy programme. This terminal descriptor answers only the population, route, exposure and outcome actually observed; it neither upgrades nor cancels a different link in BPC-157’s causal chain.
Translation rule applied
A validated proximal molecular target and human dose–exposure–target-engagement chain. Breadth across animal injury models does not create a pan-tissue human repair claim or establish the direction of angiogenesis in every biological context.
Falsification rule applied
The interpretation must change if any of its decisive propositions fail. For BPC-157, the predeclared tests are: Failure to reproduce a proximal endothelial or NO-related effect with authenticated material would weaken the proposed network. Human exposure below concentrations required for mechanistic activity would break the translation chain. No improvement in prespecified structure or function despite target engagement would falsify efficacy for that tissue and route. These are active boundaries on the conclusion, not research questions appended after a favourable narrative.
Foundational synthesis
Molecular identity and entity resolution
BPC-157 is a synthetic 15-amino-acid peptide derived from a sequence described in gastric-juice research. Identity, route and exposure must remain explicit.
Foundational synthesis
Biochemical mechanism and target architecture
Experimental literature reports gastrointestinal integrity, angiogenic signalling, nitric-oxide pathways and effects in tendon, ligament and drug-injury models. The breadth is hypothesis-rich but does not create a universal repair mechanism.
Foundational synthesis
In-vivo pharmacology and exposure
Rat and dog studies describe short plasma half-life and species-dependent bioavailability. Human pharmacokinetic characterisation remains insufficient.
Foundational synthesis
Integrated physiology and individual terrain
Repair is a coordinated tissue process involving perfusion, inflammatory timing, matrix deposition, remodelling, mechanical loading and resolution. A change in nitric-oxide or angiogenic signalling may be useful in one phase and maladaptive in another. Structure, pain and function must therefore be separated rather than compressed into a generic healing claim.
Foundational synthesis
Terminal human stress-test
A small retrospective knee-pain series reported subjective improvement without randomisation, blinding, standardised function or systematic imaging. A 2025 intravenous pilot in two previously exposed adults detected no clinical or laboratory safety signal; two participants cannot establish efficacy or safety. No adequate prospective controlled clinical trial has yet established BPC-157 efficacy for tendon, ligament, muscle, gastrointestinal injury or recovery outcomes.
Foundational synthesis
Counterevidence and causal boundaries
Human effect size, repeated-use tolerability and rare-event risk remain unresolved. Angiogenic and tissue-remodelling biology is not automatically favourable and requires context in proliferative disease, unexplained lesions and active bleeding.
Foundational synthesis
Scientific conclusion
BPC-157 has a substantive preclinical programme and a small human signal, but the identity-to-human-outcome chain is incomplete. The next meaningful evidence is prospective, route-specific, identity-verified and outcome-defined.
Extended analysis
Identity and biochemical frame
BPC-157 is the synthetic pentadecapeptide Gly-Glu-Pro-Pro-Pro-Gly-Lys-Pro-Ala-Asp-Asp-Ala-Gly-Leu-Val. The sequence is the first scientific gate: a commercial label, vial mass or generic term such as repair peptide cannot establish that the administered entity is the peptide studied in the literature.[2][4]
The frequently repeated description as a gastric-juice peptide should be interpreted carefully. The research entity is a synthetic 15-residue sequence derived from work on gastric protective compounds. Evidence for that synthetic sequence does not prove that an identical free peptide circulates physiologically at therapeutic exposure in humans.[4][8]
Extended analysis
Cell migration, adhesion and tendon biology
In tendon fibroblast and explant systems, BPC-157 increased cell survival, spreading and migration and increased phosphorylation of FAK and paxillin without changing total protein abundance. This is a coherent adhesion-and-migration mechanism relevant to repair biology, but it remains a cellular and ex-vivo finding.[4]
A separate tendon-fibroblast experiment reported increased growth-hormone-receptor expression and enhanced proliferative response when growth hormone was subsequently added. This suggests that BPC-157 can alter tissue responsiveness rather than act through one isolated pathway. It does not establish that systemic GH-axis manipulation improves a human tendon outcome.[5]
Extended analysis
Nitric-oxide, vascular and gastrointestinal mechanisms
Animal studies repeatedly interact BPC-157 with nitric-oxide biology, using L-NAME and L-arginine to probe vascular, haemostatic and gastrointestinal effects. In clopidogrel-induced gastric injury, BPC-157 modified VEGF-A/VEGFR1, AKT and p38/MAPK-associated findings while reducing mucosal damage in rats.[6]
The same literature contains effects on angiogenesis, bleeding, thrombosis and vascular adaptation that can appear directionally different across models. That breadth is not proof of a universal balancing function. It indicates a context-sensitive vascular programme whose target hierarchy and human dose-response remain unresolved.[6][7][8]
Extended analysis
Animal pharmacokinetics
Rat and beagle-dog studies found a prototype-peptide half-life below 30 minutes after intravenous or intramuscular administration. Intramuscular bioavailability was approximately 14%-19% in rats and 45%-51% in dogs, with urine and bile as major excretory routes.[2]
These data are stronger than an assumption of prolonged systemic exposure, but they remain species- and formulation-specific. They do not establish human subcutaneous, oral or intravenous pharmacokinetics, nor do they show which metabolites retain biological activity.[2]
Extended analysis
Human observations
Published human evidence is sparse. A retrospective knee-pain series reported subjective improvement after intra-articular administration without randomisation, blinding, standardised functional testing or systematic imaging. It establishes reported administration and a signal, not a causal effect size.[1]
A 2025 intravenous pilot described two previously exposed adults and did not identify acute clinical or laboratory safety abnormalities. Two participants cannot estimate efficacy, common adverse events, repeated-use effects or rare risk. The evidentiary value is descriptive and should remain labelled as such.[3]
Extended analysis
Counterevidence and translation boundaries
The main counterweight is not a large negative trial; it is the absence of adequate controlled human translation despite a broad preclinical programme. Animal success across many injury models can reflect shared experimental conditions, publication concentration and endpoints that are not equivalent to human pain, function or structural recovery.[1][2][3][4]
Angiogenesis and remodelling are not intrinsically favourable. Active bleeding, unexplained lesions, proliferative disease, infection and mechanically unstable injuries change the meaning of a repair signal. These contexts require explicit study rather than blanket reassurance or blanket prohibition.[6][7]
Evidence atlas
Every claim keeps its experimental address.
12 claim units are mapped to their evidentiary layer, experimental frame and source records. The layer describes what the evidence can answer; it is not a popularity score.
| Layer | Experimental frame | Finding and boundary | Sources |
|---|---|---|---|
| IMolecular identity | Identity and biochemical frame · claim 1 | BPC-157 is the synthetic pentadecapeptide Gly-Glu-Pro-Pro-Pro-Gly-Lys-Pro-Ala-Asp-Asp-Ala-Gly-Leu-Val. The sequence is the first scientific gate: a commercial label, vial mass or generic term such as repair peptide cannot establish that the administered entity is the peptide studied in the literature. | [2][4] |
| IMolecular identity | Identity and biochemical frame · claim 2 | The frequently repeated description as a gastric-juice peptide should be interpreted carefully. The research entity is a synthetic 15-residue sequence derived from work on gastric protective compounds. Evidence for that synthetic sequence does not prove that an identical free peptide circulates physiologically at therapeutic exposure in humans. | [4][8] |
| IIICell or ex-vivo system | Cell migration, adhesion and tendon biology · claim 1 | In tendon fibroblast and explant systems, BPC-157 increased cell survival, spreading and migration and increased phosphorylation of FAK and paxillin without changing total protein abundance. This is a coherent adhesion-and-migration mechanism relevant to repair biology, but it remains a cellular and ex-vivo finding. | [4] |
| IIICell or ex-vivo system | Cell migration, adhesion and tendon biology · claim 2 | A separate tendon-fibroblast experiment reported increased growth-hormone-receptor expression and enhanced proliferative response when growth hormone was subsequently added. This suggests that BPC-157 can alter tissue responsiveness rather than act through one isolated pathway. It does not establish that systemic GH-axis manipulation improves a human tendon outcome. | [5] |
| IVIntegrated in-vivo model | Nitric-oxide, vascular and gastrointestinal mechanisms · claim 1 | Animal studies repeatedly interact BPC-157 with nitric-oxide biology, using L-NAME and L-arginine to probe vascular, haemostatic and gastrointestinal effects. In clopidogrel-induced gastric injury, BPC-157 modified VEGF-A/VEGFR1, AKT and p38/MAPK-associated findings while reducing mucosal damage in rats. | [6] |
| VIHuman exposure or signal | Nitric-oxide, vascular and gastrointestinal mechanisms · claim 2 | The same literature contains effects on angiogenesis, bleeding, thrombosis and vascular adaptation that can appear directionally different across models. That breadth is not proof of a universal balancing function. It indicates a context-sensitive vascular programme whose target hierarchy and human dose-response remain unresolved. | [6][7][8] |
| IVIntegrated in-vivo model | Animal pharmacokinetics · claim 1 | Rat and beagle-dog studies found a prototype-peptide half-life below 30 minutes after intravenous or intramuscular administration. Intramuscular bioavailability was approximately 14%-19% in rats and 45%-51% in dogs, with urine and bile as major excretory routes. | [2] |
| IVIntegrated in-vivo model | Animal pharmacokinetics · claim 2 | These data are stronger than an assumption of prolonged systemic exposure, but they remain species- and formulation-specific. They do not establish human subcutaneous, oral or intravenous pharmacokinetics, nor do they show which metabolites retain biological activity. | [2] |
| VIHuman exposure or signal | Human observations · claim 1 | Published human evidence is sparse. A retrospective knee-pain series reported subjective improvement after intra-articular administration without randomisation, blinding, standardised functional testing or systematic imaging. It establishes reported administration and a signal, not a causal effect size. | [1] |
| VIHuman exposure or signal | Human observations · claim 2 | A 2025 intravenous pilot described two previously exposed adults and did not identify acute clinical or laboratory safety abnormalities. Two participants cannot estimate efficacy, common adverse events, repeated-use effects or rare risk. The evidentiary value is descriptive and should remain labelled as such. | [3] |
| VIIControlled human proposition | Counterevidence and translation boundaries · claim 1 | The main counterweight is not a large negative trial; it is the absence of adequate controlled human translation despite a broad preclinical programme. Animal success across many injury models can reflect shared experimental conditions, publication concentration and endpoints that are not equivalent to human pain, function or structural recovery. | [1][2][3][4] |
| IVIntegrated in-vivo model | Counterevidence and translation boundaries · claim 2 | Angiogenesis and remodelling are not intrinsically favourable. Active bleeding, unexplained lesions, proliferative disease, infection and mechanically unstable injuries change the meaning of a repair signal. These contexts require explicit study rather than blanket reassurance or blanket prohibition. | [6][7] |
Methods and coverage
A reproducible boundary, not an implied exhaustive review.
This is a curated mechanistic monograph built from primary records. It does not claim systematic-review completeness. Absence of a retrieved result is not evidence that an effect is absent.
- SEARCH SURFACE
- PubMed/MEDLINE primary records, Russian-language indexed abstracts, and reference-chaining for entity-defining, historical and translational studies. Language, geography, regulatory status and journal fashion are not evidence grades.
- QUERY FRAME
- BPC-157 · GEPPPGKPADDAGLV · FAK paxillin · nitric oxide · angiogenesis · tendon fibroblast · wound biology
- EVIDENCE CUTOFF
- 23 August 2026
- ENTITY EXCLUSION
- BPC-157 is not gastric juice, a generic collagen peptide or a universal healing factor. Sequence, route, formulation and achievable tissue exposure must remain explicit.
- TRACEABILITY
- 12 extended claim units plus 2 historical records and 3 mechanistic-calculus records; 8 cited records in the complete corpus.
- INTERPRETIVE RULE
- Identity → physical constraint → measured reaction or binding → reachable exposure → system response → outcome. Statistical significance does not manufacture mechanism; mechanistic plausibility does not manufacture a patient outcome.
Falsification map
What could make the proposition wrong?
A scientific monograph must state not only why a mechanism may hold, but what observation would weaken or overturn the inference.
- 01Failure to reproduce a proximal endothelial or NO-related effect with authenticated material would weaken the proposed network.
- 02Human exposure below concentrations required for mechanistic activity would break the translation chain.
- 03No improvement in prespecified structure or function despite target engagement would falsify efficacy for that tissue and route.
Research agenda
What would materially change the conclusion?
A useful research agenda resolves the weakest link in the identity-to-outcome chain. It does not merely add more observations of the same type.
- 01Human route-specific PK and metabolite mapping
- 02Prospective controlled tendon or ligament trials
- 03Imaging plus function plus pain as separate outcomes
- 04Repeated-dose safety and immunogenicity
- 05Mechanistic studies outside the dominant research groups
- 06Context-specific angiogenesis and haemostasis studies
Cited primary evidence corpus
8 source records used in this publication.
Reference numbers resolve directly to indexed records. This corpus documents the claims made here; it is not presented as an exhaustive systematic-review search. Index access does not upgrade the evidentiary class of a study.
- [1]Lee & Padgett. Altern Ther Health Med. 2021;27:8–13.PMID 34324435 ↗
- [2]He et al. Front Pharmacol. 2022;13:1026182.PMID 36588717 ↗
- [3]Lee & Burgess. Altern Ther Health Med. 2025;31:20–24.PMID 40131143 ↗
- [4]Chang et al. BPC-157 promotes tendon fibroblast outgrowth, survival and migration through FAK-paxillin. J Appl Physiol. 2011.PMID 21030672 ↗
- [5]Chang et al. BPC-157 enhances growth hormone receptor expression in tendon fibroblasts. Molecules. 2014.PMID 25415472 ↗
- [6]Wu et al. Clopidogrel-induced gastric injury is attenuated by BPC-157 in rats. Int J Mol Sci. 2020.PMID 33376304 ↗
- [7]Sikiric et al. BPC-157 and colocutaneous fistula healing in rats: role of the nitric-oxide system. J Pharmacol Sci. 2008.PMID 18818478 ↗
- [8]Sikiric et al. Gastroprotection by BPC-157 in experimental gastric lesions. Dig Dis Sci. 1996.PMID 8769287 ↗
Permanent citation
Cite this publication
Wibowo V. BPC-157: Biochemical, Mechanistic and Physiological Monograph. Clinic Terrain Clinical Intelligence Unit. Version 5.0. Updated 23 August 2026.
https://www.clinicterrain.com/research/peptides/bpc-157Publication and reference-verification method →Authorship and accountability
Valentine Wibowo, Clinical Researcher, Clinic Terrain Clinical Intelligence Unit, performed the conceptualisation, evidence synthesis, scientific interpretation and drafting. dr. I Wayan Cahyadi Surya Distira Putra, MPH completed medical review and approved the publication. Internal medical review is not represented as independent peer review.
No new patient data or human research were undertaken. The author and reviewer declare no personal financial competing interests. Clinic Terrain and Klinik Pratama Cepat Sehat provide clinical services; that institutional relationship is disclosed.