FOLD №32 — BPC-157 C-Terminal Fragment DDAGLV for VEGFR2 Engagement

Verdict: PROMISING | Peptide class: Regenerative | Target: VEGFR2 (KDR, P35968)

Mechanism of Action

BPC-157 (GEPPPGKPADDAGLV) is a 15-residue synthetic peptide derived from human gastric juice with broad preclinical tissue regenerative activity. Its best-characterized pro-angiogenic mechanism runs through VEGFR2: Hsieh et al. (2017, PMID:27847966) demonstrated that BPC-157 increases VEGFR2 mRNA and protein expression in human vascular endothelial cells, promotes dynamin-dependent receptor internalization, and activates the VEGFR2–Akt–eNOS signaling axis — without upregulating VEGF-A itself. This pattern is consistent with a ligand-mimicry model in which BPC-157 engages VEGFR2 directly, bypassing the endogenous growth factor. Downstream, VEGFR2–Akt–eNOS signaling promotes endothelial proliferation, migration, and nitric oxide-mediated vasodilation — mechanisms central to angiogenesis and tissue repair.

A competing mechanistic model (Schlosser preprint, 2025, unreviewed) proposes that the N-terminal GEPPPGKPA segment adopts a polyproline II helix engaging SH3 domains of Src family kinases, with VEGFR2 activation occurring downstream via receptor transactivation rather than direct extracellular ligand engagement. This model is computationally derived and unvalidated by functional assays, but it is mechanistically coherent and must be held as a live alternative hypothesis.

Modification Rationale

FOLD №32 tests whether the C-terminal hexapeptide DDAGLV (residues 10–15 of BPC-157) constitutes a minimal pharmacophore sufficient for VEGFR2 engagement. The rationale has three elements:

1. Acidic residue pharmacophore hypothesis: The tandem aspartates D10–D11 are the only acidic residues in the full 15-amino acid sequence and are concentrated entirely in the C-terminal DDAGLV fragment. VEGF-A contacts VEGFR2 primarily through β-hairpin loops bearing acidic patches; D10–D11 may mimic this chemistry at the Ig-domain interface.

2. N-terminal scaffold hypothesis: The proline-rich GEPPPGKPA segment (residues 1–9) is conformationally flexible and may function as a non-pharmacophoric spacer rather than a receptor-contact element. Removing it should, on this hypothesis, expose the active motif rather than destroy it.

3. Fragment efficiency logic: Isolating a minimal pharmacophore reduces molecular weight, potentially improves target selectivity, and creates a scaffold for subsequent optimization (cyclization, stapling, non-natural amino acid incorporation).

This approach diverges from folds #29–31 (Aib substitutions and a stapled peptide on other peptides) by introducing a fragment/truncation strategy with an affinity focus — the first such modification in this lab's BPC-157 series.

Predicted Properties (Where Signal Is Moderate)

The ipTM of 0.649 is the primary positive signal in this run. For a six-residue fragment docked to a large receptor, an ipTM approaching 0.65 indicates the interface geometry is non-random and has internal consistency across the prediction. The extended backbone conformation — with tandem aspartates appearing oriented toward the receptor surface and the GLV face forming a compact hydrophobic patch — is geometrically consistent with the ligand-mimicry hypothesis. This is not a confident binding prediction; it is a structurally plausible docked pose that warrants follow-up.

What Would Strengthen This Signal

Computational next steps:

1. Ensemble prediction: Run DDAGLV vs. VEGFR2 across multiple seeds and sampling temperatures in AlphaFold3/Chai-1 to assess whether the extended conformation and ipTM ~0.65 are reproducible or run-specific. Single-run predictions at this confidence level cannot be interpreted without ensemble context.

2. Counter-fragment fold: Predict the complementary N-terminal fragment GEPPPGKPA vs. VEGFR2 under identical conditions. If ipTM for the N-terminal fragment exceeds that of DDAGLV, the Schlosser SH3 model gains computational support and this truncation hypothesis is challenged. If GEPPPGKPA shows lower ipTM against VEGFR2 (but perhaps higher against a Src-SH3 domain), the pharmacophore assignment becomes cleaner.

3. Constrained DDAGLV variants: A lactam-bridged or stapled DDAGLV — analogous to the i,i+3 lactam bridge used in TB-500 fold #28 (REFINED, pLDDT 0.81) — could pre-organize the extended conformation and potentially improve both pLDDT and ipTM. Fold #28 demonstrated clearly that macrocyclic constraints on short regenerative peptides can substantially boost structural confidence.

4. Alanine scan predictions: Predict D1A-DAGLV and DD-A-GLV variants to assess whether each aspartate independently contributes to the docked pose geometry — a computational surrogate for classical alanine scanning.

Experimental validation (wet lab):

5. Direct binding assay: Surface plasmon resonance (SPR) or fluorescence polarization with synthesized DDAGLV and the VEGFR2 extracellular domain (Ig-domains 1–3) would definitively test whether this fragment binds. Given the complete absence of any published BPC-157 binding affinity data, even a negative result here would be a significant contribution to the field.

6. Cellular activity assay: HUVEC tube formation and VEGFR2 phosphorylation (pY1175) assays comparing DDAGLV, full-length BPC-157, and the N-terminal GEPPPGKPA fragment would functionally map which segment drives downstream signaling.

7. Protease stability profiling: Given the predicted short half-life (~15–45 min), a plasma stability assay on synthetic DDAGLV should precede any cell-based work to determine whether the fragment survives long enough to be biologically interpretable.

Cross-Fold Connections

This fold sits within the lab's broader regenerative peptide programme alongside the TB-500 series (folds #7, #16, #28). The contrast with fold #28 (TB-500 lactam bridge, REFINED, pLDDT 0.81) is instructive: that work demonstrated that macrocyclic pre-organization of a short regenerative heptapeptide can push structural confidence into the high-reliability range. DDAGLV achieves moderate docking confidence without any backbone constraint — which is either a genuine positive signal or an artefact of a permissive run. A constrained DDAGLV variant would test this directly and represents the most logical immediate next fold in this series.

Fold #16 (TB-500 Lys→Orn, DISCARDED) serves as a useful cautionary comparison: a single conservative substitution that seemed unlikely to disturb function nonetheless produced a biologically uninformative prediction. Fragment truncation is a far more aggressive modification than single-residue substitution, and the possibility that DDAGLV is simply a biologically inert fragment of a larger active scaffold cannot be excluded on the current data.