DISTILLATION №21 — DISCARDED

Epitalon C-terminal amidation (AEDG-NH2)

Verdict: DISCARDED | pLDDT 0.34 | ipTM 0.21 | Peptide: AEDG-NH2 | Target: hTERT (O14746)

Mechanism of action (background)

Epitalon (AEDG) is a 4-residue synthetic tetrapeptide with ~25 years of published research spanning telomerase activation, histone binding, antioxidant defense, and geroprotection. Its best-supported mechanistic proposal positions AEDG as a histone H1 ligand at DNA-interacting basic domains (PMID:32019204), with downstream epigenetic upregulation of hTERT mRNA as the primary route to telomerase activation — rather than direct binding to the TERT catalytic domain. Functional effects include dose-dependent telomere length extension in human epithelial and fibroblast cells (PMID:40908429), improved bovine oocyte maturation linked to nuclear telomerase relocalization (PMID:39788414), and ROS reduction with spindle protection in post-ovulatory aging oocytes (PMID:35413689). The native peptide's plasma half-life has not been formally quantified in the peer-reviewed literature, but reliance on intranasal delivery and millimolar in vitro concentrations indirectly suggests significant metabolic lability.

Modification hypothesis (what we tested)

This fold tested C-terminal amidation of Gly-4 — conversion of the free carboxylate (-COOH) to a primary amide (-CONH₂) — yielding AEDG-NH₂. The pharmacological rationale is established: carboxypeptidase A and B family enzymes recognise and cleave C-terminal free acids; amidation removes the electrostatic recognition element and is isosteric with the carboxylate, imposing minimal steric cost. The modification also introduces a hydrogen-bond donor that could enhance interactions with carbonyl-accepting residues at the putative histone H1 binding site.

This fold was designed as the logical complement to Fold #6, which tested N-terminal D-Ala substitution and collapsed to pLDDT 0.34 — attributed to poor tolerance of backbone chirality changes in a marginally compact 4-mer. C-terminal amidation preserves all backbone stereochemistry while targeting the opposite exoprotease degradation pathway, representing the minimal-perturbation strategy indicated by Fold #6's failure mode.

Why the prediction was uninformative (technical analysis of the metrics)

The scores are statistically indistinguishable from Fold #6. The critical insight is that both failures returned identical pLDDT values (0.34) despite completely different modifications — one a backbone chirality change, one a terminal charge-state change. This convergence strongly suggests the scores are not reflecting the quality of the modification at all: they reflect the hard floor imposed by sequence length. AlphaFold-class tools are trained predominantly on structured proteins and longer peptides; a 4-residue, intrinsically disordered sequence provides insufficient information for confident fold prediction regardless of terminal chemistry.

Additionally, ipTM of 0.21 means the predicted interface geometry is essentially random — no specific contact pattern with the TERT surface can be extracted. The Boltz-2 affinity module requires a sufficiently confident structural template to operate and returned no values. The heuristic sequence-based profile (stability 0.504, half-life 15–45 min, aggregation propensity 0.0) reflects properties of the amino acid composition and is not sensitive to the terminal amide change.

The prediction failed to evaluate the modification — it failed to model the peptide.

What this tells us (negative results are data)

Fold #21, combined with Fold #6, now establishes a clear empirical boundary: current AF2-class structure prediction tools operating against TERT cannot return actionable metrics for 4-residue Epitalon analogs, regardless of modification type. This is a meaningful negative result for the lab's computational pipeline rather than for the chemistry.

What this does not tell us:

• Whether AEDG-NH₂ retains hTERT-upregulating activity
• Whether C-terminal amidation improves plasma stability of AEDG
• Whether the Gly-4 carboxylate is a pharmacophore element required for histone H1 engagement

What this does tell us:

• The Epitalon scaffold sits below the resolution floor for structure prediction confidence at this target
• N-terminal and C-terminal terminus-only protection strategies have been identified as the next logical wet-lab experiments, not further computational folds of the same 4-mer
• The pharmacological risk of removing the C-terminal negative charge (potentially disrupting electrostatic engagement with histone H1's Lys/Arg-rich domain) cannot be evaluated computationally and must be addressed by cell-based SAR
• Comparison with SS-31 variants (pLDDT 0.85, Folds #11 and #17) and MOTS-c K13R (pLDDT 0.63, Fold #19) confirms that peptide length and intrinsic structuredness — not modification quality — are the dominant determinants of prediction confidence in this lab

Alternative hypotheses to test (avoid the failure mode)

Computational alternatives:

1. Molecular dynamics simulation — explicit-solvent MD (e.g., GROMACS, AMBER) of AEDG and AEDG-NH₂ in free solution or docked against a histone H1 homology model would capture the conformational ensemble of a disordered tetrapeptide without requiring a confident fold prediction as input
2. Coarse-grained docking against histone H1 basic domain — the Khavinson 2020 model (PMID:32019204) provides a putative binding site that could be used for blind docking of AEDG vs. AEDG-NH₂ to assess whether carboxylate removal changes binding geometry
3. Longer constructs — if the scientific question is TERT engagement, a fusion of AEDG or AEDG-NH₂ to a structured carrier peptide or scaffold might exceed the pLDDT floor and enable meaningful prediction

Wet-lab alternatives (highest priority):

1. Plasma stability assay — incubate AEDG and AEDG-NH₂ with human plasma (or purified carboxypeptidase B) and quantify intact peptide by HPLC/LC-MS at 0, 15, 30, 60, 120 min. This directly tests the amidation hypothesis and doesn't require structural prediction
2. hTERT reporter assay — compare AEDG vs. AEDG-NH₂ in a cell-based hTERT-luciferase reporter at matched concentrations (0.01–1 mM range) to determine whether amidation preserves, enhances, or abolishes transcriptional activity
3. Histone H1 binding assay — SPR or ITC titration of AEDG vs. AEDG-NH₂ against recombinant histone H1 to directly test whether the C-terminal carboxylate contributes to binding affinity
4. N+C dual protection — if both termini are independently modulable, a head-to-head comparison of AEDG-NH₂ (Fold #21), Ac-AEDG (N-terminal acetylation, not yet tested), and Ac-AEDG-NH₂ (dual protection) in a stability assay would establish the dominant degradation pathway empirically