FOLD №13 — Tesamorelin Gln-8 → Aib Helix Stabilization

Verdict: DISCARDED | Peptide: Tesamorelin | Class: Performance | Target: GHRHR (Q02643)

Mechanism of action (background)

Tesamorelin is a synthetic 44-residue analogue of human GHRH(1-44), modified at the N-terminus with a trans-3-hexenoyl group to protect against DPP-IV cleavage at the Tyr-1/Ala-2 bond. It is the only FDA-approved GHRH analogue, indicated for HIV-associated lipodystrophy, and acts by binding to the growth hormone-releasing hormone receptor (GHRHR, UniProt Q02643) on somatotroph cells of the anterior pituitary, stimulating pulsatile GH secretion. Downstream GH and IGF-1 signaling drives lipolysis, preferentially reducing visceral adipose tissue. Efficacy has been demonstrated across multiple Phase III trials and 52-week extension studies, including on modern integrase inhibitor-based ART (Russo et al. 2024), confirming the robustness of the GHRHR pathway as a pharmacological target.

The molecular basis of GHRHR engagement depends critically on the amphipathic α-helical conformation of residues approximately 1–13 of GHRH/tesamorelin. This N-terminal helix presents a hydrophobic face for receptor contact and a solvent-exposed face of polar residues. Disruption of helical geometry in this region attenuates receptor potency. The trans-3-hexenoyl cap provides both proteolytic protection and potential conformational stabilization at the N-terminus — but the mid-helix region (residues 4–13) remains flexible in solution and may adopt the bioactive conformation only upon receptor encounter, incurring an entropic binding penalty.

Modification hypothesis (what we tested)

This distillation tested whether substituting Gln-8 with α-aminoisobutyric acid (Aib) could pre-organize residues 1–13 of Tesamorelin into the bioactive α-helical conformation, reducing the entropic cost of GHRHR binding and potentially adding a second axis of proteolytic protection via conformational rigidity at a mid-helix site.

Modified sequence: YADAIFT**(Aib)**SYRKVLGQLSARKLLQDIMSRQQGESNQERGARARL

The rationale was multi-layered:

• Gln-8 sits on the solvent-exposed face of the N-terminal helix and is not known to make direct receptor contacts, making it a structurally permissive site for substitution.
• Aib's geminal dimethyl group on the α-carbon sterically enforces φ/ψ dihedral angles within the canonical α-helical region (approximately −60°/−45°), and its helix-inducing properties are extensively validated in peptide chemistry literature.
• Aib-type chemistry has direct precedent in GH-axis peptides: Fold №4 (Ipamorelin N-Me-Aib, REFINED, pLDDT 0.80) demonstrated that Aib-class substitutions are compatible with GH secretagogue receptor engagement and yield high-confidence predicted structures in a related peptide.
• This strategy is chemically and positionally distinct from the DPP-IV-blocking N-terminal approaches tested in Fold №2 (Sermorelin D-Ala-2, DISCARDED) and is orthogonal to the trans-3-hexenoyl cap already present in tesamorelin — targeting conformational rigidity rather than exopeptidase blockade.

The predicted outcome was pLDDT >0.85 across residues 4–13 with preserved overall fold geometry.

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

The structure prediction returned metrics that preclude any confident interpretation:

The pLDDT of 0.491 indicates substantial uncertainty in backbone placement throughout the 44-residue analogue — the model cannot confidently assign secondary structure elements, including the very N-terminal helix that is the subject of the hypothesis. The ipTM of 0.34 means the predicted peptide-receptor docking pose is not reliably resolved; the interface geometry cannot be trusted. The absence of Chai-1 agreement and Boltz-2 affinity values removes the secondary confidence checks that would ordinarily triangulate a weak primary signal.

Notably, this pLDDT value is nearly identical to Fold №2 (Sermorelin D-Ala-2, pLDDT 0.49, DISCARDED) — suggesting a systemic pattern rather than a modification-specific failure. Both folds involved non-proteogenic substitutions in longer GHRH-class peptides complexed with GHRHR. The contrast with Fold №4 (Ipamorelin, pLDDT 0.80, REFINED) — a shorter pentapeptide GHS targeting GHSR — is instructive: peptide length and template availability appear to substantially influence prediction reliability in this receptor class.

The heuristic sequence-based properties (aggregation propensity 0.193, stability score 0.203, long half-life estimate) suggest no gross sequence-level liabilities were introduced by the Aib substitution, but these cannot substitute for structural confidence.

Critical distinction: A pLDDT of 0.49 does not mean the modification is structurally disruptive. It means the model is uncertain — the two interpretations (modification causes structural disruption vs. model lacks template fidelity for this complex) are indistinguishable with the data in hand. The GHRHR lacks a published high-resolution co-crystal structure with tesamorelin or native GHRH, depriving the prediction model of the template quality needed to confidently resolve a 44-residue peptide at a class B GPCR interface.

What this tells us (negative results are data)

Fold №13 does not falsify the Gln-8 → Aib hypothesis. It establishes that this hypothesis cannot be evaluated with single-run structure prediction at the current state of the toolchain for this peptide-receptor system. That is a meaningful, if frustrating, result.

Specifically, this fold rules out the possibility of obtaining a high-confidence computational endorsement of the helix-preorganization strategy for tesamorelin using the current pipeline. It also narrows the design space: any computational approach to tesamorelin SAR at internal helix positions will require either (a) a high-resolution structural template for the tesamorelin–GHRHR complex, (b) ensemble prediction methods that can average over conformational uncertainty, or (c) fragment-based approaches that model the N-terminal helix independently before docking.

The key mechanistic trade-off — whether Aib's helix-inducing effect outweighs the loss of Gln-8's hydrogen-bonding capacity to GHRHR residues — remains entirely unresolved. The literature is silent on position-8 SAR in any GHRH analogue, meaning this is a genuine knowledge gap rather than a resolved question with a negative answer.

In the context of the lab's emerging pattern: Folds №2 and №13 both returned pLDDT ≈ 0.49 on GHRHR-targeting peptides ≥29 residues. Fold №4 (Ipamorelin, 5-mer, GHSR target) returned pLDDT 0.80. This pattern suggests the lab should either (1) focus computational GHRH work on shorter, truncated analogues (GHRH(1-17) or GHRH(1-29) fragments) where prediction confidence may be recoverable, or (2) escalate to ensemble/multi-model approaches before investing further single-run credits on full-length tesamorelin variants.

Alternative hypotheses to test (avoid the failure mode)

Computational approaches to recover signal:

• Truncated analogue: Test Aib-8 in a GHRH(1-17) or GHRH(1-29) framework — shorter peptides yield higher pLDDT in this receptor class, and SAR mapping in truncated analogues is standard practice in the GHRH literature.
• Ensemble prediction: Run 5–10 predictions with different random seeds and report mean/SD pLDDT and ipTM; ensemble agreement would provide a confidence interval on the structural uncertainty.
• Fragment modeling: Model the N-terminal helix (residues 1–13) in isolation to assess whether Aib-8 measurably increases helical propensity before attempting full complex docking.
• Alternative helix-inducing substitutions at position 8: α-methyl-phenylalanine (MePhe) or L-α-methyl-glutamine could probe whether steric reinforcement at the α-carbon is the operative variable while partially preserving side-chain character.

Wet-lab validation path (if hypothesis is to be advanced):

• Circular dichroism (CD) spectroscopy: Direct measurement of helical content in the Aib-8 analogue vs. native tesamorelin — the most direct test of the helix-preorganization hypothesis.
• GHRHR binding assay (competitive radioligand or TR-FRET): Measures affinity change independently of structural confidence; would resolve the Aib vs. Gln trade-off question computationally intractable here.
• Plasma stability assay: Incubation in human plasma with HPLC/MS quantification would determine whether mid-helix Aib confers proteolytic protection beyond what the trans-3-hexenoyl cap already achieves.
• NMR (solution structure): Would directly confirm or deny helical preorganization in the N-terminal segment — gold standard for the conformational hypothesis.