A New Class of Ghrelin O-Acyltransferase Inhibitors Incorporating Triazole-Linked Lipid Mimetic Groups

Abstract

Inhibitors of ghrelin O-acyltransferase (GOAT) have untapped potential as therapeutics targeting obesity and diabetes. We report the first examples of GOAT inhibitors incorporating a triazole linkage as a biostable isosteric replacement for the ester bond in ghrelin and amide bonds in previously reported GOAT inhibitors. These triazole-containing inhibitors exhibit sub-micromolar inhibition of the human isoform of GOAT (hGOAT), and provide a foundation for rapid future chemical diversification and optimization of hGOAT inhibitors.

Introduction

The peptide hormone ghrelin is a promising and largely unexploited target for small-molecule therapeutics to treat obesity, diabetes, and other health conditions. Ghrelin, a 28-amino acid peptide discovered in 1999, is implicated in various physiological processes, most notably appetite stimulation, as well as regulation of insulin secretion and glucose homeostasis in rodents and humans. These findings suggest that inhibiting ghrelin activity could be a novel clinical strategy for treating obesity and type II diabetes, although such approaches remain to be fully evaluated.

Ghrelin’s biological activity depends on multiple covalent modifications, the most critical being a unique posttranslational octanoylation of the serine at position 3 of the ghrelin precursor. This modification is catalyzed by the integral membrane enzyme ghrelin O-acyltransferase (GOAT), identified in 2008. GOAT catalyzes acylation of both des-acyl proghrelin and des-acyl ghrelin. Only the acylated form of ghrelin can bind to its receptor GHSR-1a and activate signaling. Importantly, ghrelin is the only known substrate for GOAT in the human proteome, minimizing potential off-target effects from GOAT inhibition.

While ghrelin signaling has been discussed as a therapeutic target, the lack of potent and biostable inhibitors has hampered progress. Previous GOAT inhibitors have been based on mimics of ghrelin, small molecule library screening, or bisubstrate molecules combining features of both ghrelin and octanoyl-CoA. The peptide-based bisubstrate inhibitor GO-CoA-Tat, for example, effectively inhibited GOAT in cells and mice, improving glucose tolerance and reducing weight gain. However, current inhibitors are limited by low potency, poor stability, or other factors.

Design and Synthesis of Triazole-Linked Inhibitors

A major challenge in developing hGOAT inhibitors is the limited structural and mechanistic information available. Most reported hGOAT inhibitors incorporate an amide-linked lipid group to mimic the octanoyl moiety of acylated ghrelin. Structure–activity studies have shown the importance of the lipid group in product-mimetic inhibitors for potency.

To create novel hGOAT inhibitors, we replaced the amide bond linkage to the lipid chain in established inhibitors with a 1,2,3-triazole linkage. Triazoles are attractive as bioisosteres for amide bonds due to similarities in size, dipole moment, and hydrogen-bonding ability. Triazole incorporation can also enhance the biological half-life of peptidomimetic analogs by providing resistance to enzymatic hydrolysis.

Inspired by previous work on triazole-based inhibitors of other membrane-bound O-acyltransferases, we applied a synthetic strategy combining lipid mimetics and peptide fragments to generate bisubstrate inhibitors. The parent peptide structure is based on the N-terminal GSSFL pentapeptide from ghrelin, with the serine at position 3 replaced by an Alloc-protected diaminopropanoic acid (Dap). After solid-phase synthesis and orthogonal deprotection, the Dap side chain was transformed to an azide, which was then coupled to functionalized alkynes via copper-catalyzed Huisgen 1,3-dipolar cycloaddition (“click chemistry”) to yield pentapeptides with alkylated triazole side chains in place of the natural octanoyl ester or previously reported amide bonds.

Structure–Activity Relationship and Inhibitory Potency

A panel of inhibitors was synthesized with diversity in both the triazole-linked hydrophobic group and the amino acid side chain at the fourth position. Variations at position 4 were inspired by hGOAT’s broad substrate tolerance, and the triazole-linked side chain incorporated various aromatic groups to probe hGOAT’s preferences.

Initial screening using a fluorescently labeled hexapeptide hGOAT substrate (GSSFLC AcDanAcDan ) showed that the inhibitor with a phenylpropyl triazole group (8d) exhibited the greatest inhibition. Further structure–activity analysis revealed that the length of the alkyl chain connecting the triazole and phenyl rings was critical for potency. The phenylpropyl triazole inhibitor 8d had the highest potency with an IC₅₀ of 0.7 μM. Shortening or lengthening the alkyl chain reduced potency, consistent with previous findings on acyl chain length preferences in hGOAT inhibitors.

Significance and Future Directions

This work demonstrates that hGOAT tolerates a biostable triazole linkage in place of hydrolytically labile ester and amide linkages. The results also show that hGOAT inhibitors can incorporate hydrophobic groups beyond simple alkyl chains, broadening the chemical space for inhibitor design. These findings enable the rapid synthesis and optimization of new hGOAT inhibitors and provide opportunities to exploit the acyl chain binding pocket for increased potency.

As new classes of hGOAT inhibitors are identified and validated in vitro, future studies will progress to cell-based and animal models. Defining the optimal hGOAT inhibitor pharmacophore remains a key challenge for developing therapeutics targeting obesity, diabetes, and other conditions influenced by PF-06424439 ghrelin signaling.