Lani J. Davies

Heparanase is the sole enzyme responsible for breaking down heparan sulfate within the extracellular matrix, and its overexpression is linked to human diseases. Despite heparanase being a promising drug target, most efforts have focused on substrate mimetics, which have failed clinical trials, highlighting the need for new inhibitor scaffolds. Here, we employed fragment-based drug design to explore a novel chemical space to develop small molecule inhibitors of heparanase. We used a crystallographic and computational approach to identify 31 fragments that bind heparanase; five of these inhibited heparanase in the micromolar range. One of these fragments underwent two cycles of fragment growing, which resulted in a compound with a 7-fold increased potency compared to the initial hit. The results from our fragment screen unveil untapped chemical space for heparanase inhibition, paving the way for the development of potent drug leads with the potential to transform the treatment of heparanase-related diseases.

Bismuth peptides and proteins are emerging as versatile tools for medicinal chemistry and chemical biology. Bismuth(III) binds three cysteine residues in peptides and proteins with remarkable selectivity. While the thermodynamic stability of these bismuth complexes is outstanding, their kinetic lability imposes limitations. Introducing bismuth selenopeptides, we demonstrate that selenocysteine binds bismuth with substantially higher kinetic stability than cysteine. This effect was quantified by directly comparing a peptide containing three cysteine residues with an identical peptide containing three selenocysteines. Bismuth selenopeptides are not only inert to strong chelators such as ethylenediaminetetraacetic acid (EDTA) and diethylenetriaminepentaacetic acid (DTPA) but also to the natural metal‐binding protein transferrin present in human plasma. We further demonstrate biological utility by developing bismuth selenopeptides that selectively bind and inhibit unrelated target proteins. To extend this concept to a more complex system, we investigated the human epidermal growth factor (EGF), a small protein comprising three disulfide bonds. We established that precisely two bismuth atoms bind to the six cysteine or selenocysteine residues in EGF and seleno‐EGF, respectively. In the presence of EDTA, bismuth seleno‐EGF remains fully intact, unlike its cysteine‐based analog, consistent with observations from smaller peptides. Structural models confirm full preservation of the native EGF fold.

Bismuth peptides and proteins are emerging as versatile tools for medicinal chemistry and chemical biology. Bismuth(III) binds three cysteine residues in peptides and proteins with remarkable selectivity. While the thermodynamic stability of these bismuth complexes is outstanding, their kinetic lability imposes limitations. Introducing bismuth selenopeptides, we demonstrate that selenocysteine binds bismuth with substantially higher kinetic stability than cysteine. This effect was quantified by directly comparing a peptide containing three cysteine residues with an identical peptide containing three selenocysteines. Bismuth selenopeptides are not only inert to strong chelators such as ethylenediaminetetraacetic acid (EDTA) and diethylenetriaminepentaacetic acid (DTPA) but also to the natural metal‐binding protein transferrin present in human plasma. We further demonstrate biological utility by developing bismuth selenopeptides that selectively bind and inhibit unrelated target proteins. To extend this concept to a more complex system, we investigated the human epidermal growth factor (EGF), a small protein comprising three disulfide bonds. We established that precisely two bismuth atoms bind to the six cysteine or selenocysteine residues in EGF and seleno‐EGF, respectively. In the presence of EDTA, bismuth seleno‐EGF remains fully intact, unlike its cysteine‐based analog, consistent with observations from smaller peptides. Structural models confirm full preservation of the native EGF fold.

Constrained peptides possess excellent properties for identifying lead compounds in drug discovery. While it has become increasingly straightforward to discover selective high‐affinity peptide ligands, especially through genetically encoded libraries, their stability and bioavailability remain significant challenges. By integrating macrocyclization chemistry with bismuth binding, we generated series of linear, cyclic, bicyclic, and tricyclic peptides with identical sequences. Utilizing bismuth to rigidify the peptide structure allows for a better comparison of different constraint levels, reducing confounding effects of interactions often seen with hydrophobic stapling reagents. Our study facilitated the identification of a peptide–bismuth tricycle that fully withstands cellular levels of glutathione, acts as a nanomolar protease inhibitor without being proteolytically digested by its target, and is fully stable in human plasma. Importantly, this multicyclic peptide does not possess any non‐canonical amino acid modifications. Using oxime ligation, we conjugated an analogue of this tricycle to the N‐terminus of two nanobodies to demonstrate potential applications in targeted therapy.

Targeted theranostics heavily rely on metal isotopes conjugated to antibodies. Single‐domain antibodies, known as nanobodies, are much smaller in size without compromising specificity and affinity. The conventional way of conjugating metals to nanobodies involves non‐specific modification of amino acid residues with bifunctional chelating agents. We demonstrate that mutagenesis of a single residue in a nanobody creates a triple cysteine motif that selectively binds bismuth which is, for example, used in targeted alpha therapy. Two mutations create a quadruple cysteine mutant specific for gallium and indium used in positron emission tomography and single‐photon emission computed tomography, respectively. Labelling is quantitative within a few minutes. The metal nanobodies maintain structural integrity and stability over weeks, resist competition from endogenous metal binders like glutathione, and retain functionality.

Targeted theranostics heavily rely on metal isotopes conjugated to antibodies. Single‐domain antibodies, known as nanobodies, are much smaller in size without compromising specificity and affinity. The conventional way of conjugating metals to nanobodies involves non‐specific modification of amino acid residues with bifunctional chelating agents. We demonstrate that mutagenesis of a single residue in a nanobody creates a triple cysteine motif that selectively binds bismuth which is, for example, used in targeted alpha therapy. Two mutations create a quadruple cysteine mutant specific for gallium and indium used in positron emission tomography and single‐photon emission computed tomography, respectively. Labelling is quantitative within a few minutes. The metal nanobodies maintain structural integrity and stability over weeks, resist competition from endogenous metal binders like glutathione, and retain functionality.

Targeted theranostics heavily rely on metal isotopes conjugated to antibodies. Single-domain antibodies, known as nanobodies, are much smaller in size without compromising specificity and affinity. The conventional way of conjugating metals to nanobodies involves non-specific modification of amino acid residues with bifunctional chelating agents. We demonstrate that mutagenesis of a single residue in a nanobody creates a triple cysteine motif that selectively binds bismuth which is, for example, used in targeted alpha therapy. Two mutations create a quadruple cysteine mutant specific for gallium and indium used in positron emission tomography and single-photon emission computed tomography, respectively. Labelling is quantitative within a few minutes. The metal nanobodies maintain structural integrity and stability over weeks, resist competition from endogenous metal binders like glutathione, and retain functionality.

The biocompatible synthesis of constrained peptides is challenging. Oxime ligation is a bioorthogonal technique frequently used for protein bioconjugation. We report a straightforward method to install N-terminal ketones and aminooxy side chains during standard solid-phase peptide synthesis. Cyclization occurs spontaneously after acidic cleavage or in aqueous buffer. We demonstrate the facile synthesis of protease inhibitors with varying conformational constraint. The most constrained peptide displayed an activity 2 orders of magnitude higher than its linear analog.