β-hairpin motif-inspired binding peptides for biomacromolecules

Peptides are versatile molecules that combine favourable properties of small molecules and proteins, such as low immunogenicity, biocompatibility, low production cost, and structural flexibility. These features have led to their widespread application in biosensing, antimicrobial therapies, and the inhibition of protein-protein interactions. Many functional peptides are derived from proteins to mimic their specific biological functions, often by emulating short structural motifs involved in molecular recognition. This thesis presents two independent projects based on such peptide mimics: the development of silencing mediator of retinoid and thyroid hormone receptors (SMRT)-derived peptides for stabilising the biologically relevant conformation of cHDAC4, and the design of cellulose-binding peptides inspired by carbohydrate-binding modules (CBMs). In both cases, peptide design was inspired by β-hairpin motifs found in the respective precursor proteins. Histone deacetylases (HDACs) are key regulators of gene expression that promote transcriptional repression by removing acetyl groups from histone lysine residues, thereby restoring the positive charge of histones and facilitating chromatin condensation. The primary biological function of HDAC4, and other class IIa histone deacetylases, lies in the recruitment of the enzymatically active SMRT/NCoR–HDAC3 complex rather than in their own catalytic activity, which is strongly reduced as a result of a conserved substitution of a catalytically essential tyrosine residue by histidine. This interaction occurs via the structural zinc-binding domain (sZBD) of HDAC4, a structural element unique to class IIa HDACs, where SMRT binds at the rim of the active site. The aim of this project was to stabilise and characterise the biologically relevant conformation of SMRT-bound cHDAC4 wild type (WT) using peptides derived from the SMRT class IIa binding region. Initial designs focused on SMRT motif 1 (SM1), a short nine-residue sequence (GSITQGIPR) identified by Park et al.1 as the minimal HDAC4- binding motif. Linear peptides based on SM1 showed binding and stabilisation properties comparable to longer SMRT-derived sequences. To further improve stabilisation and affinity, cyclic peptides were designed to preorganise the structure into a β-hairpin-like conformation, inspired by the fold adopted by SMRT peptide 1 (SP1), a peptide containing the SM1 sequence, which was co-crystallised with the HDAC4 gain-of-function (GOF) variant, an H976Y variant that restores the catalytic activity of HDAC4. Side-chain to tail cyclisation of the core GSI sequence, by linking the side chain of a glutamic acid to the N-terminus and including one additional N-terminal amino acid, yielded peptides with improved inhibition and stabilisation effects. Binding to the rim of the active site was confirmed through tracer displacement experiments and thermal shift analyses of HDAC4 alanine variants. Hydrogen-deuterium exchange mass spectrometry (HDX-MS) was used to compare conformational flexibility between HDAC4 WT and GOF, confirming that the GOF variant exhibited lower deuterium uptake, which - consistent with its higher thermal stability - indicates a more rigid structure. HDX-MS analysis of peptide–cHDAC4 WT complexes demonstrated that side chain variations at the peptide N-terminus influence local protein dynamics, with peptides carrying smaller side chains inducing reduced deuterium uptake (suggestive of local masking effects), while peptides with bulkier side chains led to increased uptake, possibly reflecting local structural rearrangements to accommodate the side-chain. The peptide cV-SM1-E displayed contextdependent effects, with reduced deuterium uptake in WT and increased uptake in GOF complexes, suggesting its potential as a probe for conformational differences between HDAC4 variants. The peptides characterised in this study may serve as starting points for the development of HDAC4 inhibitors that target the structural zinc-binding domain and adjacent rim regions of the active site, potentially interfering with SMRT/NCoR–HDAC3 complex formation. The second part of this thesis focused on developing short cellulose-binding peptides as synthetic alternatives to carbohydrate-binding modules (CBMs). CBMs are protein domains commonly found in cellulolytic enzymes, where they mediate specific attachment to cellulose, and are widely applied in biotechnology for cellulose recognition and material functionalisation. However, CBMs such as Cel7A-CBM1 depend on defined folding and disulphide bridges, which limit their flexibility for chemical modifications. To overcome this, fluorescently labelled short peptides were designed based on the β-hairpin binding region of Cel7A-CBM1 and evaluated for their ability to bind nanocrystalline cellulose (NCC) and cotton linter paper, which contains both crystalline and amorphous cellulose regions. Peptide-cellulose interactions were characterised using fluorescence anisotropy to determine dissociation constants toward NCC, CD spectroscopy to assess peptide folding, and fluorescence microscopy to visualise adsorption on cotton linter paper. To investigate the influence of sequence features on binding, variants were synthesised with different fluorophore positions, tyrosine-to-alanine substitutions, and a scrambled peptide control. The tested peptides displayed affinities towards NCC in the high nanomolar range. The only non-binding peptide displayed a distinctly folded structure in solution, which altered upon interaction with NCC, while all other peptides remained unstructured random coils. All peptides readily adsorbed to cotton linter paper even in the presence of pre-bound Cel7A-CBM1, suggesting that they interact with alternative or partially accessible sites between protein molecules. These results demonstrate that cellulose binding by short peptides does not require defined secondary structure, tyrosine residues, or specific sequence motifs. Instead, unstructured, flexible peptides appear sufficient for effective interaction. Further studies should explore how the fluorophore itself or the peptide backbone interactions contribute to binding affinity, with the aim of developing short, synthetically accessible peptides for cellulose-based applications such as biosensing, surface functionalisation, or targeted protein immobilisation, where their smaller size and potential for chemical modification may offer advantages over CBMs. Although the biological targets and applications of the two projects differ, both highlight the potential of short peptides derived from β-hairpin motifs to mimic protein interactions. In the HDAC4 project, structural preorganisation through cyclisation enhanced inhibition and thermal stabilisation properties, reflecting the importance of defined folding for binding to a structured protein surface. In contrast, the cellulose-binding peptides remained unstructured yet retained affinity, suggesting that conformational flexibility can be advantageous for interactions with extended, less-defined targets. These findings emphasise that the relevance of secondary structure motifs in peptide design is strongly influenced by the nature of the target.