131 Review Received: 1 July 2008 Revised: 23 September 2008 Accepted: 24 September 2008 Published online in Wiley Interscience: 3 December 2008 (www.interscience.com) DOI 10.1002/psc.1082 Role of side chains in collagen triple helix stabilization and partner recognition ‡ Rita Berisio, a Alfonso De Simone, a,b Alessia Ruggiero, a Roberto Improta a and Luigi Vitagliano a* Collagen is a widespread protein family involved in a variety of biological processes. The complexity of collagen and its fibrous nature prevent detailed investigations on the full-length protein. Reductionist approaches conducted by dissecting the protein complexity through the use of model peptides have proved to be quite effective. There are, however, several issues regarding structure–stability relationships, aggregation in higher-order assemblies, and partner recognition that are still extensively investigated. In this review, we discuss the role that side chains play in triple helix stabilization and in partner recognition. On the basis of recent literature data, we show that collagen triple helix stability is the result of the interplay of different factors. As a general trend, interactions established by amino/imino acid side chains within the triple helix scaffold effectively modulate the intrinsic residue propensity for this common structural motif. The use of peptide models has also highlighted the role that side chains play in collagen self-association and in its interactions with receptors. Valuable examples in these fields are illustrated. Finally, future actions required to obtain more detailed information on the structure and the function of this complex protein are also delineated. Copyright c  2008 European Peptide Society and John Wiley & Sons, Ltd. Keywords: collagen; iminoacids; triple helix; protein stability; molecular recognition Background Collagen is the most abundant protein in vertebrates as it is an essential structural component of all connective tissues such as cartilage, bones, ligaments, and skin. In these species, it accounts for one-third of the total protein weight. Collagen-like molecules are also widespread in other organisms. Indeed, they have been found in lower eukaryotic, prokaryotic, and viral genomes [1,2]. The unique shape and properties of collagen molecules are due to their peculiar amino acid composition and sequence. Collagen sequences are characterized by the repetition of triplets of the type Gly-Xaa-Yaa. Although all types of amino acids may be located at positions X and Y of the triplets, they are frequently occu- pied by iminoacids [Pro and its post-translationally modified form 4R-hydroxy-2S-proline (4RHyp)]. In vertebrate collagens, as a con- sequence of the specificity of the prolyl-hydroxylation process, Pro and 4RHyp are almost exclusively located at the X and Y positions, respectively. From the structural point of view, collagen single molecules are composed of three polypeptide chains in polypro- line II (PPII) conformation, each containing hundreds of amino acid triplets, wrapped around a common axis (triple-helix motif). The complexity of the collagen molecule and its fibrous nature prevent detailed investigations on the full-length protein. To overcome this limitation, a number of diversified strategies have been adopted. Because of the repetitive nature of collagen sequence/structure, the use of peptide models embedding specific motifs has been successful [3 – 8]. Through this approach important milestones have been achieved. Structural characterizations of collagen-like peptide models have provided an atomic resolution picture of collagen triple helix [9–28] (Table 1). Crystallographic analyses have also shown the relationship between sequence and global/local triple helix structure. These investigations have paved the way for subsequent theoretical investigations. In this framework, X-ray models were used as starting structures in several MD simulations [29 – 34]. Furthermore, quantum mechanics studies have been performed to address specific unsolved issue related to collagen structure and stability [29,35 – 38]. In addition, the design and the characterization of host–guest peptide models embedding naturally encoded amino acids have led to the generation of a detailed propensity scale for collagen triple helix [4,39–41]. These data have been of fundamental importance for unveiling sequence–stability and structure–stability relationships. Finally, through the use of the model peptide approach, new compounds with a triple helical scaffold endowed with special properties (themostability, inhibition of metalloproteases, and ability to mimic and modulate collagen interactions with its partners) have been developed [42–44]. These peptides have shown an important potential for applications in diversified fields (medicine, biology, and bioengineering). Although investigations on peptide models have generated a wealth of information on collagen structure, stability, and function, there are several aspects of these fundamental issues that are highly debated. Over the years, several excellent reviews have been reported on the use of triple-helical peptides for elucidating collagen structure and function [3,5,6,8,40,45 – 49]. Here we illustrate, on the basis of ∗ Correspondence to: Luigi Vitagliano, Istituto di Biostrutture e Bioimmagini, CNR via Mezzocannone 16, I-80134 Napoli, Italy. E-mail: luigi.vitagliano@unina.it a Istituto di Biostrutture e Bioimmagini, CNR via Mezzocannone 16, I-80134 Napoli, Italy b Department of Chemistry, University of Cambridge, Lensfield Road, CB2 1EW Cambridge, UK ‡ 11th Naples Workshop on Bioactive Peptides. J. Pept. Sci. 2009; 15: 131–140 Copyright c  2008 European Peptide Society and John Wiley & Sons, Ltd.