• 2018-07
  • 2019-04
  • 2019-05
  • 2019-06
  • 2019-07
  • 2019-08
  • 2019-09
  • 2019-10
  • 2019-11
  • 2019-12
  • 2020-01
  • 2020-02
  • 2020-03
  • 2020-04
  • 2020-05
  • 2020-06
  • 2020-07
  • 2020-08
  • 2020-09
  • 2020-10
  • 2020-11
  • 2020-12
  • 2021-01
  • 2021-02
  • 2021-03
  • 2021-04
  • 2021-05
  • 2021-06
  • In this study EDC NHS crosslinked


    In this study, EDC/NHS crosslinked collagen films were derivatised with VWFIIINle-containing THPs and the effect on VWF and DDR2 activity was examined. First, we investigated restoration of the binding of DDR2 and VWF A3. Cell behavior beyond simple binding was also investigated through DDR2 5-Formyl-UTP triggered in transfected HEK293 and COS-7 cells by THP-derivatised collagen films. In parallel, we studied the ability of such films to support thrombus deposition from flowing blood, as well as promote Human Umbilical Vein Endothelial Cell (HUVEC) activity. Our data suggest that this approach will restore or enhance physiological function of collagen-based biomaterials.
    Materials and methods
    Discussion EDC/NHS crosslinking creates covalent bonds between lysine residues and glutamate or aspartate residues within the collagen sequence [47]. This affects integrin-binding sites directly, which are composed of a Gxx'GEx'' generic sequence [13], and the resulting loss of integrin binding was demonstrated in our previous work [[8], [9], [10],15]. The binding site for DDR2 and VWF A3 was located within residues 572 to 583 of the collagen helix using homotrimeric collagen II and III Toolkits. It contains the sequence GARGQOGVMGFO in collagen II and GPRGQOGVMGFO in collagen III. Neither contains any residues likely to form crosslinks after EDC/NHS treatment. Collagen I, however, used to form the films in this study, is a heterotrimer, and the corresponding site, GARGQAGVMGFO, in its α1 chain is not competent to bind VWF A3, since it lacks a critical hydroxyproline residue. A composite VWF-binding site has been proposed [48] which requires all three strands of collagen I, and the α2 chain contributes the missing O residue. Crucially, the corresponding α2 chain sequence, GARGEOGNIGFO, contains a glutamate residue that, if crosslinked, would obstruct incoming binding partners including DDR2 and VWF A3. This binding site in collagens I, II and III is also adjacent to a reactive lysine (K586 in the collagen helix). Chemical crosslinking of this nearby residue might also restrict access of the collagen-binding proteins, VWF and DDR2. We adapted the method of Khew et al. [37] to synthesise THP ligands. Peptide strands were elongated on resin, with an Ahx linker added at the N-terminus to provide flexibility to favour subsequent end-stapling. This linker will also further separate the end-stapling site and the active binding site of the THP, thus limiting undesired interactions between the two. End-stapling was carried out by coupling the tri-acid hexapeptide Fmoc-GFGEEG to three resin-supported strands. Addition of N-terminal templates in an attempt to covalently link the three peptide strands is known to be a difficult and slow reaction [49]. Previous attempts in our lab to end-staple peptides at their C-terminus by introducing two lysine residues gave poor results. The end-stapling strategy presented here has two main advantages. Firstly, it stabilises the active triple-helical conformation increasing the transition temperature by 4 °C (or 9 °C for an end-stapled THP containing GFOGER) [15]. Secondly, it provides a single reactive site, the primary amine located at the end-terminus side, per end-stapled THP. Molecules of interest, such as fluorophores, or, in this case, diazirine, can then be grafted to THPs with a 1:1 stoichiometry. Diazirine is a photoactive group able to react readily with any carbon skeleton after UV exposure, making this technology applicable to a wide range of substrates. Furthermore, in contrast to usual strategies for grafting ligands, this allows collagen film derivatisation with pre-assembled and functional THPs. We used the conditions for covalent linkage of photoreactive THPs that were previously optimised for the integrin-binding THP, GFOGER [15], where, after 5 min of UV treatment and at a THP concentration of 5 μg/ml, maximal and specific covalent linkage of the photoreactive THP was observed. We anticipate the same specific reaction here, using GPRGQOGVNleGFO. DDR2 and VWF binding to films derivatised in this way was assessed using recombinant DDR2-Fc and VWF A3-GST fusion proteins. In both cases, best binding was obtained after treatment with diazirine-containing peptides, and only after UV exposure. Compared to crosslinked collagen not treated with peptides, the binding of VWF A3 and DDR2 increased seven-fold and three-fold respectively. Increased affinity was also observed without UV treatment or in the presence of non-photoreactive peptides. This may be due to passively-adsorbed peptides still present despite stringent washings. Nonetheless, for long-term applications, covalent attachment to our collagen substrate remains desirable to eliminate peptide elution.