Professor Jurriaan Huskens, Molecular Nanofabrication group, Department for Molecules & Materials, Molecules Centre & MESA+ Institute, University of Twente, Enschede, The Netherlands

Multivalent Interactions in the Detection of DNA and Viruses

Multivalency describes many interactions in Nature, for example, the interaction between viruses and cell membranes [1]. The influenza virus binds through multiple sialyl-terminated carbohydrates (SLNs) non-covalently interacting with hemagglutinin coat proteins of the virus particle. This interaction is weakly multivalent in nature, and therefore it responds very sensitively to the density of carbohydrates. This behavior explains the large differences between virus affinities observed for mutations in the receptor binding domain.

A key aspect of the multivalent interaction of viruses at cell membranes is its strong, non-linear dependence on the receptor density displayed at the surface. We here show the use of surface gradients of receptor-modified supported lipid bilayers (SLBs) to visualize and quantify the receptor density dependence in one microscopic image. This technique is called “Multivalent Affinity Profiling” [2]. The fitting of the data by a thermodynamic model allows quantification of the threshold density, comparison of binding selectivities for different virus strains, and thus offers a molecular understanding of the supramolecular binding energy landscape [3]. This supramolecular and nanoscopic picture links fundamental molecular aspects of binding to biological processes of antigenic drift and zoonosis.

The same principles of weak multivalent and superselective binding can be used in the design of a device that selectively captures the cancer biomarker hypermethylated DNA (hmDNA) [4]. The selectivity is achieved by careful density control over the surface coverage of the proper receptor for hmDNA, the methyl-binding domaln (MBD) protein [5]. Microfluidic devices coated with monolayers of MBD thus allow purification and up-concentration of hmDNA from solution mixtures with large backgrounds of non-methylated DNA, from cell extracts and from patient samples. This technique promises a faster and cheaper method to employ hmDNA as a biomarker in liquid biopsies, with potential applications as a screening method and for follow-up of treatment.

References
[1] N. J. Overeem, E. van der Vries, J. Huskens, Small 2021, 17, 2007214
[2] N. J. Overeem, P. H. Hamming, M. Tieke, E. van der Vries, J. Huskens, ACS Nano 2021, 15, 8525
[3] N. J. Overeem, P. H. Hamming, O. C. Grant, D. Di Iorio, M. Tieke, M. C. Bertolino, Z. Li, G. Vos, R. P. de Vries, R. J. Woods, N. B. Tito, G.-J. P. H. Boons, E. van der Vries, J. Huskens, ACS Central Science 2020, 6, 2311
[4] R. Kolkman, L. I. Segerink, J. Huskens, Adv. Mater. Interfaces 2022, 9, 2201557
[5] R. W. Kolkman, S. Michel-Souzy, D. Wasserberg, L. I. Segerink, J. Huskens, ACS Appl. Mater. Interfaces 2022, 14, 40579