Abstract

Nanoparticles are increasingly important in biotechnology as they are extensively used as drug
delivery carriers and in biosensors. In both these two contexts, protein-nanoparticle interactions
are often involved. Proteins that are present in body fluids inevitably interact with nanoparticle
based drug carriers and typically surround them forming the so called “protein corona”.
Biosensors that are based on nanoparticles often have proteins deliberately attached to their
surface, for example antibodies that bind specific analytes. The understanding of the assembly
mechanisms at the protein-nanoparticle interface and the ability to engineer proteins that
interact with nanoparticles in the desired way, are therefore two essential requisites for the
future development of nano-medicines and nano-biosensors.
In this work, we focused on the interaction of proteins with gold nanoparticles (GNPs). GNPs
are available with a broad range of surface chemistries, suitable for the conjugation of many
biomolecules. Although there are at least three decades of studies on gold colloids with
different surface chemistries, there is still quite little known about what are the exact features
of a protein that determine its adsorption onto gold. We developed methods to study this and
applied them to characterise the adsorption on GNPs of Glutathione-S-Transerase (GST),
which was reported previously as a protein that strongly binds gold. We determined its affinity
and kinetics of binding and unravelled the mechanism of its thiol-mediated chemisorption. We
found that GST binds to GNPs even more efficiently than other known gold-binding proteins,
such as Bovine Serum Albumin (BSA). We concluded that GST could be considered a very
useful gold-protein interface, especially considering that GST fusion is routinely used for
affinity purification of recombinant proteins and therefore well established.
We also fused self-assembling proteins to GST or chemically cross-linked them to BSA. The
scope was to explore the feasibility of hierarchical and ordered assembly of designer proteins
onto GNPs, with the ultimate goal of providing a convenient tool for modular assembly of
proteins onto nanomaterials. It is known that proteins tend to denature and lose their function
when in contact with GNPs, which is not optimal for biosensors or in nanomedicine. We found
that it is possible to use GST or BSA to form a sacrificial layer on gold, which exposes linked,
self-assembling proteins that are able to bind their counterpart, unaffected by the GNP surface.
We reported two proof-of-concepts: the first based on mimics of the self-assembling neuronal
SNARE proteins and the second based on the pair SpyCatcher/SpyTag, derived from
Streptococcus pyogenes proteins and used in bio-conjugation for their ability to self-catalyse
the formation of isopeptidic bonds.
We believe that the novel methods and original results presented in this thesis apply to both the
understanding and the engineering of the protein-nanoparticle interface and will be beneficial
for the broad nanobiotechnology community. In fact, our findings have potential applications
in a broad range of fields, spanning from the improvement of the circulation life-time of
nanomedicines by preventing the binding of serum protein and opsonisation, to the
improvement of the manufacturing of GNPs-based immune-biosensors such as those used in
lateral flow devices.