Exploring the MOF-peptide interface: from phage display to materials synthesis, thin films and composites

Lead Research Organisation: University of Southampton
Department Name: Sch of Chemistry

Abstract

The synthesis of materials with complex structures and well-defined properties is a central focus of the 'Directed assembly of extended structures with targeted properties' grand challenge, given their importance to economic growth and role in addressing key societal challenges. In natural biomineralisation processes the self-assembly and recognition properties of peptides are exploited to deposit inorganic materials with exquisite structures and highly specialist functions (e.g. teeth, bones, armour), and these principles can be used by synthetic chemists, nanotechnologists and surface scientists for control over materials structure and properties.

Due to their high surface areas, tuneable compositions and functionality microporous metal-organic frameworks (MOFs) assembled from metal ions and organic linkers have demonstrable applications across numerous sectors (e.g. energy, sustainability and healthcare), but a degree of processing is often required in order to facilitate their practical use. This is a rapidly growing area of MOF chemistry and while significant progress has been made, challenges in the preparation of thin films and MOF-based composites still remain. If peptide sequences that can specifically recognise these important framework materials could be readily identified, then greater control over MOF structure, properties and deposition could be afforded.

In this work we will use combinatorial libraries of viruses called bacteriophages to identify such peptides. Each phage displays a unique peptide on its surface, and the library contains millions of different viruses and hence potential binding sequences. This process is known as phage display. By exposure of MOF crystal surfaces to the phages over several cycles the strongest binding sequences can be determined as a function of framework composition, connectivity and particle size/shape. Once identified, the peptides can be synthesised and exploited for biomineral-inspired MOF synthesis.

The ability of the identified peptides to direct MOF growth will be investigated permitting control over physical aspects of the MOF crystals such as size and shape, with potential to further influence the network structure and porosity of the framework itself. These are important properties for gas storage, catalysis and drug delivery. An understanding of MOF-peptide interactions will be beneficial to the latter, and our studies of the binding interface will provide valuable data.

Biomineralisation processes are characterised by the ability of organic molecules, including peptides, to deposit inorganic materials under mild conditions. Some MOFs such as those based on titanium remain challenging to make, but are a highly desirable synthetic target for their photoactive properties and clear applications in photocatalysis, light harvesting and energy generation. To overcome some of these synthetic barriers we will use peptides that specifically recognise the mineral titania as a strategy to discover new titanium-based MOF photocatalysts for sustainable applications.

The peptides derived from phage display will also be able to specifically recognise frameworks based on composition, functionality and crystal face. This recognition capability will be exploited for enhanced MOF interfacing with other functional components such as metal nanoparticles and biomolecules to yield new composites optimised for catalysis and adsorption. By patterning surfaces with multiple peptides, the ability to localise different MOFs into pre-defined positions will allow the preparation of multifunctional MOF thin films, a major step toward realising MOF-based devices for electronic, optical, sensing and energy applications.

The project outlined is necessarily and strongly interdisciplinary in nature, spanning combinatorial biology, materials science, surface science and nanotechnology, further supported by computational chemistry, to advance the science and technology of MOFs.

Planned Impact

Due to their high surface areas, tuneable compositions and functionality MOFs have potential applications across numerous sectors including energy, sustainability and catalysis. It is recognised however that a degree of processing is required in order to facilitate their practical use, with composites and coatings being important targets. This work addresses this by applying combinatorial biology methods to identify peptides that strongly bind to MOFs, where the ability to use sequence-specific peptides to control physical aspects of MOFs and improve their interfacing with other functional components, including their spatially controlled deposition onto surfaces, will be extremely enabling toward these goals. A number of impacts are expected, and these are detailed below.

The economic impacts to be realised first will be the entry of highly trained individuals into the labour market. Their interdisciplinary skills and research training in materials will allow the PDRAs to compete effectively for jobs and contribute directly to the UK economy in this clear sector of wealth creation. The work proposed is focussed on developing enabling technology for MOFs, and as such direct measureable impacts on business are likely to be in the medium-longer term. Of particular significance will be the development of strategies toward multivariant MOF thin films. The ability to deposit multiple MOFs of differing functionality onto a single surface in a pre-defined way will be a major step toward realising MOF-based devices for electronic, optical and sensing applications, and businesses with these divisions in their wider portfolios are likely to benefit most. Improvement of MOF-composite properties through optimisation of porosity and activity via peptide recognition will further facilitate their potential in these and related areas. The targeting of new MOF-based photocatalysts using our biomineralisation-inspired framework discovery approach will also be of interest for those businesses engaged in catalysis, fuel cells, biofuel production and other clean energy generation. In all cases, knowledge and technology transfer coupled with an assessment of user needs is required and our engagement strategy is outlined in the pathways to impact document.

Sustainability to reduce energy inputs, reliance on fossil fuels and by-product waste is a major societal challenge, and many industries have initiatives in place to improve their credentials in this area. Biomineralisation approaches permit materials synthesis to be conducted under mild reaction conditions, and in this proposal we apply these principles to 'hard-to-make' MOFs. The benefits of this approach are twofold: reduce energy and waste for these MOFs (initially at the lab scale) and increase their synthetic compatibility for composite and thin film formation. Related to this is the growing area of photocatalysis, where light is harnessed for atom economical chemical reactions and efficient clean energy generation, and the MOF photocatalysts outlined will make direct contributions to these areas and the wider sustainability agenda.

MOFs also have significant potential in the healthcare sector as drug delivery vectors and imaging agents. While both applications have been demonstrated, our understanding of how MOFs interact with important biological proteins remains poor. This work will increase our knowledge of the MOF-peptide interface, allowing the surfaces of MOFs to be more carefully designed in future biomedical formulations to optimise their biodistribution and transport pathways while reducing toxicity and side effects. These impacts on healthcare will necessarily be longer term, as complementary biochemical studies of MOFs on natural systems are needed.

It is our intention to make the public and policy makers aware of these impacts and the clear link to basic science through appropriate outreach and networking events as outlined in the pathways to impact document.

Publications

10 25 50
publication icon
Crickmore TS (2021) Toward sustainable syntheses of Ca-based MOFs. in Chemical communications (Cambridge, England)

publication icon
Lupica-Spagnolo L (2018) Pollen-like ZIF-8 colloidosomes via emulsion templating and etching. in Chemical communications (Cambridge, England)

 
Description The key findings of this project so far are summarised below:

The combinatorial phage display (Ph.D.) method can be used to identify strongly-binding peptide sequences for metal-organic frameworks (MOFs), as demonstrated for a number of common MOF materials including UiO-66, ZIF-8, HKUST-1 and Ti-MIL-125. Full sequence analysis supported by simulation approaches allow us to identify which specific amino-acid residues or combination of residues are most recurring and important to MOF binding, which may permit future targeted peptide design.

In collaboration with the team of Hendrik Heinz at Boulder, we have developed simulation approaches to investigate the binding mechanism of peptides toward MOFs. This has focused on the UiO-66-X system and both on- and off-target binding are considered. The simulations have been highly successful in reproducing the order of binding observed during Ph.D., allowing us to visualize which MOF-peptide interactions are important and effective modelling of the surface incluing the interactions of water molecules with the amino-functional groups in UiO-66-NH2.

Framework functionalisation can lead to a unique-binding sequence: amino-functionalization of UiO-66 results in a unique binding sequence being identified from a phage library of 10^9 possible sequences whereas the non-amino derivative yields a smaller number of binding sequences. This is verified using non-functionalized ZIF-8, and implies that frameworks bearing additional functional groups have a greater potential to interact strongly with phage-bound peptides. Different sequences have also been identified when the framework crystallinity, topology or crystal morphology is changed but composition remains constant such as in the polymorphic ZIF system. In this latter case, two sequences are identified, one which binds to all polymorphs investigated (a general-binder) and one that is specific to sodalite. We have also observed some degree of discrimination between sodalite crystals of differing morphology, which suggests that the Ph.D. peptides are able to bind to the MOF in a facet-selective manner. This may open the door to highly controlled surface functionalisation of MOF crystals in the future.

The binding of peptide sequences identified by Ph.D. to MOFs can be determined using isothermal calorimetry (ITC) and changes in zeta-potential (ZP) on peptide-binding. Typical dissociation constants are in the range 10-6 M-1, and it is possible to extract thermodynamic properties such as ?G values from the ITC data. A complete study was performed on UiO-66 and the amino-functionalised analogue UiO-66-NH2 and while peptide-binding was found to be enthalpically-driven the differences between on- and off-target MOF-peptide binding were more similar than expected. This is attributed to the differences in peptide behaviour between the Ph.D. screening and the binding tests in solution and the high degree of aggregation of the relatively hydrophobic peptides when in solution. The experimental binding observations have been consistently reproduced using simulation in most cases.

Binding studies are further supported by synchrotron radiation circular dichroism (SRCD) spectroscopy, applied to study the MOF-peptide interface in this project for the first time. By comparison of the peptides alone in solution and in the presence of the desired MOF (UiO-66-X or ZIFs) then conformational changes upon binding can be monitored. This is further supported by computational approaches, where simulated Ramachandran plots of the torsional angle landscape significantly aids interpretation of the experimental data. Overall this has been an excellent addition to the toolbox of methods to study how peptides interact with MOFs.

An extensive study of the binding of titania-binding peptides toward Ti-MIL-125 has also been undertaken as part of this work. We aimed at a comparison of binding affinities of the MOF PH.D. sequences against the two MOFs in comparison to anatase and rutile. Additionally, we started a study with three well-known Titania binding peptides (Ti-1, Ti-2, Ti-1231) as references sequences and compare their binding affinity against anatase, rutile, Ti-Mil-125 and Ti-MIL-125-NH2. In some cases we find that the titania peptides also bind strongly to the MOFs, perhaps indicating structural similarity between the minerals and the inorganic MOF secondary-building unit clusters. Again, SRCD revealed that we find the strongest pronounced conformational changes between titania-peptides and their interaction with Ti-Mil-125-NH2. The atomistic interpretation is however complex due to the strong self-agglomeration behavior of the investigated peptides in solution, which was further confirmed by ITC and dynamic light scattering experiments.

Addition of Ph.D. peptides to MOF syntheses in a biomimetic mineralisation approach has been found to influence crystallinity and particle morphology in the UiO-66-X system, when an aqueous synthesis is employed. Interestingly, the peptides can guide framework assembly under modified conditions where this otherwise forms only amorphous products, directly mirroring natural biomineralisation approaches. Control experiments with just the amino acids or under conditions where the Ph.D. peptides fragment do not yield crystalline material indicating there is clear added value in using the peptides for MOF synthesis and that the effect of pH is minimal.

Addition of the Ph.D. peptides to the ZIF-8 synthesis, especially when starting from a point that otherwise yields an amorphous project, has the capacity to influence framework topology, and in some cases can increase the stability and persistence of metastable phases so can also modulate the kinetics of framework assembly. This follows the order of frequency in which the peptides were found during phage display screening, and general-binding peptides allow phase transformation between ZIF polymorphs but where a sequence was specifically identified for a single topology this is significantly stabilized vs. the control. In one case, we observed the recently reported katsenite phase, a topology that has only previously been accessed through ball-milling of the precursors, which further demonstrates the power of this biomimetic approach. As expected the best biomimetic framework assembly results are obtained under mild reaction conditions where the peptide is demonstrably intact.

We have found that biologically benign Ca-MOFs can be prepared directly from a range of calcium carbonate polymorphs in high yield in water under ambient conditions. Peptides previously found by Ph.D. to bind strongly to calcium carbonate are currently under investigation to determine if they can influence framework topology, morphology or crystallinity. Control over Ca-MOFs synthesis and shape is directly relevant to the use of MOFs in biomedicine, and given the prevalence of CaCO3 in natural systems provides an excellent start point to further investigate biomimetic MOF synthesis approaches.

Overall the above findings clearly demonstrate that the application of genetically modified peptide libraries can identify strong binding sequences for MOFs, and that use of these peptides in biomimetic mineralisation can direct the topology, modulate assembly kinetics and influence the crystallinity of these important framework materials in a predictable manner. This is important, as it demonstrates that small biological units can also direct materials assembly and that the mechanisms of action go beyond simple charge effects as reported for MOF-protein systems. The ability to use peptides and identify specific residues of importance through screening and simulation offers new vistas in biomimetic MOF synthesis and an increased understanding of how this important class of material interacts with biological building blocks relevant to their use in drug delivery and bio-applications more widely.
Exploitation Route The ability to direct MOF assembly with peptides and the potential to exploit their recognition capabilities will be of interest to materials scientists, surface scientists, nanotechnologists and those interested in healthcare; indeed it is the aim of our proposal to take our results forward into these areas. In such a complex project it is expected that the outcomes/impact will be mostly academic. Future funding applications based on these data to build-up a true picture of how biomolecules interact with MOF surfaces, from the simplest building blocks to tertiary protein structure are currently in preparation.
Sectors Electronics,Energy,Environment,Healthcare

 
Description This project involved the application of combinatorial biology methods for the identification of peptides which bind strongly to porous frameworks to investigate both the interactions between them and how these could be applied and as such sits firmly at the interface of the two disciplines. Given the novelty of this approach, requiring completely new experimental protocols and computational methods, as well as the application of new analytical techniques, the major impacts of this project at this stage are in developing new fundamental science. From an academic viewpoint the combination of biological methods to porous metal-organic frameworks conducted in this work remains unique in the UK and we have developed new experimental protocols and simulation methods that will be of significant use to the wider academic community once fully disseminated. In particular simulation methods developed with US colleagues at the University of Colorado (Boulder) to study the MOF-peptide interface at the atomic level will be particularly enabling for academics working with porous materials for biomedical applications which is a key route for their development as effective drug delivery and imaging agents. Combining these simulation results with bulk experimental techniques has further allowed us to build up a key understanding of how peptides behave at the surface of MOF crystals and how this can be related to their recognition capability. Ultimately the influence of the peptide sequences on the outcome of framework synthesis can be carried out in a predictable manner with high levels of control that are not accessible using small molecule additives. Although this work remains at an early technology readiness level, we are confident that the methods we have established within this project could be taken higher with appropriate development in the future. At present the most tangible non-academic output from the project are highly trained people entering the workforce. Two researchers that contributed directly to and were funded by the project are now working in chemical consultancy and the education/outreach sectors following conclusion of the project funding. This is fully in line with our expected non-academic impacts for this project. Our planned outreach programme has thus far been limited to two events (a schools talk and a nanotech related workshop) and a further series of outreach events were planned following conclusion of the project when the full implications of the research were known. Unfortunately, this has been restricted by the COVID-19 pandemic but we hope to reach out more widely in the near future.
First Year Of Impact 2020
Sector Chemicals,Education
Impact Types Economic

 
Description School visit (Salisbury) 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Schools
Results and Impact I visited Bishops Wordsworth school in Salisbury to talk to a group of 40 year 12/13 chemists and their teachers at an after-school science seminar series. The students were very engaged with the material and asked pertinent questions afterward, including a discussion with one student who had recently completed his EPQ in a directly related area.
Year(s) Of Engagement Activity 2019