Initial focus on displaying enzymes around prototypical nanoparticulates such as semiconductor quantum dots (QDs) and gold nanoparticles (AuNPs) and evaluating their subsequent catalytic activity revealed the presence of several fascinating phenomena including the ability to both stabilize an enzyme's structure along with significantly enhancing its kinetic activity in many cases.^1-5 These properties manifest as a result of how the enzyme is attached to the NPs along with the unique interfacial environment found at the NP-enzyme interface, the latter of which arises from the ability of colloidal NPs to structure their surrounding matrix.  In our work, we utilize the pendant His₆ motifs expressed at the termini of protein monomers to self-assemble to the NPs via metal affinity coordination as our primary NP attachment or bioconjugation chemistry.⁶ When utilizing multimeric enzymes displaying multiple pendant His₆ motifs, it was observed that the proteins would crosslink the NPs into nanoclusters and this would eventually lead into our current research efforts and this publication.⁷

Assembling two enzymes with coupled catalytic steps into these self-assembled nanoclusters revealed that they could engage in concerted intermediary or probabilistic channeling.^7-9 Probabilistic channeling occurs when enzymes are positioned close enough such that the intermediary product of the first enzyme has a high probability of immediately being taken up as substrate by the next enzyme rather than diffusing away.  This can in turn substantially increase the overall rate of coupled kinetic flux since molecular diffusion rates are typically orders of magnitude quicker than the enzyme's catalytic rates in bulk solution. As such, channeling represents one of the most efficient strategies for increasing the catalytic flux in a diffusion-limited multienzymatic reaction.  There has been continuous controversy surrounding how to demonstrate that two enzymes are indeed engaged in channeling and several groups have provided convincing evidence that the ubiquitous glucose oxidase-horse radish peroxidase coupled enzyme system does not engage in channeling.¹⁰ Given this, we set out to evaluate other systems and to determine how many enzymatic steps could be incorporated into the self-assembled NP-enzyme systems we were exploring along with trying to understand their functional limitations so that a first set of working design principles could be elucidated.   The enzymes of oxidative glycolysis along with some additional upstream saccharification steps were utilized along with QDs as a model system for this study.¹¹  This allowed us to create nanoclustered-cascades incorporating from 4 to 10 enzymatic steps, see Figure 1.

^{Figure 1. Self-assembled catalytic NP-enzyme clusters.  (a)  Multiple His₆-termini (purple) on the multimeric enzymes coordinate to the NP surfaces and functionally crosslink them into nanoclusters as shown with phosphofructokinase I (PFK, PDB #1PFK) and the 3 different sizes of QDs along with nanoplatelets (NPLs) used at scale relative to PFK. NPLs are shown angled for perspective. (b) Schematic depicting the self-assembled QD enzyme clusters forming multienzyme cascades that are the focus of this study.  QDs are mixed with stoichiometric ratios of enzymes that constitute a targeted cascade and self-assemble into nanoclusters. Addition of initial substrate such as linear starch is then processed to product by the multienzyme cascade in the cluster, which exploits localized intermediary channeling. (c)  Forming into NP-enzyme clusters and engaging in multistep channeling increases the overall catalytic flux by orders of magnitude over that of freely diffusing enzymes, which encounter significant diffusion limitations. Channeling manifests by substantially reducing the overall transient time (t) for that reaction. Loosely defined, t is the time it takes an initial substrate to be processed to final product in a multistep enzymatic reaction.}

A variety of classical experimental formats designed to target different mechanistic aspects of channeling were implemented to confirm its presence in these nanoclusters. Channeling increased catalytic flux in selected enzymatic cascades from ~10 to more than 100 times that of the same concentration of enzymes in control samples lacking QD or other NP presence (see Figure 2).  It was also observed that channeling efficiency in the nanoclusters could be further enhanced several fold more by optimizing relative enzymatic stoichiometry using numerical simulations of the reaction kinetics, switching from spherical QDs to 2-D planar nanoplatelets (NPLs), and by ordering the NP-enzyme assembly process.  

^{Figure 2. Nanoparticles and catalytic performance in the 7 enzyme (7E) glucose to 3-phosphoglycerate (3-PG) cascade.  (a) Representative progress curve measuring NADH conversion over time for 520 QD clusters assembled with the 7E system (glucose to 3-PG) at empirical enzyme ratios (red) vs. same enzyme free in solution (blue). 10 uM indicates the final amount of NADH converted in the free enzyme assay. Progress curves assembled using optimized (Opt) enzyme ratios per QD determined after two consecutive rounds of numerical simulation (pink - Opt 2). Free enzyme controls for optimization had identical results as that of the empirical sample. 520 QD concentration = 2.5 nM. (b)  Representative TEM micrographs of clusters formed with 660 QDs and 585 NPL materials using the 7E cascade at Opt 2 ratios with QD = 6.25 nM and NPL = 1.25 nM.  Average cluster size is given above the micrograph along with the number of QDs counted in that determination. Inset, representative high-resolution micrograph of an individual cluster. In interpreting these images it should be remembered that changes may have occurred either in deposition on the TEM grids or in the high vacuum of the TEM. Enzymes can be seen in the TEM images as the shaded areas around some of the 660 QDs. (c)  Representative progress curve assay data from the 7E system at Opt 2 ratios with enzyme concentration fixed (Glk 5.5, PGI 1, PFK 9, FBA 12, TPI 1, GPD 27, PGK 7.5 nM) as assembled with the indicated increasing concentrations of 520 QD present in the reaction. (d) Representative TEM images with corresponding cluster analysis bar plots for the 0.63, 2.5, and 25 nM samples in panel c revealing that cluster size increases with added QD.}

Detailed physicochemical analyses including especially TEM imaging of clusters were used to characterize assembly formation and clarify structure-function properties; these showed that increasing nanocluster size by adding more QDs relative to enzyme acted to further enhance channeling by incorporating more enzymes into each cluster (see Figure 2).  For extended cascades with unfavorable kinetics, channeled activity could still be maintained in the overall multienzyme catalytic process by splitting at a critical step, purifying end-product from the upstream sub-cascade, and feeding it as substrate to the downstream sub-cascade.  Lastly, the generalized applicability of this approach towards accessing channeling was verified by extending to assemblies incorporating other hard and soft NP materials.  

We believe that such self-assembled biocatalytic nanoclusters offer many potential benefits towards enabling minimalist cell-free synthetic biology. Along with providing an easily accessible platform to help understand many other properties associated with enzymatic channeling, these structures suggest themselves as possible components for artificial cells as they recapitulate some of the metabolic functionality of cells without requiring the confinement of surrounding membranes. They can also act as artificial de novo catalytic pathways or metabolons for use in enzyme-based biosynthesis of non-natural products, which is something not generally accessible within cell-based synthetic biology.¹² The ability to self-assemble enzyme clusters for small volume, high-efficiency reactions and then screen their capabilities against many different natural and non-natural substrates would allow these systems to be coupled into combinatorial chemistry approaches for the synthesis of pharmaceutical libraries.

References:

1. Johnson, B. J., Algar, W. R., Malanoski, A. P., Ancona, M. G. & Medintz, I. L. Understanding enzymatic acceleration at nanoparticle interfaces: approaches and challenges. Nanotoday 9, 102-132 (2014). doi: 10.1016/j.nantod.2014.02.005
2. Vranish, J. N., Ancona, M.G., Walper, S. A. & Medintz, I. L. Pursuing the promise of enzymatic enhancement with nanoparticle assemblies. Langmuir 34, 2901-2925 (2017). doi: 10.1021/acs.langmuir.7b02588
3. Ellis, G. A., Diaz, S. A. & Medintz, I. L. Enhancing enzymatic performance with nanoparticle immobilization: improved analytical and control capability for synthetic biochemistry. Curr. Opin. Biotechnol. 71, 77-90 (2021). doi: 10.1016/j.copbio.2021.06.021
4. Vranish, J. N., Ancona, M. G., Oh, E., Susumu, K. & Medintz, I. L. Enhancing coupled enzymatic activity by conjugating one enzyme to a nanoparticle. Nanoscale 9, 5172-5187 (2017). doi: 10.1039/C7NR00200A
5. Díaz, S. A., Choo, P., Oh, E., Susumu, K., Klein, W. P., Walper, S. A., Hastman, D. A. Odom, T. W. & Medintz, I. L. Gold nanoparticle templating increases the catalytic rate of an amylase, maltase, and glucokinase multienzyme cascade through substrate channeling independent of surface curvature. ACS Catalysis 11, 627-638 (2020). doi: 10.1021/acscatal.0c03602
6. Díaz, S. A., Breger, J. C. & Medintz, I. L. Monitoring enzymatic proteolysis using either enzyme-or substrate-bioconjugated quantum dots. Methods in enzymology 571, 19-54 (2016). doi: 10.1016/bs.mie.2016.01.001
7. Vranish, J. N., Ancona, M. G., Oh, E., Susumu, K., Lasarte Aragonés, G., Breger, J. C., Walper, S. A. & Medintz, I. L. Enhancing coupled enzymatic activity by colocalization on nanoparticle surfaces: kinetic evidence for directed channeling of intermediates. ACS Nano 12, 7911-7926 (2018). doi: 10.1021/acsnano.8b02334
8. Ellis, G. A., Klein, W. P., Lasarte-Aragones, G., Thakur, M., Walper, S. A. & Medintz, I. L. Artificial multienzyme scaffolds: pursuing in vitro substrate channeling with an overview of current progress. ACS Catalysis 9, 10812-10869 (2019). doi: 10.1021/acscatal.9b02413  
9. Hooe, S. L., Breger, J. C., Dean, S., Susumu, K., Oh, E., Walper, S. A., Ellis, G. A. & Medintz, I. L., Benzaldehyde lyase kinetic improvements, potential channeling to alcohol dehydrogenase, and substrate scope when immobilized on semiconductor quantum dots. ACS Appl. Nano Materials 5, 10900-10911 (2022). doi: 10.1.21/acsanm.2c02196
10. Abdallah, W., Hong, X., Banta, S. & Wheeldon, S. Microenvironmental effects can masquerade as substrate channelling in cascade biocatalysis. Curr. Opin. Biotech. 73, 233-239 (2022). doi: 10.1016/j.copbio.2021.08.014
11. Breger, J. C., Vranish, J. N., Oh, E., Stewart, M. H., Susumu, K., Lasarte-Aragonés, G., Ellis, G. A., Walper, S. A., Díaz, S. A., Hooe, S. L., Klein, W. P., Thakur, M. G., Ancona, M. G., & Medintz, I. L. Self assembling nanoparticle enzyme clusters provide access to substrate channeling in multienzymatic cascades, Nat. Commun. 14, 1752 (2023). doi: 10.1038/s41467-023-37255-9
12. Hooe, S. L., Ellis, G. A. & Medintz, I. L. Alternative design strategies to help build the enzymatic retrosynthesis toolbox. RSC Chem. Biol. 3, 1301-1313 (2022). doi: 10.1039/D2CB00096B