The rational design of coordination-driven supramolecular artificial enzymes: From catalysis to biomedicine

Perspective The rational design of coordination-driven supramolecular artificial enzymes: From catalysis to biomedicine Agnieszka Bajer, 1,2 Venkateswarulu Mangili, 1, *and Artur R. Stefankiewicz 1,2, * 1 Center for Advanced Technologies, Adam Mickiewicz University, Uniwersytetu Pozna� nskiego 10, 61-614 Pozna� n, Poland 2 Faculty of Chemistry, Adam Mickiewicz University, Uniwersytetu Pozna� nskiego 8, 61-614 Pozna� n, Poland *Correspondence: venkateswar[email protected].pl (V.M.), [email protected] (A.R.S.) https://doi.org/10.1016/j.chempr.2025.102784 SUMMARY Self-assembled coordination architectures are emerging as powerful platforms for creating artificial enzymes that emulate the structural and functional complexity of natural biocatalysts. By combining well-defined cavities, tunable host-guest interactions, and catalytic control within confined nanoscale spaces, these systems offer unique opportunities for advancing sustainable catalysis, molecular recognition, and biomedical innovation. In this perspective, we highlight recent advances in the design and function of coordination-driven artificial enzymes, focusing on how metal-organic architectures (MOAs) can be engineered to stabilize reactive intermediates, direct substrate selectivity, and respond to external stimuli. We outline the principles behind these supramolecular systems and explore their growing potential in both industrial and therapeutic contexts. INTRODUCTION Enzymes play a central role in biology, enabling complex chemical transformations with exceptional efficiency, selectivity, and control. 1,2 Their defining feature is the presence of active centers embedded within multifunctional substrate-binding cavities. Their activity often stems from well-defined active centers embedded within multifunctional substrate-binding cavities that orient, stabilize, and activate bound molecules. These features have inspired the widespread use of enzymes in industrial processes—from wastewater treatment and detergent formulation to pharmaceutical synthesis. 3,4 However, despite their versatility, natural enzymes are often fragile, expensive, and sensitive to environmental conditions, which limit their broader use in practical settings. 5 This has motivated the development of artificial enzyme synthetic systems designed to replicate the function of their biological counterparts while offering greater robustness and tunability. A landmark moment in this field came in 1970, when Breslow and Overman introduced the concept of an ‘‘artificial enzyme’’ combining a catalytic metal center with a hydrophobic binding pocket. 6 Since then, a wide range of supramolecular hosts, including cyclodextrins, cucurbiturils, crown ethers, calixarenes, and cryptands, has been explored for enzyme-mimetic catalysis in solution. 7–10 Yet many of these systems are limited by small cavity volumes, rigid geometries, or synthetic complexity. As catalytic applications diversify from asymmetric synthesis to reactions in aqueous or biological media, there is an increasing demand for architectures that are both more adaptable and readily scalable. 11–13 Coordination-driven self-assembly offers a compelling solution. Metal-organic architectures (MOAs), including metallocycles, cages, and prisms, can be formed from metal ions and organic THE BIGGER PICTURE Challenges and opportunities: •Many natural enzymes are powerful but fragile, limiting their use in industrial or medical settings. Designing artificial enzymes that can match nature’s precision while being more robust remains a major challenge. •This work shows how self-assembled metal-based architectures can mimic natural enzymes’ confined spaces and selective reactivity, enabling clean reactions, targeted drug delivery, and biomimetic transport for health and sustainability. •By combining modular design with responsiveness to light, pH, or other stimuli, this approach opens the door to smart, tunable catalysts that could transform how we make and control molecules supporting innovation across green chemistry, materials science, and biomedicine. Chem 11, December 11, 2025 © 2025 The Authors. Published by Elsevier Inc. 1 This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Please cite this article in press as: Bajer et al., The rational design of coordination-driven supramolecular artificial enzymes: From catalysis to biomedicine, Chem (2025), https://doi.org/10.1016/j.chempr.2025.102784 ll OPEN ACCESS ligands to create well-defined, nanoscale cavities. 14,15 These dynamic and often modular structures exhibit high stability, tunable geometry, and the ability to incorporate functional groups directly into their framework. Sometimes referred to as ‘‘enzyme-mimicking synthetic molecular containers,’’ 16 MOAs offer unique microenvironments that resemble enzyme active sites in both structure and function. 17–19 Within these confined spaces, reactions can be accelerated, selectivity enhanced, and reactive intermediates stabilized—features that are challenging to achieve in conventional homogeneous systems. 20 Although significant progress has been made in developing MOAs for catalytic purposes, their full potential in mimicking enzymatic function, particularly in asymmetric catalysis and biomedical applications, remains underexplored. 21 Recent examples demonstrate that MOAs can replicate not only the structural elements of enzymes but also their ability to direct complex chemical transformations with high fidelity. 22,23 Artificial Diels-Alderases, hydrolases, and redox catalysts encapsulated within MOAs have shown remarkable performance, offering a new approach to supramolecular catalysis. 24–26 Three primary strategies have emerged in this field: (1) designing enzyme-like confined cavities that stabilize or activate guest molecules, (2) installing catalytically active sites directly into the MOA framework, and (3) encapsulating enzyme-like catalysts within preassembled MOA hosts. 27–29 This perspective explores these approaches, highlighting recent advances and outlining future directions for the rational design of MOA-based artificial enzymes. ENZYME-MIMICKING NANOPOCKETS IN MOAs Cavities within coordination-driven molecular architectures can serve as biomimetic environments that protect, stabilize, or regulate the behavior of guest molecules, catalytic species, or transport processes without requiring built-in catalytic sites. These artificial cavities, typically formed by the spatial arrangement of metal-ligand building blocks, create confined microenvironments that resemble the active pockets of natural enzymes. Unlike the surrounding bulk solution, these pockets isolate encapsulated guests, providing a distinct chemical space. 26 Within these confined environments, guest molecules often exist in low-solvation states and experience restricted motion due to specific non-covalent interactions with the cavity walls. These include π-π stacking, van der Waals forces, hydrogen bonding, and metal coordination. Such interactions can be finely tuned to influence guest stability control substrate orientation and conformation, and they promote chemical transformations. 24,25 By lowering activation barriers, concentrating reactants, and stabilizing key intermediates, these pockets significantly enhance catalytic efficiency and selectivity. In cases where the cavity is hydrophobic but enclosed within a hydrophilic framework, these MOAs can facilitate reactions that are otherwise challenging in aqueous media. 30 This design enables catalytic processes without the need for organic solvents, offering a route toward greener chemistry. Additionally, spatial confinement can suppress side reactions, stabilize reactive intermediates, and improve regioselectivity and stereoselectivity. 27,29 Despite these advantages, only a limited number of MOAs have been successfully applied in enzyme-like catalysis. Designing host-guest interactions with the required precision for high specificity, strong yet reversible binding, and efficient catalytic turnover remains a challenge. Many current systems rely on hydrophobic effects for encapsulation and reactivity, but optimizing these features for broader substrate scope and reaction types is still an active area of research. Broadly, the catalytic roles of these enzymemimetic nanopockets can be classified into three categories: (1) modulating guest solubility and stability, (2) promoting enzyme-like transformations within the cavity, and (3) facilitating selective transport and recognition in artificial channels. Artificial nanopockets for modulating guest solubility and stability In biological systems, compartmentalization plays a crucial role in regulating the activity of reactive species by sequestering them until needed. For instance, eukaryotic cells use organelles to isolate genetic material and metabolic processes, while viruses encapsulate nucleic acids within protective capsids. 24,31 Ferritin offers another example, storing iron ions safely within a protein shell to prevent oxidative stress. 32 Inspired by these natural strategies, supramolecular chemists have developed artificial MOAs that bind guest molecules with high affinity and control their chemical behavior. 25,33,34 These confined structures can modulate solubility, 35,36 enhance reactivity, 26,37 and stabilize short-lived intermediates through selective encapsulation. 38–40 Despite the large number of reported coordination cages, only a small subset has been tailored for catalysis, and the diversity of reactions explored remains relatively limited. One striking example involves the stabilization of highly reactive white phosphorus (P 4 ). Normally pyrophoric, P 4 can be rendered both air stable and water soluble by encapsulation in a tetrahedral Fe(II)-based MOA (MOA 1), as shown by the Nitschke group. 40 The tetrahedral geometry of MOA 1 is crucial, providing a shape-complementary cavity that precisely accommodates the tetrahedral P 4 molecule. This geometric match, combined with a rigid, hydrophobic interior and enforced constrictive binding, prevents oxidation and stabilizes P 4 in both air and water (Figure 1A). The resulting water-soluble host-guest complex remains stable in air for at least 4 months, yet it releases intact P 4 rapidly upon displacement with benzene. This design emulates the protective function of enzyme pockets, which bind and release reactive species in a controlled manner. 41 Another key challenge in biomedicine is improving the solubility and photostability of natural products like curcumin. Mukherjee and co-workers addressed this by encapsulating curcumin in a Pd (II) open-barrel cage (MOA 2), featuring a hydrophobic interior and hydrophilic exterior. 42 The rational design of the tetrafacial architectures of MOA 2 stabilizes curcumin through both shape and chemical complementarity. Its hydrophobic cavity, defined by four aromatic panels, forms a tightly fitting microenvironment that enhances the aqueous solubility of curcumin while shielding it from hydrolysis and photodegradation (Figure 1B). Meanwhile, the water-soluble exterior of the barrel facilitates biological delivery, rendering curcumin bioavailable and pharmacologically effective, without the need for toxic organic solvents. The encapsulated complex retained its anticancer activity in cells, showing efficacy comparable to free curcumin in DMSO, thereby Please cite this article in press as: Bajer et al., The rational design of coordination-driven supramolecular artificial enzymes: From catalysis to biomedicine, Chem (2025), https://doi.org/10.1016/j.chempr.2025.102784 2 Chem 11, December 11, 2025 Perspective ll OPEN ACCESS underscoring its potential as a therapeutic delivery platform. While coordination cages are known to stabilize various reactive species, their use in stabilizing radical initiators remains largely unexplored. 43 A landmark study by Yoshizawa and co-workers demonstrated that a Pd(II)-type capsule can encapsulate and stabilize a radical initiator in a mixed-solvent medium. 43 However, coordination cages with closed windows and pre-encapsulated templates offer limited accessibility for externally introduced guests. To address these limitations, the Chand group recently reported a water-soluble Pd(II)-based barrel (MOA 3) capable of encapsulating and stabilizing radical initiators. 44 The rigidity of its pyrene core and the steric bulk from of the TMEDA (N,N,N′,N′-tetramethylethylenediamine) ligands direct the formation of a single, discrete trigonal prismatic barrel, without the need for a templating agent. MOA 3 selectively encapsulates guests based on size and polarity, exhibiting a clear preference for anionic over polar or non-polar guests, driven by electrostatic and π-π interactions within its hydrophobic cavity. Notably, this system has been used to stabilize radical initiators, specifically azobisisobutyronitrile (AIBN) and 2-hydroxy-2-methylpropiophenone (HMPP), by shielding them from UV-induced decomposition and extending their half-lives more than 300-fold (Figure 1C). Biologically, this behavior parallels that of protective enzymes, which bind and release substrates in a regulated fashion. 45 This was exemplified by initiating a radical bromination reaction with AIBN released from the cage on demand. In another example, the Dasgupta group encapsulated an Fe(III) complex within a hydrophobic, water-soluble Pd(II) nanocage (MOA 4), enabling the formation of a rare Fe(IV)-superoxo intermediate at room temperature. 46 The species was stable, reversible, and retained reactivity toward weak C–H bonds. Without the cage, the complex decomposed or failed to bind oxygen reversibly. Inside MOA 4, however, the Fe(IV) intermediate showed remarkable spectroscopic and thermal stability (Figure 1D) and selectively reacted with suitable substrates. The rational design of MOA 4 draws inspiration from heme iron O OH O OOH O HO O + M O A 5 M O A 1 H 2 O A E B C O 2 O 2 P d F e D H i g h l y a c t i v e , u n s t a b l e a t a m b i e n t c o n d i t i o n s I m p r o v e d w a t e r s o l u b i l i t y , b i o a v a i l a b i l i t y a n d s t a b i l i t y R e m a i n e d i n t a c t u p o n i r r a d i a t i o n M a c r o c y c l e “ H o s t - i n - h o s t ” c o m p l e x a t i o n I m p r o v e d s t a b l i l i t y P h o t o d e g r a d a t i o n D e g r a d a t i o n H 2 O H 2 O H 2 O / M e C N R e a d i l y r e a c t w i t h O 2 h i g h l y e x p l o s i v e P 4 r e n d e r e d a i r - s t a b l e a n d w a t e r - s o l u b l e h ν M O A 2 M O A 3 M O A 4 H 2 O P t ≡ ≡ ≡ O P O PO PO O O O P O O O P P P P + + + G e o m e t r y c o n t r o l N NN NNR O O O O Fe O O N NN NN R O O O O Fe Cl Figure 1. Examples of artificial nanopockets for modulating guest solubility and stability (A) Spatial confinement of white phosphorus within a self-assembled cage to prevent oxidation. (B) Supramolecular encapsulation of curcumin to enhance its aqueous solubility and photostability. (C) Selective encapsulation of radical initiators such as HMPP to shield them from decomposition. (D) Stabilization of a reactive Fe(IV)-superoxo intermediate via encapsulation in a water-soluble Pd(II) nanocage. (E) Encapsulation of flexible macrocyclic hosts in a host-in-host architecture to mimic enzyme substrate binding. Please cite this article in press as: Bajer et al., The rational design of coordination-driven supramolecular artificial enzymes: From catalysis to biomedicine, Chem (2025), https://doi.org/10.1016/j.chempr.2025.102784 Chem 11, December 11, 2025 3 Perspective ll OPEN ACCESS oxygenases, such as cytochrome P450, 47 which activate oxygen for selective oxidation of C–H bonds in biological molecules. The rigid architecture and well-defined cavity stabilize the Fe(IV)- superoxo intermediate by suppressing competing decomposition pathways and reproducing the polarity, electrostatics, and geometric constraints of enzymatic pockets. Moreover, the nanocage’s water solubility and structural homogeneity enable detailed spectroscopic characterization of reactive intermediates under ambient conditions—a significant advance over traditional organic solvent-based systems. The Fujita group introduced a strategy to transform traditionally inactive or weakly binding macrocyclic hosts into highly selective and functional recognition systems by encapsulating them within a structurally preorganized coordination cage. 48 In this approach, flexible, weakly binding organic macrocycles (such as cyclotriveratrylene), when encapsulated within a large Pd(II)- or Pt(II)-based coordination cage (MOA 5), form host-in-host complexes with enhanced molecular recognition capabilities. MOA 5 provides a rigid, preorganized environment that enforces conformational control and optimally aligns the macrocycles’ binding sites, thereby mimicking receptor-like behavior or the substrate-binding domains of enzymes (Figure 1E). This design specifically emulates the adaptive substrate-binding characteristics of natural enzymes, allowing the host system to reshape and accommodate diverse guest molecules. 49 These host-in-host complexes have shown promise in molecular sensing, selective recognition, and catalysis—achieving these enhanced properties without chemical modification, relying solely on structural confinement. Artificial nanopockets for catalysis Supramolecular hosts with open, charged, hydrophobic pockets can bind a wide variety of guest molecules, including organic ions and organometallic complexes. These confined environments can mimic the catalytic performance of natural enzymes by modulating substrate orientation, stabilizing intermediates, and enabling reaction pathways not accessible in bulk solution. Notable examples include catalysis of Diels-Alder reactions, 50–52 dehalogenation, 53,54 Michael additions, 52,55 aza-Cope rearrangements, 56 Knoevenagel condensations, 57 cyclizations, 38 hydrolysis, 58,59 Kemp eliminations, 60,61 cascade reactions, 62,63 amide hydrolysis, 64–66 and allosteric catalysis. 67,68 A seminal study by Fujita and co-workers demonstrated that hexanuclear Pd(II) coordination cages (MOAs 4 and 6) can significantly accelerate Diels-Alder reactions by confining substrates in a π-stacked, preorganized geometry. 50 Inside MOA 4, anthracene and phthalimide undergo regioselective coupling to form rare 1,4-adducts instead of the typical 9,10-isomers (Figure 2A). This unusual selectivity stems from the cage’s fully enclosed, shape-persistent cavity, which emulates enzymatic active sites by enforcing strict substrate preorganization, lowering the entropic barrier to the transition state, and directing regioslectivity and stereoselectivity via topochemical control. In contrast, MOA 6 possesses a bowl-shaped geometry that enhances catalytic turnover. Although it similarly stabilizes reactants and the early transition state through π-π interactions, the bent geometry of the product disrupts these interactions, facilitating its release and regenerating the active host. These findings highlight the dual role of molecular confinement in enhancing both reactivity and selectivity while ensuring efficient catalytic turnover. Collectively, the rational design of these Pd(II)-based MOAs emulates key enzymatic strategies—substrate preorganization, transition-state stabilization, and controlled product release—paralleling the function of natural Diels-Alderases such as the SpnF enzyme, which achieves similar effects through a highly evolved active site architecture. 69 The same group extended this confinement strategy to the more challenging tetradehydro Diels-Alder (TDDA) reaction. 70 Typically requiring high temperatures and complex substrates, these reactions proceeded rapidly and with high regioselectivity when linear alkynes were encapsulated in MOA 4 (Figure 2B). The cage’s rigid geometry and size-selective cavity preorganize substrates to overcome entropic and conformational barriers, enabling efficient [4 + 2] cycloadditions under mild conditions. In addition, the combination of a hydrophobic interior and a water-soluble exterior harnesses the hydrophobic effect to trap and stabilize guest molecules, improving the reaction rate up to 10 2 -fold. This substrate folding within the cage parallels the action of pericyclase enzymes, which direct complex cyclizations through conformational control. 71 Lusby and Beves introduced a photoswitchable heteroleptic Pd(II) cage (MOA 7), 55 inspired by enzymatic catalysis, specifically the allosteric control observed in aldolase-type enzymes that mediate Michael addition reactions. 72 Their design addresses structural limitations of homoleptic cages (e.g., twisted conformations that inhibit catalysis) by combining a rigid ligand with a photoswitchable azobenzene-based ligand to generate a symmetric cavity lined with inward-directed hydrogen bond donors. This cavity mimics enzymatic active sites by stabilizing charged intermediates through CH⋅⋅⋅O hydrogen bonding and promoting substrate binding through shape complementarity and electrostatic interactions. Notably, the catalytic activity can be reversibly switched ON and OFF using visible light (405 and 530 nm), providing precise spatiotemporal control (Figure 2C). This work represents a significant step toward light-controlled artificial enzymes, demonstrating how self-assembled molecular systems can regulate catalysis in a biologically relevant and non-invasive manner. Raymond and co-workers developed a distinct class of supramolecular catalysts based on Ga(III) tetrahedral cages (MOA 8). 73 These anionic, water-soluble hosts preferentially encapsulate cationic and hydrophobic guests—unlike most other MOAs, which favor neutral or anionic species. This strong selectivity arises from the overall anionic nature of the cage, which provides an electrostatic driving force for guest binding. Encapsulation is further favored by the exclusion of solvent from the hydrophobic interior, a process both enthalpically and entropically favorable. In one application, Ga(III) cages catalyzed the aza-Cope rearrangement of propargyl enammonium cations with up to 184-fold rate enhancement (Figure 2D). 74 The products were released and hydrolyzed post-reaction, allowing catalytic turnover. This process mimics natural [3,3]-sigmatropic rearrangements, which rely on conformational control and transition-state stabilization. 75 Asymmetric catalysis has also benefited from coordination-driven self-assembly. 76 The Mukherjee group demonstrated this encapsulating anthrone in two structurally distinct Pd(II) cages (MOAs 9 and 10). 77 Within octahedral MOA 10, anthrone dimerized to yield dianthrone, Please cite this article in press as: Bajer et al., The rational design of coordination-driven supramolecular artificial enzymes: From catalysis to biomedicine, Chem (2025), https://doi.org/10.1016/j.chempr.2025.102784 4 Chem 11, December 11, 2025 Perspective ll OPEN ACCESS whereas in the double-square MOA 9, oxidation occurred to form anthraquinone (Figure 2E). These outcomes were reproducible across different cage pairs, with the shape and internal geometry of each cage determining guest positioning and stabilization through π-π stacking and hydrophobic effects. Such preorganization directs the substrate along distinct reaction P d G a ≡ ≡ B + + M O A 4 F l e x i b l e s u b s t r a t e H 2 O H 2 O E n f o r c e d f o l d e d c o n f o r m a t i o n C o n t r o l l e d c y c l i z a t i o n 9 . 7 : 1 C E A n t h r a q u i n o n e D i a n t h r o n e M O A 1 0 M O A 9 H 2 O H 2 O A n t h r o n e c h e m i c a l t r a n s f o r m a t i o n D F G a z a - C o p e r e a r r a n g e m e n t H y d r o l y z e d p r o d u c t M O A 8 H 2 O [ 3 , 3 ] - N H 2 ( C H 3 ) 2 H 2 O A C a g e - c o n t r o l l e d D i e l s - A l d e r r e a c t i o n s + H 2 O H 2 O H 2 O H 2 O H 2 O M O A 4 M O A 6 H 2 O O O O O O E n d o - b i c y c l o [ 3 . 2 . 1 ] o c t a n o n e d e r i v a t i v e E x o - b i c y c l o [ 3 . 2 . 1 ] o c t a n o n e d e r i v a t i v e M O A 1 2 M O A 1 1 D M S O / H 2 O D M S O / H 2 O D M S O / H 2 O D M S O / H 2 O A c y l o i n R e a r r a n g e m e n t O O O DO OO D O O N O O HO OH N O O N OO OH O O O O O NO2 O O NO2 O C a t a l y s i s o f a M i c h a e l a d d i t i o n r e a c t i o n M O A 7 D C M / M e C N + O R O O R O O HO H H Ph R O O OH H H Ph F e R u ≡ ≡ E l e c t r o r e d u c t i o n o f N O 2 - M O A 1 3 [ N O 2 ] - N H 4 + D M S O 6 e - C a t a l y t i c r e d u c t i o n D C M / M e C N Figure 2. Examples of artificial nanopockets for catalysis (A) Cage-controlled Diels-Alder reactions enhancing regioselectivity and reactivity through molecular confinement. (B) Supramolecular confinement facilitating TDDA reactions under mild conditions. (C) Visible-light-controlled photoswitchable catalysis within a MOA, allowing precise, programmable control of activity through cavity preorganization and ligand exchange. (D) Ga(III)-based tetrahedral assembly accelerating the aza-Cope rearrangement of propargyl enammonium cations and enabling catalytic turnover. (E) Shape-dependent chemical transformation of anthrone inside Pd(II) cages, yielding distinct products through divergent reaction pathways. (F) Enantioselective photocycloaddition and rearrangement affording bioactive bicyclo[3.2.1]octanes through precise microenvironmental control. (G) Mimicking nitrite reductase by efficiently converting nitrite to ammonium with high selectivity and enzyme-like behavior. Please cite this article in press as: Bajer et al., The rational design of coordination-driven supramolecular artificial enzymes: From catalysis to biomedicine, Chem (2025), https://doi.org/10.1016/j.chempr.2025.102784 Chem 11, December 11, 2025 5 Perspective ll OPEN ACCESS pathways, demonstrating that the cage functions not merely as a passive container but as an active participant in defining reaction outcomes. Su and co-workers reported chiral, photosensitive Ru (II) cages (MOAs 11 and 12) incorporating multiple functional chiral pockets that mimic enzymatic confinement. 78 The precise shape and size of the cavities promoted heterocycloaddition between α,β-unsaturated ketones and cyclic 1,2-diones while suppressing undesired dimerization. This microenvironmental control enabled a visible-light-induced asymmetric [2 + 2] photocycloaddition followed by a 1,2-acyloin rearrangement, affording chiral bicyclo[3.2.1]octanes (Figure 2F)—structural motifs common in bioactive natural products. The cage mirrors the cytochrome P450 (CYP) family 47 in that its pocket geometry governs substrate binding and selectivity, although the catalytic mode is photochemical rather than redox based. This represents the first example of an enantioselective visible-light-driven cascade [2 + 2] cycloaddition/acyloin rearrangement enabled by a supramolecular architecture, opening avenues for complex photochemical transformations under biomimetic confinement. Building on the concept of biomimetic catalysis in confined spaces, Duan and co-workers presented a tetrahedral Fe(II)-based organic cage (MOA 13) 79 that mimics the biological function of nitrite reductase enzymes. 80 Featuring a large internal cavity and hydrazide-functionalized ligands, the cage forms a confined microenvironment with hydrogen bond acceptor sites, enabling selective recognition and thermodynamic activation of nitrite ions (NO 2 − ). These host-guest interactions facilitate the selective electrocatalytic reduction of nitrite to ammonium (NH 4+ ) with high yield (85.6%) and selectivity and minimal byproduct formation (Figure 2G). Kinetic analysis confirmed Michaelis-Menten behavior, consisted with biomimetic catalysis. The confined space and non-covalent interactions function analogously to enzyme active sites by enhancing substrate specificity and electron/proton transfer. MOA 13 thus acts as a synthetic nitrite reductase, with potential applications in environmental remediation and sustainable ammonium production. Artificial biomimetic channels Biological transport channels play a vital role in numerous physiological processes by enabling the selective passage of ions and small molecules across cellular membranes. 27 Their efficiency arises from finely tuned structural features such as chirality, internal polarity, and conformational responsiveness that allow for precise control of substrate flow. 28 These natural systems respond to chemical gradients or molecular signals and regulate transport with high specificity. However, their complexity, instability outside biological environments, and limited availability hinder broader technological applications. 81 Since the first report of a synthetic ion channel in 1982, 82 supramolecular chemists have pursued artificial systems that mimic these functions. 83 Among the most promising are MOAs, which combine structural precision, modularity, and self-assembly into well-defined ion-conducting units. 29,84,85 Successful design of MOA-based channels requires balancing several key features, i.e., sufficient lipophilicity for bilayer insertion, appropriate length to span a typical phospholipid membrane (∼3.5 nm), and wellpositioned ion recognition sites to ensure selective and directional transport. A foundational example comes from Fyles and co-workers, who constructed Pd(II)-based coordination squares (MOA 14) by combining amphiphilic units with bipyridine ligands. 84 These square-planar complexes bear long alkyl chains that enable insertion into lipid bilayers. Conductance studies revealed three distinct phases of ion transport: erratic activity; short-lived channel openings; and eventually, stable, highly conductive states. These final states likely result from the formation of toroidal pores via dynamic aggregation with lipid molecules. This system is particularly relevant for chloride anion transport, as the high positive charge of the palladium centers in the assembled complexes promotes anion-selective conduction, including efficient transmembrane transport of chloride ions (Cl − ) through the resulting toroidal pores (Figure 3A). Functionally, this behavior mimics natural anion channels such as ClC-type chloride channels, which regulate ion flow through membrane-spanning conduits. 86 The Liu group developed a rationally designed synthetic ion channel 87 that emulates the function of biological channels such as pendrin or cystic fibrosis transmembrane conductance regulator (CFTR). 88 These artificial channels are constructed via coordination-driven self-assembly using BINOL-based ligands and Zn(II) metal ions. The resulting cages can be chemically modified to tune the pore microenvironments. Among three variants, the ethoxyl-functionalized cage (MOA 15) exhibited the highest performance, achieving exceptional iodide transport efficiency and I − /X − selectivity up to 38, attributed to its narrow, hydrophobic, and well-defined pore structure (Figure 3B). This design mirrors the biological strategy of using hydrophobic, size-exclusion-based pathways for efficient and selective anion transport. Operated via a channel— rather than carrier mechanism—its activity was confirmed in lipid bilayers by fluorescence assays and planar bilayer conductance measurements. These findings underscore the potential of MOA 15 for biomimetic ion transport, sensing, and membrane-based separations. More recently, Xu and co-workers designed a family of benzo[c][1,2,5]thiadiazole (BTZ)-based metallohelicates (MOA 16) specifically for ion transport. 89 These systems feature tunable 3D structures with internal cavities and varied alkyl side chains to modulate membrane compatibility and control transport efficiency. MOA 16 selectively transports Cl − and displays size-dependent anion selectivity (Figure 3C). Functionally, it mimics chloride channels such as CFTR and may hold potential for addressing chloride transport disorders. 90 The same group also demonstrated that metallacage-based systems can function as unimolecular ion channels. 91 Unlike earlier systems that rely on self-assembled aggregates, these structures span the lipid bilayer as single, well-defined entities. In one example, a naphthalene diimide-based metallacage (MOA 17) was constructed via subcomponent self-assembly and shown to insert into membranes and to mediate chloride transport (Figure 3D). 91 The cage was rationally designed to ensure lipophilicity for membrane insertion; sufficient length to span the bilayer; and rigid, π-acidic surfaces for specific ion recognition via anion-π interactions. Zn(II) was selected as the coordination center for its biocompatibility as an essential trace element in humans. Molecular dynamics and density functional theory (DFT) studies confirmed that the cage forms a continuous ion-conducting conduit through the bilayer. Functionally, this system disrupted ion gradients in cancer cells and inhibited proliferation Please cite this article in press as: Bajer et al., The rational design of coordination-driven supramolecular artificial enzymes: From catalysis to biomedicine, Chem (2025), https://doi.org/10.1016/j.chempr.2025.102784 6 Chem 11, December 11, 2025 Perspective ll OPEN ACCESS at low concentrations, highlighting its therapeutic potential. Like CFTR, MOA 17 provides a synthetic strategy for modulating cellular ion balance in disease contexts. 90 ENZYME-LIKE ACTIVE SITES INCORPORATED INTO MOAs Chemists have developed a range of strategies to integrate enzyme-like active sites into ligands that self-assemble into catalytically active coordination architectures. 92–95 These designs aim to replicate not only the structure but also the functionality of enzyme active sites within MOAs. In addition to creating catalytically competent cavities through self-assembly, active sites can be introduced into ligands either pre-synthetically or post-synthetically, enabling more versatile reactivity. While the cavity itself may not always be catalytically active, its role in preorganizing substrates and stabilizing reactive intermediates significantly enhances overall reaction efficiency. Despite their promise, relatively few MOAs with embedded active sites have been reported to date. Notable examples include systems that mimic the function of redox enzymes, 96,97 oxidases, 22,98 peroxidases, 99 NADH analogs, 100,101 acetyltransferases, 102,103 artificial green fluorescent proteins, 104 and even cytolytic peptides. 95 These systems demonstrate the potential of MOAs to emulate not only catalytic activity but also the complexity of enzyme regulation and selectivity. Su and co-workers recently reported a discrete Os(II)-based metal-organic nanocage (MOA 18) designed for multifunctional cancer theranostics. 105 Mimicking the activity of natural oxidase enzymes, 106 MOA 18 integrates three synergistic therapeutic modalities—radiotherapy (RT), chemotherapy (through encapsulated coumarin), and X-rayinduced photodynamic therapy (X-PDT)—and is particularly effective against triple-negative breast cancer (TNBC), a form resistant to conventional treatment. The nanocage incorporates a near-infrared (NIR)-emitting Os(II) metalloligand for imaging and high-Z Os centers for efficient X-ray sensitization, enabling both diagnosis and therapy within a single nanoplatform. Its host-guest properties allow encapsulation of hydrophobic drug molecules, improving delivery efficiency, therapeutic performance, and biocompatibility while minimizing toxicity (Figure 4A). Building on the concept of redox-active MOAs for biomedical applications, the Duan group presented a Co(III)- based coordination capsule (MOA 19) 100 that mimics oxidoreductase enzymes, utilizing nicotinamide adenine dinucleotide (NADH/NAD + ) co-factors. 107 NADH mimics are integrated into the ligand backbone, while redox-active Co(III) centers serve as electron relays, accepting single electrons from the electrode and enabling two-electron transfer to regenerate active NADH species. The MOA’s inner pocket acts as an enzymatic active site, encapsulating substrates (such as α-keto esters) and stabilizing the intermediates through hydrogen bonding (Figure 4B). This architecture shifts the redox potential anodically by 0.4 V, allowing selective and efficient electrocatalytic hydrogenation. The confined environment promotes Michaelis-Menten-type kinetics, and coupling MOA 19 with horse liver alcohol dehydrogenase, yielded high turnover numbers and reaction rates, underscoring its promise for bio-inspired, electricity-driven synthesis of α-hydroxy and α-amino esters. The Wang group developed a Zr(IV)-based tetrahedral metal-organic cage (MOA 20) for nanozyme-based antioxidant therapy. 108 C D A B N a + O H - M O A 1 4 M O A 1 6 M O A 1 7 L u m i n e s c e n t m e t a l l o h e l i c a t e s f o r i o n t r a n s p o r t U n i m o l e c u l a r i o n t r a n s p o r t c h a n n e l S e l e cti v e i o n t r a n s p o r t a cti v i t y C l - C l - C l - C l - P d ≡ Z n ≡ C l - ≡ N a + ≡ I - ≡ I - , C l - M O A 1 5 S e l e cti v e a n i o n t r a n s m e m b r a n e t r a n s p o r t Figure 3. Examples of artificial biomimetic channels (alkyl chains omitted for clarity) (A) Coordination squares forming short-lived ion transport channels. (B) Synthetic ion channel with a narrow, hydrophobic, and well-defined pore structure, enabling highly selective and efficient iodide transport. (C) Metallohelicates engineered for efficient ion transport and membrane compatibility. (D) Metallacage-based Cl − channel constructed from a naphthalene diimide framework, spanning the phospholipid bilayer for selective Cl − transport. Please cite this article in press as: Bajer et al., The rational design of coordination-driven supramolecular artificial enzymes: From catalysis to biomedicine, Chem (2025), https://doi.org/10.1016/j.chempr.2025.102784 Chem 11, December 11, 2025 7 Perspective ll OPEN ACCESS Constructed from bipyridyl ligands and Cp 2 ZrCl 2 (Cp = η 5 -C 5 H 5 ) via self-assembly, MOA 20 was post-synthetically modified with m-chloroperbenzoic acid to introduce pyridine N-oxide groups, essential for its enzyme-mimicking reactive oxygen species (ROS) scavenging activity. MOA 20 mimics superoxide dismutase (SOD) and catalase (CAT) 109 by sequentially converting superoxide radicals into hydrogen peroxide and then into water and oxygen, producing a cascade antioxidant effect (Figure 4C). In a renal ischemia-reperfusion (I/R) injury model, it significantly reduced oxidative stress and apoptosis, preserving kidney function. This design addresses key limitations of traditional nanozymes, such as undefined structure and weak catalytic activity—while providing a stable, biocompatible, and efficient alternative to natural antioxidant enzymes. Mukherjee and coworkers reported a Pd(II)-based coordination cage (MOA 21) featuring two triangular domes connected by a hexagonal base and functionalized with a benzothiadiazole photosensitizer. 22 Upon white-light irradiation, MOA 21 generates ROS in aqueous solution, imparting oxidase-like activity (Figure 4D). This photocatalyst was effective against methicillin-resistant Staphylococcus aureus (MRSA), demonstrating potential for lightcontrolled antibacterial applications. Cages with endohedrally oriented functional groups allow reactions to occur within a well-defined confined interior, enabling control, selectivity, and M O A 2 5 M O A 2 6 E S O D T M B 2 + H 2 A n t i c a n c e r d r u g t r a n s p o r t X - r a y N I R i m a g i n g R a d i o s e n s i t i z a t i o n a n d p h o t o s e n s i t i z a t i o n X - P D T t h e r a p y O x i d a s e - l i k e a c t i v i t y P r o t o n r e d u c t i o n A C A T M O A 1 8 M O A 2 0 C a s c a d e a n t i o x i d a n t n a n o e n z y m e B M O A 1 9 H 2 O H 2 O I n t r a m o l e c u l a r h y d r i d e t r a n s f e r T M B 2 H + 2 e - C M O A 2 1 M O A 2 2 H2NNH2 H2NNH2 P d S ≡ O s ≡ ≡ Z n ≡ F e ≡ O O O O 3 O 2 1 O 2 O O O OH O O H ∙ O 2 ∙ O 2 H 2 O 2 H 2 O 2 H 2 O N OMe OMe N OMe OMe D N o a c e t a l s o l v o l y s i s A c e t a l s o l v o l y s i s M O A 2 4 M O A 2 3 M e C N M e t h y l b e n z o y l f o r m a t e M e t h y l m a n d e l a t e M e C N T a n d e m c a t a l y s i s M e C N M e C N N O F G I r ≡ C D K 1 S e n s i t i v e t o o x a l i p l a t i n , a n t i c a n c e r a c t i v i t y R e s i s t a n t t o o x a l i p l a t i n , c a n c e r d r u g r e s i s t a n c e S t r u c t u r a l c o m p a t i b i l i t y w i t h C D K 1 N o s t r u c t u r a l c o m p a t i b i l i t y w i t h C D K 1 C D K 1 Figure 4. Examples of enzyme-like active sites incorporated into MOAs (A) Os(II)-based cage integrating NIR-emitting ligands, high-Z centers, and encapsulated drugs into a multifunctional platform for bioimaging, drug delivery, and multimodal cancer therapy. (B) Capsule with embedded NADH mimics and a confined enzymatic pocket that mimics oxidoreductase function to enable selective, efficient electrocatalytic hydrogenation. (C) Zr(IV)-based MOA mimicking natural antioxidant enzymes by sequentially scavenging ROS within a well-defined, stable nanozyme structure. (D) Water-soluble Pd(II) coordination cage incorporating a benzothiadiazole unit for light-triggered ROS generation and oxidase-like activity. (E) Pd(II) cage with a tailored second coordination sphere enhancing proton reduction by encapsulating synthetic hydrogenase models. (F) Fe(II) cage with internal acid groups catalyzes reactions in a proton-rich environment, whereas the unfunctionalized analog binds substrates but is catalytically inactive. (G) MOA 25, a chiral Ir(III) metallohelix, selectively binding and inhibiting CDK1 to enhance cancer therapy, while the non-binding MOA 26 shows no anticancer activity. Please cite this article in press as: Bajer et al., The rational design of coordination-driven supramolecular artificial enzymes: From catalysis to biomedicine, Chem (2025), https://doi.org/10.1016/j.chempr.2025.102784 8 Chem 11, December 11, 2025 Perspective ll OPEN ACCESS tandem transformations that are often incompatible with the bulk environment. However, directing functional groups inward during self-assembly remains challenging. The Reek group addressed this by designing a system inspired by the second coordination sphere of [Fe–Fe]-hydrogenases. Di-iron hydrogenase mimics were incorporated into self-assembled Pd(II) cages (MOA 22), 110 creating a proton-rich nanoenvironment around the active site. Protonated amine-functionalized ligands provided effective proton preorganization, lowering the catalytic overpotential by 250–290 mV and increasing reaction rates by up to two orders of magnitude (Figure 4E). 111 Compared with earlier Fe(II)-based systems, proton reduction occurred at just 350 mV overpotential—a 290 mV improvement—without sacrificing activity. 112 The modular cage architecture also allows precise tuning of functional group density and spatial arrangement, mimicking both the active site and the surrounding protein matrix. 113 Hooley and co-workers designed two Fe(II)-based self-assembled cages, one endohedrally functionalized with carboxylic acid groups (MOA 23) and second one unfunctionalized (MOA 24)—to mimic enzymatic catalysis. 59 A V-shaped dianilinofluorene ligand ensures proper geometry and orientation of acid groups toward the cavity interior, enabling substrate binding and acceleration of acid-catalyzed acetal solvolysis. MOA 23 exhibited up to a 1,000-fold rate enhancement owing to its proton-rich microenvironment, analogous to hydrolase enzymes, whereas MOA 24, lacking acid groups, bound guests but showed no catalytic activity (Figure 4F). MOA 23 also promoted tandem processes, such as acetal deprotection followed by cage-to-cage transformations, under mild conditions that would normally degrade sensitive assemblies. This work underscores the potential of endohedrally functionalized cages for compartmentalized, selective catalysis. 114 Follow-up studies have expanded this concept across diverse catalytic transformations. 115–118 While endohedrally functionalized cages are typically explored as enzyme mimics for catalysis, structurally similar MOAs can serve entirely different functions. The Duan group recently designed flexible, homochiral dinuclear Ir (III) metallohelices using a programmable modular assembly (PMA) strategy, allowing precise control over conformation and chirality. 119 Incorporating dithiourea linkers and chiral spacers, these metallohelices were optimized for binding to protein pocket of cyclin-dependent kinase 1 (CDK1), a key enzyme in cancer drug resistance. 120 Among them, MOA 25 showed strong CDK1 inhibition and enhanced PDT efficacy, inducing oncosis-mediated cell death in oxaliplatin-resistant colorectal cancer cells. In contrast, MOA 26 showed no CDK1 binding and no inhibition of tumor proliferation. This comparison highlights the critical role of chirality in targeting CDK1 and demonstrates how supramolecular design principles developed for catalysis can be redirected toward biological inhibition, offering a pathway to next-generation metallodrugs capable of overcoming drug resistance. ENCAPSULATION OF ENZYME-LIKE ACTIVE CATALYSTS WITHIN MOAs Supramolecular coordination assemblies have emerged as powerful tools for mimicking the sophisticated functions of natural enzymes, which often achieve high catalytic efficiency and selectivity through spatially defined active sites and dynamic microenvironments. 121 Prominent examples include hydrogenases, which catalyze proton reduction 112 ; oxidases involved in selective oxygenation 122 ; and C–H activation enzymes such as cytochrome P450s, which perform regioselective and stereoselective transformations. 47 Inspired by these systems, MOAs have been developed to encapsulate enzyme-like catalysts within confined and often chiral cavities, recreating key features of enzymatic catalysis in a synthetic context. Such encapsulation allows for precise control over substrate access, orientation, and reactivity. These enzyme-mimetic environments enhance catalytic performance by increasing local substrate concentration, preventing catalyst deactivation, and stabilizing highenergy transition states. As a result, MOA-based systems can introduce regioselectivity, stereoselectivity, and charge selectivity typically unachievable in traditional homogeneous catalysis. One of the earliest demonstrations came from the Raymond group, who encapsulated a half-sandwich Ir(III) complex within a chiral Ga(III)-based tetrahedral cage (MOA 8). 123 This system retained the C–H activation ability of the free complex and catalyzed reactions with aldehydes, yielding diastereomeric products in ratios ranging from 55:45 to 70:30 (Figure 5A). The cage imparted outer-sphere effects such as size, shape, and electrostatic selectivity—illustrating how encapsulation can mimic the secondary coordination sphere of enzymes. 41 Its tetrahedral framework, constructed from Ga(III) ions and catecholamide ligands around a naphthalene-based scaffold, forms a water-soluble, highly anionic, rigid chiral cavity (∼0.5 nm 3 ) that selectively encapsulates monocationic hydrophobic guests. This confined space enables diastereoselective C–H bond activation with a level of control unattainable in bulk solution. More recently, the Mukherjee group demonstrated that encapsulation can also enhance oxidase-like activity. A benzothiadiazole dye encapsulated in a Pd(II) molecular barrel (MOA 27) generated significantly more ROS under light irradiation than the free dye. 36 Host-guest interactions within the cage stabilized the dye and facilitated photoinduced electron transfer (PET) to molecular oxygen (Figure 5B). Only the encapsulated dye, not the free dye, was able to catalytically detect 5-hydroxytryptophan, highlighting the cage’s ability to amplify function. MOA 27’s cuboid structure, with open rhombic apertures and a spacious internal cavity, prevented dye aggregation and nonradiative decay, representing the first example of enhanced oxidase-like catalysis via dye encapsulation in a water-soluble MOA. Duan and co-workers designed a urea-functionalized Fe(II) cage (MOA 28) to co-encapsulate a sulfonated gold catalyst and a urea-functionalized substrate via hydrogen bonding. 124 Urea groups at each vertex provided strong, selective binding, enabling π-activation-based catalysis independent of catalyst concentration (Figure 5C). Compared with the free catalyst, encapsulation significantly improved turnover frequency and yield, with enhancement attributed to transition-state stabilization rather than substrate preorganization. The modular design allows incorporation of any sulfonate-bearing catalyst, mimicking the compartmentalized reactivity of enzymes. 41 The same group also constructed an oxidoreductase mimic by encapsulating an Ir(III) photosensitizer within a tetrahedral Please cite this article in press as: Bajer et al., The rational design of coordination-driven supramolecular artificial enzymes: From catalysis to biomedicine, Chem (2025), https://doi.org/10.1016/j.chempr.2025.102784 Chem 11, December 11, 2025 9 Perspective ll OPEN ACCESS