Metal coordination at biological interfaces: Mechanisms, therapeutics, and translation

Abstract

Metal ions are integral to biological systems, serving as catalytic cofactors, structural components, and mediators of key cellular signaling pathways. Understanding their interactions with bioactive molecules at biointerfaces provides valuable insights into the development of metal-based drugs, therapeutic biomaterials, and innovative nanomedicines. This review adopts a bio-interface-centered framework that integrates coordination chemistry, surface chemistry, and biomolecular recognition, underscoring how coordination geometry, ligand kinetics, redox behavior, and hard/soft acid/base preferences affect the interactions of metal complexes with DNA, proteins, lipids, and glycoconjugates. These interactions regulate diverse cellular processes, such as mitochondrial function, signal transduction cascades, and programmed cell death mechanisms, including apoptosis and ferroptosis. This manuscript outlines the major classes of clinically significant metal complexes, along with emerging systems, including metal-organic frameworks, polymeric carriers, and artificial metalloenzymes, designed to provide targeted delivery and catalytic activity at biological interfaces. In particular, this review focuses on the combination of coordination chemistry, nanotechnology, and chemical biology to develop stimuli-responsive frameworks capable of targeted delivery and multifunctional therapeutic effects. This review further examines the role of metal coordination in biomolecular recognition, biointerface regulation, pharmacokinetic behavior, toxicity, and therapeutic performance, thereby linking atomic-level coordination chemistry with cellular and translational outcomes. This bio-interface-centered perspective distinguishes the present review from prior work focused separately on metallodrugs, bioinorganic mechanisms, or coordination chemistry.

Graphical abstract

1. Introduction

Metal ions play essential roles in biointerfaces, serving as catalysts, structural supports, and dynamic signaling mediators. Their significance as cofactors for the activation and function of enzymes further underscores the fundamental role of metal surface interactions in living systems [1,2]. The widespread applications of these ions reflect the chemical versatility of metal-ligand interactions: small cations, such as Na⁺, K⁺, and Ca²⁺, maintain ionic homeostasis and electrical excitability, whereas transition metals (for example, Fe, Cu, Mn, and Zn) impart various structural and catalytic capabilities to proteins and nucleic acids [[3], [4], [5]]. A well-established example is Zn²⁺-dependent catalysis, in which the metal center activates a coordinated water molecule, generating a stronger nucleophile that can promote peptide bond hydrolysis [3,6]. Similarly, redox-active metals with multiple oxidation states form the backbone of electron transfer chains, such as cytochromes, Fe–sulfur clusters, and Cu proteins, which drive respiration and redox metabolism [3,[7], [8], [9], [10]].

Beyond catalysis, metals are involved in regulation, serving as reversible switches that link binding events to conformational changes and downstream signaling. Ca²⁺ and Zn²⁺, for example, modulate protein and nucleic-acid functions within cells, thereby regulating cellular pathways [3]. Transcription factors involved in the regulation of metal ions further support this principle by directly sensing intracellular metal availability and modulating gene expression accordingly, demonstrating their roles as active participants in signaling rather than passive cofactors [3,11]. Overall, metal coordination provides a flexible mechanism for biological control, refined over billions of years, from enzymatic chemistry to genome regulation [1,6,12,13].

1.1. Bio-interface: where coordination chemistry intersects biomolecular function

The biological interface is a dynamic frontier where coordination chemistry intersects with the complex organization of biomolecules and cells (Fig. 1). At the biological interface, synthetic or inorganic metal centers interact directly with biological scaffolds, including proteins, nucleic acids, and membranes, and modulate their molecular structure and cellular function [15,16]. Because these interactions span molecular and cellular levels, the field is inherently interdisciplinary. In medicine, the biological interface encompasses metallodrugs and metal-based diagnostic tools. Coordination complexes, ranging from platinum (Pt) and ruthenium (Ru) compounds to gadolinium chelates, can bind to specific biomolecular targets or amplify imaging signals, allowing coordination chemistry to be applied in therapy and disease diagnosis [17,18]. Biocatalysis includes artificial metalloenzymes and nanozymes, where engineered metal centers are incorporated into protein scaffolds or nanomaterials to drive chemical transformations under physiologically compatible conditions [19,20]. In cellular engineering, biological interface strategies are being extended to metal-responsive genetic circuits, intracellular bio-orthogonal metal catalysis, and engineered systems that modulate the uptake, cellular transport, or availability of metal cofactors [21].

Fig. 1.

A unified multiscale framework linking metal center properties and ligand architecture → coordination environment → biomolecular recognition → cellular regulation → pharmacokinetic behavior → toxicological outcomes → therapeutic efficacy. The framework explicitly distinguishes between metal centers, structural coordination environments, catalytic sites, biomolecular targets, and biological functions. The metal-center properties and ligand architecture govern the coordination behavior, including the geometry, redox activity, ligand-exchange kinetics, and molecular selectivity. These features determine biomolecular recognition at biological interfaces, including interactions with DNA, proteins, membranes, glycans, and the ECM. Such interactions subsequently influence cellular transport, catalytic processes, signaling pathways, and biological responses, ultimately affecting pharmacokinetics, toxicological outcomes, therapeutic efficacy, and clinical translation of the drug. This framework provides a unified perspective connecting atomic-level coordination chemistry with biomolecular recognition, cellular regulation, and therapeutic performance, highlighting how metal coordination regulates recognition, delivery, catalysis, and therapeutic responses across multiple biological scales. The illustration is a conceptual summary synthesized from published studies on metal–biomolecule interactions, coordination chemistry, metalloproteins, and metal-based therapeutics [3,14].

Among these applications, metal coordination confers new functionalities. For example, metal-organic frameworks (MOFs) can be coupled with enzymes to form hybrid catalytic platforms, and peptide assemblies can be tuned through selective metal-ion binding to adjust their stability and structures [22]. Notably, metals at the biological interface may be native components of biology (e.g., Zn²⁺ in zinc-finger DNA-binding domain) or deliberately introduced non-native species (e.g., intracellular Ru complexes), thereby expanding the chemical repertoire available to living systems’ beyond what evolution typically employs [3,23]. Fundamentally, a biological interface is where design principles, such as ligand geometry, redox control, and coordination strength, intersect with biological frameworks, including macromolecular structure, cellular dynamics, and signaling [17,24,25]. This intersection results in a broad design landscape in which metal coordination motifs can regulate enzyme activity, drive metal-mediated assembly of biomaterials, and shape metal ion circulation within signaling pathways. Living cells exploit the distinctive chemistry of metals to maintain homeostasis and coordinate metabolism. Hence, materials chemists are now developing “life-like” catalysts and therapeutics that mimic or manipulate these bioinorganic mechanisms [9,[26], [27], [28], [29]]. Thus, the biological interface framework highlights the strong coupling between coordination chemistry and biological performance, with higher relevance to medicine, biocatalysis, and cellular engineering [6].

1.2. Current gaps and challenges at the metal bio-interface

Despite its potential, metal-focused biology and medicine have significant gaps that hinder progress. In particular, a major limitation is the lack of a unified, multiscale mechanistic framework that can connect atomic-level coordination to clinical outcomes [30]. Coordination chemists often describe metal–ligand interactions under controlled in vitro conditions, structural biologists explain isolated metalloprotein structures, and clinicians assess organism-level outcomes; however, these layers are rarely integrated into a single explanatory model [31]. Consequently, we still lack robust models linking a metal's local binding geometry to downstream consequences, such as changes in protein conformation, cellular trafficking, and whole-organism pharmacodynamics [24,[32], [33], [34]]. This disconnect is particularly evident in clinical translation, where ligand frameworks may be optimized for affinity or reactivity without parallel consideration of pharmacokinetics and safety, meaning that even well-designed complexes can fail once tested in vivo. The biological fate of metallodrugs is particularly challenging to predict because their efficacy and toxicity are determined not only by their chemical structure, but also by absorption, distribution, metabolism, and excretion (ADME), as well as metal-specific transporter interactions and the synthesis of distinctive metal-derived metabolites [25,[35], [36], [37]].

Current metal-based therapies also face intrinsic limitations, most notably limited selectivity and dose-limiting side effects. For example, Pt drugs, such as cisplatin, are potent anticancer agents; however, their non-specific interactions, such as widespread DNA binding, lead to nephrotoxicity, neurotoxicity, and ototoxicity [11,35,38]. Similar concerns arise with many investigational complexes, where unintended interactions with normal biomolecules or uncontrolled release of reactive species (e.g., Cu²⁺ or Fe²⁺) can trigger collateral injury. These limitations underline the need for safer coordination motifs and more targeted delivery approaches [39]. Overall, coordination chemistry, structural biology, and clinical studies remain largely siloed, lacking a unified set of rules that translate ligand field features or redox states into predictable effects on protein structure, cellular pathways, and therapeutic performance [12,40]. For example, it is not yet fully understood how the precise coordination geometry of a drug governs transporter-mediated uptake, endosomal release, or nuclear accumulation, and how these processes modulate signaling pathways. Without an integrated perspective, designs remain incomplete, and therapeutic translation often proceeds by trial and error [41].

Efforts to enhance the potency of metal complexes often undermine their translation. A metal-centered inhibitor may perform well against an enzyme in vitro but may prove ineffective in vivo if it is rapidly sequestered by serum proteins or eliminated before reaching its target tissue [[42], [43], [44]]. As emphasized earlier, “metallodrugs have more complex mechanisms of metabolism and chemical transformation in the body compared to traditional small molecules, and these processes can be more critical for clinical performance [35,45]. Bridging this gap requires incorporating pharmacokinetics and toxicity into the early stages of design and linking coordination-level measurements to realistic models of cellular uptake, distribution, and clearance. However, many metal compounds have narrow therapeutic windows because their intrinsic reactivity is difficult to confine to diseased cells [46]. Cisplatin is the striking example; its nonselective DNA binding yields potent anti-tumor efficacy but also leads to dose-limiting nephrotoxicity, neurotoxicity and ototoxicity [11,12,35]. Furthermore, certain emerging metal-based agents often generate reactive metabolites or interact with endogenous biomolecules, which reduces their safety and hinders predictability. Therefore, these limitations highlight the need for the field to evolve beyond “serendipitous” metal drug discoveries toward rationally engineered complexes that balance therapeutic efficacy with safety and control [47].

Another important limitation is the insufficient integration of dynamic biointerface phenomena into current design frameworks. Processes e.g., protein-corona formation, metal-ion release, extracellular matrix (ECM) remodeling, interfacial degradation, and immune recognition are increasingly recognized as major determinants of biological performance; however, they remain difficult to predict from conventional coordination-chemical descriptors alone. Bridging this gap requires multiscale models that connect atomic-level coordination behavior with the evolving physicochemical and biological properties of material interfaces [48,49].

1.3. Scope of this review

This review establishes a mechanistically unified and clinically relevant framework for understanding how metal coordination regulates biomolecular recognition, transport, catalytic activity and therapeutic responses at biological interfaces. Throughout this review, metal centers, coordination environments, catalytic sites, biomolecular targets, and biological functions are treated as distinct but interconnected levels of organization to avoid conceptual ambiguities. Rather than treating metallodrugs, biomaterials, catalytic systems, diagnostics, and emerging biointerface technologies as independent topics, these areas are discussed as interconnected manifestations of coordination-mediated biointerface regulation. To achieve this objective, we examined metal–biointerface interactions across multiple biological scales, linking atomic-level coordination chemistry to cellular behavior and clinical outcomes. Particular emphasis is placed on how metal-center properties and ligand architecture, including hard–soft acid–base (HSAB) preferences, coordination geometry, ligand exchange kinetics, and redox activity, govern interactions with biomolecular targets and determine biological specificity. We further discuss how the biological context, including binding-site architecture, transport and trafficking pathways, cellular redox buffering, and microenvironmental factors, modulates metal reactivity and therapeutic performance. This review highlights the key challenges in the field, including limited multiscale integration, the disconnect between molecular design and pharmacological behavior, and dose-limiting toxicities that constrain many current metal-based therapeutics. Emerging strategies, including targeted delivery systems, stimuli-responsive coordination platforms, artificial metalloproteins, and emerging coordination-engineered systems, are discussed in the context of improving selectivity, safety, and therapeutic efficacy. By integrating concepts from coordination chemistry, chemical biology, and bioinorganic medicine, this review provides guiding principles for the rational design and clinical translation of next-generation coordination-based therapeutics, diagnostics, and biomaterials across oncology, regenerative medicine, immunotherapy, and bioelectronic applications.

To connect the diverse physicochemical, molecular, cellular, pharmacokinetic, toxicological, and clinical dimensions of metal-ion biology, a unified multiscale framework is proposed that links metal-center properties and ligand architecture to biomolecular recognition, cellular regulation, pharmacokinetic behavior, toxicological outcomes, and therapeutic efficacy (Fig. 1). This framework serves as a conceptual roadmap for understanding how atomic-level coordination chemistry propagates across biological scales to influence the biological function and clinical performance. Collectively, the concepts discussed in this review highlight how metal coordination regulates recognition, transport, catalytic activity, and therapeutic responses at biological interfaces, providing a foundation for the rational design and clinical translation of next-generation metallotherapeutics.

2. Molecular foundations of metal coordination at the bio-interface

The rational development of metal-based drugs depends on a thorough understanding of the principles of coordination chemistry. This section outlines how HSAB preferences, ligand exchange kinetics, and coordination geometry provide valuable insights into the interpretation of metal behavior at biological interfaces, where metals encounter heterogeneous physicochemical environments defined by proteins, membranes, extracellular matrices, and cellular compartments. These principles govern the target selectivity, intracellular stability, and chemical reactivity of metal complexes inside living systems, while simultaneously determining how metal species interact with the spatial and dynamic characteristics of biological interfaces [50].

2.1. Hard and soft acids and bases in biological context

The HSAB theory is a fundamental concept that predicts the preferred metal-ligand pairing based on polarizability and charge density [13,51,52]. From a biological perspective, many metal centers explicitly follow these preferences for selecting potential binding targets at biological interfaces, where donor-atom availability, hydration structure, molecular crowding, and compartment-specific chemistry collectively influence coordination behavior (Fig. 2). Hard acids, defined by high charge density and low polarizability, such as Fe³⁺, Al³⁺, and Ca²⁺, tend to bind to hard bases, including ligands with electronegative donor atoms, most commonly oxygen donors, such as carboxylates or water [3]. Consistent with this, Fe³⁺, with its high charge-to-radius ratio, usually forms particularly stable complexes with O-donor ligands [3]. In contrast, soft acids are generally larger and more polarizable, with lower charge densities (Pt²⁺, Au⁺, and Hg ²⁺) that preferentially bind soft bases, such as sulfur-containing donors in thiolates [14]. Metals with intermediate and borderline characteristics, such as Zn²⁺, Cu²⁺, or Fe²⁺, commonly exhibit a significant affinity for nitrogen donors (notably histidines) and oxygen donors, depending on the local environment [3]. Overall, the HSAB concept helps rationalize why metalloproteins utilize specific amino acid chains for specific metals and why metal-based drugs show particular reactivity patterns. Importantly, at biological interfaces, these preferences are not governed solely by intrinsic metal–ligand affinity, but also by local interface properties, including ligand density, solvent accessibility, electrostatic gradients, and competing biomolecular interactions. For example, thiophilic agents are sequestered by glutathione (GSH) or cysteine residues [53,54].

Fig. 2.

Molecular foundations of metal coordination at the biointerface. Schematic illustration of the key coordination-chemical principles that regulate metal–biomolecule interactions, including HSAB logic, ligand-exchange kinetics, coordination geometry, and competition with endogenous small-molecule and macromolecular ligands. These mechanisms define the structures and reactivities of metal coordination cores in biological environments. This figure is a conceptual illustration synthesized from published studies on coordination chemistry, the HSAB theory, ligand exchange kinetics, and metallodrug design [13,14].

Therefore, hard acids such as Fe³⁺and Ca²⁺are commonly coordinated by hard bases, notably the carboxylate groups of aspartate or glutamate, in many metalloproteins [3]. Within biological interfaces, these interactions are further shaped by the local organization of proteins, glycoconjugates, ECM components, and membrane-associated ligands, which regulate metal accessibility and retention. This helps explain why Ca often stabilizes protein structures via carboxylate contacts and why Fe³⁺ is sequestered by hard oxygen donor ligands, such as citrate or phosphate, in cells [55]. Borderline metals, including Cu²⁺, Zn²⁺, and Fe²⁺, generally prefer intermediate donors, such as imidazole N from histidine in enzyme active sites and, in some contexts, oxygen donors; they often achieve stabilizing coordination environments (commonly octahedral) depending on the protein structure [3]. Finally, soft acids (for example, Pt²⁺, Au⁺) form strong interactions with sulfur donors, especially cysteine thiols in proteins. Therefore, Pt drugs are inactivated by cysteine-rich metallothioneins, and Au drugs (auranofin) target thiol-dependent enzymes. For example, cisplatin contains a relatively soft Pt²⁺ center that targets the N7 sites of purine bases in DNA, but it can also be readily intercepted by thiolate donors from GSH or cysteine residues, consistent with the HSAB-driven thiophilicity [14]. Understanding these preferences enables the more predictive design of metal complexes in biological settings and provides a framework for engineering metal behavior at biological interfaces, where the selective recognition of interface-specific ligands can improve targeting efficiency, biological compatibility, and therapeutic selectivity. For example, by tuning metal-ligand “softness” to reduce strong interactions with ubiquitous GSH and thereby enhance bioavailability and selectivity [14,56,57]. Understanding these HSAB-driven preferences provides a foundational principle for engineering metal coordination at biological interfaces, where selective recognition of interface-specific donor groups is the first determinant of targeting efficiency and therapeutic selectivity.

2.2. Ligand-exchange kinetics in vivo

Kinetics plays a crucial role in the behavior of metal complexes in vivo, as biological interfaces continuously expose metal complexes to dynamic ligand-exchange environments involving proteins, membranes, extracellular matrices, metabolites, and intracellular trafficking compartments. Biological fluids are rich in competing ligands (water, chloride, thiols, etc.); therefore, the rate of ligand exchange often determines whether a complex reaches its intended target or is derailed. These interface-associated ligands create highly competitive coordination environments that strongly influence metal speciation and biological fate [33,34,58,59]. Kinetically labile complexes with fast ligand exchange can readily react with off-target species and can be quickly deactivated before reaching their primary sites of action. They may undergo rapid ligand substitution at the plasma, membrane, or intracellular interfaces, resulting in altered biodistribution, premature deactivation, and unintended interactions with surrounding biomolecules (Fig. 3). For instance, many d¹⁰ metal complexes and certain square-planar d⁸ systems (e.g., some Pd²⁺ or more labile Pt²⁺ compounds) undergo rapid aquation or thiol binding, which can lead to non-selective reactivity inside cells. Conversely, kinetically inert complexes with slow exchange rates can persist longer in the biological milieu and act as prodrugs that are activated only under specific intracellular or environmental conditions [33,58,60,61,63].

Fig. 3.

Ligand exchange kinetics of metal complexes in vivo. A comparison of kinetically labile and inert metal complexes in the bloodstream and intracellular environment illustrates how premature ligand exchange leads to deactivation and off-target reactivity, whereas kinetically inert or prodrug strategies enable controlled activation, DNA binding, and selective cytotoxicity within cells. This illustration was synthesized from published studies on metallodrug pharmacology, platinum-based drugs, ruthenium complexes, and prodrug activation mechanisms [11,[60], [61], [62]].

Classic examples of kinetically inert systems include low-spin d⁶ octahedral complexes, such as Ru²⁺ polypyridyl complexes (RuPOPs) or “piano-stool” arene complexes, and many Rh³⁺ or Ir³⁺ compounds, which show extremely slow ligand substitution. Water ligands in hexacoordinated Rh³⁺ or Ir³⁺ complexes can have residence times on the order of 10⁸–10¹⁰ s, effectively locking the coordination sphere until a specific trigger (e.g., reduction or light) prompts ligand dissociation [14]. Drug design strategies increasingly exploit these characteristics. One recent Pt-based study explicitly emphasized the “strong interconnection between kinetic lability and undesired shortcomings of clinical Pt drugs, motivating the development of more inert analogs that retain antitumor efficacy while reducing side effects [63].

Furthermore, cisplatin is a square-planar Pt²⁺ complex with intermediate lability, which remains relatively intact in the bloodstream for several hours due to the high chloride concentration but associates with water inside cells. These characteristics support its interaction with DNA, ultimately facilitating cytotoxicity in the cells. However, the same lability enables off-target quenching by cellular thiols [6,14]. In contrast, Pt⁴⁺ prodrugs (octahedral, kinetically inert) are stable during circulation and are reduced to active Pt²⁺ only within a reductive intracellular environment. Likewise, many Ru²⁺ and Os²⁺ complexes are essentially inert in the dark and require photoactivation (phototriggered ligand release) to become cytotoxic. Therefore, understanding kinetic lability is central to balancing the stability and reactivity profile of metallodrugs across multiple biological interfaces, including circulation, cellular uptake pathways, endosomal compartments, and intracellular target sites. A labile complex can interact with targets rapidly but is more susceptible to side reactions and premature deactivation. An inert complex can avoid early inactivation but may require a reliable activation mechanism at the disease site [63]. Therefore, understanding kinetic lability is central to balancing the stability and reactivity profile of metallodrugs across the full spectrum of biological interfaces — from circulating plasma to intracellular target sites — and directly determines whether metal coordination leads to productive recognition and therapeutic response or off-target deactivation.

2.3. Coordination geometry and biological consequences

Coordination geometry determines how a metal complex spatially interacts with biological interfaces, including membrane surfaces, protein binding pockets, extracellular matrices, and nucleic-acid architectures — and this spatial interaction is the primary mechanism by which geometry governs recognition, binding orientation, and overall therapeutic activity. Metal ions adopt coordination geometries that are largely absent from conventional organic drugs and are generally limited to relatively simple linear or planar small structures (Table 1). The three-dimensional (3D) geometry, such as octahedral, square planar, and tetrahedral, profoundly affects the electronic properties of a metal complex and determines how the complex spatially interacts with biological interfaces, including membrane surfaces, protein binding pockets, extracellular matrices, and nucleic-acid architectures. This disrupts interactions with biomolecular targets [[89], [90], [91]]. Furthermore, the geometry of metal complexes can influence cellular uptake, binding orientation, and overall activity. Many biologically relevant metal ions, such as Fe²⁺, Co³⁺, Ru²⁺, and Pt⁴⁺, preferentially adopt octahedral coordination geometries. Low-spin d⁶ centers, including Fe²⁺, gain significant crystal-field stabilization, which contributes to the robustness of these complexes [92,93].

Table 1.

Metal coordination interactions at the biointerface.

Metal center	Biomolecular target	Biological/Chemical process	Therapeutic outcome	Translational outcome	Ref.
Platinum	DNA, proteins	DNA crosslinking	Chemotherapy	Resistance, toxicity	[64,65]
Gold	Thiol proteins, TrxR	Enzyme inhibition	Anticancer, anti-inflammatory	Off-target binding	[66,67]
Ruthenium	DNA, histones	Ligand exchange, redox activation	Precision oncology	Limited clinical translation	[68,69]
Iron	Oxygen carriers, ROS networks	Redox cycling	Catalytic therapy	Oxidative toxicity	[[70], [71], [72]]
Copper	Redox proteins, ECM enzymes	Electron transfer, oxidation	Catalytic therapeutics	Redox imbalance	[73,74]
Zinc	Enzymes, transcription factors	Structural coordination	Signaling regulation	Limited redox activity	[[75], [76], [77], [78]]
Manganese	Antioxidant enzymes	ROS scavenging	Nanozymes	Accumulation concerns	[79]
Lanthanides	Water coordination sites	Chelation-mediated contrast	MRI imaging	Retention concerns	[80]
Ni/Co/Mo/V	Metalloenzymes	Cofactor catalysis	Metabolic regulation	Toxicity concerns	[[81], [82], [83], [84], [85], [86]]
erium oxide	ROS networks	Redox cycling	Antioxidant nanozymes	Biodistribution	[87]
Titanium	ECM, bone	Surface mineralization	Implants	Long-term performance	[88]

Octahedral coordination also allows the development of ligand cages around the metal. Pt⁴⁺, for example, is octahedral and relatively inert, carrying axial ligands, such as chlorides or acetates, which can be functionalized with targeting groups or co-drugs (Fig. 4). Upon reduction to square-planar Pt²⁺ inside the cell, these axial ligands are released, and the complex is converted into an active DNA crosslinking agent [97]. The 3D layout of octahedral complexes further enables them to engage multiple binding sites simultaneously across structurally heterogeneous biological interfaces, facilitating multivalent interactions with proteins, membranes, and nucleic acids. For instance, a RuPOP may intercalate an aromatic ligand into DNA, while another ligand coordinates a protein residue, achieving multivalent interactions [98,99]. The rigid octahedral scaffold can also promote the selection of biomolecules through steric complementarity with binding pockets that require a particular spatial arrangement. Reflecting this, octahedral half-sandwich Ru²⁺ arene complexes are being developed as enzyme inhibitors, exploiting their 3D geometry to access binding clefts that are difficult for flat organic molecules to reach [3].

Fig. 4.

Influence of coordination geometry on biomolecular recognition and biological function at the biointerface. Representative coordination geometries influence the accessibility, binding orientation, and recognition preferences of DNA, proteins, and other biomolecular interfaces. These geometry-dependent interactions contribute to distinct biological outcomes, including DNA crosslinking, protein regulation, catalysis, redox control, and structural stabilization. Coordination geometries are presented as structural determinants of molecular recognition rather than as biological functions. The framework highlights coordination geometry as a key design parameter linking metal-complex structures to biomolecular recognition, biological activity, and therapeutic function at the biointerface. This figure is a conceptual illustration synthesized from published studies on coordination geometry, structure–activity relationships, and metal-based therapeutics [89,[94], [95], [96]].

Square-planar (4-coordinate) geometry is characteristic of many d⁸ centers, including Pt²⁺, Pd²⁺, and low-spin Ni²⁺, which define two orthogonal coordination axes (Table 2). In cisplatin, the cis arrangement of ammine ligands and labile chlorides enforce the formation of intrastrand DNA crosslinks caused by adjacent guanine base interactions [125]. The planar structures also allow these complexes to “slip” between base pairs (intercalation) and efficiently access confined interfacial environments, such as DNA grooves and protein–DNA contact regions, or occupy relatively flat binding regions on proteins. Square-planar Pt²⁺ centers tend to undergo associative ligand substitution, temporarily expanding to a five-coordinate intermediate, which facilitates direct coordination with biomolecules. This geometry underlines the unique DNA kinking and unwinding produced by cisplatin adducts that block replication forks. Notably, small geometric changes can alter biological outcomes: trans-configured Pt²⁺ complexes interact with DNA in a different manner and show much lower anticancer activity, whereas the cis geometry is crucial for clinically relevant 1,2-intrastrand crosslinks [11,[60], [61], [62]].

Table 2.

Modes of metal coordination at biomolecular interfaces and their functional consequences.

Metal center	Coordination Site/Donor Environment	Primary Biomolecular Target	Coordination Mechanism	Immediate Molecular Consequence	Cellular Outcome	Therapeutic Implication	Ref.
Pt(II)	N7 (guanine)	DNA	1,2-intrastrand crosslink	Helical bending (∼30°)	Replication arrest	Cytotoxic chemotherapy	[95,96]
Pt(II)	Cys thiolate	GSH	Thiolate substitution	Drug inactivation	Resistance development	Need for thiol-resistant ligands	[95,96]
Ru(II) polypyridyl	N7 + π stacking	DNA	Monofunctional + intercalation	DNA unwinding	Transcription inhibition	Alternative to Pt drugs	[94]
Rh(III)	Mismatch base	Damaged DNA	Metalloinsertion	Selective lesion binding	Synthetic lethality	Targeting MMR-deficient tumors	[100,101]
Au(I)	Cys/Sec	Thioredoxin reductase catalytic site	Covalent coordination	Redox enzyme inhibition	ROS accumulation	Anticancer & anti-inflammatory	[102]
Cu(II)	Histidine	Metal-dependent enzymes	Metal displacement	Enzyme inhibition	Metabolic stress	Proteostasis targeting	[103]
Fe(II)	Phosphate oxygen	DNA backbone	Electrostatic coordination	Local condensation	Chromatin modulation	Epigenetic intervention	[103]
Ga(III)	Transferrin O-donors	Iron transport proteins	Fe mimicry	Iron deprivation	Growth inhibition	Antimicrobial/anticancer	[104,105]
Zn(II)	His/Cys	Zinc-finger structural sites	Metal displacement	Transcription factor inhibition	Gene regulation shift	Antiviral strategy	[106]
Co(III)	Protein His	Hypoxia-associated protein targets	Redox-activated release	Ligand liberation	Hypoxia-selective toxicity	Tumor targeting	[107]
Ir(III)	ETC-associated protein complex	ETC complexes	Protein coordination	ΔΨm collapse	Apoptosis	Mitochondrial-targeted therapy	[108]
Au(III)	Cys clusters	Metallothionein metal-binding sites	Thiolate cluster formation	Metal sequestration	Detoxification	PK modulation	[109]
Ni(II)	Histidine clusters	His-tag proteins	Multidentate binding	Enzyme inhibition	Proteome stress	Selective inhibition	[110,111]
Cu(I)	GSH	Intracellular thiols	Redox cycling	ROS generation	Ferroptosis	Redox therapy	[112,113]
Fe(III)	Catechol ligands	Siderophores	Chelation	Metal transport hijack	Bacterial starvation	Antimicrobial design	[114,115]
Ru(II) arene	Protein pockets	Enzyme clefts	Multivalent binding	Allosteric modulation	Signaling suppression	Kinase targeting	[116]
Pt(IV)	DNA (post-reduction)	Genomic DNA	Reductive activation	Pt(II) release	Site-selective cytotoxicity	Prodrug design	[117,118]
Os(II)	Protein backbone	Kinase regulatory sites	Photolabile coordination	Light-triggered inhibition	Spatiotemporal control	Phototherapy	[119,120]
Lanthanide(III)	Phosphate	ATP-rich regions	Hard-hard interaction	Subcellular retention	Organelle targeting	Imaging + therapy	[121,122]
Ferrocene (FeII)	Lipid bilayer	Membrane	Redox-active insertion	Lipid oxidation	Apoptosis	Redox-responsive drugs	[123,124]

Note: Metal centers, coordination sites, biomolecular targets, coordination mechanisms, and biological outcomes are presented as distinct but interlinked organizational levels. Coordination sites describe the immediate metal-binding environment, whereas cellular outcomes and therapeutic implications represent the downstream biological consequences of metal–biomolecule interactions.

Tetrahedral coordination often occurs for d¹⁰ ions, such as Zn²⁺ and Cd²⁺, as well as for some high-spin d⁵ or d³ metal centers, and usually yields flexible, relatively labile binding. For example, in Zn²⁺ enzymes, the metal is most often found in a tetrahedral environment with histidines and cysteines, plus water or hydroxide, and has essentially zero-ligand field stabilization energy. This lack of CFSE renders the Zn site kinetically labile, which is crucial for catalysis because substrates and products must interact and dissociate efficiently [126,127]. Generally, tetrahedral complexes are often more fluxional and therefore exhibit enhanced adaptability when interacting with dynamic biological interfaces, particularly membrane-associated and protein-rich environments, which can facilitate membrane transport through partial dissociation and reassociation; however, it also reduces target specificity. In cells, tetrahedral Zn²⁺ complexes require dedicated protein carriers and chaperones to prevent uncontrolled interactions. In the context of drugs, compact tetrahedral shapes, such as organogold⁺ or Cu⁺, can readily penetrate cell membranes; however, their high reactivity must be carefully managed to prevent off-target damage [39,128].

Common coordination geometries in biology and metal-based drugs are summarized below.

• •Octahedral (coordination number 6): This geometry is commonly observed for Fe²⁺, Ru²⁺, and Pt⁴⁺. Low-spin d⁶ ions (for example, Fe²⁺) gain significant LFSE [3], which favors chelating ligands that enforce 6-coordinate structures. Octahedral drugs, such as Pt⁴⁺ prodrugs or Ru²⁺ arene complexes, utilize this 3D geometry to achieve plasma stability and simultaneously contact multiple binding elements at their targets [129].
• •Square planar (coordination number 4): This is a characteristic geometry of d⁸ metals (Pt²⁺, Pd²⁺). The planar geometry results in a distinct electronic structure and a reactivity profile. The square-planar Pt center of cisplatin is particularly effective in crosslinking DNA. These complexes often undergo an associative mechanism, transiently opening a fifth coordination site and forming strong 2D ligand fields that drive their substitution chemistry [130].
• •Tetrahedral (coordination number 4): This structure is typical of Zn²⁺ (d¹⁰) and some high-spin d-block metals. The lack of significant LFSE generally leads to labile, conformationally flexible complexes. Many Zn-dependent enzymes, such as carbonic anhydrase, depend on tetrahedral Zn sites that can readily interchange ligands. Tetrahedral shape also influences membrane transport: larger, hydrophilic Zn complexes usually require specific transporters or chaperones to move safely across membranes and within cells [127,131].

Additional geometries appear in metallodrugs, including trigonal-bipyramidal and square-pyramidal intermediates during ligand substitution and sandwich complexes, such as metallocenes (e.g., ferrocene), used in redox-active drugs [89]. Each geometry confers unique steric and electronic properties to the complexes. For example, ferrocene, which consists of an Fe²⁺ center between two cyclopentadienyl rings, has been incorporated into estrogen-like molecules to create redox-responsive hormone analogs [39,132]. Ultimately, the coordination geometry affects not only the binding selectivity but also the overall reactivity. A low-spin d⁶ metal in an octahedral environment can enjoy a stabilization energy (∼27 kcal/mol for Fe²⁺ low-spin), making it less likely to undergo ligand dissociation and improving kinetic inertness. The same metal in a less strongly stabilizing, 5-coordinate or 4-coordinate environment will be more reactive and more prone to substitution [3,90,133]. By designing ligands that support a given geometry, chemists can modulate a complex along the stability-reactivity spectrum. For instance, wrapping Pt²⁺ in a tetradentate ligand that rigidly enforces square-planar geometry can slow aquation by limiting available coordination sites, whereas designing Pt⁴⁺ precursors adopting octahedral coordination in circulation but converting to square-planar Pt²⁺ upon intracellular reduction provides a built-in prodrug activation mechanism [97].

2.4. Competition from endogenous biomolecules

When a metal complex enters a biological system, it is immediately exposed to various endogenous ligands that can bind to metals [134]. This competition primarily occurs at biological interfaces, where proteins, metabolites, membranes, extracellular matrices, and transport systems collectively shape the coordination landscape encountered by metal species. Cells and body fluids are rich in thiols, amines, phosphates, and carboxylates, which can outcompete the intended target, diverting the complex before it approaches its actual site of action [135]. Understanding these competitive interactions is crucial for designing metal-based drugs with sufficient stability to achieve optimal therapeutic performance [136].

GSH is a prevalent intracellular tripeptide (typically 1–10 mM) with a highly nucleophilic thiolate that strongly coordinates with soft metal centers. It serves as a primary detoxifier for electrophiles and metal species. Soft and borderline metal ions, such as Pt²⁺, Pd²⁺, Au(⁺/³⁺), and Cu(⁺/²⁺), are particularly susceptible to forming adducts with GSH. In the case of cisplatin, GSH can replace chloride ligands and bind to Pt²⁺, forming stable Pt–SG complexes and thereby inactivating the drug. Overexpression of GSH and GSH-related enzymes in tumors is a well-recognized mechanism of resistance to Pt-based chemotherapy [14,[137], [138], [139]]. Cysteine-rich proteins, such as metallothioneins, which contain numerous thiol groups, sequester metals, including Pt, Au, and Cd, by coordinating them in thiolate clusters. In the bloodstream, human serum albumin, present at ∼0.6 mM, provides another crucial thiol site at Cys-34, which frequently captures circulating metal complexes such as Pt²⁺ or Au ⁺ species [140]. These interactions can either detoxify a metallodrug or act as a delivery mechanism, as albumin-bound complexes may be preferentially delivered to tumors via the enhanced permeability and retention effect.

One strategy to reduce competition from cellular thiols is to design metal complexes that remain kinetically inert until they reach their targets. Aurofin, a Au⁺ phosphine-thiolato drug, is highly prone to thiol attack; in recent approaches, excess innocuous thiols are co-administered to saturate serum binding sites and thereby elevate the fraction of drug that can reach tumors [[141], [142], [143]]. Another approach is to equip the metal complex with an affinity for its target site that exceeds its affinity for GSH. Chelation therapy for heavy metal poisoning illustrates this principle in reverse: agents such as BAL (British Anti-Lewisite) outcompetes endogenous proteins for As³⁺ by forming stronger chelate complexes [14,144]. Similarly, cysteine-rich proteins, such as metallothionein, albumin (via Cys34), and histones, can strongly bind and sequester metal ions, often redirecting them away from their intended targets [53].

Histidine side chains, with their imidazole nitrogen donors, are less nucleophilic than cysteine thiolates; however, clusters of histidines (e.g., His-tags or poly His motifs) can cooperatively bind metal ions. Consequently, transition metals such as Ni²⁺, Cu²⁺, and Zn² can be readily captured by these sites. Although free histidine in plasma exists at relatively low concentrations, it, along with other amines (amino acids and biogenic polyamines), can still provide additional coordination partners. In practical terms, imidazole-containing drugs or buffering agents can coordinate Zn or Cu and reduce the bioavailable fraction of metallodrugs [145,146].

ATP, ADP, and the phosphate backbone of DNA and RNA offer hard anionic oxygen donors that strongly coordinate with hard metal ions. Lanthanides and Fe³⁺ readily associate with the phosphate groups. Inside cells, many metal complexes tend to localize near nucleic acids or ATP-rich regions through electrostatic attraction to polyphosphate structures. For example, certain Ru complexes preferentially accumulate in the mitochondria, partly because the high ATP and polyphosphate content provides binding sites that effectively retain the metal [147,148].

The body also relies on specialized metal transporter proteins that can unintentionally bind to metal-based drugs (Fig. 5). Transferrin, the principal Fe³⁺ carrier, can also bind other hard metal ions such as Ga³⁺ and Ti⁴⁺, and can capture metallodrugs such as titanocene by accommodating Ti⁴ in its Fe-binding sites. [14,149]. This behavior was observed for titanocene dichloride, where Ti⁴⁺ interaction with transferrin altered the pharmacokinetic profile of the drug. Similarly, ceruloplasmin participates in Cu management by oxidizing and binding Cu, with potential consequences for the accumulation of Cu-based drugs [14,150,151]. Furthermore, human serum albumin (∼0.6 mM in plasma) contains multiple metal-binding sites, including the Cys34 thiol, and can bind Cu²⁺, Zn²⁺, and even Pt complexes [14]. Albumin binding may reduce the freely diffusible fraction of a metal-based drug, effectively sequestering it, but can also serve as a transporter that delivers the drug to cells [14]. In general, endogenous ligands and carrier proteins vastly outnumber drug molecules; therefore, a metallodrug must effectively “win” in a competitive interaction landscape to reach its target. Consequently, biological interfaces act as selective coordination filters that strongly influence metal transport, accumulation, bioavailability, and therapeutic efficacy.

Fig. 5.

Metal platforms, biointerface targets, functional outcomes and translational applications at biological interfaces. Coordination design features, including geometry, redox activity, kinetics, HSAB preference, and targeting capability, influence the interactions of metal-based platforms with key biomolecular targets at the biointerface. The comparative matrix summarizes the representative associations of therapeutic metals (Pt, Ru, Au), catalytic metals (Cu, Fe), diagnostic lanthanide-based systems, and advanced coordination platforms with DNA, proteins, enzymes, membranes, reactive oxygen species (ROS) networks, and immune components. These interactions contribute to major functional outcomes, including DNA damage, protein regulation, ROS modulation, and immune modulation, which underpin diverse biomedical applications, such as chemotherapy, precision therapy, immunotherapy, catalytic therapy, diagnostic imaging, and theranostic drug delivery. The matrix illustrates the relationships among metal categories, biomolecular targets, functional outcomes, and translational application scenarios at the biological interfaces. This figure is a conceptual illustration synthesized from published studies on metallodrugs, metal-mediated biological mechanisms, and diagnostic metal complexes [121,122].

Several strategies have been developed to bias competitive binding. One approach is to use robust protective carrier ligands that shield the metal center and only open under specific conditions at the target site. For example, Ga³⁺ or Cu²⁺ theranostic agents commonly use macrocyclic ligands such as DOTA or NOTA, whose high stability prevents metal stripping by transferrin or albumin, enabling the intact complex to reach tumors and be internalized through specific receptors [152]. Another approach is to exploit kinetic inertness of the metal. As mentioned, Pt⁴⁺ complexes are sufficiently stable in circulation to prevent rapid thiol attack and are only reduced to more thiol-reactive Pt² ions within the reductive intracellular environment, where therapeutic activity is expected [14,53].

2.5. Redox microenvironments and speciation control

Biological microenvironments, including pH, O₂, and redox state, profoundly influence metal speciation in vivo [153]. These parameters vary considerably across biological interfaces, creating localized coordination environments that regulate metal activation, transport, and their biological activity. Many tumors and metabolically active cells exhibit a more acidic and reducing environment, with increased GSH:GSSG ratios, which can be used strategically (Fig. 6). Under physiological conditions in blood and normal tissues (pH ∼7.4, moderate redox potential), some metal-based drugs may remain inert, whereas within the acidic, hypoxic tumor core (pH ∼6.5–6.8, GSH/GSSG ∼50–100:1), they can be selectively activated [14]. The physicochemical properties of the surrounding biological interface determine whether metal complexes remain stable, undergo ligand exchange, or are activated in a spatially controlled manner. A well-established example of this is the Pt⁴⁺ anticancer prodrugs. Octahedral Pt⁴⁺ complexes are usually inert during circulation but are reduced intracellularly by GSH to active Pt²⁺ species [14,62]. Recent studies indicate that in normoxic tissues, Pt⁴⁺ is rapidly re-oxidized and thereby deactivated, whereas in hypoxic tumor regions, re-oxidation proceeds slowly, which effectively retains the reduced Pt²⁺ drug in the tumor microenvironment [14,[155], [156], [157], [158], [159]]. Thus, reductive activation contributes to spatially localized drug action.

Fig. 6.

Coordination-enabled biomaterial design and translation roadmap at biological interfaces. Metal–ligand design principles, including metal selection, ligand engineering, coordination stability, and stimuli responsiveness, guide the development of coordination-based biomaterials, such as metal–organic frameworks (MOFs), nanocarriers, hydrogels, and ECM-mimetic materials. These systems mediate key biointerface functions, including targeting, controlled release, and catalytic activity, leading to biological responses such as drug delivery, regeneration, immunomodulation, ROS regulation, and tissue repair. The resulting applications span oncology, regenerative medicine, inflammation management, theranostics and precision medicine. This framework highlights the progression from biomaterial design to clinical translation and provides a roadmap for the rational development of next-generation coordination-enabled biomaterials. This figure is a conceptual illustration synthesized from published studies on coordination-based biomaterials, nanomedicine and biointerface engineering [69,154].

Many metallodrugs are designed as redox-activated prodrugs. For example, Pt⁴⁺ complexes are relatively stable in the more oxidizing environment of the bloodstream but are reduced inside cells rich in GSH and ascorbate to release active Pt²⁺ [160]. In well-oxygenated healthy tissues, Pt²⁺ can be reoxidized to Pt⁴⁺, limiting sustained activation. Conversely, within hypoxic tumors, re-oxidation is inefficient, so Pt²⁺ persists, and drug release is favored locally [161]. Experimental Pt⁴⁺ drugs have demonstrated this behavior: normoxic conditions support re-oxidation, whereas hypoxic tumor cells lead to reduction and effectively “trap” the active species in the tumor. Additional redox-active metals, such as Co³⁺, Cu²⁺, and Ru³⁺, can undergo in vivo redox changes that activate their activity. For instance, Co³⁺–bisphosphine complexes can be reduced to Co²⁺ under low-oxygen conditions, inducing the selective release of a cytotoxic ligand in tumor tissues. Au³⁺ complexes are typically unstable in the cellular environment and can act as prodrugs that are immediately reduced to Au⁺, which then establishes strong bonds with protein thiols [14,162,163].

The mildly acidic pH of tumor interstitial fluid (∼6.5) and the even lower pH of endosomes/lysosomes (∼5–6) can trigger ligand release or hydrolysis for appropriately designed metal complexes [153,164]. For example, Zn–imidazole frameworks such as ZIF-8 dissolve rapidly at pH < 6.5, making them useful as carriers that release their drugs in acidic endosomal compartments. Early Ti-based drugs, such as titanocene dichloride, hydrolyze rapidly in aqueous solutions. By incorporating acidic ligands, chemists have been able to slow this process until reaching acidic environments, where Ti⁴⁺ is then released. Conversely, attaching acid-labile groups to a metal complex can make it stable at physiological pH but allow it to dissociate selectively in the low pH of tumors or endolysosomal compartments, ensuring site-specific activation [14,85]. Similarly, chelators can be designed to become unstable at low pH, allowing them to dissociate within lysosomes and liberate their metal ions [165].

Tumor hypoxia affects not only the redox balance but also signaling pathways, such as HIF-1. Several metal complexes have been engineered to exploit these mechanisms, including hypoxia-activated Co complexes and Pt⁴⁺ azidoprodrugs which are reduced under low-oxygen conditions to release cytotoxic species [166]. Concurrently, metal ions such as Fe²⁺/Fe³⁺ and Cu⁺/Cu²⁺ participate in Fenton-type reactions that generate ROS, with their cycling and ROS output differing between oxygen-rich and oxygen-poor regions. Consequently, Fe-based drugs can be more toxic in well-oxygenated tissues, where peroxide-driven ROS production is enhanced, while exhibiting distinct behaviors in hypoxic tumor cores [167]. By carefully adjusting the sensitivity of a complex to redox potential and pH, chemists can design “smart” metallodrugs that remain inert but activate only when they encounter specific biological interfaces characterized by distinct redox gradients, pH profiles, metabolic activity, and molecular compositions [14,168].

Beyond redox activation, the biological performance is governed by dynamic processes occurring at the metal–biointerface. Following exposure to biological fluids, coordination-based materials undergo protein adsorption, interfacial ligand exchange, metal-ion release, and surface restructuring, which collectively influence their biological identity and function. Surface coordination sites regulate biomolecular recognition, cellular uptake, and local reactivity, whereas protein corona formation, ECM interactions, material degradation, and immune recognition further shape biodistribution, biosafety, and therapeutic efficacy [13,169]. Consequently, understanding the dynamic evolution of metal–biointerfaces remains a major challenge for the rational design and clinical translation of coordination-based biomaterials and metallotherapeutics [170].

Despite these advantages, the clinical translation of microenvironment-responsive coordination systems remains a challenge. Although frameworks such as ZIF-8 provide efficient pH-triggered cargo release, their physiological stability may be lower than that of more robust coordination architectures, creating a trade-off between responsiveness and systemic persistence of the carrier. Furthermore, the adsorption of serum proteins can generate a protein corona that alters biodistribution, reduces targeting efficiency, and promotes immune clearance. Additional concerns include uncontrolled metal ion leakage, variability in degradation behavior, and batch-to-batch reproducibility during large-scale manufacturing. Consequently, balancing framework stability, biodegradability, safety, and therapeutic efficacy remains a major challenge for the successful clinical translation of coordination-enabled biomaterials and metallodrugs into clinical practice.

3. Coordination of metal complexes with biomolecules

The interactions of metal complexes with biomolecules are the first level of the meta– biological interface, where coordination chemistry directly modulates the structure and function of biological macromolecules [171]. These interactions occur at highly organized biological interfaces, where molecular recognition, spatial confinement, local ligand availability, and physicochemical gradients collectively govern metal binding, transport, activation, and biological response. Exploring these mechanisms is essential for understanding the native roles of metals in biology and rationally designing metal-based therapeutic agents.

3.1. Metal–DNA interaction

Cisplatin binds to DNA by coordinating to the N7 positions of adjacent guanine bases, forming 1,2‐intrastrand crosslinks (predominantly GG and AG) that induce a pronounced kink in the helix [56,62,172]. Within the nuclear bio-interface, DNA represents a densely packed and highly charged coordination landscape that strongly influences metal accessibility, diffusion and binding selectivity. These bifunctional Pt–guanine adducts bend the helix toward the minor groove and interfere with both replication and transcription [[173], [174], [175]]. The square-planar geometry of Pt²⁺ favors adjacent-site crosslinking, whereas trans-configured or bulkier Pt complexes tend to form a higher proportion of interstrand or monofunctional adducts [65]. In contrast, many octahedral Ru²⁺ and Ir³⁺ complexes often form a single covalent bond with DNA (often with G‐N7) and rely on ancillary aromatic ligands for noncovalent interactions. For example, RuPOPs can coordinate G-N7 while extended π‐systems stack between base pairs or align in the minor groove via π–π interactions and H-bonding [69,154]. Positively charged complexes, including multinuclear Pt agents or cationic Ru species, also interact with DNA through electrostatic attraction to the polyanionic phosphate backbone, which helps reorganize the complex for subsequent covalent binding. Consequently, the metal center geometry and ligand determine whether a complex predominantly intercalates, groove binds, or cross-links DNA [65].

Metal complexes can also selectively interact with abnormal DNA structures. These structurally altered regions represent specialized biological interfaces characterized by disrupted base stacking, altered electrostatic environments, and increased accessibility for metal coordination. For instance, metalloinsertors, typically based on Rh³⁺ or Ru²⁺, preferentially bind to thermodynamically unstable sites, such as mismatches or abasic regions. These regions introduce local distortion and excess space within the helix, weakening base stacking and facilitating the intercalation or coordination of bulky complexes with exposed bases. Given this specificity, Rh/Ru/Pt complexes can trigger cleavage or form permanent adducts at aberrant DNA sites, demonstrating how interface-specific structural features can be exploited to achieve selective therapeutic targeting [176,177].

3.2. Metal-protein interaction

As illustrated in Fig. 7, metal complexes interact with proteins via multiple mechanisms. Proteins constitute some of the most abundant and chemically diverse biological interfaces, providing highly organized coordination environments that regulate metal recognition, binding affinity, catalytic activity, and downstream signaling responses. This direct active‐site coordination is a common mode of interaction; labile metal centers such as Pt²⁺, Au⁺, Ru²⁺, or Co³⁺ bind catalytic residues such as cysteine, histidine, or selenocysteine, thereby inhibiting enzymatic function [178]. For example, auranofin is a Au⁺ complex that irreversibly binds to the selenocysteine–cysteine motif in the active site of thioredoxin reductase (TrxR), blocking its redox activity. metal complexes can act as cofactors that bind to metalloproteins. Pt²⁺ or Ru²⁺ may coordinate with Zn²⁺ or Mg²⁺⁻dependent sites, displacing the native metal or competing with the natural substrate [179,180]. These interactions at non-catalytic metal-binding sites or regulatory cysteines can result in allosteric changes, ultimately disrupting protein conformation and modulating their activities, highlighting how protein biointerfaces translate coordination events into functional biological outcomes [171].

Fig. 7.

Protein coordination and functional modulation at the biological interfaces. Metal coordination with amino acid residues, including cysteine, selenocysteine, and histidine, alters protein structure and function through multiple mechanisms. Coordination at catalytic sites can inhibit protein activity, whereas binding to regulatory regions can induce allosteric modulation. Metal complexes may also compete with endogenous metal ions, disrupt disulfide bonds, perturb protein conformation, or displace structural metal cofactors, such as Zn²⁺. These coordination-mediated interactions collectively regulate protein activity, stability, and biological function, highlighting protein coordination as a central mechanism through which metal-based therapeutics modulate biological processes at the biointerface. This figure is a conceptual illustration synthesized from published studies relevant to the mechanisms and applications depicted [102].

Metal-protein interactions also occur via disulfide bonds and thiol groups. These sulfur-rich microenvironments create highly favorable coordination interfaces for soft metal centers and strongly influence the biological activity of metallodrugs. Consistent with the HSAB principles discussed in Section 2.1, soft metal centers preferentially coordinate with sulfur-rich protein interfaces [181]. Auranofin and related Au⁺ complexes form S–Au–S adducts, effectively “mimicking” or reshaping disulfide patterns and disrupting protein structure. Because of this broad thiol reactivity, many metallodrugs inhibit multi-target inhibition; auranofin not only inactivates TrxR but also binds to cysteine proteases, protein kinases, and GSH transferase [182,183]. It can further target Zn-finger proteins by coordinating with cysteine/histidine ligands and displacing Zn²⁺ [14].

3.3. Metal-lipid interaction

Metal complexes can strongly interact with lipid bilayers [184,185]. The lipid membrane is a dynamic biological interface that controls cellular entry, molecular transport, signal transduction, and metal distribution in biological systems. Cationic complexes, such as many Ru²⁺ and Pt²⁺ compounds, are attracted to negatively charged phospholipid headgroups (Fig. 8). At this membrane interface, electrostatic interactions, lipid composition, membrane curvature, and local hydration layers collectively determine metal accumulation and membrane-associated activities. Cations can coordinate the phosphate oxygens on membrane lipids, altering the bilayer organization and mechanical properties [186]. For example, Li⁺ ions bind to phospholipid phosphate groups and increase membrane stiffness, and a metal complex with an accessible coordination site can bind to a lipid phosphate group. Amphiphilic metal complexes have hydrophobic ligands that enter the membrane interior, whereas the charged metal remains near the aqueous surface. This amphiphilic behavior promotes partitioning into a bilayer and disruption of the lipid matrix [187,188].

Fig. 8.

Metal–lipid interactions at the biological interfaces. Cationic, amphiphilic, and multinuclear metal complexes interact with lipid membranes via electrostatic association, membrane insertion, coordination-mediated binding, and supramolecular assembly. These interactions alter membrane organization, increase membrane stiffness, disrupt membrane integrity, and promote the generation of ROS and ion leakage. This membrane perturbation contributes to cellular dysfunction and apoptosis. The figure illustrates the representative mechanisms through which metal-based systems modulate the membrane structure and function at biological interfaces. This was synthesized from published studies relevant to the depicted mechanisms and applications [14,149].

Many metallodrugs generate ROS that induce lipid peroxidation (Table 3). As lipid membranes serve as major oxidative interfaces, membrane-associated ROS generation frequently amplifies biological damage and therapeutic efficacy. Polynuclear Pt, Au, and Ru complexes have been found to catalyze ROS formation, especially under light irradiation, thereby promoting oxidative damage to the membrane [217]. Studies on antibacterial metallodrugs have shown that membrane rupture often results from a combination of ROS-driven lipid oxidation and strong cationic/hydrophobic interactions with the bilayer [218,219]. The positively charged, lipophilic nature of these complexes leads to lipid rearrangement, increased permeability, and leakage, as their amphiphilic structures drive the initial electrostatic association, followed by destabilization of the bilayer [220].

Table 3.

Microenvironment-responsive coordination platforms and translational design strategies.

Trigger Condition	Metal Platform	Coordination State in Circulation	Activation Mechanism	Active Species Generated	Target Compartment	Therapeutic Outcome	Translational Advantage	Ref.
High GSH	Pt(IV)	Octahedral inert	Reductive activation	Pt(II)	Nucleus	DNA crosslinking	Tumor selectivity	[189]
Hypoxia	Co(III) complexes	Stable Co(III)	Reduction to Co(II)	Cytotoxic ligand release	Tumor core	Hypoxia-specific killing	Reduced systemic toxicity	[190]
Acidic pH (∼6.5)	ZIF-8 (Zn)	Stable at pH 7.4	Framework dissolution	Zn2+ + payload	Endosome	Drug release	Endosomal targeting	[191,192]
Low O2	Ru(III)	Inert Ru (III)	Reduction to Ru(II)	Active Ru(II)	Hypoxic tumor	Selective cytotoxicity	Microenvironment targeting	[193,194]
Light (λ = 450 nm)	Ru(II) polypyridyl	Dark-stable	Photo-ligand dissociation	Aquated Ru(II)	Cytosol	DNA binding	Spatial control	[117]
Elevated ROS	Cu(II) complexes	Stable	Fenton-like cycling	•OH radicals	Lipid membranes	Ferroptosis	Synergistic ROS therapy	[195,196]
High ATP	Ga(III) -DOTA	Chelated stable	Receptor-mediated uptake	Intact complex	Tumor cell	Imaging + therapy	Stability in blood	[195,196]
Lysosomal pH (∼5)	Acid-labile Pt complex	Stable neutral	Ligand hydrolysis	Pt(II)	Lysosome	Organelle damage	Intracellular activation	[197,198]
High transferrin	Titanocene	Labile Ti(IV)	Transferrin binding	Ti–Tf complex	Tumor	Altered PK	Receptor targeting	[199,200]
Elevated polyphosphate	Ln(III) complexes	Hard–hard stable	Electrostatic retention	Ln3+ localized	Mitochondria	Organelle imaging	Subcellular selectivity	[201]
High NADH	Ir(III)	Inert	Redox-triggered activation	Reduced Ir species	Mitochondria	ROS burst	Metabolic targeting	[202,203]
GSH depletion	Fe(II)-salen	Labile Fe2+	Redox cycling	Lipid peroxides	Membrane	Ferroptosis	Bypass apoptosis resistance	[204,205]
Enzyme overexpression (MMP)	Metal–peptide hybrids	Shielded	Peptide cleavage	Active metal core	Tumor ECM	Local release	Protease specificity	[206]
High chloride (blood)	Cisplatin	Stable (Cl-rich)	Aquation intracellular	[Pt(NH3)2(H2O)2]2+	Nucleus	DNA damage	PK modulation	[207]
HIF-1α signaling	Os(II) arene	Inert	Protein binding	HIF inhibition	Hypoxic cells	Anti-angiogenesis	Pathway-specific action	[208,209]
Reducing mitochondria	Au(III)	Partially stable	Reduction to Au(I)	Thiol-binding Au(I)	Matrix	TrxR inhibition	Organelle selectivity	[210]
Elevated lipid peroxides	Cu-phenanthroline	Stable	Redox cycling	ROS amplification	Membrane	Ferroptosis	Synthetic lethality	[211,212]
Acidic tumor stroma	MOF-Pt hybrid	Stable MOF	Framework degradation	Pt(II)	Tumor	Sustained release	Controlled kinetics	[212,213]
Light + O2	Ir photosensitizer	Dark stable	Singlet oxygen generation	1O2	Tumor	Photodynamic therapy	Non-invasive control	[214,215]
Inflammatory ROS	Fe3O4 nanozyme	Stable	Peroxidase-like catalysis	ROS	Macrophages	Immune activation	Immunotherapy synergy	[216]

The role of metal complexes as synthetic transmembrane ion channels or carriers well established [221]. This strategy directly exploits the membrane biointerface to regulate ion flux, cellular homeostasis, and therapeutic responses through coordination-driven transport mechanisms. Recent studies have shown that Pt²⁺ complexes with H-bond donor ligands are efficient chloride anionophores. Depending on the design, these complexes either exchange ligands with Cl– via a labile mechanism or retain their original ligands and bind Cl– via H-bonding in a fixed-ligand mechanism. In vesicle assays, these Pt-based carriers transport Cl– across lipid bilayers and induce apoptosis by disrupting ionic gradients [158,159]. More generally, MOFs, such as coordination cages and self-assembling metal helices, have been engineered to be inserted into membranes to facilitate ion or small-molecule transportation. Consequently, these metal-based systems demonstrate how coordination chemistry can be used to create synthetic membrane channels and modulate transmembrane flux for therapeutic purposes [222].

Although membrane-targeting metal complexes can enhance therapeutic efficacy through improved cellular interactions, the widespread distribution of membrane lipids in healthy tissues may reduce selectivity and increase off-target toxicity. Whether membrane disruption primarily contributes to therapeutic benefits or dose-limiting toxicity remains dependent on the disease context and is still incompletely understood.

3.4. Metal interaction with glycocalyx and carbohydrate

The cell surface glycocalyx, composed of glycoproteins, glycolipids, and polysaccharides, is increasingly recognized as a valuable target for metal complexes (Fig. 9). As the outermost biological interface separating cells from their surrounding microenvironment, the glycocalyx plays a critical role in molecular recognition, cellular communication, immune regulation and therapeutic targeting. One strategy involves the binding of carbohydrate ligands to metallodrugs. Many tumor cells overexpress glucose transporters (GLUTs) and lectins; therefore, conjugating sugar to a metal complex can utilize these natural uptake pathways. Pt²⁺ and Ru²⁺ complexes bearing glucose or galactose moieties have been developed to mimic native sugars [62]. These “glycoconjugates” show enhanced accumulation in highly glycolytic cancer cells via GLUT-mediated transport, illustrating how interface-specific carbohydrate recognition can improve selective delivery and cellular uptake. Beyond cancer, metal–sugar conjugates are being explored for targeting pathogens and other disease states by leveraging specific lectin–carbohydrate recognition pathways [223]. As shown in Fig. 9, metal-binding fragments can also be used to target carbohydrate-binding proteins within the glycocalyx. These metal-binding fragments effectively block the sugar recognition site and impede interactions with sialylated or mannosylated glycans. This demonstrates how coordination chemistry can be used to disrupt glycan–protein interactions at the glycocalyx biointerface, thereby modulating cell adhesion, immune recognition, and disease-associated signaling pathways [224]. Despite the attractive targeting potential of glycocalyx-directed systems, substantial heterogeneity in glycan expression among tissues and disease states may limit the reproducibility of targeting and therapeutic precision. The differences between experimental models and human glycocalyx architecture further complicate the translational prediction of targeting performance. These coordination-mediated interactions with glycocalyx components illustrate how interface-specific carbohydrate recognition can be engineered to improve selective delivery and modulate cell adhesion, immune recognition, and disease-associated signaling at the outermost biological interface.

Fig. 9.

Glycocalyx-targeting strategies at the biological interface. Glycoproteins, glycolipids, mucins, lectins, and cell surface receptors within the glycocalyx are key recognition elements that can be exploited by coordination-based systems. Metal centers, glycan ligands, and responsive linkers collectively influence glycan recognition, receptor binding, and receptor-mediated cellular uptake. These interactions affect intracellular trafficking and target engagement, ultimately contributing to enhanced therapeutic performance through improved efficacy, selectivity, and safety of the drug. The figure illustrates the relationships between glycocalyx recognition, coordination design, cellular uptake and therapeutic response at biological interfaces. This is a conceptual illustration synthesized from published studies relevant to the depicted mechanisms and their applications [152].

4. Coordination chemistry in cellular pathways

Once inside the cell, metal complexes can modulate critical pathways and trigger cell death. This section highlights the central mechanisms through which coordination-based therapeutics exert their biological effects. These biological effects are governed by a series of intracellular biointerfaces, including mitochondrial membranes, organelle boundaries, protein-signaling networks, and immune cell communication interfaces, which collectively determine the localization, activation, and biological consequences of metal complexes.

4.1. Metal-mitochondria interaction and redox disruption

Mitochondria are common targets for metal-based drugs [225]. The mitochondrial membrane is a highly specialized biointerface characterized by a strong transmembrane potential, unique lipid composition, and densely organized redox-active proteins, making it particularly susceptible to coordination-driven therapeutic interventions. Their strongly negative membrane potential (∼–150 mV) attracts lipophilic cations and creates an electrochemical interface that preferentially accumulates positively charged metal complexes in the mitochondrial matrix. Therefore, positively charged and moderately lipophilic metal complexes, such as cyclometalated Ir³⁺, RuPOPs, or Au(⁺/³⁺) complexes, tend to accumulate within the mitochondrial matrix [226]. Inside the mitochondria, these complexes encounter mitochondrial DNA and a dense set of redox-active proteins. Many complexes are specifically designed to interfere with the electron transport chain [227]. For instance, a mitochondria-localized Au³⁺ thiosemicarbazone has been reported to disrupt respiratory complexes, leading to a loss of membrane potential (ΔΨm) and a significant increase in ROS [228]. High ROS levels damage the mitochondrial lipids and DNA. At the mitochondrial biointerface, these oxidative events disrupt membrane integrity, electron transport, and metabolic signaling, ultimately amplifying therapeutic responses. Thus, it promotes cytochrome c release and induces caspase-dependent apoptosis. Similarly, various Ir³⁺ and Ru²⁺ complexes cause mitochondrial swelling and membrane disruption, triggering intrinsic apoptotic pathways (Erxleben, 2019). This “mitochondrial poisoning” strategy is particularly effective because many cancer cells are heavily dependent on mitochondrial metabolism and experience increased oxidative stress [227,229]. By converting mitochondria into ROS-generating hubs, metal complexes preferentially kill tumor cells. Importantly, numerous mitochondria-directed metallodrugs have been engineered as prodrugs that remain relatively inert until they reach the reductive, thiol-rich mitochondrial matrix, where they are converted into active species, for example, Pt⁴⁺ is reduced to Pt²⁺ [230,231]. Although mitochondrial targeting can improve therapeutic selectivity, excessive ROS generation can damage healthy tissues and contribute to dose-limiting toxicity. Achieving an optimal balance between mitochondrial disruption and systemic safety remains an important challenge in translational development.

4.2. Metal-mediated modulation of signaling pathways

Metal complexes affect oxidative stress and modulate key kinase and transcriptional pathways. These signaling networks function as dynamic molecular biointerfaces that integrate extracellular stimuli, intracellular redox status, and coordination-mediated biochemical events. For instance, RuPOPs can suppress the MAPK/ERK and PI3K/Akt signaling pathways. RuPOP inhibits focal adhesion kinase (FAK) activity in breast cancer cells, leading to a significant decrease in ERK and Akt phosphorylation. Suppression of the PI3K/Akt-MAPK cascades impairs cell migration and improves TRAIL-induced apoptosis [232]. Au⁺ complexes are well-recognized modulators of the NF-κB pathway. The antirheumatic drug auranofin covalently binds IKKβ at Cys-179, irreversibly blocking IKK activity, thereby preventing IκBα phosphorylation and NF-κB nuclear translocation [233]. Similarly, Au⁺ and Au³⁺ species can inhibit TrxR, thereby disrupting redox signaling upstream of NF-κB. In contrast, some metal complexes have been designed to interfere with hypoxia-responsive signaling. For example, an organo-osmium²⁺ arene complex was engineered to target the HIF-1α–p300 interface and inhibit HIF-1α-driven gene transcription under low oxygen conditions [234].

Overall, metal–ligand complexes modulate signaling through two key mechanisms, as summarized in Sections 2.4, 2.5: interface-dependent redox regulation and direct coordination with signaling proteins. These combined effects modulate key pathways, including PI3K/Akt, MAPK, NF-κB, and HIF-1α, shifting cellular outcomes from proliferation and survival toward stress responses and apoptosis, demonstrating how metal complexes reshape intracellular signaling interfaces to redirect cellular fate from survival toward stress adaptation and cell death [232].

4.3. Metal-mediated apoptosis, autophagy, and ferroptosis

Metal complexes can trigger several distinct forms of programmed cell death, including apoptosis, autophagy, and ferroptosis. Many coordination drugs induce the intrinsic (mitochondrial) apoptotic pathway [171]. For instance, cyclometalated Ir³⁺ and Au³⁺ complexes selectively accumulate in mitochondria, causing organelle swelling, inner membrane rupture, cytochrome c-release, and ultimately activating caspases [235]. These complexes often cause significant mitochondrial dysfunction and increase ROS production, which induces damage to DNA and membrane lipids, ultimately promoting apoptosis. Furthermore, Au³⁺–tetrahydroisoquinoline compounds can induce a classic ER stress–CHOP response, leading to apoptosis and pro-death autophagy [225,235]. Dual induction of apoptosis and autophagy is crucial, since autophagy can clear damaged components; however, when overactivated, it amplifies cell death signaling [236]. In addition, several metallodrugs can induce autophagy (cell self-digestion), which may function either as a pro-survival response or a cell death mechanism [236]. An earlier study reported that Au³⁺ compounds promote autophagosome formation alongside classical apoptotic markers [235]. In cases where apoptosis is impaired, autophagy can act as an alternative route to eliminate damaged cells.

Ferroptosis is an Fe-dependent, non-apoptotic form of cell death characterized by GSH depletion and extensive lipid peroxidation. Metal complexes can induce ferroptosis by disturbing the cellular redox balance [237,238]. For example, Cu²⁺-phenanthroline and Fe²⁺-salen complexes undergo redox cycling with H₂O₂ to produce hydroxyl radicals via Fenton chemistry, rapidly depleting GSH and inactivating GPX4. This promotes the uncontrolled peroxidation of polyunsaturated membrane lipids. As described earlier, ferroptotic cells exhibit increased mitochondrial membrane density and loss of cristae, with depleted GSH and inactive GPX4, enabling Fe²⁺-mediated lipid oxidation in a Fenton-like manner [237]. Hence, redox-active Cu and Fe complexes serve as ferroptosis inducers by simultaneously generating ROS and disabling antioxidant defenses, resulting in lipid ROS accumulation and irreversible membrane injury [237,239].

4.4. Immune cell response to metal complexes

Coordination-based therapeutics can profoundly influence immunity [238]. A key phenomenon is immunogenic cell death (ICD), in which certain metal drugs trigger tumor cell death in a way that releases danger-associated molecular patterns (DAMPs) and activates both innate and adaptive immunity. For instance, cancer cells treated with oxaliplatin or Ru complexes expose calreticulin and release ATP and HMGB1, which in turn stimulate dendritic cells and cytotoxic CD₈⁺ T cells- [240]. Recent evidence highlights that several metal complexes are validated ICD inducers and are capable of provoking protective antitumor immune responses in vivo [240,241]. Therefore, the tumor-metal “nexus” can act as a self-vaccinating system for the tumor.

On the innate immune side, metal ions and particulate forms can prime macrophages and inflammasomes. Particulate metal debris, such as TiO₂, Co, or Ni, is known to activate the NLRP3 inflammasome in macrophages (often after TNF-α priming), resulting in IL-1β secretion and an M1-like inflammatory phenotype [242]. Classical adjuvants, such as aluminum hydroxide (alum), act in part via macrophage activation and inflammasome signaling [243]. Similarly, metal oxide nanoparticles (NPs) (e.g., Fe oxide and Co ferrite) are being developed as vaccine adjuvants to enhance antigen presentation [244]. Thus, metal coordination compounds can reshape the tumor-immune interface by eliciting ICD and pro-inflammatory signaling in tumor cells while modulating T cells, macrophages, and dendritic cells. Notably, Fe-based coordination nanomaterials can induce ferroptosis to promote DC maturation and increase CD8⁺ T cell-infiltration in mouse models [245], indicating how metal–biological interface chemistry links direct cytotoxic effects with immunotherapy [240]. However, the long-term immunological consequences of chronic metal exposure remain unclear. Furthermore, the differences between experimental tumor models and human immune systems continue to complicate the translation of immunomodulatory metal complexes into clinical practice.

5. Coordinated therapeutics: metal-based drugs at the bio-interface

5.1. Platinum complexes

Pt²⁺ complexes remain the standard for coordination-based chemotherapy. Their therapeutic activity is primarily mediated at the nuclear DNA interface, where platinum coordination alters the genomic structure, transcriptional activity, and DNA repair processes. Cisplatin (cis-[PtCl₂(NH₃)₂]) exerts its cytotoxicity by forming 1,2-GpG interstrand DNA crosslinks, thereby inhibiting replication. It is effective against a wide range of solid tumors; however, its poor selectivity results in dose-limiting nephrotoxicity, neuropathy, and the development of resistance. Second-generation analogs have been developed to address these limitations [246]. Carboplatin replaces the two chlorides with a cyclobutane-1,1-dicarboxylate chelate, which slows aquation and reduces off-target nonspecific reactions [247]. Oxaliplatin uses a bulky DACH ligand to generate distinct DNA adducts that evade recognition and repair pathways, making it effective in settings where cisplatin is ineffective, such as in colorectal cancer. Satraplatin is an orally bioavailable Pt⁴⁺ prodrug carrying acetate ligands; it is reduced in vivo to activate Pt²⁺ species with broader tissue distribution and reduced renal toxicity compared with cisplatin [247].

As discussed in Section 2.5, Pt(IV) complexes exploit interface-dependent redox activation to achieve the controlled intracellular release of Pt(II), enabling selective activation within tumor-associated microenvironments. Many Pt⁴⁺ designs seek to mitigate cisplatin-like side effects by adding tumor-targeting moieties or solubilizing groups [97]. Coordination chemistry has also been applied to enhance the delivery of Pt. Liposomal cisplatin (Lipoplatin) and polymer-Pt conjugates, such as NC-6004 (a PEG-carboxylate-Pt framework), improve tumor accumulation and reduce systemic toxicity, in part, through the enhanced permeability and retention effect. Additional strategies involve the use of Pt-loaded NPs or MOFs that combine controlled Pt release with photothermal therapy. Across these approaches, the aim is to deliver active Pt²⁺ preferentially to tumor tissue while reducing exposure to healthy organs [97]. However, despite these advances, platinum-based delivery systems continue to face challenges related to heterogeneous tumor accumulation, premature drug release, manufacturing complexity, and variable clinical translation.

5.2. Ruthenium complexes

Ru offers highly versatile chemistry, including accessible oxidation couples and rich photophysical properties. Importantly, Ru complexes interact with multiple biological interfaces, including the ECM, DNA, mitochondrial membranes, and tumor-associated proteins, providing mechanistic versatility beyond that of conventional platinum drugs. These complexes follow two major therapeutic themes: redox-active Ru(²⁺/³⁺) and photoactivated Ru²⁺ [248]. In the redox-active Ru(²⁺/³⁺) approach, Ru³⁺ compounds such as NAMI-A and KP1019 (and its sodium salt KP1339), which carry indazole ligands, have progressed to clinical trials. NAMI-A exhibited significant anti-metastatic effects in mouse models, with modest direct cytotoxicity. Its activity is considered to depend on the in vivo reduction of Ru³⁺ to Ru²⁺, followed by interaction with proteins and disruption of metastasis-related pathways. KP1019/KP1339 exhibits broader tumor cell cytotoxicity and has been investigated particularly in liver cancer [249]. Photoactivated Ru²⁺ is another commonly followed therapeutic theme when using Ru complexes, where luminescent RuPOPs can be designed to remain largely inert in the dark but become highly cytotoxic upon light exposure [250]. These systems act as photodynamic or photochemotherapeutic agents; for example, [Ru(bpy)₂(dppz)]²⁺ derivatives can generate singlet oxygen or cause direct DNA cleavage upon irradiation [251]. Arene-Ru²⁺ complexes such as RAPTA compounds generally act as stable prodrugs, demonstrating antiangiogenic effects through interactions with the ECM and inhibition of metalloproteinases. These effects illustrate how extracellular biointerface properties regulate tumor invasion, angiogenesis, and therapeutic responses. Similarly, RuPOPs intercalate into DNA and can be structurally modulated to reduce their toxicity in normal cells [232].

5.3. Copper and iron complexes

Cu²⁺ and Fe³⁺ complexes are often employed as Fenton catalysts to overwhelm cancer cells. These systems primarily function at redox-active biological interfaces characterized by elevated H₂O₂ levels, altered metabolism, and disrupted antioxidant homeostasis. For example, Cu²⁺-phenanthroline and related diamine complexes can undergo redox cycling between Cu²⁺ and Cu⁺, where intracellular reductants (e.g., GSH) reduce Cu²⁺ to Cu⁺, which then reacts with O₂ or H₂O₂ to produce superoxide and hydroxyl radicals. This process depletes cellular reducing equivalents (GSH/NADPH) and causes oxidative damage to biomolecules [228]. Similarly, organoiron (ferrocene-based) complexes and Fe³⁺ porphyrins catalyze the conversion of H₂O₂ into •OH, promoting the lethal oxidation of membranes and DNA. As tumors often show increased H₂O₂ levels, these Fenton-active metal agents can achieve a degree of selectivity for cancer cells [252]. This selectivity arises from interface-specific redox characteristics rather than intrinsic metal toxicity. In addition to direct ROS generation, some Fe- and Cu-based drugs chelate thiols and inhibit antioxidant enzymes, such as catalase or TrxR, further amplifying oxidative stress [237].

5.4. Gold complexes

Au-based anticancer drugs predominantly function through the covalent modification of thiol groups. Their biological activity is governed by sulfur-rich protein interfaces that provide highly favorable coordination environments for Au(I) and Au(III) species. Linear, strongly thiophilic Au⁺ complexes carrying phosphine or N-heterocyclic carbene ligands effectively target cysteine- and selenocysteine-containing proteins. Auranofin, which is approved for the treatment of rheumatoid arthritis, exemplifies this model [253]. It primarily inhibits TrxR, resulting in thioredoxin oxidation, disruption of the redox balance, and induction of apoptosis. Au complexes also inhibit the NF-κB pathway by alkylating IKKβ (Cys-179). Next-generation Au⁺–NHC complexes are frequently engineered as lipophilic cations that accumulate in mitochondria, where they induce ROS production [254]. Au³⁺ species generally act as prodrugs that are reduced by cellular thiols to Au⁺, which then forms covalent bonds with the protein cysteine. Au³⁺–porphyrin and Au³⁺–biotin conjugates have been used to selectively deliver Au³⁺ to cancer cells, where they bind to intracellular proteins [255,256]. Ultimately, both Au⁺ and Au³⁺ complexes inhibit key redox enzymes (e.g., TrxR and GSH reductase) and deubiquitinases, causing oxidative and proteotoxic stresses. Notably, Au drugs also suppress anti-inflammatory activity by modulating toll-like receptor pathways and NF-κB signaling, thereby reducing cytokine production [254]. Therefore, Au (⁺/³⁺) anticancer agents work by “sulfur trapping” selectively targeting thiol-rich protein interfaces involved in redox regulation, inflammatory signaling, and cellular survival [257].

5.5. Lanthanide complexes

Lanthanide³⁺ ions, particularly Gd³⁺, Eu³⁺, and Tb³⁺, are commonly used in imaging and theranostic applications because of their magnetic and optical characteristics [258]. These systems function at diagnostic biointerfaces, where coordination-controlled magnetic and optical responses enable the visualization of biological processes. Gd-based chelates such as Gd–DOTA and Gd–DO3A are well-established T₁ MRI contrast agents, and recent innovative approaches incorporate photosensitizers or radioisotopes into these chelators to create combined imaging-therapy (theranostic) agents [258]. Eu³⁺ and Tb³⁺ complexes are intrinsically luminescent in the visible and near-infrared ranges and can be used as optical probes. For example, Eu³⁺ cryptates and Tb³⁺ phenanthroline complexes can be used to monitor cellular drug uptake using fluorescent tracking [259]. Lanthanides have also been incorporated into NPs for energy conversion [73]. Upconversion NPs (UCNPs) co-doped with Yb³⁺/Er³⁺ or Yb³⁺/Tm³⁺ absorb NIR light and emit in the visible or UV range, enabling activation of conjugated photosensitizers for photodynamic therapy (PDT) [260]. Hybrid systems, such as lanthanide–porphyrin frameworks, merge MRI contrast (from the lanthanide core) with PDT from porphyrin ligands. Similarly, Gd-oxyfluoride nanocrystals doped with thulium or erbium have been used for simultaneous MRI and PDT, taking advantage of both the magnetic relaxation properties and characteristic lanthanide emissions [261]. Hence, lanthanide coordination compounds serve as multifunctional imaging–therapy platforms, in which the lanthanide center offers high relaxivity and distinctive luminescence, while the attached ligands or nano-scaffolds provide the therapeutic function.

5.6. Enzyme-mimetic coordination therapeutics

A new paradigm is to use coordination-based assemblies as enzyme-mimicking catalysts (“nanozymes”) that exert therapeutic reactions in situ [262]. These systems create artificial catalytic biointerfaces that emulate natural enzymatic functions while enabling localized control of biochemical reactions. Fe- and Ce-containing materials are classic examples of this emerging approach (Fig. 10). For instance, Fe₃O₄ NPs exhibit peroxidase-like activity and convert endogenous H₂O₂ into •OH radicals via Fenton chemistry within tumors. Similarly, MOFs and coordination polymers (e.g., Fe–MIL-53 or Mn–based MOFs) can be engineered to exhibit combined oxidase, catalase, and peroxidase activities [263]. A recent study used a modified quasi–Fe–MIL-53(Fe) nanozyme generated by partial ligand removal, which showed enhanced peroxidase/catalase mimicry and GSH-depleting capability [245]. This Fe-based nanozyme simultaneously generates ROS and depletes cellular GSH, thereby inducing ferroptosis and ICD. In vivo, it promotes dendritic cell maturation and enhances CD8⁺ T cell infiltration [245].

Fig. 10.

Catalytic design framework linking metal centers, catalytic modes, ROS modulation, biological outcomes, and therapeutic applications at the biological interfaces. Catalytic metals (Fe, Cu, Mn, and Co), noble metals (Pt, Au, and Ru), and advanced nanozyme platforms (MOFs, single-atom catalysts, and hybrid systems) regulate biological activity through ROS generation, scavenging, and stimuli-responsive activation. These catalytic processes drive ROS elevation, reduction, and redox reprogramming, leading to tumor cell death, inflammation resolution, immune activation, and tissue protection. The resulting therapeutic applications include catalytic oncology, nanozyme therapy, immunotherapy, theranostics and precision medicine. The framework highlights key design principles linking metal selection, catalytic efficiency, ROS control, target specificity, and therapeutic outcomes for the rational development of next-generation catalytic metallotherapeutics. This is a conceptual illustration synthesized from published studies relevant to the depicted mechanisms and their applications [87].

Other designs integrate multiple catalytic steps into a single framework. One such system combined a Cu²⁺-based glucose oxidase mimic (which produces H₂O₂ from glucose) with an Fe³⁺–porphyrin peroxidase mimic, which converts H₂O₂ into •OH). This cascade depletes glucose and GSH while generating increased ROS levels, effectively starving and oxidizing cancer cells [264]. Beyond Fe and Cu systems, single-atom catalysts (e.g., Co or Mn atoms doped in carbon matrices) can exhibit SOD-like or oxidase-like activities, enabling context-dependent ROS scavenging or generation [265]. In all these systems, coordination chemistry defines enzyme-like catalytic centers, and metal ions in specific ligand environments replicate the active sites of natural oxidases, peroxidases, or phosphatases [171]. These catalytically active therapies blur the distinction between drugs and enzymes, using metal coordination to drive redox transformations (oxidation of cellular substrates and reduction of O₂) at the biological interface [245]. They provide a versatile platform for synergy, as oxidoreductase-mimicking nanozymes can sustain ROS cascades and ICD, complementing the effects of conventional metal-based drugs [245]. Compared to conventional metallodrugs, nanozymes offer catalytic amplification, multifunctionality, and prolonged therapeutic activity. However, challenges related to biodistribution, long-term biosafety, and clinical scalability remain significant barriers to their translation.

5.7. Comparative perspectives on metallodrugs and delivery platforms

Different classes of metallodrugs and coordination-based delivery systems have distinct strengths and limitations. Platinum drugs remain the most clinically validated agents; however, they suffer from resistance and cumulative toxicity. Ruthenium complexes offer broader mechanistic flexibility and anti-metastatic potential, although clinical evidence remains limited [266]. Gold complexes preferentially target sulfur-rich proteins and redox pathways but require improved selectivity [267]. Copper- and iron-based systems enable catalytic ROS amplification but may induce off-target injuries. Similarly, MOFs, nanozymes, ferritin nanocages, and polymeric carriers offer superior payload loading and multifunctionality, although manufacturing complexity, scalability, and regulatory approval remain major challenges [268]. Therefore, platform selection should be guided by the disease type, target bio-interface, safety profile, and therapeutic objectives.

6. Coordination-driven nanomedicine and drug delivery

6.1. Metal–organic frameworks

MOFs are highly porous and adjustable coordination networks that can be designed as smart, stimulus-responsive drug delivery systems [269]. As programmable nano-biointerfaces, MOFs regulate transport, molecular recognition, and controlled release in response to local biological signals. Prototypical crystalline scaffolds, such as ZIF-8, MIL-100, HKUST-1, and UiO-66, can be engineered to respond to biochemical signals characteristic of tumor microenvironments. These systems exploit the interface-responsive activation mechanisms summarized in Section 2.5. Zn–imidazolate-like ZIF-8 is stable at neutral pH but rapidly decomposes under weakly acidic environments. In practice, ZIF-8 particles loaded with doxorubicin (DO) break down in endosomes (pH ~ 5–6), releasing their active chemicals [270]. A previous study reported a ZIF-8 system in which DOX was sequestered through a pH‐labile linker and shielded by a folic-acid shell; at low pH, the shell dissociates, the ZIF-8 cage collapses, and the acid-sensitive DOX prodrug is cleaved, triggering drug release [270]. Introducing disulfide bonds or redox-active metal nodes enables MOFs to respond to high GSH levels in cancer cells. For instance, a UiO-66-NH₂ MOF bearing folate and disulfide-containing linkers was engineered such that intracellular GSH cleaved the disulfides, oxidizing GSH to GSSG and promoting the controlled release of co-loaded drugs [271]. Similar designs use Cu²⁺ or Fe³⁺ nodes that can be reduced or participate in Fenton chemistry, simultaneously depleting GSH and liberating the bound drugs. The MOF surface can be grafted with peptides or polymer gates that serve as substrates for tumor-associated enzymes, including matrix metalloproteinases and hyaluronidase [272]. These molecules keep the pore closed until they encounter a specific protease, at which point cleavage opens the framework and enables local drug release. MOFs are “stimuli-responsive platforms” whose payloads can be programmed to occur under the acidic, reductive, or enzymatic conditions characteristic of tumor microenvironments [14].

Despite their considerable potential, MOFs face several translational challenges. Compared with highly stable frameworks such as UiO-66, ZIF-8 provides superior pH responsiveness but reduced physiological stability, creating a trade-off between controlled release and systemic persistence in vivo. Furthermore, the adsorption of serum proteins can generate a protein corona that alters biodistribution, reduces targeting efficiency, and promotes immune clearance. Additional concerns include metal ion leakage, framework degradation variability, batch-to-batch reproducibility, large-scale manufacturing, and regulatory approval. Consequently, balancing stability, biodegradability, safety, and therapeutic efficacy remains a challenge for the clinical translation of MOF-based therapeutics [273].

6.2. Polymeric coordination nanocarriers

Metal–ligand crosslinks can also be used to stabilize polymer-based drug delivery systems [160]. Amphiphilic polymers or polysaccharides equipped with ligand groups, such as carboxylates, catechols, or histidines, can be interconnected by divalent or trivalent metal ions to form self-assembled networks [274]. Representative examples of this system include the following:

Catechol–Fe³⁺ hydrogels: Chitosan modified with catechol groups (mimicking mussel adhesive chemistry) forms an injectable, self-healing hydrogel (CAT-Gel) upon the addition of Fe³⁺. This Fe³⁺-catechol hydrogel can encapsulate both hydrophilic (DOX·HCl) and hydrophobic (docetaxel) drugs and release them sequentially over >40 days [275]. Fe³⁺ coordination offers shear-thinning behavior, strong adhesion, and dynamic network reformation after mechanical disruption [276].

Zn²⁺ carboxylate polymer networks: Polymers rich in carboxylate sidechains, such as acrylic/siloxane hybrids, can be crosslinked by Zn²⁺to form rigid matrices. A PDMS polymer containing –COO^- groups, when crosslinked by Zn²⁺, became extremely stiff at room temperature [277]. Crucially, Zn–carboxylate interactions are thermally labile, as heating shifts the equilibrium toward bond dissociation, softening the gel, which then rapidly re-stiffens upon cooling [[277], [278], [279]]. This thermoresponsive modulus switching, spanning several orders of magnitude, indicates how dynamic metal-ligand bonds can enable tunable mechanics and self-healing properties [280,281].

Ion-switchable matrices: Classical systems include Ca²⁺-crosslinked alginate “egg-box” gels and Fe²⁺/Fe³⁺–crosslinked alginate scaffolds. In such systems, changing the pH or adding a competing chelator (e.g., EDTA) disrupts the metal crosslinks, enabling controlled gel dissolution and the release of the encapsulated drug [282].

6.3. Coordination-enabled targeting ligands

Transition metals and their chelating ligands are commonly used to attach targeting and imaging moieties to nanocarriers [283]. A well-established approach uses macrocyclic chelators, including DOTA, NOTA, and NODA, to complex radiometals. When these ligands are conjugated to peptides or NPs, they can stably chelate isotopes such as ⁶⁴Cu, ⁶⁸Ga, or ¹⁷⁷Lu, thereby enabling PET/SPECT imaging or radiotherapy of the carrier system [284]. Likewise, nitrilotriacetic acid (NTA) is often conjugated with Ni²⁺ and is frequently grafted onto particle surfaces to capture His-tagged proteins. In such designs, an NP bearing Ni–NTA can be rapidly functionalized by simple incubation with a His₆-tagged targeting ligand or cytokine, forming a stable coordination bond [285].

Engineered metalloproteins also function as nanoscale modules. The 24-subunit protein cage of ferritin can bind thousands of Fe atoms, and heavy-chain ferritin intrinsically targets transferrin receptor 1 (TfR1), which is overexpressed in many tumors [286]. Ferritin nanocages can encapsulate small-molecule drugs or imaging metals within their 8 nm internal cavity and can be genetically or chemically modified on the exterior to display targeting peptides. This combination of intrinsic metal-binding capability and receptor specificity makes ferritin a particularly versatile coordination-based carrier [286,287]. Similarly, other metallothioneins (small Cys-rich metal-binding proteins) or synthetic poly His/metal-binding peptides can sequester ions such as Zn²⁺ or Cu²⁺ and have been explored as dynamic handles for assembling carriers or enabling stimulus-responsive release. Hence, coordination motifs, such as DOTA/NTA, metal-affinity tags (His-tag/Ni²⁺), and protein–metal scaffolds (ferritin, metallothionein, and designed peptides), provide modular anchors for targeting ligands and imaging reporters, supporting the construction of multifunctional, self-assembling nanocarrier platforms [286,288].

6.4. Coordination in ECM engineering

Metal coordination is increasingly used to construct ECM-mimicking scaffolds for regenerative medicine and localized cytokine delivery. The ECM is a dynamic biological interface that regulates cell adhesion, migration, signaling, and tissue regeneration, making it an attractive target for coordination-based engineering. For instance, Ca²⁺-crosslinked alginate or pectin hydrogels, where “egg-box” interactions between Ca²⁺ and clustered carboxylate groups form biocompatible matrices capable of encapsulating cells and growth factors. These ionic gels are intrinsically dynamic, as divalent cations can diffuse out or be exchanged, enabling gradual biodegradation and sustained drug release [289]. Emerging systems utilize catechol–metal interactions to mimic natural adhesive proteins. Hyaluronic acid modified with gallol (catechol-like) groups can be cross-linked by Fe³⁺ to form mussel-inspired bioadhesive hydrogels [6,14]. For example, adjusting the pH and Fe³⁺ concentration enables control over the mono-, bis-, and tris-catechol coordination states within the HA gel. The resulting hydrogels exhibited rapid self-healing and strong wet adhesion, which are crucial for tissue integration, with bis-coordinated networks providing optimal printability and stability [290]. Coordination bonds can also confer redox or enzymatic responsiveness to ECM-like scaffolds. Incorporating disulfide or thiol–metal linkers makes gels sensitive to high intracellular GSH levels or oxidative environments in healing tissue, while embedding matrix-metalloproteinase-cleavable peptides allows scaffolds to soften as invading cells secrete proteases [290].

7. Catalytic bio-interface: artificial metalloenzymes and intracellular catalysis

7.1. Nanozymes for oxidative therapy

Nanomaterials with metal-ligand active sites can act as enzyme mimics for therapeutic response. These systems establish artificial catalytic biointerfaces capable of reproducing or augmenting natural enzymatic processes within living systems. As described in Section 4.6, Fe₃O₄ and related NPs exhibit peroxidase-like activity, and surface Fe²⁺ enhances the Fenton conversion of H₂O₂ to •OH [291,292]. Cu- and Mn-based nanomaterials can perform superoxide dismutase (SOD)- or oxidase-like functions, converting superoxides into O₂/H₂O₂ or H₂O₂ into O₂/H₂O. Their catalytic activities can be tuned by coordination engineering, including the choice of dopants and axial ligands used. For example, introducing an axial chloride ligand to a Zn–N₄ site substantially increases its peroxidase-like turnover rate [293]. These nanozymes are typically designed to localize in tumors, where increased H₂O₂ levels and a reduced microenvironment maintain continuous ROS generation and GSH depletion. The net effect is a catalytic form of “chemotherapy,” in which sustained, locally amplified ROS production drives oxidative damage, preferentially in cancer cells, exploiting their abnormal redox balance [294]. Compared with natural enzymes, nanozymes offer superior stability and design flexibility but often exhibit lower catalytic specificity and more complex pharmacokinetic behaviors. Whether these systems can maintain long-term catalytic performance under physiologically relevant conditions remains an active research area.

7.2. Photoactivated and redox-responsive catalytic drugs

Stimuli-activated metallocatalysts are another class of therapeutic agents that have been developed. For example, light-activated RuPOPs remain inert until irradiated with visible light. Upon photoexcitation, they can produce singlet oxygen or undergo ligand dissociation to create reactive species on demand, enabling spatially confined therapy limited to the illuminated tumor region [295,296]. Similarly, redox-activated prodrugs exploit the distinct chemical milieu of tumors. Au³⁺ and Ir³⁺ complexes can be inert in the bloodstream but are reduced by high intracellular thiol levels to Au⁺ or Ir⁺, releasing active metal fragments. Pt⁴⁺ azide complexes and Co³⁺ prodrugs follow related principles; under hypoxic tumor conditions, they are reduced and trigger drug release [293,294]. Through rational ligand design and the incorporation of photochromic units or redox-sensitive bonds, it is possible to achieve precise spatiotemporal control of metal-centered catalysis. Expanding the π-systems in Ru²⁺ photosensitizers shift their absorption into therapeutically useful wavelength ranges, whereas bioreducible linkers ensure activation by intracellular GSH. These strategies help minimize off-target toxicity by keeping metal centers functionally “dormant” until they encounter tumor-specific stimuli [295,[297], [298], [299]] [291,300].

7.3. Fenton and Fenton-like catalytic chemotherapy

Catalytic chemotherapy utilizes transition metal redox cycling (Fe²⁺/Fe³⁺ or Cu⁺/Cu²⁺) to amplify intracellular oxidative stress. In acidic, H₂O₂-rich tumor environments, Fe²⁺ centers rapidly convert H₂O₂ into highly reactive hydroxyl radicals (•OH) through the Fenton reaction [293,[301], [302], [303]]. Cu-based systems function in a similar manner, where intracellular GSH reduces Cu²⁺ to Cu⁺, which then reacts with H₂O₂ to form •OH and regenerate Cu²⁺. This Cu⁺/Cu²⁺ Fenton-like cycle simultaneously produces ROS and oxidizes GSH (Cu²⁺ + 2GSH → Cu⁺ + GSSG + 2H⁺), progressively depleting the antioxidant capacity of the cells and driving ROS to harmful levels in tumor cells [304]. Previous evidence indicates that H₂O₂ bound to Fe₂⁺ on Fe₃O₄ surfaces accepts electrons to dissociate into •OH radicals [293]. As discussed in Sections 5.3, 7.1, catalytic Fe- and Cu-based systems exploit tumor-associated redox interfaces to sustain ROS generation, deplete antioxidants, and induce oxidative damage.

Tumor selectivity is another crucial aspect of catalytic chemotherapy [305]. Cancer cells typically exhibit high levels of H₂O₂ and reducing agents [306]. Coordination-based Fe or Cu nanomedicines are specifically designed to attain these behaviors [307]. Fe/Cu–containing MOFs or hybrid systems selectively deliver metal centers to tumor tissues, where the acidic, hypoxic conditions support sustained conversion of endogenous H₂O₂ into ROS while concurrently depleting GSH (mimicking a GPx-like reaction) [297]. The outcome is a targeted oxidative attack that preferentially damages and kills the tumor cells.

7.4. Enzyme-inspired coordination catalysts for oxidative stress regulation

Using artificial nanozymes as antioxidants is a complementary approach that mimics catalase (CAT), SOD, or GSH peroxidase (GPx), thereby buffering excess ROS and protecting healthy tissues. Pt-group nanozymes, including Pt NPs or Pt–N₄ single-atom sites, catalyze the breakdown of H₂O₂ into O₂ and H₂O in a catalase-like manner [297]. Similarly, nanoceria and Cu-based frameworks promote the dismutation of superoxide (O₂•^-) to H₂O₂ and O₂, resembling the activity of SOD [297]. Furthermore, vanadium oxide (V₂O₅) nanowires and mixed-valence Mn₃O₄ use cellular GSH to reduce peroxides in a GPx-like manner, consuming ROS while oxidizing GSH [297]. The catalytic performance of these nanozymes depends on the metal valence dynamics and coordination environment. For example, a higher Ce³⁺ fraction in nanoceria enhances SOD-like activity, whereas mixed Mn²⁺/Mn³⁺ phases enable broader multi-enzyme mimicry [294]. Rational doping approaches or defect (vacancy) engineering can be used to optimize these redox cycles. Furthermore, these antioxidant nanozymes often accumulate in ROS-enriched compartments, such as mitochondria and lysosomes. By acting in situ, these agents directly re-establish redox homeostasis and limit oxidative damage. In preclinical models, such antioxidants have shown promise in protecting normal tissues and modulating inflammatory responses [297]. Recent studies have indicated that coordination-based catalytic systems may regulate emerging forms of regulated cell death, including pyroptosis and PANoptosis, by modulating ROS-dependent inflammatory signaling pathways.

7.5. Metal-induced pyroptosis, PANoptosis, and cGAS–STING activation

Emerging evidence suggests that metal-based therapeutics can activate inflammatory forms of regulated cell death, in addition to apoptosis and ferroptosis. Excessive ROS generation, mitochondrial dysfunction, lysosomal disruption, and DNA damage induced by metallodrugs may activate inflammasome-associated caspases and gasdermin-driven pyroptosis. Recent studies have further indicated that coordination-based systems can simultaneously engage apoptotic, pyroptotic, and necroptotic pathways, collectively termed PANoptosis. In parallel, metal-induced DNA damage and mitochondrial DNA leakage can activate the cGAS–STING pathway, promoting type I interferon production, dendritic cell activation, and enhanced antitumor immunity. These emerging mechanisms provide new opportunities to overcome therapeutic resistance and improve immune-mediated tumor eradication.

8. Coordination in diagnostics and theranostics

Paramagnetic chelates of Gd³⁺, Mn²⁺, and Fe³⁺ remain central to MRI contrast design, with relaxivity determined by a finely balanced combination of coordination geometry, intersphere hydration number, water-exchange kinetics, and rotational dynamics [308]. In Gd-based contrast agents (GBCAs), increasing the number of coordinated water molecules can enhance T₁ relaxivity, but only within the limits set by thermodynamic stability and kinetic inertness. Ligand designs that permit higher hydration numbers, often by reducing denticity or rigidifying the scaffolds, can efficiently enhance relaxivity while simultaneously increasing the risk of Gd³⁺ dissociation [308]. This trade-off has resulted in a clinical shift toward macrocyclic chelators, whose pre-organized geometry confers superior kinetic inertness [171,308]. Modern complex design strategies have followed this framework by introducing rigid, chiral, or second sphere–modulating substituents to reduce molecular tumbling and optimize water exchange, thereby improving relaxivity without compromising safety [309].

The regulatory restriction of linear GBCAs underscores the significance of kinetic stability in clinical imaging and reinforces macrocyclic ligands as the dominant platform for future GBCA development [310]. Concurrently, Mn²⁺ complexes have been reported as Gd-free alternatives, leveraging the endogenous nature of Mn²⁺ and its favorable high-spin state for T₁ shortening. However, the inherently weak ligand field around Mn²⁺ enables adequately stable chelation, which is particularly challenging [310].

Evidence has demonstrated that rigid polyaminocarboxylate and macrocyclic ligands can prevent the dissociation and trans-metalation (exchange of ligands between metals) of Mn, while still allowing the coordination of at least one inner-sphere water molecule [308]. Advances in ligand rigidification strategies, such as cyclohexyl-based backbones, pyridyl substitutions, and bispidine-type architectures, have yielded Mn²⁺ chelates with substantially enhanced kinetic inertness and clinically relevant relaxivity, some of which have progressed to advanced clinical testing [308]. These systems indicate that for Mn²⁺ contrast agents, the principal design hurdle is not achieving relaxivity but maintaining robust metal retention under physiological competition from endogenous cations [308]. These observations highlight a broader translational challenge in coordination-based imaging, namely, balancing imaging performance with long-term biosafety, metal retention, and regulatory acceptance.

Fe-based MRI agents play a limited but mechanistically promising role compared to gadolinium or manganese systems. In clinical practice, superparamagnetic Fe oxide NPs are predominant, whereas small-molecule Fe³⁺ chelates typically exhibit only modest T₁ relaxivity because of rapid electronic relaxation and fully saturated coordination environments. [18,[311], [312], [313]]. Nonetheless, carefully designed catecholate and other oxygen-donor ligands can introduce inner- or second-sphere water interactions that are sufficient to yield measurable T₁ contrast. Given Fe's favorable biocompatibility and recycling pathways, Fe chelates remain attractive for certain specialized applications, provided that challenges such as aggregation, excessive T₂ effects, and nonspecific protein binding are carefully controlled [311]. Radiometal-based imaging increases the existing requirements because any in vivo instability directly compromises image fidelity. PET and SPECT probes using ⁶⁴Cu, ⁶⁸Ga, ⁸⁹Zr, or 99mTc require chelators that precisely match the metal's ionic radius, oxidation state, and preferred coordination geometry while providing near-absolute kinetic inertness over the timescales of biological imaging [314].

For Cu²⁺, rigid cross-bridged macrocycles outperform flexible chelators, primarily because they resist transchelation to the serum proteins. Ga³⁺ is a complex and strongly oxophilic cation that is effectively stabilized by hexadentate macrocycles, such as NOTA and its derivatives, providing both rapid room-temperature radiolabeling and exceptional in vivo stability. In contrast, Zr⁴⁺ requires complete saturation of its coordination sphere; substituting hexadentate desferrioxamine with octadentate hydroxypyridinone ligands dramatically reduces in vivo metal release and skeletal accumulation, demonstrating that complete coordination is a key determinant of radiometal safety [315]. Furthermore, technetium-based radiopharmaceuticals follow similar principles, using pre-organized donor sets tailored to specific Tc cores to maintain circulation [316]. The same coordination geometry also supports the design of fluorescent and multimodal imaging probes.

In metal-responsive fluorophores, metal binding alters the electronic structure or conformational flexibility of the ligand, suppressing non-radiative decay or enabling processes such as Förster resonance energy transfer (FRET) [316]. Zn and Ca parameters utilize these mechanisms to achieve highly sensitive cellular imaging, whereas lanthanide complexes harness the antenna effect to produce sharp, long-lived emissions suitable for time-gated detection [316]. Across all these systems, selectivity and signal modulation result from careful adjustment of donor atoms, ligand preorganization, and control of solvent quenching, efficiently reflecting the design-related constraints encountered in MRI chelate development [316].

Coordination chemistry also supports the development of probes that combine multiple imaging signals, termed multimodal imaging probes. For example, in dual-mode MRI/fluorescence agents, a fluorescent dye is usually integrated with paramagnetic metal chelates or NPs. One common strategy involves incorporating Gd³⁺ chelates into conjugated polymers or dendrimers that also carry fluorophores. This macromolecular scaffold provides a high local density of Gd sites (amplifying r₁ relaxivity) and an aggregation-induced emission (AIE)-active dye for fluorescence. Alternatively, small-molecule scaffolds have been developed in which one segment chelates Gd³⁺ and another carries a fluorophore (e.g., BODIPY or cyanine dye) [317]. However, key design challenges include reducing fluorescence quenching by the paramagnetic center (often achieved by inserting a rigid spacer between the dye and metal) and maintaining MR efficacy by preserving at least one inner-sphere water molecule on Gd. Reported examples of this architecture include Gd complexes carrying anthracene or coumarin units that yield robust cellular fluorescence along with T₁ contrast, as well as TPE-based Gd constructs that exhibit aggregation-induced emission and high r₁ relaxivity [318].

PET/MRI dual probes integrate radiometals and MRI-active metals within a single construct. A common approach uses a bifunctional chelator capable of coordinating both ⁶⁴Cu and Gd or an NP that bears each metal in distinct domains. Prominent examples of this system include Fe oxide cores providing MRI contrast and loaded on the surface with ⁶⁴Cu-labeled chelates or liposomes incorporating Gd–chelate in the lipid bilayer while ⁸⁹Zr is bound to a targeting antibody. Such constructs leverage coordination multiplexing; likewise, a DOTA-PEG scaffold may interact with Gd³⁺ for MRI, while a separate chelating motif on the same backbone binds ⁶⁴Cu for PET. The critical requirement is the orthogonality of the coordination sites (or controlled sequential radiolabeling) so that each metal remains securely complexed [319]. Another alternative strategy is to use a single lanthanide complex (e.g., ¹⁵⁵Tb or ₁₇₇Lu), which is both paramagnetic and radioactive, thereby enabling SPECT/MRI or PET/MRI with a single molecular entity.

Tri-modal probes extend this modality by combining optical, MRI, and nuclear signals into a single system. For example, lanthanide UCNPs can be doped with Gd³⁺ to provide MRI contrast, radiolabeled on the surface with chelated ¹⁸F or ⁶⁴Cu for PET, and simultaneously used for their intrinsic NIR emission as an optical readout [320]. Alternatively, “metal–organic polyhedra” have also been proposed, in which different faces bear distinct functionalities: one face carries a fluorescent ligand, another coordinates Gd³⁺, and another complexing a PET radionuclide. In these modalities, fluorescence (intensity or lifetime), T₁ relaxivity, and radionuclide decay are all driven by the coordinated metal center working in tandem [317].

Notably, the chelation strategy and overall molecular structures affect all modalities. For instance, increasing the molecular size to boost r₁ relaxivity may also slow pharmacokinetics and affect PET signal profiles, warranting careful optimization. The key performance metrics for multimodal probes include detection sensitivity, in vivo stability, and degree of spatial co-localization between signals. Well-designed systems have yielded sub-nanomolar PET detection limits, sub-micromolar MRI relaxivity, and low background fluorescence by precisely adjusting their coordination-based assemblies [321,322].

9. Pharmacology, toxicology, and clinical landscape

The coordination of chelate design has a primary effect on metal biodistribution and clearance. Rigid macrocyclic ligands, such as DOTA, are highly pre-organized and typically provide complexes with both high thermodynamic stability and pronounced kinetic inertness. In contrast, flexible acyclic ligands (for example, DTPA analogs) often form less tightly bound complexes that are easier to dissociate in vivo [319]. From a safety perspective, kinetic stability (inertness) is usually regarded as more critical than the absolute stability. Consequently, macrocyclic and high-denticity chelates generally remain intact until they are cleared, most often through renal excretion, whereas more labile complexes may release the metal and undergo nonspecific interactions [323]. This finding is consistent with the clinical perspective that specific linear Gd³⁺ contrast agents (such as gadodiamide or gadopentetate) are associated with nephrogenic systemic fibrosis (NSF) due to Gd³⁺ release. Recent in vivo studies support this observation; in mice administered different Gd³ +-chelates, rigidified complexes such as Gd(PIP-DOTA) were excreted intact, whereas more flexible chelates, Gd(AZEP-DTPA) and Gd(NPTA), showed in vivo Gd release. This evidence highlights that relatively subtle changes in ligand backbone rigidity, donor set, and overall charge can dramatically affect in vivo behavior, emphasizing the crucial role of coordination stability in driving biodistribution and off-target metal release [323].

The protein corona is another key determinant of in vivo behavior of nanomaterials. As one of the most important nano-biointerfaces, the protein corona defines the biological identity of coordination-based nanomaterials and strongly influences their biodistribution, immune recognition, and clearance. When introduced into biological fluids, metal-based NPs or even small coordination complexes rapidly adsorb abundant serum proteins through their surfaces. This absorbed layer can mask the original ligand shell and may also promote ligand exchange [324]. For instance, a metal-ligand complex may release its chelator, which is replaced by a coordinating residue from albumin, transferrin, or other proteins, or the reverse may also occur. The resulting corona effectively defines the biological identity of the particle, altering immune recognition, circulation time, and clearance [325,326]. In extreme cases, coordinated metal ions can be directly transferred to the corona proteins, changing the metal speciation. These dynamic interactions of adsorption, exchange, and corona reorganization complicate pharmacokinetics; a nanostructure engineered to be stealthy may become rapidly opsonized if its protective ligand layer is displaced [327,328]. Minimizing detrimental corona effects often requires “stealth” coatings (for example, PEGylation) or highly inert chelates, highlighting that in vivo coordination conditions are not static but are continuously reshaped by biomolecular interactions.

Coordination chemistry also contributes to organ-specific uptake and the associated toxicity. Poorly chelated or free manganese readily crosses the blood–brain barrier and accumulates in the basal ganglia, resulting in a Parkinson-like “manganism” syndrome. In contrast, Cu and Fe complexes tend to localize in the liver; intravenously injected ⁶⁴Cu²⁺ is rapidly sequestered by hepatocytes and eliminated via the biliary route, leading to pronounced hepatic accumulation [329,330]. Gd-based contrast agents are typically cleared renally; however, any Gd³⁺ released in vivo can bind to phosphates and proteins and deposit in the kidneys, bone, and brain, contributing to NSF and long-term Gd retention. Interestingly, these coordination-dependent distribution mechanisms can be used for the targeted imaging. Chelated Mn²⁺ (as in MnDPDP) behaves as a Ca²⁺ analog and is taken up by the metabolically active myocardium [331,332]. In rat myocardial infarct models, manganese-enhanced MRI (MEMRI) revealed smaller infarct areas than conventional gadolinium-contrasted MRI because Mn²⁺ selectively enters viable cardiomyocytes via L-type Ca²⁺channels, whereas gadolinium remains limited to the extracellular space [333,334].

The significance of coordination principles and structural arrangements is evident from clinical experience with metallodrugs. For example, cisplatin, a Pt²⁺-based DNA crosslinker approved >50 years ago, remains one of the most commonly used anticancer agents despite its significant nephrotoxicity, ototoxicity, and neurotoxicity. Its efficacy and underlying adverse events illustrate the balance between metal reactivity and off-target injuries to the tissue. Ferumoxytol, a superparamagnetic Fe³⁺ NP approved for treating anemia and used off-label as an MRI contrast agent, shows efficient uptake by the reticuloendothelial system (liver and spleen) with relatively minimal long-term toxicity [125,335]. In contrast, unsuccessful candidates underscore the risk of unanticipated metal-metal interactions. PledOx (calmangafodipir, a Mn²⁺ chelate) was designed to reduce oxaliplatin-induced neuropathy in phase III clinical trials. Mechanistic analyses have indicated that the in vivo redox interplay between Mn²⁺ (from PledOx) and Pt²⁺ (from oxaliplatin) contributed to this detrimental outcome [336,337]. Thus, co-administered metal centers can participate in unforeseen redox processes that intensify toxicity if such interactions are not carefully controlled.

Ongoing advances in coordination chemistry have driven more rational drug design. Novel metal-based drugs increasingly rely on rigid ligand frameworks and targeting moieties. Second- and third-generation Pt agents, including carboplatin and oxaliplatin) integrate a chelating ring system that slows deactivation and thereby minimizes dose-limiting toxicities [247]. Similarly, innovative MRI contrast agents, such as gadobutrol and gadopiclenol, favor macrocyclic chelators to lock Gd³⁺ with high kinetic inertness. In oncology, cytotoxic metal complexes are conjugated with antibodies or peptides, thereby enhancing tumor selectivity while reducing systemic exposure [[338], [339], [340]].

10. Challenges and bottlenecks

A critical concern in metal-based drug design and clinical efficacy is the thermodynamic stability–inertness paradox. Generally, these complexes must be intact and sufficiently inert in the bloodstream to prevent premature metal release. However, they must be sufficiently reactive at the target site to exert their therapeutic effects. Notably, highly inert metal complexes are uniformly distributed across living systems and cleared without exerting meaningful biological effects, whereas overly labile complexes release metals indiscriminately and cause off-target toxicity [323]. This behavior is particularly crucial for stimulus-activated prodrugs (e.g., pH- or enzyme-activated systems) and catalytic agents such as metalloenzyme mimics, which remain intact during circulation and are selectively activated when they reach the target tissue. However, efficient design is extremely challenging, and robust, generalizable rules are limited; for example, increasing ligand rigidity with a macrocycle may enhance stability but completely abolish activity, necessitating several compromises [341]. In clinical settings, this balance is usually maintained using certain approaches, such as stimuli-responsive chelation and hierarchical ligand shells, with the stability-reactivity dilemma being a persistent bottleneck during in vivo applications.

Redox mismatch is another limitation of metal complexes design. Biological tissues vary widely in redox potential, and hypoxic or inflamed microenvironments can reduce or oxidize metal centers in unexpected ways. For instance, a strongly reducing tumor microenvironment may convert a paramagnetic lanthanide or transition metal to a distinct oxidation state, thereby altering its relaxivity profile or toxicity [342,343]. Conversely, an increased oxidative environment during inflammatory responses can induce the activation of a premature redox-sensitive metal prodrug. These in vivo redox mechanisms are difficult to predict; for example, Mn²⁺ may oxidize to Mn³⁺ or Mn⁴⁺ under certain conditions, resulting in precipitation as oxides or enhanced binding proteins [161,344]. These shifts can disrupt the intended coordination state, enabling complexes that appear stable in vitro to behave differently in vivo. Consequently, the incorporation of redox-active metals (Ce, Ru, or Fe) into therapeutic agents requires careful assessment of tissue-specific redox conditions [14,345].

Furthermore, the off-target accumulation of metal-based therapeutics presents another layer of complexity in the development of effective designs. Mononuclear phagocytes in the liver, spleen, and bone marrow can efficiently sequester several metal complexes and NPs, which can reduce on-target exposure while increasing toxicity. For example, intravenously injected ionic ⁶⁴Cu²⁺ shows pronounced hepatic uptake, followed by biliary excretion, with limited tumor accumulation in the absence of a targeting approach [346,347]. Similarly, Fe oxide NPs, such as ferumoxytol, are selectively absorbed by Kupffer cells. This sequestration can also disrupt endogenous metal balance; excessive Cu or Zn accumulation in the brain or liver has been linked to neurological and hepatic disorders, and the unintended displacement of native metals by therapeutic agents may further aggravate the disease [348,349]. Although targeting ligands and careful adjustment of particle size and surface charge are utilized to reduce uptake by these off-target sinks, achieving organ-selective delivery of metal payloads remains highly challenging.

The computational prediction of metallodrug behavior is still minimal. Standard physiologically based pharmacokinetic (PBPK) models usually overlook dynamic ligand exchange and metal binding in various compartments. Quantum chemistry or molecular dynamics can estimate static parameters, such as bond strengths or selective geometries; however, they are poorly suited for capturing complex, time-dependent phenomena, such as protein corona formation, enzymatic degradation, or organ-specific clearance [350]. Data-driven QSAR and ML approaches are also limited because training data for in vivo pharmacokinetics and toxicity are scarce. Consequently, the in silico prediction of the biodistribution and clearance of metal-based drugs remains unreliable, and most drug development programs still depend heavily on experimental biodistribution studies in rodent models, which are both expensive and time-consuming. Moreover, discrepancies between animal models and human physiology frequently complicate translation, emphasizing the need for more predictive experimental and computational frameworks for coordination-based therapeutics [351,352].

Finally, regulatory and safety considerations are major concerns in the development of metal-based drugs. Regulatory agencies increasingly require extensive data on long-term metal retention because of the well-documented risks. After evidence of Gd deposition in the human brain, the EMA banned several linear Gd chelates in 2017, effectively shifting clinical use towards macrocyclic agents [353]. Consequently, newly developed metallodrugs are usually subjected to extended retention and clearance studies, often over months to years, to demonstrate safe elimination. Therefore, the approval threshold is higher than that for conventional small molecules; they must show not only acceptable toxicity but also safe long-term metal handling [14,354]. These crucial requirements increase both the timelines and development costs, but they are crucial given the potential for metals to persist indefinitely in the body.

Future progress will require standardized terminology and classification frameworks that clearly distinguish metal center properties, coordination environments, catalytic functions, biomolecular recognition events, and downstream biological outcomes. Such distinctions are essential for establishing mechanistic relationships across molecular, cellular, pharmacological and clinical scales.

10.1. Clinical translation and development status of metal-based systems

Despite substantial advances in coordination chemistry and metal-based therapeutics, successful clinical translation remains limited by challenges related to toxicity, pharmacokinetics, biodistribution, metal ion release, long-term biosafety, manufacturing reproducibility, and regulatory approval. To provide a translational perspective, representative approved, clinical-stage, discontinued, and preclinical metal-based systems are summarized in Table 4; it highlights the metal centers, ligand platforms, primary indications, mechanisms of action, major toxicities, development stages, and key translational challenges that have influenced clinical success or limited broader implementation.

Table 4.

Clinical translation and development status of representative metal-based systems (revised). The newly added column “Basis of Limitation” distinguishes mechanistic potential, clinical evidence, and specific reasons for translational barriers.

Collectively, these examples illustrate the diverse translational trajectories of metal-based therapeutics, diagnostic agents and coordination-enabled nanomaterials. Clinically successful systems are typically characterized by well-defined pharmacokinetics, manageable toxicity profiles, and robust manufacturing pathways. In contrast, emerging platforms frequently face challenges associated with biological complexity, long-term biosafety, reproducibility, and regulatory approvals. These observations emphasize the importance of integrating coordination chemistry, biointerface interactions, pharmacology, and translational considerations in the development of next-generation metallotherapeutics.

11. Future directions; next-generation medicines

Short-term priorities (0–5 years) are expected to focus on predictive design, biointerface characterization, and data-driven optimization of coordination-based therapeutics. AI-assisted coordination design represents one of the most immediately translatable opportunities for next-generation metallotherapeutics development. Machine learning and other data-driven models are increasingly being applied to predict metal–ligand binding affinities, kinetic inertness, and redox potentials. Graph neural networks trained on coordination complex databases can suggest novel chelator scaffolds or identify potentially problematic chemistries [355]. Currently, AI-assisted approaches are used to optimize conjugation sites and target ligands for metallodrug delivery systems. While still evolving, predictive models proposed using these innovative approaches hold potential for developing advanced metal complexes design with finely adjusted stability, reactivity, and target selectivity, thereby reducing the reliance on purely experimental screening [355]. However, major challenges include limited high-quality datasets, insufficient predictive models for metal speciation and bio-interface interactions, and poor integration between computational predictions and experimental validation methods. Future progress will require standardized coordination chemistry databases, explainable machine-learning frameworks, and closed-loop experimental–computational workflows (Fig. 11).

Fig. 11.

Roadmap for the future development of coordination-based biointerface systems. The roadmap highlights short-, medium-, and long-term priorities for next-generation coordination-enabled therapeutics and biomaterials, spanning AI-assisted design, protein assemblies, nucleic-acid delivery, immunomodulation, hybrid biologics, bioelectronic interfaces, and logic-gated systems. The framework illustrates the progression from biointerface engineering and advanced therapeutic platforms to clinical translation and future biointerface medicines, supported by data-driven design, safety assessment, scalable manufacturing, and regulatory improvements. This is a conceptual illustration synthesized from emerging trends in coordination chemistry, biointerface engineering, nanomedicine, immunotherapy, bioelectronics, and translational medicine [355].

Medium-term priorities (5–10 years) are expected to focus on advanced therapeutic platforms, including metal–protein assemblies, nucleic acid delivery systems, immunomodulatory metallotherapeutics, and hybrid biologics. Metal–protein assemblies and engineered metalloproteins offer opportunities to develop highly specific and biologically integrated therapeutic systems. Protein cages, such as apoferritin and viral capsids, can be loaded with metal ions or chelates to create biocompatible delivery pathways. For example, Gd-loaded ferritin nanocages exhibit extremely high relaxivity because of their slow rotational dynamics, which can target cells through ferritin receptors, making them highly sensitive MRI probes [356,357]. Beyond the imaging applications of metalloenzymes, engineered and artificial metalloproteins are being developed for therapeutic catalysis, including ROS production within tumors and the degradation of toxins. These biohybrid systems combine the specificity and biodegradability of proteins with the distinctive reactivity of metal centers, providing promising avenues for smart and self-regulating therapies [358,359]. Key bottlenecks include maintaining structural stability under physiological conditions, controlling metal loading and release, reducing immunogenicity, and enabling the scalable manufacturing of these nanomaterials. Addressing these challenges will require modular assembly strategies and an improved understanding of protein–metal biointerface interactions (Fig. 11).

Coordination-enabled nucleic acid delivery platforms may expand the therapeutic scope of gene editing, RNA therapeutics, and precision medicine. Divalent and trivalent metal ions, including Zn²⁺, Ca²⁺, and Fe³⁺, can link negatively charged nucleic acids, condensing DNA and RNA into compact assemblies. Utilizing this principle, MOFs and coordination polymers are being developed as delivery systems for gene therapies, in which metal–ligand scaffolds can encapsulate oligonucleotides, protect them from degradation, and be engineered to release them in response to specific triggers (pH, redox conditions, and light) [360]. For example, Zn-based MOFs have been used to efficiently load siRNA and trigger its release under an endosomal acidic pH. Such metal-mediated delivery systems offer both structural strength and the potential to integrate metal-based imaging or therapeutic functions into a single multifunctional platform [361,362]. Major challenges include endosomal escape, long-term cargo stability, protection from biological degradation, and reproducible and large-scale production. Future efforts should focus on programmable release mechanisms and clinically translatable carrier architectures.

Coordination-based immunomodulation represents a rapidly emerging medium-term strategy. Some metal complexes can act as anti-inflammatory agents by scavenging ROSs, such as Mn-based SOD mimetics [363,364]. Other complexes are designed to stimulate immunity rather than suppress it. Several Ru and Ir complexes have been reported to reprogram tumor-associated macrophages towards an M1-like phenotype (anti-tumor state) or enhance T-cell activation [365,366]. Recent evidence has further demonstrated that metal-mediated activation of pyroptosis, PANoptosis, and cGAS–STING signaling enhances antitumor immunity and improves responsiveness to immunotherapy. Additionally, metal NPs and coordination cages are being investigated as vaccine adjuvants, where they can function as both antigen delivery and provide innate danger signals to elicit immune responses [367,368]. These strategies exploit the redox properties and coordination chemistry of metals to modulate inflammatory pathways and antitumor immunity in unconventional manners. Key challenges include achieving selective immune regulation, minimizing systemic toxicity, identifying predictive biomarkers, and understanding the long-term consequences of immune cell reprogramming. Future development should prioritize biomarker-guided patient stratification and rational combination therapy.

Hybrid biologics that integrate metals with antibodies, peptides, and nucleic acids may improve targeting precision and therapeutic selectivity. These complexes include metal-binding peptides and antibodies that deliver metal payloads with high specificity. For instance, conjugating a cytotoxic metal chelate to an antibody, such as antibody–drug conjugates, can localize metal-based toxicity to cancer cells [340,369,370]. DNA or RNA aptamers linked to metal centers offer similar targeting potentials. The metal component of these complexes supports accurate imaging or catalysis, while the biomolecular part provides cell-type specificity [371]. These cellular processes can be used to efficiently deliver metal complexes that cleave DNA inside target cells, thereby reducing overall exposure. Challenges include manufacturing complexity, stability during storage and circulation, and regulatory evaluation of multifunctional systems. Standardized production and characterization protocols are essential for successful clinical translation.

Long-term priorities (>10 years) are expected to focus on adaptive bioelectronic interfaces, programmable coordination systems, and autonomous therapeutic platforms. Emerging bioelectronic interfaces represent a potential avenue for metal-based drug development. Redox-active coordination complexes and conductive metallopolymers are being designed as core materials for biosensors and neural interface applications. For example, electrodes coupled with metal-binding ligands can detect analytes, including glucose and neurotransmitters, via metal-ion-mediated alterations in conductivity. Similarly, coordination polymers incorporating redox centers, such as tetrathiafulvalene-MOFs, can integrate electronic circuits into biological tissues, potentially paving the way for new generations of implantable or wearable devices [[372], [373], [374]]. These approaches utilize the rapid electron transfer kinetics and tunable electronic properties of metal centers to develop dynamic, responsive interfaces between tissues and electronic devices (Fig. 11).

Logic-gated multi-metal systems may enable programmable and autonomous therapeutic responses that are based on complex biological inputs. These coordination networks are developed only for defined combinations of stimuli, such as pH, enzymes, or light, or those that require multiple metal cofactors for activation. For instance, a prodrug can be designed with two metal centers arranged such that drug release occurs only when both metals attain specific oxidation states, with one metal behaving as a hypoxia sensor and the other as a ROS sensor [375,376]. Alternatively, multi-metal NPs can be designed for synergistic therapy, in which one metal provides photothermal heating and the other drives ROS production under precise temporal control. Embedding Boolean-like logic into metal coordination chemistry offers the possibility of highly conditional “if–then” therapeutic responses, thereby minimizing off-target effects [376,377]. Key challenges include system integration, predictable multi-metal coordination behavior, safety control, and validation in complex biological environments. Future development will require quantitative design frameworks that can link coordination chemistry with systems-level biological responses.

Collectively, these opportunities can be viewed as a translational roadmap spanning the short-, medium-, and long-term horizons. Near-term efforts should prioritize predictive design and biointerface characterization, medium-term development should focus on multifunctional therapeutic platforms, nucleic acid delivery, and immunomodulation, and long-term progress will depend on programmable coordination systems, adaptive bioelectronic interfaces, and autonomous precision medicine technologies.

11.1. Coordination therapeutics combined with immune checkpoint blockade

Combination strategies integrating metallodrugs, nanozymes, and coordination-based nanomedicines with immune checkpoint inhibitors have emerged as promising approaches in cancer therapy. Metal-induced ROS generation, immunogenic cell death, pyroptosis, PANoptosis, and cGAS–STING activation can increase tumor immunogenicity, improve antigen presentation, and enhance T cell infiltration. Consequently, these mechanisms may sensitize tumors to PD-1/PD-L1 and CTLA-4 blockade while overcoming resistance associated with immune checkpoint therapy. Future coordination-based combination therapies are expected to simultaneously modulate tumor metabolism, redox homeostasis, and antitumor immunity [378,379].

12. Conclusion

Coordination chemistry is more than just a traditional tool for medicinal chemists. This molecular language has been extensively applied in biological applications, particularly in metal-based precision therapy. Living systems rely on metal coordination for structural integrity, signaling, and catalysis, from oxygen carriers such as hemoglobin to Zn-finger gene regulators to diverse metalloenzymes. By decoding the molecular language of metal coordination, chemists and biologists can now design metal-based therapeutics that interact with biological systems (cells and tissues) using the same natural dialect. Importantly, the biological interface serves as a central framework linking coordination chemistry to biomolecular recognition, cellular communication, therapeutic activity, and clinical translation. The future of medicine will increasingly exploit these interdisciplinary concepts, yielding metal-based drugs and diagnostic procedures that mimic biological approaches for interactions and reactivity. When successfully integrated, coordination-driven therapeutics can offer distinctive mechanisms of action, including redox catalysis, multivalent recognition, and stimulus responsiveness, and explore entirely new paradigms, such as living metal-protein assemblies and biohybrid devices. Therefore, by applying the basic principles of metal interactions in biological systems, coordination chemistry has the potential to revolutionize early diagnosis and managements in multiple disciplines. Future advances are expected to increasingly integrate catalytic therapeutics, PANoptosis, pyroptosis, cGAS–STING activation, and immune checkpoint blockade into next-generation coordination-based precision medicine platforms.

Funding

Basic Research Project for Universities of Liaoning Provincial Education Department (LJ232410159083); Health Commission of Liaoning Province, “Xingliao Talents Program” Project (XLYC2412088)

CRediT authorship contribution statement

Kuanbing Chen: Investigation, Methodology, Writing – original draft. Xiaofeng Wang: Data curation, Investigation, Methodology, Writing – original draft. Yutao Wang: Data curation, Formal analysis, Investigation, Writing – original draft. Seid Mahdi Jafari: Supervision, Validation, Writing – review & editing. Feng Guo: Supervision, Validation, Writing – review & editing.

Declaration of competing interest

All authors declare that there is no conflict of interest.

Data availability

No data was used for the research described in the article.