Bioinorganic chemistry

Intuition [Beginner]

About one-third of all proteins require a metal ion to function. These metals are not contaminants — they are essential cofactors that the protein evolved to use. Iron carries oxygen in haemoglobin, zinc catalyses reactions in carbonic anhydrase, copper shuttles electrons in cytochrome c oxidase, and magnesium holds ATP in place for every kinase reaction in the cell.

The metal ion provides chemical capabilities that amino acid side chains alone cannot achieve. Iron switches between Fe${}^{2+}$ and Fe${}^{3+}$, enabling reversible oxygen binding. Zinc, with its filled d-shell (d${}^{10}$), acts as a powerful Lewis acid without the complications of d-orbital chemistry. Copper has accessible +1 and +2 oxidation states that make it ideal for single-electron transfers.

Haemoglobin is the paradigmatic metalloprotein. Each subunit contains a haem group — an iron(II) porphyrin complex. In the lungs, O${}_2$ binds to Fe${}^{2+}$ without oxidising it to Fe${}^{3+}$ (the binding is reversible). The binding of one O${}_2$ molecule increases the affinity of the remaining subunits for O${}_2$ — cooperative binding. This is the molecular basis for efficient oxygen loading in the lungs and unloading in the tissues.

Visual [Beginner]

The haem group is a flat porphyrin ring with Fe${}^{2+}$ at the centre. The iron sits slightly below the plane of the ring. When O${}_2$ binds, the iron moves into the plane.

Worked example [Beginner]

Haemoglobin oxygen binding: the Fe${}^{2+}$ porphyrin complex, cooperative binding, the Bohr effect.

The haem group consists of Fe${}^{2+}$ coordinated by four nitrogen atoms of the protoporphyrin IX ring (equatorial plane), one nitrogen from the proximal histidine (below the plane), and one site for O${}_2$ binding (above the plane). This is a six-coordinate octahedral complex.

In deoxyhaemoglobin, Fe${}^{2+}$ is high-spin d${}^6$ with a relatively large ionic radius. The iron sits about 0.4 Angstroms below the porphyrin plane because it is too large to fit in the ring's central cavity.

When O${}_2$ binds, it creates a strong-field ligand that shifts the iron to low-spin d${}^6$ (smaller ionic radius). The iron now fits in the porphyrin plane and slides into the ring. This small movement pulls the proximal histidine and the F-helix of the protein, triggering a conformational change that increases the O${}_2$ affinity of the other subunits.

Cooperative binding. The binding curve for haemoglobin is sigmoidal (S-shaped), not hyperbolic. The Hill equation models this: $Y=p{O}_2^n/(P_{50}^n+p{O}_2^n)$, where $n\approx 2.8$ (the Hill coefficient) indicates positive cooperativity. $n=1$ would mean no cooperativity; $n=4$ would mean all-or-nothing binding.

The Bohr effect. Lower pH (higher [H${}^{+}$]) and higher [CO${}_2$] decrease haemoglobin's oxygen affinity. In metabolically active tissues (low pH, high CO${}_2$), haemoglobin releases O${}_2$ more readily. In the lungs (high pH, low CO${}_2$), it binds O${}_2$ more tightly. This pH-dependent tuning ensures oxygen delivery exactly where it is needed.

Check your understanding [Beginner]

Formal definition [Intermediate+]

Metalloenzymes are enzymes that contain a metal ion cofactor essential for catalytic activity. The metal may be tightly bound (as in haem proteins) or more loosely associated (as in many Mg${}^{2+}$-dependent kinases).

Classification of metal cofactors:

1. Metalloenzymes with porphyrin cofactors: Haem proteins (haemoglobin, myoglobin, cytochromes, peroxidases, catalases). The porphyrin ring provides a rigid, planar tetradentate ligand.

2. Metalloenzymes with non-haem iron: Iron-sulphur proteins (ferredoxins, aconitase), mono-oxygenases (cytochrome P450).

3. Zinc enzymes: Carbonic anhydrase, carboxypeptidase, alcohol dehydrogenase, zinc-finger transcription factors.

4. Copper enzymes: Cytochrome c oxidase, superoxide dismutase, tyrosinase.

5. Molybdenum and tungsten enzymes: Nitrogenase, xanthine oxidase, sulphite oxidase.

6. Magnesium and manganese enzymes: Kinases (Mg-ATP), photosystem II (Mn${}_4$CaO${}_5$ cluster).

Cooperative binding models. The Monod-Wyman-Changeux (MWC) model: the haemoglobin tetramer exists in equilibrium between T (low-affinity) and R (high-affinity) states. Ligand binding shifts the equilibrium toward R. The Hill coefficient $n_H=\frac{d\log [Y/(1-Y)]}{d\log p{O}_2}$ quantifies cooperativity.

Counterexamples to common slips

• Haemoglobin and myoglobin are not the same. Haemoglobin is a tetramer with cooperative O${}_2$ binding; myoglobin is a monomer with hyperbolic (non-cooperative) binding.

• Iron does not change oxidation state when O${}_2$ binds to haemoglobin. The binding is formally a coordination of O${}_2$ to Fe${}^{2+}$, not an oxidation to Fe${}^{3+}$. The O${}_2$ is best described as bound as a singlet with partial electron transfer, maintaining the ferrous state.

• Not all metals in biology are transition metals. Na${}^{+}$, K${}^{+}$, Mg${}^{2+}$, Ca${}^{2+}$ are alkali/alkaline-earth metals with no d-electron chemistry, yet they are essential.

Key theorem with proof [Intermediate+]

Proposition (Hill equation for cooperative binding). For a protein with $n$ cooperative binding sites, the fractional saturation $Y$ as a function of ligand concentration $[L]$ is approximately $Y=[L{]}^{n_H}/(K_d^{n_H}+[L{]}^{n_H})$, where $n_H$ (the Hill coefficient) satisfies $1\le n_H\le n$.

Proof. The Hill model assumes that the protein exists in two states: fully unliganded (P) and fully liganded (PL${}_n$), with no partially liganded intermediates at significant concentration. The equilibrium is:

$$\mathrm{P}+n\mathrm{L}\rightleftharpoons {\mathrm{PL}}_n,K=\frac{[{\mathrm{PL}}_n]}{[\mathrm{P}][\mathrm{L}{]}^n}$$

The fractional saturation is:

$$Y=\frac{n[{\mathrm{PL}}_n]}{[\mathrm{P}]+[{\mathrm{PL}}_n]}=\frac{K[\mathrm{L}{]}^n}{1+K[\mathrm{L}{]}^n}$$

Defining $K_d^n=1/K$ (the apparent dissociation constant) and $n_H=n$ in the idealised case gives the Hill equation. In practice, partially liganded states exist, so $n_H<n$ — the Hill coefficient underestimates the true number of sites. For haemoglobin ($n=4$), $n_H\approx 2.8$. $■$

Bridge. The Hill equation builds toward 16.03.01 crystal field theory by linking the macroscopic binding cooperativity to the microscopic spin-state transition of Fe${}^{2+}$ in the haem pocket. The foundational reason cooperativity exists is the high-spin-to-low-spin transition that moves the iron into the porphyrin plane and pulls the F-helix; this is exactly the T-to-R structural switch that the MWC model treats thermodynamically at the Master tier. The bridge is between the coordination chemistry of a single metal centre 16.04.01 and the emergent allosteric behaviour of the tetrameric protein, and the pattern generalises to all cooperative metalloproteins where a local electronic change at the metal propagates through the protein scaffold to affect distant binding sites.

Exercises [Intermediate+]

Haemoglobin allostery: the MWC model and cooperative binding thermodynamics [Master]

The Monod-Wyman-Changeux (MWC) model ^{[Monod Wyman Changeux 1965]} treats haemoglobin as existing in a pre-existing equilibrium between two quaternary conformations: the T (tense) state with low O${}_2$ affinity and the R (relaxed) state with high O${}_2$ affinity. The allosteric constant $L_0=[T_0]/[R_0]$ describes the ratio of unliganded T to unliganded R in the absence of oxygen. For human haemoglobin, $L_0\approx 10^4$ — the T state is heavily favoured before any O${}_2$ binds.

The fractional saturation under the MWC model is:

$$Y=\frac{(1+c\alpha {)}^n}{L_0(1+\alpha {)}^n+(1+c\alpha {)}^n}$$

where $\alpha =p{O}_2/K_R$ is the normalised ligand concentration (ratio of oxygen partial pressure to the dissociation constant for the R state), $c=K_R/K_T$ is the ratio of dissociation constants (R to T), and $n=4$ is the number of binding sites. When $c\ll 1$ (T state binds O${}_2$ much more weakly than R) and $L_0\gg 1$ (T state dominates at zero saturation), the model produces sigmoidal binding with positive cooperativity without invoking any interaction between binding sites — cooperativity emerges from the concerted T-to-R switch.

The MWC model is concerted: all four subunits switch between T and R simultaneously. There are no hybrid conformations with some subunits in T and others in R. This distinguishes it from the Koshland-Nemethy-Filmer (KNF) sequential model (1966), in which each binding event induces a conformational change in the neighbouring subunit one at a time, allowing mixed conformations. Both models reproduce sigmoidal binding curves, but they differ in their predictions about the population of partially liganded intermediates. The KNF model predicts detectable concentrations of species such as Hb(O${}_2$)${}_2$ with mixed conformations; the MWC model predicts that such species are either T-like or R-like, not hybrids. This distinction has been probed experimentally by trapping partially liganded haemoglobin using Fe(II)/Co(II) hybrids and measuring their O${}_2$ affinities, with results favouring the concerted MWC picture for haemoglobin.

The Adair equation provides a model-independent description of stepwise binding. For a tetramer with four sites:

$$Y=\frac{K_1[L]+2K_1{K}_2[L{]}^2+3K_1{K}_2{K}_3[L{]}^3+4K_1{K}_2{K}_3{K}_4[L{]}^4}{4(1+K_1[L]+K_1{K}_2[L{]}^2+K_1{K}_2{K}_3[L{]}^3+K_1{K}_2{K}_3{K}_4[L{]}^4)}$$

where $K_i$ is the intrinsic association constant for the $i$-th binding step. Positive cooperativity corresponds to $K_1<K_2<K_3<K_4$ — each successive binding event is stronger. For haemoglobin, the measured Adair constants show that the first O${}_2$ binds weakly and the fourth binds most tightly, consistent with a T-to-R transition driven by progressive ligand binding. The Hill coefficient is derivable from the Adair constants: $n_H$ equals the slope of the Hill plot at half-saturation.

Perutz's stereochemical mechanism ^{[Perutz 1970]} links the structural changes to the thermodynamic model. In the T state, the Fe-N${}_{\text{His}}$ bond is elongated because high-spin Fe${}^{2+}$ sits below the porphyrin plane. The T-state salt bridges — between His146 $\beta$ and Asp94 $\beta$; between Lys40 $\alpha$ and His146 $\beta$; between Val1 $\alpha$ and Arg141 $\alpha$ — constrain the tetramer. When O${}_2$ binds and iron moves into the porphyrin plane, it pulls the F-helix and the FG corner, rupturing the T-state salt bridges. After two or more O${}_2$ molecules bind, the accumulated strain overcomes the T-state stabilisation and the tetramer snaps to the R conformation. The Perutz mechanism is the structural realisation of the MWC thermodynamic model: the salt bridges are the molecular embodiment of $L_0$.

2,3-Bisphosphoglycerate (2,3-BPG) binds in the central cavity of deoxyhaemoglobin, forming salt bridges with the beta chains and stabilising the T state. At high altitude, erythrocyte 2,3-BPG concentration increases over hours to days, shifting the O${}_2$ dissociation curve rightward and enhancing O${}_2$ release to tissues. Fetal haemoglobin (HbF, ${\alpha }_2{\gamma }_2$) has reduced 2,3-BPG affinity compared to adult haemoglobin (HbA, ${\alpha }_2{\beta }_2$) because the $\gamma$ chain substitutes Ser for His143 at the 2,3-BPG binding site. This difference gives HbF higher O${}_2$ affinity than HbA, allowing efficient placental oxygen transfer from mother to fetus despite the similar haem groups in both proteins.

Iron-sulphur clusters and biological electron transfer [Master]

Iron-sulphur (Fe-S) clusters are among the most ancient metal cofactors, likely present in the last universal common ancestor. They mediate electron transfer in ferredoxins, aconitase, and the respiratory complexes. Their ubiquity across all domains of life, and their synthesis by proteins that themselves contain Fe-S clusters, suggests that Fe-S chemistry predates the evolution of modern enzyme active sites.

The simplest clusters are:

• [2Fe-2S]: Two iron atoms bridged by two sulphide ions, each iron also coordinated by cysteine thiolates (or histidine in Rieske proteins). The cluster cycles between the [2Fe-2S]${}^{2+}$ (both Fe${}^{3+}$) and [2Fe-2S]${}^{+}$ (one Fe${}^{3+}$, one Fe${}^{2+}$) states, transferring one electron. The iron atoms are antiferromagnetically coupled through the bridging sulphides.

• [4Fe-4S]: A cube of alternating Fe and S atoms, coordinated by four cysteine thiolates. This cluster can access three oxidation states: [4Fe-4S]${}^{2+}$ (2Fe${}^{3+}$ + 2Fe${}^{2+}$), [4Fe-4S]${}^{+}$ (1Fe${}^{3+}$ + 3Fe${}^{2+}$), and [4Fe-4S]${}^{3+}$ (3Fe${}^{3+}$ + 1Fe${}^{2+}$). The electron is delocalised over all four iron atoms, not localised on a single site — this delocalisation lowers the reorganisation energy and enables fast electron transfer.

Fe-S clusters illustrate a general principle: biology uses metal clusters, not single metal ions, when multi-electron chemistry or fast electron transfer is needed. The cluster distributes the charge change over multiple atoms, reducing the structural reorganisation that would slow the transfer. This connects directly to Marcus theory ^{[Marcus 1993]}.

Marcus theory of electron transfer. The rate constant for non-adiabatic electron transfer between a donor and acceptor separated by distance $r$ is:

$$k_{ET}=\frac{2\pi }{\hslash }|V{|}^2\frac{1}{\sqrt{4\pi \lambda k_{B}T}}\exp \left(-\frac{(\Delta G^{\circ }+\lambda {)}^2}{4\lambda k_{B}T}\right)$$

where $V$ is the electronic coupling matrix element (which decays exponentially with distance: $V=V_0\exp (-\beta (r-r_0))$), $\lambda$ is the total reorganisation energy (inner-sphere, from bond-length changes at the metal, plus outer-sphere, from solvent/ protein dielectric rearrangement), and $\Delta G^{\circ }$ is the driving force. The Marcus prediction that the rate reaches a maximum at $-\Delta G^{\circ }=\lambda$ and then decreases for more exergonic reactions (the "inverted region") was confirmed experimentally by millimetre-scale electron-tunnelling measurements in modified Ru-haem proteins by Gray and Winkler.

For biological electron transfer, the outer-sphere reorganisation energy is suppressed by the protein environment: the hydrophobic interior and rigid scaffold reduce dielectric relaxation, keeping $\lambda$ low (typically 0.5–1.0 eV) compared to aqueous solution ($\lambda \approx 1$–2 eV). The distance decay constant $\beta$ for electron tunnelling through protein is approximately 1.4 per Angstrom, meaning the rate drops by a factor of about 10 for every 1.7 Angstrom increase in donor-acceptor distance. This distance dependence explains why biological electron-transfer chains position redox centres at carefully controlled intervals of 10–14 Angstroms — long enough to prevent short-circuit reactions, short enough to maintain transfer rates of $10^3$–$10^6$ s${}^{-1}$.

Rieske proteins contain a [2Fe-2S] cluster in which one iron is coordinated by two cysteine thiolates and the other by two histidine imidazoles. The histidine ligation raises the reduction potential of the cluster by approximately 300 mV relative to all-cysteine [2Fe-2S] ferredoxins (from about $-400$ mV to about $+300$ mV vs NHE), because the harder histidine donors stabilise the reduced Fe${}^{2+}$ state less effectively than the soft cysteine thiolates. This elevated potential positions the Rieske cluster as the electron-acceptor in the cytochrome bc${}_1$ complex (Complex III) of the mitochondrial electron transport chain, where it accepts electrons from ubiquinol and donates them to cytochrome c${}_1$. The Rieske domain is mobile, moving between a position close to the quinol-binding site (for electron uptake) and a position close to cytochrome c${}_1$ (for electron delivery), and this motion is coupled to proton translocation across the membrane.

Aconitase provides a remarkable example of an Fe-S cluster serving a non-redox role. In its [4Fe-4S]${}^{2+}$ form, aconitase catalyses the stereo-specific isomerisation of citrate to isocitrate via a cis-aconitate intermediate. One iron of the cluster is not coordinated by a cysteine but instead binds the substrate directly, acting as a Lewis acid. Loss of this iron converts the enzyme to the inactive [3Fe-4S]${}^{+}$ form. In mammalian cells, cytosolic aconitase doubles as iron-regulatory protein 1 (IRP1): when cellular iron is low, the [4Fe-4S] cluster disassembles, and the apo-protein binds iron-responsive elements (IREs) in mRNA to regulate translation of ferritin and transferrin receptor. The same protein thus functions as a metabolic enzyme when iron is abundant and as a transcriptional regulator when iron is scarce — a direct molecular link between metal availability and gene expression.

Photosynthetic water oxidation: the oxygen-evolving complex [Master]

The most demanding redox reaction in biology is the four-electron oxidation of water to O${}_2$, catalysed by the oxygen-evolving complex (OEC) of Photosystem II. The OEC is a Mn${}_4$CaO${}_5$ cluster that cycles through five oxidation states (S${}_0$ through S${}_4$, the Kok cycle ^{[Kok 1970]}), accumulating four oxidising equivalents before O${}_2$ is released in the S${}_3$ $\to$ S${}_0$ transition. Each photon-driven charge separation at the reaction centre P680 advances the cluster by one S-state.

The Kok cycle proceeds as follows. Starting from S${}_1$ (the dark-stable state), each photon removes one electron from the Mn cluster via the redox-active tyrosine Y${}_Z$ (Tyr161 of the D1 subunit), which mediates proton-coupled electron transfer between the OEC and P680${}^{+}$. The S${}_0\to$ S${}_1$ and S${}_1\to$ S${}_2$ transitions involve Mn-centred oxidations (Mn${}^{III}\to$ Mn${}^{IV}$). The S${}_2\to$ S${}_3$ transition is more contentious — it may involve oxidation of a Mn centre or oxidation of a bridging oxo ligand to an oxyl radical. The S${}_3\to$ S${}_0$ transition is the O-O bond-forming step: the transient S${}_4$ state (never directly observed) is thought to involve nucleophilic attack of a water-derived hydroxide on an electrophilic Mn${}^{IV}$=O or Mn${}^V$=O oxo, or radical coupling between two oxo/oxyl ligands on adjacent Mn centres. The O-O bond forms, O${}_2$ is released, and two water molecules refill the substrate sites, returning the cluster to S${}_0$.

The crystal structure of Photosystem II at 1.9 Angstrom resolution (Umena et al., Nature 2011) revealed the OEC geometry: a distorted cube of three Mn, one Ca, and four bridging O atoms, with a fourth Mn (the "dangling Mn") connected to the cube via two bridging oxo groups. The five bridging oxygen atoms and the substrate water molecules are positioned for proton-coupled electron transfer that avoids high-energy intermediates.

The key design principle is that the OEC stores the four holes needed for the four-electron oxidation of two water molecules. Without this multi-nuclear metal cluster, partial oxidation products (superoxide O${}_2^{\bullet -}$, hydrogen peroxide H${}_2$O${}_2$, or hydroxyl radical HO${}^{\bullet }$) would be released, damaging the photosynthetic apparatus. The cluster achieves complete four-electron water oxidation at ambient temperature with quantum efficiency near 1.0 — each absorbed photon produces one oxidising equivalent with near-unit yield.

The synthetic challenge of replicating this chemistry with artificial catalysts remains unsolved. Ruthenium-based and cobalt-based water-oxidation catalysts achieve multi-electron water splitting but require substantial overpotentials (200–500 mV) and lack the self-repair mechanisms of the biological system. The OEC undergoes damage and replacement every 30 minutes of illumination in vivo; the D1 protein that houses the cluster is the most rapidly turned over protein in the thylakoid membrane.

Zinc enzymes and Lewis acid catalysis at atomic resolution [Master]

Zinc is the second most abundant transition metal in biology (after iron) and is the only metal used exclusively as a Lewis acid in enzymatic catalysis. Its d${}^{10}$ electronic configuration renders it redox-inert: Zn${}^{2+}$ has no accessible higher or lower oxidation states, so it participates only in substrate activation through electrostatic polarisation, never in electron transfer. This property, combined with its high charge density (ionic radius 74 pm, charge +2), makes Zn${}^{2+}$ an exceptionally strong Lewis acid with a labile coordination sphere — ligands exchange readily, which is essential for catalytic turnover.

Carbonic anhydrase II is the fastest enzyme known, with $k_{cat}\approx 4\times 10^5$ s${}^{-1}$ at 25 degrees C, approaching the diffusion-controlled limit. The active site contains Zn${}^{2+}$ coordinated by three His residues (His94, His96, His119 in human CA II) and one water molecule in a distorted tetrahedral geometry. The mechanism has two half-reactions that alternate:

In the hydration direction, Zn${}^{2+}$ lowers the pK${}_a$ of the coordinated water from 15.7 to approximately 6.9, so at physiological pH the active species is the zinc-bound hydroxide (Zn-OH${}^{-}$). This nucleophilic hydroxide attacks CO${}_2$ bound in a hydrophobic pocket adjacent to the zinc, forming bicarbonate (HCO${}_3^{-}$). Bicarbonate dissociates and is replaced by water. In the second half-reaction, a proton transfer from the newly bound Zn-H${}_2$O to buffer in the bulk solvent restores the Zn-OH${}^{-}$ nucleophile. This proton transfer is rate-limiting and is facilitated by His64, which shuttles protons between the zinc-bound water and the solvent. Mutation of His64 to Ala reduces $k_{cat}$ by 10–20-fold, demonstrating the proton shuttle's role.

Carboxypeptidase A uses Zn${}^{2+}$ for peptide bond hydrolysis at the C-terminus of protein substrates. The zinc is coordinated by two His residues, one Glu (bidentate), and one water molecule. The mechanism involves Zn${}^{2+}$ activating the carbonyl oxygen of the scissile peptide bond (Lewis acid activation), a Glu270 general base that deprotonates the attacking water, and an Arg127 that stabilises the developing negative charge on the carbonyl oxygen. The transition state is a tetrahedral oxyanion intermediate stabilised by Zn${}^{2+}$ and Arg127 — a classic example of dual electrophilic/nucleophilic catalysis by a metalloenzyme.

Alcohol dehydrogenase uses a catalytic Zn${}^{2+}$ and a structural Zn${}^{2+}$. The catalytic zinc is coordinated by two Cys thiolates, one His, and the substrate alcohol. Zn${}^{2+}$ polarises the alcohol O-H bond and positions the substrate for hydride transfer to NAD${}^{+}$. The structural zinc, coordinated by four Cys residues, stabilises the protein fold but does not participate directly in catalysis. The distinction between catalytic and structural metal sites is a recurring theme in metalloprotein chemistry: not every metal in a protein is at the active site, and misidentification of structural metals as catalytic (or vice versa) has led to incorrect mechanistic proposals.

The Irving-Williams series (Ba${}^{2+}$ < Sr${}^{2+}$ < Ca${}^{2+}$ < Mg${}^{2+}$ < Mn${}^{2+}$ < Fe${}^{2+}$ < Co${}^{2+}$ < Ni${}^{2+}$ < Cu${}^{2+}$ > Zn${}^{2+}$) orders divalent metal ions by their stability constants for complex formation with a given ligand. Cu${}^{2+}$ forms the most stable complexes; Zn${}^{2+}$ is slightly less stable. This ordering governs metal selectivity in biology: proteins that need a strongly-bound, substitutionally-inert metal (like Cu${}^{2+}$) must prevent the more weakly-bound metals (like Mg${}^{2+}$) from occupying the site, while proteins that need a labile metal for catalysis (like Zn${}^{2+}$) benefit from the moderate binding strength that allows rapid ligand exchange.

Metal ion homeostasis and metallochaperones [Master]

Cells must maintain precise concentrations of metal ions despite wide variations in environmental availability. The total cellular concentration of most transition metals is in the micromolar range, but the concentration of free (unbound) metal ions is orders of magnitude lower. For copper, the free Cu${}^{+}$ concentration in a yeast cell is estimated at less than $10^{-18}$ M — fewer than one free copper ion per cell. This near-zero free pool is maintained because free copper ions catalyse Fenton-type reactions that generate hydroxyl radicals from hydrogen peroxide, damaging DNA, proteins, and lipids.

Copper trafficking illustrates the principle of metallochaperone-mediated delivery. In Saccharomyces cerevisiae, the copper chaperone Atx1 binds Cu${}^{+}$ using a CXXC motif (two cysteines that coordinate the soft Cu${}^{+}$ ion) and delivers it to the Cu-transporting ATPase Ccc2 in the trans-Golgi network, which pumps copper into the secretory pathway for incorporation into Fet3 (a multi-copper oxidase required for iron uptake) and into extracellular superoxide dismutase (SOD1). The copper chaperone for SOD1 (CCS) directly inserts Cu${}^{+}$ into the active site of Cu/Zn-SOD1 and facilitates the oxidation of the catalytic disulphide bond. Mutations in the human homologue of CCS or in the copper-transporting ATPases (ATP7A, ATP7B) cause Menkes disease and Wilson disease, respectively — both severe disorders of copper metabolism.

Iron homeostasis is controlled at multiple levels. Ferritin stores iron as a ferrihydrite mineral core inside a 24-subunit protein shell capable of holding up to 4500 Fe atoms. Transferrin transports Fe${}^{3+}$ in the bloodstream, binding two iron atoms per molecule with high affinity ($K_d\approx 10^{-22}$ M at pH 7.4). Cellular iron uptake occurs via transferrin receptor-mediated endocytosis, followed by Fe${}^{3+}$ reduction to Fe${}^{2+}$ and transport across the endosomal membrane by DMT1. Iron export from cells is mediated by ferroportin, with hephaestin or ceruloplasmin serving as ferroxidases that re-oxidise Fe${}^{2+}$ to Fe${}^{3+}$ for transferrin loading.

The IRP/IRE regulatory system provides post-transcriptional control. Iron regulatory proteins (IRP1, IRP2) bind iron-responsive elements (IREs) — stem-loop structures in the untranslated regions of mRNAs encoding ferritin, transferrin receptor, DMT1, and ferroportin. When iron is scarce, IRPs bind IREs: binding to the 5-prime UTR of ferritin mRNA represses translation (less storage), while binding to the 3-prime UTR of transferrin receptor mRNA stabilises the transcript (more uptake). When iron is abundant, IRP1 assembles a [4Fe-4S] cluster and loses IRE-binding affinity (becoming cytosolic aconitase instead), while IRP2 is degraded by the proteasome. The result is a coordinated up-regulation of storage and down-regulation of uptake when iron is plentiful.

Zinc homeostasis is mediated by metallothioneins (small, cysteine-rich proteins that bind up to 7 Zn${}^{2+}$ ions in thiolate clusters) and by the ZIP (Zrt/Irt-like Protein, influx) and ZnT (Zn Transporter, efflux) families of membrane transporters. The human genome encodes 14 ZIP transporters and 10 ZnT transporters, each localised to specific subcellular compartments. Unlike copper and iron, zinc has no redox chemistry, so zinc homeostasis is primarily a matter of buffering and compartmentalisation rather than managing oxidative damage.

The selectivity of metal delivery — ensuring that Zn${}^{2+}$ goes to zinc enzymes and Cu${}^{+}$ goes to copper enzymes, rather than the reverse — depends on a combination of thermodynamic preferences (the Irving-Williams series) and kinetic control by metallochaperones. The cell maintains the free concentration of each metal at a level matched to the affinity of the correct recipient protein. Mis-metallation is prevented not by making every protein exquisitely selective, but by controlling the metal availability so that the thermodynamically preferred metal reaches the protein first.

Disruption of metal homeostasis underlies several human diseases. Haemochromatosis (excess iron absorption due to mutations in the HFE gene or hepcidin pathway) leads to iron deposition in the liver, heart, and pancreas, causing cirrhosis, cardiomyopathy, and diabetes. Wilson disease (mutations in ATP7B, the copper-exporting ATPase) causes copper accumulation in the liver and brain, producing hepatic failure and neurodegeneration with Kayser-Fleischer rings (copper deposits in the cornea). These diseases illustrate the toxic consequences of losing control over metal distribution, even for metals that are essential in regulated amounts.

Vitamin B12 and organometallic biochemistry [Master]

Vitamin B12 (cobalamin) contains a Co ion coordinated in a corrin ring — similar to porphyrin but with one fewer double bond in the conjugated system and a more flexible macrocycle. The key structural feature is the direct Co-C bond to the upper axial ligand, making B12 the only known organometallic cofactor in biology. Two coenzyme forms exist: adenosylcobalamin (AdoB12, with a 5-prime-deoxyadenosyl group) and methylcobalamin (MeB12, with a methyl group).

Adenosylcobalamin-dependent rearrangements. In AdoB12-dependent enzymes, homolysis of the Co-C bond generates a 5-prime-deoxyadenosyl radical (Ado${}^{\bullet }$) and Co(II). The adenosyl radical abstracts a hydrogen atom from the substrate, generating a substrate radical that undergoes a 1,2-rearrangement (migration of a functional group from one carbon to the adjacent carbon). The rearranged product radical then abstracts a hydrogen back from 5-prime-deoxyadenosine, regenerating the adenosyl radical and forming the product.

The best-characterised example is methylmalonyl-CoA mutase, which converts methylmalonyl-CoA to succinyl-CoA via a 1,2-migration of the thioester carbonyl-CoA group. The radical intermediate is the hallmark of AdoB12 catalysis: the Co-C bond homolysis provides a low-energy pathway for radical generation that avoids the high bond-dissociation energy of a C-H bond. The enzyme accelerates Co-C homolysis by a factor of approximately $10^{12}$ relative to the free cofactor in solution, through a combination of ground-state strain (distortion of the corrin ring and the Co-C bond angle) and stabilisation of the radical pair through electrostatic interactions with the protein.

The radical rearrangement mechanism was established by isotopic labelling and EPR spectroscopy, which detected the Co(II) radical pair intermediate. The substrate radical is generated within 100 ns of Co-C homolysis, and the overall reaction is completed in milliseconds. The enzyme controls the radical tightly: the substrate radical does not escape the active site, preventing damage to the protein or to other cellular components.

Methylcobalamin-dependent methyl transfer. In MeB12-dependent enzymes, the methyl group is transferred from methylcobalamin to a substrate via an S${}_N$2 mechanism. The cobalt cycles between Co(I) and Co(III): Co(I) is an extremely strong nucleophile (a "supernucleophile" with $k\approx 10^9$ M${}^{-1}$s${}^{-1}$ for methyl transfer from methyl halides) that attacks a methyl donor (typically N${}^5$-methyltetrahydrofolate or methyltetrahydrofolate), forming MeB12 with Co(III)-CH${}_3$. The methyl group is then transferred to the substrate (homocysteine to form methionine, in the case of methionine synthase) by a second S${}_N$2 displacement, regenerating Co(I).

The Co(I)/Co(III) redox cycle avoids radical intermediates entirely, in contrast to the AdoB12 radical mechanism. This dichotomy — radical mechanism for rearrangements, polar S${}_N$2 mechanism for methyl transfer — is a striking example of how the same metal centre (cobalt in a corrin ring) supports two fundamentally different catalytic strategies depending on the axial ligand and the protein environment.

B12 deficiency in humans causes megaloblastic anaemia (impaired DNA synthesis due to folate trapping as methyl-THF) and neurological damage (impaired methylmalonyl-CoA mutase leads to accumulation of methylmalonic acid and abnormal fatty acid synthesis in myelin). The biochemical basis is that methionine synthase requires MeB12 to regenerate methionine and THF from homocysteine and methyl-THF; without B12, folate becomes trapped as methyl-THF, depleting the THF pool needed for nucleotide biosynthesis. Pernicious anaemia (autoimmune destruction of gastric parietal cells that produce intrinsic factor, the B12-binding protein required for intestinal absorption) is the most common cause of B12 deficiency.

Cytochrome P450 and biological hydroxylation [Master]

Cytochrome P450 enzymes catalyse the insertion of one oxygen atom from O${}_2$ into an unactivated C-H bond of a substrate, reducing the second oxygen atom to water. The overall stoichiometry is:

$$\text{RH}+{\text{O}}_2+\text{NADPH}+{\text{H}}^{+}⟶\text{ROH}+{\text{H}}_2\text{O}+{\text{NADP}}^{+}$$

The active site contains a haem iron (iron protoporphyrin IX) with a cysteine thiolate as the proximal axial ligand — the defining feature of P450 enzymes. The cysteinate ligand is a strong electron donor that stabilises the high-valent iron-oxo intermediate responsible for C-H bond activation. This contrasts with haemoglobin (proximal His) and peroxidases (proximal His), which use nitrogen donors.

The catalytic cycle proceeds through six steps. (1) Substrate binds to the resting Fe${}^{3+}$ state, displacing a water molecule and shifting the spin state from low-spin to high-spin Fe${}^{3+}$. (2) One-electron reduction by a reductase (usually a flavoprotein) produces Fe${}^{2+}$. (3) O${}_2$ binds to Fe${}^{2+}$, forming an Fe${}^{2+}$-O${}_2$ complex analogous to oxyhaemoglobin. (4) A second electron is delivered, generating an Fe${}^{3+}$-peroxide (Fe${}^{3+}$-OO${}^{2-}$) species. (5) Protonation of the distal oxygen produces a hydroperoxo intermediate (Fe${}^{3+}$-OOH), and a second protonation triggers heterolytic O-O cleavage, releasing water and generating the reactive iron(IV)-oxo porphyrin radical cation, Compound I (Fe${}^{4+}$=O with a porphyrin pi-cation radical). (6) Compound I abstracts a hydrogen atom from the substrate C-H bond, generating a substrate radical and Fe${}^{4+}$-OH. The substrate radical recombines with the hydroxyl on the iron ("oxygen rebound"), forming the hydroxylated product and regenerating Fe${}^{3+}$.

The oxygen-rebound mechanism, proposed by Groves in 1978, explains the stereochemistry of hydroxylation: the substrate radical does not have time to racemise before rebound occurs (radical lifetime approximately 10${}^{-12}$ s), so retention of configuration is observed for most P450 reactions. However, some substrates with particularly stable radical intermediates (benzylic, allylic positions) show partial racemisation, providing direct evidence for the radical intermediate.

The power of P450 chemistry is that it activates inert C-H bonds (bond dissociation energy 370–440 kJ/mol) under mild conditions using the energy of NADPH oxidation. The catalytic competency of Compound I is the iron-oxo equivalent of the high-valent manganese intermediates in the OEC: both use multi-valent metal centres to achieve thermodynamically demanding oxidations. The difference is that P450 uses two reducing equivalents from NADPH plus the two oxidising equivalents of O${}_2$ to generate one hydroxylated product, while the OEC accumulates four oxidising equivalents from light-driven charge separation to split water.

Connections [Master]

• Crystal field theory 16.03.01 (pending). The high-spin to low-spin transition of Fe${}^{2+}$ upon O${}_2$ binding to haemoglobin is a direct application of crystal field theory to a biological system. The CFSE difference between the two spin states, and the resulting change in ionic radius that moves the iron into the porphyrin plane, is the atomic-level event that triggers the T-to-R conformational switch underlying cooperative binding.

• Coordination chemistry 16.04.01. Every metalloenzyme active site is a coordination complex with amino acid side chains (His, Cys, Asp, Glu) and sometimes exogenous ligands (water, O${}_2$, CO) as ligands. The geometry (octahedral, tetrahedral, square-planar), ligand field strength, and redox properties of the metal centre are all determined by the principles of coordination chemistry developed for inorganic model complexes.

• Enzyme mechanism 15.14.01 pending (pending). Metalloenzyme catalytic strategies — Lewis acid activation of substrates (Zn enzymes), redox catalysis (Fe-S clusters, cytochrome P450), and substrate positioning through coordination geometry — are extensions of the general catalytic strategies (acid-base, covalent, proximity/orientation) treated in the enzyme mechanism unit, specialised for the electronic properties of metal ions.

• Organometallic chemistry 16.05.01. The back-bonding concept that explains CO toxicity and haem-O${}_2$ binding is shared with organometallic pi-back-bonding in metal-carbonyl complexes. Vitamin B12's direct Co-C bond is an organometallic bond, and the catalytic radical chemistry of adenosylcobalamin parallels the radical mechanisms proposed for certain organometallic reactions.

• Acid-base chemistry 14.10.01. The Lewis acid catalysis by Zn${}^{2+}$ in carbonic anhydrase — lowering the pK${}_a$ of coordinated water from 15.7 to 7 — is a biological application of the Lewis acid-base framework. The Bohr effect is a Bronsted acid-base phenomenon (protonation of His side chains modulates O${}_2$ affinity), and proton-coupled electron transfer in Photosystem II and cytochrome c oxidase couples proton transfer to redox chemistry.

Historical & philosophical context [Master]

The recognition that metals are essential to biology dates to the 17th century (Lemery and Boyle demonstrated iron in blood by chemical analysis), but the structural basis of metal function emerged only with X-ray crystallography. Max Perutz solved the haemoglobin crystal structure over 22 years of work ^{[Perutz 1970]}, revealing the T-to-R conformational transition and providing the first molecular mechanism for cooperative ligand binding by an allosteric protein. Perutz shared the 1962 Nobel Prize in Chemistry with John Kendrew (myoglobin structure).

Archibald Hill introduced the Hill equation in 1910 (J. Physiol. 40, iv–vii) as an empirical description of the sigmoidal oxygen-binding curve. Hill's original paper used the equation as a phenomenological model; the interpretation in terms of cooperative binding sites came later. The Monod-Wyman-Changeux allosteric model ^{[Monod Wyman Changeux 1965]}, published in Journal of Molecular Biology in 1965, provided the first thermodynamically rigorous framework for understanding cooperative binding. The competing Koshland-Nemethy-Filmer sequential model appeared in 1966 (Biochemistry 5, 365). The MWC and KNF models represent two fundamentally different physical pictures — concerted vs sequential conformational change — and the debate between them stimulated decades of experimental work on partially liganded haemoglobin intermediates.

Rudolph Marcus developed the theory of electron-transfer rates between 1956 and 1965, for which he received the 1992 Nobel Prize in Chemistry ^{[Marcus 1993]}. Application of Marcus theory to biological electron transfer by Dutton, Gray, and McLendon in the 1980s and 1990s established the distance dependence and reorganisation energy parameters that govern electron tunnelling through protein. The discovery that electron tunnelling in proteins follows a single exponential distance dependence with $\beta \approx 1.4$ per Angstrom, regardless of the specific protein fold, was a key finding: it demonstrated that the protein medium supports coherent electron tunnelling rather than requiring specific pathways.

Bessel Kok and coworkers established the S-state cycle for photosynthetic water oxidation in 1970 ^{[Kok 1970]}. The Kok model explained the periodic flash-dependent oscillation of O${}_2$ yield with a maximum at the third flash (dark-adapted samples start in S${}_1$) and the observation that each flash advances the OEC by one state regardless of flash intensity. The structural determination of the Mn${}_4$CaO${}_5$ cluster by Umena, Kawakami, Shen, and Kamiya (Nature 2011) at 1.9 Angstrom resolution confirmed the cubane-plus-dangling-Mn architecture proposed from spectroscopic data and opened the way for computational studies of the O-O bond-forming transition state.

Bibliography [Master]

• Lippard, S. J. & Berg, J. M. Principles of Bioinorganic Chemistry. Mill Valley: University Science Books, 1994.

• Frausto da Silva, J. J. R. & Williams, R. J. P. The Biological Chemistry of the Elements. Oxford: Oxford UP, 2001.

• Bertini, I., Gray, H. B., Stiefel, E. I. & Valentine, J. S. Biological Inorganic Chemistry. Sausalito: University Science Books, 2007.

• Perutz, M. F. "Stereochemistry of cooperative effects in haemoglobin." Nature 228 (1970), 726–739.

• Monod, J., Wyman, J. & Changeux, J.-P. "On the nature of allosteric transitions: a plausible model." J. Mol. Biol. 12 (1965), 88–118.

• Marcus, R. A. "Electron transfer reactions in chemistry: theory and experiment." Rev. Mod. Phys. 65 (1993), 599–610.

• Kok, B., Forbush, B. & McGloin, M. "Cooperation of charges in photosynthetic O${}_2$ evolution — a model for the S-state transitions." Photochem. Photobiol. 11 (1970), 457–475.

• Umena, Y., Kawakami, K., Shen, J.-R. & Kamiya, N. "Crystal structure of oxygen-evolving photosystem II at 1.9 Angstrom resolution." Nature 473 (2011), 55–60.

• Gray, H. B. & Winkler, J. R. "Electron tunneling through proteins." Q. Rev. Biophys. 36 (2003), 341–372.

• Koshland, D. E., Nemethy, G. & Filmer, D. "Comparison of experimental binding data and theoretical models in proteins containing subunits." Biochemistry 5 (1966), 365–385.

• Berg, J. M., Tymoczko, J. L. & Stryer, L. Biochemistry, 7th ed. New York: W. H. Freeman, 2012. Ch. 7 (Haemoglobin) and Ch. 27 (Electron transport and oxidative phosphorylation).